FormalPara Key Points

The baseline noise in normal peripheral blood and formalin-fixed paraffin-embedded samples detected by next-generation sequencing (NGS) is dominated by C:G > T:A mutations, which are signature mutations of cytosine deamination

Consistent with this, treatment of samples with an enzyme designed to remove uracil reduced the frequencies of these mutations, suggesting that one source is biologic. It was also demonstrated that the heat of thermocycling (in the absence of polymerase) can increase the frequencies of these mutations

It was concluded that the major sources of baseline noise in NGS are both biologic and laboratory-induced cytosine deamination

1 Introduction

Next-generation sequencing (NGS) has been revolutionizing biomedical research and clinical practice for the past 9 years, since its introduction [1]. With substantial improvements in accuracy, read length, and depth of coverage, and remarkable reductions in costs, NGS is becoming a major sequencing platform for clinical diagnostic laboratories. In the field of oncology, NGS has allowed detection of all potential driver gene mutations that cause a particular malignancy and thus clinically reports the status of all genes known to predispose to a particular type of cancer [2–7]. To embrace the era of targeted therapy and personalized medicine, NGS will be an essential tool to evaluate all related genes that predict drug response and clinical outcome, and will offer comprehensive genetic profiling, especially if target-based treatment covers mutations in multiple gene/pathway family members [8, 9].

A promising clinical area in which NGS will likely play a role is minimal residual disease (MRD) testing of solid tumors. This will allow clinicians to initiate second- or third-line therapy more quickly on the basis of residual tumor molecules, rather than waiting until the tumor progresses radiographically [10, 11]. On the other hand, a negative MRD test could be used to avoid the risk and expense associated with additional chemotherapy, when the patient is actually in remission.

Because of its extremely high depth of coverage, NGS will likely also be the tool of choice for early detection of cancer through measurement of critical oncogene-activating mutations [12] and structural abnormalities [13, 14]. While NGS holds tremendous promise for clinical use, it also faces challenges associated with technical complexity and result interpretation. Reproducible sequence artifacts constitute the baseline noise of sequencing and may interfere with detection of true gene mutations. This is a critical concern to address when extremely low mutation frequencies need to be detected unambiguously for these and other applications.

It has been demonstrated that cytosine deamination contributes to background noise in DNA sequencing of ancient and formalin-fixed paraffin-embedded (FFPE)-treated DNA [15, 16]. Since it can manifest as either a base substitution of C to T (C > T) on the sense strand or as a G > A mutation on the sense strand (arising from a C > T deamination on the antisense strand), these mutations are collectively designated C:G > T:A. Williams and colleagues [17] observed formalin fixation-induced C:G > T:A transitions. The same artifact was also observed in freshly prepared samples in a recent study validating detection of gene mutations in an Ion AmpliSeq™ Cancer Hotspot Panel v2, using an Ion Torrent Personal Genome Machine® (PGM™) [both from Life Technologies, Carlsbad, CA, USA] in our laboratory [18]. We observed significantly higher C:G > T:A mutation frequencies than background noise levels during NGS in both normal peripheral blood and FFPE samples, by checking all common KRAS [Kirsten rat sarcoma viral oncogene homolog] (codons 12 and 13), BRAF [B-Raf proto-oncogene, serine/threonine kinase] (V600E), and EGFR [epidermal growth factor receptor] (T790M and L858R) gene point mutations. These C:G > T:A mutations must be either biologic (intrinsic to the sample prior to isolation) or an artifact of the molecular biology, including DNA isolation, polymerase chain reaction (PCR) amplification, and/or sequencing.

During routine NGS, we noted an increase in C:G > T:A transitions in normal samples. In this report, we identify the etiology of NGS baseline noise, explore its contributing factors, and provide a partial solution. To confirm cytosine deamination as the source, we treated DNA from peripheral blood with uracil N-glycosylase (UNG) and demonstrated that it significantly reduced the frequencies of C:G > T:A mutations, by checking all common KRAS (codons 12 and 13), BRAF (V600E), and EGFR (T790M and L858R) gene point mutations. Consistent with a previous finding that prolonged heating may induce deamination of DNA [19], we found that the heat associated with thermocycling induced a significant increase in the C:G > T:A mutation frequency. We next showed that UNG pretreatment of positive control samples does not interfere with the capacity of NGS to detect real mutations. Finally, we attempted to include a thermostable UNG in PCR reactions, but we were unable to identify conditions that would allow both enzymes to work.

