Volume 198 Number 1 BIOLOGICAL BULLETIN FEBRUARY 2000 ilic Marine Biological Laboratory Hf MARINE BIOLOGICAL LABORATORY COURSE OFFERINGS JULY - OCTOBER, 2000 Digital Imaging Platform Meet the challenges of the digital age with Axioplan 2 imaging. Not just a new microscope but a flexible system de- signed for the future - the digital imaging platform. Combine AxioVision software, with its built-in microscope control, and the new AxioCam ultra high resolution digital camera, and you have pure synergy - microscopy and more. Make yours the future-proof choice for high resolution image capture, image processing, archiving - even tele-microscopy. The digital age is with us. Are you? Carl Zeiss, Inc. Microscopy & Imaging Systems One Zeiss Drive Thornwood, NY 10594 800.223.2243, ext. 7858 micro@zeiss.com www.zeiss.com/micro i ZEISS Methods in Computational Neuroscience (July 30 - August 26, 2000) Directors: William Bialek and Rob de Ruyter van Steveninck, NEC Research Institute Workshop on Molecular Evolution (]uly 30 - August 11, 2000) Director: Michael Cummings, Marine Biological Laboratory Molecular Biology of Aging sponsored by the Ellison Medical Foundation (August 1 - August 18, 2000) Directors: Leonard Cuarente, Massachusetts Institute of Technology and Douglas Wallace, Emory University Molecular Mycology: Current Approaches to Fungal Pathogenesis (August 7 - August 25, 2000) Directors: Aaron Mitchell, Columbia University; P.T. Magee, University of Minnesota; and John Edwards, Harbor-UCLA Medical Center Fundamental Issues in Vision Research sponsored by the National Eye Institute (August I 3 - August 25, 2000) Director: David Papermaster, University of Connecticut Health Center Neural Development at Genetics of Zebrafish (August 1 3 - August 26, 2000) Directors: John Dowling, Harvard University; and Nancy Hopkins, Massachusetts Institute of Technology Neurobiology 6C Development of the Leech (August 13 - September I, 2000) Directors: Ronald Calabrese, Emory University; Christie Sahley, Purdue University; and Martin Shankland, University of Texas, Austin Medical Informatics sponsored by the National Library of Medicine (October 1 - October 7, 2000) Director: James Clmino, Columbia University Optical Microscopy and Imaging in the Biomedical Sciences (October 1 1 - October 19, 2000) Director: Colin Izzard, State University of New York, Albany For more information, contact: Carol Hamel, Admissions Coordinator 508-289-7401 /email: admissions@mbl.edu OR Visit our website: courses.mbl.edu The MBL is jn EEO/Affirnutive Action Institution THE BIOLOGICAL BULLETIN FEBRUARY 2000 i Editor Associate Editors Section Editor Online Editors Editorial Board Editorial Office MICHAEL J. GREENBERG Louis E. BURNETT R. ANDREW CAMERON CHARLES D. DERBY MICHAEL LABARBERA SHINYA INOUE, Imaging and Microscopv JAMES A. BLAKE, Keys to Marine Invertebrates of the Woods Hole Region WILLIAM D. COHEN. Marine Models Electronic Record and Compendia PETER B. ARMSTRONG ERNEST S. CHANG THOMAS H. DIETZ RICHARD B. EMLET DAVID EPEL GREGORY HINKLE MAKOTO KOBAYASHI DONAL T. MANAHAN MARGARET MCFALL-NGAI MARK W. MILLER TATSUO MOTOKAWA YOSHITAKA NAGAHAMA SHERRY D. PAINTER J. HERBERT WAITE RICHARD K. ZIMMER PAMELA CLAPP HINKLE VICTORIA R. GIBSON CAROL SCHACHINGER PATRICIA BURNS Ft. 3 2 5 2000 The Whitney Laboratory, University of Florida Grice Marine Laboratory, College of Charleston California Institute of Technology Georgia State University University of Chicago Marine Biological Laboratory ENSR Marine & Coastal Center, Woods Hole Hunter College, City University of New York University of California, Davis Bodega Marine Lab., University of California, Davis Louisiana State University Oregon Institute of Marine Biology, Univ. of Oregon Hopkins Marine Station, Stanford University Cereon Genomics, Cambridge. Massachusetts Hiroshima University of Economics, Japan University of Southern California Kewalo Marine Laboratory, University of Hawaii Institute of Neurobiology, University of Puerto Rico Tokyo Institute of Technology, Japan National Institute for Basic Biology. Japan Marine Biomed. Inst.. Univ. of Texas Medical Branch University of California, Santa Barbara University of California, Los Angeles Managing Editor Staff Editor Editorial Associate Subscription & Advertising Secretary Published by MARINE BIOLOGICAL LABORATORY WOODS HOLE, MASSACHUSETTS Cover Several species of sea cucumbers in the order As- pidochirotida respond to irritation by a predator in an unusual but effective way. The aboral (or anal) end is turned toward the irritant, the sea cu- cumber contracts, and a number of tubules detach from the left respiratory tree and are expelled, first through a tear in the cloaca, and then through the anus. These so-called Cuvierian tubules become sticky in the seawater and entangle the predator, immobili/ing it. The sea cucumber then crawls away to safety and, at its leisure, repairs the tear in the cloaca and regenerates the lost tubules. In this issue (pp. 34-49), the complex processes of wound healing and Cuvierian tubule regeneration and growth are described by Didier VandenSpiegel and his colleagues Michel Jangoux and Patrick Flammang. Their observations have been carried out on Holothuriaforskcili from the Bay of Morlaix in Brittany. The cover photo illustrates a set of Cuvierian tu- bules from H. forskali that are regenerating after a strong stimulus has produced a massive expulsion. Notice that some tubules are completely regrown, while others are at various stages of regeneration (see legend to Fig. I on p. 37). The measured re- sponse of this spcciali/ed response system and the phased process of regeneration lead the authors to conclude that these sea cucumbers have the advan- tage of "an almost inexhaustible line of defense maintained at limited energy cost." CONTENTS VOLUME 198, No. 1: FEBRUARY 2000 IMAGING AND MICROSCOPY Zochowski, Michal, Matt Wachowiak, Chun X. Falk, Lawrence B. Cohen, Ying-wan Lam, Srdjan Antic, and Dejan Zecevic Concepts in Imaging and Microscopy: Imaging mem- brane potential with voltage-sensitive dyes 1 RESEARCH NOTES Lambert, Charles C. Germ-cell warfare in ascidians: sperm from one species can interfere with the fertilization of a second species 22 Tsutsui, Hidekazu, and Yoshitaka Oka Light-sensitive voltage responses in the neurons of the cerebral ganglion of dona savignyi (Chordata: Ascidiacea) 26 Bavestrello, Giorgio, Attilio Arillo, Barbara Calcinai, Riccardo Cattaneo-Vietti, Carlo Cerrano, Elda Gaino, Antonella Penna, and Michele Sara Parasitic diatoms inside Antarctic sponges 29 PHYSIOLOGY VandenSpiegel, Didier, Michel Jangoux, and Patrick Flammang Maintaining the line of defense: regeneration of Cu- vierian tubules in the sea cucumber Holothuriaforskali (Echinodermata. Holothuroidea) Sikes, C. S., A. P. Wheeler, A. Wierzbicki, A. S. Mount, and R. M. I hll.nn.ni Nucleation and growth of calcite on native versus pyrolyzed oyster shell folia 34 50 Ereskovsky, Alexander V. Reproduction cycles and strategies of the cold-water sponges Halisarca dujardini (Demospongiae, Halisar- cida), Myxilla incrustans and lophon piceus (Demo- spongiae, Poecilosclerida) from the White Sea 77 ECOLOGY AND EVOLUTION Leddy, Holly A., and Amy S. Johnson Walking versus breathing: mechanical differentiation of sea urchin podia corresponds to functional spe- cialization 88 Voight, Janet R., and Anthony Grehan Egg brooding by deep-sea octopuses in the North Pacific Ocean 94 Pivkin, Michael V. Filamentous fungi associated with holothurians from the Sea of Japan, off the Primorye Coast of Russia. . 101 CELL BIOLOGY Fitt, William K. Cellular growth of host and symbiont in a cnidarian- zooxanthellar symbiosis 110 SYSTEMATICS Scheltema, Amelie H., and Christoffer Schander Discrimination and phylogeny of solenogaster spe- cies through the morphology of hard parts (Mol- lusca, Aplacophora, Neomeniomorpha) 121 DEVELOPMENT AND REPRODUCTION Hadfield. Michael G., Ella A. Meleshkevitch, and Dmitri Y. Boudko The apical sensory organ of a gastropod veliger is a receptor for settlement cues 67 STATISTICS Dewdney, A. K. A dynamical model of communities and a new spe- cies-abundance distribution 152 THE BIOLOGICAL BULLETIN THE BIOLOGICAL BULLETIN is published six limes a year by the Marine Biological Laboratory. 7 MBl. Street. Woods Hole. Massachusetts 02543. Subscription- and similar matter should be addressed to Subscription Manager. THE BIOLOGICAL BULLETIN, Manrk- Hi, 'logical 1 ahor.itory, 7 MBL Street. Woods Hole. Massachusetts 02543 Subscription per year (six issues, two volumes*: f.225 tor libraries; S95 for individuals. Subscription per volume ithree issues): $1 12.50 for libraries: S47.50 tor individuals. Back and single issues (subject to availability): S40 for libraries; S20 for individuals. Communications relative to manuscripts should be sent to Michael J. Greenberg. hditor-in-Chief, or Pamela Clapp Hmkle. Managing Editor, at the Marine Biological Laboratory . 7 MBL Street. Woods Hole. Massachusetts 02543. Telephone: (5081 289-7428. FAX: 508-289-7922. E-mail: pclapp@mbl.edu. http://www.mbl.edu/BiologicalBulletin/ The home page for the electronic companion to Tm BIOLOGICAL BULLETIN the Marine Models Electronic Record and other BIOKXUC vi BULLFTIN publications is available on the World Wide Web at the address shown above. THE BIOLOGICAL BULLETIN is indexed in bibliographic services including Index Median, and MEDLINE. Chemical Abstracts, Current Cuntenis. Elsevier BlOBASE/Curreni Amin-ness in Biological Sciences, and Ceo Abstracts. Printed on acid free paper, effective with Volume 180. Issue I. 1991. POSTMASTER: Send address changes to THE BIOLOGICAL B; 1 1 HIV Marine Biological Laboratory. 7 MBL Street. Woods Hole, MA 02543. Copyright 2000. by the Marine Biological Laboratory Periodicals postage paid at Woods Hole. MA. and additional mailing offices. ISSN 0006-3185 INSTRUCTIONS TO AUTHORS The Biological Bulletin accepts outstanding original research reports of general interest to biologists throughout the world. Papers are usually of intermediate length (10-40 manuscript pages). A limited number of solicited review papers may he accepted after formal review. A paper will usually appear within four months after its acceptance. Very short, especially topical papers (less than 9 manuscript pages including tables, figures, and bibliography) will he published in a separate section entitled "Research Notes." A Research Note in The Biological Bulletin follows the format of similar notes in Nature. It should open with a summary paragraph of 150 to 200 words comprising the introduction and the conclusions The rest of the text should continue on without subheadings, and there should be no more than 30 references. References should he referred in in the text by number, and listed in the Literature Cited section in the order that they appear in the text I'nlikc releu-in.es in ,\titurc. references in the Research Notes section should conform in punctuation and arrangement to the style of recent issues of Tin- Biological Bulletin. Materials and Methods should he incorpo- rated into appropriate figure legends. See the article by Loh- mann el al. (October I'J'io. Vol. 179: 214-218) for sample style. A Research Note will usually appear within two months after its acceptance. The Editorial Board requests that regular manuscripts con- form to the requirements set below: those manuscripts that do not conform will be returned to authors for correction before review. 1 . .Manuscripts. Manuscripts, including figures, should be submitted in quadruplicate, with the originals clearly marked. (Xerox copies of photographs are not acceptable for review pur- poses.) The submission letter accompanying the manuscript should include a telephone number, a FAX number, and (if possible) an E-mail address tor the corresponding author. The original manu- script must be typed in no smaller than 12 pitch or 10 point, using double spacing (including figure legends, footnotes, bibliography, etc. l on one side of \h- or 20-lb. bond paper. 8 by 11 inches. Please, no right justification. Manuscripts should be proofread caiclullv and errors corrected legibly in black ink. Pages should be numbered consecutively. Margins on all sides should be at least 1 inch (2.5 cm). Manuscripts should conform to the Council of Hioldg\ Editors Style Manual, 5th Edition (Council of Biology Editors. 1983) and to American spelling. Unusual abbreviations should he kept to a minimum and should be spelled out on first reference as well as defined in a footnote on the title page. Manuscripts should he divided into the following components: Iitle page. Abstract (of no more than 200 words). Introduction. Materials and Methods. Results, Discussion, Acknowledgments. Literature < 'iiol. I ahles. and Figure Legends. In addition, authors should supplv a list ol words and phrases undei which the article should he indexed. 2. Title page. The title page consists of a condensed title or running head of no more than 35 letters and spaces, the manuscript title, authors' names and appropriate addresses, and footnotes listing present addresses, acknowledgments or contribution num- bers, and explanation of unusual abbreviations. 3. Figures. The dimensions of the printed page, 7 by 9 inches, should be kept in mind in preparing figures for publication. We recommend that figures be about 1 times the linear dimensions of the final printing desired, and that the ratio of the largest to the smallest letter or number and of the thickest to the thinnest line not exceed 1:1.5. Explanatory matter generally should be included in legends, although axes should always be identified on the illustra- tion itself. Figures should be prepared for reproduction as either line cuts or halftones. Figures to be reproduced as line cuts should be unmounted glossy photographic reproductions or drawn in black ink on white paper, good-quality tracing cloth or plastic, or blue-lined coordinate paper. Those to be reproduced as halftones should be mounted on board, with both designating numbers or letters and scale bars affixed directly to the figures. All figures should be numbered in consecutive order, with no distinction between text and plate figures and cited, in order, in the text. The author's name and an arrow indicating orientation should appear on the reverse side of all figures. Color: The Biological Bulletin will publish color figures and plates, but must bill authors for the actual additional cost of printing in color. The process is expensive, so authors with more than one color image should consistent with editorial concerns, especially citation of figures in order combine them into a single plate to reduce the expense. On request, when supplied with a copy of a color illustration, the editorial staff will provide a pre-publi- eation estimate of the printing cost. 4. Tables, footnotes, figure legends, etc. Authors should follow the style in a recent issue of The Biological Bulletin in preparing table headings, figure legends, and the like. Because of the high cost of setting tabular material in type, authors are asked to limit such material as much as possible. Tables, with their headings and footnotes, should be typed on separate sheets, num- bered with consecutive Roman numerals, and placed after the Literature Cited. Figure legends should contain enough informa- tion to make the figure intelligible separate from the text. Legends should be typed double spaced, with consecutive Arabic numbers, on a separate sheet at the end of the paper. Footnotes should be limited to authors' current addresses, acknowledgments or contri- bution numbers, and explanation of unusual abbreviations. All such footnotes should appear on the title page. Footnotes are not normally permitted in the body of the text. 5. Literature cited. In the text, literature should be cited by the Harvard system, with papers by more than two authors cited as Jones et a/., 1980. Personal communications and material in prep- aration or in press should be cited in the text only, with author's initials and institutions, unless the material has been formally accepted and a volume number can be supplied. The list of references following the text should be headed Literature Cited, and must be typed double spaced on separate pages, conforming in punctuation and arrangement to the style of recent issues of The Biological Bulletin. Citations should include complete titles and inclusive pagination. Journal abbreviations should normally follow those of the U. S. A. Standards Institute (USASI), as adopted by BIOLOGICAL ABSTRACTS and CHEMICAL ABSTRACTS, with the minor differences set out below. The most generally useful list of bio- logical journal titles is that published each year by BIOLOGICAL ABSTRACTS (BIOSIS List of Serials; the most recent issue). Foreign authors, and others who are accustomed to using THE WORLD LIST OF SCIENTIFIC PERIODICALS, may find a booklet published by the Biological Council of the U.K. (obtainable from the Institute of Biology. 41 Queen's Gate. London, S.W.7, England, U.K.) useful, since it sets out the WORLD LIST abbreviations for most biological journals with notes of the USASI abbreviations where these differ. CHEMICAL ABSTRACTS publishes quarterly supplements of addi- tional abbreviations. The following points of reference style for THE BIOLOGICAL BULLETIN differ from USASI (or modified WORLD LIST) usage: A. Journal abbreviations, and book titles, all underlined (for italics) B. All components of abbreviations with initial capitals (not as European usage in WORLD LIST e.g., J. Cell. Comp. Physiol. NOT J. cell. comp. Physiol. ) C. A/I abbreviated components must be followed by a period, whole word components must not (i.e., J. Cancer Res.} D. Space between all components (e.g.. J. Cell. Comp. Physio/., not J.Cell.Comp.Physiol.) E. Unusual words in journal titles should be spelled out in full, rather than employing new abbreviations invented by the author. For example, use Kit Visindafjelags Is/eiulinga without abbreviation. F. All single word journal titles in full (e.g.. Veliger, Ecol- ogv. Brain}. G. The order of abbreviated components should be the same as the word order of the complete title (i.e., Proc. and Trans. placed where they appear, not transposed as in some BIOLOGICAL ABSTRACTS listings). H. A few well-known international journals in their preferred forms rather than WORLD LIST or USASI usage (e.g.. Nature, Science, Evolution NOT Nature, Loud., Science. N.Y.; Evolution, Lancaster, Pa. ) 6. Reprints, page proofs, and charges. Authors of articles in black and white (no color figures) receive their first 50 reprints (without covers) free of charge. Color reprints and additional black-and-white reprints may be purchased; authors will receive order forms. Reprints normally will be delivered about 2 to 3 months after the issue date. Authors (or delegates for foreign authors) will receive page proofs of articles shortly before publi- cation. They will be charged the current cost of printers' time for corrections to these (other than corrections of printers' or editors' errors). Other than these charges for authors' alterations. The Biological Bulletin does not have page charges. Reference: Biol. Bull. 198: 1-21. (February 2000) Concepts in Imaging and Microscopy Imaging Membrane Potential With Voltage-Sensitive Dyes MICHAL ZOCHOWSKI, 1 -* MATT WACHOWIAK, 1 CHUN X. FALK, 2 LAWRENCE B. COHEN, 1 2 YING-WAN LAM, 1 SRDJAN ANTIC,' AND DEJAN ZECEVIC 1 1 Department of Cellular and Molecular Physiology, Yale University School of Medicine New Haven, Connecticut 06520: ami 2 RedShirtlmaging, LLC, Fairfield. Connecticut 06432 Abstract. Membrane potential can be measured optically using a variety of molecular probes. These measurements can be useful in studying function at the level of an indi- vidual cell, for determining how groups of neurons generate a behavior, and for studying the correlated behavior of populations of neurons. Examples of the three kinds of measurements are presented. The signals obtained from these measurements are generally small. Methodological considerations necessary to optimize the resulting signal-to- noise ratio are discussed. Introduction An optical measurement of membrane potential using a molecular probe can be beneficial in a variety of circum- stances. One advantage is the ability to measure from many locations simultaneously. This is especially important in the study of nervous systems in which many parts of an indi- vidual cell, or many cells, or many regions of the nervous system are active at the same time. In addition, optical recording offers the possibility of recording from processes that are too small or fragile for electrode recording. Several optical properties of membrane-bound dyes are sensitive to membrane potential, including fluorescence, absorption, dichroism, birefringence, fluorescence reso- nance energy transfer, nonlinear second harmonic genera- tion, and resonance Raman absorption. However, because the vast majority of applications have involved fluorescence or absorption, these will be emphasized in this review. All of the optical signals described here are "fast" signals (Co- hen and Salzberg, 1978) that are presumed to arise from membrane-bound dye; they follow changes in membrane Received 22 June 1999; accepted 19 October 1999. * To whom correspondence should be addressed. E-mail: mrz@ fred.med.yale.edu This is the fourth in a series of articles entitled "Concepts in Imaging and Microscopy." This series is supported by the Optical Imaging Association (OP1A) and was introduced with an editorial in the April 1998 issue of this journal (Biol. Bull. 194: 99). Other articles in the series are listed on The Biological Bulletin's, website at http://www.mbl.edu/html/BB/home. BB.html. potential with time courses that are rapid compared to the rise time of an action potential. Studies of the molecular mechanisms that result in po- tential-dependent optical properties have produced evidence supporting three mechanisms (for different dyes): dipole rotation, electrochromism, and a potential-sensitive mono- mer-dimer equilibrium. Dye mechanisms are discussed in Waggoner and Grinvald (1977), Loew et al. (1985), and Fromherz et al. (1991). We begin with examples of results obtained from mea- surements addressing three quite different neurobiological problems. In all three instances the camera was a photo- diode array with only 464 pixels (NeuroPlex; RedShirt- lmaging, LLC, Fairfield, CT). Despite this low spatial res- olution, the camera resolution was not limiting in any of the examples. On the positive side, this camera has an outstand- ing dynamic range; with an incident intensity of more than 10' photons per frame, it measures signals that are a fractional change (AI/I) of one part in 10 5 . In addition, it has a frame rate of 1.6 kHz (fast enough to measure most neurobiologically important signals). On the other hand, recently introduced CCD cameras not only have similar frame rates but also have lower dark noise. Thus, in one of the three measurements, we think that improved signal-to- noise ratios could be obtained with a cooled CCD camera. The optical signals in the example measurements are not large they represent fractional changes in light intensity (AI/I) of from 1(T 4 to 3 X 1CT 2 . Nonetheless, they can be measured with an acceptable signal-to-noise ratio after at- tention to details of the measurement that are described in the second part of the paper. Figure 1 illustrates three qualitatively different areas of neurobiology in which imaging membrane potential has been useful. First (left panel), to know how a neuron inte- grates its synaptic input into its action potential output, one must be able to measure membrane potential wherever synaptic input occurs and at the places where spikes are initiated. Second (middle panel), to understand how a ner- vous system generates a behavior, the action potential ac- tivity of many (all) of the participating neurons must be PARTS OF A NEURON Many Detectors - One neuron Potential changes in dendntes. Microinject dye: stain one neuron. M ZOCHOWSKJ ET A/- INDIVIDUAL NEURONS One Detector - One Neuron Follow spike activity of many individual neurons. Bath applied dye. POPULATION SIGNALS One Detector - Many Neurons Signals are population average. Vertebrate brain. Bath applied dye. 0.2 mm 2 mm single neuron 0.5 mm nvertebrate ganglion Figure I. Schematic of the three kinds of measurements described as examples. (Lett) An indmdual cortical hippocampal CAI pyramidal cell. Each pixel of the -JM-elcment photodiode array receues light from a small part of the dendnte. axon. or cell body of the neuron. An optical measurement of membrane potential pro\ ides information about how the neuron converts its synaptic input into its spike output. (Middle) A slice through ,m imertebrate ganglion with its cell bodies m a cortex around the outside and neuropil in the middle. Here each detector receives light from one or a small number ol cell bodies A voltage-sensitive dye measurement of spike activity while the ganglion is generating a behavior provides information about how the heha\ lor is generated. iRighn A vertebrate brain with the superimposed 4M-element photodiode array (used in all three examples) Each pixel of the array receives light from thousands of cells and processes. The signal is the population average of the change in membrane potential in those cells and processes. The image of the hippocampal neuron \\.is taken from Mainen el at. (1996). measured simultaneously. Third (right panel), responses to sensory stimuli and generation of motor output in the ver- tebrate brain are often accompanied by synchronous activa- tion of many neurons in widespread brain areas; voltage- sensitive dye recordings allow simultaneous measurement of population signals from many areas In these three in- stances, optical recordings have provided kinds of informa- tion about the function of the nervous system that were previously unobtainable. In the second half of the article, we describe the experi- mental details that are important in obtaining the signal-to- noise ratios achieved in the experiments described in the first section. We discuss signal type, dyes, light sources, photodetectors. and optics. Three Kxamples /. Processes of an individual neuron (Fig. 1, left panel) Understanding the biophysical properties of single neu- rons and how they process information is fundamental to understanding how the brain works. With the development of new measuring techniques ih.ii ;illow more direct inves- tigation of individual nerve cells, it became widely recog- nized, especially during the last 20 years, that dendritic membranes of many vertebrate CNS neurons contain active conductances such as voltage-activated Na" 1 , Ca : *. and K ' channels (e.g., Stuart and Sakmann. 1 994; Spruston et al., 1995: Magee and Johnston. I995: Magee el ai. 1W5). An important consequence of active dendrites is that regional electrical properties of branching neuronal processes will be extraordinarily complex, dynamic, and. in the general case, impossible to predict in the absence of detailed measure- ments. To obtain such a measurement one would, ideally, like to monitor, at multiple sites, subthreshold events as they travel from the sites of origin on neuronal processes and summate at particular locations to influence the initiation of action potentials. This goal has not been achieved in any neuron, vertebrate or invertebrate, due to the technical limitations of measurements that employ electrodes. Better spatial resolu- tion necessitates a turn from direct electrical recording to indirect, optical measurements using voltage-sensitive dyes. Recently, the sensitivity of intracellular voltage-sensitive dye techniques for monitoring neuronal processes in ^itu has IMAGING MEMBRANE POTENTIAL SOMA STIMULATION Br 1 Br 3 ___v tlO> 70 ms V / 15ms Figure 2. (A) Giant metacerebral neuron from the left cerebral ganglion 12 h after injection with the fluorescent voltage-sensitive dye JPW1 1 14. The cell body and main processes are clearly visible in the unfixed preparation. Excitation wavelength: 540 30 nm; dichroic mirror: 570 nm. long-pass barrier filter: 610 nm. (B) (Ba) Voltage-sensitive dye recording of action potential signals from elements of the phntodiode array positioned over the image of axonal arborizations of a metacerebral cell in the left cerebral ganglion. Axonal branches are marked Br 1-4. Spikes were evoked by transmembrane current steps, as shown in (Bb), delivered through the recording microelectrode in the soma. Each optical trace in (Ba) represents 70 ms of recording centered around the peak of the spike as indicated by the time bar in (Bb). Each diode received light from a 50 x 50 /nm area in the object plane. (Be) Recordings from four locations indicated in panel (Ba). scaled to the same height, are compared to determine the site of the origin of the action potential and the direction of propagation. (Bd) Color-coded representation of the data shown in (Ba) indicating the size and location of the primary spike-tngger zone and the pattern of spike propagation. Consecutive frames represent data points that are 1.6 ms apart. The color scale is relative, with the peak of the action potential for each detector shown in red (modified from Zecevic, 1996). been improved by a factor of roughly 150, allowing direct recording of subthreshold and action potential signals from the neurites of invertebrate neurons (Antic and Zecevic, 1995; Zecevic, 1996). The improvement in the signal-to- noise ratio is based on previous experience from other laboratories (Davila et al., 1974; Grinvald et ai, 1987) and on ( 1 ) finding an intracellular dye that provides a relatively large fractional change in fluorescence and (2) improve- ments in the apparatus to increase the incident light inten- sity, to lower the noise, and to filter more efficiently. En- couraging results have also been obtained in initial studies on vertebrate CNS neurons in brain slices (see below). An invertebrate neuron. A typical result of a multi-site voltage-sensitive dye recording is shown in Figure 2. The fluorescent image of a metacerebral Helix neuron follow- ing injection with the fluorescent voltage-sensitive dye JPW1 1 14 is shown in Figure 2A. The image of the cell was projected onto the array of photodiodes as indicated in M ZOCHOWSKI ET AL Figure 2Ba. This panel represents multi-site recording of action-potential signals from axonal branches Br2. Br3. and Br4. evoked by a transmembrane current step (Fig. 2Bh>. Optical signals associated with action potentials, expressed as fractional changes in fluorescent light intensity (AF/F). were between 3 x 10 ' and 3 X 10 : in recordings from the processes. With these measurements it is straightfor- ward to determine the direction and velocity of action- potential propagation in neuronal processes. In Figure 2Bc. recordings from different locations, scaled to the same height, are compared to determine the site of origin of the action potential. The earliest action potential, in response to MIIIKI stimulation, was generated near location 2, in the axonal branch Br4. situated in the cerebral-buccal connec- tive outside the ganglion. The spike propagated orthodromi- cally from the site of initiation toward the periphery in branch Br4 and antidromically toward the soma and into the branch Br3 in the external lip nerve. The direction of propagation is clear from the color-coded representation of the data (Fig. 2Bd). This figure shows the potential changes in the branching structure at nine different times separated bv 1 .6 ms. Red corresponds to the peak of the action potential. The panels show the position of the action-poten- tial trigger /one at location 2 and orthodromic and anti- dromic spread of the nerve impulse from the site of initia- tion. The earliest spike was evoked about I nun from the soma. The spike-initiation segment in the a\on is roughly 300 /xni in length and remote from the soma. It appears that under normal conditions, slow depolari/ing voltage pulses applied to the soma are electrotonically spread into the processes with little attenuation. These depolari/ing pulses initiate action potentials in the processes at remote sites that are more excitable than neighboring segments. Light scattering in the ganglion limits the maximum useful spatial resolution in this kind of measurement. Thus the 24 X 24 pixel resolution of NeuroPlex appears to he adequate in this circumstance. On the basis of similar measurements, we recently deter- mined that the Helix neuron in Figure 2 has multiple trigger zones that can be independently activated. The precise pat- tern of action-potential initiation and propagation within the whole branching structure of a neuron can be analy/ed by multi-site recording. The information about the spatial and temporal dynamics of neuronal signals can be used to con- strain the choice of channel densities and geometrical fac- tors in biophysical models thai are used to descnbe tune tional properties of neurons. A vertebrate neuron. It is of considerable interest to apply the same technique to .], -mlntes of vertebrate CNS neurons in brain slices. Apart from our preliminary experiments (Kogan i-t ai. 1995). there is no previous experience in this field. Experiments were carried out on pyramidal neurons in slices from the neocortex of 14- to 18-day-old rats. The fluorescent image of the cell was projected onto the octag- onal photodiode array. In the example shown in Figure 3A, the neuion was stimulated, by depolari/ing the cell hodv. to produce a burst of two action potentials. Each trace in Figure 3B represents the output of one photodiode for 44 ms. Optical signals associated with action potentials, ex- pressed as fractional changes in fluorescent light intensity (AF/F). were between 10 and 3 X 10 : in recordings from the processes. In Figure 3C. the electrical recordings from the soma (smooth line) were compared with the optical signals filtered to eliminate high-frequency noise (dashed line I. The time courses of electrical and optical recordings agree well. In panel 1). recordings from different locations. scaled to the same height, are compared on an expanded lime scale. Each trace is a spatial average from two adjacent detectors. Both spikes in the burst originated near the soma and propagated centrifugally along the apical and basolat- eral dendriles (action potential back-propagation: Stuart and Sakmann. 1994). These initial experiments demonstrate several important methodological results. First, it is possible to deposit the dye into the cell without staining the surrounding tissue (keep- ing background fluorescence low). Second, the pharmaco- logical effects of the dye were completely reversible if the staining pipette was withdrawn, and the cell was allowed to recover for 1-2 h. Third, the level of photody namic damage alreadv allows meaningful measurements and could be re- duced further. Finally, the sensitivity of the dye was com- parable to that achieved in the experiments on invertebrate neurons (Zecevic. 1996). In these preliminary experiments. the ilye spread for roughly 500 /xm into dendritic processes within 2 h. One way to improve the staining is to attach the dye electrode to the distal region on a dendrite as done previously w-ith calcium-sensitive dyes (e.g.. Markrarn and Sakmann. 1994; Schiller camera might also be useful. 2. Action potentials from imliriiliuil nfiiron\ -'ix. I. iniililli- /mud) Aplysia abdominal ganglion. Nervous systems are made up of large numbers of neurons, and main of these are active during the generation of behaviors. The original motivation for developing optical methods was the hope that they could he used to record all of the action-potential activity of all the neurons in simpler invertebrate ganglia during behaviors (Davila ft . Oregon ing appeared to he specitic. hut nonetheless the signals were small. I 1 1,- . I'li.rv xciiMm c dvc signals discussed in this article are presented as the fractional intensity change (AI/I). These signals give information about the time course of the po- tential change hut no direct information about its magni- tude. In some instances, indirect information about the magnitude of the voltage change can he obtained (e.g., Orbach ct til.. 19K5: Delaney ct til.. 1994: Antic and Zecevic. 1995). .Another approach is the use of ratiometric measurements at two independent wavelengths (Gross et (i/.. 1994). However, to determine the amplitude of the voltage change from a ratio measurement, one must know the fraction of the fluorescence that results from dye not bound to active membrane, a requirement that is onlv ap proximately met in special circumstances (r.t,'.. tissue cul- ture). Measuring Technology Noise. 1 . Shot noise. The accuracy with which light can he measured is limited by the shot noise arising from the statistical nature of photon emission and detection. The number of photons emitted per unit time lluclu ates, and if an ideal light source (tungsten lilament at 33(K)"F) emits an average ot !()"' photons/ms. the root-mean-square (RMS) deviation in the number emitted is the square root of this number, or 10 s photons/ms. The signal-to-noise ratio is directly pro- portional to the square nmi ol the number of mea- sured photons and inverse!) proportional to lite square root of the bandwidth of the photodetection svstem (Braddick. I960; Malmstadt ct r (top of figure); when a 500-/nm-thick slice of olfactory bulb is present, the light from the small spot is spread to about 200 jim (bottom of figure i. Scattering ap- pears to be greater with mammalian cortex than with olfactory bulb. Thus, light scattering blurs signals in adult vertebrate preparations. A second source of blurring is signal from regions that are out of focus. For example, if the active region is a cylinder (a column ) perpendicular to the plane of focus, and the objective is focused at the middle of the cylinder, then the light from the in-focus plane has the correct diameter at the image plane. How- ever, the light from the regions above and below is out of focus, and its diameter is too large. The middle section of Figure 12 illustrates the effect of moving the small spot of light 500 /nm out of focus: the light is spread to about 200 /nm. Thus, in preparations with considerable scattering or with out-of-focus signals, the spatial resolution may be limited by the prepara- tion and not by the number of pixels in the imaging device. Comparison of voltage-sensitive dye and local field po- tential measurements of population signals. The results in Figure I 2 suggest that the spatial resolution of a population signal in a vertebrate brain is limited to about 200 /nm by light scattering and out-of-focus signals. This spatial reso- lution is substantially better than that obtained by recording local field potentials from the surface of the brain. Because the current sources driving these potentials can be below the surface, the spatial patterns are smoothed by the volume- conductor properties of the cortical tissue. Freeman (1978) SPATIAL RESOLUTION OF LFP AND OPTICAL RECORDINGS; TWO LOCATIONS ON CORTEX SEPARATED BY 2.3 MM. LOCAL FIELD-POTENTIAL (Correlation Coellicient= 0.9) VOLTAGE-SENSITIVE DYE RECORDINGS (Correlation Coefficient = 0.6) Looming Stimulus 400 ms Fi|>urv 13. ( 'ompanson nl simultaneous optical and local-lield-polcntial recordings from two positions on tunic visual cortex thai were separated by 2.3 mm. The top pair of superimposed traces are the local-licld- polential recordings from the two sites. The bottom pair of traces are the vollage-sensitive-dye recordings from llu- iw<> Mk-s 'I here is much less overlap in llie optical recordings. Thus, the voltage-sensitive dye recordings have heller spaiial resnlnii.ni lii.ih sets of recordings were hand-pass tillered (10-30 H/l. (J. Prechll. I H Cohen, .nul I ) Mrml.-lil. impiihl data i IMAGING MEMBRANE POTENTIAL 17 reported that spatial Fourier transforms from local field potential measurements had a sharp cutoff at about 1 cycle/ mm. suggesting a resolution on the order of 1 mm. Simi- larly, in the frequency range below 30 Hz, Bullock and McClune ( 1989) reported a correlation of about 0.9 between two local field potential electrodes spaced 2-3 mm apart on the surface of the brain. We compared the spatial resolution of the two kinds of measurements by examining simultaneous pairs of record- ings from two positions on turtle visual cortex that were separated by 2.3 mm. The top pair of superimposed traces in Figure 1 3 are the local field potentials. The recordings from the two sites show considerable overlap: the correlation coefficient was 0.9. The bottom pair of traces are the volt- age-sensitive dye recordings. These show less overlap their correlation coefficient was 0.6. Thus, the spatial dif- ferences are less blurred with the optical recordings than with the electrode recordings. We estimated the relative spatial resolution of the two methods by determining the distance on the cortex at which the correlation coefficients would be equal for a pair of voltage-sensitive dye recordings and a pair of local field potential recordings. In six trials from two preparations the correlation coefficients were equal when the two optical measurements were a distance apart that was 0.21 0.05 (SEM) of the distance of the two electrodes. Thus, the optical measurement has a linear spatial resolution about 5 times better than the electrode measurement and a two- dimensional resolution about 25 times better. Confocal and two-photon microsco/H's. The confocal mi- croscope (Petran and Hadravsky, 1966) substantially re- duces both the scattered and out-of-focus light that contrib- utes to the image. A recent modification using two-photon excitation of the fluorophore further reduces out-of-focus fluorescence and photobleaching (Denk el al., 1995). The spatial resolution of images from intact vertebrate prepara- tions is much better with these types of microscope than with ordinary microscopy. These microscopes have been successfully used to monitor changes in calcium concentra- tion inside small processes of neurons (Eilers el al., 1995; Yuste and Denk, 1995). However, at present the sensitivity of these microscopes is relatively poor and they are rela- tively slow; there are no reports of their use to measure the small signals obtained with voltage-sensitive dyes of the type discussed in this article. On the other hand, slower voltage-sensitive dye signals have been measured confo- cally (Loew. 1993). Random-access fluorescence microscopy. Bullen el al. (1997) have used acousto-optic deflectors to construct a random-scanning microscope and were able to measure signals from parts of cultured hippocampal neurons. To reduce the effects of fluctuations in the laser output, the fluorescence signals were divided by the output of a pho- todetector sampling the incident light. Relatively large signal-to-noise ratios were obtained using voltage-sensitive dyes. This method has the advantage that only a small proportion of the preparation is illuminated, thereby reduc- ing photodynamic damage from the very bright laser light source. However, this method will probably be restricted to preparations such as cultured neurons, in which there is relatively little light scattering. Photoileiectors. Because the signal-to-noise ratio in a shot-noise-limited measurement is proportional to the square root of the number of photons converted into pho- toelectrons (see above), quantum efficiency is important. Silicon photodiodes have quantum efficiencies approaching the ideal ( 1 .0) at wavelengths at which most dyes absorb or emit light (500-900 nm). In contrast, only specially chosen vacuum photocathode devices (phototubes, photomultipli- ers, or image intensifiers) have a quantum efficiency as high as 0.15. Thus, in shot-noise-limited situations, a silicon diode has a signal-to-noise ratio that is at least 2.5 times larger. Photographic film has an even smaller quantum efficiency (0.01; Shaw, 1979), and thus has not been used for the kinds of measurements discussed in this paper. Imaging devices. Many factors must be considered in choosing an imaging system. Perhaps the most important are the requirements for spatial and temporal resolution. Because the signal-to-noise ratio in a shot-noise-limited measurement is proportional to the number of measured photons, increases in either temporal or spatial resolution reduce the signal-to-noise ratio. Our discussion considers systems that have frame rates near 1 kHz. In most of these systems, the camera has been placed in the objective image plane of a microscope. However. Tank and Ahmed ( 1985) suggested a scheme by which a hexagonal close-packed array of optical fibers is positioned in the image plane, and individual photodiodes are connected to the other end of the optical fibers. NeuroPlex, a 464-pixel photodiode array camera (RedShirtlmaging, LLC, Fairfield, CT) is based on this scheme. Silicon diode arru\x. I. Parallel readout arrays. Diode arrays with from 124 to 1 020 elements are now used in several laboratories (e.g., lijima el al.. 1989; Zecevic el al., 1989: Na- kashimatv*//.. 