The mitochondrial localized CISD-3.1/CISD-3.2 proteins are required to maintain normal germline structure and function in Caenorhabditis elegans

Reproductive organs and developing tissues have high energy demands that require metabolic functions primarily supported by mitochondria function. The highly conserved CISD/NEET iron-sulfur (Fe-S) protein family regulates iron and reactive oxygen homeostasis, both of which are important for mitochondrial function. Disruption of iron and reactive oxygen homeostasis typically leads to detrimental effects. In humans, CISD dysfunction is associated with human health issues including Wolfram syndrome 2. Using C. elegans, we previously determined that the cisd-1, cisd-3.1 and cisd-3.2 have an overlapping role in the regulation of physiological germline apoptosis through the canonical programmed cell death pathway. Here, we isolated the cisd-3.2(pnIs68) mutant that resulted in physiological and fitness defects including germline abnormalities that are associated with abnormal stem cell niche and disrupted formation of bivalent chromosomes. The cisd-3.2(pnIs68) mutation led to complete disruption of the cisd-3.2 gene expression and a decrease in expression of genetically intact cisd-1 and cisd-3.1 genes suggesting an indirect impact of the cisd-3.2(pnIs68) allele. The CISD-3.2 and CISD-3.1 proteins localize to the mitochondria in many tissues throughout development. The cisd-3.2(pnIs68) mutant displays phenotypes associated with mitochondrial dysfunction, including disruption of the mitochondrial network within the germline. These results further support the idea that the CISD protein family is required for mitochondrial function that supports important functions in animals including overall fitness and germline viability.


Introduction
The highly conserved NEET proteins represented in humans as mitoNEET/CISD1, NAF-1/ CISD2, and MiNT/Miner2/CISD3 constitute a novel class of iron-sulfur cluster proteins germline [16]. Combined, these studies support the idea that the CISD proteins in C. elegans have a role in germline function and metabolic processes. The function of mitoNEET/CISD1 and NAF-1/CISD2 has been studied in various model systems, however little is understood about the function of MiNT/CISD3. Here we used CRISPR to produce a cisd-3.2 deletion mutant (cisd-3.2(pnIs68)) and the CISD-3.1 and CISD-3.2 protein reporters. The cisd-3.2(pnIs68) results in complete disruption of the cisd-3.2 gene and this dysfunction led to severe phenotypes including germline defects, reproductive dysfunction, reduced fitness, and disruption of mitochondrial network within the C. elegans germline. Due to the severe defects observed in the pnIs68 allele relative to the previously described analysis of the cisd-3.2(RNAi) animal, we assessed the expression of the cisd-1 and cisd-3.1 genes. We observed through RT-PCR a decrease in cisd-1 and cisd-3.1 expression despite the loci for these genes being intact. This suggests an indirect impact on the of the cisd-3.2(pnIs68) allele on the cisd gene family. This work further supports the role the CISD family has in mitochondrial functions and proper germ cell differentiation in the germline of C. elegans.

