Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Jul 12.
Published in final edited form as: Cell Metab. 2016 Jul 12;24(1):91–103. doi: 10.1016/j.cmet.2016.06.008

Homeostatic responses regulate selfish mitochondrial genome dynamics in C. elegans

Bryan L Gitschlag 1,2, Cait S Kirby 1,3, David C Samuels 4,5, Rama D Gangula 6, Simon A Mallal 6,7,8, Maulik R Patel 1,9,*
PMCID: PMC5287496  NIHMSID: NIHMS796406  PMID: 27411011

Summary

Mutant mitochondrial genomes (mtDNA) can be viewed as selfish genetic elements that persist in a state of heteroplasmy despite having potentially deleterious metabolic consequences. We sought to study regulation of selfish mtDNA dynamics. We establish that the large 3.1kb deletion-bearing mtDNA variant uaDf5 is a selfish genome in Caenorhabditis elegans. Next, we show that uaDf5 mutant mtDNA replicates in addition to, not at the expense of, wildtype mtDNA. These data suggest existence of homeostatic copy number control that is exploited by uaDf5 to ‘hitchhike’ to high frequency. We also observe activation of the mitochondrial unfolded protein response (UPRmt) in uaDf5 animals. Loss of UPRmt causes a decrease in uaDf5 frequency whereas its constitutive activation increases uaDf5 levels. UPRmt activation protects uaDf5 from mitophagy. Taken together, we propose that mtDNA copy number control and UPRmt represent two homeostatic response mechanisms that play important roles in regulating selfish mitochondrial genome dynamics.

Graphical abstract

Using genetic and molecular approaches in C. elegans, Gitschlag et al. show that mutant mtDNA exploits copy number control to achieve high levels. At high levels, mutant mtDNA causes mitochondrial stress and activates UPRmt. By alleviating this stress, UPRmt promotes proliferation of mutant mtDNA.

graphic file with name nihms796406u1.jpg

Introduction

Many ecological systems are predicated upon mutualistic interactions between different organisms (Momeni et al., 2011). Similar interactions between different genetic entities can underlie fundamental aspects of cell biology. For instance, the nuclear and the mitochondrial genome (mtDNA) work in a coordinated fashion to ensure optimal cellular fitness. However, just like in ecology, mutualistic interactions between these genomes are fraught with the risk that selfish mtDNA could arise. Selfish mtDNA can be defined as mutant mtDNA variants that have a transmission advantage over wildtype mtDNA despite jeopardizing cellular fitness and even organismal survival (Clark et al., 2012). Such selfish mtDNA variants often replicate and are co-transmitted along with wildtype mtDNA in a state of heteroplasmy (Taylor et al., 2002). The presence of such mutant mtDNA at frequencies above a critical threshold can be pathogenic and is thought to be one of the causes underlying inherited and age-related degeneration and metabolic diseases (Stewart and Chinnery, 2015, Wallace and Chalkia, 2013).

Selfish mtDNA are most studied in yeast, where they arise at very high rates and result in formation of ‘petite’ colonies with mitochondrial respiration failure (Bernardi, 2005, Williamson, 2002). Despite harboring major deletions and rearrangements, and causing severe cell growth defects, many of these mutant mtDNA are able to outcompete wildtype mtDNA in yeast cells and are hence dubbed ‘hypersuppressive’ mtDNA (MacAlpine et al., 2001, Jasmin and Zeyl, 2014, Harrison et al., 2014). While these selfish mtDNA were at first thought to outcompete wildtype mtDNA by virtue of possessing more origins of replication, this interpretation has been questioned in light of new data (Williamson, 2002). In metazoans, mutant mtDNA are largely characterized by deletions rather than the massive genomic rearrangements observed in hypersuppressive mtDNA. These data suggest that the mechanisms employed by selfish mtDNA in yeast are fundamentally different than those in metazoans. A deletion-harboring mtDNA was recently discovered in natural populations of the nematode species C. briggsae. Interestingly, this mutant mtDNA variant is found across C. briggsae strains of diverse geographic origins, suggesting that it has stably persisted on evolutionary time-scales (Howe and Denver, 2008, Clark et al., 2012). The presence of this mtDNA variant decreases fecundity and pharyngeal pumping rates, suggesting that it negatively impacts organismal fitness (Estes et al., 2011). While this mutant mtDNA was shown to have a transmission advantage over wildtype mtDNA in mutation accumulation lines and in small populations, the cellular and molecular bases underlying its competitive success are not well understood (Clark et al., 2012, Phillips et al., 2015).

In this study, we establish the mutant mtDNA variant called uaDf5 as a selfish genetic element in C. elegans. uaDf5 coexists as a heteroplasmy with wildtype mtDNA despite being slightly deleterious. Next, using droplet digital PCR (ddPCR) to quantify mtDNA copy number dynamics, we show the existence of homeostatic copy number control for wildtype mtDNA. Our data on uaDf5 mtDNA copy number dynamics are most consistent with uaDf5 exploiting this mtDNA copy number control to ‘hitchhike’ to high frequency. Finally, we observe activation of the mitochondrial unfolded protein response (UPRmt) in uaDf5 heteroplasmic animals. Loss of UPRmt results in a decrease in uaDf5 levels, while constitutive activation leads to its increase. Hence, besides mtDNA copy number control, uaDf5 also exploits UPRmt, together suggesting that homeostatic cellular processes are important determinants of selfish mtDNA dynamics.

Results

A bona fide ‘selfish’ mtDNA in C. elegans

We focused on identifying potentially ‘selfish’ mtDNA in a genetically tractable metazoan. C. elegans is an ideal model system to study mtDNA dynamics in a multicellular organism. Besides offering a powerful genetic toolkit, C. elegans mtDNA shares many conserved features with its mammalian counterpart (Okimoto et al., 1992). First, like mammalian mtDNA, mtDNA in C. elegans is uniparentally inherited through the oocyte (Tsang and Lemire, 2002, Sato and Sato, 2011, Al Rawi et al., 2011). Second, C. elegans mtDNA encodes 12 of the 13 proteins encoded by mammalian mtDNA. Finally, like in mammals but in contrast to yeast, most large-scale mutations in C. elegans mtDNA are deletions rather than genomic rearrangements (Denver et al., 2000).

Previous studies have identified a C. elegans strain harboring mutant mtDNA called uaDf5, which has a 3.1kb deletion that removes four protein-coding genes and seven tRNAs (Fig. 1A) (Tsang and Lemire, 2002). Due to this large deletion, individuals carrying only uaDf5 mtDNA would not be expected to be viable. Indeed, animals homoplasmic for uaDf5 have not been reported (Tsang and Lemire, 2002, Liau et al., 2007). Remarkably however, animals that have lost uaDf5 have also not been previously observed, even after passaging over hundreds of generations (Tsang and Lemire, 2002, Liau et al., 2007). Moreover, uaDf5 levels steadily increase in individuals that inherit it at a low frequency (Tsang and Lemire, 2002).

Figure 1. Mutant mtDNA uaDf5 can be forced out from a stably persisting heteroplasmy in C. elegans.

Figure 1

Open in a new tab

(A) Schematic of C. elegans mtDNA showing the uaDf5 and mptDf1 deletions (long and short red bars, respectively). Grey arrows show protein and rRNA-encoding genes and their orientation. White boxes show genes encoding tRNAs. (B) Schematic illustrating the selection strategy to force loss of uaDf5 mtDNA from a heteroplasmic C. elegans line. Each generation, the progeny of individuals with the lowest uaDf5 levels were selected for subsequent propagation. (C) Single worm PCR of wildtype and uaDf5 mtDNA. Successive propagation of individual worms with low uaDf5 levels (red boxes) results in complete loss of uaDf5 mtDNA from the population over multiple generations. (D) ddPCR data from single worms confirming complete loss of uaDf5. Positive droplets containing uaDf5-specific PCR product exhibit increased fluorescence intensity (blue) compared to negative droplets that contain no uaDf5 mtDNA (gray). For each droplet, the droplet reader detects droplet size, shape, and fluorescence intensity, and automatically distinguishes positive from negative droplets on the basis of these criteria. Sample 1, control containing uaDf5.

One explanation for the stable maintenance of uaDf5 over many generations is balanced heteroplasmy, in which two mtDNA haplotypes possess lethal but non-overlapping mutations. In this scenario, neither mtDNA type can be lost because neither mtDNA is capable of fully supporting viability. In order to test this hypothesis, we sought to sequence the entire non-uaDf5 mtDNA in the heteroplasmic animals. By designing primers inside the uaDf5 deletion, we were able to specifically amplify the entire genic region of non-uaDf5 mtDNA as two large PCR products (Fig. S1). Under the ‘balanced heteroplasmy’ hypothesis, we expected to find deleterious mutations in the non-uaDf5 mtDNA in the uaDf5 heteroplasmic individuals. In contrast, upon sequencing, we found that the non-uaDf5 mtDNA in uaDf5 heteroplasmic animals is wildtype. Next, to sequence an approximately 500bp highly AT-rich non-coding region that was not captured within the two PCR products, we amplified just the non-coding region of the mtDNA using primers that are common to both uaDf5 and non-uaDf5 mtDNA. Upon sequencing however, we did not observe any apparent heteroplasmic mutations, as we might expect if there were any mutations specific to the non-uaDf5 mtDNA. Thus, our sequencing data do not support balanced heteroplasmy as an explanation for uaDf5 persistence since the non-uaDf5 mtDNA is wildtype, suggesting that uaDf5 is not critical for organismal viability.

