Transcription Activities of RUNX1 Mutants Predict sAML Transformation in Patients with High Risk De Novo Myelodysplastic Syndrome

RUNX1 is an essential transcription factor for normal hematopoiesis, and also the most frequent mutation target in leukemia. RUNX1 mutations have been described in myelodysplastic syndrome (MDS), and related to the clinicohematologic features. However, the biological activities of the RUNX1 mutants were not thoroughly investigated or even correlated to the progression of the disease. Therefore, we analyzed RUNX1 gene mutations in de novo MDS and studied the biological functions of the mutants. We sought to quantify the biological activities of the RUNX1 mutants and correlate the activities with the disease outcome.

Bone marrow samples from 134 patients with high risk de novo MDS (54 RAEB1, 52 RAEB2, 24 RCMD and 4 others) were analyzed for RUNX1 mutations. Mutational analysis of RUNX1 was performed by RT-PCR amplification of whole coding region of RUNX1 followed by sequencing of the PCR products. Each mutation was reconfirmed by PCR with alternative primers and genomic DNA as template. Expression plasmids of wild-type RUNX1 and the mutants were constructed and applied for functional study. Immunopreciptation followed by Western blotting and electromobility shift assay (EMSA) were performed to examine the association with co-factor and DNA of the mutants. Transactivities of the mutants were quantified by luciferase reporter assay and statistically correlated to the outcome of the disease.

Among the 134 patients, 23 (17.2%) harbored RUNX1 mutations at MDS stage. In total, 21 RUNX1 mutations were identified in 23 patients, including 3 silent mutations (I87I, S195S and L445L), 5 nonsense mutations (R139X, Q245X, M283X, R293X and S383X), 9 missense mutations (S73F, H78Y, R135K, R139G, R139Q, A160T, R174Q, V425G and M439L) and 4 frameshift mutations (Q186fsX209, T219fsX233, Ser329fsX572 and D424fsX572). The mutant, R174Q, lost DNA binding ability but interacted with co-factor, CBFβ, exhibited very low transactivities and served as dominant-negative inhibitors for wt-RUNX1. R139 was demonstrated an important site in RUNX1 transactivity; therefore, R139X, R139G and R139Q mutants exhibited low activities in our assay system. H78Y near the DNA binding site R80 of RUNX1 hindered DNA binding and showed low transactivities. The mutants of Q186fsX209, T219fsX233, Q245X, M283X and R293X lacked tansactivation domain (TAD) exhibited less than 50% transactivities compared to wt-RUNX1. A160T and S73F mutants did not interact with CBFβ but showed comparable transactivity to wt-RUNX1. The C-terminal frameshift mutant, D424fsX572, and the 4 mutants, R135K, S383X, V425G and M439L, neither affected DNA and CBFβ binding nor TAD, displayed similar transactivities to wt-RUNX1. S329fsX572 mutant lost part of TAD, also showed similar transactivities to wt-RUNX1. Seventeen of 23 (73.9%) RUNX1 mutation-positive patients progressed to sAML compared with 66 of 111 (59.5%) RUNX1 mutation-negative patients (P=0.242), indicating mutation of RUNX1 may not be a crucial factor in disease progression. RUNX1 mutants were subsequently divided to high activity (> 50% transactivity compared to wt-RUNX1) and low activity (<50% transactivity) groups based on the results of luciferase reporter assay. Patients with the lower RUNX1 transactivities showed a higher risk (P=0.05) and a shorter time to sAML transformation (median 17.1 vs.10.1 months, P=0.03) compared to patients with higher activities. Our finding suggested that the transactivities of the RUNX1 mutants, rather than its mutation status were more important and precise to predict the sAML evolution in patients with MDS.

Our study determined the biological activities of the RUNX1 mutants and showed that RUNX1 mutants with lower transactivities conferred a higher risk and more rapid progression to sAML in high risk MDS patients.

Supported by grants NHRI-EX99-9711SI, DOH100-TD-C-111-006, BMRPG380031 and MMH-E-100-09.

No relevant conflicts of interest to declare.