|Home | About | Journals | Submit | Contact Us | Français|
Histone deacetylase inhibitors (HDACi) are an emerging class of cancer therapeutics that induces growth arrest and apoptosis in a variety of malignant cells. HDACi also induce the differentiation of multiple transformed cell types, a feature that distinguishes the effects of these agents from those of cytotoxic agents. The HDACi, SAHA, is FDA approved for the treatment of cutaneous T-Cell lymphoma, and several HDACi are presently undergoing clinical trials for the treatment of other hematological malignancies as well as for solid tumors. As a result, there has been considerable interest in understanding the biological function and role in tumorigenesis of the targets of HDACi, the class I HDACs (1, 2, 3 and 8) and class II HDACs (4, 5, 6, 7, 9 and 10). Identification of the HDAC paralogs most critical for tumor promotion could lead to the design of more targeted therapies and reduced toxicities.1
HDACs are enzymes which catalyze the deacetylation of lysine residues on a variety of protein substrates by acting in opposition to histone acetyltransferases (HATs). While DNA bound histone proteins were the initial substrates of HDACs identified, and the basis of the naming of this enzyme family, numerous other cellular proteins, including multiple transcription factors and cytoplasmic proteins are now known to be deacetylated by HDACs. Changes in acetylation status have been linked to altered protein function including the modulation of protein-protein complex formation and DNA binding capability.2
Consistent with the anti-proliferative and pro-differentiation and apoptotic effects of HDACi, several lines of evidence now point to a role for individual HDACs in tumor progression and survival. The case for HDAC3 stems from the observation that its expression is frequently increased in tumors relative to adjacent normal tissue, while HDAC3 downregulation results in reduced proliferation and survival of tumor cells.3,4 The importance of HDAC3 in normal development is illustrated by the early embryonic lethality of HDAC3 deficient mice. Its role in normal cell proliferation and survival is evidenced by the delay in cell cycle progression, cell cycle dependent DNA damage, and increased apoptosis of HDAC3−/− mouse embryonic fibroblasts.5 Similarly, lethality at the larval stage due to cell-autonomous induction of apoptosis is observed in HDAC3 loss of function mutants in Drosophila.6
In this issue of Cancer Biology & Therapy, Godman et al. provide new insight into mechanisms by which HDAC3 facilitates growth and survival of colon cancer cells in vitro.7 By stable transfection of an HDAC3-targeting shRNA, the authors generated colon cancer cell lines with markedly reduced HDAC3 levels. To identify genes and pathways regulated by HDAC3, the authors then performed microarray analyses on control and HDAC3 deficient clones. A large percentage of genes (26.8%) were modulated in expression (>5-fold), with the majority upregulated in expression. Importantly, the authors demonstrate that approximately 50% of these sequences are similarly altered in response to 18 hour HDACi treatment. Using a pathway analysis tool, the authors show that the silencing of HDAC3 results in increased expression of several components of the TGFβ and interferon signaling pathways and in the modulation of components of the Wnt signaling pathway.
Further analysis of the Wnt signaling pathway demonstrated reduced nuclear β-catenin levels in shHDAC3 clones. This was notable as it was observed despite these cells harboring a truncating mutation in the Apc gene, the primary regulator of β-catenin abundance and hence activation and nuclear localization of β-catenin in colon cancer cells. Extrapolating from the microarray data, the authors suggest several mechanisms for the observed reduction in nuclear β-catenin levels. These include increased expression of the TCF co-repressors TLE1 (Transducin-like enhancer of split 1) and TLE4 upon HDAC3 knockdown. TLE1 and TLE4 compete with β-catenin for TCF4 binding, and the authors suggest that HDAC3-mediated repression of these genes in tumor cells could enhance the availability of TCF4 for β-catenin binding. Additional studies however are needed to directly establish this possibility.
Consistent with the reduced nuclear β-catenin levels, downregulation of the pro-proliferative β-catenin-TCF target gene, c-myc, was observed, providing a possible explanation for the reduced cell proliferation observed upon HDAC3 knockdown. Prior studies by the same group as well as others have also linked HDAC3-induced growth promotion with the direct transcriptional repression of the key growth inhibitory genes, p21 and p15(INK4b),3,4,8 suggesting HDAC3 promotes cell growth by simultaneously facilitating expression of pro-proliferative and repressing expression of growth inhibitory genes.
