The genetic architecture of parallel armor plate reduction in threespine sticklebacks

Abstract

How many genetic changes control the evolution of new traits in natural populations? Are the same genetic changes seen in cases of parallel evolution? Despite long-standing interest in these questions, they have been difficult to address, particularly in vertebrates. We have analyzed the genetic basis of natural variation in three different aspects of the skeletal armor of threespine sticklebacks (Gasterosteus aculeatus): the pattern, number, and size of the bony lateral plates. A few chromosomal regions can account for variation in all three aspects of the lateral plates, with one major locus contributing to most of the variation in lateral plate pattern and number. Genetic mapping and allelic complementation experiments show that the same major locus is responsible for the parallel evolution of armor plate reduction in two widely separated populations. These results suggest that a small number of genetic changes can produce major skeletal alterations in natural populations and that the same major locus is used repeatedly when similar traits evolve in different locations.

Figure 1. Mapping the Genetic Basis of Lateral Plate Reduction in Different Natural Populations of Threespine Sticklebacks

Crossing a completely plated Japanese marine fish with a low-plated fish from Paxton Lake, British Columbia, produced a mixture of complete, partial, and low morph phenotypes in F2 progeny animals (Cross 1). In contrast, crossing a completely plated fish and a low-plated fish from an inland freshwater stream in Friant, California, produced only complete and low-plated progeny (Cross 2). Red dots show the geographic origins of the populations studied. Scale bars equal 1 cm. AA, Aa, and aa refer to genotypes at Gac4174 (a microsatellite marker) near the major plate locus on LG 4. The genotype at Gac4174 is missing in ten of the 360 F2s in Cross 1. All fish were stained with alizarin red to reveal bony structures.

Figure 2. Comparison of QTL Positions for Different Traits

LOD scores are shown as a function of genetic distance along different stickleback linkage groups. QTL affecting qualitative plate pattern (red line), total plate number (black lines), or plate size (blue lines) show similar shapes on several linkage groups, suggesting that the same or linked genes control multiple aspects of plate phenotype. Points in LOD plots correspond to the following microsatellite markers from left to right along each linkage group: (A) LG 4: Pitx2 (Stn220), Stn38, Gac62, Stn42, Gac4174, Stn45, Stn183, Stn46, Stn47, Stn184, Stn39; (B) LG 7: Stn70, Stn72, Stn76, Stn71, Stn78, Stn79, Stn75, Stn81, Stn80 Stn82, Pitx1; (C) LG 10: Stn119, Stn120, Stn211, Stn121, Stn124, Stn23, Stn125; (D) LG 25: Stn212, Stn213, Stn214, Stn215, Stn216, Gac1125, Stn217; (E) LG 26: Stn218, Stn219, Bmp6, Stn222, Stn223. Note that markers Stn183 and Stn184 from LG 18 in the Priest Lake cross (Peichel et al. 2001) map together with LG 4 markers in the larger Cross 1.

Figure 3. Cumulative Effects of Freshwater Alleles on the Number, Pattern, and Size of Lateral Plates in Cross 1

Increasing the total number of Paxton benthic freshwater alleles at modifier QTL on LGs 7, 10, and 26 significantly reduces plate number in animals with one marine (complete morph) and one Paxton benthic (low morph) allele near the major QTL on LG 4 (Aa progeny) (A). The same modifier QTL have little effect on fish with two marine alleles near the major QTL (AA animals) (B) and smaller phenotypic effects on animals with two benthic alleles near the major QTL (aa animals) (C). Increasing the number of benthic alleles also significantly increases the proportion of Aa fish whose overall plate pattern is classified as partial instead of complete (D). (E–F) show plate size effects. Increasing the number of benthic alleles at plate size QTL on LGs 4, 7, and 25 significantly reduces mean plate width of F2 progeny (E). (F) shows the schema of plate size measurements. Lateral plates are shown numbered from anterior to posterior. Error bars in (A–E) represent standard error.