ABSTRACT
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.
INTRODUCTION
As individuals wait longer to have families, reproductive senescence has become an increasingly prudent topic (1, 2). Decline in oocyte quality is well-documented with age and can result in fertility issues when older couples try to conceive (3). Furthermore, pregnancies at an older age pose risks for higher incidences of birth defects and miscarriages. In humans, female reproduction ceases at menopause at an average age of 41-60, with the onset of menopause (4). The C. elegans ‘wild type’ is hermaphroditic and self-fertilizing; however, they are capable of making and maintaining Mendelian ratios of male (sperm-only) animals in their populations. Like humans, C. elegans experience a decline in fecundity with age by halting oocyte production at roughly one-third of their lifespan (5). In addition, regulators of reproductive aging insulin/IGF-1 and sma-2/TGF-β signaling are conserved regulators of reproductive aging from worms to human (6). While the majority of studies in reproductive senescence have focused on maternal effects, male factors contribute to a large portion of fertility complications with increasing evidence of an inverse relationship between paternal age and sperm health (2). In fact, studies in mammals have shown an age-related decline in sperm quality with increased incidences of DNA damage, reduced motility, abnormal morphology, and decreased semen volume (7-9).
Mitochondria are essential for their role in creating energy that fuels all cellular functions; however, this process generates reactive oxygen species (ROS) as a byproduct. Low levels of ROS have an important role in cell signaling, hypoxia adaptation, aging, autophagy, immunity, and cell differentiation (10, 11); while high levels of ROS can be detrimental to cellular function and can lead to cell death. Multiple studies in humans and mice have implicated different aspects of mitochondrial function in sperm quality including: mitochondria ultrastructure (12, 13), mitochondrial genome and copy number (14-18), mitochondria protein levels (19-21), and enzyme activity of ETC complexes (22-24). While all these studies imply that mitochondrial integrity and activity are important for proper sperm function, the mechanism behind this relationship is unclear.
Mammalian sperm require a low amount of ROS for multiple aspects of sperm function and successful fertilization of an egg, including: capacitation, hyperactivation, acrosome reaction, and sperm-oocyte fusion (25-27). Interestingly, many studies have found elevated ROS in sperm to be associated with lipid peroxidation, DNA damage, reduced motility, and reduced viability in sperm; although the source of ROS and the mechanism behind ROS-induced sperm defects are unknown (28, 29). Recent studies show that mitochondria-generated ROS through inhibition of electron transport chain results in spermatozoa with reduced motility and lipid peroxidation in vitro (30, 31). Since the level of ROS in semen increases with age (7), understanding ROS-mediated sperm defects can provide insight into male reproductive senescence.
Several studies have documented fertility defects in C. elegans mitochondrial mutants. Mutation in nuo-1, a complex I component of the mitochondria respiratory chain, results in reduced brood size caused by impaired germline development (32). Similarly, clk-1 mutation affects the timing of egg laying, resulting in reduced brood size (33). Both of these mitochondrial mutations impact fertility, but their role(s) in spermatogenesis are unclear. alh-6, the C. elegans ortholog of human ALDH4A1, is a nuclear-encoded mitochondrial enzyme that functions in the second step of the proline metabolism pathway, converting 1-pyrroline-5-carboxylate (P5C) to glutamate (34). We previously revealed that alh-6(lax105) loss-of-function mutants display altered mitochondrial structure in the muscle accompanied by increased level of ROS in adult animals (35). Furthermore, mutation in alh-6 results in the activation of SKN-1/NRF2 (36), an established regulator of oxidative stress response, likely through the accumulation of toxic P5C disrupting mitochondrial homeostasis (35-39). Interestingly, SKN-1 was recently shown to respond to accumulation of damaged mitochondria by inducing their biogenesis and degradation through autophagy (40). Here, we identify a genetic pathway for regulating male reproductive decline stemming from perturbation of mitochondrial proline metabolism leading to redox imbalance, cofactor depletion, and altered mitochondria dynamics; all of which play a part in sperm dysfunction.
