AMER ZOOL, 16:573-591 (1976).
Molluscan Metamorphosis: A Study in Tissue Transformation
DALE B. BONAR.
Department of Zoology, University of Maryland, College Park, Maryland 20742
SYNOPSIS. Metamorphosis of the pelagic larvae of benthic marine invertebrates is often a
cataclysmic event in which a rapid loss of organs specialized for larval life occurs simultaneously with the renewal or increased rate of development of potential adult organs. In the
nudibranch gastropod Phestilla sibogae this change involves loss of the velum, shell, operculum, larval kidney, some retractor muscles, and some of the pedal mucous glands. Exit
from the larval shell at metamorphosis is rapid and is correlated with the spread of
epidermis from the larval foot over the visceral mass as the visceral mass emerges from the
shell aperature. This spreading of epipodial epidermis to cover the entire body has not been
previously reported for other nudibranchs. Neither cell proliferation nor active cell motility
are responsible for this epidermal migration. Rather it appears that the action of larval
muscles pulls the visceral organs out of the shell and simultaneously causes the epipodial
epidermis to cover the visceral mass. This epidermis becomes the definitive adult epidermis.
and potential adult tissues undergo renewed growth and differentiation.
Metamorphosis may be the most critical
Molluscan species which have freepoint in the life history of marine molluscs swimming larval stages show varying dewhich exhibit this phenomenon (Thomp- grees of cataclysmic metamorphosis. For
son, 1958; Fretter, 1969). Since the transi- many prosobranchs (Fretter and Graham,
tion from a pelagic existence to a benthic 1962; Fretter, 1969) and bivalves (Cole,
one involves a change in physical environ- 1938; Hickman and Gruffydd, 1971)
ment and results in substantial changes in metamorphosis involves very little change
an animal's diet, locomotion, and methods in external morphology, and is noticeable
of avoiding predation, changes must only in the loss of the velum (a ciliated
likewise occur in the functional morphol- cephalopedal locomotory and feeding orogy of that animal to accommodate the gan), a modified pattern of shell deposition,
ecological demands of the new environ- and the development of tentacles and funcment and life style. To overcome this vul- tional gills. Internally, there may be further
nerable period of transition, many species changes, such as the development of a
undergo a rather sudden, often termed radula (Fretter, 1969) and final developcataclysmic, metamorphosis in which the ment of adult heart and kindey, along with
larval morphology is quickly lost and the the loss of secondary incorporation of the
adult morphology quickly attained. This ' larval heart and kidney (Fretter and
transition can be accomplished rapidly be- Graham, 1962). But for the most part the
cause of the presence in many larvae of premetamorphic and postmetamorphic
partly or fully differentiated potential adult stages are not radically different. By conorgans, some of which do not function in trast, metamorphosis of nudibranchiate
larval life. The onset of metamorphosis molluscs involves a much greater degree of
thus results in a complex of events in which morphological change since both the shell
uniquely larval tissues are lost or destroyed and operculum are cast off at this time. For
these animals, the elongate, shell-less,
benthic juvenile bears little resemblence,
This work is a portion of a thesis submitted to the
Department of Zoology, University of Hawaii, in par- either superficially or internally, to the
tial fulfillment of requirements for the Ph.D. degree. premetamorphic shelled, pelagic larva.
INTRODUCTION
It was supported in part by NSF Grant GB-36702 to
Dr. Michael Hadfield, whose guidance, criticism and
support I sincerely appreciate.
The most detailed investigations of opisthobranch metamorphosis are those of
573
574
DALE B. BONAR
Thompson (1958, 1962), Rao (1961), of this nudibranch are given elsewhere
Horikoshi (1967), Thiriot-Quievreux (Harris, 1970; Bonar, 1973; Bonar and
(1970), Tardy (1970), Bonar (1973, 1976) Hadfield, 1974), but will be briefly reviewed
and Bonar and Hadfield (1974, 1976). here. Following an embryonic period of
Complete reviews of other, less extensive about 9 days (Hadfield, 1972) lecithotropic
investigations are given in Tardy (1970) larvae hatch from their egg masses and
begin a variable period of planktonic existand Russell (1971).
Metamorphosis of all species of nudi- ence. Larvae must reach a certain point of
branchs with pelagic larvae which have development before they are competent to
been studied to date is a rapid event, the metamorphose and this point is usually
transition from larva to postlarval juvenile reached by the time of hatching. Compebeing completed in a few hours (Bonar and tent larvae will metamorphose only in presHadfield, 1974) to a few days (Thompson, ence of Porites compressa, the adult host, al1962; Tardy, 1970). This abbreviated though presence of the coral mucus in the
period of condensed, or accelerated, de- larval milieu is sufficient to induce
velopment (Berrill, 1971) allows the inves- metamorphosis (Hadfield and Karlson,
tigator to observe a great deal of mor- 1969). Characterization of the metamorphogenesis, both constructive and destruc- phic inducer is currently under way in Dr.
tive, in a short time span.
Hadfield's laboratory (personal communiOne of the most exciting aspects of cation).
metamorphosis in nudibranchs, and many
Metamorphosis of Phestilla sibogae is a
other marine species, is the fact that this sequence of developmental changes which
phenomenon may be induced by exogen- follow a consistant pattern (Bonar and
ous environmental factors. These factors Hadfield, 1974): 1) settlement and temporare usually related to some feature of the ary attachment to the substratum; 2) loss of
preferred adult environment, such as pres- the velum; 3) loss of attachment between
ence of other individuals of the same larval body, shell, and operculum; 4) castspecies, algal or bacterial films, specific ing of shell and operculum; 5) migration of
types of sands or muds, or certain plant or definitive epidermis over the visceral mass;
animal species, often those upon which the 6) rearrangement of internal organs; 7) hismetamorphosing species will feed as an tolysis of parts of the larval musculature
adult (Thompson, 1958, 1962; Hadfield and loss of certain pedal mucous glands; 8)
and Karlson, 1969; also, see Meadows and development of the adult kidney, heart,
Campbell, 1972, for general review). In the tentacles, and cerata; and 9) growth and
absence of the appropriate metamorphic elongation of the postlarval body. Much of
inducer, many species will delay metamor- the information concerning these processes
phosis indefinitely (Thompson, 1958, has been recently published (Bonar and
1962; Bayne, 1965; Bonar and Hadfield, Hadfield, 1974) or is in press or review
1974).