The cisd-3.2(pnIs68) mutant displays abnormal germline development and function
Previously we determined that deletion of the cisd-1 gene and knock-down of cisd-3.1 and cisd-3.2 via RNAi resulted in various germline abnormalities including an increase in cell corpses within the adult germline and abnormal distal tip cell migration [16]. Given that RNAi results in a gene knock-down and such results in less severe phenotypes in comparison to genetic null mutants, we used CRISPR to produce a cisd-3 mutant to better assess the phenotype(s) associated with CISD3 dysfunction. The cisd-3.2(pnIs68) mutant was isolated (S1 Fig) however, we were unable to obtain a stable cisd-3.1 deletion mutant. RT-PCR analysis indicates that the cisd-3.2(pnIs68) allele is a cisd-3.2 null mutation ( Fig 1A). Molecular techniques (RT-PCR and genome sequencing) were used to assess if the cisd-3.1 or cisd-1 genes are impacted by the cisd-3.2(pnIs68) allele. DNA sequencing of the cisd-3.1 and cisd-1 loci in the cisd-3.2(pnIs68) mutant confirmed that there was not untargeted deletion of the cisd-3.1 or cisd-1 genes. However, we did observe a decrease in expression of cisd-3.1 and cisd-1 within the cisd-3.2(pnIs68) mutant (Fig 1B and 1C). We did not observe decreased expression of cisd-3.1 or cisd-3.2 in the cisd-1 mutant (Fig 1D and 1E). In fact, there was a slight increase in cisd-3.2 expression in the cisd-1(pnIs27) mutant. Combined, these data suggest that 1) cisd-1 and cisd-3.1 genes do not compensate for deletion of cisd-3.2; 2) deletion of the cisd-3.2 gene indirectly impacts the expression of cisd-3.1 and cisd-1 and 3) cisd-3.2 may compensate for disruption of cisd-1.
Phenotypic analysis was conducted to gain a deeper understanding of the impact of the cisd-3.2(pnIs68) allele. The cisd-3.2(pnIs68) adult hermaphrodite exhibits an abnormal adult germline morphology (Fig 2A). In N2 wild-type hermaphrodites, the germline progresses through developmental stages that ultimately produces mature sperm and oocytes [18]. The cartoon diagram in Fig 2B communicates the progression of N2 wild-type germline development based on the extent of germ cell proliferation, bend of the germline, and presence of sperm and oocytes. At specific time points (48,65,71, and 91 hours post hatching) we visualized and classified the germline development for the cisd-3.2(pnIs68) and N2 wild-type hermaphrodite (Fig 2C-2G). The cisd-3.2(pnIs68) animal exhibits a significant delay in germline development relative to the N2 wild-type hermaphrodite (Fig 2C). At 48 hours post hatch, the N2 wild-type germline displays a bend in the gonad, which is indicative of the L4 stage of development. Whereas, the germline of the cisd-3.2(pnIs68) animal has minor germ cell proliferation (Fig 2C and 2D). By 65 hours post hatch, the N2 wild-type animal contains mature oocytes and embryos within the uterus (Class V). Whereas, the germline of the cisd-3.2 (pnIs68) animal is at various stages and the germline contains some germ cell proliferation (Class III) and some of the animals display a small germline with a bend (Class IV) (Fig 2C  and 2E). At 71 and 92-hours post hatching, the germline within the cisd-3.2(pnIs68) animal continues to show a developmentally delayed, abnormal germline that visibly lacks oocytes (Fig 2C, 2F and 2G). These data suggest that complete disruption of cisd-3.2 and consequential decreased expression of cisd-3.1 and cisd-1 leads to severe germline defects.
To further evaluate the germline defect in the cisd-3.2(pnIs68) mutant we used Hoescht 33342 to examine the germline nuclei. This approach allows one to assess meiotic progression, in particular the stage of meiotic prophase, based on nuclei morphology within the hermaphrodite germline (Fig 3A) [19,20]. The nuclei located in the mitotic region can be distinguished from the crescent-shaped nuclei within the transition zone ( Fig 3A). Also, the length of the mitotic region can be determined by counting the number of nuclei diameters between the distal cell tip and the transition zone; the mitotic region length is approximately 20 nuclei diameters [20]. The mitotic region within the cisd-3.2(pnIs68) mutant appears to have less cells ( Fig  3B). We quantified the length of the mitotic region and determined that it is variable and significantly reduced in the cisd-3.2(pnIs68) germline by 27.7% in comparison to the N2 wildtype germline (Fig 3D). Furthermore, within the cisd-3.2(pnIs68) germline the number of   (B) Representative image of the mitotic region and transition zone, within the N2 wild-type or cisd-3.2(pnIs68) 1-day old adult hermaphrodite germline. The white arrowhead points to the distinct crescent shaped nuclei that notes the switch from mitosis to meiosis. Scale bar represents 10 μm. (C) Representative images of the pachytene, diplotene and diakinesis region, within the germline of N2 wild-type or cisd-3.2(pnIs68) 1-day old adult hermaphrodite animal. The number of oocytes within the diakinesis region are numbered; -1 indicates the primary oocyte. Scale bar represents 10 μm. (D) Quantification of the number of nuclei from the DTC region oocytes within the diakinesis region is significantly reduced by 35.0% relative to N2 wild-type animals (Fig 3C and 3E); note that a reduced number of oocytes within the diakinesis region was observed in the cisd-3.2(RNAi), cisd-3.1(RNAi) and cisd-1(tm4993) animal [16]. Further inspection of the oocyte found that a significant proportion of the cisd-3.2(pnIs68) oocytes contain a number of chromosome fragments or univalent chromosomes instead of the typical six bivalent chromosomes (Fig 3F and 3G). This differs from the six bivalent chromosomes observed within the oocytes of the N2 wild-type adult hermaphrodite (Fig 3F and 3G). Although the cisd-3.2(pnIs68) population can be maintained, there is a significant decrease (91.5%) in fecundity ( Fig 4A), a 44% increase in embryo lethality (Fig 4B), and an 8% increase in incidence of males (HIM phenotype) ( Fig 4C) relative to the N2 wild-type control. Combined, these results further support the idea that the CISD family is essential for germline function and in particular germ cell development.

