Distinctive Cytoskeletal Organization in Erythrocytes of the Cold-Seep Vesicomyid Clam, Calyptogena kilmeri

7 Pages·1998·2.9 MB·English

by  W D Cohen

Reference: Biot. Bull. 194: 7-13. (February, 1998)
Distinctive Cytoskeletal Organization in
Erythrocytes of the Cold-Seep Vesicomyid Clam,
Calyptogena kilmeri
WILLIAM D. COHEN'* AND MARIO N. TAMBURRI"
' Department ofBiological Sciences. Hunter College ofCUNY, 695 Park Ave., New York,
New York 10021: and ' Monterey Bay Aquarium Research Institute. P.O. Box 628,
Moss Lcuuling, California 95039
Abstract. Erythrocytes have long served as model cells, the erythrocytes are typically anucleate discoids, with a
useful for analyzing cytoskeletal structure and function. cytoskeletal system consisting principally of the mem-
In non-mammalian vertebrates, erythrocytes are typically brane skeleton (MS). In adult non-mammalian vertebrates
highly flattened, nucleated ellipsoids in which a marginal and mammalian embryos the mature erythrocytes are nu-
band (MB) of microtubules interacts with the membrane cleated, flattenedellipsoidsordiscoids, and the cytoskele-
skeleton (MS) to generate and maintain cell shape. ton characteristically includes a marginal band (MB) of
Though relatively rare, erythrocytes also occur in repre- microtubules. The MB resists mechanically and osmoti-
sentatives ofmany invertebrate phyla, including the arcid cally induced shape deformation in normal mature cells,
and vesicomyid molluscs, but the structure and function and abnormally pointed MBs generate abnormally
ofthese cells are not well understood. Previous work has pointed cells, presumably by MS deformation (Joseph-
shown arcid erythrocytes to be highly flattened ellipsoids Silverstein and Cohen, 1984). Mechanical interaction be-
containing the MB-MS cytoskeletal system, similar to tween these two cytoskeletal components is thus believed
vertebrates but with an additional interesting feature; a to be important for morphogenesis and maintenance of
functional centriole-containing centrosome associated cell shape in nucleated erythrocytes (Cohen, 1991;
with each MB. In the present study we have examined, Winckler and Solomon. 1991).
forthe firsttime,erythrocyte morphology andcytoskeletal Invertebrates with erythrocytes are relatively rare, but
structure in a vesicomyid. Using Calyptogena kilmeri, the phylogenetically diverse; they include representatives of
dominant invertebrateatmany Pacificcold seeps, wehave the annelids, brachipods, echinoderms. echiurans, mol-
found that the erythrocytes are only slightly flattened and luscs, priapulids. and sipunculans. These invertebrate
do not contain MBs. Rather, their cytoskeletons display a erythrocytes vary with respect to morphology and type
peripheral centriole-containing centrosome with radiating of respiratory protein, and the MB-MS cytoskeletal sys-
fibers, a distinctive type oforganization notobserved pre- tem may or may not be present. The role ofthese cells is
viously in mature erythrocytes from any species. incompletely understood, with several workers suggesting
that erythrocytes may be multifunctional in some species
Introduction (e.g., Terwilliger et al.. 1985). Comparative studies of
cytoskeletal organization in invertebrate erythrocytes
Hemoglobin-bearing erythrocytes are found in all ver- might, therefore, help to shed light on the functions of
tebrates and have long been utilized for studies of cy-
these cells.
toskeletal organization and function. In adult mammals Among the molluscs, erythrocytes appeal" in representa-
tives of the arcid and vesicomyid bivalves (Terwilliger
Received 2 July 1997; accepted 5 December 1997. et al.. 1983; Cohen and Nemhauser,—1985; Nagel, 1985).
