Expression of FoxA and GATA transcription factors correlates with regionalized gut development in two lophotrochozoan marine worms: Chaetopterus (Annelida) and Themiste lageniformis (Sipuncula)
- 7.7k Downloads
A through gut is present in almost all metazoans, and most likely represents an ancient innovation that enabled bilaterian animals to exploit a wide range of habitats. Molecular developmental studies indicate that Fox and GATA regulatory genes specify tissue regions along the gut tube in a broad diversity of taxa, although little is known about gut regionalization within the Lophotrochozoa. In this study, we isolated FoxA and GATA456 orthologs and used whole mount in situ hybridization during larval gut formation in two marine worms: the segmented, polychaete annelid Chaetopterus, which develops a planktotrophic larva with a tripartite gut, and the non-segmented sipunculan Themiste lageniformis, which develops a lecithotrophic larva with a U-shaped gut.
FoxA and GATA456 transcripts are predominantly restricted to gut tissue, and together show regional expression spanning most of the alimentary canal in each of these lophotrochozoans, although neither FoxA nor GATA456 is expressed in the posterior intestine of Chaetopterus. In both species, FoxA is expressed at the blastula stage, transiently in presumptive endoderm before formation of a definitive gut tube, and throughout early larval development in discrete foregut and hindgut domains. GATA456 genes are expressed during endoderm formation, and in endoderm and mesoderm associated with the midgut in each species. Several species-specific differences were detected, including an overlap of FoxA and GATA456 expression in the intestinal system of Themiste, which is instead complimentary in Chaetopterus. Other differences include additional discrete expression domains of FoxA in ectodermal trunk cells in Themiste but not Chaetopterus, and expression of GATA456 in anterior ectoderm and midgut cells unique to Chaetopterus.
This study of gene expression in a sipunculan contributes new comparative developmental insights from lophotrochozoans, and shows that FoxA and GATA456 transcription factors are part of an ancient patterning mechanism that was deployed during early evolution of the metazoan through gut. The common utilization of FoxA and GATA456 throughout gut formation by species with contrasting life history modes indicates that both genes are core components of a gut-specific gene regulatory network in spiralians. Despite a highly conserved pattern of early development, and probably similar ontogenic origins of gut tissue, there are molecular differences in gut regionalization between lophotrochozoan species.
KeywordsExpression Domain Gata Factor GATA456 Expression FoxA Gene FoxA Expression
Formation of a through gut was a key innovation during the radiation of the Metazoa. Although the origins of an internal digestive tract are uncertain, it probably occurred early in animal evolution [1, 2]. Considering this, molecular mechanisms that specify cells and tissues of such an ancient organ system are expected to show some level of conservation across a broad range of taxa [1, 3, 4]. Multiple studies support this expectation, especially for several transcriptional regulators of endoderm [3, 5, 6]78. Over evolutionary time, animals generated a variety of digestive organs and adaptive feeding mechanisms that enabled them to utilize a wide range of food sources . How different gut anatomies develop, how planktotrophic and lecithotrophic life histories influence gut evolution, whether larval and adult intestinal systems are fundamentally different, and which molecular mechanisms explain the variation in gut morphology between closely related animals are still not well understood. In a previous study , we described morphologic and molecular aspects of gut development in the marine polychaete, Capitella teleta Blake, Grassle & Eckelbarger 2009, known for years as Capitella sp. I. In this paper, we extend that work with developmental data obtained from two additional marine worms.
The 'segmented' annelid worms are one of the most speciose and ecologically diversified protostome animal groups, and are well represented within one of three metazoan superclades, the understudied Lophotrochozoa . One of our study species is the parchment worm, Chaetopterus variopedatus Cuvier 1827, sensu Enders (1909), which M. E. Petersen (personal communication) suggested is most likely not C. variopedatus, but an undescribed species. We refer to this polychaete as Chaetopterus, a marine annelid that has been used both in classic [12, 13, 14] and more recent [15, 16, 17, 18] developmental studies. Chaetopterus lives in U-shaped tubes within sublitorral, mud-bottom habitats along the Northeast Atlantic coastline and presumably is a broadcast spawner . Recent phylogenetic analyses consider the Chaetopteridae as one of the more basal annelid groups .
The second species is a sipunculan worm, Themiste lageniformis Baird 1868, for which there are a few studies describing its reproduction [21, 22] and general patterns of development [21, 23]. Adult females of T. lageniformis are considered to reproduce by facultative parthenogenesis, and both sexes are distributed in benthic marine habitats along coastlines of Africa, India, Indo-Pacific, Hawaii and Florida [21, 24]. The occurrence of fossil sipunculans from the Lower Cambrian implies they are an ancient group of animals , and may be the sister group to annelids . However, an accumulation of data from morphologic [27, 28, 29, 30], mitochondrial [31, 32] and molecular phylogenetic sources [20, 33, 34, 35] suggest that the monophyletic Sipuncula  belong within Annelida, although they are not consistently affiliated with any particular annelid group.
These two species provide a valuable comparison for examining the evolutionary development of gut formation because Chaetopterus and T. lageniformis develop very different larval and adult gut morphologies. Descriptions of larval morphology , anatomy  and development [15, 39, 40] demonstrate that Chaetopterus larvae form a functional, tripartite intestinal system consisting of the pharynx, midgut and hindgut. The larvae of T. lageniformis form a U-shaped gut without distinct compartments . Moreover, their respective life histories are dissimilar. Chaetopterus larvae are planktotrophic, begin feeding within 36 hours of development, and they are capable of extended planktonic periods exceeding 30 days . The unique pelagosphera larvae of T. lageniformis are lecithotrophic, swimming on or near the bottom in laboratory cultures for up to 9 days , and do not feed until after metamorphosis. In both species, metamorphosis includes substantial changes in gut morphology between larval and juvenile forms.