2 Materials and Methods

2.1 Materials

This study was conducted with institutional review board approval. The specimens consisted of four peripheral blood specimens from normal donors and seven FFPE samples from patients carrying distinctive EGFR, KRAS, and BRAF gene mutations. DNA was isolated as described previously [20, 21]. DNA concentrations were determined by a Qubit® 2.0 Fluorometer, and UNG enzyme 0.5 µL [1 unit/µL] (both from Life Technologies) was added into each reaction (30 ng DNA, 20 µL for total volumes) and incubated for 30 min at 50 °C prior to thermocycling for library preparation.

2.2 The NGS Platform

NGS was conducted using the Ion AmpliSeq™ Cancer Hotspot Panel v2 for targeted multigene amplification, as described in our previous study [18]. An Ion AmpliSeq™ Library Kit 2.0 was used for library preparation (PCR thermocycling at 65 °C for 30 min and 95 °C for 2 min, 20 cycles at 95 °C for 15 s and 60 °C for 4 min, with a hold at 10 °C), with an Ion PGM™ Template OT2 200 Kit and an Ion OneTouch™ ES Instrument for emulsion PCR and enrichment, an Ion PGM™ Sequencing 200 Kit v2, Ion 318™ Chips, and the PGM™ sequencing platform for NGS [all from Life Technologies], as recommended by the manufacturers’ protocols, without modification. The DNA input for targeted multigene PCR was 30 ng. Eight specimens were barcoded using Ion Xpress™ Barcode Adapters (from Life Technologies), pooled, and run on a single Ion 318™ Chip. For samples treated with extra heat, we cycled them at 99 °C for 20 min, with 40 cycles at 99 °C for 2 min and 60 ° for 4 min, and a hold at 10 °C prior to library preparation.

2.3 Data Analysis

The sequencing data were analyzed using Ion Torrent Suite™ Version 3.2.0 (from Life Technologies). The frequency (percentage) of all of the common KRAS (codons 12 and 13), BRAF (V600E), and EGFR (T790M and L858R) gene point mutations and those at five randomly picked base positions with nucleotide G or C in amplified regions of chromosomes 7 and 12 were calculated. The C:G > T:A mutations include KRAS G12D (GGT > GAT), G12S (GGT > AGT), G13D (GGC > GAC), G13S (GGC > AGC); and EGFR T790M (ACG > ATG). The non-C:G > T:A mutations include KRAS G12A (GGT > GCT), G12C (GGT > TGT), G12R (GGT > CGT), G13A (GGC > GCC), G13C (GGC > TGC), G13R (GGC > CGC); and BRAF V600E (GTG > GAG).

3 Results

3.1 C:G > T:A Mutation Frequencies in Peripheral Blood Are Reduced by Uracil DNA Glycosylase

In our previous study, we found that C:G > T:A mutations were significantly more common than other mutations in both peripheral blood and FFPE specimens [18]. As shown in Fig. 1a, the frequencies of C:G > T:A mutations—including KRAS G12D, G12S, G13D, and G13S; EGFR T790M; and those at five randomly picked positions in PCR amplicons (shown by black circles)—were significantly (about 8-fold) higher than other baseline noise levels (shown by black triangles) (Fig. 1a, d; p < 0.01). To test whether cytosine deamination had occurred prior to library construction, we treated normal peripheral blood specimens with uracil DNA glycosylase (UNG) before we conducted the initial AmpliSeq™ PCR. The glycosylase activity of UNG excises the uracil base from DNA, leaving the sugar–phosphate backbone intact, thereby functionally removing that strand from the PCR reaction, as the polymerase cannot synthesize across the abasic site. UNG treatment reduced the frequencies of C:G > T:A mutations at all of the above sites (black circles), except for two of them (shown by black arrows in Fig. 1b, c), demonstrating that some of the C:G > T:A mutations arose from deamination of cytosine to uracil prior to PCR. The overall reduction in C:G > T:A mutation frequencies was approximately 30 % and statistically significant (Fig. 1d; p < 0.05), although the C:G > T:A mutation frequencies following UNG treatment were still significantly higher than those of the other mutations (Fig. 1d; p < 0.01). In FFPE specimens, we also observed a 22 % reduction in mutation frequencies, but without statistical significance, probably because of higher levels of variation in FFPE samples (data not shown).

