MUC1 in normal and impaired spermatogenesis

Molecular Human Reproduction Vol.7, No.6 pp. 505–512, 2001 MUC1 in normal and impaired spermatogenesis Folker E.Franke1,5, Sigurd Kraus2, Claus Eiermann2, Katharina Pauls1, El-Nasir Lalani4 and Martin Bergmann3 Departments of 1Pathology, 2Urology and 3Veterinary Anatomy, Justus-Liebig University Giessen, Germany and 4Department of Histopathology, Imperial College, Hammersmith Campus, London, UK 5To whom correspondence should be addressed at: Department of Pathology, Justus-Liebig University of Giessen, Langhansstrasse 10, D-35392 Giessen, Germany. E-mail: Folker.E.Franke@patho.med.uni-giessen.de The MUC1 mucin [also known as episialin, epithelial membrane antigen (EMA) or polymorphic epithelial mucin (PEM)] is a component of the mucosal glycocalyx, contributing to anti-adhesive and protective cell functions. MUC1 has been shown in a variety of epithelial cell types in the reproductive tracts of males and females, but this is the first report of its expression in human testis and non-epithelial cells of the germ cell lineage. Analysing 65 testes with normal or impaired spermatogenesis, we identified MUC1 protein in maturing germ cells by immunohistochemistry using the monoclonal antibodies HMFG1, HMFG2 and SM3 binding to different glycosylation variants. MUC1 expression was confirmed by reverse transcription-polymerase chain reaction (RT-PCR) and Western blot analysis on tissue extracts of human testis, and RT–PCR of selected germ cells after UV laser-assisted cell picking. MUC1 glycosylation variants were selectively distributed during normal spermatogenesis. Whereas HMFG1 labelled certain groups of pachytene spermatocytes, HMFG2 labelled only spermatids. Low glycosylated forms of MUC1 mucin, recognized by SM3, were not found. In contrast to its weak expression during normal spermatogenesis, the HMFG1 glycosylation variant accumulated markedly in all spermatocytes showing abnormal or arrested maturation. These results suggest a variable glycosylation of MUC1 mucin in differentiating germ cells, which is aberrant in pathological conditions. Key words: immunohistochemistry/MUC1/mucins/RT–PCR/spermatogenesis Introduction MUC1 was the first gene to be discovered of the mucin family (Gendler et al., 1990), which now includes 12 different members, numbered from MUC1 to MUC12 (Williams et al., 1999). Both the MUC1 gene and its products, synonymously known as episialin, epithelial membrane antigen (EMA), polymorphic epithelial mucin (PEM) and DF3- or human milk fat globule (HMFG) antigen, are pleomorphic in many respects. Intron polymorphism occurs and variable numbers (21-125) of tandem repeats (VNTR region) within exon 2 of the MUC1 gene contribute to allelic variation (Gendler et al., 1990; Pratt et al., 1996). Alternative splicing events lead to several isoforms of MUC1, which are either bound to the cell membrane or secreted (Ligtenberg et al., 1990; Williams et al., 1990). The VNTR coded domains in particular are responsible for the extensive but variable O-glycosylation of MUC1 proteins in a tissue-specific manner (Zotter et al., 1988; Müller et al., 1997), but short transmembrane variants, devoid of the VNTR region and mucin-like features, have been described (Zirhan-Licht et al., 1994a; Oosterkamp et al., 1997). Proteolytic cleavage of transmembrane MUC1 proteins may also occur, adding further variability to the secreted forms © European Society of Human Reproduction and Embryology (Boshell et al., 1992). Thus, the molecular organization of MUC1 is complex and MUC1 probably represents a group of related molecules with different potential functions (Figure 1). Like other mucins, MUC1 is expressed in various epithelial cells to form the glycocalyx at mucosal surfaces. It acts as a lubricant, a barrier protecting cells from dehydration, proteolysis and infection. Furthermore it can act as an anti-adhesion molecule (Ligtenberg et al., 1992), serve as a receptor and its cognate binding protein (Baruch et al., 1999) and contribute to signal transduction (Zirhan-Licht et al., 1994b; Quin and McGuckin, 2000). Considerable work has focused on mucins in the reproductive tract of females, where the epithelial expression of MUC1 and its state of glycosylation are likely to be involved in the transit of oocytes and spermatozoa, and in the uterine implantation of embryos (Lagow et al., 1999). In contrast, little is known about mucins in the genital tract of males, except for some reports concerning the epithelium of prostate (Ho et al., 1993), epididymis (Zotter et al., 1988), and the sporadic finding of MUC8 antigenicity on spermatozoa (D’Cruz et al., 1996). In this study we have investigated the expression of MUC1 in the human testis with special regard to germ cells of normal and impaired spermatogenesis. 