CA2192175A1 - Surface protein of staphylococcus aureus - Google Patents

Abstract

Methods and compositions are provided for preventing Staphylococents infections associated with the use of catheters and similar devices. They are based on the neutralization of ability of surface proteins of staphylococcis to adhere to the surface of the catheters at the skin-catheter junction as the first step in the systemic invasion of the host by the organism.

Description

Wo 9513~65a Pcrlus9alo7loo ~ '-- Jp.' SURFACE PROTEIN OF STAPHYLOCOCCUS AUREUS
FI~LD OF TH~ lNV~ L~
This invention relates to methods and compositions for preventing ini~ections by Staphylrcocc~1c aureus ~S. aureus), iAl ly those triggered by the use of catheters prosthetic devices and heart valve rep~ ore specirically, it relates to methods and compositions useful ror inhibiting the ability of a surface protein of S. aureus to i, l- t a&esion of the organism to endoth~ ic~l cells or to catheters at the skin-catherter junction thereby initiating infection.
BA~ KUUNL~ OF THE lr~ v~;h L10~
S. aureus is one of the most ~requently encountered pathogens in infections acquired in the hospital. It accounts for 25% of all hospital acquired infections resulting from the use of catheters or similar :~LLU-LUL~:5. Since a great majority of patients entering a hospital re~uire some sort of intravenous device, there is a very high probability of infection. The same type of risk applies to the use of prosthetic devices such a as hip and other joint rPrl~r ~ because the ability of staphylococci to a&ere to such devices.
.

One of the initial reactions of the r~ n host to the presences of a catheter is to coat the object with fibrinogen and other matrix proteins as a prelude to the systemic reaction which is intended to expel the invasion. In a hospital setting, this provides an U~U~U~Lity of infection by s. aureus which attaches 2 1 9 2 1 ~ 5 ~ oo wo s~l3~6s~
itself to the fibrinogen at the s3cin-catheter injunction and thereafter worXs its ~ay through the skin and into the blood.
Since the strains prevalent in hospitals, nursing hcmes and other patient care facilitieS are o~ten antibiotic resistant, these types of infections are ~L-~ -ly serious and very ~ iclllt to contain. Accordingly, the art has t~Ypt~nt~Pd much effort to prevent such uyyU~ L-l..istic infections.
BRIEF S~RY ûF T~IE INVENTION
A surface protein, and the gene which expresses it, have now been discuv,~ d. This protein enables the invading bacteria to Adhe're to the f ibrinogen . Antibodies to f ibrinogen or to the surface protein will prevent bacterial ~hP~i~n and thereby inhibit infection. The protein and segments of the protein are useful as vaccines or for the pro~ n of antibodies useful for passive protection of patients prior to the use of a catheter or equivalent device.
This inYention, therefore, comprises the protein itself and segments thereof, the gene and segments thereof which produce such products, vectors for the gene and its useful segments, organisms transformed by such vectors, monoclcnal and polyalonal antibodies to the protein and its useful s~ Ls, vaccines produced utilizing the protein and its segments and methods of preventing s. aureus infections utilizing such products. The invention also includes diagnostic probes utilizing the gene products described herein.

