An ultrasensitive DNA-enhanced amplification method for detecting cfDNA drug-resistant mutations in non-small cell lung cancer with selective FEN-assisted degradation of dominant somatic fragments

Plasmids

The flap endonuclease 1 gene from the genomic DNA of A. fulgidus was synthesized with the additional sequence encoding a hexahistidine tag at the 3′ end by BGI BioTech (China) (GenBank ID: 1483479). The synthetic DNA product flanked by NcoI and BamHI restriction sites was inserted into the NcoI/BamHI sites of the expression vector pRSFDuet-1 (Novagen, USA).

FEN overexpression and purification

The expression plasmids were transformed into E. coli BL21(DE3) for protein overexpression and purification. The cells were grown to an optical density of 0.6–0.8 at 600 nm in Luria-Bertani medium, and protein was induced by adding 0.1 mM of isopropyl-β-D-thiogalactoside (IPTG). After incubation at 20 °C for 16 h, the cells were harvested by centrifugation and suspended in lysis buffer containing 20 mM Tris-HCl (pH 8.0) and 500 mM NaCl. The cells were lysed by sonication. After removal of cell debris by centrifugation (10,000×g, 30 min, 4 °C), the supernatant was heat-treated at 80 °C for 30 min and immediately cooled on ice, and then centrifuged (10,000×g, 30 min, 4 °C). Polyethylenimine was added slowly to the supernatant to a final concentration of 0.3 %. After stirring for 1 h, the mixture was centrifuged at 10,000×g for 30 min at 4 °C. The supernatant was adjusted to 70 % saturation with (NH₄)₂SO₄, stirring while adding for 1 h, letting it stand still for 1 h, and centrifuged at 10,000×g for 30 min at 4 °C. The pellet was resuspended in the above lysis buffer and dialyzed against the same buffer. The dialyzed material was then centrifuged (10,000×g, 30 min, 4 °C), the supernatant was loaded onto a Ni-Sepharose chelating column (1 mL; Qiagen) equilibrated with lysis buffer. The column was washed with lysis buffer, and bound proteins were eluted with an imidazole gradient (0–250 mM) in lysis buffer. Target protein fractions were pooled and subsequently loaded on a HiTrap desalting column (5 mL; GE Healthcare) equilibrated with desalting buffer containing 20 mM Tris-HCl (pH 8.0), 50 mM NaCl, and 20 % glycerol. All purification steps were carried out at 4 °C. The purified proteins were stored at −80 °C. Protein concentrations were determined by the BCA method with bovine serum albumin (BSA) (Beyotime Biotech, China) as the standard.

Oligonucleotides

The forward primer (EGFR FP) and reverse primer (EGFR RP) were employed to amplify a 174-base pair EGFR wild-type or T790M DNA fragments, respectively. Primers T790M forward primer 1, T790M forward primer 2, and EGFR RP were used to prepare T790M fragments. Probe 1 (5′-GGCAGCCGAAGGGCATGAGCTGCGGCCGCGCTGAGTGCCTTTTGGCACTCAGCGCGGCT-3′) (the sequence complementary paired with the positive-stranded template of the wild-type EGFR gene is underlined) and probe 2 (5′-CCTCCACCGTGCAGCTCATCACCGCCTGGACCTGGCGGTTTTCCGCCAGGTCCAGGCGT-3′) (the sequence complementary paired with the negative-stranded template of the wild-type EGFR gene is underlined) were annealed to the positive-strand and negative-strand of the wild-type EGFR gene to form two 5′-flap structures, respectively. The oligonucleotides were all synthesized at Sangon BioTech (Shanghai, China) (Supplemental Material, Table S1).

Reagents

Tris-base, sodium chloride, glycerol, magnesium chloride, potassium chloride, bovine serum albumin, ammonium sulphate and IPTG were purchased from Sangon BioTech. Polyethylenimine was purchased from Sigma-Aldrich Company.β-Mercaptoethanol (β-Me) used for flap endonuclease assay were purchased from Merck. 2 × Phanta Flash Master Mix used in EGFR DNA fragment amplification reaction was purchased from Vazyme. 2 × Taq Polymerase Mix was gifted by Meilun.

