AU3553399A

AU3553399A - Recombinant proteins of treponema pallidum and their use for a syphilis vaccine

Info

Publication number: AU3553399A
Application number: AU35533/99A
Authority: AU
Inventors: Caroline E. Stebeck Cameron; Glaber A. Centurion-Lara; Sheila A. Lukehart; Wesley C. Van Voorhis
Original assignee: University of Washington
Current assignee: University of Washington
Priority date: 1998-04-10
Filing date: 1999-04-09
Publication date: 1999-11-01
Also published as: CA2325576A1; JP2002511275A; WO1999053099A1; EP1071819A1

Description

WO 99/53099 1- PCT/US99/07886 RECOMBINANT PROTEINS OF TREPONEMA PALLIDUM AND THEIR USE FOR A SYPHILIS VACCINE Field of the Invention Isolated nucleic acids and polypeptides from Treponema pallidum subspecies 5 pallidum, pertenue, and endemicum and the use of these molecules to elicit protective immunity against this organism. Background of the Invention Primary syphilis is characterized by a painless primary ulcerative lesion called a "chancre" that generally develops at the site of inoculation after sexual contact with an 10 infected person. The chancre is the site of proliferation of the spirochete Treponema pallidum subspecies pallidum (T. p. pallidum), which causes syphilis. The chancre gradually resolves, and weeks to months later a rash characteristic of secondary syphilis usually develops. Syphilis also can be transmitted congenitally. Without appropriate antibiotic treatment, T. p. pallidum establishes a lifelong 15 chronic infection. Approximately 30% of patients in late stages of the disease develop tertiary neurologic, bony, hepatic, or circulatory system manifestations which may occur decades after the primary infection event. Pathogenic members of the genus Treponema include at least, four natural human pathogens and one natural rabbit pathogen. Based in part upon saturation 20 reassociation kinetics assays (Miao, R.M., and A.H. Fieldsteel, J Bacteriol. 141:427 429, 1980) three of the human pathogens are currently classified as subspecies of Treponema pallidum. These are Treponema pallidum subspecies pallidum, Treponema pallidum subspecies pertenue, and Treponema pallidum subspecies endemicum, which, respectively, cause venereal syphilis, yaws, and bejel. A fourth WO 99/53099 PCT/US99/07886 -2 treponeme, Treponema carateum, causes a disease called pinta. Yaws and bejel occur primarily in warm, humid, tropical areas of the world, primarily in children, and are transmitted by direct non-sexual contacT. Like syphilis, these diseases are characterized by primary lesions that heal within days or weeks, followed by a more 5 serious secondary phase that is systemic. Some cases of bejel exhibit tertiary symptoms as well. In addition, poorly characterized spirochetes have been isolated in plaque associated with gingivitis and periodontal lesions, and are believed to be etiologic agents of that condition. These oral treponemes are known to be reactive with a monoclonal antiserum specific for a 47 kDa protein found in T. p. pallidum, 10 thus appear to be subspecies or strain of T. p. pallidum (Riviere et al., N. Eng. J. Med. 325:539-543, 1991). Another treponeme, Treponema paraluiscuniculi, causes venereal syphilis in rabbits, and is non-infectious to humans. These various pathogenic treponemes are morphologically identical and are antigenically highly cross-reactive, e.g., currently available serological tests cannot distinguish yaws 15 infection from syphilis. Pathogenic varieties of T. pallidum, including subspecies pallidum, and endemicum, have remained refractory to being propagated in culture for more than a few passages, a circumstance that has hampered efforts to fully characterize these organisms and their pathology. However, these bacteria all can be propagated by 20 serial inoculation of rabbit testes. Moreover, the rabbit provides a good experimental model for treponemal disease, in that rabbits develop primary chancres much like humans and also develop persistent infection in their lymph nodes and central nervous systems (Turner, T.B., and D.H. Hollander, Biology of the Treponematoses, World Health Organization, Geneva, 1957). Rabbits, however, do not manifest secondary or 25 tertiary syphilis. A syphilis vaccine clearly is needed due to a recent upsurgence worldwide in the frequency of occurrence of this disease. Between 1985 and 1990, the number of reported syphilis cases in the United States increased from 27,131 to 50,578 (Rolfs, R.T., MMAIWR 42:13-19, 1993). Worldwide, over 3 million cases annually are 30 estimated to have occurred during that time period. To exacerbate the problem, syphilis infections appear to increase the risk of acquisition and transmission of human immunodeficiency virus (HIV) (Greenblatt, R.M., et al., AIDS 2:47-50, 1988; Simonsen, J.N., et al., N Engl. J. Med. 319:274-278, 1988; Darrow, W.N., et al., Am. J. Public Health 77:479-483, 1987). These circumstances have spurred efforts 35 to develop a vaccine for syphilis, but as of yet no practical vaccine effective against WO 99/53099 PCT/US99/07886 -3 this disease has been reported. Moreover, no vaccines exist for yaws or bejel, both of which are serious treponemal diseases that take a heavy toll in tropical and subtropical regions of the world. To enable rational vaccine design more information is needed about 5 treponemal interaction with the immune system and, specifically, the immune evasion mechanisms employed by T. p. pallidum. One of the central paradoxes of syphilis is the induction of a rapid humoral and cellular immune response that is capable of eliminating millions of treponemes from primary syphilitic lesions, but incapable of eradicating the few organisms that remain during latency. Macrophages are believed 10 to be responsible for this rapid clearance of T. pallidum from early lesions, presumably through antibody-mediated treponemal opsonization and subsequent phagocytosis and killing by macrophages (e.g., see Lukehart and Miller, J. Immunol. 121:2014-2024, 1978; Baker-Zander and Lukehart, J. InfecT. Dis. 165:69-74, 1992). In support of this, antibody has been demonstrated 15 to enhance phagocytosis of treponemes by macrophages (Lukehart and Miller, J. Immunol. 121:2014-2024, 1978) and is required for macrophage-mediated killing of T. pallidum (Baker-Zander and Lukehart, J. InfecT. Dis. 165:69-74, 1992). In addition, the systemic appearance of opsonic antibody has been shown to immediately precede bacterial clearance in the rabbit model (Baker-Zander et al., FEMS Immunol. 20 Med. Microbiol. 6:273-280, 1993). Although no success has been reported for efforts to protect animals by immunization with defined antigens of T. p. pallidum, complete protection against homologous challenge with T. p. pallidum was achieved in at least one instance following 60 injections of y-irradiated T. p. pallidum (Miller, J. N., 25 J Immunol. 110:1206-1215, 1973). Moreover, persons infected with the highly related T. p. pertenue, which causes yaws, exhibit partial immunity to T. p. pallidum, and similarly, persons infected with one strain of T. p. pallidum exhibit partial immunity against infections with other strains (Turner and Hollander, Biology of the Treponematoses, World Health Organization, 1957). These observations indicate that 30 a vaccine that induces protective immunity against syphilis is a plausible goal, but that antigens useful for such a vaccine have not yet been discovered. To date, most T. p. pallidum antigens considered as vaccine candidates have been selected simply on the basis of their reactivity with immune rabbit serum (IRS), i.e., the serum of rabbits that are immune to syphilis by virtue of having been 35 previously infected with T. p. pallidum. This approach has led to the identification of WO 99/53099 PCT/US99/07886 -4 a number of interesting and important lipoprotein and protein antigens, but has failed so far to provide any protein capable of protecting experimental animals from challenge with T. p. pallidum. T. p. pallidum is a highly motile spirochete containing an outer membrane, a 5 periplasmic space, a peptidoglycan-cytoplasmic membrane complex, and a protoplasmic cylinder. Proteins associated with the outer membrane are more likely to be exposed to the host immune system, and thus are more likely than other treponemal proteins to elicit an immune response by the infected hosT. However, studies have indicated that T. p. pallidum has about 100-fold fewer trans-membrane 10 proteins than does a typical gram negative bacterium (Radolf, J.D., et al., Proc. Natl. Acad Sci. USA 86:2051-2055, 1989; Walker, E.M., et al., J. Bacteriol 171:5005-11, 1989). Because of their paucity, some investigators have assigned T. p. pallidum outer membrane proteins a special name, "T. pallidum rare outer membrane proteins," or "TROMPS." Candidate TROMPS include 65-, 31- (basic and acidic pI forms), and 15 28- kDa proteins that are found in the outer membrane fraction (Blanco, D. R., et al., J. Bacteriol., 176:6088-6099, 1994; Blanco, D. R., et al., Emerg. InfecT. Dis. 3:11 20, 1997). However, no TROMP nor any other T. p. pallidum protein has definitively been identified as being located in the outer membrane, nor has any candidate outer membrane protein been shown to induce a protective immune 20 response (Radolf JD, et al., InfecT Immun. 56:490-498, 1988; Radolf et al., InfecT Immun. 56:1825-1828, 1988; Cunningham et al., J Bacteriol., 170:5789 5796, 1988;[?]11; Blanco et al., J Bacteriol. 176:6088-6099, 1994; Cox et al., Molec. Microbiol., 15:151-1164, 1995; Radolf, J. D., Molec. Microbiol., 16:1067 1073, 1995). For example, a recent report suggests that TROMP 1 is localized to the 25 cytoplasmic membrane, suggesting it is not surface exposed (Akins, D. R., et al., J. Bacteriol., 179:5076-5086, 1997). Moreover, neither of the two TROMP genes so far identified is found in greater than one copy and therefore neither appears to function in antigenic variation. In addition, several of the highly immunogenic lipoprotein antigens (47, 34, 17, and 15 kDa) already identified for T. pallidum have 30 been shown to not be exposed on the outer membrane (Radolf, J.D., Mol. Microbiol. 16:1067-1073, 1995). On June 24, 1997, a preliminary copy of the entire genome of T. p. pallidum, Nichols strain, was posted on the Internet at http://utmmg.med.uth.tmc.edultreponemaldocs/update.html. This copy of the 35 Tp. pallidum genome was not annotated to denote the positions of open reading WO 99/53099 PCT/US99/07886 -5 frames, though subsequent updates to this original posting have noted open reading frames and have provided other information. Summary of the Invention Two genes and one multi-membered gene family have been identified that are 5 useful for eliciting a protective immune response against infection by T. p. pallidum, the bacterium that causes syphilis. The nucleotide sequences of these new genes have been determined. In an experimental rabbit model, immunization with the protein products of several of these genes elicited significant protection upon subsequent challenge with virulent T. p. pallidum. These proteins represent the first 10 immunoprotective antigens that have been reported for T. pallidum subsp. pallidum. Comparative sequence analysis has indicated that one of the genes identified here (SEQ ID NO:1) encodes a 356 amino acid protein (SEQ ID NO:2) that is a glycerophosphodiester phosphodiesterase (hereafter called "Gpd"), a glycerol metabolizing enzyme previously identified in other bacteria, e.g., Haemophilus 15 influenzae, Escherichia coli, Bacillus subtilis and Borrelia hermsii (Janson, H., et al., InfecT Immun., 59:119-125, 1991; Munson, R.S., et al., J.1 Bacteriol., 175:4569 4571, 1993; Tommassen, J., et al., Mol. Gen. Genet., 226:321-327, 1991; Schwan, T.G., et al., J. Clin. Microbiol. 34:2483-2492, 1996; Shand, E.S., et al., J. BacterioL, 179:2238-2246, 1997; Nilsson, R.P. et al., Microbiol., 140:723-730, 1994). The 20 identification of this protein (SEQ ID NO:2) has been previously published (Stebeck et al., FEMS Microbiol. Letters, 154:303-310, 1997; Shevchenko et al., InfecT Immun., 65:4179-4189, 1997). Experiments were conducted to confirm that the product of the T. pallidum Gpd homologue (SEQ ID NO: 1) exhibited the expected Gpd activity, and anti-Gpd antibodies were used to confirm that 25 T. p. pallidum indeed expresses a cross-reactive protein of the predicted molecular size. Injection of recombinant Gpd (SEQ ID NO:2) into rabbits was shown to elicit a partially protective immune response upon subsequent challenge with T. p. pallidum. In addition to Gpd (SEQ ID NO:2), the invention provides another protein believed to be associated with the outer membrane, and that has homology with the surface 30 exposed D15 protein from Haemophilus influenzae (Flack, F.S., et al., Gene, 156:97 99, 1995), and Oma87 from Pasteurella multocida (Ruffolo and Alder, InfecT Immun., 64:3161-3167, 1996). This protein is herein referred to as the "D15/Oma87 homologue" and is encoded by the nucleic acid molecule having the sequence set forth in SEQ ID NO:3. The amino acid sequence of the D15/Oma87 35 homologue is set forth in SEQ ID NO:4. SEQ ID NO:5 sets forth the nucleic acid WO 99/53099 PCT/US99/07886 -6 sequence of a portion of the coding region of the D15/Oma87 homologue gene (SEQ ID NO:3) that was expressed to yield a D15/Oma87 homologue polypeptide fragment (SEQ ID NO:6) that was recovered and used for vaccine testing, as more fully described herein. 5 In addition to Gpd (SEQ ID NO:2) and D15 (SEQ ID NO:4), a novel polymorphic, multicopy gene family (called Msp) has been identified in T. p. pallidum, T. p. pertenue and T. p. endemicum. Members of this gene family have homology to the major outer sheath protein (Msp) of T. denticola. The members of this gene family are divided into several subfamilies, and present within each subfamily are 10 regions that are highly conserved as well as variable regions that are far less conserved. Analysis of their amino acid sequences suggests that many of these molecules are likely to be outer surface exposed. Furthermore, injection of rabbits with several of these proteins has resulted in partial protective immunity of the rabbits upon challenge with a large dose of T. p. pallidum, thus these proteins are useful as 15 vaccine antigens. The nucleic acid sequences of cloned T. p. pallidum Msp genes (or portions thereof), and the proteins encoded by the T. p. pallidum Msp genes, are disclosed in the following sequence listing entries: Mspl (SEQ ID NO:7), Mspl protein (SEQ ID NO:8); Msp2 (SEQ ID NO:9), Msp2 protein (SEQ ID NO:10); Msp3 (SEQ ID 20 NO: 11), Msp3 protein (SEQ ID NO:12); Msp4 (SEQ ID NO:13), -Msp4 protein (SEQ ID NO:14); Msp5 (SEQ ID NO:15), Msp5 protein (SEQ ID NO:16); Msp6 (SEQ ID NO:17), Msp6 protein (SEQ ID NO:18); Msp7 (SEQ ID NO:19), Msp7 protein encoded by open reading frame A (SEQ ID NO:20), Msp7 protein encoded by open reading frame B (SEQ ID NO:21); Msp8 (SEQ I NO:22), Msp8 protein (SEQ 25 ID NO:23); Msp9 (SEQ ID NO:24), Msp9 protein (SEQ ID NO:25); Mspl0 (SEQ ID NO:26), Mspl0 protein (SEQ ID NO:27); Mspll (SEQ ID NO:28), Mspll protein (SEQ ID NO:29); and Mspl2 (SEQ ID NO:30), Mspl2 protein (SEQ ID NO:31). The amino acid sequence of a highly conserved amino acid motif found within all of the Msp genes of T. p. pallidum is set forth in SEQ ID NO:32. The 30 nucleic acid sequence encoding the conserved amino acid sequence motif disclosed in SEQ ID NO:32 is set forth in SEQ ID NO:33. The nucleic acid sequences of cloned T. p. pertenue Msp genes, and the proteins encoded by the T. p. pertenue Msp genes, are disclosed in the following sequence listing entries: T. p. pertenue Msp homologue 1 (SEQ ID NO:34), Msp 35 homologue 1 protein (SEQ ID NO:35); T. p. pertenue Msp homologue 2 (SEQ ID WO 99/53099 PCT/US99/07886 -7 NO:36), Msp homologue 2 protein (SEQ ID NO:37); T. p. pertenue Msp homologue 3 (SEQ ID NO:38), Msp homologue 3 protein (SEQ ID NO:39); T. p. pertenue Msp homologue 4 (SEQ ID NO:40), Msp homologue 4 protein (SEQ ID NO:41). The amino acid sequence of a highly conserved amino acid motif found within all of the 5 Msp genes of T. p. pertenue is set forth in SEQ ID NO:42. The nucleic acid sequences of a cloned T. p. pallidum Msp gene (T.P. 1.6) is disclosed in SEQ ID NO:43, and the protein encoded by the nucleic acid sequence disclosed in SEQ ID NO:43 is disclosed in SEQ ID NO:44. SEQ ID NO:45 shows the nucleotide sequence of a subportion of the T.P. 1.6 DNA fragment (SEQ ID 10 NO:43) that was expressed to obtain a polypeptide (SEQ ID NO:46) to be tested for efficacy in eliciting a protective immune response against T. p. pallidum (see Example 10). SEQ ID NO:47 shows a highly conserved motif present in the amino acid sequence of SEQ ID NO:43. Detailed Description of the Preferred Embodiment 15 This invention relates to isolated nucleic acids, polypeptides and methods that are useful for preparing vaccines to protect against infection by Treponema spp., particularly Treponema pallidum subspecies pallidum, Treponema pallidum subspecies pertenue, and Treponemapallidum, subspecies endemicum. As used here, the term "isolated" refers to a biological molecule that is separated from its natural 20 milieu, i.e., from the organism or environment in which it is normally presenT. In certain embodiments, the invention provides isolated polypeptides capable of inducing a protective immunologic response to T. p. pallidum, T. p. pertenue, and T. p. endemicum when administered in an effective amount to an animal hosT. Preferred embodiments of such polypeptides include those whose amino acid 25 sequences are shown in SEQ ID NOS:2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 21, 23, 25, 27, 29, 31, 32, 35, 37, 39, 41, 42, 44 and 46. The invention provides representative examples of nucleic acid molecules capable of encoding these polypeptides in SEQ ID NOS:1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 22, 24, 26, 28, 30, 33, 34, 36, 38, 40, 43 and 45. 30 Isolated polypeptides and nucleic acids according to the invention maybe prepared by use of recombinant DNA techniques, or may be synthesized using widely available technology. The use of recombinant methods to prepare the subject vaccines provides the advantage that the immunogenic components of the vaccines can thus be prepared in substantially purified form free from undesired contaminants.

WO 99/53099 PCT/US99/07886 -8 The invention, in one aspect, provides isolated nucleic acids capable of encoding the polypeptides whose amino acid sequences are disclosed herein. In another aspect, the invention provides a nucleic acid molecule (SEQ ID NO: 1) encoding a newly identified T. p. pallidum protein (SEQ ID NO:2) that has 5 glycerophosphodiester phosphodiesterase activity (Gpd), and functional equivalents thereof Also encompassed by the present invention is a polypeptide encoded by the nucleic acid of (SEQ ID NO: 1), and whose amino acid sequence is shown in (SEQ ID NO:2). The term "functional equivalent," as used herein, is intended to include all immunogenically active substances capable of evoking an immune response in 10 animals, including humans, to which the equivalent polypeptide or nucleic acid has been administered, wherein the resulting antibody has immunologic reactivity with the indicated polypeptide. Thus, equivalents of T. p. pallidum Gpd (SEQ ID NO:2) may include mutant or recombinantly modified forms of the protein, or subportions of the Gpd molecule that retain sufficient epitopic similarity to the native protein (SEQ ID 15 NO:2) to evoke an antibody response similar to that evoked by the epitope when present in the native protein. The invention further provides nucleic acids (such as that shown in SEQ ID NO:3) that encode a protein that has significant homology both with the D15 protein previously identified in H. influenzae and with the Oma87 protein previously 20 identified in Pasteurella multocida. This T. p. pallidum protein hereafter is referred to as the "D15/Oma87 homologue"), and its amino acid sequence is shown in SEQ ID NO:4. Provided also is the nucleic acid molecule shown in SEQ ID NO:5, which encodes a subportion of the amino acid sequence shown in SEQ ID NO:4. The polypeptide encoded by the nucleic acid molecule of SEQ ID NO:5 encodes the 25 polypeptide of SEQ ID NO:6, which is useful as a vaccine against syphilis. The invention encompasses the D15/Oma87 polypeptides whose amino acid sequences are shown in SEQ ID NO:4 and SEQ ID NO:6, and functional equivalents thereof. In other aspects of the invention, SEQ ID NOS:7, 9, 11, 13, 15, 17, 19, 22, 24, 26, 28 and 30 depict nucleic acids encoding portions of 12 different T. p. pallidum 30 polypeptides (having amino acid sequences set forth in SEQ ID NOS:8, 10, 12, 14, 16, 18, 20, 21, 23, 25, 27, 29 and 31) that have homology with the previously described major sheath protein of T. denticola. These T. p. pallidum Msp homologues hereafter are referred to as "T.p. pallidum Msp proteins (or "homologues" or polypeptides)," whether the reference is to the full-length protein, or 35 to a subportion of the protein. The invention therefore provides the polypeptides WO 99/53099 PCT/US99/07886 -9 having the amino acid sequences shown in SEQ ID NOS:8, 10, 12, 14, 16, 18, 20, 21, 23, 25, 27, 29 and 31, and functional equivalents thereof The terminology used for the T. p. pallidum genome project (posted at http://utmmg.med.uth.tmc.edu/treponema/docs/update.html) refers to the Msp genes 5 as "treponemal pallidum repeats" rather than "Msp" genes, and designates them as "TPR A-L". The nomenclature used herein refers instead to Tpr A-L as Msp 1-Msp 12. Msps 1-12 correspond, respectively, to Tgr G, F, E, D, C, B, A, L, K, J, I and H. The full-length open reading frames for these 12 genes, according to the present version of the T. p. pallidum genome project, encode proteins of the 10 following sizes: Msp 1, 756 amino acids; Msp 2, 364 amino acids; Msp 3, 762 amino acids; Msp 4, 598 amino acids; Msp 5, 598 amino acids; Msp 6, 644 amino acids; Msp 7 (ORF A), 253 amino acids; Msp 7 (ORF B), 389 amino acids; Msp 8, 443 amino acids, Msp 9, 480 amino acids; Msp 10, 758 amino acids; Msp 11, 609 amino acids; Msp 12, 693 amino acids. 15 All of the T. p. pallidum Msp homologues contain a highly conserved peptide motif encoded by the nucleic acid molecule whose nucleotide sequence is shown in SEQ ID NO:33, and whose amino acid sequence is shown in SEQ ID NO:32. In view of its high degree of conservation, this conserved peptide (SEQ ID NO:32) may be important in eliciting antibodies that will cross-react with all of the T. p. pallidum 20 Msps. To facilitate the expression of useful amounts of T. p. pallidum Msp proteins, the invention further provides the PCR primers shown in Table 1, in which "S" indicates the sense primer, and "AS" indicates the primer binding to the opposite strand, i.e., the antisense primer.