The CISD-3 proteins are expressed in various tissues including the germline
To examine cisd-3.2 expression we used CRISPR to produce a cisd-3.2 promotor (cisd-3.2p) transcriptional reporter (cisd-3.2(pnIs64)) (S1 Fig). The cisd-3.2p::mKATE expression is present in the germline of the adult hermaphrodite and male ( Fig 5A). Within the hermaphrodite, cisd-3.2p::mKATE is expressed throughout the germline including the distal gonad region, oocytes, and spermatheca ( Fig 5B). Within the adult male cisd-3.2p::mKATE expression is observed in the distal gonad, spicule and vas deferens ( Fig 5B). Note that the mKATE reporter disrupts cisd-3.2 function (S1 Fig) and thus the male germline looks abnormal ( Fig 5B). The cisd-3.2p::mKATE expression occurs within other tissues including the pharynx and the anterior body wall muscle of the adult hermaphrodite ( Fig 5C). The cisd-3.2p::mKATE expression was also observed in the embryo blastomeres and within the developing germline and intestine of larvae ( Fig 5D). These expression profiles support the idea that cisd-3.2 has a functional role in both the male and hermaphrodite and within multiple tissues including the germline.
To assess the localization of CISD-3.2 and CISD-3.1, we used CRISPR to produce the cisd-3.2p::cisd-3.2::mKATE and cisd-3.1p::cisd-3.1::mYPET translational reporter strains (S2A and S2B Fig) [26]. The CISD-3.2::mKATE reporter is detected in many tissues of the adult hermaphrodite ( Fig 6A). Specifically, the CISD-3.2::mKATE expression is observed in the distal gonad, oocytes, spermatheca, and sperm within the hermaphrodite germline as well as throughout the embryo blastomeres ( Fig 6B). Additionally, CISD-3.2::mKATE expression is observed in the pharynx and body wall muscle within the head region of the adult hermaphrodite ( Fig 6C). Similar to the expression observed for the CISD-3.2::mKATE reporter, the CISD-3.1::mYPET reporter is detected in various tissues throughout the adult hermaphrodite ( Fig 6D). Specifically, the CISD-3.1::mYPET reporter is detected in the germline including the distal gonad region, spermatheca, sperm and oocytes ( Fig 6E). Furthermore, the CISD-3.1:: mYPET expression is observed throughout the embryo blastomeres and throughout the adult pharynx and body wall muscle in the head region ( Fig 6F).

