SciELO - Scientific Electronic Library Online

 
vol.21 issue2Nucleolar organizer regions and a new chromosome number for Rhea americana (Aves: Rheiformes)Microphotometrical image analysis of the subtelomeric region of T-banded endoreduplicated chromosomes of Chinese hamster ovary cells author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand

Journal

Article

Indicators

Related links

Share


Genetics and Molecular Biology

Print version ISSN 1415-4757On-line version ISSN 1678-4685

Genet. Mol. Biol. vol. 21 n. 2 São Paulo June 1998

https://doi.org/10.1590/S1415-47571998000200007 

Genetic variation in body weight gain and composition in the intercross of Large (LG/J) and Small (SM/J) inbred strains of mice

 

Melissa G. Kramer, Ty T. Vaughn, L. Susan Pletscher, Kelly King-Ellison, Emily Adams, Christopher Erikson and James M. Cheverud
Department of Anatomy & Neurobiology, Washington University School of Medicine, 660 S. Euclid Ave., St. Louis, MO 63110, USA. Send correspondence to M.G.K. c/o J.M.C. Fax: 314-362-3446, E-mail: cheverud@thalamus.wsutl.edu 

 

 

ABSTRACT

Strain intercross experiments provide a powerful means for mapping genes affecting complex quantitative traits. We report on the genetic variability of the intercross of the Large (LG/J) and Small (SM/J) inbred mouse strains as a guide to gene mapping studies. Ten SM/J males were crossed to 10 LG/J females, after which animals were randomly mated to produce F1, F2, and F3 intercross generations. The 1632 F3 animals from 200 full-sib families were used to estimate heritabilities and genetic correlations of the traits measured. A subset of families was cross-fostered at birth to allow measurement of the importance of post-natal maternal effects. Data was collected on weekly body weight from one to 10 weeks and on organ weights, body weight, reproductive fat pad weight, and tail length at necropsy in the intercross generations. There was no heterosis for age-specific weights or necropsy traits, except that one-week weight was the highest in the F2 generation, indicating heterosis for maternal effect in the F1 mothers. We found moderate to high heritability for most age-specific weights and necropsy traits. Maternal effects were significant for age-specific weights from one to four weeks but disappeared completely at ten-week weight. Maternal effects for necropsy traits were low and not statistically significant. Age-specific weights showed a typical correlation pattern, with correlation declining as the difference in ages increased. Among necropsy traits, reproductive fat pad and body weights were very highly genetically correlated. Most other genetic correlations were low to moderate. The intercross between SM/J and LG/J inbred mouse strains provides a valuable resource for mapping quantitative trait loci for body size, composition, and morphology.

 

 

INTRODUCTION

Recent work in quantitative and molecular genetics makes it possible to determine the individual loci affecting quantitative traits using known markers dispersed throughout the genome (Lander and Botstein, 1989). In many ways, mice are ideal candidates for such studies because of their relatively fast generation time; their relationship to humans including both common mammalian heritage and extensive synteny; the existence of a large number of specialized and genetically characterized strains for experimental work (Lyon et al., 1996), and availability of a multitude of highly polymorphic microsatellite markers (Dietrich et al., 1996).

Two specially inbred mouse strains, Large (LG/J) and Small (SM/J), are an excellent choice for mapping body growth and composition genes (Chai, 1956a,b, 1957, 1961, 1968; Cheverud et al., 1996). Both strains have been selected for size at 60 days: LG/J for large size (Goodale, 1938, 1941) and SM/J for small size (MacArthur, 1944), and both have been maintained by inbreeding using sib-matings. These strains have been well characterized by extensive work on the weight variation between them (Chai, 1956a,b, 1957, 1961, 1968; Cheverud et al., 1996). The 60-day weight difference between the strains is approximately 9.5 within-strain standard deviations, and the large amount of genetic variation responsible for this difference is known to be due to quantitative genetic properties found in natural populations (Chai, 1956a,b). Twenty-seven QTLs (quantitative trait loci) affecting growth in weight and 10-week body weight were mapped in the F2 intercross of these strains (Cheverud et al., 1996). In this study, we examine the quantitative genetics of growth, body size and body composition in the intercross between the LG/J and SM/J strains in order to determine the causal basis of these phenotypes. These quantitative genetic results provide a context for evaluating the comprehensiveness of the ongoing single gene mapping work.

 

MATERIAL AND METHODS

Mating protocol

Ten SM/J males and ten LG/J females were mated to yield 52 F1 progenies. Due to the inbred nature of the parental stocks, all F1 progenies are genetically identical and heterozygous at all autosomal loci differing between the LG/J and SM/J strains. The F1 generation was then intercrossed to yield 510 F2 progenies that upon random mating yielded 1632 F3 animals distributed across 200 independent full-sib families. Quantitative genetic statistics were estimated using these 200 families.

