Temporal expression of membrane antigens during mouse spermatogenesis.

The temporal expression of cell surface antigens during mammalian spermatogenesis has been investigated using isolated populations of mouse germ cells. Spermatogenic cells at advanced stages of differentiation, including pachytene primary spermatocytes, round spermatids, and residual bodies of Regaud and mature spermatozoa, contain common antigenic membrane components which are not detected before the pachytene stage of the first meiotic prophase. These surface constituents are not detected on isolated populations of primitive type A spermatogonia, type A spermatogonia, type B spermatogonia, preleptotene primary spermatocytes, or leptotene and zygotene primary spermatocytes. These results have been demonstrated by immunofluorescence microscopy, by complement-mediated cytotoxicity, and by quantitative measurements of immunoglobulin (Ig) receptors on the plasma membrane of all cell populations examined. The cell surface antigens detected on germ cells are not found on mouse thymocytes, erythrocytes, or peripheral blood lymphocytes as determined by immunofluorescence and by cytotoxicity assays. Furthermore, absorption of antisera with kidney and liver tissue does not reduce the reactivity of the antibody preparations with spermatogenic cells, indicating that these antigenic determinants are specific to germ cells. This represents the first direct evidence for the ordered temporal appearance of plasma membrane antigens specific to particular classes of mouse spermatogenic cells. It appears that at late meiotic prophase, coincident with the production of pachytene primary spermatocytes, a variety of new components are inserted into the surface membranes of developing germ cells. The further identification and biochemical characterization of these constituents should facilitate an understanding of mammalian spermatogenesis at the molecular level.

Generation of flagella by cultured mouse spermatids.

During the short-term culturing of mouse spermatogenic cells, flagella were generated by round spermatids previously lacking tails. Unseparated germ cells were obtained by enzymatic treatments and round spermatids (greater than 90% pure) were purified by unit gravity sedimentation. As determined by Nomarski or phase-contrast microscopy, no cells had flagella immediately after isolation; flagella were first clearly detected after 6 1/2 h of culture in Eagle's minimal essential medium containing 10% fetal bovine serum and 6 mM lactate. After 24 h, approximately 20% of round spermatids had formed flagella. Multinucleated round spermatids often formed multiple flagella, the number never exceeding the number of nuclei per symplast. Round spermatids were the only spermatogenic cells capable of tail formation. Flagella elongation was blocked by 1 microM demecolcine, an inhibitor of tubulin polymerization. Indirect immunofluorescence localized tubulin in the flagella. As seen by scanning electron microscopy, flagella developed as early as 2 h after culture and continued to elongate over the next 20 h, reaching lengths of at least 19 micron. Transmission electron microscopy demonstrated that flagella formed in culture resembled flagella from Golgi-phase round spermatids in situ; the flagella consisted of "9+2" axonemes lacking other accessory structures such as outer dense fibers and the fibrous sheath. As determined by acridine orange staining of the developing acrosomes, all spermatids that formed flagella in culture were Golgi-phase spermatids. By these criteria, the structures are indeed true flagella, corresponding in appearance to what others have described for early mammalian spermatid flagella in situ. We believe this is the first substantiated report of limited in vitro differentiation by isolated mammalian spermatids.

The development of regionalized lipid diffusibility in the germ cell plasma membrane during spermatogenesis in the mouse.

The lipids and proteins of sperm cells are highly regionalized in their lateral distribution. Fluorescence recovery after photobleaching studies of sperm membrane component lateral diffusibility have shown that the sperm plasma membrane is also highly regionalized in the extents and rates of diffusion of its surface components. These studies have also shown that regionalized changes in lateral diffusibility occur during the differentiative processes of epididymal maturation and capacitation. Unlike mammalian somatic cells, sperm cells exhibit large nondiffusing lipid fractions. In this paper, we will show that both regionalized lipid diffusibility and nondiffusing lipid fractions develop with the morphogenesis of cell shape during spermatogenesis in the mouse. Pachytene spermatocytes and round spermatids show diffusion rates and the nearly complete recoveries (80-90%) typical of mammalian somatic cells. In contrast, stage 10-11 condensing spermatids, testicular spermatozoa, cauda epididymal spermatozoa, as well as the anucleate structures associated with these later stages of spermatogenesis (residual bodies and the cytoplasmic droplets of condensing spermatids and testicular spermatozoa), exhibit large nondiffusing fractions. Both the diffusion rates and diffusing fractions observed on the anterior and posterior regions of the head of stage 10-11 condensing spermatids are the same as the values obtained for these regions on testicular spermatozoa. Possible mechanisms of lipid immobilization and possible physiological implications of this nondiffusing lipid are discussed.

