Annelids as models of germ cell and gonad regeneration

Abstract

Germ cells (reproductive cells and their progenitors) give rise to the next generation in sexually reproducing organisms. The loss or removal of germ cells often leads to sterility in established research organisms such as the fruit fly, nematodes, frog, and mouse. The failure to regenerate germ cells in these organisms reinforced the dogma of germline–soma barrier in which germ cells are set-aside during embryogenesis and cannot be replaced by somatic cells. However, in stark contrast, many animals including segmented worms (annelids), hydrozoans, planaria, sea stars, sea urchins, and tunicates can regenerate germ cells. Here I review germ cell and gonad regeneration in annelids, a rich history of research that dates back to the early 20th century in this highly regenerative group. Examples include annelids from across the annelid phylogeny, across developmental stages, and reproductive strategies. Adult annelids regenerate germ cells as a part of regeneration, grafting, and asexual reproduction. Annelids can also recover germ cells after ablation of germ cell progenitors in the embryos. I present a framework to investigate cellular sources of germ cell regeneration in annelids, and discuss the literature that supports different possibilities within this framework, where germ–soma separation may or may not be preserved. With contemporary genetic-lineage tracing and bioinformatics tools, and several genetically enabled annelid models, we are at the brink of answering the big questions that puzzled many for over more than a century.

Graphical Abstract

1 |. INTRODUCTION

Animals establish a set of cells, the germline, that pass on their genetic material from one generation to the next. Germline is typically viewed as strictly separate from the rest of the cells that make up the organism’s body, collectively called soma (Klein, 1917; Nussbaum, 1880; Weismann, 1893). This germline–soma separation has been argued as necessary for several reasons including for maintaining genome integrity in the germ cell lineage, for epigenetic reprogramming, and for maintaining telomere length so that organisms do not age with each successive generation (Bloom et al., 2019; Castañeda et al., 2011; Eguizabal et al., 2016; Fernández & González-Vasconcellos, 2018; Gleason et al., 2018; Gruhn et al., 2023; Hajkova, 2011; Labbé et al., 2017; Ozturk, 2015; Wright et al., 1996). In contrast, somatic cells are viewed as an aging cell population and once the germline is segregated from soma during embryogenesis, somatic cells do not contribute to the reproductive cell pool (Buss, 1983; Johnson et al., 2011; Monaghan & Metcalfe, 2019; Weismann, 1893). While this dichotomous view of germline and soma holds true for many extensively studied model organisms (e.g., Drosophila, Caenorhabditis elegans, mouse), many organisms do not fit into this strict mold (Bounoure, 1940; Buss, 1988; Heys, 1931; Özpolat et al., 2016; Stéphan-Dubois, 1964; Stephenson, 1930; Tadokoro, 2018; Weisblat, 2006; Wessel et al., 2020). In some organisms, the germline can be recovered upon loss of reproductive organs or ablation of the primordial germ cells. In some of these cases, the source cells that replace the lost germline are of somatic origin (Kawamura et al., 2011; Wessel et al., 2014a; Yajima & Wessel, 2011; Yoshida et al., 2017). In other organisms, germline can be re-established after many cycles of asexual, agametic reproduction (Brusca & Brusca, 2003; Ghiselin, 1987; Sköld et al., 2009; Sunanaga et al., 2006). Different groups of animals have evolved a range of strategies for this plasticity. Sponges, Hydra, and planarian flatworms have pluripotent stem cell populations in adults that continuously produce and regenerate germ cells (Holstein, 2023; Isaeva, 2011; Issigonis & Newmark, 2019; Wang et al., 2007). Sea star larvae recover lost germ cells after larval bisection, and as adults, they regenerate gonads during whole-body regeneration (Inoue et al., 1992; Wessel et al., 2014b). Tunicate larvae also recover lost germline upon tail amputation and become fertile as adults (Kawamura et al., 2011; Yoshida et al., 2017). Zebrafish and salamanders can regenerate gonads if a part of the gonad containing the germline stem cells remains after surgery (Erler et al., 2017; White et al., 2011). Finally, plants repeatedly make their germline from somatic cells, as they do not have germline segregation (Schmidt et al., 2015; Walbot & Egger, 2016). Altogether, these organisms defy many of the assumptions about the germline that originate via the studies on established model organisms, mainly that the germline is a set-aside lineage that needs to be segregated and protected, and if lost, it cannot be replaced. However, while we know many organisms can regenerate their germ cells, in many of these examples, we still do not know how germ cell regeneration is accomplished, and whether the source cells are of somatic origin.

The focus of this review is a group of highly regenerative organisms, annelids (or segmented worms), which regenerate their germ cells and gonads, and go through asexual–sexual cycles where the germline appears discontinuous. Therefore, annelids show high plasticity around making and remaking their germline in different contexts and at different developmental stages. Annelid germ cell and gonad regeneration have been reviewed previously focusing on a limited number of annelids (Özpolat & Bely, 2016; Özpolat et al., 2016; Tadokoro, 2018; Weisblat, 2006; Wessel et al., 2020). In this review, I cover extensive literature going back to the late 1800s and published in many languages. To my knowledge, there is no other animal group this topic has been studied in such breadth. Annelids are diverse and highly regenerative, many are easy to access and rear for the full life cycle and easy to observe under a microscope. I believe this rich history of studies reflect the feasibility of annelids as models of germ cell and gonad regeneration. With the contemporary tools and technologies available to developmental biologists, we can now add more depth to these studies. Therefore, I present a framework to investigate the possible cellular sources of the regenerated germline in the context of whether germline continuity across generations (i.e., germ–soma separation) is preserved or not. This approach will allow identifying cellular sources, and eventually the molecular underpinnings of germ cell and gonad regeneration.

2 |. GENERAL ANATOMY AND DIVERSITY OF REPRODUCTIVE SYSTEMS IN ANNELIDS

Annelids are a large group and over 21,000 species have been described (Rouse et al., 2022). Most annelids have vermiform body morphology, with repeating organs and tissues in each segment. Segments are capped with asegmental tissues on either anterior or posterior ends (Figure 1a). In some annelids, segments are physically separated from each other via a membrane called septum, while some annelids do not have this structure, therefore, have a continuous coelomic (fluid-filled) cavity. Recent molecular phylogenetic analyses group annelids into two major clades, Sedentaria and Errantia, while several annelid lineages branch basally (Struck et al., 2011; Weigert & Bleidorn, 2016). Clitellata (“oligochaetes” and leeches) is nested within Sedentaria, which also includes sedentary “polychaetes,” while Errantia is composed mainly of “polychaetes.” Within and across these clades, annelids show a large diversity of reproductive anatomies and strategies (Dorresteijn & Westheide, 1999; Giese & Pearse, 1975; Schroeder & Hermans, 1975), making them a rich group to investigate germline plasticity and regeneration in different contexts and across different life stages.

FIGURE 1.

