Establishment of Etv5 gene knockout mice as a recipient model for spermatogonial stem cell transplantation

Transplantation of spermatogonial stem cells (SSCs) is an alternative reproductive method to achieve conservation and production of elite animals in livestock production. Creating a recipient animal without endogenous germ cells is important for effective SSC transplantation. However, natural mutants with depletion of SSCs are difficult to obtain, and drug ablation of endogenous germ cells is arduous to perform for practical use. In this study, we used mouse models to study the preparation of recipients with congenital germ cell ablation. We knocked out (KO) Ets-variant gene 5 (Etv5) in mice using the CRISPR/Cas9 system. The testicular weight of Etv5-/- mice was significantly lower than that of wild-type (WT) mice. The germ cell layer of the seminiferous tubules gradually receded with age in Etv5-/- mice. At 12 weeks of age, the tubules of Etv5-/- mice lacked germ cells (Sertoli cell-only syndrome), and sperm were completely absent in the epididymis. We subsequently transplanted allogeneic SSCs with enhanced green fluorescent protein (EGFP) into 3-(immature) or 7-week-old (mature) Etv5-/- mice. Restoration of germ cell layers in the seminiferous tubules and spermatogenesis was observed in all immature testes but not in mature adult testes at 2 months post-transplantation. The presence of heterologous genes Etv5 and EGFP in recipient testicular tissue and epididymal sperm by PCR indicated that sperm originated from the transplanted donor cells. Our study demonstrates that, although Etv5-/- mice could accommodate and support foreign germ cell transplantation, this process occurs in a quite low efficiency to support a full spermatogenesis of transplanted SSCs. However, using Etv5-/- mice as a recipient model for SSC transplantation is feasible, and still needs further investigation to establish an optimized transplantation process.

protein expression in Etv5 -/male and WT mice by Western blot. The anti-β-actin monoclonal antibody was used as the loading control.

Hematoxylin-eosin staining and immunohistochemistry
Whole testes were fixed in 4% paraformaldehyde and then paraffin sectioned for hematoxylin-eosin (H.E.) staining and immunohistochemistry. For H.E. staining, paraffin sections were counterstained with hematoxylin and eosin and then observed under an electron microscope to compare spermatogonium between Etv5 -/mice and WT littermates. Immunohistochemistry was performed after antigen retrieval with EDTA (pH 9.0), 3% hydrogen peroxide was used for blocking endogenous peroxidase, and 3% BSA was used to block non-specific binding. Sections were incubated with specific antibodies. Anti-promyelocytic leukemia zinc-finger (PLZF) mouse antibody (sc-28319, Santa Cruz, Dallas, TX), a marker of undifferentiated spermatogonia, was used at 1:100 dilutions. HRP-labeled goat anti-mouse IgG diluted at 1:200 was used as the secondary antibody. The sections were then washed and developed using DAB color rendering and nuclear staining. Hematoxylin-stained nucleus that was blue and brownish yellow color indicated positive expression of DAB.

Real time PCR
Total RNA samples were extracted using Eastep Super Total RNA Extraction Kit (LS1040, Promega, Madison, WI) according to the manufacturer's instructions. Total RNA was converted to cDNA using PrimeScript RT reagent Kit with gDNA Eraser (RR047A, Takara, Dalian, China). The mRNA expression levels of the Ccl9, Prm2, and Cyp17a1 genes were then measured by quantitative PCR using PowerUp™ SYBR™ Green Master Mix (A25742, Thermo Fisher Scientific, Austin, TX). β -Actin served as an internal control. Each gene from control and experimental samples was tested in triplicate. Relative gene expression was calculated using the 2 -∆∆CT method.

Measurement of testosterone concentrations
Blood was collected from mice via the orbital vein and then centrifuged for serum collection. Testosterone concentrations in Etv5 -/male and WT control littermates at 12 weeks of age were determined using the mouse testosterone ELISA kit (E05101m, CUSABIO, Wuhan, China). In brief, 50 μ L of standards or samples were added to each well, with blank wells left empty. About 50 μ L of HRP conjugate was added to each well except for the blank well. Subsequently, 50 μ L of antibody was added to each well. The wells were mixed and then incubated for 1 h at 37 °C. Contents from each well were aspirated and washed three times. The assay plate was blotted dry, and 50 μ L of substrate A and 50 μ L of substrate B were added to each well and incubated for 15 min at 37 °C in the dark. After incubation, 50 μ L of stop solution was added to each well, and the optical density (OD) of each well was recorded within 10 min using a microplate reader set to 450 nm.