505 F.E.Franke et al. Figure 1. Molecular organization of MUC1. The diagram shows the genomic polymorphism of this mucin, the different mRNA and protein products and glycosylation variants. The positions of intron 1-spanning primers used in this study are indicated by arrows. Exon 2 codes for variable numbers of tandem repeats (VNTR), responsible for allelic variation. In addition, intron polymorphism occurs. Alternative splice sites and stop codons within the transcript lead to different isoforms of MUC1 mucin, either of full size, or a short molecule lacking the VNTR region (MUC1 Y/Z), or a secreted form lacking the transmembrane region (TM). Post-translationally, proteolytic cleavage of membrane-anchored MUC1 protein may also occur, but the main modifications of the final products are due to extensive O-glycosylation within the protein core of tandem repeats (TR). Here, glycosylation is mediated by N-acetylgalactoaminyl transferases (GalNAc) T1 to 4, which transfer oligosaccharides to threonine and serine residues (marked by bold types), depending on cell type and sequence of preglycosylation. This affects the binding properties of the mAb used in this study. The epitopes, minimally required for recognizing the TR core of MUC1 mucin, are indicated by double arrows. Materials and methods Patients, tissue specimens and diagnosis Morphological analyses were performed on 65 testicular tissue samples of 63 patients (mean age 34 years, range 20–72 years), archived during the period of 1996 to 1999 as routinely formalin- or Bouin-fixed and paraffin-embedded material. In 27 cases orchidectomy was performed for germ cell tumour. The tumour-free tissue was analysed showing spermatogenesis with normal or only slightly altered seminiferous epithelium. Biopsy specimens, taken for routine diagnosis in impaired fertility or obstructive azoospermia, were available in 36 cases. Here the histological findings were pure or combined arrest at the level of spermatogonia (n ⫽ 6), spermatocytes (n ⫽ 17), or spermatids (n ⫽ 5) including focal Sertoli cell-only features (n ⫽ 5), mixed testicular atrophy (n ⫽ 8), and normal spermatogenesis (n ⫽ 11; score 艌8), (Holstein and Schirren, 1983). In addition, normal fresh testicular tissue from three patients orchidectomized for prostatic cancer, small peripheral teratoma, and epididymitis (patients’ ages of 53, 18 and 32 years respectively) were obtained from surgery at 4°C and were either directly processed or immediately snap-frozen in liquid nitrogen. Reverse transcription-polymerase chain reaction (RT–PCR) of tissue extracts and selected spermatocytes after cell picking Consecutive cryostat sections of the frozen samples from two different cases were performed at 5 µm for (i) the histological control of pure testicular tissue with regular appearance of spermatogenesis and absence of rete testis or epididymis, (ii) direct mRNA isolation using Oligo (dT)25-coupled magnetic particles (Dynal, Oslo, Norway), and (iii) RT–PCR of a few selected cells by UV laser-assisted cell picking (UV-LACP). The methods employed were described in detail previously (Pauls et al., 1999; Steger et al., 2000). In brief, cDNA synthesis from mRNA extracts and picked cell profiles was performed using Perkin-Elmer reagents (Perkin-Elmer, Weiterstadt, Germany) in standard equivalents using random hexamers and MuLV reverse 506 transcriptase at 43°C for 75 min. Subsequent PCR was performed using the GeneAmp Kit and AmpliTaq Gold according to the manufacturer’s instructions (Perkin-Elmer). For the detection of MUC1 cDNA, the following primer pair was designed: 5⬘-CCTTTCTTCCTGCTGCTG-3⬘ (forward) and 5⬘-TGGGCACTGAACTTCTCTG-3⬘ (reverse). These primers span intron 1 and an alternative splice site of 27 bp at the 5⬘-end of exon 2 (variants A and B) (Ligtenberg et al., 1990), characteristically leading to two different cDNA products of 118 and 145 bp length. The annealing temperature used was 62°C and, using a GeneAmp PCR system 2400 (PerkinElmer), either 42 PCR cycles for mRNA extracted from single tissue sections or 60 PCR cycles for picked cell profiles were performed. PCR was also performed on genomic DNA, extracted from sections of the same tissues according to the instructions of the applied QlAmp Blood Kit (Qiagen, Hilden, Germany), to exclude the presence of pseudogenes. Controls performed included omission of the MuLV reverse transcriptase from the assay and picking clusters of Leydig cells outside the seminiferous tubules. Monoclonal antibodies, immunohistochemistry and immunoblot The monoclonal antibodies (mAb) HMFG1, HMFG2 and SM3 were used at concentrations of 10 µg/ml each. These mAb recognize closely related epitopes within the VNTR sequence of the MUC1 core protein (Pro-Asp-Thr-Arg, Asp-Thr-Arg, and Pro-Asp-Thr-ArgPro respectively), which are selectively exposed depending on the properties of carbohydrates added at glycosylation sites nearby (Burchell and Taylor-Papadimitriou, 1993). Immunohistochemistry was performed on consecutive 2–3 µm tissue sections, applying microwave pretreatment to restore the proper antigenicity of the fixed and paraffin-embedded tissues, and using the sensitive alkaline phosphatase-anti-alkaline phosphatase (APAAP) technique (Dako, Hamburg, Germany) according to the already published protocol (Pauls et al., 1999). In addition, Western blot analysis was used to confirm the reactivity of mAb. Fresh testis tissue was homogenized MUC1 in human spermatogenesis Figure 2. Molecular detection of MUC1 mucin in tissue extracts of testis (A–C) and in spermatocytes (D). (A) Immunoblot analysis shows two signals in the high molecular weight range (⬎200 kDa) for HMFG1 (lane 1) and HMFG2 (lane 2). Low glycosylated MUC1 variants were not detected by SM3 (lane 3). Lanes 4 and 5: negative controls; lane 6: angiotensin-converting enzyme recognized by CG2 at 170 kDa. (B) Using intron 1-spanning primers, polymerase chain reaction (PCR) of genomic DNA (lanes 4 and 5) produced two very long DNA fragments. Pseudogenes, involving the expected MUC1 cDNA lengths of 118 or 145 bp, were not detected. Lane 3: negative control; lanes 1 and 2: size markers. (C) Reverse transcription (RT)–PCR of mRNA extracts shows the MUC1 cDNA of 118 bp (variant B; Ligtenberg et al., 1990) and the splice variant (variant A) of 145 bp (lanes 3-6). Note that samples from two different individuals have been analysed (lanes 3 and 4: case no. 1; lanes 5 and 6: case no. 2). Lanes 7 to 10: negative controls; lanes 1, 2, 11 and 12: size markers. (D) UV laserassisted cell picking from a seminiferous tubule from case no. 2 (upper left, rectangle indicates the area of interest). A group of spermatocytes is dissected by a UV laser microbeam (upper right, arrows). The cell profiles are selected (below left) by a micromanipulatordriven steel-needle (insert) and are transferred to a reaction tube. RT-PCR performed on six such cell clusters, each containing 10–15 spermatocyte cell profiles (below right, lanes 2-7), detected the 118 bp (variant B) of MUC1 cDNA in most cases. Lane 8: positive control; lanes 9 and 10: negative controls; lanes 1 and 11: size marker. on ice in 20 mmol/l HEPES (pH 7.4), containing 1 mmol/l EDTA, 0.1 mmol/l pepstatin, 10 µmol/l phenylmethylsulphonyl fluoride, and 0.1 mmol/l dithiothreitol (Sigma, Deisenhofen, Germany), and the homogenate was cleared of debris by centrifugation (12 000 g for 15 min at 4°C). The protein samples were boiled in sodium dodecyl sulphate (SDS) buffer containing mercaptoethanol, and analysed on 6% SDS-polyacrylamide gel electrophoresis (PAGE). The gels were transferred to nitrocellulose membranes that were subsequently blocked in Tris buffer containing 5% skimmed milk. The immunoblot was performed in a miniblotter system (MN25; Biometra, Goettingen, Germany), allowing the simultaneous analysis of different mAb in