Full allogeneic fusion of embryos in a holothuroid echinoderm

Abstract

Whole-body chimaeras (organisms composed of genetically distinct cells) have been directly observed in modular/colonial organisms (e.g. corals, sponges, ascidians); whereas in unitary deuterostosmes (including mammals) they have only been detected indirectly through molecular analysis. Here, we document for the first time the step-by-step development of whole-body chimaeras in the holothuroid Cucumaria frondosa, a unitary deuterostome belonging to the phylum Echinodermata. To the best of our knowledge, this is the most derived unitary metazoan in which direct investigation of zygote fusibility has been undertaken. Fusion occurred among hatched blastulae, never during earlier (unhatched) or later (larval) stages. The fully fused chimaeric propagules were two to five times larger than non-chimaeric embryos. Fusion was positively correlated with propagule density and facilitated by the natural tendency of early embryos to agglomerate. The discovery of natural chimaerism in a unitary deuterostome that possesses large externally fertilized eggs provides a framework to explore key aspects of evolutionary biology, histocompatibility and cell transplantation in biomedical research.

1. Introduction

Intraorganismal genetic heterogeneity (IGH), and the natural occurrence of chimaeras, is perplexing [1], because it challenges the traditional biological notion of unit of selection, which relies on genetic homogeneity [2]. While IGH remains a biological puzzle, stem-cell-derived chimaeras have proven useful to study complex in vivo processes of developmental biology and regenerative medicine [3,4]. However, artificially produced interspecies chimaeras are the subject of ethical discussions, whereas the study of spontaneous intraspecific chimaeras may hold untapped potential [4]. So far, whole-body fusion among conspecifics has been documented chiefly in modular/colonial organisms with the capacity to reproduce asexually, such as sponges [5], corals [6,7], bryozoans [8] and ascidians [9–11], whereas its occurrence in unitary organisms has only been evidenced indirectly in mammals [12–15] through microsatellite DNA markers, and directly in brooded propagules of sea anemones [16] (a basal group of marine invertebrates). Here, we provide evidence of full allogeneic fusion among embryos of another unitary metazoan that sits closer to humans in the animal tree of life (in the deuterostome clade), the holothuroid Cucumaria frondosa (Echinodermata: Holothuroidea).

Cucumaria frondosa is distributed throughout the Arctic and North Atlantic oceans [17]. It is a suspension-feeding gonochoric species with an obligatory sexual reproduction. During the annual breeding season, oocytes and spermatozoa are released and fertilization occurs externally [18,19]. The large maternally provisioned (yolky) oocytes/eggs are buoyant and develop into non-feeding (lecithotrophic) larvae over approximately 45 d, until metamorphosis and settlement [18]. While feeding (planktotrophic) larvae of certain holothuroids and other echinoderms have been shown to undergo budding [20,21], embryonic fusion in an echinoderm species has never been recorded.

As evidence supporting the occurrence of natural chimaerism accumulates across many taxa, chimaeric entities are receiving considerable attention in bids to explore the unit of selection, genetic heterogeneity and allorecognition mechanisms. The present study explores the development of chimaeras in C. frondosa embryos and constitutes the first direct investigation of fusibility in an echinoderm and in a unitary deuterostome.

2. Methods

(a). Initial detection of fusion and formation of chimaeras

Given that chimaerism had never been documented in echinoderms, the occurrence of paired embryos (attached to each other) was initially thought to be a case of larval cloning by asexual budding, previously reported in planktotrophic larvae of Echinoidea and Holothuroidea [21]. To investigate the phenomenon, 15 pairs of embryos showing evidence of external tissue connection were incubated in PVC culture vessels (5 cm diameter) in which walls were removed and replaced with 800 µm mesh to allow water circulation. The vessels were placed inside a 24 l tank with running ambient seawater. Pictures and measurements of the embryos were taken daily under a motorized stereo-microscope (Leica M205FA) for a total of 7 days, which confirmed the progressive fusion of embryos and the development of whole-body chimaeras. Based on these initial observations, experiments were devised to explore morphological changes in embryonic shape and developmental time, as well as to assess the prevalence of chimaeras among the offspring population (see below). Experiments were conducted during two spawning seasons (2016 and 2017) following the same experimental protocol, but with propagules from distinct progenitors in order to confirm the findings.

