Loci on chromosomes 2, 4, 9, and 16 for body weight, body length, and adiposity identified in a genome scan of an F2 intercross between the 129P3/J and C57BL/6ByJ mouse strains.

Abstract

Mice have proved to be a powerful model organism for understanding obesity in humans. Single gene mutants and genetically modified mice have been used to identify obesity genes, and the discovery of loci for polygenic forms of obesity in the mouse is an important next step. To pursue this goal, the inbred mouse strains 129P3/J (129) and C57BL/6ByJ (B6), which differ in body weight, body length, and adiposity, were used in an F2 cross to identify loci affecting these phenotypes. Linkages were determined in a two-phase process. In the first phase, 169 randomly selected F2 mice were genotyped for 134 markers that covered all autosomes and the X Chromosome (Chr). Significant linkages were found for body weight and body length on Chr 2. In addition, we detected several suggestive linkages on Chr 2 (adiposity), 9 (body weight, body length, and adiposity), and 16 (adiposity), as well as two suggestive sex-dependent linkages for body length on Chrs 4 and 9. In the second phase, 288 additional F2 mice were genotyped for markers near these regions of linkage. In the combined set of 457 F2 mice, six significant linkages were found: Chr 2 (Bwq5, body weight and Bdln3, body length), Chr 4 (Bdln6, body length, males only), Chr 9 (Bwq6, body weight and Adip5, adiposity), and Chr 16 (Adip9, adiposity), as well as several suggestive linkages (Adip2, adiposity on Chr 2, Bdln4 and Bdln5, body length on Chr 9). In addition, there was a suggestive linkage to body length in males on Chr 9 (Bdln4). For adiposity, there was evidence for epistatic interactions between loci on Chr 9 (Adip5) and 16 (Adip9). These results reinforce the concept that obesity is a complex trait. Genetic loci and their interactions, in conjunction with sex, age, and diet, determine body size and adiposity in mice.

Full text

Page 1

Loci on Chromosomes 2, 4, 9, and 16 for body weight, body length,
and adiposity identified in a genome scan of an F2 intercross
between the 129P3/J and C57BL/6ByJ mouse strains
Danielle R. Reed1, Xia Li1, Amanda H. McDaniel1, Ke Lu1, Shanru Li1, Michael G. Tordoff1,
R. Arlen Price2, and Alexander A. Bachmanov1
1 Monell Chemical Senses Center, 3500 Market Street, Philadelphia, Pennsylvania 19104, USA
2 University of Pennsylvania, Philadelphia, Pennsylvania 19104, USA
Abstract
Mice have proved to be a powerful model organism for understanding obesity in humans. Single
gene mutants and genetically modified mice have been used to identify obesity genes, and the
discovery of loci for polygenic forms of obesity in the mouse is an important next step. To pursue
this goal, the inbred mouse strains 129P3/J (129) and C57BL/6ByJ (B6), which differ in body weight,
body length, and adiposity, were used in an F2 cross to identify loci affecting these phenotypes.
Linkages were determined in a two-phase process. In the first phase, 169 randomly selected F2 mice
were genotyped for 134 markers that covered all autosomes and the X Chromosome (Chr). Significant
linkages were found for body weight and body length on Chr 2. In addition, we detected several
suggestive linkages on Chr 2 (adiposity), 9 (body weight, body length, and adiposity), and 16
(adiposity), as well as two suggestive sex-dependent linkages for body length on Chrs 4 and 9. In
the second phase, 288 additional F2 mice were genotyped for markers near these regions of linkage.
In the combined set of 457 F2 mice, six significant linkages were found: Chr 2 (Bwq5, body weight
and Bdln3, body length), Chr 4 (Bdln6, body length, males only), Chr 9 (Bwq6, body weight and
Adip5, adiposity), and Chr 16 (Adip9, adiposity), as well as several suggestive linkages (Adip2,
adiposity on Chr 2; Bdln4 and Bdln5, body length on Chr 9). In addition, there was a suggestive
linkage to body length in males on Chr 9 (Bdln4). For adiposity, there was evidence for epistatic
interactions between loci on Chr 9 (Adip5) and 16 (Adip9). These results reinforce the concept that
obesity is a complex trait. Genetic loci and their interactions, in conjunction with sex, age, and diet,
determine body size and adiposity in mice.
Introduction
The mouse is a well-established model organism to study human obesity genetics, both because
mice and humans develop spontaneous and diet-induced obesity, and because the sequencing
of the mouse genome is nearly complete (Mouse Genome Sequencing Consortium 2002).
Given the adequacy of the model and the available genetic resources, investigators have begun
to identify genomic regions containing loci that influence mouse body size and adiposity
(Brockmann et al. 1996, 1998, 2000; Cheverud et al. 1996; Collins et al. 1993; Corva et al.
2001; Dragani et al. 1995; Ishikawa et al. 2000; Keightley et al. 1996, 1998; Kirkpatrick et al.
1998; Klein et al. 1998; Kluge et al. 2000; Lembertas et al. 1997; Mehrabian et al. 1998; Moody
et al. 1999; Morris et al. 1999; Rance et al. 1997a, 1997b; Reifsnyder et al. 2000; Taylor and
Phillips 1996, 1997; Taylor et al. 2001; Warden et al. 1993, 1995; West et al. 1994a, 1994b,
1995; York et al. 1996, 1997). In our previous studies we found that males of the C57BL/6ByJ
Correspondence to: D.R. Reed; E-mail:reed@monell.org.
NIH Public Access
Author Manuscript
Mamm Genome. Author manuscript; available in PMC 2006 April 13.
Published in final edited form as:
Mamm Genome. 2003 May ; 14(5): 302–313.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Page 2

