Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Immunol Rev. Author manuscript; available in PMC 2009 Aug 2.
Published in final edited form as:
PMCID: PMC2719766

Autoantibodies to tumor-associated antigens: reporters from the immune system


Although autoantibodies have been recognized as participants in pathogenesis of tissue injury, the collateral role of autoantibodies as reporters from the immune system identifying cellular participants in tumorigenesis has not been fully appreciated. The immune system appears to be capable of sensing aberrant structure, distribution, and function of certain cellular components involved in tumorigenesis and making autoantibody responses to the tumor-associated antigens (TAAs). Autoantibodies to TAAs can report malignant transformation before standard clinical studies and may be useful as early detection biomarkers. The autoantibody response also provides insights into factors related to how cellular components may be rendered immunogenic. As diagnostic biomarkers, specific TAA miniarrays for identifying autoantibody profiles could have sufficient sensitivity in differentiating between types of tumors. Such anti-TAA profiles could also be used to monitor response to therapy. The immune system of cancer patients reveals the immune interactive sites or the autoepitopes of participants in tumorigenesis, and this information should be used in the design of immunotherapy.

Keywords: tumor-associated antigens, TAA miniarrays, anti-TAA profiles, therapeutic response monitors, immune system reporters, auto-epitope immunotherapy


Autoantibodies are hallmark features of patients with autoimmune diseases such as systemic lupus erythematosus (lupus), scleroderma, Sjögren’s syndrome, dermato/polymyositis, and other disorders. Research for several decades in these disorders has been devoted to identifying the autoantibodies, characterizing their cognate antigens, and elucidating pathogenic mechanisms. These studies have shown that autoantibodies are biomarkers helpful in clinical diagnosis and are biological agents useful for isolating and studying the function of intracellular molecules which happen to be target autoantigens (reviewed in 1, 2). Further work in this area has shown that the occurrence of autoantibodies is not uncommon, and autoantibodies have been identified in type 1 diabetes (3), autoimmune thyroid disease (4), bullous skin diseases, liver and other gastrointestinal diseases, neurological diseases, and many other disorders (reviewed in 5). Autoantibodies have also been described in cancer (69), and a key observation was the report that some patients with breast cancer had antibodies to p53, an important tumor suppressor protein (10).

Although studies of autoantibodies in the systemic autoimmune diseases such as lupus and in organ-specific autoimmunity such as autoimmune thyroiditis and type 1 diabetes pre-date studies of autoantibodies in cancer by several years, the observations in the cancer and non-cancer areas are converging and are providing insights into questions which each area by itself was not able to answer (11). In this review, we discuss how antibodies to tumor-associated antigens (TAAs) in cancer are very much like autoantibodies to nuclear and cytoplasmic antigens in lupus and other systemic autoimmune diseases, in that anti-TAAs also have the potential of being highly useful diagnostic markers in cancer. Secondly, identification of TAAs in cancer patients has led to studies into mechanisms by which molecular and other alterations of intracellular proteins could have driven autoimmune responses. Finally, a paradigm that has been observed in lupus and other autoimmune diseases is that each disorder has a profile of autoantibodies characteristic of that disorder, although there are other autoantibodies which are shared (1). It appears this situation also applies to cancer, and there are profiles of autoantibodies unique for a certain type of cancer and other antibodies that are shared among many cancer types. These autoantibody profiles may be useful diagnostic markers for cancer differentiation (11).

Autoantibodies as reporters of early or incipient carcinogenesis

Many investigators are focused on the role that autoantibodies might play in the pathogenic or protective aspects of malignancy and overlook the collateral benefits to be derived by recognizing them as reporters from the immune system identifying cellular participants in carcinogenesis. Our appreciation of this surrogate role of autoantibodies came from studies of autoantigens in lupus and other rheumatic autoimmune diseases. Autoantigens such as Sm antigen(s) in lupus are intrinsic components of small nuclear ribonucleoprotein particles (also known as snRNPs), and Scl-70, an autoantigen in scleroderma, is a natural degradation product of the 100 kDa DNA topoisomerase I. A number of autoantigens in dermato/polymyositis are aminoacyl-tRNA synthetases associated with protein biosynthesis in ribosomes. These and other autoantigens (reviewed in 1) have important functions in all cell types and are not the so-called ‘housekeeping’ components of the cell. Similarly, many of the TAAs identified by cancer patient autoantibodies have important cellular biosynthetic functions that might be related to carcinogenesis.

Hepatocellular carcinoma

Hepatocellular carcinoma (HCC) is a cancer that has provided unique opportunities to examine autoantibody responses to TAAs (12). It has been well-documented that chronic hepatitis (hepatitis B virus- or hepatitis C virus-related) and liver cirrhosis are precursor conditions that are predisposed to develop into HCC (13, 14). About one-third of patients with chronic hepatitis or liver cirrhosis ultimately develop HCC, although the incubation period may take up to two decades or more. Such patients have frequent medical examinations that include diagnostic studies with the aim of detecting early signs of liver cancer, so that treatment might be instituted at an early stage of malignancy. In some patients, opportunities were available to us to examine serial serum samples taken many years before development of malignancy, a situation which is rarely available in other types of cancer.

Fig. 1 shows a patient in whom HCV-related chronic hepatitis was diagnosed in 1978, and liver cancer was clinically detected 10 years later (15). Serum samples were collected annually during this period and at more frequent intervals after cancer detection. The serum samples were negative for antibodies to nuclear antigens (ANAs) from 1978 to mid-1986, when there was sudden seroconversion and ANAs were detected in high titers (1/320–1/1280 dilution of serum). Another marker of liver cancer, α-fetoprotein (AFP), also appeared in high titer (normal < 20 ng/ml). The patient received chemotherapy with transhepatic arterial embolization but eventually succumbed to the malignancy. The target antigens of the novel antibodies associated with malignancy were doublet proteins of 170 and 180 kDa, which were shown to be DNA topoisomerase II-α and -β, nuclear proteins involved in DNA replication and preferentially expressed in G2 and M phases of the cell cycle (1517).

Fig. 1
Appearance of novel autoantibody associated with carcinogenesis

Another study illustrates other features in such patients. Fig. 2 shows a patient who already had circulating ANAs in low titer, 1/40 dilution of serum, but there was a substantial increase in ANA titer with appearance of malignancy that was not associated with the appearance of AFP (12). These features of either a novel autoantibody response before or coincident with detection of cancer (as in Fig. 1) or increase in titer of pre-existing antibody (as in Fig. 2), with or independent of AFP, have been observed in several patients and are not unusual phenomena (12). Fig. 2 also shows that before detection of liver cancer, pre-existing antibodies were detected by immunofluorescent antibody and Western blot assays. This observation cautions against the presumption that all antigens identified with autoantibodies from cancer patients are tumor associated because of pre-existing antibodies which were not tumor related. Many examples of such situations have been reported (12, 15).