RESULTS
Mutation in mitochondrial alh-6 results in diet-independent reduction in fertility
Altered mitochondrial structure and function have been correlated to loss of proper sperm function in different species (16, 41-43). In addition, proper sperm function requires a low level of ROS (25-27), although a specific role for endogenous mitochondrial derived ROS is undefined. ALH-6/ALDH4A1, is a nuclear-encoded mitochondrial enzyme that functions in the second step of proline catabolism, converting 1-pyrroline-5-carboxylate (P5C) to glutamate (Figure S1). We anticipated that mutation of alh-6 may affect the germline, based on our previous assessment of the premature aging phenotypes in somatic cells in alh-6 mutants (35). Using an UV-integrated alh-6::gfp strain under its endogenous promoter, we saw that alh-6 is localized to the mitochondria in the germline of hermaphrodites (Figure S2). We then assessed progeny output of alh-6(lax105) hermaphrodites until egg laying ceased and found a reduction in self-fertility brood size (−12.9%) (Figure 1A). Since the somatic phenotypes of alh-6(lax105) mutants are known to be diet-dependent (35, 36), we examined self-fertility of animals fed the E. coli K-12 bacteria HT115, to determine if the reduced reproductive output is also dependent on the type of bacterial diet ingested. Surprisingly, we found that the self-fertility of alh-6 animals was markedly reduced (−20.7%), when animals were fed the K-12 diet (Figure 1B). alh-6 mutants have similar timing in their progeny output as compared to wild type animals on both diets (Figures S3A-B). Since alh-6 mutants display normal development and reproductive timing, the progeny deficit is not a result of an attenuated reproductive span which reveals the differential impact of alh-6 loss in the soma (diet-dependent) (35) and the germline (diet-independent).
alh-6 fertility defects are sperm-specific
We noted that alh-6 mutant hermaphrodite animals laid twice as many unfertilized oocytes as wild type animals over their reproductive-span (Figure 1C), suggesting a loss of sperm function (44-46). It is notable that alh-6 mutant hermaphrodites lay very few, if any, dead eggs (Figure 1C), suggesting that the loss of ALH-6 activity is not lethal. To determine whether the reduced brood size of alh-6 mutants are due to a general loss of germ cells or a specific defect in oocytes or sperm, we examined the mated-fertility of these animals by mating wild type young adult (day 0-1) males to either wildtype or alh-6 mutant virgin hermaphrodites (in wild type C. elegans, male sperm outcompetes hermaphrodite sperm >99% of the time (47, 48)). We found that the reduced fertility in alh-6 mutant hermaphrodites is fully rescued by wild type sperm, which confirmed that oocyte quality is not impaired but rather, alh-6 hermaphrodite sperm is dysfunctional (Figures. 1D-E).
To better assess the quality of alh-6 mutant sperm, we measured the ability of alh-6 mutant sperm to compete with wild type sperm (49). To differentiate between progeny resulting from mating and progeny that arise from hermaphrodite self-fertilization, we made use of male animals harboring a GFP transgene such that any cross-progeny will express GFP while progeny that arise from hermaphrodite self-sperm will not (Figure 1D). We found that wild type hermaphrodites when mated to alh-6 mutant males have significantly more self-progeny as compared to wild type hermaphrodites mated to wild type males (Figure 1F). This finding indicates a sperm competition deficit of alh-6 males resulting in a brood derived from self-fertilization, which is uncommon after mating has occurred (47). As hermaphrodite C. elegans produce a set amount of sperm during L4 stage before switching exclusively to oogenesis, eventually depleting its reservoir of sperm (47, 50). To assess whether alh-6 mutant sperm are generally-dysfunctional, we mated older hermaphrodites that had depleted their complement of self-sperm and found that alh-6 mutant males are able to produce equal numbers of progeny as wild type males when the need for competition with hermaphrodite sperm is abated (Figure S4A); thus, although alh-6 mutant sperm are impaired for competition, they remain competent for reproduction. This is similar to recent study on comp-1, a mutation which results in context-dependent competition deficit in C. elegans sperm (51). Similarly, older alh-6 mutant hermaphrodites mated to young wild type males yield similar level of progeny as age-matched WT hermaphrodites, which further supports a model where sperm, but not oocytes, are defective in alh-6 mutants (Figure S4B).