Our sequencing data are instead consistent with uaDf5 behaving like a selfish genetic element. We reasoned that if uaDf5 is not critical for viability, then we should be able to recover healthy individuals that have lost it. In order to test this hypothesis, we carried out additional experiments in which we selected individuals with progressively lower uaDf5 levels across multiple generations (red boxes in schematic in Fig. 1B). Under this ‘selection’ regime, we were able to recover healthy individuals homoplasmic for wildtype mtDNA that had undetectable levels of uaDf5 (Fig. 1C), suggesting that these individuals have lost uaDf5 mutant mtDNA. To confirm complete loss of uaDf5, we conducted droplet digital PCR (ddPCR) with uaDf5 specific primers. ddPCR relies on performing thousands of independent PCR reactions simultaneously in small droplets that are designed to hold an average of one template per droplet (Hindson et al., 2011). Hence, it provides a highly sensitized way to detect rare variants. We observed complete loss of uaDf5 using ddPCR (Fig. 1D). Together with our sequencing results, our finding that uaDf5 can be eliminated rules out balanced heteroplasmy as an explanation for uaDf5 maintenance. Instead, together with the previously published data that uaDf5 has a transmission advantage when inherited at low frequencies (Tsang and Lemire, 2002) and is associated with decreased organismal fitness (Liau et al., 2007), we consider the second alternative that uaDf5 is a selfish mtDNA.

uaDf5 has ‘runaway’ copy number dynamics while wildtype mtDNA levels are tightly controlled

Although uaDf5 mtDNA levels are variable across individuals, previous reports have suggested they can reach high frequency in individuals (Tsang and Lemire, 2002). Indeed, we were able to stably maintain laboratory populations in which individual worms are heteroplasmic for uaDf5 mtDNA, where uaDf5 comprises 60-80% of overall mtDNA present in each worm (Fig. 2A). To ascertain mechanisms that regulate uaDf5 levels, we next examined how levels of mutant uaDf5 mtDNA affected the levels of wildtype mtDNA. One simple ‘direct competition’ model suggests that wildtype mtDNA levels should be inversely proportional to uaDf5 mtDNA levels, while total mtDNA levels remain consistent. Alternatively, we might find that the copy number of wildtype mtDNA is maintained independent of uaDf5 levels.

Figure 2. Quantification of mtDNA copy number dynamics reveals mtDNA copy number control.

Figure 2

Open in a new tab

(A) Histogram showing uaDf5 frequency (%) distribution in individuals from a population stably maintaining uaDf5 mtDNA. Heteroplasmy frequency was determined using ddPCR to quantify wildtype and uaDf5 mtDNA copy number in single individuals. (B) mtDNA levels in individual day 4 adult worms, normalized to actin and rank-ordered by uaDf5 mtDNA copy number. (C) Wider variation in uaDf5 relative to wildtype copy number (p<0.05) suggests that wildtype mtDNA, but not uaDf5 mtDNA, is subject to homeostatic copy number control. Grey data points show mtDNA copy number from single individuals. Box and whisker plot shows the median, lower and upper quartile (boxes), and minimum and maximum (error bars) mtDNA copy number. (D) mptDf1 frequency distribution obtained from single individuals from a population stably maintaining mptDf1 heteroplasmy. (E) mtDNA levels in individual L4 worms, normalized to actin and rank-ordered by mptDf1 copy number. (F) Similar to uaDf5, wider variation in mptDf1 relative to wildtype copy number (p<0.05) suggests that wildtype mtDNA, but not mptDf1 mtDNA, is subject to homeostatic copy number control. AU, arbitrary units.

Here, we sought to distinguish between these possibilities of maintenance of total mtDNA copies versus wildtype mtDNA copies. Using ddPCR, we quantified mtDNA copy number from individual day 4 adults. mtDNA replication in C. elegans does not start until the L4 stage, after which it greatly increases during the first few days of adulthood and reaches steady state levels by day 4 (Fig. S2) (Bratic et al., 2009). mtDNA copy number measurements in day 4 adults reflect germline mtDNA since more than 95% of the mtDNA content in an adult hermaphrodite is in the germline syncytium, the shared cytoplasm of germ cells (Bratic et al., 2009). The ‘direct competition’ model predicts relatively constant levels of total mtDNA amongst individuals, with a trade-off between the amount of wildtype and uaDf5 mtDNA copies. In contrast, our results show that wildtype mtDNA levels are remarkably consistent across individuals while uaDf5 levels span a wide range. These findings are especially pronounced when animals are rank-ordered from lowest to highest uaDf5 levels (Fig. 2B). Analysis of the range of wildtype and uaDf5 mtDNA copy number shows substantially wider range in uaDf5 mtDNA copy number compared to wildtype mtDNA (Fig. 2C). Taken together, our data show that uaDf5 mtDNA levels increase in addition to, not at the expense of, wildtype mtDNA levels. These data also provide an explain for previously published data showing that total mtDNA levels in uaDf5 heteroplasmic animals are higher than in homoplasmic wildtype animals (Tsang and Lemire, 2002). In conclusion, our data are consistent with the hypothesis that wildtype mtDNA levels, but not the total mtDNA levels, are well regulated.

Small deletion bearing mptDf1 mtDNA also has ‘runaway’ copy number dynamics

One attractive hypothesis for maintenance of mutant mtDNA with large deletions like uaDf5 invokes replicative advantage over wildtype mtDNA because of their smaller genome size (Wallace, 1992). The replicative advantage of a smaller genome has been shown under limited physiological contexts (Moraes et al., 1999, Diaz et al., 2002); nevertheless, this advantage could potentially give rise to the ‘runaway’ dynamics that we observe for uaDf5. If this hypothesis were correct, we would expect to see different mtDNA dynamics in mtDNA harboring smaller deletions. To address this possibility, we investigated mtDNA copy number dynamics in animals that are heteroplasmic for mptDf1, a mutant mtDNA carrying a 179bp deletion (Fig. 1A), significantly smaller than the 3.1kb uaDf5 deletion. We identified mptDf1 deletion in a heteroplasmic strain from whole genome sequencing data of mutagenized worms from the Million Mutation Project (Thompson et al., 2013). In contrast to the uaDf5 deletion, which reduces mtDNA genome size by ∼20%, the mptDf1 deletion removes only ∼1% of the genome. Interestingly, despite backcrossing to N2 reference strain five times, we noticed that high levels of mptDf1 cause gonadal defects, and likely explain the lower average heteroplasmy frequency of mptDf1 compared to uaDf5 (Fig. S3). Despite this lower heteroplasmy frequency, and the fact that we had to use L4 stage individuals given germline defects in adults, we find that the mtDNA copy number dynamics in uaDf5 and mptDf1 heteroplasmies are remarkably similar, that is, wildtype mtDNA levels were relatively steady across individuals despite increasing levels of mutant mtDNA (Fig. 2D-F). While not ruling out replicative advantage, the data suggest that other mechanisms contribute to regulation of uaDf5 and mptDf1 copy number dynamics.

mtDNA transcriptional imbalance in uaDf5 heteroplasmic animals

Since heteroplasmic animals carry uadf5 mtDNA in addition to wildtype mtDNA, we ascertained whether the uaDf5 mtDNA contributed to the mtDNA transcripts. We reasoned that expression of all genes from the wildtype mtDNA but only some from the uaDf5 genome would result in a stoichiometric imbalance in mtDNA-encoded transcript levels (see schematic Fig. 3A). We found that CYTB and ND1, genes deleted from uaDf5, were expressed at the same levels in homoplasmic wildtype animals as in uaDf5 heteroplasmic animals (Fig. 3B). Transcript levels of the nuclear-encoded electron transport chain component NUO2, as well as actin, were also unaltered in the heteroplasmic animals (Fig. 3B). However, compared to homoplasmic wildtype animals, transcript levels of COI, COII, COIII, ND4, ND5, and ND6, which are still encoded by uaDf5 mtDNA, were significantly elevated in the uaDf5 heteroplasmic population (Fig. 3B). This implies that uaDf5 mtDNA are transcriptionally active and their expression contributes to substantial transcriptional imbalances in mtDNA-encoded genes.

Figure 3. Mitochondrial function is perturbed in uaDf5 animals.

Figure 3

Open in a new tab

(A) Schematic showing expected expression of mtDNA-encoded transcripts. The presence of uaDf5 mtDNA is expected to result in stoichiometric imbalance of gene expression, as the expression of uaDf5 and wildtype mtDNA copies (red and blue lines, respectively) combine to generate total expression (black line) at elevated levels for genes located outside the deletion but at wildtype levels for genes missing from the uaDf5 mtDNA. (B) Animals heteroplasmic for uaDf5 exhibit expression levels similar to that of wildtype animals for mtDNA-encoded genes affected by the deletion (CYTB and ND1), as well as a nuclear-encoded mitochondrial gene (NUO2) and actin. However, uaDf5 heteroplasmy results in overexpression for mtDNA-encoded genes located outside the uaDf5 deletion (COXI, COXII, COXIII, ND4, and ND5). All transcript levels are normalized to wildtype. Error bars represent standard deviation. (C) Mitochondrially targeted GFP (GFPmt), but not cytosolic GFP (cGFPcyt), is significantly reduced in uaDf5 heteroplasmic individuals. (D) Western blot analysis of wildtype and uaDf5 heteroplasmic animals expressing GFPmt reveals reduced levels in uaDf5 heteroplasmic individuals relative to actin. Data are shown from two biological replicates each for wildtype and uaDf5 strain. (E) Fluorescence increase in uaDf5 animals stained with mitochondrial membrane potential independent dye MitoTracker Green FM and (F) membrane potential dependent dye TMRE. Error bars represent standard deviation. AU, arbitrary units.

