Vestigial prototroch in a basal nemertean, Carinoma tremaphoros
(Nemertea; Palaeonemertea)
S. A. Maslakova,a,c,
 M. Q. Martindale,b and J. L. Norenburga
aSmithsonian Institution, National Museum of Natural History, MRC 163, P.O. Box 37012, Washington, DC 20013-7012, USA
bKewalo Marine Laboratory, University of Hawaii, 41 Ahui Street, Honolulu, HI 96813, USA
cDepartment of Biology, George Washington University, 2023 G Street NW, Washington, DC 20052, USA

Author for correspondence (email: maslak@gwu.edu)
SUMMARY Nemerteans have been alleged to belong to a
protostome clade called the Trochozoa that includesmollusks,
annelids, sipunculids, echiurids, and kamptozoans and is
characterized by, among other things, the trochophore larva.
The trochophore possesses a prototroch, a preoral belt of
specialized ciliary cells, derived from the trochoblast cells.
Nemertea is the only trochozoan phylum for which presence of
the trochophore larva possessing a prototroch had never been
shown. However, so little is known about nemertean larval
development that comparing it with development of other
trochozoans is difficult. Development in the nemertean clade
Pilidiophora is via a highly specialized planktonic larva, the
pilidium, and most of the larval body is lost during a drastic
metamorphosis. Other nemerteans (hoplonemerteans and
palaeonemerteans) lack a pilidium, and their development is
direct, forming either an encapsulated or planktonic
??planuliform?? larva, producing a juvenile without a dramatic
change in body plan. We show that early in the development
of a member of a basal nemertean assemblage, the pal-
aeonemertean Carinoma tremaphoros, large squamous cells
cover the entire larval surface except for the apical and
posterior regions. Although apical and posterior cells continue
to divide, the large surface cells cleavage arrest and form a
contorted preoral belt. Based on its position, cell lineage, and
fate, we suggest that this belt corresponds to the prototroch of
other trochozoans. Lack of differential ciliation obscures the
presence of the prototroch in Carinoma, but differentiation of
the trochoblasts is clearly manifested in their permanent
cleavage arrest and ultimate degenerative fate. Our results
allow a meaningful comparison between the development of
nemerteans and other trochozoans. We review previous
hypotheses of the evolution of nemertean development and
suggest that a trochophore-like larva is plesiomorphic for
nemerteans while a pilidium type of development with drastic
metamorphosis is derived.
INTRODUCTION
Nemerteans (ribbon worms) belong to one of the least studied
bilaterian phyla (Turbeville 2002). Some controversy still ex-
ists in the literature as to the phylogenetic af?nities of ne-
merteans. Although some authors traditionally view them as
related to platyhelminthes based on similarity in structure and
origin of the nervous system and larval ciliated bands (Nielsen
2001), others consider them related to coelomate animals,
such as annelids and mollusks, based on ultrastructural sim-
ilarity of the nemertean rhynchocoel and blood vessels to the
coelomic cavities (Turbeville and Ruppert 1985; Turbeville
1986). DNA sequence data support the coelomate af?nities of
nemerteans and place them within the clade of coelomate
protostomes, often referred to as the Trochozoa, which in-
cludes animals with such diverse adult body plans as annelids,
mollusks, sipunculids, echiurids, and kamptozoans (Turbe-
ville et al. 1992; Peterson and Eernisse 2001; Turbeville 2002).
Traditionally, the Trochozoa united animals that possess
spiral cleavage and trochophore-type larvae. Although the
core composition of the Trochozoa (annelids, mollusks, sip-
unculids, echiurids, and kamptozoans) remained undisputed
over the last few decades, various authors argued for the
inclusion of phoronids, bryozoans, and arthropods (Beklemi-
shev 1964) or platyhelminthes and nemerteans (Ivanova-
Kazas 1985). Rouse (1999) suggested an apomorphy-based
taxon, Trochozoa, that includes the ?rst ancestor to have
evolved a prototroch and all its descendants. Based on this
definition, the Trochozoa encompasses annelids (sensu lato),
echiurids, sipunculids, mollusks, and kamptozoans. Here we
adopt the more recent scope of the Trochozoa based on the
combined phylogenetic analysis of morphological and molec-
ular data (Peterson and Eernisse 2001). These authors apply
the name Trochozoa to a clade comprising annelids, echiu-
rids, sipunculids, mollusks, kamptozoans, and nemerteans.
The only difference in terms of the scope of the Trochozoa is
that Rouse?s Trochozoa does not include the Nemertea.
However, if nemerteans are shown to possess the prototroch
EVOLUTION & DEVELOPMENT 6:4, 219?226 (2004)
& BLACKWELL PUBLISHING, INC. 219
(as is the conclusion of this manuscript), they become part of
the Trochozoa by Rouse?s definition as well. All trochozoans
possess spiral cleavage. However, not all spiralians are cur-
rently included in the Trochozoa; the Platyhelminthes (some
of which possess typical spiral cleavage) appear to be sepa-
rated from the Trochozoa by a number of nonspiralian taxa,
such as Lophophorata (Phoronida and Brachipoda), Cyclyo-
phora, Acanthocephala, Rotifera, Gnathostomulida, and
Ectoprocta (Peterson and Eernisse 2001).
