MCP
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/mcp.R600016-MCP200 on March 20, 2007.
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
R600016-MCP200v1
6/7/1115    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Glossary
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Caron, M.
Right arrow Articles by Joubert-Caron, R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Caron, M.
Right arrow Articles by Joubert-Caron, R.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
Molecular & Cellular Proteomics 6:1115-1122, 2007.
© 2007 by The American Society for Biochemistry and Molecular Biology, Inc.

Review

Cancer Immunomics Using Autoantibody Signatures for Biomarker Discovery*

Michel Caron{ddagger},§, Geneviève Choquet-Kastylevsky and Raymonde Joubert-Caron{ddagger}

From the {ddagger} Protein Biochemistry and Proteomics Laboratory, UMR CNRS 7033 (BioMoCeTi), Unité de Formation et de Recherche Santé-Médecine-Biologie Humaine, Paris 13 University, 93017 Bobigny cedex, France and New Markers Department, Advance Technology Unit, bioMérieux, 69280 Marcy-l'Etoile, France


   ABSTRACT
TOP
 ABSTRACT
SEROLOGICAL PROTEOME ANALYSIS
 MULTIPLE AFFINITY PROTEIN...
 MICROARRAY ANALYSES
 FROM AUTOANTIBODY SIGNATURES TO...
 REFERENCES
 
The increased incidence of autoantibodies in malignancies has been described since the 1970s. Thus the ability to determine molecular fingerprinting of autoantibodies (antibody signatures) may provide useful clinical diagnostic and prognostic information. This review describes the use of several proteomics approaches for the identification of antigens recognized by these autoantibodies. Serological proteome analysis combines separation of tumor cell proteins on two-dimensional gel electrophoresis gels, Western blotting with sera of patients and healthy subjects, and identification of the detected antigens by MS. Alternatively multiple affinity protein profiling combines isolation of the antigens recognized by patient antibodies by two-dimensional immunoaffinity chromatography and identification by MS/MS. The use and limitations of reverse phase protein microarrays for testing patient serum containing autoantibodies are also considered. Lastly the most important difficulty of any proteomically identified autoantibody signature is validation in patient cohorts or clinical samples.

Tumors are thought to induce the release of many proteins into the blood. Diagnosing cancer based on serum profiling is therefore an attractive concept. Particularly there is clear evidence that the immune system, in addition to defending us against pathogens, is also on guard against other threats, including tumors (1, 2). During our lifetime, it is quite possible that we experience several undetected cancerous or precancerous lesions that can further be eliminated from the organism thanks to self-defenses, including the immune system. Uncontrolled malignant growth will therefore be characterized by the presence of autoantibodies that precede clinical findings by months or years. The generation of circulating antibodies that bind self-protein can be regarded as the systemic amplification by the immune system of a signal that indicates the presence of the tumor (3, 4).

Several hypotheses have been proposed to explain the increased incidence of autoantibodies in malignancies, including host-immune reactions to tumor-associated antigens, antigenic stimulation as a result of the destruction of malignant cells, or immune dysregulation induced by the neoplastic process. At least four main categories of tumor-associated antigens can be proposed: those encoded by genes with tumor-specific expression, those resulting from point mutation (such as p53), differentiation proteins, and those encoded by genes that are overexpressed in certain tumors. Many of these antigens have first been identified on melanomas, and skepticism had been voiced regarding the applicability to other tumors (5). In addition to changes in the level of gene expression in cancer, an aberrant expression of tissue-restricted gene products has also been associated with the development of a humoral immune response (6, 7). Antigens resulting from point mutations have been found to be recognized by autologous T cells on renal cell carcinomas, head and neck squamous cell carcinomas, and bladder carcinomas (5). More generally, multiple studies have shown that patients with cancer produce detectable autoantibodies to certain tumor-associated antigens (8, 9). For example, published studies have identified autoantibodies to p53 (10), L-myc (11), glycosylated annexins I and II (12), or HER2-neu (13) in patients with malignancies.

Practically the serological identification of human tumor antigens that could be at the origin of patients’ autoantibodies was pursued as soon as the 1970s by Old and co-workers (14, 15). For this purpose, they developed an approach called autologous typing. Using autologous serum and cultured cell lines from cancer patients and extensive absorption analyses, a few antigens were detected and further defined with the techniques available in those days. Molecular characterization of these antigens was generally beyond reach primarily because the antibodies were not of high titer, and their clinical value remained subject to debate. Nevertheless the observation that patients’ sera sometimes specifically recognize tumor-associated antigens raises the possibility that detection of this humoral immune response could provide both diagnostic and prognostic information. Thus, the ability to determine a molecular fingerprinting of autoantibodies (autoantibody signatures) or of complementary antigens would be relevant for serological screening tests for malignancies. This ability can be termed "cancer immunomics" (12, 16–18).