*To whom correspondence should be addressed. E-mail: cohen@ Previous work on arcid species including Anadara
genectr.hunter.cuny.edu transversa. A. ovalis. and Noetia ponderosa from coastal
8 W. D. COHEN AND M. N. TAMBURRr
—
watersoftheeastern United States hasdemonstrated that
the MB-MS system is present and functional in mainte-
nance of cell shape (Cohen and Nemhauser, 1980; Nem-
hauser et ai. 1983; Joseph-Silverstein and Cohen. 1984,
MB
1985). Each also has an associated centrosome con-
taining a pairofcentrioles, an unusual feature not observed
in mature erythrocytes of vertebrates (Cohen and Nem-
hauser, 1980; Cohen, 1991). This centrosome remains
—
functional as a microtubule organizing center as demon-
MB
strated by experimentally induced reassembly in the
—
living cells (Nemhauser etai, 1983) and is presumed to
MB
be the same centriole that was involved in biogenesis
during erythrocyte differentiation. Erythrocytes ofthe Aus-
tralian species A. trapezia and several related Japanese
bivalves have a virtually identical centrosome-containing Figure 1. Calyptogena kilmeri. Animals were collected by a re-
cytoskeleton (Ochi, O,. and Cohen, W. D., 1984, and Co- motely operated vehicle (ROV) and maintained in aquaria in a closed
hen, W. D., 1989. unpubl. obs.). indicating that this feature filtered seawater system at 6°C.
is a general characteristic of arcid erythrocytes.
In contrast with the arcids. cytoskeletal structure in
vesicomyid erythrocytes has never been described. Of as "Clam Field" (Barry et ai. 1997b). They were main-
those vesicomyids with erythrocytes, Calyptogena mag- tained forone to several days in laboratory seawatertanks
nifica is best known because it populates well-explored at 6°C, approximately the temperatureoftheirnative habi-
Pacific hydrothermal vent communities (Boss andTurner, tat. The animals were opened with a scalpel, and the
1980; Johnson et ai, 1988b; Fisher et ai, 1988) and bright red. flowing blood (hemolymph) was collected in
because its erythroid hemoglobin has been studied (Ter- a plastic tray and used immediately in experiments. De-
williger et al, 1983). Most (perhaps all) other Calypto- pending on the experiment, hemolymph was collected
gena species with erythrocytes inhabit "cold seeps," sites either at 22°C (room temperature) or at 6°C (coldroom).
that have high levels of hydrogen sulfide, low levels of Erythrocytecytoskeletons were prepared, in general, by
oxygen, and high hydrostatic pressures in common with lysis ofcells with nonionic detergents under microtubule-
hydrothermal vents (McHugh et ai, 1992; Barry et ai. stabilizing conditions. Immediately after collection
1997a; Vrijenhoek et ai, 1994). The principal difference (<30 s) the hemolymph was diluted about 1:20 into Brij
between the two environments is that hydrothermal vents lysis medium or Triton lysis medium. The Brij lysis me-
mM
are characterized by fluctuating and frequently elevated dium contained 100 piperazine-N,N'-bis(ethanesul-
temperatures (Johnson et ai, 1988a), whereas cold-seep fonic acid) [PIPES buffer], 5 niA/ ethylene glycol-bis-
mM
temperatures are similar to those of surrounding waters (/5-aminoethyl ether) n,n'-tetraacetic acid [EGTA], 1
and are relatively constant. Vesicomyid—bivalves inhab- MgClo, pH 6.8, plus 0.6% Brij-58. The Triton lysis me-
itingcold seeps in the Monterey Canyon including Ves- dium was similar except that it contained 0.4% Triton X-
icomya —gigas, V. steamsii, Calyptogena kilmeri, and C. 100 instead of Brij-58. In some cases, 0.1% glutaralde-
pacifica experience a temperature range ofonly 4-6°C hyde was included in the medium for additional rapid
(Barry, J. P., pers. comm.). Of these species, C. kilmeri post-lysis stabilization. These media had been effective
is the most abundant at several cold-seep sites routinely previously for preparing erythrocyte cytoskeletons from
sampled by the remotely operated vehicles (ROVs) ofthe a wide range of vertebrates and invertebrates, including
Monterey Bay Aquarium Research Institute. the archid genera Noetia andAnadara (Nemhauser etai,
Seeking to determine whether molluscan erythrocytes 1983; Cohen and Nemhauser, 1985; Cohen, 1991).
