Longevity-promoting mitochondrial unfolded protein response activation requires elements of the PeBoW complex

In this study, Adedoja et al. show that reductions in mitochondrial membrane protein VDAC1 increase life span in C. elegans by activating the mitochondrial unfolded protein response (UPR^mt). They identify the PeBoW complex as a broadly required, critical mediator of UPR^mt activation upon mitochondrial stress, including that caused by VDAC1 loss, suggesting that PeBoW plays an important role in mitochondrial function and quality regulation and in animal longevity.

Abstract

Mitochondria play a crucial role in cellular energy metabolism and homeostasis and are strongly implicated in aging and age-related diseases. The outer mitochondrial membrane protein voltage-dependent anion channel (VDAC) plays multiple roles in mitochondrial homeostasis, including transport of metabolites, ATP, and Ca²⁺. Dysregulation of VDAC levels has been associated with cancer, neurodegeneration, metabolic disorders, and aging. Previously, we demonstrated that elevated VDAC-1 levels in Caenorhabditis elegans lead to increased mitochondrial permeability and reduced life span. Here we demonstrate that reduced VDAC-1 function extends life span through the activation of the mitochondrial unfolded protein response (UPR^mt), a conserved stress response that maintains mitochondrial proteostasis and is linked to life span extension in multiple species. Leveraging unbiased genomic discovery, we identified genes encoding several proteins in the PeBoW complex as a critical mediator of UPR^mt activation following VDAC-1 loss. More broadly, we demonstrated a universal requirement for several PeBoW component genes across diverse mitochondrial stressors in order to fully animate the UPR^mt. Our findings reveal a heretofore unappreciated role for PeBoW components in UPR^mt induction and life span extension in response to mitochondrial stress, highlighting its essential function in mitochondrial quality control and longevity pathways.

Mitochondria are essential organelles in eukaryotic cells and are the energetic center that drives cellular function and homeostasis. Proper mitochondrial function is essential for cellular viability across eukarya and for prevention of premature aging and age-associated human diseases such as cancer, cardiac disease, and neurodegenerative diseases (Newmeyer and Ferguson-Miller 2003; Wallace 2005). In humans, disorders in mitochondrial energy metabolism that are a consequence of defects in one or more subunits of the electron transport chain (ETC) lead to lethality, early demise, or a variety of metabolic diseases (McInnes 2013). However, and seemingly paradoxically, partial inhibition of mitochondrial function extends life span in many species like yeast, Drosophila, mice, and roundworms (Lakowski and Hekimi 1996; Kirchman et al. 1999; Dell'agnello et al. 2007; Copeland et al. 2009). In these organisms, reduced mitochondrial function promotes a signaling pathway between the mitochondria and the nucleus, which increases life span, driven by a subset of mito-protective genes from the mitochondrial unfolded protein response (UPR^mt) (Zhao et al. 2002; Yoneda et al. 2004).

The mitochondrial unfolded protein response (UPR^mt) is a conserved transcriptional response that maintains mitochondrial protein homeostasis during states of energetic or mitochondrial stress and dysfunction. In Caenorhabditis elegans, depletion of ETC components, mitochondrial DNA, reduced mitochondrial translation, or exposure to mitochondrial toxins activates the UPR^mt (Fiorese and Haynes 2017). Central to the UPR^mt regulation in the nematode is the ATF family transcription factor ATFS-1. Under unstressed conditions, an N-terminal mitochondrial targeting signal prompts import of ATFS-1 into the mitochondrial inner matrix, where it is constitutively degraded. However, under conditions of mitochondrial stress, mitochondrial import fails, prompting ATFS-1 to be imported into the nucleus, where it upregulates the entire cadre of genes involved in the UPR^mt (Nargund et al. 2012, 2015). In higher eukaryotes, including mammals, the UPR^mt similarly involves a mito–nuclear signaling relay and ATF family transcription factors, though the precise proteins transducing this signal remain incompletely characterized (Fiorese et al. 2016; Quirós et al. 2017). UPR^mt activation also requires other components, including mitochondrial protease ClpP (Haynes et al. 2007), the ubiquitin-like protein UBL-5 (Benedetti et al. 2006), and several chromatin regulatory factors (Merkwirth et al. 2016; Tian et al. 2016). UBL-5 is a chromatin modifier that binds to other transcription factors such as DVE-1 to activate the expression of mitochondrial chaperones hsp-6 and hsp-60 and other UPR^mt components (Tian et al. 2016; Melber and Haynes 2018).

Also known as the mitochondrial porin, the voltage-dependent anion channel (VDAC) is a multifunctional protein located on the outer mitochondria membrane (OMM) that is involved in the transport of ions (Na⁺, K⁺, and Cl⁻) and metabolites (ATP/ADP, pyruvate, succinate, malate, and glutamate) and the maintenance of cellular homeostasis (Mihara and Sato 1985; Kleene et al. 1987). VDAC plays a central role in cellular metabolism by interacting with several cytosolic and mitochondrial proteins. These interactions make VDACs essential for ATP production, cell growth, cell death (apoptosis), and cell survival signaling pathways (Abu-Hamad et al. 2006; Magrì et al. 2018). VDAC levels appear to be tightly regulated by the cell, as elevated VDAC levels have been linked to different forms of cancers and neurodegenerative diseases (Shteinfer-Kuzmine et al. 2019), while low VDAC levels have been reported in cardiac diseases and mitochondrial myopathies (Huizing et al. 1996; Tian et al. 2023). Consequently, VDACs are both potential therapeutic targets and disease biomarkers (Brahimi-Horn and Mazure 2014; Reina and De Pinto 2018; Fang et al. 2022). We have previously reported that in C. elegans, high levels of VDAC-1 cause increased mitochondrial permeability transition, shorten life span in wild-type animals, and prevent life span extension in longevity paradigms dependent on increased levels of autophagy (Zhou et al. 2019).