Semen collection and analysis
Mice were euthanized by cervical dislocation. The unilateral epididymis was surgically removed, cut into pieces, and immersed in 1 mL of SpermRinse™ (Vitrolife, Göteborg, Sweden). After incubation at room temperature, 10 μ L of sperm solution was transferred to a hemocytometer to determine the number and vitality of sperm. The average of three data records from different mice in each group was then determined. The sperm smear test was used to compare the sperm concentrations between Etv5 -/male and WT mice. In brief, approximately 1 mL of sperm solution was centrifuged at 2,000 rpm for 10 min. The resulting precipitate was re-suspended with 500 μ L of SpermRinse™, followed by the addition of trypan blue (1:2 in volume). Slides were sealed using glycerol gelatin after 10 μ L of the sample was placed onto glass slides.

SSC preparation
Testes were harvested from 6 to 8 days postpartum C57BL/6 male pups and then digested using a two-step enzymatic digestion protocol. In brief, testes were washed in DPBS with 2% penicillin-streptomycin, and the tunica albuginea and convoluted epididymis were removed. The seminiferous tubules were digested in 5 mL of DPBS

SSC transplantation
Mice with homozygous mutations of the Etv5 gene were used as recipients for transplantation. Every recipient was transplanted into only one side of the testis, with the other side of the testis used as the non-transplanted control. Mice were anesthetized by an intraperitoneal injection of anesthetic. An appropriate amount of 1.25% 2,2,2-tribromoethanol (M2910, Easycheck, Nanjing, China, 0.2 mL/10 g body weight) was absorbed in a 1 mL sterile syringe. The needle tip of the syringe was faced upward and then pierced the abdominal cavity at a 45° angle. The drug was slowly injected when the tip part could be moved handily. The mouse's toe or paw was stimulated, and waiting for it to faint but could keep breathing steadily. At this moment, the mouse was placed dorsally under a stereomicroscope for the transplantation procedure.
The cuticular layer at the midline of the abdomen, not far from the genitals, was lifted using small forceps. A transverse 0.3 cm incision was made, followed by an incision to the peritoneum and abdominal muscle layer until the peritoneal cavity was visible. The lateral fat pad attached to the testis was gently pulled until the testis was exteriorized. The efferent duct that connected the testis to the epididymis was identified, and fat tissues around the duct were gently removed. The SSC suspension was then carefully transferred to a 100 µL volume microinjection syringe. The syringe was connected into a capillary glass tube with an inner diameter of 40-50 µm at the tip. The cell suspension was gently forced into the seminiferous tubules of the testis via the efferent duct by applying pressure to the syringe. The injection pipette was held parallel to the ordinate axis of the efferent duct. The injection rate and cell suspension flow rate were controlled manually by monitoring the movement of the cell suspension in the tubules. Approximately 5-10 μ L of the donor cell suspension was injected into each recipient testis. The other testis was not surgically manipulated and served as a control. After transplantation, the testis was placed back into the peritoneal cavity. The incision was closed with one stitch of a 7-0 absorbable suture from the inside out, with the abdominal muscle layer first, followed by the peritoneum layer, and finally the cuticular layer. Mice were returned to their cages, monitored for distress, and assessed regularly until testis samples were collected.

Identification of allogeneic SSCs
Testes of recipient mice were harvested 2 months after transplantation to determine the functional recovery of spermatogenesis. Recipient testes were fixed in 4% paraformaldehyde and embedded in paraffin wax for histomorphology and immunohistochemistry analyses. Heterologous spermatogenesis was observed under a fluorescence stereomicroscope by using EGFP-positive sperm obtained from the epididymis of recipient mice. Presence of exogenous genes including Etv5 and EGFP in the testes and spermatozoa were measured by PCR.

Statistical analysis
PASW Statistics 21 (IBM SPSS, Chicago, IL) was used to determine statistical significance and standard deviation. Body weight, testis weight, gene expression level, testosterone concentrations, and cell counts between WT and KO groups were analyzed using two-tailed unpaired t-test. Differences were considered significant at P<0.05 (*) and P<0.01(**).

