Mitonuclear interactions shape both direct and parental effects of diet on fitness and involve a SNP in mitoribosomal 16s rRNA

Adam J Dobson; Susanne Voigt; Luisa Kumpitsch; Lucas Langer; Emmely Voigt; Rita Ibrahim; Damian K Dowling; Klaus Reinhardt

doi:10.1371/journal.pbio.3002218

. 2023 Aug 21;21(8):e3002218. doi: 10.1371/journal.pbio.3002218

Mitonuclear interactions shape both direct and parental effects of diet on fitness and involve a SNP in mitoribosomal 16s rRNA

Adam J Dobson ^1,^2,^*, Susanne Voigt ², Luisa Kumpitsch ², Lucas Langer ², Emmely Voigt ², Rita Ibrahim ¹, Damian K Dowling ³, Klaus Reinhardt ^2,^*

Editor: Nick Lane⁴

PMCID: PMC10441796 PMID: 37603597

Abstract

Nutrition is a primary determinant of health, but responses to nutrition vary with genotype. Epistasis between mitochondrial and nuclear genomes may cause some of this variation, but which mitochondrial loci and nutrients participate in complex gene-by-gene-by-diet interactions? Furthermore, it remains unknown whether mitonuclear epistasis is involved only in the immediate responses to changes in diet, or whether mitonuclear genotype might modulate sensitivity to variation in parental nutrition, to shape intergenerational fitness responses. Here, in Drosophila melanogaster, we show that mitonuclear epistasis shapes fitness responses to variation in dietary lipids and amino acids. We also show that mitonuclear genotype modulates the parental effect of dietary lipid and amino acid variation on offspring fitness. Effect sizes for the interactions between diet, mitogenotype, and nucleogenotype were equal to or greater than the main effect of diet for some traits, suggesting that dietary impacts cannot be understood without first accounting for these interactions. Associating phenotype to mtDNA variation in a subset of populations implicated a C/T polymorphism in mt:lrRNA, which encodes the 16S rRNA of the mitochondrial ribosome. This association suggests that directionally different responses to dietary changes can result from variants on mtDNA that do not change protein coding sequence, dependent on epistatic interactions with variation in the nuclear genome.

Why do genetically distinct individuals show differing responses to dietary change? This study investigates how genetic differences in mitochondria and the nucleus interact to determine fitness response to direct manipulation of diet, finding that the effects depend on essential amino acids and lipids, and a SNP in the mitochondrial 16S rRNA.

Introduction

Nutrition and genotype underpin variation in health and biological fitness. They can also interact, resulting in different responses among genotypes to the same nutritional changes [1,2]. In humans, there is interest in leveraging this variation to optimize nutrition by personalizing diet to individual consumers’ needs [3]. To realize this ambition, we must understand the genetic drivers of variation in response to nutrition, but this is challenging because independently segregating loci have nonadditive, epistatic interactions [4], which may modulate responses to nutrition, i.e., diet-by-genotype-by-genotype variation [5]. Consequently, the genetic loci involved in these responses remain elusive.

Mitochondria are critical metabolic hubs, with their own small genome, and variation in their function can contribute to variation in dietary optima. The mitochondrial genome segregates independently of the nuclear genome, and the combination of mitochondrial and nuclear variants can generate “mitonuclear” epistasis [6,7]. This epistasis is thought to occur because mtDNA is transcribed, processed, and translated by nuclear-encoded proteins, and mtDNA-encoded proteins function in pathways and complexes that include nuclear-encoded proteins [8]. Reciprocally, outputs of genetic variation in the nucleus depend on how mitochondrial metabolites and signals feed into broader cellular networks. Mitonuclear epistasis has been reported for numerous traits and processes [9–11], but diet-by-mito-by-nuclear (DMN) interactions are less well characterized, despite evidence for mitochondrial modulation of nutrient signaling [12]. These interactions could be critical determinants of individual response to diet and, therefore, health. So far, DMN interactions have been shown for development time, life span, fecundity, and gene expression in Drosophila melanogaster [5,13–16]. This study addresses 4 main questions about DMN interactions, also using D. melanogaster: (A) How much phenotypic variation do DMN interactions cause, relative to lower-order interactions (i.e., nuclear-diet, mitochondria-diet, mitochondria-nuclear) and main effects (i.e., diet, nuclear, mitochondria)? (B) Parental nutrition can modulate offspring fitness, independent of offspring diet [17]—is this impact of nutritional variation modulated by mitonuclear variation? (C) Which specific dietary nutrients cause DMN variation? And, perhaps most importantly, (D) which mitochondrial polymorphisms underpin DMN interactions?

Here, in Drosophila, we study how variation among mitochondrial genotypes (mitogenotypes) modulates reproductive response to specific nutrients, in distinct populations of nuclear genotypes (nucleogenotypes). We identify variants in these genomes to characterise each population’s specific combination of mitochondrial and nuclear variation (mitonucleogenotypes). We study reproductive traits because of their relevance to biological fitness, expanding on preceding studies [5,13–16] through multidimensional analysis of reproductive phenotype, and manipulating specific dietary nutrients (essential amino acids and lipid). We show that diets expected to promote fitness can in fact be lethal to specific mitonucleogenotypes. We also show that effects of parental nutrition on offspring performance are mitonucleogenotype specific. Effect sizes of DMN interactions were large for some traits, even exceeding those for diet:mitogenotype or diet:nucleogenotype interactions, implicating mitonuclear epistasis as a more important determinant of the response to nutrition than variation in either genome alone, and showing that DMN variation can be a major source of phenotypic variation. Importantly, we observe DMN interactions among a subset of populations differentiated only by an mtDNA polymorphism in a nonprotein-coding gene, long ribosomal RNA (mt:lrRNA), which encodes the mitoribosomal 16S rRNA. This gene is a structural component of the mitochondrial ribosome (mitoribosome) implicating a nonprotein-coding mitochondrial gene with roles in protein translation in DMN effects. Altogether, these results suggest that mitonuclear epistasis can be a leading determinant of optimal diet, that this variation maps to variants on mtDNA that do not change protein coding sequence, and that the consequences can be a matter of life or death.

Results

Establishing and sequencing a panel of diverse mitonucleogenotypes

Mitochondria are inherited exclusively from mothers through eggs. Mitonucleogenotype can therefore be manipulated by backcrossing virgin females of a given mitogenotype to males possessing a target nucleogenotype of interest and then iteratively crossing daughters produced by this cross to males with the target nucleogenotype across successive generations. This procedure is expected to dilute and eventually purge the F0 mother’s nucleogenotype in the mitonucleogenotype lineages produced, substituting it with the paternal nucleogenotype while retaining the F0 mother’s mitogenotype. We used this approach to produce D. melanogaster populations with varied mitonucleogenotypes (Fig 1A), comprising replicated and fully factorial combinations of mitochondrial and nuclear genomes from Australia, Benin, and Canada (A, B, and C, respectively). With 45 females mating to 45 males in each iteration, the crossing scheme was designed to produce distinct mitochondrial backgrounds bearing equivalent pools of standing nuclear variation, by introgressing populations either reciprocally or to themselves. The use in F0 of multiple females from outbred ancestral populations maintained the potential for multiple mtDNA haplotypes to segregate in each of the introgressed populations. For brevity, we abbreviated population names, giving mitochondrial and then nuclear origin (e.g., AB = Australian mitochondria, Beninese nuclei). Each combination was triplicated (e.g., AB₁, AB₂, AB₃) at the beginning of the introgression, with triplicates maintained in parallel for more than 160 introgressions, altogether generating 27 populations, comprising 9 triplicated mitonucleogenotypes (Fig 1A) [18].

After >100 introgressions (S1 Table), assuming no mitochondrial incompatibility, we expected maternal nucleogenotypes to be purged by introgression and that nuclear variation among populations with co-originating nuclear backgrounds would be indistinguishable, regardless of mitogenotype. For a proof of principle, single nucleotide polymorphisms (SNPs) in the nuclear genome were identified by Pool-seq, sampling 2 populations per mito-nuclear pairing (i.e., 18/27 total). Within each nucleogenotype, principal components analysis (PCA) suggested negligible differentiation by mitogenotype, indeed samples with each nucleogenotype sat on top of one another in an ordination plot (Fig 1B), revealing 3 nucleogenotypes with no visible variation, on axes that explained 93% total variance (S1A Fig). For an orthogonal analysis of the grouping of nuclear genomes, we conducted continuous structure (“conStruct”) analysis [19]. conStruct analysis produces a statistical model of population genetic structure, inferring similarity among a set of discrete genomes [19]. The method assigns genetic variation to a set of K possible user-specified states (“layers”) in the model. Each sample is then depicted as a contribution from each layer, aka proportion admixture of the different hypothetical layers. We specified K = 3 possible layers to reflect 3 founder populations at the beginning of the introgression. The conStruct analysis complemented the PCA, assigning co-originating nuclear genomes almost entirely to the same layers (Fig 1C), and each individual population was assigned between 94.4% and 99.9% to its respective dominant layer (S2 Table). Together, the PCA and conStruct analyses indicated high degrees of similarity between co-originating nucleogenotypes, and differentiation between nucleogenotypes of different geographic origin.

We then examined among-mitogenotype population structure, reanalyzing previously reported mitochondrial Pool-seq data [18]. PCA (Fig 1D) revealed consistent within-mitogenotype clustering, except population AA₃, which was distinct from other A mitogenotypes, and intermediate between the B and C mitogenotypes on the first PC. Little differentiation was apparent between mitogenotypes B and C on the first PC. On the second PC, C mitogenotypes were equivalent to A. The majority of mitogenotype variance was explained by these 2 PCs (S1B Fig). The AA₃ mitogenotype was strikingly intermediate between the 2 major clusters of other mitogenotypes; however, this was not wholly surprising, given that (A) the ancestral population originated from the middle of a cline on which 2 main mitochondrial haplotypes have been reported [20] and that (B) D. melanogaster settled relatively recently in Australia, likely founded by both European and African lineages [21], which may have introduced haplotypes with the intermediate genotype we observed.

We also conducted conStruct analysis on the mitochondrial genomes—although we interpret results conservatively because mtDNA SNPs are unlikely to segregate independently (due to the presence on this diminutive genome of only a few SNPs, in high linkage disequilibrium (LD) because of lack of recombination). Despite these caveats, the conStruct analysis complemented the PCA analysis, with populations CA₁ and CA₃ dominated by admixture with the Beninese mitogenotypes (Fig 1E) and population AA₃ appearing again as an outlier, intermediate between other mitogenotypes (Fig 1E). These findings suggested that phenotyping both B and C mitochondria in the same experiment would likely prove redundant, and so we decided to eliminate one of these mitogenotypes from the study, reducing the number of populations to be phenotyped, for experimental tractability. We retained mitochondria with Beninese origins because of this background’s widespread use in fly research but excluded mitochondria with Canadian origins. We also excluded Canadian nucleogenotypes (AC, BC) because of lack of coevolutionary history with A or B mitogenotypes. This reduced number of populations for phenotyping from 27 to 12 (i.e., studying only AA_1–3, AB_1–3, BA_1–3, and BB_1–3). To confirm that this reduced panel contained DMN variation necessary for subsequent phenotyping, we conducted a preliminary phenotypic analysis, focusing on fecundity (i.e., egg laying/24 hours). We chose fecundity because of the relevance of reproductive traits to Darwinian fitness. We applied both an established dietary manipulation that promotes fecundity by enriching essential amino acids (EAAs) [22], and a novel manipulation that represses fecundity (S2A Fig), by enriching plant-based lipids (Text A in S1 Text). Feeding these EAA-enriched and lipid-enriched diets to the focal panel of 12 populations (S2B Fig and Text B in S1 Text) revealed mitonuclear variation in fecundity response (S2C Fig, Text B in S1 Text). This motivated further study of the specific SNPs that differentiated these populations, and how these SNPs predicted variation in a comprehensive analysis of reproductive phenotype.

Among the focal Australian and Beninese populations, we characterized mtDNA polymorphisms in detail to identify genetic information that could be used to model phenotypic responses to diet. We tested for SNPs at significantly different frequencies (Fisher’s exact test, FDR < 0.001), finding 28 altogether (S3 Table), which were predominantly biallelic (S3 Fig). Positions of these SNPs on a map of the mitochondrial genome are shown in S4 Fig. Only one SNP was significantly differentiated between mitogenotypes B and C, which further validated our decision to exclude Canadian populations from our analysis. By contrast, 27 of the 28 SNPs were significantly differentiated between mitogenotypes A and B (S3 Table), i.e., the majority of mitochondrial diversity in the full panel was represented by these 2 mitogenotypes. Even though these alleles could still potentially segregate, major alleles for the majority of SNPs (70%) were fixed, 89% of SNPs were at a frequency ≥0.99, and the lowest observed frequency for a SNP (position 17,255) was nevertheless still high at 0.8 (S3 Fig and S3 Table). Thus, allele frequency differed significantly among populations, but major alleles were at high frequency within each population.

Since the populations were not isogenic, and the backcrossing regime was designed to permit segregating variation in each genome, we expected mtDNA variation might segregate within individual populations, and so allele frequencies could potentially drift. We therefore resequenced mtDNAs at a 2-year interval, but the maintenance of allele frequency over these 2 years suggested that any drift was negligible (S5 Table) and, therefore, that the populations were a legitimate resource for genetic association studies.

We examined the known functions of mtDNA loci bearing SNPs. Of the 27 SNPs, 19 were in protein-coding regions, in genes encoding subunits of electron transport chain complexes and ATPase subunit 6. Only 2/19 were predicted to be nonsynonymous (S3 Table). Interestingly, one nonsynonymous C/A SNP, at mtDNA position 9,065, coding for a valine/leucine substitution in ND4, has previously been characterized, including enhanced sensitivity to a high-protein diet [23]. In our populations, it cosegregated with variation at multiple other positions and so cannot be characterized further here. We are not aware of previous reports of the other nonsynonymous variant, an A/T SNP at mtDNA 4,616, predicted to encode a methionine/isoleucine substitution in ATPase subunit 6. The remaining 17/19 SNPs in protein-coding sequence were predicted to be synonymous. The other 8/27 SNPs were in nonprotein-coding regions (in the origin of replication, a tRNA, and the mitoribosomal 16S rRNA, lrRNA) (S3 Table). Thus, altogether, the populations were differentiated by 27 mtDNA SNPs, 25 of which were predicted to not affect protein amino acid sequences.

We explored how these SNPs were distributed among the populations to identify groups for subsequent genetic associations. Pool-seq analysis identifies alleles, but we emphasize the potential by experimental design for variation to segregate in our populations and that we did not sequence haplotypes: Alleles of distinct SNPs may theoretically segregate independently of one another among our populations. We therefore studied the differentiation of populations by major allele cosegregation using a network analysis (Fig 1F). This reinforced findings of our previous analysis of the full set of populations (i.e., including Canadian mitogenotypes) [18]. The network revealed a punctuated continuum of among-population variation, independent of nucleogenotype. Some populations had unique mitogenotypes (AB₃, AA₃). Others had identical mitogenotypes even in the presence of distinct nucleogenotypes (AA₁, AA₂, AB₁, AB₂; and BA₂, BA₃, BB₁, BB₃; and BA₁, BB₂). Australian mitogenotypes and Beninese mitogenotypes were largely dichotomous, except that population AA₃ was notably intermediate (Fig 1F), consistent with the preceding PCA and admixture analyses. Thus, the mitonuclear populations could be grouped by frequencies of major alleles on mtDNA, revealing 4 groups of mitogenotypes.

How did mtDNA alleles intersect with nucleogenotypes—what mitonucleogenotypes were present in the panel? We clustered the populations by major alleles on mtDNA, using hierarchical clustering, and examined the intersection with nucleogenotype. Because our PCA and admixture analyses of nDNA suggested that co-originating nucleogenotypes were homogenized and shaped by introgression (i.e., not incompatibility with mtDNA) in this particular panel, we viewed nucleogenotype as a dichotomous factor (A or B) for this analysis. The clustering separated A mitogenotypes from B (Fig 1G). However, within this geographic differentiation, the sequencing revealed more granular among-population differentiation, with 5 distinct clusters of unique mitogenotypes. These mitogenotypes were not nested within nucleogenotype, and some co-occurred with both A and B nucleogenotypes, which indicated that mitonuclear incompatibilities were not at play during the introgression process (consistent with PCA and admixture). To generate a final, sequence-informed mitonucleogenotype assignment, for genetic associations, we concatenated sequenced-based mitogenotype with nucleogenotype (i.e., A or B, since our sequencing data suggested nuclear homogenization independent of mitogenotype (Fig 1)). This revealed 8 distinct mitonucleogenotypes (Fig 1G). We then phenotyped the responses of each mitonucleogenotype to diet and examined how traits were shaped by mitonucleogenotype.

We noticed that a subset of populations (mitonucleogenotypes 5, 6, 7, and 8) (Fig 1G) were distinguished by only one mitochondrial polymorphism, suggesting that any mitonuclear or DMN variation in this subset was attributable to this SNP. This contrasted other populations, which bore confounding variation at other positions, and so effects could not be attributed to a single locus. The specific SNP was a C/T polymorphism in mt:lrRNA (mtDNA position 13934), which occurred at high frequencies (between 0.99 and 1; S3 Table) in each nuclear background. mt:lrRNA encodes the 16S RNA of the mitochondrial ribosome, which seemed like a good candidate to mediate mitonuclear effects, because (1) the proteins that this RNA forms complexes with in the mitochondrial ribosome (mitoribosome) are encoded by the nuclear genome; (2) the mitochondrial ribosome translates mitochondrial proteins, so variation in its function has the potential to generate a bioenergetic bottleneck, with metabolic consequences for the rest of the cell, and consequences for penetrance of nucleogenotype variation; and (3) preceding work found a role for a tRNA in mitonuclear effects [24]. We hypothesized that the C/T polymorphism in mt:lrRNA may provide an illustrative example of how nonprotein-coding variation in the mitochondrial genome could underpin DMN variation in phenotype. We therefore decided to investigate the effects of this SNP specifically in subsequent genetic mapping, alongside analyses of the full panel of populations.