(Bonar, 1976; Bonar and Hadfield, 1976).
The specific, irreversible stimulation of These papers describe the overall sequence
complex morphogenetic events by an ex- of metamorphosis and analyze in detail the
ternal factor offered a dynamic system in structure and loss of both the velum and the
which to study preprogramming of in- contacts between larval body, shell and
duced cell death and the release of de- operculum. The present paper will deal
velopmental arrest on partly differentiated primarily with the mechanism involved in
tissues. With these thoughts in mind, inves- exit from the larval shell, destruction of
tigations were begun on the characteriza- parts of the larval epidermis, and formation and fates of larval and potential adult tion of the definitive adult epidermis at
tissues of the nudibranch Phestilla sibogae. metamorphosis.
The Hawaiian aeolidacean nudibranch
The work of Thompson on dorid (1958,
Phestilla sibogae feeds on the polyps of the 1967a) and tritonid (1962) nudibranchs,
stoney coral Porites compressa. Details of the supported by the recent work of Tardy
ecology, life history, and laboratory culture (1970) on aeolids, has led to the generally
MOLLUSCAN METAMORPHOSIS
RESULTS
accepted view that, as Hyman (1967, p. 497)
notes, "In nudibranchs and other naked
forms the mantle turns back and grows to
The epidermal cell layer which covers the
form the eventual dorsal surface or be- visceral mass and part of the head of Phescomes incorporated in the wall of the vis- tilla sibogae larvae is entirely replaced at
ceral mass." This opposes earlier views metamorphosis by the definitive epidermis
(Fischer, 1880-1887; Garstang, 1890) that (Bonar and Hadfield, 1974). Comparison
the dorsal integument, and consequently of sagittal sections through a premetamorthe papillae developing from this dorsum, phic larva and a postlarva clearly shows the
were derived from epipodial tissues.
dissimilar character of the visceral epiderIn an initial investigation of metamor- mis in the two stages (Fig. 1). Detailed exphosis of Phestilla sibogae, (Bonar, 1972), it amination of the larva reveals two distinct
was also assumed for this species that the types of epidermal tissue: first, the very thin
definitive epidermis of the dorsal surface layer of perivisceral epidermis (also called
was a product of pallial tissues. However, a mantle or shell gland epidermis) which enmore detailed comparison of larval and compasses the larval body within the shell;
postlarval epidermis has revealed a strik- and second, the much thicker epidermis of
ingly different origin of adult dorsal in- the foot and part of the head. An outtegument in this species. Discovery that en- growth of the perivisceral epidermis, from
closure of the body by definitive epidermis the dorsal to the right lateral side of the larat metamorphosis is also more rapid (Bonar val body, forms the mantle fold. The periand Hadfield, 1974) than has been de- visceral epidermis and most of the mantle
scribed for other species led to an investi- fold, which are visible in the light microgation of the possible mechanisms by which
migration of epidermal tissue could be •
brought about.
Thompson (1958, 1962, 1967a) and
Tardy (1970) implicate cellular proliferation as being responsible for the
spread of epidermis over the dorsal surface
of tritonid, dorid and aeolid nudibranchs.
In this study, on Phestilla sibogae, colchicine,
a mitotic inhibitor (Goodman and Gilman,
1970), was added to cultures of metamorphosing larvae to test the possible relationship between cell division and epidermal
migration in this species. Similarly,
cytochalasin B was employed to determine
if its reported action in disrupting active
cellular migration and contraction (Carter,
1967) would affect the spread of tissues
reported here. In the many systems reviewed by Wessels, et al., (1971) contractile
cytoplasmic microfilaments are sensitive to
cytochalasin B and can be reversibly inhibited by very low concentrations of that
drug. An alternate method of tissue migration would be a passive one, involving extracellular forces, such as muscular action.
Several muscular relaxants and paralytic FIG. 1. Near sagittal sections of a) a competent larva,
agents were used to test this last possibility. and b) a postlarva (just after exit from the shell) of
The results are discussed in relation to Phestilla sibogae. fe, foot epidermis; pve, perivisceral
epidermis. The arrows denote the margins of vacuolother developmental systems.
ated epidermis. X340.
576
DALE B. BONAR
scope only as a very dark line (Fig. la), are
composed of squamous cells which are generally less than 0.5 fim in thickness. Only
occasionally are nuclei seen in sections
through this tissue, suggesting that relatively few cells are involved in covering the
viscera. These cells generally have a diffuse
cytoplasm with very few organelles;
mitochondria are sparse and there is no
evidence of endoplasmic reticulum or
Golgi apparatus. Small yolk droplets are
scattered throughout the cytoplasm, and
many small vacuoles are present. The vacuoles are usually less than 0.1 /xm in diameter and appear either to be empty or to
contain a homogeneous, light staining, fine
granular material (Fig. 2). On the floor of
the mantle cavity the perivisceral epidermis
is composed of cuboidal cells which do not
exceed 2.5 to 3.0 fim in thickness, (Bonar
and Hadfield, 1974, Fig. 8). Some of these
cuboidal cells contain cytoplasm which is
denser and more homogeneous than cells
of the rest of the perivisceral epidermis and
nuclei are seen more often in sections
through the cuboidal cell layer (Bonar and
Hadfield 1974, Fig. 8). The only cells of the
perivisceral epidermis which differ significantly in morphology from those described above are the cells which attach the
larval retractor muscles to the larval shell
(Bonar, 1976).
The epidermis of the foot and portions
of the lateral surfaces of the head are quite
different from the perivisceral epidermis.
Head and foot epidermis is composed of
large columnar cells (12 to 15 jam thick)
interspersed with at least two morphologically distinct types of epidermal mucous
glands. The columnar cells have basal nuclei and in the apical cytoplasm are found
large vacuoles, each of which contains a
peculiar dumbbell-shaped inclusion (Fig.
