The ZEB2-dependent EMT transcriptional programme drives therapy resistance by activating nucleotide excision repair genes ERCC1 and ERCC4 in colorectal cancer.

Resistance to adjuvant chemotherapy is a major clinical problem in the treatment of colorectal cancer (CRC). The aim of this study was to elucidate the role of an epithelial to mesenchymal transition (EMT)-inducing protein, ZEB2, in chemoresistance of CRC, and to uncover the underlying mechanism. We performed IHC for ZEB2 and association analyses with clinical outcomes on primary CRC and matched CRC liver metastases in compliance with observational biomarker study guidelines. ZEB2 expression in primary tumours was an independent prognostic marker of reduced overall survival and disease-free survival in patients who received adjuvant FOLFOX chemotherapy. ZEB2 expression was retained in 96% of liver metastases. The ZEB2-dependent EMT transcriptional programme activated nucleotide excision repair (NER) pathway largely via upregulation of the ERCC1 gene and other components in NER pathway, leading to enhanced viability of CRC cells upon oxaliplatin treatment. ERCC1-overexpressing CRC cells did not respond to oxaliplatin in vivo, as assessed using a murine orthotopic model in a randomised and blinded preclinical study. Our findings show that ZEB2 is a biomarker of tumour response to chemotherapy and risk of recurrence in CRC patients. We propose that the ZEB2-ERCC1 axis is a key determinant of chemoresistance in CRC.

Alternatively spliced MEFV transcript lacking exon 2 and its protein isoform pyrin-2d implies an epigenetic regulation of the gene in inflammatory cell culture models.

The function of gene body DNA methylation in alternative splicing, and its relation to disease pathogenesis is not fully elucidated. The gene for familial Mediterranean fever (MEFV) encodes the pyrin protein and contains a 998 bp CpG island, covering the second exon, which is differentially methylated in FMF patients compared to healthy controls. Our further observation of increased exon 2-spliced MEFV transcript in leukocytes of FMF patients provoked us to test the role of exon methylation in alternative splicing using inflammatory cell culture models. First, in vitro exon methylation triggered an increased level of exon 2 exclusion using a splicing cassette in a promyelocytic leukemia cell line (HL-60). HL-60 cells subjected to methylating and demethylating agents, as well as cells differentiated to neutrophil-like cells, exhibited different levels of spliced/unspliced transcripts. We observed increased levels of spliced transcripts in neutrophil-like (p = 0.0005), activated (p = 0.0034) and methylated cells (p < 0.0001), whereas decreased levels in demethylated cells (p = 0.0126) compared to control untreated HL-60 cells. We also showed that the protein isoform of pyrin lacking the exon 2 has an adverse subcellular localization in neutrophil-like cells. Therefore, it remains in the cytoplasm rather than the nucleus. This may point to an epigenetic involvement in an important inflammatory gene.

Alternatively spliced MEFV transcript lacking exon 2 and its protein isoform pyrin-2d implies an epigenetic regulation of the gene in inflammatory cell culture models

Abstract The function of gene body DNA methylation in alternative splicing, and its relation to disease pathogenesis is not fully elucidated. The gene for familial Mediterranean fever (MEFV) encodes the pyrin protein and contains a 998 bp CpG island, covering the second exon, which is differentially methylated in FMF patients compared to healthy controls. Our further observation of increased exon 2-spliced MEFV transcript in leukocytes of FMF patients provoked us to test the role of exon methylation in alternative splicing using inflammatory cell culture models. First, in vitro exon methylation triggered an increased level of exon 2 exclusion using a splicing cassette in a promyelocytic leukemia cell line (HL-60). HL-60 cells subjected to methylating and demethylating agents, as well as cells differentiated to neutrophil-like cells, exhibited different levels of spliced/unspliced transcripts. We observed increased levels of spliced transcripts in neutrophil-like (p = 0.0005), activated (p = 0.0034) and methylated cells (p < 0.0001), whereas decreased levels in demethylated cells (p = 0.0126) compared to control untreated HL-60 cells. We also showed that the protein isoform of pyrin lacking the exon 2 has an adverse subcellular localization in neutrophil-like cells. Therefore, it remains in the cytoplasm rather than the nucleus. This may point to an epigenetic involvement in an important inflammatory gene.