Phenotyping

Encouraged by initial fecundity results (S2 Text), we characterized a more extensive panel of fitness traits, examining how they responded to dietary variation and how those responses associated with mitonucleogenotype. Using a total of >25,000 individual flies, we assayed fecundity, fertility, and development time—traits that have previously been shown to be sensitive to DMN variation [13,16]—as well as number of adult progeny as a direct fitness measure. As with our preliminary investigation of fecundity, we fed flies either a control medium, an EAA-enriched medium that promotes egg laying, or a lipid-enriched medium that reduces egg laying (Texts A and B in S1 Text). Flies were maintained on a distinct “development” medium prior to experiments, before switching to experimental diets in adulthood (Fig 2B), so that all flies including controls experienced a novel diet upon switch to experimental food, to distribute any novelty effects evenly among conditions. Fig 2A shows approximate nutrient content of this diet, along with control diet, EAA-enriched diet, and lipid diet. We then varied feeding on these experimental diets in 2 different ways (Fig 2B). Fitness effects of long-term dietary changes for both parent and offspring are to be expected, and previous work [13] has shown that such changes can elicit DMN variation. However, diet can also influence offspring health when manipulated only in parents, independent of offspring diet [17,25–27]. To study whether such effects of parental diet are mitonucleogenotype dependent, we exposed flies to either a chronic feeding paradigm, in which both parents and offspring were fed experimental diets, or a parental feeding paradigm, in which diets were fed transiently to parents before eggs were laid and developed on a standardized medium, distinct from parental diet (Fig 2B). In the latter context, DMN interactions can only result from parental effects. To ensure genetic consistency, the same parents were used in each paradigm, by laying eggs for 24 hours after 1 week on experimental media (chronic paradigm), then switching to a universal standardized medium (the medium that the flies developed on) for another 24 hours of egg laying (parental paradigm).

We assessed how phenotypic variation in fecundity, progeny, fertility, and development partitioned by mitonucleogenotype and diet. No trait varied as a linear function of caloric density in any mitonucleogenotype (S5 Fig), and so diet was modeled as an unordered factor. To visualize variation, we calculated estimated marginal means (EMMs; [28]) with confidence intervals. For fecundity, progeny, and fertility, EMMs are statistical coefficients, approximating the trait values imputed to the model. For development, EMMs are a coefficient of a model representing both time to emergence and whether or not an egg developed to adulthood, integrating both parameters into a single development index. Importantly, the EMMs are calculated from statistical models, enabling visual comparisons among conditions, which was useful for our multicondition, multitrait study.

Plotting EMMs per mitonucleogenotype indicated considerable variation in response to diet (Fig 2C). Plotting by geographic origin of mitochondria and nuclei confirmed the same (S6B Fig). Statistical models revealed ubiquitous diet-mitonucleogenotype variation (generalized linear models (GLMs) for fecundity, progeny, and fertility; Cox models for development time), except for fecundity in the parental feeding paradigm (S6 Table). For development time models, we also included interactions with offspring sex, because of reports of sex-biased mito:nuclear variation [16]. However, sex did not modify diet:mitonucleogenotype interactions (all p> 0.05; S6 Table), suggesting that DMN effects on this trait were not sex biased in these populations. Chronic lipid feeding was deleterious for all traits, but mitonucleogenotype shaped magnitude. We were surprised that chronic EAA feeding promoted fecundity, across all populations, but suppressed fertility (Fig 2C), reducing progeny counts to below those of control diet, with the consequence that fitness was not ultimately not enhanced by EAAs. The magnitudes of changes induced by EAAs were mitonucleogenotype dependent (Fig 2C).

Mitonucleogenotypes 3 and 4 stood out in the chronic feeding paradigm, because their progeny counts after chronic EAA feeding were even lower than after chronic lipid enrichment, to near lethality in mitonucleogenotype 3 (Figs 2C and S6D). Again, this effect in mitonucleogenotype 4 was not observed if either mitochondria or nuclei were switched (mitonucleogenotypes 2 or 8), confirming another mitonuclear effect. Mitonuclear incompatibility is widely reported [29], as are DMN effects on physiology and life history [5,13–16]: The present data now indicate that mitonuclear incompatibility can be diet dependent, under nutrient-enriched conditions that we had expected to promote fitness.

DMN variation was also apparent in the parental feeding paradigm, albeit less pronounced than after chronic feeding. To our knowledge, this is the first evidence that mitonucleogenotype modulates effects of parental nutrition. Lipid was less universally toxic upon parental feeding than chronic feeding, and mitonucleogenotypes 3 and 4 stood out, exhibiting a benefit of parental lipid feeding, developing on average 1 day earlier (S6D Fig). However, mitonucleogenotype 4 shared a mitogenotype with mitonucleogenotype 2, and a nucleogenotype with mitonucleogenotype 8, but neither mitonucleogenotypes 2 or 8 showed the same behavior, indicating that the fitness benefit of parental lipid feeding is a mitonuclear interaction effect. The mitonuclear variation in response to parental diet was not universal among all populations. Variation in response to parental diet was pervasive in offspring traits (i.e., progeny, fertility, development index). Not all populations showed statistically significant effects of parental diet, e.g., for number of progeny, these effects were restricted to mitonucleogenotypes 1, 3, 5, 7, and 8 (Fig 2C). These populations were not all the same as those that showed an effect of parental diet on fertility (1, 2, 4, 5, 7, and 8; Fig 2C). Thus, impacts of parental diet on offspring fitness appear to manifest at the level of integration between distinct traits, and mitonuclear variation means that not all populations respond to parental nutritional variation.

For an aggregate view of traits per mitonucleogenotype, we conducted a PCA of trait values (EMMs), which showed that mitonucleogenotype 4 had a response to EAA-enriched food that was distinct not only from mitonucleogenotype 3 but in fact distinct from all other populations in the experiment (S7 Fig). Our sequencing had shown that this population’s mitogenotype was intermediate between other groups of populations at an mtDNA-wide level (Fig 1D), intermediate in the network of significantly differentiated SNPs (Fig 1F), and intermediate in the clustering of significantly differentiated SNPs (Fig 1G). Thus, this population’s mitogenotype is atypical for either Australia or Benin (Fig 1G), with some loci bearing alleles at high frequency in other populations bearing Australian mitogenotypes, and other loci bearing alleles at high frequency in other populations bearing Beninese mitogenotypes. This population bore Australian nuclei, and its mitochondria originated from Australia but were clearly distinct from mitonucleogenotype 3, and its phenotype responded to diet differently. Therefore, we speculate that the lethality of our specific nutrient treatments indicates incompatibility between the Australian nuclear genome and mtDNA loci bearing Benin-like alleles. This incompatibility appears to be diet dependent in this population. We also noted that mitonucleogenotype 5 had a distinct response to high-lipid diet, showing a compromised development index (discussed below).

We applied statistical analysis to confirm diet:mitonucleogenotype effects. We excluded mitonucleogenotype 4 from some statistical analysis because its extreme trait values complicated modeling (see Texts A-D in S1 Text). Among the other populations, ANOVA tests revealed significant mitonucleogenotype:diet interactions (S6 Table). To estimate variability in response to dietary change, we calculated F-ratios and P values for effect of diet per mitonucleogenotype (Fig 2C). F-ratios varied up to 10-fold, depending on trait (Fig 2C). Diet effects were significant for all mitonucleogenotypes in the chronic feeding paradigm (p < 0.001 in all cases) but not in the parental feeding paradigm. These analyses suggest that variance in response to diet can be partitioned by sequence-based mitonucleogenotype [22].

Phenotyping mt:lrRNA SNP

We were particularly interested by the paucity of nonsynonymous mtDNA polymorphisms, which suggested that DMN effects may be underpinned by variation outside of protein-coding regions. As detailed above, mitonucleogenotypes 5, 6, 7, and 8 bore fully factorial variation in nucleogenotype and the mt:lrRNA SNP. Plots of phenotypes in mitonucleogenotypes 5, 6, 7, and 8 (Fig 2C) revealed both quantitative variation in fertility effects of diet but also qualitative changes in the sign of the response to dietary change. Specifically, in populations with the mt:lrRNA T allele, chronic EAA feeding decreased fertility in both nucleogenotypes (mitonucleogenotypes 5 and 6). However, the mt:lrRNA C allele unleashed nucleogenotype-dependent responses to chronic diet: C allele populations with nucleogenotype A showed decreased fertility after EAA feeding (mitonucleogenotype 8) but increased fertility with nucleogenotype B (mitonucleogenotype 7) (Fig 2C). Indeed, mitonucleogenotype 7 was the only mitonucleogenotype that increased fertility upon chronic EAA feeding. In the parental feeding paradigm, nucleogenotypes A and B responded to diet equivalently in the presence of the mt:lrRNA C allele (mitonucleogenotypes 7 and 8). However, nucleogenotype-specific responses to parental diet were unleashed by the mt:lrRNA T allele: Fertility was impaired by parental feeding on either EAA or lipid in the presence of nucleogenotype A (mitonucleogenotype 5) but not in the presence of nucleogenotype B (mitonucleogenotype 6). This altered fertility had apparent consequences for progeny count and development index (Fig 2C). Statistical tests (S7 Table) confirmed interactions of the mt:lrRNA polymorphism, nucleogenotype, and diet, for all traits except egg laying. This exclusively postembryonic variation indicated impacts on offspring performance but not parental reproductive effort.

We also analyzed how the geographic origin of mitochondria and nuclear genome modulated the response to diet (Text C in S1 Text) because this allowed us to assess variance explained by mitochondria and nuclei separately (this is not possible when information is concatenated into mitonucleogenotype, and the fully factorial variation is required to fit a 3-way DMN interaction term). This analysis accorded with our sequence-based analysis of mitonucleogenotype (Text C in S1 Text), revealing DMN variation for all traits except fecundity in the parental paradigm, effects of lipid feeding, and effects of parental diet.

Effect size calculations

Our final analysis assessed the extent to which mitonucleogenotype:diet interactions, and lrRNA:nucleogenotype:diet interactions, shaped phenotypic variation in each respective analysis. We calculated an estimate of effect size (partial η²) that allowed us to compare impacts of predictive variables (Fig 3). We calculated this measure from test statistics [30] using the R effectsize library [31], deriving F values from post hoc EMM tests. (This method of calculating partial η² differs from η² in that the resulting values do not necessarily sum to 1.) We calculated partial η² for all traits, in both feeding paradigms, for each of the 3 different types of analyses we had conducted (i.e., sequence-driven mitonucleogenotype assignment (Fig 3A), the specific analysis of the SNP in lrRNA (Fig 3B), and geographic origin of populations (Fig 3C)). We compared the higher-order interactions we were interested in to lower-order effects, anticipating that diet would be the largest source of variation for most traits, but that this this might be modified by DMN interactions, mitonucleogenotype, or the lrRNA:nucleogenotype interaction. However, the magnitude of DMN effects approached or equaled the direct effects of diet for some traits, indicating that DMN effects constitute a major source of variation.

In the mitonucleogenotype analysis (Fig 3A), the effect of the diet:mitonucleogenotype interaction even exceeded that of diet for fertility in both feeding paradigms. For development, in both feeding paradigms, the effect of the diet:mitonucleogenotype interaction equaled the effect of diet. Again, this suggested that response to diet in these populations could not be understood without first accounting for mitonucleogenotype.

The lrRNA:nucleogenotype:diet interaction was an important source of variation among mitonucleogenotypes 5, 6, 7, and 8 (Fig 3B). For fertility during chronic feeding, lrRNA:nucleogenotype:diet effects were bigger even than for nucleogenotype. Most strikingly, for development in both feeding paradigms, effect size for the lrRNA:nucleogenotype:diet interaction was large, approaching or even equal to main effects of diet. Altogether, these results suggest that epistasis between nucleogenotype and a SNP in noncoding mtDNA can dictate response to diet, which can produce more phenotypic variation than the main effects of mitochondrial or nuclear genotype and can equal the effect of diet.

In the “geographic” analysis (Fig 3C), after chronic feeding, DMN effect sizes were greater than, or equal to, mito:diet and nuclear:diet effects for egg laying, progeny, and fertility. For development in both paradigms, DMN effect sizes were large, on par with diet, diet:mitogenotype, and diet:nucleogenotype, revealing DMN interactions as a major source of variation for developmental impacts of chronic dietary change. Fertility effects were more pronounced after chronic feeding than after parental feeding, but in both paradigms, DMN effect sizes were approximately 75% of diet’s main effect, as were mito:diet and nuclear:diet terms, suggesting that these factors do not simply modulate effect of diet, but their interaction with diet is a substantial source of variation outright. In fact, in the parental nutrition paradigm, DMN effect sizes for fertility outranked the main effect of diet, although with overlapping confidence intervals and modest effect size. However, for development, effect sizes for DMN terms exceeded nearly all other terms (except for the lower-order diet:mito and diet:nuclear interactions), without overlapping confidence intervals—this exceeded even the main effect of diet, suggesting that dietary regulation of this development could not be properly understood without accounting for mitonucleogenotype.

We also validated our effect size calculations orthogonally by assessing how a range of alternative models described the data (AIC analyses) and by calculating variance explained (r²) by each model, which gave congruent results (Text D in S1 Text and S9 and S10 Tables). These analyses suggest that mitonucleogenotype can modulate response to dietary variation, and the emergent interaction can produce as much phenotypic variation as the main effect of diet.

Thus, the effect of diet cannot be understood in the present panel of populations without accounting for mitonucleogenotype. A SNP in a subset of the panel of populations (populations 5 to 8) bore mitochondria differentiated only by a SNP in lrRNA, thereby associating variation in this mtDNA locus to diet-mito-nuclear effects. Our finding that these populations were distinguished only by this SNP, and no other, suggests that this SNP alone can be sufficient to underpin epistatic interactions with the nucleus, which can lead to distinct fitness impacts of altering diet.

Discussion

Predicting phenotype from genotype is a long-standing challenge. To this end, genome-wide association studies (GWASs) have flourished. Two overarching findings of the era of GWAS are that nonprotein-coding variation is more important than previously expected and that additive effects of independently segregating variants do not fully explain quantitative trait variation [4,32]. This latter finding implies “missing heritability,” suggesting that additional processes are at work. Two hypothetical explanations are that genotype-by-genotype epistasis (G*G) and genotype-by-environment (G*E) interactions create nonadditive effects. Speculation about epistasis has led to the “omnigenic model,” which posits that variation in a given trait is likely explained by G*G between a few “core genes,” and the sum effect of many (or all) small-effect variants throughout the rest of the genome [4,32]. Mitonuclear interactions may be a useful illustration of the omnigenic model, with epistasis between the few genes on the mitochondrial genome and the sum of nuclear genomic variants producing substantial phenotypic variation [7]. The omnigenic model predicts that identifying core genes may enable explanation of substantial phenotypic variance and that certain SNPs in core genes may limit or accentuate penetrance of nucleogenotype variation. Our results suggest that core genes with respect to metabolism are to be found on mtDNA and that in a subset of populations the outcome of dietary variation for specific nucleogenotypes depends on a single allele in the mitoribosomal 16s rRNA.

Why should variation in the mitoribosome affect how specific nucleogenotypes respond to specific diets? It is perhaps logical that variation in factors that affect regulatory processes like protein translation should affect penetrance of nuclear variation. mt:lrRNA sits high in a hierarchy of factors that control cellular function, since it encodes a structural unit (16s rRNA) of an organelle that translates proteins, and those proteins are responsible for ATP production for use by the whole cell. Furthermore, mt:lrRNA forms the mitoribosome in complex with nuclear-encoded proteins, so there is clear scope for mitonuclear interactions to mediate function of the mitoribosome. Our findings join others showing that mitochondrial protein translation is a mechanistic fulcrum of mitonuclear interactions. In the copepod Tigriopus californicus, mitoribosomal proteins encoded by the nuclear genome show apparently compensatory evolution in response to rapid mtDNA evolution [33]. These genetic effects can also be environment dependent: In Drosophila, an SNP in mtDNA-encoded tRNA^Tyr can cause male sterility, dependent on nuclear context—specifically the tyrosyl tRNA synthetase Aatm [24]—and this interaction is subject to thermal variation [34]. This interaction can be replicated by modifying photoperiod during development [35], which alters metabolic requirements, therefore suggesting that the epistatic effect of the tRNA variant is mediated by environment-specific energetic requirements and not by more general effects of temperature. Moreover, increasing the ratio of dietary yeast/sugar among approximately isocaloric diets can partially rescue the high-temperature sterility [15]. Since dietary yeast is the fly’s source of protein, this effect could be mediated by essential amino acids, although the role of yeast as a major source of other nutrients (e.g., lipids and vitamins) [36] mean that other nutrients may have been causal [15]. The present study has examined the effect of nutrient-specific variation, which modifies both availability of specific nutrients and total calorie availability. While we were unable to detect any linear effect of calories (S5 Fig), a study designed explicitly to quantify the relationship between caloric effects and mitochondrial translation [37] may be required to discern whether the effects of the mt:lrRNA SNP are due to total energy availability or caused by qualitative differences in proportions of specific nutrients. Prior studies have tended to focus on the relationship between mitochondrial haplotype or SNP, and male fertility [15,24,38,39], and our data now suggest that parental diet can underpin diet-mito-nuclear effects in offspring fitness. We note that the traits modulated by interactions with the mt:lrRNA SNP—fertility, progeny, and offspring development time—are all potentially subject to male fertility effects, though further work is required to conclude such a connection. Overall, our results extend the repertoire of environmental manipulations (specific nutrients) and mtDNA genes (lrRNA) that imply connections between mitochondrial protein translation, metabolism, and fitness. Translation is a critical cellular process with systemic impacts far beyond reproductive traits; for example, modest impairment of general translational machinery can prolong organismal health into old age [40,41]. We suggest that it will be interesting and important to investigate more extensively the phenotypic space affected specifically by mitochondrial protein translation, as this may provide means to individualize therapeutic interventions. Much further work is required to elucidate the molecular, biochemical, and metabolic processes that underpin mitonuclear variation, but our work suggests that a focus on mitoribosomal function may prove illuminating.

Only 2 of the 27 mtDNA SNPs we identified were predicted to change protein coding sequence. It remains to be seen if nonprotein-coding mtDNA variation is as important as nonprotein-coding nDNA variation appears to be [4,32], though our effect size calculations (Fig 3) indicate potentially large roles. Importantly, synonymous and nonprotein-coding mtDNA variation has also been reported in latitudinal clines among wild populations [20], suggesting that this variation may be an important component of natural variation in fitness. More generally, we have added to the growing body of evidence for diet-mito-nuclear interactions [5,13,16], in the context of a literature showing that outcomes of mitonuclear epistasis are environment dependent [7,15]. Thus, altogether, DMN interactions show many of the hallmarks of a major source of phenotypic variation, and we have demonstrated this for Drosophila fitness traits. The Darwinian view that reproduction subjugates all other processes, and the central role of mitochondria in cellular function, suggest that these interactions may be important but underappreciated sources of variation for many further traits and not just in flies.