2 ~ ~ 2 i 7 ~ L~ ~ ~ ~ Ioo o 9~l3~6~
1~} FIGURES
There follows a brief descripticn of the figures.
Fig. 1 Western blot of cellular i~ractions of clone number 14 probes with fibrinogen l~ollowed by anti-fibrinogen antibody cu..j, ~y~t~. The control contA i nC the lysate of an E. coli clone with pBR322 insert in ~ Zap. The arrows indicate a 34 kl~
reactive band together with an upper band which may be a dimer.
Results of a duplicate blot probed with 125I fibrinogen were similar.
Fig. 2 Nestern blot (A~ and silver-stained gel (B) of the pe_iplasmic extract of clone number 14 fractionated in a fibrinogen column. The Western blot was probed with fibrinogen/anti-fibrinogen conjugate. The short arrow indicates the fibrinogen-reactive band from crude lysate positive control.
The long arrow marks a 34 kD protein that reacts with fibrinogen t4th lane in (A) ] and is eluded with 3 M potassium thiocyanate, but not with PBS with o . s M ~2Cl nor with acid elusion.
Fig. 3 Is the complete sequence of clone 36 (C36~ ~n'-n~;n':T
fibrinogen reactive protein.
Fig. 4 Se~uence comparison of strain DB to that of coagulases from strains 8325-4, 213 and BB with the Pileup prosram under the GCG package. The arrows indicate the 11 amino acid sequence that is unique to the protein. This sequence shares homology with a cell wall anchor motiJ found in other gram WO 9al316aa 21~ 21~ S ~ c ]
pcsitive Yall protein Eowever, there i5 no ~oto identify to t~is mctif i~ the Genban3c. The arrows at rosiues 4~9 and 419 indicate a unique secuenc2 segrlent of this protein.
The aminc half (residues 59-325) c~ this protein is primarily helical as predictsd by the Gar~ier analysis. Two areas (residues 58-194 and 264-2g7) reveal a 7 residue periodicity in which residues in positicn 'a' and 'd' in a heptad mcti1' 'abcdefg' are either ~y~ u~hobic cr ncnpolar. This pattern is consistent with a coiled-coil c~n1'n~-tional ~ L~. The sa~ is analyzed by the Matcher ProSrall ~22).
Fig. 5 Se~uence comparison of cur protein (strain DB) to that o~ coagulases from strains 8325-4, 213 and BB with the Pileup program under the GCC package. The arrows indicate the 11 amino acid sequence that is unicue to cur protein. ~his se~uence shares hcmclogy with a cell wali anchor mctif found in cther gram pcsitive cell wall prctein. Xo~ever, there is no ~ let~
identi~y to this mctif in the Genbank.
The follcwin5 abbreviations are employed in the description cf this invention:
Strain DB - a ~ild strain of S. aureus N2Y broth - a co~mercially available gro~th medium I i3 - a commercially available growth medium IPTG - Isopropyl-beta-D-th i o~ tu~yLunoside BS~ - bcvine serum albumin ~ 2 1 7 ~ PCrlUssslo~loo ~ woss/3463~ ~1 .
X-gal - 5 bL 4-chlero-3-indclyl-beta-D-~lA~ ~ vl~yL~:ncside 525 - sodium dodecyl sulfate NP40 - a com~ercially available nvn i r~n; C deterge~:t pBR322 - a conmercially available pla5mid ol' known ~rLLl ~_-UL'~
rRInos~ipt - a commercially available ~hArJamiri of known '--~LU I_ULC
SSPE _ 0.l5~aCl, 10~ ~a~2P4' Im ~EDTA pH7 . 4 'rSE buffer - 0.1~ ~ris, 20% sucrose, 5~ EDTA p}I8 PBS - Fhosphate buffered saline Strain DB has been deposited at the A~erican Type Colture Collection under the A~cracsir~n nu~ber The ~ollowing Materials and ~ethods section is provided for convience and ease o~ understanding of this invention.
M~ ?T~r C ANI? MET~ICDS
Bacteria, plasmids and pha~e Strain DB which has been phenotypically characterized (5 was used in the construction of a geno~ic library. E. coli strain Sure (stratagene) was the host cell for the~~ Zap vector.
Clones 14 and 3 6 were phagemids derived ~rom ~ibrinogen-reactiVe ~ ~ n~ r pcrlus9slo7loo wo ss/346s i ~ J
.
pla~ues containing the insert. A pBluescript rh~g~m;~, pAC8, whic'~ cont2ined a protsin A clone was derived from a Zap gencmic library of DB.
Media and Antibiotics Unless otherwise indicated, the following media wew used:
NZY broth and LB (23) were used for the growth of E. coli strains .
Prer~arations of affinity purified qoat anti-fibrinoqen conjusate Goat 2nti-human fibrinogen antibody obtalned commercially ~Cappel, West Chester PA) was a~finity-purified on a fibrinogen column as previously descri~ed (7~. The fibrinogen column was LlL 2~C~ ed with glutardialehyde activ2ted beads (Boehringer M~nnhoim, Tn~;anaroliS, IN) as déscribed in the manufacturer's insert (4). The monospecificity of the affinity purified goat anti-human fibrinogen IgG was verified by an irlmunoblot using purified fibrinogen and plasma as antigens (7). The protein concentration of the affinity purified antibody was detarm;np~ by the BCA protein assay reagent (Pierce ChF~m~ c~ Rocke~ord, IL).
Affinity purified goat anti-fibrinogen antibody was conjugated to bovine intestinal alkal;no phosphates (Sigma, St. Lousis, MO) as described by Vo ller ( 3 2 ) .
Construction and screeninq of staphylococcal qeno_ic library A genomic library of strain DB was constructed with the Zap vector that has been digested with EcoRI and dephosphorylated (Stratagene cloning kit). DB ~ LI _ ~1 DNA was extracted from WO 9513'~6~5 f~
l~a~a~phin-lysed cells as pr2viously described (5,29). Genomic DNA was sheared w ith a 2 6 gauge syringe and sub j ect to gel filtration on Sepharose C~23 to re ove fL~ q s~aller than Lkb. Fractions c~nt:~;n;ng 4~5 kb fL, Ls were po~led, treated with T4 polymerase to produce blunt ends, methylate with EcoRI
~:ethylase (New Eng7and Biolabs, Boston, MA), insert into the EcoRI site of the Zap vector with EcoRI linkers and packaged in vitro with ~:i g~raC-k packing extracts (statagene) . Over 90% of the r~ ~;n~nt phages were r~-or;7F~ as white plaques when plated on lac host strain Sure in the presence of IPTG and X-gal.
For the screening of f ibrinogen reactive plaques, rF 'in~nt phage was incubated with the E. coli host on NZY agar (23) at 42C for gene expression. Following transfer to duplicate nitrocellulose filters (82 h7m in diameter, Schleicher &
Schuell, Keene, N}~), the filters were blocked with lOml of TNT
buffer (10 m~ Tris with 0.15 M NaCl and 0.05% Tween 20) containing 1% BSA for 1 hour at RT and then incubated at 37C for 3h with 25 ug of fibrinogen (Sigh7a ~4883). This fibrinogen preparation had been further purified over a protein A sepharose column to re~ove contaminated IgG and was found to be easentially free of contaminants as determined by a silver-stained SDS-gel containing this protein . The f ilters were then washed once with TNT containing O .1% BSA and O . l~c NP40, and twice with TNT with 0.1~ BSA for 5 min. each. Affinity purified goat anti-fibrinogen antibody ~lk;.lin~ phosphatase conjugate diluted 1:1000 in TNT
buffer with 1% BSA was then incubated with the filter for 1 hour at RT. After washing the filters twice with TNT with 0.1% BSA
and O.1% NP40 and three times with TNT with O.1% BSA for 5 min each, fibrinogen-reactive plaques were visualized with 5-bromo-r~,~l1.J,.,.~l~CO
,~ Wo 9sl3~6~s 2 1 9 2 1 7 5 4-cloro-3-indolyl phosphate as a substrate as described by Blake et al (1). A~Zap vector w~th a pB~322 insert plated on E.coli or E. coli cells alone served as negative controls. Positive clones detected by fibronogen/anti-~ibrinogen conjugate were cnrtf i ~ by allowing plaques on duplicate filters to react with 125I fibrinogen. This screeni~g p~ du~ was similar to the method described above except that 125 fibrinogen (lO0,000 cpm) was used in place of the cold llnl:~h l~d fibrinogen. Following reaction, the nitroc~lllllose filters were washed twice with TNT
with 0.1% BSA and 0. l~c NP40 and three times with TNT with ~.1%
BSA for 5 min each, and finally subjected to autoradiography.
Purif ied plaques were isolated by rescreening each positive clone at least 4 times.
DNA seauencinq of the f ibrinoqen reactive clones By infecting the E. coli host strain simultaneously with a fl helper phage tR408~ and the~ Zap vector containing the insert, a single strand DNA containing the pBluescript phagemid with the insert can be packaged for recirculation, thereby generating subclones in E. coli cells when plated in LB/ampicillin agar (Statagene cloning kit instructions). Plasmids for sequencing -were purified from E. coli by equilibrium centrifugation in Cesium chloride-ethidium bromide gradients (23). The purity of the plasmid was confirmed by digestion with restriction enzymes (New England BioLab, Beverly, MA). By using both T3 and T7 primers flanking the insert, plasmid sequencing of clones 14 and 36 were perfor~ed with the Sequenase Xit (U.S. Biochemicals, Cleveland, o~) following the mznufacture's instruction (27).
Additional primers were obtained for sequencing from within the insert .

