Species-Specific Traits in the Genus Hydra

Hydra species differ in morphology, development, physiology, and ecology [33,34]. As shown in Figure 1B–1F, Hydra species differ in the presence or absence of symbionts, in carotenoid metabolism, and in the presence or absence of a conspicuous stalk structure (labeled with an asterisk in Figure 1C). Hydra species vary in genome sizes, ranging from 380 to 1,450 Mbp. Hydra species also differ in the shapes and sizes of their nematocysts. For example, as schematically shown in Figure 1F, in Hydra oligactis in contrast to other Hydra species the inverted tube of holotrichous isorhiza is coiled irregularly. Other species-specific traits include a ring of neurons running circumferentially around the hypostome in H. oligactis but not in other Hydra species [35]. Hydra species can also be identified by the order in which tentacles arise on young buds. Whereas the mean number of tentacles per polyp does not differ in H. oligactis and H. magnipapillata (Figure 1G and 1H), in H. oligactis two lateral tentacles arise early and are conspicuously longer during bud development (Figure 1I–1M) in contrast to H. magnipapillata and H. vulgaris, where tentacles arise synchronously and are all of the same length (Figure 1N–1R).

Identification of TRGs in H. oligactis and H. magnipapillata

To gain insight into the molecular mechanisms controlling phenotypic differences in the genus Hydra, we compared the transcriptomes of H. oligactis and H. magnipapillata by suppression subtractive hybridization (SSH) of complimentary DNAs (Figure 2A and 2B). The goal of SSH was to identify quickly evolving genes that are present (and transcriptionally active) in one species and absent (or transcriptionally silent) in the other, as well as to uncover homologous genes that are expressed in both species but have considerably diverged sequences. As shown schematically in Figure 2A and 2B, two subtractive hybridizations were performed using cDNA pools from H. oligactis and H. magnipapillata polyps, resulting in two cDNA libraries enriched for species-specific transcripts. As expected, a considerable proportion of clusters and singletons in each of the library had no homologs in the NCBI protein database (Figure 2A and 2B), representing, therefore, genes putatively restricted to the genus Hydra (for clustering and sequence analysis see Tables S1–S3 and Figure S1). The expression of these genes was analyzed by in situ hybridization. Genes expressed in taxonomically informative structures (tentacles, nematocytes, and stalk; see Figure S1) were selected for further analysis. Here, we focus on genes represented by clusters CL223 and CL87. Both genes show no similarity to any gene outside the genus Hydra. Both genes, however, are weakly similar (see Tables S1 and S2) to the Hym301 protein, which was identified earlier in a systematic approach to isolate morphogenetically active peptides in H. magnipapillata [36]. Interestingly, as shown in Figure 2C and 2D, in the two species analyzed these Hym301-like genes have complementary expression patterns: whereas in H. oligactis expression is restricted to tentacles (Figure 2C), in H. magnipapillata Hym301-like transcripts are absent in tentacle tissue but abundant throughout the body column (Figure 2D). Because tentacle development is one of the major morphogenetic processes in Hydra, differs between H. oligactis and H. magnipapillata (Figure 1M and 1R), and is under strong selective constraints as tentacles are the only structures by which polyps can take up food, we focused on the identified Hym301-like genes to investigate their function.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. Identification of Species-Specific Genes in Hydra by SSH

(A) Using H. magnipapillata cDNAs as driver and H. oligactis cDNAs as tester, we generated a library enriched for the transcripts of H. oligactis-specific genes. 1,022 cDNA clones were sequenced and clustered: 33% of the clusters and singletons had strong similarity to known proteins (dark grey segment, E value ≤ 1e–20), 26% had weak similarity (light grey segment, 1e–20 < E value < 1e–5); 41% had no similarity (white segment, E value ≥ 1e–5), representing, therefore, putative TRGs.

(B) Using H. oligactis cDNAs as driver and H. magnipapillata cDNAs as tester we generated a library enriched for the transcripts of H. magnipapillata-specific genes. 1,104 cDNA clones were sequenced and clustered: 30% of the clusters and singletons had strong similarity to known proteins (dark grey segment, E value ≤ 1e–20), 28% had weak similarity (light grey segment, 1e–20 < E value < 1e–5) and 42% had no similarity (white segment, E value ≥ 1e–5), representing, therefore, putative TRGs.