Generation of Etv5 -/mice by using CRISPR/Cas9
We designed two gRNAs targeting the introns on both sides of exon 1 (gRNA1) to exon 5 (gRNA2) to delete a 5610 bp fragment of the Etv5 gene to generate Etv5 -/mice by embryo injection of Cas9 mRNA and gRNAs (Fig. 1A). The efficiency of CRISPR-induced founder mice is shown in Table 1. The genotypes of the founder and their offspring were identified using primers F/R (Fig. 1A). DNA gel electrophoresis results demonstrated that homozygous KO mice generated a 710 bp band, heterozygotes had two bands (710 and 6326 bp), and wild-type mice had a single band of 6326 bp (Fig. 1B). PCR and DNA sequencing demonstrated that homozygous KO mice had a 5610 bp deletion (Fig. 1C). In addition, Western blot analysis confirmed a lack of Etv5 expression in homozygous Etv5 -/mice (Fig. 1D).
Interestingly, Etv5 -/male mice at 3 weeks of age were obviously smaller than WT male mice, and this difference was more pronounced at 12 weeks of age ( Supplementary Fig. 1A). The body weights of Etv5 -/male mice were markedly lower at each point from 14 postnatal days to 84 days (P<0.05) compared with WT controls, with the magnitude of body weight decline between 34.3% and 45.7% ( Supplementary Fig. 1B).

Lack of SSCs in the seminiferous tubules of Etv5 -/mice
We measured the weight of testes in Etv5 -/mice from day 4 to 12 weeks of age and compared them with that in WT mice of the same age. The weights were almost similar in mice aged before 1 week. However, the testis weights of Etv5 -/mice were markedly lower from 3 to 12 weeks of age (P＜0.01) compared with those of WT controls, with the magnitude of testis weight declining between 43.1% and 67.4% ( Fig. 2A). The testis-to-body-weight ratios were also significantly lower in Etv5 -/mice than in WT mice after 3 weeks of age (Fig. 2B). The difference was also reflected in the sizes of testes; the testes of Etv5 -/mice were much smaller than those of WT mice at 3 weeks of age, and the difference became more evident at 12 weeks of age (Fig. 2C). H.E. staining demonstrated a remarkable decrease in the number and type of testicular cells in the seminiferous tubules of Etv5 -/mice at 3 weeks of age.
Testicular cells were absent in the tubules of Etv5 -/mice, and Sertoli cell-only tubules were observed at 12 weeks of age (Fig. 2D). Additionally, we measured the protein expression levels of PLZF, which is a marker for SSCs. We observed that PLZF was still expressed in Etv5 -/mice in some seminiferous tubules at 3 weeks of age.
However, by 12 weeks of age, SSCs in Etv5 -/had completely disappeared but were present in WT mice (Fig. 2E). In addition, the numbers of premeiotic germ cells, meiotic germ cells, and round spermatids significantly declined in Etv5 -/mice and were undetectable by 12 weeks of age (Fig. 2F). Additionally, the percentage of vacuolation observed in seminiferous tubules in both 3-and 12-week-old Etv5 -/mice were higher than that in WT control mice, reaching 100% at 12 weeks of age (Fig.   2G). The diameter of seminiferous tubules of Etv5 -/mice was significantly smaller than that of WT controls at both 3 and 12 weeks of age (P<0.01; Fig. 2H).
We next examined sperm production in KO mice at different ages. The results were consistent with the findings of H.E. staining and immunohistochemistry. At the time of sexual maturity (6 weeks of age), there were numerous motile sperm in WT mice compared with age-matched Etv5 -/mice, whose sperm numbers were extremely low ( Table 2 and Supplementary Fig. 2). At 12 weeks of age, numerous motile sperm were observed in the epididymis WT mice, but there was a lack of spermatozoa in age-matched Etv5 -/mice ( Table 2 and Supplementary Fig. 2). In addition, we found that the expression levels of the Etv5 target gene Ccl9, spermatid-specific gene Prm2, and interstitial gland-specific gene Cyp17a1 were significantly decreased (P<0.01, P<0.01, and P<0.05, respectively) in Etv5 -/mice (Supplementary Figs. 3A,   3B, and 3C). We also tested testosterone concentrations in WT and Etv5 -/littermates at 6, 8, and 12 weeks. A significant reduction in testosterone concentrations was observed in Etv5 -/mice compared with WT mice, and a maximum difference was reached at 12 weeks of age (P<0.01; Supplementary Fig. 3D).

