WO2001035981A1 - Methods and compositions for treatment of disease - Google Patents

Abstract

Disclosed are methods and compositions of matter useful in the treatment of infection and disease caused or mediated by non-eukaryotic pathogenic microorganisms. Microorganisms are characterized by regulation of one or more virulence genes by a protein containing a Sarcoma homology domain 3 region.

Description

METHODS AND COMPOSITIONS FOR TREATMENT OF DISEASE Priority

This application claims priority from United States Provisional Application No. 60/166,60, filed November 19, 1999, the contents of which are hereby incorporated by reference in their entirety. Technical Field

The present invention relates to the treatment of disease and infection caused by non-eukaryotic microorganisms, particularly bacteria and mycoplasma. Background of the Invention Infectious disease remains the largest cause of mortality in the world. A significant proportion of infectious disease-associated morbidity and mortality results from prokaryotic pathogens, particularly bacteria. The process and underlying mechanisms of the infectious process have been the subjects of intensive study for several decades.

Bacteria respond to nutritional stress by the coordinated expression of different genes. This facilitates their survival in different environments. Among these differentially regulated genes are the genes responsible for the expression of virulence determinants. The selective expression of these genes in a sensitive or susceptible host allows for the establishment and maintenance of infection or disease. Virulence genes include those which encode toxins, colonization factors and genes required for siderophores production or other factors that promote this process.

The expression of virulence genes in bacteria therefore enables the organism to invade, colonize and initiate an infection in humans and/or animals. These genes are not necessarily expressed constantly (constitutively), however. That is, the bacterium is not always orchestrating gene expression patterns to maximize "infectious" potential. In many circumstances, the expression of virulence genes is controlled by regulatory circuitry that include repressor proteins and a corresponding operon or operator. One class of repressors that are activated upon binding to or forming a complex with a transition metal ion such as iron, zinc or manganese, is thought to control the expression of a subset of genes in a number of Gram positive organisms. When such repressors are activated and associated with virulence gene expression in pathogens, they bind to operator sites thereby preventing production of virulence determinants. Virulence determinants are most often expressed when the bacterial pathogen is exposed to environmental stress such nutritional restriction. An iron-poor environment is an example of such a condition. In many eukaryotes, insufficient free iron is present to maintain the repressor in its active state. In the inactive form, the repressor cannot bind to target operators. As a result, virulence genes are de-repressed and the bacterium is able to initiate, establish, promote or maintain infection.

The expression of these virulence determinants is in many bacterial species co-regulated by metal ions. In most instances, the metal co-factor that is involved in vivo is iron but can include zinc, nickel, manganese and cobalt. In the presence of iron, the repressor is activated and virulence gene expression is halted.

This pattern of gene regulation is illustrated by the following example. The bacterium that causes diphtheria produces one of the most potent toxins known to man. The toxin is only produced under conditions of iron deprivation. In the presence of iron, the bacterial repressor (which in this species is known as diphtheria toxin repressor protein, abbreviated "DtxR") binds iron and undergoes conformational changes that activate it and allow it to dimerize and bind a specific DNA sequence called the tox operator. The tox operator is a specific DNA sequence found upstream of the gene that produces the diphtheria toxin, thereby preventing its expression. Typically, during infection of a human host the diphtheria bacillus (or other pathogenic/opportunistic bacteria) grows in an environment that rapidly becomes restricted in several key nutrients. Paramount among these essential nutrients is iron, and when iron becomes limiting the diphtheria bacillus begins to produce the toxin. Moreover, the constellation of virulence genes that DtxR controls become de-repressed and the diphtheria bacillus becomes better adapted to cause an infection. In the case of diphtheria, the toxin kills host cells thereby releasing required nutrients including iron.

Antibiotic therapy has been the accepted mode of treatment for bacterial infections and diseases. As a consequence of the widespread use and perhaps even misuse of antibacterial drugs, however, strains of drug-resistant pathogens have emerged. Antibiotic-resistant bacterial strains have been associated with a variety of infections, including tuberculosis, gonorrhea, staphylococcal and pneumococcal infections, and the bacteria most commonly associated with pneumonia, ear infections and meningitis. More importantly, infectious disease remains the largest cause of mortality in the world.

The typical response to an ineffective antibiotic has simply been change antibiotics. Unfortunately, this alternative no longer offers a guarantee of success. For example, certain strains of enterococci are resistant to vancomycin — a drug formerly considered as the ultimate weapon against many different types of bacteria. The World Health Organization has expressed concern that the development of new drugs is not keeping pace with the numbers of antibiotics that become ineffective. World Health Report 1996: Fighting Disease, Fostering Development, Executive Summary (World Health Organization 1996). Despite ongoing research, there remains a pressing need to develop new antibiotics. There is also a need for anti-bacterial agents that are effective in treating disease while not stimulating the emergence of resistant strains. Summary of the Invention

The present invention is directed to compositions and methods for treating infection and disease in mammals caused or mediated by non-eucaryotic pathogenic microorganisms such as bacteria and mycoplasma. The therapeutic agents administered to the mammals promote activation of a protein (such as a repressor protein) that regulates virulence gene expression in the pathogen. The activation of the protein results in attenuated or reduced infectiousness of the pathogens. The agents of the present invention are Sarcoma homology domain 3 (SH3) ligands; they bind SH3 domains present in the native proteins.

Proteins possessing SH3 domains are common in eucaryotes. They play a role in controlling the activity of certain enzymes that transmit signals between the eucaryotic cell and its external environment. As their name implies, they have received considerable interest as potential targets for the development of drugs to treat malignancies in humans. SH3 domains were not known to exist in prokaryotic or bacterial proteins, or for that matter, to help regulate virulence gene expression in prokaryotic pathogens. The therapeutic agents include peptides and non-peptides alike. Known agents that target the SH3 domain in eucaryotic proteins may be used in the present methods. Newly discovered SH3 ligands that contain a proline-rich peptide are also provided. Brief Description of the Figures Fig. 1. (a) Chemical shift deviation (CSD) between the measured H* chemical shift (14) and the random coil H" value (25). Clusters of positive and negative values suggest P- strands and helical structures, respectively, (b) Backbone ^l5N-{ Η} heteronuclear NOE for each residue. NOE values for some residues were not presented because of resonance overlap, (c) The number of proton-proton NOEs per residue (No. NOE/res). (d) The backbone rms deviation per residue of the 20 structures from the average structure. The helical (open ellipses) and β-strand (solid squares) regions of DtxR(130-226) are depicted at the top. Fig. 2. Structure of DtxR(130-226) with the unstructured residues 130-145 omitted. Stereoview of the backbone superposition of the 20 refined structures determined as described in the text. The N terminus of the structure is located at lower left. Fig. 3. Nondenaturing polyacrylamide gels indicating the degree of oligomerization of DtxR(130-226) (lane A) and DtxR(144-226) (lane B). The reduced electrophoretic mobility of DtxR( 144-226) monomer compared to DtxR(133-226) monomer reflects the additional residues at the N terminus of DtxR( 144-226) arising from the expression construct (see Materials and Methods).

Fig. 5. Stereoview of the C^α trace of the SH3-like domain of DtxR showing the residues implicated in peptide binding. Residues that shifted upon addition of the peptide are depicted as red balls. Best Mode of Carrying out the Invention

SH3 stands for Sarc homology domain 3 as the structure was originally identified in a protein kinase that when deregulated is associated with the development of sarcoma. It is well known that eukaryotic SH3 domains are important regulatory elements that function through the recognition of proline-rich motifs that specify distinct regulatory pathways important for cell growth, migration, differentiation, and responses to the external milieu. In general, the Src homology 3 (SH3) domain is a 50 amino acid modular element that is found in a number of eukaryotic non-receptor tyrosine kinases (e.g., Src, Fyn, Lyn, Yes, PI3K, Hck, Itk/Tsk). The SH3 domain has been proposed to provide a regulatory function in Src and related tyrosine kinases. In the inactive forms of Src (Xu et al, 1997), Hck (Sicheri et al, 1997) and Itk (Andreotti et al, 1997), the SH3 domain was found to be bound to an internal, proline-containing region that links the SH2 and catalytic domains and thereby stabilize the inactive form of the kinase. The SH2 domain is composed of three antiparallel beta-sheets with two shorter beta sheets, betaA and betaG. SH2 domains bind phospho-tyrosine-containing peptides having the sequence pTyr-Glu-Glu-ϋe. The SH3 domain is composed of a five stranded up-and-down antiparallel □ structure that is twisted into a barrel such that they form two anti-parallel sheets that pack against each other. While some SH3 domains have been shown to contain small regions of secondary structure, this fold is common to all known SH3 domains. SH3 domains specifically bind proline-rich peptides of approximately 10 amino acid residues in length. Two distinct classes of peptides have been described, namely class I (RXXPXXP) and class π (PXXPXR). Dalgarno & Kaye. These ligands each bind to SH3 domains in one of two pseudo- symmetrical orientations. This functional interaction between proline-rich peptides and SH3 domains from Src, Fyn, Lyn, Yes, PI3K, Hck, and Itk/Tsk kinases has been successfully examined by M13 phage display (Bunnell et al, 1996; Rickles et al, 1995; Schumacker et al, 1996; Sparks et al, 1995; Feng et al, 1995). In preferred embodiments of the present invention, the therapeutic agents are administered to mammals to treat caused or mediated by gram positive bacteria having virulence gene expression regulated, at least in part, by DtxR or a DtxR homolog. Applicants have established that the C-terminal domain of DtxR folds into an SH3 domain, and like its eucaryotic counterparts, binds proline-rich peptides. Applicants have also established that disruption of normal C-terminal SH3 domain function modulates DtxR activation. In other words, SH3 ligands with sufficiently high affinity promote activation of DtxR, which in turn, leads to suppression of virulence gene expression. Applicants have further established that DtxR homologs also possess SH3 domains and corresponding polyproline-rich docking sites. DtxR is a metal dependent repressor which under limiting concentrations of metal ions becomes inactivated permitting the derepression of a number of virulence genes including diphtheria toxin. The repressor contains a metal binding domain that binds iron and subsequently allows the dimerization of DtxR and repression of virulence gene expression in vivo. Manabe, et al., Proc. Natl. Acad. Sci. USA 96:12844-12848 (1996). More specifically, DtxR is a metal iron-dependent DNA-binding protein having a deduced molecular weight of 25,316 and which functions as a global regulatory element for a variety of genes on the C diphtheriae chromosome. See Tao, et al, Proc. Natl. Acad. Sci. USA 59:5897-5901 (1992); Schmitt, et al, Infect. Immun. 59:1899-1904 (1994). For example, DtxR regulates the expression of the diphtheria toxin structural gene (tox) in a family of closely related Corynebacteriophages. The DtxR gene has been cloned and sequenced in E. coli and its DNA and amino acid sequences have been reported. See Boyd, et al, Proc. Natl. Acad. Sci. USA 87:5968-5972 (1990); Schmitt, et al, supra. DtxR is activated by divalent transition metal ions (e.g., iron). Once activated, it specifically binds the diphtheria tox operator and other related palindromic DNA targets. See Ding, et al, Nature Struct. Biol. 3(41:382-387 (1996); Schiering, et al Proc. Natl. Acad. Sci. USA 92:9843-9850 (1995); White, et al, Nature 394:502-506 (1998). DNA sequences encoding DtxR from various C. diphtheria strains are defined by accession numbers M80336.M80337, M80338, and M34239.

DtxR homologs are prevalent in Gram-positive bacterial species, particularly those listed in Table 1. The diseases caused by the mycobacterial staphylococcal, and stieptococcal species are particularly preferred indications for the purposes of the present invention. Mycobacterium that cause significant disease include M. tuberculosis, M smegmatis and M. leprae.