Similar to mammals, the contribution of sperm to fertility in C. elegans is dictated by distinct functional qualities, which include: sperm number, size, and motility (49, 52, 53). In C. elegans, male sperm are larger and faster than hermaphrodite sperm, which affords a competitive advantage (53). We next sought to define the nature of the sperm competition defect in alh-6 mutants by measuring sperm size, motility, and number in alh-6 mutants compared to wild type animals. One day after spermatogenesis initiation (at the L4 larval stage of development), alh-6 adult hermaphrodites have a reduced number of sperm in the spermatheca as compared to wild type (Figure S5A), which is correlated with the reduced self-fertility observed (Figures 1A-B). In contrast, age-matched alh-6 mutant males have similar numbers of sperm as WT males, suggesting that they have a similar rate of production (Figure 2A). We next examined sperm size in day 1 adult males and discovered that alh-6 mutant spermatids are significantly smaller as compared to wild type (Figure 2B). To achieve motility, C. elegans sperm must be activated to allow pseudopod development, and this development requires protease activation (54) (Figure 5B). In vitro, sperm activation can be recapitulated by treatment of isolated spermatids with Pronase (55). In addition to reduced size, the percentage of activated spermatozoa was significantly reduced in alh-6 mutants as compared to wild type (Figure 2C). Taken together, the reduction of sperm quantity and quality (size and activation) are contributors to the reduced fertility in alh-6 mutants.
Transcriptional signatures define temporal phenotypes of alh-6 activity
We first identified alh-6 mutant in a screen for activators of the cytoprotective transcription factor SKN-1/NRF2 using gst-4p::gfp as a reporter (35, 36). When activated, SKN-1 transcribes a variety of gene targets that collectively act to restore cellular homeostasis. However, this can come with an energetic cost with pleiotropic consequences (35, 36, 56-61). alh-6 mutants have normal development, but display progeric phenotypes towards the end of the normal reproductive span (35) indicating a temporal switch in phenotypic outcomes. We reasoned that the temporally controlled phenotypes in the alh-6 mutants could be leveraged to identify potential mechanisms by which alh-6 loss drives cellular dysfunction. As SKN-1 is activated in alh-6 mutants after day 2 of adulthood (35), we defined genes that display differentially altered expression in the L4 developmental stage, when spermatogenesis occurs, as compared to Day 3 adults (post SKN-1 activation). We performed RNA-Seq analyses of worms with loss of alh-6 and identified 1935 genes in L4 stage animals and 456 genes in Day 3 adult animals that are differentially expressed (+/- Log2 (fold change), 0.05 FDR). Intriguingly, the gene expression changes at these two life periods had distinct transcriptional signatures (Figures 3A-B; Figure S6). Because the loss of alh-6 drives compensatory changes in normal cellular metabolism, which later in life results in the activation of SKN-1, we expected to identify significant changes in both metabolic genes and SKN-1 target genes. Supporting this hypothesis, the gene ontology (GO) terms most enriched include oxidoreductases and metabolic enzymes in L4 stage animals (Figure 3A) and SKN-1-dependent targets such as glutathione metabolism pathway genes in Day 3 adults (Figure 3B). Importantly, our transcriptomic analysis recapitulated the temporally-dependent phenotypic outcomes resulting from alh-6 loss; genes in the pseudopodium and germ plasm GO terms class displayed reduced expression in L4 (Figure 3A), which could impact C. elegans spermatogenesis. In contrast, genes in the muscle-specific GO term class displayed increased expression in day 3 adults (Figure 3B), which is when SKN-1 activity is enhanced in the muscle of alh-6 mutants (36). Taken together, the transcriptome analysis of alh-6 mutants is diagnostically relevant and informative for defining drivers of organism-level phenotypic changes in animals with altered proline catabolism.
FAD mediates sperm functionality and competitive fitness
The strong enrichment of genes whose protein products utilize and/or bind co-factors or coenzymes was intriguing as the maintenance of metabolic homeostasis and the redox state of the cell requires a sophisticated balance of multiple cofactors (Figure 4A). In fact, the proline catabolism pathway utilizes multiple cofactors to generate glutamate from proline; PRDH-1 uses FAD as a co-factor while ALH-6 utilizes the reduction of NAD+. In the absence of ALH-6, PRDH-1 would continue to deplete FAD, which would activate compensatory pathways to maintain metabolic homeostasis in addition to activating pathways to detoxify P5C (oxidoreductases, P5C reductase, etc.). In light of this hypothesis we measured FAD and found a significant reduction in alh-6 mutant animals (Figure 4B). As such, we predicted that restoration of FAD levels might alleviate the sperm-specific phenotypes of alh-6 mutants. Dietary supplementation of riboflavin has been shown to increase cellular FAD levels (62, 63), and when fed to alh-6 mutants, it restored sperm function. We found that wild type hermaphrodites mated to alh-6 mutant males that were fed a diet supplemented with 2.5mM riboflavin produced significantly more total progeny than alh-6 males fed the standard OP50 diet (Figure S7). Moreover, riboflavin supplementation was sufficient to restore male sperm size (Figure 4C) and also rescued the impaired activation (Figure 4D) of male sperm in alh-6 mutants. Taken together, these data suggest that loss of alh-6 leads to a decrease in cellular FAD levels that drives sperm dysfunction.