(b). Egg collection and incubation

Similarly sized adults of C. frondosa weighing 4.8 ± 0.8 g immersed weight (approx. 290 g wet weight) and measuring 14 ± 1.4 cm contracted body length (±s.d.; n = 30) were hand collected by divers from two sites along the Avalon Peninsula, Newfoundland, eastern Canada (47°17′44.6″N: 52°46′8.9″W and 47°05′22.7″N: 52°55′10.0″W), at depths between 5 and 10 m. In the laboratory, visual inspection confirmed that holothuroids were healthy; they displayed normal pigmentation, firm attachment to the substrate and no skin lesions. They were acclimated in two 500 l tanks (approx. 100 individuals each) with running ambient seawater (40 l h⁻¹) at a temperature of 3 ± 1°C. An attempt to equally distribute males and females (1 : 1) in each tank was made based on gonopore morphology [18]. Continuous input of new seawater in the tanks provided natural plankton and particulate organic matter as food sources. Light was provided through fluorescent bulbs with a maximum intensity of 150 lux and photoperiod adjusted weekly to match natural fluctuations.

Upon spontaneous spawning during the normal spring breeding season, buoyant eggs were gently collected from the tanks. Egg density was estimated by collecting five 50 ml aliquots and performing counts under a stereo-microscope (Leica M205FA). Fertilization was confirmed by the elevation of the fertilization envelop and/or cleavage and eggs/embryos were then transferred to three incubators, each containing 0.8 egg ml⁻¹. Since the propagules of C. frondosa float on the surface of the water, their density was also calculated based on the surface area they covered, resulting in a concentration of 14 egg cm⁻². Incubators consisted of 5 l round black plastic vessels placed inside a 500 l tank with running ambient seawater (40 l h⁻¹; i.e. the same seawater source used in parental tanks). In order to ensure a constant water flow inside the incubators, four equal openings (40 cm² each) were made on the walls of the vessels and covered with a 400 µm mesh. Light was provided as mentioned previously.

(c). Developmental kinetics and frequency of chimaerism in the propagule population

Embryonic and larval development were monitored by collecting a subsample of 200 propagules from each of the three incubators daily until the settlement of the pentactula larvae (for a total of 45 measurements). Propagules were measured at their maximum length and photographed under the previously described microscope. A new developmental stage was scored when greater than 50% of the embryos/larvae had reached it. In order to identify the life stage(s) during which chimaeric fusion may appear, and to quantify fusion rates in the offspring population, each incubator was examined daily for evidence of fusion (for total of 45 observations). Owing to the large size of the unitary embryos of C. frondosa (approx. 700 µm), the initial stages of fusion were large enough to be visually identified with a magnifying glass and separated from unitary propagules (figure 1). Fusing embryos were photographed under the stereo-microscope (described above) and the frequency of chimaerism determined.

Figure 1.

Normal development and formation of chimaeras in embryos of the holothuroid Cucumaria frondosa. Fertilized eggs started cleavage within a few hours post fertilization. The successive cell divisions generated a blastula embryo which hatched from the fertilization envelop 5 d post fertilization. The free-swimming ciliated blastula started to elongate and developed into gastrula larva 15 d post fertilization. After 30 d, the primary tentacles and two ambulacral podia were visible in the vestibule and the embryo became vitellaria. After 45 d post fertilization, the pentactula larva was ready to settle on the substrate. Allogeneic fusion strictly occurred between hatched blastula embryos. The first stage of fusion started 5 d post fertilization with the development of yellowish bond between two embryos. In Stage 2, the blastoderms of both partners were connected by a narrow tissue connection. This connection between embryos thickened progressively through Stages 3 and 4. Full fusion (Stage 5) was generally completed inside 15 d post fertilization. Some individuals at Stage 3 fused again to form multi-chimaeras, i.e. organisms composed of three (left path) or more (right path) embryos. Scale bar represents 300 µm. See figure 2b for length of each stage of fusion. (Online version in colour.)

