Eppin: An effective target for male contraception

Molecular and Cellular Endocrinology 250 (2006) 157–162

Eppin: An effective target for male contraception M.G. O’Rand ∗ , E.E. Widgren, Zengjun Wang, R.T. Richardson Department of Cell and Developmental Biology, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA

Abstract Eppin (epididymal protease inhibitor) is a member of the whey acidic protein (WAP)-type four-disulfide core (WFDC) gene family. This study provides updated information on Eppin and the Eppin-like genes within the Eppin cluster on human chromosome 20. A virtual structural model of the Eppin protein demonstrates that the C-terminal half of Eppin is structurally homologous to the Kunitz-type trypsin inhibitor. The Eppin N-terminal may have structural similarities to defensin-type molecules, rather than to that of the WAP consensus sequence. Human spermatozoa have a receptor for Eppin. When recombinant semenogelin (Sg) is digested with PSA many low molecular weight fragments are produced. However, when Eppin is bound to Sg, digestion by PSA is modulated. Addition of antibodies to the C-terminal of Eppin resulted in blocking PSA activity modulation. We can hypothesize from our analysis of anti-Eppin epitopes on Eppin that when anti-Eppin antibodies are bound to Eppin on the sperm surface they block the binding site for semenogelin. © 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: Contraception; Eppin; Spermatozoa; Seminal plasma

Eppin (SPINLW1, serine protease inhibitor-like, with Kunitz and WAP domains 1) is a member of the whey acidic protein (WAP)-type four-disulfide core (WFDC) gene family. The WFDC genes are on human chromosome 20q12-q13 in two clusters, one centromeric and one telomeric (Clauss et al., 2002). Eppin is WFDC 7 in the telomeric cluster and is the archetype of WFDC genes characterized by encoding both Kunitz-type and WAP-type four-disulfide core protease inhibitor consensus sequences (Richardson et al., 2001). 1. The Eppin genes Tables 1 and 2 provide updated information on Eppin and the Eppin-like serine protease inhibitor genes within the Eppin cluster on human chromosome 20, showing that both mouse and human genes express WAP-type and Kunitz-type inhibitor domains. Many of the WFDC genes are expressed in numerous tissues, however, WFDC 6, 7 and 8 genes are predominantly expressed in the epididymis and testis with the caveat that WFDC 7 transcripts have been reported in the trachea (Clauss et al., 2002). Whether these tracheal transcripts translate into actual secreted proteins remains to be determined because Eppin ∗

Corresponding author at: Department of Cell and Developmental Biology, CB #7090, 212 Taylor Hall, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-7090, USA. Tel.: +1 919 966 5698; fax: +1 919 966 1856. E-mail address: [email protected] (M.G. O’Rand). 0303-7207/$ – see front matter © 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.mce.2005.12.039

encodes three mRNAs encoding two isoforms, one of which (Eppin 2, Fig. 1B) lacks a secretory signal sequence (Richardson et al., 2001). To determine if transcripts corresponding to the predicted gene products of WFDC 6 and 8 are expressed and to identify any additional splice variants, we used RT-PCR on cDNAs from epididymis and testis to generate sequences and compared these to the predicted gene products in GenBank. As shown in Fig. 1, RT-PCR using testis cDNA produced a transcript that was identical to the WFDC 6 gene product Q9BQY6 except that it contained a fourth exon (AV378987.2). Both testis and epididymis cDNAs yielded sequences matching the WFDC 8 gene product, Q8IUA0, which is transcribed from five exons. However, an additional transcript (Transcript 1.8) from testis was identical to Q8IUA0 except that it lacked exon 4 (Fig. 1C). The known splice variants for WFDC 7 are also shown in Fig. 1 (Richardson et al., 2001). The mouse homologues of WFDC 6, 7 and 8 are expressed only in the epididymis and/or testis, except for the mouse gene WFDC 6b (XP 485099) that does not have a human homologue, which is present in several somatic tissues. 2. The Eppin protein Two isoforms of the Eppin protein are expressed: one is secreted and one lacks a signal sequence. The secreted form of Eppin (CAB37635) has a theoretical pI of 8.52 and a calculated molecular weight of 15,283.72. Analysis of native Eppin

158

M.G. O’Rand et al. / Molecular and Cellular Endocrinology 250 (2006) 157–162

Table 1 Characteristics of Eppin and the Eppin-like serine protease inhibitor genes within the Eppin cluster on human chromosome 20 WFDC family number

Accession number (human)