Rearing protocol

Males were removed from the mating cages once the female was gravid. F2 females giving birth on the same day were paired and half of the offspring from each litter were switched to be fostered by the other dam. Mice were weaned at 21 days and placed randomly in single-sex cages with five animals each. At 10 weeks animals were either mated to produce the F4 generation (Cheverud, J.M., Vaughn, T., Pletscher, L.S., King-Ellison, K., Bailiff, J., Adams, E., Erickson, C. and Bonislawski, A., unpublished results) or sacrificed. Mating animals were sacrificed after their litters were weaned.

Data collected

Beginning seven days after birth, each mouse was weighed weekly using a digital balance linked to a computer. At 10 weeks of age the mice that were not to be mated were sacrificed and necropsied. At necropsy, the animal's total weight and the weights of several organs, including the heart, kidney, spleen, and liver, were collected. Male testicular and female parametrial fat pad weights were measured as an indicator of fatness. Tail length was measured as an indicator of skeletal size.

Data treatment

The weekly weight data were inspected for anomalous entries. As a correction for missing weights, the expected weight based on the midpoint of flanking weight measurements was substituted for the missing record. The necropsy data were also inspected for anomalous entries using SYSTAT (Wilkinson, 1997). Extreme outlying values were removed to avoid biasing the analysis.

The F2 data were analyzed using a MANOVA to check for significant effects on the weekly weight and necropsy data due to dam, litter size at birth, litter size at weaning, date of birth, birth cohort, age at necropsy and sex. Litter sizes at birth, litter sizes at weaning, and birth dates were found to have no significant effects on any of the variables. The F1 dam affected all variables except heart and spleen weight. This must be an environmental maternal effect because F1 dams were genetically identical. Birth cohort affected fat pad, heart, spleen, liver, and week 1-3 weights. Sex differences were found for all measurements except spleen and liver weights. Those variables significantly affected by these covariates (P < 0.05) were corrected. To correct variable X for the effects of factor Y, the overall mean of X was added to each individual's X value and the mean of individuals sharing a common Y was subtracted. The F3 data were corrected in a similar fashion except that dam effects, which could be both genetic and environmental, were retained. Further analyses were done using these corrected F2 and F3 data.

Quantitative genetic analyses

Three different analyses were performed to calculate the heritability, phenotypic variance and coefficient of variation for the necropsy data and weekly weight data, as well as the genetic, environmental, phenotypic, and in some cases maternal correlations between the traits.

Parent-offspring analysis

The necropsy and weekly weight data were averaged for each of the full-sib families and then paired with their respective sire and dam of the litter. A midparent average was also calculated for each of the traits in each family. The sibship and midparent averages were used for this analysis. Covariances and Pearson product moment correlations between sibship means and midparent values were calculated using SYSTAT (Wilkinson, 1997) for each of the traits. The additive genetic variance was estimated by the midparent-offspring covariance for a single trait. The heritability of each trait was found by a regression of sibship average phenotype on midparent phenotype using SYSTAT (Wilkinson, 1997). The phenotypic variance could be calculated from these two values as additive genetic variance divided by heritability. Coefficients of variation were calculated as the standard deviation divided by the mean of the trait.

Genetic covariances were estimated as (1/2) [crcov(OX,MY) + crcov(OY,MX)], where crcov is a cross-relative covariance, O represents the sibship mean and M represents the midparent value for the specified trait (X or Y). Genetic correlations were calculated in the ordinary way by dividing the genetic covariance by the geometric mean of the genetic variances of the traits. The phenotypic correlation between traits X and Y for the midparent was calculated by the formula: rP = cov(X,Y)/Ö[var(X) var(Y)]. The environmental correlations were calculated as the weighted difference between the phenotypic correlation and genetic correlation for each trait (Falconer and Mackay, 1996).

Sibling analysis (without maternal effects)

ANOVA was performed using SYSTAT (Wilkinson, 1997) with family as the treatment and the measured traits as effects. The family variance measures the variance due to the differences between sibships, while the residual variance measures the variance due to differences within sibships. Variance components were calculated as in a one-way, unbalanced design (Sokal and Rohlf, 1995). The genetic variance/covariance was calculated as twice the between-sib variance/covariance. The total variance/covariance was calculated as the sum of the between-sib and within-sib components. The environmental variance/covariance matrix was then calculated as the difference between the total variance/covariance matrix and the genetic variance/covariance matrix. These results were used to calculate the genetic, phenotypic, and environmental correlations: r(XY) = cov(X,Y)/Ö[(var(X) var(Y)]. Heritability was calculated by dividing the genetic variance by the phenotypic variance for each of the traits.

