Sertoli cell ablation and replacement of the spermatogonial niche in mouse

Spermatogonia, which produce sperm throughout the male lifetime, are regulated inside a niche composed of Sertoli cells, and other testis cell types. Defects in Sertoli cells often lead to infertility, but replacement of defective cells has been limited by the inability to deplete the existing population. Here, we use an FDA-approved non-toxic drug, benzalkonium chloride (BC), to deplete testis cell types in vivo. Four days after BC administration, Sertoli cells are preferentially depleted, and can be replaced to promote spermatogenesis from surviving (host) spermatogonia. Seven days after BC treatment, multiple cell types can be engrafted from fresh or cryopreserved testicular cells, leading to complete spermatogenesis from donor cells. These methods will be valuable for investigation of niche-supporting cell interactions, have the potential to lead to a therapy for idiopathic male infertility in the clinic, and could open the door to production of sperm from other species in the mouse.

T he efficient production of sperm from spermatogonia requires the coordinated interplay between germ cells and various somatic support cells making up the spermatogenic niche. Approximately 1:100 adult males is azoospermic, and in~40% of these cases the defect is idiopathic and may lie with the germ cell itself or any of the somatic cell types making up the spermatogenic niche. Of these cell types, Sertoli cells reside within the testis tubules attached to the same basal lamina as the spermatogonia, while Leydig cells, peritubular myoid cells (PMCs) and macrophages lie immediately outside the tubules in the supporting stroma. Since the Sertoli cells interact directly with germ cells and are vital to provide morphological and nutritional support for spermatogenesis, their dysfunction is often the cause of spermatogenic failure [1][2][3][4][5] .
To compensate for Sertoli cell dysfunction, several therapeutic approaches have been investigated, such as transplantation of healthy Sertoli cells 6 , supplementation of missing growth factors by viral transduction 7-10 , or explant and culture of testis tissue in vitro 11 . Although researchers showed evidence that offspring did not carry the transgene 7-10 , it is difficult to exclude the possibility of viral transmission to the next generation, and no human testis culture system has yet been established. Since auto-transplantation of testicular cells from Hodgkin's lymphoma patients was reported 12,13 , Sertoli cell transplantation has been a possible approach for the clinic. However, long-term success depends upon sufficient space being made available along the basal lamina for repopulation of the tubules by donor Sertoli cells. For in vivo feasibility studies, some groups have used transgenic mouse models expressing cytotoxic genes or specific toxin receptors to eliminate Sertoli cells prior to transplantation 14 . However, since these approaches require genetic manipulations, they are not suitable for clinical studies. Other studies used cadmium injection into the testis to deplete Sertoli cells, but cadmium administration has many adverse side effects and could not be used in the clinic 15 .
Here, we demonstrate the use of a well-described FDAapproved nontoxic drug, benzalkonium chloride (BC), to remove host Sertoli cells and enable cell transplantation into the testis in vivo. BC is a quaternary ammonium that acts as a detergent to disrupt the lipid membrane of cells. Originally developed as a germicide to kill microorganisms, it was first used in the ophthalmic industry in the 1940s as a preservative in hard contact lens solutions.
We show that treatment of the testis with BC leads to rapid depletion of Sertoli cells. We describe two windows of opportunity for engraftment of various testis cell types after BC injection into the rete testis. Within 4 days of administration, BC preferentially depletes only Sertoli cells, leaving a scaffold into which healthy, isolated Sertoli cells can be engrafted to allow rescue of spermatogenesis from surviving (host) spermatogonia. At later times after BC treatment, some germ cell and stromal support cell types are also lost, most likely secondary to the loss of the trophic support of Sertoli cells. Significantly, these cells can be replaced by engraftment of fresh or cryopreserved testicular cells, leading to recovery of spermatogenesis from donor cells.
We show that the effects of BC extend to a large animal model (canine). As BC is a common component of eye drops, we anticipate that our Sertoli/niche cell ablation method could be safely adapted for the clinic to overcome the weaknesses of previous approaches. We anticipate that this method could be used to rescue host fertility, and possibly to produce xenogeneic sperm in mouse. This approach will also open up opportunities for in vivo studies of niche cell interactions and function.