The trochophore larva is the ?rst swimming stage of the
Trochozoa. The preoral belt of specialized ciliary cells, called
the prototroch, characterizes the trochophore larva and is
derived from specific founder cells, called trochoblasts. The
prototroch is the primary locomotory organ of the trocho-
phore larva and is derived from subsets of the same cell lineages
across all trochozoan taxa. As emphasized by Rouse (1999,
2000), a more strictly de?ned trochophore, with prototroch,
metatroch, gastrotroch, telotroch, a pair of protonephridia,
and a downstream mechanism of larval ciliary feeding (Ni-
elsen 1995), is only present in a few annelid taxa. Morpho-
logical cladistic analysis (Rouse 1999) suggests that metatroch
and other ??trochs?? might have evolved multiple times inde-
pendently within the Trochozoa, whereas the prototroch had
evolved only once. Here we adopt the broad definition of the
trochophore (Rouse 1999) based on the presence of the pro-
totroch, derived from the trochoblast cell lineage. Upon de-
velopment of a functional gut, the prototroch sometimes
becomes involved in feeding. It is often suggested that the
prototroch?s function in downstream larval feeding is
plesiomorphic for all trochophore larvae (Nielsen 2001). In
contrast, recent cladistic evidence (Rouse 2000) suggests that
the prototroch evolved primarily as a locomotory organ and
became involved in feeding secondarily on several occasions
during the radiation of the Trochozoa.
The central cells of the prototroch are referred to as pri-
mary trochoblasts and derive from the four vegetal daughters
of the ?rst quartet micromeres from each primary cell quad-
rant (A?D) at the 16-cell stage. These cells are designated as
1q2 micromeres in the classical nomenclature of spiral cleav-
age (Wilson 1892, modi?ed by Child 1900). A unique feature
of the primary trochoblasts is their ability to autodifferentiate.
Isolated primary trochoblasts of annelids and mollusks follow
the normal course of development observed in the intact em-
bryo. They undergo one or two divisions, become cleavage
arrested, and differentiate as ciliated cells (Wilson 1904; Cos-
tello 1945). Additional cells derived from the animal daugh-
ters of the ?rst quartet micromeres (called accessory
trochoblasts, derived from the 1q1 micromeres) and the sec-
ond quartet micromeres (called secondary trochoblasts, de-
rived from the 2q micromeres) also contribute to the
prototroch, but they require additional cell?cell interaction
for their differentiation (Damen and Dictus 1994a). In all
trochozoans, the prototroch is a transitory larval structure
that is shed or reabsorbed during development and does not
contribute to the adult body plan (Buckland-Nicks et al.
2002a,b; Jaeckle and Rice 2002; Nielsen 2002; Pernet et al.
2002; Pilger 2002; Zardus and Martel 2002). Although most
authors assume homology of trochophore larvae throughout
the Trochozoa (Nielsen 2001), some consider appearance of
the prototroch-bearing larvae in various taxa to be the result
of parallel evolution, driven by the functional requirements of
a swimming larva and constrained by the stereotypic nature
of spiral cleavage (Ivanova-Kazas 1986). Morphological clad-
istic evidence supports homology of the prototroch in the
Trochozoa and thus the trochophore larva, as broadly de?ned
by the presence of the prototroch (Rouse 1999).
Nemertea is the only trochozoan phylum for which pres-
ence of a trochophore larva (with prototroch derived from
trochoblasts) had never been shown. There also are no clearly
recognized homologies between nemertean larvae and a typ-
ical trochophore (Norenburg and Stricker 2002), although
many authors compare the ciliated band of the nemertean
pilidium larva to the prototroch (McBride 1914; Lacalli and
West 1985; Nielsen 2001). Peterson and Eernisse (2001) in
their morphological analysis of metazoan phylogeny coded
nemerteans as possessing a trochophore, basing this on Ni-
elsen?s (1995) inference that trochoblasts underlie the ciliary
band in pilidium larvae of heteronemerteans (Nielsen 1995).
This particular hypothesis of homology was later refuted by
cell lineage data for a pilidium larva (Henry and Martindale
1998). Nemerteans display a variety of developmental modes,
the best studied of which is the development of the pilidium
(Salensky 1886; Dawydo? 1940; Cantell 1969; Lacalli and
West 1985; Henry and Martindale 1998 and references there-
in). The juvenile worm develops from a series of ectodermal
invaginations (imaginal discs) and hatches out of the larval
body in a drastic metamorphosis in which most larval ecto-
dermal and mesodermal tissues are destroyed. Members of the
Hoplonemertea and the Palaeonemertea have direct develop-
ment where either a juvenile or a nonspecialized uniformly
ciliated ??planuliform?? planktonic larva hatches out of the egg
envelopes (Delsman 1915; Hammarsten 1918; Iwata 1960;
Ja?gersten 1972; Stricker and Reed 1981; Martindale and
Henry 1995; Maslakova and Malakhov 1999; Norenburg and
Stricker 2002). The development of the planuliform larvae of
palaeo- and hoplonemerteans is considered to be ??direct??