Transplantation of allogeneic SSCs through the efferent duct
To determine whether the Etv5 -/mouse can serve as a recipient model for SSC transplantation, we transplanted allogeneic SSCs (expressing EGFP by lentiviral transduction) through the efferent duct into immature (3 weeks of age) and mature (7 weeks of age) mice ( Fig. 3A and Table 3). Our preliminary experiments showed that almost all lentiviral transduced SSCs were positive for EGFP expression at a multiplicity of infection of 5:1 (viruses to cells). In nine immature recipients, seven harbored spermatozoa in epididymis at 2 months post-transplantation (Table 3 and Supplementary Fig 4). The average sperm number of the seven restored mice was 9.58 ± 5.05×10 4 ( Table 3). Morphological analysis of the transplanted testes found that the proportion of tubules with restored spermatogonia was 29.51 ± 0.08% in the seven restored testes ( Table 3). Reconstruction of germ cell layers (Fig. 3B, left lower) and regeneration of elongated spermatozoa (Fig. 3B, right) could be observed in these tubules, and spermatogonia were positive for PLZF protein expression (Fig. 3C, transplanted groups) in Etv5 -/mice after SSC transplantation. No spermatozoa and germ cells were detected in non-transplanted immature testes (Figs. 3B and 3C). To determine the origin of the sperm from post-transplanted immature testes, we examined them under a fluorescence microscope. The regenerated sperm emitted green light, indicating they were of donor SSC origin (Figs. 3D). Furthermore, the presence of heterologous genes Etv5 and EGFP was measured in recipient epididymal sperm by PCR. Etv5 and EGFP were not detected in non-transplanted testes; however, they were present in spermatozoa of transplanted Etv5 -/mice (Fig. 3E). This finding further indicated that the sperm were derived from donor cells. In addition, no spermatozoa were detected in all nine mature recipients with SSC transplantation (Table 3).  Chen., 2005). In our study, we successfully generated Etv5 -/mice by using CRISPR/Cas9 system for embryo injection. The generated KO mice gradually lost SSCs but preserved an intact structure of seminiferous tubules with a similar time frame as previous report, indicating that they could be potential germ cell-free models for SSC transplantation.

Discussion
Our KO manipulation revealed that undifferentiated and differentiated spermatogonia were lost in the majority of seminiferous tubules, and only a small region of the tubules retained multilayer germ cells in immature mice. Etv5 -/mice at 3 weeks old had partial germ cell loss in the seminiferous tubules and their testicular sizes were slightly smaller than those of age-matched WT controls. By 12 weeks of age, the testicular sizes of Etv5 -/mice were much smaller than those of WT controls, with a severe germ cell lost in the tubules. Interestingly, 3-week-old Etv5 -/mice were smaller in body size and weight compared with WT mice. At 12 weeks of age, this trend was even more obvious. The reduced body weights in Etv5 -/mice indicated that Etv5 had an influence on overall growth. Etv5 mRNA has been detected in a variety of tissues, including the heart, lungs, thymus, lymphocytes, kidneys, and skeletal muscles (Liu., 2003;T'Sas., 2005). The widespread expression of Etv5 during development may be crucial for growth. In a previous study that investigated the viability of Etv5 -/mice, Etv5 mRNA was found to be abundantly expressed in the brain, lungs, and colon, but it was most abundantly expressed in the testes (Schlesser., 2008 spermatogonia to mature spermatozoa in WT mice (Oakberg., 1956). Transplanted SSCs usually need 2 weeks for a complete colonization. Meiosis is usually initiated within the second month, with several spermatids being observed after 2-month transplantation. The degree of germ cell differentiation will continue to increase, with normal spermatogenesis being observed 3 months post-transplantation (Nagano., 1999). Therefore, the time point of 2 months in our experiment could be not enough for detection of a full round of spermatogenesis following SSC transplantation.
Our results showed SSC transplantation effect between immature and mature Etv5 -/testis is significantly different. SSC transplantation in immature Etv5 -/testis is more feasible than mature Etv5 -/testis, which is in line with previous studies which showed that SSC transplantation in immature pup testes is more efficient than adult testis (Shinohara.,2001;Ishii.,2013). As a lack a fully formed blood-testis barrier, transplantation of testis cells into pup testis has a 5-to 10-fold increased colonization efficiency compared with mature testis. The area for colonization per donor stem cell is also 4 times larger in recipient pups than adults (Shinohara., 2001). These factors facilitate a more efficient restoration of fertility by SSC transplantation in infertile immature recipients. Furthermore, immature Etv5 -/testis at the age of 3 weeks harbors residual endogenous spermatogonia which could facilitate maintenance of testicular function and help restoration of fertility after transplantation (Kanatsu-Shinohara., 2016).
In summary, we generated Etv5 -/mice with a genetic ablation of germline. SSC donors transplanted into recipient testes could partially restore heterologous spermatogenesis and produce donor-derived sperm, even though their quantity and vitality were not optimal. SSC transplantation in this genetically modified mouse models is possible but remains in a low efficiency in the present report. Further investigations are required to optimize the SSC transplantation process in the modified models to obtain ideal transplantation outcomes.      29.51 ± 0.08 7/9 (77.78%) 9.58 ± 5.