3). These unique vacuoles, of which there
are several per cell, are roughly oblong in
shape with average axes of 3.0 x 1.1 /im,
although shapes ranging from biconcave to
spherical are regularly seen. Most
dumbbell-shaped inclusions are oriented
with the long axis tending toward perpendicular to the cell surface. The columnar
cells otherwise have a highly vesiculated
cytoplasm with abundant mitochondria,
FIG. 2. Electron micrograph of a section through the
visceral wall of a P. sibogae larva. The arrows delimit
the thin layer of squamous perivisceral epidermis typical of that found surrounding the viscera, int, intestine, x 8,760.
sparse Golgi apparatus and endoplasmic
reticulum, and an occasional large yolk
platelet. The apical surface of this cell type
is covered with microvilli, and on the ventral face of the foot, with cilia. Much of the
lateral surface of the foot is unciliated. The
basal lamina of the foot epithelium is quite
dense and can be clearly discerned with the
light microscope. Longitudinal, circular,
and oblique muscle fibers of the
cephalopedal muscle complex insert on the
basal lamina.
Within an hour of loss of the larval shell
at metamorphosis, definitive adult epidermis is present over almost the entire body of
the postlarva. This epidermis has all the
characteristics of those cells originally seen
MOLLUSCAN METAMORPHOSIS
FIG. 3. Electron micrograph of a section through the
foot of a P. sibogae larva. Note the many vacuoles
which contain fibrous inclusions present in the epidermis, x 7,730.
on sides of the larval foot, except that these
cells are thinner (3 to 4 fjum as opposed to
12-15 yu,m on the foot) on the dorsal and
lateral surfaces of the body, and the vacuoles with dumbbell-shaped inclusions are
now oriented with their long axes parallel
to the body surface (Fig. 4). Most of this
epidermis on the lateral and dorsal body
surfaces is unciliated, although occasional
patches or bands are present. On the left
postero-lateral surface of the body there is a
clump of tissue, seen clearly from a dorsal
view. Sections through this clump of tissue
(Fig. 5) show it to be composed of cells
which have the same morphological characteristics as those originally seen in the
perivisceral epidermis and mande fold;
these characteristics include diffuse, vac-
577
uolated cytoplasm and absence of abundant cell organelles. This aggregation of
cells is rather disorganized, as shown by the
buckling of membranes between cells.
Within another 18 to 24 hours, the postlarva loses this clump of cells (whether by
rejection or resorption is not known) and is
entirely covered with definitive epidermis
that is almost indistinguishable, except in
ciliation, from the pedal epidermis.
The two possible origins of definitive
adult epidermis are 1) the transformation
of existing pallial tissues, or 2) the enclosure
of the visceral mass of epipodial tissue migration over the lateral and dorsal sides of
the body. The morphological similarities of
the epidermis of the entire postlarval body
to that of die larval foot make it clear that
the latter mode is the most likely one. If
perivisceral epidermis were indeed transformed, it would require de novo synthesis,
within one hour, of the many large,
inclusion-containing vacuoles, as well as
other organelles such as mitochondria and
cilia, seen in epidermal cells of newly
metamorphosed postlarvae. Also, this
mode of formation would not explain the
origin of the aggregation of cells on the left
postero-lateral side of the early postlarval
body. Since the definitive epidermal cell
type is already present in the larval foot, it is
reasonable to conclude that these cells migrate to enclose the entire body. Sections
through the early postlarva which show the
large vacuoles oriented parallel to the body
surface support the concept that the epidermal layer is moving relative to the inter-
FIG. 4. Electron micrograph of epidermis from the
dorsolateral body wall of a metamorphosing/', sibogae.
Note that the inclusion-containing vacuoles (vac) are
oriented parallel to the body surface, x 7,730.
578
DALE B. BONAR
FIG. 5. a) Light micrograph of a frontal section originally part of the thin layer of squamous perivisthrough a metamorphosing P. sibogae which has just ceral epithelium which is left as a crumpled aggregate
lost its shell, b) An electron micrograph of an area of cells following shell loss, a) x300;b) x 5,610.
similar to the one outlined in "a." These cells were
MOLLUSCAN METAMORPHOSIS
S70
nal organs in the process of enclosure. This glands. The visceral organs are not rigidly
concept also explains the nature of the fixed within the body, but are loosely held
posteriorly situated clump of tissue which in position so that they are able to slide with
resembles morphologically the cells of the relative freedom beneath overlying
larval perivisceral epidermis. During the epidermis. This freedom of visceral moveprocess of epipodial enclosure of the body, ment allows the organs to then squeeze
those cells have been pushed into a mass at through the narrow shell aperture so that
the visceral and cephalopedal elements
the terminal point of migration.
It is clear that the definitive epidermis of merge into a single mass. The shell, no
Phestilla sibogae is epipodial in origin, but longer surrounding the viscera, simply
the question which requires investigation is drops away. Attachment between the operhow the rapid migration of definitive culum and larval retractor muscle is lost at
epidermis about the visceral hump takes about the same time as the shell-retractor
muscle attachment, and when the shell is
place.
Epidermal migration is closely associated cast the operculum is generally lost simulwith casting of the larval shell, both pro- taneously. Uniting the visceral and
cesses beginning with the loss of the at- cephalopedal masses flattens the body; this
tachment of the larval retractor muscles to is accompanied by a slight elongation of the
the larval shell. Subsequently, a series of foot, a process that becomes increasingly
visceral movements and changes in body pronounced with further growth and deshape take place which are diagrammed in velopment. Examination of the epidermis
Figure 6 (also see Fig. 1 in Bonar and at this stage shows that definitive epidermis
Hadfield, 1974). When attachment to the covers one-half to two-thirds of the body, as
shell is lost, the viscera round up inside the represented by the stippled surface in Figshell, but cannot exit immediately from the ure 6d. The back of the foot, or
shell since the visceral diameter is larger metapodium, which was previously covthan that of the shell aperture. At this stage, ered by the operculum, is now covered
the foot of the larva is firmly anchored to with definitive epidermis. From the origithe substratum, presumably by secretions nal point of shell and operculum articulaof the propodial and metapodial mucous tion, the epidermis spreads upward over
the hump, while laterally it migrates dorsally and posteriorly. The arrows from the
leading edge of-the epidermis in Figure 6
show the direction and extent of travel that
this tissue must cover as it encloses the
body. In the hour or so during which enclosure occurs, the visceral mass continues to
flatten and the body becomes more
smoothly rounded. The reorientation of
visceral organs toward their eventual adult
position has been reported previously (Bonar & Hadfield 1974).
