ABSTRACT
Pectoral fins, the anterior paired fins in fish, have enhanced maneuvering abilities due to morphological changes. Teleosts have fewer radial bones in their pectoral fins than basal species, resulting in more-elaborate fins. The mechanism behind this radial constraint change in teleosts is unclear. Here, we found that mutations in hhip, which encodes an antagonist of Hedgehog signaling, led to an increase in radial bones in a localized region. Expression of the Shh genes, encoding ligands of Hedgehog signaling, coincided with notable hhip expression specifically during early development. We suggest that a negative feedback effect of Hedgehog signaling by hhip regulates the constraint of the pectoral fin in zebrafish. Additionally, re-analysis of hhip-related gene expression data in zebrafish and basal species revealed that the notable hhip expression during early development is a characteristic of zebrafish that is not observed in basal species. Region-specific expression of Hox13 genes in the zebrafish pectoral fin indicated that the median region, analogous to the region with abundant radials in basal species, is expanded in hhip−/− zebrafish. These data underscore potential morphological evolution through constrained diversity.
INTRODUCTION
Fins in fishes are a product of evolutionary adaptation and enhance diverse hydrodynamic maneuvers, including speed adjustments, hovering, and directional changes in the water. It is particularly noteworthy that pectoral fins, the anterior paired fins, have undergone morphological transformations in actinopterygians, resulting in a transition from their ancestral state to a derived form within the skeletal framework. In the ancestral state, pectoral fins possess a large number of radial bones, as observed in basal species, such as chondrichthyans, some basal actinopterygians, placoderms, and acanthodians (Coates, 1994, 2003; Davis et al., 2004; Onimaru et al., 2015; Starck, 1979; Tanaka, 2018). The pectoral fins of basal species are composed of multiple basal radials, comprising the tribasal bones (propterygium, mesopterygium, and metapterygium) (Gegenbaur, 1865; Onimaru et al., 2015; Tanaka, 2018). These fishes use wide pectoral fins for steady swimming or vertical directional changes (Wilga and Lauder, 1999, 2000). By contrast, pectoral fins in teleosts, derived fishes in actinopterygians, have lost these ancestral characteristics and have only a limited number of radials remaining (Grandel and Schulte-Merker, 1998; Tanaka et al., 2022; Woltering et al., 2018). While almost all teleosts have only four radials, some (e.g. Osteoglossiformes, Siluriformes, Stomiiformes) have fewer than three radials and only a few others have more than four radials (Tanaka et al., 2022). As a consequence of the radial constraint, teleosts have elaborated their pectoral fins into compact ones for speed adjustment, hovering, or horizontal directional changes (Drucker and Lauder, 2002, 2003).
Elaboration of pectoral fins in teleosts has occurred along the anterior-posterior (AP) axis. In the development of vertebrate appendages, Hedgehog signaling regulates skeletal morphology along the AP axis by promoting cell proliferation and patterning/differentiation of the appendicular skeleton. Sonic hedgehog (Shh) is one of the genes encoding a Hedgehog signaling ligand, and it regulates the number of skeletal components along the AP axis in paired appendages (Riddle et al., 1993). In limbs, which are paired appendages in tetrapods (amphibians and amniotes), and homologous to paired fins, disruption or mutation of a limb-specific Shh enhancer results in digit loss or polydactyly along the AP axis (Lettice et al., 2003, 2017; Sagai et al., 2005). In paired fins, disruption of Shh expression also causes severe defects in paired fin skeletons (Letelier et al., 2018; Neumann et al., 1999). Therefore, it is plausible that Shh regulation along the AP axis is involved in elaboration of pectoral fins in teleosts.
Based on previous reports, we hypothesized that repression of Shh signaling in teleost pectoral fins moderately reduces the number of pectoral fin radials. To investigate whether repression of Shh signaling affects the number of radials, we focused on repressive mediators of the Hedgehog signaling pathway. We generated knockout models of repressive mediators of Hedgehog signaling in zebrafish and studied their development and the changes in their pectoral fin skeletons. We found that overactivation of Hedgehog signaling by mutation of a repressive mediator increased the number of pectoral fin radials. Our results suggest that the AP constraint on Hedgehog signaling by the repressive mediator has reduced pectoral fin radials compared to those of basal species and resulted in pectoral fin elaboration in teleosts.
RESULTS
Overactivation of Hedgehog signaling by hhip mutations induces enlargement of endochondral components in pectoral and other fin skeletons
As candidate genes involved in the repression of Shh signaling, we examined two genes mediating the repression of Hedgehog signaling, hedgehog interacting protein (hhip) and GLI family zinc finger 3 (gli3). Zebrafish with hhip mutations are known to exhibit an increase in the number of fin rays and an enlarged pectoral fin primordia, but the endochondral components in paired fin skeletons have not been characterized (Koudijs et al., 2005; van Eeden et al., 1996). gli3 mutant medaka have been reported to exhibit an increase in the number of paired fin radials (Letelier et al., 2021), but the fin radial phenotype in gli3 mutant zebrafish has not yet been reported. Using the CRISPR-Cas9 genome-editing approach in zebrafish, we introduced a mutation in exon 4 of hhip (Fig. S1A) and in exon 5 of gli3 (Fig. S2A) to replicate previously reported mutants (Koudijs et al., 2005; Letelier et al., 2021). We isolated two hhip mutants (hhipH215RfsX1, hhipF242SfsX8; Fig. S1A) and one gli3 mutant with a 122-bp deletion in exon 5 (gli3Δ; Fig. S2A). hhip−/− zebrafish showed an enlargement of fin skeletons (Fig. S1B-E), while gli3Δ/Δ zebrafish did not show any radial fin abnormalities (Fig. S2B-E). Previous studies on gli3 in zebrafish have also not reported any phenotypes in fin radials (Devine et al., 2009; Wei et al., 2023 preprint). In hhip−/− zebrafish, the endochondral disk of the pectoral fin was elongated along the AP axis compared to the proximal-distal (PD) axis (Fig. S1B-E). Therefore, we focused on the fin skeletons of adult hhip mutants.
In the pectoral fin of hhip mutants, the position of the pectoral girdle was shifted laterally compared to that of wild-type (WT) zebrafish (Fig. 1A,B), and the pectoral fin skeletons of hhip mutants showed an increased number of radials compared to WT zebrafish (Fig. 1C,D; Fig. S1F). The pelvic fin of hhip mutants showed no superficial differences in the endochondral skeletons compared with WT (Fig. 1E,F). The hhip mutation also affected median fin skeletons: dorsal, anal, and caudal (Fig. 1G-L). The dorsal fin of hhip mutants showed an increased number of radials compared to WT zebrafish (Fig. 1G,H; Fig. S1H, Fig. S3A,C). In contrast, the anal fin skeleton was completely absent in hhip mutants (Fig. 1I,J; Fig. S3B,D). In the caudal fin of hhip mutants, the number of hypurals, bones formed ventral to the ural centra and supporting the caudal fin, was increased compared to WT zebrafish (Fig. 1K,L). The number of the hypurals in the region ventral to the hypural diastema, the gap between the second and third hypurals (Fig. 1K,L, black dashed line), increased from two to five (Fig. 1K). These results indicate that hhip−/− zebrafish have an increased number of radials or hypurals in almost all fin skeletons.