(C, D) Whole-mount in situ hybridization shows differences in the expression patterns of two homologous Hydra-specific genes in H. oligactis and H. magnipapillata. CL223 and CL87 have weak similarity to Hym301 protein [37] and exhibit complementary expression patterns that correlate with differences in the mode of tentacle formation between H. oligactis and H. magnipapillata.

https://doi.org/10.1371/journal.pbio.0060278.g002

Characterization of the TRG Family Hym301

Detailed analysis of the Hym301-like genes CL223 (in H. oligactis) and CL87 (in H. magnipapillata) revealed that they belong to a gene family that exhibits variation in gene number and expression patterns among Hydra species (Figure 3). In H. magnipapillata we discovered four CL223-related genes (see Materials and Methods for detail). Because one of them has been identified earlier [36] and was described as a gene Hym301 in H. magnipapillata [37], we named the four genes mHym301A to mHym301D. In H. magnipapillata, the genes mHym301A (Figure 3A), mHym301C (Figure 3C), and mHym301D (Figure 3D) are expressed in the head, whereas mHym301B (Figure 3B) is expressed along the body column. None of the four H. magnipapillata genes is expressed in the tentacles. In contrast, H. oligactis has only two Hym301-related genes, oHym301A and oHym301B, with both genes expressed in tentacles (Figure 3E and 3F). In H. vulgaris AEP, only one gene is present (aepHym301A) and expressed in the basis of tentacles (Figure 3G). Expression of aepHym301A is considerably weaker than expression of the Hym301-like genes in the other two species (Figure 3I).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. Analysis of the Taxonomically Restricted Gene Family Hym301 in Different Hydra Species

(A–D) Expression of the four mHym301 genes in H. magnipapillata. mHym301B and oHym301A identified in the SSH screening are boxed in yellow.

(E–F) Expression of the two oHym301 genes in H. oligactis.

(G) Expression of the single aepHym301 gene in H. vulgaris AEP.

(H) Exon–intron structure of the four mHym301 genes in H. magnipapillata. Note that mHym301C and mHym301D are clustered in the genome.

(I) RT-PCR confirms the expression data in (G) and shows a much lower level of Hym301 transcripts in H. vulgaris AEP than in H. oligactis and H. magnipapillata.

(J) Phylogenetic tree showing the divergence of the Hym301 proteins in the three Hydra species. H. magnipapillata, red; H. vulgaris AEP, black; H. oligactis, blue. mHym301A is identical to the previously described Hym301 protein [37] and is labeled with an asterisk.

(K) Alignment of the Hym301 amino acid sequences from three species of Hydra. The position of the Hym301 peptide described in [37] is underlined and in orange letters. The putative amidation signal (GKK) present in mHym301A, oHym301A, and oHym301B is underlined in green. SP, signal peptide.

https://doi.org/10.1371/journal.pbio.0060278.g003

A phylogenetic tree based on the amino acid sequence alignment (Figure 3J) places the members of Hym301 gene family in three distinct groups, which fits with the phylogenetic tree of the genus Hydra [34] (see Figure 1A). Predicted proteins from H. oligactis and H. magnipapillata form two distinct clades, whereas the single protein from H. vulgaris AEP forms a separate clade (Figure 3J). The phylogeny clearly indicates that gene duplications occurred separately in the H. oligactis and H. magnipapillata lineages. In H. magnipapillata the available draft genome revealed that all four genes share common genomic organization, with four exons (Figure 3H). Interestingly, mHym301C and mHym301D are clustered in the genome within 70 kb (Figure 3H), suggesting a recent duplication event. Amino acid alignment (Figure 3K) indicates that all members of the gene family encode short proteins (87–106 amino acids) with a putative signal peptide sequence at the N terminus followed by a relatively conserved region. The C-terminal part of the protein is highly diverged, not only among species, but also among the different Hym301 proteins in H. oligactis and H. magnipapillata. In mHym301A the C terminus harbors the peptide (marked in orange in Figure 3K) reported previously [37] to affect tentacle formation in H. magnipapillata. According to the sequence information available, the genes of the Hym301 family have no homologs in eukaryotic organisms outside the genus Hydra. They are also not present in the genome of Nematostella, which, as a representative of the class Anthozoa, occupies a more basal position in the phylogenetic tree than Hydra. Therefore, we consider the members of the Hym301 family as TRGs, which have most likely originated within the class Hydrozoa and expanded in the genus Hydra.