Fig. 2
Rise in ANA titer with malignancy

In 37 patients, serial serum samples were available at the initial diagnosis of chronic liver disease and when liver cancer was detected (12). The time span was 6–107 months (mean, 41 months). Serawere examined for the presence or absence of ANAs by indirect immunofluorescence using HEp 2 cells as substrate. ANA positivity included antibodies reactive with nucleoplasm, nuclear membrane, and nucleolus. Ten of the 37 patients (27%) were ANA positive in the pre-cancer stage. Of the 27 ANA-negative patients, eight (30%) converted to ANA positivity when they had cancer. Of the 10 who had pre-existing ANAs, four of the 10 (40%) showed increase in ANA titer with cancer. Thus, in this group of 37 patients who acquired liver cancer, 12 (32%) showed either a novel autoimmune reaction or increase in titer of existing antibodies. These patients’ immune systems appeared to be reacting to factors involved in carcinogenesis, because in cases where the antigens have been identified, they were cellular molecules involved in cell functions directly or indirectly related to cell proliferation or gene regulation. The 32% increase in frequency of the immune system response associated with malignant transformation should be taken as a conservative or minimal finding, since only a limited dimension of the immune response (novel appearance or increased titer of ANA) was examined.

p53 and other TAAs

In addition to the chronic liver disease to HCC model, there are several examples where autoantibodies to p53 have been found to be reporters of early or incipient carcinogenesis. A heavy smoker developed lung cancer a few years after detection of anti-p53 autoantibodies, and this patient enjoyed a long period of remission or cure because of early detection of lung cancer (18, 19). Anti-p53 has been detected in chronic obstructive pulmonary disease, which is prone to development of lung cancer (20), and also in workers occupationally exposed to vinyl chloride who have a similar predisposition (21). In uranium workers who are also at high risk for lung cancer, anti-p53 was detected 17–47 months before clinical tumor manifestation (22). There is a group of diseases called paraneoplastic neurological disorders (PNDs) in which neurological symptoms are present in patients with malignancies of the lung, breast, ovary, and testis, and the PNDs have been shown to be related to ectopic expression of neuron-specific proteins in the malignant cells (2325). These patients make autoantibodies to the ectopically expressed neuronal proteins. In such patients, a high index of suspicion led to early detection of the relevant tumors.

In lupus, medical treatment has not changed significantly since the 1960s and still consists largely of the use of corticosteroids and immunosuppressive drugs. Survival times of patients, however, have increased dramatically, and there is general agreement that this improvement is related to earlier diagnosis with the use autoantibody markers. The use of autoantibodies to TAAs for earlier diagnosis of cancer is discussed below.

Factors related to immunogenicity of TAAs

A question of great interest in autoimmunity is how intracellular molecules such as TAAs acquire immunogenicity. Some of the special features of TAAs are that they have functions known to be involved in proliferation, transformation, and other processes associated with malignancy. Although these molecules are present in most cell types, any deviation from the normal state might be expected to be detectable in malignant cells but not in normal cells. A study of p53 in lung cancer showed that missense point mutations leading to altered protein products were highly correlated with production of autoantibodies whereas stop, splice, and frameshift mutations did not (26). The point mutations occurred in the hot spot regions of p53, in exons 4–9, giving rise to abnormal proteins with altered functions and increased stability compared with wildtype protein, which is normally rapidly degraded. It might have been expected that altered p53 proteins from these missense mutations had acquired neo-antigenic determinants that stimulated antibody production and that the autoimmune response might be targeted against novel epitopes. However, the anti-p53 autoantibodies reacted equally well with wildtype and mutated p53, and the reactive antigenic determinants were located at the amino-and carboxy-terminal regions, away from the central regions where the point mutations had occurred (27, 28). These findings are highly informative and suggest that immunogenicity of TAAs is more than a function of gene mutations.

In attempting to identify TAAs with antibodies from patients with chronic hepatitis who developed novel immune responses associated with malignant conversion, we isolated a cytoplasmic protein containing two types of RNA-binding motifs, a consensus-sequence RNA-binding motif and the hnRNP-K homology motif (29, 30). This protein of 62 kDa is a splice variant of a family of insulin-like growth factor II mRNA binding proteins (IMPs) (31). The interest in the IMPs is that insulin-like growth factor II itself has been shown to be involved in carcinogenesis (32, 33). One of the mRNA binding proteins, IMP-1 also called CRD-BP (c-myc coding region determinant binding protein), has been shown to bind c-myc mRNA and protect an unstable coding region site from degradation (34). High expression of IMP-1 has been found in mesenchymal tumors and Ewing’s sarcoma (35) and in breast cancer (36). Another family member, IMP-3 also called Koc, was found to be overexpressed in human pancreatic and other cancers (37).

After finding that antibodies to p62/IMP-2 splice variant were present in patients with HCC, we addressed the question how p62 might be involved in liver carcinogenesis by examining liver tissue of patients with HCC, liver cirrhosis, biopsies of normal liver, and specimens of fetal liver (38). In reverse transcriptase polymerase chain reaction studies, mRNA encoding p62 was shown to be present in normal fetal liver but not in normal adult liver, and protein expression was confirmed by immunohistochemistry to be present in fetal liver but not in adult liver. Of 27 HCC specimens, nine showed uniform staining of cytoplasm for p62 protein in malignant hepatocytes but no staining in normal hepatocytes adjacent to HCC nodules or in other areas of normal liver (Fig. 3). Twenty-three specimens of cirrhotic livers and nine normal livers showed no staining for p62, except for focal p62 staining in a few cirrhosis specimens. Because liver cirrhosis is a common precursor of HCC, the latter finding may be of some significance and may be a prelude to carcinogenesis. The observation that p62 is developmentally regulated, expressed in fetal liver but not in adult liver, and aberrantly expressed in malignant liver cells suggests that this TAA has aspects of an oncofetal antigen. Cyclin B1, which has a key role in progression of the cell cycle from G2 to M, has been identified as a TAA in many tumor patients, and both B- and T-cell immune responses are present (39, 40). Cyclin B1 was found to be overexpressed in tumor cells and dislocated from nuclei to cytoplasm (41). It is not clear whether features such as the aberrant expression of a fetal protein or overexpression and altered localization of cellular proteins are factors sufficient to induce autoimmune responses or whether additional factors are required. However, observations such as these are accumulating and point to future directions of investigation.