Paired and median fin skeletons in hhip−/− zebrafish. (A,B) Ventral views of the pectoral fin skeleton of hhip−/− (A) and WT (B) zebrafish. White dashed lines indicate the pectoral girdles. (C,D) Medial views of the pectoral fin skeleton of hhip−/− (C) and WT (D) zebrafish. White dashed lines indicate the first proximal radial (PR1). (E,F) Pelvic fin skeletons of hhip−/− (E) and WT (F) zebrafish. (G,H) Dorsal fin skeletons of hhip−/− (G) and WT (H) zebrafish. (I,J) Anal fin skeletons of hhip−/− (I) and WT (J) zebrafish. Black brackets indicate the post-anal region where the anal fin is formed. (K,L) Caudal fin skeletons of hhip−/− (K) and WT (L) zebrafish. Black dashed lines represent the hypural diastema, and the white arrowheads indicate hypurals in the region ventral to the hypural diastema. Numbers indicate the first to fifth hypurals. The white asterisk marks the first ural vertebra. Ph, parhypural. All observations were performed on five hhip−/− and five WT fish. Double arrows indicate the anterior (A)-posterior (P) axis and the dorsal (D)-ventral (V) axis. Scale bars: 500 µm (C-F); 1 mm (A,B,G,H,K,L); 2 mm (I,J).
Paired and median fin skeletons in hhip−/− zebrafish. (A,B) Ventral views of the pectoral fin skeleton of hhip−/− (A) and WT (B) zebrafish. White dashed lines indicate the pectoral girdles. (C,D) Medial views of the pectoral fin skeleton of hhip−/− (C) and WT (D) zebrafish. White dashed lines indicate the first proximal radial (PR1). (E,F) Pelvic fin skeletons of hhip−/− (E) and WT (F) zebrafish. (G,H) Dorsal fin skeletons of hhip−/− (G) and WT (H) zebrafish. (I,J) Anal fin skeletons of hhip−/− (I) and WT (J) zebrafish. Black brackets indicate the post-anal region where the anal fin is formed. (K,L) Caudal fin skeletons of hhip−/− (K) and WT (L) zebrafish. Black dashed lines represent the hypural diastema, and the white arrowheads indicate hypurals in the region ventral to the hypural diastema. Numbers indicate the first to fifth hypurals. The white asterisk marks the first ural vertebra. Ph, parhypural. All observations were performed on five hhip−/− and five WT fish. Double arrows indicate the anterior (A)-posterior (P) axis and the dorsal (D)-ventral (V) axis. Scale bars: 500 µm (C-F); 1 mm (A,B,G,H,K,L); 2 mm (I,J).
Next, to identify the role of hhip in fin development, we examined the expression pattern of hhip. The expression pattern of hhip in the early stages has been previously reported: hhip is expressed in the entire pectoral fin bud at 48 h post-fertilization (hpf) (Koudijs et al., 2005). We observed that hhip was expressed only at a low level in the posterior margin of the pectoral fin after 5 days post-fertilization (dpf) (Fig. S4B,C). In median fins, hhip was highly expressed in the fin rays, but not in the endochondral regions (Fig. S4D). In pelvic fin buds, hhip was strongly expressed (Fig. S4E), similar to what was observed in pectoral fin buds at 48 hpf. These results suggest that hhip functions specifically during early development of the paired fins in zebrafish.
We then examined whether Hedgehog signaling is overactivated in hhip mutants. As a marker gene of Hedgehog signaling activity, one of the Hedgehog downstream target genes, ptch1, was selected for the study (Chiang et al., 2001; Matsubara et al., 2016). In WT, ptch1 was expressed in the posterior margin of pectoral fin buds (Fig. S4F, black arrowhead). Interestingly, in hhip mutants, ptch1 expression was not only observed in the posterior region but also expanded to the anterior region (Fig. S4G). Particularly, ptch1 expression was high in the anterior and posterior margins (Fig. S4G, black arrowheads). Therefore, ptch1 expression in hhip mutants expanded more anteriorly than in WT. Considering that phenotypes of hhip−/− zebrafish can be rescued by Hedgehog-repressive treatment (Koudijs et al., 2005), these data indicate that Hedgehog signaling overactivation by hhip mutations induces the enlargement of fin skeletons.
Enlargement of the pectoral skeleton and other fin skeletons in hhip mutants occurs in a region-specific manner
Although hhip−/− zebrafish exhibited an increase in the number of fin skeletal elements, adult fin skeletons were often fused and unrecognizable. To demonstrate how fin radials form and enlarge in detail, we examined fin skeletons during the development of hhip−/− zebrafish (Fig. 2).
Developmental processes of the fin skeletons in hhip−/−zebrafish. (A-F) Pectoral fin development in hhip−/− (A-C) and WT (D-F) zebrafish. Filled and unfilled white arrowheads indicate new and previous subdivisions, respectively, and numbers mark the predicted first to third subdivisions. White dashed lines indicate the first proximal radial (PR1). (G-L) Pelvic fin development in hhip−/− (G-I) and WT (J-L) zebrafish. White arrowheads indicate the radial bones. AnR, anterior large radial; MeR, medial small radial; Pg, pelvic girdle; PoR, posterior elongated radial. (M,N) Caudal fin development in hhip−/− (M) and WT (N) zebrafish. Numbers indicate the first to fifth hypurals. The white asterisk marks the first ural vertebra. Ph, parhypural. These observations were conducted on more than ten larvae at each developmental stage. Green fluorescence indicates the endochondral skeleton marked by col2a1a:EGFP. Double arrows indicate the anterior (A)-posterior (P) axis and the dorsal (D)-ventral (V) axis. The standard length of individuals (in mm) is shown in the bottom left of each panel. Scale bars: 250 µm.
Developmental processes of the fin skeletons in hhip−/−zebrafish. (A-F) Pectoral fin development in hhip−/− (A-C) and WT (D-F) zebrafish. Filled and unfilled white arrowheads indicate new and previous subdivisions, respectively, and numbers mark the predicted first to third subdivisions. White dashed lines indicate the first proximal radial (PR1). (G-L) Pelvic fin development in hhip−/− (G-I) and WT (J-L) zebrafish. White arrowheads indicate the radial bones. AnR, anterior large radial; MeR, medial small radial; Pg, pelvic girdle; PoR, posterior elongated radial. (M,N) Caudal fin development in hhip−/− (M) and WT (N) zebrafish. Numbers indicate the first to fifth hypurals. The white asterisk marks the first ural vertebra. Ph, parhypural. These observations were conducted on more than ten larvae at each developmental stage. Green fluorescence indicates the endochondral skeleton marked by col2a1a:EGFP. Double arrows indicate the anterior (A)-posterior (P) axis and the dorsal (D)-ventral (V) axis. The standard length of individuals (in mm) is shown in the bottom left of each panel. Scale bars: 250 µm.
