Copyright Ó 2006 by the Genetics Society of America
DOI: 10.1534/genetics.105.054049
Identification of Genetic Regions of Importance for Reproductive
Performance in Female Mice
Maria Liljander,*,1 Mary-Ann Sällström,* Sara Andersson,* Patrik Wernhoff,† Åsa Andersson,‡
Rikard Holmdahl† and Ragnar Mattsson*
*Lund Transgenic Core Facility and †Section for Medical Inflammation Research, Lund University, SE-221 84 Lund, Sweden and
‡
Department of Pharmacology, Danish University of Pharmaceutical Sciences, DK-2100 Copenhagen, Denmark
Manuscript received November 30, 2005
Accepted for publication March 10, 2006
ABSTRACT
Both environmental and genetic factors can dramatically affect reproductive performance in mice. In this
study we have focused on the identification of genetic regions, quantitative trait loci (QTL), which affect the
breeding capacity of female mice. We have identified polymorphic microsatellite markers for the mouse
strains used and performed a genomewide scan on 237 females from a gene-segregating backcross between
a high breeder and a relatively poor breeder. The high-breeder mouse strain we used is the inbred NFR/N
mouse (MHC haplotype H-2q), which has extraordinary good breeding properties. The moderate breeder
chosen for F1 and N2 progeny was B10.Q, which is a genetically well-characterized MHC-congenic mouse of
the H-2q haplotype. Each of the 237 females of the N2 generation was allowed to mate twice with MHCcongenic B10.RIII (H-2r) males and twice with B10.Q males. A predetermined number of phenotypes
related to reproductive performance were recorded, and these included litter size, neonatal growth, and
pregnancy rate. Loci controlling litter size were detected on chromosomes 1 (Fecq3) and 9 (Fecq4). The
neonatal growth phenotype was affected by Fecq3 and a locus on chromosome 9 (Neogq1). On chromosome
11 two loci affecting the pregnancy rate (Pregq1 and Pregq2) were identified. Furthermore, on chromosomes
13 and 17 we found loci (Pregq3 and Pregq4) influencing the outcome of allogeneic pregnancy (allogeneic by
means of MHC disparity between mother and fetuses). A locus on chromosome 1 affecting maternal body
weight was also identified and has been denoted Bwq7. It is well known that reproductive performance is
polygenically controlled, and the identification of the major loci in this complex process opens the
possibility of investigating the natural genetic control of reproduction.
R
EPRODUCTIVE success in mice, as in all mammals, is dependent on several environmental and
genetic factors. These factors are normally very complex and are often difficult to define. Most scientists
who have worked with mice in different animal houses
have experienced that the same inbred strains can show
different reproductive performance. This discrepancy
can be due to differences in more or less definable environmental factors, such as differences in cages, food,
bedding material, routines, health status, and handling
of the mice. In research there is often a considerable
problem when there is a low reproductive performance
of genetically modified model mice. However, some
strains of mice have, after many years of selection,
become more resistant to different types of stress and
produce large litters in most types of environments.
Such strains are often denoted ‘‘high breeders,’’ and
they carry genes ensuring large litter size, high stress
resistance during pregnancy, and good nursing properties during the lactation period. The NFR/N strain is
1
Corresponding author: Lund Transgenic Core Facility, BMC C13, Lund
University, SE-221 84 Lund, Sweden. E-mail: [email protected]
Genetics 173: 901–909 ( June 2006)
an example of an inbred high-breeder strain, while common C57BL strains normally produce lower numbers
of litters and are often denoted ‘‘moderate breeders.’’
Females of the NFR/N strain are known to produce
several large litters during a long period of time. Furthermore, they are known to be excellent mothers (good
nursing properties, high milk production, rapid neonatal growth, etc.); it is likely that these females carry
several genes that are valuable for successful reproduction in mice.
A limited number of reported studies have applied
genetic mapping and linkage analyses in their search for
genetic regions of significant importance for successful
reproduction in mice. In the early 1980s a classical genetic analysis aimed at finding genetic regions critical
for litter size was performed, but no clear-cut results
were obtained (Horstgen-Schwark et al. 1984). The
discovery of new techniques for genotyping during recent years, together with the continued development
of more advanced linkage analysis programs, has made
it possible to perform more exact genetic analyses and
eventually also to identify single genes. Kirkpatrick
et al. (1998) applied modern methods to map gene regions critical for litter size in an F2 progeny between
902
M. Liljander et al.
the outbred Quackenbush-Swiss mouse line and ordinary C57BL/6 mice. They found significant linkages at
specific segments of chromosomes 2 and 4 (Fecq1 and
Fecq2) and a suggestive linkage at a region of chromosome 9. Furthermore, Peripato et al. (2002, 2004) presented QTL data for maternal reproductive performance
in an F2 progeny between LG/J and SM/J mice, and
they particularly point out genetic regions on chromosomes 7 and 12 as critical for the litter size phenotype.
This group also showed that several reproductive phenotypes are very complex and subjected to epistatic
interactions from genetic regions of several chromosomes. Rocha et al. (2004) also performed a QTL study
of pregnancy-associated traits in mice, and they report
that loci on chromosome 2 are of particular importance.
Furthermore, Everett et al. (2004) reported that loci
on chromosomes 1 and 9 control the ovulation of
primary oocytes in mice, a phenotype related to litter
size. In this context it should also be mentioned that
Spearow et al. (1999a,b; Spearow and Barkley 1999)
have mapped genes critical for differences in hormoneinduced ovulation rate (ORI genes) between A/J and
C57BL/6 mice, and QTL for this trait were identified on
chromosomes 2, 6, 9, 10, and X.