Loss of cellular proline catabolism is not causal for sperm defects in alh-6 mutants
We were curious to uncover additional molecular mechanisms that underlie the loss of sperm function in alh-6 mutants. To do this, we performed an EMS mutagenesis screen to identify suppressors of the induced gst-4p::gfp expression phenotype in alh-6 mutants (Figure 5A) (35). We identified one suppressor allele, lax228, which we mapped to right arm of chromosome IV between F49E11 and Y57G11B SNPs. We then generated a list of candidate genes in this region with non-synonymous mutations in the exons of protein coding genes using whole genome sequencing data of the alh-6 mutant compared to the suppressor mutant alh-6(lax105); lax228 (64). We tested each of these genes by RNA interference (RNAi) in the alh-6;gst-4p::gfp strain to phenocopy the suppressor. RNAi of B0513.5, hereafter referred to as prdh-1 as it encodes for proline dehydrogenase, was the only RNAi target that phenocopied the lax228 mutant (Figures S8A-B). PRDH-1 catalyzes the first enzymatic step of proline catabolism (Figure 5B), converting proline to P5C. Importantly, this enzyme is linked to several of the phenotypes of alh-6 mutants including the generation of P5C (9) and the continued reduction of FAD, documented above (Figure 4B). We also examined the expression of the proline catabolism pathway genes from our RNA-Seq analysis and discovered a significant increase in the expression of enzymes that would prevent the accumulation of P5C in alh-6 mutant L4 animals, before irreparable damage occurs (Figure 5C). Specifically, there was an increase in expression of pyrroline-5-carboxylate reductase (M153.1/PYCR) that converts P5C back to proline and ornithine transaminase(oatr-1/OAT) that converts P5C to ornithine. Surprisingly, the expression of pyrroline-5-carboxylate synthase (alh-13/P5CS) was also increased, however P5CS has two enzymatic functions: glutamate kinase (GK) and γ-glutamyl phosphate reductase (GPR) activities that impact additional nodes of cellular metabolism. Moreover, since proline itself has important roles in cellular protection, the increased expression of P5CS might be an important stress response, but with pleiotropic consequences as it would deplete glutamate and increase an already accumulating pool of P5C.
To determine how the total loss of proline catabolism would affect C. elegans reproduction, we examined the alh-6; prdh-1 double mutant in our panel of reproduction and sperm quality assays. The reduction in spermatid size (Figure 5D) and impairment of spermatid activation (Figure 5E) in alh-6 mutants are both suppressed by loss of prdh-1. In addition, the prdh-1 mutation restored the reduced self-fertility (Figure S9A), lower hermaphrodite sperm count (Figure S9B), and suppressed the increased laying of unfertilized oocytes of the alh-6 single mutant (Figure S9A). Finally, the reduced ability of alh-6 male sperm to compete against wild type hermaphrodite sperm was abrogated in the alh-6;prdh-1 double mutant (Figure S9C). These results were surprising as they reveal that loss of flux through the mitochondrial proline catabolism pathway is benign for animal reproductive fitness, but suggests instead that P5C accumulation is instrumental in driving sperm dysfunction in alh-6 animals.
