Leukemia is the most common form of childhood cancer and cancer is the leading cause of death from disease of American children. Hence, improving leukemia therapy is of utmost importance in pediatric health. This is particularly relevant to AML in which progress has lagged significantly in comparison to childhood acute lymphoblastic leukemia. Resistance to cytarabine-based chemotherapy is a major cause of treatment failure in this disease 
. Therefore, new therapies for children with AML must be developed.
HDACIs represent a promising new class of anti-cancer agents and can induce apoptosis in leukemia cells but not normal cells 
. In our previous study, we demonstrated that VPA, an anti-epileptic agent in both children and adults and a potent HDACI, synergistically enhanced cytarabine sensitivities in both pediatric AML cell lines and diagnostic blasts, suggesting that HDACs are promising therapeutic targets for pediatric AML 
. However, individual HDAC family members that are involved in synergistic cytarabine response in the disease have not been identified.
This study was designed to begin to address this important question and to select the optimal HDACIs which show the greatest enhancement on cytarabine sensitivities in pediatric AML cells. Such information is mechanistically important and has significant clinical implications, as well. To begin to identify which HDAC isoforms are involved in cytarabine sensitivity, we examined the expression profiles of class I, II, and IV HDACs in 4 pediatric AML cell lines. Our results suggested that HDACs 5 and 11 are unlikely involved in cytarabine sensitivities due to the lack or marginal expression of these enzymes. Using THP-1 cells which express high levels of both classes I and II HDACs, we then used equal doses of three different HDACIs (MS-275, VPA, and SAHA) with different substrate specificities to further narrow down the HDAC isoforms likely to be involved in augmenting cytarabine sensitivity. Results from these studies suggested that HDAC8 is unlikely to be involved in cytarabine-induced apoptosis in THP-1 cells since none of the HDACI treatments resulted in significant enzyme inhibition, although they all enhanced cytarabine-induced apoptosis.
Results from our shRNA knockdown studies unequivocally demonstrated that inhibition of HDACs 1 and 6 was pivotal for sensitizing pediatric AML cells to cytarabine. This could, at least partly, be mediated by Bim, a BH3-only pro-apoptotic protein. Bim was classified as an “activator” in view of its purported ability to act directly and to activate Bax and Bak 
. It has been well documented that Bim is critical for HDACI-induced apoptosis of both solid tumor and leukemia cells 
. Our previous study strongly suggested that Bim is also critical for cytarabine-induced apoptosis in pediatric AML cells 
. Surprisingly, down-regulation of HDAC2 resulted in significantly decreased apoptosis induced by cytarabine, even though it was previously reported that down-regulation of HDAC2 is critical for inducing apoptosis in cancer cells 
. In contrast, down-regulation of HDACs 3 and 4 had no effects on cytarabine-induced apoptosis in THP-1 cells. Together, our results strongly implicate both HDACs 1 and 6 as the most relevant therapeutic targets for treating pediatric AML with HDACIs and cytarabine. Studies are underway to determine the detailed molecular mechanisms responsible for the effects of HDACs 1, 2, and 6 on cytarabine sensitivities in the disease.
It has been a long-standing debate as to whether isoform specific- or pan-HDACIs result in better anti-cancer activities 
. The perception is that isoform-specific HDACIs may offer clear therapeutic advantages over non-specific classical HDACIs. Specifically, the premise is that the greater specificity will involve the modulation of a smaller number of disease-focused genes with a reduced toxicity profile 
. However, recent microarray studies suggested that the pleiotropic antiproliferative and apoptotic effects of the broad-spectrum HDACIs may be more beneficial than an isoform-specific drug 
. Our results from shRNA knockdown studies strongly favor the latter opinion, at least in pediatric AML, since both HDACs 1 and 6 appear to be critical factors in determining cytarabine sensitivities in the disease. This was further supported by our in vitro
treatments of pediatric AML cells with both class I selective- and pan-HDACIs. At clinically achievable concentrations, only the drugs (LBH-589 and PXD101) which simultaneously inhibited both HDACs 1 and 6 showed the best antileukemic activities and significantly enhanced cytarabine-induced apoptosis. Again, our mechanistic studies suggest that induction of DNA damage and Bim is critical for the activities of LBH-589 and PXD101 and their combinations with cytarabine.
Altogether, our results not only confirmed that HDACs are promising therapeutic targets for pediatric AML, but also identified HDACs 1 and 6 as the most relevant drug targets. Accordingly, treating pediatric patients with pan-HDACIs may be more beneficial than HDAC isoform-specific drugs. Our study provides compelling rationale for the combination of cytarabine and HDACIs in pediatric AML clinical trials. It also provides a strong molecular basis for selecting the optimal HDACIs to combine with cytarabine. Since many biological features of AML are shared by adults and children, our results should also apply to the treatment of adult AML patients, as well. It is extremely important to note that we used Cmax concentrations for the HDACIs to combine with cytarabine to prove the concept. However, Cmax or the maximally tolerated doses of these HDACIs may not be the optimal doses for combination therapy with cytarabine. Detailed preclinical studies will be needed to establish the optimal scheduling and dosing for the combinational therapies involving cytarabine plus LBH-589 or PXD101 for treating pediatric AML.