Sertoli cell binding to isolated testicular basement membrane.

We have examined the adhesion of primary Sertoli cells to a seminiferous tubule basement membrane (STBM) preparation in vitro. The STBM isolation procedure (Watanabe, T.K., L.J. Hansen, N.K. Reddy, Y.S. Kanwar, and J.K. Reddy, 1984, Cancer Res., 44:5361-5368) yields segments of STBM that retain their histotypic form in both three-dimensional tubular geometry and ultrastructural appearance. The STBM sleeves contain two laminae: a thick, inner basal lamina that was formed in vivo between Sertoli cells and peritubular myoid cells; and a thinner, outer basal lamina that was formed between myoid cells and sinusoidal endothelial cells. Characterization by immunofluorescence and SDS PAGE revealed that the isolated STBM retained fibronectin, laminin, and putative type IV collagen among its many components. When the STBM sleeves were gently shaken with an enriched fraction of primary Sertoli cells, the Sertoli cells bound preferentially to the lumenal basal lamina at the ends of the STBM sleeves. Few Sertoli cells bound to either the outer basal lamina of the STBM sleeves or to vascular extracellular matrix material which contaminated the STBM preparation. 3T3 cells, in contrast, bound to all surfaces of the STBM sleeves. Pretreatment of the STBM sleeves with proteases, 0.1 M Na metaperiodate, 4 M guanidine HCl, or heating to 80 degrees-90 degrees C inhibited lumenal Sertoli cell binding, but binding was not inhibited by chondroitinase ABC, heparinase, hyaluronidase, or 4 M NaCl. The lumenal Sertoli cell binding occurred in the presence or absence of added soluble laminin, but not fibronectin. The addition of soluble laminin, but not fibronectin, restored random binding of Sertoli cells to trypsinized STBM sleeves. Our in vitro model system indicates that Sertoli cells recognize differences in two basal laminae produced in vivo on either side of myoid cells.

Chemical dissection of mammalian spermatozoa.

Spermatozoa from several mammalian species have been dissected by chemical methods to yield free heads, tails with attached midpieces, and tails from which the mitochondrial components of the midpiece were removed. Mouse and rat spermatozoa were cleaved by brief treatment with trypsin to yield free heads and tails, while human, guinea pig, and rabbit spermatozoa were cleaved by trypsin only after incubation with 2-mercaptoethanol or dithiothreitol. Spermatozoa were also cleaved at the junction of the head and the tail by treatment with acid and base. Mitochondria were removed from intact spermatozoa or isolated tails by mechanical shear after treatment with 2-mercaptoethanol or dithiothreitol. The dissected components of spermatozoa were fractionated with good yield and high purity by density gradient centrifugation. Ultrastructural analysis indicates that proteolytic cleavage to yield separated heads and tails occurs at a specific location in the neck of the spermatozoon, leaving the basal plate attached to the head of the cell. In contrast, after acid cleavage the basal plate remains with the midpiece. Proteolytic treatment has no apparent effect on any other spermatozoan structures, whereas acid or base treatment results in damage to the plasma membrane, the acrosome, and other structures. The specificity of the proteolytic cleavage suggests that a particular protein or group of proteins may be responsible for the linkage between the sperm head and tail.

Spermatogenic cells of the prepuberal mouse. Isolation and morphological characterization.

A procedure is described which permits the isolation from the prepuberal mouse testis of highly purified populations of primitive type A spermatogonia, type A spermatogonia, type B spermatogonia, preleptotene primary spermatocytes, leptotene and zygotene primary spermatocytes, pachytene primary spermatocytes and Sertoli cells. The successful isolation of these prepuberal cell types was accomplished by: (a) defining distinctive morphological characteristics of the cells, (b) determining the temporal appearance of spermatogenic cells during prepuberal development, (c) isolating purified seminiferous cords, after dissociation of the testis with collagenase, (d) separating the trypsin-dispersed seminiferous cells by sedimentation velocity at unit gravity, and (e) assessing the identity and purity of the isolated cell types by microscopy. The seminiferous epithelium from day 6 animals contains only primitive type A spermatogonia and Sertoli cells. Type A and type B spermatogonia are present by day 8. At day 10, meiotic prophase is initiated, with the germ cells reaching the early and late pachytene stages by 14 and 18, respectively. Secondary spermatocytes and haploid spermatids appear throughout this developmental period. The purity and optimum day for the recovery of specific cell types are as follows: day 6, Sertoli cells (purity>99 percent) and primitive type A spermatogonia (90 percent); day 8, type A spermatogonia (91 percent) and type B spermatogonia (76 percent); day 18, preleptotene spermatocytes (93 percent), leptotene/zygotene spermatocytes (52 percent), and pachytene spermatocytes (89 percent), leptotene/zygotene spermatocytes (52 percent), and pachytene spermatocytes (89 percent).