The technical difficulties in defining human autoantibody signatures related to the development of tumors conducted to search for strategies that allowed the exploitation of the antibody repertoire for the systematic identification of complementary antigens. In the 1990s, Pfreundschuh and co-workers (8) developed the SEREX1 approach; the acronym stands for the serological analysis of tumor antigens by recombinant cDNA expression cloning. In the SEREX approach, a cDNA library constructed from fresh tumor specimen is expressed recombinantly, and the recombinant proteins transferred onto membranes are identified as tumor-associated antigens by their reactivity with IgG antibodies present in the patient's serum. This methodology allowed identification of several antigens. For example, NY-ESO-1, which is a candidate for immunotherapy, exhibits restricted expression patterns and the ability to elicit cell-mediated as well as humoral immune responses (19). Naora et al. (6) applied the SEREX methodology using ovarian cancer patient sera and isolated HOXB7, a tumor antigen that could have a significant role in the growth of ovarian carcinoma. In a study that applied the same methodology to colon cancer, Scanlan et al. (20) showed that the majority of the SEREX-defined antigens have no known association with cancer or with autoimmunity. With regard to patterns of gene expression, most of these antigens showed a large expression. Most of the SEREX-defined antigens are intracellular and normally not exposed to the immune system, and their release by tumor cells may render such proteins immunogenic (8, 20). A disadvantage of the SEREX approach is the necessity to construct expression libraries. In addition this approach fails to detect post-translational modifications.

More recently, proteomics, using a combination of sophisticated methods, provided new opportunities for screening and identifying autoantigens (12, 18, 21–23). Based on the different combinations that can be useful for immunomics, different types of workflows corresponding to those currently used in proteomics can be proposed (Fig. 1). Even if there are no absolute boundaries between these workflows, they can be helpful to define a strategy for immunomics marker discovery. Two different situations can be considered: either the aim of the analytical scheme is to discover new putative biomarkers (autoantigens and autoantibodies), or the aim is to assess the relevance of putative markers already described. On this basis, one may distinguish two ways for research: discovery-driven immunomics or hypothesis-driven immunomics. Current discovery-driven approaches follow usually two different workflows to characterize proteins within a sample: "top-down" or intact protein analysis approaches and "bottom-up" or complex proteolytic digestion approaches. Top-down approaches include differential 2-D gel electrophoresis and the combination of 2-D gel and 2-D Western blotting for autoantigen identification (serological proteome analysis (SERPA); see below). Bottom-up approaches include various protein and peptide separation schemes, such as combined ion exchange, reverse phase, and affinity chromatography. For immunomics studies these separation schemes focus on the purification of putative autoantigens; the mixture of these antigens are then digested and analyzed by nano-LC-MS/MS.


Figure 1
View larger version (22K):
[in this window]
[in a new window]

  		FIG. 1. Different types of workflow for screening and identifying antigens complementary to autoantibodies by immunomics. Aff. Chr., affinity chromatography.

 
Hypothesis-driven immunomics can potentially use two main approaches: immunotests (from classical ELISA to protein biochips) and research of autoantigens using a peptide signature. Protein biochips or protein arrays are fundamentally an arrangement of multiple miniaturized probe sites in a single, compact analytical surface. Arrays can be thought of as two-dimensional structures (glass slide or nitrocellulose surfaces, modified microtiter plates, etc.) as well as three-dimensional devices (interaction beads, nanoparticles, etc.) (24, 25). Detection of adsorbed proteins can be accomplished using a variety of direct or indirect schemes such as induced fluorescence, surface plasmon resonance, or mass spectrometric detection.


   SEROLOGICAL PROTEOME ANALYSIS
TOP
 ABSTRACT
 SEROLOGICAL PROTEOME ANALYSIS
MULTIPLE AFFINITY PROTEIN...
 MICROARRAY ANALYSES
 FROM AUTOANTIBODY SIGNATURES TO...
 REFERENCES
 
The term SERPA (21) has been proposed for a top-down method usable for discovery-driven immunomics. SERPA (also called PROTEOMEX by Seliger et al. (26, 27)) is based on a classical proteomics workflow associating an effective separation on 2-DE gels and an identification by MS. In SERPA, Western blotting in combination with 2-DE gels (2-D blotting) (18, 26, 28–30) permits the transfer and immobilization of proteins from tumor tissue or tumor cell lines to a semirigid support. Sera from cancer patients or healthy subjects are screened individually allowing immunodetection of relevant antigens among the several thousand individual proteins separated using 2-DE. Comparative probing of blots with sera from patients and healthy subjects may allow the characterization of proteins that react specifically with sera from cancer patients (17, 22, 23, 31, 32).

Our laboratory has embarked on a long term immunomics study of breast cancer (18, 32, 33) as this disease is the most common malignancy among women (34). During this study, we applied the SERPA approach to the identification of breast cancer proteins that elicit a humoral response. The methodology used to identify these proteins in cancer patients consisted of three essential steps: (i) the preparation of cancer cell extracts, (ii) the screening of autoantigens, and (iii) the identification of the relevant molecules (Fig. 2). Well characterized serum specimens were obtained at the time of diagnosis after informed consent and for controls from healthy volunteer women. To facilitate the comparative analysis of different sets of 2-D blots probed with individual sera, the 2-D blots were stained with colloidal gold after blocking (0.2% Tween 20) and immunodetection steps. The transfer efficiency and the total number of spots transferred onto the PVDF membrane were determined on colloidal gold-stained membrane using image analysis. Then it was possible to match the image of the blot against a preparative gel used for the identification by MS or against a reference map (18). The matching permitted exact localization of particular relevant protein spots hybridized by antibodies on the 2-D blots and their subsequent identification by MS.