Still, there is a need for identification of additional
loci that are critical for successful breeding of mice. This
study focuses not only on female-dependent differences
in litter size in normal mating, but also on a number of
other female-associated traits critical for reproduction
in mice, such as vaginal plug frequency during 96-hr
mating periods, ratio of number of pregnancies to the
number of vaginal plugs, growth of pups, neonatal mortality, and amount of maternal IgG transmitted to the
offspring, etc.
A unique approach used in our study is that we have
included an analysis of the possible influence of MHC
differences between mother and fetuses. Mammalian
allogeneic pregnancies deal with the classical immunological problem, i.e., that the mother should avoid
immunological rejection of her genetically different offspring and at the same time she should mount an optimal defense against pathogens to provide the fetuses
with passive immunity. Since the days of Medaware
(1953) the immunological enigma of mammalian pregnancy has been highlighted several times, and a number
of possible protective mechanisms have been presented,
such as low placental MHC expression (Sunderland
et al. 1981; Mattsson et al. 1992), protective properties of
the placenta (Petroff et al. 2003; Aluvihare et al. 2005),
placental complement inhibitory factors (Thurman et al.
2005), cytokine balance (Raghupathy 1997; Svensson
et al. 2001), placental tryptophan catabolism (Munn
et al. 1998; Mellor and Munn 2000), etc. Although
many of the suggested mechanisms of protection might
be of importance during different situations it is still
uncertain which of these might be most critical, or if
products of genes that have yet to be identified might
fulfill a more significant protection. We have in this
study chosen to study mouse strains to evaluate differences in pregnancy success depending on differences
in MHC between mother and fetuses (allogeneic vs.
syngeneic pregnancies by means of MHC). The two
parental strains chosen (NFR/N and B10.Q) carry
MHC alleles of the H-2q haplotype. The females of the
genotyped N2 generation (backcross to B10.Q) have
been allowed to mate both with B10.Q males (syngeneic
pregnancy) and with B10.RIII males (allogeneic pregnancy). In this study we provide genetic marker information for the NFR/N strain and define genetic regions
containing alleles that influence the reproductive performance in female mice.
MATERIALS AND METHODS
Animals: NFR/N mice were originally obtained from the
National Institutes of Health (Bethesda, MD) and the B10.Q
mice were originally bought from The Jackson Laboratory
(Bar Harbor, ME). (B10.Q 3 NFR/N)F1 hybrids and (B10.Q 3
NFR/N) 3 B10.Q N2 mice were bred in IVC cages in the BMC
barrier animal house and at the biomedical center and the
pathology animal house at Lund University, Sweden. All mice
were fed ad libitum with standard rodent pellets (LAB FOR
R36, irradiated breeding food for rats and mice; Lactamin
AB, Sweden) and water in a climate-controlled environment
with a light:dark photoperiod of 12:12. The mice used in this
study had clean health monitoring protocols according to the
Federation of European Laboratory Animal Sciences Association recommendations. Ethical permissions are M125-04
(embryo transfer) and M290-03 (reproduction and arthritis).
Phenotype measurement and experimental design: A total
of 237 female mice of the (NFR/N 3 B10.Q) 3 B10.Q N2
backcross were used in this study. The females were kept with
males on four occasions, each mating period lasting for 96 hr
to approximate one estrus cycle. First, the females were allowed
to mate twice with B10.RIII males (allogeneic mating by means
of MHC) and finally twice with B10.Q males (syngeneic mating
by means of MHC). Phenotype data such as litter size and pups’
and mothers’ weights, etc., were collected during and after each
pregnancy. The reproductive performances for the parental
strains (NFR/N, B10.Q) and the F1 generation (NFR/N 3
B10.Q) were tested in the same way.
Description of phenotypes recorded:
1. Litter size: The number of pups born on the expected day
of delivery.
2. Frequency of pregnancies: Number of pregnancies
started after four possible mating occasions each being
96 hr long (96 hr is the expected period of the estrus
cycle).
3. Frequency of plugs: Number of vaginal plugs obtained
after four possible mating periods each of 96 hr. Plugs
were observed every day for each animal during this 4-day
period.
4. Pregnancy rate: The ratio between detected number of
pregnancies and detected number of vaginal plugs.
5. Birth weight per pup: Mean weight (grams) per pup on
the day of delivery.
6. Neonatal weight per pup, day 7: Mean weight (grams) per
pup on day 7 postpartum.
7. Neonatal growth per pup, days 1–7: Mean weight per pup
on day 7 minus mean weight per pup on day 1 (day 1 is day
of delivery).
Genetic Control of Mouse Reproduction
903
TABLE 1
Median litter size, mean pup weight (grams) days 1 and 7 in the parental strains B10.Q and NFR/N (F0), F1, and the N2 generation
(NFR/N 3 B10.Q) 3 B10.Q
Males used
in mating
No. of
females
Litter size:
Median (range)a
Pup weight day 1:
Mean 6 SD
Pup weight day 7:
Mean 6 SD
F0 B10.Q
F0 NFR/N
B10.Q
NFR/N
10
10
F0 NFR/N
F0 NFR/N
F1 NFR/N 3 B10.Q
F1 NFR/N 3 B10.Q
N2 (NFR/N 3 B10.Q) 3 B10.Q
N2 (NFR/N 3 B10.Q) 3 B10.Q
B10.Q
B10.RIII
B10.Q
B10.RIII
B10.Q
B10.RIII
10
10
10
10
237
237
7.0 (4, 8.5)
9.0 (7.5, 11)
P ¼ 0.0008b
9.0 (6.5, 12)
9.5 (7, 13)
11.5 (9, 13)
11.5 (10, 13.5)
7.8 (1, 13)
8.1 (0, 14)
1.46 6 0.15
1.63 6 0.93
P ¼ 0.007b
1.9 6 0.08
1.9 6 0.21
1.68 6 0.10
1.67 6 0.10
1.6 6 0.16
1.5 6 0.14
4.2 6 0.94
5.2 6 0.38
P ¼ 0.05b
5.9 6 0.51
5.3 6 0.21
4.87 6 0.38
4.9 6 0.36
4.6 6 0.74
4.7 6 0.68
Females
a
b
Range, maximum and minimum values in the series.