Generation of Transgenic Hydra Overexpressing mHym301A

To explore the biological significance of the Hym301 gene family in vivo, we produced transgenic polyps that overexpress mHym301A in all ectodermal epithelial cells. As shown in Figure 4, we introduced the H. magnipapillata mHym301A gene into polyps of the closely related species H. vulgaris AEP, in which the expression level of the endogenous gene (aepHym301A) is low (see Figure 3G and 3I). Hydra embryos were microinjected (see Materials and Methods for details) with an expression construct in which mHym301A including the sequence encoding the signal peptide was fused in frame to enhanced green fluorescent protein (eGFP) and under the control of the Hydra β-actin promoter (Figure 4A). In transgenic animals producing mHym301A:eGFP, secretion of the fusion protein was detectable during embryonic development in the egg shell of the embryo starting 3 d after injection (Figure 4B and 4C). After hatching, fully transgenic animals producing mHym301A:eGFP in all their ectodermal epithelial cells were generated by asexual propagation of founder polyps (Figure 4D–4F). Figure 4G indicates that, in nontransgenic H. magnipapillata polyps, mHym301A transcripts are restricted to the head tissue and absent in tentacle cells. Transgenic line H. vulgaris AEP^A14 was used in all further experiments. As a control, a transgenic line was produced (Ecto-1) using the same construct without the sequences encoding signal peptide and the mHym301A sequence. As shown in Figure 4I, the mHym301A:eGFP fusion protein is visually detectable by fluorescence microscopy in transgenic animals along the whole body column and in the tentacles. Fluorescence, however, is much weaker than in the control Ecto-1 line (Figure 4H). Confocal laser scanning microscopy shows (Fig 4J–4L) the presence of the mHym301A:eGFP fusion protein in secretory vesicles of ectodermal epithelial cells. In contrast, the eGFP protein in control Ecto-1 animals is not secreted and resides in the cytoplasm of epithelial cells (Figure S2). Thus, by a gain-of-function approach we generated H. vulgaris AEP transgenic animals that strongly overexpress the H. magnipapillata mHym301A gene in all ectodermal epithelial cells, including the tentacle cells. In these transgenic H. vulgaris AEP animals, the level of Hym301 expression is not only greatly enhanced compared with wild-type H. vulgaris AEP, but expression also extends to the tentacles—thereby mimicking oHym301A transcript localization in H. oligactis.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 4. Transgenic H. vulgaris AEP Overexpressing mHym301A

(A) Expression constructs for generation of transgenic Hydra. Top: control construct with eGFP driven by 1,386 bp actin 5′ flanking region. Below: construct with mHym301A fused to eGFP driven by the actin 5′ flanking region. SP, signal peptide.

(B) Hydra embryo during the first cleavage stage used for microinjection of expression construct.

(C) mHym310A:eGFP expression in Hydra embryos at cuticle stage (fluorescence microscopy). Arrows indicate accumulation of mHym310A:eGFP protein at the outer membrane, indicating secretion of the fusion protein.

(D–F) RNA in situ hybridization reveals that mHym301A expression is activated in ectodermal epithelial cells all over the body column and includes tentacle epithelial cells. Founder polyps (D) are mosaics and were used to generate a mass culture of polyps such as that shown in (F) expressing mHym301A in most of the epithelial cells.

(G) mHym301A expression in wild-type H. magnipapillata polyps is restricted to the head region and tip of the emerging bud.

(H) Control transgenic polyp with eGFP expression under control of the β-actin promoter.

(I) Transgenic polyp in which most of the ectodermal epithelial cells are expressing mHym301A:eGFP. Note that the eGFP signal is localized in ectoderm.

(J) Confocal analysis of a representative transgenic polyp showing mHym301A:eGFP protein localization in secretory vesicles of ectodermal (ecto) epithelial cells (arrows). Green, eGFP protein; blue, DAPI stained nucleus; red, actin filaments. Endo, endodermal epithelium.

(K) mHym301A:eGFP protein localizes to secretory vesicles in ectodermal epithelial cells.

(L) Top view of ectodermal epithelial cell containing numerous eGFP-positive vesicles (see also Figure S2).

https://doi.org/10.1371/journal.pbio.0060278.g004

Do these perturbations in Hym301 expression affect morphological traits of transgenic H. vulgaris AEP^A14 polyps? To address this issue, we compared tentacle development in transgenic animals with that in wild-type H. vulgaris AEP, transgenic H. vulgaris AEP line Ecto-1, and wild-type H. magnipapillata and H. oligactis.

mHym301A Overexpression Affects Both the Speed of Tentacle Regeneration and the Pattern in Which Tentacles Arise

When a hydra's hypostome and tentacles are removed, the animal promptly develops new ones. “Head regeneration” is a morphallactic process involving the reorganization of existing tissue; it follows species-specific rules, with tentacles arising synchronously in H. magnipapillata and H. vulgaris but asynchronously in H. oligactis. Intact transgenic H. vulgaris AEP^A14 animals exhibit normal morphology without any obvious disturbances (Figure 4I). The number of tentacles seems equal in nontransgenic H. vulgaris AEP^wt, transgenic H. vulgaris AEP^A14, and control transgenic line Ecto-1 animals (Figure 5A; see also Materials and Methods). Differences, however, become obvious when a polyp is bisected in the body column (Figure 5B) and a head regenerates at the apical end of the lower half within 96 h. Figure 5C indicates that in H. vulgaris AEP^A14, tentacles begin to appear much earlier and in larger numbers than in nontransgenic H. vulgaris AEP^wt or control transgenic line Ecto-1 (Table S4). Forty-two h after bisection, transgenic H. vulgaris AEP^A14 had 4.2 ± 1.5 or 3.6 ± 1.61 tentacles compared with only 1.71 ± 1.24 in nontransgenic H. vulgaris AEP^wt or 1.6 ± 1.64 in the transgenic Ecto-1 strain (Figure 5C). Figure 5D shows a representative transgenic animal, with well developed tentacles 42 h after decapitation; in the nontransgenic control (Figure 5I), tentacle formation is just at its beginning. By 66 h after decapitation, the mean number of tentacles in transgenic H. vulgaris AEP^A14 was observed (Figure 5C) to vary from 5.6 ± 1.52 to 5.8 ± 1.44 compared with 3.68 ±1.51 in nontransgenic H. vulgaris AEP^wt and 3.0 ± 1.64 in the control Ecto-1 line. The difference in tentacle number was maintained until 130 h after bisection (7.2 ± 1.3 in H. vulgaris AEP^A14 versus 6.1 ± 1.05 and 5.7 ± 1.31 tentacles, respectively, in nontransgenic and Ecto-1 line controls).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 5. Overexpression of mHym301A Alters Timing and Order of Tentacle Development during Regeneration