When identified, chimaeric embryos were collected and distributed into three separate culture vessels (1 vessel per incubator) as described above, and their development further monitored. Among all chimaeras collected, 15 were sampled from each vessel (15 chimaeras x 3 culture vessels, for a total of 45 chimaeras) daily and pictures were taken under the stereo-microscope (see above). The rotational swimming of chimaeric and non-chimaeric embryos was documented as an indicator of health because it is expected that healthy propagules exhibit circular path movement through their development at the blastula, gastrula and vitellaria stages [18]. The 9600 propagules studied measured 500–1300 µm and their size frequency distribution (50 µm bins) was established. Size classes and developmental tempo of chimaeric embryos were compared with non-chimaeric blastulae and gastrulae using one-way analysis of variance (ANOVA; α = 0.05) followed by pairwise Holm–Sidak tests (electronic supplementary material, table S1).

(d). Histology of chimaeric embryos

In order to better understand the cellular organization of fusing embryos, chimaeras at various stages of fusion were preserved in 3% formaldehyde and processed using standard histological procedures for methacrylate embedding. Sections (3 µm) were stained with haematoxylin and eosin and digitalized using an automated slide scanner (Axio Scan Z1) with a 20× objective.

(e). Effect of density and agglomeration on the development of chimaeras

The hypothesis that high propagule density might favour fusion was tested across two density treatments (high and low). Each treatment consisted of three 50 ml glass vials placed inside a 24 l tank with running ambient seawater. The low- and high-density treatments were chosen to, respectively, represent half and twice the initial egg density in the incubators where chimaerism had already been detected (see above). The low-density treatment was prepared using 0.4 egg ml⁻¹ or 7 egg cm⁻², whereas the high-density treatment consisted of 1.6 egg ml⁻¹ or 28 egg cm⁻². These values are in line with the concentration of eggs that can occur after natural mass spawning of C. frondosa in the field, which was reported to reach approximately 0.2 egg ml⁻¹ 17 h post release while the majority of eggs were dispersed in the water column and in the process of accumulating at the surface [19]. The occurrence of fusion was monitored in all the treatments daily for 21 d post fertilization. Chimaeric embryos were identified as mentioned previously, and their total number was compared between treatments using Student's t-test (electronic supplementary material, table S2).

Moreover, it was hypothesized that the gregarious behaviour of blastulae at the surface of the water favoured physical contact and fusion. Hatched blastulae were released in a 2 l vessel at a density of 0.8 propagule ml⁻¹ or 14 propagule cm² (same as in the initial egg incubators where fusion was recorded) and gently agitated to ensure uniform dispersion in the water column. A digital camera (Olympus Tough TG-3) placed above the container took pictures every 10 s for a total of 15 min. Movement and agglomeration of the embryos were monitored.

3. Results

Fertilized oocytes (eggs) of C. frondosa floated at the surface of the water a few minutes after their natural release. They measured 700 ± 150 µm and cleavage started within a few hours post fertilization (figure 1). The successive cell divisions generated a blastula (equivalent to blastocyst in mammals) approximately 3 d post fertilization. Swimming blastulae hatched from the fertilization envelop approximately 5 d post fertilization, while still gregariously floating close to the water surface, sometimes in great numbers (electronic supplementary material, figure S1).

