| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrine Unit (H.T.K.) and Reproductive Endocrine Unit (A.L.S., Y.S.), Massachusetts General Hospital, Boston, Massachusetts 02114
Address all correspondence and requests for reprints to: Henry T. Keutmann, Endocrine Unit, Wellman 501, Massachusetts General Hospital, Boston, Massachusetts 02114. E-mail: Keutmann{at}helix.mgh.harvard.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
The domain structure of FS (2) is characteristic of a large number of mosaic proteins derived originally through a process of exon shuffling. Following a signal peptide and a 63-residue N-terminal segment, the majority of the molecule (residues 64–288) consists of three successive 73–75 residue, 10-cysteine FS domains, precisely defined by exon-intron junctions (Fig. 1
). Aside from the conserved cysteine alignments, the three FS domains share about 50% primary sequence homology. At the C terminus, two variants of the molecule can be generated through alternative splicing, the first (FS-288) terminating at residue 288 and a second, longer form (FS-315) extended to 315 residues by an acidic C-terminal extension encoded by an extra exon (2).
|
Ten-cysteine FS domains have been found in a number of extracellular matrix proteins, among which osteonectin [SPARC (secreted protein, acidic and rich in cysteines)/BM40] and agrin have been most studied (18, 19, 20, 21). Each FS domain forms an autonomous folding unit, as confirmed by the crystal structure of the single FS domain from SPARC/BM40 (22) and, recently, of domain 1 from FS itself (23) localizing all disulfide linkages exclusively to intradomain cysteines. Although our mutational analyses show that the three FS domains are inherently unable to bind and bioneutralize activin (14), there is evidence that several of the other extracellular proteins interact with growth factors and other ligands, possibly in part through their FS domains (18, 21, 22, 24, 25). Nonetheless, the contribution of the FS domains to the structure and function of their respective holoproteins—including FS—remains to be defined.
We report here the use of domain-deletion and exchange mutants and selective point mutations to determine the minimum domain structure and organization required for FS activity, work that may offer insight into features important to FS domain structure and activity in other proteins as well.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Analysis of Separate N-Terminal and FS Domain Regions
The inability of the FS domain region (residues 64–288) alone to bind activin has been documented previously (14). Preliminary studies (28) suggested the N-terminal domain alone (residues 1–63), despite harboring essential activin-binding determinants, was likewise insufficient for activity. To establish this and identify any synergism between the respective regions in solution, the two regions were expressed and assayed separately. Because the N-domain sequence alone was poorly expressed, a longer sequence was prepared linking the domain to the full-length, 84-residue PTH molecule through a Factor X cleavage sequence (29). Besides improving expression, the PTH sequence enabled us to monitor and quantitate secretion through a PTH chemiluminescence immunoassay (Advantage; Nichols Institute, San Juan Capistrano, CA). The protein was isolated through the C-terminal poly-His tag as with the full-length FS preparations, and cleaved by Factor X to yield the separate N-domain protein. Cleavage was confirmed by a shift in the mass of myc-tagged peptide products by gel electrophoresis and Western blotting.
When assayed for activin binding and inhibition of activin transcriptional activity (Fig. 2), the N-domain-deleted FS preparation (residues 64–288) was again devoid of activity. The isolated N-domain showed weak binding, less than 2% of FS (1–288) wild-type (Fig. 2A), and no effect on activin-mediated transcription (Fig. 2B). To address whether the separate regions can cooperate to bind activin in solution, the two preparations were coincubated at equimolar concentrations for 1 h before assay. As shown in Fig. 2, no increase in activin binding or transcriptional suppression above that of the N-domain alone was observed. Therefore, the N-terminal domain is necessary but not sufficient for activin binding; the N-domain and FS domains must be integrated within a single molecule for FS activity.
|
). Removal of both FS domains 2 and 3, leaving only the N-domain and FS domain 1, had a comparable effect (not shown), emphasizing the need for more than one FS domain for activity. After deletion of domain 3, binding was retained with an affinity 60% that of the wild-type preparation. Hence, in addition to the N-domain, both FS domains 1 and 2 are required for activin binding, but not domain 3.
|
). Similar effects of domain deletion were observed in the more physiological setting of suppression of pituitary-cell FSH secretion. No decrease in FSH secretion was observed after domain 1 or 2 deletion at doses 1000-fold higher than those eliciting a significant response from wild type FS-288 (Fig. 3C). The domain 3 deletion mutant suppressed FSH completely with a potency 20% that of the wild type, consistent with the other assays.