The cisd-3.2(pnIs68) animal displays phenotypes associated with mitochondrial dysfunction
The phenotype and expression analyses shown thus far support the idea that the CISD-3 proteins functions within the mitochondria of many tissues including the germline. To assess if cisd-3.2(pnIs68) impacts mitochondrial structure within the germline, we stained the cisd-3.2 (pnIs68) mutant hermaphrodite with MitoTracker Red CMXRos, which is a membrane-potential-dependent mitochondrial dye. The mitochondria within the N2 wild-type animal, has a consistent pattern within the syncytium and a mesh-like structure within the oocytes (Fig 8A). The localization pattern is consistent with previous MitoTracker Red CMXRoS staining experiments conducted in C. elegans [21][22][23]. Within the germline of the cisd-3.2(pnIs68) hermaphrodite, the MitoTracker Red CMXRoS staining pattern is abnormal with an aggregated distribution (Fig 8A, white arrowhead). Furthermore, within the cisd-3.2(pnIs68) oocyte, the MitoTracker Red CMXRos staining was not uniform, for example, regions were void of staining ( Fig 8A, white arrow) and there is a significant decrease in fluorescent level within the oocytes by 51.6% (Fig 8B).

PLOS ONE
cisd-3.2 is essential for germline function unfolding response, to assess if the mitochondrial unfolding protein response (UPRmt) is induced in animals (L4 stage) with decreased cisd function, relative to control [24]. We opted to use RNAi so we could assess knockdown of cisd-3.1 and cisd-3.2. Relative to control, the expression of the hsp-6p::GFP reporter is increased in the cisd-3.2(RNAi) and cisd-3.1(RNAi) animal relative to control (Fig 8D and 8E). The combination of cisd-3.2 and cisd-3.1 knockdown did not further increase the expression of the hsp-6p::GFP reporter relative to knockdown of a single cisd-3 gene (Fig 8E). Additional phenotypes were observed in cisd-3.2(pnIs68) mutants including a reduced lifespan (S3 Fig). Furthermore, there were defects in pharyngeal pumping ( Fig 8F) and motility (Fig 8G) in the cisd-3.2(pnIs68) mutant. Taken together, these results indicate that CISD function is essential for mitochondrial function. Furthermore, the cisd-3.1(pnIs68) allele, which results in complete disruption of cisd-3.2 and partial decrease in cisd-3.1 and cisd-1 expression, impact functions associated with energy-requiring processes.

Discussion
The CISD/NEET proteins are highly conserved and present within the three domains of life [12,13]. Previous biochemical studies have shown that MiNT/CISD3 can transfer it's [2Fe-2S] clusters to the mitochondrial matrix proteins FDX1 and FDX2 [5]. Mammalian FDX1 is involved with adrenal steroidogenesis, bile acid, vitamin D synthesis and the reduction of mitochondrial CYP 450 proteins (CYP11A1, CYP11B) [25,26]. FDX2 can accept electrons from NADPH, contains a [2Fe-2S] cluster but does not reduce cytochrome P450 [25]. Similar to knockdown of CISD3/MiNT in cells, knockdown of FDX1 and FDX2 results in iron accumulation within the mitochondria [25,27]. Thus, it is likely that the disruption of CISD3 function in a whole organism can have detrimental impacts due to altered iron homeostasis and mitochondrial dysfunction.
Here we show that the C. elegans CISD-3 proteins (CISD-3.2 and CISD-3.1) are localized to the mitochondria in many tissues, throughout development, and in both males and hermaphrodites. Knockdown of cisd-3.2 and cisd-3.1 using RNAi results in an increase in germline apoptosis [16], and an increase in the hsp-6:GFP reporter for mitochondrial unfolded protein response (mtUPR) (Fig 8D and 8E). The cisd-3.2(pnIs68) mutant has a more severe phenotype relative to the cisd-3.2(RNAi) animal. This could be due to the complete reduction in cisd-3.2 transcript and an indirect impact on the expression of both cisd-3.1 and cisd-1 genes (Fig 1). Additional studies are needed to determine why the transcript levels of cisd-3.1 and cisd-1 are reduced in the cisd-3.2(pnIs68) animal. One intriguing idea to consider is that the absence of cisd-3.2 induces mitochondrial dysfunction (e.g. mtUPR) which in turn impacts mitochondria-to-nuclear communication. A decrease in cisd-3.1 or cisd-3.2 was not observed in the cisd-1 mutants (Fig 1) therefore, the cisd genes may not have overlapping function. It remains to be seen if reduction of cisd-3.1 impacts the expression of the other cisd genes. We hypothesize that disruption of all three cisd genes leads to severe mitochondrial defects (perhaps lethality). The cisd-3.2(pnIs68) mutant has a severely dysfunctional germline and a significantly reduced viability and lifespan. Furthermore, of the few cisd-3.2(pnIs68) animals that do survive and reach adulthood, the mitochondria within their germline are abnormally distributed (Fig 8). Within C. elegans, mitochondrial dysfunction will impact the electron transport chain, upregulate the mtUPR response, decrease oxygen consumption rate, and at times alter lifespan [28][29][30][31]. We hypothesize that cisd dysfunction disrupts mitochondria homeostasis which in turn impacts iron homeostasis and metabolic processes, thus, negatively impacting energy requiring processes including development and meiosis. It will be of interest to examine the specific proteins that the C. elegans CISD proteins interact with and if the function of these proteins is vital for normal germline function.
The C. elegans hermaphrodite germline rapidly produces a population of sperm, followed by oocytes over a period of several days [18,19,32]. The distal end of the adult germline

