This action might not be possible to undo. Are you sure you want to continue?
JOURNAL OF BIOLOGICAL CHEMISTRY
8 1989 by The American Society for Biochemistry and Molecular Biology, Inc.
Vol. 264, No. 35, Issue of December 15, pp. 21080-21086,1989 Printed in U.S.A.
Isolation and Characterization of a Putative Collagen Receptor from Staphylococcus aureus Strain Cowan l*
(Received for publication, July 17, 1989)
Lech M. SwitalskiS, Pietro Speziales, and MagnusHijijk$Ti
From the Departmentsof $Microbiologyand TBiochemistry, University of Alabama, Birmingham, Alabama 35294 and the $Department of Biochemistry, Uniuersity of Pauia, Pauia, Italy 27100
In a previous study we demonstrated that cells of Staphylococcus aureusstrain Cowan bind ‘2sI-collagen in a receptor-ligand type of interaction (Speziale, P., Raucci, G., Visai, L., Switalski, L. M., Timpl, R., and Hiiiik, M. (1986) J . Bacteriol. 167,77-81). In the present communication we report on the isolation and preliminary characterization of a putative collagen receptor from a lysateof S. aureus strain Cowan. Antibodies raised against a collagen receptor positive strain inhibit the binding of ‘2sI-collagen to bacterial cells, whereas antibodies raised against a collagen receptor negative strain were without effect. Solubilized cell surface components did not exhibit anymeasurable affinity for collagen-Sepharose. However, the inhibitory effect of the antibodies against bacterial cells was neutralized by the lysate from a receptorpositive but not receptor-negative strain. A collagen receptor assay was designed based on this observation and used to develop a receptor purification protocol involving anion exchange chromatography, ammonium sulfate precipitation, and gel chromatography. Using this procedure a protein with anapparent M, of 135,000 was purified. This protein which was present on a collagen receptor-positive strain butnoton a receptor-negative strain could completely neutralize the inhibitory activity of the antibodies raised against S. aureus strain Cowan. Furthermore, antibodies raised against the 135-kDa protein inhibited the binding of collagen to bacteria, and this protein is tentatively identified as a collagen receptor.
A bacterial cell may simultaneously express receptors (adhesins) forseveral matrixproteins. Staphylococcusaureus strains may independently bind fibronectin (Kuusela, 1978; Switalski et al., 1983), fibrinogen (Hawiger et al., 1982), laminin (Lopes et al., 1985), vitronectin (Chhatwal, 1987), and collagen. In some instances these proteins have been shown in in vitro assays toserve as substrates for bacterial adherence (Herrmann et al., 1988; Kuusela et al., 1985; Vaudaux et al., 1984; Vercellotti et al., 1984; Wadstrom et al., 1987). So far the best characterized bacterial receptor for a matrix protein is the S. aureus receptor for fibronectin. This receptor isolated from bacterial cells solubilized with lysostaphin is multivalent and migrates as a protein with an apparent M , of 210,000 when analyzed on PAGEZin SDS (Froman et al., 1987). The gene coding for this has been cloned (Flock et al., 1987) and sequenced (Signas et al., 1989). The fibronectin binding activity of the receptor has been localized to a domain which is composed of a 38-amino acid unit repeated3 times. Synthetic peptides mimicking these threesequences interact with fibronectinasindicated by theirabilitytoinhibitbinding of fibronectin to intact bacteria (Signas et al., 1989). Previous studies on the binding of collagen to S. aureus cells have shown that staphylococci may recognize collagens regardless of type and species of origin (Carret et al., 1985; Holderbaum et al., 1986; Mamo et al., 1988; Vercellotti et al., 1985). Moreover, collagen receptor-positive strains of staphylococci also recognize isolated a chains and peptides generated by CNBr cleavage of a chains of type I collagen. Synthetic polypeptides (PCP), mimicking the structure of collagen were found to inhibit the binding of collagen to bacteria, Many potentially pathogenic microorganisms bind to extracellular matrix proteins such as fibrinogen, fibronectin, and suggesting that staphylococci recognize repetitive structures laminin. Itis believed that these interactions represent mech- in the collagen molecule (Speziale et al., 1986). A collagen receptor from bacteria has previously not been anisms of host tissue adherence. Bacterial cells possess specific receptor’ molecules on their surface which bind to dis- identified. In this communication we report on the isolation and characterization of a S. aureus surface protein which tinct sites in the matrix proteins of the host. tentatively is identified as a collagen receptor.
* This work was supported in part by Grant AI 20624 from the National Institutes of Health and grants from the A h a Lava1 AB, Tumba, Sweden, the Arthritis Foundation Biomedical Project, NATO (389/84), and Minister0 della Saniti. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate thisfact. Aswe have pointed out in our previous paper (Froman et al., 1Y87), we use the term “receptor”referring to the components on the bacterial cell surface interacting with collagen. This is contrary to some other reports where this term describes components of the host tissue, which bacterial “adhesins” recognize. Although our receptors do not fulfill all the criteria traditionally associated with the receptors ofeukaryoticcells, we feel that referring toextracellularmatrix proteins asreceptors is improper; they resemble the eukaryoticreceptors functionally in their ability to interact with host components. After provingthat this interaction plays arole in bacterial attachment to host tissues, bacterial “collagen receptor” can be referred to as a “collagen adhesin.”
