Database search and bioinformatic analysis identifies 15 Hydra FGFs

The core domains of five previously identified Hydra vulgaris AEP FGFs (Hv-FGF-a, Hv- FGF-b, Hv-FGF-c, Hv-FGF-e, and Hv-FGF-f) (Lange et al., 2014) were used for protein BLAST and identified ten additional potential Hydra FGFs in the EST databases. Their sequence IDs (NCBI and Hydra AEP Genome Project Portal sequence) are listed in (supplemental Table S1A). Analysis of the Hydra FGFs properties using Expasy ProtParam (https://web.expasy.org/protparam) characterized all of them as small hydrophilic and highly basic proteins with molecular weights between 18,6 and 32,7 kDa and an estimated isoelectric point (pI) between 8.1 and 10.0 (supplemental Table S1B). Presence of the characteristic FGF core domain (Mohammadi et al., 2005) was verified for each of the predicted FGF candidate proteins in comparison to vertebrate FGF1, FGF2 and FGF8 (Supplement Fig. S2_1) using SMART (http://smart.embl-heidelberg.de) and InterPro (https://www.ebi.ac.uk/interpro). Structure prediction by modelling with the Expasy Swiss Model server (https://swissmodel.expasy.org) yielded the typical FGF beta-trefoil / donut-like structures for all Hydra FGFs similar to the vertebrate FGFs (Fig. 1A).

Fig. 1. Predicted protein structure of the fifteen Hydra fibroblast growth factors (FGFs) and genomic arrangement of the ten FGFs located on chromosome 03.

(A) The predicted Hydra FGF protein structures (https://swissmodel.expasy.org) show a beta trefoil structure like Homo sapiens (H.s.) FGF1, FGF2 and FGF8. (B) Distribution of the Hydra FGF genes on chromosome 03 is indicated by using the 5’ end position of the respective gene (BLAST analysis of the cDNAs in the two genomic libraries, Hydra T2T_AEP (NCBI RefSeq assembly GCF_038396675.1) and Hydra T2T_105 (GCF_037890685.1-RS_2024_06) representing the respective haplotype. Neighboring genes in green (FGFR and FGFR-like in both genomes; FGF-j and FGF-i in the Hydra T2T_105 genome, only). Features and ranges including the remaining five Hydra FGFs on chromosomes 09, 12 and 13 are given in Supplement Fig. S3_1 and Supplement Fig. S3_2).

Figure

Fig. 1. Predicted protein structure of the fifteen Hydra fibroblast growth factors (FGFs) and genomic arrangement of the ten FGFs located on chromosome 03.

(A) The predicted Hydra FGF protein structures (https://swissmodel.expasy.org) show a beta trefoil structure like Homo sapiens (H.s.) FGF1, FGF2 and FGF8. (B) Distribution of the Hydra FGF genes on chromosome 03 is indicated by using the 5’ end position of the respective gene (BLAST analysis of the cDNAs in the two genomic libraries, Hydra T2T_AEP (NCBI RefSeq assembly GCF_038396675.1) and Hydra T2T_105 (GCF_037890685.1-RS_2024_06) representing the respective haplotype. Neighboring genes in green (FGFR and FGFR-like in both genomes; FGF-j and FGF-i in the Hydra T2T_105 genome, only). Features and ranges including the remaining five Hydra FGFs on chromosomes 09, 12 and 13 are given in Supplement Fig. S3_1 and Supplement Fig. S3_2).

Close

Chromosomal distribution of Hydra FGF genes

Since some of the 22 vertebrate FGFs are clustered in the genome (reviewed in Itoh and Ornitz, 2004), we analysed the distribution of the 15 Hydra FGFs in the genomic libraries of two Hydra strains, Hydra T2T_105 versus Hydra T2T_AEP. BLAST search revealed a dispersed localization on four chromosomes. In both genomes, ten FGF genes are localized on chromosome 3, one each on chromosome 09 (Hv-fgf-n) and chromosome 13 (Hv-fgf-a), respectively, and three FGF genes (Hv-fgf-g, -o and -k) on chromosome 12 (Fig. 1B and Supplement S3). Moreover, a fragment of 409 nucleotides (nt) with 91% identity to Hv-fgf-c, spanning the 3’ coding and 3’ untranslated region, was revealed on chromosome 08 in the HydraT2T_AEP library (loc136084282 isoform x2 and loc124817691 isoform x2, see Supplement S3_1). This fragment is not annotated as a gene.

In the Hydra AEP genome, all 15 FGF genes are separated from each other by several other genes. In contrast, in the Hydra magnipapillata wt105 genome Hv-fgf-j and Hv-fgf-i, are placed as neighbors at a distance of 22.388 nt with no other genes identified in between. Interestingly, the gene for FGFRa, Kringelchen, (Sudhop et al., 2004) and a gene sequence encoding several alternative frames for smaller FGFR-like proteins are placed between the Hv-fgf-m and Hv-fgf-h genes. While the FGFR and FGFR-like sequences are direct neighbors, they are separated from the two FGFs by other genes (Fig. 1B).

Several Hydra FGFs have been annotated with identical/redundant names in the genomic libraries e.g. FGF1 is given three times in the Hydra T2T_AEP and 4 times in the Hydra 2T2_105 libraries to different genes (Supplement S3_1 and Supplement S3_2). We therefore propose to rename (annotate) the FGFs according to the letter code we had started to use for the first 5 FGFs in 2014 (Lange et al., 2014). This should help to unequivocally assign the expression pattern and functions of each Hydra FGF and to avoid the direct reference to the 22 FGFs numbered in vertebrates

Five Hydra FGFs contain a well-recognizable N-terminal signal sequence

Since most bilaterian FGFs act in a paracrine, extracellular way (reviewed in Ornitz and Itoh, 2022) and use an N-terminal signal sequence for secretion, we searched for signal sequences in the Hydra FGFs. SignalP 6.0 (https://services.healthtech.dtu.dk/services/SignalP-6.0) predicted with 93 - 99.9 % probability a signal peptide for five Hydra FGFs (see supplemental Table S1B and Supplement Fig. S2_1) (Hv-FGF-e, Hv-FGF-f, Hv-FGF-h, Hv-FGF-l, and Hv-FGF-m). Another four FGFs (Hv-FGF-c, Hv-FGF-g, Hv-FGF-i, and Hv-FGF-p) were assigned a medium to low probability for a signal sequence (54 % - 69 %) and in the remaining six Hydra FGFs presence of a signal sequence is questionable (Hv-FGF-j and Hv-FGF-n) or denied (Hv-FGF-a, Hv-FGF-b, Hv-FGF-k, and Hv-FGF-o).

