•
.
,
•
s
.
:
•
1
.
.
-
. '
.
.
, -
I
,
. .
I
.
i
'
I
•
• -
3 ,
. .
! • j • i • • ' . • • ' s . • . • • : ^ •• .
. i : • . : tl K : - '
THE
BIOLOGICAL BULLETIN
AUGUST 1999
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 INDUE, Imaging and Microscopy
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
The Whitney Laboratory, University of Florida
Grice Marine Biological 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, LIniv. 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 Assistant
Subscription & Advertising Secretary
Published by
MARINE BIOLOGICAL LABORATORY WOODS HOLE, MASSACHUSETTS
Cover
The ascoglossan sea slug, Elysia chlorotica (Gould), shown on the cover (photograph by S. K. Pierce), seeks out and specifically eats a chromo- phytic alga, Vancheria litorea. Like certain other species of sea slugs, E. chlorotica has developed the ability to acquire the chloroplasts from its algal foodstuff and to utilize them for nutrition. The plastids are usually engulfed by particular epithelial cells in the digestive gland where they photosyn- thesize and, in some species, provide sufficient nu- trients to sustain life and reproduction — even when no other food is available.
In E. chlorotica, the function of the captured chlo- roplasts is maintained for up to 8 months — a surpris- ingly long period, surpassing similar chloroplast sym- bioses by many months. Throughout this period, furthermore, plastid proteins are continuously synthe- sized, and some of the proteins appear to be encoded by the slug genome. Another remarkable feature of these slug populations is the abrupt end of the annual life cycle — all of the animals dying synchronously, whether in the laboratory or in the field.
In this issue. Skip Pierce and his colleagues (p. 1 ) report a widespread viral infection of the slug pop- ulation; this phenomenon also occurs annually and is coincident with the mass mortality. The viruses (see inset) seem to be endogenous and have many characteristics in common with retroviruses. The report suggests that the viruses may not only be involved in the regulation of the slug's life cycle, but may be the means by which algal genes are transferred to the slug genome.
CONTENTS
VOLUME 197, No I: AUGUST 1999
RESEARCH NOTES
Pierce, Sidney K., Timothy K. Maugel, Mary E. Rumpho, Jeffrey J. Hanteii, and William L. Mondy
Annual viral expression in a sea slug population: life cycle control and symbiotic chloroplast maintenance . 1
Thomas, Florence I.M., Kristen A. Edwards, Toby F.
Bolton, Mary A. Sewell, and Jill M. Zande
Mechanical resistance to shear stress: the role of echinoderm egg extracellular layers 7
Rinkevich, B., S. Ben-Yakir, and R. Ben-Yakir
Regeneration of amputated avian bone by a coral skeletal implant 11
ECOLOGY AND EVOLUTION
Miner, Benjamin G., Eric Sanford, Richard R. Strath- mann, Bruno Fernet, and Richard B. Emlet
Functional and evolutionary implications of opposed bands, big mouths, and extensive oral ciliation in larval opheliids and echiurids (Annelida) 14
Johnsen, Sonke, Elizabeth J. Balser, Erin C. Fisher, and
Edith A. Widder
Bioluminescence in the deep-sea cirrate octopod Staurotnithis syitensis Verrill (Mollusca: Cephalopoda) . 26
NEUROBIOLOGY AND BEHAVIOR
Ganter, Geoffrey K., Ralf Heinrich, Richard P. Bunge, and Edward A. Kravitz
Long-term culture of lobster central ganglia; expres- sion of foreign genes in identified neurons 40
Hanlon, Roger T., Michael R. Maxwell, Nadav Shashar, Ellis R. Loew, and Kim-Laura Boyle
An ethogram of body patterning behavior in the biomedicallv and commercially valuable squid Loligo
/mild off Cape Cod, Massachusetts 49
Bushmann, Paul J.
Concurrent signals and behavioral plasticity in blue crab (Callinectr* uipidiu Rathbun) courtship 63
PHYSIOLOGY
Engebretson, Hilary P., and Gisele Muller-Parker
Translocation of photosynthetic carbon from two algal symbionts to the sea anemone Anthopleura elegantissima 72
DEVELOPMENT AND REPRODUCTION
Grabowski, Gregory M., John G. Blackburn, and Eric R. Lacy
Morphology and epithelial ion transport of the alka- line gland in the Atlantic stingray (Dasyatis sabina) ... 82
Krug, Patrick J., and Adriana E. Manzi
Waterborne and surface-associated carbohydrates as settlement cues for larvae of the specialist marine herbivore Ahitri/i moritsta 94
Chaparro, O.R., R.J. Thompson, and C.J. Emerson The velar ciliature in the brooded larva of the Chil- ean oyster Ostrea cltiltnsis (Philippi, 1845) 104
Annual Report of the Marine Biological Laboratory .... R 1
THE BIOLOGICAL BULLETIN
THE BIOLOGICAL BULLETIN is published six limes a year by the Marine Biological Laboratory, 7 MBL Street, Woods Hole, Massachusetts 02543.
Subscriptions and similar matter should be addressed to Subscription Manager, THE BIOLOGICAL BULLETIN, Marine Biological Laboratory, 7 MBL Street, Woods Hole. Massachusetts 02543. Subscription per year (six issues, two volumes): $205 tor libraries; $95 for individuals. Subscription per volume (three issues): $102.50 for libraries; $47.50 for individuals. Back and single issues (subject to availability): $40 for libranes; $20 for individuals.
Communications relative to manuscripts should be sent to Michael J. Greenberg, Editor-in-Chief, or Pamela Clapp Hinkle. Managing Editor, at the Marine Biological Laboratory, 7 MBL Street. Woods Hole. Massachusetts 02543. Telephone: (508) 289-7428. FAX: 508-457-1924. E-mail: pclapp@mbl.edu.
http://www.mbl.edu/BiologicalBulletin/
The home page for the electronic companion to THE BIOLOGICAL BULLETIN — the Marine Models Electronic Record — and other BIOLOGICAL BULLETIN publications is available on the World Wide Web at the address shown above.
THE BIOLOGICAL BULLETIN is indexed in bibliographic services including Index Medicus and MEDL1NE, Chemical Abstracts, Current Contemn, Elsevier BIOBASE/Current Awareness in Biological Sciences, and Geo Abstracts.
Printed on acid free paper, effective with Volume ISO, Issue 1, 1991.
POSTMASTER: Send address changes to THE BIOLOGICAL BULLETIN, Marine Biological Laboratory,
7 MBL Street, Woods Hole. MA 02543.
Copyright © 1999. 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 be 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 be 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 be referred to in the text by number, and listed in the Literature Cited section in the order that they appear in the text. Unlike references in Nature, references in the Research Notes section should conform in punctuation and arrangement to the style of recent issues of The Biological Bulletin. Materials and Methods should be incorpo- rated into appropriate figure legends. See the article by Loh- iTumn et al. (October 1990. Vol. 179: 2I4-2IX) 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 for 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.) on one side of 16- or 20-lb. bond paper, 8 by 1 1 inches. Please, no right justification. Manuscripts should be proofread carefully 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 Biology Editors Style Manual. 5th Edition (Council of Biology Editors, 1983) and to American spelling. Unusual abbreviations should be 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 be divided into the following components: Title page. Abstract (of no more than 200 words), Introduction, Materials and Methods, Results, Discussion, Acknowledgments, Literature Cited, Tables, and Figure Legends. In addition, authors should supply a list of words and phrases under which the article should be 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- cation 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 al.. 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 (USASII. 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. All 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 Rit Visindafjelags Islendinga without abbreviation.
F. All single word journal titles in full (e.g., Veliger, Ecol- ogy. 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. Land., 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. 197: 1-6. (August 1999)
Annual Viral Expression in a Sea Slug Population: Life Cycle Control and Symbiotic Chloroplast
Maintenance
SIDNEY K. PIERCE1 *, TIMOTHY K. MAUGEL1. MARY E. RUMPHO2. JEFFREY J. HANTEN1, AND WILLIAM L. MONDY1
Department of Biologv. University of Man-land, College Park, Maryland 20742: and 2 Department of Horticultural Sciences, Texas A & M University, College Station, Texas 77843
In a few well-known cases, animal population dynamics are regulated by cyclical infections of protists, bacteria, or viruses. In most of these cases, the pathogen persists in the environment, where it continues to infect some percentage of successive generations of the host organism. This persis- tent re-infection causes a long-lived decline, in either pop- ulation size or cycle, to a level that depends upon pathogen density and infection level (1-4). We have discovered, on the basis of 9 years of observation, an annual viral expres- sion in Elysia chlorotica, an ascoglossan sea slug, that coincides with the yearly, synchronized death of all the adults in the population. This coincidence of viral expres- sion and mass death is ubiquitous, and it occurs in the laboratory as well as in the field. Our evidence also sug- gests that the viruses do not re-infect subsequent genera- tions from an external pathogen pool, but are endogenous to the slug. We are led, finally, to the hypothesis that the viruses may be involved in the maintenance of symbiotic chloroplasts within the molluscan cells.
Populations of the ascoglossan sea slug Elysia chlorotica occur in salt marshes from the Chesapeake Bay to Nova Scotia. The life cycle of the slug lasts about 10 months. The hermaphroditic adults lay egg masses in the spring of each year, and all of the adults die shortly afterward (5, 6). A week or so after the egg deposition, veliger larvae hatch. These larvae spend a few weeks in the plankton and, it filaments of the chromophytic alga Vaucheria litorea are present, each veliger homes in on one of them and attaches to it. During the next 24 h, the larva metamorphoses into a
Received 1 June 1999; accepted 17 June 1999.
* To whom correspondence should be addressed: E-mail: sp30@ umail.umd.edu
juvenile slug while still attached to the algal filament. If the algal filaments are not present, metamorphosis rarely oc- curs, at least not in laboratory cultures (5. 6). Vaucheria is the only alga that E. chlorotica eats, and it is the source of the symbiotic chloroplasts that are acquired during feeding. The juvenile slug immediately begins eating the algal fila- ments and taking on its first load of chloroplasts, which are sequestered by specialized cells in the epithelium lining the digestive diverticula (5, 6). During the next several months. the slugs continue to eat Vaucheria and grow, until winter temperatures cause them to become inactive. As the salt marshes warm in the spring, the slugs become active again, begin laying egg masses, and then die. By the time the egg masses have hatched in May. all the adults are gone. This mass mortality occurs synchronously in the laboratory as well as in the field and regardless of the time of year that the slugs were collected.
Symbioses in which chloroplasts — usually from a partic- ular species of alga — are taken up and retained within the cytoplasm of an animal cell occur in several phyla, but they are most commonly encountered in molluscan sea slugs, particularly in the order Ascoglossa ( = Sacoglossa) (Opis- thobranchia). Certain of the molluscan cells can capture chloroplasts from algal food (usually from a specific species of either Rhodophyceae or Chlorophyceae), and these or- ganelles retain some degree of photosynthetic function for a time (e.g.. 7, 8, 9). Whether this intracellular association is a symbiosis in the strict sense is debatable; some authors prefer terms like chloroplast symbiosis, chloroplast reten- tion, or kleptoplasty (7. 9, 11, 12, 13), which indicate that the benefit of the association is entirely to the animal.
The duration of the association between the molluscan cell and the algal plastid varies from species to species.
S. K. PIERCE ET AL
Some associations last less than a week [e.g., Hermaea hifida and Elysia hedgpethi (see 14)]; and although the plastids are initially functional, photosynthesis stops or is greatly reduced after a week of starvation. In contrast, plastid function continues for more than a week in starved specimens of several species, most belonging to the asco- glossan family Elysiidae (e.g., 7, 10, 15, 16). In the species with the longer duration symbioses, the algal chloroplasts within the molluscan cells fix I4CO:, and the I4C appears in a variety of compounds in the animal tissues ( 15, 17. 18, 19, 20, 21). Thus, there is no doubt that the captured chloro- plasts are photosynthetically active within the molluscan cell cytoplasm, and that the products of the synthesis are utilized by the host animal. Indeed, once the symbiosis is established, the species with the longer-lived chloroplast associations can be starved and, as long as light is provided, will maintain or actually gain body weight until the chlo- roplast function finally fails (7).
In plant cells, continued photosynthetic activity requires continuous synthesis of chloroplast proteins because several proteins, including those used in light harvesting, are rap- idly turned-over during the process and must be replaced (22, 23, 24). But, several of these photosynthetic proteins (or their subunits) are encoded in the nucleus, so plastid protein synthesis requires the integration of two distinct genomes, that of the chloroplast and that of the plant cell nucleus (reviewed in 25). Because normal plastid photosyn- thetic function is dependent upon major nuclear and cyto- plasmic support, the ultimate failure of photosynthesis by the symbiotic chloroplasts within the molluscan cells is not surprising. But the symbiotic plastids of E. chlorotica stay photosynthetically functional for 8-9 months (5, 26) — many months longer than those of any other species yet described. This remarkable persistence of chloroplast func- tion during starvation indicates that replacement of at least the essential photosynthetic proteins must be occurring within the plastid while it is housed in the molluscan cyto- plasm; and indeed, synthesis of plastid proteins, primarily those associated with photosynthesis, occurs in the E. chlo- rotica plastids (26. 27. 28).
The plastid proteins synthesized within the cells of E. chlorotica are of two pharmacologically distinguishable sorts: those whose synthesis is blocked by chloramphenicol, and those blocked by cycloheximide (26). Protein synthesis on pltistul ribosomes is blocked by chloramphenicol, whereas synthesis on cytoplasmic ribosomes, usually di- rected by nuclear genes, is blocked by cycloheximide. These results suggest strongly that some plastid proteins are synthesized upon slug cell ribosomes (26). If this is the case, the genetic information coding for these proteins must somehow be present in the molluscan DNA or must be acquired by all the slugs from the alga in every generation. Our findings presented below suggest a vehicle for this transfer of genetic information from the alga to the slug.
In 1990, as part of an ongoing ultrastructural study of morphological changes in the fine structure of the dying slugs from a population on Martha's Vineyard, Massachu- setts, we discovered viruses in digestive cells and hemo- cytes (29; also Fig. 1). Every year since then, every animal examined near the end of the life cycle has had viruses present, whether it was fixed within hours of collection from the field or maintained for months in our aquarium. No evidence of viruses has been found in any slug earlier in the year, except occasionally within the confines of the plastid (Fig. IE, see below).
Viruses in the spring slugs are present in the nucleus and cytoplasm of several cell types. They appear to be assem- bled in the nucleus, then move into the cytoplasm, and finally bud out either into the extracellular space or into a vacuole (Fig. 1A). The diameters of the icosahedral viral capsids in the nuclei average 89 nm (±1.0); but in the cytoplasm, after they have picked up an envelope, they are 109 nm (±1.6) (Fig. IB). The shapes and sizes of the capsids and envelopes are very similar to those in the Retroviridae, but other viral types have similar dimensions, and known retroviruses are not assembled in the nucleus (30, 31). In addition to these larger viruses, some chloro- plasts within the cells of spring slugs contain structures (diameter = 20 nm ± 0.4) that could either be smaller viruses or viral cores. These particles occur loosely col- lected in areas between the plastid membranes (Fig. 1C). They also occur as particle clumps in the cytoplasm (Fig. ID), sometimes near material that could be the remnants of chloroplasts. In some instances, in a few fall animals, these particles are present in crystalline arrays (Fig. IE). Ribu- lose-l,5-bisphosphate-carboxylase oxygenase (RuBisCO) occurs in crystalline form in some plant plastids fixed under hyperosmotic conditions (32, 33). However, the particles in such RuBisCO crystals are much smaller and usually lack a visual substructure (32, 33), whereas the array particles in the symbiotic plastids in Elysia clearly have a substructure (Fig. IE). The particle arrays in the E. chlorotica plastids are more reminiscent of mosaic viruses in plants (34), although almost nothing is known about such crystals in the viruses and plastids of algae. In summary, the morphology suggests that either more than one viral type is present in the slug cell and captured plastids, or we have found several stages of a single viral type.
In addition to the morphology, we have some biochemi- cal information about the identity of the viruses. Using differential and sucrose-gradient centrifugation, we have isolated a fraction from slug homogenates at a density of 1.18 g/ml that has reverse transcriptase (RT) activity. This activity, which is considered diagnostic for retroviruses, is two orders of magnitude higher in spring slugs than in slugs tested in the fall (Fig. 2). In addition, the RT activity of fall animals, but not spring animals, is inhibited by rifampicin (Fig. 3). This inhibition indicates that the activity measured
SEA SLUGS, PLASTIDS. AND VIRUSES
Figure 1. Electron micrographs of viruses in Elysiu chlorulica. (A) Viral particles in various stages of maturation are present in the nucleus (n) and cytoplasm of a hemocyte from a dying slug (magnification = 33,750 X, scale bar = 1 /urn). (B) Higher magnification ( 85.000 x, scale bar = 0.? /j.m) of viruses budding into cytoplasmic vacuoles. Icosahedral shape and double envelopes of the mature virus are apparent. (C) Viral aggregates (arrows) within the symbiotic chloroplast of a spring Elysiu (magnification = 16,880, scale bar = 1.0 fim). (D) Viral particles very similar in appearance to those in (C) located in the cytoplasm of a chloroplast (cp)-containing digestive cell of a spring slug. The diffuse, gray areas in the cytoplasm are lipid produced by the plastid (magnification = 27,000, scale bar = 1 .0 /j.m). (E) Viral crystal contained in a symbiotic chloroplast from a fall slug (magnification = 54,000, scale bar = 0.5 jim).
3.
E
15(1(1(111-1
100000-
50000-
A-fall animals
S. K. PIERCE ET AL. 150000-
100000- 50000-
B-spring animals
i i i I i i i i i i i i i i
1.12 1 12 I 14 118 I 18 1 18 I 19 1 19 1.19 I 20 1.20 1.21 121 1 22
~1 I I I I I 1 I I I I I 1 I I
106 1.10 I 12 I 14 I 15 1 17 1 17 1 18 I 18 1 IS I 19 1.19 1.19 120 1.20
Gradient fractions (gm/ml)
Figure 2. Comparison of reverse transcriptase activity in sucrose-gradient fractions from a typical extract of fall slugs (A) and spring slugs (B). The results are essentially the same whether the animals were freshly collected in the spring or collected in the fall and overwintered in aquaria.
in the fall animals is due to DNA dependent-RNA polymer- ase (which utilizes the same substrates in the RT assay) rather than RT, and confirms the absence of viruses in the fall animals. Taken together, the morphology, the buoyancy, and the presence of RT all suggest that a retro-like virus is present in the cells of the dying slugs.
150-1
100-
3
50-
O-1
Fall
Spring
Figure 3. Rilampicin inhibits reverse transcriptase activity. Enzyme activity was assayed in pooled gradient fractions with densities from 1.16 to 1.18 g/ml prepared from fall and spring slugs. The effect of the inhibitor is expressed as a percentage of control values.
The relationships between the nuclear, cytoplasmic, and plastid viruses are not known at present. However, for the last 9 years, the viruses have been found in every dying slug examined. Since some of the slugs had been maintained in the laboratory, in aquaria containing artificial seawater and with no access to Vauclierui for 8 months before the viruses appeared, the infection is unlikely to have been opportunis- tic. Instead, the results suggest either that the demise of the entire population is caused by an endogenous virus or that the virus can be expressed only as the defense systems of the aging animals begin to fail. Furthermore, if the effect on the life cycle is due to a retrovirus, as our data suggest, then the viral genome is probably transmitted to the next gener- ation of slugs in the molluscan DNA. Infection of germ cells by retroviruses produces endogenous proviruses that are inherited as Mendelian genes (33). Alternatively, the viruses might enter all of the slugs via the sequestered chloroplasts. either as part of the plastid genome or as constituted viral particles. In either case, viral expression is coincident with increases in environmental temperature, at least in the field slugs. Both the onset of egg laying and the death of the population are associated with the rise of water tempera- tures in the spring. We can delay the demise of the slugs by maintaining them at very cold temperatures (2°-5°C), but eventually — 6-8 weeks after the warmer-maintained slugs have died — the cooled slugs also die with viruses in their cells, indicating that temperature is not the only expression stimulus.
It will take some time to sort out both the types of viruses involved here and the molecular relationships between the
SEA SLUGS. PLASTIDS. AND VIRUSES
slugs, the algae, the plastids. and the viruses. Nevertheless, the nine-year, annual occurrence of the association between population demise and viral expression at the end of the E. chlorotica life cycle strongly suggests that the virus has a role in regulating this coincidence. Indeed, we speculate that the viral infection may have caused the transfer, from the alga to the slug, of algal genes that allow the molluscan cells to assist in plastid maintenance. Although the transfer, in- tegration, and expression of such a group of genes between species should succeed only rarely, those successes that did occur should have profound, immediate, heritable effects on the phenotype of the infected species. Such heritable effects must certainly be associated with the mechanism of the widely accepted endosymbiotic origin of intracellular or- ganelles such as mitochondria and chloroplasts; a variety of genes must have been transferred from the symbiont into the host cell nucleus to consummate such a relationship. In addition, a retrovirus as a gene transmission vehicle might have merit as a hypothesis to explain genetic similarities between distantly related or unrelated species (e.g., 35) and is the basis of some genetic therapies (36). If a successful interspecies gene transfer between an alga and a slug me- diated by an endogenous virus could be demonstrated in the case of E. chlorotica. then an exemplary mechanism for this process would have been provided.
Methods
Viral isolation
The fractionation procedure was carried out using au- toclaved equipment. All reagents were molecular biolog- ical grade (DNAase-, RNAase-. and protease-free; from Sigma Chemicals, unless otherwise noted) and filtered (0.2 /xm pore). Approximately 3.0 g of E. chlorotica was homogenized in an ice-cold buffer (450 mM NaCl, 1.0 mM EDTA, 5.0 mM 3-[/V-morpholino]propane-sulfonic acid (MOPS), 2.3 yuM leupeptin. 1.0 mM dithiothreitol (DTT), 500 /xM phenylmethylsulfonyl fluoride (PMSF), pH 7.5) containing the mucolytic agent H-acetyl cysteine (500 mM), which is necessary to disperse the copious mucus produced by the slug (26). The homogenate was filtered through six layers of cheesecloth, then through one layer of Miracloth (Calbiochem), and finally through two layers of Miracloth. The filtrate was centrifuged for 5 min at 4300 x g (4°C), and the supernatant was centrifuged at 20,000 X g for 30 min (4°C). The super- natant from the second spin was layered over a 20% sucrose cushion and centrifuged at 180,000 X g for 2 h (4°C) in a swinging bucket rotor. The supernatant was discarded, and the pellet was resuspended in ice-cold homogenization buffer (without H-acetyl cysteine). This suspension was layered on the top of a 15%-50% con- tinuous sucrose gradient and centrifuged at 180,000 X g for 43 h (4°C).
The gradients were then fractionated by piercing the bottom of the centrifuge tube and raising the gradient out of the tube with 65% sucrose. Twenty 600-/xl fractions were collected, and the density of each was determined by weigh- ing 50 /xl with an analytical balance. The sucrose in each fraction was then diluted with homogenization buffer (with- out /i-acetyl cysteine), and each fraction was centrifuged a final time at 180,000 X g for 2 h. The supernatants from this last spin were discarded, and RT assays (see below) and protein assays (37) were run on the pellets.
Reverse transcriptase assav
The final pellets (above) were treated with a detergent buffer (50 mM Tris-HCl, 5 mM KC1, 0.2 mM EDTA. and 0.02% Triton X-100, pH 8.2). Fifteen-microliter aliquots of this digest were added last to 50 /xl of buffer (100 mM Tris-HCl, 200 mM KC1. 10 mM MgCl2, pH 8.2), ft /xl 100 mM DTT, 9.75 LI! 100 mM thymidine triphosphate (TTP), 1.25 /xl RNA-guard (Pharmacia), 1.0 /xl poly(rA)-p(dT) (Pharmacia), and 10 /xCi of 32P-TTP (ICN, 10 /xCi//xl). The final volume was adjusted to 100 /xl with buffer, as neces- sary, to compensate for 12P half life. This solution was mixed and incubated at 37°C for 65 min on a shaker table. The reaction was terminated by adding 30 /xl of 10 mM EDTA in 5% TCA and placing the reaction mixture on ice for 20 min. DN A was then precipitated by the addition of 9 /xl of 72% TCA, and the precipitate was pelleted by cen- trifugation at 6610 X g for 10 min. The pellet was washed three times in 5% TCA, the final pellet dissolved in 0.1 N NaOH. and the radioactivity determined by scintillation counting. The protein concentration of a separate aliquot was determined, and RT activity was expressed as counts per minute per microgram (cpm//xg) protein (38).
