| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Reproductive Endocrine Unit, Massachusetts General Hospital, Boston, Massachusetts 02114
Address all correspondence and requests for reprints to: Patrick M. Sluss, Ph.D., Reproductive Endocrine Unit, Director, Immunodiagnostics Laboratory, Bulfinch Building, Suite 051, Massachusetts General Hospital, Boston, Massachusetts 02114. E-mail: sluss.patrick{at}mgh.harvard.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To test this hypothesis, we developed an immunoassay that is specific for full-length FS315. This assay was validated for use with human serum and follicular fluid samples and then used to examine FST in these fluid compartments. Our results indicate that FS315 is indeed the major circulating FST isoform but is undetectable in follicular fluid samples aspirated from normal women or women with polycystic ovary syndrome. These observations confirm the compartmentalization of FST isoforms according to their biochemical properties and biological actions so that the most soluble form is found in the circulation, whereas the forms that bind to cell surface proteoglycans are found in tissue compartments such as the ovarian follicle. They also confirm that the source of FST in human serum is not the ovarian follicle.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although originally purified from gonadal extracts, it is now clear that FST is produced in numerous adult and embryonic tissues (6, 10). FST is clearly essential for normal development, because disruption of the FST gene in mice results in death shortly after birth due to multiple musculoskeletal and skin defects (11, 12). In addition, the FST sequence is extraordinarily conserved across vertebrates, and the Drosophila homolog is 31% identical, contains a domain structure similar to that of human FST, and binds activin (13). Thus, functions performed by FST seem to be critical for normal physiological processes in many organs and organisms throughout a substantial portion of biological evolution.
The vertebrate FST gene contains six exons, which through alternate splicing, gives rise to two mRNA species termed FS288 and FS315 (14). The longer FS315 form includes an acidic C-terminal tail domain encoded by exon 6, whereas the FS288 protein ends after exon 5 due to a stop codon inserted as a result of alternative splicing. A third FST form, FS303, produced by proteolytic processing of the C-terminal tail domain, was discovered when FST isoforms were purified from porcine follicular fluid and sequenced (4). Thus, at least three different forms of FST exist in cells or physiological fluids. Importantly, FST contains a heparin binding sequence (15), which affords an ability to bind to cell surface proteoglycans on many cells (16). Moreover, the C-terminal domain in FS315 seems to neutralize the basic residues of the heparin binding sequence, so that this form exhibits weak cell-surface binding capability, whereas the FS303 form has an intermediate level of cell surface binding (4). Thus, by cleaving or eliminating the C-terminal domain, cells producing FST can secrete alternate forms of FST with different cell associating activity.
Although the physiological role(s) of FST in vivo remains to be established, the compartmentalization of different FST isoforms within the body has been previously proposed. The first evidence for differential distribution came from the observation that ovarian FST did not contribute to circulating FST, as measured by several newly developed RIAs (reviewed in Ref. 17). Serum was later demonstrated to contain several activin-binding proteins, including FST and 
-2 macroglobulin (18, 19, 20). Upon further analysis, this serum FST was found to differ biochemically from ovarian FST in terms of sensitivity to precipitation by heparin or antibodies (21), suggesting that local or tissue FST was biochemically and functionally distinct from circulating FST. Taken together, these observations suggest the hypothesis that local or tissue FST is composed of the FS288 or FS303 isoforms that can bind cell surface proteoglycans, whereas circulating FST is limited to the FS315 isoform, which will not bind substantially to cell surfaces and thus remain soluble.
