Download Early Events in Annelid Regeneration: A Cellular Perspective

Integrative and Comparative Biology
Integrative and Comparative Biology, volume 54, number 4, pp. 688–699
doi:10.1093/icb/icu109
Society for Integrative and Comparative Biology
SYMPOSIUM
Early Events in Annelid Regeneration: A Cellular Perspective
Alexandra E. Bely1
Department of Biology, University of Maryland, College Park, MD 20742, USA
From the symposium ‘‘The Cell’s View of Animal Body Plan Evolution’’ presented at the annual meeting of the Society
for Integrative and Comparative Biology, January 3–7, 2014 at Austin, Texas.
1
E-mail: [email protected]
Synopsis The ability to regenerate extensive portions of the body is widespread among the phylum Annelida and this
group includes some of the most highly regenerative animals known. Knowledge of the cellular and molecular basis of
regeneration in this group is thus important for understanding how regenerative processes have evolved both within the
group and across animal phyla. Here, the cellular basis of annelid regeneration is reviewed, with a focus on the earliest
steps of regeneration, namely wound-healing and formation of the blastema. Information from a wide range of annelids
is compiled in order to identify common and variable elements. There is a large body of valuable older literature on the
cellular basis of regeneration in annelids and an effort is made to review this literature in addition to more recent studies.
Annelids typically seal the wound through muscular contraction and undergo some autolysis of tissue at the site of the
wound. Bodily injury elicits extensive cell migration toward the wound, involving several different types of cells. Some
migrating cells form a tissue-clot and phagocytize damaged tissues, whereas others are inferred to contribute to regenerated tissue, specifically mesodermal tissue. In one annelid subgroup, the clitellates, a group of mesodermal cells,
sometimes referred to as neoblasts, is inferred to migrate over considerable distances, with cells moving to the wound
from several segments away. Epidermis and gut epithelia severed upon amputation typically heal by fusing with like
tissue, although not always. After amputation, cellular contacts with the extracellular matrix are disrupted and major
changes in cell morphology and adhesion occur within tissues near the wound. Interactions of tissues at the wound
appear key for initiating a blastema, with a particularly important role suggested for the ventral nerve cord, although
species are variable in this regard; longer-distance effects mediated by the brain are also reported. The anterior–posterior
polarity of the blastema can be mis-assigned, leading most commonly to double-headed worms, and the dorsal–ventral
polarity of the blastema appears to be induced by the ventral nerve cord. The blastema is thought to arise from
contributions of all three tissue layers, with each layer replacing itself in a tissue-specific manner. Blastemal cells originate
mostly locally, although some long-distance migration of source-cells is suggested in clitellates. A number of important
questions remain about the cellular basis of regeneration in annelids and addressing many of these would be greatly aided
by developing approaches to identify and isolate specific cell types and techniques to image and trace cells in vivo.
Introduction
Annelids (segmented worms) are a large and diverse
phylum that includes many members capable of extensive regeneration (Hyman 1940; Berrill 1952;
Herlant-Meewis 1964; Bely 2006). Numerous species
can regenerate a complete head, a complete tail, or
both at once, in some cases starting from just a small
fragment of the original individual. Understanding
the cellular and molecular basis of regeneration in
annelids can help to address important questions
such as: what is the relationship between regeneration and embryogenesis? What processes fail in
poorly regenerating species? How have regenerative
abilities and mechanisms evolved within and across
phyla? For this reason, regeneration of annelids has
been a subject of growing interest.
Regeneration in annelids has been studied for well
over a century and in a variety of representatives
from across the phylum (Randolph 1892;
Stephenson 1930; Hyman 1940; Herlant-Meewis
1964; Goss 1969; Bely 2006). Recent studies are
focusing increasingly at the molecular level but
there is a large, mostly older, literature that focuses
at the cellular level. Understanding the cellular and
Advanced Access publication August 13, 2014
ß The Author 2014. Published by Oxford University Press on behalf of the Society for Integrative and Comparative Biology. All rights reserved.
For permissions please email: [email protected].
689
Annelid regeneration
tissue-level dynamics that occur during regeneration
in annelids is key for interpreting the emerging molecular data and, more broadly, to provide a complete understanding of regenerative processes in
annelids.
Here, I review literature on cellular and tissuelevel processes that occur during regeneration in annelids. I focus on early phases of regeneration, specifically wound-healing and formation of the
blastema (the mass of seemingly undifferentiated
cells that forms at the site of the wound and differentiates into the new structures), as these early
phases are likely to be most relevant for addressing
the kinds of questions outlined above; many studies
describe later phases of tissue morphogenesis but
these are not reviewed here. Within the wound-healing section, I review information on muscle contraction, autolysis, cell migration, re-epithelialization,
and remodeling of tissue and the extracellular
matrix. Within the blastema-formation section, I
review information on the initiation, polarity, and
cellular origins of the blastema. Although the major
events of early regeneration are discussed in a rough
temporal sequence, it should be noted that many of
these events are interconnected and many occur simultaneously. Throughout, I focus on regeneration
by epimorphosis (involving addition of new tissue),
primarily following transverse amputation; regeneration by morphallaxis (remodeling of existing tissue)
is common among annelids (Berrill 1952; HerlantMeewis 1964) but few studies have addressed its cellular underpinnings.
This review aims to tie together findings from a
range of approaches, including histological and descriptive studies, experimental surgeries, grafting experiments, and proliferation assays. Many of the
sources reviewed here represent older literature; it
is my hope that by reviewing these works here they
will be more broadly known. It should be noted,
however, that many of these older studies were performed before the availability of methods to track
cells using cell tracers, molecular markers, or timelapse imaging. Conclusions regarding the sources and
fates of cells during regeneration or the presence or
absence of certain tissues in manipulative experiments often are based on light microscopy and histology of fixed tissue alone. Although these studies
have yielded valuable information, their limitations
should be borne in mind and their conclusions merit
being revisited with modern methods.
Findings from a broad range of annelids are reviewed here, with the goal of identifying features of
regeneration that appear to be general among annelids, as well as identifying features that merit being
studied in a broader sampling of species. Cellular
studies of regeneration have been carried out in representatives from both major annelid clades, the
Errantia (many ‘‘polychaetes’’) and the Sedentaria
(many ‘‘polychaetes’’ and all clitellates, including ‘‘oligochaetes,’’ leeches, and relatives), as well as basal annelid lineages (Struck et al. 2011; Weigert et al. 2014).
Most of the data discussed here were obtained from
the following annelids: Errantia clade—Nereis/
Neanthes (Nereididae), Nephtys (Nephtyidae),
Eulalia (Phyllodocidae), Eurythoe (Amphinomidae),
Dorvillea (Dorvilleidae); Sedentaria clade—Capitella
(Capitellidae), Sabella and Branchiomma (Sabellidae),
Hydroides and Salmacina (Serpulidae); Lumbriculus
(Clitellata: Lumbriculidae), Eisenia and Allolobophora
(Clitellata: Lumbricidae), Enchytraeus (Clitellata:
Enchytraeidae), Tubifex, Limnodrilus, Dero, and
Pristina (Clitellata: Naididae, including former
Tubificidae), Hirudo (Clitellata: Hirudinidae); and
basal annelids—Owenia (Oweniidae) and Chaetopterus
(Chaetopteridae). For brevity, and because species-level
taxonomic revisions are not uncommon, we refer only
to generic names in this review.
