Evaluation of a panel of tumor-associated antigens in breast cancer

Abstract

BACKGROUND:

Recent studies indicate that serum from cancer patients contains auto-antibodies against oncoproteins so called tumor-associated antigens (TAAs), which represent promising diagnostic and prognostic biomarkers.

OBJECTIVES:

In this study we searched for breast cancer-associated auto-antibodies against individual TAAs. Also we evaluated if a panel of multiple TTAs would improve the detection of auto-antibodies. We screened CEA, CCBN1, c-Myc, p53, Ki-67, Nm23, PRDX6, eIF5A, PARK7, GLIO-1, Hsp27 and Hsp70 proteins, previously detected as up-regulated in breast tumors of Mexican patients.

METHODS:

Enzyme-linked immunosorbent assays (ELISA) were performed to detect auto-antibodies in sera from a cohort of 104 breast cancer patients and 50 sera from healthy individuals.

RESULTS:

Our data showed that antibodies frequency to any individual TAA was low and ranged from 0.96% to 4.8%. However, the successive addition of multiple TAAs represented by panels of three-to-five TAAs resulted in increased ELISA positive reactions. The first panel of three combined TAAs (p53/PRDX6/CEA) had a sensitivity of 19%, while a second set of four TAAs (p53/PRDX6/c-Myc/Hsp70) reached 28% sensitivity. Likewise, a third panel of five antigens (p53/PRDX6/c-Myc/Hsp70/Nm23) showed 34% sensitivity.

CONCLUSIONS:

Our data showed that detection of individual autoantibodies against TAAs in the cohort of patients analyzed here was low, which was enhanced by adding multiple TAAs. Data support the notion that frequencies of autoantibodies could be impacted by geographical and heterogeneous genetic factors of breast cancer patients.

Keywords

Breast cancer tumor-associated antigens autoantibodies immunodiagnosis

1. Introduction

Breast cancer is the neoplasia with the highest incidence and mortality affecting women worldwide [1]. Despite of recent advances in diagnosis and genomic analysis of tumors translated to personalized therapies about 40% of breast cancer patients succumb to the disease. The high mortality rate is attributed to late diagnosis associated to the presence of no clinically detectable metastases at the time of the initial diagnosis. This scenario may be worst by a lack of effective and specific diagnostic tests allowing early detection. Although, the widespread use of screening mammography has resulted in increased detection of early-stage breast disease, an important number of women with abnormalities are not diagnosed [2]. Therefore, there is an important need to ameliorate screening and diagnosis of early-invasive and advanced tumors in order to improve breast cancer outcomes. The potential for molecular tools based on early diagnosis has driven the search for diagnostic biomarkers particularly in serum. Hence, numerous genomic and single-gene studies are being conducted to identify and validate novel early biomarkers in order to increase the sensitivity and specificity of breast cancer detection [3, 4].

Recent studies have demonstrated that serum from breast cancer patients may contain specific antibodies that react against a group of autologous antigens so called tumor-associated antigens (TAAs) [5]. These tumor-associated autoantibodies are detectable in patients with a variety of human malignancies [6]. Identification of tumor-specific markers capable of eliciting an immune response in the early stages of tumor development seems to provide an effective approach for early diagnosis. The potential utility of TAA-autoantibody systems as early cancer biomarker tools to monitor therapeutic outcomes or as indicators of disease prognosis has been investigated. Here we evaluated whether multiple TAAs would enhance autoantibody detection and be an effective tool in the immunodiagnosis breast cancer.

2. Materials and methods

2.1 Serum samples

Sera from 104 patients with breast cancer and 50 normal individuals were obtained from the National Institute of Cancerology, Mexico and 20 November Hospital, IMSS, Mexico, respectively. Samples were obtained following the regulations approved by the Institute Ethics Committee, including patient informed consent and anonymatization prior to release for research use. Based on clinical information all cancer sera were collected at the first time of diagnosis and none of the patients recruited in this study received any anti-neoplastic therapy prior to surgery. Collected sera were snap frozen in liquid nitrogen until analysis.

2.2 Recombinant TAAs

All recombinant proteins corresponding to potential TAAs used in the present study were obtained from the next commercially companies: c-MYC, CCNB1, Ki-67, Hsp70, andPRDX6 were purchased from Abcam; CEA, eIF5A, Nm23, Hsp27, and PARK7 were obtained from Prospec, whereas GLIO-1 and p53 were purchased from Abnova and Millipore, respectively.