MATERIALS AND METHODS
Collagen-Collagen type I1 isolated from chicken sternum (Reese and Mayne, 1981) was a gift from Dr. Richard Mayne, University of Alabama a t Birmingham. NalZ51,specific activity 15 mCi/pg, was obtained from Amersham Corp. Collagen was labeled with lZ51 by the chloramine-T method (Hunter, 1978). Although the content of tyrosine in collagen is low, we have observed that extending the labeling time to 3 min yields ‘251-collagenof high specific activity, which migrates identically with unlabeled collagen on PAGE in SDS. The specific activity of the labeled protein was typically 0.5-1.0 X lo6 cpmlrg. Antibodies-Antibodies against whole cells of S. aureus strains Cowan and Newman were raised in New Zealand rabbits employing
* The abbreviations used are: PAGE, polyacrylamide gel electrophoresis; SDS, sodium dodecyl sulfate; PBS, phosphate-buffered saline.
at 37 "C overnight.14 M sodium chloride.. and the tubes were immediately centrifuged at 1350 X g for 20 min. Bacteria were collected by centrifugation and resuspended in PBS to a final density of 1 X 10" cells/ml. 10 m M N-ethylmaleimide and incubated under constant rotation a t 37 "C. 10 g (wet weight) of bacteria grown as described above were suspended in 100 ml of TBS (0.1% milk for 2 h at 37 "C.01% bromphenol blue. 1987) to contain e. aureus Cells Staphylococcal cells were digested with different enzymes in attempts to release the collagen receptor. If we assume that the effect of the antibodies is the consequence of an antibody binding to collagen receptors. Similar assays have previously been used to quantify adhesion molecules on Dictyostelium discoideum (Muller and Gerisch. 0. lysostaphin. and Fc fragments were separated on protein ASepharose. Data are expressed relative to control. and background values representing radioactivity recovered in the tubes incubated in the absence of bacteria were subtracted. The number of cells was determined by comparing the absorbance of the sample with a previously prepared standard curve relating Am to the cell number determined by counting cells in a Petroff Hausser chamber. The suspension was supplemented with 1 mg of lysostaphin. On the other hand digestion with lysozyme cleaving only the carbohydrate chain of the peptidoglycan did only marginally reduce the '251-collagenbinding activity of the treated bacteria. lZ5Z-Collagen Binding Assay-The amount of 1251-collagen bound to bacteria was quantified essentially as described previously (Froman et al. and washed with TBS. The digestion was stopped by boiling the samples for 10 min at 37 'C. and papain. The amount of material reversing the inhibitory activity of antibodies by half (i. the amount of a n t i 4 aureus Cowan Fab fragments causing 80% inhibition of 1251-collagen binding to intact bacterial cells in a standard assay described abovewas determined and used as a base line. Bacteria were harvested by centrifugation. were incubated with 20 pgof each enzyme for 2 h at 37 "C. except that the number of bacterial cells/incubation mixture was reduced. Briefly. 1987). Electrophoresis and Western Blotting-Electrophoresis in polyacrylamide gel was performed according to Blobel and Dobberstein (1975). Assay of CollagenReceptor-In initial experiments lysostaphinsolubilized bacterial cell surface components were found not to inhibit the binding of 1251-collagen to S. Alternative receptor assays therefore hadto be explored. andheated a t 88 "C for 20 min to kill bacteria and inactivate hydrolases. we can measure molecules containing the active epitope (e. 1 0 ' cells of S. aureus strains Cowan and Newmanwere cultured under constant rotation for 15 h a t 37 "C in brain heart infusion broth (Difco). and an alternative method had to be developed. Additional binding sites were blocked by incubating the membrane with 1% defatted bovine milk in TBS for 2 h a t 22 "C. The gels werestained with Coomassie Brilliant Blue R-250 and dried. aureus cells. NY). thermolysin. pH 7. and stored in small aliquots. and the of the supernatant (referred to in the text as the "lysate") was adjusted to 7.02% azide. The supernatant was stored at -20 "C until used. 1978) and vertebrate cells (Ocklind and Obrink. Enzyme-conjugated antibody was detected by a color reaction using 4-chloro-l-naphthol as a substrate according to the manufacturer's instructions. Depending on the batch of antibodies (preabsorbed with cells of strain Newman) this amount ranged from 9 to 17 pg of Fab.1% bovine serum albumin. The membranes were incubated with purified Fab fragments (100-200 pg/lOO ml) in TBS. Bacterial pellets suspended in the initial volume of PBS were assayed for their ability to bind '251-collagen in the standard assay. Material released from bacteria during enzyme digestion was tested for its ability to inhibit '251-co11agen binding to intact cells. 0 20 40 60 60 100 RELATIVE Yo OF' I-COLLAGENBINDING * FIG. Residual binding by enzyme-treated cells was quantified and compared with that of untreated bacteria (Fig. Plainview. lost essentially all collagen binding activity as did bacteria treated withpeptidoglycan degrading enzyme. Proteins separated by electrophoresis were electroblotted for 2. Antibodies against collagen receptor-positive cells of S. The samples were boiled in a buffer containing 4% sodium dodecyl sulfate.4) and sonicated for 3 min (Heat Systems Ultrasonics. . trypsin. After 30 min a second 1 mg of lysostaphin was added. 