separately formed lanes. Primary antibodies were incubated for 30 min at room temperature and the reaction was visualized by a single-step APAAP procedure using BCIP/NBT (Sigma). For immunohisto- 507 F.E.Franke et al. Figure 3. Immunohistochemical detection of MUC1 mucin in testicular tissue sections with normal spermatogenesis (A–C), mixed tubular atrophy (D and E) and spermatocytic arrest (F and G). (A) HMFG1 revealed only a weak signal in some groups of pachytene spermatocytes (arrows), whereas spermatocytes, displaying an abnormal meiosis, were strongly labelled (arrow head). Some groups of elongated spermatids were also immunoreactive for HMFG1 (upper left). (B) The MUC1 glycosylation variant indicated by HMFG2 was demonstrated in almost all spermatids with the strongest signal occurring in elongated spermatids (consecutive tissue section of A). (C) Intensely HMFG1-labelled spermatocytes with abnormal morphology and missing generations of germ cells on the left-hand side of the seminiferous tubule (detail of A). (D) In mixed tubular atrophy with only qualitatively intact spermatogenesis, HMFG1 revealed a pattern of immunostaining similar to that in tubules with normal spermatogenesis. Certain groups of spermatocytes showed only weak immunostaining (arrows). This was in contrast to the occasional spermatocytes with maturation arrest which demonstrated strong immunoreactivity (arrow head). Compare with A. (E) In mixed tubular atrophy the MUC1 glycosylation variant recognized by HMFG2 was also localized to spermatids (consecutive tissue section of D). Compare with normal spermatogenesis in B. (F) Spermatocytes in tubules with maturation arrest consistently demonstrated strong immunoreactivity for the MUC1 glycosylation variant recognized by HMFG1. (G) In the case of spermatocytic arrest, even the HMFG2 variant became detectable in spermatocytes (consecutive tissue section of F). (A–G) APAAP immunohistochemistry; original magnification (A, B, D–G) ⫻120, (C) ⫻360. chemistry and Western blot analyses, the mAb MR12/53 to rabbit Ig (Dako) served as a negative control, and the mAb CG2 (BMA, Augst, Switzerland) as an internal positive control detecting the somatic isoform of human angiotensin I-converting enzyme (Pauls et al., 1999). 508 Results MUC1 in normal spermatogenesis Both mRNA and protein of MUC1 mucin were found in normal adult testes, histologically devoid of admixtures of rete MUC1 in human spermatogenesis men, the immunohistochemical results were summarized according to the stage of germ cell differentiation (Table I). Table I. MUC1 immunoreactivity in normal and impaired human spermatogenesisa Morphology Analysed cases HMFG1 HMFG2 SM3 Sertoli cells SCO Spermatogonia atrophyc spermatogonial arrest Spermatocytes lepto-/zygotene pachytene spermatocytic arrestd Spermatids round elongated spermatid arrest Spermatozoa 65 5b 65 8 6b – – – – – – – – – – – – – – – 65 53 17 – ⫺/⫹a ⫹⫹⫹ – – ⫺/⫹⫹a – – – 53 48 5b 48 – ⫺/⫹a – – ⫹ ⫹⫹/⫹a ⫹ – – – – aSummarized data, as differences between the testes of different individuals were not observed. However, a variable immunoreactivity was consistently noted in certain groups of differentiating germ cells. This local variability between different seminiferous tubules is indicated by a slash (/). bAnalysed in combined disorders only. cRefers to a reduced number of spermatogonia in mixed tubular atrophy. dSingular spermatocytes of normal spermatogenesis, but showing degeneration and strong HMFG1 reactivity, are not listed. SCO ⫽ Sertoli cell-only characteristics. testis or epididymis. MUC1 expression was demonstrated at both the transcriptional level using RT–PCR of mRNA extracts, and at the translational level applying immunoblot analyses. Whereas Western blot analyses revealed two protein alleles in the high molecular weight range (⬎200 kDa) with HMFG mAb recognizing the VNTR region of MUC1, RT-PCR revealed two alternatively spliced products (Figure 2A–C). In addition, MUC1 cDNA was confirmed by RT–PCR after UV-LACP on selectively picked cell profiles, which indicated specific localization in spermatocytes (Figure 2D), but not Leydig cells (data not shown). This localization was consistent with immunohistochemical data showing