Hyaluronic acid synthesis is required for zebrafish tail fin regeneration

Abstract

Using genome-wide transcriptional profiling and whole-mount expression analyses of zebrafish larvae, we have identified hyaluronan synthase 3 (has3) as an upregulated gene during caudal fin regeneration. has3 expression is induced in the wound epithelium within hours after tail amputation, and its onset and maintenance requires fibroblast growth factor, phosphoinositide 3-kinase, and transforming growth factor-ß signaling. Inhibition of hyaluronic acid (HA) synthesis by the small molecule 4-methylumbelliferone (4-MU) impairs tail regeneration in zebrafish larvae by preventing injury-induced cell proliferation. In addition, 4-MU reduces the expression of genes associated with wound epithelium and blastema function. Treatment with glycogen synthase kinase 3 inhibitors rescues 4-MU-induced defects in cell proliferation and tail regeneration, while restoring a subset of wound epithelium and blastema markers. Our findings demonstrate a role for HA biosynthesis in zebrafish tail regeneration and delineate its epistatic relationships with other regenerative processes.

Fig 1. Zebrafish hyaluronan synthases are differentially expressed during larval and adult tail regeneration.

Expression patterns of has1, has2, and has3 in regenerating larval (A-C; 1 dpa) adult (D-E; 2 dpa) tails, as determined by whole-mount in situ hybridization. (A’-C’) Equivalently stained uncut controls. At least 30 larvae or 10 adult zebrafish were analyzed for each experimental condition, and phenotypic descriptions were based on a penetrance of > 80%. Scale bars: A-C and A’-C’, 100 μm; D-E: 300 μm.

Fig 2. Zebrafish hyaluronidases are differentially expressed during larval and adult tail regeneration.

Expression patterns of hyal2, hyal3, hyal4, and hyal6 in regenerating larvae (A-D; 1 dpa) and adult (A’-D’; 2 dpa) tails. At least 30 larvae or 10 adult zebrafish were analyzed for each experimental condition, and phenotypic descriptions were based on a penetrance of > 80%. Scale bars: A-D, 100 μm; A’-D’: 300 μm.

Fig 3. Dynamics of has3 expression during larval tail regeneration.

(A-F) Expression patterns of has3 at different time points after caudal fin amputation at 2 dpf. Arrows mark the initial appearance of has3 transcripts at 6 hpa, localized to dorsal and ventral sides regions of the regenerative bud. (A’-F’) Equivalently stained uncut controls at the same time points. At least 30 larvae were analyzed for each experimental condition, and phenotypic descriptions were based on a penetrance of > 80%. Scale bar: 100 μm.

Fig 4. Multiple signaling pathways regulate the onset and maintenance of has3 expression during larval tail regeneration.

Expression of has3 in 1-dpa (3-dpf) larval tails treated with the following signaling pathway inhibitors for the first 24 hours after amputation: (A, A’) 0.5% DMSO. (B, B’) 75 μM PD173074 (FGF). (C, C’) 10 μM LY294002 (PI3K). (D, D’) 50 μM SB431542 (TGFß). (E, E’) 5 μM SP600125 (JNK). (F, F’) 50 μM DAPT (Notch). At least 30 larvae were analyzed for each experimental condition, and phenotypic descriptions were based on a penetrance of > 80%. Scale bar: 100 μm.

Fig 5. 4-MU inhibits hyaluronan synthase-dependent HA production.

Agarose gel electrophoresis and cationic dye staining of HA produced by HEK293 cells transfected with either zebrafish has3 (lanes 2–5) or human HAS2 (lanes 7–10) and then treated with varying doses of 4-MU. The structural identity of the stained HA was confirmed by Streptomyces hyalurolyticus hyaluronidase treatment (lanes 6 and 11).

Fig 6. 4-MU inhibits larval tail regeneration.

(A) Representative micrographs of larval tails that were amputated at 2 dpf and then treated 0.5% DMSO or 150 μM 4-MU for 3 days. Dotted lines indicate the amputation plane, and micrographs of uncut larval tails subjected to the same inhibitor regimen are shown for comparison. Scale bar: 100 μm. (B-C) Time-course analysis of 4-MU action on larval tail regeneration. Caudal fin sizes at 5 dpf (3 dpa) after the indicated amputation and 4-MU treatment regimens. Data are the average caudal fin areas of 15 larvae ± s.e.m., normalized to the average fin size of uncut larvae treated with 0.5% DMSO. ***, P < 0.001.

Fig 7. 4-MU inhibits adult tail regeneration.

Representative micrographs of adult tail fins that were amputated and then treated with 0.5% DMSO or 150 μM 4-MU for 7 days. White dotted lines indicate the amputation site. 10 adult zebrafish were analyzed for each experimental condition, and phenotypic descriptions were based on a penetrance of > 80%. Scale bar: 1 mm.

Fig 8. 4-MU inhibits regenerative cell proliferation.

(A) Mitotic cells in the larval tail after the indicated amputation and 4-MU treatment regimens, as visualized with anti-pH3 immunostaining at 2 dpa (4 dpf). R1 and R2 demarcate distinct regions within the larval tail, with R1 corresponding to a highly proliferative 100-μm-wide zone associated with tail regeneration. (B) Quantification of pH3-positive cells in the R1 and R2 regions under the indicated treatment conditions. Data are the average number of pH3-positive cells in 30 larval tails ± s.e.m. **, P < 0.01.

Fig 9. GSK3 inhibition rescues 4-MU-induced larval tail regeneration and cell proliferation defects.

(A) Representative micrographs of larval tails that were amputated at 2 dpf and treated with 0.5% DMSO, 100 nM BIO, 150 μM 4-MU, or 150 μM 4-MU with 100 nM BIO for the next 24 hours. Caudal fins of 5-dpf (3-dpa) larvae are shown. Scale bar: 100 μm. (B) Caudal fin sizes at 5 dpf (3 dpa) for the indicated amputation and inhibitor treatment regimens (compound administration from 2 to 3 dpf). Data are the average caudal fin areas of 15 larvae ± s.e.m., normalized to the average fin size of uncut larvae treated with 0.5% DMSO. (C) Cell proliferation within the 4-dpf caudal fin in response to the indicated amputation and inhibitor treatment regimens. Data are the average number of pH3-positive cells in 30 larval tails ± s.e.m. (R1 + R2 regions; see Fig 9). *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Fig 10. 4-MU and GSK3 inhibition differentially control the expression of wound epithelium and blastema markers in larval tails.

Effects of 0.5% DMSO, 100 nM BIO, 150 μM 4-MU, or 150 μM 4-MU with 100 nM BIO on junba (A), dlx5a (B), aldh1a2 (C), and junbb (D) expression in 1-dpa (3-dpf) larval tails.