because the larva, where present, does not possess a dis-
tinctly different body plan from that of the juvenile and
adult. However, many ??planuliform?? larvae of hoplo- and
palaeonemerteans possess larval structures that do not con-
tribute to the juvenile, such as eyes, apical organ, transitory
larval ectoderm, and ciliary bristles or cirri (Maslakova and
Malakhov 1999; Chernyshev 2000; Norenburg and Stricker
2002 and references therein). The traditional view is that the
pilidium is derived within nemerteans, whereas the develop-
ment with a nonspecialized planuliform larva is ancestral
220 EVOLUTION & DEVELOPMENT Vol. 6, No. 4, July^August 2004
(Ivanova-Kazas 1985; Norenburg and Stricker 2002). A re-
cent molecular phylogeny of nemerteans (Thollesson and
Norenburg 2003) suggesting that palaeonemerteans represent
a basal paraphyletic assemblage, whereas heteronemerteans
(together with nonheteronemertean species possessing pilidia)
and hoplonemerteans represent monophyletic sister groups
supports this hypothesis (Fig. 1). However, evidence for a
transitory larval ectoderm that is lost or reabsorbed during
development of ??planuliform?? larvae, in several hoplonemer-
tean (Delsman 1915; Reinhardt 1941; Hickman 1963; Ma-
slakova and Malakhov 1999) and palaeonemertean (Iwata
1960) species, has led several authors to propose that a pilidial
type of development with a drastic metamorphosis might be
ancestral for the phylum (Ja?gersten 1972; Starobogatov 1979;
Maslakova and Malakhov 1999). It is dif?cult to directly
compare nemertean larvae to trochophore larvae because so
little is known about the development of nonpilidial nemer-
tean larvae. We provide evidence that supports the notion
that the transitory larval epithelium of palaeonemerteans is
homologous to the transitory larval structures of other
trochozoans (i.e., trochoblast derivatives) and not to tissues
involved in pilidial metamorphosis (i.e., the derivatives of
pretrochal and trochal lineages).
MATERIALS AND METHODS
Confocal microscopy
Embryos of C. tremaphoros were obtained and raised as described
in Maslakova et al. (2004). For confocal microscopy, larvae were
?xed in 4% paraformaldehyde at 41C at 10, 15, 20, 25, 30, and
36.5h after fertilization and washed and stored in phosphate bu?er
saline (PBS) at 41C for up to 2 months. Larvae of the older ages (30
and 36.5h) were relaxed in a 1:1 mixture of 7.5% MgCl2 and
seawater before ?xation to avoid muscular contraction. Fixed lar-
vae were subsequently washed in PBT (0.1% Triton X-100 in PBS)
to permeabilize the cell membranes, counterstained with the
F-actin-binding Alexa Fluor Phalloidin (Molecular Probes Inc.,
Eugene, OR, USA) for 40min to facilitate visualization of the
larval morphology, washed in PBS, mounted in Vectashield on
poly-L-lysine pretreated slides, and viewed on a Nikon E-800
Eclipse microscope with the Bio-Rad Radiance 2100 confocal
system (Bio-Rad, Hercules, CA, USA) equipped with argon-ion and
helium-neon lasers. To visualize the green (Alexa Fluor Phalloidin)
label, argon laser was set to emit at 488nm.
Scanning electron microscopy
For scanning electron microscopy, 10-, 20-, 30-, and 36.5-h larvae
were ?xed by the semisimultaneous method described in Lacalli
and Gilmour (2001). An equal volume of 2% glutaraldehyde in
0.2M sodium cacodylate with 6% sucrose was added dropwise to
the seawater containing larvae. After 2min, the total volume was
doubled again by addition of 4% osmium tetroxide. Specimens
were left for 1h in the mixed ?xative and then washed thoroughly
in distilled water, gradually dehydrated in alcohol series and then in
acetone, and critical-point dried before being mounted on stubs.
Larvae of older ages (30 and 36.5h) were relaxed in a 1:1 mixture
of 7.5% MgCl2 and seawater before ?xation to prevent muscular
contractions. Specimens were viewed on a JSM 6400 Visions scan-
ning microscope by JEOL USA (Peabody, MA, USA).
RESULTS
Carinoma tremaphoros is a typical example of a benthic in-
faunal nemertean with a planktonic dispersal stage (larva).
Adults inhabit the top 10?15 cm of sand between high and
low tide marks. Females release up to several hundred rela-
tively small (100?110mm) eggs protected by a very thin egg
chorion and are initially embedded loosely in gelatinous mu-
cus. Eggs are fertilized externally and develop into planktonic
larvae. Although we were able to keep the larvae alive in the
culture for up to a month, we did not observe feeding or
settlement. Unfed larvae gradually diminished in size and
degenerated.
The early development of C. tremaphoros is described in
detail elsewhere (Maslakova et al. 2004). At 241C, embryos
became ciliated at 4h and began gastrulating at 5h after fer-
tilization. Uniformly ciliated spherical gastrulae possessing a
long apical tuft hatched out of the egg envelope at approx-
imately 5.5h after fertilization. At 10h after fertilization, an
actively swimming, uniformly ciliated, spherical gastrula was
formed (Fig. 2A). Gastrulae possessed a vegetal blastopore,
gut, apical tuft, and a pair of lateral invaginations at the
apical pole, most likely corresponding to the rudiments of the
nervous system (Fig. 2, E and I). Dorsoventral polarity of the
larva is apparent from the shape of the gut that is curved
toward the dorsal side (Fig. 2E) and the asymmetric arrange-
ment of the cells around the blastopore, with the group of
small cells located in the dorsal quadrant (Fig. 3). The surface
of the larva is covered by 40 large squamous cells except for
HOPLONEMERTEA
PILIDIOPHORA*
} PALAEONEMERTEA
} NEONEMERTEA?
Fig. 1. Nemertean phylogeny after Thollesson and Norenburg
(2003). Black rectangle indicates the hypothesized origin of the
necrotic type of development. Gray rectangle indicates the hy-
pothesized origin of development with the modi?ed trochophore
as found in Carinoma tremaphoros. 