Sibling analysis (with maternal effects)

ANOVA was also performed using the subset of F3 sibships that participated in cross-fostered pairs. This two-way ANOVA was calculated separately for each of the 51 pairs of cross-fostered sibships and then pooled across pairs. This analysis provides a variance due to dam, a variance due to post-natal maternal effects, and a residual variance (Sokal and Rohlf, 1995). A bootstrap analysis with 1000 iterations was performed with the 51 F3 cross-fostered pairs as the sampling unit in order to obtain standard errors for the parameters of the model. The inclusion of the nurse variance in the model measures variability due to post-natal maternal effects and allows the calculation of maternal effect variances and correlations.

QTL data

The additive and dominance variance due to QTLs for age-specific weights was estimated using a multiple regression with age-specific weight as the dependent variable and the additive and dominance genotypic values of all identified QTLs (Cheverud et al., 1996) as the independent variables. These calculations were done using the genotypes and phenotypes from 535 F2 animals from an earlier intercross experiment (Cheverud et al., 1996). The data collection for this is described in detail in Cheverud et al. (1996). Separate analyses were also performed including only additive and only dominance genotypic values.

 

RESULTS

Age-specific weights

Age- and sex-specific body weights and their standard errors were calculated (Table I). The parental strains, LG/J and SM/J, were significantly different from three weeks on up to 10 weeks of age. From five weeks on, differences were on the order of six to eight within-strain standard deviation units for each sex. In general, differences between the mid-strain values (P0) and intercross generation means were small. There was no evidence of heterosis for body weight in this cross, the only significant differences involved P0 being slightly larger than the F1 or F2 at some ages. The P0 generation was never significantly different from the F3 values. F1 females were generally slightly larger than F2 females, although none of these differences reached Bonferroni-adjusted levels of significance. F3 and P0 females were slightly larger than the F1 and F2 females and this approximately 0.25 standard deviation difference was statistically significant for the F2 to F3 comparison at most ages. The exception to these size relationships occurred at the first week of age. Seven-day weight was the highest for the F2 females and the difference was significant for the F2 to F3 comparison. At most ages, F3 and P0 males were the largest ones, followed by F2 and F1 males in that order. However, these differences were rarely statistically significant. In contrast to the other ages, F2 males were also significantly larger than both F1 and F3 males at one week of age.

 

Table I - Age- (in weeks) and sex-specific mean weights in grams, standard deviations (SD), and sample sizes (N) for males and females of the parental strains (LG/J and SM/J), the midparent value (P0), and the first three intercross generations (F1, F2, and F3).

Males

LG/J

SM/J

P0

F1

F2

F3

 

N

Significant
differences

Age

Mean

SD

N

Mean

SD

N

Mean

SD

N

Mean

SD

N

Mean

SD

N

Mean

SD

1

4.02

0.77

10

4.43

0.54

7

4.23

0.66

17

3.88

0.97

29

4.61

0.99

258

4.26

0.83

866

F2 > F1, F2 > F3

2

7.75

1.29

10

8.05

1.14

7

7.90

1.22

17

7.00

1.56

29

7.19

1.23

258

7.33

1.28

866

 

3

12.50

2.11

10

10.95

1.00

7

11.72

1.55

17

10.70

2.68

29

11.33

2.08

258

11.29

1.92

866

 

4

21.77

3.41

10

15.55

1.22

7

18.66

2.32

17

17.48

3.59

29

17.97

2.87

258

18.30

2.66

866

 

5

30.29

2.91

10

18.09

1.02

7

24.19

1.97

17

22.05

3.36

29

23.08

2.76

258

23.54

2.61

866

F1 > F3

6

34.12

2.73

10

18.91

1.26

7

26.52

1.99

17

25.04

2.26

29

25.44

2.84

258

25.81

2.95

866

 

7

36.49

3.16

10

20.35

2.40

7

28.42

2.78

17

26.40

2.20

29

27.13

3.13

258

27.58

3.29

866

 

8

38.69

3.39

10

20.82

1.31

7

29.75

2.35

17

27.74

2.44

29

28.49

3.46

258

29.16

3.64

866

 

9

41.07

3.68

10

22.03

1.44

7

31.55

2.56

17

28.73

2.59

29

30.00

3.73

258

30.54

3.96

866

P0 > F1

10

42.25

4.24

10

22.76

1.36

7

32.50

2.80

17

29.77

2.58

29

31.16

3.94

258

31.72

4.23

866

 