In this study, given the detrimental effects of excess VDAC-1 on life span and age-related morbidity (Zhou et al. 2019), we set out to determine the impact of reduced VDAC-1 function on longevity. As hypothesized, the knockdown of VDAC-1 extends life span but not in the expected manner; that is, dependent on altered mitochondrial permeability. Alternatively, we found that VDAC-1 knockdown or depletion by auxin-inducible degradation (AID) activates the UPR^mt and increases life span, mechanistically similar to reduced function of classic electron transport chain (ETC) components. Further investigation into mechanisms by which VDAC-1 knockdown induces the UPR^mt identified a genetic requirement for PeBoW, a previously known but understudied protein complex previously invoked in coordinating ribosome biogenesis. We found that the PeBoW complex is essential to the induction of UPR^mt following VDAC-1 loss and furthermore is broadly required for UPR^mt activation in response to a number of mitochondrial insults, including reduced function of the ETC. Importantly, not only is PeBoW required for UPR^mt induction, but as logically follows from this requirement, genes encoding protein members of the complex are also necessary for life span extension in response to ETC loss and paraquat survival. These results invoke heretofore unappreciated governance of the UPR^mt in response to VDAC-1 loss and energetic stress and a critical requirement for the specific components of the PeBoW complex in longevity paradigms associated with compromised mitochondrial function.

Results

Reduced VDAC-1 function promotes longevity requiring activation of the mitochondrial unfolded protein response

We have previously reported that high levels of VDAC-1 are detrimental in aging through increased mitochondrial permeability transition (mPT), which results in shortened life span (Zhou et al. 2019). Thus, we reasoned that low levels of VDAC-1 may conversely promote longevity. As vdac-1 loss of function leads to lethality, using RNA interference (RNAi), we knocked down vdac-1 in C. elegans and observed no effect on life span (Fig. 1A). vdac-1 RNAi efficiency was ∼50% as determined by quantitative RT-PCR for vdac-1 mRNA (Supplemental Fig. S1A). Considering that RNAi efficiency is limited in some C. elegans tissues (Tavernarakis et al. 2000; Kamath et al. 2001) and that VDAC is among the most abundant mitochondrial proteins (Mannella and Bonner 1975; Yamamoto et al. 2006), we conducted two-generation RNAi knockdown, whereby progeny from worms treated with vdac-1 RNAi from the L1 stage to adulthood were used to obtain a more potent reduction in VDAC-1 levels (Fig. 1B). This two-generation knockdown of vdac-1 by RNAi resulted in a 44% additional decrease in VDAC-1 protein levels compared with the VDAC-1 protein in the first generation (Supplemental Fig. S1B) and an increase in median life span versus vector control RNAi (Fig. 1C). To confirm that the life span extension observed with vdac-1 knockdown was due to specific inactivation of vdac-1, we performed a two-generation RNAi treatment in rde-1 mutant animals, which are deficient in RNAi. Loss of RNAi activity abolished the life span extension typically induced by vdac-1 knockdown (Supplemental Fig. S1C). Next, we investigated whether the two-generation vdac-1 phenotype was a result of transgenerational epigenetics. To test this, we conducted two-generation knockdown of vdac-1 in animals carrying deletions in set-25 and set-32—genes essential for the establishment of transgenerational inherited phenotypes (Woodhouse et al. 2018). We found that reduced vdac-1 still resulted in an increased life span (Supplemental Fig. S1D,E).

Figure 1.

Longevity with reduced vdac-1 function depends on the UPR^mt. (A) Reduced vdac-1 does not affect survival in P0 generation (one-generation knockdown) animals. (B) One-generation and two-generation knockdown experimental workflow for life span assays for vdac-1 RNAi. (C) Two-generation knockdown of vdac-1 (F1 progeny of animals with reduced vdac-1 RNAi) led to a 30% increase in longevity versus control. (D) Reduced vdac-1 (two-generation knockdown) significantly activates the UPR^mt. Shown for reference are known UPR^mt activators cco-1, mrps-5, or nuo-6 RNAi. (E) Like other genetic disruptions of mitochondrial function (cco-1 and nuo-6 RNAi), atfs-1 loss-of-function mutants are sterile following two-generation knockdown of vdac-1. (F) Loss of function of the UPR^mt effector ubl-5 prevents longevity resulting from two-generation vdac-1 knockdown. Data are represented as mean ± SD. (****) P < 0.0001, (ns) nonsignificant. Assays were by a one-way ANOVA (D,E) and log rank analysis (A,C,F). For life spans, results represent at least two biological replicates. See Supplemental Table S1 for tabular statistical survival data and biological replicates.

As our prior studies indicate that VDAC-1 excess shortens life span via promotion of mPT (Zhou et al. 2019), we hypothesized that the mechanism by which vdac-1 depletion extends life span could be by lowering mPT. Prior work has shown that mPT is associated with pathogenic activation of the mitochondrial unfolded protein response (UPR^mt) (Angeli et al. 2021); therefore, we asked whether vdac-1 depletion lowers UPR^mt activation. We were surprised to see the exact opposite: Longevity-promoting two-generation vdac-1 RNAi knockdown, but not one-generation knockdown, promoted strong induction of a hsp-6p::GFP (mitochondrial heat shock protein HSPA9 ortholog) reporter for UPR^mt activation (Fig. 1D; Supplemental Fig. S1E). When compared with other forms of mitochondrial stress, we found that two-generation vdac-1 knockdown induces hsp-6p::GFP expression significantly, albeit to a slightly lesser extent than electron transport chain inhibition (nuo-6 and cco-1 RNAi) and inhibition of mitochondrial translation (mrps-5 RNAi) (Fig. 1D).

