GB2367294A - Intimin and Tir binding in bacteria-host cell interactions - Google Patents

Abstract

The Tir binding site of intimin and its use to design or select agents which are capable of interfering with or preventing Tir-intimin binding and which thereby may be useful in inhibiting bacterial colonisation of host cells. Also claimed is a method of identifying an agent such as an antibody which is capable of binding all or part of the Tir binding site of intimin comprising providing an intimin molecule containing the Tir binding site and comparing the <SP>1</SP>H-<SP>15</SP>N HSQC (NMR) spectrum with and without the agent being present. Amide resonance which move or broaden in the presence of the agent indicating intimin residues involved in binding the agent.

Description

2367294 BIOLOGICAL MATERLALS AND METHODS FOR USE IN THE PREVENTION OR

TREATMENT OF INFECTIONS This invention relates to polypeptides involved in the interaction between bacteria and host cells.

5 Enteropathogenic (EPEC) and enterohaemorrhagic (EHEQ Escherichia coli constitute a significant risk to human health worldwide. EPEC are the cause of severe infantile diarrhoeal disease in many parts of the developing world, while EHEC are the etiological agents of a food-bome disease that can cause acute gastro- enteritis, bloody diarrhoea, haemorrhagic colitis and haemolytic uraemic syndrome (HUS) (reviewed by 10 Nataro, and Kaper, 1998).

EPEC colonise the small intestinal mucosa and, by subverting intestinal epithelial cell function, produce a characteristic histopathological feature known as the "attaching and effacing" (A/E) lesion (Moon et aL, 1983). The A/E lesion is characterised by localised destruction (effacement) of brush border microvilli, intimate attachment of the 15 bacillus to the host cell membrane and the formation of an underlying pedestal-like structure in the host cell consisting of polymerised actin, cc-actinin, ezrin, talin and myosin (reviewed by Frankel et aL, 1998a). A/E lesion formation is essential for pathogenicity and similar lesions have been associated with several other bacterial mucosal pathogens, most notably in EHEC (Donnenberg et aL, 1993 a; Donnenberg et aL, 20 1993b).

The first gene to be associated with A/E activity was eae encoding the intimate EPEC and EHEC adhesin, intimin (Jerse et al., 1990). Intimin exists as at least five antigenically distinct subtypes that have been named intimin (x, P, y, 8 and 6 (Adu-Bobie et al., 1998; Oswald et al., 2000). EPEC/EHEC intimins exhibit homology at their amino5 tennini to the invasin polypeptides of Yersinia (Isberg et al., 1987) and like Yersinia invasin (Leong et al., 1990), intimin harbours receptor binding activity at the C-terminus of the polypeptide (Frankel et al., 1994; Frankel et al. , 1995). A 76-ainino acid motif enclosed by a disulphide bridge between two cysteines lies within the C-terminal domain of intimin. This is absolutely required for intimin binding to the host cell, A/E lesion 10 formation and colonisation of mucosal surfaces (Frankel et al., 1995; Frankel et al., 1996; Frankel et al., 1998b; Hicks et al., 1998; Higgins et al., 1999a; Higgins et al., 1999b).

The C-terminal domain of invasin also harbours two Cys residues, in similar locations to those found in intimin (Leong et al., 1993).

Recently we have determined the global fold of the C-terminal 280 amino acids of 15 intimin a (Int280cc) by a combination of perdueteration, site-specific protonation and multidimensional nuclear magnetic resonance (NMR) (Kelly et al., 1998; 1999). T he structure shows that Int280a is approximately 90 A in length and is built from three globular domains. The first two domains (residues 1-91 and 93-181) each comprise P sheet sandwiches that resemble the immu-noglobulin super family (IgSF). Despite no 20 significant sequence homology, the topology of the C-terminal domain (residues 183280) is reminiscent of the C-type lectins, a family of proteins responsible for cell-surface carbohydrate recognition. A domain between residues 558-650 within the extracellular portion of intin-lin aligns with 35% identity (44% conservations) with the first domain of Int280, identifying a further Ig- like domain (Kelly et al.; 1999). This produces at least four domains that protrude from the bacterial membrane for interaction with the host cell.

5 For the purposes of future discussion, the domains in Int280 are therefore renamed as D2, D3 and D4 respectively. The presence of a disulpbide bridge in the C-type lectin-like module of Int280cc is essential for correct folding of this domain and for carbohydrate binding by other C-type lectins (Weis and Drickamer, 1996). Modelling other intimin types (including the EBEC intimin 7) (Adu-Bobie et A, 1998; Yu and Kaper, 1992) 10 would suggest they have similar structures, and define a new family of bacterial adhesion molecule.

The intimin-encoding eae gene is part of the large pathogenicity islands found in EPEC and EHEC, termed the LEE (Locus of Enterocyte Effacement) (McDaniel et aL, 1995; Perna. et aL, 1998). hi addition to intimin, the LEE also encodes a type III secretion 15 system (Jarvis et A, 1995; Jarvis and Kaper, 1996), an intimin receptor (Tir/EspE) (Deibel et aL, 1998; Kenny et aL, 1997) and three secreted proteins EspA, EspB, and EspD which are required for signal transduction in host cells and A/E lesion formation (Donnenberg et aL, 1993c; Kenny et A, 1996; Lai et aL, 1997). EspA is a structural protein and a major component of a large filamentous organelle that is transiently 20 expressed on the bacterial surface and interacts with the host cell during the early stage of A/E lesion formation (Ebel et aL, 1998; Knutton et al., 1998). EspA-filaments may contribute to bacterial adhesion but of greater significance, they appear to be a component of a translocation apparatus and as such are essential for the translocation of ESPB (Knutton et A, 1998; Wolff et al., 1998) and Tir (Kenny et al., 1997) into host cells.

Int280 can bind directly to uninfected host cells (Frankel et al., 1994; Frankel et 5 al., 1996). Moreover, a recent study has demonstrated that intimin can induce a T helper cell-type I immune response in the colonic mucosa and colonic hyperplasia, in mice (Higgins et A, 1999a). In addition, the bacterial protein Tir (EspE) can also act as an EPEC/EHEC intimin receptor (Deibel et al., 1998; Kenny et aL, 1997). Recently the intimin- binding region of Tir has been localised to a stretch of amino acids residues, 10 which resides between the two putative membrane-spanning domains of the polypeptide (termed Tir-M) (de Grado et al., 1999; Hartland et aL, 1999; Kenny, 1999). Immunofluoreseence staining of infected cells using polyclonal antisera raised against the amino and carboxy terminal peptides of Tir (Tir-N and Tir-C, respectively) demonstrate that these two regions are located within the host cell where they can induce the 15 polymerisation of actin and other cytoskeletal proteins to produce the characteristic pedestal-like structure ( Kenny et al, 1997; Hartland et A, 1999).

We report that an active Tir-binding fragment of intimin spans the IgSFlike domain, D3, and the lectin-like domain, D4. We have also determined the highresolution structure of the portion, comprising the C-tenninus 190 amino acids, which 20 form a super domain capable of interaction with Tir. Using chemical shift titration experiments we have also Rirther localised, for the first time, the binding site of Tir to a concerted patch of residues at the tip of the structure (D4) and propose a plausible model for the intimin-Tir interaction.

Aspects of the invention are defined in the accompanying claims and are also set out below.

5 A first aspect of the invention provides the use of the Tir binding site of intimin in a method for the design and/or selection of an agent capable of interfering with or preventing Tir-intimin binding, the Tir binding site being identifiable by comparing the HYN HSQC spectrum of an intimin molecule with and without saturating amounts of the Tir55 peptide, amide resonances which move and/or broaden in the presence of the 10 Tir55 peptide being indicative of intimin residues involved in binding Tir.

The Tir binding site may comprise, or consist of, one or more, or all, of the following residues of IntI90; Y140, Y142, 1147,1148, S149, W150, T154, Q156, D157, A158, V162, A163, S164, T165, K170, Q171, W176, 1177, S180, E181, N183, A184, Y185, T187, and V189; orthe corresponding residues of another intimin molecule.

15 Preferably, the Tir binding site comprises or consists of 5 or more of the residues of Intl 90, or the coffesponding residues of another intimin molecule.

Still more preferably, the Tir binding site comprises or consists of 5 to 14 of the residues of Int 190, or the corresponding residues of another intimin molecule.

The agent may be designed in a method involving computer-aided chemical modelling techniques.

Alternatively, the agent may be selected using an affmity purification technique.

The agent may be a peptide or polypeptide. For example, the polypeptide may be 5 an antibody or an antigen binding fragment thereof, the antibody or fragment being capable of specifically binding all or part of the Tir binding site of intimin.

A further aspect of the invention provides a method of identifying an agent which is capable of binding all or part of the Tir binding site of intimin comprising providing an intimin molecule containing the Tir binding site and comparing the 'H-"N HSQC 10 spectrum with and without the agent, amide resonance which move or broaden in the presence of the compound being indicative of intimin residues involved in binding the agent and determining whether one or more of such residues are the same as the residues of intimin involved in binding Tir; any of which are the same being indicative of an agent which is capable of binding all or part of the Tir binding site of intimin.

15 By "design" we include the meaning that the agent is made or modified so that it contains one or more functional groups which it is expected or known will bind all or part of the Tir binding site of intimin.

Skilled persons will be aware of a range of methods which can be used to design agents of the invention based on the knowledge of the Tir binding site hereby provided.

Exemplary methods include computer aided design. In all cases, the aim of using a computer for drug design is to analyse the interactions between the drug and its receptor site and to "design" molecules that give an optimal fit. The central assumption is that a good fit results from structural and chemical complementarity to the target receptor. The 5 techniques provided by computational methods include computer graphics for visualisation and the methodology of theoretical chemistry. By means of quantum mechanics the structure of small molecules can be predicted to experimental accuracy. Statistical mechanics permit molecular motion and solvent effects to be incorporated.

If the molecular model of the binding site is precise enough, one can apply 10 docking algorithms that simulate the binding of drugs to the respective receptor site. In a first step these programs create a negative image of the target site, place the putative ligands into the site and finally they evaluate the quality of the fit.

An example of the use of such design techniques is the design of potent HIV protease inhibitors (Science, 263, 1994, 380). The design was based on knowledge of the 15 target structure.

Exemplary docking algorithm references include:- I. B. K. Shoichet, D.L Bodian & I.D. Kuntz, "Molecular docking using shape descriptors", J. Comp. Chem. 13 (1992) 380-397.

8 2. 1. D. Kuntz, J. M. Blaney, S J Oatley, R. L. Langridge & T.E. Ferrin, "A geometric approach to macromolecule-ligand interactions", J. Mol. Biol. 161 (1982) 269 288.

3. 2T.J.A. Ewing & 1. D. Kuntz, "Critical evaluation of search algorithms for 5 automated molecular docking and database screening", J. Comp. Chem. 18 (1997) 1175 1189.

4. E.C. Meng, B K. Shoicet & 1. D. Kuntz, "Automated docking with gridbased energy evaluation", J. Comp. Chem. 13 (1992) 505-524.

5. A. R. Leach & I. D. Kuntz, "Conformational analysis of flexible ligands in 10 macromolecular receptor sites", I Comp. Chem 13 (1992) 730-748.

Examples of software programs for use in chemical design include MSi biosym, Tripos Sybyl, AutoDock, Software Dock, Dock and GRASP.

US Patents Nos. 5,891,643; 5,804,390; 5,698,401; and 5,989,827 relate to methods for designing compounds which bind to a specific target molecule, involving 15 two-dimensional "N-'H NMR correlation spectroscopy techniques. Such methods may be used in accordance with the present invention and the disclosure of those patents incorporated herein by reference.

By "selection" we include affinity purification techniques for identifying agents which are capable of binding to all or part of the Tir binding site of intimin. Examples include affinity chromatography and phage display technologies.

Examples of agents which are capable of binding to all or part of the Tir binding 5 site of intimin include antibodies and antigen binding fragments thereof wherein the antigen comprises all or part of the Tir binding site of intimin.

The invention further provides an agent capable of binding all or part of the Tir binding site of intimin, the agent being identifiable by a method or use of any of the preceding aspects of the invention.

10 The invention further provides the use of an agent of the invention in the manufacture of a composition for use in medicine, preferably for use in the prevention or treatment and/or diagnosis of a bacterial infection.