Electron microscopy
Small pieces of tissue were prepared for microscopy by fixation in 2% glutaraldehyde in 0.15 M cacodylate-0.58 M sucrose buffer (pH 7.3) at 900 mosm. The tissue pieces were post-fixed in 1 .0% OsO4 in the same buffer followed by 2.0% aqueous uranyl acetate. The fixed tissue was dehy- drated in an ethanol series, infiltrated with propylene oxide, and embedded in Spurr's medium. Silver sections were cut with an ultramicrotome (Reichert), mounted on 75 X 300 mesh copper grids, and stained with uranyl acetate and lead citrate. The sections were viewed and photographed with a transmission electron microscope (Zeiss EM 10).
Acknowledgments
This work was supported by an NSF grant to SKP and MER. We thank Margaret Palmer, Ulrich Mueller, and Jeffery DeStefano for critically reading early versions of
6
S. K. PIERCE ET AL
this paper. Contribution #91 from the Laboratory of Bio- logical Infrastructure.
Literature Cited
1
D'Amico, V., J. S. Elkington. G. Dwyer, R. B. Willis, and M. E. Montgomery. 1998. Foliage damage does not affect within-season transmission of an insect virus. Ecology 79: 1 104-1 1 10.
2. Hawkins, B. A., H. V. Cornell, and M. E. Hochburg. 1997. Pred- ators. parasitoids and pathogens as mortality agents in phytophagous insect populations. Ecology 78: 2145-2152.
3. Kohler, S. L., and M. L. Wiley. 1997. Pathogen outbreaks reveal large-scale effects of competition in stream communities. Ecology 78: 2164-2176.
4. Rothman. L. D. 1997. Immediate and delayed effects of a viral pathogen and density on tent caterpillar performance. Ecology 78: 1481-1493.
5. West, H. H. 1979. Chloroplast symbiosis and development of the ascoglossan opisthohranch Elysia ch/ororicu. Ph.D. dissertation. Northeastern University, Boston. 161 pp.
6. West, H. H., J. F. Harrigan, and S. K. Pierce. 1984. Hybridization of two populations of a marine opisthohranch with different develop- mental patterns. Veli^er 26: 199-206.
7. Hinde, R., and D. C. Smith. 1974. "Chloroplast symbiosis" and the extent to which it occurs in Sacoglossa (Gastropoda: Mollusca). Biol. J. Linn. Soc. 6: 349-356.
8. Clark. K. B., and M. Busacca. 1978. Feeding specificity and chlo- roplast retention in four tropical Ascoglossa. with a discussion of the extent of Chloroplast symbiosis and the evolution of the order. J. Molluxcan Stud. 44: 272-282.
9. Clark, K. B., K. R. Jensen, and H. M. Strits. 1990. Survey for functional kleptoplasty among West Atlantic Ascoglossa (= Saco- glossa) (Mollusca: Opisthobranchia). Veliger 33: 339-345.
Clark. K. B., K. R. Jensen, H. M. Strits, and C. Fermin. 1991.
Chloroplast symbiosis in a non-elysiid mollusc, Costasicl/a lilianac
Marcus (Hermaeidae: Ascoglossa (=Sacoglossa): effects of tempera-
ture, light intensity and starvation on carbon fixation rate. Biol. Bull.
160: 43-54.
Taylor, D. L. 1970. Chloroplasts as symbiotic organelles. Int. Re\:
Cytol. 27: 29-64.
Gilyarov, M. S. 1983. Appropriation of functioning organelles of
food organisms by phytophagous and predatory opisthohranch mol-
lusks as a specific category of food utilization. Zli. Ohxlich. Binl. 44:
614-620.
13. Waugh, G. R., and K. B. Clark. 1986. Seasonal and geographic variation in chlorophyll level of Elysia tuca (Ascoglossa: Opistho- branchia). Mar. Biol. 92: 483-488.
14. Greene, R. W. 1970. Symbiosis in sacoglossan opisthobranchs: functional capacity of symbiotic chloroplasts. Mar. Biol. 7: 138-142.
15. Greene, R. W., and L. Muscatine. 1972. Symbiosis in sacoglossan opisthobranchs: photosyntheUc products of animal-chloroplast associ- ations. Mar. Biol. 14: 253-259.
16 Graves, D. A., M. A. Gibson, and J. S. Bleakney. 1979. The
digestive diverticula of Alderia inmlexiu and Elysia clilorotica. Veliger
21: 415-422. 17. Trench, R. K. 1969. Chloroplasts as functional endosymbionts in
the mollusc Tridachiu crispalu (Bcrgh), (Opisthobranchia, Saco-
glossa). Nature 222: 1071-1072. IX Trench, M. E., R. K. Trench, and L. Muscatine. 1970. Utilization
ol photosynthetic products of symbiotic chloroplasts in mucous syn-
10.
11
12.
thesis by Placobranchus iantlmharixiis (Gould). Opisthobranchia, Sacoglossa. Comp. Biochem. Physio/. 37: 113-117.
19. Greene, R. W. 1970. Symbiosis in sacoglossan opisthobranchs: translocation of photosynthetic products from Chloroplast to host tis- sue. Malacologia 10: 360-380.
20. Trench, R. K., J. E. Boyle, and D. C. Smith. 1973. The association between chloroplasts of Codium fragile and the mollusc, Elysia viridis II. Chloroplast ultrastructure and photosynthetic carbon fixation in £. )•//•/<//,(. Proc. R. Soc Land. B 184: 63-81.
21. Trench, R. K. 1975. Of "Leaves That Crawl": functional chloro- plasts in animal cells. Soc. Exp. Biol. Cambridge Svmp. 29: 229-266.
22 Greenberg, B. M., V. Gaba, O. Cananni, S. Malkin, A. T. Matoo, and M. Edelman. 1989. Separate photosensitizers mediate degrada- tion of the 32 kDa photosystem II reaction center protein in the visible Lind UV spectral regions. Proc. Nail. Acad. Sci. USA 86: 6617-6620.
23. Matoo, A. K., J. B. Marder, and M. Edelman. 1989. Dynamics of the photosystem II reaction center. Cell 56: 241-246.
24. Barber. J., and B. Andersson. 1992. Too much of a good thing: Light can be bad for photosynthesis. Trends Biochem. Sci. 17: 61-66.
25 Berry-Lowe, S. L., and G. W. Schmidt. 1991. Chloroplast protein transport. Pp. 257-302 in The Molecular Biology of Plastids: Cell Culture and Somatic Cell Genetics of Plants Vol. 7A. L. Bogorad and I. K. Vasil, eds. Academic Press, New York.
26. Pierce, S. K., R. W. Biron, and M. E. Rumpho. 1996. Endosym- biotic chloroplasts in molluscan cells contain proteins synthesized after plastid capture. J. £v/». Biol. 199: 2323-2330.
27. Mujer, C. V., D. L. Andrews, J. R. Manhart, S. K. Pierce, and M. E. Rumpho. 1996. Chloroplast genes are expressed during in- tracellular symbiotic association of Vaucheriu litorea plastids with the sea slug Elysiii clilorotica. Proc. Nail. Acad. Sci. USA 93: 12333- 12338.
28. Green, Brian J., W-y Li, J. R. Manhart, T. C. Fox, R. A. Kennedy, S. K. Pierce, and M. E. Rumpho. 1999. Mollusc-algal chloroplast endosymbiosis: chloroplast gene expression, thylakoid protein main- tenance and synthesis, and photosynthetic activity continue for many months in the absence of the algal nucleus. Plant Phvsiol. (in press).
29. Maugel, T. K.. and S. K. Pierce. 1994. Is the life cycle of Elysia clilorotica ended by disease? Proc. 52nd Ann. Meeting Microsc. Soc. Am. 266-267.
30. Murphy, F. A., C. M. Fauquet, D. H. L. Bishop, S. A. Ghabrial, A. W. Jarvis, G. P. Martelli, M. A. Mayo, and M. D. Summers. 1995. Virus taxonomy. Arch. Vim/. Suppl. 10, 586 pp.
31. Goto, T.. M. Nakai, and K. Ikuta. 1998. The life-cycle of human immunodeficiency virus type 1. Micron 29: 123-138.
32. Baker, T. S., S. W. Suh, and D. Eisenberg. 1977. Structure of ribulose- 1 ,5-bisphosphate carboxylase-oxygenase: Form III crystals. Proc. Nail. Acad. Sci. USA 74: 1037-1041.
33. Shumway, L. K., T. E. Weier, and C. R. Stocking. 1967. Crystal- line structures in Vicia faba chloroplasts. Plaiita 76: 182-189.
34 Francki. R. I. B., R. G. Milne, and T. Hatta. 1985. Atlas of Plant Virusex. Vol. I and II. CRC Press. Boca Raton. FL.
35. Pennisi, E. 1999. Is it time to uproot the tree of life? Science 284: 1305-1307.
36. Sikorski, R., and R. Peters. 1998. Treating with HIV. Science 282: 1438.
37. Peterson, G. L. 1977. A simplification of the protein assay method ot Lowry. ft ai. which is more generally applicable. Anal. Biochem. 83: 346-356.
38. Adolph, K. W. 1996. Viral Ccnomc Methods. CRC Press, Boca Raton, FL. p. 122.
Reference: Bio/. Bull. 197: 7-10. (August 1999)
Mechanical Resistance to Shear Stress: The Role of Echinoderm Egg Extracellular Layers
FLORENCE I. M. THOMAS1'2'* f, KRISTEN A. EDWARDS1 2, TOBY F. BOLTON1'2'*, MARY A. SEWELL\ AND JILL M. ZANDE1
1 The Marine Environmental Sciences Consortium, Dauphin Island Sea Lab. Dauphin Island. Alabama:
2 The Department of Marine Sciences. The University of South Alabama. Mobile. Alabama: and
3 The Department of Biology, University of Southern California. Los Angeles. California
Extracellular layers (jelly coats) on echinodenn eggs are composed of a fibrous network imbedded in a gelatinous material. This type of fibrous net\vork has the potential to protect eggs from mechanical stress. To determine the ef- fects of shear stress and the role of jelly coats in protecting eggs from these stresses, eggs of the sea urchin Ly (echinus variegatus, both with and without intact jelly coats, were exposed to shear stresses ranging from 0.3 to 2 Pa in a cone and plate viscometer. The percentage of eggs remaining intact after exposure to the shear stress was assessed. The results indicate that shear stress can damage eggs and that jellv coats ma\ play a role in decreasing the effects of these stresses. Eggs with jell\ coats remained intact and fertili'- able at greater shear stresses than those with the coats removed. This is the first evidence that extracellular layers on invertebrate eggs can provide protection from mechan- ical forces.
In free-spawning invertebrates, the eggs and sperm, hav- ing passed through a gonoduct and gonopore, are released directly into the water column. During the spawning pro- cess, gametes are exposed to shear stresses (force per area) in the gonoduct and in the external environment. If gametes are harmed by shear stress before fertilization, irreversible damage will effectively lower the probability of fertilization by decreasing the number of viable gametes in a volume of seawater. Moreover, if gametes are damaged by shear stress, they may survive and be fertilized, but the resultant embryos may not develop normally. Therefore, the number of larvae
Received 16 April 1998; accepted 2 June 1999.
* Present address: Department of Biology, University of South Florida. Tampa. Florida 33620-5 1 50
+ To whom correspondence should be addressed. E-mail: fthomas@ chuma 1 .cas.usf.edu
produced by a given number of eggs may be reduced by exposure to shear stress.
Eggs are first exposed to shear stresses as they traverse the gonoduct on their way to the gonopore. As fluid moves down a tube such as the oviduct, a shear gradient develops in the fluid and imposes a shear stress on the fluid and eggs within the duct. Estimates of shear stresses in the oviduct of one species of sea urchin, Arbacia punctulata, ranged from near zero at the center of the duct to over 41 Pa near the wall of the duct ( 1 ). This magnitude of shear stress exceeds that estimated to occur in the external environment (2), where gametes meet their second challenge from shear stress after they are released from the gonopore. After eggs are re- leased, they encounter shear stress produced by the interac- tion of eddies within the water column or within the mo- mentum boundary layer at the surface of the spawning adult. The effects of such shear stress on fertilization and devel- opment have only recently been addressed in a hydrody- namic study of fertilization in the purple sea urchin Strongy- locentrotus purpuratits (2). High shear stress produced experimentally in a Couette cell resulted in low fertilization success and abnormal embryo development. Mead and Denny (2) postulate that high shear stress, as produced in turbulent environments, might reduce the encounter rates of gametes during fertilization, decrease the ability of sperm to remain attached to the egg after contact, and damage eggs prior to fertilization and embryos after fertilization.
Given the two sources of shear stress that eggs experience prior to fertilization, it is possible that there has been selec- tion for egg properties that could reduce the damage caused by these stresses. This idea is supported by the fact that at least one property of echinoid eggs, their viscosity, is pos- itively correlated to wave exposure in three species (3). One
F. I. M. THOMAS ET AL.
10C
1.0 1.5
Shear Stress (Pa)
Figure 1. Percentage of Lylechinus variegatus eggs remaining intact after exposure to shear stresses. Sea urchins were collected from St. Joseph's Bay. Florida, transported to the Dauphin Island Sea Lab, Ala- bama, and maintained in laboratory aquaria. Additional animals were obtained from the Carolina Biological Supply Company. Urchins were maintained in seawater collected from St. Joseph Bay (salinity 29-30 ppt). aerated continuously, and fed either spinach or lettuce. We obtained eggs by injecting urchins with 0.5 M KC1. Eggs were used immediately after spawning. Paired samples of eggs with and without jelly coats from each of five females were prepared by removing the jelly coats from half of the eggs collected from a female. The coats were removed using a technique commonly employed in embryology (19). Eggs were poured through a plankton screen (Nytex) with 202-;u.m-diameter pores. To determine whether the jelly coats had been removed from these eggs, a 200- /xl aliquot was pipetted onto a depression slide and Sunn ink was used to visualize the edges of the jelly coats under a compound microscope at 100 x magnifi- cation (20). To minimize the probability of damaging the eggs by subject- ing them to unnecessary passes through the plankton screen, the eggs were checked for the presence of jelly coats after every two pours.
To assess the effects of shear stress and the potential effect of jelly coats on egg survival, eggs were exposed to a range of shear stresses in a cone and plate viscometer (Brookfield Digital Viscometer, DV-II) attached to a constant temperature bath (Neslab RTE-8). Shear stress was changed by adjusting the shear rate and the viscosity of the fluid (/j.). The viscosity was altered by adding hydroxyethyl cellulose (Proto-slo) to filtered seawater. Eight viscous fluids were tested: KY jelly (chlorhexidine gluconate); hydroxyethyl cellulose; polyvinylpyrrolidone; polyvinylpolypyrrohdonc; Percoll; methylcellulose; Dextran; and egg homogentate. Two criteria were used to select a fluid: (1) after exposure to the fluid, eggs both with and without jelly coats remained viable; and (2) no cell leakage was detected with visual inspection. Hydroxyethyl cellulose was the only fluid that met both criteria, so it was chosen for use in the experiment. Experiments at shear stresses up to 1 .5 Pa were also conducted in seawater only, and the results indicated that use of the hydroxyethyl cellulose could not account for all of the egg loss.
A known number of eggs in 0.5 ml filtered seawater were exposed to a given shear stress for 2 min. The shear stress significantly affected the survival of eggs with (open circles: arcsine % intact = 1.57-0.07* shear stress, r = 0.48, P < 0.001; 95% confidence limits for the slope are 0.04 to 0.10) and without jelly coats (closed circles: arcsine % intact = 1.53- 0.28* shear stress, r = 0.91. P < l/.OOOl, 95% confidence limits for the slope are 0.24 to 0.32). For those with icily coats, there was little effect of shear stress: percent survival ranged from 100% to 96.7% over the entire range of shear stresses. For those without jelly coats, the effect of shear stress was more apparent: percent survival ranged from 100% to 82.0%.
property of eggs that has the potential to decrease the effects of shear stress is the extracellular layer, or jelly coat, that encapsulates the eggs of echinoderms (4). These coats are composed of a fibrous network imbedded in globular gly- coprotein (5) and can account for a significant portion of the maternal energy invested in an egg (Bolton and Thomas, unpubl. data). Jelly coats are described for both echinoids and asteroids and consist of several concentric layers of complex fibrous networks (5-10) that are reminiscent of engineering materials designed to withstand shear stresses (11, 12). Although the jelly coat in echinoderms is involved in many parts of the fertilization process (13-18), it seems unlikely that this complex network is required for any of these functions. Thus it is possible that the structure of jelly coats plays a role in protecting eggs from shear stresses experienced in the external environment after spawning, in the oviduct during spawning, or both. The purpose of the research presented here is to explore whether jelly coats can provide this protection in the sea urchin Lytecliinus varie- gatus.
To examine the potential role of jelly coats in resisting shear stress, we exposed eggs with and without jelly coats to shear stresses in a cone and plate viscometer and determined the percentage of those eggs that remained intact and fer- tilizable. A greater percentage of eggs with jelly coats remained intact after exposure to shear stress than did those with jelly coats removed (Fig. 1. Table I). A test for homo- geneity of slopes indicated that shear stress affected eggs with jelly coats differently than eggs with coats removed. Eggs with jelly coats suffered a maximum loss of less than 4% as shear stresses approached 2 Pa. No eggs were dam- aged until shear stresses in excess of 1.7 Pa were reached. For eggs without jelly coats, damage occurred at lower shear stresses than for those with coats, and maximum loss reached nearly 20% as shear stress approached 2 Pa.
The fertilization success of eggs exposed to shear de- creased with increasing shear stress both for eggs with jelly coats and for those without (Fig. 2). For both egg types, fertilization success was near 85% with no shear (the de- Table I
Effect of shear stress on percentage of eggs remaining intuci
Source F P
With/without coats (same intercepts) 1.99 0.165
Shear stress (slope = 0) 0.20 <0.001
Interaction (same slopes) 0.75 <0.001
Results from a test for homogeneity of slopes (1 degree of freedom I performed on the arcsine-transfonned percentage of eggs intact after ex- posure to shear stress versus the magnitude of shear stress for eggs with and without jelly coats. There was a significant effect of shear stress on eggs remaining intact (slope is significantly different from 0). Eggs with and without jelly coats were affected by shear stress to a different degree (significant interaction between shear stress and jelly coat presence).
STRESS RESISTANCE IN SEA URCHIN EGGS
sired initial fertilization success). As shear increased, fertil- ization success dropped from 85% to 68% for eggs with coats and from 85% to 59% for eggs without coats. A test for homogeneity of slopes indicated that the shear stress affected fertilization differently for eggs without intact coats than for those with coats present (Fig. 2. Table II). The slope of the regression of fertilization versus shear stress was steeper for eggs with jelly coats removed than for those with intact coats (Fig. 2). After fertilization, developing embryos were assessed for developmental stage and normal devel- opment up to the pleuteus stage. Development was assessed throughout this period. All fertilized eggs, regardless of
80-1
75-
£ 70
•a
I 65
60-
55
Tank II
Effect oj shear stress on fertilization
50
0.0
0.5
1.0 1.5
Shear Stress (Pa)
2.0
— I 2.5
Figure 2. Percentage of eggs successfully fertilized after exposure to shear stress. The fertilizability of eggs with and without intact jelly coats was determined after eggs were exposed to a given shear for 2 nun Gametes for all experiments were obtained by injecting urchins with 0.5 M KCl. Females were allowed to spawn into filtered seawater; males spawned into a dry petri dish. Sperm were stored undiluted in a scintillation vial over ice until being used in the experiments. Eggs were used immediately after spawning and sperm within I h after spawning. The concentration of sperm yielding 80%-85% successful fertilization (I04 dilution of dry spawned sperm in seawater) was determined from serial dilution experiments. We used this sperm dilution in all fertilization experiments to ensure that any fertilization decrease caused by egg damage would not be concealed by an overabundance of sperm (2). After exposure to shear, the eggs were added to 100 ml of seawater with the sperm solution. Un-sheared eggs were also fertilized as a control. A sample of approximately 60-100 eggs was examined for fertilization success. Fertilization was assessed at the four- cell stage of development, and normal development was determined after embryos reached the early pluteus stage.
There was a negative relationship between fertilization success and exposure to shear stress, both for eggs with jelly coats (open circles: arcsine % fertilized = 0.92-0.07* shear stress, r = 0.88, P < 0.001. 95% confidence limits for the slope are -0.04 to -0.09) and for those without (closed circles: arcsine % fertilized = 0.87-0.1 1* shear stress, r = 0.90. P < 0.001. 95% confidence limits for the slope are -0.08 to -0.15). Only eggs exposed to shear in the viscometer were used in these analyses. Results of a test for homogeneity of slopes indicates that shear stress affects fertilization success of eggs with jelly coats less than that of those without jelly coats (Table II).
|
Source |
F |
P |
|
With/without coats (same intercepts) |
2.6 |
0.13 |
|
Shear stress (slope = 0) |
0.001 |
<0.001 |
|
Interaction (same slopes) |
5,90 |
0.032 |
Results from a test for homogeneity of slopes (1 degree of freedom) performed on the arcsme-transformed percentage of eggs fertilized after exposure to shear stress versus the magnitude of shear stress for eggs with and without jelly coats. There was a significant effect of shear stress on fertilization (slope is significantly different from 0). Eggs with and without jelly coats were affected by shear stress to a different degree (significant interaction between shear stress and jelly coat presence).
whether they had jelly coats or not. developed normally. This result indicates that if eggs are damaged by shear stress they are not fertilizable, but if they survive they apparently have not sustained any damage that limits normal develop- ment.
The results of these experiments indicate that eggs can be damaged by shear stress and that the jelly coats on echino- derm eggs can provide mechanical strength, reducing the negative effects of shear stress on free-spawned eggs. If jelly coats are absent, survivorship (Fig. 1) and fertilization success (Fig. 2) are significantly less than when coats are present.
The shear regime experienced in the viscometer most closely approximates that seen in the gonoduct. The exper- imental shear is unidirectional, constant, and well within the range estimated to occur in the gonoduct during spawning ( 1 ). In contrast, the shear stresses imposed in the external environment not only are lower (2) than those used in these experiments, but are not constant and are unlikely to be unidirectional for sustained periods of time. Thus, the data presented here indicate strongly that eggs are susceptible to shear stress and can be damaged at stress levels in the range experienced during egg release. Therefore, the data show that jelly coats on echinoderm eggs do protect eggs from damage caused by shear stress. This result is significant because it is the first evidence that extracellular layers on invertebrate eggs can protect eggs from physical stress and may provide mechanical strength to eggs.
Acknowledgments
Research support was provided by a National Science Foundation (NSF) grant (IBN-9723770) and an NSF PECASE award to F. I. M. Thomas (OCE-9701434). Kris- ten Edwards initiated the research while participating in a research experience for undergraduate program funded by a NSF Grantto Judy Stout (REU. OCE-9619862). We also thank three anonymous reviewers for helpful suggestions. This is DISL contribution 309.
10
F. I. M. THOMAS ET AL
Literature Cited
1. Thomas, F. I. M., and T. F. Bolton. In Press. Shear stress expe- rienced by echinoderm eggs in the oviduct during spawning: potential role in the evolution of egg properties. / Exp. Biol.
2. Mead, K. S., and M. W. Dennv. 199S. The effects of hydrodynamic shear stress on fertilization and early development of the purple sea urchin Strongylocentrotus purpuratus. Biol. Bull 188: 46-56.
3. Thomas, F. I. M. 1994. Physical properties of gametes in three sea urchin species. J. £.171. Biol. 194: 263-2X4.
4. Hoshi. M. 1985. Lysins. Pp. 431-462 in Biology of Fertilisation. Vol. 2. C. B, Metz and A. Monroy. eds. Academic Press. Orlando. FL.
5. Bonnell. B. S., S. H. Keller, V. D. Vacquier, and D. E. Chandler. 1994. The sea urchin jelly coat consists of globular glycoproteins hound to a fibrous fucan superstructure. Dev. Biol. 162: 313-324.
6. Kidd. P. 1978. The jelly and vitelline coats of the sea urchin egg: new ultrastructural features. J. Ultrasln/ct. Res. 64: 204-215.
7. Holland, N. D. 1980. Electron microscopic study of the cortical reaction in eggs of the starfish (Paririti minima). Cell Tissue Res. 205: 67-76.
8. Crawford, B., and M. Abed. 1986. Ultrastructural aspects of the surface coatings of eggs and larvae of the starfish. Pisaster ochraceus, revealed by Alcian Blue. J. Morpliol. 187: 29-37.
9. Sousa, M., R. Pinto, P. Moradas-Ferreira, and C. Azevedo. 1993. Histochemical studies of jelly coat of Marthasterias glacialis (Echi- nodermata, Asteroidea) oocytes. Biol. Bull. 185: 215-224.