have a common cytoskeletal organization, we undertook
to compare the structure ofvesicomyid erythrocytes with Microscopy
that known for erythrocytes ofthe arcids. We report here
the first examination ofcell morphology and cytoskeletal Cells for morphological examination were either un-
structure in vesicomyid erythrocytes, those of C kilmeri. fixed, or fixed immediately (<30s) by dilution (about
1:20)ofhemolymph into marine molluscanRinger's solu-
Materials and Methods
tion (Cavanaugh, 1975) containing 0.1% glutaraldehyde.
Experimental material Information on cell shape was obtainedby observingcells
Clams (C. kilmeri. Fig. 1) were collected by the ROV while they were stationary in the medium on a slide, as
Ventana from the Monterey Bay cold-seep locale known well as while they were tumbling in flow. Observations
VESICOMYID ERYTHROCYTE CYTOSKELETON
and photomicrographs ofthe cells and theircytoskeletons of uniform size (Fig. 3a-d: arrowheads). In highly flat-
were made with a Zeiss Axioscope equipped with an tened samples, this structure was clearly resolvable as a
Olympus 35-mm camera system with focusing eyepiece, centriole-containing centrosome from which straight fi-
and phase contrast optics including a lOOx Plan-Neofluar brous material radiated (Figure 3e-g). However, circum-
objective (NA 1.3). ferential MBs of microtubules were not present.
Fortransmission electron microscopy ofcytoskeletons, The cytoskeletal structure observed was not induced
fresh hemolymph obtained at 6°C was diluted into Triton by temperature during the experimental procedure. Pe-
lysis medium containing 0.1% glutaraldehyde, incubated ripheral pairs ofcentrioles with radiating fibers were ob-
1 h at about 22°C, stored 3 days at 0°C with glutaralde- served in the cytoskeletons whether the hemolymph was
hyde addedto 1%, andpost-fixed 1 h in 1% OsOabuffered collected at ~22°C or —6°C, or whether lysis was
with 0.1 M KH2PO4-KOH at pH 6.8. Afterethanol dehy- achieved with media at either temperature. We did note,
dration, the material was embedded in Polybed 812 however, that cytoskeletons tended to collapse closer to
(Polysciences, Inc.), thin sectioned with a diamond knife, the nucleus more frequently when thecells were collected
stained with uranyl acetate and lead citrate, and examined and lysed at the lower temperature; thus, our stabilization
with a Hitachi H-600 transmission electron microscope. media were not as effective at that temperature.
Although the paired, phase-dense "dots'" were similar
Results to those observed previously in phase contrast and subse-
quently identified as centrioles in Noetia and Aiuuiara
Cell morphology (Cohen and Nemhauser, 1980, 1985), transmission elec-
tron microscopy of thin sections was used to verify their
The erythrocytes were generally ellipsoidal, but irregu- identification in Calyptogemi. The cytoskeletons were
lar in size and contour (Fig. 2a). Observations made as found to contain classic pairs of centrioles measuring
they tumbled in flow under coverslips showed that the about 0.22 X 0.32 ^m (Fig. 4a-c), with typical 9-triplet
cells were somewhat flattened, but relatively thick (Fig. ultrastructure (Fig. 4d), and microtubules observed fre-
2b, c). The shape of the C. kilmeri erythrocytes was the quently in their vicinity did not emanate directly from the
same whether examined in living cells or in cells fixed centriolar triplets (Fig. 4a-c). In many cases a mass of
immediately upon collection of the hemolymph. electron-dense material was observed in association with
one or both centrioles (e.g.. Fig. 4b, d, arrowheads).