Phylogeny of annelid species and distribution of processes. (a) Annelid body plan shows generalized regions: Segments are capped by two asegmental regions called prostomium (pro.) and pygidium (pyg.). Anterior to the pygidium resides the posterior growth zone (PGZ) where new cells and segments are generated by two concentric rings of stem cells called teloblasts. Some annelids have physical separation of segments via membranes called septae. Segment numbering shown for a few anterior segments. (b) Phylogenetic relationship of species and processes covered in this review (Struck et al., 2011; Weigert et al., 2014). (b′) Examples of gonad/germ cell distribution and general features are depicted for specific groups (indicated by the black bars). Cartoons are simplified annelids illustrating the general distribution of gonads (not the exact segment location in the species). AC, anterior cluster; MPC, multipotent progenitor cells.

Clitellates are exclusively hermaphroditic, have localized gonads that typically appear in the same segments in a species-specific manner (Brinkhurst & Jamieson, 1971; Christensen, 1994; Lasserre, 1975; Loden, 1981). For example, in the hermaphrodite earthworm Eisenia, two pairs of testes are found in segments 10 and 11, and a pair of ovaries are found in segment 13 (Gates, 1949). Therefore, it is possible to amputate worms into small fragments that are (in theory) completely somatic except for these specific gonad-bearing segments, and ask if somatic fragments can regenerate into sexually mature adults. Errant and Sedentary “polychaetes,” on the other hand, are generally gonochoric (either male or female) but hermaphroditism is also observed (Schroeder & Hermans, 1975). Polychaetes typically have germ cells and gonads present more widely across the body. Some polychaetes have segment-specific locations of gonads similar to clitellates, while some do not have localized gonads and instead use the entire body (the coelomic cavity) as a gamete incubator. Meanwhile, others such as Syllids undergo schizogamy (Franke, 1999), where they produce sexual clones (stolons) that bud off (Ponz-Segrelles et al., 2018).

Annelids have stereotypical embryonic development, similar to C. elegans and Ciona development, where cells in the embryo undergo precise cleavages and give rise to cell lineages with predetermined fates (Conklin, 1897; Nishida & Satoh, 1985; Sulston et al., 1983; Wilson, 1892). As a part of this embryonic program, primordial germ cells are thought to arise via maternally inherited factors and from the 4d cell lineage (Fischer & Arendt, 2013; Kang et al., 2002; Oyama & Shimizu, 2007; Özpolat et al., 2017; Rebscher, 2014; Rebscher et al., 2012; Stéphan-Dubois, 1964). This appears to be the case in all annelids studied to date (Rebscher, 2014). The primordial germ cells are first detected in the gonadal segments in clitellates. In errant and sedentary polychaetes, primordial germ cells form and then migrate to their final location. It is currently not known whether primordial germ cells form both somatic and germ components of the gonads, or whether the somatic contribution is made via other tissues. In the species that form gonads, these organs may be formed on the septae or the coelomic epithelium as pear-shaped structures, and gradually develop into fully mature sexual organs. Secondary sexual characteristics such as seminal vesicles, ovisacs, ducts, clitellum, and so on, may not appear until full sexual maturation (Brinkhurst & Jamieson, 1971; Brusca & Brusca, 2003).

3 |. GERMLINE REGENERATION IN THE CONTEXT OF EMBRYONIC REGULATION

To test if embryos can recover from and grow into fertile adults after the ablation of the germ lineage during early embryogenesis, the embryonic cell that gives to primordial germ cells (PGCs) was ablated (either 4d or 3D) in Capitella, and to my knowledge, this is the only example in annelids in this context to date (Amiel et al., 2013; Dannenberg & Seaver, 2018) (Figure 2e). In this sedentary polychaete, 4d gives rise to the PGCs and the anus only, however, this cell is very small and hard to ablate. Therefore, the majority of manipulation experiments were carried out on its precursor (3D), which gives rise to part of the endoderm, in addition to PGCs and the anus. Germline-ablated (3D-deleted) samples were first raised to stage 7 larvae (about 7 days postfertilization) and these larvae overall had similar morphology to controls. When 3D-deleted larvae were analyzed further for piwi1 expression, there was an overall expansion of piwi1 expression but with loss of the characteristic PGC signal (Amiel et al., 2013). In the controls, piwi1 is expressed in the PGCs and the posterior growth zone only. When these experiments were repeated in a later study, authors found that in stage 9 larvae, about 87% of the larvae lacked the PGCs, but 13% regenerated them even though the blastomere that normally produces them was ablated (Dannenberg & Seaver, 2018). Interestingly, the rate of individuals with regenerated germline significantly increased in juvenile stages (nearly 100%). Authors suggest there are two phases of germline regeneration: one during embryonic/larval stages with limited success, and another phase postmetamorphosis with higher success. When raised to adulthood, the 3D-deleted individuals are fully fertile with all parts of the reproductive system present and can mate to produce viable progeny. To get more insight into which lineage may be contributing to the regenerated germline, authors hypothesized that the source would be of mesodermal origin, and tested this hypothesis by deleting the 2D blastomere. This experiment effectively ablates the left mesoderm and the germ-line (the right mesoderm originates from another blastomere, and therefore still present). When raised to adulthood, while the survival was lower (only 6%), both males and females were fertile and able to mate. Therefore, authors conclude that 2D (precursor of left mesoderm) is not the source of the regenerated germline, and the precursor of the right mesoderm (3c) remains to be tested as the potential source. Capitella has genetic tools established, and, therefore, in future experiments it is feasible to combine long-term lineage tracing and blastomere ablation experiments, dissecting the origins of the regenerated germline in this embryonic germline ablation model.

FIGURE 2.

Examples of gonad and germ cell regeneration in annelids. (a) Gonads regenerate via epimorphosis as a part of body axis regeneration, in this example during anterior regeneration. (b) Gonads regenerate via morphallaxis (tissue reorganization) in original but previously nongonadal segments. (c) Germ cells regenerate via epimorphosis. (d) Gonads regenerate in transplanted segments. (e) The germline regenerates after the ablation of the germline precursor in the embryo. Blue color denotes gonadal tissues or germ cells. Gray color denotes newly regenerated segments. Anterior to the left, dorsal up.

4 |. GERM CELL AND GONAD REGENERATION AS A PART OF WHOLE-BODY AXIS REGENERATION

Most annelids regenerate at least posteriorly after bisection, many annelids regenerate both anteriorly and posteriorly from small body fragments (Figure 2) (Bely, 2006; Bely et al., 2014; Kostyuchenko & Kozin, 2021). During regeneration, annelids may employ blastema-based regeneration (epimorphosis) and tissue remodeling (morphallaxis) (Özpolat & Bely, 2016). Epimorphic regeneration is a process that results in the creation of new tissues through cell proliferation. These cells form a regeneration blastema and eventually differentiate into new tissues. This type of regeneration involves a series of stages, including wound healing, blastema formation, blastema patterning, differentiation, and growth. Morphallactic regeneration restores missing structures by remodeling existing tissues, without blastema formation (Özpolat & Bely, 2016). In the context of germ cell and gonad regeneration, epimorphic regeneration would involve regenerating the germ cells and gonads in the newly regenerated segments, while morphallaxis would be reorganizing previously nongonadal segments into gonadal segments.

It is important to note that, even if gonadal structures regenerate, this may not be sufficient proof for germ cell regeneration (e.g., perhaps only somatic components of gonads regenerate without any germ cells). Ideally, gametes would be observed in regenerated gonads and the ultimate test would be whether the regenerated individuals can sexually reproduce and give rise to fertile progeny. Some studies reviewed below have been able to show fertility after regeneration, while others could not (but always for technical reasons such as worms only asexually reproduce in laboratory conditions) (Figure 1b).

