Isolation and characterization of Streptoverticillium anticoagulant (SAC), a novel protein inhibitor of blood coagulation produced by Streptoverticillium cinnamoneum subsp. cinnamoneum

Three Streptoverticillium anticoagulants, SAC I, II, and III, which strongly inhibit human intrinsic blood coagulation, were each isolated in a homogeneous form from a culture fluid of Streptoverticillium cinnamoneum subsp. cinnamoneum IFO 12852. SAC I, II, and III are simple proteins with molecular weights of around 12,000, and with isoelectric points of 9.7, 9.7, and 9.9, respectively. Their amino acid compositions are similar and each SAC possesses two disulfide bonds. The COOH-terminal residue of each of these proteins is phenylalanine. Together with the similarity of their protein chemical properties, the results of NH2-terminal amino acid sequence analysis of these SAC proteins strongly suggested that the deletion of Ser-Leu and Ser-Leu-Tyr from the NH2-terminus of SAC I (Ser-Leu-Tyr-Ala-Pro-...) results in the generation of SAC II and III, respectively. The amount of each SAC necessary to double the partial thromboplastin time was around 5 micrograms/ml. SAC I inhibited activated human factor XII and human plasma kallikrein. It also inhibited, but to a lesser extent, activated factor X. The inhibition constants (Ki) of SAC I toward activated factor XII and plasma kallikrein were 5.3 x 10(-8) and 7.2 x 10(-9) M, respectively. The SACs also inhibited some microbial serine proteases such as subtilisin Carlsberg and, to a lesser extent, mammalian serine proteases including bovine trypsin and alpha-chymotrypsin. Of these three inhibitors, only SAC I inhibited metalloproteases such as thermolysin in addition to these serine proteases.

Autoinduction of activin genes in early Xenopus embryos

Activin exhibits a potent mesoderm inducing activity towards the ectodermal tissue (animal cap) of Xenopus laevis blastulae. Thus in order to investigate the role of activin in morphogenesis of early Xenopus embryos, activation of genes for activin beta A and beta B was examined by the reverse transcription polymerase chain reaction. In vivo, activin beta B mRNA appears to be present in embryonic stage 1 whereas beta A mRNA is undetectable prior to gastrulation. beta B and beta A mRNAs were noted to accumulate after stages 9 and 15 respectively. Activin gene expression in Xenopus animal caps was examined after treatment with various concentrations of activin A. Under these treatment conditions, both activin beta A and beta B mRNAs accumulated in a dose-dependent fashion after 24 h. The same effect was noted for treatment with similar concentrations of activin B. Accumulation of mRNAs was inhibited by the addition of cycloheximide to the culture medium, consistent with the proposition that activin gene expression requires certain protein factors. In total, therefore, these data suggest that an autoinduction mechanism is involved in the regulation of activin mRNA levels in normal Xenopus embryos and that this mechanism may play a pivotal role during early embryonic development.

Identification of an imprinted U2af binding protein related sequence on mouse chromosome 11 using the RLGS method

A new imprinted gene has been discovered in mice using the technique of restriction landmark genomic scanning (RLGS) with methylation sensitive enzymes. Eight out of 3,100 strain-specific NotI and BssHII spots were identified as imprinted in reciprocal F1 hybrids. Subsequently, we isolated a genomic clone for one locus on proximal chromosome 11 near the Glns locus, an imprinted region in uniparental disomic mice, and its corresponding cDNA clone. Expression of this transcript from the paternal allele was established using RT-PCR of reciprocal F1-hybrid mice. The amino-acid sequence deduced from the cDNA showed significant homology to the U2 small nuclear ribonucleoprotein auxiliary factor 35 kDa subunit.

moaR, a gene that encodes a positive regulator of the monoamine regulon in Klebsiella aerogenes

We cloned and sequenced a Klebsiella aerogenes gene (moaR) for activation of arylsulfatase synthesis by tyramine. This gene was cloned by complementation of a K. aerogenes mutant in which tyramine fails to relieve the arylsulfatase repression caused by sulfur compounds. The moaR gene also activated induction of the synthesis of both tyramine oxidase and the 30-kDa protein that is specifically induced by high concentrations of tyramine or catecholamines. The moaR gene on the chromosome of the wild-type strain of K. aerogenes was disrupted by homologous recombination with a plasmid containing the inactivated moaR. The resultant mutant showed the same phenotype as previously isolated atsT mutant strains that are negative for the derepressed synthesis of arylsulfatase. In this mutant strain, tyramine also failed to induce the synthesis of tyramine oxidase or the production of a 30-kDa protein. The moaR gene is capable of encoding a protein of 26,238 Da. The putative MoaR protein has a helix-turn-helix motif in its C terminus. Thus, it seems likely that the MoaR protein regulates the operons by binding to the regulatory region of the monoamine regulon. The MoaR protein is subject to autogenous control, which was shown by use of a moaR'-lacZ transcriptional fusion.