We are interested in understanding the molecular basis for distinct regional differences in gut morphology. Important candidate genes likely to be involved in gut regionalization include both Fox and GATA factors. The FoxA ('Forkhead box' A) gene belongs to a multi-family group of 'winged-helix' transcription factors with structurally similar 110-amino acid DNA-binding domains [42, 43]. GATA genes belong to a family of transcription factors with one or two class IV zinc-finger motifs . Individual GATA factors are typically assigned to one of two subclasses, GATA1/2/3 or GATA4/5/6 , and their evolution has been described in detail . Both FoxA and GATA456 gene members are part of a larger endomesoderm specification network of genes that have been shown to have regulatory roles in endoderm and mesoderm formation during embryogenesis and larval morphogenesis in a nematode , fly , sea urchin [48, 49, 50] and frog , and are expressed in similar tissues in many other animals. Within Annelida, gene expression data suggest that two paralogs of FoxA are in gut precursor cells of Hydroides elegans , and there is mesodermal, although no endoderm expression, of a GATA456 ortholog in Platynereis dumerilii . Only for the polychaete C. teleta are there expression data for transcription factors from both gene families, and their patterns are clearly implicated in gut formation and regionalization .
In this study, we isolated and identified FoxA and GATA456 orthologs from Chaetopterus and T. lageniformis. We then characterized their expression patterns to help clarify whether these genes in annelids have the potential to act as conserved or novel key regulators of the gut developmental process [50, 54, 55]. We also describe gut formation for both species to guide interpretation of in situ hybridization patterns. In each worm, FoxA and GATA456 factors show a pattern of regionalized expression in both ectodermal and endodermal domains of the developing gut tube. We interpret these expression patterns in the context of life history mode and the inherent morphogenesis associated with gut formation. This is the first report of any gene expression from a member of the Sipuncula.
Larval gut formation in Chaetopterus
Embryonic and larval gut formation in Themiste lageniformis
Approximately 2 hours after chemically activating (see Methods) mature oocytes of T. lageniformis, a unipolar cleavage furrow moves across the egg surface, initiating a conserved program of spiral, holoblastic cleavage. At the completion of first cleavage, a single polar body is visible on the animal pole and the egg is divided into a large CD cell and a smaller AB cell (Figure 2B). The D cell is the largest blastomere at the four-cell stage, and at eight cells, micromeres of the A, B and C quadrants are larger than their respective macromeres, with the D macromere being the largest of all eight blastomeres (not shown). In mid blastula (stage 1), a single polar body identifies the animal pole, and relatively large, yolky cells are positioned at the vegetal pole (Figure 2B). During late stage 2, gastrulation occurs by epiboly, and a blastopore forms on the posterior ventral face of the embryo . The stomodeum is detectable at stage 3 as a pronounced indentation on the embryo's ventral posterior face, and initially extends in a dorsal-anterior direction under the posterior rim of a band of conspicuously large, non-ciliated prototroch cells (Figure 2B, Figure 3A-C). The pretrochal anterior end of the stage 3 embryo is marked by an apical groove  between the animal pole and the prototroch. Internally, there is a centralized yolk-rich region visible in both live and fixed specimens (Figure 2B, Figure 3C), and presumptive mesoderm cells flank each side of the endoderm (Figure 3C). Stage 4 is characterized by an episphere with a pair of larval eyes, a band of large prototroch cells, bilateral sets of retractor muscle fibers on each side of the endoderm, and a circumferential band of non-ciliated metatrochal cells directly posterior to the prototroch (Figure 2B, Figure 3D-F). The stomodeum is rimmed with ciliated cells, and forms a 'bowl-shaped' cavity that interrupts the ventral-most pair of prototroch cells (Figure 2B, Figure 3D). The stomodeum is closed at its dorsal end, where it is lined with a rosette of larger cells (Figure 3F). Endoderm cells are situated in the yolky interior, posterior to the stomodeum, although they are difficult to visualize at this stage. Between stages 4 and 5, the hyposphere extends in a posterior direction, producing a conical-shaped embryo. The stomodeum is broader and extends deeper interiorly; however, it is still closed at its dorsal end where it abuts a group of interior cells (Figure 2B, Figure 3G-I). The ventral pair of head retractor muscles wrap around the stomodeal cavity at their anterior end (Figure 2G). Presumptive endoderm cell nuclei are located to the interior of the dorsal wall of the stomodeum, but are not detected in the central trunk region posterior to the metatrochal band (Figure 3H, I).
Stage 6 represents a transition to the pelagosphera larva, and is considered analogous to the first of two 'metamorphic' stages in many sipunculans . The large prototroch cells are still present but reduced in size, and the metatroch is now ciliated. Repeated extensions and contractions of the body gradually cast off part of the egg envelope on the anterior end, and a larval cuticle is formed around the body . At this stage, the stomodeum is open to the environment and lined with cilia at the surface. The interior of the stomodeum is packed with cells and now is oriented along the anterior-posterior (A/P) axis (Figure 2B, Figure 3J). The midgut contains numerous endoderm nuclei, and extends from the stomodeum through the trunk to the dorsal posterior side of the larva, narrowing along the A/P axis (Figure 3J). At stage 7, the large prototroch cells are reduced or lost, and ciliated cells surround the mouth and form a new prototrochal band that encircles the head. The metatroch is active, and propels the larva during crawling or gliding motions along the substrate surface. The ventral nerve cord is apparent as a condensation of nuclei along the ventral midline of the trunk. The foregut is packed with cells that will form the esophagus, and the midgut sweeps from ventral anterior to dorsal posterior (Figure 2B). Stage 8 larvae swim into the water column. These larvae now have an elongate esophagus region that extends posteriorly from the mouth to the intestine, which broadens at its posterior end and then narrows as it curves toward the larva's dorsal-posterior body wall (Figure 2B, Figure 3K, L). Williams  described both yolk granules and lipid droplets within membrane-bound gut cells, and we see yolky endoderm cells and cell nuclei (Figure 3K, L) dispersed along the length of the intestine. It is also clear from viewing live specimens that there is a considerable amount of yolk moving freely within the coelom. The pelagosphera larva at stage 9 has numerous cilia surrounding the mouth and ventral face of the head, also present in stages 7 to 8, and an actively beating metatrochal band (Figure 2B, Figure 3M). A thick cuticle covers the body, and at the posterior end there is a terminal organ, which serves the larva as an adhesive mechanism for attachment to the substrate . The relatively compact esophagus is oriented in an anterior-posterior direction, and is constricted at its junction with the intestine (Figure 3M). The intestine is broader along its anterior section, narrows at the posterior end of the descending arm, and then curves and ascends toward its terminus where the rectum and anus will develop (Figure 2B, Figure 3M). Larvae at stages 8 and 9 are capable of fully retracting their anterior ends, which temporarily repositions most of the gut tube within the posterior end of the coelomic cavity.