(A) Number of tentacles per polyp in H. vulgaris AEP^wt, transgenic H. vulgaris AEP^A14, and transgenic H. vulgaris AEP Ecto-1 lines. In all three lines tentacle number was counted in 100 adult polyps with buds.

(B) Experimental scheme for the regeneration experiments. Polyps were cut at 1/3 of the body length.

(C) Time kinetics of tentacle formation in transgenic H. vulgaris AEP^A14 compared with that in wild-type H. vulgaris AEP and transgenic H. vulgaris AEP Ecto-1 animals. Mean number of tentacles per polyp in AEP^A14, AEP^wt and Ecto-1 at 42, 66, and 130 h after decapitation (two asterisks, p < 0.01 according to ANOVA test).

(D–F) In transgenic H. vulgaris AEP^A14 newly formed tentacles arise asynchronously.

(G) Scheme summarizing the order in which tentacles arise in transgenic H. vulgaris AEP^A14.

(H) Scheme depicting the ectopic expression pattern of mHym301A in transgenic H. vulgaris AEP^A14 (see Figure 4F).

(I–K) In control H. vulgaris AEP newly formed tentacles arise synchronously and considerably later than in transgenic animals.

(L) Scheme summarizing the order in which tentacles arise in wild-type H. vulgaris AEP^wt.

(M) Scheme depicting the expression pattern of aepHym301A in H. vulgaris AEP (see Figure 3G). Note the absence of transcripts in tentacles.

(N–P) In H. oligactis newly formed tentacles arise asynchronously.

(Q) Scheme summarizing the order in which tentacles arise in H. oligactis.

(R) Scheme depicting the expression pattern of oHym301A in H. oligactis. Note the presence of the transcript in tentacles (see Figure 3E).

(S–U) In H. magnipapillata newly formed tentacles arise synchronously.

(V) Scheme summarizing the order in which tentacles arise in H. magnipapillata.

(W) Scheme depicting the expression pattern of mHym301A in H. magnipapillata. Note the absence of transcripts in tentacles (see Figure 3A).

https://doi.org/10.1371/journal.pbio.0060278.g005

When examining regeneration in nontransgenic H. vulgaris AEP^wt and transgenic H. vulgaris AEP^A14, we noticed differences not only in the time kinetics of tentacle development, but also in the order in which tentacles arise on the regenerating tip (Figure 5D–5M). Unexpectedly, in transgenic H. vulgaris AEP^A14 (Figure 5D–5F), in contrast to nontransgenic H. vulgaris AEP^wt (Figure 5I–5K), tentacles arise asynchronously with two tentacles developing first. This mode of tentacle formation is similar to the order in which tentacles arise in regenerating H. oligactis polyps (Figure 5N–5P) and different from the mode of tentacle formation in the closely related species H. magnipapillata (Figure 5S–5U). Thus, overexpression of mHym301A in H. vulgaris AEP body column and tentacle cells (Figure 4I) causes reorganization of the order in which tentacles appear, from the H. vulgaris AEP pattern towards the H. oligactis-specific pattern. Given that in transgenic H. vulgaris AEP^A14 as in H. oligactis polyps, mHym301A protein is present in tentacle cells (Figures 4I, 5H and 5R), these observations suggest that Hym301 genes are involved in determining the symmetry of tentacle formation.

To lend further credibility to this hypothesis, we next examined tentacle development during budding, a developmental process by which de novo head formation occurs in an adult hydra (Figure 6). As shown in Figure 6A–6D, in nontransgenic H. vulgaris AEP^wt the pattern of tentacle formation is similar to that observed in H. magnipapillata (see Figure 1N–1Q) and different from that in H. oligactis (see Figure 1I–1L). In H. vulgaris AEP^wt, five tentacle buds appear almost synchronously at the earliest stages of bud formation (arrowheads in Figure 6A) in a regular pentagonal pattern. In contrast, in transgenic H. vulgaris AEP^A14 overexpressing mHym301A (Figure 6F–6I), the tentacles arise in a significantly larger number than in nontransgenic controls and in a slightly staggered sequence. As depicted in Figure 6G and 6H, transgenic H. vulgaris AEP^A14 develop six to seven tentacle buds in a rather irregular pattern, with some of the tentacles arising even outside the head in the upper part of the body column (open arrowhead in Figure 6H). Given that in adult polyps the number of tentacles is similar (varying from five to eight; Figure 5A) in nontransgenic and transgenic H. vulgaris AEP, the full set of tentacles appears to be established early in transgenic polyps and late (after detachment from the mother polyp) in nontransgenic controls. Taken together, these results substantiate the view that mHym301A is a molecule that affects the timing and order in which tentacles arise in Hydra.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 6. Overexpression of mHym301A Alters Tentacle Development during Budding and Mimics a Mutant Phenotype of H. magnipapillata

(A–D) Representative scanning electron micrographs of bud evagination in control H. vulgaris AEP^wt. Yellow arrowheads indicate emerging tentacles.

