All animals were housed under 12 hr light-dark cycles with food and water provided ad libitum. All mice were maintained as required under the National Institutes of Health guidelines for the Care and Use of Laboratory Animals. All studies have been approved by the Animal Care and Use Committee of the University of Louisville. All the mice were sacrificed under ketamine anesthesia and all efforts were made to minimize their suffering.

To specifically investigate the role of each Fgfr1 and Fgfr2 in germ cells, a transgenic Cre mouse line expressing an improved Cre (iCre) recombinase driven by the mouse Tex101 promoter (Tex101-iCre) was used. This transgenic line was previously generated in our laboratory (30). The expression of iCre was specifically detected in the prespermtogonia within seminiferous tubules of postnatal eight-day-old testes. In adult mice, there were robust iCre activities in spermatocytes and spermatids and a weak activity in spermatogonia. For germ cell selective deletion of Fgfr1 or Fgfr2, Tex101-iCre female mice were first bred with Fgfr1^flox/flox or Fgfr2^flox/flox males to obtain bigenic heterozygous females (i.e.,Tex101-iCre;Fgfr1^flox/+ and Tex101-iCre;Fgfr2^flox/+). Then, these heterozygous females were bred with Fgfr1^flox/flox or Fgfr2 ^flox/flox to generate male germ cell-specific Fgfr1 or Fgfr2 mutant mice (male Tex101-iCre;Fgfr1^flox/flox or Tex101-iCre;Fgfr2^flox/flox). Floxed Fgfr1 and Fgfr2 mice were kindly provided by Dr. Juha Partanen (floxed Fgfr1 mice, University of Helsinki, Finland) and Dr. David M. Ornitz (floxed Fgfr2 mice, Washington University in St Louis) and details were described elsewhere (23–25). To determine the efficiency of Tex101-iCre in excision of floxed Fgfr1 and Fgfr2 alleles, Tex101-iCre;Fgfr1^flox/flox and Tex101-iCre;Fgfr^flox/flox male mice were mated with wild-type females.

Genomic DNA was isolated from mouse tails using proteinase K and phenol chloroform extraction method as described previously (30). The presence of iCre or LacZ was determined by PCR using the primer pairs listed in Table 1. The primer sets Fgfr1^Δ and Fgfr2^Δ were used to determine the deletion of floxed Fgfr1 and Fgfr2 alleles.

Testicular cells were isolated from three 2-month-old mice using the procedure described previously (31) with a little modification. Briefly, the testes were decapsulated and incubated with a collagenase type II solution (0.5 mg/ml, Sigma, St. Louis, MO) to separate interstitial cells and seminiferous tubules. The interstitial cells were pelleted by centrifugation. To obtain the germ cells and Sertoli cells, the dispersed seminiferous tubules were cut into small pieces and digested with a solution containing 1 mg/ml trypsin (Sigma) and 10 μg/ml DNase I (Sigma) at 32^oC for 30 min. The reaction was stopped by adding trypsin inhibitor (Sigma) and Hank's Balanced Salt Solution (HBSS, Invitrogen, Carlsbad, CA). The supernatant that contained germ cells was collected after precipitation by unit gravity. The pellet was incubated with a collagenase type II solution at 32^oC for 15 min and settled down by unit gravity. The cell pellet that contained Sertoli cells was rinsed with HBSS three times and cultured with Dulbecco's Modified Eagle's Nutrient Mixture/F12 Ham Medium supplemented with 10% fetal bovine serum (Invitrogen) overnight. Sertoli cells were harvested the next day after residual germ cells were hypotonically removed. The purity of isolated interstitial, Sertoli and germ cells was evaluated by performing RT-PCR using several putative marker genes, which included cholesterol side-chain cleavage enzyme and 17α-hydroxylase (interstitial cells), follicle stimulating hormone receptor and Pem homeobox gene (Sertoli cells), alkaline phosphatase and fibronectin (myoid cells) and protamine 2 and stimulated by retinoic acid gene 8 homolog (germ cells) (32, 33). The results showed that the contamination of each cell type by the others was minimal (data not shown).

Total RNA was extracted from the testes and the isolated testicular cells using Trizol Reagent (Invitrogen) according to manufacturer's instructions. Total RNA was adjusted to a concentration of approximately 1.0 μg/μl. Two microgram of total RNA was reverse transcribed into cDNA with random primers (Invitrogen) and avian myeloblastosis virus (AMV) reverse transcriptase (Promega Corporation, Madison, WI). The cDNA was amplified by PCR with the primer sets of the target gene and a housekeeping gene, ribosomal protein large subunit 19 (Rpl19). PCR primers, as listed in Table 1, were designed according to the sequences obtained from GenBank using the Vector NTI 12.0 program (Invitrogen) and synthesized by Operon Technologies (Alameda, CA). All primers were designed to amplify all variants of Fgfr1 and Fgfr2 and the products covered one or more exons. The amplified products were separated by electrophoresis and the intensity of specific bands was scanned and semi-quantified using the image analysis software, TotalLab (Nonlinear USA Inc, Durham, NC). The results were presented as the ratio of target gene over Rpl19.

