IntroductionSince its introduction in 19851, polymerase chain reaction (PCR) has been extensively utilized in molecular biology for the research and quantitation of DNA and RNA samples2,3. Numerous factors influence the efficiency of PCR amplification, such as primers4, GC-rich regions5, secondary structures6,7, and reagents present in the sample8. The impact of GC-rich regions on PCR amplification efficiency has been widely studied and documented5. GC-rich sequences are generally defined as regions with greater than 65% GC content within a stretch of 200 bp or more9. Various reagents, including dimethyl sulfoxide (DMSO), betaine and 7-deaza-dGTP (7-d-dGTP), and other additives have been employed to mitigate the GC-rich effect10,11,12,13,14,15. However, most studies demonstrated the effect with agarose gels but did not provide quantitative analysis. GC-rich regions affect the efficiency of PCR amplification when located within the amplicon5. The extent to which GC-rich regions affect neighboring amplicons remains unclear. Quantitative PCR is typically conducted using two platforms: real-time quantitative PCR (qPCR)16,17 or digital PCR18. In qPCR, quantitation depends on a standard curve, and the GC-rich effect may vary depending on the standard material employed in the assay, potentially leading to over- or underestimation of results. Digital PCR quantifies nucleic acids by partitioning a sample into thousands of individual reactions and counting positive and negative partitions at endpoint using Poisson statistics, without requiring a standard curve19. Digital PCR offers advantages over qPCR, including lower coefficients of variation20 and greater tolerance to inhibitors21. Among the commercially available platforms22,23, droplet digital PCR (ddPCR, Bio-Rad)24 and nanoplate digital PCR (ndPCR, QIAGEN QIAcuity)25,26 were used in this study. Currently, no published data quantitatively compares the impact of GC-rich regions on qPCR, ddPCR, and ndPCR results. Many vendors offer PCR kits claiming to address the impact of GC-rich sequences on amplification efficiency. However, most reagent components are proprietary, making it necessary to conduct thorough comparison studies to determine which kits contain reagents capable of improving results with GC-rich samples.During the development of multiplex ndPCR methods for rAAV (recombinant adeno-associated virus) genome integrity and impurity characterization27,28, we designed 43 amplicons spanning the ~ 150 kb recombinant baculoviral–rAAV vector genome. When tested on sonicated baculoviral–rAAV DNA (~ 1 kb fragments), all amplicons performed equivalently27. A subset of these amplicons, targeting the ~ 7 kb Tn7L–Tn7R region shared between the baculoviral–rAAV vector and the GFP-6.0 Kb vector plasmid, was then used to cross-validate the absolute copy number of the plasmid. However, these amplicons yielded markedly different copy numbers from the plasmid DNA, and the discrepancy correlated with the proximity of each amplicon to a 300 bp, ~ 79% GC region within the CBA promoter. Here, we systematically characterize this GC-rich proximity effect on neighboring amplicons across ndPCR, qPCR, and ddPCR platforms, and evaluate strategies to mitigate it.ResultsGC-rich region affects amplification efficiency of neighboring amplicons in ndPCRSix amplicons (A–F) spanning the ~ 7 kb Tn7L-Tn7R region of the GFP-6.0 Kb vector plasmid (Fig. 1) were used to determine the absolute copy number of the plasmid DNA, which served as the standard curve material for rAAV genome integrity studies28. These amplicons, designed using IDT’s PrimerQuest tool, had been validated on sonicated template DNA (~ 1 kb fragments), where they produced equivalent copy numbers27. We reasoned that concordant copy number measurements across multiple amplicons would provide a more reliable determination. Figure 1a shows a map of the GFP-6.0 Kb vector plasmid, illustrating the rAAV genome flanked by inverted terminal repeats (ITRs), the restriction sites of the endonuclease enzyme NotI, six amplicons, the CMV enhancer and