(E) Scheme of tentacle formation during budding in H. vulgaris AEP^wt.

(F–I) Representative scanning electron micrographs of bud evagination in transgenic H. vulgaris AEP^A14 overexpressing mHym301A. Note the slightly staggered order in which tentacles arise (marked with red arrowheads).

(J) Scheme of tentacle formation during budding in H. vulgaris AEP^A14.

(K–N) Scheme of tentacle formation during budding in H. magnipapillata mutant E6. Note that more tentacles arise.

(O) Scheme of tentacle formation during budding in H. magnipapillata mutant E6.

(P) Number of tentacles per polyp in wild-type H. magnipapillata 105 (n = 100) and mutant H. magnipapillata E6 (n = 100).

(Q) Expression of mHym301A in mutant E6 and wild-type H. magnipapillata. Note the particularly high level of expression in mutant E6 buds compared with wild type (marked with red brackets).

(R) Expression of mHym301B in mutant E6 and wild-type H. magnipapillata.

(S) RT-PCR analysis indicates higher level of expression of transcription factor HyAlx in transgenic H. vulgaris AEP^A14 overexpressing mHym301A than in control H. vulgaris AEP^wt. H, head region; G, gastric region; –C, negative control.

(T) Whole-mount in situ hybridization confirms the expression data in (S) and shows higher level of HyAlx transcripts in buds of transgenic H. vulgaris AEP^A14 than in control buds.

(U) Close up of HyAlx expression in buds of transgenic H. vulgaris AEP^A14 and in control H. vulgaris AEP^wt buds. Note that, consistent with data shown in (F–I), tentacles in transgenic buds arise asynchronously in contrast to nontransgenic controls (marked with red arrowheads). Fixation conditions and time of development during in situ procedure were identical for mutant E6 and wild-type H. magnipapillata (in Q and R) as well as for AEP^A14 and AEP^wt (in T and U).

https://doi.org/10.1371/journal.pbio.0060278.g006

Hym301 Expression in a Mutant with Altered Number of Tentacles

To further explore the role of Hym301 in tentacle development, we studied a H. magnipapillata mutant, whose number of tentacles is significantly increased [38]. As shown in Figure 6P, adult polyps of H. magnipapillata line E6 have on average seven to eight tentacles, whereas polyps of the H. magnipapillata wild-type line 105 have five to six tentacles. Analysis of tentacle formation during early stages of budding showed that in the mutant H. magnipapillata line E6 tentacles arise earlier and in significantly larger numbers (Figure 6K–6N) than in H. magnipapillata wild-type line 105 (Figure 1N–1Q). In E6 mutant six to seven tentacles appear simultaneously at very early stages of bud development (Figure 6L and 6M), whereas in wild-type H. magnipapillata only four to five tentacles are present at this time point (Figure 1P).

Tentacle development in buds of H. magnipapillata mutant line E6 (Figure 6K–6N) strikingly parallels the pattern of tentacle development in transgenic H. vulgaris AEP^A14 (Figure 6F–6I). Is Hym301 responsible for the alterations observed in H. magnipapillata mutant line E6? To examine this possibility we investigated the expression of mHym301A and mHym301B in H. magnipapillata mutant line E6 by in situ hybridization (Figure 6Q and 6R). Given the results described above, we anticipated an increased level of mHym301A expression in the mutant line. As predicted, in situ hybridization showed that the expression domain of mHym301A (Figure 6Q) in the head is considerably enlarged in the mutant and extends down into the body column. In evaginating buds, the differences in Hym301 expression between wild-type and mutant polyps are even more striking. Whereas in buds of the same developmental stage of wild-type H. magnipapillata line 105 mHym301A expression is restricted to the distal part of the evagination, buds in H. magnipapillata mutant line E6 express mHym301A throughout the whole cylindrical bud protrusion (Figure 6Q). Interestingly (Figure 6R), the increased expression domain of mHym301A is accompanied by reduced expression of mHym301B. Taken together, these studies indicate that altered spatiotemporal control of Hym301 expression strongly correlates with altered tentacle development in H. magnipapillata mutant E6, and suggest that the taxonomically restricted gene family Hym301 is essential in head morphogenesis as it seems to ensure the development of tentacles in a correct number and species-specific order.