The finding that multiple pathways involved in the regulation of proliferation and differentiation, including Wnt, TGFβ and interferon signaling, are regulated by HDAC3 suggests specific targeting of this enzyme may have significant therapeutic benefits. However, data from several laboratories clearly indicate that the magnitude of growth inhibition and apoptosis induced upon HDAC3 downregulation in tumor cells is relatively modest compared to the effects induced by HDAC inhibitors, which simultaneously inhibit multiple class I and II HDACs.3,4,9 Studies into the mechanism of transcriptional regulation of the p21 gene provide some insight into the basis of this effect, where ChIP experiments have demonstrated the simultaneous recruitment of multiple class I HDACs (HDAC1, 2 and 3) as well as the class II HDAC, HDAC4, to the proximal p21 promoter. Downregulation of any one HDAC results in derepression of p21 expression, however, in each case the magnitude of p21 induction is markedly less that that induced by HDACi.3,8,10 At the p21 promoter, therefore, HDAC3 participates as one of several HDACs involved in p21 repression. Therefore, while specific targeting of HDAC3 may yield some therapeutic benefit, findings to-date suggest the use of pan-HDAC inhibitors is likely to yield a stronger therapeutic response.
A further important observation reported by Godman et al. was the upregulation of the VDR gene following either HDAC3 downregulation or HDACi treatment. VDR expression has been shown to be downregulated in late stage colon cancers,11 and is a gene of particular interest to cancer. Uppon binding to its ligand, vitamin D (1α,25(OH)2D3), VDR can promote growth arrest and differentiation of select breast, prostate and colon cancer cell lines.12 In vivo, epidemiological studies have found an inverse correlation between exposure to sunlight, which produces vitamin D in the skin, with the incidence of several cancers including colorectal cancer,12 while dietary vitamin D supplementation has been shown to have chemopreventive activity in mice.13 The majority of cell lines, however, are insensitive to 1α,25(OH)2D3, limiting its therapeutic potential. Indeed, in the present study the authors fail to observe any growth inhibitory response of the control SW480 cell line to vitamin D-treatment. Impressively, the authors demonstrate that both transient and stable knockdown of HDAC3 induces expression of VDR mRNA and protein, and importantly, confers sensitivity of SW480 cells to vitamin D-induced growth inhibition. Notably, while the present findings link HDAC3 with repression of VDR expression, HDAC3 may also contribute to vitamin D3 resistance through its role as a critical component of the NCoR/SMRT co-repressor complex. Prior studies in prostate cancer cells have demonstrated that in the absence of ligand, VDR interacts with the N-CoR/SMRT/HDAC3 co-repressor complex to mediate transcriptional repression of VDR target genes.12 Indeed, basal levels of SMRT have been linked to resistance to 1α,25(OH)2D3 in prostate cancer cell lines, while knockdown of SMRT results in increased response to vitamin D3 as assessed by induction of the VDR target gene GADD45α.14 Thus the downregulation of HDAC3 may contribute to restoration of the vitamin D response both by restoring the expression of VDR and potentially by the disruption of the SMRT-HDAC3 co-repressor complex. This finding raises the intriguing possibility that the combinatorial treatment of HDACi with vitamin D may provide additive therapeutic benefit.
While the current study identifies a number of novel and interesting transcriptional changes induced upon HDAC3 knockdown, several questions remain regarding the specific mechanisms by which these transcriptional changes are mediated. First, clearer distinction between direct versus indirect HDAC3-regulated genes is important. In this regard, use of the emerging technology of genome-wide chip on chip analysis may provide important insight into the promoters directly regulated in a HDAC3-dependent manner. At least two fundamental mechanisms by which HDAC3 directly regulates gene expression can then be envisioned (Fig. 1). First, as HDAC3 does not directly bind DNA, it must be recruited to promoters by sequence-specific transcription factors where it would mediate transcriptional repression by catalyzing the deacetylation of surrounding core histones. Deacetylated histones are positively charged and therefore have greater affinity for DNA, resulting in a closed chromatin conformation that is less-permissive for transcription. Determining the identity of the sequence-specific transcription factors which recruit HDAC3 to promoter regions would be of particular interest. To-date, ChIP analysis of the HDAC3-regulated p21 and p15 loci indicate HDAC3 is recruited to these promoters by the Sp1/Sp3 family of transcription factors.3,8,10 Again, by revealing the specific promoter regions to which HDAC3 is recruited, ChIP on chip analyses could unveil the identity of the recruiting transcription factors. Second, HDAC3 may alter transcription by directly deacetylating transcription factors themselves as recently demonstrated for MEF2 and the RelA subunit of NFκB.15,2 This can have multiple effects on the biological function of the transcription factor, including alteration of its ability to interact with other proteins or on its DNA binding affinity. Interestingly, the present study identified a number of genes whose expression was decreased following HDAC3 knockdown, suggesting that HDAC3 may also function to facilitate transcription in certain contexts. While the basis for this effect is unknown, deacetylation of a specific transcription factor which results in enhancement of its transcriptional activity can be envisioned as one possible mechanism. Identification of novel non-histone substrates directly deacetylated by HDAC3 and determination of the consequence of this post translational modification is therefore an important area of future investigation.
In summary, the clinical efficacy of HDACi now clearly establishes the HDAC family of enzymes as a bone fide therapeutic target in the treatment of cancer. The present gene expression profiling study by Godman et al. provides new insights into the genes and pathways altered in expression and function by an important member of this family, HDAC3, providing a launching pad for subsequent investigations.