Loss of mitochondrial proline catabolism depletes FAD, impairing sperm function, and male reproductive advantage

Exposure to environmental stress has a clinically established influence on male reproductive health, but the impact of normal cellular metabolism on sperm quality and function is less well-defined. Here we show that homeostatic changes in mitochondrial dynamics driven by defective mitochondrial proline catabolism result in pleiotropic consequences on sperm quality and competitive fitness. Disruption of alh-6, which converts 1-pyrroline-5-carboxylate (P5C) to glutamate, results in P5C accumulation that drives oxidative stress, activation of SKN-1, and a reduction of energy-storing flavin adenine dinucleotide (FAD) levels. These molecular changes lead to premature male reproductive senescence by reducing sperm quality. These sperm-specific defects are suppressed by abating P5C metabolism, by treatment with antioxidants to combat reactive oxygen species (ROS), or by feeding diets that restore FAD levels. Our results define a role for mitochondrial proline catabolism and FAD homeostasis on sperm function and specify strategies to pharmacologically reverse unintended outcomes from SKN-1/Nrf transcriptional activation.


128
Mutation in mitochondrial alh-6 results in diet-independent reduction in fertility 129 Altered mitochondrial structure and function have been correlated to loss of proper sperm function 130 in different species (16,(41)(42)(43). In addition, proper sperm function requires a low level of ROS (25-27), 131 although a specific role for endogenous mitochondrial derived ROS is undefined. ALH-6/ALDH4A1, is a 132 nuclear-encoded mitochondrial enzyme that functions in the second step of proline catabolism, 133 converting 1-pyrroline-5-carboxylate (P5C) to glutamate ( Figure S1). We anticipated that mutation of alh-134 6 may affect the germline, based on our previous assessment of the premature aging phenotypes in 135 somatic cells in alh-6 mutants (35). Using an UV-integrated alh-6::gfp strain under its endogenous 136 promoter, we saw that alh-6 is localized to the mitochondria in the germline of hermaphrodites ( Figure  137 S2). We then assessed progeny output of alh-6(lax105) hermaphrodites until egg laying ceased and 138 found a reduction in self-fertility brood size (-12.9%) ( Figure 1A). Since the somatic phenotypes of alh-139 6(lax105) mutants are known to be diet-dependent (35, 36), we examined self-fertility of animals fed the 140 E. coli K-12 bacteria HT115, to determine if the reduced reproductive output is also dependent on the 141 type of bacterial diet ingested. Surprisingly, we found that the self-fertility of alh-6 animals was markedly 142 reduced (-20.7%), when animals were fed the K-12 diet ( Figure 1B). alh-6 mutants have similar timing 143 in their progeny output as compared to wild type animals on both diets ( Figures S3A-B). Since alh-6 144 mutants display normal development and reproductive timing, the progeny deficit is not a result of an 145 attenuated reproductive span which reveals the differential impact of alh-6 loss in the soma (diet-146 dependent) (35) and the germline (diet-independent). 147 148 alh-6 fertility defects are sperm-specific 149 We noted that alh-6 mutant hermaphrodite animals laid twice as many unfertilized oocytes as wild 150 type animals over their reproductive-span (Figure 1C), suggesting a loss of sperm function (44)(45)(46). It is 151 notable that alh-6 mutant hermaphrodites lay very few, if any, dead eggs ( Figure 1C), suggesting that 152 the loss of ALH-6 activity is not lethal. To determine whether the reduced brood size of alh-6 mutants are 153 due to a general loss of germ cells or a specific defect in oocytes or sperm, we examined the mated-154 fertility of these animals by mating wild type young adult (day 0-1) males to either wildtype or alh-6 mutant 155 virgin hermaphrodites (in wild type C. elegans, male sperm outcompetes hermaphrodite sperm >99% of 156 the time (47,48)). We found that the reduced fertility in alh-6 mutant hermaphrodites is fully rescued by 157 wild type sperm, which confirmed that oocyte quality is not impaired but rather, alh-6 hermaphrodite 158 sperm is dysfunctional ( Figures. 