ANTIBODY RECOGNIZING ENDOTHELIAL CELL LIGAND FOR LEUKOCYTE CR3

This invention was made with government support under Grant Number AI23459 awarded by the National Institutes of Health. The Government has certain rights in the invention.

RELATED APPLICATION

This application is a continuation in part of copending application serial number 695,613 filed May 3, 1991 the disclosure of which is incorporated herein by reference.

BACKGROUND OF THE INVENTION

Filamentous hemagglutinin (FHA) is a 220 D, non-fimbrial surface associated protein produced and secreted by Bordetella pertussis (BP) . It is a necessary factor for BP to adhere to ciliated respiratory epithelial cells during whooping cough, Tuomanen et al, J. Infec. Pis. 152, 118-125 (1985). FHA has also been shown to interact with the integrin complement receptor 3 (CR3) on macrophages and other leukocytes, Relman et al, Cell 61, 1375-1382 (1990). CR3 is also known as Kac-l , o ^f and CD llb/CD18. The portions of FHA responsible for these two binding properties are distinct.

BP binds to two cell types during infection: leukocytes and cilia. Adherence to cilia depends on recognition of carbohydrates such as galactose containing glycolipids. Tuomanen et al, J. Exp. Med. 168:267-277(1988) .

The organism binds to leukocytes by two means. For the first, it binds to the integrin CR3, a step which precedes entry of the bacteria into the leukocyte, as discussed in more detail below. For the second, BP binds to leukocyte carbohydrates. BP also binds to cilia by recognizing a carbohydrate analogous to that on leukocytes. Galactose is the minimum requirement for a carbohydrate receptor. It is found, for example, in such blood group determinants as Lewis a.

There are two aspects of a BP infection. One is the invasion of the leukocytes which takes place when BP binds to the integrin of leukocytes. It is a protein/protein interaction. The other, adhesion of BP to the leukocyte or the cilia through a protein/carbohydrate interaction.

The FHA gene of BP has been sequenced, Relman et al, Proc. Natl. Acad. Sci. USA 86, 2637-2641 (1989) and Domenighini et al, Molee. Micro. 4,487-800 (1990) , and a number of expression products have been produced, Delisse-Gathoye et al. Infect. Immun. 58, 2895-2905 (1990) .

All of the above publications and any other publications identified in this disclosure are incorporated herein by reference.

FHA and the integrin on leukocytes interact in a protein-protein recognition event. The interaction between FHA and leukocyte integrin involves recognition of the arginyl-glycyl-aspartyl sequence at positions 1097 to 1099 in FHA. This sequence will hereinafter be identi ied as the RGD sequence or

simply, RGD, and any region of FHA or FHA segment on which it occurs as the RGD region. R, G and D are the standard one letter abbreviations for arginine, glycine and aspartic acid.

There are two natural molecules which bind to the integrin CR3 in a manner which appears to involve recognition of RGD. One molecule is the serum complement component C3bi, an opsonin. The other is the x-molecule(s) on endothelial cells.

It has long been known that leukocytes can invade or pass through vascular endothelial tissue by a process in which integrins, such as CR3, bind to receptors for integrins on the surface of endothelial cells as a step in a sequence of reactions which results in a widening of the junctions between such cells to permit passage by the leukocytes. The receptors for CR3 on endothelia are referred to herein as x-molecules. There may be one or more than one such -molecule.

In addition to the ability to recognize C3bi and endothelial cells, CR3 also binds to Factor Ten of the coagulation cascade (Altieri et al, Science 254:1200-1202, 1992). The coagulation cascade is involved in inflammation since procoagulant activity arises on endothelial cells during infection or other noxious stimuli. Three regions of Factor Ten participate in recognition of CR3 and all three have homologous regions in FHA.

To summarize, FHA appears to represent a molecule with many binding regions for CR3, each region representing a natural analog in a eukaryotic

protein. FHA or its regions can serve an anti- inflammatory function by acting as competitive inhibitors of CR3 recognition of endothelial cells (one region) , C3bi (one region) , or Factor Ten (four regions) . Another separate and distinct region is the carbohydrate recognition domain (CRD) which allows the bacteria to adhere to cilia. The CRD and segments and analogs thereof can serve as a vaccine against whooping cough, particilarly if they are dissociated from the other domains of FHA which will generate toxic cross reactive antibodies if included in vaccines because they are homologs of human proteins.

For convenience in discussing these various properties of FHA, each region is defined according to the scheme in Figure 1. Regions A, B, C and D are defined by antibodies as in Delisse-Gathoye et al, Infect Immun 58:2895-2905, 1990 and are bounded by the amino acid residues as shown. Region A contains the CRD. Region B contains the domain mimicking C3bi. Regions consistent with Factor Ten-like domains are represented by *. The region spanning the RGD (residues 1097-9) mimics the endothelial cell ligand for CR3 which participates in leukocyte transmigration.

SUMMARY OF THE INVENTION

It has been discovered that leukocytes use the same receptor, the integrin CR3, to bind to FHA, C3bi, Factor Ten and the x-molecule of endothelial cells. This establishes that FHA and these molecules are functionally related. They are also antigenically related and therefore may be

structurally related in that FHA contains the amino acid triplet RGD and the x molecule may contain the RGD triplet or a segment which acts like the RGD triplet. As a result, antibodies to FHA cross react with endothelial cells. Similarly, FHA contains 4 regions with sequence similarity to Factor Ten. The specific leukocyte integrin involved in the binding is believed to be CR3, otherwise known as Mac-l, CDllb/CD18. Moreover, there is species cross reactivity. Thus antibodies to FHA raised in a goat, mouse, guinea pig or human will react with and bind to C3bi, Factor Ten or endothelial cells of rats, rabbits and humans. There may be one or several distinct molecules on endothelial cells which function as x-molecules by binding anti-FHA antibodies.

There are a number of significant consequences of this important discovery. These are:

1. Peptides which contain or mimic the RGD region, a Factor Ten region or the

C3bi region of FHA will bind to the integrin of leukocytes, thereby preventing adherence of the leukocyte to the endothelial cells as a procedure for inhibiting the inflammatory process.

2. Antibodies to FHA will bind to the homologous eukaryotic proteins disturb their function. In the case of the C3bi-like region, antibodies will bind C3bi or other related complement components and render them less

effective for opsonization. In the case of the Factor Ten like regions, the antibodies will disturb coagulation and prevent amplification of inflammation by the coagulation cascade. In the case of the RGD region, antibodies to FHA will bind to endothelial cells of, for example, the blood brain barrier and selectively open the junction between the cells in a manner analogous to the opening of the endothelia during leukocyte diapedesis thereby making possible the entry of desirable therapeutic agents into the cerebrospinal fluid of the subarachnoid space and into the brain parenchymsa without silumtaneously admitting the entrance of leucocytes.

3. Peptides containing or simulating the CRD of region 1141-1279 are optimal vaccines for whooping cough because they generate antibodies which block adherence of bacteria to the respiratory tract and thereby prevent disease. Elimination or modification of the other regions of FHA is important in optimizing the vaccine in order to prevent generation of antibodies which will cross react with natural CR3 ligands such as Factor Ten,

C3bi, or endothelial cells.

4. Peptides of each of the domains of FHA are useful in vaccine quality control. They can be used to detect the ability of a vaccine candidate to generate antibodies in serum reactive with the endothelial cell. Factor Ten like domains or the C3bi domain. Such antibodies would be deemed toxic.

Those skilled in the art will recognize that there are three fundamental procedures for taking advantage of the discovery upon which this invention is based. One involves the utilization of FHA or regions of FHA or antibodies thereto to prevent CR3 functions in inflammation. These are exemplified by procedures 1 and 2. In procedure 3, antibodies to FHA or to regions of FHA prevent adhesion of BP to the respiratory tract. Procedure 4 may be employed to detect toxic vaccines.

This invention is directed to peptides derived from FHA, analogs of such peptides and antibodies to such peptides which are capable of inhibiting binding between CR3 and its natural ligands. It is also directed to pharmaceutical compositions containing the products and to therapeutic use of such products to inhibit or prevent such adhesion and to other uses which flow from these basic properties. The invention relates also to genes which may be used in accordance with known techniques to produce the products employed in the invention.

