Fitness consequences of polymorphic inversions in the zebra finch genome

Abstract

Background: Inversion polymorphisms constitute an evolutionary puzzle: they should increase embryo mortality in heterokaryotypic individuals but still they are widespread in some taxa. Some insect species have evolved mechanisms to reduce the cost of embryo mortality but humans have not. In birds, a detailed analysis is missing although intraspecific inversion polymorphisms are regarded as common. In Australian zebra finches (Taeniopygia guttata), two polymorphic inversions are known cytogenetically and we set out to detect these two and potentially additional inversions using genomic tools and study their effects on embryo mortality and other fitness-related and morphological traits.

Results: Using whole-genome SNP data, we screened 948 wild zebra finches for polymorphic inversions and describe four large (12-63 Mb) intraspecific inversion polymorphisms with allele frequencies close to 50 %. Using additional data from 5229 birds and 9764 eggs from wild and three captive zebra finch populations, we show that only the largest inversions increase embryo mortality in heterokaryotypic males, with surprisingly small effect sizes. We test for a heterozygote advantage on other fitness components but find no evidence for heterosis for any of the inversions. Yet, we find strong additive effects on several morphological traits.

Conclusions: The mechanism that has carried the derived inversion haplotypes to such high allele frequencies remains elusive. It appears that selection has effectively minimized the costs associated with inversions in zebra finches. The highly skewed distribution of recombination events towards the chromosome ends in zebra finches and other estrildid species may function to minimize crossovers in the inverted regions.

Keywords: Embryo mortality; Heterosis; Inversion polymorphism; Miscarriage; Overdominance; Structural variant.

Fig. 1

Linkage disequilibrium (LD; left panel) and principal component analysis (PCA; right panel) results along chromosomes Tgu5 (a, b), Tgu11 (c, d), Tgu13 (e, f) and TguZ (g, h). In the right panels, the letters A, B, and C identify the combination of inversion types (alleles) that individuals (n = 948) carry, with A referring to the most frequent and C to the least frequent allele. Above the LD plots marker positions in Mb are given. PCA included all SNPs on the respective chromosome. Note that h includes a few (n = 18) females that were called as heterozygous for the inversion. These are carriers of occasional double crossovers

Fig. 2

a Pooled heterozygosity (ZH_p) in 50-kb windows along each chromosome in the zebra finch genome. b–e For the highlighted areas in a, which are the presumed inversion breakpoints on the autosomes and the entire inversion interior on the sex chromosome, the minor allele count frequency (MAC) spectra are shown for chromosome Tgu5 with a local maximum at 0.34–0.36 and a frequency of the minor (B) haplotype in the sample of 0.35 (b), Tgu11 with a local maximum at 0.48–0.50 and a frequency of minor (B) haplotype in the sample of 0.47 (c), Tgu13 with a local maximum at 0.48–0.50 and a frequency of minor (B) haplotype in the sample of 0.50 (d), and TguZ with two local maxima at 0.28–0.30 and 0.42–0.44 and a frequency of the B haplotype in the sample of 0.30 and frequency of the major (A) haplotype in the sample of 0.63 (e). f For comparison, the MAC of all remaining SNPs peaks at an allele frequency of around 0.1 because SNPs with a lower frequency were not unambiguously called

Fig. 3

Relationship between the size of an inversion (as percentage of the total chromosome size) and its effect on embryo mortality (meta-analytic summary of dominance effects across three captive populations (“Seewiesen”, “Bielefeld”, and “Cracow”). Shown are the odds ratios ± 1 standard error. An odds ratio >1 indicates an increased rate of embryo mortality in offspring produced by females or males that are heterozygous for one of the four inversions on chromosomes Tgu5, Tgu11, Tgu13, and TguZ. The effects on chromosome Tgu13 and TguZ for males both translate into a 4.5 % increase in embryo mortality rate. Only males can be heterozygous for chromosome TguZ. For visibility, values on the abscissa were moved 1 % up and down for females and males, respectively

Fig. 4

Dominance effects (±95 % confidence intervals) on different fitness parameters (RS = reproductive success) in three captive populations (S = “Seewiesen”, B = “Bielefeld”, C = “Cracow” and M = meta-analytic summary). Effect sizes are the factor level estimates of square-rooted and Z-transformed fitness components over inversion heterozygosity (while simultaneously fitting additive effects).The point sizes reflect log-transformed sample sizes

Fig. 5

Additive effects of the minor inversion allele ± 95 % confidence intervals on morphological phenotypes across three captive (white filled circles; S = “Seewiesen”, B = “Bielefeld”, C = “Cracow”) and two wild zebra finch populations (grey filled circles; Sy = “Sydney”, F = “Fowlers Gap”). M = meta-analytic summary (diamond symbol; yellow if significant after strict Bonferroni correction). Effect size estimates are regression slopes of Z-transformed phenotypes over inversion genotypes (while simultaneously fitting dominance effects) and show the effect of replacing one copy of allele A with allele B (or C in the rightmost panel). The point sizes reflect log-transformed sample sizes

Fig. 6

Summary of additive (left column) and dominance (right column) effect sizes from association studies between inversion genotypes and morphological traits (40 estimates = 8 phenotypes × 5 inversions; top row) and of the additive and dominance effect sizes from associations between inversion genotypes and fitness traits (20 [16] estimates = 4 fitness parameters × 5 inversions [minus 4 TguZ dominance effects in females]; bottom row). Empirical effect sizes are shown as the light grey bars overlaid with the null distribution as a black line. Effects that survived strict Bonferroni correction are highlighted in yellow. Power for a given effect size is overlaid in purple with its corresponding y-axis on the right. We estimated the null distribution (and the power values) by permuting the inversion genotypes within sexes (and adding/subtracting the corresponding effect sizes to/from the phenotypic values) and fitting the same mixed models as for the empirical data set (see “Methods” for details). For illustration, the null distribution was scaled to overlap the first bar in the histogram of the empirical estimates completely. Partial regression coefficients of additive and dominance effects are not directly comparable the way we standardized and fitted them and thus their null distributions differ (dominance effects reach higher values than additive effects because their variance is smaller; see also [114, 115])