The timing of dorsal fin development is different from LMFF reduction.
We first examined the initiation process of dorsal fin formation in the LMFF (Fig. 1). In 4.2 mm standard length (SL) zebrafish, the continuous LMFF looks to have no protrusions or outgrowth that imply a fin primordial structure at the dorsal, caudal, and ventral midline (Fig. 1a, a’). When the larval size reaches 5.6 mm SL, a small dorsal area in the LMFF began outgrowth (white arrowhead in Fig. 1b’). In 6.3 mm SL zebrafish, the outgrowth site continues to grow distally and expand along the rostral-caudal axis (white arrowhead in Fig. 1c’). By 7.2 mm SL, the dorsal fin primordium with fin rays is visible. In 7.2 mm SL zebrafish, the height of the LMFF looks reduced from the anterior side both before and behind the dorsal fin; thus, the dorsal and caudal fins separated to form independent structures (Fig. 1d, d’). Taken together, we estimated that the dorsal fin outgrowth at 5.6 mm SL may be the initial dorsal fin primordium.
To examine the detailed process of dorsal fin formation and LMFF reduction, we quantitatively analyzed the height of the LMFF every five days in the transition process from 5 days post-fertilization (dpf) to 20 dpf (n=6, Fig. 1e–h). We measured the height of the LMFF at the future dorsal fin-appearing position (red lines in Fig. 1e, f) and at the LMFF-disappearing position behind the dorsal fin (blue lines in Fig. 1e, f). The height of the LMFF at the dorsal fin position increased moderately as a protrusion between 4.5–5.0 mm SL and then started to increase rapidly for outgrowth at around 5.0 mm SL (red dots and lines in Fig. 1g, h). The rapid increase continued until 7.0 mm SL. In contrast, the height of the LMFF at the presumptive disappearing position behind the dorsal fin was almost constant between 4.0–5.5 mm SL (blue dots and lines in Fig. 1g, h). Then, the height of that region started decreasing at around 5.5 mm SL and later decreased rapidly until 7.0 mm SL.
These data suggest that the LMFF protrusion and outgrowth at the future site of the dorsal fin precedes the reduction of the LMFF. In other words, the emergence of the dorsal fin may not be due to the process of LMFF reduction, but rather depends on its protrusion.
Signals for apoptotic cell death could not be detected in the reduction process of the LMFF.
Since dorsal fin appearance precedes the LMFF reduction, this raises the question of whether cell death in the LMFF provokes dorsal fin formation. We therefore investigated cell death in the process of the LMFF reduction. First, we performed acridine orange staining, which identifies cell death in living specimens28,29, at the LMFF-reducing stage (6.0–6.5 mm SL, n = 6; 6.5–7.0 mm SL, n = 6; 7.0–7.5 mm SL, n = 5) (Fig. 2a–c”). We found no obvious signal for cell death in the reducing LMFF area in any samples (Fig. 2a, a’, b, b’, c, c’). We detected some signals in different areas such as at the base of the caudal fin, indicating that our experimental procedure for the staining was fine. We further performed whole-mount immunohistochemistry with an anti-active caspase antibody at the LMFF-reducing stage (6.0–6.5 mm SL, n=6; 6.5–7.0 mm SL, n=5; 7.0–7.5 mm SL, n=5) (Fig. 2d–f”)30. We detected some positive cells at the base of the caudal fin (Fig. 2d’’, e’’, f’’), indicating that the immunohistochemistry detected dying cells correctly. We found very few positive signals in the reducing LMFF area (Fig. 2d’, e’, f’) in all samples.
These data show that cell death did not occur in the reducing LMFF area. In conclusion, cell death may not play a role in LMFF reduction and median fin segregation.
Cell shape change and cell migration in LMFF reduction.
Our above results suggest that cellular behaviors other than cell death may play a role in the process of the LMFF reduction. We focused on epithelial cell migration and cell shape change in LMFF reduction and performed in vivo cell-tracking analysis with an epidermal cell-specific cre-expressing vector (krt8-p:cre)31. We injected the krt8-p:cre vector into Tg(actbp-loxP-DsRed-loxP-EGFP)32 embryos (Fig. 3a) and observed EGFP-positive epidermal cells that were distributed as mosaic patches in the LMFF of the injected specimens (Fig. 3c–f’). We traced the epidermal cell behavior from the stages when the LMFF reduction started (12 dpf, 5.8–6.1 mm SL, n = 3), and specimens were observed every two days until 16 dpf, when their size was around 7.0 mm SL (Fig. 3b). Figure 3 shows a specimen in which two GFP-positive populations of epidermal cells can be detected in the reducing LMFF (yellow dotted area in Fig. 3c). During the observation period, GFP-positive populations became narrow along the proximal-distal axis by shape change, in which each cell began proximo-distal shrinkage (magenta bracket in Fig. 3d’, e’, f’) and migrated down to the trunk (magenta arrowheads in Fig. 3d’, e’, f’).
These results indicate that epidermal cells collectively deform proximo-distally and migrate from the LMFF to the body trunk, suggesting that deformation and migration of LMFF epidermal cells contributes to the process of LMFF reduction.
Mesenchyme growth in the dorsal fin primordium Our morphological observations (Fig. 1) also suggest that the protrusion and outgrowth of the LMFF is a key process in dorsal fin formation, rather than the LMFF reduction process. Previous studies have shown that somite-derived mesenchymal cells are condensed in the future site of the dorsal fin, and these mesenchymal cells develop into dorsal fin skeletal elements18, 27. These studies, however, have not shown when mesenchymal cells emerge or the differentiation process of the mesenchyme. Thus, we investigated mesenchymal cell behavior during the formation of the dorsal fin primordium by analyzing a reporter transgenic fish line.