Wound-healing
Muscle contraction
Following transverse amputation, nearly all annelids
investigated can seal the body effectively by rapid
muscular contraction (Hyman 1940; HerlantMeewis 1964; Bilej 1994). Observations of amputations on live specimens of many species indicate that
constriction of circular body muscles, and in some
cases slight extrusion of gut tissue, quickly closes the
wound, stemming the loss of body fluids and presumably protecting against infection. Even annelids
that cannot regenerate following transverse amputation often effectively seal such an injury by muscular
constriction and later re-epithelialization of tissue
(Bilej 1994; Bely 2006; Bely and Sikes 2010).
However, at least a few annelids cannot accomplish
this; acanthobdellidans (relatives of leeches) have a
stiff outer cuticle and fail to seal wounds following
transverse amputation, dying shortly after any serious breach of the body wall (Bely 2006).
In some species, transverse amputation is accompanied by, and even may be preceded by, autotomy
(self-amputation) effected by muscular contraction.
In a species of Enchytraeus, transverse amputation
typically is followed by autotomy at a stereotypical
intra-segmental plane, corresponding to the position
of one of the segmental nerve rings (Yoshida-Noro
et al. 2000; Kawamoto et al. 2005). Autotomy thus
leads to regeneration being initiated from a
690
consistent, predictable intraspecific position. In
Lumbriculus, high-speed video analysis indicates
that autotomy can even occur prior to severing of
the body wall (Lesiuk and Drewes 1999). Severe constriction of the body wall for greater than about
80 ms induces an autotomy reflex that severs the
body just anterior to the plane of constriction
within a few hundreds of a millisecond after compression. The process is so efficient that no loss of
blood is apparent. In both of these annelids, anesthesia inhibits the autotomy response (Lesiuk and
Drewes 1999; Kawamoto et al. 2005). Both the
Enchytraeus species and the Lumbriculus species investigated reproduce asexually by fragmentation (a
type of fission) and thus routinely autotomize to
reproduce. It would be valuable to determine how
common pre-amputation and post-amputation autotomy are among annelids and whether these processes are found primarily among animals that
reproduce by fission.
Autolysis
In several annelid species, histological analyses of
fixed samples have shown that some tissues at the
site of the wound undergo autolysis soon after
injury. In Limnodrilus, autolysis is evident both in
ectoderm and in mesoderm within hours after transverse amputation (Cornec et al. 1987), and in an
earthworm both the epidermis and the underlying
muscle near that wound are broken down within
an hour after injury, inflicted by burning a hole in
the body wall (Cameron 1932). Additional studies
are needed to determine how common autolysis of
tissue is following injury, as well as whether the
extent of autolysis depends on the severity or type
of the injury.
Cell migration: phagocytosis, tissue-plug formation,
and source-cells
Histological studies of tissues fixed at multiple time
points after injury suggest that amputation triggers a
major migration of cells in the hours following
wounding. Migration of cells to the wound appears
to be a general feature of annelids, having been inferred in a range of annelids and following a range of
injuries (transverse amputation, breach of the body
wall, removal of terminal asegmental structures)
(e.g., Stephan-Dubois 1954; Hill 1970; Cornec et al.
1987; Bilej 1994; Tettamanti et al. 2004). The period
of extensive movements of cells has been documented within minutes to hours after amputation
and typically persists at least through the first day
after wounding (Cornec et al. 1987; Bilej 1994;
A. E. Bely
Huguet and Molinas 1994). Several cell types are
inferred to migrate toward the wound and perform
different functions. Most broadly, these include cells
that migrate to form a tissue plug at the wound, cells
that migrate to the wound and phagocytize cellular
debris, and largely undifferentiated cells that arrive at
the wound and which may contribute to regenerated
structures, as described below. It is important to note
that homologizing cell types across annelid groups
remains challenging: the names used to refer to migratory cells often are inconsistent across annelid
groups and when the same names have been applied
to cells from different groups those cells typically
have been homologized only on the basis of morphology, position, and behavior (Cameron 1932;
Stephan-Dubois 1954; Cornec et al. 1987; Bilej
1994; Tettamanti et al. 2004).
One of the most detailed analyses of cell migration
in annelids is that by Cornec et al. (1987) who investigated cellular and tissue dynamics following posterior amputation in Limnodrilus. Based on
observations of fixed histological sections, they inferred that a number of cell types migrate in response to wounding and that these fall into two
general categories: phagocytes and dissepimentary
cells. Migrating phagocytes themselves are of at
least two types: coelomocytes and splanchnopleural
cells. Coelomocytes are round cells that are free
within the coelom and often have numerous inclusions and vesicles. Within the first few hours after
amputation, coelomocytes migrate to the site of the
wound and phagocytize damaged tissue, especially
degenerating muscle fibers generated by an early
wave of autolysis. Coelomocytes extend long pseudopods and engulf these damaged cells. The second
type of migrating phagocyte is the splanchnopleural
cell. These cells normally surround the gut and
dorsal blood vessel but, following amputation, they
are thought to be the source of two general kinds of
migrating cells. The first are cells that have some
inclusions and are frequently observed engulfing
other cells at the wound. The second are chloragocytes that have many inclusions and appear to transform into eleocytes (with many dark inclusions),
which rarely, if ever, engulf other cells. In addition
to coelomocytes and splanchnopleural cells, Cornec
et al. (1987) also refer to, but do not further elaborate on, the migration of what they call ‘‘free cells,’’ a
possible third type of phagocyte that may correspond
to what has been called ‘‘amoebocytes type I’’ in
Lumbricillus, ‘‘hyalocytes’’ in Eisenia, and/or ‘‘macrophages’’ in Eisenia. In addition to engulfing cells and
cellular debris at the wound, phagocytes also form a
mass at the wound that helps to physically plug it.
691
Annelid regeneration
This temporary wound-plug accumulates by about 1
day post-amputation (dpa) and is then degraded by
autolysis once the overlying epidermis is healed, at
about 2 dpa.
The second general category of migrating cells described by Cornec et al. (1987) is the dissepimentary
cells, which are cells originating from the segmental
septa. Segmental septa are composed of a thick basal
lamina with flattened cells on either side (i.e., on the
septum’s anterior and posterior faces). In the ventral
and posterior part of the septum are round cells with
a large nucleus-to-cytoplasm ratio and inferred to be
undifferentiated or weakly differentiated. Within 4 h
after posterior amputation, in the 10 or so segments
nearest the wound, some of these round septal cells
increase considerably in size. Over the next 2 days,
spindle-shaped cells with similar cytological properties to those on the septa are seen on the dorsal
surface of the ventral nerve cord, often between
septa; these cells are inferred to have originated
from the septa and then migrated toward the
wound. Such cells accumulate at the severed tip of
the nerve cord, under the healing epidermis. Cornec
et al. (1987) suggest that these cells are likely source
mesodermal cells for the regenerate, but this is necessarily a preliminary conclusion, as the limitations
of using static images to infer movement and fates of
cells are considerable.
