Journal o/Reproduettve Immunology, 20 (1991) 15--26 EIsewer Sc~entlfic Pubhshers Ireland Ltd
AJ-p97: A novel antigen of the human sperm tail fibrous sheath detected by a neurofilament monoclonal antibody A. J a s s i m Department of lmmunology, The London Hospttal Medical College, Tmner Street, London E1 2AD ( U K ) (Accepted for pubhcauon 28 November 1990)
Summary Using indirect immunofluorescence (I1F), the RT97 anti-neurofilament monoclonal antibody (MoAb) detected an intracellular antigen an the principal piece of human ejaculated sperm tails. Its locahsation to the tail fibrous sheath (FS) was confirmed by immunoelectron microscopy (IEM), which showed the binding of the gold particles to the outer FS surface. During spermatogenesis the antigen was first expressed on the spermatid FS, and its expression was continued on ejaculated mature sperm. In Western blotting of sperm lysates, the RT97 reacted with a 97 kDa protein (A J-p97) which lacked disulphide bonds. This antigen was not detected on mouse or rat sperm tail FS, suggesting a sequence dwergence of the A J-p97 during evolution. The significance of these results and the relationship of A J-p97 to neurofilaments are discussed, together with the use of the antibody as a probe for the structural dissection of the FS and for analysing the molecular events that take place during spermiogenesis, especially those involved in sperm tail morphogenesis. Key words: fibrous sheath," sperm," sperm tail," monoclonal antibody; tmmunoelectron microscopy; neurofilaments.
Introduction The tails o f m a m m a l i a n sperm have complex ultrastructures. Their central axonemes are s u r r o u n d e d by out er dense fibres which are encompassed by Correspondenee to A Jasslm, Dept of Immunology, The London Hospltal Medical College, Turner Street, London El 2AD, U K 0165-0378/91/$03.50 © 1991 Elsevier Scientific Pubhshers Ireland Ltd. Published and Printed in Ireland
the FS in the principal piece and the mitochondrial sheath in the middle piece. The FS, which is unique to the sperm tail, is thought to provide some support to the flagella during sperm movement through the female genital tract (Fawcett, 1970; Phillips, 1972). Ultrastructurally, the FS is made up of two longitudinal columns running alongside the tail with a large number of connecting transverse ribs. Developmental studies of rat sperm tail FS showed that during spermiogenesis FS formation proceeds from the distal towards the proximal end of the principal piece (Irons and Clermont, 1982). The longitudinal columns start their assembly at step 2 spermatid and are complete at step 17. The formation of the transverse ribs, on the other hand, starts at step 11 and ends at step 15, suggesting that their development is independent of that of the longitudinal columns (Irons and Clermont, 1982). In humans, the FS appears at the Sc spermatid stage, which is the fourth step of spermiogenesis (De Kretser, 1969). Biochemical analysis of isolated FS showed a major 80 kDa protein in rat (Olson et al., 1976) and a 74 kDa product in mouse (O'Brien and Bellve, 1980), although later studies demonstrated several polypeptides in the protein extracts of rat FS (Oko, 1988). MoAbs have been used to analyse the molecular structure of rodent FS, and its ontogenic development during spermiogenesis (Sakai et al., 1986; Fenderson et al., 1988). In humans, no information is available regarding the biochemical structure of the FS or its antigenic components. The present work was therefore undertaken and herein is described the identification of a novel human FS antigen, A J-p97, which is the first to be biochemically characterised. This antigen was detected using the RT97 anti-neurofilament MoAb (Wood and Anderton, 1981). Materials and Methods
Specimens Sperm were obtained from 7 normospermac donors, while nucleated cells other than sperm (NCOS) were recovered from 5 oligospermic donors. Samples were handled within 1 h of donation and the NCOS were separated on ficoll/triosil density gradients (Jassim and Festenstein, 1987a). Normal human adult testes were obtained from cadaveric donors and snap-frozen m liquid nitrogen. Rodent sperm were isolated from testes of adult BALB/c and BALB/c D2 Ma inbred mice and Sprague--Dawley randomly bred rats.