Fig. 3
Expression of p62/IMP-II in HCC

The autoantibody-defined epitope is a highly conserved structure, conformation dependent, and a functioning site

For detection of ANAs in patients with systemic autoimmune diseases, the substrate used in the immunohistochemical assay does not have to be of human origin but could as well be tissue sections or tissue culture cells from non-human species, such as from mouse, rat, rabbit, bovine, or monkey origins (1). This feature alludes to the highly conserved nature of the antigenic determinants, but the more profound significance of this characteristic is often not fully appreciated. In this section, we review studies showing that the autoantibody-defined epitope, in addition to being highly conserved, is frequently a conformation-dependent structure and at or near the binding or functioning sites.

Patients with Sjögren’s syndrome produce autoantibodies to a nuclear protein called SS-B/La. This protein is known to be associated with RNA Pol III transcripts, which are precursors of tRNA and 5S RNA, and is involved in the processing of these precursors to mature RNAs (42). The SS-B/La autoantibody is present in up to 60% of patients with Sjögren’s syndrome and 20–30% of patients with lupus (1). The sites in SS-B/La that are reactive with autoantibodies have been identified (43, 44), but the unique feature of the epitope was revealed only after studies comparing spontaneously occurring autoantibodies and antibodies produced by experimental immunization (45). SS-B/La was affinity purified from bovine thymus nuclear extracts using antibodies from a patient with Sjögren’s syndrome and was used to immunize mice. Five clones producing monoclonal antibodies were generated from the spleen of the immunized mouse and analyzed for their immunoreactivity with a standard human SS-B/La autoantibody (Ze) available from the Centers for Disease Control, Atlanta (45). All five monoclonal antibodies were highly reactive with bovine SS-B/La in immunostaining of bovine MDBK tissue culture cells. However, there were clear differences in epitope or antigenic determinant recognition between monoclonal antibodies and human autoantibody with respect to proteolytic fragments of the antigen (45). The most significant finding was in immunostaining using tissue culture cell lines from different species (Table 1).

Table 1
Species-specific reactivities of monoclonal antibodies to SS-B/La antigen as detected by immunofluorescence

The murine monoclonal antibodies were reactive only with SS-B/La present in the nuclei of cells from hamster, rat, mouse, and rat kangaroo. In contrast, the human autoantibody (Ze) was reactive with all species tested. This study demonstrated that the autoantibody-defined determinant was a region of the molecule that was highly conserved, whereas experimentally induced antibodies recognized random or less-conserved determinants.

Proliferating cell nuclear antigen (PCNA) (46) is an auxiliary protein of DNA polymerase-δ, a multi-unit particle that is involved in DNA replication (47, 48). Autoantibody to PCNA occurs in a small proportion of lupus patients, and its usefulness as a diagnostic marker is limited. However, studies related to properties of the epitope(s) reactive with autoantibody have provided interesting insights. The immune epitopes on PCNA were initially examined using human autoantibody, murine monoclonal antibodies to purified PCNA, and rabbit polyclonal antibody to an amino-terminal synthetic peptide (49). By using proteolytic peptide fragments of PCNA as antigens, the major regions containing antigenic determinants could be mapped, and this information was used to design shorter fragments for more refined epitope mapping. A cDNA clone encoding full length human PCNA was used to generate in vitro translated products with different carboxy- and amino-terminal deletions (50). Fourteen human anti-PCNA sera were examined for their patterns of reactivity. None of the 14 lupus sera reacted with 15-mer continuous sequence synthetic peptides covering full-length PCNA. In contrast, all 14 lupus sera reacted with recombinant proteins of the engineered cDNA constructs, either in Western blotting or in immunoprecipitation assays. There were variable patterns of immunoreactivity, and this heterogeneity was interpreted as possibly consistent with immune responses to conformation-dependent epitopes related to protein folding or to complexes formed with other molecules in its DNA replication function. To test this hypothesis, the following studies were performed.

When non-synchronized tissue culture cells are used as substrate in the immunolocalization of PCNA with human autoantibodies from lupus patients, different patterns of nuclear staining in the same field are observed (Fig. 4). This was determined to correspond to cells in different phases of DNA synthesis (51). Cells that were in G1 phase of the cell cycle showed little or barely detectable staining. In early S phase, there was strong staining of nucleoli with weak speckled staining of nucleoplasm. In mid and late S phase, PCNA staining manifested as generalized small or large speckles in the nucleoplasm with absence of nucleolar staining. In G2 phase, immunostaining was again absent or minimal as in G1. PCNA staining correlated very closely with radiolabeled thymidine uptake by the cells (51).

Fig. 4
PCNA immunolocalization in non-synchronized HEp-2 tissue culture cells

Biochemically, it was shown that in the cycling cell, there was an excess of PCNA over that involved in DNA replication and that the total amount of PCNA protein did not fluctuate with different phases of the cell cycle (52). Because the patterns of expression shown by immunostaining corresponded with thymidine uptake, the implication was that human anti-PCNA antibody was recognizing that fraction of PCNA which was engaged in DNA synthesis but not with the fraction not so engaged. This additional evidence led to a study to determine whether a conformation-dependent peptide of PCNA might be able to elicit an antibody with the immunostaining properties of human autoantibody (53). Based on earlier studies (49, 50), four regions of PCNA that had been found to be reactive with human serawere selected to construct compound peptides. These compound peptides were synthesized to contain two discontinuous linear sequences from these four regions and were 14–15 amino acid residues in length.

A total of six such compound peptides are presented in Fig. 5, showing the composition and the orientation of the amino acid residues. The objective was to determine whether any of these compound peptides, which theoretically might represent native conformation-dependent epitope(s) recognized by human autoantibody, would elicit an antibody response with properties in immunostaining like that shown in Fig. 4.

Fig. 5
Construction of compound peptides of PCNA

All six compound peptides in immunized rabbits elicited antibody responses that reacted with nuclei of HEp-2 cells (53). This result was not unexpected, because all the peptides were from regions of PCNA previously identified to contain antigenic determinants. With the exception of anti-15–25, the other antibodies stained nuclei weakly and in patterns different from human autoantibody. Only antibodies to compound peptide 15–25 (residues 159–165 + 255–261) produced an antibody response similar to human anti-PCNA (Fig. 6). This response included a variegated pattern of nuclear staining in non-synchronized HEp-2 cells: staining of nucleoli in some cells (Fig. 6A, arrow) and speckled nucleoplasmic staining in other cells (Fig. 6A, arrowhead). This same field was double-stained with human lupus autoantibody and rhodamine-conjugated second antibody (Fig. 6B), and the identical pattern of nuclear staining or lack of staining was observed. In Fig. 6C, another cell spread after in vitro bromodeoxyuridine (BrdU) labeling was reactive with anti-15–25 and rhodamine-conjugated second antibody. The same specimen was reactive with anti-BrdU followed by fluorescein-conjugated second antibody (Fig. 6D). Fig. 6C and D shows that identical intranuclear structures reacted with anti-15–25 and anti-BrdU, demonstrating that anti-15–25 was reacting with nuclear components capable of DNA synthesis. The compound peptide 15–52, which has the same amino acid composition as 15–25 but with the last seven residues in reverse orientation, did not elicit a similar antibody response.