Characterization and orthology assignment of FoxA and GATA genes
Developmental expression of Cht-FoxA and Cht-GATA456a
Developmental expression of Tl-FoxA, Tl-GATA456a and Tl-GATA456b
In this study, we provide novel information on gut development in the sipunculan, Themiste lageniformis, which supplements previous work . Using labeling techniques and confocal microscopy, we identified presumptive endoderm and mesoderm cells relative to positional landmarks such as the apical groove, stomodeum, muscle retractors and metatrochal cells. In the trochophore-like stage before elongation, endodermal cells are located adjacent and posterior to brain cells in the anterior of the animal. In lecithotrophic pelagospherae, there is a compact aggregation of stomodeal cells occluding the presumptive esophagus, whereas a symmetric epithelium lines the foregut canal in the planktotrophic larval digestive system of Chaetopterus. The posterior end of the larval gut in Themiste is initially positioned midway along the dorsal trunk body wall, and both the intestine and larval body elongate posterior to the anal region resulting in the U-shaped gut in adults, a morphology that is considered to be an important adaptation among tubular burrow-dwelling sipunculans . By contrast, the linear gut of polychaetes usually terminates within or immediately adjacent to the pygidium at the worm's posterior end , even when the anus is positioned dorsally, such as in the larvae of tubiculous species (such as serpulids and chaetopterids). By comparing gut formation in Chaetopterus, Themiste and Capitella teleta , it is clear that a distinct digestive architecture develops in the larvae of each worm, which may correspond with divergent life history modes and feeding mechanisms.
Our results reveal that differences among species in gut architecture are mirrored in the expression patterns of regulatory 'gut genes.' During early development, FoxA family orthologs are expressed in the vegetal plate blastomeres of Chaetopterus, Themiste and Capitella; however, FoxA expression in the endoderm of Themiste is more persistent and extensive. In endodermal cells of Chaetopterus and Capitella, FoxA appears to be downregulated shortly after gastrulation and well before a gut tube is detectable, yet FoxA transcripts are expressed in the endoderm of Themiste during midgut morphogenesis. This could reflect a longer duration of lecithotrophic development, later onset of feeding, and/or the unique anatomy of sipunculan coiled-helix intestinal systems [56, 57, 59, 60]. Species variation of FoxA expression in the foregut is less apparent and may simply be related to the time for completion of a functional gut, which requires 1.5 days, 9 days and 2 to 3 weeks in Chaetopterus, Capitella and Themiste, respectively. In Chaetopterus, FoxA is expressed in the stomodeum 11 hours after fertilization, well before a distinct foregut epithelium is apparent (27 to 28 hours post-fertilization) and later FoxA is expressed at the oral surface in feeding larvae. By contrast, the stomodeum and foregut epithelium are formed after approximately 4 and 5 days, respectively, in the lecithotrophic metatrochophore of Capitella, and a distinct foregut epithelium is not even visible in the lecithotrophic pelagosphera of Themiste. The FoxA gene is also expressed throughout foregut morphogenesis in these two worms, although expression does not extend to the stomodeal surface cells as it does in Chaetopterus. Different polychaete species of Hydroides (for example, H. hexagonus; H. uncinata) generate planktonic trochophores with a tripartite gut 20 hours after fertilization http://hdl.handle.net/1912/295 and H. elegans has the ability to feed after just 9 hours (B. Nedved, personal communication). Accordingly, Arenas-Mena  found high levels of FoxA (He-FoxA1 and He-FoxA2) expression in foregut and endoderm precursors in 3 to 4 hour blastulae of H. elegans, although endoderm expression was transient and declined shortly after gastrulation in that polychaete as well. Thus, in the foregut domains of the species mentioned here, FoxA is commonly expressed during a broad temporal window from cleavage through larval development, regardless of developmental mode or tempo. In contrast, in the midgut endoderm of Themiste, there seems to be a heterochronic shift in the duration of FoxA expression relative to expression in the endoderm of polychaetes.
The FoxA gene is also expressed in the hindgut of Chaetopterus, Themiste, Capitella  and H. elegans , yet it extends much further internally into presumptive intestinal endoderm in Themiste. During stages 6 to 8 in the pelagosphera of Themiste, FoxA is expressed along the majority of the length of the gut, contrasting with the more restricted, separate foregut and hindgut expression domains of FoxA in Chaetopterus and Capitella. Sipunculans generally have a complex hindgut region in which highly folded, ciliated columnar epithelia and more extensive musculature develop . A FoxA ortholog is expressed in both ciliated columnar epithelia and highly muscularized tissues of the pharynx in Capitella ; however, there is no comparable pharyngeal anatomy in Themiste or adult sipunculans . It will be informative to examine FoxA expression in the larvae of sipunculan species that have a pharynx, which includes muscularized feeding structures such as the buccal organ and lip gland [29, 56, 59], to see if there is a significant shift in the amount and extent of FoxA expression between anterior and posterior regions in different sipunculans.