The testes were homogenized by sonication in an ice-cold lysis buffer. The protein concentrations were measured by the Bradford method (Bio-Rad laboratories, Hercules, CA). Protein aliquots were separated on SDS-PAGE gels, transferred to PVDF membranes, blocked with 3% non-fat milk, and then incubated overnight with rabbit polyclonal antibodies against FGFR1 (sc-121, 1:400) and FGFR2 (sc-122, 1:600) (Santa Cruz Biotechnology, Santa Cruz, CA), respectively. Peroxidase-conjugated anti-rabbit IgG (1:2000, Vector Laboratories, Burlingame, CA) was used as the secondary antibody. Immunoblotting signals were detected by Amersham ECL plus Western blotting detection system (GE healthcare Biosciences, Pittsburgh, PA). All membranes were re-blotted with β-actin or β-tubulin antibodies (Sigma) as the loading control. The intensity of specific bands was scanned using image analysis software, TotalLab (Nonlinear USA Inc). The results were presented as the ratio of target protein over β-actin or β-tubulin.

Tissues were fixed in 10% formalin and embedded in paraffin. The procedure was performed by an avidin-biotin immunoperoxidase method as described previously (34). Briefly, sections were de-waxed, rehydrated and then incubated with 1% H₂O₂ for 30 min. After rinsing with phosphate buffered saline (PBS), sections were treated with 0.025% trypsin (Sigma) for 30 min at room temperature and incubated with rabbit polyclonal antibodies against FGFR1 (1:50) and FGFR2 (1:200) (Santa Cruz Biotechnology) overnight at 4°C, respectively. Sections were then incubated with biotinylated goat anti-rabbit IgG (Vector Laboratories, Burlingame, CA) for 1 hr. After rinsing with PBS, sections were incubated with avidin-biotin-horseradish peroxidase complex using a Vectastain ABC kit (PK-4000, Vector Laboratories) for 1 hr and rinsed with PBS. Immunostaining was detected by incubation of the sections with the substrate 3'3-diaminobenzidine. All sections were counterstained with hematoxylin. Replacement of the primary antibody with PBS was performed at the same time as a procedure control. We also used irrelevant rabbit IgG instead of the primary antibody later to check the specificity of the immuno-staining. Immunostained sections were evaluated by two researchers independently. The stage of the seminiferous epithelium cycle was determined by the morphology and localization of spermatocytes and spermatids within the seminiferous tubule as described by Hess and Renato de Franca (35).

Four Tex101-iCre;Fgfr1^flox/flox and four Tex101-iCre;Fgfr2^flox/flox male mice at age of 3 months were housed individually with proven female breeders. The females were separated from males once they were pregnant. The breeding test for males continued for a period of 4 months. The resultant pregnancies were noted and the number of pups for each litter was recorded.

The data presented are the means±SEM. All results were analyzed by Kruskal-Wallis test and Dunn's post-test using a version 3.06 Instat program (Graphpad Software, San Diego, CA). A p-value <0.05 was considered statistically significant.

To identify which cell types expressed Fgfr1 and Fgfr2 in adult mouse testes and whether the expression of these two receptors changed during postnatal development, RT-PCR was performed. The results indicated that the transcripts of Fgfr1and Fgfr1 were present in germ cells, Sertoli cells and interstitial cells in adult mouse testes (Figure 1A). The mRNAs of these two receptors were readily detected in the testes during the entire postnatal development period examined ranging from neonatal (day-1), immature (day-10), peripubertal (day-20), pubertal (day-30) to adulthood (day-60) (Figures 1B and C). The testicular mRNA levels of both Fgfr1 (Figure 1B) and Fgfr2 (Figure 1C) remained constant from neonatal to pubertal period and then significantly decreased in adult testes (n=3, p<0.05 compared to day-1 to -30).

To determine whether the levels of FGFR1 and FGFR2 proteins were also postnatally regulated in the testes, Western blot analyses were carried out. In contrast to their mRNA profiles, immuno-blotting demonstrated that the protein levels of FGFR1 were low during the neonatal and immature stage, increased in the peripubertal period and then maintained a steady level to adulthood (n=3, p<0.05 compared to day-20 to-60, Figure 2A). The protein levels of FGFR2, on the other hand, showed no apparent changes from the neonatal period to adulthood (n=3, p>0.05, Figure 2B). The mechanism by which the mRNA and protein levels of Fgfr1 and Fgfr2 in sexually mature animals is differentially regulated is currently unknown.