Fig. 6
Immunostaining of HEp-2 cells by 15–25 antiserum

Extensive data from studies of antigen–antibody reacting systems demonstrate that many B-cell epitopes are discontinuous sequences and highly conformation-dependent (54). Antibodies against foot-and-mouth disease of cattle have been mapped to conformation-dependent sites (55). In human choriogonadotropin, a region of the α subunit (residues 41–60) and a region of the β subunit (residues 101–121) were selected for construction of a series of compound peptides (56). Polyclonal antibodies produced against a compound peptide (α 46–55, β 106–116) were specific for human choriogonadotropin and competitively inhibited the binding of human gonadotropin to this receptor. Autoreactive epitopes defined by antibodies in type I diabetes were mapped to the mid- and carboxy-terminal domains of GAD 65 (57), which were conformation-dependent chimeric peptides (58).

The observations that autoantibody-reactive epitopes are highly conserved regions and conformation-dependent structures raise questions concerning the biological significance or implications of these findings. It is known that these properties are characteristic of functioning sites of proteins. If the autoepitope of the antigen was indeed the functioning site or in proximity to it, autoantibodies would be likely to inhibit function.

There are many studies showing that human autoantibodies are capable of inhibiting the function of their target antigens. These findings have been demonstrated in vitro and should not be considered evidence that a similar in vivo activity was taking place. Antibodies are in the extracellular milieu, and autoantigens under discussion are intracellular. The possibilities of in vivo antigen-antibody reactions would depend on a complex of events including cell-mediated interactions, cytokine release, participation of inflammatory mediators, and others. Although we do not fully understand how or to what extent autoantibodies are involved in pathogenesis, many have antigen neutralizing properties in vitro. Autoantibodies to snRNP particles in lupus inhibit the splicing of precursor mRNAs (59), and autoantibodies in scleroderma inhibit function of DNA topoisomerase I (6062) and RNA polymerase I (63). In fact, human anti-topoisomerase I antibody inhibited activity of plant DNA topoisomerase I (62), providing further evidence for conservation of the autoantibody-defined epitope. Further evidence includes human autoantibodies that inhibit function of tRNA synthetases (64, 65) and anti-centromere antibodies that disrupt chromosome movement in mitosis (66).

It was possible to further examine the different properties between autoantibodies and antibodies elicited by immunization. We compared the ability of various anti-PCNA antibodies to inhibit complementary strand DNA nucleotide synthesis using a single-strand template with added DNA polymerase δ (67). Lupus autoantibody efficiently inhibited DNA synthesis, whereas two murine monoclonal antibodies raised against purified PCNA and a rabbit polyclonal antibody to PCNA-N terminal peptide were non-inhibitory. In follow-up studies described above (49, 50), it was shown that the autoantibody-reactive epitopes were different from those recognized by the two monoclonal antibodies and the anti-peptide antibody. The uniqueness of the auto-epitope was also found with autoantibody to threonyl-tRNA synthetase from apatient with polymyositis when compared with rat polyclonal antibody to purified synthetase (68). The autoantibody reacted with native but not denatured synthetases, whereas the rat antibody recognized both forms of the enzyme. The autoantibody recognized an epitope associated with the conserved catalytic site of threonyl-tRNA synthetase and inhibited its function, but the rat antibody not.

In autoimmunity, what the immune system sees in vivo and how it reacts to a self-antigen may be quite different from how an experimental animal sees and reacts to the same antigen purified in vitro and used for elicitation of an immune response. One obvious, but not the only, reason for the difference in reaction is that the antigen in vivo is not likely to be an isolated or purified molecule but is more likely to be a component of multi-unit subcellular particle. How a component in this particle is processed or metabolized before being confronted with an autologous immune response is a question of great interest, and much work has been devoted to whether cellular apoptosis or necrosis might play a role. Studies include examination of cleavage of known nuclear autoantigens during apoptosis and insufficient clearance of dying cells (6971). A convincing proof would be the demonstration that potential immunogens isolated from apoptotic or necrotic cells are capable of eliciting antibodies recognizing the same epitopes as spontaneously occurring autoantibodies.

Anti-TAA profiles as diagnostic markers and therapeutic response monitors

As discussed above, there are many similarities between autoantibody responses in cancer and systemic autoimmune disorders (11). Lupus and other autoimmune illnesses are characterized by the presence of multiple antibodies in an individual patient, and the same feature appears to be present in cancer. In systemic autoimmunity, it was observed that there were autoantibody profiles that appeared to be immune signatures of certain disorders, and this signature consisted of some antibodies which were unique and others which were shared (1, 72). Because these autoantibody profiles have been useful as diagnostic markers, it was of interest to determine whether anti-TAA profiles might also be present in cancer.

Koc, a member of a family of IGF-II mRNA-binding proteins, was reported to be overexpressed in pancreatic and other cancers, and it was later shown that autoantibodies were present in 12.2% of 777 cancer patients (73). In this group of patients, 11.6% also had antibodies to p62, but because Koc and p62 had 80% similarity in protein sequence, it was important to determine that the results were not due to the same antibodies binding to cross-reactive epitopes. Recombinant polypeptides were made from cDNA constructs, and the antigenic epitopes were identified by Western blotting, enzyme-linked immunoassay, and immunoprecipitation. The epitopes were shown to be present at the amino-terminal of both proteins, but absorption studies showed that the autoantibodies were not cross-reactive.

Table 2 shows the frequency of autoantibodies to p62 and Koc in 10 different cancers and normal controls. There was limited range in autoantibody frequency among the different cancers, and the highest number for antibodies to p62 was 17.9% in lung cancer and to Koc was 17.3% in liver cancer. In each of the 10 different cancers, there were some patients with antibodies to both proteins and some to one but not the other. Another observation of interest was that when the frequency of antibodies to the panel as a whole was determined, there was an increase in frequency in every case. This finding raised the question whether a larger panel of TAAs might further increase antibody detection, making this approach feasible for use in immunodiagnosis.