5 mg of papain/100 mg of IgG). The enzyme digestion was stopped by heating the incubation mixture for 10 min at 88 "C. Solubilization of '29-collagen binding components from S . aureus cells.g. suspended in PBS. 5000 X g). i. aureus the immunization protocol of Lofivist and Sjoquist (1963). 0. 1987) except that bacterial cells were sonicated before treatment with enzymes. 1.. pH 7. aureus strain Cowan in a total volume of 0. Somewhat surprisingly noneof the preparations tested including the lysate generated by lysostaphin digestion (shown in the past (Froman et al. 10 m M phosphate.1% Tween 80 in PBS were incubated for 1 h at 20 "C. dialyzed toPBS (0.e. Bacterial pH debris was removed by centrifugation (20 min. aureus strain Cowan were found to inhibit the binding of '251-collagen to bacteria. 2 m M dithiothreitol. RESULTS Binding of '251-Collagen to Enzyme-digested S.5 ml containing 5 X lo4 cpm of '251-collagen type 11. containing 0. 1 m M phenylmethylsulfonyl fluoride.1% Tween 80. MA). 50 m M Tris-HC1. and the clear supernatant was collected and stored at -20 "C. receptors) in an assay where soluble cell surface components are used to neutralize the inhibitory activity of the antibodies. 1980). The incubation mixture with papain was supplemented with 2 m M EDTA and 10 m M cysteine. Solubilization of Bacterial Cell Surface Component-Bacterial components were solubilized with lysostaphin in the presence of protease inhibitors essentially as described previously for the isolation of fibronectin receptor (Froman et ai.1% milk. Fab fragments were generated by digestion ofIgG with papain coupled to Sepharose (PBS buffer. bacteria incubated in the absence of enzymes. 1982.Collagen Receptor of S.14 M NaCl. 82 m M Tris-HC1 followed by alkylation with iodoacetamide. Duplicate samples were analyzed. The reaction was stopped by the addition of 3 mlof ice-cold PBS containing 0. The degree of neutralization of inhibitory activity of antibodies plotted uersus the log of added lysate typically gave a linear graph. In some experiments Fab fragments were absorbed with S. a direct inhibition assay receptors in thesedigests is notpossible.4).4.5 h. Briefly. and the mixtures were centrifuged. Urushihara and Takeichi. Solubilized receptors could therefore not be detected or quantified in a conventional inhibition assay. The incubation mixtures were centrifuged. The mixture was subsequently added to intact bacteria and incubated with '251-collagen as described above.Polyclonal antibodies against purified collagen receptor were also raised in rabbits by three intramuscular injections of 50 pg of receptor protein emulsified in Freund's incomplete adjuvant (Difco) at 10-day intervals.g. Since these digests should contain solubilized receptors or to quantify collagen their fragments. Bedford. 0. active fibronectinreceptorsorprotein A) significantly interfered with the binding of collagen to bacteria (data not shown). 2 mg of DNase. (1979). 1). eluted with 3 M MgC12. Blotting followed basically the procedure of Towbin et al. After aspiration of the supernatant the pellet remaining in the tubes was analyzed for radioactivity in a y counter. 10'" cells in 1 ml of PBS. andthe incubation was 21081 continued for an additional 2. Rabbits were bledat 2-week intervals. Subsequently membranes were incubated with goat anti-rabbit IgG conjugated with horseradish peroxidase (Bio-Rad) diluted 1:3000 with TBS containing 0. Antibodies were preincubated with varying amounts of solubilized receptor preparations for 30 min at 37 "C. 30% sucrose. from 80% inhibition to 40% inhibition) was arbitrarily defined as one receptor unit. Bacteria. Bacteria-S. The IgG fraction of serum was purified by affinity chromatography on protein A-Sepharose. Fivehundred pg of Fab fragments were absorbed with 50 mg (wet weight) of heat-killed bacteria for 1 h at 37 "C and overnight a t 4 "C.. and 0.e. To make this assay as sensitive as possible. aureus Cowan cells.5 h at 400 mA onto Immobilon-P membranes (Millipore.Bacterial cells digested with proteolytic enzymes.
Hence. the same data asa semilogarithmic plot. anti-S. Inhibition of collagen binding was observed with both intact IgG (not shown) and Fab fragments. respectively. An amount of receptor . d FIG. In this assay. Fab ADDED (111) FIG. aureus Cowan collagen receptor ( 0 ) . 4. had no effect (data not shown). Subsequently cells of s. 3 in that the inhibitory activity of anti-Cowan antibodies was neutralized only by preincubation of the Fab fragments with lysate of strain Cowan but not with lysate of strain Newman." The results are expressed as a percentage relative to binding in the absence of antibodies. aureus Inhibition of '251-CollagenBinding to S. details of which are described under "Materials and Methods. while absorption with cells of a collagen receptor-negative strain (Newman) only marginally reduced the inhibitory activity of the antibody preparation. 3. 1 X lo8 cells. L l O ob 20 30 40 5 I O 20304 PROTEIN ADDED ( u g . Symbols: unabsorbed antibodies.e. it may also be possible to absorb this activity using solubilized bacterial components. aureus Cowan cells (Fig. The incubation is continued. 