1D-E). 159 To better assess the quality of alh-6 mutant sperm, we measured the ability of alh-6 mutant sperm 160 to compete with wild type sperm (49). To differentiate between progeny resulting from mating and 161 progeny that arise from hermaphrodite self-fertilization, we made use of male animals harboring a GFP 162 transgene such that any cross-progeny will express GFP while progeny that arise from hermaphrodite 163 self-sperm will not ( Figure 1D). We found that wild type hermaphrodites when mated to alh-6 mutant 164 males have significantly more self-progeny as compared to wild type hermaphrodites mated to wild type 165 males ( Figure 1F). This finding indicates a sperm competition deficit of alh-6 males resulting in a brood 166 derived from self-fertilization, which is uncommon after mating has occurred (47). As hermaphrodite C. 167 elegans produce a set amount of sperm during L4 stage before switching exclusively to oogenesis, 168 eventually depleting its reservoir of sperm (47,50). To assess whether alh-6 mutant sperm are generally-169 dysfunctional, we mated older hermaphrodites that had depleted their complement of self-sperm and 170 found that alh-6 mutant males are able to produce equal numbers of progeny as wild type males when 171 the need for competition with hermaphrodite sperm is abated ( Figure S4A); thus, although alh-6 mutant 172 sperm are impaired for competition, they remain competent for reproduction. This is similar to recent 173 study on comp-1, a mutation which results in context-dependent competition deficit in C. elegans sperm 174 (51). Similarly, older alh-6 mutant hermaphrodites mated to young wild type males yield similar level of 175 progeny as age-matched WT hermaphrodites, which further supports a model where sperm, but not 176 oocytes, are defective in alh-6 mutants ( Figure S4B). 177 Similar to mammals, the contribution of sperm to fertility in C. elegans is dictated by distinct 178 functional qualities, which include: sperm number, size, and motility (49,52,53). In C. elegans, male 179 sperm are larger and faster than hermaphrodite sperm, which affords a competitive advantage (53). We 180 next sought to define the nature of the sperm competition defect in alh-6 mutants by measuring sperm 181 size, motility, and number in alh-6 mutants compared to wild type animals. One day after spermatogenesis 182 initiation (at the L4 larval stage of development), alh-6 adult hermaphrodites have a reduced number of sperm 183 in the spermatheca as compared to wild type ( Figure S5A), which is correlated with the reduced self-fertility 184 observed (Figures 1A-B). In contrast, age-matched alh-6 mutant males have similar numbers of sperm as 185 WT males, suggesting that they have a similar rate of production (Figure 2A). We next examined sperm size 186 in day 1 adult males and discovered that alh-6 mutant spermatids are significantly smaller as compared to 187 wild type ( Figure 2B). To achieve motility, C. elegans sperm must be activated to allow pseudopod 188 development, and this development requires protease activation (54) ( Figure 5B). In vitro, sperm activation 189 can be recapitulated by treatment of isolated spermatids with Pronase (55). In addition to reduced size, the 190 percentage of activated spermatozoa was significantly reduced in alh-6 mutants as compared to wild type 191 ( Figure 2C). Taken together, the reduction of sperm quantity and quality (size and activation) are contributors 192 to the reduced fertility in alh-6 mutants.