Isolation and characterization of Xenopus follistatin and activins

Xenopus follistatin and activins were purified from a Xenopus laevis cell line (XTC-F1) by four purification steps consisting of consecutive affinity chromatography on dextran sulfate-Sepharose and Sulfate Cellulofine, fast protein liquid chromatography gel permeation, and reverse-phase high performance liquid chromatography (HPLC). Our results thus obtained indicated that almost equimolar amounts of activins A, AB, and B were found to be present as a complex with follistatin (activin-binding protein) in the conditioned medium of XTC-F1 cells. Reverse-phase HPLC of the complex gave Xenopus follistatin and activins A, AB, and B. The purified Xenopus follistatin showed four major bands in a molecular mass range from 34 to 39 kDa by SDS-polyacrylamide gel electrophoresis under nonreducing conditions. The ability of each form of the protein to specifically bind activin was determined by activin-binding assay and ligand blotting analysis. Each protein was found to have the same NH2-terminus and its sequence was very homologous to that of mammalian follistatin. Several criteria including immunoblotting analysis and various functional assays revealed the existence of three isoforms of activins A, AB, and B in Xenopus, as in mammals. Xenopus activins significantly induced both ventral and dorsal mesoderm in explants of Xenopus blastula cells that would otherwise form epidermis. In a dose-dependent manner of each isoform of activin, the induced explants were able to differentiate into blood-like cells, coelomic epithelium, mesenchyme, muscle, and notochord. The induction patterns of three Xenopus activins were essentially the same. The mesoderm induction by the purified Xenopus activins was shown to be inhibited stoichiometrically by the purified Xenopus follistatin. These results indicate that Xenopus XTC-F1 cells secrete several molecular forms of follistatin/activin-binding protein and three isoforms of activins AB and B in addition to activin A.

Molecular cloning of polymorphic markers on RLGS gel using the spot target cloning method

A new method for target cloning of DNA fragments corresponding to spots on the two-dimensional restriction landmark genomic scanning (RLGS) profile has been developed (targeted spot cloning). We used a Not I restriction trapper to select target DNA fragments from Not I, Eco RV double digests of genomic DNA. The use of the restriction trapper substantially reduces the background clones that are established from the direct recovery of RLGS spot DNA from the two-dimensional gels. Genomic DNA clones were isolated in this study as mouse genome markers for 58 spot loci that were previously characterized using RLGS spot mapping. This method provides a powerful tool for isolating DNA clones after their identification by RLGS system.

Molecular heterogeneity of follistatin, an activin-binding protein. Higher affinity of the carboxyl-terminal truncated forms for heparan sulfate proteoglycans on the ovarian granulosa cell

Follistatin (FS), an activin-binding protein, is a monomer derived from two polypeptide core sequences of 315 (FS-315) and 288 (FS-288) amino acids originated from alternatively spliced mRNA. To define the structural heterogeneity of native FS, we purified six molecular forms of FS from porcine ovaries. Protein chemical analysis revealed that the structural differences among the six isoforms were caused by truncation of the carboxyl-terminal region and/or the presence of carbohydrate chains, resulting in the formation of FS-315, FS-288, and FS composed of 303 amino acids (FS-303) in various forms of glycosylation on the two potential Asn-linked glycosylation sites. The majority of FS isolated from porcine ovaries was FS-303, which may have been derived from FS-315 by proteolytic cleavage of the 12 COOH-terminal amino acids. All six molecular species have almost the same activin binding activity (Kd = 540-680 pM). By contrast, the COOH-terminal truncated form, FS-288, showed much higher affinity for the rat granulosa cell surface than FS-303, whereas FS-315 had no affinity. FS-288 bound to heparan sulfate-Sepharose CL-4B, but FS-315 did not, suggesting that the truncated forms of FS bind to heparan sulfate proteoglycans on the cell. COS cells transfected with the FS-288 DNA expressed the FS-288 protein, which adhered to the cell surface, but cells transfected with the FS-315 DNA secreted the expressed protein into the medium, which did not bind to the cell surface. In rat anterior pituitary culture, FS-288 (ED50 = 2 ng/ml) was more potent in suppressing follicle-stimulating hormone release than FS-303 (ED50 = 10 ng/ml) and FS-315 (ED50 = 20 ng/ml). These results suggest that cell-associated FS traps activin more tightly in the matrix, thereby more effectively blocking the activity of activin on heparan sulfate proteoglycans of the cell surface and that cell-associated FS plays an important role in controlling the various actions of activin in a paracrine or autocrine manner.