4 B ) .R E C E P T O R ) IO I A Collagen Receptor Assay If the '251-collagenbinding inhibitory activity of anti-Cowan antibodies may be specifically absorbed using whole cells of collagen receptor-positive bacteria.L Y S A T E . The absorbed antibodies were tested in the neutralization assay as indicated in the legend to Fig. aureus Cowan were preincubated with 50 mg (wet weight) of bacteria. The amount of bound '2sII-collagenwas determined as described under "Materials and Methods. Antibodies against other previously identifiedstaphylococcal cell wall components. we routinely used Fab fragments rather than IgG in ourassays. aureus and anti-collagen receptor antibodies with homologous and heterologous bacteria. and caused 90% inhibition of binding. Dashed line indicates the concentration of material required to restore half of the inhibition caused by the addition of the indicated amount of antibodies (40% of inhibition. Inhibition of '2sI-collagen binding to cells of S. IgG and Fab fractions of these antibodies were testedaspotentialinhibitors of '251-collagen binding to S. This amount of material is defined as one receptor unit. 2). 200 pg/ml (O). and a linear relationship between neutralization activity and amount of lysate added was observed when the data were plotted on a semilogarithmic scale (Fig. 3). aureus Newman immune (0) antibodies. 4 ) were consistent with those presented in Fig. To avoid any possible interference of protein A with the antibodies. aureus Cowan Fab fragments (60 pl. Neutralization of inhibitory activity of polyclonal anti-S. ' O O k5 0 2 5 Fab 75 A D D E D (1111 100 2 0 0 . The remaining antibodies were tested for their ability to inhibit the binding of '251-collagento cells of strain Cowan (Fig. Panel A presents the data in the linear scale. Anti-Cowan andanti-NewmanFabfragments were absorbed with cells of strain Cowan or Newman.g. aureus Cowan and '251-collagenare added. the amount that reduces collagen binding by 80%. see Fig." predetermined amountsof anti-Cowan antibodies which cause 80% inhibition of '251-collagen bindingtobacteriaareinitially incubated withvarying amounts of solubilized receptor. Absorbed and unabsorbed antibodies were checked as potential inhibitors of binding of '251-collagento S. antibodies absorbed with cells of S. aureus Cowan by antibodies. andthen subsequentlyassayedfor theirabilityto FIG. No significant inhibition of '251-collagenbinding was obtained by preimmune antibodies or by antibodies raised against cells of strain Newman. Absorption of inhibitory activity of polyclonal antis. i. 2 ) were incubated for 1 h at 37 "C with the indicated . 12 pg. e. Bacteria. aureus Cells by Antibodies Antibodies against whole cells of collagen receptor-positive strain Cowan and collagen receptor-negative strain Newman were raised in rabbits. recognize epitopes on thecollagen receptorlocated at or close to the collagen binding site. fibronectin receptor or protein A. which block '251-co11agenbinding. To test this hypothesis anti-Cowan antibodies were incubated with bacterial components solubilized by lysostaphin digestion of bacterial cells of strains Cowan and Newman. 2. Fab fragments (500 pg/2. aureus antibodies withlysates ofhomologousand heterologous bacteria. 26 pg/ml). inhibit the binding of'251-co11agen to intact bacteria. 60% of binding). 2. i. 4A). and the amount of labeled ligand bound to the bacteria is determined. aureus Cowan ( 0 .u g x IO2.e. Symbols: a n t i 3 aureus Cowan preimmune (A) and immune (0). The observed neutralization of antibody activitydepended on the concentration of the lysates (Fig. Based on these observations an assay wasdesigned for quantitating solubilized collagen receptors. We here assumed that the component(s) responsible forthe neutralizing activity is identical to the collagen receptor. This observation suggested thatanti-Cowanantibodies. Only immune antiCowan antibodies effectively inhibited binding of '251-collagen to bacteria. The amount of collagen bound was reduced by half when 1 0 ' bacterial cells were incubated with 15 pg of 50 pg of the same antibodies anti-Cowan Fab fragments.5 ml) against S. 2 5 pg/ml).antibodies absorbed with cells of S. The results (Fig. Immune anti-S. panel B . The inhibitory activity of anti-Cowanantibodies could be neutralized by absorption with homologous bacteria. were incubated with the indicated amounts of Fab fragments (200 pg/ml) for 1 h priortoincubationwith '251-collagen for 30 min. the epitope(s) which the inhibiting antibodies recognize is present on a collagen receptor-positive strain but noton a receptor-negative strain. aureus Newman (A. aureus Cowan.21082 Collagen Receptor of S. aureus Cowan (0)or amounts of lysostaphinlysates of strain 5' Newman (0) or the purified 5'.
After extensive dialysis the material not precipitaof protease inhibitors as described previously for solubiliza. Peaks containing appreciable receptor activity.ble with 60% ammonium sulfate (pool IIa) was freeze-dried tion of fibronectinreceptors of S. although after prolonged digestion (18 h) with these three enzymes a minor band with M. aureus strain Newman and subjected to the next step of purification. 7. in SDS (Fig. 7. DEAE-Sephacel chromatography of S. The gel chromatography step of puribuffer in which the solubilized material was applied to the fication resulted in loss of about half of the active material.receptor activity is limited only to the high molecular weight Sepharose column nor did lowering the ionic strength of the component (135. Rechromatography of out that heat treatment may lower the affinity of collagen to the 135-kDa protein(Fig. 9. However. The molecular freeze-dried.7 M). Support for this conof DEAE-Sephacel when applied at neutral pH and low ionic clusion was obtained in an experiment where samples from strength.4. dialyzed to water. 