FSHR-1/GPCR Regulates the Mitochondrial Unfolded Protein Response in Caenorhabditis elegans

Abstract

The mitochondrial unfolded protein response (UPR^mt) is an evolutionarily conserved adaptive response that functions to maintain mitochondrial homeostasis following mitochondrial damage. In Caenorhabditis elegans, the nervous system plays a central role in responding to mitochondrial stress by releasing endocrine signals that act upon distal tissues to activate the UPR^mt. The mechanisms by which mitochondrial stress is sensed by neurons and transmitted to distal tissues are not fully understood. Here, we identify a role for the conserved follicle-stimulating hormone G protein-coupled receptor, FSHR-1, in promoting UPR^mt activation. Genetic deficiency of fshr-1 severely attenuates UPR^mt activation and organism-wide survival in response to mitochondrial stress. FSHR-1 functions in a common genetic pathway with SPHK-1/sphingosine kinase to promote UPR^mt activation, and FSHR-1 regulates the mitochondrial association of SPHK-1 in the intestine. Through tissue-specific rescue assays, we show that FSHR-1 functions in neurons to activate the UPR^mt, to promote mitochondrial association of SPHK-1 in the intestine, and to promote organism-wide survival in response to mitochondrial stress. We propose that FSHR-1 functions cell nonautonomously in neurons to activate UPR^mt upstream of SPHK-1 signaling in the intestine.

THE mitochondrial unfolded protein response (UPR^mt) functions to maintain mitochondrial protein homeostasis in response to mitochondrial dysfunction caused by mitochondrial DNA damage, incorrect mitochondrial protein folding, or impaired oxidative phosphorylation. Failure to appropriately control and maintain protein homeostasis in the mitochondria is associated with the development of numerous diseases, neurodegeneration, and ageing (Durieux et al. 2011; Liu et al. 2014; Pellegrino et al. 2014; Nargund et al. 2015; Fiorese et al. 2016; Martinez et al. 2017). The UPR^mt is initiated when mitochondrial proteostasis is disrupted, the detection of which by mitochondrial and cytosolic factors leads to epigenetic modifications and transcriptional responses in the nucleus that restore mitochondrial function. A critical sensor and activator of the UPR^mt in Caenorhabditis elegans and in mammals is the leucine zipper transcription factor ATSF-1/ATF5 (Fiorese et al. 2016). ATFS-1 is normally targeted to mitochondria where it is degraded, but upon mitochondrial stress, mitochondrial import is disrupted and ATFS-1 is targeted instead to the nucleus where it regulates the expression of a cascade of genes including the conserved HSP70-like chaperone, hsp-6, which is targeted to the mitochondria to restore protein folding (Nargund et al. 2012). The intestinal UPR^mt can be activated by mitochondrial stress originating either cell autonomously in the intestine or cell nonautonomously in the nervous system. Mitochondrial stress in neurons activates the UPR^mt in the intestine through the release of neuropeptides, serotonin, and/or Wnt ligands (Berendzen et al. 2016; Shao et al. 2016; Zhang et al. 2018).

Genetic screens for additional factors that activate the UPR^mt have revealed important roles for mitochondrial ceramide produced by SPTL-1/serine palmitoyltransferase and sphingosine-1-phosphate (S1P) produced by SPHK-1/sphingosine kinase in the activation of the UPR^mt (Liu et al. 2014; Kim and Sieburth 2018a). SPHK-1 recruitment to mitochondria from cytoplasmic pools may serve as an early signal to activate UPR^mt (Kim and Sieburth 2018a). SPHK-1 mitochondrial recruitment is positively regulated by mitochondrial stress originating either from the intestine or from the nervous system. Neuronal mitochondrial stress activates intestinal SPHK-1 by a mechanism that involves neuropeptide, but not serotonin, signaling (Kim and Sieburth 2018a). Neuropeptides exert their biological functions primarily through activating G protein-coupled receptors (GPCRs) on target cells to trigger downstream signaling events (Frooninckx et al. 2012). However, the specific neuropeptides and the GPCRs functioning in SPHK-1-mediated UPR^mt activation have not been identified.

FSHR-1 is a GPCR-containing extracellular Leucine-Rich Repeats (LRRs) protein that is homologous to the follicle-stimulating hormone receptor family (Powell et al. 2009). FSHR-1 plays a critical role in activating innate immunity in response to infection by pathogenic bacteria, and functions in the intestine to promote protection against pathogenic infection and to regulate antimicrobial gene expression. (Cho et al. 2007; Powell et al. 2009; Miller et al. 2015). Interestingly, infection by pathogenic bacteria leads to multiple cellular responses in the intestine, including the activation of the UPR^mt (Liu et al. 2014; Pellegrino et al. 2014).

