The Platelet Integrin αIIbβ3 Has an Endogenous Thiol Isomerase Activity
2000; Elsevier BV; Volume: 275; Issue: 47 Linguagem: Inglês
10.1074/jbc.m003279200
ISSN1083-351X
AutoresSarah O’Neill, Aisling Robinson, Adele Deering, Michelle Ryan, Desmond J. Fitzgerald, Niamh Moran,
Tópico(s)Cell Adhesion Molecules Research
ResumoIntegrins are cysteine-rich heterodimeric cell-surface adhesion molecules that alter their affinity for ligands in response to cellular activation. The molecular mechanisms involved in this activation of integrins are not understood. Treatment with the thiol-reducing agent, dithiothreitol, can induce an activation-like state in many integrins suggesting that cysteine-cysteine dithiol bonds are important for the receptor's tertiary structure and may be involved in activation-induced conformational changes. Here we demonstrate that the platelet-specific integrin, αIIbβ3, contains an endogenous thiol isomerase activity, predicted from the presence of the tetrapeptide motif, CXXC, in each of the cysteine-rich repeats of the β3 polypeptide. This motif comprises the active site in enzymes involved in disulfide exchange reactions, including protein-disulfide isomerase (EC 5.3.4.1) and thioredoxin. Intrinsic thiol isomerase activity is also observed in the related integrin, αvβ3, which shares a common β-subunit. Thiol isomerase activity within αIIbβ3 is time-dependent and saturable, and is inhibited by the protein-disulfide isomerase inhibitor, bacitracin. Furthermore, this activity is calcium-sensitive and is regulated in the EDTA-stabilized conformation of the integrin. This novel demonstration of an enzymatic activity associated with an integrin subunit suggests that altered thiol bonding within the integrin or its substrates may be locally modified during αIIbβ3 activation. Integrins are cysteine-rich heterodimeric cell-surface adhesion molecules that alter their affinity for ligands in response to cellular activation. The molecular mechanisms involved in this activation of integrins are not understood. Treatment with the thiol-reducing agent, dithiothreitol, can induce an activation-like state in many integrins suggesting that cysteine-cysteine dithiol bonds are important for the receptor's tertiary structure and may be involved in activation-induced conformational changes. Here we demonstrate that the platelet-specific integrin, αIIbβ3, contains an endogenous thiol isomerase activity, predicted from the presence of the tetrapeptide motif, CXXC, in each of the cysteine-rich repeats of the β3 polypeptide. This motif comprises the active site in enzymes involved in disulfide exchange reactions, including protein-disulfide isomerase (EC 5.3.4.1) and thioredoxin. Intrinsic thiol isomerase activity is also observed in the related integrin, αvβ3, which shares a common β-subunit. Thiol isomerase activity within αIIbβ3 is time-dependent and saturable, and is inhibited by the protein-disulfide isomerase inhibitor, bacitracin. Furthermore, this activity is calcium-sensitive and is regulated in the EDTA-stabilized conformation of the integrin. This novel demonstration of an enzymatic activity associated with an integrin subunit suggests that altered thiol bonding within the integrin or its substrates may be locally modified during αIIbβ3 activation. protein-disulfide isomerase cysteine-rich repeats reduced and denatured RNase 3-(N-morpholino)propanesulfonic acid polyacrylamide gel electrophoresis Integrins are cell-surface, calcium-dependent, heterodimeric adhesion molecules that play a critical role in cell-cell and cell-substrate adhesion. In cells at rest, integrins are present in a latent or resting conformation. Following cellular activation, they undergo conformational changes to become high