Table 2
Frequency of autoantibodies to mRNA-binding proteins p62 and Koc in human malignancies

An enlarged panel of TAAs was selected on the basis of studies from many investigators. This panel consisted of a restricted number of six TAAs, which included three IGF-II mRNA binding proteins and p53, c-myc, and cyclin B1 (74) (Table 3). In 546 sera from 11 different types of cancer (breast, lung, colorectal, gastric, prostate, esophageal, pharyngeal, ovarian, thyroid, uterine, and HCC), there was stepwise increase of positive antibody reactions with addition of successive TAAs. The highest percentage of positive reactions was in lung cancer (64.3%) and the lowest in ovarian cancer (22.7%). Although this assay was undertaken with a limited panel of TAAs, certain anti-TAA profiles could be distinguished. Anti-cyclin B1 and anti-p62 were the highest scores in lung cancer, anti-c-myc and anti-Koc in breast cancer. In prostate cancer (75), there was preferential antibody response to p62 and a newly identified TAA called p90 with antibody frequencies of 22.6% and 30.8%, respectively. In this study where a panel of 10 TAAs was used, the likelihood of finding antibody in prostate cancer was 92.5%.

Table 3
Frequency of autoantibodies to six cancer-related antigens in human malignancies

Recursive partitioning, a multivariate statistical approach, was used to ascertain whether cancer patients and normal subjects could be classified separately on the basis of anti-TAA profiles (76). Recursive partitioning determines for each variable (in this case each of seven TAAs) a cut-off point that optimally splits all the individuals into cancer and controls using the variable that performs best. It then takes the remaining subpopulation and repeats the process until no additional partitioning is warranted: meaning that the final subpopulation contains one class of individuals or is too small to subdivide further. Cyclin B1 was the first discriminator for lung cancer, gastric cancers, and HCC, and the subsequent discriminator for all other cancer cohorts. C-myc was the initial discriminator in breast cancer, p62 in prostate cancer, and IMP-1 in colon cancer. Recursive partitioning demonstrated that no more than three of the seven TAAs were needed for any cancer cohort to arrive at sensitivities ranging from 0.77 to 0.92 and specificities ranging from 0.85 to 0.91. These sensitivity/specificity values pertain to discrimination between cancer cohorts and normal subjects, but the additional feature appears to be the ability to differentiate between cancer cohorts on the basis of TAAs found to be optimal discriminators.

There is increasing awareness of the importance of TAAs and autoantibodies to TAAs in many aspects of cancer, including diagnosis, surveillance, and therapy, and some of the focus has been on the identification of the potentially large number of TAAs. The initial approach consisted of using cDNA expression libraries from immortalized cell lines and screening with antibodies from cancer sera (77). Subsequently, a modification called SEREX (serological analysis of recombinant cDNA expression libraries) (78) was introduced in which the cDNA expression library was constructed from tumor tissue specimens. However, it was found that such expression libraries contained cDNA clones not only from tumor cells but also from inflammatory and other cellular infiltrates, and the method has reverted to the use of cDNA libraries from cell lines (79). Tumor cell lines have also recently been used as source of cDNA to construct phage display libraries to identify candidate TAAs in colorectal (80), breast (81), prostate (82), and ovarian cancer (83).

A proteomics type approach has been recently used (8486), where lysates from tumors or cancer cell lines are separated on two-dimensional gel electrophoresis, transferred to membranes, and probed with cancer sera and immunoreactive proteins identified by mass spectrometry. Much of this protein microarray technology has been successfully used for identifying autoantibodies in systemic autoimmunity, but in this instance, the investigators had previous knowledge of the identity of autoantigens (87, 88). In cancer, the major task ahead is the continuing identification of TAAs, and the challenging problem is the separation tumor-associated from non-TAAs, because autoantibodies other cellular antigens can be present before appearance of new antibodies occurring with malignancy (12, 15, 29, 89). In our experience, it has been necessary to validate a candidate TAA by testing not only with cancer sera but also with pre-cancer sera when available and with sera from other autoimmune disorders.

Discussion and conclusion

There is persuasive evidence from the studies cited above that the immune system of the cancer patient has the capability sensing abnormalities in structure, function, intracellular location, and other alterations of cellular participants in tumorigenesis. This can be manifested either by humoral or cell-mediated immune responses and can often be the earliest signal of carcinogenesis. The autoimmune response has been variously described as reporters from the immune system (90), sentinels within (91), and immunosurveillance (92). This unexpected help from the immune system has provided the cancer researcher with powerful tools to connect the transformation event in a particular cancer patient with a specific cellular participant. There are many questions that need to be elucidated, including why the frequency of the humoral immune response to p53, a key player in carcinogenesis, and to many other TAAs appears to hold in the range of 10–30% for all types of cancer. It is likely that further work will discover an autoantibody that is present in higher frequency and is more specific for a certain tumor type. Most of the current emphasis is being placed on cancer autoantibodies as diagnostic biomarkers, but it is likely that autoantibodies can also be used as monitors of therapeutic response. In a patient where a certain anti-TAA antibody has been detected, change in antibody levels might reflect change in tumor status or tumor burden related to therapy. Precedence for this notion comes from reports in lupus, where clinical observations indicate that a fall in levels of autoantibodies to DNA and Sm antigen follow a good response to treatment. With the recent availability of reliable quantitative measurements of antibody levels, a multi-center study is in progress in lupus to scientifically document relationships between therapeutic responses and changes in levels of certain autoantibodies.

Cancer immunotherapy is largely based on the use of peptide antigens derived from amino acid sequences of tumor antigens, and most of the work has been directed at modulating T-cell responses (93, 94). A major problem is the selection of candidate peptides, which must be strongly immunogenic and be able to induce the desired T-cell response. Different strategies have included modifying amino acid residues on candidate epitopes in a proinsulin peptide to induce either regulatory CD4+ or cytotoxic CD8+ T cells in non-obese diabetic mice (95, 96). If observations from studies of autoepitopes on lupus antigens apply to cancer, it would be important to identify regions of TAAs that are recognized by the immune system of the patient, taking this response to indicate realistic in vivo targets and to design immunotherapy directed at such autoepitopes. The feasibility of this approach is that such T-cell autoepitopes can be identified, as shown in the isolation of a cyclin B1 peptide from breast adenocarcinoma major histocompatibility complex class I molecules (97). This autoepitope immunotherapy approach is making use of intelligence reported by the immune system.


Supported by grants from the National Institutes of Health CA56956, 2SO6GM008012-37 and 5G12RR008124.


