Genetic Architecture of Testis and Seminal Vesicle Weights in Mice

Corresponding author: Pierre L. Roubertoux, UPR CNRS 9074, Génétique, Neurogénétique, Comportement, CNRS, 3 B rue de la Férollerie, 45071 Orléans Cedex, France., rouber{at}cnrs-orleans.fr (E-mail)

Communicating editor: J. A. M. VAN ARENDONK

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Comparisons across 13 inbred strains of laboratory mice for reproductive organ (paired seminal vesicles and paired testes) weights indicated a very marked contrast between the C57BL/6By and NZB/BINJ mice. Subsequently these strains were selected to perform a quantitative genetic analysis and full genome scan for seminal vesicle and testis weights. An F₂ population was generated. The quantitative genetic analyses indicated that each was linked to several genes. Sixty-six short sequences for length polymorphism were used as markers in the wide genome scan strategy. For weight of paired testes, heritability was 82.3% of the total variance and five QTL contributed to 72.8% of the total variance. Three reached a highly significant threshold (>4.5) and were mapped on chromosome X (LOD score 9.11), chromosome 4 (LOD score 5.96), chromosome 10 (LOD score 5.81); two QTL were suggested: chromosome 13 (LOD score 3.10) and chromosome 18 (LOD score 2.80). Heritability for weight of seminal vesicles was 50.7%. One QTL was mapped on chromosome 4 (LOD score 9.21) and contributed to 24.2% of the total variance. The distance of this QTL to the centromere encompassed the distance of the QTL linked with testicular weight on chromosome 4, suggesting common genetic mechanisms as expected from correlations in the F₂. Both testis and seminal vesicle weights were associated with a reduction in the NZB/BINJ when this strain carried the Y^NPAR from CBA/H whereas the Y^NPAR from NZB/BINJ in the CBA/H strain did not modify reproductive organ weights, indicating that the Y^NPAR interacts with the non-Y^NPAR genes. The effects generated by this chromosomal region were significant but small in size.

REPRODUCTIVE organ size (testes and seminal vesicles) occupies a place of special importance among morphological measures because of its direct implication in fertility. Reproductive organ sizes are markers of the timing of puberty (ARGYROPOULOS and SHIRE 1989 ) and testicular weight is connected with total sperm count in mice (KRZANOWSKA 1971 ; HUNT and MITTWOCH 1987 ; CHUBB 1992 ). A correlated response to selection also appeared for ovulation rates in lines of mice selected for testicular weight (ISLAM et al. 1976 ). Weight is also an excellent index of either the biochemical or the anatomical state of reproductive organs. The balance and quantity of phosphocreatine and creatine in the fluid secreted by the epithelial cells of seminal vesicles varies according to their developmental stage (LEE et al. 1991 ). Small testis size is often associated with low androgen concentration (MCKINNEY and DESJARDIN 1973 ; CARLIER et al. 1990 ; FRANCOIS et al. 1990 ) and reactivity to testosterone (MICHARD-VANHEE and ROUBERTOUX 1990 ). Testis size may also provide an indirect evaluation of the functionality of connected systems as it was demonstrated for hypothalamic activities (JEAN-FAUCHER et al. 1983 ). Previously reported associations between the fine anatomical structure of reproductive organs and their weight (mice with small testes having a high percentage of abnormal tubules and of reduced Sertoli cells; CHUBB 1992 ) suggest that this simple measure might be used as a marker for complex morphological events. Finally, correlations between reproductive organ sizes and neurobehavioral measures have been reported. Testis weight is higher in mice with the largest brain asymmetries and low direction of laterality assessed from a paw preference test (see COLLINS 1985 , for a review). Mice initiating attack behavior against conspecific males or initiating successful mating behavior (MCKINNEY and DESJARDIN 1973 ) have higher testicular weight.

As reproductive organ weights are the sum of different physiological events, elucidating the putative genes implicated in weight should help to dissect the biological bases of these phenotypes and to understand the mechanisms behind the correlations that have been reported above. Moreover, identifying the genes linked with testis and seminal vesicle weights in mice should have the practical outcome of facilitating the discovery of corresponding human genes via the use of comparative maps and subsequently developing animal models for sterility. Although more than 70 genes are known to be at work in cell activities of testes and 12 for seminal vesicles (MOUSE GENOME DATABASE 1999), their putative involvement in reproductive organ weights remains to be demonstrated.