1992; Hirota et al., 1995). In addition, Hammamatsu has constructed a system with 2500 elements. These arrays are designed for parallel read- out; each detector is followed by its own amplifier, whose output can be digitized at frame rates of 1 kHz. Although the need to provide a separate ampli- fier for each diode element limits the number of pixels in parallel read-out systems, it contributes to the very large ( 10 ? ) dynamic range that these systems can achieve. Amplifiers have been discussed by Wu and Cohen (1993). Two parallel readout array sys- tems are commercially available: Argus-50 (256 pix- 18 \1 /.OCHOWSKl ET AL l.ll.lc I Characteristics of fast CCD camera systems tas reported h\ the manufacturer) Frame rale (HZ) Well size Read noise Back Bits Camera system full frame 1 KKXI cl (electrons) ilium. .1 in il Pixels Brain Vision' 1600 12 92 x 64 DalsaCA-DI 756 12 128 X 128 LSR. NeuroCAM UltraPix* RedShinlmaging NeuroCCD 4 TILL Photonics. IMAGO' 500 2700 140 300 200 (5 5 @100H/ 35 @ 100(1 H/ 14 @|40 Hf yes yes 110 12 14 12 sn < 80 74 x 74 160 x 120 A blank space indicates that the data were not available. The World Wide Web sites listed helim wcie accessed on 15 November 1999. ' www.brainVision.co.jp : www.dalsa.com ' www.lsr.co.uk/cameras/ultragraph.htm 4 \vww. redshirtimaging.com 5 www.till-photonics.com els), manufactured by Hammainatxu Photonics K.K. (www.hpk.co.jp). and NeuroPlex (464 pixels), by RedShirtlmaging. LLC (www.redshirtimaging.com). 2. Serial readout arrays. Use of a serial readout greatly reduces the number of amplifiers. In addition, it is simple to cool CCD (charge-coupled device) chips to reduce the dark noise. However, because of satura- tion, presently available CCD cameras are not opti- mal for the higher intensities available in some neurobiological experiments (Fig. 11). The high-in- tensity limit of the CCD camera is set by the light intensity that tills the electron wells on the ('CD chip. This accounts tor the holding over of the CCD camera performance at segment I? in Figure II. A dynamic range ol e\en 10' is not easily achieved with a CCD camera. A CCD camera will not he optimal for measurements of absorption or for fluo- rescence measurements on in vitro slices and intact brains in measurements where all of the membranes are stained (Fig. 1 1 1. The light intensity would have to be reduced, with a consequent decrease in signal- to-noise ratio. On the other hand. CCD cameras are close to ideal for measurements from individual neu- rons stained with internally injected dyes. Table I compares several CCD cameras with frame rates near 1 kH/. 3. Vacuum photocathixle cameras. Although the lovvei quantum efficiencies of vacuum photocathodes are a disadvantage, these devices may have lower dark noise. One such device, a Radechon (Ka/an and Knoll. I96S). has an output proportional to the changes in the input. Vacuum photocathode cameras have been used in rel- atively low time resolution recordings from mammalian cortex (Blasdel and Salama. 19X6) and salamander ol- factory bulb iKauer. I9NX; Cmelli cl nl.. 1995) and are used in the Imager 2(X)1. Optical Imaging. Inc. (www.opt-imaging.com i. Kuture Directions Because the apparatus for measuring light is already reasonably optimi/cd (see above), large improvements in the sensitivity of the techniques that use voltage-sensitive dyes must come from the development of better dyes or from investigating signals from other optical properties of the dyes. The dyes in Figure 9 and most of those synthesized are of the general class named polyenes (llamer. 1964), a group that was lirst used to extend the wavelength response of photographic tilm. It is possible thai improvements in signal si/e can he obtained with new polyene dyes (see Waggoner and Grinvald. 1977. and Fromher/ cl nl., 1991. for a discussion of maximum possible fractional changes in absorption and fluorescence). On the other hand, the frac- tional change on squid axons has not increased in recent years (Gupta cl nl.. l l >XI; I.. B. Cohen. A. Grinvald. K. Kamino. and B. M. Sal/berg, unpub. data I. and most mi provemcnis (e.g., (irinvald ct nl.. 1982b: Momose-Sato cl ,//.. 1 995: Tsati I Grunald. A.. R. Hildeshtim. I. C. Farber. and I . \mjlixu-r. I982b. Improved fluorescent probes for the measurement of rapid changes in membrane potential. Biophys. J. 39: 301-308. i.rim.il.l. A.. B. M. Salzherg. V. Lev-Ram, and K. IliUlislu-im. 1987. Optical recording of synaplic potentials from processes ot single ML-II rons using intracellular potentiomelric dyes Biophys I 5l:iv4.-65l i .nm.ilil. A.. R. D. Frostig. K. I.ieke, and R. Hildt-shcim. 19X8. Op- tical imaging of neuronal activity. Phyxiol. K< 'i 68: I 2S5 1 '(.6 Gross. F... R. -S. Bedlack Jr.. and I.. M. l.orw. 1994. Dual wavelength ratiometric fluorescence measurements ot the membrane dipolc poten- tial. Biophys. J. 67: 208-216. i ,n|.i.j. R. K.. B. M. Sal/berg. A. (irinvald. I.. B. Cohen. K. Kamino. S. Lesher, M. B. Boy It. A. S. Waggoner, and C. 11. Wang. 1981. Improvements in optical meihoils lor tneasuung rapid changes in membrane potential. J. Mcmhi. Hint. 58: 123-137. llamer. F. M. 1964. The Cytininc /Ho uiul Kclutcil Coinpnuihh. Wiley. New York. 1 Hickie. (.. I*. \\cnner. M. O'Donovan. Y. Isau. .17 Fang, and I.. B. Cohen. 1996. Optical monitoring ol activity trom individual and identified populations ot neurons relrogradelv labeled with voltage sensitive dyes. Soc. Neurosci. Absrr. 22: 321 Ilirota, A.. K. Sato, Y. Momosc-Sato. I . Sakai. and K. Kamino. 1995. A new simultaneous 1020-site opne.il tccording system lor monitoring neural activity using voltage-sensitive dyes ./ Yrrf>w i. Methods 56: 187-144. Hughes, .1. R.. and ,|. \. Ma/urowski. 1962. Studies on the supracal- losal mesial cortex ol unaiicsthcli/cd. conscious mammals. II Monkey B. Responses trom the ollactorv bulb. Elciirofnii'i>hiiln^i Clm \< u rophysiol. 14: 635-645. lijima, 'I'.. M. Ichikawa, and (i. Malsuinoto. 1989. Optical momlonng of LTF and related phenomena. 5oi Veurosci \/>\;/ 15: ;I| N I noiie, S. 1986. Video Micrp\. Plenum Press. New ^i Hrk p I2S. Kautr.J. S. 1988. Real-time imaging ol evoked activnv in local ciicuits of the salamander olfactory bulb Vatur* 331: IM> Id.s Ka/an. B.. and M. Knoll. 1968. //,,;/.'/, /im/uc Sinruvf. Atademn. Press. New York. K I. ml. hi I)., and K. R. Delanev. 1996. Disinbiited icpiesenlalioii ol vibrissa movement in the upper layers of somatoseiisoiv coite\ ic vealed with voltage-sensitive dyes. J. Crimp. \rnn>l. 375: ,S l )-HIS Kogan, A.. \V. N. Ross. I). /.ece>ic. and N. I.asscr-Ross. 1995. Optical recording Iroin cerehellar Purkinje cells using inliaiellularlv injected voltage-sensitive dyes. Bruin Rt'\. 700: 235-239. Kii|ili i III.IMII. I.. H. I'insker. V. Casttllueei, and K. R. Kandel. 1971. Central and peripheral control of gill movements in Aply.^ia. S< " < i 174: 1252 1256. l.am. Y.-., Lawrence B. Cohen, Matt \\achowiak, and Michal R. /.iH-hoHski. 20(10. Odors elicit three different oscillations in the turtle olfactory bulb. J. Neumsci. In press. Laurent, (i., and M. Naraghi. 1994. Odoranl induced oscillations in the mushroom bodies of the locust. / Neurosci. 14(5 I'l 2i: 2')') \ 3(XM. Locw, L. M. 1993. Conlni'.il miv.iosi.opv ot poteiitiometrie fluorescent dyes. Methods Cell Hi,,!. 38: 195 I.IH-W. L. M.. I. It Coht-n. It. M. Sal/beru, A. I.. Ohaid, and 1. I'n /.i mil. i. 1985. Charge-shift probes of membrane potential. C'har- acteri/aiion of aminoslyrylpv 1 1. iiniiiin dyes on the sc|ind giant axon. i. V ''m. J. 47: 71-77. I.w. L. M.. I . It. Cohen. .1. l)i\. K. N. Fluhler. \. Montana, (i. Salania. and J. Y. Uu. 1992. A napthyl analog of the aminostyryl pyridinium class of potentiometric membrane dyes shows consistent sensitivity in a variety of tissue, cell, and model membrane prepara- tions. J. Memb. Hi,-l. 130: 1-10. London, J. A.. 1). /.i-cnic. and L. B. Cohen. 1987. Simultaneous optical recoidmg ot aclivuv trom many neurons during feeding in Navanax. J. Neurosci 7: i>4" 661. Maui-e. J. C., and I). Johnston. 1995. Synaplic activation of voltage- gated channels in the dendrites of hippocampal pyramidal neurons. Science 268: 301-104. ManiT, J. C'., <;. Christoti, H. Mi\ukawa, It. Christie. N. I.asser-Ross. and I). Johnston. 1995. Subthreshold svnaptic activation ot voltage- gatevl calcium channels mediate .1 loi.ili/ed calcium influx into den driles vit hippocampal pyramidal neurons. ./. ,\europliysiol. 74: 335- 342. Mainen. /.. F., N. T. Canu-vallt. A. M. /.ador, B. J. ( laiborne, and T. M. Brown. 1996. Electronic aiclntcctuiv of hippocampal CAI pvtamidal neurons based on three-dimensional reconstructions. 7. Neu- /,yi/nw,-/. 76: 1404-1923. Malmstadl. H. \ ., C. (J. Fnke. S. R. Crouch, and (J. Harlick. 1974. Eleciii'iin Measurements for Scientists. Benjamin. Menlo Park. CA. Markram, H.. and B. Sakmann. 1994. Calcium transients in dendrites ol neocortical neurons evoked by single Subthreshold excitatory postsy napiic potentials via low -v ullage-activated calcium channels. /'/,., Nat. \KI.I Sfi. 91: 5207-5211. Momosi'-Sato, Y., K. Sato, T. Sakai. V Ilirota. K. Matsutani. and K. Kamino. 1995. Evaluation of optimal voltage-sensitive dyes for op- tical measurement ol cmhivonic neural activity. J. Mcmbr. Hiol 144: 167-176. Nakashima, M., S. Yamada. S. Shiono. M. Mai-da, and F. Saloh. 1992. 448-deiector optical recording system: development and application to .\lil\-.\iit gill-withdrawal reflex. IEEE Trans. Biomed. Eng. 39: 26-36. Nirenln-rn, S.. and C. Cepko. 1993. Targeted ablation of diverse cell classes in the nervous system in vivo. J. Neurosci. 13: 3238-3251. Ohaid. A. L., T. Koyano. J. Linclstrom, T. Sakai. and B. M. Sal/berg. 1999. Spatiolemporal patterns ol activit) in an intact mammalian netwoik wiih single cell resolution: optical studies of nicotinic activity in an enteric plexus. ./. .\r/i>M;. 19: 3073-3093. Orhach. 11. S., and L. B. Cohen. 1983. Optical monitoring of activity liom many areas of the in vitro and in riro salamander olfactory bulb: a new method toi studving functional organi/ation in the vertebrate central nervous system ./ Veurosci. 3: 2251-2262. Orhaeh. 11. S.. L. It. Cohen, and A. (.rinvald. 1985. Optical mapping ot eleclncal activity in 1.11 somatosensory and visual cortex. 7. Neuro- v,, 5: issi, [895 IVIran, M.. and M. lladravskx. 1966. C/echoslovakian patent 7720. I'rechtl. J. C .. L. B. Cohen. B. IVsiran. P. P. Milra, and I). Kleinfeld. 1997. Visual stimuli induce waves ol eleclncal activity in turtle ex. t'roc. Nat. Acad. Sci. USA 94: 7621 "(.26 K.. i/l. ill I M .mil \. (rinvald. 1991. A landem-lens epitluoresccnce microscope: hundred-fold brightness adv antage for wide-held imaging. / \,-nn>s,-i. Mali. 36: 127-137. Rohr, S.. and B. M. Sal/herg. 1994. Multiple sue optical recording of Iransmembrane voltage in patterned growth heart cell cultures: assess- in ' electrical behavior, with microsecond resolution, on a cellular and subcellular scale. Hio/iln* .1 67: 1301-1315. Ross. \\. S.. and I.I. Reiehaidl. 1979. Species-specific effects on the optical signals of voltage-sensitive dyes. J. Memhr. Hiol. 48: 343-356. Ross, VV. N.. B. M. Sal/l>erg. L. B. Cohen. A. Criinald. II. \ . Davila. A. S. Waggoner, and ( . II. \\ang. 1977. Changes in absorption, fluorescence, dichroism. and birefringence in stained giant axons: op- tical measurement of membrane potential. ./. Mfinhr. Kid. 33: 141 183. IMAGING MEMBRANE POTENTIAL 21 Sakai, T., A. Hirota, H. Komuro. S. Fujii. and K. Kamino. 1985. Optical recording of membrane potential responses from early embry- onic chick ganglia using voltage-sensitive dyes. Brain Res. 349: 39- 51. Salama. G. 1988. Voltage-sensitive dyes and imaging techniques reveal new patterns of electrical activity in heart and cortex. Pmc. SPIE (Int. Sue. Optical Eng. ) 94: 75- 86. Salzberg, B. M. 1983. Optical recording of electrical activity in neurons using molecular probes. Pp. 139-187 in Current Methnji in Ccl/nlur Neurobiology. J. L. Barker and J. F. McKelvy, eds. Wiley. New York. Salzberg, B. M., H. V. Davila, and L. B. Cohen. 1973. Optical record- ing of impulses in individual neurones of an invertebrate central ner- vous system. Nature 246: 508-509. Salzberg, B. M., A. Grinvald, L. B. Cohen, H. V. Davila. and \V. N. Ross. 1977. Optical recording of neuronal activity in an invertebrate central nervous system: simultaneous monitoring of several neurons. J. Neuroplnsioi 40: 1281-1291. Schiller, J., F. Helmchen, and B. Sakmann. 1995. Spatial profile of dendritic calcium transients evoked by action potentials in rat neocor- tical pyramidal neurons. J. Physiol. 487: 583-600. Shaw, R. 1979. Photographic detectors. Appl. Optics Optical Eng. 7: 121-154. Siegel, M. S., and E. Y. Isacoff. 1997. A genetically encoded optical probe of membrane voltage. Neuron 19: 735-41. Spruston, N., Y. Schiller, G. Stuart, and B. Sakmann. 1995. Actmty dependent action potential invasion and calcium influx into hippocam- pal CA1 dendrites. Science 268: 297-300. Stuart, G. J., and B. Sakmann. 1994. Active propagation of somatic action potentials into neocortical pyramidal cell dendrites. Nature 367: 69-72 Tank, D., and Z. Ahmed. 1985. Multiple-site monitoring of activity in cultured neurons. Biophys. J. 47: 476A. Tsau, Y., J. Y. Wu, H. P. Hopp, L. B. Cohen. I). Schiminovich, and C. X. Falk. 1994. Distributed aspects of the response to siphon touch in Apl\sia: spread of stimulus information and cross-correlation anal- ysis. J. Neurosci. 14: 4167-4184. Tsau, Y., P. Wenner. M. J. O'Donovan, L. B. Cohen, L. M. Loew, and ,1. P. \\uskell. 1996. Dye screening and signal-to-noise ratio for retrogradely transported voltage-sensitive dyes. / Neurosci. Metli. 7(1: 121-129. Waggoner, A. S. 1979. Dye indicators of membrane potential. Annii. Rev. Binplm. Bioenx. 8: 47-68. Waggoner, A. S., and A. Grinvald. 1977. Mechanisms of rapid optical changes of potential sensitive dyes. Annn. NY Acini. Sci. 303: 217-241. Wu, J. Y., and L. B. Cohen. 1993. Fast multisite optical measurements of membrane potential. Pp. 389 104 in Fluorescent and Luminc\< cut Probes for Biological Activity. W. T. Mason, ed. Academic Press. London. Wu, J. Y., L. B. Cohen, and C. X. Falk. 1994a. Neuronal activity during different behaviors suggests a distributed neuronal organization in the Aplysia abdominal ganglion. Science 263: 820-823. Wu. J. Y., Y. Tsau. H. P. Hopp, L. B. Cohen, A. C. Tang, and C. X. Falk. 1994b. Consistency in nervous systems: trial-to-trial and ani- mal-to-animal variations in the response to repeated application of a sensory stimulus in Aplysia. J. Neurosci. 14: 1366-1384. Wu, J. Y., Y. W. Lam, C. X. Falk, L. B. Cohen. J. Fang, L. Loew, J. C. Prechtl, D. Kleinfeld, and Y. Tsau. 1998. Voltage-sensitive dyes for monitoring multi-neuronal activity in the intact CNS. Histochem. J. 30: 169-187. Yuste, R., and W. Denk. 1995. Dendritic spines as basic functional units of neuronal integration. Nature 375: 682-684. Zecevic, D. 1996. Multiple spike-initiation zones in single neurons re- vealed by voltage-sensitive dyes. Nature 381: 322-325. Zecevic, D.. J. Y. Wu, L. B. Cohen, J. A. London, H. P. Hopp, and C. X. Falk. 1989. Hundreds of neurons in the Aplysia abdominal ganglion are active during the Bill-withdrawal reflex. ./. Neurosci. 9: 3681-3689. Reference: Bio/. Bull. I'M: :: :> il.-hru.ir> 2000> Germ-Cell Warfare in Ascidians: Sperm From One Species Can Interfere With the Fertilization of a Second Species CHARLES C. LAMBERT* Department of Biological SCICIKC. California Stale I'niversity. Fullerton. California 92834-6850 Ascidians (invertebrate chordates) arc very abundant in man\ marine sitbtidal areas. They often live in dense multi- species clumps; thus, interspecific competition for space ma\ be intense. Although most noncolonial species are broadcast spawners, their eggs can be fertilized only by sperm of the same species ill. Multiple fertilization is lethal and all animals have evolved blocks to polyspcrmy. Ascid- ian eggs block polssperm\ b\ enzymatic 1 2) and electrical mechanisms (3). Sperm hind to N-acetylglucosamine groups on the vilelline coal 14. 5, 6. 7). i'ollicle cells surrounding the vitelline coat release N-acetylglucosamini- dase during egg activation l,\). preventing the binding of HII sperm but a few (2). I show here that this interaction is not species-specific; sperm from one species can cause glyco- sidase release from follicle cells of a second species, fur- thermore, once glycosidu.se release has been induced, the subsequent addition of sperm from the egg-producing spe- cies fails to fertilize a substantial proportion l these eggs. This leads to the hypothesis that sperm from one species of ascidian can interfere with fertilization oj a set nd species. While \nlT-d.\peciJic sperm competition has been well docu- mented in several ta\a (*). Ill), this is the fust record o/ sperm competition between species, or inlci S/XV//K sperm competition. As previously shown by numerous authors ( 1 i. ascidian fertili/.ution is species-specific. Eggs of I'hallusia mammil- lata, Phallusni julinea. Ascidia ( = Phallu.sia } nigra. and Ascidia svdneiensis arc induced In undergo cleavage only when fertili/ed with sperm from the same species (Table 1 1. In I'hallusia mammillata eggs, the failure to fertili/e is the result of sperm from U< /) of the recorded neurons showed various types of voltage responses to light stimuli. Almost all (27/28) of the recorded voltage responses \vere "on" responses that included hyperpolarizing and depolar- izing responses and < aitld he categori:ed into jive I\pcs. except lor a complex response recorded in one cell: the remaining one 11/28) \va\ a depolarizing "off" response. This is the first report of <'lectroph\siologieal recordings of light-sensitive voltage responses from ascidian cerebral ganglion neurons. Membrane potentials were recorded with intracellular microelcclnnles from 1 15 neurons in the ventral siile of the isolated cerebral ganglion of 47 ascidians I sec Fig. 1 legend for detail). Phe resting membrane potential was 4X * 12 mV (mean SI), n = 52). anil 52'- (n 1151 ot the recorded neurons showed spontaneous activity consisting of low-frequency (O.I . < ll/i icgular discharges ot action Received 13 July 1999; accqiiol : Novembi i I'' 1 ' 1 ' * To whom correspondent i- shmilil he .ulclresM-il I. in.nl c>k;i\<" iniiihs. s.u-tokyo.ai potentials (Fig. IB. 1)1. Seventeen percent of the neurons showed spontaneous irregular or bursting discharges of ac- tion potentials. The remaining 3(Kr of the neurons were silent (Fig. 1A. C. E). In addition. 24', <2S cells I of the total 115 recorded neurons, including spontaneously active and silent ones, showed \oltage responses of various types to light stimuli (Fig. I ). Most of these light responses (27/2S cells) were "on" responses (Fig. 1A-E). and we observed a transient "off depolarization with a few spikes in only one neuron (not shown). Twenty -six out of the 27 "on" re- sponses could he categori/ed into five types as follows (Fig. 1): (A) transient hyperpolari/ation (13 cells); (B) suppres- sion of spontaneous discharge of action potentials (7 cells): ((.') transient depolari/ation (2 cells); (D) transient high- frequency excitatory synaptic inputs (2 cells); (E) sustained depolari/ation I 2 cells). The remaining "on" response con- sisted of a complex pattern: transient hyperpolari/.ation fol- lowed by sustained depolari/ation accompanying a number of spikes during the light stimulus. This was similar to a mixture of response types A and E. There seemed to be no significant differences in the resting potentials for the live response types. Even the quickest of the live types of responses, type A, showed a latency longer than 500 ms. and ihe latency of the other types was 3-5 s. There seemed to be no specific distribution ol these types of neurons on Ihe ventral surface of the cerebral ganglion. The ability of a cell to respond with more than one response type when presented different light intensities or wavelengths was not investigated. Hyperpolari/ing receptor potentials of the visual cells in the ocellus of ascidian larvae has been described previously (9). Because our type A and B responses also are hyperpo- lari/ing. il is tempting, to interpret them as receptor poten- tials. However, there is also a significant difference between 26 LIGHT-SENSITIVE NEURONS IN THE ASCIDIAN CEREBRAL GANGLION 27 Neural Complex Branchial Aperture Branchial Sac Gonads_ _' D B E Figure 1. (Top left) Schematic illustration of adult dona savignyi showing location of major organs. The illustration was modified from Mackie (13). Neural complex is located in the body wall between the two siphons (boxed). (Top right) Ventral view of the isolated neural complex preparation. The neural complex consists of the cerebral ganglion and the neural gland. Recordings were made from neurons of the ventral side of the cerebral ganglion. (Bottom) The 27 "on" voltage responses to light stimuli could be categorized into five types, except for one complex response (see text). The typical response for each type is shown. (A) Transient (<5 s) hyperpolarization (13/27). Longer or continuous light stimulation did not evoke any additional voltage changes after the transient hyperpolarization. (B) Suppression of spontaneous regular discharges (7/27). Small irregular fluctuations of membrane potential indicative of synaptic inputs seemed to increase in frequency during light stimuli. Some of them are enlarged in the inset corresponding to 1 s X 10 mV. (C) Transient depolarization accom- panied by a few spikes (2/27). Longer or continuous light stimulation did not evoke any voltage changes after the transient depolarization. (D) Transient high-frequency excitatory synaptic inputs (arrow), which some- times build up to give rise to an action potential in spontaneously beating neurons. Unlike the type B response, the spontaneous regular discharges were not affected in these cells (2/27). (E) Sustained depolarization during light stimuli, accompanied by a number of spikes with frequency accom- modation (2/27). Voltage responses of each neuron against repeated light stimuli were reproducible during the recording period. Arrowheads in C and D indicate stimulus artifacts. Methods: Ciona savignvi was used for the present electrophysiological study because of the advantageous morphology of its neural complex the two; the receptor potential of the larva is fast and sustained, while the type A and type B responses recorded in the present study are slow and transient. We cannot conclude at present which type of response represents the photoreeeptor potential, because chemical and electrical synaptic transmissions were not blocked in the present recordings. Actually, the increase in frequency of synaptic inputs in the type D response is a second-order or higher order response that arises from synaptic pathways. How- ever, because all of these responses were recorded from neurons within 'isolated' cerebral ganglion preparations, it is evident that there should be a certain population of photoreeeptor cells within the cerebral ganglion of Ciona savignyi. Light-induced gamete release is a well-known behavior of ascidians (3. 4). However, removal of the whole neural complex in vivo does not affect this behavior in Ciona intestinalis (Tsutsui and Oka, unpubl. data). Similar behav- ior was reported previously in Chelyosoma production ( KM. These results may support the existence of an ascidian photoreceptive system or systems other than the cerebral ganglion. It may be an interesting future study to compare the action spectra of the voltage responses of the light- sensitive neurons in the present cerebral ganglion prepara- tion and those of light-induced spawning behavior, such as that measured by Lambert et al. (3). Such a comparison would test whether similar photoreeeptor processes underlie the control of spawning as well as the voltage responses reported here and would further our understanding of the properties of ascidian photoreceptive systems. The cerebral ganglion of ascidians has been shown to have many neurons immunoreactive for vertebrate neu- ropeptides such as GnRH. bombesin, and neurotensin (11. (composed of closely apposed cerebral ganglion and neural gland). The neural gland of this species is located in the anterior ventral part of the cerebral ganglion (Tsutsui and Oka, unpubl. obs.). and a large ventral surface area of the cerebral ganglion is easily accessible in vitro with microelectrodes. The neural complex was dissected out, and muscular tissue and the thin wall of the blood sinus that cover the ventral side of the cerebral ganglion were gently removed. Then, the isolated neural complex preparation was pinned ventral side up in a recording chamber (volume = 400 jil) lined with Sylgard and was perfused with filtered seawater (1 ml/min). The experiments were carried out at room temperature (18-22 C), and the preparation was viable at least for 6 h under these conditions. The neurons on the ventral side of the posterior part of the cerebral ganglion were recorded intracellularly by sharp microelectrodes (pulled from borosilicate glass of o.d. 1.5 mm with inner filaments and tilled with 2 M KC1: resistance = 40-70 Mil) under a dissecting microscope. Signals were amplified (by MEZ-8300, Nihon Kohden) and digitized at 3-5 kHz and stored digitally (using Axotape software, Axon Instruments). Light stimuli (3.0 x IO J lux) were delivered to the preparations through an optic fiber from a conventional light source of a 150 W halogen bulb (type 6423. Philips) equipped with a heat-absorbing filter. The light stimulus was controlled manually by a solenoid relay switch. The preparation was kept in the dark (<0.1 lux) for at least 1 min between the light stimuli. 28 H. TSUTSUI AND Y. OKA 12). Especially, (unicute GnRHs have been biochemically characteri/.ed. and the distrihu':,ms i>f (iiiRH neurons have been studied recently (11, 13. 14). The recordings of light- sensitive voltage responses in the present study were made mainly from the pi cuor ventral part of the cerebral gan- glion, the region \\here substantial numbers of immunore- active GnRH neurons are distributed in Citma intestimilis (11) and C. savignyi (Tsutsui and Oka, unpubl. data). There- fore, the present results show that light-sensitive neurons and GnRH-secreting neurons are located in an overlapping region, and it may be suggested that GnRH release is controlled by light. This possibility is interesting from the viewpoint of the general neuromodulatory functions of GnRH neurons (15. 16). The ascidian cerebral ganglion preparation may. therefore, provide a chance to test this hypothesis. Acknowledgments We thank the staff of the Education and Research Center of Bio-resources in Onagawa for their supply of ascidians. We also thank Dr. Arikawa for lending us a luminometer and for a helpful discussion. This study was supported by grant-in-aid for Fundamental Scientific Research from the Ministry of Education. Sports. Science and Culture of Japan to Y. Oka. Literature Citi-d 1. Hecht, S. 1918. The photic sensitivitv ul Ci\. Vol. 1 1. F. S. Russel. ed. Academic Press. London. 3. Lambert, C. C., and ('. L. Brandt. 1467. I IK- effect of light on the spawning of Cinna intestinalin. Bid. Hull. 132: 222-22X. 4. Numakunai. T., and /.. Hoshino. 1473. Diologs mi the ascidian. Halocynlhia rorelzi (Drasche). in Mutsu Bay I Diltcicnccs ol spawn- ing lime and e\tcrnal features. Bull. Mar. Bio/. Sin. Astimiishi. 14i4i: 191-197. s l)ill\. P. V. and .1. .1. VYnlkcn. 1973. Studies on (he receptors in dona iniestinalis. IV. The ocellus in the adult. Micron 4: 1 1-29. 6. \Vixilaooit.R.M. 1974. Microfilamenls and the mechanism of light- triggered spemi release in ascidians. De\. Biol 40: 186-195. 7. Woolacott. K. M.. and M. R. Porter. 1977. A synchroni/ed mul- ticellular movement initiated by light and mediated by microtilaments. De\: Biol. 61: 41-57. S K.i|i.ii.i. S., S. Tamotsu. Y. Murita. and I. Nuinakiiiuii. 1991). Retinal isomers in the cerebral ganglion of the uscidian. Halocynthia rorelzi. Invenehr. AY/'/<>,/ /)ri 17:2: 155-158. 9. Gorman. A. L. K.. .1. S. Reynolds, and S. V Barnes. 1971. Pho- toreceptors in primitive chordales: line structure, hyperpolari/ing re- ceptor potentials, and evolution. .SVic/ici' 172: 1052-1054. 10. Hisavi. 1 . L. Jr.. C\ R. Botticelli, and K. L. Hisavv. 1966. A study of the relation of the neural glaiul-ganghonic complex to gonadal development in an ascidian. CVir/mvoimi produclum Simpson. Gen. Comp. Endocrinol. 7: 1-9. 11 Tsutsui. H.. N. Yamamolo. H. llo. and Y. Oka. 1998. GnRH immunoreactive neuronal system in the presumptive ancestral chor- dale. Ciniiii inlestinalis ( Ascidian i. (it-n. Cam/i. Endoi'rinol. 112: 42(1-432. 12 I ritsch. H. A. F,.. S. V. Noorden. and A. (;. K. Pearse. 1982. (iastro-mtestinal and neurohormonal peptides in the alimentary tracl and cerebral complex ot ( innu intestinalis (Ascidiaceae). Cell Tissue AVv 223: 36') -402. ! ; Mackie. (i. O. 1995. 'On the visceral nervous system' of Ciona. J. Mar. Biol. AS.WI . IK. 75: 141-151. 14 Pimell. .1. K. K.. S. M. Ki-ska-SkiiiniT. M. O. Prakash. \\. 11. Fischer. M. Park. J. K. Rivier. A. (i. C 'rain. (i. O. Mackie. and N. M. Sherwood. 1996. lvu> new lorms ol gonadotiopin-releasing hoimone in a protochordate and the evolutionary implications. Proc. Nail. A,, nl. Sci. USA. 43: HUM 10464. 15. Oka, Y., and T. Mutsiishiina. 1993. ( ionadoiropin-releasing hor- mone (CinRHl-immunoieaclive terminal nerve cells have intrinsic rhythmicitv and project widely in the brain. J. M'I/II'MI 13: 2161- 2 1 76. 16. Oka, Y. 1997. The u'onadoliopin releasing hormone iGnRH I neuro- n, il sy stem of tish brain as a model system for the sludy of peptidergic ncuromodulation. I'p. 245 -276 in linRH Neunms: Genes to Behavior. I S Parph.it and ^ S.ikum.i. eds. Brain Shuppan. Tokyo. Reference: Biol. Bull. 198: 29-33. (February 2(1(10) Parasitic Diatoms Inside Antarctic Sponges GIORGIO BAVESTRELLO 1 -*, ATTILIO ARILLO 2 , BARBARA CALCINAI 2 . RICCARDO CATTANEO-VIETTI 2 . CARLO CERRANO 2 , ELDA GAINO 3 , ANTONELLA PENNA 4 . AND MICHELE SARA 2 [stitiito di Science del Mure University di Ancona, Via B recce Blanche. 1-60131 Ancona. Italy; Dipartimento per lo Studio del Territorio c delle sue Risorse Universita di Geneva, Viale Benedetto XV, 5 1-16132 Genova, Italv; and Dipartimento di Biologia animate ed Ecologia Universita di Perugia, Via Elce di Sotto 1-06123 Perugia, ltal\; and 4 lstituto di Biochinu'ca Universita di Urbino. 1-60118. Italv Antarctic sponges may host large populations of plank- tonic and bentliic diatoms. After settling on the sponge. these diatoms enter its body through pinacocytes (1) and form, there, large mono- or paiici-specific assemblages. Yet the total amount of carbohydrates in the invaded sponge tissue is inverselv correlated with that of chlorophyll-^.. We suggest, therefore, that endobiont diatoms utilise the prod- ucts of the metabolism of their host as an energy source. This is the first evidence indicating that an endobiotic autotrophic organism may parasiti-e its animal host. More- over, this unusual symbiotic behavior could be a successful strategy that allows the diatom to survive in darkness. Heterotrophic bacteria, autotrophic cyanobacteria, zoochlorellae, and zooxanthellae are common symbionts in Porifera, where they may actually constitute most of the sponge tissue. Sponges can also harbor fungal popu- lations, the significance of which has been poorly inves- tigated (2). The association of sponges with autotrophic symbionts occurs mainly in tropical (3) and temperate waters (4); it is a fruitful strategy, allowing the sponges to utilize the symbionts as a food source complementary to filter feeding (5, 6). Diatoms in sponge tissues have rarely been recorded (7), but such associations are widespread in Antarctic waters ( 1 ). In fact, most sponges from Terra Nova Bay (Ross Sea, Antarctica) host large populations of both plunktonic and Received 29 June 1999; accepted 5 November 1999. * To whom correspondence should be addressed. E-mail: bavestrellofs 1 popcsi.unian.it benthic diatoms, and these microorganisms are actively taken up by the pinacoderm and thus enter the sponge choanosome ( 1 ). This could be an adaptive supplementation of the food supply in the Antarctic ecosystem, because diatoms produce extracellular polysaccharides (8, 9) that sponges might use as energy substrates. On the other hand, the integrity of the diatom shells, the presence of cytoplasm inside the shells, and the formation of large mono- or pauci-specific assemblages inside a single sponge suggest that diatoms are able to complete their entire biological cycle inside the sponge body. The aim of this study was to investigate the energy source that symbiotic diatoms might provide to Antarctic sponges. To this end, we evaluated the variation in the total carbo- hydrate content of sponges as a function of the biomass of endobiotic diatoms; biomass was estimated from chloro- phyll-o concentration. We examined 39 specimens representing 17 species of sponges (Table 1 ). The material was collected at Terra Nova Bay (Ross Sea), between 100 and 120 m depth, during the XIII Italian Antarctic Expedition (Austral Summer 1997- 1998). The sponges were prepared for ultrastructural study as follows. Immediately after collection, small pieces from each specimen were fixed for 2 h in 2.5% glutaraldehyde in artificial seawater (ASW). The fixed tissues were rinsed, stored in ASW, and then dehydrated in a graded series of ethanol concentrations. Critical point drying was achieved with a CO, Pabish CPD 750 apparatus. Samples were examined with a Philips EM 515 scanning electromicro- scope, and several species of diatoms were observed (Fig. 29 30 G. BAVliSTRKU.O H Al.. I ..I.I. 1 Chlorophyll-a concentration and carbohydrate content for the sponge specimens analyzed Specimens Chlorophyll-a dwaf) Total sugars (mg/g d\val) Artemisina rubulosa 12.07 0.9 is..;.. Axociella nidijicala 27.29 2.5 8.120.2 Dendrilla membranosa 29.13 3.6 5 43 0.4 30.47 2.1 1 1 .40 2. 1 31.65 1.8 7.01 0.6 21.34 1.9 I(v29 2.8 7.31 1 22.56 1.8 9.96 0.9 27.06 1.4 14.43 !.9 9.59 1.1 Ectyodorix ramilobosa - 1.2 19.90 3.1 20 ~n 3.1 6.91 1.8 Gellius sp. 17.05 - 24 7.25 - 2 Gellius sp. 1 37. 10 6.9 6.91 1.5 Haliclona dancoi 16.67 2.4 9.92 ! 2.4 15.05 2.6 21 28 3.2 15.14 0.9 8.10 1.4 21.85 1.9 15.92 1.9 Homaxinella balfuurt'ii\i\ 14.37 3.1 12.73 2.1 Inflatella belli 5.75 1.6 24.30 3.1 2.21 0.8 24.41 3.6 Isodictya conulosa K..4S 1.6 12.29 2.2 13.59 2 9.96 1 Isodictya erinacea 11.86 1.8 15.25 2.8 Microxina benedeni 11.75 3.2 11.21 1.2 Phorhas xlaherrima 18.22 2.9 13. 13 2.8 19.04 4 12.73 4.1 Pseudosuberites antarcticus 44.Mi ' < 5 3.32 0.9 Pseudosuberites nudus 27.79 5.6 10.23 1.4 45.71 5.9 6.23 0.8 9.13 1.7 17.71 2.5 11.88 2.1 12.13 1.3 24. S3 4.8 9.23 1 .9 25.65 4.9 5.93 1.4 19.74 3.6 7.41 2 S 11.54 2.2 10.48 2.3 Suberiles caminatus 32.94 3.8 X.9I 1.7 6.35 + 2.1 21.08 3.4 17.25 5.7 7.11 1.5 Tedania chart nil 16.67 3.6 7.11 0.9 Values are (he mean of three replicates . Si). ilwal. ili\ weisjlii ash l I). Among pennate diatoms. Fragilariopsis curia (Fig. la). Fraftilariopsis sp. (Fig. la. b). Achnanthex sp. (Fig. Id. ami Pseudogomphonema sp. were the most recurrent species; centric diatoms belonging to the genera Porosira (Fig. Ic). Coscinodiscus, and Khi-oxolenia were rarer. For transmis- sion electron micmsu>py. fragments of selected fixed ma- terial were postfixcd in \.(Y7i (KO, in ASW at 4 C'. After repeated rinsing, these fragments were dehydrated in etha- nsiiberires nudus. High \alues were found also in Pseudosuberites antarcticus, Gellius sp. I.. Suberites cuminaius. and Dendrilla membranosa. Strong differences emerged when different specimens of the same species were compared (e.g.. from 7.3 to 31.6 /ig/g in /). membranosa; and from 9.1 to 4?. 7 ju.g/g in P. nudus). This variation suggests that diatom uptake is fortuitous, depend- ing primarily upon exposure of the sponge surface to dia- toms settling upon it. rather than to some species-specific entrapping mechanism. Carbohydrate in the sponge tissues (Table 1 ) ranged between 3.3 mg/g in Pseudosuberites antarcticus and 27.1 mg/g in Dendrilla membranosa. As with the chl--\ ii-tti. and M. Sara. 1994. spuming electron microscope evidence tor diatom uptake by two Antarctic sponges. Polar Bid. 14: 1-4. 2. Sara. M.. G. Ba\estrello. R. Cattanco-Vivtti. and C. Orrano. 1998. Endosymbiosis in sponges: relevance for epigenesis and evolution Symbiosis 25: 5 ~ < \\ ilkmson. C. R. 1980. Cyanobacteria symbiotic in marine sponges. Pp. 553-563 in Endocytobiology, Endosymhio.ti* mul Cell Biology. W. Schwemmler and H. L : . A. Schenk. eds. Walter DC Gruyter. New York. 4 Sara, M. 1971. Ultrastructural aspects of the symbiosis between two species of the genus Aphanocapsa (Cyanophyceae) and Ircinia \'uria- bilis. Mar. Bio/. 11: 214-221. 5. Wilkinson, ('. R. 1983. Nel primary productivity in coral reef sponges. Science 219: 410-412. (>. Arillo, A., G. Bavcstrello. B. Burlando. and M. Sara. 1983. Met abolic integration between symbiotic cyanohacteria and sponges: a possible mechanism. Mar. Bin/. 117: 159-162. 7 Co\. (i.. and \. W. I). I arkuni. 1983. A diatom apparently living in symbiosis with a sponge. Bull. Mar. .Sri. 33: 943-945. S. Myklestad, S.. (>. llolm-llansen. K. M. \aruin. and B. \olcani. 1989. Rale nt it-lease of extracellular ammoacids and carbohydrates 20 30 40 50 from Ihe marine diatom Cliiu'innim iilliiin. J. Plankton Res. 11: 763-773. 9. Mykk-st.id. S. 1995. Release ol extracellular products by phyto- plankton with special emphasis on pohsacchandcs .SVi. Tnhil l-jn-iron. 165: 155-164. Hi. llfllfhusl, .1. A., and .1. I.ein. 1997. Heterotrophic nutrition I'p 169 197 in The Bitilngy of Diatoms. D. Werner, ed. University of California Cress. Berkeley. II. While. A. \\. 1974. Growth of 2 facultatively heterotrophii marine centri< diatoms / /'/,i,,./. Id: 242-300. 12 I.ein. ,|. ('.. and .1. A. Helli-hust. 1976. Heterotrophic nutrition of maniK- peiinau- ihaiom Nilzschio angularis \ r ar. a/finis. Mar. Biol. 36: 313 ^20. 13. I illin.iiiii. I'. 1998. Chagotrophy by a plastidic haptophyte. Prym- nesiutn patellifemm, Ai/tmt \liti<>h. /-(h. !:,,>!. 13: 177 IS6 15 Lcyrand. ('., K. (iraiu-li. and P. Carlsson. 1998. Induced phago- tiophy in the photosynthetic dinoflagellate Heterocapsa iriqueini. \,/ii,ii \h, rob Ecol 15: 65-75. lii Skn>Kiiiird. \. 1996. \li\oiiophy in Fragiliiliiun siihglohosum I Di- noph\ceael: growth and gra/ing responses as functions of light inten- sitj Wai B ol Prog Sei 143: M7-253. 17. I i . INI I'., and M. Kiiileiistii'rna. 1996. Cuowih and decline of a dialoni spun;' hloom. phytoplankton species tomposiiion. lonnation of marine snow and the role of heterotrophic dinoflagellates. J. I'lunk- t,m Ki's. 18: 133-155. I (,cr\ais. !. 1997. Light dependent growth, dark survival and glu- cose uptake by cryptophytes isolated from a freshwater chemocline. / Phycol 33: is 25. I" V/ani. I., and I!. 1. \iilcani. 1974. kulr nl silico ..... diaioin metabolism VI. Active transport of germanic acid in the helerotrophic diatom Nilzfchio tilhti. Arch. Microbiol. 101: 1 S. M harlcy. \V. M., B. B. \\impee. and C. I. Olilman. 1981. lletero- trophic and pliotohclerolrophic utili/alion of lactate by the diatom CylimJmtheca fusifiirmn. Hi. I'ltycul. ./. 16: 423-428. 10 20 30 40 Chlorophyll-a (pg/g dwaf) 5Q 2. Concentration of chl-o as a function of Ihe total annum! ul carbohydrate. Data from all the sponge specimens tested la), from ditu-u -m specimens of Dendrilla membranosa (b), and from Pseudosuberites nu,ln\ (C). Each point represents a sponyc specimen The analyses \\crc tamed out on three samples of sponge tissue in mi eat h specimen. Samples were cylindrical cores of choanosomal tissue. I cm in diametei. 2 cm in height. Cores were longitudinally cut. and the two hall pmiiiiiis weie lesied to evaluate the chl-i; concentration (as an estimate ol diatom population bioma.ss). as well as the total amount ot taibuln.li.iii- ( im- Mibsample was extracted in 90% acetone. The chl-rt was then assayed by speclrolluonm- etry with excitation at 43(1 nm and emission nic.i m >l al i'iiS nm. On ilk- other subsample. the total carbohydrates were assayed: the method was based on the colon metric reaction ol sugais w Hh phenol and sulphuric aeid. and D( + ) glucose was used as the standard (26). siliceous particles, and the diatoms, with their siliceous shells, may thus "deceive" the sponge. In conclusion, our dala widen the ecological role of diatoms which, by exploiting the niche ptmidcd In sponges une ol (he most important Antarctic henthic organisms illusiiate .1 strong functional link between plankton and benthos (25). DIATOMS PARASITIC ON SPONGES 33 21. Tan, C. K., and M. R. Johns. 1996. Screening of diatoms for heterotrophic eicosapentaenoic acid production. J. Appl. Phycol. 8: 59-64. 22. Bavestrello, G., A. Arillo, U. Benatti. C. Cerrano, R. Cattaneo- Vietti, L. Cortesogno, L. Gaggero, M. Giovine, M. Tonetti, and M. Sara. 1995. Quartz dissolution by the sponge Chondrosia reniforinis (Porifera, Demospongiae). Nalnir 378: 374-378. 23 Palmisano, A. C., J. B. SooHoo, D. C. White, G. A. Smith, G. R. Stantun, and R. H. Burckli'. 1985. Shade-adapted benthic diatoms beneath Antarctic sea ice. J. Pliviol. 21: 664-667. 24. Vanucci, S., and M. Innamorati. 1997. Available radiation profiles recorded in the Ross Sea. Pp. 2X1-290 in Rossmi;e /W-95 Data Report I. F. M. Faranda. L. Guglielmo and P. Povero. eds. Lang Arti Grafiche. Genoa. Italy. 25. Marcus, N. H., and F. Boero. 1998. The importance of benthic- pelagic coupling and the forgotten role of the life cycles in coastal aquatic systems. Linmol. Oceanogr. 43: 763-768. 26 Dubois, M., K. A. Gilles, J. K. Hamilton, P. A. Rebers, and F. Smith. 1956. Colorimetric method of determination of sugars and related substances. Anal. Clu'iii. 28: 350-356. Reference: fli.i/. Bull. 198: 34-49. ll-ebruarv 2(XX)> Maintaining the Line of Defense: Regeneration of Cuvierian Tubules in the Sea Cucumber Holothuria forskali (Echinodermata, Holothuroidea) DIDIER VANDENSPIEGEL 1 . MICHEL JANGOUX 1 ' 2 , AND PATRICK FLAMMANG i Laboratoire de Biologic marine. Univcrsite de Mons-Hainaiit. 6 Avenue du Champ de />' ~IHHI M,,,i\. Hel^iuin: and ^Laboratoire tie Hii>l,>ic marine iCI' lhO/15). I ni\er\iie l.ihre tie Kntxclles. .^tl Avenue r. I). Roosevelt. H-IO.V) Hrn.\elle\. llclgium Abstract. When irritated, individuals of the sea cucumber Holothuria forskali expel a few Cuvierian tubules which lengthen, instant!) become sticks, and rapidls iinmohili/e most organisms with which thev come into contact. After expulsion, the lost tubules are readily regenerated. When onls a few tubules have been expelled, there is often a latent period before the regeneration starts. In contrast, when many tubules have been expelled, the regenerative process starts immediately but proceeds in successive waxes of 10 to 30 tubules that begin to regenerate at 10-das intervals. However, in all cases, the complete regeneration of a given tubule takes about 5 weeks and mas be divided into three successive phases: an initial repair phase including the overall 48-h post-auioloins period, a true regenerative phase taking about 4 weeks to complete, and a growth phase of about one more week. Initial regeneration events occur by epimorphosis. cell proliferation being essential lo the regen- erative process, whereas late events occur mainls hs mor- phallaxis. with migration of the newly differentiated cells. The mesothelium is the tissue layer in which cell prolifer- ation is the most precocious and the most important, involv- ing both pcnloncocslcs and imdiliercnliated cells (which seem to be dedifferentiated peritoneocytes). As regeneration proceeds, the percentage of undifferentiated cells regularls decreases in parallel with the ditlcicntiation ol granular (adhesive-secreting i cells and myocytes. The myocytes then separate oil IKHM i| u - mcsothclmm and ungrate within the Received I June i ited 22 Novembei I' 11 ' 1 ' * Present address: Rojal MUM-HIM l,.i (Vnlral Africa, liucrlchrate Sec- tion. B-30HO Tervurcn. Belgium. t To whom correspondence should he addressed. I in.ul Patrick. I lammang@umbjc.be connective tissue laser. Three tvpes of pseudopodial cells tol low one another in the tubule connective tissue during regeneration. TV pc I cells have all the characteristics of echinoderm phagocvtcs and mas have a libroclastic func- tion, cleaning the connective tissue compartment before new collagen svnthesis starts. Type 2 cells are rather undif- ferentiated and divide actisels. The presence of type 3 cells is closelv associated with the appearance of collagen libers. and it is suggested that they have a lihroblastic function. In the inner epithelium, cells also divide actively, hut onls those in which spherules have not vet differentiated in the basal intraconneclive processes. It appears, therefore, that in the three tissue lasers of the tubules, regeneration proceeds bs cell dedifferentiation, then proliferation, and linalls by dilferentiation. Cuvierian tubules thus constitute a very ef- licient defensive mechanism: (heir large number, sparing use. and particular regeneration dynamics make them an almost inexhaustible line of defense maintained at limited energs cost. Introduction Several species of holothuroids (sea cucumbers), all be- longing to the order Aspidochirotida. possess a very spe- ciali/ed defensive system: the so-called Cusierian tubules il lammang. 