(B6) strain are heavier, longer, and fatter than those of the 129P3/J (129) strain (Bachmanov
et al. 2001), but this pair of strains has not been previously used to search for obesity-related
trait loci. The goal of the present experiment was to identify regions of the genome that harbor
loci for body weight, body length, and adiposity in mice of the F2 generation of hybrids between
the B6 and 129 parental strains. In Phase 1 of a genome scan, we genotyped 169 (B6 × 129)
F2 mice, using polymorphic markers selected at less than 20-cM intervals along the mouse
genome. In Phase 2, we genotyped 457 F2 mice for markers near linkages found in Phase 1.
Materials and methods
Mice
B6 and 129 inbred mice used for breeding and for phenotyping were obtained from The Jackson
Laboratory (Bar Harbor, Me.). The B6 × 129 F1 and F2 hybrids were bred at the Monell
Chemical Senses Center. The mice were housed in a temperature-controlled vivarium at 23°C
on a 12:12-h light:dark cycle and had free access to water and pelleted Teklad Rodent Diet
8604 (4.4% fat). All protocols were reviewed and approved by the Institutional Animal Care
and Use Committee of the Monell Chemical Senses Center.
F2 pups were weaned at 21–30 days of age and reared in same-sex groups (in most cases, 4–6
mice per cage, and never more than 6 mice in one cage). A total of 457 F2 mice (228 female
and 229 male mice) were bred from three types of reciprocal crosses: (B6♀ × 129♂) F1♀ ×
(B6♀ × 129♂) F1♂, (129♀ × B6♂) F1♀ × (129♀ × B6♂) F1♂, and (B6♀ × 129♂) F1♀ ×
(129♀ × B6♂) F1♂. The number of pups born to each dam was recorded, and the litter size
was used as a variable in some analyses.
Prior to the measurement of body length and adiposity, mice were tested to determine their
preference for taste solutions. Results of these experiments are reported in a separate
publication (Bachmanov et al. 2002). All mice (B6, 129, and F2) were treated identically during
these tests.
Body weight, body length, and adiposity phenotypes
Body weight, body length, and adiposity measures were collected when the average age of the
F2 mice was 8.9 ± 1.1 months (range 6.8–12.6 months). Variables measured were body weight
(to the nearest 0.1 g), body length (base of the lower incisors to anus, distance to the nearest
0.1 cm), and right and left retroperitoneal and gonadal adipose depot weights (four depots total,
weighed individually to the nearest 0.01 g). The weight of dissected adipose depots is highly
correlated with measures of whole-body carcass composition (Bachmanov et al. 2001). These
measures (body weight, body length, and adiposity) were the phenotypes used in the linkage
analyses.
Mean group differences for the traits were compared by ANOVA, with sex and genotype (B6,
129, and F2) as factors. Between-genotype differences were then evaluated with LSD post-hoc
tests.
Heritability in the broad sense (the degree of genetic determination) was estimated based on
variances in the parental strains and F2 (Bachmanov et al. 2002). The environmental
(nongenetic) variance was calculated as an average between the trait (total) variances for the
two parental strains, VARE=1
difference between the trait variance of the F2 generation and the environmental variance
(VARG = VARF2−VARE). The heritability estimate was calculated as a proportion of the
genetic variance from the trait variance of the F2, h2=VARG/VARF2 (Falconer 1989).
2(VARB6+ VAR129). The genetic variance was calculated as a
Reed et al.Page 2
Mamm Genome. Author manuscript; available in PMC 2006 April 13.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Page 3

