Role of Posttranslational Protein Modifications in Epididymal Sperm Maturation and Extracellular Quality Control

Phosphorylation is the addition of a phosphate group to serine, threonine, tyrosine (O-linked) or histidine (N-linked) amino acids. This involves the transfer of phosphate groups from ATP or GTP to the protein by the action of protein kinases while dephosphorylation utilizes protein phosphatases. In general, changes in protein phosphorylation are associated with the activation or inactivation of protein function during cell signaling cascades. Both protein kinases and phosphatases are present in epididymal spermatozoa. Furthermore, changes in the phosphorylation status of proteins have been observed in spermatozoa during epididymal transit with resulting alterations in protein localization and/or functions suggesting that this PTM plays an important role during sperm maturation.

Although activation of a cAMP/PKA pathway with an upregulation of protein tyrosine phosphorylation is a hallmark of the sperm capacitation process and necessary for the acrosome reaction, less is known about specific phosphorylation events that occur in spermatozoa during epididymal transit and how these events affect maturation. Early studies in both rat and boar suggested maturational changes in sperm protein tyrosine phosphorylation with caput spermatozoa showing a more dispersed localization of phosphotyrosine labeling over the acrosome while cauda spermatozoa exhibited a pattern of protein phosphorylation confined to the posterior aspect of the acrosome, suggesting that dephosphorylation of some sperm proteins was associated with epididymal sperm maturation [5, 6]. Increased tyrosine phosphorylation labeling was then observed over the whole boar sperm head after capacitation suggesting a reversal of some of the maturation-associated modifications or activation of new populations of proteins [6]. The marked decrease in tyrosine phosphorylation in rat and boar spermatozoa during epididymal transit is in marked contrast to that observed in the mouse where overall protein tyrosine phosphorylation, as determined by Western blot analysis, was relatively low in both caput and cauda spermatozoa [7]. Individual sperm proteins that have been shown to be tyrosine phosphorylated in the epididymis include the ATP-binding cassette membrane transporter G2 (ABCG2). ABCG2 is present in the plasma membrane overlying the sperm acrosome and is proposed to play a role in the translocation of cholesterol across the plasma membrane in epididymal spermatozoa [8]. The tyrosine phosphorylation of ABCG2 has been shown to be necessary for its function within epididymal spermatozoa while dephosphorylation is believed to result in inactivation of ABCG2 function.

In addition to tyrosine phosphorylation, phosphorylation of sperm proteins at serine and threonine residues is also associated with epididymal maturation. The beta subunit of F1 ATPase in spermatozoa was shown to be serine phosphorylated during epididymal transit perhaps as a means to stabilize the protein [9]. This enzyme has been shown in a lymphoma cell line to undergo cAMP-dependent phosphorylation suggesting that the F1 ATPase may be a target of a PKA signaling pathway (cAMP-dependent kinase activity) that becomes active in spermatozoa as they mature in the epididymis [10, 11, 12]. Functionally, phosphorylation of the mitochondrial F1 ATPase may then increase ATP production, facilitating sperm motility maturation.

The use of mass spectroscopy to identify phosphoproteins in spermatozoa has revealed the breadth this particular modification plays in sperm maturation and has identified a number of sperm associated proteins that are differentially phosphorylated depending on the epididymal region from which they are isolated [13]. Using titanium dioxide to enrich for phosphopeptides, Baker et al. [14] identified several proteins including the fertilization molecule IZUMO, testis lipid binding protein, A-kinase anchoring protein 4, outer dense fiber I, sodium bicarbonate cotransporter, dynein intermediate chain I, and several others that showed changes in phosphorylation between epididymal regions. In this study, the majority of phosphopeptide changes occurred in spermatozoa between the corpus and cauda regions where significant maturational changes in motility are observed. Furthermore, of the proteins that showed maturational changes in phosphorylation status, many of these had previously described roles in sperm capacitation. It may be that in the more distal epididymal regions, important signaling complexes are becoming activated in spermatozoa in preparation for downstream fertilization events. Together, these studies show that sperm proteins with proposed functions not only in cell signaling but in other cellular processes as well are modified by phosphorylation during epididymal transit as an integral part of the maturation process.