In all three species we have examined, FoxA expression in the posterior gut is interior to surface ectoderm and may have a role in specifying the transition between hindgut/rectum and anus. In Drosophila, there is a 'constellation' of regulatory and signaling genes (caudal, FoxA/forkhead, brachyenteron, wingless, hedgehog, dpp) that comprise a regulatory network controlling gastrulation and hindgut development [62, 63, 64, 65]. In Capitella, there is FoxA expression in posterior endoderm and ectoderm, where the hindgut meets cells of the anus. Here, FoxA appears to be co-expressed with caudal (cdx) in this posterior domain , and it both precedes and overlaps with posterior expression of hedgehog (hh), which is adjacent to the expression of wingless (wg/Wnt1) in the anus . This pattern implies that a gene regulatory network similar to that in Drosophila may control hindgut formation in polychaetes. Based on the spatially and temporally conserved patterns of FoxA expression in the hindgut and foregut of Chaetopterus, Themiste and Capitella, we hypothesize that FoxA acts as a key regulator of anterior and posterior gut formation, and is not restricted to a single germ layer.
Outside the gut, FoxA is expressed in a discrete patch of ectodermal cells on the ventral midline in stages 4 to 8 in Themiste. Although FoxA is also expressed in several ectoderm cells on the ventral midline of Chaetopterus, this expression is transient and not detectable in L2 and L3 larvae when nerve fibers appear in this region (N. Meyer, personal communication). In Themiste, the patch of FoxA-positive cells on the ventral midline persists through 'first metamorphosis' and appears to correspond with serotonergic cells of the ventral nerve cord (VNC) in stages 6 to 7 pelagospherae (N. Meyer, personal communication). It is possible that these patterns reflect temporal differences in VNC specification between planktotrophic (Chaetopterus) and lecithotrophic (Themiste) development. In Themiste, FoxA is also expressed in bilaterally symmetric cell clusters along lateral margins of the trunk in stages 6 and 7 (not shown). This and the ectodermal expression at the ventral midline probably correspond to components of the sipunculan nervous system, and represent a co-option of the FoxA gene outside its conserved role in gut development.
GATA transcription factors are expressed in both the endoderm and mesoderm of Chaetopterus and Themiste. We isolated a single ortholog in the GATA4/5/6 subclass from Chaetopterus and two orthologs from Themiste, although additional subclass members may exist. Expression of GATA factors in Themiste and Chaetopterus are consistent with our previous findings for Capitella in which we identified three GATA4/5/6 subclass genes , each with a distinct expression pattern: Ct-gataB1 is expressed in endoderm, Ct-gataB2 in endoderm and visceral mesoderm, and Ct-gataB3 is restricted to visceral mesoderm. One of the GATA456 factors from Themiste, Tl-GATA456a, appears to be more similar to one of three GATA456 factors in Capitella than it is to known GATA orthologs from any other polychaetes. Our amino acid sequence analysis indicates that these genes (Tl-GATA456a and Ct-gataB1) are closely related orthologs of each other. Both genes are expressed exclusively in the endoderm, which may indicate a conserved role after gene duplication events within this subclass of GATA factors [10, 45, 53]. In addition, one GATA factor each from Chaetopterus (Cht-GATA456a), Themiste (Tl-GATA456b) and Capitella (Ct-gataB2) is simultaneously expressed in both endodermal and mesodermal cell types associated with the midgut, showing conservation of an additional GATA456 ortholog in annelid and sipunculan gut development. In the polychaete Platynereis dumerilii, Gillis et al.  identified a single GATA factor (PdGATA456) whose expression pattern does not include any endoderm (see below), which contrasts with endoderm expression for at least one GATA456 factor in Chaetopterus, Themiste and Capitella.
GATA456 expression in mesoderm varies between polychaete annelids. During early development, a stereotypic spiral cleavage program is common among annelids, sipunculans and other lophotrochozoans (for example, echiurans, molluscs, nemerteans, polyclad turbellarians), in which the 4 d cell is termed the 'mesentoblast,' and typically gives rise to adult mesoderm [68, 69, 70]. The single GATA factor (PdGATA456) identified in Platynereis  is expressed in cells of the paired mesodermal bands, "most likely in descendants of the mesodermal progenitor cells 4d1 and 4d2." An important contrast between their study and ours is that in Chaetopterus, Cht-GATA456a is expressed in seven vegetal-plate blastomeres, which appear to include all fourth quartet micromeres except 4 d, the progenitor of 4d1 and 4d2. A similar pattern was found in Capitella where Ct-gataB1 expression is reduced in 3 D just before the birth of 4 d . Thus, in Chaetopterus and Capitella, transcription of one GATA 456 factor seems to be segregated away from the 4 d lineage, indicating that there is species-specific variation of GATA456 expression in annelid mesodermal precursors during early development. At later stages, the expression of PdGATA456 is detectable throughout the entire mesodermal band , yet the expression of a GATA456 ortholog is clearly discontinuous among presumptive mesoderm cells flanking the endoderm in Chaetopterus, Themiste and Capitella, which indicates expression within a mesoderm sublineage. Moreover, Gillis et al.  reported that PdGATA456 expression is associated with the segmentally repeated chaetal sacs. The segmental expression of GATA456 is not associated with chaetal development in Capitella or in epidermal development in the achaetous non-segmented sipunculan, Themiste; however, GATA456 expression has not been examined during chaetal development in Chaetopterus.