Females

LG/J

SM/J

P0

F1

F2

F3

 

N

Significant
differences

Age

Mean

SD

N

Mean

SD

N

Mean

SD

N

Mean

SD

N

Mean

SD

N

Mean

SD

differences

1

4.20

0.87

10

4.49

0.46

6

4.35

0.66

16

3.86

1.00

23

4.37

0.85

243

4.18

0.88

766

F2 > F3

2

8.04

1.34

10

7.49

0.84

6

7.77

1.09

16

7.16

1.77

23

6.85

1.27

243

7.25

1.21

766

F3 > F2

3

12.16

2.35

10

10.55

0.52

6

11.35

1.43

16

10.63

2.87

23

10.70

1.94

243

10.95

1.76

766

-

4

19.75

2.68

10

13.59

0.44

6

16.67

1.56

16

16.01

3.34

23

15.67

2.29

243

16.14

2.18

766

F3 > F2

5

26.02

2.61

10

15.40

0.42

6

20.71

1.51

16

19.40

2.70

23

19.02

2.29

243

19.68

2.25

766

P0, F3 > F2

6

28.18

2.63

10

16.11

0.35

6

22.15

1.49

16

20.65

2.31

23

20.40

2.34

243

20.96

2.44

766

P0, F3 > F2

7

29.51

3.26

10

17.00

1.05

6

23.26

2.15

16

21.44

2.11

23

21.34

2.59

243

21.97

2.67

766

F3 > F2

8

31.48

3.01

10

17.55

0.81

6

24.52

1.91

16

22.43

2.19

23

22.45

2.94

243

23.07

2.93

766

P0 > F2

9

32.76

2.86

10

17.96

1.04

6

25.36

1.95

16

24.09

3.20

23

23.35

3.09

243

24.02

3.21

766

P0, F3 > F2

10 34.41 3.58 10 18.66 1.35 6 26.53 2.46 16 26.09 4.41 23 24.18 3.50 243 24.89 3.40 766 -

 

The heritabilities of age-specific weights obtained using midparent-offspring regression, full-sib analysis, cross-fostering analysis, and heritability and percent dominance variance explained by identified single loci were determined (Table II). All heritabilities were statistically significant. Heritability estimates were often highest for the full-sib analysis, followed by the cross-fostering analysis and the midparent-offspring regressions. The midparent-offspring regressions directly estimate the additive genetic variance for a trait. The sibship analysis is biased in that its heritability estimates also contain twice the maternal effect variance and half of the dominance effect variance. The cross-fostered sibship analysis is less biased in that maternal effects are separated from the additive effects, although half the effects of dominance variance persist in these estimates.

 

Table II - Heritabilities (h2) and variance components (Vp) of age-specific weights calculated from parent-offspring regressions, from variation among full-sibships, and from pairs of cross-fostered sibs. Proportions of variance accounted for by additive (%VA) and dominance (%VD; %G = (%VA + %VD)) effects of quantitative trait loci (QTLs) in the earlier F2 intercross of the SM/J and LG/J strains are also included (Cheverud et al., 1996).

Midparent-offspring

Full-sibships

Cross-fostered pairs

QTLs

Week

h2

VP

h2

VP

h2

m2

VP

%VA

%VD

%VG

1

0.21

0.508

0.99

0.654

0.60

0.21

0.608

0.12

0.12

0.24

2

0.17

0.895

0.91

0.925

0.46

0.30

0.889

0.16

0.12

0.26

3

0.28

2.250

1.03

2.472

0.54

0.24

2.082

0.23

0.15

0.35

4

0.33

4.217

0.84

4.938

0.58

0.15

4.788

0.25

0.09

0.33

5

0.42

5.618

0.82

5.216

0.69

0.08

5.415

0.31

0.10

0.39

6

0.52

6.500

0.82

6.610

0.76

0.04

7.076

0.37

0.13

0.44

7

0.55

8.233

0.80

8.188

0.81

0.03

8.883

0.39

0.10

0.45

8

0.50

10.250

0.75

10.030

0.82

0.02

10.845

0.38

0.10

0.44

9

0.56

11.030

0.75

11.900

0.81

0.01

12.948

0.40

0.10

0.46

10

0.57

12.641

0.76

13.510

0.82

0.01

14.631

0.40

0.11

0.47

 

Judging from the midparent-offspring and cross-fostered sibship analyses, heritability tends to increase with age. This may be a function of increased measurement repeatability at higher ages and/or a true tendency for heritable variance to increase with age. However, after accounting for differences in means at different ages by using the coefficient of variation, the additive genetic coefficient of variation levels remained stable across ages at about 7-8% for the midparent-offspring regression and 9-10% for the cross-fostered sibship analysis. The apparent decrease in heritability and genetic coefficients of variation with age in the sibship analysis is due to the inclusion of double the maternal effect variance with the additive genetic variance when estimating heritability. When this effect is controlled using age-specific percent maternal effect variances from the cross-fostered sibship analysis, the trend for decreased heritability with age is eliminated. Taking all factors into account, heritability increased with age from a moderate to a high level.