F0 (B10.Q 3 B10.Q) compared with (NFR/N 3 NFR/N). P-values were calculated with Student’s unpaired t-test.
8. Maternal weight, day 7: Maternal weight (grams) day 7
postpartum.
9. Maternal growth, days 1–7 postpartum: Maternal weight
(grams) after delivery day 1 minus maternal weight day 7
postpartum.
10. Maternal body weight: Maternal weight (nonpregnant) at
the age of 15 months.
11. IgG concentration per pup, day 7: IgG concentration in
neonatal blood.
12. Total IgG transmission, day 7: Total IgG concentration in
neonatal blood multiplied by 10% of the total weight of
the pups in the litter, day 7. Approximately 10% of the
body weight consists of blood. Approximately 80% of
the IgG detected in pups on day 7 can be expected to
originate from maternally transmitted milk (Gustafsson
et al. 1994).
13. Ovary weight: Mean wet weight of maternal ovaries at 15
months of age.
14. Uterus weight: Wet weight of maternal uterus (two horns
plus cervix) at 15 months of age.
Microsatellite genotyping and linkage analysis: Tail biopsies
were collected from all N2 females and the parental strains.
DNA was isolated according to a previously described protocol (Laird et al. 1991). Since NFR/N has not been subjected
previously to this type of genetic analysis, the first step in this
study was to identify NFR/N-specific microsatellite marker alleles (Table 2). After screening of parental DNA with 450 mouse
fluorescence-labeled microsatellite markers (INTERACTIVA,
Ulm, Germany), 115 (25%) informative polymorphic markers
were selected, covering the genome. Two hundred thirty-seven
female N2 mice were genotyped, covering all chromosomes
except for the Y chromosome. PCR amplification for the markers was performed in a final volume of 10 ml in a 96-well
V-bottom microtiter plate using 20 ng of DNA, 10 mm KCl,
20 mm Tris–HCl, 10 mm (NH4)2SO4, 2 mm MgCl2, 0.1% Triton
X-100, pH 8.8 (New England Biolabs, Beverly, MA), 3 mm of
each primer, 2 mm dNTPs (Advanced Biotechnologies, Surrey,
UK), and 0.25 units Taq DNA Polymerase (New England Biolabs).
The following program was used to amplify the DNA: denaturation at 95° for 3 min; annealing at 56° for 45 sec; polymerization at 72° for 1 min; 30 cycles of 95° for 30 sec, 56° for
45 sec, and 72° for 1 min; and a final extension step of 7 min at
72°. The PCR products were analyzed on a MegaBACE 1000
(Amersham Pharmacia Biotech, Piscataway, NJ) according to
the manufacturer’s protocol. Data were analyzed with Genetic
Profiler 1.1 through comparison from parental mouse strains.
The linkage analyses were performed using R and R/qtl
(Ihaka and Gentleman 1996; Broman et al. 2003). Ninety
percent of the mouse genome was within a 20-cM intermarker
distance. The marker map was generated using the Kosambi
map function and 1000 permutations were performed for
every phenotype (P , 0.05). The two-locus interaction was performed using the imputation model in R/qtl. Here the calculated LODjoint score values compare a full model, if
including covariates (y ¼ m 1 bq1 1 bq2 1 bq1 3 q2 1 Ag 1
Zdq1 1 Zdq2 1 Zdq1 3 q2 1 e), to a null model (y ¼ m 1 Ag 1 e).
The epistasis, LODint score, compares the full model to an
additive model (y ¼ m 1 bq1 1 bq2 1 Ag 1 Zdq1 1 Zdq2 1 e).
q1 and q2 are unknown QTL genotypes at two different locations, a matrix of covariates and a Z-matrix of QTL interacting
covariates. Permutation tests (n ¼ 1000) were done subsequently to establish empirical significance thresholds for the
interactions. A threshold equal to or above the 95th percentile
was considered significant. Figures, illustrating the interactions, were created using the image function in scanqtl. The new
loci symbols reflect the trait measured and are named according to the rules for nomenclature for mouse and rat strains in
the mouse genome informatics (MGI) (http://www.informatics.
jax.org/mgihome//nomen/index.shtml).
Enzyme-linked immunosorbent assay: The pups were killed
at the age of 7 days and sera from each litter were pooled. A
sandwich ELISA was performed to determine the amount of
IgG that the mother has transferred to the pups (Engwall
1980) Briefly, goat anti-mouse IgG (3 mg/ml) was coupled to
immunosorbent plates overnight at 4°. After blocking with
bovine serum albumin (Sigma, St. Louis) different dilutions of
purified mouse IgG (Sigma), test sera, and positive and negative controls were added. The murine IgG levels were detected
with peroxidase-conjugated goat anti-mouse IgG (Sigma). The
ELISA plates were analyzed in a spectrophotometer at 405 nm.
The adult females were killed at 15 months of age and sera
analyzed as above.
Statistical analysis: Statistical analyses were performed by
using the two-tailed Mann–Whitney U-test (for nonparametric
data) or Student’s t-test.
