All blood samples were obtained between 8 and 10 a.m. and were immediately transferred to the laboratory for further processing. Hormonal analyses included the determination of FSH, LH, TT and hPRL. FSH and LH analysis was performed using a standard two step immunoassay (Architect FSH, Abbott; ref. B7K750; Architect LH, Abbott; ref. 34-4522/R8). hPRL levels were determined by the two-step immunoassay Architect Prolactin (Abbott; ref. B7K760), TT levels by the Architect Testosterone Assay (Abbott; ref. B7K730). All assays were performed according to manufacturer's instructions.

Human semen samples were collected after 4 days of ejaculatory continence from 45 male patients undergoing semen analysis as part of an infertility workup at the Andrology Division of the Department of Dermatology, Univ. of Erlangen-Nuremberg, Germany, between 1995 and 1997. Semen samples were obtained and processed as recommended in the WHO Laboratory Manual for the Examination of Human Semen and Semen-Cervical Mucus Interaction 15 . None revealed the presence of antisperm antibodies in their seminal plasma or sera as shown by the immunobead technique 16 . The seminal fluid remaining after routine testing was aspirated from the spermatozoal pellet and stored at -20°C until use. After thawing, the seminal fluids were centrifuged at 40,000 × g to remove residual debris and then assayed. All patients gave written informed consent.

Semen analysis was performed by determination of pH, liquefaction time, volume, total and progressive motility at 60 min after ejaculation, sperm/round cell/germ cell/leukocyte counts, and the total and specific head/midpiece/tail percentage of abnormal forms according to the WHO criteria of 1995. Morphology was determined on methanol-fixed (2.5 min) and giemsa-stained (10 min) sperm smears by light microscopy and by two independent observers by using the strict Kruger criteria. Viability of sperm was measured by eosin Y staining. The concentration of polymorphonuclear neutrophil white blood cells was measured by peroxidase staining.

Patients were grouped according to their semen analysis; abnormal semen results (n = 29): oligoasthenoteratozoospermia (OAT, n = 7), asthenoteratozoospermia (n = 2), asthenozoospermia (n = 3), teratozoospermia (n = 1), oligozoospermia (n = 2), cryptozoospermia (n = 8), azoospermia (n = 6), and normozoospermic men (n = 16).

Generation and characterization of monoclonal Antibodies (mAbs) against intact hCG, hCGβ, hCGβcf, hCGα and FSH and GH, hPRL and PL has been described in detail 17 18 19 20 . The mAbs against hCG/hCG-variants were previously used as reference reagents in the international TD-7 Workshop on antibodies to hCG and hCG-related molecules 11 . Sensitive and specific IFMAs for hCG, free hCGα, free hCGβ and the hCGβcf, respectively, were developed on the basis of our panel of mAbs 21 22 . MAb pairs for each of the four IFMAs were selected on the basis of antigen specificity, epitope localization and compatibility 11 17 18 23 24 25 .

IFMAs were performed as published previously 26 27 . Samples were diluted in 0.01 mol/l NaHCO₃ containing 0.1% bovine serum albumin (BSA).

The National Institute for Biological Standards and Control (NIBSC; South Mimms, UK) kindly provided the International Standards (IS) for hCG IS75/537, hCGβ IS 75/551, hCGα IS 74/569. Highly purified hCGβcf was a gift by Drs. Klaus Mann and Rudy Hoermann (Essen, Germany).

Coating mAbs were coded as INN(sbruck)-hCG-45 (hCG+hCGn-assay), -68 (hCGβ-assay), -106 (hCGβcf-assay) and INN-hCG-72 (hCGα-assay). The detection mAbs (INN-hFSH-158 for the hCGα-assay were directed against epitope α3, presumably located on loop 3 of GPHα. To improve assay homogeneity, a single mAb (code: INN-hCG-22) recognizing a broad spectrum of hCG and hCG-like molecules, i.e. hCG, nicked hCG (hCGn), hCG lacking the carboxyl-terminal peptide of the β-subunit (-CTPhCG), hCGβ, hCGβn, -CTPhCGβ and hCGβcf, was labeled with isothiocyanatophenylene triamintetraacetic acid-europium (Wallac, Turku, Finland) according to the manufacturer's recommendations 27 and used as a detection reagent in the 3 assays for hCG, hCGβ and hCGβcf, respectively.