Exons

Protein AA residues

Number of WAP-type domains

Number of Kunitz-type domains

Mouse homologue

% Homology

WFDC 6 WFDC 7 WFDC 8

AV378987.2 O95925 (Eppin-1) Q8IUA0

4 4 5

131 133 241

1 1 3

1 1 1

XP 130716 NP 083601 Eppin XP 204940

48 62 53

Table 2 Characteristics of Eppin and the Eppin-like serine protease inhibitor genes within the Eppin cluster on mouse chromosome 2 Accession number/gene designation (mouse)

Human WFDC gene family number

Exons

Protein AA residues

Number of WAP-type domains

Number of Kunitz-type domains

XP NP XP XP

WFDC 6 WFDC 7 WFDC 8 None

4 4 6 4

136 134 311 137

1 1 3 1

1 1 1 1

130716 Gm122 083601 Eppin 204940 Gm706 485099 WFDC 6b

by SDS-PAGE indicates that the monomer form has an apparent molecular weight of 16–18 kDa and the dimer an apparent molecular weight of 33–36 kDa (Wang et al., 2005). Higher multimer forms of Eppin are often seen in seminal plasma and analysis of recombinant human Eppin indicates that multimer forms (2×, 3×, 4×), which are stable to boiling in SDS in the presence of reducing agents, easily form in vitro (Wang et al., 2005). A virtual structural model of Eppin constructed on Swiss-model, an automated comparative-modeling server http://swissmodel.expasy.org/ demonstrates that the C-terminal half of Eppin (amino acids 73–128) is structurally homologous to the Kunitz-type trypsin inhibitor (1KTHA.pdb). The

Kunitz-type inhibitor P1 reactive site is a leucine residue (Leu87 ) that is indicative of specificity for chymotrypsin-like inhibition (Laskowski and Kato, 1980). The N-terminal half is similar to the elastase inhibitor, elafin (2REL.pdb). However, the Eppin Cys40 -Cys65 residues are not close enough to one another to form a disulfide bridge between the inner and outer loops as the Cys23 -Cys49 residues do in elafin, indicating that Eppin’s N-terminal is not structurally identical to elafin. Each half of Eppin appears as a compact double looped structure held in place by disulfide bonds anchoring one loop to the core ␤-sheet of the other loop thereby exposing the WAP and Kunitz-type inhibitor binding faces at opposite ends of the equatorial plane of the molecule. Further modeling will be needed to determine

Fig. 1. Splice variants of WFDC 6, 7 and 8. (A) RT-PCR on cDNAs from testis produced a transcript that was identical to the WFDC 6 gene product Q9BQY6 except there was the addition of a fourth exon (AV378987.2’). (B) There are three known splice variants of Eppin. (C) RT-PCR on cDNAs from epididymis and testis produced a transcript that was identical to the WFDC 8 gene product Q8IUA0, which is transcribed from five exons. However, an additional transcript (Transcript 1.8) from testis was identical to Q8IUA0 except that it lacked exon 4.

M.G. O’Rand et al. / Molecular and Cellular Endocrinology 250 (2006) 157–162

159

how disulfide bridges configure the N-terminal’s double loop structure and how the two halves of Eppin fit together. The recent demonstration that Eppin has antimicrobial activity (Yenugu et al., 2004), which depends upon intact disulfide bridges, together with previous reports of antimicrobial activity of WAP containing proteins (cited in Yenugu et al., 2004) implies that Eppin may have structural similarities to other defensin-type molecules. Such similarities may have structural implications for the N-terminal in which disulfide bridges could be aligned in an alpha or beta defensin configuration, rather than the assumed four-disulfide core of the WAP consensus sequence. 3. Eppin binding to spermatozoa Human ejaculate spermatozoa are coated with Eppin over both head and tail regions (Richardson et al., 2001; Wang et

Fig. 3. Binding of Eppin to the surface of live human spermatozoa. Saturation kinetics of 125 I-recombinant human Eppin (rEppin) binding to live human “swim up” spermatozoa from two different patients (Experiments #1 and #2) attending the UNC IVF clinic. Non-specific (background) binding has been subtracted from each sample. Competition with unlabeled rEppin reduced the specific binding to approximately 0.3 pmole.

