Genetic dissection of testicular weight in the mouse with the BXD
recombinant inbred strains
Institute of Physiology, Academy of Sciences of the Czech Republic, Viden
ska´ 1083, 14220 Prague 4, Czech Republic
Institute of Biology, 1st Medical Faculty, Charles University, Prague, Czech Republic
Received: 23 January 1998 / Accepted: 16 March 1998
Abstract. Testicular weights were studied in the mouse BXD
recombinant inbred (RI) strains. These strains were derived from
DBA/2J and C57BL/6J progenitors that differ significantly in their
testicular weights (0.224 g ± 0.015 vs. 0.161 g ± 0.03, P < 0.0001).
The heritability of testicular weights was calculated to be 0.53, and
the minimum number of responsible effective factors was esti-
mated to be 5.7. The total genome scanning of the BXD RI strains
with over 1000 markers revealed a quantitative trait locus (QTL)
on mouse Chromosome (Chr) 13 near the D13Mit3 marker (LOD
score 6.9). This QTL region was designated Twq1 and associated
with over 75% of genetic variability.
The C57BL/6J and DBA/2J progenitor strains differ significantly
in their testicular weights, and it has been demonstrated that ge-
netic variation that affects this phenotype is under control of sev-
eral genes (Chub 1992). Since genetic dissection of testicular
weights could shed light on the processes of testicular develop-
ment, we decided to use the BXD recombinant inbred (RI) strains
for positional mapping of responsible quantitative trait loci (QTL).
Results of the current study demonstrate the presence of a QTL
with a major effect on testicular weight on mouse Chr 13.
Materials and methods
Males of the C57BL/6J (N ס 8) and DBA/2J (N ס 9) pro-
genitor strains were purchased from VELAZ (Prague, Czech Republic).
The males of the 24 BXD RI strains (four to five males per each strain)
were purchased from The Jackson Laboratory (Bar Harbor, Maine). Ani-
mals were bred under standard laboratory conditions. At the age of 11
weeks, the animals were sacrificed and their testicular weights were de-
termined. Since testicular weights showed no significant correlation with
body weight, absolute testicular weights were used for the analysis.
Values are expressed as means ± SD. Heritability of the tes-
ticular weight was estimated according to the method of Plomin and Mc-
Clearn (1993) from the variance in mean values between and within the RI
strains. The additive genetic variance was estimated to be 50% of the total
variance between the mean testicular weight of the RI strains; the envi-
ronmental variance was estimated to be the average variance in mean
testicular weight within the RI strains. Narrow heritability was calculated
by dividing the additive genetic variance by the sum of the additive genetic
variance and the environmental variance. The number of effective factors
in RI strains was estimated according to the equation L ס D
D is the difference between the highest and the lowest mean and V is the
variance of the means of all RI strains and progenitor strains of the set
(Bailey 1981). Map Manager QT program (version b18; Manly 1997) was
used to test for single locus associations by regression analysis, and the
significance of each potential association was measured with the likelihood
ratio statistics (LRS) of Haley and Knott (1992). The interval regression
method of Map Manager QT was used to test for QTLs within marker
intervals. The significance threshold for the genome-wide scan was em-
pirically determined by the Map Manager QT permutation test with the
informative markers and 1000 permuted data sets as recommended by
Doerge and Churchill (1996): the likelihood ratio scores of 7.4, 19.5, and
31.5 were used for detection of suggestive, significant, and highly signifi-
cant associations between gene markers and the phenotype, respectively.
Significant linkage was defined in accordance with the guidelines of
Lander and Kruglyak (1995) as statistical evidence occurring by chance in
the genome scan with a probability of 5% or less. One-half of the fraction
of variance attributable to a QTL in the RI strains was used to estimate the
QTL effect expected in a comparable F
population derived from the
progenitors (to correct for the doubling effect of inbreeding on additive
genetic variance; Belknap et al. 1996; Buck et al. 1997). The fraction of
genetic variance contributed by the QTL was determined by dividing the
estimated QTL effect by the heritability. Strain distribution patterns of
more than 1000 markers in the BXD set of RI strains were obtained from
the Mouse Genome Database (see the reference).
When absolute testicular weights were compared with body
weights in all recombinant inbred (RI) and progenitor strains, there
was no significant correlation observed (r ס 0.25; p ס 0.24). This
result shows that testicular weights are largely independent of
body weights in 11-week-old mice of the BXD RI strains. There-
fore, the absolute testicular weights were used for genetic analysis
described in the current study. The C57BL/6J strain had signifi-
cantly lower testicular weight when compared with the DBA/2J
progenitor: 0.161 g ± 0.030 vs. 0.224 g ± 0.015, P < 0.0001. The
distribution of testicular weights among RI strains is shown in Fig.
1. As can be seen, the distribution of testicular weights is largely
continuous, suggesting the involvement of more than one gene.
The minimum number of effective factors responsible for this
variation was 5.7, and the narrow heritability was 0.53. Genome
scan with more than 1000 markers that were available for the BXD
RI strains, revealed the strongest association of testicular weights
with markers on Chr 13; suggestive associations were observed
also with markers on Chr 3 and 18 (Table 1). Figure 2 shows the
simple interval mapping of testicular weights with Chr 13 markers
depicting the likelihood ratio statistics. The narrow QTL peak
reflects the enhanced mapping resolution afforded by the fourfold
increase in recombination events provided by RI strains (Darvasi
1998). This QTL region, designated Twq1, accounts for 75% of the
additive genetic variance of the phenotype and suggests the exis-
tence of a major gene regulating testicular weight in the mouse.
Correspondence to: M. Pravenec
© Springer-Verlag New York Inc. 1998Mammalian Genome 9, 503–505 (1998).