It would be highly desirable to use purified Hydra FGFs for bioassays. To obtain a first clue whether the Hydra FGFs are expressed and secreted in a heterologous system, we used the eukaryotic HEK293T expression system. The coding region of two Hydra FGFs with high probability for a signal sequence (Hv-FGFe und Hv-FGFf), of one with a medium probability (Hv-FGFc) and one with no identifiable signal peptide (Hv-FGFb) were cloned in the pcDNA 3.1-myc-His A expression vector and transfected in HEK293T cells. The vertebrate cells expressed the tagged Hydra FGFs at high levels despite their alkaline pI. All four FGFs were detected within the cells by an anti-myc antibody in the cytoplasm, either dispersed within the whole cell or in granulae below the cell membrane and/or in filopodia and varicosity-like thickenings of filopodia (Supplement Fig. S2_2). Hv-FGF-e and Hv-FGF-f were isolated from cell extracts by affinity purification via the His-tag, but none of the Hydra FGF proteins was detected in the cell culture medium (Supplement Fig. S2_3). On a Western Blot, however, multiple bands were detected (not shown). Taken together, HEK293T is a promising system to express Hydra FGFs for future bioassays, but their secretability has to be investigated and purification steps optimized.

All Hydra FGFs are transcribed locally restricted

In contrast to the previously reported almost ubiquitous expression of the Hydra FGFs Hv-FGFa, Hv-FGFc and Hv-FGFe (Lange et al., 2014), a reduced antisense RNA probe concentration as well as additional destaining steps revealed localized expression domains in Hydra tissue (Fig. 2 for Hv-fgf-b and Hv-fgf-c; Supplement Fig. S4, plate I for Hv-fgf-a and Hv-fgf-e and plate II for Hv-fgf-f). All of the newly identified FGFs are as well expressed in restricted expression domains. Supplement Fig. S4, plates III and IV provide a first overview. Their expression pattern as well as that of Hv-fgf-a and Hv-fgf-e will be summarized here only shortly since our further focus was on Hydra FGF-b and FGF-c. As documented in Supplement Fig. S4 several of the new Hydra FGFs are upregulated in terminal structures (Supplement Fig. S4, Plate III – IV), particularly in the basal disc or in tentacles (Hv-fgf-b, Hv-fgf-f, Hv-fgf-I, Hv-fgf-j, Hv-fgf-p). Some are expressed complementary to others as are the ectodermal Hv-fgf-i and the endodermal Hv-fgf-j or Hv-fgf-p in tentacle tips. Some are strongly upregulated in testes like Hv-fgf-h (early testis) and Hv-fgf-k (maturing testis) or weakly like Hv-fgf-f at the testis base. Dynamic changes with switches between ecto and endodermal patterns are obvious like, for example, in the Hv-fgf-l transcription in the early bud field and developing young polyp. The Hv-fgf-l expression domain partially overlaps and resembles Hv-fgf-f, an FGF which is transcribed in cells adjacent to both Hydra FGFRs at the bud base (Supplement Fig. S4, plate II; (Lange et al., 2014)). For Hv-FGF-n only a sense signal (speckled upper body column), but no antisense signal was detected (not shown).

Fig. 2. RNA transcription patterns of Hv-fgf-b and Hv-fgf-c.

In situ hybridization of antisense RNA probes for Hv-fgf-b (A - I’’) and Hv-fgf-c (J – J3’). Budding stages are given for Hv-fgf-b according to Otto and Campbell (1977). Abbreviations: bb bud base, ec ectoderm, en endoderm, fp foot pore, m mesoglea, st stalk region between basal disk and budding zone, tb tentacle bud, tez prospective tentacle zone. (J1’ – J2’) Hv-fgf-c positive tentacle base cells (arrows) are numbered to highlight their chain-like arrangement. (J3, J3’) Optical focus on the strongest staining cell with cytoplasmic extensions and an unstained central part (arrow). Scale bars, 100 µm unless indicated otherwise.

Figure

Fig. 2. RNA transcription patterns of Hv-fgf-b and Hv-fgf-c.

In situ hybridization of antisense RNA probes for Hv-fgf-b (A - I’’) and Hv-fgf-c (J – J3’). Budding stages are given for Hv-fgf-b according to Otto and Campbell (1977). Abbreviations: bb bud base, ec ectoderm, en endoderm, fp foot pore, m mesoglea, st stalk region between basal disk and budding zone, tb tentacle bud, tez prospective tentacle zone. (J1’ – J2’) Hv-fgf-c positive tentacle base cells (arrows) are numbered to highlight their chain-like arrangement. (J3, J3’) Optical focus on the strongest staining cell with cytoplasmic extensions and an unstained central part (arrow). Scale bars, 100 µm unless indicated otherwise.

Close

This first overview documents clear patterns which suggest locally restricted functions in specialized, terminally differentiated or differentiating cells (tentacle tip, foot pore, testis) or in cells changing their shape like at the tentacle base, in the evaginating bud or later at its base.

Transcription patterns of Hv-fgf-b and Hv-fgf-c

In the following we focus on Hv-FGF-b and Hv-FGF-c (Fig. 2), their mRNA expression patterns and a first functional analysis using siRNA. Hv-fgf-b mRNA is localized endodermal with the exception of a transient weak ectodermal domain at the tip of the early bud, stage 1-2 (Fig. 2 B-B’’). The gene is transcribed strongly in the tentacle endoderm and detected at a prolonged incubation time in a few cells of the basal disc endoderm (Fig. 2 A–I’’). A staining protocol with a highly diluted probe and short exposure times revealed small endodermal tentacle cells resembling round nematocytes (stenoteles) and few nests of small cells plus a triangular signal right between the huge epithelial cells which resembles neurons (Fig. 2A’ and Fig. 3 A,B). Recently published data reported the lack of neurons in tentacle endoderm (Keramidioti et al., 2024). We therefore carried out a single cell analysis of the upper body region. A few stained stenoteles and smaller cells, resembling desmonemes (Fig. 3 C,C’) were detected, but no neurons. As alternative explanation we take into consideration a triangular signal resulting from the cytoplasm of the highly vacuolized endodermal epithelial cells which is compressed into one corner.