We observed several other differences between species in the expression patterns of FoxA and GATA456. Neither FoxA nor Cht-GATA456a is expressed in the hindgut of Chaetopterus, although there was GATA456 expression in the corresponding regions of Themiste and Capitella. There may be another GATA ortholog or even a different regulatory gene in Chaetopterus that is expressed in this hindgut tissue. Additionally, because there are no discrete compartments in the larval gut of Themiste, GATA456 expression does not correlate with distinct gut compartment boundaries, whereas the segregation of FoxA and GATA456 expression into distinct gut compartments is definitive in Chaetopterus. The GATA456 expression pattern in the gut of Themiste may instead represent the embryonic origin of this tissue (that is, in endodermal precursors). Finally, in Chaetopterus, Cht-GATA456a appears to be expressed in specialized cells in the midgut (endoderm) and head ectoderm (neural) that are not directly comparable to any patterns of GATA456 expression in Themiste, Capitella or Platynereis. This represents a possible co-option of the Cht-GATA456a gene in cell types that are present in Chaetopterus but not in the other animals. Collectively, it appears that particular developmental roles for members of the GATA456 subclass have been partitioned among species and between germ layers during the diversification of annelids.
It is likely that both transcription factors are near the top of gut-specific gene regulatory networks in spiralians. Orthologs of FoxA and GATA factors are integral components of endomesoderm networks in a nematode [46, 71] and sea urchin [50, 55] and required for digestive tract development in fly [72, 73] and mouse embryos [74, 75]. If we also consider the expression patterns in embryos as diverse as a hemichordate , mollusc [77, 78] or an anthozoan cnidarian , there is evidence of ancient conservation or 'deep homology'  for the roles of FoxA and GATA456 genes in patterning metazoan intestinal systems. Regionalized expression of FoxA and GATA456 orthologs along the gut tube of annelid lophotrochozoans, taken together with conserved expression patterns in ecdysozoan and deuterostome animals, implies that a through gut may have been present during the earliest stages of bilaterian evolution, even in the bilaterian ancestor itself. Although it is clear that lophotrochozoans use a common set of core gut patterning components, our results reveal significant molecular differences between species and could indicate modifications leading to the evolution of diverse intestinal systems. Investigations of genes further downstream may uncover more dramatic species-specific differences in the process of gut-related tissue specification or cell differentiation.
Animal collection and handling
Adult polychaete worms of Chaetopterus variopedatus Cuvier 1827, sensu Enders (1909) were collected near Woods Hole, MA, USA by the Marine Resources Center at Marine Biological Laboratory (MBL). M. E. Petersen (personal communication) examined the material from mud bottoms in the Woods Hole region and has suggested that it probably is not C. variopedatus, but a new, undescribed species. Mature worms were maintained in their tubes in flowing seawater for 1 to 2 weeks. Gamete collection, fertilization and the handling of embryos and larvae were conducted according to published protocols .
Adult sipunculan worms of Themiste lageniformis were extracted from coral rubble found in the shallow subtidal region along the south shore of Oahu, Hawaii. Coral 'boulders' were maintained in flowing seawater tables at Kewalo Marine Laboratory, University of Hawaii at Manoa for up to 1 to 2 months and broken open to collect worms. Mature specimens were placed into glass finger bowls containing filtered seawater (FSW; 22 μm filter) for 1 to 2 hours, followed by dissection of coelomic contents into plastic petri dishes coated in a gelatin solution (1.0% gelatin; 0.37% formaldehyde; dH2O). Mature oocytes were isolated from other coelomic materials by washing in FSW. Embryonic development of T. lageniformis was activated by treatment with 1.0 mM cAMP (N, 2'-O-dibutyryladenosine 3': 5- cyclic monophosphate; Sigma, St. Louis, MO, USA) in coated petri dishes containing oocytes and FSW. cAMP stock solution was prepared at 0.5 M in dimethylformamide (DMF). cAMP treatments were terminated after 1.5 hours by several FSW exchanges. Developing embryos and larvae were maintained at room temperature (RT) with daily exchanges of FSW containing antibiotics (0.6 mg/ml penicillin; 0.5 mg/ml streptomycin). A detailed protocol is available upon request.
Cloning of FoxA and GATA456 genes
Total RNA was collected from Chaetopterus oocytes, embryos at 3 to 9 hours post-fertilization, and mixed larval stages. 5' and 3' cDNA templates were synthesized using rapid amplification of cDNA ends (RACE (SMART RACE system; Clontech, Mountainview, CA, USA). DNA fragments of Chaetopterus FoxA and GATA456 were recovered from the cDNA template by degenerate PCR. A 249 bp fragment of the Cht-FoxA winged-helix domain was isolated with the following nested primer set: A1Fout: 5'-AAR CCN CCN TAY WSN TAY AT -3' and A2Rout: 5'-TAR CAN CCR TTY TCR AAC AT-3', followed by A2Fin: 5'-TAY ATH WSN YTN ATH CAN ATG -3' and A1Rin: 5'-CCR TTY TCR AAC ATR TTN CC -3'. A 348 bp fragment of the Cht-GATA456 C4 zinc-finger domain was isolated using the following semi-nested primer set: GTF1out: 5'-GAR TGY GTN AAY TGY GGN GC -3' and GTR1out: 5'-GGY TTN CKY TTN CKN GTY GT -3', followed by GTF2in: 5'-GGN CAY TAY YTN TGY AAY GC -3' and GTR1out. Gene-specific RACE primers were designed (MacVector Inc., Cary, NC, USA), synthesized (IDT, San Diego, CA, USA) and used (Advantage® 2 PCR Kit; Clontech). RACE fragments of 1389 bp (3') and 737 bp (5') were recovered for Cht-FoxA and fragments of 1038 bp (3') and ~ 1490 bp (5') were recovered for Cht-GATA456a.
