Diagnostic implications of loss of 5-hydroxymethylcytosine for melanoma

Keywords::

• 5-hydroxymethylcytosine
• 5-methylcytosine
• DNA methylation
• epigenetic
• isocitrate dehydrogenases
• melanoma
• murine–human melanoma xenograft model
• ten–eleven translocation family of DNA hydroxylases

Despite significant advances in our understanding of the genetic mechanisms of melanoma tumorigenesis over the past decade, melanoma remains among the few human cancers with steadily rising incidence worldwide. Tumor thickness, as determined by the evaluation of hematoxylin and eosin stained sections, was introduced by Breslow in 1970 and remains the single most powerful independent prognostic indicator for primary cutaneous disease, for which surgery remains the mainstay of therapy [1]. Once melanoma has metastasized, the prognosis is dismal with a median survival of 7 months and a 5-year survival rate of 15% [2]. In this regard, metastatic melanoma is one of the most therapeutically challenging malignancies, with significant chemotherapy and targeted therapy resistance [3].

Presently, there is a dearth of molecular markers that facilitate detecting differences between benign and malignant melanocytic lesions and assisting in prediction of their biological behavior. Moreover, melanoma is a type of aggressive cancer with numerous somatic mutations that vary among individual tumors. Recent evidence indicates that epigenetic events also play critical roles in melanoma tumorigenesis. Thus, there is a pressing need for novel biomarkers focused on both genetic and epigenetic defects that will better define the malignant potential of primary lesions, predict clinical outcome and forecast therapeutic responses.

Regulatory epigenetic events include the effects of DNA methylation at the carbon-5 position of cytosine on gene expression, post-translational modifications of core histones during chromatin packaging and remodeling, and noncoding RNA regulation on gene expression. Such events are fundamental to the regulation of normal cellular processes as well as to many diseases. Among all epigenetic events, DNA methylation is the most extensively studied epigenetic modification in cancer. 5-methylcytosine constitutes 2–8% of the total cytosines in human genomic DNA and impacts a broad range of biological functions and pathological processes, including gene expression, maintenance of genome stability, genomic imprinting, X-chromosome inactivation, development regulation, aging-related processes and cancer. Recently, 5-hydroxymethylcytosine (5-hmC), an oxidized 5-methylcytosine by TET family/isocitrate dehydrogenases, was discovered in mammalian cells and found to exist at high levels in self-renewing and pluripotent stem cells [4–6]. 5-hmC may be an intermediate in a pathway of active DNA demethylation either by conversion to cytosine under certain conditions or by replacement of 5-hmC by specific DNA repair mechanisms.

Global hypomethylation within the bulk genome and local hypermethylation at specific tumor suppressor genes have been shown to be involved in melanoma [7–9]. But the degree of global hypomethylation is not sufficient to distinguish benign nevi from melanoma [10]. Recent studies indicate that multilocus DNA-methylation signature genes may be better discriminators between melanomas from nevi, but their detection requires sophisticated molecular biological tools [11,12]. In addition, small biopsy sizes of melanocytic lesions and the need for complete histologic assessment of excised tumors present additional technical limitations. Recent evidence, however, demonstrates that 5-hmC has potential to be the first epigenetic mark that can be used immunohistochemically for diagnosis and for evaluation of melanoma virulence. Specifically, 5-hmC was profoundly reduced in over 400 melanoma cases compared with nevi [13]. In addition, we have evaluated 5-hmC distribution in a study of the full spectrum of melanocytic lesions: from benign dermal nevi; mildly, moderately and severely dysplastic nevi; so-called ‘borderline’ melanocytic neoplasms; and melanoma. The data show loss of 5-hmC as one progresses from benign nevi to low and high-grade dysplastic nevi, to borderline lesions and to melanoma [Data not shown]. This is consistent with the potential value of ‘loss of 5-hmC’ as a barometer of melanoma tumor progression.

