EVOLUTION & DEVELOPMENT
2:6, 340–347 (2000)
Molluscan engrailed expression, serial organization, and shell evolution
David K. Jacobs,a* Charles G. Wray,a,1 Cathy J. Wedeen,b Richard Kostriken,b Rob DeSalle,c
Joseph L. Staton,a,2 Ruth D. Gates,a and David R. Lindbergd
a
Department of Organismal Biology, Ecology, and Evolution, University of California, Los Angeles, 621 Charles East Young
Drive, South Los Angeles, CA 90095-1606, USA; bDepartment of Cell Biology and Anatomy, New York Medical College,
Valhalla, NY 10595, USA; cDepartment of Entomology, American Museum of Natural History, Central Park West at 79th Street,
New York, NY 10024, USA; and dMuseum of Paleontology and Department of Integrative Biology, University of California,
Berkeley, CA 94720, USA
1
Current address: Jackson Laboratory, 600 Main Street, Bar Harbor, ME 04609-1523, USA
2
Current address: Belle W. Baruch Institute for Marine Biology and Coastal Research, University of South Carolina, Columbia,
SC 29208, USA
*Author for correspondence (email: [email protected])
SUMMARY Whether the serial features found in some molluscs are ancestral or derived is considered controversial.
Here, in situ hybridization and antibody studies show iterated
engrailed-gene expression in transverse rows of ectodermal
cells bounding plate field development and spicule formation
in the chiton, Lepidochitona caverna, as well as in cells surrounding the valves and in the early development of the shell
hinge in the clam, Transennella tantilla. Ectodermal expression
of engrailed is associated with skeletogenesis across a range
of bilaterian phyla, suggesting a single evolutionary origin of in-
vertebrate skeletons. The shared ancestry of bilaterian-invertebrate skeletons may help explain the sudden appearance of
shelly fossils in the Cambrian. Our interpretation departs from
the consideration of canonical metameres or segments as units
of evolutionary analysis. In this interpretation, the shared ancestry of engrailed-gene function in the terminal/posterior addition of serially repeated elements during development explains
the iterative expression of engrailed genes in a range of metazoan body plans.
A class controversy persists regarding the nature of the evolutionary relationship between metameric features found in
some molluscs and segments of annelids and arthropods
(Wingstrand 1985; Haszprunar 1996). The discovery, in the
1950s, of living monoplacophorans with coordinated repeated
organ systems (Fig. 1A) supports the view that serial repetition in molluscs, such as chiton plates (Fig. 1B), share ancestry
with the segments of annelids and arthropods. However, another school of thought argues for a simple, nonmetameric
shelled or shell-less molluscan ancestor (Haszprunar 1992;
Haszprunar 1996), a thesis popularized by Lankester in the
last century (Lankester 1881). These “models” conflict—–the
first requires shared ancestry of the serial features in the Mollusca with the segments of annelids and arthropods, whereas
the second model requires the de novo evolution of serial features within the Mollusca. With these arguments in mind, we
examined the ectodermal expression of engrailed genes in two
classes of molluscs that have multiple shells: chitons, which
have eight plates (often thought of as metameric) as well as
separate spicules in the dermis; and clams, which have separate bilateral valves that are not considered metameric.
In the development of Drosophila, the homeobox-containing regulatory gene engrailed functions in each segment
to produce ectodermal boundaries, to differentiate specific
neurons, and to pattern appendages (Kornberg et al. 1985).
Genes similar in sequence to Drosophila engrailed (subsequently referred to as “en”) have been found in all bilaterian
metazoans examined, including flatworms and diverse protostome and deuterostome phyla (Wray et al. 1995). Arthropods with a variety of developmental modes, including short
germ-band insects and crustaceans, express en in ectodermal
stripes during segment formation (Patel et al. 1989a; Scholtz
1992; Rogers and Kaufman 1996). In addition, en is thought
of as a “segmentation marker” in arthropod development (Rogers
and Kaufman 1996). Leeches (annelid) (Wedeen and Weisblat
1991; Lans et al. 1993), Onychophora (Wedeen et al. 1997), amphioxus (Holland et al. 1997), and ophiuroid echinoderms (Lowe
and Wray 1997) all exhibit serial expression of en.
Chitons are the most “metamerically” organized mollusc that
is susceptible to developmental study (Fig. 1). In chitons, eight
shell plates armor the chiton dorsum, and muscles attach to each
plate in a stereotypical fashion (Fig. 1B). Seven of the eight
plates develop simultaneously (Watanabe and Cox 1975; Kniprath 1980), and posterior addition of the eighth plate field occurs later in development (Figs. 2A and 2B). We examined the
expression of an en gene in the trochophore stage of chiton de-
© BLACKWELL SCIENCE, INC.
340
Jacobs et al.