PLOS ONE
cisd-3.2 is essential for germline function provides a stem cell niche necessary for mitotic cell proliferation (Fig 3A). The germ cells located in the mitotic region replenish themselves and generate cells that leave the niche and enter into meiosis and gametogenesis; thus they fit the criteria of a stem cell [19]. The cisd-3.2 (pnIs68) mutant has a variable and significantly reduced number of nuclei from the DTC region to the transition zone (Fig 3B and 3D). This suggests that the cisd-3.2(pnIs68) mutant

PLOS ONE
cisd-3.2 is essential for germline function has abnormalities within the stem cell niche region of the gonad. It will be of interest to further examine the role of mitochondria homeostasis with the maintenance of the stem cell niche. The stages of meiotic prophase occur along the distal-proximal axis of the adult gonad ( Fig  3A) [33]. Within N2 wild-type hermaphrodites, during diakinesis, the six individual bivalents are visualized in oocytes as the nuclear volume increases and chromosome condensation proceeds. Unattached univalent chromosomes can be observed when meiotic dysfunction occurs such as abnormal chiasmata [34]. The pairing of homologous chromosomes and the formation of bivalents, are essential for accurate segregation of homologous chromosomes and requires homologous recombination and synapsis [35]. The cisd-3.2(pnIs68) mutant has a significant increase in the number of unpaired univalent chromosomes (Fig 3F and 3G). This suggests that disruption of mitochondria function can impact the proper formation of meiotic chromosomes.
Within the C. elegans germline a high level of mitochondrial DNA replication occurs indicating high number of mitochondria [36]. Mitochondrial morphology is shown to change throughout germ cell differentiation in the C. elegans germline [21]. Mutations in germlinespecific mitochondrial ATPase subunits impact germline development, mitochondrial biogenesis, and disrupts the balance between mitosis and oocyte differentiation resulting in reduced fecundity [37]. Furthermore, perturbed mitochondrial bioenergetics affect germ cell differentiation and globular (immature) mitochondria become elongated [21]. The regulation of the two core signaling pathways associated with oocyte maturation (MAPK/ERK and MSP) are associated with mitochondria bioenergetics [21]. This, combined with our work, show that mitochondrial homeostasis and function are essential for progression through meiosis. It will be of interest to determine if disruption of the CISD-3 proteins lead to iron imbalance which then impacts meiosis.
In mammals including humans, the mature oocyte contains many mitochondria (estimated to be in the millions) to support the growth and development of the early embryo. Critical factors, such as mitochondrial size, function, and number as well as ROS defense processes, are associated with successful oogenesis, fertilization, and embryo development [38]. Studies suggest oxidative stress can partially cause infertility and defective spermatogenesis [39][40][41][42]. Within C. elegans, cisd-1 dysfunction leads to hyperfused mitochondrial morphology, and higher levels of generated mitochondrial superoxide [17]. Biological cell culture studies with CISD3/MiNT knockdown revealed an increase in mitochondrial reactive species formation [5]. It will be of interest to examine the role ROS homeostasis has in C. elegans meiosis. Thus,

