An enzymatic on/off switch‐mediated assay for KRAS hotspot point mutation detection of circulating tumor DNA

Abstract Background To detect the mutations of KRAS gene in colorectal cancer patients and other cancer patients, it is of value to develop non‐invasive, sensitive, specific, easy, and low‐cost assays. Methods Templates harboring hotspot mutations of the KRAS gene were constructed, and primers were designed for evaluation of the specificity, and sensitivity of detection system consisted of exonuclease polymerase‐mediated on/off switch; then, gel electrophoresis and real‐time PCR were performed for verification. The assay was verified by testing the DNA pool of normal controls and circulating DNA (ctDNA) samples from 14 tumor patients, as compared to Sanger sequencing. Results A specific and sensitive assay consisted of exonuclease polymerase‐mediated on/off switch, and multiplex real‐time PCR method has been established. This assay could detect <100 copies of KRAS mutation in more than 10 million copies of wild‐type KRAS gene fragments. This assay was applied to test KRAS gene mutations in three cases of fourteen ctDNA samples, and the results were consistent with Sanger sequencing. However, this PCR‐based assay was more sensitive and easier to be interpreted. Conclusion This assay can detect the presence of KRAS hotspot mutations in clinical circulating tumor DNA samples. The assay has a potential to be used in early diagnosis of colorectal cancer as well as other types of cancer.

The mutations are mainly located in codons 12 and 13 with approximate 80% occurrence in the 12th codon, 15% occurrence in the 13th codon, and 5% occurrence in codon 61 and others. 5 The most majority of KRAS mutations identified are single nucleotide point mutations. The common patterns include G12D, G12A, G12R, G12C, G12S, G12V, and G13D. The mutation pattern of KRAS provides guidance of individualized treatment, and the KRAS wild-type colorectal cancer patients can benefit from treatment with anti-epidermal growth factor cetuximab. 6,7 KRAS mutation identification is strongly recommended by FDA to guide usage of cetuximab. Therefore, a sensitive, specific, and convenient method with detection capability of multi-mutations is highly desirable. In the past several decades, Sanger sequencing has been served as golden standards in the field. 8,9 Sanger sequencing interrogates every base and identifies known and unknown variants, yet it requires post-PCR manipulations such as cleanup of PCR.
Another more critical disadvantage is its low detection sensitivity.
It requires an abundance of 15% or more of a somatic DNA variant to be detected, which may lead to a false-negative result in some cases. [8][9][10][11] Allele-specific real-time PCR identifies KRAS mutations are sensitive enough but prone to false positive because of taq DNA polymerase. 9,[11][12][13] High-resolution melting analysis is the quantitative analysis of PCR products melting curve at different temperatures. Whereas it is highly sensitive and requires no post-PCR process, melting curve analysis has no multiplex capacity and the result is difficult to be interpreted. 14,15 SnaPShot and next-generation sequencing, the sensitivity of SnaPshot for KRAS mutation detection was about 10%, but this method is also laborious and time-consuming. 16,17 Next-generation sequencing is relatively expensive, time-consuming, and complicated data interpreting. 18,19 Choong et al detected the KRAS mutations using the isothermal-based optical sensor for companion diagnostics, 20 a rapid, specific, and sensitive method, but the preparation process is relatively complex and the result is not easy to interpret. Therefore, a high-throughput, economical, sensitive, and easy data analysis assay is urgently needed. Mutation-sensitive on/off switch, 21,22 consisting of high-fidelity DNA polymerase and phosphorothioate-modified allele-specific primers, has been established to have a higher specificity compared with assays mediated by low-fidelity DNA polymerase. Table 1 shows the comparison of the KRAS mutation detection assay.
Cell-free DNA (cfDNA) is defined as extracellular DNA dissociating in the circulating system. The consistency of DNA alterations between a tumor and cfDNA proves the tumoral origin of the cfDNA. 23 Circulating tumor DNA is derived from the release of nucleic acids during the apoptosis and necrosis of tumor cells or from tumor-derived exosomes. 24 In cancer patients, ctDNA presents as a variable fraction of cfDNA (ranging from <0.01% to more than 50%). 25 The lengths of ctDNA from tumor patients range from 200 nt to longer than 1 kb. 26 As biomarkers of cancer, 27 the qualitative and quantitative analyses of the ctDNA in the blood have been used for early diagnosis of cancer, individualized therapy, prognosis, etc A highly sensitive, specific method might be

| The Statistical analysis of the database
The gene mutation data and survival data of the colorectal cancer patients were downloaded from The Cancer Genome Atlas (TCGA) database. The overall survival rates between patients with KRAS mutations and patients without KRAS mutations were analyzed by Kaplan-Meier, and P < .05 was considered statistically significant.