5. This material could Purification of the Collagen Receptor be precipitated with 60% ammonium sulfate. (1986) pointed k ) . Practically all the receptor activity could be recovered at this purification step ubilization of collagen receptorsfrom staphylococcalcells involves digestion of bacteria with lysostaphin in the presence (Table I). of 115. this of the high and thelow molecular weight materials.000) as shown by PAGE 1251-collagen to bacteria. These data suggest that the epitope which the inhibitory antibodies recognize is of protein nature.000). Ib. lane b). Digestion with lysozyme. were combined. 21083 of material purified by ion exchange chromatography (Fig. Affinity chromatography was therefore not a viable but overall recovery of receptor activity was 33%. lane j ) . the immunoreactive 135-kDa protein is not present in lysate generated from thecollagen receptor-negaE tive strain Newman nor do antibodies raised against strain 0 m N Newman react with the 135-kDa protein in lysate of strain " Cowan. Hatched bars indicate units of receptor activity in pooled fractions after dialysis to water. while the collaSolubilization of the Receptors-The protocol used for sol. contained one protein (Mr 135.1% n-octyl @-D-glucopyranoAttempts to absorb the collagen receptor on a Sepharose side. These data constitutes the main component in the lysate which neutraloped. we attempted to isolate the receptors from bacteria which were notheat-treated. pool the inability of solubilized receptors to inhibit the binding of IIIa.e. This result is perhaps not surprising in view of was limited to the first and second peaks. weight of this material corresponds to that of the purified Ammonium Sulfate Precipitation-Electrophoretic analysis receptor. 8) (Table I). trypsin. In contrast. dialyzed to distilled water. mutanolysin. contained We have routinely used heat-treated bacteria as a source of components of apparent lower molecular weight in addition the collagen receptor. while receptor activity unsuccessful. and this step was therefore of Sephacryl S-200 equilibrated with 6 M guanidinium hydroincluded in the protocol. 7. column. The bound material was eluted with NaCl gradient (0-0. Pools Ia through IC were combined. Ion Exchange Chromatography-Practically all protein izes the inhibitory activity of antibodies and therefore precomponents in the bacterial lysate were adsorbed on a column sumably represents a collagen receptor.3 X 30 cm) of DEAE-Sephacel equilibrated with 50 mM Tris-HC1. i. 0 (0 m a VOLUME (ml) FIG. lune c) and complete loss of its activity in the antibody neutralization assay (data not shown). 6 material monitored by absorbance at column substituted with type I or I1 collagen or gelatin were 280 nm appeared as a series of peaks. The purification procedure resulted in a 179 times increase of specific method of collagen receptorpurification. Lysate (550 ml) of S. pool IIIb. The material remained virtually intact.activity (Table I). when used in Westernblot analysis (data not shown). The main portion resis (as in Fig. brane and thenprobed with anti-Cowan antibodies absorbed although some residual activity was still detected in fractions with cells of strain Newman. p H 7.000 appeared in addition to the main135-kDa band (Fig. 5). Since Holderbaum et al. Ion exchange chro. 9. 7) and electroblotted into Immobilon-P memof receptor activity was eluted with less than 0. lane pool IIIb under the same conditions resulted in a separation receptor for its ligand. As shown in Fig. chloride supplemented with 0.0 I-la-IllbllldCld-It-le-I 110 Furthermore. 2. aureus Cowan obtained after digestion of 50 g of bacterial cells dialyzed to the startingbuffer was passed a t a rate of 25 ml/h through a column (3. aureus lysate. V-8 protease. or thermolysin resulted in degradation of the receptor (Fig. W h 0 z m [r 1 a Preliminary Characterization of the Receptor Protein Digestion of the receptor material with chymotrypsin. clearly show that only one major protein band in the whole pools Ia. and lysate is recognized by the inhibitor antibodies. The second peak. A three-step procedure was devel.gen receptor activity remained insolution. lane g) revealed substantial amountsof high molecularweight material which migrated as smeary bands. The adsorbed material eluted from the column with different stages of purification were separated by electrophoa NaCl gradient as a series of peaks (Fig. 1987). aureus neutralizing half of the inhibitory activity of the antibodies has been arbitrarily defined as onereceptor unit. and IC. Collagen change did not resultin abinding of receptors to the collagen. Preliminary experiments showed that Gel Filtration-Further purification of the collagen receptor a brief sonication of bacteria (3 min) substantiallyimproves was achieved by gel filtration chromatography on a column the yield of collagen receptors. Moreover.2 M NaCl. this material was resistant to digestion by peptidoglycan degrading enzymes. The results (Fig.andalternative suggest that a 135-kDa protein methods had tobe sought. ensure that the protein bands in gel the are indeed transferred accompanied by high recovery (75%) of the active material onto the membrane (data not shown).Collagen Receptor of S. . the occurrence of this band is limited to the pools which contain receptor activity.. The first peak. and freezedried. In a parallel experimentelectroeluted with higher salt concentrations. or Streptomyces globisporus N-acetylmuramidase did not cause any significant loss of activity. (Froman et al.blotted material was directly stained with Amido Black to matography provided an excellent initial purification step.