In this study, through the analysis of fshr-1 null mutants, we show that FSHR-1 positively regulates UPR^mt activation and promotes the mitochondrial association of SPHK-1 in the intestine. Through tissue-specific rescue experiments, we find that FSHR-1 functions in the nervous system to exert its function in UPR^mt activation. We propose that FSHR-1 is part of a neuroendocrine signaling network that functions to regulate the UPR^mt through intertissue signaling.

Materials and Methods

C. elegans strains

Strains used in this study were maintained at 22° following standard methods. Young adult hermaphrodites derived from the wild-type reference strain N2 Bristol were used for all experiments. The following mutant strains were used. SJ4100: zcIs13[Phsp-6::GFP], OJ4113: vjIs138[Pges-1::sphk-1::gfp], OJ4143: vjIs148[Pges-1::tomm-20::mCherry], ST66: ncIs17[Phsp-16.2::gfp], OJ2329: vjIs208[Pgst-4::gfp], uthIs375[unc-119p::cco-1 HP + rol-6(su1006)](Zhang et al. 2018), OJ997: sphk-1(ok1097), GR2250: mgIs73[cyp-14A4p::gfp], and KP3397: fshr-1(ok778). The sphk-1(ok1097) and fshr-1(ok778) strains were outcrossed at least six times with wild-type animals prior to analysis.

Molecular biology

fshr-1 complementary cDNA (cDNA) was cloned from C. elegans wild-type cDNA and then inserted into the pPD49.26 expression vector using standard molecular biology techniques. The following plasmids were generated: pSK55[Pges-1::fshr-1], pSK56[Prab-3::fshr-1], and pSK98[Prgef-1::fshr-1]. Sequence files of plasmids are available upon request.

Transgenic lines

Transgenic strains were generated by injecting expression constructs (10–25 ng/μl) and the co-injection marker pJQ70 [Pofm-1::mCherry, 40 ng/μl], KP#708[Pttx-3::rfp, 40 ng/μl], or KP#1106 [Pmyo-2::gfp, 10 ng/μl] into N2 worms or the indicated mutants using standard techniques (Mello et al. 1991). At least three lines for each transgene were tested and a representative transgene was used for further experiments. The following transgenic lines were made: vjEx1448[Prab-3::fshr-1], vjEx1449[Pges-1::fshr-1], vjEx1000[Pges-1::gfp], and vjEx1831[Prgef-1::fshr-1].

Microscopy and analysis

Fluorescence microscopy experiments were performed following previous methods (Kim and Sieburth 2018a). Briefly, L4-stage or young adult worms were immobilized by using 2,3-butanedione monoxime (BDM, 30 mg/ml; Sigma [Sigma Chemical], St. Louis, MO) in M9 buffer then mounted on 2% agarose pads for imaging. To quantify the fluorescence intensity of Phsp-6::GFP or Pgst-4::GFP, Z stacks of the intestine posterior to the vulva were selected as a representative region because of low basal expression in the absence of stress. Images were captured with the Nikon (Garden City, NY) eclipse 90i microscope equipped with a Nikon PlanApo 40×, or 60× or 100× objective (NA = 1.4), and a PhotometricsCoolsnap ES2 or a Hamamatsu Orca Flash LT + CMOS camera. Metamorph 7.0 software (Universal Imaging/Molecular Devices) was used to capture serial image stacks, and the maximum intensity was measured (Kim and Sieburth 2018b). Intensity quantification analysis was performed on the same day to equalize the absolute fluorescence levels between samples within the same experimental set. For Figure 1A, each sample cluster of worms was captured using a 4× objective lens for the representative images as shown in Figure 1A, then captured by a 40× objective lens for fluorescence intensity quantification.

Figure 1.