Exchange of FS Domains
The effect of transposition of FS domains 1 and 2 within the otherwise intact FS-288 molecule was studied using a series of exchange mutations. Reversal of FS domains 1 and 2 (FS 2/1/3), or provision of two copies of FS domain 1 (FS 1/1/3), diminished binding activity to levels comparable to outright domain deletion (Fig. 4A) and abolished the ability to suppress pituitary cell FSH secretion (Fig. 4B). The mutant containing two domain 2 sequences (FS 2/2/3) showed partial activin binding with an inhibition curve qualitatively different from wild-type FS288 (Fig. 4A) and a weak effect on pituitary cell FSH secretion at doses above 8 nM. None of these mutants were active in the activin transcriptional assay at 6.25 nM, a 40-fold excess relative to added activin (data not shown). The mutant FS 3/1/2, in which FS domain 3 separated the N domain from FS domains 1 and 2, was likewise markedly impaired in activin binding (Fig. 4A) and transcriptional activity. These results show that, despite their common cysteine alignment and conservation of certain key residues (see below), the sequential order of the FS domains and their orientation relative to the N domain is essential to activin binding and FS bioactivity.
|
). This suggests that one or more of these hydrophobic residues might fulfill a structural role in maintaining domain conformation or stability. In FS, they include Tyr-110, Leu-116, Leu-127, and Val-129 in the C-terminal subdomain of FS domain 1.
Mutation of either Tyr-110 or Leu-116 to Asp markedly diminished activin binding, as summarized in Table 1. A partial decrease was observed after mutation of L-127 and V-129. Effects on activin transcriptional activity (not shown) and on pituitary cell FSH suppression (Fig. 5A) were comparable to those on activin binding. Mutation of Tyr-185 and Leu-191 in FS domain 2, corresponding to residues 110 and 116 in FS domain 1, likewise impaired binding (Table 1) and FS bioactivity (Fig. 5B). A less drastic mutation to Ala also reduced binding (Table 1) and transcriptional activity in most of these mutants, although the decreases were generally less marked, consistent with the importance of both side-chain size and hydrophobicity at these positions.
|
|
).
Figure 6 shows the effects of domain 1 mutation on binding to heparin-Sepharose affinity matrix. Compared with the complete retention observed for wild-type FS-288 and control mutants with substitutions in the N-domain, the mutants of FS domain 1 hydrophobic residues bound weakly to the matrix, in some cases approaching the low retentions observed for mutants in which the heparin binding sequence was replaced or deleted altogether. This bespeaks a distant effect of mutation on the ability of the sequence to recognize heparin.
|
), in contrast to wild-type preparations in which concentrations are typically equivalent by the two methods (14). Absence of recognition of the Y185D and L191D mutants by antibody 7FS30 was confirmed by Western blotting (not shown). Thus, disruption of these epitopes after mutation is consistent with a role for the conserved hydrophobic residues in stabilizing the overall shape of the domain.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Binding activity at higher doses of the FS 2/2/3 protein indicates that domain 2 can partially fulfill the requirements for domain 1, but even this interaction appears compromised in its ability to neutralize activin. Hence, each domain clearly contributes distinct determinants despite their conserved disulfides and other regions of sequence homology. FS domain 3, on the other hand, is not required; its deletion had only modest effects on FS activity. It remains possible that FS domain 3 contributes to FS function in a way not discerned by the assay systems used here, such as promoting stability in vivo through, for example, its N-linked carbohydrate side-chain. The FS domain 3 deletion mutant replicates the experiment of nature provided by the recently-described FS homolog, FSTL3, also known as FS-related gene product (FLRG) (35, 36) or FS-related protein (FSRP) (37, 38). FSTL3 has an N-domain and two FS domains (Fig. 1) but lacks a third FS domain, yet binds activin and inhibits its transcriptional activity (36, 38, 39).