020.800 2:17. In general. le.8 <5.5. IIIb. The data presented in the tahle refer to the amount of lysate ohtained from 50 g (wet weight) of hacteria. reducing conditions) of fractions at different purification steps.or antibodies have subsequentlybeenused in a series of experiments analogous to theseperformed with antiCowan antibodies. lanr h.115 t77 5. Id. la. Ih. k.2 4.9 39. freezedried. Also the pattern of immunostaining of Western blots with anti-receptor antibodies was analogous to that presented in Fig. activity of antireceptor antibodies did not substantially differ from the activity of anti-Cowanantibodies preabsorbedwith cells of collagen receptor-negative bacteria. The inhibitory pattern followed that observed with anti-Cowan antibodies (Fig.9523 43. lanr i.4 122 117.GFO 12.7 30. Polyacrylamide gel electrophoresis (SDS-PAGE.710 19. Characterization of Anti-receptor Antibodies The 135-kDa material has been injected into a rabbit to generate polyclonal antibodies. and their titerwas approximately 2 times higher compared with that of anti-Cowan antibodies (data not shown).4 93 . More than 9 0 5 of receptor activity was released from hacterial cells under the conditions employed.73 I . Material not precipitahle with 60".6.5. we have so far not been able to use the blocking monoclonal antibodies in Western blots.5 1 :3 36 " . Materials from the differentpurification steps were subjected to electrophoresis in 4.3 1.ysate la Ih Ic Id le la/h/c Iln 550 1 G" 62 94 124 :396 318 II h Illa 1111.857 1. IIId.5-1070 polvacrylamide gel..676 783 44 1 <I . aureus Cowan 1 binds "'I-collagen in a time-dependent reversible reaction. X12 67 44 37 120 28 1 >:100 52 >a00 22 c 8. aureus Cowan. Sephacryl S-200 chromatography.8 3..5 2. Studies on the specificity of collagen .4 100 44. Nurnb4r.7 7.2 649.209 3. - -21.4 9.8 <0.3 97.. pools from gel filtration chromatography:. combined pool of la.5.4 <1.6 4.600 :18.6 15. 0. 7.3 12.5 t0.4 131.570 1 8. Of severalclones recognizing purified receptor some interfered with lZ'I-collagen binding to bacteria (data not shown).5 <0. x a q-t OO 20 40 60 VOLUME ( m l ) 80 100 1 1 0 FIG. 450 35 20. The bacterial cells could be saturated with labeled collagen. proteins in the gel were stained with Coomassie Brilliant Blue R-250.0 19. and an analysis of the binding data by Scatchard plot suggested that a maximum of 3 x lo4 collagen type I1 molecules K d of the could be bound per cell and that the apparent interaction was 10" M. material not precipitahle with 60% ammonium sulfate.0 t. I 7 I a b c d e f g h i j k l m IZOO.9 69. d . Lane a.1 >400 15 22.3 5.5 17. Illa.0 A H * " I 16.0 1. fractions from DEAE-Sephacel chromatographv: 6. Receptors present in pools IIIh and IIIc were rechromatopraphed at the same conditions.4 "Determined according to Bradford (1976) using human IgG as a standard. untreated lysate of's.0303 91. 4) and was linear when plotted semilogarithmically.7 3 13.834 24.5 <0.6 9. lane g. Anti-receptor antibodies inhibited binding of ""I-collagen to bacteria.2 1084 Collagen Receptor of S.9 20.ypically fractions eluted in the first peak contained chromatographically pure material and most of the receptor artivitv.4 I FIG. and IC.12.2 0.5 160. ammonium sulfate was dialyzed to water.5 t1.286 615 4 3 3 800 33. In addition to polyclonal anti-receptor antibodies monoclonal antibodies havealso been raised. However. m. n-octvl . T.600 264 1 2 lIlh/c. Fab fragments of generated anti-recept.lfio 13. 14.242. 5 and 6.1'. rerun ~ 40 12 70 240 22 28 1. IC. e.q and arroum in the middle of the figure indicate molecular weights ( X lo-:') and migration distances of standard proteins.250 9. Purification Pool Yield activity Specific Activity volume One-unit Protein Protein" Volume pl X IO" pplml pplpool rl unitslpool unitslmg r: I.1 74. IIIc. materialprecipitated with 60% ammonium sulfate. and dissolved in a small volume of 2 M guanidinium chloride containing 0. I . 8 for staining with anti-Cowan antibodies.682 1. aurrus Cowan 1 This tahle summarizes the data presented in Figs. c.6W 1. The inhibitory effect of these antibodies was abolished by absorption with collagen receptor-positive but nor receptor-negative bacteria (data not shown).2 1. lanrsj-rn. lanes b-f. aureus TABLE I Purification of the collagrn recrptor from S . Illc llld _ " 14. Ib.320 6.2 179.5 21 1 200 600 120 590 24 V :i 247.7 :3.4- 66.6.260 :i. whichindicates that the receptor molecule has a domain structure where a certain domain is responsible for collagen binding.500 :i.o.A lysate of a collagen receptor-positive strain as well as the purified collagen receptor neutralized the inhibitoryeffect of the antireceptor antibodies. DISCUSSION We have previously shown that a strain of S.. f .j-r)-glucopyranoside and passed at the rate of 20 ml/hthrough a column of Sephacryl S-200 ( 2 X 150 cm) equilihrated with the same huffer.
