[15 Redox Control of Integrin Adhesion Receptors

By jeffrey w. smith, boxu yan, france landry, and Christian R. Lombardo

Reagents

The following is a list of reagents that have been used in our studies of the integrin redox site. The manufacturer of each reagent is provided.

Triton X-100 (Sigma, St. Louis, MO) Phenylmethylsulfonyl fluoride (PMSF; Sigma) Leupeptin (Sigma)

Methyl-a-d-mannopyranoside (Sigma) Guanidine-HCl (Sigma) Trichloracetic acid (Sigma) Dimethyl sulfoxide (Sigma) Glutathione (reduced form; Sigma) l-Cysteine (Sigma)

Avidin-horseradish peroxidase (HRP) (Sigma)

Concanavalin A (ConA)-Sepharose (Amersham Pharmacia Biotech,

Piscataway, NJ) CNBr-Sepharose (Amersham Pharmacia Biotech) Aquacide (CalBiochem-Novabiochem, La Jolla, CA) [14C]Iodoacetamide (New England Nuclear, Boston, MA) Dimethylformamide (Aldrich, Milwaukee, WI)

Integrin Redox Site

Integrins are transmembrane adhesion receptors that play an essential role in normal tissue development and homeostasis. Many integrins bind to their li-gands through the Arg-Gly-Asp tripeptide sequence, which is displayed by a number of extracellular matrix proteins, plasma proteins, and even viruses.1 Unlike many other classes of receptors, the integrins participate in bidirectional signaling across the membrane.2 "Inside-out" signals activate, and deactivate, the integrin ligand-binding function. "Outside-in" signals are generated when integrin binds to its ligands and regulate a host of cellular processes including cell proliferation and cell death. We identified a redox site within the extracellular domain of the integrin.3 This redox site is composed of between two and five unpaired cysteines, which appear to reshuffle during conformational transitions in integrin. Therefore, we suggested that this redox site has the properties that might be expected of a conformational switch involved in bidirectional signaling.

The realization that integrins contain a redox site is so recent that only the initial steps toward establishing experimental methodology have been completed. We describe methods for (1) quantifying the number of free sulfhydryls within the integrin redox site, (2) tagging these free sulfhydryls with site-specific modification reagents, and (3) pinpointing the position of the free sulfhydryls that comprise the integrin redox site. Much of our effort has been directed toward the platelet integrin ffnb/}3, also known as platelet glycoprotein lib Ilia.4 Integrin cfnb^3 serves as an excellent paradigm for the study of other integrins because it participates in both inside-out and outside-in signaling, and because it can be purified from outdated platelets in milligram quantities. Equally as important, two conformers of this integrin can be obtained. Activation state 1 (AS-1) represents the resting integrin, has low affinity for ligand, and an "oxidized" redox site. In contrast, activation state 2 (AS-2) has high affinity for physiologic ligands and a "reduced" redox site. The purification of AS-1 and AS-2 is described as a starting point.

Purification of Resting and Active Conformers of Platelet Integrin am, $3

Our strategy for purifying AS-1 and AS-2 is rooted in methods that have been previously reported,5-7 with some modifications. A key aspect of the purification