affinity receptors for their specific ligand(s). The "switch" mechanism whereby integrins are converted from their resting conformation is critically important to their cellular function. However, the mechanisms underlying these conformational changes have not yet been deduced. The conformational changes in the platelet-specific integrin, αIIbβ3, are the composite result of at least two processes. First, intracellular signals converge on the cytoplasmic tails of the integrin conveying the intention to activate. Second, the extracellular domains, which constitute >95% of the molecules, respond with an increased affinity for ligand and an altered display of antibody epitopes suggestive of altered protein folding. We have shown that the conserved α-subunit cytoplasmic sequence, KVGFFKR, is critical for the intracellular-mediated activation of the platelet integrin (1Stephens G. O'Luanaigh N. Reilly D. Harriott P. Walker B. Fitzgerald D.J. Moran N. J. Biol. Chem. 1998; 273: 20317-20322Abstract Full Text Full Text PDF PubMed Scopus (72) Google Scholar). The precise role played by this peptide sequence remains uncharacterized. However, Vinogradova et al. (2Vinogradova O. Haas T. Plow E.F. Qin J. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 1450-1455Crossref PubMed Scopus (128) Google Scholar), have recently proposed a structural basis for this effect which proposes a protein-protein interaction with the integrin cytoplasmic tails. Deletion or mutation of this cytoplasmic sequence from the αIIb subunit were found to increase the ligand binding affinity of the mutant receptor expressed in Chinese hamster ovary cells (3O'Toole T.E. Katagiri Y. Faull R.J. Peter K. Tamura R. Quaranta V. Loftus J.C. Shattil S.J. Ginsberg M.H. J. Cell Biol. 1994; 124: 1047-1059Crossref PubMed Scopus (580) Google Scholar, 4Peter K. Bode C. Blood Coagul. & Fibrinolysis. 1996; 7: 233-236Crossref PubMed Scopus (27) Google Scholar). In contrast, ligand binding to the ectodomains of integrins has also been demonstrated to affect intracellular recognition sites on the cytoplasmic tails of integrins in a process known as outside-in signaling (5Leisner T.M. Wencel-Drake J.D. Wang W. Lam S.C. J. Biol. Chem. 1999; 274: 12945-12949Abstract Full Text Full Text PDF PubMed Scopus (65) Google Scholar). These studies provide evidence for transmembrane modulation of integrin conformation although the mechanisms mediating these changes are not understood. The subunits of integrins are rich in cysteine residues. The pairing of these residues in disulfide bonds, particularly in the platelet-specific integrin αIIbβ3, have been shown to be important in the structural integrity of the receptor (6Calvete J.J. Henschen A. Gonzalez-Rodriguez J. Biochem. J. 1991; 274: 63-71Crossref PubMed Scopus (158) Google Scholar, 7Calvete J.J. Proc. Soc. Exp. Biol. Med. 1999; 22: 29-38Google Scholar). Chemical reduction of the thiol bonds by the reducing agent dithiothreitol induces an active conformation in αIIbβ3 (8Zucker M.B. Masiello N.C. Platelet Thromb. Haemostasis. 1984; 51: 119-124Crossref PubMed Scopus (75) Google Scholar, 9Peerschke E.I. Thromb. Haemostasis. 1995; 73: 862-867Crossref PubMed Scopus (51) Google Scholar), although the physiological significance of this observation is not understood. Selective mutation of critical cysteines in the platelet integrin αIIbβ3 results in the induction of a constitutively active form of the receptor lending further support to the proposal that altered dithiol bonding is implicated in integrin activation (10Liu C.-Y. Sun Q.-H. Wang R. Paddock C.M. Newman P.J. Blood. 1998; 92: A1413Crossref Google Scholar). Dithiol reduction also causes activation of other integrins (11Edwards B.S. Curry M.S. Southon E.A. Chong A.S. Graf Jr., L.H. Blood. 