The understanding of this invention will be facilitated if certain of the terms used in the description thereof are defined.

The term "peptide" is used in the most generic sense to include those peptides which contain only a few amino acid residues, e.g. 5 or more. It also includes polypeptides containing 20 or more amino acids and proteins such as mutants and segments of FHA. The term encompasses mutants of FHA and segments of FHA as well as mutants of FHA and segments thereof including addition, deletion and substitution mutations. The term is, perhaps, best understood from the function the peptides perform as discussed herein. These functions include inhibition of binding between the CR3 integrin and its natural ligands, Factor Ten, C3bi, or an x-molecule on endothelial cells. They may do so because they contain an RGD triplet or a moiety which mimics an RGD triplet. They may also function by mimicking C3bi or Factor Ten peptides. These peptides are useful in control of inflammation. Peptides of the CRD are useful as nontoxic vaccines.

The term "antibodies" encompasses antibodies which bind to C3bi (or related complement components) Factor Ten or x-molecules on brain endothelium both polyclonal and monoclonal antibodies. The preferred antibody is a monoclonal antibody. The term antibody is also intended to encompass mixtures of more than one antibody (e.g., a cocktail of different types of monoclonal antibodies) . The term antibody is further intended to encompass whole antibodies, single chain antibodies, chimeric antibodies comprising portions from ore than one species, chimeric proteins comprising a functional portion of antibody coupled by covalent or reco binant techniques to an intact protein or functional portion thereof that is not of antibody origin (i.e. a chimeric antibody-protein),

bifunctional antibodies, biologically functional fragments of the aforementioned, etc. Biologically functional antibody fragments which can be used are those fragments sufficient for binding of the antibody fragment to FHA or the x-molecule.

The chimeric antibodies can comprise portions derived from two different species (e.g., human constant region and murine binding region) . The portions derived from two different species can be joined together chemically by conventional techniques or can be prepared as a single contiguous protein using genetic engineering techniques. DNA encoding the proteins of both portions of the chimeric antibody can be expressed as a contiguous protein.

The term "RGD region" includes any RGD containing segment of a molecule whether that molecule is one of high molecular weight such as FHA, or the x-molecule, or a peptide of relatively low molecular weight. There are molecules which mimic the RGD region and these may be involved in the practice of this invention. Native molecules or chemically treated molecules or derivatives are included in this definition.

The term ''block the RGD region" will be used to describe peptides which inhibit adhesion between leukocytes and endothelial cells for any of a number of reasons. They may do so because they bind to all three of the amino acid residues of RGD or to only one or two of them. They may also do so because they bind to molecular segment(s) neighboring the RGD region in such a manner that one or all of the RGD residues is prevented from reacting ;ith an integrin.

The term "Factor Ten like region" refers to any segment of a molecule containing sequences GΪDTK EDG, IDRSMKTRG or GLYQAKRFKVG or highly homologous sequences.

The antibodies useful in the practice of this invention can be raised against whole bacteria containing FHA or derivatives or against native FHA or chemically treated FHA or derivatives, and this will be the procedure usually employed to produce polyclonal antibodies. They can also be raised to mutants of FHA or fragments or segments of mutants of FHA thereof produced either by genetic manipulation of the FHA gene or by chemical or enzymatic cleavage of the whole protein or a fragment of the protein. They can be produced, and this is the preferred method, by raising monoclonal antibodies to the expression products from segments of the FHA gene or mutants of such segments according to the method of Delisse-Gathoye et al, supra. The expression products are proteins or peptides containing a sufficient number of amino acid residues to elicit an antibody response alone or when attached to other antigenic determinants. The antibodies can be produced in accordance with well known procedures for the production of monoclonal antibodies. These monoclonal antibodies are the preferred therapeutic agents for the practice of this invention.

The presently preferred monoclonal antibodies are 12.5D1, 12.1 F9, 13.6E2, 12.6F8, 12.2 Bll and 12.5A9 produced by the methods described by Delisse- Gathoye et al., supra. A preferred polyclonal

antibody is goat antiserum to FHA produced by the Cowell procedure as described in Tuomanen et al. cited above.

The presently preferred antiinflammatory products of this invention are peptides containing from about five to about forty-five amino acid residues, preferably about twelve to thirty five and their analogs, especially those shown in Figs. 2A and 2B. These are preferred because they are relatively small molecules which can be readily prepared in pure form by chemical synthesis. It will be apparent as the description of the invention proceeds that the invention is not limited to these peptides, but these presently appear to be preferred.

The presently preferred vaccine products of this invention are the peptides shown in Fig. 3 and their analogs.

A peptide will be useful in this invention in either of two cases; 1) if it is capable of reducing or inhibiting adhesion between leukocytes and endothelia or of raising antibodies having these properties, or 2) it is capable of inhibiting adhesion between BP and cilia or of raising antibodies having these properties.

It will be clear from this definition that peptides of this invention may contain amino acid derivatives comprising single or multiple amino acid additions, deletions and/or substitutions compared to the amino acid sequence of FHA. The peptides contemplated herein may be chemically synthesized such as by solid phase peptide synthesis or may be

prepared by subjecting the designated polypeptide to hydrolysis or other chemically disruptive processes whereby fragments of the molecule are produced. Alternatively, the peptides could be made in vitro or in vivo using recombinant DNA technology. In this case, the peptides may need to be synthesized in combination with other proteins and then subsequently isolated by chemical cleavage or the peptides may be synthesized in multiple repeat units. Furthermore, multiple antigen peptides could also be prepared according to the method of Tam et al, J. Am. Chgm. Soc. 105 6442-6445 (1988) . The selection of a method of producing the subject peptides will depend on factors such as the required type, quantity and purity of the peptides as well as ease of production and convenience.

The use of these peptides in vivo may first require their chemical modification since the peptides themselves may not have a sufficiently long serum and/or tissue half-life. Such chemically modified peptides are referred to herein as "analog". The term "analog" extends to any functional chemical equivalent of a peptide characterized by its increased stability and/or efficacy in vivo or in vitro in respect of the practice of the invention. The term "analog" is also used herein to extend to any amino acid derivative of the peptides as described herein.

Analogs of the peptides contemplated herein include, but are not limited to, modifications to side chains, incorporation of unnatural amino acids and/or their derivatives during peptide synthesis and

the use of crosslinkers and other methods which impose conformational constrainst on the peptides or their analogs.

Examples of side chain modifications contemplated by the present invention include modifications of amino groups such as by reductive alkylation by reaction with an aldehyde followed by reduction with NaBH.; amidination with ethylaceti idate; acylation with acetic anhydride; carbamoylation of amino groups with cyanate; trinitrobenzylation of amino groups with 2, 4, 6, trinitrobenzene sulfonic acid (TNBS) ; acylation of amino groups with succinic anhydride and tetrahydrophthalic anhydride; and pyridoxylation of lysine with pyridoxal-5'-phosphate followed by reduction with NaBH 4..

The guanidino group of arginine residues may be modified by the formation of heterocyclic condensation products with reagents such as 2,3- butanedione, phenylglyoxal and glyoxal.

The carboxyl group may be modified by carbodiimide activation via O=acylisourea formation followed by subsequent derivatisation, for example, to a corresponding amide.

Sulfydryl groups may be modified by methods such as carboxymethylation with iodoacetic acid or iodoacetamide; performic acid oxidation to cysteic acid; formation of a mixed disulphides with other thiol compounds; reaction with maleimide, maleic anhydride or other substituted maleimide; formation of mercurial derivatives using 4-

chloromercuribenzoate, 4-chloromercuriphenylsulfonic acid, phenylmerσury chloride, 2-chloromercuric-4- nitrophenol and other mercurials; carbamoylation with cyanate at alkaline pH.

Tryptophan residues may be modified by, for example, oxidation with N-bromosuccinimide or alkylation of the indole ring with 2-hydroxy-5- nitrobenzyl bromide or sulphenyl halides. Tyrosine residues on the other hand, may be altered by nitration with tetranitromethane to form a 3- nitrotyrosine derivative.