RESULTS
Breeding properties in parental strains NFR/N and
B10.Q: The NFR/N mouse is significantly larger in size
than the B10.Q mouse (NFR/N females ¼ 29.9 6 0.8 g,
904
M. Liljander et al.
TABLE 2
A complete list of the 115 polymorphic microsatellite markers used in the genome widescreening
Marker
cM/Mbpa
B10.Q/NFR/Nb
Marker
cM/Mbpa
B10.Q/NFR/Nb
Marker
cM/Mbpa
B10.Q/NFR/Nb
D1Mit64
D1Mit67
D1Mit373
D1Mit161
D1Mit216
D1Mit10
D1Mit45
D1Mit187
D1Mit89
D1Mit14
D1Mit146
5.0/12.9
9.0/13.4
17.0/26.7
27.0/59.6
49.7/80.3
56.6/90.6
58.5/92.9
62.0/113.6
63.1/124.4
81.6/156.8
92.3/171.1
131/129
142/134
123/136
114/116
123/127
138/132
179/170
162/149
146/136
178/195
124/113
D8Mit4
D8Mit339
D8Mit80
D8Mit186
D8Mit56
14.0/30.9
23.0/39.4
41.0/87.4
59.0/115.9
73.0/128.4
161/200
154/144
108/122
126/133
128/161
D13Mit3
D13Mit63
D13Mit24
D13Mit51
D13Mit78
10.0/—
26.0/42.0
43.0/68.7
53.0/101.2
75.0/115.2
99/189
137/145
206/196
142/135
227/207
100/108
84/114
165/138
203/237
104/134
172/176
194/180
119/148
148.6/161
180/190
75/104
120/109
154/144
184/214
151/136
122/141
119/127
108/97
17.0/27.8
37.0/66.7
52.2/114.1
69.0/—
87.0156.2
17.0/37.4
27.0/44.0
31.0/57.8
34.0/58.1
42.0/76.8
48.0/86.7
52.0/98.6
62.0/115.0
71.0/—
D14Mit11
0.7/8.7
D14Mit62 18.5/44.0
D14Mit193 40.0/63.3
D14Mit131 58.0/112.3
D2Mit365
D2Mit91
D2Mit102
D2Mit17
D2Mit48
D9Mit2
D9Mit130
D9Mit21
D9Mit48
D9Mit8
D9Mit11
D9Mit35
D9Mit17
D9Mit18
D15Mit102
D15Mit266
D15Mit167
D15Mit214
D15Mit79
155/198
122/142
127/119
123/131
110/104
D3Mit46
D3Mit40
D3Mit213
D3Mit255
D3Mit19
13.8/—
39.7/87.8
49.7/98.2
66.2/137.1
87.6/—
164/157
106/136
149/141
151/166
155/165
D10Mit80
D10Mit126
D10Mit198
D10Mit70
D10Mit145
4.0/11.4
21.0/26.8
40.0/—
59.0/103.8
70.0/125.1
153/149
128/123
127/123
143/147
135/140
D4Mit193 7.5/32.4
D4Mit164 28.6/58.7
D4Mit331 50.8/102.2
D4Mit233 75.5/143.6
137/141
139/147
128/135
155/165
D5Mit193
D5Mit387
D5Mit267
D5Mit41
D5Mit168
1.0/4.2
15.0/26.7
24.0/—
56.0/101.7
78.0/134.5
132/147
179/183
152/162
146/158
154/113
D6Mit236 3.1/—
D6Mit183 26.5/53.2
D6Mit65 46.0/101.8
D6Mit25 65.0/—
139/150
107/100
100/73
131/121
D11Mit227
D11Mit310
D11Mit23
D11Mit350
D11Mit29
D11Mit36
D11Mit196
D11Mit70
D11Mit99
D11Mit360
D11Mit258
D11Mit333
D11Mit181
D11Mit214
D11Mit48
2.0/—
24.0/53.8
28.1/53.9
34.45/62.9
40.0/69.4
47.64/—
50.0/87.7
54.0/93.8
59.5/99.3
64.0/103.0
65.0/107.4
66.0/108.4
69.0/109.0
70.0/114.8
77.0/117.8
166/170
130/136
131/120
98.7/93.6
142/148
236/220
137/142
142/136
124/107
122/100
130/167
123/105
123/113
148/150.6
133/127
D7Mit76
D7Mit83
D7Mit30
D7Mit62
D7Mit126
D7Mit371
225/199
151/147
232/243
146/134
170/177
120/123
3.4/10.5
26.5/39.2
37.0/68.6
42.6/71.8
50.0/—
65.2/119.9
D12Mit170 6.0/16.5
D12Mit36 28.0/55.8
D12Mit239 44.0/83.1
D12Mit8
58.0/108.1
115/145
122/136
92/74
167/155
6.7/65.8
11.4/17.1
28.2/60.2
47.9/80.8
66.2/104.0
D16Mit181 4.3/6.0
D16Mit57 21.0/—
D16Mit64 38.0/57.6
D16Mit86 66.0/93.6
110/123
110/91
133/127
126/119
D17Mit59
D17Mit81
D17Mit233
D17Mit10
D17Mit20
D17Mit217
D17Mit93
D17Mit123
9.91/17.2
16.4/29.5
20.0/34.5
24.5/—
34.3/55.6
38.5/64.2
44.5/72.2
56.7/91.9
142/123
123/134
99.5/113
152/139
176/180.7
169/185
170/156
134/147
D18Mit68
D18Mit36
D18Mit50
D18Mit4
11.0/21.8
24.0/47.0
41.0/68.4
57.0/84.6
115/94
150/135
158/162
210/206
D19Mit31
7.0/13.0
D19Mit100 27.0/41.7
D19Mit10 47.0/46.5
D19Mit6
55.0/60.4
127/129
156/113
150/141
109/113
DXMit105
DXMit119
DXMit173
DXMit80
151/146
147/153
125/135
135/140
14.5/42.3
29.5/62.0
50.5/120.4
65.4/146.2
a
The marker positions in centimorgans (cM) are according to MGI (http://www.informatics.jax.org). Positions in megabase
pairs (Mbp) are according to Mouse Ensemble (http://www.ensembl.org/Mus_musculus).
b
The length of the parental fragments in base pairs (bp).
and B10.Q females ¼ 26.0 6 0.7 g at 4 months of age)
and is known for its extraordinary good breeding properties. In Table 1 we have summarized differences in
litter size between these two parental strains. The F0
generations (B10.Q 3 B10.Q and NFR/N 3 NFR/N)
differ significantly in litter size and neonatal growth.