1995; 86: 2288-2301Crossref PubMed Google Scholar, 12Davis G.E. Camarillo C.W. J. Immunol. 1993; 151: 7138-7150PubMed Google Scholar). Analysis of the integrin cysteine patterns reveals that there are nine CXXC repeats in each β-integrin subunit. This sequence has previously been identified in the active site motif in the bacterial thiol reducing enzyme, thioredoxin, and in the mammalian thiol isomerase, protein-disulfide isomerase (PDI).1 A glycine residue is most often found as the first amino acid position of the repeat. Other proteins that possess this sequence include the chaperone molecules Erp61 and Erp72 (13Lundstrom-Ljung J. Birnbach U. Rupp K. Soling H.D. Holmgren A. FEBS Lett. 1995; 357: 305-308Crossref PubMed Scopus (74) Google Scholar), the gonadotropins, lutropin and follitropin (14Boniface J.J. Reichert Jr., L.E. Science. 1990; 247: 61-64Crossref PubMed Scopus (86) Google Scholar,15Chew C.C. Magallon T. Martinat N. Lecompte F. Combarnous Y. Gosling J.P. Biochem. Soc. Trans. 1995; 23: 394SCrossref PubMed Google Scholar), and fibronectin (16Langenbach K.J. Sottile J. J. Biol. Chem. 1999; 274: 7032-7038Abstract Full Text Full Text PDF PubMed Scopus (95) Google Scholar). All these proteins have demonstrable endogenous PDI-like activity. von Willebrand factor also has numerous CGXC sequences but no thiol isomerase activity was shown (17Mayadas T.N. Wagner D.D. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 3531-3535Crossref PubMed Scopus (176) Google Scholar). In the integrin β subunit, the specific CGXC sequence is repeated four times; once in each of the regions identified as the cysteine-rich repeat regions (CRRs) (6Calvete J.J. Henschen A. Gonzalez-Rodriguez J. Biochem. J. 1991; 274: 63-71Crossref PubMed Scopus (158) Google Scholar). We, therefore, investigated purified integrin αIIbβ3 for the presence of endogenous thiol isomerase activity. In addition, we evaluated its biochemical and pharmacological profile in in vitro assays and demonstrated that this enzymatic activity is a common feature of integrins. Ribonuclease A (RNase) and PDI were obtained from CalBiochem-Novabiochem (CA). αvβ3 and α5β1were supplied by Chemicon (CA). Sephadex G25 column, bacitracin, guanidine hydrochloride, EDTA, dithiothreitol, cytidine 2′,3′-cyclic monophosphate, MOPS, and Trizma base were all supplied by Sigma. CD41a (an antibody to αIIbβ3) was obtained from Immunotech, Marseilles, France, and anti-PDI from StressgenBiotechnology Corp., Victoria, British Columbia, Canada. Goat anti-mouse Alexa 488-conjugated IgG and goat anti-rabbit Alexa 546-conjuagated IgG were obtained from Molecular Probes, Leiden, Netherlands. This was purified from outdated units of human platelets as described previously (18Phillips D.R. Fitzgerald L. Parise L. Steiner B. Methods Enzymol. 1992; 215: 244-263Crossref PubMed Scopus (23) Google Scholar) and stored at −80 °C in a buffer containing 20 mmTris, 1 mm CaCl2, and 0.1% Triton X-100. Samples were concentrated to approximately 4 mg/ml by filtration through Centrex filters (Schleicher and Schuell, 10-kDa cut-off). 30 mg of purified RNase was reduced and denatured (rdRNase) at room temperature for 18 h in the presence of 0.15 m dithiothreitol and 6 mguanidine HCl in 0.1 m Tris, pH 8.6, before desalting on a Sephadex G25 column equilibrated with 0.01 n HCl. The concentration of the reduced and denatured (rd)RNase fractions were determined from its extinction coefficient 9200m−1 cm−1 at 275 nm. Fractions were stored under argon at −80 °C for up to 14 days before use. The reduced enzyme at a final concentration of 30 μm was diluted into 0.1 m Tris-HCl, pH 7.4, with 1 mm EDTA containing either 5 μm bovine protein-disulfide isomerase, purified GPIIb/IIIa at the concentrations indicated, 5 μm ovalbumin, or no isomerase. The extent of reactivation of rdRNase was monitored by removing aliquots at various time points and measuring recovered RNase activity, as described by Bonifaceet al. (14Boniface J.J. Reichert Jr., L.E. Science. 