Daniel J Weisenberger1. 1. University of Southern California, USC Norris Comprehensive Cancer Center, Department of Biochemistry and Molecular Medicine, 1450 Biggy St., NRT G511D, Los Angeles, CA 90033, USA. Electronic address: dan.weisenberger@med.usc.edu.
The current era of genomic medicine has greatly increased our understanding of the inter-connectedness of genetic, epigenetic, transcriptomic and proteomic profiles of human health and disease, none more evident than in humancancer. While humancancer genomics has primarily focused on genetic mutations and copy number alterations, epigenetic changes, namely DNA methylation and chromatin modifications are highly prevalent in every cancer type, occur more frequently than genetic alterations in humancancers (Schuebel et al., 2007) and can result in gene expression modulation. Humancancers are characterized by global DNA hypomethylation of repetitive elements and CpG-poor regions together with gene-specific, promoter CpG island DNA hypermethylation. DNA methylation changes may result in tumor suppressor gene silencing and the activation of oncogenes. This aspect of cancer epigenetics is compelling, as unlike somatic mutations and copy number alterations, cancer-specific DNA methylation changes can be reversed by treatment with DNA methylation inhibitors, such as 5-aza-cytidine (5-Aza-CR) and 5-aza-2′-deoxycytidine (5-Aza-CdR), which have been used in treatment of humancancers (reviewed in Raynal and Issa (2016)).DNA methylation-based subgroups of humancancers have been identified that show clinical utility. Evidence of a CpG Island Methylator Phenotype (CIMP) was first described in colorectal cancer as DNA hypermethylation of a subset of CpG islands in a subset of tumors (reviewed in Weisenberger (2014)). CIMP status is also identified in glioblastoma, gastric cancer, endometrial and breast cancers (reviewed in Weisenberger (2014)). ColorectalCIMP was further shown to strongly correlate with a point mutation in the BRAF gene, older patient age and female gender (Weisenberger et al., 2006). Overall, CIMPs generally do not present with a consistent genomic profile, suggesting that CIMPs can result from multiple and unique sets of aberrations (Weisenberger, 2014).In the accompanying report, Brennan and colleagues (Brennan et al., 2017-in this issue) describe a CIMP for head and neck squamous cell carcinomas (HNSCCs) after meta analyses of TCGA DNA methylation and gene expression datasets (The Cancer Genome Atlas Network, 2015). There are five HNSCC subgroups based on humanpapillomavirus infection (HPV +), CIMP status and smoking history: 1) NSD1-smoking, 2) Stem-like smoking, 3) CIMP-Atypical, 4) Non-CIMP-Atypical and 5) HPV +.The HNSCC CIMP subgroup, similar to HPV + HNSCC tumors and CIMPs of other tumor types, displays only minor copy number alterations. CIMPtumors did show moderate enrichment of CASP8 mutations, and to a lesser extent, HRAS mutations, as compared to the other HNSCC subgroups, also described by TCGA (The Cancer Genome Atlas Network, 2015), suggesting that the NF-kB and EGFR signaling pathways, respectively, may be targeted by specific therapeutic strategies that inhibit these pathways. However, further research will be required to investigate specific signaling pathways altered in CIMP HNSCC patients.The CIMP-Atypical subgroup is molecularly distinct and has translational and clinical importance. The ‘Atypical’ nomenclature reflects that this subtype is overrepresented for patients that do not have history of smoking, alcohol and other risk factors that are associated with HNSCC patients in the non-Atypical subgroups. Atypical CIMPpatients are also of older age, much like colorectalCIMPpatients, however, patient outcome differences between CIMP and the other HNSCC subgroups have not been determined. An analysis of gene expression modulation in CIMP-Atypical tumors shows strong enrichment of immuno-responsive genes activated by Type I and Type II interferons (IFNs). Moreover, the