Percents of variance accounted for by the additive and dominance effects at quantitative trait loci from the first F2 intercross were typically lower than found through biometrical analysis. This indicates the possibility of more factors of very small effect segregating in the intercross than were detected in the earlier genetic mapping experiment. Further mapping with the second F2 intercross population should help clarify the sources of the variance currently not accounted for.

The genetic and phenotypic correlations among age-specific weights were obtained from the midparent-offspring regressions (Table III). Correlations estimated in the sibship and cross-fostering analyses were virtually identical to those obtained from the midparent-offspring regression (matrix correlations > 0.96). Correlations between age-specific weights declined from values at or near one to values below 0.30 as the interval between ages increased. One-week weight shared only about 8% of its genetic variance with ten-week weight. A principal component analysis of the genetic correlation matrix produced a first component representing size at all ages and accounting for 78% of the total genetic variance and a second component contrasting early with later growth and accounting for 18% of the genetic variance (see Table IV).

 

Table III - Genetic (below diagonal) and phenotypic (above diagonal) correlations for age-specific weights.

RP

RG

Week

1

2

3

4

5

6

7

8

9

10

1

1.00

0.71

0.72

0.63

0.51

0.45

0.39

0.36

0.34

0.34

2

0.92

1.00

0.86

0.72

0.61

0.56

0.49

0.46

0.44

0.43

3

0.92

0.91

1.00

0.86

0.71

0.64

0.56

0.53

0.50

0.49

4

0.72

0.76

0.90

1.00

0.88

0.77

0.69

0.65

0.63

0.61

5

0.48

0.59

0.65

0.88

1.00

0.93

0.87

0.83

0.80

0.77

6

0.41

0.53

0.61

0.84

0.98

1.00

0.95

0.91

0.89

0.86

7

0.37

0.57

0.57

0.78

0.94

0.98

1.00

0.96

0.94

0.92

8

0.35

0.61

0.59

0.77

0.94

0.98

0.99

1.00

0.96

0.95

9

0.36

0.60

0.60

0.76

0.92

0.97

0.99

0.99

1.00

0.96

10

0.29

0.60

0.57

0.72

0.89

0.95

0.98

0.99

0.99

1.00

 

Table IV - First two principal components of the genetic correlation matrix for age-specific weights.

Trait Week

Component

1

2

1

0.62

-0.77

2

0.78

-0.56

3

0.80

-0.57

4

0.92

-0.23

5

0.95

0.17

6

0.95

0.27

7

0.94

0.31

8

0.95

0.30

9

0.95

0.29

10

0.92

0.33

Eigenvalues

7.83

1.76

 

Necropsy data

Means and heritabilities of organ sizes and necropsy weight were determined (Table V). Heritabilities were the highest for body weight and tail length and the lowest for the smaller organs, such as the heart and spleen. Fat pad, kidney and liver weight were moderately heritable. Heritability patterns were fairly consistent across the three analyses, although in general heritabilities estimated from the sibship and cross-fostering analyses were slightly larger than those obtained by midparent-offspring regression. There was little evidence of maternal effects for these necropsy characters, the 6% maternal effects for kidney weight and fat pad weight being the largest values. None of the maternal effect variances were significant at the 5% level after Bonferroni correction for multiple tests, although many probabilities approached statistical significance.

 

Table V - Means, heritabilities (h2) and variance components (Vp) of necropsy traits calculated from parent-offspring regressions, from variation among full-sibships, and from pairs of cross-fostered sibs.

 

Midparent-offspring

Full-sibships

Cross-fostered pairs

Mean

h2

VP

h2

VP

h2

m2

VP

Fat pad

1.04

0.34

0.3159

0.45

0.3678

0.39

0.06

0.3340

Weight

34.23

0.44

25.0770

0.65

25.0930

0.58

0.04

23.9650

Tail

92.41

0.46

36.3060

0.73

30.2240

0.71

0.04

31.6540

Heart

0.16

0.07

0.0007

0.25

0.0007

0.20

0.04

0.0008

Kidney

0.29

0.21

0.0028

0.46

0.0022

0.34

0.06

0.0022

Spleen

0.11

0.19

0.0008

0.23

0.0007

0.26

0.00

0.0008

Liver

2.01

0.16

0.1906

0.43

0.1212

0.45

0.02

0.1160

 