In the development of WT pectoral fins, radials are formed by the subdivisions of an endochondral disk, a large cartilage plate (Dewit et al., 2011; Grandel and Schulte-Merker, 1998). The first subdivision occurs in the middle of the endochondral disk, and the second and third subdivisions occur in the anterior and posterior halves, respectively (Fig. 2D-F). As a result, there are four radials in the WT pectoral fin skeleton. In hhip−/− zebrafish, the first subdivision shifted posteriorly (Fig. 2A, number 1), and the second subdivision occurred in the anterior third (Fig. 2B, number 2). Subsequently, other subdivisions occurred (Fig. 2C, arrowheads without numbers). The posterior half was further divided only once (Fig. 2C, number 3). These observations indicate that the three featured subdivisions in hhip−/− zebrafish are equivalent to the first to third subdivisions in WT zebrafish. The first subdivision in both WT and hhip−/− zebrafish occurred much earlier than the other subdivisions (Fig. 2A,D), suggesting that these subdivisions are equivalent (Fig. 2D, number 1). The second subdivision in WT formed the first proximal radial (PR1; Fig. 2F, dashed line), which has a curved shape different from the other proximal radials. The second subdivision in hhip−/− zebrafish also formed the PR1 (Fig. 2C, dashed line). Therefore, the second subdivision in hhip−/− zebrafish appears to be equivalent to the second subdivision in WT (Fig. 2E, number 2). Since both WT and hhip−/− zebrafish formed two radials from the posterior half of the endochondral disk by a certain subdivision, the subdivision in hhip−/− zebrafish also appears to be equivalent to the third subdivision in WT (Fig. 2F, number 3). Comparing the subdivision sequence in hhip−/− and WT zebrafish, the position where proximal radials increased in hhip−/− zebrafish appears to be in the anterior-median region, namely the second proximal radial of the WT pectoral fins.
The pelvic and caudal fin skeletons in hhip−/− zebrafish showed abnormalies during development (Fig. 2G-N). In the development of WT pelvic fins, the skeletons possessed three radials (Fig. 2J-L): an anterior large radial (AnR), a medial small radial (MeR), and a posterior elongated radial (PoR). The anterior radials in hhip−/− zebrafish were as small as the MeR in WT, but not as large as the AnR in WT (Fig. 2G-I). In hhip−/− zebrafish, although the radials were fused and indistinguishable (Fig. 1E), the number of radials increased to more than four (Fig. 2I; Fig. S1G). The most posterior radial was equivalent to the PoR in WT, suggesting that the hhip mutation did not affect the PoR (Fig. 2G-I). Based on these observations, the increase in pelvic radials in hhip−/− zebrafish occurred in the anterior region that is equivalent to the AnR and MeR in WT zebrafish. In the development of the WT caudal fin, the parhypural and first hypural were formed ventral to the first ural vertebra (Fig. 2N, asterisk), and the second hypural was generated ventral to the second ural vertebra (Desvignes et al., 2018) (Fig. 2N, number 2). In hhip−/− zebrafish, the primordium of the second hypural was branched, and three hypurals were consequently formed (Fig. 2M, number 2), suggesting that hypurals increased in the ventral region equivalent to the second hypural in WT zebrafish. Therefore, the paired and caudal fins of hhip−/− zebrafish had a region-specific increase of radials or hypurals, respectively.
Duplicated Shh genes in zebrafish show a different expression pattern during paired fin developments
Given the development process of pectoral fins, the region-specific enlargement of the endochondral disk along the AP axis in hhip−/− zebrafish is regulated by the AP regionality derived from Hedgehog signaling. Hedgehog signaling in paired appendages is repressed by antagonization of Hhip (Chuang and McMahon, 1999; Koudijs et al., 2005). In paired appendages of vertebrates, Shh is expressed as Hedgehog ligands antagonized by Hhip. However, the expression patterns of Shh genes in the late development of paired fins when the endochondral disk grows and subdivides, remain unclear. Although zebrafish have two Shh genes, shha and shhb (Fig. S5), their expression has not been reported after 7 dpf. To investigate whether shha and shhb are expressed in the late development of the paired fins of zebrafish, we visualized them using a CRISPR-Cas9 knock-in approach (Auer et al., 2014; Kayo et al., 2023; Ota et al., 2016). We introduced an EGFP reporter in shha and shhb genes (shhaegfp, shhbegfp; Fig. S6) and examined shha and shhb expression patterns in zebrafish (Fig. 3A-N). In the development of zebrafish pectoral fins, both shhaegfp and shhbegfp were expressed in the posterior margin (Fig. 3A-H). However, shhaegfp expression was observed until 4 dpf (Fig. 3A) and disappeared by 7 dpf (Fig. 3B-D), while shhbegfp expression continued beyond 4 dpf (Fig. 3E-H). This indicates that shhb is expressed for a longer period than shha until the subdivisions of the pectoral fin. In pelvic fin development, shhaegfp expression in the posterior margin was short-lived (Fig. 3I-K), while shhbegfp expression in the posterior margin persisted until the initiation of pelvic girdle formation (Fig. 3L-N). Therefore, shhb is predominantly expressed during the late stages of pelvic fin development. In median fin development, neither shhaegfp nor shhbegfp could be detected in the median fin primordia (Fig. S7).
Expression of shhaegfp and shhbegfp in zebrafish, as well as shhaegfp in medaka, during paired fin development. (A-H) EGFP expression in pectoral fin development of shhaegfp (A-D) and shhbegfp (E-H) zebrafish. (I-N) EGFP expression in pelvic fin development of shhaegfp (I-K) and shhbegfp (L-N) zebrafish. (O-R) EGFP expression in pectoral fin development of shhaegfp medaka. (S-V) EGFP expression in pelvic fin development of shhaegfp medaka. All observations were conducted on ten or more larvae at each developmental stage. Magenta fluorescence indicates paired fin skeleton marked by sox10:DsRed. Filled white arrowheads denote EGFP expression and unfilled white arrowheads denote no EGFP expression. Dashed lines outline paired fin buds. Asterisks highlight EGFP expression at the tips of fin rays. White arrows indicate autofluorescence by leucophores. Double arrows mark the anterior (A)-posterior (P) axis. The standard length of individuals (in mm) is shown in the bottom left of each panel. wpf, weeks post-fertilization. Scale bars: 100 µm (A-H); 250 µm (I-V).
Expression of shhaegfp and shhbegfp in zebrafish, as well as shhaegfp in medaka, during paired fin development. (A-H) EGFP expression in pectoral fin development of shhaegfp (A-D) and shhbegfp (E-H) zebrafish. (I-N) EGFP expression in pelvic fin development of shhaegfp (I-K) and shhbegfp (L-N) zebrafish. (O-R) EGFP expression in pectoral fin development of shhaegfp medaka. (S-V) EGFP expression in pelvic fin development of shhaegfp medaka. All observations were conducted on ten or more larvae at each developmental stage. Magenta fluorescence indicates paired fin skeleton marked by sox10:DsRed. Filled white arrowheads denote EGFP expression and unfilled white arrowheads denote no EGFP expression. Dashed lines outline paired fin buds. Asterisks highlight EGFP expression at the tips of fin rays. White arrows indicate autofluorescence by leucophores. Double arrows mark the anterior (A)-posterior (P) axis. The standard length of individuals (in mm) is shown in the bottom left of each panel. wpf, weeks post-fertilization. Scale bars: 100 µm (A-H); 250 µm (I-V).
