Page 86 - 2022_01-Haematologica-web

A. Damnernsawad et al.
inhibitors, even though the responses are transient.21 Diverse mutations underlie resistance in AML patients to the type I inhibitor, crenolanib, including rare mutations at the gatekeeper region of FLT3, as well as NRAS and IDH2 mutations in FLT3-independent subclones, and TET2 and IDH1 mutations in FLT3 mutant clones in non- responding patients.22 Single-cell DNA sequencing of specimens from patients relapsing after treatment with a novel type I inhibitor, gilteritinib, revealed shifts in clonal architecture to select for secondary mutations in NRAS, KRAS, IDH2, or BCR-ABL1, either in the context of FLT3- ITD or FLT3 wild-type clones.18 Aberrant activation of ERK either extrinsically through the bone marrow microenvironment or intrinsically in a cell-autonomous manner has been implicated in FLT3 resistance in AML.23,24 Upregulation of the RAS/RAF/ERK pathway has been observed after treatment with FLT3 tyrosine kinase inhibitors in AML cell lines and AML patients’ bone mar- row samples.23,25 Signaling through JAK/STAT5 mediated by granulocyte-macrophage colony-stimulating factor and interleukin-3 allows AML cells to survive FLT3 inhibitor treatment.26 Activation of the phosphatidylinos- itol-3 kinase (PI3K)/mammalian target of rapamycin (MTOR) pathway has also been demonstrated to pro- mote resistance to a FLT3 inhibitor.27
Sorafenib, a multi-kinase inhibitor targeting not only FLT3 but also RAF, VEGFR, FGFR, KIT and RET,28 has been evaluated in combination with azacytidine in AML patients with FLT3-ITD, who had an overall response rate of 46%.29 The combination of sorafenib and standard-of- care chemotherapy extended event-free survival in patients younger than 60 years old.30 Data from a phase I trial showed that patients harboring FLT3-ITD who were treated with allogeneic hematopoietic stem cell trans- plantation had a 1-year progression-free survival rate of 85% and a 1-year overall survival rate of 95%.31
To identify mechanisms of resistance to sorafenib we used a genome-wide CRISPR (clusters of regularly inter- spaced short palindromic repeats) knockout screen to search for genes whose loss-of-function variants can pro- mote FLT3 inhibitor-sensitive AML cells to survive in the presence of sorafenib. To confirm that aberrant signaling in the identified pathways renders cells insensitive to FLT3 inhibitors, we established AML cells resistant to both type I and type II FLT3 inhibitors. Our CRISPR screen identified genes in the MTOR and mitogen-acti- vated protein kinase (MAPK) pathways that modulate sensitivity to sorafenib. Activities of MTOR and MAPK pathways were upregulated in cells with acquired resist- ance, and these cells were sensitive to MEK inhibitors supporting the role of aberrant downstream MAPK sig- naling in resistance to FLT3 inhibitors. We found the combination of FLT3 and MEK inhibitors had synergistic efficacy in both FLT3 inhibitor-sensitive and -resistant AML cells as well as in samples from AML patients. In summary, our work identified several negative regulators of MTOR and MAPK signaling pathways, LZTR1, TSC1/2, NPRL2, NF1, not previously associated with AML, as modulators of sensitivity to sorafenib. We show that aberrations in MTOR and MAPK pathways are important mechanisms of resistance to sorafenib as well as other FLT3 inhibitors in AML and suggest that the combination of FLT3 inhibitors and MEK inhibitors could be useful for the treatment of FLT3 inhibitor-resis- tant AML.
Methods
Cell lines
Human MOLM13 cells were obtained from the Sanger Institute Cancer Cell Line Panel. All cell lines used in this study were authenticated at the OHSU DNA Services Core facility. Cell lines were maintained in 20% fetal bovine serum, RPMI medium, sup- plemented with glutamine, penicillin/streptomycin and an anti- mycotic. All cell lines were tested for Mycoplasma on a monthly schedule.
Lentivirus production and transduction
HEK293T cells were transfected using Lipofectamine-2000 (Invitrogen) with single transfer vectors in combination with pack- aging plasmids, psPax2 (Addgene, #12260) and VSVG (Invitrogen). Supernatants were collected, filtered through 0.45 mM filters and used for transduction as described previously.32
The CRISPR/Cas9 library screen and CRISPR/Cas9 gene inactivation by individual sgRNA
Cas9-expressing cells were generated using Cas9Blst (Addgene, #52962). Loss-of-function screens were performed using pooled human genome-wide single-guide (sg)RNA libraries, the Y. Kosuke library,33 purchased from Addgene (#67989), as described previously,32 which targets 18,010 genes with 90,709 sgRNA (aver- age of 5 guides per gene). High-titer lentivirus was generated using standard calcium phosphate precipitation procedures in HEK293T cells. Viral supernatant was concentrated and the titer determined using a viral titration kit (ABM good, Canada). One hundred million cells were used for viral transduction at a multi- plicity of infection (MOI) of 0.3, selected with puromycin for 5-7 days to ensure stable viral integration. Individual genes were inac- tivated by cloning sgRNA into plentiCRISPRV2 (Addgene, #52961) according to the manufacturer’s suggestions. The follow- ing sgRNA were used in the study: LZTR1: 5’ CCCATAGAC- GACGGCCGAG 3’, NF1: 5’CATATCAGTCTGTGGGATC 3’, TSC1, 5’ACGTCGTTGTCCTCACAAC 3’, TSC2: 5’ TTGAT- GCGCACGGCGCCTC 3’, NPRL2: 5’ GAACCCATCAATG- TAGGGC 3’, DEPDC 5’ GACTGTGACTCAAGTGTTCC and 5’ TGTTAATGTCGTAGACCCTA, TBC1D7 5’ GTATCGTASAG- GAGCAGTACT. Sequencing data were deposited to GEO with, accession number GSE138343.
Drug sensitivity assay
Small-molecule inhibitors, purchased from LC Laboratories Inc. (Woburn, MA, USA) and Selleck Chemicals (Houston, TX, USA), were reconstituted in dimethylsulfoxide (DMSO). Cells were seeded at 1,000 cells/well in a 384-well plate in 50 mL medium (RPMI-1640 supplemented with fetal bovine serum [15%], L-glut- amine, penicillin-streptomycin and an antimycotic) with different concentrations of drugs and cultured for 72 h. For the drug sensi- tivity assay, 5 uL of MTS reagent (CellTiter96 AQueous One; Promega Madison, WI, USA) were added to each well and incubated for 4 h. Optical density was measured at 490 nm. Relative cell viability was calculated by normalizing the readings to those of untreated control wells. Prism software (GraphPad) was used to produce non-linear fitting and determine the response to the drug, the half maximal inhibitory concentration (IC50) and the area under the curve (AUC).
Immunoblot analysis
Whole cell protein lysates were prepared using cell lysis buffer (Cell Signaling Technologies), 1 mM phenylmethylsulfonyl fluo- ride, proteasome (Roche) and a phosphatase inhibitor cocktail (Sigma-Aldrich). Proteins were resolved on 4 -15% gradient gels
78
haematologica | 2022; 107(1)