UPR^mt activation in C. elegans requires translocation of the transcription factor ATFS-1 to the nucleus to activate a group of stress response genes including mitochondrial heat shock proteins (hsp-6/HSPA9 and hsp-60/HSPD1) (Nargund et al. 2015). Therefore, we tested mutants lacking atfs-1 to determine whether it is required for longevity induced by reduced vdac-1 function. We found that concomitant loss of function of atfs-1 and vdac-1 knockdown by RNAi leads to lethality and that this phenotype exactly mirrors loss of electron transport chain (ETC) components nuo-6 (ETC complex 1) and cco-1 (cytochrome C oxidase) in atfs-1 mutants (Fig. 1E). These data are consistent with published studies indicating that atfs-1 loss increases embryonic lethality under conditions of ETC disruption (Wu et al. 2018). We next tested whether two-generation vdac-1 knockdown depends on ubl-5, a ubiquitin-like gene required for UPR^mt activation and associated longevity (Benedetti et al. 2006; Haynes et al. 2007). Indeed, loss of ubl-5 completely prevents life span extension attributable to vdac-1 RNAi (F1) knockdown (Fig. 1F). This result is similar to other forms of mitochondrial stress like reduced cco-1, reduced isp-1 (ETC complex 3), and reduced mrps-5 (mitochondrial protein translation), all of which increase C. elegans life span by activating the UPR^mt and are dependent on UBL-5 (Durieux et al. 2011; Houtkooper et al. 2013). In aggregate, these data indicate that reduced vdac-1 function by two-generation knockdown prompts mitochondrial stress, increasing longevity in a manner dependent on induction of the UPR^mt.

Reverse genetic screening identifies suppressors of the UPR^mt with reduced VDAC-1

We considered the possibility that reduced vdac-1 might be associated with other forms of cellular stress responses, as VDACs are known to function in superoxide release from the mitochondria and interact with proteins on the endoplasmic reticulum (Han et al. 2003; Lustgarten et al. 2012; Janikiewicz et al. 2018). Therefore, we examined stress response reporters for ER stress (hsp-4p::GFP) and oxidative stress (gst-4p::GFP) with vdac-1 RNAi (F1) and found that reduced vdac-1 levels did not activate these stress pathways, which is in sharp contrast to reduced levels of ETC components, which activate these pathways (Supplemental Fig. S2A,B). Similarly, a reporter for β-oxidation, acs-2p::GFP, is not activated by reduced vdac-1 levels, while other forms of mitochondrial stress (cco-1 and mrps-5 RNAi) significantly induce acs-2p::GFP expression (Supplemental Fig. S2C).

Based on these differences, we hypothesized that reduced vdac-1 function induces the UPR^mt in a mechanistically distinct manner. To identify genes involved during UPR^mt induction with reduced vdac-1, we performed a targeted RNAi screen of ∼1000 genes bearing metabolic function annotation for altered hsp-6p::GFP expression in C. elegans with VDAC-1 levels reduced by auxin-inducible degron (AID) depletion (Fig. 2A). First, confirming UPR^mt with compromised VDAC-1 function, expression of the hsp-6p::GFP reporter is increased in AID-tagged VDAC-1 animals irrespective of whether auxin is present or not (Fig. 2B). Addition of auxin further reduced VDAC-1 protein levels but did not produce a corresponding increase in UPR^mt activation (Supplemental Fig. S2G). This suggests that the introduction of a C-terminal GFP::AID::3xFLAG tag in the presence of the auxin-activated ubiquitin ligase TIR-1 alone is sufficient to alter VDAC-1 function to the extent that the UPR^mt is activated.

Figure 2.

RNAi screen of 1000 genes with annotated metabolic function identifies suppressors of the UPR^mt with reduced function of vdac-1. (A) Schematic representation of the RNAi screen of 1000 metabolic genes from the Ahringer RNAi library. RNAi treatment was from L1 with 5 μM auxin. Worms were visually scored for total or partial knockdown in GFP. (B) VDAC-1 depletion by auxin-inducible degradation triggered by 5 μM auxin activates the UPR^mt. (C) Eight primary hits from the screen and their predicted protein function. (D) Secondary follow-up, validation analysis including GFP images and quantification of GFP signal for positive hits from the primary RNAi screen. Data are represented as mean ± SD. (****) P < 0.0001, (ns) nonsignificant. (B,D) Assays were by one-way ANOVA. Quantifications in A and D represent the aggregate analysis of three biological replicates collected from worms on day 1 of adulthood.

The RNAi screen yielded eight hits, which, when knocked down, reduced activation of the hsp-6p::GFP reporter upon VDAC-1 depletion (Fig. 2C), including known UPR^mt regulators elt-2, atfs-1, and sptl-1 (Liu et al. 2014; Liu and Wang 2021).

Other RNAi screen hits were predicted to be involved in lipid metabolism, including H41C03.1 (phosphatidylinositol transfer protein), sbp-1/SREBP, and lpd-7 (lipid-depleted 7) (Fig. 2C,D). Even though sbp-1 emerged from the primary screen but did not validate in secondary confirmatory studies, the presence of these lipid regulatory genes led us to test additional genes for lipid synthesis. Monounsaturated fatty acid synthesis is governed by the transcription factor sbp-1, which in turn regulates mRNA expression of a series of fatty acid biosynthetic enzymes, including acetyl-CoA carboxylase (POD-2) and fatty acid synthase (FASN-1) (Watts and Ristow 2017). To determine whether reduced vdac-1 leads to alterations in sbp-1 activity, we assessed mRNA levels of pod-2, fasn-1, sbp-1, and fatty acid desaturases fat-6 and fat-7 following vdac-1 knockdown and found no significant change in mRNA levels (Supplemental Fig. S2D). Translational reporters for FASN-1::GFP and GFP::POD-2 similarly indicated no increase in protein levels following vdac-1 knockdown (Supplemental Fig. S2E,F). These data are consistent with other forms of mitochondrial stress that similarly show no increase in expression of fatty acid biosynthetic enzymes (e.g., RNAi of cco-1) (see Supplemental Fig. 2E,F). We also tested whether knockdown of fasn-1 or pod-2 by RNAi impacted hsp-6p::GFP expression and found that they did not (data not shown). Thus, these data do not emphasize a strong connection of lipid metabolism to animation of the UPR^mt either in response to reduced vdac-1 or upon exposure to other mitochondrial stressors.

