RUNX1/AML1 mutations have been frequently detected in patients with myeloid neoplasms, especially myelodysplastic syndromes (MDS) and chronic monocytic leukemia (CMML). Although the mutations have been analyzed thoroughly, its expression level has not been investigated. Therefore, we attempt to clarify the expression of RUNX1 in the pathogenesis of myeloid neoplasms. The study was approved by the institutional review board and patients gave written informed consent for the study, according to the Declaration of Helsinki.
Several isoforms of RUNX1 mRNA are known and we analyzed RUNX1a (including exon 7a which has stop codon) and RUNX1b (skipping exon 7a and including exon 7b and 8). Expression levels of full length isoform (RUNX1b) and short isoform (RUNX1a which has a dominant negative effect on RUNX1b) in CD34+ cells from patients with myeloid neoplasms were examined. A part of patients with MDS or myelodysplastic syndrome / myeloproliferative neoplasms (MDS/MPN) including CMML showed RUNX1a overexpression. Average of relative RUNX1a expression level in MDS patients (n=34) and MDS/MPN patients (n=20) was 7.4-fold and 8.6-fold of the level in normal bone marrow (BM), respectively, whereas most of these patients showed almost same or slight increase of expression level of RUNX1b compared with normal BM. Interestingly, some patients showed high expression of RUNX1a and repression of RUNX1b. In both disease categories, patients with excess blasts displayed a significantly higher expression level of RUNX1a compared with normal BM and patients without excess blasts. During the disease progression in a single patient with MDS or MDS/MPN, the expression of RUNX1a became higher, while azacitidine treatment reduced RUNX1a expression. Genomic mutations of RUNX1 were also examined. RUNX1 mutations were detected in 16% of MDS and 35% of MDS/MPN. Surprisingly, a part of patients had both RUNX1 gene mutation and RUNX1a overexpression, and they showed rapid progression of disease.
To evaluate the effects of RUNX1a overexpression, RUNX1a was transduced into CD34+ cells from MDS patients with low expression level of RUNX1a. RUNX1a-transduction resulted in cell proliferation on MS5 stromal cells. These results indicate that overexpression of RUNX1a may add growth advantage to CD34+ cells in patients with MDS or MDS/MPN.
We next analyzed the mechanism of RUNX1a overexpression. Gene mutations affecting exon recognition were examined in the patients. Splicing factor mutations, SRSF2 and U2AF1, were detected frequently in MDS (15%) and MDS/MPN (50%). Patients with splicing factor mutations showed higher RUNX1a expression than patients without the mutations. To confirm that the splicing factor mutations affect the expression of RUNX1a, we performed enforced expression of SRSF2 p.P95H mutant using pMYs.IRES.EGFP retrovirus vector in a MDS-derived cell line, TF-1. After a single cell sorting, independent 13 expanding clones were analyzed. Most of the clones demonstrated higher expression of RUNX1a than mock cells, whereas RUNX1b expression was reduced in all clones. Increase of RUNX1a expression in SRSF2 mutant-transduced TF-1 cells was also confirmed by Western blot. Moreover, the clones with higher GFP intensity showed higher expression level of RUNX1a, suggesting that SRSF2 p.P95H expression level may affect the expression level of RUNX1a. Furthermore, SRSF2 mutant-transduced TF-1 cells showed phenotypic changes of higher CD11b and CD14 than mock TF-1 cells, suggesting that SRSF2 mutant may induce monocytic differentiation via RUNX1a overexpression.
Gene mutations of RUNX1 in intron 6 and exon 7a were also analyzed. A 5' splice site change just after exon 6 was detected in a CMML patient with RUNX1a overexpression, which may be another mechanism of RUNX1a overexpression. Mutations of exon 7a or changes in 3' splice site just before exon 7a have not been detected yet.
In conclusion, our data suggest that overexpression of RUNX1a may play a critical role in the progression of MDS and MDS/MPN, in addition to RUNX1 mutations. Splicing factor mutations are suspected to contribute to the mechanism of the dysregulation of RUNX1.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.