Nasal bone shape is under complex epistatic genetic control in mouse interspecific recombinant congenic strains

Abstract

Background: Genetic determinism of cranial morphology in the mouse is still largely unknown, despite the localization of putative QTLs and the identification of genes associated with Mendelian skull malformations. To approach the dissection of this multigenic control, we have used a set of interspecific recombinant congenic strains (IRCS) produced between C57BL/6 and mice of the distant species Mus spretus (SEG/Pas). Each strain has inherited 1.3% of its genome from SEG/Pas under the form of few, small-sized, chromosomal segments.

Results: The shape of the nasal bone was studied using outline analysis combined with Fourier descriptors, and differential features were identified between IRCS BcG-66H and C57BL/6. An F2 cross between BcG-66H and C57BL/6 revealed that, out of the three SEG/Pas-derived chromosomal regions present in BcG-66H, two were involved. Segments on chromosomes 1 (∼32 Mb) and 18 (∼13 Mb) showed additive effect on nasal bone shape. The three chromosomal regions present in BcG-66H were isolated in congenic strains to study their individual effect. Epistatic interactions were assessed in bicongenic strains.

Conclusions: Our results show that, besides a strong individual effect, the QTL on chromosome 1 interacts with genes on chromosomes 13 and 18. This study demonstrates that nasal bone shape is under complex genetic control but can be efficiently dissected in the mouse using appropriate genetic tools and shape descriptors.

Figure 1. Dorsal view of the rostral part of the skull showing the nasal bone of 60±5 days old male mice.

A: C57BL/6 (arrows show the rostral depression); B: SEG (Mus spretus); C: IRCS strain 66H; D: caudal region of the nasal bone in C57BL/6 showing the interfrontal bone (arrow).

Figure 2. Genetic map of 66H IRCS indicating the position and sizes of the SEG-derived segments.

The segments of Mus spretus origin are displayed in solid while B6 segments are shaded. The 66H strain contains three SEG-derived segments on the Chromosome 1, 13 and 18.

Figure 3. Comparison of nasal bone shape in B6, 66H and their F1 hybrids by linear discriminant analysis (LDA) based on 15 principal components axes on a combination of Procrustes superimposition and elliptic Fourier descriptors (30 harmonics).

The first and second canonical axes were represented. The number of mice in each group is given in parentheses. Shapes drawn outside the graph describe nasal bone shape variations associated with low values (dashed line) or high values (solid line) along the axes. B6 and 66H fall into two well separated groups. F1 hybrids are distinct from either parent. Shape drawn outside the scatterplot, calculated with a multivariate regression, describes nasal bone shape variation along the canonical axes with low values (in dashed lines) and high values (in solid lines). No amplification and the nasal bone shape changes was effected.

Figure 4. Cumulative effect of Chr1 and Chr18 QTLs on nasal bone shape.

The projection on the canonical axis calculated as a LDA score (see text) is plotted against the number of SEG alleles inherited by the F2 progeny at both Chr1 and Chr18 QTLs. Parental strains are shown at extreme positions. Error bars represent s.e.m. The scores in F2 progeny decrease as a function of the number of SEG alleles and encompass the difference between 66H and B6 parental strains.

Figure 5. Comparison of nasal bone shape in B6, 66H and in Chr1, Chr13, and Chr18 congenic mice, by LDA based on 15 principal components axes on a combination of Procrustes superimposition and elliptic Fourier descriptors (30 harmonics).

The first two canonical axes are represented, totaling 80.67% of total variance. Chr1 congenics are close to 66H, while Chr13 congenics partially overlap with B6. The shape of Chr18 congenics is intermediate between that of the two parental strains.

Figure 6. Plots of the first and second canonical axes from LDA with elliptic Fourier descriptor on nasal bone shape to assess the mode of inheritance of the three chromosomal regions.

A: The first and second canonical axes for 66H, B6, Chr1 (Chr1 congenics), and Chr1H (heterozygotes for Chr1) are represented, totaling 95.5% of the total variation and showed the same phenotype between 66H and Chr1S whereas Chr1H exhibited a shape difference with B6 and Chr1S. Therefore Chr 1 QTL is semi-dominant. B: the first and second canonical axes of 66H, B6, Chr13 (Chr13 congenics), Chr13H (heterozygotes for Chr13) accounted for 92.4% of the total shape variation displayed no specific inheritance pattern. Neither heterozygotes nor homozygotes for Chr13 are different from B6 C: the first and second canonical axes of 66H, B6, Chr18 (Chr18 congenics), Chr18H (heterozygotes for Chr18) represented 95.4% of the total variance exhibited a similar intermediate shape between 66H and B6 for Chr18H and Chr18S. Therefore Chr18 QTL is dominantly inherited. For the explanation of the shape changes, see Figure 3. Shape changes were not amplified.

Figure 7. Evaluation of epistatic interactions between the three congenic fragments. Each graph represents B6, 66H, two congenic strains and the corresponding bicongenic strain.

A: The first and second canonical axes for 66H, B6, Chr1, Chr13, and Chr1+13 are represented, totaling 89.4% of the total variation and showed the same phenotype between Chr13 and Chr1+13 indicating that Chr13 segment decreases the effect of Chr1 QTL. B: The first and second canonical axes for 66H, B6, Chr13, Chr18 and Crh13+18 displayed 82.44% of the total variance and exhibited no differences in nasal bone shape between Chr 18 and Chr13+18. Therefore Chr13 QTL has no effect when combined with Chr18. C: The first and second canonical axes for 66H, B6, Chr1, Chr18 and Chr1+18 represented 81.7% of the nasal bone shape variation. Chr1+Chr18 bicongenic mice show a phenotype similar to that of the Chr1 congenic and 66H. For the explanation of the shape changes, see Figure 3. No amplification of the nasal bone shape was effected.