Abstract
The treatment outcomes for patients diagnosed with acute myeloid leukemia (AML) are still dismal. Recent advances in understanding AML indicate that the lack of efficacy is primarily due to non-specificity of currently used chemotherapeutics targeting both leukemic stem/progenitor cells (LSC) and normal hematopoietic stem cells (HSC). Thus, a critical barrier is the identification of innovative therapies that selectively target LSC. Histone deacetylase 8 (HDAC8) has been shown to enhance p53 protein deacetylation, which results in inactivation of p53, promoting LSC survival. We hypothesize that enzymatic/non-enzymatic role of HDAC8 is critical for LSC survival but not for HSCs. Then, we characterized our two tetrahydroisoquinoline (TIQ)-based selective HDAC8 inhibitors (HDAC8i) BIP and OCH3 for growth inhibition, apoptosis, activation of caspase 3, integrity of mitochondrial membrane potential (MMP), and acetylation of histone H4 in human leukemia cell lines. The growth inhibitory effects observed in cell lines were validated using bone marrow (BM) or peripheral blood (PB) cells from AML patients. Colony forming cell (CFC) assays were performed using AML BM/PB cells treated with OCH3 or BIP. OCH3 and BIP were also tested for hematotoxicity using normal CB CD34+ cells. Furthermore, we compared class I HDAC isoform engagement in human normal cord blood (CB) CD34+ cells and in SET-2 leukemia cells using our novel photoreactive probe TH1143. In CD34+ cells, TH1143 had higher level of engagement for HDAC1 and 2, whereas engagement of HDAC3 and 8 was minimal. In SET-2 cells, HDAC3 and HDAC8 displayed relatively higher engagement with TH1143 indicating HDAC engagement is likely cell type specific.
The biological efficacies of OCH3 at 50uM and BIP at 25uM were noted to exert >50% growth inhibition in KG1 and in K562 leukemia cells. Both OCH3 and BIP significantly increased the number of apoptotic cells and there was an enhanced active caspase-3 activity. Furthermore, OCH3 and BIP treated cells displayed lower red/green ratio in comparison to control, indicative of poor MMP and depolarization to induce apoptosis (Table 1.a). OCH3 and BIP were further validated by using BM/PB cells from AML patients showing growth inhibition. This was also accompanied by increase in apoptotic cells by OCH3 and BIP. In contrast to BIP, OCH3 spared CB CD34+ cells as demonstrated by notably lower growth inhibition, apoptotic cells vs control when compared with primary AML cells from patients. Both OCH3 and BIP displayed minimal inhibition of CFU growth in CD34+ cells. However, HDAC8i induced significant CFU growth inhibition in primary AML samples suggesting that HDAC8i spares normal CFU progenitors but not leukemia progenitors (Table 1.b). Notably, both BIP and OCH3 lack ability to exert acetylation of histone H4, unlike broad spectrum HDAC inhibitor TSA (MFI with OCH3=0.96±0.03, BIP=0.77±0, TSA =1.63±0.15) which is consistent with isoform selectivity of OCH3 and BIP.
The leukemia growth inhibitory effects at LSC level was demonstrated using ex vivo OCH3 treated AML patient derived BM/PB cells transplantation in humanized immunodeficient NSGS mice. After 10 to 12 weeks of transplantation mice receiving untreated AML cells had 7.73±2.18% while with OCH3 treatment mice had 4.84±1.37% human CD34+ leukemia cells, a 38% reduction in CD34+ leukemia cells, despite only a single ex vivo exposure to OCH3. Furthermore, in a second model, NSGS humanized mice were transplanted (IV) with primary leukemia cells from AML patients and after 4 weeks injected (IP) with OCH3 or vehicle control. After 12 weeks of transplantation in this second model human primary AML cell burden was 5.74±1.31% (OCH3) and 18.13±12.76% (vehicle control), while mice transplanted with normal CD34+ cells treated similarly with OCH3 or vehicle control displayed no detectable inhibition of human myeloid cell chimerism (OCH3:12.28 ± 3.31% vs vehicle control: 17.92±11.96%).
Taken together, our data indicate that HDAC8 isoform inhibitor, OCH3 displayed significant inhibition of primary AML patient derived leukemia cells growth in vitro and in vivo in contrast to normal CD34+ cells. Selective inhibition of HDAC8 is sufficient to cause growth inhibition in primary AML progenitors including LSCs in vivo while sparing normal HSCs thus offer opportunities for further development of HDAC8i as new experimental therapeutics in AML.
No relevant conflicts of interest to declare.
