Normal acinar cells, normal ductal cells as well as stromal fibroblasts showed occasional weak to moderate positiv ity for all three HDAC isoforms. Pancreatic adenocarci noma displayed strong nuclear immunoreactivity third for HDAC1, HDAC2, HDAC3 and RelA p65 in a considerable number of cases. Some expression was also evident in desmoplastic stroma cells and inflammatory cells. Raw expression scores of HDAC1, HDAC2 and HDAC3 correlated significantly with each other, suggesting a shared regulation of these isoforms. High HDAC2 expression was signifi cantly associated with poor tumor differentiation. No other correlations of HDAC isoforms with clinico pathological parameters were found. Interestingly, we found a positive correla tion between RelA p65 expression and HDAC isoform expression.

Categorized grouped HDAC scores signifi cantly correlated with the presence of nuclear RelA p65. To further elucidate the strength of the relationship, raw expression scores were compared. Here we found that both cytoplasmic as well as nuclear RelA p65 positivity was significantly linked with the expression of specific HDAC isoforms, the strength of those correlations were weak to moderate. When the grouped HDAC scores were correlated with RelA p65 expression, only the relation to nuclear RelA p65 expres sion was significant. The correlation with cyto plasmic RelA p65 expression showed borderline significance. This supports a func tional relationship between HDAC activity and RelA p65 expression and nuclear translocation.

In contrast to nuclear RelA p65 expression expression of HDAC isoforms 1, 2 and 3 did not have prognostic impact in univariate survival analyses. For the conven tional prognostic parameters tumor grade and Dacomitinib nodal sta tus a significant correlation with overall survival could be ascertained in our cohort. Inhibition of RelA p65 activity by treatment with SAHA and VPA Based on the association of class I HDAC expression and nuclear RelA p65 translocation in pancreatic adenocarci noma in vivo and the fact that RelA p65 is a putative target of HDIs we wanted to know if this link could be function ally confirmed for pancreatic cancer in vitro. First, to show that inhibition of nuclear translocation is in fact linked to a decreased activity of RelA p65, we per formed a RelA p65 specific transcription factor assay which measures the binding activity of the protein. Since the amount of activated RelA p65 was comparatively low under cytokine absence, RelA p65 activity was enhanced by stimulation with IL 1.