We conducted univariate and multiple regression analyses to evaluate the relationship between
body weight, length, adiposity, sex, age, reciprocal cross type, and litter size of F2 mice. Some
covariates that explained a significant amount of trait variance in univariate analyses were used
in multivariate regression to remove covariate effects. The residuals obtained from the multiple
regression analyses were standardized to a mean of zero and a standard deviation of 1 so that
all trait values were comparable and are referred to as ’adjusted variables’ in the linkage
analysis. When mice of only one sex were examined, i.e., in the X-linked and sex-dependent
analyses, trait values were adjusted for covariates within each sex separately. Descriptive and
linear regression analyses were computed using Statistica (StatSoft, Tulsa, OK).
DNA extraction and genotyping
Genomic DNA was purified from mouse tails either by phenol/chloroform extraction and
precipitation with ethanol (Hogan et al. 1986) or by a sodium hydroxide method (Truett et al.
2000). We selected 134 markers polymorphic between the progenitor strains, spaced at less
than 20-cM intervals throughout the mouse genome (Table 1). In addition to the semidominant
PCR-based microsatellite markers, we used several dominant coat and eye color markers,
including agouti (A) on Chr 2 and tyrosinase (Tyr, formerly albino) and pink-eyed dilution
(P) on Chr 7. The B6 mice have black eyes and fur determined by genotypes a/a, Tyr/Tyr, P/
P. The 129 mice have pink eyes and albino fur (genotype Aw/Aw, Tyrc/Tyrc, p/p) or cream fur
(light chinchilla; genotype Aw/Aw, Tyrc-ch/Tyrc, p/p) (Festing et al. website; Roderick and Guidi
1989; Withham 1990). The F2 mice had several eye and coat color phenotypes. The F2 mice
with pink eyes were albino (Tyrc/Tyrc), cream (Tyrc-ch/Tyrc), or light buff (Tyrc-ch/Tyrc-ch);
agouti alleles could not be determined in these mice and were scored as unknown. The other
variants were white-bellied agouti (Aw/−, Tyr/−, P/−), black (a/a, Tyr/−, P/−), yellow coat with
pink eyes (Aw/−, Tyr/−, p/p), blue-gray coat with pink eyes (a/a, Tyr/−, p/p), chinchilla coat
with black eyes (Aw/−, Tyrc-ch/Tyrc-ch, P/−, and Aw/−, Tyrc-ch/Tyrc, P/−), and chocolate coat
with black eyes (a/a, Tyrc-ch/Tyrc-ch, P/−) (Silvers 1979). For each coat and eye color marker,
two genotypes could be distinguished: a homozygous genotype for a recessive allele (agouti
for the B6 strain, and albino and pink-eyed dilution for the 129 strain), and a heterozygous or
homozygous genotype for a dominant allele.
Microsatellite markers were amplified by the PCR with primers purchased from Invitrogen
(Carlsbad, Calif.) or Research Genetics (Huntsville, Ala.) with a protocol modified slightly
from that of Dietrich and coworkers (Dietrich et al. 1992). The denatured PCR products were
electrophoresed on a 6% polyacrylamide, 8.3 M urea sequencing gel, and the polymorphisms
were visualized by autoradiography. Some genotyping was conducted by the Australian
Genome Research Facility (Melbourne, Australia) with fluorescently labeled primers.
Linkage analysis
In Phase 1, we randomly selected a subset of 169 F2 mice (88 females and 81 males) for
genotyping from a pool of 457 mice. Mice were selected randomly rather than for extremes of
phenotype because we used multiple traits in the analyses and therefore had no single
distribution from which to draw the extreme phenotypes.
Interval mapping based upon maximum likelihood estimation was conducted with
MAPMAKER/QTL 1.1 software (Lander et al. 1987) that calculated chromosomal positions
and the percentage of variance accounted for by quantitative trait loci. Thresholds for
suggestive and significant linkage were estimated as described previously (Lander and
Kruglyak 1995). For markers from the X Chr and to detect sex-dependent autosomal linkages
(see below), F2 mice were grouped by sex, and the analysis was conducted separately for males
and females.
Reed et al.Page 3
Mamm Genome. Author manuscript; available in PMC 2006 April 13.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Page 4