10. Bonnell, B. S., C. Larahell, and D. E. Chandler. 1993. The sea urchin egg jelly coat is a three-dimensional fibrous network as seen by intermediate voltage electron microscopy and deep etching analysis. Mol Reprod. Dev. 35(2): 181-188.
11. Sastry, A. M., S. L. Phoenix, and P. Schwartz. 1993. Analysis of
mterfacial failure in a composite microbundle pull-out experiment. Composite Sciences and Technology. 48: 237-251.
12. Sastry, A. M., X. Cheng, and C. W. Wang. 1998. Mechanics of stochastic fibrous networks. J. Thermoplastic Composite Materials. Vol. II. 288-296.
13. Vaquier, V. D., and G. W. Moy. 1977. Isolation of bindin: the protein responsible for adhesion of sperm to sea urchin eggs. Proc. Null. Aciul. Sci. USA 74: 2456-2460.
14 Tilney, L. G., D. P. Kiehart, C. Sardet, and M. Tilney. 1978. Polymerization of actin IV. Role of Ca2* and H+ in the assembly of actin and in membrane fusion in the acrosomal reaction of echinoderm sperm. / Cell Biol. 77: 536-550.
15 SeGall, G. K., and W.J. Lennarz. 1981. Jelly coat and induction of the acrosome reaction in echinoid sperm. Dev. Biol. 86: 87-93.
16. Garbers, D. L., and G. S. Kopf. 1980. The regulation of spermata- zoa by calcium and cyclic nucleotides. Adv. Cyclic Nucleotide Res. 13: 251-306.
17. Nomura, K., and S. Isaka. 1985. Synthetic study of the structure- activity relationship of sperm-activating peptides from the jelly coat of sea urchin eggs. Biochem. Biophys. Res. Commun. 126: 974-982.
18. Miller, R. L. 1985. Sperm chemo-orientation in the Metazoa. Pp. 276-337 in Biology of Fertilization. Vol. 2. C. B. Metz and A. Monroy. eds. Academic Press, Orlando. FL.
19. Hinegardner, R. 1975. Care and handling of sea urchin eggs, em- bryos, and adults (principally North American species). Pp. 10 — U in The Sea Urchin Embryo — Biochemistry and Morphogenesis. G. Czi- hak, ed. Springer- Verlag. New York.
20. Strathmann, M. F. 1987. Reproduction and Development of Marine Invertebrates of the Northern Pacific Coast. University of Washington Press, Seattle. 670 pp.
Reference: Biol. Bull. 197: I 1-13. (August 1999)
Regeneration of Amputated Avian Bone by a Coral
Skeletal Implant
B. RINKEVICH1, S. BEN-YAKIR2, AND R. BEN-YAKIR2
' National Institute of Oceanography, Tel Shikmona, P.O.B. 8030, Haifa 31080, Israel: and 2 Hod-Hasharon Veterinan< Clinic, 17 Gordon Street, Hod Hasttaron, Israel
Bone fractures are common in both wild and captive birds (I, 2). Avian bones are thin and brittle and tend to break into fragments or shatter upon a variety of natural events (midair collisions, fights with other animals; ref. 2) or anthropogenic experiences (wounding by gunfire, colli- sions with cars or fences, encounters with traps, attacks by dogs or cats, etc.; ref. 1). The prospect of full recovery following repair of avion bone fracture is often poor, and the complication rate is high (3). For wild birds, anything less than complete normal function cannot be regarded as successful, and slight malunion or a change in a few de- grees of rotation can produce a severe loss of flight function (4). Furthermore, in nomadic species, time is critical be- cause long periods of rehabilitation may prevent the birds from reuniting with their flocks. In experiments with implan- tation of fragments of skeleton from the coral Stylophora pistillata, we found the implants to be avion osteo-conduc- tive biomaterial, acting as a scaffold for a direct osteoblas- tic deposition. In the case study presented here, the bird regained complete flight activity within 2 weeks after sur- gery, with full regeneration of the amputated ulna.
The general principles for treating fractures in birds are similar to those established for mammals and include rigid stabilization (primary bone healing does not occur if there is a gap or motion at the fracture site; ref. 4). However, treatment such as external coaptation (slings, bandages, casts, splints, etc.), intramedullary pins or rods, bone plate fixation, or modifications of any of the traditional means of external skeleton fixation (3-5) not only fails for rehabili- tating wild birds, but also involves prolonged hospitaliza- tion of avian patients. Internal fixation is one of the best procedures for fracture management, but the brittle nature of
Received 26 January 1999; accepted 23 April 1999. E-mail: buki@ocean.org. il
avian bones (3-5) results in problems that are not encoun- tered in mammals.
We recently investigated the use of coral skeleton as a natural intramedullary fixation device for fractures of bird bone and found that fractured avian bones can be rehabili- tated following the internal implantation of coral skeletal pins (unpubl. data). This investigation was based on studies with mammals (including humans) documenting that coral skeletons may be employed as osseous substitutes, as scaf- folds for direct osteoblastic application, or as an artificial bone filler for repairing bone defects (6, 7). The coral tested here was of the branching species Stylophora pistillata, one of the most abundant coral species in the Gulf of Eilat (8).
Mature domestic pigeons (Colmnba Hvia domestica) were randomly assigned to a variety of treatment groups (in preparation). For all radiographic and surgical procedures, the birds were given Halothane as a general anesthetic. The skin was prepared for aseptic surgery using a septal scrub followed by a povidone-iodine wash. Whenever possible, flight feathers were not removed or clipped; others were plucked over the intended incision. Reflexes and cardiac and respiratory activities were monitored. Birds were placed on their backs and a limited ventral approach to the ulna was performed. Two of the same bones from the wings of two birds were used as experimental and control bones (ban- daged in the traditional manner; ref. 5). Small processed coral pins were obtained from SagivCoral, P.O. Box 3337, Ramot Hashavim 45930, Israel. Postoperative radiographs were taken to evaluate fracture repair and to document the status of the coral implant. Pigeons were housed in the flypen system and fed with a commercial pigeon food supplemented with vitamin D and oyster shell as a calcium source.
In one of the experiments, which is detailed in this communication, the proximal half of the ulna (which pro-
11
B. RINKEVICH ET AL
Figure 1. Progressive repair of ulna fracture treated hy implantation of an intramedullary coral pin (a-e), compared to control, an untreated amputated ulna (fl. Weeks after operation: a = 2. h = 4, c = 6, d = 8, e, f = 12.
vides primary support for the wing) was accidentally com- minuted during surgery. Full rehabilitation of this amputa- tion hone by the use of coral skeleton implant is described here.
In the case of the accidentally comminuted ulna, all brittle fragments were immediately removed and a small coral pin (24 mm length. 4 mm diameter) was first passed distally and then retrogradely until resistance was met at the proximal side of the ulna. The coral pin was then firmly wedged in place, forming an inert calcium carbonate milieu between the two separated parts, replacing the amputated portion of the ulna. This pigeon used the treated wing freely 14 days after the operation, alleviating ankylosis resulting from joint immobilization. In the control bird, the entire segment of proximal ulnar bone (cortex and medulla) was removed using Gilgi wire.
Two weeks after the operation, the coral pin was encap- sulated firmly at both ends by overgrown calcium deposits and callus formation along the pin shall, providing rota- tional stability (Fig. la). After 4 weeks (Fig. Ib). the coral implant was already overgrown by deposited material. By 6 weeks (Fig. Ic), deposited material surrounded the implant
in layers and the first sign of coral resorption was evident. During resorption. which was significantly advanced at 8 and 12 weeks (Fig. Id, e), radiography showed that the area between the two ends of the broken ulna was being filled with accumulated new bone, replacing the degradable pin. Sft'liiplwni pistillntti skeleton (although its mechanical and biological properties were not yet evaluated) was thus found to be avian osteo-conductive biomaterial, acting as a scaf- fold for direct osteoblastic deposition. The bird regained full flight activities 2 weeks after surgery, and the coral pin activated skeletal regeneration (compare with the control; Fig. If). This process ended in complete regeneration of the amputated area.
This case of regeneration of an amputated bone and our study (in prep.) demonstrate the value of coral skeletal implants for avian bone repair. Coral material (calcium carbonate) is well tolerated by bird tissue. The pin matrix is porous enough to be colonized by the birds' bony cells, is biodegradable, and is easily adjustable in size and shape to the osseous site of grafting. Previous studies employing coral implants for bone repair in mammals have shown that coral resorption rates varied with porosity of the coral
CORAL IMPLANTS IN AVIAN FRACTURES
13
species used and with host reaction (9). We used natural fragments of 5. pistillata skeletons, the first pocilloporid coral used in vertebrate skeleton rehabilitation. Each year, around the globe, veterinarians tend an ever-increasing number of wild and domestic birds with broken bones; unfortunately, at present the prognosis for many of these birds is poor. The approach described here may provide a fast and dependable method for rehabilitation of avian frac- tures, increasing the survival rate of birds treated for bone injuries.
Acknowledgments
This study was supported by the Minerva Center for Marine Invertebrate Immunology and Developmental Biol- ogy. Animal surgeries and treatments were conducted in conformance with the guidelines of the Canadian Council on Animal Care.
Literature Cited
1. Fix, A. S., and S. Z. Barrows. 1990. Raptors rehabilitated in Iowa during 1986 and 1987: a retrospective study. J. Wildl. Dis. 26: 18-21.
2. Houston, D. C. 1993. The incidence of healed fractures to wing bones of White-backed and Ruppell's Griffon Vultures Gyps africaiius and G. ntt'ppt'tln and other birds. Ibis 135: 468-475.
3. Mathews, K. G., L. J. Wallace, P. T. Redig, J. E. Bechtold, R. R. Pool, and V. L. King. 1994. Avian fracture healing following stabi- lization with mtramedullary polyglycolic acid rods and cyanoacrylate adhesive vs. polypropylene rods and polymethylmethacrylate. Vel. Comp. Orthop. Trauma 7: 15X-169.
4 Bennett, R. A., and A. B. Kuzma. 1992. Fracture management in birds. J. Zoo. Wildl Meil. 23: 5-38.
5. MacCoy, D. M. 1992. Treatment of fractures in avian species. Vet. dm. North Am. Small Aiiini. Pract. 22: 225-238.
6. Kehr, P. H., A. G. Graftiaux, F. Gosset, I. Bogorin, and K. Ben- cheikh. 1993. Coral as a graft in cervical spine surgery. Ortlmp. Traumarol. 3: 287-293.
7 Guillemin, G., J-L. Patat, and A. Meunier. 1995. Natural corals used as bone graft substitutes. Bull. lust. Oceanogr. (Monaco) 14: 67-77.
8. Loya, V. 1976. The Red Sea coral Stylophora pistillata is an r strategist. Nature (Land.) 259: 478-480.
9. Roudier, M., C. Bouchon. J. I.. Rouvillain, J. Amedee, R. Bareille, F. Rouais, J. Ch. Fricain, B. Dupay, P. Kien, R. Jeandot, and B. Basse-Cathalinat. 1995. The resorption of bone-implanted corals varies with porosity but also with the host reaction. J. Biomed. Mater. Res. 29: 909-915.
Reference: Biol. Bull. 197: 14-25. (August 1999)
Functional and Evolutionary Implications of Opposed
Bands, Big Mouths, and Extensive Oral Ciliation in
Larval Opheliids and Echiurids (Annelida)
BENJAMIN G. MINER1, ERIC SANFORD2, RICHARD R. STRATHMANN3, BRUNO FERNET3,
AND RICHARD B. EMLET 4
1 Department of Zoology, University of Florida, 223 Bartram Hall, Gainesville. Florida 3261 1;
2Department of Zoology, Cordley Hall 3029, Oregon State University, Corvallis, Oregon 97331;
3 Friday Harbor Laboratories and Department of Zoology, University of Washington, 620 University
Road, Friday Harbor, Washington 98250; and ^Department of Biology and Oregon Institute of
Marine Biology, University of Oregon, P.O. Box 5389, Charleston. Oregon 97420
Abstract. Larvae of two annelids, the opheliid Armandia brevis and the echiurid Urechis caupo, captured small par- ticles between opposed prototrochal and metatrochal ciliary bands and also captured large particles with wide ciliated mouths. The body volume of larval A. brevis increased more rapidly than the estimated maximum clearance rate as seg- ments were added. Capture of larger particles by late-stage larvae may compensate for this potentially unfavorable al- lometry. The existence of larvae that use two feeding mech- anisms at once, not previously known in annelids, suggests possible evolutionary routes between larval forms that feed only with opposed bands (e.g., serpulids and oweniids) and those that use complex oral ciliature to feed primarily on large particles (e.g.. polynoids and nephtyids). In particular, the metatroch and food groove of opposed-band feeders may have arisen as expansions of oral ciliation in ancestral large-particle feeders; alternatively, extensive oral ciliation in large-particle feeders may have originated as a modifi- cation of metatroch and food-groove cilia in ancestral op- posed-band feeders.
Introduction
The trochophore is a larval form of several phyla: Anne- lida, Sipuncula, Mollusca. and Entoprocta (Nielsen, 1995). It is largely denned by the presence of the prototroch, a preoral ciliary band with a well-defined cell lineage. Despite
Received 16 December 1998; accepted 8 April 1999. E-mail: miner@zoo.ufl.edu
this and other embryological similarities, trochophores are structurally and functionally diverse. Much of this diversity is found among the approximately 70 families of annelids in which larvae occur. Annelid larvae vary in the number and position of ciliary bands (though almost all possess a pro- totroch). and in whether or not they feed. Among annelid larvae that feed, mechanisms of capturing suspended parti- cles have been described in only a few species (Strathmann. 1987).
One of these feeding mechanisms involves capturing and transporting particles with the prototroch and several postoral ciliary bands. The prototroch beats with an anterior-to-posterior effective stroke. A postoral band, the metatroch, parallels the prototroch and beats in opposition to it, with effective strokes from posterior to anterior. Par- ticles small enough to tit between the prototroch and metatroch are captured between these two ciliary bands and transported to the mouth by a band of shorter cilia, the food groove. Particle capture by opposed bands has been de- scribed in larvae of two annelid families, the serpulids and the oweniids (Strathmann et al.. 1972; Emlet and Strath- mann, 1994), and larvae of several other families possess the ciliary bands necessary to feed in this way.
Another feeding mechanism known in annelid larvae involves active responses to individual food particles. For example, polynoid larvae lack an opposing metatroch and a ciliated food groove. These larvae swim forward until they encounter relatively large particles, then manipulate each particle individually into the mouth with a tuft of long compound cilia (Phillips and Pernet, 1996). Larvae belong-
14
ANNELID LARVAL FEEDING MECHANISMS
15
ing to related families (e.g., phyllodocids and nephtyids: Rouse and Fauchald, 1997) also lack a metatroch and a food groove (Bhaud and Cazaux. 1987), and are able to capture particles as large as bivalve larvae, but how they do this is not known. Additional feeding mechanisms are known or suspected from larvae of other annelid families (Strath- mann, 1987; Nielsen, 1998).
The structural and functional variety of trochophores in annelids and related phyla has raised questions about their evolution (Strathmann, 1993; McHugh and Rouse, 1998). There is no consensus as to whether feeding or nonfeeding larvae are ancestral, or on which feeding mechanisms are primitive (Strathmann and Eernisse, 1994). Given uncer- tainties about such key issues as the distribution of traits among clades, the functional requirements for capturing particles, and the phylogeny of annelids, inferences about ancestral character states are weak.
Our study describes ciliation and mechanisms of particle capture in larvae of two families of annelids, the Opheliidae and the Echiuridae. We use these observations to compare the feeding capabilities of different annelid larvae and to suggest possible evolutionary transitions among annelid lar- val forms. These data also augment the number of informa- tive characters available for phylogenetic inferences.
Hermans (1978) showed that larvae of the opheliid Ar- iminJia brevis possess prototrochal and metatrochal ciliary bands. Although the feeding mechanism was not described, these observations suggest that opposed-band feeding may occur. He also noted that late-stage A. brevis larvae are able to ingest large particles. A larva with 15 segments had ingested a tintinnid 80 /im in diameter and a diatom 35 ^m in diameter and 260 /xm long (Hermans, 1964). This implies that these larvae were using a different feeding mechanism, since other work on annelid and mollusc larvae indicates that opposed-band feeding is limited to particles that fit between the prototroch and metatroch (typically spaced <30 /urn apart: Strathmann et ai, 1972; Strathmann and Leise, 1979).
Thus, limited observations suggested that A. brevis larvae might use several feeding mechanisms to capture particles of a broad range of sizes. Alternative mechanisms for the capture of larger particles by later stage larvae might sup- plement the opposed-band feeding mechanism. Such versa- tility might be particularly advantageous to later stage lar- vae if unfavorable allometric relationships reduce the profitability of opposed-band feeding as development progresses. An unfavorable allometry might occur if body volume and metabolic demands increase more rapidly than ciliary band area and maximum clearance rates as segments are added during development. Therefore, in addition to observing particle captures, we examined the relationship between clearance rates and body volume.
Echiurids have sometimes been placed in the phylum Annelida and sometimes in their own phylum, the Echiura,
which is distinguished from the annelids by an apparent lack of segmentation (Nielsen, 1995). McHugh's (1997) molec- ular evidence shows that they are derived annelids, and she suggests that they should be placed in the annelid family Echiuridae. Larvae of the echiurid Urechis caupo bear pro- totrochal, metatrochal, and food-groove cilia (Newby, 1940; Suer, 1982), but how they capture particles has been un- known. We observed larval feeding in U. caupo to confirm use of the opposed-band feeding mechanism in the Echi- uridae; to our surprise, we also obtained evidence that larger particles are captured at the mouth.
Our observations demonstrate that larvae in the annelid families Opheliidae and Echiuridae are able to capture par- ticles both with opposed bands and directly at the mouth. This previously unrecognized combination of feeding mechanisms suggests hypotheses for evolutionary transi- tions among the diverse feeding larval forms of the Anne- lida.
Materials and Methods
Larval cultures
Reproductive adults of the opheliid Armandia brevis were collected in April and May 1998 in front of the Friday Harbor Laboratories. San Juan Island. Washington. Some animals were taken from beneath cobbles in the mid-inter- tidal zone and others from the plankton swarming at night to a light suspended from the laboratory dock. We isolated adults in finger bowls containing bag-filtered seawater (mesh size ^ 10 /xm) until gametes were released. Eggs were fertilized by the addition of sperm and then rinsed with filtered seawater. Fertilized eggs were placed in 450-ml beakers that held filtered seawater and were partially sub- merged in a seawater table at 1 1°-13°C. Larvae were fed a mixture of the algae Isochrysis galhana and Chaetoceros gracilis.
Adults of the echiurid Urechis caupo were dug in inter- tidal mudflats in Bodega Harbor, California, in June of 1995 and held in aquaria at the Bodega Marine Laboratory for use throughout the summer. Methods described by Gould ( 1967) were used for obtaining gametes and fertilizing eggs. We reared larvae in 800-ml beakers cooled in aquaria at 10° to 16°C (median 13.3°C), approximately the temperature of the coastal seawater. The seawater was filtered through meshes of 30 or 70 /urn and larvae were fed the alga Rhodomonas sp. and occasionally Isochrysis galbana in addition to whatever food entered with the filtered seawater.
Ciliarv bands
Light microscopy provided information about the cilia- tion of both opheliid and echiurid larvae. Larvae were viewed with differential interference contrast (DIG) optics
16
B. G. MINER ET AL
for an optical section through the prototroch, food groove, and metatroch.
Scanning electron microscopy provided additional infor- mation about the ciliation ofArmandia brevis. Larvae were relaxed in a 1:1 mixture of 7.5% MgCl2 and seawater for 30 min and fixed in 1% OsO4 in seawater. After a rinse in seawater, fixed larvae were dehydrated in ethanol. infiltrated with hexamethyldisilazane for 30 min, and air-dried. They were mounted on stubs with double-sided tape and sputter- coated with gold-palladium before viewing.
Analysis of particle capture
To record larval feeding, we used video cameras mounted on compound and dissecting microscopes. A time-date gen- erator indicated intervals between video images to the near- est 0.01 s. Larvae of Armandia brevis were presented with small and large particles in separate trials, and feeding activity was recorded at room temperature (22°C) onto VHS tape. We observed capture of small particles by placing several larvae on a slide with polystyrene-divinylbenzene spheres (Duke Scientific) of 5 and 12 /xm diameter (one size per slide), adding a raised coverslip, and viewing the larvae with a 20 X objective and DIC optics. Larvae that had tethered themselves with mucous strands and were actively feeding (indicated by beating of both the prototroch and
metatroch) were videotaped for about 10 min. We observed capture of large particles by placing larvae in a small petri dish onto a dissecting microscope and adding Sephadex beads ranging from 20 to 80 /u,m in diameter. Larvae were videotaped as they swam and fed.
For Urechis caupo larvae, feeding was observed at 15° to 20°C and recorded onto 8-mm tape. Larvae were confined within the spaces of a nylon mesh placed on a slide topped with a coverslip; they were free to rotate and change orien- tation but not to move forward continuously. We presented the larvae with three types of particles: the dinoflagellate Prorocentrum micans (length about 20 /xm), polystyrene- divinylbenzene spheres (diameter 5 to 29 ju,m), and Seph- adex beads (diameter 20 to 80 ju.ni).
The size of particles captured and ingested by U. caupo was analyzed by inspecting the gut contents of particle-fed larvae. Larvae and suspensions of particles of several sizes were placed in vials that were rotated at 15 rpm. After 5 min, the larvae were fixed with formaldehyde for gut- content examination.
Scaling of clearance rate and bod\ volume
The relationship of maximum clearance rate to body volume was estimated for larvae ofArmandia brevix with 6 to 16 setigerous segments. We counted the number of seti-
Figure 1. Scanning electron micrographs of larvae of Armandia brevis. (A) Posterolateral view ol the anterior end of an IX-setiger larva. The food groove (*) is the region between the long compound cilia of the prototrochal (p) and melatrochal (m) ciliary bands. The inner surfaces of the mouth (mo) are heavily ciliated. (B) Ventral view of the anterior end of an 18-seliger larva, showing the long compound cilia of the metatroch (m) on the lower lip, the prototroch (p). and the .short neurotroch (n). Both photos are to the same scale.
ANNELID LARVAL FEEDING MECHANISMS
17
gers and measured body length (for the entire larva), width (at the middle segment), and prototroch diameter of live larvae (n = 36) under a compound microscope with 4X objective. A video camera and image analysis program (NIH Image 1.61: available free at http://rsb.info.nih.gov/ nih-image) were used for these measurements. We esti- mated larval volume as a cylinder by the equation:
larval volume = Tr(D/2)2(L)
where D is body width and L is body length.
Maximum clearance rates were estimated as the volume of water passing through the prototroch per unit of time. To calculate these rates, we measured particle velocities and particle distances to the base of the prototroch from video- taped sequences of three larvae in each of three size classes (6-7, 11-12, and 15-16 setigers). We observed larvae and 5-;u,m particles on a compound microscope with DIC optics and 20 X objective lens, as described above. The larvae tethered themselves by mucous strands and were recorded for several minutes. The distances traveled by particles per unit of time and their distances to the base of the prototroch were measured from videorecorded sequences. Particles
were measured as they passed within the direct influence of the cilia where velocities are negligibly affected by the slide or coverslip (Emlet, 1990).
We fitted binomial regressions from the origin through the plot of particle velocity versus particle distance from the cilium base. The rationale for fitting curvilinear lines to these data was both theoretical (Sleigh, 1984) and empirical (Strathmann and Leise, 1979). The studies in both areas suggest that velocity should increase from zero near the larval body surface to a maximum near the full length of the cilia; it should then decrease beyond the tips of the cilia. Since these curves included some particles that presumably passed beyond the tips of the cilia, it was necessary to estimate the lengths of the cilia for larvae of each of the three size classes. We measured cilium lengths (15 cilia per larva) with NIH Image from videotaped, live larvae with 0-17 setigers (n = 22 larvae). The binomial regression equations relating particle velocity to particle distance from the cilium base were then integrated from the origin (the base of the prototroch) to the estimated cilium length for that size class. The resulting areas represent estimates of the area of water that, in one unit of time, passes through one optical section of the prototroch in the plane of ciliary beat.
100|jrn
Figure 2. Light micrographs of larvae of Urecfus caupo. (A) Lateral view with plane of focus through the prototroch (p), food groove (0. metatroch (m), and telotroch (t). dorsally. Ventrally, the plane of focus passes through the prototroch (p). mouth (mo), and metatroch (m). The neurotroch is not visible. The dark spot near the center is a particle in the gut. (B) The same larva when contracted. Both photos are to the same scale.
18
B. G. MINER ET AL
We then estimated maximum clearance rates for each size class by multiplying that value by the circumference of the prototroch halfway between the base of the cilium and its tip (midpoint prototroch circumference). Finally, to determine whether maximum clearance rate scaled proportionately to body size during larval growth, we divided the maximum clearance rate for a given size class by the average body volume for that size class.