Fig. 1. Serial architecture of Monoplacophora, “A,” and of chiton,
“B” after Wingstrand, and “C,” the addition of serial dermal elements in the development of an aplacophoran after Baba (1940).
(A) In Monoplacophora, there is a close correspondence in repetition of gills, nephridia, and muscles. The fossil record suggests an
independent derivation of the modern single-shelled (conchiferan) molluscs, gastropods, scaphopods, cephalopods, and bivalves,
from monoplacophoran-like Cambrian ancestors. (B) Paired repetition of gills and nephridia occur in chiton. However, these organ systems do not have a one-to-one relationship with the series of shell
and muscle shown. (C) Ventral view of trochophore development in
Epimenia verrucosa, a neomeniomorph aplacophoran, both a ciliated,
protoctroch, “pr,” and telotroch, “te,” band are present. The larva
elongates through the introduction of epidermal pattern elements at
the telotroch. The epidermal pattern consists of globose papillary
cells, much like the goblet cells of chitons, and calcareous spicules.
velopment using the monoclonal antibody 4D9 (Patel et al.
1989b), and using in situ hybridization. We also report in situ hybridization observations of en expression associated with shell
development in the brooding clam, T. tantilla. Thus, we examine
repeated features along the chiton body axis and the bilaterally
divided shell of a clam.
engrailed in molluscs
341
Fig. 2. (A) Dorsal view of trochophore larval development and plate
field formation in the ectoderm of the chiton trochophore (Mopalia
muscosa), from hatching, 1, through settlement (Watanabe and
Cox 1975). The plate fields are first manifested by rows of mucusproducing goblet cells, “g.” Plates, “p,” then form between the
rows of goblet cells, as spicules, “s,” form in the girdle surrounding
the region of plate development. (B) During development, each
plate field consists of a stereotypic suite of cell types numbered
1–4, after Kniprath (1980). The goblet cell, type-1, nucleus is lower
than the adjacent type-2 cell nuclei, forming a fan, or fleur de lis
pattern; the type-4 cell nuclei are also lower in the section that the
other cell types aiding in the identification of cell types in section.
MATERIALS AND METHODS
Unlike other chitons, L. caverna retains eggs and trochophore larvae are retained as a brood in the pallial groove, making larvae accessible for study. We collected adult L. caverna from a coastal
cave in Santa Cruz, CA. Eggs and trochophore larvae were kept in
filtered seawater in petri dishes at 10⬚C, and were examined daily
342
EVOLUTION & DEVELOPMENT Vol. 2, No. 6, November-December 2000
under the dissecting microscope. Larvae were fixed when dorsoventral flattening associated with plate field formation was observed,
usually within 24 h after hatching. For antibody staining, larvae were
fixed in 4% formaldehyde in 50 mM cacodylic acid (pH 7.4) for 1 h
at room temperature, dehydrated in an ethanol series, and stored in
methanol at ⫺20⬚C. For in situ hybridization, larvae were fixed in
4% formaldehyde in phosphate buffered saline (PBS) and similarly
dehydrated and stored. T. tantilla were collected in the shallow subtidal of Morro Bay, CA. Sequential broods characteristic of this
species were fixed for in situ hybridization as described above.
After fixation, chiton larvae were treated with chitinase [Sigma
C125 from Streptomyces griseus 4 units/ml in HEPES (pH 7.4) buffer
for 1 h at room temperature (13)] to remove an ultra-structural layer of
chitin. Larvae were rinsed in PBS and placed in a blocking solution
(2% normal goat serum, 0.5% Triton X-100, 1xPBS for 12 h at 4⬚C).
The MAb 4D9 (17) was then applied in 200-fold dilution in blocking
solution for 24 h. After six 1-h washes in blocking solution, the larvae
were incubated in a 500-fold dilution of biotin-conjugated goat antimouse secondary antibody, and then washed six more times in blocking solution (1 h each). Following incubation with strepavidin-conjugated horseradish peroxidase for 12 h at 4⬚C, the larvae were again
washed six times in blocking solution, pre-incubated for 15 min in 5
mg/ml diaminobenzedine (DAB) and treated with 0.03% NiCl2 and
CoCl2. The resulting color reaction was observed and stopped after 10
to 15 min. The larvae were then Hoechst stained (1 ␮g/ml in 100 mM
Tris (pH 7.5) for 30 min), embedded in plastic (JB4), and sagitally sectioned at 4 ␮m thickness.