Consistent with Cornec et al. (1987), in a broad
range of annelids injury induces the recruitment of
many phagocytic cells that engulf damaged cells and
cellular debris and induces the formation of a temporary cellular plug at the wound (Cameron 1932;
Burke 1974; Cornec 1984; Bilej 1994; Grdisa 2010).
Although phagocytosis mostly has been shown
through ultrastructural imaging (e.g., transmission
electron microscopy), some studies also have demonstrated phagocytosis directly by following the fate
of foreign particles (e.g., carbon particles, beads)
introduced into the animal (Cameron 1932).
Phagocytosis and the formation of wound-plugs are
thus components of the post-injury cell migration
response that are probably common to most annelids, and, indeed, some aspects may represent part of
a generalized response to injury that may be shared
with other animals (Vetvicka et al. 1994).
Only in clitellates, however, have cells similar to
the dissepimentary cells of Cornec et al. (1987) been
documented. In a broad range of ‘‘oligochaetes’’
(namely, non-leech clitellates), spindle-shaped cells
that appear to be undifferentiated or weakly differentiated are inferred (based on time-series of fixed
material) to originate from the base or surface of the
septa, become activated by injury, and migrate to the
wound typically along the ventral nerve cord
(Randolph 1891, 1892; Stephan-Dubois 1954;
Tadokoro et al. 2006). These cells often have been
referred to as ‘‘neoblasts,’’ a term first coined for
annelids (Randolph 1891, 1892) to describe these
cells in Lumbriculus (and now a term also widely
used to refer to a pluripotent type of cell in planarians). Despite efforts to identify these cells in other
annelids, dissepimentary cells/‘‘neoblasts’’ that migrate intersegmentally appear to be restricted to ‘‘oligochaetes,’’ or at least are easily recognized only in
this clade (Boilly 1969b; Hill 1970; Paulus and
Müller 2006). The inferred function of these cells
has long been debated, as discussed in the section
‘‘Cellular origins of the blastema,’’ see below.
In summary, a significant post-injury phase of cell
migration is evident in annelids, suggesting this to be
an important aspect of early regeneration, yet much
work remains to be done to characterize the types of
cells involved, their behaviors, and their functions.
Identifying and homologizing types of cells across
different groups of annelids remain persistent challenges. To date, migrating cells have been identified
almost exclusively based on morphology, yet different cell types could have similar morphologies, the
same type could present different morphologies depending on regenerative stage or context, and a particular type of cell may appear different in different
groups of annelids. Efforts to generate markers for
specific cell types could allow for important advances
in our understanding of cell migration and of the
homologies among cell types. Another important
goal for future research is to identify the cues that
trigger the migration of different types of cells.
Finally, almost all of the knowledge about migrating
cells is based on inferences from static histological
sections. Methods to assess cell migration in vivo
are greatly needed. Developing in-vivo imaging
would allow direct assessments of cell migration, including the direction and speed of movement; would
reveal cells’ behaviors and changes in shape; could
reveal less obvious movements of cells (e.g., smaller
or less numerous cells and minor migration routes);
and, if in-vivo imaging can be coupled with labeling
of cells, could prove highly valuable for tracing the
fates of migrating cells.
Re-epithelialization
Transverse amputation severs the epidermal and gut
epithelia, both of which must then be healed. Most
commonly, the severed edges of the epidermis fuse to
each other and the severed edges of the gut epithelium fuse to each other (Herlant-Meewis 1964). This
692
produces a continuous outer epithelium (referred to
as a wound-epithelium or wound-epidermis) with no
opening; the mouth or anus is then formed secondarily. This type of healing occurs following anterior
amputations and usually following posterior amputations. However, some descriptions of posterior regeneration indicate that the severed edges of the
epidermis fuse directly to the severed edges of the
gut epithelium; when this occurs the anus is reformed directly by the wound-healing process.
Wound-healing by fusion of the outer epidermis
with the epithelium of the gut has been described
following posterior amputation in Sabella (Hill
1970) and Nereis (Boilly 1969a), and in Nephtys, posterior amputation can be healed in either way, suggesting that the mode of healing can vary even
within a species (Clark and Clark 1962). Regardless
of the type of healing, the re-epithlialization process
consistently occurs by rearrangement of cells, and
without mitosis (Hill 1970; Paulus and Müller
2006; Zattara and Bely 2011, 2013). Indeed, re-epithelialization typically is complete well before the
first indication of cell proliferation at the wound
(typically at about 1–2 dpa).
Remodeling of the extracellular matrix and tissues
near the wound
Amputation leads to extensive changes in the cellular
interactions with the extracellular matrix as well as in
the contacts among cells within tissues near the
wound. These changes have been most extensively
studied in histological sections following anterior
amputation in Owenia (Fontes et al. 1983; Coulon
et al. 1989; Dupin et al. 1991). A basement membrane composed of extracellular matrix is normally
present below the epidermis, separating the epidermis from the underlying muscle layer. In uncut animals, at the base of epidermal cells, where they
contact the basement membrane, there is an extensive cytoskeleton network, made of actin, that gives
these cells a strong apical-basal polarity. These cells
also have many adherence-structures anchoring them
to other epidermal cells. Muscle-cells below the basement membrane also make numerous contacts with
the basement membrane as well as with each other.
Following anterior amputation in Owenia, the
wound-healing process brings the edges of the epidermis and underlying muscle layer in direct contact,
with no basement membrane between them. In addition, epidermal cells that are near the wound and
close to the basement membrane lose contact with
the basement membrane. Correlated with the loss of
contact with the basement membrane, epidermal
A. E. Bely
cells lose their basal cytoskeleton network and also
lose the adherence-structures anchoring them to
other epidermal cells. The muscle layer below also
changes extensively; muscle-cells near the wound
lose contact with the basement membrane and lose
the membranes that anchored them to other musclecells. They then autotomize their contractile apparatus and dedifferentiate to a myoblast-like form.
Epidermal and muscle-cells then both begin to proliferate (see below ‘‘Cellular origins of the blastema’’). The important changes both in epidermal
and in mesodermal cells, and especially their loss
of connections to other cells, presumably allow for
rearrangement and migration of cells as well as their
proliferation during wound-healing and early stages
of regeneration. Within a few days after amputation,
the extracellular matrix reforms between the woundepidermis and the underlying mesoderm. A thin
basement membrane is evident within about 2 dpa,
collagen is strongly upregulated at the wound site by
4–5 dpa, and the basement membrane has fully reformed by about 1 week after amputation. At that
time, ectodermal and mesodermal cells both re-develop contacts with the basement membrane and adherence-structures with each other (Dupin et al.
1991).
In the leech Hirudo, an annelid that wound-heals
and scars but does not regenerate structures, there
are also significant changes in cell–cell adhesion,
cytoskeletal organization, and the extracellular
matrix following injury (Huguet and Molinas
1994, 1996; Tettamanti et al. 2004). Wounds in
the body wall are sealed largely by vasocentral (fibroblast) cells; these cells migrate in to form a
wound-plug, contract the wound’s edges by modifying their cytoskeleton and their cell–cell contacts,
and secrete an extracellular scaffold of proteins, largely collagen, that further seals the wound.