Results
Cilifin hands
Scanning electron micrographs clearly show the pro- totrochal and metatrochal cilia of Armandia brevis. The prototroch is made up of several rows of compound cilia that completely encircle the larval body anterior to the mouth (Fig. 1). The rnetatroch is a postoral band of compound cilia that extends laterally from the lower lip of the mouth around the larval body to a dorsal position
(Fig. 1A). The metatrochal cilia on the lower lip are longer than the other metatrochal cilia (Fig. 1A. B). The prototroch and rnetatroch define the boundaries of a cilia- lined food groove. The width of the food groove lateral to the mouth was estimated from a scanning electron mi- crograph to be 10 jam (SEM not shown). Dorsally. the food groove narrows. The mouth is large (about 50 ju,m wide in the 18-setiger larva shown in Fig. IB) and both its upper and lower surfaces are heavily ciliated (Fig. 1 A. B). A band of neurotrochal cilia runs along the ventral surface of the larva from just behind the mouth to the third setigerous segment (Fig. IB).
Larvae of Urechis caupo also possess prototrochal and metatrochal ciliary bands (Fig. 2). The prototrochal cilia are longer than the metatrochal cilia. Again, these two ciliary bands define the boundaries of a food groove lined with simple cilia. Larvae also bear a midventral neurotroch. posterior to the mouth, and a telotroch.
0.00
Figure 3. Videorecorded capture of a 5-/xm sphere by opposed hands, and particle rejection by a self-tethered Armandia brevis larva. Time in seconds is in the upper left-hand comer. All images are at the same magnification. The larva is oriented with its dorsal side toward the top of the page. The sequence shows a particle, indicated by the black line, approach the dorsal part of the metatroch where it is captured and then transported along the food groove and deposited in the mouth. The particle is then rejected. During rejection (he mclatroch on the lower hp ceases to beat. At 0 s the larva is 120 p.m in diameter at the base of the prototrochal cilia.
ANNELID LARVAL FEEDING MECHANISMS
19
Figure 4. Videorecorded capture of a 13-jum sphere by a Llrccliia cmipo larva. Time is in seconds in the upper right-hand corner. The particle has entered a dorsolateral part of the food groove at 0 s, moves along the food groove toward the mouth at 0.2 and 0.4 s, and enters the side of the mouth at 0.55 s. Rotation of the larva moves the mouth from upper right at 0 s to center at 0.55 s. The anterior end of the larva is toward the upper left. At 0 s the larva is 175-jum wide al the base of the prototrochal cilia.
Capture h\ opposed ciliary hands
In larvae of Arniandia hrevis and Urechis caitpo. the movements of partieles and the directions of recovery strokes of cilia indicated that the effective strokes of the prototrochal cilia were from anterior to posterior, and those of the metatrochal cilia were from posterior to anterior. For larvae of each species, we observed captures of more than 50 particles of 5 and 12 /urn in diameter; particles that came within reach of the prototroch were transported into the food groove between the prototroch and metatroch and moved to the mouth via the food groove, presumably by the food- groove cilia (Figs. 3-5). Particles were captured between prototroch and metatroch on the lateral and dorsal surfaces of the larva. Particles in the food groove moved around to the mouth from both the left and the right sides and both with and against the direction of rotation of the larval body. These particle paths indicate an opposed-band feeding mechanism.
Capture of large panicles
Larvae of both species also captured particles at the mouth, without transport in the food groove. A late-stage larva of A nnandla brevis (with > 14 setigers) captured two large particles (50 jum in diameter) while videorecorded through a dissecting microscope (Fig. 6). When the swim- ming larva contacted a large particle in the vicinity of the mouth, the larva slowed and rotated so that the lower lip was aligned with the particle. The larva opened its mouth and ingested the particle, presumably using oral cilia or muscu- lature.
Swimming Urechis caitpo larvae used the mouth for direct capture of particles that passed over the episphere. Such captures occurred simultaneously with opposed-band particle captures (Fig. 5). Many of the particles caught directly by the mouth were too large to fit between opposed prototroch and metatroch, as illustrated by the gut contents
in Figure 7 and the particles being rejected in Figure 8. In some cases the mouth gaped to admit a large particle. The 9-day-old larva in Figure 4 opened its mouth to a gape of about 35 jum with a width of 95 ;j.m. The 17-day-old larva in Figure 8 opened its mouth to 70 to 95 /urn, and the mouth's width when closed was about 125 /u,m. Cilia on the mouth's lower lip (anterior to the shorter cilia of the neu- rotroch) appeared to aid the movement of large particles into the mouth. These cilia seemed to be continuous with the metatrochal band, which would account for the posterior-
0.00
•
0.13
Figure 5. Videorecorded capture ol l\u> 12-fxm spheres by a Urechis fiiiil>t> larva. Time is in seconds in the lower right-hand corner. The particle marked by an adjacent black bar has entered a dorsolateral part of the food groove at 0 s. moves along the food groove toward the mouth al 0.04 and 0.13 s, and is near the side of the mouth at 0.30 s. The second particle passes over the prototroch directly into Ihe mouth. It is near the protolro- chal cilia at 0 s, passes over the anterior edge of the mouth at 0.04 and 0. 1 3 s. and has entered the mouth at 0.30 s. The mouth is at the lower left; the anterior end toward the upper left. At 0 s the larva is 170-jLim wide at the base of the prototrochal cilia.
20
B. G. MINER ET AL
0.00
Figure 6. Videorccoided capture of a 50-/nm sphere h\ a tree swimming Anuttndiu hrevis lar\'a under a dissecting microscope. Time in seconds is in the upper left-hand comer. All images are at the same magnifi- cation A black line indicates the particle. The larva approaches the particle and then orients its mouth towards the particle, which is on the bottom of the dish. The particle is captured at the larva's mouth, presumably moved by the large oral compound cilia, and swallowed. At 0 s the larva is 85-/am wide at the center of the body.
to-anterior current past these cilia. In sonic cases a panicle was brought into the mouth over the lower lip (Fig. 7).
Larvae of U. cuiipo captured large particles from an early stage. Small 4-day-old larvae ingested Sephadex spheres almost as large as those ingested by 16-day-old larvae (Table I). Even a 3-day-old larva ingested a 42-by-35-|u,m mineral grain. Larger larvae did capture larger spheres, however. When early and later stage larvae were fed the sai suspension, as in the last two lines of Table I, the median sizes and the largest si/.es of ingested spheres were significantly greater for larger, older larvae (Mann-Whitney U tests, H, 1 0. a 2 = 5, P < 0.05). Objects larger than the spheres olio can he ingested. For example, a 49-day-old larva, 375 /u,m id :, ingested an unidentified object 366 /xm long by 40 |um wide
When larvae of U. c<iii/>n of different ages and sizes were offered smaller plastic sphctes, all 10 of the small, 3-day-
old larvae caught fewer spheres of 29-fj,m than of 1 2-jum, and all 4 of the larger, 48-day-old larvae ingested more of the 29-/j,m spheres than of the 12-/xm spheres (Table II). Small, early-stage larvae did ingest 5- and 20- /xm spheres in about the same ratio as ingested by larger larvae (Table II). Estimates of the width of the food groove of a single 5-day-old larva ranged from 22 to 34 /xm, but the width of the food groove varies with contraction of the larva. The upper limit on the sizes of particles that could be transported in the food groove was not determined.
Rejection of particles
Larvae could actively reject particles. Particle rejection often occurred after a particle had been transported to the mouth and entered the esophagus. When a larva of Ariiuin- dia brevis expelled a particle, the metatrochal cilia around
ANNELID LARVAL FEEDING MECHANISMS
21
0.15
Figure 7. Videorecorded capture of a 40-fxm sphere by a Urechis caupo larva. Time is in seconds in the lower left-hand comer. The sphere is near the metatrochal cilia at the posterior lip of the mouth at 0 s and moves over this band of cilia toward the mouth at 0.1 and 0.15 s. It is just entering the mouth at 0.25 s. The anterior end is toward the upper right. At 0 s the larva is 300-^im wide at the base of the prototrochal cilia.
the mouth stopped beating as the particle moved posteriorly down the body (Fig. 3). Metatrochal cilia at the mouth of larvae of Urechis caupo must also have altered beat during particle rejection, because large particles moved posteriorly over the lower lip and down the neurotroch during rejection (Fig. 8). in contrast to their posterior-to-anterior path over the lip during ingestion (Fig. 7).
For larvae ofArmandia h rev is. prototroch circumference and prototrochal eilium length increased with number of setigerous segments (Fig. 9A. B). Larval volume increased exponentially with number of setigers (Fig. 9C).
Particle velocities increased slightly with number of se- tigers for larvae of A. hrevis with 6-7. 11-12. and 15-16 setigers (H = 9) (Fig. 10). Increased particle velocities and eilium lengths resulted in a 30% increase in the area of water per prototrochal slice moved per second between larvae with 6-7 and 11-12 setigers and a 22% increase between larvae with 11-12 and 15-16 setigers (Table III). Maximum particle velocities were within the distal third of the eilium length (estimated for each size class from Fig. 9B). consistent with our expectations (Emlet and Strath- mann. 1994). Although Strathmann et al. ( 1993) suggested that eilium lengths might be underestimated from videore- cordings, our results indicate that this was not the case. In addition, our measurements agree with the eilium length of approximately 35 /im reported by Hermans (1964) for a larva with an unspecified number of setigers.
Although estimated maximum clearance rates increased with number of setigers, they did not increase proportion- ately to body volume (Table III). Late-stage larvae (15-16 setigers) had a maximum ratio of clearance to body volume that was less than half of that achieved earlier in develop- ment (6-7 setigers; Table III).
Prototrochal circumference and eilium length both in- creased with larval growth to a greater extent for larvae of U. caupo than for larvae of A. brevis. over the stages measured (Tables I-III). The relative increase in body length was much less for U. caupo. Early-stage larvae were nearly spherical and elongated to the shape shown in Figure 2A at later stages. Data for particle velocities are lacking for
Figure 8. Videorecorded rejection of previously ingested spheres up to 50 jum in diameter by a Urcchix ciiii/x) larva. Time is in seconds in the lower left-hand corner. At 0 and 0.3 s the mouth gapes at least l(IO-(nm wide, and the clump of spheres moves over the posterior lip of the mouth and down the midventral neurotroch. The larva in the last frame is 295-/nm wide at the base of the prototrochal cilia, and the mouth, now rotated toward the viewer, is closed and approximately 1 20-ju.m wide.
B. G. MINER ET AL
Table I
Sizes of Sephadex spheres ingested by larvae «/ Urechis caupo differing in size and age
|
Particle |
diameter (/Limit |
||||
|
Age |
Prototrochal diameter |
Cilium length |
Number |
||
|
(days) |
(fan)* |
(/Mm) |
In suspension |
Ingested |
of larvae |
|
4 |
159 |
45 |
45,26-73(50) |
36, 14-53(51) |
12 |
|
5 |
165 |
44 |
44. 30-74 (50) |
36, 19-60(34) |
10 |
|
16 |
318 |
65 |
44, 30-74 (50) |
38,21-73(104) |
5 |
* Diameter of the prototrochal band is diameter at the base of the prototrochal cilia. t Values are median, range, and (in parentheses) number of particles.
U. caupo, but the increase in prototrochal area (cilium length times prototrochal circumference) relative to body volume was greater for this species than for A. brevis.
Discussion
Our observations add the Opheliidae and Echiuridae to those annelid families known to possess larvae with op- posed-band feeding. As in other opposed-band feeders, lar- vae of both Armandia brevis and Urechis caupo possess a ciliated food groove between two parallel ciliary bands, a postoral metatroch and a prototroch. Direct observations confirm that particles are captured in the food groove (Figs. 3-5), probably through the combined action of long com- pound cilia in the prototroch (which beat anterior to poste- rior) and shorter compound cilia in the metatroch (which beat posterior to anterior). Simple cilia of the food groove may aid in retention of particles as well as in transport. This system is very effective in capturing relatively small parti- cles (5-12 /urn), regardless of which part of the prototrochal circumference is contacted (ventral, lateral, or dorsal). How common this feeding method is in larvae of other opheliids or echiurids is not known, but larvae of at least one other echiurid bear opposed bands of cilia (Salensky, 1876; Hatschek, 1880).
Larvae of both A. brevis and U. caupo also ingested particles larger than the space between prototrochal and
metatrochal bands. For A. brevis, it was later stage (14-17 setiger) larvae that ingested large (50-ju.m) particles. These larvae approached large particles so that contact was di- rectly at the mouth. This behavior was not observed in larvae at earlier stages. In contrast, larvae of U. caupo ingested particles greater than 50 /urn at early stages. Larvae of U. caupo did not appear to change orientation as they approached large particles; however, their movements were constrained by mesh cages. Particles that were captured directly at the mouth entered either over the episphere and prototroch or over the extension of the metatroch on the lower lip. In both species the mouths were large, could be opened to a wide gape, and were heavily ciliated. The cilia bordering the lower lip of the mouth appear to be a contin- uation of the metatroch. The oral cilia of A. brevis may include additional compound cilia (Fig. 1). For both A. brevis and U. caupo, the large ciliated oral field and the large mouth aid in the capture of large particles.
The combination of two ciliary feeding mechanisms in individual larvae suggests hypotheses for evolutionary tran- sitions among the feeding larvae of annelids. Some larvae, such as those of serpulids, appear to be restricted to captur- ing small particles between opposed bands; other larvae, like those of polynoids, lack opposed bands and appear to capture mostly large particles one by one, using complex oral ciliature (Phillips and Fernet, 1996). Our results dem-
Tahle II
Sizes of plastic spheres iiixcMctl h\ lamie oj Urechis caupo differing in size and age
Particle diameter
|
Age (days) |
Prototrochal diameter (fun)* |
Cilium length (fj.ni) |
Ratio in suspension (29:12 /Mm) |
Ratio ingested (29:12 jum) |
Number of larvae |
|
3 |
151 |
46 |
1.43:1 |
39/146 = 0.27 |
10 |
|
4S |
347 |
76 |
1.43:1 |
206/112 = 1.84 |
4 |
|
(20:5 /im) |
(20:5 /urn) |
||||
|
4 |
161 |
45 |
11 |
146/30 = 4.9 |
8 |
|
15 |
310 |
67 |
1:1 |
99/37 = 2.7 |
8 |
Diameter of the prototrochal band is diameter at the base of the prolotrochal cilia.
ANNELID LARVAL FEEDING MECHANISMS
23
500
4 6 8 10 12 14 16 18
40-i
U. 35-
OfJ
u
I
30-
25H
15
R2 = 0.84
0 2 4 6 8 10 12 14 16 18
U
T
0 2 4 6 8 10 12 14 16 18
# of Setigers
Figure 9. Binomial regression of various larval parameters vs. number of setigers for Armandia brevis. For all equations. X = number of setigers. The R2 value is reported in the lower right-hand corner of each plot. (A) Inner prototroch perimeter (;i = 36 larvae); larval circumference = 178.36 + 23.24x - 0.44x2. (B) Cilium length (n = 22 larvae); cilium length = 18.67 + 2.16x - 0.07x:. (C) Larval volume (n = 36 larvae); larval volume = 105 48 + 008x.
onstrate that in at least two families of annelids, both types of mechanisms can be employed simultaneously by the same larva. In addition, it appears that the oral ciliature of A. brevis and U. caupo, which is responsible for the capture of large particles, is continuous with the lateral and dorsal
extensions of the metatroch and food groove. As an evolu- tionary transition, expansion of oral filiation might result in a food groove and metatroch paralleling the whole length of the prototroch to produce an opposed-band system. Alter- natively, enlargement of the mouth and elaboration of oral ciliation (with loss of the lateral and dorsal parts of the opposed-band system) could produce the variety of oral
4000-1
3000 -
2000 -
1000-
o
u
C/3
"e
^i
IT
'o
_0 U
>
— o
4000-1
3000 -
2000 -
1000-
4000n
3000 -
2000 -
1000-
A
— i 1 1 1 1 —
o 10 20 30 40 50 60 70
0 10 20 30 40 50 60 70
C
0 10 20 30 40 50 60 70
Distance To Cilium Base (urn)
Figure 10. Particle velocity vs. distance of particle from the base of the prototroch for Armandia brevis larvae with (A) 6-7, (B) 11-12, and (C) 15-16 setigers. The vertical dotted line shows the estimated cilium length taken from the binomial regression of Figure 9B.
24 B. G. MINER ET AL.
Table III
Estimated clearance rate and clearuin c rule per lamil volume fur three .w.-r ( /</vv •>, <>/ lumie o) Armandia hrcvis
|
Cilium |
Water area per |
Midpoint |
Larval |
Clearance |
||
|
#of |
lencth |
prototroch slice |
prototrochal |
Max. clearance rate |
volume |
rate/volume |
|
Seligers |
( nm I |
per unit timet (junr/s) |
circumference (jumi |
IjunvVs)- 10" |
(ju,m')± |
(1/S) |
|
6-7 |
29.9 |
32846 |
422 |
13.9 |
998309 |
13.9 |
|
11-12 |
34.8 |
42602 |
549 |
23.4 |
2526475 |
9.3 |
|
15-16 |
36.4 |
51449 |
642 |
33.4 |
5310250 |
6.3 |
* Calculated from the binomial regression in Fig. 10B.
t Calculated from the areas under the curves in Fig. 1 1. bound by the origin and the estimated cilium length lor that size class.
i Estimated from the binomial regression in Fig. IOC.
ciliature found in the diverse feeding larvae of annelids. Continued modification of such cilia might result in such unusual and functionally important structures as the group of long compound cilia on the left side of the mouth of polynoid larvae.
Estimated maximum clearance rates did not scale isomet- rically with body volume among the three size classes of A. brevis. Cilium length, prototroch circumference, and parti- cle velocities through a prototrochal slice all increased as body volume increased, but not enough for maximum clear- ance rate to increase in proportion to body volume — thus the volume of water swept by cilia decreases relative to body volume as the larva adds segments. An analogous situation has been described for the cyphonautes larva of bryozoans. in which ciliated band length does not increase proportionately to body volume during growth and devel- opment (McEdward and Strathmann. 1987). This allometry is potentially unfavorable to larger larvae. In asteroid, echi- noid, and bivalve larvae similar in size to A. brevis larvae, metabolic rates scale isometrically with body mass (Hoegh- Guldberg and Manahan. 1995). Further, in the larvae of an echinoid, metabolic demand scales isometrically with larval volume (McEdward, 1984). If these results can be general- ized to larvae of A. brevis, and if we make the reasonable assumption that the masses of these larvae are proportional to their volume, then the maximum clearance rates of A. brevis larvae decline relative to metabolic demand as the larvae increase in size. However, larger larvae of A. hrcvis (>12 setigers) can supplement the amount of small particles captured by opposed-band feeding by capturing larger par- ticles at the mouth. The increased size range of food may compensate, at least partly, for the decrease in clearance rate. This decrease in maximum clearance rate per larval volume may have selected for larvae that possess two types of feeding mechanisms.
Do other annelid larvae share this potentially unfavor- able allometry of maximum clearance rate and body volume? Some annelid larvae resemble A. hrevis in ex- treme elongation of a segmented body during the larval
stage (Bhaud and Cazaux, 1987). Some of these larvae (<'.,!,'.. spionids) possess feeding mechanisms other than the opposed prototrochal and metatrochal bands. Thus, evolutionary changes in the size range of particles cap- tured may have been favored in several groups of anne- lids as a result of a small head circumference and long larval body. Other possible solutions to this problem are opposed bands elongated on ciliated lobes, as reported for the rostraria larva of an annelid (Jagersten, 1972). or the sinuous opposed bands of mitraria larvae of oweniid annelids (Emlet and Strathmann, 1994).
The larvae of U. cinipo and some other annelids probably do not face such an unfavorable allometry of maximum clearance rate to body volume, however. The larvae of U. cuit/w develop from nearly spherical trochophores (at 3 to 5 days) to forms with more elongate bodies (at several weeks), but the elongation is not as extreme (cf. Fig. 2 to Fig. 6). Also, these larvae capture relatively large particles from an early stage. Nevertheless, the circumferential cili- ary bands are shorter, relative to body size, than similar bands that are extended on the velar lobes of many gastro- pod larvae (Richter and Thorson. 1975). Feeding on an extended size range of particles and extension of opposed, ciliary bands on lobes may be alternative ways of increasing ingestion rates.
Further analyses of larval feeding methods, as well as robust phylogenies, are required to understand the evolution and functional consequences of diverse larval feeding mechanisms in the Annelida. For example, why are opposed bands apparently used only in the capture of small particles? What functional constraints place an upper limit on the spacing of the prototroch and metatroch in opposed-band feeders? Such analyses may also reveal why some larvae (c.i>.. serpulids) use restricted opposed bands to feed on small particles, and others («'.#., polynoids) use complex oral ciliature to feed primarily on large particles instead of employing both methods, as do the opheliid and echiurid larvae described here.
ANNELID LARVAL FEEDING MECHANISMS
25
Acknowledgments
NSF grant OCE9633193. the Robert Fernald Fellowship endowment, and the Friday Harbor Laboratories of the University of Washington supported the research on Annan- din hrevis. NSF grant OCE9301665 and the Bodega Marine Laboratory of the University of California at Davis sup- ported the research on Urechis caupo. K. Uhlinger advised on collection of adults and culture of larvae of U. caupo. W. Borgeson provided algal medium and Isochrysis galbana. N. E. Phillips and C. Staude advised on analysis of video- tapes of U. caupo. We thank J. Marcus for help in printing photographs, and J. Hoffman and two anonymous reviewers for useful comments on the manuscript.
Literature Cited
Bhuud, M., and C. Cazaux. 1987. Description and identification of polychaete larvae; their implications in current biological problems. Oceania 13: 595-753.
Emlet, R. B. 1990. Flow fields around ciliated larvae: the effects of natural and artificial tethers. Mar. Ecol. Prog. Ser. 63: 211-225.
Emlet, R. B., and R. R. Strathmann. 1994. Functional consequences of simple cilia in the mitrariu of owcniids (an anomalous larva of an anomalous polychaete) and comparisons with other larvae. Pp. 143- 157 in Reproduction and Development oj Marine Invertebrates, W. H. Wilson, Jr., S. A. Strieker, and G. L. Shinn, eds. Johns Hopkins University Press, Baltimore.
Gould, M. 1967. Echiund worms: Urechis. Pp. 163-171 in Methods in Developmental Biology. F. H. Wilt and N. Wessels. eds. Crowell, New York.
Hatschek. B. 1880. Oner die Entwicklungsgeschichte von Echiurus und die systematische Stellung der Echiuridae. Arbeiten Zool. lust. Wien 3: 45-78, plates 1-6.
Hermans, C. O. 1964. The Reproductive and Developmental Biology of lite Opheliid Polychaete, Armandia brevis. M. S. Thesis, University of Washington. 131 pp
Hermans, C. O. 1978. Metamorphosis in the opheliid polychaete Ar- mandia brevis. Pp. 1 13-126 in Settlement and Metamorphosis of Ma- rine Invertebrate Larvae. F.-S. Chia and M. E. R. Rice. eds. Elsevier, New York.
Hoegh-Guldberg, O., and D. T. Manahan. 1995. Coulometric mea- surement of oxygen consumption during development of marine inver- tebrate embryos and larvae. / E\p. Biol. 198: 19-30.
Jagersten, G. 1972. Evolution of the Metazoan Life Cycle. Academic Press. New York. 282 pp.
McEdward, L. R. 1984. Morphometric and metabolic analyses of the
growth and form of an echmopluteus. J. Exp. Mar. Biol. Ecol. 82: 259-287.
McEdward, L. R., and R. R. Strathmann. 1987. The body plan of the cyphonautes larva of bryo/.oans prevents high clearance rates: compar- isons with the pluteus and a growth model. Biol. Bull. 172: 30-45.
McHugh. I). 1997. Molecular evidence that echiurans and pogono- phorans are derived annelids. Proc. Nail. Acud. Sci. USA 94: 8006- 8009.
McHugh, I)., and G. VV. Rouse. 1998. Life history evolution of marine invertebrates: new views from phylogenetic systematics. Trends Ecol. Evol. 13: 1X2-1X0.
Newby, W. W. 1940. The Embryology of the Ecliiuroid Worm Urechis caupo. American Philosophical Society. Philadelphia. 219 pp.
Nielsen. C. 1995. Animal Evolution. Oxford University Press, Oxford. 467 pp.
Nielsen, C. 1998. Origin and evolution of animal life cycles. Biol. Rev. 73: 125-155.
Phillips. N. E., and B. Fernet. 1996. Capture of large particles by suspension-feeding scaleworm larvae (Polychaeta: Polynoidae). Biol. Bull. 191: 199-208.