A 231 bp en sequence, obtained via PCR of genomic DNA
(Wray et al. 1995) and cloned into the PCR 2.1 vector (Invitrogen),
was used to generate an antisense RNA probe. In situ hybridizations
were carried out as in Tautz and Pfeifle (1989) with the exception
that larvae were treated with chitinase, as described above, prior to
proteinase K digestion (Wedeen and Weisblat 1991). Control embryos were probed with a sense RNA generated from the same en
region. Embryos were embedded in a mixture of Epon and Araldite
and sagitally sectioned at 10 ␮m intervals. T. tantilla in situ hybridizations were identical except that a T. tantilla-specific probe was
used (Wray et al. 1995). Results pertaining to en expression in shell
development are reported here for comparison with the chiton results. A more comprehensive examination of en expression through
T. tantilla development will be reported elsewhere.
RESULTS
In situ hybridized chiton trochophores express en in a transverse stripe associated with each developing plate, as well as
in the girdle area surrounding the region of plate formation
(Fig 3A and 3E). Sagital sections document that in each plate
field, the two rows of type-2 cells and the intercalated type-1,
goblet cells (see Fig. 2B) express en (Figs. 3B and 3F). It was
observed that staining concentrates around the nucleus and
basal portion of the cells. Similar localization of total RNA
has been observed in the shell gland of the gastropod Limnaea (Raven 1966), suggesting a general localization of
RNA within the cytoplasm rather than as an aspect of en
function. The expression in type-1 and type-2 cells bound the
plate fields, but there is no expression in cell types 3 and 4
that contact the plates. Spicules and developing plate fields
are evident under cross-polarized light (Fig. 3E) and Nomarski optics (Figs. 3A, 3B, and 3F), permitting one to discern
arelationship to en expression. As with plate formation, cells
directly involved in spicule formation do not express en, but
they are surrounded by cells in the girdle that do.
A ventral view of the whole mount embryo suggests repetition of en expression in cells of the developing foot, the
area of the embryo with ladder-like neural organization (Fig.
3G). This observation recalls the expression of en prior to the
organization of nervous tissue as seen in the development of
leeches (Wedeen and Weisblat 1991; Lans et al. 1993) and
arthropods (Patel et al. 1989a; Lans et al. 1993; Rogers and
Kaufman 1996). Further study is needed to confirm this interpretation.
Chiton trochophore larvae treated with the MAb4D9 protocol (Fig. 3C) show localized staining in the nuclei of cell
types 1 and 2 (Fig. 2B). These cell types form transverse
rows in each of the plate fields (Figs. 2A and 2B). Nuclear
localization of the signal reflects the DNA-binding regulatory function of the en protein.
In T. tantilla, en expression occurs around the margin of
the shell and along the hinge, reflecting the bivalved condition of the shell (Fig. 3H). Expression of en in the shell gland
of T. tantilla precedes shell formation (unpublished data), as
does the expression in the development of the shell gland in
the gastropod Illyanassa (Moshel et al. 1998).
DISCUSSION
In the chiton, en expression bounds regions of the shell, a
type of exoskeleton. In the clam, en expression surrounds
each developing valve and the uncalcified hinge. The chiton
girdle contains numerous carbonate spicules encircled by enexpressing cells (Figs. 3A and 3E). Similarly, en transcription surrounds the shell gland in developing gastropods
(Moshel et al. 1998). Thus, en expression delimits regions of
ectoderm that form the shell and spicule, the exokeletal units
in these three molluscan classes.
Aplacophora are thought to be amongst the most basal
branches in the molluscan phylum (Wingstrand 1985), and a
number of authors have argued that they are informative
with regard to the ancestral molluscan condition (SalviniPlawen 1972; Scheltema 1988). Aplacophora contain spicules (but not plates or shells) consistent with theories of
shell evolution involving coordination of the sets of cells that
originally produced spicules (Haas 1981; Bengtson 1992). It
should also be noted that shell loss has occurred in several
gastropod and cephalopod lineages. Perhaps most striking is
the loss of the shell and virtually all other adult structures in
Jacobs et al.
the enigmatic Xenoturbella, which was previously placed in
its own phylum but is now recognized through the study of
larval structures and molecular data as a reduced bivalve (Israelsson 1999). Loss of the shell would presumably entail
loss of shell-associated en expression.
Skeletogenesis and en expression
The expression of en in association with molluscan skeletogenesis, discussed above, invites comparison to association
of en expression with skeletogenesis observed in other taxa.
We find instances of en expression that correlate with skeletogenesis across the invertebrate bilaterians. This may reflect
a skeletogenetic function of en that evolved at or near the
base of the bilaterian clade. In addition, absence of en expression appears to be correlated with the absence or loss of
invertebrate skeletons. To date, en sequences are only known
from Bilateria (Wray et al. 1995). Current phylogenies divide the Bilateria into three major groups: Lophotochozoa,
Ecdysozoa, and deuterostomes.
Lophotrochozoa include the lophophorates, molluscs, annelids (Halanych et al. 1995), phyla potentially related to annelids (such as echiurans, pogonophorans, and sipunculans)
(McHugh 1997), and possibly some flatworms (Ruiz-Trillo
et al. 1999). Within the lophophorates, brachiopods and
bryzoans have external skeletons and the phoronids do not.