WO 99/53099 PCT/US99/07886 -10 0 0 ofzC z z Z ~ N z0 0Z(7 o Z 00 0 Z 0/ 0 2 w Ol 0 mQ Qu E-4~ HE u 0< 0 H< < F U Q H 0 H E-4H - H QH QQ <U F-4O 000 OON C F- F U ~u WO 99/53099 PCT/US99/07886 -11 Each of the primer pairs in Table 1 can be used to specifically amplify a portion of the T. p. pallidum Msp gene(s) as indicated in the last column of the table. In addition, the invention provides a PCR primer pair having the following nucleotide sequences: 5'-ACCAGTCCTTCCTGTGTGGTTAA (sense) (SEQ ID 5 NO:60), and 5'-ACTCCTTGGTTAGATAGGTAGCTC (antisense) (SEQ ID NO:61). This primer pair is useful for amplifying not only one of the Msp genes of T. p. pallidum, i.e., TP 1.6 (SEQ ID NO:43), but also for amplifying a portion of at least four different T.p. pertenue Msp genes, thus defining four genes in the

T

.p. pertenue genome that are highly related to the T. p. pallidum Msp gene family, 10 and that are encompassed by the present invention. These four amplified T. p. pertenue Msp DNA fragments have the nucleotide sequences shown in SEQ ID NOS:34, 36, 38 and 40, and the predicted amino acid sequences translated from these four amplicons are shown, respectively, in SEQ ID NOS:35, 37, 39 and 41. Three of these amplicons (SEQ ID NOS:36, 38 and 40) contain the same number of 15 nucleotides, but differ somewhat in nucleotide sequence, thus appear to represent fragments from different Msp homologues. The primer pairs shown in Table 1 as well as the primer pair 5'-ACCAGTCCTTCCTGTGTGGTTAA (sense) (SEQ ID NO:60), and 5'-ACTCCTTGGTTAGATAGGTAGCTC (antisense) (SEQ ID NO:61) can be used 20 in accord with this invention to amplify portions of the T. p. pallidum genome. The resulting amplified DNA (amplicons) can be expressed as recombinant proteins in E. coli or another suitable host, and the recombinant proteins thus derived used to formulate vaccines useful for eliciting a protective immune response against syphilis, yaws, bejel, or other treponemal diseases. For example, the primers designated as 25 "Set 1" in Table 1 are useful for amplifying portions of at least three Msp genes found in the genome of T. p. pertenue, and three Msp genes in the genome of T. p. endemicum (Example 7). In addition to the aforementioned nucleic acids, PCR primers and polypeptides, the invention provides two novel methods for identifying T. p. pallidum 30 proteins useful as vaccine candidates. The first of these methods involves the identification of T. p. pallidum proteins that are immunologically reactive with an opsonizing serum against T. p. pallidum but that are immunologically unreactive with a non-opsonizing serum (Stebeck et al., FEMS Microbiol. Lett., 154:303-310, 1997). Such proteins are likely to elicit protective immunity, hence are vaccine candidates, 35 i.e., useful for vaccine trials and for eventual inclusion in a vaccine. Vaccine WO 99/53099 PCT/US99/07886 -12 candidates are tested in a suitable host, i.e., one susceptible to infection with T. p. pallidum, for their ability to elicit an immune response that is protective against challenge by this organism. Rabbits, for example, can provide a suitable host for this purpose. Proteins that prove to be capable of eliciting such an immune response are 5 determined to be vaccine candidates. This method for selecting vaccine candidates can be applied to identify polypeptides capable of eliciting a protective immune response against yaws, bejel, or any other disease caused by a subspecies of T. pallidum that is susceptible to opsonizing antibodies. The rationale for the above-described strategy for obtaining vaccine 10 candidates is that opsonizing antibodies are known to be involved in treponeme clearance during primary syphilis, thus a vaccine containing antigens capable of eliciting opsonizing antibodies should produce resistance or immunity against infection with T. p. pallidum. The disclosed method for identifying T. p. pallidum proteins that are targets for opsonizing antibody requires the use of both opsonic and 15 non-opsonic antisera. One means of preparing opsonic serum is to use the rabbit model system. To prepare opsonic rabbit serum (ORS) using this system, serum from rabbits infected with T. p. pallidum is adsorbed to remove activity against the major known treponemal antigens, none of which is capable of eliciting protective immunity. Opsonic activity can be assessed by applying the rabbit macrophage phagocytosis 20 assay (Lukehart and Miller, J. Immunol., 121:2014-2024, 1978). Non-opsonic rabbit serum (NORS) can be derived from rabbits injected with heat-killed T. p. pallidum. To obtain clones corresponding to proteins that are targets for ORS, an expression library is constructed from T. p. pallidum genomic DNA, and the proteins thereby expressed are screened using both ORS and NORS. Plaques that interact with ORS 25 but not with NORS are isolated and the proteins they express are tested to determine whether they are capable of eliciting protective immunity in a susceptible hosT. In the representative examples given below, the application of this method has identified three different T. p. pallidum proteins, the above-described Gpd (four independent clones), the D15/Oma87 homologue, and one member of the T. p. pallidum Msp 30 family. Because of the method by which they were obtained, each of these three proteins appears to be a target for opsonizing antibodies, and all three likely are to be exposed on the surface of T. p. pallidum cells and capable if included in a vaccine of eliciting a protective immune response against syphilis. Prior efforts to identify the potential targets of opsonic antibody have focused 35 primarily on direct isolation of these proteins from the syphilis bacteria themselves.

WO 99/53099 PCT/US99/07886 -13 However, such efforts have been hampered because the T. p. pallidum outer membrane is extremely fragile and has a relatively low number of surface proteins (Walker et al., J. Bacteriol., 171:5005-5011, 1989; Radolf et al., Proc. Natl. Acad. Sci., 86:2051-2055, 1989 5 The invention further provides another method for obtaining vaccine candidates that involves identifying proteins that are expressed by genes that are present in the genome of T. p. pallidum but that are not present in the genome of the closely related treponeme, T. paraluiscuniculi, a pathogen that causes syphilis in rabbits but that does not infect humans. The genes thus isolated are presumed to 10 provide some function that enables T. p. pallidum to infect human cells. Accordingly, genes present in T. p. pallidum but absent from T. paraluiscuniculi are considered to be effective as a vaccine for syphilis, because antibodies directed against them are expected to protect against infection by T. p. pallidum. This method is applicable for identifying pathogenicity-related genes present in the genomes of other treponemes 15 that infect humans but not rabbits, e.g., the genomes of T. p. pertenue and T. p. endemicum. Genes identified by either of the aforementioned methods are tested to determine whether their gene products are capable of eliciting in an animal host an immune response that is protective against challenge with T. p. pallidum. This test 20 may be performed by any convenient means, for example, by inoculating rabbits intradermally or intramuscularly according to standard immunologic procedures with the protein being tested, then challenging the rabbit with a dose of T. p. pallidum that is capable of causing syphilis in an uninoculated rabbiT. One means for identifying proteins present in subspecies of T. pallidum but 25 absent from T. paraluiscuniculi is to use representation difference analysis (RDA), a PCR-based technique that selectively amplifies nucleic acid molecules that are present in one population of nucleic acids but absent from another. This method is effective using DNA from any subspecies of T. pallidum, including T. p. pallidum, T. p. pertenue, and T. P. endemicum. In the study described in Example 5, RDA was 30 used to obtain clones that permitted the isolation of a fragment of DNA, called herein "TP 1.6," (SEQ ID NO:43) that was found to be unique to the T. p. pallidum genome. The protein encoded by the nucleotide sequence shown in SEQ ID NO:43 is set out in SEQ ID NO:44. Both are included within the scope of this invention. Sequence analysis of TP 1.6 (SEQ ID NO:44) indicated that it shared a significant 35 degree of homology with Mspl 1 (SEQ ID NO:8) and Msp2 (SEQ ID NO: 10) of the WO 99/53099 PCT/US99/07886 -14 T. p. pallidum Msp gene family. It should be noted that another member of the Msp family, Msp 9 (SEQ ID NO:25), was also identified as described above by virtue of its specific reactivity with opsonizing antibody against T. p. pallidum. Thus, members of the T.p. pallidum Msp family have been identified by two independent methods 5 designed for isolating syphilis vaccine candidates. Experiments using the rabbit model system have borne out the expectation that the T. p. pallidum proteins reactive with ORS but not NORS are capable of eliciting antibodies that protect against T. p. pallidum (see Example 10). Accordingly, the subject invention provides a vaccine that includes a physiologically 10 acceptable carrier together with an effective amount of an isolated T. p. pallidum polypeptide capable of inducing a protective immunologic response to T. p. pallidum when administered to a suitable host, the isolated polypeptide being immunologically reactive with an opsonizing serum against T. p. pallidum but immunologically unreactive with a non-opsonizing serum against T. p. pallidum. 15 A rabbit model was used to test the capacity of these newly identified T. p. pallidum proteins to elicit protective immunity against T. p. pallidum because proteins that elicit protective immunity in rabbits are expected to have a similar effect in humans. This is because the clinical course of the disease is similar in both hosts and also because the range of antibody reactivities, measured by immunoblot, appears 20 to be the same in both rabbits and humans following infection with T. p. pallidum. For example, in both hosts, reactive IgM becomes detectable within days after the appearance of clinical disease, and declines after clearance, while IgG responses rise somewhat later, peak at about the time of clearance, and persist for a long period thereafter at relatively high levels (e.g., see Baker-Zander et al., 25 J. InfecT. Dis., 151:264-272, 1985; Baker-Zander et al., Sex. Trans. Dis., 13:214 220, 1986; Lukehart et al., Sex. Trans. Dis., 13:9-15, 1986). Moreover, these same studies indicated that antibodies directed against many of the same antigenic proteins appeared in both hosts during corresponding stages of the disease. These observations demonstrate that the human immune system sees basically the same 30 antigens for this pathogen as seen by the rabbit immune system, and that both hosts' immune systems attack the pathogen in a similar fashion. Similarly, rabbits are a suitable animal model for testing the efficacy of yaws or bejel vaccines prepared according to the above-discussed methods. The present studies confirm that the rabbit and human immune systems 35 respond similarly to infection with T. p. pallidum. Sera from rabbits infected with WO 99/53099 PCT/US99/07886 -15 T. p. pallidum, Nichols strain, or from human syphilis patients infected with unknown strains both were observed here to contain antibodies against several members of the Msp family, and both exhibited especially high levels of activity against Msp 9 (SEQ ID NO:25) and the D15/Oma87 homologue (SEQ ID NO:4). Moreover, immune 5 rabbit serum (IRS) was observed to react with Gpd (SEQ ID NO:2). As detailed in Example 10, T. p. pallidum proteins to be tested in rabbits for their protective capacity were expressed in E. coli, and the corresponding recombinant molecules were purified and used as immunizing antigens. In all cases, rabbits were immunized three times with 200 gg of the recombinant antigen. The 10 rabbits were subsequently challenged with 10 3 or 10 T. p. pallidum at multiple dermal sites three weeks after the last boost, and lesion development was monitored by comparison to a control group of rabbits that had received no immunization prior to challenge. Typical red, indurated ulcerating lesions appeared in the control unimmunized animals at days 5-7 post-challenge in animals that had received 10 5 15 treponemes, or at days 12 to 14 post-challenge for animals that had received 103 treponemes (Gpd challengers). The rabbits immunized with four of the Msp proteins were protected from challenge and did not exhibit typical development of progressive lesions at the corresponding time points. The mild lesions that did develop in the immunized rabbits healed very quickly compared to control animals, and 20 T. p. pallidum could not be detected by darkfield analysis in most of these atypical lesions. The term "vaccine" as used herein is understood to refer to a composition capable of evoking a specific immunologic response that enables the recipient to resist or overcome infection when compared with individuals that did not receive the 25 vaccine. Thus, the immunization according to the present invention is a process of causing increased or complete resistance to infection with Treponema species. The vaccines of the present invention involve the administration of an immunologically effective amount of one or more of the polypeptides described above, i.e., the entire proteins, or a functional equivalent thereof, in combination with 30 a physiologically acceptable carrier. This carrier may be any carrier or vehicle usually employed in the preparation of vaccines, e.g., a diluent, a suspending agent, an adjuvant, or other similar carrier. Preferably, the vaccine will include an adjuvant in order to increase the immunogenicity of the vaccine preparation. For example, the adjuvant may be selected from Freund's complete or incomplete adjuvant, aluminum 35 hydroxide, a saponin, a muramyl dipeptide, an immune-stimulating complex (ISCOM) WO 99/53099 PCT/US99/07886 -16 and an oil, such as vegetable oil, or a mineral oil, though other adjuvants may beused as well. In another aspect of the invention, the immunogenicity of the immunogenic protein may be coupled to a macromolecular carrier, usually a non-toxic biologically 5 compatible polysaccharide or protein, e.g., bovine serum albumin. One route by which the syphilis treponemes can enter the body is through the mucosal membranes, thus an effective vaccine optimally will prime the immune response at mucosal surfaces to recognize T. p. pallidum. Strategies that may be used to administer the subject vaccines in order to elicit a mucosal immune response 10 include using E. coli heat labile enterotoxin as an adjuvant, expression of immunogenic antigens by plasmids carried in attenuated Salmonella spp., microsphere or liposome delivery vehicles, ISCOMS, or naked DNA encoding antigenic proteins (Staats et al., Curr. Opin. Immunol., 6:572-583, 1994). DNA vaccines stimulate strong CTL responses, as well as helper T cell and B cell responses. Since CTL are 15 known to be present in syphilis primary and secondary lesions, and since infection with T. p. pallidum itself is known to be associated with the generation of protective immunity, a DNA vaccine thus is a preferred embodiment of the subject vaccine compositions. In a further aspect of the invention, genes encoding the vaccine polypeptides 20 of the present invention may be inserted into the genome of a non-pathogenic organism to provide a live vaccine for administration of the vaccines of the subject invention. For example recombinant vaccinia viruses have been employed for this purpose, as well as attenuated Salmonella spp. Efficient vaccines can be prepared by inserting a variety of immunogenic genes into the same live vaccine, thus providing 25 immunity against several different diseases in a single vaccine vehicle, e.g., a vaccine against many different sexually transmitted diseases. A particularly advantageous live vaccine is one that is engineered to express one or more of the subject immunogens on the outer surface of the bacteria expressing the vaccine proteins, thus maximizing the recipient's exposure to the immunogens in an orientation likely to resemble that 30 found in the treponemal pathogen, thereby eliciting an appropriate immune response. The amount of immunogenically effective component used in the vaccine will of course vary, depending on the age and weight of the vaccine recipient, as well as the immunogenicity of the particular vaccine componenT. For most purposes, a suitable dose will be in the range of 1-1000 jig of each immunogen, and more 35 preferably, 5-500 jig of each immunogen.

WO 99/53099 PCT/US99/07886 -17 The present invention provides vaccines that include the T. p. pallidum glycerophosphodiester phosphodiesterase, D15/Oma87 homologue, and the members of the Msp family, each to be administered alone or in various combinations in amounts sufficient to induce a protective immunologic response to infection by 5 T. p. pallidum in a host animal that is normally susceptible to syphilis. It is understood that the vaccine of the subject invention may contain one or more of the aforementioned proteins, as well as additional T. p. pallidum proteins identified by the above described methods. For example, the vaccine may include T. p. pallidum glycerophosphodiester phosphodiesterase in combination with one or more of the 10 Msps, or may include in addition the D15/Oma87 homologue. With regards to the T. p. pallidum glycerophosphodiester phosphodiesterase, this may be provided by expressing in a suitable expression vector system a nucleic acid having the nucleotide sequence shown in SEQ ID NO:1. The isolated T. p. pallidum D15/Oma87 homologue may be obtained by expressing in a suitable 15 vector system a nucleic acid molecule having the nucleotide sequence shown in SEQ ID NO:4. The isolated T. p. pallidum Msp may be derived by expressing in a suitable vector the full-length T. p. pallidum Msp genes, as their positions in the genome are now known, or alternatively, may be derived by PCR from the variable portions of the Msp genes, as set out in the Examples below. The variable regions of Msps 1, 3, 4, 5, 20 6, 7, 8, 9, 10, 11 and 12 are shown in SEQ ID NOS:7, 11, 13, 15, 17, 19, 22, 24, 26, 28 and 30, respectively, and polypeptides corresponding to these sequences can be obtained by standard recombinant technology, i.e., by expression in a suitable bacterium, yeast, or other expression system. Alternatively, the Msp polypeptide for use in a vaccine of the subject invention may be provided by the nucleic acid 25 molecules shown in SEQ ID NO:43 or SEQ ID NO:45, or their polypeptide products, shown in SEQ ID NO:44 and SEQ ID NO:46, respectively. In one embodiment of the invention, the vaccine includes several different Msps or may even include all of the Msps. In a preferred embodiment, the vaccine includes Msps 2 (SEQ ID NO: 10), 9 (SEQ ID NO:25) and 11 (SEQ ID NO:29). In other embodiments of the invention, 30 the vaccine may consist of a polypeptide that includes both conserved and variable regions of one or more Msps. For vaccines including the D15/Oma87 homologue, this may be provided by expressing in a suitable host a nucleic acid molecule having the nucleotide sequence as shown in SEQ ID NO:3 or SEQ ID NO:5. In addition to vaccines for T. p. pallidum, the present invention provides 35 vaccines to protect against yaws, which is caused by the treponeme T. p. pertenue.