Cytoskeletal structure
Discussion
In different experiments, erythrocyte cytoskeletons
were prepared from hemolymph collected either at room With their generally ellipsoidal and partially flattened
temperature (~22°C) or at cold-seep temperature (cold- irregular shape, the C. kilmeri erythrocytes differed con-
room at ~6°C); the lysis media were also maintained siderably in morphology from those of arcids, which are
eitherat ~22°C or ~6°C. Examinationby high resolution, much flatter and smoother in contour when first removed
phase contrast light microscopy revealed, in most cells, from the animal (Cohen and Nemhauser, 1985). In addi-
a peripheral pair of closely opposed phase-dense "dots"" tion, the shape of the vesicomyid cells remained stable.
Figure 2. Fixed erythrocytes of CalypUit>emi kilmeri, observed by phase contrast microscopy, (a)
Erythrocytes were generally ellipsoidal, but irregular in size and contour; (b, c) face and edge views,
respectively, of the same cell, observed as it tumbled in flow. The erythrocytes are thus found to be
somewhat flattened but relatively thick, and the shape ofunfixed cells is similar (not shown).
10 W. D. COHEN AND M. N. TAMBURRI
^
2.5^m
d
f
Figure3. CytoskeletonsofCiilypt(>i>eniikilnwrierythrocytes; phasecontrast microscopy, (a-d) Exam-
ples ofcytoskeletons in which centrioles are resolved as a pair ofphase-dense "dots" (arrowheads); (e-
g) cytoskeletons flattened under the coverslip to improve visualization of fibers (f) radiating from the
centrosomal region. Marginal bands of microtubules are not present. Conditions: (a) and (e-g). Brij lysis
medium, no glutaraldehyde; (b. cl, Triton lysis medium + glutaraldehyde; (d). Brij lysis medium +
glutaraldehyde.
whereas arcid erythrocytes, if left in their own hemo- erythrocytes, but. in contrast to the arcid cells, the vesico-
lymph. undergo a spontaneous, reversible morphological myid erythrocytes contained no MB. So far as we have
transformation to lumpy spheroids within about 5 min been able to determine, this kind of organization has not
(Sullivan. 1961: Dadacay et al.. 1996). been observed previously in erythrocytes of any other
Cytoskeletal structure in these vesicomyid erythrocytes species. Our survey encompasses annelid, brachiopod,
was distinctive. Fibers radiated from prominent centro- echinoderm, echiuran. molluscan. priapulid. and sipuncu-
somes containing centriole pairs similar to those of arcid lan representatives among the invertebrates (Table 1 ). as
mt f^
.25tJm
Jf'
Figure4. Centrioles incytoskeletonsasobservedby transmissionelectron microscopyofthin sections,
(a-c) Longitudinal and oblique views; (d) cross-sectional view showing 9-triplet cylindrical ultrastructure.
Adjacent microtubules or microtubule bundles (ml) did not emanate from the centriolar triplets (a-c).
Electron-dense material was sometimes observed close to the centrioles (b. d; arrowheads).
VESICOMYID ERYTHROCYTE CYTOSKELETON 11
Table I
Sun'ev ofcywskeletal orgunizauoi] in imertehrate erythrocytes
Phylum.' species
12 W D COHEN AND M. N. TAMBURRI
daughter cell having received one pair. This implies the unlysed bivalve erythrocytes (both arcid and vesicomyid)
existence ofcontrol mechanisms that reprogram or switch traps sufficient hemoglobin to obscure cytoskeletal ele-
centrosome function from mitotic spindle organization to ments in thin sections for transmission electron micros-
erythrocyte morphogenesis and shape maintenance. copy and to render immunofluorescence light microscopy
Among the species in Table I. in addition to C. kilmeri. ineffective. Indeed, blood mustbe withdrawn andthecells
there are five in which erythrocyte MBs are lacking. How- lysed. as in the present work, ifthe entire cytoskeleton is
ever, these species also lack the centrosomal organization to be viewed. Such procedures would be problematic at
found in C. kilmeri erythrocytes, verified by examination the depths and under the challenging physical conditions
of cytoskeletons as in the present work (Nemhauser, characteristic of the cold seeps.