194
Transcriptional signatures define temporal phenotypes of alh-6 activity 195 We first identified alh-6 mutant in a screen for activators of the cytoprotective transcription factor 196 SKN-1/NRF2 using gst-4p::gfp as a reporter (35,36). When activated, SKN-1 transcribes a variety of 197 gene targets that collectively act to restore cellular homeostasis. However, this can come with an 198 energetic cost with pleiotropic consequences (35,36,(56)(57)(58)(59)(60)(61). alh-6 mutants have normal development, 199 but display progeric phenotypes towards the end of the normal reproductive span (35) indicating a 200 temporal switch in phenotypic outcomes. We reasoned that the temporally controlled phenotypes in the 201 alh-6 mutants could be leveraged to identify potential mechanisms by which alh-6 loss drives cellular 202 dysfunction. As SKN-1 is activated in alh-6 mutants after day 2 of adulthood (35), we defined genes that 203 display differentially altered expression in the L4 developmental stage, when spermatogenesis occurs, 204 as compared to Day 3 adults (post SKN-1 activation). We performed RNA-Seq analyses of worms with 205 loss of alh-6 and identified 1935 genes in L4 stage animals and 456 genes in Day 3 adult animals that 206 are differentially expressed (+/-Log2 (fold change), 0.05 FDR). Intriguingly, the gene expression changes 207 at these two life periods had distinct transcriptional signatures (Figures 3A-B; Figure S6). Because the 208 loss of alh-6 drives compensatory changes in normal cellular metabolism, which later in life results in the 209 activation of SKN-1, we expected to identify significant changes in both metabolic genes and SKN-1 210 target genes. Supporting this hypothesis, the gene ontology (GO) terms most enriched include 211 oxidoreductases and metabolic enzymes in L4 stage animals ( Figure 3A) and SKN-1-dependent targets 212 such as glutathione metabolism pathway genes in Day 3 adults ( Figure 3B). Importantly, our 213 transcriptomic analysis recapitulated the temporally-dependent phenotypic outcomes resulting from alh-214 6 loss; genes in the pseudopodium and germ plasm GO terms class displayed reduced expression in L4 215 ( Figure 3A), which could impact C. elegans spermatogenesis. In contrast, genes in the muscle-specific 216 GO term class displayed increased expression in day 3 adults (Figure 3B), which is when SKN-1 activity 217 is enhanced in the muscle of alh-6 mutants (36). Taken together, the transcriptome analysis of alh-6 218 mutants is diagnostically relevant and informative for defining drivers of organism-level phenotypic 219 changes in animals with altered proline catabolism. 220 221 FAD mediates sperm functionality and competitive fitness 222 The strong enrichment of genes whose protein products utilize and/or bind cofactors or co-223 enzymes was intriguing as the maintenance of metabolic homeostasis and the redox state of the cell 224 requires a sophisticated balance of multiple cofactors ( Figure 4A). In fact, the proline catabolism 225 pathway utilizes multiple cofactors to generate glutamate from proline; PRDH-1 uses FAD as a co-factor 226 while ALH-6 utilizes the reduction of NAD+. In the absence of ALH-6, PRDH-1 would continue to deplete 227 FAD, which would activate compensatory pathways to maintain metabolic homeostasis in addition to 228 activating pathways to detoxify P5C (oxidoreductases, P5C reductase, etc.). In light of this hypothesis 229 we measured FAD and found a significant reduction in alh-6 mutant animals ( Figure 4B). As such, we 230 predicted that restoration of FAD levels might alleviate the sperm-specific phenotypes of alh-6 mutants. 