Very few studies have been performed with seminal vesicle weight in mice, notwithstanding its potential medical interest for modeling prostate cancer. Weight could be related to the abnormal vesicle shape described by SHUKRI et al. 1988 , who identified Svs, one of the corresponding genes on chromosome 7. The implication of the H-2 complex on chromosome 17 has also been suggested (GREGOROVA et al. 1977 ). More attempts have been made to elucidate genes linked with testicular weight. Although four genes are known to contribute to testis differentiation signaling in mice (tdy, tda1, tda2, tda3; MOUSE GENOME DATABASE 1999), the genes underlying the variation in testis weight in mice remain unknown. Marked polymorphisms for weight have been reported in inbred strains of laboratory mice (SHIRE and BARTKE 1972 ; SHUKRI and SHIRE 1989 ; CHUBB 1992 ) and several analyses either using recombinant inbred strains (ARGYROPOULOS and SHIRE 1989 ; SHUKRI and SHIRE 1989 ; CHUBB 1992 ) or generating segregating generations between inbred strains (HUNT and MITTWOCH 1987 ; WASHBURN and EICHER 1989 ) have confirmed that testis weight follows a polygenic transmission. Unfortunately, the few chromosomal linkages that were suggested now appear controversial.

Probably due to its major implication in testis differentiation signaling, the specific part of the Y chromosome was considered as encompassing putative candidates for variation in testis weight, one of these candidates being tdy itself. Results from backcross or intercross designs fit with an involvement of the specific part of the Y chromosome interacting with non-Y genes (HUNT and MITTWOCH 1987 ; WASHBURN and EICHER 1989 ) in testis weight. This Y-chromosome effect was interpreted as a possible contribution of tdy. However, a more recent study (CHUBB 1992 ) has shown that three strains differing in testis size did not reveal any polymorphism for the Y chromosome checked with a Y-specific probe (pY2). This result, indicating no implication of the tdy gene in testis weight, also excluded the contribution of the other genes carried by the Y chromosome in this phenotype. This conclusion might be limited to the C57BL substrains that were used in Chubb's article. Contrasting with other morphological measures that have widely used wide genome scan strategies, only one attempt was made to investigate other regions. With recombinant inbred mice derived from C57BL/6J and DBA/2J (B x D), ZIDEK et al. 1998 found only one highly significant quantitative trait locus (QTL) linked to testis weight that was labeled Twq1, close to the D13Mit3 marker, on chromosome 13. At the D13Mit3 locus, the alleles from C57BL/6J were, surprisingly, linked with high testicular weight and those from DBA/2J with low testicular weight, although C57BL/6J parental strain had lower testis weight than DBA/2J. With the C57BL/6J and DBA/2J strains having almost extreme weights among the set of B x D recombinant inbred strains, few "minoring" alleles in DBA/2J and few "majoring" alleles in C57BL/6J would have been expected. Considering that Twq1 contributed to 75% of the genetic variance, little room would remain for other QTL, indicating an incrementation of testis weight associated with DBA/2J alleles.

For these reasons, we performed a wide genome scan for paired testis and seminal vesicles weights in an F₂ population derived from C57BL/6By (B) and NZB/BINJ (N) strains. These two strains were selected among 13 inbred strains of laboratory mice because they showed large differences for both the measures. Moreover, B and N appeared suitable for the wide genomic scan or QTL mapping because they differed for 92% of the tested simple sequence length polymorphisms (SSLPs; LE ROY et al. 1998 ). We had derived a quartet of congenic strains for the specific part of the Y chromosome (Y^NPAR) from NZB/BINJ (N) and CBA/H (H). This quartet was used to test for the implication of this region in testis and seminal vesicle weights.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

General rearing conditions:
All the mice had been maintained in a pathogen specific free area of our animal facilities under brother x sister mating for several generations when the experiment began. Mice from parental strains, F₁ and F₂ generations, were raised contemporarily. They were reared under the following general conditions: cages, 42 x 20 x 18 cm; bedding, dust-free sawdust; food, IM UAR and tap water ad libitum; temperature, 23° ± 0.5°; photoperiod, 12:12 with lights on at 8 AM; weaning at 29 ± 2 days. Pregnant mothers were isolated from the mating cages. Litters with fewer than seven pups were discarded and the others culled to seven pups to prevent possible litter size effect on reproductive organ weights. At weaning, each male was housed with one NMRI female.

Reproductive organ weights:
Male mice were killed after cervical dislocation. Testes were removed and excised of adhering tissues. To prevent the loss of secretory fluid, the base of each seminal vesicle was grasped with forceps before removing. Paired seminal vesicle and paired testis weights were recorded to the nearest 0.1 mg.

Comparisons across 13 inbred strains of mice:
Reproductive organs were weighed in male mice at 90 ± 5 days of age. They belonged to strains of mice developed from identified breeders and currently used for our experiments on intermale aggression (LE ROY et al. 1999 ). We used 13 strains: A/J, BALB/cJBy, CAST/Ei, C57BL/6By, DBA/2J, and NZB/BINJ purchased from The Jackson Laboratory (Bar Harbor, ME); CBA/H, C57BL/6J, and XLII from CSEAL (Orléans, France); BA and CPB-K provided generously by Dr. Hans van Abeelen (Nijmegen, The Netherlands); and DBA/1Bg and C57BL/10Bg given by Dr. Stephen C. Maxson (Storrs, CT).