The phosphorylation of proteins is mediated by the activity of kinases that are present in spermatozoa including those of the MAP kinase pathway [15, 16, 17], CaMKII alpha [18], FYN kinase [19], JAK/STAT [20] and many others. Most of these kinases have been studied from the standpoint of their roles in sperm capacitation and fertilization, and will not be addressed here. However, several other sperm-associated kinases have been shown to undergo modifications themselves during epididymal transit and others have been implicated in specific maturation events. For example, the testis specific serine kinase I undergoes phosphorylation during epididymal transit, perhaps as an activation step to ultimately allow additional downstream phosphorylation of target proteins [14].

Other sperm kinases include glycogen synthase kinase-3 alpha (GSK3-apha), a signaling kinase that becomes serine phosphorylated in spermatozoa during epididymal transit. GSK-3 was shown to be more active in caput than in cauda spermatozoa, suggesting that phosphorylation of GSK-3 is associated with a loss of its function. Stimulation of sperm motility by the phosphodiesterase inhibitor isobutyl-methyl-xanthine also caused an increase in GSK-3 serine phosphorylation, suggesting a role for this kinase in the regulation of sperm motility [21]. Upstream signaling molecules protein kinase B and phosphoinositide 3-kinase, that are involved in GSK-3 phosphorylation, are also present in spermatozoa [21].

The tyrosine kinase cSrc has been proposed as the kinase responsible for protein tyrosine phosphorylation during sperm capacitation since this kinase is activated during capacitation and specific inhibitors blocked sperm capacitation and its associated protein tyrosine phosphorylation [22, 23]. However, despite exhibiting reduced motility and fertility in vitro, protein tyrosine phosphorylation still increased during capacitation of spermatozoa from cSrc null mice suggesting that cSrc does not play a direct role in this process [24]. Recently, it was shown that cSrc is secreted by the epididymis and is acquired by spermatozoa during epididymal transit [24]. Together these studies suggest an important role for cSrc in epididymal sperm maturation possibly as a mediator of critical signaling events that contribute to the acquisition of sperm motility.

Protein kinase A RI, RIα, and RIIα subunits are active in regulating tyrosine phosphorylation in both caput and cauda epididymal spermatozoa yet caput sperm exhibit lower and different patterns of tyrosine phosphorylation when exposed to dbcAMP and PTX, suggesting other maturational changes are also involved [5, 7, 25]. RIIα itself also appears to undergo region-dependent serine phosphorylation [14].

The removal of phosphate groups by the action of phosphatases has also been implicated in epididymal sperm maturation. The serine–threonine protein phosphatase I gamma 2 (PP1γ2) plays an important role in the development and regulation of sperm motility. Specifically, high PP1 activity is present in immotile caput spermatozoa with lower activity in motile cauda spermatozoa. Furthermore, inhibition of PP1 activity by okadaic acid and calyculin A stimulated motility in the normally quiescent caput spermatozoa, suggesting that the potential for motility is already present in caput epididymal spermatozoa and that the high PP1 activity keeps this motility in check [26]. PP1 has also been proposed to regulate sperm motility by the suppression of the full activation of PKA [27]. A population of phosphorylated PP1γ2 was shown to localize to the sperm head, suggesting additional roles for this phosphatase in fertilization [28].

Glycosylation refers to the enzymatic process that attaches glycans/sugars to proteins, lipids, or other organic molecules and is distinct from glycation which is the nonenzymatic attachment of sugars. N-linked glycosylation involves the transfer of sugars to a nitrogen of asparagine or arginine residues and occurs in the lumen of the endoplasmic reticulum. O-linked glycosylation is the transfer of sugars to the hydroxy oxygen of serine, threonine, tyrosine, hydroxylysine, or hydroxyproline residues and occurs in the Golgi. Cellular processes associated with glycosylation include modifications of the cell surface and host defense. The glycan modifying enzymes glycosyltransferases, which add sugars to proteins, and glycosidases, which cleave sugar residues from glycoproteins, are present in the epididymal fluid. Both N-linked and O-linked glycosylation has been observed in proteins present in the epididymal fluid and associated with spermatozoa; thus glycosylation appears to be a significant PTM that modifies epididymal sperm surface proteins affecting functions important for sperm maturation. Indeed, using a variety of different approaches including treatment with glycosidases or by lectin binding, it appears that the sperm surface, like other cells, is dominated by the presence of carbohydrates. These carbohydrates linked to protein, lipids or other molecules create what is referred to as the glycocalyx on the sperm surface and which plays an integral role in fertilization [29].

