Sulf1 and Sulf2 exhibit partially overlapping expression in the adult testes. (A) In situ hybridization detects Sulf1 and Sulf2 mRNA expression in seminiferous tubules of 3-month-old mice. Sulf1 mRNA is detected in all spermatocytes, but is largely excluded from the basal layer, whereas Sulf2 mRNA is present in the basal layer and a subset of spermatocytes. (B) RT–PCR detects both Sulf1 and Sulf2 mRNA expression in primary Sertoli cells. (a–c) Sertoli cells were isolated from the testes of 2-month-old mice. Primary cultures contain a majority of Sertoli cells, identified by WT1 immunoreactivity in more than 95% cells. (d) RT–PCR assay to detect Sulf1 and Sulf2 mRNA expression in primary Sertoli cell cultures. Samples that were not RT serve as negative controls. (C) Immunohistochemistry using antibodies against Sulf1 and Sulf2 confirmed partially overlapping Sulf expression. Arrowheads mark Sertoli cells in the basal layer that express WT1; note the distinct punctate heterochromatin labeling of the nuclei by Hoechst dye (Supplementary data, Figure S1). Arrows point to the acrosome of spermatid. Asterisks indicate spermatogonia in the basal layer that are distinguishable from Sertoli cells by their more even nuclei labeling with Hoechst dye and by the expression of Oct4 and Ki67 (Supplementary data, Figure S1). Sulf1 is present in Sertoli cells and is localized to the acrosome of spermatid (a–c), whereas Sulf2 is detected in all cells in the basal layer and in the acrosome of a subset of spermatids (d–f). Scale bars, 50 μm.

Sulf mutant mice exhibit accelerated depletion of SSCs at 8 months. (A) Cross-sections of the testes from Sulf1+/−;Sulf2+/− and Sulf1−/−;Sulf2−/− mice at P10 and 3 and 8 months of age were stained with H&E and immunolabeled with antibodies against WT1 to detect Sertoli cells and Ki67 to detect spermatogonia in the basal layer of seminiferous tubules. Asterisks indicate atrophic seminiferous tubules which are devoid of spematogonia and sperm, with only Sertoli cells left in the basal layer. Scale bars, 50 μm. (B) Testicular atrophy in Sulf single- and double-mutant mice was quantified. The X-axis shows the number of mice with the testis phenotype (numerator) and the total number of mice examined for each genotype (denominator), and the Y-axis represents the percentage of atrophic seminiferous tubules for individual mice. More than 300 tubules were quantified for each animal.

Sulfs have a gene dose-dependent effect on testicular atrophy in GDNF+/− background, but not in Ret+/− background. Testicular atrophy was quantified at 3 months after crossing Sulf mutant mice with GDNF+/− mice and Ret+/− mice. The numerator of the ratio shown in the X-axis represents the number of mice with the testis phenotype, whereas the denominator represents the number of mice examined. The percentages of atrophic seminiferous tubules of affected individual mice are shown on the Y-axis. No GDNF+/−;Sulf1−/−;Sulf2−/− mice survived to adulthood. A loss of three Sulf wild-type alleles combined with GDNF haplodeficiency resulted in a synergistic effect, shown by higher levels of testicular atrophy and more affected mice than combined phenotypes from GDNF+/− mice and mice with one Sulf wild-type allele. Ret heterozygocity had no effect on testes phenotypes observed in Sulf mutant mice.

WT1 binds to promoters of the Sulf1 and Sulf2 genes to regulate their transcription. (A) Alignment of promoter sequences of the Sulf1 and Sulf2 genes from mouse (M), rat (R) and human (H) identifies a conserved WT1-binding site (in red) at ∼100 and ∼125 bp upstream of the transcription initiation site, respectively. Conserved DNA sequences are in black and unconserved nucleotides in gray. (B) WT1 binds to the Sulf1 and Sulf2 promoter sequences. ChIP assay was performed using the TM4 Sertoli cell line. The same amount of sonicated chromatin was used for the IgG control (−), a WT1 antibody (Ab) and the positive control for PCR (+). The WT1 antibody enriched Sulf promoter sequences by 2- to 5-fold in Ab pulldowns than IgG control (−), shown by more abundant PCR products using two separate primer sets (A and B). (C) The conserved WT1-binding sites in the Sulf1 and Sulf2 promoters were mutated by converting G/C to A. (D) Wild-type Sulf1 and Sulf2 promoters (S1 and S2) and mutated promoters with a disrupted WT1-binding site (S1(m) and S2(m)) were cloned into the pGL3-luciferase vector, and their transcriptional activities compared by dual luciferase assay in TM4 Sertoli cells. Cells were co-transfected with a GFP expression vector to test transcription by endogenous WT1 and with WT1(−KTS) and WT1(+KTS) to test overexpressed WT1 for transcriptional regulation and RNA binding, respectively. Results were normalized to the basal activity of empty pGL3-luciferase vector. Data shown represent the mean and standard deviations of three independent experiments with samples in triplicates. **P < 0.01. (E) Sulf1 and Sulf2 mRNA levels are reduced in Wt1+/− Sertoli cells compared with wild-type Sertoli cells, assayed by real-time RT–PCR. Error bars represent the standard deviation from three independent experiments in triplicates. **P < 0.01. (F) A model of WT1-regulated Sulf expression in Sertoli cells and Sulf regulation of GDNF signaling in the SSC niche. WT1 activates mRNA expression of both Sulf1 and Sulf2 genes by directly binding to their promoters. Sulfs enzymatically reduce GDNF binding to HS in the extracellular matrix of Sertoli cells and SSCs after secretion, leading to increased GDNF binding to receptors for signaling in SSCs. Together with other gene targets of WT1, Sulfs function to maintain the SSCs and fertility of adult males.