FSHR-1 promotes UPR^mt activation. (A) Representative low-magnification images of several wt adults or fshr-1 mutants expressing the zcIs13 [Phsp-6::GFP] transgene (GFP driven by the hsp-6 promoter), following treatment with M9 (control) or PQ for 24 hr. (B and C) Left: representative high-magnification images of intestines posterior to the vulva in young adult wt animals or fshr-1 mutants expressing Phsp-6::GFP following treatment with M9 (control) or PQ (B), or DMSO (control) or antimycin (C) for 24 hr. Right: average GFP fluorescence intensity is quantified. (D) Left: representative images of Phsp-6::GFP fluorescence in intestines of wt or fshr-1 mutants following e.v. control or cco-1/cox-5B RNAi treatment by feeding. Right: average GFP fluorescence intensity is quantified. (E and F) Left: representative intestine images of wt or fshr-1 mutants expressing the zcIs9 [Phsp-60::GFP] transgene following M9 and e.v. (G) (control), or PQ or spg-7/AFG3L2 RNAi treatment. Right: average GFP fluorescence intensity is quantified. (G) Left: representative images of wt or fshr-1 mutants expressing the mgIs73 [Pcyp-14A4::gfp] transgene in the absence or presence of PQ. Right: average GFP fluorescence intensity is quantified. (H) Left: representative images of posterior intestines from animals expressing the antioxidant reporter transgene [Pgst-4::GFP] in wt or fshr-1 mutants in the absence or presence of As. Right: average GFP fluorescence intensity is quantified. (I) Left: representative images of posterior intestines from animals expressing heat shock transgene ncIs17 [Phsp-16.2::GFP] in wt or fshr-1 mutants following 22° (control) or 37° exposure for 20 min. Right: average GFP fluorescence intensity is quantified. Bar represents 10 µm. Error bars indicate ± SEM. One-way ANOVA with Bonferroni post-test *P < 0.1; **P < 0.01; ***P < 0.001. As, arsenite; e.v., empty vector; ns, not statistically significant; PQ, paraquat; RNAi, RNA interference; UPR^mt, mitochondrial unfolded protein response; wt, wild-type.

RNA interference

A feeding RNA interference (RNAi) knockdown assay was performed following established protocols (Kamath and Ahringer 2003). Briefly, gravid adult animals were placed on RNAi plates seeded with HT115(DE3) bacteria transformed with L4440 vector containing a genomic fragment of the gene to be knocked down (or empty L4440 vector as a control) to collect eggs, then removed after 4 hr to obtain an age-matched synchronized worm population. Young adult animals were used for subsequent assays.

Stress induction assays

For drug-induced stress, transgenic L4 animals were transferred to fresh NGM plates seeded with HB101 bacteria, then 80 μl of stock solutions dissolved in M9 buffer of paraquat were added to plates for a final concentration of 0.4 mM paraquat. After 24 hr, adults were selected for fluorescence microscopy analysis. For imaging of animals that had been subjected to RNAi-induced knockdown, L4 animals grown on RNAi plates were transferred to new RNAi plates to obtain synchronized animals and imaged 24 hr later. For Pgst-4::GFP imaging, young adult animals were incubated with 5 mM arsenite or 50 mM paraquat in liquid solution (in M9 buffer) for 1 hr, then drug was washed out and worms were transferred onto new NGM plates for 4 hr before imaging. M9 buffer was used as a control for paraquat and arsenite treatment.

For Phsp-16.2::GFP imaging, young adult animals were incubated at 22° or 37° for 20 min, and then allowed to recover at 22° for 24 hr prior to imaging. For paraquat survival assays, young adult animals were placed onto NGM plates containing 10 mM paraquat for 17 hr. After 17 hr, the percentage of surviving animals was counted every 3 hr over the course of 15 hr. Survival assays were done in experimental duplicate for each biological duplicate.

Statistical analysis

We utilized one-way ANOVA with a Bonferroni post-test to evaluate the statistical significance for multiple sample comparisons including Figure 3B (26 hr), while Student’s t-tests were used for two-sample comparison using Graphpad Prizm 8. P-values < 0.01 or 0.001 are indicated with asterisks (**P < 0.01 and ***P < 0.001, respectively). Error bars in the figures indicate the SE of the mean (± SEM). At least 15 animals per sample were analyzed for fluorescence quantification and data points, and averages are presented as shaped box plots in the figures.

Figure 3.

FSHR-1 functions in a common pathway with SPHK-1 to activate the UPR^mt. (A) Left: representative images of Phsp-6::GFP expression in the intestines of wild-type, sphk-1 mutants, fshr-1 mutants, and sphk-1;fshr-1 double mutants in the absence or the presence of PQ for 24 hr. Right: average GFP fluorescence intensity is quantified. (B) Survival rate curves of the indicated strains on plates containing 10 mM PQ. Bar represents 10 µm. One-way ANOVA with Bonferroni post-test: @P < 0.001 compared to sphk-1; &P < 0.001 compared to sphk-1; #P < 0.001 compared to fshr-1; $P < 0.001 compared to fshr-1; and ^P < 0.001 compared to sphk-1;fshr-1; ns: P > 0.05 compared to fshr-1;FSHR-1 neuron rescue; *P < 0.05 compared to wt; and ***P < 0.001 compared to wt. ns, not statistically significant; PQ, paraquat; UPR^mt, mitochondrial unfolded protein response; wt, wild-type.