Mitochondrial perturbations in uaDf5 heteroplasmic animals

uaDf5 affects organismal fitness; uaDf5 animals have decreased egg-laying and defecation rates, reduced lifespan, and decreased sperm motility (Liau et al., 2007). At the molecular level, we observe overexpression of uaDf5 mtDNA-encoded transcripts, resulting in transcriptional imbalance. Given these effects, we sought to determine whether uaDf5 has cellular consequences. Mitochondrially targeted green fluorescence protein (GFPmt) has previously been used as a model matrix protein to assess mitochondrial proteostasis (Yoneda et al., 2004, Benedetti et al., 2006). We reasoned that if uaDf5 animals have altered proteostasis, it might result in decreased fluorescence of GFPmt. Indeed, we observe a significant decrease in fluorescence in uaDf5 animals that express GFPmt in the intestinal cells under the control of the ges-1 promoter (Fig. 3C) (Benedetti et al., 2006). No such decrease is observed in fluorescence of cytoplasmic GFP (GFPcyt) in uaDf5 animals (Fig. 3C). These data suggest that the fluorescence of mitochondrially targeted GFP is specifically affected in uaDf5 animals. Consistent with fluorescence data, western blot analysis shows decreased GFPmt protein levels in uaDf5 animals (Fig. 3D). In contrast to decreased GFPmt, we observe increased staining in uaDf5 animals with fluorescent dye MitoTracker Green FM, which localizes to mitochondria independent of the membrane potential (Dingley et al., 2014, Hicks et al., 2012), as well as with tetramethyl rhodamine ethyl ester (TMRE) (Dingley et al., 2014, Yoneda et al., 2004, Billing et al., 2011, Palikaras et al., 2015), which localizes to mitochondria in a membrane potential dependent manner (Fig. 3E-F). These data suggest an increase in mitochondrial organelle mass in uaDf5 animals, correlating with an increase in total mtDNA levels. These data also suggest that the decreased GFPmt levels and fluorescence in uaDf5 animals are due to alterations in mitochondrial proteostasis rather than due to decreased mitochondrial organelle mass. This altered proteostasis might reflect decreased GFPmt import efficiency into mitochondria or a compromised protein-folding environment inside the mitochondrial matrix. Either scenario might result in degradation of the unfolded/misfolded GFPmt. Taken together, these data suggest mitochondrial alterations in uaDf5 animals.

High levels of uaDf5 activate the mitochondrial unfolded protein response

The mitochondrial unfolded protein response (UPRmt) has emerged as an important protective stress response that is activated under a variety of conditions that affect mitochondrial proteostasis (Houtkooper et al., 2013, Yoneda et al., 2004, Haynes and Ron, 2010, Runkel et al., 2013, Baker et al., 2012). This homeostatic response involves expression of hundreds of target genes including chaperones and proteases that eliminate misfolded and nonfunctional complexes inside mitochondria (Nargund et al., 2015, Nargund et al., 2012). To determine whether UPRmt is induced in uaDf5 animals, we quantified hsp-6 and hsp-60 transcript levels in uaDf5 animals. HSP-6 and HSP-60 are mitochondrial chaperones induced by UPRmt (Yoneda et al., 2004, Nargund et al., 2012). We observed significant increase in hsp-6 and hsp-60 transcript levels in uaDf5 animals (Fig. 4A). We were able to confirm this induction with a transgenic fluorescence reporter in which the hsp-60 promoter drives GFP expression, although UPRmt induction is variable across individuals (Fig. 4B-C) (Yoneda et al., 2004). These data suggest that UPRmt might be appreciably induced in animals harboring uaDf5 levels above a certain threshold. Consistent with this notion, we were able to observe a weak but positive relationship between UPRmt activation and uaDf5 levels (Fig. 4D). The lack of a stronger correlation might be a product of the fact that most of the animals we analyzed have uaDf5 levels within a very narrow range of 60-80%. Taken together, these results suggest that presence of the selfish mtDNA uaDf5 induces UPRmt.

Figure 4. UPRmt is activated in heteroplasmic animals carrying uaDf5 mtDNA.

Figure 4

Open in a new tab

(A) Transcription of two UPRmt-activated molecular chaperones (hsp-60 and hsp-6) is increased in individuals with uaDf5 compared to wildtype individuals. (B) Quantification of fluorescence between wildtype homoplasmic and uaDf5 heteroplasmic animals shows increased activation of the UPRmt marker hsp-60∷GFP in the presence of uaDf5 mtDNA. Each data point is from a single individual picked randomly from a population. (C) Visual comparison of GFP fluorescence between uaDf5 and wildtype animals, each expressing hsp-60∷GFP. Wildtype animals were picked at random from a population but only uaDf5 animals with apparent fluorescence were picked to show UPRmt activation. (D) Positive relationship between uaDf5 frequency and hsp-60∷GFP fluorescence (trendline) indicates that UPRmt activation increases at higher uaDf5 frequency. Each data point corresponds to a single individual. Error bars represent standard deviation. AU, arbitrary units.

UPRmt modulates uaDf5 heteroplasmy levels

UPRmt plays a protective role under conditions that affect mitochondria (Nargund et al., 2012, Baker et al., 2012, Runkel et al., 2013). Might the protective role for UPRmt inadvertently create the conditions that allow uaDf5 to sustain high levels in certain individuals? According to this hypothesis, we predict that loss of UPRmt would result in a decrease in uaDf5 levels. To test this hypothesis, we conducted RNAi-mediated knockdown of atfs-1, a gene that encodes a transcription factor central to UPRmt induction (Nargund et al., 2012). We observed a significant decrease in the average uaDf5 frequency in atfs-1 knockdown animals (Fig. 5A). We similarly observe a decrease in uaDf5 levels in atfs-1 loss-of-function mutants (Fig. 5B). Quantification of mtDNA dynamics shows that wildtype mtDNA levels are well regulated in atfs-1 loss-of-function mutants, thus ruling out loss of mtDNA copy number control as a potential explanation for the decrease in uaDf5 levels (Fig. 5C-D, compare Fig. 2B-C). Given that mutant mtDNA molecules typically disrupt mitochondrial function and become pathogenic only when their levels are high enough to pass a critical threshold (Rossignol et al., 2003, Stewart and Chinnery, 2015), cells may become tolerant to mutant mtDNA, even with a relatively modest reduction in mutant levels, relaxing further selection against mutant mtDNA. Moreover, because UPRmt is likely induced in animals harboring uaDf5 levels above a certain threshold, and given the hypothesis that uaDf5 may exploit additional mechanisms to persist, such as mtDNA copy number control, we predict that although uaDf5 levels decrease in absence of UPRmt, the mutation would still be able to persist. Indeed, uaDf5 is still present, albeit at low levels, in atfs-1 loss-of-function animals that we have continuously maintained for more than 30 generations (Fig. 5E). These data are consistent with the hypothesis that loss of UPRmt exposes uaDf5 to more stringent selection, which causes a decrease in uaDf5 levels but not its complete elimination.

Figure 5. Loss of UPRmt activation results in decreased uaDf5 levels but does not affect mtDNA copy number control.

Figure 5

Open in a new tab

(A) Growth under RNAi-mediated knockdown of atfs-1, required for UPRmt activation, results in a shift to lower uaDf5 frequency relative to growth under control conditions (p<0.05). (B) uaDf5 frequency decreases in heteroplasmic animals homozygous for the atfs-1(tm4525) loss-of-function allele compared to heteroplasmic animals that express wildtype atfs-1, in which high uaDf5 levels are stably maintained. uaDf5 frequency decreases further in the atfs-1 null animals after multiple generations but is not lost completely. (C) Quantification of mtDNA copy number in individual day 4 adult animals homozygous for the atfs-1 loss-of-function allele, normalized to actin and rank-ordered by uaDf5 mtDNA copy number. (D) Wider variation in uaDf5 relative to wildtype copy number (p<0.05) in atfs-1 null animals suggests that mtDNA copy number control persists in absence of UPRmt. (E) PCR of single heteroplasmic individuals against the atfs-1 wildtype or atfs-1 null nuclear background shows that uaDf5 is retained in both lines but is at lower levels in the null animals after about 30 generations. Note that because mutant and wildtype templates compete for amplification, the wildtype band appears fainter when uaDf5 levels are high but does not actually reflect reduced wildtype mtDNA levels (see Fig. 2B). Error bars represent standard deviation. AU, arbitrary units.