231 Dietary supplementation of riboflavin has been shown to increase cellular FAD levels (62, 63), and when 232 fed to alh-6 mutants, it restored sperm function. We found that wild type hermaphrodites mated to alh-6 233 mutant males that were fed a diet supplemented with 2.5mM riboflavin produced significantly more total 234 progeny than alh-6 males fed the standard OP50 diet ( Figure S7). Moreover, riboflavin supplementation 235 was sufficient to restore male sperm size ( Figure 4C) and also rescued the impaired activation ( Figure  236 4D) of male sperm in alh-6 mutants. Taken together, these data suggest that loss of alh-6 leads to a 237 decrease in cellular FAD levels that drives sperm dysfunction. 238 239 Loss of cellular proline catabolism is not causal for sperm defects in alh-6 mutants 240 We were curious to uncover additional molecular mechanisms that underlie the loss of sperm 241 function in alh-6 mutants. To do this, we performed an EMS mutagenesis screen to identify suppressors 242 of the induced gst-4p::gfp expression phenotype in alh-6 mutants ( Figure 5A) (35). We identified one 243 suppressor allele, lax228, which we mapped to right arm of chromosome IV between F49E11 and 244 Y57G11B SNPs. We then generated a list of candidate genes in this region with non-synonymous 245 mutations in the exons of protein coding genes using whole genome sequencing data of the alh-6 mutant 246 compared to the suppressor mutant alh-6(lax105); lax228 (64). We tested each of these genes by RNA 247 interference (RNAi) in the alh-6;gst-4p::gfp strain to phenocopy the suppressor. RNAi of B0513.5, 248 hereafter referred to as prdh-1 as it encodes for proline dehydrogenase, was the only RNAi target that 249 phenocopied the lax228 mutant ( Figures S8A-B). PRDH-1 catalyzes the first enzymatic step of proline 250 catabolism ( Figure 5B), converting proline to P5C. Importantly, this enzyme is linked to several of the 251 phenotypes of alh-6 mutants including the generation of P5C (9) and the continued reduction of FAD, 252 documented above ( Figure 4B). We also examined the expression of the proline catabolism pathway 253 genes from our RNA-Seq analysis and discovered a significant increase in the expression of enzymes 254 that would prevent the accumulation of P5C in alh-6 mutant L4 animals, before irreparable damage 255 occurs ( Figure 5C). Specifically, there was an increase in expression of pyrroline-5-carboxylate 256 reductase (M153.1/PYCR) that converts P5C back to proline and ornithine transaminase(oatr-1/OAT) 257 that converts P5C to ornithine. Surprisingly, the expression of pyrroline-5-carboxylate synthase (alh-258 13/P5CS) was also increased, however P5CS has two enzymatic functions: glutamate kinase (GK) and 259 γ-glutamyl phosphate reductase (GPR) activities that impact additional nodes of cellular metabolism. 260 Moreover, since proline itself has important roles in cellular protection, the increased expression of P5CS 261 might be an important stress response, but with pleiotropic consequences as it would deplete glutamate 262 and increase an already accumulating pool of P5C. 263 To determine how the total loss of proline catabolism would affect C. elegans reproduction, we 264 examined the alh-6; prdh-1 double mutant in our panel of reproduction and sperm quality assays. The 265 reduction in spermatid size ( Figure 5D) and impairment of spermatid activation ( Figure 5E) in alh-6 266 mutants are both suppressed by loss of prdh-1. In addition, the prdh-1 mutation restored the reduced 267 self-fertility ( Figure S9A), lower hermaphrodite sperm count ( Figure S9B), and suppressed the increased 268 laying of unfertilized oocytes of the alh-6 single mutant ( Figure S9A). Finally, the reduced ability of alh-269 6 male sperm to compete against wild type hermaphrodite sperm was abrogated in the alh-6;prdh-1 270 double mutant ( Figure S9C). These results were surprising as they reveal that loss of flux through the 271 mitochondrial proline catabolism pathway is benign for animal reproductive fitness, but suggests instead 272 that P5C accumulation is instrumental in driving sperm dysfunction in alh-6 animals.