While basal teleosts, including zebrafish, have both shha and shhb genes, teleosts in Acanthomorpha, including Japanese medaka, lost shhb and retained only shha (Fig. S5). If the shha expression pattern in Japanese medaka is similar to that in zebrafish, teleosts in Acanthomorpha might have lost late developmental Shh gene expression corresponding to that of zebrafish after 7 dpf in paired fins. However, in most teleosts in Acanthomorpha, four radials are conserved in the pectoral fin skeleton, similar to basal teleosts, which have both shha and shhb remaining (Tanaka et al., 2022). To explore how shha is expressed in Acanthomorpha, we generated a knock-in medaka line (shhaegfp; Fig. 3O-V; Fig. S8A). The shhaegfp medaka showed EGFP fluorescence in the posterior margin of the pectoral fin (Fig. 3O-R). In the pectoral fin development of Japanese medaka, shhaegfp was expressed not only in the early stage (Fig. 3O,P) but also in the late stage when proximal radials begin to form (Fig. 3Q,R). In pelvic fin development, shhaegfp was not expressed during the early budding (Fig. 3S,T) but was expressed from the mid-stage to the late stage (Fig. 3U,V). Similar to shhaegfp and shhbegfp zebrafish, shhaegfp medaka also showed no fluorescence in the median fin primordia (Fig. S8B,C). Therefore, the shha expression pattern in Japanese medaka was not equivalent to that of shha in zebrafish, but rather to that of shhb in zebrafish. These results suggest that teleosts largely retain Shh function until the late development of paired fins.
The loss of shhb expression induces defects in paired fin skeletons
The knock-in analysis of Shh genes in zebrafish showed that the expression of shhb lasts longer than that of shha. There is a possibility that shhb expression primarily contributes to the growth of the endochondral disk and the number of radials in the pectoral fin skeleton. To investigate the functions of late Hedgehog signaling, the function of shhb in zebrafish was studied (Fig. 4). Using CRISPR-Cas9 genome editing in zebrafish, we introduced mutations in shhb and identified one shhb mutant that produced a non-functional truncated protein (Fig. S9).
Paired fin skeletons in shhb−/−zebrafish. (A,B) Pectoral fin development in shhb−/− (A; n=7/10) and WT (B; n=10/10) zebrafish. For shhb−/− zebrafish, 7/10 larvae exhibited complete fusion of the third and fourth proximal radials in either the left or right pectoral fins. Some shhb−/− zebrafish exhibited a mild defect in the subdivision between the third and fourth proximal radials (A′,A″). Numbers indicate first to fourth proximal radials. (C,D) Pelvic fin development in shhb−/− (C; n=7/10) and WT (D; n=10/10) zebrafish. For shhb−/− zebrafish, 7/10 larvae exhibited pelvic fin loss, with 6/10 larvae exhibiting complete loss of both left and right pelvic fins, and 1/10 larvae exhibiting complete loss of only the left pelvic fin. The white bracket indicates the pre-anal region where the pelvic fins are formed in WT zebrafish. Green fluorescence indicates the endochondral skeleton marked by col2a1a:EGFP. The standard length of individuals (in mm) is shown in the bottom left of each panel. Scale bars: 250 μm (A-A″,B); 500 μm (C,D).
Paired fin skeletons in shhb−/−zebrafish. (A,B) Pectoral fin development in shhb−/− (A; n=7/10) and WT (B; n=10/10) zebrafish. For shhb−/− zebrafish, 7/10 larvae exhibited complete fusion of the third and fourth proximal radials in either the left or right pectoral fins. Some shhb−/− zebrafish exhibited a mild defect in the subdivision between the third and fourth proximal radials (A′,A″). Numbers indicate first to fourth proximal radials. (C,D) Pelvic fin development in shhb−/− (C; n=7/10) and WT (D; n=10/10) zebrafish. For shhb−/− zebrafish, 7/10 larvae exhibited pelvic fin loss, with 6/10 larvae exhibiting complete loss of both left and right pelvic fins, and 1/10 larvae exhibiting complete loss of only the left pelvic fin. The white bracket indicates the pre-anal region where the pelvic fins are formed in WT zebrafish. Green fluorescence indicates the endochondral skeleton marked by col2a1a:EGFP. The standard length of individuals (in mm) is shown in the bottom left of each panel. Scale bars: 250 μm (A-A″,B); 500 μm (C,D).
In almost all shhb−/− zebrafish, three proximal radials were present in the pectoral fins (Fig. 4A). Some shhb−/− zebrafish showed an incomplete subdivision of the posterior endochondral part of the pectoral fin, where most posterior radials, the third and fourth proximal radials, were fused (Fig. 4A′,A″). Although the shhb−/− phenotype was mild compared to the severe phenotype of the shha complete loss mutants (Neumann et al., 1999), these data suggest that late Hedgehog signaling by shhb also contributes to the number of radials in the pectoral fin skeleton. Interestingly, in almost all shhb mutants, the entire pelvic fin skeleton, including the pelvic girdle, was absent (Fig. 4C). A similar loss of the pelvic fin skeleton has been reported in Japanese medaka without the shha fin enhancer (Letelier et al., 2021). This indicates that the shhb mutation in zebrafish leads to the loss of Hedgehog signaling, resulting in the absence of pelvic fin bud initiation. In zebrafish pelvic fin development, shha was rarely expressed while shhb expression dominated (Fig. 3I-N). These results suggest that shhb constitutes the entirety of Shh gene expression in the pelvic fin of zebrafish, and shha cannot compensate for the absence of shhb. Therefore, it was considered that the expression of shhb, rather than shha, regulates the development of the pelvic fin. Taken together, we conclude that shhb in zebrafish contributes to pectoral fin development along with shha and dominates pelvic fin development independently.