Data availability

Strains and plasmids used in this study are available upon request. The authors confirm that all data necessary for confirming the conclusions of the findings are present within the article and figures.

Results

FSHR-1/GPCR signaling regulates the UPR^mt

Prior studies found that fshr-1 mutants have normal life spans (Powell et al. 2009), but exhibit enhanced sensitivity to lethality caused by the mitochondrial toxin and UPR^mt activator paraquat (Miller et al. 2015). Because impairing UPR^mt activation causes sensitivity to paraquat-induced lethality (Nargund et al. 2012; Gatsi et al. 2014; Liu et al. 2014; Kim and Sieburth 2018a), we speculated that FSHR-1 may positively regulate the UPR^mt. To monitor UPR^mt activation, we quantified the intestinal fluorescence of the UPR^mt transcriptional reporter zcIs13, in which GFP is expressed under control of the hsp-6 promoter (Phsp-6::GFP). Paraquat is an oxidant that interferes with electron transport at the inner mitochondrial membrane, and has been used widely to acutely activate the UPR^mt in both C. elegans and in mammalian cells (Nargund et al. 2012; Runkel et al. 2013; Fiorese et al. 2016; Kim and Sieburth 2018a). Wild-type animals treated with paraquat for 24 hr exhibited a roughly 10-fold increase in Phsp-6::GFP expression in the intestine compared to nontreated controls (Figure 1, A and B). In contrast, fshr-1 mutants showed a slight increase in baseline Phsp-6::GFP expression, and either no significant change or a small increase in Phsp-6::GFP expression following paraquat treatment (Figure 1, A and B). Antimycin is a mitochondrial electron transport chain (ETC) stressor, and antimycin treatment for 24 hr robustly induces the UPR^mt (Liu et al. 2014). fshr-1 mutations attenuated the antimycin-induced increase in Phsp-6::GFP expression (Figure 1C). cco-1 (aka cox-5B) encodes the cytochrome c oxidase subunit in complex IV, which is the terminal electron acceptor of the ETC, and RNAi-mediated cco-1 knockdown is a potent activator of the UPR^mt (Nargund et al. 2012; Berendzen et al. 2016; Merkwirth et al. 2016; Tian et al. 2016; Kim and Sieburth 2018a). cco-1 RNAi by feeding, which induces mitochondrial stress in the intestine (Zhang et al. 2018), significantly increased Phsp-6::GFP expression in wild-type animals but failed to increase Phsp-6::GFP expression in fshr-1 mutants (Figure 1D). fshr-1 mutations also blocked or attenuated the stress-induced increase in expression of two additional UPR^mt reporters: Phsp-60::GFP and Pcyp-14A4::GFP [Figure 1, E–G, Nargund et al. (2012), and Mao et al. (2019)].

fshr-1 mutants were not defective in the induction of the antioxidant reporter Pgst-4::GFP in response to treatment with the mitochondrial reactive oxygen species (ROS) generator arsenite [Figure 1H; 1.6-fold increase in wild-type and 1.9-fold increase in fshr-1 mutants by arsenite (Inoue et al. 2005; Choe et al. 2009; Ruiz-Ramos et al. 2009; Prakash et al. 2015; Wu et al. 2016)]. In addition, fshr-1 mutants exhibited normal induction of the heat-shock reporter Phsp-16.2::GFP in response to heat shock at 37° for 20 min (Figure 1I). Together, these results suggest that FSHR-1 plays a specific role in activating the UPR^mt in response to mitochondrial stress.

FSHR-1/GPCR functions cell nonautonomously in neurons to regulate UPR^mt

FSHR-1 is expressed primarily in the intestine and in a subset of neurons (Sieburth et al. 2005; Cho et al. 2007; Cao et al. 2017). To determine in which tissue FSHR-1 functions to regulate the UPR^mt, we examined Phsp-6::GFP induction in transgenic fshr-1 mutants expressing fshr-1 cDNA in either the nervous system (using the rab-3 or the rgef-1 promoter) or the intestine (using the ges-1 promoter). Pan-neuronal fshr-1 either partially (rab-3 promoter) or fully (rgef-1 promoter) rescued the paraquat-induced hsp-6 expression defects of fshr-1 mutants (Figure 2, A and B). In contrast, intestinal fshr-1 cDNA expression failed to rescue the paraquat-induced hsp-6 expression defects of fshr-1 mutants (Figure 2A). RNAi-mediated knockdown of fshr-1 by bacterial feeding, which efficiently knocks down genes in the intestine but not in the nervous system (Kamath et al. 2001; Timmons et al. 2001; Asikainen et al. 2005), did not attenuate paraquat-induced hsp-6 expression, confirming that fshr-1 does not function in the intestine to regulate the UPR^mt (Figure 2D). Taken together, these results suggest that FSHR-1 primarily functions cell nonautonomously in the nervous system to positively regulate UPR^mt activation.