19%). This system consists of several hundred tubules whose proximal ends allach lo I he basal part of the lei! respiratory tree and whose distal, blind ends lloat freely in (he coelomic cavity. When irritated, the animal curses us aboral end toward I he irritating object and undergoes a general contraction. The anus opens, the wall of the cloaca tears, and the free ends of the Cuvierian tubules, together with coelomic lluid. are expelled through ihe lear and the 34 HOLOTHUROID CUVIERIAN TUBULE REGENERATION 35 anus. The emitted tubules lengthen, instantly become sticky, and rapidly immobilize most organisms with which they come into contact (VandenSpiegel and Jangoux. 1987). Finally, the expelled tubules autotomize at their attachment point on the left respiratory tree and are left behind as the holothuroid crawls away (Miiller et nl., 1970; VandenSpie- gel and Jangoux. 1987). As only a portion of the tubules are emitted at one time, the total number may suffice for several responses. After expulsion, the lost tubules are readily re- generated (Hyman. 1955). The fine structure of the Cuvierian tubules in the species Holothuria forskali has been described by VandenSpiegel and Jangoux (1987). Quiescent Cuvierian tubules consist of an outer mesothelium. an inner epithelium, and between them, a thick connective tissue layer that includes muscle fibers. The mesothelium is a pseudostratified epithelium made up of two cell layers, namely an apical layer of ciliated peritoneocytes and a basal layer of granular cells filled with densely packed lipoproteic granules. Together, these layers form conspicuous transverse folds that pene- trate the underlying connective tissue. Circular and longi- tudinal myocytes occur under the tips of the mesothelial folds. These cells divide the connective tissue into a thin outer layer (where collagen fibers are directed perpendicu- larly to the tubule long axis) and a thick inner layer (where collagen fibers are arranged in helixes parallel to the tubular long axis). Nerve processes are associated with the longi- tudinal muscles; they also occur between the spirally ar- ranged collagen fibers. The tubule lumen is narrow with a highly convoluted limiting epithelium. Epithelial cells are not ciliated; they have basal processes containing enlarged spher- ules whose contents include mucosubstances and proteins. The histological organization of Cuvierian tubules, which is quite different from that of the left respiratory tree that bears them, appears to be functionally important for their correct operation (VandenSpiegel and Jangoux. 1987). However, as pointed out by Smiley ( 1994). the histogenesis of these intriguing organs is virtually unknown. Similarly, nothing is known about the renewal of the Cuvierian tubule stock after expulsion and, consequently, on the way holo- thuroids maintain their line of defense. We have thus inves- tigated the process of Cuvierian tubule regeneration in the holothuroid Holothuria forskali. Our aims were (i) to study the dynamics of the regenerative process and to describe macroscopically the different regeneration stages; (ii) to detail the installation of the different tissue layers, using electron microscopy; and (iii) to determine, through immu- nohistochemistry and autoradiography, the origin of the cells involved in tubule regeneration. Materials and Methods Individuals of Holothuria forskali (Delle Chiaje, 1823) were collected at a depth of 20 to 30 m by scuba diving in the bay of Morlaix (Brittany, France). They were trans- ported to the Marine Biology Laboratory of the University of Mons-Hainaut. where they were kept in a marine aquar- ium with closed circulation (13C, 339ft salinity). Induction of tubule expulsion and sampling of regenerating tubules The expulsion of Cuvierian tubules was induced mechan- ically by pinching the dorsal integument of the holothuroids with forceps. Two sets of regenerating animals were inves- tigated: the first set comprised 15 individuals that had been stimulated only once and had expelled about 15 tubules (gentle stimulation); the second set consisted of 25 individ- uals that had been stimulated repeatedly to induce the ex- pulsion of about 300 tubules (strong stimulation). After tubule expulsion, all the animals were returned to the aquar- ium and dissected at various times to provide samples for observation macroscopically and with transmission electron microscopy (TEM). Tubules in the process of regeneration were dissected from each set of cucumbers every day for 6 days, and then weekly for 8 weeks. Before dissection, the animals were anesthetized for 1 h in 0.2% propylene phe- noxytol (Nipa laboratories. UK; see Hill and Reinschmidt. 1976) in seawater. For the immunohistochemical and auto- radiographical studies of Cuvierian tubule regeneration, 10 of the strongly stimulated sea cucumbers were put aside in a separate aquarium for 1 month. Light and transmission electron microscopy Regenerating tubules were fixed for 3 h at 4C in 3% glutaraldehyde in cacodylate buffer (0.1 M, pH 7.8; adjusted to 1030 mOsin with sodium chloride), then rinsed in caco- dylate buffer and post-fixed for 1 h in l'7r osmium tetroxide in the same buffer. After a final wash in buffer, the podia were dehydrated in graded ethanol and embedded in Spurr. Semithin sections (1 jum in thickness) were cut on a Reichert Om U2 ultramicrotome equipped with a glass knife and then stained according to the method of Humphrey and Pittman (1974). Ultrathin sections (40-70 nm) were cut with a Leica UCT ultramicrotome equipped with a diamond knife, stained with uranyl acetate and lead citrate, and observed with a Zeiss EM 10 transmission electron micro- scope. Immunohistochemistry The BrdU/anti-BrdU method was used to study DNA synthesis in the regenerating tubules. After one month of regeneration, five of the strongly stimulated holothuroids were injected intracoelomically with 2 ml of a 2% solution of BrdU (5-bromo-2-deoxyuridine; Aldrich Chemie) in fil- tered seawater (0.45 jiim; Millipore). One hour after injec- tion, the individuals were dissected and their regenerating 36 D. VANDKNSPIEGEL ET AL tubules fixed for 6 h in Allcn-Bouin's fluid (Gabe. 1968). They were then rinsed thoroughly in running tap water. dehydrated in graded ethanol. embedded in Paraplast. and sectioned (5 /umi. BrdU-containing DNA in the nuclei of proliferating cells was detected by the ininiunogold-silver staining (K.SSi method (Hacker ft til.. 19K5: adapted lor echinodenn tissues in VandenSpiegel ct til.. 1491). The sections were treated with trypsin (0.1% trypsin in 0.01 \l sodium barbital. 0.15 M NaCl, pH 8.0) for 50 min at 37C. followed by acid hydrolysis in 3 M HC1 at 60C for 30 min. This treatment denatures the DNA and unmasks hidden epitopes. After rinsing in distilled water, the sections were preincubated for 20 min in 5*7r normal goat serum in 0.01 A/ sodium barbital. 0.15 M NaCl. pH 7.4 i \eronal buffer- NGS). then incubated overnight at 4 C with mouse anti- BrdU monoclonal antibodies (DAKO A/S) diluted 1:30 in veronal buffer-NGS. After several washes in phosphate- buffered saline (PBS), the sections were incubated for 1 h at room temperature in goat anti-mouse Lmmunoglobulins con- jugated to 5 nm gold panicles (Auroprobe LM: Amersham) diluted 1:40 in PBS-NGS. Gold particles bound to the immunocomplexes were visuali/ed by a silver precipitation reaction, using a commercial kit (Intense SE TMI1: Amer- sham). The reaction was stopped with distilled water and the sections were counterstained \\ith hemalun and eosin. A utorodiography Cell proliferation was also monitored using autoradiog- raphy. Tritiated thymidine ( 150 /LiCi of a I mCi/ml aqueous solution of 5'- 3 HT; Amersham) was mixed with 1 ml of tillered seawater and injected, through the holothuroid in- tegument, into the coelomic cavity. This injection was per- formed on five strongly stimulated individuals that had started regenerating Cuvierian tubules one month earlier. One hour after injection, the holothuroids were dissected, and the regenerating tubules were fixed and embedded as tor TEM (see above). Semithin sections I l-^m thick) were collected on glass slides and stored in distilled water. All subsequent manipulations took place in a darkroom. The slides were dipped in photographic emulsion (Nuclear re- search emulsion L5; 1 1 ford) diluted 1:1 in distilled water at 45C. After air-drying, the coated slides were stored in light-proof boxes at 4C for 10 days. Exposed slides were developed for 4 min at room temperature in Kodak hi 1 ' diluted \:2 in distilled water, transferred to a stop-bath (2% acetic acid) for 30 s, followed by fixation (107r Acidolixi for 5 min. After rinsing in running tap water, the autoradm graphed sections were stained with toluidinc blue. i "iinting tind Identification of proliferating cells Labeled nuclei were counted in sections of regenerating tubules that were processed either by immunohistochcmis- try or by autoradiography. For each regeneration stage, a labeling index (LI) was determined in the three main tissue layers of the tubules (i.e., the mesothelium. the connective tissue layer, and the inner epithelium). This index represents the percentage of labeled nuclei (LN) compared to the total number of counted nuclei (IN i: it is calculated according to the formula: LI = (LN/TN) X 100. For each stage, four counts, each comprising a maximum of 100 nuclei, were done on sections coming from tour different regenerating tubules from a single holothuroid. To identify proliferating cells we used the procedure of Laurent ct til. (1988). which relies on the fact that tissues labeled with *HT and embedded in Spurr resin can he observed by both light and electron microscopy. For each regeneration stage, we used three to four ultrathin sections directly followed by one semithin section. The former were observed in TEM. whereas the latter was processed for autoradiography (see above). In most cases, the shape of the labeled cells, as it appeared on the semithin section, was characteristic enough to allow their locali/ation on the ul- trathin section. It was therefore possible to identify the cell types involved in DNA synthesis. Results Macroscopic observations Regeneration of the Cuvierian tubules in Holothuria /<>/ \Lili was observed after two types of stimulations: gentle stimulations inducing the expulsion of about 15 tubules and strong stimulations inducing the expulsion of about 300 tubules. After a gentle stimulation, tubule expulsion is not always followed by immediate regeneration of new tubules: rather, a latent period of up to 20 days was observed In addition, the beginning of the regeneration could not he detected without sacrificing the animal. Therefore, a regen- eration that followed a gentle stimulation was difficult to sluily. However, data collected from the staggered dissec- tions have allowed us to estimate that the duration of the regeneration is at least 4 weeks, and that all of the autoto- mi/ed tubules regenerate synchronously. Alter a strong stimulation, on the other hand, regeneration starts immedi- ately, although only fora limited number of tubules. Indeed, the regeneration proceeds by groups of 10 to 30 tubules: the beginning of the regeneration in one group being separated from (hat of the following one by 5 to 10 days. Therefore. about one month after ihc induction ol a massive tubule autotomy. holothuroids have tubules at several stages ol regeneration (Fig. I ). Ihe observation of Cuvierian tubule regeneration after a strong stimulation has allowed us to describe live succes- sive stages in the expulsion-autotomy-regeneration cycle (Fig. 2i : Stage I corresponds to the healing of the damaged tissues I ol lowed by the appearance. 2 to ft days after autotomy. ol HOLOTHUROID CUVIERIAN TUBULE REGENERATION 37 Figure I. General aspect of regenerating Cuvierian tubules in Halo- thuria forskali 3 weeks after a massive tubule expulsion, showing tubules at different stages of regeneration (bar = 0.5 cm). C. cloaca; FT. fully developed tubule; I, intestine; RT. left respiratory tree; S1-S4. stages 1-4. a thickening of the wall of the basal part of the left respi- ratory tree, at the place where the expelled tubules were attached (SI; Figs. 1 and 2). Stage 2 corresponds to the formation. 8 to 10 days after autotomy, of a translucent spherical structure measuring from 0.1 to 0.4 mm in diameter (S2: Figs. 1 and 2). Stage 3 corresponds to the development. 10 to 14 days after autotomy, of a translucent club-shaped structure rang- ing from 1 to 3 mm in length for a maximum diameter of 0.3 mm (S3; Figs. 1 and 2). Stage 4 corresponds to the transformation of the regen- erating structure, 14 to 21 days after autotomy, into a small pale-pink tubule measuring from 4 to 12 mm in length (S4; Figs. 1 and 2). From this stage on, the shape of the regen- erating tubule remains constant, only the length and diam- eter of the tubule increase. Stage 5 corresponds to the differentiation, 21 to 28 days after autotomy, of a functional tubule that possesses a short basal peduncle and measures from 14 to 22 mm long for a diameter of about 0.65 mm (S5: Fig. 2). Although they are only half the size of a fully developed tubule, regenerating tubules at stage 5 are able to lengthen and become sticky. They reach their full size 28 to 35 days after autotomy. Ultrastnicture of regenerating tubules Immediately after their expulsion and lengthening, Cu- vierian tubules are autotomized at the narrow peduncle that attaches them to the left respiratory tree. The resulting wound is relatively limited and is rapidly closed by migra- tion of mesothelial cells on the coelom side and epithelial cells on the side of the respiratory tree lumen. At the same time, the connective tissue is invaded by very active phago- cytes that remove cell debris from the tissue. These large cells have many cytoplasmic inclusions, such as phago- somes, secondary lysosomes, and residual bodies. Stage 1. Regeneration sennit stricto takes place right after wound healing and starts with the formation of a small mesothelial bud that appears as a local thickening of the wall of the basal part of the left respiratory tree. This bud consists of a mass of cells forming a pseudostratitied me- sothelium in which an outer layer of ciliated adluminal cells (peritoneocytes) covers a few layers of undifferentiated cells (Fig. 3). Peritoneocytes are tack-shaped cells whose apical part lines the coelomic cavity. They are bound to- Expulsion and 01 autotomy s S2 Figure 2. Chronology of the regeneration stages of Cuvierian tubules in Holothuria forskali. Open arrows point to the regenerating tubules. FT, fully-developed tubules; I, intestine; RT, left respiratory tree; T, Cuvierian tubules; S1-S5. stages 1-5. - 3 ^fcft - P . C -'"** vgy $ <~^^ && 2$ t& ^^ ( D. VANDENSPIEGEL ET AL PC M "^ wl Figures 3-9. Ili'luiliiiini tim). Figure 4. Longitudinal section through the wall of a stage 2 regenerating tubule (bar = 5 /urni Figure 5. Detailed view of the basal pan of the mcsothclium showing the myocytes and the nerve plexus ih.n J 1 /tm). Arrows indicate bundles nl tmnMla Figiircfi. Intraconnective spherulocyte (stage 2; bar - 2finu. Figure 7. [ntraconnective type 1 i iiiliipodial cell (stage 2; bar = 4 /nin Figure K. I vtailod view of the cytoplasm of a type I pseudopodial cell i stage 2; bar 0.5 jun). Figure 9. l.oiiL'itudiual section through the inner epithelium (stage 2; bar = gcthcr by an apical /unula ailhaercns. Each periluiK-ocylc bears a single ciliuni (about 10 /j.m long) and a few scattered microvilli (Fig. 3). The apical pan of the cell houses the nucleus, a juxtanuclear Golgi complex, a well-developed rough endoplasmic reticulinn (RER). numerous free ribo- somes. some lipid granules, and many mitochondria. Large HOLOTHUROID CUVIERIAN TUBULE REGENERATION 39 secondary lysosomes ( 1 to 1 .5 fj,m in diameter) also occur in the apical cytoplasm. The basal part of the cell, which is thin and elongated, traverses the undifferentiated cell layers be- fore contacting the basal lamina through a hemidesmosome- like structure. Undifferentiated cells are cube-shaped and have a high nuclear-cytoplasmic ratio (Fig. 3). However, although reduced, the perinuclear cytoplasm encloses a well-developed Golgi apparatus and RER, many mitochon- dria and free ribosomes, lipid granules, and numerous sec- ondary lysosomes (Fig. 3). Undifferentiated cells are con- nected to the basal lamina through hemidesmosome-like structures, but they are neither bound together nor bound to the peritoneocytes. Stage 2. At the end of this stage, the regenerating struc- ture already has the three tissue layers characteristic of a non-regenerating Cuvierian tubule, i.e., a mesothelium. a connective tissue layer, and an inner epithelium (Figs. 4, 10). The general organization of the mesothelium is similar to the one observed in stage 1 except that some myocytes occur among the undifferentiated cells (Figs. 5. 10). Myo- cytes are cube-shaped. They have a large nucleus and en- close one bundle of myofilaments in their cytoplasm (Fig. 5). which also contains a well-developed RER and numer- ous mitochondria. No junction connecting the myocytes to the other mesothelial cells was observed. However, each myocyte is anchored to the basal lamina by a hemidesmo- some-like structure. A discrete nerve plexus is associated with the mesothelium of stage 2 regenerating tubules. It consists of nerve processes, preferentially located close to the myocytes (Figs. 5. 10). MV Figure 10. Holothuria forskali. Reconstruction of a longitudinal sec- tion through the wall of a stage 2 regenerating Cuvierian tubule (not to scale). BL. basal lamina; C, cilium; CL, connective tissue layer; IE, internal epithelium; G, Golgi apparatus; M. mesothelium; MC. myocyte; MI. mitochondrion; MV. microvillus; NP, nerve plexus; PI. type 1 pseudo- podial cell; PC, peritoneocyte; SC, spherulocyte, SL, secondary lysosome; UC. undifferentiated cell. Encompassed between the basal lamina of the mesothe- lium and that of the inner epithelium, the thin connective tissue layer (about 7.5 jam in thickness) is poor in collagen fibers, but contains many mesenchymatous cells, both spherulocytes and type 1 pseudopodial cells (Figs. 6, 7, 10). Spherulocytes are ovoid to spherical cells, 5 to 10 jam in diameter. Their cytoplasm is packed with spherules that average 3 /urn in diameter and enclose an electron-lucent material (Fig. 6). Type I pseudopodial cells possess a cen- tral, nucleus-containing cell body from which two to three thin cell processes radiate (Figs. 4, 7). Their cytoplasm encloses one or two juxtanuclear Golgi stacks, a well- developed RER, a few lysosomal bodies, many mitochon- dria, and numerous clear vesicles of various sizes. In some micrographs, large secondary lysosomes enclosing partly decomposed collagen fibrils were observed (Fig. 8). The inner epithelium delimits a quite large lumen. It consists of only one cell type that strongly resembles the epithelial cells of the respiratory tree at the level where Cuvierian tubules are inserted. These cells are non-ciliated cells that bear a few apical microvilli (Figs. 4, 9, 10). They are connected together through junctional complexes con- sisting of an apical zonula adherens and a subapical septate desmosome, but they do not attach to the basal lamina. Epithelial cells are flat (2.4 /im in the nucleus-containing thickest part). Their cytoplasm contains a Golgi apparatus, a few scattered mitochondria, and some lipid granules. Stage 3. The mesothelium is pseudostratified and still comprises peritoneocytes, undifferentiated cells, and myo- cytes (Figs. 11. 12, 16). Peritoneocytes and undifferentiated cells are similar to those observed in stages 1 and 2. In contrast, myocytes have clearly increased in number and can be divided into two sets according to the orientation of their bundle of myofilaments, either circular or longitudinal. Circular myocytes are always located beneath longitudinal myocytes. They possess a basal cytoplasmic process that penetrates the underlying connective tissue layer and en- closes the bundle of myofilaments (Figs. 12, 13). The nerve plexus is more developed than in the previous stage and still preferentially associated with the myocytes (Figs. 12, 16). The connective tissue layer has increased in thickness (to about 30 /nm) and, at this stage, makes up three-quarters of the total thickness of the regenerating tubule wall. As in the previous stage, the connective tissue layer is poor in colla- gen fibers and contains numerous mesenchymatous cells. Spherulocytes, however, are no longer present; only pseu- dopodial cells remain. These cells are of two types: type 1 cells, identical to those observed at the previous stage, and type 2 cells; the latter being, by far, the more abundant. Type 2 pseudopodial cells have a particular distribution: they are mainly located just below the basal lamina of the mesothelium (Figs. 13, 16). They have a teardrop shape, the tapered part forming a single process that always faces the basal lamina. These cell processes were sometimes ob- 40 I) \ \NDI \SPII i ,11 / / W Figures 11-15. Hulmhurui lurxkuli. IMliastiucture ul stage 3 regenerating Cuvierian tubules. BL. basal lamina: BP. epithelial cell hasal process; CE, centriolc. CL. connccli\c tissue layer: CM. circular myocyte; IE. internal epithelium; G. Golgi apparatus: I., lumen: LM. longitudinal myocyte; MC. myocyte; N, nucleus; NP, nerve plexus: ]'2. l\pe 2 pseudopodial cell: PC. pciitoin'oc\te: SI., secondary lysosumc. TC. unditterenlialed cell. Figure 11. Longitudinal section through the mesothelium (bar 3 fimi Figure 12. Detailed view of ihe myocytes (bar = 3 /nm). Figure 13. Intraconnective type 2 pseudopodial cells (bar = 3 jtm). Figure 14. Detailed \ieu ul a t\pe 2 pseudopodial cell (bar = 0.5 /J.m). Figure 15. Transverse section through the inner epithelium (bar = 3 jxm). served contacting the intraconnective hasal processes of the circular myocytes (Fig. 13). Type 2 pseudopodial cells have a high nuclear-cytoplasmic ratio. Their reduced perinuclear cUoplasm encloses a Golgi apparatus as well as a pair of centrioles (Fig. 14). The inner epithelium is only slightly different from that seen in stage 2. The epithelial cells have a more irregular shape and develop two to three hasal processes penetrating the underlying connective tissue layer (Figs. 15. 16). These processes are lined hy the hasal lamina hut. as previously. there is no attachment structure hetween the epithelial cells and this hasal lamina. Slum' 4. At this si. i .eneration. the mesothelium develops transversal folds that penetrate the connective tis- sue layer (Figs. 17, 20). Three cell types may he recogni/.ed in ihe mesothelium: peritoneocytes, granular cells, and mxocytes. The undifferentiated cells are no longer present. IVnioneocytes form the coelomic lining and always cover the granular cells. They are tuck-shaped \\nli a llatlened apical part and a thin elongated hasal part. They are now attached to each other through junctional complexes con- sisting of an apical /onulu adherens and a suhupical septate desmosome. and to ihe hasal lamina through hemidesmo- some-like structures. The nucleus is located in ihe apical part of ihe cell where the cytoplasm contains a Golgi appa- ratus and a few mitochondria (Fig. 17). Granular cells are very flattened and lorm a suhapical cell layer that is accor- dion pleated perpendicularly to the long axis of the Uihule. each pleat penetrating the underlying connective tissue. The cytoplasm ol these cells contains many electron-dense gran- HOLOTHUROID CUVIER1AN TUBULE REGENERATION 41 Figure 16. Hololhuria forskali. Reconstruction of a longitudinal sec- tion through the wall of a stage 3 regenerating Cuvierian tuhule (not to scale). BL. basal lamina: BP, epithelial cell basal process; C, cilium; CL. connective tissue layer; IE, internal epithelium; M. mesothelium; MC, myocyte; ME, bundle of myofilaments; MV. mierovillus; NP. nerve plexus; P2. type 2 pseudopodial cell: PC. peritoneocyte: UC. undilt'erentiated cell. ules that range from 0.1 to 0.3 /urn in diameter and often occur in the vicinity of a large Golgi apparatus and of a well-developed RER (Fig. 17). Many of the cisternae of the RER are distended and filled with an amorphous material. The nucleus is egg-shaped and usually located in the lower- most part of the cell (i.e., at the base of the pleats). No cellular junctions were observed between the granular cells or between the granular cells and peritoneocytes. However, the granular cells attach to the basal lamina through a hemidesmosome-like structure. Myocytes are located either in the most basal part of the mesothelium (i.e.. at the base of the folds), or here and there within the connective tissue layer, where they form an intraconnective muscle layer (Figs. 17, 20). The circular myocytes are always basal to the longitudinal ones. A single basal lamina lines both types of myocytes as well as the rest of the mesothelium. A well- developed nerve plexus is present and is mostly associated with the longitudinal myocytes, whether they are intrame- sothelial or within the connective tissue (Figs. 17, 20). The thick connective tissue layer is, from place to place, divided into an outer area and an inner area by the presence of the intraconnective muscle layer. The outer area is thin (about 12 /urn thick) and encloses very few collagen fibers and cells. The inner area, on the other hand, is much thicker (about 120 jum thick): it contains many collagen fibers, which show no tridimensional organization, and numerous pseudopodial cells (Fig. 20). Most of the latter are type 3 cells, although a few type 2 cells may still be observed. Type 3 pseudopodial cells possess a central nucleus-con- taining cell body from which one to three cell processes radiate (Fig. 18). Their characteristic feature is an extensive RER with some very distended cisternae filled with an electron-lucent amorphous material (Fig. 18). Their cyto- plasm also contains numerous mitochondria and a well- developed Golgi apparatus. On some micrographs, the Golgi stacks show cisternae ending with spheroidal disten- tions that enclose a fuzzy material of low electron density (Fig. 19). The inner epithelium is similar to the one described in stage 3 (Fig. 20). Stage 5. At this stage, the ultrastructure of the regener- ating tubule is very similar to that of a non-regenerating quiescent Cuvierian tubule (or fully developed tubule); the tubules in these two stages differ mostly in their size. The mesothelium consists of peritoneocytes that cover elongated granular cells. Together they form the conspicu- ous transversal folds that are characteristic of the nonregen- erating tubule mesothelium (Figs. 21. 25). The muscle layer is now completely separated from the mesothelium and is located in the outer part of the connective layer, just under the tips of the mesothelial folds (Figs. 21, 25). This layer comprises outer longitudinal and inner circular myocytes, and it divides the connective tissue layer into a thin outer area and a thick inner area. Numerous nerve processes are associated with the longitudinal myocytes (Figs. 21, 25). An additional cell type, vacuolar cells, has appeared at this stage (Figs. 21, 25). These cells, which are not numerous, are directly connected neither to the mesothelium nor to the muscle layer, but are scattered within the connective tissue outer area, occurring singly or in small clusters of two or three cells. Vacuolar cells are limited by a basal lamina. Their cytoplasm is packed with vacuoles about 0.6 ju,m in diameter that enclose an electron-lucent heterogeneous ma- terial (Fig. 2 1 ). The connective tissue layer still contains numerous type 3 pseudopodial cells that are widely distributed from the mesothelium to the inner epithelium (Figs. 22-25). In its inner area, it encloses densely packed collagen fibers that start to be organized in helixes parallel to the long axis of the tubule. Intermingled with the collagen fibers are slender neurosecretory-like cell processes whose cytoplasm is filled 42 D. VANDUNSPIEGEL ET AL. 17' PC . Figures 17-19. Hulntlniiiii lt\kJ 1 1 II I'H' The distribution and abundance of proliferating cells at each regeneration stage were investigated In immunoliistochemis- try after BrdU incorporation, and by attloradiography after '] IT incorporation. The intensity and pattern of cellular labeling were identical with both methods. The results show that, in the four lirst stages of regeneration, labeled nuclei are unilormh distributed in the three tissue layers of the regenerating tubule: /.<'.. the mesothelium. the connective tissue layer, and the inner epithelium (Figs. 2d-2 1 )). There is no evidence of a well- delined cell proliferation site. Labeling indices indicated that, in the three tissue layers, DNA-synthesi/ing cells are most abundant at the beginning of the regenerative process, their number then decreasing regularly in the successive regenera- tion stages (Fig. 26). In the connective tissue layer, however, the labeling index stays muxinUim until stage 3 and then regularly decreases (Fig. 26). At stage 5. very few labeled cells are observed in the three tissue layers, although the regenerat- ing tubule measures only half of its linal si/e. Labeled cells were never observed in a non regenerating Cuvierian tubule. Proliferating cells were identified by observation of con- secutive semithm and uliraihin sections, the former being treated by autoradiography after *HT incorporation. In the HOLOTHUROID CUVIERIAN TUBULE REGENERATION 43 MV Figure 2(1. Holothuriu frrxkuli. Reconstruction of a longitudinal sec- tion through the wall of a stage 4 regenerating Cuvierkm tubule (not to scale). BL. basal lamina; BP. epithelial cell basal process; C, cilium: CL. connective tissue layer; CM. circular myocyte; IE. internal epithelium; GC. granular cell; LM, longitudinal myocyte; M. mesothelium; MV. microvil- lus; NP, nerve plexus; P3. type 3 pseudopodial cell; PC. peritoneocyte: SG, secretory granule. mesothelium. proliferating cells are present from regenera- tion stage 1 to stage 4. In all these stages, labeled nuclei were always observed in undifferentiated cells and perito- neocytes (Figs. 30. 31 ). No labeling was detected in myo- cytes and granular cells, whatever the regeneration stage considered. Labeled cells were observed in the connective tissue layer until the regenerating tubule becomes a small functional tubule (i.e., end of stage 5). These cells were always pseudopodial cells (Figs. 30, 31) and never spheru- locytes. Among the three types of pseudopodial cells, la- beled nuclei were observed principally, but not exclusively, in type 2 cells (Figs. 29-31 ). As for the inner epithelium, the labeling was always detected in epithelial cells that had not yet developed piles of spherules in their intraconnective basal processes. Discussion Echinoderms in general, and holothuroids in particular, exhibit a remarkable ability to regenerate a missing part of the body (Hyman. 1955). Regeneration in echinoderms gen- erally occurs after evisceration, autotomy, or fission (Emson and Wilkie, 1980). In sea cucumbers, the best studied re- generative model is the intestine regeneration after eviscer- ation (Dawbin, 1949; Mosher. 1956; Bai. 1971; Tracey, 1972; Garcia-Arraras et al., 1998). In this model, formation of the new gut may be subdivided into three successive phases: initial repair, true regeneration, and growth. The same phases were observed in the replacement process of autotomized Cuvierian tubules in Holothuria forskali. Phases of Cuvierkin tubule rc^cncnition The repair phase, which includes the overall 48-h post- autotomy period, comprises wound closure and histolysis of the damaged tissues. During this phase, no clotting of coe- lomocytes was observed, contrary to what usually occurs during repair of injured tissues in echinoderms (Smith, 1981). This lack of coelomocyte plugging may reflect the rapid closure of the relatively limited wound by the con- traction of the circular muscles of the tubule peduncle. Wound closure is followed by re-epithelialization on both the coelomic and luminal sides of the basal part of the left respiratory tree. Phagocytosis of the damaged tissues then takes place within the connective tissue layer. This is the role of the many large and active phagocytic cells observed during this phase. Such cells have been described at wound sites in the five echinoderm classes (Gibson and Burke, 1983; Mladenov et al.. 1989; Candia Carnevali et al.. 1993; Dubois and Ghyoot, 1995). These cells are generally consid- ered to be derived from coelomic amoebocytes that migrate to the wound site (Smith, 197 la; Dubois and Ghyoot, 1995). The regenerative process seusit stricto starts right after wound healing and takes 3 to 4 weeks to complete. Five stages of regeneration may be recognized. Stage 1 can be defined by the appearance of a thickening of the mesothe- lium at the site where autotomy took place. This thickening will be the starting point of the new tubule formation. Stage 2 corresponds to the initial appearance of a lumen within the regenerating tubule. The tubule wall therefore acquires its characteristic trilayered structure, which consists of an outer mesothelium. a connective tissue layer, and an inner epithe- lium. Stage 3 involves a tissue layer proportioning in which the connective tissue layer grows to become the thickest of the three tissue layers, as in non-regenerating Cuvierian tubules. Stage 4 is characterized by the acquisition of the typical tubule shape. Finally, stage 5 is reached when the regenerating tubule becomes functional. Regeneration is followed by a growth phase. Indeed, the functional tubule formed at the end of the last regeneration stage is only half the size of a non-regenerating Cuvierian tubule. It thus continues to grow for about one more week until it reaches its final size. 44 D. VANDENSPIEGEL T AL \ !* CM Kiuurt's 21-24. llnlntlmriu forskali. UltrastructUTC ol' stage S icgcnciating Cuvierian tubules. HI-', bundle of collagen; BL. basal lamina: BP. epithelial cell basal process; CM. circular myocyte; IE. internal epithelium: Of, granular cell: L. lumen: LM. longitudinal myocyle: NC. neurosecretory-like cell; OA. outer area of the connective tissue layer; P3. type 3 pseudopodial cell: PC', periloneocyle; RF.R. rough endoplasmic reticulum; S, rule; SO, secretory granule: VC. vacuolar cell. Figure 21. Longitudinal section through the mesothe- lium and the muscle layer (bar = 3 fim). Figure 22. Longitudinal section through the inner area of the connective tissue layei iii.n 4/im). Kiuurc 23. Longitudinal section through the inner epithelium (bar = 3 iun\. Kinurr 24. Detailed \ lew ol .in MIII.IOMIIK-I ti\c h.is.il piocess ..I .1 cell ol I he inner epithelium (bar = 3 im). Cell cycle activity uml origin <>t ris\ti<- ln\i-rs The use of BrclU and 'MT incorporiilion luis provided further details concx-inm;' ihc role of cell proli feral ion in Cuvierian tubule regeneration and also, indireetly. concern- ing the possible contributions of migration and differentia- tion during regeneration. In // Ini^kiili. our ivsulis showed that cell proliferation occurs within each of the three constitutive tissue layers of ilir regenerating tubule and during the \\hole regenerative phase, although it was observed only in the mesothelmm at stage I and on!\ in I!IL- connective tissue layer at stage 5. Conversely, no labeled nucleus was observed after stage 5 HOl.OTHUROID CUV1ERIAN TUBULE REGENERATION 45 Mesothelium Connective tissue Inner epithelium Figure 25. Holotlmriii forskali. Reconstruction of a longitudinal section through the wall of a stage 5 regenerating Cuvierian tubule (not to scale). BL. basal lamina; BP, epithelial cell basal process: C. cilium; CM. circular myo- cyte; IA, inner area of the connective tissue layer; IE. internal epithelium; GC, granular cell; LM, longitudinal myocyte; M. mesothelium; MV. microvillus; NP. nerve plexus; OA, outer area of the connective tissue layer; P3, type 3 pseudopodial cell; PC, peritoneocyte; S. spherule; VC. vacuolar cell. X u TJ C Dfi 2 Regeneration stage Figure 26. Labeling indices (mean SD. n = 4) for proliferating cells in the different tissue layers at each stage of Cuvierian tubule regeneration (BrdU/anti-BrdU method). (i.e., when the regenerating tubule becomes functional), indicating that the growth phase proceeds only by an in- crease of cell volume, without cell division. Moreover, the absence of DNA-synthesizing cells in non-regenerating Cu- vierian tubules indicates that there is no cellular "turnover" and that the cell populations which constitute the tubule tissues must be considered as static (Messier and Leblond. 1960). During regeneration, the mesothelium is the tissue layer in which cell proliferation is the most precocious and the most important, involving both undifferentiated cells and peritoneocytes. At stage 1 . there is an accumulation of actively proliferating undifferentiated cells that form the mesothelial thickening characteristic of this stage. As re- generation proceeds, the percentage of labeled nuclei in the mesothelium regularly decreases in parallel with the differ- entiation of granular cells on the one hand and myocytes on the other hand. Nucleus labeling was never observed in these two cell types, which therefore do not divide. We were not able to determine the origin of undifferentiated cells, but they probably originate from peritoneocytes. These cells would thus contribute significantly to the regenerative pro- cess through dedifferentiation, proliferation, and redifferen- tiation into other cell types (in this case, granular cells and myocytes). Their relative totipotence in echinoderms is manifested by their ability to proliferate actively during regeneration (Candia Carnevali ct ul., 1997; present study) and to transdifferentiate into free coelomocytes or myocytes (Vanden Bossche and Jangoux, 1976; Dolmatov et al., 1996; respectively). In their description of the ultrastructure of non-regener- ating Cuvierian tubules, VandenSpiegel and Jangoux (1987) suggested that the intraconnective myocytes were derived 46 D. VANDENSPlWil I / / W 28 29 30 31 NIC jtmm 7 PC uc Figures 27-31. Holoihuriti fm \knli. Detection of proliferating cells in the regenerating C'uMcriun tubules. CL, connective tissue layer; IE. internal epithelium; I., lumen: M. mesolhehum; MC". mvocyte; P2, type 2 pseiKlnpiulial cell. PC. periloneocyte; UC, unditTerenliatcil cell. Figure 27. I .ongiludinal section ihnmgh a st.ige 2 regenerating tuhule (BrdU/anti-BrdU method i showing labeled nuclei in the three tissue layers (arrowheads: bar = ?() fimi. Figure 28. Detailed view of the wall ol a stage 2 regenerating tuhule (Brdtl/anli-BrdU method) showing labeled nuclei in the mcsothclmm (single arrowhead) and in the inner epithelium (double arrowhead) (bar = 5 /urn). Figure 29. Section through the wall of a stage 3 regenerating tuhule (BrdU/anti-Brdf method) showing labeled nuclei in the mesothelium (peritoneocyte. single arrowhead: and unditterentiated cell, double arrowhead) and in the conncclixc tissue laser (type 2 pseudopodial cell, arrow) (bar = 8 ^tm). Figures 3(1 and 31. Consecutne scniilhm and ultialhin sections through a stage 3 I. rating tuhule. *HT incorporation in the tuhule tissues was icxealed In auloradiography on the semithin section (arrowheads. I ig 2-: bar = 5 Jim) and the DNA-synthesi/ing cells wcie unambiguously identified by observation of the ultralhm section in I1-A1 il-'ig. 30: bar = 4 ^m). from the mesothelium. However, they had no direct e\ i deuce to corroborate their hypothesis. Our studs of regen- erating luhules clearly iiulicales ihc mesothelial origin of the myocytes which, during the regenerative process, first dif- ferentiate as myoepithelial cells and then migrate in the connective tissue layer. This is consistent with the hypoth- esis of Kicger and Lombard! (19X7) on the mesothelial origin ol nonepiihelial myocyles in echinoderms. This spec- ulative hypothesis was critici/ed hy Cavey and Wood (I WO), hut has since been strengthened by the work of Stauber ( 1993) on echinoid lantern muscles, lhai of Dolma- tov el ul. ( I Wo) on hololhuroid longitudinal muscle hands, and now by ours on holothuroid Cuvierian tuhule muscles. The nerve plexus appears early in regeneration and con sists only of nerve processes; no cell body \\as observed. This plexus is first basimesolhelial and then migrates in the connective tissue layer with the myocytes. This explains why, in non-regenerating tubules. .the nerve plexus is never observed within the mesothelium. Vacuolar cells were orig- inally described as egg-shaped structures and interpreted as sections through a spiral nerve ( VandenSpiegel and Jangoux, 1987). However, our reexamination of these struc- tures in the regenerating tubules suggests that ihey are clusters of vacuolar cells. Because these cells tirst appear close to the mesothelium and are surrounded by a basal lamina, their origin is probably in the mesothelium. I. ike myocytes and granular cells, they presumably differentiate from undifferentiated cells. These cells have also been oh- HOLOTHUROID CUV1ERIAN TUBULE REGENERATION 47 served in the Cuvierian tubules of other holothuroid species (VandenSpiegel and Jangoux, 1988), but their function re- mains enigmatic. In the connective tissue layer, spherulocytes and pseu- dopodial cells co-occur at the beginning of the regeneration. However, although the pseudopodial cells appear to prolif- erate actively, spherulocytes show no sign of DNA synthe- sis and disappear early in the regenerative process. Never- theless, they presumably play an important role in the early stages of Cuvierian tubule restoration. Indeed, most studies of wound repair in holothuroids have revealed the impor- tance of spherulocytes (Cowden, 1968; Menton and Eisen, 1973). The function generally ascribed to these cells is the formation and maintenance of the ground substance of the extracellular matrix (Fontaine and Lambert. 1977; Byrne. 1986; Jans et ui, 1996). In addition, spherulocytes have been put forward as producers of antibacterial compounds and chemotactic agents for other mesenchyniatous cells (see Smith, 1981. for review). The latter could account for the recruitment of type 1 pseudopodial cells at the beginning of Cuvierian tubule regeneration. Three types of pseudopodial cells follow one another in the tubule connective tissue during regeneration. Type 1 cells have all the characteristics of echinoderm phagocytes (e.g., see Dubois and Ghyoot. 1995). The collagen-contain- ing secondary lysosomes observed in some of these cells suggest they may have a tibroclastic function, cleaning up the connective tissue compartment before new collagen synthesis starts. Type 2 cells are characterized by a more rounded shape, a high nuclearcytoplasmic ratio, and a conspicuous juxtanuclear centriole all features suggesting that these cells are rather unditterentiated (Fontaine and Lambert. 1977; Smith, 1981). This notion is further sup- ported by the active division of these cells. Type 3 cells possess distended RER cisternae and Golgi stacks ending in spheroidal distentions. RER with distended cisternae has been described both in proven and presumptive echinoderm fibroblasts (Dubois and Ghyoot, 1995; Heinzeller and Welsch, 1994; respectively). Golgi with spheroidal disten- tions, on the other hand, are a diagnostic feature of verte- brate fibroblasts (Weinstock and Leblond, 1974). We have not observed all the stages of Golgi distentions, including the final collagen granule. Nevertheless, the ultrastructural characteristics of type 3 pseudopodial cells, together with their presence in the regenerating connective tissue close to the time when the collagen fibers appear, strongly support a fibroblastic function for these cells. According to the se- quence of appearance of the three types of pseudopodial cells, type 2 cells are probably dedifferentiated type 1 cells that then actively proliferate before differentiating into type 3 cells. Phagocytic cells, undifferentiated mesenchymatous cells, and fibroblasts would thus share lineage relationships, as is the case in echinoids (Dubois and Ghyoot, 1995). Once the regenerating tubule has reached its definitive size, the collagen synthesis presumably stops, as evidenced by the lack of mesenchymatous cells in non-regenerating Cuvier- ian tubules (VandenSpiegel and Jangoux. 