Sperm plasma membrane glycoproteins that likely contribute to the glycocalyx are extensively modified during epididymal transit as shown by the profound changes in the binding of the lectin PNA to spermatozoa from different epididymal regions. Rat cauda epididymal spermatozoa showed a loss of PNA binding compared to that in caput spermatozoa which was thought to reflect either the masking of galactosyl residues by the addition of other sugars or the cleavage of the galactosyl residues by β-d-galactosidase [30]. In the rhesus monkey, regional modifications of epididymal glycoproteins (O- and N-linked) including those associated with spermatozoa was observed using biotinylated lectins [31, 32]. An analysis of the boar sperm glycocalyx was recently performed by lectin binding and increases in sperm galactose, glucose/mannose, and N-acetylglucosamine were observed in spermatozoa from distal epididymal regions [33]. Similar studies have also been performed in the cat [34]. Examples of specific proteins that undergo changes in glycosylation during epididymal transit include SPAM 1 (PH-20). The hyaluronidase activity of SPAM1 in caput spermatozoa is much lower than in cauda epididymal spermatozoa. Furthermore, the increase in enzymatic activity in spermatozoa during epididymal transit was correlated with a reduction in the molecular weight of SPAM1. It was proposed that the deglycosylation of the N-linked glycosylated SPAM1 may be important for the activation of its activity [35]. The transmembrane glycoprotein basigin belongs to the immunoglobin superfamily and is localized to the sperm tail. During epididymal transit, basigin localization changes from the principal piece to the midpiece, with a concurrent reduction in molecular mass that is thought to represent deglycosylation [36]. In most species, there is a general increase in negative surface charge during epididymal sperm maturation [37, 38]. This is likely due to the acquisition of sialic acid in spermatozoa during epididymal transit since increased binding of the lectin WGA, specific for sialic acid and N-acetylglucosamine, is observed [39]. Two sialoproteins, proteins D and E, are secreted by the epididymal epithelium and associated with spermatozoa [40]. Other glycoproteins that become associated with the sperm surface during epididymal transit that may contribute to the sperm glycocalyx include CD59, fertilin, HE2, HE4, and HE5 (reviewed in [29]).

Of the several glycosyltransferase and glycosidase activities that have been examined in the rat epididymis, most are in the luminal fluid fraction with only a small percentage of enzyme activities associated with spermatozoa [41]. Namely, galactosyltransferase, glucosaminyltransferase, fucosyltransferase, and sialyltransferases were primarily detected in the supernatant after low and high speed centrifugation to pellet spermatozoa and insoluble components in the luminal fluid. Similarly, the glycosidases β-d-galactosidase, β-d-glucuronidase, α-d-mannosidase, and β-N-acetylglucosaminidase were present in the supernatant fractions [41]. While glycosyltransferases exhibit pH optima at neutral pH, compatible with the epididymal luminal pH of 6.6–6.8, most glycosidases are optimally active at acidic pH (pH 3–5). However, examination of the luminal fluid β-d-galactosidase showed optimal activity at pH 6.8, suggesting that the epididymal enzymes may have unique properties allowing them to function within the epididymal lumen [42]. While galactosyltransferase and glucosaminyltransferase exhibited similar activities in spermatozoa isolated from the different epididymal regions, both fucosyltransferase and sialyltransferase activities showed regional differences with highest activity present in caput spermatozoa and very low activity in cauda spermatozoa [43]. These results are consistent with the idea that addition of sialic acid to spermatozoa occurs during epididymal transit as well as that fucosylation appears to also contribute to region-dependent changes in glycosylation [43]. Although the functional significance for the addition of sialic acid to the sperm surface has not been established, studies in several species including the mouse and bull suggest that fucose residues on sperm surface glycoproteins play a role during sperm–oocyte fusion [44, 45]. Addition of fucose during sperm–oocyte fusion assays or exposure of capacitated spermatozoa to fucosidase, which removes fucose, inhibited sperm–oocyte fusion during fertilization [46].

Using a proteomic approach, several other enzymes involved in glycoprotein metabolism have been identified in the epididymal fluid from the bull including β-mannosidase, β-hexosaminidase, and sialidase [47].

In addition to changes in phosphorylation and glycosylation, a critical component of the sperm maturation process is the remodeling of the sperm surface by proteolytic processing. Proteolytic processing involves the breaking of peptide bonds between amino acids in proteins which is carried out by peptidases and proteases. These proteolytic processing events are associated with changes in protein localization in spermatozoa and/or activation of protein function and thus protein processing is thought to contribute to the formation of sperm cell surface domains [48].