Five different stages of fusion were determined based on morphological characters, including maximum length (figure 2a,b) and shape, as well as percentage fusion over time. Fusion started to occur after hatching from the fertilization envelop, through the formation of a bond between the blastoderms of two touching embryos (figure 1; fusion Stage 1). This early attachment was easily broken during gentle handling, but it provided enough strength for the paired embryos to spin and move together. The total length of embryos at Stage 1 of fusion was approximately 1300 µm (figures 1 and 2b). In Stage 2 (approx. 9 d post fertilization), a stronger bond approximately 150 µm thick was established between blastulae (figure 1; Stage 2) and the chimaeric entity measured 1150 ± 40 µm (figure 2b). As fusion progressed to Stage 3 (approx. 11 d post fertilization) the cellular connection between embryos thickened to 400 µm and total length of the merging pair decreased (1050 ± 55 µm; figures 1 and 2b). At this stage, fusion was about one-third complete and embryos acquired a peanut shape (figure 1). Histology revealed that there were still two visible blastocoels, but that the blastoderms were completely fused (figure 3a).

Figure 2.

Size of chimaeric and non-chimaeric embryos of Cucumaria frondosa. (a) Size frequency distribution of non-chimaeric and chimaeric embryos. (b) Size structure of chimaeric blastulae in the five stages of fusion described in figure 1. Data shown as mean ± s.e. (n = 3).

Figure 3.

Histological sections of chimaeric embryos of the holothuroid Cucumaria frondosa. (a) Fusing embryos (Stage 3) can be individually distinguished by the presence of two distinct blastocoels (B); however, blastoderm cells (BC) are mingling at the fusing area (FA). (b) Fusing embryos at Stage 5 of fusion are no longer distinguishable from each other and exhibit a single blastocoel (B), indicating complete fusion. Scale bars represent 100 µm. See figure 1 for illustrations of each stage of fusion. (Online version in colour.)

Chimaeric embryos acquired a smoother oval shape when fusion was two-thirds completed in Stage 4, approximately 13 d post fertilization (figure 1). The bond thickened to 550 µm and there was a further decrease in overall length to 1000 ± 35 µm (figures 1 and 2b). After approximately 15 d post fertilization, non-chimaeric blastulae started to elongate and developed into gastrulae (normal development illustrated in figure 1) and reached a length of 930 ± 50 µm. Concurrently, full fusion (Stage 5) was completed in chimaeric propagules (figure 1), which displayed a round shape, homogeneous ciliation and a single blastocoel/gastrocoel (figure 3b). These whole-body chimaeras had a mean length of 980 ± 40 µm (figure 2b), which was significantly larger than non-chimaeric blastulae (700 ± 60 µm) and gastrulae (930 ± 30 µm; F_2,31 = 24.6, p = 0.011). Moreover, their volume was 0.5 mm³, which was roughly twice that of non-chimaeric blastulae.

Further fusion events were detected at Stage 3, either among chimaeras or between chimaeras and blastulae, resulting in multi-chimaeric propagules (figure 1). The frequency of occurrence of these multi-chimaeras was low (0.02% of the population). However, some multi-chimaeras composed of three entities reached the full fusion stage (figure 1; Stage 5), resulting in notably large propagules with a length of 1100 ± 38 µm and a volume of approximately 0.7 mm³ (approx. 5 times the volumetric size of non-chimaeric blastulae). Multi-chimaeras involving more than three entities failed to develop.

Apart from directly monitoring fusion events, we measured the size structure of embryos at the post-hatching blastula stage and detected two dominant cohorts (figure 2a). The smaller cohort (maximum length of 700 ± 60 µm, volume of approx. 0.18 mm³) included those that did not undergo allogeneic fusion. The second cohort was represented by larger embryos (1120 ± 51 µm; figure 2a), subsamples of which were microscopically shown to be the result of fusion (figure 1). Confirmed fusion occurred in 8.6 ± 0.6% of the entire propagule population (9600 embryos). It was seen in all independent cultures of C. frondosa (n = 12) across two separate breeding seasons (2016 and 2017) involving progenitors collected from different locations, supporting the fact that allogeneic fusion in this species is a natural widespread phenomenon. The post-metamorphic just-settled juveniles still displayed marked size variations, between 910 and 1160 µm in length.