If UPRmt activation allows for tolerance to high levels of uaDf5, then we predict that restoring atfs-1 will allow uaDf5 levels to recover after atfs-1 RNAi treatment. Indeed, uaDf5 levels recover within a single generation when atfs-1 expression is restored in uaDf5 animals after growing for several generations on atfs-1 RNAi conditions (Fig. 6A). These data suggest that UPRmt is required for uaDf5 to attain high levels. We next tested if the converse was also true, i.e., whether constitutive activation of UPRmt decreases selection against uaDf5, thereby driving uaDf5 to higher frequency. For this, we tested animals heterozygous for an atfs-1 gain-of-function allele to constitutively activate UPRmt (Rauthan et al., 2013). We did not observe any significant increase in the uaDf5 levels in animals heterozygous for the atfs-1 gain-of-function allele (Fig. 6B). However, given that the starting uaDf5 levels in our uaDf5 strain are already very high (approximately 80%), we speculated that they might not be able to increase substantially due to an upper threshold effect. It is also be possible that given the induction of UPRmt in individuals with high levels of uaDf5, the atfs-1 gain-of-function allele might not induce significant further UPRmt activation in these animals. To overcome these limitations, we tested whether uaDf5 levels can rise in atfs-1 gain-of-function heterozygotes in a population with lower starting uaDf5 frequency (approximately 30%). In this case, we observed a significant increase in uaDf5 frequency in atfs-1 gain-of-function heterozygotes relative to animals homozygous for wildtype atfs-1 (Fig. 6C). Together, our results show that atfs-1 can modulate uaDf5 levels.

Figure 6. Persistence of uaDf5 at high frequency depends in part on UPRmt activation.

Figure 6

Open in a new tab

(A) Growth under RNAi-mediated knockdown of atfs-1 across seven generations reduces average uaDf5 frequency. However, restoration of atfs-1 expression by returning atfs-1 knockdown animals to control conditions results in recovery of elevated uaDf5 frequency in a single generation. (B) When starting uaDf5 frequency is high (75-80%), constitutive UPRmt activation in individuals heterozygous for an atfs-1 gain-of-function allele causes no further rise in average uaDf5 frequency; (C) however, uaDf5 frequency rises when the atfs-1 gain-of-function allele is crossed into a strain harboring lower uaDf5 levels (∼30%). Error bars represent standard deviation.

Loss of UPRmt does not select against uaDf5 at the organismal level

Increased sensitivity of animals with high uaDf5 levels to atfs-1 loss at the organismal level provides one potential explanation for the observed decrease in uaDf5 levels. RNAi knockdown of atfs-1 enhances developmental delay in animals exposed to paraquat (Runkel et al., 2013). Mitochondrial-stressed isp-1(qm150) and clk-1(qm30) mutants similarly fail to develop under atfs-1 RNAi conditions (Nargund et al., 2012). We sought to determine whether knockdown of atfs-1 RNAi causes similar developmental delay in uaDf5 animals. While we observe a mild developmental delay in uaDf5 animals, it is not enhanced by atfs-1 knockdown (Fig. 7A). We also did not observe appreciable levels of embryonic lethality in uaDf5 animals raised on atfs-1 RNAi (Fig. 7B). Finally, there is no increase in lethality up to day 4 of adulthood (Fig. 7C), which is when we assess and observe a decrease in uaDf5 levels on atfs-1 RNAi. The absence of an atfs-1 knockdown-dependent effect on reproduction suggests that selection against high uaDf5 levels is unlikely to operate at the organismal level.

Figure 7. UPRmt protects uaDf5 from mitophagy.

Figure 7

Open in a new tab

(A) Heteroplasmic individuals exhibit delayed growth: as 100% of progeny from wildtype parents reach adulthood in three days, approximately 10% of uaDf5 progeny remain in the larval stage. Knockdown of atfs-1 showed no effect on development in homoplasmic wildtype animals and did not further enhance developmental delay in uaDf5 heteroplasmic animals. (B) No significant difference was observed between uaDf5 and wildtype animals, or between atfs-1 knockdown and control conditions, on the percentage of embryos that remain unhatched after one day or (C) on the percentage of lethality among day 4 adults. (D) Quantification of Pink-1∷GFP fluorescence shows increased mitophagy in uaDf5 animals upon pdr-1;atfs-1 double knockdown compared to knockdown of pdr-1 alone. AU, arbitrary units. (E) Crossing scheme employed to isolate uaDf5 animals in wildtype, atfs-1 null, pdr-1 null, and atfs-1;pdr-1 double mutant backgrounds. (F) Quantification of uaDf5 levels shows recovery of uaDf5 levels in atfs-1;pdr-1 double mutants compared to atfs-1 single mutant animals. uaDf5 recovers to the highest levels in pdr-1 single mutants. Error bars represent standard deviation. AU, arbitrary units.

UPRmt modulates uaDf5 levels via mitophagy

Organelle level selection provides an alternate possibility for the observed decrease in uaDf5 levels in absence of atfs-1. According to this hypothesis, uaDf5 might be more susceptible to mitophagy in absence of UPRmt. Mitophagy is initiated by stabilization of Pink-1 on dysfunctional mitochondria, which in turn recruits parkin to mediate mitophagy (Randow and Youle, 2014). Here, we observe an increase in Pink-1∷GFP fluorescence in uaDf5 animals with RNAi knockdown of atfs-1 compared to control (Fig. 7D; these experiments were conducted in parkin homolog pdr-1 RNAi background to stabilize Pink-1∷GFP localization to mitochondria). The increase in Pink-1∷GFP fluorescence suggests that loss of UPRmt promotes mitophagy in uaDf5 animals. If mitophagy mediates decrease in uaDf5 levels in absence of UPRmt, then we predict that uaDf5 levels should recover in animals without UPRmt that also have compromised mitophagy. In order to test this prediction, we quantified uaDf5 levels in atfs-1;pdr-1 double mutants (Fig. 7E-F). We observed significant recovery of uaDf5 levels in atfs-1;pdr-1 double mutant animals compared to atfs-1 single mutants. Relative to the wildtype nuclear background, the highest uaDf5 levels were observed in pdr-1 single mutants (Fig. 7F), consistent with the recently published observation that uaDf5 frequency increases in pdr-1 mutants (Valenci et al., 2015). In summary, our genetic analyses indicate that UPRmt facilitates high levels of uaDf5 by protecting the mutant genome from mitophagy, with loss of UPRmt exposing it to mitophagy.

Discussion

Mechanisms that regulate selfish mutant mtDNA dynamics are poorly understood. Here, we address this issue by establishing that an mtDNA mutation in C. elegans, uaDf5, behaves as a selfish genetic element. We show that the ‘runaway’ copy number dynamics of uaDf5 are consistent with it exploiting the host's homeostatic mtDNA copy number control mechanism to achieve high frequency. We further show that the UPRmt, a mitochondrial stress response system, is activated in animals with high uaDf5 levels. uaDf5 frequency decreases in absence of UPRmt, suggesting that the uaDf5 mutant also exploits UPRmt to persist at high levels. Thus, uaDf5 can be viewed as exploiting the very homeostatic responses that its presence activates to persist and proliferate (Fig. S4). Such exploitation of endogenous pathways provides a potential explanation for how mutant mtDNA are able to achieve high frequency to cause disease in humans.

mtDNA copy number control

Because mitochondrial respiration depends on the expression of mtDNA-encoded proteins, mtDNA copy number can serve as a proxy indicator of metabolic capacity. Furthermore, altered mtDNA copy number is implicated in numerous human diseases and prognosis (Clay Montier et al., 2009, Reznik et al., 2016). Thus, maintenance of mtDNA copy number is a subject of great research interest, and several models have been proposed. One study found that mtDNA levels in cultured human cells depend on the number of replication origins in mtDNA molecules, with replication favoring mutant molecules that contain a larger number of replication origins (Tang et al., 2000a).

Conversely, comparison of mutant and wildtype mtDNA content in homoplasmic cell lines revealed that total mtDNA mass was constant even when mutant and wildtype genomes differed in size and in number of replication origins, suggesting a possible role for the availability of nucleotides in regulating mtDNA copy number (Tang et al., 2000b). More recently, computational modeling and experimental observation support a previously proposed “maintenance of wild type” model, whereby the cell induces mtDNA replication to establish optimal mtDNA levels, which can stochastically proliferate mutant mtDNA in a heteroplasmy (Chinnery and Samuels, 1999, Capps et al., 2003, Durham et al., 2007, Tam et al., 2015).

Our investigation of mtDNA copy number dynamics in a uaDf5 heteroplasmic C. elegans strain shows regulation of wildtype mtDNA levels but ‘runaway’ dynamics of uaDf5 levels. We find that our data are most consistent with the “maintenance of wild type” model, which requires a feedback mechanism whereby mtDNA replication is inversely related to wildtype mtDNA levels, allowing cells to maintain mtDNA at optimal levels (Fig S5). When mutant mtDNA fail to support normal mitochondrial function, insufficient levels of wildtype mtDNA in a heteroplasmy induce mtDNA replication. However, random sampling by the replication machinery can lead to replication of mutant mtDNA, requiring additional rounds of replication and, in turn, production of more mutant mtDNA. Thus, this negative feedback model of copy number homeostasis predicts ‘runaway’ dynamics of mutant mtDNA, exactly like those observed for uaDf5. Based on these data, uaDf5 can be viewed as taking advantage of homeostatic mtDNA copy number control to hitchhike to high frequency. Similar over-proliferation of mtDNA is observed in individuals with heteroplasmic mtDNA diseases (Durham et al., 2007). Taken together, these data suggest that exploitation of homeostatic mtDNA copy number control by mutant mtDNA might be a conserved and widespread strategy. Ours is the first report to show that this strategy can operate in the germline to regulate heteroplasmy transmission dynamics.