While we partitioned phenotypic variance to SNPs, we have not attempted systematic GWAS. In general, GWAS to test genome-wide epistasis is not tractable due to the enormous sample that would be required to maintain statistical power. For mitonuclear epistasis, testing consequences of interactions between “only” every mtDNA variant and every nuclear variant would be simpler than testing every pairwise combination of SNPs in the genome (i.e., a*b, rather than (a+b)²), but an enormous sample would still be required. However, if the omnigenic hypothesis is correct [4], such a systematic approach would fail to recognize the underlying biology, which is better modeled as epistasis between a subset of core genes (i.e., mitochondrial genes) and nuclear genomic background (e.g., represented by “background” as in our study, or, alternatively, dimension reduction, pedigrees, marker loci, or pathway-level variation). If candidate mtDNA variants can be identified, methods to test their role conclusively, by mtDNA editing, are on the horizon. mtDNA editing is in its infancy [42] but would facilitate powerful tests of how mitochondria affect outputs of nuclear variation, including response to diet. An additional question raised by our study is the mechanistic role of SNPs outside of protein-coding regions on mtDNA: Are these variants regulatory? Our analyses suggest statistical associations, which would also be testable by mtDNA editing. These tools would also enable mechanistic investigation of how mtDNA variants impact mitochondrial function (e.g., respiration, proteome), their consequences for cellular processes (e.g., metabolism, epigenome), and how their impact on phenotype and response to diet varies among nuclear backgrounds. We do not dismiss the importance of coding variation, but our data suggest that noncoding variation may yet prove important.

Coevolution between mitochondrial and nuclear genomes is expected to optimise fitness. This implies that novel combinations of mitochondrial and nuclear genomes, before coevolution, would bear a fitness cost. While we have focussed mostly on sequence-informed mitonucleogenotype and its relationship to phenotype, at the same time, we can see how phenotypes parse according to geographic origins of mitochondria and nuclei. Our findings have implications for understanding variation that could emerge when novel combinations of mitochondria and nuclei arise. To date, predictions of a disadvantage to “mis-matched” mitonuclear pairings, without coevolutionary history, have received equivocal support [9,24,43–45]. In the present data, a naturally co-occurring mito-nuclear pair (AA) responded uniquely poorly to dietary EAAs. In the case of population AA₃, (mitonucleogenotype 4), EAAs were lethal. This cost was not replicated in either population bearing the constituent mitogenotype (AB) or nucleogenotype (BA), nor in another naturally co-occurring mito-nuclear pair (BB). However, since AA was a naturally co-occurring combination, with probable past coevolution, it does not seem that the detriments of this combination were caused by a novel and poorly matched pairing. These findings indicate that costs and benefits of novel mitonucleogenotypes are not necessarily straightforward functions of mito-nuclear matching or mis-matching. We suggest that novel, non-coevolved mitonucleogenotypes may be variously deleterious, beneficial, or have unforeseen costs and benefits [46].

Diet is a major source of biological variation. But the importance of genotype-by-diet variation is increasingly recognized, with genetic variation manifesting phenotypically only under certain dietary conditions and genotype-specific responses to diet [1,47]. We manipulated 2 specific nutrient classes (EAAs and lipid) normally derived from yeast in fly food, offering greater specificity than previous DMN studies. The nutrients we identify are of particular interest, because EAAs regulate many life history and health traits, while lipid consumption is associated with the pandemic of human metabolic disease [3]. We found that impacts of dietary lipid depend on mitonuclear genotype, which may be relevant to understanding variation in impacts of high-fat human diets. The high-EAA diets that we used have parallels to high-protein diets used to increase yields of livestock and human muscle mass, and our finding that EAAs can decrease offspring quality may give pause for thought in use of these diets. We were surprised that EAA enrichment did not enhance offspring development, because we interpreted increased parental egg laying on this food to indicate parental preference, presumably in anticipation of fitness benefits. However, the discrepancy with development and fertility may indicate that EAAs function as signals of food quality as well as metabolites, which could drive deleterious outcomes when EAA levels are not representative of the composition of food that would be found in the yeasts that flies are thought to consume in nature.

An important finding of our study is that mitonucleogenotype can modify the sign of the response to dietary variation and can even result in lethality for some genotypes on EAA enriched-diet. For traits where lethality was evident, effect size calculations suggested that impacts of DMN interactions were equal in magnitude to the main effect of diet, suggesting not only that mitonucleogenotype modulates response to diet but also that DMN interactions can be major sources of variation outright. Our results thus suggest that mitonuclear incompatibility can be diet dependent. Proportion of flies surviving from egg to adult has previously been reported to be dependent on diet-mito-nuclear interactions [13]. However, in that study, it does not seem that flies experienced widespread lethality.

We have revealed a relationship between parental nutrition and mitonucleogenotype. This is a novel finding. Transient dietary alterations and metabolic disease can drive persistent molecular and phenotypic change, within and across generations [17,48]. In our study, transient parental feeding on a high-lipid diet even accelerated offspring development in AA mitonucleogenotypes—a surprising finding, given that we expected this diet to be largely toxic to parents and offspring. A direct role for diet in selecting embryos can be excluded in parental effects in our study because of the standardized postembryonic environment. Instead, these effects are likely explained by (A) mitonucleogenotype-specific parental allocation of development-accelerating factors, after feeding on specific diets, or (B) mitonucleogenotype-specific selection on offspring from such factors. Further investigation will be required to discriminate between these possibilities; for example, it may be illuminating to investigate variation in the metabolome of offspring whose parents had distinct mitonucleogenotypes and were fed on varied diets. More generally, after both chronic and parental feeding, effects manifested most strongly in postembryonic traits, i.e., for fertility, development time, and total progeny, and a dietary and nuclear interaction with a C/T polymorphism in mt:lrRNA was sufficient to cause these effects. Interestingly, in embryos, lrRNA has been localized outside the mitochondria, in polar granules, suggesting functions in germline determination [49] and highlighting this noncoding RNA as a potential mechanistic link to postembryonic variation. Additional or alternative candidate mechanisms to mediate mitonuclear variation in parental effects include altered epigenetic marks, nutrient provision from mother to offspring, or microbiota. It may be illuminating in the future to ask if effects of parental diet are modulated by parental mitonucleogenotype (e.g., differential nutrient allocation to eggs, gamete epigenome) or mitonucleogenotype-dependent processes in offspring (e.g., differential response to altered maternal nutrition). More generally, our study suggests that nonprotein-coding variation in mitochondria may modify cellular function in ways that are not yet understood but appear to depend on dietary and nuclear genetic context. The SNP in mt:lrRNA may, for example, modify protein translation. A role for small RNAs encoded by the mitochondrial genome is also emerging [50,51], which may, for example, modulate posttranscriptional gene regulation. Altered mitochondrial metabolism resulting from altered regulatory processes will alter overall cellular metabolism, which can have myriad downstream consequences. DMN effects on appetite, feeding rate, and nutrient allocation may be at play. Much further work is now required to elucidate these mechanisms.

In summary, our study shows that (A) specific nutrients’ fitness effects are shaped by interplay of mitochondrial and nuclear genetic variants, that (B) effect sizes of DMN interactions can equal effects of diet, and that (C) a single-nucleotide substitution in mitochondrial 16S rRNA is sufficient to elicit these effects. This suggests that variation in mtDNA does not need to change the protein sequence to interact with the nuclear genome to dictate optimal nutrition.

Materials and methods

Flies

D. melanogaster populations were established as described in Fig 1A. The ancestral Australian population was isolated in Coffs Harbour, New South Wales, Australia [52]. The Benin population is the widely used Dahomey population, isolated in the 1970s in Dahomey (now Benin). The cytoplasmic endosymbiont Wolbachia was cleared by tetracycline treatment 66 generations prior to experiments. For each population, 45 females of the desired mitochondrial background were crossed to 45 males of the desired nuclear background per generation, sampling the daughters of each cross and backcrossing these to males of the desired nuclear background. Iterating this process over many generations led to introgression of the desired nuclear background (from males) into each mitochondrial background. Fly populations were maintained at 25°C on development medium (see below) throughout their history prior to experimentation. For experiments, flies were collected upon eclosion to adulthood and fed fresh developmental medium before being assigned at random to experimental medium in groups of 5 males and 5 females at 3 to 5 days posteclosion. Experimental flies were maintained at 25°C and transferred to fresh media every 48 to 72 hours for 1 week. Flies were transferred to fresh medium 24 hours before egg laying experiments. For development experiments, eggs were incubated at 25°C and pupation and eclosion were scored daily. Eclosing adults were lightly CO₂ anaesthetised before counting and sexing.

Diets

Our study used 2 distinct types of base media. Drosophila populations were constructed and sequenced while feeding on development medium. To manipulate nutrient content, experimental media (see below) were fed. To assay impacts of parental feeding on offspring, parents were returned to developmental medium after an interval of feeding on experimental medium. Using distinct base media ensured that, upon switching to experimental media, all flies were feeding on a new diet, and so any novelty effects were distributed evenly among conditions.

Development medium contained 1.4% agar and 4.5% brewer’s yeast (both Gewürzmühle Brecht, Germany), 10% cornmeal and 11.1% sucrose (both Mühle Milzitz, Germany) (all w/v), 0.45% propionic acid, and 3% nipagin (v/v).

Experimental media built on published protocols [53–55]. These media contained a final concentration of 10% brewer’s yeast, 5% sucrose, 1.5% agar (w/v), 3% nipagin, and 0.3% propionic acid (v/v). EAAs were purchased as powder (Sigma) and supplemented by dissolving into a 6.66× solution in ddH₂0 (pH 4.5) (final media concentrations: L-arginine 0.43 g/l, L-histidine 0.21 g/l, L-isoleucine 0.34 g/l, L-leucine 0.48 g/l, L-lysine 0.52 g/l, L-methionine 0.1 g/l, L-phenylalanine 0.26 g/l, L-threonine 0.37 g/l, L-tryptophan 0.09 g/l, L-valine 0.4 g/l). We added margarine (15% w/v, after [56]) to ensure that lipid supply was plant based, because wild fly physiology is likely influenced by their consumption of what appears to be a vegan diet [57,58] and because margarine sets in agar (in contrast to oils). Margarine (Ja! Pflanzenmargarine from Rewe Supermarkets, Germany; manufacturer’s analysis 720 kcal/100 g; 80/100 g fat from 23/100 g saturated fatty acids, 40/100 g monounsaturated fatty acids, 17/100 g polyunsaturated fatty acids) was briefly melted and then mixed thoroughly into the food (15% w/v). Final nutrient contents of rearing and control media were estimated using the Drosophila diet content calculator [59], with additional protein, lipid, and caloric content after nutrient supplements calculated according to margarine nutrient content report, and by assuming caloric equity between EAAs and protein at a caloric value of 4 calories/g (USDA). Vials contained approximately 5 ml of food and were stored at 4°C for up to 1 week before use.

Characterizing impact of high-lipid diet

The egg laying response to the novel high-lipid diet was characterized in an outbred population of Benin flies. These flies were reared on development medium until day 3 of adulthood, fed on lipid-enriched food or control food for 7 days, and then allowed to lay eggs on development medium overnight. This “switch” design was used to ensure that we measured the physiological capacity to lay eggs following diet treatment.

Genome sequencing

Input DNA controls from a ChIP-Seq experiment were used for whole genome sequence analysis. Genomes of 2 replicates for each mitonuclear combination were sequenced. Pools of 50 adult flies were subjected to a standard native ChIP protocol [60]. The protocol included an MNase digestion step of 6 minutes at 37°C using 15U of the enzyme (Thermo Fisher Scientific) per sample, which yielded fragments between 284 and 300 bp in length. DNA was extracted with the QIAquick PCR purification kit (Qiagen), and 100 ng genomic DNA of the unChIPped input/negative controls were used for library preparations with NEB Next Ultra DNA lib Prep kit for Illumina. Over 50 million 2 × 75bp (PE) reads per sample were sequenced on an Illumina Nextseq 500 platform in High Output (150 cycles) mode.

Nuclear genome analysis

Raw FASTQ reads were trimmed and filtered to remove low-quality reads (minimum base PHRED quality of 18 and minimum read length of 50 bp) prior to mapping using cutadapt (version 2.4; [61]). Reads were mapped to the reference genome of D. melanogaster (Flybase Release 6.28) with bwa aln (version 0.7.12; [62]) using parameters optimized for Pool-seq data [63]. Mapped reads were filtered for proper pairing and a mapping quality of at least 20 using samtools (version 1.9; [64]). Duplicates were removed with Picard (version 2.18.11; http://broadinstitute.github.io/picard/), and sequences flanking indels were realigned with GATK (version 3.8.1.0; [65]). Sequencing depth was assessed using Qualimap (version 2.2.1; [66]) and ranged from 46 to 58× for autosomes and 22 to 28× for X chromosomes. Individual bam files from all samples were then combined into a single mpileup file using samtools (version 1.9; [64]). SNPs were called with the PoolSNP pipeline (version 1.05; https://github.com/capoony/PoolSNP) from the DrosEU project [67], which was specifically developed for SNP detection in Pool-seq data. Parameters for SNP calling were as those used and optimized by the DrosEU project (except minimum count was set to 10) as their dataset closely resembled the present one. The resulting vcf file was converted into a sync file using the python script VCF2sync.py from the DrosEU pipeline.

General genetic differentiation among nuclear genomes was assessed by PCA using the R package LEA (version 3.4.0; [68]) and by estimating admixture proportions with the R package ConStruct (version 1.0.4; [19]). Both approaches were based on major allele frequencies of nuclear SNPs on all major chromosome arms (2L, 2R, 3L, 3R, and X). In order to minimize the effects of LD, only SNPs at least 1 kb apart and outside regions of no recombination [69] were considered. Major allele frequencies were calculated with the python script sync2AF.py from the DrosEU pipeline. Admixture proportions (K = 3) for each population were inferred by nonspatial modeling with 3 MCMC chains per run and 10,000 iterations.

Mitochondrial genome analysis

Sequence data of mitochondrial genomes were retrieved from data of [18] who had sequenced all 27 mitonuclear populations in pools of 150 flies including a prior mitochondrial enrichment protocol. These data were reanalyzed with more stringent sequencing depth criteria. Bam files from this previous study were combined into a single mpileup file using samtools (version 1.9; [64]), which was then converted into a sync file with Popoolation2 [70]. As for nuclear SNPs, a PCA using the R package LEA (version 3.4.0; [68]) was performed, and admixture proportions with the R package ConStruct (version 1.0.4; [19]) were estimated based on the major allele frequencies of all mitochondrial SNPs. Major allele frequencies were calculated with the python script sync2AF.py from the DrosEU pipeline. Admixture proportions (K = 3) for each population were inferred by nonspatial modeling with 3 MCMC chains per run and 10,000 iterations. Genetic differentiation for each mitochondrial SNP was assessed by estimating F_ST according to [71] and Fisher’s exact tests to estimate the significance of the allele frequency differences. Pairwise F_ST and Fisher’s exact tests per SNP were calculated between mitonuclear genotypes (replicates were pooled) with Popoolation2 [70]. Resulting P values from the Fisher’s exact tests were corrected for multiple testing using FDR correction [72], and SNPs significant at an FDR < 0.001 were considered as significantly differentiated between mitonuclear genotypes. The mitochondrial genome map (S4 Fig) was drawn by downloading the full mitochondrial genome sequence from flybase, uploading into SnapGene Viewer as a plasmid sequence, and adding annotations manually.

Quantitative trait analysis

Phenotype data were analyzed in R 4.2.3. Fecundity, progeny, and fertility data were all analyzed per vial of 5 females and 5 males. Development indices were analyzed per fly. Fit of fecundity data to a negative binomial distribution was determined with firdistrplus::descdist and firdistrplus::fitdist. For “geographic” analysis of phenotypes, generalised linear mixed models of the form

y \sim d i e t * m i t o c h o n d r i a * n u c l e a r + (g e n o t y p e)

were fit with lme4::glmer.nb (egg counts and progeny, negative binomial distribution) or lme4::glmer (fertility, binomial of progeny and egg counts); in which diet (control/EAA/lipid), mitochondria (A/B), and nuclear (A/B) were fixed factors, genotype was a random factor denoting fly population (AA1, AA2, AA3, AB1, AB2, BA1, etc.). Where relevant (S2 Fig), experimental replicate was also included as an additional random factor. An observation-level random effect was also included for fertility under chronic feeding to ameliorate overdispersion. ANOVA tests (type 3) were conducted with car::Anova, and post hoc analyses were applied with the functions emmeans::joint_tests, emmeans::pairs, emmeans::emmip [28]. Options for contrasts were set to orthogonal polynomials and sum-to-zero contrasts.

For geographic analysis of development, Cox mixed effects models of the form

y \sim d i e t * m i t o c o n d r i a * n u c l e a r * s e x + r e a r i n g d e n s i t y + (g e n o t y p e)

were fit to the data with coxme::coxme. Diet, mitochondria, nuclear, and genotype terms were as in models of egg laying. Rearing density coded number of eggs laid in the vial in which the individual developed, to account for variation in rearing density. We chose to omit vial as a random effect from the development models because each vial had a matching egg count, included as a fixed effect, and, therefore, including vial would have constituted redundant information that compromised modeling: Indeed, attempting to include vial as a random effect led to poor or failed model fitting. ANOVA and post hoc EMM tests were conducted as per fecundity analyses.

For analyses focussing on the subset of lines bearing only the mt:lrRNA SNP, the relevant subset of the data was modeled as above, replacing the “mitochondria” term with a factor denoting whether the mitochondria bore the C or T variant.

For “mitonucleogenotype” analysis, fecundity, progeny, and fertility data were analyzed with general linear models of the form

y \sim d i e t * m i t o n u c l e o g e n o t y p e

using a negative binomial model (MASS::glm.nb) for fecundity and progeny counts and a binomial model for fertility data (stats::glm, specifying binomial error family). For “mitonucleogenotype” analysis of development, a model of the form

y \sim d i e t * m i t o n u c l e o g e n o t y p e * s e x + r e a r i n g d e n s i t y

was fit using survival::coxph. PCA of phenotype data was conducted with prcomp on scaled EMMs for each trait on each diet. r² was calculated with MuMIn::r.squaredGLMM. Akaike weights were calculated with MuMIn::dredge.

To overcome challenges in calculating effect sizes for 3-way DMN interactions, for each trait, we used EMMs and joint tests to calculate F ratios, from which a measure of effect size within the sample population (partial η²) can be estimated [28,31,73]. Effect sizes were calculated with custom functions built around effectsize::F_to_eta2. Application of this function was necessarily specific to the type of model in question. In all cases, F statistics were taken from ANOVA tables returned by emmeans::joint_tests. For GLMMs and Cox mixed-effect models, degrees of freedom were taken from ANOVA tables returned by emmeans::joint_tests, with residual degrees of freedom calculated by df.residual for GLMMs, or taken from model output for Cox mixed-effect models. For GLMs, degrees of freedom and residual degrees of freedom were taken from stats::anova.