FIG. 6. Diagrammatic representation of movements of
the larval musculature and epipodial epidermis during exit from the shell. Stippling represents the area of
the body covered by vacuolated epidermis, lrm, left
larval retractor muscle; oprm, opercular retractor
muscle; rrm, right larval retractor muscle; s, shell.
With these topographical changes to account for, three mechanisms can be proposed which could cause the spread of
definitive epidermis seen at metamorphosis: 1) cellular proliferation; 2) active
cell migration via an intracellular contractile system; or 3) passive movement of the
tissues by forces generated outside the tissues, such as muscular action. These possible mechanisms are individually considered below.
580
DALE B. BONAR
of the foot. As a second check for the possible role of active cellular motility, cytoNo mitotic figures were ever detected in chalasin B, a fungal metabolite which has
metamorphosing larvae; however to test been shown to reversibly inhibit cellular
the possible role of cell division in epider- contractility (Carter, 1967), apparently by
mal migration, several groups of P. sibogae disrupting microfilaments (Wessels, et al.,
larvae were stimulated to metamorphose in 1971), was added to cultures in which P.
the presence of a mitotic inhibitor, col- sibogae larvae were then induced to
chicine. The results of these experiments metamorphose. Initial experiments reare shown in Table 1. Larvae were able to sulted in very low rates of metamorphosis,
metamorphose and undergo epidermal thus, subsequent experiments were conmigration in the presence of even the high- ducted with larvae which had begun metaest concentration used (0.1%), although morphosis (noted by loss of the velum) in
further postmetamorphic development normal sea water and were then rapidly
was arrested. Separate experiments on transferred to appropriate concentrations
early cleavage stages of Phestilla sibogae still of cytochalasin B in sea water. Concentrawithin the egg mass have demonstrated tions of cytochalasin B that ranged from 0.1
that the presence of 0.001% colchicine will to 10.0 /xg/ml were considered effective
block cell division (Hadfield, personal com- dosages for disrupting contractility. The
munication). Assuming postmetamorphic fact that 85% of the larvae of P. sibogae were
growth is dependent on cell division, the able to complete metamorphosis at this
fact that the metamorphosed larvae do not concentration seemingly rules out a
elongate in 0.01% or 0.1% colchicine fur- microfilament-contractile system as being
ther suggests cell division was blocked.
the active force in epidermal migration.
Cellular proliferation
Active cellular motility
Migration motivated by forces outside the
epidermis
Ultrastructural examination of metamorphosing Phestilla sibogae larvae reveal- The third and most plausible mechanism
ed no organized system of contractile mi- proposed for epidermal migration is based
croniameiiLs cither in the thin perivisreral on a duality of muscular movement of
epidermis or in the vacuolated epidermis epidermal tissues and a sliifl in internal organs relative to the position of epipodial
epithelium. A close examination of placeTABLE 1. Treatment of competent" larvae with colchiane.
ment of muscle bundles and fibers during
Percent metamorphosis'1
exit from the shell and spread of definitive
epidermis will aid in clearly understanding
this process.
Colchicine
Colchicine
Colchicine
with coral
cone.
alone
Two systems constitute the larval musculature.
The first consists of the large left
c
0.001%
(75%)
and right larval retractor muscles which in0.01%
(60%)d
sert on the shell posteriorly and on the
0.1%
cephalic
and pedal tissues anteriorly (BoControls
nar and Hadfield, 1974). These muscles are
used primarily to retract the body into the
Percent metamorphosis'1
shell during larval life. The second part of
MPF-SW with coral
the larval musculature consists of a more
MPF-SW
diffuse, subepidermal muscle system which
has longitudinal, circular, and oblique fibers
a
All experiments run with 100 larvae per bowl.
throughout the larval body. These musb
Approximate percentages only.
cle fibers are seen predominantly in the
c
Postlarval growth and elongation very slow as
cephalic
and pedal regions and have been
compared to controls.
referred
to by Thompson (1958) as the
" No postlarval growth or elongation evident.
MOLLUSCAN METAMORPHOSIS
581
cephalopedal muscle complex, although in exit from the shell cannot occur. Figure 8
Phestilla sibogae a few Fibers are present sub- shows a larva which is being held on its side
epidermally in the visceral region as well. at this stage and the arched shape of the
Unfortunately, it is quite difficult to see any foot graphically demonstrates contraction
of these muscle fibers in the living larvae, of at least the left retractor muscle. With the
due in part to the opacity of yolk-rich larval foot well anchored, contraction of the retissues, and, in part to the small size (less tractor muscles, as well as those few subthan 5 /xm in diameter, and often less than 1 epidermal fibers which extend from the
fim) of the individual fibers. Consequently, foot to the viscera, pulls the visceral organs
most of the placement and suggested func- out of the constricting shell. Simultanetion of muscle bundles and fibers reported ously, contraction of the opercular retrachere are extrapolated from viewing these tor pulls definitive epidermis from the posmuscles in sections with light, and to a great terior tip of the foot to cover the back of the
extent, electron microscopy. Figure 7 is a metapodium which is now freed from the
tangential section through the upper part operculum (Fig. 9). The squamous cells on
of the larval foot showing the extensive which this muscle inserts are pulled into the
subepidermal fibers present there. These larval body at the point of original shellfibers are not part of the left larval retrac- operculum articulation.
tor, but are individual muscle fibers which
When the visceral hump is pulled into the
insert on the basal lamina of pedal and shell aperture, it presses against the apercephalic epidermis. Some of the fibers, as tural rim on all sides. Observations of
seen here, insert on the squamous epider- metamorphosing larvae vitally stained with
mal cells which underlie the operculum. neutral red indicate that the epidermis is
These muscle fibers lie in diverse orienta- pinched against the shell rim as the visceral
tions and undoubtedly effect the complex organs are pulled through the aperture
movements of which the foot is capable. In and that the constricture holds the epiderparasagittal sections of larvae, a few muscle mis stationary while the viscera move relafibers are seen passing over the shell lip and tive to it. The apparent movement of visinto the ventral portions of the visceral cera relative to overlying epidermis is repmass (Fig. 7b); these few fibers attach to the resented diagrammatically in Figure 10.