Richter, G., and G. Thorson. 1975. Pelagische Prosobranchier-Larven des Golfes von Neapel. Ophelia 13: 109-185.
Rouse, G. W., and K. Fauchald. 1997. Cladistics and polychaetes. Zool. Scr. 26: 139-204.
Salensky, W. 1876. Uber die Metamorphose des Echiurus. Morpholo- gisches Jahrhiich 2: 319-327.
Sleigh, M. A. 1984. The integrated activity of cilia: function and coor- dination. J. Proto-ool. 31: 16-21.
Strathmann, R. R. 1987. Larval feeding. Pp. 465-550 in Reproduction of Marine Invertebrates. Vol. 9. General Aspects: Seeking Uniiv in Diversity, A. C. Giese, J. S. Pearse, and V. B. Pearse, eds. Blackwell, Palo Alto, CA.
Strathmann, R. R. 1993. Hypotheses on the origins of marine larvae. Ainni. Rev. Ecol. Syst. 24: 89-1 17.
Strathmann, R. R., and D. J. Eernisse. 1994. What molecular phylog- enies (ell us about the evolution of larval forms. Am. Zool. 34: 502- 512.
Strathmann. R. R., and E. Leise. 1979. On feeding mechanisms and clearance rates of molluscan veligers. Biol. Bull. 157: 524-535.
Strathmann, R. R., T. L. Jahn, and J. R. C. Fonseca. 1972. Suspen- sion feeding by marine invertebrate larvae: clearance of particles by ciliated bands of a rotifer, pluteus. and trochophore. Biol. Bull. 142: 505-519.
Strathmann. R. R., L. Fenaux, A. T. Scwell, and M. F. Strathmann. 1993. Abundance of food affects relative size of larval and postlarval structures of a molluscan veliger. Biol. Bull. 185: 232-239.
Suer, A. 1,. 1982. Larval Settlement, Growth, and Reproduction of the Marine Ecluuran Urechis caupo. Ph.D. Thesis. University of Califor- nia. Davis. 198 pp.
Reference: Biul. Bull. 197: 26-39. (August I9Q9|
Bioluminescence in the Deep-Sea Cirrate Octopod Stauroteuthis syrtensis Verrill (Mollusca: Cephalopoda)
SONKE JOHNSEN1, ELIZABETH J. BALSER2, ERIN C. FISHER1, AND EDITH A. WIDDER1
' Marine Science Division, Harhor Branch Oceanographic Institution. Ft. Pierce. Florida: and 2 Department of Biology, Illinois Wesleyan University, Bloomington, Illinois
Abstract. The emission of blue-green bioluminescence (Am.ix = 470 nm) was observed from sucker-like structures arranged along the length of the arms of the citrate octopod Stauroteuthis syrtensis. Individual photophores either glowed dimly and continuously or flashed on and off more brightly with a period of 1-2 seconds. Examination of the anatomy and ultrastructure of the photophores confirmed that they are modified suckers. During handling, the photo- phores were unable to attach to surfaces, suggesting that, unlike typical octopod suckers, they have no adhesive func- tion. The oral position of the photophores and the wave- length of peak emission, coupled with the animals' primary postures, suggests that bioluminescence in S. syrtensis may function as a light lure to attract prey.
Introduction
Bioluminescence is a common and complex characteris- tic in coleoid cephalopods. A large percentage of these animals are bioluminescent, many possessing complicated light organs utilizing lenses, reflectors, irises, interference filters, pigment screens, and shutters (Harvey, 1952: Her- ring, 1988). The diversity of the morphologies and anatom- ical distributions of cephalopod photophores is unparalleled among invertebrate phyla (Voss, 1967; Herring, 1988). However, despite this extraordinary radiation, biolumines- cence appears to be rare among octopods. Although 63 of the 100 genera of squid and cuttlefish have bioluminescent species, only 2 of the 43 genera of octopods have species confirmed to be bioluminescent —the bolitaenids Japetella and Eledonella (Robison and Youn«. 1981; Herring el <//.,
Received 16 March 1999; accepted 27 May !')'»>
Address tor correspondence: Dr. Sonke Johnsen, MS #33, Woods Hole Oceanographic Institution, Woods Hole, MA 02543-1049. E-mail: sjohnsen@whoi.edu
1987; Herring, 1988). In these genera, the light organs are found only in breeding females (Robison and Young, 1981 ) and are restricted to tissues associated with the oral ring and the base of the arms (Herring et ai. 1987). In the case of citrate octopods, bioluminescence has been suggested but never confirmed (Aldred et ai, 1982, 1984; Vecchione. 1987).
This study provides the first description of biolumines- cence in the cirrate octopod Stauroteuthis syrtensis. We also describe the anatomy and ultrastructure of the photophores in comparison with the morphology reported for cephalopod photophores (Herring et til.. 1987) and octopod suckers (Kier and Smith. 1990; Budelmann et ai, 1997). In addi- tion, we present a hypothesis to explain how the presence of light organs relates to the feeding behavior postulated for these animals. A preliminary account of this research has been presented by Johnsen et al. ( 1999).
Materials and Methods
Source and maintenance of animals
Three specimens of Stauroteuthis s\rtensis were obtained during a cruise of the R.V. Edwin Link to Oceanographer Canyon (on the southern rim of Georges Bank, USA) in August and September 1997. The animals were collected at depth using the research submersible Johnson-Sea-Link out- fitted with acrylic collection cylinders (11-liter volume) with hydraulically activated, sliding lids. The three speci- mens were caught during daylight at depths of 755 m (225 m from bottom), 734 m (246 m from bottom), and 919 m (165 m from bottom) (dive numbers 2925 and 2927) and maintained for up to 2 days at 8°C in water collected at depth.
26
BIOLUMINESCENCE IN A DEEP-SEA OCTOPOD
27
Video ami photography
Specimens were videotaped in two situations. First, the behavior of two animals was recorded from the submersible. Second, the captured animals were filmed aboard ship in the dark by using an intensified video camera (Inte vac's Nile- Mate 1305/1306 CCTV intensifier coupled to a Panasonic Charge Coupled Device). During shipboard filming, the animals were gently prodded to induce bioluminescence. Representative video frames were digitized (ITSCE capture board, Eyeview Software, Coreco Inc.). The animals were also placed in a plankton kreisel (Hamner. 1990) and pho- tographed with a Nikon SLR camera with Kodak Elite 100 color film. Data from a previously recorded i';i xitu video of a specimen of S. svrtensis from the slope waters near Cape Hatteras at 840 m (35 m from bottom; August 1996; R.V. Edwin Link; dive 2777) are also reported in this paper.
Spectrophotometry
Bioluminescent spectra were measured using an intensi- fied optical multichannel analyzer (OMA-detector model 1420. detector interface model 1461, EG&G Princeton Ap- plied Research) coupled to a 2-mm-diameter fiber optic cable. The detector was wavelength calibrated using a low- pressure mercury spectrum lamp (Model 6047. Oriel Inc.) and intensity calibrated using a NIST referenced low-inten- sity source (Model 310. Optronics Laboratories) intended for the calibration of detectors from 350 nm to 800 nm. For further details on the theory of operation and calibration of the OMA detector, see Widder et al. ( 1983). Three emission spectra were recorded from one animal, and an average spectrum was calculated.
Microscopy of photophores
The fixation, dehydration, and infiltration procedures were performed at room temperature aboard the R. V. Edwin Link. The animals were sacrificed by over-anesthesia with MS222 (Sigma Chemicals Inc.). One specimen was fixed in 10% formalin in seawater and dissected to confirm species identification. One arm of a different specimen was fixed in 2.5% glutaraldehyde in 0.2 M Millonig's buffer at pH 7.4, adjusted to an osmolarity of 1000 mOs with NaCl. After an initial 1-h fixation, several photophores were dis- sected from the arm and placed in fresh fixative for an additional 5.5 h. Postfixation of the dissected photophores in \c/c osmium tetroxide in Millonig's buffer for 70 min was followed by dehydration through a graded series of ethyl alcohols. Over a period of 6 days, the specimens were slowly infiltrated with propylene oxide and Polybed 812 (Polysciences) and then embedded in Polybed 812.
Semithin ( 1 /urn) and thin sections of embedded material were cut with a diamond knife (Diatome) and a Sorvall MT2 rotary ultramicrotome. Semithin sections were stained
with 2% toluidine blue in 1% sodium borate and photo- graphed with a Zeiss Photomicroscope II using Kodak Tmax 100 black-and-white film. Intact arms fixed in 10% formalin in seawater were photographed with a Tessovar photographic system. Thin sections for ultrastructural eval- uation were stained with aqueous 3% uranyl acetate and 0.3% lead citrate. Stained sections were viewed and photo- graphed with a Zeiss EM 9 transmission electron micro- scope.
For scanning electron microscopy, an arm with suckers was fixed in 2.5% gluteraldehyde (as described above). dehydrated with ethyl alcohol, infiltrated with hexa- methyldisali/ane (Pellco), and air-dried. Micrographs were obtained with a JEOL JSM 5800LV scanning electron mi- croscope using Kodak Polapan 400 film.
Results
General description of animal and distribution of photophores
Figures 1A and IB show the largest of the three captured specimens of S. syrtenxis. The appearance of the specimen is typical for the species (Vecchione and Young, 1997). The mantle length is about 9 cm (mantle lengths of other two specimens ~ 6 cm), suggesting that all three animals were immature (Collins, unpubl. data). The measurements are highly approximate because the mantle in the living animal is easily deformed. The primary webbing extends for about three-quarters of the length of the arms. The arms are oral to the primary web and attached to it by a secondary web. The photophores are arranged in a single row along the oral surface of each arm, situated between successive pairs of cirri (Fig. IB). Each arm supports about 40 photophores. The distance between photophores decreases from the base to the tip of the arm, with the greatest distance being 4 mm and the smallest less than O.I mm. The diameter and the degree of development of the photophores located at the tip are less than those located at the base of the arm. The fresh tissue of the entire animal had a gelatinous consistency typical of many deep-sea cephalopods (Voss, 1967). Al- though orange-red under the photo-floodlights, the color of the animal was closer to reddish-brown in daylight.
Bioluminescence
When mechanically stimulated, S. svrtensis emitted mod- erately bright, blue-green light (Amax = 470 nm) from the sucker-like photophores along the length of each arm (Fig. 2). With continuous stimulation, these photophores pro- duced light for up to 5 min, though the intensity of biolu- minescence decreased over time. Individual photophores either glowed dimly and continuously or blinked on and off brightly at 0.5 to I Hz. The blinking photophores cycled asynchronously, producing a twinkling effect. All suckers
28
S. JOHNSEN ET AL
Outer epithelium
Figure 1. Photographs under artificial light of the deep-sea finned octopod Stauroteuthis syrtenxis with the wehhed arms in swimming pos- ture (A) and spread (B) displaying the photophores/suckers (arrowheads) that appear as white spheres along the length of the inner surface of the arms. The posture shown in (B) may he one of extreme withdrawal intended to startle intruders with the sudden appearance of hioluminescent suckers, ar, arm; ey, eye; fi, fin; wb, webbing between arms. Scale bars = 4 cm.
(except possibly the very small ones at the tips of the arms) appeared capable of luminescence. No other portion of the body was observed to emit light.
Morphology of photophores
Each photophore is a raised papilla-like structure partially embedded in the connective tissue of the arm. The photo- phores are composed of three layers of cells: an outer epithelium modified to form a collar, infundibulum, and acetabulum: a capsule-like mass of muscle and neural tissue beneath the epithelium; and a thin layer separating the capsule from the dermis of the arm (Figs. 3, 4, 5). The collar epithelium is continuous with the epidermis and is folded inward, forming a rim around the central portion of the photophore (Figs. 3B, C; 4A). In both formalin- and glut- araldehyde-preserved specimens, the photophores appear to be either everted above (Fig. 3B) or retracted below (Fig. 3C) the outer edge of the collar.
The outer and inner folds of the collar epithelium are morphologically distinct and are different from the epider- mis covering the arm (Figs. 5, 6). The epidermis of the arm is squamous to cuboidal in character and consists of epithe- lial cells possessing scattered apical microvilli (Fig. 6A). The outer edge of the collar is composed of columnar cells with apical microvilli, numerous electron-lucent and elec- tron-dense vesicles, and large, apically placed, elongated nuclei (Fig. 6B). Like the epidermis, this region of the collar is not covered by a cuticle.
The inner edge of the collar is similar in cellular mor- phology to the outer collar epithelium except that the mi- crovilli are more densely arranged and are covered by a cuticle (Figs. 6C, D). In this region, the cuticle is composed of at least three layers: an outer lamina 0.3 yum thick with irregular projections; a second electron-dense lamina, also 0.3 /am thick: and an inner layer approximately 1 /urn thick consisting of amorphous material.
The epithelium and the overlying cuticle of the inner edge of the collar continue as the epithelium of a flat recessed region of the photophore corresponding to the infundibulum of typical octopod suckers (Figs. 3B; 4 A: 5 A, B). The outer edge of the infundibulum is ringed by hook-shaped den- ticles (Fig. 4B-D), which are elaborations of the cuticle (Fig. 6C, D). In addition to the presence of denticles, the cuticle covering the infundibular epithelium differs from that described for the inner part of the collar in that the outer layer contains more irregular projections and the innermost lamina is greatly expanded. The cuticle in this region is apparently secreted by the infundibulum and, as supported by Figure 6C and D, is periodically molted and replaced by a new, pre-formed cuticle. Subcuticular spaces were ob- served in association with what appear to be newly forming denticles.
Three cell types — gland cells, columnar epithelial cells, and multiciliated cells — were observed in the infundibular epithelium. Gland cells with narrow apical necks and a reduced number of apical microvilli are situated between columnar epithelial cells, which are characterized by a brush border of branched microvilli, rounded apical nuclei, apical endocytic vesicles, and mitochondria (Fig. 7 A). Both co- lumnar cells and gland cells have a tine granular cytoplasm replete with Golgi bodies and electron-dense and electron- lucent vesicles of varying sizes (Fig. 7B. C). Electron-dense granules, not bounded by a membrane, were observed be- tween microvilli. These presumably originate from the in- fundibular cells and are incorporated into the cuticle (Fig. 7C). Multiciliated columnar cells were infrequently ob- served as part of the infundibulum. Cilia were not found in epidermal or collar cells. The cilia of the infundibular cells have two nearly parallel striated rootlets and appear to have reduced axonemes that do not project above the level of the
BIOLUM1NESCENCE IN A DEEP-SEA OCTOPOD
29
Figure 2. Digitized frames from a video sequence of light emission (white spots) from photophores/suckers taken from video of an animal filmed in the dark using an intensified video camera (Inlevac's NiteMatc 1305/1306 CCTV Intensifier coupled to a Panasonic CCD). Two amis are shown. For scale, their closest approach is approximately 1 cm.
microvilli (Fig. 7C). All three cell types are interconnected by apical adherens and subapical septate junctions (Fig. 7D).
At the center of the light organ, the infundibular epithe- lium invaginates to form the acetabulum, which is seen externally as an opening, or pore (Figs. 3B, 4A). This central opening continues internally as a blind canal (Fig. 5C). The acetabular cells differ from those of the infundib- ulum primarily in the basal position of the nuclei, the highly interdigitated lateral membranes, and the diminution of the outer two layers of the cuticle (Figs. 5C: 8A. B).
The infundibulum and the acetabulum rest on a basal lamina beneath which is located an expanded layer of con- nective tissue with a maximum thickness of 1.5 jam (Figs. 5C; 8C, D). Fibers, presumably collagen, although confir- mation of this is not provided by the data, are arranged in alternating directions in multiple layers, giving the tissue a herringbone appearance. Occasional breaks, traversed by nerve axons, were observed in this otherwise continuous connective tissue sheath.
Muscle and neural tissue
Beneath the connective tissue underlying the epithelium of the infundibulum and acetabulum is a mass of tissue consisting of muscle and neural cells; this surrounds and encapsulates the outer epithelium (Figs. 5; 8A, E; 9). The myofilaments. which include thick filaments (25 and 50 nm in diameter) and thin filaments consistent with the size of myosin and actin, are oriented in three planes — circular.
radial, and longitudinal with respect to the axis of the photophore. Although all sections were taken in the longi- tudinal plane of the photophore. the precise plane of each section for these transmission electron micrographs was not known. Thus, the differentiation of the fibers seen in Figure 8C (shown in cross-section) and Figure 9A (horizontal fibers shown in longitudinal section) as circular or radial cannot be determined.
Intermingled with the muscle cells are nerve cells char- acterized by electron-dense granules O.I jum in diameter. Nerve axons are located throughout the capsule and espe- cially in the basal region closest to the dermis (Fig 9B). Although a direct connection was not documented, fluores- cent images of the photophores indicate that axons originat- ing from the large branchial nerve traverse the dermis and connect to the photophore.
The innermost layer of the photophores is an epithelium that separates the muscular capsule from the dermis of the arm. The cells of this layer have interdigitated lateral mem- branes and a cytoplasm that appears more granular than that of the outer epithelium. This layer is associated with extrin- sic (to the photophore) muscle cells (Fig. 9A) and a blood vessel located in the dermis (Fig 9B).
In situ behavior
Animal I (from Cape Hatteras) was first seen in a bell posture with its fins sculling (Fig. 10A). It then moved away from the submersible, using a slow medusoid locomotion. After one contraction/expansion cycle, the animal closed its
30
S. JOHNSEN ET AL
CO
ct
Figure 3. (A) Photograph of part of an arm of Slauroteiilhis syrtensis with the webbing removed. Photophores (arrowheads) are arranged in a single row along the length of the arm and are unequally spaced with decreasing distance between light organs at the proximal tip of the arm. The positions of the photophores alternate with the positions of the cirri (cit. Scale bar = 0.5 cm. (B) Light micrograph of a fluorescent image of a single formalin-fixed photophore in the extended position. Like octopus suckers, the photophore is elevated above the epidermis (ep). is surrounded by a collar of epidermal cells (co), and consists of an infundihulum (in) and central acetabular canal (ac). (C) Light micrograph of a retracted photophore that has been bisected longitudinally. Internally, a capsule-like mass of tissue (ca) underlies the epithelium of the infundibulum and acetabulum. ct, dermal connective tissue of the arm. Scale bar for B and C = 0. 1 mm.
web and assumed a highly distended balloon posture with motionless fins (Fig. 10B). After several minutes in this posture, the arms opened to a bell posture, and then closed to a considerably smaller balloon posture (Fig. IOC) re- ferred to as the "pumpkin posture" by Vecchione (pers. comm.). After 2 min, the fins began sculling and the animal made one more medusoid contraction and then again closed its web to the pumpkin posture with fins sculling. After a minute, the animal made about seven more medusoid con- tractions and then closed to the pumpkin posture with fins sculling and head down.
Animal 2 (from Oceanographer Canyon) was first seen with its arms spread in the horizontal plane with the mouth oriented upwards (Fig. 10D). It underwent one medusoid contraction and then inflated to a highly distended balloon posture with fins motionless and cirri extended and pressed against the primary web. After several minutes, the fins
began sculling and the animal simultaneously twisted its body and opened its arms (Fig. 10E).
Animal 3 (from Oceanographer Canyon) was first seen in a bell posture. Then, using slow medusoid locomotion, moved away from the submersible. During the escape, its fin sculled continuously and sometimes vigorously. During expansion of the primary web, the cirri could be seen and were extended perpendicular to the arms and pressed against the primary web.
Discussion
Morphology of photophores and homology with octopus suckers
Although the anatomical position and morphology of the light organs of S. svrtensis indicate their homology with octopod suckers, other aspects of their structure are consis-
BIOLUMINESCENCE IN A DEEP-SEA OCTOPOD
31
Figure 4. Scanning electron micrographs of photophores. (A) Externally, each photophore has three main recognizable parts: outer wall or collar (co). infundibulum (in), and acetabulum (arrow indicates the opening to the acetabiilar canal). Scale bar = 100 /Mm. (B) The junction between the infundibulum and the infolded collar is ringed by a row of denticles (arrowheads). Scale bar = 10 ^im. (C-D) These hook-like denticles (de). which are atypical of octopod suckers, appear to be elaborations of the cuticle covering the infundibulum and acetabulum. Scale bars for C and D = 1 jum. A and B adapted from Johnsen el ill ( 1999). with permission from Nature, copyright 1999 Macmillan Magazines Ltd.
tent with those reported for simple photophores in other cephalopods (Young and Arnold, 1982; Herring et al., 1987, 1994). Definitive structural characteristics of octopod suck- ers are given by Kier and Smith (1990) and Budelmann et al. (1997). Like the suckers of other citrate octopods, the photophores of S. syrtensis are arranged in a single row along the oral surface of the arm with the largest, most developed organs located at the base of the arm, nearest the mouth. Suckers and these photophores both consist of three layers of tissue: an outer epithelium, an intrinsic muscular layer, and an extrinsic layer associated with muscle cells. The outer epithelium is covered by a cuticle that, as in suckers, appears to be periodically molted. Moreover, the epidermis associated with the photophore is modified to form the columnar epithelial cells of the recessed infundib- ulum and the invaginated acetabulum. The arrangement of
myofilaments in the muscular capsule are consistent with the three-dimensional array of contractile fibers typical of suckers. Although this may be an artifact of fixation, the morphology and the arrangement of myofilaments would allow for the retraction and extension, as well as a change in the diameter, of the photophore and may be important in regulating the intensity of the emitted light.
Although denticles are not common in octopod suckers (Nixon and Dilly, 1977; Budelmann et al., 1997), hooks and denticles of various sizes are found in decapod cephalopods. The functional significance, especially with the apparent loss of an adhesive function for the suckers, of the denticles on the photophores of S. syrtensis is unknown. They may, however, be vestigial structures indicating an evolutionary connection to the decapods.
Although definitive morphological characteristics are
32
S. JOHNSEN ET AL
-
**?•>
Figure 5. Light micrographs of a series of semithin sections from the outer edge of the infundihulum (A), through the middle region of the inl'undihulum (B), lo the center of the acetabular canal (C). Each photo- phore consists of an outer epithelium that is recessed below the level of a supporting epidermal collar (co). This epithelium forms the infundihulum (inland the acetabulum (ac)and is covered by a cuticle (cu). A capsule-like mass of tissue (ca) is located below the outer epithelium and is separated from the connective tissue (ct) of the arm by a third layer of cells (tl). Arrowheads, denticles; arrow, putative reflector. Scale bars = 0.2 mm
lacking for photocytes in general (Herring, 1988). the epi- thelium of the acetabulum (and possibly of the infundihu- lum) is presumed to be the bioluminescent region of the
photophores in S. syrtensis. Characteristics that identify photocytes in the octopod Japetella diaphana ( Herring et ai 1987) and the squid Ahralia trigonura (Young and Arnold, 1982) and are also found in the photophores of S. syrtensis include the presence of an amorphous, finely granular cy- toplasmic ground substance containing numerous electron- dense vesicles, large basal nuclei, highly interdigitated lat- eral plasma membranes, ciliary rootlets, and abundant Golgi bodies. To some degree, this cellular morphology is found in the cells of both the infundihulum and the acetabulum. Since these ultrastructural traits are also typical of secretory epithelia. one hypothesis is that the infundibular epithelium secretes the cuticle, and the acetabular epithelium is in- volved in light production.
Reflectors in cephalopod photophores are typically com- posed of collagen fibers arranged in layers beneath the photocytes (Young and Arnold, 1982; Herring et <//., 1994). The layer of connective tissue separating the outer epithe- lium and the underlying capsule of muscle and nerve tissue in S. syrtensis conforms to this description. As in J. di- aphana (Herring et ui, 1987), the fibrous layers appear concentric with respect to the axes of the photophore and alternate in orientation. In a longitudinal tangential section, this arrangement would produce the herringbone effect seen in the putative reflector of S. syrtensis.
The functional significance of bioluminescence
The replacement of adhesive suckers with photophores might have occurred during colonization of the pelagic deep sea from the shallow-water benthos. However, the signifi- cance of this transition and the function of the light emis- sions are, at present, unknown. One function of light emis- sion common to almost all bioluminescent animals is defense (reviewed in Widder, 1999). The vast majority of bioluminescent animals emit light when disturbed, perhaps to startle predators or to attract larger animals that may prey upon the predator. Feeding experiments have shown that feeding rates of predators are diminished in the presence of bioluminescence (see Widder, 1999, for further details). Since S. syrtensis emitted light in response to physical disturbance, it is likely that defense is at least one function of its light emission.