We observed reporter expression in the gt1116A transgenic line (Fig. 4). In gt1116A, the gene-trapping gal4 construct was integrated within the prdm16 gene, and the UAS:EGFP reporter was found to be expressed in mesenchymal cell populations of the early pectoral fin bud33,34. In addition, UAS:EGFP in gt1116A has been reported to be expressed in other developing median fins, including the dorsal fin34. We first examined whether EGFP expression patterns of gt1116A are valid as a reporter of the fin mesenchyme in the developing dorsal fin primordium. The reporter EGFP was detected in median fin formation at the stage when the LMFF of the future dorsal fin site starts protruding (Fig. 4a; SL = 5.0–5.5 mm). Transverse sections showed UAS:EGFP expression distributed in cells of the middle at the LMFF protrusion, sandwiched by two outer layers of cells (presumably epidermal layers) where no signal was detected (Fig. 4b–b’’). These data confirmed that UAS:EGFP in the gt1116A line is expressed in the mesenchymal cells of the dorsal fin primordium.
We next assessed the expression pattern of the UAS:EGFP reporter in the gt1116A line in the transition from LMFF to dorsal fin primordium while observing chondrocytes by sox10:DsRed (4.0–4.5 mm SL, n = 6; 6.5–7.0 mm SL, n= 5; 7.0–7.5 mm SL, n = 5) (Fig. 4c–f’’ ). In 4.0–4.5 mm SL zebrafish, which are before the protrusion of the LMFF (Fig. 1h), no EGFP-positive cells were observed in the future site of the dorsal fin (white bracket in Fig. 4d). In 4.5–5.0 mm SL zebrafish, when the LMFF begins protruding (Fig. 1h), a mass of EGFP-positive cells was observed in the future site of the dorsal fin (white bracket in Fig. 4e). In 5.0–5.5 mm SL zebrafish with the LMFF outgrowth, the EGFP-positive mesenchymal cell population expanded distally (white bracket in Fig. 4f). The DsRed-positive cartilaginous elements, which become pterygiophores (basal elements of the dorsal fin skeleton), emerged at the lower part of the mesenchymal cell population (Fig. 4f’). In addition, because the site of the cartilage formation matched the distal expansion of the EGFP-positive mesenchymal cell population, these cells may give rise to the fin rays (white arrowheads in Fig. 4f’’). These data suggest that EGFP-positive cells of the gt1116A line develop dorsal fin skeletal elements, such as pterygiophores and fin rays. Thus, the dorsal fin-specific developmental mechanisms with mesenchymal cells, at least as defined by expression of prdm16A, appeared simultaneously with the protrusion of the LMFF at 4.5–5.0 mm SL, and we define the protrusion with prdm16A-positive mesenchyme as the initial dorsal fin primordium.
Cell proliferation in dorsal fin mesenchyme
An analysis of the gt1116A line suggested that the development of the dorsal fin primordial mesenchyme is associated with the protrusion of the LMFF, which raises the question of whether cell proliferation in the fin mesenchyme may contribute to dorsal fin primordium development. To identify the distribution of proliferating cells in the dorsal fin primordium, we performed wholemount immunohistochemistry with an anti-phospho histone H3 (pH3) antibody30,34. In 4.5–5.0 mm SL zebrafish (n=6) (Fig. 5a, b), a few pH3-positive cells were detected within the dorsal fin primordium in some samples (n=2/6). In many samples (n=10/12) of 5.0–5.5 mm SL zebrafish, pH3-positive cells were detected, though the number of positive cells was still few (Fig. 5c, d). In 5.5–6.0 mm SL zebrafish, all samples (n=7/7) showed many pH3-positive cells in the dorsal fin primordium (Fig. 5e, f). Fig. 5g (optical sections of the dorsal fin primordium) shows that pH3-positive cells were located in the middle of the dorsal fin primordium, indicating that these are mesenchymal (Fig. 5g). Quantitative analysis confirmed that the number of pH3-positive cells in the dorsal fin primordium increases along with body growth from 4.5 to 6.0 mm SL (Fig. 5h). Interestingly, the stage when the proliferation of the mesenchymal cells starts increasing corresponds with the stage when the height of the dorsal fin primordium starts increasing rapidly for outgrowth (5.0 mm SL, Fig. 1h). This suggests that mesenchymal cell proliferation plays a role in the outgrowth of the dorsal fin primordium.
In some fish such as sharks and cichlids, it has been shown that FGF signaling works in dorsal fin development19,35. In larval zebrafish, it has also been shown that FGF signaling play roles in early processes of LMFF development10. Therefore, we sought to investigate the role of FGF signaling in the cell proliferation of the dorsal fin primordium. SU5402 is a chemical inhibitor of Fgfr that has been reported to specifically inhibit the kinase activity of nearly all types of Fgfr10,36,37. Treatment for three days with SU5402 at 5.0–5.5 mm SL resulted in no significant effect on cell proliferation in the early dorsal fin primordium (Fig. 5i, S1a). These results suggest that FGF signaling may not play a role in the early proliferation of dorsal fin mesenchymal cells (5.0–6.0 mm SL). In anal fins, on the other hand, SU5402 treatments at 5.0–5.5 mm SL inhibited cell proliferation (Fig. S1b). It is possible that FGF signaling may play a role in the later stages of median fin development, since anal fin development greatly precedes dorsal fin development. Furthermore, SU5402 treatment also did not affect the height of the dorsal fin primordium when zebrafish were between 5.0 and 6.0 mm SL (Fig. S1c, d). Taken together, these data suggest that FGF signaling does not play a role in at least the initial protrusion process of dorsal fin primordium.