Annelids, echiurans, pogonophorans, and sipunculans may
form a clade (McHugh 1997), and skeletal elements in these
taxa are largely limited to jaws and setae in some polychaete
groups. The expression of the en gene in Lophotrochozoa is
only known from molluscs and leeches, which are derived annelids lacking the minimal exoskeletal elements present in
polychaetes. In leeches, en expression is added terminally in
rows of ectodermal and mesodermal cells, but ectodermal expression is transient (Wedeen and Weisblat 1991; Lans et al.
1993) as is consistent with the loss of ectodermal en expression in association with the loss of an identifiable exoskeleton.
Ecdysozoa contains the arthropods and related phyla,
such as tardigrades and onychophorans, as well as other
molting taxa such as priapulids and nematodes (Aguinaldo
et al. 1997). In Ecdysozoa, arthropods have the most extensive exoskeleton, and transcription of en in the ectoderm occurs at, or adjacent to, segment boundaries in a later stage of
insect and crustacean development (Rogers and Kaufman
1996). Parallels can be drawn between these boundaries and
the en expressing boundaries in molluscs. At segmental
boundaries in arthropods, calcification and cuticle formation
is minimal, thereby permitting flexion between skeletonized
segments. Furthermore, the shell matrix of molluscs contains
chitin (Weiner and Traub 1980), as does arthropod cuticle.
However, en expression does not appear to correlate with
boundaries in chitin formation. Material akin to shell matrix
is present in the bivalve hinge where en is transcribed, and
chitin is present across segmental boundaries in arthropods.
engrailed in molluscs
343
Rather, regulatory activity related to en transcription appears
to demarcate regions where secretion or enzyme activity
mineralizes the exoskeleton or leads to the hardening of an
organic cuticle.
In arthropods, en expression has been interpreted in the
context of segmentation. However, several observations suggest a role of ectodermal en expression in defining skeletal
or cuticular boundaries that do not always conform to segmental boundaries. For example, sternal-tergal boundaries
are usually rigid; however, in the last abdominal segment in
crayfish, en is expressed at the sternal-tergal boundary (Patel
et al. 1989a), presumably because this boundary must accommodate tail flexion and cannot be rigid. Thus, there is an
arthropod en expression boundary comparable to the clam
hinge (Fig. 3H) where expression is parallel rather than perpendicular to the anterior/posterior axis, and is presumably
associated with a change in mechanical properties of the ectoderm. If en functions in ectodermal boundary formation,
then one would anticipate expression in the joints of arthropod limbs. Other developmental genes (e.g., fringe) are
known to be expressed at limb joints, and modification of the
ectoderm could be under the control of these other regulators
(Rauskolb and Irvine 1999). However, nearly all study of en
in Drosophila has been conducted on larvae.1 There is only
one published photograph of en expression in adult Drosophila. It documents expression at segmental boundaries and in
the posterior compartment of the adult limb consistent with en
expression in the imaginal disc. However, expression appears
to be concentrated in the joints as well (Hama et al. 1990), and
further observations are needed to assess the relationship between en expression and morphologic details of the adult cuticle.
Onychophorans are thought to be the sister taxon of arthropods and are segmented. However, onychophorans lack
en expression in their dermis (Wedeen et al. 1997). Instead,
expression is observed in the posterior half of the developing
limb and in a segmental pattern in the lateral mesoderm. The
limb staining suggests shared ancestry of the onychophoran
and arthropod limbs (Wedeen et al. 1997). However, given
the close relationship of Arthropoda and Onychophora, and
their segmented body plans, the lack of segmental ectodermal expression in Onychophora suggests that the ancestral
role of en was not segmentation; this absence may be a consequence of evolutionary loss of skeletons. Onychophoran
dermis lacks a chitinous cuticle (Storch 1984); thus, Onychophora lack an exoskeleton. Furthermore, Cambrian fos-
1
It should be noted that the long germ-band development of Drosophila is
highly derived and little control of the segmentation process via the upstream
regulation of en is likely to be ancestral. Larval structures in long germ-band
insects are also derived. Primitive insects and crustacean larvae are far more
similar to adults, especially in their cuticle and segmentation, than are larval
Drosophila.