Although we have found that the three FS domains alone cannot bind or bioneutralize activin (14), they may contribute to binding either by providing additional direct contact sites or by interacting with the N-terminal domain to facilitate binding through determinants there. Activin binding by domain 2 of FSTL3/FLRG/FSRP has been detected (36), although the affinity of this interaction could not be ascertained by the radioligand-blotting technique used. Our evidence for partial binding by the FS 2/2/3 mutant noted above also suggests a possible direct contribution by domain 2 to activin binding provided the N-terminal domain is also present.
In addition to primary sequence differences, intradomain conformation may be important in determining contact surfaces and interdomain alignment. As recognized for some time, the 10-cysteine FS domain can be separated into two subdomains: an elongated, hairpin-like N-terminal region with cysteine alignments resembling epidermal growth factor, and a globular C-terminal portion with distant similarities to the Kazal family of protease inhibitors (18, 22, 23) (Fig. 1). In comparing their recent crystallographic analysis of FS domain 1 with the structure of the single FS domain from SPARC/BM40, Innis and Hyvonen (23) have found striking differences in the orientation of the subdomains relative to each other, even though the subdomains themselves align reasonably closely. The angle between subdomains appears to be influenced in particular by the sequence of the loop segment leading into the C-terminal subdomain (residues 89–100 in FS domain 1) that differs extensively among FS domains including SPARC/BM40, FS domains 1 and 2 (23). This may render domain shapes incompatible except in their appropriate order within the full-length molecule.
A salient feature of the C-terminal subdomain is the number of FS domain hydrophobic residues that are conserved not only within FS and FSTL3 (Fig. 1), but among the domains from all FS domain containing proteins including agrin and the SPARC/BM-40 family (see Refs. 22 and 30, 31, 32 for comparisons). The importance of these residues is borne out by our mutational analyses showing near-complete loss of activin binding and bioactivity after Asp or Ala substitution for Tyr-110 and Leu-116 in FS domain 1, and their FS domain 2 counterparts at positions 185 and 191. Partial loss of bioactivity was observed after mutation of Leu-126 and Val-129. As shown in the structural model (Fig. 7) from the crystal structure of FS domain 1 (23), these appear to constitute a distinct hydrophobic core in the rather compact C-terminal subdomain. This is also apparent in models based on the earlier SPARC/BM40 structure (22). These residues may thus be important in stabilizing overall domain structure and conformation, in a manner shown for numerous other proteins (for example see Refs. 40 and 41).
|
) between the N- and C-terminal regions that corresponds with the loop segment in FS domain 1 and SPARC/BM40 proposed to influence subdomain orientation, as noted above (23). The heparin-binding site (75–84) in FS domain 1, disturbed as shown in Fig. 6, occupies the hairpin-like N-terminal subdomain. Other than the third and fourth disulfide linkages (89–121 and 93–114 in FS domain 1), there appear to be relatively few contacts between the subdomains (23) so conformational changes may be readily propagated from one to the other. Among the more distantly related extracellular matrix proteins containing FS motifs, SPARC/BM40 most closely resembles FS functionally through its ability to inhibit platelet-derived growth factor and vascular endothelial growth factor (VEGF) binding to their receptors (42, 43). Although sometimes attributed to the FS domain, the platelet-derived growth factor binding site in SPARC has in fact not been localized. Direct functional effects involving FS domains are suggested, however, by evidence that a peptide sequence from the FS domain can replicate the effects of the full-length molecule in disrupting endothelial-cell focal adhesion sites (19, 24) or inhibiting VEGF-stimulated DNA synthesis (43). At least one other sequence, within the adjacent EF-hand calcium-binding domain in SPARC, has similar effects, suggesting a cooperative or multisite interaction with the FS domain. This association appears essential also for sustaining high-affinity calcium binding.
Crystallographic data (44) reveals a close contact between a segment of the FS domain and neighboring EF-hand calcium-binding region of the SPARC/BM-40 molecule, and very likely in a growing number of other extracellular-matrix proteins bearing a distinctive paired FS domain/calcium-binding (EC) domain (22, 24, 45). Notably, two of the conserved hydrophobic residues in FS domain 1 (L116 and V129) correspond to those appearing at the interface between the FS and EC domains in the crystal structure of SPARC/BM-40 (44). Other interface residues correspond to E128 and Q130, nearby in the C-terminal subdomain, and H62 in the N-terminal subdomain (Figs. 1 and 7). FS (and FSTL3) clearly differ from the SPARC/BM40 in lacking any EC domain, but these positions in SPARC/BM40 could provide clues in future study of contact site(s) involving FS domains in ligand binding.