Preferably, the bacterial infection causes an histopathologic effect on intestinal epithelial cells, the effect being known as attachment and effacement (A/E).

15 Advantageously, the bacterial infection comprises infection by enteropathogenic Rcoli (EPEC) and/or enterohemorrhagic E.coli (EHEC), and especially E. coli 0157:117.

It will be appreciated that infection by other EHEC serotypes and shiga toxigenic E.coli (including human and bovine strains) Hafnia alvei and Citrobacterfreundii biotype 4280 are also included within the scope of the invention.

An agent of the invention is also provided for use in the manufacture of a composition for use as a food supplement or a food additive. The food is preferably a milk substitute. Preferably, the food is suitable for administration to a human baby or infant or a young animal.

5 However, it may be suitable for any individual who is susceptible to a bacterial infection, including older individuals. Exemplary animals include domestic cattle, especially calves; and poultry such as chickens and turkeys The invention also relates to a food product comprising a foodstuff and an agent of the invention.

10 It will be appreciated that the term "intiniin" includes variants, fragments and fusions that have interactions or activities which are substantially the same as those of the intimin sequences described herein and/or those disclosed in Frankel et al (1994) Infect.

Immun. 62:1853-1842 and/or public databases.

Techniques for the generation of intimin-derivatives with differing biological 15 activities using site-directed mutagenesis of the intimin C-terminal domain are described in Frankel et al (1998) Mol Microbiol 29(2), 559-540, the disclosure of which is incorporated herein by reference.

According to a fiu-ther aspect of the invention, there is provided a method of making an antibody against one or more residues of the Tir binding site of intimin as defined in accordance with the earlier aspects of the invention, comprising administering said residues of the Tir binding site to an animal and collecting and purifying the directly or indirectly resulting antibody. The antibody may be polyclonal, but is preferably monoclonal. Also provided is a hydridoma cell capable of producing such a monoclonal 5 antibody.

By."antibody" in accordance with the invention, we include molecules which comprise or consist of antigen binding fragments of an antibody including Fab, Fv, ScFc and dAb. We also include agents which incorporate such fragments as portions for targeting antigens and/or cells or viruses which display such antigens.

10 In accordance with this aspect of the invention, there is also provided an antibody raised against the Tir binding site of the invention, for use in medicine. Preferably, the antibody is for use in the manufacture of a composition for use in the prevention, treatment and/or diagnosis of a bacterial disease.

The agents identified or obtained according to the above aspects of the invention 15 may be a drug-like compound or lead compound for the development of a drug-like compound. Thus, the methods may be methods for identifying a drug-like compound or lead compound for the development of a drug-like compound.

The term "drug-like compound" is well known to those skilled in the art, and may include the meaning of a compound that has characteristics that may make it suitable for 20 use in medicine, for example as the active ingredient in a medicament. Thus, for 12 example, a drug-like compound may be a molecule that may be synthesised by the techniques of organic chemistry, less preferably by techniques of molecular biology or biochemistry, and is preferably a small molecule, which may be of less than 5000 daltons molecular weight and which may be water-soluble. A drug-like compound may 5 additionally exhibit features of selective interaction with a particular protein or proteins and be bioavailable and/or able to penetrate target cellular membranes, but it will be appreciated that these features are not essential.

The term "lead compound" is similarly well known to those skilled in the art, and may include the meaning that the compound, whilst not itself suitable for use as a drug 10 (for example because it is only weakly potent against its intended target, non-selective in its action, unstable, poorly soluble, difficult to synthesise, too toxic or has poor bioavailability) may provide a starting-point for the design of other compounds that may have more desirable characteristics.

The compounds identified in the methods of the invention may themselves be 15 useful as a drug or they may represent lead compounds for the design and synthesis of more efficacious compounds.

In a fin-ther aspect the invention provides an agent in accordance with the earlier aspects of the invention for use in medicine. Preferably, for use in the manufacture of a composition for use in the prevention, treatment and/or diagnosis of a bacterial disease, or 20 for use as a research reagent. Thus, the invention provides the use of an agent of the invention in the manufacture of a composition for use in the prevention, treatment and/or diagnosis of a bacterial disease, or for use as a research reagent.

In another aspect, the invention provides a pharmaceutical composition comprising phannaceutically effective amount of an agent in accordance with the aspects 5 of the invention, together with a pharmaceutically acceptable diluent or carrier.

In another aspect, the invention provides a method of preventing and/or treating a bacterial disease comprising administering to a subject an effective amount of an agent in accordance with the earlier aspects of the invention.

By "effective amounf 'we mean that sufficient quantities of the. agent are provided i o to produce a desired pharmaceutical effect beneficial to the health of the recipient.

Non-limiting examples embodying certain aspects of the present invention will now be described, with reference to the accompanying figures:

Figure 1:

Schematic representation of the overlapping Int280-derived polypeptides. The two IgSF 15 like domains (D2 and D3), the C-type lectin-like domain (D4) and the conserved motifs in Int280 are shown at the top. The position of W150 within int190 is indicated.

Numbers on both sides of the fragments mark the first and last amino acids of each fragment within the Int280 domain.

Figure 2:

(a) Detection of Int-Tir interactions using gel overlays. Western blots of NMP-Int derivatives were reacted with a rabbit MBP antiserum (top) or overlayed with Tir-M (bottom). Similar levels of MBP-Int280 (lane 1), NMP-IntI9O (lane 2), MBP- Intl5O (lane 3), MBP-Intl46 (lane 4), UMP-IntD3&D4 (lane 5) and NMP-IntD3 (lane 6), 5 fusion proteins were detected with polyclonal antiserurn (top), while Tir-M only bound to NOP-Int280 (bottom, lane 1) and MBP-Intl 90 (bottom, lane 2).

(b) Detection of Int-Tir interactions using the yeast two-hybrid system. P-galactosidase assays showing a 14 fold increase in enzymatic activity in strains co- expressing the whole Tir polypeptide and Int280 and a 7.5 fold increase when IntI90 was co 10 expressed with Tir compared with the parent, cured and the other single and double transformants.

Figure 3:

(a) Q, traces representing the superimposition of the 15 refined Intl 88 structures.

(b) Q, traces representing the superimposition of the 15 refined Intl 88 structures. The 15 orientations of (a) and (b) are related by a 90' rotation.

(c) Schematic representation of Intl 88 for the orientation displayed in (a).

(d) Schematic representation of Intl 88 domains for the orientation displayed in (b), (e) A 'flattened' illustration highlighting the topology of Intl 88. Helices are represented as open tubes and P-strands as arrows.

Figure 4:

The structural comparison of Intl88 topology with D4/D5 from Invasin (Hamburger et aL, 1999).

(a) Schematic representation of Intl88. Helices are shown as tubes and strands as 5 arrows.

(b) Schematic representation of D4/D5 from invasin. Helices are shown as tubes, strands as arrows.

(c) Sequence alignment and topology for Intl 88 and invasin. The approximate location of secondary structure elements is also indicated; helices delineated as open tubes and 10 P-strands as black arrows. The amino acids positions are shown for IntI90 from EPEC 0127:H6 and correspond to residues 750-939 in full-length intimin. The residues highlighted represent identity and conservation with the EPEC 0127:H6 sequence. Intimins from E. coli 012706, E. coli 026:H- and E. coli 0157:H7 have been immunologically categorised into types a, 0 and y, respectively. The two 15 invasin sequences, Yersinia pseudotuberculosis and Yersinia enterocolitica, are aligned on the basis of the structural superimposition provided by DALL Asterisks represent amide resonances that are significantly perturbed upon the addition of Tir55 peptide.

Figure 5:

(a) The 'H-'-5N HSQC NNR spectrum for Intl 88 (grey) with the identical region of the 'H-"N HSQC NNM spectrum for Intl88 with Tir55 peptide overlaid (black), Peaks, corresponding to amides that are affected upon addition of Tir55 are labelled with their residue number (For consistency these are numbered according to Intl 90).

(b) A schematic representation of the structure of Intl88. Residues shown in dark grey indicate chemical shift/line-width perturbation in the presence of Tir55, namely Y140, K142,1147,1148, S149, W150, T154, Q156, D157, A158, V162, A163, S164, T165, K170, Q171, N176,1177, S180, E181, N183, A184, Y185, T187, and V189This data is also summarised in Figure 3(a). The data illustrate a possible binding site 10 for Tir55, which is composed of highly concerted patch of residues in D4. For clarity Intl 88 is rotated by I 80'to figure with respect to figure 3a.

(c) A dot representation of the solvent accessible surface for Intl 88. Shownindarkgrey are the residues that show chemical shift/line-width perturbation in the presence of TiM. The data illustrate a possible binding site for Tir55, which is composed of 15 highly concerted patch of residues in D4.

(d) Schematic representation of the quaternary structure of the entire extracellular region of intimin in complex with Tir peptide. IGSF domains DI and D2 are shown in white. The Tir binding fragment of intimin (i.e. int 190, D3 and D4) is illustrated by the space filling representation shown in (b). Tir is shown schematically: the two predicted transmembrane helices shown as tubes and the intimin-binding peptide, which is manually docked onto the binding site, as a dark grey ribbon.

Figure 6:

FAS reaction (a, d, g), intimin staining (b, e, h) and overlaid images (c, f, and i) of 5 infected HEp-2 cells. CVD206(pCVD438) (a-c) showed both FAS and intimin staining.

No staining was observed using strain CVD206 alone (d, e) although adhering bacteria could be observed by phase contrast (f). Substitution of W899 in strain CVD206(pICC54) resulted in FAS negative staining although surface intimin expression was not affected (g-i).

10 Figure 7:

(a) Detection of Intl9O-Tir interactions using gel overlays. Western blots of NOP-Int mutants were reacted with a rabbit NMP antiserum or overlayed with Tir-M. Similar levels of NMP-Intl9O (lane 1) and MBP-Intl9OWl5OA (lane 2) fusion proteins were detected with polyclonal antisenun, while Tir-M only bound to NMP-Int 190 (lane 3).

15 (b) Detection of Int-Tir interactions using the yeast two hybrid system. P-galactosidase assays showing a 7-fold increase in enzymatic activity in strains co- expressing the whole Tir polypeptide and lntl90 compared with Tir-Int:190AI50, Tir- Int28OA240 and single plasn-d transformants.

Figure 8: details the atomic co-ordinates which define the threedimensional structure of the Tir binding site of intimin (Int 190/Int 188) . The various columns give the following infon-nation:- 2 d column atom no.

5 Tdcolumn atom type 4' column amino acid type 5'h column amino acid number (hnportant: 3 should be added to give the binding site reference numbers mentioned herein) 6', 7' and 8' columns give the x, y and z co-ordinates.

10 Using the above information, a skilled person can design and synthesise compounds which inhibit or prevent Tir-intimin binding and therefore find utility in the prevention, treatment and/or diagnosis of disease.

The atomic co-ordinates have been deposited in the Brookhaven PDB and can be obtained from s.j.matthews@ic.ac.uk.

Tables

Table 1. List of plasmids.

Plasmid Description Reference pGBT9 A yeast GAL4 DNA-BD cloning vector Clontech pGAD424 A yeast GAL4 DNA-AD cloning vector Clontech pMal-C2 Vector for expression of NIBP- tagged NEB proteins pET3d Vector for expression of proteins for Novagen purification without a tag PICCIO pGBT9 expressing Tir Hartland et aL (1999) PICC19 pGAD424 expressing Int280 Hartland et aL (1999) pICC39 pGAD424 expressing Intl 90 This study pICC40 pGAD424 expressing Intl 50 This study pICC41 pGAD424 expressing Intl 46 This study pICC42 pGAD424 expressing IntI20 This study pICC44 pGAD424 expressing Intl9OW150A This study pICC45 pMal expressing MBP-Int280 Frankel et aL (1994) pICC46 pMal expressing MBP-hitl9O This study pICC47 pMaI expressing UMP-Intl5O Frankel et aL (1994) pICC48 pMal expressing NMP-Intl46 This study pICC49 pMaI expressing NMP-Int 120 Frankel et aL (1994) pICC50 pMaI expressing MMP-IntD2+D3 TWs study pICC51 pMal expressing MMP-IntD3 This study pICC52 pMal expressing I-ntl9OWI5OA This study pCVD438 pACYC184 containing the eae gene of (Donnenberg and E2348/69 Laper, 1991) pICC53 pCVD438W899A Ibis study pICC54 pET3a expressing Intl 90 This study pICC59 pET3a expressing Intl 88 This study pICC62 pGAD424 expressing lnt28OA2,0 This study pICC63 pMaI expressing Int28OA240 This study Table 2: List of primers.