How are such novel genes incorporated into conserved signaling pathways? To address this question we examined whether mHym301A is interacting with conserved regulatory components such as transcription factor Aristaless (HyAlx). In H. magnipapillata, HyAlx is expressed exclusively in the tentacle zone, the lower part of the head from which tentacles emerge. As shown previously by expression and RNA interference (RNAi) analysis, HyAlx is directly involved in the specification of tissue for tentacle formation [39]. Figure 6 shows that in nontransgenic H. vulgaris AEP^wt (Figure 6S–6U), the expression of HyAlx is similar to that described previously [39] in H. magnipapillata. However, in contrast to nontransgenic H. vulgaris AEP^wt, in transgenic H. vulgaris AEP^A14, the expression of HyAlx is drastically increased (Figure 6S–6U). Reverse transcription PCR (RT-PCR) suggests (Figure 6S) that in polyps overexpressing mHym301A, HyAlx transcripts are not only present in higher amounts in the head region compared with wild type, but are detectable even in the gastric region of AEP^A14 animals. Given that overexpression of mHym301A causes an increased expression of HyAlx, interaction between novel gene mHym301A and a conserved transcription factor seems likely. Interestingly, our expression analysis places Hym301 upstream of HyAlx in the signaling cascade. Previously, Smith et al. have shown [39] that depletion of HyAlx RNA drastically delays tentacle appearance in buds, indicating that expression of HyAlx is required for this process. We conclude that in transgenic H. vulgaris AEP^A14 the mHym301A-induced increase in the level of HyAlx is responsible for the alteration observed in tentacle development compared with controls.

Silencing of oHym301A in H. oligactis Interferes with Tentacle Formation

We have observed drastic disturbances in head morphogenesis, leading to an increased speed of tentacle formation and altered pattern of tentacle distribution in animals overexpressing mHym301A (Figures 5C–5M and 6A–6J). We next investigated the role of Hym301 in tentacle formation by silencing oHym301A expression in H. oligactis by RNAi.

To optimize the RNAi procedure, we first performed control experiments in which GFP expression was knocked down in H. vulgaris AEP Ecto-1 animals by electroporation of GFP double-stranded RNA (dsRNA). As shown in Figure 7A, introduction of GFP dsRNA causes complete disappearance of GFP protein in large patches of ectodermal epithelial cells covering 10–70% of the polyp's surface. Ecto-1 polyps electroporated with oHym301A dsRNA showed no depletion of GFP signal (Figure 7B).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 7. Silencing of oHym301A in H. oligactis Interferes with Tentacle Formation

(A) Depletion of GFP protein in H. vulgaris AEP Ecto-1 line 3, 7, and 14 d after electroporation with GFP dsRNA. Note dark areas of ectodermal epithelium where GFP transcript and protein are completely depleted.

(B) Electroporation of Ecto-1 animals with oHym301A dsRNA has no effect on GFP expression.

(C) RT-PCR shows depletion of oHym301A transcript by RNAi in H. oligactis polyps at various time-points: 0 d, before electroporation and 4–10 d after electroporation; glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used for equilibration.

(D) Disturbed tentacle formation in H. oligactis after electroporation with oHym301A dsRNA. In situ hybridization 7 d after electroporation indicates that abnormalities in tentacle length are strongly correlated with the absence of oHym301A transcripts. Percentage of animals with normal and aberrant tentacle morphology is shown. Absolute number of experimental animals used is shown in brackets. Three independent RNAi experiments were performed.

(E) Time kinetics of tentacle formation in H. oligactis polyps 7 d after electroporation with oHym301A or GFP dsRNA. Mean number of tentacles per polyp at 42, 66 and 130 h after decapitation (two asterisks, p < 0.01 according to ANOVA test).

https://doi.org/10.1371/journal.pbio.0060278.g007

Next we examined the effect of oHym301A knock-down on the tentacle formation and morphology of H. oligactis. Polyps were electroporated with oHym301A dsRNA. Control polyps were electroporated with GFP dsRNA. The depletion of oHym301A was monitored by RT-PCR at various time points and by in situ hybridization 7 d after electroporation. As shown in Figure 7C, by introducing oHym301A dsRNA the expression of oHym301A was depleted by about 50% (7 d and 10 d in Figure 7C). Electroporation with GFP dsRNA had no effect on the expression of oHym301A (Figure 7C). The decrease of Hym301 transcript level 4 d after electroporation in GFP control is due to the loss of tentacles during the electroporation procedure (see Materials and Methods). After electroporation tentacles regenerate within 5 d and in control polyps the normal symmetrical pattern is reestablished.

Intriguingly, in H. oligactis polyps electroporated with oHym301A dsRNA, tentacle formation was found to be greatly disturbed. As shown in Figure 7D, 42.5% of polyps electroporated with oHym301A dsRNA developed one or two aberrant tentacles, which were considerably shorter than the others (labeled with arrowheads in Figure 7D). In control animals only 13.6% of polyps displayed abnormalities in tentacle morphology. Abnormal length of tentacles strongly correlates with the absence of oHym301A transcript as detected by in situ hybridization (Figure 7D). On the basis of the observation shown in Figure 7A, which indicates that RNAi depletion is restricted to patches of cells, we assume that shorter tentacles originate from areas of epithelial cells where oHym301A expression was effectively silenced. Taken together, the RNAi experiments indicate that in the absence of oHym301A mRNA tentacle development in Hydra is retarded.