Figure 2
View larger version (25K):
[in this window]
[in a new window]

  		FIG. 2. Methodology for identification of cancer-associated autoantigens using SERPA. Cells isolated from tumor or cell lines are used as sources of antigens. The proteins complementary to autoantibodies in sera from patients or healthy controls are screened individually by immunodetection (2-D blots) and then identified.

 
Examination of data shows the occurrence of autoantibodies of IgG isotype that are highly conserved between individuals. They are directed against a limited set of antigens independently of cancer status. Antibodies from breast cancer patients’ sera showed predominantly binding to proteins of ~25-kDa molecular mass (spot numbers 1, 3, 6, 11, and 14) and to proteins of ~45-kDa molecular mass (spot numbers 8, 9, 12, 13, 15, and 16). In Fig. 3, a window of a representative silver-stained 2D gel is shown. Marked are those proteins identified by antibodies. This window comprises proteins with isoelectric points ranging from pH 5.7 to 7.5 and molecular masses ranging from 60 to 23 kDa and contains all detected spots except spot number 7 detected at pH 6.9 and 100 kDa. The incidence of antibody binding to spot numbers 2, 3, 8, 9, 10, 11, 12, and 14 ranged from 65% up to 100% of breast cancer sera (Fig. 4). The incidence of antibodies binding to spot numbers 1, 6, 7, 9, 13, and 14 was particularly increased in the patients’ group compared with healthy controls (ratio >2). Only a few patients’ sera displayed antibody binding to spot numbers 15 and 16. None of the control sera displayed any immunostaining with these proteins. On the contrary, spot number 4 was recognized only by a few controls. This spot could well fit into a group of its own in terms of a protein species that may well be a surveillance target that is lost upon transition into a disease scenario.


Figure 3
View larger version (95K):
[in this window]
[in a new window]

  		FIG. 3. Window of silver-stained gel of two-dimensional separated proteins from breast cancer cell preparation. Molecular masses are indicated on the left side; isoelectric points are indicated above the gel. Those proteins identified by antibodies from patients’ sera during SERPA investigations are given spot numbers 1–16. Spot number 7 (molecular mass, ~100 kDa) is not seen in this window.

 

Figure 4
View larger version (27K):
[in this window]
[in a new window]

  		FIG. 4. Incidence of autoreactive antibodies against breast cancer cell proteins in sera from patients with breast cancer and healthy donors. RP, reverse phase.

 
The IgG isotype of these antibodies indicated a secondary response after a long incubation period characterized by the appearance of autoantibodies while lacking clinical manifestations. This study, like other studies described up until now (12, 22, 35–37), used sera taken at the time of diagnosis before any treatment. Thus, it is reasonable to presume that such antibodies could be used as an early indicator and that the SERPA approach appears promising for discovering new potential biomarkers on the basis of autoantibody signatures (2, 17, 38, 39).


   MULTIPLE AFFINITY PROTEIN PROFILING
TOP
 ABSTRACT
 SEROLOGICAL PROTEOME ANALYSIS
 MULTIPLE AFFINITY PROTEIN...
MICROARRAY ANALYSES
 FROM AUTOANTIBODY SIGNATURES TO...
 REFERENCES
 
The term multiple affinity protein profiling has been proposed for a chromatography-based method aimed at the purification of putative autoantigens; the mixture of these antigens are then digested and analyzed by nano-LC-MS/MS (40). The experimental overview is shown in Fig. 5. The proteins from lysed cells are subjected to 2-D immunoaffinity separation according to their affinity for immunoglobulins from healthy controls in the first dimension and immunoglobulins from patients in the second dimension. The affinity supports used during these separations are prepared by coupling an IgG fraction isolated from the sera of healthy volunteers (first dimension) or patients (second dimension) (41). The first immunoaffinity chromatography is crucial as it is used for selectively removing proteins (autoantigens) recognized by healthy volunteers’ antibodies. The proteins that cannot bind to these antibodies are subsequently discarded in the flow-through. After recovery of all proteins that are not bound, retained proteins can be eluted using buffer modifiers (such as acidic buffer) with subsequent analysis by tandem mass spectrometry. The second immunoaffinity chromatography is used for the selection of antigens recovered in the flow-through of the first chromatography and recognized by the antibodies of patients. Several columns can be used in parallel, allowing the analyses in the same run on different columns corresponding to different patients in comparison with columns of controls. Retained proteins can then be eluted, enzymatically digested, and analyzed by on-line tandem MS (ESI-ion trap, ESI-Q-TOF, ESI-Orbitrap, etc.). Currently the method involves an accurate and reproducible separation of tryptic digests of proteins isolated by immunoaffinity based on an integrated microfluidics device (42) with on-line detection by ESI-ion trap tandem MS (43). Starting from this simple workflow one can extend the approach to various clinical situations.


Figure 5
View larger version (26K):
[in this window]
[in a new window]

  		FIG. 5. Experimental scheme of the immunoaffinity liquid phase 2-D separation technique followed by tandem MS analysis and identification of proteins of cancer cell lines with specificity for control or patient antibodies. The proteins of the cancer cell lysate recognized by the autoantibodies of healthy controls are adsorbed during the first dimension immunoaffinity chromatography. The antigens recognized by the antibodies of each patient are purified from the depleted extract (flow-through of the first dimension) during the second dimension immunoaffinity. Several second dimension columns can be used in parallel, each corresponding to a different patient. Following enzymatic digestion of the eluted proteins, the peptides are separated in the HPLC-chip microfluidics device, and MS/MS spectra are obtained on line from the ion trap (IT).

