The role of G0S2 in FLT3+ AML bone marrow microenvironment and metabolism

Abstract Introduction: Acute myeloid leukemia (AML), which is characterized by an abnormal growth of myeloid cells, is estimated to be the second most prevalent form of leukemia this year in the United States, accounting for 33% of cases, and the deadliest, responsible for 47% of leukemia mortality. Its 5-year relative survival rate from 2015 to 2021 is, unfortunately, 32.9%, based on data from the National Institutes of Health (NIH). Although standard chemotherapy treatments are advancing, many patients still face relapses caused by common mutations, particularly in the FMS-like tyrosine kinase-3 (FLT3) gene, which is detectable in about 30% of all AML cases. Like other treatments, FLT3 tyrosine kinase inhibitors (TKIs) such as Midostaurin are used to target mutational changes and improve outcomes. However, as many as 60% of patients have resistance or considerable side effects during this therapy. Our lab previously found that the loss of G0S2 (G0/G1 switch gene 2) promotes disease progression and drug resistance in chronic myeloid leukemia (CML, Gonzalez MA, et al. Clin Transl Med. 2022 Dec; PMID: 36536477). We also observed that G0S2 expression was reduced in AML and even further downregulated in AML patients with FLT3 mutations, showing a similar phenotype compared with CML (Gonzalez-Henry MA, et al., unpublished). For this reason, we hypothesized that the G0S2 phenotype will be influenced by the bone marrow microenvironment and will impact cell metabolism as well as drug resistance in FLT3+ AML. Methods: To test this hypothesis, we used MOLM-14, a human AML cell line with an internal tandem duplication (IDT) mutation in the FLT3 gene, as our TKI-sensitive model. Then we used these paternal cells and cultured them long-term with Midostaurin at 100 nM, to generate our TKI-resistant model. G0S2 RNA and protein levels in both cell lines were measured using RT-PCR and immunoblotting. Metabolic parameters such as oxygen consumption rates (OCR) were analyzed with the Agilent Seahorse XFp Bioanalyzer Mito Stress Kit. Additionally, we sent samples from both cell lines to Parse Bioscience for single-cell RNA sequencing (scRNAseq). Furthermore, we used HS-5, a stromal fibroblast cell line, to mimic the bone marrow microenvironment by co-culturing it with MOLM-14 cells that are either Midostaurin-sensitive or -resistant, compared to regular media and HS-5 conditioned media. Both TKI-sensitive and resistant AML cell lines were transduced with small hairpin RNAs (shRNAs) targeting G0S2 (shG0S2) for knockdown or to induce ectopic G0S2 expression (pLVX G0S2), compared with a non-targeting shRNA control (shNT). All lentiviral vectors were puromycin-sensitive for selecting infected cells and were doxycycline-inducible. We used RT-PCR, immunoblotting, and colony formation assays to check the phenotypes of the transduced cell lines. Results: G0S2 mRNA and protein expression were almost completely reduced in the MOLM-14 TKI-resistant cells compared to their parental MOLM-14 TKI-sensitive cells. Regarding metabolic parameters, the MOLM-14 midostaurin-resistant cell line showed higher OCR levels than its TKI-sensitive control. In the Parse Bioscience scRNAseq, we identified three main genes of interest that were upregulated in the MOLM-14 midostaurin-resistant cells: AIG1 (Androgen Induced Gene 1), involved in lipid metabolism; SATB2 (Special AT-rich sequence Binding protein 2), which plays a role in energy metabolism; and FOXP2 (Forkhead Box Protein P2), a transcription factor implicated in cell metabolism. For the HS5-AML co-culture, we observed protection of the AML cells in both the HS5 conditioned media and the co-culture cells. Additionally, we noted more colony formations in shG0S2 compared to their control, and colony ablation with the ectopic G0S2 vector, after confirming knockdown and increased gene expression at both RNA and protein levels. Conclusions: Midostaurin-resistant cells demonstrated significant changes in metabolism and G0S2 levels, which show that the lower the G0S2 level, the higher the oxygen consumption rates. Additionally, we emphasized the importance of the bone marrow microenvironment, as it can protect cancer cells from TKI treatments. We are currently conducting similar co-culture experiments using G0S2 knockdown and ectopic expression vectors. We are planning in vivo experiments to test our in vitro results from the bone marrow microenvironment.