Fig. 1
figure 1

Uracil N-glycosylase (UNG) reduces the frequencies of C:G > T:A mutations. The background noise of next-generation sequencing for each of the common mutations within KRAS [Kirsten rat sarcoma viral oncogene homolog] (codons 12 and 13), BRAF [B-Raf proto-oncogene, serine/threonine kinase] (V600E), EGFR [epidermal growth factor receptor] (T790M and L858R), and five additional C:G sites in amplified regions of chromosomes 7 and 12 (a) without or (b) with UNG treatment prior to polymerase chain reaction (PCR) amplification was evaluated in normal peripheral blood samples from four healthy subjects. The C:G > T:A mutations are shown by black circles, whereas other potential mutations are shown by black triangles. (c) The difference between the two groups is shown by subtraction of the “No UNG” values from the “UNG” values. Note that most C:G > T:A mutation frequencies are reduced with UNG treatment, except at two sites (shown by black arrows). The reason for the lack of an effect on the latter two sites is unknown. (d) The C:G > T:A mutation frequencies and the other baseline noise levels are averaged, and the C:G > T:A mutation frequencies are significantly higher than the other baseline noise levels (p < 0.01; Student’s t test). UNG treatment prior to PCR significantly decreases the C:G > T:A mutation frequency (p < 0.05; paired t test), without affecting the frequency of the other mutations (p > 0.05; paired t test). Following UNG treatment, the C:G > T:A mutation frequency is still significantly higher than the other baseline noise level (p < 0.01; Student’s t test). The error bars represent the standard deviations

Full size image

3.2 Heat Associated with Thermocycling Induces Deamination

Considering a previous observation that prolonged heat could induce cytosine deamination of DNA [19], we hypothesized that the denaturation phase of thermocycling during library preparation and emulsion PCR might cumulatively cause such an effect. To test this, we thermocycled DNA from peripheral blood specimens (without performing PCR) prior to library preparation. As shown in Fig. 2, this treatment induced additional deamination effects at all susceptible positions (shown by black circles), except for two of them (shown by black arrows in Fig. 2b, c). Compared with the statistically significant increase in the overall C:G > T:A mutation frequencies caused by additional thermocycling (Fig. 2d; p < 0.05), the frequencies of other baseline mutations (shown by black triangles) at non-cytosine positions was not obviously affected by the treatment (Fig. 2d; p > 0.05).

Fig. 2
figure 2

Thermocycling induces deamination. The background noise of next-generation sequencing for each of the common mutations within KRAS [Kirsten rat sarcoma viral oncogene homolog] (codons 12 and 13), BRAF [B-Raf proto-oncogene, serine/threonine kinase] (V600E), EGFR [epidermal growth factor receptor] (T790M and L858R), and five additional C:G sites in amplified regions of chromosomes 7 and 12 (a) without or (b) with prolonged heat treatment prior to library preparation was evaluated in normal peripheral blood samples from four healthy subjects. (c) The difference between the two groups is shown by subtraction of the “No Heat” values from the “Heat” values. While most of the values are above zero, there is clearly some variability among the positions in the response to heat. The C:G > T:A mutations are shown by black circles and are induced with heat treatment, except at two sites (shown by black arrows). Control mutations are shown by black triangles. (d) The C:G > T:A mutation frequencies and the other baseline noise levels are averaged, and prolonged heat significantly increases the C:G > T:A mutation frequency (p < 0.05; paired t test), without affecting the frequency of the other mutations (p > 0.05; paired t test). The error bars represent the standard deviations

Full size image

3.3 UNG Does Not Interfere with Detection of Real Mutations

Since UNG reduces the frequencies of deamination-related mutations, it is a potential tool to lower NGS background noise levels. With this consideration, we wanted to exclude the possibility that UNG may reduce the ability of NGS to detect bona fide gene mutations. We selected seven positive control FFPE samples carrying a wide range of mutant allele frequencies (Table 1) of distinct EGFR, KRAS, and BRAF mutations. We then treated those DNA samples with UNG prior to library preparation. Compared with the results from untreated specimens, the percentages of the mutations detected after UNG treatment were consistent with the previously determined mutation frequencies. Thus, UNG treatment did not interfere with the capability of NGS to detect clinically important mutations (Fig. 3; Table 1), and this was consistent with the findings of other studies [16].