Characteristically, the F1 generation also shows very
good reproductive performance, which is significantly
higher than that in the B10.Q strain. The MHC disparity between mother and fetuses (mating with two
different MHC-congenic paternal C57/BL strains:
B10.RIII and B10.Q) did not significantly affect the
litter size in the parental NFR/N or NFR/N 3 B10.Q F1
females.
Genetic Control of Mouse Reproduction
905
TABLE 3
Markers associated with reproductive phenotypes in an N2 backcross involving (NFR/N 3 B10.Q) 3 B10.Q females mated
with B10.Q and B10.RIII males
Phenotype
Symbol Chr
Litter size (median value from
four pregnancies)
Fecq3
Flanking marker
1 D1Mit64–D1Mit373
Fecq4
9 D9Mit130–D9Mit48
Pregq1 11 D11Mit23–D11Mit36
Pregq2
D11Mit70–D11Mit48
Neonatal growth (g)/pup days 1–7
Fecq3
1 D1Mit64–D1Mit373
Fecq4
9 D9Mit130–D9M48
Neogq1
9 D9Mit48–D9Mit11
Maternal weight (g) at age 15 mo
Bwq7
1 D1Mit45–D1Mit89
Maternally transmitted IgG (mg/ml) day 7
1 D1Mit187–D1Mit14
13 D13Mit24–D13Mit78
Pregnancy rate
Linking marker cM/Mbpa
D1Mit67
D9Mit21
D11Mit29
D11Mit360
D1Mit67
D9Mit21
D9Mit8
D1Mit187
D1Mit89
D13Mit51
9.0/13.4
ABb
AAc
LOD
7.9
9.0
2.8**
31.0/57.8
8.0
6.8
40.0/69.4
0.89 1.0
64.0/103.0 0.89 1.0
9.0/13.4
3.38 3.95
31.0/57.8
3.30 3.04
42.0/76.8
3.41 2.95
62.0/113.6 36.6 30.3
63.1/124.4 0.41 0.60
53.0/101.2 0.64 0.43
2.5**
2.9**
2.7**
3.8**
2.9*
3.0**
4.5***
1.7*
1.7*
* Suggestive significance in a genomewide scan; **significant in a genomewide scan; ***highly significant in a genomewide
scan.
a
The position in centimorgans for the linking marker is according to the MGI database (http://www.informatics.jax.org) and
the markers’ physical positions are according to Mouse Ensemble build 33 (http://www.ensembl.org/Mus_Musculus).
b
Mean phenotypic values in the group of mice heterozygous at the linking marker and median values for litter size.
c
Mean phenotypic values in the group of mice homozygous for B10.Q at the linking marker and median values for litter size.
Plugs were recorded daily for each animal and 95% of
the plugged females became pregnant during the first
mating opportunity. During mating periods two, three,
and four, 90% of all plugged females became pregnant. None of the females became pregnant without being
plugged, confirming the accuracy of plug detection.
QTL analyses of phenotypes for reproductive performance: Table 2 shows the map positions of markers
typed across the data set. Significant QTL were found
for several phenotypes (Table 3, Figure 1). On proximal
chromosome 1 we found a region affecting litter size
and neonatal growth, denoted Fecq3. In addition, on
chromosome 9 we found another region affecting litter
size (Fecq4) and neonatal growth (Neogq1). Furthermore,
two loci on chromosome 11 (Pregq1 and Pregq2) were
found to strongly affect pregnancy rate, i.e., the frequency
of successful pregnancies after plugging (Figure 2).
Another phenotype of possible interest in reproductive contexts is the size of the mouse, since there is normally a correlation between number of litters produced
and the size of the mouse (the NFR/N mice are larger
than B10.Q and have the potential capacity to carry
more pups). A highly significant QTL (now denoted
Bwq7) for this phenotype was identified around 62 cM
on chromosome 1 (Figure 1).
Epistatic interactions: A significant two-locus interaction was found for the phenotype pregnancy rate. Here
the locus Pregq1, identified from the single-locus analysis, on chromosome 11 was significantly affected, LOD
5.6, by a locus on chromosome 10 at marker D10Mit198
Figure 1.—Chromosomal
locations of QTL for litter
size (Fecq3 and Fecq4), body
weight (Bwq7), neonatal
growth (Neogq1), and pregnancy rate (Pregq3 and
Pregq4). Solid lines within
bars denote the location of
the maximum LOD score
value.
906
M. Liljander et al.
reduced pregnancy rate (0.82) for individuals expressing heterozygosity (AB/AB) at both loci compared to
that of AA/AB individuals (0.97).
QTL specific for allogeneic pregnancy (MHC disparity between mother and fetuses): The phenotypes
listed above appeared to be unaffected by the type of
mating (syngeneic or allogeneic mating by means of
MHC disparity between mother and father). However, a
limited number of QTL were recorded exclusively in
allogeneic matings. As seen in Table 4, these QTL refer
to the phenotype ‘‘pregnancy rate’’ that seems to be
controlled by some gene(s) around 26 cM on chromosome 13 (Pregq3). Furthermore, in a region on chromosome 17 (Pregq4), the chromosome harboring MHC
in mice appears to affect the pregnancy rate; i.e., the
probability that a mating results in a successful pregnancy increases if there is a MHC disparity between
mother and father (Figure 2B). The phenotype ‘‘frequency of plugs’’ was also significantly affected by a
region on chromosome 17. Our results indicate that a
gene(s) outside the MHC is (are) involved in controlling this effect.
DISCUSSION
Figure 2.—Logarithm of odds (LOD) plots for chromosomes with identified loci. Horizontal lines represent experimentwide significance level (P ¼ 0.05) according to
permutation test (n ¼ 1000). (A) Identification of two novel
loci on chromosome 11 for pregnancy rate (Pregq1 and Pregq2).
