Pseudomonas peptide compositi on and method for produci ng the same

This application is a continuation-in-part of co- 5 pending U. S . for "Synthetic Pseudomonas aeruginosa Pilin Peptide, " Serial No . 344, 565, filed April 28, 1989.

1. Field of the Invention

The present invention relates to Pseudomonas -derived 10 polypeptide antigens, to methods of producing such anti¬ gens, and to antibodies immunoreactive against the anti¬ gens .

2. References

15 Adams, M.H. Methods of study of bacterial viruses, p. 443-452. In M.H. Adams (ed.), Bacteriophages. Inter- science Publishers, Inc., New York (1959).

Ausubel, F.M., et al.. Current Protocols in Molecu¬ lar Biology, John Wiley and Sons Inc., Media, PA (1990). 20 Beachy, E.H. 1981 J. Infect. Dis. 143:325-345 (1981) .

Boulianne, G.L. et al.. Nature 312:643-646 (1984). Carr, B., et al.. Gerontology 35:127-129 (1989) . Cwirla, S.E. et al., Proc. Natl. Acad. Sci. USA 25 87:6378-6382 (1990).

Devlin, J.J., et al.. Science 249: 404 (1990) .

Doig, P., et al.. Infect. Immun. 56:1641-1646 (1988) .

Doig, P., et al.. Infect. Immun. 58:124-130 (1990). Franklin, A.L., et al.. Infect. Immun. 55:1523-1525 (1987) .

Geyson, H.M. et al., in Synthetic Peptides as Anti¬ gens; Ciba Foundation Symposium 119:131-149 (1986) .

Irvin, R.T. and Ceri, H., Can. J. Microbiol. 31:268- 275 (1985) . Irvin, R.T., et al., Microbial Ecology Health Disease, 3:39-47 (1990).

Lee, K.K., et al.. Infect. Immun. 57:520-526 (1989). Lee, K.K., et al, Inf Immunol, 58:2727-32 (1990). Marrs, C.F., et al.. Am. J. Med. 88 (Suppl 5A) : 36S-40S (1990) .

McBride, L.J., et al., Clin Che , 35:2196-2201 (1989).

McEachran, D.W., et al. Can. J. Microbiol. 31:563-569 (1985). McEachran, D.W., et al., J. Microbiol. Meth. 5:99- 111 (1986) .

Morrison, S.L., et al., Proc. Natl. Acad. Sci. USA 81:6851-6855 (1984) .

Nieto, A., et al., Mol. Immunol. 21:537-543 (1984). Pasloske, B.L., et al., J Bacteriol, 170:3738-3741 (1988) .

Paranchych, W., et al.. Can. J. Microbiol. 25:1175- 1181.

Paranchych, W. et al., Advan. Microbiol. Phys. 29:53-114 (1988) .

Parmley, S.F. and Smith, G.P. Gene 73:305-318 (1988) .

Rabbitts, T.H. et al.. Nucleic Acids Res. 9:4509- 4524 (1981) .

Sanger, S., et al., PNAS (USA), 74:5463-5467 (1977).

Sato, H. and Okinaga, K., Infect. Immun. 55: 1774- 1778 (1987) .

Scott, J.K. and Smith, G.P., Science 249:386-390 (1990) .

Sastry, L., et al., Proc. Natl. Acad. Sci. USA 86:5728-5732 (1989) .

Sastry, P.A., et al.. Can. J. Cell Biol. 63:284-291 (1985) . Staddon, W., et al.. Can. J. Microbiol. 36:336-340 (1990) .

Todd, T., et al.. Am. Rev. Respir. Dis. 140:1585-1589 (1989).

Tsai, C, et al.. Anal. Biochem. 119:115-119 (1982). Worobec, E.A., et al., J. Biol. Chem. 260:938-943 (1985) . zn Putlitz, J., et al., Bio/Technology 8:651-654 (1990) .

3. Background of the Invention

During the past two decades, Pseudomonas aeruginosa has been recognized as a pathogen which causes between 10% and 20% of infections in most hospitals. Pseudomonas infection is especially prevalent among patients with burn wounds, cystic fibrosis, acute leukemia, organ transplants, and intravenous-drug addiction. _. aerugi¬ nosa is a common nosocomial contaminant, and epidemics have been traced to many items in the hospital environ¬ ment. Patients who are hospitalized for extended periods are frequently affected by this organism and are at increased risk of developing infection. The most serious infections include malignant-external otitis, endophthal- mitis, endoconditis, meningitis, pneumonia, and septice- mia. The likelihood of recovery from Pseudomonas infec-

tion is related to the severity of the patient's underly¬ ing disease process. The reported mortality for P_. aeru¬ ginosa pneumonia is as high as 50-80%. Even with the development of newer antibiotics, resistance remains a problem necessitating combined antibiotic treatment for severe P_. aeruginosa infections.

Various therapies for the management of severe P_. aeruginosa infections have been evaluated for many years, with particular attention focused on virulence factors. As with most bacterial pathogens, virulence of P_. aerugi¬ nosa is multifactorial and is the product of many inter¬ acting variables, involving both the bacterium and the host. Evidence suggests that the initial event in infec¬ tion is the adherence of microorganisms to epithelial cells of mucosal surfaces (Bleachy) . Organisms that are unable to adhere to mucosal surfaces fail to colonize because they are removed by the secretions that bathe the mucosal surfaces (Bleachy) . The adherence process is dependent upon the specific recognition between bacteria and epithelial cells.

For a number of gram-negative bacteria, including P. aeruginosa, attention has been directed to surface appen¬ dages as mediations of adherence. The surface of many gram-negative bacteria, e.g., Escherichia coli, P_. aeru- ginosa, Moraxella bovis, Neisseria gonorrhea, are covered with filamentous structures called pili or fimbriae. Pili are composed primarily of protein (pilin) and have been found to act as antigenic determinants when injected into test animals. In P_. aeruginosa, strain-specific pili, such as those designated PAO, PAK, and CD4, mediate the colonization of the bacteria in humans (Doig,88).

Some P_. aeruginosa bacteria lacking these pili, either through mutation or loss of the plasmid carrying the pilus gene, are incapable of colonizing mucosa.

Apparently, the pili on the surface of the bacterium adhere to the lining of the throat and trachea through specific interactions with epithelial cell receptors. P_. aeruginosa can utilize both pili and alginate (the prin- ciple component of the P_. aeruginosa capsule) as adhesins to mediate attachment to human respiratory epithelial cells (Doig) .

Equilibrium analysis of P_. aeruginosa binding to human respiratory epithelial cells indicates that the Pseudomonas pilus adhesin has a considerably higher appa¬ rent affinity or binding constant than does the alginate adhesin (McEachran, 1985, 1986) . These observations sug¬ gest that the pilus adhesin is likely the dominant Pseud¬ omonas adhesin in the initiation of an infection (Irvin) . Adhesion-mediated anchorage is a prerequisite for the induction of disease by P_. aeruginosa.

The earlier filed co-pending patent application dis¬ closes a P_. aeruginosa peptide derived from the C-termi- nal region of the P_. aeruginosa pilin protein, and speci- fically, the C-terminal region which includes two Cys residues and the intervening amino acid residues. The derived region of representative peptides vary in length between 14 and 19 amino acid residues, including the two Cys residues, and are prepared in both oxidized (disul- fide-linked) and reduced (non-cyclized) form. The pep¬ tides (in both reduced and oxidized form) were shown to have the following properties:

(a) ability to bind to human tracheal epithelial cells (TECs) and human buccal epithelial cells (BECs) ; (b) ability to inhibit binding of Pseudomonas pilin peptide to tracheal epithelial cells (TECs) and buccal epithelial cells (BECs) ;

(c) ability to elicit serum antibodies which are immunoreactive with Pseudomonas pilin peptide; and

(d) ability to elicit serum antibodies which block binding to Pseudomonas pilin peptide to BECs.

It has now been discovered that the Pseudomonas- derived peptide is able to inhibit binding of unrelated bacterial and fungal organisms to human TECs and/or BECs. Thus, the epithelial cell receptor site(s) which bind the Pseudomonas-derived peptide, and thereby inhibit binding of Pseudomonas pilin (and Pseudomonas bacteria) to TECs and BECs is also involved in binding of other bacterial and fungal organisms to these target cells. It has further been shown, in studies conducted in support of the present invention, that monoclonal antibodies pre¬ pared against the Pseudomonas-derived peptide are effec¬ tive in blocking fungal cell adherence to BECs. These combined findings show that the Pseudomonas- derived peptide, and antibodies produced in response to the peptides, are capable of inhibiting bacterial and fungal infections in which the infecting microorganism has surface proteins which are antigenically crossreac- tive with antibodies produced against the C-terminal, disulfide-linked peptide region of P. aeruginosa pilin protein.

4. Summary of the Invention The invention includes, in one aspect, a peptide having a sequence corresponding to a C-terminal region of a P_. aeruginosa pilin protein, and more specifically, to one of the sequences:

(i) K C T S D Q D E Q F I P K G C S K, (ii) K C T S D Q D E Q F I P K G C S R,

(iii) A C K S T Q D P M F T P K G C D N,

(iv) T C T S T Q E E M F I P K G C N K ,

(v) S C A T T V D A K F R P N G C D,

(vi) A C T S N A D N K Y L P K T C Q T A T T T T P,

(vii) N C K I T K T P T A W K P N Y A P A N C P K S, (viii) T C G I T G S P T N W K A N Y A P A N C P K S, (ix) T C G I T G S P T N W K T N Y A P A N C P K S, and (x) G C S I S S T P A N W K P N Y A P S N C P K S, including amino acid variations which are internally consistent among sequences (i)-(vi) and among sequences (vii)-(x) .

The peptide of the invention is further characte¬ rized by (a) a disulfide linkage between the two Cys (C) residues; (b) immunospecific binding to PK99H or PK34C monoclonal antibody; (c) specific binding to tracheal or buccal epithelial cells; and (d) absence of specific binding to P_. aeruginosa pilin protein.

In a related aspect, the invention includes a com- position for use as a vaccine against infection by bac¬ terial and fungal organisms which have surface proteins which are antigenically crossreactive with antibodies produced against the C-terminal, disulfide-linked peptide region of P_. aeruginosa pilin protein. The composition includes the above peptide and an immunogenic carrier to which the peptide is attached. In one embodiment, for use as a vaccine against Pseudomonas infection, the pep¬ tide is immunoreactive with PK34C monoclonal antibody. In another embodiment, for use as a vaccine against Candida infection, the peptide is immunoreactive with PK99H antibody.

In another aspect, the peptide has an amino acid sequence selected from the group consisting of: (i') D E Q F I P K, (ii') D E Q F I P K, (iii') D P M F T P K, (iv') E E M F I P K, (v') D A K F R P N, and (vi') D N K Y L P K,

including amino acid variations which are internally con¬ sistent among sequences (i')-(vi') and characterized by: (a) immunospecific binding to a PK99H or a PK34C mono¬ clonal antibody; (b) specific binding to human buccal or human tracheal epithelial cells; and (c) absence of spe¬ cific binding to Ε_. aeruginosa pilin protein.

The peptide composition is employed in a vaccination method of protecting an individual against infection by bacterial and fungal organisms which have surface pro- teins which are antigenically crossreactive with anti¬ bodies produced against the C-terminal, disulfide-linked peptide region of P_. aeruginosa pilin protein.