344
EVOLUTION & DEVELOPMENT Vol. 2, No. 6, November-December 2000
Fig. 3. In situ hybridization (A, B, F, G) and antibody (C, D) studies of en expression in the trochophore larva of the chiton, L. caverna,
as well as in situ hybridization study of the clam (H), T. tantilla. (A) Whole-mount chiton larva shows the repeated pattern of en expression in the dorsal surface (length approximately 250 ␮m); en is expressed adjacent to each of the 7 plate fields, “p,” and in the girdle that
surrounds the plate fields; en is not expressed in the cells associated with spicule, “s,” formation in the girdle. Thus, cells contacting plates
and spicules don’t express en but are surrounded and bounded by cells that do. (B, F) Sagital sections of a chiton in situ hybridized embryo. Staining is concentrated in the basal portions of the goblet cell, type-1 and adjacent type-2 cells. (C) Dorsal portion of a sagitally
Jacobs et al.
sils thought to be stem group onychophorans, such as Microdictyon, Hallucinogenia, and Xenusion (Ramskold and
Hou 1991; Dzik 1993), bear skeletal elements above the limb
on each segment. Therefore, the absence of en transcription
in the ectoderm of modern Onychophora could well be a
consequence of evolutionary loss of exoskeletal elements
coincident with the loss of a component of en expression associated with skeletal boundaries. In addition, the Cambrian
Paleoscolex, an ecdysozoan (Conway-Morris 1997), bears
more extensive ectodermal sclerites than modern priapulids,
and Wiwaxia and Halkeiria (Conway-Morris 1995) bear
more armor than many modern Lophotrochozoa. These observations are consistent with loss of skeletal elements in a
number of lineages since the early Cambrian Period.
In deuterostomes, en expression has been examined in
ophiuroid echinoderms (Lowe and Wray 1997), the basal
chordate amphioxus (Holland et al. 1997), and vertebrates,
including lamprey (Holland et al. 1993), zebrafish (Ekker et
al. 1992), and several tetrapods (Davis et al. 1991). In ophiuroid echinoderms, en is expressed in a dynamic pattern in
the ectoderm, bounding the ossicles that appear early in the
underlying mesoderm, as well as bounding spine ossicles,
which are formed subsequently, suggesting a relationship
between ectodermal en expression and skeleton formation
(Lowe and Wray 1997). This expression may imply an ancestry of en function in skeletogenesis basal to the deuterostome/protostome split. In other deuterostomes, association
of en expression with skletogenesis is not apparent and there
is no such relationship between en expression and skeletogenesis in vertebrates, consistent with the derived nature of the
vertebrate bone.
Small shelly fossils near the base of the Cambrian mark
the transition to a skeletonized fauna and the metazoan-dominated Phanerozoic fossil record. The recovery of articulated
specimens composed of multiple sclerites discussed above,
such as Wiwaxia, Halkieria and Microdiction, suggests that
much of the remaining “small shelly fauna” represent elements similarly employed as ectodermal armor in bilaterian
Metazoa that have yet to be recovered in an articulated form.
In addition, recent finds of cap-shaped shells composed of
fused spicules of the Early Cambrian age (Bengtson 1992)
support an argument of fusion of individual skeletal elements to form sclerites, plates, or shells (Haas 1981). Thus,
engrailed in molluscs
345
the en data reported here, in combination with previous scenarios for the formation of ectodermal armor and recent fossil discoveries, suggest a singular evolution of invertebrate
skeletons near the base of the Cambrian, followed by subsequent loss in several lophotrochozoan and ecdysozoan lineages. Such an interpretation, if substantiated, would have
important implications for the Cambrian radiation, as it
would constrain readily fossilizable exoskeletons to a single
lineage, leading to a monophyletic clade of bilaterians. This
would lead to a closer association of the bilaterian diversification with the base of the Cambrian. Further corroboration
of an ancestral role of en in delimiting exoskeletons will
require study of en expression in the development of invertebrate taxa that retain skeletons. Identification of a skeleton-specific en promotor element could also provide corroboration for the hypothesis of a shared ancestral role for en
in invertebrate skeleton formation.
en and metamerism
This work was initiated with the purpose of addressing the
evolutionary relationship between the quasi-metameric units
found in some molluscs and the canonical segments found in
annelids and arthropods. However, our data suggest a relationship between skeletogenesis and ectodermal en expression that is not always consistent with metameric boundaries.
Previous studies have emphasized similarity of en expression in
serially repeated mesodermal elements (Wedeen and Weisblat
1991; Lans et al. 1993; Holland et al. 1997; Wedeen et al. 1997),
and there is a general consensus that the serial repetition of en expression in neural development reflects shared ancestry (Lowe
and Wray 1997). These observations taken together suggest that
the focus on segments, or metameres, as units of evolution to be
used in comparing phyla could, in many cases, be too narrowly
drawn, depending on the definition of segmentation or metamerism used. It may be useful to compare components of gene expression, as we have done, in comparing instances of ectodermal
expression. Formal segments or metameres of some taxa can
then be more easily compared to serially repeated features found
in other taxa. It may also be fruitful to assess the shared ancestry
of mechanisms of addition of repeated elements in development, ignoring for the moment the serial versus metameric nature of those repeated features.