Early expression of hhip is a characteristic of the pectoral fins of zebrafish
The hhip and shh analyses indicated that hhip expression during early development, when both duplicated Shh genes were expressed, regulated pectoral fin elaboration in teleosts. However, it is unclear whether early hhip expression is characteristic of teleosts. To compare hhip expression patterns in the pectoral fins of zebrafish with those of basal species, we used previously published RNA-sequencing (RNA-seq) data of pectoral fins in the brownbanded bamboo shark (Onimaru et al., 2021), bowfin (Thompson et al., 2021), and zebrafish (Kudoh et al., 2024). For this analysis, we re-analyzed the RNA-seq data using the same pipeline (Patro et al., 2017). We selected hhip-related genes (hhip, ptch1, ptch2, and shh), and actinodin (and) genes for comparison of the developmental process (Fig. 5; Fig. S10). ptch1 and ptch2 encode Hedgehog signaling receptors that compete with Hhip. and encodes a structural protein specific to fin folds, a structure covering the outer margin of the endochondral disk and forming fin rays (Zhang et al., 2010). We used and genes to distinguish the stages before and after the beginning of fin fold formation in each species (Fig. 5; Fig. S10, bottom row). The expression level (transcripts per million, TPM) of hhip in zebrafish was higher than that of bowfin, implying that hhip expression increases from basal actinopterygians to teleosts (Fig. 5). Additionally, for inter-species comparison of hhip expression, we scaled the hhip TPM by the sum of ptch1 and ptch2 TPM (Fig. S10, top row). After fin fold formation, the scaled hhip values for each species were almost the same. Before fin fold formation, however, the scaled value of zebrafish was higher than after fin fold formation, while those of the brownbanded bamboo shark and bowfin were lower than after fin fold formation. Taken together, the significant expression of hhip before fin fold formation was characteristic of pectoral fin development in zebrafish.
Expression patterns of hhip-related genes in pectoral fins. Bar graphs displaying expression levels (transcripts per million, TPM) at each developmental stage. The TPM of hhip, ptch, shh and and are in order from the top row. The and TPM distinguishes when fin fold formation began [bowfin, from stage (st) 26; zebrafish, from 5 dpf]. Each dot on the graphs represents the value of each replicate. The RNA-seq data of pectoral fins were obtained from previously published bowfin (Thompson et al., 2021) and zebrafish (Kudoh et al., 2024) studies.
Expression patterns of hhip-related genes in pectoral fins. Bar graphs displaying expression levels (transcripts per million, TPM) at each developmental stage. The TPM of hhip, ptch, shh and and are in order from the top row. The and TPM distinguishes when fin fold formation began [bowfin, from stage (st) 26; zebrafish, from 5 dpf]. Each dot on the graphs represents the value of each replicate. The RNA-seq data of pectoral fins were obtained from previously published bowfin (Thompson et al., 2021) and zebrafish (Kudoh et al., 2024) studies.
The number of radials in the median region modulates the width of pectoral fins in fishes along the AP axis
We then examined whether the pectoral fin skeleton of hhip−/− zebrafish is comparable to those of basal species. Morphologically, the numerous radials in the pectoral fin of basal species are derived from the posterior region because they branch off from the metapterygium, the posterior tribasal bone (Gegenbaur, 1865; Hawkins et al., 2021; Mabee, 2000; Onimaru et al., 2015; Tanaka, 2018). As a result, teleosts have been considered to have lost the posterior region and radials, leading to pectoral fin elaboration. Here, we identified a genetic AP regionality in teleosts and examined the region involved in pectoral fin elaboration.
To identify the AP axis genetically, we visualized the expression patterns of the hoxa13 and hoxd13 genes in zebrafish. These genes are expressed along the AP axis in some chondrichthyans, basal actinopterygians, and teleosts (Ahn and Ho, 2008; Davis et al., 2007; Freitas et al., 2007; Sordino et al., 1995; Tulenko et al., 2016). We generated CRISPR-Cas9 knock-in zebrafish lines (hoxa13begfp, hoxd13aegfp; Fig. S11). Prior to the subdivision of the endochondral disk in the pectoral fin, hoxa13begfp was initially expressed in the distal margin (Fig. 6A), and later its expression was limited to the posterior region of the endochondral disk (Fig. 6B). Similarly, hoxd13aegfp was expressed in the posterior region of the endochondral disk (Fig. 6G). As the pectoral fin subdivided, hoxa13begfp was expressed on the distal margin covering the second, third, and fourth proximal radials (Fig. 6C), while hoxd13aegfp was expressed on the distal margin covering the third and fourth proximal radials (Fig. 6H). In hhip−/− zebrafish, hoxa13begfp was expressed in the distal margin, but EGFP was highly expressed in posterior cells compared to WT zebrafish (Fig. 6A,D, arrowheads). Throughout the subdivision, hoxa13begfp was expressed not only on the distal margin of the most posterior radials equivalent to the third and fourth radials in WT zebrafish but also on the distal margin of the increased radials equivalent to the second radial in WT zebrafish (Fig. 6E,F). These results suggest that the pectoral fin of zebrafish is regionalized by distinct patterns of hoxa13 and hoxd13 expression, and that the number of radials in hhip mutants increases in the median region where hoxa13b is expressed.
Expression of EGFP by hoxa13begfp and hoxd13aegfp in pectoral fin development of zebrafish. (A-F) EGFP expression induced by hoxa13begfp in pectoral fin development of WT (A-C) and hhip−/− (D-F) zebrafish. (G,H) EGFP expression induced by hoxd13aegfp in pectoral fin development of WT. Numbers indicate the first to fourth proximal radials. White arrowheads indicate significant expression of EGFP by hoxa13begfp. White bars represent the range of EGFP expression. White asterisks denote EGFP expression by col2a1a:egfp on the distal radials. Observations were conducted on more than five larvae at each developmental stage. Magenta fluorescence indicates the paired fin skeleton marked by sox10:DsRed. Double arrows indicate the anterior (A)-posterior (P) axis. The standard length of individuals (in mm) is shown in the top right of each panel. Scale bars: 250 µm.
Expression of EGFP by hoxa13begfp and hoxd13aegfp in pectoral fin development of zebrafish. (A-F) EGFP expression induced by hoxa13begfp in pectoral fin development of WT (A-C) and hhip−/− (D-F) zebrafish. (G,H) EGFP expression induced by hoxd13aegfp in pectoral fin development of WT. Numbers indicate the first to fourth proximal radials. White arrowheads indicate significant expression of EGFP by hoxa13begfp. White bars represent the range of EGFP expression. White asterisks denote EGFP expression by col2a1a:egfp on the distal radials. Observations were conducted on more than five larvae at each developmental stage. Magenta fluorescence indicates the paired fin skeleton marked by sox10:DsRed. Double arrows indicate the anterior (A)-posterior (P) axis. The standard length of individuals (in mm) is shown in the top right of each panel. Scale bars: 250 µm.
During development of the pelvic fins in zebrafish, hoxa13begfp was expressed throughout the entire distal region (Fig. S12A-C), while hoxd13aegfp was expressed specifically in the posterior region where the PoR was formed (Fig. S12D,E). In the caudal fin primordium, hoxa13begfp was evenly expressed until 2 dpf (Fig. S12F), but then the median expression, including the second and third hypurals, was lost (Fig. S12G,H), whereas hoxd13aegfp was only expressed in blood vessels and not in the caudal fin primordium (Fig. S12I-K). This suggests that both the pelvic and caudal fins of zebrafish exhibit regionality based on hoxa13 or hoxd13 expression.