Also forming part of the invention is a method of producing a composition for use as a vaccine against infection by bacterial and fungal organisms which have surface proteins which are antigenically crossreactive with antibodies produced against the C-terminal, disul¬ fide-linked peptide region of P_. aeruginosa pilin pro¬ tein. The method relies on the selection of random- sequence peptides produced by a vector library of random- sequence polynucleotides, typically corresponding to a random sequence of 5-10 codons. Selection is by peptide binding to a monoclonal antibody immunoreactive with the C-terminal, disulfide-linked peptide region of P_. aerugi- nosa pilin protein, and preferably to the PK34C or PK99H monoclonal antibody. The sequence of the selected bind¬ ing polypeptide can be determined from the polynucleotide coding sequence of the corresponding library vector. From this sequence, the desired polypeptide can be made by synthetic or recombinant means.

Also disclosed is a chimeric monoclonal antibody composed of the variable regions of mouse PK34C or PK99H monoclonal antibody, and the constant regions of a human immunoglobulin G antibody. The antibody can be used to

treat infection by bacterial and fungal organisms having surface proteins which are antigenically crossreactive with antibodies produced against the C-terminal, disul¬ fide-linked peptide region of P_. aeruginosa pilin pro- tein.

In still another aspect, the invention includes a method of treating an infection of the lung caused by a bacterial or fungal organism which have surface proteins which are antigenically crossreactive with antibodies produced against the C-terminal, disulfide-linked peptide region of P_. aeruginosa pilin protein. The method in¬ cludes forming an aerosol of the peptide and administe¬ ring the peptide by inhalation.

These and other objects and features of the present invention will be come more fully apparent when the fol¬ lowing detailed description of the invention is read in conjunction with the accompanying drawings.

Brief Description of the Drawings Figure 1 shows the amino acid sequences of Pseudomo¬ nas peptides of the invention;

Figure 2 shows the relative binding affinity of na¬ tive PAK peptide and PAK peptides with modified C- terminus and N-terminus and peptides containing Ala substitutions at each of the 17 peptide residues except for the Cys residues for monoclonal antibody PK99H;

Figure 3 is a plot showing the binding of synthetic peptide PAK _rβd (solid squares) and PAK _0X to human BECs;

Figure 4 is a modified Lineweaver-Burk plot of the binding of PAK pili to human BECs, showing inhibition of binding by increasing concentrations of PAK peptide;

Figure 5 is a bar graph showing the percent of PAK pili binding to BECs, after preincubation of BECs with Fab fragments of the antibodies indicated;

Figure 6 is a plot showing the inhibition of Pseud- o - monas bacteria binding to BECs by Fab fragments of mono¬ clonal antibodies PK99H (open squares) and PK34C (open triangles) ;

Figure 7 shows fractionated M. catarrhalis proteins stained by immunoblot with polyclonal anti-PAK pili anti¬ body (lane 1) , and with protein stain (lane 2) and standards (lane 3) ; Figures 8A and 8B are immunoelectron micrographs showing M. catarrhalis cells after binding with polyclonal anti-PAK pili antibody (8A) and a control antibody (8B);

Figures 9A and 9B are electron micrographs showing M. catarrhalis bacteria with PK99H antibody localized by colloidal gold;

Figure 10 shows a Western blot of several Bacteroi- des strains immunoblotted with PK99H monoclonal antibody; Figure 11 is a phase contrast micrograph of Candida cells bound to BECs;

Figure 12 shows SDS-PAGE gel patterns of purified Candida fimbrial protein (lane 3) ;

Figure 13 is a bar graph showing the inhibition by Candida fimbriae on Candida adherence to BECs; Figure 14 is a bar graph showing the inhibition by Pseudomonas pili on Candida adherence to BECs;

Figure 15 is a bar graph showing the inhibition by Candida fimbriae on Pseudomonas pili binding to BECs;

Figure 16 is a bar graph showing the inhibition by anti-PAK pili antibodies on Candida binding to BECs;

Figure 17 illustrates recombinant methods for produ¬ cing and selecting random-sequence peptides, in accord¬ ance with the invention; and

Figure 18 illustrates recombinant methods suitable for producing a chimeric monoclonal antibody in accord¬ ance with the present invention.

Detailed Description of the Invention

I. Definitions

The terms "epitope" and "epitopic, " as used herein, designate the structural component of a molecule that is responsible for specific interaction with correspon¬ ding antibody (immunoglobulin) molecules elicited by the same or related antigen.

The term "antigen, " as used herein, means an entity that is recognized by an antibody. The term "immunogen, " as used herein, describes an entity that induces antibody production in the host animal. In some instances the antigen and the immunogen are the same entity, while in other instances the two entities are different. The term "immunologically mimics" is used herein to mean that an immunogenic polypeptide of this invention is not a natural protein or a cleaved fragment of a natural protein, but a manufactured polypeptide, as by solid phase synthesis or genetic engineering techniques, which polypeptide induces production of antibodies that bind to the inducing polypeptide and also to a corresponding pilin or pilin polypeptide portion.

All amino acid residues identified herein are in the natural or L-configuration unless otherwise specified. In keeping with standard peptide nomenclature, abbrevia¬ tions for amino acid residues that have been used herein are as follows:

Symbol Amino Acid

1 Letter 3 Letter

Figure 1 shows the C-terminal amino acid sequences, and corresponding polynucleotide coding sequences, of the pilin protein from ten P_. aeruginosa strains which have been sequenced to date. The P_. aeruginosa strains or isolates from which the sequences were obtained are given at the left in the figure, and are used herein to desig¬ nate the particular pilin peptide sequence. Strains PAK, PAO, and 492c have been reported (Doig, 1990) , as have strains CD4, KB7, K122, GA1, . and TBOU1 (Posloske) . The strain designated PAK(R) is muant PAK strain containing a C-terminal Lys-to-Arg substitution (Sastry) . The C-ter¬ minal sequences of strains PAK(R), PAO, CD4, K122, KB7,

PI, 492C, and TBOUl were reported in the earlier filed co-pending application.

The corresponding polynucleotide coding sequences for the various strains were determined from published reports, or by sequencing isolated P_. aeruginosa genomic material obtained from the individual strains. The genomic material was amplified by polymerase chain reaction (PCR) methods, using degenerate probes corre¬ sponding to N-terminal and C-terminal regions of the amino acids sequences shown in the figure, according to conventional procedures (McBride) . Sequencing of the amplified genomic material was by dideoxy sequencing, according to standard procedures (Sanger) .

With continued reference to Figure 1, the preferred peptide sequences include (a) the Cys-to-Cys residues, (b) a residue immediately N-terminal to the Cys-to-Cys residues and (c) the residue immediately C-terminal to the Cys-to-Cys residues, and preferably all of the resi¬ dues C-terminal to the Cys-to-Cys residues. These pre- ferred sequences are indicated as sequences (i)-(x) below, using the conventional single-letter amino acid codes given above:

(i) K C T S D Q D E Q F I P K G C S K,

(ii) K C T S D Q D E Q F I P K G C S R, (iii) A C K S T Q D P M F T P K G C D N,

(iv) T C T S T Q E E M F I P K G C N K P, (v) S C A T T V D A K F R P N G C T D, (vi) A C T S N A D N K Y L P K T C Q T A T T T T P, (Vii) N C K I T K T P T A W K P N Y A P A N C P K S, (viii T C G I T G S P T N W K A N Y A P A N C P K S,

(ix) T C G I T G S P T N W K T N Y A P A N C P K S, and (x) G C S I S S T P A N W K P N Y A P S N C P K S,

As seen, the ten sequences can be classed into two groups: one group (sequences i-vi) containing 14 Cys-to-

Cys residues, and a second group (sequences vii-x) con¬ taining 19 Cys-to-Cys residues. The P. aeruginosa strains from which each sequence is derived are: (i) , PAK; (ii) , PAK (Lys-to-Arg mutation) ; (iii) , PAO; (iv), CD4; (v), KB7; (vi) , K122; (vii) , PI; (viii) , GA1; (ix) , 492C; and (x) , TBOUl. The peptides are referred to here¬ in by their strain designation. For example, a peptide containing the sequence (i) is referred to herein as the "PAK peptide", meaning the C-terminal, disulfide-linked peptide region of P. aeruginosa K pilin protein, having the additional constraints noted below.

Preferred disulfide-linked peptides also include amino acid variations of the given sequences which are internally consistent among sequences (i)-(vi) and among sequences (vii)-(x). Thus, for example, the first (left- hand) residue into the group of peptide (i) -(vi) contains possible variations K, A, T, and S, and the third posi¬ tion in this group contains possible variations T, K, and A. It is noted that the internal-variation substitu¬ tions are those substitutions found in nature, and thus are apparently compatible with requisite antigenic pro¬ perties of the peptide. Further, the substitutions are generally within groups of amino acids having similar properties related to one or more of the following: (1) hydrophobicity; (2) polarity; (3) size of side chain; (4) charge; (5) preference for turns; (6) preference for beta strand secondary structure; and (7) preference for helical secondary structure. For example, at several positions, the allowed substitution variations is between T (Thr) and S (Ser) , or between Y (Tyr) and F (Phe) . The amino acid variation is also supported by the Ala- substitution effects discussed below.

In addition, the peptides are characterized by:

(a) a disulfide linkage between the two Cys (C) residues;

(b) immunospecific binding to PK99H or PK34C monoclonal antibody (Mab) ;

(c) specific binding to human buccal or human tracheal epithelial cells; and

(d) absence of specific binding to P_. aeruginosa pilin protein.

The disulfide link between the two Cys residues, which effectively cyclizes the peptide, has been found to be important for the immunogenicity of the peptide, when preparing anti-sera which are cross-reactive with pili from different P_. aeruginosa strains (Lee) , as will be discussed below.

The PK99H and PK34C Mab's are prepared against the PAK peptide (in oxidized form) and are cross-reactive with various other P_. aeruginosa strain pili, as will be seen below. The requirement that the peptide of the invention have crossreactivity with at least one of these antibodies ensures that the peptide has requisite epito- pic similarity to the PAK peptide.

Also as will be seen below, the peptide of the in¬ vention has the ability to bind to a receptor site on human buccal epithelial cells (BECs) and human tracheal epithelial cells (TECs) , and this binding is effective to inhibit P_. aeruginosa binding to these epithelial cells. The requirement for peptide binding to these cells ensures that the peptide has the requisite receptor binding activity.

The absence of specific binding to _. aeruginosa pilin protein distinguishes the peptide from earlier reported C-terminal P_. aeruginosa fragments (Paranchych) which contain the C-terminal sequences of the PAK strain peptide, but in addition, contain N-terminal residues which cause peptide binding to pilin protein. Such

binding is presumably related to the self-aggregating property of the pilin protein. Such binding represents unwanted epitope(s) for the purposes of the present invention. The peptides of the invention may contain additional N-terminal or C-terminal residues, consistent with the above constraints.

The effect of Ala substitutions at each of the 15 residue positions in the PAK peptide other than the two Cys residues was examined, to identify the region of the peptide most sensitive to amino acid variations. Brief¬ ly _/ peptides with specific Ala substitutions were com¬ pared with unsubstituted PAK peptide for binding affinity to anti-PAK monoclonal antibody PK99H (described below) . The unsubstituted and substituted peptides were prepared by solid-phase synthesis, substantially as described in Example 1. Relative binding affinities of the unsub¬ stituted and each of the substituted peptides was deter¬ mined by competitive enzyme-linked immunosorbent assay (ELISA) , according to standard procedures. Relative binding was expressed as logIC ₅₀ (substituted)-logIC ₅₀ (native) , where IC ₅₀ is the peptide concentration needed to displace 50% of an enzyme-linked peptide from immobilized antibody. The results are shown in Figure 2. A positive value of logIC ₅₀ (substituted)-logIC ₅₀ (native) indicates loss of binding affinity in the substituted peptide. As seen, the residue positions most sensitive to Ala substitution are positions 7 (Asp) , 8 (Glu) , 10 (Phe) , 11 (He) , 12 (Pro) , and 13 (Lys) . As seen in Figure 1, these positions are highly conserved in the peptide sequences (i)-(vi), particularly at positions 7 (Asp and Glu), 10 (Phe and Tyr) , 12 (Pro) , and 13 (Lys and Asn) .