4.1 |. Germ cell and gonad regeneration during epimorphic regeneration

It has been possible to test whether worms have the capacity to regenerate sexual organs from parts of the body that normally do not have the gonads in clitellates, because this group generally displays restricted localization of gonads to specific segments. There are multiple examples of clitellates regenerating germ cells and gonads in previously nongonadal segments, even though studies vary in their approach. Experiments in Criodrilus, Enchytraeus, Lumbriculus, Perionyx, Rhynchelmis all show that gonads regenerate, while in some of these species, the position, number, or the sex of regenerated gonads may vary from the control worms that were not amputated.

Victor Janda extensively studied germ cell and gonad regeneration in Criodrilus lacuum and Rhynchelmis limosella (Janda, 1912a, 1912b, 1924, 1926), and these appear to be the first extensive works on the subject. Janda removed gonad-bearing segments from Criodrilus, and observed successful germ cell and gonad regeneration eventually with female or male gametes forming. In his initial studies (Janda, 1912a, 1912b), Janda used hundreds of mostly medium-sized worms, some showing clear external characteristics of sexual maturation. He amputated 17–30 anterior segments which included the entire sexual region. The anterior pieces were fixed, and the posterior pieces were let to regenerate. Some of these regenerating samples were fixed after 5 months, others were raised further. Janda observed that germ cells and gonads always regenerated, he never found worms that failed to regenerate germ cells and gonads (Figures 2c and 3c). Just as during development, the pear-shaped gonads appeared first, and the secondary sexual characteristics formed later. Some regenerated ovaries were significantly larger than normal ovaries. Interestingly, supernumerary ovaries and testes in the regenerated worms appeared at a high frequency; ovaries were always more numerous than testes, and some regenerated worms had only ovaries but never all testes (Figure 3d). In unamputated worms, ovaries are found only in segment 13 as a single pair, and there are only two pairs of testes in segments 10 and 11. In regenerated worms there were pairs of ovaries and testes found in many other segments, for example, ovaries were observed in multiple consecutive segments (e.g., segments 8–13 or 4–18). Also, supernumerary regenerated gonads appeared in segments they normally do not occupy in normal worms. These tended to be shifted anteriorly, sometimes ovaries replacing testes, which were shifted to further anterior segments. Janda’s R. limosella findings were similar to Criodrilus: worms could always regenerate germ cells and gonads during anterior regeneration, usually with a positional shift and supernumerary, but not as extensive as in Criodrilus (Janda, 1924). Positional shifts or supernumerary gonads after regeneration have also been observed in several other clitellates. In Lumbriculus gonad number and positional variability are observed in regenerated worms (Mrázek, 1906). In Enchytraeus regenerated gonads along with germ cells are always observed in the seventh and eighth segments (Myohara et al., 1999; Tadokoro et al., 2006), even though gonads develop normally in 10th and 11th segments (Christensen, 1994; Sugio et al., 2008). In Perionyx excavatus during anterior regeneration, gonads and sex apparatus regenerate but are found at variable locations compared to controls (Gates, 1927).

FIGURE 3.

Examples of reproductive anatomies in different annelids. (a) Pristina leidyi (a clitellate naidid), is hermaphroditic, and gonads always appear in segments 6 and 7. Gonads, fission zones, and the posterior growth zone all express germline/multipotency markers. arrowheads indicate some of the piwi+ cells on the ventral nerve cord. Inset (a′) shows the piwi+ cells from the ventral view (adapted from Özpolat & Bely, 2015). (b) Platynereis dumerilii (an errant polychaete) is gonochoric, and does not have localized gonads. Instead gonial clusters are distributed along the body (shown by vasa expression). Adapted from Kuehn et al. (2021). (c) Cartoon of a frontal section of Criodrilus. Ovaries (yellow) and testes (blue) as well as secondary sexual structures such as seminal vesicles and ducts are shown (Janda, 1912a). This species has septae in between segments. (d) Schematics showing the normal distribution of gonads (bottom) and the supernumerary gonads that extend beyond gonadal segments in Criodrilus (top) (Janda, 1912a).

Only a handful of studies show successful sexual maturation after regeneration in clitellates. In Criodrilus, Janda reared the regenerated worms for 2–4 years and observed that supernumerary gonads persisted long after amputation (Janda, 1912a). These worms sometimes had abnormalities in the morphology or placement of their sexual apparatus, but they were nevertheless fertile and gave rise to viable progeny even 3 years after the operation. He then carried out extended analyses in a later study where he tested the effect of different conditions on gonad regeneration in Criodrilus, including repeated amputation–regeneration cycles (Janda, 1926). In the repeated amputation–regeneration experiment, 15–17 anterior segments were removed at first (samples used were either adults or very young worms). Gonads were reformed as many times as the operation was carried out (up to seven times of amputations) in both age groups. Even when regeneration was abnormal, these animals still mated and produced viable progeny. This work in Criodrilus and Janda’s other works in Rhynchelmis are the first examples in annelids where regenerated germ cells and gonads are shown to give rise to viable progeny. Similarly, more recent studies in Enchytraeus showed that this clitellate can also regenerate germ cells and gonads and become sexually mature from fragments that do not contain previously gonadal segments (Myohara et al., 1999; Tadokoro et al., 2006). When worms were cut into three fragments (only the anterior piece containing the gonadal segments), the majority of the middle and posterior pieces and all the anterior pieces regenerated and became sexually mature (Tadokoro et al., 2006).

In “polychaetes,” studies that obtain small nongonadal fragments and grow into full worms do not exist. This is in part due to the anatomical differences between polychaetes and clitellates: polychaetes typically have gonads or germ cells spread more widely across the body (e.g., more pairs of gonads in numerous segments, or coelomic gonial clusters which are not confined to gonadal structures but instead are freely floating across the body). Another reason is the scarcity of polychaete regeneration laboratory models which can regenerate both anteriorly and posteriorly, thereby eliminating the possibility to obtain a small worm fragment that can regenerate into a complete worm. Syllids are one group of annelids that regenerate both anteriorly and posteriorly and can form complete individuals from small fragments that are only a few segments long (Ribeiro et al., 2018), however, studies in Syllids have not focused on gonad regeneration as of date. Nevertheless, experiments in Capitella, Nephtys, Platynereis, and Spirorbis, show extensive removal of gonadal/germline tissues followed by successful regeneration of germ cells and gonads.

Capitella and Platynereis juvenile worms can only regenerate posteriorly. In both groups, there is a cluster of cells anteriorly located and expressing germline multipotency markers (Figure 3b) (Juliano et al., 2010). Posterior regeneration only takes place if amputation is carried out posterior to this “anterior cluster.” Nevertheless, this amputation removes the majority of germ cells and gonads both in Capitella and Platynereis, worms regenerate the entire body axis, including the male and female gonads (in Capitella) or male or female gonial clusters (in Platynereis) (Giani et al., 2011; Hauenschild & Fischer, 1969; Hofmann, 1966; Metzger et al. in prep; Rebscher, 2014). Therefore, in either case, while complete removal of the germ cells is not possible with simple bisections, the majority of germ cell and gonadal tissues can be removed and will regenerate during body axis regeneration. Future studies that employ genetic ablation of germ cells will be able to test whether these species still regenerate germ cells without the anterior cell clusters that express germline markers.