1990; 247: 61-64Crossref PubMed Scopus (86) Google Scholar). A typical assay consisted of 1.4 μm RNase and 0.44 mm cytidine 2′,3′-cyclic monophosphate in 0.1 m MOPS, pH 7.0. Changes in absorbance at 284 nm were monitored in a Unicam split beam spectrophotometer. Activity was expressed as a percentage of a native RNase control. To determine these parameters, the RNase refolding assay described above was used. Concentrations of reduced denatured RNase ranging from 7.5 to 75 μm was used in the absence of isomerase or in the presence of protein-disulfide isomerase (1.5 μm) or αIIbβ3 (1.5 μm). The amount of active RNase was determined by measuring the initial rate of recovery of activity over 1 min at 284 nm after a 48-h incubation. Plots of substrate concentration versus initial velocity were fit to Michaelis-Menten equation using Deltagraph to determineV max and K m. The uncatalyzed initial rates, determined in the absence of a source of isomerase activity, were subtracted from the catalyzed rates determined in the presence of PDI or integrin. A standard curve was generated from the initial rates of various concentrations (0–4 μm) of native RNase. A value for micromolar of RNase regenerated per minute for V max was then determined from the slope of this curve. 2 μg/ml of either purified αIIbβ3 or protein-disulfide isomerase in 50 mm Tris, pH 7.4, with 150 mm NaCl and 1% Triton X-100 was precipitated with a rabbit polyclonal anti-human PDI antibody (a kind gift of Dr. Eric Quemeneur, Center d'Etudes Saclay, Gif-sur-Yvette, France). Precipitates were harvested with protein G beads. Samples were separated on 12.5% SDS-PAGE and transferred to polyvinylidene difluoride. Membrane was probed with 1:1000 dilution of anti-PDI antibody (rabbit anti-protein-disulfide isomerase polyclonal antibody from Stressgen Biotech Corp.) followed by 1:1500 dilution of Protein A peroxidase. Membrane was developed with ECL. Glass slides pre-coated with poly-l-lysine (Sigma) or slides coated with fibrinogen (20 μg/ml) for 2 h were blocked with 1% bovine serum albumin in Tris-buffered saline, pH 7.4, for 1.5 h in a humidified staining tray. Gel filtered platelets prepared as described previously (19Mendelsohn M.E. Zhu Y. O'Neill S. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 11212-11216Crossref PubMed Scopus (75) Google Scholar) were diluted to 2.0 × 105/ml and allowed to adhere to the poly-l-lysine-coated slides for 15 min and to the fibrinogen-coated slides for 30 min. Slides were then washed with Tris-buffered saline and fixed in ice-cold methanol (7 min). Slides were blocked with 1% bovine serum albumin for 20 min. Primary antibodies (1:200 dilution for monoclonal anti-αIIbβ3 complex specific and 1:50 for polyclonal anti-protein-disulfide isomerase) were incubated for 45 min. 1:400 dilution of goat anti-mouse Alexa 488-conjugated IgG and goat anti-rabbit Alexa 546-conjugated IgG in Tris-buffered saline were incubated for 10 min. Slides were washed and mounted in fluorescent mounting medium (Dako, Carpinteria, CA) before imaging on a Zeiss Axioplan 2 confocal microscope. All experiments were carried out at room temperature. Bacitracin (0–100 μm) was added to each reaction mixture at 0 h. Examination of the reactivation of reduced and denatured RNase in the presence of this inhibitor was carried out as described above. The results are expressed as % of maximal recovery of the refolding activity of rdRNase observed in the absence of bacitracin. Analysis