Specific expression of nuclear proto-oncogenes before entry into meiotic prophase of spermatogenesis.

The expression of proto-oncogenes representative of several functional categories has been investigated during development of mouse male germ cells. The c-raf proto-oncogene and three members of the c-ras gene family were expressed in mitotically active stem cells, throughout the prophase of meiosis and to varying extents in post-meiotic cell types. In contrast, the nuclear proto-oncogenes c-fos, c-jun, and c-myc were specifically expressed at high levels in type B spermatogonia. High levels of c-myc and c-jun RNAs were also detected in spermatocytes early in the prophase of meiosis. The type B spermatogonia represent the last mitotic cell division before entry into meiotic prophase; therefore, these nuclear proto-oncogenes may be involved in altering programs of gene expression at this developmental transition.

DNA methylation and demethylation events during meiotic prophase in the mouse testis.

The genes encoding three different mammalian testis-specific nuclear chromatin proteins, mouse transition protein 1, mouse protamine 1, and mouse protamine 2, all of which are expressed postmeiotically, are marked by methylation early during spermatogenesis in the mouse. Analysis of DNA from the testes of prepubertal mice and isolated testicular cells revealed that transition protein 1 became progressively less methylated during spermatogenesis, while the two protamines became progressively more methylated; in contrast, the methylation of beta-actin, a gene expressed throughout spermatogenesis, did not change. These findings provide evidence that both de novo methylation and demethylation events are occurring after the completion of DNA replication, during meiotic prophase in the mouse testis.

Gli family members are differentially expressed during the mitotic phase of spermatogenesis.

The Gli family of DNA binding proteins has been implicated in multiple neoplasias and developmental abnormalities, suggesting a primary involvement in cell development and differentiation. However, to date their specific roles and mechanisms of action remain obscure, and a drawback has been the lack of a model system in which to study their normal function. Here we demonstrate that Gli family members are differentially expressed during spermatogenesis in mice. Specifically, Gli and Gli3 mRNAs were detected in mouse germ cells, while Gli2 was not. Further, both Gli and Gli3 exhibited stage-dependent patterns of expression selectively in type A and B spermatogonia. Gli expression was somewhat higher in type B spermatogonia while the abundance of Gli3 transcripts was similar in type A and B cells. Gel-shift analyses also demonstrated the enrichment of DNA binding activity specific for the Gli target sequence in spermatogonial cells. These results indicate a selective role for Gli and Gli3 during mitotic stages of male germ cell development. Spermatogenesis may thus provide a unique opportunity to identify downstream targets and explore the normal function of Gli family proteins.

Identification of the protein product of the c-mos proto-oncogene in mouse testes.

The mouse c-mos proto-oncogene RNA is expressed primarily in mouse gonadal tissues and embryos. Until now, the c-mos protein has not been identified. Utilizing two different site-directed affinity purified anti-peptide antibodies, we have identified a 43 kDa c-mos protein in mouse testes and in germ cell preparations derived from testes. This 43 kDa testicular protein was found to be structurally related to a bacterially expressed c-mos protein by peptide mapping. Immunoblots of whole mouse sections were employed to establish that the c-mos protein is expressed primarily in the testes.

Transcription factor Sp1 is expressed by three different developmentally regulated messenger ribonucleic acids in mouse spermatogenic cells.

