Autophagy's varied roles in sperm survival and quality

Autophagy is a ‘self-eating’ catabolic process of degrading cytoplasmic materials in lysosomes, which plays a fundamental role in various physiological or pathological processes (Mizushima and Levine, 2020). Our knowledge of autophagy is dramatically expanding day by day. The 60-year developmental history of autophagy (Fig. 1a) reflects the progress of science and leads to a new era in our understanding of autophagy in human health. Known as a double-edged sword, autophagy can serve as protective mechanism by eliminating damaged organelles and providing energy for cellular renovation, yet, autophagy may also contribute to cell damage (Shintani and Klionsky, 2004).

Autophagy-related genes (Atgs) and enzymes are identified in the three types of autophagy: macroautophagy, chaperone-mediated autophagy (CMA) and microautophagy (Fig. 1b). Firstly, macroautophagy, as the major form of autophagy, is a conserved dynamic process, which consists of induction (phagophore formation), nucleation, elongation, closure (autophagosome formation), fusion (autophagosome and lysosome into autophagolysosome) and degradation, thus degrading organelles and proteins for the synthesis of new macromolecules or as a source of energy. Due to the difference of destructive targets, macroautophagy (hereafter called autophagy) is divided into non-selective autophagy (targeting bulk cytosol) and selective autophagy (targeting specific organelles), such as mitophagy, lipophagy, pexophagy, ribophagy, reticulophagy, aggrephagy and ferritinophagy. The involved signaling pathways and key molecules are shown in Fig. 1b. Secondly, in CMA, proteins carrying the KFERQ motif are recognized by the Hsp70 chaperone, which interacts with lysosome membrane protein LAMP-2A, leading the translocation of the bound protein into the lysosome (Boya et al., 2013). Thirdly, microautophagy is characterized by the direct engulfment of cytoplasmic material into the lysosome through invagination and pinching off. To sustain homeostasis, autophagy, together with the ubiquitin-proteasome system (UPS), constitutes the major cellular quality control systems for the degradation and disposal of organelles (Pohl and Dikic, 2019). Depending on these emerging regulatory processes and functions, autophagy regulates cell growth, survival and cell death (Dai et al., 2020).

As shown in Table I, the knockdown/knockout (KD/KO) of Atgs in mammalian testis has shown the role of specific molecules in the regulation of distinct testicular cells. Sperm is the male reproductive cell or gamete that fuses with an oocyte to form a fertilized oocyte that develops into an individual, so sperm is the key to male fertility (Kimble and Page, 2007). The fate of sperm refers to an intricate process from the origin of sperm to its disappearance. Sperm is produced by spermatogenesis in the male reproductive gland, the testis, whose functions are controlled by the hypothalamic–pituitary–gonadal (HPG) axis. Once completing spermatogenesis in the testis, sperm move into the epididymis to mature into spermatozoa. Subsequently, mature sperm-containing semen is ejaculated into the female reproductive tract and for fertilization of the oocyte. Overall, the fate of sperm may be categorized as pre-testicular, testicular, post-testicular according to the timeline (Krausz, 2011).

In pre-testicular processes, the HPG axis is the neuroendocrine network that regulates sexual development and reproduction (Navarro and Tena-Sempere, 2011). In the hypothalamus, kisspeptin (a neuropeptide), following neurokinin B and dynorphin, signals directly to GnRH neurons to orchestrate pulsatile GnRH release (Skorupskaite et al., 2014). GnRH binds to a membrane receptor in the pituitary, which secretes LH and follicle-stimulating hormone (FSH) (Kaprara and Huhtaniemi, 2018). After release from the pituitary, LH and FSH interact with the LH/FSH receptor in testicular cells, respectively, to initiate and maintain spermatogenesis.

In testicular processes, the testis is responsible for producing spermatozoa through spermatogenesis. Testicular deficiency often leads to spermatogenic failure caused by conditions other than obstruction or HPG dysfunction. The entire process of spermatogenesis can be divided into three stages (Fig. 2), as follows: (i) during spermatocytogenesis, spermatogonia undergo mitosis to develop into primary spermatocytes (PSC); (ii) during spermatidogenesis, the PSC undergo meiosis I to form secondary spermatocytes, which divide into haploid round spermatids via meiosis II; and (iii) during spermiogenesis, round spermatids are differentiated into the elongated spermatids and then spermatozoa (Cheng and Mruk, 2012; Staub and Johnson, 2018). The last step in the testis is spermiation. Elongated spermatids are then released from the Sertoli cells into the seminiferous tubule lumen.