al., 2005) and the origins of the Eppin coat maybe both testicular and epididymal. Eppin is present in mouse Sertoli cells (Sivashanmugam et al., 2003) and although this has not been confirmed in human testis, it is likely that Eppin first interacts with the spermatid surface in the testis. As spermatozoa move into the human ductuli efferentes they encounter epididymal Eppin and Eppin continues to be present in both caput and cauda epididymal epithelium (Richardson et al., 2001). Immediately following ejaculation when spermatozoa are part of the ejaculate coagulum (Fig. 2A), Eppin is an integral part of that coagulum. Fig. 2B shows that spermatozoa embedded within the coagulum are coated with Eppin. On the surface of spermatozoa there is a receptor for Eppin. This can be demonstrated by the saturation kinetics of 125 Irecombinant human Eppin (rEppin) binding to live human “swim up” spermatozoa from two different patients attending the UNC IVF clinic (Fig. 3). Both samples had >99% motility and whiplash swimming movements in each sample were observed in the microscope. Spermatozoa from both patients fertilized eggs in the clinic. Non-specific (background) binding has been subtracted from each sample. Competition with unlabeled rEppin reduced the specific binding to approximately 0.3 pmole. We may conclude from this data that Eppin has a saturable binding site (receptor) on fertile, live, human “swim up” spermatozoa. 4. Association of Eppin with semenogelin Fig. 2. Spermatozoa and coagulum of a human ejaculate. (A) Phase contrast image of the coagulum containing embedded spermatozoa. Spermatozoa are oriented with their tails in the center of the mass. (B) Fluorescent microscope image of spermatozoa embedded in a coagulum stained with affinity purified anti-Eppin (Q20E) and detected with Alexa Fluor-488 labeled goat anti-rabbit antibodies. Arrows (←) indicate the heads of spermatozoa embedded the coagulum mass. The strong fluorescent staining of the sperm tails gives the coagulum mass a honey-combed appearance.

As spermatozoa pass through the ampulla of the vas deferens they encounter the proximal extension of the seminal vesicle and enter the ejaculatory duct. At this juncture the copious secretions formed in the seminal vesicles mix with the spermatozoa. Shortly thereafter as the ejaculatory ducts pass through the prostate and enter the prostatic urethra the spermatozoa and seminal fluid mix with prostatic secretions. As spermatozoa enter the

160

M.G. O’Rand et al. / Molecular and Cellular Endocrinology 250 (2006) 157–162

Table 3 Summary of fertility trial of male monkeys immunized with recombinant human Eppin (from O’Rand et al., 2004) Immunized male monkeys N = 7 Control male monkeys N = 6

Female cycles tested N = 21 Female cycles tested N = 18

ejaculatory ducts their surface coating of Eppin should become saturated with semenogelin (Sg), the major protein constituent of seminal fluid (de Lamirande et al., 2001). The binding of Sg to Eppin (Wang et al., 2005) constitutes the major protective event that occurs to ejaculate spermatozoa in primates, providing antimicrobial activity to a tight agglutination of spermatozoa in a coagulum (Fig. 2A). The protective mass of spermatozoa and seminal vesicle proteins that constitute the coagulum necessitates its liquefaction by hydrolysis immediately after ejaculation in order to free the spermatozoa for both motility and fertility (Robert and Gagnon, 1996, 1999). The cleavage of Sg on spermatozoa and in the coagulum is accomplished by prostatic specific antigen (PSA), a serine protease in the kallikrein family (Robert et al., 1997). 5. Evidence that Eppin is involved in male fertility To study the role of Eppin in primate fertility an immunological approach was taken to answer the question: If Eppin’s function is blocked by an antibody, what will be the effect on fertility? Following the immunization of adult male Macaca monkeys, progressive sperm motility in their ejaculates dropped and coincided with the appearance of an anti-Eppin titer in their semen (O’Rand et al., 2004). Additionally, accompanying the appearance of an anti-Eppin titer in the semen was a loss of the semen coagulum. These two observations indicated that Eppin is a potential target for the development of a contraceptive and warranted fertility testing. As summarized in Table 3, the fertility study (O’Rand et al., 2004) demonstrated that effective and reversible male immunocontraception in primates is an obtainable goal. We found that a high serum titer (>1:1000) sustained over several months achieves an effective level of contraception. Seven out of nine males (78%) developed high titers to Eppin, and all these high titer monkeys were infertile. Five out of seven (71%) high anti-Eppin titer males recovered fertility when immunization was stopped. The results from the fertility study suggest that antibodies directed at Eppin have the potential of disrupting an essential step in the pre-fertilization preparation of spermatozoa in the male reproductive tract and in the ejaculate that is initially deposited in the vagina. 6. Eppin–semenogelin interaction during liquefaction As previously reported by Gagnon and co-workers (Robert et al., 1997), during liquefaction of semen, activated PSA cleaves semenogelin (Sg) bound to the sperm surface, releasing the sperm motility inhibitory factor (amino acids 69–160; Robert and Gagnon, 1996). We now know that Sg on the sperm surface is bound to Eppin and therefore the cleavage of Sg by PSA must occur while Sg is bound to Eppin. Consequently, we compared