Fig. 3. In situ hybridization signals of Hv-fgf-b in the tentacle endoderm.

(A,B) Tentacles digitally enlarged from Fig. 2A; endodermal signals resembling neurons (open arrows) and nematocytes (closed arrows) below the mesogloea (mg). (C,C‘) Single cell preparation of a Hydra head, cut close to the tentacles. st stenotele, de desmonemes.

Figure

Fig. 3. In situ hybridization signals of Hv-fgf-b in the tentacle endoderm.

(A,B) Tentacles digitally enlarged from Fig. 2A; endodermal signals resembling neurons (open arrows) and nematocytes (closed arrows) below the mesogloea (mg). (C,C‘) Single cell preparation of a Hydra head, cut close to the tentacles. st stenotele, de desmonemes.

Close

Prolongation of the staining time blurred the above-described pattern and resulted in a dark blue tentacle endoderm (mainly by staining endodermal epithelial cells). Additional Hv-fgf-b positive basal disk cells were now disclosed around the foot pore (Fig. 2 B-I,I’). Moreover, transcription was found upregulated throughout bud evagination: from bud stages 1-3 (evagination phase) in the whole bud endoderm with a maximum in the bud tip endo- and ectoderm (Fig. 2 B-C’’), in bud stages 4 and 5 (elongation phase), in a transient broad endodermal ring of cells at the bud base and in the prospective tentacle zone (Fig. 2 D-E,E’) was established. Thereafter endodermal cells in tentacle buds and transiently surrounding the mouth (Fig. 2 F-G’,I’’) as well as at the foot pore (Fig. 2 H–I’,I’’) became positive and the adult pattern was attained shortly before detachment from the parent (Fig. 2H).

Hv-fgf-c, in contrast, was transcribed ectodermal at the tentacle bases and in a broad belt surrounding the stalk, distant to the basal disc (Fig. 2J). This mature pattern was attained in growing buds from bud stage 7 onwards in tentacles and from stage 8 onwards in the stalk (not shown). In contrast, to Hv-fgf-b, the gene is not transcribed in additional regions. Interesting is the chain-like arrangement of Hv-fgf-c positive cells around the tentacle base (Fig. 2 J1-J2’) with a stronger signal in cells at the aboral tentacle base close to the body column (Fig. 2 J1-J3). Here, single cells were distinguishable which seemed to either form strongly staining cytoplasmic protrusions connecting to other cells (Fig. 2 J2’ between cells 4-5) or proximal placed, huge cells with short protrusions (Fig. 2 J1’,J3’). By analyzing single cell preparations with in situ hybridization, these cells could, unfortunately, not be further characterized (data not shown). In summary, both Hv-fgf-b and Hv-fgf-c are interesting markers for tentacle-specific endo- and ectodermal cell types.

RNA interference for Hv-fgf-b and Hv-fgf-c

Since Hv-fgf-b and Hv-fgf-c are expressed in the tentacle endoderm and in specific ectodermal cells at the tentacle base, respectively, we asked whether a knockdown using siRNA electroporation affects tentacle morphology. Using the Hydra vulgaris strain AEP which is used widely as the strain to generate transgenic polyps, no morphological changes were observed. To better visualize ecto- and endoderm, we used the Traffic Light Hydra line which is a chimaera between the Hydra vulgaris AEP watermelon strain and Hydra vulgaris AEP ss1 YPet (Wang et al., 2022).

In four independent experiments we electroporated siRNAs once in early budding (for Hv-fgfr-a) or nonbudding animals (Hv-fgf-b and -c; negative control). Using a well-established protocol (Lommel et al., 2017), Hv-fgfr-a siRNA as a positive control was electroporated in polyps bearing stage 1-3 buds (Supplement Fig. S5). Scrambled GFP siRNA (scrGFP) or siRNA pGL2 served as negative controls (see Supplement Fig. S6 for evaluation of siRNA effects). Hv-fgf-c downregulation was clearly visible (S6, A-C’). All polyps lost their tentacles within a couple of hours after electroporation but recovered within two days post electroporation (dpe) and were able to feed again on day 2 dpe. Morphology was analyzed in detail from day 5 or 6 after electroporation onwards. Neither the scrGFP nor the pGL2 nor the Hv-fgf-b siRNAs had visible effects on bud or parent morphology - except split tentacles occurring in almost all the electroporation experiments at low incidence of 1-2 animals per 30 (not shown). Between 50 - 60% of the positive control, Hv-fgfr-a siRNA treated polyps carrying stage 1-3 buds, developed the known branching phenotype (Sudhop et al., 2004; Hasse et al., 2014) as shown in Supplement Fig. S5.

A striking phenotype occurred in 40 - 70% of the Hv-fgf-c siRNA-treated animals which was not seen with any of the other siRNAs (Fig. 4; raw data in Supplement S5_2). The siHv-fgf-c – treated polyps developed tentacles with (multiple) irregular shaped battery cell arrangements and even packed battery cell zones resembling in their most expressed form nodules bulging from the tentacles without a distinguishable separation of the battery cells (Fig. 4 A-G, Fig. 5 A,B). The nodule shape was fixed and they neither contracted nor elongated. All other tested siRNAs resulted in normal tentacles (Fig. 4 H-M for scrGFP and Fig. 4 N–S for Hv-fgf-b; Supplement S5 A,B for Hv-fgfr-a).

Fig. 4. Knockdown of Hv-fgf-c by siRNA and ectodermal tissue dynamics in the Hydra head.

(A–S) Budless transgenic polyps were subject to siRNA electroporation and their morphology evaluated on day 12 (A–G) Hv-fgf-c, (H-M) scrGFP and (N–S) Hv-fgf-b. (A) Picture taken by a dissection microscope, Hydra anesthetized in 1 mM linalool (Goel et al., 2019). (A-G) Tentacle nodules (arrows) formed by a bloated endoderm with tightly clustered ectodermal battery cells; apparently normal battery cells (arrow heads) arranged like beads on a string distant from adjacent battery cells. Green fluorescence is ectodermal, red fluorescence is endodermal. (T-Y) Tissue dynamics of ecto- and endoderm towards the head analysed in transplants of normal upper body half and transgenic lower body, 15 days post transplantation (dpt). (T,U) 23 h post transplantation; (V,W) 8 dpt: the ectoderm has moved quickly towards the head, the endoderm is stalled; (X) 10 dpt: ectodermal cells cross the body-tentacle boundary at the lower tentacle base and quickly spread along the lower side of the tentacle (arrow) or have already enclosed the complete tentacle base circumference (open arrow); (Y) 15 dpt: tentacles are completely covered by transgenic ectodermal cells.