of the integrin cysteine patterns reveals that there are four highly conserved CRRs in each β-integrin subunit (6Calvete J.J. Henschen A. Gonzalez-Rodriguez J. Biochem. J. 1991; 274: 63-71Crossref PubMed Scopus (158) Google Scholar). In Fig.1, we show the alignment of the second cysteine-rich repeat from human β3, with the same region from other β-integrins ranging from sponge and coral to mammalian sequences. There is 44–63% sequence identity within this region increasing to 100% for the β3 subunits. The cross-species homology index is further increased when conservative amino acid substitutions are accounted for, giving a value of 56–83% homology. The cysteine residues are 100% conserved. Similar homology scores are obtained when the first, third, or fourth CRRs are aligned. The four CRRs of β3 subunits from human, mouse, and dog are 90–94% identical. The alignments also reveal the presence of two motifs in each repeat: motif I has the sequence "GXCXCXXCXC" and is separated by 5–12 amino acids from motif II which has the sequence "GXXCXC"(Fig. 1 b). In all integrins, the first CRR is imperfect with its first cysteine replaced by an Phe, Met, Leu, or Tyr. Similarly, the fourth CRR is incomplete in all integrins, lacking the second motif but having a similar GXX CXXCXXC in β1, β3, and β5–8 integrins. β8 integrins are an exception to the integrin family and have very corrupted CRRs, the first CRR containing a CXXC, but not the flanking residues of motif I. It lacks motif II. The second CRR also lacks motif II. The third CRR lacks C2 in motif I but has a normal motif II. The fourth CRR, like all other integrins, lacks a true motif II. Motif I contains the active site sequence (CXXC) of the ubiquitous thiol modifying enzymes, PDI and thioredoxin. Each β-integrin subunit contains nine repeats of this CXXC tetrapeptide sequence; two in the amino-terminal region, one in each of the four CRRs, and two in the modified motif II of the fourth CRR. The final CXXC is immediately COOH-terminal to the fourth CRR. Other proteins that possess this motif include the insulin-like growth factor-binding protein-3 (20Koedam J.A. van den Brande J.L. Biochem. Cell Biol. 1994; 198: 1225-1231Google Scholar), chaperone molecules Erp61 and Erp72 (13Lundstrom-Ljung J. Birnbach U. Rupp K. Soling H.D. Holmgren A. FEBS Lett. 1995; 357: 305-308Crossref PubMed Scopus (74) Google Scholar), the gonadotropic hormones lutropin and follitropin (14Boniface J.J. Reichert Jr., L.E. Science. 1990; 247: 61-64Crossref PubMed Scopus (86) Google Scholar, 15Chew C.C. Magallon T. Martinat N. Lecompte F. Combarnous Y. Gosling J.P. Biochem. Soc. Trans. 1995; 23: 394SCrossref PubMed Google Scholar), fibronectin (16Langenbach K.J. Sottile J. J. Biol. Chem. 1999; 274: 7032-7038Abstract Full Text Full Text PDF PubMed Scopus (95) Google Scholar), and von Willebrand factor (17Mayadas T.N. Wagner D.D. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 3531-3535Crossref PubMed Scopus (176) Google Scholar). Their active sites are compared with the active site sequences of PDI and thioredoxin in Fig. 1 c. All these proteins, with the exception of von Willebrand factor, have demonstrable endogenous thiol isomerase activity. Moreover, the four CXXC repeats found in the four CRRs of the β-integrins have a glycine (G) residue in their most NH2-terminal "X" position in common with PDI and many of the other proteins known to have endogenous thiol isomerase activity (13Lundstrom-Ljung J. Birnbach U. Rupp K. Soling H.D. Holmgren A. FEBS Lett. 1995; 357: 305-308Crossref PubMed Scopus (74) Google Scholar, 14Boniface J.J. Reichert Jr., L.E. Science. 1990; 247: 61-64Crossref PubMed Scopus (86) Google Scholar, 15Chew C.C. Magallon T. Martinat N. Lecompte F. Combarnous Y. Gosling J.P. Biochem. Soc. Trans. 