~ 1 n ~ oo Wogs/3~6~s f. 1~ J
Southern blot hybridization Southern blot hybr;~7~ n ~as perSormed with random primed samples of gel-purified DNA fragments as probes ~12,23).
Briefly, ~ l~z, -- 1 DNA digested with restriction enzymes was resolved on 0.7% TBE gel and transferred onto }~ybond-N membrane Anersham, Arlington }~eights, IL) (14). DNA probes were labeled with 32p ( _32p deoxycytidine tr;rh~ k~e, AD~ersham) using the rando2 pri2ed DNA lAh~lin~ kits (Boehringer M~nnhl~;m). The membrane was then hybridized uith the 32p labelea DNA probe at 65C overnight, washed twice with 2X SSPE with 0.1% SDS at RT for 10 min each followed by lX SSPE with 0.1% SDS at 65C for 15 min.
The membrane was then subject to autoradiography with an intensifying screen at -70C.
Ex~ression of fibrinoqen-reactive protein of S. aureus in E. coli one of the fibinogen-binding clones, clone 14 was evaluated for the expression of the fibrinogen binding protein in E. coli.
E. coli cells containing this clone was grown in 10ml of LB with 50 ug/21 of ampicillin at 37C until the OD 600 reached 1. 0 .
Cells were collected by cetrifugation at 7,000 g for lO min and resll~p.on~d in 1.25 ml of ice cold TSE buffer (100 _M Tris, pE~
8.0, containing 20% sucrose and 52M EDTA). Lysozy2e was added to a final ~ onc~ ction of 0.5 mg/ml and the sample was iced for 20 min. For whole cell lysate, 0 . 25 ml of this suspension was removed and 7 . 5 ul of Triton X-100 and 50 ul of DNase solution (lO 2M MgCl2 with 100 uglml of DNase) were added. The sa2ple was frozen (-70C) and thawed twice.
_g_ WO 95134655 2 1 9 2 1 7 ~ , IIU:~YSIU I 100 Magnesiu~ chloride was added to the remaining 1 r. l . cell suspension (50 m~ final Cu~ LLation) to 51 ~hi 1 i ~o the spheroplasts which were then pelted at 7,000 g for 15 min. The supernatant was filtered through a 0.45 um M;llirore membrane to obtain the periplasmic frâction.
To lyse the spheroplasts, 0 . 25 ml of DNase solution wzs added to the pellet along wit_ 0 . 75 of water. The spheroplasts were aspirated vigorously several times with a pasteur pipette, frozen and thawed twice as described above. The lysate generated by this treatment was centrifuged for 49, 000 g for 1 hr. The supernatant filtered through a 0.4 um membrane was designated cytoplasmic fraction.
The transfor~ed E. coli has been deposited at the American Type Culture Collection under the accession number SDS-PAGE and ir--ln~lhlt~t analysis Cellular extracts (lO ul each) were separated on 9% SDS-polyacryla~ide gel slabs by the method of Laemmli (21).
Prestained molecl~lar standards (BRL, Gaithersburg, ~D) were run o ù~l~ urr~ntly in adjacent wells. After ele~_LLu,uholesis, the gel was either stained with silver (Pierce `hPric~l~) or transferred onto nitro~ llos2 (30). After transfer, the nitrocPlllllose filters were allowed to react with fihrinogen followed by affinity purified goat anti-fibrinogen ;~lk~l irP phosphatase conjugate and reactive suhstrate as described for the screening of the genomic library . In some experiments, the f ilters were incubated with 125I fibrinogen (250,000 cpm~, washed and autoradiographed as descrihe~ above.

Wo 9s/3~6~ 2 1 9 2 1 ~
To detect other proteins in the cell extracts, the SpPri1';r antibcdy diluted in blorl-;n~ bu~er (Tris lOmN with O.S M NaC1 and 0.05~6 Tween 20, p~ 8.2) was incubated with the blot for 2 hours at RT. This was followed by incubation with an d"u~uruuLiate al~:~l;no phosphate Cu~Ju~:t2 for 1 hour and then processed for band v;c~ll;7ation as previously described (3).
Partial purif ication o~ the f ibrinoqen-reactive protein of S .
aureus In an attempt to pu~ify the fibrinogen-reactive protein from clone 14, peripl~-;r extracts fro~ 4 L of culture were ~Lc:~dl~d as previously described. Briefly, cells were harvested by centrifugation (7,000 g for 20 ~in) and rpc~lcpondp~ in 7S ml of TSE buffer containing 37.5 mg of lysozyme. After incubation on ice for 20 min, MgC12 was added to a final Cu.~cer ~Lc-tion of 50 mM
and spheroplasts segmented at 7,000 g for 30 min. The periplasmic fraction was aspirated from the supernatant, filtered through a 0 . 45 um membrane and immediately applied to a fibrinogen column (2.5 x 20 cm) followed by rotation at 4C
overnight. The fibrinogen column was prepared by mixing 25 mg of fibrinogen and 5 gm of glutardialdehyde beads as described (4).
After collecting the fall through, the column was washed with lS0 ml. of PBS followed by 150 ml of PBS with 0.5 NaCl. The f ibrinogen-binding protein was then eluded by rotating the column with lo ml of 3 M potassium thiocyanate at RT for 20 Min followed by collection. In prel;m;n~ry studies, a similar elusion . uce.luL e with o . ~ glycine pE~ 3 . o was not ~ rul. Fractions from the column were concentrated in a Centricon 10 (A~icon, Danvers, MA) and analyzed by SDS-PAGE and immunoblots with ~ibrinogen as described.