The genetic and phenotypic correlations for necropsy traits were calculated using the midparent-offspring regressions (Table VI). Correlation matrices based on the sibship and cross-fostering analyses were quite similar to that based on the midparent-offspring regressions. The highest genetic correlation was between the weight of the reproductive fat pad and total body weight. These two traits shared most of their genetic variability. Fat pad weight was moderately correlated with the remaining traits. Other particularly high correlations were found between weight and liver, heart and kidney, and tail and heart. Spleen had relatively low correlations with other traits. The remaining correlations were moderate. The first principal component of the genetic correlation matrix for necropsy traits (see Table VII) accounted for 61% of the genetic variance and represented overall size. The second component, accounting for 15% of the variance, contrasted the heart and kidney with the remaining traits. The third principal component, accounting for 13% of the variance, contrasted the spleen with the other traits while the fourth principal component (8% of the variance) contrasted the kidney and liver with the tail.

 

Table VI - Genetic (below diagonal) and phenotypic (above diagonal) correlations for necropsy traits.

 

RP

Fat pad

Weight

Tail

Heart

Kidney

Spleen

Liver

RG

Fat pad

1.00

0.77

0.21

0.17

0.26

0.15

0.35

Weight

0.92

1.00

0.41

0.37

0.56

0.30

0.69

Tail

0.46

0.59

1.00

0.23

0.31

0.23

0.34

Heart

0.44

0.56

0.84

1.00

0.47

0.22

0.44

Kidney

0.28

0.64

0.49

0.91

1.00

0.21

0.56

Spleen

0.21

0.43

0.35

0.19

0.32

1.00

0.38

Liver 0.59 0.79 0.47 0.59 0.59 0.55 1.00

 

 

Table VII - First four principal components of the genetic correlation matrix for necropsy traits.

 

Component

Trait

1

2

3

4

Fat pad

0.73

0.45

0.50

0.05

Weight

0.91

0.32

0.21

-0.09

Tail

0.78

-0.29

0.04

0.56

Heart

0.85

-0.54

0.06

0.02

Kidney

0.79

-0.43

-0.16

-0.39

Spleen

0.51

0.38

-0.73

0.17

Liver

0.84

0.25

-0.16

-0.22

Eigenvalues

4.29

1.07

0.89

0.56

 

Genetic correlations were calculated between midparent and offspring values for the age-specific weights with the necropsy data (Table VIII). Correlations between necropsy traits and early age-specific weights were quite low. However, the genetic correlations of 10-week weight and necropsy traits tended to be even higher than the correlations between necropsy body weight and the remaining necropsy traits. The exception to this was the relatively high correlation between fat pad weight and necropsy weight. The correlations between necropsy traits and age-specific weights tended to increase from one to five weeks, after which they stabilized at the value seen at 10 weeks.

 

Table VIII - Genetic correlations between weekly weights and necropsy data calculated using the midparent-offspring regression.

Week

Fat pad

Weight

Tail

Heart

Kidney

Spleen

Liver

1

-0.16

0.15

0.36

0.03

0.32

-0.03

0.08

2

0.41

0.61

0.58

0.14

0.38

0.26

0.32

3

0.14

0.51

0.46

0.22

0.42

0.25

0.36

4

0.35

0.63

0.63

0.62

0.56

0.22

0.52

5

0.62

0.81

0.64

0.72

0.61

0.36

0.61

6

0.63

0.86

0.65

0.74

0.65

0.40

0.65

7

0.63

0.92

0.63

0.74

0.66

0.44

0.72

8

0.66

0.95

0.64

0.79

0.71

0.43

0.78

9

0.71

0.98

0.63

0.77

0.72

0.44

0.78

10

0.73

0.97

0.62

0.82

0.78

0.48

0.78

 

DISCUSSION

The population formed by the intercross of LG/J and SM/J strains of mice displayed moderate to high levels of heritability for age-specific weights and adult morphology, making it a suitable population in which to map individual gene loci affecting these traits. In particular, moderate to high heritabilities were obtained for adult weight and weight of the reproductive fat pad, indicating the utility of these strains for investigating obesity. The high genetic correlation observed between necropsy weight and fat pad weight indicates that most genes affecting fat pad size should have already been discovered in our earlier mapping experiment for adult body weight (Cheverud et al., 1996). However, the observation that necropsy weight was less well correlated with organ weights at necropsy and a skeletal size measure (tail length) than is 10-week weight indicates that most later (post-10 week) weight gain may be due to increases in fatness.