Studies in mammalian cell culture have shown that mutant mtDNA with large deletions like uaDf5 may out-compete wildtype mtDNA for replication due to their smaller genome size (Wallace, 1992, Moraes et al., 1999, Diaz et al., 2002). However, mtDNA dynamics in mptDf1 heteroplasmic animals, which harbors a very small deletion, are also like those of uaDf5. These data suggest that this mutant mtDNA also exploits mtDNA copy number control. While these data suggest that it is not necessary to invoke replicative advantage to explain uaDf5 and mptDf1 dynamics, our data do not rule out a role for such mechanisms in contributing to persistence of selfish mtDNA. Another model that ascribes an inherent advantage to mutant mtDNA was recently proposed (Kowald and Kirkwood, 2014). According to this model, the production of the output sensed to ‘count’ mtDNA might be coupled to inhibition of replication of that specific mtDNA molecule (Kowald and Kirkwood, 2014). This feature would have the effect of hastening runaway dynamics of mutant mtDNA. Our data does not exclude this possibility and it would be interesting to investigate the role of such coupling in the future.

What is the mtDNA output that is sensed to achieve mtDNA copy number control? Given that ND1 is the only gene deleted from both uaDf5 and mptDf1 mtDNA, it is possible that some output that is dependent on ND1 function is required to ‘count’ mtDNA. However, ND1 is one subunit of the large electron transport chain complex I. Hence, it is also possible that the output is produced by this complex and mutations in any one of the mtDNA encoded complex I subunits will result in the same runaway copy number dynamics. Consistent with a role of complex I, analysis of mtDNA deletions in mammals implicates components of complex I in regulating mtDNA copy number (Kowald and Kirkwood, 2014). Finally, it is also possible that production of the sensed output is dependent on the entire electron transport chain. In this instance, mutations in any of the mtDNA-encoded genes will result in failure of the mutant mtDNA to contribute to the output. Future experiments with additional heteroplasmies with mutations or deletions in different mtDNA-encoded genes will shed light on this issue. Failure to contribute to the sensed output used for mtDNA copy number control might be a common feature of selfish mtDNA.

mtDNA copy number control has to occur during periods of mtDNA replication. Previous work has shown that mtDNA replication starts in L4 stage and continues throughout adulthood (Bratic et al., 2009). This replication coincides with germline proliferation. Furthermore, there is almost no increase in mtDNA copy number during these stages of development in germline free animals, suggesting that almost all mtDNA replication occurs in the developing germline (Bratic et al., 2009). Based on this data, we suggest that mtDNA copy number control occurs in the germline syncytium either continuously from L4 stage onwards, or during adulthood after mtDNA has reached steady state levels.

Interestingly, we observe a slight increase in wildtype mtDNA levels with increasing uaDf5 levels. These data might be reflective of overcompensation in wildtype mtDNA replication. For instance, interference with wildtype mtDNA function by uaDf5-encoded products might result in an effective decrease in the number of functional wildtype mitochondria. Such interference might result in a cellular effort to increase levels of wildtype mtDNA above baseline.

UPRmt in mtDNA heteroplasmy

We observed UPRmt activation in animals with high uaDf5 levels as determined using a transgenic reporter of the hsp-60 promoter driving GFP expression. We confirmed this UPRmt activation by measuring hsp-6 and hsp-60 transcript levels. How does uaDf5 induce UPRmt? Defective protein handling inside mitochondria is believed to trigger UPRmt (Yoneda et al., 2004, Houtkooper et al., 2013, Mouchiroud et al., 2013). Here, we observe an overexpression of mtDNA-encoded transcripts that are still intact in uaDf5 mtDNA. If these transcripts are translated, they might overwhelm the protein-handling environment in the mitochondria. Alternatively, they might result in formation of stoichiometrically imbalanced non-functional electron transport chain complexes. Decrease in GFPmt levels and fluorescence in uaDf5 animals is consistent with the hypothesis that either import of GFPmt into mitochondria or its folding inside the matrix are compromised. Future experiments aimed at characterizing the details of protein-handling environment in uaDf5 animals will help determine the mechanistic basis of UPRmt induction.

UPRmt activation relies on decreased import efficiency of ATFS-1 into mitochondria. Active protein import into mitochondria is known to rely on the mitochondrial membrane potential. In uaDf5 animals, although we see UPRmt induction and a role for ATFS-1 in modulating uaDf5 heteroplasmy levels, mitochondrial membrane potential is likely not compromised. Indeed, we actually see an increase in staining of uaDf5 animals with TMRE, a mitochondrial membrane potential-dependent dye. This increase might not be reflective of an actual increase in mitochondrial membrane potential but probably results from an overall increase in mitochondrial organelle mass since we also observe an increase in staining with MitoTracker Green FM, a dye that localizes to mitochondria independent of the membrane potential. Overall, these data suggest a different molecular basis for decreased ATFS-1 import efficiency than decreased mitochondrial membrane potential. This conclusion is perhaps not surprising given that ATFS-1 import efficiency has not been linked to decreased membrane potential (Nargund et al., 2012).

Our data show that UPRmt is an important regulator of uaDf5 levels. Why does loss of UPRmt result in reduced uaDf5 levels? We suggest that UPRmt protects uaDf5 against selection, with loss of UPRmt sensitizing the cell and exposing uaDf5 to selection. Selection can operate at the organismal level, which would result in elimination of animals with high uaDf5 levels either through developmental delay or lethality. While we observe a slight developmental delay in uaDf5 animals, this delay is not enhanced by atfs-1 RNAi. We also did not observe enhancement in embryonic lethality or increased death up till day 4 of adulthood when we assessed decreased uaDf5 levels in atfs-1 RNAi. These data suggest that UPRmt protects uaDf5 against selection, perhaps at the organelle level. Under this hypothesis, mitochondria harboring high levels of uaDf5 might be more susceptible to mitophagy in absence of UPRmt. In support of this hypothesis, we observe an increase in Pink-1∷GFP fluorescence in uaDf5 animals without the ability to induce UPRmt. We also observed recovery of uaDf5 levels in atfs-1;pdr-1 double mutants that lack the ability to induce UPRmt and mitophagy. Interestingly, pdr-1 mutants exhibit uaDf5 levels higher than either the wildtype controls or the atfs-1;pdr-1 double mutants, consistent with the recently published data that loss of the parkin homolog pdr-1 in C. elegans results in up-regulation of uaDf5 frequency (Valenci et al., 2015). Furthermore, overexpression of parkin is reported to result in selection against mutant mtDNA in a mammalian cell line (Suen et al., 2010). Together, the observation that uaDf5 levels recover in atfs-1;pdr-1 double mutants suggests that UPRmt protects selfish mtDNA like uaDf5 from organelle level selection, while the increased uaDf5 frequency in pdr-1 single mutants with intact UPRmt likely suggests additional roles for UPRmt in promoting the propagation of mutant mtDNA. These findings are consistent with recent evidence suggesting an atfs-1-dependent role for mitochondrial biogenesis, as well as organelle fusion-fission dynamics, in modulating mutant mtDNA levels (Lin et al., 2016).

Many mitochondrial stressors, such as exposure to paraquat or disruption of electron transport chain function, induce UPRmt (Runkel et al., 2013, Durieux et al., 2011). It will be interesting to determine whether uaDf5 levels are modulated under these conditions. Severe mitochondrial stressors might sensitize uaDf5 animals, or trigger additional stress response mechanisms such as mitophagy (Palikaras et al., 2015). In this instance, uaDf5 levels might decrease due to selection against high levels of uaDf5 either at the organismal or organelle level. In contrast, uaDf5 levels might be able to increase under mild mitochondrial stress conditions that activate UPRmt without inducing significant mitophagy or significantly compromising organismal fitness. An elegant series of experiments using RNAi dilution series show that the level of mitochondrial stress is an important determinant of aging (Rea et al., 2007). These studies show that mild mitochondrial stress extends lifespan while severe mitochondrial stress can shorten lifespan. Similarly, it will be interesting to test whether varying levels of mitochondrial stress result in differential modulation of uaDf5 levels, the prediction being that uaDf5 levels will increase with mild mitochondrial stress but decrease with severe stress.

Given that high levels of mutant mtDNA are required to disrupt mitochondrial function and activate the UPRmt, we anticipate that as mutant mtDNA levels decrease, they would become less phenotypically consequential and hence exposed to less stringent selection. Indeed, uaDf5 mtDNA persists, albeit at lower levels, in afts-1 loss-of-function mutants even after continuously propagating these animals for five months. These data suggest that first, while UPRmt allows uaDf5 to rise to high levels, it is not required for uaDf5 persistence at lower frequency. This interpretation is consistent with our observations that UPRmt activation generally correlates with uaDf5 levels, is likely not induced at appreciable levels in animals with uaDf5 below a certain threshold, and that loss of UPRmt fails to impact uaDf5 levels when they drop below this threshold. In support of this view, relatively modest shifts in heteroplasmic frequency of a pathogenic mtDNA point mutation were recently shown to coincide with abrupt changes in transcriptional profiles and phenotypic consequences in human patients (Picard et al., 2014). Such findings indicate that the cellular response to mutant mtDNA can vary considerably with even small changes in the frequency of the mutation. Accordingly, the abrupt but modest reduction (∼15%) that we observe in uaDf5 frequency in the absence of UPRmt may reflect a shift to levels at which the mutation is under more relaxed selection. Second, the persistence of uaDf5 mtDNA in atfs-1 loss-of-function mutants suggests roles for additional mechanisms such as exploitation of mtDNA copy number control (see above). In this instance, loss of mtDNA copy number control and UPRmt might be required to force permanent loss of uaDf5.