Figure 2.

FSHR-1 functions in the nervous system to regulate the UPR^mt. (A and B) Left: representative images of posterior intestines of young adult animals or fshr-1 mutants expressing the Phsp-6::GFP transgene following treatment with M9 (control) or PQ for 24 hr. “fshr-1 neuronal rescue” denotes fshr-1 mutants expressing fshr-1 cDNA under the neuron-specific promoter rab-3 (A) or rgef-1 (B). “fshr-1 intestine rescue” denotes fshr-1 mutants expressing fshr-1 cDNA under the intestine-specific promoter ges-1. Right: average GFP fluorescence intensity is quantified. (C) Left: representative images of intestines of wt or fshr-1 mutants expressing Phsp-6::GFP following e.v. control or atfs-1 RNAi treatment by feeding. Right: average GFP fluorescence intensity is quantified. (D) Left: representative images of intestines of wt worms expressing Phsp-6::GFP following e.v. control or fshr-1 RNAi treatment by feeding in the absence or presence of PQ. Right: average GFP fluorescence intensity is quantified. (E) Left: representative images of Phsp-6::GFP fluorescence in intestines of wt or fshr-1 mutants in the absence or presence of transgene uthIs375[unc-119p::cco-1 HP + rol-6(su1006)]). Right: average GFP fluorescence intensity is quantified. (F) Left: representative images of intestinal cells of wt or fshr-1 mutants expressing the mitochondrial marker TOMM-20::mCherry under the intestinal-specific promoter ges-1. Right: average mCherry fluorescence intensities are quantified. (G) Left: representative images of animals expressing soluble GFP in the intestine (Pges-1::GFP) in wt or fshr-1 mutants. Right: average GFP fluorescence intensities are quantified. Bar in the panel represents 10 µm. Error bars indicate ± SEM. One-way ANOVA with Bonferroni post-test ***P < 0.001; #P < 0.001 compared to fshr-1; Prab-3::fshr-1; @P > 0.05 compared to wt; $P > 0.05 compared to wt; and &P > 0.05 compared to wt + PQ. cDNA, complementary DNA; e.v., empty vector; ns, not statistically significant; PQ, paraquat; RNAi, RNA interference; UPR^mt, mitochondrial unfolded protein response; wt, wild-type.

The intestinal UPR^mt can be activated by cell-autonomous stress originating in the intestine or by cell-nonautonomous stress originating in the nervous system (Nargund et al. 2012; Berendzen et al. 2016; Merkwirth et al. 2016; Zhang et al. 2018). fshr-1 is required for proper UPR^mt activation elicited by intestinal stress (Figure 1, D and F). To determine whether fshr-1 promotes UPR^mt activation elicited by neuronal stress, we examined Phsp-6::GFP expression in animals in which cco-1 was selectively silenced in the nervous system. Neuronal cco-1 knockdown by hairpin-mediated RNAi is a potent activator of the intestinal UPR^mt (Zhang et al. 2018). Neuronal cco-1 knockdown elicited similar intestinal Phsp-6::GFP expression in fshr-1 mutants as in wild-type controls (Figure 2E). The fold induction was slightly reduced in fshr-1 mutants [4.3-fold for fshr-1(+) vs. 2.2 fold for fshr-1(-)] due to increased baseline hsp-6 expression in fshr-1 mutants (Figure 2E). Thus, fshr-1 appears to be largely dispensable for UPR^mt activation when stress originates from the nervous system.

Neuronal FSHR-1 regulates baseline hsp-6 and hsp-60 expression

Under nonstressed conditions, fshr-1 mutants exhibited small but significant increases in Phsp-6::GFP expression in the intestine compared to wild-type controls (Figure 1, A–D), and larger increases in Phsp-60::GFP expression (Figure 1, E and F) that could reach levels similar to those seen under stressed conditions (Figure 1E). The increase in Phsp-6::GFP expression was blocked by RNAi-mediated knockdown of atfs-1 (Figure 2C) and was restored to wild-type levels by pan-neuronal, but not by intestinal, fshr-1 cDNA expression (Figure 2, A and B), suggesting that the UPR^mt is activated even in the absence of stress in fshr-1 mutants. Together, these results indicate that neuronal FSHR-1 functions cell nonautonomously to keep UPR^mt activity low in the absence of stress.

