Faculty Bibliography

Components of the SWI/SNF chromatin remodeling complex, including BRG1 (also SMARCA4), are inactivated in cancer. Among other functions, SWI/SNF orchestrates the response to retinoid acid (RA) and glucocorticoids (GC) involving downregulation of MYC. The epigenetic drugs SAHA and azacytidine, as well as RA and GC, are currently being used to treat some malignancies but their therapeutic potential in lung cancer is not well established. Here we aimed to determine the possible therapeutic effects of azacytidine and SAHA (A/S) alone or in combination with GC plus RA (GC/RA) in lung cancers with either BRG1 inactivation or MYC amplification. In vitro, responses to GC/RA treatment were more effective in MYC-amplified cells. These effects were mediated by BRG1 and involved a reprogramming towards prodifferentiation gene expression signatures and downregulation of MYC. In MYC-amplified cells, administration of GC/RA enhanced the cell growth inhibitory effects of A/S which, in turn, accentuated the prodifferentiation features promoted by GC/RA. Finally, these treatments improved overall survival of mice orthotopically implanted with MYC-amplified, but not BRG1-mutant, cells and reduced tumor cell viability and proliferation. We propose that the combination of epigenetic treatments with retinoids and corticoids of MYC-driven lung tumors constitute a strategy for therapeutic intervention in this otherwise incurable disease.

The characterization of the epigenetic changes within the obesity-related adipose tissue will provide new insights to understand this metabolic disorder, but adipose tissue is not easy to sample in population-based studies. We aimed to evaluate the capacity of circulating leukocytes to reflect the adipose tissue-specific DNA methylation status of obesity susceptibility. DNA samples isolated from subcutaneous adipose tissue and circulating leukocytes were hybridized in the Infinium HumanMethylation 450 BeadChip. Data were compared between samples from obese (n = 45) and non-obese (n = 8-10) patients by Wilcoxon-rank test, unadjusted for cell type distributions. A global hypomethylation of the differentially methylated CpG sites (DMCpGs) was observed in the obese subcutaneous adipose tissue and leukocytes. The overlap analysis yielded a number of genes mapped by the common DMCpGs that were identified to reflect the obesity state in the leukocytes. Specifically, the methylation levels of FGFRL1, NCAPH2, PNKD and SMAD3 exhibited excellent and statistically significant efficiencies in the discrimination of obesity from non-obesity status (AUC > 0.80; p < 0.05) and a great correlation between both tissues. Therefore, the current study provided new and valuable DNA methylation biomarkers of obesity-related adipose tissue pathogenesis through peripheral blood analysis, an easily accessible and minimally invasive biological material instead of adipose tissue.

Epigenetics of aging is an emerging field that promises exciting revelations in the near future. Epigenetic pathways, including DNA methylation and histone modification, are determinants of normal development and can change during aging. Some of the epigenetic alterations described during aging, as hypermethylation at specific promoters and decrease of global DNA methylation, are also associated with tumor development. The epigenetic changes occurring during development and aging can be stochastic or depend on environmental factors. Future challenges in the field involve the determination of the precise molecular mechanisms that create age-dependent epigenetic variation and how these epigenetic changes affect the aging phenotype.

Elucidating the potential mechanisms involved in the detrimental effect of excess body weight on insulin action is an important priority in counteracting obesity-associated diseases. The present study aimed to disentangle the epigenetic basis of insulin resistance by performing a genome-wide epigenetic analysis in visceral adipose tissue (VAT) from morbidly obese patients depending on the insulin sensitivity evaluated by the clamp technique. The global human methylome screening performed in VAT from 7 insulin-resistant (IR) and 5 insulin-sensitive (IS) morbidly obese patients (discovery cohort) analyzed using the Infinium HumanMethylation450 BeadChip array identified 982 CpG sites able to perfectly separate the IR and IS samples. The identified sites represented 538 unique genes, 10% of which were diabetes-associated genes. The current work identified novel IR-related genes epigenetically regulated in VAT, such as COL9A1, COL11A2, CD44, MUC4, ADAM2, IGF2BP1, GATA4, TET1, ZNF714, ADCY9, TBX5, and HDACM. The gene with the largest methylation fold-change and mapped by 5 differentially methylated CpG sites located in island/shore and promoter region was ZNF714. This gene presented lower methylation levels in IR than in IS patients in association with increased transcription levels, as further reflected in a validation cohort (n = 24; 11 IR and 13 IS). This study reveals, for the first time, a potential epigenetic regulation involved in the dysregulation of VAT that could predispose patients to insulin resistance and future type 2 diabetes in morbid obesity, providing a potential therapeutic target and biomarkers for counteracting this process.

The introduction of new therapies against particular genetic mutations in non-small-cell lung cancer is a promising avenue for improving patient survival, but the target population is small. There is a need to discover new potential actionable genetic lesions, to which end, non-conventional cancer pathways, such as RNA editing, are worth exploring. Herein we show that the adenosine-to-inosine editing enzyme ADAR1 undergoes gene amplification in non-small cancer cell lines and primary tumors in association with higher levels of the corresponding mRNA and protein. From a growth and invasion standpoint, the depletion of ADAR1 expression in amplified cells reduces their tumorigenic potential in cell culture and mouse models, whereas its overexpression has the opposite effects. From a functional perspective, ADAR1 overexpression enhances the editing frequencies of target transcripts such as NEIL1 and miR-381. In the clinical setting, patients with early-stage lung cancer, but harboring ADAR1 gene amplification, have poor outcomes. Overall, our results indicate a role for ADAR1 as a lung cancer oncogene undergoing gene amplification-associated activation that affects downstream RNA editing patterns and patient prognosis.

Close to half of de novo acute myeloid leukemia (AML) cases do not exhibit any cytogenetic aberrations. In this regard, distortion of the DNA methylation setting and the presence of mutations in epigenetic modifier genes can also be molecular drivers of the disease. In recent years, somatic missense mutations of the DNA methyltransferase 3A (DNMT3A) have been reported in ~20% of AML patients; however, no obvious critical downstream gene has been identified that could explain the role of DNMT3A in the natural history of AML. Herein, using whole-genome bisulfite sequencing and DNA methylation microarrays, we have identified a key gene undergoing promoter hypomethylation-associated transcriptional reactivation in DNMT3 mutant patients, the leukemogenic HOX cofactor MEIS1. Our results indicate that, in the absence of mixed lineage leukemia fusions, an alternative pathway for engaging an oncogenic MEIS1-dependent transcriptional program can be mediated by DNMT3A mutations.