Although the cells eventually leave the site of the
wound, the collagen scaffold is not destroyed and
remains as a scar.
Available data thus suggest that the extracellular
matrix may play an important role in healing
wounds and in tissue-level changes at the wound’s
site, but only a few studies have investigated such
questions in annelids. Studies of a broader range of
annelids are needed to identify the general patterns
of change in the extracellular matrix and in the
remodeling of tissues. Furthermore, experimental
work is needed to determine whether the changes
in the extracellular matrix and nearby tissues
are merely correlated with each other or whether
they are actually causally linked, and, if the latter,
Annelid regeneration
what the cause-and-effect relationships are between
them.
Formation of the blastema
Initiation of the blastema
Studies in a range of annelids indicate that interactions of tissues at the site of the wound can be important for initiating a blastema. In particular,
several lines of evidence suggest that the severed
end of the ventral nerve cord can elicit regeneration
and proliferative outgrowth. First, surgeries in which
a severed end of the nerve cord is diverted to a
wound in the body wall in Eisenia lead to the formation of an ectopic blastema that differentiates into
a lateral head (Avel 1959). Thus, interaction between
a cut ventral nerve cord and an injured body wall
may be sufficient to initiate formation of a blastema.
Second, amputation in a body region that has had
the ventral nerve cord removed can inhibit regeneration, often completely (Avel 1961). This effect has
been shown to depend both on the timing of removal of the ventral nerve cord and on the distance
between the severed nerve cord and the site of the
wound. When the ventral nerve cord is removed following amputation in Eisenia, the degree of inhibition decreases as the time between amputation and
removal of the nerve cord increases, until, beyond a
certain time (4–5 days, or mid-blastemal stage) the
negative effect of removing the nerve cord on regeneration disappears. In Eurythoe, extirpation of the
ganglion closest to the site of the wound retards regeneration, whereas extirpation of the two nearest
ganglia abrogates it completely, leading only to healing (Müller et al. 2003). Thus, the severed nerve cord
appears to be important for initiating and sustaining
the early stages of regeneration in a temporally and
spatially dependent manner. Third, in several groups
of annelids investigated, the earliest cell proliferation
detected during regeneration occurs near the severed
ventral nerve cord (Coulon and Thouveny 1984;
Müller et al. 2003). Finally, the presence of nerve
ganglia or nerve fibers correlates with regenerative
ability in several species. In Chaetopterus, excising
the portion of a long mid-segment (segment 12)
that has the ventral ganglion leads to regeneration,
while excising a similar-sized portion of the same
segment but that does not contain the ventral ganglion leads to failure to regenerate (Hill 1972). In
Owenia, when anterior amputation occurs in a regeneration-competent region of the body, nerve
fibers extend from the ventral nerve cord into the
site of the wound, with the fibers running between
the epidermis and mesoderm. However, when the
693
cut occurs at an axial position incapable of regeneration, no nerve fibers extend subepidermally (Coulon
and Thouveny 1984).
Although a considerable amount of data suggests
that key aspects of annelid regeneration are dependent on the nerve cord, there also appear to be some
components of regeneration that occur independently of this structure as well as some species in
which regeneration can still proceed without a ventral nerve cord at the site of the wound. In Eisenia,
in the absence of the nerve cord at the wound, ectodermal and mesodermal growth is inhibited but
considerable endodermal growth still proceeds (Avel
1961). Indeed, in individuals in which the nerve cord
is removed over several segments closest to the site of
amputation, the anterior limit of the gut still grows
substantially, even though a blastemal outgrowth
never forms. Eventually the tip of the gut grows ventrally and then curves posteriorly, extending until it
contacts the ventral nerve cord several segments
back. Despite the growth and formation of a pharynx
anlagen in these animals, pharyngeal differentiation
does not occur, suggesting that signals from the
nerve cord or the blastema itself are required for
the differentiation process. Additional data also suggest a role for tissues other than the nerve cord in
initiating regeneration. In Tubifex, it is the severing
of the gut at the site of the wound that is implicated
in initiating blastemal outgrowth, although comparable studies in Eisenia show no such requirement (reviewed by Avel 1959). Finally, in Nereis, a blastemal
outgrowth does form following amputation within a
body region in which the nerve cord has been removed (although the resulting blastema is abnormal,
having apparently incomplete dorsal–ventral polarity) (Combaz and Boilly-Marer 1976; WattezCombaz 1995), in contrast to the several examples
noted in the preceding paragraph in which comparable cuts fail to elicit a blastema.
Although the kinds of surgical manipulations described above provide important insights into the
regenerative process, they should also be interpreted
with some caution. First, excisions of certain tissues
(e.g., nerve cord and gut) can alter the spatial relationships among remaining tissues in operated animals; the collapsing of tissues may still allow for
some contact between relevant tissues (Fitzharris
and Lesh 1969). Second, the degree of inhibition of
regeneration can vary even across closely related species. For example, comparable experiments performed by the same experimenter on two species
of Eisenia resulted in considerably different outcomes: amputation within a body region in which
the ventral nerve cord was removed led to
694
regeneration failing in all individuals in one species
but resulted in a small, slowly growing blastema in
about one quarter of the animals in a second species
(Avel 1961). Thus, carefully controlled experiments
and investigations in a broader range of annelids are
needed to strengthen the conclusions of these
studies.
In addition to the effect of interactions of tissues
at the site of the wound, a number of studies suggest
that the nervous system, and the brain in particular,
can have long-distance effects on regeneration, possibly through the release of hormones promoting or
inhibiting regeneration. The effect of manipulations
such as removal of the brain varies considerably
among species, however. In Nereis, removal of the
brain inhibits posterior regeneration (Golding
1967), yet in a number of other annelids
(Branchiomma, Chaetopterus, Eophila, and Eulalia)
(Gallissi 1965; Hill 1972; Olive and Moore 1975)
posterior regeneration is still possible and may even
be promoted by removing the brain. In
Allolobophora, the effect of removing the brain on
posterior regeneration has been found to depend
on the developmental stage of the animal: in juveniles, presence of the brain promotes regeneration,
whereas in sexually mature (clitellated) individuals,
presence of the brain inhibits regeneration (AlonsoBedate and Sequeros 1985). If these effects on regeneration are indeed caused by hormones emanating
from the brain, then hormonal changes at sexual
maturity may be mediating these stage-dependent
effects.
In summary, a large amount of experimental and
correlational evidence suggests that the nervous
system has a particularly important role in initiating
and sustaining at least certain aspects of blastemal
development. Many manipulations that have been
performed are ‘‘cut-and-paste’’ type experiments
and are necessarily rather crude, yet effects on regeneration can be dramatic (e.g., complete failure of
regeneration). Replicating some of these experiments
in a broader range of annelids would be useful for
assessing how general their results are. Determining
what molecules are responsible for both short-range
and long-range signals that inhibit or promote regeneration is also an important direction of research to
pursue.