Several ADAM (A Disintegrin and Metalloprotease) family members with proposed roles in fertilization are present in the epididymal lumen and/or associated with epididymal spermatozoa, and the majority of these have been shown to undergo protein processing in the epididymis [49, 50]. Specifically, ADAMs are transmembrane domain proteins composed of a signal sequence, pro- and metalloprotease, and disintegrin domains, a cysteine-rich region, transmembrane domain, and cytoplasmic tail. Both ADAM1B (fertilin α) and ADAM2 (fertilin β) have their pro- and metalloprotease domains removed with the disintegrin domain becoming the N-terminus on mature spermatozoa. The disintegrin domain has been shown to be important for sperm–oocyte membrane interactions during fertilization [48]. In the mouse, ADAM2 undergoes additional modifications in its C-terminus [51]. Following protein processing, ADAM2 is redistributed on the sperm surface and becomes localized to the posterior sperm head [52]. ADAM1B–ADAM2 and ADAM2–ADAM3 (cyritestin), which is also processed during epididymal transit, have been shown to form complexes in spermatozoa [53, 54]. ADAM3 is thought to promote sperm migration in the female reproductive tract [55]. Trimeric complexes between ADAM2–ADAM3–ADAM4, ADAM2–ADAM3–ADAM5, and ADAM2–ADAM3–ADAM6 have also been described [56]. Interestingly, ADAM3 and ADAM6 are lost from spermatozoa in mice lacking tyrosylprotein sulfotransferase-2 activity, which catalyzes the post-translational modification of tyrosine O-sulfation, suggesting additional PTM of ADAM proteins contribute to their functionality or quality control [57]. Tyrosine O-sulfation has been shown in other cell systems to play a role in protein complex formation [58].

Although ADAM15 and ADAM24 (testase1) are also proteolytically processed during epididymal transit, they only lose their prodomains with the active metalloprotease domain remaining. ADAM15 later undergoes additional protein processing during the acrosome reaction, ultimately exposing the disintegrin domain [59, 60]. While the function of ADAM15 is not clear, roles in sperm–egg interactions have been proposed. ADAM24 has been proposed to function as a protease during epididymal sperm maturation [61]. ADAM24 has a cytoplasmic PKC phosphorylation site and can be phosphorylated in vitro by PKC [61]. It was proposed that protein processing in the epididymis may be the initial activation step while later phosphorylation by PKC during capacitation/acrosome reaction is the final step for activation of ADAM24 protease function during fertilization [61].

Other sperm proteins that undergo protein processing during epididymal transit include the proteins D/E [62], guanylyl cyclase-G [63], the adaptor protein CASK [64], CE9 [65], and α-d-mannosidase [66]. For several of these proteins, protein processing is associated with changes in cellular localization. CE9 is redistributed from the posterior to the anterior tail of the rat spermatozoa following proteolytic cleavage [65]. In addition to proposed changes in glycosylation status, protein processing has also been implicated in the changes in molecular mass of SPAM1 (PH-20) and its localization on spermatozoa. The processing of SPAM1 from a 74 kDa form to a 67 kDa form correlates with its change in localization from a broad distribution on the head of caput spermatozoa to a more defined region in cauda spermatozoa [67]. This change in localization was proposed to be due to a trypsin like activity in the luminal fluid since treatment of caput spermatozoa with trypsin in vitro changed the SPAM1 localization to that characteristic of cauda spermatozoa [48].

For many of the processing events described above, the proteases involved have yet to be described. However, angiotensin-converting enzyme (ACE), a key regulator of blood pressure by its ability to cleave small peptides resulting in a change in their biological activities, has also been shown to exhibit GPIase activity by its ability to release GPI anchored proteins from the cell surface including PH-20 and TESP5 from spermatozoa [68]. ACE has also been shown to regulate the movement of ADAM3 in the sperm membrane [55]. ACE itself is shed from the sperm surface by the action of an unknown serine protease activity that appears to be in the epididymal fluid [69].