To test this hypothesis, we have developed and validated a new FS315-specific two-site immunoassay. Our results indicate that FS315 is the major FST isoform in the peripheral circulation, whereas no FS315 could be detected in follicular fluid aspirated from follicles from normal women and women with polycystic ovary syndrome (PCOS), indicating that FST isoforms in blood and follicular fluid are indeed biochemically distinct and remain compartmentalized within the body.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The cDNA for human FST recombinant human (rh)FS-315 was generously provided by Dr. Sunichi Shimasaki (University of California, San Diego), cloned into PCDNA3, and used to generate a 293-cell line secreting full-length protein. The rhFS315 was purified from cell-conditioned culture medium by affinity chromatography using a mouse monoclonal antibody (22) conjugated to Affigel-10 (Bio-Rad, South San Francisco, CA). Based upon Coomassie Blue-stained SDS-PAGE, N-terminal and C-terminal sequencing, this FS315 was determined to be full length and more than 95% pure (data not shown). Mass was determined by amino acid analysis. Recombinant human activin A was produced by transfecting the human ßA cDNA (generously provided by Genentech, South San Francisco, CA) into 293 cells. Cell-conditioned medium from stable colonies was concentrated 30-fold and dimeric activin concentration verified by immunoassay (Serotec, Raleigh, NC). This activin was added along with heparan sulfate (Sigma, St. Louis, MO) to the assay buffer as detailed below. The mass of purified FS315 used as assay calibrators was based on total amino acid content after hydrolysis with 6 N HCl. Paramagnetic particles (PMP) and dimethylacridinium ester, used to prepare capture and detection antibody reagents, respectively, were generously provided by Chiron Corp. (Bayer Diagnostics, Inc, Walpole, MA). Protein conjugation reagents were purchased from Pierce Chemical Corp. (St. Louis, MO), and general chemistry reagents were purchased from Sigma Chemical Corp.
Generation of antibodies
Hybridoma cell lines secreting monoclonal antibodies against human FST were generated by immunization of BALB-C mice with synthetic peptides matching the 292–315 sequence of the FS315 C-terminal tail domain, which were synthesized by the Protein Core Facility at the MGH. During peptide synthesis, a series of truncations were also created that began at residue 292 and ended at 300, 303, 306, 309, and 312 that were used to examine the specificity of each hybridoma clone secreting monoclonal antibodies to the peptide immunogen (see below).
Spleen cells from immunized mice were fused with syngenic SP/O2 mouse myeloma cells (ATCC, Bethesda, MD). Hybridomas were screened against recombinant FS315, and the clone with the highest affinity (FSA-103) was selected for assay development.
Free FST immunoassay
Free FST was measured using a two-site solid-phase immunochemiluminescence assay (SPICA) previously reported (23). This assay uses monoclonal antibody 6FS7 conjugated to PMP to capture FST and monoclonal antibody 7FS30 conjugated to dimethyl acridinium ester to detect all isoforms of FST. The assay does not measure FST bound to activin A or activin-like growth factors, nor does it detect any FST in serum samples from normal men and women (23).
FST 315 immunoassay
The FS315 assay format follows the same SPICA design as that used for the free FST immunometric assay described above. Thus, the 7FS30 antibody that recognizes all isoforms of FST was paired with the newly generated monoclonal antibody to the FS315 tail (FSA-103). In this assay, optimal performance was achieved using monoclonal antibody FSA-103 as the acridinium-conjugated detection antibody and monoclonal antibody 7FS30 as the PMP-conjugated capture antibody.
FS315 mutations
To test the specificity of the FS315 antibody and assay, a series of FS315 mutants were created within the pCDNA3 context and expressed in 293 cells. Stop codons were introduced after residues 300, 303, 308, 312, 313, and 314 using a site-directed mutagenesis kit (Stratagene, La Jolla, CA). Mutant proteins were then produced by transfecting mutant expression vectors into 293 cells and collecting conditioned medium. The concentration of FST in the conditioned medium was determined in the free FST SPICA.
Human specimens
Blood was obtained from human volunteers participating in a number of different protocols within the National Cooperative Center for Infertility Research, in which FST was to be evaluated as a potential marker of reproductive function. All volunteers gave their informed consent before participating in these studies. Relevant medical history and clinical evaluations were obtained to verify the classification of study subjects (e.g. reproductively healthy, cycling women; postmenopausal women; and others). All study protocols were approved and monitored by the Institutional Review Board for human studies at the Massachusetts General Hospital and Partners Healthcare Organizations. Serum was harvested after centrifugation of clotted blood and stored at –15 to –30 C until assayed for FST, reproductive hormones, or clinical endpoints.
For studies of the distribution of FST levels in a large series of men and women, unselected clinical discard specimens were used. These specimens were deidentified and unlinked to medical records before testing in the FST assay. The majority of specimens were submitted to the clinical laboratory for anemia panel testing (serum folate, vitamin B12, or ferritin), whereas the remainder had been submitted for a variety of immunodiagnostic testing.
From a series of sera assayed for FS315, we selected and pooled those that were undetectable and did the same for those that were extremely elevated. The elevated FS315 pool was used as a serum calibrator and diluted with the serum pool that had no detectable FS315, as indicated in the figures. Regression analysis of this high-FS315 serum pool, compared with recombinant FS315, indicated that it contained approximately 1.07 ng FS315/µl.