EMMs superimposed on plots were calculated by fitting GLMs of the form

y \sim d i e t * l i n e

and calculating EMM per population per diet. A distinct model was used for EMM calculation because models used to calculate statistical effects did not return main effects for each population (e.g., coefficient for mitogenotype A, nucleogenotype A, control diet) and therefore did not show among-mitonucleogenotype replication (i.e., coefficients for each of population AA₁, AA₂, AA₃ on control diet). Confidence interval of development time EMM for population AA₃ on +EAA diet could not be properly calculated (total lethality should mean no error, giving rise to infinite estimates in our statistical analyses); therefore, CIs were excluded from the plot. Values were returned to original scale by exponentiation when emmeans returned logged values.

Difference indices were calculated from EMMs per population per diet described above. For each pairwise comparison, fold-change was calculated as

\frac{(Y - X)}{X}

where Y represented posttreatment value and X represented starting value. Absolutes of these values were then logged, re-signed, and scaled to a −1:1 scale. P values for difference in EMM were calculated for each pairwise comparison using emmeans::pairs, from which FDR was returned with stats::p.adjust. Bubble plots were produced using ggplot2. Heatmap of nutrient content was plotted in R with superheat. Figures were assembled in Adobe Illustrator.

Graphics

Outlines of Drosophila in Fig 1 were drawn by hand. Mitochondria in Fig 1 were edited from a wikimedia.org representation of an animal cell distributed under Creative Commons CC0 1.0 licence. Skull and crossbone graphics were sourced from openclipart.org under a Creative Commons Zero 1.0 Licence. Map in Fig 1A was drawn in the R package “maps” and edited using Adobe Illustrator.

Supporting information

S1 Text. Text A in S1 Text.

Novel high-lipid diet represses fecundity. Text B in S1 Text. Initial fecundity experiments: Specific nutrients sufficient for DMN variation. Text C in S1 Text. Multitrait phenotyping with chronic and parental diet manipulation: Analysis by geographic origin. Text D S1 Text. AIC and r² calculations.

(DOCX)

Click here for additional data file.^{(75KB, docx)}

S1 Fig. Variance explained by PCA of SNPs sequenced by Pool-seq in each population.

Barplots show variance explained by (A) nuclear SNPs and (B) mitochondrial SNPs. Data underlying the graphs shown in the figure can be found in S16 and S17 Tables.

(PDF)

Click here for additional data file.^{(144.6KB, pdf)}

S2 Fig. Preliminary investigation of impacts of diet:mito:nuclear interactions in the present set of lines, and impact of a novel high-lipid diet.

(A) Reproductive manipulation by enriching fly medium with plant-based lipid. Egg laying by wild-type Benin flies (the ancestral population from which B populations were derived) after 7 days feeding on control medium (10% yeast, 5% sugar) and 15% added plant-based lipid source (margarine). After 1 week of feeding, flies were switched to development medium for egg-laying assay to ensure that any effects resulted from physiological impacts of manipulation and not differences in preference for oviposition on the food. Boxplots show medians, first and third quartiles, and fifth and 95th percentiles. Two-sample t test t = 1.98, df = 16, p = 0.03. Data shown per fly. (B) Key and experimental design. Flies were reared from egg to adult on rearing food and allocated at random to experimental media 6–48 hours after eclosion, at a density of 5 of each sex per vial. After 7 days, flies laid eggs on fresh food for 24 hours. (C) Mitonuclear variation in fecundity response to nutrient enrichment. Plot shows eggs laid in vial of 5 females and 5 males over 24 hours. Boxplots show medians, first and third quartiles, and fifth and 95th percentiles. Points to left of each box show raw data. Connected points to right of each box show estimated marginal means (EMMs) with 90% confidence intervals. Data shown per vial (5 females + 5 males). Egg counts are presented as x+1 to enable plotting log values. (D) Fecundity does not correlate caloric content of experimental media. Scatterplots show eggs at each caloric level, with facets per each combination of mitochondrial (columns) and nuclear (rows) genotype. Diet indicated by color. Populations show smoothed spline through points. Egg counts are presented x+1 to enable plotting log values. Trait values do not linearly correlate with calories; therefore, caloric content is no more informative than modeling diet as an unordered factor. (E) Technical repeatability of diet:mito:nuclear effect between replicate experiments. Each point shows mean egg laying per population per diet in each of 2 replicate experiments, with the replicates of each haplotype grouped by dashed populations. Means were correlated between experiments (Pearson’s r = 0.87, p = 7.6 × 10⁻¹²). (F) Biological repeatability of diet:mito:nuclear effect among replicate lines. Bubble plot shows response index—signed, logged, absolute fold-change in specified comparisons of EMMs—with point size scaled to indicate probability of observed difference (−log10 FDR), and border opacity indicating threshold of statistical significance (FDR ≤ 0.05). Fold-change calculated for conditions on y-axis relative to conditions on x-axis, e.g., bottom-right cluster of points shows increase on EAA-enriched media relative to control. Points along diagonal show comparisons within replicate genotypes on the same diet, with few significant differences among replicate genotypes. In response to lipid enrichment, the same changes were always evident among replicates of the same genotype, and in response to EAA enrichment, similar changes were evident in some replicates. Boxes indicate comparisons among replicates of the same genotype on the same diet. Data underlying the graphs shown in the figure can be found in S18–S20 Tables.

(PDF)

Click here for additional data file.^{(1.9MB, pdf)}

S3 Fig. Significantly differentiated loci between mtDNAs A and B.

In total, 146 SNPs were observed within the 12 populations, of which 27 were significantly differentiated between populations with mtDNAs of different origins. Significant allele frequency differences were assessed by Fisher’s exact test (FDR < 0.001). Stacked barplots show allele frequencies at each locus, per population. Data underlying the graphs shown in the figure can be found in S21 Table.

(PDF)

Click here for additional data file.^{(176.5KB, pdf)}

S4 Fig. Graphical representation of Drosophila melanogaster mitochondrial genome and significantly differentiated SNPs.

Red shows protein-coding genes; blue shows tRNAs; and purple shows rRNAs. Positions of significantly differentiated SNPs shown in orange, with position and alleles.

(PDF)

Click here for additional data file.^{(188.9KB, pdf)}

S5 Fig. Trait values do not correlate linearly with the caloric content of experimental media.

Trait and feeding paradigm indicated above each panel of plots. Within each panel, scatterplots show trait values at each caloric level, with rows for each mitonucleogenotype. Diet indicated by color. Populations show smoothed spline through points. Egg and progeny counts are presented x+1 to enable plotting log values. Trait values do not linearly correlate with calories; therefore, caloric content is no more informative than modeling diet as an unordered factor. Data underlying the graphs shown in the figure can be found in S22 Table.

(PDF)

Click here for additional data file.^{(2.1MB, pdf)}

S6 Fig. Impacts of mitonucleogenotype on response to chronic and parental nutritional change, with data parsed per population in the study.

(A) Key and experimental design. Flies were reared from egg to adult on rearing food and allocated at random to experimental media 6–48 hours after eclosion, at a density of 5 of each sex per vial. After 7 days, flies laid eggs on fresh food for 24 hours, followed by a further 24 hours on standardized rearing medium. (B) Mitonuclear variation in response to chronic and parental changes in nutrition. Panels show EMMs (±95% CI) for trait indicated on y-axis. Feeding paradigm and mitonuclear variation are indicated at the top of the plot. Colors encode diet as per panel A, egg and progeny counts are presented as x+1 to enable plotting on log scale. Development index shows EMMs for Cox mixed-effects models of proportion eclosed over time, excluding sex from plot. Development data are plotted in full as Kaplan–Meier plots in panels (C) and (D). Note the exclusion of EMMs for development of genotype AA₃ in chronic feeding: EAA lethality prevented meaningful estimation. (C, D) Kaplan–Meier plots of development for the indicated feeding paradigms. Plots show proportion eclosed over time. Colors encode diet as per panel (A). Data underlying the graphs shown in the figure can be found in S23 and S24 Tables.

(PDF)

Click here for additional data file.^{(736.6KB, pdf)}

S7 Fig. Diet-mitonucleogenotype interactions and the architecture of phenotype.

PCA shows ordination of populations according to mitogenotype, nucleogenotype, and diet. Results shown from PCA of scaled and mean-centered EMMs, split by facets per genotype, with mitonucleogenotype split by rows. Data underlying the graphs shown in the figure can be found in S25 Table.

(PDF)

Click here for additional data file.^{(294.1KB, pdf)}

S8 Fig. Structure of statistically significant differences among populations.

Bubble plot shows response index—signed, logged, absolute fold-change in specified comparisons of EMMs—with point size scaled to indicate probability of observed difference (−log10 FDR), and border opacity indicating threshold of statistical significance (FDR ≤ 0.05). Fold-change calculated for conditions on y-axis relative to conditions on x-axis, e.g., bottom-right cluster of points shows increase on EAA-enriched media relative to control. Points along diagonal show comparisons within replicate genotypes on the same diet, with few significant differences among replicate genotypes. In response to lipid enrichment, the same changes were always evident in replicate genotypes, and in response to EAA enrichment, similar changes were evident in some replicates. Boxes indicate comparisons within replicate populations on the same diet. Data underlying the graphs shown in the figure can be found in S26 Table.

(PDF)

Click here for additional data file.^{(9.2MB, pdf)}

S1 Table. Rounds of introgression at different phases of the study.

(XLSX)

Click here for additional data file.^{(8.8KB, xlsx)}

S2 Table. Percentage nuclear admixture (most prevalent background) per population.

(XLSX)

Click here for additional data file.^{(9.1KB, xlsx)}

S3 Table. Significantly differentiated SNPs between mtDNAs A and B.

(XLSX)

Click here for additional data file.^{(12.8KB, xlsx)}

S4 Table. Confirmatory DMN for fecundity (GLMM ANOVA tests).

(XLSX)

Click here for additional data file.^{(9.1KB, xlsx)}

S5 Table. Major allele frequencies of significantly differentiated SNPs between mtDNAs A and B derived from 2 different rounds of sequencing (in 2016 and 2018).

(XLSX)

Click here for additional data file.^{(11.8KB, xlsx)}

S6 Table. ANOVA tests (Type III) of diet*mitonucleogenotype interactions (GLMs and Cox proportional hazards).

(XLSX)

Click here for additional data file.^{(10.5KB, xlsx)}

S7 Table. ANOVA tests (Type III) of diet*lnRNA interactions (GLMs and Cox proportional hazards).

(XLSX)

Click here for additional data file.^{(12.2KB, xlsx)}

S8 Table. ANOVA tests (Type III) of diet*mito*nuclear interactions (GLMMs and Cox mixed-effects models).

(XLSX)

Click here for additional data file.^{(12.3KB, xlsx)}

S9 Table. AIC and r² values for permuted GLMMs of diet*mito*nuclear interactions, for fecundity, progeny, and fertility in chronic and parental feeding paradigms.

(XLSX)

Click here for additional data file.^{(25.5KB, xlsx)}

S10 Table. AIC for permuted Cox mixed-effects models of diet*mito*nuclear*sex interactions, for development in chronic and parental feeding paradigms.

(XLSX)

Click here for additional data file.^{(51.4KB, xlsx)}

S11 Table. Source data Fig 1B.

(XLSX)

Click here for additional data file.^{(9.3KB, xlsx)}

S12 Table. Source data Fig 1C.

(XLSX)

Click here for additional data file.^{(10.3KB, xlsx)}

S13 Table. Source data Fig 1D.

(XLSX)

Click here for additional data file.^{(9.6KB, xlsx)}

S14 Table. Source data Fig 1E.

(XLSX)

Click here for additional data file.^{(11.1KB, xlsx)}

S15 Table. Source data Fig 3.

(XLSX)

Click here for additional data file.^{(22.6KB, xlsx)}

S16 Table. Source data S1A Fig.

(XLSX)

Click here for additional data file.^{(8.9KB, xlsx)}

S17 Table. Source data S1B Fig.

(XLSX)

Click here for additional data file.^{(9.1KB, xlsx)}

S18 Table. Source data S2A Fig.

(XLSX)

Click here for additional data file.^{(8.9KB, xlsx)}

S19 Table. Source data S2C–S2E Fig.

(XLSX)

Click here for additional data file.^{(36.4KB, xlsx)}

S20 Table. Source data S2F Fig.

(XLSX)

Click here for additional data file.^{(151.6KB, xlsx)}

S21 Table. Source data S3 Fig.

(XLSX)

Click here for additional data file.^{(14.8KB, xlsx)}

S22 Table. Source data (fecundity, progeny, fertility) Figs 2C, 2D and S5.

(XLSX)

Click here for additional data file.^{(31.6KB, xlsx)}

S23 Table. Source data (development, chronic feeding) Figs 2C, 2D and S6B–S6D.

(XLSX)

Click here for additional data file.^{(806.4KB, xlsx)}

S24 Table. Source data (development, parental feeding) Figs 2C, 2D and S6B–S6D.

(XLSX)

Click here for additional data file.^{(630.7KB, xlsx)}

S25 Table. Source data S7 Fig.

(XLSX)

Click here for additional data file.^{(11.7KB, xlsx)}

S26 Table. Source data S8 Fig.

(XLSX)

Click here for additional data file.^{(1.2MB, xlsx)}

Acknowledgments

We thank L. Holman, S. Parratt, C. Selman, E. Combet, D. Marcu, D. Sannino, V. Howick, and J. Rolff for helpful discussion. C. Froschauer and R. Dobler provided invaluable advice in setting up experiments and the populations.

Abbreviations

DMN: diet-by-mito-by-nuclear
EAA: essential amino acid
EMM: estimated marginal mean
GLM: generalized linear model
GWAS: genome-wide association study
LD: linkage disequilibrium
PCA: principal components analysis
SNP: single nucleotide polymorphism

Data Availability

Phenotype data and R code are available at github.com/dobdobby, and in supplementary materials. Sequence data are available from NCBI SRA, accession PRJNA853138.

Funding Statement

This work was supported by a Dresden Fellowship funded by the Excellence Initiative of the German Federal and State Governments to A.D., a UKRI Future Leaders Fellowship (MR/S033939/1) to A.D., a University of Glasgow Lord Kelvin Adam Smith Fellowship to A.D., and Deutsche Forschungsgemeinschaft grant RE 1666/9-1 to K.R. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

