High resolution melting analysis for the rapid and sensitive detection of mutations in clinical samples: KRAS codon 12 and 13 mutations in non-small cell lung cancer

Abstract

Background: The development of targeted therapies has created a pressing clinical need for the rapid and robust molecular characterisation of cancers. We describe here the application of high-resolution melting analysis (HRM) to screen for KRAS mutations in clinical cancer samples. In non-small cell lung cancer, KRAS mutations have been shown to identify a group of patients that do not respond to EGFR targeted therapies and the identification of these mutations is thus clinically important.

Methods: We developed a high-resolution melting (HRM) assay to detect somatic mutations in exon 2, notably codons 12 and 13 of the KRAS gene using the intercalating dye SYTO 9. We tested 3 different cell lines with known KRAS mutations and then examined the sensitivity of mutation detection with the cell lines using 189 bp and 92 bp amplicons spanning codons 12 and 13. We then screened for KRAS mutations in 30 non-small cell lung cancer biopsies that had been previously sequenced for mutations in EGFR exons 18-21.

Results: Known KRAS mutations in cell lines (A549, HCT116 and RPMI8226) were readily detectable using HRM. The shorter 92 bp amplicon was more sensitive in detecting mutations than the 189 bp amplicon and was able to reliably detect as little as 5-6% of each cell line DNA diluted in normal DNA. Nine of the 30 non-small cell lung cancer biopsies had KRAS mutations detected by HRM analysis. The results were confirmed by standard sequencing. Mutations in KRAS and EGFR were mutually exclusive.

Conclusion: HRM is a sensitive in-tube methodology to screen for mutations in clinical samples. HRM will enable high-throughput screening of gene mutations to allow appropriate therapeutic choices for patients and accelerate research aimed at identifying novel mutations in human cancer.

Figure 1

Location of KRAS codon 12 and 13 mutations and PCR amplicons. Exon 2 of KRAS is shown from the ATG without the untranslated region. The position and size of the PCR amplicons used in the HRM assays in relation to exon 2 of KRAS is indicated. There are 12 possible single base mutations that can occur at codons 12 and 13 of KRAS. All possible mutations are listed along with the corresponding amino acid changes. These mutations are located at positions 64, 65, 67 and 68 of the 189 bp PCR amplicon and at positions 53, 54, 56 and 57 of the 92 bp PCR amplicon.

Figure 2

Difference plots of KRAS mutated cell lines. Panel A: A normal plot of the wild-type controls with the cell line samples. Fluorescence is on the Y-axis and temperature in degrees Celsius is on the X-axis. Panel B: A difference plot of the wild-type controls with the cell line samples. All cell line samples were compared to the median wild-type control sample to produce the plot.

Figure 3

Sensitivity of the KRAS HRM assay for different amplicons. Panel A: The HCT116 dilutions were compared to the median wild-type control sample to produce the difference plot for the 189 bp amplicon. Panel B: The HCT116 dilutions were compared to the median wild-type control sample to produce the difference plot for the 92 bp amplicon. Panel C: The A549 dilutions were compared to the median wild-type control sample to produce the difference plot for the 189 bp amplicon. Panel D: The A549 dilutions were compared to the median wild-type control sample to produce the difference plot for the 92 bp amplicon. Panel E: The RPMI8226 dilutions were compared to the median wild-type control sample to produce the difference plot for the 189 bp amplicon. Panel F: The RPMI8226 dilutions were compared to the median wild-type control sample to produce the difference plot for the 92 bp amplicon.

Figure 4

Difference plot and sequencing traces of various mutations in KRAS. Sequencing plots for each sample are below the patient numbers. The samples were compared to the median wild-type control sample to produce the difference plot. The sequencing traces for patients 3 and 6 showed the presence of a 35G > T and 35G > A mutations respectively.