Monoclonal antibodies ( MoAbs) The MoAbs used in this study included the RT97 anti-200 kDa neurofilaments (Wood and Anderton, 1981), GDA-J/F7 anti-germ cells and monocytes (Jasslm, 1990), G A P 8.3 anti-leucocyte-CD45 (Berger et al.,
1981), W6/32 anti-Class I (Parham et al., 1979) and L227 anti-Class II HLA antigens (Lampson and Levy, 1980) and LP34 anti-cytokeratins (I. Leigh). All antibodies were supernatants with the exception of RT97 which was from lyophilised ascites. Following its reconstitution, the RT97 MoAb was diluted 1:120 with R P M I medium containing 10% fetal calf serum. Fifty microlitres of undiluted supernatant antibodies or diluted RT97 were used in the IIF testing. Initially, RT97 had been raised by immunising mice with rat brain Triton X-100 insoluble material (Wood and Anderton, 1981).
IIF testing Standard techniques were used to stain sperm suspensions, dried NCOS or frozen sections with MoAbs. These have been described previously (Jassim and Festenstein, 1987a,b). Demembranation of sperm for IIF testing Sperm were demembranated by either repeated freezing and thawing (Jassim and Festenstein, 1987b) or by incubation with 1% Nonidet-P40 (NP40) or 1% Triton X-100 for 1 h at 4°C. Cells were pelleted by centrifugataon and the supernatants discarded. Sperm were washed twice with phosphate buffered saline (PBS) and screened with MoAbs using IIF. IEM techmque Sperm obtained from a normospermic donor were washed twice with PBS. Cells were solubilised in 1 ml of 1% NP40 for 1 h at 4°C, and aggregated sperm removed by centrifugation at 1000 × g (Wifug 5500 rev./min) for 10 s. Non-sedimented sperm (6--8 × 106) were washed twice with 20 mM Tris buffered saline containing 1% bovine serum albumin, pH 8.2 (this washing buffer was prepared according to Janssen Biotech instructions). Sperm were incubated with 200 IA RT97 MoAb or RPMI medium (as a negative control) for 1.5 h at 37°C. Cells were washed twice with the same buffer and then incubated with 40 ~1 colloidal gold goat anti-mouse IgG (10 nm gold particles, Janssen Biotech NV, Code No. RPN425) for 40 min at 37°C, with gentle agitation every 10 min. Samples were washed three times and cells processed for electron microscopy by fixing in 2.5% gluteraldehyde and 2% paraformaldehyde in 0.1 M cacodylate buffer. Following thorough washing, the sperm were postfixed in 1% osmium tetroxide in water, dehydrated in alcohol, stained "en bloc" with 1% phosphotungstic acid in alcohol and embedded in L o n d o n Resin White. Sections were examined using a Phillips 400 transmission electron microscope. Western blotting Sperm obtained from normospermic donors were washed twice with PBS. Cells were solubilised with 1% NP40 or 1% Triton X-100 for 1 h at 4°C or
with 1% sodium deoxycholate for 1 h at 37°C. Sperm lysates from the different detergents were mixed with reducing or non-reducing sample buffers and run through a 7.5% gel using standard techniques of SDS-PAGE electrophoresis. The proteins were blotted onto nitrocellulose paper (Towbin et al., 1979) and screened with the RT97 MoAb using a standard immunoperoxidase technique. Diaminobenzidine was used as substrate. Results
Reaction of RT97 MoAb with ejaculated human sperm Using IIF, the RT97 MoAb reacted with less than 1% of freshly isolated sperm. These were immotile and probably dead (data not shown). However, the antibody stained all the frozen-thawed sperm (data not shown) as well as those dried onto slides (Fig. l). In all experiments the reaction was restricted to the principal piece of the tail (Fig. 1).
Ontogeny of A J-p97 during spermatogenesis Frozen sections of normal adult human testes were tested with RT97 using IIF. The antibody reacted with sperm tails (Fig. 2) and also produced weak fibrillar nuclear staining of spermatocytes/spermatids (data not shown).
Reaction of RT97 with NCOS of oligospermic donors The germinal cell origin of NCOS was first ascertained using a panel of tissue-specific MoAbs. These cells reacted with the germ cell specific MoAb,
Fig. 1. Indirect immunofluorescence test of dried human sperm with RT97 MoAb. The antibody stained the principal piece of the tail only (a, phase; b, fluorescence microscopy) x 400.