Endogenous ROS drives alh-6 sperm defects
Several studies have examined the impact of exogenous ROS-inducing electrophiles on sperm function, but the impact of endogenously produced ROS on sperm function remains poorly defined. The continuous generation of P5C by PRDH-1 leads to the accumulation of this highly toxic and unstable biomolecule, which can lead to redox imbalance and impair the normal function of germ cells as it does for somatic tissues (35, 37-39, 65). If the sperm defects in the alh-6 mutants are a result of a loss of redox and/or ROS homeostasis, then we anticipated that antioxidants could alleviate these phenotypes. We supplemented the diet of alh-6 mutant males with the antioxidant N-acetylcysteine (NAC), from birth through reproductive maturity, and re-measured the reproductive parameters of these animals. NAC supplementation restored spermatid size and activation of alh-6 animals to WT levels (Figures 5D-E). Antioxidant supplementation in wild type (Figures S10A-B) or alh-6; prdh-1 double mutants had no effect (Figures S10C-D). Collectively, these data suggest that endogenous production of ROS is causative for the sperm dysfunction seen in alh-6 animals. In addition, this study reveals that antioxidant supplementation can act as a treatment to overcome reproductive deficiencies stemming from defects in specific cellular metabolic pathways.
Mitochondrial dynamics regulate spermatid function
Although there is a clear and documented role for mitophagy in the clearance of paternal mitochondria post-fertilization in C. elegans, the role(s) for mitochondrial dynamics and turnover in sperm function prior to zygote formation are unclear. We first examined mitochondrial dynamics in wild type sperm by staining with the fluorescent, mitochondrial-specific dye JC-1 (ThermoFisher), and noted that each spermatid on average contained multiple discernable spherical mitochondria that are mostly not fused (Figures 6A, 6B, 6G). Previous studies in yeast and cultured mammalian cells have shown that when cells are exposed to mild stress, the initial response of mitochondria is to fuse in an attempt to dilute damage (66-68). Indeed alh-6 mutant spermatids had mitochondria that were more interconnected (Figures 6C, 6D, 6G) as compared to wild type spermatids, which supports our finding that these sperm are under oxidative stress (Figure 5). Loss of prdh-1, which restores sperm function (Figure 5), returned spermatid mitochondria to more punctate structures (Figures 6E, 6F, 6G). Similarly, treatment with the antioxidant NAC returned alh-6 mutant mitochondria in spermatids to wild type levels of fusion (Figure 6H). The JC-1 dye accumulates in mitochondria in a membrane potential (ΔΨ)-dependent manner, and as concentration exceeds threshold, its fluorescence switches from green to red emission; thus, a higher red-to-green fluorescence ratio is indicative of healthier mitochondria, with higher ΔΨ. alh-6 mutant spermatids have reduced red:green JC-1 fluorescence that indicates a lower mitochondrial ΔΨ, and an accumulation of unhealthy mitochondria (Figure 6I).
The role of mitochondrial dynamics (fusion and fission) in the maturation of sperm has not been studied; however recent work has revealed that the mitochondrial fusion and fission machinery are important for the elimination of paternal mitochondria post-fertilization (69). FZO-1 is required for proper fusion of the mitochondrial membranes and DRP-1 is required for mitochondrial fission (70, 71). The balance of this fusion and fission machinery in the upkeep of mitochondrial homeostasis allows cells to respond to changes in metabolic needs and external stress (72, 73). RNAi of fzo-1 suppressed the enhanced fusion observed in alh-6 mutant spermatid mitochondria indicating mitochondrial fusion is active in spermatids with impaired proline catabolism (Figure 6J). We next examined spermatids from drp-1 mutant animals and observed a greater level of mitochondrial fusion as compared to wild type and alh-6 mutant spermatids (Figure 6K). We observed a synergistic level of mitochondrial fusion in spermatids derived from alh-6; drp-1 double mutants. This finding is consistent with previous studies in yeast which reveal that defects in fusion can be compensated for by changes in the rates of fission and vice versa (72, 73). In support of our model where mitochondrial dynamics act as a major driver of the sperm-specific defects in alh-6 mutants, we discovered that loss of drp-1, which results in increased mitochondrial fusion (like that observed in alh-6 mutants), also reduces sperm activation (Figure 6L). Moreover, reduced fzo-1 does not alter activation in wild type sperm, but restores activation in alh-6 sperm (Figure 6M); suggesting increased fusion in alh-6 sperm mitochondria is impairing proper function. Taken together, these data support a model where loss of mitochondrial proline catabolism induces mitochondrial stress, activating mitochondrial fusion, which can subsequently eliminate damage in order to preserve functional mitochondria (Figure 6N). These data also reveal a functional role for mitochondrial fusion and fission in maintaining proper sperm function. In conclusion, our studies define mitochondrial proline catabolism as a critical metabolic pathway for male reproductive health.
