ObjectiveEpidemiological and clinical data indicate that patients suffering from IBD with long-standing colitis display a higher risk to develop colorectal high-grade dysplasia. Whereas carcinoma invasion and metastasis rely on basement membrane (BM) disruption, experimental evidence is lacking regarding the potential contribution of epithelial cell/BM anchorage on inflammation onset and subsequent neoplastic transformation of inflammatory lesions. Herein, we analyse the role of the α6β4 integrin receptor found in hemidesmosomes that attach intestinal epithelial cells (IECs) to the laminin-containing BM.DesignWe developed new mouse models inducing IEC-specific ablation of α6 integrin either during development (α6ΔIEC) or in adults (α6ΔIEC-TAM).ResultsStrikingly, all α6ΔIEC mutant mice spontaneously developed long-standing colitis, which degenerated overtime into infiltrating adenocarcinoma. The sequence of events leading to disease onset entails hemidesmosome disruption, BM detachment, IL-18 overproduction by IECs, hyperplasia and enhanced intestinal permeability. Likewise, IEC-specific ablation of α6 integrin induced in adult mice (α6ΔIEC-TAM) resulted in fully penetrant colitis and tumour progression. Whereas broad-spectrum antibiotic treatment lowered tissue pathology and IL-1β secretion from infiltrating myeloid cells, it failed to reduce Th1 and Th17 response. Interestingly, while the initial intestinal inflammation occurred independently of the adaptive immune system, tumourigenesis required B and T lymphocyte activation.ConclusionsWe provide for the first time evidence that loss of IECs/BM interactions triggered by hemidesmosome disruption initiates the development of inflammatory lesions that progress into high-grade dysplasia and carcinoma. Colorectal neoplasia in our mouse models resemble that seen in patients with IBD, making them highly attractive for discovering more efficient therapies.

To ensure genome stability, pericentromeric regions are compacted in a dense heterochromatic structure through a combination of specific 'epigenetic' factors and modifications. A cascadal pathway is responsible for establishing pericentromeric chromatin involving chromatin modifiers and 'readers', such as H3K9 histone methyltransferases (Suv)39h and heterochromatin protein 1. Here we define how H3K64me3 on the lateral surface of the histone octamer integrates within the heterochromatinization cascade. Our data suggest that enrichment of H3K64me3 at pericentromeric chromatin foci is dependent on H3K9me3 but independent of a number of central factors such as heterochromatin protein 1, DNA methyltransferases and Suv4-20h histone methyltransferases. Our results support a model in which pericentromeric heterochromatin foci are formed along distinct pathways upon H3K9 trimethylation, involving H3K64me3 to potentially stabilize DNA-histone interactions, as well as sequential recruitment of repressive histone tail and DNA modifications. We hence suggest that multiple mechanisms ensure heterochromatin integrity at pericentromeres, with H3K64me3 as an important factor.