1981). The best studied of these cases are instructive. A second possible source ofartifact would be a sponta-
The hemoglobin-bearing erythrocytes of Pista pacificu. neous change oferythrocyte shape occuiring in the hemo-
Glyceniclibranchiata. and Urevhis caitpo are all spherical lymph shortly after collection, as observed previously in
(Terwilliger et ciL. 1985; Pierce and Maugel. 1985). and arcid erythrocytes (Cohen and Nemhauser, 1985; Dada-
the coelomic hemerythrin-containing cells of Themiste cay et ai, 1996) and in coelomic erythrocytes ofa sipun-
dxscrita are described as varying from disks to spheres culan, T. dyscritu (Terwilliger et a!., 1985). In the present
MB
(Terwilliger et ai. 1985). Absence of the is thus work, this is not a factor. Although, in the absence of
conelated withabsenceorlossofmorphological asymme- priordata, precautions were taken by preparing cytoskele-
MB
try. Conversely. presence is correlated with mainte- tons and fixed cells immediately upon obtaining the he-
nance of marked erythrocyte flatness, as in the Arcidae molymph, such shape alteration was not observed in the
(Cohen and Nemhauser, 1985). The unusual cytoskeletal vesicomyid erythrocytes.
organization of C. kilmeri erythrocytes might then be
viewed as functioning to maintain an equally unusual Acknowledgments
intermediate morphological state, that of limited cell flat-
tening. We thank K. R. Buck for patient technical assistance,
The value oferythrocyte flattening to an organism can and Drs. R. Kochevar and J. P. Barry for their efforts in
be at least twofold. First, reduction ofdiffusion distances making this work possible. We are indebted to the crew
is generally acknowledged to facilitate respiratory gas and pilots of the Pi. Lobos and the ROV Ventana for
exchange, and second, flattening reduces the work re- aid in specimen collection, and to MBARI for providing
quired to maintain "blood" flow by reducing its viscosity workspace and other support. Additional support from
(Fischer, 1978). The latter helps to explain why blood NSF MCB-91 18773 and PSC-CUNY 666180 (WDC) is
cells other than erythrocytes, including mammalian plate- also gratefully acknowledged.
lets, non-mammalian vertebrate thrombocytes, and inver-
tebrate clotting cells such as Limiiliis amebocytes (Arm- Literature Cited
strong, 1985) are also highly flattened. All of these cell
types also contain MBs. Conversely, for organisms with Armstrong, P.B. 1985. Adhesion and motility ofthie blood cells of
spherical erythrocytes, we might speculate that rapid res- Linndus. Pp. 77-124 in BloodCellsofMarineInvertebrates, W. D.
Cohen, ed. Alan R. Liss. New York.
piratory gas exchange and the requirement for reduced
Barry.J.P., H.G. Greene, D.L. Orange, C.H. Baxter. B. H. Robi-
viscosity are not as critical. son. R. E. Kochevar, J.W. Nybakkcn, D. L. Reed, and C.M.
In inteipreting the data, potential artifacts must be con- McHugh. 1997a. Biologic and geologic charactenstics of cold
sidered. Wenotethatthehabitat ofC. kilmeri is character- seep.'i in Monterey Bay, California. Deep-Sea Re.\. 43: 1739-1762.
ized in part by relatively high pressures, with a depth Barry, J.P., R.E. Kochevar, and C.H. Baxter. 1997b. The influ-
range of ~0.5-5 km for cold-seep vesicomyids in gen- ence of pore-water cheinistry and phy.siology in the distribution
of vesicomyid clams at cold seeps in Monterey Bay: implications
eral, and ~0.9 km at "Clam Field" (Orange, D. L., and for patterns of chemosynthetic community organization. Limnol.