Human follicular fluid was obtained by aspiration of follicles from well-characterized normal volunteers and women with PCOS, as previously described (24, 25). These patients also gave informed consent, and the study protocol was approved by the Institutional Review Board for human studies at the Massachusetts General Hospital and Partners Healthcare Organizations. Assay results for the free FST assay were previously reported in aggregate (24) and presented here for comparison to the new FS315 assay results.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The capture monoclonal antibody (7FS30) binds all FST isoforms because the epitope recognized is in the second FST domain (26). Thus, the specificity of this FS315 immunoassay rests upon epitope recognition of the FSA-103 detection antibody. FSA-103 was first evaluated for binding to peptides in solid phase that represent C-terminal truncations of the FS315 tail sequence. FSA-103 bound to the 292–315 peptide immunogen, as well as N-terminal truncations, but did not recognize any of the C-terminal truncated peptides (Fig. 1
). These results indicate that deletion of as few as three C-terminal amino acids (313–315) is sufficient for complete loss of detection antibody binding and could thus be useful for construction of an FS315-specific immunoassay.
|
). Moreover, as little as 300 pg/ml FS315 was easily distinguished from nonspecific binding in the immunoassay. In contrast, 300–500 ng FS288, FS300, or FS303 was undetectable in this assay (Fig. 2A), indicating a cross-reactivity of less than 0.1% for these shortened FST isoforms. To more completely define assay specificity, additional truncations were created at residues FS308, 312, 313, and 314. Although these truncated FST mutants were secreted at 50–1000 ng/ml into the medium as determined with the free-FST immunoassay, they were all undetectable in the FS315 immunoassay (Fig. 2B). Thus, the FSA-103 antibody requires the terminal tryptophan residue and can therefore distinguish FST isoforms shortened by as little as one amino acid.
|
Because some earlier FST immunoassays were altered in the presence of activin or heparin, we investigated the effects of these potential interferences on assay performance. Both activin A and heparin sulfate influenced the quantification of FS315 in the immunoassay. In the presence of a 125-ng/ml FS315 dose, the detection antibody signal increased with added activin up to 200 ng/ml (Fig. 3A). Similarly, signal detection was also increased for the 125-ng/ml FS315 dose when heparin sulfate was added up to 100 ng/ml (Fig. 3B). In fact, the signal enhancement was found to be additive over the entire range of the assay, giving a total increase in signal detection of over 10-fold when both activin and heparin were present (Fig. 3C). This enhancing effect was also observed when human serum samples are analyzed. As shown in Table 1, recovery of FS315 was significantly less than 100% until 300 ng/ml activin and heparin were added, whereas no effect of higher doses was observed. Thus, both activin A and heparan sulfate enhance the signal generated from a single dose of FS315 in this assay and therefore increase the sensitivity of the assay over its effective range. Based on these observations, the assay buffer routinely contains 300 ng/ml rh activin A and 300 ng/ml heparan sulfate.
|
|
Although recombinant FS315 added to assay buffer produced a robust standard curve, the same FS315 added to human serum resulted in a suppressed and nonlinear response curve (Fig. 4). Thus, the immunoassay was calibrated with endogenous FS315 in human serum. To generate assay standards, the human serum calibrator, a pool of human serum with high FS315 concentrations, was diluted with a second pool of human serum containing an undetectable level of endogenous FS315. As shown in Fig. 4 (inset), the measurement of FS315 in serum of individual patients was linear and parallel to the serum standard. To calibrate the approximate mass of FS in the serum standard, the FS activity in 1 µl of this standard was assigned an arbitrary value of 1 U. Dilutions of this standard were then tested in the immunoassay against a calibrator of recombinant FS315 for which the mass was determined by amino acid analysis. The resulting data fit an exponential regression with a significance of P < 0.001 (Fig. 5A); and after log transformation, a linear regression established an approximate value of 1.07 U/ng, or 935 ng FS/ml, of the serum pool calibrator (Fig. 5B).
|
|
Analytical precision was determined using three pools of human serum containing different concentrations of FS315. Analysis of these quality control pools in five separate assays demonstrated that the interassay coefficient of variation (CV) ranged from 7.9–8.5% for FS315 concentrations spanning the reportable range of the assay (Table 2). The intraassay CV was less than 5% for both recombinant FS315 and a serum pool.