Additional insights into potential interdomain interactions are also available from other proteins. Electron micrographic images of the agrin molecule (46) suggest that the multiple FS domains are oriented end-to-end to enable the molecule to span the intersynaptic space at the neuromuscular junction. A hypothetical model of FS domain organization based on the SPARC/BM40 crystal structure (22) depicts a similar extended configuration of successive FS domains. A more direct prediction may be gained from the structures (47, 48) of a group of cell-surface proteins containing multiple short consensus repeats, four-cysteine 60-residue modules linked by three to four residue segments similar to those separating the domains in FS. In the complement-regulating protein CD46, adjacent modules form a 60-degree angle with limited flexibility and a very small interdomain interface (4% of total molecular surface) involving only a few amino acids (48). It may be predicted, therefore, that the FS domains together form a relatively elongated structure, with a more flexible N-terminal domain possibly forming the majority of interdomain contacts.
An important contribution of the FS domain region to bioactivity is its ability to bind to cell-surface heparan sulfate proteoglycans via the consensus heparin-binding sequence within FS domain 1 (33). This property appears essential to FS’s function as a local cellular regulator; cell surface bound FS has been postulated to form a barrier or canopy limiting approach by exogenous activin while allowing access by endogenous, autocrine-derived activin closer to the cell surface (1, 6, 49, 50). High local FS concentrations may be sufficient to regulate the local actions of other TGFß-related factors such as the BMPs, despite their lower inherent affinity for FS (36, 51, 52). BMPs have been implicated, for example, as alternative mediators of pituitary FSH secretion (53), in cartilage maturation and bone development (54) and in muscle growth (55).
Despite their outwardly diverse actions, most of the FS domain-containing extracellular matrix proteins also exhibit some mechanism for binding to cell surfaces or other extracellular matrix components. Several, including testican and agrin, bear proteoglycan side-chains that recognize binding proteins on cell surfaces (46, 56, 57). SPARC/BM40 has been shown to interact with both collagen and vitronectin (21). Agrin also has recognition sequences for laminin and heparin, and splice variants retaining or lacking these segments appear important to its tissue localization (20, 46). Members of the expanding family of tomoregulin/S7365-related proteins are anchored integrally to cell surfaces through a membrane-spanning sequence toward the C terminus of the molecule (25, 31, 58).
FSTL3 is unusual in its lack of a heparin-binding sequence or other apparent means for cell-surface association. This accounts for its low biological activity in systems employing endogenous activin (e.g. pituitary cell FSH suppression) as opposed to its ability to bind exogenous activin that approaches that of FS itself (37, 38). This may be of considerable importance in future efforts to distinguish the physiological effects of FSTL3 from those of FS. In the current study, the very limited ability of the FS 2/2/3 mutant to suppress pituitary cell FSH secretion even at doses that bind activin (Fig. 4) may be at least in part due to absence of FS domain 1 and its heparin-binding sequence. Beyond a direct or indirect involvement in activin binding, therefore, the FS domains in FS may also fulfill a structural or scaffolding role in orienting the activin binding region appropriately at the cell surface.
To summarize our findings, a continuous sequence comprising the N-terminal domain and first two FS domains fulfill the structural requirements for activin binding and neutralization. The domains cannot be rearranged; each appears to contribute uniquely to overall FS structure and, ultimately, activin binding—either through direct contact or indirectly through interactions with or conformational effects on essential hydrophobic determinants in the N-terminal domain. A structural role in orienting the activin-binding site at the cell surface, abetted by the heparin-binding region on FS domain 1, also appears likely based on the rather limited structural information to date from other FS domain-containing proteins. A definitive picture of the organization of the respective domains and their functional contribution to activin binding and neutralization must, however, await a crystal structure of full-length FS.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Preparation of Mutant FSs
FS-288 coding sequence was removed from pHTF302R (a gift of Dr. S. Shimasaki, Whittier Institute) and subcloned into the mammalian expression vector pcDNA3.1/myc-His (Invitrogen). The resulting construct (pFS288mycHis) was then used as a template for mutagenesis. Domain deletion and domain exchange mutants were generated by PCR amplification of specific domains using extended primers containing sequence homologues to the desired neighbor domain at their 5' end. These overlapping domains were then fused together by a second PCR and cloned back into the expression vector. Site-directed mutagenesis of single residues within FS domains 1 and 2 was carried out as described (14) using the QuikChange kit (Stratagene, La Jolla, CA) after the manufacturer’s recommendations. Mutant sequences were verified by bidirectional sequencing at the DNA sequencing core facility of Massachusetts General Hospital.