WO 99/53099 PCT/US99/07886 -18 This vaccine contains an effective amount of at least one isolated Msp capable of inducing a protective immunologic response when administered to a suitable host; and a physiologically acceptable carrier as described above. The yaws vaccine includes one or more Msp homologues derived from the T. p. pertenue genome, and may be 5 obtained in isolated form by expressing in a suitable vector one of the nucleic acid sequences shown in SEQ ID NOS:34, 36, 38 or 40. Other polypeptides useful for yaws vaccines may be identified by applying the RDA method described in Example 5, wherein T. p. pertenue DNA is used as tester DNA. Similarly, polypeptides for a bejel vaccine can be identified by using T. p. endemicum DNA as tester. The efficacy 10 of polypeptides so identified can be tested for their ability to elicit protective immunity by using a rabbit model as described in Example 10 for testing syphilis vaccine candidates. The invention further encompasses vaccines against bejel, the disease caused by Treponema pallidum subspecies endemicum, and pinta, caused by Treponema 15 carateum. T. p. pallidum and T. p. pertenue, the causative agents of jaws and bejel both contain Msp genes related to those present in T. p. pallidum, by analogy, the closely related T. carateum must also contain Msp genes useful for vaccines, and these can be identified and isolated according to the methods disclosed herein. In a further aspect, the invention provides vaccines that provide protective immunity 20 against the T. p. pallidum-related treponemes that cause gingivitis and periodontal disease. The Msp genes of the oral pathogen treponemes are amplified using the primers disclosed herein (e.g., the primers of Table 1), and polypeptides expressed from the resulting amplicons are expressed and tested for their capacity to elicit protective immunity in a suitable animal host. 25 The subject invention includes methods of inducing a protective immune response against T. p. pallidum that involve administering to a susceptible host an effective amount of any of the aforementioned treponemal vaccines, e.g., the polypeptides whose amino acid sequences are shown in SEQ ID NO:2, SEQ ID NO:4, SEQ ID NO:6, SEQ ID NOS:8, 10, 12, 14, 16, 18, 20, 21, 23, 25, 27, 29 and 30 31, or any polypeptide whose coding region is amplifiable by one or more of the primer pairs of Table 1, or the primer pair 5'-ACCAGTCCTTCCTGTGTGGTTAA-3' (sense) (SEQ ID NO:60) and 5'-ACTCCTTGGTTAGATAGGTAGCTC-3' (antisense) (SEQ ID NO:61), or functional equivalents thereof. The vaccines may be administered by any of the methods well known to those skilled in the art, e.g., by 35 intramuscular, subcutaneous, intraperitoneal, intravenous injection, orally, or WO 99/53099 PCT/US99/07886 -19 intranasally. Naked DNA encoding the treponemal antigen or the treponemal polypeptide itself may be administered. The invention further provides a PCR-based method for analyzing a sample of treponemal genomic DNA to determine whether it originated from T.p. subspecies 5 pallidum, T. p. subspecies pertenue or T. p. subspecies endemicum. To carry out this method, DNA is isolated from the treponeme whose identity is at issue, or Chancre DNA is isolated, and this DNA is amplified using the PCR sense primer 5'-ACCAGTCCTTCCTGTGTGGTTAA-3' (SEQ ID NO:60) and antisense primer 5'-ACTCCTTGGTTAGATAGGTAGCTC-3' (SEQ ID NO:61), and the size 10 of the resulting DNA fragments, e.g., by gel electrophoresis, or by some other method. It is determined that the treponemal genomic DNA originated from T. p. pallidum if the size analysis of the restriction products reveals a single DNA fragment having a size of about 1.7 kb, or that the treponemal genomic DNA originated from T. p. subspecies pertenue if at least two DNA fragments having sizes 15 of about 1.7 and 1.3 kb are detected instead. If no DNA fragments result from amplification using this pair of primers, the treponeme DNA is determined to have originated from T. p. subspecies endemicum. Thus, when a patient presents with a primary lesion that appears to be caused by a treponemal infection, this test can be applied to quickly determine whether the patient suffers from syphilis, yaws, or bejel. 20 It is disclosed herein that sufficient variation exists within the Msp gene family among various clinical isolates of T. p. pallidum such that restriction fragment length polymorphism (RFLP) analysis can be used to differentiate the clinical isolates, thus providing a useful means for epidemiologic monitoring of cases of syphilis. The invention provides a method of RFLP analysis for determining whether clinical 25 isolates of T. p. pallidum from different syphilis patients are the same or differenT. This method utilizes PCR to amplify samples of genomic DNA from the clinical isolates, followed by restriction digestion and subsequent length analysis of the resulting DNA fragments. In an illustrative example of this technique, the variable domains of six alleles of the Msp gene family were amplified using the following 30 primers that bind to two short conserved regions that flank a highly variable region within the central portion of several members of the Msp family (see Example 7). The nucleotide sequences of the primers used in this example were 5'-CGACTCACCCTCGAACCA-3' (sense) (SEQ ID NO:48), and 5'-GGTGAGCAGGTGGGTGTAG-3' (antisense) (SEQ ID NO:49). After 35 amplification of the highly variable region using these primers or other primers that WO 99/53099 PCT/US99/07886 -20 amplify this same DNA region, the amplified DNA is digested with one or more restriction endonucleases that recognize a four-base cleavage site, and the resulting restriction fragments are analyzed on a gel. Experimental results presented below in Example 7 have indicated that the 5 high degree of variability observed in the RFLPs thus obtained is sufficient to distinguish many different individual isolates of T. p. pallidum. In a preferred embodiment, the restriction endonucleases used for differentiating individual isolates of T. p. pallidum are BstUI, AluI, Hhal and NlaIII, as these enzymes yielded 15 distinct patterns among 18 tested T. p. pallidum strains. The RFLP method described 10 here can be applied to clinical specimens without any need for the technically difficult and expensive isolation of the organism prior to analysis. Because aggressive contact tracing is relatively effective in the control of syphilis outbreaks, this method can provide a means for a public health entity to be able to identify a single strain of T. p. pallidum as responsible for a high proportion of incident cases versus the 15 multiple strains causing a background level of syphilis in a community, or to trace the parties involved in spreading clusters of the disease. Moreover, these same PCR primers were found also to amplify DNA segments from both T. p. pertenue and T. p. endemicum (Example 7). Digestion of these amplified DNAs with restriction enzymes has yielded distinctive patterns that 20 are sufficiently different from the patterns seen for T. p. pallidum to provide a diagnostic test for differentiating these three subspecies of T. p. pallidum. Also included in the invention is the nucleic acid molecule whose nucleotide sequence is shown in SEQ ID NO:45, and the polypeptide it encodes which is shown in SEQ ID NO:46. This polypeptide represents the amino terminal portion of the 25 TP 1.6-encoded polypeptide (SEQ ID NO:44) that is described in Example 5, and the portion of the TP 1.6 polypeptide shown in SEQ ID NO:46 matches a portion of Msp 2 (SEQ ID NO:10). It is notable that Msp 2 (SEQ ID NO:10) lacks a variable region, yet vaccine testing with the polypeptide shown in SEQ ID NO:46 provided protective immunity in rabbits, thus indicating that conserved as well as variable 30 region epitopes of Msp proteins are useful in vaccine compositions. The invention is further explained by reference to the following examples. Example 1. Production of Antisera Immune rabbit serum (IRS): WO 99/53099 PCT/US99/07886 -21 For IRS, antiserum was prepared from rabbits that had been injected with live infectious T. p. pallidum. Sera were collected at various times following infection, and were pooled. Adsorbed opsonic antiserum (ORS): 5 Two rabbits infected with T. p. pallidum (Nichols strain) for three months were boosted intradermally and intraperitoneally with 2 x 108 T. pallidum one month prior to blood collection. Sera from the two animals were pooled and shown to have opsonic activity. The antiserum was sequentially adsorbed with the following antigens that do not induce opsonizing antibodies or have been shown not to elicit 10 immune protection against syphilis: T. phagedenis, biotype Reiter (Lukehart, S.A., et al., J Immunol., 129:833-838, 1982), recombinant 47, 37, 34.5, 33, 30, 17 and 15 kDa molecules (Norris et al., Electrophoresis, 8:77-92, 1987) expressed as maltose binding protein-fusion peptides in the pMAL system (New England Biolabs, Beverly, MA) and recombinant TROMP 1 (Blanco et al., J Bacteriol., 177:3556-3562, 1995) 15 expressed as a glutathione-S-transferase-fusion peptide (Pharmacia, Piscataway, NJ). The antiserum was further adsorbed with Venereal Disease Research Laboratory (VDRL) antigen, a lipid complex that has been shown to be the target of some opsonic antibodies (Baker-Zander et al., J InfecT. Dis., 167:1100-1105, 1993). These adsorption steps were performed to reduce the number of irrelevant positive 20 clones identified by this antiserum in the expression library screening. Adsorption was repeated until no antibody reactivity against the adsorbents could be demonstrated by immunofluorescence (Reiter treponeme), immunoblot analysis (recombinant antigens) or serological testing (VDRL). The final antiserum retained significant opsonic activity as measured by our rabbit macrophage phagocytosis assay (Lukehart and 25 Miller, J Immunol., 121:2014-2024, 1978). This absorbed antiserum is hereafter termed "opsonic antiserum," or "ORS." Non-opsonic antiserum (NORS): Non-opsonic antiserum was prepared by immunization of a seronegative rabbit with 6 x 10' T. p. pallidum, Nichols strain, that had been heated at 63 0 C for 1 h, 30 followed by two boosts of 2-8 x 10 7 heat-killed organisms. All immunizations were performed using incomplete Freund's adjuvanT. The resulting antiserum was weakly reactive in the VDRL test, 4+ reactive at 1:1000 dilution in the FTA-ABS test, and non-opsonic in the phagocytosis assay. This antiserum is hereafter termed "non opsonic antiserum," or "NORS." WO 99/53099 PCT/US99/07886 -22 Anti-E. coli antibodies present in the opsonic and non-opsonic antisera were removed using standard techniques (Sambrook et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N. Y., 1989, which is hereby incorporated by reference in its entirety). Briefly, eight 5 nitrocellulose filters were incubated with an E. coli lysate prepared from 50 ml of OD 1.0 bacteria, then air dried. Following blocking of non-specific sites, four of the E. coli lysate-impregnated filters were incubated with the antiserum. A T.p. pallidum lysate was subjected to SDS-PAGE and the separated proteins were tested by immunoblot analysis for reactivity with the T. pallidum 10 specific ORS. Total T. pallidum lysate was separated by SDS-PAGE, immunoblotted onto nitrocellulose, and exposed to ORS that had not yet been adsorbed, to post adsorption ORS, or to NORS. Results of these analyses indicated that fifteen molecules with approximate molecular masses of 70, 68, 60, 55, 45, 43, 41, 39, 38, 35, 33, 32, 31, 29 and 13 kDa reacted with the adsorbed ORS. Of these fifteen, those 15 with approximate sizes of 68, 43, 41, 39, 38, 35, 31 and 29 kDa exhibited minimal immunoreactivity with the non-opsonic antiserum, thus seemed likely to encode proteins exposed on the surface of T. pallidum. Example 2. Construction and Screening with ORS of a T. pallidum Expression Library 20 Rabbit macrophages have been shown to efficiently phagocytize T. p. pallidum in vitro using antiserum from T. p. pallidum-infected rabbits, i.e., IRS as a source of opsonizing antibody (Lukehart and Miller, J. Immunol., 121:2014 2024, 1978; Baker-Zander and Lukehart, J.1 InfecT. Dis., 165:69-74, 1992). In contrast, antiserum from rabbits immunized with heat-killed T. p. pallidum fails to 25 opsonize. In addition to its opsonic potential, IRS has been shown to block T. p. pallidum adherence to host cells (Fitzgerald et al., InfecT Immun., 18:467-478, 1975; Fitzgerald etal., InfecT. Immun.,11:1133-1145, 1975; Hayes etal., InfecT Immun., 17:174-186, 1977; Wong et al., Br. J. Vener. Dis., 59:220-224, 1983) and to provide partial protection against T. p. pallidum infection in passive 30 transfer experiments (Sepetjuan et al., Br. J Vener. Dis., 49:335-337, 1973; Perine et al., InfecT Immun., 8:787-790, 1973; Turner et al., Johns Hopkins Med J., 133:241-251, 1973; Bishop and Miller, J Immunol., 117:191-196, 1976; Weiser et al., InfecT Immun., 13:1402-1407, 1976; Graves and Alden, Br. J Vener. Dis., 55:399-403, 1979; Titus and Weiser, J. InfecT Dis., 140:904-913, 1979). As a WO 99/53099 PCT/US99/07886 -23 result, antigens exhibiting reactivity with IRS may have additional functional roles in cytoadherence and immune protection. Collectively, these observations demonstrate the importance of identifying the target antigens of T. pallidum-specific opsonic antibody. Opsonic antibodies 5 generally recognize bacterial peptidoglycan, lipopolysaccharide, capsular polysaccharides or proteins, and since T. pallidum does not have an accessible peptidoglycan layer nor does it contain either lipopolysaccharide or capsular material, the opsonic targets are likely to be surface-exposed outer membrane proteins. To identify potential opsonic targets, a treponemal genomic expression library 10 was constructed and differentially screened with ORS and NORS that were prepared as described in Example 1. To prepare the library, T. p. pallidum genomic DNA was isolated from approximately 1010 organisms using the QIAamp Tissue Purification Kit (Qiagen, Chatsworth, CA) and a genomic expression library was constructed using the Lambda ZAP® II/EcoRI/CIAP cloning kit (Stratagene, La Jolla, CA) according to 15 the manufacturer's instructions. Briefly, 2 pgg of T. pallidum genomic DNA were partially digested with Tsp509I and DNA fragments in the size range of 0.5 to 4.0 kb were gel-purified using standard techniques (Sambrook et al., 1989). One hundred and forty nanograms of the size-selected Tsp509I-digested DNA preparation were ligated to EcoRI predigested Lambda ZAP II vector arms and the ligated DNA was 20 packaged using the Gigapack II packaging extract (Stratagene). The resulting bacteriophage library had a titer of 4.7 x 106 pfu/ml. E. coli XL-1 Blue (Stratagene, La Jolla, CA) was used as the host strain to plate approximately 50,000 plaques (12,500 pfu/plate) using established methods (Sambrook et al., 1989). The plates were incubated for 5.5 h at 37 0 C, overlaid with 25 10 mM isopropylthiogalactopyranoside (IPTG)-impregnated nitrocellulose filters and incubated for a further 4 h at 37 0 C. Duplicate lifts were prepared by removing the filters and overlaying the plates with fresh IPTG-impregnated filters prior to a second overnight incubation at 37 0 C. Filters were washed in Tris-buffered saline with 0.05% Tween-20 and stored moist at 4oC until the immunoscreening step. 30 Immunoblot analysis was performed as previously described (Baker-Zander et al., J. InfecT. Dis., 151:264-272, 1985). For SDS-PAGE gels, a 10 kDa protein ladder (Gibco BRL, Gaithersburg, MD) was included as a standard. Filters were screened according to the manufacturer's instructions (Stratagene's picoBluea immunoscreening kit). Briefly, blots were blocked with 3% nonfat milk in Tris 35 buffered saline and exposed to a 1:100 dilution of the anti-T. pallidum ORS with the WO 99/53099 PCT/US99/07886 -24 primary plaque lifts and a similar dilution of the NORS with the duplicate plaque lifts. Immunoreactive plaques were detected with 1 ptCi of 1 25 I-labeled protein A/nitrocellulose filter using established methods (Sambrook et al., 1989). Those clones showing reactivity with the opsonic antiserum but no reactivity with the non 5 opsonic antiserum were subjected to secondary screening with both the opsonic and non-opsonic antiserum. Clones consistently showing differential reactivity were screened yet again with the opsonic antiserum. Cloning and sequencing: Immunoreactive plaques were converted to pBluescript SK(-) phagemids by 10 in vivo excision in the E. coli host strains XL-1 Blue and SolR according to the manufacturer's instructions. Both strands of insert DNA were sequenced by a combination of single-stranded and double-stranded DNA sequencing using the Sequenase Version 2.0 and the Applied Biosystems dye terminator sequencing kits and the ABI 373A DNA sequencer according to the manufacturer's instructions. In 15 all cases both universal sequencing primers and internal primers designed from DNA sequences were used. Results of Screening: A Lambda ZAP II T. P. pallidum genomic expression library was constructed and screened in duplicate with the ORS as well as with the NORS. Ten clones were 20 identified that were immunoreactive exclusively with the opsonic antiserum. As discussed in more detail in the examples to follow, nucleotide sequence analysis has been performed for six of these clones. DNA and protein sequence analysis: Ten clones that specifically reacted with ORS were selected for DNA 25 sequence analysis. Of these, four proved to encode the same protein (see Example 3), while one encoded a putative outer membrane protein (see Example 4), and the remaining positive encoded one member of a 12-member gene family (see Example 5). Nucleotide sequences were analyzed using the SeqAppa software (Gilbert, D.G. (1992) SeqAppa, which is published electronically on the Internet, and 30 which is available via anonymous itp from ftp.bio.indiana.edu. IUBio archive of molecular and general biology software and data). Database searches were performed using the basic local alignment search tool (BLAST) algorithm (Altschul et al., J. Mol. Biol., 215:4673-4680, 1990) and either the BLASTN, BLASTX or BLASTP programs. Alignments of the protein sequences encoded by the clones were 35 performed using the Clustal Wa general purpose multiple alignment program WO 99/53099 PCT/US99/07886 -25 (Thompson et al., Nucl. Acids Res., 22:4673-4680, 1994). The percentage of positional identity and similarity between sequences was calculated from the number of identical or similar residues, respectively, between aligned sequences, but insertions and deletions were not scored. The molecular mass and pI of the translated product 5 were calculated using the MacProMassa vl.05 software (Beckman Research Institute, Duarte, CA). The Prositea protein motif database was used to access the signal peptidase I and II cleavage sites. Example 3. T.p. pallidum glycerophosphodiester phosphodiesterase (Gpd): The ten ORS-specific plaques described in Example 2 were subjected to 10 tertiary screening to obtain well-isolated plaques and to verify positivity. Analysis of one of these plaques has been reported previously in Stebeck et al., FEFMSMicrobiol. Letters, 154:303-310, 1997, which is hereby incorporated by reference in its entirety. In vivo excision of the plaque described in Stebeck et al., 1997, produced a pBluescript phagemid containing a 3.5 kb inserT. Nucleotide sequence analysis of 15 the 3.5 kb insert revealed a 1071 bp open reading frame (SEQ ID NO: 1) encoding a 356 amino acid translated product (SEQ ID NO:2). Sequence analysis of three more of the ten positive plaques described in Example 2 revealed nucleotide sequences encoding this same 41 kDa protein. The polypeptide shown in SEQ ID NO:2 has a predicted isoelectric point at pH 9.13 and a predicted molecular mass 20 of 41,014 kDa. Putative -35 (TGCACG) and -10 (TATAA) promoter regions and a ribosome binding site (GAGGAG) were noted in the nucleotide sequence encoding this protein, upstream from the ATG initiation codon. Analysis indicated that the 41 kDa protein of SEQ ID NO:2 contains a two amino acid signal peptide characteristic of previously identified prokaryotic membrane 25 lipoproteins, including an amino-terminal basic residue, a hydrophobic core and a putative Leu-Val-Ala-Gly-Cys signal peptidase II cleavage site (Hayashi and Wu, J Bioenerg. Biomembr., 22:451-471, 1990), strongly indicating that this protein itself is a membrane lipoprotein. Another group of investigators using a different gene isolation approach reported the isolation of a gene encoding this same 356 amino acid 30 protein from T. p. pallidum, but reported that the protein was anchored to the periplasmic leaflet rather than being part of the outer membrane. (Shevchenko et al., InfecT. Immun., 65:4179-4189, 1997). This predicted molecular mass corresponds with that of the 41 kDa immunoreactive protein described in Example 1 that reacts specifically with ORS 35 when this antiserum was used to develop Western blots containing treponeme lysates WO 99/53099 PCT/US99/07886 -26 (see Example 1). It was shown previously that a 41 kDa protein is among those that can be detected in treponeme lysates analyzed on Western blots with serum from human syphilis patients (Baker-Zander et al., 1985). As described in more detail below, antibody directed against the subject recombinant 41 kDa protein also reacts 5 with a 41 kDa protein present in treponeme lysates, thus this new gene may correspond to the same protein detected with human syphilis patient sera. Sequence alignment analyses: Sequence database analysis of the 356 amino acid translated sequence (SEQ ID NO:2) identified glycerophosphodiester phosphodiesterase (Gpd) from a variety of 10 bacterial species as the optimal scoring protein, the closest match being with the Gpd of Haemophilus influenzae. The T. p. pallidum Gpd homologue (SEQ ID NO:2) exhibited about 72.2% sequence similarity with the corresponding H. influenzae protein (Janson et al., InfecT Immun., 59:119-125, 1991; Munson and Sasaki, J Bacteriol, 175:4569-4571, 1993), as well as 70.5% amino acid sequence homology 15 with an E. coli enzyme having the same activity (Tommassen et al., Mol. Gen. Genet., 226:321-327, 1991). Homology was found also but to a lesser degree, with the Gpds from Borrelia hermsii (58.4%; Schwan et al., J Clin. Microbiol., 34:2483 2492, 1996; Shang et al., . Bacteriol., 179:2238-2246, 1997) and Bacillus subtilis (37.4%) (Nilsson et al., Microbiol., 140:723-730, 1994). The 41 kDa T. p. pallidum 20 protein (SEQ ID NO:2) is within the range of masses reported for Gpds from other bacterial species, and closely matches the 40-kDa T. pallidum immunoreactive antigen identified by Shang et al. using rabbit anti-B. hermsii glycerophosphodiester phosphodiesterase antiserum (Shang et al., J Bacteriol., 179:2238-2246, 1997). Taken together, these results indicated that the 356 amino acid translated sequence 25 (SEQ ID NO:2) is a Gpd encoded by T. p. pallidum. Example 4. Identification of a T. pallidum D 15/Oma 87 homologue Another of the immunoreactive lambda clones was subjected to nucleotide sequence analysis, and an open reading frame was found by sequencing the portion of the cloned insert fused with the open reading frame of j3-galactosidase in 30 pBluescripT. The cloned insert was sequenced as described in Example 2, and an open reading frame was identified that gave a 94 kDa protein, whose amino acid sequence is shown in SEQ ID NO:4. A corresponding full length ORF encoding this 94 kDa protein was identified from the T. p. pallidum genome sequence that was released June 24, 1997, by the Institute for Genomic Research (TIGR), although 35 TIGR predicted a different initiating methionine for the D15/Oma87 homologue.

WO 99/53099 PCT/US99/07886 -27 The amino acid sequence predicted from this cloned insert was found to share sequence similarity with the protective surface-exposed outer membrane antigens D 15 of H. influenzae (36.3%) (Flack et al., Gene, 156:97-99, 1995) and Oma87 of Pasteurella multocida (35.7%) (Ruffolo and Alder, Infec. Immun., 64:3161-3167, 5 1996), as well as with outer membrane proteins from Brucella abortus (37.2%, Genbank accession number U51683) and N. gonorrhoeae (35.2%, Genbank accession number U81959). The open reading frame of this clone was subcloned into expression vectors for further analysis. The T. pallidum D15/Oma87 homologue (SEQ ID NO:4) is predicted to have 10 a type I cleavable signal sequence (using rules devised by von Heinje, et al. (Nucleic Acids Res., 14:4683-4690, 1986) and McGeoch, et al. (Virus Res., 3:271-286, 1985). In addition, the protein was shown to have an 85% probability of being an outer membrane protein by the pSORT program which takes into account hydrophobic domains and secondary structure (see http://psort.nibb.ac.jp/). Moreover, the 15 Borrelia burgdorferi homologue of this clone has been identified from the B. burgdorferi genome project (Vugt et al., Nature, 390:580-586, 1997) and has been classified as a probable outer membrane protein. As described in Example 10, this protein has been expressed in E. coli and the recombinant protein used to immunize rabbits. 20 Example 5. Identification of a family of T. pallidum major sheath protein homologue Another of the ORS-reactive clones described in Example 2 was sequenced, and upon analysis the polypeptide it encoded proved to have 41.5% amino acid sequence similarity with the 53 kDa Treponema denticola major outer membrane sheath protein (Msp) (Egli et al., InfecT Immun., 61:1694-99, 1993) and, as 25 discussed further below, with a T. pallidum sequence deposited in Genbank (50.1%; Genbank accession number TPU88957, deposited by Hardham et al., Univ. N. Carolina, and corresponding to TIGR TprK, or Msp9). Fragments of another gene related to the T. denticola Msp gene were identified by a separate approach using representational difference analysis (RDA), a 30 subtractive hybridization technique in which one compares two populations of nucleic acid molecules to obtain clones of genes that are present in one population but not in the other (Lisitsyn et al., Science, 259:947-950, 1993; Lisitsyn et al., Nature Genetics, 6:57-63, 1994). For RDA, the DNA that contains the genes of interest is called the "tester," and the reference DNA is the "driver." In essence, sequences 35 present in the tester DNA but absent from the driver DNA are selectively amplified by WO 99/53099 PCT/US99/07886 -28 using PCR. In a first annealing step, an excess of driver DNA is hybridized with a small amount of tester DNA. Tester sequences common to both populations are thus selectively driven into tester-driver hybrids, while unique tester sequences will form only tester-tester hybrids. The unique tester-tester hybrid molecules are separated 5 from tester-driver hybrids as follows. Prior to the first hybridization step, short adapter oligonucleotides are ligated to the tester DNA. After the tester DNA has been hybridized with the driver DNA, the adapter sequences are annealed with PCR primers that bind to the protruding adapter sequences, and the tester-tester hybrids are thus selectively amplified. 10 For the experiments described below, the organisms used were T. p. pallidum, Nichols strain, and T. paraluiscuniculi, Cuniculi A strain. After being propagated in New Zealand white rabbits, the bacteria were extracted from infected rabbit testes in sterile saline, collected in DNAse/RNAse-free 1.7 ml microfuge tubes, and spun immediately in a microfuge at 12,000 X G for 30 minutes at 4 0 C. Bacterial pellets 15 were resuspended in 200 p. of 1X lysis buffer (10 mM Tris pH 8.0, 0.1 M EDTA, 0.5% SDS), and DNA was extracted using the Qiagen Kit for genomic DNA extraction (Qiagen Inc., Chatsworth, CA) using the manufacturer's instructions. The DNA was treated with RNAse A. RDA was carried out using the CLONTECH PCR Select Subtraction Kit (Clontech, Palo Alto, CA) following the manufacturer's 20 protocol beginning from the section describing the restriction digestion step. For RDA, DNA from T. p. pallidum served as the tester DNA, and a mixture of Treponema paraluiscuniculi (a rabbit pathogen) plus rabbit genomic DNA served as the driver. T. paraluiscuniculi was used as a driver DNA because this relative of T. p. pallidum, unlike its virulent cousin, cannot infect humans. Thus, it was surmised 25 that genes present in T. p. pallidum but absent from T. paraluiscuniculi would be involved in pathogenicity, and would provide likely candidates for vaccine testing. Rabbit genomic DNA was included in the driver to remove any traces of rabbit DNA that co-purified with the bacterial DNA. This same experimental strategy is applicable to the isolation of genes related to pathogenicity in humans from any species or 30 subspecies of Treponema that infects humans but not rabbits. For example, to isolate pathogenicity-related genes from T. p. pertenue or T. p. endemicum using RDA, one would use tester DNA from one or the other of these bacteria and driver DNA from T. paraluiscuniculi. Two separate subtraction libraries were created using the above-described 35 tester and driver DNAs. Briefly, 0.5 pg of T. p. pallidum genomic DNA (tester WO 99/53099 PCT/US99/07886 -29 DNA) and a pool of 3 pg of T. paraluiscuniculi DNA plus 3 gg of rabbit liver DNA (driver DNA) were digested to completion with Rsa I. The digestion products were purified by the phenol/chloroform/isoamyl alcohol method and the digested tester was then divided into 2 aliquots (tester-1 and tester-2) and each was ligated to one of two 5 adapters that were to serve as binding sites for PCR primers: Adapter 1. 5'-TAATACGACTCACTATAGGGCTCGAGCGGCCGCCCGGG CAGGT-3' (SEQ ID NO:62) Adapter 2. 5'-GTAATACGACTCACTATAGGGCAGCGTGGTCGCGGCCG AGGT-3' (SEQ ID NO:63) 10 These adapters are sufficiently long to accommodate binding with two different sets of primers to permit "nested PCR" as described below. No adapters were ligated to the driver DNA. For the first hybridization, two aliquots of tester DNA (tester-1 and tester-2) were heat denatured in separate reaction tubes in the presence of an excess of driver 15 and allowed briefly to reanneal. During this time, low abundance DNA fragments that are unique to the tester remained as single-stranded DNA, and common DNA fragments annealed with the driver to form double stranded DNA. For the second hybridization step, both of the first hybridization mixtures were pooled and hybridized again with additional excess denatured driver DNA. This 20 second hybridization step permitted further removal of common sequences, and permitted the single-stranded DNA fragments unique to the tester populations to form hybrids with one another, these latter hybrids including tester-tester duplexes having different adaptors at each end, i.e., tester-1/tester-2 duplexes. At this stage, the adapter sequences were single-stranded, forming overhangs at each end of the duplex 25 molecules. These overhangs were filled in with DNA polymerase, yielding unique double-stranded molecules having different primer binding sites on their 5' and 3' ends adaptor sequences. Primary PCR was then used to amplify these unique tester-tester hybrids, using a PCR primer No. 1, which binds to both adaptors 1 and 2, followed by a nested PCR (nested primer 1, 5'-TCGAGCGGCCGCCCGGGCAGGT (SEQ ID 30 NO:64), and nested primer 2, 5'--AGCGTGGTCGCGGCCGAGGT (SEQ ID NO:65)), to further enrich unique sequences, to reduce the background, and to increase the specificity of the amplification. Secondary PCR products were then cloned directly into the PCR 3.1 T/A cloning vector (Invitrogen, Sorrento, CA), and the cloned inserts subjected to DNA sequence analysis.