The functions of bioluminescence in deep-sea animals are difficult to determine because these animals are seldom observed in their own environment, and then only under bright lights that both mask natural light emissions and affect behavior. Therefore, the following discussion is based on circumstantial rather than direct evidence and is neces- sarily speculative. On the basis of the anatomical distribu- tion of the photophores, the optical characteristics of the emitted light, and the behavior of the animal, we propose two more functions for bioluminescnce in S. syrtensis: sexual signaling and light luring. Visual communication is
BIOLUM1NESCENCE IN A DEEP-SEA OCTOPOO
33
Figure 6. Transmission electron micrographs of tangential longitudinal sections through the epidermis (A), outer rim of the collar (B), and the inner rim of the collar and outer edge of the infundibulum (D and C). These, as do all TEM images, represent sections taken in the longitudinal axis of the photophore. The epidermis is composed of flat cells with scattered apical microvilli (niv). The inner rim of the collar (co) and the infundibulum (in) are covered externally by a cuticle (cu). which at the edge of the infundibulum is modified to form denticles (del. Like the cuticle of suckers, the photophore cuticle is apparently shed and replaced by a new pre-formed cuticle (nd). ct. dermal connective tissue of the arm; ne. nucleolus; nu. nucleus. Scale bars for A, B. and C = 1 fum. Scale bar for C = 3 jum.
extremely common among cephalopods (Hanlon and Mes- senger. 1996), and the suckers of certain shallow-water octopods have been implicated in (non-biolurninescent) sex- ual signaling (Packard. 1961). Indeed, sexual signaling is the proposed function of bioluminescence in the other bioluminescent genera of octopods (Robison and Young. 1981). In these animals, light organs are found only in mature breeding females. Males, immature females, and brooding females all lack this organ. In addition, Robison and Young (1981) noted that the emitted light is distinctly green (rather than the far more common blue-green) and suggested that the bioluminescence may be a private line of communication.
However, although the octoradial pattern of twinkling
bioluminescence in 5. syrtensis makes a highly species- specific signal, several factors contradict its use as a sexual signal. First, the photophores appear to be found in mem- bers of the species that, based on mantle length, are imma- ture (Collins, unpubl. data). None of the three collected specimens were sexed by dissection. However. S. .syrtensis can be externally sexed by a sexual dimorphism in suckers 9-22. In mature males, suckers 9-22 are considerably en- larged; in immature males, the enlargement is less notice- able (Collins, unpubl. data). By this method, two of the thixv animals were sexed as female; the sex of the third animal is unknown. All, including the animal observed from the submersible at Cape Hatteras. had the same character- istically reflective suckers. Second, the light organs' wave-
34
S. JOHNSEN ET AL
mv
mi
Figure 7. Transmission electron micrograph'- showing the details of the epithelial cells of the infundihulum. This epithelium contains gland cells (gc) and columnar epithelial cells (A) and multiciliated columnar cells (C). (A. B) Numerous electron-dense lev) vesicles, vesicles that are less opaque (ve), and Golgi bodies (arrowheads) arc found in finely granular cytoplasm (cy) of all cell types. Scale bars = 1 /urn. (C) Both types of columnar epithelial cells are characterized by having mitochondria (mi) and a brush-border of branched microvilli (mv). A few multiciliated cells were identified by the presence ot ciliary rootlets (cr) and short apical axonemes. Scale bar = 2/xm. (D) Infundibular cells are interconnected by apicolateral adherens (ad) and subapical septate (spi junctions. Scale bar = ().? /urn. nu. nucleus.
length of peak emission closely approximates the wave- length of maximum light transmission in the ocean (475 nm) and the usual peak wavelengths of bioluminescence and visual sensitivity in deep-sea animals (Jerlov. 1976; Herring, 1983; Widder ct ui, 1983; Frank and Case, 1988; Partridge el <//.. 1992; Kirk. 1983). Therefore, though the bioluminescent signal would be visible at long distances to other member-* of S. xyrtcnxix, it would also be visible at long distances to potential predators. In addition, since the octoradial pattern would be distorted by light scattering
after a short distance underwater (Mertens. 1970), it would be difficult to distinguish from the many other biolumines- cent signals of similar wavelength.
A more attractive hypothesis is that bioluminescence in S. xyrtenxis functions to lure potential prey. Voss (1967) has previously suggested that the photophores in some cepha- lopods function as light lures, and this function in S. syr- tensis is supported by the following line of evidence. The ciirate genera Stauroteuthis, Cirrotctithis. and Cirrm/uiiinni feed on small planktonic crustaceans, and other researchers
BIOLUMINESCENCE IN A DEEP-SEA OCTOPOD
35
.
m v
f;
" •&". ••••.^:A.«i?VJ-& ; -
-.".. •• . •' •••>'
Figure 8. Transmission electron micrographs of the acetahular epithelium ( A-C). the putative reflectoi i l>i. and distally positioned cells in the capsule of tissue beneath the outer epithelium (E). (A) A montage showing the presumptive photocytic epithelium (ph), reflector (re), and the underlying capsule (ca) of muscle and nerve cells. Scale bar = 3 /xm. (B) Like photocytes of other cephalopods, the cells of the acetabulum have highly digitized lateral membranes (arrowheads) and a finely granular cytoplasm (cy). Scale bar = 0.5 fim. (C. D) A layer of connective tissue separates the acetabular epithelium from the underlying muscle (mu) and nerve cells. Breaks in the connective tissue layer are bridged by neurons (nv). Fibers (fi) in the connective tissue are arranged in layers with alternating orientation, giving the tissue a herringbone appearance. Scale bars = 0.5 /^m. (E) Presumptive neurons with electron-dense granules (gr) are found throughout the capsule. Scale bar = 0.5 fxm.
have suggested that these are trapped within a mucous web produced by buccal secretory glands and handled by the elongated cirri (Vecchione, 1987; Vecchione and Young,
1997). Since all three genera appear to have nonfunctional suckers (Aldred ct «/.. 1983; Voss and Pearcy, 1990), this method of feeding seems likely. In all three genera, the
36
S. JOHNSEN ET AL
A
Figure 9. (A, B) Transmission electron micrographs ot Ihc medial region of the capsule and ot the third innermost cell layer of the photophore. The capsule, like the intrinsic musculature of suckers, consists, in part, ot cells with myolilaments arranged in three planes (also see inset). Longitudinal dm) and horizontal fibers (nui) alternate throughout the capsule. Inset shows thick and thin filaments consistent with the si/e and appearance of myosin and actin. Neural axons (nv) are intermingled with muscle cells. The capsule is separated from the connective tissue of the arm (ct)by a third layer of cells ( ' I. which is associated with extrinsic muscle cells (cm), bv. lumen ot a blood vessel; nu. nucleus. Scale bars = 1 fj.ni.
shape of the primary web precludes any flow-through feed- ing current. Therefore, the prey cannot be filtered from the water column, but must somehow be attracted to the mucous web. Since many deep-sea crustaceans have well-devel- oped, sensitive eyes and are attracted to light sources (Morin
ft nl.. 1475). the photophores of S. syrteiisis may provide the lure.
With the exception of the twisting behavior following ballooning (observed only once), the in situ behaviors of the three specimens of S. syrtensis reported here are similar to
BIOLUMINESCENCE IN A DEEP-SEA OCTOPOD
37
Figure 10. Digiti/ed frames of in situ video of Stauroteuthis syrtensis: (A) bell posture; (B) distended balloon posture; (C) pumpkin posture (distinguished from the balloon posture by the fact the web is not fully inflated); (D) inverted umbrella posture; (E) animal twisting and opening its arms after ballooning. The highlights on the mantle in A. C, and E are reflections from the submersible lights.
those described from videos of two other specimens ob- served by Vecchione and Young (1997). When first ap- proached. S. syrtensis is generally found with its arms spread in an umbrella or bell posture, with the mouth oriented either upwards or downwards (Roper and Brund- age, 1972; Vecchione and Young, 1997; Villeneuva et a!.. 1997). Since the animal almost certainly detects the rela- tively large and well-lit submersible well before itself being captured on video, it is difficult to know whether this is the natural posture or a defensive reaction. Given the assump- tion that the animals are in an undisturbed, non-withdrawal state when first filmed, their posture and the location of their photophores are consistent with the use of bioluminescence as a lure. As mentioned above, the wavelength of peak emission approximates the wavelength of maximum light transmission. This suggests that the emission spectrum of the photophores has been selected for maximum visibility.
Finally, because the intensity of upwelling light is only a small percentage of that of downwelling light (Demon, 1990), animals bioluminescing in the mouth-up posture would be highly visible to potential prey in shallower depths. Collectively, these observations give credence to the idea that 5. syrtensis uses photophores to attract prey.
Existence and evolution of photophores in octopods
Aside from the present study, the only other conclusive evidence of bioluminescence in octopods is restricted to the breeding females of the family Bolitaenidae (Herring. 1988). However, the existence of light organs within suck- ers (at the base of the peduncle) has been suggested in the cirrate octopod C. nuimivi (Chun. 1910, 1913). As in the photophores of 5. svrtensis, these organs have a bright white appearance due to reflection from a connective tissue layer and are found in suckers that have many reduced traits compared to typical octopod suckers (Aldred et ui, 1983). Unlike the photophores of S. syrtensis, the postulated light organs of C. mtirmyi are not found within the sucker itself. In addition, the connective tissue layer is situated such that the produced light would be reflected into the tissue of the arm. After a complex subsequent study (Aldred et ai, 1978, 1982, 1983), Aldred et nl. (1984) tentatively interpreted the organs as unusual nerve ganglia (see also Vecchione, 1987).
Because photophores and photocytes have a bewildering variety of morphologies (Buck, 1978; Herring, 1988), con- clusive determination of the presence or absence of light organs (which often emit only dim light) requires observa- tion of a healthy specimen in near-total darkness by a thoroughly dark-adapted observer (i.e., in near-total dark- ness for a minimum of 10 min) (Widder et ai, 1983). Owing to the bright lights of submersibles and remotely operated vehicles, bioluminescence is seldom observed in situ. In addition, since most bioluminescent animals produce light when disturbed, deep-sea cephalopods collected in nets generally have exhausted their light production by the time they reach the surface. Finally, observations of spontaneous luminescence are rare; most bioluminescent animals must be physically stimulated (often for a considerable period of time) before light is observed (Widder et «/., 1983).
The evolutionary history of photophores in any animal group is extremely difficult to determine because biolumi- nescence has no fossil record (Buck, 1978). The evolution of bioluminescence in the coleoid cephalopods is particu- larly intriguing because of the extraordinary diversity and complexity of photophores in deep-sea decapods and vampyromorphs and their apparent rarity and simplicity in deep-sea octopods (Herring, 1988). However, biolumines- cence in the deep-sea octopods may not be as rare as previously assumed. For the reasons given in the previous paragraph, bioluminescence may be under-reported in the deep-sea octopods. Cirrothuuimi innrniyi and Cirroteitthis
38
S. JOHNSEN ET AL
are found at abyssal depths (except in polar regions, where they can be found at the surface) (Voss, 1988), making capture of healthy specimens extremely difficult. Opistoteu- this is found at shallower depths and has been maintained in aquaria (Pereyra, 1965). but it is not known whether it was observed under the specialized conditions necessary to de- tect bioluminescence. However, given that Opistoteuthis feeds on a variety of benthic prey that it apparently captures using functional suckers (Villaneuva and Guerra, 1991 ), if it is bioluminescent, its photophores are probably in a differ- ent site.
Octopod bioluminescence may exist only in S. syrtensis and the bolitaenids. A cladistic analysis of the Octopoda involving 66 morphological characters places the citrates basal to the incirrates and the bolitaenids basal among the incirrates (Voight, 1997). This analysis also supports the monophyly of the bolitaenids and the two clades (Cirroteu- thidae and Stauroteuthidae) composing the genera Stauro- teitthis, Cirrothauma, and Cirroteuthis. The homology of the photophores in S. syrtensis and the bolitaenids is un- likely. The photophores of S. syrtensis appear to exhibit the rare trait of muscle derivation. The only other example of muscle-derived light organs has been found in the scopel- archid fish Benlhalbella infans (Johnston and Herring, 1985). Although the luminous circumoral ring in the boli- taenid Japetella diaphann is initially a muscular band, the great increase in the size of the ring in adult females apparently requires tie uoro synthesis of luminous tissue (Herring et ai, 1987). In addition, the light organs differ in almost all other anatomical and morphological characteris- tics. Finally, only mature female bolitaenids have light organs, which appear to have a sexual function, whereas the light organs of S. syrtensis are found in immature animals and may be involved in feeding. Multiple independent evo- lutions of photophores are common in decapods, at least at taxonomic ranks of subfamily or higher (Young and Ben- nett, 1988; Herring et ai, 1992. 1994). Therefore, despite the close evolutionary relationship between the cirrates and the bolitaenids. photophores in these two groups seem to have evolved independently. However, the monophyly of Stauroteuthis, Cirroteuthis, and Cirrothauma and the fact that they all have suckers with reduced traits suggest the possibility of light production by modified suckers in the latter two genera.
Acknowledgments
We thank the captain and crew of the R.V. Edwin Link and the Johnson-Sea-Link pilots, Phil Santos and Scott Olsen. for assistance with all aspects of animal collection. We also thank Dr. Tamara Frank for a critical reading of the manuscript. Dr. Michael Vecchione for aid with identifying the specimens, Dr. Janet Voight for helpful discussions on octopod evolution, and Dr. Martin Collins for use of un-
published data. The authors are grateful to the Smithsonian Marine Station at Fort Pierce, Florida, for allowing the use of the photomicroscopes. Our thanks are also extended to Dr. William Jaeckle for his assistance with the scanning electron microscopy and to Julie Piriano for help with the transmission electron microscope. This work was funded by a grant from the National Oceanic and Atmospheric Ad- ministration (subgrant UCAP-95-02b, University of Con- necticut, Award No. NA76RU0060) to Drs. Tamara Frank and EAW, a grant from the National Science Foundation (OCE-93 13872) to Drs. Tamara Frank and EAW. and by a Harbor Branch Institution Postdoctoral Fellowship to SJ. This is Harbor Branch Contribution No. 1287 and Contri- bution No. 479 of the Smithsonian Station at Fort Pierce, Florida.
Literature Cited
Aldred, R. G., M. Nixon, and J. Z. Young. 1978. The blind octopus
Cinotlnniimi. Nature 275: 547-549. Aldred, R. G., M. Nixon, and J. Z. Young. 1982. Possible light organs
in tinned octopods. J. Mo/luscan Stud. 48: 100-101. Aldred, R. G., M. Nixon, and J. Z. Young. 1983. Cirrothauma murrayi
Chun, a tinned octopod. Philos. Trans. R. Soc. Land. B 301: 1-54. Aldred, R. G., M. Nixon, and J. Z. Young. 1984. Ganglia not light
organs in the suckers of octopods. J. Molluscan Snul. 50: 67-69. Buck, J. 1978. Functions and evolutions of bioluminescence. Pp. 419-
460 in Bioluminescence in Action, P. J. Herring, ed. Academic Press,
London. Budelmann, B. U., R. Schipp, and S. Boletzky. 1997. Cephalopoda. Pp.
119-414 in Microscopic Anatomy of Invertebrates Vol. 6A, F. W.
Harrison and A. J. Kohn, eds. Wiley-Liss Publications, New York. Chun, C. 1910. Die Cephalopoden. Wiss. Ergebn. dl. Thiefsee-Exped.
"Valdn-ia" 18: 1-552. Chun. C. 1913. Cephalopoda. Pp. 1-21 in Report on the Scientific
Results of the "Michael Sars " North Atlantic Deep-Sea Expedition
1910 Vol. III. Part I. J. Murray and J. Hjort. eds. The Trustees of the
Bergen Museum. Bergen. Denton, E. J. 1990. Light and vision at depths greater than 200 meters.
Pp. 127-148 in Light ami Life in the Sea. P. J. Herring. A. K. Campbell.
M. Whittield. and L. Maddock, eds. Cambridge University Press, New
York. Frank, T. M., and J. F. Case. 1988. Visual spectral sensitivities of
bioluminescent deep-sea crustaceans. Bio/. Bull. 175: 261-273. Hamner, W. M. 1990. Design developments in the pkmktonkreisel, a
plankton aquarium for ships at sea. / Plankton Res. 12: 397-402. Hanlon, R. T., and ,1. B. Messenger. 1996. Cephalopod Behaviour.
Cambridge University Press, Cambridge.
Harvey, E. N. 1952. Bioluminescence. Academic Press, New York. Herring, P. J. 1983. The spectral characteristics of luminous marine
organisms. Pruc. R. Soc. Loud. B 220: 183-217. Herring, P. .1. 1988. Luminescent organs. Pp. 449-489 in The Molluxca
Vol. II. E. R Tru<;man and M. R. Clarke, eds. Academic Press. New
York. Herring, P. J., P. N. Dilly. and C. Cope. 1987. The morphology of the
bioluminescent tissue of the cephalopod Japetella diaphana (Cepha- lopoda: Bolitaemdael. / Zoo/. Lund. 212: 245-254. Herring, P. J., P. N. Dilly, and C. Cope. 1992. Different types of
photophore in the oceanic squids, Octopoteuthis and Tuningiti (Cepha- lopoda: Octopoteuthidae). J. Zoo/. Loud. 227: 479-491. Herring, P. J., P. N. Dilly, and C. Cope. 1994. The bioluminescent
BIOLUMINESCENCE IN A DEEP-SEA OCTOPOD
39
organs of the deep-sea cephalopod Vampyroteuthis infemalis (Cepha- lopoda: Vampyromorpha). J. Zool. Land. 233: 45-55. Jerlov, N. G. 1976. Marine Optics. Elsevier Scientific Publishing, New York. Johnsen, S., E. J. Balser, and E. A. Widder. 1999. Light-emitting
suckers in an octopus. Nature 398: 1 13-1 14. Johnston, I. A., and P. J. Herring. 1985. The transformation of muscle
into bioluminescent tissue in the fish Benlhalbella infans Zugmayer.
Proc. R. Soc. Land. B 225: 213-218. Kier, W. M., and A. M. Smith. 1990. The morphology and mechanics
of octopus suckers. Bio/. Bull. 178: 126-136. Kirk. J. T. O. 1983. Light and Photosynthesis in Aquatic Ecosystems.
Cambridge University Press, Cambridge. Mertens, L. E. 1970. In-water Photography: Theory and Practice.
Wiley Interscience, New York. Morin, J. G., A. Harrington, K. Ncalson, N. Krieger, T. O. Baldwin,
and J. W. Hastings. 1975. Light for all reasons: versatility in the
behavioral repertoire of the flashlight fish. Science 190: 74-76. Nixon, M., and P. N. Dilly. 1977. Sucker surfaces and prey capture.
Symp. Zool. Soc. Land. 38: 447-5 1 1 . Packard, A. 1961. Sucker display of Octopus vulgaris. Nature 190:
736-737. Partridge, J. C., S. N. Archer, and J. Van Oostrum. 1992. Single and
multiple visual pigments in deep-sea fishes. J. Mar. Bio/. Assoc. U.K.
72: 113-130. Pereyra, W. T. 1965. New records and observations on the flapjack
devilfish Opistoteuthis cali/orniana Berry. Pac. Sci. 19: 427-441. Robison, B. H., and R. E. Young. 1981. Bioluminescence in pelagic
octopods. Pac. Sci. 35: 39-44. Roper, C. F. E., and W. L. Brundage. 1972. Citrate octopods with
associated deep-sea organisms: new biological data based on deep benthic
photographs (Cephalopoda). Smithson. Cnntrib. Zool. 121: 1-45. Vecchione, M. 1987. A multispecies aggregation of curate octopods
trawled from north of the Bahamas. Bull. Mar. Sci. 40: 78-84.
Vecchione, M., and R. E. Young. 1997. Aspects of the functional morphology of cirrate octopods: locomotion and feeding. Vie Mi/ic'ii 47: 101 -1 10
Villaneuva, R.. and A. Guerra. 1991. Food and prey detection in two deep-sea cephalopods: Opistoteuthis aga.isizi and O. vossi. Bull. Mar. Sci. 49: 288-299.
Villaneuva, R., M. Segonzac, and A. Guerra. 1997. Locomotion modes of deep-sea cirrate octopods (Cephalopoda) based on observations from video recordings on the Mid-Atlantic Ridge. Mar. Bio/. 129: 1 13-122.
Voight, J. R. 1997. Cladistic analysis of the octopods based on anatom- ical characters. J. Molluscan Stud. 63: 311-325.
Voss, G. L. 1967. The biology and bathymetric distribution of deep-sea cephalopods. Stud. Trop. Oceanogr. 5: 51 1-535.
Voss, G. L. 1988. The hiogeography of the deep-sea octopoda. Malaco- logia 29: 295-307.
Voss, G. L., and W. G. Pearcy. 1990. Deep-water octopods (Mollusca: Cephalopoda) of the northeastern Pacific. Proc. Calif. Acad. Sci. 47: 47-94.
Widder, E. A. 1999. Bioluminescence. Pp. 555-581 in Adaptive Mech- anisms in the Ecology of Vision. S. N. Archer, M. B. A. Djamgoz. E. Loew, J. C. Partridge, and S. Vallerga, eds. Kluwer Academic Pub- lishers. Dordrecht, The Netherlands.
Widder, E. A., M. I. Latz, and J. F. Case. 1983. Marine biolumines- cence spectra measured with an optical multichannel detection system. B/o/. Bull. 165: 791-810.
Young, R. E., and J. M. Arnold. 1982. The functional morphology of a ventral photophore from the mesopelagic squid, Abralia trigonura. Ma/aco/ogia 23(1): 135-163.
Young, R. E., and T. M. Bennett. 1988. Photophore structure and evolution within the Enoploteuthidae (Cephalopoda). Pp. 241-251 in The Mollusca Vol. XII, E. R. Trueman and M. R. Clar! eds. Aca- demic Press, New York.
Reference: Biol. Bull 197: 40-48. (August 1999)
Long-Term Culture of Lobster Central Ganglia: Expression of Foreign Genes in Identified Neurons
GEOFFREY K. GANTER, RALF HEINRICH. RICHARD P. BUNGE1 *, AND EDWARD A. KRAVITZt
DC/HI rtinent of Neurohiologv, Harvard Medical School, 220 Longwood Avenue, Boston, Massachusetts 021 15: and Miami Project to Cure Paralysis, University of Miami School of Medicine,
Miami, Florida 33136
Abstract. The ventral nerve cords of lobsters (Homarus americamts) can be cultured /';; vitro for at least 7 weeks. Over this period, neurons maintain their normal electro- physiological features and continue, among other measures of neuronal health, to synthesize RNA and proteins. One application of this culture system is demonstrated: the ma- nipulation of gene expression in identified neurons. After intracellular injection of complementary RNA (cRNA) en- coding green fluorescent protein (GFP), the amount of pro- tein product measured by fluorescent confocal microscopy increases for 4 days and then decreases to background by day 10. Thus, translation of the injected message must have increased for 4 days before declining. Moreover, after in- jection of cRNA encoding j3-galactosidase, the levels of enzyme activity were measured using a fluorogenic sub- strate, revealing a peak of /3-galactosidase activity at 6 to 9 days: this activity was still detectable for at least 10 days after injection.
Therefore, either GFP or /3-galactosidase can be used as an injectable marker, allowing HI vivo quantitation of ex-
Received 16 December 1998; accepted 21 April 1999.
* This paper is dedicated to the memory of a good colleague and friend. Dr. Richard P. Bunge. Dick died on September 10. 1996, of esophageal cancer at the age of sixty-four. He was in the midst of an important project, that of rebuilding the Miami Project to Cure Paralysis, and in 1989 became the scientific director of that project. One of us (EAK) had the good fortune to work with Dick from 1968 to 1969 during his sabbatical visit to our laboratory in Boston. The organ culture system was developed at that time, and although these earlier experiments never were published, they are an important part of our present and future research activities. It is typical of Dick and his studies that they often were far ahead of their time. We are honored to include him as an author on this paper.
t To whom correspondence should be addressed. E-mail: edward_kravitz@ hms.harvard.edu
pression in individual cells over time. We measured long- lasting expression of these proteins after a single injection, suggesting that it may be possible to manipulate the levels of expression of any gene of interest.
Introduction
The ventral nerve cord preparation has been a useful tool for exploring the physiology and pharmacology of central neurons in the lobster (see Otsuka et ai. 1967; Livingstone ct <//., 1980; Ma and Weiger, 1993). The entire nerve cord, with the nerve roots sectioned, can be dissected from lob- sters and maintained in oxygenated saline for up to 36 hours. After the ganglia have been desheathed, the large cell bodies of central neurons can be identified both by their positions and by their physiological properties as revealed in electrophysiological recordings.
In this study, we report an extension of this technique that allows experiments of at least 7 weeks duration to be carried out. Longer term experiments that become possible under these conditions include an exploration of the effects of manipulation of levels of gene expression. Since activation of a gene can take several days to affect cells, experiments of this type were not practical with the short-term prepara- tions that are commonly used.