~ Wo 95/3~6~ 21~ 2 1 7 ~ JaYalul~ûû
Computer analysis of seS~uenc~ data DNA protein sPT~Pnre analysis, and se~uence compariscn with database were rnn~llr od with the Sequence Analysis Software Package from the GPn~f i ~-q Computer Group (IJniversity of w;ccnn~:in, Nadison, WI) (9). The deduced amino acid se~uence of the putative protein was ~ I:d to a se~uence database by the algorithm of Pearson and Lipman (TFASTA ;mrl~ tation of GenBank) (25). The fibrinogen reactive protein SP~onre shown in the figures has been ~c~isnpd to clone 36.
RESULTS
Isolation of f ibrinoqen reactive clones Using the foregoing ~Luce~uL~s~ a~ Zap library of strain DB, was screened for clones that were reactive with fibrinogen. Of 100,000 plaques screened, three novel clones, designated 14, 30 and 36, were found to be highly reactive with both l25I
fibrinogen and fibrinogen/antifibrinogen conjugate on immunoblots. S~hr~ onPc contalning the pBluescript phagemid together with the insert were subse~uently generated in E. coli strain Sure. Plasmid DNA from alk~l ;nP lysis minipreps of clones 14, 30 and 36, upon digestion with Eco~ which released the inserts, revealed DNA LL_, 5 of 4.6, 3.6 and 3.2kb, respectively. Using the 4 . 6, 3 . 6 and 3 . 2 kb fragments as separate probes, Southern blot analysis of Eco RI digests of these clones established that they hybridized each other. These clones did not hybridize with the EcoRI fragments of pAC8, a protein A probe of DB, thus eliminating the possibility of a false positlve reaction between expressed protein A gene product 1 9 2 1 7 ~ PCr~S95107100 WO 9513'~655 2 and gcat anti-f ibrincgen anti~ody c~...; ,.~te during the screening pL ~c- duLt:. Further analysis indicated that clone 14 comprises about 2/3 of the mature molecule, C 36, oYtPn~;ng into the C-f~n;mlq .
Expression studies of the fibrinoqen reactive protein of S.aureus Based on restriction analysis, clones 14 and 3 0 were Oimilar. Although clone 36 contained the co~plete gene as ~et orm;nP-l by SPq-lon--e analysis, expression of the fibrinogen reactive protein with this clone was found to be difficult.
Notably, a culture of clone 14 when grown to late stationary phase (i.e. OD600~m 1.5) also resulted in a significantly decreased yield in f ibrinogen-reactive protein . This result can be explained either by toxicity of this protein on E. coli or by increased proteolytic breakdown during stationary phase. For these reasons, the expression of the partial protein was evaluated in clone 14 that has been grown to early stationary phase (OD6001"mm-1. o) .
Expression studies of dif~erent fractions from clone 14 with Western Blots probed with either 125I fibrinogen or fibrinogen/anti-fibrinogen conjugate es~hl; ~hod that the protein, which has a molecular si2e of 34 kD was found in the whole cell, peri~l~C~;~ and membrane fraction (Fig 1). In contrast, a crude lysate of an E. coli clone which contained a llloe:--ript phagemid with a pR~322 insert did not react with ~ibrinogen (Fig. l) . With some fractions (e . g . membrane), there was also a higher molecular weight band, possibly a dimer, which reacted with fibrinogen. ~leither of these proteins were fusions wo s~/3~6~s 2 1 ~ 2 1 7 ~ P~
proteins as they were not inr;lt~ihl.~ with IPTG. Additionally, these bands did not react with pclyclonal and ~ 1 anti-beta-galac~si~l~ce antibody ~1:1000 ~ tion) (Boehringer Mi~nnh.~;~) cn i ~hlnts. The fibrincgen reactive band alsc did nct react with a~inity purified chicken anti-prctein A antibcdy (Accurate C~ c , Westbury , NY), thus prcviding additional evidence that prctein A was nct cloned which, by binding tc ~gG, could lead to ~alse positive results.
To conf irm the binding specif icity cf this protein to fibrinogen, periplasmic extracts which contalned fewer ccntaminating bands were harvested frcm 4 L cf E. coli cells expressing the protein of clone 14 and applied to an affinity column with f ibrinogen linked beads . The cloned proteins of interest, as analyzed by silver stain and Western blots with 125I
fibrinogen and fibrinogen/anti-fibrinogen conjugate, was found in precolumn fractions and the 3M potassium thiocyanate eluant.
However, they were not found in the fall-through, P~S with 0.5 N
NaCl eluant, nor in the acid eluant (glycine pH 3.0) Fig. 2).
Although the protein was not purif ied to homogeneity in this one step pLO~,~dL~ (Fig. 2), these results clearly indicate the binding specificity of this protein to fibrinogen.
Secuence analysis of the fibrinogen reactive protein The complete sequence of the f ibrinogen protein expressed by clone 3 o is shown in Fig . 3 . The sequence revealed an open reading frame of 1,935 nucleotides. The seguence has a guanosine-cytosine (GC) content of 34.7%, in contrast to the 30%
GC content in the staphyl~cq~ genome (10). The higher GC
content is attributable to the carboxy terminal half of the W0 951346C~ F~ 00 --ler~1P (39.7%). Putative transcription and translation signals and rih^sr--l binding sites are indicated in Fig. 3. The first 26 a;aino acids have features characteristic of a bacterial signal peptide (16). Based on the predicted cleavage site, the mature protein has a predicated size of 69,9gl Da with a deduced pI of 6.5 .
.