Polarity of the blastema
Annelid blastemas have either anterior (head) or
posterior (tail) axial polarity; no mixed-polarity phenotypes are known (Avel 1959). Indeed, even amputating in a body region in which axial polarity should
A. E. Bely
differ across the surface of the wound (following a
body-wall graft) produces a blastema with a single
polarity (reviewed by Avel 1959). Although the cues
that assign a blastema its axial polarity remain to be
elucidated, blastemas that regenerate with incorrect
polarity can be generated, indicating that polarity can
be mis-assigned. For example, heads have been
formed at posterior cut sites in Sabella following
treatment with colchicine (a tubulin-polymerization
inhibitor that blocks mitosis, among other effects)
(Fitzharris and Lesh 1969), in Dero when tiny fragments (one to three segments) are excised (Hyman
1916), occasionally in Enchytraeus following simple
amputations (Myohara et al. 1999), and at relatively
high frequencies in this same Enchytraeus species
when amputated individuals are treated with an anesthetic that inhibits corrective (post-amputation)
autotomy (Kawamoto et al. 2005). Regarding the
last of these, corrective autotomy is a post-injury
self-amputation that places the site of the wound at
a stereotypical intra-segmental plane; thus, in this
species, initiating a blastema from the wrong intrasegmental position may be what leads to mis-assigned polarity. Based on available records of biaxial
worms, double-headed worms, with a head regenerated at the posterior end, appear to be much more
common than double-tailed worms, with a tail that
regenerated at the anterior end. The only examples of
double-tailed individuals that could be found are in
Eisenia (Gates 1949, 1950).
The assignment of dorsal–ventral polarity of the
blastema has also been investigated using grafting
experiments (Combaz and Boilly-Marer 1976;
Wattez-Combaz 1995). These studies have been conducted on Nereis, in which a blastemal outgrowth
can form even in the absence of the ventral nerve
cord. Amputations made in body regions from which
the ventral nerve cord had been removed yielded an
abnormal blastema that formed a pygidium (posterior tip) and segments, but which were smaller and
with fewer structures than normal. To investigate the
dorsal–ventral polarity of these ‘‘aneurogenic’’ blastemas, patches of body wall from them were grafted
onto normal hosts. Because parapodia (segmental,
lateral projections of the body wall) are thought to
form at the junction between dorsal and ventral
body-wall tissue, ectopic parapodia are expected to
form at boundaries between patches of dorsal tissue
and ventral tissue, but not between two patches of
dorsal tissue or between two patches of ventral
tissue. Results showed that ectopic parapodia
formed at the edges of grafts when body wall from
anywhere in the aneurogenic blastema was grafted
onto the ventral region, but not onto the dorsal
Annelid regeneration
region, of the host. This, and other, experiments (including the grafting of a supernumerary ventral
nerve cord) suggest that the body wall of the blastema in Nereis does not have an inherent dorsal–
ventral polarity but instead has a default state of
dorsal polarity; the ventral nerve cord then imposes
ventral polarity in tissue that is in close proximity.
Results consistent with these findings were also
found in earthworms (Avel 1942; Okada and
Kawakami 1943; Avel 1947; cited in WattezCombaz 1995). In these annelids, a blastema with
normal dorsal–ventral polarity was produced from
a stump that was not expected to have dorsal–ventral
polarity in the body wall, a cuff of body-wall tissue
with all dorsal or all ventral polarity having been
grafted at the site prior to amputation. These experiments suggest that the dorsal–ventral polarity of the
blastema does not inherit this polarity solely from
the body wall at the site of the wound.
In summary, knowledge of the assignment of blastemal polarity remains rudimentary and stems from
very few studies. Both anterior–posterior and dorsal–
ventral polarity can be mis-assigned through experimental manipulations, but the underlying mechanisms assigning polarity remain to be elucidated.
Identifying the molecular signals that assign polarity
is an important direction to pursue. In planaria, a
single signaling pathway (Wnt/-catenin) is sufficient
for imposing the axial polarity of the blastema and,
when mis-expressed, can result in a blastema with
the wrong polarity (Gurley et al. 2008). Whether a
single pathway, and possibly even the same pathway,
is involved in this process in annelids will be an
interesting question to address.
Cellular origins of the blastema
In a broad range of annelids, the location of cells in
mitosis and the incorporation of thymidine-analogs
indicate that ectoderm, mesoderm, and endoderm all
proliferate at the wound, consistent with the hypothesis that all three tissue layers contribute to the blastema (Buongiorno-Nardelli and Thouveny 1966; Hill
1970; Coulon and Thouveny 1984; Paulus and
Müller 2006; Zattara and Bely 2011). Consistently,
there is also a significant delay (usually on the
order of at least a day) between the time of amputation and the earliest evidence of proliferation; this
delay suggests that some preparatory steps, such as
wound-healing and dedifferentiation, may need to
occur prior to blastemal growth.
The epidermis of the regenerated body wall, the
cell bodies of the nervous system, and the foregut are
ectodermal tissues that, based on time-series studies
695
of fixed tissues, appear to derive from the epidermis
at the wound. The epidermis covering the wound
and later that of the blastema is highly proliferative
and this proliferation appears to be the only source
of new epidermal cells (Boilly 1969b; Hill 1970;
Zattara and Bely 2011). The ventral epidermis of
the blastema in Limnodrilus generates some cells
with little cytoplasm that appear to move inwards
to form new nerve ganglia. In this and a number
of other annelids, the ectoderm thus appears to furnish internal cells that will become neural cells
(Cornec et al. 1987; Cornec 1990). In a broad
range of annelids, nerve fibers thought to originate
from cells in the old ventral nerve cord also invade
the blastema and form some of the neural tracts of
the new brain and ventral nerve cord (Müller 2004;
Müller and Henning 2004; Paulus and Müller 2006;
Zattara and Bely 2011). Regarding the foregut, the
regenerated buccal cavity appears to originate from
invaginated surface ectoderm at the site of the
wound (Hyman 1916, 1940; Zattara and Bely
2011). Although data from a number of sources suggest that ectodermal structures consistently arise
from the wound’s epidermis, one study does suggest,
based on the histology of fixed tissue, that dissepimentary cells may contribute to nerve ganglia
(Cornec et al. 1987); this possibility warrants further
investigation since a mesodermal source for regenerated neural cells is unexpected.
Histological studies of fixed tissue at different
stages of regeneration also suggest that the source
of new mesoderm is primarily local dedifferentiated
muscle-cells and/or dissepimentary cells (cells of the
lining of the septa) that migrate to the wound, although endothelial cells (associated with blood vessels) have also been implicated as a source. Muscle
regeneration has been studied in some detail during
anterior regeneration in Owenia (Fontes et al. 1983).