The protease(s) responsible for ADAM protein processing was proposed to be a serine protease [70], in particular, a proprotein convertase, since predicted cleavage sites for this family of serine proteases are present in ADAM1 [71]. Furthermore, proprotein convertases have been implicated in the processing of several other ADAMs [72]. ADAM15 also possesses a proprotein convertase cleavage site [60]. Several proprotein convertases are present in the epididymal fluid including furin [73] [Cornwall, unpublished observations]. Furthermore, mice lacking proprotein convertase 4, a testis-specific convertase, exhibit fertility defects including reduced zona pellucid binding suggesting a role for PCSK4 in fertilization [74]. Spermatozoa from PSCK4 null mice showed increased ADAM2 processing possibly reflecting the upregulation of PCSK7 activity in response to the loss of PCSK4, supporting a role for proprotein convertases in ADAM protein processing [75]. Furthermore, exposure of spermatozoa to a peptide inhibitor of PCSK4 resulted in decreased ADAM2 processing [76]. Finally, mice lacking PCSK4 showed reduced levels of sperm ACRBP protein processing in the epididymis, suggesting that several sperm associated proteins may be PCSK4 substrates [77]. Recent evidence suggests that in addition to PCSK4 present in spermatozoa, a population of PCSK4 is present in the epididymal fluid and may also contribute to sperm protein processing [78] [Cornwall, unpublished observations].

The serine protease PRSS21 (testin) is involved in epididymal sperm maturation since PRSS21 deficient mice show decreased motility, angulated and curled tails, and an increased susceptibility to decapitation [79]. Proteomics has also revealed a large number of proteases that are present within the epididymal luminal fluid. These include cathepsins A, D, H, L and S, dipeptidyl peptidase III, matrix metalloproteinase 2, and serine carboxypeptidase I [47]. The specific roles these proteases play in epididymal function have yet to be established.

In parallel with the number of proteases that have been identified in the epididymis that contribute to sperm protein processing and activation, many protease inhibitors have also been identified which likely control these proteolytic events. Several members of the Kazal-type serine protease inhibitor (SPINK) family such as SPINK1, SPINK2, SPINK8, SPINK10, and SPINK12 are highly expressed in the mouse epididymis [80]. These studies also discovered a new WAP 4 disulfide core domain protease inhibitor WFDC10. All of the protease inhibitors showed region-specific localizations and their expression was controlled by testis-specific factors suggesting a role for the inhibitors in the regulation of sperm maturation. Recently, SPINK13 was also identified in the epididymis [81]. SPINK13 was secreted into the epididymal lumen where it associated with the acrosomal region of maturing spermatozoa. RNAi to knockdown SPINK13 resulted in spermatozoa with an accelerated acrosome reaction and fertility defects in vitro and in vivo. These studies suggest that the protease inhibitory activity of SPINK13 may be necessary for preventing premature acrosome reaction [81]. In addition to the SPINK family, other families of protease inhibitors have been found to show epididymis-specific expression. In particular, on human chromosome 20 and mouse chromosome 2 is a cluster of genes encoding proteins with both Kunitz-type and whey acidic protein four disulfide core (WFDC) domains as well as those that possess only WFDC or only Kunitz-type domains (SPINT3, SPINT4, SPINT5) [82, 83]. While most of the protease inhibitors are predominantly expressed in the epididymis, their functional roles have yet to be established. However, roles in innate immunity have been proposed since the Kunitz and WFDC domains have been shown to inhibit pro-inflammatory proteases such as elastin [84]. Human epididymal protein 4 (HE4), is one of the WFDC proteins that comprise the protease inhibitor locus and though its function in the epididymis is unknown, it has become a marker for the diagnosis of ovarian cancer [85].

In addition to the Kunitz–WFDC protease inhibitor locus, a second cluster of unrelated protease inhibitors is also localized to human chromosome 20 and mouse chromosome 2. The cystatin family 2 of the cystatin superfamily of cysteine protease inhibitors forms a cluster of which only cystatin C is a prototypical cystatin with potent inhibitory activities against papain-like cysteine proteases. The other cystatin family members present at this locus compose the CRES (cystatin-related epididymal spermatogenic) subgroup of reproductive specific cystatins [86]. Although structurally resembling cystatins, the eight characterized CRES subgroup members lack consensus sites for cysteine protease inhibition, suggesting distinctive functions. Indeed CRES, the defining member of the CRES subgroup, did not inhibit cysteine proteases in vitro but rather inhibited the serine proteases prohormone convertase 2 (PCSK2) and prohormone convertase 4 (PCSK4), suggesting CRES was a cross-class inhibitor with a role in the regulation of protein processing [87, 78]. As described above, PCSK4 has been implicated in the processing of several sperm associated proteins including ADAM2 and ACRBP and therefore, CRES may function within the epididymal lumen to regulate this proteolytic activity. In addition to CRES, CRES2, CRES3 and cystatin E2 are also expressed in the epididymis and are present in the epididymal lumen. However, whether they also function as cross class inhibitors or have completely lost inhibitor activity has yet to be determined.