Reconstruction of the evolution of serial organization in
sectioned Hoechst stained anti-engrailed MAb 4D9 stained chiton larva observed under fluorescence. In this larva the downward waves
of cell type-4 nuclei under the plate fields is pronounced. The clear areas, denoted “p,” are the developing plates. The nuclei of cell types
3 and 4 show strong fluorescence. The DAB color reaction on the 4D9 antibody masks florescence from nuclei of cell types 1 and 2. (D)
Dorsal portion of a sagitally sectioned Hoechst stained control embryo, subjected to the antibody protocol in the absence of the primary,
MAb4D9, antibody. Note that Hoechst stain florescence is not masked in the type-1 and -2 cells. (E) Whole-mount chiton larva exposed
to cross-polarized light revealing the calcareous deposition of the spicules in the girdle and the seven developing plates. (G) A ventral
view of the whole-mount in situ hybridized chiton embryo shows a repeated pattern in cells of the foot possibly associated with the development of the ventral nervous organization. (H) Dorsal view of two in situ hybridized clam embryos (approximately 400 ␮m in length);
en expression surrounds both developing valves of the shell including expression in the hinge, “h.”
346
EVOLUTION & DEVELOPMENT Vol. 2, No. 6, November-December 2000
the molluscs suggests the existence of an ancestor that developed by a process of repeated addition of elements in a posterior growth zone. Aplacophora lack discrete metameres,
however, in the development of the aplacophoran solenogastre, Epimenia (Baba 1940), successive sets of dermal papillae are added in a subterminal growth zone in a fashion comparable to the posterior terminal addition of segments in
annelids and arthropods (Fig. 1C). By interpreting terminal
addition as ancestral in molluscs, combined with the view
that a number of other phyla in the Lophotrochozoa may be
derived from annelidous ancestors (McHugh 1997), one can
infer that terminal addition may be ancestral in the Lophotrochozoa. Terminal addition of segmental elements is thought
to be ancestral in the arthropods and occurs in onychophorans
(Anderson 1973). Similarities between the posterior/terminal
addition in these ecdysozoan taxa and annelids, now placed in
the Lophotrochozaoa, have long been recognized (Anderson
1973). Viewing the ambulacra and arms of echinoderms as
derived from bilaterians, they too experience terminal addition of serially repeated units (Ubaghs 1975; Littlewood et al.
1997).2
Expression of en is associated with the terminal addition
of serial elements in short germ-band insects (Patel et al.
1989a; Rogers and Kaufman 1996), crustaceans (Patel et al.
1989a, Scholtz 1992), leeches (Wedeen and Weisblat 1991;
Lans et al. 1993), and the arms of ophiuroid echinoderms
(Lowe and Wray 1997). In all these instances, en is expressed
in an ectodermal and/or mesodermal repeated registry, followed by serial neuronal expression. Many details of the en
expression pattern vary between the phyla examined. However, the repeated addition of posterior elements, combined
with en expression, is consistent with a terminal addition developmental process derived from a shared ancestor of Lophotrochozoa, Ecdysozoa, and, potentially, deuterostomes as
well. Therefore, the posterior/terminal addition could represent a component of the developmental process acquired at or
near the base of the bilaterian clade that has been retained in
a number of modern lineages, divergent in other respects.
In this scenario, metameres that reflect exoskeletal units
evolved in some lineages. In these instances, there is a strong
relationship between the metameric units and en expression
in the ectoderm, as has apparently occurred in arthropods
and to a more limited extent in some molluscs such as chitons. In other lineages, such as Onychophora and leeches,
metameric units have evolved that are not skeletal in nature,
2
Bilaterally organized, singly ambulacrate, stylophoran echinoderms occur
early in the Paleozoic Period (Ubaghs 1975), where the single ambulacra is
thought to be a component of the single body axis. Multiplication of an
ambulacral axis then generated the pentaradial form of modern echinoderms
from this singly ambulacrate condition (Littlewood et al. 1997). See Petersen et
al. (2000) for an alternative perspective.
and ectodermal expression of en is limited. Other taxa, such
as long germ-band insects, have lost terminal addition while
retaining a metameric exoskeleton derived from their immediate insect ancestors. In the conchiferan molluscs, loss of
terminal addition and formation of a single shell has been
followed by independent reduction of serial muscles and organ systems in bivalve, cephalopod, and gastropod lineages
(Wingstrand 1985).
CONCLUSIONS
Our results document ectodermal en expression that bound
serial skeletal elements in chiton, and both shells in a bivalve. Similar ectodermal expression at skeletal boundaries
occurs in arthropods (Patel et al. 1989a; Scholtz 1992; Rogers and Kaufman 1996) and echinoderms (Lowe and Wray
1997) as well as molluscs, suggesting homology of invertebrate skeletogenesis. Such a skeleton-generating developmental pathway must have predated, and could have been
causally related to, the sudden appearance of numerous skeletal elements in the fossil record near the base of the Cambrian. Absence of such ectodermal en expression in some
modern metazoan lineages, such as in recent Onychophora,
is correlated with the loss of skeletal structures. Terminal/
posterior addition of serial elements involving expression of
en may also have been an ancestral feature in the Bilateria
that has been lost in a number of taxa. Further work will be
needed to substantiate and clarify these views of bilaterian
body plan evolution.
