For characterization of antigen epitopes for SI-Abs, a number of human and mouse monoclonal antibodies with complement-dependent sperm-immobilizing activity were generated in our laboratory [10–12]. A human monoclonal antibody, Mab H6-3C4, with a high titer of SI activity was established using peripheral B-lymphocytes from an infertile woman [13]. A mouse monoclonal antibody, 1G12, reactive to human sperm membrane also showed a high titer of SI activity [4]. Another mouse monoclonal antibody, S19, generated by Herr’s group, showed strong sperm agglutinating and SI activities and the corresponding antigen was termed as SAGA-1 (sperm agglutination antigen-1) [14]. Campath-1 is a rat monoclonal antibody defining CD52 as an antigen [15]. It was established against human spleen cells and reacted with virtually all leucocytes. Subsequent studies showed campath-1 was cross-reactive to mature human sperm [16] with sperm agglutinating and immobilizing activities similar to monoclonal antibodies generated to sperm antigens.

Tandem mass spectrometric analysis shows that there are distinct differences in the N-linked carbohydrates between lymphocyte-CD52 and mrt-CD52 [17, 18]. Both lymphocyte- and mrt-CD52 are GPI (glycosylphosphatidylinositol) anchor glycoproteins and the molecular conformation formed by three C-terminal amino acids and the GPI anchor is recognized by campath-1 [19]. The observation that Mab H6-3C4 recognizes exclusively with sperm suggests that this monoclonal antibody reacts a sperm-specific antigen present in a carbohydrate moiety [20]. Other CD52-recognizing monoclonal antibodies such as 1G12 and campath-1 react with sperm and also with lymphocytes [4]. This suggests that the epitopes for these monoclonal antibodies are the common sites of mrt-CD52 and lymphocyte-CD52.

Figure 11.4 shows indirect immunofluorescent stainings of human sperm and lymphocytes with Mab H6-3C4 and campath-1. Both monoclonal antibodies stain the whole sperm surface but lymphocytes are stained with campath-1 only. It appears that the antigens recognized by these monoclonal antibodies are similarly distributed on the sperm surface. Mab H6-3C4 did not react with lymphocytes and exclusively recognizes mrt-specific antigen while campath-1 recognizes a core structure of CD52 shared by lymphocytes and mrt.

For detailed analysis of the epitopes, sperm extracts were subjected to high-resolution two-dimensional polyacrylamide gel electrophoresis with the first dimension in a pH 2–4 range and the second dimension in molecular sieving followed by Western blot analysis [20]. As positive control, anti-CD52 antibody produced to a core peptide comprising 12 amino acids was used. For carbohydrate analysis, mrt-CD52 extracted from sperm was treated with N-glycosidase F to remove the N-linked carbohydrate. The presence of O-linked carbohydrates was examined by mild alkaline treatment. Figure 11.5 shows that anti-CD52 peptide antibody reacts with intact mrt-CD52 molecules showing a heterogeneous staining pattern of PI <2.8 and MW 15–25 K (Fig. 11.5a). This heterogeneity is markedly reduced by deglycosylation of N-linked carbohydrate (Fig. 11.5e). Additional removal of O-linked carbohydrates results in staining of a single spot (Fig. 11.5i), suggesting that the O-linked carbohydrate contributes to molecular polymorphism of mrt-C52. Recently, the existence of the O-linked carbohydrate in mrt-CD52 has been demonstrated by lectin binding assay [21] and MALDI-TOF mass spectrometry [22]. Monoclonal antibodies, Mab H6-3C4, 1G12 and campath-1, show similar polymorphic staining pattern in the region of PI <2.8 and MW 15–25 K (Fig. 11.5b–d). The patterns of staining change after removal of the N-linked carbohydrate. In the case of Mab H6-3C4, no staining is observed after the removal of the N-linked carbohydrate (Fig. 11.5f). In the case of 1G12 and campath-1, heterogeneity is reduced but still several spots remained. 1G12 shows with six spots at different pH, while campath-1 reacted with three spots, suggesting that the epitope for 1G12 is not identical to that for campath-1 (Fig. 11.5g, h). These results show that Mab H6-3C4 recognizes the N-linked carbohydrate moiety of mrt-CD52, while 1G12 and campath-1 recognize the core portion of mrt-CD52. After further removal of the O-linked carbohydrate, 1G12 and campath-1 yield single spots like the positive control (Fig. 11.5k, l). Collectively, these results confirm that Mab H6-3C4 and 1G12 recognize mrt-CD52 but the epitopes are different. The epitope for Mab H6-3C4 is present in the N-linked carbohydrate, while the epitope for 1G12 is present in the core portion of CD52. These results indicate that SI-Abs in some infertile women produced against mrt-specific carbohydrate antigens in the mrt-CD52 molecule. Indeed, it has been reported that mrt-CD52 contains specific carbohydrate chains [18].