that MUC1 protein products occur only in germ cells. Here, the mAb HMFG1 and HMFG2, which recognize similar epitopes dependent on glycosylation, labelled different stages of germ cell differentiation. In all analysed cases, HMFG1 labelled only pachytene spermatocytes and elongated spermatids. A generally weak and, between individual groups of germ cells, variable immunoreactivity was observed with differences between neighbouring seminiferous tubules (Figure 3A). In contrast, immunoreactivity for HMFG2 was consistently found only in spermatids (Figure 3B). Singular groups of spermatocytes, however, were intensely labelled by HMFG1. These cells showed an arrest of maturation and frequently degenerative or abnormal features, and were often directly adjacent to missing generations of germ cells (Figure 3C). Intratubular spermatozoa were not immunoreactive with either of the mAb, and the mAb SM3, recognizing poorly glycosylated MUC1 variants, did not show any immunoreactivity in the testis, as was the case with all negative controls. The pattern of MUC1 immunoreactivity was similar in all analysed cases and in seminiferous tubules exhibiting normal spermatogenesis. Since no differences were observed between the testes of different MUC1 in impaired spermatogenesis As indicated by HMFG1 and HMFG2, MUC1 expression was only found in spermatocytes and spermatids irrespective of the type of spermatogenic disorder (Table I). Cases of mixed tubular atrophy, exhibiting quantitatively reduced spermatogenesis, revealed a MUC1 distribution similar to that in normal spermatogenesis (Figure 3D and E). The strongest immunoreactivity was found in tubules with maturation arrest at the level of spermatocytes. Here, the MUC1 glycosylated variant recognized by HMFG1 was abundant in all arrested cells and this was a constant finding in spermatocytes exposed to this maturation block (Figure 3F). Although normally found only in spermatids (Figure 3B), the MUC1 variant recognized by HMFG2 was also demonstrable in cases with spermatocytic arrest (Figure 3G). Moreover, HMFG1 labelled pachytene spermatocytes more frequently and was enhanced in those seminiferous tubules showing degenerative alterations in the vicinity of germ cell tumours (Figure 4A and B). The immunostaining of both mAb, HMFG1 and HMFG2, was not strictly bound to the cell membranes and was mainly localized to the cytoplasm of spermatocytes or spermatids (Figure 4B and C). SM3 was negative in all cases analysed (Figure 4 D). Discussion In this study we describe the expression of MUC1 mucins in maturing germ cells of the human testis. Since MUC1 and its protein isoforms are generally thought to be expressed only in epithelial cells, until now MUC1 has not been considered to play a role in non-neoplastic testis. Our findings demonstrate MUC1 expression in pachytene spermatocytes and markedly enhanced MUC1 immunoreactivity with maturation arrest. Many studies on MUC1 deal with its expression and subsequent glycosylation in epithelial tumours, including possible strategies for molecular targeting or use as vaccine in tumour therapy (Karsten et al., 1998; Koumarianou et al., 1999). Thus, most of the current understanding of the functional roles of MUC1 proteins comes from tumour and animal models (Lalani et al., 1991; Baruch et al., 1999). Recently, mRNA and protein of MUC1 mucin have been demonstrated in nonepithelial tumour cell lines, contradicting the belief that it is expressed in epithelial cells alone (Oosterkamp et al., 1997). It is still unclear whether non-epithelial tumours express MUC1 aberrantly through gene amplification or merely mirror the expression in those tissues from which they are derived. Chromosomal rearrangement at the 1q21 locus of the MUC1 gene is a possible cause of aberrant expression in some B-cell lymphomas (Dyomin et al., 2000), but MUC1 variants have also been detected in multiple myeloma (Takahashi et al., 1994) and, as reported recently, in regular human T-lymphocytes (Chang et al., 2000). Depending on the assay used, however, MUC1 is found only in low amounts when compared with epithelial cells (Oosterkamp et al., 1997). At least one former study, which screened different human tissues with