Since residues 7-13 are most critical for binding activity, and substitution on either side of this region has relatively little effect on binding activity, it is seen that a peptide containing this 7-mer, and optionally, N-terminal or C-terminal flanking residues, may also be used for the peptide applications described below, including its use as a vaccine against P_. aerugi¬ nosa and against bacterial and fungal bacterial and fungal organisms which have surface proteins which are antigenically crossreactive with antibodies produced against the C-terminal, disulfide-linked peptide region of Pseudomonas aeruginosa pilin protein.

The sequences of the 7mer peptide, and corresponding 17mer peptides derived from sequences (i)-(vi) above are shown at (i')-(vi') below, respectively. As with sequen¬ ces (i)-(vi), the 7mer sequences include internal varia¬ tions among the six sequences, and the peptide is further characterized by (a) immunospecific binding to PK99H or PK34C monoclonal antibody; and (b) absence of specific binding to P_. aeruginosa pilin protein. The peptide may be flanked by disulfide-linked Cys groups, preferably spaced 1-5 residues from the N-terminal D or E residue of the 7mer, and 1-2 residues from the C-terminal Lys or Asn residue of the 7mer. (i') D E Q F I P K, (ii') D E Q F I P K, (iii') D P M F T P K, (iv') E E M F I P K, (v' ) D A K F R P N, and (vi') D N K Y L P K.

The P_. aeruginosa strains from which each sequence is derived are: (i'), PAK; (ii'), PAK (Lys-to-Arg muta¬ tion); (iii'), PAO; (iv' ) , CD4; (v') , KB7; and (vi' ) , K122. These "core" peptides are referred to herein by

their strain designation. For example, a peptide containing the sequence (i) is referred to herein as the "PAK core peptide", meaning the C-terminal, disulfide- linked peptide region of P. aeruginosa K pilin protein, having the 7 core residues and additional constraints noted above.

The binding of the PAK peptide (in both reduced and oxidized form) to BECs and TECs, and the ability of the peptide to inhibit pilin protein binding to TECs and BECs has been described in the earlier-filed co-pending application. Briefly, BEC and TEC preparations were made as described in Example 2. Binding of the PAK peptide to BECs was carried out by first successively contacting BECs with (a) the PAK peptide, (b) PK99H mouse monoclonal antibody (which binds immunospecifically to PAK peptide) , and (c) enzyme-labeled goat anti-mouse antibody. The amount of peptide bound (expressed as measured enzyme activity) as a function of peptide concentration is shown in Figure 3 for reduced (solid squares) and oxidized (+) peptide. Binding to the PAK peptide to TECs was simi¬ larly shown.

The ability of the PAK peptide to inhibit pilin pro¬ tein binding to BECs was measured by competitive binding in which BECs were first incubated with one of a series of increasing concentrations of reduced-form PAK peptide, then with pilin protein, at one of a series of increasing concentrations. The amount of pilin protein bound to the cells was measured by contacting the cells successively with a PK3B mouse monoclonal antibody (which is immuno- specific against pilin protein, but does not recognize the PAK peptide) and enzyme-labeled goat anti-mouse anti¬ bodies, then assaying enzyme activity bound to the cells. The results are plotted as a Lineweaver-Burke plot in Figure 4, which shows the inverse of measured enzyme

activity plotted as a function of the inverse of pili protein concentration, at 0 (X) , 40 (£) , 80 ⁽ open triangles), and 120 (open diamonds) nmoles/ml of peptide. As seen, the peptide produced a concentration-dependent inhibition of pilin protein binding to BECs.

III. Anti-Peptide Antibodies

This section summarizes methods of production, and antibody binding characteristics of polyclonal and mono- clonal antibodies which are immunoreactive with the pep ^¬ tides of the invention. The antibodies are useful in selecting random-sequence peptides having cross-reac ^¬ tivity with P_. aeruginosa pilin C-terminal peptide, as detailed in Section V, and in preparing chi eric thera- peutic antibodies, as detailed in Section VII.

IIIA. Polyclonal Antibodies

Polyclonal antibodies specific against reduced and oxidized forms of PAK peptide were prepared as described in the earlier-filed co-pending application, and as pub ^¬ lished (Lee) . Briefly, PAK peptides were conjugated to keyhole limpet hemocyanin (KLH) , and the co jugate was used to immunize female Flemish rabbits. The peptides include the PAK peptide in reduced (PAK _r#d ) and oxidized form (PAK _0X ) form, and PAK with an Ala substitution at the N-terminal Cys residue (PAK^.) . Rabbits were given an initial immunization, two weeks later given a booster immunization, and then bled two weeks later. An immuno- globulin fraction was purified by Protein A affinity chromatography. Antibody binding to native PAK pilin protein, PAK peptide, and PAO peptide was examined by Stan _d ar _d H _I JJ OA piuueuuie_ _> vwuiuuet- . rtii iwuujf ^.t-«_.-.____._.._.._. ties were as follows:

(a) The antisera produced by both PAK _0X and PAK _rβd was able to bind native PAK pili, and the titres raised against both peptides were similar;

(b) The antisera raised against the PAK _ox peptide was strongly crossreactive with native PAO pili;

(c) The antisera raised against the PAK _rβd peptide was only weakly crossreactive with native PAO pili; and

(d) antisera prepared against the PAK^ peptide did not bind to either PAK or PAO pili protein. The results show that, although both oxidized and reduced forms of the peptide are effective to induce antibodies which are reactive with same-species pilin protein, the oxidized (disulfide-linked) form of the pep¬ tide is important for stimulating production of anti- bodies which are cross-reactive with pilin proteins from other P_. aeruginosa strains.

The ability of the polyclonal antibodies to inhibit PAK pilin binding to BECs was examined, as detailed in Example 3. Briefly polyclonal antibodies were prepared against several peptide regions corresponding to the PAK peptide and from these. Fab fragments were prepared. The Fab fragment designations are ("rl," "r2") and ("ol" and "o2") , against the reduced (r) and oxidized (o) forms of PAK peptide (residues 128-144 of the PAK pilin protein) ; "22," against residues 22-33 (of the PAK pilin protein); "41," against residues 41-49; "58," against residues 58-70; "75," against residues 75-84; "89," against resi¬ dues 89-99; "107," against residues 107-116; and "117," against residues 117-125. "Pre" refers to preimmune sera; and "99H" to monoclonal antibody PK99H. The Fab fragments were preincubated . with PAK pili before the addition of BECs, and the amount of pilin protein bound to the BECs was detected, as above, by successive binding of mouse monoclonal antibody PK3B (which is specific

against pili protein, but not the PAK peptide) , and enzyme-linked goat anti-mouse antibody. The results, expressed as percent inhibition of pili binding with respect to preimmune antibody Fab fragments, are shown in the bar graph of Figure 5.

The bar graph demonstrates that Fab fragments pro¬ duced against regions other than the C-terminal of PAK pilin are ineffective at preventing pilin binding to BECs. The most effective fragments are rl, r2, ol and 02, directed against residues 128-144, reducing pili binding to 40% to 70% of the control and preimmune serum. This is similar to the effect shown by Fab 99H which is made from anti-PAK pilin monoclonal antibody PK99H (de¬ scribed below) which is also directed at this C-terminal region.

The studies with monoclonal antibodies, presented below, confirm that antibody inhibition of pilus binding to TEC or BEC cells also inhibits P_. aeruginosa binding to these cells.

IIIB. Monoclonal Antibodies

Monoclonal antibodies against native PAK pili pro¬ tein were prepared according to methods described else¬ where by the inventors (Doig) . Briefly, BALB/c mice were immunized with weekly injections of PAK pili. Spleen cells from the animals were fused with mouse myeloma cell line NS1 (Irvin) , and successful fusions were screened by an ELISA method for ability to secrete anti-pilin anti¬ body . A library of 262 hybridoma clones that secreted antibodies immunoreactive with PAK pili were obtained. Protein A purified Mabs were then screened against pilin peptide fragments (Doig) , to determine specificities of these antibodies. Four hybridoma cell lines were selec¬ ted for further specificity studies: cell lines PK99H,

PH34C, PK3B, and PK41C.

_I muno _b lots of purified PAK and PAO pili revealed that PK99H and PK3B Mabs were specific for P ^A K pilin protein, while PK34C and PK41C Mabs were immunoreactive with both P _A K and PAO pilin peptide. PK99H and PK34C _M abs were both immunoreactive with a C-terminal fragment of PAK pilin.

_F ab fragments prepared from PK99H and PK34C were examined for their ability to inhibit Pseudomonas pili bin _d ing to _B E _C s, as detailed in Example 4B. Briefly, Fab fragments of PK99H, PK34C, and non-immune Ig ^G were preincu _b ated with PAK pili at the concentrations indica ^¬ ted in Ta _b le 1, followed by addition of BECs and further incubation at 37°C for 2 hours. Binding to the ^BE Cs was _d etected _b y an ELISA method, with the results shown in Table 1.

Table 1 Fab Concentration ^* % °f

Fragment (μg/ml) Control

** Fab fragments prepared from normal mouse Ig .

_A s seen, both PK99H and PK34C Fab fragments produced a concentration-dependent inhibition of pili binding to BECs. Non-immune IgG Fab fragments produced only a slight decrease in pili binding.

The ability of Fab fragments of PK99H and PK34C to block binding of a number of different P. aeru ^g inosa strains ₍ Table 2) to BECs was also investigated. The bacterial strains were first incubated with the Fab

fragments, then mixed with BECs. Binding of the bac- t e r i a to the cells was performed as described (McEarchran) . Figure 6 shows the inhibition of P_. aeruginosa K binding to BECs, as a function of antibody Fab concentration. The Fab fragments were prepared from the monoclonal anti¬ bodies PK99H (open squares) and PK34C (open triangles) . As seen, both antibody fragments are effective in inhibiting P_. aeruginosa binding to target epithelial cells.

The effect of the PK99H, PK34C, and non-immune control IgG Fab fragments binding of the different Pseudomonas strains is given in Table 2.

* The concentration of PK99H and PK34C Fab used in the inhibi¬ tion assay was 100 μg/ml and had a titer of 10 ^s by ELISA, using PAK pili as the antigen (coated at 1 μg per well) . Given is the mean ± the standard deviation. The percent of control is given in parentheses. ^b Control value when 100 μg of Fab fragments per ml produced from normal mouse IgG was added. No difference was noted between these values and those from tubes to which no Fab fragments were added. ° The significant difference (P <0.05) was determined by using the Student c test.

SUBSTITUTE SHEET

The data show that the PK99H antibody produces bind¬ ing inhibition of binding of strains PAK, HD1, and 492c. The PK34C antibody, by contrast, produces a significant inhibition of binding of all of the strains tested except PAK/3. The results indicate that the PK34C antibody is more crossreactive, among Pseudomonas strains, than the PK99H antibody. The data also demonstrate that anti¬ bodies effective in inhibiting Pseudomonas binding to BECs are also effective in inhibiting Pseudomonas bacterial attachment to BECs.