thin perivisceral epidermis in this region, to The basal surface of the columnar epiderthe anterior regions of the digestive diver- mis is scalloped, as seen in a section through
ticula, and probably to the stomach as well. the larval foot (Fig. 7a), giving the impresThe left digestive diverticulum is also at- sion that the cells are tightly packed, or
tached to the right larval retractor which compressed, together. Loss of the constrictpasses just above it.
ing operculum and shell may result in a loss
The apparent mechanism of migration of compression on this colurrirv epiderof definitive epidermis occurs in two parts; mis, so that the cells become more cuboidal
the first, closely associated with exit from as the interior of the foot expands to accept
the shell, is brought about primarily by con- the visceral organs being pulled into it. I
traction of the left and right larval retractor believe that a combination of friction at the
muscles, while the second, consisting of the shell mouth and expansion of previously
enclosure of the visceral hump by epipodial compressed epipodial epidermis aid the
tissue, is effected by the subepidermal mus- initial encroachment of definitive epidercle fibers. Figure 6 diagrammatically repre- mis over the body as the viscera emerge
sents the changing position of muscles, from the shell.
epidermal tissues, shell, and operculum
For the next hour following emergence
during epidermal migration. When muscle from the shell the body continues to flatten
attachment to the shell is lost, contraction of somewhat as the visceral organs assume a
the left and right larval retractor muscles more ventral position. During this hour the
pulls the viscera against the inner margin of contracted larval retractor muscles become
the shell aperture. The foot must be well increasingly disorganized by autolysis.
anchored on a substratum at this time or
The subepidermal muscle fibers which in
582
DALE B. BONAR
FIG. 7. Electron micrographs of tangential sections
through the larval body of P. sibogae, showing the
subepidermal muscle fibers (mus). a) Seaion through
the foot showing the fibers inserting on the oper-
culum. x 2,160. b) Section just below the epidermis
within the mantle cavity at the level of the shell aperature. These fibers pass just subepidermally into the
visceral mass, x 4,350.
MOLLUSCAN METAMORPHOSIS
vf
FIG. 8. Drawing copied from a light micrograph of a
larva during exit from the shell. This individual was
trapped on its side so that the foot had no surface upon
which to obtain anchorage. Apparent contraction of
the retractor muscles is suggested by the arched shape
of the ventral surface of the foot (vf).
the larva were situated in the cephalopedal
region still underlie the same epidermis,
but that epidermis has now assumed a more
lateral position in relation to the visceral
organs. The subepidermal fibers retain diverse orientations so that longitudinal, circular, and oblique bands exist. It appears
that contraction of oblique bands, as shown
in Figure 6d, causes the drawing up of
epipodial epidermis over the lateral sides,
while posteriorly, oblique, circular muscle
fibers begin to act as a contractile ring to
pull the epidermis around the visceral
hump. Finally, vacuolated epidermis encloses the entire body except for the constricted aggregation of remnant cells which
remain at the terminal point of enclosure.
The action of the subepidermal muscles is
not rapid, since fast contraction would result in the central area of a muscle fiber
cutting down into the viscera rather than
causing the epidermis to slide in the plane
of contraction. Tangential sections through
the left side of a metamorphosing larva
fixed near the end of epidermal migration
show the presence of longitudinal and oblique fibers. Electron micrographs of
cross sections through this area show the
small muscle fibers which lie under the
migrating epidermis and insert on the basal
583
lamina (Fig. lla.b). Occasionally, connections are seen in which the basal lamina is
distorted at the point of muscle fiber insertion, suggesting that force is being applied
by that fiber (Fig. 12). It is possible.that this
distortion may be the result of contraction
effects due to fixation; however, the
exerted force seen here is in a posterodorsal direction, as would be predicted if
muscular action were involved.
The attachment of epidermal cells to
basal lamina changes appreciably during
metamorphosis. In the larval foot the basal
lamina of the pedal columnar epidermis is
quite dense and closely associated with
membranes. During epidermal migration,
however, the basal lamina underlying the
migrating epipodial cells is much thinner
and the attachment much more diffuse
(Fig. lla,b). Presumably, the cells remain
attached to the basal lamina and both migrate together, but the reason for the apparently reduced adhesion between cell
membrane and basal lamina is not known.
Posteriorly, the basal lamina which was
originally beneath the perivisceral epidermis is highly convoluted (Fig. 13); apparently as cells of the larval perivisceral
epidermis accumulate in a pile they lose
their attachment to the underlying basal
lamina.
The final enclosure of the postlarval
body at the site of the posterolateral aggregation of remnant cells takes place over the
18 to 24 hours following emergence from
the shell. Although the fate of the remnant
cells is not known, proliferation of epipodial cells now in this area is probably responsible for final enclosure.
The proposed role of muscular action in
exit from the shell and migration of epodial
epidermis was tested by using several
pharmacological agents that may effect
muscular paralysis. Three classes of drugs
were employed: 1) neuromuscular blocks
such as tubocurarin chloride (curare), procaine hydrochloride, hexamethonium
chloride, and succinylcholine chloride; 2)
relaxing agents which apparently affect
ionic balance, such as magnesium chloride
and lanthanum chloride, and 3) anesthetics
such as chlorobutanol (chloretone) whose
pharmacological action is uncertain
584
DALE B. BONAR
FIG. 9. Electron micrograph of a sagittal section
through a metamorphosing P. sibogae following lossof
the operculum. This micrograph shows the area
where the shell and operculum originally articulated
(asterisk). The shell was still present when the speci-
men was Fixed but was decalcified prior to embedding.
Note the vacuolated foot epidermis (fe) which has
covered the surface previously occupied by the operculum. vm, visceral mass, x 4,580.
(Goodman and Gilman, 1970). Muscular
activity was considered blocked when animals no longer contracted in response to
sharp taps on the culture vessel or to prodding with a fine needle. Appropriate concentrations of the various drugs were determined by preliminary tests over a wide
range, and final experiments were conducted at concentrations around those at
which paralysis had been seen. The results
are presented in Table 2.
Different drugs varied considerably in
degree of paralysis and in length of time
necessary to promote paralysis. Chlorobutanol effectively blocked both muscular action and metamorphosis at any stage within
5 to 10 minutes, whereas tubocurarin
chloride was progressive, requiring several
hours before producing partial paralysis.