Pilidiophora comprises the
Heteronemertea and the nonheteronemertean species possessing
a pilidium.
Hidden trochophore in nemertean development 221Maslakova et al.
the apical region and a small area at the vegetal pole posterior
(dorsal) to the blastopore, shown in Figure 3. The large sur-
face cells are arranged more or less in ?ve rows, each con-
taining eight cells. These 40 large cells thus form a very broad
and continuous preoral (preblastoporal) belt.
During the next 10h of development, the uniformly cili-
ated larva gradually elongated and the blastopore became
shifted toward the ventral side (Fig. 2, B, F, and J). As can be
inferred from the cell lineage analysis (Maslakova et al. 2004),
shifting of the blastopore to the ventral side is a result of the
faster rates of cell proliferation of the dorsal quadrant in both
the apical and the vegetal regions. The 40 large surface cells
described above remained cleavage arrested and formed an
increasingly more distinct contorted belt. The cells of the belt
changed shape slightly and shifted but maintained the neigh-
bor relationships: the ventral region of the belt became
broader than the dorsal as the whole belt assumed an oblique
position with respect to the anterior?posterior axis of the lar-
va. The ??N??-shaped gut can be subdivided into three distinct
regions at this stage. The mouth, derived from the blastopore,
leads into the esophagus, parallel to the anterior?posterior
axis of the larva and directed upward. The short middle re-
gion connects the esophagus to the main region of the gut
situated centrally along the anterior?posterior axis (Fig. 2J).
Fig. 2. Larval development of Carinoma tremaphoros. (A?D) Scanning electron microscopy images. (E?H) Confocal microscopy images,
ventral view. (I?L) Confocal microscopy images, right side view. (A, E, I) 10h old. (B, F, J) 20h old. Scale, 25mm. (C, G, K) 30h old. (D, H,
L) 36.5h old. Scale, 50mm. Note that the large ectodermal cells forming the prototroch disappear in the 36.5-h-old larvae. 
, blastopore
(mouth); ap, apical organ; gt, gut; n, nervous system rudiments; pt, prototroch cells. White arched lines in I?K mark the dorsal area occupied
by the prototroch, which gradually diminishes over the course of larval development.
222 EVOLUTION & DEVELOPMENT Vol. 6, No. 4, July^August 2004
The gut ends blindly at this stage; neither formation nor
presence of the anus was observed.
The general architecture of the larval musculature ?rst be-
came evident around 20?24h after fertilization. Four dorsal
longitudinal muscular strands are followed by the lateral and
ventral longitudinal muscles. The circular and diagonal mus-
culature of the body wall developed by 30h. By 24h the
presumed rudiments of the cerebral ganglia (paired ectoder-
mal invaginations at the apical pole) were almost completely
closed (Fig. 2, F and J). Left?right asymmetry of the larval
body became evident at this stage as a single ventrolateral eye
developed just to the right o? the midline.
At 30h after fertilization, the uniformly ciliated larva con-
tinued to elongate, up to about 200mm. As at the earlier
stages, there was no visible difference in ciliation between the
modi?ed surface cells forming the belt and the rest of the
ectoderm (Fig. 2C). Proliferating small cells of the apical and
dorsoposterior regions moved the ventral mouth further to-
ward the anterior end. Large cleavage-arrested surface cells
diminished in size and shifted to form a relatively compact,
oblique, preoral belt (Figs. 2, G and K, and 4). The difference
in cell size between the cleavage-arrested cells of the belt and
the proliferating cells of the anterior and posterior regions is
dramatic (Fig. 4). Longitudinal, circular, and diagonal mus-
culature of the body wall differentiated, forming a muscular
cylinder. The esophagus became surrounded by the well-de-
veloped circumpharyngeal muscles. At 36.5h, the cleavage-
arrested cells forming the belt disappeared completely (Fig. 2,
H and L), presumably by apoptosis.
DISCUSSION
The position and fate of the preoral belt of specialized cells in
the larvae of C. tremaphoros strongly resemble those of the
prototroch of other trochozoans. The exact cell composition
of prototrochs in annelids and mollusks (the best-studied
trochozoans) varies somewhat between species. However, the
total number of cells commonly is 28?40, and in all known
cases these are derived from subsets of the same cell lineages
(Damen and Dictus 1994b). Carinoma?s alleged prototroch
comprises 40 cells that are cleavage arrested, gradually di-
minish in size over the course of larval development, and
disappear, presumably undergoing cell death. Furthermore,
cell lineage analysis of C. tremaphoros embryos (Maslakova
et al. 2004) reveals that these 40 cells are derived from the
same lineages of embryonic precursors as the prototrochs of
other trochozoans. As in other trochozoans, four primary
trochoblasts (1a2?1d2) of Carinoma each divide twice and
produce 16 central cells of the prototroch (Fig. 5). Various
numbers and arrangements of accessory (vegetal-most prog-
eny of the 1q1) and secondary (animal-most progeny of 2q)
trochoblasts are found among different mollusks and anne-
lids. In these animals, the dorsal (D) quadrant typically gives
rise to fewer accessory and secondary trochoblasts than other
quadrants (Damen and Dictus 1994b). Interestingly, in Car-
inoma each quadrant, including D, contributes three accessory
and three secondary trochoblasts, so that 12 cells are derived
from the 1q1 micromeres (homologous to accessory trocho-
blasts in other trochozoans) and 12 cells are derived from the
2q micromeres (homologous to secondary trochoblasts in
other trochozoans) (Fig. 5; Maslakova et al. 2004). Although
the lack of differential ciliation obscures the presence of the
prototroch in Carinoma (hence the ??hidden?? trochophore),
it can be easily recognized as a preoral belt composed of
Fig. 3. The confocal image and the diagram of cell outlines of the
10-h-old larva of Carinoma tremaphoros showing the small cells in
the dorsal quadrant (view from the blastoporal side). Scale, 25mm.