DISCUSSION
Here we investigate the effects of mutation in the mitochondrial enzyme gene alh-6, and the associated increased ROS levels on male fertility stemming from defective mitochondrial proline catabolism. We found that alh-6 mutants show a reduction in brood size that is sexually dimorphic; defects in sperm function but not oocytes contribute to reduced hermaphrodite fertility. As societal factors continue to push individuals to wait longer to have children, our studies are of critical importance to elucidate how restoring and maintaining functional amino acid catabolism during aging will promote reproductive success.
Although C. elegans is a well-established organism for studying aging and reproduction, with several studies describing hermaphrodite reproductive senescence, many questions regarding the basis of male reproductive decline remain unanswered. Decades of work have shown that exposure to pollution, toxins, xenobiotics, and other ROS-inducing compounds can prematurely drive the loss of sperm function (29, 74, 75), but the impact that normal cellular metabolism plays on sperm function and the identification of specific molecules that can mediate sperm quality are not well-defined. In this study we characterized a new role for mitochondrial proline catabolism and FAD homeostasis in the maintenance of proper sperm function. Perturbation of this pathway, through mutation of alh-6/ALDH4A1, increases ROS, causing metabolic stress and increased mitochondrial fusion in spermatids, which results in impaired sperm function and premature reproductive senescence
Mutation in proline dehydrogenase (PRODH) in humans results in hyperprolinemia type I (HPI), while mutation in delta-1-pyrroline-5-carboxylate dehydrogenase (ALDH4A1/P5CDH) results in hyperprolinemia type II (HPII). Surprisingly, in C. elegans, mutation in upstream proline dehydrogenase gene, prdh-1, is able to suppress all the reproductive defects in alh-6 mutants suggesting that overall reduction in proline catabolism is not causal for the observed reproduction phenotypes. In humans, the diagnosis of HPI and HPII are both through by elevated level of proline in plasma, with addition of high level of P5C in HPII patients. The symptoms of HPI varies on severity depending on the reduction of PRODH activity (type of mutation) and are characterized by neurological, auditory, and renal defects (53). Although proline catabolism has not previously been shown to have a direct role in fertility, human fertility studies have shown that the addition of proline in cryopreservation medium improves sperm mobility and preservation of membrane integrity upon thawing (60, 61). This study reveals that in C. elegans, proline catabolism impacts several functional qualities of sperm. Loss of proline catabolism results in smaller sperm with impaired activation, two qualities that directly impact competitive advantage. As such, proline biosynthesis, catabolism, and steady state concentrations must be tightly regulated, and the importance of proline in cellular homeostasis may help explain the transcriptional responses measured in animals with dysfunctional alh-6 (Figures 3 and 4).
Our previous work defined the age-dependent decline in function of somatic tissues, particularly muscle in animals lacking functional ALH-6 (35, 36), which is not manifested until Day 3 of adulthood. This study reveals that although somatic phenotypes are observed post-developmentally, the germline and specifically sperm are sensitive to loss of alh-6 much earlier in development (phenotypes assayed at L4 or Day 1 of adulthood), with many physiological consequences from dysregulation of metabolism. Reproductive senescence is a field of growing significance as the number of couples that choose to delay having children increases. About 30-40% of all male infertility cases are associated with increased levels of ROS, yet we don’t understand the underlying mechanism (76). Additionally, sperm quality has been shown to decline with age, as ROS content increases with age (7, 8, 77, 78), demonstrating the link between ROS and male reproductive senescence. Our study demonstrates that perturbation of mitochondrial proline catabolism, particularly mutation in alh-6/ALDH4A1, leads to redox imbalance and impaired sperm function. Importantly, addition of antioxidants to diet can abrogate this sperm dysfunction (Figures 5D-E), implicating the potential therapeutic effects of antioxidant supplement in male infertility arising from redox imbalance.
