HMGA1 Is Upregulated in Myelodysplastic Syndromes with Mutation in Pre-mRNA Splicing Genes and Inhibits Leukemia Cell Differentiation

In myelodysplastic syndrome (MDS), genes associated with pre-mRNA splicing such as SRSF2 (Serine and Arginine Rich Splicing Factor 2) and U2AF1 (U2 small nuclear RNA auxiliary factor 1) are frequently mutated. While several reports demonstrated that MDS patients with these mutations had worse prognosis, few treatments specific to these mutations are available, partly due to the dearth of material on functions of these mutated genes.

To assess the role of pre-mRNA splicing gene mutations in the pathogenesis of MDS, we performed RNA-seq analysis using CD34-postitive fractions of bone marrow samples of MDS patients. Patients with mutations in either SRSF2 or U2AF1 tended to have shorter 1-year leukemia free survival compared to patients without these mutations (66.7% vs 73.7%, respectively; p=0.067). Gene set enrichment analysis revealed that MYC target genes were enriched in patients with mutations in either SRSF2 or U2AF1 compared to patients (n=4) without these mutations (n=3). Among them, we found that HMGA1 (High Mobility Group A1) was distinctly upregulated in these patients.

Expression levels of HMGA1 are known to be higher in hematopoietic stem cells. HMGA1 is involved in various cellular processes such as transcriptional regulation, DNA repair, cell differentiation and regulated cell death. We confirmed that expression levels of HMGA1 in murine bone marrow cells were higher in Lineage marker-negative, Sca1-positive, and c-Kit-positive (LSK) fractions than other more differentiated fractions. To determine whether upregulation of HMGA1 contributed to pathogenesis of MDS, we retrovirally transduced HMGA1 to murine hematopoietic progenitor cell line 32D-cl3. We found that proportions of Gr-1 positive cells after treatment with G-CSF were lower in HMGA1-transduced 32D-cl3 cells. Further, expression levels of myeloid-associated genes were also suppressed in these 32D-cl3 cells. These results suggested that 32D-cl3 with extrinsic expression of HMGA1 were more resistant to G-CSF induced myeloid differentiation. Similarly, primary murine hematopoietic progenitor cells retrovirally transduced with HMGA1 showed increased colony-forming capacities. Considering that differentiation block plays a major role in the onset of MDS, these results supported our hypothesis that upregulation of HMGA1 would contribute to pathogenesis of MDS through blocking normal myeloid differentiation.

To investigate whether inhibition of HMGA1 could affect the pathogenesis of myeloid malignancies, we silenced expression of HMGA1 via short-hairpin RNA (shRNA) in various cell lines of acute myeloid leukemia (AML). Human acute monocytic leukemia cell line THP-1 became more easily induced to differentiate upon phorbol12-myristate13-acetate treatment when HMGA1 was silenced. Similarly, we silenced expression of HMGA1 in HL60, which is a human acute promyelocytic leukemia (APL) cell line. We treated HL60 cells with all-trance retinoic acid (ATRA) to induce myeloid differentiation. As expected, we found that expression levels of CEBP-β and Gr-1 in ATRA-treated HL60 cells were significantly higher when HMGA1 was silenced, which meant that HL60 cells became more sensitive to ATRA-induced myeloid differentiation by inhibition of HMGA1. Considering that differentiation induction therapy has improved clinical outcome not only in APL but also in AML, these results suggested that HMGA1 would be a potential therapeutic target for myeloid malignancies.

In summary, our experiments demonstrated that expression levels of HMGA1 were higher in patients with splicing mutations. Considering its inhibitory effects on differentiation demonstrated by our experiments, upregulation of HMGA1 could be involved in the pathogenesis of MDS. Further, therapeutic intervention for HMGA1 had a potential to improve clinical outcome of myeloid malignancies.