1987). The origin of the tridimensional organization of the collagen in parallel helixes is not clear, but it could be the result of the particular association observed at regeneration stage 3 between type 2 pseudopodial cells and the mesothelium. In the inner epithelium, cells also divide actively, but only when they have not yet differentiated spherules in their basal intraconnective processes ("undifferentiated epithelial cells"). It appears, therefore, that in the three tissue layers of the Cuvierian tubules, regeneration proceeds by dedifferen- tiation, then proliferation, and finally differentiation. Cuvierian tubule regeneration: morphallaxis vs. epimorphosis Regeneration has been described classically as proceed- ing by one or the other of two mechanisms; (i) morphallaxis. in which cells differentiate or migrate from existing popu- lations, and (ii) epimorphosis. in which mitosis occurs and new cells are produced that either directly replace those lost or form a blastema that goes on to differentiate and replace lost tissue (Bonasoro et ai, 1998). In echinoderms, the relative importance of morphallaxis and epimorphosis seems to be variable. Regeneration of longitudinal muscle bands in holothuroids proceeds without cell proliferation and represents a case of cellular morphallaxis (Dolmatov et ul., 1996). Similarly, regeneration of the intestine after fission in the apodid holothuroid Leptosynapta is an entirely morphallactic mechanism; in this case, the remaining por- tion of the intestine is remodeled to form a functionally complete organ (Smith, 197 la, b; Gibson and Burke. 1983). Conversely, regeneration of the arm in the crinoid Antedon mediterranea is an epimorphic process in which a regener- ative blastema is formed (Candia Carnevali et al., 1993, 1995, 1997). Most regenerative processes in echinoderms. however, appear to involve a combination of morphallactic and epimorphic mechanisms. This is the case for the intes- tine regeneration after evisceration in aspidochirote and dendrochirote holothuroids (Dolmatov, 1992: Garcia Ar- raras et al., 1998). as well as for the arm regeneration in asteroids (Mladenov et al., 1989; Moss et al.. 1998). Regeneration of the Cuvierian tubules in H. forskali also appears to involve both epimorphosis and morphallaxis. Initial regeneration events occur by epimorphosis, cell pro- liferation being essential to the regenerative process. The mesothelial thickening appearing at stage 1 could be com- pared to a blastema, that is, an accumulation of undifferen- tiated cells capable of proliferation and differentiation. However, it is not a true blastema, giving rise to all the tissue layers in the new tubule; rather it is a transitory blastema-like structure that contributes only to the regener- ation of the mesothelium. forming mesothelial cells (peri- 48 D. VANDENSPIEGEL ET AL. toneocytes and granular cells) and mesothelium-derived cells (myocytes and vactMlar cells). This structure does not contribute to the regeneration of the other two tubule lay- ers i.e.. the connective tissue layer and the inner epithe- liumthat each nclose their own population of dividing cells. As regeneration proceeds, the percentage of dividing cells decreases, and in the late stages, tubule histogenesis occurs mainly bv morphallaxis. with nevslv differentiated cells migrating from the mesotheliuin into the connective tissue laver. The dynamics of Cuvierian tubule regeneration and its implication for the defense mechanism In holothuroid species that contain Cuvierian tubules, the structures are generally present in large numbers. For ex- ample, individuals of H. forskali may ha\e between 200 and 600 tubules, depending on their sj/e ( VandenSpiegel and Jangoux. 1987). When gently stimulated they expel only a tew tubules, but if the stimulation is stronger, tubules can be discharged several times in succession (Bakus. 1968; present study). However, in both cases, only a fraction of the total tubule number will be used. Such a behavior allows a sparing use of the defensive organ (VandenSpiegel and Jangoux. 1987). After expulsion and autotomy. Cuvierian tubules are readily regenerated. In H. forskali, our study showed that the complete regeneration of the autotomi/ed tubules, namely the formation of functional. fullv developed tubules, takes about 5 weeks. This period starts with the actual beginning of the regenerative process and not with the expulsion and autotomy of the tubules. Indeed, after a gentle stimulation, when only a few (about 15) tubules have been expelled, there is often a latent period that may last up to 3 weeks before the regeneration starts. In contrast, after a strong stimulation, when many (up to 300) tubules have been expelled, the regenerative process starts immediately, but proceeds by successive waves of 10 to 30 tubules that begin to regenerate at 10-day intervals. This pattern of regeneration is advantageous lor the sea cucumber because it allows a staggered spending of the encrgv necessary for regeneration. Holothuroid Cuvierian tubules thus constitute a very efficient defensive mechanism. Indeed, in addition to their remarkable structural organi/ation. which accounts for their adhesive and mechanical properties (VandenSpiegel and Jangoux. l'JS7i. their large number, sparing use. and particular regeneration dynamics also make them an almost inexhaustible line of defense maintained at limited energy cost. Acknowledgments We thank Professor P. Lassere for providing facilities at the Ohservatoire OccanoloL'ii|iie de Roscoff (Brittany. France) and P. Postiau for technical assistance. P. F'lam- mang is Research Associate of the National Fund for Sci- entific Research (NFSR. Belgium). This research was sup- ported by a NFSR grant to M. Jangoux (no. 2.4512.95. Credit aux Chercheurs). Contribution of the Centre Inter- universitaire de Biologic Marine (CIBIM). Literature Cited liai. M. M. 1971. Rcgenciation in the holothurian. Holothuria scabra Jager. Indian J. /;/>. Hiol. 9: 467-471. Itakus. (i. .1. 1968. Defensive mechanisms and ecology of some tropical holothurians. Mar. Bial. 2: 23-32. Kmiasoro, F.. M. I). ( anclia Carne>ali, C. Moss, and M. ('. Thonuhkc. 1998. Epimorphic irmi.v morphalactic mechanisms in arm regener- ation of crinoids and asteroids: pattern of cell proliferation and cell lineage. Pp. 13-18 in Ecliinmli rm\: Sun Francisco, R. Mooi and M. Telford. eds. Balkcmu. Rotterdam l<\rnt>. M. 1986. I he ultrastructure of the morula cells of Eupentacta quinquesemita (Echinodermata. Holotluiroideai and their role in the maintenance of the extracellular matrix. J. Mor/ihol. 188: 179-189. i .null. i Carnexali. M. I).. K. I.ucca. and !'. Kmiasnrn. 1993. Mecha- nisms of arm regeneration in (he feather star Antedon mediterranea: healing of wound and early stages of development. J. Exp. Zool. 267: 299-317. < .mill. i Carnevali, M. 1).. F. linnasoru. F. I.ucca. and M. C. Thorndykf. 1995. Pattern of cell proliferation in the early stages of arm regeneration in the feather star Antedon mediterranea. J. E.\p. /,.,nr. /)<>. Kiol. 23: 40-50. l)iiliiialo\, I. Y., M. (i. I h-i il.in.i A. A. l!nk'.ii,d\ I I (iinanova, N. E. I ainash. and V. P. knrcliayin. 1996. Muscle regeneration in (he hololhurian Slichi>pu\ japonicus. Rnu.\'s Arch. l)e\: Bin/. 205: 486-493. Dulmis, P.. and M. (ilnoot. 1995. Inlciiunicntary resorption and colla- gen synthesis during regression ot headless pedicellariae in Sphaere- i hiini.s ,<,'m/iii/iif7.\ (lichinodermala. l-xhinoidea). Cell Tissiif Rf\ 282: 297-309. I nis, HI R. H.. and I. ('. \\ilkii 1 . 198(1. Fission and autotomy in echi- noderms. Oceanngr. Mm: liiol. .\nnii. Rr\: 18: 155-250. I l.imm.iir. 1 I' 1996. Adliesion in eclunoileriiis. Pp. 1-60 in Echinodfrm Stuiiie.i. Vol. 5. M. Jangouv and .1. M I awrence. eds. Balkema. Rotterdam Konluini-. A. R., and I'. l.uinluTt. 1977. 1 lie tine structure of leucocytes of (he holothurian Curumaria minima. Can. J. Zool. 55: 1530-1544. (ialii 1 , M. 1968. Techniques histologiques. Masson, Paris. (iarcia-Arraras, J. K., I.. Kstrada-RodyiTs. R. Sanliagn. 1. I. Torres. I.. I)ia/-Miranda. and I. Torres- A \illan. 1998. Cellular mechanisms of intestine regeneration in the sea cucumber, Holothnriit glaberrimo Selcnka I Holothuroidea: Echinodermata). J. /;'/>. Zonl. 281: 288-304. i.ili on A. \\.. and R. I). Kurkc. 1983. Gut regeneration by morphal- HOLOTHUROID CUVIERIAN TUBULE REGENERATION 49 laxis in the sea cucumber Leptosynapta clurki (Heding. 1428). Can. J. Zoo/. 61: 2720-2732. Hacker, G., D. Springall, S. VanNoorden, A. Bishop, L. Grimelius, and J. Polak. 1985. The immunogold-silver staining method a powerful tool in histopathology. Virchows Arch. B 406: 444-461. ll< in/, II, i . T., and U. Welsch. 1994. Crinoidea. Pp. 9-148 in Micro- scopic Anatomy uf Invertebrates, Vol. 14: Echinodermata. F. W. Har- rison and F.-S. Chia. eds. Wiley-Liss. New York. Hill, R. B., and I). Reinschmidt. 1976. Relative importance of the antioxidant and anesthetic properties of propylene phenoxetol in its action as a "preservatie" for living holothurians. J. Invenebr. Pathol. 28: 131-135. Humphrey, C. D., and F. E. Pittman. 1974. A simple methylene blue-a/,ur II-basic fushin stain for epoxy-embedded tissue secretion. Stain Teclmol. 49: 9-14. Hyman, L. H. 1955. The Invertebrate. Vol. 4 Echinodermata. McGraw- Hill. New York. Jans, D., P. Dubois. and M. Jangoux. 1996. Defensive mechanisms of holothuroids (Echinodermata): formation, role, and fate of intracoelo- mic brown bodies in the sea cucumber Ho/niliiina inhiiln\a. Cell Tissue Res. 283: 99-106. Laurent, G., G. Toubeau, J. A. Heuson-Stiennon, P. Tulkens, and P. Maldague. 1988. Kidney tissue repair after nephrotoxic injury: bio- chemical and morphological characterization. Cril. KIT. To.\ictil. 19: 147-183. Menton, I). N., and A. Z. Eisen. 1973. Cutaneous wound healing in the sea cucumber. Tlmme hriareus. J. Morphol. 141: 1X5-204. Messier, B., and C. P. Leblond. 1960. Cell proliferation and migration as revealed by autoradiography after injection of thymidine-'H into male rats and mice. Am. J. Anat. 106: 247-265. Mladenov, P. V., B. Bisgrove, S. Asotra, and R. IX Burke. 1989. Mechanisms of arm tip regeneration in the sea star. Leptusterias he.\ac- lis. Rout's An-h. Dev. Bio/. 198: 19-28. Mosher, C. 1956. Observations on evisceration and visceral regeneration in the sea-cucumber, Actinopyga agassi-i Selenka. Zoologica (N.Y.) 41: 17-26. Moss, C., A. J. Hunter, and M. C. Thorndyke. 1998. Patterns of bromodeoxyuridine incorporation and neuropeptide immunoreactivity during arm regeneration in the starfish Asterias rube/is. Phil. Tunis. R. Soc. Loud. 353: 421-436. Miiller, M. E. G., R. K. Zahn. and K. Schmid. 1970. Morphologic und Funktion der Cuvierian Organe von Holothuria forskali Delle Chiaje (Echinodermata: Holothuroidea: Aspidochirota: Holothuriidae). Z. Wiss. Zool. 181: 219-231. Rieger, R. M.. and J. Lomhardi. 1987. Ultrastructure of the coelomic lining in echinoderm podia: significance for concepts in the evolution of muscle and peritoneal cells. Zoomorphology 107: 191-208. Smiley, S. 1994. Holothuroidea. Pp. 401-471 in Microscopic Anatomy of Invertebrates. Vol. 14: Echinodermata. F. W. Harrison and F.-S. Chia. eds. Wiley-Liss. New York. Smith, G. N. 1971a. Regeneration in the sea cucumber Letosynapki. I. The process of regeneration. J. Exp. Zool. 177: 319-330. Smith, G. N. 1971h. Regeneration in the sea cucumber Letosynapta. II. The regenerative capacity. J. Exp. Zool. 177: 331-342. Smith, V. J. 1981. The echinoderms. Pp. 514-562 in Iinrrtchnite Blood Cells, N. A. Ratcliffe and A. F. Rowley, eds. Academic Press, New York. Stauber, M. 1993. The lantern of Aristotle: organization of its coelom and origin of its muscles (Echinodermata. Echinoidea). Zoomorphology 113: 137-151. Tracey, D. J. 1972. Evisceration and regeneration in Tliyone okeni (Bell. 1884). Pruc. Linn. Soc. N. S. W. 97: 72-81. Vanden Bossche, M., and M. Jangoux. 1976. Epithelial origin of star- fish coelomocytes. Nature 261: 227-228. VandenSpiegel. D., and M. Jangoux. 1987. Cuvierian tubules of the holothuroid Holoihuria forskali (Echinodermata): a morphofunctional study. Mar. Biol. 96: 263-275. VandenSpiegel. I)., and M. Jangoux. 1988. Les tubes de Cuvier d' Holothuria nianiinata Grube, 1840 (Holothuroidea, Echinodermata). Ann. Soc. R. Zool. Bc/t;. 118: 191-198. VandenSpiegel, D., D. Nonclercq, and G. Toubeau. 1991. Immunocy- tochemical detection of bromodeoxyuridine in proliferating cells of regenerating Cuvierian tubules of Holothuria forskali. Pp 81-85 in Echinoderm Research 1991. L. Scalera-Liaci and C. Canicatti. eds. Balkema. Rotterdam. VVeinstock, M., and C. P. Leblond. 1974. Synthesis, migration, and release of precursor collagen by odotonblasts as visualized by autora- diography after ('H)proline administration. / Cell Biol. 60: 92-127. Reference: Biol. Bull. 198: 50-66. (Fehnuiv 2iXKii Nucleation and Growth of Calcite on Native Versus Pyrolyzed Oyster Shell Folia C. S. SIKES 1 *. A. P. WHEELED. A. WIERZBICKI 1 . A. S. MOUNT 2 . AND R. M. DILLAMAN 4 l The Minerali-ation Center. Department of Hi CI(iiiM>n L'niver.\it\. Clcmson. South Carolina 2 Scii'iicc\: epartment of Biological Sciences, v<)3; * Department of Chemistry, I /in i inuai . 1999 i' . pti -I 16 Novemb i 1999 I" uliinii correspondence shuuld Ix- ;nltln's-.i.-il. !, mail: ssikt-s(" III II ."lllll.ll l-.lll Introduction The mineral of most molluscan shells is CaCO, that routinely contains less than a few percents and often less than \7c by weight of the shell as organic material, mostly protein. One objective of shell research is to understand how the shell is put together, particularly the relationships be- tween the mineral and the organic components. The intrin- sic appeal of exquisite biominerals like shells has been advanced in recent years by an appreciation of the material properties of the structures (Heuer et ai. 1992: Sarikaya et nl.. 1995: Mann. 1496: Stupp and Braun. 1997; Weincr and Addadi. 1997). A variety of studies have shown that bio- logical composites like shell have particularly favorable properties, including increased durability and resistance to fracture. Such properties arc considered to result from the interplay of the organic and inorganic components, even though the organic content is quite low by composite- materials standards. In the case of the oyster shell, which is composed of calcile and is representative of many kinds of molluscan shell, the mineral even within a lew millimeters of the forming edge is formed into layers of thin sheets, termed folia (Tsiijii el 1()00) on bleached shell surfaces, revealing asymmetri- cal, three-sided pyramidal crystals, perhaps the same as observed herein. Other workers (Wada. 1963. 1968; Watabe. 1965. 1981) had observed crystallites on folia that were arranged with the c axis vertical to the plane of foliation and therefore thought to be nucleated from the (00 1 ) plane. We have also sometimes seen this on control, foliar chips that have prom- inent ectopic. polygonal crystals that appeared to be nucle- ated from the basal plane (not shown). As reported herein, the control foliar surfaces may act as nucleation sites for such basal calcite crystals at relatively high levels of super- saturation. The c\cle of mineralization in vivo The formation of shells as layers of mineral interspersed with organic sheets implies a cyclic process of organic secretion and inorganic mineral growth (e.g.. Degens, 1976; Wheeler and Sikes, 1989). Details central to this process are poorly known but include the in vivo levels of fi and the organic components, as well as the relative timing of secre- tion of lattice ions and organic matrix. The AFM measure- ments of growth of calcite on shell fragments, together with measurements of shell growth in vivo, may provide a way to estimate the effective fl in vivo, as well as the time frame of lath formation, as explained below. Starting with the possible levels of lattice ions and or- ganic matrix in vivo, measurements of the ionic and organic 60 C. S. SIKES tT AL 23 composition of extrapalliul fluid (FPF. ihe medium of shell formation), are available by means of exacting microanal- ysis following attempts to sample the fluid. In the Eastern oyster, the estimate (C'renshaw, 1972) of calcium content was close to that of seawater at nearly 10 m/V/, inorganic carbon was as high as 5 mM (with seawater normally at about 2.2 mM), but the pH at about 7.4 was notably lower than typical values for seawater ( 8.2), which in turn would result in significantly lower levels of carbonate ion. These conditions of the EPF would convert to an il of 2.83, which is quite low. Studies of the EPF of other molluscs yielded roughly similar results (Wada and Fujinuki. 1976; Misi- ogianes and Chasteen. 1979; Nair and Robinson. 1998). Moreover, other solutes of the FPF such as magnesium. CALCITE GROWTH ON OYSTER SHELL FOLIA 61 Figure 19. Computer model of a calcite rhombohedron. The model is depicted with the c axis upward, in the plane of the page. Planes are cut through the eight corners. The arrows indicate the bottom and top planes (basal planes), which are (001) surfaces. The planes through the other corners are the family of ( 1 1 0) planes. Each of the sides is a member of the (104) family of planes. Figure 20. A computer diagram of the (001) surface of calcite, showing an upper plane of calcium atoms, and an underlying plane of carbonate groups. Notice how the calcium atoms coordinate with three oxygens of separate underlying carbonate groups. Each calcium atom also would coordinate with three oxygens of separate carbonate groups of a carbonate homoplane (not shown) that would overlay the calcium plane in a similar manner, with successive homoplanes of calcium and carbonate alternating to form the calcite crystal. The c axis runs perpendicular to the (001) plane. The a axes are shown as the grid of lines on the diagram. The (1 - 1 0) family of planes also run perpendicular to the (0 1 1 plane (or parallel to the c axis), emerging perpendicularly from the a axes. Scale bar = 0.5 nm. Figure 21. Atomic force (AF) micrograph of the atomic pattern of one of the uppermost surfaces of the crystals of Figure 16. The spacings of atoms are consistent with the ( 1 - 1 0) surface of calcite. The more clearly resolved rows of atoms (from upper left to lower right) are thought to correspond to protruding oxygens of carbonate groups, with the atoms that tend to blend together in the other rows (also from upper left to lower right) corresponding to calcium atoms that are somewhat recessed into the surface. Scale bar = 3 nm. Figure 22. AF micrograph of the atomic pattern of one of the sides of the ectopic crystals as in Figure 16. viewed in artificial seawater saturated with respect to calcite. The spacings of atoms are consistent with the (104) surface of calcite. In this image, each lattice position is thought to correspond to a protruding oxygen atom of the carbonate groups, which are elevated to some extent relative to calcium atoms that occur just below the (I 04) surface. Scale bar = 2.0 nm. Figure 23. A computer diagram of the ( 1 - I 0) surface of calcite, showing the oxygen atoms (white) of the carbonate groups outermost and the calcium atoms (gray) slightly recessed into the surface. The number of oxygen atoms per unit area of this surface is not nearly as dense as in the ( 1 4) surface, and therefore both the oxygens and the calcium atoms can be seen in AF micrographs of the (1 1 0) surface. The atoms within one row of oxygens are spaced 4.99 A apart, with the nearest oxygens in separate rows 8.53 A apart. This matches within 5% of the atomic spacings of the uppermost, forming surface at the corner of one of the ectopic crystals on a pyrolyzed foliar surface as seen in Figure 21. Note that if only every other row of oxygens of the ( 1 04) surface has the atoms clearly resolved, with the other rows of oxygens having atoms blurred together (again with no calcium atoms observed at all), which can be done by increasing the force of the AFM probe as well as the gain setting, a pattern somewhat similar to the (1 - 1 0) pattern can occur. However, the spacings and angles between atoms of the two surfaces still differ, as does the number of atoms per unit area. Scale bar = 1.0 nm. Figure 24. A computer diagram of the ( 1 04) surface of calcite. showing only the outermost oxygen atoms of carbonate groups, as are seen in AF micrographs of this surface. The calcium atoms are recessed by about 1 A into the surface and do not show up in AFM's due to the closeness of the surface oxygen atoms. The nearest neighbors of outermost oxygen atoms form rhombohedrons with spacings of 4.24 and 4.99 A, which matches within 5% of the spacings as seen in Figure 22 of the sides of the ectopic crystals on a pyrolyzed foliar surface. Scale bar = 2.0 nm. phosphate, small chelants that bind soluble calcium, and organic components like soluble shell proteins are all inhib- itory to calcite formation. If these constituents are taken into account, the measurements actually suggest that the EPF is undersaturated with respect to calcium carbonate. In this regard, Wilbur and Bernhardt ( 1984) attempted an in vitro assay for calcite formation based on prior estimates of the content of EPF. It was necessary to raise calcium to 28.2 mM and inorganic carbon to 7.6 mM at pH = 8.33 (including Mg at 47.2 mM, but no phosphate and no organic compounds) to obtain nucleation after about 20 min at 21C. These levels of soluble lattice ions correspond to ft = 72.4. In the absence of Mg, under otherwise similar condi- tions, we obtained nucleation after about 20 min at fi = 37.2. By comparison, seawater itself is considered to be mildly supersaturated with respect to calcite (II ~ 10). However, again, the other constituents such as magnesium, phosphate. and organic components act to stabilize seawater so that calcium carbonate does not normally precipitate. Turning to rates of shell deposition in vivo, on the other hand, direct measurements of shell enlargement revealed substantial rates of carbonate deposition that are indicative of notably high supersaturations. For example, in the scallop Argopecten imulicms and the marine clam Mercenaria mer- cenaria, shell enlargement was measured by radioisotopic approaches to range from about 0.1 to 2.0 /nm cm" 2 h~' (Wilbur and Jodrey, 1952; Wheeler et ai, 1975; Wheeler and Wilbur, 1977; Dillaman and Ford, 1982). By dividing by the density of shell, this value was converted to an increase of a single linear dimension such as thickness or elongation of about 35 to 700 nm cm~ 2 h~'. Regarding the possible levels of fi in vivo as indicated by AFM observations of calcite growth in vitro, the measure- ments of calcite formation on pyrolyzed foliar surfaces under controlled conditions of supersaturation indicated an 62 C. S. SIKES ET AL emergence of foliar crystal elements of about 1 to 2 nm h ' at (1 = 10.7. At fi = 16.3. the average crystal growth over the surface was about 100 to 150 nm h '. At 11 = 26.9, on average, the ectopic crystals emerged from pyrolyzed foliar surfaces at 300 to 500 nm h~ '. Comparison of these values to the above-referenced measurements of shell thickening suggests that the levels of fi locally \\ithin the EPF must have been in the range of about 16 to 27. and perhaps even somewhat higher, exclusive of the effects of any inhibitory constituents of the EPF that may have been present. It soluble inhibitors of calcite formation were present, the levels of fi would need to be even higher to dri\e the observed rates of shell growth. One issue that affects direct measurement of constituents of the oyster EPF is the ability to obtain fluid for analysis. The mantle lies essentially in contact with the shell and the volume of EPF available for sampling is thus minimal to nonexistent (e.g.; see figure 9 of Watabe. 1974; figure I of Erben and Watabe. 1974: and figures 2 and 4 of Bevelander and Nakahara. 1980). Consequently, direct analysis of the actual levels of lattice ions and various inhibitors or pro- moters of crystalli/ation in the fluid of minerali/ation may be impossible (Wheeler and Sikes, 1989: Wada. 19W assuming that there is a fluid as such. Having some bearing on the question of whether fluid is present is GaltsotTs view (1964) that minerali/ation in the oyster is a kind of solidi- fication within a gel-like layer that is deposited by the mantle on the shell as it grows. He observed rhythmic back-and-forth movements by the mantle on the shell along the direction of foliation. These movements were accompa- nied by deposition of a highly viscous coating, which sub- sequently became minerali/ed (see also Carriker el cii. I ''SO; Carriker. 1996). Crenshaw (1990) and Calvert and Crockett (1997) have further reviewed aspects of hiomineral formation within a possibly gel-like precursor. When extracted by slow dissolution of shell, the shell matrix proteins, most of which are fundamentally similar in amino acid composition, are collected over a continuum of molecular weights (M u ) that ranges into the millions and higher (Wheeler el <;/., 1988). The higher M w fractions are gelling materials that are highly interactive with water but not actually water-soluble (Wheeler and Koskan. 1993: Wheeler el nl.. uiipubl. ohs.). Some of the lower M u mate- rials that are released from shell as soluble fractions might also be linked to the gelling matrix when in the shell, but this is not known. The possible significance of soluble versus gelling pro- teins bears on the question of the effective supersaturation of the mincrali/ing environment. Even low doses of soluble protein from oyster shell are strongly inhibitory to calcite nucleation (0.1 /ag protein ml ' at fi = 42.7; Wheeler and Sikes. 1989) or growth ol e.ileiie seed crystals (0.015 /u.g protein ml" 1 cm 2 at fi < 5; Low, 1990: Wheeler el al.. 1991). The incorporation of (he soluble protein into the crystals under these conditions was measured at 0.4 to 0.5 ^ig pmole ' calcite (Sikes and Wheeler, 1986). which is approximate!) equal to the amount of the protein that occurs in the folia. However, if the shell is to grow at the rates that have been measured in vivo, which would require levels of fi that are significantly elevated relative to seawater. it would seem that the shell protein, if present at all during intervals of mineral deposition, would occur in the minerali/ation fluid in soluble form at levels too low to account for the amount of protein in shell. However, this problem could be oxer- come if the secretion of shell protein by the mantle were episodic such that intervals of shell growth, during which no inhibitory proteins were present at the site of crystalli/ation, were followed by intervals during which the newly formed mineral layer became covered by proteinaceous material. In this framework, the measurements of inhibition of calcite formation in vitro at low doses of samples of whole EPF (Wilbur and Bernhardt. 1984) may be understood. When 210 /j.1 of EPF from M. mercenaria was added to a calcite growth assay, inhibition was observed at roughly the level of 0.1 jug protein ml ' established in other studies (Wheeler and Sikes. 1989). Given the volumes used in the assay, this would suggest that the EPF sample contained about 3 /u.g of protein, or roughly 15 /jig protein ml ' of EPF. This is such a high level that it would totally inhibit shell growth, even at very high fi. if the protein molecules were present in a soluble form that could interact with the cry sial surfaces. On the other hand, it a pulse of protein into the liPF was layered quickly onto a fresh mineral layer of shell and followed by an interval of protein-free EPF of fi in the range of 16 to 27. the pattern of mineral deposition that is observed in foliar layers could occur. If such intervals of pulsed secretion are on the order of minutes, then samples of EPF". which consist of pooled \olumes collected over periods of hours (Wada. 1990). would always contain shell proteins. There is evidence to support these speculations, as follows First, the foliar laths have thicknesses in the range 100 to 350 nm. As cited above, the shells of various molluscs have been measured to thicken at rates of 100 to 700 nm h ', thus formation of individual laths appears to occur at a rale of about 1 or more per h. Each" lath is covered by a layer of foliar globules, which are thought to consist of shell protein and amorphous CaCO,. This morphology is consistent w ith pulsed secretion of the protein, perhaps during an interval of only a lew minutes per hour, when the protein is removed from solution by binding to the previously formed mineral. Following that, a period on the order of an hour of protein- free crystal growth would occur. Next, a new organic layer, which appears initially to be a gel (Wheeler el cil.. unpubl. ohs.). is secreted, hinds to the lath, and subsequently he comes covered by calcite. Thus, the gelled organic layer also becomes minerali/ed during this process, apparently in the form of amorphous CaCO, of the Foliar globules. Conceivably, the next layer of calcite could nucleate from CALCITE GROWTH ON OYSTER SHELL FOLIA 63 the organic layer, from exposed areas of calcite of the underlying lath, or both. That there are exposed areas of calcite even after the organic layer is bound seems probable because the pyrolyzed foliar chips were firm and did not crumble on handling following removal of the organic ma- terials. This suggested that the mineral was continuous throughout the layers of folia. Direct evidence of continuity of this type has been observed between aragonitic tablets of the nacreous layers of shells (Watabe, 1981; Giles et al., 1995). If there are areas of exposed calcite on the foliar surfaces after the organic layer is bound, these would serve as ready nucleation sites for new laths. Such exposed areas of calcite were not visually evident in the AF micrographs of fractured surfaces of folia, and therefore would necessar- ily be very small, presumably on the scale of nanometers. The foliar globules themselves were not particularly ef- fective as nucleation sites, and calcite crystals that were nucleated from them by raising the level of 11 did not appear to have the crystallographic orientation and morphology of the pyrolyzed lath. Hence, each succeeding foliar layer seemed to track the mineral of the underlying layer rather than to derive from the amorphous character of the foliar globules. Another line of evidence that is consistent with the notion of a cycle of mineralization as presented above involves measurements of interactions of the soluble protein from oyster shell with calcite in vitro. When the protein was bound to calcite crystals, the maximum capacity of binding was 0.18 ;u,g cm 2 (Wheeler et al., 1991 ). Taking a value of 0.5% protein by weight in foliar shell, which agrees well with measurements of protein content of shell extracts, including foliar layers (reviewed in Sikes et /., 1998). and with radioisotopic incorporation of the protein into calcite in vitro (Sikes and Wheeler. 1986), there would be 0.5 /ng protein per jumole (100 ^.g) calcite. Using 2.6 as a typical value for the density of shell (Wheeler et til.. 1975). there would be 13 mg protein crrT 3 of shell. If this protein were deployed in layers at 0.18 fig protein cirT : , there would be enough protein to form a stack of 72,222 layers of 1 cm 2 in each cm\ This corresponds to about 140 nm per layer (including the protein), which is at the low end of the range of thicknesses of a foliar lath. One problem with these conjectures about binding and coverage involves drawing inferences from in vitro studies and applying them to in vivo processes. For example, the binding studies were based on measurements of the inter- action of the protein with (104) surfaces of calcite. More- over, the incorporation studies permitted the protein to interact freely with the growing crystal and likely involved several surfaces. The in rivo interaction, on the other hand, appeared mainly to be with the ( 1 - 1 0) surface. The binding affinity of the protein to the ( 1 - 1 0) surface has been calculated to be higher than to the (104) surfaces (Wierzbickma/.. 1994; Sikes and Wierzbicki. 1995, 1996). However, the binding capacity and coverage of the surfaces would probably be similar. That is, taking the average oyster shell protein as a globule at 50 kD with a linear dimension of about 5.4 nm (Sikes et al., 1998), each unit of protein would cover about 30 nm 2 . The maximum capacity of 0.18 /ng of protein per cm 2 of calcite would represent 2.17 X 10 12 molecules (50 kD = 8.3 X 10~ 2 " g), or a coverage of 6.5 X 10 13 nm 2 . Comparing this to 1 x 10 14 nm 2 per cm 2 suggests that 0.18 ju,g of the protein would cover most of 1 cm 2 of calcite regardless of the specific surface. Of course, it is also possible that the protein mol- ecules (and the foliar globules) can form more than a monolayer on the calcite of the laths. This in turn would translate into laths of greater thickness as calculated above, which is often seen in the AFM and SEM images, both herein and in prior studies. Conclusions Overall, the evidence can be interpreted to support a hypothesis of pulsed secretion of the shell proteins, deple- tion of the soluble protein from the EPF upon binding to the mineral of laths, and subsequent intervals of growth of new mineral at relatively high fi. The pulses of shell proteins might occur in a time frame of minutes, with the interval of mineral growth more on the order of an hour. The hypoth- esis would include secretion of the proteinaceous material primarily as a gelatinous covering that terminates the growth of the underlying lath and defines its surface as a member of the calcite (1 1 0) family of planes. In view of the need to elevate fl to support the observed rates of folia formation in vivo, as well as to overcome the influence of soluble inhibitors in the EPF, it seems neces- sary for the organism to make significant metabolic invest- ment in mechanisms to generate the mineralizing microen- vironment. As indicated in many prior studies and reviews, although perhaps not at the magnitude suggested herein, the necessary supply of lattice ions may be provided by mech- anisms such as direct active transport of inorganic carbon or calcium, or a combination of transport processes. Another factor to consider is the formation of the crystals in thin, unstirred layers that are diffusion-limited. Watabe and coworkers (Watabe et al.. 1958; Watabe and Wilbur, 1961) noted that folia formation produced tablet-like mor- phologies that were similar to CaCO-, crystals grown in vitro in unstirred layers, probably initiated in an area of highest fl with growth toward regions of lower 1, thus producing directional laths. Although crystallization assays often involve high rates of stirring to eliminate diffusion as a rate-limiting variable, some assays have been run under diffusion-limited conditions with concentration gradients into unstirred layers. It is interesting to note that calcium carbonates so produced frequently do exhibit tabular mor- phologies that resemble foliar laths to some extent, includ- ing crystals grown in the presence of shell protein (Fallini et nl.. 1996; Weiner and Addadi. 1997). In related observa- 64 i s MM s I I AL. lions, Gower (Gower and Tirrell. 1998; Gower. pcis. comm.i has noted thai at higher concentrations under spe- cific conditions of unstirred supersaturation. low M w s, poKaspartates and acidic polypeptides can induce \\\ ' droplets of a liquid-like, mineral-precursor phase that con- verts to calcitic tablets and other CaCO, struct urev It thus may also be useful to keep in mind thai in i only can the shell matrix proteins be arranged as the suptamo- lecular assemblage of the gelling material, but aNo even the soluble molecules of relatively low M, A can agglomerate both in solution and when binding to mineral in the torm ol ellipsoids of linear dimensions up to about I no nm i \Vieiv- bicki et al.. 1994; Sikes et al.. I998t. Such associatise behavior of other biomineral proteins such as dentin phos- phophoryn. as well as other polvamomc proteins such as casein, has been reviewed hv Maish i |9,X9a. hi. who also observed agglomerations ot phosphoproteins and mineral salts including amorphous calcium phosphate in nonminer- ali/ing compartments ol some molluscs i Marsh and Sass. 1983. 1984. |985i. None ol these protein-mineral agglom- erations appeared to be very effective as nucleators ol crystals of shell or teeth (Marsh. I98h. I989a. b. 1994). Similarly, the anielogeiims. the principal proteins of the formative stage of tooth enamel, have been shown to form gels composed of agglomerative "nanospheres." These nanospheres have been observed in viva with dimensions of 15 to 20 nm (Fincham et til.. 1994. I99S) and in \-itru with dimensions that range to 100 nm and higher (Wen et al.. 1999). The amelogenins are mainly hydrophobic. hut they do have anionic C-termini that appear to occur at the surface of the nanospheres and interact there with the calcium hydroxyapatite of enamel. The amelogenin nanospheres are thought to function less in apatite nucleation than in local- i/ing and orienting the developing enamel crystallites prior to their maturation (Fincham .miK-iali/ation. Proc. Nail. Acad. Sri. USA H2 lil. nli I... ,|. Moradian. K. Shay. N. d. Maroiidas. and S. \\eiiuT. 1987. A chemical model lor the cooperation of sultales and carboxy- lales in calcite crystal nucleaiion. Relevance lo biominerali/ation. Proc. Nail. Acad. Sci. USA 84: 2732-2736. Vi/cnbiT)!. .1.. (J. l.amhiTl. I,. Addadi. and S. \Yvinrr. 1996. Slahili- /.iin in ot amorphous calcium carhonale by specialised macromolecules in biological and synthetic precipitates. Ad\: Mater. 8: 222-226. Vllnik. S.. S. Weiner, and I.. Addadi. 1996. Polysaccharides of in- tracrvstalline glycoproteins modulate calcite crystal growth in vitro. Chem. Ear. J. 2: 278-284. Htniash. K.. .1. Aizvnberg, I.. Addadi. and S. Wcintr. 1997. Amor- phous calcium carbonate transforms into calcite during sea urchin larval spicule growth. Proc. K. Soc. Land. B 264: 461-465. I'.. MM. in A., I.. Vilil.nli. and S. \\cintr. 1988. Interactions of sea- uicliin skeleton macromolecules with growing calcite crystals a study nl inlracrvsialline proteins, \aiure 331: 546 54S. Bi-v dander. (;., and H. Nakahara. 198(1. Compartment and envelope lormation in the pu>ces\ ol biological minerali/ation. Pp. 19-27 in The Mechanisms of Biomineralization in Animals and Plains. M. Omori and N. Watahe. eds. Tokai l.'niversity Press. Tokyo. Bi> nton. K. S. 1 98(1. ( 'hcmisiry and Technology of Lime ami Limestone. I \\ilev. New York. C'alxiTt, I'., and R. Crockett. 1997. Chemical solid free-form fabrica- tion: making shapes without molds. Chem. Mater. 9: 650-663. Carrikcr. M. R. 1996. I lie shell and ligament. Pp. 75-168 in The /,M/rc;i Ovster. V. S. Kennedv. R. I. t. Newell, and A. F. Eble. eds. M. 11 viand Sea Grant College Publication. College Park. ( arriker, M. R., and R. K. Palmer. 1979. Ultrastructural morphogen- esis ot prodissoconch and early dissoconch valves of the oyster Cras- wnr/,d \'i/\ini(,i. I'l-nc. \atl. Shelltuh. AXMH: 69: 103-128. Carriki-r, M. R.. R. K. PalmiT. and R. S. Prezant. 198(1. I unclional ultramorpholog) of the dissoconch valves of the oyster Crassastrea virginit-a. /'/.', \<;;/ s/i. l/tnh. A.\\oc. 70: 139-183. Cri'iishaw. M. A. 1972. The inorganic composition of molluscan extra- palhal Mind. H/../. Hull. 143: 506-512. Cri'iishaH. M. A. 1990. Biommerali/alion mechanisms. Pp. 1-9 in Skel- etal Biomineralization: Paitern.i. Processes, and Evolutionary Trends. Volume I. J. G. Carter, eel. Van N'ostrand Reinhold. New York. Dean. .1. A. 1992. Lange'f llandl;>k t 'Chemistry. McGraw-Hill. New ^ ork Drums. K. I. 1976. Molecular Mechanisms on Carbonate, Phosphate. ,in,l S///1H l>i'i>ii\itiini in tin- l.iviiii; Cell. Springer. Berlin. Didvmus. .1. M., P. Oliver. S. Mann. \. 1 . De\ rics. P. V. Maiisclika. and P. \\fsllirock. 1993. Intluence ot low-molecular weight and macromolecular organic additives on the morphology ol calcium car- honale. ./. Chem. Soc. /'7 69. I in. li. MM \. (i.. J. Moradian-Oldak. J. P. Simmer. P. Sarte, K. C. I .m T. Dii'kwiscli. and II. C. Slav kin. 1994. Self-assembly of a recom- hinani .unelogenin protein generates supramoleculai structures. J. MI-IHI /.'/,./ 112: MM 109. I imham. \. (,.. ,|. Mdiadian-Oldak. and H. C. Slavkin. 1995. I M dence tor amelogenin "nanospheres" as functional components of secretory-staj'r enamel matrix. ./. Struct. Hiol. 115: 50-59. Hiirhain. \. (... J. Miirailian-Olilak. anil .1. P. Simmer. 1999. The trUCtUral biolog) ol ihe developmi: dental enamel matrix. J. Struct. Kn.l 126: ' n |99 CALCITE GROWTH ON OYSTER SHELL FOLIA 65 Galtsoff, P. S. 1964. The American oyster Crassostrea in^inica Gmelm. Fish. Bull. 64: 1-480. Giles, R., S. Marine, S. Mann, D. E. Morse, G. D. Stuck), and P. K. Hansnia. 1995. Inorganic overgrowth of aragonite on molluscan nacre examined by atomic force microscopy. Bioi Bull. 188: 8-1?. Gower, L. A., and D. A. Tirrell. 1998. Calcium carbonate films and helices grown in solutions of polyl aspartate ). J. Cryst. Growth 191: 153-160. Heiier, A. H., D. J. Fink, V. J. Laraia, J. L. Arias. P. D. Calvert, K. Kendall, G. L. Messing, P. C. Rieke, D. H. Thompson, A. P. Wheeler, A. Veis, and A. I. Caplun. 1992. lnno\ alive materials processing strategies: a biomimetic approach. Science 255: 1098- 1105. Low, K. C. 1990. Adsorption of oyster shell matrix and its peptide analogs onto calcium carbonate: mechanisms of inhibition of crystal growth. Thesis. University of South Alabama. Mobile. 75 pp. Lowenstam, H. A., and S. Weiiier. 1989. On Biomineralization. Oxford University Press. New York. Mann, S., ed. 1996. Biomimetic Approaches in Materials Science. VCH Publishers. New York. Mann, S.. .1. M. Didymus, N. P. Sanderson, B. R. Heywood, and E. J. A. Samper. 1990. Morphological influence of t'unctionalized and non- functionali/.ed a, O-dicarboxylates on calcite crystallization. J. C/iem. Soc. Fiiniiluy Trans. 86: 1 873- 1 880. Marsh, M. E. 1986. Biomineralization in the presence of calcium- binding phosphoprotein particles. J. Exp. Zoo/. 239: 207-220. Marsh, M. E. 1989a. Self-association of calcium and magnesium com- plexes of dentin phosphophoryn. Biochemist 28: 339-345. Marsh, M. E. 1989b. Binding of calcium and phosphate ions to dentin phosphophoryn. Biochemistry 28: 346-352. Marsh, M. E. 1994. Polyanions and biominerali/ation. Bull. lust. Ocewiogr. {Monaco) 14(1): 121-128. Marsh, M. E., and R. L. Sass. 1983. Calcium-binding phosphoprotein particles in the extrapallial fluid and innermost shell lamella of clams. J. E\p. Zool. 226: 143-203. Marsh, M. E., and R. L. Sass. 1984. Phosphoprotein particles: calcium and inorganic phosphate binding structures. Biochemistry 23: 1448- 1456. Marsh, M. E., and R. L. Sass. 1985. Distribution and characterization ol mineral-binding phosphoprotein particles in bivalvia. J. Exp. Zool. 234: 237-242. Misogianes. M., and N. D. Chasteen. 1979. A chemical and spectral characterization of the extrapallial fluid of Mylilus euulis. Anal. Bio- chem. 100: 324-334. Nair, P. S., and W. E. Robinson. 1998. Calcium spcciation and ex- change between blood and extrapallial fluid of the quahog Mercenaria merccnaria (L.). Bioi Bull. 195: 43-51. Okazaki, K., and S. Inoue. 1976. Crystal property of the larval sea urchin spicule. Dev. Growth Differ. 18: 413-434. Okazaki, K., R. M. Dillaman, and K. M. Wilbur. 1981. Crystalline axes of the spine and test of the sea urchin Strongylocentrotus purpu- ratus: determination by crystal etching and decoration. Bioi. Bull. 161: 402-415. Paine, R. T. 1964. Ash and calorie determinations of sponge and opis- thohranch tissues. Ecology 45: 384-387. Price, T. J., G. W. Thayer, M. W. LaCroix, and G. P. Montgomery. 1976. The organic content of shells and soft tissues of selected estuarine gastropods and pelecypods. Proc. Nat/. Shellfish. Assoc. 65: 26-31. Runnegar, B. 1984. Crystallography of the foliated calcite shell layers of bivalve molluscs. Alcheringa 8: 273-290. Sarikaya, M.. J. Liu, and I. A. Aksay. 1995. Nacre: properties, crys- tallography, and formation. Pp. 35-90 in Bininimctics: Design and Processing of Materials. M. Sarikaya and I. A. Aksay. eds. American Institute of Phvsics, New York. Sikes, C. S., and A. P. Wheeler. 