Results
Neonatal Sertoli cell selective depletion in vitro. In our initial experiments, we incubated a whole neonatal mouse testis in 0.02% BC solution for 10 min and, after thorough washing, cultured the organ for 4 days at the surface/air interface in wells atop an agarose block (Fig. 1a). The testis was then fixed and stained for whole-mount immunofluorescence analysis using an antibody against the Sertoli cell marker SOX9 (Sry-box 9). In this assay, SOX9 + Sertoli cells were depleted throughout the testis (Fig. 1b,  c). Neonatal testes incubated in PBS, 0.02% or 0.03% BC were sectioned, and Sertoli and other cell types were counted ( Fig. 1; Supplementary Fig. 1a-l). A solution of 0.02% BC completely eliminated Sertoli cells in 82% of the tubule cross-sections analyzed (719/873 tubule sections, n = 4 independent testes), while fewer Sertoli cells were left in the remainder (952 cells/154 tubule sections;~6 cells/tubule section; n = 4 independent testes) compared with control (3070 cells/115 tubule sections;~27 cells/ tubule section; n = 4 independent testes) (P = 0.029). Examples of most-to-least affected tubules are presented in Supplementary  Fig. 1m-o.
We performed H&E staining on samples treated with 0.02% BC to confirm that Sertoli cells (and not only SOX9 protein) were lost. These assays showed that by day 3, there was a severe depletion of Sertoli cell nuclei along the basal lamina of seminiferous cords ( Supplementary Fig. 2a, b). Apoptotic cell death increased from day 2 to day 4 based on staining with cleaved caspase 3 ( Supplementary Fig. 2c, d). Loss of SOX9 + cells (Fig. 1b, c) was associated with elevated numbers of F4/80 + macrophages. However, despite the severe depletion of Sertoli cells based on both histology and SOX9 staining, the normal distribution of Laminin (LMN) showed that the structure of the seminiferous tubule was well maintained (Fig. 1d, e). Importantly, other cell types in the testis, 3βHSD (3β-hydroxysteroid dehydrogenase)-positive Leydig cells (Fig. 1f, g) were spared. Immunohistochemistry for smooth muscle actin, alpha (αSMA) suggested that PMCs were intact (Fig. 1h, i), and antibody staining with both the germ-cell-specific monoclonal antibody (TRA98) 16 and GDNF family receptor alpha-1 (GFRα1) revealed that some germ cells remained along the basement membrane in Sertoli-ablated tubules ( Fig. 1, j-m). Testes treated with 0.02% or 0.03% BC were sectioned, and the number of germ cells per tubule cross-section was counted. In samples treated with 0.02% BC, germ cell numbers were significantly reduced (~4 cells/tubule cross-section in a total of 968 cross-sections analyzed; n = 4 independent testes) compared with controls (~9 cells/tubules cross-section in a total of 340 tubules counted; n = 4 testes) (P = 0.029) (Fig. 1n), likely a secondary effect of Sertoli cell depletion. However, the numbers of PMCs and Leydig cells outside the tubules were not affected: 1303 PMCs/125 tubule cross-sections in treated testis (n = 4 independent testes) versus 963 cells/101 tubule cross-sections in control testis (n = 4 independent testes); and 3253 Leydig cells/873 tubules in treated testis (n = 4 independent testes) versus 1883 cells/541 tubules in control testis (n = 4 independent testes) (Fig. 1n). Although incubation of the neonatal testis with 0.03% BC led to a more severe loss of Sertoli cells, it also caused disruption of cord structure and loss of germ cells, PMCs, and Leydig cells (Supplementary Fig. 1c and Supplementary Fig. 2e, f).

Time course of cell depletion in vivo.
To test the efficacy of BC in vivo, we injected a 0.02% BC solution or PBS into the rete testis of anesthetized adult male mice carrying the Sox9-ECFP transgene, which marks Sertoli cells (Fig. 2a). H&E staining and immunohistochemistry showed that many Sertoli cell nuclei disappear by day 4 ( Supplementary Fig. 3a-d). This result was  ARTICLE confirmed by loss of SOX9 + Sertoli cells from 27% of the tubule cross-sections analyzed (248/908, n = 4 independent testes) ( Fig. 2b-l). In 6% of the tubule cross-sections (59/908, n = 4 testes), fewer Sertoli cells remained (an average of 6 Sertoli cells/ tubule cross-section compared with an average of 28 Sertoli cells/ tubule cross-section in testes injected with PBS) (P = 0.029).