IV. Inhibiting Bacterial and Fungal Infections

As demonstrated above, antibodies produced against the C-terminal disulfide-linked peptide region of B_. aeruginosa K pilin protein, such as monoclonal antibody PH34C, are immunoreactive with pilin protein from a variety of P.. aeruginosa strains, as evidenced by the ability of the antibody to block binding of different Psuedomonas strains to BECs. In accordance with one aspect of the invention, it has been discovered that antibodies produced against the C-terminal, disulfide-linked peptide region of _. aerugi¬ nosa are crossreactive with surface proteins on a variety of bacterial and fungal microorganisms. Antibody binding to bacterial and fungal proteins and/or cells are pre¬ sented below. Additional studies on the ability of the antibodies to inhibit Candida albicans binding to BECs, also presented below, demonstrate that such binding is effective to inhibit cell binding to target epithelial cells, such as BECs.

The invention thus includes, in another aspect, a method of blocking attachment to target epithelial cells, of bacterial and fungal organisms which have surface proteins which are antigenically crossreactive with anti-

bodies produced against the C-terminal, disulfide-linked peptide region of Pseudomonas aeruginosa pilin protein. The method includes contacting the bacterial or fungal microorganism with such antibodies produced against the C-terminal, disulfide-linked peptide region of £. aerugi¬ nosa, to bind to the crossreactive surface protein. This binding is then effective to block binding of the micro¬ organism with target epithelial cells, such as TECs and BECs. The peptide used to produce the antibody is prefer¬ ably selected from the peptide disclosed in Section II, including a peptide selected from the group of peptides identified as (i)-(vi), (vii-x) , and (i')-(vi'). Alter¬ natively, the peptide used to produce the antibody is one selected for its cross-reactivity with P_. aeruginosa, as described in Section V.

The antibody, either a polyclonal or monoclonal, may be generated by standard methods, such as those outlined in Section III. One antibody useful in the method is the PK99H or the PK34C -monoclonal antibody described in Sec¬ tion III. For therapeutic purposes, i.e., where the antibody is administered parenterally, the antibody is preferably a chimeric antibody containing the variable region of a mouse monoclonal antibody, such as antibody PK99H or PK34C, and the constant region of a human immu- noglobulin gene. Details of preparing such an chimeric antibody of this type are given in Section VII below. Alternatively, the antibody may be produced by vaccina¬ tion with a C-terminal P_. aeruginosa peptide, as de- scribed in Section VI.

IVA. Moraxella catarrhalis

Moraxella catarrhalis produces two morphological forms of pili (Marrs) and binds to human respiratory epi-

SUBSTITUTE SHEET

thelial cells (Carr) . It has been suggested ⁽ Marrs ⁾ that at least one of the pili produced is an N-methylphenyl alanine ₍ N-MePhe) pilus, the same class of pili that is produced by Neisseria, P_. aeruginosa, Moraxella bovis, Bacteroides nodosus, and vibrio cholerae ⁽ Paranchych, 1988 ₎ based on agar corrosion, twitching motility, and probing with a M. bovis pilin gene probe.

Experiments conducted in support of the present invention have confirmed the earlier observations ⁽ Marrs ⁾ that M. catarrhalis produces two morphological forms of pili, ^" designated alpha and beta pili. Further studies in support of the invention have shown that a 1.2 kb Hindi11 £. aeruginosa PAK pilin gene probe hybridizes with rea ^¬ sonable stringency to restriction endonuclease fragments of a number of M. catarrhalis clinical isolates. Addi ^¬ tional studies have established that rabbit polyclonal anti-Pseudomonas PAK pili antisera (Section III above ⁾ reacts specifically with an 18 kD protein ⁽ lane 2 in Figure 7 ₎ in immunoblots. This 18 kD protein constitutes the structural subunit of the beta pili.

The beta pilus type is significantly associated with virulence in M. catarrhalis, being found with high fre ^¬ quency in virulent strains in a retrospective epidemiolo- gical study. In one study, the distribution of alpha and beta pili types among 43 clinical isolated of colonized and infected patients showed, for alpha pili, 67% and 87% in colonized and infected patients, respectively, and for beta pili, 42% and 81% in colonized and infected pili, respectively. The immunolocalization study described in Example 5 shows that polyclonal anti-PAO pili antisera binds with high affinity to the surface and surface appendages of M. catarrhalis.

IVB. Porphyromonas gingivalis

SHEET

Monoclonal antibody PK99H was found to cross-react with a Porphyromonas (previously referred to as Bacterio- ides) gingivalis 40 kD cellular protein in a number of isolates (obtained from Dr. R. Ellen, Faculty of Den- tistry. University of Toronto, Toronto, Ontario) on the basis of an i munoblot. Briefly, total cellular protein of P. gingivalis colonies cultured anaerobically in an anaerobe jar on BHI agar was solubilized, separated by SDS-PAGE, electrophoretically transferred onto nitrocel- lulose, and immunoblotted with monoclonal antibodies PK99H and PK34C as previously described. PK99H was observed to bind specifically to a 40-50 kD protein depending on the isolate, as seen in Figure 10. The legend for the Western blot is as follows: 1. Bacteroides intermedius, ATCC 25611

2. Bacteroides intermedius, NTCC 9336

3. Porphyromonas (Bacteroides) gingivalis 381

4. Porphyromonas gingivalis 9-14k-l

5. Bacteroides melaninogenicus 20/30 6. Porphyromonas gingivalis 33277

7. Bacteroides melaninogenicus VPI 2381

The lack of reaction with bacteria other than P. gingivalis, and the weak reaction with B. intermedius

NTCC 9336 may be due to low pilus expression by the other strains, rather than lack of reactivity with the PK99H antibody.

IVC. Candida albicans

The binding of C. albicans to TECs and BECs was examined in studies on inhibition of fungal cell attach¬ ment to target epithelial cells. Several C. albicans strains identified in Example 6A were used. BECs and TECs were obtained as described in Example 6B and 6C, respectively. The binding to C. albicans cells to TECs

EET

and BECs can be shown by microscopic methods, such as described in Example 6D. Figure 11 is a phase-contrast photomicrograph showing C. albicans cells ⁽ light cells ⁾ bound to TECs (dark cells) . Quantitative binding of C^ albicans cells to BECs and TECs was also demonstrated by the adhesion assay detailed in Example 6E.

The adhesion protein in C. albicans is a fimbrial protein which forms surface self-polymerized fimbriae on the cell surface of the yeast cells. Fimbriae were ob- tained in substantially purified form by the isolation method detailed in Example 6F. Fractionation of purified fimbriae by sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE) gave the single 64 kD protein shown by silver staining in lane 3 of Figure 12. The purified fimbriae consists of about 15% (w/w ⁾ protein and 85% carbohydrate (w/w) on the basis of colorometric assays.

The ability of purified fimbriae to inhibit Candida attachment to BECs was studied by a direct competition method, as outlined in Example 6H. The results, given in the bar "graph in Figure 13, show that increasing concen ^¬ trations of fimbriae produce increasing inhibition of Candida binding to the epithelial cells. These findings are consistent with the role of Candida fimbrial protein in fungal cell attachment to target epithelial cells.

Similar competitive inhibition studies were carried out with purified P. aeruginosa K pili, with the results shown in the bar graph of Figure 14. As seen, relatively high concentrations of PAK pilin protein inhibited Can- dida binding to BECs in a concentration-dependent manner. Results of the reciprocal study, showing inhibition by Candida fimbriae protein of Pseudomonas pilin protein binding to BECs, are shown in Figure 15.

BSTITUTE SHEET

The inhibition results just discussed indicate that attachment of both Pseudomonas cells through pili, and Candida cells through fimbriae, occur at a common epithe¬ lial cell receptor which is at least partially blocked by either the pilin or fimbrial protein.

These results, like the results obtained above for surface proteins from Pseudomonas, Moraxella, and Bacte¬ roides bacterial cells, suggest a conservation of anti- genic epitopes between Pseudomonas C-terminal pili pep- tide and fungal cell proteins. This conservation of sites is demonstrated by reciprocal competitive ELISA studies on pilin and fimbrial proteins for binding to antibodies specific against the C-terminal pilin peptide (PK99H and PK34C Mabs) and an antibody specific against fimbrial protein. The latter antibody is a polyclonal antibody prepared against purified fimbriae.

Details of the binding method are given in Examples 7A and 7B. Briefly, fimbrial or pilin protein were immobilized on a solid support. The competitor antigen and antibody are mixed together, then added to the solid support, at an antibody concentration such that about 50% of the immobilized antigen would be bound to the anti¬ body, in the absence of the competitor. The amount of antibody actually bound to the solid support was deter- mined by a standard ELISA method. The results of the study are given in Table 3 below. The similar binding affinity values of each of the three antibodies for the two different antigens indicates a strong conservation of epitopic sites between the two antigens.

Table 3 Fim Fim Pili Pili

Competitor Fim Pili Fim Pili

Antibody

In a related study, rabbit polyclonal antibodies prepared against the PAK peptide in oxidized or reduced form were examined for binding affinity to the fimbrial and pili proteins. Antibody was added to the immobilized antigen, either pilin or fimbrial protein, at one of the 4 antibody dilutions shown at the left in Table 4. The amount of antibody bound to the support was assayed by the above ELISA method, with the results given in Table 4. Each polyclonal antibody against the C-terminal, disulfide-linked pili proteins showed high affinity for both pilin and fimbrial protein, again demonstrating a high conservation of epitopes between the two proteins.

Table 4

Anti-Oxidized Ab Anti-Reduced

A Antigen Immobilized Dilution Fim Pili Pirn Pili

10- ¹ >2.0" >2.0 >2.0 >2.0 10- ² >2.0 >2.0 >2.0 >2.0 10 ^" * 0.450 >2.0 0.598 >2.0 10- 0.430 0.496 0.480 1.28

1. The values are ELISA A _«os values.

The ability of the monoclonal antibodies PK99H and

PK34C to cross react with a variety of C. albicans strains was examined by dot blotting, according to the method described in Example 7C. Briefly, cells of a selected Candida strain were immobilized on a nitrocel ^¬ lulose filter, and exposed successively to the PK99H or PK34C antibody, and goat anti-mouse antibody conjugated

to alkaline phosphatase. The amount of antibody bound was measured by color change of a nitro blue tetrazolium substrate. The antibody levels measured are given in Table 5 below. It is evident from the results that all of the Candida strains were highly immunoreactive with the two anti-pili antibodies.

Table 5

Immunospecific binding of PK99H, PK34C, and of polyclonal anti-fimbrial antibody has also been demonstrated by in¬ direct immuofluoresecence, after antibody binding to Can¬ dida cells.

The apparent common binding receptors for Pseudomo¬ nas pili and Candida fimbria and the conservation of epi- topes between the two proteins indicate that the anti-

SUBSTITUTE SHEET

Pseudomonas antibodies would be effective in blocking fungal cell binding to target epithelial cells. This effect has in fact been observed with the PK34C and PK99H antibodies. Figure 16 shows the percent inhibition of Candida binding to BECs after initial exposure of the fungal cells to the antibody indicated. Details of the inhibition method are given in Example 7D. Significant inhibition was seen with both antibodies which are spe¬ cific against Pseudomonas pilin protein C-terminal se- quence.