Mitochondria marked by the mitochondrial localization signal of TOMM-20 fused to mCherry (TOMM-20::mCherry) are distributed in a highly reticulated pattern within the cytoplasm of intestinal cells [expressed under the ges-1 promoter; Figure 2E, Palikaras et al. (2015), and Kim and Sieburth (2018a)]. fshr-1 mutants exhibited a normal overall pattern of TOMM-20:mCherry fluorescence in intestinal cells (Figure 2E), but the average fluorescence intensity of TOMM-20:mCherry associated with mitochondria was significantly increased by about threefold in fshr-1 mutants compared to wild-type controls (Figure 2F). In control experiments, we found that fshr-1 mutants exhibited no change in the fluorescence intensity of soluble GFP expressed under the ges-1 promoter (Figure 2G), suggesting that transgene expression was not altered in fshr-1 mutants. The increase in TOMM-20 fluorescence intensity in fshr-1 mutants could reflect either an increase in mitochondrial mass, a decrease in TOMM-20 turnover, or an increase in TOMM-20 targeting to the outer mitochondrial membrane. Alterations in mitochondrial mass, morphology, or mitochondrial protein import have been linked to UPR^mt activation (Ungvari et al. 2011; Houtkooper et al. 2013; Lerner et al. 2013; Mouchiroud et al. 2013; Mao et al. 2019).

FSHR-1 and SPHK-1 function in a common pathway to activate the UPR^mt

SPHK-1 functions in the intestine to activate the UPR^mt in response to a variety of mitochondrial stressors, including paraquat (Kim and Sieburth 2018a). To test whether FSHR-1 functions in a common pathway with SPHK-1 to activate the UPR^mt, we examined genetic interactions between sphk-1 and fshr-1 mutants. sphk-1 mutants exhibited a significant reduction in paraquat-induced expression of Phsp-6::GFP (89% reduction), which is similar to that exhibited by fshr-1 mutants [92% reduction; Figure 3A and Kim and Sieburth (2018a)]. Double mutants lacking both sphk-1 and fshr-1 exhibited increased baseline expression of hsp-6, which was similar to that of fshr-1 mutants, and had defects in paraquat-induced hsp-6 induction that were no more severe than those seen in single mutants (93% reduction, Figure 3A).

sphk-1 mutants exhibit reduced survival rates when exposed to toxic levels of paraquat (Kim and Sieburth 2018a), which are similar to those of fshr-1 mutants (Figure 3B). The survival rates of double mutants lacking both sphk-1 and fshr-1 were no more severe than those of either single mutant (Figure 3B). Expression of sphk-1 cDNA in the intestine fully rescued the increased paraquat-induced lethality of sphk-1 mutants (Kim and Sieburth 2018a). Expression of fshr-1 cDNA in the nervous system fully rescued the increased paraquat-induced lethality of fshr-1 mutants, whereas expression of fshr-1 cDNA in the intestine partially restored paraquat sensitivity to fshr-1 mutants [Figure 3B and Miller et al. (2015). Taken together, FSHR-1 and SPHK-1 may function in a common pathway to activate the UPR^mt, and to promote organism-wide protection from mitochondrial stress-induced lethality. Moreover, FSHR-1 functions in the nervous system to activate the UPR^mt and promote survival, whereas SPHK-1 functions exclusively in the intestine.

FSHR-1/GPCR regulates mitochondrial association of SPHK-1 in the intestine

SPHK-1 localizes to intestinal mitochondria and is recruited to mitochondria from cytosolic pools by mitochondrial stress (Kim and Sieburth 2018a). SPHK-1::GFP fusion proteins, which are fully functional in rescuing the UPR^mt defects of sphk-1 mutants, colocalize with TOMM-20::mCherry (Figure 4A). Mitochondrial stress induces a significant increase in mitochondrial SPHK-1::GFP fluorescence, as well as an increase in the number of intestinal cells in which SPHK-1::GFP adopts a mitochondrial distribution pattern [referred to as mitochondrial occupancy; Figure 4B and Kim and Sieburth (2018a)]. To determine whether FSHR-1 regulates SPHK-1 mitochondrial association, we examined the mitochondrial abundance of SPHK-1::GFP fusion proteins before and after paraquat exposure. fshr-1 mutants exhibited two defects in SPHK-1::GFP mitochondrial association. First, fshr-1 mutants exhibited a nearly twofold reduction in SPHK-1::GFP mitochondrial association compared to wild-type controls in the absence of paraquat (Figure 4B). Second, the paraquat-induced increase of mitochondrial SPHK-1::GFP association was completely abolished in fshr-1 mutants (Figure 4B). Thus, FSHR-1 positively regulates SPHK-1 association with mitochondria under nonstressed conditions and FSHR-1 is required for stress-induced SPHK-1 mitochondrial targeting.