274
Endogenous ROS drives alh-6 sperm defects 275 Several studies have examined the impact of exogenous ROS-inducing electrophiles on sperm 276 function, but the impact of endogenously produced ROS on sperm function remains poorly defined. The 277 continuous generation of P5C by PRDH-1 leads to the accumulation of this highly toxic and unstable 278 biomolecule, which can lead to redox imbalance and impair the normal function of germ cells as it does 279 for somatic tissues (35,(37)(38)(39)65). If the sperm defects in the alh-6 mutants are a result of a loss of redox 280 and/or ROS homeostasis, then we anticipated that antioxidants could alleviate these phenotypes. We 281 supplemented the diet of alh-6 mutant males with the antioxidant N-acetylcysteine (NAC), from birth 282 through reproductive maturity, and re-measured the reproductive parameters of these animals. NAC 283 supplementation restored spermatid size and activation of alh-6 animals to WT levels ( Figures 5D-E). 284 Antioxidant supplementation in wild type (Figures S10A-B) or alh-6; prdh-1 double mutants had no effect 285 (Figures S10C-D). Collectively, these data suggest that endogenous production of ROS is causative for 286 the sperm dysfunction seen in alh-6 animals. In addition, this study reveals that antioxidant 287 supplementation can act as a treatment to overcome reproductive deficiencies stemming from defects in 288 specific cellular metabolic pathways. 289 290 Mitochondrial dynamics regulate spermatid function 291 Although there is a clear and documented role for mitophagy in the clearance of paternal 292 mitochondria post-fertilization in C. elegans, the role(s) for mitochondrial dynamics and turnover in sperm 293 function prior to zygote formation are unclear. We first examined mitochondrial dynamics in wild type 294 sperm by staining with the fluorescent, mitochondrial-specific dye JC-1 (ThermoFisher), and noted that 295 each spermatid on average contained multiple discernable spherical mitochondria that are mostly not 296 fused ( Figures 6A, 6B, 6G). Previous studies in yeast and cultured mammalian cells have shown that 297 when cells are exposed to mild stress, the initial response of mitochondria is to fuse in an attempt to 298 dilute damage (66-68). Indeed alh-6 mutant spermatids had mitochondria that were more interconnected 299 ( Figures 6C, 6D, 6G) as compared to wild type spermatids, which supports our finding that these sperm 300 are under oxidative stress ( Figure 5). Loss of prdh-1, which restores sperm function (Figure 5), returned 301 spermatid mitochondria to more punctate structures (Figures 6E, 6F, 6G). Similarly, treatment with the 302 antioxidant NAC returned alh-6 mutant mitochondria in spermatids to wild type levels of fusion ( Figure  303 6H). The JC-1 dye accumulates in mitochondria in a membrane potential (DY)-dependent manner, and 304 as concentration exceeds threshold, its fluorescence switches from green to red emission; thus, a higher 305 red-to-green fluorescence ratio is indicative of healthier mitochondria, with higher DY. alh-6 mutant 306 spermatids have reduced red:green JC-1 fluorescence that indicates a lower mitochondrial DY, and an 307 accumulation of unhealthy mitochondria ( Figure 6I). 308 The role of mitochondrial dynamics (fusion and fission) in the maturation of sperm has not been 309 studied; however recent work has revealed that the mitochondrial fusion and fission machinery are 310 important for the elimination of paternal mitochondria post-fertilization (69). FZO-1 is required for proper 311 fusion of the mitochondrial membranes and DRP-1 is required for mitochondrial fission (70, 71). The 312 balance of this fusion and fission machinery in the upkeep of mitochondrial homeostasis allows cells to 313 respond to changes in metabolic needs and external stress (72, 73). RNAi of fzo-1 suppressed the 314 enhanced fusion observed in alh-6 mutant spermatid mitochondria indicating mitochondrial fusion is 315 active in spermatids with impaired proline catabolism ( Figure 6J). We next examined spermatids from 316 drp-1 mutant animals and observed a greater level of mitochondrial fusion as compared to wild type and 317 alh-6 mutant spermatids ( Figure 6K). We observed a synergistic level of mitochondrial fusion in 318 spermatids derived from alh-6; drp-1 double mutants. This finding is consistent with previous studies in 319 yeast which reveal that defects in fusion can be compensated for by changes in the rates of fission and 320 vice versa (72,73). In support of our model where mitochondrial dynamics act as a major driver of the 321 sperm-specific defects in alh-6 mutants, we discovered that loss of drp-1, which results in increased 322 mitochondrial fusion (like that observed in alh-6 mutants), also reduces sperm activation ( Figure 6L). 