however. Also eukaryotic cells have beenshown to bind types tested. namely streptococci (Kostrzynska mutanolvsin: lane c.- a b c 200. in:' L. 1987). Similar changes in (see Fig. Data presenting et al. 9 ) (50 pg of Fabfragments in 100 ml of solution). FracIt is also possible that the conformation of the solubilized tions from various purilication steps (as in the legend to Fig.of 135. 1987). receptor S. lane b. Consegested for 18 h with 5 pg of different enzymes. Although a direct binding of the 135-kDa protein tocollagen has notbeen demonstrated in this study it should be pointed out that this protein is FIG. Furthermore. Urushihara and Takeichi. 1989). and Escherichia coli (Ljungh and Wadstrom.. Thepreviously measured Kd (IO-' M ) would. quantification. 1978. if i t reflects multiple receptors interacting with one ligand molecule. 1988. A collagen receptor of the integrin type from fibrosarof '""I-labeled type I1 collagen to bacteria (Speziale et al. MA) and probed with a n t i 3 directly affect conformation of the collagen ligand binding aureus Cowan antibodies absorbed with cells of S. Speziale.397. bacteria of some other species were In addition to digestedwith themixture of lysozyme.0 absorb to an affinity matrix containing immobilized collagen. in the legend to Fig. 1988). In these cases. The affinity of such an inter. After inactivation of butabsenton the enzymes (89 "C. Nagy et al.ration.and reported to bind collagen. 9.. Purified receptor.. 1988). was dia collagen receptor-negative strain.0+ I 16. In fact. and isolation of the collagen receptor. The active antibodies also specifically recognize the 135-kDa protein when used inWestern blot assaydemonstrating that the epitopes recognized by the inhibitory antibodies are located on the135-kDa protein. isolated collagen a chains. 1988. 1987). protein from chondrocytes has been also isolated. Switalski.1988. 8. 1987. Sugrue..sequenced (Pilar Fernandez et al. Digestion of purified collagen receptor with mural. In light of these observations it is perhaps not surprising that solubilized collagen receptors do not inhibit the binding of '2511-collagen to bacterial cells or 200. Indirect methods had tobe employed. 1982. M. isolated. 8.bacteria. (Gonzilez et the digestions with other proteolytic enzymes are not included. manuscript in prepadicate that an individual receptor-ligand interaction may ex. cloned and Since the collagen receptor appears to be a major staphy. Of special interest in this context is a with several receptormolecules. 50 pg.collagen receptor present on hepatocytes which binds collagen action is an exponential function of the number of interac..:' FIG. The the antigenic epitopes in fibrinogen molecule following its memhrane was incubated with horseradish peroxidase-conjuaated enzymatic digestion were reported by others (Cierniewski and goat anti-rabbitantibodies(Rio-Rad) followed by theappropriate substrate. andprocessed as described coccal collagen receptor. Another consequence of the apparent low affinity that the solubilized staphylococcal collagen receptor has for its ligand is that direct binding methods are essentially useless in the identification. or charnot recognize noncollagenous proteins but all the collagen acterized. and some of the receptors involved were characterfragments.4- . 1987. aureus Newman site and hence theaffinity for the ligand. Salmonella sp. Hoiik. Westerlund et al. preliminary experiments have shown that polystyrenebeadscoatedwith isolated collagen receptors bind '""I-collagen in a reaction that 14.present on the surface of a collagen receptor-positive strain ytic and proteolytic enzymes. aureus. h i j k . Collagen binding lococcal cell surface protein it is possible that the density of proteins have been identified or isolated also from other cells the receptor on the bacterial cell surface is high enough to (Santoro et al..000. al. a u r e u l m 21085 hibit a fairly modest affinity. 20 min) in the undigested and digested materials quently we propose that the 135kDa protein is a staphylowere separated by electrophoresis. 1989.4 resembles that between staphylococcal cells and collagen. receptor digested with trypsin. Wayner allow a collagen ligand molecule to simultaneously interact and Carter. 7) after receptordiffers from that of areceptor on the surface of separation in the polyacrvlamide gel were electroblotted into Immo. These data suggested that the staphylococcal receptor type I collagen (Dedhar et al. untreated receptor. Detection of putative collagenwith antibodies. and syntheticcollagen analogs inhibit the binding ized. could we design experiments where several receptors were simultaneously available to bind one collagen molecule it should be possible to demonstrate an interaction with reasonable affinity. Ocklind 00 and Obrink.. aureus cells demonstrated that the receptor does sponding receptorshave not been identified. the correbinding toS. and M.Hedford.with a specificity similar to that of the staphylococcal receptor tions.. In this study we adopted an immunologicalassay. The affinity of an individual receptor molecule may simply be too low to allow us to detect an interaction with these methods. blotted. P. 1985). 1980). N-acetylmuramidase. coma cells appears to recognize an Arg-Gly-Asp sequence in 1986).Receptor Collagen a b c d e f g of S. Mamo et al. Lune a. A collagen binding recognizes a fairly simple structure in the collagen molecule... Velge et al. collagen collagen. which previously has been used to identify cell surface molecules involved in eukaryotic cell-cell adhesion (Muller andGerisch. This protein has been used to generatepolyclonal as well as monoclonal antibodies which inhibit the binding of collagen to bacteria. In this assay solubilized proteins without apparent biological activity are used to neutralize the inhibitory effect of the crude antibodies. Changes in the tertiary structure of a protein may bilon-P membrane (Millipore. With thehelp of this assay we have isolated a putative collagen receptor protein of an apparent M. Consequently. Budzynski.