 

   MICROARRAY ANALYSES
TOP
 ABSTRACT
 SEROLOGICAL PROTEOME ANALYSIS
 MULTIPLE AFFINITY PROTEIN...
 MICROARRAY ANALYSES
FROM AUTOANTIBODY SIGNATURES TO...
 REFERENCES
 
Protein microarray formats can be divided into two major classes: forward phase arrays and reverse phase arrays (44). The reverse phase format immobilizes an individual sample (i.e. antigen) in each array spot, and each array is incubated with one detection test sample (i.e. antibody or serum). Several studies have reported on the application of protein microarrays using arrayed major antigens. In most protein microarrays, a capture or bait molecule (like peptide or protein antigen) is immobilized onto a substratum. The array of bait molecule is thereafter incubated with a test sample (i.e. patient serum) containing analytes of interest (i.e. autoantibodies).

Protein (antigen) arrays have been shown to be well suited for the study of autoantibody responses (45). Miniaturized and highly parallelized microarrays reduce costs by decreasing reagent consumption and improve efficiency by greatly increasing the number of assays that can be performed with a single serum sample. Microarrays are produced either by using on-chip synthesis strategies or with an arrayer based on contact printing or ink jet technology. The first groups who described the development of antigen arrays for the specific purpose of analyzing autoantibodies were Lueking et al. (46) and Joos et al. (47). These antigen arrays used different concentrations of known autoantigens within one array to determine the titer of different autoantibodies as serological markers in autoimmune diseases. Robinson et al. (45) developed and validated a 2304-feature array containing 232 peptide and protein antigens to study multiple sclerosis. A major limitation of this kind of approach is in cases when proper post-translational modification and protein folding are necessary for antibody recognition. Considering that using recombinant proteins as arrayed antigens may not reveal true antigenicity, Hanash and co-workers (36, 48) proposed to array proteins isolated from tumors or cell lines and fractionated by liquid-based techniques. Ubiquitin C-terminal hydrolase isozyme 3 showed enhanced reactivity with sera from patients with colon cancer relative to healthy controls. Recently Qin et al. (49) proposed a "reverse capture" autoantibody microarray based on a dual antibody sandwich ELISA, which allows the antigens to be immobilized in their native configuration. Unlike the strategy using liquid-based protein separation procedures to separate intact proteins in lysates where fractions containing unknown proteins are arrayed, the reverse capture approach provides information on known antigens that are immobilized. However, the approach is limited to antigens with an affinity for previously arrayed antibodies. As "proof of principle" the authors described its use for antigen-autoantibody profiling with sera from patients with prostate cancer and benign cancer hyperplasia. To circumvent potential limitations of array systems such as alteration of immunologic determinants that can result from attachment to solid surfaces, fluid phase systems are also being developed. In such systems antigens can be labeled with beads (24) or nanoparticles (25).

A limitation with the different types of protein microarrays is that presently they provide only a relative comparison between samples for each spot and do not provide any quantitative estimation of the level of autoantibody. With the microarray approach, as with SEREX of SERPA approaches, even if some reactive antigens can be identified, no specific autoantigen (or corresponding autoantibody) has been shown to have without ambiguity a clinical usefulness. Larger scale and multisite validation studies need to be carried out to confirm the reliability of the results obtained during the discovery phase of the research.


   FROM AUTOANTIBODY SIGNATURES TO BIOMARKER VALIDATION
TOP
 ABSTRACT
 SEROLOGICAL PROTEOME ANALYSIS
 MULTIPLE AFFINITY PROTEIN...
 MICROARRAY ANALYSES
 FROM AUTOANTIBODY SIGNATURES TO...
REFERENCES
 
To make potential biomarkers valuable, several points have to be considered if we define a biomarker as a characteristic that is objectively measured and evaluated as an indicator of the pathogenic process. First it will be necessary to compare the profile of autoimmunity in patients with different pathologies and healthy volunteers, and that means well characterized specimens. To be effective, clinically useful cancer-related biomarkers should be measurable in an accessible body fluid, such as serum or plasma. Consequently antibodies are good biomarker candidates. And finally, it is necessary to use well established and robust approaches to be able to evaluate in a multicentric study the impact of technical variations on the discovery.

However, such approaches raise many practical and ethical questions. During sample collection, one may keep in mind that the patient is the focus of all the efforts with a requirement to consider the social responsibility that is given when handling human samples (26). In particular, the pathological classification of the samples and the availability of patient data (age, gender, information on treatment or surgery, and grade and cytogenetic type of the tumor) are a prerequisite. Another important question is the amount of sample (serum, plasma, etc.) available, which may significantly affect the research strategy. For instance, larger volumes are needed for chromatography-based strategies such as multiple affinity protein profiling compared with SERPA. A critical need for systematic management of these data, and later of large sets of results, is the development of user-friendly computer-based tools.