Analysis of the deduced amino acid sequence revealed three distinct domains in this protein. With the exception of residues 27-58, the N-tpr~n;n~l half (residues 58-325) of the protein is primarily helical as predicted by the Garnier analysis (15).
Two areas (residues 58-lg4 and 264-294) within the helical portion of the molecule reveal a 7 residue periodicity in which residue in positions 'a' and d' in a heptad motif 'abcdefg' are either ~ly~,~hobic or nonpolar (Fig. 4). This finding is suggestive of a stable coiled-coil conformational structure in these areas (14.22). The second domain between residues 326 and 505 denotes a proline and glycine rich region (20~). Of the 180 residues present, there are 17 proline and 19 glycine residues.
This contrasts with the N-~Pnm;nAl portion of the molecule in which only 3~6 of the residues are either proline or glycine while the r^-~;n;nr carboxyl portion reveals a composition of 14~
proline/glycine residues. The carboxyl-terminal domain (residues 506-645) consists of S tandem, direct repeats of 27 amino acids each followed by 5 tP~m;nAl amino acids (PRVTK). Div~ ce is observed mainly in the outside repeats. Conformational analysis indicated that this repeat region is nrnllP1 ir~1 and contains mostly beta-sheets. Comparison of the protein sequence with others in the Gen~ank dataoase revealed significant homology to three p~ l; ChP~ 5 . aureus coagulases from 5 . aureus strains 8325-4, BB and 213 (18,19,26). With the exception of residue 7 W0 9513~6~a 2 1 9 2 1 7 ~ P~ on .
in the primary translaticn product of the fibrinogen reactive prctein, the N-tP7~inA~ 33 amino acid residue which include the leader peptides among all four se~uencss were i~pnt;rAl and therefore are likely to possess j~Pntir~1 sequence cleavage sites (see Fig. 3). Comparing residues 1 to 422 in the fibrinogen reactiYe protein to coA~llA~P~ in strains 8325-4, BB and 2~3, there are 56.2%, 73.2% and 56.2% identity, respectively. The identity between residues 423 and 645 comprising the five repeated units to homologous regions in the coagulase sPrll~PnrDc increased markedly to 93.9~. 95.1% and 96.9% for strains 8325-4, BB and 213, respectively (Fig. 5). LiXe that of fibrinogen-reactive protein, the C-termini of coagulase sequences of strains 8325-4, BB and 213 are -s~d of repeating units of 27 homologous, but no irlPn~icAl, amino acids followed by the tPrm~nAl sequence PRVT~ (Fig. 6). ~Iowever, the number of repeating units dif f er among strains . Although the f ibrinogen reactive protein sequence displayed features that are common to the coagulase sequence, a careful comparison revealed a unique stretch of 11 amino acids between residues 409 and 419 (S~ITLPSITGES) in the middle of the proline/glycine rich region (Fig. 5). Of interest is the fact that the motif LPSITGES shares homology with a cell ~all anchor motif ~r PXTGX) found in other gram positive surface protein (12,28). ~owever, there is no complete identity to this heptad motif among se~uences in the Genbank database.
Based on all of the foregoing, it is clear that a novel f ibrinogen reactive protein of S . aureus has been cloned. This protein is both structurally and functionally different from other apparently similar proteins such as the coagulases. The protein and the gene which expresses it are illustrated in Fig.

W0 9s/3~6~ 219 2 ~ 7 5 F~l~u~ oO

3. The ~igure shows the complete clone 36 and the protein it expresses. Clone 14 runs from nucleotide 6a4 to nucleotide 1935 in Fiq. 3. The protein expressed ~y clone 14 runs from amino acid residue 229 to 645. Clone 30 is substantially the same as clone 14 and expresses substantially the same protein.
Sequence analysis clearly indicates that the fibrinogen-reactive proteins of this invention shares ~ign ~ f; cAnt homology with staphylococcal coagulases. Recent evidence by Boden and Flock also suggested that the fibrinogen binding protein of S.
aureus may possess cross-reactivity with anti-coagulase antibody (2). ~owever, several lines of evidence indicate that clone 36 expresses a unique f ibrinogen-binding protein as do clones 14 and 30. First, eYpression studies of clone 14 which expresses amino acid residues 229 to 645 show that the ~Y~essed protein is nPCPe:=Ary for fibrinogen binding. In contrast, ~1 AcsicAl coagulase has been found to co~pLex with ~Lul~"U~in to from staphylothrombin which subsequently converts f ibrinogen to f ibrin (10,17). Secondly, there is a unique stretch of 11 amino acids in the sequence (residues 409-419) that are not found in any of coagulase sequence described. Third, this unique amino acid sequence shares homology with a cell wall anchor motif (LPXTGX) that is found to be nPcPs~ Ary for anchoring in a variety of gram positive surface proteins (13,28). Based on these findings, it would appear that the f ibrinogen reactive protein may belong to a ~amily of coagulase-like proteins, yet it is both structurally and functionally distinct from any of the cqa~lA~D~ previously describ~d .