As with the “maintenance of wild type” model of mtDNA copy number control, successful proliferation of mutant mtDNA via UPRmt activation does not invoke a replicative advantage of the mutant genome. In both cases, the proliferation of mutant mtDNA is linked to a deleterious effect on mitochondrial function, underscoring their behavior as selfish genomes. Consequently, we can expect that an array of mtDNA mutations of varying sizes, including small deletions and point mutations, would proliferate selfishly as long as they induce UPRmt similarly to uaDf5.

Exploitation of homeostatic processes might also underlie persistence of naturally occurring selfish mtDNA. Many natural isolates of C. briggsae are heteroplasmic for mutant mtDNA (nad5∆ mtDNA) with a ∼900bp deletion that disrupts an essential gene (Howe and Denver, 2008). This nad5∆ mtDNA is widespread in C. briggsae natural populations despite its deleterious organismal effects (Estes et al., 2011). It has also been shown to have a strong drive to increase in frequency when bottlenecked through small populations (Clark et al., 2012). It will be interesting to determine in the future whether homeostatic mtDNA copy number control and UPRmt activation contribute to persistence of nad5∆ mtDNA in C. briggsae.

Mitochondrial stress is known to promote longevity (Dillin et al., 2002, Rea et al., 2007). In particular, UPRmt is thought to mediate this extension in lifespan (Durieux et al., 2011, Houtkooper et al., 2013). Given the role of UPRmt in alleviating proteotoxic stress in mitochondria (Schulz and Haynes, 2015), it has been proposed to protect mitochondria from age-related decline in proteostasis (Taylor and Dillin, 2011, Jensen and Jasper, 2014). Here we show that UPRmt is activated in the context of high mutant mtDNA frequency, which works to restore mitochondrial function that can be affected by mutant mtDNA. However, this indicates that the longevity-promoting protective role of UPRmt can also be viewed as a double-edged sword, by creating the conditions that allow the mutant mtDNA to propagate. Consistent with this interpretation, uaDf5 animals expressing an atfs-1 gain-of-function mutation that constitutively activates UPRmt exhibit reduced metabolic activity compared to uaDf5 animals with wildtype atfs-1 (Lin et al., 2016).

Although UPRmt is less characterized in mammals, previous reporting suggests that the mechanism by which UPRmt modulates heteroplasmy levels in C. elegans reported here may be widely conserved. Specifically, the accumulation of unfolded protein was shown to result in the up-regulation of the chaperone HSPD1 (HSP-60) and the mitochondrial protease CLPP in mammalian cells (Zhao et al., 2002). More recently, pharmacologically induced mitochondrial stress from rapamycin treatment was reported to correlate with UPRmt activation and increased longevity in both C. elegans and mice (Harrison et al., 2009, Houtkooper et al., 2013). Finally, numerous UPRmt factors, including HSPD1 and CLPP, are transcriptionally regulated by the CHOP-C/EBPβ heterodimer in mammalian cells (Aldridge et al., 2007), providing an attractive target for future studies characterizing the cellular mechanisms underlying mutant mtDNA dynamics in the context of human disease. Overall, the data presented here suggests that exploitation of homeostatic responses may represent a general strategy underlying the proliferation of pathogenic mtDNA mutations.

Materials & Methods

Strains

Worm strains were maintained on nematode growth medium (NGM) seeded with OP50 E. coli at 20°C under standard laboratory conditions.

Mutants

LGIII, pdr-1(gk448); LGV, atfs-1(tm4525), atfs-1(et15); mtDNA, uaDf5, mptDf1.

Transgenic lines

zcIs9 [hsp-60∷GFP + lin-15(+)], zcIs17 [ges-1∷GFP(mit)], zcIs18 [ges-1∷GFP(cyt)], byEx655 [pink-1∷Pink-1:GFP + myo-2∷mcherry + herring sperm DNA].

Droplet digital PCR

Day 4 adults were lysed in 50μL lysis buffer. Lysates were diluted 1:50 or 1:100 in water (same dilution factor across all samples in an experiment) before using 2μL for a ddPCR reaction. For ddPCR quantification of wildtype and uaDf5 mtDNA copy number and heteroplasmy frequency, we amplified product specifically off of wildtype mtDNA as well as a product common to wildtype and mutant templates; uaDf5 copy number was then determined by subtracting wildtype from total mtDNA copy number. Quantification of mptDf1 mtDNA was accomplished in a similar manner but from L4 animals. For quantification of actin, we used lysates diluted 1:5 fold.

Gene expression quantification

Worms were lysed by briefly incubating them at 65°C for 10 minutes to generate RNA, which was then used to synthesize cDNA according to manufacturer's protocol (ThermoFisher). Quantification of gene expression was performed using ddPCR, using primers listed in supplementary material.

Fluorescence microscopy

Worms were imaged using Zeiss Axio Zoom V16 stereo zoom microscope. Fluorescence intensities were quantified using Image J. 25 animals were used to calculate average fluorescence intensity for each group. To correlate hsp-60∷gfp fluorescence with uaDf5 levels, worms heteroplasmic for uaDf5 mtDNA and expressing the hsp-60∷gfp marker were individually picked as day 4 adults and immobilized on unseeded NGM plates treated with 250μL 10mM levamisole. Worms were individually imaged, followed by lysis and ddPCR quantification of mtDNA heteroplasmy as described above.

TMRE and Mitotracker green FM staining

Adult animals were grown overnight on 250μL of 10μM TMRE or 50μM Mitotracker Green FM (Molecular Probes). They were allowed to recover on new plates without the dye for 1 hr before imaging.

RNAi

RNAi bacterial cultures were grown at 37°C overnight. 750μL of the overnight culture was diluted in 75mL LB with ampicillin and grown at 37°C until OD550-600 > 0.8. An additional 75mL LB was added to the culture along with 1M IPTG to induce RNAi expression. Cultures were incubated an additional 3.5-4 hours at 37°C before pelleting and resuspending in M9 buffer with IPTG. These RNAi cultures were seeded onto IPTG-containing plates. L4 worms were grown on these RNAi plates and their progeny used for experimentation.

Fitness assays

Three adults picked from population of animals growing on RNAi plates since L4 stage were allowed to lay eggs for 3 hours on corresponding fresh RNAi plates. Number of unhatched and hatched embryos were counted one day later to determine fraction of unhatched embryos. After additional two days, total number of larva and adults were counted to determine fraction of animals with delayed growth. Subsequently, all animals were transferred to fresh RNAi plates every other day until day four of adulthood. Total number of dead animals were counted until day four of adulthood to determine fraction of dead animals.

Western blot analysis

One hundred worms in 10μL M9 were boiled in 10μL 2× SDS sample buffer for 10 minutes. SDS PAGE gel and transfer were performed according to standard protocol. Mouse monoclonal anti-beta-actin (sc-47778, Santa Cruz Biotechnology) or mouse monoclonal anti-GFP (sc-9996) were used at 1:500 dilution overnight at 4°C. HRP-conjugated goat anti-mouse antibody (sc-2005) was used at 1:5000 dilution for 90 minutes at room temperature. SuperSignal West Pico Chemiluminescent Substrate (Thermo Fisher) was used to detect HRP.

Supplementary Material

supplement

Highlights.

  • Mutant mtDNA uaDf5 behaves as a selfish genetic element in C. elegans

  • Mutant mtDNA exploit mtDNA copy number homeostasis to hitchhike to high frequency

  • Mitochondrial proteostasis defects in heteroplasmic animals induce UPRmt

  • UPRmt protects mutant mtDNA from parkin-mediated mitophagy

Acknowledgments

uaDf5 and mptDf1 heteroplasmic strains, pdr-1(gk448) and atfs-1(e15) mutant strain, and zcIs9, zcIs17, zcIs18, and byEx655 transgenic lines were kindly provided by Caenorhabditis Genetics Center (CGC). atfs-1(tm4525) mutant strain was kindly provided by the Mitani Lab through the National Bio-Resource Project of the MEXT, Japan. We would like to thank Sarah Sturgeon for technical assistance. We would like to thank O. Thompson and R. Waterston (University of Washington) for identifying mptDf1 deletion from the Million Mutation Project worm collection. We would like to thank Harmit Malik (HHMI/Fred Hutchinson Cancer Research Center) for valuable advice and support. We thank Harmit Malik, Nitin Phadnis, Janet Young, and Sarah Zanders for providing critical comments on the manuscript. Funding for this work was provided in part by Helen Hay Whitney Foundation Fellowship (MRP), by grants from the Mathers Foundation and HHMI (to Harmit Malik), startup funds from Vanderbilt University (MRP), the NIH-sponsored Cellular, Biochemical and Molecular Sciences Training Program (5T32GM008554-18) (BLG), and the NIH-funded Tennessee Center for AIDS Research (P30 AI110527) (to SAM).