Analysis of affected testis tubules on day 3 showed defects in the differentiating germ cell layer, with many irregularly condensed nuclei ( Supplementary Fig. 3e, f). Cleaved Caspase 3 + cells were rarely seen on day 3 ( Supplementary Fig. 3e), but some cells, presumably spermatocytes, were positive for Caspase 3 on day 4 ( Supplementary Fig. 3g). There was evidence for accumulation of To quantify the effect of BC on other cell types in adult testis in vivo, the number of HuC/ D + spermatogonia, Leydig cells, or PMCs per cross-section of BC-affected seminiferous tubules was counted (n = 4 independent testes). On day 4, the number of spermatogonia (positive for the spermatogonial marker, HuC/D) per tubule cross-section was decreased in drug-treated testes (P = 0.029). However, the numbers of Leydig cells and PMCs were similar to controls (Fig. 2l). In summary, 4 days after 0.02% BC treatment, Sertoli cells were significantly depleted, but the interstitial compartment, testis cord structure, and some spermatogonia were still present, and there was no evidence of an effect on the behavior or viability of the treated animals.
Canine Sertoli cells are also depleted by drug treatment. To test whether BC can deplete Sertoli cells from the testis of other mammals, we used the same in vitro procedure on 6 to 8-weekold canine testis. Testis tissue fragments were incubated with PBS or 0.02% or 0.03% BC for 10 min, washed, and cultured for 3 days ( Supplementary Fig. 5a, b). At the end of this time, testis fragments were fixed and stained for immunofluorescence analysis using antibodies against SOX9 and the germ cell marker MVH (mouse vasa homolog). In control tissue, both Sertoli cells and germ cells were detected in seminiferous tubules ( Supplementary  Fig. 5c). However, in 0.02% BC-treated testis tissue, Sertoli cells were depleted, particularly from the peripheral regions (Supplementary Fig. 5d). In 0.03% BC-treated tissues, all Sertoli cells were absent, but some spermatogonia survived in Sertoli-depleted seminiferous tubules (Supplementary Fig. 5e). Tubule structure was well maintained ( Supplementary Fig. 5f, g). However, compared with controls ( Supplementary Fig. 5f, h), the numbers of GATA4-positive Leydig cells were reduced in 0.03% BC-treated testes ( Supplementary Fig. 5g), and αSMA-positive PMCs were almost completely lost ( Supplementary Fig. 5i), similar to our observations in 0.03% BC-treated mouse testis ( Supplementary  Fig. 1c).

Repopulation of Sertoli cells in BC-treated testis.
To determine whether donor Sertoli cells can efficiently engraft in drugdepleted cords, we transplanted a purified population of Sertoli cells isolated by fluorescence-activated cell sorting (FACS) from an 6.5-10.5 days postpartum (dpp) Sox9-ECFP mouse testis (for analysis of this population, see Supplementary Fig. 6a, b) into an adult mouse testis prepared by injection of BC into the rete 4 days earlier (Fig. 3a). Soon after transplantation, some clusters of donor cells were found in the lumen (Fig. 3b). However, after 12 days, donor Sertoli cells colonized the basement membrane in some tubules (2-6/section) and surrounded host germ cells (Fig. 3c, d). Thirty-three days after transplantation, donor Sertoli cells were present near host spermatogonia that had differentiated to primitive spermatids based on PNA (peanut agglutinin) staining (Fig. 3e). Elongating spermatids were detected 8 weeks after transplantation (Fig. 3f, g) in two of four engrafted testes.
Engraftment of multiple cell types into BC-treated testis. We used a DND1 (dead end homolog 1)-EGFP mouse to investigate spermatogonia kinetics after BC treatment. In this transgenic line, all spermatogonia, including undifferentiated spermatogonia and differentiating spermatogonia express EGFP ( Supplementary  Fig. 7, a-d). Seven days after BC injection, in 24% of tubule crosssections (205/868; n = 4 independent testes), both Sertoli cells and germ cells, including spermatogonia, had disappeared ( Fig. 4a, b). Although Sertoli cells and germ cells/spermatogonia were lost 7 days after treatment, cord structure, Leydig cells, and PMCs were intact (Fig. 4, c-e), and their numbers surrounding BC-affected seminiferous tubules were similar to controls (n = 4 independent testes) (Fig. 4f).