Fig, 2. Indirect immunofluorescence test of frozen sections of h u m a n adult testes with RT97 MoAb. The antibody strongly reacted with the tails but produced variable fibrillar nuclear staining of the spermatocytes/spermatids (data not shown) ×400.
GDA-J/F7, but showed no reaction with antibodies against leucocytic, epithelial (cytokeratin) or Class I or II HLA antigens. Following identification as germ cells, NCOS were screened with RT97 using IIF. The antibody produced variable fibrillar staining of some germ cell nuclei (Fig. 3) in addition to its reaction with the principal piece of mature sperm tails.
Biochemical characterisation of A J-p97 Western blotting of NP40 or Triton-treated sperm lysates failed to demonstrate any specific reaction with RT97 (data not shown). More than 90% of these detergent-treated sperm retained their immunofluorescence upon testing with RT97 (Fig. 4). Western blotting of sperm lysed with sodium deoxycholate, on the other hand, revealed a 97 kDa protein. This was observed in both reducing and non-reducing gels although the intensity of the reaction was stronger in the reducing gels (Fig. 5). Ultrastructural localisation of A J-p97 in sperm NP40-treated sperm were tested with RT97 using IEM. In the demembranated sperm the antibody stained the FS of 60--70% of tails. Gold par-
Fig. 3. Indirect immunofluorescence test of dried h u m a n NCOS with RT97 MoAb. In addition to staining the principal piece of the sperm tails, the antibody also produced fibrillar nuclear staining of the germ cells (a, phase; b, fluorescence microscopy) x 400.
ticles were seen localised only to the outer FS surface (Fig. 6). N o reaction was observed with the sperm head, mitochondrial sheath, axonemal ultrastructures or outer dense fibres. Sperm incubated with RPMI as negative controls also showed no binding (data not shown).
Fig. 4. Indirect immunofluorescence test of NP40-treated h u m a n sperm with RT97 MoAb. Note that the antibody still stains the principal piece (a, phase; b, fluorescence microscopy) x 400.
Fig. 5. Western blotting of human sperm lysates with RT97 MoAb. The antibody recognised a 97 kDa protein (a,c). The intensity of the reaction was stronger under reducing (a) than non-reducing conditions (c); (b,d) are the corresponding negative controls.
Reaction of RT97 MoAb with demembranated rodent sperm Frozen-thawed as well as NP40-treated rodent sperm were tested with RT97 using IIF. The antibody did not react with the principal piece o f mouse (data not shown) or rat sperm (Fig. 7). Instead, RT97 stained the middle piece of rat sperm only (Fig. 7).
Fig. 6. lmmunogold electron microscopy of detergent-treated h u m a n sperm with RT97 MoAb. The antibody stained the fibrous sheath only (a, longitudinal section through the principal piece). In crosssections (b,c) the gold particles were localised to the outer surface of the fibrous sheath. In (c), the top section was not stained, probably due to solubilisation of the antigen; note its ill-defined uhrastructure. (a, x41,125; b, ×68,750; c, x67,850).
Fig. 7. Indirect immunofluorescence test of NP40-treated rat sperm with RT97 MoAb. The antibody stained the middle piece instead of the principal piece. (a, phase; b, fluorescence microscopy) x 400.