Mean values were compared using a one-way ANOVA procedure (SAS INSTITUTE 1987). Nine male mice per strain were used.

Crosses and components of the mean differences:
Parental N and B strains and reciprocal F₁'s were separately compared with a Student's t-test. We used the four reciprocal F₂'s: NB.NBF₂'s and NB.BNF₂'s vs. BN.NBF₂'s and BN.BNF₂'s to test for a maternal contribution on the one hand and NB.NBF₂'s and BN.NBF₂'s vs. NB.BNF₂'s and BN.BNF₂'s to test for an effect of the specific part of the Y chromosome (Y^NPAR) on the other hand (CARLIER et al. 1991 ). This design was analyzed with a two-way ANOVA with the origin of the mother and the origin of Y^NPAR as main factors. Sample sizes are indicated in Table 2.

Components of the mean differences (MATHER and JINKS 1971 ) were estimated using seven parameters: [m] (mean), [d] (additivity), [h] (dominance), [i] (interaction between homozygous loci), [j] (interaction between homozygous and heterozygous loci), [l] (interaction between heterozygous loci), and [om] (origin of the mother). Heritability in the broad sense was estimated as

where V_E = V_N + V_B6 + V_F1 and V_G = V_F2 - V_E.

Mapping QTL:
QTL linked with paired testis and seminal vesicle weights were investigated with a full genome scan in the F₂ intercross population derived from N and B mice. We used 193 males (17 B, 16 N, 18 NBF₁'s, 12 BNF₁'s, 34 NB.NBF₂'s, 28 NB.BNF₂'s, 30 BN.NBF₂'s, and 38 BN.BNF₂'s) maintained under the general conditions described above until they were killed at 150 ± 6 days old. Given that the correlations between body weight and testis or seminal vesicle weight were not significant in the F₂'s (r = 0.09 and r = -0.03), respectively, the absolute paired testis and seminal vesicles weights were used for subsequent analyses. Individual measures were transformed (square root transformation) to ensure homoscedasticity in the nonsegregating generations (untransformed values were reported in the tables).

DNA scoring:
Tails and spleens were collected and stored at -80° until DNA extraction. DNA was extracted from tails and amplified with the usual procedure (SAMBROOK et al. 1989 ): initial denaturation (94° for 3 min and then 94° for 30 sec during each of the 40 amplification cycles), annealing (1 min 15 sec at 42° to 55°, according to the primers), extension (1 min 15 sec at 72°), and final extension step of 3 min. We used a 4% agarose gel (3% NuSieve, 1% Sigma type II agarose) stained with ethidium bromide for visualization. The three possible genotypes were read with an UVP PMW 20 computer system on a screen with 4.5 magnification. The first and last authors performed blind and independent genotyping. Discordant observations resulted in a second amplification.

Full genome scan:
Genotyping was performed individually with the DNA from the F₂ male mice. Sixty-six SSLPs were selected as genetic markers: 5 (chromosomes 1 and 2), 4 (chromosomes 4, 5, 17, 19), and 3 on the others (average interval length 22.5 cM). Significant differences between the three genotypes N//N, N//B, and B//B were assessed with the Kruskal-Wallis test. The chromosomes where these differences reached a P < 0.05 threshold were selected for QTL mapping. Subsequent likelihood ratios and LOD score computations were calculated with the interval mapping method using MapQTL package (VAN OOIJEN and MALIEPAARD 1996 ; MapQTL-tm-version 3.0). The expected confidence interval is expressed as

(DARVASI and SOLLER 1997 ), where 530 is a constant obtained from simulations, N is the number of informative meioses, and v is the proportion of variance explained by the QTL.

We tested for epistatic effects between two QTL using a two-way ANOVA with the values of the three genotypes at the closest SSLPs to the peaks of the QTL and the two QTL as main sources of variation. Epistasis was deduced when an interaction occurred.

