Quantification of rare somatic single nucleotide variants by droplet digital PCR using SuperSelective primers

Abstract

Somatic single-nucleotide variants (SNVs) occur every time a cell divides, appearing even in healthy tissues at low frequencies. These mutations may accumulate as neutral variants during aging, or eventually, promote the development of neoplasia. Here, we present the SP-ddPCR, a droplet digital PCR (ddPCR) based approach that utilizes customized SuperSelective primers aiming at quantifying the proportion of rare SNVs. For that purpose, we selected five potentially pathogenic variants identified by whole-exome sequencing (WES) occurring at low variant allele frequency (VAF) in at-risk colon healthy mucosa of patients diagnosed with colorectal cancer or advanced adenoma. Additionally, two APC SNVs detected in two cancer lesions were added to the study for WES-VAF validation. SuperSelective primers were designed to quantify SNVs at low VAFs both in silico and in clinical samples. In addition to the two APC SNVs in colonic lesions, SP-ddPCR confirmed the presence of three out of five selected SNVs in the normal colonic mucosa with allelic frequencies ≤ 5%. Moreover, SP-ddPCR showed the presence of two potentially pathogenic variants in the distal normal mucosa of patients with colorectal carcinoma. In summary, SP-ddPCR offers a rapid and feasible methodology to validate next-generation sequencing data and accurately quantify rare SNVs, thus providing a potential tool for diagnosis and stratification of at-risk patients based on their mutational profiling.

Figure 1

Diagram of the SP-ddPCR (SuperSelective primers in droplet digital PCR) optimization method to test SuperSelective primer accuracy and selective amplification of the allele of interest using DNA templates containing 10,000 molecules of the target loci bearing either the wild-type allele or the SNV. These measurements correspond to the estimation of 100% and 0% VAF by ddPCR. Suitable SuperSelective primers designed to detect APC c.4128T>A are shown as an example. WT, wild-type.

Figure 2

SP-ddPCR performance for quantification of rare SNVs measured in synthetic plasmid mixtures simulating 1.00, 0.50, 0.25, 0.13 and 0.00% VAFs. VAFs (%) measured by ddPCR are plotted versus expected theoretical VAF (%) for serial dilutions of plasmids containing the corresponding SNV sequence into the respective wild-type background consisting of plasmids containing the wild-type allele. Each mixture was measured in triplicate. Dots express mean values and error lines indicate standard deviations. R squared values of Pearson correlation between measured and expected VAF are displayed within each plot. (a) APC c.2626C>T, p-value 0.002 (b) APC c.4128T>A, p-value 0.0007 (c) FABP4 c.105G>A, p-value 0.001 (d) NTRK2 c.220G>A, p-value 0.000002 (e) LAMC3 c.1241G>A, p-value 0.00005 (f) NRXN3 c.1421G>A, p-value 0.000002 (g) ASNA1 c.193C>T, p-value 0.003.

Figure 3

Validation of SNV-associated VAFs by SP-ddPCR in clinical patient samples. Plots show raw data output for fluorescence amplitude. Standard, total copies of the target gene measured by standard primers; SSP, quantification of SNV copies by SuperSelective primers; NTC, non-template control well containing equivalent amounts of DNA but not the target sequence. (a) Surrounding mucosa from APC c.4128T>A patient (b) Distal mucosa from APC c.4128T>A patient (c) CRC lesion from APC c.4128T>A patient (d) Surrounding mucosa from NRXN3 c.1421G>A patient (e) Distal mucosa from NRXN3 c.1421G>A patient (f) CRC lesion from NRXN3 c.1421G>A patient (g) Surrounding mucosa from NTRK2 c.220G>A patient (h) Distal mucosa from NTRK2 c.220G>A patient (i) AAD lesion from NTRK2 c.220G>A patient.