Discussion The A J-p97 protein described herein is clearly an FS antigen. This is based on: (1) its intracellular localisation - - the RT97 antibody did not react with viable sperm but showed strong reactions with dried or demembranated spermatozoa (frozen-thawed or detergent-treated); (2) its limited expression on the principal piece of the sperm tail; (3) IEM which confirmed its localisation to the FS. The binding of gold particles to the FS of only 60--70% of sperm tails was probably due to partial solubilisation of the antigen by the detergent. In IEM the restricted distribution of gold particles to the FS outer surface suggests the outward orientation of the antigenic epitope and possible polarity in the antigenic structure of the two FS surfaces. The mechanism and biological importance of this interesting phenomenon requires further investigation. It may be related to the type of filaments attached to the two FS surfaces. It is known that the outer FS surface is linked to the sperm plasmalemma by cross-filaments which are different from those linking the inner FS surface to the outer dense fibres (Escalier, 1984). Biochemically, A J-p97 is a 97 kDa protein, insoluble or only partially soluble in 1% Triton or 1% NP40, and seems to lack disulphide bonds as its electrophoretic mobility in gels was not affected by reducing reagents. The stronger reaction of RT97 with the "reduced" compared with the "nonreduced" antigen was probably due to the effect of the reducing agent on the solubilisation of the FS; such reagents are known to enhance FS solubilisation (Olson et al., 1976). The same antigen was also recovered from purified human FS isolated by urea extraction (Gillott and Jassim, unpublished data) using a technique similar to that described by Olson et al. (1976). Although A J-p97 was detected using the RT97 anti-neurofilament MoAb, its 97 kDa molecular weight rules out identity with the 200 kDa neurofilament subunit recognised by the same antibody in nervous tissue (Wood and Anderton, 1981). Furthermore, it as unlikely that A J-p97 is either a precusor or a degradation product of the 200 kDa neurofilament subunlt as germ cells lack intermediate filaments, including neurofilaments (Franke et al., 1981; Battifora et al., 1984; Miettinen et al., 1985). The reactiyity of RT97 with both A J-p97 and the 200 kDa neurofilament is therefore a cross-reaction, probably due to sharing of a homologous sequence or a conformational epitope (Lane and Koprowski, 1982). Amino acid sequencing of AJ-p97 could help solve this problem. Cross-reactions of MoAbs are not u n c o m m o n (Lane and Koprowski, 1982; Jassim et al., 1989b) and, previously, RT97 was reported to cross react with nuclear histones, staining the nuclei of various tissues, including brain, liver and heart, as well as nuclei of cultured cells, e.g. mouse fibroblasts, human embryo lung and rat cerebellum (Wood et al., 1985). This cross-reactivity with histones was probably responsible for the
nuclear staining of the testicular germ cells and ejaculated NCOS. The failure of RT97 to stain sperm nuclei could be due to replacement of their histones by protamines during nuclear condensation. Ontogenic studies showed that during spermatogenesis, A J-p97 appeared first on developed FS (seen as linear immunofluorescence) and its expression was continued on mature sperm. The antigen was not detectable in the cytoplasm or on the surface of spermatogonia, spermatocytes or spermatids (apart from their FS). This was also demonstrated in the NCOS which were germinal in origin as shown by their reactivity with GDA-J/F7 MoAb (Jassim, 1990) and their lack of epithelial and leucocytic markers including the Class I and II HLA antigens (Jassim and Festenstein, 1987a; Jassim et al., 1989a). Reaction of RT97 with the developed FS indicates either (1) posttranslational modification of the native A J-p97 after the assembly of the FS framework, e.g. glycosylation or phosphorylation; in fact, RT97 reacts with the neurofilaments only after phosphorylation (Haugh et al., 1986) or (2) an addition of a cytoplasmic antigen to the FS framework. In this case, however, the antibody recognises the antigen in its bound form only, as A J-p97 is not detectable in the cytoplasm. This is