(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Mapping mitonuclear epistasis using a novel recombinant yeast population [1]

['Tuc H. M. Nguyen', 'Department Of Biological Sciences', 'Binghamton University', 'Binghamton', 'New York', 'United States Of America', 'New York University', 'Austen Tinz-Burdick', 'Meghan Lenhardt', 'Margaret Geertz']

Date: 2023-04

Genetic variation in mitochondrial and nuclear genomes can perturb mitonuclear interactions and lead to phenotypic differences between individuals and populations. Despite their importance to most complex traits, it has been difficult to identify the interacting mitonuclear loci. Here, we present a novel advanced intercrossed population of Saccharomyces cerevisiae yeasts, called the Mitonuclear Recombinant Collection (MNRC), designed explicitly for detecting mitonuclear loci contributing to complex traits. For validation, we focused on mapping genes that contribute to the spontaneous loss of mitochondrial DNA (mtDNA) that leads to the petite phenotype in yeast. We found that rates of petite formation in natural populations are variable and influenced by genetic variation in nuclear DNA, mtDNA and mitonuclear interactions. We mapped nuclear and mitonuclear alleles contributing to mtDNA stability using the MNRC by integrating a term for mitonuclear epistasis into a genome-wide association model. We found that the associated mitonuclear loci play roles in mitotic growth most likely responding to retrograde signals from mitochondria, while the associated nuclear loci with main effects are involved in genome replication. We observed a positive correlation between growth rates and petite frequencies, suggesting a fitness tradeoff between mitotic growth and mtDNA stability. We also found that mtDNA stability was correlated with a mobile mitochondrial GC-cluster that is present in certain populations of yeast and that selection for nuclear alleles that stabilize mtDNA may be rapidly occurring. The MNRC provides a powerful tool for identifying mitonuclear interacting loci that will help us to better understand genotype-phenotype relationships and coevolutionary trajectories.

Mitochondrial functions require genes from nuclear and mitochondrial genomes that must work together. These interactions influence organismal fitness and coevolutionary processes yet it is difficult to identify the genes involved. Here, we created a novel collection of yeast designed explicitly for mapping mitonuclear genes. We used this collection to reveal genes influencing the maintenance of mitochondrial DNAs (mtDNAs), a trait important for human health. The mapping population presented here is an important new resource that will help to understand genotype-phenotype relationships and coevolutionary trajectories. Additionally, this work provides insight into mechanisms underlying mtDNA stability.

Funding: This research was supported by an NIH award (GM101320) to HLF, ACF and KC that provided salaries to THMN, ML, JFW, KC, ACF and HLF. MS was supported by an NSF REU (EEC 1757846). ATB, BB, MT, and BG received Binghamton University Undergraduate Research awards. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Copyright: © 2023 Nguyen et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Here, we show that mtDNA stability in wild S. cerevisiae yeasts is influenced by mitonuclear epistasis and by independent effects of both nuclear and mitochondrial genomes. We describe the construction of the Mitonuclear Recombinant Collection, and how we were able to use this to detect nuclear loci that participate in the mtDNA stability through main (independent) effects and mitonuclear interactions. We identified mitonuclear epistatic loci involved with mitotic growth, correlating rates of mitotic growth with mtDNA stability. This novel strategy and mitonuclear recombinant strain collection provide new tools for identifying mitonuclear loci that are important in nature.

To test our novel mapping strategy, we focused on mapping naturally-occurring alleles that contribute to the stability of the mitochondrial genome. Large scale deletions within mtDNAs are a common form of mtDNA instability, leading to mitochondrial heteroplasmies (defined as having more than one mitotype present) and deterioration of organismal health [ 55 – 59 ]. Saccharomyces yeasts do not tolerate mtDNA heteroplasmies and will fix for a single mitotype after just a few mitotic divisions [ 60 ]. Yeast cells fixed for these truncated mtDNAs can grow via fermentation and will form small, petite, colonies in comparison to the larger grande colonies formed by respiring cells, making it possible to quantify rates of mtDNA deletions [ 61 ]. Laboratory strains have accumulated multiple genetic variants leading to increased petite frequencies [ 62 ]. Mitochondrial research was heralded through curiosity of genetic causes of the petite phenotype in yeast [ 63 ]. Detailed mechanistic studies have, over decades, revealed a large number of genes contributing to mitochondrial respiration [[ 64 ] and refs therein], including those involved in replication and maintenance of mtDNAs [ 62 , 64 – 67 ]. This long list of genes is likely incomplete, given that studies typically focus on laboratory strains in singular, controlled environments. While these genes all interact with the mitochondria, it is not obvious which exhibit specific mitotype effects. Additionally, these studies are unable to provide insight into the potential for natural selection to be operating. We reasoned that natural genetic variation would lead to differences in petite frequencies among wild isolates of S. cerevisiae through mitonuclear epistasis and that our recombinant population would enable the identification of mitonuclear epistatic loci.