Procaine hydrochloride appeared to produce an all-or-nothing effect. In concentrations of 0.01 % procaine, larvae were seen to
metamorphose normally, whereas at con-
centrations above 0.01%, virtually all animals were quickly paralyzed in a contracted
state and died within 24 hours. When concentrations of succinylcholine or tubocurarin were adjusted so as to cause rapid
paralysis, animals were invariably paralyzed in a contracted or semicontracted
state, and death followed within 4 to 24
hours. At these concentrations animals
could not be revived. Only with chlorobutanol could rapid paralysis be reversed and
metamorphosis subsequently be completed.
DISCUSSION
Derivation of the dorsal integument
from the cells of the mantle, as has been
shown for several nudibranch species, does
not occur in Phestilla sibogae. In briefly describing metamorphosis of five nudibranch
species which have a developmental pattern similar to that of P. sibogae, Tardy
MOLLUSCAN METAMORPHOSIS
FIG. 10. Diagrammatic representation of the movement of visceral elements (vm) in relation to the
epipodial epidermis during exit from the shell (s).
Figures 10 A-D are equivalents of sections oriented as
shown in the uppermost drawing. Asterisks show the
original alignment positions of visceral mass and
perivisceral epidermis and their respective movements during exit from the shell.
585
cavity floor prior to metamorphosis. This
figure, a parasagittal section of a larva of
Teneillia ventilabrum, is almost identical to a
similar section through a Phestilla sibogae
larva, but as shown in the present study,
none of the pallial elements contribute to
the definitive epidermis in P. sibogae. Cells
of the mantle cavity floor are not sufficiently thick or extensive to produce the
definitive epidermis in as short a time as
occurs in P. sibogae, nor are these cells morphologically similar to the definitive epidermal cells. A reexamination with the electron microscope of the five species observed by Tardy would establish whether
the adult epidermis is a derivative of the
epithelial floor of the mantle cavity or is, as
in P. sibogae, derived from the lateral surfaces of the foot.
Of particular interest to the question of
pallial versus epipodial origin of dorsal
epidermis is the work of Herdman and
Clubb (1891) on innervation of the nudibranch dorsum. Their investigation revealed that nerves from the pleural ganglia
innervated papillae on the dorsal body surface of tritonid and two dorid species,
whereas in Eolis, the single aeolid they
studied, nerves from the pedal ganglia
were found to innervate the cerata. The
authors suggested that these results reflected the epipodial nature of the dorsal
surface ofEolis, and the pallial nature of the
tritonid and dorid dorsum. Later work by
Russell (1929) on Aeolidia papillosa revealed
that the cerata of this aeolid were innervated by nerves from the pleural ganglion.
Unfortunately, the origin of adult dorsal
epidermis has not been described for
Aeolidia papillosa, but on the basis of the
pleural innervation, it is interesting to
speculate that, like Aeolidiella alderi (Tardy,
1970), the dorsal epidermis would have a
pallial origin.
(1970, pp. 339-340) notes that the mantle
Cell proliferation is the mechanism by
and "bourrelet palleal" are involved in
forming the larval shell and do not evert to which most of the epidermal migration in
cover the visceral hump. He suggests, how- other species is reported to occur. The cells
ever, that elements of the "bourrelet pal- in the "bourrelet palleal" of Aeolidiella alleal" slide across the floor of the mantle deri, as a result of localized cell division,
cavity where they multiply and thus form spread posteriorly over the viscera as the
the definitive epidermis. One of his figures shell is lost (Tardy, 1970).
(15a, p. 340) shows what he considers to be
Apparently, the epidermal migration
definitive epidermis covering the mantle seen in Phestilla sibogae is not produced by
TABLE 2. Treatment of competent and metamorphosing larvae with muscular relaxants and paralytics.'
DRUG
Cone.
Curare
0.1 mg/ml
1.0 mg/ml
Drug alone
Drug with Degree of
coral
paralysis
+ (85%)
+*(75%)
10.0 mg/ml
Procaine
hydrochloride
partial
total,
contracted
0.01%
0.1%
+*(40%)
slight
total,
contracted
total,
contracted
Remarks
normal metamorphosis.
slightly abnormal after metamorphosis, but
covered with definitive epidermis.
all die within 24 hours.
all die within 24-48 hours.
50% die within 5 hours.
all die within 8 hours.
Succinylcholine
chloride
0.01%
0.1%
0.5%
1.0%
25 min.
60 min.
+ (90%)
+(2-42%)
+i
+(22%)
slight
slight
total,
contracted
normal metamorphosis,
normal metamorphosis, a few larvae die.
develop abnormally after metamorphosis,
die within 24 hours.
The following were "pulsed" in l.i 7c succinylcholine for times shown:
contracted during recover to normal appearance,
+ (18%)
n.a.
treatment
contracted during recover to normal appearance,
+(10%)
n.a.
treatment
Hexamethonium
chloride
0.01%
0.01%
1.0%
+ (90%)
+ (85%)
+ (60%)
none
none
none-slight
normal metamorphosis.
normal metamorphosis.
some nonmetamorphosing larvae
lost attachment to the shell.
Chlorobutanol (Chloretone)
Saturated in distilled water, diluted 1 3 with MPF-SW
1. Effectively stops metamorphosis at any stage within
2. Reversible effects if larvae removed within 30 min.
00
(Approx. 0.24%, according to Cavanaugh, 1956).
5-10 minutes. Paralysis total.
r
W
O
>
MOLLUSCAN METAMORPHOSIS
3
3
o
s
o
c '"p
o o
V CO
1
I
c
c
u
3
o-
o
s.
s -E
a.
u
ra
T3
c
o
-a
•»
ra
o
i
o
i I
•3
_3
o
1.
£
-5
s
TJ
O
ra
•5
2
v
c
—
n "P5
a - I 8
1 1=3- 8 1
-a
«
~
Q
—-
c^i
en
o o ^ v ^~
E
fe I
S«
p e r t " be
« _ E o-S
S « o a c
c P 2 *J ^
-
-
II
II
11
»
—
•
=
1 + + c.
