CD80 down-regulation is associated to aberrant DNA methylation in non-inflammatory colon carcinogenesis

Immune surveillance of nascent cancer cells is a fundamental process mediated by T-cells, macrophages, and natural killer cells that work to protect host against the unrestrained proliferation of transformed cells [1]. Full T-cell activation requires the engagement of T-cell receptors with specific major histocompatibility complexes at the surface of antigen-presenting cells in the presence of adequate co-stimulatory signals provided by CD80 or CD86 [2]. Oncogenic insults inducing CD80 expression seem then to be able to modulate the anti-tumor immune response [3]. The significant overexpression of CD80 in the colonic mucosa of patients with ulcerative colitis (UC) and dysplasia as opposed to CD80 down-regulation in non-inflammatory colon cancer [4‐6] have led to the hypothesis that the lack of positive co-stimulatory molecules is one of the main mechanisms by which colorectal cancer (CRC) escapes immune surveillance [7]. The molecular mechanisms underlying immune surveillance remain, nevertheless, largely unknown.

An altered methylation pattern in cancer cell genomes is a well recognized characteristic of tumor cells, and specific aberrant methylation events take place during the early stages of colorectal carcinogenesis leading to profound modifications in gene expression [8]. The aberrant methylation of H-cadherin (CDH13) commencing at an early stage of colorectal tumorigenesis frequently silences, in fact, the expression of this tumor suppressor gene in colorectal adenomas and cancers [9]. Besides germ-line mutations associated with hereditary familial adenomatous polyposis and somatic mutations in sporadic colorectal tumors, hypermethylation also provides an important mechanism underlying impaired APC function [10]. Moreover, the hypermethylation of the CpG island within the DNA-repair protein O-6-methylguanine-DNA-methyltransferase (MGMT) gene [11] and in the MLH1 gene is associated with a reduced gene expression observed in the majority of sporadic primary CRC with microsatellite instability [12]. Finally, RUNX3 hypermethylation decreases TGF-β/BMP signaling in gastrointestinal cancer cells [13].

In addition to oncogenes or oncosuppressors, some immune stimulatory molecules are regulated in cancer cells by DNA methylation in their promoter regions. MHC class I and its antigen presentation machinery have been shown to be regulated by DNA methylation [14‐17]. Adhesion molecules [14‐18], such as ICAM-1 and LFA-3, and costimulatory molecules [17, 18], such as CD40 and CD86, can be regulated by DNA methylation in cancer cells. Thus, demethylation treatment can dramatically increase the susceptibility of some tumor cells to T-cell destruction [15, 19‐21]. Our first step in investigating the role of DNA methylation in CRC immune surveillance consisted in examining the expression of CD80; we then went on to analyze its relationship with genomic methylation during the early stages of colon carcinogenesis.

Cancer immune surveillance is a fundamental process by which immune cells protect against cancer formation by identifying and eliminating tumor cells on the basis of their expression of tumor-specific and stress-induced antigens [1]. While it has been suggested that the lack of positive co-stimulatory molecules is the mechanism by which tumor cells evade immune surveillance, the molecular events underlying this process remain elusive.

In this study, we show that CD80 mRNA expression, the co-stimulatory molecule found in colon carcinogenesis, inversely correlates with the inducible DNMTs expression in healthy tissue of patients with adenoma or cancer and with the number of genes that have promoters which are often aberrantly hypermethylated at early stages of colorectal tumorigenesis and, as a consequence, thought to contribute to CRC pathogenesis [8]. Indeed, APC, CDH13, MGMT1 and RUNX3 methylation frequency of gene promoters as well as DNMTs expression progressively increases along the pathway towards invasive cancer in sporadic colonic carcinogenesis. Taken together these data suggest that DNA methylation plays a role in CD80 expression. This association has been found to be notable at the dysplastic stage of carcinogenesis when the CD80-driven immune surveillance process might be critical in preventing cancer cells from escaping [6, 22]. But although two prospective studies characterized by adequate follow-up and precise definition of the adenoma site failed to show complete regression or spontaneous reduction of polyps [23, 24], the findings of a more recent epidemiological investigation suggest that dysplastic adenoma prevalence in humans is caused by a dynamic process including both adenoma formation and regression [25]. Our data are, indeed, the first to suggest that DNA methylation of CD80 gene may explain why immune surveillance mechanisms can at times fail in patients with non-inflammatory colorectal carcinogenesis.

Interestingly, a recent investigation reported hypermethylation of the CD80 promoter in mice tumours, and treatment with decitabine, a DNA methyltransferase inhibitor, was found to enhance CD80 expression in EL4 cells via demethylation of CpG dinucleotide sites in the promoter of CD80 gene resulting in T lymphocyte infiltration into tumors and, ultimately, in tumor rejection [26]. Likewise, epigenetic silencing of CD70, a CD80 related co-stimulatory molecule, by DNA methylation was noted during breast cancer progression in an in vitro model [27].

A limit of the present study is the lack of a direct evidence of DNA methylation of the CD80 promoter. Usually, but not exclusively, methylation occurs at the so-called CpG islands and if hypermethylation involves the promoter region it may result in gene silencing. CD80 promoter do not contain any sequence with characteristics making methylation easy to detect (CD80 at chr3:119524293-119559634 Genome Browser http://​genome.​ucsc.​edu/​index.​html). Given the difficulty in directly assessing whether and at what stage CD80 methylation takes place along the colorectal carcinogenesis pathway and in order to assess if hypermethylation plays a role in CD80 down-regulation in human colorectal tumorigenesis, we examined the correlation between expression of CD80 and the methylation of genes involved in the early stages of CRC carcinogenesis and the expression of DNMTs and attempted to verify our hypothesis by developing an in vitro assay. Finally, this study did not explore the interactions of CD80 and PD-1 with B7H1 (programmed death ligand 1 [PD-L1]) [28]. In fact, although they are both located on antigen presenting cells’ surface, CD80 also has an appreciable affinity for the PD-L1 [29]. Their interplay has a pivotal role in controlling T cell activation, proliferation, anergy, and apoptosis and methylation might play a role in their regulation. However, the interactions between the two pathways remain still unknown.

In conclusion, gene methylation seems to be associated to the CD80 down-regulation observed in dysplasia of sporadic colonic carcinogenesis. In vitro inhibition of DNA methyltransferases significantly enhances CD80 expression in colon cancer cells. We hypothesize that the failure of immune surveillance mechanisms in colonic non-inflammatory carcinogenesis is linked to genomic methylation directly or indirectly affecting CD80 expression.

5AzadC, 5-Aza-2′-deoxycytidine; APC, Adenomatous Polyposis Coli; CDH13, H-cadherin; CRC, colorectal cancer; DNMT, DNA methyltransferase; MGMT, O-6-methylguanine-DNA methyltransferase; MLH1, MutL homolog 1; RUNX3, Runt-related transcription factor 3; UC, ulcerative colitis