Robust and Sensitive Detection of Insertions, Deletions and Point Mutations In CEBPA, a GC-Rich Content Gene, Using 454 Next-Generation Deep-Sequencing (NGS).

Abstract 1657

CCAAT/enhancer binding protein alpha (CEBPA) is an essential transcription factor for granulocytic differentiation and encodes a protein exclusively expressed in the myelomonocytic lineage. Mutations are seen in 6% to 19% of acute myeloid leukemia (AML) and biallelic CEBPA mutations have been associated with a favorable clinical outcome. Today, screening of CEBPA mutations in AML patients is usually performed combining fragment length analysis to detect insertions and deletions, denaturing high-performance liquid chromatography (DHPLC), and subsequent direct Sanger sequencing. Notably, each assay has its strengths and weaknesses, i.e. fragment length analysis is not able to detect substitutions (25% of all mutations in our selected cohort), and DHPLC misses rare mutations, especially those located at the end of the amplicons or those resulting from base duplications in AT- or GC-rich content regions. Finally, Sanger sequencing, while being able to detect all sorts of mutations, has an accepted lower cut-off value of 20% diagnostic sensitivity. This study aimed at establishing a robust assay for detecting CEBPA mutations in AML patients using 454 Titanium amplicon NGS. 454 deep-pyrosequencing technically includes an emulsion PCR (emPCR) step that allows a massively parallel clonal amplification of PCR products, thereby permitting a highly sensitive detection of CEBPA mutations. Initially, we tested this procedure on two patients using the standard emPCR condition according to the manufacturer's recommendation on four overlapping CEBPA fragments. In this setting, only amplicons 1 and 4 generated reads. This was due to 454 Titanium chemistry laboratory procedures that, so far, lacked efficient amplification of GC-rich amplicons. In detail, the GC-content for the respective CEBPA amplicons was as follows: amplicon 1: 73%, amplicon 2: 76%, amplicon 3: 77%, and amplicon 4: 69%. Therefore, in order to improve the amplification reactions, we investigated six distinct emPCR conditions. We could define a robust amplification method of all four CEBPA fragments, even amplicon 3 with the highest GC-content of 77%. Subsequently, the performance of this assay was tested on a larger independent cohort of 24 AML patients, which were preselected according to their known CEBPA mutation status. All patients had been investigated first with conventional methods, i.e. DHPLC or fragment length analysis followed by Sanger sequencing. After excluding silent mutations and polymorphisms, we observed 35 distinct mutations with NGS. In particular, 454 next-generation sequencing allowed a highly sensitive detection of variances. In comparison to the data previously known from our conventional methods, i.e. 30 mutations in 24 patients, we detected additional 5 mutations (n=3 <15% of sequencing reads). These five novel mutations were not observed before due to technical limitations of the routine methods as described above. Interestingly, most CEBPA-mutated AML cases carried two mutations, which often involved a combination of N-terminal and bZIP mutations. As only these biallelic mutations in CEBPA were shown to be associated with favorable clinical outcome, the detection of all mutations is critical. In the cohort of 24 patients analyzed here 13 cases harbored more than one mutation. In three cases these mutations were detected in the same amplicon and in ten cases the mutations occurred in separate amplicons. Moreover, in 3 cases with mutations that occurred in the same amplicon, 454 deep-sequencing allowed a differentiation between monoallelic or biallelic status. In conclusion, an efficient screening of CEBPA mutations currently requires a combination of different methods and therefore is labor-intensive. Due to the high GC-content, NGS was not able to fully sequence the complete gene. Using our adjusted emPCR protocol we present a modified master mix and reaction condition to amplify GC-rich content amplicons and to overcome this technical limitation. Therefore, adjusted NGS is a suitable method that allows the detection of point mutations, insertions, duplications, or deletions in CEBPA with important clinical relevance in AML, and, furthermore, represents the most sensitive assay available thus far for screening of CEBPA mutations in a diagnostic setting. Moreover, this assay potentially offers a reliable assessment of minimal residual disease status for patient-specific CEBPA mutations.