| Preparation of KRAS hotspot mutation templates
Preparation of KRAS hotspot mutation templates was carried out by the method of overlapping extension PCR. Sixteen primers were listed in Table S1. These primers were designed according to the reference sequences of Gene ID 3845 in Genbank. The primers were designed to introduce hotspots mutations at codon 12 and codon 13 of KRAS gene. The primer-extended products were purified again and then subcloned into PGEM-T Easy Vector (Invitrogen). Subcloned products were sequenced for confirmation. The seven mutation templates of KRAS gene, including G12D, G12V, G12S, G12C, G12R, G12A, and G13D, are shown in Figure S1.

| The condition of the detection of KRAS mutations
The following nine mutation-specific primers were designed for mutation analysis (

| PCR conditions for amplifying targets for KRAS mutations sequencing analysis
Taq polymerase-mediated regular PCR was used for sequencing of the DNA samples from the normal controls and cancer patients.
The primers for amplifying the products for sequencing are listed in Table S1. This regular PCR was performed in a volume of 25

| A PCR-based assay of KRAS mutation detection was established
We established the detection assays of single or multiplex PCR system combination with high-fidelity enzyme "on/off"switch. As shown in Figure 2A Figure 2B. Figure 2C shows that real-time PCR is more sensitive than conventional PCR. The sensitivity of this assay by real-time PCR for the KRAS seven hotpot mutations is at least 10 times higher than conventional PCR without compromising its specific-

| KRAS mutations cannot be detected in healthy human DNA samples
This assay was further tested for its application in clinical samples. The efficient extension of KRAS mutation primers would not be observed in healthy human genomic DNA samples. As shown in Figure 3, we used one pooled DNA sample of 200 normal individuals for evaluating the specificity of our method. Wt template were diluted in 10-fold serial and amplified by PCR with mutation-specific primer. The length of the PCR products was both 155 bp. C, Establishment of a multiplex PCR system with codon 12, as the melt curve and amplification plot shows, the detection limit reach to 10 2 copies, and the specificity can reach to 10 6 Mutation-specific primers failed to amplify products in these samples, while products were amplified by wild-type-specific detection primers. Codon 13 primers of KRAS could amplify a primer dimer, and we could distinguish the objective fragments from the dimer according to the Tm value. To further confirm this method, the PCR products of the KRAS gene were sent for sequencing and there were no point mutations in codon 12 and codon 13 of KRAS gene.

| Application of the current assay in KRAS hotspot mutation detection with cancer patients
CtDNA samples derived from fourteen tumor patients were tested with the newly developed KRAS hotspot mutation assay, two samples were detected carrying at least one type mutations of codon 12 and one sample was detected carrying codon abundance of mutated templates. 10,28 Screening the KRAS mutations by HRM had a sensitivity of 5%-6%. 10,28 Hillary et al were able to detect KRAS mutations by allele-specific hybridization-induced aggregation (HIA) of oligonucleotide probe-conjugated microbeads with a sensitivity of 25%. 29 Our multiplex real-time PCR method can detect KRAS mutation as few as 100 copies in at least 10 5 wild-type counterparts. Currently, ctDNA concentrations vary from 0 to 1000 ng/mL blood, with the median concentration of cfDNA 790 ng/mL. 30,31 In general, 1000 ng cfDNA includes approximately 150 000 copies of the genome 31 ; accordingly to this, it is difficult to cause false-positive amplification in one PCR system by our assay.
In the present study, efficient primer extension was exclu-

| CON CLUS ION
The present study has successfully established a non-invasive diagnostic method for KRAS mutation detection in circulat- ing DNA of cancer patients. The allele-specific real-time PCR is mediated by mutation-sensitive "molecular switch" composed of high-fidelity DNA polymerase and phosphorylated-modified primers. We have demonstrated the success of using the assay in micro-mutation detection with ctDNA. This technology has a great potential in screening KRAS gene mutations thanks to its high specificity, sensitivity, low-cost, multiplex detection, and easy data interpretation.

ACK N OWLED G M ENTS
This study was supported by The Priority Academic Program

CO N FLI C T O F I NTE R E S T
No competing financial interests exist.