Ubiquitination involves the modification of target proteins through the covalent attachment of ubiquitin via an isopeptide bond with target lysine residues of a substrate protein. This process requires the presence and activity of a set of ubiquitin activating and conjugating enzymes. In general, ubiquitination of proteins marks them for degradation by the 26S proteasome. While typically ubiquitination occurs intracellularly, studies have now established that the ubiquitin–proteasome system is also functional extracellularly [88, 89]. Specifically, defective spermatozoa have been shown to become ubiquitinated within the epididymal lumen, suggesting a mechanism for quality control [88]. The ubiquitin–proteasome pathway may also play an important role in zona pellucid penetration by acrosome reacted spermatozoa [90]. Further discussion on the role of ubiquitination in reproduction is presented in other chapters.

Methylation is the addition of a methyl group to proteins usually at a lysine or arginine residue. Typically, methylation is associated with the cellular process of development and differentiation. A functional methylation system includes a methylating enzyme such as protein carboxyl methylase (PCM), a demethylating enzyme protein such as methylesterase (PME), and methyl acceptor proteins (MAPs). This system reversibly modifies by methylation the carboxyl groups of proteins affecting their charge, structure, and function. PCM and MAPs are present in mature spermatozoa and the MAP/PCM ratio increases by 20-fold as rat germ cells differentiate into mature and motile cauda epididymal spermatozoa. PME has also been found in spermatozoa with the majority of activity associated with the detergent insoluble component of the sperm flagella which is consistent with a role for protein carboxyl methylation in sperm motility maturation [91, 92]. Subsequent to these early studies, little work has been done to examine the role of protein methylation in epididymal sperm maturation.

Disulfide bonds are formed between the sulfur atoms of pairs of cysteine residues within or across proteins and can affect protein structure and function. Spermatozoa from the caput epididymal region are rich in sulfhydryls and lack disulfides while cauda epididymal spermatozoa are rich in disulfides, suggesting that sperm disulfide bond formation is part of the sperm maturation process in the epididymis [93, 94]. Within spermatozoa, disulfide bond formation occurs in both the head, particularly in nuclear protamines as part of sperm chromatin condensation, as well as within the sperm tail [95]. The sulfhydryl rich organelles in the sperm tail that undergo oxidation to disulfides include the outer dense fibers, including the protein outer dense fiber 1, connecting piece, outer mitochondrial membranes, and the fibrous sheath [96]. Disulfide bond formation is thought to be important for sperm motility maturation since incubation of spermatozoa with a sulfhydryl-specific membrane impermeant dye inhibited the motility of goat spermatozoa [97]. Furthermore, induction of motility in hamster caput epididymal spermatozoa resulted in angulated spermatozoa that were prevented by the addition of the sulfhydryl oxidant diamide [98]. Together, these studies suggest that sulfhydryl oxidation is important for sperm tail structure stabilization and contributes to normal wave patterns of sperm motility. Sperm disulfides are also important for fertility since treatment of guinea pig spermatozoa with dithiothreitol (DTT) to reduce sperm disulfides to sulfhydryls resulted in inhibition of sperm capacitation, acrosome reaction and sperm–oocyte interactions [99]. The thiol status of spermatozoa is regulated by androgens since cauda epididymal spermatozoa isolated from castrated or anti-androgen treated rats showed increased levels of sulfhydryls, including those in protamines that correlated with reduced sperm fertilizing ability [100]. While the enzymes involved in sperm sulfhydryl oxidation are not well characterized, a sulfhydryl oxidase activity is present in epididymal luminal fluid which prevented flagellar angulation in hamster caput epididymal spermatozoa induced to acquire motility [98, 101]. Protein disulfide isomerase which catalyzes disulfide bond formation is also present in the epididymal fluid and may participate in sperm disulfide bond formation [102]. Finally, sperm sulfhydryl oxidation has been shown to facilitate protein tyrosine phosphorylation. Spermatozoa exposed to diamide to form disulfides exhibited enhanced tail protein phosphorylation while the reduction of disulfides by treatment with DTT decreased phosphorylation [103]. Further studies also showed a correlation of protein phosphotyrosine phosphatase activity with sperm thiol status [103].