Footnotes

Author contributions: MRP and BLG conceived and designed the study. BLG and MRP carried out the experiments. CSK carried out experiments with mptDf1 mtDNA heteroplasmy. BLG and RDG performed ddPCR experiments in the laboratory of SAM. MRP, BLG, and DCS analyzed the data. BLG and MRP prepared the manuscript.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Al Rawi S, Louvet-Vallee S, Djeddi A, Sachse M, Culetto E, Hajjar C, Boyd L, Legouis R, Galy V. Postfertilization autophagy of sperm organelles prevents paternal mitochondrial DNA transmission. Science. 2011;334:1144–7. doi: 10.1126/science.1211878. [DOI] [PubMed] [Google Scholar]
  2. Aldridge JE, Horibe T, Hoogenraad NJ. Discovery of genes activated by the mitochondrial unfolded protein response (mtUPR) and cognate promoter elements. PLoS One. 2007;2:e874. doi: 10.1371/journal.pone.0000874. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Baker BM, Nargund AM, Sun T, Haynes CM. Protective coupling of mitochondrial function and protein synthesis via the eIF2alpha kinase GCN-2. PLoS Genet. 2012;8:e1002760. doi: 10.1371/journal.pgen.1002760. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Benedetti C, Haynes CM, Yang Y, Harding HP, Ron D. Ubiquitin-like protein 5 positively regulates chaperone gene expression in the mitochondrial unfolded protein response. Genetics. 2006;174:229–39. doi: 10.1534/genetics.106.061580. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bernardi G. Lessons from a small, dispensable genome: the mitochondrial genome of yeast. Gene. 2005;354:189–200. doi: 10.1016/j.gene.2005.03.024. [DOI] [PubMed] [Google Scholar]
  6. Billing O, Kao G, Naredi P. Mitochondrial function is required for secretion of DAF-28/insulin in C. elegans. PLoS One. 2011;6:e14507. doi: 10.1371/journal.pone.0014507. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bratic I, Hench J, Henriksson J, Antebi A, Burglin TR, Trifunovic A. Mitochondrial DNA level, but not active replicase, is essential for Caenorhabditis elegans development. Nucleic Acids Res. 2009;37:1817–28. doi: 10.1093/nar/gkp018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Capps GJ, Samuels DC, Chinnery PF. A model of the nuclear control of mitochondrial DNA replication. J Theor Biol. 2003;221:565–83. doi: 10.1006/jtbi.2003.3207. [DOI] [PubMed] [Google Scholar]
  9. Chinnery PF, Samuels DC. Relaxed replication of mtDNA: A model with implications for the expression of disease. Am J Hum Genet. 1999;64:1158–65. doi: 10.1086/302311. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Clark KA, Howe DK, Gafner K, Kusuma D, Ping S, Estes S, Denver DR. Selfish little circles: transmission bias and evolution of large deletion-bearing mitochondrial DNA in Caenorhabditis briggsae nematodes. PLoS One. 2012;7:e41433. doi: 10.1371/journal.pone.0041433. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Clay Montier LL, Deng JJ, Bai Y. Number matters: control of mammalian mitochondrial DNA copy number. J Genet Genomics. 2009;36:125–31. doi: 10.1016/S1673-8527(08)60099-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Denver DR, Morris K, Lynch M, Vassilieva LL, Thomas WK. High direct estimate of the mutation rate in the mitochondrial genome of Caenorhabditis elegans. Science. 2000;289:2342–4. doi: 10.1126/science.289.5488.2342. [DOI] [PubMed] [Google Scholar]
  13. Diaz F, Bayona-Bafaluy MP, Rana M, Mora M, Hao H, Moraes CT. Human mitochondrial DNA with large deletions repopulates organelles faster than full-length genomes under relaxed copy number control. Nucleic Acids Res. 2002;30:4626–33. doi: 10.1093/nar/gkf602. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dillin A, Hsu AL, Arantes-Oliveira N, Lehrer-Graiwer J, Hsin H, Fraser AG, Kamath RS, Ahringer J, Kenyon C. Rates of behavior and aging specified by mitochondrial function during development. Science. 2002;298:2398–401. doi: 10.1126/science.1077780. [DOI] [PubMed] [Google Scholar]
  15. Dingley SD, Polyak E, Ostrovsky J, Srinivasan S, Lee I, Rosenfeld AB, Tsukikawa M, Xiao R, Selak MA, Coon JJ, Hebert AS, Grimsrud PA, Kwon YJ, Pagliarini DJ, Gai X, Schurr TG, Huttemann M, Nakamaru-Ogiso E, Falk MJ. Mitochondrial DNA variant in COX1 subunit significantly alters energy metabolism of geographically divergent wild isolates in Caenorhabditis elegans. J Mol Biol. 2014;426:2199–216. doi: 10.1016/j.jmb.2014.02.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Durham SE, Samuels DC, Cree LM, Chinnery PF. Normal levels of wild-type mitochondrial DNA maintain cytochrome c oxidase activity for two pathogenic mitochondrial DNA mutations but not for m.3243A-->G. Am J Hum Genet. 2007;81:189–95. doi: 10.1086/518901. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Durieux J, Wolff S, Dillin A. The cell-non-autonomous nature of electron transport chain-mediated longevity. Cell. 2011;144:79–91. doi: 10.1016/j.cell.2010.12.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Estes S, Coleman-Hulbert AL, Hicks KA, De Haan G, Martha SR, Knapp JB, Smith SW, Stein KC, Denver DR. Natural variation in life history and aging phenotypes is associated with mitochondrial DNA deletion frequency in Caenorhabditis briggsae. BMC Evol Biol. 2011;11:11. doi: 10.1186/1471-2148-11-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Harrison DE, Strong R, Sharp ZD, Nelson JF, Astle CM, Flurkey K, Nadon NL, Wilkinson JE, Frenkel K, Carter CS, Pahor M, Javors MA, Fernandez E, Miller RA. Rapamycin fed late in life extends lifespan in genetically heterogeneous mice. Nature. 2009;460:392–5. doi: 10.1038/nature08221. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Harrison E, Maclean RC, Koufopanou V, Burt A. Sex drives intracellular conflict in yeast. J Evol Biol. 2014;27:1757–63. doi: 10.1111/jeb.12408. [DOI] [PubMed] [Google Scholar]
  21. Haynes CM, Ron D. The mitochondrial UPR - protecting organelle protein homeostasis. J Cell Sci. 2010;123:3849–55. doi: 10.1242/jcs.075119. [DOI] [PubMed] [Google Scholar]
  22. Hicks KA, Howe DK, Leung A, Denver DR, Estes S. In vivo quantification reveals extensive natural variation in mitochondrial form and function in Caenorhabditis briggsae. PLoS One. 2012;7:e43837. doi: 10.1371/journal.pone.0043837. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hindson BJ, Ness KD, Masquelier DA, Belgrader P, Heredia NJ, Makarewicz AJ, Bright IJ, Lucero MY, Hiddessen AL, Legler TC, Kitano TK, Hodel MR, Petersen JF, Wyatt PW, Steenblock ER, Shah PH, Bousse LJ, Troup CB, Mellen JC, Wittmann DK, Erndt NG, Cauley TH, Koehler RT, So AP, Dube S, Rose KA, Montesclaros L, Wang S, Stumbo DP, Hodges SP, Romine S, Milanovich FP, White HE, Regan JF, Karlin-Neumann GA, Hindson CM, Saxonov S, Colston BW. High-throughput droplet digital PCR system for absolute quantitation of DNA copy number. Anal Chem. 2011;83:8604–10. doi: 10.1021/ac202028g. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Houtkooper RH, Mouchiroud L, Ryu D, Moullan N, Katsyuba E, Knott G, Williams RW, Auwerx J. Mitonuclear protein imbalance as a conserved longevity mechanism. Nature. 2013;497:451–7. doi: 10.1038/nature12188. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Howe DK, Denver DR. Muller's Ratchet and compensatory mutation in Caenorhabditis briggsae mitochondrial genome evolution. BMC Evol Biol. 2008;8:62. doi: 10.1186/1471-2148-8-62. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Jasmin JN, Zeyl C. Rapid evolution of cheating mitochondrial genomes in small yeast populations. Evolution. 2014;68:269–75. doi: 10.1111/evo.12228. [DOI] [PubMed] [Google Scholar]