Table 1 Mutation frequencies with and without uracil N-glycosylase (UNG) treatment
Full size table
Fig. 3
figure 3

Uracil N-glycosylase (UNG) does not interfere with detection of real mutations by next-generation sequencing. The mutation frequencies detected in UNG-treated positive control samples (shown on the y-axis) are compared with those in untreated ones (shown on the x-axis)

Full size image

4 Discussion

In this study, we demonstrated that the baseline noise of NGS is mainly attributable to cytosine deamination and that the source is both biologic and thermocycling induced. UNG eliminates the uracil that results from deamination and thus is a tool to reduce biologic background noise levels in NGS. Another possible method to eliminate PCR-induced mutations should be to avoid heating by employing isothermal amplification technologies. Pretreating samples with UNG does not inhibit the ability to detect known positive control mutations. Our conclusions are all based on the Ion Torrent platform; however, our findings are consistent with similar work on the MiSeq system analyzing FFPE samples [22].

NGS is a powerful tool to discover novel disease-related genetic variations, to clinically diagnose and predict disease on the basis of comprehensive genetic profiling, and to reveal therapeutic targets. Extremely high depth of coverage allows NGS to be highly sensitive and accordingly suitable for discovery of rare genetic variants, including early detection of cancer and monitoring of MRD in cancer patients. In conducting NGS, reproducible sequence artifacts may produce false positives or interfere with detection of true gene mutations. For early detection of cancer or MRD monitoring, where extremely low mutation frequencies need to be identified unambiguously, nonspecific background noise needs to be minimized, if not eliminated.

Cytosine deamination is actually one of the most prevalent point mutations spontaneously occurring in nature, thereby contributing to background noise for sequencing [15, 16]. The two major underlying mechanisms include deamination of 5-methylcytosine, resulting in thymine and ammonia. In DNA, this reaction can be corrected by the enzyme thymine-DNA glycosylase prior to passage of the replication fork, otherwise a cytosine to thymine base substitution is generated [23, 24]. The other mechanism of deamination involves hydrolysis of cytosine into uracil. This deamination in DNA is corrected by the DNA glycosylase UNG, which removes the uracil base to generate an abasic site, which is then repaired by adding back a cytosine opposite the guanine. However, if the pro-mutagenic G–U mispair is not repaired prior to the next round of DNA replication, a U:A mutation is generated [23, 24], which results in a T:A during the next round of synthesis.

Deamination can be attributed to multiple factors, including biologic factors (intrinsic to the sample prior to isolation) or an artifact of the molecular biology [23–25]. For studies of ancient DNA, it is a major source of sequencing artifacts [26]. In addition to age, it has been observed that formalin induces C:G > T:A transitions in Sanger sequencing [17]. However, we identified the same artifact in freshly prepared samples in our recent study validating detection of gene mutations using NGS [18]. This adds evidence that deamination may also result from polymerase-induced errors along with the lack of DNA repair, or directly from the heat associated with thermocycling [19]. Nevertheless, it appears that the frequencies of these mutations are higher in FFPE samples than in peripheral blood [18], indicating that the process of specimen fixation also contributes to the noise level.

In the current study, we demonstrated that biologic deamination contributes to C:G > T:A mutations in background noise, given that treating peripheral blood samples with UNG prior to NGS led to a significant reduction in the C:G > T:A mutation frequency. While we favor a biologic source to explain the reduction by UNG treatment, we cannot eliminate the possibility that deamination is induced during DNA isolation. However, the reduction is only approximately 30 % (Fig. 1d), suggesting that the other 70 % of the mutations are already fixed (fully converted to T:A) prior to DNA isolation, or it is occurring during the process of PCR. In this regard, we found that the heat from the denaturation phase of thermocycling induced a significant increase in this background noise, consistent with the prolonged heating used by Ehrlich et al. [19]. To test whether it was solely an artifact of PCR, we attempted to add thermostable UNG from an extreme thermophile organism, Archaeoglobus fulgidus [27], during PCR. However, we were unable to identify conditions where this enzyme and the polymerase were both active (data not shown). An alternative approach may be to use an alternate thermostable UNG [28] or to replace traditional PCR with isothermal amplification.

5 Conclusion

A major cause of baseline noise in NGS is cytosine deamination. This appears to be pre-analytic (i.e., biologic in origin), but it can also be induced by the heat associated with thermocycling. Routine use of UNG pretreatment and isothermal amplification are viable strategies to reduce the background noise level in NGS.