Genetics of evolved load resistance in the skeletons of unusually large mice from Gough Island

Abstract

A primary function of the skeleton is to resist the loads imparted by body weight. Genetic analyses have identified genomic regions that contribute to differences in skeletal load resistance between laboratory strains of mice, but these studies are usually restricted to 1 or 2 bones and leave open the question of how load resistance evolves in natural populations. To address these challenges, we examined the genetics of bone structure using the largest wild house mice on record, which live on Gough Island (GI). We measured structural traits connected to load resistance in the femur, tibia, scapula, humerus, radius, ulna, and mandible of GI mice, a smaller-bodied reference strain from the mainland, and 760 of their F2s. GI mice have bone geometries indicative of greater load resistance abilities but show no increase in bone mineral density compared to the mainland strain. Across traits and bones, we identified a total of 153 quantitative trait loci (QTL) that span all but one of the autosomes. The breadth of QTL detection ranges from a single bone to all 7 bones. Additive effects of QTL are modest. QTL for bone structure show limited overlap with QTL for bone length and width and QTL for body weight mapped in the same cross, suggesting a distinct genetic architecture for load resistance. Our findings provide a rare genetic portrait of the evolution of load resistance in a natural population with extreme body size.

Keywords: bone geometry; island rule; load resistance; skeleton.

Fig. 1.

Structural measurements of bone. A micro-CT image from a cross-section of the femur of a GI mouse is shown. Cortical area is estimated as the summed areas of bone in cross-section (lighter shade). Lines indicate orientations for calculating maximum area moment of inertia (Imax) and minimum area moment of inertia (Imin). Polar moment (J) is computed as Imax + Imin.

Fig. 2.

Pairwise correlations between bone structural traits across F2s. Magnitudes of Pearson's correlations are indicated by color and size of circles. n = 760 F2s.

Fig. 3.

Relationship between principal component (PC) scores and 16-week body weight. Principal components were derived from the pairwise Pearson's correlation matrix of 18 postcranial traits (see Materials and methods).

Fig. 4.

Positions of QTL for bone structure. QTL positions are designated by vertical lines. 1.5 LOD intervals are denoted by horizontal lines.

Fig. 5.

Phenotypic effects of bone structure QTL. Additive effects divided by F2 phenotypic standard deviations are shown. Traits are distinguished by colors. Points within bones and chromosomes are randomly separated for visual clarity (depicted positions are unrelated to actual positions along chromosomes). Vertical lines display ± 1.96 SEs. Dashed line denotes 0. QTL with |d/a| > 1 are excluded.