The specificities of the applied IFMAs for hCG and hCG variants are summarized in Table 1.

Seminal plasma samples from three normozoospermic men were pooled and 1 ml of the pool was diluted 1:2 with modified RIPA buffer (10 mM Tris-HCl pH 7.4; 150 mM NaCl; 1% NP-40; 0.25% Na-deoxycholate, complete Mini Protease Inhibitor Cocktail (Roche Diagnostics) and 15 μl monoclonal antibody (4 mg/ml; Code INN-hFSH-132; 18 ). The mixture was incubated overnight at 4°C on a rotary shaker. A Protein-G agarose resin (Upstate) was added and samples were incubated for further 2 h at 4°C. Thereafter, samples were centrifuged, supernatant removed and the Protein-G agarose resin was washed 5 times with modified RIPA buffer. The Protein-G agarose resin was heated to 95°C for 10 min in 50 μl modified RIPA buffer and centrifuged at 16,000 g for 10 min. The pellet was discarded and the hCGα-containing supernatant stored at -20°C.

hCGα purified from seminal plasma by IP, as described above (2.4.) and, for comparative purposes, large free hCGα purified by HPLC from supernatant of HEK293 stably transfected with β₂AR and specifically stimulated with 10 μM isoproterenol 28 , as well as the frozen carrier-free concentrate (FC 862) of the WHO adopted 1^st International Reference Preparation for Immunoassay of hCGα (1^st IRP hCGα 99/720) were deglycosylated with peptide N-glycanase PNGase F (New England BioLabs). For digestion under non-reducing conditions to remove the glycan at Asn⁵² 29 30 , 15 μl samples equivalent to 150 ng hCGα were incubated with 1.5 μl 0.5 M sodium phosphate buffer (pH 7.5), 1.5 μl NP-40 and 1 μl Enzyme (PNGase F, 500 U) for 2 hrs at 37°C. For digestion under reducing conditions to remove both glycan moieties, samples were incubated with 1.5 μl 10× denaturing buffer (5% SDS, 10% β-mercaptoethanol) for 10 min (95°C) and put on ice prior to deglycosylation.

Samples were diluted with sample buffer to 25 μl, separated via gel electrophoresis on 4% stacking and 13% separating gels and transferred to an Immun-Blot™ polyvinylidene difluoride (PVDF) membrane (Bio-Rad Laboratories). Membranes were probed with mouse monoclonal antibody INN-hFSH-132 at a dilution of 1:2,000 for blots under non-reducing conditions and rabbit hCGα antiserum at a dilution of 1:10,000 for reduced proteins, respectively. Detection was performed with HRP-conjugated secondary antibodies (Promega), chemoluminescent substrate (Amersham ECL™ Western Blotting Analysis System, GE Healthcare) and exposure to ECL Hyperfilm (GE Healthcare).

hCGα purified from seminal plasma by IP, as described above (2.4.), was analyzed by SDS-PAGE, and the protein band at 24 kDa was excised from the gel and digested with endoproteinase Lys-C [EC 3.4.21.50] (Sigma-Aldrich, 1/20 w/w) in 100 mM H₄HCO₃ buffer (pH = 8.0) for 2 hours at 37°C. The digest was analyzed using nano-HPLC consisting of an UltiMate 3000 System (Dionex Corporation) connected online to a linear iontrap mass spectrometer ThermoElectron Finnigan LTQ) equipped with a nanospray ionization source. The nanospray voltage was set at 1.6 kV, the heated capillary was held at 200°C.MS/MS, and spectra were searched against a human protein database using SEQUEST (LCQ BioWorks; ThermoFinnigan).

Results are expressed as mean values ± SEM. Statistical differences among groups were calculated by unpaired Student's t-test and considered significant at P < 0.05.