To determine whether the host testes depleted for both Sertoli and germ cells could be used as a scaffold for engraftment of multiple cell types, we performed testicular cell transplantation 7 days post-BC treatment (Fig. 5a). Nineteen days after transplantation into a BC-treated host testis, donor Sertoli cells and spermatogonia had colonized in~4-8% of host tubules (Fig. 5b). A highly colonized region of a testis isolated 59 days posttransplantation is shown in Supplementary Fig. 8a, b. Donor spermatogonia proliferated with support from donor Sertoli cells (Fig. 5c), initiated meiosis, and progressed to the spermatocyte stage. EGFP-positive leptotene and pachytene spermatocytes were detected (Fig. 5d). Other donor testicular cells, negative for SOX9, were observed around some tubules where donor Sertoli cells and spermatogonia colonized (Fig. 5e-l). Most of these interstitial EGFP-bright donor cells were Leydig cells (Fig. 5g-i), but some donor PMC colonization was also detected (Fig. 5j-l). Because these cells were observed specifically around the seminiferous tubules where donor Sertoli cells and germ cells colonized, we Fig. 2 Four days after BC treatment in vivo, only Sertoli and germ cells were severely depleted. a Schematic representation of experimental procedures for in vivo adult Sertoli cell ablation. b-k Sox9-ECFP adult mouse testis 4 days after PBS or BC injection into seminiferous tubules. Tissues were stained with antibodies against ECFP (green; SOX9-ECFP, in this transgenic line, ECFP is present throughout the nucleus and cytoplasm of Sertoli cells) and Hoechst (blue). b Antibody staining of endogenous SOX9 (red); c, d αSMA (peritubular myoid cells; white; arrow). BC-affected tubule is marked "A", and BCunaffected tubule is marked "U". e LMN-positive basement membrane (red). f Leydig cells (3βHSD-positive, red). g Vascular structures (PECAM1-positive, red) are shown. The left bottom corner of each frame (white box) shows a magnification of a vessel. h MVH-positive germ cells (red). i STRA8-positive spermatogonia (red). j HuC/D-positive spermatogonia (magenta) on the basement membrane in treated or untreated control (inset). k C-KIT-positive differentiated spermatogonia (magenta) in treated or untreated control (inset). The rectangular area surrounded by the broken line is enlarged on the right. Ten independent experiments. Scale bar: 100 µm. l Quantification of BC affect on Sertoli, germ cells, Leydig, and peritubular myoid cells. Data were analyzed from four biologically independent samples examined over three independent experiments and expressed as mean ± SD; (NS) not significant. Statistical analysis was performed using unpaired t test, Kolmogorov-Smirnov test.

Transplantation of cryopreserved testicular cells.
To test the efficiency of engraftment using cryopreserved donor tissue, we dissociated testis tissue from CAG-EGFP pups and cryopreserved the cells for 39 days. Seven days after BC treatment of wild-type host mice, cryopreserved donor testicular cells were thawed and introduced into the rete testis. The results were similar to the results using freshly prepared cells for transplantation. Seventy-one days after transplantation, we found that cryopreserved Sertoli cells and spermatogonia had colonized testis tubules and donor supporting cells, presumably Leydig cells and PMCs, were found in the interstitium (Fig. 6, a-d; Supplementary Fig. 8c, d).
In some tubules, cryopreserved Sertoli cells supported host spermatogenesis (Fig. 6b), and in other tubules, cryopreserved donor Sertoli cells colonized with donor spermatogonia and supported donor spermatogenesis (Fig. 6c, d). In 3/3 host testes, cryopreserved Sertoli cells and spermatogonia colonized the tubules, and advanced spermatogenesis from donor cells was observed in two of these hosts.

Discussion
Defects in Sertoli cells and other somatic cell types that form the spermatogonial niche in the testis are responsible for many cases of infertility 2-4 , but efforts to replace defective Sertoli cells in vivo have been hampered by the inability to achieve engraftment in competition with resident cells 14 . Ablation of Sertoli cells was previously reported using diphtheria toxin receptor-mediated conditional and targeted cell ablation systems 14,17,18 . However, while these methods have significantly advanced our knowledge of testis biology, they are not feasible in the clinic.
We have devised a technique to deplete Sertoli cells using a nontoxic agent, benzalkonium chloride. BC was previously reported to remove mouse and rat enteric nerve plexi when the GI tract was treated in vitro [19][20][21][22][23][24][25] . In this study, 0.02% BC injection into the rete testis caused no discernable side effects in the mouse, but did cause severe Sertoli cell depletion and partial depletion of spermatogonia and differentiating sperm. BC is a cationic surfactant that works preferentially on cells with a strong negative membrane charge 26 , raising the possibility that the Sertoli cell membrane might have a more negative charge than other cell types in the testis. Future experiments will be designed to address the mechanism of BC action. We hypothesize that the most depleted tubules were those most closely connected to the rete site of injection. We show that the effect of incubation with BC is similar in canine testes, which suggests that the treatment can be used in large mammals. In all, 0.03% BC treatment of mouse and canine testes in vitro led to Sertoli cell elimination, but also caused PMC depletion and reduced the number of Leydig and germ cells. These results indicate that the dose and time to engraftment will require optimization for each species.
