Detection of High-Frequency and Novel DNMT3A Mutations in Acute Myeloid Leukemia by High Resolution Melting Curve Analysis
Rajesh R Singh, Ashish Bains, Keyur P Patel, Hamed Rahimi, Bedia A Barkoh, Abhaya Paladugu, Tigist Bisrat, Farhad Ravandi-Kashani, Jorge E Cortes, Hagop M Kantarjian, Jeffrey L Medeiros, Rajyalakshmi Luthra. University of Texas MD Anderson Cancer Center, Houston, TX
Background: Acute myeloid leukemia (AML) is characterized by extensive deregulation of gene expression as a result of genetic abnormalities, including gene mutations, chromosomal translocations and aberrant epigenetic modification. Recently, mutations in DNMT3A, which encodes a DNA methyltransferase have been identified as markers of poor overall and event-free survival in de novo AML. DNMT3A has 23 exons making routine screening for mutations using standard Sanger sequencing both expensive and labor-intensive. We designed and validated an alternative, high-throughput screening method of using high resolution melting (HRM) curve analysis which identifies sequence variants by detecting subtle changes in mutant DNA melting properties in comparison with wild-type sequence.
Design: Using 108 AML DNA samples, twelve exons of DNMT3A (exons 8 to 10, 15 to 23), which encode the three functional domains (PWWP, ADD and Methyltrasferase domains) were tested by both HRM and Sanger sequencing analysis. M13 tagged primers were designed to amplify the exonic sequences and HRM analysis was performed using the Roche LightCycler® 480. PCR products from HRM analysis were used for Sanger sequencing. Predictive modeling studies were used to investigate the functional effects of detected mutations by using the PopMuSic V2.1 and Swiss-Pdb Viewer V 4.0.4 programs.
Results: In 20 of 108 (18.5%) AML patient samples, HRM analysis identified variant sequences with possible mutations in five exons (8, 10, 15, 18 and 23). Sanger sequencing confirmed these samples to harbor mutations. Seven of twenty mutations detected were novel and previously unreported. Mutation of codon 882 of exon 23 was detected with highest frequency in 11 of 20 variant samples (55%). Complete concordance (100%) was observed between HRM and Sanger sequencing with no false negative results by HRM. Structural modeling showed that 7 of the 8 missense mutations increased the free energy (DG), destabilized the protein, and altered solvent accessibility indicating their 'loss-of-function' nature.
Conclusions: Successful detection of sequence variants by HRM with high sensitivity and specificity established HRM as an effective high-throughput screening method to detect DNMT3A mutations in a clinical diagnostic laboratory.
Tuesday, March 20, 2012 2:00 PM
Platform Session: Section G, Tuesday Afternoon