Fig 8. CISD dysfunction results in abnormal mitochondrial function. (A)
The mitochondria, detected using MitoTracker Red CMXRos, within the germline of the young gravid cisd-3.2(pnIs68) animal has abnormal structure and distribution in comparison to the developmentally matched N2 wild-type germline. Shown is the distal gonad, oocytes, and spermatheca (sp) region of the gonad; a single oocyte is indicated by dashed lined and enlarged in the right panel. The mitochondria are within the cytoplasm and surround the nuclei region (N). Arrowhead shows regions where "globular" structures accumulate. Arrow points to oocyte region void of MitoTracker Red CMXRos staining in the cisd-3.2(pnIs68) mutant germline. Scale bar indicates 50 μm (whole germline image) or 25 μm (oocyte image). Analysis of 15 animals from three independent experiments was conducted. (B) The cisd-3.2(pnIs68) animal (gravid adult) has a decrease in mean fluorescence intensity per oocyte relative N2 wild-type. The fluorescence, using MitoTracker Red CMXRos, was quantified in single oocytes of the cisd-3.2(pnIs68) mutant and the N2 wild-type animal; � indicates P value<0.0001, (unpaired t-test with Welch's correction). Oocytes from three independent experiments for a total of 15 oocytes per genotype were analyzed. The error bar represents standard deviation. (C) The cisd-3.2(pnIs68) mutant has a decrease in oxygen consumption rate relative to N2 wild-type animals; � indicates P<0.003, (unpaired t-test with Welch's correction). Animals assayed were non-gravid young adults. (D) Representative group images of the hsp-6p::gfp reporter strain fed specified RNAi food. HT115 is the control RNAi strain (empty vector). The images shown include fluorescent and fluorescent merged with DIC. Animals assayed were L4 larvae. Scale bar indicates 100 μm. (E) Quantification of hsp-6::GFP intensity in cisd-3.2(RNAi) and/or cisd-3.1(RNAi) relative to control animals fed HT115 control RNAi strain (n = 30). Bar indicates P<0.0001 (ordinary one-way ANOVA, Tukey's multiple comparisons test). Error bars represent standard deviation. (F) The pharyngeal pumping rate per minute is decreased in the cisd-3.2(pnIs68) animal relative to N2 wild-type (n = 10); � indicates P<0.0001, (two-tailed student t-test with Welch's Correction). Animals assayed were non-gravid young adults. (G) The rate of locomotion, determined by the number of body bends per minute, is reduced in the cisd-3.2(pnIs68) animal relative to N2 wild-type (n = 10); � indicates P = 0.0006, (two-tailed student t-test with Welch's Correction). Animals assayed were non-gravid young adults.
https://doi.org/10.1371/journal.pone.0245174.g008 the work conducted in C. elegans can translate to a greater understanding of the role mitochondrial homeostasis has in human fertility issues.

Germline development assay
The developmental progression of the germline was examined in the N2 wild-type and cisd-3.2 (pnIs68) hermaphrodite using DIC Normarski microscopy. Worms were placed on a 3% agarose pad with an anesthetizing solution of .1% tricane and .01% levinmosole. Animals were analyzed using the motorized Zeiss Axioscope mot plus 2 microscope. Images were captured using the Zeiss Axiocam camera and Axio Vision 4.7.1 software. Morphological features examined include extent of germ cell proliferation, bend within the germline, and the presence of sperm and oocytes. The stage of germline development was categorized at specific time points post egg hatching (48,65, 71 and 91 hours) (Refer to Fig 2B for categorization). Animals were categorized as Class II if the germline contained several proliferated cells beyond the Z2 and Z3 cells. Class III refers to a germline with a higher presence of proliferated cells and no bend in the germline; at this stage, the presence of vulva precursor cells was observed. Class IV refers to a germline in which the bend is observed but no oocytes are present. Class V refers to a germline in which visible oocytes are present in the diakinesis region. In three independent experiments, at least 10 animals for each genotype were examined at each time point.
Fecundity, embryo lethality, and incidence of male assays As previously described [16], to assess fecundity, synchronized N2 wild-type or cisd-3.2 (pnIs68) animals (n = 4) were collected at the L4-to-adult molt and placed individually onto a NGM plate and allowed to lay eggs. Every 24 hours the adults were transferred to a new NGM plate and the number of progeny produced for each animal was quantified after hatching. Animals were examined until no progeny were produced. To assess embryo lethality and the high incidence of males, synchronized N2 wild-type or cisd-3.2(pnIs68) mutant hermaphrodite (N = 9) were collected at the L4-to-adult molt and transferred daily to new NGM plates until reproduction ceased. Each day, the embryo lethality was determined by quantifying the number of eggs laid and the number of unhatched eggs 24-hours post egg lay. To assess the high incidence of male (HIM) phenotype the hatched offspring were allowed to develop and the number of hermaphrodite and male progeny was quantified 48 hours after hatching.

