ANAs have been known to be present in BC sera for several decades [
25] but their significance remains unknown [
24]. This is likely because autoantibodies are part of the normal immune response, and sera from healthy subjects exhibit a plethora of autoantibodies not related to cancer [
30-
32]. The application of genomics and proteomics to biomarker discovery allowed the identification of multiple autoantibodies in BC sera recognizing TAAs [
3-
10]. These studies strongly suggested the possibility that autoantibodies in cancer sera were potentially useful biomarkers for the early diagnosis of BC. The seminal work establishing the role of autoantibodies as diagnostic biomarkers in the rheumatic ADs [
16-
20] suggested the hypothesis that the model epitomized by the rheumatic ADs is highly relevant to explain the plethora of autoantibodies detected in cancer sera. Importantly, PBC as an organ-specific autoimmune disease is characterized by a set of autoantibodies with mitochondrial specificity with recognized diagnostic value [
21-
23]. In this work we attempted to demonstrate that autoantibodies in BC sera have unique immunological features, as in those found in the rheumatic and some organ-specific ADs. We show here that autoantibodies reacting with antigens located in several important cell organelles (including mitochondria, centromeres, nucleoli, centrosomes, and the mitotic spindle) are consistently found in sera from women with suspicious mammography findings. The level of these autoantibodies was highest in women with IDC, lesser in women with DCIS, and still lower but above background levels in healthy women with BBD [Table
2]. Moreover, AMAs, anti-centromere, and anti-centrosome antibodies are not components of the autoantibody repertoire of normal healthy women [
30-
32]. These findings suggest that, in the future, the combination of suspicious mammography and autoantibody signatures could potentially identify a group of women in the early stages of breast carcinogenesis. Here we provide evidence that most of the antigens targeted by autoantibodies in BC sera differ from those involved in the rheumatic and organ-specific ADs. Sera from patients with SLE frequently exhibit a homogeneous pattern and anti-dsDNA or anti-histone antibodies [
18,
20]. Although anti-dsDNA antibodies were not found in sera from women with homogeneous pattern of nuclear fluorescence, we have detected anti-histone antibodies in BC sera by immunoscreening a cDNA library of BC proteins with BC sera [unpublished data, FFM et al.]. The speckled pattern is found in scleroderma and other rheumatic ADs [
16,
20] and very frequently in IDC of the breast. None of the BC sera exhibiting the speckled pattern [Table
2 and Figure
2] reacted with ENAs on ELISA as is frequently the case in the rheumatic ADs. Anti-centromere antibodies which are characteristic of limited scleroderma, or CREST syndrome [
34], were found in the sera from cases of BC and BBD controls. Most of the anti-CENPs detected in BC sera in this work [Table
2 and Figure
4] were anti-CENP-B, which have been reported in BC sera [
35] and are prevalent in the rheumatic ADs [
16,
20]. Centromere protein abnormalities reflected by the presence of CENP antibodies are clearly common in BC. The consistent finding of CENP antibodies in BBD sera is notable, since these antibodies are seldom found in healthy women [
23,
30-
32].
Autoantibodies reacting with centrosome antigens are common in women with BC and BBD
Autoimmune sera contain autoantibodies targeting epitopes found in a family of proteins located on centrosomes [
36]. Centrosome aberrations have long been reported in invasive and pre-invasive cancer [
37,
38]. In our study, anti-centrosome antibodies were detected frequently in both BC and BBD sera [Table
2 and Figure
8], but these autoantibodies were not prominent features in previous studies in healthy subjects [
30-
32]. Abnormal centrosome amplification and supernumerary centrosomes, as well as abnormalities in centrosome number, size, and morphology, have been observed in nearly all human tumor types including BC [
38]. Centrosome defects have been associated with genetic instability [
37,
38] but the role of the centrosome in tumorigenesis is yet to be defined. The significance of anti-centrosome antibodies and autoantibodies reacting with proteins in the spindle apparatus in cancer sera is unclear. The novel findings reported here suggest that autoantibodies in BC sera are promising probes to identify centrosome proteins likely to be implicated in both autoimmunity and cancer. Chromosomal aberrations are the hallmark of cancer and autoantibodies develop early in carcinogenesis. Thus, the possibility should be investigated that autoimmunity to centrosome and mitotic spindle proteins may be involved in inducing genetic instability.