323 Moreover, reduced fzo-1 does not alter activation in wild type sperm, but restores activation in alh-6 324 sperm ( Figure 6M); suggesting increased fusion in alh-6 sperm mitochondria is impairing proper function. 325 Taken together, these data support a model where loss of mitochondrial proline catabolism induces 326 mitochondrial stress, activating mitochondrial fusion, which can subsequently eliminate damage in order 327 to preserve functional mitochondria ( Figure 6N). These data also reveal a functional role for 328 mitochondrial fusion and fission in maintaining proper sperm function. In conclusion, our studies define 329 mitochondrial proline catabolism as a critical metabolic pathway for male reproductive health. 330 331

332
Here we investigate the effects of mutation in the mitochondrial enzyme gene alh-6, and the 333 associated increased ROS levels on male fertility stemming from defective mitochondrial proline 334 catabolism. We found that alh-6 mutants show a reduction in brood size that is sexually dimorphic; 335 defects in sperm function but not oocytes contribute to reduced hermaphrodite fertility. As societal factors 336 continue to push individuals to wait longer to have children, our studies are of critical importance to 337 elucidate how restoring and maintaining functional amino acid catabolism during aging will promote 338 reproductive success. 339 Although C. elegans is a well-established organism for studying aging and reproduction, with 340 several studies describing hermaphrodite reproductive senescence, many questions regarding the basis 341 of male reproductive decline remain unanswered. Decades of work have shown that exposure to 342 pollution, toxins, xenobiotics, and other ROS-inducing compounds can prematurely drive the loss of 343 sperm function (29, 74, 75), but the impact that normal cellular metabolism plays on sperm function and 344 the identification of specific molecules that can mediate sperm quality are not well-defined. In this study 345 we characterized a new role for mitochondrial proline catabolism and FAD homeostasis in the 346 maintenance of proper sperm function. Perturbation of this pathway, through mutation of alh-6/ALDH4A1, 347 increases ROS, causing metabolic stress and increased mitochondrial fusion in spermatids, which results 348 in impaired sperm function and premature reproductive senescence 349 Mutation in proline dehydrogenase (PRODH) in humans results in hyperprolinemia type I (HPI), 350 while mutation in delta-1-pyrroline-5-carboxylate dehydrogenase (ALDH4A1/P5CDH) results in 351 hyperprolinemia type II (HPII). Surprisingly, in C. elegans, mutation in upstream proline dehydrogenase 352 gene, prdh-1, is able to suppress all the reproductive defects in alh-6 mutants suggesting that overall 353 reduction in proline catabolism is not causal for the observed reproduction phenotypes. In humans, the 354 diagnosis of HPI and HPII are both through by elevated level of proline in plasma, with addition of high 355 level of P5C in HPII patients. The symptoms of HPI varies on severity depending on the reduction of 356 PRODH activity (type of mutation) and are characterized by neurological, auditory, and renal defects (53). 357 Although proline catabolism has not previously been shown to have a direct role in fertility, human fertility 358 studies have shown that the addition of proline in cryopreservation medium improves sperm mobility and 359 preservation of membrane integrity upon thawing (60, 61). This study reveals that in C. elegans, proline 360 catabolism impacts several functional qualities of sperm. Loss of proline catabolism results in smaller 361 sperm with impaired activation, two qualities that directly impact competitive advantage. As such, proline 362 biosynthesis, catabolism, and steady state concentrations must be tightly regulated, and the importance 363 of proline in cellular homeostasis may help explain the transcriptional responses measured in animals 364 with dysfunctional alh-6 (Figures 3 and 4).