It will be appreciated from the foregoing that a variety of bacterial and fungal cells have surface pro ^¬ teins which are immunoreactive with antibodies prepared against the Pseudomonas C-terminal pili peptide of the invention. Since binding of these cells to target epi ^¬ thelial cells is inhibited by antibodies prepared against the C-terminal pili peptide, such antibodies and vaccines for their production can be used to prevent and treat infection by the crossreactive microorganisms. Bacteria and fungi which are responsive to such treatment can be readily identified by the methods de ^¬ scribed above, for example, by showing binding of anti¬ bodies prepared against Pseudomonas pili C-terminal peptide to the microorganism, or by showing crossreacti- vity of isolated adhesins to the antibodies.

V. Random-Sequence Antigens

The studies described above demonstrate that an antibody produced against the C-terminal pili peptide of the invention is specific against an epitope present in Pseodomonas pili as well as surface protein present in unrelated bacterial and fungal microorganisms. This finding can be exploited, in accordance with another aspect of the invention, for producing a generalized.

SUBSTITUTE SHEET

random-sequence peptide which contains the epitope common to the different surface proteins. Such a generalized peptide has use in a vaccine composition, for provoking antibodies agaisnt the common epitope (Section VI) , for preparing chimeric antibodies (Section VII), and for therapeutic use in a peptide aerosol method of treatment (Section VIII) .

Methods for generating and identifying generalized, random-sequence peptides having a selected epitopic site have been reported recently (Scott; Cwirla) . Both studies demonstrate that a large population of random- sequence peptides containing random-sequence peptides of 5-10 residues in length can be successfully screened, by immunospecific binding to a selected antibody (or other receptor) for the presence of peptides having a selected binding activity with respect to the receptor molecule. This method is thus effective to generate and identify novel sequences which are predicted to be alternate im unogens for generation of immunity against P_. aerugi- nosa and other microbial species having a common immuno¬ reactive site.

Example 8 describes a method by which random se¬ quence peptides can be prepared and selected for useful¬ ness as immunogens according to the invention. In the preferred method, approximately 10 ⁷ - 10* novel heptapep- tides are generated through construction of an epitope library using the filamentous phage fUSE5 as a vector. Other filamentous phage vectors are considered to be equally efficacious in developing such a library. Alter- natively, similar epitopic libraries can be generated in bacterial expression systems or in mechanically generated peptide systems (Geyson et al. CIBA Foundation Symp. 119: 131-149) .

SUBSTITUTE SHEET

Figure 17 shows schematically the sequence of steps necessary to generate and screen a fUSE5 filamentous phage epitopic library. Briefly, fUSE5 RF DNA is subjec ^¬ ted to digestion with restriction endonuclease Sfil to create an insertion site for insertion of foreign DNA. A synthetic (15+3m) base pair (bp) Bgll DNA fragment is prepared which contains a degenerate sequence of the form ₍ NNK)m, where N represents A, G, C, or T; K represents G or T; and m can vary from 2 to 15. In the preferred em- bodiment of the invention, m ranges from 5-10, typically 6-7, and the bases are randomly added in single addition events to the template primer. An alternative method of achieving random addition of codons coding for the twenty amino acids is to randomly attach trinucleotide codons representing each amino acid to the template primer.

Following ligation of the insert to the cloning vector, amplification of the filamentous phage vector is achieved by transfection of E. coli cells. Successful transfection is measured by the presence of vector borne markers. In the preferred embodiment of the invention, this marker is tetracycline resistance. Recombinant phage are then isolated from bacterial cells. Phage bearing sequences of interest are isolated by an antibody panning method in which phage are incubated with the antibody of interest, e.g., PK34C or PK99H. Biotinylated second antibody (goat anti-mouse IgG) is then added, and complexes containing biotinylated second antibody, anti¬ body PK34C or PK99H, and immunoreactive peptide bearing phage are separated from unreacted antibodies and phage by adhesion onto a streptavidin coated plate. After eluting phage-bearing immunoreactive sequences, the corresponding DNA coding sequences are determined.

The coding sequences corresponding to the selected- epitope peptide(s> are exploited using conventional pep-

TiTU ^■ 6 SH-fcf

tide synthesis methods to produce the epitopic peptide. This may involve solid-phase synthesis, as described in Example 1, or recombinant peptide expressiion according to known methods. Foreign DNA sequences present in the filamentous phage fusion protein pill determine the sequence of the immunoreactive peptide. Peptides discovered to be immu¬ noreactive through this procedure can then be synthesized by standard peptide synthetic methods and prepared as immunogens by conjugation to an appropriate peptide carrier.

VI. Vaccine Compositions

Also included in the invention is a vaccine composi- tion containing a C-terminal Pseudomonas pili peptide and an immunogenic peptide carrier to which the peptide is bound. The composition is used as a vaccine against infection by bacterial and fungal organisms which have surface proteins which are antigenically crossreactive with antibodies produced against the C-terminal, disul¬ fide-linked peptide region of P_. aeruginosa pilin pro¬ tein.

In one embodiment, the peptide includes the se¬ quence: (i) K C T S D Q D E Q F I P K G C S K, (ii) K C T S D Q D E Q F I P K G C S R, (iii) A C K S T Q D P M F T P K G C D N, (iv) T C T S T Q E E M F I P K G C N K P, (V) S C A T T V D A K F R P N G C T D, (vi A C T S N A D N K Y L P K T C Q T A T T T T P, (vii) N C K I T K T P T A W K P N Y A P A N C P K S, (viii T C G I T G S P T N W K A N Y A P A N C P K S, (ix) T C G I T G S P T N W K T N Y A P A N C P K S, and (x) G C S I S S T P A N W K P N Y A P S N C P K S,

SUBSTITUTE SHEET

including amino acid variations which are internally consistent among sequences (i)-(vi) and among sequences ₍ vii ₎ -(x). The peptide is further and characterized by: ₍ a ₎ a disulfide linkage between the Cys (C) residues and (b) immunospecific binding to a PK99H or a PK34C mono ^¬ clonal antibody.

In another embodiment, the peptide includes the sequence:

(i') D E Q F I P K, (ii') D E Q F I P K, (iii') D P M F T P K, (iv') E E M F I P K, (v') D A K F R P N, and (vi') D N K Y L P K, including amino acid variations which are internally consistent among sequences (i')-(vi'). The peptide is further characterized by: (a) immunospecific binding to a PK99H or a PK34C monoclonal antibody; (b) specific bind¬ ing to human buccal or human tracheal epithelial cells; and (c) absence of specific binding to . aeruginosa pili adhesin."

Particularly useful immunogenic carriers include keyhole limpet hemocyanin (KLH) , tetanus toxoid, poϊ ^' y- L- ₍ LYS:GLU), peanut agglutinin, poly-D-Lysine, diphtheria toxoid, ovalbumin, soybean agglutinin, bovine serum albu ^¬ min (BSA), human serum albumin, and the like.

The peptide may be conjugated to the carrier by a variety of known methods, including chemical derivatiza- tion or by standard genetic engineering techniques (e.g., Ausubel) .

Vaccines and inocula o the present invention may be administered by injection, usually intramuscularly or subcutaneously, orally by means of an enteric capsule or

SUBSTITUTE SHEET

tablet, as a suppository, as a nasal spray, and by other suitable routes of administration. For a human patient, a suitable dose of the polypeptide depends, in part, upon the chosen route of administration and a number of other factors. Included among those factors are the body weight of the subject to be immunized, the carrier used, the adjuvant used, and the number of inoculations desired to be used.

Individual inoculations for a human patient typi- cally contain unit doses of about 10 micrograms to about 100 milligrams of polypeptide, exclusive of any carrier to which the polypeptide may be linked. If desired, a series of doses may be administered over a period of time for optimum immunity. Unit dosage forms of the vaccine can also be provided, if desired, containing the afore¬ mentioned amounts of the polypeptide.

VII. Chimeric Antibodies

The present invention also contemplates a chimeric antibody having variable (antigen-reactive) regions which are immunospecific for the C-terminal region of P. aeru¬ ginosa pilin protein, preferably derived from the vari¬ able regions of the above mouse monoclonal antibodies PK99H or PK34C, and constant antibody regions from human immunoglobulin constant regions. The chimeric antibody described here is an IgG antibody, it being recognized that other immunogluobulin types, such as IgM antibodies are also suitable.

Antibody molecules of the IgG class consist of two heavy (H) chains and two light (L) chains, linked toge¬ ther by disulfide bonds as indicated. As shown in Figure 18, the variable regions of the antibody molecule consist of portions of both the heavy and light chain polypep- tides (V _H and V _t ) . Likewise the constant regions of the

SUBST

molecule consist of portions of both the heavy and light chain polypeptides <C„ and C . Therefore, in order to construct a chimeric molecule comprising variable regions derived from mouse monoclonal antibodies and constant regions derived from human IgG, partial genes coding for the appropriate portions of the polypeptides must be joined prior to expression of the polypeptide.

Methods for construction of chimeric mouse-human antibodies by recombinant methods are known in the art ₍ Boulianne; Morrison). Suitable expression systems include, but are not limited to prokaryotic and eukaryotic expression systems known in the art. Prefer ^¬ ably, the expression system is an insect cell ⁽ Spodoptera fru _g iperda ₎ system, which is infected by a recombinant baculoviral vector. Likewise, it will be understood that the recombinant DNA sequences coding for the polypeptide chains of the chimeric antibody can be inserted into separate vectors which are then co-transfeeted into cells, or they can be inserted into the same expression vector. Preferably, the recombinant DNA sequences are sequentially inserted into a coexpression baculovirus- derived vector (pACVC3) which contains two polyhedrin promoters in opposite orientation which drive the tran ^¬ scription of the inserted gene sequences ⁽ zu Pulitz ⁾ . Figure 18 illustrates a preferred construction suitable for use in the baculoviral expression system illustrated. As indicated in the figure, mRNA isolated from a suitable hybridoma cell line, such as cell line PK34C or PK99C, is incubated with a 3' primer to a con- stant region flanking the variable region of interest and incubated with reverse transcriptase. Gene amplification by polymerase chain reaction (PCR) is carried out using a 5' primer selected from a constant flanking region. This procedure, including appropriate primers for the variable

SUBSTIT

region of the mouse heavy chain (V _H ) , is described by Sastry et al.

The gene coding for VH is preferably subjected to digestion by restriction endonucleases BamHI and SacI to produce a BamHI/SacI (5'-3') fragment. A similar proce¬ dure is carried out to obtain the gene fragment coding for the mouse VL region, which is preferably digested with restriction endonucleases BamHI and HindiII to produce a BamHI-HindiII (5'-3') fragment. Likewise, fragments of genes coding for human heavy and light chain constant regions (CH and CL, respective¬ ly) are obtained by methods known in the art (Rabbitts) . The C„ containing gene is preferably subjected to digestion by restriction endonucleases Hindlll and BamHI to produce a Hind III-Bam HI (5'-3') fragment. The C _L containing gene is preferably subjected to digestion by Hindlll and BamHI to produce a Hindlll-BamHI ((5'-3') fragment.

A gene coding for a chimeric heavy chain is then obtained by joining the BamHI-SacIDNA fragment coding for V _H with the Hindlll/BamHI fragment containing human IgGl or IgG2 heavy chain constant region (C _H ) using a Sail linker. A chimeric kappa light chain gene is constructed by joining the BamHI/HindiII fragment containing mouse PK99H or PK34C VL to the Hindlll site of a Hindlll/BamHI fragment containing human C _L (Boulianne et al. (Nature 312: 643-646; 1984); Morrison et al (PNAS 81: 6851-6855; 1984)) .