1.Reed LK, Williams S, Springston M, Brown J, Freeman K, DesRoches CE, et al. Genotype-by-Diet Interactions Drive Metabolic Phenotype Variation in Drosophila melanogaster. Genetics. 2010;185:1009–1019. doi: 10.1534/genetics.109.113571 [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Liao C-Y, Rikke BA, Johnson TE, Diaz V, Nelson JF. Genetic variation in the murine lifespan response to dietary restriction: from life extension to life shortening. Aging Cell. 2009;9:92–95. doi: 10.1111/j.1474-9726.2009.00533.x [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Simpson SJ, Raubenheimer D. The Nature of Nutrition: A Unifying Framework from Animal Adaptation to Human Obesity. 2012. doi: 10.1515/9781400842803 [DOI] [Google Scholar]
4.Boyle EA, Li YI, Pritchard JK. An Expanded View of Complex Traits: From Polygenic to Omnigenic. Cell. 2017;169:1177–1186. doi: 10.1016/j.cell.2017.05.038 [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Zhu C-T, Ingelmo P, Rand DM. G×G×E for Lifespan in Drosophila: Mitochondrial, Nuclear, and Dietary Interactions that Modify Longevity. PLoS Genet. 2014:10. doi: 10.1371/journal.pgen.1004354 [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Rand DM, Mossman JA. Mitonuclear conflict and cooperation govern the integration of genotypes, phenotypes and environments. Philos Trans R Soc B. 2020;375:20190188. doi: 10.1098/rstb.2019.0188 [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Rand DM, Mossman JA, Spierer AN, Santiago JA. Mitochondria as environments for the nuclear genome in Drosophila: mitonuclear G×G×E. J Hered. 2021. doi: 10.1093/jhered/esab066 [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Ali AT, Boehme L, Carbajosa G, Seitan VC, Small KS, Hodgkinson A. Nuclear genetic regulation of the human mitochondrial transcriptome. elife. 2019;8:e41927. doi: 10.7554/eLife.41927 [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Ma H, Gutierrez NM, Morey R, Dyken CV, Kang E, Hayama T, et al. Incompatibility between Nuclear and Mitochondrial Genomes Contributes to an Interspecies Reproductive Barrier. Cell Metab. 2016;24:283–294. doi: 10.1016/j.cmet.2016.06.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Hill GE. Reconciling the mitonuclear compatibility species concept with rampant mitochondrial introgression. Integr Comp Biol. 2019;59:912–924. doi: 10.1093/icb/icz019 [DOI] [PubMed] [Google Scholar]
11.Healy TM, Burton RS. Strong selective effects of mitochondrial DNA on the nuclear genome. P Natl Acad Sci U S A. 2020;117:6616–6621. doi: 10.1073/pnas.1910141117 [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Santiago JC, Boylan JM, Lemieux FA, Gruppuso PA, Sanders JA, Rand DM. Mitochondrial genotype alters the impact of rapamycin on the transcriptional response to nutrients in Drosophila. BMC Genomics. 2021;22:213. doi: 10.1186/s12864-021-07516-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Mossman JA, Biancani LM, Zhu C-T, Rand DM. Mitonuclear Epistasis for Development Time and Its Modification by Diet in Drosophila. Genetics. 2016;203:463–484. doi: 10.1534/genetics.116.187286 [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Rand DM, Mossman JA, Zhu L, Biancani LM, Ge JY. Mitonuclear epistasis, genotype-by-environment interactions, and personalized genomics of complex traits in Drosophila: Mitonuclear G x G x E. IUBMB Life. 2018;70:1275–1288. doi: 10.1002/iub.1954 [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Montooth KL, Dhawanjewar AS, Meiklejohn CD. Temperature-sensitive reproduction and the physiological and evolutionary potential for Mother’s Curse. Integr Comp Biol. 2019;59:890–899. doi: 10.1093/icb/icz091 [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Camus MF, O’Leary M, Reuter M, Lane N. Impact of mitonuclear interactions on life-history responses to diet. Philos Trans R Soc B. 2020;375:20190416. doi: 10.1098/rstb.2019.0416 [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Deas JB, Blondel L, Extavour CG. Ancestral and offspring nutrition interact to affect life-history traits in Drosophila melanogaster. Proc R Soc B. 2019;286:20182778. doi: 10.1098/rspb.2018.2778 [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Vaught RC, Voigt S, Dobler R, Clancy DJ, Reinhardt K, Dowling DK. Interactions between cytoplasmic and nuclear genomes confer sex-specific effects on lifespan in Drosophila melanogaster. J Evol Biol. 2020;33:694–713. doi: 10.1111/jeb.13605 [DOI] [PubMed] [Google Scholar]
19.Bradburd GS, Coop GM, Ralph PL. Inferring Continuous and Discrete Population Genetic Structure Across Space. Genetics. 2018;210:33–52. doi: 10.1534/genetics.118.301333 [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Camus MF, Wolff JN, Sgrò CM, Dowling DK. Experimental Support That Natural Selection Has Shaped the Latitudinal Distribution of Mitochondrial Haplotypes in Australian Drosophila melanogaster. Mol Biol Evol. 2017;34:2600–2612. doi: 10.1093/molbev/msx184 [DOI] [PubMed] [Google Scholar]
21.Bergland AO, Tobler R, González J, Schmidt P, Petrov D. Secondary contact and local adaptation contribute to genome-wide patterns of clinal variation in Drosophila melanogaster. Mol Ecol. 2016;25:1157–1174. doi: 10.1111/mec.13455 [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Grandison RC, Piper MD, Partridge L. Amino-acid imbalance explains extension of lifespan by dietary restriction in Drosophila. Nature. 2009;462:1061–1064. doi: 10.1038/nature08619 [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Aw WC, Towarnicki SG, Melvin RG, Youngson NA, Garvin MR, Hu Y, et al. Genotype to phenotype: Diet-by-mitochondrial DNA haplotype interactions drive metabolic flexibility and organismal fitness. PLoS Genet. 2018;14:e1007735. doi: 10.1371/journal.pgen.1007735 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Meiklejohn CD, Holmbeck MA, Siddiq MA, Abt DN, Rand DM, Montooth KL. An Incompatibility between a Mitochondrial tRNA and Its Nuclear-Encoded tRNA Synthetase Compromises Development and Fitness in Drosophila. PLoS Genet. 2013;9:e1003238. doi: 10.1371/journal.pgen.1003238 [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Dearden L, Ozanne S. You are what your mother ate: The effects of maternal obesity during pregnancy on offspring obesity and cardiovascular diseases. Physiology News. 2017:26–28. doi: 10.36866/pn.106.26 [DOI] [Google Scholar]
26.Camilleri T, Piper MDW, Robker RL, Dowling DK. Maternal and paternal sugar consumption interact to modify offspring life history and physiology. Funct Ecol. 2022. doi: 10.1111/1365-2435.14024 [DOI] [Google Scholar]
27.Camilleri-Carter T-L, Dowling DK, Robker RL, Piper MDW. Transgenerational Obesity and Healthy Aging in Drosophila. J Gerontol A Biol Sci Med Sci. 2019. doi: 10.1093/gerona/glz154 [DOI] [PubMed] [Google Scholar]
28.Searle SR, Speed FM, Milliken GA. Population Marginal Means in the Linear Model: An Alternative to Least Squares Means. Am Stat. 1980;34:216–221. doi: 10.1080/00031305.1980.10483031 [DOI] [Google Scholar]
29.Hill GE, Havird JC, Sloan DB, Burton RS, Greening C, Dowling DK. Assessing the fitness consequences of mitonuclear interactions in natural populations. Biol Rev. 2018;94:1089–1104. doi: 10.1111/brv.12493 [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Friedman H. Simplified Determinations of Statistical Power, Magnitude of Effect and Research Sample Sizes. Educ Psychol Meas. 1982;42:521–526. doi: 10.1177/001316448204200214 [DOI] [Google Scholar]
31.Ben-Shachar M, Lüdecke D, Makowski D. effectsize: Estimation of Effect Size Indices and Standardized Parameters. J Open Source Softw. 2020;5:2815. doi: 10.21105/joss.02815 [DOI] [Google Scholar]
32.Mathieson I. The omnigenic model and polygenic prediction of complex traits. Am J Hum Genet. 2021;108:1558–1563. doi: 10.1016/j.ajhg.2021.07.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Barreto FS, Burton RS. Evidence for Compensatory Evolution of Ribosomal Proteins in Response to Rapid Divergence of Mitochondrial rRNA. Mol Biol Evol. 2013;30:310–314. doi: 10.1093/molbev/mss228 [DOI] [PubMed] [Google Scholar]
34.Hoekstra LA, Siddiq MA, Montooth KL. Pleiotropic Effects of a Mitochondrial–Nuclear Incompatibility Depend upon the Accelerating Effect of Temperature in Drosophila. Genetics. 2013;195:1129–1139. doi: 10.1534/genetics.113.154914 [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Hoekstra LA, Julick CR, Mika KM, Montooth KL. Energy demand and the context-dependent effects of genetic interactions underlying metabolism. Evol Lett. 2018;2:102–113. doi: 10.1002/evl3.47 [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Piper MD, Blanc E, Leitão-Gonçalves R, Yang M, He X, Linford NJ, et al. A holidic medium for Drosophila melanogaster. Nat Methods. 2014;11:100–105. doi: 10.1038/nmeth.2731 [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Camus MF, Moore J, Reuter M. Nutritional geometry of mitochondrial genetic effects on male fertility. Biol Lett. 2020;16:20190891. doi: 10.1098/rsbl.2019.0891 [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Clancy DJ, Hime GR, Shirras AD. Cytoplasmic male sterility in Drosophila melanogaster associated with a mitochondrial CYTB variant. Heredity. 2011;107:374–376. doi: 10.1038/hdy.2011.12 [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Patel MR, Miriyala GK, Littleton AJ, Yang H, Trinh K, Young JM, et al. A mitochondrial DNA hypomorph of cytochrome oxidase specifically impairs male fertility in Drosophila melanogaster. elife. 2016;5:e16923. doi: 10.7554/eLife.16923 [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Filer D, Thompson MA, Takhaveev V, Dobson AJ, Kotronaki I, Green JW, et al. RNA polymerase III limits longevity downstream of TORC1. Nature. 2017;552:263. doi: 10.1038/nature25007 [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Corrales GM, Filer D, Wenz KC, Rogan A, Phillips G, Li M, et al. Partial Inhibition of RNA Polymerase I Promotes Animal Health and Longevity. Cell Rep. 2020;30:1661–1669.e4. doi: 10.1016/j.celrep.2020.01.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Gammage PA, Moraes CT, Minczuk M. Mitochondrial Genome Engineering: The Revolution May Not Be CRISPR-Ized. Trends Genet. 2018;34:101–110. doi: 10.1016/j.tig.2017.11.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Baris TZ, Wagner DN, Dayan DI, Du X, Blier PU, Pichaud N, et al. Evolved genetic and phenotypic differences due to mitochondrial-nuclear interactions. PLoS Genet. 2017;13:e1006517. doi: 10.1371/journal.pgen.1006517 [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Chang C-C, Rodriguez J, Ross J. Mitochondrial–Nuclear Epistasis Impacts Fitness and Mitochondrial Physiology of Interpopulation Caenorhabditis briggsae Hybrids. G3. 2016;6:209–219. doi: 10.1534/g3.115.022970 [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Ellison CK, Burton RS. Disruption of mitochondrial function in interpopulation hybrids of Tigriopus californicus. Evolution. 2006;60:1382–1391. doi: 10.1554/06-210.1 [DOI] [PubMed] [Google Scholar]
46.Dobler R, Dowling DK, Morrow EH, Reinhardt K. A systematic review and meta-analysis reveals pervasive effects of germline mitochondrial replacement on components of health. Hum Reprod Update. 2018;24:519–534. doi: 10.1093/humupd/dmy018 [DOI] [PubMed] [Google Scholar]
47.Reed LK, Lee K, Zhang Z, Rashid L, Poe A, Hsieh B, et al. Systems Genomics of Metabolic Phenotypes in Wild-Type Drosophila melanogaster. Genetics. 2014;197:781–793. doi: 10.1534/genetics.114.163857 [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Holland ML, Lowe R, Caton PW, Gemma C, Carbajosa G, Danson AF, et al. Early-life nutrition modulates the epigenetic state of specific rDNA genetic variants in mice. Science (New York, NY). 2016;353:495–498. doi: 10.1126/science.aaf7040 [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Kobayashi S, Amikura R, Okada M. Presence of Mitochondrial large Ribosomal RNA Outside Mitochondria in Germ Plasm of Drosophila melanogaster. Science. 1993;260:1521–1524. doi: 10.1126/science.7684857 [DOI] [PubMed] [Google Scholar]
50.Pozzi A, Dowling DK. New insights into mitochondrial-nuclear interactions revealed through analysis of small RNAs. Genome Biol Evol. 2022;14:evac023–. doi: 10.1093/gbe/evac023 [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Pozzi A, Dowling DK. Small mitochondrial RNAs as mediators of nuclear gene regulation, and potential implications for human health. BioEssays. 2021;43:e2000265. doi: 10.1002/bies.202000265 [DOI] [PubMed] [Google Scholar]
52.Dowling DK, Williams BR, Garcia-Gonzalez F. Maternal sexual interactions affect offspring survival and ageing. J Evol Biol. 2014;27:88–97. doi: 10.1111/jeb.12276 [DOI] [PubMed] [Google Scholar]
53.Bass TM, Grandison RC, Wong R, Martinez P, Partridge L, Piper MDW. Optimization of Dietary Restriction Protocols in Drosophila. J Gerontol A Biol Sci Med Sci. 2007;62:1071–1081. doi: 10.1093/gerona/62.10.1071 [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Emran S, Yang M, He X, Zandveld J, Piper MD. Target of rapamycin signalling mediates the lifespan-extending effects of dietary restriction by essential amino acid alteration. Aging. 2014;6:390–398. doi: 10.18632/aging.100665 [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Dobson AJ, He X, Blanc E, Bolukbasi E, Feseha Y, Yang M, et al. Tissue-specific transcriptome profiling of Drosophila reveals roles for GATA transcription factors in longevity by dietary restriction. NPJ Aging Mech Dis. 2018:4. doi: 10.1038/s41514-018-0024-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Woodcock KJ, Kierdorf K, Pouchelon CA, Vivancos V, Dionne MS, Geissmann F. Macrophage-Derived upd3 Cytokine Causes Impaired Glucose Homeostasis and Reduced Lifespan in Drosophila Fed a Lipid-Rich Diet. Immunity. 2015;42:133–144. doi: 10.1016/j.immuni.2014.12.023 [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Brankatschk M, Gutmann T, Knittelfelder O, Palladini A, Prince E, Grzybek M, et al. A Temperature-Dependent Switch in Feeding Preference Improves Drosophila Development and Survival in the Cold. Dev Cell. 2018;46:781–793.e4. doi: 10.1016/j.devcel.2018.05.028 [DOI] [PubMed] [Google Scholar]
58.Knittelfelder O, Prince E, Sales S, Fritzsche E, Wöhner T, Brankatschk M, et al. Sterols as dietary markers for Drosophila melanogaster. Biochim Biophys Acta Mol Cell Biol Lipids. 2020;1865:158683. doi: 10.1016/j.bbalip.2020.158683 [DOI] [PubMed] [Google Scholar]
59.Lesperance DNA, Broderick NA. Meta-analysis of Diets Used in Drosophila Microbiome Research and Introduction of the Drosophila Dietary Composition Calculator (DDCC). G3 Genes Genomes. Genetics. 2020;10:g3.401235.2020. doi: 10.1534/g3.120.401235 [DOI] [PMC free article] [PubMed] [Google Scholar]
60.Cosseau C, Azzi A, Smith K, Freitag M, Mitta G, Grunau C. Native chromatin immunoprecipitation (N-ChIP) and ChIP-Seq of Schistosoma mansoni: Critical experimental parameters. Mol Biochem Parasitol. 2009;166:70–76. doi: 10.1016/j.molbiopara.2009.02.015 [DOI] [PubMed] [Google Scholar]
61.Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. Embnet J. 2011;17:10–12. doi: 10.14806/ej.17.1.200 [DOI] [Google Scholar]
62.Li H, Durbin R. Fast and accurate short read alignment with Burrows–Wheeler transform. Bioinformatics. 2009;25:1754–1760. doi: 10.1093/bioinformatics/btp324 [DOI] [PMC free article] [PubMed] [Google Scholar]
63.Kofler R, Orozco-terWengel P, Maio ND, Pandey RV, Nolte V, Futschik A, et al. PoPoolation: A Toolbox for Population Genetic Analysis of Next Generation Sequencing Data from Pooled Individuals. PLoS ONE. 2011;6:e15925. doi: 10.1371/journal.pone.0015925 [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, et al. The Sequence Alignment/Map format and SAMtools. Bioinformatics. 2009;25:2078–2079. doi: 10.1093/bioinformatics/btp352 [DOI] [PMC free article] [PubMed] [Google Scholar]
65.McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A, et al. The Genome Analysis Toolkit: A MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res. 2010;20:1297–1303. doi: 10.1101/gr.107524.110 [DOI] [PMC free article] [PubMed] [Google Scholar]
66.Okonechnikov K, Conesa A, García-Alcalde F. Qualimap 2: advanced multi-sample quality control for high-throughput sequencing data. Bioinformatics. 2016;32:292–294. doi: 10.1093/bioinformatics/btv566 [DOI] [PMC free article] [PubMed] [Google Scholar]
67.Kapun M, Barrón MG, Staubach F, Obbard DJ, Wiberg RAW, Vieira J, et al. Genomic Analysis of European Drosophila melanogaster Populations Reveals Longitudinal Structure, Continent-Wide Selection, and Previously Unknown DNA Viruses. Mol Biol Evol. 2020;37:2661–2678. doi: 10.1093/molbev/msaa120 [DOI] [PMC free article] [PubMed] [Google Scholar]
68.Frichot E, François O. LEA: An R package for landscape and ecological association studies. Methods Ecol Evol. 2015;6:925–929. doi: 10.1111/2041-210x.12382 [DOI] [Google Scholar]
69.Comeron JM, Ratnappan R, Bailin S. The Many Landscapes of Recombination in Drosophila melanogaster. PLoS Genet. 2012;8:e1002905. doi: 10.1371/journal.pgen.1002905 [DOI] [PMC free article] [PubMed] [Google Scholar]
70.Kofler R, Pandey RV, Schlötterer C. PoPoolation2: identifying differentiation between populations using sequencing of pooled DNA samples (Pool-Seq). Bioinformatics. 2011;27:3435–3436. doi: 10.1093/bioinformatics/btr589 [DOI] [PMC free article] [PubMed] [Google Scholar]
71.Hudson RR, Slatkin M, Maddison WP. Estimation of levels of gene flow from DNA sequence data. Genetics. 1992;132:583–589. doi: 10.1093/genetics/132.2.583 [DOI] [PMC free article] [PubMed] [Google Scholar]
72.Benjamini Y, Hochberg Y. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. J R Stat Soc Ser B Methodol. 1995;57:289–300. doi: 10.1111/j.2517-6161.1995.tb02031.x [DOI] [Google Scholar]
73.Cohen J. Eta-Squared and Partial Eta-Squared in Fixed Factor Anova Designs. Educ Psychol Meas. 1973;33:107–112. doi: 10.1177/001316447303300111 [DOI] [Google Scholar]

PLoS Biol. 2023 Aug 21;21(8):e3002218. doi: 10.1371/journal.pbio.3002218.r001

Author response to previous submission

14 Jul 2022

Attachment

Submitted filename: 127022_1_rebuttal_2389737_rdsq53.docx

Click here for additional data file.^{(86.6KB, docx)}

PLoS Biol. doi: 10.1371/journal.pbio.3002218.r002

Decision Letter 0

Ines Alvarez-Garcia

9 Aug 2022

Dear Adam,

Thank you for submitting your revised Review Commons manuscript entitled "Diet's impact dictated by synonymous mitochondrial SNP interacting with nucleotype" for consideration as a Research Article by PLOS Biology.

Your manuscript has now been evaluated by the PLOS Biology editorial staff, as well as by an academic editor with relevant expertise, and I'm writing to let you know that we would like to send your revised submission out for re-review. Please accept my apologies for the extraordinary time that it has taken us to secure expert advice at this challenging time of year.

IMPORTANT: The Academic Editor has assessed the two existing Review Commons reviews, but (based on the expertise of the reviewers) feels that some aspects of your study have not yet been fully assessed. S/he has therefore asked us to recruit one or two new reviewers to fill this expertise gap (this is not unprecedented with Review Commons manuscripts).

However, before we can send your manuscript to reviewers, we need you to complete your submission by providing the metadata that is required for full assessment. To this end, please login to Editorial Manager where you will find the paper in the 'Submissions Needing Revisions' folder on your homepage. Please click 'Revise Submission' from the Action Links and complete all additional questions in the submission questionnaire.

Once your full submission is complete, your paper will undergo a series of checks in preparation for further peer review. After your manuscript has passed the checks it will be sent out for review. To provide the metadata for your submission, please Login to Editorial Manager (https://www.editorialmanager.com/pbiology) within two working days, i.e. by Aug 11 2022 11:59PM.

Feel free to email us at plosbiology@plos.org if you have any queries relating to your submission.

Kind regards,

Roli

Roland G Roberts, PhD

Senior Editor

PLOS Biology

rroberts@plos.org

on behalf of

Ines Alvarez-Garcia, PhD

Senior Editor

PLOS Biology

ialvarez-garcia@plos.org

PLoS Biol. 2023 Aug 21;21(8):e3002218. doi: 10.1371/journal.pbio.3002218.r003

Author response to Decision Letter 0

22 Aug 2022

Attachment

Submitted filename: 127022_1_rebuttal_2389737_rdsq53.docx

Click here for additional data file.^{(86.6KB, docx)}

PLoS Biol. doi: 10.1371/journal.pbio.3002218.r004

Decision Letter 1

Ines Alvarez-Garcia

23 Oct 2022

Dear Dr Dobson,

Thank you for your patience while your revised manuscript entitled "Diet's impact dictated by SNP in mitoribosomal 16S rRNA interacting with nucleotype" was peer-reviewed at PLOS Biology and please accept my sincere apologies for the delay in providing you with our decision. The manuscript has now been evaluated by the PLOS Biology editors, an Academic Editor with relevant expertise, one of the original reviewers from Review Commons and a new one that we recruited after advice from the Academic Editor.