Females pregnant 0% Females pregnant 4/6 = 67%

in vitro the digestion of Sg by PSA in the presence or absence of recombinant Eppin. As shown in Fig. 4, when recombinant Sg (Sg, lane 4) is digested with PSA many low molecular weight fragments are produced (lane 3). However, when Eppin is bound to Sg, digestion by PSA is modulated, producing incomplete digestion and a 15–16 kDa fragment (asterisk, lane 2). Our understanding of Eppin’s essential role in sperm survival during transfer from male to female reproductive tracts prior to fertilization stems from an analysis of anti-Eppin antibody binding sites (epitopes) on Eppin. As described previously (O’Rand et al., 2004), sera from the infertile male monkeys immunized with Eppin recognized two predominant epitopes, one N-terminal (QGPGLTDWLFPRRCPKIRE; amino acids 20–38) and another C-terminal (TCSMFVYGGCQGNNNNFQSKANCLN; amino acids 101–125). Production of antibodies to N-terminal amino acids 20–39 (anti-Q20E; Wang et al., 2005) and to recombinant Eppin (Richardson et al., 2001) have been described and their linear B-cell epitopes are shown in Fig. 5. To test the effect of specific anti-Eppin antibodies on the PSA hydrolysis of Sg, as it might occur in vivo, either anti-Q20E or anti-recombinant Eppin was incubated with Eppin, incubation continued with the addition of Sg, and finally PSA added for a final incubation period. Addition of anti-Q20E had no effect on Eppin–Sg binding as monitored by PSA digestion of Sg. PSA activity was modulated with a result identical to that shown in Fig. 4, lane 2. Therefore, antibody binding to the N-terminal of Eppin did not affect Sg digestion. However, addition of antibodies to the C-terminal of Eppin resulted in blocking PSA activity modulation. Consequently digestion with PSA produced many low molecular weight fragments and notably the protected

Fig. 4. Digestion of recombinant semenogelin (Sg) by PSA in the presence or absence of recombinant Eppin. SDS-PAGE gel stained with Coomassie Blue. Lane 1: recombinant Sg bound to Eppin digested with PSA in the presence of antibodies to the C-terminal of Eppin resulted in blocking PSA activity modulation, producing low molecular weight fragments (Sg{) and notably the loss of the protected 15–16 kDa fragment (asterisk, lane 2). IgG (h) and IgG (l) are the heavy and light chains of the IgG antibody. Lane 2: recombinant Sg bound to Eppin digested with PSA producing incomplete digestion (arrows) and a 15–16 kDa fragment (asterisk). Lane 3: recombinant Sg digested with PSA producing low molecular weight fragments (Sg{). Lane 4: recombinant semenogelin (Sg). Lane 5: recombinant Eppin.

M.G. O’Rand et al. / Molecular and Cellular Endocrinology 250 (2006) 157–162

161

Fig. 5. Linear B-cell epitopes of human Eppin. Human Eppin has 133 amino acids containing a conserved WAP and Kunitz domain. Recombinant proteins were made to N-terminal fragment 18-75 and C-terminal fragment 76-133 (Wang et al., 2005). Analysis of Eppin linear B-cell epitopes was performed on peptide pin blocks as described previously (O’Rand et al., 2004; O’Rand and Widgren, 1994). The Eppin sequence from amino acids 20 to 133 was divided into 36 sequential overlapping decapeptides (seven amino acids overlap) and synthesized. A 1:2000 dilution of serum was used and baseline reactivity was established with control macaque serum as described (Lea et al., 1997). Immunodominant linear B-cell epitopes are determined by their z scores (Lea et al., 1997).