4. Discussion

Natural fusion of early life stages has only been reported previously in cnidarians and sponges. In brood-protected larvae and embryos of the sponge Chalinula sp. and the sea anemone U. felina, natural fusion occurred in 40% and 3% of the propagules, respectively [16,22]. Fusion rates at later stages (newly settled polyps) were 78% in the soft coral Clavularia hamra, 90% in Nepthea sp., 40% in Heteroxenia fuscescens and 80% in Parerythropodium fulvum fulvum [7]. Natural fusion was also documented among modular adults, including 3–5% of colonies in the coral Acropora millepora [6], 40–90% of colonies in four species of soft corals [7] and 8–73% of colonies in botryllid ascidians [23].

The mechanisms that lead to the establishment of chimaeras revolve around allorecognition [24,25]. Studies have identified allogeneic responses controlling fusion or rejection in sponges; however, the molecular mechanism is still poorly understood [26]. According to grafting trials in hydroids, fusion may occur among colonies when they share at least one of the histocompatibility loci alr1 and alr2 [27]. In ascidian, fusion among oozoids is controlled by the self-recognition fusibility/histocompatibility (Fu/HC) locus [28–30]. Allograft studies suggest that echinoderms have a non-adaptive, activation type immune response based on coelomocyte activity; however, the specificity of the response is still unclear [31]. Hence, whether fusion in C. frondosa occurs among embryos that share at least one histocompatibility locus (i.e. between full and half siblings) or non-related embryos remains to be determined through controlled fertilization and molecular analyses. Such information might provide insights into the potential use of chimaeric embryos produced by C. frondosa in biomedical research, especially with the recent publication of the holothuroid genome [32].

Allogeneic fusion has been associated with maturation of the alloimmune system. Natural chimaeras have been detected among free-swimming larvae and embryos of sponges [22]. In the brooding sea anemone U. felina, fusion of either embryos or larvae seemed to occur inside the brooding female [16]. In soft corals, fusion occurred among young polyps a few weeks after settlement [7]. Although the stages and mechanisms involved in the formation of natural chimaeras in mammals are still elusive [14,15], interspecies chimaeras produced artificially are generated only during the blastocyst stage [3]. Similarly, natural fusion in C. frondosa only occurred among hatched blastulae; no evidence of fusion was found in either earlier (unhatched) or more developed embryonic and larval stages (i.e. gastrula, vitellaria and pentactula). This result indicates that the alloimune system might not be completely developed in blastula embryos of C. frondosa, preventing them from recognizing genetically distinct (non-self) tissues/embryos. This also suggests that allorecognition mechanisms in echinoderms may develop within a brief window during early ontogeny (i.e. between just-hatched blastulae and full development of gastrulae, similar to evidence from mammals). Fusion at earlier stages (unhatched embryos) was presumably prevented because the elevated fertilization envelope acted as a physical barrier. The present study documented the very early process of natural fusion in C. frondosa, with the appearance of a yellowish bond linking two blastulae, which may be the first evidence of alloimmune interaction, leading to the development of a tissue connection between blastoderm cells. Although this bond might represent the first communication between fusing embryos, its nature and composition remain unclear. Further studies using molecular markers may help to characterize its origin. Rejection was never detected, unlike reports in corals where a visible inflammatory response occasionally developed in the contact area, sometimes leading to the death of one of the partner [6].

After hatching, blastulae of C. frondosa showed a strong tendency to aggregate with each other at the surface of the water column (electronic supplementary material, figure S1), a phenomenon that was not seen in the older stages, where no fusion was detected. Moreover, experimental trials revealed that the formation of chimaeras was density dependent. There were statistically higher proportions of confirmed chimaeras in the high-density blastula treatment (1.6 ± 0.2%) than in the low-density treatment (0.6 ± 0.3%; t₂ = 3.5, p = 0.02). While this result indicates that, all else being equal, higher propagule density favoured fusion in embryos of C. frondosa, these small-scale experimental trials were not optimal for fusion. The maximum fusion rates (8.6%) were obtained at medium density in the much larger culture vessels (5 l) where 4000 propagules were incubated, probably because it increased the chance of embryos sharing fusibility/histocompatibility loci. Aggregation of hatched blastulae at the surface of the water may involve the hyaline layer, which is an extracellular matrix surrounding early echinoderm embryos that is composed of glycoproteins and is necessary for blastomere adherence [33]. Although the nature and interactions of these adhesive molecules remain uncertain, they may favour contact agglomeration of blastulae leading to the first step of fusion in C. frondosa.