Expression of alpha, beta A and beta B subunits of inhibin or activin and follistatin in rat pancreatic islets

We first detected the mRNA expression of follistatin and three subunits of inhibin/activin in rat pancreatic islets by reverse transcription-polymerase chain reaction (RT-PCR). Immunohistochemistry using anti-follistatin serum (against residues 123-134) revealed that follistatin was localized only in insulin-producing B cells. Although the beta A subunit was detectable in the islets, the immunostainable cell types were completely different with two beta A antisera, i.e. anti-beta A (1-10)-Tyr stained B cells, while anti-beta A (87-99) stained glucagon-producing A cells. This inconsistent immunoreactivity was probably related to follistatin binding to beta subunits of inhibin/activin. This study indicates that follistatin and inhibin/activin in the islet serve as paracrine or autocrine modulators in the endocrine pancreas.

Isolation and characterization of activin receptor from mouse embryonal carcinoma cells. Identification of its serine/threonine/tyrosine protein kinase activity

The activin receptor protein was isolated from the mouse embryonal carcinoma (EC) cell line P19 by three cycles of affinity chromatography on an activin A-immobilized column. The purified receptor had a specific and high affinity for activins A, AB, and B (Kd = 345 pM), but not for transforming growth factor beta. The purified activin receptor was identified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and ligand blotting analysis as a single protein of 70 kDa. The amino acid sequence of the first 18 NH2-terminal residues revealed that the receptor is a member of the activin receptor family. The purified receptor phosphorylated itself and exogenous substrate proteins on serine, threonine, and tyrosine residues, indicating that the activin receptor is a transmembrane serine/threonine/tyrosine protein kinase. These results suggest that signal transduction of activin employs a novel pathway via a new class of cellular receptor in EC P19 cells.

Competitive protein binding assay for activin A/EDF using follistatin determination of activin levels in human plasma

A sensitive and specific protein binding assay for activin A/EDF (activin) was developed using follistatin as a binding protein and [125I] labelled activin as a tracer. As 50% acetonitrile (CH3CN) separated free and follistatin-bound activin, plasma pretreated with an equal volume of CH3CN was used as the assay sample and B/F separation was also done with 50% CH3CN. The recovery of the assay was 85.0% and its sensitivity was 0.5 ng/ml. Crossreactivity with inhibin A was 1.8%. The mean plasma level of follistatin-free activin in normal subjects was 1.3 +/- 0.7%. (M +/- SD) ng/ml. Plasma free activin levels were generally elevated in patients with chronic renal failure or hematological diseases associated with anemia.

Paracrine effect of folliculo-stellate cells on the growth factor-like action of activin A in anterior pituitary cultures

Several studies have shown that pituitary folliculo-stellate (FS) cells exhibit local functions within the pituitary gland. On the other hand, we have shown previously that activin A increases the number of FSH-producing gonadotropes in cultured rat anterior pituitary cells. In this study, we investigated whether FS cells exert an influence on the action of activin A. FS cells were prepared by culturing the dispersed rat anterior pituitary cells in media containing 15% fetal calf serum and 6 mM glutamine for 15 days. Most cells had the morphological characteristics of FS cells and S-100 protein immunoreactivity, a specific marker of FS cells. The number of FSH cells, which was higher in activin A-treated than in control cultures, was reduced to the control level by incubation with activin A plus conditioned media from FS cell-enriched cultures (FSCM). This inhibitory effect of FSCM was neutralized by a follistatin antibody, but not by anti-S-100 protein or anti-basic fibroblast growth factor. Furthermore, follistatin suppressed activin A stimulated increases in the number of FSH cells in a similar inhibitory pattern to that of FSCM. Meanwhile, the number of FSH cells was not affected by FSCM or follistatin in the absence of activin A. These results suggest that FS cells are involved in the regulation of the function and/or the morphogenesis of the FSH cell-lineage by affecting the action of activin A, and that this paracrine effect of FS cells is mediated by follistatin.