Congenic strains for the nonpairing region of the Y chromosome:
To test directly an effect of the nonpairing region of the Y chromosome (Y^NPAR) on reproductive organ weights, the paired testis and seminal vesicle weights were measured in a quartet of congenic strains for this chromosomal region that we had developed from NZB/BINJ (N) and CBA/H (H) (ROUBERTOUX et al. 1994 ). Briefly, the N-Y^NPAR was substituted in the place of the H-Y^NPAR in the H strain to obtain its congenic H.N-Y^NPAR for this region of the Y chromosome and second, the H-Y^NPAR was substituted in the place of the N-Y^NPAR in the N strain to obtain its congenic N.H-Y^NPAR. The congenic N.H-Y^NPAR was developed with N as recipient strain and H as donor, the N females being sired by NHF₁ males and the backcross females subsequently sired again with N. The same design was used to transfer the Y^NPAR from N onto H and to obtain the second congenic H.N-Y^NPAR. At every generation it was assumed that the congenic progeny had lost 50% of the alleles from the congenic donor. Thus, for the measure of reproductive organ weights, few residual allelic forms located throughout the genotype including the Y^PAR of the parental donor strain were expected in the congenic strains at 35 and 37 backcross generations for N.H-Y^NPAR and H.N-Y^NPAR, respectively. Mean values were compared using a two-way ANOVA procedure with the recipient strain (H vs. N) and the origin of the Y^NPAR (H-Y^NPAR vs. N-Y^NPAR) as main sources of variation.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Measurement of reproductive organ weights in 13 inbred strains:
Table 1 presents means ± SEM for the two measures. Strain differences appeared for paired testis weight (F = 6.99, P < 0.0001) and seminal vesicle weight (F = 3.84, P < 0.001).

The N and B strains presented a significant (t = 4.27, P < 0.0001) and marked contrast for testis weight. The difference for seminal vesicle weight was not the largest across the 13 strains even if it remained significant (t = 3.31, P < 0.001).

Components of the mean differences in populations derived from N and B strains:
Reproductive organ weights were measured in new individuals from the N and B strains. Sample sizes and values in parental strains and reciprocal F₁'s and F₂'s are shown in Table 2. Parental strains differed for the two measures (t = 4.75, P < 0.0001, and t = 3.63, P < 0.001, for testis and seminal vesicle weights, respectively), the reproductive organs being heavier in the N than in the B strain. No significant difference was shown (NB.NBF₂'s and BN.NBF₂'s did not differ from NB.BNF₂'s and BN.BNF₂'s) for the origin of the Y^NPAR in the F₂'s. Heritabilities were 0.823 and 0.507 for testis and seminal vesicle weights, respectively, in this population. The difference between the two values may result from the lower contrast between seminal vesicle weights in B and N strains. The best fitting models (² = 8.140; P < 0.10 for testicular weight and ² = 3.015; P < 0.70 for seminal vesicle weight) for estimation of the components of the mean differences were selected according to KERBUSCH et al. 1981 (Table 3). They indicated a polygenic mode of inheritance for both testis and seminal vesicle weights because of a significant [l] interaction between heterozygous loci for the two measures.

The parameters contributing to the mean differences differed for the two phenotypes since a dominance effect was shown for testis and not for seminal vesicle weight, suggesting partially different genetic bases. A contribution of the origin of the mothers appeared for testis weight and not for seminal vesicle weight. The NBF₁ males had higher testis weight than BNF₁ males (t = 4.75, P < 0.01). Although the testes were heavier in males from NB.NBF₂ or NB.BNF₂ mothers than those in males from BN.NBF₂ or BN.BNF₂ mothers, the difference failed to reach significance (P < 0.057). In F₁'s and F₂'s the greater weights were observed for the males from the N mothers or grandmothers. A global estimate for the effect of the origin of the mother was significant for testis weight (Table 3). The product-moment correlation coefficient between testis and seminal vesicle weights in the F₂ population was r = 0.45, P < 0.001.

Mapping QTL:
The three genotypes N//N, N//B, and B//B differed for testis weight (Kruskal-Wallis test) at the P < 0.05 threshold for the following SSLPs: chromosome 4, D4Mit205a (K = 20.89, P < 0.0001), D4Mit12 (K = 16.94, P < 0.0005); chromosome 10, D10Mit20 (K = 6.94, P < 0.05); D10Mit14 (K = 14.57, P < 0.005); chromosome 13, D13Mit3 (K = 8.36, P < 0.01), D13Mit13 (K = 6.76, P < 0.05); chromosome 18, D18Mit17 (K = 9.83, P < 0.005); and chromosome X, DXMit25 (K = 8.87, P < 0.005), DXMit223 (K = 19.29, P < 0.0005). For seminal vesicle weight, differences between the genotypes N//N, N//B, and B//B appeared only for chromosome 4 at D4Mit205a (K = 35.21, P < 0.0001), D4Mit12 (K = 20.71, P < 0.0001). Hence, QTL mapping was performed with MapQTL for chromosomes 4, 10, 13, 18, and X (Fig 1). For paired testis weight, highly significant linkages were reached for the QTL found on chromosomes 4, 10, and X (LOD scores 4.3; LANDER and KRUGLYAK 1995 ) and suggested linkage on chromosomes 13 and 18 (LOD scores 2.8). A highly significant linkage emerged also on chromosome 4 for paired seminal vesicle weight (Fig 1). The sum of the contributions of the QTL to the phenotypic variance was lower for each of the two measures (72.8% for testis weight and 24.2% for seminal vesicles weight) than the corresponding estimated total genetic variance (82.3 and 50.7% for testis and seminal vesicle weights, respectively). The values that were computed for the three genotypes at the SSLP closest to the peak of the QTL indicated that alleles from the N strain contributed to increased testis and seminal vesicles weights (Table 4).