Total RNA was collected from T. lageniformis oocytes and from assorted embryonic and larval stages from the first 3 days of development. 5' and 3' RACE-Ready cDNA templates were prepared as described for Chaetopterus. A degenerate DNA fragment of 226 bp from the Tl-FoxA winged-helix domain was isolated using the following semi-nested primer set: A1Fout and A2Rout, followed by A1Fout and A3Rin: 5'-CCA RAA NSW NCC YTT NCC NGG -3'. A 339 bp degenerate fragment of Tl-GATA456 was isolated with the primer set: GTF1out and GTR1out. Gene-specific RACE primers were designed and used as previously described. RACE fragments of 1569 bp (3') and 779 bp (5') were recovered for Tl-FoxA. Two distinct GATA456 genes were recovered from T. lageniformis using one set of 5' and 3' RACE primers designed from the same conserved, C4 zinc-finger domain: RACE fragments of 1485 bp (3') and 1102 bp (5') for Tl-GATA456a and 1068 bp (3') and 881 bp (5') for Tl-GATA456b were recovered. Additional gene-specific primers were then used to confirm the identity of each distinct Tl-GATA factor by amplification of sequences connecting each of the individual Tl-GATA456a and Tl-GATA456b RACE fragments. Degenerate and RACE gene fragments obtained from both species were subcloned into pGEM-T Easy vectors (Promega, Madison, WIm USA) and sequenced at Macrogen Inc. (Seoul, South Korea).
Gene orthology analyses
Sequence data for all recovered gene fragments from Chaetopterus and T. lageniformis were examined for similarity to public database protein sequences using the blastx program of NCBI http://www.ncbi.nlm.nih.gov. Sequence data for DNA-binding domains of the Fox winged-helix and GATA C4 zinc-finger proteins from a subset of additional organisms were obtained from the NCBI Protein database (Entrez Protein). All amino acid sequences were aligned using the ClustalW algorithm in MacVector (version 10.6), and manually adjusted for alignment errors.
Phylogenetic analyses of selected Fox winged-helix and GATA zinc-finger domains were performed with MrBayes (version 3.1.2) . The Fox protein analysis used the WAG model of amino acid replacement for 3,000,000 generations sampled every 100 generations, with four chains and four independent runs. A 50% majority-rule consensus tree was generated from the last 90,004 trees. The GATA protein analysis used the JTT model of amino acid replacement for 3,000,000 generations sampled every 100 generations, with four chains and four independent runs. A 50% majority-rule consensus tree was generated from the last 80,004 trees. Gamma shape parameters and the proportion of invariable sites were uniformly distributed for each analysis. The resulting consensus trees were viewed with FigTree (version 1.1.2; http://tree.bio.ed.ac.uk/) and edited in Adobe illustrator CS3.
Gene accession numbers
Nexus files of the Fox and GATA amino acid alignments are available upon request.
Amino acid sequences used in the alignments are available with the GenBank accession numbers from NCBI http://www.ncbi.nlm.nih.gov/ or the identification numbers of predicted proteins from JGI http://www.jgi.doe.gov/:
Fox transcription factors:
Pvfkh [GenBank:AJ507424]; LgFoxA [JGI protein ID:54814]; TlFoxA [GenBank:GU357822]; CtFoxA [GenBank:EF651787]; SpFoxA [GenBank:DQ459376]; PfHNF3 [GenBank:AB023019]; ChtFoxA [GenBank:GU357819]; Pdfkh [GenBank:AM114771]; HrFoxA [JGI protein ID:137725]; Dmfkh [GenBank:J03177]; Bmsgf1 [GenBank:NM_001043864];
Tcfkh [GenBank:NM_001039414]; Atfkh [GenBank:BAC24088]; DrFoxA1 [GenBank:AAH65668]; MmFoxA1 [GenBank:AAH96524]; DrFoxA2 [GenBank:BC086703]; MmFoxA2 [GenBank:NM_010446]; CsHNF3 [GenBank:AB04958]; BfHNF3-1 [GenBank:X96519]; Nvfkh [GenBank:AY457634]; LgFoxB [JGI protein ID:75592]; Pdest [GenBank:CT033411]; Spfkh1 [GenBank:NP_999797]; XlFoxB [GenBank:AJ487619]; BfFoxB [GenBank:AJ506162]; CtFoxB [JGI protein ID:225366]; NvfkhB [GenBank:DQ17368]; SpFoxABlk [GenBank:XM_001186527]; CtFoxAB [JGI protein ID:131123]; CtFoxC [JGI protein ID:199610]; SpFoxC [GenBank:XM_001183474]; Dmcroc [GenBank:NM_079478]; BfFoxC [GenBank:CAH69694]; SpFoxG [GenBank:DQ286739].
GATA transcription factors:
CeELT1 [GenBank:BAE06473]; CbELT1 [GenBank:XM_002633918]; DmGRAIN [GenBank:NP_731211]; TcGRAIN [GenBank:NM_001164788]; SpGATAc [GenBank:AAC62960]; CiGATAb [GenBank:BAE06473]; DrGATA1 [GenBank:NM_131234]; MmGATA1 [GenBank:NP_032115]; DrGATA2 [GenBank:NM_131233]; MmGATA2 [GenBank:BC107009]; DrGATA3 [GenBank:BC162401]; MmGATA3 [GenBank:NM_008091]; LgGATAc [JGI protein ID:160340]; ChtGATA123a [GenBank:GU357821]; PdGATA123 [GenBank:ABK32792]; CtGATAa1[GenBank:EF651791]; NvGATA [GenBank:AY465174]; CtGATAb2 [GenBank:EF651789]; PdGATA456 [GenBank:ABK32792]; SpGATAe [GenBank:NM_001005725]; ChtGATA456a [GenBank:GU357820]; TlGATA456b [GenBank:GU357824]; DmPANNIER [GenBank:NM_057337]; AgPANNIER [GenBank:AF395080]; AaGatad [GenBank:AY745809]; DmSERPENT [GenBank:NP_732098]; AmSERPENT [GenBank:XM_001121273]; AaGata4 [GenBank:EAT37835]; TcGATA [GenBank:XM_968040]; DrGATA4 [GenBank:NM_131236]; MmGATA4 [GenBank:AF179424]; DrGATA5 [GenBank:BC116537]; MmGATA5 [GenBank:AK142213]; DrGATA6 [GenBank:BC067710]; MmGATA6 [GenBank:NM010258]; LgGATAb [JGI protein ID:167550]; LgGATAa [JGI protein ID:129788]; CtGATAb3 [GenBank:EF651790]; CtGATAb1 [GenBank:EF651788]; TlGATA456a [GenBank:GU357823].