The timeline of gonad and germ cell regeneration appears to be uncoupled from body axis regeneration in many polychaetes. In Platynereis, posterior regeneration is completed in about 5–7 days (Planques et al., 2019), however, regeneration of gonial clusters takes a lot longer, often several weeks (Metzger et al., in prep). Similar observations were made in Nephtys, where after 3 weeks of regeneration, Clark observed rudiments of nephridia and a complex of gonadal blood vessels appear but no gonads and no signs of gametogenesis until later (Clark, 1968). Contradicting with these observations, Stagni reported oocytes regenerated only a few days into anterior body axis regeneration in the sedentary polychaete Spirorbis (Stagni, 1959). However, it is unclear if these worms had some remaining germ cells in the stump that went under an accelerated maturation; therefore, this example may not represent true gamete regeneration (Hauenschild, 1966; Schenk et al., 2016, 2019). Successful sexual maturation after germ cell regeneration has not been tested in these studies except for Platynereis. Platynereis reaches sexual maturation after regeneration, there is no significant change in oocyte number and success of fertilization in females, and F1 progeny of the regenerated males and females are also fertile (Metzger et al., in prep).

4.2 |. Germ cell and gonad regeneration as a part of morphallaxis and dynamic tissue maintenance

Annelids dynamically maintain body proportions as they grow posteriorly and when they regenerate. In the context of germ cell and gonad regeneration, dynamic maintenance could mean making new gonads in previously nongonadal segments, or regressing and regrowing gonads based on nutrient availability. A striking example of this is morphallaxis. Morphallaxis in annelids is a regenerative process where pre-existing tissues and organs are reorganized following injury or damage to form new structures. Unlike epimorphosis, which generates new tissues mainly via the formation of a blastema and extensive cell proliferation, morphallaxis is thought to involve reorganizing the existing cells, tissues, and organs without blastema formation (Özpolat & Bely, 2016). For example, in Pristina, worms only regenerate four segments via epimorphosis during anterior regeneration, even when more than four anterior segments are removed. This results in a worm that has abnormal body proportions: depending on where the amputation is made, the worm may have posterior segments starting immediately after the head segments, while missing the mid-body segments along with the gonads (Figure 2b). The correct body organization and proportions are re-established via morphallaxis following epimorphic regeneration (Özpolat & Bely, 2016; Özpolat et al., 2016). This has also been shown in Enchytraeus using regional gut markers (Myohara et al., 1999; Takeo et al., 2008) and in Eisenia which only regenerates five to six anterior segments irrespective of the number of anterior segments removed (Gates, 1949). Therefore, in the “oligochaetes” Enchytraeus, Pristina, and Eisenia, if anterior amputation removes a large enough region that includes the gonadal segments, the number of segments regenerated anteriorly will not include the gonadal segments during the epimorphic phase. Instead, the worms will reorganize the more posterior and previously nongonadal segments into gonadal segments via morphallaxis (Figure 2b).

Similar plasticity and dynamic maintenance of gonads are also observed in experiments where nutrient availability was altered (Figure 4c). In Eisenia, starving and refeeding the worms induces supernumerary gonads in previously nongonadal segments, sometimes along with sex inversion where testes turned into ovotestes (Herlant-Meewis, 1962; Relexans, 1970). In Enchytraeus and Pristina gonads regress upon starvation and reform upon refeeding in lab conditions (Myohara et al., 1999; Özpolat & Bely, 2015). While these are not experiments involving injury and regeneration, they demonstrate dynamic maintenance and the ability to grow gonads that are naturally occurring in annelids.

FIGURE 4.

Examples of germ cell plasticity and formation without embryogenesis in annelids. (a) Asexual–sexual cycles, and the formation of new gonads without embryogenesis are shown for paratomic fission. In some species, the asexual cycles and new gonad formation during asexual reproduction may happen for 100 s of cycles before the organism undergoes sexual reproduction. (b) Stolonization in Syllids depicted. The stock worm (most anteriorly located) has gonads, and the clones (called stolons) forming at the tail end contain mature gametes. These stolons detach from the stock, reproduce sexually, and die. (c) Dynamic regression and reformation of gonads during starvation–refeeding cycles. In many annelids, gonads regress upon starvation, and reform once nutrients are available again. Blue color denotes gonadal tissues or germ cells.

5 |. GERM CELL AND GONAD REGENERATION UPON SURGICAL REMOVAL AND TRANSPLANTATION

Another way to approach gonad and germ cell regeneration is to surgically remove these organs and cells and test whether the operated worms regenerate them. However, in practice, these are difficult experiments in many annelid models due to small organismal size, or the presence of numerous pairs of gonads or germ cells across the body that would have to be removed as is the case in many polychaetes. Therefore, experiments involving surgical removal or transplantation have been carried out mainly in the earthworm Eisenia, because this species is large and easy to surgically manipulate (André & Davant, 1977; Avel, 1929).

A neat set of grafting experiments in Eisenia tested whether gonads would regenerate along with the germ cells after surgical removal (André & Davant, 1977). A several-segment-long piece containing the gonads was obtained. This piece was then flipped inside out (like a sock), and all organs including gonads were removed. Then the piece was reversed back and grafted onto a control worm (Figure 2d). These grafts were able to regenerate gonads, in some cases the regenerated gonads in the grafts reached full sexual maturity including gametes. In other cases, different stages of oogenesis and spermatogenesis were observed, or there were small buds of gonads that failed to fully differentiate where it was not possible to determine the sex. Regeneration of sexual organs was always partial, and different parts of the reproductive system (including the secondary sexual characteristics, such as seminal vesicles) regenerated independently of each other. In rare cases, all the sexual organs and apparatus were present at the same time, but in reduced numbers. For example, there was one testis instead of four. Interestingly, the digestive tract never regenerated, and the nervous system only regenerated in some grafts. However, the absence of these tissues did not affect reproductive organ regeneration. Nephridia usually regenerated along with the gonads. But when nephridia did not regenerate, gonads nevertheless regenerated and, in some cases, reached full sexual maturity including gametes. The authors also tested different grafting locations (gonadal region vs. nongonadal) in the host, which did not influence the success of gonad regeneration. Overall, gonads and germ cells could robustly regenerate in these experiments, especially compared to all the other organs and tissues.

6 |. GERMLINE RE-ESTABLISHMENT WITHOUT EMBRYOGENESIS

An organism may undergo various morphological, physiological, and behavioral changes that are essential for its growth, survival, and reproduction as a part of postembryonic development. Examples include metamorphosis in insects, growth, and differentiation of organs and tissues in animals, and asexual reproductive processes such as fission and budding in many organisms. I will primarily focus on asexual reproduction (paratomic and architomic fission), stolonization, and posterior growth in this section (Figure 4a,b). These are processes during which new germ cells and gonads are produced postembryonically.

