Abstract
Background
The t(5;11)(q35;p15.5) translocation resulting in fusion of the nucleoporin NUP98 and methyltransferase NSD1 (NUP98-NSD1) genes is a recurrent aberration observed in pediatric and adult AML. The NUP98-NSD1 fusion often co-occurs with the FLT3-ITD mutation and characterizes a group of cytogenetically normal AML patients with very poor prognosis. Despite advances in the understanding of the biology of NUP98-NSD1-positive AML, its therapeutic success rate has remained low. We aimed to identify novel candidate drugs for NUP98-NSD1-positive AML by testing primary patient cells and in vitro cell models with a high-throughput drug sensitivity platform.
Methods
Leukemic blasts were Ficoll separated from bone marrow (BM) aspirates of an AML patient positive for t(5;11)(q35;p15.5) and FLT3-ITD. RNA extracted from primary cells was used for RNA sequencing and gene expression analysis. NUP98-NSD1 cDNA was amplified from primary cell RNA and expressed from a lentiviral vector (LeGO-iCer2) also encoding the cerulean fluorescent marker. The NUP98-NSD1/LeGo-iCer2 and empty LeGo-iCer2 viruses were used to establish stably expressing Ba/F3 cell lines. Primary murine (BALB/c) BM cells were transduced with NUP98-NSD1 and FLT3-ITD retroviruses alone or in combination (NNF) in vitro (“preleukemic”) or passaged in vivo (“leukemic”) as previously described (Thanasopoulou et al, 2014). For screening, 309 small molecule inhibitors including FDA/EMA-approved and investigational oncology drugs were plated on 384-well plates in a 10,000-fold concentration range. Cells were dispensed on the pre-drugged plates and incubated at 37°C for 72h, and then cell viability measured using the CellTiter-Glo® luminescent assay. Drug response curves were generated and a drug sensitivity score determined (Yadav et al, 2014). Select drug sensitivity was calculated for each drug by comparing results between primary leukemic and healthy donor BM cells or between the cell constructs and empty vector transduced controls cells.
Results
Primary patient cells and murine BM cells expressing FLT3-ITD alone or in combination with NUP98-NSD1 were selectively sensitive to specific FLT3 inhibitors (e.g. quizartinib, sorafenib and lestaurtinib), and broad-spectrum receptor tyrosine kinase inhibitors targeting FLT3-ITD (e.g. cabozantinib, crenolanib, foretinib, midostaurin, MGCD-265 and ponatinib). Furthermore, these cells were highly sensitive to checkpoint kinase 1/2- inhibitor AZD7762. The primary murine cells expressing both NUP98-NSD1 and FLT3-ITD showed higher sensitivity to all of the above-mentioned drugs compared to cells expressing either of the events alone indicating functional synergy. A very distinct drug response pattern was observed in the leukemic NNF cells cultured in vivo compared to the same cells cultured in vitro suggesting that microenvironment may also affect the observed drug responses. Interestingly, the preleukemic murine cells expressing NUP98-NSD1 with or without FLT3-ITD as well as the primary patient cells showed extreme vulnerability to BCL2/BCL-xL inhibitor navitoclax. Furthermore, primary murine cells expressing NUP98-NSD1 alone showed high select sensitivity to JAK-inhibitors ruxolitinib, BMS-911543, AZD1480 and tofacitinib indicating the fusion may stimulate JAK/STAT-signaling. Similar sensitivity was also observed in the Ba/F3-cells expressing NUP98-NSD1. In support of these findings, gene expression analyses showed high expression of anti-apoptotic factors BCL2, BCL-xL and MCL1 in the patient cells. MCL1 is regulated by STAT3 while BCL-xL is regulated by STAT5, which were also highly expressed.
Conclusions
In summary, we have observed an enhanced response to specific and non-specific FLT3 inhibitors in cells expressing NUP98-NSD1 and FLT3-ITD together compared to cells expressing either of the two alone. This coincides with previous findings that functional co-operation between NUP98-NSD1 and FLT3-ITD is important in AML (Thanasopoulou et al, 2014). We have seen high in-vitro-in-vivo correlation between primary patient cells and murine cells expressing NUP98-NSD1 and FLT3-ITD. Moreover, we have identified potential candidate compounds targeting oncogenic signaling activated by these two events. These data form a basis for clinical evaluation of candidate compounds for NUP98-NSD1-positive AML.
Disclosures
Porkka: Bristol-Myers Squibb: Honoraria, Research Funding; Novartis: Honoraria, Research Funding. Heckman:Celgene: Research Funding.
MOLI: Multi-Omics Late Integration with deep neural networks for drug response prediction