Figure

Fig. 4. Knockdown of Hv-fgf-c by siRNA and ectodermal tissue dynamics in the Hydra head.

(A–S) Budless transgenic polyps were subject to siRNA electroporation and their morphology evaluated on day 12 (A–G) Hv-fgf-c, (H-M) scrGFP and (N–S) Hv-fgf-b. (A) Picture taken by a dissection microscope, Hydra anesthetized in 1 mM linalool (Goel et al., 2019). (A-G) Tentacle nodules (arrows) formed by a bloated endoderm with tightly clustered ectodermal battery cells; apparently normal battery cells (arrow heads) arranged like beads on a string distant from adjacent battery cells. Green fluorescence is ectodermal, red fluorescence is endodermal. (T-Y) Tissue dynamics of ecto- and endoderm towards the head analysed in transplants of normal upper body half and transgenic lower body, 15 days post transplantation (dpt). (T,U) 23 h post transplantation; (V,W) 8 dpt: the ectoderm has moved quickly towards the head, the endoderm is stalled; (X) 10 dpt: ectodermal cells cross the body-tentacle boundary at the lower tentacle base and quickly spread along the lower side of the tentacle (arrow) or have already enclosed the complete tentacle base circumference (open arrow); (Y) 15 dpt: tentacles are completely covered by transgenic ectodermal cells.

Close

Fig. 5. Morphology of Hv-fgf-c by siRNA treated tentacles.

(A-D) Hydra AEP traffic light. (A) Overview of several elongated tentacles with aberrant battery cell arrangements or nodules formed by bloated endoderm and tightly packed battery cell clusters (arrows) besides normal-appearing battery cells (arrow heads); stenoteles (s) and desmonemes (d) are labelled. (A’) scheme of a normal battery cell (open arrowhead); a contracted (ct) normal tentacle is visible on the right. (B) Tentacle with a central nodule and flanking aberrant battery cells (arrows) as well as with an almost normal battery cell (arrowhead). (s) stenotele (the biggest, toxin-filled nematocyte). (B’) Scheme of one of the aberrant elongated battery cells (open arrow). Hydra anesthetized in 1 mM linalool for max. 10 min. (A,B) 40 x magnification, DIC optics; (C,D) fluorescence detected (C) ectodermally using a GFP/FITC or (D) endodermally by a Rhodamine/TRITC fluorescence filter set. (E) Wild type Hydra AEP, 6 dpe. One tentacle with minor irregularities in battery cells (arrows). Left: missing nematocytes, right: three stenoteles in one position.

Figure

Fig. 5. Morphology of Hv-fgf-c by siRNA treated tentacles.

(A-D) Hydra AEP traffic light. (A) Overview of several elongated tentacles with aberrant battery cell arrangements or nodules formed by bloated endoderm and tightly packed battery cell clusters (arrows) besides normal-appearing battery cells (arrow heads); stenoteles (s) and desmonemes (d) are labelled. (A’) scheme of a normal battery cell (open arrowhead); a contracted (ct) normal tentacle is visible on the right. (B) Tentacle with a central nodule and flanking aberrant battery cells (arrows) as well as with an almost normal battery cell (arrowhead). (s) stenotele (the biggest, toxin-filled nematocyte). (B’) Scheme of one of the aberrant elongated battery cells (open arrow). Hydra anesthetized in 1 mM linalool for max. 10 min. (A,B) 40 x magnification, DIC optics; (C,D) fluorescence detected (C) ectodermally using a GFP/FITC or (D) endodermally by a Rhodamine/TRITC fluorescence filter set. (E) Wild type Hydra AEP, 6 dpe. One tentacle with minor irregularities in battery cells (arrows). Left: missing nematocytes, right: three stenoteles in one position.

Close

Within the nodules evoked by siHv-fgf-c, the endoderm was bloated and covered with multiple ectodermal battery cells. The pattern of nematocytes in the nodules was irregular with an extreme amount of stenoteles packed within each nodule (Fig. 5 A,B) instead of the regular pattern consisting of one or two central stenoteles per battery cell surrounded by multiple, smaller isorhizas and desmonemes (compare to Fig. 5 A’,B’). Feeding the animals evoked normal reactions with multiple stenoteles ejected from the nodules (not shown). The tentacle endoderm in the unchanged tentacle regions appeared smooth and without any signs of endodermal bloating (Fig. 4 G,M,S and Fig. 5D). First aberrant tentacles were observed on day 5 dpe and persisted until at least day 21. Re-evaluation of the Hydra AEP siHv-fgf-c experiments yielded a single case of tentacles with irregular shaped / equipped battery cells (1/30) and none in the positive and negative controls (Fig. 5E). The phenotype of tentacle nodules thus occurred in the Hydra AEP Traffic Light strain while the wild type Hydra AEP showed only minor disturbances of the battery cell pattern if any.

Since battery cells enter the tentacles from the body column and not from the hypostomal side (Hobmayer et al., 2012), we monitored the dynamics of tissue movement towards the head and into the tentacles of transplants from transgenic and nontransgenic Hydra polyps. To this end polyps were bisected at 50% body length and a lower transgenic half transplanted to an upper nontransgenic one. Transplants joined and healed completely within 2 hours. Monitoring tissue movement after 1, 8, 10 and 15 days revealed a quick movement of ectodermal tissue and a very slow one of endodermal tissue (Fig. 4 V,W) as described by others (Campbell 1967). The ectodermal epithelial cells entered the tentacles from the lower side no matter whether control polyps or Hv-fgf-c-siRNA polyps (6 dpe) were used. They moved quickly towards the tentacle tip forming a transient green stripe on its lower side (Fig. 4X) for 2-3 hours. About 4 hours later epithelial cells had reached the upper part of the tentacle base and closed in forming a complete ring. On day 15, the tentacles were completely covered by ectodermal transgenic cells (Fig. 4Y). Interesting is the dynamics of this tissue movement – a very fast entrance into the lower side of the tentacles, a trailing movement on the upper side of the tentacle and a slow movement of tissue between the tentacles and into the hypostome. This kinetics was similar in control and siHv-fgf-c animals.

