Characterization of the lhx4 mutation. (A) A 5 bp deletion mutation (denoted by red dashes) at the end of exon 1 of the lhx4 gene was generated by the CRISPR-Cas9 system. (B) The cDNA sequences derived from WT and lhx4-mutant brains indicate a frameshift caused by the deletion mutation (underlined in WT sequence), leading to 13 altered amino acids (aas) (red) and an early stop codon in exon 2. (C) Gel Electrophoresis of PCR products amplified from WT and lhx4-KO cDNAs using a primer set targeting exons 1 and 3 yielded a 265 bp WT and 260 bp mutant product. The similar product lengths confirm that the splicing of the lhx4 mRNA was not altered by the mutation. (D) The lhx4 mutation resulted in a predicted 41 aa truncated protein (bottom), including altered aas (orange), as compared with the 391 aa WT protein (top). The positions of the LIM1 and LIM2 domains (yellow) and homeobox domain (green) are designated in the WT protein.

Whole-mount ISH analysis of pituitary hormone-encoding transcripts in 48 hpf embryos. (A,C–E) Transcript levels measured by the integrated densities of the whole-mount-ISH-staining signals of tshb ((A); N = 10 homozygotes and 13 WTs), gh ((C); N = 18 homozygotes and 10 WTs), pomca ((D); N = 13 homozygotes and 11 WTs) and prl ((E); N = 10 homozygotes and 6 WTs) in 48 hpf lhx4 mutants and their WT siblings from heterozygous intercross. Homozygous lhx4 mutants expressed lower amounts of tshb compared with their WT siblings ((A); *** p < 0.001, Mann–Whitney test), while the expressions of gh, pomca and prl remained unaltered (C–E). Error bars indicate s.e.m. (B) Representative samples (heads of 48 hpf embryos, dorsal view) of WT sibling (left) and homozygous lhx4 mutant (right), analyzed by whole-mount ISH using the tshb probe. Tshb ISH signal (denoted by arrows) is substantially reduced in the pituitary of homozygous lhx4 mutant at 48 hpf as compared with its WT sibling. Scale bar = 100 µm.

Whole-mount ISH analysis of pituitary hormone-encoding transcripts in 7 dpf larvae. (A,C–E) Transcript levels measured by the integrated densities of whole-mount-ISH-staining signals of tshb ((A); N = 10 homozygotes and 10 WTs), gh ((C); N = 11 homozygotes and 10 WTs), pomca ((D); N = 9 homozygotes and 4 WTs) and prl ((E); N = 9 homozygotes and 9 WTs) in 7 dpf lhx4 mutants and their WT siblings from heterozygous intercross. Homozygous lhx4 mutants expressed lower amounts of tshb compared with their WT siblings ((A); ** p < 0.01, Mann–Whitney test), while the expressions of gh, pomca and prl remained unaltered (C–E). Error bars indicate s.e.m. (B) Representative samples (heads of 7 dpf larvae, dorsal view) of WT sibling (left) and homozygous lhx4 mutant (right), analyzed by whole-mount ISH using the tshb probe. Tshb ISH signal (denoted by arrows) is considerably reduced in the pituitary of homozygous lhx4 mutant at 7 dpf as compared with its WT sibling. Scale bar = 100 µm.

Reduced body size of adult lhx4 mutants. (A) Bar chart representing head-to-tail-base length (mm) of 5-month-old lhx4-mutant fish and WT siblings that were mutually raised under controlled conditions. Homozygous lhx4 mutants (N = 42) are significantly shorter than their WT siblings (N = 59; *** p < 0.001, Mann–Whitney test). Error bars indicate s.e.m. The presented results were pooled from four independent repeats. (B) Representative adult homozygous lhx4 mutant (top) and WT sibling (bottom), lateral views. Bar scale = 2 mm.

qRT-PCR analysis of pituitary hormone-encoding transcripts in adult fish. The relative expressions of pituitary hormone-coding mRNAs in pituitary glands dissected from 4-month-old lhx4 mutants (N = 13) and their WT siblings (N = 10), as measured by qRT-PCR. Adult homozygous lhx4 mutants express significantly lower mRNA levels of tshb (A), gh (B), pomca (C) and fshb (D) in the pituitary, compared with their WT siblings (* p < 0.05, Mann–Whitney test), while the lhb mRNA levels are insignificantly reduced (E). Error bars indicate s.e.m

Reduced pomca-driven expression in corticotrophs of adult lhx4 mutants. (A) Tg(pomca:GFP) reporter line was utilized for evaluating corticotroph development in 4-month-old lhx4 mutants (N = 21) and their WT siblings (N = 18). The integrated fluorescence in the pituitaries of homozygous lhx4 mutants was reduced (* p < 0.05, t-test) as compared with that of their WT siblings. (B) Representative pomca promoter-driven fluorescence in the pituitaries of adult WT (left) and homozygous lhx4 mutant (right). Scale bar = 100 µm.

Immature ovaries and depleted gonadotrophs in lhx4 KO females. (A) Hematoxylin and eosin histology performed on adult fish ovary sections, demonstrating an undeveloped ovary of homozygous lhx4-mutant female (top), compared with a properly developed WT sibling ovary (bottom). Scale bar = 200 µm. (B) Expression of RFP under the tilapia lhb promoter in the pituitary of adult WT sibling female (left) and homozygous lhx4-mutant female (right), indicating severely impaired gonadotroph development in the mutant. Scale bar = 100 µm.

Unillustrated author statements