Recent studies have focused on the role of NAD+ metabolism in cellular health, while the impact of FAD has received less attention. FAD levels are diminished in alh-6 animals specifically at the L4 stage when spermatogenesis is occurring (Figure 4B). Riboflavin (Vitamin B2) is a precursor to FAD and flavin mononucleotide (FMN) cofactors that are needed for metabolic reactions (like proline catabolism and mitochondrial oxidative phosphorylation) to maintain proper cellular function. Despite its importance, humans lack a riboflavin biosynthetic pathway and therefore require riboflavin from exogenous sources (79). Insufficient intake can lead to impairment of flavin homeostasis, which is associated with cancer, cardiovascular diseases, anemia, neurological disorders, fetal development, etc. (79). Our study suggests that riboflavin and FAD play critical roles in reproduction as alh-6 mutants suffer from sperm dysfunction driven by a reduction in FAD levels (Figure 4). Importantly, these sperm specific defects can be corrected by dietary supplementation of vitamin B2, which in light of the exceptional conservation of mitochondrial homeostatic pathways, suggest the nutraceutical role vitamin B2 could play in sperm health across species.
Our study also demonstrates that spermatids lacking alh-6 have increased mitochondrial fusion; a perturbation at the mitochondrial organelle structure-level that contributes to the sperm-specific phenotypes observed. In addition to prior work showing fzo-1/MFN1/MFN2 and drp-1/DRP-1 to be important for mitochondrial elimination post-fertilization (69), our work reveals that mitochondrial fission and fusion machinery are present and active in spermatids and that perturbation of these dynamics can affect sperm maturation and competitive fitness (Figure 6). Future work to define how alh-6 spermatids use mitophagy, which can clear damaged mitochondria, will be of interest. In conclusion, our work identifies proline metabolism as a major metabolic pathway that can impact sperm maturation and male reproductive success. Moreover, these studies identify specific interventions to reverse the redox imbalance, cofactor depletion, and altered mitochondria dynamics, all of which play a part in sperm dysfunction resulting from proline metabolism defects.
AUTHOR CONTRIBUTIONS
S.P.C. designed the study; C-A.Y., D.L.R., N.M., S.P., and S.P.C. performed the experiments; C-A.Y. and S.P.C. analyzed data and wrote the manuscript. All authors discussed the results and commented on the manuscript.
COMPETING INTERESTS
The authors declare no competing interests.
DATA AVAILABILITY
All relevant data are available from the authors.
METHODS
C. elegans strains and maintenance
C. elegans were cultured using standard techniques at 20°C. The following strains were used: wild type (WT) N2 Bristol, CB4856 (HW), SPC321[alh-6(lax105)], SP326[alh-6p;alh-6::gfp], CL2166[gst4-p::gfp], SPC223[alh-6(lax105);gst-4p::gfp], and CU6372[drp-1(tm1108)]. Double and triple mutants were generated by standard genetic techniques. E. coli strains used were as follows: B Strain OP50(80) and HT115(DE3) [F-mcrA mcrB IN(rrnD-rrnE)1 lambda- rnc14::Tn10 λ(DE3)](81). For dietary supplement assays, the following were added to the NGM plate mix to final concentration: 5mM NAC, 2.5mM riboflavin. RNAi experiments done using OP50 RNAi E. coli B strain as described in (82) yielded similar results as HT115-based RNAi. All strains were adapted to diets for at least three generations and strains were never allowed to starve.
EMS mutagenesis
Ethyl methanesulfonate (EMS) mutagenesis was performed as previously described (57). Briefly, SPC223[alh-6(lax105);gst-4p::gfp] was mutagenized with EMS, and F2 worms with reduced GFP expression (indicating suppression of SKN-1 activation) were selected. prdh-1(lax228) was isolated and mapped to chromosome IV. Whole genome sequencing and injection rescue confirmed mutant sequence identity.
Microscopy
Zeiss Axio Imager and ZEN software were used to acquire all images used in this study. For GFP reporter strains, worms were mounted in M9 with 10mM levamisole and imaged with DIC and GFP filters. For sperm number assay samples were imaged with DIC and DAPI filters in z-stacks. For sperm size and activation assays, dissected sperm samples were mounted with coverslip and imaged at 100x with DIC filter on two different focal planes for each field to ensure accuracy. For sperm mitochondria assays, dissected sperm samples were covered with coverslip and imaged at 100x with DIC, GFP, and RFP filters in z-stacks to assess overall mitochondria content within each spermatid.
Fertility assay
Worms were egg prepped and eggs were allowed to hatch overnight. The next day synchronized L1 larvae were dropped on NGM plates seeded with either OP50 or HT115. 48 hrs later, at least ten L4 hermaphrodites for each genotype were singled onto individual plates and moved every 14 hours until egg laying ceased. Progeny were counted 48 hours after the singled hermaphrodite was moved to a different place. Plates were counted twice for accuracy.