Modification of the imidazole ring of a histidine residue may be accomplished by alkylation with iodoacetic acid derivatives or N- carbethoxylation iwth diethylpyrocarbonate.

Examples of incorporating unnatural amino acids and derivatives during peptide synthesis include, but are not limited to, use of norleucine, 4-amino butyric acid, 4-amino-3-hydroxy-5-phenylpentanoic acid, 6-aminohexanoic acid, t-butylglycine, norvaline, phenylglycine, ornithine, sarcosine, 4- amino-3-hydroxy-6-methylheptanoic acid, 2-thienyl alanine and/or D-isomers of amino acids.

Crosslinkers can be used, for example, to stabilize 3D conformations, using homo-bifunctional crosslinkers such as the bifunctional imido esters having (CH ₂ ) spacer groups with n=l to n=6, glutaraldehyde, N-hydroxysuccinimide esters and hetero-bifunctional reagents which usually contain an amino-reactive moiety such as N-hydroxysuccinimide and another group specific-reactive moiety such as

maleimido or dithio moiety (SH) or carbodiimide (COOH) . In addition, peptides could be conformationally constrained by, for example, incorporation of C and N_x--methylamino acids, introduction of double bonds between C and C atoms of amino acids and the formation of cyclic peptides or analogs by introducing covalent bonds such as forming an amide bond between the N and C termini, between two side chains or between a side chain and the N or C terminus.

The peptides of the invention or their analogs may be single length or in tandem or multiple repeats. A single type of peptide or analog may form the repeats or the repeats may be composed of different molecules including suitable carrier molecules.

The present invention, therefore, extends to peptides or polypeptides and amino acid and/or chemical analogs thereof corresponding to regions of the RGD domain or the CRD.

BRIEF DESCRIPTION OF THE FIGURES

Fig. 1 shows regions of FHA which mimic human proteins that bind to CR3 with the mapping relative to the CRD.

Figs. 2A and 2B are representative of peptides which will block leukocyte adherence to endothelia or interfere with coagulation.

Fig. 3 is a representation of peptides suitable for vaccines.

Figs. 4 - 6 illustrate the use of FHA, peptides or antibodies of the invention as antiinflammatory agents.

Figs. 7 and 8 illustrate the use of antibodies to FHA to enhance vascular permeability.

Fig. 9 is a representation of Fragment 7 of FHA as defined by Delisse-Gathoye et al, supra.

Fig. 10 shows the sequence for the gene encoding Fragment 7 and for Fragment 7 itself. Sequence ID No. 1.

Fig. 11 shows deletions of the FHA gene which will result in truncated FHA molecules useful for vaccines.

Fig. 12 shows the relative ability of anti-FHA antibodies to bind to C3bi.

Fig. 13 and 14 refer to Examples 5 and 6 respectively.

DETAILED DESCRIPTION OF THE INVENTION

The utilities which may be realized from the recognition that FHA, C3bi, Factor Ten and the x- molecule are structurally and functionally related and the recognition of the important consequences mentioned above will now be discussed.

1. FHA and peptides which contain or mimic the RGD region of FHA or the Factor Ten region of FHA will bind to integrins of leukocytes to inhibit the inflammatory process.

Leukocytes, such as monocytes, and polymorphonuclear (PMN) leukocytes (PMN) , circulate in the blood and do not adhere to the endothelium. Upon the introduction into the tissue of (1) an infectious agent, (2) fragments that result from the death of an infectious agent, or (3) another inflammatory substance, leukocytes, such as PMN are induced to bind to the endothelium and then migrate into the tissues. This is a two step process in which the leukocyte initially binds to receptors on the endothelium, including the x-molecule. One effect of the binding is that the cell junctions in the endothelium open. This permits the leukocyte, in the second step, to move from the x-molecule, through the junctions and into the tissue. Since PMN can recognize and kill many infectious agents, the passage of leukocytes through the endothelium and into the tissue is a protective mechanism. However, in many disease circumstances, leukocytes react in an exaggerated and deleterious fashion. They may bind so avidly to endothelium as to occlude blood flow. Once in the tissues, they secrete proteases, reactive oxygen intermediates, and other toxic molecules which not only kill infectious agents, but also can result in extensive tissue damage. In addition, they trigger release of inflammatory mediators that alter vascular tone and permeability, and that recruit additional leukocytes to the site, thus perpetuating inflammation.

As stated above, FHA binds to CR3 leukocytes through three domains RGD, Factor Ten like regions and C3bi like regions for which the integrin on the leukocyte surface acts as a receptor. When it does

so, it prevents the leukocytes from using the integrins to adhere to the endothelium and initiating the inflammatory process.

This process is schematically illustrated for the x-molecule in Fig. 4. The figure shows a portion of the endothelium constructed from adjacent endothelial cells with apposed cell junctions. The sketch shows leukocytes in the blood stream together with FHA or segments thereof.

The left side of the Figure illustrates the normal adhesive reaction between an integrin such as CR3 on a leukocyte and the RGD or RGD mimicking region on x-molecule of the endothelium. The result of the reaction, as shown by the arrow, is that a cell junction opens and the leukocyte moves from the RGD into the tissue.

The right side of the sketch shows that the presence of FHA or peptides acting like FHA in the blood stream prevents this reaction because they react with the integrin and prevent it from binding to the x-molecule. This has the effect of preventing the interaction of the x-molecule and the leukocyte integrin so that the leukocyte cannot pass through the endothelium into the tissue.

The agent which will achieve this desirable result may be native FHA or a segment thereof. Typically, it will be a relatively low molecular weight peptide containing one of two motifs; 1) an RGD region or an RGD mimicking region or 2) a Factor Ten like region. It may contain, for example from about 5 to 20 amino acid residues or even more. One

example of such a peptide is described by Relman et al in the Cell article cited above. It is Thr-Val- Gly-Arg-Gly-Asp-Pro-His-Gln. Other examples are shown in Fig. 2. Fig. 7 illustrates that intravenous administration of FHA in an experimental model of meningitis decreases inflammation as evidenced by lower numbers of leukocytes in cerebrospinal fluid.

An infection in which leukocyte mediated damage contributes to morbidity and mortality of disease is bacterial meningitis. Depending upon the infecting organism, thirty percent of the cases of meningitis per year die despite sterilization of the infection by antibiotics. Over fifty percent of the survivors have permanent and severe sequelae such as paralysis, deafness, and learning disabilities. Obviously, the prevention and/or diminishment of such damage would greatly enhance the quality of life for the survivors of this disease.

Activated leukocytes also contribute to cerebral edema and blood-brain barrier injury. Neutropenic animals (animals in which the leukocytes have been artificially diminished) have been found to have improved survival rates in experimentally induced disease. A high amount of inflammation in the subarachnoid space correlates directly with a poor outcome of disease. Inhibition of the accumulation of leukocytes in cerebrospinal fluid directly correlates with improved morbidity and mortality of experimental pneumococcal meningitis and of Haemophilus influenzae meningitis and bacteremia in children.

Clearly, an agent which would inhibit the influx of leukocytes into infected sites would be a therapeutic tool of immense value particularly if non-leukocyte mediated defense systems are left functionally intact. It would further be beneficial to block leukocyte diapedesis only at inflam ed sites and not at other sites throughout the body. Thus treatment directed at inflammed endothelia would be advantageous over that directed to leukocytes.

The use of antibiotics magnifies the deleterious effects of inflammation during infectious diseases. This is due to the mechanism by which such agents exert their antiinfective effects. For example, following the administration of a beta-lactam antibiotic (or another cell-wall directed antibiotic) , the bacteria disintegrate due to lysis by the antiinfective agents. The resulting fragments of bacteria initiate a dramatically enhanced inflammatory response. Earlier research has indicated that inhibition of this enhanced level of inflammation correlates with improved morbidity and mortality, Tuomanen et al., J. Infect. Pis., 155, 985-990 (1985) and Kadurugamuwa, Program and Abstracts of the 27th ICAA Meeting, p. 205 (1987) . In penumococcal meningitis, for instance, mortality can be directly correlated with the amount of meningeal inflammation, McAllister et al., J. Infect. Pis. , 132, 355-360 (1975). Thus, a method of dampening inflammation during the course of therapy with an antibiotic would be advantageous in treating infections, particularly meningitis, septic arthritis, and endophathalmitis.