|
). These results indicate that the variability in FS315 concentrations detected over the sampling interval was less than or equal to the intraassay CV and thus showed no obvious physiological pattern or pulsatility. For females, blood was collected daily over the course of one menstrual cycle. Results for two reproductively normal women are shown in Fig. 6B. The day-to-day variation was greater than the assay CV, but there was no discernable pattern throughout the menstrual cycle, indicating that variation in circulating FS315 is not related to follicular changes during a menstrual cycle.
|
Serum concentrations of FS315 in 250 individual men and women ranged from undetectable (<2 ng/ml) in 12.8% of these specimens to greater than 30 ng/ml in 8% of the samples (Fig. 7). Results for the whole sample set were distributed in a non-Gaussian fashion, with a median between 5 and 10 ng/ml.
|
); but in all cases, FS315 was undetectable. However, FS315 added to follicular fluid was detected with a recovery of greater than 90%, indicating that, if FS315 were present in significant concentrations, it would be detected by this immunoassay (Fig. 8). These observations strongly suggest that the majority of FST in follicular fluid is processed to isoforms smaller than 315, and that serum FS, which is primarily FS315, is not derived from ovarian sources. This, in turn, indicates that the different FST isoforms are compartmentalized within the body according to their biochemical properties.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Measurements of FS315 in serum from normal men and women revealed no evidence for variation as a function of gonadal status, confirming previous observations that serum FST does not reflect changes in gonadal function (reviewed in Ref. 17). Thus, despite the fact that FST was originally isolated from gonadal extracts (1), it seems that serum FST is derived primarily, if not exclusively, from nonovarian sources that might include endothelial cells in blood vessels (29) or from cells involved with immune defense (30). This compartmentalization of FST isoforms is supported by the observation that the FST proteins found naturally in porcine follicular fluid were mainly variably glycosylated forms of FS303 (4), which must be derived proteolytically from the FS315 mRNA rather than from the alternatively spliced FS288 mRNA form. Moreover, our results demonstrate that the FS315 isoform is undetectable in human follicular fluid samples derived from small and large normal follicles as well as those from women with PCOS, all of which have substantial levels of free FST as determined by the free FST SPICA assay that does not discriminate between FST isoforms. Thus, the FST in follicular fluid is biochemically distinct from serum FS, remains compartmentalized within the follicle, and does not contribute appreciably to circulating FST concentrations. In addition, alteration of FST isoform production does not seem to occur in PCOS follicles relative to normal follicles.
However, the FST isoform distribution is consistent with FST activity in vitro. FS315 binds poorly to cell surface proteoglycans (4) and thus would be expected to remain soluble in the circulation. FS288 and 303, on the other hand, bind well to cell surfaces and thus remain with the tissue(s) that produces them. We have previously shown that the cell surface binding of FS288 gives it the ability to neutralize endogenous (autocrine) activin as well as exogenous (paracrine/endocrine) activin, whereas molecules that do not bind to cell surface proteoglycans, like follistatin-like 3 (follistatinrelated protein, follistatin-related gene), only inhibit exogenous activin (8, 31). This would allow for shorter forms of FST to remain on or near cells to regulate activin action locally, whereas FS315 is able to enter the circulation, perhaps to restrict the activity of circulating activin or other TGFß superfamily members and/or aid in their disposal. Further experimentation in mice will be necessary to more fully elucidate the physiological significance of each FST isoform.
Measurements of FS315 in samples from 10-min sampling protocols in normal men or in daily blood samples from normally cycling women indicate that FST levels vary substantially between subjects, as well as within subjects, although this within-subject variation was not pulsatile, nor did it vary with follicular development. Moreover, whereas the majority of FS315 levels in a large number of individual samples from normal men and women was between 3 and 20 ng/ml, a number of samples had substantially higher FS315 levels, ranging up to 60 ng/ml in this series. Indeed, the FST level can reach as high as approximately 1,000 ng/ml, based on specimens (about 10 specimens of at least 600 screened before the assay had been validated) selected for pooling and used as an assay calibrator. It is presently unclear whether these higher FST concentrations reflect normal human variation or some unknown pathophysiology resulting in elevation of FS315. For example, both FST and activin are elevated in response to acute infection (30). Thus, additional experiments will be required to determine whether FS315 levels are altered by disease states or treatments. Such studies might be invaluable in determining precisely which organ(s) is responsible for secreting FS315 into the circulation and is presently underway.