Mated reproductive assay
Males were synchronized by egg laying, picked as L4 larvae for use as young adults for mating experiments. L4 hermaphrodites were each put on a plate with 30ul of OP50 seeded in the center together with three young adult males. 24hrs post mating, males are removed, and each hermaphrodite was moved to a new plate every 24 hr until egg laying ceases. Progeny were counted 48 hours after hermaphrodite was moved from the plate. For sperm competition assay progeny with GFP fluorescence were counted from the cohort. Plates were counted twice for accuracy.
Cofactor Measurements
Worms were egg prepped and eggs were allowed to hatch overnight. Next day, synchronized L1s were dropped on NGM plates seeded with concentrated OP50. FAD levels are measured following directions in FAD Colorimetric/Fluorometric Assay Kit (K357) from BioVision.
Sperm Number Assay
Worms were egg prepped and eggs were allowed to hatch overnight. Next day, synchronized L1s were dropped on seeded NGM plates. 72 hrs post drop, day 1 adult animals were washed 3x with 1xPBST, fixed with 40% 2-propanol, and stained with DAPI for 2 hrs. Samples were washed for 30min with PBST, mounted with vectashield mounting medium, and covered with coverslip to image.
Sperm Size Assay
Males were isolated at L4 stage 24 hour before assay. For each strain, five day 1 adult males were dissected in 35µL pH 7.8 SM buffer (50mM HEPES, 50mM NaCl, 25mM KCl, 5mM CaCl2, 1mM MgSO4, 10mM dextrose) with DAPI to release spermatids and imaged.
Sperm Activation with Pronase
Males were isolated at L4 stage 24 hour before assay. For each strain, five day 1 adult males were dissected in 35µL pH 7.8 SM buffer (50mM HEPES, 50mM NaCl, 25mM KCl, 5mM CaCl2, 1mM MgSO4, 1 mg/ml BSA) supplemented with 200µg/mL Pronase to release spermatids. Another 25ul of same solution was added and the spermatids were incubated at RT for 30 min for activation to occur.
Sperm Mitochondria Staining
Males were isolated at L4 stage 24 hour before assay. For each strain, five day 1 adult males were dissected in 35µL pH 7.8 SM buffer (50mM HEPES, 50mM NaCl, 25mM KCl, 5mM CaCl2, 1mM MgSO4, 1mg/ml BSA) with JC-1(Thermo Fisher Scientific T3168) added to 10µM final concentration. Another 25ul of same solution is added and the spermatids are incubated at RT for 10 min. Then the slide was washed three times with 100ul SM buffer before imaging.
RNA-sequencing
Worms were egg prepped and eggs were allowed to hatch overnight. Next day, synchronized L1s were dropped on NGM plates seeded with concentrated OP50. 48 and 120 hrs post drop, L4 animals and day 3 adult animals, respectively, were washed 3 times with M9 and frozen in TRI Reagent at −80°C. Animals are homogenized and RNA extraction is performed following protocol in Zymo Direct-zol RNA Isolation Kit. RNA samples were sequenced and analyzed by Novogene.
Statistical analysis
Data are presented as mean ± SEM. Comparisons and significance were analyzed in Graphpad Prism 7. Comparisons between two groups were done using Student’s Test. Comparisons between more than two groups were done using ANOVA. For sperm activation assays, Fisher’s Exact Test was used and p-values are adjusted for multiple comparisons. *p<0.05 **p<0.01 *** p<0.001 ****<0.0001
ACKNOWLEDGEMENTS
We thank K. Han and L. Thomas for technical assistance, and H. Dalton and A. Hammerquist for critical reading of the manuscript. Some strains were provided by the CGC, which is funded by the NIH Office of Research Infrastructure Programs (P40 OD010440). This work was funded by the NIH (GM109028) and the Ellison Medical Foundation (S.P.C.) and the American Federation of Aging Research (C-A.Y. and S.P.C).
REFERENCES
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.
- 16.↵
- 17.
- 18.↵
- 19.↵
- 20.
- 21.↵
- 22.↵
- 23.
- 24.↵
- 25.↵
- 26.
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.
- 39.↵
- 40.↵
- 41.↵
- 42.
- 43.↵
- 44.↵
- 45.
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.
- 59.
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