1995; 23: 394SCrossref PubMed Google Scholar, 16Langenbach K.J. Sottile J. J. Biol. Chem. 1999; 274: 7032-7038Abstract Full Text Full Text PDF PubMed Scopus (95) Google Scholar, 20Koedam J.A. van den Brande J.L. Biochem. Cell Biol. 1994; 198: 1225-1231Google Scholar). We therefore assessed the platelet integrin, αIIbβ3, for the presence of endogenous thiol isomerase activity. Platelet αIIbβ3 was purified as described by Phillips et al. (18Phillips D.R. Fitzgerald L. Parise L. Steiner B. Methods Enzymol. 1992; 215: 244-263Crossref PubMed Scopus (23) Google Scholar). Coomassie staining of reducing SDS-PAGE gels shows the presence of pure protein bands corresponding to the β3 subunit and the heavy chain of αIIbat 95 and 115 kDa, respectively. The light chain of αIIbis observed at 25 kDa. There is no evidence for any additional co-purifying or contaminating proteins in this preparation (Fig.2 a). The ability of the purified αIIbβ3 to express thiol isomerase activity was determined in an assay described by Pigiet et al. (21Pigiet V.P. Schuster B.J. Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 7643-7647Crossref PubMed Scopus (123) Google Scholar). Purified αIIbβ3 (15 μm) was shown to have equivalent thiol isomerase capacity to PDI (5 μm). In contrast, ovalbumin, a cysteine-rich protein that lacks a CXXC motif, had no isomerase activity above baseline levels when compared with an uncatalyzed reaction (Fig.2 b). This activity in αIIbβ3 was shown to be saturable and both time- and concentration-dependent (Fig. 2 c). To ascertain that the activity observed in these assays was not due to contaminating or co-purifying platelet PDI, the purified integrin preparation was probed for the presence of PDI immune-reactive proteins. Platelet PDI has a molecular mass of 57 kDa and is readily seen in immunoblot analysis of platelet lysates. Only one isoform of PDI is identified in human platelets as determined by immunoblotting. Similarly, platelet cDNA scanning with degenerate primers to conserved regions of published PDI shows evidence for only one platelet isoform of PDI in agreement with Chen et al. (22Chen K. Detwiler T.C. Essex D.W. Br. J. Haematol. 1995; 90: 425-431Crossref PubMed Scopus (90) Google Scholar) (data not shown). PDI is not apparent in Coomassie-stained gels or in Western blots of purified αIIbβ3, nor could it be immunoprecipitated from large volumes of the purified integrin (Fig. 2 d). From these studies, we conclude that our integrin preparation is devoid of contaminating platelet PDI. Furthermore, in indirect immnuofluoresence studies it was found that, regardless of the activation state of the platelet, PDI did not colocalize with αIIbβ3 (Fig.3). Activated platelets, which spread on a fibrinogen-coated surface, or resting platelets, immobilized on poly-l-lysine, stain for both αIIbβ3 and for PDI. However, these proteins do not co-localize at any stage as shown in the overlay images (Fig. 3,D and H) where colocalization would be detected by the appearance of yellow pixels.Figure 3αIIbβ3and protein-disulfide isomerase do not co-localize in adhered platelets. Gel filtered platelets were allowed to adhere to poly-l-lysine (PLL, A-D) or fibrinogen (Fg, E-H)-coated slides as described under "Experimental Procedures." These were dual stained for αIIbβ3 using a monoclonal antibody (CD41a; B and F) which is complex specific and for protein-disulfide isomerase (anti-PDI; C andG) using a polyclonal antibody along with the appropriately labeled Alexa secondary antibodies. It was observed that resting platelets were obtained on the poly-l-lysine slides and spread or "activated" platelets on the fibrinogen-coated slides using differential interference contrast (DIC) mode (A