Moreover statistical analyses are desirable, not to say necessary, to estimate and to compare the predictive sensitivity and specificity of the candidate biomarkers (50). Any potential marker identified during the discovery phase has to be evaluated in an {alpha}-testing using a prototype assay test. This first clinical testing is to confirm the intended use of the assay with a verification of its specificity and its sensitivity (its robustness) before entering development phase. At this crucial moment, the main question is "how accurate is the test for identifying diseased cases?". This ability of a test to discriminate diseased cases from normal cases is evaluated using receiver operating characteristic (ROC) curve analysis (51) (Fig. 6). When one considers the data resulting from a test in two populations, one population with the disease and a control population without the disease, it is rare to observe a perfect separation between the two groups. Indeed the distribution of the test data will overlap as shown in Fig. 6A. For every possible cutoff point or criterion value selected to discriminate between the two populations, there exist some cases with the disease correctly classified as positive, denoted true positive fraction; but some cases with the disease are classified as negative, denoted false negative fraction. Then some cases without the disease are correctly classified as negative, denoted true negative fraction; but some cases without the disease are classified as positive, denoted false positive fraction. The sensitivity is defined as the probability that test assay data will be positive when the disease is present (true positive rate, expressed as a percentage), and the specificity is the probability that test assay data will be negative when the disease is absent (true negative rate, expressed as a percentage). Positive and negative likelihood ratios can be calculated as well as positive and negative predictive values. These values allow determination of the probability of a good discrimination between the two populations (i.e. diseased/normal) (Table I). In a ROC curve diagram (Fig. 6B) the true positive rate (sensitivity) is plotted as a function of the false positive rate (100 – specificity) for different cutoff points of a parameter. Each point on the ROC plot represents a couple (sensitivity/specificity) corresponding to a particular decision threshold. The area under the ROC curve is a measure of how well a parameter can distinguish between two diagnostic groups (diseased/normal).


Figure 6
View larger version (20K):
[in this window]
[in a new window]

  		FIG. 6. Evaluation of the specificity and sensitivity of a test assay for a given marker. A, the values for the two populations obtained using a test assay were plotted. In that case, the distribution shows an overlap between the data of the controls and those of the patients. A cutoff or criterion value is chosen to obtain the best discrimination between the two populations (decision threshold). B, on the ROC curve sensitivity is plotted as a function of the false positive rate (100 – specificity) for different cutoff points of a parameter. Each point represents a sensitivity/specificity pair corresponding to a particular decision threshold. The area under the ROC curve is a measure of how well a parameter can distinguish between two diagnostic groups (diseased/normal). TN, true negative fraction; TP, true positive fraction; FN, false negative fraction; FP, false positive fraction.

 

View this table:
[in this window]
[in a new window]

  		TABLE I Schematic outcomes of an assay test

Sensitivity is calculated by the ratio a/(a + b), and specificity is the ratio d/(c + d). Positive likelihood is the ratio sensitivity/(1 – specificity), and negative likelihood is the ratio (1 – sensitivity)/specificity. Positive predictive value is the probability that the disease is present when the test is positive. Negative predictive value is the probability that the disease is absent when the test is negative.

 
To conclude, it is crucial for potential biomarkers to be validated by the use of independent techniques using e.g. immunochemical measurement or protein arrays. Several distinct areas of research are necessary to identify and define relevant autoantigens, including discovery of candidate autoantigens and characterization of the sensitivity and specificity of reactivity against individuals or combinations of candidate autoantigens in cohorts of patients and controls. This type of study focus is to obtain new diagnostic tests. Unfortunately most diseases are the results of the derailment of heterogeneous and multiple regulatory processes. As a result, it is not a single biomarker that needs to be elucidated. Multiplexed biomarker protein patterns have a significantly higher positive predictive value than single markers in discriminating diseased patients from non-cancer. For instance, to identify antibodies that could be used as markers of non-small cell lung cancer, Zhong et al. (52) measured antibodies against five phage-expressed proteins by ELISA. No single antibody response measured showed overwhelming sensitivity of specificity (estimated using ROC curves). Different combinations of several markers showed either mild improvement or no improvement. They were able to increase significantly both sensitivity and specificity only by combining data from the five proteins. They concluded that multiple antibody measurements improve predictive accuracy even if further work would be necessary to improve the statistical power and to validate autoantibody measurements as a clinically reliable approach. It remains the goal for the future to test the suitability and clinical usefulness of identified marker panels in blood samples of large and independent patient cohorts.


  FOOTNOTES
 
Received, October 20, 2006, and in revised form, March 13, 2007.

Published, MCP Papers in Press, March 20, 2007, DOI 10.1074/mcp.R600016-MCP200

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

1 The abbreviations used are: SEREX, serological analysis of tumor antigens by recombinant cDNA expression cloning; 2-D, two-dimensional; 2-DE, two-dimensional gel electrophoresis; IgG, immunoglobulin G; SERPA, serological proteome analysis; ROC, receiver operating characteristic. Back

§ To whom correspondence should be addressed: Protein Biochemistry and Proteomics Laboratory, CNRS UMR 7033 (BioMoCeTi), UFR SMBH Leonard de Vinci, University Paris 13, 74 rue Marcel Cachin, 93017 Bobigny cedex, France. Tel.: 33-1-48-38-77-54; E-mail: caron_prot{at}yahoo.fr


   REFERENCES
TOP
 ABSTRACT
 SEROLOGICAL PROTEOME ANALYSIS
 MULTIPLE AFFINITY PROTEIN...
 MICROARRAY ANALYSES
 FROM AUTOANTIBODY SIGNATURES TO...
 REFERENCES
 

