MOLECULAR REPRODUCTION AND DEVELOPMENT 29:347-356 (1991)
Alterations in Distribution of Surface and Intracellular Antigens During Epididymal Maturation of Rat Spermatozoa DAVID M. PHILLIPS,’ ROY JONES: AND RUTH SHALG13 ‘The Population Council, New York, New York; 2Department of Molecular Embryology, AFRC Institute of Animal Physiology and Genetics Research, Cambridge, United Kingdom; 3Department of Embryology and Teratology, Tel Aviv University Medical School, Ramat Aviv, Israel
ABSTRACT The surface membrane of mammalian spermatozoa is known to undergo considerable conformational and organizational changes during epididyma1 maturation. However, much less is known about remodelling of intracellular membranes. In this communication we have used specific immunological markers to study the behavior of several antigens both on and within rat spermatozoa as they mature in the epididymis. Four monoclonal antibodies (McAbs)designated 5B1, 165,2D6, and 156 were used to probe testicular and caput and cauda epididymal spermatozoa by indirect immunofluorescence and immunogold labeling techniques. None of the McAbs bound to testicular spermatozoa; in all cases, they became reactive only on spermatozoa which had reached the caput epididymis. McAb 551 was restricted to the outer acrosomal membrane (OAM) of the acrosomal cap domain. The epitope first appeared on antigen(s) with molecular mass (Mr) of -200 kDa in immature spermatozoa, but later in mature spermatozoa the antigen(s)had Mr of -160 kDa. The antigen(s)recognized by 1B5 McAb on the other hand was initially distributed over the OAM of the entire acrosomal domain (cap + equatorial segment), but during maturation it became progressively more restricted in area until in cauda spermatozoa only the anterior tip of the OAM bound the McAb. McAb 2D6 also bound to the entire OAM and acrosomal contents of caput spermatozoa, but, unlike 5B1 and 1B5 McAbs, reactivity was transient. That is, staining was first detected in caput spermatozoa but then disappeared in corpus and cauda spermatozoa. In contrast to all of the above, 1B6 McAb bound to the surface membrane overlying the entire head domain of caput spermatozoa, but during maturation it became restricted to the postacrosomal domain. These results indicate that, in addition to remodeling of the surface membrane during epididymal maturation, extensive processing of intracellular membrane antigens also takes place and that it is very active within the acrosome. The nature of these intracellular processing events remains to be elucidated, but they may have important consequences for membrane fusion and cell recognition phenomena during fertilization. Key Words: Sperm maturation, lntracellular membranes, Outer acrosomal membrane
0 1991 WILEY-LISS, INC.
INTRODUCTION Mammalian spermatozoa undergo numerous physiological, morphological, and biochemical “maturation” changes as they pass through the epididymis (Bedford, 1975; Amann, 1987). Some of the most often studied aspects of sperm maturation have been the changes in the sperm surface which accompany epididymal transit (reviewed by Jones et al., 1990). Early investigations employed relatively nonspecific markers such as surface change (Bedford, 1963), lectins (Koehler, 19781, or enzyme-mediated labeling techniques, e.g., lactoperoxidase iodination (Voglmayr et al., 1980; Jones et al., 1983; Dacheux et al., 1985). Molecular surface changes have also been documented by sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE)of solubilized plasma membranes isolated from spermatozoa, although the purity of such membrane preparations is often questionable (Peterson et al., 1980; Noland et al., 1983). More recently, a number of laboratories have used poly- and monoclonal antibodies to follow maturational changes in surface antigens (Eddy et al., 1985; Jones et al., 1985; Primakoff et al., 1987; Lakoski et al., 1989). There are, however, problems with interpreting results of cytochemical and biochemical experiments of sperm surfaces because of the unusual nature of the mammalian sperm cell. The sperm plasmalemma is fragile and often becomes disrupted by treatments such as centrifugation, yet spermatozoa retain their shape after the plasma membrane is broken. Thus molecular probes may pass through broken areas of the plasma membrane and bind to internal organelles. In immunolocalization studies at the light microscopic level, it is not possible to distinguish between surface and internal antigens. This problem is exacerbated by the finding that spermatozoa from different regions of the epididymis react differently to the same treatment. In general, spermatozoa from the cauda epididymis tend to be more fragile than spermatozoa from the caput epididymis. Thus, for example, differences in immunoReceived January 21, 1991; accepted March 18, 1991. Address reprint requests to Dr. David M. Phillips, The Population Council, 1230 York Ave., New York, NY 10021.
348
D.M. PHILLIPS ET AL.
staining between caput and caudal spermatozoa could be caused simply by their differential response to incubation and centrifugation. Another complication is that antibodies frequently recognize epitopes common to several different antigens. To clarify the extent to which antigens on rat spermatozoa change their distribution during posttesticular development, we have compared quantitative light microscopic observations with ultrastructural localization of four monoclonal antibodies (McAbs) directed against different antigens on rat spermatozoa. These McAbs were chosen because they had previously all been found by immunofluorescence staining to display different areas of distribution on spermatozoa from the cauda than from the caput epididymis. A comparison between light and electron microscopic analysis of these antigens has allowed us to gain perspective on how posttesticular development of spermatozoa in the epididymis involves modifications to intracellular as well as surface membrane components.
Ltd., London, U.K.) and examined with a Zeiss Axiophot photomicroscope using epifluorescence illumination. Percentages of stained cells were determined from counts of a least 100 spermatozoa.
Immunogold Labeling Spermatozoa were collected from the epididymis as described above, washed in RFM, and incubated in hybridoma supernatant for 30 min at room temperature. They were then washed twice and resuspended in gold-conjugated antimouse IgG (5 nm gold; Janssen Pharmaceuticals, Piscataway, NJ) for 30 min at room temperature. Following two washes in RFM to remove the excess gold, spermatozoa were fixed in 2.5% glutaraldehyde in 0.1 M phosphate buffer, pH 7.4, and processed for transmission electron microscopy as previously described (Shalgi et al., 1989). Western Blotting Washed sperm pellets were resuspended in phosphate-buffered saline (PBS) containing 0.5% sodium deoxycholatei2 mM phenylmethylsulphonylfluoride pH 8.5 and incubated on ice for 30 min. Suspensions were centrifuged at 10,OOOg for 10 min and solubilized proteins were separated by reducing SDS-PAGE (815% polycrylamide) (Laemmli, 1970). Proteins were blotted onto nitrocellulose paper (0.45 km) and blots incubated with McAbsiperoxidase rabbit antimouse IgG as described previously (Jones et al., 1985).
MATERIALS AND METHODS Production of Antibodies McAbs to a variety of antigens on spermatozoa from the cauda epididymis of adult male rats (Wistar strain) were produced in female Balb/c mice as described previously (Gaunt et al., 1983). Supernatants were collected from confluent hybridoma cell cultures and frozen at -20°C until use. The type of immunoglobulin was assessed on double diffusion agar gels against Ligation of the Epididymis class-specificantisera (Gaunt et al., 1983).Ascites fluid Rats were anaesthetized with diethyl ether, and the was produced in female Balbic mice that had been previously primed intraperitoneally with 0.5 ml pris- testis and epididymis were exposed through an abdominal incision. The epididymal duct was occluded on one tane (Aldrich Chemicals). side with a ligature at the junction of the distal caput Indirect Immunofluorescence and proximal corpus regions. The contralateral epididSpermatozoa were collected from the testis by minc- ymis served as a control. Five days later, animals were ing the tissue in medium composed of 110 mM NaC1, killed, and spermatozoa were recovered proximal to the 4.1 mM KC1, 3.4 mM CaCl,, 1.2 mM KH2P04,1.2 mM ligature on the experimental side and from the equivMgS04, 25.1 mM NaHC03, 25.1 mM Hepes, 0.1% alent region on the control side, washed, and incubated bovine serum albumin (BSA), pH 7.0 (Jones et al., with McAbsiFITC-RAM as described above. 1990). We will refer to this medium as RFM. Sperm RESULTS were obtained from different regions of the epididymis by puncturing the duct several times with a sharp New antigenic determinants that appear on spermascissors. Suspensions (10 ml) were centrifuged (6008 for tozoa during passage through the epididymis are gen5 mid, sperm pellets were resuspended to a suitable erally referred to as “maturation antigens.” In theory, concentration in RFM, and aliquots were treated in such antigens may arise from modifications to existing three different ways: 1) untreated (control), 2) two membrane components or from interactions with epidcycles of freezing (-20°C) and thawing, 3) exposure to idymal secretions. If located on the surface, they may 0.2%NP-40 for 20 min at room temperature followed by remain unchanged within the particular domain or two washes in RFM. domains on which they were initially detected, or they Antigens were visualized on cells in suspension by may actively migrate to new domains, or they may indirect immunofluorescence microscopy using undi- disappear altogether only to reappear on a different luted hybridoma culture supernatants as a first layer domain. Antigens on intracellular membranes could and 1:100 diluted fluorescein-conjugated rabbit anti- behave similarly, although the extent of internal promouse IgG (FITC-RAM;Bio-Yeda, Rehovot, Israel) as a cessing or rearrangement of antigens within spermasecond layer (Jones et al., 1985). Four McAbs selected tozoa is largely unknown. Superimposed on these variwere designated 1B6, 1B5,5B1, and 2D6. Spermatozoa ables are artifacts caused by the relative fragility of the were suspended in Citifluor AF1 antifadent (Citifluor plasmalemma and the differential response of imma-
SPERM SURFACE ANTIGENS ture vs. mature spermatozoa to incubation and centrifugation. The latter problems are particularly acute when it comes to distinguishing between internally and externally located antigens by indirect immunofluorescence techniques (IIF) (Jones et al., 1983). With these reservations in mind, we have followed the posttesticular maturation of several surface and intracellular antigens on rat spermatozoa using four different McAbs, designated 1B5, 5B1, 1B6, and 2D6. Our approach has been to stain spermatozoa in suspension for IIF under mild conditions in order to avoid damaging their plasma membranes and then deliberately to permeabilize them (by freezinghhawing or brief exposure t o 0.2% NP-40 detergent) to determine if this alters their staining pattern. Finally, at the ultrastructural level, we have localized the type of membrane to which each McAb binds and correlated this with previous evidence obtained from IIF experiments.
349
was obtained on spermatozoa that had lost their plasma membrane. Unfortunately, 1B6 McAb did not give a reaction on Western blots, so it was not possible to determine if its epitope is present on one or several antigens.
2D6 McAb 2D6 McAb (an IgM) did not bind to spermatozoa from the testis or proximal caput epididymis (Fig. 11, even after exposure of intracellular antigens by freezethawing or treatment with NP-40. However, -30% of spermatozoa from the distal caput epididymis showed strong fluorescence all over the acrosomal domain. If caput spermatozoa were freeze-thawed, then >80% became positive on the acrosome, suggesting that the McAb was binding to an intraacrosomal antigen (Fig. 2). This conclusion was confirmed by electron microscopy, which showed gold particles on the outer acrosoma1 membrane. The plasma membrane on distal caput spermatozoa remained unlabeled (Fig. 5). In contrast to the above results, 2D6 McAb bound only to the postacrosomal and tail domains of corpus and cauda epididymal spermatozoa (Fig. 1). Under the present labeling conditions fluorescence was strong and uniform, but if the concentrations of first and second layer antibodies were reduced to 1/16 and 1/50, respectively, then the antigen became patchy in appearance (Gaunt et al., 1983). Interestingly, patchiness could not be induced within the acrosome of distal caput spermatozoa (R. Jones, unpublished observations). In agreement with the above results, only the plasma membrane of cauda epididymal spermatozoa labeled with gold particles (Fig. 6). Likewise, freeze thawing and NP-40 extraction failed to expose intracellular antigens for 2D6 McAb in corpus or cauda epididymal spermatozoa. To investigate if the disappearance of 2D6 binding to the intracrosomal antigen in distal caput sperm was related to maturation, spermatozoa were retained within the distal caput region by ligation of the proximal corpus epididymis. This procedure has been shown in several species to induce progressive motility and to increase the percentage of fertile sperm (Orgebin-Crist et al., 1975). After 5 days there was a 50-60% decrease in the proportion of spermatozoa proximal to the ligature that were positive for 2D6 inside the acrosome. This suggests that processing of this antigen is correlated with sperm maturation. When Western blots containing solubilized proteins from testicular, caput, and cauda epididymal spermatozoa were probed with 2D6 McAb/peroxidase RAM, no reaction was obtained over testicular or caput spermatozoa, but a reaction was detected over a protein at approximately 24 kDa on cauda spermatozoa (Fig. 7).
1B6 McAb In this and in all subsequent experiments, specific staining of spermatozoa by IIF was judged relative to controls that had been incubated with HAT medium/ FITC-RAM. It was found that very weak, nonspecific, second-layer antibody binding was present on the midpiece of -10% of testicular and distal caput epididymal spermatozoa. These spermatozoa probably represent damaged cells since after freeze-thawing or extraction with 0.2% NP-40, >90% of cells showed very weak fluorescence on the midpiece domain. This phenomenon was not observed with cauda epididymal spermatozoa. 1B6 McAb (an IgG1) did not bind specifically to any region of testicular spermatozoa (Fig. 1)and could not be induced to do so even after permeabilization by freeze-thawing or NP-40 extraction. However, 40% of spermatozoa from the proximal caput epididymis showed weak patches of fluorescence on the head (Figs. 1and a), whereas >90% of spermatozoa from the distal caput, corpus, and cauda epididymal regions showed strong uniform fluorescence over the postacrosomal domain. It was noticeable that distal caput and corpus spermatozoa also had a patchy or “network-like” fluorescence on the acrosome cap and the equatorial segment; this fluorescence was not detected on spermatozoa from the cauda epididymis (Fig. 2). Thus, during early stages of maturation, the epitope recognized by 1B6 McAb appears on three different membrane domains on the sperm head but becomes restricted to one domain on fully mature spermatozoa. That 1B6 McAb is recognizing surface membrane antigen(s) was shown by 1)failure to “expose” reactive epitopes on testicular epididymis by permeabilizing them and 2) localization of immunogold particles on the plasma membrane of both caput and cauda epididymal spermatozoa. In agreement with the IIF results,. gold 5B1 McAb was observed over the entire surface of the head of No reaction was obtained on testicular or proximal caput spermatozoa but only on the postacrosomal domain of cauda spermatozoa (Figs. 3 and 4). No reaction caput epididymal spermatozoa (intact or permeabi-
350
D.M. PHILLIPS ET AL.
TESTIS
2 D6
D.CAPUT CORPUS
>90%
0 o/o
40%
0K
0 o/o
6 O/o
0 vo
60o/o
1
lB5
PCAPUT
r 0 o/o
0 */!a
30 O/o
> 90%
CAUDA
>90%
50%
80 Yo
> 90%
>9 o x
> 90%
Fig. 1. Diagram representing changes in staining patterns of monoclonal antibodies. Shaded regions represent FITC positive regions. Numbers represent averages of four to seven experiments rounded to the nearest factor of ten. AC = acrosomal cap; ES = equatorial segment; PA = postacrosomal area.
lized) with 5B1 McAb (an IgG2,) using either IIF or immunogold labeling procedures (Figs. 1and 2). In the distal caput epididymis, -6% of spermatozoa showed strong, irregular areas of fluorescence over the acrosoma1 cap; no other domains of the spermatozoon were stained. In the corpus and cauda epididymal regions, however, 20% and 50% of spermatozoa, respectively, showed strong uniform fluorescence over the entire acrosomal cap regions. Immunogold labeling of the latter spermatozoa revealed particles localized on the
outer acrosomal membranes and the acrosomal contents but not on the plasma membrane (Fig. 8a,b). Further evidence for the intraacrosomal location of 5B1 antigen(s) was provided by freeze-thawing corpus and cauda spermatozoa. In both cases, >90% of cells became fluorescent specifically over the acrosomal cap region. In agreement with the above results, 5B1 McAb did not react on Western blots of testicular spermatozoa, but several antigens with molecular masses of -160 kDa were present in cauda spermatozoa (Fig. 7). A
SPERM SURFACE ANTIGENS
Fig. 2. Photomicrographs of FITC-McAb labeled cauda (left) and caput (right) rat spermatozoa.
351
352
D.M. PHILLIPS ET AL.
Figs. 3,4. Immunogold localization of antibody 1B6. Gold is found on the plasma membrane over the entire surface of the sperm head in caput epididymal spermatozoa (3a-c). However gold is observed only on the plasma membrane over the postacrosomal region in spermatozoa removed from the tail of the epididymis (4a-c). The arrows in 3c and 4c indicate the regions of the sperm head that are enlarged in Figures 3a and b and 4a and b. Figures 3a,b and 4a,b, X75,OOO. Figures 3c and 4c, x 13,000.
(Figs. 1and 2). As spermatozoa progressed through the corpus epididymis, the binding of 1B5 McAb became more and more restricted in distribution until on cauda 1B5 McAb spermatozoa, only the anterior half of the acrosomal The antigenb) carrying the epitope for 1B5 McAb (an domain was stained. Permeabilizing cauda spermatoIgG2,) was first detected on the whole acrosomal do- zoa did not expose any 1B5 antigenb) in the posterior main of spermatozoa from the distal caput epididymis acrosomal domain (i.e., they could not be made to look
weak diffuse reaction was detected at -200 kDa on caput spermatozoa (Fig. 7).
SPERM SURFACE ANTIGENS
353
Fig. 5. Immunogold labeling of caput epididymal spermatozoa with 2D6. Sperm that show label are those with missing or fragmented plasmalemma. The plasmalemma (arrows) is unlabeled. Gold is found on the outer acrosomal membrane. ~ 6 5 , 0 0 0 . Fig. 6. Antibody 2D6 stains the plasma membrane of the sperm tail. It is not observed on the fibrous sheath or mitochondria1 membrane in cells with missing plasma membranes (upper and lower right). ~75,000.
labeling was localized to the outer acrosomal membrane and acrosomal contents (Figs. 9 and 10).
DISCUSSION This work has shown that, in addition t o remodeling of the surface membrane of spermatozoa during epididymal maturation, there is also significant processing of intracellular antigens, particularly in the head domain. The use of specific immunological probes combined with their localization at the ultrastructural level has enabled us to investigate this processing of membrane antigens in detail and to interpret artifacts arising from inadvertently damaging the cells. One cannot assume that an antibody is recognizing a cell surface antigen from immunocytochemical evidence at the level of the light microscope, because there is no way to do immunocytochemistry of spermatozoa without breaking membranes. There are a number of Fig. 7. Western blots of detergent extracted proteins separated by different methods for immunocytochemistry. Many non-reducing SDS-PAGE from testicular (c), distal caput (b), and workers air dry cells from phosphate-buffered saline. cauda (a) spermatozoa probed with 2D6 and 5B1 McAb. Control blots This produces extreme hypertonic conditions and cryswere probed with HAT medium. Approximate molecular masses of tal formation as water evaporates and the solution known protein standards are as shown in kDa. dries. Fixation certainly causes membrane breakage, and immunocytochemical localization of intercellular like caput spermatozoa), suggesting that either 1B5 proteins can be carried out only on fixed cells. To antigen(s) had migrated to the anterior tip of the circumvent these problems, we carried out our reacacrosome, or else it had been selectively masked/ tions on living spermatozoa before fixation, but even so modified in the posterior acrosomal area. Immunogold the incubations and slow centrifugations must have
354
D.M. PHILLIPS ET AL.
Fig. 8. 5B1 McAb is localized on the outer acrosomal membrane.
permeabilized some cells, because in certain instances we found the reaction product on intracellular membranes. The fact that immunocytochemistry of spermatozoa can detect intracellular as well as surface antigens drastically affects the interpretation of IIF observations. Consider McAb 2D6; it is present on the head of caput epididymal spermatozoa and the tail of cauda epididymal spermatozoa. If one assumed that 2D6 was recognizing a plasma membrane antigen it could be concluded that the antigen migrated from the head to the tail during epididymal maturation. However, this is clearly not the case. The antibody recognizes a plasma membrane protein in spermatozoa from the cauda epididymis, but it identifies an intracellular (outer acrosomal) membrane antigen in caput epididyma1 spermatozoa. Since the outer acrosomal membrane and plasma membrane are not connected, the McAb is most likely recognizing two separate antigens. A plasma membrane antigen over the tail is being added (Jones et al., 1985) and simultaneously an outer acrosome antigen recognized by the same antibody is being removed or masked. It is also possible that one or both of these antigens is not being added but that the epitope that is recognized by 2D6 is becoming available due to a change in tertiary structure of glycosylation. Whatever the case, it is clear that the different staining patterns obtained with 2D6 McAb on caput and cauda spermatozoa cannot be interpreted as evidence of antigen migration or redistribution. Stage-specific expression of antigens during spermatogenesis has been documented in several species (e.g., J1, C6, and A5 antigens in the mouse [Fenderson et al., 1984; Symington et al., 19841 and RSA antigens
in the rabbit [O’Rand et al., 198811, but less is known about this phenomenon during epididymal maturation. Like J1, C6, and A5 McAbs, 2D6 McAb is of the IgM subclass. It binds to surface as well as intracellular antigens and, according to our most recent evidence, recognizes epitopes on O-linked carbohydrate moieties, probably of the polylactosamine type (Jones, unpublished observations). The acquisition and subsequent loss of the 2D6 epitope from the sperm acrosome during a specific stage of epididymal maturation emphasizes that there are subtle changes taking place within the acrosome that are likely to be of major importance in relation to zona recognition and egg fusion, two properties that are not found in testicular or caput spermatozoa. Precisely how such modifications might be mediated remains a puzzle. The Golgi contribution to the developing acrosome is thought to cease at stage 8 of spermiogenesis (Clermont and Tang, 19851, so it would seem likely that new terminal sugars could be added to existing polysaccharide chains. However, it could be argued that, since the mature acrosome contains a variety of glycosidases and hydrolases (Bellve and O’Brien, 19831, modifications could still occur. These questions are difficult to answer without more information on the chemistry of the 2D6 epitope itself and the factors that are influential in switching it “on” or ‘
Alterations in Distribution of Surface and Intracellular Antigens During Epididymal Maturation of Rat Spermatozoa DAVID M. PHILLIPS,’ ROY JONES: AND RUTH SHALG13 ‘The Population Council, New York, New York; 2Department of Molecular Embryology, AFRC Institute of Animal Physiology and Genetics Research, Cambridge, United Kingdom; 3Department of Embryology and Teratology, Tel Aviv University Medical School, Ramat Aviv, Israel
ABSTRACT The surface membrane of mammalian spermatozoa is known to undergo considerable conformational and organizational changes during epididyma1 maturation. However, much less is known about remodelling of intracellular membranes. In this communication we have used specific immunological markers to study the behavior of several antigens both on and within rat spermatozoa as they mature in the epididymis. Four monoclonal antibodies (McAbs)designated 5B1, 165,2D6, and 156 were used to probe testicular and caput and cauda epididymal spermatozoa by indirect immunofluorescence and immunogold labeling techniques. None of the McAbs bound to testicular spermatozoa; in all cases, they became reactive only on spermatozoa which had reached the caput epididymis. McAb 551 was restricted to the outer acrosomal membrane (OAM) of the acrosomal cap domain. The epitope first appeared on antigen(s) with molecular mass (Mr) of -200 kDa in immature spermatozoa, but later in mature spermatozoa the antigen(s)had Mr of -160 kDa. The antigen(s)recognized by 1B5 McAb on the other hand was initially distributed over the OAM of the entire acrosomal domain (cap + equatorial segment), but during maturation it became progressively more restricted in area until in cauda spermatozoa only the anterior tip of the OAM bound the McAb. McAb 2D6 also bound to the entire OAM and acrosomal contents of caput spermatozoa, but, unlike 5B1 and 1B5 McAbs, reactivity was transient. That is, staining was first detected in caput spermatozoa but then disappeared in corpus and cauda spermatozoa. In contrast to all of the above, 1B6 McAb bound to the surface membrane overlying the entire head domain of caput spermatozoa, but during maturation it became restricted to the postacrosomal domain. These results indicate that, in addition to remodeling of the surface membrane during epididymal maturation, extensive processing of intracellular membrane antigens also takes place and that it is very active within the acrosome. The nature of these intracellular processing events remains to be elucidated, but they may have important consequences for membrane fusion and cell recognition phenomena during fertilization. Key Words: Sperm maturation, lntracellular membranes, Outer acrosomal membrane
0 1991 WILEY-LISS, INC.
INTRODUCTION Mammalian spermatozoa undergo numerous physiological, morphological, and biochemical “maturation” changes as they pass through the epididymis (Bedford, 1975; Amann, 1987). Some of the most often studied aspects of sperm maturation have been the changes in the sperm surface which accompany epididymal transit (reviewed by Jones et al., 1990). Early investigations employed relatively nonspecific markers such as surface change (Bedford, 1963), lectins (Koehler, 19781, or enzyme-mediated labeling techniques, e.g., lactoperoxidase iodination (Voglmayr et al., 1980; Jones et al., 1983; Dacheux et al., 1985). Molecular surface changes have also been documented by sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE)of solubilized plasma membranes isolated from spermatozoa, although the purity of such membrane preparations is often questionable (Peterson et al., 1980; Noland et al., 1983). More recently, a number of laboratories have used poly- and monoclonal antibodies to follow maturational changes in surface antigens (Eddy et al., 1985; Jones et al., 1985; Primakoff et al., 1987; Lakoski et al., 1989). There are, however, problems with interpreting results of cytochemical and biochemical experiments of sperm surfaces because of the unusual nature of the mammalian sperm cell. The sperm plasmalemma is fragile and often becomes disrupted by treatments such as centrifugation, yet spermatozoa retain their shape after the plasma membrane is broken. Thus molecular probes may pass through broken areas of the plasma membrane and bind to internal organelles. In immunolocalization studies at the light microscopic level, it is not possible to distinguish between surface and internal antigens. This problem is exacerbated by the finding that spermatozoa from different regions of the epididymis react differently to the same treatment. In general, spermatozoa from the cauda epididymis tend to be more fragile than spermatozoa from the caput epididymis. Thus, for example, differences in immunoReceived January 21, 1991; accepted March 18, 1991. Address reprint requests to Dr. David M. Phillips, The Population Council, 1230 York Ave., New York, NY 10021.
348
D.M. PHILLIPS ET AL.
staining between caput and caudal spermatozoa could be caused simply by their differential response to incubation and centrifugation. Another complication is that antibodies frequently recognize epitopes common to several different antigens. To clarify the extent to which antigens on rat spermatozoa change their distribution during posttesticular development, we have compared quantitative light microscopic observations with ultrastructural localization of four monoclonal antibodies (McAbs) directed against different antigens on rat spermatozoa. These McAbs were chosen because they had previously all been found by immunofluorescence staining to display different areas of distribution on spermatozoa from the cauda than from the caput epididymis. A comparison between light and electron microscopic analysis of these antigens has allowed us to gain perspective on how posttesticular development of spermatozoa in the epididymis involves modifications to intracellular as well as surface membrane components.
Ltd., London, U.K.) and examined with a Zeiss Axiophot photomicroscope using epifluorescence illumination. Percentages of stained cells were determined from counts of a least 100 spermatozoa.
Immunogold Labeling Spermatozoa were collected from the epididymis as described above, washed in RFM, and incubated in hybridoma supernatant for 30 min at room temperature. They were then washed twice and resuspended in gold-conjugated antimouse IgG (5 nm gold; Janssen Pharmaceuticals, Piscataway, NJ) for 30 min at room temperature. Following two washes in RFM to remove the excess gold, spermatozoa were fixed in 2.5% glutaraldehyde in 0.1 M phosphate buffer, pH 7.4, and processed for transmission electron microscopy as previously described (Shalgi et al., 1989). Western Blotting Washed sperm pellets were resuspended in phosphate-buffered saline (PBS) containing 0.5% sodium deoxycholatei2 mM phenylmethylsulphonylfluoride pH 8.5 and incubated on ice for 30 min. Suspensions were centrifuged at 10,OOOg for 10 min and solubilized proteins were separated by reducing SDS-PAGE (815% polycrylamide) (Laemmli, 1970). Proteins were blotted onto nitrocellulose paper (0.45 km) and blots incubated with McAbsiperoxidase rabbit antimouse IgG as described previously (Jones et al., 1985).
MATERIALS AND METHODS Production of Antibodies McAbs to a variety of antigens on spermatozoa from the cauda epididymis of adult male rats (Wistar strain) were produced in female Balb/c mice as described previously (Gaunt et al., 1983). Supernatants were collected from confluent hybridoma cell cultures and frozen at -20°C until use. The type of immunoglobulin was assessed on double diffusion agar gels against Ligation of the Epididymis class-specificantisera (Gaunt et al., 1983).Ascites fluid Rats were anaesthetized with diethyl ether, and the was produced in female Balbic mice that had been previously primed intraperitoneally with 0.5 ml pris- testis and epididymis were exposed through an abdominal incision. The epididymal duct was occluded on one tane (Aldrich Chemicals). side with a ligature at the junction of the distal caput Indirect Immunofluorescence and proximal corpus regions. The contralateral epididSpermatozoa were collected from the testis by minc- ymis served as a control. Five days later, animals were ing the tissue in medium composed of 110 mM NaC1, killed, and spermatozoa were recovered proximal to the 4.1 mM KC1, 3.4 mM CaCl,, 1.2 mM KH2P04,1.2 mM ligature on the experimental side and from the equivMgS04, 25.1 mM NaHC03, 25.1 mM Hepes, 0.1% alent region on the control side, washed, and incubated bovine serum albumin (BSA), pH 7.0 (Jones et al., with McAbsiFITC-RAM as described above. 1990). We will refer to this medium as RFM. Sperm RESULTS were obtained from different regions of the epididymis by puncturing the duct several times with a sharp New antigenic determinants that appear on spermascissors. Suspensions (10 ml) were centrifuged (6008 for tozoa during passage through the epididymis are gen5 mid, sperm pellets were resuspended to a suitable erally referred to as “maturation antigens.” In theory, concentration in RFM, and aliquots were treated in such antigens may arise from modifications to existing three different ways: 1) untreated (control), 2) two membrane components or from interactions with epidcycles of freezing (-20°C) and thawing, 3) exposure to idymal secretions. If located on the surface, they may 0.2%NP-40 for 20 min at room temperature followed by remain unchanged within the particular domain or two washes in RFM. domains on which they were initially detected, or they Antigens were visualized on cells in suspension by may actively migrate to new domains, or they may indirect immunofluorescence microscopy using undi- disappear altogether only to reappear on a different luted hybridoma culture supernatants as a first layer domain. Antigens on intracellular membranes could and 1:100 diluted fluorescein-conjugated rabbit anti- behave similarly, although the extent of internal promouse IgG (FITC-RAM;Bio-Yeda, Rehovot, Israel) as a cessing or rearrangement of antigens within spermasecond layer (Jones et al., 1985). Four McAbs selected tozoa is largely unknown. Superimposed on these variwere designated 1B6, 1B5,5B1, and 2D6. Spermatozoa ables are artifacts caused by the relative fragility of the were suspended in Citifluor AF1 antifadent (Citifluor plasmalemma and the differential response of imma-
SPERM SURFACE ANTIGENS ture vs. mature spermatozoa to incubation and centrifugation. The latter problems are particularly acute when it comes to distinguishing between internally and externally located antigens by indirect immunofluorescence techniques (IIF) (Jones et al., 1983). With these reservations in mind, we have followed the posttesticular maturation of several surface and intracellular antigens on rat spermatozoa using four different McAbs, designated 1B5, 5B1, 1B6, and 2D6. Our approach has been to stain spermatozoa in suspension for IIF under mild conditions in order to avoid damaging their plasma membranes and then deliberately to permeabilize them (by freezinghhawing or brief exposure t o 0.2% NP-40 detergent) to determine if this alters their staining pattern. Finally, at the ultrastructural level, we have localized the type of membrane to which each McAb binds and correlated this with previous evidence obtained from IIF experiments.
349
was obtained on spermatozoa that had lost their plasma membrane. Unfortunately, 1B6 McAb did not give a reaction on Western blots, so it was not possible to determine if its epitope is present on one or several antigens.
2D6 McAb 2D6 McAb (an IgM) did not bind to spermatozoa from the testis or proximal caput epididymis (Fig. 11, even after exposure of intracellular antigens by freezethawing or treatment with NP-40. However, -30% of spermatozoa from the distal caput epididymis showed strong fluorescence all over the acrosomal domain. If caput spermatozoa were freeze-thawed, then >80% became positive on the acrosome, suggesting that the McAb was binding to an intraacrosomal antigen (Fig. 2). This conclusion was confirmed by electron microscopy, which showed gold particles on the outer acrosoma1 membrane. The plasma membrane on distal caput spermatozoa remained unlabeled (Fig. 5). In contrast to the above results, 2D6 McAb bound only to the postacrosomal and tail domains of corpus and cauda epididymal spermatozoa (Fig. 1). Under the present labeling conditions fluorescence was strong and uniform, but if the concentrations of first and second layer antibodies were reduced to 1/16 and 1/50, respectively, then the antigen became patchy in appearance (Gaunt et al., 1983). Interestingly, patchiness could not be induced within the acrosome of distal caput spermatozoa (R. Jones, unpublished observations). In agreement with the above results, only the plasma membrane of cauda epididymal spermatozoa labeled with gold particles (Fig. 6). Likewise, freeze thawing and NP-40 extraction failed to expose intracellular antigens for 2D6 McAb in corpus or cauda epididymal spermatozoa. To investigate if the disappearance of 2D6 binding to the intracrosomal antigen in distal caput sperm was related to maturation, spermatozoa were retained within the distal caput region by ligation of the proximal corpus epididymis. This procedure has been shown in several species to induce progressive motility and to increase the percentage of fertile sperm (Orgebin-Crist et al., 1975). After 5 days there was a 50-60% decrease in the proportion of spermatozoa proximal to the ligature that were positive for 2D6 inside the acrosome. This suggests that processing of this antigen is correlated with sperm maturation. When Western blots containing solubilized proteins from testicular, caput, and cauda epididymal spermatozoa were probed with 2D6 McAb/peroxidase RAM, no reaction was obtained over testicular or caput spermatozoa, but a reaction was detected over a protein at approximately 24 kDa on cauda spermatozoa (Fig. 7).
1B6 McAb In this and in all subsequent experiments, specific staining of spermatozoa by IIF was judged relative to controls that had been incubated with HAT medium/ FITC-RAM. It was found that very weak, nonspecific, second-layer antibody binding was present on the midpiece of -10% of testicular and distal caput epididymal spermatozoa. These spermatozoa probably represent damaged cells since after freeze-thawing or extraction with 0.2% NP-40, >90% of cells showed very weak fluorescence on the midpiece domain. This phenomenon was not observed with cauda epididymal spermatozoa. 1B6 McAb (an IgG1) did not bind specifically to any region of testicular spermatozoa (Fig. 1)and could not be induced to do so even after permeabilization by freeze-thawing or NP-40 extraction. However, 40% of spermatozoa from the proximal caput epididymis showed weak patches of fluorescence on the head (Figs. 1and a), whereas >90% of spermatozoa from the distal caput, corpus, and cauda epididymal regions showed strong uniform fluorescence over the postacrosomal domain. It was noticeable that distal caput and corpus spermatozoa also had a patchy or “network-like” fluorescence on the acrosome cap and the equatorial segment; this fluorescence was not detected on spermatozoa from the cauda epididymis (Fig. 2). Thus, during early stages of maturation, the epitope recognized by 1B6 McAb appears on three different membrane domains on the sperm head but becomes restricted to one domain on fully mature spermatozoa. That 1B6 McAb is recognizing surface membrane antigen(s) was shown by 1)failure to “expose” reactive epitopes on testicular epididymis by permeabilizing them and 2) localization of immunogold particles on the plasma membrane of both caput and cauda epididymal spermatozoa. In agreement with the IIF results,. gold 5B1 McAb was observed over the entire surface of the head of No reaction was obtained on testicular or proximal caput spermatozoa but only on the postacrosomal domain of cauda spermatozoa (Figs. 3 and 4). No reaction caput epididymal spermatozoa (intact or permeabi-
350
D.M. PHILLIPS ET AL.
TESTIS
2 D6
D.CAPUT CORPUS
>90%
0 o/o
40%
0K
0 o/o
6 O/o
0 vo
60o/o
1
lB5
PCAPUT
r 0 o/o
0 */!a
30 O/o
> 90%
CAUDA
>90%
50%
80 Yo
> 90%
>9 o x
> 90%
Fig. 1. Diagram representing changes in staining patterns of monoclonal antibodies. Shaded regions represent FITC positive regions. Numbers represent averages of four to seven experiments rounded to the nearest factor of ten. AC = acrosomal cap; ES = equatorial segment; PA = postacrosomal area.
lized) with 5B1 McAb (an IgG2,) using either IIF or immunogold labeling procedures (Figs. 1and 2). In the distal caput epididymis, -6% of spermatozoa showed strong, irregular areas of fluorescence over the acrosoma1 cap; no other domains of the spermatozoon were stained. In the corpus and cauda epididymal regions, however, 20% and 50% of spermatozoa, respectively, showed strong uniform fluorescence over the entire acrosomal cap regions. Immunogold labeling of the latter spermatozoa revealed particles localized on the
outer acrosomal membranes and the acrosomal contents but not on the plasma membrane (Fig. 8a,b). Further evidence for the intraacrosomal location of 5B1 antigen(s) was provided by freeze-thawing corpus and cauda spermatozoa. In both cases, >90% of cells became fluorescent specifically over the acrosomal cap region. In agreement with the above results, 5B1 McAb did not react on Western blots of testicular spermatozoa, but several antigens with molecular masses of -160 kDa were present in cauda spermatozoa (Fig. 7). A
SPERM SURFACE ANTIGENS
Fig. 2. Photomicrographs of FITC-McAb labeled cauda (left) and caput (right) rat spermatozoa.
351
352
D.M. PHILLIPS ET AL.
Figs. 3,4. Immunogold localization of antibody 1B6. Gold is found on the plasma membrane over the entire surface of the sperm head in caput epididymal spermatozoa (3a-c). However gold is observed only on the plasma membrane over the postacrosomal region in spermatozoa removed from the tail of the epididymis (4a-c). The arrows in 3c and 4c indicate the regions of the sperm head that are enlarged in Figures 3a and b and 4a and b. Figures 3a,b and 4a,b, X75,OOO. Figures 3c and 4c, x 13,000.
(Figs. 1and 2). As spermatozoa progressed through the corpus epididymis, the binding of 1B5 McAb became more and more restricted in distribution until on cauda 1B5 McAb spermatozoa, only the anterior half of the acrosomal The antigenb) carrying the epitope for 1B5 McAb (an domain was stained. Permeabilizing cauda spermatoIgG2,) was first detected on the whole acrosomal do- zoa did not expose any 1B5 antigenb) in the posterior main of spermatozoa from the distal caput epididymis acrosomal domain (i.e., they could not be made to look
weak diffuse reaction was detected at -200 kDa on caput spermatozoa (Fig. 7).
SPERM SURFACE ANTIGENS
353
Fig. 5. Immunogold labeling of caput epididymal spermatozoa with 2D6. Sperm that show label are those with missing or fragmented plasmalemma. The plasmalemma (arrows) is unlabeled. Gold is found on the outer acrosomal membrane. ~ 6 5 , 0 0 0 . Fig. 6. Antibody 2D6 stains the plasma membrane of the sperm tail. It is not observed on the fibrous sheath or mitochondria1 membrane in cells with missing plasma membranes (upper and lower right). ~75,000.
labeling was localized to the outer acrosomal membrane and acrosomal contents (Figs. 9 and 10).
DISCUSSION This work has shown that, in addition t o remodeling of the surface membrane of spermatozoa during epididymal maturation, there is also significant processing of intracellular antigens, particularly in the head domain. The use of specific immunological probes combined with their localization at the ultrastructural level has enabled us to investigate this processing of membrane antigens in detail and to interpret artifacts arising from inadvertently damaging the cells. One cannot assume that an antibody is recognizing a cell surface antigen from immunocytochemical evidence at the level of the light microscope, because there is no way to do immunocytochemistry of spermatozoa without breaking membranes. There are a number of Fig. 7. Western blots of detergent extracted proteins separated by different methods for immunocytochemistry. Many non-reducing SDS-PAGE from testicular (c), distal caput (b), and workers air dry cells from phosphate-buffered saline. cauda (a) spermatozoa probed with 2D6 and 5B1 McAb. Control blots This produces extreme hypertonic conditions and cryswere probed with HAT medium. Approximate molecular masses of tal formation as water evaporates and the solution known protein standards are as shown in kDa. dries. Fixation certainly causes membrane breakage, and immunocytochemical localization of intercellular like caput spermatozoa), suggesting that either 1B5 proteins can be carried out only on fixed cells. To antigen(s) had migrated to the anterior tip of the circumvent these problems, we carried out our reacacrosome, or else it had been selectively masked/ tions on living spermatozoa before fixation, but even so modified in the posterior acrosomal area. Immunogold the incubations and slow centrifugations must have
354
D.M. PHILLIPS ET AL.
Fig. 8. 5B1 McAb is localized on the outer acrosomal membrane.
permeabilized some cells, because in certain instances we found the reaction product on intracellular membranes. The fact that immunocytochemistry of spermatozoa can detect intracellular as well as surface antigens drastically affects the interpretation of IIF observations. Consider McAb 2D6; it is present on the head of caput epididymal spermatozoa and the tail of cauda epididymal spermatozoa. If one assumed that 2D6 was recognizing a plasma membrane antigen it could be concluded that the antigen migrated from the head to the tail during epididymal maturation. However, this is clearly not the case. The antibody recognizes a plasma membrane protein in spermatozoa from the cauda epididymis, but it identifies an intracellular (outer acrosomal) membrane antigen in caput epididyma1 spermatozoa. Since the outer acrosomal membrane and plasma membrane are not connected, the McAb is most likely recognizing two separate antigens. A plasma membrane antigen over the tail is being added (Jones et al., 1985) and simultaneously an outer acrosome antigen recognized by the same antibody is being removed or masked. It is also possible that one or both of these antigens is not being added but that the epitope that is recognized by 2D6 is becoming available due to a change in tertiary structure of glycosylation. Whatever the case, it is clear that the different staining patterns obtained with 2D6 McAb on caput and cauda spermatozoa cannot be interpreted as evidence of antigen migration or redistribution. Stage-specific expression of antigens during spermatogenesis has been documented in several species (e.g., J1, C6, and A5 antigens in the mouse [Fenderson et al., 1984; Symington et al., 19841 and RSA antigens
in the rabbit [O’Rand et al., 198811, but less is known about this phenomenon during epididymal maturation. Like J1, C6, and A5 McAbs, 2D6 McAb is of the IgM subclass. It binds to surface as well as intracellular antigens and, according to our most recent evidence, recognizes epitopes on O-linked carbohydrate moieties, probably of the polylactosamine type (Jones, unpublished observations). The acquisition and subsequent loss of the 2D6 epitope from the sperm acrosome during a specific stage of epididymal maturation emphasizes that there are subtle changes taking place within the acrosome that are likely to be of major importance in relation to zona recognition and egg fusion, two properties that are not found in testicular or caput spermatozoa. Precisely how such modifications might be mediated remains a puzzle. The Golgi contribution to the developing acrosome is thought to cease at stage 8 of spermiogenesis (Clermont and Tang, 19851, so it would seem likely that new terminal sugars could be added to existing polysaccharide chains. However, it could be argued that, since the mature acrosome contains a variety of glycosidases and hydrolases (Bellve and O’Brien, 19831, modifications could still occur. These questions are difficult to answer without more information on the chemistry of the 2D6 epitope itself and the factors that are influential in switching it “on” or ‘