and E). When these images were overlaid the absence of yellow pixels indicates that αIIbβ3 did not co-localize with protein-disulfide isomerase (D and H). Images were made on a Zeiss Axioplan 2 confocal microscope with a × 40 oil-immersion lens (1.4 n/a). Data shown is representative ofn = 3 independent experiments.View Large Image Figure ViewerDownload Hi-res image Download (PPT) A direct comparison of the abilities of αIIbβ3 and protein-disulfide isomerase to catalyze the reactivation of rdRNase was made by determining values for the k cat and K m of each protein from Michaelis-Menten plots of kinetic data (Fig.4 and TableI). Both proteins showed typical equilibrium kinetics. The affinity (K m) of the integrin for its substrate (RNase) is less for αIIbβ3 than for PDI although the catalytic capacity of the enzymes (k cat) is equivalent. The value reported here for the K m of PDI (24.7 μm) is identical to that reported by Langenbach and Sottile (16Langenbach K.J. Sottile J. J. Biol. Chem. 1999; 274: 7032-7038Abstract Full Text Full Text PDF PubMed Scopus (95) Google Scholar) in their comparative study of the intrinsic isomerase activity of PDI and fibronectin. In contrast, however, to the thiol isomerase activity of fibronectin, we show that the catalytic capacity (k cat) of αIIbβ3 is equivalent to that of PDI (2.17 × 10−4 mol min−1/μm enzyme and 2.45 × 10−4 mol min−1/μm enzyme, respectively). Thus, the intrinsic thiol isomerase activity within αIIbβ3 is potent and has high capacitance, equal to that observed in PDI. The k cat values obtained in this study are different to those obtained by Langenbach and Sottile (16Langenbach K.J. Sottile J. J. Biol. Chem. 1999; 274: 7032-7038Abstract Full Text Full Text PDF PubMed Scopus (95) Google Scholar). This may be explained by the fact that we took into consideration the contribution of spontaneous refolding of rdRNase and subtracted this away from the catalyzed reaction in our studies.Table IEnzymatic parameters of αIIbβ3 and PDIK mk catk cat/K mμmμmmin −1 / μmenzymeμm −1 min −1αIIbβ336.622.17 × 10−45.93 × 10−6PDI24.722.45 × 10−49.91 × 10−6The standard RNase refolding assay was performed using 7.5–75 μm rdRNase in the presence of 1.5 μm PDI or 1.5 μm αIIbβ3. Analysis of Michaelis-Menten data was performed using the curve-fitting programs of DeltaGraph 4.0. Values are derived from the mean data from four separate experiments performed in duplicate. Open table in a new tab The standard RNase refolding assay was performed using 7.5–75 μm rdRNase in the presence of 1.5 μm PDI or 1.5 μm αIIbβ3. Analysis of Michaelis-Menten data was performed using the curve-fitting programs of DeltaGraph 4.0. Values are derived from the mean data from four separate experiments performed in duplicate. To examine the pharmacological profile of this activity, we looked at the effects of bacitracin, a known inhibitor of protein-disulfide isomerase (23Essex D.W. Li M. Br. J. Haematol. 1999; 104: 448-454Crossref PubMed Scopus (120) Google Scholar). Bacitracin dose dependently inhibited the isomerase activity of both PDI and αIIbβ3 (Fig.5). The IC50 value is 11 μm for αIIbβ3 and 4 μm for PDI. Maximal inhibition of αIIbβ3 activity is observed at 100 μm bacitracin. Moreover, in vitro incubation of platelets with bacitracin results in a dose-dependent abolition of platelet aggregation to all platelet agonists and an inhibition of fibrinogen binding (data not shown). Thus, the endogenous thiol isomerase activity of the platelet integrin can be inhibited by the pharmacological agent, bacitracin, resulting in an inhibition of functional responses in the platelet. The isomerase activity of platelet integrin αIIbβ3 can also be regulated by divalent cations. αIIbβ3 forms a calcium-dependent heterodimer on the platelet surface and both subunits contain calcium-binding motifs. Divalent cations are required for ligand recognition and influence the affinity of interactions with the integrin receptor (24Gulino D. Boudignon C. Zhang L.Y. Concord E. Rabiet M.