LPD-7/PES1 and other components of the PeBoW complex are required for UPR^mt animation in response to reduced VDAC-1 function

Thereafter, we returned to our screen data to examine lpd-7, RNAi knockdown of which suppressed hsp-6p::GFP with a potency approaching that of atfs-1 (Fig. 2D). lpd-7 (lipid-depleted 7) has a moniker related to its identification in a targeted RNAi screen for alterations in fat storage in C. elegans using the neutral lipid stain Sudan Black (McKay et al. 2003). First, we tested the previously reported effect of lpd-7 RNAi on fat levels in C. elegans using fixative Nile Red staining on day 3 of adulthood. Contrary to the published results, we found an increase in fat levels with lpd-7 RNAi compared with the control (Fig. 3A). Furthermore, adding oleate, which is a key intermediate in lipid synthesis, did not rescue low hsp-6 mRNA levels observed with lpd-7 RNAi with VDAC-1 knockdown, further suggesting that altered UPR^mt activation is not the consequence of disrupted lipid homeostasis (Fig. 3B). Finally, we considered whether other genes involved in fat storage were suppressors of the vdac-1 knockdown-induced UPR^mt. We found that RNAi of dgat-2, which converts diacylglycerol to triglycerides for storage in lipid droplets (Xu et al. 2012; Walther et al. 2017), had no effect on UPR^mt activation with VDAC-1 knockdown (Supplemental Fig. S3A). In aggregate, these data, together with the analysis of sbp-1 and other genes encoding lipid biosynthetic enzymes, indicate that suppression of the UPR^mt upon lpd-7 knockdown is unlikely to be associated with lipid synthesis and storage.

Figure 3.

lpd-7 and other components of the PeBoW complex are required to upregulate the UPR^mt following vdac-1 knockdown. (A) Fat mass is higher in wild-type worms treated with lpd-7 RNAi, as assayed using fixative-based Nile Red staining. (B) hsp-6 mRNA levels in animals with reduced vdac-1 (using auxin) with or without supplementation with 0.2 mM oleate determined by qRT-PCR. (C) P_hsp-6:GFP levels in animals with reduced vdac-1 (using auxin) combined with RNAi of either atfs-1 or members of the PeBoW complex (lpd-7, bop-1, and wdr-12). Data are represented as mean ± SD. (****) P < 0.0001, (ns) nonsignificant. (B,C) Assays were by one-way ANOVA.

LPD-7 is an evolutionarily conserved protein, and its mammalian homolog, Pescadillo (PES1), associates with block of proliferation 1(BOP1) and WD repeat domain 12 (WDR12) to form the PeBoW complex, located in the nucleolus (Hölzel et al. 2005). The PeBoW complex is essential for maturation of the large ribosomal subunit and cell proliferation in mammalian cells (Rohrmoser et al. 2007). We wanted to know whether other components of the complex acted as suppressors of the UPR^mt with reduced vdac-1. Like lpd-7, RNAi of the C. elegans homolog of Bop1, bop-1, suppressed the UPR^mt, whereas wdr-12 RNAi further increased the intensity of the UPR^mt reporter (Fig. 3C). To assess the impact of PeBoW complex reduction on cell proliferation in C. elegans, we measured body size following RNAi knockdown of individual components. Depletion of lpd-7, bop-1, and, to a lesser extent, wdr-12 resulted in reduced body size without observable delays in developmental timing (Supplemental Fig. S3B). To determine whether the phenotypic differences reflected variable knockdown efficiencies, we quantified mRNA levels of each PeBoW component using qPCR. Knockdown resulted in an ∼10-fold reduction in lpd-7, a fivefold reduction in bop-1, and a 2.2-fold reduction in wdr-12 transcripts (Supplemental Fig. S3C–E). Next, we asked whether the knockdown of the PeBoW complex was specific to UPR^mt activation or whether the PeBoW complex also regulated other cellular stress response pathways. RNAi of PeBoW complex proteins did not activate the cytosolic heat shock response (HSR) reporter hsp-16.2p::GFP in unstressed conditions and did not prevent activation in response to 2 h of heat stress (Supplemental Fig. S3F). Thus, generally speaking, the PeBoW complex appears to specifically regulate the UPR^mt.

PeBoW complex components are required for UPR^mt activation in response to diverse mitochondrial stressors

Even though we identified PeBoW components in a screen for UPR^mt activation in response to reduced vdac-1 function, it remained possible that PeBoW is broadly required across multiple mitochondrial stressors to activate defense mechanisms. Thus, we tested whether reduced levels of PeBoW complex proteins affected UPR^mt induction in response to other genetic or chemical mitochondrial insults. Treating C. elegans with a low dose of the mitochondrial free radical generator paraquat affects mitochondrial respiration and increases life span in C. elegans (Lee et al. 2010). We found that RNAi knockdown of lpd-7 and bop-1 prevented the activation of the UPR^mt; in contrast, wdr-12 knockdown further increased UPR^mt activation with the addition of 0.5 mM paraquat (Fig. 4A). lpd-7 and bop-1 RNAi also reduced the survival rate after 24 h for animals treated with a high dose (40 mM) of paraquat, which causes short-term lethality (Fig. 4B). Consistently, lpd-7 and bop-1 also prevented life span extension from lower-dose (6 mM), chronic paraquat exposure, while wdr-12 RNAi had no effect on survival (Supplemental Fig. S4A).

Figure 4.