Serum levels of TT, FSH, LH and hPRL were analyzed in 21 patients with abnormal semen analyses (OAT n = 5, asthenozoospermia n = 2, teratozoospermia n = 1, oligozoospermia n = 1, cryptozoospermia n = 8, azoospermia n = 4) and compared with 6 control subjects with normal semen analyses (Figure 2A). Patients with pathologic semen analysis findings showed no significant differences in TT and LH serum levels, while hPRL levels were significantly elevated (6.4 ± 0.7 vs. 4.5 ± 0.3 ng/ml; P = 0.02). Mean serum FSH levels were significantly increased in the abnormal group compared with controls (9.7 ± 1.4 vs. 3.1 ± 0.1 mIU/ml; P = 0.0001), and elevation of FSH levels (Figure 2B) was more pronounced in patients diagnosed with cryptozoospermia (12.4 ± 2.5 mIU/ml, P = 0.003) or azoospermia (12.5 ± 3.8 mIU/ml, P = 0.04) than those diagnosed with OAT (6.6 ± 1.5 mIU/ml, P = 0.04).

Hormone and hormone derivative levels were determined by respective IFMAs in seminal plasma specimens from 29 patients with abnormal semen analyses and compared with samples from 16 control subjects with normal semen analyses (Table 2). hPL and hCGβcf levels were similarly low in both groups, with several samples below the detection limit preventing statistical evaluation. hCGβ, holo-hCG, hGH and hPRL levels were not significantly different in the two cohorts (Figure 2B). However, free hCGα levels were significantly lower in patients with pathologic semen analysis findings (1346 ± 191 vs. 2753 ± 533 ng/ml; P = 0.022).

In patients with impaired semen quality, the highest variability of all analyzed hormones was observed for holo-hCG (0.279 ± 0.101 ng/ml, Table 2). Patients with astheno-, terato- or asthenoteratozoospermia had higher, but statistically insignificant, hCG seminal plasma levels. On the other hand, patients diagnosed with reduced sperm counts (oligo-, oligoasthenoterato-, crypto- or azoospermia; n = 23) had significantly reduced hCG levels compared with normozoospermic men (0.097 ± 0.022 vs. 0.203 ± 0.040 ng/ml, P = 0.028, Figure 3A). While hCGβ levels were not significantly different (1.80 ± 0.27 vs. 2.05 ± 0.29 ng/ml, P = 0.55, Figure 3B), mean hCGα levels were significantly reduced in patients with reduced sperm counts (1237 ± 197 vs. 2753 ± 533 ng/ml, P = 0.015, Figure 3C). Given the fact that hCG as well as hCGα were similarily reduced by approximately 50-60% in both cohorts, the ratio of hCGα/hCG was analyzed and found to be comparable (normozoospermia: 25.7 ± 6.1 × 10³, reduced sperm count: 17.9 ± 3.2 × 10³, Figure 3D). Patients with astheno-, terato- or asthenoteratozoospermia had a strongly reduced hCGα/hCG ratio (2.6 ± 0.5 × 10³, Figure 3D).

Given the high free hCGα levels in seminal fluid compared with serum (approximately 10,000-fold higher, as described previously 13 ) and the decreased levels in seminal plasma of patients with poor semen analysis, the hormone derivative from these patients was purified and investigated in more detail.

Under non-reducing conditions, hCGα dissociated from pregnancy-derived holo-hCG migrated as an approximately 22 kDa band (Figure 4A). Large free hCGα, which is unable to associate with β-subunits due to larger N-linked sugar chains 31 , had an apparent molecular mass (M_r,app) of 24 kDa. hCGα isolated form seminal plasma migrated comparable with HEK293-derived large free hCGα. After digestion of the glycan at Asn⁵², the dissociated standard and HEK293-derived large free hCGα showed identical M_r,app of 18 kDa, while protein isolated from seminal fluid migrated as a diffuse band of approximately 19 - 20 kDa (Figure 4A). When both glycan moieties were removed, all three hCGα variants tested resulted a 15 kDa deglycosylated protein.

The biochemical identity of hCGα purified from seminal fluid was verify by nano-HPLC MS/MS analysis after digestion with endoproteinase Lys-C (Table 3).