Flap endonuclease assay

The flap endonuclease assay was carried out in a PCR instrument with a 10-μL reaction mixture consisting of hairpin probe 1 (10 pmol or indicated concentration), hairpin probe 2 (10 pmol or indicated concentration), wild-type EGFR DNA fragment (120 fg), EGFR T790M DNA fragment (12 fg), Tris-HCl, pH 8.0 (50 mM), MgCl₂ (7.5 mM), KCl (30 mM),β-Me (0.5 mM), and BSA (50 μg/mL). Before FEN enzyme was added, the mixture was incubated at 95 °C for 5 min and annealed to 65 °C or indicated temperature. Then, FEN was added or not, and the reaction was incubated at 65 °C or indicated temperature for 20 min or longer. Under the indicated conditions, FEN was added directly into the reaction system to withstand the temperature of DNA template denaturation.

T790M DNA enhanced amplification

After the flap endonuclease assay, the reaction products were added to a 50-μL PCR amplification system in the presence of 1 × DNA Polymerase Mix, EGFR FP (20 pmol), and EGFR RP (20 pmol). The PCR reaction was subjected to a denaturation step (94 °C for 30 s), 30 cycles of a denaturation step (94 °C for 20 s), a hybridization step (55 °C for 20 s), and an extension step (72 °C for 20 s). At the end of PCR cycling, the reaction was extended for 5 min at 72 °C and lowered to 4 °C hold. Next, the samples were subjected to electrophoresis in 2 % agarose gel in 0.5 × TAE buffer. The gel was exposed to a gel imager (GE Healthcare) and recovered by DNA purification kit (Genfine BioTech). Then, the purified PCR products were confirmed by first-generation sequencing (Sangon BioTech).

T790M detection assay in clinical samples

Plasma cfDNA was prepared from sodium citrate-anticoagulated plasma samples using Apolstle Minimax High Effiency cfDNA Isolation Kit according to the World Medical Association Declaration of Helsinki. First, diluted cfDNA was used as the template for amplification by the high-fidelity DNA polymerase, and the amplified product was purified by gel electrophoresis. Then, the purified cfDNA was gradient diluted to participate in flap endonuclease assay, in which cfDNA cleavage by FEN was performed as described above except that cfDNA was added to the reaction mixture to replace the wild-type EGFR DNA fragment and EGFR T790M DNA fragment. Finally, T790M DNA-enhanced amplification was conducted as described above, followed by first-generation sequencing (Sangon BioTech). Meanwhile, as a control, cfDNA samples collected were sequenced by NGS to identify the mutation status of the EGFR gene. All samples were sequenced using Illumina Miseq in a 150-bp paired-end pattern. Adapter sequence and low-quality region were removed with Trimmomatic (version 0.38), and the clean reads were aligned to the hg19 reference genomes using ‘bwa mem’ (version 0.7.17). BAM files were realigned and base quality score were conducted using GATK (version 4.1). Potential mutations were identified through analysis of data across targeted regions of interest using SAMtools mpileup. Mutations were annotated using ANNOVAR and true mutations were verified using IGV (version 2.3.34).

Principle of the mutated DNA-enhanced amplification system

The overall experimental procedures of the EGFR T790M DNA-enhanced amplification detection assay are depicted in Figure 1. To selectively degrade the dominant wild-type DNA and increase the proportion of T790M DNA in the EGFR DNA mixture, we constructed a signal amplification system based on A. fulgidus flap endonuclease (Afu FEN). First, according to the DNA base sequence of wild-type T790, a pair of hairpin probes were designed to complement the base of a positive and negative strand of wild-type T790 duplex DNA, respectively, thus form the specific structure that FEN can cleavage. As a target DNA, wild-type T790 dsDNA fragment was denatured at 95 °C for 5 min to completely separate the double strand. Probe 1 (green) and probe 2 (blue) were hybridized to each of the two strands in the target DNA by annealing from 95 to 65 °C, so as to capture both strands of the separated target DNA (Figure 1A and B). The captured DNA complex produced two 5′-flap structures, which were specifically recognized and cleaved by Afu FEN at the specified sites (indicated by orange arrows). Next, FEN was added to the reaction system to cleave the 5′ flap downstream of the first base pair between the probe and the target DNA. Theoretically, FEN should be added to the reaction system after annealing the probe and target DNA, so that FEN would not be deprived of activity by denaturing at 95 °C; however, we also tried to add FEN to the reaction system at the beginning of the denaturation step and found that the activity of FEN was not significantly affected.

