Mitonuclear compatibility is maintained despite relaxed selection on male mitochondrial DNA in bivalves with doubly uniparental inheritance

Abstract

Mitonuclear coevolution is common in eukaryotes, but bivalve lineages that have doubly uniparental inheritance (DUI) of mitochondria may be an interesting example. In this system, females transmit mtDNA (F mtDNA) to all offspring, while males transmit a different mtDNA (M mtDNA) solely to their sons. Molecular evolution and functional data suggest oxidative phosphorylation (OXPHOS) genes encoded in M mtDNA evolve under relaxed selection due to their function being limited to sperm only (vs. all other tissues for F mtDNA). This has led to the hypothesis that mitonuclear coevolution is less important for M mtDNA. Here, we use comparative phylogenetics, transcriptomics, and proteomics to understand mitonuclear interactions in DUI bivalves. We found nuclear OXPHOS proteins coevolve and maintain compatibility similarly with both F and M mtDNA OXPHOS proteins. Mitochondrial recombination did not influence mitonuclear compatibility and nuclear-encoded OXPHOS genes were not upregulated in tissues with M mtDNA to offset dysfunction. Our results support that selection maintains mitonuclear compatibility with F and M mtDNA despite relaxed selection on M mtDNA. Strict sperm transmission, lower effective population size, and higher mutation rates may explain the evolution of M mtDNA. Our study highlights that mitonuclear coevolution and compatibility may be broad features of eukaryotes.

Keywords: bivalvia; evolutionary rate covariation; molecular evolution; phylogenetics; proteomics; transcriptomics.

Figure 1.

Summary of mitonuclear evolutionary rate covariations (ERCs) based on root-to-tip branch lengths. (A) Points and trendlines for ERCs between mitochondrial proteins and nuclear OXPHOS proteins or nuclear BUSCO proteins using either female (left) or male (right) mtDNA. Shading represents the 95% confidence interval of regression fits. (B) Distribution of ERC strength (r_s) for female (top) or male (bottom) mitochondrial with nuclear OXPHOS or BUSCO proteins. Distributions are based on 10,000 bootstrap replicates. Black dots represent the mean r_s and lines represent the 95% confidence interval.

Figure 2.

Summary of protein-protein interactions estimated in this study. Individual cells represent support for an interaction between female (top) or male (bottom) mtDNA and nuclear OXPHOS proteins. Coloration is scaled based on acceptable support based on developers’ recommendations (0.23), corresponding to supported and unsupported interactions. Gray cells represent interactions where nuclear proteins could not be identified, or model generation failed. Interactions are grouped by complex: Complex 1 (CI), Complex 3 (CIII), and Complex 4 (CIV). Presented phylogenetic reconstructions are based on female (top) and male (bottom) mtDNA OXPHOS proteins.

Figure 3.

Summary for support for female and male mitochondrial OXPHOS protein interactions with nuclear OXPHOS proteins. (A) Points and trendline for female–male protein–protein interactions for all doubly uniparental inheritance (DUI) bivalves and (B) between DUI bivalves without recent mitochondrial DNA recombination and those with recent recombination. In both A and B, colored points represent a pairwise interaction, colored lines are linear regression lines, gray coloring represents the 95% confidence interval of regression fits, and black lines represent a 1:1 relationship. (C) Violin and box plots summarizing female (left) and male (right) OXPHOS protein interactions with nuclear OXPHOS proteins. (D) Violin and box plots summarizing male OXPHOS protein interactions with nuclear OXPHOS proteins in lineages without recent mitochondrial DNA recombination (left) and those with recent recombination (right).

Figure 4.

Summary of nuclear OXPHOS gene expression comparisons among female and male gonadal tissue across bivalves. Top panel shows female–male linear regression lines based on gene expression in gonadal tissue. Read counts from female (y-axis) and male (x-axis) gonadal tissue samples are reported in log-scale. Lines correspond to three groups: (1) non-recombining (Unionida), (2) recombining (Ruditapes philliparum), and (3) strictly maternal inheritance (SMI; Magallana gigas and Mizuhopecten yessoensis). Gray coloring represents the 95% confidence interval of regression fits. The black line represents a 1:1 relationship. The bottom panel includes violin and box plots summarizing female and male nuclear OXPHOS gene expression for all DUI (Unionida and Ruditapes philliparum; left) and SMI (right) species.