Sex-dependent linkages for autosomal markers
The results of previous studies indicate that some autosomal obesity loci are sex dependent
(Ishikawa et al. 2000; Taylor et al. 1999). For the current analysis, if the LOD score was
suggestive or significant in one sex but at least 1 LOD score lower in the other sex, the locus
was considered sex dependent.
Phase 2 linkage analysis
Chromosomes that showed suggestive or significant evidence of linkage in Phase 1 (i.e., Chr
2, 4, 9, and 16; LOD > 2.8 for one or more phenotypes) were genotyped by using 288 additional
F2 mice. The analyses were conducted again using all 457 genotyped F2 mice. For each LOD
score peak, the mode of inheritance was determined by comparing LOD scores obtained under
unconstrained, additive, recessive, and dominant models. The percentage of trait variance
explained by several loci was determined with MAPMAKER/QTL by fitting all linked loci
simultaneously under the unconstrained model. With this value and heritability estimates
(described above), the percentage of genetic variance explained by several loci was calculated
as: “% of trait variance explained by loci”/“heritability” × 100.
A confidence interval for each locus was defined as LOD drops of 1.0 proximal and distal to
the LOD maximum. Where there was an indication of multiple linkages, we declared separate
peaks if the drops in LOD score on both sides were >2.0. When multiple peaks were detected,
composite interval mapping (Zeng 1993), computed with QTLCartographer (Basten et al.
2001), was employed to better resolve the location of the trait loci. Possible epistatic
interactions were evaluated by two-way ANOVA for the pairs of markers nearest to the loci
identified in the Phase 2 scan.
Results
Analysis of phenotypes
Genotype, (parental strain or F2), sex (male or female), and their interaction were evaluated.
Regardless of genotype, male mice were heavier, longer, and had more fat than female mice
[effect of sex on body weight, F(1,483) = 35.7, p < 0.01; body length, F(1,483) = 20.1, p <
0.01; adiposity, F(1,472) = 14.0, p < 0.01] (Table 2). Regardless of sex, B6 and F2 mice were
heavier (effect of genotype, F(2,483)=7.78, p < 0.001) and had more dissectible fat (effect of
genotype, F(2,483) = 3.72, p =0.025) than the 129 mice, but there was no main effect of
genotype on body length (p > 0.05) (Fig. 1). However, post hoc comparisons of body length
revealed that male B6 and F2 mice were longer than male 129 mice. The three groups of female
mice were similar in body length. Heritability estimates calculated by using the phenotype
information from the parental strains and the F2 mice were 0.80 (body length), 0.74 (body
weight), and 0.73 (adiposity).
Body weight, body length, and adiposity were affected by sex, age, and litter size in the F2
population (Table 3). Older mice were longer and had more adipose tissue than younger mice.
Mice from small litters were longer, heavier, and had more adipose tissue than mice from large
litters. Mice from different reciprocal cross groups did not differ in body weight, length, or
adiposity (p > 0.05).
Body weight, body length, and adiposity unadjusted for covariates were highly correlated
(Table 4). Using multiple regression, body weight was adjusted for sex and litter size; body
length was adjusted for sex, age, and litter size; adiposity was adjusted for sex, age, litter size,
body length, and body weight. After adjustment, body weight unadjusted and body length
variables were correlated with adiposity, whereas adjusted adiposity reflected the amount of
adipose tissue independent of body weight and body length. The distributions of all adjusted
Reed et al.Page 4
Mamm Genome. Author manuscript; available in PMC 2006 April 13.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Page 5