in contrast to the FS antigens reported in rodents. For instance, in mice, the K32 MoAb recognises an antigen which first appears in the cytoplasm of stage 14 and 15 spermatids, and later becomes localised to the FS at stage 16 (Sakaa et al., 1986). Similar results were also obtained in the rat using the ATC MoAb which recognises a 67 kDa protein (Fenderson et al., 1988). In these instances, the cytoplasmic antigen was probably utilised in the final assembly of the FS. Although the biological importance of the A J-p97 has yet to be determined, its late appearance on the FS and continued expression on mature sperm (seen on ejaculated sperm) suggests that the antigen may be involved in stabilising the FS structural framework. Phylogenetically, RT97 was previously reported to react with neurofilaments of all vertebrates, including man, mouse, rat and chick, and to cross-react with nuclear histones of several species (Wood et al., 1985). In the present study, however, RT97 showed species specificity for human sperm tail FS and &d not react with those of mouse or rat. This suggests sequence divergence of A J-p97 during evolution. Investigation of sperm from other species would help to establish the time of appearance of A J-p97 epitope. The restricted reactivity of RT97 with human sperm FS also argues against the possibility that A J-p97 and the 200 kDa neurofilament subunit are structurally related (discussed previously). Its reaction with the middle piece of the rat sperm, however, may be another cross-reaction, in this case with a mitochondrial antigen. Recently, a MoAb, GDA-J/F3, which recognises an intracellular antigen inside the principal piece (Jassim and Festenstein, 1987b) has been further
characterised and found to react with a human sperm FS antigen (Jassim et al., 1990a). This antigen, like A J-p97, appears on developed FS, shows species specificity for human sperm and has a similar IEM ultrastructural distribution. However, without biochemical characterisation, the exact relationship of the GDA-J/F3 antigen to the A J-p97 remains unknown. Previously, the GDA-J/F3 MoAb was found useful in the detection of abnormal germ cells in human semen, including those with tail aprotrusion (Jassim and Festenstein, 1988) and tail agenesis (Jassim et al., 1991). Similarly, RT97 detected spermatids with aprotruded tails (unpublished data) and the antibody could therefore be used for their identification. In conclusion, the RT97 MoAb is a useful probe for the molecular analysis of human sperm FS and its anomalies. Purification and sequencing of A Jp97 are currently in progress to determine any sequence homology with the 200 kDa neurofilament but, more importantly, to produce synthetic oligonucleotides for the isolation of the coding gene. Since A J-p97 appears first during the haploid phase of spermatogenesis, the understanding of the molecular biology of its coding gene would enlighten our understanding of the timing of its transcription, whether meiotic or post-meiotic. Also, it would help to assess the regulatory mechanisms involved in the RNA translation and any post-translational modification (for reviews, see Bellve and O'Brien, 1983; Handel, 1987).
Acknowledgements The author is grateful to Dr J.N. Wood (Sandoz Institute, London) for providing the RT97 MoAb and to D. Auger (Oral Pathology) for helping with the electron microscopy. The Leverhulme Trust is acknowledged for the financial support.
References Batnfora, H , Shelbam, K , Tubbs, R R , Koplnskl, M I and Sum T - T (1984) Antlkeratln antibodies in tumour diagnosis Cancer 54, 843 848 Bellve, A.R and O'Brlen, D A (1983) The mammalian spermatozoon Structure and temporal assembly In Mechamsm and Control of Ammal Fertlllsanon (Hartmann, J F , ed ). pp 55--137, Academic Press, New York Berger, A E., David, J E and Cresswell, P (1981) A human leukocyte antigen identified by a monoclonal antibody Hum. Immunol 3, 231--245 DeKretser, D M. (1969) Ultrastructural features of human spermlogenesls Z Zellforsch 98, 477--505 Escaher, D (1984) The cytoplasmic matrix of the human spermatozoon Cross-filaments hnk the various cell components. Btol Cell 51, 347--364 Fawcett, D W. (1970) A comparative view of sperm ultrastructure Blol Reprod Suppl 2, 90 -127 Fenderson, B A , Toshamon, K , Muller, C.H, Lane, T.F and Eddy, E M (1988) Identification of a protern m the fibrous sheath of the sperm flagellum. Blol Reprod 38, 345--357
26 Frank. W M , Schmld, E , Schiller, D L . Winter, S , Jarasch, E D . Moll, R , Denk, H , Jackson, B W and Illmensee, K (1981) Differentiation-related patterns of expression of proteins of intermedtate-slze filaments in tissues and cultured cells Cold Spring Harbor Syrup Quant Blol 46, 4 3 1 ~ 4 5 3 Handel, M.A (1987) Genetic control of spermatogenesls in mice Results Probl Cell Differ 15, 1--62 Haugh, M C , Probst, A , Ulrich, J , Kahn, J and Anderton, B H (1986) Alzhelmer neurofibrlllary tangles contain phosphorylated and hidden neurofilament epltopes J Neurol Neurosurg Psyschlat 49, 1213--1220 Irons, M J. and Clermont, Y (1982) Klnehcs of fibrous sheath formation in the rat spermatld A m J Anat 165, 121--130. Jasslm. A (1990) GDA-J/F7 monoclonal antibody A new marker for sperm cell precursors m h u m a n semen J Reprod Immunol 18. 123--137 Jasslm, A and Festensteln, H (1987a) Immunological and morphological characterlsatlon of nucleated cells other than sperm m semen of ohgospermlc donors J Reprod lmmunol 11, 77--89 Jasslm, A and Festensteln, H (1987b) Molecular dissection of h u m a n testlcular germ cell differentiation with monoclonal antibodies J Reprod lmmunol 12, 1 7 3 - 1 8 9 Jasslm, A and Festensteln, H (1988) Identification of spermatlds with aprotruded tails in h u m a n semen by monoclonal antibodies and electron microscopy In. Molecular and Cellular Endocrinology of the Testis (Cooke, B A and Sharpe, R M , eds ), pp 203--208, Ra'~en Press. New York Jasslm, A , OIher, W , Payne, A , Blro. A , Oliver, R T D and Festenstem, H (1989a) Analysis of HLA antigens on germ cells in h u m a n semen Eur J Immunol 19, 1215--1220 Jasslm, A , Payne, A , Oliver, R T D and Festenstem, H (1989b) Charactemsatlon of h u m a n male germ cells with monoclonal antibodies against mltogen-actlvated lymphocyte antigens and C D markers Tissue Antigens 33, 106 Jasslm, A , Auger. D , Ohvel, R T D and Sachs, J A (1990) GDA-J/F3 monoclonal antibody as a novel probe for h u m a n sperm taft fibrous sheath and its anomahes H u m Reprod 5, 990--996 Jasslm, A , Oliver, R T D , Blandy, J P and Festenstem, H ( 1991 ) Agenesls of h u m a n sperm tall revealed with GDA-J/F3 monoclonal antibody In The Treatment of Infertility (Boutaleb, Y and Gzouh, A . eds ) Parthenon Pubhshmg, U K , m press Lampson, L A and Levy, R (1980) Two populations of la-hke molecules on a h u m a n t] cell line J lmmunol 125, 293--299 Lane, D and Koprowskl, H (1982) Molecular recogmtton and the future of monoclonal antibodies Nature 296, 200--202 Mlettlnen, M , Vwtanen. I and Talerman. A (1985) Intermediate filament proteins m h u m a n testis and testlcular germ-cell tumours A m J Pathol 120, 402--410 O'Brlen. D A and Bellve, A R (1980) Protein constituents of the mouse spermatozoon 1 An electrophoretlc charactemsatlon Dev Bzol 75. 3 8 6 4 0 4 Oko, R (1988) Comparative analysis of proteins from the fibrous sheath and outer dense fibres of rat spermatozoa Blol Reprod 39, 169--182 Olson, G E , Hamilton, D W and Fawcett, D W (1976) Isolation and characterlsat~on of the fibrous sheath of rat epadldymal spermatozoa Blol Reprod 14, 517--530 Parham, P , Barnstable, C J and Bodmer, W F (1979) Use of a monoclonal antibody (W6/32) in structural studies of HLA-A.B and C antigens J Immunol 123, 342--349 Phllhps, D M (1972) Comparative analysis of mammalian sperm mobility J Cell Blol 53. 561--573 Sakm, Y , Koyama, Y - 1 , Fujlmoto, H , Nakamoto, T and Yamashma, S (1986) Immunocytochemlcal study of fibrous sheath formation m mouse spermlogenes~s using a monoclonal antibody Anat Rec 215, 119--126 Towbln, H , Staehehn, T and Gordon, J (1979) Electrophoretlc transfer of proteins from polyacrylamlde gels to nitrocellulose sheets Procedure and some apphcatlons Proc Natl. Acad Scl U S A 76, 4350~354 Wood, J N., Lathangue, N B , McLachlan, D R , Smath, B J , Anderton, B H and Dowdlng, A J (1985) C h r o m a t m proteins share antigenic determinants with neurofilaments J Neurochem 44, 149--154 Wood, J N, and Anderton, B H (1981) Monoclonal antibodies to m a m m a h a n neurofilaments BlOSCl Rep 1, 263--268