Genes from practically any source can be injected into Xenopits oocytes and the resulting heterologous expression analyzed (for review, see Dawid and Sargent, 1988). Use of this technique allows, for example, the effects of point mutations or deletions in genes to be examined. Although expression can be characterized in heterologous systems, the function of cloned genes might be more informatively explored if they could be expressed in cells of their native organism. Our studies show that this important option is
40
ORGAN CULTURE AND INJECTION OF NEURONS
41
now available for investigations of the roles of neuronal genes in the central neurons of the lobster. Moreover, we can use this technique to express proteins or peptides in neurons and investigate both the physiological responses of the neuron and the effectiveness of the introduced sub- stance.
Injection of DNA or RNA into identified neurons can be a direct and effective means of altering gene expression in identified cells. Several investigators have used microinjec- tion techniques to express genes in invertebrate neural sys- tems; in particular, the effects of various proteins on syn- aptic transmission and electrical excitability of neurons have been determined. For example, Kaang and coworkers (1992) found that microinjection of DNA encoding an A- type Shaker potassium channel into Aplysia neurons short- ened action potentials and thereby had a profound effect on transmitter release. Expression of a noninactivating potas- sium channel in another type of Aplysia neuron abolished spontaneous bursting activity (Zhao et til.. 1994). RNA encoding the leech homeobox protein Loxl, when injected into certain types of leech motoneurons, introduced active spike propagation to proximal neurites (Aisemberg et al., 1997). In another system. Dearborn and coworkers (1998) injected rat synapsin 1A1 into crayfish neurons and found an increase in transmitter release.
In this communication, we present an organ culture sys- tem that extends the time frame of nerve cord experimen- tation, allowing for such long-term experiments as the ex- pression of foreign RNAs in lobster central neurons. We have induced the expression of green fluorescent protein (GFP) from the jellyfish Aequoria victoria (Prasher et al., 1992; Chalfie et al., 1994) and the enzyme /3-galactosidase from the bacterium Escherichia coli (MacGregor et al., 1987). Levels of both proteins were measured repeatedly in individual living cells over several days. Neither of these proteins interfered with the viability of the cells. These reagents can be injected either as continuous expression monitors or as fusion proteins linked with other proteins of interest (Lalumiere and Richardson, 1995; Gerdes and Kaether. 1996). This study confirms the effectiveness of the gene delivery system and our long-term organ culture sys- tem, a combination that promotes connections between the disciplines of molecular biology and physiology in the study of the lobster nervous system.
Materials and Methods
Dissection and neuron identification
Nerve cords were removed from ice-anesthetized adult lobsters and pinned out in lobster saline as previously de- scribed (Otsuka et al., 1967; Harris-Warrick and Kravitz, 1984; Ma et al., 1992). After desheathing the abdominal ganglia and removing the glial layers, we targeted two cell types for injections. These were identified according to
criteria defined by Otsuka et al. (1967) and included large glutamate-containing motoneurons (M6/M7). with their ax- ons projecting through ipsilateral 3rd roots, and GABA- containing neurons (12), displaying prominent excitatory synaptic input and projecting through contralateral 3rd roots to the periphery.
Lobster organ culture system
After experimental manipulation, preparations were rinsed several times in sterile saline containing penicillin (50 /u,g/ml). streptomycin (50 /ig/ml) and neomycin (100 fig/ml) (GIBCO/BRL). Nerve cords were rinsed twice in a modified Leibovitz's L-15 medium, pinned out in ethanol- and UV-sterilized Sylgard-coated (Dow Corning) petri dishes (60 mm diameter), and then incubated in the modi- fied medium. The culture medium contained Leibovitz's L-15 medium with L-glutamine at 300 mg/1 (GIBCO/BRL) and with the following additions made to bring the salt concentration to levels suitable for incubation of lobster tissues: NaCl was adjusted to 462 mM; KC1 to 16 mM; CaCU to 26 mM; MgCl2 to 8 mM; and glucose to 1 1 mM; HEPES buffer (pH 7.4) at 10 mM (all salts from Sigma); fetal bovine serum (GIBCO/BRL) at a final concentration of 10%. Antibiotics and antifungal agents were added at the same concentrations specified above. Nerve cords were cultured at 16°C in an air incubator (Isotemp, Fisher) and the culture medium was changed twice weekly.
Lobster ganglion labeling and autoradiography
Labeling and autoradiography were carried out according to the method of Hendelman and Bunge (1969). Lobster ganglia were incubated in medium containing 10 jaM H- uridine (RNA labeling) or 10 /J.M 3H-leucine (protein label- ing) at high specific activity (New England Nuclear). Tis- sues were fixed in 2% osmium tetroxide buffered with veronal acetate (pH 7.4) with added CaCl: (0.05%) (all chemicals from Sigma). After several washes, ganglia were dehydrated in a graded ethanol series and embedded in a mixture of 10 parts Epon 812 (Structure Probe, Inc. (SPD); 10 parts Aralidite 6005 (SPD: 24 parts DOS A (dodecenyl succinic anhydride. SPI): and 2% DMP-30 (2,4,6-tris(dim- ethylaminomethyl (phenol, Sigma). Sections of 1-1.5 /am were cut and autoradiographed at 4°C in Ilford L-4 emul- sion (Ferranti-Dege). Slides were developed with D-19 (Kodak) and mounted in glycerin. Two experimental sets were examined in detail using this method. In one, pairs of ganglia were incubated in organ culture for 2 days or for 49 days, before incubation for 4 h in H-leucine or H-uridine, followed by 20 h of washout in nonradioactive medium, fixation and processing all as described above. In the second experimental set, ganglia were cultured for 30 and 45 days in vitro before labeling, fixation, and autoradiography.
42
G. K. GANTER ET AL
GABA measurements
GABA in single cells was measured by the original assay method of Jakoby and Scott ( 1959). This method uses the enzymes GABA-glutamic transaminase and succinic semi- aldehyde dehydrogenase to generate spectrophotometrically detectable triphosphopyridine nucleotide (reduced) in direct proportion to the amount of GABA in a sample. Single-cell- body isolation, extraction, and enzyme measurements were carried out as described in Otsuka ct ul. ( 1967).
RNA reagents
Capped GFP cRNA was transcribed from Xhol -linear- ized pSEM/GFP (generously provided by Richard Dear- born, see Dearborn ct ul.. 1998) (enzyme from New En- gland Biolabs) using the CapScribe SP6 kit (Boehringer Mannheim) and following the manufacturer's instructions. /3-galactosidase-encoding cRNA was transcribed from pCMV-SPORT-0-gal (GIBCO/BRL. linearized with Xmnl. New England Biolabs) by the same method. RNA quality and quantity were monitored by gel electrophoresis and UV spectrophotometry. The solutions used for injection con- tained cRNA at 0.5 to 1 MJ?V alH' included 0.1 U/ju,l of RNase inhibitor (RNasin. Promega).
RNA was introduced into the cytoplasm of lobster neu- rons by pressure injection through an intracellular recording electrode (pressure at compressor: 5-20 psi). Glass capil- laries (Drummond Scientific Company) were baked over- night at 300°C to facilitate filling and to remove possible sources of RNase activity; microelectrodes were pulled (Na- rishige PE-2) and the tips were broken to less than 4 /im by tapping them against a fine glass rod under a compound microscope at 200 x magnification. Electrodes were filled by first dipping their tips into 2 jul of the electrode solution (see below) on a small piece of Parafilm (American Na- tional Can). The remaining solution was picked up using a pipette tip and injected into the back of the electrode, where it quickly ran to the tip by capillarity. Electrodes were mounted in an electrode holder (MEH2RW, World Preci- sion Instruments (WPI)) fitted with a silver wire long enough to make electrical contact with the solution in the tip of the electrode. The electrode wires were treated with RNaseZap! solution (Ambion), with care taken in loading the RNA solution to avoid RNase contamination. Potassium acetate was added to the RNA samples to a final concen- tration of 0.5 M, along with 0.05% phenol red (Sigma) for visualization. Control injections revealed that phenol red did not interfere with the fluorescence detection or the viability of the cells, and appeared to diffuse out of injected cells within a few hours. Cells were injected slowly with RNA solution until the somata appeared slightly red.
DNA reagents
Plasmid DNA was injected in the same way as was the RNA. Purified DNAs including the following constructs were injected at concentrations ranging from 0.1 to 5 /J.g//-tl: (i) Drosophila melanogaster neuron-specific elur promoter (Yao and White, 1994) driving a /3-galactosidase reporter gene (gift from Dr. Thomas Schwarz): (ii) human CMV immediate early enhancer/promoter (Thomsen et ai. 1984) driving a fusion of the human j3-2 adrenergic receptor and the At'i/itoriti victoria GFP (gift from Dr. Timothy Mc- Clintock); and (iii) Drosophila melanogaster heat-shock inducible promoter from hsp70 (Pelham, 1982) driving GFP.
Physiological recordings
Intracellular recordings from neuronal somata were per- formed with glass microelectrodes filled with 1.0 M potas- sium acetate (12-25 MO resistance). For injections, this solution was replaced by a mixture of 0.5 M potassium acetate and the molecular constructs to be injected (3-5 Mil resistance). Electrical signals were amplified with an Axo- probe 1A amplifier (Axon Instruments). The cells were physiologically identified by their synaptic input and anti- dromic stimulation of their axons in the 3rd roots following criteria defined by Otsuka and coworkers (1967). Nerve roots were stimulated by placing their cut ends into closely fitting suction electrodes. Electrical stimuli were generated and delivered by a Master-8 stimulus generator (A. M.P.I.) with built-in isolation units.
Detection of cRNA expression
Injected neurons were periodically monitored, using con- focal fluorescence microscopy, for GFP expression or for /3-galactosidase activity. Nerve cords were transferred from culture vessels to sterile, medium-filled, deep-well micro- scope slides that had a thin coat of Sylgard on their lower surfaces. The cords were pinned out, coverslipped. and observed with a scanning laser confocal microscope equipped with FITC filters (MRC 600, Bio-Rad). For /3-ga- lactosidase activity determinations, fluorescein di-(/3-D-ga- lactopyranosideK Sigma) dissolved in dimethyl sulfoxide to make a final concentration of 0.67 mg/ml was added to the medium on ice after the preparation had been transferred to the slide.
For GFP measurements. 6.5-/xm optical sections were taken through the cell body region of the injected cell and images recorded using the confocal microscope's photomul- tiplier tubes at various times after injection. For /3-galacto- sidase measurements, a single section was scanned as quickly as possible after transferring the slide to the micro- scope, and the increase in fluorescence with time was mea- sured over the next 35-40 min in periodic scans. Fluores-
ORGAN CULTURE AND INJECTION OF NEURONS
43
cence was quantified by measuring average pixel intensity for comparable single sections using NIH Image software (Shareware by Wayne Rasband. National Institutes of Health). Following each imaging session, preparations were rinsed again in sterile saline and lobster L- 1 5 and repinned m the culture dish; the culture was continued at 16°C.
Results
Sunivul of central neurons in lon^-terin orifun culture
The ventral nerve cord was removed from an adult lobster and maintained in a modified Leibovitz L-15 culture me- dium for at least 7 weeks. Over this period, identified neurons maintained features typical of their condition in the short-term preparation routinely used for physiological ex- periments (usually from 12 to 36 h).
Phvsiology of neurons. The positions of the somata of many neurons in the lobster ventral nerve cord have been mapped (Otsuka el al. 1967: Beltz and Kravitz. 1983; Schwarz el al., 1984; Schneider et al., 1993). Cell bodies can be reliably and readily identified from preparation to preparation by their relative sizes, their position in a gan- glion, their endogenous and nerve-evoked synaptic input, their spontaneous activity, and by backfiring their axons. To illustrate, 12, a large GABA-containing inhibitory motor neuron innervating the fast flexor muscles, was studied in a preparation maintained for 49 days in culture (Fig. 1). As shown in the schematic, an intracellular recording electrode was placed in the putative 12 cell in this ganglion, and stimulating electrodes were placed on the anterior connec- tive (point 1) and contralateral 3rd root of the ganglion (point 2). Stimulation of the contralateral 3rd root led to the appearance in the cell body of small action potentials (ac- tion potentials do not invade these somata) that retained a constant amplitude with increasing intensities of stimula- tion. No other large cell bodies showing this property were seen in that location, hence the identification of the cell as 12 is highly likely and further supported by the demonstra- tion that these neurons contain GABA. In contrast, with increasing levels of stimulation of the anterior connective (point 1, Fig. 1, left), excitatory postsynaptic potentials measured in 12 increased in amplitude, finally generating action potentials in these cells. This suggests that a progres- sive recruitment of presynaptic inputs to 12 is occurring, and that these inputs, too. have survived the long-term organ culture. When the connective between electrode 1 and the ganglion was crushed, excitation of 12 via this route was abolished (Fig. 1, top right), thereby assuring that the re- sponses seen were not due to current spread from the stimulating electrode. Resting membrane potentials in 12 and other cells in this and in other experiments were within normal ranges (—50 to —65 mV, data not shown). GABA was detected in 12 cell somata dissected from organ-cul- tured ganglia at concentrations normally found in 12 cell
crush ant. conn.
ins
mv
ms
Figure 1. Normal actmt\ in a lobster ganglion maintained in organ culture for 49 days. While recording intracellularly from an abdominal GABA-containing cell. \2. the preparation was stimulated from the anterior connective (point 1 ). Increasing stimulation intensity (left panel, bottom to top) elicited an increasing excitatory response in 12 (at point 1 1 that finally triggered an action potential. After the anterior connective was crashed between the stimulating electrode and the ganglion, the EPSPs in 12 disappeared (upper right). When the 3rd root was stimulated directly (point 2 at right), an antidromic action potential was triggered in 12 and recorded in the cell body. As expected, the action potential showed no size change with increasing stimulus intensity.
bodies from freshly dissected preparations [fresh prepara- tions— 0.79 ± 1.75 X 10~'" moles/cell body (mean ± SD, „ = 14) (Otsuka et al.. 1967); 6-week cultures — 1.1 ± 7.1 1 X 10"'° moles/cell body (mean ± SD, n = 4)].
Evidence of RNA synthesis by neurons. RNA synthesis in cultured nerve cords was measured by radiolabeled pre- cursor incorporation and autoradiography. In Figure 2, one example is presented. Lobster ganglia maintained in organ culture for 1 and 49 days were incubated in medium con- taining 'H-uridine for 4 h. followed by a washout with unlabeled medium for 20 h. Ganglia were fixed, embedded, and sectioned, and autoradiography was performed. Silver grains are readily seen over the nuclei of neurons and glial cells (Fig. 2). These represent the location of newly synthe- sized RNA and demonstrate that neurons in long-term cul- ture still actively transcribe RNA. Qualitatively, we saw
44
G. K. CANTER ET AL
Figure 2. Autoradiography of newly synthesized RNA in lobster gan- glia maintained for 2 and 49 days in vitro. Ganglia were treated with 3H-uridine lor 4 h. followed by a chase with unlabeled medium for 20 h. Following fixation, samples were embedded and sectioned, and autora- diography was performed. The resulting silver grains clustered primarily at the nuclei of cells cultured for 2 and 49 days indicate the presence ol nascent RNA.
little difference in grain density between the 2-day and 49-day organ cultures.
Similar labeling experiments were performed after other times of incubation with both 3H-uridine and 3H-leucine. In the latter experiments, a cytoplasmic rather than nuclear distribution of silver grains was noted after autoradiogra- phy. indicating the synthesis of protein in the cultured ganglia (data not shown).
Cytology of ganglia. Some indication of overall health can be seen in the cytology of cultured ganglia. Cells of ganglia cultured for the longer time period appeared similar to cells of recently dissected ganglia, although the former usually had larger spaces between cells (not shown). Rarely, opaque or darkly colored cells were observed in cultured lobster ganglia, such as those shown in Figure 3 (bottom panel). These cells typically had low or no resting potentials and were likely to be dead or dying. Another feature of unhealthy cells is their slight autofluorescence (Fig. 3, top). The bright cell in Figure 3. top panel, is a GFP-expressing cell. As seen in bright-field illumination (Fig. 3, bottom panel), the GFP-expressing cell is transparent, whereas the two weakly fluorescent cells are opaque. The death of these two cells resulted from experimental DNA injection, de- scribed below. The numbers of spontaneously unhealthy
cells was generally low throughout the culture period: most cells and their processes appeared morphologically undis- tinguishable from their counterparts in freshly dissected ganglia.
Expression of introduced RNAs in organ-cultured neurons
A single injection of cRNA for either GFP or /3-galacto- sidase into central neurons in cultured ganglia led to the synthesis of these proteins. Their presence was detected with fluorescence and confocal microscopy, either directly in the case of GFP, or indirectly for (3-galactosidase, by adding a fluorogenic substrate for the enzyme. Measure- ments were made by periodically locating the same injected cell in organ-cultured cords.
Induction of GFP expression. An identified cell, usually 12. M6, or M7, was pressure-injected with the cRNA for GFP and maintained in organ culture for periods of up to 10 days. At various time points the injected ganglia were imaged using a confocal fluorescence microscope to detect and measure GFP expression. GFP fluorescence in this cell was detected after one day, increased until day 4. then decreased to background
Figure 3. Comparison of a GFP-expressing neuron with two cells injected with high concentrations of plasmid DNA, The highly fluorescent cell seen with epifluorescent illumination (top panel) was injected with cRNA encoding green fluorescent protein (GFP). After 3 days of incuba- tion, this cell appeared clear under bright-field illumination (bottom) In contrast, two cells injected with plasmid DNA fluoresced only \\euklv (lop) and were opaque under bright-field illumination (bottom). These cells had poor (or no) resting membrane potentials and resembled untreated cells that occasionally died during prolonged organ culture.
ORGAN CULTURE AND INJECTION OF NEURONS
45
100-
r 75-
50-
25-
0
012 4 6 8 10 days post-injection
Figure 4. Time course of GFP expression after injection of cRNA. (Left) A single neuron injected with GFP-encoding cRNA was photographed, using a confocal fluorescent microscope, at various time points (in days) following injection. (Right) The intensity of fluorescence increased to a maximum at day 4, then declined to background at day 10.
by day 10 (Fig. 4). Other cells injected with this cRNA showed similar patterns of expression.
As the level of GFP fluorescence declined from its peak, its distribution in injected cells became patchy. In a few injections. GFP distribution was uniform in the cytoplasm but later became punctate, and the protein product appeared to be excluded from certain areas of the cytoplasm. In many cases, GFP fluorescence was seen in the primary neurite leaving the cell body (Fig. 5). However, we were unable to detect a GFP signal in deeper layers of the neuropil. in connectives, or in peripheral nerves.
Induction of f)-galactosidase expression. Injection ot the cRNA for /3-galactosidase into cell somata resulted in enzyme activity detectable in cultured neurons within 2 days. Expres- sion of this enzyme was determined by using the fluorogenic substrate fluorescein di-(/3-D-galactopyranoside) to measure activity. Hence, we did not directly measure the amount of protein synthesized. Since enzyme activity probably was di- rectly related to the amount of protein present, levels of ex- pression of /3-galactosidase apparently peaked between 6 and 9 days after injection. Strong /3-galactosidase activity was still easily detectable in cultured ganglia 10 days after cRNA in- jection (Fig. 6). The fluorescence intensity increased rapidly and close to linearly in a /3-galactosidase-expressing cell fol- low ing addition of the fluorogenic substrate. Since the fluores- cent cleavage product leaves the cells within a few hours, the same cells can be repeatedly tested for /3-galactosidase activity on consecutive days.
Injection of DNA into lobster neurons
We attempted to induce reporter gene expression by the cytoplasmic microinjection of plasmid DNA. Identified
cells were injected with DNA constructs containing three different promoters: the human CMV immediate early en- hancer/promoter; Drosophila melanogaster hsp70; and D. melanogaster elav promoter driving either GFP or /3-galac- tosidase. Injections of low concentrations of DNA (below 1 jug//u,l in the electrode) had no detectable toxic effect on cells, whereas injections of DNA at concentrations above 1 jug/ju,l led to cell death. Most cells injected with higher levels of DNA became opaque and autofluorescent within 1 to 2 days (see Fig. 3), and all such cells showed low or no resting membrane potentials, indicating that they were dead or in the process of dying. In no case (low or high concen- trations of DNA). however, did we observe any protein product expression.
Discussion
We have used an organ culture method to maintain the isolated central nerve cord of the lobster for up to 7 weeks, a significant extension beyond the 1 to 2 days previously possible. During this time RNA and protein synthesis, as well as physiological activity in identified neurons, ap- peared normal. This technique makes possible a range of long-term experiments, including study of the effects of pharmacological and hormonal treatment on the central nervous system, and the effects of manipulation of gene expression in central neurons.
Long-term culture of lobster ventral nerve cord
Ventral nerve cord preparations of crustaceans are valuable for exploring central circuitries because they allow absolute identification of neurons (see Otsuka et ui, 1967; Kennedy et
46
G. K. CANTER ET AL
Figure 5. Lobster neuron injected with GFP cRNA. (Top) A desheathed lobster central ganglion at low magnification. (Middle) High- magnitication view of an injected cell (center) under bright Held illumina- tion. (Bottom) The same view in dark field, showing fluorescent signal from the expressed GFP protein. The signal fills the injected soma and can be seen in the proximal section of the primary neurite as it descends into the neuropil.
ill., 1969; Roberts ct al.. 1982; Beltz and Kravitz, 1987; Ma et til.. 1992; Yehf/«/.. 1996; Homer et al., 1997). Once exposed by removal of the connective tissue sheath, the somata of these neurons are easily visible and accessible to microelectrodes. Cells occur in predictable arrangements and can be unambigu- ously identified by their position, size, and activity, and by backfiring their axons from roots or connectives (Otsuka et al., 1967). In our laboratory, preparations of this type have been used to explore the roles of amines in the neural networks involved in postural regulation ( Harris- Warrick and Kravitz, 1984; Beltz and Kravitz, 1987; Ma et al.. 1992; Weiger and Ma, 1993). In crayfish, similar preparations have been used to define changes in synaptic properties accompanying changes in social status at particular synaptic sites (Yeh et nl.. 1996).
In the organ culture system described here, we showed that neurons are suitable for electrophysiological experi- mentation for at least 7 weeks. We used only one example for illustration. In Figure 1 we showed that the large inhib- itory motoneuron 12 could be activated by backfiring its axon at a distance of several centimeters from the cell body, and that interneurons upstream of 12 still could relay signals via the release of transmitters. Moreover, in single 12 cells dissected from long-term organ cultures of ganglia, the intracellular levels of GAB A were comparable to those found in cells dissected from fresh preparations.
The autoradiographic studies demonstrated that neurons in ganglia cultured for 49 days still synthesized RNA (Fig. 2) and protein (data not shown). In addition, at a light microscopic level, the cytology of cultured neurons ap- peared normal, except for a somewhat more frequent vacu- olar appearance of cell somata, and for larger spaces be- tween cells and neuropil processes. These differences may result from looser cell packing in the excised and desheathed ganglia, from degeneration of sensory fibers from the periphery (Barker et al., 1972), or both.
Cells that do not survive in organ culture appear dark with light microscopy and their autofluorescence is weak, making them easy to distinguish from surviving cells. Autofluores- cence of unhealthy or dead cells has been reported elsewhere (Linnik et al., 1993; Kosslak et al, 1997), and is a useful way to prevent making comparisons between these cells and healthy cells in the cultures (O'Brien et al., 1995). It is impor- tant to identify such cells so that their fluorescence is not confused with that of experimental markers like GFP.
GFP and LacZ are suitable expression markers in lobster neurons
GFP and LacZ (the gene encoding jS-galactosidase) have been used as reporter and marker genes in studies involving a wide variety of organisms (see Chalfie. 1995). Lobsters and crayfish (see Dearborn et al., 1998) now can be added to the list of organisms capable of expressing these proteins. The studies presented here, and those reported by Dearborn et nl. ( 1998), suggest that it may be possible to use GFP and j8-galactosidase as injection markers, as reporters for tran- scriptional studies, and as tags that can be fused to proteins under study in crustacean systems. To illustrate the latter, in preliminary studies (G. Ganter. unpubl. obs.), we have found that intracellular injections of cRNAs encoding a human (3-2 adrenergic receptor/GPP fusion and a lobster amine receptor/GFP fusion result in fluorescent signals. Although different cRNAs might show differences in their rates of translation, thus far all cRNAs that we have injected have been expressed. Indeed, the different times we noted for peak detection of jS-galactosidase and GFP could be due to differences in the translation rates of these two transcripts.