The process of this invention will be useful in treating inflammation caused by any of a variety of infective agents, including gram-positive and gram- negative bacteria as well as viruses and fungi. Particularly targeted infections are those which are susceptible to treatment with beta-lactam antibiotics, or antiviral agents such as Haemphilus influenzae B; N. meningitidis b; pneumococci, e.g.. Streptococcus pneumoniae; Escherichia coli; Staphylococcus epidermidus; Staphylococcus aureus; group B Streptococci; Salmonella; Bacillus subtillis; Pseudomonas aeruginosa; and Herpes virus.

The infected tissue which is the target of the present invention can likewise be tissue in any body site susceptible to inflammation caused by the above-described infective agents. The method of the present invention is, however, particularly adaptable to the treatment of infected tissue of the central nervous system, lung, kidney, joints, endocardium, eyes and ears, with the treatment of the cerebrospinal fluid and articular fluid being highly preferred embodiments.

One particularly susceptible tissue for which the present invention is uniquely suited is the tissue of the central nervous system. The vascular endothelium in the brain is morphologically different from that in other tissues in that endothelial cells are joined by tight junctions thereby creating a blood-brain barrier which prevents molecules the size of proteins from passing from blood into the cerebrospinal fluid. They are particularly useful in the treatment of meningitis infections, including

those arising from pneumococci, Haemophilus influenzae B, N. meningitidis b and Escherichia coli, group B Streptococcus and Staphylococci.

Additionally, the ingress of leukocytes into articular fluid can be prevented by administration of a therapeutic amount of a selected anti-FHA antibody of the present invention. In cases where the inflammation of an infection migrates to the joints, e.g. arthritis, this method can be utilized to alleviate the inflammation by preventing the ingress of leukocytes into the articular fluid.

The process of this invention is useful in the control of inflammation arising from substantially any source including, for example autoimmune disease.

A further method of the present invention is that of reducing or eliminating inflammation in an infectious disease caused by the administration of an antiinfective agent for that disease which comprises the simultaneous administration of an effective amount of antiinfective agent and an effective amount of an anti-FHA antibody or an active fragment thereof to a patient in need of such therapy. Pue to the mechanism of their therapeutic activity, antiinfective agents, and particularly beta-lactam antibiotics, cause additional inflammation as a result of their therapeutic effect. Although such antiinfective agents sterilize a given infection, they cause release of the cell wall and/or endotoxin of the infecting agent. Such bacterial components initiate an inflammatory response in the infected

tissue. It is this inflammation which contributes significantly to the tissue damage that is the long- term consequence of most infections.

The term "simultaneous administration" as used herein means that therapeutic amounts of the antiinfective agent and the anti-FHA antibody are administered within a time period where they influence each other. Thus the anti-infective agent may be administered at the same time or before or after the antibody.

Reduction or elimination of inflammation in infectious diseases results in a diminution of the neurological damage that usually accompanies such infections. Since the antibodies of this invention possess the unique ability to block movement of leukocytes across the blood-brain barrier, they are uniquely suited to treat infections where the causative infective agent is Haemophilus influenza B, N. meningitidis b, or a pneumococci such as Streptococcus pneumoniae. Such infections are generally treated with an aminoglycoside such as gentamicin or a beta-lactam antibiotic such as a penicillin or cephalosporin.

Examples 5 and 6 illustrate the ability of the antibodies of this invention to assist in the penetration of the blood brain barrier by penicillin.

Pue to the ability of FHA to reduce or eliminate inflammation in an infectious disease caused by the administration of an anti-infective agent, FHA or active fragment thereof can be combined in a single unit dosage form with the anti-infective agent for

convenience of administration. Such dosage form is most preferably an intravenous dosage form since most anti-infective agents, particularly the beta-lactam antibiotics, are available in a suitable chemical form for administration via the intravenous route. This is also the preferred route of administration for FHA or its peptides of the invention. Typically, the anti-infective agent and one or more antibodies can be combined in a single ampoule solution. Where this is not possible, the anti-infective agent and the peptide can be packaged separately and mixed just prior to injection. Administration can likewise be via a mixture with any standard intravenous solution, i.e., normal saline.

The amount of anti-infective agent in the dosage form is dependent upon the particular anti-infective agent being utilized and the particular infection being treated. The amount of the peptide utilized in dosage form can range from about 1 to about 1,000 mg, with 10-100 mg per dosage unit being highly preferred. Dosages can be administered one to four times daily, with continued therapy for as long as the infection persists.

This process is also applicable to targeting therapeutic agents to leukocytes. The desired therapeutic agent, for example an anti-leukemic agent, an immunomodulator or other known drug may be bonded to a selected peptide by any selected process, and it will be carried to the leukocyte by the peptide because the peptide binds to CR3. CR3 is restricted to leukocytes and thus limits distribution of peptide linked agents to leukocytes. Furthermore,

ligands bound to CR3 initiate ingestion of the ligand so the peptide linked agent could be delivered to and then taken up by the leukocyte.

2. Antibodies to FHA may by employed to block inflammation or induce blood brain barrier permeability.

Since FHA contains domains that resemble C3bi, Factor Ten and the x molecules of endothelial cells, antibodies to FHA bind to these natural molecules and can be used to disturb their function. The data and description for antibodies binding to C3bi is reported in section 4. This section will illustrate two phenomena:

2a) Antibodies to Factor Ten domains block leukocyte recruitment, the preferred antibody being 12.1B11.

2b) Antibodies to the RGD region bind to endothelial cells and enhance permeability of the blood brain barrier, the preferred antibody being 13.6E2.

2a) . Antibodies to FHA which block leukocyte recruitment, i.e, 12.1B11.

During an inflammatory process the coagulation cascade is activated by the expression of coagulation factors on the surface of an endothelium. This leads to a net pro-coagulant state at the endothelial surface and promotes the deposition of fibrin and clotting. These processes result in the localized thrombotic events characteristic of advanced

inflammation and contribute to tissue damage by occluding blood flow leading to tissue anoxia. Factor Ten is a coagulation cascade component and it interacts with CR3 on leukocytes to promote the association of leukocytes with cells which harbor procoagulant proteins on their surface, such as an endothelium. Factor Ten has three regions which bind to CR3 as shown by the ability of three peptides to competitively inhibit the binding of CR3 bearing tissue culture cells to purified Factor Ten. These three peptides are GYDTKQEDG (366-373), IDRSMKTRG (422 to 430) and GLYQAKRFKVG (238-246) as described by Altieri et al. Science, supra.

FHA contains four regions of significant sequence similarity to these regions of Factor Ten. They are shown in Figure 2B. Based on this sequence similarity, FHA has antiinflammatory activity as documented in Fig. 5. This similarity makes it clear that antibodies to FHA which bind to these regions would disturb leukocyte recruitment. This is to be distinguished from anti-FHA antibodies that bind to endothelia by recognizing an x molecule which is an endothelial cell component (see below) . In the case of anti-FHA antibodies which bind to the Factor Ten like regions, the antibodies do not bind directly to endothelial cell components but rather to the coagulation components captured on the endothelial cells during inflammation. This is illustrated by the activity of anti-FHA antibody 12.1B11 which does not bind to capillaries but is highly anti¬ inflammatory when administered intravenously into rabbits with meningitis. This is shown in Figure 6. Various doses of antibody were given to rabbits and the number of leukocytes recruited into the

cerebrospinal fluid in response to pneumonoccal meningitis was determined. Those animals receiving even low doses of antibody 12.1B11 showed inhibition of leukocyte accumulation in cerebrospinal fluid.

2b) . Antibodies to the RGD region bind to endothelial cells and enhance blood brain barrier permeability

The antibodies of this invention bind to and block the RGD or RGD mimicking region of endothelial cells including those of the blood brain barrier (BBB) and facilitate the passage of water soluble molecules including, for example, therapeutic agents through the BBB and into the cerebrospinal fluid (CSF) .