the HMFG mAb, failed to record MUC1 immuno509 F.E.Franke et al. Figure 4. MUC1 immunoreactivity for the HMFG glycosylation variants in impaired spermatogenesis of seminiferous tubules localized in the vicinity of a germ cell tumour. (A) Spermatocytes were more frequently and intensely labelled by HMFG1 in tubules with impaired spermatogenesis (ISP), as compared with normal spermatogenesis (see Figure 3A). This was demonstrated in the vicinity of a germ cell tumour (GCT, seminoma) and its compression zone showing atrophy, fibrosis (TF), and Sertoli cell-only (SCO) characteristics of the encompassed tubules. (B) Immunoreactivity for the HMFG1 glycosylation variant was not restricted to the spermatocyte cell membrane, but was also found within the cytoplasm of spermatocytes. (C) HMFG2, recognizing the MUC1 glycosylation variant of spermatids, exhibited a cytoplasmic staining pattern in these cells and labelled elongated spermatids in particular. (D) SM3, recognizing only low-glycosylated MUC1 variants, did not show any immunoreactivity in germ cells or in any other cells of the human testis. (A–D) APAAP immunohistochemistry; original magnification (A) ⫻25, (B–D) ⫻480. reactivity in testis and spermatogenesis (Zotter et al., 1988). This difference from our results is probably best explained by the sensitivity of the APAAP technique which we used in this study and the weak signals obtained regarding the majority of cells in normal spermatogenesis. These immunoreactive signals appeared to be largely cytoplasmatic, suggesting that maturing germ cells express a secreted MUC1 isoform rather than a transmembrane one, either due to an alternative splice variant (Williams et al., 1990) or direct cleavage of the cytoplasmic anchor (Boshell et al., 1992). Thus, immunohistochemistry 510 can fail to demonstrate the actual amount of MUC1 mucin, whereas the gross mass escapes detection. Nevertheless, MUC1 immunoreactivity of germ cells has already been reported in transgenic mice, which express the human MUC1 protein and its variants in tissues and cells in a similar pattern as in humans (Peat et al., 1992). In this model, the glycosylation variants of human MUC1 appear in a similar sequence as indicated by HMFG1 in spermatocytes and HMFG2 in spermatids (Peat et al., 1992). This sequence is probably a consequence of the complex O-glycosylation of MUC1 mucin being dependent on MUC1 in human spermatogenesis access to N-acetylgalactoaminyltransferases (GalNAc) T1 to T4 (Bennett et al., 1998). GalNAc are expressed in human spermatogenesis, with the pattern of GalNAc-T3 immunoreactivity strictly limited to early stages of germ cell maturation and ejaculated spermatozoa (Mandel, 1998). In the normal human testis, we confirmed the presence of MUC1 mRNA and protein in tissue extracts and maturing germ cells by different techniques. Marked accumulation of MUC1 immunoreactive products, however, was found in all spermatocytes with maturation arrest. HMFG1 in particular easily detected the pathologically altered seminiferous tubules showing a complete block of terminal germ cell differentiation and, in normal spermatogenesis, those individual spermatocytes prone to degeneration (see Figure 3). The latter corresponds to the rather frequently occurring defects of meiosis, which may lead to abnormal cell forms (Holstein and Eckmann, 1986) but regularly contribute to missing generations of germ cells in normal human spermatogenesis (Johnson et al., 1992). In this context, it is noteworthy that the glycosylation variant of MUC1 recognized by HMFG1 was found more frequently and was consistently expressed in spermatocytes of degenerating tubules located in the vicinity of germ cell tumours (see Figure 4). Both its abundance in spermatogenic arrest and its rather gradual accumulation in spermatocytes exposed to irritation, suggest the HMFG1 variant of MUC1 is a useful tool for the study of impaired human spermatogenesis. The pattern in which the HMFG1 variant was detected in normal and impaired spermatogenesis may reflect a functional meaning. MUC1 isoforms, containing the O-glycosylated VNTR regions, are long and rigid molecules mainly transferring anti-adhesive cell properties (Ligtenberg et