Based on biochemical and immunological analyses, a hypothetical structure of mrt-CD52 is presented in Fig. 11.6. The core peptide of CD52 is composed of just 12 amino acids and common to lymphocytes and mrt. This suggests that the peptide portion is a scaffold for supporting carbohydrate moieties. The core peptide contains³Asn(N)⁴Asp(X)⁵Thr(T), a consensus sequence, for N-linked carbohydrate binding. It has been reported that mrt-CD52 is heavily glycosylated with heterogeneous carbohydrate chains comprising more than 50 different glycoforms which are almost completely sialylated and fucosylated in 10–15 % of total mrt-CD52 [18]. In contrast, the carbohydrate moieties in lymphocyte CD52 are much smaller and only lightly sialylated but not fucosylated [17]. Another distinct structure of the mrt-CD52 carbohydrate is [GluNAcβ1-6Man] in the N-linked carbohydrate chain. The presence of the β1-6 bond possibly allows branching of a carbohydrate chain to the backbone [23]. The carbohydrate branching via this bond has been reported to contribute to the metastatic potential of tumor [24].

The GPI anchor portion is bound to Ser residue at the carboxyl terminal through ethanolamine linked to three mannose, inositol residue and glycerolipid. Approximately 80 % of the inositol residue are acylated (mainly palmitoylate) at the 2-position as shown by ** in Fig. 11.6. The low susceptibility of mrt-CD52 to phospholipase C may be due to this acyl group anchoring into the cell membrane because this anchoring is known to be refractory to phospholipase C. Another structure difference from lymphocyte CD52 is that the glycerolipid portion of mrt-CD52 is a sn-1-alkyl-lyso-glycerol type (single-footed) in which only one fatty acid chain at the 1-position is linked as shown by * in Fig. 11.6. 1-Alkyl structure is reported to be synthesized by sperm as a 1-alkyl-2-acetyl-sn-glycero-3-phosphocholine but mono-alkyl structure has not been documented in mammalian species. Phospholipase A_2, which is detected abundantly in seminal plasma, removes the acylation at the 2-position, although inhibitory factors are also detected in the seminal plasma. The lyso- (single-footed) glycerolipid anchor may play an important role in transportation of this molecule to sperm from the epithelium in the cauda epididymis or from seminal plasma.

mrt-CD52 has been reported easily to be transferred from epithelial cells to mature sperm in the epididymis [25]. Epididymosome, exosome from epithelial cells, may be a possible transporter of mrt-CD52 as reported by Sullivan et al. [26]. More recently, ACE (Angioteisin-converting enzyme) found in the epididymis and testis has been shown to exhibit GPI anchor protein-releasing activity (GPIase) [27]. Considering the crucial role of this enzyme for fertilization, GPI anchoring proteins may play important roles at different stages of reproduction.

CD52 is a GPI anchor glycoprotein present in lymphocytes and male reproductive tissues including mature sperm and seminal plasma [28, 29]. It has been reported that CD52 on the lymphocyte surface induces regulatory T cells with immunosuppressive activities [28], while soluble mrt-CD52 from epididymis induces clot formation and liquefaction of human semen [30]. However, the biological significance of mrt-CD52 anchoring to the sperm membrane is not well understood.