In this annelid, histological studies suggest that longitudinal muscles at the site of the wound dedifferentiate into myoblast-like cells, proliferate, and then
redifferentiate into new muscle-cells. During dedifferentiation, the muscle-cells, which are mononucleate, lose their membrane-anchoring structures and
autotomize, separating into a nucleated part and
contractile part. The shed, enucleated part degenerates in the coelom or is phagocytized by coelomocytes; the nucleated part gives rise to dedifferentiated
muscle-cells that migrate to the wound, proliferate
there, and redifferentiate. Interestingly, according to
this study, this process involves major changes in
cellular morphology but little change in molecular
expression, suggesting that morphological dedifferentiation of muscle-cells may not alter the terminal
696
differentiation program. While results from Owenia
are largely consistent with findings in several other
annelids (Hill 1970), an alternative mechanism of
muscle formation has been described in Hirudo
(Grimaldi et al. 2006, 2009). In this annelid, regeneration of muscle has been investigated following the
making of lesions in the body wall (since Hirudo
does not regenerate segments) and occurs considerably longer after injury (1 month after body lesions
in Hirudo versus a few days after transverse amputation in many other annelids). In Hirudo, the
mononucleated muscle-cells can proliferate slowly
but most of the regenerated muscle-cells are thought
to arise from a population of precursor cells, normally associated with the blood vessels in uninjured
animals, which migrate to the wound and differentiate into new muscle. Recruitment of these cells to the
regenerating areas is thought to be mediated, at least
in part, by a growth factor (Grimaldi et al. 2009).
Origin of new mesoderm from cells associated with
blood vessels has also been suggested for some polychaetes (Hill 1970). Additional studies are needed to
determine the generality of either of these processes
in the regeneration of muscles.
Migration of dissepimentary cells to the wound
has been inferred in a range of annelids based on
histology of fixed tissue. Many studies have suggested
that these cells contribute to the coelomic lining (including septum walls) of new segments and to new
muscles (reviewed by Cresp 1964; Hill 1970).
Although migration of such cells in ‘‘polychaetes’’
is reported only over short distances (within one
segment from the wound) (Cresp 1964; Hill 1970),
in a number of clitellates a much more extensive
migration is inferred. In clitellates, amputation appears to induce activation of dissepimentary cells
over a number of segments and these cells then
appear to migrate intersegmentally along the ventral
nerve cord to the wound, where they finally proliferate (reviewed by Stephan-Dubois 1954; Cornec
et al. 1987). As mentioned above, these migrating
cells are sometimes referred to as neoblasts. It remains to be determined whether these cells and
this process are characteristic of clitellates or merely
most easily seen or more extensive in this group. The
fate of these intersegmentally migrating dissepimentary cells is also difficult to ascertain and has been
heavily debated (Stephan-Dubois 1954; Hill 1970;
Jamieson 1981; Myohara 2012). To help resolve
this long-standing issue, studies that directly trace
the fates of these migrating cells are critically needed.
The source of the germ line in regenerated segments is an important question that also needs
greater study. A study in Enchytraeus suggests,
A. E. Bely
based on histology and gene expression, that germline cells are distributed along the length of the body
and that, during anterior regeneration, such cells migrate anteriorly to the wound from several segments
away to re-establish the germ line in the regenerated
gonad-bearing segments (Tadokoro et al. 2006). By
contrast, in Capitella, expression patterns for the
same molecular marker used in the Enchytraeus
study (a piwi homolog) do not suggest migration
of marker-positive cells to the wound (Giani et al.
2011). Whether this difference is because the mechanism of germ-line regeneration differs between these
two species or because the markers themselves are
expressed differently between them remains to be
determined. Additional markers and methods to
identify germ-line cells, as well as studies in a
broader range of annelids, are needed to address
the important question of how the germ line is
regenerated in annelids.
Finally, new endodermal cells appear to come
from old endodermal cells in annelids. Proliferation
assays indicate that the cut tip of the gut proliferates
and histological studies suggest this tissue gives rise
to the new gut (Hill 1970). Importantly, a recent
study employing a cell-tracing technique in
Lumbriculus has confirmed that old gut tissue does
indeed contribute to new gut tissue (Tweeten and
Reiner 2012). Cell-tracing studies such as this one
are key to validating inferences that previously have
been based solely on histological data.
In summary, although the question of what cells
contribute to regenerated structures in annelids has
long been debated (Randolph 1891, 1892; Boilly
1969b; Hill 1970), a working model that appears to
be general for annelids is emerging. Specifically, all,
or most, of the blastema is thought to arise from
cells at, or very near, the site of the wound; only
some portion of the mesoderm in one group of annelids (clitellates) is thought possibly to originate
from cells that migrate from several segments away
(with no evidence for such a long-distance migration
from polychaetes). Source-cells appear to dedifferentiate, proliferate at the site of the wound, and then
re-differentiate, with each layer of tissue, apparently
giving rise to its own layer in the regenerate. Several
key questions remain regarding the cellular origins of
the blastema, however. First, whether source-cells
fully dedifferentiate before proliferating is an important question that needs further investigation; several
authors have indicated that cells (specifically of the
epidermis and of the muscle layer) may maintain
some level of differentiation, and even retain characteristic gene expression, as they proliferate (Fontes
et al. 1983; Cornec et al. 1987). Second, the breadth
697
Annelid regeneration
of the potential fates of the source-cells is unknown.
This question is particularly relevant for regeneration
of the mesoderm, for which available data suggest
some source-cells can differentiate into several types
of mesodermal structures, and the same type of cell
(i.e., muscle) can arise from several types of sourcecells. Methods to trace fates of specific cells or
groups of cells are needed to address this question,
in mesoderm as well as in other tissue layers. Finally,
whether stem cells contribute to regenerated structures in annelids, as is now well documented in some
other animals, most strikingly in planaria (Tanaka
and Reddien 2011), remains debated and in need
of further study. There is strong evidence that in
annelids local, dedifferentiated tissues make a significant contribution to the regenerate, but there is still
no strong evidence that stem cells (pluripotent or
tissue-specific) contribute to regenerated somatic
structures in this group.
Conclusions
The cellular events of regeneration have been studied
in a remarkably broad range of annelid species.
Unlike the study of regeneration in many other
groups, efforts have not concentrated on just one
or two model systems but instead have been distributed over many species. Although this distribution of
effort is not without drawbacks (i.e., the picture that
has emerged is broad yet still relatively shallow), a
benefit is that generalities are apparent for a number
of aspects of regeneration across the phylum. Thus, it
is possible to compile a broad summary picture of
how regeneration typically proceeds in annelids, as
has been done here. Whether regeneration in annelids is, in fact, more uniform as a process across the
phylum than it is in other phyla remains to be seen.
Many important questions remain regarding the
early cellular events of regeneration in annelids.
Some of these are fundamental, yet understudied
or difficult to address. For example, how does the
early cellular response to injury differ depending on
the type of injury, the severity of injury, and the
potential for regeneration? What is the source and
nature of cells that contribute to the regenerated
structures? Have the major events of regeneration
and the cellular origins of regenerated structures
evolved across annelid phylogeny, and if so, how?
To help address these, and other, important questions, new techniques and approaches should be applied to the study of regeneration in annelids. There
is a particular need to develop and apply methods to
identify and manipulate specific cell types (e.g.,
Grimaldi et al. 2011), to image the dynamics of
cells and tissues over long durations in vivo (e.g.,
Zattara 2012), and to trace the fates of cells (e.g.,
Weisblat et al. 1978; Meyer et al. 2010; Tweeten
and Reiner 2012) during regeneration. Employing
such new approaches is likely to yield major advances in our understanding of the cellular basis of
regeneration in annelids.