Studies of biomarkers predictive of clinical outcome are impeded by latent periods for the detection of metastases that may range from several years to more than a decade due to the slow progression of melanoma in some patients. Accordingly, 2–3-year follow-up may not be sufficient to exclude eventual evolution to more advanced stages. Thus, clinically-annotated biospecimen archives serve as valuable surrogates for otherwise impractical prospective approaches. When the association between 5-hmC levels and survival probability was analyzed in a large cohort clinically annotated for long-term outcome, Kaplan–Meier curves revealed that patients with 5-hmC-positive melanomas had significantly higher survival probabilities than patients with 5-hmC-negative melanomas at diagnosis [13]. Thus, the preliminary data at hand suggest that the loss of 5-hmC in melanoma may have prognostic, as well as diagnostic value.

The ability to more definitively evaluate ‘loss of 5-hmC’ in relationship to melanoma virulence will require further testing in animal models. The first SCID model for growth and metastasis of xenografts from fresh human melanoma tissue was described in 1991 [14], and it was adapted to the more relevant model of melanoma growth intradermally in human skin xenografts in 1993 [15]. Using a modification of this established xenograft bioassay, melanoma behavior has recently been correlated with clinical outcome, further supporting the predictive validity of this in vivo model system [16]. Care must be taken, however, in extrapolating melanoma behavior in immunocompetent humans from profoundly immunocompromised NSG mice potentially incapable of differentiating among more virulent tumorigenic melanoma subpopulations [17]. Moreover, harsh cell preparation techniques that potentially skew results by modifying molecular phenotypes [18] must also be rigorously taken into account [19]. Given these admonitions, the xenograft model continues to hold promise as a preclinical approach to defining the diagnostic and prognostic utility of biomarkers like ‘loss of 5-hmC’. Indeed, in our xenograft studies to date, we find an inverse relationship between metastatic melanoma cells expressing certain virulence-conferring biomarkers and 5-hmC, raising the possibility that a multiplex approach may be more sensitive for profiling melanomas in terms of revealing their molecular/epigenetic phenotypes.

Amazingly, more than 80% of the human sequences highly conserved across evolution do not encode exons in the genome [20]. DNA methylations play critical roles in normal cellular biology and pathology of diseases by regulating gene expression. Epigenetic studies in cancer biology were initiated 20 years ago by the discovery of altered DNA methylation in certain tumors. Despite recent advances in our understanding of epigenetic mechanisms in melanoma, the most pressing questions remain unanswered. For example, no DNA demethylase has yet been identified. Furthermore, although advances have been made in our understanding of genes that specify developmental pathways, comparatively little is known as to how epigenetic events maintain or reprogram differentiation. Both epigenetic aberrations and genetic mutations, either individually or most likely in concert, can result in loss of control over cell growth and development, with progression of cancer. Thus, another key question relates to identification of the pathways that connect genetics and epigenetics in cancer.

The causes of cancer are more complicated than a binary model involving presence or absence of specific genetic mutations. Epigenetic studies have gained momentum in recent years due to two important characteristics. First, epigenetic DNA methylation changes in cancer appear to be considerably more frequent and earlier events than genetic mutations. More importantly, given the reversible nature of epigenetic regulation, as opposed to genetic mutations, such epigenetic studies could lead to new strategies for cancer therapy. Moreover, epigenetic approaches may provide clues that link environmental factors with cancer-promoting genetic alterations. The ability to understand and control reversible epigenetic changes before irreversible mutations ensue may play an increasingly important role in further cancer prevention and treatment. Moreover, new insights into the epigenetic landscape that will permit the use of more patient-specific therapeutic agents could provide important inroads into personalized medicine. For now, rigorous examination of the biomarker utility of epigenetic events, like ‘loss of 5-hmC’ in human melanoma, may provide novel adjuncts to diagnosis and prognosis in this most lethal of human cancers.

Acknowledgements

We extend our sincere apology to those colleagues whose studies were not cited in this review due to space constraints.

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.