  1. Tan, E. M. (2001) Autoantibodies as reporters identifying aberrant cellular mechanisms in tumorigenesis. J. Clin. Investig. 108, 1411 –1415[CrossRef][Medline]

  2. Finn, O. J. (2005) Immune response as a biomarker for cancer detection and a lot more. N. Engl. J. Med. 353, 1288 –1290[Free Full Text]

  3. Canevari, S., Pupa, S. M., and Menard, S. (1996) 1975–1995 revised anti-cancer serological response: biological significance and clinical implications. Ann. Oncol. 7, 227 –232[Abstract/Free Full Text]

  4. Purcell, A. W., and Gorman, J. J. (2004) Immunoproteomics: mass spectrometry-based methods to study the targets of the immune response. Mol. Cell. Proteomics 3, 193 –208[Abstract/Free Full Text]

  5. Boon, T., and Old, L. J. (1997) Cancer tumor antigens. Curr. Opin. Immunol. 9, 681 –683[CrossRef][Medline]

  6. Naora, H., Yang, Y. Q., Montz, F. J., Seidman, J. D., Kurman, R. J., and Roden, R. B. (2001) A serologically identified tumor antigen encoded by a homeobox gene promotes growth of ovarian epithelial cells. Proc. Natl. Acad. Sci. U. S. A. 98, 4060 –4065[Abstract/Free Full Text]

  7. Scanlan, M. J., Welt, S., Gordon, C. M., Chen, Y. T., Gure, A. O., Stockert, E., Jungbluth, A. A., Ritter, G., Jager, D., Jager, E., Knuth, A., and Old, L. J. (2002) Cancer-related serological recognition of human colon cancer: identification of potential diagnostic and immunotherapeutic targets. Cancer Res. 62, 4041 –4047[Abstract/Free Full Text]

  8. Sahin, U., Tureci, O., and Pfreundschuh, M. (1997) Serological identification of human tumor antigens. Curr. Opin. Immunol. 9, 709 –716[CrossRef][Medline]

  9. Bradford, T. J., Wang, X., and Chinnaiyan, A. M. (2006) Cancer immunomics: Using autoantibody signatures in the early detection of prostate cancer. Urol. Oncol. 24, 237 –242[Medline]

  10. Soussi, T. (2000) p53 Antibodies in the sera of patients with various types of cancer: a review. Cancer Res. 60, 1777 –1788[Abstract/Free Full Text]

  11. Yamamoto, A., Shimizu, E., Sumitomo, K., Shinohara, A., Namikawa, O., Uehara, H., and Sone, S. (1997) L-Myc overexpression and detection of auto-antibodies against L-Myc in both the serum and pleural effusion from a patient with non-small cell lung cancer. Intern. Med. 36, 724 –727[Medline]

  12. Brichory, F. M., Misek, D. E., Yim, A. M., Krause, M. C., Giordano, T. J., Beer, D. G., and Hanash, S. M. (2001) An immune response manifested by the common occurrence of annexins I and II autoantibodies and high circulating levels of IL-6 in lung cancer. Proc. Natl. Acad. Sci. U. S. A. 98, 9824 –9829[Abstract/Free Full Text]

  13. Ward, R. L., Hawkins, N. J., Coomber, D., and Disis, M. L. (1999) Antibody immunity to the HER-2/neu oncogenic protein in patients with colorectal cancer. Hum. Immunol. 60, 510 –515[CrossRef][Medline]

  14. Shiku, H., Takahashi, T., Resnick, L. A., Oettgen, H. F., and Old, L. J. (1977) Cell surface antigens of human malignant melanoma. III. Recognition of autoantibodies with unusual characteristics. J. Exp. Med. 145, 784 –789[Abstract/Free Full Text]

  15. Old, L. J., and Chen, Y. T. (1998) New paths in human cancer serology. J. Exp. Med. 187, 1163 –1167[Free Full Text]

  16. Stockert, E., Jager, E., Chen, Y. T., Scanlan, M. J., Gout, I., Karbach, J., Arand, M., Knuth, A., and Old, L. J. (1998) A survey of the humoral immune response of cancer patients to a panel of human tumor antigens. J. Exp. Med. 187, 1349 –1354[Abstract/Free Full Text]

  17. Wang, X., Yu, J., Sreekumar, A., Varambally, S., Shen, R., Giacherio, D., Mehra, R., Montie, J. E., Pienta, K. J., Sanda, M. G., Kantoff, P. W., Rubin, M. A., Wei, J. T., Ghosh, D., and Chinnaiyan, A. M. (2005) Autoantibody signatures in prostate cancer. N. Engl. J. Med. 353, 1224 –1235[Abstract/Free Full Text]

  18. Canelle, L., Bousquet, J., Pionneau, C., Deneux, L., Imam-Sghiouar, N., Caron, M., and Joubert-Caron, R. (2005) An efficient proteomics-based approach for the screening of autoantibodies. J. Immunol. Methods 299, 77 –89[CrossRef][Medline]

  19. Jager, E., Jager, D., and Knuth, A. (1999) CTL-defined cancer vaccines: perspectives for active immunotherapeutic interventions in minimal residual disease. Cancer Metastasis Rev. 18, 143 –150[CrossRef][Medline]