DISCUSSION
Our genetic and developmental analyses in zebrafish demonstrated the detailed growth pattern of the endochondral disk of the pectoral fin along the AP axis. In hhip mutants, the endochondral disk expanded along the AP axis, and the number of radials increased (Fig. 1C; Fig. S1F). hhip was expressed in the entire pectoral fin bud during early development, but only in the posterior margin during late development (Fig. S4A-E; Koudijs et al., 2005). This pattern was associated with the early excessive growth of the endochondral disk in hhip mutants until approximately 7 dpf (standard length=5.0 mm; Fig. S1B-E). In contrast, shhb mutation induced a decrease in Hedgehog signaling during both early and late development, resulting in a defect in pectoral fin radials (Fig. 4A). From these results, we suggest that the endochondral disk has two growth phases: allometric growth during early development, and isometric growth during late development (Fig. 7). During early development, the ratio of the AP axis to the PD axis (AP-PD ratio) in the endochondral disk can change, resulting in allometric growth (Fig. 7). Therefore, the number of pectoral fin radials is largely determined by the early growth of the endochondral disk. Conversely, late growth of the endochondral disk may result in an isometric AP-PD ratio. Although the number of radials is largely dependent on the early allometric growth of the endochondral disk, the Hedgehog signaling-promotive growth is buffered by hhip negative feedback. Both shha and shhb were expressed in the pectoral fins of zebrafish, while only shha was expressed in Japanese medaka (Fig. 3). Nevertheless, both fishes have the same number of radials in their pectoral fin skeletons. Although other differences, such as the activities of receptors or regulatory enhancers, may also regulate the strength of Hedgehog signaling in detail (Letelier et al., 2018; Lettice et al., 2017; Lopez-Rios et al., 2014), we suggest that the negative feedback effect of hhip buffers the different strengths of Hedgehog signaling to a certain level, generating four radials across teleosts. The inter-species analysis between zebrafish and basal species showed that the early hhip expression is characteristic of zebrafish (Fig. S10). Therefore, the varying number of pectoral fin radials in basal species may reflect non-buffered Hedgehog signaling activity. In summary, the data suggest that pectoral fin elaboration in teleosts depends on hhip-mediated negative feedback of Hedgehog signaling during early development.
Summary of pectoral radial changes in zebrafish. Diagram illustrating the pectoral fin skeleton at each developmental stage. The length under the dpf stage indicates approximate standard length. The color bars with gene names indicate gene expression periods. Colors in the 28 dpf section indicate the anterior region where neither Hoxa13 nor Hoxd13 is expressed (red), the median region where Hoxa13 is expressed but Hoxd13 is not expressed (orange), and the posterior region where both Hoxa13 and Hoxd13 are expressed (green).
Summary of pectoral radial changes in zebrafish. Diagram illustrating the pectoral fin skeleton at each developmental stage. The length under the dpf stage indicates approximate standard length. The color bars with gene names indicate gene expression periods. Colors in the 28 dpf section indicate the anterior region where neither Hoxa13 nor Hoxd13 is expressed (red), the median region where Hoxa13 is expressed but Hoxd13 is not expressed (orange), and the posterior region where both Hoxa13 and Hoxd13 are expressed (green).
The phenotypes in hhip−/− zebrafish support a potential developmental mechanism in the evolution of paired fins from the ancestral form to the teleost-specific derived form. First, based on observation of the developmental process (Fig. 2A-F), the increase in radials in the pectoral fins of hhip−/− zebrafish occurred in the median region occupied by the second proximal radial, rather than in the posterior region. During the development of pectoral fins in chondrichthyans and basal actinopterygians, the anterior (Fig. S13A, red) and posterior basal radials (Fig. S13A, green) are formed first, followed by the other median radials (Fig. S13A, orange), which fill in the gap between the anterior and posterior radials (Davis et al., 2004; Dillman and Hilton, 2015; Mabee and Noordsy, 2004; Onimaru et al., 2016). The subdivision order of the endochondral disk is similar between basal species and hhip−/− zebrafish, with the radials in the median region being the last to form (Fig. 2C). Second, Hox gene expression supports the AP regionality based on endochondral subdivisions during pectoral fin development in hhip−/− zebrafish (Fig. 6). Hoxa13 is expressed in the distal margin of the median and posterior radials during late development of paddlefish (Davis et al., 2007). Hoxd13 is expressed in the posterior region of the pectoral fins in catshark and paddlefish (Davis et al., 2007; Freitas et al., 2007; Tulenko et al., 2016). These Hox expression patterns correspond to those in zebrafish (Fig. 6; Fig. S12A-E, Fig. S13A). Overall, the expression of Hoxa13 and Hoxd13 determines the anterior (Fig. 7, red; neither Hoxa13 nor Hoxd13 are expressed), median (Fig. 7, orange; Hoxa13 is expressed but Hoxd13 is not expressed), and posterior (Fig. 7, green; both Hoxa13 and Hoxd13 are expressed) regions in the pectoral fin skeleton. EGFP expression induced by hoxa13begfp in hhip−/− zebrafish showed anterior expansion, eventually reaching the posterior border of the first proximal radials, supporting the above regional distinctions (Fig. 6F). Lastly, our expression and mutation analyses highlighted the importance of the median region for pectoral fin elaboration. ptch1 was expressed only in the posterior margin in WT (Fig. S4F), while ptch1 expression reached both the median and posterior regions in hhip−/− zebrafish (Fig. S4G). Based on these expression patterns, Hedgehog signaling affected only the posterior region in shhb−/− zebrafish, which exhibited fusion of posterior radials, and affected both the median and posterior regions in hhip−/− zebrafish, which had excessive radials in the median region. We suggest that the median region is susceptible to Hedgehog signaling, which can increase the number of radials response to the signaling. By contrast, the posterior region cannot increase radials but requires a certain degree of Hedgehog signaling to generate normal third and fourth radials. Although it has been believed that the pectoral fins of teleosts lost posterior radials branched from the metapterygium in the ancestral form and decreased to only four radials, we suggest that pectoral fin elaboration along the AP axis occurs through the reduction of radials in the median Hedgehog-susceptible region (Fig. 7, orange). In addition, pelvic fins also exhibited an increase in radials in hhip−/− zebrafish (Fig. 2I; Fig. S13B). The pelvic fin skeletons of chondrichthyans, sturgeon, and paddlefish possess multiple radials, while there are fewer radials in the pelvic fin skeletons of polypterus, bowfin, and teleosts (Grandel and Schulte-Merker, 1998; Hilton, 2011; Jarvik, 1980; Mabee and Noordsy, 2004; Molnar et al., 2017). Therefore, Hedgehog signaling may be involved in variation of the number of paired fin radials.