Whole mount in situ hybridization
Chaetopterus embryos were pretreated with a 1:1 mixture of 1.0 M sucrose and 0.25 M sodium citrate (Sigma, St. Louis, MO, USA) for 1 to 2 minutes, washed in FSW to terminate pretreatment, and fixed with 3.7% formaldehyde in FSW overnight at 4°C. Chaetopterus larvae were relaxed in a 1:1 mixture of 0.37 M MgCl2 in FSW for 10 to 15 minutes and fixed in 3.7% formaldehyde in FSW overnight at 4°C.
T. lageniformis stage 3 to 4 embryos were pretreated with a 1:1 mixture of 100 mM sucrose:0.25 M Na citrate for 2 to 3 minutes, washed in FSW and fixed in 4% paraformaldehyde in FSW. Stage 5 pre-larvae were anesthetized with 1:1 0.37 M MgCl2 in FSW for 30 to 40 minutes, followed by a 'slow-fix' with drop-wise (5 to 10 μl) additions of 4% paraformaldehyde every 10 minutes until no movement was detected, then fixation with 4% paraformaldehyde in FSW. Stage 6 to 8 pelagosphera larvae were anesthetized with a 1:1:2 mixture of 0.37 M MgCl2: 0.25% bupivicaine hydrochloride (BH; Sigma) in FSW for 1 hour, followed by fixation with 4% paraformaldehyde in FSW. All fixations were carried out overnight at 4°C.
Specimens of both species were washed out of fixative into phosphate-buffered saline (PBS), dehydrated stepwise into 100% methanol, and stored at -20°C. Before hybridization, specimens were rehydrated stepwise from methanol into PBS with 0.1% Tween-20 detergent (PTw). Whole mount in situ hybridization was performed according to published protocols [18, 67] with several modifications: proteinase-K digestion was terminated after 2 minutes (Chaetopterus) or 15 minutes (T. lageniformis); acetic anhydride treatments were 6.0 μl/ml and 12.0 μl/ml triethanolamine (Chaetopterus) or 4.0 μl/ml and 8.0 μl/ml triethanolamine (T. lageniformis); pre-hybridization consisted of 50% hybridization buffer for 10 minutes followed by 100% hybridization buffer overnight (Chaetopterus) or 100% hybridization buffer overnight (T. lageniformis). For both species, in situ hybridization was conducted at 65°C for 72 hours, with the following color development reaction: [4.4 μl nitroblue tetrazolium (NBT) 75 mg/ml: 3.3 μl 5-bromo-4-chloro-3-indolyl phosphate (BCIP) 50 mg/ml]/ml alkaline phosphatase (AP) buffer. A detailed in situ hybridization protocol is available upon request.
A 5' riboprobe of 737 bp was synthesized for Cht-FoxA; a 5' riboprobe of approximately 1490 bp was synthesized for Cht-GATA456a (each used at 0.5 ng/μl). A 3' riboprobe of 1569 bp was synthesized for Tl-FoxA (used at 1.0 ng/μl); and two 5' riboprobes of 1102 and 881 bp were synthesized for Tl-GATA456a and Tl-GATA456b, respectively (each used at 3.0 ng/μl). All of the antisense digoxigenin-labeled riboprobes (dig-11-UTP) were synthesized with a T7 MEGAscript kit (Ambion Inc, Austin, TX, USA) and detected by chromogenic staining (NBT/BCIP) of an anti-digoxigenin antibody-alkaline phosphatase conjugate (Roche, Indianapolis, IN, USA). Color reactions were terminated in PTw, followed by a graded dilution series of hybridization buffer/PTw washes. Specimens were equilibrated and stored in glycerol (80% glycerol, 10% 10X PBS, 10% diH20) and mounted on coated slides (Rainex®; Sopus Products, Houston, TX, USA). Microscopic analyses were performed on a compound light microscope (Axioskop 2; Carl Zeiss, Inc. Thornwood, NY, USA) with DIC optics. Digital micrographs of riboprobe-labeled and non-labeled, fixed animals were captured with a stem-mounted, 4.0 megapixel digital camera (Coolpix 4500; Nikon Inc., Melville, NY, USA).
Confocal imaging and analysis
T. lageniformis embryonic and larval stages were pretreated as previously described, fixed for 1.0 hour in 4% paraformaldehyde in FSW at RT, and washed in phosphate-buffered saline with 0.1% Triton X-100 (PBT). Embryos at stages 3 to 5 were blocked in PBS-T with 10% heat-inactivated goat serum (Sigma) for 1 hour at RT, followed by consecutive overnight treatments at 4°C with primary mouse anti-histone antibody (1:250 dilution; F152.C25.WJJ; Millipore, Billerica, MA, USA) and secondary goat anti-mouse rhodamine antibody (1:200 dilution; Molecular Probes, Eugene, OR, USA) in blocking buffer. Secondary antibody was co-incubated with 1:200 BODIPY FL-phallacidin (Molecular Probes). Labeled embryos at stages 3 to 4 were rinsed in PBS and mounted in SlowFade Gold (Molecular Probes). Pelagosphera larvae at stages 6 to 8 were treated with RNase A at 1.0 mg/ml PBS-T for 1 hr at 37°C, washed in PBS-T, and then labeled with propidium iodide (Sigma)] 5 μg/ml in PBS-T and 1:200 BODIPY FL-phallacidin for 2 hours at RT. Stage 5 pre-larvae and pelagosphera larvae were washed in PBS, dehydrated stepwise in isopropanol, cleared by emersion in 2:1 benzyl benzoate:benzyl alcohol, mounted on non-coated glass slides and sealed with clear nail polish. Visualization and imaging of all stages were performed using a confocal microscope (Axioplan 2 LSM510; Zeiss).