Taken together, the knockdown of Hv-fgf-c induces tentacle nodules and irregular shaped battery cell arrangements in Hydra AEP Traffic light. Their occurrence on day five correlates with the kinetics of epithelial cells (= battery cells) covering a tentacle completely.

The properties of Hydra FGFs resemble bilaterian FGFs, but are they secreted?

All fifteen Hydra FGFs show typical characteristics of vertebrate FGFs (for recent review see Ornitz and Itoh, 2022) including the beta-trefoil structure, a well-identifiable core domain, an alkaline pI and a molecular weight between 20 and 37 kDa. Differing from vertebrates is that only five of the fifteen Hydra FGFs, within them Hv-FGFe and Hv-FGFf, carry an N-terminal signal peptide predicted with high probability. A signal peptide is usually a good indicator for a secreted protein and most bilaterian FGFs act as paracrine or autocrine growth factors either using a signal sequence or, rarely, a noncanonical secretion mechanism like FGF1 and FGF2 (Ornitz and Itoh, 2022). In pilot experiments to obtain FGF proteins for bioassays, we expressed four FGFs in HEK293T cells. These have been used successfully to express Hydra proteins including secreted Wnt proteins (Vogg et al., 2019; Lengfeld et al., 2009). Hydra FGF localization in the cytoplasm and in filopodia of HEK293T cells was as expected, but most of the proteins located to submembrane granulae. Furthermore, despite affinity-purification, Hv-FGFe and Hv-FGFf were detectable within cell extracts, exclusively, and not in the cell culture medium. Whether the granular accumulation is due to secretion problems as known for vertebrate FGFs when flanking or internal sequences affect secretion (Miyakawa and Imamura, 2003) remains to be shown.

Given the lack of well-supported signal peptides in ten of the fifteen FGFs and considering the close spatial expression of Hydra FGFRa and – b with Hv-FGF-f (shown in S4, Plate II by a double labeling), we cannot exclude that Hydra FGFs act on a short range or even within the cells. This will be discussed in the following, mainly with respect to the Hydra head.

Hydra FGFs are expressed strongly in restricted domains and might act at short distances or even intracellular

Locally upregulated expression in testis, terminal and/or boundary regions is a hallmark of the FGFs analysed here and in previously studies (Krishnapati and Ghaskadbi, 2013, Lange et al., 2014).

FGF-j and FGF-i

The two genomic neighbors in the Hydra T2T_105 genome, Hv-fgf-i and Hv-fgf-j are both expressed in the extreme tentacle tips of Hydra AEP - however in different tissue layers: Hv-fgf-i in ectodermal and FGF-j in endodermal epithelial cells. Both proteins are closely related, with a similar pI of 9,6 and 9,3 and a low probability for a signal peptide (53,2 % for FGF-i, 30,6% for FGF-j). It is likely that they were generated by a duplication, but what is their potential function in this tissue? Tentacle tip cells are terminally differentiated and strongly bent to keep the tissue closed even if wiggling prey is fixed. They are thus in a very exposed and mechanically stressed position. Tentacle tip cells are sloughed off eventually without disturbing the integrity of the tentacle. A local FGF function might therefore be in cell physiology or by acting on a short range, e.g. enabling the woundless detachment of cells at the tip. This hypothesis fits with the FGFR-dependent smooth detachment of cells from adjacent ones when bud and parent separate (Hasse et al., 2014; Holz et al., 2017). Another function might be prevention of apoptosis which is a main function of vertebrate FGF2 and FGF8 (Katoh and Nakagama, 2014). Apoptotic cells occur frequently along the tentacles as shown by Tunel analysis, but they are rare in the tentacle tips (Chera et al., 2009). However, both genes are not expressed in the testes where apoptosis occurs frequently and mostly basal where stem cells enter the testis (Kuznetsov et al., 2001). FGF-f, FGF-h and FGF-k are expressed in basal and apical positions or across the whole testis.

Whether FGF-j and FGF-i exert similar (non)redundant functions in the ecto- and endoderm of Hydra remains an interesting question particularly with respect to their genomic neighborship in Hydra magnipapillata wt 105 and their separation in the Hydra AEP genome.

FGF-b

A second typical example for local transcription is Hv-fgf-b in endodermal tentacle cells and in just 10-12 cells at the foot pore. FGF-b is identical to FGF1 (t12060-AEP, gene ID G005481, 688 nucleotides) which was identified in the Hydra 2.0 Genome Project Portal as endodermal tentacle and basal disc gene (Siebert et al., 2019). The overall expression pattern is congruent. Our study now adds a detailed spatial resolution like presence of FGF-b positive stenotele and desmonemes nematocytes in the tentacle endoderm.

The Hydra endoderm contains only degenerating nematocytes engulfed by endodermal processes and phagocytosed (Lyon et al., 1982). The FGF-b positive huge stenotele nematocyte-like cells had no visible capsule or thread-structure, but their size and shape with the stained cytoplasm in a peripheral position suggests that they are degenerating stenotele nematocytes as described in the 1980s using ultrastructural methods. Function of the FGFb in these degenerating, but not in the functional ectodermal nematocytes is obscure. Vertebrate FGFs prevent apoptosis, induce proliferation, control cell shape, their migration or differentiation (Ornitz and Itoh, 2022; Katoh and Nakagama, 2014). None of these functions seems relevant for degenerating nematocytes.

We were unable to track an FGF-b function for morphogenesis by siRNA knockdown. Given the reliability of the siRNA knockdown for Hv-fgfra and Hv-FGF-c, we propose to consider localized, cell type-specific (physiological) function(s) for FGF-b in the terminally differentiated tentacle and foot pore cells. Hv-FGF-b lacks a signal sequence. Although there are alternative secretion mechanisms for FGFs (Ornitz and Itoh, 2022; Biadun et al., 2024), Hv-FGF-b, is a potential candidate as intracellular FGF. In vertebrates, FGFs 11-14 fail to bind FGFR and act intracellular in excitable cells like neurons and cardiomyocytes where they modulate voltage-gated sodium channels and calcium-induced calcium release (Hennessey et al., 2013). The tentacle endoderm lacks neurons according to recent investigations (Keramidioti et al., 2024) and channel modulating activities in epithelial cells are therefore an interesting aspect. Although the ectodermal neurons and battery-epitheliomuscle cells are sufficient to induce tentacle contractions (the ectodermal myonemes run longitudinally), an autonomous endodermal epithelial excitability would be helpful to support tentacle elongation and movement. Endodermal epitheliomuscle cells constrict the circumference and elongate the Hydra body.