587
cell division, as shown by the experiments
with colchicine. The highest concentrations
of the drug used in this study, 0.1%, is a very
high dosage, on the order of that used to
disrupt microtubules in organisms which
have cuticles that greatly restrict permeability, such as in the heliozoan Actinosphaerium
nucleophilum (Tilney,1968). Larvae of Phestilla sibogae have no apparent restrictions to
permeability, yet were able to complete
epidermal migration even at that high dosage. The fact that further postlarval growth
and elongation in 0.1 % or 0.01 % colchicine
did not occur indicates that cell division had
indeed been halted.
Active cellular motility or contractility
has been shown in many systems to be related to an intracellular, contractile, microfilamental system, such as has been visualized in widely divergent phenomena
(amoeboid movement, Komnick, Stockem,
and Wolfarth-Botterman, 1970; sea urchin
cleavage, Schroeder, 1972; ascidian tadpole tail absorption, Cloney, 1972; and
other systems reviewed by Wessels, et al.,
1971). Cytochalasin B concentrations of
0.25 to 0.50 Mg/ml have been reported to
rapidly block sea urchin cleavage
(Schroeder, 1972) and tail resorption of ascidian tadpoles (Cloney, 1972). Concentrations of 10 /i,g-/ml of the drug have been
shown to cause toxic effects in mouse
L-cells (Carter, 1967), several fibroblast and
epithelial cell lines (Estensen, 1971),
Xenopus laevis eggs (Estensen, Rosenberg,
and Sheridan, 1971) and ascidian tadpoles
(Cloney, 1972). Together, these results
suggest that concentrations of 1.0 /xg/ml
should be completely effective in blocking
cell migration mediated by an intracellular
contractile system. The fact that 85% of the
Phestilla sibogae larvae in this concentration
of cytochalasin B successfully metamorphose indicates that no intracellular contractile system is involved in this process. Of
peripherai interest is the apparent ability of
cytochalasin B to block the induction of
metamorphosis, even at low concentrations. Since cytochalasin B has been shown
to effect transport of glucose and glucosamine across membranes of cultured cells,
apparently by simple competition (Estensen and Plagmann, 1972), it may in some
DALE B. BONAR
FIG. 11. Electron micrographs of sections through the after shell loss. Note the close relationship between
dorsolateral epidermis of a postlarval P. sibogae just musde fibers (mus) and overlying epidermis, x 5,880.
similar way effect the induction stimulus in
P. sibogae.
Thompson (1958, 1962) notes that during the obligatory swimming phase in
Adalaria proximo, and Tritonia hombergi, muscular elements develop in the thickened
mantle and run from it down into the foot
on both sides of the body. He does not state
MOLLUSCAN METAMORPHOSIS
589
Until the discrepancies in results produced
by these reportedly similar drugs can be
explained, interpretation of those results
must be with due caution. The fact that
these drugs reduced the rate of metamorphic induction suggests there are hidden
side effects for which we cannot account.
Mention should be made of the anomolous artificial induction of metamorphosis
seen in some larvae treated with succinyl
choline chloride (Table 2). As noted earlier, Phestilla sibogae larvae will normally
metamorphose only in the presence of Porites compressa, the coral upon which adult
Phestilla feed. Treatment with 0.1% suc-
•*V»?vtfWB$s
.
FIG. 12. Electron micrographs of a section through
the dorsolateral epidermis of a metamorphosing P.
sibogae. Muscle fibers (mus) are seen connecting to the
basal lamina (bl) of epidermal cells and appear to be
exerting a pulling force on those cells, x 20,300.
whether these muscles ever function in the
spread of the mantle over the body to form
definitive epidermis.
It appears that for Phestilla sibogae the
processes of exit from the shell and migration of definitive epidermis are based almost entirely on muscular action, initially
with the aid of the left and right larval retractors, then finally by action of the subepidermal nuscle fibers. Attempts to experimentally block muscle action indicate
that when contractility is blocked, exit from
the shell and epidermal migration are also
blocked. Because the total effect which
these drugs have on invertebrates is not
known, a rigorous interpretation of these
results is not possible. The neuromuscular
blocks used here are known, from their action in mammalian physiology, to compete
with acetylcholine or to block its release at
neuromuscular junctions (Goodman and
Gilman, 1970), but one must be careful in
extrapolating to invertebrate tissues.
Beeman (1968), investigating anesthetics
for opisthobranchs, found that curare had FIG. 13. Electron micrograph of a section of a
very little effect on adult Aplysia californica, metamorphosing larva showing the crumpled basal
while succinylcholine, with a similar phar- lamina (bl) beneath the posterior aggregate of perivismacology (Goodman and Gilman 1970), ceral epidermal cells following shell loss. Note the lack
attachment between the basal lamina and adjacent
apidly and effectively caused paralysis. of
epidermal cells, x 13,400.
590
DALE B. BONAR
cinylcholine (in the absence of coral) was
found to induce between 2 and 42%
metamorphosis of Phestilla larvae. Although its basis of action in this instance is
not known, one would expect some interaction with the larval nervous system by which
the various events of metamorphosis could
be integrated. Further studies on the action
of various neurotransmitters and analogs
are in progress.
REFERENCES
Bayne, G. L. 1965. Growth and delay of metamorphosis of the larvae of Mytilus edulis L. Ophelia 2:147.
Beeman, R. D. 1968. The use of succinylcholine and
other drugs for anesthetizing or narcotizing gastropod molluscs. Pubbl. Staz. Zoot. Anpoli 36:267270.
Berrill, N. J. 1971. Developmental biology. McGraw-Hill,
New York.
Bonar, D. B. 1972. Fate of larval organs at metamorphosis in a gastropod. Amer. Zool. 12:722.
Bonar, D. B. 1973. An analysis of metamorphosis in
Phestilla sibogae Bergh 1905 (Gastropoda, Nudibranchia). Doctoral Diss., University of Hawaii.
Bonar, D. B. 1976. Metamorphosis of the marine gastropod Phestilla sibogae (Aeolidacea, Nudibranchia).
II. Hemidesmosomes and cytoplasmic filaments in
attachments between larval body, shell, and operculum. Tissue and Cell. (In press)
Bonar, D. B. and M. G. Hadfield. 1974. Metamorphosis <~>f the marine gastropod Phestilla sibogae
(Aeolidacea, Nudibranchia). I. Light ami dcclrcr.
microscopic analysis of larval and metamorphic
stages. J. Exp. Mar. Biol. Ecol. 16:227-255.