1 E. Ruoslahti, Annu. Rev. Cell Dev. Biol. 12,697 (1996).

3 B. Yan and J. W. Smith, J. Biol. Chem. 275, 39964 (2000).

4 D. R. Phillips, I. F. Charo, and R. M. Scarborough, Cell 65, 359 (1991).

5 L. A. Fitzgerald, B. Leung, and D. R. Phillips, Anal. Biochem. 151,169 (1985).

6 R. Pytela, M. D. Pierschbacher, S. Argraves, S. Suziki, and E. Ruoslahti, Methods Enzymol. 144,475 (1987).

7 W. C. Kouns, P. Hadvary, and B. Steiner, J. Biol. Chem. 267, 18844 (1992).

is the separation of AS-1 from AS-2 by ligand affinity chromatography. Outdated human platelets (100 units) are subjected to centrifugation at 5500g for 30 min at 4° in a Beckman (Fullerton, CA) JA-10 rotor (or equivalent). Red blood cells also pellet with the platelets but are removed by gently inverting the centrifugation tube. The platelet pellet is further washed by gentle resuspension in 20 mM Tris-HCl (pH 7.4), 150 mM NaCl, containing 0.1 mM EDTA, followed by centrifugation at 5500g for 30 min at 4°. This wash is repeated one additional time. Platelets are subsequently lysed at 4° by addition of 5 volumes of 20 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1 mM CaCl2 and containing 1% Triton X-100, 5 mM PMSF, and 10~5 M luepeptin. This sample is maintained at 4° for 14 hr with gentle rocking. The lysate is cleared of insoluble material by centrifugation at 50,000g in a Beckman JA-25.5 rotor at 4° for 1 hr. The supernatant, containing aiib/63, is collected and used for subsequent chromatography. We have found it convenient to store the supernatant from the platelet lysate at -70° for up to 4 months.

The platelet lysate is thawed rapidly at 37° and immediately centrifuged again at 50,000g in the Beckman JA-25.5 rotor. The supernatant is adsorbed to concanavalin A-Sepharose in batch. Aliquots of fresh PMSF (1 mM) and leupeptin (10~5 M) are added to inhibit proteolysis during batch adsorption. The lysate is incubated with ConA-Sepharose for 18 hr at 4°. The ConA-Sepharose resin is packed into a low-pressure column and washed with 20 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1 mM CaCl2, 1 mM MgCl2 containing 1% Triton X-100 (buffer A) until the optical density of the eluate stabilizes. Although Triton X-100 absorbs at 280 nm, we have found monitoring of the wash at this wavelength to be an excellent indicator of the level of nonspecifically bound proteins eluting during the wash. Integrin (both AS-1 and AS-2) is subsequently eluted from ConA-Sepharose in buffer A containing 200 mM methyl-a-d-mannopy-ranoside. The eluate is examined by Coomassie staining of sodium dodecyl sulfate (SDS)-polyacrylamide gels, and fractions containing integrin ailhp3 are pooled.

The partially purified integrin is recirculated for 18 hr at 4° over an affinity resin containing the peptide KYGRGDSP linked to CNBr-agarose. We have found that a column containing 8 ml of resin linked to approximately 1 mg of peptide per milliliter is sufficient to retain the vast majority of the AS-2 from 100 units of platelets. The flowthrough fraction, containing the nonbinding AS-1, is retained for further purification (see below). To obtain AS-2, the KYGRGDSP affinity resin is washed with 10 column volumes of buffer A, and is then eluted by addition of 0.5 mM GRGDSP-amide in buffer A. Fractions containing AS-2 are pooled and dialyzed exhaustively against buffer A to remove free RGD peptide, and are then stored at —70° in 5% (v/v) glycerol.

The flowthrough from the KYGRGDSP affinity resin is then used as a source for AS-1. This material is recirculated over heparin-agarose for 18 hr at 4° to remove heparin-binding proteins as described.5 The nonbinding fraction, containing AS-1, is concentrated to approximately 8 ml by placing it into a dialysis bag and overlaying with Aquacide (Mr 500,000 cutoff). The concentrated sample is dialyzed against buffer A, and then passed over a Sephacryl S-300 column equilibrated in the same buffer. Highly purified AS-1 elutes as a single broad peak, and can be detected by Coomassie-stained SDS-polyacrylamide gel electrophoresis (PAGE). Fractions containing AS-1 are pooled and stored with 5% (v/v) glycerol as a stabilizing agent at —70° until use.