Age-specific weights were not significantly different between the parental strains until 3 weeks of age indicating no substantial differences in pre-natal growth. After 3 weeks, the LG/J strain grew at a much higher rate than the SM/J strain, so that from 5 weeks on the parental strains differed by 6 to 8 within-strain standard deviation units. This interstrain difference is slightly less than the 9.5 intra-strain standard deviation difference measured by Chai (1956a,b) over 40 years ago. The reduction in difference is due to the SM/J males and females being 7 and 5 grams larger today, respectively, than they were 40 years ago. It is possible that over the course of approximately 120-150 generations, the SM/J strain has responded to fecundity and/or viability selection for increased size. Genetic variability for this selection response would arise through new mutations in the SM/J strain. The intercross populations have average body weights similar to the midpoint of the parental strains. Age-specific weights showed little change over the first several generations of the intercross experiment. One striking pattern, repeated across the sexes, is that one-week weight was greater in the F2 generation than in the P0, F1 or F3 generations. This pattern is expected under heterosis for the maternal contribution to her offspring's one-week weight (Falconer and Mackay, 1996) and suggests overdominance at maternal effect loci acting on offspring size at birth. This pattern disappeared by two and three weeks of age, suggesting that heterosis is for pre-natal maternal effects.

Levels of heritability rose from a low of about 20% at one week of age to a high of about 60% at ten weeks. Heritability estimates based on the between sibship variance are severely positively biased at the younger ages by including twice the maternal effect variance in their estimate. The cross-fostered sibship analysis accounts for this bias but still leaves one-half the dominance variance percentage confounded with heritability. Even so, the cross-fostered between-sibship analysis provides higher heritability estimates than are obtained from midparent-offspring regressions. The percent of additive genetic variance accounted for by identified QTLs in our previous mapping experiment (Cheverud et al., 1996) is about 60-80% of the estimate provided by the parent-offspring analysis. This indicates that while the majority of the additive variance in body weight is accounted for by identified QTLs there may be other loci segregating in this cross with effects too small to be easily mapped.

Genetic correlations between age-specific weights followed the common pattern noted for such trait sets, with the level of the correlation being based on the length of time between weight measurements. Early and late weights shared less than 10% of their genetic variation. This corresponds to the findings of our earlier mapping study, in which we determined that variation in early and later growth was largely due to different sets of genes (Cheverud et al., 1996). While a balance between positive and negative pleiotropy could also produce the observed pattern, we did not find this for individual loci in the earlier study (Cheverud et al., 1996).

Heritabilities of necropsy traits were moderate to high, corresponding to the results for later age-specific weights. The relatively low heritabilities of spleen and heart are likely due to the small size and range of variability of these organs relative to the accuracy with which they can be weighed during necropsy. Heritability levels for fat pad, necropsy weight, and tail length indicate that QTL mapping for body composition in the F2 intercross population should be successful. Genetic correlations between body composition variables indicate that most genetic variation for skeletal size and fatness is independent. We hypothesize that this relative lack of correlation is due to the two traits being affected by largely separate sets of genes. This hypothesis will be tested in the F2 intercross population described here. Genetic correlations between 10-week weight and necropsy traits were typically similar to the correlations between necropsy weight and the other necropsy traits. Fat pad weight, however, had a notably lower correlation with 10-week weight (rG = 0.73) than it did with necropsy weight (rG = 0.92). This was not true for tail length (rG = 0.62 with 10-week weight and rG = 0.59 with necropsy weight). This indicates that weight gain after 10 weeks may be largely due to gain in fat, rather than gain in skeletal growth.

We have shown here that the intercross of LG/J and SM/J inbred strains had great potential for mapping genes affecting body growth and composition. The further investigation of these strains should lead to the discovery of several new genes affecting body composition.

 

ACKNOWLEDGMENTS

We wish to thank all those who have helped on this project over the years, including Dr. Francisco A.M. Duarte, Dr. Eric Routman, Christy Perel, Dr. Natalia Vesey, Dr. Shemelis Beyene, Kilinyaa Cothran, Eirik Cheverud, Bruno van Swinderen, and Jeff Bailiff. Melissa G. Kramer is a Howard Hughes Medical Institute Predoctoral Fellow. We also gratefully acknowledge the support of Washington University School of Medicine, NSF grant DEB-9419992 and NIDDKD grant DK52514 for this work.