6.1 |. Paratomic and architomic fission

Fission is an asexual reproduction mode where an individual divides into two or more new individuals, and this type of reproduction is observed across annelid taxa. In paratomic fission, worms first start developing new tail and new head portions, and then split into two complete individuals, or sometimes form chains of clonal worms that eventually separate. In architomic fission, worms fragment themselves into small pieces (sometimes as small as a single segment) via constriction of circular muscles, then each of these pieces develop (“regenerate”) into new individuals. There are many fissioning genera in clitellates (Berrill, 1952; Christensen, 1994; Dehorne, 1915, 1916; Harper, 1904; Lasserre, 1975; Loden, 1981; Mrázek, 1906; Myohara et al., 1999; Özpolat & Bely, 2015; Schroeder & Hermans, 1975; Stephenson, 1930; Zattara & Bely, 2011, 2016), but the ability is not restricted to clitellates and examples in polychaetes are also found (Faulkner, 1930; Gibson & Harvey, 2000; Malaquin, 1905, 1925; Marescalchi et al., 2019).

In natural and lab populations of several clitellate genera such as Enchytraeus, Lumbriculus, Pristina, and Stylaria (Christensen, 1973, 1994; Dehorne, 1916; Harper, 1904; Mrázek, 1906) and in some polychaetes such as Pygospio (Gibson & Harvey, 2000) asexual and sexual reproduction alternate (Figure 4a). In some species, population density (or culture density in the lab) can trigger the shift: in low-density conditions (a few worms) worms may sexually mature, while in high-density they will primarily asexually reproduce (Christensen, 1973). Sometimes there may be many generations of asexual reproduction before some worms in a population become sexual again. In Pristina lab cultures, it was estimated that even after thousands of asexual generations, worms still make gonads during each asexual cycle (Figure 3a). Sexual individuals are observed although very rare (Özpolat & Bely, 2015). In many Naididae, including Pristina, gonads are made on the septae of newly produced segments during paratomic fission (Christensen, 1973; Özpolat & Bely, 2015). How gonads and the germline are maintained and/or re-established through numerous asexual cycles is unknown.

In Enchytraeus, sexually mature worms undergo architomic fission (fragment into small pieces), and then the fragments that develop into whole worms can eventually become sexually mature again (Myohara et al., 1999). In species that fission via paratomy, new gonads are gradually produced as new head segments develop, and in some of these species, cells that express germline markers are observed to migrate toward segments where new gonads are made (Özpolat & Bely, 2015). It is unclear if architomic species “expect” an upcoming fission event, and distribute germline progenitors across all segments, so that fission results in pieces that can re-establish gonads. Alternatively, worms can remake gonads from cells present in the segments, without relying on the germline specifically (see below for more discussion on source cells.)

Genital segment regeneration is observed after autotomy as a part of the natural life cycle in the errant polychaete Eunice (Hofmann, 1972, 1974). In this species, there are about 70–150 genital segments containing gonads and gametes. These genital segments are located posteriorly and they separate from the rest of the body via autotomy, leaving worms with only nongonadal segments. Upon regeneration of missing segments, gonads regenerate along with gametes, and worms can reproduce again (Hofmann, 1972, 1974).

6.2 |. Stolonization

A fascinating group of errant polychaetes called Syllids make clones (stolons), sometimes forming chains of clones, in a process called stolonization (Franke, 1999) (Figure 4b). Some Syllids also alternate between asexual fragmentation and sexual stolonization (Berrill, 1952; Okada, 1929; Ribeiro et al., 2020). While superficially stolonization appears similar to what happens in asexual fission, there is a key twist: unlike asexual fission, the clones in stolonization are produced with the sole purpose of sexual reproduction in Syllids (Berrill, 1952; Franke, 1999; Malaquin, 1893; Schroeder & Hermans, 1975). In some Syllids, it has been suggested that oocytes or sperm are produced by and passed to the stolons from the gonads of the anterior “stock” worm (Franke, 1999; Gidholm, 1963; Wissocq, 1966). In some of these older studies, histological analyses show that ovaries are present only in the stock worm and not the stolons, and that gametes from the stock are transferred to the stolons via a special canal (Gidholm, 1963). Therefore, there is strong histological evidence of gamete transfer into stolons in Syllids. However, there is a more recent study using molecular markers that suggest de novo formation of gametes and gonads in the stolons simultaneously with the stock worm (Ponz-Segrelles et al., 2018). Further studies are needed to show the exact origins of gametes in the stolons and investigate whether some germ cells are established de novo in the stolons via cell lineage analyses.

6.3 |. Posterior growth

One of the annelid hallmarks is posterior growth from a localized ring of cells at the tail end (segment addition zone) (Balavoine, 2014). Most annelids retain their growth zone throughout their life and continuously add new segments. Cells in the growth zone express germline/multipotency markers (Gazave et al., 2013; Özpolat & Bely, 2016). Because of the molecular signature similarity of the growth zone to germ cells, and the pluripotent nature of these stem cells as a population (Álvarez-Campos et al., 2023a), they cannot be ruled out as a potential secondary source of germ cells in addition to the primordial germ cells established during embryogenesis (Özpolat et al., 2017; Rebscher, 2014; Rebscher et al., 2012). As the new segments are being established, new gonads or germ cells could be made as a part of the posterior growth process. While there is no direct evidence supporting this possibility yet, it has been suggested in the polychaete Perinereis (Maceren-Pates et al., 2015). Overall, annelids can continuously produce germ cells and gonads via asexual fission, stolonization, and posterior growth. Several examples presented here highlight that the clones and segments produced this way contain germ cells, but it is still unclear whether the germline is transferred to the clones and new segments, or whether it is made de novo.

7 |. CELLULAR MECHANISMS OF REGENERATING THE GERM CELLS

Cells that participate in regeneration have been of great interest in this field of research because determining the source cells (e.g., whether regeneration involves pluripotent stem cells such as neoblasts or cell dedifferentiation and/or transdifferentiation) is critical for understanding the mechanisms of regeneration (Ereskovsky et al., 2019; Jopling et al., 2011; Pesaresi et al., 2019; Poss, 2010; Tanaka & Reddien, 2011; Vierbuchen et al., 2010; Vogt, 2012). Source cells can tell us how cells respond to injury and are reprogrammed, or how organisms fail or succeed in regenerating based on the availability of cell types and their potencies. While the source cell question is one of many approaches to the mechanistic understanding of regeneration in general, I argue that it becomes a central question for germ cell regeneration: because the germline has been viewed as a set-aside and continuous cell lineage in metazoans, depending on the cell type(s) that replace the germ cells during regeneration, continuity of this lineage may be preserved or broken (MacCord & Ozpolat, 2019) (Figure 5). It is important to know whether this is the case, because, if somatic cells are the source cells (i.e., discontinuity in germ lineage), mutations they carry may be passed on to the regenerated germ cells, and consequently they may have poor genome integrity. On the other hand, organisms that regenerate their germ cells primarily from somatic tissues may have evolved mechanisms to retain genomic integrity, and understanding how they manage high levels of genome integrity in their somatic cells will be greatly informative. Finally, if organisms evolved ways to keep the germline intact and continuous across injury–regeneration cycles, or asexual–sexual cycles, how they achieve maintaining a dormant germline across the body through these cycles will also be greatly informative.

FIGURE 5.