In post-testicular processes, after spermatogenesis and spermiation, spermatozoa are transported from the testis to the epididymis for obtaining progressive motility and fertility (Sullivan and Mieusset, 2016). Subsequently, mature sperm are ejaculated into the female reproductive tract to undergo capacitation and activation (Jin and Yang, 2017). Eventually, the fusion of the sperm and the oocyte into a zygote, namely fertilization, represents the disappearance of sperm and appearance of a new individual, which indicates the final step for the fate of sperm.

Recently, increasing evidence has shown that autophagy functions in multiple mammalian organs or systems, including the embryo (Yoshii et al., 2016), placenta (Agrawal et al., 2015), liver (Komatsu et al., 2005), brain (Kaushik et al., 2011), heart (Nakai et al., 2007; Lu et al., 2016), skeletal muscle (Masiero et al., 2009), intestine (Cadwell et al., 2008), pancreas (Ebato et al., 2008), kidney (Hartleben et al., 2010), and the male (Gao et al., 2018) and female reproductive systems (Gawriluk et al., 2014) (Table II). Atgs have been implicated in numerous cellular events in the fate of sperm, such as GnRH secretion, LH, FSH and testosterone production, ectoplasmic specialization (ES) assembly, acrosome biogenesis and paternal mitochondria elimination (PME), shedding light on new therapeutic targets in the autophagy pathway for male subfertility or infertility (Nian et al., 2019). To date, there has been no effective and timely summary on the role of autophagy in the entire fate of sperm. Given that autophagy is known as a double-edged sword, the most fundamental question of whether autophagy plays a protective or a harmful role in the regulation of sperm remains unclear. In this narrative review, we will summarize the role of autophagy in the entire fate of sperm, including pre-testicular (HPG axis), testicular (spermatogenesis and spermiation) and post-testicular (sperm maturation and fertilization), and evaluate the clinical impact of autophagy on the sperm and male fertility. Our review will provide a better understanding of autophagy in the fate of sperm, laying the foundation for the further study of sperm, which can help us identify more therapeutic targets for male subfertility or infertility.

Atg5 deficient mice show neuronal dysfunction and hypogonadism (Yoshii et al., 2016), implying the key role of autophagy in HPG axis. Warburg Micro syndrome (WARBM) is a rare autosomal recessive genetic disease characterized by defective neurodevelopmental and ophthalmological phenotypes, such as microcephaly, microcornea, optic atrophy, lower limb spasticity and hypogonadotropic hypogonadism (Nian et al., 2019). RAB18, a member of small G protein, has been demonstrated to be a causative gene of WARBM in human. Rab18−/− mice neurons exhibit abnormal lysosomal transport and autophagosome marker LC3-II expression, suggesting the involvement of aberrant autophagy activities. Furthermore, RAB18 protein colocalizes with the lysosomal regulator RAB7 protein, which is upregulated in Rab18-deficient neurons, indicating a compensatory effect of Rab7 (Nian et al., 2019).

Rab3GAP2, another causative gene of WARBM, is reported to be a guanosine nucleotide exchange factor for activating Rab18. Rab3GAP2 mutation in Drosophila leads to the disorganization of autophagosomals/late endosomal compartments and lysosomal transport, finally perturbing autolysosome morphology. Meanwhile, Rab3GAP2, RAB18 and ATG6/BECN1 (subunits of Vps34 complexes) are co-located on autophagosomal and autolysosomal membranes. Consequently, the Rab3GAP-RAB18 module regulates autolysosomal maturation via its interaction with the Vps34 complexes. Taken together, Rab3GAP-RAB18-VPS34-mediated autolysosomal maturation may contribute to the development of WARBM, affecting gonadal function (Takats et al., 2021).

In addition to Rab3GAP and RAB18, mutations in GnRHR are also associated with hypogonadotropic hypogonadism. Houck et al. (2014) show that a missense mutation E90K in GnRHR induces misfolded proteins due to molecular chaperones Hsp70 and Hsp40 JB12. Interaction between the Hsp40 and Vps34 (autophagy initiation) complex permits the selective degradation of membrane proteins via an autophagy pathway.

Microcystin-leucine arginine (MC-LR), a kind of toxin produced by cyanobacterial, is reported to induce GnRH neurons apoptosis, resulting in a reduction of serum testosterone and spermatogenesis disruption in mice. An autophagy inhibitor (3-MA, PI3K inhibitor) aggravates MC-LR-induced apoptotic cell death in GT1-7 mouse hypothalamic GnRH neuronal cell line, implying a potential protective role of PI3K from apoptosis of GnRH neurons upon MC-LR exposure (Jin et al., 2021).