A si_Hv-fgf-c knockdown evokes battery cell nodules

The third FGF example to discuss, Hv-fgf-c is interesting as a morphogenesis controlling growth factor. Hv-fgf-c mRNA is upregulated in a ring around the basal disc and in a few ectodermal cells at the tentacles bases which seem to be connected to each other, resembling a chain. Their position is strategic, because actively migrating cells coming in from the body column have to pass this region and must receive signals to terminally differentiate and reach their target position in the battery cells.

The siRNA of Hv-fgf-c did not affect the foot region morphologically, but evoked a striking tentacle phenotype: tentacles developed nodules, from day five after siRNA electroporation. This corresponds to about the time of the maximal siRNA effect in Hydra (Lommel et al., 2017). In nodules the endoderm was bloated and the ectodermal battery cells were tightly clustered on the outside and equipped abundantly with a chaotic arrangement of, mostly, stenotele nematocytes. A milder phenotype occurred flanking to bloated regions: here, the nematocyte charge of battery cells appeared irregular, but the distance to neighboring cells or the nodule was kept.

The results suggest that Hydra FGFc is involved in the regulation of arrangement and / or equipment of battery cells, particularly with respect to stenoteles. FGFc might control the canonical pattern within a battery cell with 1-2 huge central stenoteles and surrounding small desmonemes as well as isorhiza. This pattern determines the battery cell shape. Central stenoteles raise battery cells in the middle and the surrounding smaller nematocytes allow a flattening towards adjacent battery cells. Empty battery cells at the tentacle base or deprived of nematocytes are flat and the tentacles appear smooth. As a first simple hypothesis, we propose that battery cell positions along the tentacles are unchanged and cell-to-cell contacts developed in their normal way.

Nematocytes integrate in battery cell pouches (Novak and Wood, 1983), and tightly anchor in the mesoglea. If too many stenoteles integrate in adjacent battery cells their shape would change. Provided, the tissue has a limited flexibility, an overload with stenoteles in adjacent battery cells would raise the cells and compact them closely resembling a cluster. Nodules are the result when the signal for stenotele immigration is too strong for a limited time span.

FGFc is, therewith, a candidate to control stenotele nematocyte distribution to battery cells.

The FGF-c knockdown tentacle phenotype is not obtained by an FGFR knockdown and occurs strain specific

Although the FGF-c knockdown phenotype is striking and mechanistically interesting, there are two obvious problems. First, the FGFc knockdown phenotype is not obtained by a knockdown of Hydra FGFRa. Both FGFRa (Kringelchen) as well as the slightly divergent FGFRb are co-expressed weakly throughout the body column, but upregulated only during budding at the constriction and detachment site and in early tentacle development (Supplement Fig. S4; Sudhop et al., 2004; Suryawanshi et al., 2020). This does not exclude an FGF/FGFR interaction in low expression regions like at the mature tentacles, but 54% probability for a signal sequence is poor and release of FGFc from tentacle base cells questionable. To affect e.g. stenotele immigration or battery cell load a diffusible signal even on short distance is a most likely mechanism.

Second, the phenotype was strain-specific - obtained in the Hydra AEP traffic light line (Wang et al., 2022) only and not in Hydra vulgaris AEP. As one possibility to explain this discrepancy we consider its genotype. Hydra AEP traffic light was generated in several steps from chimaera of Hydra vulgaris AEP water melon and Hydra magnipapillata A10 (Wang et al., 2022). A10 is a hybrid between the wild type Hydra magnipapillata wt 105 and nematocytes of H. magnipapillata sf1, a mutant with temperature-sensitive i-cells, generated by transplantation of interstitial cells to epithelial wild type Hydra (Kasahara and Bosch, 2003). These I-cells were first introduced and then removed using colchicine in the steps to generate the tricolor Traffic light Hydra. It cannot be excluded that a few very slow cycling A10 stem cells survived this process and persist in Hydra AEP traffic light - now allowing for effects not seen in Hydra AEP. As well, the insertion of the YPet construct in Hydra AEP ss1 used to generate the Traffic Light ectoderm is a factor which might be responsible for the susceptibility of Hydra AEP Traffic light to the FGF-c knockdown.

As a third possibility, we have to discuss potential off-target effects of the Hv-fgf-c knockdown in Hydra AEP Traffic Light. Both Hv-fgf-c siRNAs were selected carefully by (i) using the suggestions of Kaneka Eurogentec for optimal siRNAs and (ii) controlling these sequences against the Hydra nucleotide databases to avoid interference with other genes. This approach sometimes requires a trade-off, because the best-suited siRNA sequences might not be appropriate. In the case of Hv-fgf-c, siHv-fgf-c_1 was designed, as usual, for a sequence close to the 5’end. For a second siRNA, no reasonable match was possible within the coding sequence. We instead had to use for siHv-fgf-c_2 a sequence in the 3’UTR of the mRNA. Since this siRNA yielded an identical tentacle phenotype, although at a lower incidence, we consider an off- target effect to be unlikely.

Although the strong effect (tentacle nodules) is strain-specific, the basal ability to react on the Hv-fgf-c knockdown and therewith FGF-c signals altering tentacle morphology and cell composition is obviously present in both, Hydra AEP as well as Hydra AEP Traffic light. In summary, despite some puzzling results, the whole set of Hydra FGFs shows very interesting features and expression patterns for future investigations. Elucidating the role of this ancestral signaling system in the “simple” fresh water polyp Hydra remains a fascinating task.

Hydra culture and manipulation

The Traffic light strain of Hydra AEP was obtained from Eva Collins, Swarthmore College, formerly UC San Diego (Wang et al., 2022). Hydra vulgaris AEP was obtained from Thomas Bosch, Kiel and both were cultured in hydra medium (HM) under standard conditions (Klimovich et al., 2019). All manipulations of Hydra were done with polyps unfed for at least 24 hrs.