chicken beta-actin promoter (CBA), and Tn7. All amplicons except amplicon F are in the rAAV region. Figure 1b presents the GC content (%) of each amplicon; all amplicons have GC content ranging from 38 to 61%, below the 65% threshold used to define GC-rich sequences. Figure 1c shows the % GC content distribution across the Tn7L-Tn7R fragment of this vector. Supplementary Table S3 provides the partial sequence of the GFP-6.0 Kb vector plasmid. Published data showed supercoiled plasmid DNA and AAV ITRs affecting the PCR amplification29,30, thus the GFP-6.0 Kb vector plasmid DNA was purposely predigested with restriction endonuclease NotI to avoid these effects. However, although the same amount of plasmid DNA was used for each amplicon, the copy numbers varied substantially among different amplicons. Figure 2a shows a linear map of the region between Tn7L and Tn7R. Figure 2b displays one-dimensional (1D) scatterplots for all six amplicons, each analyzed across six three-fold serial dilutions together with two negative controls. The resulting distributions differed among amplicons, with substantial rain (partitions exhibiting intermediate fluorescence amplitudes between clearly negative and positive clusters) observed for amplicons A-E but not for amplicon F. The density of positive partitions was lower, especially for amplicon A and B compared to amplicon F. Figure 2c presents the quantitative ndPCR results. Amplicon F yielded a concentration of 7.3 × 1013 copies/mL with a 40-cycle amplification program, while amplicon B yielded the lowest concentration at 6.7 × 1012 copies/mL (9% of the amplicon F value). % Amplification efficiency is defined as (amplicon copy number ÷ amplicon F copy number) × 100, where amplicon F serves as the reference. To ensure adequate separation between positive signals and background noise, the amplification was extended by an additional 20 cycles following the initial 40-cycle protocol. Extending to 60 cycles increased the copy numbers 18% to 75% for amplicons A–E, except for amplicon F. After 40 cycles, % amplification efficiency ranged from 9 to 78%. At 60 cycles, these values increased to a range of 15% to 92%. Detailed data are provided in Supplementary Table S5.Fig. 1Full size imagePlasmid map and GC content of the Tn7L to Tn7R fragment including the amplicons used in this study. (a) Circular map of the GFP-6.0 Kb vector plasmid (10,300 bp) showing the positions of the six amplicons A-F (green) used in this study, the CMV enhancer (blue), chicken beta-actin (CBA) promoter (magenta), inverted terminal repeats (ITRs, orange), Tn7L and Tn7R transposon elements, f1 origin of replication (f1 ori), ampicillin resistance gene (AmpR), origin of replication (ori), and two NotI restriction sites (positions 363 and 6081). Amplicons A-E are located within the recombinant adeno-associated virus (rAAV) genome region between the two ITRs, while amplicon F is located outside the rAAV region between the 3ʹ ITR and Tn7R. (b) GC content (%) of each amplicon. All amplicons have GC content ranging from 38 to 61%, below the 65% threshold used to define GC-rich sequences. (c) GC content plot showing the %GC content in a 200 bp sliding window for the Tn7L-Tn7R fragment including the amplicons A-F used in this study. Positions for amplicons A through F are indicated by a green bar, the 65% GC-rich sequence threshold is shown as a blue dotted line.Fig. 2Full size imagendPCR amplification of NotI-digested GFP-6.0 Kb vector plasmid DNA with amplicons A–F. (a) Linear map of the NotI-digested GFP-6.0 Kb vector showing the region between Tn7L and Tn7R (yellow), including the positions of the six amplicons (A–F, green) used in this study, CMV enhancer (blue), chicken beta-actin (CBA) promoter (magenta), inverted terminal repeats (ITRs, orange), and two NotI restriction sites. (b) Representative 1D scatterplots from ndPCR for each amplicon (A–F). Each plot shows six three-fold serial dilutions (decreasing concentration from left to right) plus two negative controls. Positive partitions are shown in blue; the red line indicates