After translation, proteins fold into their native and functional three-dimensional conformations. Proteins that do not fold correctly or that unfold after adopting their native state can aggregate if the cell does not degrade the misfolded protein or assist in protein refolding. Unfolded proteins with exposed hydrophobic sites will attempt to stabilize by interacting with the exposed hydrophobic regions of other unfolded proteins leading to protein aggregation. Cells have the ability to sort misfolded proteins into two different compartments including the JUNQ (juxta-nuclear quality control compartment) for soluble or ubiquitinated misfolded proteins and the IPOD (insoluble protein deposit) for non-ubiquitinated insoluble proteins [104]. This differential sequestration of misfolded proteins into two quality control compartments is conserved from yeast to mammals.

Proteins that self-aggregate and form higher ordered structures with a cross β sheet fibrillar structure are known as amyloids. Although typically associated with neurodegenerative diseases including Alzheimer’s and Parkinson’s disease, amyloid structures have also been shown to be nonpathological and carry out biological functions. Functional amyloids have long been known to exist in bacteria and yeast. However, it has been only recently that functional amyloids were described in mammals. The PMEL protein forms amyloid structures that contributes to the formation of a stable scaffold within melanosomes and is involved in melanin synthesis [105]. Also, in the anterior pituitary gland several protein hormones are stored as amyloids in secretory granules [106].

Within the epididymis, the cystatin CRES contributes to the formation of a functional amyloid structure present within the epididymal lumen [107, 108]. Specifically, CRES localized to a film-like amyloid-containing material that was present in the lumen throughout the epididymis. Gel filtration and Western blotting experiments showed that monomeric forms of CRES were prevalent in the proximal caput epididymal lumen and then disappeared by the distal caput region. The loss of monomeric forms of CRES correlated with the appearance of CRES in SDS-resistant high molecular mass complexes, presumably reflecting a transition of CRES to an aggregated amyloid state [107]. Because the CRES structures were part of the epididymal mileu from normal mice with no pathologies or fertility defects, a functional role for the CRES amyloid structures in epididymal function was implied. The role of CRES amyloid in the epididymis has yet to be determined. One possibility is that the aggregation of CRES is a means to regulate CRES protease-inhibitory function. Surprisingly, in vitro the CRES dimer was a more potent inhibitor of the proprotein convertase 4 (PCSK4) than the monomer [78]. However, the related protein cystatin C loses its cysteine protease-inhibitory function following dimerization [109]. Alternatively, CRES amyloid may function as a scaffold structure within the epididymal lumen perhaps as an organizational center for proteins that may then be transferred to discrete domains of the spermatozoa during epididymal maturation.

The formation of aggregate structures seems to be a common theme within the epididymal lumen. In addition to the CRES amyloid, several other particulate structures have been described. These include epididymosomes which are small membrane bound structures released from the epididymal epithelium by apocrine secretion and which are thought to be a means to transfer hydrophobic proteins to the maturing spermatozoa [110]. The prion protein was found in both the epididymosomes and in a soluble high molecular mass lipophilic complex with the chaperone clusterin [111]. Other structures in the lumen include dense bodies, which appear larger than epididymosomes, and which contain the heat shock proteins HSPD1/HSP60 and tumor rejection antigen 1 (TRA1), a member of the heat shock protein 90 family [112]. Proposed functions for TRA1 include the folding of denatured proteins and multimer assembly. While the epididymosomes are thought to be released from multivesicular bodies/apical blebs that bud off from the epididymal epithelium indicating an intracellular packaging of the proteins into these structures, the presence of several nonmembrane bound proteinaceous structures in the lumen suggests that protein aggregation of several other proteins, like CRES, may form following their secretion into the lumen. In support, like prion protein, SPAM1 (PH20) has been found in both epididymosomes as well as in monomeric and oligomeric complexes in the epididymal lumen [113].

The presence of several distinct types of aggregate structures in the epididymal lumen raises the possibility that some of these may be extracellular equivalents of the JUNQ and IPOD inclusions that are present intracellularly and which are used by the cell to differentially sort aggregated proteins depending on their solubility and ubiquitination state. The fact that the ubiquitin–proteasome pathway is present in the epididymal lumen would suggest that, in addition to ubiquitination of spermatozoa, ubiquitination of luminal proteins also occurs and these proteins may be targeted to the luminal inclusions for downstream removal by endocytosis. Another possibility is that these inclusions represent preassembled protein complexes that are then delivered to the spermatozoa during maturation. In support of this idea, several proteins with proposed roles in fertilization have been found in large molecular mass complexes with chaperone proteins, perhaps as a means to maintain appropriate folding/stability of the complex [112, 114, 115, 116]. Also, several proteins, including prion protein and SPAM1, that become associated with spermatozoa during epididymal transit, are present in the luminal fluid with GPI-anchors. It is possible that these proteins associate with lipoproteins/chaperones such as clusterin or with lipid raft structures in luminal inclusions as a means to keep the proteins stable and soluble prior to their interactions with spermatozoa [113].