View larger version (19K): In this window In a new window Download PPT slide   Figure 1. LOD plot of paired testis and seminal vesicle weights as calculated by MapQTL (tm) version 3.0 package (VAN OOIJEN and MALIEPAARD 1996 ). Genetic distances from centromere are indicated on the x-axis with SSLPs used as markers for maximum likelihood tests.

No deviation due to dominance was detected for the QTL linked with testis weight. This conclusion fits with inspection of Table 4 indicating that B//N genotype had midvalues between B//B and N//N but not with the results from biometrical analysis. As epistasis is suspected to induce false QTL detection, we performed square root raw data transformation to eliminate the interaction between homozygous loci that appeared in the analyses of the components of the means for the two measures (Table 3). The subsequent analysis showed that epistatic contribution disappeared [l] (did not reach the significance for testis or for seminal vesicle weights) when this transformation was performed. A new QTL analysis with the transformed individual values was carried out. The LOD scores and chromosomal positions of the QTL did not differ from those reported above. The absence of epistasis between the QTL that we detected for testis weight was confirmed by calculating the interactions between the SSLP closest to the peak of the QTL because none of these interactions reached the P < 0.05 threshold.

Contribution of the Y^NPAR:
No effect of the Y^NPAR was detected from the analysis of reciprocal F₂'s, whereas it appeared in a quartet of congenic strains for the Y^NPAR (Table 5). The recipient strain (N vs. H) contributed to testis and seminal vesicle weights (F = 55.23, P < 0.0001 and F = 96.19, P < 0.0001, respectively). The Y^NPAR effect was present in interaction with the recipient strain for seminal vesicle weight (F = 8.97, P < 0.0037) and testis weight (F = 7.18, P < 0.009).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We reported large differences related to genetic variability in testis and seminal vesicle weights in a population of 13 inbred strains of mice. Genetic analyses were performed with different complementary approaches.

Contribution of the Y^NPAR to paired testis and seminal vesicle weights:
The Y^NPAR contributed to reproductive organ weights as shown by the results obtained from a quartet of congenic strains, for this chromosomal region, using H and N as parental strains. The Y^NPAR from the N strain that had heavier testes increased testicular weight in the H strain and the transfer of Y^NPAR from the H strain reduced testicular weight in N. For the seminal vesicle weight, the N strain was more reactive because the Y^NPAR from the H strain reduced this weight on the N background, whereas it had no significant effect on H. The sizes of the effects were small and contributed to only a small part of the difference observed between the parental strains, suggesting that non-Y^NPAR regions contribute to both testis and seminal vesicle weight.

The non-Y^NPAR contribution to reproductive organ weights was investigated generating the two reciprocal F₁'s and the four reciprocal F₂'s between the N and B parental strains, these strains being among the most contrasted strains out of a sample of 13 inbred strains.

QTL for paired testis weights:
As in recent studies that had used backcross and intercross generations, we found a polygenic inheritance of testis weight, encouraging a wide genome scan to investigate QTL. In the present study, five QTL contributing to 72.8% of the total variance were detected for testis weight, the total genetic variance being 0.823.

Three QTL involved in testis weight had the values corresponding to a highly significant level (LOD score 4.3; LANDER and KRUGLYAK 1995 ). The strong genetic contribution that appeared for the QTL mapped on the X chromosome is compatible with the previously published results from a biometrics analysis (HUNT and MITWOCH 1987) and with the observed difference between our reciprocal F₁'s (Table 2). The other highly significant QTL mapped on chromosomes 4 and 10. We have also found two "suggested" QTL on chromosomes 13 and 18. The QTL mapped on chromosome 13 was at 13.4 ± 15.10 cM from the peak of Twq1 (ZIDEK et al. 1998 ), which is a debated QTL, for the reasons presented above. It is, moreover, at the limit of the confidence interval of our suggested QTL. ZIDEK et al. 1998 had also reported a suggested difference between the genotypes B//B and D//D for D18Mit19 at 2 cM from the centromere. The mice carrying D//D alleles had, as expected, heavier testes than those with B//B. However, even if the LOD score found by Zideck and the LOD score detected in our study were low, the confidence intervals of the two QTL overlapped.