WO 99/53099 PCT/US99/07886 -30 For sequencing, single colonies were selected and plasmid DNA was digested with Eco RI to identify the clones containing inserts. Double-stranded plasmid DNA was extracted with the Qiagen Plasmid Kit (Qiagen, Chatsworth, CA), and 500 ng of each DNA was used for fully automated sequencing by the dye terminator method 5 (Perkin Elmer, Foster City, CA) according to the manufacturer's instructions but with the addition of 1 pl of molecular grade dimethylsulfoxide (Sigma, ST. Louis, MO) per reaction, giving a final concentration of 5% vol/vol. Cloned DNAs were sequenced in both directions using the T7 and reverse sequencing primers homologous to plasmid regions flanking the cloned inserts. The cloned inserts were found to range in size 10 from 100 bp to 500 bp. Two clones of particular interest were obtained, clones 3 and 33, each of which was isolated from an independently constructed subtraction library made as described above. The sequences obtained from clones 3 and 33 were used to do Blast searches in the nucleotide and protein databases. No significant homologies were found at the 15 nucleotide sequence level, but the predicted amino acid sequences encoded by both clones indicated that these polypeptides were related to the Msp protein of T. denticola, an oral treponeme associated with periodontal disease (Genbank accession No. U29399). Alignment using the Clustal W program indicated that the inserts of clones 3 and 33 aligned, respectively, with regions near the amino and 20 carboxyl ends of the T. denticola Msp protein. These clones were subsequently used as described below for hybridization with Southern blots of the T. p. pallidum genomic DNA, and to design oligonucleotides for PCR amplification of longer pieces of the T.p. pallidum Msp homologue from which they appeared to be derived. To determine the specificity of the cloned unique sequences for T. p. pallidum, 25 as well as their hybridization patterns to digested genomic DNA, approximately 3 pg each of T. p. pallidum (Nichols strain) and rabbit DNA were digested with Eco RI, Pst I, and Bam HI, then separated in 1% TBE agarose gels, denatured with 0.5 M NaOH and transferred to Hybond N membrane (Amersham Laboratories, Arlington Heights, IL). The inserts of clones 3 and 33 were labeled as follows to use as 30 hybridization probes. The inserts were PCR amplified from the cloning vectors using the nested primers described above under the same conditions as for the nested PCR during the subtraction experiments, and purified using the Qiaquick PCR Purification Kit (Qiagen, Chattsworth, CA). Fifty ng of the purified amplicons were then labeled by random priming with a- 32 P using the Random Priming labeling Kit (Boehringer 35 Manheim, Indianapolis, IN) according to the manufacturer's protocol.

WO 99/53099 PCT/US99/07886 -31 The labeled inserts of clones 3 and 33 were hybridized under high stringency conditions to the above-described Southern blots. Each probe was allowed to bind the PCR products on a separate filter for 12 hours at 37 0 C in hybridization solution (50% formamide, 5X SSC, 50 mM NaPO 4 , 1% SDS, 5X Denhardt's solution). The 5 blots were then subjected to stringent washes at 65 0 C in buffers containing 2X SSPE, 0.1% SDS, and 0.2X SSPE, 0.1% SDS, for 20 minutes each (SSPE: 150 mM NaC1, 10 mM NaPO 4 , 1 mM NaEDTA, p 7.4). Hybridization was detected by autoradiography. No hybridization of these probes with rabbit DNA was observed, indicating that the probes were specific for T. p. pallidum. The results of these 10 Southern blots disclosed several hybridizing DNA fragments, thus suggesting that the cloned genes belonged to a multigene family. The Eco RI digests yielded bands of about 8 and 5 kb, the Pst I digests bands of about 1 kb, 800 bp, and 500 bp, and the Bam HI digests bands of about 8, 5 and 3 kb. Isolation of TP 1.6 (SEQ ID NO:43): 15 As explained above, the inserts of clones 3 and 33 were homologous, respectively, to the 5' and 3' ends of the Msp gene of T. denticola, thus primers were designed to amplify that portion of the T. p. pallidum Msp homologue that presumably lay between the two clones. Primers used were the S-3 sense primer corresponding to the 5' end of the insert of clone 3 and having the sequence 20 5'-ACCAGTCCTTCCTGTGTGGTTAA (SEQ ID NO:66), and the antisense primer As-33, corresponding to the 3' end of the insert of clone 33, and having the sequence 5'-ACTCCTTGGTTAGATAGGTAGCTC (SEQ ID NO:67). A hot start PCR amplification was performed as described above using as templates approximately 1 pg of genomic DNA of T. p. pallidum, Nichols strain. The DNA was amplified in a 25 total volume of 100 gl per tube, each containing 200 gM dNTPs, 50 mM TRIS-HCI (pH 9.0 at 2000 C), 200 mM ammonium sulfate, 1 pM each primer and 2.5 units of Taq polymerase (Promega, Madison, WI). MgCl 2 beads (Invitrogen, San Diego California) were added giving a final MgCl 2 concentration of 1.5 mM. The following cycling conditions were used: an initial step of 4 minutes denaturation at 94oC 30 followed by 40 cycles at 94 0 C for 1 minute, 65 0 C for 2 minutes, 72 0 C for 1 minute, and a final elongation step of 10 minutes at 72 0 C. The PCR products were then kept at 4°C and directly cloned into T/A cloning vectors for sequencing and for further analysis on agarose gels. The PCR was repeated several times, and each time yielded one band that proved to contain 1687 bp (TP 1.6)(SEQ ID NO:43).

WO 99/53099 PCT/US99/07886 -32 The sequence of TP 1.6 (SEQ ID NO:43) was found later to have high homology with a newly released T.p. pallidum Msp-like sequence in Genbank (TPU88957), and to at least 10 different ORFs that were present in the initial release on June 24, 1997 of the TIGR T. p. pallidum genome project (posted at 5 http://med.uth.tmc.edu/Treponema/tpall.html/). When first posted on the Internet in June, 1997, the TIGR T. p. pallidum sequence was not annotated, i.e., the locations of open reading frames were not indicated. The August 18, 1997 update was annotated, but not until the January 1, 1998 update were all 12 Msp family members (Tprgenes by TIGR Terminology) identified according to their coordinates. It should 10 be noted that all versions of the T. p. pallidum genome posted at the TIGR site are regarded as preliminary in nature and may contain misassembled genes, mutations and frameshifts, particularly within the Msp family. Nonetheless, comparisons were conducted to determine whether the posted sequences contained any sequences similar or identical to the nucleotide sequence of TP 1.6 (SEQ ID NO:43). A search 15 located an open reading frame in the posted T. p. pallidum sequence at positions 73,979 - 75,665 (based on the August 18, 1997 version) that is 90.21% identical to the sequence of TP 1.6 (SEQ ID NO:43). The aligned sequences contained 55 amino acid mismatches spread throughout the 5' end from amino acid positions 1 through 123. Beyond this point to the 3' end, the identity of both amino 20 acid sequences is 100%. The Msp genes are arranged into five regions on the T. p. pallidum chromosome. There are three major subfamilies of Msps as defined by homology of their predicted amino acid sequences. Subfamily I includes Msps 2 (SEQ ID NO:9), 4 (SEQ ID NO:13), 5 (SEQ ID NO:15), and 11 (SEQ ID NO:28), which are highly 25 homologous to one another at their 5' and 3' termini. Msps 4 (SEQ ID NO: 13) and 5 (SEQ ID NO:15) and 11 (SEQ ID NO:28) have central variable regions of about 600 bp, while Msp 2 (SEQ ID NO:9) lacks any variable region. Msp 4 (SEQ ID NO:13) and 5 (SEQ ID NO:15) are identical. Subfamily II includes Msps 1 (SEQ ID NO:7), 3 (SEQ ID NO:11) and 10 (SEQ ID NO:26), and has larger variable 30 regions of about 1000 bp. This subfamily shares significant homology at the 5' and 3' ends with the Subfamily I. Subfamily III includes Msps 6 (SEQ ID NO: 17), 7 (SEQ ID NO:19), 8 (SEQ ID NO:22), 9 (SEQ ID NO:24) and 12 (SEQ ID NO:30), all of whose sequences are comparatively distinct from the two other groups and from one another. Msp 7 (SEQ ID NO:19) appears to have a premature termination, in that at 35 the termination of ORF A (SEQ ID NO:20), in another reading frame, there is another WO 99/53099 PCT/US99/07886 -33 ORF encoding another 368 amino acids (ORF B (SEQ ID NO:21)) that is homologous to the other Msps. The TP 1.6 sequence (SEQ ID NO:43) was found by comparison to the TIGR Tpr sequences to be a hybrid gene. The amino terminus, i.e., the first 152 amino 5 acids, of the TP 1.6 polypeptide (SEQ ID NO:44) matches the amino terminus of Msp 2 (SEQ ID NO:9), and differs in only two amino acids from the amino terminus of Msp 4 (SEQ ID NO:13) and 5 (SEQ ID NO:15), while the 410 amino acids at the carboxyl terminus of TP 1.6 (SEQ ID NO:43) match the corresponding portion of Msp 1 (SEQ ID NO:7). The significance of this finding is not presently known. 10 One Msp gene is predominantly transcribed by T. p. pallidum Nichols strain: T. p. pallidum Nichols that was isolated on days 5, 7, and 15 after infection transcribes predominantly Msp 9 (SEQ ID NO:24) mRNA, as determined by reverse transcriptase PCR (RT-PCR), a procedure that amplifies cDNA synthesized from total RNA, including mRNA, found in the bacteria, thus reflecting transcribed genes. To 15 perform RT-PCR, a group of oligonucleotide primers were prepared that are specific to the variable regions of Msps 1 (SEQ ID NO:7), 3 (SEQ ID NO: 11), 4 (SEQ ID NO:13), 5 (SEQ ID NO:15), 6 (SEQ ID NO:17), 7 (SEQ ID NO:19), 8 (SEQ ID NO:22), 9 (SEQ ID NO:24), 10 (SEQ ID NO:26), 11 (SEQ ID NO:28), and 12 (SEQ ID NO:30)(see Table 1), thus providing specific amplification of transcripts of those 20 Msps. Using T. p. pallidum RNA extracted from infected rabbit testes, RT-PCR analysis of the Msp transcription pattern was conducted beginning at day 5 after infection. At day 5, a strong signal for Msp 9 (SEQ ID NO:24) was evident with a weak signal for Msps 6 (SEQ ID NO:17) and 11 (SEQ ID NO:28). Transcripts from Msps 1 (SEQ ID NO:7) or 12 (SEQ ID NO:30) mRNA were detected, but signals 25 were weak and variable. After 5 more PCR cycles, signal was discernible for all the Msps, indicating that transcripts from all of them were present, but at relatively low levels. The preponderance of Msp 9 (SEQ ID NO:24) product was not due to an overly efficient Msp 9 (SEQ ID NO:24) PCR, because when these same primers were used to amplify T. p. pallidum genome DNA, it was found that the primers for Msp 9 30 (SEQ ID NO:24) were less efficient than the primers for Msp 7 (SEQ ID NO:19) or 4 (SEQ ID NO: 13) or 5 (SEQ ID NO: 15). Moreover, the PCR products obtained from the RT-PCR RNA likely reflected mRNA and not contaminating T. p. pallidum genome DNA because the RNA preparation was extensively pre-treated with DNAse before the cDNA synthesis step. Furthermore, omitting reverse transcriptase from the 35 reactions led to no producT.

WO 99/53099 PCT/US99/07886 -34 The most likely explanation for these results is that a majority of the treponemes express Msp9 (SEQ ID NO:24), and that a minority of them express Msps 1 (SEQ ID NO:7), 6 (SEQ ID NO:17), 11 (SEQ ID NO:28), or 12 (SEQ ID NO:30). Alternatively, it may be the case that each individual treponeme cell 5 expresses high amounts of Msp 9 mRNA and lower quantities of Msps 1 (SEQ ID NO:7), 6 (SEQ ID NO: 17), 11 (SEQ ID NO:28), and 12 (SEQ ID NO:30). Other strains of T. p. pallidum have been similarly analyzed by RT-PCR, and proved to express other Msp preferentially, i.e., the pattern of expression appears to be strain-specific. 10 Identification of an Msp homologue in Treponemapallidum pertenue: The primers described above for amplification of TP 1.6 (SEQ ID NO:43) were used to amplify a fragment of DNA from the closely related treponeme, T. p. pertenue, the etiologic agent of yaws. A hot start PCR amplification was performed as described above using as templates approximately 1 pg of genomic 15 DNA of T. p. pallidum, Nichols strain, and T. p. pertenue, Gauthier strain, using the same cycling conditions described in Example 5 for these primers. The PCR products were then kept at 4°C and directly cloned into T/A cloning vectors for sequencing and for further analysis on agarose gels. PCR amplification with these primers reproducibly yielded the expected 1687 bp band using T. p. pallidum DNA, and for 20 the T. p. pertenue DNA, a band of 1705 bp, as well as smaller bands of 1291 bp. When attempts were made to amplify the DNA of T. p. endemicum with this same primer pair, no DNA fragment was amplified. Analysis of the T. p. pallidum and T. p. pertenue Msp Homologues: Sequencing was done by the primer walking approach, using the T7, PCR 3.1 25 reverse, the INT-S, 5'-GGCTTCCGCTTCTCCTTCG (SEQ ID NO:68), and the INT-As, 5'-GTTTCGAGCTTAAGGAATCC (SEQ ID NO:69). The following clones were sequenced: T. p. pallidum, clones 1,2,4,5,7, and T. p. pertenue, clones 6 and 16 of the larger amplicons (-1.7 kb), and clones 2, 3, 5, 7, and 8 of the shorter amplicons (-1.3 kb). 30 Automated sequencing of the 1.6 kb amplicon of T. p. pallidum and of the 1.7 kb and 1.3 kb amplicons of T. p. pertenue revealed four different copies in T. p. pertenue, one 1.6 kb (clones 6 and 16) and three 1.3 kb homologues (homologue 13Ty 238, Ty5, and Ty7; from clones 2/3/8, 5 and 7, respectively), and a single DNA sequence in T. p. pallidum (clones 1, 2, 4, 5, and 7) among the 19 clones 35 of T.p. pertenue and the five from T. p. pallidum that were examined. The WO 99/53099 PCT/US99/07886 -35 T. p. pallidum DNA fragment (TP 1.6)(SEQ ID NO:43) has 1687 bp, thus predicting a peptide sequence of 562 amino acids (frame +1). The long homologue of T. p. pertenue (17Ty) had a DNA sequence of 1705 bp, and encodes a putative polypeptide of 568 amino acids (SEQ ID NO:35). 5 The shorter amplicons (13Ty 238, 13Ty5, and 13Ty7) all were 1291 bp long, and predicted polypeptides having the same length, 438 amino acids, but differing at their carboxyl termini (SEQ ID NOS:37, 39 and 41). When the deduced peptide sequences of amplicons identified in both subspecies were aligned, i.e., TP 1.6, 17 Ty and 13 Ty, it was found that the T p. pertenue Msp homologue, like those of T. p. pallidum, 10 have highly conserved regions located at the amino and carboxyl terminal ends, separated by a central variable region. For the three 438 amino acid polypeptides, the amino terminal conserved regions extend from amino acid positions 1 through 153, the carboxyl terminal conserved regions from positions 444 through 592, and the internal variable region from positions 154 through 443. As compared with the 15 polypeptide encoded by 17 Ty, the central variable portions of the 438 amino acid polypeptides lack the 161 amino acids present at positions 241 through 400 of the 17 Ty polypeptide. When the peptides encoded by the three 1.3 kb short fragments of T. p. pertenue were compared, it was found that they are highly conserved in almost 20 their entire length, except at their 3' regions where sequence variation was found in a short region from amino acids 354 through 381. Based on the differences in their Msp regions, it is clear that PCR using the above described primer pair can differentiate the treponemes responsible for syphilis, yaws, and bejel, as the results for the three treponemes yield DNA fragments that 25 differ in size and number, and of course, nucleotide sequence. As shown in Example 7, this approach has been extended to develop an RFLP-based method for the differentiation of other strains and subspecies of Treponema. The various subspecies of T. pallidum, including the etiologic agents of human syphilis, yaws, and bejel, possess very small and highly-related genomes, yet all are 30 able to produce lifelong infection in untreated patients. The past inability to differentiate subspecies and strains of T. p. pallidum using serologic methods has led some investigators to hypothesize that these pathogens actually are identical, with only environmental factors dictating different clinical manifestations (Hudson, E.H., Treponematosis Perspectives Bull., WHO 32:735-748, 1965). However, this view is 35 contraindicated, e.g., by differences in the pathogenesis of the infections, and by the WO 99/53099 PCT/US99/07886 -36 co-existence of more than one distinct treponemal disease in some locales. Moreover, there is experimental evidence for antigenic heterogeneity between subspecies and strains. More specifically, this heterogeneity must lie in the "protective" antigen or antigens, since hosts infected with one of the strains is only partially resistant to the 5 other strains. To date, the molecular bases for differences in pathogenesis and immunity have not been identified. The present findings provide an additional means of differentiating the strains of T. pallidum responsible for syphilis, yaws and bejel, and moreover may be directly relevant to the antigenic variations that are responsible for the differences in pathogenicity among these treponemes. 10 Transmembrane Topology Analysis: Like the protein encoded by the T. p. pallidum TP 1.6 (SEQ ID NO:44), the proteins predicted from all four of the T. p. pertenue DNA fragments described above were found to have significant homology to the Msp protein of T. denticola. The T. p. pallidum and T. p. pertenue peptide sequences were analyzed for indications of 15 transmembrane topology using the TmPred program (Hofmnan and Stoffel, A Database of Membrane Spanning Protein Segments, Biochem. Hoppe-Seylor 348, 166). For T. p. pallidum, results indicated three possible amphipathic transmembrane helices at amino acid positions 46-65, 389-409, and 415-438. For T.p.pertenue, three transmembrane helices have were determined for the translate of the 1.7 kb 20 homologue at similar positions, i.e., at amino acids 46-65, 394-412, and 421-444, and two transmembrane regions were found for the short 1.3 kb copies of T. p. pertenue at amino acid numbers 46-65, and 291-314. The T. p. pallidum Msp homologue described by another laboratory (GenBank accession number U88957) was similarly analyzed to determine whether it 25 also has a predicted transmembrane topology to the sequences disclosed here. The three transmembrane regions in the proteins encoded by the 1.6 kb clone of T. p. pallidum and the 1.7 kb clone of T. p. pertenue, and the three in the 1.3 kb homologues of T. p. pertenue were found to overlap extensively with the corresponding predicted transmembrane regions of the GenBank Msp homologue. 30 Interestingly, the differences found between the syphilis and yaws Msp proteins are located in the variable, middle portion of the protein, which is relatively hydrophilic, and thus may be exposed to the extracellular environmenT. The pathogenic treponemes are classified based upon the distinct clinical infections they produce, as well as their host specificity and very limited genetic 35 studies. The syphilis and the yaws treponemes have been classified as subspecies of WO 99/53099 PCT/US99/07886 -37 T. pallidum based upon saturation reassociation assays, methods of low sensitivity to detect small differences. All attempts to show species or subspecies-specific signatures had failed until it was recently shown that these two organisms differ in the 5' and 3' untranslated regions of their 15 kDa lipoprotein genes. However, 5 the 15 kDa lipoprotein gene is neither a protective antigen or a molecule related to differential pathogenesis because the open reading frame is identical in T. p. pallidum and in T. p. pertenue. Furthermore, immunization of rabbits with recombinant 15 kDa lipoprotein has failed to provide any evidence of protection against virulent challenge. The above-described studies on a novel Msp gene family in the Genus 10 Treponema describes for the first time extensive differences in the coding regions of putative outer membrane antigens in two subspecies of T. pallidum. These differences can serve as the basis for the diagnostic differentiation, e.g., using PCR, for determining whether one of these two treponemes, or the treponeme responsible for bejel, is present in a primary lesion. 15 Attachment and invasion are the first steps for a successful treponemal infection, as in vitro studies have shown that T. p. pallidum penetrates eucaryotic cells and localizes to the cytoplasm (J.A. Sykes, et al., Br. J. Vener. Dis., 50:40-44, 1974). The molecules involved in attachment and invasion of eukaryotic cells have not yet been identified, but outer surface proteins are likely to be involved. In 20 T. denticola, an oral spirochete associated with periodontal disease, the Msp antigen has been shown to be involved in cell adhesion, and has porin and extracellular matrix binding activities (Egli et al., InfecT. Immun., 61:1694-9, 1993; Fenno et al., J. Bacteriol., 178:2489-97, 1996). The transmembrane topology analyses (see above) have indicated that there are three overlapping amphipathic regions in the 1.6 kb 25 sequences of

T

. p. pallidum and T.p. pertenue and two in the 1.3 kb fragments of T. p. pertenue, leaving in both cases a large, intermediate hydrophilic segment that includes part of the conserved region and the whole internal variable region. These analyses suggest that the Msp homologous proteins ofT. p. pallidum identified in this study, as well as the other members of the T. p. pallidum Msp family, probably are 30 membrane-spanning molecules located in the outer sheath, making them likely candidates for cell attachment and invasion, as demonstrated for the Msp of T. denticola, and suggesting that they are useful as vaccine candidates. It should be noted that no Msp homologue completely identical to the one described here is present in the current version of the Internet-posted T. pallidum 35 genome sequence, with the best match being the Msp 1 homologue (SEQ ID NO:7), WO 99/53099 PCT/US99/07886 -38 which is only 90.21% identical. As compared with the posted Msp 1, the ORF of TP 1.6 (SEQ ID NO:43) has mismatches throughout the 5' end from amino acid position 1 until amino acid 123 and is completely identical in the rest of the sequence. The present finding that PP 1.6 is a "hybrid" of two of the posted Msp genes, i.e., 5 Msp 1 and 2, may indicate that homologous recombination may be occurring between two homologues so that the 5' region corresponds to one gene in which the downstream portion has been replaced by the corresponding piece of another gene, creating a hybrid molecule with different antigenic characteristics. Mechanisms of this type have been described in Borrelia (e.g., Zhang et al., Cell, 89:275-85, 1997). 10 Alternatively, the Msp genes in the current version of the T. pallidum genome may simply be misassembled, or the results described here may have resulted from copying errors during the PCR amplification. Example 6. PCR Amplification ofMsp Homologues in Various Treponemes These same PCR primers used originally to amplify TP 1.6 (SEQ ID NO:43) 15 were tested also with DNA from several other species and subspecies of the Treponema genus, including genomic DNA from T. pallidum subspecies endemicum, Bosnia A strain, T. paraluiscuniculi, Cuniculi A strain, and a Treponema sp. Simian strain. As a control, aliquots of the DNAs were amplified using primers specific for a 15 kDa lipoprotein gene common to all treponemal species. Results with these 20 control primers yielded bands for all the DNA templates, thus indicating that sufficient amounts of DNA for PCR were present in all of the DNA preparations. Using the TP 1.6 primers, no amplification was seen for T. p. endemicum (which causes bejel), Bosnia A strain, or for T. paraluiscuniculi, Cuniculi A strain. Treponema sp, Simian strain, which is capable of infecting humans, yielded two bands of the same sizes as 25 those noted previously when these primers were used to amplify T. p. pertenue DNA, i.e., 1.6 kb, 1.3 kb. Thus, this group of pathogens can be distinguished using PCR with this primer pair. Example 7. RFLP Strain Differentiation of T. p. pallidum Infection of rabbits with one strain of T. p. pallidum is completely protective 30 against homologous strain challenge, but only partially protective against heterologous strain challenge (Egli et al., InfecT Immun., 61:1694-1699, 1993). This may be because treponemal surface proteins vary from strain to strain, possibly due to variation in Msps. Strain variation in the Msp region was investigated by comparing the Msp variable regions from 18 different clinical isolates of T. p. pallidum, which 35 were isolated from different geographical locations and at different times. These were WO 99/53099 PCT/US99/07886 -39 as follows: Bal 73-1; Bal-2; Bal-3; Bal-5; Bal-6; Bal-7; Bal-8; Bal-9; Chicago; Mexico A; Nichols; Sea 81-1; Sea 81-2; Sea 81-3; Sea 81-4; Sea 81-8; Sea 83-1; Sea 83-2; Sea 84-2; Sea 85-1; Sea 86-1; Sea 86-2; Sea 87-1; Sea 87-2; Street 14; Yobs. 5 The alignment of the amino acid sequences for Msps 1, 3, 4, 5, 10, and 11 indicated a middle region of high heterogeneity flanked by conserved regions. Within these conserved regions are short stretches of identity in all of these Msp alleles. The short highly conserved stretches of sequence were used to design the following primers for PCR amplification of the variable regions of these 6 Msps: sense, 5' 10 CGACTCACCCTCGAACCA (SEQ ID NO:48); antisense, 5' GGTGAGCAGGTGGGTGTAG (SEQ ID NO:49) (corresponding to Set 1 in Table 1). The 18 strains of T. p. pallidum were propagated and their DNA extracted. PCR was performed using a 100 p.l reaction containing 200 pM dNTPs, 50 mM TRIS-HC1 (pH 9.0 at 20 0 C), 1.5 mM MgCl 2 , 200 mM NHI4SO 4 , 1 pM of each primer, and 15 2.5 units of Taq polymerase (Promega, Madison, WI). The cycling conditions were as follows: denaturation at 94oC for 3 minutes, then 40 cycles of 94 0 C for 1 minute, 60 0 C for 1 minute and 72 0 C for 1 minute. Amplicons were purified away from primer-dimers using the QuiaQuick Kit extraction (Qiagen Inc., Chatsworth, CA), and the purified DNAs were quantitated by 20 spectrophotometry. Restriction digests of amplicons were performed with 10 pg of purified PCR product from each treponemal strain, according to the manufacturer's instructions (New England Biolabs, Beverly, MA), using the following 13 restriction endonucleases, all of which recognize four base cleavage sites: BstUI, AluI, Tsp509I, MseI, NheI, Taq*I, HhaI, Nlaml, BfaI, RsaI, MspI, MboI, and AciI. The resulting 25 DNA fragments were separated by electrophoresis in 2.5% TBE/ethidium bromide NuSieve agarose gels. PCR amplification was optimized so that no smearing of bands was detected on the gels. For all T. p. pallidum strains tested, these primers gave bands at about 650 bp and 1.1 kb and about 1 kb. After cleavage with the above-listed restriction enzymes, 30 it was apparent that the 650 bp and 1.0 kb bands actually were quite heterogeneous. The restriction digestion patterns could be divided into 15 distinct "RFLP" patterns. This degree of polymorphism is remarkable in an organism with a small genome of only 1.2 MB. Each enzyme identified a different number of restriction patterns in the 18 T. p. pallidum strains. Msp I and Nhe I each recognized three groups of 35 organisms that gave the same RFLP pattern for that enzyme. Mbo I, Rsa I, and Bfa I, WO 99/53099 PCT/US99/07886 -40 each recognized four groups, Taq a I, five; Hha I, Tsp 509 I, BstU I, six; and Alu I and NLA HI, seven groups. Combining the data from these enzyme digests permitted the division of the 18 strains into 15 distinguishable groups, based upon RFLP differences. Further analysis of the restriction patterns of the T. p. pallidum strains 5 showed that digestion with only four individual enzymes, BstUI, Alul, HhaI, and NlaIII, was sufficient to differentiate the 15 groups. Using these four enzymes, three of the groups were especially easily differentiated from the other strains. These three groups each contain strains that have the same RFLP patterns with these four enzymes. Group I comprises the strains 10 Bal 9, Sea 81-8, and Sea 84-2; group II, Nichols and Yobs strains, and group Il includes the Bal 2 and Bal 8 strains. The strains in each subgroup do not represent unique geographical areas, year of isolation or tissue tropism. Unlike the isolates of these three subgroups, the other 11 T. p. pallidum strains tested showed distinct, specific patterns. Some strains, such as Sea 81-1 and 81-3, were collected in the same 15 city, year, and from the same site in the body, yet showed very different RFLP profiles. Although three groups were identified with at least two strains each, overall, these results indicate that there is a very high degree of heterogeneity in the variable regions of these Msp homologues of these bacterial isolates. In summary, the RFLP patterns demonstrate that there is marked 20 heterogeneity in the variable regions of the different strains of T. p. pallidum. Table II shows the distribution the variability of Msp variable domains amongst the different strains and restriction enzymes tested to date. One of the strains appeared identical to T. p. pallidum Nichols strain, but the other 16 differed from Nichols in their variable domains. Thus, these results demonstrate that the variable domains 25 differ in different strains of T. p. pallidum and this may be the basis for the lack of complete protection of infected animals after heterologous strain challenge. Accordingly, a fully effective vaccine may require a combination of several or all of the Msp proteins. In other experiments, it was found that the above-described PCR primers used 30 for RFLP analysis of T. p. pallidum strains also primed the amplification of Msp genes in T. p. pertenue and in T. p. endemicum, in each case yielding bands of 600 bp, 630 bp, 600 and 1.1 kb. For RFLP analysis, these amplicons were digested with Mbo 1, Rsa 1, Hae HI, Alu 1, Nla III, Hha 1, Msp 1, Taq 1(oa), and Tsp 509. The resulting DNA fragment patterns permitted these two subspecies of T. pallidum to be 35 easily distinguished from one another and from T. p. pallidum.

WO 99/53099 PCT/US99/07886 -41 Moreover, the primer pair used to amplify the DNA fragments for these RFLP analyses, i.e., Set 1 from Table 1, appears to be useful for identifying Msps from many or perhaps all species of Treponema, including pathogens associated with gingivitis and periodontitis. For example, when this primer pair was used with DNA from 5 Treponema denticola (an oral pathogen not reactive with antibodies for the 47 kDa protein of T. p. pallidum; Riviere et al, 1991) or from Treponema phagedenis (not considered a pathogen), bands of about 1 and 0.6 kb were obtained. Example 8. Expression in E. coli of Recombinant Gpd and D15 To further characterize the clones described in Example 2, efforts were made 10 to express in E. coli the genes contained in all 10 of the immunoreactive lambda plaques. However, the products these positive lambda plaques proved to be difficult to obtain because of apparent toxicity to E. coli of the proteins expressed from these clones. Such toxicity is typical of outer membrane proteins. During the original immunoscreening of the lambda expression library (Example 2), protein expression 15 from the Lambda ZAP protein did not depend upon survival of the E. coli host, thus the toxicity to E. coli of these proteins was not apparent during the initial screening. However, in order to obtain cloned DNA for nucleotide sequence analysis, the immunoreactive plaques identified in this screen were subsequently subjected to in vivo excision to recover the positives as pBluescript phagemids, a process that is 20 strictly dependent upon survival of the E. coli host strain. Of the ten positive plaques, seven were successfully converted to pBluescript phagemids only after several attempts, while the remaining three so far have not been converted successfully. With regard to these last three clones, though their inserts have not yet been identified, it has been shown that they do not encode Gpd because it has not been possible amplify 25 their inserts using PCR primers corresponding to the Gpd sequences. Methods expected to ultimately obtain expression of the remaining clones will involve minimal bacterial growth times to prevent accumulation of the toxic protein, lowering the growth temperature to 30 0 C instead of the standard 37 0 C to prevent bacterial overgrowth, immediate purification of recombinant proteins from recently 30 transformed bacterial constructs rather than purification from previously frozen bacterial construct stock cultures, and additional experimental approaches. In addition to Gpd (SEQ ID NO: 1), the T.p. pallidum homologue of D15/Oma 87 (SEQ ID NO:3) was expressed in E. coli with the pRSET expression vector system. The expressed D15 homologue was used to immunize rabbits, as 35 described below in Example 10. Antibodies to this protein are being prepared.

WO 99/53099 PCT/US99/07886 -42 Example 9. Characterization of T p. pallidum Gpd protein The T. p. pallidum Gpd protein (SEQ ID NO:2) was expressed in E. coli BL21 (DE3) pLysS using the pET-3a expression system by inserting the entire coding region of Gpd (SEQ ID NO: 1). This yielded a full-length, 41 kDa recombinant 5 protein molecule. To verify that the T. p. pallidum Gpd (SEQ ID NO:2) indeed possessed the predicted enzymatic activity, Gpd activity was measured in crude lysates of E. coli that were expressing the recombinant molecule. (Larson et al., .1 Biol. Chem., 258, 5428-5432, 1983). A glycerophosphodiester phosphodiesterase functions by 10 hydrolyzing glycerophosphodiesters from phospholipid and triglyceride metabolism to glycerol 3-phosphate. The assay used here measures the conversion of the substrate glycerophosphocholine, a glycerophosphodiester, to dihydroxyacetone phosphate (DHAP) via glycerol 3-phosphate with the concomitant reduction of NAD to NADH. This reduction of NAD is followed spectophotometrically by measuring the increase 15 in absorbance at 340 nm. In brief, aliquots of a sonicated lysate of E. coli expressing the recombinant T. p. pallidum Gpd were added to a hydrazine/glycine 0.5 ml assay mixture containing NAD, CaCl 2 , and glycerol-3-phosphate dehydrogenase. The substrate glycerophosphocholine was then added to 0.5 ptm. A background control to account 20 for the E. coli intrinsic Gpd activity (a.k.a. "GlpQ") was provided by a sonicated lysate of E. coli transformed with only the pET-3a vector, i.e., the vector with no T. p. pallidum Gpd inserT. A positive assay was considered one in which an increase in absorbance at 340 nm was observed in E. coli expressing the recombinant T. p. pallidum Gpd over the absorbance at 340 nm observed in the background 25 control sample. The results of these assays indicated a three fold increase in absorbance in E. coli transformed with the T. p. pallidum Gpd (SEQ ID NO: 1). These assay results thus demonstrated that the recombinant Gpd was enzymatically active and, at least within the context of the enzyme's active site, conformationally correct, a characteristic important to various manipulations involving the recombinant 30 T. p. pallidum Gpd (SEQ ID NO: 1). Inclusion bodies containing recombinant T. p. pallidum Gpd (SEQ ID NO:2) were recovered from transformed E. coli and used as an immunogen to generate polyclonal antiserum. This antiserum failed to induce opsonization of T. p. pallidum appreciably compared to nonimmune rabbit serum. One possible reason for this result 35 may be that Gpd is not involved in opsonization, but alternatively, it may be that Gpd WO 99/53099 PCT/US99/07886 -43 is an opsonic target antigen, but that for opsonization to occur addition opsonic target antigens must also be present. A 1:1000 dilution of the rabbit anti-Gpd antiserum was used to develop Western blots containing lysates of T. p. pallidum before and after washing by 5 centrifugation. The washes are know to partially remove the bacterium's outer membrane. Blots were developed with 1:3000 dilution of goat anti-rabbit IgG (peroxidase-conjugated Fab fragment, Amersham), using the chemiluminescence protocol provided by Amersham. An immunoreactive band was observed that had a size of 41 kDa, the approximate molecular weight predicted for Gpd from the open 10 reading frame identified in the cloned DNA. The 41 kDa band was not observed in control blots developed with normal rabbit serum collected from the same rabbits prior to immunization. The signal for Gpd was observed in lysates obtained from unwashed, once-washed, and from thrice-washed treponemes, but signal strength diminished noticeably with increasing numbers of washes. These results thus imply 15 that Gpd is associated with the outer membranes of T. p. pallidum. The polyclonal antiserum to Gpd was used in further studies to analyze the surface disposition of Gpd using a previously described immunofluorescence assay (Cox et al., Mol. Microbiol., 15:1151-1164, 1995). Because of the fragility of the T. pallidum outer membrane, special precautions to preserve this membrane were 20 employed (Cox et al., 1995) Briefly, virulent 7'. pallidum were encapsulated in gel microdroplets to preserve the treponemal molecular architecture prior to immunofluorescence analysis, thus ensuring an accurate cellular localization for Gpd within T. pallidum. Preliminary results using the anti-Gpd antiserum showed uniform surface immunofluorescence on both intact and detergent-treated T. pallidum, as did 25 immune rabbit serum collected from chronically infected rabbits. To ensure that the integrity of the T. pallidum cellular architecture had been maintained despite the experimental manipulations, the level of immunoreactivity was examined for pre immune serum and serum prepared against the periplasmic 37 kDa endoflagellar sheath protein (Isaacs et al., InfecT. Immun., 57:3403-3411, 1989). The pre-immune 30 serum lacked immunoreactivity against either the intact or the detergent-treated treponemes, while the anti-37 kDa serum was reactive only against detergent treated treponemes, a finding consistent with its periplasmic location. These studies thus support a cell surface disposition for the T. p. pallidum Gpd. Because the H. influenzae Gpd homologue has been reported to have 35 IgG-binding capability (Janson et al., InfecT. Immun., 59:119-125, 1991; Sasaki and WO 99/53099 PCT/US99/07886 -44 Munson, InfecT. Immun., 61:3026-3031, 1993), the immunoglobulin binding capacity of the recombinant T. pallidum Gpd was investigated. To analyze the immunoglobulin binding capability of recombinant T.p. pallidum Gpd, inclusion bodies were purified from E. coli transformants using standard techniques, subjected 5 to SDS-PAGE analysis, and transferred to Immobilon-PVDF. The blots were exposed first to one of several types of immunoglobulin (primary immunoglobulin), washed, and then to the corresponding peroxidase-conjugated secondary antibody, followed by use of the Enhanced Chemiluminescence (ECL) Detection system (Amersham). The antibody pairs used were: (i) human IgA followed by goat F(ab') 2 10 anti-human IgA (u-chain specific); (ii) human IgD followed by goat F(ab') 2 anti human IgD (8-chain specific); (iii) human IgG followed by goat F(ab') 2 anti-human IgG (y-chain specific); and (iv) human IgM followed by goat F(ab') 2 anti-human IgM (u-chain specific). For control blots, the primary incubation was conducted in the absence of any primary immunoglobulin. As expected, no signal was observed for the 15 control blots. Results of these binding studies showed that the recombinant T. p. pallidum Gpd bound specifically with human immunoglobulins A, D and G but not M. The immunoglobulin binding was specific for the T. p. pallidum Gpd and did not represent spurious binding by a contaminating E. coli protein, as no immunoglobulin binding 20 was observed for similarly prepared inclusion bodies from E. coli expressing the pET-3a vector alone. The IgG binding of T. pallidum Gpd was further characterized by IgG fractionation studies. For these studies, Fab and Fc fragments of human IgG were prepared by papain digestion, and purified using a standard procedure (Harlow and 25 Lane, Eds., Antibodies: A Laboratory Manual, Cold Spring Harbor, NY, 1988, which is hereby incorporated by reference in its entirety). Immunoblots were incubated with either the Fab or Fc fragment, then developed with horseradish peroxidase/goat anti-human IgG (F(ab')2 fragment) and the Enhanced Chemiluminescense Reagent (Amersham, Cleveland, OH). Results of binding assays 30 with these IgG fragments revealed that the T. p. pallidum Gpd specifically binds the Fc fragment of human IgG with an intensity similar to that observed for intact IgG, while no binding to either the Fab fragment of human IgG or the secondary antibody was detected. Control lanes containing inclusion bodies prepared from E. coli transformed with the pET-3a vector alone once again did not exhibit binding to intact 35 IgG, IgG Fc and Fab fragments or the secondary antibody.

WO 99/53099 PCT/US99/07886 -45 In H. influenzae, the Gpd homologue has been linked to pathogenesis, as Gpd knockout mutants for that organism have been shown to be 100-fold less virulent in animal models (Janson et al., InfecT. Immun., 62:4848-4854, 1994). Similarly, Gpd may be relevant to the pathogenesis of T. pallidum. It has been proposed that the 5 coating of T. pallidum by host IgG is a factor in long-term treponeme survival in the host (Alderete and Baseman, InfecT. Immun., 26:1048-1056, 1979), a hypothesis that is consistent with the present indications that Gpd is disposed on the treponeme surface and that Gpd avidly binds the Fc region of IgG. The binding of T. pallidum Gpd to IgA and IgG is significant also because IgA and IgG represent much of the 10 immunoglobulins at mucosal surfaces where syphilis is sometimes transmitted. Example 10. Induction of Protective Immunity by Gpd, D15, and MSP Gpd: If Gpd contributes to treponemal evasion of the host immune system, the introduction of excess high affinity Gpd-specific antibodies through recombinant Gpd 15 vaccination may provide protective immunity to T. p. pallidum infection. The protection afforded by immunization with Gpd was tested in the rabbit syphilis model in two separate experiments. In the first experiment, one rabbit was immunized with inclusion bodies purified from E. coli expressing the pET-3a-Gpd construct emulsified in RIBI® adjuvant prior to intradermal challenge. A control rabbit received no prior 20 immunization and served as a comparison animal for intradermal challenge. The test rabbit was immunized intramuscularly, subcutaneously, and intradermally three times at three-week intervals with RIBI adjuvant using 200 Rg recombinant Gpd per immunization. One week after the final boost, the immunized and unimmunized control rabbits were challenged intradermally at each of six sites with 103 T. pallidum 25 Nichols strain per site. The Gpd immunized rabbit developed atypical pale, flat, slightly-indurated and non-ulcerative lesions within several days of challenge at two out of the six challenge sites, with no lesions observed at the remaining four challenge sites. In contrast, the control rabbit developed typical red, raised, highly-indurated and ulcerative lesions at 30 five of six challenge sites at 12 to 14 days post-challenge. In a second vaccination trial, the above immunization and challenge protocol was repeated using four rabbits immunized with the pET-3a-Gpd inclusion body preparation prior to intradermal T. pallidum challenge. Four control rabbits were similarly immunized with inclusion bodies purified from E. coli expressing the pET-3a 35 vector alone. As an additional control, another four rabbits received no prior WO 99/53099 PCT/US99/07886 -46 immunization. After challenge, all eight control rabbits developed typical red, raised, highly-indurated and ulcerative lesions at each of the six challenge sites, while all four of the Gpd-immunized rabbits developed atypical pale, flat, slightly-indurated and non-ulcerative reactions at each of the six challenge sites. In all cases, the reactions in 5 the Gpd-immunized animals resembled delayed type hypersensitivity responses more than typical syphilis chancres and resolved before lesions appeared in the control animals. This is the first time a defined vaccine has been shown to be protective against T. pallidum challenge, in marked contrast to previous experiments where no 10 protection was observed when rabbits were immunized with a variety of recombinant T. pallidum proteins. Dark field examination of the challenge sites were performed 31 days following the infection, and revealed treponemes in four of four unimmunized control rabbits and three of four control pET-3a vector-immunized rabbits. No treponemes 15 were observed in the three pET-3a-Gpd construct-immunized animals. The fourth pET-3a-Gpd rabbit could not be evaluated at this point, as it had expired. The absence of treponemes in one of the control rabbits may reflect an adjuvant effect and/or animal to animal variability. In summary, these results indicate that immunization with the Gpd antigen is 20 significantly protective for challenge with T. p. pallidum. Gpd represents the first surface-exposed, immunoprotective antigen reported for T. p. pallidum, and thus is valuable for a human syphilis vaccine. T. p. pallidum D15/Oma87 Homologue: In T. p. pallidum infected rabbits, anti-D15 antibodies were observed to 25 develop between days 13 and 17, and to peak at about day 30 after infection, after which time the level of anti-D15 activity decreased slightly and plateaued. Thus the appearance of antibodies to the T. p. pallidum D15 corresponds to the appearance of antibodies that opsonize and block cytoadherence of the organism, and to the time of treponemal clearance from the syphilis lesions in these animals. Thus, immunization 30 with the D15/Oma87 homologue is likely to elicit protective immunity, especially given that D15 of H. influenzae and the Oma87 protein of Pastuerella multocida are protective against infection by those organisms (Flack et al., Gene, 156:97-99, 1995; Loosmore et al., InfecT. Immun., 65:3161-3167, 1996; Ruffolo and Alder, InfecT. Immun., 64:3161-3167, 1996).

WO 99/53099 PCT/US99/07886 -47 To determine directly whether D15 is capable of eliciting protective immunity, a sector of the coding region corresponding to base pairs 76-2514 of the D15/Oma87 homologue (SEQ ID NO:3) that does not include the cleavable signal sequence, was cloned into the pRSET-C expression vector, and was expressed in E. coli BL21 5 (DE3) pLysS. The amino acid sequence of this portion of the D15/Oma87 homologue is shown in SEQ ID NO:6. The T. p. pallidum recombinant D15 was purified using Ni-NTA matrices according to the manufacturer's instructions (Qiagen, Valencia, CA). Using 200 tg of the recombinant D15, one rabbit was immunized using the 10 vaccination protocol described above for Gpd. This rabbit will be challenged with T. p. pallidum as described above for Gpd. Msp: Because of the methods by which the Msp homologues of T. p. pallidum were here identified, this protein family was thought likely to provide an effective syphilis 15 vaccine. The 785 bp at the 5' end of TP 1.6 (SEQ ID NO:45), which corresponds to the 5' half of Msp 2 (SEQ ID NO:9), was expressed with a 6-histidine tag in the pRSET system (Kroll et al., DNA & Cell Biol., 12:441-453, 1993) to yield a polypeptide having 261 amino acids (SEQ ID NO:46). Recombinant protein was 20 purified by nickel chromatography and a rabbit was immunized subcutaneously, intramuscularly and intradermally with RIBI adjuvant and 200 tg recombinant TP 1.6 protein. Injections were given three times at three-week intervals, as for Gpd immunization. The rabbit that was immunized with the polypeptide corresponding to TP 1.6 25 (SEQ ID NO:46) was challenged with 10 5 T. p. pallidum, Nichols strain, intradermally in eight sites on the back. A control rabbit that was not immunized was also challenged. The TP 1.6-immunized rabbit developed small, slightly indurated patches which cleared in seven days. These lesions were not typical of syphilis chancres, but rather resembled delayed type hypersensitivity responses. The control 30 rabbit developed red, indurated nodules at the sites of inoculation at 5 days. These persisted and reached a maximum size of 2 cm and ulcerated at approximately 20 days. At 21 days, the VDRL (Venereal Diseases Research Laboratory cardiolipin-antibody test) serology of the TP 1.6-immunized animal remained negative, but the VDRL serology of the control rabbit was positive at 35 a 1:2 dilution. At 28 days, the TP 1.6-immunized animal was sacrificed, and its lymph WO 99/53099 PCT/US99/07886 -48 nodes and testes were minced and extracted for treponemes. Treponemes were found on dark field examination, implying that the TP 1.6 immunization was only partially protective. This experiment has now been now repeated with 2 additional rabbits and 5 darkfield examination of challenge sites revealed treponemes in only 2 out of 12 sites in immunized rabbits, but in 6 out of 6 sites in unimmunized rabbits. These results indicate that significant protection was achieved with TP 1.6 immunization against an extremely large challenge of T. p. pallidum (the ID50 for rabbits is 51 treponemes). To further explore the ability of Msp polypeptides to elicit protective 10 immunity, PCR primers were devised to specifically amplify the central variable region present in all of the Msps that contain a variable region. Because some of the Msp variable regions share short stretches of identity even within their variable regions, it was possible to amplify all of the variable regions using primers sets shown in Table 1. The amplified variable region DNAs were prepared from a T. p. pallidum 15 genomic DNA template using the primers in Table 1 to amplify all of the Msps (except for Msp 2), and each of the DNAs thus obtained was expressed in E. coli, and the recombinant polypeptides recovered in order to test their capacity to induce protective immunity against T. p. pallidum. The nucleotide sequences of these amplified DNA fragments are shown in SEQ ID NO:7 and SEQ ID NOS:11, 13, 15, 20 17, 19, 22, 24, 26, 28 and 30, and the amino acid sequence of each of the corresponding variable region recombinant polypeptides are shown in SEQ ID NO:8 and SEQ ID NOS:12, 14, 16, 18, 20, 21, 23, 25, 27, 29 and 31. In tests conducted so far, variable region polypeptides corresponding to Msps 1 (SEQ ID NO:8), 9 (SEQ ID NO:25) and 11 (SEQ ID NO:29) have been used 25 to immunize a single rabbit as described above for the first test conducted with the TP 1.6 amino terminus polypeptide (SEQ ID NO:46). Upon challenge with T. p. pallidum, immunization with Msp 9 (SEQ ID NO:25) and Msp 11 (SEQ ID NO:29), but not Msp 1 (SEQ ID NO:8), were found to have conferred protective immunity as compared with controls. Although Msp 1 (SEQ ID NO:8) failed to yield 30 positive results in this preliminary trial, it cannot be ruled out that the single rabbit inoculated here with Msp 1 (SEQ ID NO:8) was unusually susceptible to syphilis, or that Msp 1 (SEQ ID NO:8) could contribute to immunity if injected in combination with other Msp antigens. In other experiments, antiserum was withdrawn from rabbits immunized as 35 described above with Msp polypeptides 1 (SEQ ID NO:8), 9 (SEQ ID NO:25), 11 WO 99/53099 PCT/US99/07886 -49 (SEQ ID NO:29), and TP 1.6 (SEQ ID NO:46) and these antisera were tested in an opsonization assay. For this assay, in brief, rabbit macrophages were mixed with the test antiserum, treponemes added, then incubated for 4 hours. At that time, the cells were fixed and stained using an immunofluorescent tag specific for T. p. pallidum. 5 Macrophages containing ingested treponemes were scored by microscopy. All four test antisera were found to have promoted opsonization over negative control serum from unimmunized rabbits. IRS provided a positive control. In one such experiment, the 90 percentages of macrophages containing ingested treponemes were: unimmunized control, 16.9%; IRS, 45.3%; Msp 1 antiserum, 67.9%; Msp 9 10 antiserum, 47.4%; Msp 11 antiserum, 33.5%; and TP 1.6 32.7%. These values are the averages of triplicate plates for each antiserum. It is of note that the protection seen after inoculation with the Msp 9 (SEQ ID NO:25) polypeptide was more complete than the protection seen after injecting polypeptides corresponding to the variable regions of TP 1.6 (SEQ ID NO:46) or 15 Msp 11 (SEQ ID NO:29). These results are consistent with other observations indicating that Msp 9 is expressed at relatively high levels during the early stages after infection of rabbits with the Nichols strain of T. p. pallidum (see Example 5). These experiments are being repeated in additional rabbits, and with the remaining Msp variable region polypeptides. This result is in marked contrast to previous 20 experiments in which no protection was observed when rabbits were immunized with a variety of recombinant T. p. pallidum proteins, including Tp47, Tp37, Tp34.5, Tp33, Tp30, Tpl7, Tpl5, Tpl90 (4D), Tp44.5 (TmpA), Tp34 (TmpB), Tp37 (TmpC), Tp 29-35 (TpD) (Tp terminology refers to MW consensus according to Norris et al., 54), and TROMP1 (Blanco et al., J. Bacteriol. 178:? 199?). Clearly, the 25 Msp family provides a group of antigens useful for vaccination against syphilis. As indicated above (see Examples 6 and 7), experiments have indicated that the pathogens T. p. pertenue and T. p. endemicum each contain several Msp genes. These are exploited for vaccine production by expressing these Msp homologues using a suitable vector, and the resulting polypeptides are used in combination with a 30 pyhsiologically acceptable carrier as vaccines to protect against yaws or bejel. By combining the Msp polypeptides derived from several different subspecies of Treponema pallidum, a vaccine is made whose administration to a suitable animal host confers protective immunity to syphilis, yaws and bejel. Such a vaccine may include the T. p. pallidum Gpd (SEQ ID NO:2) and D15/Oma87 homologues (SEQ 35 ID NO:4) disclosed above, and may further include Msp genes from pathogenic WO 99/53099 PCT/US99/07886 -50 spirochetes that cause oral disease. Due to the high degree of relatedness among these subspecies of T. pallidum, and because infection with any one of them has been noted to confer partial immunity against the other two, a vaccine comprising at least one Msp from any one of the three subspecies should confer at least partial protection 5 against infection with either of the other two. Example 11. Sequence Conservation of Glycerophosphodiester Phosphodiesterase Among Treponemapallidum Strains The suitability of the glycerophosphodiester phosphodiesterase (Gpd) as a potential syphilis vaccine candidate was further investigated by determining the degree 10 of Gpd sequence conservation among pathogenic treponemes. Bacterial species. The Gpd coding sequence was PCR amplified from genomic DNA isolated from a variety of treponemal strains. All strains were propagated in New Zealand white rabbits as previously described (Lukehart, S. A., S. A. Baker-Zander, and S. Sell. 1980. Characterization of lymphocyte responsiveness in 15 early experimental syphilis. I. In vitro response to mitogens and Treponema pallidum antigens. J. Immunol. 124:454-460). T. pallidum subsp. pallidum, Nichols strain, was originally sent to the University of Washington by James N. Miller (University of California, Los Angeles) in 1979, and T. pallidum subsp. pertenue, Gauthier strain, was supplied by Peter Perine (Centers for Disease Control, Atlanta, GA) in 1981. 20 T. pallidum subsp. pallidum, Bal-3, Bal-7 and Bal 73-1 strains; T. paraluiscuniculi, Cuniculi A strain; T. pallidum subsp. pertenue, Haiti B strain; T. pallidum subsp. endemicum, Iraq B strain; and the Simian isolate were supplied by Paul Hardy (John Hopkins University, Baltimore, MD). T. pallidum subsp. pallidum, Sea 81-3 and Sea 83-1 strains, were isolated by Sheila A. Lukehart from the cerebrospinal fluid of 25 untreated syphilis patients. PCR amplifications. To obtain the entire gpd open reading frame, primers were designed from the 5' (5'-TGCACGGTGACGATCTGTGC-3')(SEQ ID NO:70) and 3' (5'-GGTACCAGGCGACACTGAAC-3')(SEQ ID NO:71) non-coding regions flanking the gpd gene (Fraser, C. M.,et al., 1998, Science 281:375-388). These 30 primers are located 48 bp upstream and 51 bp downstream, respectively, of the gpd open reading frame. PCR amplification of the gpd gene was performed using a 100 pl reaction containing 200 pM dNTP's, 0.25 pM of each primer, lx Taq polymerase buffer (50 mM Tris-HCl, pH 9.0 at 20 0 C, 1.5 mM MgCl2, 20 mM NH4SO4), and 1 pl of genomic DNA containing 5,000-10,000 treponeme equivalents for each strain. 35 The PCR reaction conditions were 30 cycles of 1 minute denaturation at 94oC, WO 99/53099 PCT/US99/07886 -51 1 minute annealing at 60 0 C, and 2 minutes extension at 74 0 C. For each reaction, "hot start" PCR (Chou, Q., M. Russell, D. E. Birch, J. Raymond, and W. Bloch. 1992. Prevention of pre-PCR mis-priming and primer dimerization improves low-copy number amplifications. Nucleic Acids Res. 20:1717-1723) was performed by adding 5 2.5 units of Taq polymerase after the initial denaturation step. Following PCR, the amplification products were cloned into the pGEM-T vector (Promega, Madison, WI) and each insert was sequenced in its entirety in both directions. To reduce the possibility of PCR- or sequencing-induced errors, two clones derived from independent PCR amplifications were sequenced for each strain. 10 Sequence analysis. Double-stranded plasmid DNA was extracted using the Qiagen Plasmid Mini Kit (Qiagen, Chatsworth, CA) and both strands of insert DNA were sequenced using the Applied Biosystems dye terminator sequencing kit (PE Applied Biosystems, Foster City, CA) and the ABI 373A DNA sequencer in accordance with the manufacturer's instructions. In all cases both universal 15 sequencing primers and internal primers designed from the insert sequence were used. Nucleotide sequences were translated and analyzed using the Sequencher m Version 3.1RC4 sequence analysis software (Gene Codes Corporation, Ann Arbor, MI). Alignment of protein and DNA sequences was performed using the Clustal W general purpose multiple alignment program (Thompson, J. D., D. G. Higgins, and T. J. 20 Gibson. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, positions-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22:4673-4680). Restriction fragment length polymorphism (RFLP) analysis. RFLP analysis was performed on the gpd open reading frame amplified from each treponeme strain. 25 One microgram of each of the amplified templates was digested with PleI (New England Biolabs, Beverly, MA) for four hours at 37 0 C prior to electrophoresis on a 1.5% NuSieve ® (FMC BioProducts, Rockland, ME) agarose gel. Nucleotide accession numbers. The nucleotide sequences of the gpd genes from the Nichols, Bal-3, Bal-7, Bal 73-1, Sea 81-3, Sea 83-1, Mexico A, Haiti B, 30 Gauthier, Iraq B, Simian, and Cuniculi A strains have been assigned GenBank accession numbers AF004286 and AF127415-AF127425, respectively, each of which nucleotide sequences, accorded the foregoing GenBank accession numbers, are incorporated herein by reference. As shown in Table 2, all six strains of T. pallidum subsp. pallidum have 35 identical Gpd gene sequences, while the other human subspecies (pertenue and WO 99/53099 PCT/US99/07886 -52 endemicum) and the animal pathogens (Simian strain and T. paraluiscuniculi) have a silent A to G change at base pair 579. Table 2. Summary of Gpd sequence conservation between T. pallidum subsp. pallidum (Nichols strain) and various pathogenic treponeme strains. 5 Subspecies Strain Sequence Divergence from Nichols (%) nucleotide amino acid pallidum Bal-3 none none pallidum Bal-7 none none pallidum Bal 73-1 none none pallidum Sea 81-3 none none pallidum Sea 83-1 none none pallidum Mexico A none none pertenue (?) Haiti B none none pertenue Gauthier base pair none 579, A to G endemicum Iraq B base pair none 579, A to G ? Simian base pair none 579, Ato G paraluiscuniculi Cuniculi A base pair residue 88, R to H 263, G to A none base pair none 459, A to G none base pair none 579, A to G none base pair 711, Ato G base pair 960, C to T base pair 999, G to C Interestingly, T. paraluiscuniculi (the only different species represented) has 5 additional base pair changes, one of which (base pair 263) results in a conservative WO 99/53099 PCT/US99/07886 -53 amino acid substitution at residue 88. This demonstrates genetic divergence of the nonvenereal treponemal strains and the rabbit pathogen away from the syphilis strains, consistent with their different clinical diseases and host ranges. The Simian strain has been thought to be very closely related (or identical) to the human pertenue 5 subspecies (Felsenfeld, O., and R. H. Wolf 16:294-305(1971); Sepetjian, M., F. T. Guerraz, D. Salussola, J. Thivolet, and J. C. Monier 40:141-151(1969)), and this study supports this hypothesis. The base pair change at position 579 in the non-syphilis strains introduces a PleI restriction site that creates different RFLP patterns between the T. pallidum 10 subsp. pallidum strains and the other human and animal pathogens. PleI digestion of the T. pallidum subsp. pallidum strains generates three restriction fragments of sizes 766, 241 and 163 base pairs. The presence of the additional PleI site in the non syphilis strains generates four restriction fragments of sizes 635, 241, 163 and 131 base pairs. These characteristic RFLP patterns provide a means of genetically 15 differentiating between infections caused by the pallidum subspecies and those caused by the various other pathogenic treponemes. The finding that the Haiti B strain, which is reportedly a T. pallidum subsp. pertenue strain, shows sequence identity with the pallidum subspecies and not with the non-syphilis strains supports the proposal by Centurion-Lara et al. (Centurion 20 Lara, A., C. Castro, R. Castillo, J.M. Shaffer, W. C. Van Voorhis, and S. A. LukeharT., J. InfecT. Dis. 177:1036-1040(1998)) that this strain is misidentified and should be classified as a T. pallidum subsp. pallidum strain. Similar sequence analyses performed on the tpr K (Centurion-Lara, A., C. Castro, W. C. Van Voorhis, and S. A. LukeharT. Unpublished data) and tp92 (Cameron, C. E., C. Castro, S. A. 25 Lukehart, and W. C. Van Voorhis. Unpublished data) sequences from the Haiti B strain further support its identification as a T. pallidum subsp. pallidum strain. Homologues of Gpd from other bacterial species also demonstrate remarkable conservation of amino acid sequence. The enzyme from Haemophilus influenzae, designated Protein D, is 98% conserved among eight strains (Song, X., A. Forsgren, 30 and H. Janson., InfecT. Immun. 63:696-699(1995)). The corresponding molecule from the relapsing fever spirochete Borrelia hermsii, GlpQ, exhibits a range of 96.5% to 100% amino acid sequence similarity among 26 B. hermsii isolates (Schwan, T. G., and S. F. Porcella. Personal communication). Similarly, results reported here show Gpd is highly conserved among twelve strains that encompass a total of five 35 pathogenic treponemes. The invariant nature of the Gpd, combined with the WO 99/53099 PCT/US99/07886 -54 immunoprotective capability previously described for this molecule in the experimental syphilis model (Cameron, C. E., C. Castro, S. A. Lukehart, and W. C. Van Voorhis, InfecT. Immun. 66:5763-5770 (1998)), make it an attractive candidate for inclusion in a universal subunit vaccine against T. pallidum infection. 5 Example 12.Opsonic Potential, Protective Capacity and Sequence Conservation of the Treponema pallidum subsp. pallidum Tp92 The Tpallidum D15/Oma87 homologue protein is referred to as Tp92 in the present example. As discussed more fully herein, Tp92 is protective against challenge with T pallidum. As disclosed more fully herein, the predicted Tp92 amino acid 10 sequence from a variety of different strains of Tpallidum is almost identical. This observation suggests that immunization with Tp92 should protect against many strains of syphilis. Additionally, as discussed more fully herein, Tp92 is a target of opsonizing antibodies for Tpallidum, and thus Tp92 is likely to be a surface antigen. Bacterial Strains. All T. pallidum subspecies and strains were propagated in 15 New Zealand white rabbits as previously described (Lukehart, S.A., S.A. Baker Zander, and S. Sell. J Immunol. 124:454-460 (1980)). E. coli XL-1 Blue, SolR and BL21 (DE3) pLysS were obtained from Stratagene (La Jolla, CA). Expression Library Screening. The T. pallidum subsp. pallidum tpa92 gene was identified using the previously published method of differentially screening a 20 T. pallidum genomic expression library (Stebeck, C.E., et al. FEMS Microbiol. Lett. 154:303-310 (1997)). Briefly, the library was prepared using the Lambda ZAP® II cloning kit (Stratagene) according to the manufacturer's instructions. Approximately 200,000 plaques (12,500 pfu/plate) were plated and duplicate lifts prepared and screened using established methods (Sambrook, J., E.F. Fritsch, and T. Maniatis. 25 1989. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY). Filters were differentially screened with a T. pallidum-specific immune rabbit serum depleted of activity against the major known treponemal antigens but still retaining its opsonic capacity (termed opsonic rabbit serum; ORS), and a non-opsonic antiserum prepared using heat-killed 30 T. pallidum (termed non-opsonic rabbit serum; NORS). The ORS was prepared by sequential adsorption of pooled syphilitic rabbit serum with T. phagedenis, biotype Reiter, recombinant T. pallidum 47, 37, 34.5, 33, 30, 17 and 15 kDa molecules (as designated in Table 3 in Norris, S.J. et al., Electrophoresis 8:77-92 (1987), incorporated herein by reference) and recombinant Tromp 1 (Blanco, D.R., C.I. 35 Champion, M.M. Exner, H. Erdjument-Bromage, R.E. Hancock, P. Tempst, J.N.

WO 99/53099 PCT/US99/07886 -55 Miller, and M.A. Lovett. 1995. Porin activity and sequence analysis of a 31-kilodalton Treponema pallidum subsp. pallidum rare outer membrane protein (Trompl1). J Bacteriol. 177:3556-3562). In unpublished studies from our laboratory, antisera raised against electroeluted or recombinant forms of these antigens failed to 5 demonstrate opsonic function. The antiserum was further adsorbed with VDRL antigen, a lipid complex that has been shown to be the target of a minor portion of opsonic antibodies (Baker-Zander, S.A., J.M. Shaffer, and S.A. Lukehart. J. Infect. Dis. 167:1100-1105 (1993)). These adsorption steps were performed to reduce the number of irrelevant positive clones identified by this antiserum in the expression 10 library screening. Immunoreactive plaques were detected with 1 ptCi of 12 5 I-labeled protein A on nitrocellulose filters using established methods (Sambrook, J., E.F. Fritsch, and T. Maniatis. 1989. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY). Plaques showing reactivity with the ORS but no reactivity with the NORS were subjected to secondary 15 screening with both the ORS and the NORS. Those plaques showing consistent differential reactivity were screened a third time with ORS and converted to pBluescript SK(-) phagemids by in vivo excision in the E. coli strains XL-1 Blue and SolR according to the manufacturer's instructions. DNA Sequencing. Double-stranded plasmid DNA was extracted using the 20 Qiagen Plasmid Mini Kit (Qiagen, Chatsworth, CA) and both strands of insert DNA were sequenced using the Applied Biosystems dye terminator sequencing kit (PE Applied Biosystems, Foster City, CA) and the ABI 373A DNA sequencer in accordance with the manufacturer's instructions. In all cases both universal sequencing primers and internal primers designed from the insert sequence were used. 25 DNA and Protein Sequence Analyses. Nucleotide sequences were translated and analyzed using the SequencherT M Version 3.1RC4 sequence analysis software (Gene Codes Corporation, Ann Arbor, MI). Database searches were performed using the basic local alignment search tool (BLAST) algorithm (Altschul, S.F., et al. J. Mol. Biol. 215:403-410 (1990)) and either the blastn, blastx or blastp programs. The 30 published T. pallidum genome (http://utmmg.med.uth.tmc.edu/treponema/tpall.html) was used to obtain the complete tpa92 open reading frame and the corresponding non-coding flanking regions. Alignment of protein and DNA sequences was performed using the Clustal W general purpose multiple alignment program (Thompson, J.D. et al. Nucleic Acids Res. 22:4673-4680 (1994)). The percentage of 35 positional identity and similarity between sequences was calculated from the number WO 99/53099 PCT/US99/07886 -56 of identical or similar residues, respectively, between aligned sequences; insertions and deletions were not scored. For the predicted amino acid sequence of Tpa92, the molecular mass was calculated using the Compute pl/MW Tool (http://www.expasy.ch/ch2d/pitool.html), transmembrane topology analysis was 5 performed using the TMpred program (http://ulrec3.unil.ch/software/TMPRED), and signal sequence and cellular location predictions were performed using the PSORT program (http://psort.nibb.acjp:8 800 ). PCR Amplification of tpa92 from T. pallidum Subspecies and Strains. The Tpa92 coding sequence was PCR amplified from genomic DNA isolated from a 10 variety of T. pallidum subspecies and strains. To obtain the entire open reading frame, primers were designed from the 5' (5'-GGGTGTCGTGGAGTTTTGCG 3')(SEQ ID NO:72) and 3' (5'-CTTGCCTGGTGGACGCAGC-3')(SEQ ID NO:73) non-coding regions flanking the tpa92 gene. These primers are located 55 bp upstream and 49 bp downstream, respectively, of the tpa92 open reading frame. PCR 15 amplification of tpa92 was performed using a 100 gl reaction containing 200 pM dNTP's, 0.25 pM of each primer, lx Taq polymerase buffer (50 mM Tris-HC1, pH 9.0 at 20 0 C, 1.5 mM MgCl2, 20 mM NH4SO4), and 1 ptl of genomic DNA containing 5,000-10,000 treponeme equivalents for each T. pallidum subspecies and strain. The PCR reaction conditions were as follows: 30 cycles of 1 minute denaturation at 94oC, 20 1 minute annealing at 60 0 C, 2 minutes extension at 74C for T. pallidum Bal 73-1, Bal-3, Bal-7, Sea 81-3, Sea 83-1, Haiti B and Simian templates; 35 cycles of 1 minute denaturation at 94 0 C, 1 minute annealing at 55 0 C, 2 minutes and 30 seconds extension at 74C for the T. pallidum Gauthier template; and 35 cycles of 1 minute denaturation at 94oC, 1 minute annealing at 60 0 C, and 2 minutes and 30 seconds 25 extension at 741C for the T. pallidum Cuniculi A template. For each reaction, "hot start" PCR (Chou, Q. et al., Nucleic Acids Res. 20:1717-1723 (1992)) was performed by adding 2.5 units of Taq polymerase after the initial denaturation step. Following PCR, the amplification products were cloned into the pGEM-T vector (Promega, Madison, WI) and each insert was sequenced in its entirety in both directions. To 30 reduce the possibility of PCR- or sequence-induced errors, two clones derived from independent PCR amplifications were sequenced for each T. pallidum subspecies and strain. Overexpression Studies. The open reading frame encoding Tpa92 was PCR amplified from T. pallidum subsp. pallidum (Nichols strain) genomic DNA using 35 primers designed from the 5' (5'-CGGGATCCACAATTGGTACGAGGGAAAGCC- WO 99/53099 PCT/US99/07886 -57 3'; contains a BamHI site)(SEQ ID NO:74) and 3' (5'

CGGAATTCCTACAAATTATTTACCGTGAACG-

3 '; contains an EcoRI site)(SEQ ID NO:75) ends of the Tpa92 coding region. PCR amplification was performed as outlined above, using 30 cycles of 1 minute denaturation at 94 0 C, 1 minute annealing 5 at 60 0 C, and 2 minutes extension at 74C. To ensure optimal expression of the recombinant molecule within E. coli, the DNA sequence encoding the N-terminal 25 amino acids, which include the predicted signal sequence, were excluded from the primer design and, thus, from the resulting expressed recombinant molecule. Following PCR, the 2457 bp amplification product was digested with BamHI and 10 EcoRI, ligated to a similarly digested pRSETc T7 expression vector (Invitrogen, Carlsbad, CA) and transformed first into E. coli XL-1 Blue and then into the E. coli expression strain BL21 (DE3) pLysS. The reading frame and sequence of the expression construct was verified by DNA sequencing using the T7 promoter primer (Pharmacia, Piscataway, NJ) and internal primers designed from the tpa92 DNA 15 sequence, the Applied Biosystems dye terminator sequencing kit and the ABI 373A DNA sequencer according to the manufacturer's instructions. Expression of the recombinant T. pallidum Tpa92 was performed using 500 ml of LB broth seeded with 50 ml of OD 0.6 E. coli transformed with the Tpa92-pRSETc construct. Cells were grown for 3 hours at 30 0 C prior to induction of protein expression from the T7 20 promoter by the addition of 0.4 mM IPTG and a further 4 hour incubation at 30 0 C. Cells were harvested by centrifugation, and the histidine-tagged recombinant Tpa92 protein was purified from the bacterial pellet according to the manufacturer's instructions (Invitrogen). Antisera. Immune rabbit serum (IRS) was collected from rabbits that had 25 been chronically infected with T. pallidum for >90 days. Anti-Tpa92 polyclonal antiserum was raised in four New Zealand white rabbits (#5061, #5200, #5202, and #5207) by immunizing three times with 100 jtg each of the purified recombinant Tpa92 emulsified in the Ribi adjuvant MPL + TDM + CWS (Monophosphoryl lipid A + Trehalose dicorynomycolate + Cell wall skeleton; Sigma, St. Louis, MO). 30 Immunizations were administered intradermally (ID), subcutaneously (SC), intramuscularly (IM) and intraperitoneally (IP) at three week intervals as outlined by the Ribi adjuvant system, and antiserum was collected one week after the final immunization. Opsonization Assay. IRS, anti-Tpa92 polyclonal antiserum collected from 35 rabbit #5061, and the corresponding control pre-immune serum were tested in three WO 99/53099 PCT/US99/07886 -58 separate experiments with a total number of replicate assays of 9 (IRS), 7 (anti-Tpa92 serum) and 8 (pre-immune serum) for their ability to opsonize T. pallidum using a standard phagocytosis assay as previously described (Shaffer, J.M. et al., Infect. Immun. 61:781-784 (1993)). All antisera were used at a 1:100 dilution and 5 incubated for four hours with rabbit peritoneal macrophages and T. pallidum prior to determination of the percentage of macrophages phagocytosing treponemes. Statistical analysis was performed using the two-tailed Student t-test. PAGE and Immunoblot Analyses. Sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting were performed as previously 10 described (Baker-Zander, S.A. et al., J. Infect. Dis. 151:264-272 (1985)), except that samples were blotted to Immobilon-PVDF membrane (Millipore Corp., Bedford, MA). Heterologous expression of the recombinant T. pallidum Tpa92 was monitored by SDS-PAGE analysis of approximately 5 gg of total bacterial lysate or 2 gg of purified recombinant protein and subsequent staining with Coomassie blue R-250. 15 The level of immunoreactivity of anti-Tpa92 polyclonal antiserum on purified recombinant Tpa92 was assayed by electrophoresis and blotting of 2 gg of purified recombinant protein, and probing with a 1:200 dilution of anti-Tpa92 polyclonal rabbit serum followed by a 1:3000 dilution of alkaline phosphatase-labeled goat anti rabbit IgG (Fc; Promega). For analysis of the level of immunoreactivity of anti 20 Tpa92 antiserum on washed and unwashed treponemes, T. pallidum was extracted from infected testes as previously described (Lukehart, S.A. et al., J. Immunol. 121:2014-2024 (1978)) and either immediately resuspended in SDS-PAGE sample buffer (unwashed preparation) or washed one time or three times with 10mM Tris HC1 pH 7.5 by centrifugation (15,000 xg) prior to resuspension of the treponemes in 25 sample buffer. Approximately 1.4 x 107 T. pallidum were electrophoresed for each sample (unwashed, washed one time, washed three times), blotted and probed with a 1:200 dilution of anti-Tpa92 polyclonal rabbit serum (collected from rabbit #5061) followed by a 1:3000 dilution of alkaline phosphatase-labeled goat anti-rabbit IgG (Fc). All immunoblots were blocked with 5% milk powder in Tris-buffered saline 30 with 0.1% Tween-20 and developed using BCIP/NBT color substrate detection (Promega). RainbowTM high range molecular weight markers (Amersham, Cleveland, OH) were used as standards. Protection Experiments. Four New Zealand white rabbits, as designated above, were immunized three times (IM, SC, IP and ID) at three week intervals with 35 the Ribi MPL + TDM + CWS adjuvant and 100 gg purified recombinant Tpa92.

WO 99/53099 PCT/US99/07886 -59 Three weeks after administration of the final immunization, the immunized rabbits and two unimmunized control rabbits were intradermally challenged at each of eight sites on their shaved backs with 10 5 T. pallidum subsp. pallidum (Nichols strain) per site. The rabbits were examined daily to monitor the development, morphological 5 appearance and progression of lesions appearing at the challenge sites. Lesion development was designated for each individual rabbit as typical if lesions were red, raised, indurated and generally progressed to ulceration, and atypical if lesions were pale, flat, only slightly indurated and generally non-ulcerative. Prior to lesion ulceration on the control animals (19 days post-challenge), lesion aspirates were 10 collected from all challenge sites and examined by darkfield microscopy for viable treponemes. The serological status of all challenged rabbits was determined using the Venereal Disease Research Laboratory (VDRL) and the FTA-ABS tests at 4 weeks post-challenge. Statistical analyses were performed using the two-tailed Student t-test and analysis of variance with repeated measures. 15 Results Identification of T. pallidum subsp. pallidum tpa92. A Lambda ZAP II T. pallidum subsp. pallidum genomic expression library was constructed and screened with a T. pallidum-specific, antigen-adsorbed opsonic antiserum preparation. As the name implies, immunoreactivity against known 20 T. pallidum antigens had been adsorbed from this preparation, although the opsonic capability of the antiserum was retained as demonstrated by phagocytosis assays (data not shown). To aid in distinguishing plaques specifically reacting with opsonic antibodies from background immunoreactive plaques, duplicate plaque lifts were differentially screened with a T. pallidum-specific non-opsonic antiserum. Plaques 25 exhibiting consistent immunoreactivity with the opsonic antiserum but no immunoreactivity with the non-opsonic antiserum on the primary and secondary screens were selected for further study and subjected to tertiary screening to obtain well isolated plaques. In vivo excision of one immunoreactive plaque produced a pBluescript 30 phagemid containing a 3.0 kb insert, as shown by restriction digest analysis (data not shown). Nucleotide sequence analysis of the insert revealed a 2439 bp open reading frame encoding an 812 amino acid translated product. Comparison of the insert sequence with an early version (July, 1997) of the released T. pallidum genome sequence (http://utmmg.med.uth.tmc.edu/treponema/tpall.html) identified 75 bp at the 35 5' end of the open reading frame that were missing from the insert sequence of the WO 99/53099 PCT/US99/07886 -60 immunoreactive clone. This DNA sequence was downstream from a putative ribosome binding site and thus was presumed to encode the N-terminal 25 amino acids of the translated protein product. Subsequent release of the completed T. pallidum genome identified the putative open reading frame between base pairs 5 344,276 and 346,834 of the genome, corresponding to open reading frame TP0326 (genbank accession number AE001212; Fraser, C.M. et al., Science 281:375-388 (1998)). This open reading frame encodes a slightly larger translated protein containing an extra 16 amino acids at the N-terminus, a discrepancy that arises due to the assignment of an alternative initiator methionine. 10 PSORT analysis (http://psort.nibb.acjp:8800) performed on the complete 837 residue translated protein predicts a 21 amino acid cleavable N-terminal signal sequence and an 84.6% likelihood that this putative protein is located in the T. pallidum outer membrane. The mature translated protein, lacking the 21 residue signal sequence, has a predicted molecular mass of 92,040 Da. This translated protein 15 was designated Tpa92 (T. pallidum antigen, 92 kDa). The DNA sequence of Tpa92 is incorporated herein by reference and is available from EMBL/Genbank/DDBJ under accession number AF042789. Sequence Analyses. As shown in Table 3, sequence database analysis using the blastp algorithm 20 (Altschul, S.F. et al., Basic local alignment search tool. J Mol. Biol. 215:403-410 (1990)) revealed the T. pallidum Tpa92 shares the highest degree of sequence similarity with a putative outer membrane protein identified by genome sequencing of the related spirochete, Borrelia burgdorferi (28.1% identical, 44.7% similar; Fraser, C.M. et al., Nature 390:580-586 (1997)).

WO 99/53099 PCTIUS99/07886 -61 6 .0 I-) knI-C .~0 4)C) D 1C -D 1 00 o Cd '-4 tl t 0 -++ 000 C)m C 0 0 0*' 4) 0 F- M w~ N4 r- C W , C0 C 0 cq - c0 W N-t w0 z S C t~- t- - wt 4) 0 r, W ) W 4 000w000000Qo 0~ o L 0 U0 0) 0 0 WO 99/53099 PCT/US99/07886 -62 The T. pallidum Tpa92 also shares approximately equal levels of sequence similarity with high molecular weight outer membrane proteins identified from a large variety of bacterial species (18.6-22.1% identical, 35.1-40.9% similar). The observed sequence similarity within this group of bacterial proteins is evenly distributed 5 throughout the coding sequence of Tpa92, with the exception of a stretch of serine residues at the C-terminal end of the translated protein that is unique to the T. pallidum Tpa92. The presence of transmembrane segments within Tpa92 was analyzed using the TMPred program, resulting in the prediction of three transmembrane helices (data not shown). In this putative model, the C-terminal 10 serine-rich stretch of Tpa92 is predicted to be located within an external loop on the outer face of the outer membrane. Sequence Conservation of Tpa92 Among T. pallidum Subspecies and Strains. To assess the degree of sequence conservation of Tpa92 among T. pallidum subspecies and strains, the tpa92 open reading frame was PCR amplified and 15 subsequently sequenced from six additional T. pallidum subsp. pallidum strains, two 7. pallidum subsp. pertenue strains (causative agent of the disease Yaws), one T. pallidum subsp. paraluiscuniculi strain (causes venereal syphilis in rabbits), and the Simian strain. The sequence divergence observed for each of these strains from the Tpa92 sequence of T. pallidum subsp. pallidum Nichols strain is tabulated in 20 Tables 4-7, and the overall percentage of sequence conservation for each strain compared to the Nichols strain is summarized in Table 8.

WO 99/53099 PCTIUS99/07886 -63

C-

4 0) 0 00 000 0 0 4-44 00 r -4 m 00 0 00 0- O ~ 0 o ~ or orLf o 0 0 0 0 0-~ 0 -~~~~~~~ ___ -___ (UU d) 0) 0) t 0 00 G -o kn W) t .00G 644 -oq ~) Cd Ud WO 99/53099 PCTIUS99/07886 -64 Nz u 0 0 00 0- W) ) 4 o en00C0 64 W) 0 t- 00 0 .)0 0 0 0 0 0* 0 0 C.) in C) W) t I-nID N C od 4 .~- '~ - 0 -4 -4 i- 4 o , 'r q * ~ ~ ~I N00 000 WO 99/53099 PCTIUS99/07886 -65 -5 2 00 m )e oiC 10C r- 'N t - 0 0 04 u 00 00 - C 00C r-000 W)) 00 WO 99/53099 PCTIUS99/07886 -66 N 00 Zz 0 0 000 co 0 4) 0 \ '. ~ .0 Nf M,. 00 .0 4) ___ __ - -o U, 0'.n WO 99/53099 PCTIUS99/07886 -67 rAI 4 ~~00 o) 0 r- C4 n ) t w~ o 0000 m) W) W n )W W W ) CA I WO 99/53099 PCTIUS99/07886 -68 0 V 0 00 CN00 C1 0- 0o W) _ _ _ _W F- u_ u_ _ __ E__ _4 o 0- 00 0-0 ir0 00 00 00-C N C 0 r W CD 0i ~- c c - c o edW ) W n k )c 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0041 IM C44 WO 99/53099 PCTIUS99/07886 -69 0 0 4. 0 ~ ~ +2 O~ W) Cde ~ .o \o C.o I>4 > ___1 C_4 N- _' N- ' c ' WO 99/53099 PCT/US99/07886 -70 0 0 C4 4) .=n cn -c-nU 0 0 -ow 4:3 Z1 CC - WO 99/53099 PCT/US99/07886 -71 The amino acid sequence of Tpa92 is highly conserved, with a range of 95.5 100% identity and 96.8-100% similarity shared between the Nichols Tpa92 sequence and that of the various other T. pallidum strains. However, several of the amino acid sequence changes that do exist are of particular interest. First, parallel sequence 5 divergence is observed between Bal-2 and Sea 81-3 strains and again with Gauthier and Simian strains, thus suggesting a common strain origin for each of these groups. Second, and most importantly, a distinctive sequence deletion pattern is present in the Tpa92 sequences from non-T pallidum subsp. pallidum strains. The tpa92 genes of the Gauthier and Simian strains have base pairs 2336-2350 deleted (data not shown), 10 which corresponds to deletion of the amino acids that comprise the end of the T. pallidum Tpa92 signature serine stretch, residues 780-784. The tpa92 gene sequence of the Cuniculi A strain possesses an additional complexity, in that base pairs 2293-2352, which encode the characteristic serine stretch comprising amino acid residues 765-784, are deleted. This DNA sequence is replaced with 30 base pairs that 15 encode an alternative 10 amino acids that, although serine-rich, represents a minimal serine content compared to that of the same stretch of amino acids in the other T. pallidum strains. All DNA sequence deletions are in-frame and do not introduce premature termination codons into the tpa92 open reading frame. Overexpression of the T. pallidum Tpa92. 20 Heterologous expression of the mature 816 residue T. pallidum Tpa92 open reading frame in E. coli BL21 (DE3) pLysS using the IPTG-inducible pRSETc T7 expression system resulted in production of a recombinant molecule with an approximate molecular mass of 70 kDa, as assayed by SDS-PAGE and subsequent Coomassie blue staining. Expression of the 70 kDa recombinant protein was 25 significantly decreased in E. coli lysates in which protein expression from the pRSETc T7 promoter had not been induced by IPTG addition. The 70 kDa molecular mass of the recombinant protein is unexpectedly lower than the 97 kDa molecular mass predicted for the histidine-tagged recombinant molecule (92 kDa for the T. pallidum Tpa92 plus 5 kDa extra for the N-terminal hexa-histidine tag). This low molecular 30 mass is not the result of truncated expression of the tpa92 open reading frame, as sequencing of the tpa92-pRSETc construct verified the entire 2451 bp insert encoding the 816 residue open reading frame was present, but likely represents sequence induced aberrant migration of the recombinant molecule on SDS-PAGE. Nickel resin chromatography performed on E. coli expressing the Tpa92-pRSETc construct 35 allowed purification of the histidine-tagged recombinant molecule away from WO 99/53099 PCT/US99/07886 -72 contaminating E. coli proteins. The recombinant 70 kDa molecule represented the major protein in the resulting preparation (approximately 90% of the total protein). Proteins of a smaller molecular mass present in the nickel-purified preparation represent breakdown products of the 70 kDa recombinant Tpa92. 5 The recombinant T. pallidum Tpa92 was used to generate polyclonal antiserum, and subsequent immunoblot analysis showed an immunoreactive 70 kDa protein in both the nickel-purified recombinant protein preparation and lysates of E. coli expressing the Tpa92-pRSETc construct. No corresponding immunoreactive protein was observed using either control pre-immune serum on the nickel-purified 10 recombinant protein preparation or the anti-Tpa92 antiserum on preparations of E. coli expressing the pRSETc vector alone. Characterization of Anti-Tpa92 Imunoreactivity on T. pallidum Lysates. The level of reactivity of the anti-Tpa92 polyclonal antiserum on lysates of washed and unwashed T. pallidum preparations was investigated by immunoblot 15 analysis. An immunoreactive band corresponding to the 92 kDa T. pallidum Tpa92 was present in lysates of unwashed treponemes extracted directly from infected rabbit testes. In contrast, no immunoreactive 92 kDa bands were observed in equal quantities of lysates prepared from T. pallidum washed one time and three times following extraction from rabbit testes, or in lysates of unwashed treponemes using 20 control pre-immune serum. Previous investigations have demonstrated that the fragile outer membrane is partially removed during washing of T. pallidum by centrifugation (Cox, D. L. et al., Mol. Microbiol. 15:1151-1164 (1995)), and thus the above results suggest concurrent loss of anti-Tpa92 immunoreactivity, and therefore loss of Tpa92 itself, with the treponeme outer membrane during washing. 25 Opsonic Potential of the T. pallidum Tpa92. The anti-Tpa92 antiserum was also investigated for its ability to opsonize T. pallidum in three separate experiments using a standard phagocytosis assay. The anti-Tpa92 polyclonal antiserum was significantly opsonic for the Nichols strain of T. pallidum, as compared with control pre-immune serum (p=0.0089). The level of 30 opsonic activity observed for anti-Tpa92 approximated that observed with serum collected from rabbits chronically infected with T. pallidum (immune rabbit serum; p<0.0001). Immunoprotective Capacity of T. pallidum Tpa92. The protection afforded by immunization with the T. pallidum Tpa92 was 35 tested in the rabbit syphilis model. In these experiments, four rabbits were immunized WO 99/53099 PCT/US99/07886 -73 three times each with the purified recombinant Tpa92 emulsified in Ribi adjuvant. Rabbit #5061 and #5200 demonstrated approximately equal levels of immunoreactivity against the recombinant Tpa92, while rabbit #5202 showed slightly less anti-Tpa92 immunoreactivity and rabbit #5207 demonstrated no detectable 5 reactivity. No immunoreactivity was observed using control pre-immune sera collected from each of the rabbits prior to immunization. Three weeks following administration of the final immunization, rabbits were intradermally challenged at eight independent sites with 105 T. pallidum per site. Two control rabbits received no prior immunization but underwent the same 10 intradermal challenge. Table 9 summarizes the post-challenge analyses performed on the rabbits to determine the degree of protection provided by immunization with the T. pallidum recombinant Tpa92.

WO 99/53099 PCTIUS99/07886 -74 4C4 00 H Cfd 4-4) 04 0 0~ 4 Nc cc ) cc cc cc V) P 0 4) 0 WO 99/53099 PCT/US99/07886 -75 As shown in the table, the control animals developed typical red, raised and highly-indurated lesions, the majority of which progressed to ulceration. In contrast, the rabbits immunized with the T. pallidum recombinant Tpa92 prior to challenge all demonstrated alteration of lesion development. However, the degree of protection 5 varied amongst the immunized rabbits, with the highest levels of protection observed for those rabbits exhibiting strong anti-Tpa92 immunoreactivity in immunoblot analysis. Significant attenuation of lesion development was observed in rabbits #5061 and #5200, with atypical pale, flat, slightly-indurated and non-ulcerative lesions appearing at the sites of challenge. The lesions of rabbit #5202 also were 10 morphologically atypical, although two of the eight challenge sites progressed to ulceration. This value, however, still represents a statistically significant difference from the occurrence of ulceration in the control unimmunized animals (p=0.0047), and thus these lesions received an atypical designation. In contrast, although rabbit #5207 developed lesions that were paler, flatter and less indurated than those of the 15 control rabbits, all lesions progressed to ulceration and therefore were designated as typical. The results of darkfield microscopy examination of the challenge sites performed 19 days following infection paralleled the observed range of clinical manifestations of lesion development in the challenged rabbits. Analysis of the control 20 unimmunized rabbits (#5111 and #5228) revealed treponemes in all eight challenge sites. Similarly, analysis of the Tpa92-immunized rabbits #5202 and #5207 showed the presence of treponemes in six out of eight lesions. In contrast, the Tpa92 immunized rabbits that demonstrated the most impressive clinical alteration in lesion development, #5061 and #5200, had significantly lower numbers of lesions containing 25 treponemes (one and three out of eight, respectively). Serological examination of the rabbits four weeks post-challenge revealed a high VDRL and FTA-ABS test titer for normal, unimmunized animals compared to significantly reduced titers observed for the Tpa92-immunized rabbits (P<). Parallel experiments revealed that immunization with an unrelated, non-treponemal recombinant molecule in Ribi adjuvant provided no 30 protection (data not shown), thus demonstrating that the adjuvant did not contribute to the protection observed in the Tpa92-immunized rabbits. The data reported herein describes the identification and characterization of a 92 kDa T. pallidum protein that shares sequence similarity with outer membrane proteins from a wide range of bacterial species, including the related spirochete B. 35 burgdorferi and two STD-causing bacterial species, N. gonorrhoeae and WO 99/53099 PCT/US99/07886 -76 C. trachomatis. Although the majority of these proteins have been identified through genome sequencing of the bacterial species in which they are found, and thus are hypothetical, six have been independently isolated using molecular biological or protein immunochemical approaches. These include an unknown protein from E. coli 5 (genbank accession number P39170), OMP1 from B. abortus (genbank accession number U51683), Omp85 proteins from N meningitidis and N. gonorrhoeae (Manning, D.S., D.K. Reschke, and R.C. Judd. 1998. Omp85 proteins of Neisseria gonorrhoeae and Neisseria meningitidis are similar to Haemophilus influenzae D-15 Ag and Pasteurella multocida Oma87. Microb. Pathog. 25:11-21), Oma87 from P. 10 multocida (Ruffolo, C.G., and Adler, B. 1996. Cloning, sequencing, expression, and protective capacity of the oma87 gene encoding the Pasteurella multocida 87 kilodalton outer membrane antigen. Infect. Immun. 64:3161-3167) and D15 from H. influenzae (Flack, F.S., S. Loosmore, P. Chong, and W.R. Thomas. 1995. The sequencing of the 80-kDa D15 protective surface antigen of Haemophilus influenzae. 15 Gene 156:97-99). Characterization of the latter four proteins confirms they are present on the bacterial surface (Ruffolo, C.G., and Adler, B. 1996. Cloning, sequencing, expression, and protective capacity of the oma87 gene encoding the Pasteurella multocida 87-kilodalton outer membrane antigen. Infect. Immun. 64:3161-3167; Manning, D.S., D.K. Reschke, and R.C. Judd. 1998. Omp85 proteins 20 of Neisseria gonorrhoeae and Neisseria meningitidis are similar to Haemophilus influenzae D-15-Ag and Pasteurella multocida Oma87. Microb. Pathog. 25:11-21.; Thomas, W.R., M.G. Callow, R.J. Dilworth, and A.A. Audesho. 1990. Expression in Escherichia coli of a high-molecular weight protective surface antigen found in nontypeable and type b Haemophilus influenzae. Infect. Immun. 58:1090-1913), and 25 passive immunization of antiserum against Oma87 and D15 has been shown in animal models to be protective against P. multocida and H. influenzae challenge, respectively (Ruffolo, C.G., and Adler, B. 1996. Cloning, sequencing, expression, and protective capacity of the oma87 gene encoding the Pasteurella multocida 87-kilodalton outer membrane antigen. Infect. Immun. 64:3161-3167; Thomas, W.R., M.G. Callow, R.J. 30 Dilworth, and A.A. Audesho. 1990. Expression in Escherichia coli of a high molecular weight protective surface antigen found in nontypeable and type b Haemophilus influenzae. Infect. Immun. 58:1090-1913; Yang, Y., W.R. Thomas, P. Chong, S.M. Loosmore, and M.H. Klein. 1998. A 20-kilodalton N-terminal fragment of the D15 protein contains a protective epitope(s) against Haemophilus influenzae 35 type a and type b. Infect. Immun. 66:3349-3354; and 32.Loosmore, S.M., Y. Yang, WO 99/53099 PCT/US99/07886 -77 D.C. Coleman, J.M. Shortreed, D.M. England, and M.H. Klein. 1997. Outer membrane protein D15 is conserved among Haemophilus influenzae species and may represent a universal protective antigen against invasive disease. Infect. Immun. 65:3701-3707.). Results reported here suggest that Tpa92 is a similar protective 5 outer membrane antigen of T pallidum. Evidence for the surface location of Tpa92 in T. pallidum comes from the observation that antibodies directed against Tpa92 have significant opsonic activity for living T. pallidum, thus demonstrating that this protein is accessible on the surface of intact treponemes. Indirect evidence for the presence of Tpa92 in T. pallidum 10 outer membranes was obtained by immunoblot analysis using the anti-Tpa92 antiserum on T. pallidum lysate preparations. A loss of immunoreactivity was observed in lysates prepared from treponemes whose outer membranes had been partially removed by washing prior to lysis, compared to lysates prepared from unwashed treponemes in which the fragile outer membrane and its constituent 15 proteins remain intact prior to lysis. Analysis of the amino acid sequence of Tpa92 also provides supporting evidence for the presence of Tpa92 on the bacterial surface. The first 21 amino acid residues at the N-terminus of Tpa92 comprise a cleavable signal sequence that is characteristic of proteins translocated across the bacterial inner membrane (Von Heijne, G. 1983. Patterns of amino acids near signal-sequence 20 cleavage sites. Eur. .1 Immunol. 133:17-21). In addition, analysis reveals the C-terminus of Tpa92 is not a perfect match for the consensus hydrophobicity pattern predicted for bacterial outer membrane proteins of hydrophobic residues at positions 1 (Phe), 3 (preferentially Tyr), 5, 7 and 9 from the C-terminus (Struyve, M., M. Moons, and J. Tommassen. 1991. Carboxy-terminal phenylalanine is essential for 25 the correct assembly of a bacterial outer membrane protein. J Mol. Biol. 218:141 148.), but does contain hydrophobic residues at positions 1, 5 and 7 from the C terminus and thus loosely conforms to this pattern. Furthermore, PSORT analysis predicts an 84.6% probability that Tpa92 resides in the T. pallidum outer membrane, and the TMPred program identified three potential transmembrane helices within the 30 Tpa92 amino acid sequence. These combined results suggest Tpa92 is associated with the T. pallidum outer membrane, and additional biochemical studies are currently underway to investigate the potential cell surface disposition of this molecule. PCR amplification and subsequent sequence analysis of the Tpa92 open reading frame from twelve T. pallidum strains revealed minimal amino acid sequence 35 divergence between the various strains. Similarly, the D15 antigen is conserved WO 99/53099 PCT/US99/07886 -78 among H. influenzae strains and thus also represents an invariant antigen (Loosmore, S.M., Y. Yang, D.C. Coleman, J.M. Shortreed, D.M. England, and M.H. Klein. 1997. Outer membrane protein D15 is conserved among Haemophilus influenzae species and may represent a universal protective antigen against invasive disease. Infect. 5 Immun. 65:3701-3707). Of the divergence that does occur in the Tpa92 sequence, the majority is found in non-T. pallidum subsp. pallidum strains and lies within a serine-rich sequence that is unique to Tpa92. The C-terminal end of this seine stretch is deleted in the Tpa92 sequences from both the Simian strain and the T. pallidum subsp. pertenue Gauthier strain. Surprisingly, this sequence is not deleted in the 10 Tpa92 sequence from the Haiti B strain, suggesting its classification as a T. pallidum subsp. pertenue strain is a misnomer. Similar sequence analyses performed on other T. pallidum antigens, including glycerophosphodiester phosphodiesterase (C.E. Cameron, unpublished observations) and Tpr K (A. Centurion-Lara, unpublished observations), also suggest the Haiti B strain should be re-classified as a T. pallidum 15 subsp. pallidum strain. It is interesting to note that the entire C-terminal serine-rich sequence has been deleted from the Tpa92 sequence of the rabbit-infective T. pallidum subsp. paraluiscuniculi strain Cuniculi A, although the relevance of this sequence divergence is not known at this time. The potential significance of the serine-rich sequence present in Tpa92 20 becomes apparent when one considers similar serine-rich sequence stretches are observed in proteins involved in attachment to cells or cellular substances, including the Saccharomyces cerevisiae A-agglutinin attachment subunit precursor (Roy, A. et al., Mol. Cell. Biol. 11:4196-4206 (1991)) and the Candida albicans chitinase 3 precursor (McCreath, K.J. et al., Proc. Natl. Acad. Sci. USA. 92:2544-2548 (1995)). 25 Numerous studies have shown IT. pallidum attaches to host cells (Fitzgerald, T.J., J.N. Miller, and J.A. Sykes. Infect. Immun. 11:1133-1140 (1975); Fitzgerald, T.J. et al., Infect. Immun. 18:467-478 (1977); Hayes, N.S. et al., Infect. Immun. 17:174-186 (1977); Baseman, J.B., and E.C. Hayes., J. Exp. Med. 151:573-586 (1980); Baseman, J.B., and J.F. Alderete, Pathogenesis and immunology of Treponema infections, Vol. 30 20., (1983), R. Schell and D. Musher, editors. Marcel Dekke, Inc., New York. 229 239; Wong, G.H.W., B. Steiner, and S. Graves., Br. J. Vener. Dis. 59:220-224 (1983); Fitzgerald, T.J. et al., Br. J Vener. Dis. 60:357-363 (1984); Rice, M., and T.J. Fitzgerald, Can. J. Microbiol. 31:62-67 (1984); Thomas, D.D. et al., J. Exp. Med. 161:514-525 (1985); Thomas, D.D. et al., Proc. Natl. Acad. Sci. USA. 85:3608 35 3612 (1988)), although the T. pallidum proteins mediating such attachment have not WO 99/53099 PCT/US99/07886 -79 yet been identified. As a putative outer membrane protein, Tpa92 could be hypothesized to constitute one such attachment ligand. In this scenario, the stretch of serine residues present in the C-terminal end of the Tpa92 sequence, which have been predicted to reside within an external loop on the outer face of the outer membrane, 5 could act as potential sites for hydrogen bonding to carbohydrates present on the surface of host cells. In support of this, preliminary investigations conducted in our laboratory show Tpa92-specific antiserum can inhibit T. pallidum attachment to rabbit epithelial cells (E.S. Sun, unpublished observations). Studies are currently underway to further investigate this putative functional role of Tpa92 as a T. pallidum adhesion. 10 The immunoprotective potential of the T. pallidum Tpa92 was also investigated in this study for several reasons. First, antiserum raised against the analogous proteins Oma87 and D15 from P. multocida and H. influenzae, respectively, have been shown to induce protection in animal models (Ruffolo, C.G., and Adler, B., Infect. Immun. 64:3161-3167 (1996); Thomas, W.R. et al., Infect. 15 Immun. 58:1090-1913 (1990); Yang, Y., W.R. et al., Infect. Immun. 66:3349-3354 (1998); Loosmore, S.M. et al., Infect. Immun. 65:3701-3707 (1997)). Second, the invariant nature of Tpa92 among various T. pallidum subspecies and strains makes it an attractive candidate for design of a universal subunit vaccine against T. pallidum infections. And lastly, the identification of Tpa92 as a target of opsonic antibodies, 20 through both the differential immunologic expression library screen and the phagocytosis assays, combined with the central role opsonization and phagocytosis plays in bacterial clearance, suggests this antigen may have immunoprotective capability. Indeed, immunization of rabbits with the T. pallidum Tpa92 resulted in partial protection from subsequent T. pallidum challenge, with alteration of lesion 25 development at the sites of challenge compared to unimmunized control rabbits. Not surprisingly, the level of protection achieved strongly corresponded to the antibody response generated in the immunized rabbit, with rabbits exhibiting the highest level of Tpa92-specific immunoreactivity demonstrating significant protection upon challenge. These rabbits developed atypical small, pale, fiat, slightly indurated and non-ulcerative 30 lesions at the sites of challenge. Darkfield examination of aspirates collected from the sites of challenge in these rabbits showed a lower number of lesions contained viable treponemes compared to control unimmunized animals. Alternative methods of antigen delivery will be investigated in an attempt to generate higher levels of anti Tpa92 reactivity and, correspondingly, more significant protection against T. pallidum 35 challenge.

WO 99/53099 PCT/US99/07886 -80 In summary, the T. pallidum Tpa92 represents a target of opsonic antibodies and an invariant, immunoprotective antigen. Further studies will be performed to determine whether co-vaccination of Tpa92 with other promising immunoprotective antigens, such as glycerophosphodiester phosphodiesterase (Cameron, C.E., et al., 5 Infect. Immun. 66:5763-5770 (1998)) and Tpr K (Centurion-Lara, A., C., et al., J. Exp. Med In Press (1999)), can achieve complete immunity against T. pallidum challenge. Example 13. DNA-mediated vaccination with a vector expressing Gpd is partially protective against challenge with T. palladium. 10 We have constructed a Gpd DNA vaccine based on the high-expression CMV promoter vector pCR3.1 (Invitrogen, San Diego, CA) expressing Gpd. We have shown that the rabbit epithelial cell line, Sf-1 Ep (American Type Culture Collection), transfected with pCR3.1-Gpd expresses Gpd detectable by Western blot (data not shown). Intradermal and intramuscular immunization of rabbits with 200 gg of 15 pCR3.1-Gpd performed every three weeks led to easily detectable antibodies to Gpd after 3 injections in 2 of 2 rabbits (data not shown). Twenty one days after the 4th injection of DNA, the two DNA-injected rabbits and one control (uninjected) rabbit were challenged intradermally with 105 T. palladium Nichols strain at each of eight separate sites on their shaved backs. Unlike the control rabbit, which developed 20 progressive large chancres that ulcerated by 28 days, the DNA-injected rabbits developed only small papules at the sites of challenge which cleared before the control rabbit developed ulcerated lesions. Darkfield examination of aspirates from the challenge sites on day 21 after challenge demonstrated that 8 of 8 (100%) lesions on the control rabbit contained 25 treponemes but only 2 of 16 (12.5%) of the challenge sites of the two DNA-injected rabbits had demonstrable treponemes. Both DNA-injected rabbits seroconverted by day 36 after challenge and were judged infected. Thus, although the DNA-injected rabbits did become infected, lesion development in these animals was drastically altered and treponeme growth was limited. These results demonstrate that DNA 30 vaccination with a vector expressing Gpd was partially protective against T. palladium challenge. This is the first time that DNA vaccination has been shown to be protective against challenge with T. pallidum. This mechanism of immunization could be advantageous because DNA vaccination stimulates humoral, CD4 and CD8 immunity. 35 Both CD4 and CD8 cells have been found in the primary and secondary syphilis WO 99/53099 PCT/US99/07886 -81 lesions at the time of treponemal clearance and are probably responsible with production of IFN-y necessary for the activation of macrophages. The results confirm that Gpd is a protective immunogen against challenge with T. palladium using both standard and alternative vaccination approaches. 5 Example 14. Protection Studies Using Recombinant Msp Peptides The variable domains of the msp-homologues have been expressed in E. coli as 6 his-fusion proteins, purified and used to immunize rabbits before intradermal challenge with 105 T. pallidum per site. Table 10 describes immunizaion of single animals with recombinant variable domains from msp 3, 4/5, 6, 10, and 12; of these, 10 msp 4/5 showed evidence of protection, as measured by lesion appearance and lack of treponemes on darkfield microscopy of lesion aspirates. In addition, the recombinant carboxyl-terminal conserved domain from Subfamily II appears to confer significant protection.

WO 99/53099 PCTIUS99/07886 -82 0

-

0 914 00 008 - 00 00 ' -~ 00~ 00~ 000 1--,0 0 00 00 0 00 040 000 00 00P o F-4 0 00 ceU P4~~; 4 ;- :3 WO 99/53099 PCT/US99/07886 -83 Because of the expense, we have chosen to immunize and challenge single animals with each of our recombinant peptides as a screening procedure. Those antigens that appear to be protective are then examined using larger groups of animals. For example, msp 9 appeared to be protective in the first animal tested, so 5 we immunized and challenged a group of four additional rabbits, along with four unimmunized controls. The composite results for msp 9 are also shown in Table 10, indicating that msp 9 variable domain induces significant protection against infectious challenge with 10 5 T. pallidum, Nichols strain. Example 15 Opsonization of T. pallidum Nichols Strain by Antisera to Msp 10 Homologue Variable Domains Opsonization data for antisera raised against recombinant variable domains of msp 1, 9, 11, and 2/1 have already been provided. Antisera raised against recombinant variable domains of msp 3, 4/5, 6, and 12 have now been tested. Only antisera to msp 4/5 and 12 have statistically significant opsonic activity against the Nichols strain 15 (p=0.02 and p=0.0 5 , respectively) compared to NRS, but the levels of phagocytosis with these antisera are lower than with IRS and lower than previously seen with antisera to msp 1, 9, 11, and 2/1. These results suggest that several msp-homologues are expressed on the surface of T. pallidum, or on subpopulations of organisms within the Nichols strain suspension, but that the level of expression in the individual cell or 20 in the population may be lower for msp 4/5 and 12 than for the msps tested previously. The failure of anti-msp 3 and 6 to opsonize T. pallidum suggests that these molecules are not expressed on the surface of the target orgamnism. Example 16 Heterogeneity in Msp 9 (TprK) Among Strains ofT. pallidum Msp 9 (tpr K) is the gene that is preferentially transcribed and expressed in the 25 Nichols strain (laboratory strain) of T. pallidum. It codes for the msp antigen that is most protective in our studies. To examine its structure in other strains, an issue that is highly relevant to its ability to confer broad protection in a natural setting, we amplified msp 9 genomic DNA in a number of strains from our T. pallidum strain bank. The gene could be amplified in all strains tested, but the amplicons showed 30 significant variability in size compared to the Nichols strain (from which the primer sequences were derived). In addition, many strains had multiple amplicons using these primers. We are currently exploring msp 9 heterogeneity in other strains by cloning and sequencing the amplicons from selected strains. All strains other than the Nichols 35 strain contain multiple alleles of tpr K, and their sequences differ from the published WO 99/53099 PCT/US99/07886 -84 tpr K sequence. The sequence differences are limited to defined "hypervariable" regions. Given the nature of the sequence diversity, it is highly unlikely that these differences are due to PCR-induced errors. It is particularly interesting that this heterogeneity is seen in msp 9, which is a protective and opsonic antigen in the 5 Nichols strain, and is the msp-homologue that is predominantly transcribed and expressed. The amino acid sequences of tpr K hypervariable regions from 34 different T. pallidum strains are set forth in: SEQ ID NO:76 (strain [N); SEQ ID NO:77 (strain 1-n); SEQ ID NO:78 (strain 1-1-Bal2); SEQ ID NO:79 (strain 2-1-Bal2); SEQ 10 ID NO:80 (strain 1-1-Bal3); SEQ ID NO:81 (strain 1-1-Bal7); SEQ ID NO:82 (strain 1-2-Bal7); SEQ ID NO:83 (strain 2-3-Bal7); SEQ ID NO:84 (strain 1-1-Bal8); SEQ ID NO:85 (strain 1-2-Bal8); SEQ ID NO:86 (strain 1-3-Bal8); SEQ ID NO:87 (strain 1-1-Bal73-1); SEQ ID NO:88 (strain 1-2-Bal73-1); SEQ ID NO:89 (strain 1-3 Bal73-1); SEQ ID NO:90 (strain 2-1-Bal73-1); SEQ ID NO:91 (strain 1-2-sea81-3); 15 SEQ ID NO:92 (strain 1-3-sea81-3); SEQ ID NO:93 (strain 1-1-sea81-4); SEQ ID NO:94 (strain 1-2-sea81-4); SEQ ID NO:95 (strain 1-3-sea81-4); SEQ ID NO:96 (strain 2-1-sea81-4); SEQ ID NO:97 (strain 1-1-sea84-2); SEQ ID NO:98 (strain 1-2 sea84-2); SEQ ID NO:99 (strain 1-3-sea84-2); SEQ ID NO:100 (strain 1-1-h); SEQ ID NO:101 (strain 1-2-h); SEQ ID NO:102 (strain 1-4-h); SEQ ID NO:103 (strain 2 20 1-h); SEQ ID NO:104 (strain 2-2-h); SEQ ID NO:105 (strain 1-1-ch); SEQ ID NO:106 (strain 1-2-ch); SEQ ID NO:107 (strain 1-3-ch); SEQ ID NO:108 (strain 1 4-ch); SEQ ID NO:109 (strain 1-5-ch). Example 17 Identification of a New Msp Homologue in Some T. pallidum Strains 25 We have identified a new msp-homologue in approximately 50% of T. pallidum subsp. pallidum strains. Primers targeted to conserved regions of Subfamilies I and II were used to amplify DNA from these two strains, the products were cloned, and inserts were sequenced. A new sequence, called msp 13 or tpr M (SEQ ID NO:110), was identified. All sequences were identical and this sequence is 30 not found in the Nichols genome. Primers, specific for msp 13 (SEQ ID NO: 110), were then designed: sense 5' cactagtcttggggacacgc (SEQ ID NO:111); antisense 5' tacgtgattgcaaccagga (SEQ ID NO: 112). Msp 13 appears to be most closely related to msp 4/5 (tpr C/D) in Subfamily II.

WO 99/53099 PCT/US99/07886 -85 While the preferred embodiment of the invention has been illustrated and described, it will be appreciated that various changes can be made therein without departing from the spirit and scope of the invention.

Claims

1. An isolated nucleic acid molecule that encodes a polypeptide having an amino acid sequence selected from the group consisting of the amino acid sequences of SEQ ID NOS:2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 21, 23, 25, 27, 29, 31, 32, 35, 37, 39, 41, 44 and 46.

2. An isolated nucleic acid molecule according to Claim 1, selected from the group consisting of the nucleotide sequences of SEQ ID NOS: 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 22, 24, 26, 28, 30, 33, 34, 36, 38, 40, 43 and 45.

3. A pair of PCR primers comprising two nucleic acid molecules selected from the group consisting of one of the following sets of two nucleic acid molecules: Set 1: 5'-CGACTCACCCTCGAACCA-3' (SEQ ID NO:48) (sense) 5'-GGTGAGCAGGTGGGTGTAG-3' (SEQ ID NO:49) (antisense) Set 2 5'-CGCGTTTGACGCTTTCCCCG-3' (SEQ ID NO:50) (sense) 5'-ACACAAGCTTAGAAAGAGAATCG-3' (SEQ ID NO:51) (antisense) Set 3 5'-CTTITTCTCGCTGACGCTTTGT-3' (SEQ ID NO:52) (sense) 5'-TGCAAGGCATGGGTGTAATCA-3' (SEQ ID NO:53) (antisense) Set 4 5'-CGGCTGACGCTGACCCCG-3' (SEQ ID NO:54) (sense) 5'-CAAGTAGTCTGTAAGCTGCCTG-3' (SEQ ID NO:55) (antisense) Set 5 5'-ATATTGAAGGCTATGCGGAGCTG-3' (SEQ ID NO:56) (sense) 5'-CCTCAAGGAAAGAAGTATCAGG-3' (SEQ ID NO:57) (antisense) Set 6 5'-CGCGCATAACGCTCACTCC-3' (SEQ ID NO:58) (sense) 5'-GTCTATAAGGTGTGTATACGCG-3' (SEQ ID NO:59) (antisense) Set 7 5'-ACCAGTCCTICCTGTGTGGTTAA-3' (SEQ ID NO:60) (sense) 5'-ACTCCTTGGTTAGATAGGTAGCTC-3' (SEQ ID NO:61) (antisense)

4. An isolated polypeptide encoded by a nucleic acid molecule according to Claim 1, and functional equivalents thereof WO 99/53099 PCT/US99/07886 -87

5. An isolated polypeptide capable of inducing a protective immunologic response to T. p. pallidum, T. p. pertenue, or T. p. endemicum when administered in an effective amount to an animal host.

6. A vaccine comprising: an effective amount of an isolated polypeptide according to Claim 5; and a physiologically acceptable carrier.

7. A vaccine according to Claim 6, wherein the isolated polypeptide is a polypeptide of Claim 4.

8. A method of identifying a T. p. pallidum vaccine candidate comprising the steps: identifying a T. p. pallidum protein that is immunologically reactive with an opsonizing serum against T. p. pallidum but that is immunologically unreactive with a non-opsonizing serum against T. p. pallidum; testing said protein to determine whether it is capable of eliciting in an animal host an immune response that is protective against challenge with T. p. pallidum; and determining that said protein is a vaccine candidate if test results indicate that it elicits in said host an immune response that is protective against challenge with T. p. pallidum.

9. A vaccine comprising: a vaccine candidate identified according to the method of Claim 8; and a physiologically acceptable carrier.

10. A method of identifying a T. pallidum vaccine candidate comprising the steps: identifying a protein that is expressed by a gene that is present in the genome of T. pallidum but that is not present in the genome of Treponemaparaluiscuniculi; testing said protein to determine whether it is capable of eliciting in a suitable host an immune response that is protective against challenge with T. pallidum; and determining that said protein is a vaccine candidate if test results indicate that it elicits in said host an immune response that is protective against challenge with T. pallidum.

11. A vaccine comprising: WO 99/53099 PCT/US99/07886 -88 a vaccine candidate of Claim 10; and a physiologically acceptable carrier.

12. The vaccine of Claim 11, wherein the vaccine candidate is capable of eliciting in a suitable host an immune response that is protective against challenge with T. p. pallidum, T. p. pertenue, and T. p. endemicum.

13. A vaccine according to Claim 7, which comprises at least two different T. p. pallidum Msp polypeptides.

14. A vaccine according to Claim 7, which comprises: an isolated T.p. pallidum glycerophosphodiester phosphodiesterase polypeptide; and an effective amount of at least one isolated T. p. pallidum Msp polypeptide.

15. A vaccine according to Claim 14, which further comprises an effective amount of an isolated T. p. pallidum D15/Oma87 homologue.

16. A method of inducing a protective immune response against T. pallidum comprising administering to a host a vaccine of Claim 6.

17. A method of inducing a protective immune response against T. pallidum comprising administering to a host a vaccine of Claim 9.

18. A method of inducing a protective immune response against T. pallidum comprising administering to a host a vaccine of Claim 11.

19. A method for analyzing a sample of DNA to determine whether it originated from T. p. subspecies pallidum, T. p. subspecies pertenue or T. p. subspecies endemicum, comprising the steps: amplifying the DNA with the PCR sense primers 5'-ACCAGTCCTTCCTGTGTGGTTAA-3' (SEQ ID NO:60) (sense) and 5'-ACTCCTTGGTTAGATAGGTAGCTC-3' (SEQ ID NO:61) (antisense); analyzing the size of the resulting DNA fragments; and determining that the DNA originated from T. p. pallidum if a single DNA fragment having a size of about 1.7 kb is detected, and that the DNA originated from T. p. subspecies pertenue if at least two DNA fragments are detected, one having a WO 99/53099 PCT/US99/07886 -89 size of about 1.7 and the other having a size of about 1.3 kb, and that the DNA originated from T. p. subspecies endemicum if no DNA fragment is detected.

20. A method of determining whether a first and a second clinical isolate of T. p. pallidum are the same or different comprising: amplifying a first sample of genomic DNA from the first clinical isolate and a second sample of DNA from the second isolate using the PCR primers 5'-CGACTCACCCTCGAACCA-3' (SEQ ID NO:48) (sense), and 5'-GGTGAGCAGGTGGGTGTAG-3' (SEQ ID NO:49) (anti-sense); digesting the first and the second amplified DNAs with a restriction endonuclease that recognizes a four-base cleavage site; analyzing the size of the DNA products in the first and the second amplified DNAs; and determining that the first and the second clinical isolates are different if the DNA fragments produced by restriction digestion of the first and the second DNA samples are not the same.

21. The method of Claim 20, wherein the genomic DNA of the first and second clinical isolates are digested with the restriction endonucleases BstUI, AluI, HhaI and NlaIII.