(B) Identification of a novel locus (Pregq4) on chromosome 17
for pregnancy rate (allogeneic pregnancies only). Analyses
were performed using the imputation model in R/qtl.
(Figure 3). The interaction was strongest under a complete model (q1 3 q2), allowing both strict epistasis and
additive interaction. The genetic interaction between
markers D10Mit198 and D11Mit29 (Figure 4) showed a
A limited number of linkage analyses based on
QTL mapping have been performed previously for
identification of gene regions controlling pregnancy
success in female mice (Kirkpatrick et al. 1998;
Peripato et al. 2002, 2004) To our knowledge this is
the first study of this type that also focuses on the
identification of genetic regions that specifically affect
the outcome of allogeneic pregnancy (by means of MHC
disparity), a question that has been discussed by reproductive and transplantation immunologists for decades (Medaware 1953; Billington 1993; Gustafsson
et al. 1994; Bulla et al. 2004).
The fact that NFR/N and B10.Q mice carry the same
MHC haplotype (H-2q) also made it possible to study
the possible influence of fetal/maternal MHC disparity
Figure 3.—Pregplug (interacting model q1 3
q2). Interaction graph for pregnancy rate showing
interacting loci for chromosomes 10 and 11. Significant interaction, LOD ¼ 5.6, was found for locus Pregq1, marker D11Mit29 on chromosome 11,
and marker D10Mit198 on chromosome 10. Significance levels according to genomewide permutation tests, n ¼ 1000, were P(0.95) ¼ 5.3. The
values written in the graph are LOD values from
the interaction analysis in R/qtl using the scanqtl
function and the imputation model. Distances between markers are according to recombination
frequency.
Genetic Control of Mouse Reproduction
Figure 4.—Interaction plot for markers D11Mit29 and
D10Mit198. Genetic interaction between marker D11Mit29
on chromosome 11 and marker D10Mit198 on chromosome
10 for pregnancy rate is shown. A represents B10.Q and B represents NFR/N alleles. The phenotypic variation is shown and
expressed as error bars representing 1 SD.
by allowing the females to mate with either B10.Q or
B10.RIII males (‘‘syngeneic’’ and ‘‘allogeneic’’ pregnancies, respectively, by means of MHC). It is known that
MHC class I molecules are expressed on the trophoblast
cells of the spongio-trophoblast region of the placenta
(Mattsson et al. 1992) although previously it has been
difficult to recognize immune reactivity against paternal MHC during normal allogeneic pregnancy in most
strains of mice (Billington 1993). However, other studies claim that such immune reactions are distinguishable (Munn et al. 1998; Mellor and Munn 2000) and
that suppression of Tcell responses against fetal/placental
MHC is a significant mechanism of protection against
host vs. graft reactivity in allogeneic pregnancies in mice.
In our analyses we have distinguished between phenotypes that are unaffected or affected by the type of
907
mating. The sequential mating, first with B10.RIII males
and later with B10.Q males, may influence the outcome
of the frequency of pregnancies, but despite the order
of the mating the results clearly indicate that fetal/
maternal MHC disparity influences other reproductive traits such as pregnancy rate and neonatal growth
(Table 4). In Table 3 we present QTL that are of general
validity for pregnancy (data from all pregnancies, independent of type of mating), while the QTL shown in
Table 4 refer to observed differences between syngeneic
and allogeneic pregnancy by means of maternal/fetal
MHC disparity.
It is obvious from the results in Table 3 that many
genetic regions are involved in the regulation of reproductive performance; furthermore, the majority of these
loci have not been reported previously. We have found
a locus (Fecq3) of significant negative influence of the
NFR/N alleles at 9 cM on chromosome 1 (Figure 1).
Several other QTL of suggestive significance for many
different reproductive phenotypes were present at different regions of this chromosome. However, the genetic analyses indicated a negative influence of NFR/N
alleles on reproductive parameters. It should be noted
that the maternal body weight phenotype, which was
mapped at 62 cM on chromosome 1 (now denoted
Bwq7), shows a positive dominant effect from the NFR/N
fragment (Figure 1). Two loci controlling body weight
have previously been detected in the vicinity of this
region, i.e., Bw17 (Anunciado et al. 2000) and Bwtq1
(Morris et al. 1999). As to what extent Bwq7, which is a
highly significant QTL, affects litter size is still unclear
and no other QTL for litter size have been found in this
region of chromosome 1. However, earlier studies indicate some overlap between QTL affecting both litter size
and mature weight on chromosome 4 (Kirkpatrick
et al. 1998), which indicates the possibility of some correlation between weight and litter sizes.
TABLE 4
Markers associated with reproductive phenotypes in an N2 backcross involving (NFR/N 3 B10.Q) 3 B10.Q females mated
with B10.RIII (allogeneic mating) and B10.Q males (syngeneic mating)
Phenotype
Symbol
Pregnancy rate
Pregq3
Pregq4
Neonatal growth/
pup days 1–7 (g)
Flanking marker
Linking
marker
Allogeneic
pregnancy
a
cM/Mbp
b
c
Syngeneic
pregnancy
b
c
LOD score
AB
AA
AB
AA
D13Mit3–D13Mit24
D13Mit63 26.0/42.0
D17Mit20–D17Mit123 D17Mit93 44.5/72.2
D6Mit236–D6Mit65
D6Mit183 26.5/53.2
0.64
1.0
3.19
0.74
0.85
3.69
0.70
0.97
2.95
0.66
0.95
3.10
3.1**
4.8***
3.0*
0.2
0.3
0.4
D9Mit48–D9Mit11
2.92
3.42
2.90
3.13
3.1**
2.6
D9Mit8
42.0/76.8
Allo preg. Syn preg.