QTL for paired seminal vesicle weights:
Heritability was lower for this organ (50.7%) than for testis weight, this difference resulting, probably, from the smaller contrast between the N and B6 strains. The biometrics analysis performed with the mice that were used for testicular weights indicated a polygenic inheritance because an epistatic component ([l], interaction between heterozygous loci) reached significance. For this reason, chromosomal regions linked to paired seminal vesicle weights were investigated. In the present analysis, a substantial QTL (24.2% of the total variance, half of the genetic variance) has been detected on chromosome 4. Its distance from the centromere (47.5 cM) did not differ from the distance obtained for testicular weight on chromosome 4. Thus testis and seminal vesicle weights could be linked to the same QTL as expected from correlations in the F₂ population.

Candidate genes:
Investigation of candidate genes always implies uncertainty in F₂'s or recombinant inbred strains because the confidence interval is large, due to the reduced number of informative meioses, and hence encompassing a high number of genes. However, given that the physiological bases of the reproductive organ weights are documented, they may pave the way to suggesting potential candidates. Briefly, the hypothalamic-pituitary axis is related to testes via two pathways. The morphogenic effect of follicle-stimulating hormone (FSH) on testes is activated by the development of epithelial tissues in seminiferous tubules and proliferation of Sertoli cells. This process is perinatal because the Sertoli cells' proliferation stops at about 12 days after birth (KLUIN et al. 1984 ). Inhibin, testosterone, 17ß-estradiol, and dihydrotestosterone ensure a feedbackloop from Sertoli cells to hypothalamic-pituitary axis. A second pathway starting from this axis via luteinizing hormone reaches Leydig cells with a feedback loop to the hypothalamic-pituitary axis by testosterone and 17ß-estradiol. The implication of hypothalamic-pituitary axis on these pathways is modulated by several factors. Leptin reduces its sensitivity, modifying the fasting-induced inhibition of gonadotropin releasing hormone (BARASH et al. 1996 ) acting on frequency and amplitude of pulses of FSH. The implication of leptin in reproductive organ weights has been directly observed because leptin injection increases epithelial heights, producing an augmentation of both testes and seminal vesicles in mice (BARASH et al. 1996 ). Candidate genes implicated in these mechanisms and included in the confidence intervals of putative QTL were screened using the MOUSE GENOME DATABASE (1999).

For the QTL linked with testis and seminal vesicle weights, the leptin receptor gene (Lepr) may be a common candidate, as leptin is involved in reproductive organ weights. The Lepr gene is mapped on chromosome 4 at 46.7 cM from the centromere, this distance being included in the confidence interval of both these QTL (48 and 47.5 cM from the centromere, respectively).

The confidence interval of the QTL mapped on the X chromosome (38.6 cM) encompasses the androgen receptor gene (Ar), which is mapped at 36 cM from the centromere on the X chromosome (MOUSE GENOME DATABASE 1999). Ar gene is a member of the nuclear receptor superfamily that acts as a ligand-dependent tissue, specific transcription factor (MAGELSDORF et al. 1995 ). It is activated by binding testosterone or dihydrotestosterone. A single point mutation in the N-terminal region of Ar results in a premature stop codon leading to the expression of a nonfunctional truncated form of Ar (GASPAR et al. 1991 ). Due to X-link, Tfm mice are genetically males with smaller testes that are consistently found in the inguinal region upon dissection (HUTSON et al. 1994 ). During development, the fetal testes secrete testosterone and antimulerian hormones, which are essential to proper differentiation and growth of the male reproductive tract. During later sexual maturation, dihydrotestosterone, the more potent testosterone metabolite, results in virilization of external genitalia.

The susceptibility to testicular teratomas depends on the Ter gene in the 129 strain. It was mapped on chromosome 18 near D18mit 62 (ASADA et al. 1994 ; SAKURAI et al. 1995 ), which is close to the peak of the QTL that we found on this chromosome. As the Ter gene reduces germ cells in many strains of mice and was mapped in crosses derived from males having small vs. normal-sized testes, it could be a candidate for the QTL that we mapped on chromosome 18. The contribution of glucocorticoid receptor 1 (Grl-1) remains open (SAKURAI et al. 1995 ).

Conclusions:
The present QTL analysis led to describing five chromosomal regions implicated in testis weight and one implicated in seminal vesicle weight. The genetic contribution of the QTL linked to testis weight reached the heritability value, suggesting that a small number of QTL were undetected in the F₂ population. The picture was different for seminal vesicle weight for which 26.5% of the genetic variance was uncovered by the QTL described herein. The contribution of the specific part of the Y chromosome to reproductive organ weights is not in contradiction with this conclusion. Several genes are linked to Y^NPAR and other genes such as Tdy could be implicated in this phenotype.

	ACKNOWLEDGMENTS

This work was supported by the CNRS (UPR 9074), Ministry for Research and Technology (Paris V-René Descartes and University of Orléans), Région Centre and Préfecture de la Région Centre, and Fondation pour la Recherche Médicale. UPR 9074 is affiliated with INSERM and the University of Orléans.