Lifespan measurement
Synchronized N2 wild-type or cisd-3.2(pnIs68) mutant, at the L4-to-adult hermaphrodite molt stage, were transferred to fresh NGM plates (time = 0). The N2 wild-type and cisd-3.2(pnIs68) mutant was scored daily for survivorship. The gravid hermaphrodites were transferred to a new plate every day. Non-gravid animals as they aged were transferred less frequently to minimize inducing injury. Animals were scored as dead if the worm failed to move after touching and was subsequently removed from the plate. The animals that crawled off the plate, bagged out, or showed uterine rupture were nulled. Three biological replicates were completed with a total of N = 50 animals for each experiment.

Pharyngeal pumping and locomotion analysis
Pharyngeal and locomotion were assessed in synchronized 1-day old adult N2 wild-type or cisd-3.2(pnIs68) mutant animals (n = 10). The number of pharyngeal pumps was quantified over a 30-second period to assess pumping rate per minute. To assess locomotion, the number of body bends within one-minute intervals was quantified. A body bend is defined as a change in the direction of propagation, or the completion of one sinusoidal movement of the animal's tail.

Live animal microscopy analysis
To visualize worms of the specified genotypes, animals were placed in an anesthetizing solution of 0.1% tricane and 0.01% levinmosole and mounted on 3% agarose slides. Synchronized animals, at the stage indicated, were analyzed using a motorized Zeiss Axioscope mot 2 plus microscope or a Zeiss LSM 710 Confocal Scanning Microscope. To assess the impact cisd-3.1 (RNAi) and/or cisd-3.2(RNAi) has on the hsp-6::GFP fluorescent reporter, synchronized animals of the F1 generation were grown to L4 larval stage on the specified RNAi food and analyzed by epifluorescence microscopy. The integrated density, area, and mean background fluorescence was computed using ImageJ. As described by others [48], the Corrected Total Worm Fluorescence (CTWF) was calculated using the following formula for Corrected Total Cell Fluorescence as described, CTWF = Integrated Density-(area of selected worm � Mean background fluorescence). Three independent experiments were conducted analyzing a total of 30 animals per RNAi treatment. Images were processed using ImageJ and Adobe Photoshop.

Hoescht 33342 germline staining
The germline nuclei in 1-day old adult hermaphrodites were analyzed using Hoescht 33342 staining; nuclei analyzed includes those within the mitotic region, meiotic stage, oocyte region, and sperm containing spermatheca. Briefly, a synchronized population of adult hermaphrodites were fixed in ethanol for 4 minutes, rinsed with M9, and stained with Hoescht 33342 (10 μg/ml) mixed within anti-fade reagent. Animals were placed on a slide and visualized using a motorized Zeiss Axioscope microscope. Images were captured using the Zeiss Axiocam camera and Axio Vision 4.7.1 software. The mitotic region/transition zone (MT/TZ) boundary was determined by counting the number of mitotic nuclei from the distal tip to the region that contains distinct crescent-shaped nuclei morphology indicative of the leptonene/zygotene prophase I nuclei (Fig 2A) [19,49]. Hoescht 33342 staining allows detection of nuclei within the transition zone and other meiotic stages (pachytene, diplotene, and diakinesis and the mature gametes) (Fig 2A). The oocyte nuclei, within the diakinesis region, were analyzed to quantify the number of bivalent chromosomes [50].