Previously, we showed that statistical estimates of mitonuclear epistasis explained over 30% of the phenotypic variances observed in a panel of S. cerevisiae yeasts consisting of 225 unique mitonuclear genotypes [ 15 ]. In the current study, our goal was to map naturally occurring alleles leading to mitonuclear epistasis in yeast populations. Detecting mitonuclear epistatic loci (or any g x g interaction) using association approaches can be challenging due to low allele frequencies, the large number of tests required to detect pairwise epistasis, dominance effects, and the potential for the environment to affect genetic interactions [ 37 , 50 , 53 , 54 ]. To overcome some of these challenges, we created a multiparent advanced intercrossed recombinant population of S. cerevisiae haploids designed explicitly to detect mitonuclear epistasis through association testing. Inheritance of mtDNA was controlled, resulting in a mapping collection where each nuclear genotype is paired with up to three different mitotypes. This should increase statistical power to detect mitonuclear interactions across the full mapping population, detect nuclear effects within a given mitotype and provide finer mapping resolution. Mitonuclear interactions could be integrated into phenotype-genotype association models allowing detection of loci contributing to complex traits.

Other genetic approaches can also be used to reveal mitonuclear epistatic loci. Chromosomal replacements in interspecific Saccharomyces yeast hybrids, followed by plasmid library screening, allowed the identification of mitonuclear incompatibilities between nuclear genes encoding intron splicing factors from one species and their mitochondrially encoded targets in the other [ 17 , 41 , 42 ]. Quantitative trait loci (QTL) mapping approaches using genotype-phenotype associations of recombinant progeny containing different mtDNAs have also been used to seek intraspecific mitonuclear incompatibilities [ 43 – 46 ]. Analysis of meiotic segregants following forced polyploidy enabled the detection of QTLs for interspecific mitonuclear incompatibilities contributing to sterility barriers in Saccharomyces [ 47 ]. Due to the low resolution of QTL mapping, specific loci were not identified, but in some cases, regions of mitonuclear genomic interest were implicated. An association study in wild yeast isolates identified mitochondrial variants of ATP6 that associated with sensitivity to the ATP synthase inhibitor, oligomycin [ 20 ]. When mtDNAs with these alleles were introduced into iso-nuclear heterozygous genetic backgrounds, oligomycin sensitivity was sometimes dependent on the nuclear background in addition to the ATP6 allele, suggesting mitonuclear epistasis. The interacting nuclear loci were not identified. (Interestingly, these isonuclear diploids also demonstrated non-respiratory phenotypes that were dependent on both mitotypes and nuclear genotypes, suggesting pervasive mitonuclear epistasis [ 20 ]). Other approaches to uncovering mitonuclear epistatic loci include differential expression analysis [ 5 , 48 – 50 ] and experimental evolution [ 51 , 52 ]. Because relatively few genetic backgrounds are used for most mapping approaches, even if single gene pairs are identified, it is not clear if these approaches will reveal a general picture of mitonuclear epistasis or lineage specific idiosyncrasies.

Selection for mitonuclear interactions may contribute to speciation [ 29 , 37 ]. Because of the large interest in uncovering speciation loci, strategies to identify mitonuclear epistasis often focus on analyzing mitonuclear incompatibilities in interspecific and inter-population hybrids or in mitonuclear hybrids where nuclear genomes from one population are paired with the mtDNAs from another. Sometimes candidate genes for these incompatibilities can be revealed through deductive reasoning. For example, in crosses between populations of Tigriopus californicus where mtDNA inheritance was controlled, F2 hybrids showed mitonuclear-specific OXPHOS enzyme activities and mtDNA copy number differences, prompting investigations into mitochondrially encoded electron transport proteins and the mtRNA polymerase [ 38 , 39 ]. In Drosophila, a mtDNA from D. simulans paired with a nuclear genome from D. melanogaster resulted in a mitonuclear genotype with impaired development and reproductive fitness [ 40 ]. Fortuitously for mapping purposes, the mtDNA sequences in this mitonuclear panel had very few polymorphisms, enabling the causative alleles behind this incompatibility (a mitochondrially encoded tRNA and a nuclear encoded tRNA synthetase) to be identified [ 36 ].