Manuscript received March 28, 2000; Accepted for publication January 24, 2001.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ARGYROPOULOS, G. and J. G. M. SHIRE, 1989  Genotypic effects on gonadal size in fetal mice. J. Reprod. Fertil. 86:473-478[Abstract/Free Full Text].

ASADA, Y., D. S. VARNUM, W. N. FRANKEL, and J. H. NADEAU, 1994  A mutation in the Ter gene causing increased susceptibility to testicular teratomas maps to mouse chromosome 18. Nat. Genet. 4:363-368.

BARASH, I., C. C. CHEUNG, D. S. WEIGLE, H. REN, and E. B. KABIGTING et al., 1996  Leptin is a metabolic signal to the reproductive system. Endocrinology 137:3144-3147[Abstract].

CARLIER, M., P. L. ROUBERTOUX, M.-L. KOTTLER, and H. DEGRELLE, 1990  Y-chromosome and aggression in strains of laboratory mice. Behav. Genet. 20:137-156[Medline].

CARLIER, M., P. L. ROUBERTOUX, and C. PASTORET, 1991  The Y chromosome effect on intermale aggression in mice depends on the maternal environments. Genetics 129:231-236[Abstract].

CHUBB, C., 1992  Genes regulating testis size. Biol. Reprod. 47:29-36[Abstract].

COLLINS, R. L., 1985 On the inheritance of the direction and the degree of asymmetry, pp. 41–71 in Cerebral Lateralization in Nonhuman Species, edited by S. GLICK. Academic Press, New York.

DARVASI, A. and M. SOLLER, 1997  A simple method to calculate resolving power and confidence interval of QTL map location. Behav. Genet. 27:125-132[Medline].

FRANÇOIS, M. H., M. NOSTEN-BERTRAND, P. L. ROUBERTOUX, M. L. KOTTLER, and H. DEGRELLE, 1990  Opponent strain effect on eliciting attacks in NZB mice: physiological correlates. Physiol. Behav. 47:1181-1185[Medline].

GASPAR, M. L., T. MEO, P. BOURGAREL, J. L. GUENET, and M. TOSI, 1991  A single base deletion in the Tfm androgen receptor gene creates a short-lived messenger RNA that directs internal translation initiation. Proc. Natl. Acad. Sci. USA 19:8606-8610.

GREGOROVA, S., P. IVANYI, M. SIMONOVA, and M. MICHOVA, 1977  H-2 associated differences in androgen-influenced organ weights of A and C57BL/10 mouse strains and their crosses. Immunogenetics 4:301-313.

HUNT, S. E. and U. MITTWOCH, 1987  Y-chromosomal and other factors in the development of testis size in mice. Genet. Res. 50:205-211[Medline].

HUTSON, J. M., M. BACKER, M. TERADA, B. ZHOU, and G. PAXTON, 1994  Hormonal control of testicular descent and the cause of cryptorchidism. Reprod. Fertil. Dev. 6:151-156[Medline].

ISLAM, A. B. M., W. G. HILL, and R. B. LAND, 1976  Ovulation rate of lines of mice selected for testis weight. Genet. Res. 27:23-32[Medline].

JEAN-FAUCHER, C., M. BERGER, M. DE TURCKHEIM, G. VEYSSIERE, and C. JEAN, 1983  Testicular response to HCG stimulation and sexual maturation in mice. Horm. Res. 17:216-221[Medline].

KERBUSCH, J. M. L., F. J. VAN DER STAAY, and N. HENDRICKS, 1981  A searching procedure for transformation and models in a classical mendelian cross breeding study. Behav. Genet. 11:239-254[Medline].

KLUIN, P. H. M., M. F. KRAMER, and D. G. DE ROOIJ, 1984  Proliferation of spermatogonia and sertoli cells in maturing mice. Anat. Embryol. 169:173-178.

KRZANOWSKA, H., 1971 Influence of Y chromosome on fertility in mice, pp. 370–386 in Edinburgh Symposium on the Genetics of Spermatozoon, edited by R. A. BEATLY and S. GLUECKSOHN-WALSH. The University Press, Edinburgh.

LANDER, E. S. and L. KRUGLYAK, 1995  Genetic dissection of complex traits: guidelines for interpreting and reporting linkage results. Nat. Genet. 11:241-247[Medline].

LEE, H., C. GONG, S. WU, and M. R. IYRENGAR, 1991  Anulation of phosphocreatine and creatine in the cells and fluid of mouse seminal vesicles is regulated by testosterone. Biol. Reprod. 44:540-545[Abstract].