As you will see, the reviewers have mixed opinions. While the original reviewer appreciates the improvements done in the manuscript during the revision and asks for several clarifications, the new reviewer raises concerns regarding the general presentation of the data, the statistics and the experimental design – particularly, the mtDNA genetics. In addition, the reviewer thinks that several statements regarding novelty are overstated and should be toned down. After discussing the reviews with the Academic Editor and the rest of the team, we think that all the issues raised by Reviewer 3 would have to be satisfactorily addressed in order for us to consider the manuscript further for publication and that we will consider it only as a Short Report (please look at our guidelines regarding this format - https://journals.plos.org/plosbiology/s/what-we-publish#loc-Research-based-content). We also think that the identification of a candidate SNP in the mtDNA is interesting and follow up experiments investigating this as a causal SNP would strengthen the manuscript.

In light of the reviews, which you will find at the end of this email, we would like to invite you to revise the work to thoroughly address the reviewers' reports.

Given the extent of revision needed, we cannot make a decision about publication until we have seen the revised manuscript and your response to the reviewers' comments. Your revised manuscript is likely to be sent for further evaluation by a subset of the reviewers.

We expect to receive your revised manuscript within 3 months. Please email us (plosbiology@plos.org) if you have any questions or concerns, or would like to request an extension.

At this stage, your manuscript remains formally under active consideration at our journal; please notify us by email if you do not intend to submit a revision so that we may withdraw it.

**IMPORTANT - SUBMITTING YOUR REVISION**

Your revisions should address the specific points made by each reviewer. Please submit the following files along with your revised manuscript:

1. A 'Response to Reviewers' file - this should detail your responses to the editorial requests, present a point-by-point response to all of the reviewers' comments, and indicate the changes made to the manuscript.

*NOTE: In your point-by-point response to the reviewers, please provide the full context of each review. Do not selectively quote paragraphs or sentences to reply to. The entire set of reviewer comments should be present in full and each specific point should be responded to individually, point by point.

You should also cite any additional relevant literature that has been published since the original submission and mention any additional citations in your response.

2. In addition to a clean copy of the manuscript, please also upload a 'track-changes' version of your manuscript that specifies the edits made. This should be uploaded as a "Revised Article with Changes Highlighted" file type.

3. Resubmission Checklist

When you are ready to resubmit your revised manuscript, please refer to this resubmission checklist: https://plos.io/Biology_Checklist

To submit a revised version of your manuscript, please go to https://www.editorialmanager.com/pbiology/ and log in as an Author. Click the link labelled 'Submissions Needing Revision' where you will find your submission record.

Please make sure to read the following important policies and guidelines while preparing your revision and fulfil the editorial requests:

a) *PLOS Data Policy*

Please note that as a condition of publication PLOS' data policy (http://journals.plos.org/plosbiology/s/data-availability) requires that you make available all data used to draw the conclusions arrived at in your manuscript. If you have not already done so, you must include any data used in your manuscript either in appropriate repositories, within the body of the manuscript, or as supporting information (N.B. this includes any numerical values that were used to generate graphs, histograms etc.). Please also indicate in each figure legend where the data can be found. For an example see here: http://www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.1001908#s5

b) *Published Peer Review*

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out. Please see here for more details:

https://blogs.plos.org/plos/2019/05/plos-journals-now-open-for-published-peer-review/

c) *Blot and Gel Data Policy*

Please provide the original, uncropped and minimally adjusted images supporting all blot and gel results reported in an article's figures or Supporting Information files. We will require these files before a manuscript can be accepted so please prepare them now, if you have not already uploaded them. Please carefully read our guidelines for how to prepare and upload this data: https://journals.plos.org/plosbiology/s/figures#loc-blot-and-gel-reporting-requirements

d) *Protocols deposition*

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosbiology/s/submission-guidelines#loc-materials-and-methods

Thank you again for your submission to our journal. We hope that our editorial process has been constructive thus far, and we welcome your feedback at any time. Please don't hesitate to contact us if you have any questions or comments.

Sincerely,

Ines

Ines Alvarez-Garcia, PhD

Senior Editor

PLOS Biology

ialvarez-garcia@plos.org

------------------------------------------------------------------

Reviewers' comments

Rev. 1:

This paper investigates mitochondrial x nuclear x diet interactions in Drosophila melanogaster. They create fully-factorial combinations of mitochondrial and nuclear genomes from Australia, Benin, and Canada. These populations are then exposed to various diets, including a control diet, a diet high in essential amino acids, and a diet high in plant-based lipids. They screened for evidence of repeatable mitonuclear effects on fecundity and additional fitness traits looking for influence of mito-nucleotype on response to chronic vs parental dietary changes, which they found evidence of. However, the effect of mito-nucleotype on traits was variable. Certain mito-nucleotypes exhibited lower to near lethality progeny counts after amino acid feeding (which is thought to promote fecundity), while others exhibited normal counts. When quantifying the size of various effects, they found that mitonucleotype interactions often had comparable effect size to that of diet:mitotype interactions, diet:nucleotype interactions, and diet on its own. Finally, they were able to associate this mito-nucleotype interaction with a mt:lrRNA C/T polymorphism that had nucleotype-dependent effects on fertility.

I have previously reviewed this manuscript for a different journal, and I find the author's response to my review comments satisfactory. In the new draft, I noticed rewording of several sections, which I think improve the manuscript and will help readers understand the methods and results. I appreciate the additional commentary on the identified mt:lrRNA that housed an mtDNA SNP associated with phenotypic differences between lines.

I believe this study and the associated results will obviously be of interest to those studying the evolution of mitochondrial x nuclear x environmental interacts but also to those interested in the effects of nutrients/diet in Drosophila. I also find the comparison to the omnigenic model quite compelling, and I appreciate the idea that mitochondrial genes could contribute to the "core gene" set for several notable phenotypes - potentially through a regulatory mechanism.

I have only a few minor comments (nearly all typos/clarifications).

Page 15, line 75 - This sentence confused me - "This approach is expected to dilute with the paternal nucleotype and before eventually purging the F0 mother's nucleotype, while retaining the F0 mother's mitotype." I think the "and" is erroneous.

Page 17, line 118 - 120 - This may be a naive question. To ensure genetic consistency you use the same parents in each paradigm. These parents spend one week on experimental media (chronic paradigm) and then are allowed to lay eggs for 24 hours. They are then switched to a universal medium for another 24h of egg laying (parental paradigm). Is 24 hours on the universal medium sufficient to remove/minimize the effects of the experimental media? Could diet effects from the experimental media persist into egg laying while on the universal medium?

In my previous review, I asked about co-adaptation between the mitochondrial and nuclear genome. The response to reviewers mentions that a commentary on co-evolution between the mitochondrial/nuclear genomes was added on lines 289 - 297. These lines do not mention co-evolution/adaptation, and I could not find any commentary in the rest of the discussion. I'm curious whether the referenced text was not added to the version we have.

Rev. 3:

Major comments:

The main text should more clearly describe the underlying genetics. Was the Australia source stock/population segregating mtDNA haplotypes that are very similar to Benin? If so, why? Was a single female sampled to provide the mtDNA for each replicate mitochondrial-introgression line? If not, do you have evidence that there are not segregating mtDNAs within each replicate line? In the end, this is fortuitous because it enables a candidate SNP to be identified. However, once the reader gets to figure 2, it makes it seem inappropriate to include AA3 as an "A" mitotype in the analyses done in figure 1. Why is C included in the paper at all and what is the relationship between mitotypes in B and C populations? Figure S1 assigns the mtDNA in AA3 to Canada, but the figure 2 alignment makes the mtDNA appear to be very similar to Benin. Does the Benin/Canada-like mitotype(s) segregate in the "A" stocks used for the Canadian and Australian cytotype? In figure S1, the A nuclear genome has more admixture from C and the A mitotype appears to have C mtDNAs. Is this from past contamination of a C/B-like mtDNA into the A source population? Is the remnant C nuclear contribution to the A nuclear introgression lines what randomly remains after introgression or is it potentially selected for by the presence of the B/C mtDNA? Is this from evolutionary history in the source populations in the lab or from the history of introgression during the experiment? I write B/C-like because it was not clear from the combination of information from fig 2 sequence alignment and the fig S1 structure analyses what is the source of the AA3 mtDNA. Maybe a mtDNA haplotype map might help here?

One could imagine a more streamlined presentation of the data where the mtDNA genotyping is presented first with all subsequent analysis presented using the sequence-based mitonuclear genotypes rather than the AA,AB,BA,BB categorization.

Were these cytotypes cleared of or checked for the presence of wolbachia?

The manuscript describes mito-nuc-diet quantitative genetic variation for a number of traits related to fitness in flies. This is an important area of research and because of what appears to be a Benin-like mtDNA haplotype segregating in the Australian population (or the original lab stocks), there is the ability to identify a candidate mitochondria SNP in a mt-ribosomal RNA. This is an interesting result. As pointed out by reviewer 2 from review commons, this adds to a growing number of studies in flies that demonstrate mito-nuc-diet interactions as an important source of variation for phenotypes, including reproduction, development and survival. A strength of the design is the ability to contrast chronic versus parental diet effect, which is novel, particularly how this implicates mito-nuclear-diet interactions as important for provisioning gametes. This could be discussed more.

To treat this three-way interaction as "unprecedented" incorrectly represents the literature and the paper ultimately misses the opportunity to focus on the interesting biology of the candidate SNP that is identified. I fail to see the distinction that you make in response to reviewer #2 from reviewer commons that other studies have "focused on physiology and evolution, and we are not aware of prior studies that have shown that mitonuclear incompatibility is diet-dependent." Other studies in flies, including some that you cite, have shown diet can modify mito-nuclear incompatibilities that cause sterility and impact egg-to-adult survival. The diet manipulations in other studies are no less specific than those used in this study, just different.

The paper could be more focused on the past literature and the underlying biology -- how a mitochondrial SNP putatively affecting mitochondrial protein translation may interact with nuclear genome and the environment to affect gametogenesis and development. There is supporting research in this area. The discussion is framed around regulatory versus coding effects. lrRNA SNPs are not typically described as regulatory. The lrRNA function in the ribosome to catalyze protein synthesis. SNPs in these genes have the potential to affect the function of the molecule that they encode. The framing with the omnigenetic model also seems out of place. A single mtDNA SNP with detectable effects on fitness is not exactly the type of variation that the omnigenetic model describes. Although, clearly any GxGxE interaction indicates variation that does not have deterministic effects on fitness and thus may be a source of complex genotype-phenotype relationships.

The main text has no methods and the supplement is extensive, making it hard for the reader to find the key methodological and results needed to understand the main text.

Additional comments:

The questions in the abstract are not clear. I do not know what is meant by: "Are these "mitonuclear" effects deterministic with regard to optimal nutrition?" Very few genetic effects are deterministic; GxGxE variation is by nature not particularly deterministic, as they are genetic effects that are modifiable by environmental context. Or is the question whether mitonuclear genotype determine optimal nutrition, which I also find to be an odd question.

I fail to understand the repeatability = deterministic argument and this seems like a contrived framing of the introduction.

It was not clear what is meant by "the majority of mitotype diversity was represented by populations bearing only Australian or Beninese mitochondria." Figure S1 did not clarify this for me. What do you think is the source of the Australian population having a Benin (or Canada?)-like mtDNA?

Line 126, I suggest including why you have excluded a genotype from an analysis in the main text rather than the supplement.

Line 129-130, it is unclear whether diet is being considered as a categorical factor with caloric density as a co-factor. Are any of the diet modifications isocaloric? Another statistical question is how vial is being accommodated in the analyses. I was surprised by how low the p-values are in the tables given the means and variances in the fig 1 and 2 plots. Are individual vials in the experiment being treated as random factors such that individuals are not the unit of replication?

Line 142-144, the language of genetic replacement is out of place. The experiment involves a reciprocal set of mitonuclear genotypes made through introgression and not a design where you identified a genetic effect and then did a replacement to rescue the effect. I recommend sticking to the language of epistasis (GxG) where a particular phenotype is observed only in one combination of two-locus introgression genotypes.

Line 145-146, I fail to see the distinction being made here between some of the studies you cite and this study. Some of the studies that you cite provide evidence that mitonuclear incompatibilities interact with diet to impact reproductive traits, which is the same fitness trait in this study. Other studies (I think not cited) have shown effects of diet on egg-to-adult survival in mitonuclear genotypes.

Line 197-198, I do not understand what is meant by "because introgression maintained variation and therefore phenotypic variation could be associated with preceding sequence data."

Line 202-213 "how the clustering grouped populations (Figure 2A). This separated A mitotypes from B, validating our initial approach of encoding mitotype by geographic origin (Figures 2, 3)"; was a single female used to establish the two mtDNA's in the study? Are the populations segregating mtDNA variation? This should be more clearly explained in the results (or the methods provided before the results). The statement "these mitotypes were not nested within nucleotype" indicates that the designation of AA, AB, BA, and BB is by geography, but does not reflect genotype? I was surprised by this. Lines 211-213 sounds as if you have genotype data for each individual that was phenotyped; is this the case? Or, rather, do the replicate populations have a single mitonucleotype (one of the 8?)?

Lines 215-217, Why is this described as having a qualitative effect? Was it not statistically significant quantitative effect? The main text should provide statistical evidence for a nucleotype x C/T mtDNA genotype x diet interaction to support the conclusions that you draw.

Figure 1. If the three replicate populations are segregating different mtDNA (as I infer from the 5,6,7,8 mitotypes results section and figure 2 and S1), then I am not sure why categorizing the replicates as having A or B mitochondria is valid. It is hard to see how AA3 is such an outlier in the PC clustering given the data in plot 1B. From the phenotypes in 1B, the three replicate populations in the AA category behave similar to each other.

PLoS Biol. 2023 Aug 21;21(8):e3002218. doi: 10.1371/journal.pbio.3002218.r005

Author response to Decision Letter 1

8 Jan 2023

Attachment

Submitted filename: response to reviews.pdf

Click here for additional data file.^{(119.8KB, pdf)}

PLoS Biol. doi: 10.1371/journal.pbio.3002218.r006

Decision Letter 2

Ines Alvarez-Garcia

24 Mar 2023

Dear Adam,

Thank you for your patience while we considered your revised manuscript entitled "Mitonuclear interactions dictate both direct and parental effects of diet on fitness, and involve a SNP in mitoribosomal 16s rRNA" for consideration as a Short Reports at PLOS Biology. Please also accept again my sincere apologies for the delay in providing you with our decision. Your revised study has now been evaluated by the PLOS Biology editors, the Academic Editor and one of the original reviewers.

The review is attached below. We are pleased to offer you the opportunity to address the remaining points raised by the reviewer in a revision that we anticipate should not take you very long. We will then assess your revised manuscript and your response to the reviewer' comments with our Academic Editor aiming to avoid further rounds of peer-review, although might need to consult with the reviewers, depending on the nature of the revisions.

We expect to receive your revised manuscript within 1 month. Please email us (plosbiology@plos.org) if you have any questions or concerns, or would like to request an extension.

At this stage, your manuscript remains formally under active consideration at our journal; please notify us by email if you do not intend to submit a revision so that we withdraw the manuscript.

**IMPORTANT - SUBMITTING YOUR REVISION**

Your revisions should address the specific points made by each reviewer. Please submit the following files along with your revised manuscript:

You should also cite any additional relevant literature that has been published since the original submission and mention any additional citations in your response.

3. Resubmission Checklist

When you are ready to resubmit your revised manuscript, please refer to this resubmission checklist: https://plos.io/Biology_Checklist

Please make sure to read the following important policies and guidelines while preparing your revision and fulfil the editorial requests:

a) *PLOS Data Policy*

We need you to provide the data underlying the graphs shown in the following figures:

Fig. 1B-E; Fig. 2C, D; Fig. 3A-C; Fig. S1A, B; Fig. S2A, C-F; Fig. S3; Fig. S5; Fig. S6B-D and Fig. S7

Please also indicate in each figure legend WHERE THE DATA CAN BE FOUND. For an example see here: http://www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.1001908#s5

b) *Published Peer Review*

https://blogs.plos.org/plos/2019/05/plos-journals-now-open-for-published-peer-review/

c) *Protocols deposition*

Sincerely,

Ines

Ines Alvarez-Garcia, PhD

Senior Editor

PLOS Biology

ialvarez-garcia@plos.org

----------------------------------------------------------------

Reviewers' comments

Rev. 3:

I appreciate the extensive revision which has improved the presentation of and made more clear aspects of the experimental design and results. Below are my detailed comments,

1) In the response to review, the authors indicate that they have downplayed determinism given the quantitative genetic nature of the variation they are describing. However, the title uses the term "dictate," which is synonymous with and perhaps even stronger than the term determine. Also, the word "determinant" is used throughout the paper including the first and last sentences of the introduction. At a minimum, can the authors please use terms more appropriate to the type of complex genetic effects that they are describing such as "impact," "affect" or "underlie"?

2) Showing that mito-nuc-diet effects are on the same order of effect size is an interesting result that I think should be highlighted in the paper. This conclusion rests on the use of partial eta-squared statistics, which I understand to be SS(of focal effect) divided by the SS(error and all other factors). In Figure 3B and C, which I think are the stronger analyses for evidencing DMN effects in the study, how can so many of the factors have a partial eta-squared statistic so close to a value of one? This is particularly evident in the analysis of development. It seems to me like the statistics should sum to something on the order of 1. The other piece of information to report in the main text along with Figure 3 would be the percent of the total variation explained by the model, which would complement the point being made about equivalent relative effect sizes while also providing the reader with a sense of how much total variation is explained by diet, genetic compartments, culture/vial effects, and interactions between factors.

Also with respect to development, how is the "index" calculated and why was development time not used as the measure? The methods indicate that every individual in a vial is being treated as an individual data point for development index? This is fine, but to avoid pseudo replication, a random vial effect needs to be included in the model to account for non-independence of individuals developing within the same culture vial.

The data seem not normally distributed with one DMN combination in each of the two feeding paradigms to be strong outlier. To what extent are the results in Fig 3 robust to versus driven by these particular combinations? These data are not included in some statistical models, but in the main text it would be good to summarize what patterns in Fig 3 are robust to versus driven by inclusion of particular genotypes in the statistical models.

Does it not cause issues with the statistical models to have redundant information within mtDNA, nuclear genome, and the random factor of genotype?