Anti-FHA antibodies apparently bind to molecules on capillary endothelial cells known as endothelial cell ligands (to distinguish them from the FHA receptors found on leukocytes) . These EC ligands have determinants that are exposed on the vascular surface of endothelial cells. This feature is inferred from the ability of intravenously administered anti-FHA antibodies to bind to vessels as measured immunohistochemically. These EC ligand determinants do not require stimulation by cytokines in order to be expressed or incorporated on the vascular surface of capillary endothelial cells. Antibodies to ICAM-1 do not appear to bind to these EC ligands. This indicates that the EC ligands are not ICAM-1. Immune blot analysis of the proteins from purified human cerebral capillaries with anti-

FHA antibodies indicates that these antibodies bind to two novel polypeptides of apparent molecular size of 64 and 52 kilodaltons.

When the anti-FHA antibody binds to EC ligands, an apparently transient opening of the BBB to therapeutic agents results. That is, in response to intravenous administration of anti-FHA antibodies, the penetration of therapeutic agents into the brain is enhanced in a time dependent and reversible manner. In addition, binding of intravenously administered anti-FHA antibodies to EC ligands results in an inhibition of leukocyte diapedesis into the brain even though the BBB permeability to therapeutic agents is increaesd as a result of such binding. These features of anti-FHA antibodies will now be illustrated.

It is a particular advantage of the invention that certain antibodies within its scope permit such passage without concurrent penetration of leukocytes into brain or cerebrospinal fluid.the CSF.

The BBB is a continuous boundary between the blood and both the interstitial fluid (IF) and the CSF of the brain. It is composed of a layer of endothelial cells, the cerebral capillary endothelium, that serves as an effective barrier against the entry into the brain's tissue of serum components of both high and low molecular sizes. The restriction against entry of such substances into the brain and the CSF is due to the structure of the cerebral capillary endothelium in which the anatomically tight junction seals spaces between adjacent endothelial cells.

In a normal (healthy) state, the only substances capable of traversing the BBB to enter the CSF tend to be relatively hydrophobic (lipid-like) . Substances which are hydrophylic (water-soluble) penetrate the BBB much less effectively or not at all. Such water-soluble and poorly penetrating substances encompass a whole range of molecules extending from molecules as large as albumin to ions as small as sodium.This poor permeability of BBB by many potentially useful drugs poses a severe limitation on the treatment of diseases of the brain tissue and CSF. It is therefore of paramount clinical significance to develop products and methods which would "open" the BBB and allow access to the brain tissues and CSF by agents which are known to be effective in treating or diagnosing brain disorders but which, on their own, would not be able to traverse the BBB. Certain antibodies of this invention will achieve that end.

This invention provides a method for increasing the permeability of the blood-brain barrier of a host to a molecule present in the host's bloodstream. The host can be any animal which possesses a central nervous system (i.e., a brain). Examples of hosts include mammals, such as humans and domestic animals (e.g., dog, cat, cow or horse), as well as animals intended for experimental purposes (e.g., mice, rats, rabbits) .

The molecule in the host's bloodstream can be exogenous to the host. For example, it can be a neuropharmaceutical agent which has a therapeutic or prophylactic effect on a neurological disorder. Examples of neurological disorders include cancer

(e.g., brain tumors), Autoimmune Deficiency Syndrome (AIDS) , epilepsy, Parkinson's disease, multiple sclerosis, neurodegenerative disease, trauma, depression, Alzheimer's disease, migraine, pain, or a seizure disorder.

Classes of nueropharmaceutical agents which can be used in this invention include antibiotics, adrenergic agents, anticonvulsants, nuσleotide analogs, chemotherapeutic agents, anti-trauma agents and other classes of agents used to treat or prevent a neurological disorder. Examples of antibiotics include amphotericin B, gentamycin sulfate, pyrimethamine and penicillin. Examples of adrenergic agents (including blockers) include dopamine and atenolol. Examples of chemotherapeutic agents include adriamycin, methotrexate, cyclophosphamide, etoposide, carboplatin and cisplatin. An example of an anticonvulsant which can be used is valproate and an anti-trauma agent which can be used is superoxide dismutase. Nucleotide analogs which can be used include azido thymidine (AZT) , dideoxy Inosine (ddl) and dideoxy cytodine (ddc) .

The molecule in the host's bloodstream can also be diagnostic imaging or contrast agents. Examples of diagnostic agents include substances that are labelled with radioactivity, such as 99-Tc glucoheptonate.

The administration of exogenous molecules and/or antibody to FHA to the host's bloodstream can be achieved parenterally by subcutaneous, intravenous or intramuscular injection or by absorption through a bodily tissue, such as the digestive tract, the

respiratory system or the skin. The form in which the molecule is administered (e.g., capsule, tablet, solution, emulsion) will depend, at least in part, on the route by which it is administered.

The administration of the exogenous molecule to the host's bloodstream and the administration of antibody to FHA can occur simultaneously or sequentially in time. For example, a therapeutic drug can be administered orally in tablet form while the intravenous administration of the antibody is given 30 minutes later. This is to allow time for the drug to be absorbed in the gastrointestinal tract and taken up by the bloodstream before the antibody is given to increase the permeability of the blood- brain barrier to the drug. On the other hand, the antibody can be administered before or at the same time as an intravenous injection of the drug. Thus, the term "co-administration" is used herein to mean that the blood-brain barrier permeabilizing antibody and the exogenous molecule will be administered at times that will achieve significant concentrations in the blood for producing the simultaneous effects of increasing the permeability of the blood-brain barrier and allowing the maximum passage of the exogenous molecule from the blood to the cells of the central nervous system.

Figs. 7 and 8 illustrates the process by which the antibodies of this invention will enhance permeability of the BBB to water soluble therapeutic agents. The term "therapeutic agents" is used herein as a convenient term to define all of the various materials which a physician or veterinarian will wish to pass through the BBB into the CSF or the brain.

It includes, for example anti-infective agents such as antibiotics, antineoplastic agents, diagnostic agents, imaging agents, and immuno-suppressive agents, nerve or glial growth factors and other such products.

Figure 7 shows anti-FHA antibodies in the blood stream adhering to and blocking the RGD region in the x-molecule of endothelial cells. By analogy to the first step in the passage of leukocytes through the endothelia, the blocking of the x-molecule has the effect of opening the cell junctions as shown to the left of the figure. The cell junctions, however, are not open to leukocytes because, as explained above, in order to pass through the opening in the cell wall, a leukocyte must first adhere to the cell by reaction with the x-molecule. It is prevented from so doing because the x-molecule is blocked by the anti-FHA antibody. Figure 8 shows that in the example of Figure 7, other molecules, e.g. the soluble therapeutic agents described above can pass through the open junctions.

Example 1 describes the binding of polyclonal anti-FHA antibodies to the BBB.

Example 2 illustrates this same property for monoclonal antibodies and further teaches that the antibody must bind near the RGD region in a manner like mAb 13.6E2. This property is clearly distinguished from the antibodies 12.5A9 and 12D1 which block the carbohydrate recognition. It also teaches that the antibody 13.6E2 has a preferred

affinity for cerebral vessels as opposed to peripheral vessels exemplified by umbilical vein cells.

Example 3 illustrates that the 13.6E2 antibody functions by the mechanism of Figs. 7 and 8 since it blocks entry of leukocytes into CSF at the same time as passage of serum proteins into CSF is enhanced.

Example 4 illustrates that the 13.6E2 antibody enhances blood brain barrier permeability in otherwise healthy animals when injected intravenously.

Examples 5 and 6 show the BBB permeability enhancing properties of 13.6E2 for a therapeutic agent 3H labelled penicillin. The effect is time and dose dependent.