Barry, J. P., pers. comm.). Since our findings were Oceanogr. 42: 3IS-,328.
obtained from animals brought to sea level, it might Boss, K.J.,and R.D. Turner. 1980. The giant white clam from the
be hypothesized that arcid-like MBs with associated GalapogosRift.Calyptogenamagnifieaspeciesnovum.Malacologia
20: 161-194.
centrosomes occur in vesicomyid erythrocytes under na- Cavanaugh, G.M. 1975. Formulae and Methods VI. The Manne
tive conditions and become disorganized by the pressure Biological Laboratory. Woods Hole, MA.
reduction. We believe this to be quite unlikely; as shown Clifford, C.H. 1969. Morphological and biochemical features ofthe
in many studies (summarized in Dustin, 1978, 1984), it is coelomocytes of the sea cucumber Molpadia arenicola (Echino-
increasedhydrostatic pressure that can cause microtubule dermata: Holothuroidea). Master's Thesis, Scripps Institution of
Oceanography. University ofCalifornia, San Diego. CA.
disorganization. Data relevant to this question might be Cohen, W.D. 1991. The cytoskeletal system of nucleated erythro-
obtainable by fixing material in situ at the time of speci- cytes. Im. Rev. Cxtol. 130: 37-83.
men collection, but we have found that fixation ofwhole. Cohen, W.D., and 1. Nemhauser. 1980. Association of centrioles
VESICOMYID ERYTHROCYTE CYTOSKELETON 13
with the marginal band of a molluscan erythrocyte. / Cell. Biol. Mattison,A.,andR.Fange. 1973. Ultrastructureoferythrocytesand
86: 286-291. leucocytesofPriapuluscaudatus(DeLamarck)(Priapulida).J.Mor-
Cohen, W.D., and I. Nemhauser. 1985. Marginal bands and the phol. 140: 367-380.
cytoskeleton in blood cells of marine invertebrates. Pp. 1-49 in McHugh, C.M., W.B.F. Ryan, and B. Hecker. 1992. Contempo
BloodCellsofMurineliivertehrates. W. D.Cohen,ed. AlanR.Liss. rary sedimentary processes in the Monterey Canyon-Fan System.
New York. Mar. Geol. 107: 35-50.
Cohen,W.D.,M.F.Cohen,C. H.Tyndale-Biscoe,J.L.VandeBerg, Meglitsch, P.A. 1967. Inrerlebrate Zoology. Oxford University
and G.B. Ralston. 1990. The cytoskeletal system of mammalian Press. New York.
primitiveerythrocytes: studiesindeveloping marsupials. CellMoiil. Nagel, R.L. 1985. Molluscan hemoglobins. Pp. 227-247 in Blood
Cytoskeleton 16: 133-145. CellsofMarine Invertebrates. W. D. Cohen, ed. Alan R. Liss. New
Dadacay, A-V. M., J.C. Huerta, C.J. Theiner, S. Swarnakar, and York.
W.D. Cohen. 1996. Reversible alteration of molluscan erythro-
cyte morphology by a natural hemolymph activity. Biol. Bull. 191: Nemhauser,I. 1981. Marginalbandsystemsinbloodcellsofinverte-
brates. Doctoral dissertation. Graduate Center. City University of
276-277.
New York, New York.
DDuussttiinn,, PP.. 11997848.. MMiiccrroottuubhuullecss.. S(p2rnidngeed.r)-.VeSrplraign,geBre-rVleirnl.ag. Beriin. Nemhauser, I., R. Ornberg, and W.D. Cohen. 1980. Marginal
Fisher, C. R., J.J. Childress, A.J. Arp, J.M. Brooks, D.L. Distel, bands in blood cells of invertebrates. / Ultrasiruct. Res. 70: 308-
J.A. Dugan, H. Felbeck. L.W. Fritz, R. R. Hessler, K.S. John- 317.