Acknowledgments
We thank L. Holland, B. Runnegar, R. Flores, and anonymous reviewers for their comments as well as D. Weisblatt for his support.
This work was supported by the National Aeronautics and Space
Administration Exobiology Program (NAGW 4223, NAG-7207).
REFERENCES
Aguinaldo, A. M. A., Turbeville, J. M., Linford, L. S., Rivera, M. C., Garey,
J. R., Raff, R A., and Lake, J. A. 1997. Evidence for a clade of nematodes, arthropods and other moulting animals. Nature 387: 489–493.
Anderson, D. T. 1973. Embryology and Phylogeny in Annelids and Arthropods. Pergamon Press, Oxford, p. 495.
Baba, K. 1940. The early development of a Solenogastre, Epimenia verrucosa (Nierstrasz.) Annot. Zool. Japon. 19: 107–113.
Bengtson, S. 1992. The cap-shaped cambrian fossil Maikhanella and the relationship between coeloscleritophorans and molluscs. Lethaia 25: 401–420.
Conway-Morris, S. 1997. The cuticular structure of the 495-Myr-old type
species of the fossil worm Palaeoscolex, P-piscatorum. (?Priapulida).
Zool. J. Linn. Soc. 119: 69–82.
Conway-Morris, S., and Peel, J. S. 1995. Articulated halkieriids from the
lower Cambrian of North Greenland and their role in early protostome
evolution. Phil. Trans. Roy. Soc. Lond. Ser. B Biol. Sci. 347: 305–358.
Davis, C. A., Holmyard, D. P., Millen, K. J., and Joyner, A. L. 1991. Examining pattern formation in mouse, chicken and frog embryos with an
EN-specific antiserum. Development 111: 287–298.
Dzik, J. 1993. Early metazoan evolution and the meaning of its fossil
Jacobs et al.
record. In M. K. Hecht, R. J. MacIntyre, and M. T. Clegg (eds). Evolutionary Biology. New York, pp. 339–386.
Ekker, M., Wegner, J., Akimentko, M. A., and Westerfield, M. 1992. Coordinate embryonic expression of three zebrafish engrailed genes. Development 116: 1001–1010.
Haas, W. 1981. Evolution of calcareous hardparts in primitive molluscs.
Malacologia 21: 403–418.
Halanych, K. M., Bacheller, J. D., Aguinaldo, A. M. A., Liva, S. M., Hills,
D. M., and Lake, J. A. 1995. Evidence from 18S ribosomal DNA that
the lophophorates are protostome animals. Science 267: 1641–1643.
Hama, C., Ali, Z., and Kornberg, T. B. 1990. Region-specific recombination and expression are directed by portions of the Drosophila engrailed
promoter. Genes Dev. 4: 1079–1093.
Haszprunar, G. 1996. Origin and evolutionary radiation of the Mollusca. In
J. Taylor (ed.). Oxford University Press, Oxford, pp. 1–28.
Haszprunar, G. 1992. The first molluscs—small animals. Bollettino di Zoologia 59: 1–16.
Holland, L. Z., Kene, M., Williams, N. A., and Holland, N. D. 1997. Sequence and embryonic expression of the amphioxus engrailed gene
(AmphiEn): the metameric pattern of transcription resembles that of its
segment-polarity homolog in Drosophila. Development 124: 1723–1732.
Holland, N. D., Holland, L. Z., Honma, Y., and Fujii, T. 1993. Engrailed expression during development of a lamprey, lampetra-japonica: a possible clue to homologies between agnathan and gnathostome muscles of
the mandibular arch. Dev. Growth Differ. 35: 153–160.
Israelsson, O. 1999. New light on the enigmatic Xenoturbella (phylum uncertain): ontogeny and phylogeny. Proc. Roy. Soc. Lond. Ser. B Biol.
Sci. 266: 835–841.
Kniprath, E. 1980. Ontogenetic plate and plate field development in two chitons, Middendorffia and Ischnochiton. Wilhelm Roux’s Arch. 189: 97–106.
Kornberg, T., Siden, I., O’Farrell, P., and Simon, M. 1985. The engrailed locus of Drosophila: In situ localization of transcripts reveals compartment-specific expression. Cell 40: 45–54.
Lankester, E. R. 1881. Mollusca. In Encyclopaedia Britannica, 9th ed. A. & C.
Black, Edinburgh, pp. 632–695.