* Suggestive significance in a genomewide scan; **significant in a genomewide scan; ***highly significant in a genomewide
scan.
a
The marker position in centimorgans for the linking marker is according to MGI (http://www.informatics.jax.org) and the
markers’ physical positions are according to Mouse Ensemble build 33 (http://www.ensembl.org/Mus_Musculus).
b
Mean phenotypic values in the group of mice heterozygous at the linking marker.
c
Mean phenotypic values in the group of mice homozygous for B10.Q at the linking marker.
908
M. Liljander et al.
A genetic region of high interest in the context of
litter size is chromosome 9 at 42 cM (Fecq4). The presence of NFR/N alleles in this region has a positive
influence on litter size. What makes this region even
more interesting is that another trait (neonatal growth
days 1–7), which is phenotypes closely related to litter
size, shows significant linkage to the very same region. It
should also be noted that Kirkpatrick et al. (1998)
actually have identified a QTL of suggestive significance
for the litter size trait in the same region of chromosome
9. Although speculative, it is worth mentioning that a
highly interesting candidate gene, the fork-head box
gene 1b (Fox1b), is present in this region (41 cM). Fox1b
is a part of a gene family and codes for a transcription
factor, which has been reported to influence reproductive parameters, such as embryonic lethality, postnatal
mortality, and lactation (Labosky et al. 1997).
If male mice are fertile, the vaginal plug formed in the
female mouse after copulation is a fairly certain indication of copulation and initiation of pregnancy.
However, the plug is never a 100% certain indication
of pregnancy onset and genetic differences exist between different strains of mice. Failure to enter pregnancy after plugging can be due to fertilization failure,
implantation failure, or early embryonic lethality. The
phenotype pregnancy rate is the ratio between the number of pregnancies and the number of recorded plugs
for each mouse, which in turn reflects the frequency of
early pregnancy failure. Two significant QTL for this
phenotype were found on chromosome 11 (Figure 2A),
one peak at 40 cM (Pregq1) and one peak at 64 cM
(Pregq2). The Ace gene coding for angiotensin-converting
enzyme, located at 65 cM on this chromosome, is claimed
to affect male fertility (Kondoh et al. 2005). Furthermore, Kirkpatrick et al. (1998) reported a suggestive
significant locus for the trait litter size at around marker
D11Mit231 1 11 cM. To our knowledge, no other locus
affecting female fertility has previously been identified
on this part of chromosome 11.
Pregq1 also shows an interaction effect (epistatic interaction) with another locus on chromosome 10 (D10Mit198).
In our study we used the N2 backcross of highly inbred
mice (B10.Q and NFR/N), which gives less genetic
variation than the F2. This may explain why our interaction analysis did not detect more than one clearly
significant interaction. Since most reproductive traits
are very complex it is not surprising that other studies
of intercrossed mice have resulted in the identification of
several gene interactions (Peripato et al. 2002, 2004).
Neonatal growth is a phenotype that can be expected
to be dependent on both litter size and milk production
(nutrient supply via milk). We have found several loci
for these traits on chromosome 9 (Neogq1 and Fecq4) and
Fecq3 on chromosome 1 mentioned above. Interestingly,
Everett et al. (2004) recently reported that they have
identified genetic regions affecting ovulation of primary
oocytes on chromosomes 1 and 9 (in the vicinity of the
regions we denoted Fecq3, Fecq4, and Neogq1). It is possible that we have identified the same loci, since ovulation rate, litter size, and neonatal growth may be related
phenotypes.
All linkage analyses described above are based on
mean and median values obtained from all pregnancies
(phenotypes that were not obviously affected by the
MHC of the male). We have also grouped our material
in another way (mean and median values of syngeneic
vs. allogeneic pregnancies) to clarify whether some phenotypes are particularly affected by the presence of
foreign MHC in the feto-placental unit (Table 4). As
shown in Figures 1 and 2B, there are regions on chromosomes 13 and 17 that appear to be specifically affected
by the type of mating. Here we have particularly focused
our interest on a region outside the MHC (44 cM) on
chromosome 17 (Pregq4) that is showing a highly significant QTL, indicating that this region contains genes
that are very important for pregnancy success in allogeneic mating (maternal/fetal MHC disparity).
From the data obtained we have decided to pay
special attention to regions where significant QTL have
been observed for more than one phenotype, i.e., Fecq3
and Fecq4 (litter size); Neogq1 (neonatal growth); Pregq1,
Pregq2, Pregq3, and Pregq4 (pregnancy rate); and Bwq7
(maternal body weight). The production of congenic
mice for these chromosomal regions is in progress and
will be an essential tool in the future identification of
specific genes. The data reported in this study give the
starting point for identifying individual genes that may
have an impact on breeding success. The identification
of specific genes that could then be crossed into
genetically modified model mice would have not only
economical benefits, but also ethical benefits as there
would be fewer mice needed.
We thank A. Bäcklund for comments and linguistics corrections.
This study was supported by Österlund’s fund, Crafoord’s fund, the
Gustav V 80 year foundation, The Royal Physiographic Society in
Lund, and Swegene.
LITERATURE CITED
Aluvihare, V. R., M. Kallikourdis and A. G. Betz, 2005 Tolerance, suppression and the fetal allograft. Mol. Med. 83: 88–96.
Anunciado, R. V., T. Ohno, M. Mori, A. Ishikawa and S. Tanaka,
2000 Distribution of body weight, blood insulin and lipid levels
in the SMXA recombinant inbred strains and the QTL analysis.
Exp. Anim. 49: 217–224.
Billington, W. D, 1993 Species diversity in the immunogenetic relationship between mother and fetus: Is trophoblast insusceptibility to immunological destruction the only essential common
feature for the maintenance of allogeneic pregnancy? Exp. Clin.
Immunogenet. 10: 73–84.