-J. Marguerie G. J. Biol. Chem. 1992; 267: 1001-1007Abstract Full Text PDF PubMed Google Scholar). Furthermore, extracellular divalent cations regulate integrin function by controlling the integrin ligand binding events (25Hu D.D. Barbas III, C.F. Smith J.W. J. Biol. Chem. 1996; 271: 21745-21751Abstract Full Text Full Text PDF PubMed Scopus (82) Google Scholar). We investigated the effects of calcium on the thiol isomerase activity of αIIbβ3. The observed isomerase activity in αIIbβ3 occurs in the presence of the calcium chelator EDTA (1 mm). Removal of this EDTA causes the αIIbβ3 to loose the ability to reactivate rdRNase while the enzymatic activity of PDI is unaffected by the presence or absence of EDTA (Fig. 6). These data suggest that the isomerase activity associated with αIIbβ3 is dependent upon an EDTA-stabilized conformation of the integrin. This is supported by the observation that calcium is displaced from the integrin as the complex between receptor and ligand becomes stabilized (26D'Souza S.E. Haas T.A. Piotrowicz R.S. Byers-Ward V. McGrath D.E. Soule H.R. Cierniewski C. Plow E.F. Smith J.W. Cell. 1994; 79: 659-667Abstract Full Text PDF PubMed Scopus (204) Google Scholar). Furthermore, EDTA has been demonstrated to induce a high affinity, activated conformation in αIIbβ3 (27Kouns W.C. Wall C.D. White M.M. Fox C.F. Jennings L.K. J. Biol. Chem. 1990; 265: 20594-20601Abstract Full Text PDF PubMed Google Scholar) and other integrins (28Bazzoni G. Shih D.T. Buck C.A. Hemler M.E. J. Biol. Chem. 1995; 270: 25570-25577Abstract Full Text Full Text PDF PubMed Scopus (248) Google Scholar) equivalent to that induced by agonists or ligand mimetics. A similar EDTA-stabilized active conformation is observed in thrombospondin permitting the exposure of an RGD sequence necessary for ligand binding (29Huang E.M. Detwiler T.C. Milev Y. Essex D.W. Blood. 1997; 89: 3205-3212Crossref PubMed Google Scholar). Finally, in order to determine if this enzymatic activity is a common property of integrins, we examined commercial preparations of purified αvβ3 and α5β1for the presence of endogenous thiol isomerase activity. Fig.7 shows that the vitronectin receptor, αvβ3 has thiol isomerase activity. The activity of the αvβ3 is greater than that of αIIbβ3. This data suggests that the activity does indeed reside in the common β3 integrin subunit. The leukocyte fibronectin receptor, α5β1, also expresses thiol enzymatic activity but, under the conditions tested, its potency appears less than that of the β3 integrins. It may be that the conditions of time, temperature, pH, and Ca2+concentrations optimized for αIIbβ3 are not ideal for α5β1. The distinct potencies of the different integrins may suggest that the enzymatic activity of each integrin is regulated by their individual α-subunits, or that each may operate under discrete optimal conditions not addressed in this study. The contribution of thiol bonds to the integrin structure and function has been implied by many other groups (7Calvete J.J. Proc. Soc. Exp. Biol. Med. 1999; 22: 29-38Google Scholar, 30Yan B. Hu D.D. Knowles S.K. Smith J.W. J. Biol. Chem. 2000; 275: 7249-7260Abstract Full Text Full Text PDF PubMed Scopus (61) Google Scholar, 32Ni H. Li A. Simonsen N. Wilkins J.A. J. Biol. Chem. 1998; 273: 7981-7987Abstract Full Text Full Text PDF PubMed Scopus (95) Google Scholar). In platelet functional assays, it has been shown that treatment with the reducing agent dithiothreitol causes platelet aggr
Referência(s)