1986. The organic matrix from oyster shell as a regulator of calcification in \-ivo. Bioi Bull. 170: 494-505. Sikes, C. S., and A. Wierzbicki. 1995. Mechanisms of regulation of crystal growth in selected biological systems. Pp. 183-206 in Mineral Scale Formation ami Inlnhition, Z. Amjad. ed. Plenum Press. New York. Sikes, C. S., and A. Wierzbicki. 1996. Polyamino acids as antiscalants, dispersants. antifreez.es. and absorbent gelling materials. Pp. 249-278 in Biomimetic Approaches in Materials Science. S. Mann, ed. VCH Publishers. New York Sikes, C. S., A. P. Wheeler, A. Wierzbicki, R. M. Dillaman, and L. DC Luca. 1998. Oyster shell protein and atomic force microscopy of oyster shell folia. Bioi. Bull 194: 304-316. Sunk is-.. K. 1994. Amorphous minerals in biology. Bull. lust. Occanogr. (Monaco I 14(1): 49-54. Simkiss, K., and K. M. Wilbur. 1989. Biomineralization. Academic Press, San Diego, CA. Stupp, S. I., and P. V. Braun. 1997. Molecular manipulation of micro- structures: biomaterials, ceramics, and semiconductors. Science 277: 1242-1248. Taylor, J. D., W. J. Kennedy, and A. Hall. 1969. The shell structure and mineralogy of the bivalvia. Introduction. Nuculacea-Trigonacea. Bull. Br. Mns. {Nat. Hist.) Zool.. Supplement}: 1-125, plates 1-29. Tsujii, T., G. D. Sharp, and K. M. Wilbur. 1958. Studies on shell formation. VII. The submicroscopic structure of the oyster Crussosirca rirginica. J. Biophys. Biochem. Cytol. 4: 275-279. Wada, K. 1963. Studies on the mineralization of the calcified tissue in molluscs. VI. Crystal structure of the calcite grown on the inner surface ol calcitostracum. /. Electron Microsc. 12: 224-227. Wada, K. 1968. Spiral growth of calcitostracum. Nature 219: 62. Wada. K. 1990. Initiation of mineralization in bivalve molluscs. Pp. 79-92 in The Mechanisms of Biomineralization in Animals ami Plants. M. Omori and N. Watabe. eds. Tokai University Press. Tokyo. Wada, K. and T. Fujinuki. 1976. Biomineralization in bivalve molluscs with emphasis on the chemical composition of the extrapallial fluid. Pp. 1 75-190 in The Mechanisms of Mineralization in the Invertebrates and Plants. N. Watabe and K. M. Wilbur, eds. University of South Carolina Press. Columbia. Watabe, N. 1965. Studies on shell formation. XI. Crystal-matrix rela- tionships in the inner layers of mollusk shells. J. Ultras/met. Res. 12: 351_370. Watabe, N. 1974. Crystal growth of calcium carbonate in biological systems. J. Cryst. Growth 24/25: 1 16-122. Watabe, N. 1981. Crystal growth of calcium carbonate in the inverte- brates. Prog. Crysl. Growth Charact. 4: 99-147. Watabe, N., and K. M. Wilbur. 1961. Studies on shell formation. IX. An electron microscope study of crystal layer formation in the oyster. J. Biophys. Biochem. Cytol. 9: 761-771. Watabe, N., D. G. Sharp, and K. M. Wilbur. 1958. Studies on shell formation. VIII. Electron microscopy of crystal growth of the nacreous layer of the oyster Crassostrea virginit-u. J. Biophys. Biochem. Cytol. 4: 281-284. plates 152-156. Weiner, S., and L. Addadi. 1997. Design strategies in mineralized biological materials. J. Mater. Chem. 1: 689-702. Wen. H. B., J. Moradian-Oldak, W. Leung, P. Bringas Jr., and A. G. Fincham. 1999. Microstructures of an amelogenin gel matrix. J. Struct. Bioi. 126: 42-51. Wheeler, A. P., and L. P. Koskan. 1993. Large scale thermally s\ nthe- sized polyaspartate as a biodegradable substitute in polymer applica- tions. Mater. Res. Soc. Synip. Proc. 292: 277-283. Wheeler, A. P., and C. S. Sikes. 1989. Matrix-crystal interactions in CaCO, biomineralization. Pp. 95-133 in Biomineralization: Chemical and Biochemical Perspectives. S. Mann. J. Webb, and R. J. P. Wil- liams, eds. VCH Publishers. Weinheim. Germain Wheeler, A. P., and K. M. Wilbur. 1977. Shell growth in the scallop 66 C. S. SIKES T .4Z. Argapeclen im.. Besses of shell growth. J. Molluscan Stud. 43: 155-16 \\hevk-r. \. r ' kddrr. and K. M. \\ilhur. 1975. Shell growth in .rradians. \. Isotope incorporation I growth. Bitil. Bull. 148: 472-482. \\hi-vler. \ I V. KiiM'iik... D. M. Swift, and C. S. Sike.s. 1988. Reti Tmatmn h\ tractions of oyster shell matrix. Mar Wheeler. \ I'.. K. i . I ... and < . S. Sikes. 1991. CaCO, crystal- hindinj; pn>pertieMi| peptides and ihcir influence on crystal growth. Pp. f Reactive Peptides ami l',il\mt-r\- /)M.U/\ unJ Commercialization, Symposium Series 444. C. S. Sikes and A. P. Wheeler, eds. ACS Books. Washington. DC. Wier/bicki. A.. C. S. Sikes. B. Drake, and .1. Madura. I'W4. Atomic force microscopy and molecular modeling of protein and peptide binding to calcite. Calcif. Tissue Int. 54: 133-141. \\ilhur, K. M.. and \. \l. Kernhardl. 1984. Effects of amino acids, magnesium, and molluscan extrapallial fluid on cryslalli/ation of cal- cium carbonate: in vitro experiments. Biol. Bull. 166: 251-259. \\ilhur. k. M., and L. H. Jodrey. 1952. Studies on shell formation. I. Measurement ot the rate of shell formation using Ca js . Biol. Bull. 103: 269-276. Reference: Bio/. Bull. 198: 67-76. (February 20(10) The Apical Sensory Organ of a Gastropod Veliger Is a Receptor for Settlement Cues MICHAEL G. HADFIELD 1 -*, ELLA A. MELESHKEVITCH 1 , AND DMITRI Y. BOUDKO 1 1 Kewalo Marine Laboratory. University of Hawaii. 41 Ahui Street. Honolulu. Hawaii 96813; ami 2 Department ofZoolog\. University of Hawaii at Manoa. Honolulu. Hawaii 96822 Abstract. On the basis of anatomy and larval behavior, the apical sensory organ (ASO) of gastropod veliger larvae has been implicated as the site of perception of cues for settlement and metamorphosis. Until now, there have been no experimen- tal data to support this hypothesis. In this study, cells in the ASO of veliger larvae of the tropical nudibranch Phestilla sibogae were stained with the styryl vital dye DASPEI and then irradiated with a narrow excitatory light beam on a fluo- rescence microscope. When its ASO cells were bleached by irradiation for 20 min or longer, an otherwise healthy larva was no longer able to respond to the usual metamoiphic cue. a soluble metabolite from a coral prey of the adult nudibranch. The irradiated cells absorbed the dye acridine orange, suggest- ing that they were dying. When larvae not stained with DAS- PEI were similarly irradiated, or when stained larvae were irradiated with the light beam focused on other pans of the body, there was no loss of ability to metamorphose. Together these data provide strong support for the hypothesis. Potassium and cesium ions, known to induce metamorphosis in larvae of many marine-invertebrate phyla, continue to induce metamor- phosis in larvae that have lost the ability to respond to the coral inducer due to staining and irradiation. These results demon- strate that ( 1 ) the ASO-ablated larvae have not lost the ability to metamorphose and (2) the ions do not act only on the metamorphic-signal receptor cells, but at other sites down- stream in the metamorphic signal transduction pathway. Introduction Abundant data demonstrate that most invertebrate larvae succeed in locating appropriate habitats for settlement, metamorphosis, and growth by responding to site-specific Received 1 July 1999: accepted 3 November 1999. * To whom correspondence should be addressed. E-mail @hawaii.edu hadrield chemical cues associated with conspecific individuals, req- uisite prey, microbial films, or algal or other benthic sub- strata (Crisp, 1974, 1984: Pawlik, 1992: Hadfield, 1998). Much current research on the subject of larval settlement focuses on the chemical identity of settlement cues and the signal-transduction mechanisms by which external cues stimulate morphogenetic transformations in the larva (e.g., see papers in Biofoiding 12( 1 ), 1998). Despite more than 50 years of research on metamorphic induction in marine- invertebrate larvae, experimental definition of the exact location on the larval body where these interactions take place that is, where the chemoreception that results in site-specific larval settlement occurs is lacking for most groups. The anterior pole of at least some cnidarian planulae includes a region that must contact stimulatory surfaces for metamorphosis to occur (Miiller el ai. 1977: Freeman and Ridgway, 1987), and there is good evidence that the anten- nules of barnacle cyprid larvae are the location of receptors for settlement inducers (summarized by Clare, 1995). How- ever, the site of induction has eluded strong inference for the trochophores and trochophore-derived larvae of large ma- rine-invertebrate clades. Larvae of many phyla behave prior to settlement in a manner that indicates that they are "testing" or "sampling" substrata. For larvae of groups as diverse as phoronids, polychaetes, and chitons, all of which possess an apical tuft of elongate and somewhat stiff cilia, presettlement behavior typically includes swimming near the substratum with their apical ends downward so that the apical tuft brushes or is pressed against the substratum (Barnes and Conor. 1973; Nott. 1973; Zimmer, 1991; unpublished personal observa- tions on the serpulid polychaete Hydroides elegans). Thus the apical tuft, together with its underlying cells, has long been suspected to be the site of detection of substrate- associated cues for settlement. As others have, we refer to 67 68 M C HUM HIP / / M.. the apical tuft and its associated cells as the apical sensory organ, henceforth abbreviated ASO. Larvae of opisthobranch gastropods bear a probable sen- sory structure, derived developmental!) and evolutinarily from the ASO of trochophore larvae (Bonar. 19~Si. Recent papers by Kempf etui. ( 1997) and Marois ami (.'aiv\\ i 1997) provide elegant details concerning the cellular o nposition of the ASOs of opisthobranch larvae Thi^ : opistho- branch larvae consists of a set of cili. beat mg receptor cells. some of which send axons direciK mu< the cerebral com- missure of the brain: neurons fro \SO cells inner- vate the velum. Bonar (19"- how well situated the structure is, in veliger larva..- >>! the nudihranch Phesiilla sibogae Bergh 1905. to detect the water-borne chemical cue that arises from the nudihranch'x coral prey and induces metamorphosis. An .W > IMS been demonstrated in proso- branch larvae as well .1 the. 1995). where Leise <1996i reported that it can he stained with DASPEI. a styryl Hun rescent dye. The only previous experimental data supporting the hy- pothesis that the ASO is the site of cell-surface receptors tor inducers of metamorphosis in invertebrate larvae were pro- vided b> Baxter and Morse ( 1992) for larvae of the gastro- pod Haliotis rufescens. Their experiments revealed that receptors tor lysine. a compound known to modifv the effects of a metamorphic inducer for which GABA is a receptor agonist, lie on cilia harvested from the larvae of //. micM-cin: among these cilia were those from the apical tuft. Larvae of the nudibranch Om-luiloris hildincllata were re- ported to detect a water-borne cue from the barnacle prey of the adult nudibranch via a pair of lateral propodial "gan- glia": however, the barnacle factor induces only reversible settlement behavior, but not metamorphosis, which is de- pendent on a surface-bound cue (Arkett . Because all authors who have described details of ASOs in gastropod larvae have argued for a probable role for this ni-jan in induction of settlement and metamorphosis, it is timely to perform robust experimental tests of the hypoth- esis. In the research presented here, the vital dvc DASPEI, known to vitally stain mitochondria (Haugland. 1996) and thus mitochondria-rich sensory cells (Bereiter-Hahn. 1976; Nurse and Faraway. 1989: Balak et al.. 1990: Leise. 1996: Moudko ct ul.. 1999). was employed to vilallv stain the ASO in veliger larvae of l'lic\lil/n \ihn^(ic. followed by photo- ablation of the stained organ by fluorescent excitation Balak el al. (1990) reported that labeling lateral-line hair cells of amphibians with DASPLI and exposing them (o cpifluorscent illumination at 450-490 nm resulted in cell death due to pi millions arising from photoexei- tation of DASH I We attempted to ol-.i.iin additional evidence of cell death in the ASO by applying the \iial d\c acrulme orange to treated larvae. Acridine orange, long used as a fluorchrome indicator of cell death, becomes intercalated into uncoiled I )\.\ of dying cells when applied at relatively low (i.e., en. 10 '' M) concentrations (Delic el al.. 1991 ). Although acri- dine orange can stain RNA. as well as DNA. the red-orange (650 nm) emission color of the RNA-acridine-orange complex distinguishes it from DNA-acridine-orange emis- sion, which is green (525 nm) (Haugland. 1996). Although some experimentalists have reported acridine-orange stain- ing to be specific to cells undergoing apoptosis (e.g.. Abrams el al.. 1993). others report similarity of staining between different forms of induced cell death, at least in later stages (reviewed by Dar/ynkiewic/ and Traganos. 1998). Among the major tools used to study metamorphosis in marine invertebrates has been a growing list of so-called artificial inducers (reviewed by Crisp. 1974. 1984: Pawlik. 1992: Hadtield. 199S). The most useful of these are the cations potassium and cesium, which induce metamorphosis in larvae from seven pin la in the absence of other stimuli (summari/ed by Herrmann. 1995: Woollacott and Hadlield. 1996). While noting that the entire larval bodies are bathed in elevated potassium or cesium when larvae are treated with these ions, some authors have proposed that K ' or Cs + act bv depolari/ing the sensory cells that typically bind the natural inducer and thus initiate spikes in neurons extending from those cells (e.g., Baloun and Morse. 1984: Yool ct ul.. 19X6; I.eit/ and Klingmann. 1990; Herrmann. 1995: Wool- lacott and Hadtield. 1996; Carpi/o-Ituartc and Hadlield. I99S). The possibilitv that K or C's acts on the entire nervous system (Todd et ai. 1991 ). directly on target tissues (Yool -100% meta- morphosis within 24 h of exposure to coral water were used in the experiments. To test that the treatment effects (i.e.. DASPEI staining plus irradiation of the stained ASO) were those responsible for observed results: ( 1 ) larvae were stained with DASPEI but not exposed to 450-490 nm light; (2) unstained larvae were exposed to irradiation focused on the ASO for 20 min: and (3) stained larvae were exposed to irradiation focused on regions of the larvae body other than the ASO. In all of these controls, the larvae were subjected to the same treatments as the experimental larvae, including having their larval shells decalcified and being embedded in low-melting-point agarose. Acridine orange staining to detect cell death. Untreated control larvae (also decalcified and embedded in agarose) and larvae that had been stained with DASPEI and exposed to radiation focused on the ASO were placed in a seawater solution of acridine orange (2.7 X 10~ 6 M) for 20 min. After exposure to acridine orange, the larvae were examined and photographed on a Zeiss fluorescence microscope 70 M. G. HADRELD ET AL Decalcification pH 5 6 FSW Labeling 5% DASPEI FSW Immobilization Photoablation 1%Agar FSW Aperture Bandpass filter 428 LA/ Light Induction Cocal Inducer & FSW experimental 7 Cora! Inducer & FSW Figure 1. Diagram representing the methods used in photoablalion of cells in the apical sensory organ of veliger larvae of l'ln:\iillu w/>,pu ill\ 5 o) of stained cells are consistent with their identity as components of the ASO. Unstained larvae displayed no uniolliiorescencc in the ASO. The large prototrochal cells ol UK \ela stained lightly with DASPEI. and various structures in the visceral hump, exposed In decalcitication of the larval shell, also absorbed DASPEI. but these regions were easily excluded from the irradiated area. DASPEI was retained by larvae in FSW for periods in excess of one week, even through metamorphosis. Larvae that hail been stained with DASPFI. embedded in low-melting-point agarose. and irradiated with a narrow hand of 450-490 nm light focused on the ASO survived well (>907r). When released from agarose and exposed to coral seawatcr. the typical metamorphic inducer. these lar- \ac metamorphosed in percentages inversely proportional to the duration of irradiation (Fig. 3). The excitatory illumina- tion resulted in photobleaching of the ASO cells; thai is. they became colorless. DASPEI-stained cells in non-illumi- nated parts of the larvae continued to show brilliant II no rescence (Fig. 5A). Very few larvae that hail been exposed for as long as 20 nun were capable of metamorphosing in response to natural inducer. Excitatory illumination alone dul not reduce the ability of larvae to metamorphose in response to coral inducer (Fig. 3. differences in responses in the control labeled P20-CI were not significantly different [P > .19()| from (hose in the control labeled C'l). DASPEI- stained larvae exposed to the irradiation beam focused on the larval foot instead of on the ASO retained full compe- tence to metamorphose when exposed to the coral inducer (Fig. 4). There were no DASPEI-stained cells in the region SETTLEMENT-CUE PERCEPTION IN VELIGERS 71 Figure 2. Veliger larvae of Phi:\lilln vhtigM- stained with DASPEI. (top) Bright-field image, (middle) Same specimen photographed with fluo- rescent light, (bottom) Enlargement of anterior region of the same speci- men showing DASPEI staining in 5-6 cells of the apical sensory organ. Scale bars, top and middle = 50 /im. bottom = 25 yum. irradiated. We conclude that exposure of the stained ASO to intense excitatory irradiation for 20 min resulted in photo- ablation of the ASO cells, and that subsequent loss of the ability of these larvae to respond to the metamorphic in- ducer contained in coral seawater is evidence that the irra- diated cells were the site of receptors for the inducer. These larvae were otherwise unharmed. They continued to swim normally in FSW. and they had not lost the capacity to metamorphose in response to artificial inducers (see below). The yellow DASPEI fluorescence disappeared from the ASO following excitatory illumination, although it persisted in other regions of the larval body. When larvae whose ASO cells had been stained with DASPEI and exposed for 20 min to fluorescent irradiation and thus photobleached were placed in a seawater solution of acridine orange, the ASO cells uniquely absorbed the dye and emitted a green fluorescence (Fig. 5). Comparable stain- ing in other cells and organisms has been found specific to induced cell death (Delic et a!.. 1991; Abrams et al., 1993). ASO cells of untreated larvae did not take up acridine orange. These observations support the conclusion that DASPEI staining followed by fluorescent irradiation of the ASO cells led to their ablation. When larvae that had undergone the photoablation treat- ment described above were exposed to seawater containing 20 mM cesium ion for 20 min and then transferred to FSW. they underwent normal metamorphosis in large numbers within 24 h (Fig. 6). Similarly treated larvae exposed to seawater containing 20 mM excess potassium also meta- morphosed in numbers much greater than controls, reaching more than 50% after 48 h (Fig. 6). The latency of the metamorphic response of larvae of Phestilla sibogae to potassium ion has been reported previously (Pechenik et al., 1995). Control exposures of treated larvae to coral inducer demonstrated that they were, as in the experiments de- scribed above, unresponsive. Low percentages of larvae that metamorphosed after photoablation without inducer (Fig. 6, C-2) or with inducer (Fig. 6, CI in Photoablation bracket) were not significantly different from those in untreated larvae (Fig. 6. CI) (t tests, P 0.05). We conclude that the site of action of K + and Cs + in inducing metamorphosis is not on the primary receptor cells, which had been destroyed in the experimental treatment. However, the possibility of multiple sites of metamorphic stimulation by these cations is not eliminated. Discussion The data presented here provide compelling evidence that cells in the apical sensory organ bear the receptors for the dissolved molecular inducer of settlement and metamorpho- sis in larvae of the gastropod Phestilla sibogae. Evidence presented by Wodicka and Morse (1991) and Baxter and Morse (1992) demonstrated that receptors for a related receptor pathway are found on cilia harvested from compe- tent larvae of another gastropod. Huliotis rufescens. Yool (1985) had earlier demonstrated that these larvae have an apical ciliary structure that is presumably homologous to the ASO of P. sibogae. It remains to be demonstrated if the key pathway, that for the settlement/metamorphic inducer (for which GABA is an agonist), is also on the same cilia in larvae of H. rufescens. Apical sensory organs have been demonstrated in all opisthobranch larvae that have been examined appropriately (Bonar, 1978; Chiaand Koss, 1984; Kempf et al., 1997; Marois and Carew, 1997). as well as in two prosobranch species (Uthe, 1995; Leise, 1996). While it is logical to assume that ASOs in other gastropods have a role in metamorphic induction, experimental evidence for this sensory function has not yet been provided. 72 M. Ci HADFIELD tT .\L 100 - T T 90 - ;x 1 % Metamorphosis to UJ J< O\ -J OO ~ ~ ~ oooo .- :: 1 I !%'' :: " 1 ~T 0. , , | | [~~ 1 , , , , , 1 ' C CI 1 5 10 15 20 DAS P20 P20 DASPEI + Photoexposure (min) C C CI Coral Inducer Figure 3. Percent metamorphosis in larvae of Phestilla sibogae that have been stained with DASPEI. exposed to fluorescent illumination focused on the stained cells in the apical sensory organ, and exposed to coral metamorphic inducer: bars = mean and SD. Experimental and control larvae embedded in agarose (n = 3 replicates. 15 larvae/replicale): DASPEI + Photoexposure. DASPEI-slained larvae exposed to a beam of fluorescent light focused on stained cells of the apical sensory organ for 0-20 min and then placed in coral inducer for 24 h; DAS-C. DASPEl-stained larvae that were not photo-treated nor exposed to coral inducer: P20-C'. unsiamed larvae exposed to fluorescent light for 20 min and not placed in coral inducer; P20-CI. unstained larvae exposed 10 lluuiescent light for 20 min and then placed in coral inducer. CI. untreated larvae exposed to coral inducer for 24 h. Controls with larvae nut stained nor embedded in agarose (n = 3 replicate-,. 25 larvae/replicate): C. control to determine that the hatch ol larvae did not show high levels of spontaneous metamorphosis; CI. control to determine that the hatch u! lai\.ie metamorphosed normally upon exposure to the coral inducer without other treatment. P20-CI was not significantly different from CI U test. P > 0.140). The homology of key elements of the apical sensory organ of gastropod veligers. and presumably those of other molluscan larvae (Raven. 1966. pp. 143-145). with the apical -tuft/apical-organ complex of the trochophore larvae of polychaetous annelids appears to be sound. The cell lineage for the ASO of larvae of the gastropod Crepitiulu fornicata, described by Conklin ( 1897). appears to be iden- tical to that for the apical organ of polychaete trochophores; in both larval types, the apical organs arise from descen- dants of the first-quartet micromeres la 1 Id' (summari/ed by Kume and Dan. 1968). Cells bearing apical-tuft cilia appc.ii ii .n isc from the same microtnere quartet in polyclad flatworms tSelenka. 1881) and nemerteans (Horstaditis. 1937) that produce pelagic larvae. There thus appears to he- great antiqun apical sensory organs of larvae of spiralian meta/<> st cases. The chemical ilnit trigger metamorphosis in larvae ot Phestilla sihi>%ae. . ; HIK- other gastropods, and a number of other invertchraic types are dissolved in seawater (nudibranchs: Thompson, I95X: lladheld and Scheuer. 1985: Bahamondes-Rojas and Dherbonez. 1990: Lambert and Todd. -1994: prosobranchs: Scheltema. 1961: McCiee and Targett. 1989: Boettcher and Targett. 1998; bivalves: Zimmer-Faust and Tamburri. 1994; sipunculans: Rice. 1986; echinoids: Burke. 1984; Pearce and Scheihling. I99()a. I990b; ascidians: Young and Braithwaite. l l (); barnacles: Rittschof. 1985; crabs: Welch el ^iw that were stained with DASPEI, embedded in low-melting-point agarose, ex- posed for 20 mm to fluorescent irradiation aimed either at the apical sensory organ (ASO) or the foot, freed from the agarose. and then placed in coral inducer for 24 h (n = 3 replicates. 15 larvae/replicate!: bars = mean and SD. ASO, fluorescent light locused on the ASO; Foot, fluores- cent light focused on the foot. Not embedded in agarose were larvae used in C, untreated controls; CI, untreated larvae exposed to coral inducer (H = 3 replicates, 25 larvae/replicate). lished personal observations on larvae of H\dnudes el- egans). larvae of a chiton (Barnes and Conor. 1973). and veligers of the abalone Haliolis rufescens (Wodicka and Morse, 1991 ) apparently detect absorbed settlement cues by brushing the surfaces of substrata with their apical ciliary Figure 5. (A) Veliger larva of Phestilla sihagae that was stained with DASPEI. embedded in low-melting-point agarose. and subjected to exci- tatory irradiation focused on the apical sensor) organ (ASO) for 2(1 nun. The ASO cells are bleached (compare to Fig. 2B. C). although the DASPEI stain remains in other parts of the larval body. ( B ) The same larva after 20-min immersion in acridine orange in filtered seawater. The ASO cells emit a pale green fluorescence. Other bright areas, some not visible in both photos, retain DASPEI stain. Scale bars = 100 ju.m. Photoablation (20 min) Figure 6. Percent metamorphosis in larvae of Phestilla sihogac that had undergone photoablation of apical-sensory-organ (ASO) cells and then exposure to the ionic inducers potassium or cesium; bars = mean and SD. Percent metamorphosis was determined after 24 h in all cases, plus at 48 h for potassium-treated larvae. Experimental treatments (larvae embedded in low-melting-point agarose for light treatment; n = 3 replicates. 15 larvae/ replicate) in Photoablation (20 min) bracket, all stained with DASPEI and subjected to 20-min excitatory illumination focused on the ASO: C-2. larvae not placed in coral inducer; CI, larvae exposed to coral inducer; Cs + , larvae exposed to 20 mM cesium chloride in seawater for 20 min; K*. larvae continuously exposed to seawater augmented with 20 mM excess potassium chloride. Controls (not embedded in agarose): C-l. untreated larvae not exposed to coral inducer; CI. untreated larvae exposed to coral inducer for 24 h (n = 3 replicates. 25 larvae/replicate). Neither C-2 nor CI with photoablation treatment were significantly different from the untreat- ed-larvae control C-l d tests. P 58> 0.05). tufts during a period before settlement and attachment. We presume that during what has been interpreted as pre-attach- ment "searching behavior," the larvae apply chemically sensitive cilia to potential settlement sites in a manner that will bring together stimulatory ligands and their specific receptors and initiate neurological signaling for settlement and metamorphosis. Because cilia protrude very little, if at all. from the sensory cells in the apical sensory organ of veligers of P. sibogae, they would be of little use in contact chemoreception; additionally, the cilia of these cells are structurally similar to stereocilia as contrasted with motile cilia such as those found in other sensory organs (Bonar, 1978). The same is true of the apical sensory cells of veligers of the prosobranch Littorimi littorea (Uthe. 1995). Kempf et al. (1997) provided detailed ultrastructural and immunocytochemical evidence for the presence of at least three cell types in the apical sensory organs of nudibranch larvae. Three sensory serotonergic neurons are intricately associated with the innervation of the velum, leading the 74 M. G. HADI-'ltl.D ET .\L authors to conjecture that these elements of the ASO are mechanosensory and serve as a compensatory system ot velar control to "modulate the position of the velar lohes in response to deformations of the pretrochal surface Caused by changes in velar orientation." It appears most likely ihat the cells responsible for chemosensory detection ol the metamorphic cue in larvae of Pliextilltt \ihogae are the so-called flask-shaped cells (Bonar. ampullary cells of Chia and Koss. 1984). The . i number (five or six in veligers of P. sibogae accor.mii: to Bonar' s |1978] ultrastructural study) and location with those that were stained by DASPEI in the current study (Eig. 2). and cells with very similar structure ha\e been shown to be chemo- sensory in a cephalopod d.ucero <; us system to bring about morpho- genesis (e.g., Baloun ;mil Morse. 1984). Cesium, acting as a potent blocker ol pu- . sium channels, has been thought to have a similar action (e.g., Carpi/o-Ituartc and Hadlield. 1998). However, nearly all those who have carried out such experiments have acknn , dg d tli.it entire larvae are bathed in the seawater with elevated potassium or cesium ions and that the ions could be acting downstream from the primary chemosensory cells (Todd ct ai. 1991 ). We have presented here data consistent with the latter hypothesis, although exactly where downstream awaits clarification. The possibility thai K and Cs act on external receptors as well as downstream sites in intact larvae is. of course, not disproven by these experiments. There are undoubtedly numerous synapses in the central nervous system between the ASO and the responding tissues where potassium and cesium could act. Because it focused on a single gastropod species, the current research provides no support for homology of ASO cells among gastropods or across phyla. However, it does point to profitable future research on this subject. ( I ) Elec- tron-microscopic study of the ASOs of larvae of P. sihogae after the photoablation treatment could reveal exactly which cells were affected, and thus confirm that the flask-shaped cells with nonmotile cilia are the critical receptor cells in metamorphic signaling. (2) If the experiments described here were performed with larvae of other species, across the Mollusca and other spiralian phyla, and similar results ob- tained, the data would be further evidence for homology. The new data presented here, considered together with experimental data on the molecular attributes of metamor- phic-cue receptors in larvae of the abalone (Baxter and Morse. 1987) anil the polychaete Hydroides eleganx (Carpi/o-ltuarte and Hadtield. 1998: Holm et <;/.. 1998). provide a framework for investigating the molecular nature of the receptor molecules used by these larvae in settlement- cue perception and the homology of these receptor mole- cules across phyla. Data on the abalone. H. elegant, and P. \il>in;(i(' (Hadfield. unpubl. obs.) strongly suggest that these receptor molecules are not members of the G-protein-trans- duced. 7-transmembrane-domain peptides typical of much chemical perception in animals. Once the specific location of the actual receptor molecules is demonstrated across phyla, and their molecular sequences are known, molecular probes (e.g.. /// \itii hybridi/ation) could be employed to provide additional evidence for homology and phylogcny. Similar experimental data obtained Irom other phyla, espe- cially the noncoelomate spiralians. would greatly expand our understanding of the evolution of the ASO complex and. perhaps, evolutionary relationships among the spiralians, Acknowledgments This research was supported by ONK grants NOOOI4-94- 1-0524 and N()()OI4-95-l-l()15. Many members of our re- search group at the University of Hawaii's Kewalo Marine Laboratory contributed significantly to maintaining research animals, culturing larvae, and assisting in the research at important moments; we thank them all. We thank Caiole Hickman (University of California. Berkeley) for her con- SETTLEMENT-CUE PERCEPTION IN VELIGERS 75 structive criticisms and suggestions for improving the manuscript, Richard Strathmann (University of Washing- ton) for assisting the authors in recalling older references on larval settlement behavior, and two anonymous reviewers for suggestions that clarified significant parts of this paper. Literature Cited Abrams, J. A., K. White, L. I. Fessler, and H. Steller. 1993. Pro- grammed cell death during Drusophila embryogenesis. Development 117: 29-43. Arkett, S. A., F.-S. Chia, J. I. Goldberg, and R. Koss. 1989. Identified settlement receptor cells in a nudibranch veliger respond to specific cue. Biol. Hull. 176: 155-160. Bahamondes-Rojas, I., and M. Dherbonez. 1990. Purification partielle de substances glyco-conjuguees capahles d'induire la metamorphose des larves competentes d' Eubranclius doriae (Trinchese, 1X7), mol- lusLjue nudibranche. J. E.\p. Mar. Biol. Ecol. 144: 17-27. Balak, K. J., J. T. Corwin, and J. E. Jones. 1990. Regenerated hair cells can originate from supporting cell progeny: evidence from pho- totoxicity and laser ablation experiments in the lateral line system. / Neiimxci. 10: 2502-2512. Baloun. A. J., and D. E. Morse. 1984. Ionic control of settlement and metamorphosis in larval Huliotis rufexcens (Gastropoda). Biol. Bull. 167: I24-I3S. Barnes, J. R., and J. J. Gonor. 1973. The larval settling response of the lined chiton Tonicella linculii. Mar. Biol. 20: 259-264. Baxter, G., and I). E. Morse. 1987. G protein and diacylglycerol reg- ulate metamorphosis of planktonic molluscan larvae. Proe. Null. Acail. Sci. USA 84: 1867-1870. Baxter, G. T., and D. E. Morse. 1992. Cilia from abalone larvae contain a receptor-dependent G protein transduction system similar to that in mammals. Biol. Hull. 183: 147-154. Bereiter-Hahn, J. 1976. Diniethylaminostyryl-methyl-pyridinium-io- dine (DASPEI) as a fluorescent probe of mitochondria in situ. Biocliim. Bin/thy.*. Acta 423: 1-14. Boettcher. A. A., and N. M. Targett. 1998. Role of chemical inducers in larval metamorphosis of queen conch. Strombus gigus Linnaeus: relationship to other marine invertebrate systems. Biol. Bull. 194: 132-142. Bonar, I). B. 1978. Ultrastructure of a cephalic sensory organ in larvae of the gastropod Phestilla sibogae (Aeolidacea. Nudibranchia). Tissue Cell 10: 153-165. Boudko, D. Y., M. Switzer-Uunlap, and M. G. Hadneld. 1999. Cel- lular and subcellular structure of anterior sensory pathways in Phcxiilla xibague (Gastropoda, Nudibranchia). J. Coinp. Neural. 403: 39-52. Burke, R. I). 1984. Pheromonal control of metamorphosis in the Pacific sand dollar. Dendraster excentricus. Science 225: 442-443. Carpizo-Ituarte, E. J., and M. G. Hadfield. 1998. Stimulation of meta- morphosis in the polychaete Hydroides elfgans Haswell (Serpulidae). Biol. Bull. 194: 14-24. Chia, F.-S., and R. Koss. 1984. Fine structure of the cephalic sensory organ in the larva of the nudibranch Rostanga pulchra (Mollusca, Opisthobranchia, Nudibranchia). Znomorphology 104: 131-139. Clare. A. S. 1995. Chemical signals in barnacles: old problems, new approaches. Pp. 49-67 in New Frontiers in Barnacle Evolution, F. R. Schram and J. T. Hoeg, eds. Balkema, Rotterdam. Conklin, E. G. 1897. The embryology of Crepidulu. J. Marpliol. 13: 1-226. Crisp, D. J. 1974. Factors influencing the settlement of marine inverte- brate larvae. Pp. 177-265 in Chemoreceprion in Marine Orgcinismx. P. T. Grant and A. M. Mackie, eds. Academic Press, New York. Crisp, D. J. 1984. Overview of research on marine invertebrate larvae. 1940-1980. Pp. 103-126 in Murinc Biodeterioration: . Cell Kes. 194: 147-153. Freeman, G., and E. B. Ridgway. 1987. Endogenous photoprotems. calcium channels and calcium transients during metamorphosis in hydro/.oans. Kou.\'s Arch. /Vr. liiol. 196: 30-50. Hadfield, M. G. 1978. Metamorphosis in marine molluscan larvae: an analysis of stimulus and response. Pp. 165-175 in Settlement ami Metamorphosis of Marine Invertebrate Larvae. F.-S. Chia and M. E. Rice, eds. Elsevier, New York. Hadfield, M. G. 1998. The D. P. Wilson Lecture: research on settlement and metamorphosis of marine invertebrate larvae: past, present and future. RiofoulinK 12: 9-29. Hadfield, M. G., and J. T. Pennington. 1990. The nature of the meta- morphic signal and its internal transduction in larvae of the nudihranch Phestilla sibogae. Bull. Mar. Sci. 46: 455-464. Hadfield, M. G., and I). Scheuer. 1985. Evidence for a soluble meta- morphic inducer in Phexiilla: ecological, chemical and biological data. Bull. Mar. Sci. 37: 556-566. Haugland, R. P. 1996. Handbook of Fluorescent Probes and Research Chemicals, 6th edition. Molecular Probes, Eugene, Oregon. Herrmann, K. 1995. Induction and regulation of metamorphosis in planktonic larvae: Phoronis niiilleri (Tentaculata) as archetype. Helgol. Meeresunters. 49: 255-281. Holm, E. R., B. T. Nedved, E. Carpizo-Ituarte, and M. G. Hadfield. 1998. Metamorphic-signal transduction in Hydroides eleguns (Polychaeta: Serpulidae) is not mediated by a G protein. Biol. Bull. 195: 21-29. Horstadius, S. 1937. Experiments on determination in the early devel- opment of Cerebniiulux luctcns. Biol. Bull. 73: 317-342. Jensen, R. A., D. E. Morse. R. L. Petty, and N. Hooker. 1990. Arti- ficial induction of larval metamorphosis by free fatty acids. Mar. Ecol. Prog. Ser. 67: 55-7 1 . Kempf, S. C., L. R. Page, and A. Pires. 1997. Development of seroto- nin-like immunoreactivity in the embryos and larvae of nudihranch mollusks with emphasis on the structure and possible function of the apical sensory organ. J. Camp. Neurol. 386: 507-528. Kume, M., and K. Dan. 1968. Invertebrate Embryology. English trans- lation by Jean C. Dan. National Library of Medicine, Public Health Service, U.S. Dept. of Health, Education and Welfare and National Science Foundation, Washington, DC and NOLIT Publishing House, Belgrade. Lambert, W. J., and C. D. Todd. 1994. Evidence for a water-borne cue inducing metamorphosis in the dorid nudibranch mollusc Aduluriu proxima (Gastropoda: Nudibranchia). Mar. Biol. 120: 265-271. Leise, E. M. 1996. The fluorescent dye DASPEI selectively stains the apical ganglion in larval Ilyanassu obsoleta (Mollusca. Gastropoda). Microsc. Res. Tech. 33: 496-500. Leitz, T., and G. Klingmann. 1990. Metamorphosis of Hydractiniu: studies with activators and inhibitors aiming at protein kinase C and potassium channels. Roux's Arch. Dev. Biol. 199: 107-113. Lucero, M. T., F. T. Horrigan, and W. F. Gilly. 1992. Electrical responses to chemical stimulation ol squid olfactory receptor cells. J. Exp. Biol. 162: 231-249. Marois, R., and T. J. Carew. 1997. Fine structure of the apical ganglion and its serotonergic cells in the larva of Aplyxia califomica. Biol. Bull. 192: 388-398. McGee, B. L., and N. M. Targett. 1989. Larval habitat selection in 76 M. G. HADFIELD ET AL Crepidula (L.I and ils el feel on adull distribution patterns. J. >/' Mar. Biol. Ecol. 131: 195-214. Milk-r. S. K_ and M. (i. Hadlifld. IIS*.. Ontogeny ol phocouxi- .UK! nxxphK competence in larvae of the nudibranch PhestilUi sihogat ! (Gastropoda: Opisthobranchia i ./ l.y l/.ir. flnrf. & MiilltT. \V. A.. A. Mit/e. J. I', \\ii-khin .ml II. M. Meier-Menjje. I 'l"" 1 . I'.iljr morphogenesis in eat' did ile\elopment: action of caesium, of ncurotransmitter- nmnsic head activator on pattern formation. Roux's Arch. D> 182: ; ll ; -s> \otl.J.A.1973. Settlement ol 'the of Spirorbis spirorbis. J. \t. Mm: Hinl. Annii. Rev. 30: 273-335. Pearce. C. M.. and K. K. ScheihlinK. 1990a. Induction of metamorpho- sis of larvae of the jicen sea urchin. Strongylocentrotus droebochiensis by coralline red algae. //. //. 179: 304 311. Pearce, C. M., and R. K. Schi-ihliiij>. 19901). Induction of settlement and metamorphosis in the sand dollar Echinarachinius parma: evidence lor an adult-associated factor. Mar. Biol. 107: 363 V>'> Pechenik. J. A.. M. (i. Hadheld. and I.. S. K>sli-r. 1995. Assessing whether larvae of the opisthobranch gastropod Pheslilla sihufiiie Bergh become responsive to three chemical cues at the same age. J. L\/>. Mar. Biol. Ecol. 191: 1-17. Penninuton. .). T.. and M. (i. lladlic-ld. 1989. A simple nontoxic method for deealcitk.ition ol living invertebrate larvae. ./. / >/' Mur H,,>l. Ecol. 130: 1-7. Pires. A., and M. G. Hadlifld. 1993. Responses of isolated vela of nudibranch larvae to inducers of metamorphosis. ./. /.'./>. Ziiol. 266: 234-239. Pi rts. A., S. L. Coon, and M. (i. Hadlifld. 1997. Calecholainincs and dihydroxyphenylalanine in metamorphosing larvae ol the nudihranch Pheslilla sibogae Bergh. J Ci-iii/>. /'/nwW. .A 181: 1X7 I'M. Raven. C. P. 1966. \torpho nesis: The Analysi* "I Mi>llu\ftm /)/ i. / opment. 2nd ed. Pergamon. Oxford. Rite. M. K. 1986. Factors influencing larval metamorphosis in <;,'///;/i>< misakiana (Sipuncula). Bull. Mar. Sci. 39: 362-375. Kitlschol. I). 1985. Oyster drills and the frontiers of chemical ecology- unsettling ideas. .Am. Malaciil. Bull. Special Ed. 1: 111-116. Scheltema. R. S. 1961. Metamorphosis of the veliger larvae ofNassarius nh\i>leiii\ ( Gastropixla ) in response to bottom sediment. Binl. Bull 12(1: 92-109. Selvnka. K. 1881. Zoologische Snulien. II. Zur EntwicUungsgeschichU tier Seeplanarien. Ein Beilrag ~ur Keimblaltlehre and De:enilenz Thto- rie. Leip/ig, 36 pp. 1 hompson. I. !'.. 1958. The natural history, embryology, larval biol- ogy, and post-larval development of Adalaria proximo (Alder and Hancock) (Gastropoda Opisthobranchia). Philo\. /;.;/n K. Sue. Lund. 242: 1-58. Todd. ( . I).. M. (i. Hentle>. and J. N. Havi-nhand. 1991. Larval metamorphosis of the opisthohranch mollusc Adulann /'/.u/nhj (Gas- tiopoda: Nudibranchia): the effects of choline and elevated potassium ion concentration. 7. Mar. Biol. Assoc. U.K. 71: 53-7 I tin-. I). 1995. I me structure of the cephalic sensory organ in veliger larvae of Liltorina liltorea, (L.) (Mesogastropoda. Liltorinidae). H\- drhim fnccus (Vosmaer, 1881) (Demospongiae, Poe.ilo-elcnda). found in deeper water ( 15-25 m). were sampled tw KV a month from June to September. 1982-19S4. u^.iMonal samples of these spe- cies were taken in spring and summer from 1985 through 1991 and in 1997. Subiid.il sponges \\ere sampled using scuba. For quantitative studies. 90 specimens of H. dnjar- dini (sampled from January to August in 1986-1989) and 29 specimens of /. />icciu and 25 specimens of A/, in- crustans (sampled from June to October in 1983-1984) were investigated. Other specimens of these three species were analyzed for the presence of sexual reproductive ele- ments. To ascertain the minimum lifespan of the three species, in July 1992 a few specimens of each species were marked in their locality. Every year these specimens were monitored for survival. Immediately alter specimens were removed from the water, sponge fragments were fixed in Bouin's and Car- noy's fixatives for light microscopy. Tissue fragments were dehydrated through an ethanol series, placed in celloidin blended with castor oil and then in chloroform, and embed- ded in paraffin. Sections were cut to a thickness of 6 /^.m and Meier's hematoxylin, eosin, and Heidenhein ferric hema- toxylin. Ten microscopic slides of each specimen were examined. The number of gametes, embryos, and larvae in parental tissues were calculated using Elvin's equation (Elvin. 1976): where N is the number of objects per cubic millimeter of tissue; 8N is the average number of objects (gametes, em- bryos, larvae and spermatic cysts) in the microscope lield: t is the thickness of the histological section (here 0.006 mini. I) is the diameter of the object; and K is a constant for converting the number of objects in a square millimeter to the number in a cubic millimeter: it is equal to 166.7. Egg, spermatic cyst, embryo, and unreleased larva vol- umes were calculated using the equation: V= (2) where V is an object volume and I) is its diameter. The volume of reproductive elements per cubic millime- ter of parental tissue was obtained by multiplying A' by I'. The total sponge volume in breeding season was measured by immersing each specimen into a graduated cylinder tilled with water. The volume of water displaced was a measure ol the volume of the sponge tissue. Data are presented as mean + standard error. A hydrologic thermometer was used to measure water temperature during sampling. Additional information on r, os'i Kigurv I. ( l,u|..i lulri. shown)).' iln- 10-m depth contour; insel Kandalaksh.i w lnu- Sea. uslralcs the localion of the Inlcl in COLD-WATER SPONGE REPRODUCTION STRATEGIES 79 seasonal temperature changes at different depths in Chnpa Inlet (the White Sea) was obtained from the White Sea Biological Station, Zoological Institute. Russian Academy of Sciences. Study Area Seasonal sampling was performed in the Chupa Inlet located in the innermost part of the Kandalaksha Bay, the White Sea (Fig. 1). All oceanologic and climatic data were obtained from Babkov (1982, 1984) and Babkov and Go- likov (19841. The region has a long, severe winter and a short, relatively warm summer. From December until mid- May, Chupa Inlet is covered with ice. The average annual water temperature in the Inlet, which has a mean depth ot about 20 rn, is about 5C. ranging from 1.5C in winter to I7C in summer (Fig. 2a). In autumn and spring the water column is homothermal: the temperature throughout all layers in November does not exceed +2C and in March ranges from -1.0 to -0.5C (Babkov, 1982). Except for the open part of the Kandalaksha Bay, Chupa Inlet is characterized by reduced salinity. The seasonal variation in the 10-m surface layer is between 15%< and 26%c; fluctuations decrease with depth, and in the bottom layer are not more than \%o. The minimum surface-layer Figure 2. Seasonal changes in temperature (A) and salinity (B) at different depths in Chupa Inlet. SI) i . i kl si-., i\ SK'l salinity was detected in April, and the maximum in Novem- ber (Fig. 2b)(Babkov. 