Gene expression during spermatogenesis is highly cell- and stage-specific and involves the complex interplay of multiple developmentally regulated transcription factors. Recent evidence suggests that the DNA-binding protein Sp1 functions as an important trans-activator during cell development and differentiation. In the present study, the developmental expression of Sp1 was characterized during mouse spermatogenesis. Three distinct Sp1 transcripts were detected in mouse spermatogenic cells, each with a distinct developmental pattern; an 8.2-kilobase (kb) messenger RNA (mRNA) identical in size to the somatic mRNA expressed in spermatogonial cells, a larger mRNA approximately 8.8 kb in size present in meiotic cells, and a 2.4 kb mRNA in meiotic and postmeiotic germ cells. The 8.8- and 2.4-kb Sp1 transcripts were not observed in somatic cells and, thus, are male germ cell specific. Northern, ribonuclease protection, and RT-PCR assays revealed that the 2.4-kb Sp1 transcript is truncated in both the 5'- and 3'-untranslated regions relative to the somatic mRNA and lacks a short segment of the N-terminal coding region. Polysome analysis further indicated that these germ cell-specific Sp1 mRNAs are translated, albeit with a lower efficiency than the somatic transcript. Consistent with these results, spermatogenic cells were shown to contain approximately 9-fold lower concentrations of Sp1 proteins that are approximately the same size as the somatic form. Of particular interest, the apparent affinity of Sp1 DNA-binding activity in nuclear extracts from mouse germ cells was 5-fold greater than that in extracts from mouse somatic tissues. This may reflect the existence of mechanisms within mouse spermatogenic cells that compensate for the lower nuclear concentrations of Sp1 protein. These results suggest that cell- and stage-specific regulation of Sp1 gene expression and activity may be an important component of the mouse spermatogenic cell developmental program.

Germ cells of the mouse testis express P450 aromatase.

Estrogen production within the testis has been a subject of considerable controversy for many years. Several studies have shown that both Sertoli and Leydig cells produce estrogen during different stages of development. Therefore, we have conducted experiments to localize aromatase, a cytochrome P450 enzyme that converts androgen to estrogen, within the testis. First, P450 aromatase (P450arom) was localized in germ cells of the adult mouse testis by immunocytochemistry, using an antiserum generated against purified human placental cytochrome P450arom. In the germinal epithelium, P450arom was located primarily in the Golgi region of round spermatids, throughout the cytoplasm of elongating spermatids, and along the flagella of late spermatids. Second, localization of P450arom within the germinal epithelium was supported by Western blot analysis of isolated germ cells. Third, Northern blot analysis using a mouse P450arom cDNA probe indicated that the mRNA for the mouse P450arom was present in testicular germ cells. Fourth, P450arom activity was measured in germ cells by the 3H2O water assay. Based upon these observations, we conclude that germ cells are a site of estrogen synthesis in the adult mouse testis.

Fucosylation events during mammalian spermatogenesis.

It will be necessary to conduct further studies to establish more precisely the localization of FT on mouse male germ cells. Antibodies to FTs are not yet available, so an immunocytochemical approach is not currently feasible. Additional cell fractionation protocols can be designed to compare plasma membrane fractions with enriched fractions of Golgi apparatus and to compare directly the activities of multiple glycosyltransferase enzymes and Golgi-specific markers in these preparations. Schachter et al. and Nyquist and colleagues have already provided experimental techniques for the isolation of Golgi fractions of good purity from rodent pachytene spermatocytes and spermatids. Ample opportunity exists, then, for a detailed analysis of the number, specificity, and localization of FT enzymes during mammalian spermatogenesis. All available data imply that these enzymes will prove to be vital components in the differentiation of cells within the seminiferous epithelium.

Glycosidic specificity of fucosyltransferases present in rat epididymal spermatozoa.

We have recently demonstrated multiple fucosyltransferase (FT) activity in rat spermatogenic cells. To complement these findings, here we identify and partially characterize the glycosidic linkage specificity of FTs present in spermatozoa from caput and cauda epididymides. Analysis of the acceptor substrate specificity of the FTs by thin-layer chromatography indicated that both caput and cauda sperm expressed alpha(1-2)-, alpha(1-3)-, alpha(1-4)-FTs as demonstrated by fucose incorporation into phenyl-beta-D-galactoside, 2'-fucosyllactose, and lacto-N-fucopentaose-I, respectively. Spermatozoa from the cauda epididymidis exhibited significant decreases in the levels of alpha(1-2)-, alpha(1-3)-, alpha(1-4)-FTs, and of total soluble FTs in comparison to spermatozoa from the caput epididymidis. The relative ratio of alpha(1-3)-FT to total FT activity appeared to be significantly higher than those of alpha(1-2)- or alpha(1-4)-FTs, in spermatozoa both from caput and cauda epididymides. Using different types of low molecular weight acceptors and the selective inhibition of the FT by N- ethylmaleimide, we have demonstrated that at least alpha(1-2)-FT is different from alpha(1-3)- or alpha(1-4)-FTs. Kinetic studies also showed that alpha(1-2)-FT is different from alpha(1-3)- or alpha(1-4)-FTs as demonstrated by apparent Km and Vmax values. Moreover, alpha(1-3)- and alpha(1-4)-FT activities in cauda sperm were found to be highly sensitive to Mn2+ but showed differential responses to divalent cations. In contrast, both alpha(1-3)- and alpha(1-4)-FTs seemed to be relatively less sensitive to Mg2+. Thus, these results not only demonstrate the presence of multiple FTs in rat epididymal sperm but also differentiate individual FTs with regard to their kinetic properties and sensitivity to both inhibitor and divalent cations.