Banding profiles of LTR of human endogenous retrovirus HERV-A in 24 chromosomes in somatic cell hybrids

The human genome carries multiple copies of sequences related to endogenous retroviral genomes. We investigated the distribution of one of these sequences, HERV-A, in 24 human chromosomes by Southern analyses using DNAs from flow-sorted chromosomes or rodent cells carrying a single human chromosome. The results showed that HERV-A is distributed among all human chromosomes and that each chromosome has a specific Southern blot profile. The chromosome-specific pattern did not show significant polymorphism, except in a few cases, when the same chromosome obtained from different individuals was compared. These chromosome-specific Southern hybridization profiles may be useful for chromosome karyotyping. This would allow the integrity of human chromosomes in human-rodent somatic cell hybrids to be monitored without using conventional cytogenetic methods.

Follistatin inhibits activin-induced differentiation of rat follicular granulosa cells in vitro

The effect of follistatin on activin-induced granulosa cell differentiation was investigated in freshly harvested granulosa cells from diethylstilbestrol (DES)-treated rats. Activin induced a remarkable change in granulosa cellular morphology from elongated fibroblast-like to round cells, which follistatin prevented. Follistatin itself had no influence on the cellular morphology. We studied the action of follistatin on activin-induced differentiation of granulosa cells cultured in a chemically defined medium. Addition of activin (30 ng/ml) to the culture increased the FSH binding site approximately 2-fold compared with the control (spontaneous expression) level, whereas follistatin reduced the activin-induced expression level to the control level in a concentration-dependent manner. Activin (30 ng/ml) markedly augmented FSH-induced hCG binding and progesterone production by approximately 20-fold, and these effects were suppressed by follistatin in a concentration-dependent manner. Similarly, addition of follistatin to the culture induced a concentration-dependent decrease of activin-enhanced inhibin-producing activity, but had no effect on FSH-induced inhibin production. These results suggest that follistatin/activin-binding protein binds to activin stoichiometrically to suppress the activin-induced differentiation of rat granulosa cell in vitro, but follistatin itself has no direct effect on activin-independent reactions.

Follistatin is a developmentally regulated cytokine in neural differentiation

Activin acts mitogenically on P19 cells as well as being inhibitory of the differentiation of retinoic acid-treated P19 cells and some neuroblastoma cell lines. Here, we show some lines of evidence that follistatin, an activin-binding protein, is also involved in neural differentiation. Counteracting the activity of activin, addition of follistatin suppresses the anchorage-independent growth of P19 cells in soft agar and stimulates neurite outgrowth of a neuroblastoma cell line, IMR-32 cells. While activin does not seem to be expressed significantly, follistatin is demonstrated in the conditioned medium of these cells. Furthermore, the expression of follistatin in P19 cells is subject to dynamic fluctuations in response to retinoic acid treatment. These neural cells may produce follistatin in a cell stage-specific manner in order to interact with exogenously derived activin.

A monoamine-regulated Klebsiella aerogenes operon containing the monoamine oxidase structural gene (maoA) and the maoC gene

The Klebsiella aerogenes gene maoA, which is involved in the synthesis of monoamine oxidase, was induced by tyramine and the related compounds, subjected to catabolite and ammonium ion repression, and cloned. The nucleotide sequence of the region involved in monoamine oxidase synthesis was determined. Two open reading frames, the maoA gene and a hitherto unknown gene (maoC), were found. These are located between a potential promoter sequence and a transcriptional terminator sequence. A region of the Escherichia coli chromosome that was highly homologous to the Klebsiella maoA gene was found. The potential maoA gene is located at 30.9 min on the E. coli chromosome. Analysis of the amino acid sequences of the first 11 amino acids from the N terminus of the purified monoamine oxidase agrees with those deduced from the nucleotide sequence of the maoA gene. The leader peptide extends over 30 amino acids and has the characteristics of a signal sequence. Primer extension and S1 nuclease mapping of transcripts generated in vivo suggests that the tyramine-induced mRNA starts at a site 62 bases upstream from the ATG initiation codon of the maoC gene. In the putative promoter region, a high degree of similarity to the consensus sequence for the binding site of cyclic AMP receptor protein was found. Thus, the mao region is composed of two cistrons, and the mao operon is regulated by monoamine compounds, glucose, and ammonium ions.