PeBoW complex components are required for UPR^mt induction following varied forms of mitochondrial impairment. (A) P_hsp-6:GFP levels in wild-type animals treated with 0.5 mM paraquat and lpd-7, bop-1, or wdr-12 RNAi. (B) Percentage survival at 24 h for wild-type animals treated with high-dose (40 mM) paraquat and lpd-7, bop-1, or wdr-12 RNAi. (C) P_hsp-6:GFP levels in wild-type animals treated with tandem RNAi constructs for cco-1 and PeBoW complex proteins. (D) Survival curve for wild-type animals treated with RNAi for lpd-7 and cco-1 + lpd-7. (E) Survival curve for wild-type animals treated with RNAi for bop-1 and cco-1 + bop-1. (F) Survival curve for wild-type animals treated with RNAi for wdr-12 and cco-1 + wdr-12. (G) P_hsp-6:GFP levels in isp-1(qm150) animals treated with RNAi for PeBoW complex proteins. (H) Survival curve for wild-type and isp-1(qm150) animals treated with RNAi for lpd-7. (I) Survival curve for wild-type and isp-1(qm150) animals treated with RNAi for bop-1. (J) Survival curve for wild-type and isp-1(qm150) animals treated with RNAi for wdr-12. Data are represented as mean ± SD. (*) P < 0.05, (**) P < 0.01, (****) P < 0.0001, (ns) nonsignificant. Assays were by one-way ANOVA (A–C,G) or log rank analysis (D–F,H–J). For life spans, results represent at least two biological replicates. See Supplemental Table S1 for tabular statistical survival data and biological replicates.

To study the effect of reduced PeBoW levels in combination with genetic disruption of mitochondrial function, we built tandem RNAi plasmid constructs consisting of sequences for cco-1 with either lpd-7, bop-1, or wdr-12. We measured cco-1 RNAi efficiency levels for all three construct levels by qPCR to ensure consistent knockdown of cco-1 in these constructs (Supplemental Fig. S4B). Next, we treated hsp-6p::GFP reporter worms with the tandem cco-1:PeBoW RNAi to quantify UPR^mt activation. The results show that in spite of equivalent cco-1 knockdown, reducing levels of all three PeBoW complex components mitigated activation of the UPR^mt, with lpd-7 and bop-1 having a larger effect compared with wdr-12 (Fig. 4C). To further validate the results from the hsp-6p::GFP reporter, we measured the levels of cdr-12, another gene upregulated during the UPR^mt. Compared with animals treated with cco-1 RNAi alone, cdr-12 levels were lower in the animals treated with tandem cco-1 + PeBoW (Supplemental Fig. S4C). Next, we conducted life span assays with wild-type worms following treatment with control, cco-1, and tandem cco-1 + PeBoW RNAi. Once again, reduced levels of all three PeBoW complex proteins were sufficient to prevent increased life span with reduced cco-1 (Fig. 4D–F). As complete loss of PeBoW complex components is lethal in C. elegans, we evaluated the impact of partial lpd-7 reduction on life span by performing cco-1 RNAi in a balanced lpd-7 heterozygous strain. Like the tandem cco-1 + lpd-7 RNAi condition, reduced lpd-7 levels in the heterozygote suppressed the life span extension typically observed with cco-1 knockdown (Supplemental Fig. S4D). Finally, we crossed hsp-6::GFP worms with isp-1(qm150) mutant animals and treated them with RNAi knockdown of lpd-7, bop-1, or wdr-12. Like with reduced vdac-1 and paraquat treatment, we found that RNAi knockdown of lpd-7 and bop-1 reduced the animation of the UPR^mt; in contrast, wdr-12 knockdown further increased UPR^mt activation (Fig. 4G). We followed these results with life span assays with wild type and isp-1(qm150) on PeBoW RNAi. RNAi of both lpd-7 and bop-1 proteins was sufficient to prevent increased life span with isp-1(qm150), while wdr-12 had no effect on life span (Fig.4H–J). These results, albeit inconsistent with regard to the specific role of WDR-12, generally indicate that proteins in the PeBoW complex positively regulate the UPR^mt activation initiated by both genetic and chemical activators and that PeBoW complex activity is essential for survival from varied forms of mitochondrial stress.

PeBoW complex components regulate the UPR^mt downstream from ATFS-1

To understand how components of the PeBoW complex were influencing the activity of the UPR^mt, we used a group of available reporters and mutants to study PeBoW complex knockdown. We found that reduced levels of lpd-7 and bop-1 robustly increased the protein levels and nuclear localization of the UPR^mt regulatory factor UBL-5, but reduced wdr-12 did not (Fig. 5A). Next, we obtained a C. elegans strain with constitutively active ATFS-1 owing to an atfs-1 gain-of-function mutation and treated these worms with lpd-7 and bop-1 RNAi. We found that reducing the function of LPD-7/PES1 or BOP-1/BOP1 returned hsp-6 mRNA back to wild-type levels when compared with the worms fed the control RNAi (Fig. 5B). These data suggested that the PeBoW complex components interfere, either directly or indirectly, with the transcriptional activity of ATFS-1. To determine whether this occurred in part via direct regulation of ATFS-1 levels, we used a translational reporter for ATFS-1 to measure the protein levels following RNAi knockdown of individual PeBoW complex components and with and without concomitant cco-1 knockdown. Here we found that the components of the PeBoW complex had no effect on AFTS-1 levels but also prevented the increase in ATFS-1 protein resulting from cco-1 knockdown (Fig. 5C). Thus, these data suggest that the PeBoW complex may regulate the UPR^mt at several levels, including the transcriptional activity of the ATFS-1/UBL-5/DVE-1 complex as well as through direct governance of ATFS-1 protein abundance. To investigate whether additional components of the ribosome biogenesis pathway influence UPR^mt activation, we performed RNAi knockdown of selected genes and assessed induction of the HSP-6::GFP reporter following treatment with 1 mM paraquat. Knockdown of fib-1 and nath-10 significantly reduced HSP-6::GFP expression, whereas garr-1 and drsh-1 had no detectable effect (Fig. 5D). As the ribosome biogenesis pathway involves >200 proteins (Jiao et al. 2023), a more systematic analysis will be necessary to define the broader regulatory relationship between this pathway and the UPR^mt. Finally, we sought to assess whether other canonical longevity pathways require the PeBoW complex. To this end, we examined the effects of RNAi-mediated knockdown of PeBoW components in daf-2 mutants, which exhibit extended life span independent of the mitochondrial unfolded protein response (Durieux et al. 2011). Notably, knockdown of lpd-7 and bop-1 attenuated the life span extension in daf-2 animals, suggesting that PeBoW function contributes to longevity even outside the context of UPR^mt activation (Fig. 5E).