Acknowledgments
The author thanks Eduardo Zattara and Duygu
Özpolat for extensive and helpful discussions about
regeneration in annelids and Mansi Srivastava,
Deirdre Lyons, and Mark Martindale for the opportunity to participate in this symposium.
Funding
Participation in the symposium was supported by
the Society for Integrative and Comparative Biology.
References
Alonso-Bedate M, Sequeros E. 1985. Suggested regulatory
mechanisms for caudal regeneration in Allolobophora molleri (Annelida: Oligochaeta). Comp Biochem Physiol
81:225–8.
Avel M. 1942. Sur l’autonomie de la differenciation des regenerats cephaliques chez les Lombrics. CR Acad Sci Paris D
215:333–4.
Avel M. 1947. Les facteurs de la regeneration chez les
Annelides. Rev Suisse Zool 54:219–35.
Avel M. 1959. Classe des Annélides Oligochètes.
In: Grassé PP, editor. Traité de Zoologie, Vol. V. Paris:
Masson et Cie. p. 224–470.
Avel M. 1961. L’influence du systeme nerveux sur la regeneration chez les urodeles et les oligochetes. Bull Soc Zool Fr
86:464.
Bely AE. 2006. Distribution of segment regeneration ability in
the Annelida. Integr Comp Biol 46:508–18.
Bely AE, Sikes JM. 2010. Latent regeneration abilities persist
following recent evolutionary loss in asexual annelids. Proc
Natl Acad Sci USA 107:1464–9.
Berrill NJ. 1952. Regeneration and budding in worms. Biol
Rev 27:401–38.
Bilej M. 1994. Cellular defense mechanisms. In: Vetvicka V,
Sima P, Cooper EL, Bilej M, Roch P, editors. Immunology
of annelids. Ann Arbor: CRC Press. p. 167–200.
Boilly B. 1969a. Origine des cellules regeneratrice chez Nereis
diversicolor O. F. Muller (Annelide Polychete). Wilhelm
Roux’s Arch Dev Biol 162:286–305.
Boilly B. 1969b. Sur l’origine des cellules regeneratrices ches
les annelides polychetes. Arch Zool Exp Gen 110:127–43.
Buongiorno-Nardelli M, Thouveny Y. 1966. Donnees histochimiques et enzymologiques sur la regeneration anterieure
de l’Annelide Polychete Owenia fusiformis (Delle Chiaje).
Bull Biol Fr Belg 100:487–517.
Burke JM. 1974. Wound healing in Eisenia foetida
(Oligochaeta). III. A fine structural study of the role of
non-epidermal tissues. Cell Tissue Res 154:83–102.
698
Cameron GR. 1932. Inflammation in earthworms. J Pathol
Bacteriol 35:933.
Clark ME, Clark RB. 1962. Growth and regeneration in
Nephtys. Zool Jahrb 70:24–90.
Combaz A, Boilly-Marer Y. 1976. Mise en evidence de la
nature dorsale de la paroi de corps du regenerat aneurogenique de Nereis pelagica L. (Annelide Polychete).
[Experimental evidence of dorsal nature of body
wall in aneurogenic regenerate of Nereis pelagica
L. (Annelida: Polychaeta)]. C R Hebd Seances Acad Sci
Serie D 283:785–8.
Cornec JP. 1984. Modifications ultrastructurales apres amputation dans le territoire de regeneration posterieure de l’hirudine rhynchobdelle Helobdella stagnalis; mise en place et
evolution du bouchon cicatriciel. Arch Anat Microsc
73:269–89.
Cornec JP. 1990. Ultrastructural study of regenerative buds
after amputation of posterior sucker of an adult rhynchobdellid hirudinean, Helobdella stagnalis. Can J Zool
68:303–12.
Cornec JP, Cresp J, Delye P, Hoarau F, Reynaud G. 1987.
Tissue responses and organogenesis during regeneration
in the oliogochaete Limnodrilus hoffmeisteri (Clap.). Can J
Zool 65:403–14.
Coulon J, Thouveny Y. 1984. Relation entre l’innervation et
l’activite proliferatrice des cellules blastematique au cours
de la regeneration de l’annelide polychete Owenia fusiformis: etudes ultrastructurale et autoradiographique. Arch
Anat Microsc 73:45–56.
Coulon J, Diano M, Arsanto J-P, Thouveny Y. 1989.
Remodeling processes during anterior regeneration of
Owenia fusiformis (Polychaeta, Annelidae): a morphological
and immunocytochemical survey. Can J Zool 67:994–1005.
Cresp J. 1964. Etudes experimentales et histologiques sur la
regeneration et le bourgeonnement chez les serpulides
Hydroides norvegica (Gunn.) et Salmacina incrustans
(Clap.P). Bull Biol Fr Belg 98:3–152.
Dupin F, Coulon J, Le Parco Y, Fontes M, Thouveny Y. 1991.
Formation of the extracellular matrix during the epimorphic anterior regeneration of Owenia fusiformis: autoradiographical and in situ hybridization studies. Int J Dev
Biol 35:109–19.
Fitzharris TP, Lesh GE. 1969. Gut and nerve-cord interactions
in sabellid regeneration. J Embryol Exp Morphol
22:279–93.
Fontes M, Coulon J, Delgrossi MH, Thouveny Y. 1983.
Muscle dedifferentiation and contractile protein synthesis
during post-traumatic regeneration by Owenia fusiformis
(polychaete annelid). Cell Differ 13:267–82.
Gallissi A. 1965. Vacuite du tube digestif et regeneration
caudale en l’absence des segments anterieurs chez le lumbricide Eophila dollfusi Tetry. C R Hebd Seances Acad Sci
261:817.
Gates GE. 1949. Regeneration in an earthworm, Eisenia foetida (Savigny), 1826. I. Anterior regeneration. Biol Bull
96:129–39.
Gates GE. 1950. Regeneration in an earthworm, Eisenia foetida (Savigny) 1826. II. Posterior regeneration. Biol Bull
98:36–45.
Giani VC Jr, Yamaguchi E, Boyle MJ, Seaver EC. 2011.
Somatic and germline expression of piwi during
A. E. Bely
development and regeneration in the marine polychaete
annelid Capitella teleta. EvoDevo 2:10.
Golding DW. 1967. Neurosecretion and regeneration in
Nereis. I. Regeneration and the role of the supraesophageal
ganglion. Gen Comp Endocrinol 8:348–55.
Goss RJ. 1969. Principles of regeneration. New York (NY):
Academic Press.
Grdisa M. 2010. Mechanism of wound healing in annelids.
Invert Surviv J 7:192–7.
Grimaldi A, Banfi S, Gerosa L, Tettamanti G, Noonan DM,
Valvassori R, de Eguileor M. 2009. Identification, isolation
and expansion of myoendothelial cells involved in leech
muscle regeneration. PLoS One 4:e7652.
Grimaldi A, Banfi S, Vizioli J, Tettamanti G, Noonan DM, de
Eguileor M. 2011. Cytokine loaded biopolymers as a novel
strategy to study stem cells during wound-healing processes. Macromol Biosci 11:1008–19.
