Lipid nanoparticle-based delivery of gene therapy targeting the RUNX1::RUNX1T1 fusion oncogene in acute myeloid leukemia

Abstract Acute myeloid leukemia (AML) with RUNX1::RUNX1T1 was one of the first types of AML to be defined by a translocation leading to the generation of a fusion oncogene. Constituting a gain-of-function aberration, RUNX1::RUNX1T1 blocks hematopoietic differentiation while promoting myeloid progenitor cell proliferation. Despite an upfront favorable prognosis, an estimated 30% of patients relapse following intensive chemotherapy, with a survival probability of 35% for this population. Previously, we have demonstrated that a CRISPR-based dual-gRNA knock-out design can disrupt RUNX1::RUNX1T1 and induce an anti-leukemic effect suggesting its therapeutic potential, while pending a suitable vector for delivering the CRISPR components to the leukemic cells in vivo. In this study, we investigated the emerging non-viral delivery method, lipid nanoparticles (LNP), as a putative vector for in vivodelivery of CRISPR-based gene therapy for treatment of AML with RUNX1::RUNX1T1. We used human AML cell line Kasumi-1 as model for RUNX1::RUNX1T1 driven AML, and the THP-1 AML cell line (negative for RUNX1::RUNX1T1) as control. Fresh and unstimulated primary bone marrow- (BM) and peripheral blood (PB) cells from healthy patients (n=10) and newly diagnosed MDS/AML patients (n=5) were used for ex vivo LNP transfection evaluation. We encapsulated SpCas9 mRNA and gRNAs into LNPs, comprising four types of lipids with molar ratios of 50:10:39:1 (ionizable lipid:helper lipid:sterol:PEG). We also compared the efficiency of two types of Ionizable lipids: D-Lin-MC3-DMA (LNPMC3)and SM-102 (LNPSM-102). Particles were loaded with cargo: (i) eGFPmRNA; (ii) Cas9mRNA; or (iii) dual-gRNAs targeting RUNX1::RUNX1T1. By multiparameter flow cytometry, we detected the uptake in hematopoietic stem- and progenitor cells (HSPCs), monocytes, granulocytes, B-cells, and T-cells by quantification of GFP-expression in each subset. Synthesized LNPs exhibited high RNA-encapsulation, with no difference in particle size, zeta potential nor polydispersity depending on the cargo. Following confirmation of LNP capacity for transfecting AML cell lines, LNP-mediated delivery of dual-gRNAs and Cas9mRNA targeting RUNX1::RUNX1T1 resulted in consistent fusion oncogene knock-out, with an estimated 3.5-fold increase (95% CI: 2.3 – 4.9) in edited RUNX1::RUNX1T1 per 24 hours. Furthermore, functional assays demonstrated inhibited proliferation rate (p<.009), clonogenic potential (p=.001), and decreased viability (p<.001) in knock-out Kasumi-1 cells as compared with vehicle treatment, while no difference was observed in the control cell line. In addition, using RNA sequencing-based transcriptomics and mass spectrometry-based proteomics, we showed that in vitro knock-out of the RUNX1::RUNX1T1 fusion oncogene results in changes in gene- and protein expressions correlated to key cellular processes such as proliferation and myeloid differentiation, and importantly, when evaluating this in the control, no such correlations were observed. Ex vivo transfection of healthy primary patient BM cells revealed a myeloid bias of LNP delivery efficiency (p<.001). The highest GFP-expression was detected in HSPCs (median: 62%; 95% CI: 49% – 77%), monocytes (median: 56%; 95% CI: 40% – 86%), and T-cells (median: 29%; 95% CI: 19% – 37%). Across hematopoietic subsets, LNPSM-102 performed superior as compared to LNPMC3. Primary patient leukemic blasts (n=4) displayed a median GFP expression of 45% (95% CI: 44% – 49%) and 69% (95% CI: 40% – 92%) in the blasts of the BM and PB, respectively. Morphological variation of the LNP construct, with incorporation of DOPE and b-sitosterol, increased the ex vivo transfection efficiency to 97% in the leukemic blast subset in primary cells from one newly diagnosed patient. Collectively, we provide a detailed characterization of LNP capacity to transfect primary patient hematopoietic cells and demonstrate LNPs as a putative vector for delivery of our dual-intron targeted CRISPR-based gene therapy for AML with RUNX1::RUNX1T1.