Transglutaminases (TGase) are a family of calcium-dependent enzymes that catalyze the formation of a covalent bond either through protein cross-linking via e-(g-glutamyl) lysine bonds or through incorporation of primary amines at selected peptide-bound glutamine residues. These bonds can be intra- or inter-molecular, and result in extremely stable cross-linked products that are resistant to proteolysis [117]. TGase activities have previously been documented in the male reproductive tract including testis, epididymis, prostate and spermatozoa, and have been suggested to play a role in stabilization of FSH-receptor complexes [118], formation of the seminal coagulum [119], sperm motility [120], and suppressing sperm antigenicity [121, 122]. Little is known regarding the expression or function of TGase in the epididymis other than that spermatozoa and fluid from the head of the epididymis exhibited higher levels of activity than the tail [107]. Studies by von Horsten et al. [107] showed that the TGase activity present in the epididymal luminal fluid was a tissue-type TGase, as evidenced by inhibition of its activity in the presence of a tissue type TGase-specific peptide inhibitor. In addition, although in vitro studies indicated that tissue-type TGase activity is optimal at basic pH, ¹⁴C-putrescine incorporation studies to assess TGase activity in the epididymis revealed TGase activity at pH 6.8, which closely approximates luminal pH in the caput epididymis [107]. Thus, these studies indicated that a functional TGase activity is present within the caput epididymal lumen. Although TGase activity has also been detected in spermatozoa, including the activity associated with the sperm head and the cytoplasmic droplet, its roles in sperm maturation are unclear [123]. However, TGase was shown to mediate the binding of spermidine as well as a seminal vesicle secreted protein to the rat sperm surface [124].

The epididymal lumen is a complex microenvironment that is continually being modified by the addition and removal of proteins along the tubule. The proximal or caput region is the most metabolically active region secreting approximately 80 % of the total overall protein secretion in the lumen. Within this same region, more than 99 % of the fluid accompanying the testicular spermatozoa is removed resulting in an extreme concentration of spermatozoa and luminal contents. This concentration of luminal components appears to be necessary for normal sperm maturation events. However, the loss of water content can result in macromolecular crowding which leads to protein misfolding and aggregation [125]. Since an important role of the epididymis is to protect the maturing spermatozoa, it is likely that extracellular quality control mechanisms are in place. Several lines of evidence suggest that this indeed is the case. First, as mentioned above, components of the ubiquitin–proteasome pathway are present in the epididymal lumen suggesting that this well-characterized intracellular quality control system also is functional extracellularly [90]. These observations in the epididymis were one of the first to describe a mechanism for extracellular quality control in any organ system. A number of chaperones are also present in the epididymal lumen where they may bind to misfolded proteins and either prevent their aggregation or assist in refolding. It is intriguing that the chaperone clusterin contributes approximately 41 % of the total luminal protein content in the rat caput epididymis, possibly functioning as a mediator of extracellular quality control [126]. The observation that many of the particulate structures found in the epididymal lumen contain chaperones suggests an active role for these proteins in quality control systems. Studies have also suggested that the PTM by transglutaminase crosslinking may contribute to quality control in the epididymis. The cystatin CRES is a substrate for transglutaminase cross-linking [107]. Furthermore, exposure of CRES to TGase resulted in the formation of an aggregate structure that was distinct from the highly organized structures of amyloid that CRES forms in the absence of TGase. These data suggest that TGase cross-linking may be a PTM that shifts potentially cytotoxic amyloidogenic structures into stable cross-linked amorphous structures that are not inherently cytotoxic and which may then be taken up by endocytosis [107].

It is likely that within the epididymal lumen, there is a delicate balance between monomeric and oligomeric forms of proteins with aggregate structures carrying out active roles in sperm maturation as well as in the removal or sequestration of luminal proteins [127]. Although extracellular quality control mechanisms are not well described in any organ system, the presence of the ubiquitin–proteasome system, significant levels of TGase activity, and extremely highly levels of chaperones suggest that the epididymal luminal environment has adopted extreme and perhaps unusual measures to ensure viability of the maturing spermatozoa and thus provides an interesting model system for dissecting out the roles these components play in quality control.