W0 95~346aa 219 21~ ~ r~ oo .
The isolation cf fibrinogen binding protein from staphyl"rocr~1 whcle CPll lysates with conv~n~i~,n~l chromatographic methcds has been reported in two studies (11,31).
}Iowever, the molecular weight (420 vs 62 kD) and the amino acid composition differs widely between the two studies. In c~
to the 62 kD protein, methionine and tyrosine, ~ut not cysteine residues, are present in the protein of this invention. The fibrinogen reactive protein described here as e~ sced by clone 36 comprises ~)L~ ' jnltely lysine (11.2%), threonine (9.3~s) and glutamic acid (9%) while glycine (16 . 8%), glutamic acid (15%) and lysine (13 . 3%) were the abundant amino acids in the 62 XD protein (31). In addition, the deduced isoelectric point (pI-6.5) of the fibrinogen reactive protein also differs from the basic pI (about 10 . 2) of the 62XD protein.
Previous studies have revealed that the f ibrinogen binding -nt of 5. aureus is a cell wall constituerlt because it is absent in staphylorr,cr 11 ~ form (10) and because bacterial clumping in the presence of fibrinogen is abolished upon whole cell digestion with proteinases (4). In reviewing the molecular architecture of the C-terminal region of other gram positive surface-anchored proteins, it is evident that they contain several conserved features (13,14). In the C-tc~m;n~l of these proteins, a charged tail (4-7 amino acids) is usually preceded by a highly hydrophobic membrane anchor (about 16-20 amino acids), the hexamer LPXTGX, a proline-glycine rich domain and a C-f~-~m;n~l repeat region (14). Clearly, the C-terminal region of the fibrinogen reactive protein is different from the model described. In particular, the region preceding the stop codon lacks a charged tail and a hydrophobic membrane anchor. Instead, the five t~r~n;n~l amino acids (PRVTK) are preceded by five WO 9~3~16aa 21~ 2 i 7 ~ r~ oo repeats of 27 amino acid each. In addition, the region N-tP~;n~l to these repeats i5 a broad proline/glycine region (residues 326-505) in the middle of which is a unique 5Dqn~-n~-e (IPSITGE) that shares homclogy with a cell wall anchor motif (LPXTGX) found in other gram positive surfac2 proteins. Notably, this molecular architecture at the C-terminus is similar to those described for E -: 1 surface protein A (35). The amino tPrm;nAl half of E -~ 1 surface protein A, like that of fibrinogen reactive protein, is ~-helical and is consistent with an ~-helical coiled protein conformation. The ~ -helical region is followed by a proline-rich domain and a repeat domain consisting of ten 20-amino-acid repeats. In addition, it also lacks a classic membrane anchor and a charyed tail. In contrast to the f ibrinogen reactive protein, however, there is no LPXTGE
motif in the c-tPrm;n~1 region of 1 :: 1 surface protein A.
The novel proteins of this invention are surf ace proteins of S. aureus. No such proteins have previously been detected, isolated and characterized. They are pr;n~;r~lly characterized by their ability to bind fibrinogen. The ~olecular weight of the protein ex~,Lessed by clone 36 is about 6g,9gl Da and its isoelective point 6 . 5 . The gene which expresses this protein containS about lg35 nucleotides. Other characteristic features of the protein and of s_ L-. C14 and C30 are described above.
Because of the difficulty in expressing protein from clone 36, the preferred clone of this invention is clone 36. The preferred f ibrinogen binding protein is the protein expressed by this clone .

Wo gsl3~6~ 2 1 9 2 ~ 7 ~ P~ 00 .
The protein of tbis inventicn, as specifically described herein, will be re~-o~n i 70~ by the s~cilled artisan as re ~L~as~:nLltive of a class of protein which may dif~er amongst the various strains of S. aureus, but will all be ~-hAract~ized as having substantially the same number of amino acid residues, the 8ame tertiary :~LLU~,LUL~:: and binding activity. They may differ slightly in the identity of the amino acid residues at 5p-~-; f i r pOSitions in the protein chain . All such proteins are i n~ d within the scope of this invention.
The genes which generate the proteins of this invention may also dif~er slightly amongst various strains, but they all have the common characteristic of producing a protein of this invention .
The genes of this invention may be employed, as will be rPro~Tn; 70~ by the sXilled artisan, to produce plasmids or other vectors which, in turn are useful for transforming organisms such as E. coli to produce novel strains of this bacteria which will express the proteins of the invention.
Inhibition of the binding of S. aureus to endothelial cells is a major factor in preventing infection. Accordingly antibodies to proteins ~c~ssed by clone 36, its segments clone 14 and clone 30 and even smaller segments are important factors in controlling infection. The proteins and protein s~e Ls of this invention are therefore useful to form vaccines to inhibit S. aureus infections of mammals, inr~ in~ humans, by administering an amount of the selected protein which will stiriulate the production of protective ouantities of antibodies .