CIMPtumors also showed increased levels of CD8 + T cells, another marker for immune response. These data indicate that chronic inflammation may be a driving force for these patients.The immune-based signature of CIMP-Atypical tumors is also important with respect to precision medicine based approaches for HNSCC treatments. In general, HNSCC patients have poor outcome and diminished anti-tumor immune response (Economopoulou et al., 2016). The activation of IFN-related genes in patients with CIMP-Atypical tumors suggests that these patients may benefit from anti-inflammatory and immune-based therapeutics, however, patient survival and response to therapy of CIMP-Atypical patients compared to other HNSCC subgroups remain to be determined.The DNA hypermethylation status of the CIMP-Atypical subgroup is intriguing, since these patients may respond to epigenetic therapies. Recent evidence of treating cancer cells with 5-Aza-CdR resulted in activation of tumor suppressor genes, reduction in oncogene expression, and activation of endogenous retroviral elements (ERVs) that stimulate an immune response (Chiappinelli et al., 2015, Roulois et al., 2015). Moreover, the addition of vitamin C, a required co-factor for TET enzyme based DNA demethylation provides a synergistic effect of 5-Aza-CdR treatment (Liu et al., 2016). Most cancerpatients are vitamin C deficient, and therefore, combining DNA methylation inhibitors with vitamin C may prove efficacious for CIMP-positive HNSCC patients. Pre-clinical studies and clinical trials of these drug combinations are necessary to determine clinical utility, however, epigenetic therapies may provide promise for a subset of HNSCC patients.Taken together, these findings provide strong evidence for a unique subset of HNSCC patients defined by DNA methylation profiles. Further research is required to determine the connection between CIMP and CASP8 mutations, however, the correlations between individual CIMPs and specific somatic gene mutations across several cancer types are striking.
Disclosure
Daniel J. Weisenberger is a consultant for Zymo Research Corporation (Irvine, CA USA), a company that distributes DNA methylation-related products. Zymo Research does not have an interest in Dr. Weisenberger's research or the writing of this manuscript.
Authors: David Roulois; Helen Loo Yau; Rajat Singhania; Yadong Wang; Arnavaz Danesh; Shu Yi Shen; Han Han; Gangning Liang; Peter A Jones; Trevor J Pugh; Catherine O'Brien; Daniel D De Carvalho Journal: Cell Date: 2015-08-27 Impact factor: 41.582
Authors: Katherine B Chiappinelli; Pamela L Strissel; Alexis Desrichard; Huili Li; Christine Henke; Benjamin Akman; Alexander Hein; Neal S Rote; Leslie M Cope; Alexandra Snyder; Vladimir Makarov; Sadna Budhu; Sadna Buhu; Dennis J Slamon; Jedd D Wolchok; Drew M Pardoll; Matthias W Beckmann; Cynthia A Zahnow; Taha Merghoub; Taha Mergoub; Timothy A Chan; Stephen B Baylin; Reiner Strick Journal: Cell Date: 2015-08-27 Impact factor: 41.582
Authors: Daniel J Weisenberger; Kimberly D Siegmund; Mihaela Campan; Joanne Young; Tiffany I Long; Mark A Faasse; Gyeong Hoon Kang; Martin Widschwendter; Deborah Weener; Daniel Buchanan; Hoey Koh; Lisa Simms; Melissa Barker; Barbara Leggett; Joan Levine; Myungjin Kim; Amy J French; Stephen N Thibodeau; Jeremy Jass; Robert Haile; Peter W Laird Journal: Nat Genet Date: 2006-06-25 Impact factor: 38.330
Authors: Minmin Liu; Hitoshi Ohtani; Wanding Zhou; Andreas Due Ørskov; Jessica Charlet; Yang W Zhang; Hui Shen; Stephen B Baylin; Gangning Liang; Kirsten Grønbæk; Peter A Jones Journal: Proc Natl Acad Sci U S A Date: 2016-08-29 Impact factor: 11.205
Authors: Kornel E Schuebel; Wei Chen; Leslie Cope; Sabine C Glöckner; Hiromu Suzuki; Joo-Mi Yi; Timothy A Chan; Leander Van Neste; Wim Van Criekinge; Sandra van den Bosch; Manon van Engeland; Angela H Ting; Kamwing Jair; Wayne Yu; Minoru Toyota; Kohzoh Imai; Nita Ahuja; James G Herman; Stephen B Baylin Journal: PLoS Genet Date: 2007-07-31 Impact factor: 5.917