LE ROY, I., P. L. ROUBERTOUX, L. JAMOT, F. MAAROUF, and S. TORDJMAN et al., 1998  Neuronal and behavioral differences between Mus musculus domesticus (C57BL/6JBy) and Mus musculus castaneus (CAST/Ei). Behav. Brain Res. 95:135-142[Medline].

LE ROY, I., S. MORTAUD, S. TORDJMAN, E. DONSEZ-DARCEL, and M. CARLIER et al., 1999  Correlation between expression of the steroid sulfatase gene, mapped on the pairing region of the Y-chromosome, and initiation of attack behavior in mice. Behav. Genet. 29:131-136[Medline].

MAGELSDORF, D. J., C. THUMMEL, M. BEATO, P. HERRLICH, and G. SCHUTZ et al., 1995  The nuclear receptor superfamily: the second decade. Cell 83:835-839[Medline].

MATHER, K., and J. L. JINKS, 1971 Biometrical Genetics, Ed. 2. Chapman & Hall, London.

MCKINNEY, T. D. and C. DESJARDIN, 1973  Postnatal development of the testis, fighting behavior, and fertility in house mice. Biol. Reprod. 9:279-294[Abstract].

MICHARD-VANHEE, C. and P. L. ROUBERTOUX, 1990  Genetic analysis of differences in behavioral reactivity to neonatal injection of testosterone in female mice. Behav. Genet. 19:63-71.

MOUSE GENOME DATABASE, 1999 Mouse Genome Informatics, The Jackson Laboratory, Bar Harbor, ME. World Wide Web (URL: http://www.informatics.jax.org).

ROUBERTOUX, P. L., M. CARLIER, H. DEGRELLE, M. C. HAAS-DUPERTUIS, and J. PHILLIPS et al., 1994  Co-segregation of the pseudoautosomal region of the Y chromosome with aggression in mice. Genetics 136:225-230[Abstract].

SAKURAI, T., T. IGUCHI, K. MORIWAKI, and M. NOGUCHI, 1995  The ter mutation first causes primordial germ cell deficiency in ter/ter mouse embryos at 8 days of gestation. Dev. Growth Differ. 37:293-302.

SAMBROOK, J., E. FRITSCH and T. MANIATIS, 1989 Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

SAS INSTITUTE, 1987 SAS/STAT User's Guide, Version 6. SAS Institute Inc., Cary, NC.

SHIRE, J. G. M. and A. BARTKE, 1972  Strain differences in testicular weight and spermatogenesis with special references to C57bl/10J and DBA/2J mice. J. Endocrinol. 55:163-171[Abstract/Free Full Text].

SHUKRI, N. M. and J. G. SHIRE, 1989  Genetic variation in testicular development in mice of the C57BL/10ScSn, C57BL/6By and BALB/cBy strains and the CXB recombinant inbred lines. J. Reprod. Fertil. 87:587-592[Abstract/Free Full Text].

SHUKRI, N. M., F. GREW, and J. G. SHIRE, 1988  Recessive mutation in a standard recombinant-inbred line of mice affects seminal vesicle shape. Genet. Res. 52:27-32[Medline].

VAN OOIJEN, J. W., and C. MALIEPAARD, 1996 MapQTL (tm) version 3.0, Software for the calculation of QTL positions on genetic maps. CPRO-DLO, Wageningen, The Netherlands.

WASHBURN, L. L. and E. M. EICHER, 1989  Normal testis determination in the mouse depends on genetic interaction of a locus on chromosome 17 and the Y chromosome. Genetics 123:173-179[Abstract/Free Full Text].

ZIDEK, V., A. MUSILOVA, J. PINTIR, M. SIMAKOVA, and M. PRAVENEC, 1998  Genetic dissection of testicular weight in the mouse with the BXD recombinant inbred strains. Mamm. Genome 9:503-505[Medline].

This article has been cited by other articles:

				J. M. Good, M. D. Dean, and M. W. Nachman A Complex Genetic Basis to X-Linked Hybrid Male Sterility Between Two Species of House Mice Genetics, August 1, 2008; 179(4): 2213 - 2228. [Abstract] [Full Text] [PDF]

				D. L'Hote, C. Serres, P. Laissue, A. Oulmouden, C. Rogel-Gaillard, X. Montagutelli, and D. Vaiman Centimorgan-Range One-Step Mapping of Fertility Traits Using Interspecific Recombinant Congenic Mice Genetics, July 1, 2007; 176(3): 1907 - 1921. [Abstract] [Full Text] [PDF]

				P. L. Roubertoux, I. Le Roy, S. Tordjman, A. Cherfou, and D. Migliore-Samour Analysis of Quantitative Trait Loci for Behavioral Laterality in Mice Genetics, March 1, 2003; 163(3): 1023 - 1030. [Abstract] [Full Text] [PDF]