  20. Scanlan, M. J., Chen, Y. T., Williamson, B., Gure, A. O., Stockert, E., Gordan, J. D., Tureci, O., Sahin, U., Pfreundschuh, M., and Old, L. J. (1998) Characterization of human colon cancer antigens recognized by autologous antibodies. Int. J. Cancer 76, 652 –658[CrossRef][Medline]

  21. Klade, C. S., Voss, T., Krystek, E., Ahorn, H., Zatloukal, K., Pummer, K., and Adolf, G. R. (2001) Identification of tumor antigens in renal cell carcinoma by serological proteome analysis. Proteomics 1, 890 –898[CrossRef][Medline]

  22. Le Naour, F., Misek, D. E., Krause, M. C., Deneux, L., Giordano, T. J., Scholl, S., and Hanash, S. M. (2001) Proteomics-based identification of RS/DJ-1 as a novel circulating tumor antigen in breast cancer. Clin. Cancer Res. 7, 3328 –3335[Abstract/Free Full Text]

  23. Xiang, Y., Sekine, T., Nakamura, H., Imajoh-Ohmi, S., Fukuda, H., Nishioka, K., and Kato, T. (2004) Proteomic surveillance of autoimmunity in osteoarthritis: identification of triosephosphate isomerase as an autoantigen in patients with osteoarthritis. Arthritis Rheum. 50, 1511 –1521[CrossRef][Medline]

  24. Fulton, R. J., McDade, R. L., Smith, P. L., Kienker, L. J., and Kettman, J. R., Jr. (1997) Advanced multiplexed analysis with the FlowMetrix system. Clin. Chem. 43, 1749 –1756[Abstract/Free Full Text]

  25. Nicewarner-Pena, S. R., Freeman, R. G., Reiss, B. D., He, L., Pena, D. J., Walton, I. D., Cromer, R., Keating, C. D., and Natan, M. J. (2001) Submicrometer metallic barcodes. Science 294, 137 –141[Abstract/Free Full Text]

  26. Seliger, B., and Kellner, R. (2002) Design of proteome-based studies in combination with serology for the identification of biomarkers and novel targets. Proteomics 2, 1641 –1651[CrossRef][Medline]

  27. Seliger, B., Lichtenfels, R., and Kellner, R. (2003) Detection of renal cell carcinoma-associated markers via proteome- and other ‘ome’-based analyses. Brief. Funct. Genomics Proteomics 2, 194 –212[Abstract/Free Full Text]

  28. Ishida, K., Kaneko, K., Kubota, T., Itoh, Y., Miyatake, T., Matsushita, M., and Yamada, M. (1997) Identification and characterization of an anti-glial fibrillary acidic protein antibody with a unique specificity in a demented patient with an autoimmune disorder. J. Neurol. Sci. 151, 41 –48[CrossRef][Medline]

  29. Geissler, S., Sokolowska-Kohler, W., Bollmann, R., Jungblut, P. R., and Presber, W. (1999) Toxoplasma gondii infection: analysis of serological response by 2-DE immunoblotting. FEMS Immunol. Med. Microbiol. 25, 299 –311[Medline]

  30. Petersen, A. (2003) Two-dimensional electrophoresis replica blotting: a valuable technique for the immunological and biochemical characterization of single components of complex extracts. Proteomics 3, 1206 –1214[CrossRef][Medline]

  31. Stulik, J., Hernychova, L., Porkertova, S., Pozler, O., Tuckova, L., Sanchez, D., and Bures, J. (2003) Identification of new celiac disease autoantigens using proteomic analysis. Proteomics 3, 951 –956[CrossRef][Medline]

  32. Canelle, L., Bousquet, J., Pionneau, C., Hardouin, J., Choquet-Kastylevsky, G., Joubert-Caron, R., and Caron, M. (2006) A proteomic approach to investigate potential biomarkers directed against membrane-associated breast cancer proteins. Electrophoresis 27, 1609 –1616[CrossRef][Medline]

  33. Pionneau, C., Canelle, L., Bousquet, J., Hardouin, J., Bigeard, J., Caron, M., and Joubert-Caron, R. (2005) Proteomic analysis of membrane-associated proteins from the breast cancer cell line MCF7. Cancer Genomics Proteomics 2, 199 –208

  34. Jemal, A., Tiwari, R. C., Murray, T., Ghafoor, A., Samuels, A., Ward, E., Feuer, E. J., and Thun, M. J. (2004) Cancer statistics, 2004. CA Cancer J. Clin. 54, 8 –29[Abstract/Free Full Text]

  35. Brichory, F., Beer, D., Le Naour, F., Giordano, T., and Hanash, S. (2001) Proteomics-based identification of protein gene product 9.5 as a tumor antigen that induces a humoral immune response in lung cancer. Cancer Res. 61, 7908 –7912[Abstract/Free Full Text]

  36. Nam, M. J., Madoz-Gurpide, J., Wang, H., Lescure, P., Schmalbach, C. E., Zhao, R., Misek, D. E., Kuick, R., Brenner, D. E., and Hanash, S. M. (2003) Molecular profiling of the immune response in colon cancer using protein microarrays: occurrence of autoantibodies to ubiquitin C-terminal hydrolase L3. Proteomics 3, 2108 –2115[CrossRef][Medline]