GnRH and kisspeptin/neurokinin B/dynorphin neurons in the hypothalamus regulate the age-related physiological decline in energy metabolism, hormone regulation and reproduction. Mammalian target of rapamycin (mTOR), autophagy and SIRT1 have been recognized as critical factors or pathways in hypothalamus-mediated aging progression (Kim and Choe, 2019).

ZNF216 is an identified causative gene for Gordon Holmes syndrome, characterized by ataxia, dementia and hypogonadotropic hypogonadism in humans. KD of RNF216 leads to the augmentation of BECN1 and migration defects in the GN11 immature GnRH neuronal cell line, which can be rescued by the autophagy inhibitors chloroquine (CQ) and 3-MA. Meanwhile, rapamycin (an autophagy activator) can suppress the GN11 cell migration. Therefore, RNF216 regulates GnRH neuron migration via inhibiting BECN1-mediated autophagy, suggesting a potential contribution of autophagy to hypogonadotropic hypogonadism (Li et al., 2019).

KiSS1 plays a key role in the activation of the HPG axis in regulating the onset of puberty and reproductive function. KiSS1-derived peptides, i.e. kisspeptins, signal through the G-protein coupled receptor GPR54 (Corno and Perego, 2019). In breast cancer cells, KiSS1 has been reported to down-regulate two Atgs (ATG5 and ATG7) and inhibit conversion of LC3-I to LC3-II, whereas KiSS1 KD leads to the down-regulation of p62 and up-regulation of Beclin 1. Therefore, KiSS1 inhibition is associated with promotion of autophagy (Kaverina et al., 2017).

Kisspeptin preserves mitochondrial function by inducing mitophagy and autophagy in the hippocampus of aging rat brains and in a human neuronal cell line via a series of signaling pathways, including Ca2+/CaM-dependent protein kinase β (CaMKKβ), AMP-activated protein kinase (AMPK) and Unc-51 like autophagy activating kinase (ULK1) (Mattam et al., 2021).

Environmental endocrine disruptors can disturb HPG axis by autophagy. Cadmium (Cd) decreases the serum concentrations of GnRH, FSH, LH and testosterone. Naringenin (Nar) suppresses MDA and H₂O₂ production and protects the testis from Cd-induced autophagy by downregulating P62 and LC3-II expression. Therefore, Nar protects the testis from Cd-induced toxicity (Wang et al., 2021). Paternal exposure to arsenic results in oxidative stress, autophagy and mitochondrial impairment in the HPG axis of pubertal male offspring. Specifically, autophagic cell death-related genes and proteins, such as Atg3, Atg5, Beclin 1, p62, Atg12, PI3K and mTOR, are disturbed in the HPG tissues of the pubertal male mice offspring (F1-generation) (Ommati et al., 2019). Similarly, in the mature male offspring, arsenic induces autophagic alterations and mitochondrial impairments in the HPG-sperm axis through AMPK/tuberous sclerosis complex (TSC) (tuberous sclerosis complex)/mTOR and LC3-related pathways (Ommati et al., 2020). Zearalenone, a non-steroidal estrogen mycotoxin, can induce mitochondrial dysfunctions through overproduction of reactive oxygen species (ROS) and aberrant autophagy pathways, finally disturbing the synthesis and secretion of mammalian sex steroid hormones (Zheng et al., 2019).

Autosomal dominant familial neurohypophyseal diabetes insipidus (adFNDI) is a progressive and inherited neurodegenerative disorder. The mutation of vasopressin (VP) from posterior pituitary nerve terminals can cause adFNDI, which is rescued by the inhibition of autophagy. Consequently, autophagy is a prosurvival mechanism in cells expressing an adFNDI mutant VP transgene (Castino et al., 2005a). Subsequently, Castino et al. (2005b) demonstrated that autophagy-mediated cell death is a two-hit process: the first hit of autophagy by cellular stress degrades the misfolded mutant protein, thus autophagy is pro-survival; whereas, a second insult triggers an autophagy-dependent apoptosis.