2.3 Enzyme-linked immunosorbent assay (ELISA)

Purified recombinant TAAs were individually diluted in PBS to a final concentration of 0.5 $\mu$ g/ml and 200 $\mu$ l were pipetted into each well to coat Immulon 2 microtiter plates (Fisher Scientific, Houston, TX, USA) overnight at 4 ${}^{\circ}$ C. The human serum samples were diluted at 1:200, incubated with the antigen-coated wells at 37 ${}^{\circ}$ C for 90 min followed by washing with PBS containing 0.05% Tween-20. The samples were then incubated with horseradish peroxidase (HRP)-conjugated goat anti-human IgG (Caltag Laboratories, Burlingame, CA, USA) as a secondary antibody diluted 1:2,000 for 90 min followed by washing with PBS containing 0.05% Tween-20. A solution of 3, 3’, 5, 5’-tetramethyl benzidine (TMB)-H2O2-urea was used as the detecting agent. The OD of each well was read at 450 nm. Each sample was tested in duplicate. The cut-off value for determining a positive reaction was designated as the mean absorbance of the 82 normal human sera (NHS) plus 2 standard deviations (mean $+$ 2SD). Each ELISA run included at least 8 NHS samples and 2 positive control samples. These 8 NHS samples, representing a range of 2SD above and below the mean of the 82 NHS, were used in each experiment and the average value of the 8 NHS samples was used in each run to normalize all absorbance values to the mean of the entire 82 normal samples. In addition, all positive sera were confirmed with repeat testing, as were certain negative sera.

Table 1
Frequency, sensitivity and specificity of anti-TAAs antibodies in breast cancer patients and healthy individuals

	Frequency (%)
Tumor-associated	Breast cancer		Healthy	Sensitivity		Specificity	Youden’s
antigens (TAAs)	( $n=$ 104)		( $n=$ 50)	(%)		(%)	index
P53	5	(4.8) ${}^{\text{a}}$	0 (0.0)		4.8	100		0.048
DJ1	2	(1.9)	1 (2.0)		1.9	98	$-$	0.001
Hsp27	0	(0.0)	0 (0.0)		0	–	–
Hsp70	3	(2.8)	2 (4.0)		2.88	96	$-$	0.011
c-Myc	4	(3.8) ${}^{\text{a}}$	1 (2.0)		3.8	98		0.018
Ki67	3	(2.8)	1 (2.0)		2.88	98		0.008
eIFG5A	2	(1.9)	2 (4.0)		1.9	96	$-$	0.021
PRDX-6	4	(3.8) ${}^{\text{a}}$	1 (2.9)		3.8	98		0.018
CEA	2	(1.9)	2 (4.0)		1.9	96	$-$	0.021
GLIO-1	1	(0.96) ${}^{\text{a}}$	2 (4.0)		0.96	96	$-$	0.030
Nm23	3	(2.8) ${}^{\text{a}}$	1 (2.0)		2.88	98		0.00
Cyclin B1	1	(0.96) ${}^{\text{a}}$	0 (0.0)		0.96	100		0.96

Relative to healthy individuals: ${}^{\text{a}}P<$ 0.01.

2.4 Statistical analysis

To determine whether the frequency of autoantibodies binding to select TAAs in the cancer sera was significantly higher than that in NHS, the data were analyzed using the $\chi^{2}$ tests with Yates’ correction. Two levels of statistical significance (0.05 and 0.01) were used and $P<$ 0.05 was considered to indicate statistically significant differences. Specificity was calculated as: number of negative healthy individuals/number of negative healthy individuals plus number of positive healthy individuals. Sensitivity was calculated as: number of positive patients/number of positive patients plus number of negative patients. The comprehensive evaluations of the testing results for each anti-TAA antibody, including the methods for calculating the sensitivity, specificity, Youden’s index (YI), positive and negative likelihood ratio, positive (PPV) and negative predictive value (NPV) were calculated.

Table 2
Sensitivity and specifity of TAAs

Tumor-associated antigens	P (21)	Sensitivty (%)	Specifity (%)	IY	$+$ VP (%)	$-$ VP (%)
p53/PRDX6/CEA	4	19	100	0.19	100	29.2
p53/PRDX6/c-Myc/Hsp70	6	28	100	0.28	100	31.8
p53/PRDX6/c-Myc/Hsp70/Nm23	5	24	100	0.24	100	30.4