Head regeneration experiments (Figure 7E) also show that H. oligactis polyps in which the oHym301A was silenced by RNAi develop new tentacles considerably more slowly than control polyps that were electroporated with GFP dsRNA (see Table S4). Differences in regeneration kinetics between H. oligactis knock-down and control animals are not as drastic as between animals of H. vulgaris AEP¹⁴ and Ecto-1 lines (Figure 5C). We assume that the mosaic distribution of cells with complete knock-down of oHym301A is the reason for this, as tentacles develop more slowly only in the areas in which oHym301A transcript was depleted.

In summary, oHym301A expression in tentacles is required for correct development in H. oligactis. Knock-down of oHym301A gene does not fully abolish tentacle formation, but slows it down considerably (Figure 7D and 7E). Hence, a “novel” gene belonging to the Hym301 gene family is causally involved in the process of tentacle formation in Hydra.

Animals and culture conditions.

Experiments were carried out with H. oligactis, H. vulgaris strain AEP, H. magnipapillata strain 105, and H. magnipapillata strain E6. Transgenic animals were generated using H. vulgaris strain AEP [44]. Animals where cultured according to standard procedures at 18 °C.

Molecular techniques.

Nucleic acid isolation, cDNA synthesis, RT-PCR, cloning, and sequencing were done following standard protocols. RT-PCR was performed using the following primer sets: oHYM301A_F(40) and oHYM301A_R(144), AEP301A_F and AEP301A_R, gmHYMA_E1F and gmHYMA_E3R, Alx_F(422) and Alx_R(631) (HyAlx), GAPDH_F and GAPDG_R (GAPDH), ACTIN35 and ACTIN34 (β-actin). Sequencing was done with Li-COR 4200 and Li-COR 4300 sequencers and manually verified using e-Seq V2.0 and e-Seq V3.0 software. Primer sequences are shown in Table S5.

SSH and cDNA libraries.

For SSH, double-stranded cDNA was synthesized using 2 μg mRNA from H. oligactis and H. magnipapillata strain 105 polyps. SSH was performed in both directions using PCR-Select cDNA Subtraction Kit (Clontech) according to the manufacturer's protocol. Two subtractive libraries were generated: Kiel 6, enriched for H. oligactis specific transcripts and Kiel 7, enriched for H. magnipapillata specific transcripts (see Figure 2). cDNAs were cloned into pGEM-T vector (Promega) and transformed into DH5α Escherichia coli cells. Bacterial clones were picked into 384-well plates using Q-Pix robot and plasmid inserts were sequenced at the Washington University Genome Sequencing Centre (St. Louis, Missouri, United States). Raw sequences were submitted to NCBI (http://www.ncbi.nlm.nih.gov/) under accession numbers CV284311–CV284646 and CV2845050–CV285735 (Kiel 6) and CV284048–CV284310, CV284647–CV285049, and CV285736–CV286173 (Kiel 7).

Isolation of Hym301-like genes from H. magnipapillata, H. oligactis, and H. vulgaris AEP.

To identify all Hym301-like genes in H. magnipapillata, we searched Hydra ESTs at http://blast.ncbi.nlm.nih.gov/Blast.cgi (about 170,000 sequences) and single whole-genome shotgun reads deposited at CompaGen server in Kiel (http://www.compagen.org) using translated BLAST and the mHym301A and mHym301B protein sequences as queries. To isolate Hym301-like genes in H. oligactis we screened macroarrays containing 30,000 cDNA clones with the mixture of mHym301A, mHym301B and oHym301A probes using low stringency washing conditions. Out of eight positive clones, four were identical to the oHym301A cDNA already known from the SSH library, two represented transcripts of a new member of a gene family designated oHym301B, and the other two clones contained sequences not related to Hym301. To isolate Hym301-like genes in H. vulgaris AEP, we performed 3′ RACE PCR according to the previously published method using primer oHym301_F(32) (Table S5) directed against sequence highly conserved in both H. oligactis and H. magnipapillata. In three independent RACE experiments only one type of Hym301-related cDNA was identified, designated later as aepHym301A. Sequences of Hym301 genes were submitted to NCBI under accession numbers EU787492–EU787498.

Gene expression analysis.

For assessment of gene expression, whole-mount in situ hybridization was carried out as described previously [45].

Generation of transgenic H. vulgaris AEP expressing mHym301A:eGFP.