Figure 1:

Scheme for DNA-enhanced amplification-based detection of EGFR T790M mutation using selective FEN-assisted degradation of dominant somatic DNA fragments. (A) The main elements of the standard cleavage reaction mixture include EGFR wild-type T790 (blue) and mutated T790M (purple) DNA fragments, Archaeoglobus fulgidus flap endonuclease (Afu FEN) (orange), and a pair of 59 nt hairpin probes (green for probe 1 and light blue for probe 2). (B) Probe 1 and probe 2 form two 5′-flap structures by denaturation and annealing to EGFR wild-type T790 target DNA. Then, FEN specifically recognizes the structure and cleaves the T790 target DNA strands to release the flap fragments. The cleavage sites are indicated by orange arrows. (C) DNA-enhanced amplification reaction by PCR. Template: cleavage products by FEN; primers: EGFR FP and EGFR RP. (D) Agarose gel electrophoresis and purification of the amplified 174 bp DNA products. (E) First-generation sequencing chromatograms of the purified products.

After FEN cleavage of wild-type EGFR DNA, the mutated T790M DNA signal amplification reaction was performed by PCR, in which the product from the first reaction system was added as a template and then the amplification reaction was conducted with a pair of EGFR primers (Figure 1C). Here, since the amplification product is only 174 bp long, both high-fidelity and Taq DNA polymerases can be selected in the amplification system, which will not affect the DNA amplification. When PCR was completed, the amplification products were analyzed and purified on an agarose gel for first-generation sequencing (Figure 1D and E), which showed that, in the presence of FEN in the reaction system during the first step, the sequencing profile predominantly showed the ATG bases encoding methionine at position 790. In contrast, the negative control, which lacked FEN, showed the vast majority of the ACG bases encoding threonine. Thus, our method can easily detect the T base that may not be detected otherwise through first-generation sequencing. As a result, a highly sensitive strategy was achieved for the detection of EGFR T790M DNA.

Feasibility of the DNA-enhanced amplification system for detecting EGFR T790M

To investigate the applicability of the DNA-enhanced amplification system, the mutated EGFR T790M DNA fragments were spiked into wild-type EGFR DNA fragments at different ratios. Afu FEN protein was successfully expressed and purified (Supplementary Material, Figure S1). We first prepared short fragments of wild-type 790T and mutated 790M EGFR DNA by site-directed mutagenesis PCR (Supplementary Material, Figure S2), respectively. These fragments were 174 bp in length, and their sequences were shown in Figure 2A. According to the assay procedure, we mixed different ratios of the two DNA fragments, and then sequentially performed the cleavage reaction of FEN and the PCR-enhanced amplification reaction (Figure 2B). When the ratio of 790T to 790M was 10:1 (120 fg:12 fg), it was clearly seen on agarose gel electrophoresis that the amount of the reaction product was lower in the presence of FEN compared to its absence, indicating that FEN cleaved 790T, leading to a decrease in the amount of 790T template and consequently a reduction in the amount of product (Figure 2C). The products after electrophoresis were purified and subjected to first-generation sequencing. The sequencing results showed that the coding base of the amino acid at position 790 in the amplification product containing FEN was ATG, of which a small proportion of C base in the ACG was visible on the first-generation sequencing chromatogram. In contrast, the coding base of the amino acid at position 790 in the amplification product without FEN was ACG. However, the T base of ATG was not visible on the sequencing chromatogram (Figure 2D). This method can detect the mutant T base with initial frequency as low as 0.01 % (Supplementary Material, Figure S3). Therefore, the DNA-enhanced amplification assay involving the use of Afu FEN showed good performance in amplifying and detecting the signal of low-abundance EGFR T790M DNA.