Bonar, D. B. and M. G. Hadfield. 1976. Metamorphosis of the marine gastropod Phestilla sibogae
(Aeolidacea, Nudibranchia). III. Velar morphology
and destruction of the velum at metamorphosis. (In
press)
Carter, S.B.I 967. Effects of cytochalasins on mammalian cells. Nature (London) 21B:261-264.
Cloney, R. A. 1972. Cytoplasmic filaments and morphogenesis: effects of cytochalasin B on contractile
epidermal cells. Zeit. Zellforsch. Mikrosk. Anat.
132:167-216.
Cole, H. A. 1938. The fate of the larval organs in the
metamorphosis of Ostrea edulis. J. Mar. Biol. Assn.,
U.K. 23:469-484.
Estensen, R. D. 1971. Cytochalasin B. I. Effect on
cytokinesis in Novikoff hepatoma cells. Proc. Soc.
Exp. Biol. Med. 136:1256-1258.
Estensen, R. D. and P. G. Plagemann. 1972. Cytochalasin B: Inhibition of glucose and glucosamine transport. Proc. Nat. Acad. Sci., U.S.A. 69:1430-1434.
Estensen, R. D., M. Rosenberg, and J. D. Sheridan.
1971. Cytochalasin B: Microfilaments and "Contractile processes." Science 173:356-357.
Fischer, P. H. 1880-1887. Manuel de Conchyologie et de
Paleontologie Conchyologique. Paris, Savy.
Fretter, V. 1969. Aspects of metamorphosis in prosobranch gastropods. Proc. Mala. Soc, London
38:375-386.
Fretter, V. 1972. Metamorphic changes in the velar
musculature, head and shell of some prosobranch
gastropods. J. Mar. Biol. Assn., U.K. 52:161-178.
Fretter. V. and A. Graham. 1962. British prosobranch
molluscs. Ray Society, London.
Garstang, W. 1890. A complete list of the opisthobranchiate Mollusca found at Plymouth. J. Mar. Biol.
Assn., U.K. 1:399-457.
Goodman, L. and A. Gilman. 1970. The pharmacological
basis of therapeutics. 4th Ed. MacMillan, New York.
Hadfield, M. G. 1972. Flexibility in larval life patterns.
Amer. Zool. 12:721.
Hadfield, M. G. and R. Karlson. 1969. Externally induced metamorphosis in a marine gastropod.
Amer. Zool. 9:317.
Harris, L. G. 1970. Studies on the biology of the
aeolid nudibranch Phestilla melanobranchia Bergh
1874. Doctoral Diss., University of California,
Berkeley.
Herdman, W. A. and J. A. Clubb. 1891. On the innervation of the epipodial processes of some nudibranchiate Mollusca. Rept. Brit. Ass. Adv. Sci.
(1891):692-693.
HickmaYi, R. W. and LL. D. Gruffydd. 1971. The histology of the larva of Ostrea edulis during metamorphosis. In: E. J. Crisp (ed.) Fourth european marine
biology symposium, pp. 281-294. University Press,
Cambridge.
Horioshi, M. 1967. Reproduction, larval features, and
life history of Philine denticulata. Ophelia 4:42-84.
Hymani L. H. 1967. The invertebrates, Vol. VI. Mollusca
I. McGraw-Hill, New York.
Komnick, H., W. Stockem, and K. E. WohlfarthRottermann. 1970. Weitreichende fibrillare protoplasmadifferenzierungen unu ilnc Ecdeutung fur
die Protoplasmastromung. VII. Experimented Induktion, Kontraktion, und Extraktion der Plasmafibrillen von Physarum polycephalum. Zeit.
Zellforsch. Mikrosk. Anat. 109:420-430.
Meadows, P. S. and J. I.Campbell. 1972. Habitat selection by aquatic invertebrates. Adv. Mar. Biol.
10:271-382.
Rao, K. V. 1961. Development and life history of a
nudibranch gastropod Cuthona adyarensis Rao. J.
Mar. Biol. Assn., India 3:186-197.
Russell, H. D. 1971. Index nudibranchia 1554-1965.
Delaware Mus. Nat. Histo. Greenville, Delaware.
Russell, L. 1929. The comparative morphology of the
elysioid and aeolidoid types of the molluscan nervous system and its bearing on the relationships ot the
ascoglossan nudibranchs. Proc. Zool. Soc. London
14:197-233.
Schroeder, T. 1972. The contractile ring. II. Determining its brief existence, volumetric changes and
vital role in cleaving Arbacia eggs. J. Cell Biol.
53:419-449.
Tardy, J. 1970. Contribution a l'etude des
metamorphoses chez les nudibranches. Ann. Sci.
Nat., Zool. Paris 12:299-370.
MOLLUSCAN METAMORPHOSIS
591
nudibranch Cadlina laevis, with a discussion on deThiriot-Quievreux, C. 1970. Transformations histologiques lors de la metamorphose chez Cymbulia
velopmental processes in Opisthobranchia. J. Mar.
Biol. Assn., U.K. 47:1-22.
peroni de Bainville (Mollusca, Opisthobranchia).
Zeit. Morph. Tiere 67:106-117.
Tilney, L. G. 1968. Studies on the microtubules in
Heliozoa. IV. The effect of colchicine on formation
Thompson, T. E. 1958. The natural history, embryoland maintainence of the axopodia and the redeogy, larval biology, and post-larval development of
velopment pattern in Actinosphaerium nucleofilum
Adalaria proxima (Alder and Hancock) (Gastropoda,
Opisthrobranchia). Phil. Trans. Roy. Soc, London
(Barrett). J. Cell. Sci. 3:549-562.
B 242:1-58.
Wessels, N. K., B. S. SpoonerJ. F. Ash, M. O. Bradly,
Thompson T. E. 1962. Studies on the ontogeny of
M. A. Luduena, E. L. Taylor, J. T. Wrenn, and K. M.
Tritonia hombergi Cuvier (Opisthobranchia). Phil.
Yamada. 1971. Microfilaments in cellular and deTrans. Roy. Soc, London B 24.:171-218.
velopmental processes. Science 171:135-143.
Thompson, T. E. 1967. Direct development in a