Primer Sequence 280-F 5GGA ATT CAT TAC TGA GAT TAA GGC T 190-F YGGA ATT CTT TTT TAC AAC GCT TAC AAT T 190-R 5'CGG GAT CCT TAT TTT ACA CAA GTG GC 188-F YAAA CAT ATG ACG ACG CTT ACA ATT GAT G 146-F YGGA ATT CGG TAA TAT TGA AAT TGT TGG AAC 146-R 5'CGG GAT CCT TAA GCT GTT TGT TGT ACC CAT D2+D3 -F YGGA ATT CAT TAC TGA GAT TAA GGC T D2+D3 -R 5'CGG GAT CCT AAT GCA GCC CCC CAT GC D4 -F YGGA ATT CCT TAC AAT TGA TGA CGG T D4 -R 5'CGG GAT CCT AAT GCA GCC CCC CAT GC F - forward primer, R- reverse primer.

Table 3: Results from structure calculations for Intl 88 Statistic <SA>' Restraints 5 Medium and short range (1 < ji-j I < 5) 771 0,030 0.003 Long range (ji-jj > 4) 455 0.025 0.002 Hydrogen bonds 151 0.027 0.003 Dihedral angles 307 0.94 0.03 Idealised geometry 10 Bonds b 0.0024 0.0003 Anglesr 0.58 0.009 Improper Anglesc 0.47 0.02 Coordinate PoSitionS d 15 Backbone atoms for secondary structure residues in D3:

(5-8,8-14,20-23,32-38,43-48,54-57,61-65,70-77,82-87) 0.7 0.1 Heavy atoms for secondary structure residues in 133:

(5-8,8-14,20-23,32-38,43-48,54-57,61-65,70-77,82-87) 1.2 0.1 2o Backbone atoms for secondary structure residues in D4:

(92-96,104-113,122-132,146-151,162-167,173-175,186-190) 0.9 0.1 Heavy atoms for secondary structure residues in 134: (92-96,104-113,122-132,146-151,162-167,173-175,186-190) 1.4 0.1 25

Backbone atoms for all residues 3-190 1.5 0.2 All heavy atoms for all residues 3-190 1.9 0.2 a The average root-mean square deviations for the final 15 structures.

bAngstroms cDegrees Numbered according to Intl 90, see figure 4c.

The average root-mean square deviations from the average structure MATERIALS AND MIETHODS Bacterial strains and plasmids Bacterial strains used in this study included wild-type EPEC strain E2348/69 (0127:H6), eae deletion mutant derivative CVD206 (Donnenberg and Kaper, 1991) and 5 E. coli strains TG1 and BL21. Bacterial strains were grown in L-broth. Media were supplemented with 50 gg/ml kanamycin or 100 pg/ml ampicillin where appropriate. The plasmids are listed in Table 1.

Preparation of MBP-Int280 derivatives DNA segments encoding Intl9O, IntI46, Int235, Intl34, IntDl&2 and IntD2 10 (Figure 1; Table 1) were amplified by PCR (primers are listed in Table 2) using E2348/69 DNA as template (Frankel et aL, 1995). The truncated Int280 fragments were cloned into pMal-c vector in E. coli TG1 and the MBP-1nt280 derivatives purified and analysed by 12% SDS-polyacrylamide gel as described (Frankel et aL, 1995).

Gel overlays 15 Purified M13P-Int280 derivatives or MBP alone were separated by SDS- PAGE, blotted onto a nitrocellulose membrane and blocked with 10% skim-milk in PBS, 0.1% Tween-20 overnight as described (Frankel et aL, 1995; Hartland et aL, 1999). His-Tir-M was purified as described (Hartland et aL, 1999). The nitrocellulose membranes were reacted with 5 pglml of His-Tir-M in PBS, 0. 1% Tween-20 for 2 h and washed twice for 24 min in PBS, 0.1% Tween-20. I-Es-Tir-M binding to the different MBP-Int280 derivatives was detected with anti-His antiserum, (1:2,000 for 2 h) and then anti-rabbit antibodies conjugated to alkaline phosphatase (1:2,000 for I h). Binding of the individual MBP-Int280 derivatives to immobilised Tir-M was performed as described (Hartland et 5 aL, 1999).

Yeast two-hybrid system The yeast two-hybrid host PJ694A (MATa trpl-901 leu2-3,112 ura3-52 his3- 200 gaND ga180D LYS2::GALI-HIS3 GAL2-ADEI met2::GAL7-IacZ) was selected for use in this study which confers the advantage of three independent reporter genes under the 10 control of three different GAL promoters (James et A, 1996). DNA fragments encoding the Int280 derivatives and Tir derivatives were cloned, following PCR amplification (Table 1), into the ADHI driven fusion vectors pGBT9 (carrying the GAIL4 binding domain, BD) and pGAD424 (carrying the GAL4 activation domain, AD) respectively (Clontech). The constructs were used to transform PJ694A with initial selection for the 15 plasn-lid encoded TRP1 and LEU2 genes (Hartland et A, 1999). The resulting transformants were replica plated onto 3-aminotriazole containing medium to select for the HIS3 reporter, and onto SC minus Trp, Leu, Ade to select for the ADE2 reporter. The function of lacZ reporter was quantified in cell extracts by assaying for P-galactosidase activity using o-nitrophenyl b-D-galactopyranoside as a substrate (Miller, 1972).

Site-directed mutagenesis Site-directed mutagenesis of the eae gene was perfonned using the QuickChange Site-directed mutagenesis kit (Stratagene) following manufacturers' instructions, using double-stranded pCVD438 as template. Complementary mutagenesis oligonucleotide 5 pairs incorporating single amino acid substitutions are as follows; Sense oligonucleotides:

51 GACTATAAMCAGCTGTACAACAAACAGC (A150) A,ntisense oligonucleotides; 5'GCTGTT-rGTTGTACAGCTGAAATTATAGTC (A150) 10 Mutated plasmid containing staggered nicks was generated by extension of primers annealed to opposite strands of the denatured plasmid by temperature cycling (I cycle of 950C, 30 seconds, then 16 cycles of 950C, 30 seconds, 550C, I minute, 680C, 18 minutes) in the presence of the high fidelity Pfu DNA polyinerase. Synthesised DNA containing the desired mutation was selected from the original DNA template by is incubation with Dpn I at 370C for I h,'on the basis that dam methylated parental DNA template would be susceptible to digestion whereas the newly synthesised unmethylated mutated plasmid would not. Nicks in the plasmid were repaired following transformation of 1 d of the synthesised products into competent E.coli X-L I -Blue cells.

Chloramphenicol-resistant transformants were randomly selected and inoculated to ovemight L-broths for preparation of plasmid mini-preps (Qiagen). Correct incorporation of each mutation was monitored by DNA sequencing using an automated DNA sequencer. Mutated pCVD438 derivatives were also used as PCR templates for 5 subcloning the DNA fragments encoding IntI90 and Int280 into pMAL-c (pICC52 and pICC63 respectively) and pGAD424 (pICC44 and pICC62, respectively) for expression as M13P fusion in E. coli and for the yeast two hybrid-system, respectively (Table 1).

Mutated plasmid was transformed to competent CVD206 cells (eae-, Donnenberg:

and Kaper, 1991) for analysis of phenotypic effects. The mutated pCVD438 derivative 10 was also used as PCR templates for subeloning the DNA fragments encoding Intl 90 into pMAL-c2 (pICC52) and pGAD424 (pICC44) for expression as MBP fusion in E. coli and for the yeast two hybrid-system, respectively (Table 1).

Detection of intimin expression by Western blotting and fluorescent actin staining (FAS) 15 Expression of the intimin derivatives was determined by Western blotting.

Briefly, stationary L-broth cultures were diluted 1: 100 in DMEM and incubated for 3 h at 370C. An equivalent of an optical density (600) (OD600) of 0.5 was loaded onto 7.5 % SDS-polyacrylamide gel electrophoresis as described (Adu-Bobie et al., 1998). The electrophoresed polypeptides were transferred to a nitrocellulose membrane and 20 immunodetection of intimin was performed using an intimin a antiserum, diluted 1:500.

Fluorescence actin staining (FAS) test (using FITC Phalloidin) (Knutton et al., 1989) and intimin immuno-staining (using anti Int28O and TRITCconjugated anti rabbit antibodies) were employed to detect A/E lesion formation and intimin expression on infected HEp-2 cells by confocal microscopy.

5 Recombinant expression of Int190 and Int188 The 570 bp (encoding IntI90) and 564 bp (encoding Intl88) fragments of the eae gene from EPEC E2348/69 (and a TAA stop condon) were cloned, following PCR amplification (the primer sequences are listed in Table 2), into the prokaryotic expression vector pET3a using restriction sites Ndel and BamHI, generating plasmid pICC54 and 10 pICC59. The plasmid was transformed into E. coli strain BL21 containing pLysS and grown in Luria-Bertani broth supplemented with 100 [tg/ml ampicillin and 30 iglml chloramphenicol. For NNM, the cultures were grown in minimal media containing 0.07 % 15NH4CL and 0.4% 13C-glucose supplemented with 100[tg/ml ampicillin and 30 jig/ml chloramphenicol. Cultures were grown until mid log phase, and then protein 15 expression induced by the addition of 0.4 mM isopropyl-b-D- tMogalactopyroside (][PTG).

Growth was continued overnight at 300C after which the cultures were harvested by centrifugation. Since Intl 88 was found to be insoluble, the cell pellets were resuspended in lysis buffer (20 mM Tris pH 8.0, 1 m-NM EDTA, 0.5 nim pMSF), sonicated and centrifuged to pellet the insoluble and inclusion body material. The supernatant fraction 20 was discarded.

Refolding and purification of Int188 The pellets were washed (50 mM Tris pH 8.0), a nd spun for 15 minutes at 20,000 g. This was repeated twice, firstly washing in TritonTM buffer (SOMM Tris pH 8.0, 1% TritonTm), and secondly guanidine (50 mM Tris pH8.0, 0.5 mM guanidine). Following 5 the final wash the inclusion body pellets were dissolved in 50 MI of solubilisation. buffer (50 mM Tris pH8.0, I mM EDTA, 6 M Guanidine) by stirring overnight at room temperature. The solubilised inclusion bodies were then passed slowly by peristaltic pump into I L of refolding buffer (50 mM Tris pH8.0, I mm EDTA, 1 M arginine, 1 mM glutathione (reduced), 0.5 mM glutathione (oxidised) at 40C overnight, stirring. The 10 refolded protein was then concentrated to 50 mls using an Amicon apparatus with a 10 kDal cut-off membrane. This was then dialysed against 50 mM Tris pH 8.0, 1 MM EDTA. The refolded protein was purified by cation exchange chromatography using sulphopropyl sepharose in 20 mM sodium acetate pH 5.2. Pure intimin was eluted in 15% NaCl, the samples pooled, dialysed into 20 mM sodium acetate pH 5.2 and concentrated for NMR.

NMR spectroscopy For NMR spectroscopy, 1 mM samples of "C/"N Intl 8 8 was prepared in 0. 5 ml sodium acetate buffer at pH 5.2. Approximately 10% v/v D.0 was added to the sample. All NMR spectra were recorded at 500 Nfflz proton frequency on a four-channel Bruker 20 DRX500 spectrometer equipped with a z-shielded gradient triple resonance probe. 'ne temperature was maintained at 3 1 OK throughout the experiments. The sequence-specific backbone 'HN, "N, "Coc, "CP and "C' assignments were completed using HNCA, HN(CO)CA, CBCA(CO)NH and BBHA(CO)NH experiments (Grzesiek & Bax, 1992a; Grzesiek & Bax 1992b; Muhandiram. & Kay, 1994; Kay et aL, 1994). The secondary 5 structure and side chain assignments were determined using 'HYN nuclear Overhauser (NOESY) heteronuclear single quantum coherence (HSQQ and HCCH- total correlation (TOCSY) spectroscopy (Bax et aL, 1990). Distance restraints for the structure calculation were obtained from 'HYN NOESY HSQC (Norwood et aL, 1990; Marion et al, 1989) (80 ms mixing time), 'H-"C NOESY-HSQC (80 ms mixing time) and double 10 13 C edited HMQC-NOESY-HSQC (Vuister et aL, 1993) (80 ms mixing time) spectra recorded on a fully exchanged D20 sample. All the experiments used gradients for coherence selection together with the sensitivity enhancement protocol (Kay et aL, 1992).