TABLE I

S. pneumoniae S. agalactia S.equisimillis S. meningitis S. bovis S.anginosus S. pyogenes S. salivarius S. sanguis S. suis S. mutans Enterococcus faecalis

Staphylococcus species S. aureus S. epidermitis

Mycobacterium species M. tuberculosis M. avium complex M. kansasii M. leprae M. scrofulaceum M. fortuitum M. ulcerans M. marinum M. bovis M. microtii M. africanum

Actinomyces species A. pyogenes A. israelii A. bovis A. viscosus A. hordeovulneris A. gerencseriae A. naeslundii A. odontolyticus Listeria monocytogenes Proprionibacterium acnes Erysipelothrix rhusiopathiea

A collection of accession numbers for sequences that are either homologous to DtxR or contain a consensus tox O/P is presented in Table 2. See also http://www.ncbi.nlm.nih.gov/BLAST and http://www.ncbi.nlm.nih.gov/unfinishedgenomes.html. See also, Altschul, et al, J. Mol. Biol. 275:403-410 (1990); Gish, et al, Nature Genet. 3:266-272 (1993); Madden, et al, Meth. Enzymol. 266:131-141 (1996); Altschul, et al, Nucleic Acids Res. 25:3389-3402 (1997); and Zhang, et al, Genome Res. 7:649-656 (1997). This high degree of sequence similarity and homology indicates that the iron regulatory pathway that employs the DtxR- family of repressors is conserved in many important human and animal pathogens.

Table 2 DtxR Homologs and Species with DtxR Binding Sites

Pathogenic Human/Veterintary Applications Other

CAA67572 S. epidermitis L35906 C. glutamicum Gi 1777937 T. pallidum Z50048 S. pilosus

CAA15583 M. tuberculosis Z50049 S. lividans

U14191 M. tuberculosis U 14190 M. smegmatis

L78826 M. leprae M50379 M. jannaschi

M80336 C. diphtheriae Gi 2621260 M.thermoautotrophicum M80337 C. diphtheriae Gi 2622034 M. thermoautotrophicum

M34239 C. diphtheriae 033812 S. xylosus

M80338 C. diphtheriae Q57988 M. jannaschi

AAD18491 C. pneumoniae Gi 264870 Afulgidus

Gi 3328463 C. trachomatis Gi 2648555 Afulgidu TIGR 1280 S. aureus Gi 2650396 Afulgidus Stanford 382 S. meliloti Gi2650706 Afulgidus AE001439 H. pylori BAA79503 A. pernix TIGR 1752 V. cholera CAB49983.1 P. abyssi TIGR1097 C. tepidum BAA30263 P. horikoshi OUACGT S. pyogenes AL109974 S. coelicolor Sanger 518 B. bronchoseptica L35906 B. lactofermentum Sanger 1765 M. bovis AE000657 A. aeolius Sanger 520 B. pertusis TIGR 920 T. ferrooxidans WUGSC K. pneumoniea TIGR 76 C. crescentus TIGR 24 S. putrificacieus TIGR 1351 E. faecalis AE000783 B. burgdorferi TIGR1313 S. pneumoniea Sanger 632 Y. pestis

Table 3 depicts a sequence alignment that illustrates the high degree of conservation in DtxR type repressors from a number of clinically important species, including DtxR from Brevibacterium lactofermentum (Bl), DtxR from Corynebacterium diphtheriae (Cd); IdeR from Mycobacterium segmatus (Ms), IdeR from Mycobacterium tuberculosis (Mt); DesR from Streptomyces lividans (SI), DesR from Streptomyces pilsous (Sp) and SirR from Staphylococcus aureus (Sa). The consensus amino acid sequences between these members of the DtxR family of iron-dependent repressors is indicated. *, metal ion coordination residues m the Primary site; #, metal ion coordination residues in the Ancillary site; @, the single amino acid residue that interacts with a base in the binding of DtxR dimers to the tox operator. Grey area is the highly conserved iron and DNA binding domain. The C-terminal domains exhibit a high degree of structural homology and exhibit an extremely high degree of similarity in the functionally significant polyproline region (shown in black).

sasirR u ^^^ sm s^ s^^^ s^^m m ^^a^s CONSENSUS M-L-DTTEM YLRTI-LEE EGV-P-RARI AERL-QSGPT VSQTV-RMER DGL-V-DR

CONSENSUS -L--T-GR- LA- •VMR-R LAE-LL-D-I VH-E ACRWEHVMS- -VER — L

# #

CONSENSUS - — SP-GN PIPGL-EL-

190 200 210 220 230

Bl DtxR ALTDAGVEIG TEVDIINEQG RVVITHNGSS VELIDDLAHA VRVEKVEG Cd DtxR QLLDADIRVG SEVEIVDRDG HTTLSHNGKD VELLDDLAHT IRIEEL Ms IdeR IGRLKEAGVV PNARVTVEAN NNGGVMIVIP GHEQVELPHH MAHAVKKKVE KVEKV Mt IdeR ITRLKDAGVV PNARVTVETT PGGGVTIVIP GHENVTLPHE MAHAVKVEKV SI DesR QLMYTLRRAG VQPGSWS VT ESAGGVLVGS GGEAAELEAD TASHVFVAKR Sp DesR QLMYTLRRAG VQPGSWS VT EAAGGGVLVG SSGEAAELET DVASHVFVAK P Sa SirR VYLSSKDIYI GNTVEIVSKD DTNKVIILKR NDIVTILSYE NAMNIFAEK CONSENSUS

Sequence similarity among DtxR homologs is also reported in Schmitt, et al, Infect Immun.

63(77):4284-4289 (1995); Doukhan, et al, Gene 165(l):67-70 (1995); Oguiza, et al, J.

Bacteπol. 777(2 -465-467 (1995); Gunter, et al, J. Bacteriol. 775:3295-3302 (1993); and Schmitt, et al, Infect. Immun. 63:4284-4289 (1995).

Diseases caused or mediated by other non-eucaryotic pathogenic microorganisms, including Gram-negative bacteria and mycoplasma, and which are also characterized by SH3 domain-mediated modulation of virulence gene expression, are included within the scope of the present invention. Under physiological conditions in the presence of iron, the DtxR-type repressors are activated and suppress iron dependent gene regulation. When iron becomes limiting both in vitro and in vivo, the repressors surrender iron and undergo a conformational shift and deactivation, a process that involves the SH3 domain.

Deactivation of the repressors permits the expression of iron dependent genes which in many human and animal pathogens, encode virulence factors that promote the establishment, growth and maintenance of infection. Activation of the SH3 domain, and in turn, suppression of virulence gene expression leading to attenuation of infectiousness, may be achieved by displacement of the SH3 domain from its native or endogenous polyproline docking station, even in iron-poor environments. Therefore, compounds that mimic the endogenous polyproline sequence and/or bind the SH3 domain contained in the repressor with sufficient affinity to inhibit binding with the native docking station, are useful as therapeutic antimicrobial agents. The sequence of the SH3 docking site in DtxR that Applicants have identified is as follows: VSRSPSGNPIPGLDELGV. In more preferred embodiments, the therapeutic agents of the present invention contain a polyproline peptide sequence that reproduce potential recognition motifs for the SH3 domains of bacterial repressors. The peptides described below share common properties of the expanding library of proline peptides that appear to be involved in the regulation of protein associations in eukaryotic cells. These characteristics include the presence of one or more peptides within a hydrophobic stretch of amino acids often containing one or more possible phosphorylation sites, serine or threonine residues.

The sequences described below have been identified by distinct methods. By virtue of the techings in example 1, we have cloned and identified a number of DtxR homologues. Nucleotide sequence analysis and determination of primary amino acid sequence confirms the presence of a conserved proline containing sequence in region equivalent to amino acids 126-136 of DtxR. Each of these stretches of amino acids are putative SH3 ligands for the development of high affinity competitive agonists of DtxR or DtxR homologues. At least two distinct families of sequences can be obtained. Synthesis of a degenerative library of oligo-nucleotides encoding all possible derivations of amino acids or subspecies of this library can be created by in vitro synthesis. Such a library is systematically created by randomizing the addition of bases in an oligo-nucleotide library and then expressing these peptides in the PSDT system as described below in Example 3. A general method of preparing a randomized population of molecules based upon synthesized oligonucleotides is described in Park and Raines Nature Biotechnology (2000) Genetic Selection of for dissociative inhibitors of designed protein-protein interactions 18 847-851. This paper also presents a strategy that could be employed to screen for additional ligands that would bind to the DtxR SH3 domain.

Alternatively, a series of parental oligo-nucleotides encoding the conserved proline sequences in DtxR [or any homologue] may be created and cloned into a suitable expression vector such as those described below in Example 3. Alternatively, the peptide- encoding mingenes could also be cloned into a vector such as M13KE [New England Biolabs] or pSKAN [Mo Bi Tech] which express the peptides via phage display such as on the minor coat protein pDI of M13. Suitable primers can be prepared from the flanking regions to allow the amplification of the intervening peptide encoding nucleic acid sequence. By employing saturation mutagenesis as described by Vartainian a PCR generated library of all possible combinations of peptide minigene is created. These minigenes can be used to replace the sequences encoding the native proline containing sequences in pRCD, pBADT [a,b,c], pSKAN, or pM13KE. Functional screening as described in Example 3. or affinity selection of peptides by phage display Example 2. can be employed to generate additional peptides that both bind the C-terminal SH3 domain and activate this family of repressors.

The PSDT screen described below can also be modified such that any repressor/operator couple from any species employing a DtxR type repressor can be used in functional screening. Using the standard molecular biology techniques as in Example 1, it is also possible to express and utilize any species specific repressor C-terminal domain for phage display and peptide library panning.

We have initially employed PCR approaches using oligo-nucleotide primers to conserved regions of the N-terminal domains of DtxR to isolate additional DtxR repressors from other species. Once cloned we have analyzed the nucleotide and amino acid sequence to determine if the putative homologues contain the conserved proline region found in DtxR. With the advent of high through put publicly sponsored sequencing of microbial genomes it is now possible to scan for DtxR homologues through NCBI. Identification of the potential SH3 domain- internal proline ligand switches is readily completed by performing searches for DtxR homologues and multiple sequence alignments to identify internal proline ligands. Below is a list of swequences which we have obtained by cloning and sequencing of DtxR homologues [CD/SE/SA/SM/EF] or from the NCBI database. The sequences below are from microbes which are associated with bacterial disease with the exception of 'BL' Brevibacterium lactofermtum and 'MS' Mycobacterium smegmatis. An additional list of internal sequences is presented below the Group 1-3 sequences from recently reported DtxR homologues of predominantly non pathogenic bacteria.

List of Internal Sequences broken into three families

Alanine Ala A Cysteine Cys C Aspartic AciD Asp D Glutamic Acid Glu E Phenylalanine Phe F Glycine Gly G

Histidine His H

Isoleucine He I

Lysine Lys K

Leucine Leu L

Methionine Met M

AsparagiNe Asn N

Proline Pro P

Glutamine Gin Q

ARginine Arg R

Serine Ser S

Threonine Thr T

Valine Val V

Tryptophan Trp W

TYrosine Tyr Y

Hydrophobic AA A/V/F/P/M/I/L

Charged AA D/E/K/R

Polar AA S7T/Y/H/C/N/Q/W

Group 1. Where Pro=proline, Gly = glycine, ϋe= isoleucine, P*=a polar amino acid which is thr or ser, H = a hydrophobic amino acid, [+/-] = a charged amino acid and P= a polar amino acid. Consensus sequence for group 1 have a length of 12-14 amino acids of the general sequence:

Pro/P*/P*/P*/Pro/H/Gly/P/Pro/Ile/Pro/Gly/H or [+/-]/[+/-]/H/Gly

Natural examples of Group 1. ligands from DtxR homologues.