Figure 5.

PeBoW complex components regulate the UPR^mt downstream from ATFS-1. (A) UBL-5::GFP nuclear intensity in animals treated with lpd-7, bop-1, or wdr-12 RNAi. (B) hsp-6 mRNA levels in atfs-1 gain-of-function (atfs-1 GOF) animals treated with lpd-7 and bop-1 RNAi. (C) Immunoblots of ATFS-1::GFP in animals treated with RNAi for individual PeBoW components or combined with cco-1 RNAi. (D) HSP-6::GFP fluorescence intensity in animals treated with RNAi targeting selected components of the ribosome biogenesis pathway following exposure to 1 mM paraquat. (E) Data survival curve for daf-2-defective animals treated with RNAi for lpd-7, bop-1, or wdr-12. Data are represented as mean ± SD. (*) P < 0.05, (***) P < 0.001, (****) P < 0.0001, (ns) nonsignificant. Assays were by one-way ANOVA (A–D) or log rank analysis (E).

Discussion

Activation of the UPR^mt is central to the maintenance of mitochondrial, cellular, and organismal homeostasis (Melber and Haynes 2018). In the work presented here, we show that reduced vdac-1 expression prompts mitochondrial stress and activates the UPR^mt. The consequences of this UPR^mt activation are increased life span dependent on central effectors of the UPR^mt. To expand our understanding of these effector mechanisms, using unbiased genomics, we uncovered components of the PeBoW complex as essential to full animation of the UPR^mt in response to loss of vdac-1. Demonstrating the broader importance of this finding, PeBoW complex components are required across a wide array of mitochondrial stressors in order to transcriptionally activate the UPR^mt. Knockdown of PeBoW complex components prevented increased longevity from genetic or chemical disruption of mitochondrial function by acting downstream from transcription factor ATFS-1 to prevent transcription of UPR^mt stress response genes and reduce levels of ATFS-1. Importantly, this response did not generalize to other stress response pathways, indicating a highly specific mechanistic interaction of PeBoW with the UPR^mt.

Voltage-dependent anion channel (VDAC) is an outer mitochondrial membrane protein essential for metabolite transport in and out of mitochondria and plays a key role in cellular metabolism by interacting with >100 different proteins (Shoshan-Barmatz et al. 2010). We previously demonstrated that high VDAC-1 levels in C. elegans reduce longevity by activating mitochondrial permeability transition (Zhou et al. 2019). Based on this finding, we hypothesized that lower VDAC-1 levels might promote longevity by reducing mitochondrial permeability transition. While we cannot determine conclusively that reduced VDAC-1 impacts aging through effects on mitochondrial permeability, the dominant impact of VDAC-1 loss appears to be induction of mitochondrial unfolded protein stress. In this way, reduced vdac-1 function is similar to other forms of mitochondrial stress, such as impaired mitochondrial translation and reduced electron transport chain function (Durieux et al. 2011; Houtkooper et al. 2013; Haynes and Hekimi 2022), because vdac-1 knockdown activates the UPR^mt in a mechanistically similar manner with similar genetic dependencies. However, how precisely reduced function of vdac-1 induces mitochondrial stress remains unknown, though we speculate that its function in metabolite transport and mitochondrial quality control through mitophagy could play a role (Ham et al. 2020; Varughese et al. 2021).

Evidence presented here suggests that reduced vdac-1 has distinct differences from other forms of mitochondrial stress. Explicitly, when compared with other forms of mitochondrial stress, reduced vdac-1 did not activate parallel cellular stress defense pathways, including oxidative stress, endoplasmic reticulum (ER) stress, or increased fatty acid β-oxidation. While this led us to search for specific features of vdac-1 deficiency that distinguish it from other mitochondrial stressors, our unbiased search identified components of the PeBoW complex as generally required across multiple paradigms of mitochondrial stress for full animation of the UPR^mt. As we did not identify more specific machinery required for specific animation of the UPR^mt in response to reduced VDAC-1 function, additional efforts will be required to uncover specifically how its loss compromises mitochondrial biology.

Reducing vdac-1 levels induced relatively mild UPR^mt activation compared with other forms of mitochondrial stress, such as reduced electron transport chain (ETC) capacity and impaired mitochondrial translation. This may explain why the suppression of hsp-6p::GFP by lpd-7 RNAi was readily detectable in our vdac-1 suppression screen, whereas cco-1 knockdown with lpd-7 RNAi still exhibited residual hsp-6::GFP activation. For the same reasons, we speculate that lpd-7 or other proteins in the PeBoW complex did not appear in previously published RNAi suppression screens for hsp-6::GFP inducers, including antimycin, atp-2 (RNAi), and isp-1(qm150) (Liu et al. 2014). Nonetheless, lpd-7 was identified in a genome-wide RNAi suppressor screen for regulators of the UPR^mt caused by reduced mitochondrial fusion but was not further mechanistically examined (Haeussler et al. 2021).