  37. Nam, M. J., Kee, M. K., Kuick, R., and Hanash, S. M. (2005) Identification of defensin {alpha}6 as a potential biomarker in colon adenocarcinoma. J. Biol. Chem. 280, 8260 –8265[Abstract/Free Full Text]

  38. Livingston, P. O., Ragupathi, G., and Musselli, C. (2000) Autoimmune and antitumor consequences of antibodies against antigens shared by normal and malignant tissues. J. Clin. Immunol. 20, 85 –93[CrossRef][Medline]

  39. Tontsch, D., Pankuweit, S., and Maisch, B. (2000) Autoantibodies in the sera of patients with rheumatic heart disease: characterization of myocardial antigens by two-dimensional immunoblotting and N-terminal sequence analysis. Clin. Exp. Immunol. 121, 270 –274[CrossRef][Medline]

  40. Caron, M., Joubert-Caron, R., Canelle, L., and Hardouin, J. (2005) Serological proteome analysis (SERPA) and multiple affinity protein profiling (MAPPING) to discover cancer biomarkers. Mol. Cell. Proteomics 4, (suppl.)S142

  41. Hardouin, J., Duchateau, M., Canelle, L., Vlieghe, C., Joubert-Caron, R., and Caron, M. (2007) Thiophilic adsorption revisited. J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 845, 226 –231[CrossRef][Medline]

  42. Yin, H., Killeen, K., Brennen, R., Sobek, D., Werlich, M., and van de Goor, T. (2005) Microfluidic chip for peptide analysis with an integrated HPLC column, sample enrichment column, and nanoelectrospray tip. Anal. Chem. 77, 527 –533[Medline]

  43. Hardouin, J., Duchateau, M., Joubert-Caron, R., and Caron, M. (2006) Usefulness of an integrated microfluidic device (HPLC-Chip-MS) to enhance confidence in protein identification by proteomics. Rapid Commun. Mass Spectrom. 20, 3236 –3244[CrossRef][Medline]

  44. Sheehan, K. M., Calvert, V. S., Kay, E. W., Lu, Y., Fishman, D., Espina, V., Aquino, J., Speer, R., Araujo, R., Mills, G. B., Liotta, L. A., Petricoin, E. F., III, and Wulfkuhle, J. D. (2005) Use of reverse phase protein microarrays and reference standard development for molecular network analysis of metastatic ovarian carcinoma. Mol. Cell. Proteomics 4, 346 –355[Abstract/Free Full Text]

  45. Robinson, W. H., Steinman, L., and Utz, P. J. (2003) Protein arrays for autoantibody profiling and fine-specificity mapping. Proteomics 3, 2077 –2084[CrossRef][Medline]

  46. Lueking, A., Horn, M., Eickhoff, H., Bussow, K., Lehrach, H., and Walter, G. (1999) Protein microarrays for gene expression and antibody screening. Anal. Biochem. 270, 103 –111[CrossRef][Medline]

  47. Joos, T. O., Schrenk, M., Hopfl, P., Kroger, K., Chowdhury, U., Stoll, D., Schorner, D., Durr, M., Herick, K., Rupp, S., Sohn, K., and Hammerle, H. (2000) A microarray enzyme-linked immunosorbent assay for autoimmune diagnostics. Electrophoresis 21, 2641 –2650[CrossRef][Medline]

  48. Qiu, J., Madoz-Gurpide, J., Misek, D. E., Kuick, R., Brenner, D. E., Michailidis, G., Haab, B. B., Omenn, G. S., and Hanash, S. (2004) Development of natural protein microarrays for diagnosing cancer based on an antibody response to tumor antigens. J. Proteome Res. 3, 261 –267[CrossRef][Medline]

  49. Qin, S., Qiu, W., Ehrlich, J. R., Ferdinand, A. S., Richie, J. P., O'Leary M, P., Lee, M. L., and Liu, B. C. (2006) Development of a "reverse capture" autoantibody microarray for studies of antigen-autoantibody profiling. Proteomics 6, 3199 –3209[CrossRef][Medline]

  50. Wilkins, M. R., Appel, R. D., Van Eyk, J. E., Chung, M. C., Gorg, A., Hecker, M., Huber, L. A., Langen, H., Link, A. J., Paik, Y. K., Patterson, S. D., Pennington, S. R., Rabilloud, T., Simpson, R. J., Weiss, W., and Dunn, M. J. (2006) Guidelines for the next 10 years of proteomics. Proteomics 6, 4 –8[CrossRef][Medline]

  51. Zweig, M. H., and Campbell, G. (1993) Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clin. Chem. 39, 561 –577[Abstract/Free Full Text]

  52. Zhong, L., Peng, X., Hidalgo, G. E., Doherty, D. E., Stromberg, A. J., and Hirschowitz, E. A. (2004) Identification of circulating antibodies to tumor-associated proteins for combined use as markers of non-small cell lung cancer. Proteomics 4, 1216 –1225[CrossRef][Medline]


Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?



This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
R600016-MCP200v1
6/7/1115    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Glossary
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Caron, M.
Right arrow Articles by Joubert-Caron, R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Caron, M.
Right arrow Articles by Joubert-Caron, R.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Journal of Biological Chemistry 
 Journal of Lipid Research   ASBMB Today