Spermatocytogenesis is the first stage of spermatogenesis. In this process, a diploid spermatogonial stem cell (SSC) undergoes mitosis to renew the stem cell pool or differentiate into two diploid PSC (Staub and Johnson, 2018). In non-primates, spermatogonia are composed of A-single (As), A-paired (Apr) and A-aligned (Aal), and Type B spermatogonia (Lie et al., 2009). In primates, this compartment consists of three subtypes (Fig. 2): (i) Type A dark (Ad), namely SSCs, renew themselves or generate differentiating Type A pale (Ap) spermatogonia; (ii) Type A pale (Ap) undergo mitosis to acquire Type B spermatogonia; and (iii) Type B spermatogonia produce PSC (Clermont, 1966; Di Persio et al., 2017). In terms of the causal relationship of Ap and Ad spermatogonia, other seemingly contradictory views are also proposed, such as that Ap spermatogonia are the SSCs, or that both of Ap and Ad cells are the SSCs simultaneously (Fouquet and Dadoune, 1986; Hermann et al., 2009). Regardless of the debate on Ap and Ad, what can be confirmed is that Type B spermatogonia will differentiate into PSC via mitosis. In this process, autophagy delicately exerts its bilateral effects.

During spermatidogenesis, each diploid PSC develops into four haploid round spermatids through meiosis, which occupies an absolute central position in the process. One PSC undergoes one round of DNA replication and cell division (Meiosis I) to produce two haploid secondary spermatocytes, which subsequently proceed through the second cell division (Meiosis II) to produce four equal haploid round spermatids (Wang et al., 2017a). Meanwhile, a series of distinctive cellular events occur, including programmed DNA double-strand break formation, homologous recombination, crossover formation and resolution (Sung and Klein, 2006).

The study on the role of autophagy in spermatidogenesis originates in 1986. Chemes (1986) proved that Sertoli cells phagocytized and digested meiotic spermatocyte residual bodies by autophagy in a testosterone-independent manner. Recent research showed that autophagy and apoptosis are synchronously provoked in the heat-treated mouse spermatocyte cell line, GC-2 cells. Dramatically, Atg7-mediated downregulation of autophagy reduced apoptosis in heat-induced GC-2 cells, indicating that autophagy and apoptosis act as partners to promote cell death (Zhang et al., 2012).

Apart from being pro-death, autophagy seems to undertake a dual mission in regulating chromatoid bodies (CBs) during spermatidogenesis. CBs are a typical cytoplasmic features of haploid round sperm cells, consisting of RNA and RNA-binding proteins, and are unique ribonucleoprotein (RNP) granules (Meikar et al., 2011). Despite appearing in late pachytene spermatocytes, the CB-like granules are immediately condensed into a single granule after meiosis and maintain their character throughout the differentiation of round spermatids (Kotaja and Sassone-Corsi, 2007). Due to the accumulation of plentiful PIWI-interacting RNA (piRNA) in the CBs of round spermatids, the CB is deemed to be responsible for piRNA-targeted RNA regulation (Gomes Fernandes et al., 2018). Remarkably, piRNA-directed cleavage of meiotic transcripts regulates spermatogenesis (Goh et al., 2015). Therefore, it is rational to suspect that CBs govern meiotic transcripts during spermatidogenesis.

Interestingly, Da Ros et al. (2017) discovered that the LC3B-interacting protein FYCO1, a novel CB component, initiates the intracellular transport of autophagic vesicles, which regulates the integrity of RNP granules in haploid male germ cells. Furthermore, both the agonists and antagonists of autophagy aggravate the cellular defects (fragmented CB) in the Fyco1 cKO germ cells, manifesting that autophagy devotes itself to two distinct events: the clearance of CB materials and the maintenance of CB homeostasis, synchronously (Da Ros et al., 2017). Consistent with the perception, autophagy can protect genomic stability by degrading retrotransposon RNA (Guo et al., 2014), and PIWI proteins/piRNA in the CB are the targets of degradation after autophagy activation (Siomi et al., 2011; Da Ros et al., 2017). Taken together, autophagy plays three roles during spermatidogenesis: (i) pro-death in spermatocyte, (ii) clearance of CB materials in round spermatids and (iii) maintenance of genomic stability in CBs by degrading PIWI proteins/piRNA in round spermatids.

The regulation of spermatidogenesis by autophagy was confirmed in Ppp1cc and Raptor KO mice, respectively. Ppp1cc2, as a substrate of the CMA regulator Hsc70 (Bonam et al., 2019; Itoh et al., 2019), was required for chromatin condensation and acrosome development (Forgione et al., 2010). Ppp1cc KO mice exhibited meiosis arrest during spermatidogenesis and spermiogenesis, presenting polyploid spermatids (Varmuza et al., 1999). Similarly, conditional ablation of mTORC1 component Raptor in the male germline caused meiotic arrest (Xiong et al., 2017). Therefore, the cross-talk between meiosis and autophagy-related proteins needs further exploration during spermatidogenesis.