al., 1992). Whether secreted to form a viscous gel at the cellular surface or membrane-bound, any accumulation of glycosylated MUC1 proteins would enlarge the space between Sertoli- and expressing germ cells, interfere with physiological cell functions, and probably accelerate the detachment and transit of delayed or abnormal germ cells. In such a scenario, the accumulation of anti-adhesive MUC1 molecules would serve as a safeguard for the exact timing of maturing germ cells, protecting the germinal epithelium from occasional meiotic defects and retention of abnormally differentiated germ cells. Considering the variability of MUC1 isoforms, its glycosylation properties, and the ability of short variants to act as cognate binding receptors for the secreted MUC1 protein (Baruch et al., 1999), MUC1 may have other potential functions. Mucins or mucin-like molecules have already been reported on spermatozoa (D’Cruz et al., 1996) and in the epididymis (Zotter et al., 1988; Osterhoff et al., 1997), the female genital tract and ovary (Lagow et al, 1999). Thus, MUC1 isoforms expressed in human spermatids (see Figure 3B), contributing to the glycocalyx of the sperm surface (Schroter et al., 1999), may, in addition, be involved in maturation, transport, and ovum recognition of spermatozoa. Acknowledgements The authors thank Professor Andreas Schulz and Professor Wolfgang Weidner, Heads of the Departments of Pathology and Urology of the Justus-Liebig University of Giessen, and the members of the SFB 547 for constant support during the course of this study. Furthermore, we like to thank Dr Alison Kraus, Dr Klaus Steger and Dr Joachim Woenckhaus for helpful comments on improving the manuscript. References Baruch, A., Hartmann, M., Yoeli, M. et al. (1999) The breast cancer-associated MUC1 gene generates both a receptor and its cognate binding protein. Cancer Res., 59, 1552–1561. Bennett, E.P., Hassan, H., Mandel, U. et al. (1998) Cloning of a human UDP-N-acetyl-alpha-D-galactosamine:polypeptide N-acetylgalactosaminyltransferase that complements other GalNAc-transferases in complete Oglycosylation of the MUC1 tandem repeat. J. Biol. Chem., 273, 30472– 30481. Boshell, M., Lalani, E.N., Pemberton, L. et al. (1992) The product of the human MUC1 gene when secreted by mouse cells transfected with the fulllength cDNA lacks the cytoplasmic tail. Biochem. Biophys. Res. Commun., 185, 1–8. Burchell, J. and Taylor-Papadimitriou, J. (1993) Effect of modification of carbohydrate side chains on the reactivity of antibodies with core-protein epitopes of the MUC1 gene product. Epithelial Cell Biol., 2, 155–162. Chang, J.F., Zhao, H.L., Phillips, J. and Greenburg, G. (2000) The epithelial mucin, MUC1, is expressed on resting T lymphocytes and can function as a negative regulator of T cell activation. Cell. Immunol., 201, 83–88. D’Cruz, O.J., Dunn, T.S., Pichan, P. et al. (1996) Antigenic cross-reactivity of human tracheal mucin with human sperm and trophoblasts correlates with the expression of mucin 8 gene messenger ribonucleic acid in reproductive tract tissues. Fertil. Steril., 66, 316–326. Dyomin, V.G., Palanisamy, N., Lloyd, K.O. et al. (2000) MUC1 is activated in a B-cell lymphoma by the t(1;14) (q21;q32) translocation and is rearranged and amplified in B-cell lymphoma subsets. Blood, 95, 2666–2671. Gendler, S.J., Lancaster, C.A., Taylor-Papadimitriou, J. et al. (1990) Molecular cloning and expression of human tumor-associated polymorphic epithelial mucin. J. Biol. Chem., 265, 15286–15293. Ho, S.B., Niehans, G.A., Lyftogt, C. et al. (1993) Heterogeneity of mucin gene expression in normal and neoplastic tissues. Cancer Res., 53, 641–651. Holstein, A.F. and Schirren, C. (1983) Histological evaluation of testicular biopsies. In Schirren, C. and Holstein, A.F. (eds), Fortschritte der Andrologie 8, Grosse Verlag, Berlin, pp. 108–117. Holstein, A.F. and Eckmann, C. (1986) Megalospermatocytes: indicators of disturbed meiosis in man. Andrologia, 18, 601–609. Johnson, L., Chaturvedi, P.K. and Williams, J.D. (1992) Missing generations of spermatocytes and spermatids in seminiferous epithelium contribute to low efficiency of spermatogenesis in humans. Biol. Reprod., 47, 1091–1098. Karsten, U., Diotel, C., Klich, G. et