Expression of Recombinant FSs
The pFS288mycHis vectors bearing mutant or wild-type FSs were transfected into CHO cells using polybrene (Specialty Media, Phillipsburg, NJ) and stably secreting cells were selected using geneticin. Expression levels were typically 150–300 ng/ml. For production of certain mutants, we used the 293F cell suspension system in serum-free medium ("Free-style"; Invitrogen) that became available during later stages of the work. Expression of wild-type and mutants from these cells was considerably higher (8–12 µg/ml). Neither of the cell types used secreted detectable levels of activin. FS secretion was monitored by SPICA for FS and/or immunoassay for C-terminal myc tag as described below. The recombinant FSs were isolated from medium by binding to nickel-Sepharose affinity columns (QIAGEN, Valencia, CA) via the C- terminal poly-His tag. Following stepwise elution with imidazole, products (typically eluting between 50–150 nM imidazole for CHO medium and 50–300 mM for Freestyle medium at pH 6.8) were concentrated and exchanged into Dulbecco’s PBS by filter centrifugation (Centriprep-10 tubes; Amicon, Bedford, MA). Conditioned medium from nontransfected CHO cells was processed similarly for use as a control preparation in all assays.
Quantitation of Secreted FSs
FS concentrations in media and affinity eluates were established by two independent immunological assays.
The two-site solid-phase immunochemiluminescent assay (SPICA) for free FS is described in detail by Wang et al. (59) and McConnell et al. (27). Monoclonal antibody 6FS7, recognizing determinants largely in FS domain 3, is coupled to paramagnetic beads for use as capture antibody, with chemiluminescent-labeled antibody 7FS30, directed toward the sequence (168–178) in FS domain 2 (34), employed for detection.
A solution-phase assay directed toward the C-terminal myc tag (14) was used for quantitation of preparations lacking FS domain SPICA epitopes (e.g. FS domain 2 and/or FS domain 3 deletions and mutations) and, in conjunction with the SPICA, for probing conformational changes after point mutagenesis. This assay recognizes FS in either free or activin-bound form; estimates of concentration of expressed wild-type FS-288 by SPICA and myc-tag assay are comparable. A synthetic peptide incorporating the myc epitope, linked by a poly-Gly spacer to an N-terminal tyrosine for (125I) labeling, was used as radioligand. A rabbit polyclonal anti-myc tag antibody from Upstate Biotechnology (Lake Placid, NY) was used at a final dilution of 1:3500 in PBS/0.5% BSA (pH 7.20) (20 h, 20 C) in an assay volume of 0.5 ml/tube. For phase separation, 0.2 ml antirabbit IgG (1:16) with 16% polyethylene glycol was added for 2 h at 4 C. Detection limit of the assay is 0.1 nM myc-epitope peptide or C-terminal myc-tagged FS, with a linear inhibition curve between 0.5 and 50 nM. For repeated measurements of a myc-peptide reference standard at successive doses over this working range, the intraassay and interassay coefficients of variation were 3.81–4.52% and 4.76–12.8%, respectively.
Activin Binding Assay
Binding of expressed FSs to labeled activin was determined by competition assay as described (4). Successive doses of mutant or wild-type FSs in duplicate were incubated with (125I)-labeled activin in binding buffer (10 mM PBS, 0.1% gelatin, 0.05% Tween; 200 µl) for 2 h at 20 C, then added to 96-well plates (Immulon-2; Dynatech Laboratories, Chantilly, VA) coated with 25 ng of affinity-purified FS288. After incubation at 20 C for 90 min, wells were washed and counted in a gamma counter. Each mutant preparation was assayed in at least three independent experiments. Relative potencies were calculated by comparison of half-maximal inhibition of labeled activin binding by mutant and wild-type FSs, respectively.