Figure 4.

FSHR-1 regulates mitochondrial association of SPHK-1. (A) High-magnification images of SPHK-1::GFP (left), TOMM-20::GFP (middle), and merge (right) in wt animals coexpressing Pges-1::sphk-1::gfp and Pges-1::tomm-20::mCherry transgenes. (B) Left: representative images of the indicated mutants expressing intestinal SPHK-1::GFP in the absence or presence of PQ for 24 hr. fshr-1 neuron and intestine rescue denote transgenic fshr-1 mutants expressing fshr-1 cDNA under the pan-neural promoter rab-3 or the intestinal promoter ges-1, respectively. Middle: average fluorescence intensities in the indicated strains expressing SPHK-1::GFP are quantified. Right: mitochondrial occupancy of indicated transgenes is quantified in the absence or presence of PQ. H, M, or L indicates the percentage of animals showing mitochondrially localized SPHK-1::GFP in more than two-thirds of the intestine (H), one-half of the intestine (M), or only in a few intestinal cells (L). Bar represents 2 µm. One-way ANOVA with Bonferroni post-test **P < 0.01; ***P < 0.001; @P > 0.05 compared to wt; #P > 0.05 compared to fshr-1 mutants; $P > 0.05 compared to wt + PQ; and ^P > 0.05 compared to fshr-1 + PQ. cDNA, complementary DNA; ns, not statistically significant; PQ, paraquat; wt, wild-type.

Neuronal fshr-1 cell-nonautonomously regulates mitochondrial association of SPHK-1 in the intestine

To determine in which tissue FSHR-1 functions to regulate the mitochondrial association of SPHK-1, we quantified SPHK-1::GFP mitochondrial fluorescence in fshr-1 mutants expressing fshr-1 cDNA in either the nervous system or in the intestine. We found that pan-neuronal expression of fshr-1 cDNA in fshr-1 mutants restored both the mitochondrial occupancy and fluorescence intensity of SPHK-1::GFP to wild-type levels in the absence of stress. Similarly, neuronal fshr-1 cDNA also restored normal paraquat-induced SPHK-1::GFP mitochondrial recruitment to fshr-1 mutants. In contrast, intestinal fshr-1 cDNA expression failed to rescue the SPHK-1::GFP fluorescence defects of fshr-1 mutants (Figure 4B). These results suggest that FSHR-1 functions exclusively in the nervous system to positively regulate baseline SPHK-1 abundance in the intestine in the absence of stress, and to promote paraquat-induced SPHK-1 mitochondrial association.

Discussion

The nervous system coordinates various stress responses by releasing diffusible factors that act upon distal tissues to activate cellular defense programs (Nussbaum-Krammer and Morimoto 2014; Berendzen et al. 2016; Shao et al. 2016). Here, we show that the conserved GPCR FSHR-1 is part of an integrated organism-wide response to mitochondrial stress that functions to activate the UPR^mt in the intestine. fshr-1 promotes UPR^mt activation in response to either acute exposure to mitochondrial toxins or to chronic mitochondrial dysfunction, and functions in a common pathway with sphk-1 to promote survival in response to toxic mitochondrial stress and UPR^mt activation. fshr-1 positively regulates the mitochondrial association of SPHK-1 in the intestine in the absence of stress, as well as stress-induced SPHK-1 mitochondrial recruitment. fshr-1 functions in the nervous system, but not in the intestine, to promote UPR^mt activation and stress-induced SPHK-1 mitochondrial recruitment. Finally, neuronal fshr-1 regulates intestinal mitochondrial homeostasis in the absence of stress by a mechanism that may involve regulating mitochondrial mass and/or mitochondrial protein abundance. We propose a model whereby neuronal fshr-1 signaling keeps UPR^mt activity low in the absence of stress and positively regulates the UPR^mt in response to intestinal stress. The underlying mechanism by which fshr-1 regulates the UPR^mt may be by establishing or maintaining proper mitochondrial homeostasis in the intestine.

fshr-1 functions cell nonautonomously in the nervous system to activate the UPR^mt; however, our genetic evidence suggests that it is not likely to function in one of the previously identified “mitokine” pathways that signal neuronal stress to the intestine. Disruption of these pathways (e.g., egl-20/Wnt or flp-2/neuropeptide) blocks the ability of stress originating in the nervous system to promote intestinal UPR^mt activation (Shao et al. 2016; Kim and Ewbank 2018; Zhang et al. 2018). We found that fshr-1 is required for UPR^mt activation in response to intestinal stress but not neuronal stress. Thus, fshr-1 is not likely to be a component of a neuronal mitokine pathway but instead may promote UPR^mt activation by a mechanism that does not involve the direct response to neuronal stress.