son, M.C. Kennicutt II, R.A. Lutz, S.A. Macko, A. Newton, Nenihauser,I.,J.Joseph-Silverstein,andW.D.Cohen.1983. Centri-
M.A.Powell,G.N.Somero,and T.Soto. 1988. Variation in the oles as microtubule organizing centers for marginal bands of mol-
hydrothermalventclam.Calyptogenainagnifieu. atthe RoseGarden luscan erythrocytes. / Cell. Biol. 96: 979-989.
vent on the Galapogos spreading center. Deep-Sea Res. 35: 1811- Ochi, O. 1969. Blood pigments and erythrocytes found in some ma-
1831. rine annelids. Mem. Ehiine Univ. Nat. Sci.. Ser. B (Biol.) VI: 23-
Fischer, T.M. 1978. A comparison of the flow behavior of disc 131.
shaped versus elliptic red blood cells (RBC). BlootI Cells 4: 453- Pierce, S.K., and T. Mangel. 1985. A comparison of the water-
461. regulating responses ofbivalve and polychaete red cells to osmotic
Fontaine, A.R., and P. Lambert. 1973. The fine structure of the stresses.Pp. 167-189inBloodCellsofMarineInvertebrates. W. D.
haemocyte ofthe holothurian Cucunuiria miniata (Brandt). Can. ,/. Cohen, ed. Alan R. Liss. New York.
Zool. 51: 323-332. SulHvan.G.E. 1961. Functionalmorphology,microanatomy,andhis-
Fontaine,A.R.,andB.D. Hall. 1981. The hemocyteoftheholothu- tology ofthe "Sydney cockle". Anadara trapezia (Deshayes) (La-
rian Eupentacta quinquesemita: ultrastructure and maturation. Ciin. mellibranchia: Arcidae). Aiist. J. Zool. 9: 219-257.
J. Zool. 59: 1884-1891. Terwilliger, R.C, N.B. Terwilliger, and A. Arp. 1983. Thermal
Johnson, K.S., J.J. Childress, and C.L. Beehler. 1988a. Short- vent clam [Calxptogenaina.iiiiifica)hemoglobin. .Science 219: 981-
term temperature variability in the Rose Garden hydrothermal vent 983,
Held: an unstable deep-sea environment. Deep-Sen Res. 35: 1711- Terwilliger,N.B.,R.C.Terwilliger,andE.Schabtach. 1985. Intra-
1721.
Johnson, K.S., J.J. Childress, R. R. Hessler, C.M. Sakomoto-Ar- cPepl.lu1la9r3-r2e2sp5iriantoBrlyoopdroCteelilnssooffMSairpiunnecuIlnav.erEtcehbiruatreas..aWn.dDA.nnCeolhiedna,.
ntioolnds,ianntdheC.RoLs.eBGeaerhdleern.h1y9d8r8obt.hermCahlemvienctalhealnd.dGbaiollaopgoigcaolsisnptreeraadc-- ed. Alan R. Liss, New York.
ing center. Deep-Sea Res. 35: 1723-1744. Vrijenhoek, R.C, S.J. Schutz, R.G. Gustafson, and R.A. Lutz.
Joseph-Silverstein.J.,andW.D.Cohen. 1984. Thecytoskeletalsys- 1994. Crypticspeciesofdeep-seaclams(Mollusca; Bivalvia:Vesi-
temofnucleatederythrocytes. 111. Marginal bandfunctioninmature comyidea) from hydrothermal vent and cold-water seep environ-
cells. J. Cell. Biol. 98: 21 18-2125. ments. Deep-Sea Res. 41: 1171-1189.
Joseph-Silverstein,J.,and W.D.Cohen. 1985. Roleofthemarginal Winckler,B.,and F.Solomon. 1991. Aroleformicrotubulebundles
bandinaninvertebrateerythrocyte:evidenceforauniversalmechan- in themorphogenesisofchickenerythrocytes. Proc. Natl. Acad. Sci.
ical function. Can. J. Biochem Cell Biol. 63: 621-630. 88: 6033-6037.