(A) Confocal projection. (B) Diagram showing cell outlines.

, blastopore (mouth).
Fig. 4. The confocal image and the diagram of cell outlines of the
30-h-old larva of Carinoma tremaphoros showing the difference in
cell size of the epidermal cells. (Ventral-left view). Scale, 50mm. (A)
Confocal projection. (B) Diagram showing cell outlines. 
, blasto-
pore (mouth); pt, prototroch cells.
Hidden trochophore in nemertean development 223Maslakova et al.
40 specialized cells on confocal projections. Differentiation of
the nemertean trochoblasts is manifested in their cleavage
arrest and ultimate degenerative fate, characteristic of trocho-
blasts of other trochozoans. Although it is beyond the scope
of this study, staining with acridine orange and ethidium
bromide or a TUNEL assay could determine whether this is
an example of programmed cell death, as we would predict.
Although primary trochoblasts of annelids and mollusks
have been experimentally shown to undergo autodi?erentia-
tion, nothing is known about the mode of trochoblast spec-
i?cation in nemerteans. Blastomere isolation experiments on
the hoplonemertean Nemertopsis bivittata, which also makes a
planuliform larva, suggest that cell fates are speci?ed as early
as the four-cell stage (Martindale and Henry 1995). More
comprehensive cell isolation experiments in the heteronemer-
tean Cerebratulus lacteus (Henry and Martindale 1994; Mar-
tindale and Henry 1995) show an uncharacteristic ability of
nemertean embryos to regulate development, a feature not
seen in any other Trochozoan. An analysis of the develop-
mental potential of blastomeres, including the trochoblasts in
palaeonemerteans, is clearly needed.
Experiments on trochoblast speci?cation in the gastropod
mollusk Patella vulgata revealed the leading role of the 3D
macromere (spiralian organizer) in the modi?cation of the
dorsoventral organization of the prototroch (Damen and Dic-
tus 1996). Trochoblast speci?cation appears to occur in all
four quadrants initially, and then accessory and secondary
lineages become modi?ed by the in?uence of the 3D organ-
izer. In this context, radial symmetry (in terms of the lineage
contributions) of the prototroch in Carinoma, con?rmed by
the cell lineage data (Maslakova et al. 2004), suggests that the
role of the spiralian organizer in development of the nemer-
tean prototroch is either lacking/reduced or that all the
trochoblasts in Carinoma (including the primary, accessory,
and secondary lineages) are ??immune?? to the in?uence of the
organizer, just as the primary trochoblasts of other spiralians
are ??immune?? to the action of the organizer. If the latter is
true, this may be a case of lineage fate change, that is, the
accessory and secondary trochoblast lineages ??accept the
fate?? of the primary trochoblasts.
In most other trochozoans (except sipunculids), the pro-
totroch remains perpendicular with respect to the anterior?
posterior axis of the larva, even following the morphogenetic
movements of gastrulation (Buckland-Nicks et al. 2002a,b;
Jaeckle and Rice 2002; Nielsen 2002; Pernet et al. 2002; Pilger
2002; Zardus and Martel 2002). In Carinoma, the prototroch
becomes increasingly more distorted by morphogenetic move-
ments during the course of larval development. Just after
hatching and until about 10h after fertilization, the cells of the
prototroch occupy most of the larval ectoderm and the pro-
totroch is perpendicular to the anterior?posterior axis of the
body. However, later in development, the dorsal region of the
prototroch is pushed more and more toward the posterior end
by proliferating cells of the D quadrant in the pretrochal re-
gion (1q1 derivatives), whereas the ventral portion of the pro-
totroch retains its original position (Fig. 5). Faster rates of
proliferation of the dorsal ectodermal domain are character-
istic of spiralian development. However, in all other
trochozoans they are restricted to the posttrochal area (e.g.,
the primary somatoblast, 2d). Fast rates of proliferation of
the D quadrant in both pretrochal and posttrochal areas
(implied from the position of the blastopore and the pro-
totroch) are unique features of Carinoma (Maslakova et al.
2004). Labeling experiments on the hoplonemertean Nemerto-
psis bivittata (Henry and Martindale 1994) also indicate pos-
sible faster rates of proliferation of the D and C quadrants in
the pretrochal area. However, because the labeling was per-
formed at the four-cell stage (and not later), it is hard to
Fig. 5. Cell lineage of the developing prototroch in Carinoma tremaphoros. (A) The 16-cell stage. (B) 10-h-old larva. (C) 20-h-old larva.
(D) 30-h-old larva. The 16 central cells of the prototroch are derived from the primary trochoblast cell lineage (1q2). The 12 accessory
trochoblasts are derived from the 1q1. The 12 secondary trochoblasts are derived from the second quartet micromeres (2q).
224 EVOLUTION & DEVELOPMENT Vol. 6, No. 4, July^August 2004
determine whether this expansion is driven by asymmetric
growth of pre- or posttrochal derivatives.