Bl DtxR VHRSPFGN PIPGLGEIGL

Cd DtxR VSRSPFGN PIPGLDELGV Mt IdeR PTTSPFGN PIPGLVELGV

Ms IdeR PTTSPFGN PIPGLTELAV

Ml IdeR PTTSPFGN PIPGLLDLGA

SI DesR PTESPYGN PIPGLEELGE Mtb SirR PQRDPHGD PIPGADGQVP

Group 2 Consensus sequence for group 2 have a length of 12-14 amino acids of the general sequence shown below with the fourth polar residue preferably as cysteine the sixth polar residue a histidine:

Pro/[-/+]/P*/P/Pro/P/GIy/Gly/Val/ne/Pro/[+/-]/P or [+/-]/[+/-]

Natural examples of Group 2. ligands froms DtxR homologues.

Se SirR PKTCPHGG VIPRGNSDAA

Sa SirR PETCPHGG VIPRNNEYKE

Ef PEFCPHGG VIPEDNQPIH

Group 3 Consensus sequence for group 3 have a length of 12-18 amino acids of the general sequence shown below with the second amino acid being lysine, the fourth amino acid preferably being cysteine,and he sixth polar amino acid preferably being histidine.

Pro/[-/+]/P* or H/H/Pro/P/Gly/Glyy hr/Ile/Pro/H/P or [+/-]/Gly/[+/-]/H/H Natural examples of Group 3. ligands froms DtxR homologues.

Sg PKACPHGG TIPAKGELLV

Sm PKVCPHGG TIPGHGQPLV Spn PKTCPHGG TIPAKGELLV

Spy PKTCPHGG TIPAKGELLV

*Additional Ligands could be developed by the methods disclosed herein from the following internal proline rich regions from the DtxR homologues found in these species. The following list presents non pathogens and homologues identified by partial sequence analysis in unfinished genomes at NCBI

Methanobacterium thermoautotrophicum D69126 pgecpdekpipacefk

Rhodococcus erythropolis AAF36925 ttspygnpipgldqlg Sulfolobus solfataricus CAB57634 pttcphghpignrikv

Deinococcus radiodurans C75261 pthdphgdpiptlege

Thermoplasma acidophilum C AC 12001 vdrcphgnpipdpegn

Archaeoglobus fulgidus G69497 refcpcgkripevkk

Mycobacterium avium pttspfgnpipglldlgvgpesg Mycobacterium bovis pttapfgnpipglvelgvgpepg

In Example 1, figure 3 displays the polymerization of the C-terminal SH3 domain region of DtxR resolved by PAGE in native and denaturing conditions. Putative SH3 ligand competitive inhibitors [of internal SH3 ligands] could be screened by incubation with the C-terminal domain including residues 120-140 [the internal ligand residues] and then analysis by PAGE gel under native conditions. If the peptides or synthetic compounds being tested disrupt normal association between the SH3 domain and the internal ligand the multimeric complexes observed in the left lane of the gel depicted in figure 3 of Example 1 will not be observed. The C-terminal domain will be resolved essentially as shown in the right lane of the gel, as a single monomeric form.

Similar to the methods disclosed above any competitive binding assay which measures the association of labeled internal SH3 ligand based peptide [residues 125-140 of DtxR or analogous peptide from a DtxR homolog] to the cognate DtxR or DtxR homolog could be used to test for competitive inhibitors of this association. Labeled peptides can readily be obtained by 1125 labeling peptides or by purchasing fluorescently labeled peptides from vendors. Competition between the SH3 domain and the labeled peptide by unlabeled test substances constitutes a method of identifying potential repressor activators.

To demonstrate that the C-terminal domain of DtxR bound proline containing ligands we employed phage display to affinity select peptides by phage display from a random peptide library. The C-terminal domain of DtxR was immobilized onto a substrate through a poly-histidine tail placed N-terminal to a GSG space fused to residue 129. This immobilized SH3 containing domain served as a trap for M13 phage expressing epitopes built from a random peptide library. Multiple rounds of panning and amplification were performed and a set of affinity purified phage particles was obtained. Sequence analysis of the pin protein [where the random peptides are fused] revealed the sequences depicted in Example 2. This technique has been employed extensively in the study of eukaryotic SH3 domains and has provided an affinity based approach for classifying different SH3 domains and in identifying potential lead compounds targeted at eukaryotic proteins modulated by SH3 mediated interactions. The approach is reviewed in Zarrinpar and Lim (Nature Struct Biol. 2000 7:611-613, and Kay, Williamson and Sudol FASEB (2000) 14:231-241) and examples of the technique are presented by Weng et al (MCB (1995) 15:5627-5634) and Sparks et al (PNAS (1996) 93:1540-1544 in which rational dissection of core and specificity residues are pursued and discussed in detail. These approaches can also be used to identify ligands as building blocks for combinatorial chemistry efforts to produce compounds which exhibit specificity from prokaryotic SH3 domains (Dalgarno & Kaye; Parks AB, Adey NB, Quilliam LA, Thorn JM, & Kay BK. Methods in Enzymology, 1995; 255:498-509., Rickles RJ, Botfield MC, Zhou XM, Henry PA, Brugge JS, & Zoller MJ. Proc Natl Acad Sci, USA, 92:10909-12913. ,Kapor, TM, Andreotti, AH, and Schrieber, SL (1998) JACS 120(1)., Feng S, Kasahara C, Rickles RJ, & Schreiber SL. Proc Natl Acad Sci, USA, 1995; 92:12408-12415.) Thus we have employed phage display to demonstrate first that the SH3 domain of DtxR and by analogy this class of repressors indeed bind proline peptide ligands and that in the presence of the internal polyproline [PIP] sequence that additional [high] affinity proline peptides can be obtained.

Grouped generically as P=polar AA, P*=Serine or Threonine, H=hydrophobic, G=glycine, - =negatively charged AA, += positively charged AA, Pr=proline These peptides have a general sequence of: H/P/[+/-]/P/[+/-]/H Pr H/P/H/G H/Pr/Pr H/G/[+/-]P/H/Pr/Pr

P*/[+/-]/H/P*/H/Pr/H [+/-]/P[+/-]/H/Pr/P/Pr P/Pr/P/H/H/P/Pr

Other therapeutic agents useful in the present invention, both peptide and non-peptide alike, may be identified following methods and screening assays reported in the literature in connection with eucaryotic SH3 domains. Dalgarno, et al., Biopolymers 43:383-400 (1997), for example, reviews the nature of several well-characterized intracellular SH3-ligand interactions found in eucaryotic systems, as well as current approaches for design and synthesis of SH3 ligands. One such approach entails mimicking the preassembly of the polyproline helices observed in proteins by replacing the proline-rich core with a rigid organic moiety (referencing Witter, D., et al., (1997) presented at 5^th Chemical Congress of North America, Cancun, Mexico, Poster Presentation 1080). Another main strategy discussed involves exploration of the specificity pocket of the SH3 domain binding site using combinatorial chemistry (referencing Combs, et al. (1996) J. Am. Chem. Soc. 118, 287-288, and Feng, et al. (1996) Chem. Biol. 3, 661-670). Dalgarno further describes a phage display approach using a synthetic D-amino acid Src SH3 domain (referencing Schumacher, et al., (1996) Science 271, 1854-1855). The technique, named "mirror-image phage display", involves inverting the chirality of the SH3 domain by producing a D-enantiomic form of the protein from D-amino acids. L-amino acid phage libraries are screened with the D-SH3 domain, and are equivalent to screening a D-amino acid phage library with the native L-SH3 protein. Kapoor, et al, J. Am. Chem. Soc. 120:23- 29 (1998), describes the design of non-peptide SH3 ligands using structure-based, split-pool synthesis and affinity-based selection.

Other peptides and peptide mimetics targeted to the SH3 domain in this class of procaryotic repressors may represent a useful approach to developing antimicrobial compounds by virtue of their ability to activate DtxR. Peptides can be type I or type ϋ eucaryotic ligands or derivatives thereof. References and strategies are supported by the annotated patents and references, particularly Delgarno, et al. and Nguyen, et al. Synthetic organic ligands may also be produced and screened with the screen previously described by Sun, et al, 1998. In addition, potential compounds can be screened for their ability to inhibit the activation of DtxR [or any homologue] provided that expressed fragment contains the putative SH3 domain and the highly conserved poly-proline sequence endogenous to that repressor. Fluorescence assays can also be developed in which immobilized test compounds can be used to fish out radio- or fluorescent labeled C-termial DxtR [or homologue] Sh3 targets.

The peptides of the invention may be provided in the form of pharmaceutically acceptable salts. Suitable salts include base salts such as alkali metal salts (e.g., sodium or potassium salts), ammonium salts, and acid addition salts such as hydrochloride and acetate salts. D-Peptides (as opposed to peptides containing naturally occurring L-amino acid residues) may also be synthesized as a pure population in an effort to produce more stable and effective therapeutics. The peptides may also be modified to increase binding specificity using the strategy described by Nguyen, et al, [Science 1998], including cyclization. The active form of the peptides is generally phosphorylated, but it may be advantageous to administer a peptide in unphosphorylated form and allow the peptide to become phosphorylated inside the body of the patient. Peptides may be more easily taken up into cells when unphosphorylated. The therapeutic agents of the invention may contain the peptide and at least one non-peptide synthetic moiety.

The peptides of the invention can be synthesized according to standard methods such as those described in Escobedo, J. A., et al, Mol. Cell. Biol. 11:1125-1132 (1991) or Turck, C. W. Peptide Res. 5: 156-160 (1992), for example, using a protected prephosphorylated tyrosine residue. In particular, the peptides can be prepared by liquid or solid-phase methodologies known to those skilled in the art. (Schroeder, et al, "The Peptides", Vol. I, Academic Press 1965, or Bodanszky, et al, "Peptide Synthesis", Interscience Publishers, 1966, or McOmie (ed.) "Protective Group in Organic Chemistry", Plenum Press, 1973, or Barany et al, "The Peptides: Analysis, Synthesis, Biology" 2, Chapter 1, Academic Press, 1980). In the case of solid-phase synthesis any manual or automatic peptide synthesizer can be used and the peptides can be assembled in a stepwise- manner on a resin support using either Boc or Fmoc strategies. The mode of administration of the therapeutic agents of the present invention depends may depend upon the nature and degree of the disease. In general, these routes are topical (e.g., cream or ointment), nasal (e.g., aerosol inhaler), parenteral (e.g., subcutaneous, intramuscular and intravenous) and ionophoretic. The agents may be conjugated to another moiety in order to increase enzymatic stability and cell permeability. The route of administration, as well as the dosage amount and frequency of dosing depend upon numerous factors including, for example, the purpose of the administration, the age and weight of the patient being treated and the condition of the patient. Humans and animals, particularly livestock and domestic animals, may be treated in accordance with the present invention.

The therapeutic agents may be formulated in a pharmaceutical composition suitable for any of the described routes of administration using standard procedures and ingredients. The pharmaceutical composition also comprises a pharmaceutically acceptable carriers or diluents, solubilizers, stabilizers, etc. Aqueous based carriers are preferred for the peptide agents. Any appropriate carrier or diluent may be employed, depending upon the route of administration. See generally, Remington's Pharmaceutical Sciences, Mack Publishers (Easton, PA).

The invention will be further described by reference to the detailed examples. These examples are provided for purposes of illustration only, and are not intended to be limiting unless otherwise specified.

Example 1 The purpose of these experiments was to obtain a more complete understanding of the function of the intact repressor protein, particularly the C-terminal domain. From the sequential assignment of resonances in heteronuclear NMR spectra of a recombinant C-terminal domain (residues N130-L226), we have shown that this isolated domain contains five P-strands and three helices (14). Here, we present the three- dimensional (3D) structure of DtxR( 130-226) determined in solution by using multidimensional NMR spectroscopy and show that it adopts an SH3-like conformation. We also present evidence that this prokaryotic SH3-like domain binds to a proline-rich segment that is located in the region linking the N- and C-terminal domains of DtxR and is conserved in all known DtxR homologues. NMR chemical shift perturbation studies demonstrate that a synthetic peptide corresponding to this internal ligand interacts with specific amino acid residues in the C-terminal domain. The demonstration of peptide binding by the C-terminal domain suggests a mechanism for regulating the activity of the intact repressor protein.