wo 9~ 16~ 2 ~ g 2 1 7 ~ oo to limit adhesion cf S. aureus. The proteins may also be used to produce anti ho~ c in vitro which may be employed for passiYe ; ~ation.
The proteins and polypeptide or peptide s-, L:, of this invention may be o_tained by any of a number of known process in~ in~ the re ~;n~nt DNA techniques described above.
Polypeptide and peptides within the scope of the invention containing, for example from a~out 6 to 20 or more amino acid Sr~ ~, may be synthesized be standard solid phase p~ uce~uL~:s with appropriate amino acids using the protection, deprotection and cleavage techniques and reagents appropriate to each specif ic amino acid or peptide. A combination of manual and automated (e.g., Applied Biosystem 430A) solid phase techniques can be used to synth--ci ~e the novel peptides of this invention. Although less convenient, classical methods~ of peptide synthesis can also be employed. For ba~ L-,u.,d on solid phase techniques, reference is made to Andreu, D., Merrifield, R.B., Steiner, H. and Boman, H.G., (1983~ Proc. Natl. Acad. sci ~JSA 80, 6475-6479; Andreu, D.
~errifield, R.B., Steiner, H. ~nd Boman, El.G., (1985) Biochemistry 24, 1683-1688; Fink, J., Boman, A., Boman, H.G., and ~errifield, R.B., tJune 1989) Int. J. Peptide Protein Res. 33, 412-421; Fink, J. Merrifield, R.B., Boman, A. and Boman, H.G., (1989) J. Biol. Chem. 264-6260-6267; each of ~-hich being hereby incorporated herein by ref erence .
The products of the inYentiOn are amphoteric. They can exist and be utilized as free bases or as pharmaceutically acceptable metallic or acid addition salts. Suitable metallic salts include alkali and Alki21 in~ earth metal salts, preferably wo ss/346ss 2 19 2 1 7 ~ F~ / I00 sodiuht or potassium salts. Acid addition salts may be p~
from a wide variety of organic and inorganic acids ; n~ ; n~
mineral acids, for example citric, lactic, maleic, tartaric, phosphoric and hydrochloric acids. These salts can be prepared by ~ duL~s well 3cnown to those skilled in the art.
Por use as a vaccine, it is presently preferred to ad~tinister the selected product in a rhA~-~-e~ l ly acceptable carrier such as a bui~fer. Mice or other Dlam~als, ;nrlur~in~
huhtans, when so ; ; 7Pt?l are protected from colonization and ~-~q-lPnt infection by S. aureus.
Typically, the patient to be protected ~ill be treated with product of the invention in an amount which is effective to elicit a protective imhtune response. The sa~ e~-~Pd agent may be ad~tinistered alone or in a rhA~-^ellti~Ally acceptable liquid or solid carrier in which it ~say be qispersed, dissolved or SIlCpPn~P~. If, for example, the patient is to be treated inL~ clY~ uu5ly~ the peptide may be sllcpPn~p~ as a free base or dissolved as a ~-~A l l i 1~ salt in isotonic asueous buf f er . Other htethods of treatment and rhAnr-~ ttically acceptable carriers will be apparent to the skilled artisan.
The proteins, polypeptides and peptides of this invention and the genes or oligonucleotides which are employed in their expression are useful as probes for genes and proteins. They are also useful to raise antibodies by which specif ic strains of streptococci can be identif ied . The sequence of nucleotides which elicit the unique segment from position 409 to position 419 and modification of this sequence are ~CpP~ iAlly useful as diagnostic probes to identify 5. aureus strains. The procedure Wo 9S/3165a 2 1 9 2 t r~ u., 1 ,~ r ls well known to the s3cilled artisan fcr identifying other infectious organisms. It involves the prep2ration of labeled cligonucleotides which aRe used to probe the DNA released fro the ~ gram positive bacteria by all lysis, either n i ~ l l y or enzymatically .

Wo gs/3~6ss ~ 1 ~ 2 ~ 7 ~ P~ C
The follcwing citatlons are mentioned in the application.
They are inccrporated by reference.
t~T'rll~
1. Blake, M.S., K.~.Johnston, G.J.~ussell-Jones, and E.C.
Got crhl i- h. 1984. A rapid sensitive methcd for detection of Allr~lin~ phospha~r~ _o~ju~c~ed anti-body on Western blots.
Anal . Biochem 13 6 :175-179 .
2. Boden, ~.R. and J.I. Flock. 1989. Fibrinogen-binding protein/clumping factor from Staphylococcus aureus. Infect.Immun.
57: 2358-2363 .
3. Cheung, A.L. and V.A. Fischetti. 1989. The role of surface proteins in staphylaco~rA 1 adherence to f ibers in vitro .
J.Clin.Invest. 83:2041-2049.

4. Cheung, A.L. and V.A. Fischetti. l990. The role of fibrinogen in staphylc~oc ~l adherence to catheters in vitro. J.Infect.Dis.
161: 1177-lla6.

5. Cheung, A.L., J.~. Xoamey, C.A. Butler, S.J. Projan, and V.A.
Fischetti. 1992. Regulation of exprotein expression in Staphylacoccus aureus by a locus (sar) distinct from aqr.
Proc . Natl . Acad . Sci . USA . 8 9: 64 62 -64 66 .

