In mature lymphoid malignancies including T-cell lymphoma and multiple myeloma (MM), aberrant acetylation status has been strongly linked to their tumorigenesis. Thus, the modulation of acetylation through targeting histone deacetylases (HDACs) is considered to be a viable therapeutic strategy. Vorinostat (SAHA) is the first HDAC inhibitor approved by the FDA for the patients with cutaneous T-cell lymphoma (CTCL). Anti-tumor activity of SAHA against CTCL, adult T-cell leukemia/lymphoma (ATLL), and MM cell lines was examined. Most of these cells were found to be sensitive to this drug at IC50 of less than 1–2uM. For further clarification of its mechanism of action, we established five SAHA-resistant cell lines consisting of 3 CTCL and 2 MM using dose stepwise increase method over six months. IC50 of the SAHA-resistant cells was 4-to 14-fold higher than that of their parental cells. These cell lines also showed cross-resistance of 2.8- to 17.5-fold against another pan-HDAC inhibitor, panobinostat (LBH589). Regarding HDAC activity, it was greatly inhibited by SAHA in parental cells, whereas it was only partly inhibited in SAHA-resistant cells. That is, SAHA-resistant cells have lost apart of the HDAC inhibiting function caused by SAHA. Moreover, SAHA-resistant cells showed higher anti-apoptosis ability when exposed with SAHA than their parental cells with acetylation status of histone H3 being remained low. Next, we performed microarray analysis to compare expression levels of various HDACs and other related genes between parental and SAHA-resistant cells. Results indicated that the expression level of HDAC3 being obviously low in resistant cells among various HDACs, which was also confirmed by real-time PCR. In line with mRNA analysis, protein level of HDAC3 was also decreased in SAHA-resistant cells compared with their parental cells, while other HDAC expression remained unchanged. We assumed that HDAC3 could be a main target of SAHA. To examine this possibility, we established both HDAC3 knocked-down and over-expressing cell lines, and examined the sensitivity of these cells to SAHA. HDAC3 knocked-down cells showed obviously SAHA-resistant feature in MTS assay, however, HDAC3 over-expressing cells showed higher sensitivity to SAHA. Knocking out other HDACs (1, 2 and 8) in parental cell lines did not change the sensitivity to SAHA. Thus, our results suggest that SAHA induced apoptosis depends on the inhibition of HDAC3. To search for other possible mechanisms, we screened for the mutations in HDAC2, 3, 4 and 8, but did not find them. Finally, we supposed that HDAC3 expression was epigenetically silenced by promoter methylation in SAHA resistant cells, and attempted to restore HDAC3 expression in the presence of 5-azacytidine, a DNA demethylase. We incubated SAHA-resistant cells with non-toxic levels of 5-azacytidine (4uM) for 9 days, and confirmed that HDAC3 expression was restored during 6–9 days after exposure. When HDAC3 expression being restored, the resistant cells showed higher sensitivity to SAHA. It suggests that hyper-methylation of promoter sequences of HDAC3 contributed to the mechanism of SAHA-resistance. From these results, we conclude that anti-tumor effect of HDAC inhibitors depends on the expression level of HDCA3 in mature lymphoid malignancies, and HDAC3 might provide a useful biomarker for identifying favorable response to HDAC inhibitors, and the overcoming the resistance of HDAC inhibitors.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.