Assay for Transcriptional Response to Activin
Inhibition of activin-mediated transcriptional responses was assayed in a luciferase-coupled reporter system using HEK-293 cells transfected with a pARE-GL3 and pFAST1 reporter system as previously described (38). After a 16-h posttransfection incubation, cells were treated with fresh media containing 5 ng/ml activin, alone or preincubated (60 min) with 10–200 ng/ml of various FS preparations for an additional 24 h in triplicate. Cell extracts were assayed for luciferase activity using the Dual-Luciferase Reporter Assay system from Promega (Madison, WI). All mutants were tested in at least three independent assays.
Bioassay for Pituitary FSH Secretion
Assay for suppression of basal FSH secretion in cultured rat anterior pituitary cells was based on the method of Scott et al. (60). The anterior pituitary glands of adult male Sprague-Dawley rats were mechanically and enzymatically dispersed with 0.4% trypsin and 0.25% deoxyribonuclease and plated at 2.5 x 105 cells/0.5 ml well in 48-well trays in 
MEM containing 21 mM NaHCO3, 10% heat inactivated fetal bovine serum, and 10% penicillin/streptomycin solution (pH 7.4). Cells were incubated at 37 C in 95% air/5% CO2 for 72 h, washed with PBS, and reincubated in 0.5 ml fresh media containing the various FS and control preparations at the specified concentrations. After 72 h, the conditioned media were assayed for rat FSH using reagents and protocols provided by Dr. A. F. Parlow through the National Hormone and Pituitary Program, National Institute of Diabetes and Digestive and Kidney Diseases. The antirat FSH antibody (FSH-RIA-11) was used at a final dilution of 1:125,000. All preparations were tested in three independent assays, from which a representative assay is illustrated.
Heparin Affinity Chromatography
Partially purified FS preparations (1 ml, 100–200 ng) were applied to columns of heparin-agarose gel (Affi-Gel, Bio-Rad, Hercules, CA; 0.3-ml bed volume) in PBS (pH 7.2) in presence of 0.1% BSA. Columns were washed (0.5 ml) then eluted with 1.5 ml PBS/BSA. Aliquots of applied and eluted FSs were measured by SPICA and/or myc-tag assay to determine the proportion retained on the column.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication March 28, 2003. Accepted for publication October 10, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  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]
|
|
|
|
 |
|
 A. L. Schneyer, Y. Sidis, A. Gulati, J. L. Sun, H. Keutmann, and P. A. Krasney Differential Antagonism of Activin, Myostatin and Growth and Differentiation Factor 11 by Wild-Type and Mutant Follistatin Endocrinology, September 1, 2008; 149(9): 4589 - 4595. [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]
|
|
|
|
|
|
C. A. Harrison, K. L. Chan, and D. M. Robertson Activin-A Binds Follistatin and Type II Receptors through Overlapping Binding Sites: Generation of Mutants with Isolated Binding Activities Endocrinology, June 1, 2006; 147(6): 2744 - 2753. [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]
|
|
|
|
|
|
S. Saito, Y. Sidis, A. Mukherjee, Y. Xia, and A. Schneyer Differential Biosynthesis and Intracellular Transport of Follistatin Isoforms and Follistatin-Like-3 Endocrinology, December 1, 2005; 146(12): 5052 - 5062. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Xia, Y. Sidis, A. Mukherjee, T. A. Samad, G. Brenner, C. J. Woolf, H. Y. Lin, and A. Schneyer Localization and Action of Dragon (Repulsive Guidance Molecule b), a Novel Bone Morphogenetic Protein Coreceptor, throughout the Reproductive Axis Endocrinology, August 1, 2005; 146(8): 3614 - 3621. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Sidis, A. L. Schneyer, and H. T. Keutmann Heparin and Activin-Binding Determinants in Follistatin and FSTL3 Endocrinology, January 1, 2005; 146(1): 130 - 136. [Abstract] [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 |
FSD-1, 