Mitochondria analysis using MitoTracker
Young gravid adult animals, of the specified genotype, were incubated overnight on OP50 seeded plates with stock solution of MitoTracker freshly diluted in M9 containing 10 μM Mito-Tracker Red CMXRos (Invitrogen Cat #M7512) or swam for 20 minutes in 10 μM Mito-Tracker Green FM (Invitrogen Cat# M7514) respectively [51,52]. After incubation with the MitoTracker molecule, animals were transferred to OP50 seeded plates to allow a 1-hour clearance of the dye from their guts. For colocalization studies, at least twenty cisd-3.2(pnIs25):: mKATE or cisd-3.1(pnIs39)::mYPET animals, were stained with the specified MitoTracker, and analyzed using Zeiss Azioscope. Images were captured using the Zeiss Axiocam camera and Axio Vision 4.7.1 software.
To analyze the mitochondria membrane potential (ΔCM) within the oocytes of N2 wildtype and cisd-3.2(pnIs68) hermaphrodite, whole animals were stained with Mitotracker Red CMXRos as described above, and were imaged with a Zeiss LSM 710 Confocal Scanning Microscope and captured with a Zeiss LSM ZEN camera. The z-stack images (2 μm slices) of the germline were collected and combined into a single image; at least 15 animals per genotype were analyzed. Images were processed using ImageJ and Adobe Photoshop. The mean fluorescence intensity was measured for the individual secondary oocyte for the indicated genotypes [21].

Basal oxygen consumption rate measurements
The oxygen consumption rate (OCR) was analyzed in young non-gravid hermaphrodites using the Seahorse-X24 instrument [53]. Approximately 100 N2 wild-type or cisd-3.2(pnIs68) animals were collected for each well and washed three times with M9 and allowed to settle by gravity. In the Seahorse 24-well assay plate, 500 μl/well of M9 was aliquoted; 4 blank wells and 3 wells per genotype. Experiments were repeated 10 times and oxygen consumption rates were performed using the following program: 10 cycles of (2 min mix, 2 min wait, 2 min measure). After the Seahorse program, samples were collected and the total number of animals for each well was counted to determine the OCR rate per animal.

Statistical analysis
Animal experimental values were compared to control animals first by analyzing for Gaussian distribution using D' Agostino & Pearson or Shapiro-Wilk normality test (alpha = 0.05, P<0.05). If the normality test was passed, a parametric statistical test was performed: unpaired t-test with Welch's correction or ordinary one-way ANOVA followed by a Dunnet's multiple comparison test. If the normality test was not passed, data was analyzed using nonparametric statistical tests; unpaired t-test or Kruskal-Wallis test followed by a Dunn's multiple comparisons test. In the instance of the total % population, animal experimental values were compared to control animals by conducting a Fisher's Exact Two-tailed test. For the lifespan experiment we conducted a comparison of survival curves (P<0.0001, Log-rank (Mantel-Cox) test and Gehan-Breslow-Wilcoxon test. The statistical test for each experiment is noted and the P-values are reported in the figure legends. All data sets are expressed as mean ± standard deviation (SD), or as percent of animals with the presence or absence of the noted phenotype of interest. Statistical analysis was conducted using Prism 8 software.

S1 Fig. Use of CRISPR to generate the cisd-3.2(pnIs68) mutant (strain PM91) and cisd-2 (pnIs64) transcriptional reporter (strain PM58).
Illustration of the cisd-3.2 locus predicted to produce the cisd-3.2 transcriptional reporter and mutant. The mKATE^SEC^3XMyc sequence was inserted upstream from the first exon disrupting cisd-3.2 function. The reporter strain was exposed to heat shock to remove the SEC and gene resulting in the non-roller strain.