Figure 2:

Construction of an enhanced amplification assay for EGFR T790M DNA based on flap endonuclease (FEN) cleavage. (A) DNA sequences of human wild-type EGFR fragment (790T) and T790M mutated fragment (790M), both 174 bp in length. (B) Schematic representation of the Afu FEN-specific cleavage reaction of the EGFR WT DNA fragments. (C) Agarose gel electrophoresis of 174-bp PCR reaction products mediated by FEN cleavage. 790T (120 fg) and 790M DNA fragments (12 fg) were incubated at 95 °C for 5 min and cooled down to 65 °C with probe 1 and probe 2 in the standard flap endonuclease cleavage assay mixture. After the probes were annealed to 790T, FEN was added (lane 2) or not (lane 1) into the standard reaction mixture at 65 °C for 20 min. Later, the standard reaction mixture was subjected to enhanced amplification by PCR. Samples were then resolved by electrophoresis in 2 % agarose gel, exposed to a gel imager, and recovered. M, DNA markers, 5,000 bp, 3,000 bp, 2,000 bp, 1,000 bp, 750 bp, 500 bp, 250 bp and 100 bp in length (from top to bottom). (D) First-generation sequencing of the174-bp DNA purified from the gel was performed to detect the mutated base encoding T790M (C → T). 1, no FEN in the reaction system; 2, 20 ng of FEN per reaction system. The mutated DNA base site ‘X’ to be detected is indicated by red arrows. The chromatographic peaks in different colors represent different nucleotides (blue for C, red for T, black for G, and green for A). Each experiment was repeated three times.

Optimization of the DNA-enhanced amplification assay for EGFR T790M DNA based on FEN cleavage

The cleavage efficiency of FEN in the flap endonuclease assay is important for the degradation of the dominant somatic DNA fragments. Therefore, the flap endonuclease assay was performed at 55 °C, 60 °C, 65 °C, and 70 °C, and the enhanced amplification results are shown in Figure 3A. At these four temperatures, the amount of all the final 174-bp DNA amplification products gradually decreased as the working concentration of FEN increased, suggesting that FEN degraded the wild-type dominant DNA templates. The amplification products were subsequently sequenced and the results showed that the addition of FEN enabled the detection of T bases encoding methionine. A significant increase in the proportion of distinct T base (red) sequencing chromatogram was observed when the working concentration of FEN reached 20 ng per reaction, far exceeding the proportion of C base (blue) (Figure 4B–E). The FEN enzyme was active at 55–70 °C, suggesting that FEN-assisted dominant somatic DNA fragments degradation system had a wide range of reaction temperatures, which provided a more relaxing environment for probe design and selection. To achieve faster FEN-mediated degradation of wild-type DNA, the flap endonuclease assay was performed at 60 and 65 °C, respectively, for 10 min. T bases were observed in the final first-generation sequencing results, but not at a higher percentage than that of C bases. This may be due to insufficient cleavage of the FEN within 10 min. When FEN cleavage reaction was conducted at 65 °C, the proportion of T bases in the final sequencing results was still slightly higher than that at 60 °C (Supplementary Material, Figure S4). Considering the specificity of the flap endonuclease reaction and the annealing efficiency of the probes to the templates, we performed the FEN cleavage reaction at 65 °C.

Figure 3:

Effects of temperature on the cleavage of dominant wild-type EGFR DNA fragments. (A) Agarose gel electrophoresis of the 174-bp PCR reaction products mediated by FEN cleavage at 55 °C, 60 °C, 65 °C, and 70 °C, respectively. 790T (120 fg) and 790M DNA fragments (12 fg) were incubated at 95 °C for 5 min and cooled down to indicated temperatures with probe 1 and probe 2 in the standard flap endonuclease cleavage assay mixture. After the probes were annealed to 790T, FEN was added or not into the standard reaction mixture at indicated temperatures for 20 min. Later, the standard reaction mixture was subjected to enhanced amplification by PCR. Samples were then resolved by electrophoresis in 2 % agarose gel, exposed to a gel imager, and purified. Lanes 1, 2, and 3: FEN cleavage reactions were incubated at 55 °C for 20 min. Lanes 4, 5, and 6: FEN cleavage reactions were incubated at 60 °C for 20 min. Lanes 7, 8, and 9: FEN cleavage reactions were incubated at 65 °C for 20 min. Lanes 10, 11, and 12: FEN cleavage reactions were incubated at 70 °C for 20 min. Lanes 1, 4, 7, and 10: no FEN. Lanes 2, 5, 8, and 11: 10 ng FEN per reaction system. Lanes 3, 6, 9, and 12: 20 ng FEN per reaction system. M: DNA markers, 5,000 bp, 3,000 bp, 2,000 bp, 1,000 bp, 750 bp, 500 bp, 250 bp, and 100 bp in length (from top to bottom). (B–E) First-generation sequencing was performed on 174-bp DNA purified from the gel to detect mutated bases (C → T) encoding T790M mediated by FEN cleavage at 55 °C (B), 60 °C (C), 65 °C (D), and 70 °C (E), respectively. The mutated DNA base site ‘X’ to be detected is indicated by red arrows. The chromatographic peaks in different colors represent different nucleotides (blue for C, red for T, black for G, and green for A). Each experiment was repeated three times.

Figure 4:

Effects of probe/template ratio on the cleavage of dominant wild-type EGFR DNA fragments. (A) Agarose gel electrophoresis of 174-bp PCR reaction products mediated by FEN cleavage at 65 °C for 20 min. 790T (10⁻⁶ pmol) and 790M DNA fragments (10⁻⁷ pmol) were incubated at 95 °C for 5 min and cooled down to 65 °C with indicated concentrations of probe 1 and probe 2 in the standard flap endonuclease cleavage assay mixture. After annealing, FEN (20 ng) was added into the standard reaction mixture at 65 °C for 20 min. Later, the reaction mixture was subjected to enhanced amplification by PCR. Samples were then resolved by electrophoresis in 2 % agarose gel, exposed to a gel imager, and purified. The working concentrations of probe1 and probe 2 were 10⁻⁷ pmol (lane 1), 10⁻⁵ pmol (lane 2), 10⁻³ pmol (lane 3), 10⁻¹ pmol (lane 4), and 10 pmol (lane 5). M: DNA markers, 5,000 bp, 3,000 bp, 2,000 bp, 1,000 bp, 750 bp, 500 bp, 250 bp, and 100 bp in length (from top to bottom). (B) First-generation sequencing was performed on 174-bp DNA purified from the gel to detect mutated bases (C → T) encoding T790M mediated by FEN cleavage. 1–5 correspond to the sequencing results of lanes 1–5 in (A). The mutated DNA base site ‘X’ to be detected is indicated by red arrows. The chromatographic peaks in different colors represent different nucleotides (blue for C, red for T, black for G, and green for A). Each experiment was repeated three times.

In the above reaction system, the molar concentration of wild-type 790T was 10⁻⁶ pmol. To explore the impact of probe/template molar ratio on the degradation reaction of wild-type dominant EGFR DNA fragments, a series of probes (probe 1 and probe 2) with molar concentration gradients were added to the reaction system. The working concentrations of the probes were 10⁻⁷ pmol, 10⁻⁵ pmol, 10⁻³ pmol, 10⁻¹ pmol, and 10 pmol. Accordingly, the mole ratios of the probe to the template were 0.1, 10, 10³, 10⁵, and 10⁷. The sequencing results showed that at a probe concentration of 10⁻¹ pmol, a small proportion of T base became detectable; however, when the probe concentration reached 10 pmol, T base became the dominant base (Figure 4). Therefore, an excess probe/template molar ratio ensures the formation of stable 5′-flap structures, which facilitates efficient degradation of the template DNA.