PeBoW is a heteromeric, evolutionarily conserved complex classically associated with ribosomal RNA (rRNA) processing and biogenesis (Hölzel et al. 2005). However, its individual components have also been implicated in distinct cellular functions outside of their role in rRNA biology. PES1/LPD-7 plays a role in telomerase assembly, BOP1/BOP-1 is involved in endocytosis, and WDR-12 is a component of Notch signaling (Nal et al. 2002; Kellner et al. 2015; Cheng et al. 2019; Li et al. 2022; Keer et al. 2024). With this study, we have identified an additional critical function for protein members of the PeBoW complex as regulators of the UPR^mt.

Importantly, inactivation of individual PeBoW complex proteins differed with regard to impact on UPR^mt activation and survival. Reduced levels of lpd-7 and bop-1 shortened survival of paraquat, cco-1 knockdown-treated animals and isp-1(qm150) mutants. However, reductions in WDR-12 had variable impact, sometimes further increasing UPR^mt transcriptional responses upon exposure to mitochondrial stress and other times mitigating UPR^mt-dependent life span extension (i.e., with combined cco-1 knockdown). Thus, in response to distinct mitochondrial stressors, previously identified PeBoW proteins may regulate the UPR^mt in opposing ways, with lpd-7 and bop-1 acting as positive regulators and wdr-12 acting as a negative regulator. Speculatively, in situations where lpd-7 and bop-1 demonstrate opposite phenotypes from wdr-12, this could indicate either that the canonical PeBoW complex has distinct outputs in addition to its role in rRNA biogenesis or that a complex with distinct membership from PeBoW governs the UPR^mt. In the case of the latter, formation of PeBoW (containing all three members) might compete with the UPR^mt regulatory complex (consisting of LPD-7 and BOP-1). Thus, loss of wdr-12 would encourage more assembly of the UPR^mt regulatory LPD-7/BOP-1 complex, potentially explaining why wdr-12 appeared to be a negative regulator of the UPR^mt. As these conclusions are speculative, future studies will be required to understand the precise protein complex involving LPD-7 and BOP-1 that is required in the specific context of mitochondrial stress defense animation and in response to discrete mitochondrial stressors.

Based on variable and disparate effects of lpd-7 and bop-1 versus wdr-12, we speculate that PeBoW's UPR^mt regulation may not be related to its canonical functions in ribosome biogenesis. In keeping with this possibility, previous studies have reported that LPD-7/PES1 and BOP-1/BOP1 directly interact and that this interaction with BOP1 is crucial for PES1 function (Lapik et al. 2004). Additionally, BOP-1/BOP1 localization to the nucleolus depends on LPD-7/PES1, whereas WDR12 can translocate independently (Rohrmoser et al. 2007). Identifying the specific subcellular compartment where their activity is required may provide insight into their mechanisms of action in mitochondrial stress defense activation. Further work will be required to substantiate this possibility.

Our results suggest that knockdown of PeBoW complex components reduced UPR^mt animation by multiple, complementary mechanisms. First, PeBoW knockdown reduced ATFS-1 protein levels under mitochondrial stress. Given the role of PeBoW in ribosome biogenesis, it is plausible that reduced ATFS-1 protein levels result from an impact on protein translation. Polysome analysis under conditions of reduced ribosome biogenesis genes has shown reductions in polysome abundance, as expected following reductions in rRNA synthesis (Wu et al. 2018). However, previous work did not show an impact of blocking translation initiation on UPR^mt activation, as measured by an hsp-6p::GFP reporter (Tjahjono et al. 2022). An alternative explanation for reduced ATFS-1 levels is that PeBoW may control the selective translation of atfs-1 mRNAs during mitochondrial stress or regulate ATFS-1 protein turnover. This also suggests the exciting possibility that the impact of PeBoW complex reduction on survival from mitochondrial stress is independent of potential changes in global protein translation.

Second, the knockdown of PeBoW complex components interferes with transcriptional induction of the UPR^mt upon genetic activation of the master transcriptional regulator ATFS-1. This invokes a promoter-proximal interference mechanism by which PeBoW loss, either directly or indirectly, blocks transcriptional activity of ATF family transcription factors. Additional evidence presented here is consistent with the possibility of a promoter-proximal mechanism of PeBoW components in UPR^mt activation. Knockdown of lpd-7 and bop-1 increased nuclear localization of UBL-5, a chromatin modifier that interacts with DVE-1 to promote UPR^mt gene transcription and is essential for life span extension in response to mitochondrial impairment (Haynes et al. 2007). Precise molecular determination of this proposed impact of LPD-7 and BOP-1 will be the subject of future studies. Finally, the ability of lpd-7 and bop-1 knockdown to suppress life span extension in daf-2 animals presents an exciting opportunity to explore how regulators of ribosome biogenesis influence longevity across distinct genetic and physiological contexts.

Activation of the UPR^mt in response to stress improves mitochondrial function by increasing mitochondrial biogenesis and ATP production (Shpilka and Haynes 2018). Inducing mitochondrial stress has been proposed as an approach to treating various mitochondrial, neurodegenerative, and cardiac diseases and cancer (Torres et al. Inestrosa 2024). Although the fundamental mechanism of UPR^mt is known, the details of its transcriptional regulatory mechanisms remain incompletely understood. Our work has introduced a new set of regulators for the UPR^mt. Loss of mammalian lpd-7 (PES1) has been reported to regulate senescence in cultured cells and in vivo (Cheng et al. 2019). Further studies are required to explore whether genetic or pharmacological targeting of this pathway could have implications for mitochondrial and other diseases.