ORGAN CULTURE AND INJECTION OF NEURONS
47
0 5 10 15 20 25
min. after addition of fluorescein di-(B-D-galactopyranoside)
Figure 6. |3-galactosidase activity in a lobster neuron maintained 10 days in organ culture alter injection with LacZ cRNA. /3-galactosidase activity was detected in a living injected cell by addition of a fluorogenic substrate, fluorescein di-(fJ-D-galactopyranoside). (Left) The cell was photographed, using a confocal fluorescent microscope, at various time points (in minutes) following addition of the substrate. (Right) The intensity of fluorescence increased almost linearly
Future uses of the organ culture method
Long-term organ culture offers a controlled environment in which to perform experiments that were difficult to carry out in short-term studies using isolated nerve cords. This should allow us to explore the consequences of applying test sub- stances to central ganglia in more biologically relevant time frames. For example, the lobster molting hormone, 20-hy- droxyecdysone, is likely to have actions at both membrane and genomic levels (see Zakon, 1998). The genomic effects may take days to bring about observable changes at synaptic or circuit levels. The organ culture system offers enough time for such changes to be seen, in an environment in which tissues will continue to synthesize the RNA and protein required to trigger the changes. Long-term drug effects relating to amin- ergic function also can be examined in the cultured ganglia. At present we are particularly interested in chronic exposure of ganglia to Prozac (fluoxetine). In behavioral studies, acute exposure to Prozac has little effect on agonistic behavior in lobsters (Huber el ai, 1997); in contrast, chronic exposure increases the amount of time that animals are willing to fight (A. Delago, unpubl. obs.).
Another exciting application of the organ culture method is in the experimental manipulation of gene expression in identified neurons. The extended time frame makes it pos- sible to examine the effects of gene expression on the physiology of cRNA-injected neurons and to analyze cloned genes in their native environment. For example, the seroto-
nin-containing neurosecretory neurons of lobster appear to lack the mRNA for the shab form of the potassium channel (H. Schneider, unpubl. obs.). It would be interesting to express shab in these neurons, and then examine the con- sequences of this manipulation on the intrinsic properties of the injected cell and on the network in which it functions. Other applications could include injections of sense, anti- sense, or double-stranded RNAs coding for particular pro- teins into cells to ask how such manipulations alter function. These and other applications await further exploitation of this important system.
Summary and conclusions
We describe an organ culture method that maintains isolated lobster ganglia in viable states for up to 7 weeks. We have validated the method, showing that evidence of gene expres- sion and electrophysiological activity persist throughout the culture period. Applications include long-term experiments to examine the consequences of chronic treatment with pharma- cological or hormonal reagents and of changes to the levels of expression of particular genes in single neurons and in net- works of neurons. The ability to introduce genes into lobster nerve cells should allow analysis of the function of these genes in their native environment, narrowing the gap between mo- lecular methods and the study of physiology in this important system.
48
G. K. CANTER ET AL.
Acknowledgments
We acknowledge Dr. Gary N. Cherr, Dr. Frederick J. Griffen. and Ms. Carol Vines for help with confocal mi- croscopy: the staff of the Bodega Marine Laboratory, Dr. Ernest S. Chang, and Ms. Sharon Chang for providing technical support, and a rich environment; and the Univer- sity of California. Davis, for the fellowship support that funded these studies. We also thank Drs. Margaret Bradley and Stuart Cromarty for their ideas about the project and their helpful comments on the manuscript. This research was supported by a grant from the National Science Foun- dation (to EAK) and by a post-doctoral fellowship from the Alexander von Humboldt Foundation (to RH).
Literature Cited
Aisemberg, G. O., T. R. Gershon, and E. R. Macagno. 1997. New
electrical properties of neurons induced by a homeoprotein. J. Neuro- biol. 33: 11-17.
Barker. D. L., E. Herbert. J. G. Hildebrand. and E. A. Kravitz. 1972. Acetyleholine and lobster sensory neurons. J. Physitil. 226: 205-229.
Beltz. B. S.. and E. A. Kravitz. 1983. Mapping of serotonin-like im- munoreaetivity in the lobster nervous system. J. Neiirosci. 3: 585-602.
Beltz. B. S., and E. A. Kravitz. 1987. Physiological identification, morphological analysis and development of identified serotonin-proc- tolin containing neurons in the lobster ventral nerve cord. J. Neitroxci. 7: 533-546.
Challie, M. 1995. Green fluorescent protein. I'lioincheni. Pliniohinl. 62: 651-6.
Chalfie, M., Y. Tu, G. Euskirchen, VV. VV. Ward, and D. C. Prasher. 1994. Green fluorescent protein as a marker for gene expression. Science 263: 802- S04.
Davtid, I. B., and T. I). Sargent. 1988. AVmyw.v lucvis in developmental and molecular biology. Science 240: 1443-1448.
Dearborn, R. E., Jr., B. G. Szaro, and G. A. Lnenicka. 1998. Micro- injection of mRNA encoding rat synapsm la alters synaptic physiology in identified motoneurons of the crayfish. Prucambarua clarkii. J. Neu- rohiol. 37: 224-236.
Gerdes, H. H.. and C. Kaether. 1996. Green fluorescent protein: appli- cations in cell biology. FEBS Lett. 389: 44-47.
Harris-Warrick, R. M., and E. A. Kravitz. 1984. Cellular mechanisms for modulation ol posture by octopamine and serotonin in the lobster. J. M-»/-.H,7. 4: 1976-1993.
Hendelman, W. J., and R. P. Bunge. 1969. Radioautographic studies of choline incorporation into peripheral nerve myelin. J. Cell Binl. 40: 190-208.
Horner, M., W. A. Weiger, D. H. Edwards, and E. A. Kravitz. 1997. Excitation of identified serotonergic neurons by escape command neu- rons in lobsters ./ Ay. Binl. 200: 2017-2033.
Huber, R., K. Smith. A. Delago, K. Isaksson, and E. A. Kravitz. 1997. Serotonin and aggressive motivation in crustaceans: altering the deci- sion to retreat. Proc. Nail. Acud. Sci. USA 94: 5939-5942.
Jakoby. VV. B., and E. M. Scott. 1959. Aldehyde oxidation. III. Succmic semialdehyde dehydrogenase. J. Binl. Client. 234: 937-940.
Kaang, B.-K., P. J. Pfaffinger, S. G. N. Grant, and E. R. Kandel. 1992. Overexpression of an Aplysia Shaker K ' channel gene modifies the electrical properties and synaptic efficacy ol identified /\/>/v,v?(/ neurons. Proc. Nail. Acud. Sci. USA 89: 1 133-1 137.
Kennedy. I)., A. I. Selverstcm, and M. P. Remler. 1969. Analysis of restricted neural networks. Science 164: 1488-1496.
Kosslak, R. M., M. A. Chamberlin, R. G. Palmer, and B. A. Bowen.
1997. Programmed cell death in the root cortex of soybean root necrosis mutants. Plant ./. 11: 729-745.
I .iliumc n . M., and C. I). Richardson. 1995. Production of recomhi- nant baculoviruses using rapid screening vectors that contain the gene for beta-galactosidase. Mali. Mol. Binl. 39: 161-177.
Linnik, M. D., M. D. Hatfield, M. D. Swope. and N. K. Ahmed. 1993. Induction of programmed cell death in a dorsal root ganglia X neuro- hlasioma cell line. J. Netirohiol. 24: 433-446
Livingstone, M. S., R. M. Harris-Warrick, and E. A. Kravitz. 1980. Serotonin and octopamine produce opposite postures in lobsters. Sci- ence 208: 76-79.
Ma, P. M., and W. A. Weiger. 1993. Serotonin-containing neurons in lobsters: the actions of y-aminobutyric acid, octopamine. serotonin, and proctolin on activity of a pair of identified neurons in the first abdom- inal ganglion. J. Nenropli\.\iol. 69: 2015-2029.
Ma, P. M., B. S. Beltz, and E. A. Kravitz. 1992. Serotonin-containing neurons in lobsters: their role as "gain-setters" in postural control mechanisms. / Neiiniphyxiol. 68: 36-54.
MacGregor, G. R., A. E. Mogg. J. F. Burke, and C. T. Caskey. 1987. Histochemkal staining of clonal mammalian cell lines expressing E. culi beta-galactosidase indicates heterogeneous expression of the bac- terial gene. Somatic Cell Mol. Genet. 13: 253-265.
O'Brien, M. C., and VV. E. Bolton. 1995. Comparison of cell viability probes compatible with fixation and pemieabilization for combined surface and intracellular staining in flow cytometry. Cyitnnetry 19: 243-255.
Otsuka, M., E. A. Kravitz. and D. D. Potter. 1967. The physiological and chemical architecture of a lobster ganglion with particular reference to gamma-aminobutyrate and glutamate. J. Neurophysiol. 30: 725-752.
Pelham, H. R. 1982. A regulatory upstream promoter element in the DriKopliilu h.\p70 heat-shock gene. Cell 30: 517-528.
Prasher, D. C., V. K. Eckenrode, VV. W. Ward, F. G. Prendergast, and M. J. Cormier. 1992. Primary structure of the Act/norm \-tctoria green-fluorescent protein. Gene 111: 229.
Roberts. A., F. B. Krasne, G. Hagiwara. J. J. Wine, and A. P. Kramer. 1982. Segmental giant: evidence for a driver neuron interposed be- tween command and motor neurons in the crayfish escape system. J. Neiimnliyuol. 47: 761-781.
Schneider, H.. B. Trimmer, J. Rapus, M. Eckert. D. Valentine, and E. A. Kravitz. 1993. Mapping of octopamine-immunoreactive neu- rons in the central nervous system of the lobster. ./. Com/'. Nenrol. 329: 129-142.
Schwarz, T. L., G. M.-H. Lee, K. K. Siwicki, D. G. Standaert, and E. A. Kravitz. 19S4. Proctolin in the lobster: the distribution, release, and chemical characterization of a likely neurohormone. J. Ncurotci. 4: 1300-131 I.
Thnmsen, D. R.. R. M. Stenberg, VV. F. Goins. and M. F. Stinski. 1984. Promoter-regulatory region of the major immediate early gene of human cytomegalovirus. Proc. Null. Aaul. Sci. L/SA 81: 659-663.
Weiger, W. A., and P. M. Ma. 1993. Serotonin-containing neurons in lobsters: the actions of ganima-aniinobutync acid, octopamine. seroto- nin and proctolin on ncuronal activity. ./. iVcnro/iliy\iol 69: 2015-2029.
Vao, K.-M.. and K. White. 1994. Neural specificity ol <7<n expression: defining a Drosupliilu promoter for directing expression to the nervous system. J. Nettrochem. 63: 41-51.
Yeh. S., R. Fricke. and D. Edwards. 1996. The eflect of social expe- rience on serotonergic modulation of the escape circuit of crayfish. Science 271: 366-369.
/akon. H. H. 1998. The effects of steroid hormones on electrical activity of excitable cells. Trcml\ Nctirosci. 21: 202-207.
/hao, B., F. Rassendren, B.-K. Kaang, Y. Furukawa, T. Kuho. and E. R. Kandel. 1994. A new class of noninactivating K' channels from Apl\sia capable of contributing to the resting potential and firing patterns of neurons. Neuron 13: 12(15-1213.
Reference: Biol. Bull 197: 49-62. (August
An Ethogram of Body Patterning Behavior in the
Biomedically and Commercially Valuable Squid
Loligo pealei off Cape Cod, Massachusetts
ROGER T. HANLON, MICHAEL R. MAXWELL, NADAV SHASHAR. ELLIS R. LOEW1,
AND KIM-LAURA BOYLE
Marine Resources Center, Marine Biological Laboratory, Woods Hole, Massachusetts 02543: and ' Department of Biomeilical Sciences, Cornell University; Ithaca, New York 14853
Abstract. Squids have a wide repertoire of body patterns; these patterns contain visual signals assembled from a highly diverse inventory of chromatic, postural, and loco- motor components. The chromatic components reflect the activity of dermal chromatophore organs that, like the pos- tural and locomotor muscles, are controlled directly from the central nervous system. Because a thorough knowledge of body patterns is fundamental to an understanding of squid behavior, we have compiled and described an etho- gram (a catalog of body patterns and associated behaviors) for Loligo pealei. Observations of this species were made over a period of three years (>440 h) and under a variety of behavioral circumstances. The natural behavior of the squid was filmed on spawning grounds off Cape Cod (northwest- ern Atlantic), and behavioral trials in the laboratory were run in large tanks. The body pattern components — 34 chro- matic (including 4 polarization components). 5 postural, and 12 locomotor — are each described in detail. Eleven of the most common body patterns are also described. Four of them are chronic, or long-lasting, patterns for crypsis: an example is Banded Bottom Sitting, which produces disrup- tive coloration against the substrate. The remaining seven patterns are acute; they are mostly used in intraspecific communication among spawning squids. Two of these acute patterns — Lateral Display and Mate Guarding Pattern — are used during agonistic bouts and mate guarding; they are visually bright and conspicuous, which may subject the squids to predation; but we hypothesize that schooling and diurnal activity may offset the disadvantage presented by
Received 1 February 1999; accepted 20 April 1999. E-mail: rhanlon@nihl.edu
increased visibility to predators. The rapid changeability and the diversity of body patterns used for crypsis and communication are discussed in the context of the behav- ioral ecology of this species.
Introduction
Cephalopods have a highly developed system of visual communication that is expressed mainly through the skin. The distinguishing features of this remarkable chromato- phore system are its speed of change and the diversity of body patterns that each individual uses for either crypsis or communication (Hanlon and Messenger. 1996). A body pattern is defined as the total appearance of the animal at any given time, and includes the expression of the full complement of chromatic (i.e.. color or visual), textunil. postural, and locomotor components (see Packard and Hochberg. 1977; Hanlon and Messenger, 1996). Among the components of the body pattern, the most conspicuous are chromatic, although squids probably perceive intraspecific signals monochromatically because cephalopods are thought to be color blind (Hanlon and Messenger, 1996). These chromatic components are produced primarily by chromatophore organs and various reflective cells in the dennis, and they are discrete neural entities (just as postural, textural. and locomotor components are) because the chro- matophore organs are controlled by radial muscles under the direct control of the posterior chromatophore lobes in the brain (e.g.. Dubas et al.. 1986). Most of the reflective cells are also controlled by the squid (Cooper et al., 1990). This neural control enables the cephalopod to change its appear-
49
50
R. T. HANLON ET AL
ance in a fraction of a second, depending upon the visual sensory input it receives during behavioral interactions. Few. if any. animals can match the speed of change and diversity of cephalopod signals, and the body patterns are used in most behavioral interactions, whether they be for competition for resources or mates, or interactions between predators and prey. We are documenting these diverse body patterns, focusing primarily on adult squids during their inshore migration every year.
Squids, like other cephalopods, are sensitive to the partial polarization characteristics of light (Saidel et ai. 1983; Hanlon and Messenger. 1996; for a description of polarized light see Kattawar. 1994; Wolff and Andreou. 1995). Shashar and Hanlon (1997) described a few specific polar- ization components of squid and correlated these patterns with the distribution of iridophore cells in the animals' skin. In cuttlefish, partial polarization patterns have been associ- ated with communication (Shashar et ai, 1996). Since squids may use polarization patterns for intraspecific com- munication, and since polarization-sensitive predators may be looking for polarization contrasts to locate squid prey, we also document here some polarization components pre- sented by the squid.
The long-tinned squid Loligo pealei Lesueur, 1821, is a renowned model in neuroscience research. The third-order giant axon. its attendant giant synapse, the complex eye, and several other organ systems in L. pealei have been studied in detail for over 50 years at the Marine Biological Labo- ratory (MBL) in Woods Hole (see Gilbert et til., 1990). Although a great deal is known about the peripheral nervous system of L. pen lei. little is known about the behaviors of this squid, which like most cephalopods, has an enormous brain relative to its body size. Loligo pealei is also a valuable commercial resource in the northeastern United States — worth about $30 million annually (McKiernan and Pierce. 1995; NEFSC, 1995). Curiously, little is known about the ecology, life history, and behavior of this species (e.g., Verrill. 1880; Drew. 191 1; Stevenson. 1934; Griswold and Prezioso, 1981; Summers, 1983; Gilbert et al. 1990; Brodziak and Macy. 1996). The present report is part of a broad-based study that focuses on sexual selection pro- cesses in L. pealei from two perspectives: as a test of sexual selection theory (e.g.. Hanlon. 1996; Hanlon et til.. 1997) and as a study of the role that reproductive behavior plays in the life history and population dynamics of the species (Hanlon. 1998).
Materials and Methods
The behavior of Loligo pealei can be observed both in a natural setting and in the laboratory because the squids habituate quickly to divers and to laboratory surroundings.
Overall, 27.5 h of videotape were analyzed for body pat- terning and behavior.
During the months of May 1996. May 1997, and May 1998. 103 scuba dives were made on squid spawning grounds by RTH and NS off the southern arm of Cape Cod, Massachusetts. Depths ranged from 3-10 m and most sites were within 2 km of shore between Hyannis and Chatham. Spawning squids were found mostly in or near commercial weir traps whose inner pocket dimensions (or capture arena) were roughly 20 m2; often there were many thousands of squids in these traps, with a proportion of them actively engaged in reproductive behavior. Water temperatures ranged from about 4° to 13°C. currents were often strong, and visibility was usually poor. On about one-third of the dives, conditions were suitable for video. In total. 16.5 h of dive video were recorded, using video cameras (either an- alog or digital) in underwater housings, and analyzed, with multi-motion playback machines and high-resolution mon- itors.
Laboratory trials of mating behavior were performed from May through October in 1996, 1997. and 1998 in the Marine Resources Center of the MBL. Three large tanks were used, each measuring 3 m (diameter) by 1 m (height) and containing about 28,000 1 of seawater. Each tank had a substrate of mixed gravel and sand, and a continuous supply of ambient seawater. Animals were acquired by squid jig- ging (both at night and during the day) off the MBL re- search vessel Gemma in Vineyard and Nantucket Sounds. This method minimizes skin damage for maximal survival in captivity (see Hanlon el al., 1983). Squids were fed live fish (Fundiihis sp.) daily. Trials involved from three to eight squids in various combinations of males and females. One set of trials was performed in an outdoor pond, 20 m X 20 m X 1 m deep, at the Environmental Systems Laboratory of the Woods Hole Oceanographic Institution. The squids were observed for 440 h in captivity. 1 1 of which were recorded on video.
All videos were reviewed multiple times, each time look- ing for only one category of component (i.e.. first viewing for chromatic components, second viewing for postural components, third viewing for locomotor components). In the laboratory, chromatic, postural, and locomotor compo- nents were recorded on separate data sheets each time they were seen. A chromatic component was recorded if it was expressed for at least 2 s; locomotor and postural compo- nents were recorded if they were performed for at least 3 s. All chromatic components were illustrated using a computer graphics program.
Polarization components were recorded using a video polarimeter based on a standard three-tube ENG camera (JVC BY-110) that uses a dichroic prism block for color separation. The dichroic prism has been replaced with a custom-made neutral prismatic splitter (Richter Enterprises,
SQUID BODY PATTERNING AND BEHAVIOR
51
Manhattan Beach, CA) such that each of the three video channels receives 1/3 of the broad-spectrum image input. Since this assembly lacks the color-trimming filters ce- mented to the original dichroic prism, magnification errors due to pathlength differences were corrected with small quartz discs of appropriate thickness. A small disc of sheet polarizer (Polaroid, HNP'B) was placed immediately in front of each camera tube to impart polarization sensitivity to the channels. The orientation of the polarizers was ad- justed so that the color channels now encoded 0°, 45°, and 90° polarization images. The camera electronics encode the three polarization channels as if they were color, making it possible to store all the data on a regular portable videocas- sette recorder and allowing for immediate viewing of a pseudocolor polarization image on a color monitor. Nonpo- larizing elements of the scene have no color, whereas po- larizing elements do. The signal in all three channels is identical, and the output of the tubes was adjusted to give white for a saturating faceplate intensity. A polarizer placed in front of the lens such that horizontally polarized light is freely transmitted produces the following normalized sig- nals in the three "color" channels: the R channel signal is 1, the G channel is 0.707, and the B channel is 0. Monochro- matic images of the same scene, taken from the three channels separately, were transferred through a frame grab- ber into the computer and their linear polarization charac- teristics were analyzed following procedures in Cronin et al. (1994). This camera is better suited than previously de- scribed polarimeters (Cronin et al., 1994; Wolff and An- dreou, 1995; Horvath and Varju, 1997) for recording the polarization patterns of moving animals, because it provides true instantaneous measurements. Technological limitations made it impossible to get the camera in an underwater housing; thus measurements were limited to the laboratory. Furthermore, the light conditions during measurements had to be precisely controlled, thereby allowing only 3 h of recorded footage. During these periods, the squids exhibited only a few behaviors that included fighting, mate guarding, and egg laying.
Ethogram
We constructed an ethogram for Loligo pealei on the basis of our field and laboratory observations. The compo- nents and body patterns identified (Table I) represent a segment of all behaviors, especially those related to repro- duction. In fact, because of the size of the sample, most of the patterning components of the species were probably identified. The more than 440 h of observation far exceed the observation periods in other published accounts of Lo- ligo spp. (e.g., Hanlon, 1982, 1988; Hanlon et al., 1983, 1994; Porteiro et al., 1990).
The chromatic components of the ethogram are illustrated in Figures 1 and 3, and some of the postural components are shown in Figure 2. Unlike octopuses and cuttlefishes, loli- ginid squids do not show textural components in the skin. Table I, which lists all components, includes the number of times that we counted a component on videotape or from observation notes, giving an impression of how commonly it occurs. Unless otherwise indicated, all components and body patterns were shown by both sexes.
Light
' components
Chromatic components are produced mainly by the action of dermal chromatophore organs, which number in the hundreds of thousands in an adult squid. Loligo pealei has three color classes of chromatophores: yellow, red, and brown. Expansion of the chromatophores darks the skin, while retraction of the chromatophores (and the resultant expression of underlying iridophores) produces a lightening or even brightening effect. Intense darkness produced by maximal expansion and intense brightness produced by maximal retraction mark two ends of a chromatic contin- uum, and thus it is somewhat arbitrary to assign a compo- nent to light or dark. Some of these components are com- mon to other Loligo spp., as described by Hanlon ( 1982) for Loligo plei, by Porteiro et ul. ( 1990) for Loligo forbesi, and by Hanlon et al. ( 1994) for Loligo vulgaris reynaudii.
Clear is retraction of all or most chromatophores, thus rendering the animal translucent in clear water or white in murky water. In clear water, when viewed against a sand bottom or laterally against the aquatic background (Fig. 2B), the translucence renders the squid cryptic, or camou- flaged, and often the Dorsal iridophore splotches are ex- pressed simultaneously. Internal organs, such as the red accessory nidamental gland in females, are often visible. In murky water. Clear appears bright white in most lighting circumstances (i.e., the brightness surpasses the albedo of the greenish water, producing a whitish color). In the im- mediate vicinity of egg beds, the white form of Clear seems to function as an intraspecific signal to repel other squids; a squid displaying this component is almost always engaged in mate guarding, egg laying, or agonistic bouts (see Fig. 2C). White arms/head results from variable retraction of chromatophores on the head and arms (three variations are illustrated in Fig. 1 ). This component sometimes preceded all white (or clear) in intraspecific encounters; thus, it ap- pears to be a milder signal of alarm or repellent to approach- ing squids (Fig. 2G). White head/arms is most common in paired females near eggs and is seen when unpaired males approach. White dorsal stripe is retraction of chromato- phores along a dorsal mantle that is otherwise dark; the stripe may be short or long (Fig. 1). It has been seen in
52
R. T. HANLON ET AL
Table I
Body patients and their components in the squid Loligo pealei; compare Figure 1
BODY PATTERNS
Chronic (mm to hours)
1. Basic Amber Pattern
2. Clear Body Pattern
3. Countershading
4 Chronic AM Dark
5. Banded Bottom Sitting
6. Chronic Bright White Pattern
Acute (seconds)
1. Very Dark
2. Blanch-Ink-Jet Maneuver
3. Lateral Display
4. Mate Guarding Pattern
5. Accentuated Testis
COMPONENTS* Chromatic
Light:
1 . Clear
2. White arms/head
3. White dorsal stripe
4. Accentuated testis (m)
5. Accentuated oviducal gland (f) Iridescent:
6. Dorsal mantle collar indophores
7. Iridescent sclera
8. Dorsal iridophore splotches
9. Iridescent arm stripes 10. Dorsal iridophore sheen
Light polarization components:
1 . Polarized arms
2. Skin surface polarization
3. Polarized eyes
4. Polarized dorsal sheen
(861)
(769)
(194)
(1179)
(183)
(a 1000)
(167) (500) (338) (32)
Dark:
1. All dark
2. Dark arms/head
3. Dark head
4. Dark dorsal stripe
5. Ventral mantle stripe
6. Mantle margin stripe
7. Dark arm stripes
8. Fin spots
9. Arm spots