Spermiogenesis refers to the final stage of spermatogenesis when round spermatids become elongated and then develop into spermatozoa through drastic morphological changes. Spermiogenesis involves the reshaping of the nucleus, rearrangement of mitochondria and development of flagellum and acrosome (Tanaka and Baba, 2005). These orchestrated physiological processes require a cellular homeostasis between degradation and recycling of cytoplasmic components. Recently, growing evidence indicates that autophagy, as a unique catabolic pathway (Ktistakis and Tooze, 2016), is involved in spermiogenesis. Autophagy-related proteins (LC3, ATG5, ATG16, BECN1, p62, mTOR, AMPKα 1/2 and PINK1) and their upstream regulators are functionally active in human spermatozoa, suggesting that autophagy may regulate sperm motility (Aparicio et al., 2016). Specifically, LC3 and ATG7 are increased dramatically from the round to elongated spermatids (Yang et al., 2017). Besides, autophagosomes in spermatozoa originate from the bilayer separation membrane of the chrysanthemum flower center, which is converted from endoplasmic reticulum (Yang et al., 2017), implying the regulation of spermiogenesis by autophagy.

The acrosome is also a unique lysosome-related membranous organelle (MO) in the anterior part of the sperm nucleus (Berruti and Paiardi, 2011). The acrosome carries hydrolytic enzymes to facilitate sperm penetrating the zona pellucida (Jin et al., 2011; Ozturk et al., 2017). For the first time, Hartree proposed the correlation between the acrosome and the lysosome in 1975 (Hartree, 1975). Controversially, Martínez-Menárguez et al. argued in 1996 that acrosomes were independent of endosomes or lysosomes due to the absence of late endosomal marker cation-dependent and non-dependent mannose 6-phosphate receptor in acrosomal vesicles and preantral vesicles (Martinez-Menarguez et al., 1996). Ten years later, both Berruti et al. (2010) and Hu et al. (2007) demonstrated that the acrosome was indeed a lysosome-related organelle (LRO). Furthermore, acrosome biogenesis, as a hinge of spermiogenesis (Kang-Decker et al., 2001), was linked closely with autophagy.

Generally, acrosome biogenesis is divided into four major phases: Golgi, cap, acrosome and maturation phases (Khawar et al., 2019). The first phase corresponds to the perspective that acrosome is derived from Golgi (Berruti and Paiardi, 2011). Research verifies that Golgi can be regulated by TBC1D20 (Haas et al., 2007), and the disruption of TBC1D20 results in testicular abnormalities in mice (Park et al., 2014), as well as in humans (Liegel et al., 2013). Thus, TBC1D20 may mediate the testicular function via regulating Golgi. TBC1D20, as a member of GAPs, interacts with GTP and subsequently, the ‘active’ GTP-bound RAB-GTPase is returned to the ‘inactive’ GDP-bound state (Frasa et al., 2012). Strikingly, Sidjanin et al. (2016) showed that TBC1D20 facilitates autophagy flux by its RAB1B GAP function, and regulates the formation of acrosome in mice, suggesting that TBC1D20 may regulate acrosome biogenesis via autophagy.

In parallel, Sirt1 regulates acrosome biogenesis by modulating autophagy flux during spermiogenesis in mice. Liu et al. found that the depletion of Sirt1 undermined spermiogenesis by stimulating autophagy in mice spermatids, including three successional cellular events: (i) the depletion of Sirt1 disrupted LC3 and ATG7 deacetylation, leading to the redistribution of LC3 from the nucleus to the cytoplasm; (ii) Golgi-derived vesicles failed to recruit LC3; and (iii) nucleus-associated acrosomal vesicles were unable to recruit Golgi-associated PDZ- and coiled-coil motif-containing protein (GOPC) and PICK1 (Liu et al., 2017a). In addition, for this comprehensive research, further studies proved that ATG7 not only partially targeted GOPC to control acrosome biogenesis (Wang et al., 2014), but is also essential for activation of LC3, which initiated autophagy (Tanida et al., 2012). Even, ATG7 was required for spermatozoa flagella biogenesis and cytoplasm removal during mice spermiogenesis (Shang et al., 2016). In terms of the mechanism, research elucidated that the autophagy-lysosome pathway regulates spermatid differentiation by degrading negative cytoskeleton regulator PDZ and LIM domain 1 (PDLIM1) to facilitate cytoskeleton organization (Shang et al., 2016). Coincident with this assertion, in moss, ATG5- and ATG7-mediated autophagy promote the flagellated motile sperm differentiation and cytoplasmic elimination (Sanchez-Vera et al., 2017). Therefore, whether in mice or moss, autophagy is indispensable for spermatid differentiation, especially in acrosome biogenesis and flagella biogenesis.