3. Results

3.1 Prevalence of antibodies against individual TAAs in breast cancer

In order to evaluate the frequency of autoantibodies against individual TAAs, we collected sera from breast cancer patients. The age range of patients was 26–69 years. Clinical stage of patients was stage II ( $n=$ 19), stage III ( $n=$ 80), and stage IV ( $n=$ 5), and tumors corresponded to ductal ( $n=$ 94) and lobulillar ( $n=$ 4) carcinoma. Classification of breast tumors based on expression of hormonal receptors was luminal ( $n=$ 69), HER2 positive ( $n=$ 17), and triple negative ( $n=$ 18). To asses the presence of anti-TAAs antibodies in serum we analyzed by ELISA a panel of twelve selected recombinant proteins, five of them previously reported as TAAs including Myc proto-oncogene (c-Myc), cellular tumor antigen p53 (p53), carcynoembrionic antigen (CEA), antigen Ki-67, and cyclin B1 (CCNB1). Myc is a transcription factor commonly upregulated in breast cancer, whereas p53 is a tumor suppressor inactivated in diverse types of cancer [7, 8]. Nuclear antigen Ki67 is proliferation marker, whereas CEA is a glycoprotein involved in cell adhesion, both used as progression and therapy response markers in cancer [9, 10]. Cyclin B1 is cell cycle related protein related to tumor progression and prognosis [11]. These proteins are well known markers in diverse types of tumors including breast cancer and greatly contribute to tumorigenesis.

Also, we tested five additional potential TAAs representing proteins previously detected as overexpressed in tumors from Mexican breast cancer patients [12] such as peroxirredoxin-6 (PRDX6) and glyoxalase 1 (GLO1) metabolism proteins; deglycase DJ-1 (PARK7); nucleoside diphosphate kinase A (Nm23) metastasis-associated protein, oncogenic heath shock protein-27 (Hsp27) and heath shock protein-70 (Hsp70) and eukaryotic translation initiation factor 5A-1 (eIF5A). These proteins were identified from comparative proteomic profiling of ductal breast tumors ( $n=$ 7) versus non-tumoral tissues ( $n=$ 5) using two-dimensional electrophoresis and LC/ESI-MS/MS tandem mass spectrometry and have been related to tumor progression and prognosis [12]. Our data showed that antibody frequency for any individual TAA in breast cancer was lower in comparison to those previously reported from diverse geographic locations. Frequencies were variable ranging between 7.3 and 22.0%. The highest antibodies frequencies were p53 (4.8%) and c-Myc (3.8%) followed by Hsp70 (2.88%), Ki-67 (2.88%), nm23 (2.88%), eIF5A (1.9%), CEA (1.9%), CCNB1 (0.96%), and GLO1 (0.96%), (Table 1). Intriguingly, no autoantibodies against Hsp27 were detected in serum samples.

3.2 Multiple TAAs enhances the detection of autoantibodies

As autoantibodies were detected in a low range, we then evaluate whether the addition of multiple TAAs yields higher sensitivity for the detection of autoantibodies in breast cancer. Combinations of three, four and five selected recombinant TAAs were tested in serum from breast cancer patients and healthy individuals. Our data showed that with the successive addition of TAAs to a total of four (p53/PRDX6/c-Myc/Hsp70) and five antigens (p53/PRDX6/c-Myc/Hsp70/Nm23) resulted in a stepwise increase in sensitivity up to 28% and 24%, respectively. Moreover, diagnostic specificity was 100% in both panels. The addition of c-Myc and Hsp70 antigens had a positive effect as we detected a significant increase in sensitivity up to 24%–28%. In contrast, when metastasis associated Nm23 protein was added to the panel of four antigens, there was no further increase in sensitivity in Table 2.

4. Discussion

Numerous studies have confirmed that sera from cancer patients may contain auto-antibodies that react with a group of autologous cellular antigens, thus they have been dubbed tumor-associated antigens (TAAs) [12]. In addition, the available data show that a number of the target antigens are cellular proteins whose aberrant regulation or overexpression are capable of leading to tumorigenesis, and should be used as potential biomarkers as its levels are increased in serum samples [13, 14]. It has been also demonstrated that the immune surveillance system in certain cancer patients has the capability of recognizing these antigenic can amplify the response signal to the TAAs [13, 14]. Thus, cancer-associated autoantibodies might be considered as reporters identifying aberrant or disregulated cellular mechanisms in tumorigenesis. However, because of the low sensitivity, no single marker has been used for early diagnosis. Many studies have shown that the sensitivity can be enhanced by using a mini-array of multiple TAAs as target antigens. For example, a previous study showed that the frequency of antibodies against any individual antigen rarely exceeded 15–20% [14], but with the successive addition of antigens to the panel, there was a stepwise increase in the percentage of positive reactors between 44% and 68% against a combination of seven antigens [14].