Cloning and nucleotide sequence of a negative regulator gene for Klebsiella aerogenes arylsulfatase synthesis and identification of the gene as folA

A negative regulator gene for synthesis of arylsulfatase in Klebsiella aerogenes was cloned. Deletion analysis showed that the regulator gene was located within a 1.6-kb cloned segment. Transfer of the plasmid, which contains the cloned fragment, into constitutive atsR mutant strains of K. aerogenes resulted in complementation of atsR; the synthesis of arylsulfatase was repressed in the presence of inorganic sulfate or cysteine, and this repression was relieved, in each case, by the addition of tyramine. The nucleotide sequence of the 1.6-kb fragment was determined. From the amino acid sequence deduced from the DNA sequence, we found two open reading frames. One of them lacked the N-terminal region but was highly homologous to the gene which codes for diadenosine tetraphosphatase (apaH) in Escherichia coli. The other open reading frame was located counterclockwise to the apaH-like gene. This gene was highly homologous to the gene which codes for dihydrofolate reductase (folA) in E. coli. We detected 30 times more activity of dihydrofolate reductase in the K. aerogenes strains carrying the plasmid, which contains the arylsulfatase regulator gene, than in the strains without plasmid. Further deletion analysis showed that the K. aerogenes folA gene is consistent with the essential region required for the repression of arylsulfatase synthesis. Transfer of a plasmid containing the E. coli folA gene into atsR mutant cells of K. aerogenes resulted in repression of the arylsulfatase synthesis. Thus, we conclude that the folA gene codes a negative regulator for the ats operon.

Evidence for the participation of endogenous activin A/erythroid differentiation factor in the regulation of erythropoiesis

Activin A/erythroid differentiation factor (EDF) is a human protein that induces differentiation of a murine erythroleukemia cell (the Friend cell). In this study, we demonstrate that endogenous activin A/EDF activity is present in murine bone marrow and spleen. In addition, this activity is secreted by bone marrow and spleen cells in primary culture. Administration of follistatin (a specific binding protein for activin A/EDF) to mice results in a decrease of erythroid progenitors in the bone marrow and spleen. These findings support the concept that activin A/EDF and follistatin have opposing actions in the regulation of erythropoiesis.

Purification and characterization of high molecular weight forms of inhibin from bovine follicular fluid

High molecular mass forms [95 kilodaltons (kDa)] of bovine inhibin-A as well as the known forms of intermediate (55 kDa) and low (32 kDa) mass were purified from bovine follicular fluid by ion exchange chromatography on DEAE-Sepharose, immunoaffinity chromatography using a monoclonal antibody directed against bovine 32-kDa inhibin-A, gel permeation HPLC on TSK-gel, and reverse phase HPLC. The 95-kDa inhibin-A had similar suppressive activity on FSH secretion from cultured rat anterior pituitary cells as the 55- and 32-kDa inhibins. There is, however, a possibility that the inhibin activity detected with larger forms may be due to that of the 32-kDa form that results from proteolytic processing during incubation with rat pituitary cells. Both sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting analysis using monoclonal antibodies specific for 32-kDa inhibin alpha- or beta A-subunits revealed that the 95-kDa inhibin preparation contained two forms of inhibin (105 and 95 kDa), which were composed of either a 50- or a 40-kDa alpha-subunit linked by a disulfide bond(s) to a 55-kDa beta A-subunit. Amino-terminal sequence analysis showed that the 50-kDa alpha-subunit and the 55-kDa beta A-subunit were generated by removal of a signal peptide from each corresponding primary translation product [the first NH2-terminal 17 residues of the inhibin alpha-subunit (residues 1-360) and the first 20 residues of the inhibin beta A-subunit (residues 1-425)] and suggested that the 40-kDa alpha-subunit was formed by proteolytic processing of the 50-kDa alpha-subunit. On the basis of our findings, we propose that in bovine follicular fluid, the larger 105-kDa form of inhibin is processed successively to form the lowest molecular mass form, 32 kDa inhibin, through the smaller 95- and 55-kDa forms.