When it comes to flagella biogenesis, it is important to mention intraflagellar transport 20 (IFT20). During mouse spermiogenesis, IFT20, as a Golgi transport protein, interacts with the sperm flagellar 2 (SPEF2) to regulate the development of the sperm flagella (Lehti et al., 2017). SPEF2 is critical for microtubule-mediated transport in axonemal CP assembly and sperm head reshaping in mouse (Lehti et al., 2017). Consistently, in humans, homozygous mutations in SPEF2 lead to multiple morphological abnormalities of the sperm flagella and male infertility (Liu et al., 2020). These solid data support that IFT20 and SPEF2 are required for sperm tail formation and head reshaping during spermiogenesis. Although LC3 and ubiquitin are well-balanced in the testis of the Ift20 mutant mice (Zhang et al., 2016), it is not enough to deny the correlation between IFT20 and autophagy because of the scarcity of autophagy flow evaluation in this limited study. Remarkably, IFT20 contributes to autophagosome formation by delivering ATG16L (Pampliega et al., 2013) and to lysosome biogenesis by regulating the post-Golgi transport of acid hydrolases (Finetti et al., 2020).

Hence, autophagy participates in acrosome biogenesis, flagella assembly, shaping of the head and the removal of elongating spermatid cytoplasm during spermiogenesis (Fig. 3).

Spermiation is the release of elongated spermatids from Sertoli cells into the seminiferous tubule lumen prior to the epididymis (Cheng and Mruk, 2010). This process is guided by a testis-specific, actin-based anchoring junction, apical ES (aES) in the Sertoli cell-spermatid interface (Mruk et al., 2008). Apart for spermatids movement and release, aES also contributes to shaping the spermatid head (Toyama et al., 2003; Mruk and Cheng, 2004). To achieve this goal, aES needs to undergo tightly and timely managed restructuring (Berruti and Paiardi, 2014).

Currently, autophagy has been demonstrated to be required for ES assembly. ES encompasses two parts: (i) aES, at the Sertoli cell-spermatids interface and (ii) basal ES (bES), at the Sertoli–Sertoli cell interface (Cheng and Mruk, 2010, 2012). Sertoli cell-specific KO of Atg5 or Atg7 results in spermatozoa with malformed heads and low motility, affecting the fertility of male mice (Liu et al., 2016). Meanwhile, defective autophagy in Sertoli cells perturbs the degradation of PDLIM1, which is vital for cytoskeleton assembly (Liu et al., 2016). Consequently, through PDLIM1, autophagy not only affects bES to regulate BTB, but also mediates aES to govern spermatids movement and release during spermiation (Fig. 4a).

For fertilization, elongated spermatids have to undergo maturation. Sperm maturation contains three highly orchestrated processes: (i) obtaining motility and fertility in the epididymis, (ii) capacitation in the female reproductive tract and (iii) activation upon the approach of ovary and sperm.

Firstly, sperm flows into the rete testis in Sertoli cell-secreted fluid. With the action of their flagella and smooth muscle contraction, sperm then enter the caput epididymis. In the epididymis, sperm maturation involves both changing flagellum beat and acquiring the characteristics necessary for effective contact with the oocyte (Sullivan and Mieusset, 2016). The sperm cannot attain fertilization ability until they reach the proximal end of the epididymis. The motility and fertility capacities of sperm are completed in the cauda epididymis.

Secondly, capacitation, as the penultimate step of mammalian sperm maturation, occurs in the female reproductive tract. When semen is ejaculated into a female vagina, the sperm leaves the semen and enters the uterine cavity and fallopian tube via the cervical canal. Then female genital tract secrets alpha and beta-amylase, which degrades the acrosome surface glycoprotein. Meanwhile, the sperm launches a series of orchestrated biochemical reactions, including phosphorylation, alkalinization and hyperpolarization, dependent on CFTR and PKA, preparing for penetrating out layer of the oocyte (Puga Molina et al., 2017). Additionally, the increase in Ca2+ influx allows the sperm tail greater motility (Ren et al., 2001). Finally, capacitation renders sperms competent to fertilize an oocyte.