High titer AMAs are frequently found in BC sera
AMAs detected by IFA in sera from women with BC and BBD [Figure
5] are indistinguishable from the mitochondrial fluorescence typically detected in the sera from patients with PBC [
21-
23]. AMAs reacting with the M2 mitochondrial antigen complex are diagnostic markers for PBC [
22]. One serum from a woman with IDC of the breast reacted with the M2 antigen complex, suggesting that most of the AMAs in BC sera have specificities other than those found in PBC. The cases and BBD controls in our cohort that displayed AMAs in their sera did not have liver disease. Although liver function was normal in the only BC patient whose serum reacted with the M2 antigen complex, it is possible that this result could be due to coincidental PBC developing in a patient with BC. AMAs, MNDs, and anti-centromere antibodies as seen in BC sera in our study are classically detected in PBC [
21-
23,
33] [Table
2, Figures
4,
5A-D and
7B]. Another similarity between PBC and BC is that ANAs are found in both conditions. The resemblance between the AMAs in BC and PBC, however, is limited to the mitochondrial origin of the antigens since most of the AMA-positive sera in BC did not react with the M2 antigen complex. This is in agreement with our report of two mitochondrial proteins, peripheral benzodiazepine associated protein-1 [PRAX-1] [
39] and diazepam binding inhibitor related protein [
40] recognized as autoantigens by BC sera [
9,
41]. Furthermore, although AMAs predominate over ANAs in PBC [
21,
23], ANAs predominate over AMAs in BC sera. MNDs are commonly found in PBC and with less frequency in the sera from BC patients. While in PBC, MNDs are associated with the NSP1 antigen [
33], the MNDs seen in BC sera [Table
2 and Figure
7B] do not seem to be related to NSP1 antibodies. Thus, the distinctive serologic findings (AMAs, MNDs, centrosome, and nucleolar staining) observed in BC sera appear to reflect a distinct autoantibody repertoire, suggesting that autoimmunity to TAA residing in breast tissue is a prominent feature in BC.
The autoantibodies found in some healthy women with BBD may be generated during the pre-malignant phase
Although many normal subjects exhibit low titer ANAs in their sera, relatively few healthy individuals have positive ANA tests at high titers [
20,
30-
32]. The results of our study are not strictly comparable to previous reports since all our cases and controls were women, and women are known both to have higher levels of autoreactivity and to develop more robust immune responses than men [
42]; this may partly explain the higher levels of autoantibodies in our group of healthy women with BBD. Nevertheless, the results from IFA in the sera from women with suspicious mammography assessment were notable for two reasons: the findings of relatively high ANA titers [≥1:320–640, Table
1] with high frequency of mixed patterns [Table
2] in some women with BBD, and the detection of AMAs and antibodies to both centromeres and centrosomes in these sera [Table
2 and Figures
4C,
5C and D, and Figure
8, lower row]. These results indicate that a group of healthy women undergoing annual screening mammography and shown by breast biopsy not to have BC have serologic findings not commonly found in healthy women [
30-
32]. These data suggest the hypothesis that the sera from some women in the control group having BBD may contain signatures that could in the future help identifying women at high risk for BC.
Are autoantibodies participants in breast carcinogenesis as well as being biomarkers?
Both rheumatic ADs and cancer develop after a considerable latent period [
43,
44]. The evidence for multiple cooperating events in the pathogenesis of the rheumatic ADs resembling those observed in multistep carcinogenesis has been noted [
43]. Somatic mutations are considered to be major factors in the pathogenesis of both autoimmunity and carcinogenesis [
43,
44]. Our findings indicate that autoantibodies in BC sera have unique immunological features, as do rheumatic and organ-specific ADs; this suggests that autoantibodies in BC sera are not epiphenomena and that they may, as in the rheumatic ADs, be participants in the process of carcinogenesis. Until recently, the participation of B cell activation and autoantibodies in the anti-cancer immune response attracted little attention [
2]. However, infiltrating lymphocytes including B and T cells have been well described in breast tumor tissue, suggesting an antigen-driven immune response triggered by TAAs [
45-
47]. A causal relationship between chronic inflammation and cancer has been established for many solid tumors [
48]. Recent studies have proposed a possible role of antibody and/or cytokine-mediated effects of B cells on cancer cells, paradoxically potentiating disease progression [
11]. Although the consistent presence of activated B cells in BC tumor tissue has been clearly demonstrated, the mechanism by which B cells may promote cancer progression has not been established. In view of our findings, investigation of a hypothesis in the context of an appropriate genetic background is warranted regarding tumor antigen-triggered autoimmunity, and how it may inflict epithelial damage in the breast as the target organ through promotion of chronic inflammation and cancer progression. Future research based on this study may clarify whether immunotherapy in the treatment of BC should attempt to stimulate or suppress immunogenicity of TAAs in the efforts to modulate the immune system.