A. H. Nguyen1, S. Patel1, M. Vogelauer2, P. A. Toste1, N. Wu1, J. Williams3, L. Li1, D. W. Dawson4, S. Kurdistani2, T. R. Donahue1 1University Of California – Los Angeles,Department Of Surgery,Los Angeles, CA, USA 2University Of California – Los Angeles,Department Of Biological Chemistry,Los Angeles, CA, USA 3Harbor-UCLA Medical Center,Department Of Surgery,Torrance, CA, USA 4University Of California – Los Angeles,Department Of Pathology And Laboratory Medicine,Los Angeles, CA, USA
Introduction:
Histone deacetylase inhibitors (HDACi) are currently being investigated in early phase clinical trials for patients with pancreatic ductal adenocarcinoma (PDAC). Although there have been measurable responses in patients with hematologic malignancies treated with HDACi, similar results have not been demonstrated for solid organ tumors. We hypothesize that tumor associated fibroblasts (TAFs), the predominant cell type in the PDAC stoma, may contribute HDACi therapy resistance.
Methods:
Primary TAFs were isolated from human PDAC tumor samples. MTT assay was used to determine cell viability. Western blot was performed of whole cell lysate and acid extracted histone-enriched samples. Gene expression was determined by qRT-PCR. A modified Boyden chamber was utilized to assess tumor cell (TC) invasion. ChIP-seq was performed for HDAC2 and was overlaid with publically available H3K4me1 and H3K27ac ChIP data from the NIH Roadmap Epigonics Mapping Consortium website.
Results:
As has been described, PDAC TC (PANC-1, MIA PaCa-2) viability was significantly decreased (p<0.01) by all tested class I and II HDACi's (vorinostat, entinostat, panobinostat). However, slowly dividing primary TAFs showed neither growth arrest nor an upregulation of CDKN1A, despite an appropriate increase in global levels of acetylated histones H3 and H4. HDACi treatment of TAFs induced a tumor supportive secretory milieu, as conditioned media from treated cells increased TC invasion (p<0.001) and viability (p=0.05). HDACi treatment on primary TAFs increased secretion of a series of pro-inflammatory proteins (p=0.05) including CXCL1, IL-6, IL-8, and SPP1, by upregulating NFκB target genes (p<0.01). By western blot, we identified markers suggesting the cytosolic activation of NFκB, STAT3, and p38 MAPK pathways, yet, inhibition of STAT3 and p38 MAPK failed to completely abrogate the HDACi-induced inflammatory response and NFκB inhibition proved to be lethal to HDACi-treated fibroblasts. To elucidate how HDAC inhibition may directly affect gene regulation, pro-inflammatory gene enhancers were identified from published H3K4me1 ChIP-seq on normal fibroblasts. We overlaid our HDAC2 ChIP-seq from naïve primary PDAC TAFs to these poised enhancer regions. Our ChIP-seq shows HDAC2 binds these enhancer regions, and importantly, in cells that normally express the active enhancer mark H3K27ac. The promoter regions for these genes both show motif and ChIP-seq data supporting expression from STAT3 and NFκB mediated factors, suggesting HDAC inhibition provides a permissive chromatin landscape for the expression of inflammatory genes in primary PDAC TAFs.
Conclusion:
HDACi's are effective against PDAC TCs in culture, but induce a tumor supportive phenotype in primary PDAC TAFs, which may explain their disappointing results in solid organ tumors. These studies are beginning to uncover the mechanism of this detrimental response in PDAC TAFs