Notably, a dramatic change of ROS occurs in sperm during capacitation (Fig. 4b). Despite the detrimental effect of high concentrations of ROS, low concentrations of ROS are beneficial for sperm fertilization. It is pivotal to balance the homeostasis between ROS and autophagy. ROS regulates autophagy by transcriptional and post-transcriptional regulation, including ROS-FOXO3-LC3/BNIP3-autophagy, ROS-NRF2-P62-autophagy, ROS-HIF1-BNIP3/NIX-autophagy and ROS-TIGAR-autophagy (Li et al., 2015). In turn, autophagy can regulate ROS levels through CMA, the mitotic pathway and the P62 delivery pathway (Li et al., 2015). Also, autophagy can affect sperm motility by inhibiting the fusion of autophagosomes and lysosomes through the PI3K/AKT-dependent pathway in leucine-treated zebrafish (Zhang et al., 2017).

Given that cholesterol efflux alters lipid raft stability and distribution during capacitation of boar spermatozoa (Shadan et al., 2004), cholesterol efflux may have close relationship with capacitation. Recent studies have shown a strong link between the PKA pathway and sperm capacitation. PKA can stimulate protein phosphorylation and capacitation or acrosomal exocytosis by activating tyrosine kinase and/or inhibiting protein phosphatase (Puga Molina et al., 2017).

Nonetheless, so far, there is no direct evidence that autophagy participates in the process of capacitation (Fig. 4b). It may be a good potential target for exploring the molecular mechanism in capacitation dysfunction-induced male infertility.

Fertilization is a fundamental process in reproduction (Ohto et al., 2016). In this process, the oocyte and sperm achieve mutual recognition, and fuse into a zygote, which then develops into a new individual, allowing for the continuity of a species (Okabe, 2013).

Classically, the acrosome reaction is one of the most critical steps in fertilization. This may be due to the fact that acrosome reaction allows the sperm to release acrosomal enzymes, dissolve the radiation corona and penetrate the zona pellucida (Tosti and Menezo, 2016). Recent research shows that NAADP and the two-pore channel (TPC) protein 1 participate in the acrosome reaction by Ca2+ release in mammalian spermatozoa (Arndt et al., 2014). The regulation of TPCs by autophagy is involved in multiple cells, such as neural cells (Pereira et al., 2017), cardiomyocytes (Garcia-Rua et al., 2016) and cancer cells (Sun and Yue, 2018). TPC2 mediates mTORC1 (Chang et al., 2020), which is the main gateway to autophagy (Rabanal-Ruiz et al., 2017). In turn, mTORC1 controls lysosomal Ca2+ release by TPC2 (Ogunbayo et al., 2018). Despite the interaction of TPC2 and mTORC1, there is no direct evidence that autophagy regulates the acrosome reaction through TPCs. This is waiting to be clarified.

During fertilization, most animals only inherit the mitochondrial genome from the maternal parent (Ankel-Simons and Cummins, 1996). The paternal mitochondria are eliminated in the embryo, although the sperm provides DNA, centrioles, some cytoplasm and organelles for the offspring (Kaneda et al., 1995). The underlying mechanism of PME during fertilization remains unclear.

Fertilization-triggered autophagy can degrade paternal mitochondria (Sato and Sato, 2011; Zhou et al., 2011), and post-fertilization autophagy of sperm organelles prevents paternal mitochondrial DNA (mtDNA) transmission in Caenorhabditiselegans (Al Rawi et al., 2011). Zhou et al. (2016) found that CPS-6 migrated from the parent mitochondrial intermembrane to the matrix after fertilization for the degradation of mtDNA. Loss of cps-6 disrupts autophagy, PME and embryogenesis, revealing that CPS-6 regulates PME by interacting with maternal autophagy and proteasome machinery upon fertilization in C. elegans (Zhou et al., 2016). Tsukamoto et al. discovered that autophagy-deficient sperm (sperm-specific KO ATG5) and oocytes without ATG5 in mice could undergo regular fertilization and fuse into a zygote, but failed to develop into four-cell and eight-cell stages. In contrast wild-type sperm and autophagy-deficient oocytes were available to develop normally into embryos, suggesting that fertilization-triggered autophagy was essential for the development of mammalian early embryos (Tsukamoto et al., 2008).