These patterns can be extended to the morphogenesis of the caudal fin in teleosts. In the caudal fin skeleton of teleosts, the number of hypurals is decreased in parallel with the elaboration of paired fins (Desvignes et al., 2018). In hhip−/− zebrafish, hypural 2 was branched and three hypurals were formed (Fig. 2M), suggesting that caudal fin morphogenesis is regulated by axis formation similar to the AP regulation for paired fins. alx4, an anterior-specific gene in the developing paired fins, is expressed at both the dorsal and ventral margins during caudal fin development, suggesting that the caudal fin possesses bipolar anterior regulation (Desvignes et al., 2022). Additionally, gdf5 mutation in zebrafish, which affects the formation of posterior radials in pectoral fins, disrupts the formation of the dorsal hypurals 3-5, suggesting that the dorsal region of the caudal fin may correspond to the posterior region of pectoral fins (Waldmann et al., 2021). These reports on molecular markers suggest that the dorsal-ventral axis in the caudal fin is regulated by a mechanism corresponding to the AP axis in paired fins (Fig. S13B). Previous studies have identified various phenotypes in the caudal fin skeleton (Abe et al., 2014; Moriyama et al., 2012), but our results are the first to report Hedgehog-susceptible and region-specific phenotypes. Although neither shha nor shhb expression was detected in the endochondral primordium of the caudal fin skeleton (Fig. S7A,B,E,F), Shh ligands secreted by other organizers such as the notochord may regulate hypural development (Hadzhiev et al., 2007). We suggest that Hedgehog susceptibility in the caudal fin is involved in the morphological evolution of the caudal fin skeleton. The number of hypurals decreased in the morphological change from the heterocercal form of chondrichthyans and basal actinopterygians to the homocercal form of teleosts (Desvignes et al., 2018), and this morphological change occurred approximately concurrently with paired fin elaboration. Therefore, repression of Hedgehog signaling might affect not only paired fin evolution but also caudal fin evolution from the ancestral form to the teleost form. However, the caudal fin skeletons in basal species have more hypurals than those in hhip−/− zebrafish (Cumplido et al., 2024). An increase in the number of ural vertebrae itself will be required to replicate the complete pattern of the caudal fin skeleton in basal species. Additional mutations modulating axial elongation (Aires et al., 2019; Cumplido et al., 2024; Kingsley et al., 2024) may generate sufficient ural vertebrae and hypurals to obtain a caudal fin skeleton of teleosts that is closer to that of basal species.
In hhip−/− zebrafish, the dorsal fin radials increased, but the anal fin skeleton was completely absent (Fig. 1G-K). The opposite phenotypes in the median fin skeletons seem inexplicable, but the function of hhip on the dorsal-ventral regionalization in the somites may resolve the inconsistency. hhip is expressed in the horizontal boundary cells of somites in teleosts and regulates the dorsal-ventral boundary of somites (Fig. S4A; Abe et al., 2019). Mutation of hhip in Japanese medaka expands the expression of zic1, a dorsal marker of somites, largely into the ventral part of somites (Abe et al., 2019). Since the somite cells contribute to the median fins as skeletal components, the median fin phenotypes presented here may be induced by the dorsalized somites, giving rise to an increase in dorsal fin components and a loss of the ventral ones that differentiate into anal fin skeletons. It is possible that regulation within somites might be as important as that within fin buds for median fin development, although the effect of Shh within median fin buds has been the focus of studies so far (Dahn et al., 2007; Höch et al., 2021; Letelier et al., 2018, 2021).
Shh regulation in paired fin development differs in several aspects between basal teleosts (such as zebrafish) and Acanthomorpha teleosts (such as medaka). The first difference lies in the mechanism underlying the repression of Hedgehog signaling. Our results showed that hhip mutant zebrafish exhibited an increase in radials, while gli3 mutant zebrafish exhibited no abnormalities in the fin skeletons (Fig. S2). In contrast, a previous study demonstrated that gli3 mutant medaka exhibited an increase in radials in the paired and dorsal fin skeletons (Letelier et al., 2021). We hypothesized that the genes involved in the repression mechanism of Hedgehog signaling may vary while still maintaining the number of radials. Knockout of hhip in Japanese medaka will help elucidate the generalizability and variability in Hedgehog repression that regulates pectoral fin development. The second difference lies in how Shh gene expression is regulated between zebrafish and Japanese medaka (Fig. 3). In zebrafish, both shha and shhb regulate paired fin development (Fig. 3A-N); however, only shha regulates development in medaka (Fig. 3O-V). The expression of Shh genes at the posterior margin of paired fins is regulated by a single cis-regulatory element called the ZPA regulatory sequence (ZRS) (Letelier et al., 2018; Lettice et al., 2003; Sagai et al., 2005). The ZRS is located in intron 5 of lmbr1, a gene located on the same chromosome as Shh genes, and is common among gnathostome genomes. In basal teleosts, the lmbr1 paralog has been conserved in the genomic region upstream of shha (Fig. S5). Conversely, the genomic region upstream of shhb has lost the lmbr1 paralog and some neighboring genes, leading to a gene desert. Considering that shhb in the paired fins of zebrafish was expressed for a similar duration as shha in medaka, it is possible that the genomic region upstream of shhb in zebrafish contains a conserved paired fin enhancer equivalent to the ZRS located within the lmbr1 intron. However, the expression of shha in the paired fins of zebrafish has likely been modulated to be shorter than that of shhb in zebrafish and shha in Japanese medaka. In summary, the regulation system of Hedgehog signaling during fin development varies even between zebrafish and Japanese medaka. In this study, although we suggest that fin elaboration in zebrafish is regulated by negative feedback of hhip, the developmental regulation of Hedgehog signaling appears to have diversified among teleosts. Further genetic analyses in other teleosts will help uncover the precise evolutionary mechanism of fin skeleton development.
MATERIALS AND METHODS
Fish care and strains
The following transgenic zebrafish lines were used in this study: col2a1a:EGFP (Hamada et al., 2019) and sox10:DsRed (Yoshida et al., 2020). Zebrafish and medaka were housed at 28°C with a 14 h light cycle. The standard length of individuals was used as an indicator of individual body size instead of the date of development since zebrafish of the same age often have different body sizes (Parichy et al., 2009). The hhip−/− zebrafish had two different mutations (H215RfsX1, F242SfsX8). Homozygous alleles of both mutations (H215RfsX1/H215RfsX1, F242SfsX8/F242SfsX8) and compound heterozygous alleles (H215RfsX1/ F242SfsX8) induced the hhip mutant phenotypes. All shhb−/− zebrafish that we observed possessed homozygous alleles (G122LfsX1/G122LfsX1).
Preparation of sgRNAs
All single guide RNAs (sgRNAs) were designed using CRISPRscan (Moreno-Mateos et al., 2015) and CHOPCHOP (Labun et al., 2019) CRISPR online tools. Template DNA for sgRNA synthesis was PCR-amplified using the crRNA/tracrRNA sequence primer AAAAGCACCGACTCGGTGCCACTTTTTCAAGTTGATAACGGACTAGCCTTATTTTAACTTGCTATTTCTAGCTCTAAAAC with the forward primer AAAAGCACCGACTCGGTGCC and the reverse primer TAATACGACTCACTATAggxxxxxxxxxxxxxxxxxxGTTTTAGAGCTAGAAATAGCA (for T7 polymerase). Lowercase letters indicate genome-targeting sequences (either 19 or 20 bases) in sgRNAs. The genome-targeting sequences in sgRNAs used in this study are shown in Table S1. After PCR amplification with KOD -plus- neo polymerase (Toyobo), PCR products were purified using a PCR purification kit (Cica). The obtained template DNA was used for in vitro transcription of sgRNAs using a CUGA®7 gRNA synthesis kit (Nippon Gene). sgRNAs were purified using the CUGA®7 gRNA synthesis kit (Nippon Gene).