Alongside this discovery, we developed methods for engraftment of Sertoli cells or multiple testis cell types. An ultrasoundguided seminiferous tubule transplantation technique has already been established for primates and humans 27 , thus our method is technically feasible for clinical applications. Sertoli cells used for engraftment could be derived via an induced pluripotent stem cell (iPSC) strategy 28 to avoid immune rejection. In cases where the defect in Sertoli cells is identified through whole-exon sequencing and cross-referencing with known Sertoli essential genes (eg. KITL), it could be corrected via CRISPR-mediated genome editing strategies in iPSCs prior to transplantation following BC depletion. A proof-of-principle experiment demonstrating this approach in mice would be an important follow-up experiment. Testing the fertility of sperm generated with support from donor Sertoli cells is not feasible in this experiment, because we cannot distinguish host sperm matured in the company of donor Sertoli cells from sperm matured in regions of the host testis where host Sertoli cells survived BC treatment. In future experiments, we will replace Sertoli cells in the Kitl Sl/Sld mutant mouse, which lacks spermatogenesis due to a Sertoli cell defect, and perform micro insemination to test fertility.
Since seminiferous transplantation was developed by Brinster 29 , xenogeneic transplantation has been investigated. Although human and piglet spermatogonia attach to the mouse testis basement membrane and remain for several months, they do not show further maturation [30][31][32] , suggesting that the mouse germ cell niche cannot support xenogeneic spermatogonia. Here, we show that cryopreserved Sertoli cells, Leydig cells, and PMCs can colonize the BC-treated testis, regenerate the germ cell niche, and support spermatogenesis. Replacement of many cell types comprising the germ cell niche might make it possible to produce human sperm in an immunocompromised BC-treated mouse testis from cryopreserved tissue of pre-pubertal donors.
In summary, we have identified an FDA approved, safe drug that induces severe Sertoli cell depletion when added to isolated mouse or canine testes in vitro or injected into the rete testis of mice in vivo. Using this drug, we have developed techniques for Sertoli-only transplantation and for transplantation of spermatogonia plus multiple supporting cell types. We hope to use these approaches to analyze the function of Sertoli cells, design a therapy for human idiopathic male infertility, and investigate the cross talk between niche cells and stem cells. Finally, if xenogeneic spermatogenesis can be achieved, it may not only be valuable for treatment of infertility for young cancer patients but may also be . e, f In addition to donor cells inside tubules, some of which express SOX9 (red), many SOX9-negative cells colonized outside of tubules. g-i Donor Leydig cells expressing 3βHSD (red) colonized the interstitium, and j-l donor PMCs (arrow; αSMA (red)) surrounded testis cords (region within dotted rectangle is magnified to the right in the green channel; green myoid cell is outlined in red). SOX9 (magenta). m Recipient testis 61 days after 7 dpp CAG-EGFP testicular cell transplantation. SOX9 (magenta) labels both donor Sertoli cells (arrows) and a remaining host Sertoli cell (arrowhead). PNA (red) co-labels many donor spermatogonia (green). n Host testis 59 days after testicular cell transplantation from 10 dpp CAG-EGFP donors. Donor Sertoli cells (arrow, green) surround elongating sperm labeled with MVH (red). Donor-derived spermatogonia persist on the basement membrane (arrowhead). o Higher magnification view of the region surrounded by broken lines shows elongating sperm. Donor spermatogenesis with multiple cell type colonization was observed in 9 of 13 biologically independent testes examined over four independent experiments. Scale bar: 25 µm.
adapted to produce sperm from humans or endangered species in mice.