2) The last line in the abstract refers to the mt variant as a synonymous variant. It is also presented in figure 1 as a synonymous/non-coding variant. In the response to review, the authors acknowledge that the change in the rRNA is likely functional. Yet, this remains described as a synonymous or non-coding change in the abstract and at several place in the revised manuscript, including important arguments in the discussion. The synonymous/non-synonymous categorization is used to describe nucleotide changes in protein-coding genes. Non-coding is generally used to describe parts of the genome that do not encode genes (e.g., intergenic and potentially regulatory). tRNA and rRNA genes encode functional molecules, and, unlike in a protein-coding gene, we cannot assume that any variants are synonymous. Can the authors just call it a SNP in the lrRNA when they describe it and in figure 1, change the black highlighting to be nonsynonymous/RNA-encoding? I suggest updating the last line of the introduction. Also, on Line 176, this should be non-protein coding. For line 394 of discussion ... I think this non-coding variation is different than SNPs in ribosomal RNA genes; the non-coding in this sentence is referring to the observation that there is much regulatory variation such as in intergenic regions of the genome that may have small RNAs or binding sites for regulatory factors or influence chromatin structure. Maybe a strict definition on "coding" means parts of the genome that use the genetic code to make proteins, but RNA genes are genic and encode functional RNA products that are distinct from the inter-genic, non-coding and presumably regulatory parts of the genome. Lines 449-453 of discussion should be reconsidered in light of this.

3) Lines 56-61; I would remove question "A." The authors can highlight this result from this study in the results and discussion, but the prior studies cited in line 55 use designs that address this question.

4) While I understand what the authors intend by using nucleotype (to contrast with mitotype), I suggest using the term nuclear genome. My understanding of the term nucleotype is that it is used to describe ploidy variation in the nucleus (https://doi.org/10.1086/700636; https://pubmed.ncbi.nlm.nih.gov/6360135/). Mitonucleotypes could then be referred to as mitonuclear genotypes or genomes, which would be more consistent with the literature.

5) In the original manuscript's presentation of the experimental design, it was not clear that the lines were genetically diverse, so I apologize that this question comes upon review of the revised manuscript. Why were genetically diverse lines (although presumably these were still from lab culture, so not as diverse as intentionally outbred populations) used? Or is the genetic variation just the residual heterozygosity that we expect in lab cultures of flies? Lines 95-96 state that "the crossing scheme was designed to produce distinct mitochondrial backgrounds bearing equivalent pools of standing nuclear variation." What is meant by "equivalent pools of standing nuclear variation?" Do you have data indicating that the A, B, and C nuclear genotypes have the same heterozygosity? What is the motivation for incorporating this genetic variation?

In figure 2 there are many genotype-diet combinations that are producing values of zero (and I think without error bars or very little variance). If the mito-nuclear genotypes are segregating in the introgressed populations and these genotypes are associated with the phenotype, shouldn't you also sample individuals with combinations of genotypes that are non-sterile or non-lethal? I understand that the mtDNA variants are primarily, although not completely, fixed differences between populations, but if there is backcrossing to the outbred nuclear background, shouldn't you expect interacting nuclear factors to be segregating in the pool of individuals that you phenotype? Otherwise, you would need to posit that nuclear factors associated with DMN effects are fixed differences between populations? At a minimum, this should be clearly explained for readers.

6) A characterized "Dahomey" mtDNA has diet-dependent effects on fitness in D. melanogaster (DOI: 10.1371/journal.pgen.1007735) and a male-sterile Dahomey mtDNA has been well characterized by co-authors on this manuscript. What is the relationship of the Benin mtDNA major alleles in this study and these characterized Dahomey mtDNA variants?

7) Line 207 indicates that there were eight distinct mitonuclear genome combinations which are labeled in Figure 1G. The mtDNA for mitonuc combination 4 is distinct at a number of positions from both other "orange" and other "purple" labeled mtDNA. It is also only present in the "orange" nuclear background. Divergent diet-dependent phenotypes between groups 2 and 3 will show that mtDNA-diet effects are conditional on the nuclear background. Divergent diet-dependent phenotypes between groups 7 and 8 will also show that mtDNA-diet effects are conditional on the nuclear background. But, it seems that the SNP that is highlighted in the manuscript is being analyzed for effects using contrasts between groups 5,6,7 and 8. This took me quite a while to piece together going back and forth between figures 1 and 2. It would be helpful for readers to clearly present what contrasts are being used for what inference in the results section. For example, mito-nuclear genotypes 1 and 4 are being used for what inference? And is a DMN interaction specifically inferred as a diet-dependent genetic interaction using the subset of genotypes 5-8? I think this is the best way to define a DMN in this experimental design. Could patterns that support DMN effects be highlighted in figure 2 somehow? I found throughout the results section that DMN was used broadly and it was difficult for me to follow exactly what genetic-by-diet effects were being used to infer DMN effects. For example, both "diet:mitonucleotype" and "diet:mt:nuc" seem to be used equivalently as evidence of "DMNs" but I think only the latter is formally testing for diet by mtDNA by nuclear effects; if otherwise, then it should be clarified.

8) Figure 2. Are the p-values reported in Fig 2 multiple-test corrected?

9) Lines 241-247; also to clarify another reviewer's question, are parents developed on the different diets or only exposed as adults to the different diets before they lay eggs on the experimental and then standardized diets? Clarity on this allows the reader to infer if the effect of diet on parents and their progeny is potentially via effects on the development of parent gonadal tissue and/or via effects of diet on gametogenesis in parents. Is the effect of chronic versus parental diet confounded with the parental age at which an offspring was laid? I am willing to believe that effects are likely dietary, but good to mention this as a caveat.

10) In the results section, include statistical evidence (p-value, preferably multiple-test corrected) for statements such as in line 259 "mitonucleotype determined magnitude," and for the diet-by-mitonucleotype interaction in the GLM. Similar in line 262. Line 328-329 is where it would be important to report p-values for the mt-by-nuclear-by-diet interaction using the subset of genotypes 5-8.

11) Multiple places in the manuscript report that the C/T SNP is "sufficient to cause" DMN variation. The SNP seems "associated with" or "involved in" DMN interactions. With an epistatic mutation, one SNP alone is not sufficient; it seems by definition to also require variation in the nuclear genome and in dietary environment. This is more of an association study so the language of association seems more appropriate.

12) A final point is that I really don't know what to make of that AA3 mtDNA. Are the authors not perplexed by how this mtDNA from a putatively Australian lineage/culture in the lab appears recombinant with the Benin mtDNA sequence? In the geographic analysis, is it considered an Australian mtDNA even though the tree in Fig 2G groups it with the Benin mtDNA? Are the statistical results and conclusions robust to leaving this mtDNA out of the geographic analyses of if you consider this to be a Benin allele? I am really curious because it differs from Australia at many sites, and in each case it has the Benin allelic state, which is not what I would expect.

Minor comments:

Lines 585-587; this description of the diet manipulation was unclear.

Methods; I have not done a ChIP-seq experiment, so was unfamiliar how sequence data from a ChIP-seq experiment can be used for estimating allele frequencies as in a pooled-sequencing design. Readers may benefit from more details on this and whether the resulting coverage is within ranges that are considered good for estimating allele frequencies.

PLoS Biol. 2023 Aug 21;21(8):e3002218. doi: 10.1371/journal.pbio.3002218.r007

Author response to Decision Letter 2

20 Apr 2023

Attachment

Submitted filename: 2023_response to reviewers.pdf

Click here for additional data file.^{(114.7KB, pdf)}

PLoS Biol. doi: 10.1371/journal.pbio.3002218.r008

Decision Letter 3

Ines Alvarez-Garcia

28 Jun 2023

Dear Dr Dobson,

Thank you for the submission of your revised Short Report entitled "Mitonuclear interactions define both direct and parental effects of diet on fitness, and involve a SNP in mitoribosomal 16s rRNA" for publication in PLOS Biology. On behalf of my colleagues and the Academic Editor, Nick Lane, I am delighted to say that we can in principle accept your manuscript for publication, provided you address any remaining formatting and reporting issues. These will be detailed in an email you should receive within 2-3 business days from our colleagues in the journal operations team; no action is required from you until then. Please note that we will not be able to formally accept your manuscript and schedule it for publication until you have completed any requested changes.

Please take a minute to log into Editorial Manager at http://www.editorialmanager.com/pbiology/, click the "Update My Information" link at the top of the page, and update your user information to ensure an efficient production process.

PRESS

We frequently collaborate with press offices. If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximise its impact. If the press office is planning to promote your findings, we would be grateful if they could coordinate with biologypress@plos.org. If you have previously opted in to the early version process, we ask that you notify us immediately of any press plans so that we may opt out on your behalf.

We also ask that you take this opportunity to read our Embargo Policy regarding the discussion, promotion and media coverage of work that is yet to be published by PLOS. As your manuscript is not yet published, it is bound by the conditions of our Embargo Policy. Please be aware that this policy is in place both to ensure that any press coverage of your article is fully substantiated and to provide a direct link between such coverage and the published work. For full details of our Embargo Policy, please visit http://www.plos.org/about/media-inquiries/embargo-policy/.

Many congratulations and thanks again for choosing PLOS Biology for publication and supporting Open Access publishing. We look forward to publishing your study.

Sincerely,

Ines

Ines Alvarez-Garcia, PhD

Senior Editor

PLOS Biology

ialvarez-garcia@plos.org

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Text. Text A in S1 Text.

(DOCX)

Click here for additional data file.^{(75KB, docx)}

S1 Fig. Variance explained by PCA of SNPs sequenced by Pool-seq in each population.

Barplots show variance explained by (A) nuclear SNPs and (B) mitochondrial SNPs. Data underlying the graphs shown in the figure can be found in S16 and S17 Tables.

(PDF)

Click here for additional data file.^{(144.6KB, pdf)}

S2 Fig. Preliminary investigation of impacts of diet:mito:nuclear interactions in the present set of lines, and impact of a novel high-lipid diet.

(PDF)

Click here for additional data file.^{(1.9MB, pdf)}

S3 Fig. Significantly differentiated loci between mtDNAs A and B.

(PDF)

Click here for additional data file.^{(176.5KB, pdf)}

S4 Fig. Graphical representation of Drosophila melanogaster mitochondrial genome and significantly differentiated SNPs.

Red shows protein-coding genes; blue shows tRNAs; and purple shows rRNAs. Positions of significantly differentiated SNPs shown in orange, with position and alleles.

(PDF)

Click here for additional data file.^{(188.9KB, pdf)}

S5 Fig. Trait values do not correlate linearly with the caloric content of experimental media.

(PDF)

Click here for additional data file.^{(2.1MB, pdf)}

S6 Fig. Impacts of mitonucleogenotype on response to chronic and parental nutritional change, with data parsed per population in the study.

(PDF)

Click here for additional data file.^{(736.6KB, pdf)}

S7 Fig. Diet-mitonucleogenotype interactions and the architecture of phenotype.

(PDF)

Click here for additional data file.^{(294.1KB, pdf)}

S8 Fig. Structure of statistically significant differences among populations.

(PDF)

Click here for additional data file.^{(9.2MB, pdf)}

S1 Table. Rounds of introgression at different phases of the study.

(XLSX)

Click here for additional data file.^{(8.8KB, xlsx)}

S2 Table. Percentage nuclear admixture (most prevalent background) per population.

(XLSX)

Click here for additional data file.^{(9.1KB, xlsx)}

S3 Table. Significantly differentiated SNPs between mtDNAs A and B.

(XLSX)

Click here for additional data file.^{(12.8KB, xlsx)}

S4 Table. Confirmatory DMN for fecundity (GLMM ANOVA tests).

(XLSX)

Click here for additional data file.^{(9.1KB, xlsx)}

S5 Table. Major allele frequencies of significantly differentiated SNPs between mtDNAs A and B derived from 2 different rounds of sequencing (in 2016 and 2018).

(XLSX)

Click here for additional data file.^{(11.8KB, xlsx)}

S6 Table. ANOVA tests (Type III) of diet*mitonucleogenotype interactions (GLMs and Cox proportional hazards).

(XLSX)

Click here for additional data file.^{(10.5KB, xlsx)}

S7 Table. ANOVA tests (Type III) of diet*lnRNA interactions (GLMs and Cox proportional hazards).

(XLSX)

Click here for additional data file.^{(12.2KB, xlsx)}

S8 Table. ANOVA tests (Type III) of diet*mito*nuclear interactions (GLMMs and Cox mixed-effects models).

(XLSX)

Click here for additional data file.^{(12.3KB, xlsx)}

S9 Table. AIC and r² values for permuted GLMMs of diet*mito*nuclear interactions, for fecundity, progeny, and fertility in chronic and parental feeding paradigms.

(XLSX)

Click here for additional data file.^{(25.5KB, xlsx)}

S10 Table. AIC for permuted Cox mixed-effects models of diet*mito*nuclear*sex interactions, for development in chronic and parental feeding paradigms.

(XLSX)

Click here for additional data file.^{(51.4KB, xlsx)}

S11 Table. Source data Fig 1B.

(XLSX)

Click here for additional data file.^{(9.3KB, xlsx)}

S12 Table. Source data Fig 1C.

(XLSX)

Click here for additional data file.^{(10.3KB, xlsx)}

S13 Table. Source data Fig 1D.

(XLSX)

Click here for additional data file.^{(9.6KB, xlsx)}

S14 Table. Source data Fig 1E.

(XLSX)

Click here for additional data file.^{(11.1KB, xlsx)}

S15 Table. Source data Fig 3.

(XLSX)

Click here for additional data file.^{(22.6KB, xlsx)}

S16 Table. Source data S1A Fig.

(XLSX)

Click here for additional data file.^{(8.9KB, xlsx)}

S17 Table. Source data S1B Fig.

(XLSX)

Click here for additional data file.^{(9.1KB, xlsx)}

S18 Table. Source data S2A Fig.

(XLSX)

Click here for additional data file.^{(8.9KB, xlsx)}

S19 Table. Source data S2C–S2E Fig.

(XLSX)

Click here for additional data file.^{(36.4KB, xlsx)}

S20 Table. Source data S2F Fig.

(XLSX)

Click here for additional data file.^{(151.6KB, xlsx)}

S21 Table. Source data S3 Fig.

(XLSX)

Click here for additional data file.^{(14.8KB, xlsx)}

S22 Table. Source data (fecundity, progeny, fertility) Figs 2C, 2D and S5.

(XLSX)

Click here for additional data file.^{(31.6KB, xlsx)}

S23 Table. Source data (development, chronic feeding) Figs 2C, 2D and S6B–S6D.

(XLSX)

Click here for additional data file.^{(806.4KB, xlsx)}

S24 Table. Source data (development, parental feeding) Figs 2C, 2D and S6B–S6D.

(XLSX)

Click here for additional data file.^{(630.7KB, xlsx)}

S25 Table. Source data S7 Fig.

(XLSX)

Click here for additional data file.^{(11.7KB, xlsx)}

S26 Table. Source data S8 Fig.

(XLSX)

Click here for additional data file.^{(1.2MB, xlsx)}

Attachment

Submitted filename: 127022_1_rebuttal_2389737_rdsq53.docx

Click here for additional data file.^{(86.6KB, docx)}

Attachment

Submitted filename: 127022_1_rebuttal_2389737_rdsq53.docx

Click here for additional data file.^{(86.6KB, docx)}

Attachment

Submitted filename: response to reviews.pdf

Click here for additional data file.^{(119.8KB, pdf)}

Attachment

Submitted filename: 2023_response to reviewers.pdf

Click here for additional data file.^{(114.7KB, pdf)}

Data Availability Statement

Phenotype data and R code are available at github.com/dobdobby, and in supplementary materials. Sequence data are available from NCBI SRA, accession PRJNA853138.

[pbio.3002218.ref001] 1.Reed LK, Williams S, Springston M, Brown J, Freeman K, DesRoches CE, et al. Genotype-by-Diet Interactions Drive Metabolic Phenotype Variation in Drosophila melanogaster. Genetics. 2010;185:1009–1019. doi: 10.1534/genetics.109.113571 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref002] 2.Liao C-Y, Rikke BA, Johnson TE, Diaz V, Nelson JF. Genetic variation in the murine lifespan response to dietary restriction: from life extension to life shortening. Aging Cell. 2009;9:92–95. doi: 10.1111/j.1474-9726.2009.00533.x [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref003] 3.Simpson SJ, Raubenheimer D. The Nature of Nutrition: A Unifying Framework from Animal Adaptation to Human Obesity. 2012. doi: 10.1515/9781400842803 [DOI] [Google Scholar]

[pbio.3002218.ref004] 4.Boyle EA, Li YI, Pritchard JK. An Expanded View of Complex Traits: From Polygenic to Omnigenic. Cell. 2017;169:1177–1186. doi: 10.1016/j.cell.2017.05.038 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref005] 5.Zhu C-T, Ingelmo P, Rand DM. G×G×E for Lifespan in Drosophila: Mitochondrial, Nuclear, and Dietary Interactions that Modify Longevity. PLoS Genet. 2014:10. doi: 10.1371/journal.pgen.1004354 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref006] 6.Rand DM, Mossman JA. Mitonuclear conflict and cooperation govern the integration of genotypes, phenotypes and environments. Philos Trans R Soc B. 2020;375:20190188. doi: 10.1098/rstb.2019.0188 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref007] 7.Rand DM, Mossman JA, Spierer AN, Santiago JA. Mitochondria as environments for the nuclear genome in Drosophila: mitonuclear G×G×E. J Hered. 2021. doi: 10.1093/jhered/esab066 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref008] 8.Ali AT, Boehme L, Carbajosa G, Seitan VC, Small KS, Hodgkinson A. Nuclear genetic regulation of the human mitochondrial transcriptome. elife. 2019;8:e41927. doi: 10.7554/eLife.41927 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref009] 9.Ma H, Gutierrez NM, Morey R, Dyken CV, Kang E, Hayama T, et al. Incompatibility between Nuclear and Mitochondrial Genomes Contributes to an Interspecies Reproductive Barrier. Cell Metab. 2016;24:283–294. doi: 10.1016/j.cmet.2016.06.012 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref010] 10.Hill GE. Reconciling the mitonuclear compatibility species concept with rampant mitochondrial introgression. Integr Comp Biol. 2019;59:912–924. doi: 10.1093/icb/icz019 [DOI] [PubMed] [Google Scholar]