Immunohistochemical staining of testicular sections revealed that FGFR1 (Figure 3A) and FGFR2 (Figure 3B) were detected in both the interstitial and seminiferous tubular compartments from neonatal to adulthood. The most prominent immunostaining of FGFR1 and FGFR2 was observed in interstitial and peritubular cells with little or no changes in all phases of postnatal development. In the seminiferous tubular compartment, weak immunostaining of both receptors was present in spermatogonia, spermatocytes and Sertoli cells throughout all phases of postnatal development. The immuno-staining for FGFR1 and FGFR2 was also evident in elongated spermatid, spermatozoa, the seminiferous tubules and sperm in the epididymis but weak immunostaining was seen in round spermatid in adult testes (Figures 3A and B). Differential immunostaining intensity of FGFR1 and FGFR2 was observed in seminiferous epithelial cycle. It appeared to be lower in stages IX and X and higher in stages I-VIII.

First, the efficiency of Tex101-iCre in deletion of floxed Fgfr1 and Fgfr2 was evaluated by breeding Tex101-iCre;Fgfr1^flox/flox and Tex101-iCre;Fgfr2^flox/flox males with wild-type females, respectively, and genotyping analysis of the progenies was performed. If iCre recombinase is active in the spermatogenic cells, the floxed Fgfr1 and Fgfr2 alleles will be converted to the recombined Fgfr1^Δ and Fgfr2^Δ allele regardless of the presence or absence of Tex101-iCre transgene in the progenies. The results showed that lack of Fgfr1^flox (Figure 4A) and Fgfr2^flox (Figure 4B) alleles and presence of Fgfr1^Δ and Fgfr2^Δ alleles in all pups indicated complete deletion of the floxed Fgfr1 or Fgfr2 alleles in the male germline.

Complete deletion of Fgfr1 and Fgfr2 in the germ cells of Tex101-iCre;Fgfr1^flox/flox and Tex 101-iCre;Fgfr2^flox/flox males was further confirmed by performing RT-PCR and immunohistochemistry. RT-PCR showed that the transcripts of testicular germ cells of Fgfr1 in Tex101-iCre; Fgfr^flox/flox (Figure 5A) and Fgfr2 in Tex101-iCre; Fgfr2^flox/flox (Figure 5B) animals were not detectable. Immunohistochemistry demonstrated the absence of immunostaining of Tex101-iCre; Fgfr1^flox/flox (Figure 6A) and FGFR2 in Tex101-iCre;Fgfr2^flox/flox (Figure 6D) mice, while the immunostaining of these two proteins in testicular somatic cells for either genotype animal was comparable to wild-type siblings (Figures 6B and E).

Both Tex101-iCre;Fgfr1^flox/flox and Tex101-iCre; Fgfr2^flox/flox mice were viable with no apparent developmental defects. To test the fertility of Fgfr1 and Fgfr2 mutant male mice, sexually mature Tex101-iCre;Fgfr1^flox/flox and Tex101-iCre; Fgfr2^flox/flox male mice were mated with wild-type female mice, respectively. All of the wild-type mice tested delivered pups with normal litter sizes. Average litters sired by Tex101-iCre; Fgfr1^flox/flox and Tex101-iCre;Fgfr2 ^flox/flox male mice during four months of fertility tests did not significantly differ from wild-type siblings (n=4, p>0.05, Table 2), indicating that germ cell-selective ablation of individual Fgfr1 or Fgfr2 in mice does not affect male fertility.

There were no gross abnormalities and size difference in the testes and accessary sex organs in either Fgfr1 or Fgfr2 mutant mice. Light microscopy revealed that in wild-type and mutant mice, all stages of spermatogenesis were present. The size of the seminiferous tubules, the histological structures of the testes and epididymides of wild-type (Figures 7C, F, I), Tex101-iCre;Fgfr1^flox/flox (Figures 7A, D, G) and Tex101-iCre;Fgfr2^flox/flox (Figures 7B, E, H) were essentially indistinguishable.

To explore possible effects of germ cell-selective deletion of Fgfr1 and Fgfr2 on the expression of other Fgfrs, RT-PCR was carried out to determine their mRNA levels in adult testes. The results showed that Fgfr1 to Fgfr5 were expressed in the germ cells of adult testes. Deletion of Fgfr1 in germ cells led to a moderate elevation of Fgfr4 mRNA levels, while the expression of Fgfr3 and Fgfr5 was not affected (Figure 5A). Semiquantitative RT-PCR also showed that deletion of Fgfr2 in germ cells did not significantly influence the expression of Fgfr1, Fgfr3, Fgfr4 and Fgfr5 in the adult testes (Figure 5B).