Although the Carinoma prototroch does not possess spe-
cialized ciliation characteristic of typical prototrochs, we ar-
gue that it functions as a primary larval locomotory organ at
the early stages of larval development (because it covers the
majority of the surface). Our preliminary observations of the
embryos around the time of onset of ciliation suggest that the
equatorial region (comprising the trochal lineages) becomes
ciliated before the apical regions, which corresponds to the
early ciliation of the trochoblasts in mollusks and annelids.
Later in development, the cells of the pretrochal and post-
trochal region gradually take over the surface and the func-
tion of locomotion from the trochoblasts, which diminish in
size and eventually are lost. Despite the common perception
of the prototroch as a narrow band of cells with long com-
pound cilia, broad uniformly ciliated prototrochs covering a
substantial part of the larval surface are also found among
other trochozoans, for example, protobranch bivalves (Drew
1899; Zardus and Morse 1998; Gusta?son and Reid 1986,
1988), neomenioid aplacophorans (Baba 1951; Thompson
1960; Okusu 2002), a sipunculid Sipunculus nudus (Gerould
1906), and nonfeeding eunicid and onuphid polychaete larvae
(Pernet et al. 2002). Judging from the taxonomic distribution
of these larval forms, they are likely to represent convergent
similarity that illustrates the ??evolvability?? of larval ciliated
cells in these animals.
What we thought to be a transitory larval ectoderm in
Carinoma (which would be homologous to the larval tissues
of a pilidium) turns out to be a vestigial prototroch. This
challenges the previously suggested idea that a wholesale ne-
crotic type of metamorphosis (as found in the pilidium) is a
primitive character for the Nemertea (Ja?gersten 1972; Star-
obogatov 1979; Maslakova and Malakhov 1999). Instead, it
suggests that development with a uniformly ciliated tro-
chophore (as found in the basal nemertean Carinoma) might
correspond to the ancestral type of development in ne-
merteans, whereas development with a highly modi?ed pi-
lidium larva and a subsequent loss of the whole pretrochal
and trochal region is an autapomorphy of the nemertean
clade Pilidiophora (Fig. 1). In this light, there is no justi?ca-
tion for direct comparisons of nonnemertean larvae, like
Mu?ller?s larva of the polyclad platyhelminthes (Nielsen 2001),
with the nemertean pilidium. Unlike the prototroch of Car-
inoma and other trochozoan larvae, the pilidial ciliary band is
formed by blastomeres of the ?rst and second micromere
quartets, as well as 3c and 3d (Henry and Martindale 1998)
and the cells of the ciliated band appear to retain the ability to
divide, judging from the cell number and size. Unlike Car-
inoma but similar to many other trochozoans, the ciliary band
of the pilidium larva bears specialized cilia, different from
those covering the rest of the larval surface. Lack of di?er-
ential ciliation of the trochal lineage appears to be the
condition found in all ??planuliform?? larvae of hop-
lonemerteans and palaeonemerteans. This implies that the
differentially ciliated band of the pilidium larva evolved de
novo and coopted novel cell lineages that do not contribute to
prototrochs in other trochozoans. Thus, the highly stereo-
typed nature of spiral cleavage results in the independent co-
option of similar and novel cell lineages for a similar function
(a larval locomotory ciliary band) as had been previously
suggested by Ivanova-Kazas (1986). It remains unclear wheth-
er the hoplonemertean transitory larval ectoderm also repre-
sents a modi?ed prototroch or whether it is homologous to
the whole pretrochal and trochal region of a pilidium. A de-
tailed morphological study of hoplonemertean larvae, similar
to that undertaken for Carinoma, would be necessary to refute
either hypothesis. Until then the possibility remains that the
transitory larval ectoderm found in some hoplonemerteans
corresponds to the tissues lost during pilidial metamorphosis
and that a necrotic metamorphosis is ancestral for the Ne-
onemertea, a clade including the Pilidiophora and the Hop-
lonemertea (Fig. 1).
The inferred ancestral presence of a trochophore in ne-
merteans ?lls the ??gap?? in the distribution of trochophore
larvae among the Trochozoa (as de?ned by Peterson and
Eernisse 2001) and allows a meaningful comparison between
larval development of nemerteans and other trochozoans. A
study of the timing and molecular mechanisms of di?erenti-
ation of the trochoblast lineage in various trochozoans could
strengthen the assertion of homology of the prototroch and
other de?ned structures across the Trochozoa.
Acknowledgments
We are grateful to Dr. Mary Rice and the sta? of the Smithsonian
Marine Station at Fort Pierce and sta? of the Friday Harbor Lab-
oratories for their support and assistance. S. A. M. acknowledges Eric
Edsinger-Gonzales and George von Dassow for guidance in confocal
microscopy techniques. Eric Edsinger-Gonzales also provided a
number of critical comments on the manuscript. This work was
partially supported by NSF PEET grant DEB 97124463 to Diana
Lipscomb and J. L. N., Link Foundation/Smithsonian Institution
Graduate Fellowship to S. A. M., and NSF and NASA grants to M.
Q. M. This is contribution no. 582 from the Smithsonian Marine
Station at Fort Pierce.
REFERENCES
Baba, K. 1951. General sketch of the development in a Solenogastre,
Epimenia verrucosa (Nierstrasz). Misc. Repts. Res. Inst. Nat. Res. 19?21:
38?46.
Beklemishev, V. N. 1964. Principles of the Comparative Anatomy of Inver-
tebrates. University of Chicago Press, Chicago.