1982). Temperature-salinity-analysis reveals two bodies of vva- ter in the region: surt'ace water boundary depth between 10 and 25 m. yearly temperature average I I ('. salinity not exeeeding 27.09? f : and bottom water water temperature not more than 5-6C. salinity exceeding: 27 iBabkov. 1982: Babkov and Golikov. 1984). Results Halisarea dujardini In the White Sea. Huli.\niru JnjurJini dwells at depths from 1.5 to 10 m. mainly on the algae / nc/n rr.w'r/i/wis and Laminaria saccharina and. raiel>. on stones. Its body shape is irregular: encrusting, pillowy or clotted in form. Its si/e varies from 3 to 40 nun in width and from 2 to 6 mm in height. The volume ol specimens averages 0.25 to 0.40 cm'. The body surface is smooth and slimy. Oscules are small. one or several in a specimen. Color varies from milky to greyish-brown. Hali.sarcii dujanlini is a gonochoristic organism. Sper- matogenesis occurs at a water temperature of around 0.8C and lasts for about 4 months. Spermatic cysts containing spermatocytes can be found in the male mesohyl in about the middle of December at a water temperature of about 0.08C. Male generative cells appear to originate from choanocytes that migrate into the lumen of choanocyte chambers, where spermatogenesis takes place. Choanocyte chambers thereby transform into spermatic cysts with a diameter between 40 and 90 pirn. During intensive sperma- togenesis (January-March), the mesohyl of the mature male differs greatly from that of an immature individual. It eon- tains neither choanocyte chambers nor channels and pores. At this time the volume of spermatic cysts containing male generative cells amounts to 0.6S ! 0.5 mm Vmm' of tissue (Fig. 3). In March and April, spermatic cysts contain only mature spermatozoa. From mid-December until mid-April, the volume of total reproductive elements (male and female) is nearly equal to the volume of spermatic cysts, differing from it only by the minute volume contributed by early oocytes d-'ig. 3). Spermatogenesis and. thereafter, tertili/a- tion cease by April or the tirst week of May. Karly oocytes, 15 to 35 ju.ni in diameter, are tirst observed during the last third of December, at a water temperature of about 0.6C. Cytoplasmic growth is recorded until the beginning of June: however, vitellogenesis begins in May. when water temperature is about -t-2C. Mature eggs. 110 130 /urn in diameter, appear in females at the end of May: their number increases rapidly and their total volume I' miii q I.OCXXX) : 0.10000 : 0.01000 : OOIOC 0.00001 2/1 M/F 2/1 M/F 5/4 M/F 15: 14 F Earh oocytcs O Mature ooo Ics 12 F ...- Embryos -o- - Larvae -A Spermatic cysts Tola I reproductive elements i i s iinpli I'rriotl (in thirds nt months) i i i -' ' ...... I ' ..... ites, iinl'iMis. and larvae per cubic millimeter ol parental tissue in Halisarea dujardini during Mi, i,-|imdnctic>n seasons for 1986- 1 WW. Numbers indicate individuals (males/ females) un.ilwr.l V nin.il bars designate standard errors of the values for total reproductive elements. COLD-WATER SPONGE REPRODUCTION STRATEGIES 81 reaches 0.49 0.01 1 mnrVmm 3 of tissue in mid-June (Fig. 3. 4a). The development of larvae, namely cleavage and mor- phogenesis, occurs within a fortnight, from late June until July. During this period water temperature averages about 10C. The even, asynchronous cleavage results in a coelo- blastula, which transforms into a disphaerula larva (Eres- kovsky and Gonobobleva, 1999), with a diameter between 120 and 150 jam (Fig. 4b. c). Larvae of H. ditjanlini (disphaerula), consisted of two flagellated sphaeras: exter- nal and internal; the internal sphaera was formed by invag- ination of laternal cells. The disphaerula was completely flagellated sparsely so on the posterior pole (Ereskovsky and Gonobobleva, 1999). The volume of reproductive ele- ments (cleaving embryos and prelarvae) reaches its maxi- mum from the end of June to the beginning of July and amounts to about 0.69 0.20 mmVl mm 3 of the tissue (i.e.. 69.5% of the total sponge volume; Fig. 3). This period is marked by the complete disorder of central and basal parts of the choanosoma, which are now filled with developing larvae (Fig. 4c). Normal tissue organization persists only in the narrow marginal zone of the sponge. Larval emergence occurs rapidly, within about a fortnight, begining before mid-July at a water temperature of about 12C. Slow postmetamorphic development of the new genera- tion and postreproduction rehabilitation of parental sponges continues until December. Some parental sponges died and underwent disruption after larval emergence. A general life-history scheme of H. diijardini in the White Sea is shown in Figure 5. Myxilla incrustans and lophon piceus White Sea populations of two Myxillidae species, M\xilla incrustans and lophon piceus, are similar in their main life-history stages. Both species are simultaneous her- maphrodites; oogenesis and spermatogenesis occur at the same time. Gametogenesis and embryogenesis take place only in the choanosoma. lophon piceus. in the White Sea, dwells at depths from 3 to 172 m on stony-muddy substrates where the water tem- perature ranges from - 1 . 1 5 to 1 1 .5C and the salinity from 18.6%c to 29.12% f , The body is irregular lumpy or flat- tened and uneven with a wrinkled and porous surface. The oscules average 0.5 cm in diameter. The body height ranges from 6 to 12 cm and the diameter from 7 to 10 cm. The volume of observed individuals averages 11-16 cm . All stages of embryogenesis and gametogenesis in /. piceus. including new gonia formation, occur simultaneously (Fig. 6). Generative cell development in /. piceus is usually initi- ated during February for female elements, and April-May for male elements. Male and female gametogenesis be- comes active in mid-May at depths of 15-25 m at an ' - x -** Figure 4. The central part of the chounosoma of Halisana diijardini. showing stages of vitellogenesis, cleaving embryos, and mature larvae. (A) Oocytes (O) in late vitellogenesis, still with prominent nucleus (N) and nucleoius (Nu), are surrounded by cells forming embryonic capsules (ar- rows). The aquiferous system in this part of the choanosoma is destroyed. There are remnants of exhalant channels (Ex) and choanocyte chambers (Cc) in upper marginal parts of the choanosoma. Scale bar = 50 mm. (B) Cleaving embryos (E) enclosed in embryonic capsules (Ec). As in (A), choanocyte chambers (Cc) remain in upper marginal parts of the choano- soma. Note the nucleus (N) and nucleoius (Nu) of blastomers. Scale bar = 50 jum. (C) Larvae (L) before emission with well-differentiated ciliated cells at their periphery. Scale bar = 50 mm. 82 A. V 1 Kl SKOVSKY Hjjure 5. Scheme of the life history of Halisarca Jujartlini at Chupa Inlet. White Sea. average water temperature of +0.1 C. Egg vitellogenesis and cleavage occur in late June and early July, \\hen water temperature is about +4.2C. During the last third of June the total volume of reproductive elements m this species amounts to about 0.05 0.021 mm /mm of tissue, whereas during the last third of July it amounts to 0.096 0.02S mm 3 (Fig. 7). The first larvae in /. piceus tissues are recorded at the end of July. These are typical parenchymulae common to the order Poecilosclerida. Their oval or oviform body (200 X 260 /j.m) is evenly covered (except the tailpiece i by llagella that are all of the same length (Ereskovsky. 1986). Larval emergence lasts from the first third of August (when water temperature is about 8C) to early October, when the larvae constitute the total volume of reproductive elements. In the initial period of this process, the volume ol the generative cells and larvae reaches O.I IS * (Mil mnr/mm of tissue (about 12%). It is notable that onlv local disintegration ol the parental tissues is observed during the period of larval emergence in /. /i/i h.is in White Sea sponges correlates with a spring temperature change in May. The disappear- ance of male spermatic cysts in April suggests that oocytes are fertilized during their cytoplasmic growth. Penetration of the oocyte by sperm during meiotic prophase has been described in the nematode Brach\coelium. the annelids Di- and Hixtriobdellu. ami the onvchophora Peripal- (Austin. 1465). Hiilixarca dujardini releases larvae during the period of temperature maximum. This is typical for marine hydro- bionts in cold waters: the release of their larvae is timed to the momentary summer period of warmest water (Kaufman. 1977: Kasyanov. 1989). Embryogenesis. larval develop- ment, and metamorphosis in the White Sea population of //. dujardini occupy only about 3 to 4 weeks from the end of June to the middle of Julv. White Sea populations of M. incrustans and /. piceus dwell mostlv in more stable and predictable conditions. Eurybathic circumlittoral-highhathyal (1.5-500 m) M. in- crusUins and /. piceus (Ereskovsky. I995a) inhabit chiefly stony-muddy substrates. The boreal-arctic species M. in <-ni\iiin.\ is eurychoric: it is found in all Arctic Seas; ihe southern Pacific boundary of its area is the northern part of i I'.l : 0.01 ; 0.001 O.IKKII - (MNKKII O. ^ V V V \ >V * > \ ) * * * I I I Early oocyles o - Mature oocytes - - -Embryos -o- Larvae - A - Spermatic cysts Total reproductive elements Sample I'rriod (in Ihirds of months) Figure 10. Mean volume of gametes, embryos, and larvae per cubic millimeter of tissue in the hermaph- rodite My.iillu iiicruMiuu during Ihe reproductive seasons for I9H3-I984. l-'or legend, see Higure -t COLD-WATER SPONGE REPRODUCTION STRATEGIES 85 Figure 11. Life-history scheme of My.wlUi incruxtan.i in the White Sea. the Sea of Japan and the California bathyal zone; in the Atlantic Ocean it extends south to the western Mediterra- nean and Cape Cod (Ereskovsky, 1994a). Thus, M. incntstans is apparently a eurythermic ( 1 .9- 16.2C) and euryhaline (25^-35. 59tr) species (Ereskovsky, 1994b). The area of the Pacific highboreal-Arctic species loplwn piceus is not so broad. I piceus is circumarctic; in the Pacific Ocean it is found near the Northern Kurile Islands, in the Sea of Okhotsk and near Vancouver Island; in the Atlantic Ocean it extends south to Baffin Sea and the Faeroe Islands. Like M. incnistans, this species is eurythermic (-1.6- 12.5C) and euryhaline (25.6%c-35.57) (Ereskovsky. 1994b). It is clear from the reported data that different stages of the sexual cycles in the investigated species correlate well with seasonal environmental changes. Most significantly, the dependence of the main stages of oogenesis (a compli- cated multi-stage process) upon water temperature should be noted. Thus, deutoplasm growth begins within hydrolog- ical spring in all of the populations investigated. Vitello- genesis and egg maturation, as well as embryogenesis, end during hydrological summer i.e., the warmest season. A similar dependence on temperature has been noted for the developmental stages of some tropical sponge species (Fromont, 1994; Fromont and Bergquist, 1994). Larval emission in some species is timed to a specific hydrological season in their habitat. On one hand, this relationship is modified by peculiarities of larval develop- ment and ecology since organisms are more temperature- sensitive during early ontogenesis than later (Kinne, 1963). On the other hand, the relationship is determined by the species genotype, and larvae are released when the temper- ature is close to the optimum for the species. But depen- dence of larval release upon water temperature is further mediated by biogeographical characteristics of the species. For poikilotherms. the temperature of the environment dur- ing speciation influences cell and tissue thermoresistance. which is considered to be a species-specific feature (Usha- kov. 1989). As a result, temperatures optimal for both life and larval release are closely connected with the conditions that prevailed during the origin of the species and, conse- quently, with its zoogeographic position (Golikov and Scar- lato, 1972; Mileikovsky, 1981). Maximal annual average temperatures most favorable for larval development are recorded within this period (Babkov, 1984). In this study, special attention was given to the state of maternal tissues in different sexual reproduction stages in each of investigated species. It became evident from previ- ous studies (Ereskovsky and Korotkova, 1997) that sexual and somatic morphogenesis con-elate closely in sponge on- togenesis. Thus, somatic tissue state is important for the attainment of different sexual reproductive stages. Sexual and somatic morphogenesis either take place as successive life-cycle stages or occur in parallel, but they vary in cor- relation with each other due to their equal dependence on internal integrative mechanisms (Simpson and Gilbert, 1974; Fell et ui. 1979; Korotkova, 1988). Owing to the vertical stratification in the Kandalaksha Bay, deep-water M. incnistans and /. piceus are less ex- posed to seasonal environmental fluctuations than is the shallow-water H. dujardini. It could be suggested then that M. incntxtiinx and /. piceus are A"-strategists, whereas H. dujardini is an /--strategist. Some features of these species may provide additional evidence for such conclusions: 1. Reproductive effort (the contribution by the organism to all parts of reproduction) is low in Myxillidae (about 7.3% of the maternal tissue volume in M. incntstans and about 12% in /. piceus) but high in H. dujardini (69.5% in females and 65% in males). 2. These different levels of reproductive effort result in different degrees of destruction of maternal tissue: only localized destruction in Myxillidae (both /. pi- ceus and M. incntstanx), but widespread destruction in H. dujardini. 3. Embryogenesis and larval development last over the hydrological summer in Myxillidae, but only 3 or 4 weeks in H. dujardini. 4. The average life span is more than 4 years in M. incnistans and /. piceus and about 7-12 months in H, dujardini. 5. M. incnistans and /. piceus inhabit a more stable environment than H. dujardini. X,, A V IKISKOVSKY Similar ecoK'gK nd i. -productive characteristics have been reported t 'onges inhabiting different regions. Thus, the volui ^productive elements in the eurybionl r-strategist .1 p. amounts to 1.57c-2Q7c under stress- ful conditit> Discover}, 1 Bay, Jamaica (Reiswig, 1973). Similar c :tstics have been reported for other /--strat- egists 51 .IN I ittoral specimens of Haliclona pennoHs from the Oregon coast of the United States (Elvin. 1976) and the estuarine species Haliclona loosanoffi and Halichondria sp. (Fell. 1976: Fell and Jacob. 1979; Fell ct ai. 1979: Fell and Lewandrovvsky. 1981; Levvandrowsky and Fell. 1981). Shallow-water Halichondria puniccd from the Barents (Ivanova. 1978). Baltic (Barthel. 1986. 1988). and White Seas (Ereskovsky. unpuhl.i could be classified as a typical opportunist species and an /--strategist. A comprehensive study of marine ecosystems is impos- sible without data on the reproductive cycles of the species of which they are composed. Knowledge of the peculiarities of both the reproduction strategies of the species and the reproduction tactics of the populations in different regions is thus of great value. The concept of reproduction strategies in sponges is still under development. In analysing sponge reproduction, it is necessary to consider the maternal tissue state and reproductive effort of the specimens throughout the period of reproduction. Ackno\\ ledn I thank Dr. N. V. Maksimovitch and Dr. S. M. Efremova for helpful discussions; Dr. A. A. Golikov for critical v ievv and help with the manuscript; and two anonymous review- ers for critically reading earlier versions of the article. This work was supported by INTAS - 97-857 program. IJtiTutuir Cited Viistin. ('. K. 1965. fertilisation. Prentice H.ill. I'.nglcwood ("lills. New Jersey. Habkm. A. I. 19X2. Uriel hydrological characteristics ol Ihc Chupu Inlet of the White Sea. Pp. 3-16 in / >/\. 1984. ll\-tlrobiocnmplexes of the While Sen. [In Russian! Zool. Inst. Acad. Sci.. Leningrad. Karthcl. I). 19X6. On the ecophysiologv ol lire sponge llnlicliomlriii Putin i-ii in Kiel Bight. I. Suhstratc specificity, growth and reproduction. Mar. Biol. 32: 2'M 2'>x. Harthfl. I). I9XX. On the ecophysiology of the sponge Halichomlria panicea in Kiel Hiuht. II li ..... lass, production, energy budget .mil integration in environmental processes. Mar. Kcol. 43: 87-93. Chen, \\. I. 1976. I- 1 , rodui ii"ii and speciation in tlatixarcu. Pp. I I ' 139 in Aspects nl -/ji;v. F. M. Harrison and R. R. Cowden. cds. Academic Press. New York. l.fn-rmiva, S. M., A. V. Kreskovsky, and I). B. I 'okiiui. !9H7a. Ciame- (ogcncsis in sponges ol th i , m ,l. M> \illtdac Irorn the While Sea. I. Oogenesis in Myxilla incrustans and lophon piceux (Demospongiae, Poecilosclerida). Ontogenez 18: 257-262. Kfremova, S. M., A. N'. Ereskovsky, and D. B. Tokina. 19X7b. Game- togenesis in sponges of the family Myxillidae from the White Sea. 2. Spermatogenesis in Myxilla incrustans and lophon piceus (Demo- spongiae. Poecilosclerida). Ontogenez 18: 263-268. Kit in. I). \\. 1976. Seasonal growth and reproduction of an interlidal sponge Haliclona permollis (Bnwcibank). Biol. Bull. 151: 108-125. F'.reskovsky. A. V. 1986. Larval formation in luplion piceus (Demo- spongiae. Poecilosclerida). |ln Russian] Zool. Zli. 65: 1614-1621 Kreskovsky. A. V. 1993. Addition to lire fauna of sponges (Porifera) of the White Sea In Russian] Vestn. St.-Pcicnhv. I 'ni\: Ser. #.?. 2 ( 101:3-12. Kresknvsk). A. V. 1994a. Materials to the faunistic study of the While Se.i and Barents Sea sponges. 2. Biogeographical and comparative-faunistic analysis. [In Russian] Vestn. St.-/', t, rsbg I nil s. ;-. #3. I (3): 13-26. Kreskovsky. A. V. 1994b. Materials to the taunistic study of the While Sea and Barents Sea sponges. 3. Dependence of sponge distribution on the temperature and salinity. [In Russian] \\-\in. Sl.-Petersbg. Univ. Ser. 3. 3(17):3-IO. Kreskovsky, A. V. 1995a. Materials to the faunistic study of the While Sea and Barvnls Sea sponges. 4. Vertical distribution. [In Russian] Vestn. St.-PetersbK. Univ. Ser. .?. 1 (3): 3-17. Kroskovxky, A. V. I995b. Materials to the faumsiic siudy of the While Sea and Barents Sea sponges S. (Juantilatne distnbulion. Herliner Ofimi^ .\hh. I- 16: 709-714. Kreskovsky, A. V. I995e. Materials to the faunistic study of the White Sea and Barents Sea sponges 6. The origin of the While and Barents seas sponge faun. is. Rerlincr (/Ymra.s. Abh. K 16: 715-730. Kreskovsky, A. V., and K. I.. (;<>nobol>li'\a. 1999. V data on em- hrvonic development of llalisurca tlujtirdini Johnston. 1S42 (Demo- spongiae: Halisarcida). Zoos\-\tcinti. In press. Kreskovsky, A. V., and (.',. P. Korotkova. 1997. The reasons of sponge sexual morphogenesis peculiarities. Modern problems of Poriferan biology. A. Ereskovsky. H. Keupp. and R. Kohring, eds. Berliner GVounv Ahh. Reihe H. Bd. 20: 25-33. Fell. P. K. 1976. The reproduction of Haliclona loosanoffi and its appar- ent relationship lo water lemperalure. liiol. Hull 150: 200 210. Fell. P. K., and \V. K. Jacob. 1979. Reproduction and development of Haliclionilnu sp. m the Mystic Hstuary. Connecticut. Biol. Bull. 156: 62-75. Fell, P. K., and K. B. l.cuandnmski. 1981. Population dynamics of lire estuarine sponge. Halithoinlna sp.. within a New England eelgrass community. ./. Kxp. Mm. Biol. l-Aiil. 55: 49-63. Fell. P. K., K. B. I.ewandrowski. and M. l.cnice. 1979. Poxtlarval reproduction and reproductive strategy in Haliclona loosamiffi and Halichomlria sp. Pp. 113-1 22 in Biologic ties Spongiaires. C. Levi and N. Boury-hsnaull. eds. CNRS. Pans I MinH.nl J. 1994. Rcpioduclnc development and timing of tropical sponges (Order Haplosclenda) from the Great Barrier Reef. Australia. Conil Kcch 13: 127 133. Frumont. .(.. and P. K. Ber|>i|iiist. 1994. Reproductive biology of three sponges species of the genus Xestospongia (Porifera: Demospongiae: Petrosiidai Mom lire Great Barrier Reef. Coral Reefs 13: 1 19-126. Colikov, A. N.. and O. A. Scarlato. 1972. Method for indirectly defin- ing optimum temperatures of inhabitancy for marine cold blooded animals. Mar. liio/. 20: I 5. l\ano>a, I.. V. 1978. Morphogeiielic processes and seasonal changes ol the anatomical and tissue organisation ol lire Barents Sea sponges Hulichoiulria panicea (Pallas). [In Russian] Arch. Anal. Histol. Km- hnol. 10: 62-71. K.I -,-IIM.I V. I.. 1989. Reproductive Strateg? oj Marine Bivalve Mol- loses and Echinodemw. [In Russian | Nauka. Leningrad. K. mini. in /,. S. 1977. Peculiarities of Sexual Cycles of White Sea Invertebrates. [In Russian) Nauka. Leningrad. COLD-WATER SPONGE REPRODUCTION STRATEGIES 87 Kinne, O. 1963. The effects of the temperature and salinity of marine and brackish water animals. I. Temperature. Oceanogr. Mar. Biol. A/iini. Rev. 1: 301-340. Korotkova, G. P. 1988. The integrative mechanisms and morphogenesis (to the problem of evolution of ontogenesis). |In Russian] Z/i. Ohshch. Biol. 69: 464-475. Korotkova, G. P., and A. V. Ereskovsky. 1984. Peculiarities of the egg cleavage of White Sea Sponge Halisarca dujardini Johnston. (In Rus- sian] Vestn. Leningr. Univ. Biol. 21: 36 12. Korotkova, G. P., and N. O. Ermolina. 1982. Period ot development of the larvae Halisarca dujanlini (Demospongia). [In Russian ] Zool. Zh. 61 (10): 1472-147". Levi, C. 1956. Etude de Halisarca de Roscoff. Embriologie et system- atique des demosponges. Arch. Zool. Exp. Gen. 93: 1-181. Lewandrovvski, K. B., and P. E. Fell. 1981. Sequential reproduction by different types of specimens of the estuarine sponge. Halichomlrni sp with an emphasis on reproduction of postlarval specimens. Int. J. In- vertebr. Reprod. 3: 227-236. Mileikovsky, S. A. 1981. Reproductive Ecology of Marine Benthos. |In Russian] Nauka. Moscow. Pianka, E. R. 1978. Evolutionary Ecology. Harper and Row, New York. Propp, M. V. 1971. The Eco/ogv of the Coastal Associations of the Murman Seaboard of the Barents Sen. [In Russian] Nauka, Leningrad. Reiswig, H. M. 1973. Population dynamics of three Jamaican Demo- spongiae. Bull. Mar. Sci. 23: 191-226. Simpson, T. L., and J. J. Gilbert. 1974. Gemmulation, gemmule hatch- ing and sexual reproduction in fresh-water sponges. I. The life cycle of Spongil/a lacuslris and Tuhella pennsylvanica. Trans. Am. Microsc. Soc. 92: 422-433. Ushakov, B. P. 1989. Cell Ph\siolo K \ and the Problem of Species in Poikilothenns. [In Russian] Nauka, Leningrad. White, IL, and D. Barthel. 1994. Reproductive cycle and oogenesis of Halichondria panicea (Pallas) in Kiel Bight. Pp. 297-305 in Sponges in Time and Space. R. van Soest, T. van Kempen, and J. Braekman. eds. Balkema. Rotterdam. Reference: Bio/. Bull. 198: 88-93. (February 2000) \Yalking Versus Kreathing: Mechanical Differentiation of Sea Urchin Podia Corresponds to Functional Specialization HOLLY A. LEDDY 1 '* AND AMY S. JOHNSON ,' v l)c pa nine in, Bowdoin College. Brunswick, Maine 0401 1 \hstruct. The podia of sea urchins function in locomo- tion, adhesion. (ceding, and respiration; but different podia on a single urchin arc often spcciali/cd to one or more of these tasks. \Ve examined the morphology and material properties of podia of the green sea urchin. Stron^yloccn- irotns Jri>chucliicii.\i.\. to determine whether, despite appar- ent similarities, they achieve functional speciali/ation along the oral-ahoral axis through the differentiation of distinct mechanical properties. We found that oral podia, which are used primarily for locomotion and adhesion, arc stronger and thicker than ahoral podia, which arc used primarily for capturing drift material and as a respiratory surface. The functional role of ambitul podia is more ambiguous: how- ever, they are longer and are extended at a lower strain rate ili. in other podial types. They are also stronger and stiller than aboral podia. In addition, all podia become stronger and stiffen when extended at faster strain rates, in some cases by nearly an order of magnitude for an order of magnitude change in strain rate. This strain-rale dependence implies that resistance to rapid loading such as thai imposed by waves is high compared to resistance to slower, sell- miposcd loads. Thus, the serially arranged podia ol ' S. ili'nc- hachiensix are functionally speciali/ed along an oral-aboral axis by differences in (heir morphology and mechanical properties. Introduction The surface of the green sea urchin, Strongylocentrotus (ln>ehcichiensi\. like that of all urchins, is covered with live Received 21 October I'm. .Kxcptcd IX November l ' Present Address . I p.iriincnl. Duke University. Durham. NC 27708. * Author lo whom conx-.poiuh-ni.-e should he addressed. E-mail: ha!2l" acpub.dukc.edu double rows of podia, and these structures are essentially hollow lubes with terminal suckers. The wall of each po- dium consists of three main layers: an outer epithelium, a middle connective tissue layer, and an inner muscular layer (Florey and Cahill. 1977). Internally, the podium is con- necled to an ampulla, a sac-like structure that antagoni/es podial movement. Contraction of the ampulla forces coelo- mic thud into the podium, and thus extends it. Conversely, contraction of the longitudinal muscles shortens the podium and forces coelomic fluid back into the ampulla. All of the podia on S. droebachiensis share these characteristics, and all of them are similarly shaped. They exemplify the typical echinoid locomotory podia. However, podia perform a variety of functions besides locomotion, including adhesion, feeding, shading, sensing, and respiration, and many podia are speciali/ed for one of these functions. I'odia in a particular position along the oral-aboral axis lend lo perform a particular function and often exhibit morphological features thai enhance pertor- mancc of (hat function. Lor example. Arhacia /nincliilatd has morphologically differentiated podia, reflecting the sep- aration of respiratory, sensory, and locomotory functions. The aboral podia lack suckers and their larger surface area and thinner walls enhance respiration, (he ambital podia are long with minimal suckers to allow sensation in ihe region around (he urchin, and the oral podia are suckered to facil- itate locomotion (L'enner. 1973). Although the podia of S. droebachiensis lack these gross morphological differences, there may still be a division of labor between the primarily locomotory oral podia and the primarily respiratory aboral podia. We hypothesi/e thai the podia of Ihis species achieve functional speciali/ation through the differentiation of disiinct mechanical properties. The goal of this study was to icsi our hypothesis thai the differences in function of oral, aboral. and ambital podia are MECHANICAL SPECIALIZATION OF PODIA 89 correlated with differences in their mechanical characteris- tics and morphology. Materials and Methods Sea urchins. Stronglyocentrotus droebachiensis, were collected subtidally near Monhegan Island. Maine. The animals, ranging in diameter from 45.4 mm to 59.7 mm, were maintained in a recirculating seawater aquarium kept at 1()C and fed Laminaria sp. every 2 weeks. Natural extension Urchins in a glass aquarium were videotaped, and the natural extension rate and initial length of podia were mea- sured from the digitized video images. The length of a podium was measured from the base to the sucker as it was extending, and just as the podium straightened. Extension rate was quantified as the increase in length (in millimeters) per second, and strain rate was quantified as true strain (the natural logarithm of the ratio of final length to initial length ) divided by the time to extend. Mechanical extension Podia from three locations on the animal were tested separately: aboral podia are those on the upper surface of the urchin; ambital podia are those occupying a narrow circumferential band at the widest diameter of the urchin; and oral podia are those on the lower surface of the ur- chin. Urchins were placed in an isotonic solution. 7.59r MgCK 6H^O, at 10C, for 15 min prior to mechanical tests of their podia and were kept in this solution during the tests (Dales. 1970). The MgCl 2 solution prevented muscle con- traction, so passive material properties could be tested. All mechanical tests were performed with an Instron material testing device (Model 4301) that loaded materials at specified rates and simultaneously measured force and extension. Force and extension signals were digitized (12- bit) at 1000 Hz and recorded in a computer. The error in the force measurements was <0.001 N. The error in the extension measurements was <0.01 mm. Tests were con- ducted as follows. An urchin was strapped in place, and a small, spring-loaded clip was attached to the distal end of a podium. The podium was then pulled perpendicular to the surface of the urchin (in the direction of normal extension) at a constant extension rate until it broke. Fifteen podia 5 aboral. 5 ambital, and 5 oral were tested on each urchin at a given extension rate. Each of five urchins were tested at three extension rates: 0.167 mm/s (minimum natural extension rate). 0.708 mm/s (halfway between the mean and the minimum natural extension rates), and 1.25 mm/s (mean natural extension rate). An- other set of podia from one urchin was also tested at 3.45 mm/s (maximum natural extension rate). In comparing lengths of different podial types and calcu- lating strain and strain rate, initial length of the podium (measured between the body wall and the end of the clip) was defined as the point at which force started to increase (above 0.001 N) during the breaking tests, thus indicating that the podium has just started to stretch (Fig. la). Nominal stress on a podium was calculated by dividing the force applied by the mean tissue cross-sectional area of podia of that type from that urchin. Nominal strength is the nominal stress at which a material breaks. True strain was calculated by taking the natural logarithm of the length divided by the initial length. Stiffness is the slope of a stress-strain curve and was taken from the upper, linear portion of the J-shaped stress-strain curves (Fig. Ib). Strain rate was calculated by dividing the final strain by the time required to get from initial length to breaking. Tissue cross-sectional area Relaxed podia from urchins soaked in 7.5% MgCK 6H,0 were cut off at the base. Although this technique depressurized the podia, it allowed us to compare the three types in as similar a state as possible. The exterior and lumenal diameters of the podia were measured halfway between the base and the sucker with an ocular micrometer (0.01 mm) on a light microscope. These diameters were used to calculate the total and lumenal cross-sectional areas, and the tissue cross-sectional area was found by subtracting the lumenal cross-sectional area from the total cross-sec- tional area. Five podia of each type were measured on each urchin. Results Structure and natural extension All oral, ambital. and aboral podia had the same gross morphology (suckered tube) and similar diameters (Table 1 ). The ambital podia, however, were 67% longer than the aboral and oral podia, and the oral podia had thicker walls and a much greater cross-sectional area of tissue than either the aboral or ambital podia (Table 1). Urchins extended their oral, ambital, and aboral podia at the same rate. None- theless, because the ambital podia had the greatest initial length, they also had the lowest strain rate (Table 1 ). Mechanical extension tests All the material properties of all podia were positively dependent on strain-rate (Fig. 2, linear regression analysis, all P < 0.05. all d.f. = 86). For the three podial types, the slopes of breaking force, nominal strength, and breaking strain with respect to strain rate were not significantly different (ANCOVA. Table 2). For breaking force and nominal strength, the r of the regressions ranged between 0.44 and 0.58; and for breaking strain, the r of the regres- "' H. A. LbDDY AND A. S. JOHNSON a. Z o 0.07 0.06- 0.05- 0.04 0.03- 0.02- 0.01 0.00 2 4 6 B 10 12 14 [ < AL > Extension (mm) 16 b. 0.6- \ Ahoral Ambital Oral ANOVA d.t. = 2. 96 Initial length (mm) 7.59 ill d ., 1 1.64(0.5) h 6.26 (0.3 1 a /> | Exterior diameter (mm) 0.6 (0.01 i 0.7: mill i 0.70 ' P -- 0.0 Wall thickness (mm) 0.06(0.01) .1 0.07(0.01) a (1 1 1 (0.01) h P^ (1110(11 Cross-sccl. (issue ai 0.12(0.007)a 0.13(0.008) a 0.18 (0.009) b P = 0.0001 d.f. = 2. 89 Extension rale imcn 1.23(0.16) a 1.30(0.17) a 1.23(0.13) a P = 0.9 Strain rale (s~'( 0.34(0.05) a 0.14(0.03) b 0.45 (0.07) a P = 0.(X)6 si'tiiiip.inn j il \K-.ms with dilleieni letters are significantly different (P < 0.05 1 b\ the Fisher PLSD a fitisterinri test. MECHANICAL SPECIALIZATION OF PODIA 91 a. 0.1- CQ 0.01- 0.01 0.1 Strain Rate (s-i) b. i - 2 i I CD lo, & 0.01 0.1 Strain Rate (s- 1 ) c. 4.0 .S a en 60 (8 i 01 1 - j CQ 0.4 ir n "u w- vv |D H, Q fl C _ n n * ^ n 0.01 0.1 Strain Rate (s- 1 ) d. g S 2 in en 01 CD z ai 1- ffi 0.01 0.1 Strain Rate i Figure 2. Breaking force (a), nominal strength (b). breaking strain (c). and nominal stiffness (d) are plotted on a log-log scale against strain rate for oral (crosses, dotted line), ambitul (filled circles, dashed line), and aboral (open squares, solid line) podia. Lines are least-squares linear regressions. Line slopes and elevations were compared by ANCOVA (Table 2). rates, and aboral podia were significantly stiffer than oral podia at all strain rates; r values for these regressions ranged from 0.30 to 0.43. Discussion Some of the morphological differences found among the podia of other urchins (e.g., presence versus absence of suckers) are absent in Strongylocentrotus droebachiensis, possibly because the aboral podia, in addition to acting as a respiratory surface, perform other functions that require suckers. For example, drift algae captured by aboral podia may be the primary food source for some urchins (De Ridder and Lawrence, 1982). Furthermore, urchins like S. droebachiensis, that live in high-flow environments, may require suckered aboral podia to secure themselves in cracks and to right themselves (Fenner, 1973). Thus, aboral podia, though performing a different set of tasks from oral podia, would still require the same gross morphology. Although all podia in S. droebachiensis do possess suck- ers, they still exhibit mechanical and morphological differ- entiation consistent with their location and functional spe- cialization for locomotion or respiration. The greater strength and thicker walls of the oral podia presumably enhance one of their primary functions, which is to resist hydrodynamically generated forces on the urchin (Denny and Gaylord, 1996). Similarly, the thin walls of the aboral 92 H. A. LEDDY AND A. S. JOHNSON Tahli- 2 Comparison of strain-rale dependent material prn/>: Material property IN. Ml ,'t elevations 1 1 ) Breaking force (N) (2) Nominal strength (N mm" : ) i ; i Breaking strain (4) Nominal stiffness (N mm~ : i inihital > aboral (Oral = ambital) > aboral i pi.il lainbital = aboral) Ambital > aboral Aboral > oral Comparisons of uta shown in Figure 2. For material properties (1-3). si not statistically different from each other ( ANCOVA. P > 0.05. u.l = 2. 259), so all comparisons of elevations we im- possible. For nominal stiffness (4). a posteriori statistical analysis by Tukey's Q re n.u slopes for umbital and oral podia were sigmli- cantly different i/' < 0.05. d.f. = 3. 259). and so thai comparison was omitted: all other slopes were statistically similar iTukey's Q. P > 0.05. d.f. = 3. 259). Overall ANCOVA for elevations \vas always significant (all P < 0.05. all d.f. = 2. 261). Symbols indicate results ol u /),. liiol. 199(3): 717-729. IK- KI.I.I. i ('., and J. M. l.awri'tiiT. 1982. Food and feeding mecha- nisms: Kchmoidea. Pp. 57-115 in Echinoderm Nutrition, M. Jangoux and J. M Lawrence, eds. A. A. Balkema. Rotterdam. MECHANICAL SPECIALIZATION OF PODIA 93 Fenner, D. H. 1973. The respiratory adaptations of the podia and am- Long, J. H. Jr., D. A. Pabst, W. R. Shepherd, and W. A. McLellan. pullae ofechinoids (Echinodermata). Binl. Bull. 145: 323-339. 1997. Locomotor design of dolphin vertebral columns: bending me- Florey, E. A., and M. A. Cahill. 1977. Ultrastructure of sea urchin tube chanics and morphology of Delpluiui.t Jclphis. J. Ev/>. Binl. 2(10: feet. Cell Tissue Res. 177: 195-214. 65-81. Koehl, M. A. R. 1977. Mechanical diversity of connective tissue of the Wilkie. I. C., R. H. Kmson, and C. M. Young. 1993. Smart collagen in body wall of sea anemones. J. Exp. Biol. 69: 107-125. sea lilies. Nature 366: 519-520. Long. J. H. Jr. 1992. Stiffness and damping forces in the intervertebral Wilson. J. A. and G. Hoyle. 1978. Serially homologous neurons as joints of blue marlin (Makairti nit'ricans'). J. E.\p. Biol. 162: 131-155. concomitants of functional specialisation. Nature 274: 377-378. Reference: Biol. Bull. 198: 44-100. (February 2(XX Egg Br oding by Deep-Sea Octopuses in the North Pacific Ocean JANET R. VOIGHT'-* AND ANTHONY J. GRKHAN 2 1 Department /ooli>^\. Field Museum of \aturnl History. Roosevelt Rd. at Lake Shore Dr.. Chicago. Illinois 60605; mill Martin Ryan Marine Science Institute, National University of Ireland. Cialway. Ireland \hstract. Videotapes made from the submersible Alrin on Bab\ Bare. .1 2N)()-m-deep North Pacific basalt outcrop. and at two other deep-sea localities document that octopuses of the genera (irtineledone and Benthoctopus attach their eggs to hard substrate and apparently brood them through development. The behavior of brooding females was gen- erally similar to that of shallow-water octopuses, bill the genera showed apparent differences. In addition to the high density of brooding females observed at Baby Bare, which may relate to the increased availability of exposed hard substrates for egg attachment and of pie\. females are suggested to increasingly associate with hard substrates as they mature. The biology of Baby Bare ma\ seem unduK unique because the outcrop is isolated on a scdimenied plain and is among the few exposures of hard substrate other than hydrothermal vents that have been explored by submersible. On the sediment-covered ocean floor, the availability of hard substrate may strongly affect the distribution of brood- ing octopuses. The si/e and shape of boreholes in 19 ol o\er 400 thyasirid clam shells collected from Bah\ Bare support the hypothesis that octopuses had preyed upon the clams. Introduction Videotapes filmed from submersihles ha\e documented the beha\ ior of deep-sea squids (Moiseev. 1991 ; Vecchione and Roper. 1991; Roper and Vecchione. 1997) and citrate octopods i cchionc and Young, 1997; Villanuc\a el til.. 1997). However, such reports have provided \iiiuallv no new informal KID cm the henthic octopuses that form the Octopodidae. despite the occurrence of nine recogni/ed Received 17 May I'W" ' > i.,her I9'W. To whom corrcspondi n ln.nl. I In- .uhlu-xM-cl I, mail: JvoighU"' lmnh.org genera at depths greater than 1000 m ( V'oss. 1988a: Gon/a- le/ el al.. 1998). Laboratory studies do not compensate for the near-absence of m \//H observations of deep-sea octo- puses. Bathypolypus arcticus is the only species reported from below 1000 m that has survhed trawling to be suc- cessfulK maintained in the laboratory (O'Dor and Maca- laster. 1983; Wood et al.. 1998). This species, however, is unusual among deep-sea octopuses (Voss. 1988a) in occur- ring at depths as shallow as 75 m at high latitudes (Wood ft til.. I99S). Trawl-collected octopodid specimens generally provide sparse information on behaviors such as maternal care or on the t\pc of pre\ taken. Because we have had to rely almost exclusively on specimens collected in trawls, our knowl- edge of these fundamental aspects of deep-sea octopus biology is minimal. The rare discovery of isolated octopo- did eggs in a trawl haul has required that generic identifi- cation be interred from developing embryos (O'Shea and Kubodera. I99di. despite our scant knowledge of embry- onic development in deep-sea octopodids. Although deep- sea octopuses are presumably predators, as are other cepha- lopods. the pre\ taken by hatchlings and adults remains generally unknown (but for exceptions sec O'Dor and Ma- calaster. 1983; Nixon. 1987). This paper reports submersible observations that docu- ment egg attachment and brooding by members of two octopodid genera. Graneledone and Hentltoctopus. al depths greater than 2000 m in the North Pacific Ocean. Documen- tation of maternal egg care among deep-sea octopuses con- firms the presence of this basic behavior in diverse members of the Octopodidae. The very high densities of brooding females discovered on the base and sides of Baby Bare, an SO in high basalt outcrop, suggest that substrate suitable tor i iiiachmenl may he extremely limiting. We also pro\ ide EGG BROODING BY DEEP-SEA OCTOPUSES 95 evidence to suggest that small octopuses drill the shells of clams on which they feed. Materials and Methods In situ obsen'ations Videotapes made by the Deep Submergence Vehicle Alvin during dives in the North Pacific Ocean at Baby Bare outcrop (47 42.64'N, 127 47.15'W; 2640 m) in Cascadia Basin in August 1995, October 1997, and July 1998 form the bulk of the observations reported here. Additional data derive from an ROV Jason dive at Baby Bare in 1999. from Alvin dive videotapes taken in July 1994 (R. A. Lutz, pers. comm.) at Middle Valley (48 27'N, 128 42'W; 2400-2430 m), a sedimented hydrothermal vent described by Juniper et ul. (1992), and from two still photographs shot from Alvin near the Oregon Subduction Zone in 1984 (J. C. Moore, pers. comm.). Becker and Wheat (1995) and Mottl et ul. (1995, 1998) describe the geology of Baby Bare. The extinct volcano is estimated to rise 300 m from a buried ridge crest, but extends only 70 m above the heavily sedimented, 2592-m- deep seafloor. Baby Bare is one of the few volcanic edifices that penetrate the nearly continuous layer of sediment that fills the northern Cascadia Basin between Juan de Fuca Ridge and the continent (Davis et ai, 1992a). Basalt expo- sures on steep, sediment-free slopes of Baby Bare occupy much less than 0.1 km 2 in total area (Mottl et ai. 1998). The Baby Bare fauna differs strikingly from that of the sur- rounding abyssal plain (Grehan et ul., 1995: Mottl et ul., 1998). Its abundant sponges and echinoderms are typical of the hard-substrate, filter- feeder- dominated seamount fauna, the composition and density of which may reflect the influ- ence of the outcrop on local currents (e.g., Tyler and Zi- browius, 1992; Genin et a!.. 1992; Rogers, 1994). The fish fauna includes zoarcids, skates, and macrourids. In addition, at Baby Bare, thyasirid clams form locally dense aggrega- tions near springs with water temperatures up to 25C (Grehan et al., 1995; Mottl et ul., 1998). Temperatures were recorded with the Alvin and SUAVE (SUbmersible System Used to Assess Vented Emissions) heat probes (Massoth et ul., 1995). Octopuses were tentatively identified as members of the genus Graneledone if the videotape showed them to have many dorsal skin warts that are especially prominent over the eyes and arm suckers arranged in a single row. Individ- uals with skin warts in which the arm suckers formed a zig-zag row were also assigned to this genus. Octopuses with smooth skin and two sucker rows (in which the suckers were conspicuously smaller relative to the size of the octo- pus than in individuals referred to Graneledone) were as- signed to the genus Benthoctopiis. Internal examination of 10 specimens collected from Baby Bare, Middle Valley, and Axial Volcano (46N 130W, 1459 m) that were identified on videotapes as members of Graneledone allowed this generic determination to be tested. These A/vz'n-collected specimens were compared to other North Pacific octopus specimens in the collections of the Field Museum of Natural History, the California Academy of Sciences, and the Uni- versity of Miami Marine Laboratory, including seven para- types of G. pacified Voss and Pearcy, 1990. The specimens lack an ink sac and crop, have irregular radulae and small posterior salivary glands, and the dorsal skin of their mantles and heads bears cartilaginous tubercles that are diagnostic of the genus Graneledone (Voss, 1988h). The zig-zag arrangement of arm suckers on some video- taped octopuses was attributed to the presence of more suckers that could be arranged in a strictly single row on arms of that length (Voight, 1993). The species determina- tion is complicated by the possible synonymy of Granele- done horeopacifica Nesis, 1982, and G. pacifica Voss and Pearcy, 1990. Comparisons of the skin texture and sucker and gill counts of these specimens with seven paratypes of G. pacifica and extended study of Northeast Pacific speci- mens of the genus suggest that more than one species exists, probably segregated by depth (Voight. 