Grafting and relaxation

60 mm Petri dishes were prepared for grafting by pouring 2-3 ml melted paraffin into a 45° tilted dish and placing a 0.1 mm insect needle head down in the soft paraffin at an angle of 20 - 30° during the hardening process. Small silicon cushions (2 x 2 mm) were prepared from silicon tubing and pierced in the middle using an injection cannula. One cushion was immediately transferred to the insect needle and later fixed the first transplant on one side. The grafting dish was filled with HM until the tip of the needle was well submerged. One transgenic and one non-transgenic animal were relaxed in 1 mM linalool in HM for 10 min (Goel et al., 2019) and bisected in 100 mm dishes. The parts of interest were transferred to the grafting dish and transplanted within a maximum of 3 min after bisection. All manipulations were done using a Dumont No 5 forceps (ROTH 10427301). The head of a recipient animal was placed on the insect needle with its hypostome towards and close to the silicon cushion. Then the open body column of the lower part of a donor was forced on the needle and the transplant slightly pressed together by a second silicon cushion. Transplants were removed from the needle after 2 hrs.

Imaging

All pictures of living Hydra were taken within 10 min following about 10 min relaxation in 1 mM Linalool/HM. Tentacle pictures should be taken as fast as possible, otherwise the battery cells flatten and are difficult to evaluate. Single polyps were transferred to a slide in 100 μl linalool/HM and overlayed by a 2 x 2 cm sheet of cellophane. This way polyps can be flipped and pictures taken from both sides for about 10 min. Thereafter, polyps were transferred to a petri dish with hydra medium and washed twice for 5 min each. Fixed polyps were prepared following standard procedures (Klimovich et al., 2019) and embedded permanently in Roti®Mount FluorCare. GFP and dsRED fluorescence were monitored on the Nikon TS2000 Eclipse fluorescence or as cLSM on a LeicaTCSSP5 microscope.

Genomic databases

We analysed the genomic distribution of the 15 Hydra FGFs in the genomic libraries of two Hydra strains, Hydra T2T_105 versus Hydra T2T_AEP (https://www.ncbi.nlm.nih.gov/datasets/genome/GCF_037890685.1; 25.7.2025). Both were generated from the haplotype of a Hydra magnipapillata wt 105/Hydra vulgaris AEP hybrid by chromosome scale telomere to telomere assemblies. Data were derived using BLAST of the cDNA sequences and the Genome Data Viewer to retrieve features and ranges of the genes (see Supplement S3). Further data like accession numbers for nucleotide sequences and proteins are available using the Genome Data Viewer.

In situ hybridization

In situ hybridization and controls were carried out as previously described (Suryawanshi et al., 2020). The Hv-fgf and Hv-fgfr-a, Hv-fgfr-b (anti)sense probes were diluted after synthesis by a factor of 1:10 for a first attempt and up to 1:20.000 to optimize the results. As an example, for hv-fgf-b, the probe was diluted 1:1000 and staining time was reduced to 5-10 min for a sensitive detection of Hv-fgf-b neurons, but to detect the positive cells close to the foot pore, staining was prolonged to 1,5 hrs (or probe used 1:100) and stopped with 1x TE buffer, pH 7,5. Additional to previously published methods, background staining was reduced by changing from 1x TE to 100% MeOH in 3 steps (75:25, 50:50, 25:75) and immersing for a maximum of 30 min in 100 % MeOH followed by rehydration in three steps to 1x PBS and a post-fixation step in 4% paraformaldehyde / 1x PBS for 30 min. Double in situ hybridization was performed according to (Hansen et al., 2000). In situ hybridization to single cell preparations as described (Hassel, 1998).

siRNA-mediated knockdown in Hydra

Appropriate siRNA sequences were proposed for each cDNA sequence by Kaneka Eurogentec and controlled by us blasting against the Hydra EST database to avoid cross reactions (sequences in Table 1). Usually, two siRNAs are placed close to the 5’ end. If no second appropriate sequence was identified a proposed sequence close to the 3’ end was chosen. Custom made RP-HPLC siRNA-duplexes were ordered as HPLC-purified, lyophilized siRNA and stored at -80°C.

Electroporation was performed according to Lommel et al., 2017 in triplicates of ten Hydra polyps each, unfed for 24 hrs. Budless polyps were used for Hydra fgf and pGL2/scrGFP siRNAs, while Hv-fgfr-a siRNA was applied to polyps carrying stage 3 buds to evoke the branching phenotype as positive control (Sudhop et al., 2004). Immediately prior to use the siRNAs were suspended in water. We either used 3 µM siRNA in 100 μl DEPC-treated, autoclaved MilliQ water for a single siRNA to be electroporated (fgfra_1 or scrGFP or pGL2 and the FGF siRNAs in preparatory experiments) or a total of 4 µM siRNA in 100 μl sterile water when two siRNAs were used at 2 µM each (fgfc_1 and fgfc_2 or fgfb_1 and fgfb_2). Electroporation was at 250 V, 25 ms using a square pulse (Fischer, Heidelberg, Square pulser (prototype 1998); 4 mm PeqLab or BioRad electroporation cuvettes). Polyps were handled very careful following electroporation, avoiding shearing by air bubbles or a strong liquid stream. 48 hours post electroporation Hydra had regenerated short to medium length tentacles. They were fed and carefully cleaned prior to two days starvation. Unless otherwise indicated, the polyps were fed 5 times a week thereafter and their morphology was analysed and documented prior to feeding.

Cloning of the Hv-fgf cDNA using the expression vector pcDNA3.1 myc-His A

The Hv-fgf cDNAs were amplified via PCR with specific oligonucleotide primers (Table 2) and ligated in the expression vector pcDNA3.1-myc-His A. This vector is suitable for mammalian cell culture systems and adds a C-terminal polyhistidine-tag plus a myc-tag to the translated protein. Both enable purification and detection of the protein of interest.

The coding sequences for Hydra vulgaris fgf-b, fgf-c, fgf-e, and fgf-f were amplified by PCR with Q5-polymerase (NEB, USA) under the following conditions: 5 min initial denaturation (98°C) followed by 35 cycles: 1 min denaturation at 98°C, 1 min annealing (fgf-b, fgf-c, fgf-e: 58°C; fgf-f: 56°C, 1 min elongation at 72°C; 10 min final elongation at 72°C and cooling/storage at 4°C. During PCR, the 5’ EcoRI and 3’ XhoI restriction recognition sites (underlined in Table 2) were incorporated via primer extension. The amplified inserts as well as the expression vector were double digested with EcoRI and XhoI under standard conditions and the opened vector was dephosphorylated with Antarctic Phosphatase (NEB) to avoid self-ligation. Ligation was performed with T4 Ligase (NEB) and the construct transformed into chemically competent cells of the Escherichia coli strain DH5α. Positive transformants were selected on Ampicillin containing Luria-Bertani (LB) agarose and the insert-containing plasmid DNA sequenced by Microsynth AG (Switzerland). The constructs were kept in DH5α for selection, amplification, and storage.