Buckland-Nicks, J., Gibson, G., and Koss, R. 2002a. Phylum Mollusca:
Polyplacophora, Aplacophora, Scaphopoda. In C. M. Young (ed.). Atlas
of Marine Invertebrate Larvae. Academic Press, San Diego, pp. 245?259.
Buckland-Nicks, J., Gibson, G., and Koss, R. 2002b. Phylum Mollusca:
Gastropoda. In C. M. Young (ed.). Atlas of Marine Invertebrate Larvae.
Academic Press, San Diego, pp. 261?287.
Hidden trochophore in nemertean development 225Maslakova et al.
Cantell, C. E. 1969. Morphology, development and biology of the pilidium
larvae (Nemertini) from the Swedish West Coast. Zool. Bidrag. F. Upp-
sala 38: 61?111.
Chernyshev, A. V. 2000. Larvae of nemerteans of the family Ototyphlo-
nemertidae in the plankton of Peter the Great Bay, Sea of Japan. Biol.
Morya 26: 50?52 (in Russian).
Child, C. M. 1900. The early development of Arenicola and Sternaspis.
Arch. Entw. Mech. 9: 587?723.
Costello, D. P. 1945. Experimental studies of germinal localization in
Nereis. J. Exp. Zool. 100: 19?66.
Damen, P., and Dictus, W. J. A. G. 1994a. Cell lineage analysis of the
prototroch of the gastropod mollusk Patella vulgata shows conditional
speci?cation of some trochoblasts. Roux?s Arch. Dev. Biol. 203: 187?198.
Damen, P., and Dictus, W. J. A. G. 1994b. Cell lineage of the prototroch of
Patella vulgata (Gastropoda, Mollusca). Dev. Biol. 162: 364?383.
Damen, P., and Dictus, W. J. A. G. 1996. Organizer role of the stem cell of
the mesoderm in prototroch patterning in Patella vulgata (Mollusca,
Gastropoda). Mech. Dev. 56: 41?60.
Dawydo?, C. 1940. Les formes larvaires de polyclades et de nemertines du
plancton indochinois. Bull. Biol. Fr. Belg. 74: 443?496.
Delsman, H. C. 1915. Eifurchung und Gastrulation bei Emplectonema
gracile Stimpson. Helder Tijdschr. Nederl. Dierk. Ver. 14: 68?114.
Drew, G. A. 1899. The anatomy, habits, and embryology of Yoldia lima-
tula. Mem. Biol. Lab. J. Hopkins Univ. 4: 1?37.
Gerould, J. H. 1906. The development of Phascolosoma (studies on the
embryology of the Sipunculidae II). Zool. Jahrb. Abt. Anat. Ont. Tiere
23: 77?162.
Gusta?son, R. G., and Reed, R. G. B. 1986. Development of the per-
icalymma larva of Solemia reidi (Bivalvia: Cryptodonta: Solemyidae) as
revealed by light and electron microscopy. Mar. Biol. 93: 411?427.
Gusta?son, R. G., and Reed, R. G. B. 1988. Larval and post-larval
morphogenesis in the gutless protobranch bivalve Solemya reidi (Crypt-
odonta; Solemyidae). Mar. Biol. 97: 373?387.
Hammarsten, O. D. 1918. Beitrag zur Embryonalentwicklung der Mala-
cobdella grossa (Mu?ll.). Dissertation thesis. Uppsala, Almqvist and Wik-
sells Boktryckery A. B., 96 pp.
Henry, J. Q., and Martindale, M. Q. 1994. Establishment of the dorsoven-
tral axis in nemertean embryosFevolutionary considerations of spiralian
development. Dev. Gen. 15: 64?78.
Henry, J. J., and Martindale, M. Q. 1998. Conservation of the spiralian
developmental program: cell lineage of the nemertean, Cerebratulus lac-
teus. Dev. Biol. 201: 253?269.
Hickman, V. V. 1963. The occurrence in Tasmania of the land nemertine,
Geonemertes australiensis Dendy, with some account of its distribution,
habits, variations and development. Pap. Proc. R. Soc. Tasmania 97:
63?75.
Ivanova-Kazas, O. M. 1985. Origin and phylogenetic significance of the
trochophoran larvae. 3. Larvae of ?atworms and nemerteans. Zool. Zh.
64: 1765?1767 (in Russian).
Ivanova-Kazas, O. M. 1986. Origin and phylogenetic significance of the
trochophoran larvae. 4. Kamptozoa larvaeFgeneral ideas. Zool. Zh. 65:
165?174 (in Russian).
Iwata, F. 1960. Studies on the comparative embryology of the nemerteans
with special reference to their inter-relationships. Publ. Akkeshi Mar.
Biol. Stn. 10: 1?51.
Jaeckle, W. B., and Rice, M. E. 2002. Phylum Sipuncula. In C. M. Young
(ed.). Atlas of Marine Invertebrate Larvae. Academic Press, San Diego,
pp. 375?396.
Ja?gersten, G. 1972. Evolution of the Metazoan Life Cycle. Academic Press,
London.
Lacalli, T. C., and Gilmour, T. H. J. 2001. Locomotory and feeding
effectors of the tornaria larva of Balanoglossus biminiensis. Acta Zool. 82:
117?126.
Lacalli, T. C., and West, J. E. 1985. The nervous system of a pilidium larva:
evidence from electron microscope reconstructions. Can. J. Zool. 63:
1901?1916.