Previous studies have established an important role for FSHR-1 in activating the innate immune response following infection with pathogenic bacteria. FSHR-1 promotes both survival following infection and the behavioral avoidance response to pathogenic bacteria, and FSHR-1 activates a number of antimicrobial and antioxidant genes in response to pathogens (Powell et al. 2009; Miller et al. 2015). FSHR-1 is proposed to not participate in the detection of the pathogenic bacteria themselves, but instead to participate in the detection of cellular damage (e.g., ROS) resulting from bacterial infection (Miller et al. 2015). Pathogenic bacterial infection is also a potent activator of the UPR^mt, and UPR^mt activation by pathogens promotes survival by inducing the expression of innate immune genes (Pellegrino et al. 2014). We speculate that fshr-1 may activate the innate immune response by contributing to activation of the UPR^mt in response to a broad array of insults (e.g., pathogenic infection, ROS, and mitochondrial dysfunction). Interestingly, we found that intestinal fshr-1 does not regulate the UPR^mt, yet intestinal fshr-1 expression protects fshr-1 mutants from paraquat toxicity, suggesting that activation of the UPR^mt may not be necessary to elicit resistance to paraquat. Paraquat is an ROS generator that also activates the intestinal antioxidant response, and fshr-1 promotes the expression of antioxidant genes (Powell et al. 2009; Miller et al. 2015). Thus, the paraquat sensitivity of fshr-1 mutants may arise from defects in the antioxidant response in the intestine.

FSHR-1 also regulates intestinal mitochondrial homeostasis in the absence of stress. Whether FSHR-1 does so by regulating mitochondrial mass or mitochondrial protein import, or both, is not clear based on our quantitative imaging of mitochondrial markers. Interestingly, FSHR signaling negatively regulates mitochondrial biogenesis in mammals (Liu et al. 2017). Enhanced mitochondrial biogenesis is correlated with UPR^mt activation. Inducers of mitochondrial biogenesis, such as nicotinamide riboside and poly(ADP-ribose) polymerase inhibitors (PARPi), disrupt mito–nuclear protein homeostasis resulting in UPR^mt activation via the sirtuin 1 (SIRT1) pathway in mammals and C. elegans. (Mouchiroud et al. 2013). In addition, rapamycin and resveratrol, which increase mitochondrial content, also induce UPR^mt activation in C. elegans (Ungvari et al. 2011; Houtkooper et al. 2013; Lerner et al. 2013). In worms, mutants with defects in the import of mitochondrial proteins into the matrix exhibit increased hsp-60 expression that is dependent upon atfs-1 (Nargund et al. 2012). We speculate that the defects in paraquat-induced UPR^mt activation in fshr-1 mutants may be due to an underlying defect in the mitochondrial association of SPHK-1 under normal conditions.

Since FSHR-1 functions to protect animals from diverse stressors, the ligand for FSHR-1 is likely to originate from the host rather than from an exogenous source (Powell et al. 2009). FSHR-1 shares homology with three human GPCRs—FSHR, TSHR, and LHCGR—by virtue of nine LRRs found in its extracellular domain (Dolan et al. 2007). These GPCRs are activated by the heterodimeric glycopeptide hormones FSHα/β, TSHα/β, and LHα/β, respectively. In humans, FSH induces the generation of S1P by stimulating SphK1 activity for granulosa cell proliferation (Hernández-Coronado et al. 2016), suggesting potential functional conservation of ligand-activated FSHR-1 signaling. An additional mammalian glycopeptide hormone family member, thyrostimulin, is a heterodimer composed of the GPB5 α-subunit and the GPA2 β-subunit, which share homology with worm T23B12.8 and FLR-2, respectively (Rocco and Paluzzi 2016). Interestingly, genetic analysis implicates flr-2 in the neuronal control of intestinal functions (Oishi et al. 2009), but whether flr-2 regulates the UPR^mt or innate immunity has not been investigated. Identifying the ligand(s) for FSHR-1 will help to clarify the cell-nonautonomous mechanisms by which FSHR-1 regulates the UPR^mt pathway.