phenotypes were skewed (body weight, 0.90; body length, −0.22; adiposity, 1.58) and kurtotic
for body weight (2.16) and adiposity (9.29) and were not normally distributed (Shapiro-Wilk’s
W test; all p < 0.01; Fig. 2).
Phase 1 linkage results
Data for both sexes analyzed together revealed several significant and suggestive linkages for
body weight, body length, and adiposity (Figure 3; Table 5). There were two significant
linkages on Chr 2 for body weight and body length. Other linkages were suggestive and were
found on Chr 9 (body weight, body length, and adiposity) and on Chr 16 (adiposity). For all
loci except for the one on Chr 16, the mode of inheritance was additive. Chromosomes with
LOD scores >2.8 (unconstrained model) for one or more phenotypes were selected for
additional genotyping in Phase 2.
Linkage analysis within each sex (81 male and 88 female F2 mice) resulted in the identification
of two regions with suggestive LOD scores in one sex, but with low LOD scores in the other
sex. A region of Chr 4 was associated with body length in male but not female mice; likewise,
a region of Chr 9 was linked to body length in male but not female mice. These two
chromosomes were selected for additional genotyping in Phase 2.
Phase 2 analyses
There were five significant linkages found in the expanded sample of F2 mice (Table 6), with
the largest having a LOD score > 7.0 for body length near D2Mit168 (Fig. 4). The other
significant linkages were found for body weight on Chr 2 (Fig. 4), for body weight and adiposity
on Chr 9 (Fig. 5), and for adiposity on Chr 16 (Fig. 6). Three suggestive linkages were found:
an adiposity locus on Chr 2, and two LOD score peaks on Chr 9 for body length. These two
peaks on Chr 9 for body length met the criteria for separate peaks. (Note that, although there
are apparently two peaks for body weight on Chr 9, the centromeric peak does not meet the
criteria for suggestive linkage). For body length, composite interval mapping with a window
size of 10 cM confirmed two peaks on Chr 9 (LOD peak 1 = 2.22 at 56 cM, LOD peak 2 =
2.37 at 70 cM), although at a lower LOD score compared with the interval LOD score results.
Symbols were assigned to each locus following guidelines approved by the Mouse Genomic
Nomenclature Committee.
Individually, the loci identified here accounted for 3–9% of the total trait variance. When loci
were considered simultaneously, they accounted for 10% of the total phenotypic variance in
body weight, 12% of the variance in body length, and 11% of the adiposity variance. Using
heritability estimates for these traits, we calculated that these loci account for 13, 15, and 15%
of the total genetic variance for body weight, body length, and adiposity, respectively.
Sex-dependent autosomal linkages
When additional mice were genotyped by using markers for Chr 4, there was >1.0 LOD score
difference between male and female mice, and for the male mice, there was evidence for
significant linkage (Table 7). For Chr 9, the evidence for a male-dependent quantitative trait
remained suggestive. The male-only LOD score for Chr 9 is nearly as high as the LOD score
for both sexes analyzed together (see Fig. 5), which indicates that the suggestive LOD score
is explained by a male-specific effect.
Epistatic interactions
For all loci that met the criteria for suggestive or significant linkage in Phase 2, the nearest
markers were examined for evidence of epistasis. Two loci that contributed to adiposity, one
on Chr 9 near D9Mit25 and one on Chr 16 near D16Mit6, interacted to increase adiposity (Fig.
Reed et al.Page 5
Mamm Genome. Author manuscript; available in PMC 2006 April 13.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript