WO2007016938A2

WO2007016938A2 - Methods for diagnosing colon cancer and streptococcus bovis proteins for use in the method

Info

Publication number: WO2007016938A2
Application number: PCT/EP2005/008322
Authority: WO
Inventors: Harold Tjalsma
Original assignee: Stichting Katholieke Universiteit
Priority date: 2005-07-27
Filing date: 2005-07-27
Publication date: 2007-02-15
Also published as: WO2007016938A3

Abstract

The present invention relates to methods for diagnosing colon cancer in an individual, especially the diagnosis of an pre or early stage of colon cancer, comprising detecting the presence of immunoglobulin molecules directed against Streptococcus bovis antigens in a sample of the individual comprising total serum immunoglobulin molecules. The present invention further relates to antigens for use in said method and to the use of the antigens as a diagnosticum and/or medicament. Preferred antigens are HlpA and RpL7/L12. The most preferred detection method involves capturing the S. Bovis antigens on a solid support, to which the immunoglobulin analytes have been previously immobilized.

Description

METHODS FOR DIAGNOSING COLON CANCER AND STREPTOCOCCUS BOVIS PROTEINS FOR USE IN THE METHOD.

The present invention relates to methods for diagnosing colon cancer in an individual and especially to methods for diagnosing colon cancer in a pre or early stage of colon cancer. The present invention further relates to antigens for use in said methods and to the use of the antigens as a diagnosticum and/or medicament. The incidence of colon cancer or colorectal cancer is highest in developed countries such as Europe, the United States and Japan, and lowest in developing countries in Africa and Asia. According to the American Cancer Society, it is the third most common type of cancer in both men and women in the United States. The incidence is slightly higher in men than women, and is highest in African American men.

The American Cancer Society estimates that about 145,000 cases of colon cancer will be diagnosed and about 56,000 people will die from the disease in 2005. The death rate from colon cancer cancer has declined over the past 15 years due to improved screening methods and advances in treatment .

Colon cancer has a long asymptomatic period, and when detected during this early stage, the 5-year survival of colon cancer patients is about 90% after surgical resection of the tumor. However, when the disease is discovered in a late stage (when spread to other organs) this number drops t -10%. Unfortunately, only 40% of the cases of colon cancer is detected at an early stage (Etzioni et al., 2003). Therefore, early detection of colon cancer, or identification of individuals at risk, is one of the great challenges in the battle against this disease. The present only available non-invasive method for detecting the presence of colon tumors is the fecal occult blood test or FOB test. The test involves the detection of minor amounts of blood, undetectable by the naked eye, in a fecal sample of an individual.

A disadvantage of this test is its sensitivity of about 70%, which means that many polyps (generally regarded as a pre-malignant stage of cancer) , and cancers are not detected. This is mainly due to the fact that many colon cancers do not cause blood to be released into the feces in an early stage. Furthermore, the test is not very specific and many people who test positive have to undergo the discomfort and risk of full bowel investigation without the actual presence of a polyp or tumor (Ahlquist, 1997) .

It is therefore an object of the present invention to provide improved methods for the detection of colon cancer, especially with respect to sensitivity, reliability, and early detection of colon cancer. This object is achieved with the present invention by providing a method for diagnosing colon cancer in an individual comprising detecting the presence of immunoglobulin molecules directed against Streptococcus bovis antigens in a sample of the individual comprising total serum immunoglobulin molecules such as for example serum itself.

The term "Streptococcus bovis" as used herein comprises bacteria which a skilled medical microbiologist would classify as a Streptococcus bovis, such as for example phylogenetic or genetically closely related species to Streptococcus bovis strain NCTC 8133. It is noteworthy to mention that Streptococcus bovis is recently reclassified as Streptococcus infantarius subsp. infantarius (Schlegel et al., 2003). Therefore, in this specification the terms "Streptococcus bovis" and "Streptococcus infantarius subsp. infantarius" are interchangeable.

The sensitivity of the above method according to the present invention is about 88% or higher meaning that the presence of colon tumors and/or the premalignant stage thereof, i.e., polyps, is detected in 88 of the 100 cases. This is an significant improvement over the about 70%, i.e., 70 of the 100 cases, when using the fecal occult blood test (FOB test) . The improvement of the method according to the present invention over the prior art is more striking when only the detection of colon tumors is taken into account, i.e, about 92% or higher.

Further, no false positive test results were observed using the method according to the present invention suggesting a very high reliability of the method. This reduces or eliminates the number of false-positive cases and the subsequent discomfortable and risky full bowel investigations without the actual presence of a polyp or tumor.

As already mentioned, the present commonly used method for the detection of colon tumors is the fecal occult blood test (FOB test) . This test is based on the detection of minor bleeding of the intestinal tract due to the presence of colon tumors. However, such detectable bleedings will only occur in a relatively late stage of the tumor development. Since the method according to the present invention allows the detection of the premalignant stage of these tumors, i.e., polyps and/or developing tumors which in an early stage do not cause bleedings, the present invention also provides means for the detection of colon cancer at a relatively- earlier stage. The strong correlation between the presence of immunoglobulin molecules directed against Streptococcus bovis antigens and the presence of colon tumors and/or polyps is highly surprising. The human colon is the natural habitat for a large and dynamic bacterial community, which is essential for the control of intestinal epithelial homeostasis and human health (Rakoff-Nahoum et al . , 2004). However, colon gut flora might also be an essential factor in certain diseases, including multisystem organ failure, inflammatory bowel diseases, and colon cancer (Guarner and Malageleda, 2003) .

Although bacterial infections were originally not considered to be a major cause of cancer, accumulating evidence suggests that bacteria can induce or promote cancer by causing inflammation. In this model, tumor formation is caused or promoted by induction of cell proliferation, and the production of mutagenic free radicals and n-nitroso compounds. In this respect, Helicobacter pylori has been the first invasive bacterium to be identified as a definite cause of gastric cancer (Parsonnet, 1995) .

Similarly, an association between Streptococcus bovis infection and colon cancer has been known for at least 25 years (Klein et al . , 1977). Moreover, recent research has shown that Streptococcus bovis can promote intestinal carcinogenesis in a rat model for colon cancer, and that certain cell surface proteins of this bacterium can induce inflammation, supporting a possible linkage between Streptococcus bovis, inflammation and colon carcinogenesis (Elmerich et al . 2000, Biarc et al., 2004). Streptococcus bovis can be detected in the gastrointestinal tract of only ~ 10% of the human population, and is considered to be a lower grade pathogen involved in bacteremia and endocarditis. The present inventor surprisingly discovered that there is a strong correlation between an infection with Streptococcus bovis and the presence of colon tumors or polyps . This is highly surprising since, based on the prior art, one would expect a rather low correlation between colon cancer and an Streptococcus bovis infection instead of the about 88% or higher correlation which the present inventor found. The expected low correlation according to the prior art would not indicate to the skilled person that testing for anti Streptococcus bovis immunoglobulin molecules would be highly predictive of the presence of colon tumors, especially in the pre or early stages of the disease.

Without limiting the scope of the invention to any theory, the positive strong correlation between the presence of immunoglobulin molecules against Streptococcus bovis and colon cancer can possibly be explained by two models.

First, polyps or colon tumors provide an excellent niche for Streptococcus bovis, which is followed by bacterial proliferation and infection. Second, a bacterial infection of epithelial cells can induce polyp and subsequently tumor formation as suggested previously (Ellmerich et al . , 2000; Biarc et al. , 2004) .

Preliminary data indicates that silent Streptococcus bovis infections are established in polyps, early during carcinogenesis. Furthermore, passive antigen presentation due to tumor/polyp bleedings is less likely because antigens from the gut bacterium E. coli could not distinguish tumor from control serum. This corroborates the lack of difference in antibody titers against lipopolysaccharide core antigens of Gram-negative bacteria in colon cancer patients and control subjects as reported by Darjee and Gibb (1993). Together, this suggests that the presence of antibodies to Streptococcus bovis antigens in colon cancer patients is not simply due to a non-specific increase in antibodies to gut bacteria. It is hypothized that Streptococcus bovis incidentally enters the gastrointestinal tract by consumption of contaminated meat (Knudtson and Hartman. 1993) , In the bowel of healthy individuals, Streptococcus bovis is quickly competed out by the established gut flora. However, in patients with polyps or early colon tumors, Streptococcus bovis has the ability to attach to these malignant sites by bacterial surface proteins that have affinity for specific surface proteins expressed by tumor cells. Thereby, Streptococcus bovis finds a niche for survival in the bowel, and can cause a local tumor associated infection.

This infection might possibly promote tumor development by stimulating the Cyclooxygenase-2 pathway, inducing hyperproliferation, invasion and angiogenesis, whereas apoptosis is inhibited (Biarc et al., 2004; Wendum et al., 2004).

In a preferred embodiment of the present invention, the sample of the individual is contacted with a preparation comprising one or more antigens of Streptococcus bovis and binding is detected between the immunoglobulin molecules and the antigens_..

The term "antigens of Streptococcus bovis" as used herein comprises any compound or molecule whose shape and/or sequence triggers the production of antibodies (immunoglobulins) against Streptococcus bovis. This comprises proteins of Streptococcus bovis itself but also fragments thereof, including peptides both naturally occurring or synthetic, and compounds able to elicit the production of immunoglobulins against Streptococcus bovis such as synthetic compounds and homologous proteins and fragments of closely related species.

The use of a preparation of one or more antigens of Streptococcus bovis, such as for example a protein extract, allows for an improved sensitivity of the method according to the present invention especially since aspecific binding of the immunoglobulin molecules against Streptococcus bovis, for example with fatty acids, nucleic acids, or other contaminants is reduced or inhibited. In a more preferred embodiment of the method according to the present invention, the method comprises:

— immobilizing the total serum immunoglobulin molecules on a solid support;

— contacting the immobilized total serum immunoglobulin molecules with the Streptococcus bovis preparation to capture Streptococcus bovis antigens;

— eluting the captured Streptococcus bovis antigens from the solid support; and — detecting the presence of Streptococcus bovis antigens in the eluate.

This indirect detection of the immunoglobulin molecules against Streptococcus bovis allows for a further improvement of the detection of these molecules. Immunoglobulin molecules are relatively large molecules and it is difficult to discriminate between empty molecules and complexes of the molecules with the usually much smaller Streptococcus bovis antigens, especially using a detection methods based on size such as size exclusion chromatography, gel electroforeses, etc.

Using the usually much smaller Streptococcus bovis antigens for detection not only allows for an improved detection due to a higher discriminative power, but also speeds up the methods since it usually takes less time to efficiently separate relatively small molecules from each other compared to larger molecules.

A particularly preferred method for detecting the presence of Streptococcus bovis antigens in the eluate is mass spectrometry (MS) .

Mass spectrometry (MS) , preferably using a surface- enhanced laser desorption/ionization time-of-flight (SELDI- TOF) reader for detection, provides a fast reproducible profile of the molecular masses of the captured Streptococcus bovis antigens. Further, due to its discriminative power, even Streptococcus bovis antigens which only differ in molecular weight by as much as 100 Da can be easily discriminated. Furthermore, the above immuno-capture mass spectrometry, preferably using a SELDI-TOF platform, provides an alternative approach for high-throughput screening of large number of serum samples. Both immunoglobulin molecule binding and the subsequent Streptococcus bovis antigen capture can be performed on one ProteinChip surface (Gruss et al., 2003) .

Moreover, immuno-capture MS provides a knowledge- based profiling approach despite the fact that identities of individual discriminative markers are not necessarily known. _ _The reproducibility of this approach is largely- increased, compared to currently employed serum profiling strategies (Baggerly et al . , 2004), by the use of serum IgG' s, which are generally regarded as stable proteins, and a separate controllable antigen source, i.e., the Streptococcus bovis preparation, to generate discriminative patterns. This reduces the risk that (small) differences in serum sample handling interfere with the diagnostic pattern of instable serum-borne biomarkers. As indicated above, the use of a separate controlled antigen source improves the discriminative power of the detection.

Therefore, in the method according to the present invention, preferably antigens of Streptococcus bovis are selected from the group consisting of HIpA, Rp L7/L12, GAPDH, class II aldolase, Eno, 3OS ribosomal protein S6, Hpr, Pkase, TIF-I, CapAB, CpnlO, 5OS ribosomal protein L9, 30 S ribosomal protein S15, 5OS ribosomal protein L20, PrP, 3OS ribosomal protein S5, 5OS ribosomal protein L29, 6-Phfk, 3OS ribosomal protein S13, 5OS ribosomal protein LIl, and 5OS ribosomal protein L15, 5OS ribosomal protein L27, phosphoenolpyruvate carboxylase, 3OS ribosomal protein S20, 5OS ribosomal protein L31, DNA-binding HTH domain-containing protein, major cold shock protein, CsbD-like protein, secreted protein, SagA-like protein, pyrimidine operon regulatory protein, 3OS ribosomal protein S16, 5OS ribosomal protein Ll, 50 S ribosomal protein L18, 5OS ribosomal protein L2, 5OS ribosomal protein L24, Glycine/serine hydroxymethyltransferase, acetyltransferase, N-acetylase of ribosomal proteins, response regulator of the LytR/AlgR family, superfamily I DNA and RNA helicase, connector protein, transcription regulator, cell division initiation protein, diguanylate cyclase/phosphodiesterase domain 1, glutamine synthetase type 1; glutamate-ammonia ligase, GroEL, nitrogen regulatory protein P-II, phophoesterase, iron(III): ABC transporter, solute-binding protein, proton-translocating ATPase: beta subunit, phosphoglycerate kinase, polypeptide deformylase, ribosome- associated protein, pyruvate kinase, ribosomal protein L15, single-stranded DNA binding protein, prohibitin-like protein, uricase, and Hsp70-like protein.

More preferred antigens of Streptococcus bovis are antigens selected from the group consisting of HIpA, Rp L7/L12, GAPDH, class II aldolase, Eno, 3OS ribosomal protein S6, Hpr, Pkase, TIF-I, CapAB, CpnlO, 5OS ribosomal protein L9, 30 S ribosomal protein S15, 5OS ribosomal protein L20, PrP, 30S ribosomal protein S5, 5OS ribosomal protein L29, 6- Phfk, 3OS ribosomal protein S13, 5OS ribosomal protein LIl, 50S ribosomal protein L15; 5OS ribosomal protein L27.

Most preferred antigens of Streptococcus bovis are HIpA and/or Rp L7/L12.

According to the present invention, the above antigens were identified by analysis of the captured antigens from a Streptococcus bovis protein extract and the subsequent partial amino acid sequence determination and comparison of this amino acid sequences with the public databases.

Because the genomic sequence or partial sequences thereof of Streptococcus bovis were largely not . known at the filing date of this application, the identity of the captured antigens of Streptococcus bovis was established by amino acid sequence similarity with proteins from other sources, thereby yielding the identity of the captured antigens of Streptococcus bovis.

The present invention also preferably relates to methods wherein the antigen of Streptococcus bovis comprises an amino acid sequence or amino acid sequences selected from the group consisting of SEQ ID Nos: 1 to 3; 4 to 6; 7 to 12; 13 to 16; 17 to 20; 21 to 23; 24; 25 to 28; 29 to 30; 31 to 32; 33 to 35; 36 to 37; 38 to 39; 40 to 41; 42 to 43; 44 to 45; 46 to 47; 48, 49; 50; 51; 52; 53; 54; 55; 56; 57; 58; 59; 60; 61; 62; 63; 64; 65; 66; 67; 68 ;69; 70; 71; 72; 73; 74; 75; 76; 77; 78; 79; 80; 81; 82; 83; 84; 85; 86; 87; 88; 89; and 90, more preferably of SEQ ID Nos: 1 to 3; 4 to 6; 7 to 12; 13 to 16; 17 to 20; 21 to 23; 24; 25 to 28; 29 to 30; 31 to 32; 33 to 35; 36 to 37; 38 to 39; 40 to 41; 42 to 43; 44 to 45; 46 to 47; 48, 49; 50; 51; and 52 and most preferably of SEQ ID Nos: 1 to 3; and 4 to 6.

According to one preferred embodiment of the present invention the presence of Streptococcus bovis antigens is detected by the presence of one or more proteins and/or protein fragments with molecular weights of 7.8 kDa, 7.9 kDa or 9.6 kDa.

As outlined above, the present invention provides improved methods for the diagnosis of colon cancer by using antigens of Streptococcus bovis for the detection. Therefore, the present invention also relates to a protein of

Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 1 to 3; a protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 4 to 6; a protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 7 to 12; a protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 13 to 16; a protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 17 to 20; a protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 21 to 23; a protein of Streptococcus bovis comprising amino acid sequence SEQ ID No: 24; a protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 25 to 28; a protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 29 to 30; a protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 31 to 32; a protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 33 to 35; a protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 36 to 37; a protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 38 to 39; a protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 40 to 41; a protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 42 to 43; a protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos : 44 to 45; a protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 46 to 47; a protein of Streptococcus bovis comprising an amino acid sequence according to SEQ ID No: 48; a protein of Streptococcus bovis comprising an amino acid sequence according to SEQ ID No: 49; a protein of Streptococcus bovis comprising an amino acid sequence according to SEQ ID No: 50; a protein of Streptococcus bovis comprising an amino acid sequence according to SEQ ID No: 51; and a protein of Streptococcus bovis comprising an amino acid sequence according to SEQ ID No: 52.

The above identified proteins are preferably used as antigens of Streptococcus bovis in a method according to the invention. Since these protein of Streptococcus bovis were not identified or isolated before the present invention, the invention also relates to the use of these proteins of Streptococcus bovis for the diagnosis of colon cancer.

Further, the present invention relates to these proteins of Streptococcus bovis for use as a diagnosticum.

Furthermore, the present invention relates to the use of the above proteins for the preparation of a medicament for the prophylaxis and/or treatment of colon cancer, wherein preferably the prophylaxes comprises vaccination and the treatment comprises drug targeting. the present invention shows for the first time that the above proteins are surface proteins of Streptococcus bovis , making them attractive proteins for vaccination purposes and attractive drug targets to fight bacterial infections.

As most these proteins, except for a putative secreted protein, SagA-like protein, iron (III) ABC transporter, solute-binding protein, and prohinitin-like protein (SEQ ID Nos 29, 30, 52 and 58 respectively) lack any of the known signals for protein export and/or surface retention (Tjalsma et al . , 2004; Tjalsma and van Dijl, 2005), it is presently not known via which mechanisms these proteins are retained at the bacterial surface

The present invention will now be further detailed in the following example and the accompanying figures wherein:

Figure 1: schematically shows one preferred embodiment of the present invention. In the figure, total serum

IgG is captured by immobilized Protein A. Next, this IgG pool is incubated with Streptococcus bovis antigens after which eluted IgG-antigen complexes are analyzed by mass spectrometry (MS) .

Figure 2: shows a SELDI-TOF MS protein profile of IgG- captured Streptococcus bovis antigens (IgG- captured antigen; second panel) using pooled serum from four colon cancer patients. Control protein profiles concerned isolated IgG without incubation with Streptococcus bovis antigens (empty IgG; upper panel) , Streptococcus bovis cell wall extract (antigen source; third panel) , and Streptococcus bovis cell wall proteins a _.. _ _. specifically bound to Protein A (non-IgG bound; fourth panel) .

Figure 3: shows in panels A and B the IgG-specific capture of three Streptococcus bovis antigens (A) m/z 7748 and 7888 and (B) m/z 9563 (middle panels), using serum from a colon cancer patient. Controls protein profiles concern isolated IgG without incubation with Streptococcus bovis antigens (empty IgG; upper panel), Streptococcus bovis cell wall extract (antigen source; lower panel. Peak intensity is given in arbitrary units; m/z, mass /charge.

Figure 4: shows a colon cancer associated bacterial antigen pattern. Serum samples from 12 (early stage) colon cancer patients, 4 polyp patients and 8 asymptomatic control subjects were used to generate antigen patterns as indicated in figure

1. Ten antigen peaks (m/z) with highest intensities in any one of the 24 examined samples were selected for hierarchical cluster analysis. Relative peak intensities are indicated in red; green indicates low intensity. Absolute values of peak intensities are given in Table II. m/z, mass /charge.

Figure 5: shows a classification tree for colon cancer. Samples are classified using antigen peaks m/z

7748 (node 1) and 7888 (node 2) and m/z 9563 (node 3) . Samples were from 12 (early stage) colon cancer patients, 4 polyp patients and 8 asymptomatic control subjects. Samples are _ _ classified as "diseased" when intensities at nodes 1, 2, or 3 are above the indicated cut-of value (+) . The samples with intensities below the cut-of value (-) are further classified at the following node, and finally at node 3 classified as "not diseased".

Figure 6: shows an immuno-capture MS using Streptococcus bovis and E. coli protein extracts. SELDI-TOF MS protein profiles using pooled serum from four control subjects and four patients with colon tumors. Upper two panels show protein profiles of Protein A-isolated IgG' s without antigen incubation (empty IgG) . Panels 3 to 6 show protein profiles of captured Streptococcus bovis antigens using IgG' s from control and tumor serum isolated with Protein A (ProtA) and Protein G (ProtG) , respectively. Bottom two panels show protein profiles of captured E. coli antigens using IgG' s from control and tumor serum isolated with Protein A. IgG-captured antigens are indicated. Peak intensity is given in arbitrary- units; m/z, mass /charge.

Figure 7: shows the identification of a histon-like antigenic protein from Streptococcus bovis. SDS- PAGE (A), and SELDI-TOF MS protein profiles (B) of Streptococcus bovis surface proteins (upper panel) and fractions thereof, eluted from sections of a 16.5% Tricine SDS PAA gel (sections denoted 10, 8.5, 7 and 6 kD as indicated). Note that changes in protein profiles can be due to distinct ProteinChip binding characteristics of _ native (upper panel) and denatured (gel-eluted fractions) variants of the same proteins. Sizes of the reference protein markers used for SDS- PAGE, and protein peaks corresponding to diagnostic Streptococcus bovis antigens are indicated. Peak intensity is given in arbitrary units; m/z, mass /charge. C. Histon-like protein A (HIpA) from Streptococcus bovis as identified by semi-quantitative peptide mapping. Number of peptide hits corresponding to HIpA sequences from Streptococcus thermophilis and Bacillus cereus in each gel-eluted protein fraction is indicated. D. Deduced partial amino acid sequence of HIpA from Streptococcus bovis. HlpA-derived peptides from S. thermophilis or B. cereus (underlined) are indicated in capitals. Amino acids printed in lower case show the HIpA sequences from S. thermophilis or B. cereus (underlined) that were not hit by peptide mapping

Figure 8: shows a heparin-affinity profile. SELDI-TOF MS protein profiles of Streptococcus bovis cell wall extracts with (heparin-bound; lower panel) , and without affinity to heparin (unbound; upper panel) . The 9563 Da peak, corresponding to HIpA from Streptococcus bovis, is indicated. Peak intensity is given in arbitrary units; m/z, mass /charge.

Figure 9: shows an immunoprecipitation of diagnostic antigens. SDS-PAGE (A), and SELDI-TOF MS protein profiles (B) of Streptococcus bovis surface proteins immunoprecipitated with serum IgC s from two colon cancer patients. Sizes of reference protein markers, and sections 10, 8.5, 7 and 6 kD, previously used for peptide mapping of Streptococcus bovis surface proteins (antigen pool) , and protein peaks corresponding to known (HIpA) and unknown (?) diagnostic antigens are indicated. Peak intensities are given in arbitrary units; m/z, mass /charge. Figure 10: shows peptide hits in HIpA from Streptococcus bovis. HlpA-derived peptides found by in-gel tryptic digestion of Streptococcus bovis surface proteins are indicated in capitals. Boxed indicates that sequences were derived from S. thermophilis, and underlined capitals indicates that sequences were derived from B. cereus. Sequence coverage is 62% (56/91 amino acids) . HlpA-derived peptides that were also found by immunoprecipitation-based peptide mapping are indicated in grey shadings . Sequence coverage is 41% (37/91 amino acid residues). In the latter case, shaded capitals indicate that sequences were derived from Enterococcus faecalis. Amino acids printed in lower case are HIpA sequences that were not hit by peptide mapping. Theoretical mass of the HIpA protein with this sequence is 9675 Da.

Figure 11: shows amino-terminal processing of Rp L7/L12. SELDI-TOF MS spectrum of antigens 7748, 7877, 7888, 7847, and 7948 immunoprecipitated from Streptococcus bovis surface proteins using serum IgG' s from a colon cancer patient. The observed mass differences (DM/Z) between the antigen peaks, and closest matching amino acid residues corresponding to these differences as a result of amino-terminal cleavage of a single precursor protein are indicated. The predicted masses are (amino acid residue minus hydroxyl group of 17 Da) : A = 72 Da, V = 100 Da, T = 102 Da. The m/z, mass /charge (B) Streptococcal Pr L7/L12 amino acid sequences containing the TAAA and VAAA motifs (boxed) between position 39 and 50. Sequences were from Streptococcus thermophilics (ref. AAV62133.1), Streptococcus agalactiae (ref. NP_688299.1) , Streptococcus pyogenes (ref. AAN58664.1), Streptococcus mutans (ref. AE014936_4), Streptococcus pneumoniae (ref. NP_358804.1) , and Streptococcus suis (ref. ZP_00331499.1) , numbers refer to the position of the first and last residues shown.

Figure 12: shows peptide hits in Ribosomal protein L7/L12 from Streptococcus bovis. Rp L7/L12-derived peptides found by in-gel tryptic digestion of Streptococcus bovis surface proteins or immunoprecipitation-based peptide mapping (grey shadings) are indicated in capitals. Boxed indicates that sequences were derived from S. thermophilis (ref. YP_140948), and underlined capitals indicates that sequences were derived from S. agalactiae (ref. NP_688299) . Putative processing sites 1 - 4 are indicated by arrows. Sequence coverage of the at position 1 truncated form of Rp L7/L12 is 69% (58/84 amino acids) . Amino acids printed in lower case are Rp L7/L12 sequences derived from S. thermophilus, but were not hit by peptide mapping. Theoretical mass of the intact RP L7/L12 protein with this sequence is 12931 Da.

Figure 13: shows the identification of tropomyosin as a protein that co-fractionates with Streptococcus bovis surface proteins. Streptococcus bovis surface proteins were separated by 16.5% Tricine SDS PAA gel electrophoresis and gel sections were applied to in-gel tryptic digestion. The eukaryotic cytoskeleton protein tropomyosin could be identified by homology searches in public protein databases. Number of peptide hits corresponding to tropomyosin sequences are printed in capital letters (sequence coverage of 43%) , lower case show the tropomyosin sequences that were not hit by peptide mapping.

EXAMPLE

MATERIAL AND METHODS

Serum samples.

Pre-operative serum samples from 12 colon cancer patients, and 4 polyp patients, who had been admitted to the Department of Surgery of the Radboud University Nijmegen Medical Centre, The Netherlands, were used in this study.

Serum samples from 8 asymptomatic, age-matched, blood donors were used as controls. The characteristics of patients, and controls are presented in Table II. None of the patient or control subjects had a known history of Streptococcus bovis infection. The study was approved by the local medical ethical committee, and informed consent was obtained from all patients. Serum samples were stored at -80 'C until use.

Table I. Characteristics of colon cancer patients, polyp patients and asymptomatic control subjects from which serum samples were used in this study.

¹¹ m, male; f, female ²⁾ NR, not reported

³¹ NA, not applicable

Bacterial extracts . The Streptococcus bovis strain used in this study was NCTC 8133, recently reclassified as Streptococcus infantarius subsp. infantarius (Schlegel et al. , 2003). Bacteria were cultured at 37 ^'C in brain-heart infusion broth (Difco

Laboratories) with 1% glucose under anaerobic conditions. The cell surface proteins were prepared from 18 h culture. Bacteria were centrifuged, the pellet washed three times with phosphate-buffered saline pH 7.4 (PBS; Sigma), and shaken for 1 h at 4 'C with 0.5 M phosphate buffer pH 6 in the presence of glass beads as has been described (Schδller et al., 1981; Biarc et al. , 2004) .

The suspension was centrifuged and the supernatant dialysed against water, lyophilized and resuspended in PBS. The E. coli strain used in this study was DH5a (Invitrogen) . Cells were grown at 37° C for 18 hours after which cells were collected by centrifugation. Cells were resuspended in spheroplast buffer (20% sucrose, 100 mM Tris-HCl pH 8.0, I mM EDTA, 40 ng/ul Lysozyme) , and incubated for 5 min at room temperature (RT) .

After addition of one volume H₂O, and incubation for another 5 min at room temperature, spheroplasts were collected by centrifugation. The spheroplasts were resuspeded in lysis buffer (PBS with 0.1% Triton X-100), and incubated for 5 minutes at RT. After centrifugation, the supernatant was collected and kept frozen at -20° C until use.

Immuno-capture of bacterial antigens .

Bacterial antigens were captured with antibodies isolated from serum samples. To do this, Protein A-Sepharose or Protein G-Sepharose (Pierce) was incubated with 40 μl serum in PBS containing 0,1% Triton-XlOO (PBS-Tx). Supernatant (containing unbound serum proteins) was decanted after centrifugation, and Protein A/G-Sepharose with bound IgG was resuspended in PBS-Tx containing 40 μg protein extract from Streptococcus bovis or E. coli. Protein A/G-Sepharose bound to IgG-captured antigens was collected by centrifugation, and washed 3 times PBS-Tx. IgG-antigen complexes were eluted from Protein A/G-Sepharose by 0,5 M acetic acid, 0,15 M NaCl (pH 2.4), collected in a fresh tube, and directly used for mass spectrometry. Optionally, the eluate was neutralized with IM Tris-HCl pH

8.8, applied to a <3 kD size-exclusion column (Milipore) , and re-dissolved in PBS. In the latter case samples were diluted to an IgG concentration of 2 μg/μl prior to mass spectrometry.

Heparin-capture of bacterial antigens .

Heparin binding bacterial antigens were captured using Heparin-Sepharose (Pierce). To do this, Heparin- Sepharose was incubated with 40 μg Streptococcus bovis surface proteins in PBS containing a final concentration of

0.25 M NaCl and 0.1% Triton-XIOO (PBS_0-25-Tx).

Supernatant (containing unbound proteins) was collected after centrifugation, and concentrated using a <3 kD siz.e_-exclusion column (Milipore) . Heparin-Sepharose was washed 3 times with PBS₀₂₅-Tx before elution of heparin binding proteins by PBS containing a final concentration of

1.5 M NaCl (PBS_1-5) .

Protein profiling by SELDI-TOF MS.

To obtain protein profiles of IgG-captured antigens, Protein A/G-Sepahrose eluates were applied on an H₂O- equilibrated spot of a normal phase chip (NP20), and allowed to air dry. To obtain protein profiles of heparin-binding proteins, Heparin-Sepharose eluates were applied to an H₂O- equilibrated spot of an NP20 chip and incubated for 30 min in a humidity chamber.

To obtain profiles of SDS-polyacrylamide gel-eluted proteins, samples were applied to an 0.1 M ammonium acetate

(pH3) -equilibrated cation exchange protein chip (CMlO), and allowed to air dry. Next, samples were- removed and/or spots were washed one to three times with H₂O or equilibration buffer, and allowed to air dry. Finally, 0.8 μl of a saturated solution of sinapinic acid in 0.5% (v/v) trifluoroacetic acid and 50% (v/v) acetonitrile was applied to each spot surface, allowed to air-dry, and reapplied. Mass/charge (m/z) spectra were generated in a Ciphergen Protein Biology System II Time-of-Flight mass spectrometer. Laser intensity was set to 200 with detector sensitivity of 9, high mass to acquire 50 kDa, with optimization range of 5-15 kDa, 400 laser shots were averaged to obtain the spectra. External calibration was performed using Hirudin BKHV (7033.6 Da), bovine Cytochrome C (12230.9 Da), Myoglobin (16951.5 Da), and bovine Carbonic Anhydrase (29023.7 Da) as standards (Ciphergen Biosystems) .

Proteinchip binding and SELDI-TOF MS within one experiment were performed on the same day with samples randomly distributed over ProteinChip arrays. ProteinChip Software 3.0 was used for analysis of the mass spectra. If applicable, spectra were normalized to total ion current before further analysis. The Biomarker Wizzard application of the ProteinChip Software was used for peak detection. Hierarchical cluster analysis was performed with Cluster 3.0 (complete linkage / city-block distance) and Tree View available at http://rana.lbl.gov/EisenSoftware.htm (Eisen et al . , 1998).

Peptide mapping by nanoLC- (FT)MS/MS.

To reveal the identify of Streptococcus bovis proteins, surface protein extracts were separated on a 16.5% Tricine SDS-polyacrylamide gel (Schagger and von Jagow, 1987). The separated proteins were in-gel treated with dithiothreitol and iodoacetamide, followed by in-gel digestion with trypsin20 as described by Lasonder et al. (2002) .

Diagnostic Streptococcus bovis antigens were captured with patient antibodies isolated from serum samples. To do this, Protein A-Sepharose was incubated with 200 μl serum in PBS-Tx for 1 hour at RT. Supernatant was decanted after centrifugation, and Protein A-Sepharose with bound IgG was resuspended in PBS-Tx containing 400 μg Streptococcus bovis surface proteins.

After incubation for one hour at RT, Protein A- Sepharose with IgG-captured antigens was collected by centrifugation, and washed 3 times with PBS-Tx to remove unbound antigens, and three times with PBS to remove Triton- XlOO. IgG-antigen complexes were eluted from Protein A- Sepharose by 50% (v/v) acetonitrile and 0.3% (v/v) trifluσroace-tic acid.

After centrifugation, the supernatant was applied to an YM-30 spin column (30 kDa cut-off filter; Millipore) to deplete the sample from IgG. The sample was concentrated using an YM-3 spin column (3 kDa cut-off filter; Millipore) in combination with liquid evaporation in a vacuum centrifuge .

Proteins were denatured by 8M urea, and incubated with 1 mM dithiothreitol and 1 mM iodoacetamide for 1 hour in two successive steps at RT. Proteins were digested by LysC for three hours, followed by the addition of trypsin20 in 50 mM ammonium hydrogen carbonate and continued cleavage at 37⁰C for 15 hours. Prior to nanoLC-MS analysis, all samples were purified and desalted after digestion using Stage tips (Rappsilber et al._f 2003).

Peptide identification experiments were performed using a nano-HPLC Agillent 1100 nanoflow system connected online to a 7-Tesla linear quadrupole ion trap-Fourier transform (LTQ-FT) mass spectrometer (Thermo Electron,

Bremen, Germany) . Peptides were separated on a 10 cm 100 μm ID PicoTip (New Objective, Woburn, USA) columns packed with 3μm Reprosil C18 beads (Dr. Maisch GmbH, Ammerbuch, Germany) using a 30 min gradient from 10% buffer A (0.5% acetic acid) to 30% buffer B (80% acetonitrile in 0.5% acetic- acid) with a flow-rate of 300 nl/min.

Peptides eluting from the column tip were electro sprayed directly into the mass spectrometer with a spray voltage of 2.1 kV. The mass spectrometer was operated in the data-dependent mode to sequence the four most intense ions per duty cycle Briefly, full-scan MS spectra of intact peptides (m/z 400-1500) with a automated gain control accumalation target value of 1E6 ions were acquired in the Fourier transform ion cyclotron resonance (FT ICR) cell, with- a re-solution of 50.000. The four most abundant ions were sequentially isolated and fragmented in the linear ion trap by applying collisionally induced dissociation using an accumalation target value of 20.000 (capillary temperature, 200⁰C; normalized collision energy, 30%. A dynamic exclusion of ions previosuly sequenced within 180 s was applied. All unassigned charge states were excluded from sequencing. A minimum of 500 counts were required for MS2 selection. RAW spectrum files were converted with the aid of Perl script algorithms to DTA files and combined into a single Mascot generic peaklist.

Peptides and proteins were identified using the Mascot (Matrix Science) algorithm to search a local version of the NCBInr database (http://www.ncbi.nlm.nih.gov). The following search criteria were applied: 30 ppm tolerance for the parental peptide and 0.8 Da for fragmentation spectra, and a fixed carbamidomethyl modification for cysteines, and an oxidized variable modification for methionine. First ranked peptides were parsed from Mascot database search html files with MSQuant

(www.msquant.sourceforge.net) to generate unique first ranked peptide lists. Proteins identified by first ranked peptides were verified by manual inspection of the MS/MS spectra in MSQuant or Mascot. Proteins identified with one or more unique peptides with a (n average) peptide score >30 were considered significant. Relative protein abundance was based on the total number of unique peptides identified for each protein.

RESULTS

Immuno-Capture Mass Spectrometry

- - To monitor the presence of anti-Streptococcus bovis immunoglobulins (IgG) in colon cancer patients, a three-step proteomics-based serological approach was developed (Fig. 1). First, total serum IgG from colon cancer patients was isolated using immobilized Protein A. In a second step, the pool of immobilized human IgG was incubated with a pool of Streptococcus bovis surface proteins (antigen source) , allowing the binding of those antigens that have induced a humoral immune response. Finally, captured Streptococcus bovis antigens were eluted and used to generate antigen profiles using mass spectrometry (MS) on a surface-enhanced laser desorption/ionisation time-of-flight (SELDI-TOF) reader. Thus, the obtained antigen patterns are diagnostic for the presence of circulating antibodies against the corresponding antigens. Fig. 2 and Fig. 3 C show typical MS profiles of Streptococcus bovis antigens captured by serum antibodies from colon cancer patients. In parallel, protein profiles of "empty" IgG' s (without antigen capture), of the crude Streptococcus bovis surface proteins (antigen source) , and of Streptococcus bovis surface proteins a specifically bound to Protein A (non-IgG bound) were generated.

These data show that Streptococcus bovis antigens can be captured with IgG from colon cancer patients using and Immuno-capture MS approach. Furthermore, it shows that Streptococcus bovis surface proteins or homologous antigens from closely related bacteria have induced an humoral immune response in these patients.

Antigen Patterns in Colon Cancer and Polyp Patients

The clinical relevance of silent Streptococcus bovis infection in colon cancer patients was performed using 24 serum -s-amples . These serum samples were from 12 (early stage) colon cancer patients, 4 polyp patients and 8 asymptomatic, age matched, control subjects (see Table I). It should be noted that all examined colon cancer patients, polyp patients, and asymptomatic control subjects had no history of Streptococcus bovis bacteraemia.

All samples were applied to the immuno-capture MS protocol to generate antigen patterns as described above. Only those peaks that were also present in the profile of the crude Streptococcus bovis surface extracts (antigen source) , but absent in the empty IgG end Protein A profiles, were used for further analysis. The ten antigen peaks with highest intensities in any one of the 24 examined samples were selected for unsupervised hierarchical cluster analysis.

As shown in Fig. 4, heterogeneous antigen patterns were obtained for all individual samples. Nevertheless, relatively high antigen peak intensities were seen in most of the tumor samples, whereas antigen peak intensities were generally low in the samples from the control subjects (see the below Table II for absolute peak intensities).

Table II Peak intensities¹ of Streptococcus bovis antigens captured by IgG' s from colon cancer patients (Tl- 8; TN1-4) _r polyp patients (Pl-4) and asymptomatic control subjects (Cl-8) .

m/z (Da) 3907 4783 4815 7748 7807 7888 9563 9675 9757 10959

Tl 3.3 2.6 1.7 7.1 5.6 3.2

T2 1.2 1.7 4.8 0.3 2.3 0.7 2.0

T3 1.7 5.2 4.2 1.7 0.0 0.5

T4 1.9 1.5 3.7 6.3 2.0 0.0 2.6

T5 0.5 2.0 1.9 2.2 0.8 1.6 1.0 0.0 0.3

T6 1.4 0.8 1.6

T7^" i:9- 3.1 1.7 0.0 2.2 1.5 0.2

T8 1.3 0.2 0.8 2.9 0.3 0.6 0.1 1.0

TNl 0.9 1.0 1.6 2.4 0.6 0.3 0.8 1.4 0.0 1.2

TN2 1.0 3.1 3.5 2.8 2.8 0.0 2.6

TN3 1.6 1.! 2.6 1.0 2.3 4.2 0.0 3.0

TN4 0.6 0.9 2.0 3.1 4.3 1.2

Pl 0.9 3.1 0.7 2.6 1.0 3.0 5.0 0.0 2.4

P2 1.1 1.0 5.0 5.2 0.5 3.9 4.8 0.0

P3 0.8 2.4 1.4 1.7 2.1 3.0 5.1

Remarkably, also samples from polyp patients contain a number of antigen peaks with intensities above those from the controls. It should be noted that about -20% of individuals over the age of 50 carry "silent" polyps (Loeve et si., 2004), thus concerning one or two of the asymptomatic control subjects in this study.

If so, likely candidates to carry "silent" polyps are controls C6 and C7 (Fig. 4). Together, these data indicate that infection with Streptococcus bovis occurs both in colon cancer and polyp patients, and that this event can serve as a marker to identify early stage colon cancer, and to identify people with (a) colon polyp (s), who are at risk of developing a colon tumor.

Classification of Colon Cancer and Polyp Patients

To further explore the discriminative power of the obtained Streptococcus bovis antigen patterns, intensities of the antigen peaks in all 24 spectra were used to determine cut-off values that fully separate control samples from those of colon cancer and polyp patients (Table II). Based on these values, three peaks (m/z 7748, 7888, and 9563 Da) were chosen to set up a classification tree (Fig. 5) .

At node 1, samples with m/z 7 748 peak intensities > 6 were classified as diseased (cancer or polyp) , which concerned 9 tumor samples and 1 polyp sample. Of the remaining 14 samples, two tumor samples and one polyp sample with m/z 9563 peak intensities > 8 could be classified as diseased at node 2.

Finally, 1 polyp sample could be classified as diseased at node 3, as it had a m/z 7888 peak intensity > 5. The remaining samples were classified as not diseased, which concerned the 8 control samples, and samples from one colon cancer patient (TNl) and a polyp patient (P4) . Thus, at a specificity level of 100%, this model identified 11 of 12 cancer patients and 3 of 4 polyp patients yielding a sensitivity of 88%, and an overall accuracy of 94% for the detection of colon polyps and tumors.

Analysis of Streptococcus bovis Antigen Presentation

When a new microorganism is encountered in the human body, IgM is usually the first antibody produced by the immune system. When the body is attacked by the same microorganism for a prolonged time, IgG antibodies will be produGe-d after a few weeks to month after initial infection

(Gray and Skarvall, 1988; Welliver et al. , 1980; Zolkowski et al._r 1992.).

It should be noted that Protein A, used for antibody capture, has in addition to affinity for IgG also a weak affinity for IgM. To determine to which extent IgM contributes to the Streptococcus bovis antigen profiles, immuno-capture MS experiments were repeated with pooled serum from colon cancer patients and control subjects using Protein G which has only affinity for IgG (see Bjorck and Kronvall, 1984) .

As shown in Fig. 6, diagnostic peaks of m/z 7748, 7888, and 9563 Da have similar intensities, irrespective of the use of Protein A or Protein G. This shows that these Streptococcus bovis antigens are primarily captured by IgG, which indicates that immune stimulation has occurred over a long period of time.

Although it seems likely that the Streptococcus bovis immune response is caused by (chronic) bacterial infection, an alternative possibility is that presentation of antigens is passive due to tumor/polyp bleedings. In the latter case, antibodies against antigens from other gut bacteria might perform equally good or better in an immuno-capture assay. To evaluate this, immuno-capture MS was performed with pooled serum from colon cancer patients and control subjects, using protein extracts from the gut bacterium Escherichia coli as antigen source. This bacterium accounts for -1% of the total gut flora of healthy individuals and polyp patients (Moore and Moore, 1995), and is known to induce an immune response in asymptomatic individuals (Davies et al. , 2005).

As shown in Fig. 6 (bottom two panels), two captured E. coli antigens (m/z 9260 and 12207 Da) have similarly intensities, irrespective of the used serum pools. This in strong- contrast to the results obtained with Streptococcus bovis antigens (see Fig. 4, panels 3-6) .

This shows that similar antibody titers against E. coli antigens are present in colon cancer patients and asymptomatic control subjects, indicating that a humoral immune response to E. coli is not correlated with the presence of a colon tumor. Taken together, these data indicate that prolonged silent Streptococcus bovis infection is a common event in colon cancer patients, and that Streptococcus bovis antigens induce a specific humoral immune response in colon cancer patients.

Identification of Diagnostic Antigens by Peptide Mapping

High accuracy peptide mapping and homology searches in public protein databases can unravel the identity of proteins from organisms of which the genome sequence is not known, as is the case for Streptococcus bovis. To identify potential antigens that elicit an immune response in colon cancer patients, Streptococcus bovis surface proteins, used as antigen pools, were separated by ID SDS-PAGE. Subsequent SELDI-TOF MS analysis of four of these fractions (10, 8.5, 7 and 6) showed that fractions 8.5 was enriched with diagnostic antigen m/z 9563 Da (Fig. 7A, and B) . Next, in-gel tryptic digestion was performed on all four fractions, after which eluted tryptic peptides were applied to a nano-HPLC system connected online to a linear quadrupole ion trap-Fourier transform (LTQ-FT) mass spectrometer for peptide identification.

Database searches with assigned peptide sequences revealed the presence of 60 proteins, or fragments thereof, in the four protein fractions. From these 60 candidate colon cancer-associated antigens, 21 were identified with at least two distinct, peptides (Tables III, IV and V) . As indicated in Fig 7C, eleven peptides from fraction 8.5 corresponded to sequences within homologous bacterial Histon-like proteins (HIpA' s) from several Streptococcus and Bacillus species (data not shown) . From these sequences, the HIpA sequence from

Streptococcus bovis could be deduced, resulting in a HIpA sequence coverage of 62% (Fig. 7D) . Notably, the molecular weights of the HIpA proteins from S. thermophilus and B. cereus (-9.6 kD) are similar to the m/z value (9563) of the observed antigen peak from Streptococcus bovis.

The HlpA family of histon-like proteins is known to bind to heparin, and can be purified by heparin affinity chromatography (Stinson et al. , 1998) . To confirm that the 9563 Da peak corresponded to the Streptococcus bovis HIpA protein, the affinity of this protein for heparin was investigated.

Therefore, immobilized heparin was incubated with Streptococcus bovis surface proteins, allowing the binding of those proteins with affinity for heparin. As shown in Fig. 8, the 9563 Da peak is completely missing from the unbound protein fraction (upper panel) , whereas it is the main peak eluted from the immobilized heparin (lower panel) . This shows that the 9563 Da peak is indeed a heparin binding protein, confirming that HIpA from Streptococcus bovis is one of the diagnostic antigens for the detection of colon cancer.

Identification of Diagnostic Antigens by an Immunoproteomic Shotgun Approach

To reveal the identity diagnostic antigens 7748 Da and 7888 Da, an immuno-capture procedure was performed on a preparative scale. Serum from patient T6, which contains high antibo-dy titers against diagnostic antigens 9563 (HIpA; positive control), and 7748 Da and 7888 Da, and patient T2 containing only antibodies against the latter two antigens (see Figure 2; Table II) was used to this purpose.

As shown in Figure 9, HIpA can be precipitated from Streptococcus bovis extracts using serum IgG' s from patient

T6, whereas it is, as expected, not immunoprecipitated with

IgG' s from patient T2. Diagnostic antigens 7748 Da and 7888

Da are, however, present in both immunoprecipitates . This confirms that immuno-capture of diagnostic antigens from Streptococcus bovis requires the presence of corresponding antibodies in serum from colon cancer patients. To identify all Streptococcus bovis antigens captured by IgG' s from patient T6, the immunoprecipitated protein fraction was used for in-solution tryptic digestion, after which tryptic peptides were applied to a nano-HPLC system connected online to a linear quadrupole ion trap-Fourier transform (LTQ-FT) mass spectrometer for peptide identification.

As indicated in Table III, assigned peptide sequences corresponded only with two of the 60 previously identified Streptococcus bovis surface proteins. HIpA was identified by 3 different tryptic peptides with a sequence coverage of 41% (Fig. 9; Table III and IV), and Rp L7/L12 was identified with 6 tryptic peptides. However, Rp L7/L12 has a predicted mass of -12 kDa, whereas the observed diagnostic antigens have a smaller size.

It should be noted that the matching peptides corresponded only to the carboxyl-terminal part of Rp L7/L12, suggesting that part of the amino-terminus of this protein is missing. To investigate whether the observed immunoprecipitated antigen peaks could be derived from a single precursor protein, the SELDI-TOF mass spectrum was analys-ed in more detail.

The diagnostic antigens 7748 Da and 7888 Da are part of a cluster of five peaks with masses of 7748, 7877, 7888, 7846, and 7947 Da, respectively (Fig HA) . If these peaks are the result of amino-terminal cleavage of a single precursor protein, the closest matching amino acid sequences corresponding to these mass differences is [T/V] -A-A-A. Strikingly, a T-A-A-A-x (1, 3) -V-A-A-A motif is present between position 39 and 50 of at least six different Streptococcal Rp L7/L12 proteins found in public databases (Fig 12, A and B) .

Together, these findings indicate that the diagnostic antigen 7748 and 7888 belong to amino-terminally truncated forms of Rp L7/L12. The sequence of this protein from Streptococcus bovis can be deduced by identified peptide sequences derived from S. thermophilis and Streptococcus agalactiae with a sequence coverage of 45% (Fig 12A) . Although we can currently not discriminate between cleavage at the T-A-A-A or V-A-A-A sequence, processing at the latter site is most probable, as the predicted masses of the five Rp L7/L12 fragments resulting from cleaveage at the V-A-A-A motif resemble most closely those of the antigen peaks observed by SELDI-TOF MS (Fig HA and 12B) .

Taken together, these data confirm the identity of HIpA as a diagnostic antigen, and reveal the identity of the 7748 Da and 7888 Da peaks as proteolytic fragments of Rp L7/L12 from Streptococcus bovis.

Identification of eukaryotic proteins with affinity for Streptococcus bovis

One possible explanation for the specific association of Streptococcus bovis with colon tumors is its ability to attach to malignant sites, thereby finding a niche for survival.- In. this respect, the identification of a surface- attached histon-like protein from Streptococcus bovis that elicits an immune response in several colon cancer patients is an interesting finding.

Although this protein belongs to a well-known family of highly conserved bacterial proteins (Drlica and Rouviere- Yaniv, 1987) , it was never reported to be exposed on a bacterial surface. However, it has been reported that extracellular, released, HIpA from Streptococcus pyogenes was complexed with soluble lipoteichoic acid (LTA) , the main component of cell walls from Gram-positive bacteria. Interestingly, it was also shown that these HIpA-LTA complexes could bind to heparan sulfate-proteoglycans on the surfaces of human epithelial cells in vitro (Stinson et al., 1998) .

Taken together, surface (LTA) -attached HIpA from Streptococcus bovis has affinity for surface components specifically expressed by colon tumor cells. The local recruitment of bacteria might eventually lead to silent tumor-associated infections as observed in the present study. To investigate whether Streptococcus bovis cells have indeed affinity for specific eukaryotic components, surface proteins from this bacterium grown in medium containing brain/heart extracts (Biarc et al., 2004), were first separated by ID SDS-PAGE. Next, in-gel tryptic digestion was performed on different fractions, after which the eluted peptides were applied to LTQ-FT MS for peptide separation and MS/MS analysis. As shown in Fig. 11, twelve peptides from these fractions corresponded to the eukaryotic tropomyosin protein (sequence coverage 43 %) .

Thus, brain/heart extract-derived tropomyosin, which is present in the bacterial growth medium, co-fractionates with Streptococcus bovis surface proteins. This indicates that, tropomyosin acts as a bacterial attachment molecule on heart tissue, which enables Streptococcus bovis to cause endocarditis .

Similarly, tropomyosin expressed and externalized by tumor cells might be involved in bacterial attachment, enabling Streptococcus bovis to cause a tumor-associated infections . g

Table III. Identification of S. bovis surface proteins by peptide mapping¹

§ a- Peptide hits / fraction

U NCBI ,_ , a- Protein ID j Mw (kDa) Organism accessions Vl ' B.S - 7- 6 IP

6-Phosphofructokinase 34.1 Lactobacillus delbrueckli subsp. bulgaricus gi|13629190 0 0 3 0 0

Chaperonin GroES 10.1 Bacillus anthracis gi|21398212 1 2 0 0 0

Class-ll aldolase 31.5 Streptococcus bovis gi| 10944298 4 2 5 2 0

Cold shock protein CapA/B 7.8 Pseudomonas sp gi| 13625473 0 2 1 0 0

Conserved hypothetical protein 9.1 Streptococcus pyogenes gi|15675694 0 1 1 0 0

DNA-binding protein HU (HIpA) 9.6 Streptococcus thermophilus gi|3334160 1 9 3 0 4

Enolase 47.5 Staphylococcus epidermidis gi|27467479 4 2 3 4 0

Glyceraldehyde-3-phosphate dehydrogenase 34.0 Streptococcus dysgalactiae subsp. equislmilis gi|2494648 6 3 2 1 0

Phosphocarrier HPr 8.8 Streptococcus mutans gi|546175 3 1 1 0 0

Putative translation initiation factor IF-1 8.3 Streptococcus pyogenes gi| 15674309 0 3 2 0 0

Pyruvate kinase 54.4 Streptococcus pneumoniae gi|15900780 0 3 2 0 0

Ribosomal protein L11 14.8 Streptococcus mutans gι|24380007 0 0 2 0 0

Ribosomal protein L20 13.6 Streptococcus pyogenes gl|15674848 0 3 0 0 0

Ribosomal protein L27 10.4 Streptococcus pyogenes gi|15674862 1 0 0 1 0

Ribosomal protein L29 8.0 Streptococcus mutans gi|24380361 1 0 1 0 0

Ribosomal protein L31 9.8 Streptococcus pyogenes gi| 15674775 0 1 1 0 0

Ribosomal protein L7/L12 12.5 Streptococcus thermophilus gi|55822507 3 5 7 3 6

Ribosomal protein L9 16.5 Streptococcus pyogenes gi|15675918 0 0 0 2 0

Ribosomal protein S13 13.4 Lactococcus lactis subsp. Lactis gi|15674052 0 0 2 0 0

Ribosomal protein S20 8.9 Streptococcus agalactiae gi|22537111 0 1 0 1 0

Ribosomal protein S6 11.0 Streptococcus pyogenes gi| 15¹675658 4 1 0 0 0

¹ Fractions 10, 8.5, 7 and 6 correspond to gel-eluted protein fractions as indicated in Figure 5. Number of peptide hits in each gel-eluted fraction corresponding to sequences present in bacterial databases, and their molecular weight, is indicated. Organism from which the seqence with the best/most protein hits were derived (and corresponding accesion numbers) are indicated. Number of peptide hits found by immunoprecipitation (IP)-based protein mapping, and the corresponding r- proteins are indicated in bold. It should be realized that several protein ID's were obtained with peptide hits in homologous proteins from different organisms.

©

Table IV. Peptide sequences from S. bovis surface proteins as identified by peptide mapping¹ fraction

Number/ Name Protein < Mw [Da] NCBl peptide sequence 10 8.5 IP

1. HIpA DNA-bindlng protein HU [Streptococcus thermophilic] 9599 gi|3334160 ANKQDLIAK 1 1

DNA-blπd(ng protein HU [Streptococcus thermophilus] 9599 gl|3334160 MANKQDLIAK 1 DNA-blnding protein HU [Streptococcus thermophilus] 9599 gl|3334160 QDLIAKVAEATELTK 1 DNA-bindlng protein HU [Streptococcus thermophilus] 9599 g!|3334160 VAEATELTK 1 1 1 DNA-binding protein HU [Streptococcus thermophilus] 9599 gi|3334160 VAEATELTKK 1 1 DNA-binding protein HU [Streptococcus thermophilus] 9599 gi|3334160 VQLIGFGNFEVR 1 1 DNA-binding protein HU [Bacillus cereus ATCC 14579] 9723 gl|30021B22 NVAQNAEISQK 1 Bac_DNA_binding, Bactenal DNA-blnding protein [Bacillus anthracis A2012] 9636 gl|21399426 GRNPQTGEEIEIAASK 1 1 DNA-bindlng protein HU [Enterococcus faecalls V583] 9650 gi|29376113 GRNPQTGQEIQIAASK 1 Bac_DNA_bindlng, Bacterial DNA-bindlng protein [Bacillus anthracis A2012] 9636 gl|21399426 NPQTGEEIEIAASK 1 1

1 9 5 0 2

UJ

OO

2.Rp L7/L12 rlbosomal protein L7/L12 [Streptococcus pneumoniae TIGR4] 12435 gl|15901208 AKLEEAGASVTLK 1 rlbosomal protein L7/L12 [Streptococcus pneumoniae TIGR4] 12435 gl|15901208 LEEAGASVTLK 1 1 1 nbosomal protein L7/L12 [Streptococcus agalactlae 2603V/R] 12322 gl|22537448 DSFDVELTAAGDK 1 1 1 1 1 ribosomal protein L7/L12 [Streptococcus agalactlae 2603V/R] 12322 gl|22537448 DSFDVELTAAGDKK 1 1 1 1 nbosomal protein L7/L12 [Streptococcus agalactlae 2803V/R] 12322 gi|22537448 EITGEGLK 1 1 1 1 5OS nbosomal protein L7/L12 [Streptococcus pyogenes M1 GAS] 12249 gl|15675065 EGVAAAEAEEIK 1 1 1 1 , 5OS ribosomal protein L7/L12 [Streptococcus mutans UA159] 12406 gi|24379403 EGVAAAEAEELK 1 1 1 1 ribosomal protein L7/L12 [Coxiella burnetii] 4278 gi|3660648 KSLEEAGAK 1

6 6 7 5 2

3. GAPDH glyceraldehyde-3-phosphate dehydrogenase [Streptococcus thermophllus] 36174 gl|17066732 AGAANIVPNSTGAAK 1 glyceraldehyde-3-phosphate dehydrogenase [Streptococcus thermophilus] 36174 gi|17066732 DGGFEVNGK 1 glyceraldehyde-3-phosphate dehydrogenase [Streptococcus thermophilus] 36174 gi|17066732 INDLTDPVMLAHLLK 1 glyceraldehyde-3-phosphate dehydrogenase [Streptococcus thermophilus] 36174 gl|17066732 KWITAPGGNDVK 1 glyceraldehyde-3-phosphate dehydrogenase [StreptococcΛis thermophllus] 36174 gi|17066732 NVTVDEVNAAMK 1 glyceraldehyde-3-phosphate dehydrogenase [Streptococcus thermophilus) 36174 gl|17066732 WITAPGGNDVK 1 glyceraldehyde-3-phosphate dehydrogenase [Streptococcus pneumoniae!^' 33010 gi|12619268 FDGTVEVK 1 glyceraldehyde-3-phosphate dehydrogenase - Bacillus coagulans 7267 gl!80089 VGIDGFGR 1

6 0 2 0 0

4. CII-AId class-ll aldolase [Streptococcus bovfe] 31467 gl| 10944298 FVAEYEANEEEYDK 1 1 class-ll aldolase [Streptococcus bovts] 31467 gl|10944298 FVAEYEANEEEYDKKK 1 class-ll aldolase [Streptococcus bovis] 31467 gi|10944298 IDVFGSANK 1 class-ll aldolase [Streptococcus bovis] 31467 gl|10944298 IDVFGSANKA 1 1 1 class-ll aldolase [Streptococcus bovis] 31467 gl|10944298 KAPILIQTSMGAAK 1 1 1 1

3 1 5 2

5. Eno enolase [Streptococcus pneumdniae TIGR4] 47131 gi|15900994 IEDQLGEVAEYR 1 1 SGETEDSTIADIAVATNA enolase [Streptococcus pneumoniae TIGR4] 47131 gl|15900994 GQIK 1 enolase [Streptococcus agalactiae 2603V/R] 47159 gi|22536801 GLETAVGDEGGFAPK 1 enolase [Streptococcus agalactiae 2603V/R] 47159 gi|22536801 SIITDVYAR 1 enolase [Staphylococcus epidermidis ATCC 12228] 47275 gi|27467479 FEGTEDAVETIIK 1 1 1 1

3 1 2 3

5. 3OS Rp S6 3OS ribosomal protein S6 [Streptococcus pyogenes M1 GAS] 11075 gll15675658 AKYEILYIIRPNIEEEAK 1 U) 3OS ribosomal protein SS [Streptococcus pyogenes M1 GAS] 11075 gi|15675658 VD

FDSILTDNGATWESK 1 1 3OS ribosomal protein S6 [Streptococcus pyogenes M1 GAS] 11075 gl(15675658 INGDILR 1 3OS ribosomal protein SS [Streptococcus pyogenes M1 GAS] 11075 gl|15675658 YEILYIIRPNIEEEAK 1

4 1 0 0

ASKDFHIVAETGIHARPA

7. HPr heat-stable phosphocarrier protein, HPr [Streptococcus mutans, Ingbritt, Peptide, 86 aa] 8800 gi|546175 TLLVQTASK 1 heat-stable phosphocarrier protein, HPr [Streptococcus mutans, Ingbritt, Peptide, 86 aa] 8800 gl|546175 FASDITLDYK 1 1 1 heai-stable phosphocarrier protein, HPr [Streptococcus mutans, Ingbritt, Peptide, 86 aa] 8800 gi|546175 FASDITLDYKGK 1

3 1 1 C o o

8. PKaεe pyruvate kinase [Streptococcus mutans UA159] 54390 gi|24379618 FNFSHGDHAEQGER 1 pyruvate kinase [Streptococcus mutans UA159] 54390 gi|24379618 IVATLGPAVEIR 1 1 pyruvate kinase [Streptococcus mutans UA159] 54390 gll24379618 TEWASAVK 1

0 2 2 C

9. TIF-1 putative translation Initiation factor IF-I [Streptococcus pyogenes Mt GAS] 8267 gi|15674309 AKEDVIEIEGK 1 putative translation initiation factor IF-1 [Streptococcus pyogenes M1 GAS] 8267 gi|15674309 VTVEMSPYDLTR 1 1 translation initiation factor IF-1 [Bradyrhizoblum japonicum USDA 110] 10838 gi|27376557 VTVEMSPYDLEK 1

0 2 2 C

10. CapAB cold acclimation protein CapBi[Pseι!ιdomonas sp. 30/3] 7691 gl|13625473 EGQQVSFIATR

COLD SHOCK PROTEIN CAPA (COLD ACCLIMATION PROTEIN A) (C7.0) 7011 gi|2493771 AIESDGFK major cold-shock protein [Bacillus cereus] 5004 gi|1402737 TLEEGQEVTFEVEQGNR

11. Cpn10 cpπ10, Chapβromn 10 Kd subunit [Bacillus anthracls A2012] 10063 gi|21398212 WIELVQAEEK cpn10, Chaperonin 10 Kd subun'it {Bacillus anthracis A2012] 10063 gl]21398212 YEGTDYLILR 10 kDa chaperonin (Protein Cpn10) (groES protein) 9736 gl|29839352 TLTGELIALSVAAGDK

12. 5OS Rp L9 5OS ribosomal protein L9 [Streptococcus pyogenes M1 GAS] 16502 gι|15675918 EVPTGYAQNFLIK 5OS ribosomal protein L9 [Streptococcus pyogenes M1 GAS] 16502 gi|15675918 VIFLADVK πbosomal protein L9 [Streptococcus agalactiae 2603V/R] 16763 gl|22538274 EVPTGYAQNFLLK

13. 30S Rp S15 3OS ribosomal protein S15 [Streptococcus pyogenes M1 GAS] O

10497 gi|15675753 KNEIIAQYAR 3OS ribosomal protein S15 [Streptococcus pyogenes M1 GAS] 10497 gi|15675753 NEIIAQYAR 3OS ribosomal protein S15 [Streptococcus pyogenes M1 GAS] 10497 gi|15675753 NLLAYLR

14. 5OS Rp L20 5OS ribosomal protein L20 [Streptococcus pyogenes M1 GAS] 13614 gl|15674848 KLWITR 5OS ribosomal protein L20 [Streptococcus pyogenes M1 GAS] 13614 gi|15674848 LAEIEVNR 5OS ribosomal protein L20 [Streptococcus pyogenes M1 GAS] 13614 gi|15674848 LAEIEVNRK

0 0 0

15. PrP DNA binding protein starved cells-like peroxide resistance protein - Streptococcus mutans 19605 gi|11356483 AVLNQAVADLSK putative peroxide resistance protein [Streptococcus pyogenes M1 GAS] 19335 gi|15675430 TIWMLQAER

1 0 0

16. 3OS Rp S5 3OS ribosomal protein S5 [Streptococcus pyogenes M1 GAS] 17017 gl|15674304 SLGSNTPINIVR 1 1 ribosomal protein S5 [Desulfovibrio vulgaris subsp. vulgaris str. Hildenborough] 17144 gi[46579731 AQEVPEALR

1 1 0

17. 5OS Rp L29 50s ribosomal protein L29 [Streptococcus mutaπs UA159] 8025 gl|24380361 FQAMGQLDQTAR

5OS πbosomal protein L29 [Streptococcus pyogenes M1 GAS] 7957 gl|15674295 LQEIKDFVK

18. 6-Phfk 6-phosphofructokiπase [Lactobacillus delbrueckli subsp. bulgaricus] 34158 gi|547755 IGILTSGGDAPGMNAAVR

6-phosphofructokinase [Streptococcus (hemophilus] 35981 gi|13629190 IAVLTSGGDAPGMNAAVR

19. 3OS Rp S13 3OS ribosomal protein S13 [Lactococcus lactis subsp. lactis 111403] 13483 gi|15674052 DLTSDQEDAIR 3OS ribosomal protein S13 [Lactococcus lactis subsp. lactis 111403] 13483 g!|15674052 DLTSDQEDAIRR

20. 5OS Rp L11 5OS ribosomal L11 protein [Streptococcus mutans UA159] 14807 gi|24380007 SFDFVTK 5OS ribosomal L11 protein [Streptococcus mutans UA159] 14807 gl|24380007 TADQAGMIIPWISVYEDK

21. 5OS Rp. L15 5OS ribosomal protein L15 [Streptococcus pyogenes M1 GAS] 15420 gi[15674306 VLGNGELTK

COG0200: Ribosomal protein L15 [Pediococcus pentosaceus ATCC 25745] 15434 gi|48871237 VLGNGEITK

0 0 ^

22 5OS ribosomal protein L27 [Streptococcus pyogenes M1 GAS] 10398 gi|15674862 AADGQTVSGGSILYR 1 23 phosphoenolpyruvate carboxylase [Synechococcus sp. WH 8102] 114491 gi|33866579 QLEEEIR 24 30S ribosomal protein S20 [Streptococcus pyogenes M1 GAS] 8307 gi|15675196 AFEANPSEELFR 25 5OS ribosomal protein L31 [Streptococcus pyogenes M1 GAS] 9848 gl|15674775 ETVEFEGETYPLIR 26 COG2771: DNA-binding HTH domain-containing proteins [Burkholderia cepacia R18194] 20322 gi|46311120 VAVELTPR 27 fructose-Wsphosphate aldolase [Streptococcus mltis] 18993 gi|3777455 GLHLDHLQK 1 28 hypothetical protein SPy2005 [Streptococcus pyogenes M1 GAS] 6957 gl|15675792 EAVEGAVDAVK 29 hypothetical protein; putative signal peptide [Acinetobacter sp. ADP1] 19876 gi|50085542 VATPVEIR 30 putative secreted protein [Streptococcus mutans] 22790 gl|4098503 SVSDAINR 1 31 pyrimidine operon regulatory protein [Streptococcus agalactiae 2603V/R] 19788 gi|22537511 NLDNIVLAGIK 32 30S ribosomal protein S16 [Lactococcus lactis subsp. lactis 111403] 10280 gl|15673551 VLEWLSK 1 33 5OS ribosomal protein L1 [Streptococcus pyogenes M1 GAS] 24379 gi|15674580 AAGADFVGEDDLVAK 34 5OS ribosomal protein L18 [Streptococcus pyogenes M1 GAS] 12858 gl|15674303 TEQAWVGK 1 35 50S ribosomal protein L2 [Streptococcus pyogenes M1 GAS] 29915 gl|15674290 TANIALVHYTDGVK 1 36 5OS ribosomal protein L24 [Streptococcus mutans UA159] 10964 gi|24380358 VWEGVAIVK

!

COG0112: Glycine/serine hydiOxymethyltransferase [Clostridium thermocellum ATCC

37 27405] 45408 gi|48859801 AIELEVNR

COG1670: Acetyltransferases,' including N-acetylasβs of ribosomal proteins [Oenococcus

38 oeni PSU-1] ' 23772 gl!48866060 DEIEFINYAR 39 COG3279: Response regulatof of the LytR/AlgR family [Desulfitobacterium hafniense] 28022 gl|23113590 DYAMVGSARIPVSK COG3973: Superfamlly I DNA and RNA helicases [Pediococous pβntosaoeus ATCC

40 25745] 88419 gi|48869854 TEWQLSK 41 connector protein [Streptococcus phage Cp-1] 39535 gi|9629535 VIEELHK 42 conserved hypothetical protein [Streptococcus agalactlae 2603V/R] 9558 g!|22537688 EKITNDFEK 43 conserved hypothetical protein [Streptococcus agalactiae 2603V/R] 9785 gi|22536455 LVDDEGNDVTPEK 44 conserved hypothetical protein [Streptococcus pyogenes M1 GAS] 9141 gi|15675694 VFYQETK 45 conserved hypothetical protein [Streptococcus pyogenes M1 GAS] 12437 gi|15675518 DIFEQEFK

Diguanylate cyclase/phosphodiesterase domain 1 (GGDEF) [Thermoanaerobacter

46 tengcongensis] 77518 giI20807566 EMEKELK 47 glutamine synthetase type 1; glutamate— ammonia ligase [Streptococcus mutans UA159] 50094 gi|24378861 AMFDGSSIEGFVR 48 GroEL [Streptococcus Intermedius] 56240 gi|21666293 SFGSPLITNDGVTIAK 49 hypothetical protein [Streptococcus bovls] 28135 gi|40714540 GYFSGNDYILEK 50 hypothetical protein [Streptococcus thermophllus] 12135 gi|8574401 IVPTLLAK 51 hypothetical protein TM0737 [Thermotoga maritime MSB8] 46581 gi|15643500 ASITNVLGK to 52 iron(lll) ABC transporter, solute-binding protein [Methanosarcina acetivorans str. C2A] 45995 giI20091600 IVIDSRGVEVK 53 proton-translocating ATPase, beta subunit [Streptococcus bovis] 51187 gi|2662326 TIAMESTDGLTR 54 putative phσsphoglycerate kinase [Streptococcus intermedius] 8581 gi|14571812 SIIGGGDSAAAAINLGR 55 putative polypeptide deformylase [Streptococcus pyogenes M1 GAS] 22962 gi|15675756 WEGYWR 56 putative ribosome-associated protein [Streptococcus mutans UA159] 21073 gl|24378984 YFNAEQELDAR 57 pyruvate kinase [Streptococcus pneumoniae TIGR4] 54777 gi|15900780 NAQALLNEYGR 58 ribosomal protein L15 [Streptococcus agalactiae 2603V/R] 15441 gi|22536262 SAEAAITAK 59 single-stranded DNA binding protein [Bacillus phage Nf] 13139 gl|7960759 SSKGNEFFSLLLVG 60 unnamed protein product [Kluyveromyces lactis] 34313 gl|50309305 QIAQQDAQK 61 unnamed protein product [Kluyveromyces lactis] 35185 gl|50308875 LGSLETLEAGK 62 unnamed protein product [Kluyveromyces lactis] 69808 gi|50306557 VFTPEEISSMILSK

' Fractions 10, 8.5, 7 and 6 correspond to gel-eluted protein fractions as indicated in Figure 5. peptide sequences, identified in each of these fractions and/or by immunoprecipitation (IP)-based protein mapping, corresponding to protein sequences present in bacterial databases, their molecular weight, and corresponding accesion numbers are indicated.

Table V Partial amino acid sequences of identified antigens of Streptococcus bovis

~» ' u \J o J

45

REFERENCES

Ahlquist, D. A. (1997) Fecal occult blood testing for colorectal cancer: can we afford to do this? Gastroenterol. Clin. North Am. 26: 41-55.

Baggerly, K. A., Morris, J. S., Coombes, K. R. (2004) Reproducibility of SELDI-TOF protein patterns in serum: comparing datasets from different experiments. Bioinformatics 20: 777-785.

Biarc, J., Nguyen, I. S., Pini, A., Gosse, F., Richert, S., Thierse, D., van

Dorsselaer, A., Leize-Wagner, E., Raul, F., Klein, J. P., Scholler-Guinard, M (2004) Carcinogenic properties of proteins with pro-inflammatory activity from Streptococcus infantarius (formerly S. bovis) . Carcinogenesis. 25: 1477-1484.

Bjorck, L. and Kronvall, G. (1984) Purification and some properties of streptococcal protein G, a novel IgG-binding reagent. J. Immunol. 133: 969-974.

Darjee, R., and Gibb, A. P. (1993) Serological investigation into the association between Streptococcus bovis and colonic cancer. J. Clin. Pathol. 46: 1116-1119.

Davies, D. H., Liang, X., Hernandez, J. E., Randall, A., Hirst, S., Mu, Y., Romero, K.M., Nguyen, T. T., Kalantari-Dehaghi, M., Crotty, S., Baldi, P., Villarreal, L. P., and Feigner, P. L.. (2005) Profiling the humoral immune response to infection by using proteome microarrays: High-throughput vaccine and diagnostic antigen discovery. Proc. Natl. Acad. Sci. U.S.A. 102: 547-552

Drlica, K., and Rouviere-Yaniv, J. (1987). Histone-like proteins of bacteria. Microbiol. Rev. 51: 301-319.

Eisen, M. B., Spellman, P. T., Brown, P.O., and Botstein, D.

(1998) Cluster analysis and display of genome-wide expression patterns. Proc. Natl. Acad. Sci. U.S.A. 95: 14863-14868.

Ellmerich, S., Scholler, M., Duranton, B., Gosse, F.,

Galluser, M., Klein, J. P., Raul, F. (2000) Promotion of intestinal carcinogenesis by Streptococcus bovis.

Carcinogenesis 21: 753-756.

Etzioni, R., Urban, N., Ramsey, S., Mcintosh, M., Schwartz,

S., Reid, B., Radich, J., Anderson, G., and Hartwell, L.

(2003) The case for early detection. Nat. Rev. Cancer. 3:

243-252.

Gray D, and Skarvall, H. (1998) B-cell memory is short-lived in the absence of antigen. Nature 336: 70-73.

Grus, F. H., Joachim, S. C, and Pfeiffer, N. (2003) Analysis of complex autoantibody repertoires by surface-enhanced laser desorption/ionization-time of flight mass spectrometry. Proteomics 3: 957-961.

Guarner, F., and Malagelada, J. R. (2003) Gut flora in health and disease. Lancet 361: 512-519. Kesari, K. V., Yoshizaki, N., Geng, X., Lin, J.J., and Das,

K. M. (1999) Externalization of tropomyosin isoform 5 in colon epithelial cells. Clin. Exp. Immunol. 118: 219-227.

Klein, R. S., Recco, R. A., Catalano, M. T., Edberg, S. C, Casey, J.I., and Steigbigel, N. H. (1977) association of Streptococcus bovis with carcinoma of the colon. N. Engl. J. Med. 297: 800-802.

Knudtson, L.M., and Hartman, P.A. (1993) Comparison of fluorescent gentamicin-thallous-carbonate and KF streptococcal agars to enumerate enterococci and fecal streptococci in meats. Appl. Environ. Microbiol. 59: 936-938

Lasonder, E., Ishihama, Y., Andersen, J. S., Vermunt, A.M., Pain, A., Sauerwein, R.W., Eling, W.M., Hall, N., Waters, A. P., Stunnenberg, H. G., Mann, M. (2002) Analysis of the Plasmodium falciparum proteome by high-accuracy mass spectrometry. Nature 419: 537-542.

Loeve, F., Boer, R., Zauber, A. G., Van Ballegooijen, M., Van Oortmarssen, G.J., Winawer, S.J., and Habbema, J. D. (2004) National Polyp Study data: evidence for regression of adenomas. Int. J. Cancer 111: 633-639.

Moore, W. E., and Moore, L. H.. (1995) Intestinal floras of populations that have a high risk of colon cancer. Appl. Environ. Microbiol. 61: 3202-3207. Parsonnet, J. (1995) Bacterial infection as a cause of cancer. Environ. Health Perspect. 103:263-268.

Rappsilber, J., Friesen, W. J., Paushkin, S., Dreyfuss, G., and Mann, M. (2003) Detection of arginine dimethylated peptides by parallel precursor ion scanning mass spectrometry in positive ion mode. Anal Chem. 75: 3107-3114.

Rakoff-Nahoum S., Paglino, J., Eslami-Varzaneh, F., Edberg, S., and Medzhitov, R. (2004) Recognition of commensal microflora by toll-like receptors is required for intestinal homeostasis. Cell 118: 229-241.

Schagger, H., and von Jagow, G. (1987) Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 166: 368-379.

Schlegel, L., Grimont, F., Ageron, E., Grimont, P.A., and Bouvet, A. (2003) Reappraisal of the taxonomy of the

Streptococcus bovis/Streptococcus equlnus complex and related species: description of Streptococcus gallolyticus subsp. gallolyticus subsp. nov. _r S. gallolyticus subsp. macedonicus subsp. nov. and S. gallolyticus subsp. pasteurianus subsp. nov. Int. J. Syst. Evol . Microbiol. 53: 631-645.

Schδller, M., Klein, J. P. and Frank, R.M. (1981) Common antigens of streptococcal and non-streptococcal oral bacteria: immunochemical studies of extracellular and cell-wall-associated antigens from Streptococcus sanguis, Streptococcus mutans, Lactobacillus salivarius and Actinomyces viscosus. Infect. Immun. 31: 52-60.

Stinson, M. W., McLaughlin, R., Choi, S. H., Juarez, Z. E., and Barnard, J. (1998) Streptococcal histone-like protein: primary structure of hlpA and protein binding to lipoteichoic acid and epithelial cells. Infect Immun. 66: 259-265. Tjalsma, H., and van Dijl, J.M. (2005) Proteomics-based consensus prediction of protein retention in a bacterial membrane. Proteomics, in press.

Tjalsma, H., Antelmann, H., Jongbloed, J. D. H., Braun, P. G. Darmon, E., Dorenbos, R., Dubois, J.Y.F., Westers, H., Zanen, G., Quax, W.J., Kuipers, O.P., Bron, S., Hecker, M., and van Dijl, J.M. (2004) . Proteomics of protein secretion by Bacillus subtilis: separating the "secrets" of the

Welliver, R. C, Kaul, T.N., Putnam, T.I., Sun, M., Riddlesberger, K., and Ogra, P. L. (1980) The antibody response to primary and secondary infection with respiratory syncytial virus: kinetics of class-specific responses. J. Pediatr. 96: 808-813.

Wendum, D., Masliah, J., Trugnan, G., and Flejou, J. F. (2004) Cyclooxygenase-2 and its role in colorectal cancer development. Virchows Arch. 445: 327-333. Zolkowski, K., Gruszewska-Lewczuk, L., and Zwierz, K. (1992) The serum immunoglobulins level in maxillofacial bacterial infections. Rocz. Akad. Med. Bialymst. 37: 64-70.

Claims

1. Method for diagnosing colon cancer in an individual comprising detecting the presence of immunoglobulin molecules directed against Streptococcus bovis antigens in a sample of the individual comprising total serum immunoglobulin molecules.

2. Method according to claim 1, wherein the sample of the individual is contacted with a preparation comprising one or more antigens of Streptococcus bovis and binding is detected between the immunoglobulin molecules and the antigens .

3. Method according to claim 2 comprising:

— immobilizing the total serum immunoglobulin molecules on a solid support;

— contacting the immobilized total serum immunoglobulin molecules with the

Streptococcus bovis preparation to capture Streptococcus bovis antigens;

4. Method according to claim 3, wherein the presence of Streptococcus bovis antigens in the eluate is detected by mass spectrometry (MS) .

5. Method according to claim 4, wherein the mass spectrometer is equipped with a surface-enhanced laser desorption/ionization time-of-flight (SELDI-TOF) reader.

6. Method according to any of the claims 1 to 5, wherein the diagnosis of colon cancer comprises diagnosis of an pre or early stage of colon cancer.

7. Method according to any of the claims 1 to 6, wherein the diagnosis of colon cancer is an in vitro diagnosis of colon cancer.

8. Method according to any of the claims 1 to 7, wherein the sample of the individual is serum.

9. Method according to any of the claims 1 to 8, wherein the antigens of Streptococcus bovis are selected from the group consisting of HIpA, Rp L7/L12, GAPDH, class II aldolase, Eno, 3OS ribosomal protein S6, Hpr, Pkase, TIF-I, CapAB, CpnlO, 5OS ribosomal protein L9, 30 S ribosomal protein S15, 5OS ribosomal protein L20, PrP, 3OS ribosomal protein S5, 5OS ribosomal protein L29, 6-Phfk, 3OS ribosomal protein S13, 5OS ribosomal protein LIl, and 5OS ribosomal protein L15, 5OS ribosomal protein L27, phosphoenolpyruvate carboxylase, 3OS ribosomal protein S20, 5OS ribosomal protein L31, DNA-binding HTH domain-containing protein, major cold shock protein, CsbD-like protein, secreted protein, SagA-like protein, pyrimidine operon regulatory protein, 3OS ribosomal protein S16, 5OS ribosomal protein Ll, 50 S ribosomal protein L18, 5OS ribosomal protein L2, 5OS ribosomal protein L24, Glycine/serine hydroxymethyltransferase, acetyltransferase, N-acetylase of ribosomal proteins, response regulator of the LytR/AlgR family, superfamily I DNA and RNA helicase, connector protein, transcription regulator, cell division initiation protein, diguanylate cyclase/phosphodiesterase domain 1, glutamine synthetase type 1; glutamate-ammonia ligase, GroEL, nitrogen regulatory protein P-II, phophoesterase, iron (III): ABC transporter, solute-binding protein, proton-translocating ATPase: beta subunit, phosphoglycerate kinase, polypeptide deformylase, ribosome- associated protein, pyruvate kinase, ribosomal protein L15, single-stranded DNA binding protein, prohibitin-like protein, uricase, and Hsp70-like protein.

10. Method according any of the claim 9, wherein the antigens of Streptococcus bovis are selected from the group consisting of HIpA, Rp L7/L12, GAPDH, class II aldolase, Eno, 3OS ribosomal protein S6, Hpr, Pkase, TIF-I, CapAB, CpnlO, 5OS ribosomal protein L9, 30 S ribosomal protein S15, 50S ribosomal protein L20, PrP, 3OS ribosomal protein S5, 5OS ribosomal protein L29, 6-Phfk, 3OS ribosomal protein S13, 5OS ribosomal protein LlI, 5OS ribosomal protein L15; 5OS^' ribosomal protein L27.

11. Method according to claims 10, wherein the antigens of Streptococcus bovis are HIpA and/or Rp L7/L12.

12. Method according to any of the claims 1 to 11, wherein the antigen of Streptococcus bovis comprises an amino acid sequence or amino acid sequences selected from the group consisting of SEQ ID Nos : 1 to 3; 4 to 6; 7 to 12; 13 to 16; 17 to 20; 21 to 23; 24; 25 to 28; 29 to 30; 31 to 32; 33 to 35; 36 to 37; 38 to 39; 40 to 41; 42 to 43; 44 to 45; 46 to 47; 48, 49; 50; 51; 52; 53; 54; 55; 56; 57; 58; 59; 60; 61; 62; 63; 64; 65; 66; 67; 68 ;69; 70; 71; 72; 73; 74; 75; 76; 77; 78; 79; 80; 81; 82; 83; 84; 85; 86; 87; 88; 89; and 90.

13. Method according to claim 12, wherein the antigen of Streptococcus bovis comprises an amino acid sequence or amino acid sequences selected from the group consisting of SEQ ID Nos: 1 to 3; 4 to 6; 7 to 12; 13 to 16; 17 to 20; 21 to 23; 24; 25 to 28; 29 to 30; 31 to 32; 33 to 35; 36 to 37; 38 to 39; 40 to 41; 42 to 43; 44 to 45; 46 to 47; 48, 49; 50; 51; and 52.

14. Method according to claim 13, wherein the antigen of Streptococcus bovis comprises amino acid sequences selected from the group consisting of SEQ ID Nos: 1 to 3; and 4 to 6.

15. Method according to any of the claims 3 to 14, wherein the presence of Streptococcus bovis antigens is detected by the presence of one or more proteins and/or protein fragments with molecular weights of 7.8 kDa, 7.9 kDa or 9.6 kDa.

16. Protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 1 to 3.

17. Protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 4 to 6.

18. Protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 7 to 12.

19. Protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 13 to 16.

20 Protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 17 to 20.

21. Protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos : 21 to 23.

22. Protein of Streptococcus bovis comprising amino acid sequence SEQ ID No: 24.

23. Protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 25 to 28.

24. Protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 29 to 30.

25. Protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 31 to 32.

26. Protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 33 to 35.

27. Protein of Streptococcus bovis comprising ^' amino acid sequences SEQ ID Nos: 36 to 37.

28. Protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 38 to 39.

29. Protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 40 to 41.

30. Protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 42 to 43.

31. Protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 44 to 45.

32. Protein of Streptococcus bovis comprising amino acid sequences SEQ ID Nos: 46 to 47.

33. Protein of Streptococcus bovis comprising an amino acid sequence according to SEQ ID No: 48.

34. Protein of Streptococcus bovis comprising an amino acid sequence according to SEQ ID No: 49.

35. Protein of Streptococcus bovis comprising an amino acid sequence according to SEQ ID No: 50.

36. Protein of Streptococcus bovis comprising an amino acid sequence according to SEQ ID No: 51.

37. Protein of Streptococcus bovis comprising an amino acid sequence according to SEQ ID No: 52.

38. Method according to any of the claims 2 to 15, wherein the antigens of Streptococcus bovis are selected from the group consisting of the proteins as defined in claims 16 to 37.

39. Use of a protein of Streptococcus bovis according to any of the claims 16 to 37 for the diagnosis of colon cancer.

40. Use according to claim 39, wherein the diagnosis of colon cancer is in vitro diagnosis of colon cancer.

41. Protein of Streptococcus bovis according to any of the claims 16 to 37 for use as a diagnosticum.

42. Use of a protein according to any of the claims 16 to 37 for the preparation of a medicament for the prophylaxis and/or treatment of colon cancer.

43. Use according to claim 42, wherein the prophylaxes comprises vaccination.

44. Use according to claim 42, wherein the treatment comprises drug targeting.