Broman, K. W., H. Wu, S. Sens and G. Churchill, 2003 R/qtl:
QTL mapping in experimental crosses. Bioinformatics 19: 889–
890.
Bulla, R., F. Fischetti, F. Bossi and F. Tedesco, 2004 Feto-maternal
immune interaction at the placental level. Lupus 13: 625–629.
Engwall, E., 1980 Enzyme immunoassay ELISA and EMIT.
Enzymology 70: 419–439.
Everett, C. A., C. A. Auchincloss, M. H. Kaufman, C.M. Abbott
and J. D. West, 2004 Genetic influences on ovulation of
Genetic Control of Mouse Reproduction
primary oocytes in LT/Sv strain mice. Reproduction 128: 567–
571.
Gustafsson, E., A. Mattsson, R. Holmdahl and R. Mattsson,
1994 Pregnancy in B-cell-deficient mice: postpartum transfer
of immunoglobulins prevents neonatal runting and death. Biol.
Reprod. 51: 1173–1180.
Horstgen-Schwark, G., E. J. Eisen, A. M. Saxton and T. R. Bandy,
1984 Reproductive performance in a dialer cross among lines
of mice selected for litter size and body weight. J. Anim. Sci.
58: 846–862.
Ihaka, R., and R. Gentleman, 1996 R: a language for data analysis
and graphics. J. Comput. Graph. Stat. 5: 299–314.
Kirkpatrick, B. W., A. Mangelt, N. Schulman and I. C. A. Martin,
1998 Identification of quantitative trait loci for prolificacy and
growth in mice. Mamm. Genome 9: 97–102.
Kondoh, G., H. Tojo, Y. Nakatani, N. Komazawa, C. Murata et al.,
2005 Angiotensin-converting enzyme is a GPI-anchored protein
releasing factor crucial for fertilization. Nat. Med. 11: 160–166.
Labosky, P. A., G. E. Winner, T. L. Jetton, L. Hargett, A. K. Ryan
et al., 1997 The winged helix gene, Mf3, is required for normal development of the diencephalon and midbrain, postnatal
growth and the milk-ejection reflex. Development 124: 1263–
1274.
Laird, W. P., A. Zijderveld, K. Linders, M. A. Rudnicki and R.
Jaenisch, 1991 Simplified mammalian DNA isolation procedure. Nucleic Acids Res. 15: 4293.
Mattsson, R., A. Mattsson, R. Holmdahl, A. Scheynius and P. H.
Van Der Meide, 1992 In vivo treatment with IFN-gamma
during early pregnancy in mice induces strong expression of
MHC class I and II molecules in uterus and decidua but not
in extra-embryonic tissues. Biol. Reprod. 46: 1176–1186.
Medaware, P. B., 1953 Some immunological and endocrinological
problems raised by the evolution of viviparity in vertebrates.
Symp. Soc. Exp. Biol. Evol. 11: 320–338.
Mellor, A. L., and D. H. Munn, 2000 Tryptophan catabolism and
T cell tolerance. Immunol. Today 21: 250.
Morris, K. H., A. Ishikawa and P. D. Keightley, 1999 Quantitative
trait loci for growth traits in C57BL/6J x DBA/2J mice. Mamm.
Genome 10: 225–228.
909
Munn, D. H., M. Zhou, J. T. Attwood, I. Bondarev, S. Conway et al.,
1998 Prevention of allogeneic fetal rejection by tryptophan
catabolism. Science 281: 1191–1193.
Peripato, A. C., R. A. De Brito, T. T. Vaughn, L. S. Pletscher,
S. R. Matioli et al., 2002 Quantitative trait loci for maternal performance for offspring survival in mice. Genetics 162: 1341–1353.
Peripato, A. C., R. A. De Brito, S. R. Matioli, L. S. Pletscher, T. T.
Vaughn et al., 2004 Epistasis affecting litter size in mice. J. Evol.
Biol. 17: 593–602.
Petroff, M. G., L. Chen, T. A. Phillips, D. Azzola, P. Sedlmayr et al.,
2003 B7 family molecules are favorably positioned at the human maternal-fetal interface. Biol. Reprod. 68: 1496–1504.
Raghupathy, R., 1997 Th1 type immunity is incompatible with successful pregnancy. Immunol. Today 18: 478–482.
Rocha, J. L., E. J. Eisen, F. Siewerdt, L. D. Van Vleck and D. Pomp,
2004 A large-sample QTL study in mice: III. Reproduction.
Mamm. Genome 15: 878–886.
Spearow, J. L., and M. Barkley, 1999 Genetic control of hormoneinduced ovulation rate in mice. Biol. Reprod. 61: 851–856.
Spearow, J. L., P. A. Nutson, W. S. Mailliard, M. Porter and M.
Barkley, 1999a Mapping genes that control hormone-induced
ovulation rate in mice. Biol. Reprod. 61: 857–872.
Spearow, J. L., P. Doemeny, R. Sera, R. Leffler and M. Barkley,
1999b Genetic variation in susceptibility to endocrine disruption by estrogen in mice. Science 285: 1259–1261.
Sunderland, C. A., M. Naiem, D. Y. Mason, C. W. Redman and
G. M. Stirrat, 1981 The expression of major histocompatibility antigens by human chorionic villi. J. Reprod. Immunol. 3:
323–331.
Svensson, L., M. Arvola, M.-A. Sällström, R. Holmdahl and
R. Mattsson, 2001 The Th2 cytokines IL-4 and IL-10 are not
crucial for the completion of allogeneic pregnancy in mice.
J. Reprod. Immunol. 51: 3–7.
Thurman, J. M., D. M. Kraus, G. Girardi, D. Hourcade, H. J. Kang
et al., 2005 A novel inhibitor of the alternative complement
pathway prevents antiphospholipid antibody-induced pregnancy
loss in mice. Mol. Immunol. 42: 87–97.
Communicating editor: C. A. Kozak