al. (1998) Enhanced binding of antibodies to the DTR motif of MUC1 tandem repeat peptide is mediated by sitespecific glycosylation. Cancer Res., 58, 2541–2549. Koumarianou, A.A., Hudson, M., Williams, R. et al. (1999) Development of a novel bi-specific monoclonal antibody approach for tumour targeting. Br. J. Cancer, 81, 431–439. Lagow, E., DeSouza, M.M. and Carson, D.D. (1999) Mammalian reproductive tract mucins. Hum. Reprod. Update, 5, 280–292. Lalani, E.N., Berdichevsky, F., Boshell, M. et al. (1991) Expression of the gene coding for a human mucin in mouse mammary tumor cells can affect their tumorigenicity. J. Biol. Chem., 266, 15420–15426. Ligtenberg, M.J., Vos, H.L., Gennissen, A.M. and Hilkens, J. (1990) Episialin, a carcinoma-associated mucin, is generated by a polymorphic gene encoding splice variants with alternative amino termini. J. Biol. Chem., 265, 5573– 5578. Ligtenberg, M.J., Buijs, F., Vos, H.L. and Hilkens, J. (1992) Suppression of cellular aggregation by high levels of episialin. Cancer Res., 52, 2318–2324. Mandel, U. (1998) 4th Copenhagen Workshop on ‘Carcinoma In Situ and Cancer of the Testis’, Copenhagen May 18–21, 1997, organized by Rajpert-De Meyts, E., Grigor, K.M. and Skakkebeak, N.E. APMIS, 106, p. 126. Muller, S., Goletz, S., Packer, N. et al. (1997) Localization of O-glycosylation sites on glycopeptide fragments from lactation-associated MUC1. All putative sites within the tandem repeat are glycosylation targets in vivo. J. Biol. Chem., 272, 24780–24793. Oosterkamp, H.M., Scheiner, L., Stefanova, M.C. et al. (1997) Comparison of MUC-1 mucin expression in epithelial and non-epithelial cancer cell 511 F.E.Franke et al. lines and demonstration of a new short variant form (MUC-1/Z). Int. J. Cancer, 72, 87–94. Osterhoff, C., Ivell, R. and Kirchhoff, C. (1997) Cloning of a human epididymis-specific mRNA, HE6, encoding a novel member of the seven transmembrane-domain receptor superfamily. DNA Cell Biol., 16, 379–389. Pauls, K., Fink, L. and Franke, F.E. (1999) Angiotensin-converting enzyme (CD143) in neoplastic germ cells. Lab. Invest., 79, 1425–1435. Peat, N., Gendler, S.J., Lalani, E.N. et al. (1992) Tissue-specific expression of a human polymorphic epithelial mucin (MUC1) in transgenic mice. Cancer Res., 52, 1954–1960. Pratt, W.S., Islam, I. and Swallow, D.M. (1996) Two additional polymorphisms within the hypervariable MUC1 gene: association of alleles either side of the VNTR region. Ann. Hum. Genet., 60, 21–28. Quin, R.J. and McGuckin, M.A. (2000) Phosphorylation of the cytoplasmic domain of the MUC1 mucin correlates with changes in cell–cell adhesion. Int. J. Cancer, 87, 499–506. Schroter, S., Osterhoff, C., McArdle, W. and Ivell, R. (1999) The glycocalyx of the sperm surface. Hum. Reprod. Update, 5, 302–313. Steger, K., Pauls, K., Klonisch, T. et al. (2000) Expression of protamine-1 and -2 mRNA during human spermiogenesis. Mol. Hum. Reprod., 6, 219–225. 512 Takahashi, T., Makiguchi, Y., Hinoda, Y. et al. (1994) Expression of MUC1 on myeloma cells and induction of HLA-unrestricted CTL against MUC1 from a multiple myeloma patient. J. Immunol., 153, 2102–2109. Williams, C.J., Wreschner, D.H., Tanaka, A. et al. (1990) Multiple protein forms of the human breast tumor-associated epithelial membrane antigen (EMA) are generated by alternative splicing and induced by hormonal stimulation. Biochem. Biophys. Res. Commun., 170, 1331–1338. Williams, S.J., McGuckin, M.A., Gotley, D.C. et al. (1999) Two novel mucin genes down-regulated in colorectal cancer identified by differential display. Cancer Res., 59, 4083–4089. Zotter, S., Hageman, P.C., Lossnitzer, A. et al. (1988) Tissue and tumor distribution of human polymorphic epithelial mucin. Cancer Rev., 11–12, 55–101. Zrihan-Licht, S., Vos, H.L., Baruch, A. et al. (1994a) Characterization and molecular cloning of a novel MUC1 protein, devoid of tandem repeats, expressed in human breast cancer tissue. Eur. J. Biochem., 224, 787–795. Zrihan-Licht, S., Baruch, A., Elroy-Stein, O. et al. (1994b) Tyrosine phosphorylation of the MUC1 breast cancer membrane proteins. Cytokine receptor-like molecules. FEBS Lett., 356, 130–136. Received on November 2, 2000; accepted on March 16, 2001