Sertoli cell plasma membrane polypeptides involved in spermatogenic cell-Sertoli cell adhesion.

This study concerns Sertoli cell-spermatogenic cell adhesive interactions in the seminiferous tubule. Sertoli cell surface polypeptides involved in germ cell-Sertoli cell adhesion were identified by serological inhibition of an in vitro Sertoli-germ cell adhesion assay. This assay was modified from a previously reported adhesion assay, and employs a scanning laser cytometer for quantification of adherent cells. Reactivity of the polyclonal antiserum raised against rat Sertoli cells was also assessed via immunofluorescent microscopy. The addition of antiserum to the adhesion assay resulted in a 42% to 66% inhibition of cell-cell adhesion. Moreover, preincubation of antiserum with Sertoli cell monolayers resulted in a significant reduction of spermatogenic cell binding. Conversely, preincubation of antiserum with germ cells resulted in no reduction. Western blot analysis of the antiserum against purified Sertoli cell membranes indicated reactivity with four polypeptides. The data suggest that one or more of these polypeptides are directly involved in the adhesion of germ cells to Sertoli cell monolayers in vitro.

Toxic effects of polychlorinated biphenyls on cultured rat Sertoli cells.

Polychlorinated biphenyls (PCBs) are ubiquitous and persistent environmental contaminants. In mammals, PCBs affect spermatogenesis and may be associated with Sertoli cell changes. Therefore, our aim was to evaluate in vitro toxic effects of hydroxylated PCB (PCB-22; 2',3',4',5'-tetrachloro-4-biphenylol) and PCB congener (PCB-77; 3,3',4,4'-tetrachlorobiphenyl) on Sertoli cells isolated from 19- to 21-day-old male rats. Sertoli cells incubated for 24 hours in 10(-7) M PCB-22 and 10(-8) M PCB-77, but not in 0.05% ETOH or 10(-7) M 17beta-estradiol (E2) showed morphological changes. Sertoli cells demonstrated progressive damage with higher concentrations of PCB-77 (10(-7) M). After 24 hours, 10(-7) M PCB-22 killed 20% of the Sertoli cells and equimolar PCB-77 killed 45% of the Sertoli cells in culture. At 10(-8) M, PCB-22 did not kill any significant number of Sertoli cells, whereas PCB-77 killed 40% of cells in culture. This result showed differential effects of PCB compounds, with PCB-77 being more cytotoxic than PCB-22 as tested on Sertoli cells. Because PCB-77 produced greater toxic effects, we further tested this congener on Sertoli cell lactate production. After 24 hours, lactate production by Sertoli cells treated with 10(-7) or 10(-8) M PCB-77 was significantly increased. Finally, Sertoli cells exposed to 10(-7) M PCB-77 showed disorganized and less intense F-actin staining. The results demonstrate that PCBs, but not E2, is directly toxic to Sertoli cells in vitro, and suggest this toxic effect is independent of estrogenic action.

Surface-associated glycosyltransferase activities in rat Sertoli cells in vitro.