The antibodies of the invention can also serve as carriers for targeting therapeutic agents to endothelial cells in mammals. For this purpose, the therapeutic agent will be chemically bonded to the antibody and the combined product administered to the patient in eed of such treatment. These therapeutic agents include, for example coagulation cascade modifiers or immunomodulators such as cytokines. They may include also immunotoxins such as Pseudomonus exotoxin A or ricin attached to an anti- FHA antibody of the invention to produce products capable of killing tumors supplied by or involving vascular endothelium. Procedures for combining such therapeutic agents with proteins such as antibodies are well known.

In certain patients, a potential problem with the use of a murine anti-FHA monoclonal antibody, such as those employed in this invention exists since the patient may generate an immune response against a murine monoclonal antibody. This effect may be ameliorated or obviated by using active fragments of the monoclonal antibody so as to minimize the amount of foreign protein injected. Another alternative is to employ genetic engineering techniques to make a chimeric antibody in which the binding region of the murine anti-FHA antibody is combined with the constant regions of human immunoglobin.

3. Peptides consisting of the CRD region of FHA may function as nontoxic vaccines.

Several lines of evidence indicate that a CRD exists in FHA and that interference with its function by inhibitors or antibodies decreases colonization of the lung by B. pertussis in an animal model.

1) Inhibition of adherence of B. pertussis to human ciliated cells can be achieved by soluble receptor analogs (lactosamines) or anti-carbohydrate antibodies (anti-Lewis a) (Tuomanen et al, J Exp Med supra) .

2) FHA binds to lactosylceramide on thin layer chromatography plates containing ciliary extracts

(Tuomanen et al, J Exp Med supra) .

3) Lactose and antibody to Lewis a decrease colonization of rabbit lung with virulent B. pertussis (Saukkonen et al, J Exp Med 173:1143-1149).

This CRD has been found to lie between amino acids 1141 and 1279 in the FHA sequence by three kinds of experimental evidence.

1) One antibody to FHA, 12.5A9, which binds to this region blocks bacterial adherence to ciliated cells. Antibodies to other regions of FHA do not.

2) The DNA sequence from position 3674 to 4088 (bounded by Xhol sites) which corresponds to the CRD was amplified by the polymerase chain reaction and ligated into one of the pET expression vectors (Rosenberg, et al (1987) Gene, 56, 125-135). Expression of the polypeptide in E. coli utilized the T7 RNA polymerase promoter system (Tabor, S. &

Richardson, C.C. (1985) Proc. Acad. Natl. Sci. U.S.A., 82,). [ 35S] Methionine labeled protein preparations from cell lysates showed a band at the expected size of 18kD when run on SDS polyacrylamide gels. Cell lysates containing the expressed protein were overlaid on thin-layer chromatography plates containing glycolipid standards and the binding pattern was compared to that of FHA. The CRD region protein bound to these carbohydrates in a pattern similar to FHA, especially in the ability to bind to lactosylceramides. Cell lysates not containing the expressed protein did not bind lactosylceramide.

3) B. pertussis mutants lacking the CRD region did not bind to ciliated cells or macrophages. Several mutants of B. pertussis were created which produce a truncated form of FHA that lacks the CRD region. A 0.4 kb Xhol-Xhol fragment which encompasses the DNA sequence corresponding to the CRD region was deleted from the gene for FHA thus

creating an inframe deletion. This was accomplished genetically with a plasmid vector designed for gene replacement of an unmarked allele. Mutants were created in either a wild type (BP536) or ptx- (BP Tox6) background. Colony immunoblots of bacteria containing the truncated FHA gene do not react with antibody 12.5A9 which is specific for the CRD region. Antibody binding was determined by standard procedures. The following controls were evaluated:

A truncated form of FHA was readily detected by Western blot analysis of culture supernatants from the mutant strains of B. pertussis. The FHA from these strains migrated with a slightly smaller apparent molecular weight than whole FHA. The truncated form of FHA did not cross react with antibody 12.5A9 to the CRD region but did cross react with antibody 12.6F8 to another separate and distant region of FHA. These mutants also produced the same amount of FHA as the parental strain.

B. pertussis mutants producing the truncated form of FHA failed to bind to Macrophages and Ciliated Cells. The FHA mutants were tested in a ciliated cell adherence assay as described in Tuomanen et al J Infec Pis supra. Binding for these mutants was not detectable; the parental strain BP536 bound 4 BP/cell.

A second adherence test involved the binding of

•7

FHA mutants to macrophages. Approximately 10 fluorescein-labeled bacteria were incubated with 10 macrophages and examined under the microscope for evidence of binding. The mutants bound in the range of 60-80 per 100 macrophages as opposed to the wild type's binding of 300 per 100 macrophages.

These studies indicate the CRP of FHA lies in the region 1141 to 1279. Taken together with the efficacy of antibodies to the receptor for this region in blocking colonization of the lung in animal models and the efficacy of antibodies to this region in blocking adherence of bacteria to cilia, this region constitutes an immunogen which would generate antibodies effective in protecting against whooping cough. It would have several advantages over current whole cell or FHA-containing vaccines. It is well known that pertussis vaccines are toxic including reactions such as death and encephalophathy. Vaccines containing the entire FHA molecule engender antibodies which are cross reactive with endothelial cells, C3bi and Factor Ten which can contribute to these toxic reactions. Presentation of a vaccine which contained no toxic epitopes is preferred. Furthermore, generation of antibodies that block adherence, such as the antibodies generated by the

CRP as an immunogen, would block colonization of the respiratory tract, a desirable property not characteristic of present vaccines. This vaccine can be formulated by presenting the CRD alone or in combination with other proteins or other segments of FHA which have been chemically or genetically altered to eliminate generation of cross reactive antibodies. Testing for this property will be described below. The preferred peptides for vaccines are shown in Fig. 3.

Pelisse-Gathoye et al., supra have described a number of expression products of the FHA gene. One of them. Fragment 7 is defined by the Bam HI site at position 2837 and the Pst 1 site at position 6581 as shown in Fig. 9 which also shows other segments of the fraction defined by other restriction enzymes. This region (Fig. 10) expresses an FHA segment which contains the RGD region at the positions corresponding to positions 1097 to 1099 of FHA. Fragment 7 also contains at least one carbohydrate recognition site. This site lies between amino acid residues 1141 through 1279. The presence of such site is established by the fact that antibodies to Fragment 7 such as 12.5A9 will react with FHA and prevent adherence of FHA to cilia or purified glycoconjugates. The Xhol-Xhol segment of Fragment 7 , therefore, will be a suitable peptide for producing a vaccine in accordance with this invention as will be segments thereof lacking the RGD region byt containing a carbohydrate recognition regions or amino acid sequences mimicking such regions. Such segments may be contained in FHA mutants.

Alternatively, truncated FHA's which delete the RGD region may be produced genetically as shown in Fig. 11.

Antibodies that block the function of the CRD, such as 12.5A9 or antisera generated with CRD vaccine candidates detailed above constitute prophylactic or a therapeutic agents for whooping cough. These antibodies when delivered to the respiratory epithelium decrease colonization by interfering with adherence of the bacteria to cilia and promoting clearance.

4. Detection of Toxic Antibodies elicited by Vaccines

It will be apparent to the skilled artisan that some of the same antibodies to FHA elicited by vaccines to protect against BP infections as explained in Item 2 above may also (a) cross-react with the x-molecule of endothelia thereby to open the endothelia to the passage of serum components or (b) bind complement components such as C3bi so that they are not available to participate in inflammation. Both activities might be regarded as toxic reactions for a product intended for use as a vaccine. Accordingly, it is preferred for the production of vaccines to select peptides generating antibodies which will block the carbohydrate dependent interactions of BP with mammalian cells but will not react with the x-molecule and open cell junctions. Such peptides will be selected from amongst the peptides of the invention so as to eliminate regions

of FHA mimicking the eukaryotic molecules but preserving the carbohydrate recognition site. Examples are shown in Fig. 3.

The peptides of this invention particularly those mimicking eukaryotic molecules as listed in Fig. 2 or REgion B of FHA will be valuable for quality control procedures in the production of BP vaccines as well as other vaccines, for example antiviral or antibacterial vaccines to eliminate components which generate antibodies which cross- react to endothelium, C3bi or Factor Ten. Such peptides may be employed to test for the presence of antigens in vaccines which will generate toxic antibodies. For example, anti-pertussis vaccines based on FHA or FHA fragments may contain molecules which will generate antibodies reactive with the x- molecule, or an analogous molecule on the endothelial surface and open the cell junctions to penetration by unwanted substances or otherwise generate an inflammatory response. The technique is illustrated in Examples 1 and 2.

To test for such potential toxicity, the candidate vaccine antigen will be employed to immunize an animal such as a rabbit. The antiserum from the immunized animal will be used to overlay human brain tissue slices or another surface coated with an RGD, or RGD mimicking peptide. Binding of the testing antibody to the tissue section, or the peptide indicates that the vaccine antigen produces antibodies which cross react with the x-molecule and that the vaccine will be toxic.

Conversely, the antibodies in a serum to be tested can be bound to a surface of a plate coated with Protein A. Addition of C3bi coated particles will elad to capture of the particles if the serum contains anti C3bi antibodies. Serums with a high capture capacity indicate the toxicity of a vaccine. This test is illustrated in Example 8. Its application to anti-FHA monoclonal antibodies is shown in Fig. 12.

Any of a variety of tests may be employed to detect the binding of toxic antibodies. Typical tests include radioimmunoassay, enzyme linked immunoassay, as well as, direct and indirect immunofluorescence. These tests may employ competitive and sandwich type assays. Typically, the tests will employ detectable labels on an indicator antibody. Useful labels include fluorescent labels such as fluorescein, rhodamine or auramine. Radiosotopes such as 14C, 131I, 125I and 25S may be employed. Enzyme labels which may be utilized include, for example, s-,y -glucamidase,- ₇ _ -D- glucosidase,^_5-D-galactosidase, urease, glucose oxidase plus peroxidase, and acid or alkaline phosphatase. Methods for labeling biological products such as cells, antibodies, antigens and antisera are well known and need not be described.

There are several currently available procedures for detecting these labels including, for example colorimetric, spectrophotometric, fluorospectrophotometric, photometric and gasometric techniques, as well as various instrumental methods of detecting isotopes.

All of these tests involve the formation of a detectable reaction product between the indicator antibody and the test toxic antibody which will react with the RGD or analogous region on endothelial cells or 3Cbi and is generated by an immunological response to a toxic antigen in a vaccine. The indicator antibody, i.e. the labeled antibody, may react directly with the toxic antibody as in the enzyme linked immunoassay procedure (ELISA) or other sandwich type test.

The peptides of the invention may be provided as parenteral compositions, for example vaccines, for injection, infusion, or other parenteral procedures, such compositions comprising a prophylactically effective amount of the selected peptide and a pharmaceutically acceptable carrier. They can, for example be suspended in an inert oil, or in alum or other suitable adjuvant. Alternatively they can be suspended in an aqueous isotonic buffer solution at a pH of about 5.6 to 7.1. Useful buffers include sodium phosphate-phosphoric acid.

The desired isotonicity may be accomplished using sodium chloride or other pharmaceutically acceptable agents such as dextrose, boric acid, sodium tartrate, propylene glycol or other inorganic or organic solutes. Sodium chloride is preferred particularly for buffers containing sodium ions.

If desired the solutions may be thickened with a thickening agent such as methyl cellulose. They may be prepared in emulsified form, either water in oil or oil in water. Any of a wide variety of pharmaceutically acceptable emulsifying agents may be

employed including, for example acacia powder, or an alkaryl polyether alcohol sulfate or sulfonate such as a Triton.

The therapeutically useful compositions of the invention are prepared by mixing the ingredients following generally accepted procedures. For example, the selected components may be simply mixed in a blender or other standard device to produce a concentrated mixture which may then be adjusted to the final concentration and viscosity by the addition of water or thickening agent and possibly a buffer to control pH or an additional solute to control tonicity.

The dosage and method of administering the peptides of the invention may be varied due to patient condition, the result sought to be acheived and other factors readily evaluated by the attending physician or veterinarian. For example, while the presently preferred method of administering peptides comprising vaccines of the invention is parenteral administration, certain of them will be administered orally in a pharmaceutically acceptable oral carrier. The antigenic peptides thus administered will generate antibodies in the lymph nodes of the intestine. These antibodies will be distributed systemically to produce a prophylactic effect.

The following non-limiting examples are given by way of illustration only.

Example 1

Rabbit and goat polyclonal anti-FHA antibodies bind to cerebral capillary endothelia.

Cerebral capillary endothelial cells harvested from rabbit brains or cryostat sections from human brains were mounted onto glass slides (detailed method appears after the table) . Preparations from at least two different individuals were incubated with the antibody at 1:20 dilution (or greater) at room temperature for 2 hrs, rinsed, and incubated for 30 min with either a Vector biotinylated secondary antibody (Vector Elite ABC Immunohistochemistry Kit) for human specimens or a fluresceinated anti-Fc antibody for rabbit specimens. After rinsing, rabbit specimens were viewed with a fluorescence microscope; for human specimens, the Vector avidin-peroxidase mixture was applied for 30 min, followed by rinsing and application of the substrate. The stained tissued preparaitons were viewed under a light microscope. (+ indicates detectable staining comparable to control) .

The results are shown below.

TABLE

Cross reactivity antibodies to B. pertussis antigens with cerebral capillaries of mammalian brain.

Rabbit Human

Goat antisera to: native purified FHA native purified pertussis toxin nd

Rabbit antisera to: native purified FHA glutaraldehyde-treated

FHA 0 0

Human antisera: pooled cord serum 0 0 pre-immune (n=8) 0 0 post-primary DPT (n=2) + + post-infection (n=5) 3 of 5+ 3 of 5+ post-infection absorbed with BP 0 pertussis immune globulin

(PIG) + +

PIG absorbed with BP 0 0

Controls: anti-human transferrin receptor nd rabbit, goat or horse serum 0

Details for Table

Human brain samples were quick frozen 1-3 hrs post mortem; those of animals were frozen at the time of serifice. For preparation of human and animal tissue slices, crebral cortex was cut in 20u sections; for some experiments, rabbit cerebral capillaries were extracted from homogenized tissues by centrifugation in 15% dextran. Capillaries or tissue slices were fixed onto glass slides with acetone at room temperature for 10 min and stored at -80°C until used. Antisera were diluted into phosphate buffered saline (PBS) at 1:20. Rabbit capillary staining was visualized with a 1:40 dilution of fluorescenated anti-Fc antibody to the appropriate speices (Boeringher Chemical, Indianapolis, IN). For tissue slices, the Vector ABC Elite immunohistochemistry kit for peroxidase was used to visualize capillary staining. Brain slices were incubated at room temperature in the following: PBS (10 min), 0.3% peroxide/methanol (30 min), PBS (10 min), 0.25% a-methyl mannoside/0.25% nonfat dry milk in PBS (15 min) , PBS (10 min) . Primary antibody was then applied either at room temperature for 2 hrs orat 4°C for 16 hrs. Slices were rinsed for 10 min in PBS and then treated with the following at room temperature: Vector biotinylated secondary antibody (30 min) , PBS (10 min) , Vector avidin-peroxidase mixture (30 min) , PBS (10 min) , peroxidase substrate until color was seen (2-7 min) . Purified monoclonal antibodies to human or rat transferrin receptor were always used as a positive control and all observed experimental staining was graded with respect to the positive. Negative controls included goat serum (because the biotinylated anti-

rabbit antibodies are made in goat) and horse serum (because the biotinylated anti-rabbit antibodies are made in horse) .

Example 2

The binding of anti-FHA antibodies to cerebral capillaries involves recognition of the RGD region of FHA.

Monoclonal antibodies were used undiluted as supernatant fluids. Staining of human cerebral capillaries was detected as in Example 1. Each antibody was tested at least two times at 24 C and at 4°C. Staining of capillaries was accompanied by staining of large vessels. The number of + indicates relative degree of staining; antibodies were tested on samples from at least two humans, nd = not determined. Staining of cultured human umbilical vein endothelial cells was performed using peroxidase-anti-peroxidase technique Muller et al, J. Exp. Med. 170:399-404, 1989. The ability of antibodies to bind C3bi was tested in an ELISA assay in which wells were coated with antibodies (10 ug/ml) , washed and then incubated for 30 min with erythrocytes coated with C3bi (made as described in Example 8) . Captured C3bi-coated erythrocytes were detected visually. The ability of antibodies to block binding of FHA to carbohydrates was performed in an overlay assay as described using glycolipid standards separated by thin layer chromatography [Tuomanen, 1988#260].

The results are shown below.

TABLE

Ability of monoclona anti-FHA antibodies to cross react with endothelium, C3bi and inhibit interaction of FHA with carbohydrates in vitro.

Bind Inhibit FHA- C3bi carbohydrate

nd

+ + nd +

+++

nd nd

Example 3

Antibody to FHA blocks influx of leukocytes into rabbit cerebrospinal fluid. g

Rabbits were inoculated intracisternally with 10 pneumococci and the generation of an inflammatory response in cerebrospinal fluid (CSF) was followed over time by measuring the appearance of leukocytes and protein in CSF. Animals were treated with the antibodies intravenously (2mg/kg of culture supernatant fluid) at the same time as the intracisternal challenge was given.

leukocyte number* protein** (x 10 ³ /ml CSF) (mg/dl)

5 hrs 6 hrs

no antibody 112 3661 6538 1.9 2.2

*normal_120; **normal _jr= l.0

+ significantly different from control

These results are interpreted to show that:

1. Intravenous anti-FHA mAB 13.6E2 significantly decreases the influx of leukocytes into CSF of rabbits challenged with an inflammatory amount of pneumococci. Antibody to pertussis toxin does not. Accumulation of protein in CSF was augmented in antibody treated animals indicating increased blood brain barrier permeability. This latter observation is consistent with Example 4.

Example 4

Antibody to FHA enhances permeability of the cerebral capillary endothelium.

Rabbits were injected intravenously with antibodies and the influx of protein into the CSF was followed over time. Such influx occurs only when the permeability of the cerebral capillary endothelium is enhanced. As shown in Example 3, such protein influx was enhanced during the inflammatory response to bacterial products in CSF in antibody treated animals. In this example, this activity was demonstrated in the absence of an inflammatory stimulus.

polyclonal anti FHA Ab 1.2 1.2

These results are interpreted to show that:

Anti-FHA antibodies, particularly the monoclonal antibody 13.6E2, enhance the permeability of cerebral capillary endothelium sufficiently as to allow passage of serum proteins into CSF.

Example 5

Comparative Activity Of Anti-FHA Antibodies

3 In Enhancing Penetration Of [ H]

Penicillin From Serum To Brain

Rabbits (n= at least 2 per condition) were injected with the selected antibodies identified in

Example 2 (10 ug/kg) intravenously; 4 hours later, ₃ Hpenicillin (1.04 mCu/5Oug/rabbit; Merck, Sharp and dohme, Rahway, NJ) was injected intravenously and after 30 min, blood and brain harvested.

Approximately one gram of right parietal cortex was excised, weighed, homogenized, solubilized with

₃ Soluene 350 (Packard, Boston, MA) , and H cpm in brain and serum samples were counted. Blood volume per gram of brain was determined to be 300+ 73ul from a set of 6 saline-treated control animals. Brain uptake index indicates the amount of radiolabel in brain tissue after subtraction of values from blood

(BUI = cpm per gram brain homogenate - [cpm per ul blood x 300 ul blood/gram homogenate]). Results are shown in Fig. 13. The horizontal line indicates the maximum uptake observed in the IV saline control

₃ group (n=6) . The H penicillin is *§0% bound to albumin and thus, accumulation of radiolabel in this model reflects permeability of the BBB to large soluble molecules the size of albumin (70kD) . Only anti-FHA antibody 13.6E2 increased the passage of penicillin into brain.

Example 6

Antibody 13.6E2 Enhances The Permeability Of The Cerebral Capillary Endothelium In A Reversible, Time Dependent Manner

Healthy Animals received antibody 13.6E2 intravenously (lOug/kg) followed at various times thereafter by 3H penicillin intravenously. Thirty minutes later, blood and brain were harvested and analyzed as in Example 5. Results from each animal are plotted individually in Figure 14; cpm/g brain was converted to percent of the initial injected dose of radiolabel/entire brain weight.

Example 7

Intravenous Formulation I Ingredient m ml

cefotaxime monoclonal antibody 12.5D1 dextrose USP sodium bisulfite USP edetate disodium USP water for injection q.s.a.d.

Intravenous Formulation II Ingredient mg/ml

ampicillin 250.0 monoclonal antibody 12.1F9 10.0 sodium bisulfite USP 3.2 disodium edetate USP 0.1 water for injection q.s.a.d. 1.00 ml

Intravenous Formulation III Ingredient mg/ml

gentamicin (charged as sulfate) 40.0 monoclonal antibody 12.1 Bll 10.0 sodium bisulfite USP 3.2 disodium edetate USP 0.1 water for injection q.s.a.d. 1.00 ml

EXAMPLE 8 C3bi capture toxicity test

Preparation of C3bi coated erythrocytes (EC3bi)

1. Put E ⁹ (amount see below) in a 15 ml tube spin 5 min 2000 rpm aspirate supernatant resuspend in DGVB++ (amount see below) add C5 deficient serum

Amount E ^ Amount of DGVB++ Amount of C5def. serum Super 250 ul 250 ul 40 ul

High 250 ul 250 ul 25 ul

Medium 500 ul 250 ul 25 ul

Low 250 ul 80 ul 8 ul

2. Incubate 60 min 37°, shake the tubes after 30 min.

3. Stop With 500 ul EDTA/GVB—

4. Incubate 10 min 0°C (on ice)

5. Wash 4x in DGVB++ therefore add 5 ml DVGB++, spin 5 min, 2000 rpm, 4°C aspirate supernatant

6. Resuspend in: Super 2.5 ml DGVB++

High 2.5 ml DGVB++

Medium 5.0 ml DGVB++ Low 2.5 ml DGVB++

Q

Concentration is now 1x10 /ml

7. Finally, add for storage to E ^ and EC3bi, lul penicillin/streptomicin per 100 ul of suspension. Store ice. Suspensions are usable for 3 weeks.

IgG capture on Protein A and Toxicity test

1. Coat each well of a 60 well high profile terasaki tray 1003-01-0) with 5 ul of commercial protein A solution (cone 50 ug/ml) for 2 hr. at RT.

2. Adjust the pH of serum/fluid to 8.0 by adding 1/10 volume of 1.0 M Tris pH 8.0.

3. Wash the wells 2 times with Tris (1M) pH 8.0, aspirate all fluid.

4. Add the undiluted test serum (5ul) and leave for 2 hr.

5. Wash 2 times with PBS pH 7.4.

6. Remove fluid by aspiration

7. Add 5 ul C3bi supra coated erythrocytes diluted in PBS as described above.

8. Incubate for 30 min at 37°C in the incubator.

9. After 30 min incubation, turn the plates upside down to allow gravity to remove unbound erythrocytes for 10 min at 37 C.

10. Wash with PBS 3 times, slam on paper, wash once more.

11. Remove the fluid by aspiration

12. Add glutaraldehyde (2.5%) in PBS for 2 minutes.

13. Wash with PBS 3 times

14. Count at magnification 400 (100 fields).

PBS=with Ca and Mg.

PBS/glutaraldehyde = l ml 25% glutaraldehyde into 9 ml PBS.

Binding at or above the level of the positive control commercially available anti-C3bi or mAbF9 indicates toxic antibodies are present.

Ability of anti-B, pertussis antisera to bind cerebral capillaries and C3bi

Antigen Species Brain Cap* C3bi Capture

preimmune guinea pig rabbit human — —

DT guinea pig +++ + (whole cell) rabbit ++ human +++ +

Formalin FHA guinea pig rabbit human

JNIH-6 guinea pig (—) (—) rabbit human

FHA, TNM guinea pig ++ rabbit human

♦Binding to cerebral capillaries was performed as described in Example 1.