Our results with the new FS315-specific immunoassay demonstrate that the FST in the circulation is primarily full-length FS315, whereas that in human follicular fluid is not, indicating that circulating FST is not derived from gonadal sources. Differences between individuals in FST concentrations further indicate that circulating FS315 is derived from unknown sources that may be differentially affected by alterations in physiology. Identification of tissue sources of circulating FST, as well as the conditions resulting in elevated levels, will advance our understanding of the physiology and perhaps pathophysiology of FST in the human and could lead to useful markers of disease. The existence of this new FS315 assay now allows such investigations to proceed.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Current address for Q.W.: R & D Systems, Minneapolis 55413, Minnesota.
A preliminary report of this immunoassay was presented at the 35th Annual Meeting of The Society for the Study of Reproduction, Baltimore, MD, July 28–31, 2002.
Abbreviations: CV, Coefficient of variation; FST, follistatin; PCOS, polycystic ovary syndrome; PMP, paramagnetic particles; rh, recombinant human; SPICA, solid-phase immunochemiluminescence assay.
Received January 30, 2004.
Accepted June 30, 2004.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2-Macroglobulin is a binding protein of inhibin and activin. Endocrinology 132:2038–2050This article has been cited by other articles:
 |
|
 |
|
  P. Florio, F.M. Reis, P.B. Torres, F. Calonaci, M.S. Abrao, L.L. Nascimento, M. Franchini, L. Cianferoni, and F. Petraglia High serum follistatin levels in women with ovarian endometriosis Hum. Reprod., October 1, 2009; 24(10): 2600 - 2606. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Xia and A. L Schneyer The biology of activin: recent advances in structure, regulation and function J. Endocrinol., July 1, 2009; 202(1): 1 - 12. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Stamler, H. T. Keutmann, Y. Sidis, C. Kattamuri, A. Schneyer, and T. B. Thompson The Structure of FSTL3{middle dot}Activin A Complex: DIFFERENTIAL BINDING OF N-TERMINAL DOMAINS INFLUENCES FOLLISTATIN-TYPE ANTAGONIST SPECIFICITY J. Biol. Chem., November 21, 2008; 283(47): 32831 - 32838. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S.-Y. Lin, R. G. Craythorn, A. E. O'Connor, M. M. Matzuk, J. E. Girling, J. R. Morrison, and D. M. de Kretser Female Infertility and Disrupted Angiogenesis Are Actions of Specific Follistatin Isoforms Mol. Endocrinol., February 1, 2008; 22(2): 415 - 429. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. F. Lerch, S. Shimasaki, T. K. Woodruff, and T. S. Jardetzky Structural and Biophysical Coupling of Heparin and Activin Binding to Follistatin Isoform Functions J. Biol. Chem., May 25, 2007; 282(21): 15930 - 15939. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Mukherjee, Y. Sidis, A. Mahan, M. J. Raher, Y. Xia, E. D. Rosen, K. D. Bloch, M. K. Thomas, and A. L. Schneyer FSTL3 deletion reveals roles for TGF-beta family ligands in glucose and fat homeostasis in adults PNAS, January 23, 2007; 104(4): 1348 - 1353. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S J. Lin, T. F Lerch, R. W Cook, T. S Jardetzky, and T. K Woodruff The structural basis of TGF-{beta}, bone morphogenetic protein, and activin ligand binding. Reproduction, August 1, 2006; 132(2): 179 - 190. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. M Bilezikjian, A. L Blount, C. J Donaldson, and W. W Vale Pituitary actions of ligands of the TGF-{beta} family: activins and inhibins. Reproduction, August 1, 2006; 132(2): 207 - 215. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Sidis, A. Mukherjee, H. Keutmann, A. Delbaere, M. Sadatsuki, and A. Schneyer Biological Activity of Follistatin Isoforms and Follistatin-Like-3 Is Dependent on Differential Cell Surface Binding and Specificity for Activin, Myostatin, and Bone Morphogenetic Proteins Endocrinology, July 1, 2006; 147(7): 3586 - 3597. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Utriainen, J. Liu, T. Kuulasmaa, and R. Voutilainen Inhibition of DNA methylation increases follistatin expression and secretion in the human adrenocortical cell line NCI-H295R J. Endocrinol., February 1, 2006; 188(2): 305 - 310. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. R. Kumar Too Many Follistatins: Racing Inside and Getting Out of the Cell Endocrinology, December 1, 2005; 146(12): 5048 - 5051. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