W0 95/3l6aa 219 217 ~ r ~ Ioo `

6. Cheung, A.L., J.~. ~oomey, S.Lee, E.A. Jafe, and V.A.
F;C~ t~i. 1991. r- 'in~tnt human tumor necrosis factor alpha promstes 2dherence cf StaDhylococcus aureus to endothelial cells.
Infect.Tmmun.59:3827-3831.
~. Cheung, A.L., l~.Rrishman, E.A. Ja~1'e, and V.A. Ficr~h~ot~;~
lg91. Fibrinogen acts as a bridging molecule in the adherellce o~
Staphylo,-o,~ c aureu5 to culture human endothelial cells. J.
Clin. Invest . B7: 2Z3 6-22 45 .
8. Cohen, M.L. 1992. Epidemiology of drug resistance:
implications for a post-antibiotic era. Science 257:1050-1055.
9. Devereux, J. ,P. ~Iaeberli, and 0. S~ithies. 1984. A
comprehensive set of sequences of analysis programs for the VAX.
.ACids Res. 12:387-395.
10 . Easom, C. S . F. and C. Adlam. 1983 . Staphylococci and staphylococcal infections, ~ mic~ Press, New York.
11. Espersen,F. ,I. t~ ^n, and V. Barkholt. 1985. Isolation of Staphylor-or-~tlc aureus clumping factor. Infect.Immun. 49:700-708 .
12. Feinbery, A.P. and B. Vogelstein. 1983. A technique for radiolabeling DNA re_triction r~ nt~ ;tce rragments to high specific activity. Anal.BioChem. 132:6-13.
13. Fischetti, V.A., V. Pancholi, and 0. Schneewind. 1990.
Conservation of a hexapeptide sequence in the anchor region of surfacs prot~ins from gra~-bacteria. Xol.~icrobiol. 4:16G3-1605.
--2~--2192i7~ r~l/u~
WO 9513465a 14. Fischetti,V.A., ~7.Pancholi, P.Sellers, J.Schmidt, G.~andau, X.Xu, and O. schneewind. 1992.Streptnco~ protein: A ccmmcn 4LL l~_~ULal mctif used by gram pcsitive bacteria ~or biot~ ;r~lly active surface ~ lpc~ p.31-38. In T.R. lrnrh~npn~ T.~ovi, and P.~I. Malela (ed.~, Mclecular recognition in hcst-parasite interactionS, Plenum Press, New York.
15. Garnier, ;r. ,D.J. Osguthorpe, and B.Robson. 1978. Analysis o~
the accuracy and implications cf si3ple methods for predicting the secondary 5~U..L..L~ cf globular proteins.J.Nol.Biol.120:97-120 .
16. EIalvorson,X.O. and D. Perlman. lg83. A putative signal peptidase recognition site and se~uence in eukaryotic and prokaryotic signal peptides.J.Mol.Biol. 167:391-409.
17. ~Iemker, H.C., M.Bas, and A.D.Muller. 1975. Biochim Biophys.Acta 379: 180-la8 .
18. Kaida, S.T.Miyata, Y.Yoshiza~a, H.Igarashi, and S.Iwanaga.
1989. Nucleotide and deduced amino acid sequences of staphylocoagulase gene from Staphyloccccus aureus strain 213.
Nucl.Acids Res.17:8871.
1 9 . Ka ida , S ., T . Miyata , Y . Yoshi zawa , S . Rawa~ata , T . Morita , El.Igarashi, and S. Iwanaga. 1987. Nucleotide sequence of the staphylocoagulase gene: its unique COOH-t~rm;n~l 8 tandem repeat~. J.Biochem. 102 :1177-1186 .
--~6--Wo 95l3~6~ 2 1 ~ 2 1 ~ ~ P~ I.a r 20. }~yte, J. and R.F. Dcclittle. 1982. A simple method for displaying the !~u~:thic character of 2 protein.
J.~Iol .Biol. 157:105--132 .
21. Laemmli, U.~;. 1970. Cleavage of ~ proteins during the ~sse bly of the head of bacteriophage T4, Nature 227:680-685.
22. Lupas, A.,l~. Van Dyke, and J. Stock. 1991. Predicting coiled coils from protein ~s~quPrll~aq. Science 252:1162-1164.
23. Manaitis, T.,E.F. Fritsch, and J. Sambrook. 1989. M~ lAr cloning, a laboratory manualj Cold Spring Harbor Laboratory, Cold spring Harbor, NY.
24 . Neu, H. C. 1992 . The crisis in antibiotic resistance . Science 257: 1064-1072.
25. Pearson, W.R. and D.L. Lipman. 1988. improved tolls for biological s~ an~-e comparison . Proc . Natl . Acad . Sci . IJSA . 88: 4781-4785 .
26. Phoni-'A~n~,P ,~5. O'Reilly, P. Nowlan, A.J. Bramley, and T.
Foster. 1990. The coagualse of Staphylococcus aureus 8325-4.
Sequence analysis and ~irulence of site-specific coagualse-deficient mutants.~ol.l~icrobiol. 4:393-404.
27. Sanger, F. ,S. Nicklen, and A.R. Coulson. 1977. DNA sequencins with chain terminatins inhibitors.
Proc.Natl.Acad.sci.usA.74:5463~
--2l--PCT/US~5107100 Wo 9sl3~6~ Z 1 ~ 2 1 7 S
28. Schneewind, o.,P.~cdel, and V.A. ~ischetti. 1992. Sorting o~
protein A tc the staphylccoccal cell wall. Cell 70:267-281.
29. So, 1~., J.H. Crosa, and S. Falkow. 1975. Polynuclectide secuence rel2~inn~h;rs among Ent pl~ ;~q and the relationship between Ent and other plasmids. J. Bacteriol. 121:234-238.
30. Towbin, H., T. S~hPlin~ and J. Gordon. 1970.
Ele~.L~ I,ur ~:Lic transfer cf proteins from polyacrylamide gels to nitrocPlll-lnse sheets:PLoceduL~ and some applications.
Proc.Natl.Acad.sci.USA. 76:4350-4354.
31. Usui, Y. 1986. Bisrh~;c~l properties of fibrinogen binding (clu~ping factor) of the staphyloroo~ l cell surface.
Zentralbl.Bakteriol.Mi~-obiol.Hyg. ~A]262:287-297.
32. Voller, A. ,D.E. Bidwell, and A. Barlett. 1976. Enzyme qs~ys in diagnostic medicine. Bull.World Health Organ.
53: 55-65 .
33. Waldvogel, F.A. 1985. Staphylccoccus aureus, p.1097-1116. In G.~. ~andell, R.G.Jr. Dcuglas, and J.E. Bennett (ed.), Prin~-;rl~q and Practice of InfectioUs Diseases, John Wiley ~ Sons, New York.
34. Wat~n~lnln~nrn, C. and I.~l. 8aird. 1977. Staphylococcus aureus ba~ rTn; ~ and endocarditis associated with a removable infec:ed intravenous device. Am.J.2~ed. 63:253-256.

r~"~
Wo 93/3463~ 2 1 9 2 1 7 ~
35. Yother, J. and ~.E. Briles. 1992. Structual proper~ies and evnl~t~nn~ry relatf~nc}ir o~ PspA, a sur~acP protein c)~
Sl,L~JtOr.~....C ~ as revealed by s~T~ e analysis J.B~ ri~l. 174:601-609.

Claims (6)

WHAT IS CLAIMED:

1. A gene of the group clone 14, clone 30 and clone 36 which expresses a fibrinogen binding surface protein on S.aureus and segment and analogs of said gene capable of expressing surface proteins on S. aureus having substantially the same activity.

2. A surface protein expressed by a gene of claim 1.

3. A plasmid vector carrying a gene of claim 1.

4. A microorganism transformed by a gene of claim 1 and capable of expressing a fibrinogen binding protein.

5. An antibody to a surface protein of claim 2.

6. A vaccine effective to inhibit the adhesion of S. aureus to fibrinogen because it contains a protein of claim 2.