Follistatin, an activin-binding protein, associates with heparan sulfate chains of proteoglycans on follicular granulosa cells

Follistatin, an activin-binding protein secreted by cultured rat granulosa cells, was shown to associate with the cell surface by affinity labeling with 125I-activin. Addition of follistatin to the cultured cells demonstrated a typical ligand-binding saturation curve, suggesting that granulosa cells have a specific binding site for follistatin. This binding was markedly inhibited by heparin and heparan sulfate, but not by chondroitin sulfates A and C, keratan sulfate, and dermatan sulfate. When granulosa cells were treated with glycosaminoglycan-degrading enzymes before or after addition of follistatin to the cultures, heparinase and heparitinase treatments resulted in significant suppression of the binding, whereas treatment with chondroitinase ABC had no effect. A competition study of the binding using heparin derivatives demonstrated that follistatin seemed to recognize O-sulfate groups in the heparin molecule. Heparitinase-treated granulosa cells exhibited almost full responsiveness to activin, indicating that the enzyme treatment had no effect on activin and receptor interaction. These results suggest that follistatin/activin-binding protein binds to heparan sulfate side chains of proteoglycans on the granulosa cell surface to regulate the various actions of activin.

Characterization of antisera directed against follistatin/activin-binding protein peptides

An attempt was made to develop immunologic probes directed against follistatin/activin-binding protein (ABP), for use in investigating the distribution of ABP in various tissues. Five oligopeptides corresponding to different regions of the predicted ABP amino acid sequence (peptides 1-12, 28-43, 123-134, 270-281 and 300-315) were synthesized chemically, and coupled to Limulus polyphemus hemolymph hemocyanin. The peptide-hemocyanin conjugates were then used to immunize rabbits, and the immunoresponses were monitored by enzyme-linked immunosorbent assay. Reactivity of the antipeptide antisera with ABP was determined by SDS-polyacrylamide gel electrophoresis and immunoblotting analysis. All of the peptides produced immune responses. The antiserum to peptide 123-134 recognized all forms of ABP, whereas the antiserum to peptide 300-315 reacted specifically and sensitively with the long forms of ABP. These two antisera exhibited only a limited cross-reaction with other proteins or none at all. Therefore, they will be useful for studying the distribution of ABP in various tissues.

Gene cloning of the maoA gene and overproduction of a soluble monoamine oxidase from Klebsiella aerogenes

We cloned the structural gene for monoamine oxidase (maoA) from Klebsiella aerogenes into a pKI212 vector in an maoA mutant strain of K. aerogenes. Deletion analysis and complementation tests of the recombinant plasmid showed that the maoA gene was located entirely within a 4.1-kb segment. In an maoA mutant strain harbouring the cloned maoA gene, synthesis of monoamine oxidase was induced by addition of tyramine and related compounds. Transfer of a plasmid containing the maoA gene into a monoamine oxidase-producing strain of K. aerogenes W70 resulted in about a 30- to 40-fold increase in total production of the enzyme. When cells of K. aerogenes carrying the plasmid containing the maoA gene were grown with tyramine, more than 85% of the monoamine oxidase was produced in soluble form, whereas the parent strain W70 produced most monoamine oxidase as the membrane-bound form.

Immunohistochemical localization of follistatin in rat tissues

We have used immunohistochemistry to localize follistatin/activin-binding protein in adult male and female rats. A polyclonal antibody directed against a follistatin peptide (residues 123-134) was used as a specific immunologic probe. Intense and specific follistatin immunoreactivity was evident in spermatogenic cells of seminiferous tubules in the testis. The predominant staining was in nuclei of spermatocytes and spermatids, but no immune reaction was observed in spermatogonia or spermatozoa. Moderate immunoreactivity was detected in Leydig cells. Sertoli cells were follistatin-negative. Significant immunoreactivity was evident in ovarian granulosa cells. The intensity of the staining changed with follicle development: no immunoreactivity was observed in granulosa cells of primordial to primary follicles, but the cells of secondary to Graafian follicles displayed moderate to strong staining and finally luteal cells of the corpus luteum became negative. The epithelial lining of the oviduct and the smooth muscle of the myometrium of the uterus were intensely immunoreactive. Immunoreactive follistatin staining was present in the pituitary: a group of round-shaped cells were specifically stained. Immunostainable follistatin was visible in the epithelial layers of renal tubules with moderate to strong staining reactivity. Hepatic cells in the liver demonstrated homogeneous immunoreactivity from moderate to strong. The cortex of the adrenal gland, white pulp of the spleen and the brain cortex were also stained weakly but distinctly with the antiserum. In conclusion, immunoreactive follistatin is widespread in rat tissues, suggesting that follistatin/activin-binding protein is a ubiquitous protein, regulating a wide variety of activin actions.