Mitonuclear epistasis, defined as the non-additive phenotypic effects of interacting mitochondrial and nuclear allele pairs, has been demonstrated across Eukarya, including vertebrates [ 4 – 6 ], invertebrates [ 7 – 12 ], plants [ 13 ] and fungi [ 14 – 21 ]. In humans, allelic variation in mitonuclear interactions contributes to human diseases [ 22 – 26 ]. Mitonuclear epistasis occurs within populations [ 27 , 28 ], between populations of the same species [ 15 , 29 – 35 ] and between closely related species [ 17 , 36 ], suggesting that physiologically-relevant mitonuclear epistasis is ubiquitous in natural populations. Mitonuclear loci are clearly influencing phenotypes that shape the structure of natural populations and are important for understanding evolutionary and coevolutionary processes, including speciation. Identifying the mitonuclear interactions that contribute to complex traits is challenging and largely a goal unmet in biology.

Interactions between mitochondrial and nuclear genomes are essential for the mitochondrial functions that power eukaryotic life. Mitonuclear interactions can be direct, as physical contacts between mitochondrial and nuclear genes and their products are needed for mitochondrial DNA (mtDNA) replication and maintenance and transcription, translation, assembly and function of mitochondrially encoded components of oxidative phosphorylation (OXPHOS) complexes [ 1 ]. Mitonuclear interactions can also be indirect, though anterograde (nucleus-to-mitochondria) and retrograde (mitochondria-to-nucleus) signaling where metabolites, biochemicals or RNAs direct gene expression in response to metabolic needs or environmental stressors [ 2 , 3 ]. Genetic variation can alter the efficiencies of these interactions leading to phenotypic differences between individuals and populations.

Results

Creation of a Mitonuclear Recombinant Collection for association studies Mitonuclear interactions explain a significant proportion of phenotypic variances in S. cerevisiae yeasts and involve numerous, as yet unmapped, loci [14,15,77]. We sought a general genome-wide mapping approach that would facilitate the mapping of both the nuclear and mitonuclear loci underlying complex traits such as mtDNA stability. We created a multiparent recombinant collection of S. cerevisiae strains specifically designed for association mapping of nuclear and mitonuclear loci (called the Mitonuclear Recombinant Collection, or MNRC) (Fig 4). To do this, we first replaced the mtDNAs in 25 wild divergent yeast isolates such that each contained an identical mitotype, and then mated them to create each possible heterozygous diploid. The diploids were sporulated and ~10,000 haploid F1 haploid progeny were isolated and then randomly mated. F1 diploids were then isolated and sporulated. Only one laboratory strain (W303, derived from the reference strain) was included as a parent strain. We noted that very few progeny after one round of meiosis contained the selectable markers found in this strain, suggesting that this genotype was quickly lost from the collection. The process was repeated for a total of 7 rounds of meiosis. A collection of 181 F7 haploids (named Recombinant Collection (RC) 1 or RC1) was isolated and fully sequenced. The mtDNAs from RC1 strains were removed (creating RCρ0) and replaced with two additional mtDNAs via karyogamy-deficient matings, creating populations RC2 and RC3. The GC-cluster content of the mtDNAs in the MNRC are classified as low (117 clusters in RC2), medium (137 clusters in RC1), or high (203 clusters in RC3). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. A Mitonuclear Recombinant Collection designed for association mapping. 25 unique genetic backgrounds of S. cerevisiae fixed for a single mitotype were systematically crossed to create each possible F1 heterozygous diploid with identical mtDNAs (284 total). The numbers of unique genotypes from each parental ancestral background are indicated; blue (Wine/European); yellow (Malaysian); green (North American); orange (West African); black (mosaic). The F1 diploids were sporulated and F1 recombinant haploid progeny isolated. Following 6 rounds of random matings, haploid F7 progeny were isolated to create RC1. The mtDNAs from RC1 were removed (RC1ρ0) and replaced with 2 different mtDNAs, creating RC2 and RC3. The mtDNAs in RC1,2, and 3 are from the wild isolates 273614N, YPS606, and NCYC110, respectively. https://doi.org/10.1371/journal.pgen.1010401.g004 To generate SNPs tables for association testing, RC1 strains were sequenced to ~40x coverage. Paired-end reads were aligned to the S. cerevisiae reference genome and the locations of nuclear SNPs and small indels were extracted from each alignment. Polymorphic sites were filtered by removing telomeric regions and SNPs/indels with low allele frequencies (MAF <5%). Following filtering, 24,955 biallelic sites across the 16 yeast chromosomes with an average of ~2200 SNPs/chromosome were available for association testing. Chromosomal polymorphic data are summarized in S6 Table. Our read alignments and subsequent analyses did not account for chromosomal rearrangements, such as copy number variants, translocations and inversions that would map to similar locations of the reference genome nor genomic regions absent in the reference strain. We validated that this novel recombinant population could be used for simple association studies. Strains from RC1 were phenotyped for growth on copper sulfate and an association test was performed to identify SNP variants associated with copper tolerance. A single peak on Chr. 8 coincided with a region containing CUP1, the copper binding metallothionein (S2 Fig). Variation at this locus is known to lead to high copper tolerances found in Wine/European isolates [78] and has previously been identified through association studies using wild isolates [68,79]. Thus, the recombinant collection produced here is successful for association studies despite a relatively low number of parental strains.