15–16 kDa fragment (asterisk, lane 2) was absent (Fig. 4, lane 1). Analysis of the protected fragment by tandem mass spectrometry (MS/MS) revealed that it contained cys239 , the residue necessary for Eppin binding (Wang et al., 2005). Moreover, the N-terminal Sg sequence containing the sperm motility inhibiting peptide (Robert et al., 1997) had been cleaved from the cys239 -containing fragment by PSA into a 10.4 kDa fragment, which would presumably no longer be anchored to Eppin. While sperm motility inhibiting peptide is bound to sperm they remain immotile and its removal is necessary for resumption of motility (Robert and Gagnon, 1999) and subsequent capacitation (de Lamirande et al., 2001). We can hypothesize from our analysis of anti-Eppin epitopes on Eppin that when anti-Eppin antibodies in the infertile male monkeys entered the epididymal fluid and bound to Eppin on the sperm surface they blocked the binding site for semenogelin. Blocking the binding of Sg had two consequences. First, as a result of not being bound to Eppin, Sg in the ejaculate was quickly hydrolyzed into small fragments; there was no modulation of PSA activity and no semen coagulum was observed. Second, having anti-Eppin bound to Eppin on the sperm surface, mimicked the physiological effect of having sperm motility inhibiting peptide bound to the surface, namely a loss of forward motility, which was observed in semen from infertile males. This second consequence predicts that the removal of anti-Eppin antibodies from the sperm surface would allow spermatozoa to recover their motility. Further studies are underway to determine if this is correct.

Acknowledgements Grant support: Supported by grant CIG-96-06 from the CICCR Program of CONRAD (Contraceptive Research and Development Program), the Andrew W. Mellon Foundation, and by D43TW-HD00627, Program for International Training and Research in Population and Health from the Fogarty International Center and the National Institute of Child Health and Human Development. ZW was supported by a Fogarty Postdoctoral fellowship. References Clauss, A., Lilja, H., Lundwall, A., 2002. A locus on human chromosome 20 contains several genes expressing protease inhibitor domains with homology to whey acidic protein. Biochem. J. 368, 233–242. de Lamirande, E., Yoshida, K., Yoshiike, T.M., Iwamoto, T., Gagnon, C., 2001. Semenogelin, the main protein of semen coagulum, inhibits human sperm capacitation by interfering with the superoxide anion generated during this process. J. Androl. 22, 672–679. Laskowski, M., Kato, I., 1980. Protein inhibitors of proteinases. Ann. Rev. Biochem. 49, 593–626. Lea, I.A., Adoyo, P., O’Rand, M.G., 1997. Autoimmunogenicity of the human sperm protein, Sp17 in vasectomized men and identification of linear B cell epitopes. Fertil. Steril. 67, 355–361. O’Rand, M.G., Widgren, E.E., 1994. Identification of sperm antigen targets for immunocontraception: B-cell epitope analysis of Sp17. Reprod. Fertil. Dev. 6, 289–296. O’Rand, M.G., Widgren, E.E., Sivashanmugam, P., Richardson, R.T., Hall, S.H., French, F.S., VandeVoort, C.A., Ramachandra, S.G., Ramesh, V., Jagannadha Rao, A., 2004. Reversible immunocontracep-

162

M.G. O’Rand et al. / Molecular and Cellular Endocrinology 250 (2006) 157–162

tion in male monkeys immunized with Eppin. Science 306, 1189– 1190. Richardson, R.T., Sivashanmugam, P., Hall, S.H., Hamil, K.G., Moore, P.A., Ruben, S.M., French, F.S., O’Rand, M.G., 2001. Cloning and sequencing of human Eppin: a novel family of protease inhibitors expressed in the epididymis and testis. Gene 270, 93–102. Robert, M., Gagnon, C., 1996. Purification and characterization of the active precursor of a human sperm motility inhibitor secreted by the seminal vesicles: identity with semenogelin. Biol. Reprod. 55, 813–821. Robert, M., Gagnon, C., 1999. Semenogelin I: a coagulum forming, multifunctional seminal vesicle protein. Cell Mol. Life Sci. 55, 944–960. Robert, M., Gibbs, B.F., Jacobson, E., Gagnon, C., 1997. Characterization of prostate-specific antigen proteolytic activity on its major physiological

substrate, the sperm motility inhibitor precursor/semenogelin I. Biochemistry 36, 3811–3819. Sivashanmugam, P., Hall, S.H., Hamil, K.G., French, F.S., O’Rand, M.G., Richardson, R.T., 2003. Characterization of mouse Eppin and a gene cluster of similar protease inhibitors on mouse chromosome 2. Gene 312, 125–134. Wang, Z., Widgren, E.E., Sivashanmugam, P., O’Rand, M.G., Richardson, R.T., 2005. Association of Eppin with semenogelin on human spermatozoa. Biol. Reprod. 72, 1064–1070. Yenugu, S., Richardson, R.T., Sivashanmugam, P., Wang, Z., O’Rand, M.G., French, F.S., Hall, S.H., 2004. Antimicrobial activity of human EPPIN, an androgen-regulated, sperm-bound protein with a whey acidic protein motif. Biol. Reprod. 71, 1484–1490.