The resulting recombinant DNAs are then preferably inserted sequentially into the coexpression baculovirus vector pACVC3 sequentially. First the BamHI fragment coding for the chimeric heavy chain (V _H -C _H ) is inserted into the Bglll site of the vector to yield pVH3. The

SUBSTITUTE SHEET

BamHI chimeric light chain gene fragment (V _L -C _t ⁾ is then inserted at a BamHI site of pVH3 to yield pVLH3.

Spodoptera frugiperda cells are infected with recom¬ binant baculovirus pVLH3 (Putlitz et al.) . Binding capa- city of secreted antibodies is analyzed by ELISA as has been described.

The chimeric antibodies produced in accordance with the invention are useful in the treatment or prevention of mammalian infections of Pseudomonas and crossreactive infectious agents, by parenteral administration of the antibodies.

VTII. Peptide Treatment

In one preferred mode of administration, peptides of the invention are delivered by nasal insufflation of pow¬ ders or atomized solutions containing the peptide. This mode of administration has the advantage that delivery of the peptide is made directly to the pulmonary mucosal epithelial surface. Yet another use of the peptides of the invention is as target molecules for drug delivery to pulmonary epi¬ thelial cells. Since the peptides bind specifically to pulmonary epithelial cells, they are construed to be use¬ ful as therapeutic adjuvants in pathological conditions involving the lungs. One such condition is carcinoma of the lung. In one preferred use, the peptides of the in ^¬ vention are conjugated to a photoactivatable chemothera- peutic agent useful in the treatment of lung carcinoma. The drug-peptide conjugate is then administered by nasal insufflation, and the drug is activated by high intensity light delivered through a bronchoscope.

The following examples illustrate methods for prepa¬ ring and using the peptide and antibody of the invention.

SUBSTITUTE SHEET

The examples are intended to illustrate, but not limit, the scope of the invention.

Example 1 Solid-Phase Synthesis of Pilin PAK Peptide

Abbreviations used in this example are BOC, tertiary butoxycarbonyl; DCM, dichloromethane; TFA, trifluoroace- tic acid; and BOC-AA-OH, amino acids protected at the alpha amino group by BOC group. Commercially available phenylacetamidomethyl resin for polypeptide synthesis was obtained from Applied Biosyste s (Foster City, CA) . BOC-AA-OH were obtained from Institute Armand Frappier (Laval, Quebec, Canada) . Side-chain protecting groups on the residues are as follows: o-(p-bromobenzoyloxycarbonyl) for tyrosine, o-benzyl for threonine, serine, aspartic acid and gluta- mic acid; S-methoxy-benzyl for cysteine, 2-chloroben- zyloxycarbonyl for lysine and formyl tryptophane.

A. Solid-phase Synthesis

In preparing a synthetic polypeptide of this inven¬ tion by the above solid-phase method, the amino acid residues are linked to a resin (solid-phase) through an ester linkage from the carboxy-ter inal residue. Reactive amino acid side chains are also protected during synthesis of the polypeptide. Couplings are typi¬ cally carried out using a 2-fold molar excess of protec¬ ted amino acid and one equivalent of dicyclohexyl car- bodiimide over the number of milliequivalents of initial N-terminal amino acid. For asparagine (N) and glutamine

(Q) , 2 molar equivalents of N-hydroxy-benzotriazole and dicyclohexyl carbodiimide were used. Coupling reactions are monitored by the ninhydrin test of Sarin (1981) and are typically more than 99% complete.

B. Oxidation and Purification of the Peptide

The peptide is cleaved from the resin and subse¬ quently cyclized to form a disulfide bond. The cleavage of the peptide and the complete removal of the side-chain protecting groups is accomplished using using anhydrous hydrogen fluoride. The resin is suspended in a mixture containing hydrogen fluoride and anisole (9:1, v/v) and the reaction is allowed to proceed in vacuo for 45 minutes at 5°C. The hydrogen fluoride is then evaporated. The resin is removed and washed with ether ( 3 x 10 ml) and the peptide is extracted with 30% acetic acid (3 x 10 ml) . The combined filtrates are diluted to give a 5% aqueous acetic acid solution and lyophilized.

The crude peptide can be purified on an analytical reversed-phase HPLC column (250 x 4.6 mm internal diame¬ ter) using a shallow gradient. The crude peptide was dissolved in the smallest volume of starting buffer possible (about 5 ml) . The highly concentrated peptide was centrifuged to sediment undissolved material. An analytical sample, 5-10 μl, was chromatographed using a linear gradient (solvent A is 0.05% aqueous TFA and solvent B is 0.05% TFA in acetonitrile) to determine the total amount of peptide present. When the crude peptide contained hydrophilic and hydrophobic impurities with retention times close to that of the peptide of interest in the analytical run (1% B/min gradient rate) , a shallow gradient of 0.2% B/min with a flow rate of 1 ml/min was employed.

The whole stock solution of 30-50 mg was injected onto the column and the run was monitored at 210 nm. Fractions (1 ml) were collected and analysed. Every third or fifth fraction was analysed to identify the region on the chromatogram with the peak of interest. Further analysis of the fractions within this region

would then be carried out. The chromatogram of each run could be compared with the initial analytical run prior to purification to ascertain the peak of interest. In this way, the shoulders of the neighboring peaks were eliminated, while fractions of interest were pooled and freeze dried. Dried peptides were stored in glass vial in a dessicator.

Mass spectrometry and HPLC anlysis were used to confirm the PAK peptide structure.

Example 2 Preparation of Epithelial Cells

A. Buccal Epithelial Cell (BEC) Preparation

BECs were collected from ten healthy non-smoking male volunteers via wooden application sticks rubbed gently on the inside of cheeks, three wooden application sticks per cheek. These sticks were rubbed gently to¬ gether in 30 mL phosphate buffered saline to suspend the BECs. These cells were washed three times with 30 mL phosphate buffered saline by successive centrifugation (650 x g) and resuspended. The final pellet was suspended in 5 mL phosphate buffered saline at pH 7.2. This suspension was filtered (prewetted 70 μm nylon mesh) and the cells were diluted to a final concentration of 2 x 10 ^s cells/mL in phosphate buffered saline at pH 7.2. This suspension was stored at 4°C until ready for use.

B. Tracheal Epithelial Cell Preparation

Human ciliated tracheal epithelial cells (TECs) were obtained from patients in the Surgical Intensive Care unit at Toronto General Hospital by bronchoscopic brush¬ ing of the bronchial mucosa as described by Franklin et al. (1987) . TECs were obtained by bronchoscopy from sur¬ gical patients (under general anesthetic) , intubated in-

tensive care unit (ICU) patients, and health volunteers. For the surgical and ICU patients, bronchoscopy was per¬ formed with a flexible Olympus Type 2 BF bronchoscope inserted through an endotracheal tube. A cytology brush was used to abrade the tracheal-bronchial mucosa, and TECs were collected in high-glucose Dulbecco modified Eagle medium containing 1% (w.v) sodium citrate.

The cell suspension obtained by bronchoscopy con¬ tained both ciliated and nonciliated cuboidal and colum- nar epithelial cells in addition to various amounts of mucus, erythrocytes, granulocytes, and cell debris and was not suitable for direct use in an adhesion assay. The cell suspension was vortexed briefly, sequentially passed through 70- and 30-micron pore size mesh nylon screens, washed twice (500 x g for 15 min at 4°c) with 10 ml of 0.01 M phosphate-buffered saline (pH 7.2) (PBS), and then resuspended in 1 ml of PBS. The cell suspension was then fractionated by density gradient centrifugation (500 x g for 15 min at 4°C in a swinging bucket rotor) on a PBS-preformed (48,000 x g for 40 min at 4°C) 65% (vol/- vol) percoll gradient.

The TEC band was collected and applied to a second percoll gradient. The ciliated TEC band was collected from the second gradient, and the cells were washed once in PBS and then resuspended in 1.5 ml of PBS. A direct cell count was performed with a hemacytometer; cell viability was determined by trypan blue dye exclusion. The cell fractionation procedure typically yielded (2.08 ± 0.34) x 10 ⁵ cells (mean t standard error), of which 32.8 ± 6.5% were ciliated TECs. The vast majority of these cells were viable, and in many cases the cilia were still beating. The fractionated TECs contained only epi¬ thelial cells, were essentially free of contaminating mucus, and were used directly for adhesion assays.

Example 3 Polyclonal Antibody Inhibition of Pilus Binding to BECs

A. Preparation of Fab Fragments

Polyclonal antibodies were prepared against the fol¬ lowing peptide regions of PAK pilin protein: "rl, " :r2," "ol" and "o2, " against the reduced (r) or oxidized (o) PAK peptide composed of residues 128-144 of native PAK pilin protein; "22," against residues 22-33; "41," against residues 41-49; "58," against residues 58-70; "75," against residues 75-84; "89," against residues 89-99; "107," against residues 107-116; and "117," against residues 117-125. "Pre" refers to preimmune sera; and "99H" to monoclonal antibody PK99H.

Fab fragments of the above polyclonal sera derived from each peptide antigen were prepared using immobilized papain (Pierce Chemical Co., Rockford, IL) . Briefly, affinity purified polyclonal antibody was dialyzed against 20 mM cysteine HC1, 10 mM tetrasodium ethylenedi- aminetetraacetic acid (EDTA) in 20 mM sodium phosphate buffer pH 6.2. Antibody (1 ml containing approximately 2 mg antibody) was added to 0.5 ml immobilized papain and incubated at 37°C for 20 h with shaking at 150 rpm. The immobilized papain was removed by centrifugation and the supernatant containing the Fab fragments was diluted with 1 ml of PBS.

The Fab fragments were purified by HPLC using a Protein G column eluted with PBS. Fab fragments were collected in the flowthrough, and Fc fragments were eluted from the column with 10 mM glycine pH 2.75. Fab fragments were concentrated by placing the Fab effluent in dialysis tubing (molecular weight cutoff of < 8000) and extracting liquid from the dialysis sack using poly-

ethylene glycol (molecular weight of 15,000 - 20,000). The fragments were then dialyzed against PBS. Activity of Fab fragments was checked by ELISA and production of Fab fragments was confirmed by SDS-PAGE.

B. Inhibition of PAK pilin Binding to BECs by Fab Fragments

PAK pilin protein was isolated according to pub¬ lished methods (Paranchych et al, 1979) . Fab fragments, prepared as above, were preincubated with PAK pili before the addition of BECs (1 x 105 cells/mL final concentra¬ tion) and pili binding was detected using monoclonal antibody PK38B (which is specific against pilin protein, but not the PAK peptide) . All Fabs were diluted such that their final titre as measured by ELISA to PAK pili was 10 ^"3 .

Example 4 Monoclonal Antibody Inhibition of P. aeruginosa and Pili Binding to BECs

A. Monoclonal Antibodies

Hybridoma cell lines PK99H and PK34C (Doig) were deposited in the cell depository of the Department of Medical and Infectious Diseases of the University of Alberta, Alberta, Canada, and are identified by cell line Nos. PK99H and PK34C. Fab fragment of the PK99H and PK34C Mabs, and non-immune IgG were prepared as described in Example 3A.