1998). Because the 1350-m depth of the type locality of G. boreopacifica Nesis, 1982, contrasts sharply with the 2706-m depth of the type locality of G. pacifica Voss and Pearcy, 1990, the specimens from Baby Bare and Middle Valley are tentatively assigned to Voss and Pearcy's species, and the one from Axial to Nesis' species. The reproductive condition of four submers- ible-collected female specimens, including the female ten- tatively assigned to G. boreopacifica. is also reported. Because no specimens that were identified from the vid- eotapes as members of the genus Benthoctopiis were col- lected, their identification remains unconfirmed. Although they might pertain to the genus Bathypolypus, the smooth dorsal mantle and the distinctly reversed countershaded coloration of these octopuses suggest otherwise. The appar- ent absence of octopodid genera other than Graneledone and Benthoctopiis from the deep Northeast Pacific Ocean (Voss, 1988a; Voss and Pearcy, 1990) and the fact that the videotaped octopodids consistently show characters shared with those of specimens of Benthoctopiis trawled from depths of over 2500 m off the Oregon margin (Voight, unpubl. data) strongly support this generic assignment. Clam shell collections Biological collections made at Baby Bare during the three cruises included sediment scoops containing thyasirid clams and clam shells. The hypothesis that octopuses preyed on the clams (Mottl et a!.. 1998) was tested by comparing boreholes made by shallow- water octopuses (Nixon et al., 1980; Nixon and Maconnachie, 1988) to those discovered in clam shells from Baby Bare. Features characteristic of oc- topus boreholes are the round or ovoid shape, the beveled '< J. R. VOIUHT AND A J. GREHAN cross-section w ith an external orifice larger than the internal orifice that penetrates the inner shell surface, and the It Ca- tion in or close to the myosiracum (Nixon and Ma- connachie. 1988). To test whether boring predators pictcr- entially preyed on clams of a given si/e, the length valve collected in 1997 was measured, am! me 'un lengths of intact and bored valves were comp proportion of bored clam shells was calculated ' ear's collec- tion and compared across collect' Resu!- In situ observations DSV Alvin videotapes made both at Baby Bare and at Middle Valley show unu- ;.il!\ high densities of octopuses. as Mottl el al. iiwsi and Juniper ct at. (1992) noted. Although octopuses are raicl> seen during most submersible dives, 28 octopuses were videotaped during 11 dues at Baby Bare in 1995. One 3-min length of videotape from a 1997 dive shims textured dorsal mantle, pruminent supraocu- lar eirri. and comparatively few arm suckers. Eggs are seen suspended from the rock just ahtne and to the left of the female's web. roughly I km from the nearest cold seep i.l. C. Moore, pers. comm.): although eggs can be seen near her upper web. the rock cannot he identified. Videotape from Mama Bare, another small basalt outcrop, also shows octopuses brood- ing eggs, although on the single dive there, fewer octopuses were seen than at Baby Bare. A single rock outcrop was \ isited in both 1998 and 1999. A female of Graneledone was present at the outcrop in both years. In 1999. her appearance was consistent with impend- ing senescence: her posture indicated a deterioration of muscle tone, with her arms limply coiled: her skin was mottled and appeared to be flaccid. Unfortunately her clutch could not be examined. Two eggs collected with the brooding female ot'Granelc- ./'//, in 1995 (egg length = 25 mm: width = 12.25 mm) contained partially developed embryos. The embryos mea- sured 9.5 mm m mantle length, with 7-mm-long arms that carried up to 30 suckers each. The similarity of these embryos to those reported anil illustrated by O'Shea and Kubodera ( 1996) supports their assignment of the eggs and embryos to Gruni'letlnne. Embryonic structures were not yet visible m the seven eggs collected with the brooding female in 1998 (si/e range: 22.0 mm long by 1 1.2 mm wide to 24. 8 by 12.2 mm). Water temperatures near the brooding sites were recorded during two dives in 1995. The ambient temperature mea- sured with the SUAVE probe was 1.83C (G. J. Massoth, pers. comm.). At the base of a 1.5 m high basalt outcrop on which an octopus brooded eggs, the Alvin heat probe re- corded a temperature of 3.0C; within the crevice where the eggs were attached, the temperature was 1.83 "C. Tempera- EGG BROODING BY DEEP-SEA OCTOPUSES 97 Figure 2. Still from a videotape of a female of Bemhoctofius brooding eggs inside a rock overhang. Alrin dive 2974, August 13. 1995. Note the double rows of arm suckers, the dark ventral web surface, and the much brighter dorsal mantle surface. The oral surface of the animal is positioned away from the egg mass and the tear-shaped eggs are suspended from the roof of the rock crevice. ture probes inserted into the sediment recorded temperatures as high as 4.5C, but the water temperature just above the sediments was 1.84C, only slightly elevated over ambient (G. J. Massoth, pers. comm.). Brooding individuals of Benthoctopus were seen much less often than were those of Graneledone. Most members of Benthoctopus were observed partially hidden under ledges or in narrow crevices (Fig. 2). on one occasion within 5 in of a congener. Once in both 1995 and 1997. females of both genera brooded eggs in opposite ends of the same cleft in the rock face at Baby Bare. The sequence in which the females moved into the cleft could not be estimated because females of Benthoc- topus were so deep in the recess of the rock that the develop- mental stage of their eggs could not be determined for com- parison to those of Graneledone. Their proximity (within an arm's length) allowed the animals to be compared in size: females of Benthoctopus seen brooding at these depths of from 2400 to 2600 in were generally one-half to one-third the size of those of Graneledone. This differences in size and in how females were positioned those of Benthoctopus were most often under rock overhangs or inside crevices while those of Graneledone were most often positioned against rocks are likely to have strongly biased the observations toward mem- bers of Graneledone (e.g., 16 of 20 observations of octopuses in which the genus could be identified in 1995). Videotapes show that in both genera the eggs are individu- ally attached to the rock by short stalks that rapidly broaden to merge with the large chorion balloon of the egg. This outer egg membrane (chorion) is transparent, allowing the developing embryos to be seen. The developmental stages observed ranged from apparently undeveloped eggs to embryos in the second inversion (terminology of Boletzky, 1987). In clutches guarded by members of Graneledone, each egg stalk could be seen to be attached to the rock by dark green cement similar to that reported for Bathvpolvpus arcticus (O'Dor and Maca- laster, 1983). Regardless of the egg's stage of development, the color of the cement appeared to be the same. The number of eggs in any given clutch could not be determined. Differences between the genera, in addition to the mor- phological characters used to distinguish them, were appar- ent. Octopuses of Graneledone behaved sluggishly, often showing virtually no response to the approach of the sub- mersible. In contrast, individuals of Benthoctopus often jetted into the water column, sometimes toward the ap- proaching submersible. While brooding eggs, octopuses of Graneledone positioned themselves with their dorsal sur- face away from the eggs and their oral surface nearest them. In contrast, octopuses of Benthoctopus positioned them- selves with their oral, sucker-carrying surface facing away from their eggs and their mantle tip pointed toward them. Although observations are few, the eggs brooded by octo- puses of Graneledone appeared to be more sausage-shaped and the eggs of Bentlioctopus more teardrop-shaped. The reproductive organs of all female specimens of Graneledone that were collected from hard substrate by submersible were enlarged, and the ovarian eggs of each showed some degree of enlargement (Table 1 ). The female brooding early-stage eggs collected from Baby Bare in 1998 had the smallest ovarian eggs. The oviducal glands of brooding females were smaller than those of nonbrooding females collected from hard substrate. No prey were appar- ent in the digestive systems of either brooding female, although the large mass of amber-colored oil in the proxi- mal intestine of the female collected from Baby Bare in 1995 suggested a partially digested egg. Clam shell collections Small boreholes were discovered in between 3.8% and 5.1% of the thyasirid clam shells (tentatively identified as Aximis sp., near A. grandis, E. Southward, pers. comm.) collected (Table 2). Round or slightly oval boreholes were located most often on or near the umbo of the valves. Bored and intact valves do not appear to significantly differ in size (bored: median = 21.75 mm; intact median = 18.75 mm), nor are bored shells biased in handedness (right n = 8; left n = 6). The outer diameters of the boreholes range from 0.75 to 2 mm, and their inner diameters from 0.4 to 1 .5 mm. Discussion In situ observations The deep-sea octopus Graneledone pacijica and at least one species of Benthoctopus attach their eggs to hard substrate and apparently brood them during development. The high densities of brooding octopuses at depths below 2600 m at Baby Bare outcrop and Middle Valley hydrothermal vent field suggest - J R VOIGHT AND A. J. GREHAN I ..I.I, I Reported from preserved female specimens collected from hard sul'\iruii-\ an inaiillr i 'iii/m ij/ gland si-e and color in nun. ih< niimhcr iiitil .Wcf of ovarian ce.o ami Mantle le> Specimen localily & Catalog number (Juniper el <;/.. 1992: Mottl ct nl.. 1998). The benefits that accrue to hatchling and maturing octopuses from the proximiu of potential prey are clear, although mature and brooding individuals of most shallow-water species that ha\e been studied generally reduce their prey intake (Wells and Wells. 1977). Because exposed hard substrate appears to be required for egg attachment, and the Hour of the Cascadia Basin is covered by a thick layer of sediment (Davis ct til.. hihlc 2 Ihr ininl niunhi'r of K//ICS of th\u\iml < lunn - nil* < I, ,1 ,ind examined : tht number and frequency of bored valves H.'ir.l \ alv.es Total (;i) dead Year \al\es coll.' i 1 n % 1995 51 : 3.9 1997 272 14 5.1 1998 * l\pn.\ tiiTticn\ develop in about 400 davs at tem- peratures of 5' to 1 1'C (Wood et til.. 1998). Eleven females of this species, and more than 100 eggs, were found en- lolded in a single plastic sheet trawled from near Sweden (O'Dor and Macalaster, 1983). We did not observe conge- neric females sharing a single substrate at Baby Bare, but lemales of different genera weie seen within an arm's length of each other under the same ledge. The distribution of maturing females in relation to sub- strate type appears to shift toward hard substrate. Both of the nonbrooding females collected from hard substrates had enlarged oviducal glands and large ovarian eggs (Table 1 i. evidence of reproductiv e maturation ( Mangold-Wir/. 1 9ft3 ). but onlv I of the 14 trawl collected female specimens in the type series (Voss and I'earcv. 1990) shows comparable enlargement ol the lemale reproductive system. The small oviducal glands of brooding females compared to those ol EGG BROODING BY DEEP-SEA OCTOPUSES 99 females that apparently have not yet spawned (Table 1 ) may be due to postspawning resorption. The small ovarian eggs in an "obviously exhausted" female of Graneledone boreo- pacifica (Nesis, 1989) may reflect this pattern. The contrast between the reproductive maturity of females collected from hard substrate and the immaturity of those trawled from soft substrate is startling and suggests that maturing females shift habitat preference to increasingly associate with hard substrate, perhaps to secure a brooding site. If increased prey availability were solely responsible for their advanced reproductive maturity, reproductively mature fe- males would occur at a wide range of sizes rather than being among the largest specimens known in the species. If hard substrate suitable for egg attachment is limiting in the deep sea, high densities of brooding octopuses should be found at sites other than Baby Bare. Despite decades of study given to seamounts, such groups of octopuses have been unreported. The nature of the research that has been conducted at Baby Bare may make its fauna appear unduly unique. Baby Bare is one of the few non-hydrothennal vent structures in the deep sea that has been examined at depth with submersibles. The trawls, rock dredges, epibenthic sleds, and baited traps and hooks that have been most frequently used to sample sea- mounts (Wilson and Kaufmann, 1987) have focused on sea- mount summits rather than their bases. Even submersible ex- plorations have most often focused on seamourit summits. For example, the Canadian submersible Pisces IV documented the biota of Cobb Seamount at 180 m and above (Parker and Tunnicliffe. 1994). depths that are inaccessible to octopuses living at 2600 m. The bases of other seamounts surrounded by heavily sedimented seafloor must be observed to determine whether Baby Bare is unique in having high densities of brooding octopuses. Clam shell col tedious The similarity of the boreholes on thyasirid clam shells collected at Baby Bare to those drilled by shallow-water octopuses in bivalve shells (Nixon et til.. 1980; Nixon and Maconnachie. 1988) support the hypothesis that octopuses prey on the clams by drilling their shells. Two caveats reduce the confidence with which octopuses can be identi- fied as predators of the clams. First, borehole morphology and size vary with prey species rather than octopus size (Nixon and Maconnachie. 1988). Second, because Baby Bare lies beneath the depth at which calcium and aragonite passively dissolve, the boreholes may have been enlarged and their edges obscured by shell dissolution. The edges of the boreholes were poorly defined when viewed with scan- ning electron microscopy (Voight. unpubl. data). Because, however, the boreholes are generally similar to those made by octopuses and no other deep-sea predators known to bore shells were collected during any cruise to Baby Bare, octo- puses are concluded to have bored the clam shells. Whether members of one, the other, or both genera bored the clams cannot be assessed. The octopus (an individual of Bentlwc- tupiix) reported to have grasped clam shells in three of its arms (Mottl et til.. 1998) was not observed to prey on them. The presence of crushed gastropod shells in the gut of the specimen of Graneledone from Axial Volcano (Voight. unpubl. data) suggests that these octopuses would not drill the thin shells of these clams. The comparatively small size (28.5 mm or less) and thin shells of the Baby Bare clams indicate that they were bored by small octopuses. Shallow-water octopuses are known to prey on bivalves either by drilling the shell and injecting toxins through the borehole or by grasping the opposing valves with their suckers and forcing them open, usually without damaging the shells (McQuaid, 1994). Small octo- puses are more likely to drill small bivalves than are large octopuses; large octopuses are more likely to force the valves apart, leaving little evidence of predation on the shells (McQuaid. 1994). In addition, greater predation pres- sure may be exerted by small or young octopuses, which may reach very high densities when clutches hatch, than by fully mature and possibly senescent adults. Nixon's (1987) review suggests that deep-sea octopuses opportunistically take small prey, although no evidence had previously sug- gested that abyssal octopuses prey on clams. Further opportunities to directly observe the biology of abyssal octopuses may resolve questions raised by this research, including whether egg-brooding sites used by deep-sea octopuses are typically extremely localized, what is the duration of development of these large eggs, and do deep-sea octopuses drill hard-shelled prey such as clams. Acknowledgments M. Mottl, chief scientist of the 1995 cruise (supported by NSF grant OCE-93- 14632), collected the brooding female with her eggs and shared videotapes. Co-chief Scientists of the 1997 cruise, K. Becker and H. P. Johnson, and J. Cowen, Chief Scientist in 1998 and 1999, graciously made oppor- tunities available to expand earlier observations and make additional collections. R. A. Lutz provided videotape and octopus specimens from Middle Valley. C. J. Moore pro- vided photographs of the Oregon specimens. Assistance of the Alvin and Jason groups and the crews of the R/V Atlantis and Thomas G. Thompson was essential in making observations on which this report is based. We thank R. A. Zierenberg. B. A. Seibel. G. J. Massoth. B. Loetel, R. Collin, N. Becker, F. E. Anderson, and two anonymous reviewers for their help. JRV's participation in the 1997 and 1998 cruises was supported by grants from the West Coast and Polar Regions Undersea Research Center at the Uni- versity of Alaska Fairbanks and NOAA's National Under- sea Research Program; her 1999 cruise participation was supported by Marshall Field Fund of the Department of KM) J. R VOIGHT AND A J. GRKHAN Zoology. The Field Museum. AJG's panicipalion in the 1995 cruise was funded by u European collaboration grant awarded to Kim Juniper from the Association of I 'imersi- ties and Colleges of Canada. Literature filed Becker. V C.. and C. I'., \\heal. 1995. \ physical investigation of Baby Bare, a site of active j venting on the Eastern Flank of Juan de Fuca Rid-.v . ophyi I rw 76: F4I9. Bok-t/kv. s. v. 1987. Embryonic ph.' ; 1 in ( />,/ / il, Cycles. Vol II P. R. Boyle, ed ress I ondon. Bnlel/ky. S. s. 1994. F.mhryonk i of cephalopoiK al low temperatures. Antarct. Sci. 6: Davis, E. E., D. S. Chapman, M I \l..itl. W. .1. Bentkowski. K. Dadey. C. ForstiT. R. Harris. S. Nagih.ii ... k. Rohr. (i. Wheat, and M. \\ hiticar. 1992a. FlankFlux: an experimenl lo study the nature ot hydrothermal circulation in young oceanic crust. ('in hydrolhermalis gen. et sp. nov. (Mollusca. Cephalopoda: an minp^d from .1 deep sea hydrothermal vent site. d;/i. Biol. M,u 39: Id" Is,-) Grerum. \. .I, C. Mover, and S. K. Juniper. 1995. Hie enhanced biological production on Baby Rare, a basalt escarpment on the I lank ol Juan de Kidge (NF, Pacific). EOS, Amer. Gtoph\s. I'nimi ,\h\i. 76: III" llwhberg, F. G., M. Nixon, and R. B. Toll. 1992. Ordei Octopoda. Pp. 213-279 in "Uin-ul" and Juvenile Cephalnpod\: A Manna/ I i l/irn Idcntiluution. M. J. Sweeney. C. 1 . I; Roper. K. M Man;jold. M. R. Clarke, and S. v Bolel/ky. eds. Smith. Contnli /,<,<< 513. Juniper. S. K.. V. Tunnii-lifTe. and E. C. Southward. 1992. llydroiher- . enls in turhidite sediments on a Northeast Pacilic spreading centii- Organ! ms and substratum al an ocean drilling site Can. .1 /.ool. 70: |7<>2 1X09. Mangold-\Vir/. K. 1963. Biologic dcs Ccphalopodcs hcnlhiqiics et nec- toniques de la Mer Catalane. Vie Milieu Supp. 13: I ^Sy Massoth. (i. J.. K. I. Baker. K. \. leel>. I). A. Butterfield. R. K. Kmhley. J. E. l.iiplon. R. K. I homson. and (i. \. Cannon. 1995. Observations of manganese and ison al the CoAxial seafloor eruption site. Juan de Fuca Ridge. GY/i/i\\ K, \ hit 22: 151 - 154. Mc(Jiiaid. C. I). 1994. I eedm-j In h.u MMII and selection of bivalve pre> by Ociopus vulgari* C'u\ier. J. .v/i. Mm Iliol I.,<>1. 177: 1X7-202. Moisec\. S. I. 1991. Ohseri.it ion ot ilu- \eitieal disinbulion and hclun 101 ol neklomc si|inds using manned suhmersihles. Hull. Mar. Sri. 49: Hi, Mottl. M.. (,. Ulu-at. I), kadko, K. Sansone. (i. Massolh, A. Gri-lian. ( . Mo>er, E. Da\is. B. Uaker. R. I i-i-l\. M. l,ille>. J. Geiidron. G. l.i-bon. I-. Olson. S. \\alkrr. and N. Bei-ker. I'W5. \\aim -pinigs .liscou-ied M.I Crust. liah> Hare oiileiop. l-.asiein I lank ol .In. in de I uca 1995. EOS. Amer. Gen/ilny Ini.m \hX. Warm springs discovered on 3.5 Ma oceanic crust, eastern I),. Hica Ridge. li>)>\ 26: 51 s| Nesis. K. N. I9K2. /(// l>,,n;ni,M-\ ,////,*> / ,h,- World Ocean I ighl ami I : Indll tT 1'iil'lisliing House. Moscow 360 pp. [In Rus Nesis, K. N. I9K'*. I ,,| Hie Okhotsk Sea. dislnhulion and biology of non-coaslai (,s : i Ni\i>n. M. 1987. Cephaloplm;iii 35: ^5| 159 Voight. .1. R. 1997. Clailistic analysis ol the oitopods based on anatom- ical chaiactets ./. A/,//in, ,m Siinl. 63: 31 I 325. Voight. J. R. 1998. Biological investigations ol the genus Graneledone from abyssal and haihyal depths ot the N'oiih Pacilic Ocean. P. 344 in \h\ti, i, t\ \\mlil Congress of Malacology, Washington. DC.. R. Bieler and P. M. Mikkelson. eds. I 'nit. is Mal.uologica. \ oss. (i. I.. 1988a. The hiogeography ol the deep-sea Octopoda. Mala- cologia 29: 2'5-307. \ oss. (i. I.. 19881). 1 v olution ami pliy logenelic relationships ol deep-sea oclopods i ( 'mala and Incirrata). Pp. 253-276 in The Mollusca. Vol 12. l'iil<-iiiitiilov\ ami \eoniol,>gy of Cephalopods, M. R. Clarke and E. R. lineman, eds. Academic Press. San Diego. \oss. (i. I... and \\. IVarcv. 1990. Deep-water octopods (Mollusca: Cepha- lopoda) of the Northeastern Pacilic. Pror. Calif. Acad. Sri. 47: 47-94. \\ells. M. J., and J. NVells. 1977. Cephalopoda: Octopoda. Pp. 291-336 in Reproduction of Marine Invertebrates Vol. IV. Molluscs: Gastropods anil Cephalopods. A. C. Giese anil .1 S IV.iise. eds. Academic Press. New York. Wilson. R. R. Jr., and R. S. kaiilmann. 1987. Seamounl biota and hiogeography. Pp. 355-377 in Seamtiunts, Islands and Atolls. H II Keating. P. Fryer. R. Bati/a. anil (i. VV Bcohlcn. eds. AGU Geophys. Monog. 43. American Geophysical Union. Washington. DC. \\ I. .1. U., E. kenchington. and R. k. O'Dor. 1998. Reproduction and embryonic development time of Balhypolypus arctirus, a deep-sea oclopod (Cephalopoda: Octopoda). Malarologia 39: II 1" >, oiiiiu. R. E. l'>72. Brooding in a hath) pelagic octopus. l\u. Sri. 26: 400-404. Young. R. E. 1995. Aspects of the natural history of pelagic cephalopods ol the Hawaiian Mesopelagic-Boundary region. I'ac, Sri. 49: 143-155. Reference: Biol. Bull. 198: 101-109. (February 2000) Filamentous Fungi Associated With Holothurians From the Sea of Japan, Off the Primorye Coast of Russia MICHAEL V. PIVKIN Pacific Institute of Bioorgamc Chemistry. Fur East Branch of the Russian Academy of Sciences, 690022. Vladivostok. Russia Abstract. Holothurians (Holothurioidea, Echinodermata) are known to contain triterpene glycosides, which show antifungal activity. Nevertheless, fungi can be isolated from all organs of holothurians. During 1995-1996. mycelial fungi from several Far-Eastern holothurians Apostichopus japonic/is. Eupentacta fraiidutrix, Cuciinuiria japonica were collected from the Sea of Japan near the coast ot Primorye (Russia) and studied. Twenty-seven species of marine fungi, mostly facultative ones belonging to the mi- tosporic fungi, were isolated from the holothurians and identified. Fungi isolated from the holothurian surface were more diverse and abundant than those from internal organs and coelomic fluids. Of the holothurians studied. Ciicum- aria japonica was poorest in abundance and diversity of fungi. The fungi Cladosporium brevicompactum and C. sphaerospermum were common in the holothurian coelom. Because of their high proteolytic activity, these fungi may be pathogenic to holothurians. The detritovorus holothurian A. japonicus was shown to modify the fungal assemblages within the marine bottom sediments. Introduction Interest 'in the interactions between echinoderms and other species of marine organisms has intensified in the past few years. In addition to the trophic interrelationships of echinoderms within marine communities (Levin and Voronova. 1979) and diseases caused by bacteria, protists, and other parasites (Jangoux. 1987; Skadsheim etal.. 1995), benign, symbiotic interactions between echinoderms and bacteria have received much attention (Bauer and Agertr. Received 21 April 1998: accepted 10 September 1999. E-mail: Elyakov@stl.ru 1994; McKenzie and Kelly. 1994; Newton and McKenzie, 1995: Kelly and McKenzie, 1995; Kelly el al. 1995; Thorsen, 1995). However, there is very little information about the fungi of echinoderms. Mortensen (1909) de- scribed a peculiar disease in the Antarctic cidaroid echi- noids, genera Rhynochocydaris and Ctenocidaris, which was caused by the fungus-like organism Echinophyces mirabilis. More recently, fungi were found to damage the spines of the sea urchins Diadema antillarum (Mortensen. 1940) and Strongylocentrotus franciscanus (Johnson and Charman. 1970). There is some evidence that fungi are present in the intestine of a sea cucumber. Apostichopus japonicus. from the Sea of Japan, but no data are available on their abundance or taxonomy (Levin, 1982). Despite the lack of data, the presence of fungi in holothu- rians is of potential importance. Holothurians are known to contain triteipene glycosides, which show fungitoxic, he- motoxic, and cytotoxic activities (Stonik and Elyakov, 1988). The diverse biological activities of the triterpene glycosides depend upon their specific binding to A-' sterols (Kalinovskaya etal. 1983; Kalinin et al. 1994, 1996). Such sterols are common in most fungi, but holothurian fungi must be assumed to contain some other sterols that will not bind to holothurian glycosides. In addition, fungi adapted to the effects of triterpene glycosides might produce secondary metabolites having a similar type of action. Thus, the iso- lation, in pure culture, of fungal strains from holothurians, and a definition of their species diversity would give us an insight into the biochemical mechanisms by which the fungi adapt to triterpene glycosides. This study had three aims. The first was to establish that fungi are found in and on holothurians a surprising phe- nomenon since holothurians are well known as producers of fungitoxic glycosides. The second aim was to prepare a list 101 1(12 M V PIVKIN of t'ungi isolated from the holothurians collected along the coast of the Primorye territory in the Sea of Japan. Such a list is basic to any mycological research. In mycologv. the substrate from which a fungus was isolated, as well .is Us area of distribution, are characteristics used in identitu .iiion. as are the morphological features and genetic distinctions of any species. The final goal was to be able to speculate rationally about the role of holothurian fungi in the envi ronment. Materials and Methods Sampling We studied the fungi from three species of Far-Kastern holothurians: Apostichopus juponictts. l-Mpcninctti fnnuhi- trix, andCucumariajaponica. In 1995-1996. 10 animals of each species were collected at each of three sites along the Russian coast of the Sea of Japan: i 1 ) the Gulf of Opritch- nik. near Cape Skalistyi i44 26'3"N, 135 59'5"E; depth 12-14 mi: (2) the Primorye coast (45 51'1"N. 13637'3"E; depth 111 m): (3) in Trinity Bay of Peter the Great Bay (4234'<> V I30 C '57'58"E: depth 1.5-15 m). Specimens were placed in sterile polyethylene bags and either pro- cesscd immediately or stored in a refrigerator at 0-3C for not more than 4S h. Two additional t\pes of material were collected at site 3. Feces from A. juponiciis were obtained by placing live animals in sterile polyethylene cages suspended in the sea at the site for 6 h. then collecting the fecal material in sterile poKetln lenc hays Bottom sediments from the habitat of this species were also collected in sterile hags. Isolation of fungi Fungi were plated anil cultured from !i\e t\pcs of male- rial: from three holothurian tissues surface muscle, intes- tine, and coelomic fluid and from feces and bottom sedi ment. Holothurian specimens were washed five times in sterile seawater. and 0.5-ctrT samples of surface muscle and inici N. il organs were dissected. Pieces of these tissues were placed on HA medium (Table I) in sterile petri dishes. Coelomic fluid was extracted with a sterile s\ringc and inoculated into HA medium. Fungi from bottom sediments. fetal material, and stomach contents were isolated by dilu- tion plaint'.' (Sicele. 1967). Colonies from all ol these cul- tures then it-seeded into tubes of Tuhaki's medium (Table I ). riiiif>nl iii: 'H and ilivcr. g ol streploimcin added after sicnli/ution. M I '< \ Potatoes-carrot agai *\iih feawoter and yeast extract: 20 g of potatoes and 2(> g ol carrot boiled for 30 min in a litei ot se.iw.iiei. t rushed, brought up to a liler of distilled water: n ^ - "I \east extract and 20 g "I ag.ii added: pH adjusted to 7.5 with NaOH: sterilised at 121 (' lor 20 min. CM Gelatinous medium: 2 g ot gelatin in UK) ml of seawater: pH adiusted lo (> s with NaOH; sterili/ed at 121 C for 20 mm iBilas. l l lN2i MA Malt seawaler agar: S'yr malt extract agar made up with seawater; pll adjusted to 7.5 with NaOH; sterili/ed at I I2"C for 30 mm iHulc ; iiilturcd and idenlilied lo species using slide culliii is well .is moipholovit al t haiat leristics of the colony. Us nit- cultures allows microscopic examination of intact tiiivj.il icprodiictKc structures. In ihis Production of extracellular proteases was determined by i'tow ing three cultures of each isolate at 22 "(' on GM media (Table I) in tubes of IS mm diameter (Hilay. 1982). Liq- iichmg /ones were measured in 7 and 14 days. HOLOTHURIAN FUNGI 103 Resistance of fungi to glycosides Resistance of fungi to the action of glycosides extracted from the holothurian Eupentacta fraudatrix was detected by the agar diffusion method. The following fungi were tested: Cladosporium sphaerospermum from coelomic fluid of E. fraudatrix, Aspergillus eburneocremeus from the surface of the holothurian, Cladosporium cladosporioides from bot- tom sediments, Aspergillus flavipes from the alga Fucus evanescens, and Candida albicans KMM 245. Solutions of the glycoside fraction at concentrations of 0.625. 1.25, 2.5, 5, 10. 20. and 40 mg/ml were used in the tests. A total fraction of glycoside, common to these animals (Kalinin et nl., 1994) was kindly provided by Dr. Avilov of the Pacific Institute of Bioorganic Chemistry. Statistical analysis The statistical significance of the proteolytic activity of the holothurian fungi and their resistance to holothurian glycosides was analyzed by the Fisher exact probability test (Sokal and Rolf. 1981 ). The results of the investigation of fungal species diversity were evaluated statistically using Student's t test (Masurran. 1988). Results Species diversity of fungi from holothurians Twenty-seven species of marine fungi were isolated from various organs and tissues of holothurians and identified (Table 2). Most of these isolates were mitosporic fungi, predominantly Hyphomycetes (24 spp.) and Coelomycetes (3 spp.). The obligate marine fungi (definition according to Kohlmeyer and Kohlmeyer, 1979) included three species: Botryophialophora sp. (Hyphomycetes). Phialophorophoma sp., and Coniothirium ohiones (Coelomycetes). The facul- tative marine fungi of holothurians were represented by 24 species of "terrestrial" fungi. All of these species belong to 13 genera of the class Hyphomvcetes. Among the holothurians investigated, the mycota of Eu- pentacta fraudatrix were the most diverse. Cladosporium oxysporum and Metarchizium anisopliae were the fungi that occurred most frequently on the surface of these animals. Cladosporium sphaerospermum was the dominant species inside of E. fraudatrix and on the surface of Apostichopus japonicus and Cucumaria ja/wnica. Cladosporium brevi- compactum dominated inside of A. japonicus and C. ja- ponica (Table 2). Three species of fungi, Pacilomyces puntonii, C. sphae- rospermum, and Beauveria alba, occurred on the gonads of A. japonicus and E. fraudatrix. Proteolytic activity of some fungal strains from holothurians The proteolytic activities of some holothurian fungal strains are listed in Table 3. Of the eight fungal species taken from the surface of the holothurian Eupentacta frau- datrix, only one, Penicillium commune, had no proteolytic activity. However, the congeneric species P. hen/uei had the highest rate of gelatinous medium liquefaction when com- pared to other fungal species found on the surface (although the rates for Acremonium striatisponnn and Altemaria al- lernata were also relatively high). Fungi of the species Cladosporium sphaerospermum and C. brevicompactum isolated from internal organs of these three holothurians demonstrated higher gelatinolytic activity than the same fungal species isolated from the surface. Indeed, the activity demonstrated by C. brevicompactum taken from the internal organs of Cucumaria japonica was the highest measured in these experiments. Resistance of fungi to glycosides of the holothurian Eupentacta fraudatrix The inhibition of fungal growth by different concentra- tions of a glycoside fraction taken from E. fraudatrix is plotted in Figure 1. These glycosides inhibited Candida albicans and fungal strains from bottom sediments and from Fucus evanescens (curves Cal, Ccl, Afl) at the lowest con- centration (0.625 mg/ml): but the growth of the fungus Aspergillus eburneocremeus, isolated from the surface of this holothurian, was stopped only at glycoside concentra- tions 5 mg/ml or higher (curve Aeb). Cladosporium sphae- rospermum, isolated from internal organs (curve Csp), was resistant to these holothurian glycosides at 40 mg/ml. a concentration many times higher than it could be in this holothurian (Kalinin et ai. 1994, 1996). That fungi from various marine habitats differ in their resistance to the glycosides of E. fraudatrix is illustrated in Figure 2. The fungi from this holothurian (top dishes) are more resistant to its glycosides than are fungi from bottom sediments and from Fucus evanescens (bottom dishes). Fungi in the intestine, feces, and habitat of the holothurian Apostichopus japonicus The abundance and species composition of fungal prop- agules (conidia, spores, and pieces of mycelia) in the intes- tine of A. japonicus. in its feces, and in the marine sediments it inhabits are set out in Table 4. The abundance of fungi in the bottom sediments was ten times higher than in the intestine of this sea cucumber. However, the abundance of fungi in feces was twice that in samples of bottom sediment taken from the habitat of these holothurians. The species composition of fungi is also different in the sediment, in- 104 M V PIVKIN 1 .ihk- 2 Fungi from holothurians: species diversity and distribution Taxon kupenuicui fraudotrix Apostichopus japimii u\ Cucumaria japonico Cladosporium sphaerospermum Pen/. C brevicompaclum Pidopl. et Deniak C atnapermum Pidopl. et Deniak C. oxysporum Berk, et M. A. Curtis X Allrrnaria allernulti (Fr. ) Keissl. Aspergilliis versicolor (Vuill.l Tiralv .4. ebiinieocremeus Sappa + Epicoccum st. Phoma sp. Ulocladium sp. Aspergilliis fiavus Link \. rt-inimium charticula il.mdaui \V. (iams \ tiiMilioides (Nicoti V\ (iams + A. strialisponim (Onions & G. L. Barron) \V. Ciams Beauveria alba iLimberl Saccas Botryophialophora sp. + Coniolhiriiim obioncs Jaap* Metarchi:ium aiiisopliae (Metschn.) Sorokin \.n anisopliae i R. Johnst. X Oidiodendron sp. + Pacilomices /'iiiiliiini (V'uill.i Nann. Penicillium commune Thorn + P. herqnci Bainier et Sartory + /'. im/i/ii ilium Biourge + /'. rtn/iicti>rii Thorn + /' f/tr/jWm'i Schmotina et Golovleva + Phialophorophoma sp.* :/idium sp. I . nit iltiiim tenerum Ness + 25 3 + X 9 2 Total number tShannon-Weiner diversity index Equitabilit) 27 24 2.95 0.43 s 1.67 0.76 3 0.82 0.74 Letters (S. I. Cl indicate the huloihurian site Irom which the tuns:i were isnl.ticd S. Mirt.nc. I. mteinal nrs;ans; C. coclomic lluid. Symbols indicate the frequency of fungal species occurrence: +.0%-10%: X, 10%-20'V. . 2(i'. W%; mmc ih.in Hi', Huldface numbers give summary information about the distribution of the fungi among the linlntliuiians. l-m example, ol the 27 lungal species idenlilied. 25 were found on Eupentacta fraudatrix; 2 of the 1 1 species found on Apostichopus japonicui were exclusive to that species, and l > weie also louiul on / fraudatrix', and 3 species ot tungi were found on 'unnn iiiponica. none exclusiseh Obligate species ni m.iniiv tunyi ^ Differences in species diversit\ ol lungi on the holoihurian species air siaiisikalh sigmlic.ini at the 'W.S'.r level. -. and luces. The intestine has the fewest species, but all of them occur in the sediment and in the Icxvs. Discussion The vast m.ii'Miix nf t'unyi isolated from the holothurians were faculia!i\ . marine species. Only three species of ob- ligate martn A civ found in the holothurians. and all of them were isol.ik-d Irom the surface ol l:ii/>i-iiliii-l o 0,4 0,2 0,625 1,25 2,5 5 10 Concentration of glycosides, mg/ml 20 Figure 1. Inhibition of fungal growth by glycosides of the holothurian Eupentacta fraudatrix. Curves: Csp, Cladosporium sphaerospeniium isolated from coelomic fluid of the holothurian. Aeb. Aspergillus ebumeocremeus isolated from the surface of the holothurian. Ccl, Cladosporium cladosporioides isolated from bottom sediments. Afl, Aspergil/us flavipes isolated from the alga Fucus evanescens. Cal, Candida a/bicans KMM 245. Differences in resistance of fungal strains to glycosides are statistically significant at the 99% confidence level. P < 0.00 1 . 1(16 M V. PIVKIN Tuble 4 Abundance and fungal specie* n>mpi>\iti<'n "I the internal contents anil feces / Aposlichopus japonicus. ,j // .;> tin- marine sediments it inhabits Materials Examined* Bottom sediments InKMiiie 1 t\ fs Fungal taxon (1.3 x Iff 1 ) I0 3 ) (2.6 x I0 4 ) Acremonium fiisidioidei it) W. Gams A. kiliense Grut/ A. iriicli\caulon \V. Gams Allernaria alternata (Fr. ) Keissl. Arthrynium sp - i//i/i tfuuis Link \ Ihtli'i'lulii ii\ C'. M Chi.. Paravizas et C. R. Bej. /A. n/'#fr Tiegh. A. sydowi (Bamier et Sarturv I Thorn et Church Chaetomium olivaccum Cooke et Ellis Cladosporium alrospermum Pidopl. el Ueniak C. brevicompaclum Pidopl et Demak C. sphaerospermum Pen/. ryphiella arenaria \\^^\ :< ndron trunitintin G. L. Barren / Hium brevicompactum Dierckx P. chrysogenum Thorn P. commune Thorn P. herqtiei Thorn P. thomii Maire Stilbella aciculosa (Ellis et Everh.) Seilert Trichoderma aureonruli' Rifai T. harzianum Rifai T. viride Rifai Wallemia sehi (Fr.) Arx Number of species 23 17 Numbers in parentheses art- the number ol lunj!al prnp.iL'iiles per grain i 'I material. have more tolerance tor these glycosides than did tungi from other marine substrates (Figs. 1. 2). This suggests that fungicidal Hycosides may not limit the ability of certain species of fungi to grow on a holothurian. More | of fungi inhabit the surface of bolothurians than the in: MIIS and coelomic lluid (Table 2). The mycota of holoihuiians i\picall\ comprises unspecitic cos- mopolitan species ili.it occur in soils and on various protein- rich substrates ("fable 5), whereas the internal fungal assemblages of hololhuriaiis are raiher specilic. Thus, three of seven fungal species isolated from internal organs and coelomic lluid were found in only one holothurian species (Table 2). Four other species were characteristic of the internal mycota of two holothurian species. But none of these fungal species occurred in all three species of these holothurians. Pacilomyces puntonii and Beauveria alhu are rare spe- cies, found on human skin (De Hoog. 1972: Samson, 1974): they have been isolated from the gonads of A. japonicus and . fraudutrix. respectively. The cosmopolitan species Cla- il\l><>riiim sphaerospermum, which occurs on various sub- strates, including ones of animal origin iHllis. 1 97 I i. was also isolated from the gonads of '. fraudutrix. Even though the amount of glycosides in gonads can be ten limes more than in other organs of these holothurians (Le\in. 1 C )S2). these species of fungi were found even in such adverse conditions. I he fungal species from E. fraudatrix are more diverse than those from any other holothurian species studied: and as indicated above, glycosides from this holothurian have the highest fungicide activity. Triterpene glvcosides from C. juponicii are active against Candida alhicanx and C. tropi- ctilis at concentrations higher than 30-50 mg/ml. hut they have lower fungicidal activity than other holothurian gly- cosides (Kalinin ft al.. 1494). We might expect that the toxicity of secondary holothurian metabolites would inllu- ence the species diversity of the fungi associated with these animals; hut in fact, the diversity of fungi increases with growing glycoside toxicity. The investigated holothurians differ in food specialisa- tion. A. japonicn*. is an animal that feeds on organic material that has settled to the marine floor. /.'. fraudatrix and C. japimica feed on organic particles suspended in the seston. Because the number and diversity of fungi in the bottom sediments are about 100 times higher than those in the water column i Steele. 1967). A. japonicus has contact with a much more diverse fungal assemblage than the two other hololhu- riaiis encounter. Nevertheless, this condition does not pro- v ide the greater diversity of fungi associated with A. ja- ponicus. Depth ol habitat might also affect fungal diversity. I'he most deep sea species. Cucumaria japonica, has the fewest fungal species. A. japonicus is common at depths down to 100 m and has more diverse mycota than C. jiiponicu. The greatest diversity and equitability of fungal species are characteristic of /.'. fraudatrix. which has an optimal depth of 1.5 m but can occur down to 40 m. Therefore, fungal diversity appears to decrease w ith the depth of the holothu- rian habitat. The regions of isolation of these holothurians are also important. In our experience, the species composition of fungi isolated from C. japonica collected Irom areas around the Kuril Islands in I UI )N was different Irom that isolated Irom C. japonicii collected in Primorye waters, fungi that HOLOTHURIAN FUNGI Table 5 Fungi isolated from hololhurians: distribution, habitats, substrates 107 Taxon Substrates and distribution References Acremonium charticola A. fusidioides A. striatisporum Alternaria altematu Aspergillus eburneocrerneits A. flavus A. versicolor Beauveria alba Botryophialophora sp. Cladosporium atrospermum C. brevicompactum C. oxvsporuiu C. sphaerospermum Coniothirium obiones MeUirclii^ntm anisopliae var. anisopliae Oidiodendron sp. Penicilliitm commune P. herquei P. implicatum P. roqueforti P. skrjabinii Pticilomices puntonii Phialophorophoma sp. Tilachlidium sp. Ulocladium sp. Verticillium lent 1 rum Cellulose substrates, human skin (-Cephalosporium ba/Iagii). widespread, unknown in the sea. Soil, widespread, unknown in the sea. Soil, widespread, unknown in the sea. On various marine substrates, widespread. Soil, Southeastern Asia, North and South America. Soil, humans, and animals, seawater, bottom sediments, and coastal sands, widespread. Soil, (-A. versicolor var. glauca) pathogenic for man and animals. bottom sediments, seawater, widespread. Insects, human skin, widespread, unknown in the sea. Allied to Botryophialophora marina that inhabits wood, test panels, and coastal sands in the Atlantic and Pacific Oceans. Soil, cosmopolitan on various substrates, widespread, unknown in the sea. Soil, cosmopolitan on various substrates, widespread, unknown in the sea. Soil, cosmopolitan on various cellulose substrates, widespread. unknown in the sea. Soil species, cosmopolitan on various substrates including ones of animal origin, widespread, unknown in the sea. Saprotrophic species. Atlantic and Indian Oceans. Insects, widespread, unknown in the sea. Soil, plants, and others. Soil, contaminants from cheese and other foodstuffs, bottom sediments, seawater, widespread. Soil, widespread, but not abundant, unknown in the sea. Soil, widespread, but not abundant, unknown in the sea. Soil, contaminants from cheese and other foodstuffs, widespread. unknown in the sea. Soil. Far East, vicinities of Blagoveshchensk, unknown in the sea. Human skin, widespread, but not abundant, unknown in the sea. Allied to Phialophorophomu litoralis that is found on cellulose substrates, and on the dead alga Avicennia marina var. resinifera. the Atlantic and Pacific Oceans Soil, plants, and others. Soil, plants, and others, unknown in the sea. Soil, widespread, unknown in the sea. Gams, 1971 Gams, 1971 Gams. 1971 Johnson and Sparrow, 1961; Jones, 1962. 