1. Tan EM. Antinuclear antibodies: diagnostic markers for autoimmune diseases and probes for cell biology. Adv Immunol. 1989;44:93–151. [PubMed]
2. Tan EM. Interactions between autoimmunity and molecular and ell biology. Bridges between clinical and basic sciences. J Clin Invest. 1989;84:1–6. [PMC free article] [PubMed]
3. Notkins A, Lernmark A. Autoimmune type 1 diabetes: resolved and unresolved issues. J Clin Invest. 2001;108:1247–1252. [PMC free article] [PubMed]
4. Rapoport B, McLachlan SM. Thyroid autoimmunity. J Clin Invest. 2001;108:1253–1259. [PMC free article] [PubMed]
5. Rose NR, Mackay IR. The Autoimmune Diseases. New York: Academic Press; 1998.
6. Whitehouse JMA, Holborow EJ. Smooth muscle antibody in malignant disease. Br Med J. 1971;27:511–513. [PMC free article] [PubMed]
7. Steiner M, Klein E, Klein G. Antinuclear reactivity of sera in patients with leukemia and other neoplastic diseases. Immunol Immunopathol. 1975;4:373–381. [PubMed]
8. Nelson DS. Autoantibodies in cancer patients. Pathology. 1977;9:155–158. [PubMed]
9. Thomas PJ, Kaur JS, Aitcheson CT, Robinson WA, Tan EM. Antinuclear, antinucleolar and anticytoplasmic antibodies in patients with malignant melanoma. Cancer Res. 1983;43:1372–1380. [PubMed]
10. Crawford LV, Pim DC, Bulbrook RD. Detection of antibodies against the cellular protein p53 in sera from patients with breast cancer. Int J Cancer. 1982;30:403–408. [PubMed]
11. Tan EM, Shi F-D. Relative paradigms between autoantibodies in lupus and autoantibodies in cancer. Clin Exp Immunol. 2003;134:169–177. [PMC free article] [PubMed]
12. Imai H, Nakano Y, Kiyosawa K, Tan EM. Increasing titers and changing specificities of antinuclear antibodies in patients with chronic liver disease who develop hepatocellular carcinoma. Cancer. 1993;71:26–35. [PubMed]
13. Blumberg BS. Hepatitis B virus, the vaccine and the control of primary cancer of the liver. Proc Natl Acad Sci USA. 1997;94:7121–7125. [PMC free article] [PubMed]
14. Berk PD, Okuda K, Wands JR. Hepatocellular carcinoma. Semin Liver Dis. 1999;19:233–311.
15. Imai H, Furuta K, Landberg G, Kiyosawa K, Liu FL, Tan EM. Autoantibody to DNA isomerase II in primary liver cancer. Clin Cancer Res. 1995;1:417–424. [PubMed]
16. Woessner RD, Mattern MR, Mirabelli CK, Johnson RK, Drake FH. Proliferation and cell cycle-dependent difference in expression of the 170 kilodalton and 180 kilodalton forms of topoisomerase II in NIH-3T3 cells. Cell Growth Diff. 1991;2:209–214. [PubMed]
17. Hsiang Y-H, Wu H-Y, Liu LF. Proliferation-dependent regulation of DNA Topoisomerase II in cultured human cells. Cancer Res. 1998;48:3230–3235. [PubMed]
18. Lubin R, et al. Serum p533 antibodies as early markers of lung cancer. Nat Med. 1955;1:701–702. [PubMed]
19. Soussi T. P53 antibodies in the sera of patients with various types of cancer a review. Cancer Res. 2000;60:1777–1788. [PubMed]
20. Trivers G, et al. Anti-p53 antibodies in sers from patients with chronic obstructive pulmonary disease can predict a diagnosis of cancer. Clin Cancer Res. 1996;2:1767–1775. [PubMed]
21. Trivers GE, et al. Anti-p53 antibodies in sera of workers occupationally exposed to vinyl chloride. J Natl Cancer Inst. 1995;87:1400–1407. [PubMed]
22. Royahem J, Conrad K, Frey M, Melhom J, Frank KH. Autoantibodies predictive parameters of tumor development. In: Konrad K, Humboldt RL, Meurer M, Shoenfeld Y, Tan EM, editors. Pathogenic and Diagnostic Relevance of Autoantibodies. Berlin: Pabst; 1988. pp. 412–414.
23. Posner JB, Dalmau J. Paraneoplastic syndromes. Curr Opin Immunol. 1997;9:723–729. [PubMed]
24. Masumuru K, Darnell RB. Paraneoplastic neurologic disease antigens: RNA-binding proteins and signaling proteins in neurological degeneration. Annu Rev Neurosci. 2001;24:239–262. [PubMed]
25. Keene JD. Why is Hu where? Shuttling of early response messenger RNA subsets. Proc Natl Acad Sci USA. 1999;96:5–7. [PMC free article] [PubMed]
26. Winter SF, Minna JD, Johnson BE, Takahashi T, Gazdor AF, Carbone DP. Development of antibodies against p53 in lung cancer patients appears to be dependent on the type of p53 mutation. Cancer Res. 1992;52:4168–4174. [PubMed]
27. Labrecque S, Naftaly N, Thomson D, Matlas K. Analysis of the anti-p53 antibody response in cancer patients. Cancer Res. 1993;53:3468–3471. [PubMed]
28. Lubin R, Schlichtholz B, Bengoufa D, Soussi T. Analysis of p53 antibodies in patients with various cancers define B cell epitopes of human p53 on primary structure and exposure on protein surface. Cancer Res. 1993;53:5872–5876. [PubMed]
29. Zhang J-Y, Chan EKL, Peng X-X, Tan EM. A novel cytoplasmic protein with RNA-binding motifs is an autoantigen in human cellular carcinoma. J Exp Med. 1999;189:1101–1110. [PMC free article] [PubMed]
30. Burd CC, Dreyfus G. Conserved structures and diversity of functions of RNA-binding proteins. Science. 1994;265:615–621. [PubMed]
31. Nielsen J, Christiansen J, Lykke-Anderson J, Johnson AH, Weaver UM, Nielsen FC. A family of insulin-like growth factor II mRNA-binding proteins represses translation in late development. Mol Cell Biol. 1999;19:1261–1270. [PMC free article] [PubMed]
32. Cariani E, et al. Differential expression of insulin-like growth factor II mRNA in human primary liver cancers, benign liver tumors and liver cirrhosis. Cancer Res. 1988;48:6844–6849. [PubMed]
33. Westley BR, May FEB. Insulin-like growth factors: the unrecognized oncogenes. Br J Cancer. 1995;72:1065–1066. [PMC free article] [PubMed]
34. Doyle GAR, Betz NA, Leeds PF, Fleisig AJ, Prokipcak RD, Ross J. The c-myc coding region determinant binding protein: a member of a family of KH domain RNA-binding proteins. Nucl Acids Res. 1988;26:5036–5044. [PMC free article] [PubMed]
35. Ioannidis P, et al. C-myc and IGF-II mRNA-binding protein (CRD-BP/IMP1) in benign and malignant mesenchymal tumors. Int J Cancer. 2001;94:480–484. [PubMed]