365
Our previous work defined the age-dependent decline in function of somatic tissues, particularly 366 muscle in animals lacking functional ALH-6 (35, 36), which is not manifested until Day 3 of adulthood. 367 This study reveals that although somatic phenotypes are observed post-developmentally, the germline 368 and specifically sperm are sensitive to loss of alh-6 much earlier in development (phenotypes assayed 369 at L4 or Day 1 of adulthood), with many physiological consequences from dysregulation of metabolism. 370 Reproductive senescence is a field of growing significance as the number of couples that choose to delay 371 having children increases. About 30-40% of all male infertility cases are associated with increased levels 372 of ROS, yet we don't understand the underlying mechanism (76). Additionally, sperm quality has been 373 shown to decline with age, as ROS content increases with age (7,8,77,78), demonstrating the link 374 between ROS and male reproductive senescence. Our study demonstrates that perturbation of 375 mitochondrial proline catabolism, particularly mutation in alh-6/ALDH4A1, leads to redox imbalance and 376 impaired sperm function. Importantly, addition of antioxidants to diet can abrogate this sperm dysfunction 377 (Figures 5D-E), implicating the potential therapeutic effects of antioxidant supplement in male infertility 378 arising from redox imbalance. 379 Recent studies have focused on the role of NAD+ metabolism in cellular health, while the impact 380 of FAD has received less attention. FAD levels are diminished in alh-6 animals specifically at the L4 stage 381 when spermatogenesis is occurring ( Figure 4B). Riboflavin (Vitamin B2) is a precursor to FAD and flavin 382 mononucleotide (FMN) cofactors that are needed for metabolic reactions (like proline catabolism and 383 mitochondrial oxidative phosphorylation) to maintain proper cellular function. Despite its importance, 384 humans lack a riboflavin biosynthetic pathway and therefore require riboflavin from exogenous sources 385 (79). Insufficient intake can lead to impairment of flavin homeostasis, which is associated with cancer, 386 cardiovascular diseases, anemia, neurological disorders, fetal development, etc. (79). Our study 387 suggests that riboflavin and FAD play critical roles in reproduction as alh-6 mutants suffer from sperm 388 dysfunction driven by a reduction in FAD levels (Figure 4). Importantly, these sperm specific defects can 389 be corrected by dietary supplementation of vitamin B2, which in light of the exceptional conservation of 390 mitochondrial homeostatic pathways, suggest the nutraceutical role vitamin B2 could play in sperm health 391 across species. 392 Our study also demonstrates that spermatids lacking alh-6 have increased mitochondrial fusion; 393 a perturbation at the mitochondrial organelle structure-level that contributes to the sperm-specific 394 phenotypes observed. In addition to prior work showing fzo-1/MFN1/MFN2 and drp-1/DRP-1 to be 395 important for mitochondrial elimination post-fertilization (69), our work reveals that mitochondrial fission 396 and fusion machinery are present and active in spermatids and that perturbation of these dynamics can 397 affect sperm maturation and competitive fitness (Figure 6). Future work to define how alh-6 spermatids 398 use mitophagy, which can clear damaged mitochondria, will be of interest. In conclusion, our work 399 identifies proline metabolism as a major metabolic pathway that can impact sperm maturation and male 400 reproductive success. Moreover, these studies identify specific interventions to reverse the redox 401 imbalance, cofactor depletion, and altered mitochondria dynamics, all of which play a part in sperm 402 dysfunction resulting from proline metabolism defects. 403 404

ACKNOWLEDGEMENTS 405
We thank K. Han and L. Thomas for technical assistance, and H. Dalton      sperm activation by Fisher's exact test with p-value cut-off adjusted by number of comparisons. *, p<0.05; 742 **, p<0.01; ***, p<0.001; ****, p<0.0001. All studies performed in biological triplicate; refer to Table S1 for  743 n for each comparison.  alh-6 mutation results in increased fusion in sperm mitochondria that is mediated by fzo-1, which results 757 in impaired sperm activation. Statistical comparisons of JC-1 Red/Green FL ratio by unpaired t-test. 758 Statistical comparisons of mitochondria fusion by ANOVA. Statistical comparisons of sperm activation by 759 Fisher's exact test with p-value cut-off adjusted by number of comparisons. *, p<0.05; **, p<0.01; ***, 760 p<0.001; ****, p<0.0001. All studies performed in biological triplicate; refer to Table S1 for n for each  761 comparison. 762 SUPPLEMENTAL FIGURE LEGENDS 763 764 Figure S1. Cartoon depiction of proline catabolism pathway in C. elegans. 765 766 Figure S2. ALH-6 expression in the germline. UV integrated alh-6::gfp strain under its endogenous 767 promoter reveal expression of ALH-6 in hermaphrodite (a-b) and male (c-d) germline. a and c are DIC 768 images while b and b are GFP images. 769 770 Figure S3. alh-6 hermaphrodite reproductive span is similar to wild type (WT) on different diets.