Framework for investigating cellular sources of germ cell regeneration. (a–c) Models of cellular sources in germ cell regeneration. Some organisms may have adult pluripotent cells that can regenerate both somatic and germ cells (a). In others, there may be lineage-restricted cell populations that can only regenerate somatic or germ cells (b). Finally, some organisms may be able to regenerate germ cells using their somatic cells via dedifferention or transdifferentiation. Blue: germline; Yellow: soma; Gray: pluri/multipotent cell. (d) Continuity of germline, and break in the continuity, if soma is the source of regeneration (adapted from MacCord & Ozpolat (2019).

In this section, I aim to first establish my definitions for possible cell sources in germ cell regeneration, and then discuss the annelid literature in this context. Overall, my definitions are strictly based on historical definitions of germ and soma as cell lineages established during embryogenesis. I regard the germline as a cell lineage that gets segregated during embryonic development with the sole purpose of eventually giving rise to the gametes in adults. Any cell type that comes from this initial set of cells and will give rise to gametes is considered the germline or germ cells (e.g., primordial germ cells, germline stem cells, mature gametes). Soma is any other cell type in the body such as muscle, neurons, gut, and so on, that do not contribute to the germline under normal circumstances (i.e., normal growth and maturation), once the germline is established (the Weismann barrier) (Extavour & Akam, 2003; Nieuwkoop & Sutasurya, 1979, 1981; Weismann, 1893).

What complicates this dichotomy is animals that have adult pluripotent stem cell types (such as i-cells in hydrozoans, or neoblasts in planarians and acoels) (see Solana, 2013 for an extensive discussion on this topic) (Varley et al., 2023). Data are scarce on the exact embryonic origins of these pluripotent cells, and, therefore, it is hard to categorize them as germline or soma in the traditional sense (Davies et al., 2017; Kimura et al., 2022; Varley et al., 2023). However, they can give rise to both soma and germ cells throughout life in these organisms (Bounoure, 1940; Buss, 1988; Gold & Jacobs, 2013; Heys, 1931; Newmark et al., 2008; Solana, 2013). Therefore, I consider these as a third cell type (pluripotent) which cannot be categorized as soma or germline until we have more information about their embryonic origins and their postembryonic potential (i.e., whether they have subpopulations that are strictly dedicated to giving rise to the germline) (Figure 5a vs. 5b).

7.1 |. Models for cellular sources of germline regeneration

The cellular basis of new tissues generated in annelids during regeneration, posterior growth, and asexual reproduction is still an open question and an active area of investigation. While the exact source cells and their potencies are not entirely known, there are several lines of evidence that suggest that annelids do not have adult pluripotent stem cells such as planarian neoblasts. In regeneration, most new cells and tissues are thought to arise from cells nearby the wound site (Bely, 2014; Bely et al., 2014; Bideau et al., 2021, 2023; Boilly, 1967, 1969; Kostyuchenko & Kozin, 2021; Planques et al., 2019; Ribeiro et al., 2021). New cells in the zones of growth (especially posterior growth) are thought to arise from a population of cells that altogether have pluripotency but subpopulations have more limited potencies, and the individual potencies of these cells are yet to be determined (Álvarez-Campos et al., 2023a; Balavoine, 2014; Gazave et al., 2013).

The term “neoblast” was originally coined in annelids upon observations of cells that appeared to have migratory characteristics and a large nucleus-to-cytoplasmic ratio, accumulating around the wound site during regeneration (Randolph, 1891, 1892). However, these cells are only found in clitellates, and generally not found in sedentary or errant polychaetes (Bely, 2014; Boilly, 1969; Potswald, 1969). Even in clitellates, the nature of neoblasts whether they are pluripotent or restricted in their potential, and whether they are even required for regeneration is still debated. Indeed, several lines of evidence suggest that annelid neoblasts are not necessary for regeneration. For example, in the clitellate Enchytraeus, two sister species were compared, one with neoblasts and another without (Myohara, 2012), but both species are able to regenerate and therefore neoblasts do not appear necessary for regeneration in this genus (in contrast to planarian neoblasts, which are necessary for regeneration). In another study, cells that fit the historical description of annelid neoblasts (i.e., cells with large nucleus/cytoplasm ratio and migratory) were found to be strongly related to fission and formation of new gonads during fission but not anterior or posterior regeneration in Pristina (Özpolat & Bely, 2015). Overall, while neoblasts have been widely studied in annelids for over a century, so far there is no strong evidence for individual pluripotent neoblasts that can give rise to all tissue types and germ layers including the germ cells in annelids. Therefore, all three models (Figure 5) still need to be tested via contemporary methods such as genetic lineage tracing or single-cell sequencing across the regeneration, asexual reproduction via fission, or posterior growth timelines. Below, I will review the existing literature and list evidence supporting each mechanism.

7.1.1 |. Mechanism 1: Pluripotent lineages (e.g., neoblasts) (Figure 5a)

Clitellate neoblasts are described as cells that reside on septa (the membrane that separates segments from each other), seen across many segments, and become migratory when there is an injury (Bely, 2014; Hill, 1970; Randolph, 1891, 1892). In their migratory form (sometimes referred to as “wandering cells”), they are typically found ventrally, especially on the ventral nerve cord. However, these cells are not universal in annelids, and even within the same species they may be observed migrating towards the posterior wound but not the anterior wound, though both axes can regenerate (Berrill, 1952; Herlant-Meewis, 1946; Zattara et al., 2016). In Enchytraeus, neoblasts have been studied extensively, and while there is evidence for these cells to migrate to the wound site and participate in blastema formation, these studies did not include lineage tracing that could resolve whether Enchytraeus neoblasts are pluripotent or not (Tadokoro et al., 2006). As mentioned above, some Enchytraeus species do not have neoblasts, but can regenerate nevertheless (Myohara, 2012).

In “polychaetes” fewer studies from the molecular era investigated neoblasts and their role in regeneration. Generally, works that studied regeneration from a neoblast perspective are relatively older, and had the limitations of using static images of histological preps or not having access to molecular tools for labeling. One recent study in Capitella suggests cell migration to the blastema, but acknowledges that these may be lineage-restricted stem cells instead of neoblasts (de Jong & Seaver, 2017). While there were several older works claiming the existence of neoblasts in polychaetes, many also disagreed (Boilly, 1969; Potswald, 1969). Specifically, several works focusing on germ cell regeneration suggested that germ cells regenerated from neoblasts (Faulkner, 1930; Malaquin, 1925; Stagni, 1959; Vannini, 1965). In Spirorbis, oocytes were suggested to regenerate in a matter of days from a conspicuous reserve of neoblasts, distributed along animal’s body, but were particularly evident at the achaetic region (a few specific segments without chaetae in Spirorbis) (Stagni, 1959). However, later studies disagreed with the neoblast-driven germ cell regeneration observations, arguing that these studies confuse blood vessel structural cells with neoblasts (Potswald, 1969, 1972). It is evident in Potswald’s language in his titles (e.g., “so-called neoblasts”) that he was strongly against the presence of neoblasts as source cells in regeneration. He quotes Hill (1970): “Perhaps it is time to stop evoking ‘reserve’ cells to explain all of the mysteries of regeneration.”