[pbio.3002218.ref011] 11.Healy TM, Burton RS. Strong selective effects of mitochondrial DNA on the nuclear genome. P Natl Acad Sci U S A. 2020;117:6616–6621. doi: 10.1073/pnas.1910141117 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref012] 12.Santiago JC, Boylan JM, Lemieux FA, Gruppuso PA, Sanders JA, Rand DM. Mitochondrial genotype alters the impact of rapamycin on the transcriptional response to nutrients in Drosophila. BMC Genomics. 2021;22:213. doi: 10.1186/s12864-021-07516-2 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref013] 13.Mossman JA, Biancani LM, Zhu C-T, Rand DM. Mitonuclear Epistasis for Development Time and Its Modification by Diet in Drosophila. Genetics. 2016;203:463–484. doi: 10.1534/genetics.116.187286 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref014] 14.Rand DM, Mossman JA, Zhu L, Biancani LM, Ge JY. Mitonuclear epistasis, genotype-by-environment interactions, and personalized genomics of complex traits in Drosophila: Mitonuclear G x G x E. IUBMB Life. 2018;70:1275–1288. doi: 10.1002/iub.1954 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref015] 15.Montooth KL, Dhawanjewar AS, Meiklejohn CD. Temperature-sensitive reproduction and the physiological and evolutionary potential for Mother’s Curse. Integr Comp Biol. 2019;59:890–899. doi: 10.1093/icb/icz091 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref016] 16.Camus MF, O’Leary M, Reuter M, Lane N. Impact of mitonuclear interactions on life-history responses to diet. Philos Trans R Soc B. 2020;375:20190416. doi: 10.1098/rstb.2019.0416 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref017] 17.Deas JB, Blondel L, Extavour CG. Ancestral and offspring nutrition interact to affect life-history traits in Drosophila melanogaster. Proc R Soc B. 2019;286:20182778. doi: 10.1098/rspb.2018.2778 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref018] 18.Vaught RC, Voigt S, Dobler R, Clancy DJ, Reinhardt K, Dowling DK. Interactions between cytoplasmic and nuclear genomes confer sex-specific effects on lifespan in Drosophila melanogaster. J Evol Biol. 2020;33:694–713. doi: 10.1111/jeb.13605 [DOI] [PubMed] [Google Scholar]

[pbio.3002218.ref019] 19.Bradburd GS, Coop GM, Ralph PL. Inferring Continuous and Discrete Population Genetic Structure Across Space. Genetics. 2018;210:33–52. doi: 10.1534/genetics.118.301333 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref020] 20.Camus MF, Wolff JN, Sgrò CM, Dowling DK. Experimental Support That Natural Selection Has Shaped the Latitudinal Distribution of Mitochondrial Haplotypes in Australian Drosophila melanogaster. Mol Biol Evol. 2017;34:2600–2612. doi: 10.1093/molbev/msx184 [DOI] [PubMed] [Google Scholar]

[pbio.3002218.ref021] 21.Bergland AO, Tobler R, González J, Schmidt P, Petrov D. Secondary contact and local adaptation contribute to genome-wide patterns of clinal variation in Drosophila melanogaster. Mol Ecol. 2016;25:1157–1174. doi: 10.1111/mec.13455 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref022] 22.Grandison RC, Piper MD, Partridge L. Amino-acid imbalance explains extension of lifespan by dietary restriction in Drosophila. Nature. 2009;462:1061–1064. doi: 10.1038/nature08619 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref023] 23.Aw WC, Towarnicki SG, Melvin RG, Youngson NA, Garvin MR, Hu Y, et al. Genotype to phenotype: Diet-by-mitochondrial DNA haplotype interactions drive metabolic flexibility and organismal fitness. PLoS Genet. 2018;14:e1007735. doi: 10.1371/journal.pgen.1007735 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref024] 24.Meiklejohn CD, Holmbeck MA, Siddiq MA, Abt DN, Rand DM, Montooth KL. An Incompatibility between a Mitochondrial tRNA and Its Nuclear-Encoded tRNA Synthetase Compromises Development and Fitness in Drosophila. PLoS Genet. 2013;9:e1003238. doi: 10.1371/journal.pgen.1003238 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref025] 25.Dearden L, Ozanne S. You are what your mother ate: The effects of maternal obesity during pregnancy on offspring obesity and cardiovascular diseases. Physiology News. 2017:26–28. doi: 10.36866/pn.106.26 [DOI] [Google Scholar]

[pbio.3002218.ref026] 26.Camilleri T, Piper MDW, Robker RL, Dowling DK. Maternal and paternal sugar consumption interact to modify offspring life history and physiology. Funct Ecol. 2022. doi: 10.1111/1365-2435.14024 [DOI] [Google Scholar]

[pbio.3002218.ref027] 27.Camilleri-Carter T-L, Dowling DK, Robker RL, Piper MDW. Transgenerational Obesity and Healthy Aging in Drosophila. J Gerontol A Biol Sci Med Sci. 2019. doi: 10.1093/gerona/glz154 [DOI] [PubMed] [Google Scholar]

[pbio.3002218.ref028] 28.Searle SR, Speed FM, Milliken GA. Population Marginal Means in the Linear Model: An Alternative to Least Squares Means. Am Stat. 1980;34:216–221. doi: 10.1080/00031305.1980.10483031 [DOI] [Google Scholar]

[pbio.3002218.ref029] 29.Hill GE, Havird JC, Sloan DB, Burton RS, Greening C, Dowling DK. Assessing the fitness consequences of mitonuclear interactions in natural populations. Biol Rev. 2018;94:1089–1104. doi: 10.1111/brv.12493 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref030] 30.Friedman H. Simplified Determinations of Statistical Power, Magnitude of Effect and Research Sample Sizes. Educ Psychol Meas. 1982;42:521–526. doi: 10.1177/001316448204200214 [DOI] [Google Scholar]

[pbio.3002218.ref031] 31.Ben-Shachar M, Lüdecke D, Makowski D. effectsize: Estimation of Effect Size Indices and Standardized Parameters. J Open Source Softw. 2020;5:2815. doi: 10.21105/joss.02815 [DOI] [Google Scholar]

[pbio.3002218.ref032] 32.Mathieson I. The omnigenic model and polygenic prediction of complex traits. Am J Hum Genet. 2021;108:1558–1563. doi: 10.1016/j.ajhg.2021.07.003 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref033] 33.Barreto FS, Burton RS. Evidence for Compensatory Evolution of Ribosomal Proteins in Response to Rapid Divergence of Mitochondrial rRNA. Mol Biol Evol. 2013;30:310–314. doi: 10.1093/molbev/mss228 [DOI] [PubMed] [Google Scholar]

[pbio.3002218.ref034] 34.Hoekstra LA, Siddiq MA, Montooth KL. Pleiotropic Effects of a Mitochondrial–Nuclear Incompatibility Depend upon the Accelerating Effect of Temperature in Drosophila. Genetics. 2013;195:1129–1139. doi: 10.1534/genetics.113.154914 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref035] 35.Hoekstra LA, Julick CR, Mika KM, Montooth KL. Energy demand and the context-dependent effects of genetic interactions underlying metabolism. Evol Lett. 2018;2:102–113. doi: 10.1002/evl3.47 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref036] 36.Piper MD, Blanc E, Leitão-Gonçalves R, Yang M, He X, Linford NJ, et al. A holidic medium for Drosophila melanogaster. Nat Methods. 2014;11:100–105. doi: 10.1038/nmeth.2731 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref037] 37.Camus MF, Moore J, Reuter M. Nutritional geometry of mitochondrial genetic effects on male fertility. Biol Lett. 2020;16:20190891. doi: 10.1098/rsbl.2019.0891 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref038] 38.Clancy DJ, Hime GR, Shirras AD. Cytoplasmic male sterility in Drosophila melanogaster associated with a mitochondrial CYTB variant. Heredity. 2011;107:374–376. doi: 10.1038/hdy.2011.12 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref039] 39.Patel MR, Miriyala GK, Littleton AJ, Yang H, Trinh K, Young JM, et al. A mitochondrial DNA hypomorph of cytochrome oxidase specifically impairs male fertility in Drosophila melanogaster. elife. 2016;5:e16923. doi: 10.7554/eLife.16923 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref040] 40.Filer D, Thompson MA, Takhaveev V, Dobson AJ, Kotronaki I, Green JW, et al. RNA polymerase III limits longevity downstream of TORC1. Nature. 2017;552:263. doi: 10.1038/nature25007 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref041] 41.Corrales GM, Filer D, Wenz KC, Rogan A, Phillips G, Li M, et al. Partial Inhibition of RNA Polymerase I Promotes Animal Health and Longevity. Cell Rep. 2020;30:1661–1669.e4. doi: 10.1016/j.celrep.2020.01.017 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref042] 42.Gammage PA, Moraes CT, Minczuk M. Mitochondrial Genome Engineering: The Revolution May Not Be CRISPR-Ized. Trends Genet. 2018;34:101–110. doi: 10.1016/j.tig.2017.11.001 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref043] 43.Baris TZ, Wagner DN, Dayan DI, Du X, Blier PU, Pichaud N, et al. Evolved genetic and phenotypic differences due to mitochondrial-nuclear interactions. PLoS Genet. 2017;13:e1006517. doi: 10.1371/journal.pgen.1006517 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref044] 44.Chang C-C, Rodriguez J, Ross J. Mitochondrial–Nuclear Epistasis Impacts Fitness and Mitochondrial Physiology of Interpopulation Caenorhabditis briggsae Hybrids. G3. 2016;6:209–219. doi: 10.1534/g3.115.022970 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref045] 45.Ellison CK, Burton RS. Disruption of mitochondrial function in interpopulation hybrids of Tigriopus californicus. Evolution. 2006;60:1382–1391. doi: 10.1554/06-210.1 [DOI] [PubMed] [Google Scholar]

[pbio.3002218.ref046] 46.Dobler R, Dowling DK, Morrow EH, Reinhardt K. A systematic review and meta-analysis reveals pervasive effects of germline mitochondrial replacement on components of health. Hum Reprod Update. 2018;24:519–534. doi: 10.1093/humupd/dmy018 [DOI] [PubMed] [Google Scholar]

[pbio.3002218.ref047] 47.Reed LK, Lee K, Zhang Z, Rashid L, Poe A, Hsieh B, et al. Systems Genomics of Metabolic Phenotypes in Wild-Type Drosophila melanogaster. Genetics. 2014;197:781–793. doi: 10.1534/genetics.114.163857 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref048] 48.Holland ML, Lowe R, Caton PW, Gemma C, Carbajosa G, Danson AF, et al. Early-life nutrition modulates the epigenetic state of specific rDNA genetic variants in mice. Science (New York, NY). 2016;353:495–498. doi: 10.1126/science.aaf7040 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref049] 49.Kobayashi S, Amikura R, Okada M. Presence of Mitochondrial large Ribosomal RNA Outside Mitochondria in Germ Plasm of Drosophila melanogaster. Science. 1993;260:1521–1524. doi: 10.1126/science.7684857 [DOI] [PubMed] [Google Scholar]

[pbio.3002218.ref050] 50.Pozzi A, Dowling DK. New insights into mitochondrial-nuclear interactions revealed through analysis of small RNAs. Genome Biol Evol. 2022;14:evac023–. doi: 10.1093/gbe/evac023 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref051] 51.Pozzi A, Dowling DK. Small mitochondrial RNAs as mediators of nuclear gene regulation, and potential implications for human health. BioEssays. 2021;43:e2000265. doi: 10.1002/bies.202000265 [DOI] [PubMed] [Google Scholar]

[pbio.3002218.ref052] 52.Dowling DK, Williams BR, Garcia-Gonzalez F. Maternal sexual interactions affect offspring survival and ageing. J Evol Biol. 2014;27:88–97. doi: 10.1111/jeb.12276 [DOI] [PubMed] [Google Scholar]

[pbio.3002218.ref053] 53.Bass TM, Grandison RC, Wong R, Martinez P, Partridge L, Piper MDW. Optimization of Dietary Restriction Protocols in Drosophila. J Gerontol A Biol Sci Med Sci. 2007;62:1071–1081. doi: 10.1093/gerona/62.10.1071 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref054] 54.Emran S, Yang M, He X, Zandveld J, Piper MD. Target of rapamycin signalling mediates the lifespan-extending effects of dietary restriction by essential amino acid alteration. Aging. 2014;6:390–398. doi: 10.18632/aging.100665 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref055] 55.Dobson AJ, He X, Blanc E, Bolukbasi E, Feseha Y, Yang M, et al. Tissue-specific transcriptome profiling of Drosophila reveals roles for GATA transcription factors in longevity by dietary restriction. NPJ Aging Mech Dis. 2018:4. doi: 10.1038/s41514-018-0024-4 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref056] 56.Woodcock KJ, Kierdorf K, Pouchelon CA, Vivancos V, Dionne MS, Geissmann F. Macrophage-Derived upd3 Cytokine Causes Impaired Glucose Homeostasis and Reduced Lifespan in Drosophila Fed a Lipid-Rich Diet. Immunity. 2015;42:133–144. doi: 10.1016/j.immuni.2014.12.023 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref057] 57.Brankatschk M, Gutmann T, Knittelfelder O, Palladini A, Prince E, Grzybek M, et al. A Temperature-Dependent Switch in Feeding Preference Improves Drosophila Development and Survival in the Cold. Dev Cell. 2018;46:781–793.e4. doi: 10.1016/j.devcel.2018.05.028 [DOI] [PubMed] [Google Scholar]

[pbio.3002218.ref058] 58.Knittelfelder O, Prince E, Sales S, Fritzsche E, Wöhner T, Brankatschk M, et al. Sterols as dietary markers for Drosophila melanogaster. Biochim Biophys Acta Mol Cell Biol Lipids. 2020;1865:158683. doi: 10.1016/j.bbalip.2020.158683 [DOI] [PubMed] [Google Scholar]

[pbio.3002218.ref059] 59.Lesperance DNA, Broderick NA. Meta-analysis of Diets Used in Drosophila Microbiome Research and Introduction of the Drosophila Dietary Composition Calculator (DDCC). G3 Genes Genomes. Genetics. 2020;10:g3.401235.2020. doi: 10.1534/g3.120.401235 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref060] 60.Cosseau C, Azzi A, Smith K, Freitag M, Mitta G, Grunau C. Native chromatin immunoprecipitation (N-ChIP) and ChIP-Seq of Schistosoma mansoni: Critical experimental parameters. Mol Biochem Parasitol. 2009;166:70–76. doi: 10.1016/j.molbiopara.2009.02.015 [DOI] [PubMed] [Google Scholar]

[pbio.3002218.ref061] 61.Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. Embnet J. 2011;17:10–12. doi: 10.14806/ej.17.1.200 [DOI] [Google Scholar]

[pbio.3002218.ref062] 62.Li H, Durbin R. Fast and accurate short read alignment with Burrows–Wheeler transform. Bioinformatics. 2009;25:1754–1760. doi: 10.1093/bioinformatics/btp324 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref063] 63.Kofler R, Orozco-terWengel P, Maio ND, Pandey RV, Nolte V, Futschik A, et al. PoPoolation: A Toolbox for Population Genetic Analysis of Next Generation Sequencing Data from Pooled Individuals. PLoS ONE. 2011;6:e15925. doi: 10.1371/journal.pone.0015925 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref064] 64.Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, et al. The Sequence Alignment/Map format and SAMtools. Bioinformatics. 2009;25:2078–2079. doi: 10.1093/bioinformatics/btp352 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref065] 65.McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A, et al. The Genome Analysis Toolkit: A MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res. 2010;20:1297–1303. doi: 10.1101/gr.107524.110 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref066] 66.Okonechnikov K, Conesa A, García-Alcalde F. Qualimap 2: advanced multi-sample quality control for high-throughput sequencing data. Bioinformatics. 2016;32:292–294. doi: 10.1093/bioinformatics/btv566 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref067] 67.Kapun M, Barrón MG, Staubach F, Obbard DJ, Wiberg RAW, Vieira J, et al. Genomic Analysis of European Drosophila melanogaster Populations Reveals Longitudinal Structure, Continent-Wide Selection, and Previously Unknown DNA Viruses. Mol Biol Evol. 2020;37:2661–2678. doi: 10.1093/molbev/msaa120 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref068] 68.Frichot E, François O. LEA: An R package for landscape and ecological association studies. Methods Ecol Evol. 2015;6:925–929. doi: 10.1111/2041-210x.12382 [DOI] [Google Scholar]

[pbio.3002218.ref069] 69.Comeron JM, Ratnappan R, Bailin S. The Many Landscapes of Recombination in Drosophila melanogaster. PLoS Genet. 2012;8:e1002905. doi: 10.1371/journal.pgen.1002905 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref070] 70.Kofler R, Pandey RV, Schlötterer C. PoPoolation2: identifying differentiation between populations using sequencing of pooled DNA samples (Pool-Seq). Bioinformatics. 2011;27:3435–3436. doi: 10.1093/bioinformatics/btr589 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref071] 71.Hudson RR, Slatkin M, Maddison WP. Estimation of levels of gene flow from DNA sequence data. Genetics. 1992;132:583–589. doi: 10.1093/genetics/132.2.583 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pbio.3002218.ref072] 72.Benjamini Y, Hochberg Y. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. J R Stat Soc Ser B Methodol. 1995;57:289–300. doi: 10.1111/j.2517-6161.1995.tb02031.x [DOI] [Google Scholar]

[pbio.3002218.ref073] 73.Cohen J. Eta-Squared and Partial Eta-Squared in Fixed Factor Anova Designs. Educ Psychol Meas. 1973;33:107–112. doi: 10.1177/001316447303300111 [DOI] [Google Scholar]

PERMALINK

Mitonuclear interactions shape both direct and parental effects of diet on fitness and involve a SNP in mitoribosomal 16s rRNA

Adam J Dobson

Susanne Voigt

Luisa Kumpitsch

Lucas Langer

Emmely Voigt

Rita Ibrahim

Damian K Dowling

Klaus Reinhardt

Roles

Abstract

Introduction

Results

Establishing and sequencing a panel of diverse mitonucleogenotypes

Fig 1. A panel of diverse mitonucleogenotypes in D. melanogaster: Population setup and grouping according to mitochondrial and nuclear SNPs.

Phenotyping

Fig 2. Mitonucleogenotypes modulate multitrait responses to chronic and parental nutritional variation and show that a single C/T polymorphism in a subset of populations is sufficient to induce diet-mito-nuclear variation.

Phenotyping mt:lrRNA SNP

Effect size calculations

Fig 3. Effect size calculations reveal substantial modification of response to diet by mt:lrRNA.

Discussion

Materials and methods

Flies

Diets

Characterizing impact of high-lipid diet

Genome sequencing

Nuclear genome analysis

Mitochondrial genome analysis

Quantitative trait analysis

Graphics

Supporting information

Acknowledgments

Abbreviations

Data Availability

Funding Statement

References

Author response to previous submission

Decision Letter 0

Ines Alvarez-Garcia

Roles

Author response to Decision Letter 0

Decision Letter 1

Ines Alvarez-Garcia

Roles

Author response to Decision Letter 1

Decision Letter 2

Ines Alvarez-Garcia

Roles

Author response to Decision Letter 2

Decision Letter 3

Ines Alvarez-Garcia

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases