THE ANATOMICAL RECORD 216:504-515 (1986) Localization of Actin in Mammalian Spermatozoa: A Comparison of Eight Species SEAN P. FLAHERTY, VIRGINIA P. WINFREY, AND GARY E. OLSON Department o f Cell Biology, School of Medicine, Vanderbilt University, Nashville, TN 37232 ABSTRACT The distribution of monomeric and polymeric actin in spermatozoa from the bull, boar, rabbit, human, rat, mouse, golden hamster, and guinea pig has been examined by using a monoclonal antiactin antibody and NBD-phallacidin. Actin was present in sperm from each species. When the monoclonal antibody was used, there was a species-specific distribution and intensity of fluorescence, but no generalized pattern. Specific fluorescence was noted in the neck and principal piece of human sperm; in the postacrosomal region, neck, and midpiece of bull and boar sperm; in the postacrosomal region, neck, and principallequatorial segment border of rabbit sperm; in the neck region of hamster sperm; and in the neck, midpiece, and principal piece of rat, mouse, and guinea pig sperm. Sperm from all eight species displayed no specific fluorescence with NBD-phallacidin, indicating that actin was present in a nonfilamentous form. SDS extracts of sperm were analyzed by SDSPAGE and Western blotting; in sperm from each species, a 42-kD protein with specific affinity for the monoclonal antibody was present. The cytoskeletal proteins actin and myosin are components of the cytoskeleton of eukaryotic cells. The cytoskeleton contributes to the determination and maintenance of cell shape, cell motility, intracellular transport of organelles, organization of the cell surface, and dynamic membrane functions such as endocytosis (Clark and Spudich, 1977; Korn, 1978; Jacobson, 1983). There have been numerous reports of cytoskeletal proteins in mammalian spermatozoa. Campanella et al. (1979) localized myosin in the acrosomal region of human sperm, but using affinity purified antibodies, Virtanen et al. (1984) found myosin sequestered primarily in the neck region. In addition, biochemical evidence has been presented for myosin in bull sperm, although its localization to specific regions of the head or tail was not established (Tamblyn, 1981).Vimentin and spectrin have also been identified by immunofluorescence and biochemical techniques in human sperm (Virtanen et al., 1984). Actin has been identified immunocytochemically in epididymal and ejaculated sperm from many mammalian species and generally is localized in either the acrosomal (Talbot and Kleve, 1978; Virtanen et al., 1984) or postacrosomal region (Clarke and Yanagimachi, 1978; Tamblyn, 1980; Greenberg and Tamblyn, 1981; Clarke et al., 1982; Welch and O’Rand, 1985). Studies of human sperm have produced conflicting results on the distribution of this protein. Clarke and co-workers demonstrated actin in the postacrosomal region (Clarke and Yanagimachi, 1978; Clarke et al., 1982), while Virtanen et al. (1984) suggested it was present in the principal segment of the acrosome and principal piece of the tail. Such differences may relate to variations in technique or the specificity of antisera. Nevertheless, a 42-kD protein which either comigrates on SDS polyacrylamide gels @ 1986 ALAN R. LISS, INC. with skeletal muscle actin, or binds to myosin, or reacts with actin antisera has been identified in extracts from human, hamster, bull, and rabbit spermatozoa (Talbot and Kleve, 1978; Greenberg and Tamblyn, 1981; Clarke et al., 1982; Ochs, 1984; Virtanen et al., 1984; Ochs and Wolf, 1985; Welch and O’Rand, 19851, suggesting that actin probably is a component of most mammalian spermatozoa. In contrast, actin has not been found with immunofluorescence or biochemical techniques in cauda epididymal sperm of the rat and guinea pig (Campanella et al., 1979; Baccetti et al., 1980; Halenda et al., 1984). In human and rabbit sperm, actin appears to be in the monomeric form, based on the absence of staining with the F-actin probes NBD-phallacidin and rhodaminephalloidin (Virtanen et al., 1984; Welch and O’Rand, 1985). Actin filaments have only been demonstrated in cauda epididymal sperm from one species, the plains mouse. However, the filaments are associated with the two ventral hooks on the the sperm head, a unique structural component of hydromyine rodent sperm (Flaherty et al., 1983; Flaherty, 1986). Peterson et al. (1978) used ionophore A23187 to induce acrosome reactions in boar sperm and noted 5-9-nm-thick microfilaments around the sperm head; however, the nature of these filaments was not determined. The presence of filamentous or tubular cytoskeletal elements on the surface of guinea pig, rat, and vole sperm has also been described (Phillips, 1975; Koehler, 1978).A recent study also revealed parallel arrays of 11-12-nm-thick filaments associated with the outer acrosomal membrane along the ventral margin of the hamster sperm head (Olson and Winfrey, 1985). Received June 24, 1986; accepted July 25, 1986. ACTIN IN MAMMALIAN SPERM 505 Prior to immunostaining, the coverslips were rinsed several times in PBS and then immersed for 20 minutes in PBS containing 1%heat-inactivated normal goat serum (NGS). They were then incubated for 90 minutes at room temperature in a 1:lOO dilution of the primary antibody (Monoclonal antiactin, Amersham Corporation, Arlington Heights, Illinois) in PBS containing l% NGS and 0.1% sodium azide (PBS/NGS). In the controls, the primary antibody was omitted. Controls for autofluorescence were incubated in PBS alone. After three washes in PBS/NGS (5 minutes each), the coverslips were incubated with the secondary antibody (FITC-conjugated affinity purified goat antimouse IgM, KPL, Gaithersburg, Maryland) at a dilution of 1:lOO in PBS for 60 minutes in the dark a t room temperature. The coverslips were washed thoroughly in PBS and mounted on slides with a 9:lmixture of glycerol and PBS, pH 7.8. The preparations were examined by using a Zeiss microscope fitted with epifluorescence, FITC filters, and a 63 x (N.A. 1.4)objective. Matched phase-contrast and fluorescence images were recorded on Kodak Tri-X film. As positive controls, AKR-2B fibroblasts were grown on glass coverslips in McCoy’s 5a medium supplemented with 5% fetal calf serum, and rat skeletal muscle myofiMATERIALS AND METHODS brils were prepared according to the method of Knight Preparation of Spermatozoa and Trinick (1982).These were then fixed and stained Adult male golden hamsters, guinea pigs, Sprague- with the antibody. Dawley rats, New Zealand white rabbits, and C57 b116 Localization of Filamentous Actin Using NBD-phallacidin mice were killed by a n overdose of Nembutal or cervical dislocation, and the cauda epididymides were removed NBD-phallacidin (7-nitrobenz-1,3-diazole-phallacidin) and rinsed carefully in phosphate-buffered saline (PBS: was obtained from Molecular Probes Inc., Junction City, 150 mM NaC1, 20 mM sodium phosphate, pH 7.6). To Oregon. A working solution (3.3x lop6 M) was prepared avoid contamination of the suspensions with blood cells, by drying the stock solution and resuspending it in PBS adherent fat and blood vessels were removed; then sperm (+0.1% azide), pH 7.6. were extruded from the large distal tubules into warm Washed sperm were attached to poly-L-lysine-coated PBS by carefully pricking the tubules with a 25G needle. coverslips, then fixed in 1-4% formaldehyde in PBS, pH The resultant sperm suspensions were washed twice by 7.6, for 15 minutes, permeabilized with cold acetone for centrifugation (500g, 8 minutes). 5 minutes, and rinsed in PBS. Some sperm were also Human semen was obtained from volunteers by mas- fixed in suspension before attachment to coverslips. The turbation and allowed to liquefy for 15 minutes prior to coverslips were incubated in the working solution for 45 washing. minutes in the dark a t room temperature, rinsed thorBull and boar epididymides were obtained from a local oughly in PBS and then in distilled water, and mounted slaughterhouse and maintained on ice. Sperm were with a 1:l mixture of PBS and glycerol, pH 7.8.The flushed from the cauda epididymidis with cold PBS or slides were examined as described in the previous Hepes-buffered saline (HBS: 145 mM NaC1, 10 mM section. Hepes, pH 7.6) as previously described (Olson et al., Control preparations consisted of sperm treated in a n 1983)and washed twice in cold PBS. identical manner but incubated in a working solution which did not contain NBD-phallacidin. As additional lmrnunofluorescence controls, cultured fibroblasts and skeletal muscle myofiSperm were allowed to attach to poly-L-lysine-coated brils were also stained with NBD-phallacidin, and the coverslips, and subsequently were rinsed in PBS. Sus- appropriate controls were prepared. pensions containing sperm from several species were Preparation of Sperm Extracts often used to demonstrate the relative intensity of fluorescence. The coverslips were then treated in one of two Proteins were extracted from cauda epididymal sperm ways: 1)fixation in acetone a t -20°C for 10 minutes, or of all species except human. Sperm counts were deter2) fixation in 1-4% formaldehyde in PBS (pH 7.6) for 15- mined for the washed sperm suspensions, and aliquots 30 minutes, followed by rinsing in PBS and permeabili- containing 1-20 x lo7 sperm were incubated for 60 minzation in cold acetone for 5 minutes. In some prepara- utes at room temperature in a n extraction medium tions, sperm were fixed in suspension with formaldehyde which consisted of 250 p1 of 1% SDS, 20 mM Tris-HC1, prior to attachment to coverslips and then treated with pH 6.8, 1 pg/ml leupeptin, 1 pg/ml pepstatin, 1 mM cold acetone. PMSF, and 0.1% sodium azide. They were pelleted a t To investigate whether actin was associated with the 7,OOOg for 3 minutes; the supernatant fluid was removed plasma membrane, sperm were incubated in 0.1% Tri- and mixed 1:l with 2~ sample buffer (Laemmli, 1970) ton X-100in PBS (pH 7.6) for 15 or 60 minutes on ice, and then boiled for 2 minutes. washed twice in PBS (700g, 8 minutes), and then atTo provide an authentic actin sample, SDS extracts of tached to coverslips and fixed in cold acetone as above. rat skeletal muscle were also prepared. The hindleg Thus, there is considerable variation in the literature with respect to the presence and distribution of cytoskeleta1 elements, in particular actin, in mammalian spermatozoa. Some of the conflict may be due to variations in the specificity of the antisera employed. An understanding of the organization of the sperm cytoskeleton would contribute to our perception of the differentiation and functional topography of sperm and the events of fertilization. The aim of the current study, therefore, was to define the distribution of actin in sperm from eight species by using specific probes and a combination of biochemical and histochemical techniques. The localization of actin has been examined by immunofluorescence microscopy using a monoclonal antiactin antibody and by fluorescence microscopy with the actin probe NBD-phallacidin (Barak et al., 1980).In addition, sperm extracts have been analyzed by SDS-PAGE, and actin has been identified on immunostained Western blots. Actin is present in bull, boar, rabbit, human, rat, mouse, hamster, and guinea pig sperm, although its abundance and distribution is highly species-specific. These results are compared to previous studies, and the possible function of actin in fertilization is discussed. 506 S.P. FLAHERTY, V.P. WINFREY, AND G.E. OLSON muscles were removed, rinsed in PBS and then macer- ment (Fig. 2). Acetone fixed boar sperm showed a similar ated in cold PBS with a razor blade. About 25 mg of staining pattern to bull sperm (Fig. 3), but after formaldehyde fixation, there was sometimes also spotty fluomuscle was extracted with SDS as described above. rescence on the principal piece. Triton X-100-treated SDS-PolyacrylamideGel Electrophoresis sperm exhibited decreased staining in the postacrosomal The sperm and muscle extracts were analyzed by SDS- region but retained intense fluorescence in the neck PAGE on 10-15% gradient gels (Laemmli, 1970). After region, annulus, and anterior midpiece (Fig. 4). Rabbit sperm displayed very intense fluorescence in electrophoresis, the proteins were visualized by the silver staining method of Wray et al. (1981)or transferred the sperm head, but it was restricted to a narrow band in the anterior part of the postacrosomal region. There to nitrocellulose for immunostaining. were also one or two fluorescent spots near the border Western Blotting and lmmunostaining between the equatorial and principal segments of the Proteins were transferred from polyacrylamide gels to acrosome. There was one or two small, brightly staining nitrocellulose a t 0.2 amp for 3 hours. The transfer buffer spots in the neck region, but the remainder of the tail consisted of 25 mM Tris, 192 mM glycine, 20% metha- was nonfluorescent (Fig. 5). After incubation of sperm in nol, pH 8.3 (Towbin et al., 1979). The blots were either Triton X-100, a similar pattern of staining was observed, but the intensity of staining in the postacrosomal region stained immediately or stored a t -20°C. For immunostaining, the blots were rinsed in PBS was diminished (Fig. 6). Cauda epididymal rat and mouse sperm were charac(+0.1% azide) and then incubated for 2 hours in a blocking solution which consisted of 2.5% bovine serum albu- terized by the absence of fluorescence in the sperm head. min, 5%NGS, 0.1% azide in PBS, pH 7.6. After rinsing Fluorescence was observed in the dorsal neck region, in PBS containing 1%NGS, 0.02% Tween 20, and 0.1% and the tail displayed spotty fluorescence in the midazide (PBSINGSPTW), they were incubated with con- piece and anterior principal piece. The intensity was stant agitation in a 1500 or 1000 dilution of the mono- greater in mouse sperm. There appeared to be regions of clonal antiactin antibody in PBS/NGS/TW for 60 increased intensity on either side of the annulus and minutes. The blots were then washed three times in sometimes in the anterior midpiece (Figs. 7, 9). Triton PBSINGSPTW and incubated for 60 minutes in a 1 5 0 0 X-100 extraction altered the distribution of fluorescence. dilution of biotinylated goat antimouse IgM Wector The midpiece displayed intense but even fluorescence, Labs., Burlingame, California). Following thorough while there was weaker fluorescence of the principal washing, they were treated with streptavidin-biotiny- piece (Figs. 8, 10). Ejaculated human sperm presented a markedly differlated horseradish peroxidase complex (Amersham Corporation, Arlington Heights, Illinois) for 60 minutes, ent distribution of fluorescence. In sperm fixed with rinsed, and developed with 0.05% diaminobenzadine, acetone, there was very intense fluorescence in the neck 0.0225% H202, 0.16% NiC12 in 50 mM Tris-HC1, pH 7.5. and principal piece of the tail, and occasionally weak, The specificity of the reactions was ascertained by patchy staining of the anterior border of the postacrosousing controls which were treated in a similar manner ma1 region (Fig. 9). Extraction with Triton X-100 did not but incubated in a solution which was devoid of the affect the fluorescence in the neck region or principal piece (Fig. 11). primary antibody. Guinea pig sperm displayed moderate, spotty fluoresRESULTS cence of the entire midpiece and anterior principal piece. lmmunofluorescence Localization of Actin Fluorescence was also present in the neck region (Figs. After staining with the monoclonal antiactin anti- 9, 12). Incubation in Triton X-100 resulted in loss of the body, sperm from all eight species showed a specific acrosome, but the pattern of fluorescence of the tail was pattern of fluorescence, but a universal pattern was not not markedly affected (Fig. 11). Acetone-fixed intact or Triton X-100-treated hamster evident. Unless otherwise stated, the staining pattern was the same after both types of fixation although the sperm had a n intensely fluorescent neck region and fluorescence was less intense after formaldehyde fixa- sometimes also staining at the level of the annulus tion. There were no apparent differences between 15-or (Figs. 8, 10, 11).Sperm fixed in formaldehyde displayed 60-minute incubations in Triton X-100. The intensity of less intense fluorescence in the neck region, but in adfluorescence varied between species, but it was constant dition, there was a strong band of fluorescence along the for each species during three replicates of each experi- anterior, concave margin of the hooked sperm head ment. Human sperm displayed the most intense fluores- (Fig. 13). The specific pattern of staining described above was cence, followed by hamster, rabbit, mouse, and rat sperm. Bull, boar, and guinea pig sperm exhibited less- never seen in control preparations. Control sperm incubated in PBS alone displayed weak autofluorescence of intense fluorescence. In bull sperm, there was strong fluorescence in the neck region, a t the annulus, and in the entire postacrosoma1 region; within the postacrosomal region there was Abbreviations a more intense band at its anterior margin (Fig. 1).After A Ann u 1us extraction in Triton X-100, the fluorescence in the post- G Guinea pig sperm Hamster sperm acrosomal region was reduced or almost abolished in H Human sperm most sperm, but the band of fluorescence at the border Hu Mouse sperm of the postacrosomal region with the equatorial segment M N Neck region was more resistant to this treatment. Fluorescence of P Principal piece Rat sperm the neck and annulus was maintained after this treat- R ACTIN IN MAMMALIAN SPERM Figs. 1-18. Matched phase-contrast (A) and fluorescence micrographs (B) of cells stained with the monoclonal antiactin antibody (Figs. 1-13, 15, 16) or NBD-phallacidin (Figs. 14, 17,18). x 1,860. Fig. 1. In acetone-fixed bull sperm, fluorescence is located in the neck region (N) and in the postacrosomal region; there is a band of fluorescence at the border of this region with the equatorial segment (arrowheads). Fig. 2. After treatment of bull sperm with Triton X-100, fluorescence 507 is reduced or abolished in the postacrosomal region, although the band of fluorescence at its anterior border is more resistent (arrowheads). The neck (N) and annulus (A) display strong fluorescence. Fig. 3. Acetone-fured boar sperm show strong fluorescence in the neck region (N), annulus (A), and postacrosomal region (arrowheads). Fig. 4. Treatment with Triton X-100 greatly reduces staining of the postacrosomal region (arrowheads) of boar sperm, but does not affect the fluorescence in the neck or annulus. 508 S.P. FLAHERTY, V.P. WINFREY, AND G.E. OLSON Fig. 5. Epididymal rabbit sperm display intense staining of the staining in the neck region or the spots in the acrosomal region anterior part of the postacrosomal region. There is also staining in the (arrowheads). neck region (N), and one or two brightly fluorescent spots at the border Fig. 7. Acetone-fixed rat sperm exhibit spotty fluorescence in the of the equatorial and principal segments of the acrosome (arrowheads). midpiece and anterior principal piece of the tail. This is most intense Fig. 6. Triton X-100 extraction of rabbit sperm reduces the intensity around the annulus (arrowheads),and there is a fluorescent spot in the of the fluorescence in the postacrosomal region but has no effect on dorsal neck region (N). the midpiece; the intensity varied between species and was most intense in rodent sperm. A similar pattern of weak fluorescence of the midpiece was observed in sperm incubated only in the secondary antibody. In addition, there was sometimes dull fluorescence of the sperm head and principal piece, but this was not consistently observed (Figs. 16, 18).The autofluorescence is not readily apparent in these figures because control micrographs were exposed at similar times to those used for recording specific fluorescence. Fibroblasts stained with the monoclonal antibody showed strong cytoplasmic fluorescence (Fig. 15). This was more intense around the cell periphery, where groups of filaments were present, and in thin cytoplasmic processes. Control cells showed weak nonspecific staining around the nucleus. Isolated skeletal muscle myofibrils stained intensely with the antibody. Fluorescence was localized primarily in the I bands. There was only a dull fluorescence of the myofibrils in control preparations. ACTIN IN MAMMALIAN SPERM 509 Fig. 8. Extraction of sperm with Triton X-100. Rat sperm (R) show neck region (N) of human sperm (Hu). The acrosome of guinea pig even staining of the midpiece and patchy staining of the principal sperm (G) has not been preserved by acetone fixation; there is weak piece; there is clear change in the pattern at the annulus (arrowheads). staining of the midpiece and anterior principal piece. Fluorescence is only located in the neck region (N) of hamster sperm Fig. 10. After incubation in Triton X-100, mouse sperm (MI display (H). even fluorescence in the midpiece and anterior principal piece; the Fig. 9. In acetone-fixed mouse sperm (M)there is intense staining of staining pattern changes abruptly at the annulus (arrowheads). Fluothe midpiece and principal piece; this is concentrated over the annulus rescence is restricted to the neck region (N) of hamster sperm (H). (arrowheads). Fluorescence is located in the principal piece (P) and 5 10 S.P. FLAHERTY, V.P. WINFREY, AND G.E. OLSON Fig. 11. After Triton X-100 extraction, fluorescence is located in the neck (N) of hamster sperm (HI, in the tail of guinea pig sperm (GI, and in the neck region (N) and principal piece (PI of human sperm (Hu!. Fig. 13. In hamster sperm fixed in formaldehyde, there is reduced fluorescence in the neck region (N), and intense staining along the concave margin of the sperm head (arrowheads). Fig. 12. Formaldehyde-fixed guinea pig sperm show the same pattern of staining as in Figure 9. Figs. 14,15. Fibroblasts stained with either NBD-phallacidin (Figure 14) or the antiactin antibody (Figure 15) show staining of stress fibres, the cell periphery and cytoplasmic extensions. ACTIN IN MAMMALIAN SPERM NBD-phallacidin Staining No specific staining with NBD-phallacidin was seen in sperm from any of the eight species. The patterns and intensity of fluorescence in the midpiece were identical to the autofluorescence exhibited by control sperm incubated in PBS (Figs. 17, 18). However, sperm incubated in NBD-phallacidin also showed very dull fluorescence in the entire principal piece, though this is not evident in Figure 17. Cultured fibroblasts displayed intense staining of stress fibers, and weak fluorescence of the remainder of the cytoplasm (Fig. 14). In control preparations, there was only weak autofluorescence of the perinuclear region of the cells. Skeletal muscle myofibrils stained with NBD-phallacidin showed intense fluorescence of the I and A bands. Electrophoretic and Western Blot Analysis of Sperm Proteins SDS extracts of bull, boar, rabbit, rat, mouse, hamster, and guinea pig sperm were analyzed by one-dimensional SDS-PAGE. The sperm extracts contained a large number of polypeptide bands, and a different pattern was evident in each species. In most extracts, there were several bands present in the 40-45-kD range (Fig. 19). SDS-PAGE-separated sperm and muscle extracts were transferred to nitrocellulose and immunostained with the monoclonal antiactin antibody. A single intensely stained 42-kD band was present in the muscle samples. Similarly, a distinct band of identical electrophoretic mobility reacted specifically with the monoclonal antibody in bull, boar, rabbit, rat, mouse, guinea pig, and hamster sperm (Fig. 20). Faint background bands were also present, but these appeared on the control blots which were not incubated with the primary antibody (not shown). In rat and mouse sperm there was a second band, of about 53 kD, which cross-reacted with the actin antibody. It was less intense than the actin band in mouse sperm but appeared as a more intense band than did actin in rat sperm (Fig. 20). There were no other such cross-reactive proteins in the sperm extracts from any of the other species. DISCUSSION In the present study, we have examined the localization of actin in sperm from eight species by immunofluorescence with a monoclonal antiactin antibody, and the configuration of actin has been studied by using the fluorescent probe NBD-phallacidin, which specifically binds filamentous actin (Barak et al., 1980). The immunofluorescence results suggest that actin is present in bull, boar, rabbit, human, rat, mouse, hamster, and guinea pig sperm. This was confirmed by the demonstration of a 42-kD protein which cross-reacted with the antibody on Western blots prepared from SDS-PAGEseparated sperm proteins. However, the intensity of fluorescence and the staining intensity of actin on Western blots suggest tiiat there may be marked species differences in the amount of actin present in sperm. This was not quantitated and may have been due to species specificity of the antibody or the availability of binding sites in sperm from the various species. The only universal site of actin localization was in the neck region. Four main patterns of localization were 511 found: 1)In bull, boar, and rabbit sperm, actin appeared to be localized primarily in the postacrosomal and neck regions. 2) In rat, mouse and guinea pig sperm, the sperm head did not appear to contain actin, but there was staining in the neck, midpiece, and anterior principal piece. 3) Hamster sperm seemed to contain actin only in the neck region. 4) Actin was localized in the neck region and principal piece of human sperm. In the present study, actin was found to occur primarily in the postacrosomal region of bull and boar sperm, thus confirming previous studies (Clarke and Yanagimachi, 1978; Tamblyn, 1980; Greenberg and Tamblyn, 1981). We found that it is also located in the neck, annulus, and anterior midpiece. Tamblyn (1980) also reported actin in the neck region of boar sperm. Several studies have also localized actin in the postacrosomal and neck regions of rabbit sperm (Clarke and Yanagimachi, 1978; Welch and O’Rand, 1985). Our results are in agreement with Welch and O’Rand (1985), who used the same monoclonal antibody to localize actin to the anterior half of the postacrosomal region. However, we also found actin in two discrete spots on the border of the principal and equatorial segments of the acrosome. These correspond closely to the location of the subacrosoma1 blisters (Phillips, 19721, suggesting that actin may be a component of the subacrosomal material. This is further supported by the resistence of these spots to Triton X-100 extraction. In human sperm, actin has previously been localized to the postacrosomal region (Clarke and Yanagimachi, 1978); to the midpiece, connecting piece, and posterior margin of the postacrosomal region (Clarke et al., 1982); or to the principal segment of the acrosome and principal piece of the sperm tail (Virtanen et al., 1984). In the present study, we found specific fluorescence in the principal piece and neck region of ejaculated human sperm. This variability may be due to differences in the specificity of antisera, and only one report (Virtanen et al., 1984) has identified human sperm actin on Western blots. In this study, we did not analyze human sperm extracts by Western blotting because high numbers of nonsperm cells are present in semen, and therefore one cannot be certain that the fluorescence we observed was due to actin. However, we recently found that only a 42kD polypeptide from human sperm stains with the antiactin antibody on Western blots (Flaherty and Olson, unpublished observation). Thus, although some of the actin may have originated from nonsperm cells, this finding suggests that proteins other than actin were probably not responsible for the fluorescence of human sperm. In rat and mouse sperm, we noted fluorescence only in the neck and tail. This confirms previous studies which have reported that although actin is located in the subacrosomal space of testicular rat sperm, it is absent from the head of epididymal sperm (Clarke and Yanagimachi, 1978; Campanella et al., 1979; Baccetti et al., 1980). However, one study also reported weak fluorescence in the postacrosomal region of mouse sperm (Clarke and Yanagimachi, 1978).A word of caution is required, however, concerning the distribution of actin in rat and mouse sperm, as two bands cross-reacted with the antibody on Western blots. The first was a 42-kD protein which comigrated with muscle actin, but another protein of 53 kD also stained, suggesting that only some of 512 S.P. FLAHERTY, V.P. WINFREY, AND G.E. OLSON Figs. 16-18. In control sperm incubated in PBS or only in the secondFig. 16. Boar sperm incubated only in the secondary antibody. ary antibody and in sperm treated with NBD-phallacidin there was only weak autofluorescence of the midpiece (arrowheads) and someFig. 17. Mouse (M) and human (Hu) sperm treated with NBDtimes dull fluorescence of the principal piece. This is not apparent in phallacidin. the following micrographs, which were exposed at times similar to those for Figures 1-12. Fig. 18. Control mouse (MI and human (Hu)sperm incubated in PBS. ACTIN IN MAMMALIAN SPERM Figs. 19, 20. Silver-stained gel (Fig. 19) and Western blot (Fig. 20) of SDS extracts of sperm. The positions of molecular weight standards are shown on the left. The lanes correspond to muscle extracts (1,9), or rat (2), mouse (31, guinea pig (41, hamster (5), rabbit (61, bull (7), and boar (8)sperm extracts. Fig. 19. In skeletal muscle extracts, there is a prominant band at 42 kD which corresponds to actin (arrowhead). There are bands of similar 513 electrophoretic mobility in some sperm extracts. Fig. 20. Western blots were immunostained with the antiactin antibody. A single 42-kD band stains in the muscle sample, and a reactive protein of identical molecular weight is present in each sperm extract (arrowheads). In rat and mouse sperm extracts, a second band of 53kD is also stained (arrowhead), The faint background bands all appeared on control blots. 514 S.P. FLAHERTY, V.P. WINFREY, AND G.E. OLSON the intense tail fluorescence was due to actin. In rat sperm extracts, this 53-kD protein reacted very intensely with the antibody, indicating that it may have been responsible for much of the fluorescence. However, neck fluorescence was seen in sperm from each species, and in bull, boar, rabbit, and hamster extracts, only a n actin band was immunolabeled on Western blots. This suggests that actin may be responsible for the fluorescence in the neck region of rat and mouse sperm. At this stage, however, this is speculative, but experiments are currently underway to identify and determine the distribution of actin and the 53-kD protein in rat sperm. It has been suggested that in hamster sperm, actin is present in the postacrosomal region (Clarke and Yanagimachi, 1978) or in the principal piece of the tail and along the concave margin of the sperm head (Talbot and Kleve, 1978). In this study, we localized actin only i n the neck region of hamster sperm. However, in formaldehyde-fixed sperm there was also a band of intense fluorescence along the concave margin of the hooked sperm head, as described by Talbot and Kleve (1978). This was not evident in acetone-fixed sperm. A cytoskeletal complex has recently been described in this region of hamster sperm, although the filaments were about twice the thickness of actin filaments (Olson and Winfrey, 1985). It is possible that the two types of filaments may share a common epitope, but this would not account for the absence of staining in this region after acetone fixation, and only a single actin band reacted specifically with the antiactin antibody on Western blots of hamster sperm extracts. The pattern of staining of sperm from the other species was not dependent on the method of fixation, although the fluorescence was generally more intense after acetone fixation. It has previously been reported, based on staining with rhodamine-phalloidin and immunoblot staining with a monoclonal antibody, that epididymal guinea pig sperm do not contain actin (Halenda et al., 1984). In this study, a faint actin band was detected in guinea pig sperm extracts on Western blots. Although care was taken to avoid contamination of sperm suspensions with other cell types, the suspensions were not purified, and it therefore remains possible that the actin band may have been due to cells other than sperm. However, we also found that guinea pig sperm displayed relatively weak, but specific, fluorescence of the tail, which seems to indicate that some actin is present in mature guinea pig sperm. In accord with several other studies (Virtanen et al., 1984; Halenda et al., 1984; Welch and O’Rand, 19851, we failed to find convincing evidence of filamentous actin in mature sperm of any of the eight species with the aid of NBD-phallacidin. The lack of staining is unlikely to have been due to technical problems since NBD-phallacidin is a low molecular weight compound which easily penetrates cells (Barak et al., 19801, and it specifically stained stress fibers in cultured fibroblasts and the actin-containing regions of muscle myofibrils. There was also some very weak fluorescence of the principal piece, but a t this stage, we are not sure if this was due to nonspecific staining, or to low amounts of filamentous actin. Nevertheless, the results indicate that all or most of the actin must be present in a nonfilamentous form, and it is doubtful that the filamentous arrays observed in various regions of the sperm head (Phillips, 1975; Koehler, 1978; Olson et al., 1983; Olson and Winfrey, 1985) are composed of actin filaments. Treatment of human and hamster sperm with Triton X-100 caused little change in the intensity or distribution of fluorescence, which suggests that although the actin appears to be in a nonfilamentous form, it must be associated with stable sperm structures such as the connecting piece and fibrous sheath. Indeed, there was specific fluorescence in the neck region of sperm from all species, suggesting that actin may be localized in this region of most mammalian sperm. However, Triton X-100 extraction resulted in partial or almost complete loss of fluorescence from other sperm domains. The pattern of staining of the tail of rat and mouse sperm was altered slightly by Triton X-100, but at this stage it is not possible to speculate whether this involved extraction of actin or the 53-kD protein from the region of the annulus. Staining of the midpiece was more even after extraction in Triton X-100 and was probably due to the exposure of antibody binding sites. The postacrosomal sheath contains a single layer of parallel filamentous elements and is closely linked to the plasma membrane by periodic ridges or projections (Pedersen, 1972; Fawcett, 1975; Olson et al., 1983). We have shown that actin is a component of this region in bull, boar, and rabbit sperm but not in rodent sperm. Triton X-100 extraction of bull, boar, and rabbit sperm caused a variable reduction of the fluorescence in the postacrosomal region. Since Triton X-100 removes the plasma membrane from this region but does not affect the structure of the sheath (Olson et al., 19831, it is probable that some of the actin is associated with the plasma membrane overlying the postacrosomal sheath, possibly in a manner similar to its association with the erythrocyte membrane (Marchesi, 1985). Hence, it may link the plasma membrane and sheath. The species-specific pattern of actin localization in mammalian spermatozoa would seem to preclude a universal function for this protein in the events of fertilization. Prior to fertilization, the spermatozoon must undergo two preliminary events-capacitation and the acrosome reaction (Yanagimachi, 1981). We have been unable to demonstrate the presence of actin in the head of mouse, rat, hamster, guinea pig, or human sperm, which suggests that actin may not participate in the surface changes which constitute a major part of the capacitation process (Friend et al., 1977; Yanagimachi, 1981; Bearer and Friend, 1982).Young and Cooper (1983) demonstrated that there is a progressive stabilization of the head-tail linkages in rabbit sperm during capacitation. This correlates with the development of a hyperactivated pattern of motility (Yanagimachi, 1981; Young and Cooper, 1983).The localization of actin in the neck region of sperm suggests that it may participate in this change in the properties of the head-tail junction. The acrosome reaction follows capacitation and consists of multiple point fusions between the plasma membrane and the outer acrosomal membrane, thereby releasing the acrosomal contents which aid or effect penetration of the extracellular coats around the oocyte (Yanagimachi, 1981). A number of studies have suggested that actin may become polymerized as a result of the acrosome reaction (Peterson et al., 1978; Ochs, 1984). It has been reported that 5-9-nm-thick microfilaments of unknown composition appear in the region of vesiculation ACTIN IN MAMMALIAN SPERM after induction of the boar sperm acrosome reaction with ionophore A23187 (Peterson et al., 1978). However, in agreement with a n earlier study (Tamblyn, 1980), we found no evidence for actin, either monomeric or filamentous, in the acrosomal region of intact epididymal boar sperm. Hence, if the filaments reported by Peterson et al. (1978)were indeed actin, then the actin must have been transported from another location in the spermatozoon during capacitation. The potential occurrence and function of actin polymerization remains to be clarified. In this study, we have demonstrated that there is a species-specific distribution of actin in bull, boar, rabbit, human, rat, mouse, hamster, and probably guinea pig sperm. The function of actin in mammalian sperm is still unclear, but this study has provided the basis for future experiments which could aim to determine the ultrastructural localization of actin, any changes in its distribution