 

RESUMO

Os experimentos de intercruzamento de cepas constituem um valioso meio de mapear genes que afetam caracteres quantitativos complexos. Nós estudamos a variabilidade genética do intercruzamento das cepas procriadas por endogamia de camundongos Large (LG/J) e Small (SM/J), como um guia para estudos de mapeamento gênico. Dez machos SM/J foram cruzados com 10 fêmeas LG/J, após o que os animais foram acasalados aleatoriamente para produzir as gerações intercruzadas F1, F2 e F3. Os 1632 animais da geração F3 das 200 famílias de irmãos completos foram usados para calcular a herdabilidade e as correlações genéticas dos caracteres medidos. Um subgrupo de famílias foi submetido a adoção cruzada após o nascimento, para permitir a avaliação da importância dos efeitos maternos pós-natais. Foram coletados dados semanais de peso corporal da 1ª à 10ª semanas, assim como dados de necropsia do peso corporal, peso de órgãos, peso do coxim gorduroso reprodutor e comprimento de cauda nas gerações intercruzadas. Não houve heterose para pesos específicos para idade ou caracteres de necropsia, exceto pelo fato de que o peso da 1ª semana foi o maior na geração F2, indicando heterose dos efeitos maternos nas mães da geração F1. Encontramos herdabilidade moderada/alta para a maioria dos pesos específicos para a idade e dos caracteres de necropsia. Os efeitos maternos foram significantes para os pesos específicos para a idade da 1ª à 4ª semana, mas desapareceram completamente para o peso de 10 semanas. Os efeitos maternos para os caracteres de necropsia foram pequenos e não estatisticamente significantes. Os pesos específicos para a idade mostraram um padrão de correlação típico, com declínio da correlação à medida que a diferença em idades aumentava. Entre os caracteres de necropsia, os pesos corporal e do coxim gorduroso reprodutor mostraram alta correlação genética. A maioria das outras correlações genéticas foram pequenas ou moderadas. O intercruzamento entre cepas procriadas por endogamia de camundongos SM/J e LG/J constituem fonte valiosa para o mapeamento de loci de caracteres quantitativos para o tamanho corporal e sua composição e morfologia.

 

REFERENCES

Chai, C. (1956a). Analysis of quantitative inheritance of body size in mice. I. Hybridization and maternal influence. Genetics 41: 157-164.         [ Links ]

Chai, C. (1956b). Analysis of quantitative inheritance of body size in mice. II. Gene action and segregation. Genetics 41: 167-178.         [ Links ]

Chai, C. (1957). Analysis of quantitative inheritance of body size in mice. III. Dominance. Genetics 42: 601-607.         [ Links ]

Chai, C. (1961). Analysis of quantitative inheritance of body size in mice. IV. An attempt to isolate polygenes. Genet. Res. 2: 25-32.         [ Links ]

Chai, C. (1968). Analysis of quantitative inheritance of body size in mice. V. Effects of small numbers of polygenes on similar genetic backgrounds. Genet. Res. 11: 239-246.         [ Links ]

Cheverud, J., Routman, E., Duarte, F.M., van Swinderen, B., Cothran, K. and Perel, C. (1996). Quantitative trait loci for murine growth. Genetics 142: 1305-1319.         [ Links ]

Dietrich, W., Miller, J., Steen, R., Merchant, M., Damron-Boles, D., Husain, Z., Dredge, R., Daly, M., Ingalls, K., O'Connor, T., Evans, C., DeAngelis, M., Levison, D., Kruglyak, L., Goodman, N., Copeland, N., Jenkins, N., Hawkins, T., Stein, L., Page, D. and Lander, E. (1996). A comprehensive genetic map of the mouse genome. Nature 380: 149-152.         [ Links ]

Falconer, D.S. and Mackay, T.F. (1996). Introduction to Quantitative Genetics. 4th end. Longman Press, New York.         [ Links ]

Goodale, H. (1938). A study of the inheritance of body weight in the albino mouse by selection. J. Hered. 29: 101-112.         [ Links ]

Goodale, H. (1941). Progress report on possibilities in progeny test breeding. Science 94: 442-443.         [ Links ]

Lander, E.S. and Botstein, D. (1989). Mapping Mendelian factors underlying quantitative traits using RFLP linkage maps. Genetics 121: 185-199.         [ Links ]

Lyon, M.F., Rastan, S. and Brown, S.D.M. (1996). Genetic Variants and Strains of the Laboratory Mouse. 3rd edn. Oxford University Press, Oxford.         [ Links ]

MacArthur, J. (1944). Genetics of body size and related characters. I. Selection of small and large races of the laboratory mouse. Am. Nat. 78: 142-157.         [ Links ]

Sokal, R. and Rohlf, F.J. (1995). Biometry. W.H. Freedman and Co., New York.         [ Links ]

Wilkinson, L. (1997). SYSTAT 7.0 for Windows. SPSS Inc., Chicago.         [ Links ]

 

(Received March 9, 1998)

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License