Interestingly there is one publication that argued that germ cells give rise to all tissue types in Filograna and Salmacina, claiming they act like neoblasts (Malaquin, 1905). According to Malaquin, germ cells multiply via mitosis inside each segment, then these cells infiltrate the septa or between the epidermis and muscle, and later participate in forming the fission zone. He likens this to germline cancers, where germ cells can give rise to other types of tissues. This is the only work that I have encountered in annelids where germ cells are claimed to give rise to somatic cells during regeneration. But Faulkner disagrees with Malaquin and says that these are not gonocytes, but instead, they are neoblasts that are functionally identical with germ cells and are the only cells that retain the potential to proliferate to make new tissues (Faulkner, 1930). Similarly in another study in Filograna, Faulkner suggested that the production of the fission zone entails tissue disintegration, followed by new tissue formation (all three germ layers) via neoblasts located in the ventral body wall. Overall, there is no strong and direct evidence for pluripotent neoblasts that migrate across several segments to participate in regeneration in annelids where they can give rise to both somatic and germ cells. It is more likely that migratory cells sometimes mistakenly are referred to as “neoblasts,” and these cells may be associated with building new tissues including gonads during asexual reproduction than regeneration (discussed in the next section), and, therefore, represent a lineage-restricted cell type.

7.1.2 |. Mechanism 2: Lineage restriction (Figure 5b)

Lineage restriction as a mechanism of regeneration or producing new tissues during posterior growth and asexual reproduction has been suggested in organisms such as axolotl, zebrafish, and mouse (Cao et al., 2019, 2021; Flowers et al., 2017; Johnson et al., 2020; Jopling et al., 2010; Kikuchi et al., 2010; Kragl et al., 2009; Lehoczky et al., 2011; Sanor et al., 2020). Two levels of lineage restriction can be considered: (1) restriction at the embryonic germ layer or (2) restriction at the cell type level. If cells that originated from one germ layer give rise to cell types of only their original germ layer (e.g., mesoderm derivatives regenerate mesoderm cell types only) they can be considered lineage restricted at the germ layer level. If cells can give rise to cell types outside of their original germ layer, then they are likely to be considered pluripotent. A stricter approach to lineage restriction would focus on the individual cell types, and whether a cell type can regenerate only the same differentiated cell type (e.g., muscle cells regenerate new muscle). In the germ cell regeneration context, considering the cell-level lineage restriction is more relevant, because we are concerned with germline–soma separation, and whether the germline continuity is preserved or not. The embryonic germ layers are too broad and therefore not helpful in this case, because germ cells in annelids have mesodermal origin during development, where they share embryonic lineage with other cell types such as muscle and connective tissues (Özpolat et al., 2017; Pilon & Weisblat, 1997; Rebscher, 2014; Wilson, 1892). This makes it still possible to have a somatic to germ transdifferentiation even if there is germ layer level lineage restriction during regeneration, which has been suggested in annelids (Bely, 2014; Cornec et al., 1987). Consequently, lineage restriction at the cell type level is more relevant in the context of germ cell regeneration.

Lineage restriction in terms of germ and soma is established early in the annelid life cycle, where the germline is established and set aside early during embryogenesis (Rebscher, 2014). The mechanism of germline specification is thought to be via maternal determinants, and once the germ cells are set aside, somatic cells are believed not to contribute to this lineage anymore. It is then possible that annelids may have a dedicated germ cell lineage that specifically participates in regenerating the germ cells or re-establishing them during asexual reproduction, therefore, keeping the lineage continuous. In this section, I will review examples that support this scenario.

In Enchytraeus, piwi+ cells were suggested to migrate into regeneration blastema and possibly take part in regenerating the gonads specifically (Tadokoro et al., 2006). These piwi+ cells are thought to be different from the cells described as neoblasts in Enchytraeus, both in terms of location and behavior during regeneration (Sugio et al., 2008, 2012; Tadokoro et al., 2006; Yoshida-Noro & Tochinai, 2010). piwi+ cells proliferate in original segments after amputation, and then are thought to migrate to regenerate gonads. This migration occurs at a later stage than neoblast migration to blastema. Neoblasts participate in the formation of blastema first, while piwi+ cells arrive after the blastema has been formed, therefore, authors of the work interpret this as these two cell populations having distinct roles in regenerating separate tissues and organs, which would support the lineage restriction scenario. However, the authors also discuss, it is still possible for piwi+ cells to give rise to more than the gonads (therefore, they could be multipotent). It is also possible that the neoblasts that migrate earlier to the blastema eventually also participate in gonad regeneration. Lineage tracing with contemporary techniques is needed to resolve these open questions.

In Criodrilus, Janda alluded to the presence of “latent sex cells” that help regenerate the gonads. These cells are located on the ventral part of the body, are dormant until needed, and migrate both anteriorly and posteriorly during regeneration (Janda, 1926). He comments on some ongoing experiments to test whether he can exhaust this “stem” population that allows regenerating gonads by repeatedly amputating and obtaining small fragments from regenerates for many (re)generations. However, I was unable to find this publication and it is unclear whether he published his findings from this experiment. He notes that “it remains to be seen whether the latent sex cells result from the mere transformation of originally ‘somatic’ and differentiated cells, or whether they present themselves from the beginning as independent, even if hidden and undifferentiated sex elements, which are only activated under suitable conditions.”

In Capitella, gonads regenerate during posterior regeneration. Authors suggest anteriorly located vasa+ cluster of cells migrate posteriorly to regenerate gonads (de Jong & Seaver, 2017; Seaver & de Jong, 2021), though in an earlier study (Giani et al., 2011) it was suggested there is not any migration from this cell cluster (called multipotent progenitor cells or MPCs). Similarly, in Platynereis, if amputation removes all the germline including the cluster of cells always located anteriorly in segments 6–8 (Figure 3b), worms fail to regenerate the posterior body axis (Álvarez-Campos et al., 2023b; Hofmann, 1966; Kuehn et al., 2021; Zelada González, 2005). If the amputation is carried out leaving the anterior cluster intact, while removing the majority of the germline, worms regenerate the entire body along with the germ cells (Metzger et al., in prep). However, there is currently no direct evidence for the anterior cluster to be the source of the regenerated germline. Overall, there is some evidence that lineage restriction is a mechanism in these polychaetes to regenerate germ cells, but further lineage tracing studies are needed.

During asexual reproduction in clitellates, including in Criodrilus, Enchytraeus, Nais, Lumbriculus, Pristina, several workers observed “wandering cells” which appeared migratory especially along the ventral nerve cord (Berrill, 1952; Herlant-Meewis, 1946; Iwanow, 1903; Janda, 1926; Kostyuchenko & Smirnova, 2023; Lasserre, 1975; Özpolat & Bely, 2015; Sugio et al., 2008; Yoshida-Noro & Tochinai, 2010). Some of these works suggested these to be neoblasts that would participate in regeneration. I hypothesize that these cells possibly have a separate role, which is to transmit germline across asexual cycles of reproduction. However, most evidence is indirect and once again cell lineage tracing is required to resolve this open question. In Nais these cells are congregated along the nerve cord and accumulate in the segments where the fission zone appears (Herlant-Meewis, 1946). In two Pristina species, piwi+ or vasa+ cells that fit the histological definitions of previous studies in Nais and other Naids are also found on the ventral nerve cord (Kostyuchenko & Smirnova, 2023; Özpolat & Bely, 2015). Some of these piwi+ cells come into close contact with the gonads, and live-imaging suggests they primarily migrate toward the fission zone where new gonads are being established (Özpolat & Bely, 2015). In the polychaetes, Filograna and Salmacina, migratory cells during asexual reproduction have been observed and argued to be germ cells (Malaquin, 1905). Overall, it is possible that at least Clitellate annelids have a set-aside cell population that may be primarily involved in actively building new gonads during asexual fission, which is reminiscent of colonial organisms such as tunicates (Rodriguez et al., 2017).