On a fundamental level, the ecological and evolutionary benefits of different life-history strategies in echinoderms have been investigated on theoretical grounds. It has been suggested that larval cloning in echinoderms might be an adaptation to increase survival under favourable conditions of water temperature and food concentration [21,34], and to minimize visual detection and predation by reducing propagule size [35]. Clonal propagation may also provide evolutionary advantages by preventing internal conflicts of genetically distinct cell lineages [36]. Conversely, the evolutionary role of chimaerism (which is essentially the opposite of cloning) is rather controversial, fuelling discussions about the potential benefits of fusion [1,37]. Fusion between allogeneic entities is expected to confer genetic variability, developmental synergism, and immediate increase in size and survivorship [38,39]. An overall gain in fitness has been documented in multi-chimaeric ascidians [37,40], corals [6] and sea anemones [16]. Our results confirmed that fully fused chimaeras are significantly larger than singletons, with a volume two to five times greater than non-chimaeric blastulae. This difference in size is still clear in post-metamorphic juveniles. Hence, the occurrence of chimaerism may explain reports of high variability in the size of just-settled juveniles in C. frondosa [18,41] and other holothuroids [42], suggesting that fusion among embryos may be common in holothuroid species from other climes and with different modes of development. Because C. frondosa produces maternally provisioned lecithotrophic larvae that do not feed until settlement (when the mouth and anus are formed), a size increase through feeding can be ruled out. The greater volume of chimaeric embryos may increase buoyancy, allowing them to disperse efficiently during the long pelagic phase (approx. 45 d). Moreover, larger juveniles that are a product of fusion may gain competitive advantage during the settlement phase due to increased resistance to water turbulence and predation [38], enabling chimaeric juveniles to better cope in dynamic competitive habitats where food and space are limited. Although chimaerism may provide potential benefits, early zygote fusion before the germline is sequestered may also allow these deuterostomes to develop germ-cell competition/parasitism. The occurrence of cell-lineage parasitism was previously reported in chimaeric ascidians, and sexual parasitism was documented in anglerfish [43].

Overall, Cucumaria frondosa emerges as a promising model organism for investigating intraspecific chimaeras because natural fusion occurs among large free-swimming blastulae (i.e. not inside the body or womb, as in sea anemones and mammals). The potential costs/benefits of chimaerism in this unitary species definitely deserve further investigation due to the strategic position of Echinodermata within the deuterostome lineage [44]. Unlike more primitive clades, such as colonial and unitary cnidarians, where visual evidence of incomplete fusion can occur (electronic supplementary material, figure S2), demonstrating the existence of chimaeric holothuroids in the wild will require the development of appropriate molecular markers and techniques. Meanwhile, in light of recent genomic advances [32], the present discovery offers a unique framework to explore cellular self-recognition mechanisms and develop our understanding of the evolution of immune systems in higher metazoans.

Ethics

Animals were collected by the Field Services of the Department of Ocean Sciences of Memorial University with the required permits issued by the Department of Fisheries and Oceans.

Data accessibility

The datasets supporting this article have been uploaded as part of the electronic supplementary material.

Authors' contributions

Experiments were designed by B.L.G., J.-F.H. and A.M. Experiments and data analyses were performed by B.L.G. The manuscript was written by B.L.G., J.-F.H. and A.M. All authors gave final approval for publication.

Competing interests

We have no competing interests.

Funding

This study was supported by grants from the provincial Research and Development Corporation (RDC), the Department of Fisheries and Aquaculture (DFA) of Newfoundland and Labrador, the Canadian Centre for Fisheries and Innovation (CCFI), the Natural Sciences and Engineering Research Council of Canada (NSERC) and the Canada Foundation for Innovation (CFI).