In mice embryos, this autophagic degradation process is dependent on the E3 ubiquitin ligases PARKIN and MUL1, mitochondrial outer membrane protein FIS1, autophagy adaptor p62 and PINK1 kinase (Rojansky et al., 2016). Owing to mitochondrial depolarization stabilizing PINK1, both PINK1 and PARKIN are recruited to damaged mitochondria, and latent PARKIN is activated for mitophagy in Parkinson’s disease (Kawajiri et al., 2010; Matsuda et al., 2010). PINK1/PARKIN-mediate mitophagy through voltage-dependent anion channel 1 and p62/SQSTM1 (Geisler et al., 2010). Although in Drosophila, autophagy and endocytosis regulate PME by a PARKIN-independent pathway (Politi et al., 2014), PINK1/PARKIN-mediated mitophagy may be essential for PME in mammals.

Another study demonstrated that fertilization induces ubiquitination and recruitment of LGG-1/LGG-2 (the two LC3 homologs in worms) around the flagellum mid-piece; after fertilization, paternal mitochondria and MOs are engulfed in autophagosomes and degraded during the first zygote divisions (Al Rawi et al., 2012). Furthermore, the depletion of lgg-1 led to the death of 95% animals at or before the L1 larval stage, implying that disruption of PME may undermine embryogenesis, although there were other zygotic defects due to the loss of LGG-1 (Levine and Elazar, 2011; Sato and Sato, 2011). Djeddi et al. (2015) verified that sperm-inherited organelle clearance in C. elegans relied on LC3-dependent autophagosomes. The above results suggest that both proteasomal and autophagic degradation participate in PME and MOs clearance upon fertilization in C. elegans.

In parallel, Hajjar et al. confirmed that sperm-derived mitochondria and MOs, which cluster together, are simultaneously decreased during fertilization in C. elegans. K48-linked ubiquitin chains (K48 chains), which degrade proteasomes, arise on MOs and vanish very quickly; meanwhile, K63-linked ubiquitin chains (K63 chains), which recruit autophagosomal markers (LGG-1/2) to MOs, emerge on MOs early and remain throughout the first several cell divisions (Hajjar et al., 2014). These findings suggest that K63 chain-mediated autophagy modulates PME and MOs clearance, and K48 chains only participate in MOs clearance. However, the interaction mechanism between K63 chains and K48 chains in PME or MOs clearance remains an enigma.

Song et al. (2016) showed that sperm mitophagy in pig and rhesus monkey relied on p62-dependent autophagy, valosin-containing protein-mediated dislocation and recruitment of ubiquitin to the 26S proteasome. They concluded that autophagy and the UPS contributed to sperm mitophagy after mammalian fertilization (Song et al., 2016).

Interestingly, Al Rawi et al. (2011) and Sato and Sato (2011) claim that ubiquitin was only related to MOs but not paternal mitochondria, despite the fact that autophagosomes indeed degrade both MOs and paternal mitochondria. In view of this, Levine and Elazar put forward two potential roles of paternal MOs by suggested that (i) MOs provided membranes for autophagosome formation and (ii) MOs launched autophagy in the process of PME (Levine and Elazar, 2011).

However, this theory was challenged by Luo et al. (2013). They proposed that autophagy was not involved in PME after fertilization by using two transgenic mice, separately labeling mitochondria and LC3, although some autophagy-related proteins, such as p62/SQSTM1 and LC3 indeed localized near the sperm mitochondria before the two-cell stage. Additionally, most motile sperm eliminated their mtDNA before fertilization (Luo and Sun, 2013). Thus, in mice, maternal inheritance of mtDNA may be a passive process as a result of pre-fertilization sperm mtDNA elimination and uneven mitochondrial distribution in embryos (Luo and Sun, 2013). Whether this is the case awaits further research.

To sum up, autophagy may regulate the acrosome reaction, PME and MOs clearance during fertilization (Fig. 5).

M.W. and Y.Z.Z. designed the study and edited the final text. M.W. and P.S. collected the data from publications, developed the database and wrote the manuscript. M.W. prepared the figures. M.W. and L.Z. prepared the tables. M.W., Y.Z.Z., L.M. and M.Z. contributed to the manuscript revision and critical discussion.

This study was supported by Ministry of Science and Technology of the People’s Republic of China (National Key R&D Program of China, No. 2020YFA0803900) and Wuhan University (Wuhan University Medical Development Plan, No. TFJC2018001).

None declared.

The data underlying this article are available in the article.