Quantifying Free Sulfhydiyls in Integrin

The quantification of free sulfhydryls within the integrin redox site is an essential first step. As described previously,3 the number of unpaired cysteines changes as a consequence of integrin activation, and is likely to be mechanistically linked to activation. We have modified standard carboxymethylation procedures to enable the quantification of free sulfhydryls within integrin. Integrin, at a concentration of greater than 500 /xg/ml in buffer A (although 50-100 mM octylgucoside is also acceptable), is first denatured by the addition of 8 volumes of 6 M guanidine-HCl. After a 30-min incubation, EDTA is added to a final concentration of 0.2 M (diluted from a 2 M stock, pH 8.0). We have found this chelation step to be essential for the accurate quantification of free sulfhydryls. After a 15-min incubation at ambient temperature with EDTA, free sulfhydryls in the integrin are alkylated by addition of [14C]iodoacetamide (New England Nuclear) at a 500:1 (iodoacetamide: integrin) molar ratio. Alkylation is allowed to proceed for 1 hr at 25°. The alkylated integrin is precipitated by the addition of a 1/5 volume of ice-cold 50% (w/v) trichloroacetic acid (TCA) to bring the final concentration of TCA to 10% (w/v). The sample is maintained on ice for 1 hr, and is then precipitated by centrifugation for 20 min at 14,000 rpm in a microcentrifuge maintained at 4°. It is anticipated that in some cases it will be difficult to achieve integrin concentrations that are sufficient to promote precipitation. Therefore, the addition of carrier protein, to a concentration of 1 mg/ml, to facilitate precipitation would seem prudent. It is important, however, that the carrier be devoid of unpaired cysteines, as are found in bovine serum albumin. We suggest horse myoglobin as a suitable alternative. After centrifugation, the pellet is washed six times with 50 mM Tris-HCl (pH 7.4), 100 mMNaCl. After precipitation, the alkylated integrin can be brought into solution by the addition of 10% (w/v) SDS (in water) and heating to 50°. Frequent agitation over a period of 30 min is normally sufficient to solubilize the precipitated integrin. The sample is subjected to scintillation counting, from which the molar ratio of [14C]iodoacetamide to integrin can be derived. In our experience, it is best to compare an identical sample of integrin that is alkylated with unlabeled iodoacetamide before alkylation with [14C]iodoacetamide to control for nonspecific incorporation of the isotope.

Site-Specific Modification of Integrin Redox Site

Tagging unpaired cysteines with site-specific sulfhydryl modification reagents has also facilitated analysis of the integrin redox site. By using modification reagents linked to biotin, the modification can be readily detected by standard blotting approaches. A number of such reagents are commercially available. In our experience, only reagents greater than 29 Â in length are able to label the redox site on the two ^-integrins. This limitation may be due to the inherent depth of the site. Two biotinylated modification reagents have yielded good results. These are 1 -biotinamido-4- [4' (maleimidomethyl)-cyclohexane-carboximido]butane (biotin-BMCC) and A^-[6-biotinamido)hexyl]-3'-(2'-pyridyldithio)propionamide (biotin-HPDP), both purchased from Pierce (Rockford, IL). Biotin-BMCC generates a maleimide link with free sulfhydryl that is insensitive to reduction, whereas biotin-HPDP creates a disulfide link with free sulfhydryls that is readily released by reduction. Although we have found both reagents to reproducibly label «iibfo, our analysis indicates that not all the unpaired cysteine residues within the redox site are modified by these reagents. Rather, alkylation of these cysteine residues of AS-1 with biotin-BMCC blocks the incorporation of between 35 and 60% of the [14C]iodoacetamide incorporated during alkylation. Hence, biotin-BMCC is tagging about one-half of the free sulfhydryls within the redox site. Because biotin-HPDP is sensitive to reduction, we have been unable to measure what percentage of the free cysteines this reagent will modify.