Finally, the posterior growth zone has been suggested as a potential postembryonic source of germ cells in Perinereis. A model is proposed where primordial germ cells stay in the posterior growth zone instead of migrating and forming an anterior cluster of germ cells, in contrast to Platynereis (Maceren-Pates et al., 2015). In this model, PGCs that remain in the growth zone give rise to the gonial clusters each time a new segment is produced.

7.1.3 |. Mechanism 3: Somatic lineages (Figure 5c)

Several lines of evidence suggest somatic cell dedifferentiation during regeneration and asexual reproduction is widespread in annelids. In the context of germ cell regeneration, this widespread dedifferentiation becomes a potential source of regenerated germ cells from formerly somatic cells. One work that directly supports soma-to-germ transition is the embryonic ablations in Capitella as discussed above. However, all works carried out in juveniles and adults are indirect lines of evidence.

In the clitellate Eisenia, transplantation of segments containing only somatic tissues results in gonad regeneration (André & Davant, 1977). Authors suggest nongonadal mesodermal structures (somatopleure and septae) within the graft as the source of regenerating gonads, and that the host tissues do not contribute to regeneration, therefore, germ cells are not migrating from the host into the grafted segments to regenerate the gonads. In Lumbricillus, there are no reserves of unspecialized cells in the form of neoblasts; however, ordinary mesoderm cells were suggested to regenerate the genital organs (Herlant-Meewis, 1946). In Enchytraeus, Myohara either artificially amputated worms to make small fragments from all along the body, or induced the worms to naturally fragment (Myohara et al., 1999). Depending on which region needed to be regenerated, gonads regenerated either via epimorphosis or morphallaxis. Even though this work only systematically looked at the gut as a read out of normal regeneration from the small fragments, some of these fragments were previously nongonadal, these fragments regenerated and became sexually mature. While this can be seen as a line of evidence for soma-to-germ conversion, it is also possible these worms have a dormant lineage of germ cells dedicated to regenerating/re-establishing the germline (as discussed above) (Yoshida-Noro & Tochinai, 2010). Similarly, in older studies mentioned above, at the histological level, Janda and others did not find evidence of gonads or related structures outside of the segments that have been described as gonadal and concluded that nongonadal segments can produce gonads from somatic tissues. But these workers did not have access to molecular tools, therefore, the possibility of a germ cell reserve across the body cannot be ruled out. In the polychaete Nephtys, during the late stages of regeneration, gonads were observed to regenerate via histological analyses of a series of frontal sections (Clark, 1968). Dividing fibroblasts with epithelial origin are suggested as the source cells for regenerating different tissues, including gonads. Overall, while there is compelling evidence for somatic origin for regenerated germ cells, future studies using genetic lineage tracing are required to show direct lineage relationships of somatic cells with the regenerated germ cells.

8 |. WHAT LIES AHEAD

Discussions of whether somatic cells contribute to germ cell and gonad regeneration go back to the late 19th century when Weismann’s and Nussbaum’s theories of germplasm or germline continuity were first postulated (Klein, 1917; Nussbaum, 1880; Stockberger, 1913; Weismann, 1893). These theories became widely accepted, but there were many researchers who also challenged them (Berrill & Liu, 1948; Bounoure, 1940; Herlant-Meewis, 1946; Heys, 1931; Janda, 1926). These people made observations that did not fit into the simpler model of germ–soma separation. Investigations (including some of Weismann’s himself!) revealed gonad regeneration or postembryonic germ cell re-establishment where germline continuity appeared broken (Berrill & Liu, 1948). As the modern synthesis of evolution gained more traction, the idea of a continuous germline (protected from somatic contribution) helped geneticists explain inheritance. Leo Buss notes that “the modern synthesis was virtually an event unattended by embryologists,” plant biologists, mycologists, and people who studied colonial invertebrates, which resulted in the exclusion of these “oddities” from the modern synthesis of evolution (Buss, 1988). Therefore, we left an interesting line of inquiry—a more complex model of animal biology with germ cell regeneration and discontinuity—for another one—a simpler dichotomous germ–soma separation that conveniently explained inheritance and genetics. I argue it is now the right time we go back to these ignored complexities where, in some of these organisms, germline is not strictly segregated and perhaps not even as protected as we have come to accept. It will be interesting to see if this complexity has genomic integrity and evolutionary repercussions, or whether animals have evolved to overcome potential deleterious effects of regenerating their germline, and what these mechanisms are. It is time, because several techniques and technology available today enable us: Determining source cells in regeneration is not only possible in an increasing number of new genetically tractable organisms using genetic lineage tracing, but also possible with bioinformatics lineage-building approaches via single-cell RNAseq time series in those organisms where genetic lineage tracing may not be feasible. Identifying precise sources of germ cell regeneration will allow us to uncover the cellular programming that underlies this process. As the regeneration of this cell type has the potential to affect the next generation, investigations using sequencing and genomics are possible to test whether the regenerated germ cells have higher mutation accumulation and whether this is a driving factor in evolution via introducing variation. From cellular reprogramming to population genetics many exciting discoveries are within reach. (Box 1)

BOX 1. DEFINITIONS.

Germ cells (germline):

The terms germ cells or germline are used for cells that include fully mature gametes, as well as any cell type leading up to gametes such as the embryonic precursor of the germline, primordial germ cells, and germline stem cells.

Somatic cells (soma):

Any cell type that does not contribute to the germ cell lineage under normal developmental conditions such as muscle, neurons, and epidermis.

Gonads:

This term is used to denote the reproductive organs (ovaries and testes). Depending on the organism, these organs may have both somatic and germ components. In some examples reviewed here, it is difficult to know whether the gonads that regenerate include both cell types. I use this term when regeneration of this organ is tested, and I additionally indicate when the germ cells within gonads are also observed to regenerate.

Asexual reproduction (agametic):

Reproduction without involvement of gametes, mating of different sexes, and embryogenesis. In annelids, agametic asexual reproduction is observed as either paratomic fission or architomic fission. In paratomy, worms first build a new tail and a head, individuals still attached, these new tissues develop and clones eventually separate. In architomy, worms fragment themselves into several pieces, then rebuild missing parts. In the old annelid literature, sometimes this process is referred to as “regeneration.” In this review, I use the term “regeneration” strictly for when an external insult removes a significant amount of tissue/body axis.

Stolonization:

A mode of sexual reproduction found in Syllid annelids, where stock worms produce a series of clones specifically for sexual reproduction.

Regeneration:

Renewal and restoration of lost tissues, organs, body parts and axes. I use regeneration when there is injury/removal of tissues/organs/body axes via an external factor and exclude rebuilding segments as a result of asexual reproduction from the definition of regeneration.

DATA AVAILABILITY STATEMENT

Data sharing not applicable to this article as no data sets were generated or analyzed during the current study.