Transgenic founder polyps overexpressing the mHym301A:eGFP fusion protein under control of the β-actin promoter (construct ligE) were produced at the University of Kiel Transgenic Hydra Facility (http://www.transgenic-hydra.org/). Briefly, a 291 bp fragment of mHym301A coding for the full-length protein including the signal peptide was amplified from H. magnipapillata strain 105 cDNA using Platinum High Fidelity Taq polymerase (Invitrogen) and primers Hym301_F(2)Pst and Hym301_R(295)Pst. The cDNA was cloned into the modification of HoTG expression vector using the PstI cutting site (see Figure 4A). The resulting transfection construct was sequenced, plasmid DNA was purified using Qiagen MidiPrep Kit and injected into H. vulgaris AEP embryos as described previously [44]. Embryos began to express the reporter gene 2–3 d after injection. Founder transgenic animals bearing the mHym301A:eGFP construct started to hatch 14 d after microinjection. Out of 24 embryos injected with the LigE construct, two transgenic lines were generated. One of them (line A7) showed stable integration of mHym301A:eGFP in ectodermal and endodermal epithelial cell lineages. Another (line A14) showed integration of the construct in ectodermal epithelial cells only. Initial founder transgenic animals of A14 line were further expanded into a mass culture by clonal propagation by budding and used in all further experiments.

Sequence analysis.

TIGR Indices Clustering Tools [46] were used for clustering the sequences of Kiel 6 and Kiel 7 libraries. Nucleotide and translated BLAST engines at the NCBI server (http://blast.ncbi.nlm.nih.gov/Blast.cgi) were used for homology searches in public databases. The Seqtools program (S.W. Rasmussen, http://www.seqtools.dk) was used for sequence analysis and batch BLASTX and BLASTN searches. DNA and protein sequences were aligned with DnaMan Version 4.12. Phylogenetic tree analysis was performed by the neighbor-joining method with standard settings and a bootstrap value of 1,000 using the Mega 3.1 software. Local assemblies of H. magnipapillata genome were performed manually using single whole genome shotgun reads deposited at NCBI trace archive (http://www.ncbi.nlm.nih.gov/Traces/) and the Compagen server in Kiel (http://www.compagen.org).

Microscopic analysis.

Fluorescent images were taken on a Zeiss Axioscope fluorescence microscope with an Axiocam (Zeiss) digital camera. Confocal laser microscopy was done using a LEICA TCS SP1 CLS microscope. A Zeiss S420 microscope was used for scanning electron microscopy.

Gene silencing by RNAi in Hydra.

Double-stranded RNAs for GFP and oHym301A were synthesized using a MEGAscript RNAi Kit (Ambion). DNA templates for RNA synthesis were generated by PCR using primers T7_GFP_F, T7_GFP_R, o301Ai_T7_F, o301Ai_T7_R (see Table S5). DsRNA was introduced into Hydra polyps by electroporation as described previously [47] with minor modifications. For each experiment 30–60 polyps were placed into chilled electroporation cuvettes with a 4 mm gap (Peqlab). Polyps were washed twice with 1 ml sterile ice-cold Millipore water. Electroporation was carried out in 200 μl water containing 15 μg dsRNA. DsRNA was added to polyps just before electroporation and cuvettes were briefly shaken to insure regular distribution of animals and dsRNA. Bio-Rad Gene Pulser (Bio-Rad) was adjusted to an electric field strength of 0.5 kV/cm and 50-μF capacitance. Polyps were electroporated two times successively with each pulse lasting for about 10–12 ms. Immediately after the second pulse polyps were transferred into 10 ml ice-cold hydra medium that was supplemented with 20% hyperosmotic dissociation medium (DM) containing 6 mM CaCl₂, 1.2 mM MgSO₄, 3.6 mM KCl, 12.5 mM N-tris-[hydroxymethyl] methyl-2 aminoethanesulfonic acid, 6 mM sodium pyruvate, 6 mM sodium citrate, 6.0 mM glucose, and 50 mg/ml rifampicin, pH 6.9. Electroporation causes cell loss and tissue damage with all the animals losing their tentacles (note the absence of tentacles in Figure 7A and 7B, 3 d after electroporation). To facilitate recovery, polyps were kept at 10 °C for up to 3 d. Hydra medium supplemented with 20% DM was exchanged every 12 h; viable polyps were separated from cell debris and transferred into new Petri dishes. Thirty-six h after electroporation the medium was exchanged for standard hydra medium. Five d after electroporation polyps were fully recovered and used for experimentation. Survival rate for the electroporated polyps varied between 40% and 60%.

Ecto-1 transgenic line which expresses GFP in all ectodermal epithelial cells was used to directly visualize RNAi in vivo. Electroporation of Ecto-1 animals with GFP dsRNA allowed for the first time to observe the disappearance of the RNAi target protein in living Hydra polyps. Because of the high stability of the GFP protein, the first areas of completely GFP-negative epithelial cells become clearly visible 5 d after electroporation. Complete down-regulation of the target gene was observed in areas of cells that, in the best case, covered about 70% of the polyp surface. The RNAi effect is stable for the period of over 14 d (Figure 7A).