Grimaldi A, Tettamanti G, Perletti G, Valvassori R, de
Eguileorl M. 2006. Hematopoietic cell formation in leech
wound healing. Curr Pharm Des 12:3033–41.
Gurley KA, Rink JC, Sánchez Alvarado A. 2008. -catenin
defines head versus tail identity during planarian regeneration and homeostasis. Science 319:323–7.
Herlant-Meewis H. 1964. Regeneration in annelids. Adv
Morphogen 4:155–215.
Huguet G, Molinas M. 1994. The pseudoblastema in the
wound healing process of the leech Hirudo medicinalis L.
(Hirudinea): changes in cell junctions. J Exp Zool
269:23–36.
Huguet G, Molinas M. 1996. Myofibroblast-like cells and
wound contraction in leech wound healing. J Exp Zool
275:308–16.
Hill SD. 1970. Origin of the regeneration blastema in polychaete annelids. Am Zool 10:101–12.
Hill SD. 1972. Caudal regeneration in absence of a brain in
two species of sedentary polychaetes. J Embryol Exp
Morphol 28:667.
Hyman LH. 1916. An analysis of the process of regeneration
in certain microdrilous oligochaetes. J Exp Zool 20:99–163.
Hyman LH. 1940. Aspects of regeneration in annelids. Am
Nat 74:513–27.
Jamieson BGM. 1981. The ultrastructure of the oligochaeta.
New York (NY): Academic Press Inc.
Kawamoto S, Yoshida-Noro C, Tochinai S. 2005. Bipolar head
regeneration induced by artificial amputation in
Enchytraeus japonensis (Annelida, Oligochaeta). J Exp
Zool Part A Comp Exp Biol 303A:615–27.
Lesiuk N, Drewes C. 1999. Autotomy reflex in a freshwater
oligochaete,
Lumbriculus
variegatus.
Hydrobiologia
406:253–61.
Meyer NP, Boyle MJ, Martindale MQ, Seaver EC. 2010. A
comprehensive fate map by intracellular injection of identified blastomeres in the marine polychaete Capitella teleta.
EvoDevo 1:8.
Müller MCM. 2004. Nerve development, growth and differentiation during regeneration in Enchytraeus fragmentosus
and Stylaria lacustris (Oligochaeta). Dev Growth Differ
46:471–8.
Müller MCM, Berenzen A, Westheide W. 2003. Experiments
on anterior regeneration in Eurythoe complanata
(‘‘Polychaeta’’, Amphinomidae): reconfiguration of the
Annelid regeneration
nervous system and its function for regeneration.
Zoomorphology 122:95–103.
Müller MCM, Henning L. 2004. Ground plan of the polychaete brain—I. Patterns of nerve development during regeneration in Dorvillea bermudensis (Dorvilleidae). J Comp
Neurol 471:49–58.
Myohara M. 2012. What role do annelid neoblasts play?
A comparison of the regeneration patterns in a
neoblast-bearing and a neoblast-lacking enchytraeid oligochaete. PLoS One 7:e37319.
Myohara M, Yoshida-Noro C, Kobari F, Tochinai S. 1999.
Fragmenting oligochaete Enchytraeus japonensis: a new material for regeneration study. Dev Growth Differ 41:549–55.
Okada YK, Kawakami IK. 1943. Transplantation experiments
in the earthworm Eisenia foetida (Savigny) with special remarks on the inductive effects of the nerve and on the
differentiation of the body wall. J Fac Sci Tokyo Imp
Univ Sec IV 6:25–96.
Olive PJW, Moore FR. 1975. Hormone independent regeneration in Eulalia viridis (Polychaeta—Phyllodocidae). Gen
Comp Endocrinol 26:259–65.
Paulus T, Müller MCM. 2006. Cell proliferation dynamics and
morphological differentiation during regeneration in
Dorvillea bermudensis (Polychaeta, Dorvilleidae). J Morph
267:393–403.
Randolph H. 1891. The regeneration of the tail in
Lumbriculus. Zool Anz 14:154–6.
Randolph H. 1892. The regeneration of the tail in
Lumbriculus. J Morph 7:317–44.
Stephan-Dubois F. 1954. Les neoblastes dans la regeneration
posterieur des oaligochetes microdriles. Bull Biol Fr Belg
88:182–247.
Stephenson J. 1930. The oligochaeta. Oxford: Clarendon
Press.
Struck TH, Paul C, Hill N, Hartmann S, Hosel C, Kube M,
Lieb B, Meyer A, Tiedemann R, Purschke G, et al. 2011.
Phylogenomic analyses unravel annelid evolution. Nature
471:95–U113.
Tadokoro R, Sugio M, Kutsuna J, Tochinai S, Takahashi Y.
2006. Early segregation of germ and somatic lineages
699
during gonadal regeneration in the annelid Enchytraeus
japonensis. Curr Biol 16:1012–7.
Tanaka EM, Reddien PW. 2011. The cellular basis for animal
regeneration. Dev Cell 21:172–85.
Tettamanti G, Grimaldi A, Rinaldi L, Arnaboldi F, Congiu T,
Valvassori R, de Eguileor M. 2004. The multifunctional role
of fibroblasts during wound healing in Hirudo medicinalis
(Annelida, Hirudinea). Biol Cell 96:443–55.
Tweeten KA, Reiner A. 2012. Characterization of serine proteases of Lumbriculus variegatus and their role in regeneration. Invert Biol 131:322–32.
Vetvicka V, Sima P, Cooper EL, Bilej M, Roch P. 1994.
Immunology of annelids. Ann Arbor (MI): CRC Press.
Wattez-Combaz A. 1995. Chaine nerveuse et polarite dorsoventrale des regenerats posterieurs chez l’Annelide
Polychete Nereis pelagica [Involvement of the ventral
nerve cord in determination of dorsoventral polarity
during caudal regeneration in the polychaete annelid
Nereis pelagica]. Ann Sci Nat Zool Paris 16:97–103.
Weigert A, Helm C, Meyer M, Nickel B, Arendt D,
Hausdorf B, Santos SR, Halanych KM, Purschke G,
Bleidorn C, et al. 2014. Illuminating the base of the annelid
tree using transcriptomics. Mol Biol Evol 31: 1391–401.
Weisblat DA, Sawyer RT, Stent GS. 1978. Cell lineage analysis by
intracellular injection of a tracer enzyme. Science 202:1295–8.
Yoshida-Noro C, Myohara M, Kobari F, Tochinai S. 2000.
Nervous system dynamics during fragmentation and regeneration in Enchytraeus japonensis (Oligochaeta, Annelida).
Dev Genes Evol 210:311–9.
Zattara EE. 2012. Regeneration, fission and the evolution of
developmental novelty in naid annelids. PhD Thesis.
Biology Department. University of Maryland, College Park.
Zattara EE, Bely AE. 2011. Evolution of a novel developmental trajectory: fission is distinct from regeneration in the
annelid Pristina leidyi. Evol Dev 13:80–95.
Zattara EE, Bely AE. 2013. Investment choices in postembryonic development: quantifying interactions among
growth, regeneration, and asexual reproduction in the
annelid Pristina leidyi. J Exp Zool Part B Mol Dev Evol
320:471–88.