IntroductionClassically, cancer therapy approaches lead to tumor cell death or an alternative fate such as senescence. Senescent cells remain viable but without proliferation markers. Senescent cells secrete factors with diverse actions that influence surrounding cells, the extracellular matrix (ECM) and the immune system, a phenomenon termed the senescence‐associated secretory phenotype (SASP). Cyclin-dependent kinase 4/6 inhibitors (CDKi) are game-changers in the therapy of metastatic hormone receptor-positive, HER2-negative breast cancer. A potential mechanism of the action of these agents is induction of senescence in breast tumor cells, going beyond cell cycle arrest. However, senescence can be studied mainly using invasive biopsies, and therefore the prevalence and importance of senescence in patients are largely unknown. Moreover, only few circulating biomarkers exist to predict activity or measure CDKi effects. In order to develop a correlative liquid biopsy for CDKi activity, we have analyzed epigenetic (methylation) changes following therapy-induced senescence in vitro. We sought to identify (un)methylated loci that will be affected by different approaches and later can be discovered in circulating cell-free DNA.

MethodsWe have treated luminal breast cancer MCF-7 cells with Doxorubicin, CDKi (Palbociclib) or by irradiation. This protocol resulted in up to 80% of cells with senescence-related beta-galactosidase activity. DNA methylation was profiled using Illumina Infinium MethylationEPIC 850K BeadChip. Differentially methylated loci (mDNA) were identified using GenomeStudio and Minfi. Analyses were done using GREAT, CSGene, and Reactome databases. Methylation age was analyzed using the Horvath Methylation Calculator. Statistical significance was defined as p<0.05, q<0.1.

ResultsSignificantly differentially methylated sites (in comparison to untreated cells) were revealed as following: 9111 sites in Doxorubicin treated cells, 3828 sites in Palbociclib treated cells, and 694 sites in irradiated cells. These loci comprise 1%, 0.45%, and 0.08% of analyzed methylation sites, respectively. 324 loci were similarly changed following the three treatment options ('common sites'). We found that the 'geographic' distribution of intragenic and intergenic methylation sites (5' UTR, gene body etc.) in all treated cells was similar. Gene set over-representation analysis revealed that sites associated with genes of the 'collagen metabolic process' set were significantly altered in drug-treated cells. Pathway investigation of the 324 'common sites' revealed that ECM-related 'focal adhesion assembly' is the most significant pathway involved. Specific analysis of senescence-related gene sets showed that 15.1% (76/503) of senescence genes changed after Doxorubicin treatment, 6.2% (31/503) after Palbociclib and 1% (5/503) after irradiation. Pathway analysis of these genes showed that drugs affected the 'oxidative stress-induced senescence' pathway, while irradiated cells had SASP-related genes affected.Despite the above mentioned methylation changes, age calculation based on methylation clock showed that all samples had similar age (6), regardless of manipulation.

ConclusionSingle treatment of MCF-7 cells with known senescence inducers results in changes in methylation patterns. A significant number of common loci changed following all types of treatments, suggesting them as potential surrogate loci of senescence. Also, we have reproduced the well-known interplay between collagen and senescence/SASP.This preliminary data sheds light on epigenetic changes following treatment-induced senescence. Further studies are needed to validate whether these methylation changes can be found in vivo in tumors and in patients’ cell-free DNA following therapy with CDKi.

Citation Format: Albert Grinshpun, Anatoli Kustanovich, Joshua Moss, Yael Gabai, Yuval Dor, Ittai Ben-Porath. Senescence-related methylation changes following therapy as potential biomarker for CDK4/6 inhibitor activity [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS5-43.