Lans, D., Wedeen, C. J., and Weisblat, D. A. 1993. Cell lineage analysis of
the expression of an engrailed homolog in leech embryos. Development
117: 857–871.
Littlewood, D. T. J., Smith, A. B., Clough, K. A., and Emson, R. H. 1997.
The interrelationships of the echinoderm classes: Morphological and
molecular evidence. Biol. J. Linn. Soc. 61: 409–438.
Lowe, C. J., and Wray, G. A. 1997. Radical alterations in the roles of homeobox genes during echinoderm evolution. Nature 389: 718–721.
McHugh, D. 1997. Molecular evidence that echiurans and pogonophorans
are derived annelids. Proc. Natl. Acad. Sci. USA 94: 8006–8009.
Moshel, S, M., Levine, M., and Collier, J. R. 1998. Shell differentiation and
engrailed expression in the Ilyanassa embryo. Dev. Genes Evol. 208:
135–141.
Patel, N. H., Kornberg, T. B., and Goodman, C. S. 1989a. Expression of engrailed during segmentation in grasshopper and crayfish. Development
107: 201–212.
engrailed in molluscs
347
Patel, N. H., Martin-Blanco, E., Coleman, K. G., Poole, S. J., Ellis, M. C.,
Kornberg, T. B., and Goodman, C. S. 1989b. Expression of engrailed
proteins in arthropods, annelids, and chordates. Cell 58: 955–968.
Petersen, K. J., Arenas-Mena, C., and Davidson, E. H. 2000. The A/P axis
in echinoderm ontogeny and evolution: evidence from fossils and molecules. Evol. Dev. 2: 93–101.
Ramskold, L., Hou, X. 1991. New early Cambrian animal and Onychophoran affinities of enigmatic Metazoans. Nature 351: 225–228.
Rauskolb, C., and Irvine, K. D. 1999. Notch-mediated segmentation and
growth control of the Drosophila leg. Dev. Biol. 210: 339–350.
Raven, C. P. 1966. Morphogenesis: The Analysis of Molluscan Development. Pergamon Press, Oxford, p. 365.
Rogers, B. T., and Kaufman, T. C. 1996. Structure of the insect head as revealed by the EN protein pattern in developing embryos. Development
122: 3419–3432.
Ruiz-Trillo, I., Riutort, M., Littlewood, D. T. J., Herniou, E. A., and Baguna, J. 1999. Acoel flatworms: Earliest extant bilaterian metazoans, not
members of platyhelminthes. Science 283: 1919–1923.
Salvini-Plawen, L. V. 1972. Zur morphologie und phylogenie der Mollusken: Die beziehungen der Caudofauveata und der Solenogastres als
Aculifera, als Mollusca und als Spiralia. Zeitschrift für wissenschaftliche
Zoologie 184: 205–394.
Scheltema, A. H. 1988. Ancestors and descendants relationships of the
Aplacophora and Polyplacophora. Amer. Malacol. Bull. 6: 57–68.
Scholtz, G. 1992. Cell lineage studies in the crayfish Cherax-destructor
Crustacea Decapoda germ band formation segmentation and early neurogenesis. Roux’s Arch. Dev. Biol. 202: 36–48.
Storch, V. 1984. Onychophora. In J. Bereiter-Hahn, A. G. Matoltsy, and
K. S. Richards (eds.). Biology of the Integument1 Invertebrates. Springer-Verlag, Berlin, pp. 703–705.
Tautz, D., and Pfeifle, C. 1989. A non-radioactive in situ hybridization
method for the localization of specific RNA in Drosophila embryos reveals translational control of the segmentation gene hunchback. Chromosoma 98: 81–85.
Ubaghs, G. 1975. The early Paleozoic echinoderms. Ann. Rev. Earth Planet. Sci. 3: 79–98.
Watanabe, J. M., and Cox, L. R. 1975. Spawning behavior and larval development in Mopalia lignosa and Mopalia muscosa. Veliger 18 Suppl.: 18–27.
Wedeen, C. J., Kostriken, R. G., Leach, D., and Whitington, P. 1997. Segmentally iterated expression of an engrailed-class gene in the embryo of
an Australian onychophoran. Dev. Genes Evol. 207: 282–286.
Wedeen, C. J., and Weisblat, D. A. 1991. Segmental expression of an engrailed-class gene during early development and neurogenesis in an annelid. Development 113: 805–814.
Weiner, S., and Traub, W. 1980. X-ray diffraction study of the insoluble organic matrix of mollusk shells. FEBS Letters 111: 311–316.
Wingstrand, K. G. 1985. On the anatomy and relationships of recent
Monoplacophora. Galathea Reports 16: 7–94.
Wray, C. G., Jacobs, D. K., Kostriken, R., Vogler, A. P., Baker, R., and Desalle, R. 1995. Homologues of the engrailed gene from five molluscan
classes. FEBS Letters 365: 71–74.