36. Doyle GA, Bordeaux-Heller JM, Coulthard S, Meisner IF, Ross J. Amplification in human breast cancer of a gene encoding a c-myc mRNA binding protein. Cancer Res. 2000;60:2756–2759. [PubMed]
37. Mueller-Pillasch F, et al. Cloning of a gene highly expressed in cancer coding for a novel KH-domain containing protein. Oncogene. 1997;14:2729–2733. [PubMed]
38. Lu M, et al. Aberrant expression of fetal RNA-binding protein p62 in liver cancer and liver cirrhosis. Am J Pathol. 2001;159:945–953. [PMC free article] [PubMed]
39. Covini G, Chan EK, Nishioka M, Morshed SA, Reed SI, Tan EM. Immune response to cyclin B1 in hepatocellular carcinoma. Hepatology. 1997;25:75–86. [PubMed]
40. Suzuki H, Graziano DF, McKolanis J, Finn OJ. T cell-dependent antibody responses against aberrantly expressed cyclin B1 protein in patients with cancer and premalignant disease. Clin Cancer Res. 2005;11:1521–1526. [PubMed]
41. Yu M, Zhau Q, Finn OJ. Immune recognition of cyclin B1 as a tumor antigen is a result of its overexpression in human tumors that is caused by non-functional p53. Mol Immunol. 2002;38:981–987. [PubMed]
42. Rinke J, Steitz JA. Precursor molecules of both human 5S ribosomal RNA and transfer RNAs are bound by a cellular protein reactive with anti-La lupus antibodies. Cell. 1982;29:149–154. [PubMed]
43. Chan EKL, Francoeur AM, Tan EM. Epitopes, structural domains and asymmetry of amino acid residues in SS-B/La nuclear protein. J Immunol. 1986;136:3744–3749. [PubMed]
44. Chan EKL, Tan EM. The small nuclear ribo-nucleoprotein SS-B/La binds RNA with a conserved protease-resistant domain of 28 kD. Mol Cell Biol. 1987;7:2588–2591. [PMC free article] [PubMed]
45. Chan EKL, Tan EM. Human autoantibody-reactive epitopes on SS-B/La are highly conserved in comparison with epitopes recognized by murine monoclonal antibodies. J Exp Med. 1987;166:1627–1640. [PMC free article] [PubMed]
46. Miyachi K, Fritzler MJ, Tan EM. Autoantibody to a nuclear antigen in proliferating cells. J Immunol. 1978;139:2942–2946.
47. Tan C-K, Castillo C, So AG, Downey KM. An auxiliary protein for DNA polymerase delta from fetal calf thymus. J Biol Chem. 1986;261:12310–12315. [PubMed]
48. Prelich C, et al. Functional identity of proliferating cell nuclear antigen and a DNA polymerase-delta auxiliary protein. Nature (London) 1987;326:317–320. [PubMed]
49. Ogata K, Ogata Y, Takasaki Y, Tan EM. Epitopes on proliferating cell nuclear antigen recognized by human lupus autoantibody and murine monoclonal antibodies. J Immunol. 1987;139:2942–2946. [PubMed]
50. Huff JP, Roos G, Peebles CL, Houghten R, Sullivan KF, Tan EM. Insights into native epitopes of proliferating cell nuclear antigen using recombinant DNA protein products. J Exp Med. 1990;172:419–429. [PMC free article] [PubMed]
51. Takasaki Y, Deng JS, Tan EM. A nuclear antigen associated with cell proliferation and blast transformation. Its distribution in synchronized cells. J Exp Med. 1981;154:1899–1909. [PMC free article] [PubMed]
52. Morris GF, Mathews MB. Regulation of proliferating cell nuclear antigen during the cell cycle. J Biol Chem. 1989;264:13856–13864. [PubMed]
53. Muro Y, Tsai W-M, Houghten R, Tan EM. Synthetic compound peptides simulating antigenicity of conformation-dependent autoepitopes. J Biol Chem. 1994;269:18529–18534. [PubMed]
54. Laver WG, Air GM, Webster RG, Smith-Gil SJ. Epitopes on protein antigens: misconceptions and realities. Cell. 1990;61:553–556. [PubMed]
55. Parry NR, Barnett PV, Ouldridge EJ, Rowland DJ, Brown F. Neutralizing epitopes of type 0 foot and mouth disease virus. II. Mapping three conformational sites with synthetic peptide reagents. J Gen Virol. 1989;70:1493–1503. [PubMed]
56. Bidart J-M, et al. Peptide immunogen mimicry of a protein-specific structural epitope on human choriogonadotropin. Science. 1990;248:736–739. [PubMed]
57. Richter W, Shi Y, Baekkeskov S. Autoreactive epitopes defined by diabetes-associated human monoclonal antibodies are localized to the middle and C-terminal domains of the smaller form of glutamic decarboxylase. Proc Natl Acad Sci USA. 1993;90:2832–2836. [PMC free article] [PubMed]
58. Powers AC, Bavik K, Tremble J, Daw K, Scherbaum A, Banga JP. Comparative analysis of epitope recognition of glutamic acid decarboxylase (GAD) by autoantibodies from different autoimmune diseases. Clin Exp Immunol. 1999;118:349–356. [PMC free article] [PubMed]
59. Yang VW, Lerner MR, Steitz JA, Flint SJ. A small nuclear ribonucleoprotein is required for splicing of adenoviral early RNA sequences. Proc Natl Acad Sci USA. 1981;78:1371–1375. [PMC free article] [PubMed]
60. Shero JH, Bordwell B, Rothfield NF, Earnshaw WC. Autoantibodies to topoisomerase I are found in sera from scleroderma patients. Science. 1986;231:737–740. [PubMed]
61. Guldner HH, Szostecki C, Vosberg HP, Lakomek HJ, Penner E, Bautz FA. Scl 70 autoantibodies from scleroderma patients recognize a 96 kDa protein identified as DNA topoisomerase I. Chromosoma. 1986;94:132–138. [PubMed]
62. Agris PF, Parks R, Bowman L, Guenther RH, Kovacs SA, Pelsue S. Plant DNA topoisomerase I is recognized and inhibited by human Scl-70 autoantibodies. Exp Cell Res. 1990;189:276–279. [PubMed]
63. Reimer G, Rose KM, Scheer U, Tan EM. Autoantibody to RNA polymerase I in scleroderma sera. J Clin Invest. 1987;79:65–72. [PMC free article] [PubMed]
64. Bunn CC, Mathews MB. Autoreactive epitope defined as the anticodon region of alanine transfer RNA. Science. 1987;238:1116–1119. [PubMed]
65. Targoff IN. Autoantibodies to aminoacyl-transfer RNA synthetases fos isoleucine and glycine. Two additional synthetases are antigenic in myositis. J Immunol. 1990;144:1737–1743. [PubMed]
66. Bernat RL, Borisy GG, Rothfield NF, Earnshaw WC. Injection of anticentromere antibodies in interphase disrupts events required for chromosome movement at mitosis. J Cell Biol. 1990;111:1519–1533. [PMC free article] [PubMed]
