Abstract

Histone deacetylases (HDACs) perform key functions in transcriptional regulation by modifying the core histones of the nucleosome as well as non-histone targets. We have previously cloned and characterized HDAC9, a member of the Class IIa HDAC family, which also contains HDACs 4,5 and 7. These are characterized by the presence of a common N-terminal region, which mediates direct interactions with transcription factors such as BCL-6 or MEF2. The HDAC9 gene encodes multiple protein isoforms some of which display distinct cellular localization patterns and biological activities. The transcribed region of HDAC9 is very large, spanning more than 900 kilobases, with approximately 50% of these sequences being non-coding. Consistent with its size as well as the multiplicity and structural complexity of expressed HDAC9 isoforms, we report that HDAC9 expression is under the control of three independent promoter regions, one of which possesses a CpG island. This is in contrast to other HDAC genes thus far identified, each of which appears to possess a single CpG island-containing promoter. Transcripts initiating from the individual HDAC9 promoters are differentially expressed and encode specific HDAC9 isoforms. A comparison of the human and mouse HDAC9 genes reveals some important differences in both regulatory regions and coding sequences. Specifically, the human HDAC9 gene can express from its second promoter, which is not present in the mouse gene, hematopoietic-specific transcripts (i.e. only found in the lymph node, spleen and lymphocytes in normal tissue) that encode isoforms containing previously unidentified N-terminal sequences. In normal B cells, HDAC9 mRNA transcripts are initiated from the second and proximal promoters and stimulation of these cells with IL2, α-CD40 MoAb, Staphylococcus aureus Cowan strain 1, or arsenic trioxide had no appreciable affect on HDAC9 expression. However, due to a potentially abnormal differential promoter usage and alternative splicing, chronic lymphocytic leukemia (CLL) patient B-cells displayed HDAC9 isoform expression pattern that was dramatically distinct from that observed in normal B-cells. Specifically, there was significant expression from the distal (CpG island containing) promoter, which is not utilised in normal B cells, and overexpression from the second and proximal promoters. With regard to alternative splicing, the CLL patient cells lacked exon7, which contains a nuclear localization signal and exon12, which contains a site of sumoylation - a modification linked to deacetylase activity. Moreover, these changes were specific for full-length HDAC9, and not the MITR isoform that lacks catalytic domain. Stimulation of CLL patient B-cells with SAC/IL-2 caused a switch in promoter usage leading to the change in isoform specific expression to a pattern that is identical to that of normal B-cells. These results suggest that deregulation of HDAC9 expression may play a role in B-cell development and the pathogenesis of B-cell neoplasms. Given the need for development of better therapies for indolent B-cell malignancies such as CLL, the finding that HDAC9 expression in cancer cells is modulated by SAC/IL-2 points to a possible combinatorial therapy with HDAC inhibitors, which have been shown to induce apoptosis in CLL cells.