HEK293T cell culture and transfection

HEK293T cells are adherent cells. They were cultured in 10 ml Dulbeccos Modified Eagle Medium (DMEM, Gibco) with 10 % fetal bovine serum (FBS) and penicillin-streptomycin (100 U penicillin, 100 µg/ml streptomycin) in 75 cm² flasks at 37 °C and 5 % CO₂. For transfection, cells were grown in 6-well plates to a concentration of 5 x 10⁶ cells in 2 ml DMEM. The medium was replaced by 400 µl of fresh medium. For transfection, 2.5 µg plasmid DNA per Hv-fgf construct were mixed with 100 µl opti-MEM and 4 µl Lipofectamine (Thermo Fisher Scientific) and preincubated at room temperature for 15 min. The transfection mix was added to the cells for 6 hrs and thereafter diluted with 1 ml of fresh medium. Further incubation for up to four days followed at 37 °C and 5 % CO₂.

Cell harvest and determination of optimal protein expression in HEK293T

After protein expression, the cell culture medium was collected and centrifuged for 5 min at 3000 x g to remove any remaining cells. Aliquots of the supernatant were stored at – 20 °C for later analysis. The adhesive cells were treated with TrypLE^TM (Thermo Fisher) for 1 min at 37 °C and 5 % CO₂. The reaction was stopped by the addition of 2 ml fresh DMEM, the cell suspension transferred to reaction tubes and centrifuged for 5 min at 3000 x g. The supernatant was removed and 150 µl of lysis buffer containing 50 mM Tris pH 7.5, 150 mM NaCl, and 0.5 % (v/v) Triton X-100 (NP-40) and cOmplete^TM Mini protease inhibitor cocktail (Merck) was added. The pellet was resuspended in lysis buffer and stored at – 20 °C until Western Blot analysis.

The time of maximum Hv-FGF expression was determined using four individual wells of a 6 well plate per transfected Hv-FGF. Cells of one well, each, were harvested at 24, 48, 72 and 96 hours post transfection (hpt) and supernatant and cell lysate were analyzed via dot blot. To this end 2 μl of each cell lysate and supernatant sample were dotted onto a nitrocellulose membrane and dried for 1 h at room temperature (RT). The membrane was blocked with 5 % casein in 1 x TBS-T (50 mM Tris pH 7.5, 150 mM NaCl, 0.1% Tween-20) for 30 min at RT. Detection of the recombinant proteins by primary anti-myc antibody (Thermo Fisher Scientific, diluted 1:100 in blocking buffer) occurred for 30 min at RT. The membrane was washed 3 x 5 min with PBTx (1x PBS, 0.25% Triton-X 100) and incubated with the secondary HRP conjugated anti-mouse antibody (Thermo Fisher Scientific, diluted 1:100 in blocking buffer) for 30 min, RT. The antibody was removed by washing 3 x 5 min with PBTx, followed by an additional washing step in sterile H₂O. Detection of the antibody signal was performed via ECL in a LI-COR Odyssey FC.

Purification of FGFe and FGF-f by Nickel-NTA affinity chromatography

HEK293T cells were grown in 75 cm² flasks to a concentration of 5 x 10⁶ cells in 20 ml DMEM. Before transfection, the medium was removed and replaced with 4 ml of fresh medium. 8 μg plasmid DNA was mixed with 1 ml opti-MEM and 30 μl lipofectamine and incubated at room temperature for 15 min. After addition of the transfection mix the cells were placed in the incubator for at least 6 hrs before 10 ml medium was added for incubation at 37 °C, 5 % CO₂ for 72 hrs. The medium was removed from the cells, centrifuged for 5 min. at 3000xg and the supernatant stored in a fresh reaction tube at -20 °C until further analysis. The cells were treated with TrypLETM (Thermo Fisher Scientific) for 1 min at 37 °C, 5 % CO₂ and the reaction stopped by addition of 20 ml fresh DMEM. The cells were transferred into a fresh reaction tube and centrifuged for 5 min. at 3000xg. The supernatant was discarded and the cells were resuspended in 1,5 ml lysis buffer (50 mM Tris pH 7.5, 150 mM NaCl, 0.5 % (v/v) Triton X-100 and cOmpleteTM Mini protease inhibitor cocktail (Merck)).

Purification of the histidine-tagged FGFs was performed by immobilized metal affinity chromatography with Ni-NTA sepharose under native conditions as recommended by the manufacturer (QIAGEN). A 1,5 ml Ni-NTA sepharose column was prepared and washed once with 6 ml sterile H₂O and equilibrated twice with 6 ml native binding buffer (250 mM NaH₂PO₄ pH 8.0, 2.5 M NaCl). 1.5 ml cell lysate was diluted with 8 ml native binding buffer. The sample was pressed through an 18 gauge needle to reduce viscosity. The sample was centrifuged for 15 min, 3000xg, 4 °C, transferred to a fresh reaction tube and mixed with the prepared Ni-NTA Sepharose. Protein binding to the resin followed on a platform shaker (60 min, 125 rpm, 4 °C). The column was washed 5 x with 8 ml native washing buffer (250 mM NaH₂PO₄, pH 8.0, 2.5 M NaCl, 20 mM imidazole). The proteins were eluted with 8 ml native elution buffer (250 mM NaH₂PO₄, pH 8.0, 2.5 M NaCl, 250 mM imidazole). The eluate was collected in 1 ml fractions and stored at 4 °C until dot blot analysis.

Immunodetection in HEK293T cells

To detect the expression of the recombinant Hydra vulgaris AEP FGFs, transfected HEK293T cells were grown on coverslips and fixed in 4% paraformaldehyde/ 1x PBS and the proteins detected via the attached myc-tag. The mouse anti-myc antibody (Thermo Fisher Scientific) was diluted 1:100 in 1x PBS and detected with anti-mouse Alexa Fluor⁴⁸⁸ secondary antibody (BioLegend, diluted 1:500 in PBS). As a control, untransfected HEK293T cells were analysed.