or configuration that occur during the events of fertilization, and whether the variable pattern of actin localization is of functional significance. ACKNOWLEDGMENTS The authors would like to thank Tom Noland, Tina Krivacka, Mary Aakre, and Michael Holland for assistance with aspects of this work, and Vera Murphy for assistance in preparing the manuscript. This work was supported by NIH grant HD-20419, NlH Center Grant HD-05797, and a grant from the Mellon Foundation. LITERATURE CITED Baccetti, B., E. Bigliardi, and A.G. Burrini (1980) The morphogenesis of the vertebrate perforatorium. J. Ultrastruct. Res., 71r272-287. Barak, L.S., R.R. Yocum, E.A. Nothnagel, and W.W. Webb (1980)Fluorescence staining of the actin cytoskeleton in living cells with 7nitrobenz-2-oxa-l,3-diazole-phallacidin. Proc. Natl. Acad. Sci. USA, 77t980-984. Bearer, E.L., and D.S. Friend (1982) Modifications of anionic lipid domains preceding membrane fusion in guinea pig sperm. J. Cell Biol., 92r604-615. Campanella, C., G. Gabbiani, B. Baccetti, A.G. Burrini, and V. Pallini (1979) Actin and myosin in the vertebrate acrosomal region. J. Submicrosc. Cytol., llr53-71. Clark, M., and J.A. Spudich (1977) Non-muscle contractile proteins: The role of actin and myosin in cell motility and shape determination. Ann. Rev. Biochem., 46r797-822. Clarke, G.N., and R. Yanagimachi (1978) Actin in mammalian sperm heads. J. Exp. Zool., 205t125-132. Clarke, G.N., F.M. Clarke, and S. Wilson (1982) Actin in human spermatozoa. Biol. Reprod., 26r319-327. Fawcett, D.W. (1975) The mammalian spermatozoon. Dev. Biol., 44:394-436. Flaherty, S.P. (1986) Structural and Functional Aspects of Sperm Head Morphology in the Plains Mouse, Pseudomys australis. PhD Dissertation, The University of Adelaide. Flaherty, S.P., W.G. Breed, and V. Sarafis (1983) Localization of actin in the sperm head of the plains mouse, Pseudomys australis. J. Exp. Zool., 225:497-500. 5 15 Friend, D.S., L. Orci, A. Perrelet, and R. Yanagimachi (1977) Membrane particle changes attending the acrosome reaction in guinea pig spermatozoa. J. Cell Biol., 74r561-577. Greenberg, B.J., and T.M. Tamblyn (1981) Actin from mature bovine spermatozoa. J. Cell Biol., 91t191A (Abstr.). Halenda, R.M., P.M. Primakoff, and D.G. Myles (1984) Evidence for the presence of actin in guinea pig spermatogenic cells and its loss prior to formation of mature sperm. J. Cell Biol., 99;394A (Abstr.). Jacobson, B.S. (1983) Interaction of the plasma membrane with the cytoskeleton: An overview. Tissue Cell, 15t829-852. Knight, P.J., and J.A. Trinick (1982) Preparation of myofibrils. Methods Enzymol., 85t9-12. Koehler, J.K. (1978) Observations on the fine structure of vole spermatozoa with particular reference to cytoskeletal elements in the mature sperm head. Gamete Res., lt247-257. Korn, E.D. (1978) Biochemistry of actomyosin-dependent cell motility (A review). Proc. Natl. Acad. Sci. USA, 75r588-599. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, 227t680-685. Marchesi, V.T. (1985) Stabilizing infrastructure of cell membranes. Ann. Rev. Cell Biol. 1t531-561. Ochs, D. (1984) Actin in post-ejaculatory human sperm. Biol. Reprod. 30: Suool.. 1:78 (Abstr.). Ochs, D., ahd’D.P. Wolf (1985)Actin in ejaculated human sperm cells. Biol. Reprod., 33t1223-1226. Olson, G.E., and V.P. Winfrey (1985) Substructure of a cytoskeletal complex associated with the hamster sperm acroaome. J. Ultrastruct. Res., 92r167-179. Olson, G.E., T.D. Noland, V.P. Winfrey, and D.L. Garbers (1983) Substructure of the postacrosomal sheath of bovine spermatozoa. J. Ultrastruct. Res., 85t204-218. Pedersen, H. (1972) The postacrosomal region of the spermatozoa of man and Macaca arctoides. J. Ultrastruct. Res., 4Or366-377. Peterson, R.N., L. Russell, D. Bundman, and M. Freund (1978) Presence of microfilaments and tubular structures in boar spermatozoa after chemically inducing the acrosome reaction. Biol. Reprod., 19:459-466. Phillips, D.M. (1972) Substructure of the mammalian acrosome. J. Ultrastruct. Res., 38t591-604. Phillips, D.M. (1975) Cell surface structure of rodent sperm heads. J. Exp. Zool., 191.1-8. Talbot, P., and M.G. Kleve (1978)Hamster sperm crossreact with antiactin. J. Exp. Zool., 204r131-136. Tamblyn, T.M. (1980) Identification of actin in boar epididymal spermatozoa. Biol. Reprod., 22t727-734. Tamblyn, T.M. (1981)Evidence for non-muscle myosin in bovine ejaculated spermatozoa. Gamete Res., 4r499-506. Towbin, H., T. Staehelin, and J. Gordin (1979) Electrophoretic transfer of proteins from polyacrylamide to nitrocellulose sheets: Procedure and some applications. Proc. Natl. Acad. Sci. USA, 76r4350-4354. Virtanen, I., R.A. Badley, R. Paasivuo, and V.P. Lehto (1984) Distinct cytoskeletal domains revealed in sperm cells. J. Cell Biol., 99r10831091. Welch, J.E., and M.G. O’Rand (1985) Identification and distribution of actin in spermatogenic cells and spermatozoa of the rabbit. Dev. Biol. 159t411-417. Wray, W., T. Boulikas, V.P. Wray, and R. Hancock (1981) Silver staining of proteins on polyacrylamide gels. Anal. Biochem., 118r197203. Yanagimachi, R (1981) Mechanisms of fertilization in mammals. In: Fertilization and Embryonic Development In Vitro. L., Mastroianni and J.D. Biggers, eds. Plenum Press, New York, pp. 81180. Young, R.J., and G.W. Cooper (1983) Stabilization in vivo and in vitro of the spermatozoon head and tail to detachment induced by primary amine, thiol and detergent. Gamete Res., 7:277-288.