Materials And Methods

Protein Expression, Purification, and Sample Preparation. The expression vector for DtxR( 130-226) was constructed by first introducing a unique BamHX restriction endonuclease site in the dtxR structural gene before N130. The portion of dtxR cDNA encoding residues N130-L226 then was excised by digestion with Ba X and Hfndiπ, and, after purification by agarose gel electrophoresis, ligated into the Ba HX and Hfndiπ sites of the pQE30 expression vector (Qiagen, Chatsworth, CA). The final protein construct, referred to as DtxR( 130-226), contains a 13-residue extension at the N terminus that includes a six- residue His tag (MRGSHHHHHHGSG) to facilitate purification. DtxR(130-226) was expressed in Escherichia coli strain HMS174 grown in M9 minimal medium containing 1 g/liter ¹⁵NH4C1 and 4 g/liter glucose or 2 g/liter ¹³C_ό-glucose to produce uniformly ¹⁵N- or ¹⁵N/¹³C- labeled proteins, respectively. Protein expression was induced by addition of 0.4 mM isopropyl P-D-thiogalactoside to the culture at an ODβoo of ΛJθ.6 and grown for an additional 3 h before harvesting by centrifugation. The cell pellet was resuspended in 20 ml of lysis buffer (50 mM potassium phosphate, pH 7.5, containing 0.5 M NaCl, 8 M urea, 5 mM imidazole, and 1 mM PMSF) and lysed by French press. The clarified lysate was chromatographed over a Ni ⁺-chelating Sepharose Fast Flow column (Amersham Pharmacia), washed with the lysis buffer (containing no urea or PMSF), and eluted with a linear gradient of imidazole (10-600 mM). Fractions containing DtxR(130-226) (at approximately 300 mM imidazole) were pooled, dialyzed, and concentrated in a Centriprep 3 (Amicon) before exchange into phosphate buffer for NMR analysis (50 mM potassium phosphate, containing 0.4% NaN₃ and 10% D₂0, pH 6.5). A shorter construct of the C- terminal domain, corresponding to residues 144-226 [DtxR(144-226)], was generated from DtxR( 130-226) by amplifying the cDNA encoding these residues using PCR, followed by ligation into the NdeX and BamΑX sites of a pET-15b expression vector (Novagen). This construct contains a 21- residue extension at the N terminus of DtxR( 144-226), including a six-residue His tag and a thrombin cleavage site. DtxR(144-226) uniformly enriched in ¹⁵N was expressed in BL21(DE3) E. coli grown in M9 minimal medium supplemented with ¹⁵NHtCl and purified as described for DtxR( 130-226). NMR Spectroscopy. NMR spectra were collected at 30°C on a three-channel 500 MHz Varian Unityp/«_s instrument equipped with waveform generators and three-axis pulsed field gradient accessories. A 3D ¹⁵N-separated nuclear Overhauser effect spectroscopy (NOESY)-heteronuclear single quantum correlation (HSQC) spectrum (15) was collected on a uniformly ¹⁵N-enriched DtxR( 130-226) sample by using 8,333-, 1,650-, and 6,250-Hz sweep widths, and digitized as 512, 48, and 128 complex points in the ah, (^lH ), u>₂ (¹⁵N), and >ι (Η) dimensions, respectively. Two complementary 3D ¹³C-separated CCH-NOESY and HCH-NOESY spectra (16) were collected on uniformly ¹⁵N,^I3C-labeled DtxR(130-226) in the deuterated phosphate buffer by using sweep widths of 2,999.2 and 8,798.8 Hz for Η and ¹³C chemical shift dimensions, respectively. A homonuclear two-dimensional (2D) NOESY spectrum was collected on a 720-MHz Varian Unityplus spectrometer, by using excitation sculpting for solvent suppression (17). All NOESY spectra were collected with a 120-ms mixing time. The Φ-dihedral angle restraints were obtained from analysis of HNHA (18) and HMQC-J (19) spectra collected on ¹⁵N-labeled DtxR( 130-226). Slowly exchanging amide hydrogens were identified from a 2D Η-¹⁵N HSQC spectrum collected 24 h after dissolution in the deuterated phosphate buffer. Heteronuclear NOEs were measured as described (20). All NMR data were processed on Silicon Graphics workstations by using NMRPIPE (21) and analyzed with NMR VIEW (22).

Structure Calculation. Structure calculations were performed by using X-PLOR, version 3.843 (23). The interproton NOE peaks of 2D and 3D NOESY spectra were classified as

1.8-2.8, 1.8-3.5, 1.8-5.0, and 1.8-6.0 A corresponding to strong, medium, weak, and very weak NOEs. Pseudoatom and proton multiplicity corrections were made as described by

Fletcher et al. (24). Hydrogen bond restraints were added as 2.4-3.5 A and 1.5-2.8 A for N-

O and H-O internuclear distances, respectively, in regions where regular secondary structure elements were identified in initial structures calculated by using only NOE restraints. The Φ dihedral angle restraints were applied as —120 ± 30° for P-strands with ³J_HN_H* > 8 Hz and -

60 ± 30° for helical regions with ³J_HNH_Λ < 5.5 Hz, respectively (25). Hydrogen bond and dihedral angle restraints were combined with NOE restraints only in the final stage of the structural refinement. A total of 100 structures were calculated, of which 67 showed no restraint violations greater than 0.5 A and 5°. From the 67 structures, 20 structures with lowest total energy were chosen for further refinement by five additional cycles of simulated annealing by decreasing the initial temperature by 100 K in each cycle from 900 K to 500 K (26). The structures were viewed by using INSIGHTϋ (Molecular Simulations, Sacramento, CA), and analyzed by using MOLMOL (27), AQUA, andPROCHECK-NMR software (28).

Peptide Binding Experiments. A 15-residue peptide (RSPFGNPIPGLDELG; residues R125- G139 of DtxR) was synthesized by using standard solid-phase methods. The peptide showed a single peak on analytical reversed-phase HPLC and gave a mass spectrum identical to that expected. Binding experiments were performed by adding aliquots of peptide to a sample of uniformly ¹⁵N-labeled DtxR( 130-226) or DtxR( 144-226) in the phosphate buffer, pH 6.5 at 30°C. 2D Η-¹⁵N HSQC spectra (29) were collected by using 1,024 and 140 complex points over 8,333.3 and 1,650 Hz spectral widths in the Η and ¹⁵N dimensions, respectively.

Results And Discussion

Structure Determination. Chemical shift assignments for the backbone and side-chain Η, ¹³C, and ¹⁵N resonances of DtxR(130-226) were obtained by using the standard suite of triple-resonance NMR experiments Q4). The 'H* (Fig. Xa), ¹³C^α, and ¹³CO chemical shift deviations (30) suggested the presence of five P-strands and three helices, which subsequently were confirmed in the final 3D structures (Fig. 2). The structure of DtxR(130- 226) was determined from a total of 1,142 NMR restraints in the form of NOE-derived interproton distances, Φ dihedral angles, and hydrogen bonds. Structures were calculated by using a hybrid distance geometry-simulated annealing protocol (23, 3X_). A summary of the structural statistics for the final set of 20 structures is presented in Table 4. These 20 structures had the lowest total energies, no distance violations greater than 0.35 A, and dihedral angle violations less than 5°. Within this family of structures, 97.6% of residues had backbone Ψ,Φ angles located in the allowed regions (28) of the Ramachandran plot. The rms deviation for the backbone atoms in all P-strands superimposed on the average structure

Table 1. Structure statistics

NOE restraints

Total 1,086

Intraresidue 548

Sequential 263

Medium range 87 Long range 188

Φ dihedral angle restraints 26

Hydrogen bond restraints 30

Deviation from expeπmental restraints

Distance restraints, A 0.014 ± 0.003

Dihedral restraints, deg 0.17 ± 0.09 Deviation from idealized covalent geometry

Bonds, A 0.002 ± 0.000

Angles, deg 0.48 ± 0.01

Impropers, deg 0.359 ± 0.005 Backbone rms deviation, A

(SA) to (SΛ>resιdues A147-L226 1.45 ± 0.16

(SA> to (SΛ)all P-strands 0.80 ± 0.08

S A) stands for the ensemble of 20 NMR structures and

the average structure of the ensemble calculated by using X-PLOR. The parameter used to calculate the van der Waals (vdw) repulsion energy was 0.75 rather than 0.80 (47).

Description of NMR Structure. The structure of DtxR( 130-226) consists of a disordered N- terminal region (residues N130-A146) followed by a folded domain (residues A147-L226) (Fig. 3). The five P-strands identified in the final ensemble of structures include residues V163-Q167 (Pi), V193-R198 (P2), H201-H206 (P3), K209-V211 (P4), and R222-E225 (P5). These strands are organized into a P-barrel formed by two partially orthogonal antiparallel P- sheets, with strand 2 shared by the two sheets. Sheet 1 contains strands Pi, P2' (V193-I195), and P5, while sheet 2 is formed by strands P2" (V196-R198), P3, and P4. Preceding Pi in the folded domain, the polypeptide chain forms two short, extended P-like structures (T150- R151 and S158-P160) that are separated by a single-turn 3ιo helix [residues V152-A155 (HI)]. The P-like structures of these two short segments are indicated by down-field H* chemical shifts (Fig. Xa) and by long-range NOE contacts from residues T150-R151 to P5 and from residues S158-P160 to P2". Strands Pi and P2' are connected by a long loop (residues I168-G190) containing the single κ-helix [residues D177-A185 (H2)]. A short 3ι₀ helix [residues D215-A218 (H3)] is formed between strands P4 and P5, while strands P2"4*3 and P3-P4 are connected by tight turns. Many of the hydrophobic residues in helices HI, H2, andH3 (V152, 1153, A155, L182, L183, A185, and A218) showed NOE contacts with the P- barrel, forming the hydrophobic core. To obtain insight into protein chain mobility, we measured steady-state backbone ¹⁵N-{ Η} heteronuclear NOE values. Heteronuclear NOEs for a limited number of residues could not be determined because of spectral overlap. Residues preceding A 147 have negative heteronuclear NOEs (Fig. Xb), indicating high mobility (25). In contrast, residues A147-L226 have positive heteronuclear NOEs, indicating lower overall mobility and that these residues tumble in solution as a single folded domain. The slightly lower heteronuclear NOEs observed for residues I168-E175 suggest an increased mobility for these loop residues compared with other residues in the folded domain. The polypeptide chain mobility deduced from the heteronuclear NOE data correlated well with the number of proton-proton NOEs and the rms deviation per residue (Fig. 1 c and d), indicating that the limited number of interproton NOEs and low structural precision of the linker and the loop regions in the final family of structures reflect the internal motions of the polypeptide chains.

The C-terminal domain of DtxR adopts a similar fold in the crystal (9) and in solution, with a 2.6-A rms deviation obtained when superimposing the C* atoms of the two structures (residues P148-R198 and H201-L226). The largest difference between the two structures was found in residues I168-G190, consistent with their location in a long loop and their increased mobility in solution. Residues G141-A147, which were not traced in previous x-ray structures, were also highly mobile in solution and were poorly defined by the NMR data.

77ιe C-Terminal Domain of DtxR Binds a Proline-Rich Peptide. During purification and characterization of DtxR( 130-226), it was observed that highly purified protein ran as a series of bands in nondenaturing polyacrylamide gels that correlated in mass to multiples of the monomeric protein molecular weight (Fig. 3). As seen in Fig. 3, the monomeric and trimeric forms were predominant, with lower amounts of dimer and higher aggregates observed. The formation of oligomers was not altered upon incubation with EDTA or by addition of 10 mM Ni²⁺, suggesting that the oligomerization was not induced by residues of the His tag binding to metal ions leached during purification. In contrast, a single molecular weight band corresponding to monomeric DtxR( 130-226) was observed in denaturing PAGE gels (not shown). As noted previously (9), the structure of residues P160-L226 is homologous to eukaryotic SH3 domains. SH3 domains bind peptides with the consensus sequence PpXP, where P is a strictly conserved proline, p is generally a proline, and X is a hydrophobic residue (32-37). Following the His tag and additional residues associated with the cloning sites (see Materials and Methods), the DtxR( 130-226) sequence begins as NPIPGL. We reasoned that the oligomers may result from DtxR( 130-226) binding this proline-containing segment. To test this hypothesis, DtxR( 144-226), in which this internal ligand is removed, was created. NMR spectra of DtxR( 144-226) showed that the protein adopted the same fold as DtxR(130-226). However, in contrast to DtxR( 130-226), DtxR(144-226) migrated as a single band corresponding to monomer molecular weight in nondenaturingpolyacrylamide gels (Fig. 3).

The possible binding interaction between the SH3 C-terminal domain of

DtxR and the internal proline-rich sequence was further investigated by using a synthetic peptide having the sequence RSPFGNPIPGLDELG, which corresponds to residues R125- G139 of full-length DtxR. Aliquots of this peptide were added to DtxR( 130-226), and 2D HSQC spectra were collected. Because chemical shifts are extremely sensitive reporters of the local magnetic environment, ligand binding generally changes the chemical shifts of backbone and side-chain resonances. This approach is sensitive to weak binding (into the millimolar range; ref. 38) and has been used previously to demonstrate binding between proline-rich peptides and eukaryotic SH3 domains (36). When a stoichiometric amount of this peptide was added to DtxR( 130-226), a limited number of protein ^!H and/or ^l5N resonances exhibited line broadening or resonance frequency changes in 2D HSQC spectra (residues V174, 1187, E192, L204, H206, D215, D216, L217, H219, and T220), but no additional resonances appeared. At approximately a 5:1 peptide/protein molar ratio, additional residues in strands P2, P3, P5, and helix H3 were shifted in an HSQC spectrum (Fig. 5). The chemical shift perturbation data demonstrate weak binding of the peptide by the SH3-like domain, in fast exchange on the NMR time scale. The perturbed residues generate a putative peptide-binding surface located between the long loop and the P-barrel (Fig. 6). The presence of an internal partial ligand that competes with the external peptide complicates a quantitative analysis of the binding affinity for the peptide. By using the existing NMR data we estimate an apparent dissociation constant in the 100 μM-1 mM range, which is slightly higher than that obtained for eukaryotic SH3 domains binding optimized peptide ligands (32-37). DtxR(144-226) also binds the R125-G139 peptide, with the same residues being shifted upon binding.

Except for NOEs observed between side-chain protons of A 146 and 1187 that were consistent with oligomer formation, no NOEs were observed between residues at the N terminus and the folded domain of DtxR( 130-226), although some residues at the N terminus of DtxR(130-226) shifted after addition of the R125-G139 peptide. These intermolecular NOEs disappeared upon dilution of the DtxR( 130-226) sample. The absence of NOEs from the tail to the folded domain of DtxR( 130-226) may be attributed to the high flexibility of the N terminus in the monomeric species (Fig. Xb) and to the variety and low concentration of oligomeric species in solution.

A Proposed Functional Role for Peptide Binding. A working model for transcriptional regulation by DtxR is that micromolar concentrations of Fe²⁺ or other divalent metals trigger the formation of the metal-bound dimeric state, which then binds to the tox and irp operators (39-42). In the absence of divalent metal ligand, DtxR is thought to exist as an inactive, monomeric apo-protein that is incapable of binding DNA. Residues R125-G139 make numerous contacts with the three helices that constitute the dimerization interface in the N-terminal domain, thereby contributing to the stabilization of the dimeric form of DtxR (9). In the current work, we found that residues R125-G139 also can interact with the C- terminal domain of DtxR. If residues R125-G139 were to dissociate from the N-terminal domain, the dimeric structure might be destabilized and dissociate into monomers. Although not int4ending to be bound by theory, we propose that the C-terminal domain binds residues R125-G139 in the monomeric state, thereby altering the monomer-dimer equilibrium and effectively stabilizing the monomeric, inactive form. Our data is consistent with either an inter- or intramolecular binding. This model for the regulation of dimer formation by the SH3-like C-terminal domain is consistent with the weakly cooperative activation of DtxR by metal ions (4) and with the existing C-terminal domain mutants that alter repressor activity 02, 13).

Eukaryotic SH3 domains in Hck (43), Src (44), and Itk (45) regulate tyrosine kinase activities in signal transduction cascades by weak binding to an internal proline- containing peptide whose sequence differs from the high-affinity peptide sequences that activate the kinase. Here, we have postulated that binding to an internal proline-containing region by the SH3-like domain of this prokaryotic protein has significance in regulating the repressor activity of intact DtxR. According to our model, the C-terminal domain plays no direct role in the structure or function of the dimeric form of the repressor and must be flexibly linked to the N-terminal domain. This intrinsic flexibility may explain the low averaged electron density found for this domain in the existing crystal structures (6-11). Residues L120-L226 were not traced in a crystal structure of DtxR(C102D) complexed with a 33-bp DNA sequence (11), so the structure of the proline-containing region and the C- terminal domain in this state of the repressor is uncertain. The N-terminal domain of DtxR shows strong homology with the other members of Gram-positive toxin gene repressor proteins. A recent crystal structure of the DtxR homologue from M. tuberculosis, IdeR, shows the proteins are structurally homologous as well (46). Similarly, the sequence homology of the C-terminal domains in the family of DtxR homologues suggests that they will adopt SH3-like folds. Residues S126-G139 are highly conserved in all known DtxR homologues, therefore we also believe that the regulatory mechanism proposed here for DtxR is applicable to the entire family of virulence-gene repressor proteins in the Gram-positive bacteria.

A recent study by Goranson-Siekierke et al. (48) has demonstrated that single alanine substitutions for residues R80, S126, and N130 caused severely decreased DtxR activity. Crystallographic analyses of dimeric metal complexes of the native protein show that these residues coordinate an oxyanion, which has been identified as a possible co- corepressor (9, 49). In dilute solutions, the dimeric form of the protein is stabilized by low concentrations of Fe ⁺ or other divalent transition metal cations, but dimerization also is favored in the absence of the metal ions at high protein concentration under crystallizing conditions. High-resolution analyses of crystals of the metal-free DtxR (10) show a dimeric structure very similar to the metal-bound form, in which the segment including the conserved sequence S126-G139 is folded in an ordered conformation contacting the helices of the N-terminal domain involved in dimer formation; these polar interactions among the residues R80, S126, and N130 together with water and/or anion evidently contribute to the stability of the dimer interface. Our results demonstrate that the proline-rich segment, including residues S126 and N130, binds to the isolated C-terminal SH3-like domain of DtxR in a manner similar to the peptide binding by eukaryotic SH3 domains (43—45). According to our model for the regulation of the DtxR activity, binding of the proline-rich segment to the C-terminal SH3-like domain should stabilize the inactive monomeric form of the repressor. Because replacement of the polar residues R80, S126, and N130 with alanines will weaken the interaction between the S126-L138 segment and the N-terminal dimerization domain, we interpret the recent results reported by Goranson-Siekierke et al (48) to indicate that destabilization of the proline-rich segment in the N-terminal domain of the dimer consequently should favor binding of this segment to the C-terminal SH3-like domain in the inactive monomer, even in the presence of activating metal ions. Thus, the sequence S126-G139 may function as an internal molecular switch, either associated with the N-terminal domain, thereby contributing to the stability of the active, metal-bound dimeric form of the repressor, or alternatively bound to the C-terminal domain, favoring the inactive monomeric form. Abbreviations SH3, Src homology 3; DtxR, diphtheria toxin repressor; 3D, three- dimensional; NOE(SY), nuclear Overhauser effect (spectroscopy); HSQC, heteronuclear single quantum correlation; 2D, two-dimensional.

References

1. Tao, X., Schiering, N., Zeng, H.-Y., Ringe, D. & Murphy, J. R. (1994) Mol. Microbiol. 14, 191-197.

2. Pappenheimer, A. M. (1977) Annu. Rev. Biochem. 46, 69-94.

3. Schmitt, M. P. (1997) Infect. Immun. 65, 4634-4641.

4. Oguiza, J. A., Tao, X., Marcos, A. T., Martin, J. F. & Murphy, J. R. (1995) J. Bacteriol. 177, 465-467. 5. Schmitt, M. P., Predich, M., Doukhan, L., Smith, I. & Holmes, R. K. (1995) Infect. Immun. 63, 4284-4289.

6. Qiu, X., Verlinde, C. L. M. J., Zhang, S., Schmitt, M. P., Holmes, R. K. & Hoi, W. G. J. (1995) Structure (London) 3, 87-100 .

7. Schiering, N., Tao, X., Zeng, H.-Y., Murphy, J. R., Petsko, G. A. & Ringe, D. (1995) Proc. Natl. Acad. Sci. USA 92, 9843-9850. 8. Ding, X., Zeng, H., Schiering, N., Ringe, D. & Murphy, J. R. (1996) Nat. Struct. Biol. 3, 382-387 .

9. Qiu, X., Pohl, E., Holmes, R. K. & Hoi, W. G. J. (1996) Biochemistry 35, 12292- 12302 . 10. Pohl, E., Holmes, R. K. & Hoi, W. G. J. (1998) J. Biol. Chem. 273, 22420-22427 .

11. White, A., Ding, X., vanderSpek, J. C, Murphy, J. R. & Ringe, D. (1998) Nature (London) 394, 502-506 .

12. Sun, L., vanderSpek, J. C. & Murphy, J. R. (1998) Proc. Natl. Acad. Sci. USA 95, 14985-14990 . 13. Schmitt, M. P. & Holmes, R. K. (1993) Mol. Microbiol. 9, 173-181.

14. Twigg, P. D., Wylie, G. P., Wang, G., Murphy, J. R., Caspar, D. L. D. & Logan, T. M. (1999) J. Biomol. NMR 13, 197-198 .

15. Zhang, O., Kay, L. E., Olivier, J. P. & Forman-Kay, J. D. (1994) J. Biomol. NMR 4, 845-858 . 16. Vuister, G., Clore, G. M., Gronenborn, A. M., Powers, R., Garret, D. S., Tschudin, R. & Bax, A. (1993) J. Magn. Reson. 101, 210-213 .

17. Callihan, D., West, J., Kumar, S., Schweitzer, B. I. & Logan, T. M. (1996) J. Magn. Reson. 112, 82-85 .

18. Kuboniwa, H., Grzesiek, S., Delagio, F. & Bax, A. (1994) J. Biomol. NMR 4, 871- 878 .

19. Kay, L. E. & Bax, A. (1990) J. Magn. Reson. 86, 110-126 .

20. Farrow, N. A., Muhandiram, R., Singer, A. U., Pascal, S. M., Kay, C. M., Gish, G., Shoelson, S. E., Pawson, T., Forman-Kay, J. D. & Kay, L. E. (1994) Biochemistry 33, 5984-6003 . 21. Delaglio, F., Grzesiek, S., Vuister, G. W., Zhu, G., Pfeifer, J. & Bax, A. (1995) J. Biomol. NMR 6, 277-293.

22. Johnson, B. A. & Blevins, R. A. (1994) J. Biomol. NMR 4, 603-614.

23. Briinger, A. T. (1992) X-PLOR: A System for X-Ray Crystallography and NMR (Yale Univ. Press, New Haven), Version 3.1. 24. Fletcher, C. M., Jones, D. N. M., Diamond, R. & Neuhaus, D. (1996) J. Biomol. NMR 8, 292-310. 25. Wuthrich, K. (1986) NMR of Proteins and Nucleic Acids (Wiley, New York).

26. Briercheck, D. M., Wood, T. C, Allison, T. J., Richardson, J. P. & Rule, G. S. (1998) Nat. Struct. Biol. 5, 393-399.

27. Koradi, R., Billeter, M. & Wuthrich, K. (1996) J. Mol. Graphics 14, 51-55. 28. Laskowski, R. A., Rullmann, J. A. C, MacArthur, M. W., Kaptein, R. & Thornton, J. M. (1996) J. Biomol. NMR. 8, 477-486.

29. Kay, L. E., Keifer, P. & Saarinen, T. (1992) J. Am. Chem. Soc. 114, 10663-10665 .

30. Wishart, D. S. & Sykes, B. D. (1994) J. Biomol. NMR 4, 171-180 .

31. Nilges, M., Clore, G. M. & Gronenborn, A. M. (1988) FEBS Lett. 229, 317-324 . 32. Feng, S., Chen, J. K., Yu, H., Simon, J. A. & Schreiber, S. L. (1994) Science 266, 1241-1247 .

33. Lim, W. A., Richards, F. M. & Fox, R. O. (1994) Nature (London) 372, 375-379.

34. Saraste, M. & Musacchio, A. (1994) Nat. Struct. Biol. 1, 835-837 .

35. Viguera, A. R., Arrondo, J. L. R., Musacchio, A., Saraste, M. & Serrano, L. (1994) Biochemstry 33, 10925-10933 .

36. Wittekind, M., Mapelli, C, Farmer, B. T. B, Suen, K.-L., Goldfarb, V., Tsao, J., Lavoie, T., Barbacid, M., Meyers, C. A. & Mueller, L. (1994) Biochemistry 33, 13531- 13539 .

37. Yu, H., Chen, J. K., Feng, S., Dalgarno, D. C, Brauer, A. W. & Schreiber, S. L. (1994) Cell 76, 933-945.

38. Shuker, S. B., Hajduk, P. J., Meadows, R. P. & Fesik, S. W. (1996) Science 274, 1531-1534 .

39. Tao, X., Boyd, J. & Murphy, J. R. (1992) Proc. Natl. Acad. Sci. USA 89, 5897- 5901. 40. Tao, X., Zeng, H. Y. & Murphy, J. R. (1995) Proc. Natl. Acad. Sci. USA 92, 6803- 6807 .

41. Wang, Z., Schmitt, M. P. & Holmes, R. K. (1994) Infect. Immun. 62, 1600-1608 .

42. Tao, X. & Murphy, J. R. (1994) Proc. Natl. Acad. Sci. USA 91, 9646-9650 .

43. Sicheri, F., Maorefi, I. & Kuriyan, J. (1997) Nature (London) 385, 602-609 . 44. Xu, W., Harrison, S. C. & Eck, M. L. (1997) Nature (London) 385, 595-602. 45. Andreotti, A. H., Bunnell, S. C, Feng, S., Berg, L. J. & Screiber, S. L. (1997) Nature (London) 385, 93-97 .

46. Pohl, E., Holmes, R. K. & Hoi, W. G. J. (1999) J. Mol. Biol. 285, 1145-1156 .

47. Johnson, P. E., Joshi, M. D., Tomme, P., Kilburn, D. G. & Mclntosh, L. P. (1996) Biochemistry 35, 14381-14394 .

48. Goranson-Siekierke, J., Pohl, E., Hoi, W. G. J. & Holmes, R. K. (1999) Infect. Immun. 67, 1806-1811 .

49. Pohl, E., Qiu, W., Must, L. M., Holmes, R. K. & Hoi, W. G. J. (1997) Protein Sci. 6, 1114- 1118 Example 2

Phage Display

To demonstrate that the C-terminal domain of DtxR bound proline containing ligands we employed phage display to affinity select peptides by phage display from a random peptide library. The C-terminal domain of DtxR was immobilized onto a substrate through a polyhisitidine tail placed N-terminal to a GSG space fused to residue 129. This immobilized SH3 containing domain served as a trap for M13 phage expressing epitopes built from a random peptide library. Multiple rounds of panning and amplification were performed and a set of affinity purified pahge particles was obtained. Sequence analysis of the pHI protein [where the random peptides are fused] revealed the sequences depicted in Example 2 below. This technique has been employed extensively in the study of eukaryotic SH3 domains and has provided an affinity based approach for classifying different SH3 domains and in identifying potential lead compounds targeted at eukaryotic proteins modulated by SH3 mediated interactions. The approach is reviewed in Zarrinpar and Lim (Nature Struct Biol. 2000 7:611-613, and Kay, Williamson and Sudol FASEB (2000) 14:231-241) and examples of the technique are presented by Weng et al (MCB (1995) 15:5627-5634) and Sparks et al (PNAS (1996) 93:1540-1544 in which rational dissection of core and specificity residues are pursued and discussed in detail. These approaches can also be used to identify ligands as building blocks for combinatorial chemistry efforts to produce compounds which exhibit specificity from prokaryotic SH3 domains (Dalgarno and Kaye; Parks AB, Adey NB, Quilliam LA, Thorn JM, & Kay BK. Methods in Enzymology, 1995; 255:498-509., Rickles RJ, Botfield MC, Zhou XM, Henry PA, Brugge JS, & Zoller MJ. Proc Natl Acad Sci, USA, 92:10909-12913. ,Kapor, TM, Andreotti, AH, and Schrieber, SL (1998) JACS 120(1)., Feng S, Kasahara C, Rickles RJ, & Schreiber SL. Proc Natl Acad Sci, USA, 1995; 92: 12408-12415.) Thus we have employed phage display to demonstrate first that the SH3 domain of DtxR and by analogy this class of repressors indeed bind proline peptide ligands and that in the presence of the internal polyproline [PIP] sequence that additional [high] affinity proline peptides can be obtained.

Grouped generically as P=polar AA, P*=Serine or Threonine, H=hydrophobic, G=glycine, - =negatively charged AA, += positively charged AA, Pr=proline These peptides have a general sequence of:

H/P/[+/-]/P/[+/-]/H/Pr H/P/H/G/H/Pr/Pr H7G/[+/-]P/H/Pr/Pr

P*/[+/-]/H/P*/H/Pr/H [+/-]/P[+/-]/H/Pr/P/Pr P/Pr/P/H/H/P/Pr

Example 2

Identification of Repressor SH3 ligands from random peptide libraries

The X-ray and NMR analysis demonstrate that the C-terminal domains of DtxR and DtxR like repressor IdeR can fold into an SH3-like structure. The internal polyproline motif of DtxR employed in Example 1 suggests a functional consequence of internal peptide binding. These studies only utilized the C-terminal 136 amino acids of a multi-domain 226 amino acid protein, however. In support of these findings are the polyproline rich internal ligands which we have identified by cloning and sequence analysis of homologous DtxR repressors from species of clinical interest. The high degree of amino acid homology in the proline rich linker region within the DtxR family of repressors (SI 26- L138; Fig. 2) suggests that these SH3-like domains may share a common mechanism of action. Furthermore, to develop lead drug candidates, the SH3 like domain should be able to select and bind exogenous polyproline peptidic ligands. Candidate polyproline ligands of higher affinity for the unique DtxR SH3 like domain should bind the SH3 domain in the presence of the endogenous peptide. As a first step, we employed random peptide phage display libraries to determine if affinity selected peptides could be identified that would interact with the DtxR SH3 like domain. The peptides selected are disclosed below. Screening of phage displayed combinatorial libraries

Affinity selection of targets for receptors, transcription factor and protein interaction surfaces in which large numbers of random molecules are screened for their ability to interact, label or activate the protein of interest is a widely employed technique. Phage display of small random peptides having lengths of between 7-25 amino acids has provided the ability to rapidly screen a random yet representative universe of all possible combinations of amino acids. Random peptide libraries have been widely used for epitope mapping (Scott & Smith, 1990), the identification of peptide mimics of non-peptide ligands (Scott et al, 1992), and mapping protein-protein contacts (Hong & Boulanger, 1995). In general, Ml 3 phage display is a selection technique in which a peptide, or peptide library, is genetically fused to a bacteriophage coat protein. Following phage assembly, the peptide library is then presented on the surface of the virion. Most importantly, this method allows the physical linkage between each individual peptide sequence with the DNA encoding that sequence. Phage display 7-mer and 12-mer peptide libraries are commercially available and will be initially employed in these studies (New England Biolabs{ Beverly, MA}; cat #8100, #8110 and Mo Bi Tech LLC, Marco Island, FL). After multiple rounds of affinity selection and amplification, phage were plated, and individual clones were isolated and characterized by DNA sequence analysis. To demonstrate that the C-terminal domain of DtxR bound proline containing ligands we employed phage display to affinity select peptides by phage display from a random peptide library. This technique has been employed extensively in the study of eukaryotic SH3 domains and has provided an affinity based approach for classifying different SH3 domains and in identifying potential lead compounds targeted at eukaryotic proteins modulated by SH3 mediated interactions.

Protein Expression, Purification, and Assay Plate Preparation. The expression vector for DtxR( 130-226) was constructed by first introducing a unique BamHX restriction endonuclease site in the dtxXλ structural gene before N130. The portion of dtxXλ cDNA encoding residues N130-L226 then was excised by digestion with BamHX and HmdlH, and, after purification by agarose gel electrophoresis, ligated into the BamHl and H dlH sites of the pQE30 expression vector (Qiagen, Chatsworth, CA). The final protein construct, referred to as DtxR(130-226), contains a 13-residue extension at the N terminus that includes a six- residue His tag (MRGSHHHHHHGSG) to facilitate purification. DtxR(130-226) was expressed in Escherichia coli strain HMS174 grown in M9 minimal medium. Protein expression was induced by addition of 0.4 mM isopropyl P-D-thiogalactoside to the culture at an OD₆oo of ftiθ.6 and grown for an additional 3 h before harvesting by centrifugation. The cell pellet was resuspended in 20 ml of lysis buffer (50 mM potassium phosphate, pH 7.5, containing 0.5 M NaCl, 8 M urea, 5 mM imidazole, and 1 mM PMSF) and lysed by French press. The clarified lysate was chromatographed over a Ni²⁺-chelating Sepharose Fast Flow column (Amersham Pharmacia), washed with the lysis buffer (containing no urea or PMSF), and eluted with a linear gradient of imidazole (10-600 mM). Fractions containing DtxR(130-226) (at approximately 300 mM imidazole) were pooled, dialyzed, and concentrated in a Centriprep 3 (Amicon). This protein was used to determine conditions for maximal binding to Ni+ affinity micro titer plates (Qiagen, Chatsworth, CA) and wells were prepared under saturating conditions. These wells washed to remove free C-terminal domain and then used in affinity selection of random phage by biopanning. The 7-mer peptide library has been reported to carry 2 x 10⁹ independent clones, a number which is sufficiently large to represent a significant fraction of the 20 possible sequences. In contrast, the 12-mer library has been reported to also contain approximately 2 x 10⁹ independent clones, which in this instance is only a small fraction of the 20¹² possible sequences. Phage were incubated in SH3 coated microtiter wells for between 2 and 12 hrs after which unbound phage were removed and washed away. Specifically bound pahge were removed in step imidazole washes fractions were amplified by preparing new stocks of enriched M13 phage.

The selected and amplified phage were reprocessed through additional rounds of selection until a population enriched phage is derived [4-5 rounds]. After selction random phage were picked from a PFU assay and used to prepare template DNA for sequenceing. Sequencing reactions were carried out by a vendor and alignment of random peptide in pin and the amino acid sequence of the random peptides determined.

Results Since Example 1 shows that DtxR( 130-226) binds the proline rich peptide

R125-L135, it was not surprising to identify proline containing peptides by affinity selection using phage display. However, with the use of phage display one must be concerned with the complexity of the library, peptide degradation, specificity of binding, and (perhaps most importantly) assignment of function. Using a commercially available library, we have utilized phage display to isolate and characterize a number of phage isolates after 4 rounds of affinity selection on the DtxR C-terminal domain peptide, DtxR( 130-226). It is noteworthy that this target contains the C-terminal half of the internal proline sequence of DtxR including PIP. The construct is fused to a poly histidine tag by small linker region [GSG]. This means that there is the potential for the SH3 to SH3 domain interactions depicted in Example 1 above in this assay suggesting that the phage isolated to date have a higher affinity for the DtxR SH3 domain than the internal PIP containing ligand. To further expand this set of ligands affinity selection with coupled in vitro mutagenesis and affinity selection to define a consensus sequence amongst potential proline containing peptides could be performed. Affinity purification can also be sensitive to selection conditions, therefore by adjusting the pH or salt concentration during affinity purification it is possible to more readily define or differentiate ligand specificity and expand the set peptide ligands.

Poly Proline Peptides Mono Proline Peptides Proline Free Peptides SMPΓΓPP GDNAPP VPASVKS SDGEVWE AHLGFPP DHRLPSP FTNRLLP WRAMRAG

YPHAMQP

Example 3 Identification Of Random Sh3 Ligands Which Activate Dtxr A critical step is to provide evidence that one could isolate peptides that activate DtxR and contain a polyproline motif. In a recently published article (Sun et al., 1998), the assay selection system ("PSDT") was used to isolate and characterize the first hyper-repressor mutants of DtxR. In this study, PCR mutagenesis was employed to generate a library of variant DtxR genes which were then screened in the PSDT system. In this system, only those variants which maintained functional DtxR::tox operator interaction in the presence of the iron chelator 2,2"-dipyridyl were selected on medium supplemented with chloramphenicol. We have employed a similar approach to screen random peptide libraries. The results of these studies yielded a polyproline containing peptide that is capable of activating DtxR in the presence of 2,2"-dipyridyl.

To identify polyproline peptides capable of activating DtxR and DtxR homologues we employed the PSDT system. The approach utilized a minigene library created from bacterial gDNA inserted into an expression vector carrying a copy of the DtxR repressor gene. Expression of DtxR in this system results in active repression of a the Tet repressor which is carried on a second plasmid under the control of the tox operator. A chromosomal insertion in the host strain of E. coli carries the chloramphenical acetyl- transferase [CAT] gene under the conrtol of the tet repressor. When iron becomes limiting the DtxR in the cells becomes inactivated thereby de-repressing TetR and in so doing repressing CAT. The host cells switch from a chloramphenicol resistant to a chloramphenicol senstive phenotype. To screen for peptide activators of DtxR inserted random gene fragments under the control of a constitutive promoter and selected colonies on chloramphenicol in the presence of the iron chelator 2' -2 dipyridyl. Only cells co- expressing the repressor and a peptide fragment capable of activating the repressor are selected. As illustrated in the published PCT Application No. US99/22770, the PSDT system consists of a lysogenic E. coli TOP 10 host strain which carries the reporter gene cat (chloramphenicol acetyltransferase, Cat) on an integrated lambda phage, λRS65T, and a set of detector plasmids. In this system, expression of cat from λRS65T is controlled by the tetA promoter / operator (tetAPO). In the absence of the tetracycline repressor (TetR), the expression of Cat is constitutive in E. coli TOP10/λRS65T and as a result this strain is resistant to chloramphenicol (Cm^R). The detector plasmid, pSC6, carries the tetR gene under the control of the diphtheria toxPO. When pSC6 is transformed into E. coli TOP10/λRS65T, this host strain becomes Cm sensitive (Cm^s) by virtue of the constitutive expression of tetR. In this instance, TetR recognizes and binds to the tetAPO and represses cat gene expression. However, if either a functional dtxR allele or homolog is introduced into the bacterial host on a second compatible plasmid (e.g., pRCD or PBADT-A,B,C), the interaction between DtxR and the toxO will repress the expression of tetR and the bacterial host, E. coli TOP10/λRS65T/pSC6/pRDA, will then regain its Cm^R phenotype.

Several classes of peptides were identified; however, most striking was a peptide with the following sequence: MΓΓPSAQLTLTKGNKSWVPGPPSRSTVSISLISNSSSVPL.

When expressed in the beta-galactosidase reporter strain in the presence of 2' -2 dipyridyl followed by ONPG based beta galactosidase assay the clones carrying this construct failed to display beta-galactosidase activity whereas clones carrying a copy of DtxR alone yielded activity indicating that the DtxR repressor had been inactivated.

Repressor Peptide Gene BetaGal Fe + BetaGal Fe -IDPX

DxtR None - ++

DtxR SH3 Activator Peptide

The central core of this 40 amino acid peptide contains a polyproline stretch which is analogous to class I SH3 ligands employed by eukaryotic systems. Moreover, this proline rich region is related to the DtxR proline rich region that is positioned between the N- and C-terminal [AA 125-139] of the repressor and compares favorably to the peptides identified by phage display. Dissection of Peptide Activators minimal sequences: Synthetic minigenes splitting the peptide into two overlapping sequences can encode (1) the N-terminus to residue 25, (2) from the C-terminus in 25 residues to the central polyproline rich core. Each of these peptides can be used in the PSDT screen to identify additional SH3 ligands. The Minigene can also be subject to PCR mediated saturation mutagenesis in addition to 5' and 3' deletions to derive additional peptide activators. Derivatives of the peptide core can also be constructed by oligo-nucleotide assembly of a minigene and tested in the PSDT system. 7. MITPSAQLTLTKGNKSWVPGPPSRS

2. NKSWVPGPPSRS TVSISLISNSSSVPL

3. XGPP

4. PGXP

5. PGPX

6. XGPX

7. PGXX

S. XGXP

9. PGPPSX

The following list provides complete citations for all publications referenced elsewhere in the specification, other than in Example 1.

Anderotti AH, Bunnell SC, Feng S, Berg LJ, & Schreiber SL. Regulatory intramolecular association in a tyrosine kinase of the Tec family. Nature, 1997; 3S5:93-97.

Bedford MT, Chan DC, Leder P.FBP WW domains and the Abl SH3 domain bind to a specific class of proline-rich ligands. EMBO J. 1997 May l;76(9):2376-83

Boyd J, Murphy JR. Analysis of the diphtheria tox promoter by site-directed mutagenesis. J Bacteriol, 1988; 770:5949-5952.

Boyd J, Oza M, & Murphy JR. Molecular cloning and DNA sequence analysis of an iron dependent diphtheria tox regluatory element (dtxR) from Corynebacterium diphtheriae Proc Natl Acad Sci, USA, 1990; §7:5968-5972.

Bunnell SC, Henry PA, Kolluri R, Kirchhausen T, Rickles RJ, Berg LJ. (1996) Identification of Itk/Tsk Src homology 3 domain ligands. J Biol Chem.277(47):25646-56.

Combs AP, Kapoor TM, Feng S, Chen JK, Lygia F. DS, Schreiber SL. Protein Structure- Based Combinatorial Chemistry Discovery of Non-Peptide Binding Elements to Src SH3 Domain, J. Am. Chem. Soc, 1996:77S: 287-288. Cortese R, Monaci P, Luzzago A, Santini C, Bartoli F, Cortese I, Fortugno P, Galfre G, Nicosia A, & Felici F. Selection of biologically active peptides by phage display of random peptide libraries. Curr Opin Biotechnol, 1996; 7:616-621.

Dalgarno DC, Botfield MC, Rickles RJ.SH3 domains and drug design: ligands, structure, and biological function. Biopolymers. 1997;43(5):383-400.

Doukhan L, Predich M, Nair G, Dussurget O, Manic-Mulec I, Cole ST, Smith DR, & Smith I. Genomic organization of the mycobacterial sigma gene cluster. Gene, 1995; 765:67-70.

Dussurget O, Rodriguez M, Smith I.An ideR mutant of Mycobacterium smegmatis has derepressed siderophore production and an altered oxidative-stress response. Mol Microbiol. 1996 Nov;22(3):535-44.

Feese MD, Ingason BP, Goranson-Siekierke J, Holmes RK, Hoi WG Crystal Structure of the Iron-dependent Regulator (IdeR) from Mycobacterium tuberculosis at 2.0 Angstrom Resolution Reveals the SH3-like Fold and Metal Binding Function of the Third Domain. J Biol Chem. 2000 Oct 26

Feng S, Kasahara C, Rickles RJ, & Schreiber SL. Specific interactions outside the proline- rich core of two classes of Src homology 3 ligands. Proc Natl Acad Sci, USA, 1995; 92:12408-12415.

Feng S, Schreiber SL. Enantiomeric Binding Elements Interacting at the Same Site of an SH3 Protein Receptor, J. Am. Chem. Soc, 1997: 779:10873-10874.

Gottesfield JM, Neely L, Trauger JW, Baird EE, & Dervan PB. Regulation of gene expression by small molecules. Nature, 1997; 3S7/202-205. Gunter K, Toupet C, & Schupp T. Characterization of an iron-regulated promoter involved in desferrioxamine B synthesis in Streptomyces pilosus: repressor-binding site and homology to the diphtheria toxin gene promoter. J Bacteriol 1993; 775:3295-3302.

Gϋnter-Seeboth K., & Schupp T. Cloning and sequence analysis of the Corynebacterium diphtheriae dtxR homologue from Streptomyces lividans and S. pilosus encoding a putative iron repressor. Gene, 1995; 766. 117-119.

Hill PJ, Cockayne A, Landers P, Morrissey JA, Sims CM, & Williams P. 1998. SirR, a novel iron-dependent repressor in Staphylococcus epidermidis. Infect Immun. 66:4123- 4129.

Hong SS, & Boulanger P. Protein ligands of the human adenovirus type 2 outer capsid identified by biopanning of a phage-displayed peptide library on separate domains of wild-type and mutant penton capsomers. EMBO. 1995; 23-2974:4714-4727

Jakubovics NS, Smith AW, Jenkinson HF.Expression of the virulence-related sea (Mn2+) permease in streptococcus gordonii is regulated by a diphtheria toxin metallorepressor-like protein sea R. Mol Microbiol. 2000 Oct;3S(l): 140-53

Kapor, TM, Andreotti, AH, and Schrieber, SL (1998) Exploring the specificity pockets of two homologous SH3 domain using structure-based, split-pool synthesis and affinity-based selection. JACS 120(1).

Kardinal C, Konkol B, Schulz A, Posern G, Lin H, Adermann K, Eulitz M, Estrov Z, Talpaz M, Arlinghaus RB, Feller SM.Cell-penetrating SH3 domain blocker peptides inhibit proliferation of primary blast cells from CML patients. FASEB J. 2000 Aug;74(77):1529- 38. Kay BK, Williamson MP, Sudol M.The importance of being proline: the interaction of proline-rich motifs in signaling proteins with their cognate domains. FASEB J. 2000 Feb;74(2):231-41

Kitten T, Munro CL, Michalek SM, Macrina FL.Genetic characterization of a Streptococcus mutans Lral family operon and role in virulence. Infect Immun. 2000 Aug;6S(S):4441-51.

Litwin CM, & Calderwood SB. Role of iron in regulation of virulence genes. Clin Microbiol Rev, 1993; 6:137-149.

Ma J, & Ptashne M. A new class of yeast transcriptional activators. Cell, 1988; 57:113-119.

Manabe Y, Saviola BJ, Sun L, Murphy JR, & Bishai WR. Attenuation of virulence in Mycobacterium tuberculosis expressing a constitutively active iron repressor. 1999; (PNAS 96:12844-12848).

Maly DJ, Choong IC, Ellman JA .Combinatorial target-guided ligand assembly: identification of potent subtype-selective c-Src inhibitors. Proc Natl Acad Sci U S A. 2000 Mar 14;97(6):2419-24

Metcalf CA 3rd, Eyermann CJ, Bohacek RS, Haraldson CA, Varkhedkar VM, Lynch BA, Bartlett C, Violette SM, Sawyer TK.Structure-based design and solid-phase parallel synthesis of phosphorylated nonpeptides to explore hydrophobic binding at the Src SH2 domain. J Comb Chem. 2000 Jul-Aug;2(4):305-13.

Mongiovi AM, Romano PR, Panni S, Mendoza M, Wong WT, Musacchio A, Cesareni G, Di Fiore PP.A novel peptide-SH3 interaction. EMBO J. 1999 Oct l;7S(79):5300-9.

Morken JP, Kapoor TM, Feng S, Shirai F, Schreiber SL. Exploring the Leucine-Proline Binding Pocket of the Src SH3 Domain Using Structure-Based, Split-Pool Synthesis and Affinity-Based Selection, J. Am. Chem. Soc. 1998; 220:30-36. Nguyen JT, Turck CW, Cohen FE, Zuckermann RN, Lim WA.Exploiting the basis of proline recognition by SH3 and WW domains: design of N-substituted inhibitors. Science.

1998 Dec l l;2S2(5396):2088-92

Qiu X, Pohl E, Holmes RK,& Hoi WGJ. High-resolution structure of the diphtheria toxin repressor complexed with cobalt and manganese reveals an SH3-like third domain and suggests a possible role of phosphate as co-repressor. Biochemistry, 1996; 35:12292-12302.

Posey JE, Hardham JM, Norris SJ, Gherardini FC. Characterization of a manganese- dependent regulatory protein, TroR, from Treponema pallidum. Proc Natl Acad Sci U S A.

1999 Sep 14/96(79): 10887-92

Rickles RJ, Botfield MC, Zhou XM, Henry PA, Brugge JS, & Zoller MJ. Phage display selection of ligand residues important for Src homology 3 domain binding specificity. Proc Natl Acad Sci, USA, 92:10909-12913.

Rickles RJ, & Berg LJ. Identification of ITK TSK Src homology 3 domain ligands. J Biol Chem, 1966; 277:25646-25656.

Schiering N, Tao X, Zeng H, Murphy JR, Petsko G, & Ringe D. Structures of the apo-and the metal ion-activated forms of the diphtheria tox repressor from Corynebacterium diphtheriae. Proc Natl Acad Sci, USA, 1995; 92:9843-9850.

Schmitt MP. 1997. Transcription of the Corynebacterium diphtheriae hmuO gene is regulated by iron and heme. 7n/ect Immun, 1997; 65:4634-4641.

Schmitt MP, & Holmes RK. Characterization of a defective diphtheria toxin repressor (dtxR) allele and analysis of dtxR transcription in wild-type and mutant strains of Corynebacterium diphtheriae. Infect Immun, 1991; 59:3903-3908. Schmitt MP, & Holmes RK. Cloning, sequence, and footprint analysis of two promoter/operators from Corynebacterium diphtheriae that are regulated by the diphtheria toxin repressor (DtxR) and iron. J Bacteriol, 1994; 776:1141-1149.

Schumaker TN, Mayr LM, Minor DL Jr, Milhollen MA, Burgess MW, & Kim PS. Identification of D-peptide ligands through mirror-image phage display. Science, 1996; 277:1854-1857.

Sicheri F, Maorefi I, & Kuriyan J. Crystal structure of the Src family tyrosine kinase Hck. Nature, 1997; 385:602-609.

Scott JK, & Smith GP. Searching for peptide ligands with an epitope library. Science.1990; 249:386-390.

Scott JK, Loganathan D, Easley RB, Gong X, & Goldstein IF. A family of concanavalin A-binding peptides from a hexapeptide epitope library. Proc Natl Acad Sci, USA, 1992; 89:5398-5402.

Sparks AB, Adey NB, Quilliam LA, Thorn JM, & Kay BK. Screening phage-display random peptide libraries for SH3 ligands. Methods in Enzymology, 1995; 255:498-509.

Sun L, vanderSpek J, Murphy JR. Isolation and characterization of positive dominant mutants of the diphtheria toxin repressor, DtxR. Proc Natl Acad Sci, USA, 1998; 95: 14985-14990.

Tao X, Boyd J, & Murphy JR. Specific binding of the diphtheria tox regulatory element DtxR to the tox operator requires divalent heavy metal ions and a 9-base-pair interrupted palindromic sequence. Proc Natl Acad Sci, USA, 1992; 89:5897-5901.

Vartanian, P, Henry, M and Wain-Hobson, S (1996) Hypermutagenic PCR involving all four transitions and a sizeable portion of trans versions. NAR 24(14)2627-2631 Wang G, Wylie GP, Twigg PD, Caspar DLD, Murphy JR, & Logan TM. Solution structure, peptide binding, and function of the SH3 domain of the prokaryotic diphtheria toxin repressor. Proc Natl Acad Sci, USA, 1999; (in press).

White A, Ding X, Murphy JR, & Ringe D. Structure of metal ion-activated diphtheria toxin repressor / tox operator complex. Nature, 1998; 394: 502-506.

Xu W, Harrison SC, Eck ML. Three dimensional structure of the tyrosine kinase c_sSrc. Nature, 1997; 385:595-602

Industrial Applicability

The present invention is useful in the treatment of diseases and infection. All publications mentioned in this specification are indicative of the level of skill of persons skilled in the art to which this invention pertains. All these publications are herein incorporated by reference to the same extent as if each individual publication was specifically and individually indicated as being incorporated by reference.

Claims

Claims:

1. A method of treating a mammal afflicted with a disease or infection caused or mediated by a prokaryote, wherein expression of one or more virulence genes in the prokaryote is regulated by a protein containing an SH3 domain, comprising administering to the mammal therapeutic agent comprising an SH3 ligand in an amount effective to treat the disease.

2. The method of claim 1 wherein the prokaryote is a Gram-positive bacterium.

3. T e method of claim 2 wherein the bacterium is a Mycobacterial species.

4. The method of claim 3 wherein the bacterium is Mycobacterium avium.

5. The method of claim 3 wherein the bacterium is Mycobacterium tuberculosis.

6. The method of claim 3 wherein the bacterium is Mycobacterium leprae.

7. The method of claim 3 wherein the bacterium is Mycobacterium paratuberculosis.

8. The method of claim 3 wherein the bacterium is Mycobacterium bovis.

9. The method of claim 2 wherein the bacterium is a Staphylococcal species.

10. The method of claim 9 wherein the bacterium is Staphylococcus aureus.

11. The method of claim 9 wherein the bacterium is Staphylococcus epidermidis.

12. The method of claim 2 wherein the bacterium is Enterococcus faecalis.

13. The method of claim 2 wherein the bacterium is T. pallidum.

14. The method of claim 2 wherein the bacterium is a Stieptococcal species.

15. The method of claim 14 wherein the bacterium is Streptococcus penumoniae.

16. The method of claim 14 wherein the bacterium is Streptococcus pyogenes.

17. The method of claim 14 wherein the bacterium is Streptococcus meningitis.

18. The method of claim 14 wherein the bacterium is Streptococcus mutans.

19. The method of claim 1 wherein the protein is DtxR.

20. The method of claim 1 wherein the protein is a DtxR homolog.

21. The method of claim 20 wherein the DtxR homolog is IdeR.

22. The method of claim 20 wherein the DtxR homolog is DesR.

23. The method of claim 20 wherein the DtxR homolog is SirR.

24. The method of claim 1 wherein said mammal is a human.

25. The method of claim 1 wherein said mammal is a livestock animal.

26. The method of claim 1 wherein said SH3 ligand comprises a peptide represented by the consensus sequence Pro-P*-P*-P*-Pro-H-Gly-P-Pro-_Qe-Pro-Gly-H or [+/-]-[+/-]-H-Gly, wherein Pro represents a proline residue, Gly represents a glycine residue, lie represents an isoleucine residue, P* represents a threonine or serine residue, H represents a hydrophobic amino acid, [+/-] represents a charged amino acid, and P represents a polar amino acid.

27. The method of claim 1 wherein said SH3 ligand comprises a peptide represented by the consensus sequence Pro-[+/-]-P*-P-Pro-P-Gly-Gly-Val-Ile-Pro-[+/-]-P or [+/-]-[+/-], wherein Val represents a valine residue.

28. The method of claim 1 wherein said SH3 ligand comprises a peptide represented by the consensus sequence Pro-[+/-]-P* or H-H-Pro-P-Gly-Gly-Thr-Ile-Pro-H-P or [+/-]-Gly-[+/-]-H-H, wherein Thr represents a threonine residue.

29. A composition of matter comprising a peptide represented by the consensus sequence Pro-P*-P*-P*-Pro-H-Gly-P-Pro-Ile-Pro-Gly-H or [+/-]-[+/-]-H-Gly, wherein Pro represents a proline residue, Gly represents a glycine residue, lie represents an isoleucine residue, P* represents a threonine or serine residue, H represents a hydrophobic amino acid, [+/-] represents a charged amino acid, and P represents a polar amino acid.

30. The composition of matter of claim 29 wherein said peptide comprises the sequence VHRSPFGN PIPGLGEIGL.

31. The composition of matter of claim 29 wherein said peptide comprises the sequence VSRSPFGN PIPGLDELGV.

32. The composition of matter of claim 29 wherein said peptide comprises the sequence PTTSPFGN PIPGLVELGV.

33. The composition of matter of claim 29 wherein said peptide comprises the sequence PTTSPFGN PIPGLTELAV.

34. The composition of matter of claim 29 wherein said peptide comprises the sequence PTTSPFGN PIPGLLDLGA.

35. The composition of matter of claim 29 wherein said peptide comprises the sequence PTESPYGN PIPGLEELGE.

36. The composition of matter of claim 29 wherein said peptide comprises the sequence PQRDPHGD PIPGADGQVP.

37. A composition of matter comprising a peptide represented by the consensus sequence Pro-[+/-]-P* or H-H-Pro-P-Gly-Gly-Thr-Ile-Pro-H-P or [+/-]-Gly-[+/- ]-H-H, wherein Thr represents a threonine residue.

38. The composition of matter of claim 37 wherein said peptide comprises the sequence PKTCPHGG VIPRGNSDAA.

39. The composition of matter of claim 37 wherein said peptide comprises the sequence PETCPHGG VIPRNNEYKE.

40 . The composition of matter of claim 37 wherein said peptide comprises the sequence PEFCPHGG VIPEDNQPIH.

41. A composition of matter comprising a peptide represented by the consensus sequence Pro-P*-P*-P*-Pro-H-Gly-P-Pro-He-Pro-Gly-H or [+/-]-[+/-]-H-Gly, wherein Pro represents a proline residue, Gly represents a glycine residue, lie represents an isoleucine residue, P* represents a threonine or serine residue, H represents a hydrophobic amino acid, [+/-] represents a charged amino acid, and P represents a polar amino acid.

42. The composition of matter of claim 41 wherein said peptide comprises the sequence PKACPHGG TIPAKGELLV.

43. The composition of matter of claim 41 wherein said peptide comprises the sequence PKVCPHGG TIPGHGQPLV.

44. The composition of matter of claim 41 wherein said peptide comprises the sequence PKTCPHGG TIPAKGELLV.

45. The composition of matter of claim 41 wherein said peptide comprises the sequence PKTCPHGG TIPAKGELLV.

46. A composition of matter comprising a peptide having the sequence MTTPSAQLTLTKGNKSWVPGPPSRSTVSISLISNSSSVPL.