G. T. 77-81 Cierniewski.21086 Collagen Receptor of S. T. (1987) J . F. (1963) Allergy Appl. 1987) are suggestive of the relationship between the ability of staphylococci to colonize Muller. P. Obrink. 16. experiments been shown to mediate the attachment of bac571-575. Vartio... M. K. and Ehrhart. receptor reported here is unclear. and Capron. Gristina. Immun. Raucci.. Raucci. D. Ryden.. ed) pp. and Korhonen. 463REFERENCES 470 Blobel... Druguet. Jacob. 150. T. Infect. W.. S. B. and Lindberg. C. Immun. Afchain. M. Rauvala. G. A.. 153.. (1985) . Dis.. (1984) J. P. 1 5 8 . T. 262. (1988) Biomaterials 9. M. A. Vaudaux. A. D. (1984) J. The relationship between these Kuusela. and Brentani. 77-81 Liunph. G.. Clin. E. B. V.. Immunol. Peterson. Wadstrom. T. Hyg... H. (1983) Infect. Herbage.. Biochem. ... D. Sci.. D. K. S. Switalski. Cornette. A. M. Switalski. Holderbaum..Pathol.. 229-234 isolated and characterized. 6 7 .. S.Lussenhop. L. and Obrink. and Mvhre.. Lett. P. G. E. aureus collagen Kuusela. 1407-1413 Herrmann. S. B. 6. (1981) Cell 24. F. Peterson. C.. W.. M. H.. O. A. 76. Inst. M. Q. (1985)Am. Bacteriol. Speziale. T. Rozgonyi. T.628-633 Biol. 2258-2261 Wayner. 5443-5448 Rubin. Bakteriol. Chem. Emonard. Ljungh. T. 289the leadingcause of osteomyelitis and infectious arthritis.. 835-851 Santoro. and Wadstrom. Zentralbl. M. P. Acad. Immun. (1978) Nature 276. 3. G. R. and Lindberg. ~~ ~ ~ . Signas. (1987) J . .. I - ... M. H. P. 13-21 D.. Guss.. A.. Schalen. (1987) J .. 275-277 and bone. Westerlund. T.. M. J. J.. Lab. B. (1982) J . H. Natl.. H. S. Biol. (1989) Mol. H. Jacob... Microbiol... G. P. and Hall. C. and Moldow.. Myrvik. and Wadstrom. G.. and Woods. 34-43 ' and Doolittle. M. Mollenhauer. G. M. Mikrobiol. and Wadstrom. Schumacher-Perdreau. Ouaissi. K. F. aureus have in preliminary . M.. Barth. and 54.. 107-110 Holderbaum. D. Virkola. R. (1987) J . Staatz. Lew. E. M. (1975) J . Froman. (1988) Inflammatorv Bowel Disease: Current Status and Future Approach(MacDermdtt. Jonsson.. Abt. (1979) Proc. M. and Speziale. D. M. Infect. Lindholm... and Wadstrom... Lindgren. P. and Scheraga. T. C.1873-1884 Hunter.. and Froman. A. Pfaffle. Kuusela. 305 where bacteria colonize collagen-rich tissues such as cartilage Lopes. T. Blackwell Scientific Publications.882-887 We are currently examining this hypothesis. Cell Biol. Timmons. (1978) Nature 274. 86. Natl. Biophys... Wadstrom.546-553 Biol.. Immuh. and Hook. R.585-593 Towbin. (1982) Biochemistry 21. Keys. G... Mikrobiol. and Sjoquist. and Carter... Chem.. G. R.. U. A.. Visai. S. (1985) Biochemistry collagen-rich tissues and the expression of collagen receptors. Hyg. F. L. Clin. 1981). 262. J. I. Fardel. and Dobberstein. P. 105.'lO3. G.. represents a mechanism of bacterial adherence to these tisPathog.. Selmin. K.. Anatharamaiah. Y. 1878-1883 Hook.. G. P.445-449 Nagy. and Pierschbacher. B. Elsevier Science Publishers B. Sundis. P... 417-424 sues. and von der Mark. J. 4350-4354 B. Ocklind. 248-254 (1988) Biochem. Waldvogel. R.359-364 Ryden. C. K. 120. Speziale. (1978) Handbook of Experimental Immunology (Weir. Bradford. (1976)Anal. Z. S. M.. M. D.. P.. 437-446 McCarthv. A. 693-701 M. and Gerisch. Riley. A. 718-720 identified collagen binding proteins and S.217-223 Carret. Rubin. A. 241-245 S.. Chem. G. G. R. P. (1980) Cell 20. Staehlin. A.. 42... Suzuki. Pasteur (Paris) 136A. Meinwald.. 1 Suppl. J. Timpl. 2. A. i:T. Bakteriol. However. (1986) Infect.. J. Biol. The collagen receptors of S. 163-176 and others (Holderbaum et al. G. Morgenthaler.. M.. P.. and Signas.. S.. B.. K. Spech. (1987) Infect. G.. A. Risteli. Froman. F.6564-6571 Velge. 72. J . F. G.. E. G..M. L. R. (1989) Proc.. Sci.. T. Microbiol. Raucci. 50. Biol.. 104. W. Pittet. 24. C. P. K.. S. I. Preliminary data from our group (Voytek et al.. D. Microbiol. A. G. McCarthv. M. Hawiger. and Gordon. R. Infect.. K. A..699-703 Chhatwal.. Speziale. B.363-371 E M B O J. Hook.. Vercellotti. A 269. Ehrhart.. (1986) J. (1987) Microb... Strong. Hall. Flock. Nydegger.. T. this receptor has not yet been Kostrzynska. P. B.. (1988) J .. 59. T.. (1988) (1988) Parasitology 97. C. T. DosReis. Jonsson. and Timpl.. (1985) Science 2 2 9 . B. C. M. U.. 167. and Mayner. P... and Blobel. U.. 255-268 Vercellotti. M. A. Preissner. Commun. P. S. Dis. A. 6788-6795 ... Risteli.. E. it is possible that the binding of bacteria tocollagen Mamo. M. D. A. H.L. " Fernandez. Rajpara.. Muller-Berghaus. P. Peters. E. Yamada. D. Maxe. Cell Biol. 2 5 7 .. L. and Furcht. Cottrell.. (1987) Urushihara. Hook.. Vartio. 5921-5925 Reese. L. (1988) J.. C. T. 2351-2357 Vaudaux. J. Blanco.. and Budzynski... (1985) Ann.. R. Res. Amsterdam teria to collagen-coated surface^. A. Vuento. A.. (1981) Biochemistry 20. S. E.. J. L. J. J.. 14. Hook. L. G. Ljungh..1-14. P. Voytek. J. M. Cell Biol. Deutzmann. M. G... K. A. aureus (Rubin et al. and Waldvogel. W. Sugrue.. (1988) Vet. M. S. 25. 18. Gonzalez. L.. Ltd. R. 5 5 .. R. and Wadstrom. K. (1987) J.. M. G. Flandrois.. B. Hook. Furcht.83-91 (1987) J. Froman.. M. S.329-337 Oxford Acknowledgment-We would like to thank Wade Butcher for his excellent technical assistance... T... Cell Switalski.. D.3338-334Q 13896-13901 Dedhar.. J.^ Since S. aureus strains are Lofkvist. Switalski. and Takeichi. and Wadstrom. 262. R. 1988) Mamo.. Mei. (1987) Zentralbl. Baloda. Ser..... W. 2 3 . J. ed) pp. M. (1989) F E M S M i crobiol. Acad. J. .. and H. Y . Martin. M. Jr. V.. J. Chem. C..... L. Ruoslahti. Auckenthaler. Nilsson. Hook. M.40. 263.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue reading from where you left off, or restart the preview.