Microinjection for mutagenesis and knock-in
sgRNAs and Cas9 nuclease were co-injected into one-cell-stage zebrafish and medaka embryos. Each embryo received an injection of a solution containing 10 ng/μl of sgRNA for digesting genomic DNA and 250 ng/μl of Cas9 nuclease (Integrated DNA Technologies). For knock-in samples, we also added 10 ng/μl of sgRNA for digesting BaitD and 7.5 ng/μl of phenol-chloroform to extract purified plasmid. To minimize leaky EGFP expression, we used the pUC-BaitD-Xhbb-EGFP plasmid, known for stable EGFP expression in target tissues (Kayo et al., 2023). The injection volume was adjusted to result in the death of approximately 50% of injected embryos within 1 week after injection. These injection mixtures were introduced into one-cell-stage eggs, following established protocols for zebrafish (Kawakami et al., 2016) and medaka (Kinoshita, 2009).
Genotyping and sequencing of mutant fish
The genotype of mutant fish was determined using a heteroduplex mobility assay (HMA) (Ansai et al., 2014). The sequences of mutant alleles isolated from F1 or later generations were determined through direct Sanger sequencing of the PCR products using the primer pairs shown in Table S2. All mutations present in these mutants are listed in Table S3.
Genotyping and sequencing of knock-in fish
For knock-in insertion mapping, we collected fluorescent F1 fish at 2-4 dpf and extracted genomic DNA using standard protocols. The insertion status was examined from either the 5′ side or the 3′ side of the insertion. For example, when checking from the 5′ side of the insertion, a PCR reaction was performed using a 5′ primer specific to each target gene (upstream of the expected insertion site) and a 3′ primer specific to the donor plasmid, pUC-BaitD-Xhbb-EGFP. The primers used for insertion mapping in this study are shown in Table S2. The ‘knock-in check’ primers were designed on the 5′ side (fw) and the 3′ side (rv) of the donor plasmid, respectively. The sequences of mutant alleles isolated from F1 or later generations were determined by direct Sanger sequencing of the PCR products.
Bone staining with Alizarin Red
Vital bone staining was performed with Alizarin Red S (AR) solution in water, which was prepared before use, as previously described with slight modifications (Yoshida et al., 2020). In this study, zebrafish were transferred to the AR solution and incubated for 2 h at 28.5°C. Following bone staining, the zebrafish were washed several times with system water and examined under a Leica M205 FA stereomicroscope. Images were taken using a Leica DFC 360 FX camera.
Whole-mount in situ hybridization
Probe sequences for in situ hybridization were cloned into the pBluescript-II-SK(+) vector using PCR from zebrafish cDNA and the In-Fusion® HD Cloning Kit (Takara Bio Inc.). PCR primers are listed in Table S4. The zebrafish hhip probe had been previously used in research (Koudijs et al., 2005). In situ hybridization was performed following the protocol outlined in a previous study (Woltering et al., 2009) with a slight modification of the number of washes increasing one or more times during every washing process.
Phylogenetic analysis
To infer individual gene family trees, amino acid sequences were retrieved from aLeaves (Kuraku et al., 2013) and manually curated. Multiple sequence alignment was performed using MAFFT (v.7.525) (Katoh et al., 2002) with the option ‘-linsi’. The aligned sequence sets were then processed using trimAl (v.1.4.1) (Capella-Gutiérrez et al., 2009) with the option ‘-automated1’. Molecular phylogenetic trees were inferred using RAxML (v.8.2.12) (Stamatakis, 2014) with the ‘-m PROTGAMMAAUTO -f a -# 1000’ options.
RNA-seq data analysis
The RNA-seq re-analyses of pectoral fins were performed using the publicly available RNA-seq data of the brownbanded bamboo shark (Onimaru et al., 2021), bowfin (Thompson et al., 2021), and zebrafish (Kudoh et al., 2024). The Salmon pipeline was used for transcriptome assembly and quantification (Patro et al., 2017). The transcript-level counts were then summarized to gene-level counts using tximport v.1.30.0 (Soneson et al., 2015). Reference transcriptome data for the brownbanded bamboo shark (https://github.com/Squalomix/sequences), bowfin (https://github.com/AndrewWT/AmiaGenomics), and zebrafish (Ensembl GRCz11 cDNA; https://ftp.ensembl.org/pub/release-111/fasta/danio_rerio/cdna/) were employed. To identify non-annotated genes in the brownbanded bamboo shark and bowfin, BLASTN v.2.15.0 was performed against these reference transcriptome data. For the brownbanded bamboo shark, coding gene sequences from the elephant shark (Callorhinchus_milii-6.1.3) were used, and for the bowfin, sequences from the spotted gar (LepOcu1) were used.
Fluorescence imaging
Zebrafish and medaka were anesthetized by immersion in 0.025% MS222 and placed in 3% agarose gel/E3 on a glass slide. All fluorescence images were taken using a Leica M205 FA microscope system and photographed with a Leica DFC 369 FX camera.
Statistical analysis
The number of skeletal components and the length of pectoral fin primordia were measured at specific time points during the developmental process using col2a1a:EGFP fluorescence and Leica Application Suite X (LAS X, Leica). Differences in the number of skeletal elements and the length between mutant and WT zebrafish were tested using an unpaired t-test in GraphPad Prism software.
Acknowledgements
We appreciated the comments from Tamura lab members.
Footnotes
Author contributions
Conceptualization: Y.T., G.A., K.T.; Methodology: Y.T., S.A., G.A., K.T.; Validation: Y.T., S.O., K.U., K.T.; Formal analysis: Y.T.; Investigation: Y.T., S.O., K.U., H.K., S.A.; Resources: Y.T., S.A., G.A.; Writing - original draft: Y.T., G.A., K.T.; Writing - review & editing: Y.T., G.A., K.T.; Visualization: Y.T.; Supervision: Y.T., G.A., K.T.; Project administration: Y.T., K.T.; Funding acquisition: Y.T., G.A., K.T.
Funding
This work was supported by Grants-in-Aid for Scientific Research (KAKENHI) from the Japan Society for the Promotion of Science (JSPS) (JP20J21314 and JP23K19384 to Y.T.; JP22K06232 and JP20H04854 to G.A.; JP23H04301, JP22H02627, JP21H05768, JP21K19202, and JP20H05024 to K.T.), and Takeda Science Foundation Life Science Research Grants (2022036015 to G.A.). Open Access funding provided by Okayama University. Deposited in PMC for immediate release.
Data availability
All relevant data can be found within the article and its supplementary information.
Peer review history
The peer review history is available online at https://journals.biologists.com/dev/lookup/doi/10.1242/dev.202526.reviewer-comments.pdf
References
Competing interests
The authors declare no competing or financial interests.