Nuclear and mitonuclear associations for mtDNA stability To map nuclear and mitonuclear associations, petite frequencies were collected for each strain in RC1, RC2 and RC3. In RC1, the petite frequencies ranged from 0.0% to 27.7% forming a continuum, as would be expected for a complex trait involving numerous loci (Fig 5A). The same rank orderings were not observed in RC2 or RC3, revealing the influences of mitonuclear interactions. RC3, containing the GC-cluster-rich mtDNA, had a higher average petite frequency than RC1 or RC2 (S3 Fig). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 5. Nuclear SNPs associate with mtDNA stability through main effects and mitonuclear interactions. A. Petite frequency is a quantitative trait influenced by natural genetic variation. Petite frequencies for strains from RC1 (gray) were rank ordered from lowest to highest. The petite frequencies for strains from RC2 (green) and RC3 (gold) are plotted next to isonuclear strains from RC1. B. Manhattan plots show main effect nuclear SNPs associated with petite frequencies in ways that were independent of mtDNA and C. nuclear SNPs whose association is dependent on mitotype (ie. mitonuclear). The different plot profiles indicate that the main effect nuclear SNPs are different that those involved with mitonuclear interactions. FDR thresholds at 0.1% (P < 4.1 × 10−5) for nuclear associations and 5.0% (P < 1.2 ×10−5) for mitonuclear associations and a conservative Bonferroni threshold (P < 2.0 × 10−6) are shown. https://doi.org/10.1371/journal.pgen.1010401.g005 In theory, the recombinant genomes and fixed mitotypes of the RC strains should reduce effects of population structure, improve statistical power while using a smaller number of samples, limit false positives, and control for mitonuclear interactions. We performed association testing to identify nuclear loci that had both a main effect and interacted with the mtDNA to influence mtDNA stability. Mating types, auxotrophic markers and residual population structure as determined by principal component analyses were included as covariates (see METHODS). The significance profiles of associations for nuclear variants that were independent (nuclear SNP, Fig 5B) and dependent on mitotype (nuclear SNP× mtDNA, Fig 5C) were different, providing confidence that the association model is able to detect nuclear features that are unique to either main or epistatic effects. Nuclear SNPs whose effects were independent of mitotype resulted in stronger associations than mtDNA-dependent alleles. This is not surprising given that independent contributions of nuclear genotypes influence growth phenotypes to a greater extent than mitonuclear interactions [14,15,77]. At a false discovery rate (FDR) < 0.1% (Q-value = 0.001), we observed 130 mtDNA-independent associated SNPs located within or 250 bp upstream of coding sequences (S7 Table). In comparison, at FDR <5% (Q-value = 0.05), there were 3 mtDNA-dependent associated SNPs (S8 Table). Strong nuclear effects could mask mitonuclear interactions. We identified the alleles with the strongest effects by calculating the effect size of each SNP with mitotype-independent associations as the difference between the average petite frequencies of each allele weighted by its frequency in the recombinant collections (S9 Table). This revealed that the highest effect size was attributed to a SNP on Chr. 15 predicting a previously uncharacterized G50D missense mutation in MIP1, the mitochondrial DNA polymerase required for replication and maintenance of mtDNA. To improve power of detecting mitonuclear associations, we repeated the analysis including the MIP1 SNPs as covariates. This removed the mitotype-independent associations on Chr. 15 and increased the numbers of significant mitonuclear associations from 3 to 27 without changing the overall association profiles (S4 Fig and S8 Table).

[END]
---
[1] Url: https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1010401

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.