We thank Néva P. Meyer and Mark Q. Martindale for preparing fixed developmental stages of Chaetopterus at Marine Biological Laboratory, Woods Hole, MA and Mary E. Rice for helpful commentary on sipunculan terminology. We are grateful to Rich Chock for the maintenance of all facilities at Kewalo Marine Laboratory. This work was supported by grants from the National Science Foundation (NSF IOB05-44869 and NSF EF05-31558) to E.C.S.
- 8.Erwin DH: The developmental origins of animal body plans. Neoproterozoic Geobiology and Paleobiology; Dordrecht. Edited by: Xiao S, Kaufman AJ. 2006, Netherlands: Springer, Chapter XXI: 300-Google Scholar
- 15.Henry JJ: The role of unequal cleavage and the polar lobe in the segregation of developmental potential during first cleavage in the embryo of Chaetopterus variopedatus. Roux's Arch Dev Biol. 1986, 195: 10.1007/BF00456106.Google Scholar
- 19.Rouse GW, Pleijel F: Polychaetes. 2001, Oxford; New York: Oxford University PressGoogle Scholar
- 21.Pilger JF: Reproductive biology and development of Themiste lageniformis, a parthenogenic sipunculan. Bull Mar Sci. 1987, 41: 59-67.Google Scholar
- 22.Pilger JF: A cytophotometric study of parthenogenesis in Themiste lageniformis (Sipuncula). Bull Mar Sci. 1989, 45: 415-424.Google Scholar
- 23.Williams JA: Functional development in four species of the sipuncula. Doctoral Dissertation. 1977, University of Hawaii, Department of ZoologyGoogle Scholar
- 24.Pilger JF: Reproduction of a parthenogenetic sipunculan. Annual Meeting of the American Society of Zoologists. American Zoologist. 1978, 663-Google Scholar
- 33.Colgan DJ, Hutchings PA, Beacham E: Multi-gene analyses of the phylogenetic relationships among the Mollusca, Annelida, and Arthropoda. Zool Stud. 2008, 47: 338-351.Google Scholar
- 34.Dunn CW, Hejnol A, Matus DQ, Pang K, Browne WE, Smith SA, Seaver E, Rouse GW, Obst M, Edgecombe GD, Sørensen MV, Haddock SH, Schmidt-Rhaesa A, Okusu A, Kristensen RM, Wheeler WC, Martindale MQ, Giribet G: Broad phylogenomic sampling improves resolution of the animal tree of life. Nature. 2008, 452: 745-749. 10.1038/nature06614.CrossRefPubMedGoogle Scholar
- 37.Cazaux C: Developpement larvaire de Chaetopterus variopedatus (Renier). Actes Soc Linneénne de Bordeaux. 1965, 102: 1-16.Google Scholar
- 38.Bonch-Bruevich EV, Malakhov VV: Organization of early larvae of the polychaete worm Chaetopterus variopedatus (Spiomorpha, Chaetopteridae). Dokl Akad Nauk SSSR. 1987, 292: 107-109.Google Scholar
- 39.Malakhov VV: Embryogenesis of Chaetopterus variopedatus (Spiomorpha, Chaetopteridae). Zool Zh. 1984, 63: 656-661.Google Scholar
- 41.Williams JA: Development of a rock burrowing sipunculid inhabiting stony coral. Winter Meeting of the American Society of Zoologists. Am Zool. 1972, 723-Google Scholar
- 48.Davidson EH, Rast JP, Oliveri P, Ransick A, Calestani C, Yuh CH, Minokawa T, Amore G, Hinman V, Arenas-Mena C, Otim O, Brown CT, Livi CB, Lee PY, Revilla R, Rust AG, Pan Z, Schilstra MJ, Clarke PJ, Arnone MI, Rowen L, Cameron RA, McClay DR, Hood L, Bolouri H: A genomic regulatory network for development. Science. 2002, 295: 1669-1678. 10.1126/science.1069883.CrossRefPubMedGoogle Scholar
- 57.Cutler EB: The Sipuncula: their systematics, biology, and evolution. 1994, Ithaca: Comstock Pub. AssociatesGoogle Scholar
- 58.Michel C: Intestine and digestive glands. The Ultrastructure of Polychaeta. Edited by: Westheide W, Hermans CO. 1988, Stuttgart: Gustav Fischer Verlag, 4: 494-[Ax P (Series Editor) Microfauna Marina]Google Scholar
- 59.Rice ME: Sipuncula. Microscopic Anatomy of Invertebrates. Onychophora, Chilopoda, and Lesser Protostomata. Edited by: Harrison FW, Rice ME. 1993, New York: Wiley-Liss, 12: 274-296.Google Scholar
- 61.Costello DP, Henley C: Methods for Obtaining and Handling Marine Eggs and Embryos. 1971, City: Marine Biological Library; Wood Hole Oceanographic Institute, [https://darchive.mblwhoilibrary.org/handle/1912/295]CrossRefGoogle Scholar
- 75.Soudais C, Bielinska M, Heikinheimo M, MacArthur CA, Narita N, Saffitz JE, Simon MC, Leiden JM, Wilson DB: Targeted mutagenesis of the transcription factor GATA-4 gene in mouse embryonic stem cells disrupts visceral endoderm differentiation in vitro. Development. 1995, 121: 3877-3888.PubMedGoogle Scholar
- 82.NCBI GenBank. [http://www.ncbi.nlm.nih.gov/]
- 83.DOE Joint Genome Institute. [http://www.jgi.doe.gov/]
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.