We have previously demonstrated fucosyltransferase (FT) activity on mouse germ cell surfaces at different stages of spermatogenesis. To complement these findings, here we report FT activity on the Sertoli cell (SC) surface. SC isolated and cultured from 20-day-old rat testes displayed FT activity with a Vmax of 12.5 pmoles/mg protein/min and a Km of 22 microM, while purified Sertoli cell plasma membranes (SCPM) showed FT activity with a Vmax of 10 pmoles/mg protein/min and a Km of 18.2 microM for GDP-[14C]-L-fucose. Fucosyltransferase activities were 16.7 and 2.6 pmoles/mg protein/min in SC and SCPM, respectively; approximately 16% of FT activity is, therefore, on the cell surface. To test whether the expression of FT activity in SC was regulated by hormones and growth factors, SC were cultured in serum-free medium supplemented with insulin, transferrin, sodium selenite, and epidermal growth factor (medium 4F) or in 4F plus follicle-stimulating hormone, testosterone, hydrocortisone, and vitamin E (medium 8F). We found that FT activity in SC is not modulated by these hormones or growth factors (4F or 8F). For comparison with FT, galactosyltransferase (GalTase) activities in SC and SCPM were also determined. SC displayed GalTase activity with a Vmax of 50 pmoles/mg protein/min and a Km of 38.5 microM, while SCPM showed GalTase activity with a Vmax of 25 pmoles/mg protein/min and a Km of 20.8 microM for UDP-[3H]-galactose. Galactosyl-transferase activities were 29.2 and 9.6 pmoles/mg protein/min in SC and SCPM, respectively. Therefore, approximately 33% of the total cell GalTase activity was detected on the surface membranes of rat Sertoli cells.(ABSTRACT TRUNCATED AT 250 WORDS)

Cell surface marker proteins during mouse spermatogenesis: two-dimensional electrophoretic analysis.

Purified plasma membranes isolated from separated highly homogenous populations of mouse pachytene spermatocytes, round spermatids (step I-8), and residual bodies have been compared using 2-dimensional polyacrylamide gel electrophoresis. Two polypeptides apparently specific to pachytene spermatocytes have been identified. Component Pa has a molecular weight of 90 k daltons (K) and pI of 5.6. Component Pb has a molecular weight of 56.5 K and a pI of 6.0. Four polypeptides detected only in plasma membranes of round spermatids have been identified as follows: RSa, 90-95 K and pI 5.9; RSb, also 90-95 K and pI 5.9; RSc, approximately 88 K and pI 5.5; RSd, 58 K and pI 6.0-6.3. No polypeptides unique to residual body membranes were identified. Short-term culture experiments have established that separated adult mouse spermatogenic cells survive short-term culture in vitro. These cells actively synthesize numerous cellular proteins as determined by the incorporation of [3H]leucine. Investigations concerning the effect of the cell separation procedure on mouse spermatogenic cell membranes indicate that only 7 of 110-120 total plasma membrane constituents are degraded enzymically during cell purification. Only one of these constituents may correspond to the presumptive cell differentiation markers described for pachytene spermatocytes and round spermatids. These results indicate, therefore, that plasma membranes obtained immediately after cell separation are suitable for the detailed biochemical analysis of the most integral surface proteins during spermatogenesis in the mouse.

Presence of multiple fucosyltransferases in rat Sertoli cells and spermatogenic cells.

Differential expression of fucosyltransferases (FTs) on Sertoli cell and germ cell surfaces and their function as ectoenzymes may be important in the process of spermatogenesis. To determine the glycosidic linkage specificity of FTs present in cultured Sertoli cells and in germ cells, we quantified FT activities by thin-layer chromatography using both high and low molecular weight acceptors in the presence of GDP-[14C]-L-fucose. Analysis of the acceptor substrate specificity of the FTs indicated that alpha(1-2), alpha(1-3), alpha(1-4)-FTs are expressed as demonstrated by fucose incorporation into phenyl-beta-D-galactoside, 2'-fucosyllactose, and lacto-N-fucopentaose-I, respectively. In Sertoli cells, the ratios of the three FTs examined were the same for whole-cell extracts and samples of purified plasma membranes. Higher relative FT activity was observed in plasma membranes from mixed germ cells than in Sertoli cell membranes. Furthermore, alpha(1-3)-FT and alpha(1-4)-FT activities were higher in mixed germ cell membranes. Spermatogenic stage specificity of FT expression was assessed in purified populations of germ cells. With calculation on a per-cell basis, all three alpha-FTs exhibited a quantitative decrease during the transition between pachytene spermatocytes and round spermatids. The decrease in alpha(1-3)-FT activity was particularly significant. In rat germ cells, all three alpha-FT activities associated with the cell surface in pachytene spermatocytes and round spermatids were 34-53% and 52-53%, respectively, of the total cell FT activity.(ABSTRACT TRUNCATED AT 250 WORDS)