1963. 1968. 1972 Artemtchuk, 1981; Steele. 1967; Bilay and Koval, 1 988 Thorn and Raper, 1945; Bilay and Koval. 1988; Steele, 1967 Bilay and Koval, 1988; Blochwitz, 1934; Thorn and Raper, 1945; Steele, 1967 De Hoog, 1972 Linder, 1944; Steele, 1967; Wilson, 1951 Ellis. 1971 Ellis, 1971 Ellis, 1971 Ellis, 1971 Kohlmeyer and Kohlmeyer, 1979 Latch, 1965; Tulloch, 1976 Ellis, 1971 Ramirez, 1982; Raper and Thorn, 1949; Steele, 1967 Raper and Thom, 1949 Raper and Thom, 1949 Ramirez, 1982; Raper and Thom. 1949 Shmotina and Golovleva. 1974 Samson, 1974 Kohlmeyer and Kohlmeyer, 1979 Ellis, 1971 Ellis, 1971 Gams, 1971 were prevalent in isolates from Primorye holothurians were very rare in the cucumbers from the Kuril Islands, where they occurred only in the southern collection sites. Proteolytic activity is frequently suggested as a factor in fungal pathogenicity (St. Leger et ai. 1988; Monod et al., 1995). The body of any holothurian contains 1 .9%-9.0% of protein, mainly collagen, and the content of protein exceeds that of all other components of the holothurian body, ex- cluding water (Levin, 1982). Therefore, the ability of fungi to damage holothurian tissues can be estimated from their proteolytic activity, thus precluding the need to inoculate the animals. In particular, the gelatinolytic activity of ho- lothurian fungi, determined simply by measuring the lique- faction of gelatinous medium (denatured collagen), allows us to estimate their pathogenicity. The strains of Cladospo- rium sphaerospermum and C. brevicompactum isolated from the internal organs of these holothurians are twice as active in liquefying gelatin as are isolates from their sur- faces (Table 3). Note that the fungi of these species often occur in internal organs and coelomic fluid of holothurians; in several cases, the animals contained a monoculture of these fungi species. Recall, furthermore, that C. sphaero- spennum is resistant to glycosides of E. fraudatrix (Fig. 1). These facts indicate that C. sphaerospermum and C. brevi- compactum may be pathogenic for holothurians. However, the terms "pathogen" for these fungi and "host" 108 M. V. PIVKIN for those holothurians must be used with care, because (lies imply a parasitic relationship between organisms in this case between a macroorganism and a pathogenic micinni ganism. To confirm the pathogenicity of a microorganism. the classical method of Koch should be used, and this includes three steps: ( 1 ) a search for symptoms oi a disease: (2) isolation of potential pathogens into pine culture: (3) artificial inoculation of this organism. In ilic end. a micro- organism can be recognized as pathogcm> it all the symp- toms of a recogni/able disease produced by infection com- pletely coincide with those th;,t ippeai alter artificial inoculation. Such an analysis was ! ol possible in the present study. Therefore, an analysis "I gelatinolytic activity was taken as indirect evidence for the ability of a fungus to damage holothurian tissues, and on that basis, fungi isolated from coelomic fluid and internal organs were considered to be potentially ptitli<>nicit\ feeds on marine deposits (Levin. 19X2). which include propagules of fungi. A com- parison of fungi from bottom sediments and from holothu- rian feces revealed that A. japonicux, armed with the tritcr- pene glycosides. increased the abundance of dematiaceous species of Hyphomycetes and eliminated some species of Moniliaretie from the bottom sediments (Table 3). Dema- tiaceous Hyphomycetes of the genera Alternaria, Arihrv- nium. and Dentlryphiellu passed through the alimentary tract of A. japonicus practically without loss of their viabil- ity; and the number of fungi of the genus Cladosporium even increased under these conditions. But not all dematia- ceous H\phomvcete\ are unaffected by transit through a holothurian. For example. Wnllemia schi was isolated from bottom sediments, but occurred in neither the intestine nor iii fe< es of the sea cucumber. Hyphoi\< < /cs of the family Miiiiiliiiicdi' were affected more selectively: e.g., fungi of nera \* >< /nnniiiin and Trichnilcniiti were resistant to passage through the sea cucumbers, whereas fungi ol the genus A\prrt;illus were not found in feces of these animals. Of all the fungi of the genus fi-nicilliiini found in bottom sediments, only /' ln-n/nci and /'. commune did not lose viability in the intestine of these sea cucumbers. Fungi from Imloihuiians have never been investigated before. This pionen icport describes the species composi- tions of fungi from holoiluirians and their possible ecolog- ical role. Acknowledgments I acknowledge, with gratitude and affection, the generous help 1 have received from Prof. M. Jangoux in obtaining information on diseases of Echinodermata. 1 am most grate- ful to Miss N. Shepetova for helpful discussions and for translating this article into Fnglish. Dr. Kalinin supplied much constructive criticism for which I am grateful. My special thanks to Dr. Yu. Jakov lev for collecting some of the animals and to Dr. S. Avilov for providing glycoside sam- ples. This work was financially supported in part by grant #99-04-49363 of the Russian Foundation for Basic Re- search and by a grant of the Governor of Primorye. Literature filed ArlemU-hiik. N. .1. 1981. Mycoftuni train V<; <>t I'SSR. Nauka. Moscow. (in Russian). Bauer. .1. C'., and t'. J. Ajjertr. 1994. Isolalion nl potcnlially pathogenic- flora from tropical sea urchins in selected \\esi Atlantic and East Pacific sites. Hull. Mar. Sci. 55: 142-150. Bilav. V. I., ed. 1982. Methods of Experimental Mycology. Naukova Dumka. Kiev (in Russian). Bilaj. V. I., and K. /.. Koval. 1988. .\\iiergillii. Naukova. Dumka. Kiev. I in Russian I. Kim hvtii/. A. 1934. Die Gaining Aspergillus. Ill Neiic Spe/ies. Vari- anien und Mutantcn der Konidicnfarhc. Ssnoinmc und mteressante Standorte. Ann. Mycol. 32: X2 S'l Di 1 Iliiof*. (i. S. 1972. The genera Hi-mncrui. /WJMlinin\i , i, \ Commonwealth Myco- logical Insiilule. Kevv, Bngland. (iams. \V. 1971. Cephalosporium-artige Si lumim-li>il;e t Hy/>homyce- h \i. VF.B Gusiav l-ischer Verlag. Jena, (iciinanv II>d. K. D., C. A. Farranl. and K. B. (;. Jones. 1987. Isolation and culliire of marine fungi. Hoi. Mar. 30: 291 .W. .laiigonv. M. 1987. Diseases ol 1-xhmodermala. 1. Agents microorgan- isms and protistans. l)is. At/nut. Orf>. 2: 147-162. Jolinsiin. I*. '!'.. and I'. A. Charman. 1970. Inleelion uilh diatoms and other microorganisms in sea urchin spines (Strongylocentrotus fran- i/M(/fi/n). ./. Inverlebr. Puthol. 17: 94-106. .Inhnson. T. \V.. Jr., and K. K. Sparnm. Jr. |9hl. l-'ungi in Oceans and I \imo-if\ Weinheim. New York. Jones, E. B. G. 1962. Marine lungi Inun.Hi \fycol Sot 45:43-114. Jones, K. B. G. 1963. Marine lungi II Ascoiiueeies and I 'euteromycctcs from suhmerged wood and drill Spurn/hi. iiun\ Hi: Mycol. Sac. 46: 1 35- 1 44. Jones, K. B. (i. 1968. 1 lie disiiihulion of marine fungi on wood suh- merged in the sea. Pp. 460 485 in Himli'icrioniiion of Materials. A. H. Wallers and J. l ; . Llphick. eds I Kexiei. Amsierdam. Jones, K. B. (J. 1972. I lie deca> of limber in aquatic environments. Hi: Wood Pres. Assoc., Anna. Com'. I I S Kalinin. \. I.. \. S. l.e\in. and \. A. Stonik. 1994. /'/ Chemical Morf>htilog\: Trilerpene ClyciisiJes of Sea Cucumbers (Holothuri- oidea. Echinodermata), Ualnauka. Vladivostok, (in Russian). Kalinin \ . I.. M. M. \nisinim. N. (i. l'n>kolie>a, S. A. Avilov, Sh. Sh. Mialiillm. and \ . A. Stonik. 1996. Biological aclivities and biolog- ical role of triterpene glycosides from holothuroides (Echinodermata). 1'p I (4 LSI in Kcliimiilerm Studies. Vol. 5. M. Jangoux and J. M. Lawrence, eds. Balkema. Rolterdam. HOLOTHURIAN FUNGI 109 Kalinovskaya, N. I., T. A. Kuznetsova, A. M. Popov, A. S. Antonov, and G. B. Elyakov. 1983. Steroid metabolites of the Far-Eastern Holothurian Stichoptis japonicus Selenka. Comp. Biochem. Physiol. B 76(1): 167-171. Kelly, M. S., and J. D. McKenzie. 1995. A survey of the occurrence and morphology of subcuticular bacteria in shelf echinoderms from the north-east Atlantic Ocean. Mar. Biol. 123: 741-756. Kelly, M. S., M. F. Barker, J. D. McKenzie, and J. Powell. 1995. The incidence and morphology of subcuticular bacteria in the echinoderm fauna of New Zealand. Biol. Bull. 189: 91-105. Kohlmeyer, J., and E. Kohlmeyer. 1979. Marine Mycologv. Academic Press. New York. Latch. G. C. M, 1965. Metar(r)hirizium unisopliue (Metsch.) Sorok. strains in New Zealand and their possible use tor controlling pasture inhabiting insects. N. Z. J. of Agric. Res. 8: 384-346. Levin, V. S. 1982. Japanese Sea Cucumber. Dalnauka, Vladivostok (in Russian). Levin, V. S., and E. I. Voronova. 1979. Assimilation of bacterial food by Japanese sea cucumber. Pp. 121-123 in Material IV All Union Colloquium on Echinoderm. Tbilisi. Rep. of Georgia (in Russian). Linder, I). H. 1994. I. Classification of the marine fungi. Pp. 401-433 in Farlowia. 1, E. S. Barghorn and D. H. Linder. Marine Fungi: Their Taxonomy and Biology. Magurran, A. E. 1988. Ecological Diversit\ and Its Measurement. Croom Helm, London. McKenzie, J. D., and M. S. Kelly. 1994. Comparative study of sub- cuticular bacteria in brittlestars (Echinoderniata: Ophiuroidea). Mar. Biol. 120: 65-80. Monod, M., A. Fatih, K. Jaton-Ogay, S. Paris, and J. P. Latge. 1995. The secreted proteases of pathogenic species of Aspergillus and their possible role in virulence. Pp. 1081-1086 in Can. J. Bot.. Fifth Int. Mycol. Congress. Sect E-H. Aug. 14-21, 1994. Vancouver. Mortensen, T. 1909. Die Echinoiden. Dtsch. Suhpot. Ev/>. IWI-1W3. 11 (3): 1-113. Mortensen, T. 1940. A Monograf of the Echinoiden. Pan III. I Aul- oilonta. C. A. Reitzel. Copenhagen. Newton, L. C., and J. D. McKenzie. 1995. Echinoderms and oil pollu- tion: a potential stress assay using bacterial symbionts. Mar. Pollut. Bull. 31: 453-456. Ramirez, C. 1982. Manual and Atlas of the Penicillia. Elsevier Biomed- ical Press. Amsterdam. Raper, K. B., and C. Thorn. 1949. A Manual of the Penicillia. Williams & Wilkins, Baltimore. Samson, R. A. 1974. Paecilomyces and some allied Hyphomycetes. Stud. Mycol. 6: Centroalbureau voor Schimmelcultures, Baarn, The Netherlands. Shmotina, G. E., and L. A. Golovleva. 1974. A new species of Peni- cil/ium isolated from soil in the Amur region. Mikol. i Fitopahtl. 8: 530-532 (in Russian). Skadsheim, A., H. Christie, and H. P. Leinaas. 1995. Population reductions of Strongylocentrotus droebachiensis (Echinodermata) in Norway and the distribution of its endoparasite Echinomermella mat si (Nematoda). Mar. Ecol. Prog. Ser. 119: 199-209. Sokal, R. R., and R. J. Rolf. 1981. Biometry. 2nd ed. W. H. Freeman, New York. St. Leger, R. J., P. K. Durrands, A. K. Charnley, and R. M. Cooper. 1988. Role of extracellular chymoelastase in the virulence of Metha- rhi-ium anisop/iae for Manduca sexta. J. Invertebr. Pathol. 52: 285- 293. Steele, W. S. 1967. Fungus populations in marine waters and coastal sands of the Hawaiian, Line, and Phoenix Islands. Pac. Sci. 21: 317-331. Stonik, V. A., and G. B. Elyakov. 1988. Structure and biologic activities of sponge and sea cucumber toxins. Pp. 107-120 in Handbook of Natural Toxins and Venoms. A. Tu, ed. Marcel Dekker, New York. Thorn, C., and C. B. Raper. 1945. A Manual of the Aspergi/li. Williams and Wilkins, Baltimore. Thorsen, M. S. 1995. Oxygen gradients and microbiota in the gut of the irregular sea urchin Echinocardium cordatum (Spatangoida: Echino- dermata). P. 77 in Echinoderm Research. Proceedings of the Fourth European Echinoderm Colloquium, 10-13 April, London. Tubaki, K. 1969. Studies on the Japanese marine fungi. Lignicolous group. III. Annu. Rep. Inst. Ferment. Osaka 4: 12-41. Tulloch, M. 1976. The genus Metarhi-ium. Trans. Br. Mycol. Soc. 66: 407-411. Wilson, I. M. 1951. Notes on some marine fungi. Trans. Br. Mycol. Soc. 34: 540-543. Reference: Biol. Bull. 198: 110-120. (February 2000) Cellular Growth of Host and Symbiont in a Cnidarian-Zooxanthellar Symbiosis W. K. FITT /intitule of Ecology. University of Georgia, Athens, Georgia 30602 \bstract. Iho hvdroid Myrionema unihionenxe, a fast- growing cnidarian (doubling time = 8 days) found in shal- low water on tropical back -reefs, lives in symbiosis with symbiotic dinoflagellates of the genus Symbiotliniuni (here- after also referred to as zooxanthellae). The symbionts live in vacuoles near the base of host digestive cells, whereas unhealthy looking zooxanthellae are generally located closer to the apical end of the host cell. Cytokinesis of zooxanthellae occurred at night, with a peak in number of symbionts with division furrows (mitotic index, MI : 12'--20'r> observed at dawn. The Ml of /.ooxanthellae decreased to near zero by the middle of the afternoon and remained there until the middle of the next night. Densities of live zooxanthellae living inside of host digestive cells peaked following cytokinesis, whereas densities of un- healthy looking symbionts were highest just before the division peak. Mitosis of host digestive cells was highest in the evening, also preceding the peak in zooxanthellar MI. This is the first study relating phased host cell division to diel zooxanthellar division in marine cnidarians. Food vacuoles were prevalent inside of digestive cells of field-collected hydroids within a few hours after sunset and throughout the night, coinciding with digestion of captured demersal plankton. Laboratory experiments showed that food vacuoles appeared in digestive cell cytoplasm within 2 h of feeding with nauplii of Anemia. The number and size of food vacuoles per digestive cell and the percentage of digestive cells with food vacuoles all decreased 5-7 h following feeding in laboratory experiments, and by mid- day in field-collected hydroids. Light and exi<"n:il I nod supply were important in main- taining phased division of the symbionts, with a lag in response time !< Dimeters of 1 1-36 h. Altering light Received 17 June I'd 1 ' November 1999. E-mail: fitt9sparTOW.eC! ' du. and feeding during the night did not influence the level of the peak MI the next morning, though in one experiment the absence of light slowed tinal separation of daughter cells at the end of cytokinesis. In another experiment, hydroids starved for 3-7 d and "pulse-fed" Artciiiiu nauplii for 1 h at the beginning of the dark period showed continued low symbiont division (<5%) after 1 1 h. whether maintained in constant light or darkness, implying that most algal division is set more than 24 h prior to actual cytokinesis. Transferred to a 14:10 h light:dark cycle for another 24 h (36 h after feeding), the same hydroids exhibited a "normal" peak MI (<(/. !5'/( ) at dawn, but /ooxanthellae from hydroids kept in constant darkness still showed a low MI. These results show that mitosis of symbiotic dinoflagellates requires three fac- tors: external food; a minimum period of time following feeding ( I 1-36 h). presumably for digestion: and a period of light following feeding, presumably to provide carbon skel- etons necessary for completing cytokinesis. Introduction Although the ecological significance of cnidarian-zoo- xanthellar symbioses has been recogni/.ed for over 70 years (Yonge and Nichols. 1931 ). efforts to understand how these associations remain together have been sporadic. Most algal sv mbioses are characteri/ed by relatively constant densities of symbionts (Reimer, 1971; Muscatine and Pool. 1979; McAuley, 1994), giving rise to theories that symbiotic algae are "regulated" by their hosts. The three proposed modes of maintaining densities of symbionts are ( 1 ) expulsion or exocytosis of extra symbionts. (2) digestion of extra sym- bionts. and (3) control of growth of endosymbionts by superimposition of an external control, or by limiting nutri- ent supply (Muscatine and Pool, 1979). The first two meth- ods assume that "extra" algae are produced by algal ihvi sion. and that the host has some mechanism of detecting and responding to increased densities of symbionts, somehow 110 GROWTH OF HOST AND ZOOXANTHELLAE 111 culling supernumerary algae down to the steady-state den- sity. The third method relies on integration of host and symbiont division cycles. Both expulsion and synchronized division are present in most cnidarian-algal symbioses that have been investigated (refs. below), but little evidence exists for digestion of extra symbiotic algae. The approach taken in this study is patterned after the productive research conducted on hydra-zoochlorellae sym- bioses. Green algae in the genus CMorella live symbioti- cally in vacuoles at the base of host digestive cells in some types of Hydra, and this symbiosis has been widely used as a model system. In Hydra, algal volume per digestive cell is positively correlated with size of the host cell, suggesting that available space is a factor in determining the symbiont population size (Douglas and Smith, 1984). Zoochlorellae divide synchronously within the host cell following host feeding (McAuley, 1982, 1985, 1986); mitosis of digestive cells and their symbiotic algae increases about 12 h after feeding (McAuley, 1982). However, the number of zoo- chlorellae within dividing host cells increases before the host cells complete cytokinesis (McAuley. 1982, 1986), suggesting that in normal culture conditions zoochlorellae often divide before the host cells do. A similar phenomenon occurs when hydra regenerate (McAuley, 1986). To grow and survive in hydra, zoochlorellae require light as well as food, and their numbers are reduced in animals kept in the dark (Pardy, 1974a. b). There are few analogous studies on control of cellular proliferation of symbiotic dinoflagellates in marine cnidar- ian host cells. However, many studies have documented the dissociation of these symbioses; these have largely focused on recent coral "bleaching events," which involve the loss of algal symbionts or their pigments (see references in Jokiel and Coles, 1990). Research in this area indicates that each partner in the symbiosis has its own physiological requirements and tolerances, and that even subtle changes in factors influencing the physiology of either partner may radically alter the steady-state of the symbiosis (i.e.. Porter et al.. 1989; Iglesias-Prietoer <;/., 1992; Gates et til., 1992; Fitt et al., 1993. 1995). Symbiotic dinoflagellates typically show peaks of divid- ing cells at dawn or at the beginning of the light period. For instance, cultured zooxanthellae maintained on a 14:10 lightidark cycle show division peaks at the beginning of the light period; the peaks are followed by the production of motile cells (Fitt and Trench, 1983). Zooxanthellae living in host gastrodermal cells (Muscastine et al.. 1998) exhibit phased division inside of the jellyfish Mastigias sp. (Wil- kerson et al.. 1983), the hydroid Myrionema amboinense (Michael and Fitt, 1984; Fitt and Cook, 1990; McAuley and Cook, 1994), the sea anemone Aiptasia pallida (Cook et al., 1988), and five species of Indo-Pacific reef corals (Smith and Hoegh-Guldberg, 1987; Hoegh-Guldberg, 1994). In contrast, asynchronous division of symbionts was reported from nine species of reef corals from Discovery Bay, Ja- maica (Wilkerson et al.. 1988). Studies investigating syn- chrony of zooxanthellar mitosis with host cell division are generally lacking; limited data from the Caribbean staghorn coral Acropora cen'icornis indicate night-time peaks in host cell division (Gladfelter. 1983). The basis for the diel division patterns seen in zooxan- thellar symbioses is not clear. It has been suggested that diel cycling of intracellular pH, driven by photosynthetic utili- zation of intracellular carbon dioxide, may be responsible by providing pulses of diffusible ammonia/ammonium (Fitt / (//.. 1995). Pulses of nitrogen have long been thought responsible for phased division of phytoplankton in nature (i.e.. Doyle and Poore, 1974), and additions of high con- centrations of dissolved nitrogen to seawater damped out the diel rhythm of zooxanthellar division in Pocillopora damicornis (Hoegh-Guldberg, 1994). That nutrients, either dissolved or from external food, are involved in the division of host cells and zooxanthellae is neither surprising nor as interesting as the temporal relationships between host feed- ing, availability of nutrients to symbionts, and mitosis of host cells and their intracellular zooxanthellae. The tropical shallow-water marine hydroid Myrionema ambionense shares several characteristics with the green hydra symbiosis, making it a good model system for cellular studies of marine dinoflagellate symbioses: it has relatively rapid growth, it can be maintained in the laboratory, and the dynamics of host and symbiont cell relationships can be analyzed with cell maceration techniques. This study relates natural diel patterns of zooxanthellar division inside of hydroid host cells to diurnal feeding of the host, host cell division, and exposure to natural light:dark cycles. Materials and Methods Collection and maintenance of animals Colonies of the hydroid Myrionema ambionense were collected from shallow-water (< 2 m) habitats adjacent to the Discovery Bay Marine Laboratory in Jamaica and used immediately in experiments. In some experiments animals were maintained in glass petri dishes in the laboratory in unfiltered seawater (SW) obtained from the laboratory sea- water system at ambient air and water temperature (26- 28C) and light (ca. 80 pE m~~s Determination of mitotic index (MI), symbiont densities, and vacuoles Unless indicated otherwise, all experiments involved macerating (David, 1973) 5-10 polyps from each of six colonies of hydroids, each polyp including < 2 mm of stolon. Zooxanthellae from 100 digestive cells were ob- served and the number of symbionts in each cell was counted. Unhealthy looking zooxanthellae in host cells were 112 \V K FITT counted from the same 100 digestive cells. The unhealthy looking dinoflagellaes were distinguished from their live counterparts by their lack of circular symmetry and rela- tively uneven and darker coloration (see Fig. 1C). The percentage of zooxanthellae dividing (milotic index Ml) was determined from microscopic counts of cells w itli di- vision furrows (doublets). The mean |vu !i\ iding zooxanthellae from each colony of hydroi w as calculated from a minimum total of 1000 /ooxanthe'lae. To determine the time of peak division ot zoo ;llae in Myrionenui amboinense. hydroids from each ' the si\ colonies were collected from the field e\ ei _4 h. macerated w ithin 15 min of collection, and the Ml determined. The volume of the host cells was determined from their depth and surface area as described in Douglas and Smith (1984). Macerated cells are flattened and not cylindrical: they are variable in length and width hut usually have a relatively straight basal edge and rounded apex. The dis- tance between the points at which the upper and lower surfaces of the cell just go out of focus was found to be 9-12 /im. as determined from the scale on the focusing knob of the microscope, so the depth of digestive cells in macerations was taken as 10 ju.m. Surface areas were deter- mined from length and width (average of widths at top and bottom of cell), measured for rectangular hydranth cells. and diameter for circular tentacle cells. Host cell division was determined by staining mitotic figures of digestive cells in macerated preparations with 4',6-diamidino-2-phenylin- dole (DAPI) (Falkowski and Owens. 1982: McAuley. 1982). Number and estimated volume of intracellular vacu- oles were also determined from the same 100 digestive cells described above. Only relatively large ( -2 pirn in diameter) vacuoles were monitored. The number and volume ol the vacuoles was compared to the volume of the host cell in one experiment to estimate the relati\e portion of the cell taken up by vacuoles. Hvdroid growth rale Hve days before the beginning of the growth experiment. olonies of Myrionema umhionenxe. with 2 to \5 polyps. were collected and placed on microscope slides in glass petri dishes. Only colonies that had attached to the slides were subsequently used in experiments. Slides with at tached hydroids were placed in SW in glass pelri dishes in the laboratory m among natural colonies of hydroids in the field. Animals maintained in the laboratory were fed nauplii of Anemia, and the .> .iter was changed daily. The polyps were counted and the Icir'tlr. <>l the stolons were measured for all experimental animals at the start ol the expci inicni and after 7 days. Factors influencing mitotic index The relationship of light and feeding to division of svm- hiotic dinoflagellates was investigated in two experiments. In the first experiment hydroids were collected from the field in the dark just after dusk or preceding dawn. Half of the hydroids collected at each time were kept in constant dark, the other half in constant light (ca. 80 /xE m 2 s ' ). The mitotic index of the zooxanthellae was determined from macerated hydroids. as described above, every 1-2 h for about 16 h after collection. In the second experiment hydroids were unfed for 3-7 davs and then fed Anemia nauplii for I h at the beginning of the dark period: they were then removed from the food source. Hall the animals were subsequently maintained in constant light, the other half in the dark. At dawn (i.e., 1 I h following feeding) the MI of the symbionts was determined. Hydroids maintained in the dark were divided into two groups again, and held another 24 h in either a 14:10 h light:dark cycle or in constant dark. The MI was determined again at the next dawn, about 36 h after feeding. Results Distribution of -ooxanthellae Zooanthellate digestive cells of Myrionenui unihioncnsc have two general shapes: columnar from the hvpostome ( = cup) region of the hydranth (also called digestive cells here. Fig. 1 ) and circular from the tentacle portion of the hvdranth (Fig. 2). The circular form is donut-shaped with a hole where the coelenteron extends up each tentacle (Fig. 2). Intermediate morphologies are found at the base of the tentacles (i.e.. Fig. 2c). Healthy looking zooxanthellae were located near the base of host hvpostome cells, adjacent to the mesoglea (Fig. IA-I3). and usually at the periphery of tentacle cells, depending on the angle of observation (Fig. 2). Recently fed hydroids maintained in the laboratory also had /ooxanthellae at the base of their digestive cells but contained many more vacuoles (Fig. IB) than seen in unfed hydroids (Fig. 1A). Unhealthy looking /ooxanthellae were usuallv located near the apical end of the cell, between the host nucleus and the coelenteron (Fig. 1C). Most of the digestive cells in the hvpostome ol the hydroid contained one. two. or three zooxanthellae: only about a quarter harbored more than three symbionts (Figs. 1. 3). Gastrodermal cells in tentacles appear to develop from digestive cells in the hvpostome that migrate from a pre- sumed central division /one. Their bases and cell volume expand as they encircle the inside of the hollow tentacle (Fig. 2C), giving rise to their characteristic circular and semicircular shapes (Fig. 2). In contrast to digestive cells in the hypostome. tentacle cells usuallv held more than 10 /ooxanthellae. with more than 75 r * of the tentacle cells containing between 10 and 40 /ooxanlhellae (Fig. 3). The GROWTH OF HOST AND ZOOXANTHELLAE 113 xanthellae may be limited by available space or physiolog- ical parameters associated with the size of the host cell. Larger host cells contained more zooxanthellae than smaller host cells, and the number of zooxanthellae residing inside of a digestive cell was directly correlated with the relative volume of the cell (Fig. 4). The average number of zoo- xanthellae in each hydroid polyp, including 1-2 mm of stalk, was 2.4 0.2 (mean SD) X 10 5 (;i = 15 different colonies). Growth Growth of a colony of Mvrioneina was surprisingly con- stant. When maintained in the laboratory for 1 week, hy- droids grew an average of 4.3 0.8 mm stolon length/d (/( = 15) for each piece of stolon. There was no correlation between growth rate and initial size of the colony (range 1-15 polyps). Both laboratory-maintained and field-moni- tored colonies roughly doubled their number of polyps over an 8-day period (Fig. 5), regardless of colony size. d Figure 1. Gastrodermal cells from the hypostome of the hydroid of Mvritinema ambioinense. (a) Columnar digestive cell from hydroids col- lected in the afternoon containing four healthy looking zooxanthellae (/. ) and few vacuoles (v). (b) Columnar digestive cell from hydroids collected at night containing two zooxanthellae and numerous vacuoles. (c) Healthy looking /ooxanthellae at basal end of this digestive cell contrast with tour degenerate or unhealthy looking zooxanthellae (d) at the distal end of the digestive cell, (d) Row of columnar digestive cells from partially macerated polyp showing zooxanthellae at the base, away from the phagocytic distal end of their host cell. Arrows show centrally located host nucleus with prominent nucleolus. Bars = 10 /nm. highest number of symbionts in a single digestive cell was 69, observed in a tentacle cell. Estimates of the relative volume of digestive cells in relation to their population of symbionts suggest that zoo- Diel patterns of host and symbiont division Zooxanthellate division (MI = 16.6 1.5%, n = 6) from field-collected hydroids peaked at dawn, declining to near zero in the early afternoon and evening (Fig. 6C). Thirty- four percent of the host cells containing zooxanthellae held at least one dividing algal cell over the diel period. The mean density of zooxanthellae per hydranth cell was 2.67. The density of zooxanthellae per hypostome digestive cell was highest in the mid-afternoon following the division peak (ca. 2.9 0.1 zooxanthellae per host cell, n = 6), and lowest about 3 h after sunset (ca. 2.4 0.4 zooxanthellae per host cell, n = 6) (Fig. 6C). This implies an average growth rate of about 0.5 zooxanthellae per digestive cell per day. The same value is obtained by multiplying the maxi- mum number of zooxanthellae per host cell (2.9) by the Ml (16.6%), again resulting in an increase of about 0.5 zoo- xanthellae per host cell per day. In other words, both meth- ods show that the number of zooxanthellae in each hydranth digestive cell doubles about every 6 days. The number of unhealthy looking zooxanthellae also exhibited a diel cycle (Fig. 6B). The number of unhealthy looking zooxanthellae per hydranth digestive cell showed a broad peak at about midnight (four highest data points = 1.04 0.14 unhealthy looking zooxanthellae/cell). and reached a broad low during the day (four lowest data points = 0.72 0.05 dead zooxanthellae/cell). To balance the average growth rate of colonies (doubling time = 8 days) with the average growth rate in number of zooxan- thellae (doubling time = 6 days), an average digestive cell would have to lose about 1 zooxanthella every 8 days (0.14 zooxanthellae per day). This rate of loss of zooxanthellae is about half that calculated from the difference in average 114 \v. K. HTT * ' 'd Figure 2. Ciastrodermal cells from the tentacles of ihe hvdroid Myriimema ambioinense. (a. h) Circular disc-shaped digestiu- cells showing /oovmihcllae I/) in Ihe basal portion ot the cell which would normally lie adjacent to the mcsoglea in ihe tentacle, host nucleus laiiowsi near the center, and lumen leucnsion ol coelenteron) of hollow tentacle immediate!;, below allow u i 1 1. ill -moon-shaped digestive cell, wrappmi.' .IT on in I Ihe base of a tentacle idi Dist vh.iped dii:esii\e cell showing degenerate or unhealthy looking /.ooxanlhellae (d) near the center and at one side where Ihe two ends of the base of the cell presumably have met. m = looxanthella with division furrow indicalnir i \iokmesis at the end of mitosis; v = \.auoles Arrows show centrally located host nucleus with prominent nucleolus. Bars = II) /jm density ot .k\nl or monhmul Inokiiiij /(inxjiiillK-lhn. 1 (0..^2 /ooxanlhellae p daj i uhscrvcd over a 24-h PLTJOI!. Though n ! fficult to see final cytokinesis (telophase) of host digestive celN. nuclear staining with DAI'I slmued that mitosis in these trlls (pro-, mcta-. and anaphasei oe curred predcimm.intK in the evening (peak Ml = 2.4 1.3%, n = 6 colonies i. ei'-np.iied ID other limes tltiring the day (MI < 1%) (Fig. 6Ai. These d.ii.i suggest that cytoki- nesis (cell separation) occurs lirsi in the early evening, corresponding v\ith ihe ohseixed decrease in /ooxanlhellar densii\ per Imsi cell (see above). Evidence fur diel host feeding Hydroids collected at night and macerated immediately alter collection inevitably contained ingested prey items. GROWTH OF HOST AND ZOOXANTHELLAE 115 J3 - | % ol hypostome cells ^^ D % ol tentacle cells jfl 30 - ~Q> O ~ 25 - . I )i ,iiihiilion ..i symbiotic dmollagcllali-s l/oo\antlicll.iel IM|CS within individual host cells ol MMII'IU-IIIH iiiiil'inm n\i- >n I" lime til da> ciilkxlol Inun Ilir Ik-lil. I A I IVitt-nl ol hyposlniiu- digestive 1 n-IU contiiininy vannik's .uul iiiiinhei nl \aiunk-s |ici thj;evli\e cell and hiisi itolil index. IB) Percent nl li\|'cislnriK- diyesine cells comaim 01 unhealihs looking zooxanthellae and numbci pei digestive eel! ni nl /ooxanlhellae with division furrows imiimu index I and nm-J . looking zooxanthellae pei hvilianih digestive cell. Data presenh SI) n <> D.iik li.n on i .IMS driu 8 li .ilu-i Environmental factors ami relationship to die! pattern* The inlluence of photoperiod and feeding on /ooxanthel- late division was determined in two experiments. In the first experiment (Fig. 8 A), hydroids collected at dusk and main- tained without food in either constant light or dark showed the same level and timing of MI as seen in hydroids in the field (Fig. 6). Thus it appears that altering food and the light and dark periods less than 12 h before the expected peak in MI does not influence the division patterns of the /ooxan- tltdlae. Hydroids that were collected from the field at the fiul of the dark period (when MI of symbionts was highest) and maintained in constant Hunt or in constant dark showed o LL o 4O 3O O 6? LU U to LLJ _l O 3 O < LU O U > X K $ CO LU O 10 o 4 3 1 1 O too 80 60 4O 2O B 4 6 8 IO HOURS AFTER FEEDING 12 14 7. Relationship between xacuolc appearance in digestive cells of Myrioiii-imi amhiiau'iise following a l-h pulse-feeding with nauplii ol brine shrimp (time 0). lAi Volume (area) of host cells occupied by tacuolcs. iHi Au-i.iiH' numbei ol \acnoles per cell. (C) Percentage lit digestive cells containing vacuolcs. Bach data point is the mean ol MM) digestive cells from the hypostome. GROWTH OF HOST AND ZOOXANTHELLAE 117 1800 2200 0200 0600 1000 1400 1800 2200 TIME OF DAY Figure 8. Relationship between mitotic index of symbiotic dinorlagel- lates residing in digestive cells of Mvrionema ambionense when hydroids were either (A) collected at dusk ( 1900) and maintained in either constant light (open symbols) or constant dark (closed symbols) or (B) collected at the end of the night and maintained similarly in either constant light or dark. slightly different results (Fig. 8B). Zooxanthellae from hy- droids maintained in the light showed no changes in the division patterns seen in the field. In contrast, the hydroids maintained in constant darkness showed a slower decline in zooxanthellar cytokinesis during the afternoon period (Fig. 8B), suggesting that light was critical for the final stages of symbiont cytokinesis in this experiment. In the second experiment, starved hydroids (3-7 days) were pulse-fed for an hour at the beginning of the dark period in the laboratory, then either maintained in constant light or left in the dark (Fig. 9). About 12 h after the pulse-feeding, those hydroids kept under constant light dur- ing the night had a higher MI (4.2% 1.4% SD, n = 6) than those hydroids kept in the dark (1.6% 0.9% SD, n = 6). The hydroids maintained in the dark for 12 h were then divided into two groups. The group that remained in the dark for an additional 24 h continued to show low algal MI (2.1%), while those moved into a 14:10 h lightidark cycle for 24 h showed a normal level of dividing cells (ca. 14.2%) 36 h after feeding. These results show that normal mitosis of symbiotic dinoflagellates in this hydroid requires three fac- tors: external food: a minimum period of time following feeding ( 1 1-36 h), presumably for digestion to occur; and a period of light following feeding, presumably to provide carbon skeletons necessary for completing cytokinesis. Discussion This study shows that division of zooxanthellae and host cell is synchronized in the marine hydroid Myrionema am- bionense, and that host feeding and light play an important role in the phased division patterns observed. Most, if not all, zooxanthellae symbioses appear to exhibit phased sym- 16 14 - 12 - 1 8 - o o Time after feeding (h) Figure 9. Relationship between feeding, light, and darkness on divi- sion (mitotic index) of symbiotic dinoflagellates residing in the hydroid Myrionema ambionense. Unfed (3-7 days) hydroids were fed for 1 h at time (1900) and maintained for the next 1 1 h (12 h after feeding) in either constant dark (filled bars) or light (open bars). Animals kept in the light were then discarded, and those from the dark were maintained an additional 24 h (36 h after feeding) in either constant dark or on a normal 14:10 h lighcdark cycle. 118 \V. K FITT biont division in siru. with peaks in doublet cells seen around da\vn (Michael and Fitt. 1984: Smith and H Guldberg. 1987; Fitt and Cook. 1990: Cook and Fitt. 1990: Hoegh-Guldberg and Smith. 1989: McAulev and Cook, 1994: Hoegh-Guldberg. 1994). Symbiotic dm>!!agellates in hosts removed from the reef or placed in seawater contain- ing high concentrations of dissolve! ,,iorganic nitrogen (DIN) lose their division synchronv i-or instance, additions of DIN damped out natural diel patterns of zooxanthellar division in the coral Pocill<>"< ./ 86i. Data from M\rit>nenni amhioncnse suggest that dividing host cells completed cytokinesis before their symbiotic dinofiagellates finished dividing, such that decreases in svmbiont densities were observed in the late afternoon and evening. Early morning cytokinesis of zooxanthellae re- turned densities of symbionts to their afternoon peak levels. Phased division of cndothelial and calicoblastic epithelial coral cells in Acro/ioru < -en n nnii\ also occurred during the middle of the night, and MI was less than 2 r /f (Gladfelter. 1983). The peak percentage of cells containing mitotic figures in M\riui iic>i\<- was also low compared in the proportion of zooxanthellae dividing daily, due to the likelihood that during the cell cycle mitosis lasts much less than the 3-h sample interval, suggesting that mam such events were missed between sample times. I -or instance, mitosis in well-fed H\dni lasted only about 1.5 h (David and Campbell. 1972). heeding and light are two factors linked to phased mitosis (it /iiMx.inihellae and host digestive cells in Myrioiifitiii (ini/'inneiiM'. Digestive cells in fhdni also exhibit a diel periodicity in the mitotic index, with a midnight peak fol- lowing a dailv feeding regime at 1000 h each morning (see references in McAuley, 1994). Similarly, normal leedmg of tropical marine cnidarians in n.iiuie appears to occur on a diel cycle set bv the availability of demersal plankton food (Johannes and Tepley. l ( '74; Porter. 1974; Alldredge and King, 1977) or alternate sources of nitrogen (see I Hi 12 to 24 h (McAuley. 1985). The results lor both svmhioses suggest that although host cell mitosis follows host feeding h\ 12 to 24 h. the algal symbiont cell cvcle is set 12 to 24 h later (24-36 h follow ing feeding, depending on the synchronization of light and feeding cy- cles). These results also suggest that the close regulatory relationship between host and symbiont division may be easily disrupted. For instance, the Chlorellti symbionts in green hydra will overgrow and actually burst their host digestive cell when unfed hydra are placed in a complex mixture of inorganic nutrients in the light (Muscatine and Necklemann. 1981). Virtual!) all /ooxanthellate cnidarians appear to expel some of their symbiotic dinoflagellates from their coe- lenteron on a dailv basis (c.i;.. Reimer. 1971; Hoegh-Guld- berg c/ ul., 1987), whereas mass expulsion of /ooxanthellae has been alliihuied lo seveie osmotic or thermal stress GROWTH OF HOST AND ZOOXANTHELLAE 119 (Goreau. 1964; Jokiel and Coles, 1977, 1990; Jaap, 1979; Glynn, 1983). The released zooxanthellae may be of normal morphology, with dividing and motile cells, or they may appear unhealthy, as described here. Most researchers feel that expulsion from the host is a way to partially balance increases of symbionts from algal division (Reimer, 1971; Hoegh-Guldberg et a I.. 1987). and in cases where live zooxanthellae are released, a dispersal mechanism for in- fecting new hosts (Trench. 1979; Muller-Parker. 1984). Many sea anemones (Steele, 1977) and all of the giant clams (Trench et ai, 1981 ) release motile forms of zooxanthellae daily. Hoegh-Guldberg et icornis. Biol. Bull. 165: 811-815. Glynn, P. W. 1973. Ecology of a Caribbean coral reef. The Porites reef 120 \V K. FITT flat biotype Pi. II. Plankton community with evidence for depletion. Mar Biol. 22: 1 :: Glynn. P. \V. 1983. Extensive "bleaching" and death of reel corals on the Pacific coast of Panama. Environ. Consen: 10: 149-154 Goreau, T. F. 1964. Mass expulsion of /ooxanthellae tiom Jamaican reef communities after Hurricane Flora. Science 145 Hx.'i>h-Guldberg. O. 1994. Population dynamics ot symbiotic /ooxan- thellae in the coral Pocillopora damicormi exposed to elevated am- monium concentrations. Pac. Sci. 48: > Hoegh-Guldberg, O., and G. J. Smith. 1989. Influence of the popula- tion density of zooxanthellae and supply ammonium on the biomass and metabolic characteristics of the reel .S,i 29:414 4:: Johannes. R. K., and L. Tepley. 1974. I \aiiunation of feeding of ihe reef coral Ponies lobala in situ using time-lapse photography. Proc. 2nd Inl. Coral Recj S\mp. 2: 369-374. Jokiel. P. I... and S. L. Coles. 1977. F.ltects of temperature on the mortality and growth of Hawaiian reef corals. A/,// Hint. 43: 201-208. Jokiel. P. I... and S. L. Coles. 1990. Response oi Hawaiian and other Indo-Pacific reef corals to elevated temperature. Coral Reefs 8: 155- 162. Mi \uli-v. P. J. 1982. Temporal relationships ot host cell and algal mitosis in the green hydra symbiosis. ./. Cell Sci. 58: 423-431. Me \uley, P. J. 1985. The cell cycle of symbiotic Clilorella. 1 The relationship between host reeding and algal cell growth and division. J. Cell Sci. 77: 225-239. McAuley, P. J. 1986. The cell cycle ot symbiotic Chlorcllu. 111. Num- bers of algae in green hydra digestive cells are regulated at digestive cell division. J. Cell Sci. 85: 623-671. Mi \uley. P. J. 1994. Interactions between hosts and symhioms in algal invertebrate intracellular symbioses. Hot. J. Scoll. 47: 97-1 12. McAuley. P. J., and C. B. Cook. 1994. Httects of host feeding and dissolved ammonium on cell division and nitrogen status of /ooxan- thellae in the hydroid M:n^ii,;na ambionense. Mar. Biol. 121: 343- 348. Michael. \\. ('.. and \\. K. I ill. 19S4. l-llects of a water-soluble fraction of a crude oil on the coral reel hydroid M\-rionema hargitti: feeding, growth and algal symhiontx \l,n Hml 84: 143-154. Mailer-Parker, G. 1984. Dispersal of /ooxanthellae on coral reels In predators on cnidarians. Biol. Bull. 167: 159 167 Muscatine, I... and N. Neckelmaim. 1981. Regulation ol nunibcis ol algae in the Hydra-Chlorclla symbiosis. Hn: Dr\,li HIT (ic\ 94: 571-582. Muscatine, I... and R. R. Pool. 1979. Regulation of numbeis ot mli.i cellular algae Prot K S. Land 204: Ml 139 Musi-aline. I... K. P. \\ilkerson. and I,. R. McCloskey. 1986. Regula- tion of population density of symbiotic algae in a tropical marine jellyfish (Masiigias sp.). Mar. Ecol. 32: 279-290. Muscatine. I... C. Ferrier-Pages. A. Blackburn. R. I). Gates, G. Bagh- .1.1-. ii i. in and I). Allemand. 1998. Cell-specific density of symbiotic dinoflagellates in tropical antho/oans. Coral Reefs 17: 329-337. Pardy, R. I.. 1974a. Some factors affecting the growth and distribution of the algal endosymbionts of Hydra I'iridis. Biol. Bull. 147: 105-1 18. Pardy. R. I.. 1974h. Regulation of the eodosymbiotic algae in Hydra by digestive cells and tissue growth. Am. Zooi 14: 583-588. Porter. J. \\ . 1974. /.ooplankton feeding by the Caribbean reef-building coral Mantastrea annularis. Proc. 2nd Int. Coral Reef Symp. 1: III- 125. Porter. J. \\ .. \\ . K. Kill. H. J. Spero. C. S. Rugers. and M. \\ . \\ hite. 1989. Bleaching in reef corals: physiological and stable isotopic responses. Proc. Nail. A-