Performance of the DNA-enhanced amplification system in detecting T790M mutation in clinical samples

To further investigate the potential clinical value of our newly-developed assay, 50 plasma cfDNA samples from NSCLC patients were validated using the enhanced amplification system. Further, we selected two typical cfDNA reference materials for detailed analysis. The two cfDNAs are hereinafter referred to as specimen #1 and specimen #2, respectively, at a concentration of 5 ng/μL, in which the variant allele fraction (VAF) of specimen #1 was 0.1 % and VAF of specimen #2 was 0.5 %. The VAFs of EGFR T790M mutations were both marked in Figure 5A and D, respectively. Next, we used our method presented here to enrich the mutated T790M DNA for each of the two cfDNAs. First, we amplified and purified these two cfDNAs by simple PCR with EGFR FP and EGFR RP primers, respectively, to obtain the amplified cfDNAs of specimen #1 and specimen #2 at a concentration of 18 ng/μL for the convenience of subsequent operations. Then, 100-, 1,000-, and 10,000-fold dilutions of the concentrations of these two amplified cfDNAs for specimen #1 and specimen #2 were performed to participate in the FEN cleavage reaction as well as the subsequent enhancement amplification reaction.

Figure 5:

Detection of EGFR T790M DNA from clinical cfDNA reference materials with different variant allele fractions (VAFs). The mutation VAFs (measured by NGS) of cfDNAs from specimen #1 and specimen #2 are shown in (A) and (D), respectively. VAFs of T790M are labeled as blue triangles. (B–C, E–F) detecting mutated T790M of cfDNAs from specimen #1 and specimen #2 by the DNA-enhanced amplification method. cfDNA (3.6 pg) was incubated at 95 °C for 5 min and annealed to 65 °C with probe 1 and probe 2 in the standard flap endonuclease cleavage assay mixture. After annealing, FEN was added or not into the standard reaction mixture at 65 °C for 20 min, 40 min, 60 min, and 2 h, respectively. Later, the standard reaction mixture was subjected to enhanced amplification by PCR. Samples were then resolved by electrophoresis in 2 % agarose gel, exposed to a gel imager, and purified. First-generation sequencing was performed on the purified DNA. (B–C) DNA enhanced amplification electrophoresis after 2 h of FEN degradation of dominant somatic DNA (left) and first-generation sequencing chromatogram (right) of cfDNA at VAF of 0.1 % from specimen #1. (E–F) DNA enhanced amplification electrophoresis after 2 h of FEN degradation of dominant somatic DNA (left) and first-generation sequencing chromatogram (right) of cfDNA at VAF of 0.5 % from specimen #2. 1 and 4: no FEN in the detection system. 2 and 5: 10 ng of FEN per reaction system. 3 and 6: 20 ng of FEN per reaction system. M, DNA markers, 5,000 bp, 3,000 bp, 2,000 bp, 1,000 bp, 750 bp, 500 bp, 250 bp, and 100 bp in length (from top to bottom). The mutated DNA base site ‘X’ to be detected is indicated by red arrows. The chromatographic peaks in different colors represent different nucleotides (blue for C, red for T, black for G, and green for A). The degradation of dominant somatic DNA by FEN (20 ng) in specimen #1 (G) and specimen #2 (H) were performed at 65 °C for 20 min (7 and 10), 40 min (8 and 11), and 60 min (9 and 12), respectively. (I) Schematic illustration of degradation of clinically dominant wild-type somatic DNA by Afu FEN and amplification reaction of mutated DNA. Each experiment was repeated three times.

Figure 5B and C represent the cfDNA results of specimen #1, where ‘1’ is the result without FEN, and ‘2’ and ‘3’ are the results with different amounts of FEN. The first-generation sequencing chromatogram showed that the original C base was flipped to T base in ‘2’ and ‘3’ after the addition of FEN. Similarly, Figure 5E and F represent the cfDNA results of specimen #2, where ‘4’ is the result without FEN and ‘5’ and ‘6’ are the results with different amounts of FEN. The first-generation sequencing chromatogram showed that the original C base was converted to a T base in the presence of FEN (Figure 5C and F). The FEN cleavage reaction was carried out at 65 °C for 20 min to 2 h. The effective cleavage of somatic DNA fragments was achieved after 20 min when FEN was added to the reaction system (Figure 5G and H). This allows for the easy detection of a small proportion of EGFR T790M mutation in both specimens with high sensitivity, even at levels as low as pg (10,000-fold dilution). Thus, our novel mutated cfDNA enhanced amplification method is indeed a proposed strategy to detect EGFR mutation sites for selective degradation of dominant somatic fragments in NSCLC patients (Figure 5I).