67. Tan C-K, Sullivan K, Li X, Tan EM, Downey KW, So AG. Autoantibody to the proliferating cell nuclear antigen neutralizes the activity of the auxiliary protein for DNA polymerase delta. Nucleic Acids Res. 1987;15:9299–9308. [PMC free article] [PubMed]
68. Dang CV, Tan EM, Traugh JH. Myositis autoantibody reactivity and catalytic function of threonyl-tRNA synthetase. FASEB J. 1988;2:2376–2379. [PubMed]
69. Casiano CA, Martin S, Green DR, Tan EM. Selective cleavage of nuclear autoantigens during CD95 (Fas/APO-1)-mediated T cell apoptosis. J Exp Med. 1996;184:765–770. [PMC free article] [PubMed]
70. Mahoney JH, Rosen A. Apoptosis and autoimmunity. Curr Opin Immunol. 2005;17:583–588. [PubMed]
71. Gaipl US, et al. Inefficient clearance of dying cells and autoreactivity. Curr Top Microbiol Immunol. 2006;306:161–176. [PubMed]
72. Notman DD, Kurata N, Tan EM. Profiles of antinuclear antibodies in systemic rheumatic diseases. Ann Int Med. 1975;83:464–469. [PubMed]
73. Zhang J-Y, et al. Autoimmune responses to mRNA binding proteins p62 and Koc in diverse malignancies. Clin Immunol. 2001;100:149–156. [PubMed]
74. Zhang J-Y, Casiano CA, Peng X-X, Koziol JH, Chan EKL, Tan EM. Enhancement of antibody detection in cancer using panel of recombinant tumor-associated antigens. Cancer Epid Biomarkers Prev. 2003;12:136–143. [PubMed]
75. Shi F-D, et al. Preferential humoral immune response in prostate cancer to cellular proteins p90 and p62 in a panel of tumor-associated antigens. Prostate. 2005;63:252–258. [PubMed]
76. Koziol JA, et al. Recursive partitioning as an approach to selection of immune markers for tumor diagnosis. Clin Cancer Res. 2003;9:5120–5126. [PubMed]
77. Imai H, Chan EKL, Kiyosawa K, Fu X-D, Tan EM. Novel nuclear autoantigen with splicing factor motifs identified with antibody from hepatocellular carcinoma. J Clin Invest. 1993;92:2419–2426. [PMC free article] [PubMed]
78. Tureci O, Sahin U, Pfreundschuh M. Serological analysis of human tumor antigens: molecular definition and implications. Mol Med Today. 1997;3:342–349. [PubMed]
79. Fernandez-Madrid F, Tang N, Alansari H, Karoonen RL, Tonkiel JE. Improved approach to identify cancer-associated autoantigens. Autoimmun Rev. 2005;4:230–235. [PubMed]
80. Somers VA, et al. A panel of candidate tumor antigens in colorectal cancer revealed by serological selection of a phage displayed cDNA expression library. J Immunol. 2002;169:2772–2780. [PubMed]
81. Sioud M, Hansen MH. Profiling the immune response in patients with breast cancer by phage-displayed cDNA libraries. Eur J Immunol. 2001;31:716–725. [PubMed]
82. Fossa A, Alsoe L, Crameri R, Funderad S, Gaudernack G, Smeland EB. Serological cloning of cancer/testis antigen expressed in prostate cancer using a cDNA expression phage surface display. Cancer Immunol Immunother. 2004;53:431–438. [PubMed]
83. Chatterjee M, et al. Diagnostic markers of ovarian cancer by high-throughput antigen cloning and detection on arrays. Cancer Res. 2006;66:1181–1190. [PMC free article] [PubMed]
84. Klade CS, et al. Identification of tumor antigens in renal cell carcinoma by serological proteome analysis. Proteomics. 2001;1:890–898. [PubMed]
85. Canelle L, et al. An efficient proteomics-based approach for the screening of autoantibodies. J Immunol Methods. 2005;299:77–89. [PubMed]
86. Joos TO, et al. A microarray enzyme-linked immunosorbent assay for autoimmune diagnostics. Electrophoresis. 2000;21:2641–2650. [PubMed]
87. Robinson WH, et al. Autoantigen microarrays for multiplex characterization of autoantibody responses. Nature Med. 2002;8:295–301. [PubMed]
88. Balboni I, Chan SM, Kattah M, Tenenbaum JD, Butte AJ, Utz PJ. Multiplexed protein array platform for analysis of autoimmune disorders. Annu Rev Immunol. 2006;24:391–418. [PubMed]
89. Imai H, Kiyosawa K, Chan EKL, Tan EM. Autoantibodies in viral hepatitis-related hepatocellular carcinoma. Intervirology. 1993;35:73–85. [PubMed]
90. Tan EM. Autoantibodies as reporters identifying aberrant cellular mechanisms in tumorigenesis. J Clin Immunol. 2001;108:1411–1415. [PMC free article] [PubMed]
91. Anderson KS, La Baer J. The sentinel within: exploiting the immune system for cancer biomarkers. J Proteome Res. 2005;4:1123–1133. [PMC free article] [PubMed]
92. Egloff AM, Vella LA, Finn OJ. Cyclin B1 and other cyclins as tumor antigens in immuno-surveillance and immunotherapy of cancer. Cancer Res. 2006;66:6–9. [PubMed]
93. Soares MM, Mehta V, Finn OJ. Three different vaccines based on the 140-aminoacid MUC1 peptide with seven tandemly repeated tumor-specific epitopes elicit distinct immune effector mechanisms in wild-type versus MUC1 transgenic mice with different potential to tumor rejection. J Immunol. 2001;166:6555–6563. [PubMed]
94. Yu Z. Restifo NP. Cancer vaccines: progress reveals new complexities. J Clin Invest. 2002;110:289–294. [PMC free article] [PubMed]
95. Pugliese A. Peptide-based treatment for autoimmune diseases: learning how to handle a doublepedged sword. J Clin Invest. 2003;111:1280–1282. [PMC free article] [PubMed]
96. Martinez NR, et al. Disabling an integral CTL epitope allows suppression of autoimmune diabetes by intranasal proinsulin peptide. J Clin Invest. 2003;111:1365–1371. [PMC free article] [PubMed]
97. Kao H, et al. Identification of cyclin B1 as a shared human epithelial tumor-associated antigen recognized by T cells. J Exp Med. 2001;194:1313–1323. [PMC free article] [PubMed]
PubReader format: click here to try


Save items

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem chemical substance records that cite the current articles. These references are taken from those provided on submitted PubChem chemical substance records.

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...