Martindale, M. Q., and Henry, J. Q. 1995. Modi?cations of cell fate spec-
i?cation in equal-cleaving nemertean embryosFalternate patterns of
spiralian development. Development 121: 3175?3185.
Maslakova, S. A., and Malakhov, V. V. 1999. Hidden larva in the Hop-
lonemertini order of nemerteans. Dokl. Akad. Nauk. 366: 849?852.
Maslakova, S. A., Martindale, M. Q., and Norenburg, J. L. 2004. Fun-
damental properties of the spiralian developmental program are dis-
played by the basal nemertean, Carinoma tremaphoros (Palaeonemertea,
Nemertea). Dev. Biol. 267: 342?360.
McBride, E. W. 1914. Text-Book of Embryology. Macmillan and Co., Ltd.,
London.
Nielsen, C. 1995. Animal Evolution. Interrelationships of the Living Phyla.
Oxford University Press, Oxford.
Nielsen, C. 2001. Animal Evolution. Interrelationships of the Living Phyla.
Oxford University Press, Oxford.
Nielsen, C. 2002. Phylum Kamptozoa. In C. M. Young (ed.). Atlas
of Marine Invertebrate Larvae. Academic Press, San Diego, pp.
397?409.
Norenburg, J. L., and Stricker, S. A. 2002. Phylum Nemertea. In C. M.
Young (ed.). Atlas of Marine Invertebrate Larvae. Academic Press, San
Diego, pp. 163?177.
Okusu, A. 2002. Embryogenesis and development of Epimenia babai (Mol-
lusca Neomeniomorpha). Biol. Bull. 203: 87?103.
Pernet, B., Qian, P., Rouse, G., Young, C. M., and Eckelbarger, K. J. 2002.
Phylum Annelida: Polychaeta. In C. M. Young (ed.). Atlas of Marine
Invertebrate Larvae. Academic Press, San Diego, pp. 209?243.
Peterson, K. J., and Eernisse, D. J. 2001. Animal phylogeny and the
ancestry of bilaterians: inferences from morphology and 18S rDNA gene
sequences. Evol. Dev. 3: 170?205.
Pilger, J. E. 2002. Phylum Echiura. In C. M. Young (ed.). Atlas of Marine
Invertebrate Larvae. Academic Press, San Diego, pp. 371?373.
Reinhardt, H. 1941. Beitra?ge zur Entwicklungsgeschichte der einheimischen
Su?
wassernemertine Prostoma graecense (Bo?hmig). Vierteljahrsschr.
Naturf. Ges. Zu?rich. 86: 184?252.
Rouse, G. W. 1999. Trochophore concepts: ciliary bands and the evolution
of larvae in spiralian metazoa. Biol. J. Lin. Soc. 66: 411?464.
Rouse, G. W. 2000. Bias? What bias? The evolution of downstream larval-
feeding in animals. Zool. Scr. 29: 213?236.
Salensky, W. 1886. Bau und Metamorphose des Pilidiums. Z. Wiss. Zool.
43: 481?511.
Starobogatov, Y. I. 1979. Evolution of the pelagic larvae of protostomia
and problem of main body components. Zool. Zh. 58: 149?160
(in Russian).
Stricker, S. A., and Reed, C. G. 1981. Larval morphology of the nemertean
Carcinonemertes epialti (Nemertea: Hoplonemertea). J. Morphol. 169:
61?70.
Thollesson, M., and Norenburg, J. L. 2003. Ribbon worm relationshipsFa
phylogeny of the phylum Nemertea. Proc. R. Soc. Lond. B Biol. 270:
407?415.
Thompson, T. E. 1960. The development of Neomenia carinata Tullberg
(Mollusca, Aplacophora). Proc. R. Soc. B Biol. 153: 263?278.
Turbeville, J. M. 1986. An ultrastructural analysis of coelomogenesis in the
hoplonemertine Prosorhochmus americanus and the polychaete Magelona
sp. J. Morphol. 187: 51?60.
Turbeville, J. M. 2002. Progress in nemertean biology: development and
phylogeny. Integr. Compar. Biol. 42: 692?703.
Turbeville, J. M., and Ruppert, E. E. 1985. Comparative ultrastructure and
the evolution of nemertines. Am. Zool. 25: 53?71.
Turbeville, J. M., Field, K. G., and Ra?, R. A. 1992. Phylogenetic position
of phylum Nemertini, inferred from 18S ribosomal RNA sequenc-
esFmolecular data as a test of morphological character homology. Mol.
Biol. Evol. 9: 235?249.
Wilson, E. B. 1892. The cell lineage ofNereis: a contribution to the cytogeny
of the annelid body. J. Morphol. 6: 361?480.
Wilson, E. B. 1904. Experimental studies in germinal localization: II.
Experiments on the cleavage-mosaic in Patella and Dentalium. J. Exp.
Zool. 1: 197?268.
Zardus, J. D., and Martel, A. L. 2002. Phylum Mollusca: Bivalvia. In C. M.
Young (ed.). Atlas of Marine Invertebrate Larvae. Academic Press,
San Diego, pp. 290?325.
Zardus, J. D., and Morse, M. P. 1998. Embryogenesis, morphology and
ultrastructure of the pericalymma larva of Acila castrensis (Bivalvia:
Protobranchia: Nuculoida). Invert. Biol. 117: 221?244.
226 EVOLUTION & DEVELOPMENT Vol. 6, No. 4, July^August 2004