B. Inhibition of Pili Binding to BECs

BECs were prepared as described in Example 2. PAK pili were isolated according to published procedures

(Paranchych et al., 1979). PAK pilin protein was iso-

lated according to published methods (Paranchych) . Fab fragments of PK99H, PK34C, and non-immune IgG were prein- cubated with PAK pili at the concentrations indicated in Table 1 above (Doig) . After incubation, BECs were added to a final concentration of 1 x 10 ⁵ cells/mL. Pili bind¬ ing was detected using monoclonal antibody PK3B, followed by reaction with enzyme-labeled goat anti-mouse antibody, as above. Pili binding, as measured by enzyme activity associated with BECs, is expressed as percent control (no Fab fragments added) in Table 1.

C. Inhibition of P_. aeruginosa binding to BECs

P_. aeruginosa strains PAK, PAO, HD1, 492c, PI, K122- 4, and PAK/3 are as reported (Doig) . Fab fragments of PK99H, PK34C, and normal mouse non-immune IgG were pre¬ pared as above. The Fab fragments (0.5 ml of 0 to 3.2 mg/ml) were added to 0.1 ml of bacteria containing 0.7 x 10 ^s CFU/ml in PBS, and the mixture was incubated for 30 minutes at room temperature. To this mixture was then added 0.4 ml PBS and either 1 ml of BECs containing 2 x 10 ⁵ cells or 1 ml PBS (to assess non-specific binding of bacteria to filters. After incubation at 37°C for 2 hours with shaking, the cells were washed, filtered, and the filters were assayed for the presence of bound bacteria by an ELISA method employing an anti-pilin mono¬ clonal antibody and an enzyme-linked antibody, as described above. The results are shown in Table 2 above.

Example 5 Monoclonal Antibody Immunolocalization on Moraxella Cells

Immunolocalization studies employing the polyclonal anti-PAO pili antisera utilized M. catarrhalis cells which were absorbed onto the surface of carbon coated, glow discharged electron microscope grids, blocked with

1% (w/v) BSA in PBS pH 7.4 for 2 5 min, and reacted with rabbit anti-PAO pili in PBS pH 7.4 containing 0.5% (w/v) BSA for 35 min at 37° C. The grids were then washed on 5 drops of PBS pH 7.4 containing 0.5% (w/v) BSA, blocked with l%(w/v) BSA for 2 5 min, reacted with goat anti-mouse IgG-20 n colloidal gold conjugate (E.Y. Labs Inc., San Mateo, California) in PBS pH 7.4 contain¬ ing 0.5% (w/v) BSA for 30 min, washed on 5 drops of PBS pH 7.4 containing 0.5% (w/v) BSA, then washed on 5 drops of H ₂ 0 and finally blotted dry. The grids were then stained with l%(w/v) phosphotungstic acid before examina¬ tion in a Philips EM 401 transmission electron microscope operating at an accelerating potential of 80 kV. Con¬ trols included no first antibody and normal mouse IgG (Jackson Laboratories) . The polyclonal anti-PAO pili antisera was observed to bind with high affinity to the cell surface and surface appendages of the Moraxella catarrhalis.

Immunolocalization studies with monoclonal antibody PK99H were carried out with thin sections of M. catar¬ rhalis cells that had previously been embedded in LX12, sectioned, thin sections collected on the surface of 3mm copper EM grids, and the sections etched with saturated sodium metaperiodate to remove osmium and regain antige- nic activity. The thin sections were then treated as described above except that monoclonal antibody PK99H was used instead of the polyclonal antibody. Immunospecific binding of monoclonal antibody PK99H was observed, with the antibody binding both to cell surface components and to cytoplasmic components in the M. catarrhalis cell.

Example 6 Inhibition of Candida Binding to BECs

A. C. albicans culture conditions

The C. albicans strains used were isolated from the trachea of intubated intensive care unit patients at Toronto General Hospital, Toronto, Ontario (for fimbrial purification and characterization) or obtained from the Department of Microbiology, University of Alberta Hospi¬ tal, Edmonton, Alberta.

A loopful of culture from Sabouraud-dextrose agar (Gibco) was used as a source of inoculum for 10 ml of M9 medium (Adams, 1959) supplemented with 0.4% (wt/v) glu¬ cose. Two incubation protocols were used. Cultures sha¬ ken at 150 rpm were either incubated at 25° C for 19 hours or for 16 hours at 25° C followed by a 3-hour in¬ cubation at 37° C. Cultures to be used in the radio- adhesion assay were supplemented with 55 Ci/ml of [ ³⁵ S]- L-methionine (New England Nuclear, Boston Mass.) after 17 hours of incubation. Cells were harvested by centrifuga- tion (12,000 x g for 10 min) and washed 3 times with PBS pH 7.2 to remove unincorporated methionine. Washed cells were resuspended in PBS pH 7.2 at varying concentrations. Cells which had been incubated at 37° C were forcibly passed twice through an 18 gauge needle to break up clumps. No clumping was observed during the adhesion assays.

B. Buccal epithelial cells

BECs were collected with wooden applicator sticks from healthy, non-smoking, male volunteers (n=10) . BECs were removed from the applicator sticks by gentle agita- tion in PBS pH 7.2. The BECs were washed 3 times (2 000 X g for 10 min at 4oC) with PBS pH 7.2, then passed through a 70 μm nylon mesh. The cell concentration for the BECs was determined with a hemocytometer and the BEC concentration was adjusted. The viability of BECs

obtained by this procedure was generally about 5%, as determined by trypan blue dye exclusion.

C. C. albicans Adherence to Human Ciliated Tracheal Epi- thelial Cells

Human ciliated tracheal epithelial cells (TECs) were obtained as described by Franklin et al. (1987) and by Todd et al. (1989) . Briefly, cells were obtained by bronchoscopy following administration of 5 ml of 1% (w/v) xylocaine to the nasal and oral pharyngeal passages with a further 5 ml of 1% (w/v) xylocaine being administered via the suction port of the bronchoscope at the level of the glottis. An additional 5 ml was further administered within the trachea before repeated (n=10) brushing of the trachea with a disposable cytology brush. Cells were eluted from the brush by agitation in 30 ml of serum free Dubecco's Modified Eagle's Media (high glucose formu¬ lation) containing 1% (w/v) sodium citrate and stored at 40°C before use. TECs were fractionated from mucus, blood cells, microbial contaminants, and debris by gentle filtration through a 70 μm and a 30 μm nylon mesh, washing the cells three times with 10 ml of PBS pH 7.2 (500 X g, at 40C for 10 min) concentrating the cells following centrifugation (500 X g, at 40C for 10 min) in 1 ml of PBS pH 7.2, two sequential density gradient centrifugations on preformed 65% (v/v) Percoll gradients in PBS pH 7.2 (preformed by centrifugation at 48,000 X g for 40 min at 40°C) for 20 minutes at 500 X g for 20 min at 40°C, washed twice with 5 ml of PBS pH 7.2, and resuspended in 1 ml of PBS pH 7.2. TECs were quantitated by direct counting employing a hemacytometer.

D. Immunofluorescence of Candida Fimbriae

Yeast were grown at 25° C or shifted to 37° C as described above. Yeast were harvested by centrifugation.

Cells were fixed with 1.0% formaldehyde in PBS for 30 min and washed twice with PBS. Primary antibody was added and the mixture was incubated at 37° C for 1 hour, shaking at 300 rpm. Yeast were then collected by centrifugation (12 000 x g for 1 min at room temperature) and washed 3 times with PBS pH 7.2. Rabbit anti-mouse IgG (H+L) affinity purified IgG conjugated to fluorescein isothiocyanate (Jackson Laboratories) in PBS pH 7.2

(1/500 dilution) was added to the washed yeast preparations and incubated for 30 min at 37°C, agitating at 300 rpm. The yeast were washed 3 times as described above and resuspended in 0.1 mL of PBS pH 7.2. Wet mounts were prepared, and examined by epifluorescence and phase contrast microscopy using a Lietz Laborlux equipped with a MPS4 camera system. Photographs were recorded with Kodak T-Max film.

E. Adhesion assay

The adhesion assay of McEachran as modified as described by Staddon was used to determine the number of bacteria bound per epithelial cell. BECs (1 mL of 2 X 10 ^s cells per mL) were mixed with an equal volume of radio-labelled yeast suspended in PBS pH 7.2 and incubated at 37°C for 2 h, shaking at 300 rpm. Epitheli¬ al cells with bound yeast were then collected by filtration on 5 micron polycarbonate filters (Nuclepore) pretreated with 2% (w/v) bovine serum albumin (BSA) in PBS pH 7.2 to reduce nonspecific binding, washed with 15 mL PBS pH 7.2 and then placed in scintillation vials. Aquasol (5 mL) was added to each vial and the amount of radioactivity was determined by scintillation counting in

a Beckman LS-150 liquid scintillation counter. Triplicate aliquots were filtered for each sample. Binding of yeast to epithelial cells was corrected for nonspecific binding of yeast to the 12 μm filter (nonspecific binding was generally less than 15% of the experimental value) . The epithelial cell concentration was determined at the end of the assay to correct for cells lost during incubation.

Total and viable cell counts were performed before and after the adhesion assay. Total cell counts were determined using a he acytometer. Viable counts were determined by serially diluting C. albicans in PBS pH 7.2 and plating appropriate dilutions on Sabouraud-dextrose agar (Gibco) which were incubated at 37° C until visible and countable colonies formed (usually 24 to 48 hours) .

F. Purification of Candida Fimbria

C. albicans were cultured in M9 medium (Adams, 1959) supplemented with 0.4% (wt/v) glucose overnight at 37° C at 150 rpm. This culture was used to inoculate Sabouraud- -dextrose agar (Gibco) in aluminum trays (approximately 2 ml/tray) . The trays were incubated at 37° C for 5 days. Yeast cells were then scraped from the surface of the agar with a bent glass rod and suspended in PBS pH 7.2 containing 1 mM phenylmethylsulfonyl fluoride (Sigma) as a protease inhibitor. Fimbriae were then sheared from the cell surface by blending (4 x 15 second cycles using a Waring blender) . Cells were examined by phase microscopy and appeared to be intact. Cells were removed by centrifugation (12,000 x g for 20 min.) and by subsequent filtration of the supernatant through a 0.45 μm polycarbonate filter (Nuclepore Corp., Pleasanton, Calif.). The supernatant was placed in dialy¬ sis tubing (Spectrum, Los Angeles, Calif.; M.W. cut off

6000-8000 Da) and concentrated with polyethylene glycol (M.W. 15,000-20,000, Sigma) (PEG). Finally the sample was dialyzed against PBS pH 7.4. The final preparation was termed crude fimbriae (CF) and was stored at -70° C. Crude fimbriae were purified by HPLC size exclusion chromatography on a Protein-Pak 300 SW column (Millipore Inc.) having a size exclusion limit of 300,000 daltons operating at 0.5 ml/min flow rate and previously equi¬ librated with PBS pH 7.2 buffer containing ImM CaCl ₂ and eluted with the same buffer. Purified fimbriae were obtained by re-chromatographing the material that ini¬ tially eluted in the void volume of the column and col¬ lecting the material that still eluted within the void volume from the second chromatographic run. Purified fimbriae consisted of ~15% (w/w) protein and ~85% car¬ bohydrate (w/w) on the basis of colorimetric assays and consisted of a single polypeptide of ~64,000 daltons on the basis of SDS-polyacrylamide gel electrophoretic analysis utilizing silver staining (see figures) .