Materials and methods

C. elegans genetics

All C. elegans strains were maintained at 20°C on standard nematode growth medium (NGM) plates with Escherichia coli OP50 bacteria. All experiments were conducted at 20°C unless otherwise specified. The strains used were as follows: wild type (N2 Bristol ancestral), SJ4100 (zcIs13[hsp-6p∷GFP]), CL2070 (dvIs70[hsp-16.2p∷GFP]), SJ4005 (zcIs4[hsp-4p∷GFP]), VC2654 [ubl-5(ok3389)], SJ4151 (zcIs19[ubl-5p::ubl- 5::GFP]), WBM392 {Is[acs-2p::GFP + rol­6(su1006)]}, VC3201 [atfs-1(gk3094)], DG4324 (tn1765[GFP::3xFLAG::pod-2]), AG400 (av138[fasn-1::GFP]), CL2166 {dvls19[(pAF15)gst-4p::GFP::NLS]}, VC617 {lpd-7(ok870) III/hT2 [bli-4(e937) let-?(q782) qIs48]}, WM27 [rde-1(ne219)], MT17463 [set-25(n5021)], VC967 [set-32(ok1457)], MGH613 (vdac-1p::vdac-1::GFP::AID::3XFLAG), and CB1370 [daf-2(e1370)].

RNA interference and E. coli strains

RNA interference (RNAi) experiments were conducted using E. coli HT115 (DE3) bacteria (Ahringer Library). RNAi clones for genes of interest were isolated from the RNAi library by streaking onto LB agar plates containing ampicillin and tetracycline. After individual clones were verified by sequencing, they were grown in LB broth for 15 h with shaking and then concentrated 1:5 before being seeded onto NGM plates supplemented with 200 mg/mL carbenicillin and 5 mM isopropyl-β-D-thiogalactopyranoside (IPTG). Plates were used for experiments 2–4 days after seeding. For dual/tandem/mixed RNAi experiments, a 500 bp sequence of cco-1 was fused to a 500 bp sequence of either lpd-7, bop-1, or wdr-12 and cloned into an L4440 plasmid by Gibson assembly, which was then transformed into competent HT115 bacteria.

Longevity assays

Life span analysis was conducted at 20°C. Gravid adults were grown on NGM plates, and eggs were isolated using hypochlorite bleaching before being incubated overnight in M9 solution. Synchronized L1 larvae were transferred onto RNAi plates and allowed to grow until adulthood.

For each RNAi condition, 40–50 worms were transferred to three separate plates. Survival was assessed every other day by observing movement or using gentle prodding to confirm viability. Data were analyzed using the OASIS2 software package (https://sbi.postech.ac.kr/oasis2).

Nile Red staining for body fat

Synchronized day 1 adult animals were collected in M9 solution and washed three times. The washed animals were then fixed with 40% isopropanol for 3 min with shaking, followed by staining with 3 µg/mL Nile Red in 40% isopropanol for 2 h. After a 30 min wash in M9 with shaking, the animals were mounted on plain glass slides and imaged using the GFP channel with a 10 msec exposure at 5× magnification (as described by Pino et al. 2013). Body fat mass was quantified using ImageJ by measuring fluorescence intensity relative to body area.

Western blotting

Synchronized animals were collected at the required stage and washed three times in M9 solution. Worms were lysed in RIPA buffer containing a protease inhibitor cocktail (Roche) using a water bath sonicator (Diagenode Bioruptor XL 4). Lysates were then centrifuged at 21,000g for 15 min at 4°C, and the supernatants were collected.

Protein concentration was measured using the Pierce BCA assay (Thermo Fisher). Lysates were then mixed with 4× Laemmli buffer (Bio-Rad) and incubated for 10 min at 95°C. Samples were run on SDS-PAGE and transferred to a nitrocellulose membrane for 1 h at 100 V.

Data quantification and statistical analysis

Western blots were quantified using Bio-Rad Image Laboratory software. All statistical analyses, except for longevity assays, were performed using Prism 9 (GraphPad).

Quantitative RT-PCR

Worms were collected, flash-frozen in liquid nitrogen, and stored at −80°C until RNA extraction. Samples were lysed using metal beads and the TissueLyser system (Qiagen). Total RNA was extracted using RNAzol RT (Molecular Research Center).

Reverse transcription was performed using the QuantiTect reverse transcription kit (Qiagen). Quantitative RT-PCR was conducted with three biological replicates using the QuantiTect SYBR Green PCR kit (Qiagen) on a Bio-Rad CFX96 real-time PCR machine (Bio-Rad). Expression levels of tested genes were normalized to the control gene (actin) using the ΔΔCt method. The following primers were used for qPCR: vdac-1 forward (AATGGAACACCGAGAACCAG), vdac-1 reverse (CAGCGTTGATAACTGGAGCA), act-1 forward (TGCTGATCGTATGCAGAAGG), act-1 reverse (TAGATCCTCCGATCCAGACG), lpd-7 forward (CCACAAGAAGCAAGCAAACA), lpd-7 reverse (TGATTCAATTTACGCAGGAACA), Y48B6A.1/bop-1 forward (AATGATCTACGCCCGTTCCC), Y48B6A.1/bop-1 reverse (GATTGCACGGACTTGGCC), wdr-12 forward (TGGTTACGATTGGAGGGCAT), wdr-12 reverse (GGTCCTTTGATCCGCGAAAT), cco-1 forward (AGATCCACTTGAGCACGCTA), cco-1 reverse (GATCGCTGCACAAACTGGAA), hsp-6 forward (GTTATCGAGAACGCAGAAGGAG), hsp-6 reverse (CATCCTTAGTAGCTTGACGCTG), hsp-60 forward (CATGCTCGTCGGAGTCAAC), hsp-60 reverse (TTTGTGATCTTTGGGCTTCC), cdr-2 forward (CGAGCCTCATTTGGAAAGAA), and cdr-2 reverse (GCATCTGCCGCTGTAACTTT).

Microscopy

All images were obtained by mounting worms on a 2% agarose pad with 2.5 mM levamisole using the Leica Thunder imaging system. Mean fluorescence from images was analyzed using ImageJ.

Competing interest statement

A.A.S. has financial interests in Atman Health, LLC, a company developing an AI-based platform for remote clinical care. A.A.S.’s interest was reviewed and is managed by Massachusetts General Hospital and Mass General Brigham in accordance with their conflict of interest policies.