Structure calculation NOE cross peak intensities were measured at 80 ms mixing. time. The 15 structures were calculated from random starting coordinates on the basis of 1528 distance restraints, comprising 771 short range NOEs (residue i to residue I < i < 4), 455 long range NOEs (residue i to residue i >4), 151 hydrogen bonds and I disulphide bond. The calculation also included 307 dihedral angle restraints, composed of 112 and I I I T angles.. A hybrid torsion angle and Cartesian co20 ordinate dynamics protocol was executed within the program X-PLOR (Nilges et aL, 1988, Briinger, 1993). The distance restraints were calibrated internally using known distances. NOEs observed at 80 mixing time were placed in three categories on the basis of estimated cross peak intensities: strong (< 2. 8 A), medium (< 3.5 A) and weak (< 5.0 A). The dihedral angle restraints were estimated using the backbone torsion angle prediction package TALOS (Comilescu et aL, 1999). No 5 distance violation greater than 0.2 A and no dihedral violation greater than 4.0' was found.

NMR binding studies 10 The 55 amino acid peptide encompassing the intimin-binding region of Tir was synthesised using Fmoc chemistry on a Perseptive Biosysterns 9050+ automatic peptide synthesiser with customised protocols. Coupling was performed using HBTU (benzotriazol-N,N,9,9-tetramethyluronium hexafluorophosphate) chemistry. After removal of the final Fmoc group, the resin was washed with methanol and 15 dichloromethane and dried under high vacuum. The peptide was released from the resin (500mg) during 2h with a cocktail (50ml) of trifluoroacetic acid (Tfa) 92.5%, phenol 2%, water 2%, ethaneditbiol 2%, anisole 1% and triisobutylsilane 1%. Characterisation was performed using reversed-phase HPLC on a C18 silica wide-pore column (15cmx 2.Imrn) with an elution gradient of 10%-50% acetonitrile containing 0.1% Tfa over 40 20 min. Further purification where necessary was perfon-ned on a 25cm x 22mm column using the same conditions. The molecular weight was confirmed by electrospray mass spectrometry. The final peptide sample was assessed to >75% pure. Two 0.5mM Intl 88 31 samples were prepared in acetate buffer at pH 5.2. One of which was left untouched and the other freeze-dried Tir55 was added to obtain a 10:1 Tir55:Intl88 mole ratio. 'H-"N Heteronuclear single quantum coherence (HSQC) spectra were recorded on the two starting samples and on sequential mixtures in order to follow the titration of the amide 5 resonances, which facilitated their assignment. The experiment was repeated twice; again with Intl 88 and once with Int280.

Results and Discussion Localisation of the minimal Tir-binding region of intimin Based on low resolution structural information reported earlier (Kelly et al., 10 1998; 1999), a number of truncated Int280cc derivatives, expressed as maltose binding protein (M13P) fusion proteins, were constructed and used together with purified Tir-M (Hartland et al., 1999) in a gel overlay binding assay. The NMP-Int280a derivatives are shown schematically in Figure I and include Int280, Int190, IntI50 and IntI20. We have shown before that the D3 domain forms numerous hydrophobic contacts with D4 (Kelly 15 et aL, 1999). Accordingly, we included the first 42 amino acids of D4 in the derivatives designed to assess the potential Tir-binding activity of the IgSF-like domains (IntD2&D3 and IntD4; Figure 1). MBP-Intl46, which harbours D3 and 62 amino acids of D4, encompasses both conserved VVLQYGQ and WGAANKY motifs (Adu-Bobie et al., 1998). (Figure 1). The gel overlay protein-binding assays revealed that Tir-M bound to 20 the immobilised Int280cc and Intl9Occ fusion proteins (Figure 2a). However, no binding was observed with any other MBP-Int fusion proteins or MBP alone (Figure 2a and data not shown). Similar results were obtained when the binding assay was performed with inimobilised Tir-M and soluble M13P-Int fusion proteins (data not shown). These results are consistent with those reported recently for intimin y from EHEC (Liu et at, 1999).

5 The interaction of the different Int280ct derivatives with Tir was also investigated using the yeast two-hybrid system, designed to identify protein-protein interactions through the fimctional restoration of the yeast GAL4 transcriptional activator in vivo (James et at, 1996). For this, selected DNA fragments encoding truncated Int280ct polypeptides were sub-cloned into pGAD424 to generate plasmids pICC39, pICC40, 10 pICC41 and pICC42 (Table 1). We previously reported that in the yeast two-hybrid system, the interaction of Int280 (pICC19) with the whole Tir polypeptide is more efficient than the interaction of Int280 with Tir-M (Hartland et at, 1999) . Accordingly, the DNA fragment encoding the whole Tir polypeptide, expressed from the second yeast two-hybrid system vector, pGBT9 (pICC10) (Hartland et at, 1999), was used in this part 15 of the study.

Plasmid pICCIO was co-transformed with each of the different pGAD424based plasmids into a derivative of the yeast stain PJ694A selected previously as a reporter for intimin-Tir interaction (Hartland et at, 1999). Replica plating these colonies onto selective media yielded vigorously growing colonies, and hence a positive two-hybrid 20 phenotype, in yeast strains expressing both Tir and Int:280 (pICCIO and pICC19) and Tir and Int 190 (pICC 10 and plCC3 9). No yeast colonies were observed using Tir with any of the other Int280 truncations or single plasmid transformants (data not shown).

The function of the non-selective reporter, lacZ, was also assessed in these strains by measuring P-galactosidase activities (Figure 2b). The host or single plasmid-bearing 5 strains exhibited low levels of P-galactosidase activity, whereas the strains expressing Int280 or IntI90 and Tir showed a 14 or 7.5-fold induction of P-galactosidase units, respectively. Based on the P- galactosidase levels of the strains bearing the plasmids pICCIO and pICC19, and pICCIO and pICC39, InI280 appears to interact with Tir with greater affinity than IntI90. This implies that regions outside IntI90 may be involved in 10 interaction with the whole Tir polypeptide. Nevertheless, taken together, these results show that the region of intimin that spans the C- tenninal 190 amino acids is capable of interacting with Tir. In the low- resolution structure of Int280 it can be seen that the D3 and D4 junction comprises a small surface area. Therefore, it is unlikely that a truncation at this point will severely affect stability whereas truncations elsewhere will probably 15 compromise its stability. Localising the Tir-binding region of intimin to the C-terminal 190 amino acids using gel overlay and yeast two-hybrid system assays defines the minimal functional fragment.

Solution structure of Int19O Using a combined perdeuteration/site specific protonation and multidimensional 20 NMR approach we were able to define the fold of the original 280 amino acids fragment, which is 30.1 KDa (Kelly et al. 1999). The shorter Intl 90 fragment is approximately 20 kDa and its size falls within the applicability of standard multi-resonance NMR methods for determining highly defined structures. The DNA fragment encoding IntI90 was subcloned into pET3a and over expressed in 131,21. However, IntI90 was expressed at levels

5 significantly lower than Int280. We hypothesised that this might be due to the presence of two hydrophobic amino acids (Phel and Phe2) at the N terininus of the IntI90 polypeptide. In order to try and improve expression and solubility we removed these two amino acids and cloned into pET3a a DNA fragment encoding IntI88. The level of expression was vastly improved but Intl88 concentrated into insoluble inclusion bodies.

10 Accordingly, following centrifugation, the inclusion body material was precipitated, solubilised and refolded as described in Materials and Methods.

Comparison of the 'H-"N HSQC NMR spectra for Intl 88 and Int280 indicate that Intl 88 is fully structured. Peaks corresponding to amides within the N-terminal 92 amino acids of Int280 are absent in the Intl 88 spectrum but the rest remain largely unchanged.

15 'H, "N, "C sequence specific backbone and side-chain assignments were completed using the standard methodology (Grzesiek & Bax, 1992a; Grzesiek & Bax 1992b; Muhandiram & Kay, 1994; Kay et al, 1994; Bax et al, 1990). Structures were calculated on the basis of 1226 nuclear Overhauser effect (NOE) distance, 115 H-bond distance and 307 dihedral angle restraints. The calculations used a hybrid torsion angle and Cartesian 20 co-ordinate dynamics protocol, executed within the program XPLOR (Brunger, 1993; Nilges et al., 1988). A final family of 15 structures (shown in Figure 3) were produced which contained no distance violations greater than 0.2A. Intl 88 is approximately 60 A in length and comprises two intimate globular domains; D3 (residues 3-90) and D4 (residues 91-190, numbered according to Intl.90). The overall root-mean square deviations (rmsds) between the family and mean co-ordinate positions are 0.7 0.1 A and 5 0.9 0.1 A for the backbone atoms of secondary structure elements in D3 and D4, respectively. For all heavy atoms of the same region the rmsd values rise to 1.2 0. 1 A and 1.4 0.1 A, respectively. Numerous contacts between D3 and D4 are observed that define the relative orientation of the two domains. The rmsd for all the backbone atoms of D4D5 (residues 1- 188) from the average structure is 1.5 0.2 A, indicating the relative 10 orientation of these domains is well defined. The complete set of structural statistics is shown in Table 3.

The topology of Intl88 is shown in Figure 3. D3 displays little sequence homology with known IgSF members, but belongs to the Type C set of the IgSF.

Interestingly, D4 contains a unique feature that is not seen in mammalian IgSF domains.

15 A prominent P-insertion (A', A") between strands A and Aextends a platform on top of D3 that contacts D4 and helps to define the relative orientation of the two domains.

This feature was poorly defined in our earlier determination of the fold of Int280 but, in the current study, is now characterised by numerous representative P- sheet NOEs and is therefore well defined.

20 D4 comprises four helices that surround two anti-parallel P-sheets. The C terminal strand is disulphide bonded to helix 1, and together with the N- tern-.tinus of D4 forms the two principal strands of the first sheet. Helix III protrudes from the main structure into the solvent and therefore was not observed in our low-resolution structure of Int280 (Kelly et aL, 1999). Interestingly, this helix contains an unusual kink at residue 139, which is replaced by proline in other intimin types. Despite no significant sequence 5 homology (i.e. less than 10%), the a/P topology of D4 is reminiscent of the C-type lectins domains (CTLD). Seventy-seven structurally equivalent Q, atoms from D4 of intimin superimpose with a RMSD of 3.0 A on the lectin domain from E-selectin (Graves et aL, 1994). The CTLI)s form a family of calcium-binding proteins responsible for cell-surface carbohydrate recognition (CRD) that includes animal cell-receptors and bacterial toxins.

10 However, several variants on the CTLD theme have been recently characterised. One example of a CTLD that lacks calcium co-ordination but recognises carbohydrate, is the TSG-6 link module that binds haluronan (Khoda et aL, 1996). Conversely, the type H antifreeze protein from sea raven, which does co-ordinate calcium, inhibits the formation of ice crystals, hypothesised to bind ice surfaces via the second P-sheet and calcium loop 15 (Gronwald et aL, 1998). Intriguingly, the CTLD's are also structurally related to proteins that are involved in protein recognition. These include CD94/NKG2 heterodimer, in which two CTL-like domains form an extensive flat surface involving the second P-sheet that is postulated to interact with the type lb human leukocyte antigen (Boyington et aL, 1999).

20 In CTLDs, the carbohydrate recognition site lies on the exposed face of the second P-sheet and an extensive loop between strands C and D (30 residues) in which co- ordinated calcium is directly involved in binding. No evidence for calcium binding is available for intimin and moreover intimin D4 lacks the extensive calcium bind loop, which is replaced by a six-residue helix (helix IV). Further differences between the CTL-like domain from intimin and archetypal CTLDs also exist. D4 from intimin 5 contains a single disulphide bond whereas CTLDs retain several that are conserved. Intimin also contains a larger proportion of helical structure than known CTLDs and the relative orientations of these regions are subtly different. However, the similarity with CTLDs does raise the question of an intimin function involving carbohydrate recognition, but evidence has yet to be provided. Therefore, this remains highly speculative.

io Structural comparison with Invasin from Yersiniapseudotuberculosis Intimin from enteropathogenic Escherichia coli (EPEC) and invasin from enteropathogenic Yersinia pseudotuberculosis (YP) are probably the bestcharacterised bacterial cell-adhesion paradigms. Research has now culminated in the high-resolution three-dimensional structure information on both proteins.

15 Intimin and invasin belong to a family of outer membrane proteins that mediate bacterial adherence. Members of this family of proteins are approximately 900 amino acids in length and possess highly similar N-tennini; over 36% identity exists within the first 500 amino acids, part of which is hypothesised to be sequestered within the bacterial outer membrane. Despite being highly related, these proteins perforin remarkably 20 divergent tasks and this is illustrated by the marked difference in pathogenesis between the organisms. Invasin causes invasion and translocation of the bacterium across the intestinal epithelium into deeper host tissue. Intimin also promotes adherence to epithelial cells but not intemalisation; instead, it induces the attaching and effacing lesion.

Invasin recognises members of the integrin super-fwnily from mammalian cells in order to facilitate bacterial adherence to and invasion of host cells. Moreover, two aspartic acid 5 residues, reminiscent of the integrin binding and synergy regions of fibronectin, are required for recognition (Hamburger et aL, 1999). It has also been established that horno multimerisation of invasin from YP facilitates receptor clustering, tight adherence and the subsequent invasion signal (Dersch and Isberg, 1999). In contrast, intimin binds the translocated intimin receptor protein (Tir), which is introduced into the host cell 10 membrane by the bacterium via a type III protein secretion/translocation system. For both intimin and invasin a C-terminal fragment of approximately 190 amino acids is sufficient for function in which no significant sequence homology is present (less than 10%).

Remarkable likenesses can be revealed between the structures of intimin and 15 invasin (shown in Figure 4). Hamburger et aL (1999) determined the crystal structure of an extracellular portion of invasin (Inv497). Inv 497 is approximately 180A in length and composed of five distinct domains (131, D2, D3, D4 and D5). The first four of which resemble eukaryotic members of the immunoglobulin super-family (IgSF). D3- D5 in invasin are analogous to the domains of Int280 (280 amino acid C-terminal fragment of 20 intimin), that was revealed by earlier NMR studies (Kelly et aL, 1998; 1999). It has also been reported that sequence alignment indicates the existence of a further IgSF domain within intimin. Also, secondary structure algorithms suggest that residues 445-558 contain a high P-sheet content but sequence alignment with known Ig domains (maximum of 20% identity, 25% conservation) falls below the cut- off for definitive assignment of an Ig-like fold. Therefore the extracellular portion of intin-fin contains at least four Ig-like domains arranged in a rod similar to the C-terminus of invasin, which 5 conveys a highly accessible 'adhesive tip'to the target cell. The tip in both intimin and invasin is composed of the C-terminal 190 amino acids, which forms a super-domain. Despite no significant sequence homology, Intl.90 and D4/D5 from invasin are similar; 73 equivalent Q, atoms superimpose with an misd of 2.9A for D3 in intimin with D4 in invasin and 77 equivalent C atoms superimpose with an rmsd of 3.5A for D4 in intimin 10 with D5 in invasin. The most notable differences occur within the divergent C-terminal domain (D4 in intimin and D5 in invasin); these include two additional helices in Intl 90 (helix 1111 and IV), the positions of some of the secondary structure elements and the relative orientations of C-terminal. domains. Despite these differences the C- terminal 100 amino acids of both intimin and invasin possess a folding topology related to that of 15 the CTLD's.

Considering the limited sequence similarity between 1nt190 and D4-D5 of invasin (-10% identity), their different binding specificities and biological activities, it is remarkable that these proteins adopt very similar structures. These findings raise intriguing questions regarding the origin and the parallel evolution of these two bacterial 20 adhesion molecules.

Identification of the Tir binding site within intimin.

Recently, the intimin-binding region of Tir has been localised to a central region encompassing a 55 amino acid motif between two putative membrane-spanning helices (de Grado et aL, 1999; Hartland et aL, 1999; Kenny, 1997; 1999). Analysis of line-widths and chemical shifls for Intl88 amide resonances in the presence of the Tir55 peptide 5 facilitated mapping of the Tir-binding surface. Amides, which exhibit large chemical shift changes and/or line broadening in the presence of Tir55, reveal residues likely to be directly involved in binding. A comparison of the 'H-"N HSQC spectra with and without saturating amounts of Tir55 is shown in Figure 5. A number of amide resonances move or broaden whilst the majority of the spectrum remains unchanged, which is indicative of 10 a specific complex between IntI88 and Tir55. The affected residues are concentrated within a region located in the C-terminal domain, D4, and are illustrated clearly in Figure 5, in which they are coloured red on schematic and solvent accessibility representations of Int:188. Residuesinvolved include Y140, K142,1147,1148, S149, W150, T154,QI56, D157, A158, V162, A163, S164, T165, K170, Q171, N176, 1177, S180, E181, N183, 15 A184, Y185, T187, and V189. For clarity these positions are numbered according to Int:190 and marked by asterisks in the sequence aligrunent figure (Figure 4c). Principally, this region is located at the tip of the structure and forms a long thin surface that is centred on the upper, solvent exposed surface of the second P-sheet (strands C, D and E).

This putative binding site is highly localised, covering an area on intimin that measures 20 approximately 20 A by 8A. This data is also completely reproducible with Int280 (data not shown), which suggests that residues downstream of IntI90 are unlikely to be 41 involved in Tir-binding. The highly localised nature of this region with the three dimensional structure suggests that the portion of Tir in contact with intimin is likely to be shorter than the 55 amino acids currently proposed. Based on the structure, we hypothesise that the number of amino acid residues of Tir in contact with intimin may be 5 as short as 5, if the conformation is completely extended, or as many as 14 if helical.

Site-directed mutagenesis of Intimin Interestingly, based on the structural alignment provided by DALI (Figure 4), W150 (within Intl9O) is highly conser-ved in all intimin types, invasin and CTLDs.

Although W150 is buried, it is located just below surface residues involved in binding 10 and may therefore provide a structural scaffold for the Tir binding site. ID '11 NMR spectra of the W1 50A mutant (data not shown) indicate that this protein is stable and fully folded and therefore any effects of the modification are entirely due to localised structural effects and not global destabilisation. Moreover, W150 shows chemical shift and line width perturbations of its backbone and side-chain amide in the presence of Tir. This 15 residue was therefore chosen for mutation and extensive functional analysis, in tenns of intimin expression, A/E lesion formation and Tir binding. This residue in the intact intimin (W899) was subjected to site-directed mutagenesis using the cloned eae encoding intimin (pCVD438, Donnenberg and Kaper, 1991) generating plasmid pICC53 (Table 1).

The site-directed mutated eae clone was introduced into the eae mutant EPEC strain 20 CVD206 (Donnenberg and Kaper, 1991), and its biological activities was characterised in terms of intimin expression, the ability to support A/E lesion formation and binding to Tir.

Western blots of whole bacterial cell extracts (data not shown) and immunofluorescence microscopy of infected BEp-2 cells confirmed that the mutation introduced 5 into the cloned eae gene did not affect intimin expression or association with the bacterial outer membrane (Figure 6). The ability of CVD206(pICC53) to induce A/E lesions on cultured HEp-2 cells was investigated using the fluorescent actin staining (FAS) test as a marker for lesion formation. Replacement of W899 with alanine (PICC53) did not affect surface expression of intimin but resulted in adherent bacteria that are unable to initiate 10 host cytoskeletal rearrangements (Figure 6). CVD206(pCVD438) produced both FAS and intimin positive staining while CVD206 alone gave a double negative reaction (Figure 6).

To ascertain whether or not this FAS negative phenotype was due to impaired binding to Tir, gel overlay and yeast two-hybrid system binding assays were employed.

15 In contrast to the fusion protein IMP-IntI90, which showed a clear interaction with TirM in gel overlay experiments (Figure 7a), no interaction could be detected with MBPIntI90 or NOP-Int280 containing the modification (MBP-Intl9OWl5OA, Figure 7a and MBP-Int28OW240A, data not shown). For the yeast two-hybrid system, DNA encoding Intl9OW150A and Int28OW240A were cloned into pGAD424 (plasmid pICC44 and 20 pICC62 respectively) while plasmid pICC10, expressing Tir, was used as the bait. Replica plating onto selective media produced vigorously growing colonies in yeast strains co-transformed with pICCIO and plCC39 (encoding the wild type IntI90). However, co-transformation of pICCIO with pICC44 (encoding Intl9OW150A) or pICC62 (encoding Int28OW240A) did not support growth on the selective media. Measuring the P-galactosidase activities in these strains revealed background levels 5 (Figure 7b), while strains harbouring pICC 10 and pICC39 produced elevated levels of Pgalactosidase activity. (Figure 7b)

The results taken together are entirely consistent and provide further evidence for the conclusions of the NMR titration experiments. In summary, altering W150 in IntI90 or W240 in Int280 (W899 in intact intimin) prevents EPEC from forming actin pedestals

10 by specifically disrupting the intimin-Tir interaction. This mutation lies proximal to the Tir-recognition region, which comprises a 20 A by 8A patch of residues within the Cterminal 100 amino acids (134).

Intimin is a bacterial adhesion molecule involved in intimate attachment of enteropathogenic Escherichia coli to manimalian host cells. Intimin targets the 15 translocated intin-iin receptor (Tir), which is exported by the bacteria and integrated into the host cell plasma membrane. In this study, we have localised the Tir-binding region of intimin to the C-terminal 190 amino acids (Intl9O). We have also determined the highresolution solution structure of this region, which comprises an immunoglobulin domain that is intimately coupled to a novel C-type lectin domain. This fragment, which is 20 necessary and sufficient for Tir interaction, defines a new super domain in intimin that exhibits striking structural similarity to the integrin-binding domain of the Yersinial invasion and the C-type lectin family. The extracellular portion of intimin comprises an articulated rod of immunoglobulin domains that extend from the bacterium surface conveying a highly accessible 'adhesive tip' to the target cell. In addition, the interpretation of NMR-titration and mutagenesis data has enabled us to identify, the first 5 time, the binding site for Tir, which is located at the extremity of the Intl 90 moiety.

This work represents the first report of high-resolution structure information for intimin. We also provide compelling evidence for a Tirbinding site at the extremity of intimin, which extends from the bacterial membrane. Figure 5d illustrates a schematic model for the structure of the intimin-Tir peptide complex. The results will aid the 10 development of preventative therapies against EPEC.

Screening Methods A preferred screening method involves coating beads, such as Covaspheres with the Tir binding site of the invention (eg as described in Frankel et al (May 1994) Infection and 15 Immunity 62, 5, p18 35-1842); exposing the coated beads to an intestinal cell extract or other test agent; and recovering agents which bind the coated beads. The agents are then tested for their ability to affect the binding of intimin to mammalian cells, preferably intestinal epithelial cells.

Skilled persons will appreciate that a variety of affinity purification techniques can be used to obtain agents which bind the Tir binding site of the invention, for example as described below.

5 Exemplary compositions of the invention Antibody production method Methods for purification of antigens and antibodies are described in Scopes, R.K. (1993) Protein pupiJi'cation 3rd Edition. Publisher Springer Verlag. ISBN 0-387-94072-3 and 10 3-540-94072-3. The disclosure of that reference, especially chapters 7 and 9, is incorporated herein by reference.

Antibodies may be produced in a number of ways.

15 For polyclonal antibodies, this is simply a matter of injecting suitably prepared sample into the animal at intervals, and testing its serum for the presence of antibodies (for details, see Dunbar, B.S. & Schwoebel, E.D. (1990) Preparation of polyclonal antibodies. Methods Enzymol. 182, 663-670). But it is essential that the antigen (ie. the protein of interest) be as pure as possible. For monoclonal antibodies, the purity of the antigen is relatively 20 unimportant if the screening procedure to detect suitable clones uses a bioassay.

Antibodies can also be produced by molecular biology techniques, with expression in bacterial or other heterologous host cells (Chiswell, D.J. & McCafferty, J. (1992) Phage antibodies: will new "coli-clonal" antibodies replace monoclonal antibodies?" Trends Biotechnol. 10, 80-84). The purification method to be adopted will depend on the source 5 material (serum, cell culture, bacterial expression culture, etc.) and the purpose of the purification --- (research, diagnostic investigation, commercial production). The major methods are as follows:

1. Ammonium sulphate precipitation. The y-globulins precipitate at a lower concentration than most other proteins, and a concentration of 33% saturation is sufficient. Either dissolve in 200g ammonium sulphate per litre of serum, or add 0.5 vol of saturated arnmorAum sulphate. Stir for 30 minutes, then collect the y globulin fraction by centrifugation, redissolve in an appropriate buffer, and remove excess ammonium sulphate by dialysis or gel filtration.

2. Polyethylene glycol precipitation. The low solubility of y-globulins can also be exploited using PEG. Add 0. 1 vol of a 5 0% solution of PEG 6,000 to the serum, stir for 30 minutes and collect the y-globulins by centrifugation. Redissolve the precipitate in an appropriate buffer, and remove excess PEG by gel filtration on a column that fractionates in a range with a minimum around 6,000 Da.

47 3. Isoelectric precipitation. This is particularly suited for IgM molecules, and the precise conditions will depend on the exact properties of the antibody being produced.

5 4. Ion-exchange chromatography. Whereas most serum proteins have low isoelectric points, y-globulins are isoelectric around neutrality, depending on the exact properties of the antibody being produced. Adsorption to cation exchangers in a buffer of around pH 6 has been used successfully, with elution with a salt gradient, or even standard saline solution to allow immediate therapeutic use.

5. Hydrophobic chromatography. The low solubility of y-globulins reflects their relatively hydrophobic character. In the presence of sodium or ammonium, sulphate, they bind to many hydrophobic adsorbents, such as 'T-gel" which consists of mercaptoethanol coupled to divinyl sulphone-activated agarose.

6. Affinity adsorbents. Staphylococcus aureus Outer coat protein, known as Protein A, is isolated from the bacterial cells, and it interacts very specifically and strongly with the invariant region (F) of immunoglobulins (Kessler, S.W. (1975) Rapid isolation of antigensfrom cells with a staphylococcal protein A-antibody absorbent:

20 Parameters of the interaction of antibody-antigen complexes with protein A. J Inimunol, 115, 1617-1624. Protein A has been cloned, and is available in many different forms, but the most useful is as an affi:nity column: Protein A coupled to agarose. A mixture containing immunoglobulins is passed through the column, and only the immunoglobulins adsorb. Elution is carried out by lowering the pH; different types of IgG elute at different pHs, and so some trials will be needed each 5 time. The differences in the immunoglobulins in this case are not due so much to the antibody specificity, but due to different types of F,; region. Each animal species produces several forms of heavy chain varying in the F, region; for instance, mouse immunoglobulins include subclasses IgG, lgG2a, and IgG, all of which behave differently on elution from Protein A.

Some -y-globulins do not bind well to Protein A. An alternative, Protein G from G from a Streptococcus sp., can be used. This is more satisfactory with immunoglobulins from farm animals such as sheep, goats and cattle, as well as with certain subclasses of mouse and rabbit IgGs. The most specific affinity adsorbent is the antigen itself The process of purifying an antibody on an antigen adsorbent is essentially the same as purifying the antigen on an antibody adsorbent. The antigen is coupled to the activated matrix, and the antibody-containing sample applied. Elution requires a process for weakening the antibody-antigen complex. This is particularly useful for purifying a specific 20antibody from a polyclonal mixture.

49 Monoclonal antibodies (MAbs) can be prepared to most antigens. The antigen-binding portion may be a part of an antibody (for example a Fab fragment) or a synthetic antibody fragment (for example a single chain Fv fragment [ScFv]). Suitable monoclonal antibodies to selected antigens may be prepared by known techniques, for example those disclosed in 5 "Monoclonal Antibodies: A manual of techniques", H Zola (CRC Press, 1988) and in "Monoclonal Hybridoma Antibodies: Techniques and Applications ", J G R Hurrell (CRC Press, 1982).

Chimaeric antibodies are discussed by Neuberger et al (1988, 8th International 10 Biotechnology Symposium Part 2, 792-799).

Suitably prepared non-human antibodies can be "humanized" in known ways, for example by inserting the CDR regions of mouse antibodies into the framework of human antibodies.

15 The variable heavy (VH) and variable light (VL) domains of the antibody are involved in antigen recognition, a fact first recognised by early protease digestion experiments. Further confirmation was found by "humanisation" of rodent antibodies. Variable domains of rodent origin may be fused to constant domains of human origin such that the resultant antibody retains the antigenic specificity of the rodent parental antibody (Morrison et al 20 (1984) Proc. Nad. Acad. Sci. USA 81, 6851-6855).

That antigenic specificity is conferred by variable domains and is independent of the constant domains is known from experiments involving the bacterial expression of antibody fragments, all containing one or more variable domains. These molecules include Fab-like molecules (Better et al (1988) Science 240, 1041); Fv molecules (Skerra et al (1988) 5 Science 240, 103 8); single-chain Fv (ScFv) molecules where the VHand VL partner domains are linked via a flexible oligopeptide (Bird et al (1988) Science 242, 423; Huston et al (1988) Proc. Nad. Acad. Sci. USA 85, 5879) and single domain antibodies (dAbs) comprising isolated V domains (Ward et al (1989) Nature 341, 544). A general review of the techniques involved in the synthesis of antibody fragments which retain their specific io binding sites is to be found in Winter & Milstein (199 1) Nature 349, 293-299.

By "ScFv molecules" we mean molecules wherein the V. and VL partner domains are linked via a flexible oligopeptide.

The advantages of using antibody fragments, rather than whole antibodies, are several-fold.

15 The smaller size of the fragments may lead to improved phannacological properties, such as better penetration of solid tissue. Effector functions of whole antibodies, such as complement binding, are removed. Fab, Fv, ScFv and dAb antibody fragments can all be expressed in and secreted from E. coli, thus allowing the facile production of large amounts of the said fragments.

Whole antibodies, and F(abl)2 fragments are "bivalenf'. By "bivalent" we mean that the said antibodies and F(ab)2 fragments have two antigen combining sites. In contrast, Fab, Fv, ScFv and dAb fragments are monovalent, having only one antigen combining site.

5 Raising an antibody response in a patient Active immunisation of a patient is preferred. In this approach, one or more Tir binding site containing intimin peptides or full length intimin or agents of the invention are prepared in an immunogenic formulation containing suitable adjuvants and carriers and administered to the patient. Suitable adjuvants include Freund's complete or incomplete adjuvant, muramyl i o dipeptide, the "Iscoms" of EP 109 942, EP 180 564 and EP 231039, aluminium hydroxide, saponin, DEAE-dextran, neutral oils (such as miglyol), vegetable oils (such as arachis oil), liposomes, Pluronic polyols or the Ribi adjuvant system (see, for example GB-A-2 189 141). 'Tluronic" is a Registered Trade Mark.

15 In order to select antibodies which bind all or part of the Tir binding site of intimin 'HYN HSQC NMR techniques can be used to identify amide resonances which move or broaden with and without saturating amounts of the antibody and thereby indicate intimin residues involved in binding the antibody. If residues are the same as one or more of the residues involved in binding the Tir55 pcpfide, the antibody is predicted to bind all or part of the Tir 20 binding site of intimin and thereby find utility in the prevention, treatment and/or diagnosis of disease.

Exemplary pharmaceuticalformulations of the invention The formulations may conveniently be presented in unit dosage form and may be prepared by any of the methods well known in the art of pharmacy. Such methods include the step of 5 bringing into association the active ingredient (compound of the invention including antibody) with the carrier which constitutes one or more accessory ingredients. In general the formulations are prepared by uniformly and intimately bringing into association the active ingredient with liquid carriers or finely divided solid carriers or both, and then, if necessary, shaping the product.

Whilst it is possible for an agent eg compound of the invention to be administered alone, it is preferable to present it as a pharmaceutical formulation, together with one or more acceptable carriers. The carrier(s) must be "acceptable" in the sense of being compatible with the agent of the invention and not deleterious to the recipients thereof Typically, the 15 carriers will be water or saline which will be sterile and pyrogen free.

Formulations in accordance with the present invention suitable for oral administration may be presented as discrete units such as capsules, cachets or tablets, each containing a predetermined amount of the active ingredient; as a powder or granules; as a solution or a 20 suspension in an aqueous liquid or a non-aqueous liquid; or as an oil- in-water liquid emulsion or a water-in-oil liquid emulsion. The active ingredient may also be presented as a bolus, electuary or paste.

A tablet may be made by compression or moulding, optionally with one or more accessory ingredients. Compressed tablets may be prepared by compressing in a suitable machine the active ingredient in a free-flowing form such as a powder or granules, optionally mixed 5 with a binder (eg povidone, gelatin, hydroxypropyhnethyl cellulose), lubricant, inert diluent, preservative, disintegrant (eg: sodium starch glycolate, crosslinked povidone, crosslinked sodimn carboxymethyl cellulose), surfaceactive or dispersing agent. Moulded tablets may be made by moulding in a suitable machine a mixture of the powdered compound moistened with an inert liquid diluent. The tablets may optionally be coated or 10 scored and may be formulated so as to provide slow or controlled release of the active ingredient therein using, for example, hydroxypropylmethylcellulose in varying proportions to provide a desired release profile.

Formulations suitable for topical administration in the mouth include lozenges comprising 15 the active ingredient in a flavoured base, usually sucrose and acacia or tragacanth; pastilles comprising the active ingredient in an inert base such as gelatin and glycerin, or sucrose and acacia; and mouth- washes comprising the active ingredient in a suitable liquid carrier.

Formulations suitable for parenteral administration include aqueous and non-aqueous sterile 20 injection solutions which may contain anti-oxidants, buffers, bacteriostats and solutes which render the formulation isotonic with the blood of the intended recipient; and aqueous and non-aqueous sterile suspensions which may include suspending agents and thickening agents. The formulations may be presented in unit-dose or multi-dose containers, for example sealed ampoules and vials, and may be stored in a freeze-dried (lyophilised) condition requiring only the addition of the sterile liquid carrier, for example water for injections, immediately prior to use. Extemporaneous injection solutions and suspensions 5 may be prepared from sterile powders, granules and tablets of the kind previously described.

Preferred unit dosage formulations are those containing a daily dose or unit, daily sub-dose or an appropriate fraction thereof, of an active ingredient.

i o It should be understood that in addition to the ingredients particularly mentioned above the formulations of this invention may include other agents conventional in the art having regard to the type of formulation in question, for example those suitable for oral administration may include flavouring agents.

15 The following examples illustrate pharmaceutical formulations according to the invention in which the active ingredient is selected from one or more of antibodies and agents eg compounds of the invention:

ExaMple A: Tablet Active ingredient 100 mg 5 Lactose 200 mg Starch 50 mg Polyvinylpyrrolidone 5 mg Magnesium stearate, 4 mg 10 359 mg Tablets are prepared from the foregoing ingredients by wet granulation followed by compression.

15 Ex4mple B: Tablet Formulations The following formulations A. B and C are prepared by wet granulation of the ingredients with a solution of povidone, followed by addition of magnesium stearate, and compression.

Fonnulation A mg/tabl mg/tabl (a) Active ingredient 250 250 (b) Lactose B.P. 210 26 5 (c) Povidone B.P. 15 9 (d) Sodium Starch Glycolate 20 12 (e) Magnesium Stearate 5 3 500 300 Formulation B mg/table mg/t (a) Active ingredient 250 250 (b) Lactose 150 - 15 (c) Avicel PH 1010 60 26 (d) Povidone B.P. 15 9 (e) Sodium Starch Glycolate 20 12 (f) Magnesium Stearate, 5 3 20 500 300 I Formulation C mg/tablet

Active ingredient 100 5 Lactose 200 Starch 50 Povidone 5 Magnesium stearate 4 10 359 The following formulations, D and E, are prepared by direct compression of the admixed ingredients. The lactose used in formulation E is of the direct compression type.

15 Formulation D mg/capsule Active Ingredient 250 Pregelatinised Starch NF 15 150 20 400 Formulation E mg/capsul Active Ingredient 250 Lactose 150 5 Avicel 100 500 Formulation F (Controlled Release Formulation) The formulation is prepared by wet granulation of the ingredients (below) with a solution of povidone followed by the addition of magnesium stearate and compression.

mg/table (a) Active Ingredient 500 15 (b) Hydroxypropylmethylcellulose 112 (Methocel K4M Premium)@ (c) Lactose B.P. 53 (d) Povidone B.P.C. 28 (e) Magnesium Stearate 7 700 Drug release takes place over a period of about 6-8 hours and is generally complete after 12 hours.

5 Example C: Capsule Formulations Formulation A A capsule formulation is prepared by admixing the ingredients of Formulation D, in 10 Example C above and filling into a two-part hard gelatin capsule. Formulation B (infta) is prepared in a similar manner.

Formulation B mg/capsule 15 (a) Active ingredient 250 (b) Lactose B.P. 143 (c) Sodium Starch Glycolate 25 (d) Magnesium Stearate 2 20 420 Formulation C mg/capsul (a) Active ingredient 250 (b) Macrogol 4000 BP 350 600 Capsules are prepared by melting the Macrogol 4000 BP, dispersing the active ingredient in the melt and filling the melt into a two-part hard gelatin capsule.

Formulation D mg/capsule Active ingredient 250 Lecithin 100 15 Arachis Oil 100 450 Capsules are prepared by dispersing the active ingredient in the lecithin and arachis oil and 20 filling the dispersion into soft, elastic gelatin capsules.

Foimulation E (Controlled Release Capsule) The following controlled release capsule fon-nulation is prepared by extruding ingredients a, 5 b, and c using an extruder, followed by spheronisation of the extrudate and drying. The dried pellets are then coated with release-controlling membrane (d) and filled into a two- piece, hard gelatin capsule.

mg/capsule (a) Active ingredient 250 10 (b) Nficrocrystalline Cellulose 125 (c) Lactose BP 125 (d) Ethyl Cellulose 13 513 Exarnple D: Injectable Formulation Active ingredi 0.200 g Sterile, pyrogen. free phosphate buffer (pH7.0) to 10 ml The active ingredient is dissolved in most of the phosphate buffer (35- 40'C), then made up to volume and filtered through a sterile micropore filter into a sterile 10 ml amber glass vial (type 1) and sealed with sterile closures and overseals.

10 Example E: Intramuscular injectio Active ingredient 0.20 g Benzyl Alcohol 0.10 g GlucofLirol 750 1.45 g 15 Water for hijection q.s. to 3.00 ml.

The active ingredient is dissolved in the glycofurol. The benzyl alcohol is then added and dissolved, and water added to 3 ml. The mixture is then filtered through a sterile inicropore filter and sealed in sterile 3 ml glass vials (type 1).

Example F: Syrup Suspension Active ingredient 0.2500 g Sorbitol Solution 1.5000 g 5 Glycerol 2.0000 g Dispersible Cellulose 0.0750 g Sodium Benzoate 0.0050 g Flavour, Peach 17.42.3169 0.0125 ml Purified Water q.s. to 5.0000 ml The sodium benzoate is dissolved in a portion of the purified water and the sorbitol solution added. The active ingredient is added and dispersed. The thickener (dispersible cellulose) is dispersed in the glycerol. The two dispersions are mixed and made up to the required volume with the purified water. Further thickening is achieved as required by extra 15 shearing of the suspension.

Example G: Supposi mg/supposi Active ingredient (63 pm) 250 Hard Fat, BP (Witepsol H15 - Dynamit Nobel) 1770 2020 The active ingredient is used as a powder wherein at least 90% of the particles are of 63 pm. diameter or less.

One fifth of the Witepsol H15 is melted in a steam-jacketed pan at 45'C maximum. The active ingredient is sifted through a 200 pm sieve and added to the molten base with mixing, using a silverson fitted with a cutting head, until a smooth dispersion is achieved. Maintaining the mixture at 451C, the remaining Witepsol H15 is added to the suspension 15 and stirred to ensure a homogenous mix. The entire suspension is passed through a 250 PM stainless steel screen and, with continuous stirring, is allowed to cool to 40'C. At a temperature of 38'C to 40'C, 2. 02 g of the mixture is filled into suitable plastic moulds. The suppositories are allowed to cool to room temperature.

Example H: Pessaries mg/pess Active ingredient 250 Anhydrous Dextrose 380 5 Potato Starch 363 Magnesium Stearate 7 1000 io The above ingredients are mixed directly and pessaries prepared by direct compression of the resulting mixture.

Use in medicine The aforementioned active agents eg compounds of the invention or a formulation thereof 15 may be administered in a variety of ways, for non-limiting example, by any conventional method including oral and parenteral (eg subcutaneous or intramuscular) injection. The treatment may consist of a single dose or a plurality of doses over a period of time, depending on the characteristics of the patient and/or the state of the particular bacterial disease against which the treatment is directed.

66 Diagnosis of disease The agents and other compounds of the invention may also find utility as diagnostic agents. For example, the agent or other compound can be labelled with a detectable label (eg a radiolabel) and used in the diagnosis of bacterial infections. Skilled persons will 5 appreciate that the agents and other compounds of the invention can readily be provided for use in ELISA techniques.

Prevention of disease The agents of the invention may find particular utility in the prevention of bacterial 10 infections. For example, the agents can be administered to individuals at particular risk of exposure to bacterial infections. Such risks may arise when an individual is likely to, or has already, come into contact with an affected individual.

It will be appreciated that the agents of the invention can be used to treat humans and 15 animals.

References Adu-Bobie, J., Frankel, G., Bain, C., Goncaleves, A. G., Trabulsi, L. R., Douce, G., Knutton, S., and Dougan, G. (1998) Detectionof intimin a, P, y, and 8, four intimin derivatives expressed by attaching and effacing microbial pathogens. J Clin Microbiol 5 36, 662-668.

Bax, A. Clore, G. M., & Gronenborn, A. M. J. (1990) 111-111 correlation via isotropic mixing of 13C magnetization, a new three-dimensional approach for assigning 1H and 13C spectra of 13C enriched proteins Magn. Reson. 88, 425-43 1.

Boyington, J. G., Riaz, A. N., Pataniawain, A., Coligan, J., E., Brook, A. G. and Sun, P.

10 D. (1999) Structure of the CD94 reveals a novel C-type lectin fold: implications for the NK cell-associated CD94/NKG2 receptors. Immunity 10, 75-82.

Brunger, A. T. (1993).XPLOR Manual Ver 3.1 Yale University Press New Haven.

Cornilescu, G., Delaglio, F. and Bax A. (1999) Protein backbone angle restraints from searching a database for chemical shift and sequence homology. J Biomol, NMR 13, 15 289-302.

de Grado, M., Abe, A., Gauthier, A., Steele-Mortimer, 0., DeVinney, R., and Finlay, B. B. (1999) Identification of the intimin binding domain of Tir of enteropathogenic Escherichia coli. Cell Microbiol 1, 7-18.

Deibel, C., Kramer, S., Chakraborty, T., and Ebel, F. (1998) EspE, a novel secreted protein of attaching and effacing bacteria, is directly translocated into infected host cells, where it appears as a tyrosinephospborylated 90 kDa. protein. Mol Microbiol 28,463-474.

Derscb, P. and Isberg, R. R. (1999) A region of Yersinia Pseudotuberculosis invasin protein enhance integrin-mediated uptake into mammalian cells and promotes selfassociation. EMBO J 181,1199-1213.

Donnenberg, M. S., and Kaper, J. B. (1991) Construction of an eae deletion mutant of enteropathogenic Escherichia coli by using a positive- selection 10 suicide vector. Infect Immun 59: 4310-4317.

Donnenberg, M. S., Tacket, C. 0., James, S. P., Losonsky, G., Nataro, J. P., Wasserman, S. S., Kaper, J. B., and Levine, M. M. (1993a) Role of the eaeA gene in experimental enteropathogenic Escherichia coli infection. J Clin Invest 92, 1412-1417.

Donnenberg, M. S., Tzipori, S., McKee, M. L., O'Brien, A. D., Alroy, J., and Kaper, J. B. 15 (1993b) The role of the eae gene of enterohemorrhagic Escherichia coli in intimate attachment in vitro and in a porcine model. J Clin Invest 92, 1418-1424.

Donnenberg, M. S., Yu, J., and Kaper, J. B. (1993c) A second chromosomal gene necessary for intimate attachment of enteropathogenic Escherichia coli to epithelial cells. JBacteriol 175, 4670-4680.

Ebel, F., Podzadel, T., Rohde, M., Kresse, A. U., Kramer, S., Deibel, C., Guzman, C. A., and Chakraborty, T. (1998) Initial binding of Shiga toxin-producing Escherichia coli to host cells and subsequent induction of actin rearrangements depend on filamentous EspA-containing surface appendages. Mol Microbiol 30, 147-16 1.

5 Frankel, G., Candy, D. C., Everest, P., and Dougan, G. (1994) Characterization of the C- terminal domains of intimin-like proteins of enteropathogenic and enterohemorrhagic F-scherichia coli, Citrobacterfreundii, and Hafnia alvei. Infect Immun 62, 1835-1842.

Frankel, G., Candy, D. C., Fabiani, E., Adu-Bobie, J., Gil, S., Novakova, M., Phillips, A.

D., and Dougan, G. (1995) Molecular characterization of a carboxyterminal 10 eukaryotic-cell- binding domain of intimin from enteropathogenic Escherichia coli.

Infect Immun 63,4323-4328.

Frankel, G., Lider, 0., Hershkoviz, R., Mould, A. P., Kachalsky, S. G., Candy, D. C. A., Cahalon, L., Humphries, A J., and Dougan, G. (1996) The cell-binding domain of intimin from enteropathogenic Escherichia coli binds to 01 integrins. J Biol Chem 15 271, 20359-20364.

Frankel, G., Phillips, A. D., Rosenshine, L, Dougan, G., Kaper, J. B., and Knutton, S.

(1998a) Enteropathogenic and enterohaemorrhagic Escherichia coli: More Subversive Elements. Mol Microbiol 30, 911-921.

* Frankel, G., Philips, A. D., Novakova, M., Batchelor, M., I-licks, S., and Dougan, G. (1998b) Generation of Escherichia coli intimin-derivatives with differing biological activities using site-directed mutagenesis of the intimin C-terminus domain. Mol Microbiol 29, 559-570.

5 Graves, B.J., Crowther, R.L., Chandran, C., Rumberger, J.M., Li, S., Huang, K.S., Presky, D.H., Familletti, P.C., Wolitzky, B.A. and Bums, D.K. (1994). Insight into e selectin ligand interaction from the crystal-structure and mutagenesis of the lec EGF domains. Nature 367, 532-538.

Gronwald, W., Loewen, M. C., Lix, B., Daugualis, A. J., Sonnichsen, F. D., Davies, P. L.

10 and B. D. Sykes. (1998) The solution structure of type H antifreze protein reveals a new member of the lectin family. Biochemistry 37, 4712-4721.

Grzesiek, S. and Bax, A. (1992a). An efficient experiment for sequential backbone assigriment of medium-sized isotopically enriched proteins. J. Magn. Reson. 99, 201 207.

15 Grzesiek, S. and Bax, A. (1992b). Improved 3D triple resonance NMR techniques applied to a 31 kDa protein. J Magn. Reson. 96, 432-440.

Hamburger, Z. A., S., B. M., Isberg, R. R., and 13jorkman, P. J. (1999) Crystal structure of invasin: a bacterial integrin-binding protein. Science. In Press.

Hartland, E. L., Batchelor, M., Delahay, R. M., Hale, C., Matthews, S., Dougan, G., 20 Knutton, S., Connerton, I., and Frankel, G. (1999) Binding of intimin from 71 enteropathogenic Eschefichia coli to Tir and to host cells. Mol Microbiol 32, 151 158.

Hicks, S., Frankel, G., Kaper, J. B., Dougan, G., and Phillips, A. D. (1998) Role of intimin and bundle fonling pili in enteropathgenic Escherichia coli adhesion to 5 paediatric intestine in vitro. Infect Immun 66, 1570-1578.

Higgins, L. M., Frankel, G., Connerton, I., Goncalves, N. S., Dougan, G., and MacDonald, T. T. (1999a) Role of bacterial intimin in colonic hyperplasia and inflammation. Science 285, 588-591.

Higgins, L. M., Frankel, G., Douce, G., Dougan, G., and MacDonald, T. T. (1999b) 10 Citrobacter rodentium infection in mice elicits a mucosal Thl cytokine response and lesions similar to those in murine inflammatory bowel disease. Infect Immun 67, 3031-3139.

Isberg, R. R., Voorhis, D. L., and Falkow, S. (1987) Identification of invasin: a protein that allows enteric bacteria to penetrate cultured mammalian cells. Cell 50, 769-778.

Isberg, R. R., and Leong, J. A (1990) Multiple beta I chain integrins are receptors for invasin, a protein that promotes bacterial penetration into mammalian cells. Cell 60, 861-871.

James, P., Halladay, J., and Craig, E. A. (1996) Genomic libraries and a host strain designed for highly efficient two- hybrid selection in yeast. Genetics 144, 1425-1436.

Jarvis, K. G., Giron, J. A., Jerse, A. E., McDaniel, T. K., Donnenberg, M. S., and Kaper, J. B. (1995) Enteropathogenic Escherichia coli contains a putative type III secretion system necessary for the export of proteins involved in attaching and effacing lesion formation. Proc Nad Acad Sci U S A 92, 7996-8000.

Jarvis, K. G., and Kaper, J. B. (1996) Secretion of extracellular proteins by enterohemorrhagic Escherichia coli via a putative type HI secretion system. Infect Immun 64,4826-4829.

Jerse, A. E., Yu, J., Tall, B. D., and Kaper, J. B. (1990) A genetic locus of enteropathogenic Escherichia coli necessary for the production of attaching and 10 effacing lesions on tissue culture cells. Proc Nad Acad Sci U S A 87, 7839-7843.

Kay, L.E., Xu, G.Y. and Yamazaki, T. (1994). Enhanced-sensitivity tripleresonance spectroscopy with minimal water saturation. J. Magn. Reson. 109, 129-133.

Kelly, G., Prasannan, S., Daniell, S., Frankel, G., Dougan, G., Connerton, L, and Mathews, S. (1998) Sequential assignment of the triple labelled 30.1 kDa cell 15 adhesion domain of intimin from enteropathogenic E. coli. J Biomolecular NMR 12, 189-191.

Kelly, G., Prasarman, S., Daniell, S., Fleming, K., Frankel, G., Dougan, G., Connerton, L, and Matthews, S. (1999) Structure of the cell-adhesion fragment of intimin from enteropathogenic Escherichia coli. Nat Struct Biol 6, 313-318.

Kenny, B. (1999) Phosphorylation of tyrosine 474 of the enteropathogenic Escherichia coli (EPEC) Tir receptor molecule is essential for actin nucleating activity and is preceded by additional host modifications. Mol Microbiol 33, 1229-1241.

Kenny, B., DeVinney, R., Stein, M., Reinscheid, D. J., Frey, E. A., and Finlay, B. B. 5 (1997) Enteropathogenic E. coli (EPEC) transfers its receptor for intimate adherence into mammalian cells. Cell 91, 511-520.

Kenny, B., Lai, L. C., Finlay, B. B., and Donnenberg, M. S. (1996) EspA, a protein secreted by enteropathogenic Eschefichia coli, is required to induce signals in epithelial cells. MolMicrobiol 20,.313-323.

10 Kohda, D., Morton, G., Parkar, A. A., Hatanaka, H., Inagaki, F. M., Campbell, I and Day, A. (1996). Solution structure of the link module: a hyaluronan-binding domain involved in extracellular matrix stability and cell migration. Cell 86, 767-775.

Knutton, S., Baldwin, P/ H., Williams, P. H. and McNeish, A. S. (1989) Actin accumulation at site of bacterial adhesion to tissue culture cells: basis of a new 15 diagnostic test for enteropathogenic and enterohaemorrhagic Escherichia coli. Infect.

Immun. 57,1290-1298.

Knutton, S., Rosenshine, I., Pallen, A J., Nisan, I., Neves, B. C., Bain, C., Wolff, C., Dougan, G., and Frankel, G. (1998) A novel EspA-associated surface organelle of enteropathogenic Eschefichia coli involved in protein translocation. into epithelial 20 cells. EMBO J 17, 2166-2176.

Lai, L. C., Wainwright, L. A., Stone, K. D., and Donnenberg, M. S. (1997) A third secreted protein that is encoded by the enteropathogenic Escherichia coli pathogenicity island is required for transduction of signals and for attaching and effacing activities in host cells. Infect Immun 65, 2211-2217.

Leong, J. M., Founiier, R. S., and Isberg, R. R. (1990). Identification of the integrin binding domain of the Yersinia pseudotuberculosis invasin protein. EMBO J 9, 19791989.

Leong, J. M., Morrissey, P. E., and Isberg, R. R. (1993) A 76-amino acid disulfide loop in the Yersinia pseudotuberculosis invasin protein is required for integrin receptor 10 recognition. JBiol Chem 268, 20524-20532.

Liu, H, Magoun, L, Luperchio, S, Schauer, D. B, Leong_jM (1999) The Tirbinding region of enterohaernorrhagic Escherichia coli intimin is sufficient to trigger actin condensation after bacterial-induced host cell signalling. Molecular Microbiology 34, 67-81.

McDaniel, T. K., Jarvis, K. G., Donnenberg, M. S., and Kaper, J. B. (1995) A genetic locus of enterocyte effacement conserved among diverse enterobacterial pathogens. Proc Natl Acad Sci U S A 92, 1664-1668.

Marion, D., Driscoll, P. C., Kay, L. E., Wingfield, P. T., Bax, A. Gronenborn, A. M. and Clore, G. M. (1989) Overcoming the overlap problem in the assignment of 1H NMR spectra of larger proteins by use of three-dimensional hetemuclear IH-15N Hartmann-Hahn multiple quantum coherence and nuclear Overhauser-multiple quantum coherence spectroscopy: Application to interleukin 1. Biochemistry 28, 6150-6156.

Miller, J. H. (1972) Experimental in molecular genetics. Cold Spring Harbour, NY:, Cold Spring Harbour Laborator Press.

Moon, H. W., Whipp, S. C., Argenzio, R. A., Levine, M. M., and Giannella, R. A. (1983) Attaching and effacing activities of rabbit and human enteropathogenic Escherichia coli in pig and rabbit intestines. Infect Immun 41, 1340-13 5 1.

lo Muhandiram, D. R. and Kay, L.E. (1994). Gradient-enhanced tripleresonance 3dimensional NMR experiments with improved sensitivity. J Magn. Reson. B103, 203-216.

Nataro, J. P., and Kaper, J. B. (1998) Diarrheagenic Escherichia coli. Clin Microbiol Rev 11,142-201.

Nicholls, A., Sharp, K. A. and Honig, B. (199 1) Protein folding and association - insight from the interfacial and thermodynarn properties of hydrocarbon. Proteins: Struct. Funct. & Gene. 11, 281-296.

Nilges, M. Gronenborn, A. M. and Clore, G. M. (1988) Determination of 3dimensional structures of proteins by simulated annealing with interproton distance restraints - 76 application to crambin, potato carboxypeptidase inhibitor and barley serine proteinase inhibitor-2. Protein Eng. 2, 27-38.

Norwood, T.J., Boyd, J., Heritage, J.E., Soffe, N. and Campbell, I.D. (1990) Comparison of techniques for 'H-detected heteronuclear 'HYN spectroscopy. J Magn. Reson 87, 5 488-501.

Oswald, E., Schmidt, H., Morabito, S., Karch, H., Marches, 0. and Caprioli, A. (2000) Typing of intimin genes in human and animal enterohernorrhagic and enetropathogenic Escherichia coli: chracterisation of a new intimin variant. Infection and Immunity 68, 64-7 1.

Perna, N. T., Mayhew, G. F., Posfai, G., Elliott, S., Donnenberg, M. S., Kaper, J. B., and Blattner, F. R. (1998) Molecular evolution of a pathogenicity island from enterohernorrhagic Escherichia coli 0157:117. Infect 1mmun 66,3810-3817.

Vuister, G.W., Clore, G.M., Gronenbom, A.M., Powers, R., Garrett, D.S., Tschudin, R. and Bax, A. (1993) Increased resolution and improved spectral quality in 4- 15 dimensional 11C/13 C-separated 11MQC-NOESY-11MQC spectra using pulsed field gradients. J Magn. Reson. B 101, 210-213.

Weis, W. I., and Drickamer, K. (1996) Structural basis of lectincarbohydrate recognition.

Annu Rev Biochem 65, 441-473.

77 Wolff, C., Nisan, I., Hanski, E., Frankel, G., and Rosenshine, 1. (1998) Protein translocation into HeLa cells by infecting enteropathogenic Escherichia coh. Mol Microbiol28,143-155.

Yu, J., and Kaper, J. B. (1992) Cloning and characterization of the eae gene of 5 enterohaemorrhagic Escherichia coli 0 1 57:H7. Mol Microbiol 6, 411-417.

Claims (20)

1. Use of the Tir binding site of intimin in a method for the design and/or selection of an agent capable of interfering with or preventing Tir-intimin binding, the Tir binding site being identifiable by comparing the 'HYN HSQC spectrum of an intimin molecule 5 with and without saturating amounts of the Tir55 peptide, amide resonances which move and/or broaden in the presence of the Tir55 peptide being indicative of intimin residues involved in binding Tir.

2. Use an claimed in Claim I wherein the Tir binding site comprises, or consists of, one or more, or all, of the following residues of lntl90; Y140, Y142, 1147, 1148, S149, 10 W150, T154, Q156, D157, A158, V162, A163, S164, T165, K170, Q171, W176, 1177, S180, E181, N183, A184, Y185, T187, and V189; or the corresponding residues of another intimin molecule.

3. Use as claimed in Claim 2 wherein the Tir binding site comprises or consists of 5 or more of the residues of IntI90, or the corresponding residues of another intimin 15 molecule.

4. Use as claimed in Claim 3 wherein the Tir binding site comprises or consists of 5 to 14 of the residues of Int 190, or the corresponding residues of another intimin molecule.

5. Use as claimed in any preceding claim wherein-the agent is designed in a method involving computer-aided chemical modelling techniques.

6. Use as claimed in any preceding claim wherein the agent is selected using an affinity purification technique.

5

7. Use as claimed in Claim 6 wherein the agent is a peptide or polypeptide.

8. Use as claimed in claim 7 wherein the polypeptide is an antibody or an antigen binding fragment thereof, the antibody or fragment being capable of specifically binding all or part of the Tir binding site of intimin.

9. A method of identifying an agent which is capable of binding all or part of the Tir io binding site of intimin comprising providing an intimin molecule containing the Tir binding site and comparing the 'H-"N HSQC spectrum with and without the agent, amide resonance which move or broaden in the presence of the compound being indicative of intimin residues involved in binding the agent and determining whether one or more of such residues are the same as the residues of intimin involved in binding Tir; any of 15 which are the same being indicative of an agent which is capable of binding all or part of the Tir binding site of intimin.

10. An agent capable of binding all or part of the Tir binding site of intimin, the agent being identifiable by a method or use as claimed in any one of Claims I to 9.

11. An agent as claimed in Claim 10 for use in medicine.

12. Use of an agent as claimed in claim 10 in the manufacture of a medicament for use in the prevention, treatment and/or diagnosis of a bacterial infection.

13. The use as claimed in Claim 12 wherein the bacterial infection causes an histopathologic effect on intestinal epithelial cells, known as attachment and effacement (A/E).

14. The use as claimed in Claim 13 wherein the bacterial infection comprises infection by one or more of enteropathogenic Exo1i (EPEC) and/or enterohaemorrhagic Exo1i (EHEQ, Shiga toxigenic Exo1i, Halvei, and Cfireundii.

15. The use as claimed in claim 14 wherein the bacterial infection comprises E. coli io 0157:117.

16. An agent as defined in any one of Claims 10 to 15 for use in the manufacture of a composition for use as a food supplement or as an additive of food.

17. An agent for use as claimed in claim 16 wherein the food is a milk substitute.

18. A food product comprising a foodstuff and an agent as defined in any - one of 15 claims 10 to 15.

19. A food product as claimed in Claim 18 wherein the food is adapted for consumption by animals.

20. A pharmaceutical composition comprising an agent as claimed in any one of Claims 10 to 15 together with a pharmaceutically acceptable diluent or carrier.