G. Effect of Candida fimbriae on PAK pilus binding to BECs

An immunoassay was performed to assess the effect of Candida fimbriae binding of pili from PAK to BECs. BECs (1 mL at 2.0 X 10 ^s BECs/mL) , PAK pili (0.5 mL of 80 ug/mL) , and fimbriae (0.5 mL of 400 or 800 ug/mL) in PBS pH 7.2 were mixed and incubated at 37°C, shaking at 300 rpm in a New Brunswick gyroshaker. After 1 h BECs were collected by centrifugation (12 000 X g for 10 min at 4°C) and washed twice with PBS pH 7.2. Anti-PAK pilus monoclonal antibody PK3B (Doig et al. 1990) was added to the BEC pellet (1 mL of a 10 ^"4 dilution) and incubated as described above for 1 h. The BECs were then collected by centrifugation and washed twice with PBS pH 7.2. Goat

anti-mouse IgG(H+L) peroxidase conjugate (Jackson Labora¬ tories) was added to the BEC pellet (1 mL diluted per instructions for use) and the mixture was incubated as described above for 1 h. The BECs were collected by centrifugation, transferred to a clean test tube, and washed twice with PBS pH 7.2. The pellet was resuspended in 1 mL of a solution containing lmM 2,2'-azinobis[3— ethylbenzothiazoline-6-sulfonic acid), 0.03% (vol/vol) in 10 mM citrate buffer pH 4.2. The reaction was stopped by the addition of 1 mL of 4 mM NaN ₃ and the optical density at 405 nm was determined after removal of the BECs by centrifugation. The BEC concentration in each tube was determined with a hemocytometer at the end of the assay prior to the removal of BECs by centrifugation.

H. Assessment of PAK pilus or Candida fimbriae inhibi¬ tion of Candida binding

Pili and fimbriae inhibition assays were performed using a direct competition method. Direct competition of pili/fimbriae and yeast was achieved by the simultaneous addition of pili/fimbriae, yeast and BECs at the com¬ mencement of the yeast binding assay. The number of yeast bound per BEC was determined as described above.

Example 7

Pseudomonas/Candida Antibody Crossreactivities

A. Enzyme-linked immunosorbant assay (ELISA)

Antigens were coated on NUNC 96-well polystyrene wells. Antigen (10 ug/mL in 0.01 M carbonate buffer, pH 9.5) was added to each well (100 μl/well) and left for 6 h at room temperature. Wells were then washed 3 times with 250 μl of PBS pH 7.4 supplemented with 0.02% (wt/vol) BSA (buffer A) . Wells were blocked with 5%

(wt/vol) BSA in PBS pH 7.4 overnight at 4°C . Wells were washed three times and 100 μl of primary antibody was added for 2 h. Each well was then washed 3 times with

250 μl buffer A using aspiration. A goat anti-mouse IgG (H+L) immunoglobulin-horse radish peroxidase conjugate

(Jackson Laboratories) in buffer A (100 μl/well) was added and incubated for 2 h at room temperature. The wells were washed 3 times with buffer A and 250 51/well substrate solution (lmM 2,2'-azino-di-(3-ethylbenzthi- azoline sulfonic acid), 0.03% (vol/vol) hydrogen peroxide in lOmM sodium citrate buffer pH 4.2) added. The reac¬ tion was stopped by the addition of 250 μl/well of 4mM sodium azide and absorbance at 405 nm determined using an EL-407 plate reader.

B. Competitive ELISA

Competitor and antibody were mixed together in 10 mM PBS pH 7.2 buffer containing 0.05% (w/v) BSA and incuba¬ ted for 30 min at room temperature. The conditions of the assay were such that ~50% of the antigen immobilized on the ELISA plate surface would be bound with antibody if there was no competitor present. The mixture of anti¬ body and competitor was then added to wells (100 μl/well) coated with PAK pili or Candida fimbriae and blocked with BSA as described above. The ELISA was then performed as described above. The apparent affinity of the antibody for the competitor was determined as described by Nieto et al. (1984) following determination of the concentra¬ tion of competitor that would give 50% inhibition of antibody binding to antigen.

C. Whole Cell C. albicans Dot Blots

Dot blots were performed using a Bio-Rad dot blot¬ ting manifold. Whole cells of various clinical isolates

of Candida albicans were initially washed 3 times with PBS pH 7.2 buffer and 100 μl of cell suspension contain¬ ing 2 X 10 ⁶ CFU was added per well on a pre-wetted nitro¬ cellulose membrane. The cells were collected on the surface of the filter by vacuum filtration. The wells were washed 4 times with 0.1% (vol/vol) Tween 20, 50 mM Tris buffered saline pH 7.5 (TTBS) (200 μl/well) and blocked with 100 μl of 3% (wt/vol) BSA in TTBS for 1 h. The wells were then washed 4 times with TTBS and mono- clonal antibody PK99H and PK34C at various dilutions in TTBS were added (100 μl/well) and incubated for 1 h. The blot was washed 4 times with TTBS and 100 51 of a goat anti-mouse IgG (H+L) immunoglobulin-alkaline phosphatase conjugate in TTBS was added to each well for 1 h. After washing the blot 6 times with TTBS, a substrate solution consisting of 0.33 mg/mL nitro blue tetrazolium chloride, 0.165 mg/mL 5-bromo-4-chloro-3-indolyl-phosphate, 100 mM sodium chloride, 5 mM magnesium chloride, 100 mM Tris buffer pH 9.5 was added. Color development was stopped by aspiration and rinsing the membrane in distilled water.

D. Assessment of antibody Fab fragments on yeast binding: Effect of Fab fragments on binding was performed as follows. Fab fragments (0.5 mL of 800 μg/mL in PBS pH 7.2) were added to 0.1 mL of yeast in PBS pH 7.2 and incubated for 30 min at room temperature. To this 0.4 mL of PBS pH 7.2 and either 1.0 mL of BECs (2 X 10 ⁵ cells/mL) or 1.0 mL of PBS pH 7.2 was added. The mixture was then incubated at 37°C shaking at 300 rpm for 2 h. The remainder of the adhesion assay was performed as described above.

Example 8 Preparation of Random-Sequence Peptide

A. Construction of epitope library for peptide ligands An epitope library which approximately 10 ⁸ novel heptapeptide sequences is constructed as described by Scott and Smith. Alternatively a similar epitope library can be constructed as described by Cwirla et al or Devlin et al. Figure 17 shows a schematic representation of the construction of the library as described by Scott and Smith and summarized here.

Filamentous phage fUSE5 is constructed as a vector for the epitope library as described by Parmley and Smith

(1988; Gene 73: 305-318) and by Scott and Smith. This phage contains a tetracycline resistance gene, and is designed to have cloning sites, insertion into which result in addition of peptide sequence at the exposed N- terminus of the minor capsid protein pill. Addition of foreign peptide sequence at this site does not substan- tially inhibit infectivity of the phage. However its location does make it exposed on the surface of the phage and thus amenable to recognition by antibodies.

In preparation for ligation with a foreign DNA fragment, fUSE5 is digested with Sfi I (BRL; 120 units/30 ug fUSE5 RF DNA) . Following extraction with phenol and chloroform, the volume is adjusted to approximately 0.8 ml with TE buffer (10 mM Tris pH 8, lmM EDTA), and the

DNA is precipitated by additions of sodium acetate buffer

(3M; pH 6) and isopropanol. Following incubation (20 min at 0 degrees) and pelleting, the pellet is washed with

70% (v/v) ethanol, redissolved in TE buffer, ethanol precipitated and redissolved in TE.

An insert containing a random heptapeptide sequence is prepared by polymerase chain reaction (PCR) amplifica-

tion of a 73 base degenerate template shown in Figure 17 with 5' biotinylated primers corresponding to the first 20 bases at the 5'ends of both strands shown in Figure 17_(top) . The template is then cleaved at the two Bgll sites shown and then adsorbed onto streptavidin-agarose to remove the biotinylated terminal fragments along with the undigested and partially digested by-products. The PCR mixture contains 1 μg of template, 5 μg of each of the biotinylated primers, and 25 units of AmpliTaq DNA polymerase (Perkin-Elmer/Cetus) .

The mixture is subjected to five temperature cycles (2.5 min. at 95 degrees, 4 min. at 42 degrees, 4.4 min. at 72 degrees, 5 min. at 72 degrees) ,then the reaction is stopped by addition of EDTA solution (final concentration, 1 mM) , pH 8. Following precipitation with ethanol and dissolution in 0.1 ml TE buffer, a portion of the product is digested with Bgl I (Promega; 6.4 units/ul final concentration) for 2 hours at 37 degrees to produce a 36 bp degenerate fragment. The randomly selected codons in this fragment are represented by (NNK) ₇ , where N stands for an equal mixture of deoxynucleotides G, A, T, and C, and K stands for an equal mixture of G and T. M stands for an equal mixture of C and A in the complementary strand. NNK therefore represents an equal mixture of 32 triplets, including codons for the 20 amino acids plus the amber stop codon. The digestion by Bgl I is stopped by addition of EDTA solution (10 mM, final concentration) . Pre-washed streptavidin beads are then added to the solution and mixed for 30 minutes, and removed by centrifugation. This step is repeated with fresh steptavidin beads. The final product is extracted (phenol plus chloroform) and evaporated to a volume of 0.1 ml.

Ligation of the insert to the Sfil digested fUSE5 RF is carried out in a volume of 2 ml. The reaction mixture consists of 36 μl of the degenerate 36 bp insert and 10 ug of the Sfi I digest of fUSE5. The product is extrac- ted with phenol and chloroform, ethanol precipitated, and dissolved in 0.2 ml TE buffer.

The ligation product is electroporated into E. coli MCI061 cells for amplification of phage carrying degene¬ rate peptide sequences. Following electroporation, the bacteria are diluted into growth medium containing tetra- cycline. Tetracycline resistance of cells indicates successful transfection.

B. Selection of Antibody-Binding Peptide Phage are isolated from plate stocks by scraping from the agar surface, resuspending in L broth and clear¬ ing the supernatant twice by centrifugation (8000 rpm for 10 minutes in a Beckman JA10 rotor at 4 degrees) . Phage particles are precipitated with polyethylene glycol (3.3%) in 0.4 M NaCl, followed by centrifugation. Phage pellets are resuspended in TBS buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl) and stored at 4 degrees.

Phage are affinity purified using the monoclonal antibodies PK34C or PK99H in a panning procedure. Alternatively, FAb fragments of these antibodies are used in this procedure. Briefly, after incubation of phage (lO^-lO ¹² infectious particles) overnight with 1 ug of purified antibody, phage expressing peptides with af¬ finity for the antibodies or their Fab fragments are isolated using the panning method of Parmley and Smith. Biotinylated goat anti-mouse antibodies are added to the mixture, and the mixture is then added to a streptavidiin coated plate. This procedure can also be carried out without use of the second (goat) antibodies, by directly

biotinylating the primary antibodies of Fab fragments. Following incubation (10-30 minutes) the streptavidin- coated plate is washed, and adherent phage are eluted for 10 minutes in a buffer containing 0.1 M HC1 (pH 2.2, adjusted with glycine) and bovine serum albumin (1 mg/ml) . Neutralization of the eluate is achieved using an aliquot of 2 M Tris. Eluted phage are then amplified by infection of E.coli, followed by incubation on agar plates containing tetracycline. The resulting amplified phage pool is repurified twice by the same panning proce¬ dure described above.

Phage selected through 2-3 rounds of panning are cloned and propagated, and their DNA's are sequenced using standard techniques to determine the amino acid sequences of their peptide epitopes.

Although preferred embodiments of the invention are described herein in detail, it will be understood by those skilled in the art that variations may be made thereto without departing from the spirit of the inven¬ tion or the scope of the appended claims.