  27. Jensen MB, Jasper H. Mitochondrial proteostasis in the control of aging and longevity. Cell Metab. 2014;20:214–25. doi: 10.1016/j.cmet.2014.05.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kowald A, Kirkwood TB. Transcription could be the key to the selection advantage of mitochondrial deletion mutants in aging. Proc Natl Acad Sci U S A. 2014;111:2972–7. doi: 10.1073/pnas.1314970111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Liau WS, Gonzalez-Serricchio AS, Deshommes C, Chin K, Lamunyon CW. A persistent mitochondrial deletion reduces fitness and sperm performance in heteroplasmic populations of C. elegans. BMC Genet. 2007;8:8. doi: 10.1186/1471-2156-8-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Lin YF, Schulz AM, Pellegrino MW, Lu Y, Shaham S, Haynes CM. Maintenance and propagation of a deleterious mitochondrial genome by the mitochondrial unfolded protein response. Nature. 2016;533:416–9. doi: 10.1038/nature17989. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Macalpine DM, Kolesar J, Okamoto K, Butow RA, Perlman PS. Replication and preferential inheritance of hypersuppressive petite mitochondrial DNA. EMBO J. 2001;20:1807–17. doi: 10.1093/emboj/20.7.1807. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Momeni B, Chen CC, Hillesland KL, Waite A, Shou W. Using artificial systems to explore the ecology and evolution of symbioses. Cell Mol Life Sci. 2011;68:1353–68. doi: 10.1007/s00018-011-0649-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Moraes CT, Kenyon L, Hao H. Mechanisms of human mitochondrial DNA maintenance: the determining role of primary sequence and length over function. Mol Biol Cell. 1999;10:3345–56. doi: 10.1091/mbc.10.10.3345. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Mouchiroud L, Houtkooper RH, Moullan N, Katsyuba E, Ryu D, Canto C, Mottis A, Jo YS, Viswanathan M, Schoonjans K, Guarente L, Auwerx J. The NAD(+)/Sirtuin Pathway Modulates Longevity through Activation of Mitochondrial UPR and FOXO Signaling. Cell. 2013;154:430–41. doi: 10.1016/j.cell.2013.06.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Nargund AM, Fiorese CJ, Pellegrino MW, Deng P, Haynes CM. Mitochondrial and nuclear accumulation of the transcription factor ATFS-1 promotes OXPHOS recovery during the UPR(mt) Mol Cell. 2015;58:123–33. doi: 10.1016/j.molcel.2015.02.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Nargund AM, Pellegrino MW, Fiorese CJ, Baker BM, Haynes CM. Mitochondrial import efficiency of ATFS-1 regulates mitochondrial UPR activation. Science. 2012;337:587–90. doi: 10.1126/science.1223560. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Okimoto R, Macfarlane JL, Clary DO, Wolstenholme DR. The mitochondrial genomes of two nematodes, Caenorhabditis elegans and Ascaris suum. Genetics. 1992;130:471–98. doi: 10.1093/genetics/130.3.471. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Palikaras K, Lionaki E, Tavernarakis N. Coordination of mitophagy and mitochondrial biogenesis during ageing in C. elegans. Nature. 2015;521:525–8. doi: 10.1038/nature14300. [DOI] [PubMed] [Google Scholar]
  39. Phillips WS, Coleman-Hulbert AL, Weiss ES, Howe DK, Ping S, Wernick RI, Estes S, Denver DR. Selfish Mitochondrial DNA Proliferates and Diversifies in Small, but not Large, Experimental Populations of Caenorhabditis briggsae. Genome Biol Evol. 2015;7:2023–37. doi: 10.1093/gbe/evv116. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Picard M, Zhang J, Hancock S, Derbeneva O, Golhar R, Golik P, O'hearn S, Levy S, Potluri P, Lvova M, Davila A, Lin CS, Perin JC, Rappaport EF, Hakonarson H, Trounce IA, Procaccio V, Wallace DC. Progressive increase in mtDNA 3243A>G heteroplasmy causes abrupt transcriptional reprogramming. Proc Natl Acad Sci U S A. 2014;111:E4033–42. doi: 10.1073/pnas.1414028111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Randow F, Youle RJ. Self and nonself: how autophagy targets mitochondria and bacteria. Cell Host Microbe. 2014;15:403–11. doi: 10.1016/j.chom.2014.03.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Rauthan M, Ranji P, Aguilera Pradenas N, Pitot C, Pilon M. The mitochondrial unfolded protein response activator ATFS-1 protects cells from inhibition of the mevalonate pathway. Proc Natl Acad Sci U S A. 2013;110:5981–6. doi: 10.1073/pnas.1218778110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Rea SL, Ventura N, Johnson TE. Relationship between mitochondrial electron transport chain dysfunction, development, and life extension in Caenorhabditis elegans. PLoS Biol. 2007;5:e259. doi: 10.1371/journal.pbio.0050259. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Reznik E, Miller ML, Senbabaoglu Y, Riaz N, Sarungbam J, Tickoo SK, Al-Ahmadie HA, Lee W, Seshan VE, Hakimi AA, Sander C. Mitochondrial DNA copy number variation across human cancers. Elife. 2016;5 doi: 10.7554/eLife.10769. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Rossignol R, Faustin B, Rocher C, Malgat M, Mazat JP, Letellier T. Mitochondrial threshold effects. Biochem J. 2003;370:751–62. doi: 10.1042/BJ20021594. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Runkel ED, Liu S, Baumeister R, Schulze E. Surveillance-activated defenses block the ROS-induced mitochondrial unfolded protein response. PLoS Genet. 2013;9:e1003346. doi: 10.1371/journal.pgen.1003346. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Sato M, Sato K. Degradation of paternal mitochondria by fertilization-triggered autophagy in C. elegans embryos. Science. 2011;334:1141–4. doi: 10.1126/science.1210333. [DOI] [PubMed] [Google Scholar]
  48. Schulz AM, Haynes CM. UPR(mt)-mediated cytoprotection and organismal aging. Biochim Biophys Acta. 2015;1847:1448–56. doi: 10.1016/j.bbabio.2015.03.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Stewart JB, Chinnery PF. The dynamics of mitochondrial DNA heteroplasmy: implications for human health and disease. Nat Rev Genet. 2015;16:530–42. doi: 10.1038/nrg3966. [DOI] [PubMed] [Google Scholar]
  50. Suen DF, Narendra DP, Tanaka A, Manfredi G, Youle RJ. Parkin overexpression selects against a deleterious mtDNA mutation in heteroplasmic cybrid cells. Proc Natl Acad Sci U S A. 2010;107:11835–40. doi: 10.1073/pnas.0914569107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Tam ZY, Gruber J, Halliwell B, Gunawan R. Context-Dependent Role of Mitochondrial Fusion-Fission in Clonal Expansion of mtDNA Mutations. PLoS Comput Biol. 2015;11:e1004183. doi: 10.1371/journal.pcbi.1004183. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Tang Y, Manfredi G, Hirano M, Schon EA. Maintenance of human rearranged mitochondrial DNAs in long-term cultured transmitochondrial cell lines. Mol Biol Cell. 2000a;11:2349–58. doi: 10.1091/mbc.11.7.2349. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Tang Y, Schon EA, Wilichowski E, Vazquez-Memije ME, Davidson E, King MP. Rearrangements of human mitochondrial DNA (mtDNA): new insights into the regulation of mtDNA copy number and gene expression. Mol Biol Cell. 2000b;11:1471–85. doi: 10.1091/mbc.11.4.1471. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Taylor DR, Zeyl C, Cooke E. Conflicting levels of selection in the accumulation of mitochondrial defects in Saccharomyces cerevisiae. Proc Natl Acad Sci U S A. 2002;99:3690–4. doi: 10.1073/pnas.072660299. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Taylor RC, Dillin A. Aging as an event of proteostasis collapse. Cold Spring Harb Perspect Biol. 2011;3 doi: 10.1101/cshperspect.a004440. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Thompson O, Edgley M, Strasbourger P, Flibotte S, Ewing B, Adair R, Au V, Chaudhry I, Fernando L, Hutter H, Kieffer A, Lau J, Lee N, Miller A, Raymant G, Shen B, Shendure J, Taylor J, Turner EH, Hillier LW, Moerman DG, Waterston RH. The million mutation project: a new approach to genetics in Caenorhabditis elegans. Genome Res. 2013;23:1749–62. doi: 10.1101/gr.157651.113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Tsang WY, Lemire BD. Stable heteroplasmy but differential inheritance of a large mitochondrial DNA deletion in nematodes. Biochem Cell Biol. 2002;80:645–54. doi: 10.1139/o02-135. [DOI] [PubMed] [Google Scholar]
  58. Valenci I, Yonai L, Bar-Yaacov D, Mishmar D, Ben-Zvi A. Parkin modulates heteroplasmy of truncated mtDNA in Caenorhabditis elegans. Mitochondrion. 2015;20:64–70. doi: 10.1016/j.mito.2014.11.001. [DOI] [PubMed] [Google Scholar]
  59. Wallace DC. Mitochondrial genetics: a paradigm for aging and degenerative diseases? Science. 1992;256:628–32. doi: 10.1126/science.1533953. [DOI] [PubMed] [Google Scholar]
  60. Wallace DC, Chalkia D. Mitochondrial DNA genetics and the heteroplasmy conundrum in evolution and disease. Cold Spring Harb Perspect Biol. 2013;5:a021220. doi: 10.1101/cshperspect.a021220. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Williamson D. The curious history of yeast mitochondrial DNA. Nat Rev Genet. 2002;3:475–81. doi: 10.1038/nrg814. [DOI] [PubMed] [Google Scholar]
  62. Yoneda T, Benedetti C, Urano F, Clark SG, Harding HP, Ron D. Compartment-specific perturbation of protein handling activates genes encoding mitochondrial chaperones. J Cell Sci. 2004;117:4055–66. doi: 10.1242/jcs.01275. [DOI] [PubMed] [Google Scholar]
  63. Zhao Q, Wang J, Levichkin IV, Stasinopoulos S, Ryan MT, Hoogenraad NJ. A mitochondrial specific stress response in mammalian cells. EMBO J. 2002;21:4411–9. doi: 10.1093/emboj/cdf445. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

supplement
NIHMS796406-supplement.pdf (2.4MB, pdf)

RESOURCES