We have found the following method to reproducibly modify the redox site within aiibP3, and av/?3, and consequently expect that the same approach could be applied to examine the redox site in other integrins. Biotin-BMCC is made fresh as 4 mM stock in dimethyl sulfoxide. Biotin-HPDP should be made in dimethyl-formamide. Either reagent can be added directly to purified integrin (1-10 jig) in buffer A. Modification is allowed to proceed for 1 hr at 25°. We have found the redox site of arnb/*3 to label across a concentration range of these reagents, with the labeling saturating at a reagent concentration of approximately 100 ¡iM. The reaction can be quenched by addition of a 100-fold molar excess of reduced glutathione, or L-cysteine. After modification, the integrin can be separated by SDS-PAGE and transferred to polyvinylidene difluoride (PVDF) filters, and the modification assessed by probing the membrane with HRP-avidin (Sigma). It is conceivable that an enzyme-linked immunosorbent assay (ELISA) could also be devised to quantify the extent of modification with biotin-BMCC or biotin-HPDP.

This labeling strategy can also be adapted to examine the redox site within integrin on the surface of cells. For this application, the sulfhydryl modification reagents are added directly to cells in a pellet, making sure that the final concentration of dimethylsulfoxide does not exceed 5% (v/v). Best results are obtained when the modified integrin is subsequently immunoprecipitated from cell lysates, and then analyzed by SDS-PAGE and blotting to PVDF. Integrins are known to have a number of regulatory divalent cation-binding sites, but we have yet to perform rigorous experiments to determine whether the occupation of these sites influences the accessibility of the redox site to site-specific modification reagents. Therefore, to reduce variability in outcome, particular attention should be given to maintaining consistent levels and types of divalent cations.

Identification of Free Sulfhydiyls with Mass Spectrometry

All integrin fi subunits contain 56 conserved cysteine residues, many of which are located within a 200-residue cysteine-rich domain near the C terminus. The redox site appears to be positioned within this domain, where approximately 40% of all residues are cysteine. Consequently, obtaining definitive information about the position of the unpaired cysteines will be particularly challenging. To identify cysteines within the integrin redox site that are tagged with biotin-BMCC, we have begun to explore the use of precursor ion scanning with a triple quadrupole mass spectrometer.8,9 Precursor ion scanning allows parent ions with specified properties to be identified on the basis of the mass spectrum of their daughter ions. Hence, by fragmenting the parent ion, and applying mass filters to identify daughters with a fragmentation pattern consistent with the conjugate between biotin-BMCC and cysteine, the parent peptide could presumably be identified. This parent peptide could then be immediately subjected to tandem mass spectrometry (MS/MS) sequencing to pinpoint the position of the tagged cysteine within the whole integrin.

To this point we have focused on establishing a robust method for such analysis, using glutathione (GSH) as a model peptide that contains an unpaired sulfhydryl. Our objectives have been to establish the fragmentation pattern of biotin-BMCC when linked to cysteine, and to test the concept that the conjugate between biotin-BMCC and GSH can be identified from a precursor ion scan of its daughter ions. The conjugate between biotin-BMCC (Fig. 1A) and GSH was analyzed with an ABI3000 electrospray mass spectrometer (Applied Biosystems, Foster City, CA) in positive ion mode. A 20 fiM sample of BMCC-GSH conjugate was infused into the mass spectrometer, using a Harvard syringe pump, at a flow rate of 20 /xl/min. This yielded the Q1 spectrum shown in Fig. IB. This spectrum reveals the singly charged parent ion (m/z =841) of the conjugate. This conjugate was selected in the first quadrupole and then subjected to MS/MS fragmentation in the second quadrupole. The first quadrupole was set to transmit only the parent ion (m/z = 841). Collision with nitrogen gas [collisionally activated dissociation (CAD) gas setting = 7) at —80 V in the second quadrupole (R02

8 M. Mann and M. Wilm, Trends Biochem. Sci. 20, 219 (1995).

9 K. Chatman, T. Hollenbeck, L. Hagey, M. Vallee, R. Purdy, F. Weiss, and G. Siuzdak, Anal. Chem.

0 0

Post a comment