Here we evaluated if a panel of TAAs can enhance autoantibody detection in serum from Mexican breast cancer. Antigens analyzed in this study were selected based on its overexpression status revealed from proteomic studies in tumors from breast cancer patients [12]. We examined the feasibility of using a panel of autoantibodies to multiple TAAs as an easy method for early detection of breast cancer and, more particularly, in ductal locally invasive carcinoma. Thus it’s important to remark the limitations of the present study as these panels could be potentially applied only for breast cancer patients. Data showed that antibodies frequency to any individual TAA was low and ranged from 0.96% to 4.8%. Interestingly, the successive addition of multiple TAAs represented by panels of three-to-five TAAs resulted in increased reactivity. The panel of three TAAs (p53/PRDX6/CEA) had a sensitivity of 19%, while a second set of four TAAs (p53/PRDX6/c-Myc/Hsp70) reached 28% sensitivity. A third panel of five antigens (p53/PRDX6/ c-Myc/Hsp70/Nm23) showed 34% sensitivity. These results indicate that, at least for the tested antigens, the frequency was low in comparison to other TAAs described in several reports, which could be not sufficient to build a reliable screening test. We conclude that the low frequencies of TAAs found could be impacted by geographical and heterogeneous genetic/epigenetic factors inherent to Mexican breast cancer patients, which have a very heterogeneous genetic and mixed background between mestizo, indigenous and European/African ancestries, ethnic groups and races. At this point, we have no data about the genetic and genomic (epi-) features of the patients positively associated to the panel of TAAs, which remains as an interesting topic of research. Our data also confirm a serious fact indicating that serum biomarkers with higher a near to 100% of sensitivity are not easy to find and replicate in closed populations, where the inherent tumor heterogeneity in patients still represent a big challenge. Although our panel of TAAs has a limited diagnostic potential, it provides important information about the feasibility of use of these particular set of TAAs, which can be reevaluated in future studies associated to variables as therapy response, hormone receptors expression, etc. Unfortunately, no similar studies searching for TAAs in Mexican cancer patients have been reported (PubMed and Google scholar search: TAAs cancer Mexico, Oct 2019), thus we can’t make comparisons with our data. Finally, our results from this first screening orientate us to study and redefine new additional proteins for evaluation as potential prognostic markers. Our data also highlights the limitations to find biomarkers with very high frequency in mixed population of Mexican people. In conclusion, our results suggested that frequencies of autoantibodies could be impacted by geographical and heterogeneous genetic factors inherent to Mexican breast cancer patients. Thus, further research will focus on increasing the number of tested TAAs in order to identify other relevant antigens for breast cancer detection.

Footnotes

Acknowledgments

We acknowledge to Secretaría de Ciencia, Tecnología e Innovación (SECITI) del Distrito Federal for funding this research (Grant ICYTDF/214/2012). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. We also acknowledge to Universidad Autónoma de la Ciudad de México (UACM) for support. Cecilia Pagaza-Straffon was supported by CONACYT (341800) fellowship.

Conflict of interest

Authors declare no conflicts of interests.

References

Ferlay

et al., Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008, International Journal of Cancer 127 (2010), 2893–2917.

Mandelson

M.T.

et al., Breast density as a predictor of mammographic detection: comparison of interval- and screen-detected cancers, Journal of the National Cancer Institute 92 (2000), 1081–1087.

Murphy

Millar

and Lee

C.S.

, Gene expression profiling in breast cancer: towards individualising patient management, Pathology 37 (2005), 271–277.

C.I.

et al., Discovery and preliminary confirmation of novel early detection biomarkers for triple-negative breast cancer using preclinical plasma samples from the Women’s Health Initiative observational study, Breast Cancer Research and Treatment 135 (2012), 611–618.

Reuschenbach

von Knebel Doeberitz

and Wentzensen

, A systematic review of humoral immune responses against tumor antigens, Cancer Immunology Immunotherapy 58 (2009), 1535–1544.

Zhang

J.Y.

and Tan

E.M.

, Autoantibodies to tumor-associated antigens as diagnostic biomarkers in hepatocellular carcinoma and other solid tumors, Expert Review of Molecular Diagnostic 10 (2010), 321–328.

Dang

C.V.

, MYC on the path to cancer, Cell 149 (2012), 22–35.

Hong

et al., Targeting tumor suppressor p53 for cancer therapy: strategies, challenges and opportunities, Current Drug Targets 15 (2014), 80–89.

Hashmi

A.A.

et al., Ki67 index in intrinsic breast cancer subtypes and its association with prognostic parameters, BMC Research Notes 12 (2019), 605.

10.

Yang

K.L.

et al., Carcinoembryonic antigen (CEA) level, CEA ratio, and treatment outcome of rectal cancer patients receiving pre-operative chemoradiation and surgery, Radiation Oncology 8 (2013), 43.

11.

et al., Prognostic role of cyclin B1 in solid tumors: a meta-analysis, Oncotarget 8 (2017), 2224–2232.

12.

Fonseca-Sánchez

M.A.

et al., Breast cancer proteomics reveals a positive correlation between glyoxalase 1 expression and high tumor grade, International Journal of Oncology 41 (2012), 670–680.

13.

Tan