Methods
Animals. CD1 outbred, C57BL/6J and FVB inbred strains (from The Jackson Laboratory), and FVB CAG-EGFP 33 , C57BL/6J Sox9-ECFP 34   b Cryopreserved donor Sertoli cells (green), co-labeled with antibodies against SOX9 (magenta; arrows), colonized and supported host spermatogenesis (PNA (red) marks round spermatids that are not green). c, d Cryopreserved donor Sertoli cells (green, with magenta nuclei, white arrows) colonized beside donor spermatogonia (white arrowheads) and supported donor spermatogenesis (PNA (red) marks round spermatids from the donor (green)). Host Sertoli cells (yellow arrows) were also present. c Donor Leydig cells (black arrowhead) and d donor peritubular myoid cell (black arrow) were also observed. Hoechst (blue). Cryopreserved donor advanced spermatogenesis with supporting cell colonization was observed in 2/3 biologically independent testis in two independent experiments. Scale bar: 25 µm. determined using trypan blue (Thermo Scientific; 15250-061) and found to be high (80-90%). The solution was diluted with 10% FBS + DMEM + P/S to a concentration of 1 × 10 6 /ml isolated Sertoli cells or 1 × 10 8 /ml testicular cells, and 10-15 μl were injected through the rete testis into the seminiferous tubules of a host mouse on day 4 or day 7 after BC treatment. A glass needle (pulled as for drug injection) was back-loaded with the cell suspension and 0.04% Trypan blue to track the efficiency of the injection.
Whole-mount immunofluorescent cytochemistry. For cultured neonatal mouse testes, whole-mount immunofluorescence analysis was performed. Samples were fixed overnight with 4% (wt/vol) paraformaldehyde in PBS at 4°C. After several washes in PBS, samples were processed through a methanol gradient (25%, 50%, 75% and 100% in PBS) and stored in 100% methanol at −20°C until use. Samples were re-hydrated and washed in PBTx (PBS, 1% Triton X-100), samples were incubated in blocking solution (PBTx, 10% FBS and 3% BSA) for 1 h at room temperature. Primary antibodies were diluted in blocking solution and applied to samples for 1 to 2 days at 4°C. After several washes in 1% PBTx, fluorophoreconjugated secondary antibodies were applied for 2 days at 4°C as described 38 . After several washes in PBS, samples were mounted in polyvinyl alcohol (MP Biomedicals; 151937) and imaged with a Zeiss 780 upright confocal laser microscope.
Immunofluorescent cytochemistry on cryosections. Neonatal and adult mouse testis and canine testis tissues were fixed in 4% paraformaldehyde in PBS at 4°C overnight, processed through a sucrose gradient (10%, 15%, and 20% in PBS), and cryo-embedded in OCT compound (Sakura Finetek) 38 . Samples were cut into 8-12-μm cryosections. Samples were incubated with blocking solution (0.1% PBTx + 10% FBS + 3% BSA) for 30 min, and primary antibodies were diluted in blocking solution and incubated overnight at 4°C. Samples were washed several times in 0.1% PBTx and incubated with fluorophore-conjugated secondary antibodies for 1 h at room temperature. Samples were mounted after several washes in PBS. Primary and secondary antibodies used for immunofluorescence are listed in Supplementary Tables 1 and 2. Nuclei were counterstained with Hoechst 33342 dye. Samples were mounted in polyvinyl alcohol and imaged with a Zeiss 780 upright confocal laser scanning microscope.
Immunofluorescent cytochemistry on FACS-isolated cells. Testicular cells were cultured on glass coverslips for 30 min at room temperature and fixed in 4% paraformaldehyde in PBS at room temperature for 20 min. Cells were incubated with blocking solution (0.1% PBTx + 10% FBS + 3% BSA) for 20 min, and primary antibodies against GFP and MVH were diluted in blocking solution and incubated at room temperature for 20 min. Samples were washed several times in 0.1% PBTx and incubated with fluorophore-conjugated secondary antibodies for 20 min at room temperature. Samples were mounted in polyvinyl alcohol after several washes in PBS and imaged with a Zeiss 780 upright confocal laser scanning microscope.
Quantitative analysis using immunohistochemical sections. The Fiji version 1.52 g was used to count the proportion of tubules affected by BC 39 . This proportion was estimated by counting the total number of seminiferous tubule crosssections (defined by αSMA staining) in 10-24 serial sections, and determining the proportion of tubule cross-sections not containing Sertoli cells (SOX9 + cells). The number of Sertoli cells in tubules, and the numbers of germ cells (TRA98 + cells in neonatal testis and HuC/D + spermatogonia in adult samples) were determined. Leydig cells (3βHSD + cells) and PMC (αSMA + cells) were counted in the interstitial areas immediately surrounding each cord. The GraphPad Prism version 6 was used for statistical analysis. Two-thirds of the values fall within the top and bottom bars. Statistical significance was determined by unpaired t tests, Kolmogorov-Smirnov test.
Reporting summary. Further information on research design is available in the Nature Research Reporting Summary linked to this article.