the fluorescence threshold. (c) Copy number (copies/mL) for amplicons A–F at 40 cycles (grey bars) and 60 cycles (black bars). Error bars represent s.e.m. Detailed data are provided in Supplementary Table S5.To address potential issues with the primers and probes, several new amplicons (primers/probe pairs, Supplementary Table S2) were designed and synthesized near the original amplicon A and B regions. These new amplicons still demonstrated lower-than-expected amplification efficiency on NotI-digested GFP-6.0 Kb vector plasmid DNA (Fig. 3a). A 2544 bp gBlock DNA fragment containing all the relevant amplicon sequences except for amplicon F was synthesized (Supplementary Table S4). When using the gBlock DNA as a template, all amplicons exhibited similar amplification efficiency (Fig. 3b).Fig. 3Full size imagendPCR copy number determination of NotI-digested GFP-6.0 Kb vector plasmid DNA with additional amplicons and gBlock DNA template. (a) Copy number (copies/mL) of the original amplicons (A–F) and additional amplicons designed near the A and B regions (A-a1, A-a2, B-a1 through B-a4) using NotI-digested GFP-6.0 Kb vector plasmid DNA as template in ndPCR. Grey bars, 40 cycles; black bars, 60 cycles. Error bars represent s.e.m. (b) Copy number (copies/mL) of amplicons A–E using the 2,544 bp gBlock DNA fragment as template. Grey bars, 40 cycles; black bars, 60 cycles. Error bars represent s.e.m. Amplicon F is not included as it is located outside the rAAV region and is not represented in the gBlock DNA. Primer and probe sequences for additional amplicons are provided in Supplementary Table S2; the gBlock sequence is provided in Supplementary Table S4.Amplification efficiency gradually increased from amplicon B to F with the NotI-digested GFP-6.0 Kb vector plasmid DNA, suggesting that the region near amplicon B may affect PCR amplification efficiency. Amplicon B is located close to the CBA promoter. A 300 bp region surrounding the CBA promoter contains a GC content of approximately 79%, well above the 65% threshold for GC-rich sequences and other regions on the Tn7L-Tn7R DNA fragment (Fig. 1c). The presence of this GC-rich region may affect the PCR amplification of adjacent amplicons by proximity, even though these amplicons themselves do not have a high GC content. Primers and probes could not be designed within the GC-rich region (CBA region) of the GFP-6.0 Kb vector using IDT’s online PrimerQuest tool; therefore, the effect of the GC-rich region within the amplicon itself on PCR amplification efficiency could not be evaluated. To further test whether the CBA/GC-rich region was responsible for the reduced amplification efficiency of neighboring amplicons, restriction enzyme digestion was used to physically separate amplicons from the GC-rich region. Figure 4a shows the linear map of the Tn7L-to-Tn7R fragment with regional GC content (%), amplicon positions, and the relative positions of the restriction enzyme sites of NdeI, AfeI, BamHI and NotI. The precise copy number of plasmid DNA was determined by averaging values obtained from multiple amplicons, with adjustments made to minimize the influence of GC-rich regions. The result, 7.5 × 1013 copies/mL, served as the reference for calculating % recovery across all experiments. % Recovery is defined as (mean copies/mL ÷ 7.5 × 1013 copies/mL) × 100. Figure 4b shows that AfeI, which cuts between the GC-rich region and amplicon B reduced the rain observed in the 1D amplitude plots for amplicons B, C, D, and E, while having no effect on amplicon A. Digestion with BamHI, whose restriction site is 3’ of amplicon B, decreased the level of rain for amplicons C, D, and E, but not amplicons A and B, which showed rain similar to the uncut control. The combination of AfeI and NdeI, the latter cutting between amplicon A and the GC-rich region resulted in a reduction of rain across all amplicons A to E (Fig. 4b). Notably, none of the enzymes affected amplification of amplicon F. Gel electrophoresis validation of restriction digests is shown in Supplementary Fig. S1. Figure 4c shows the ndPCR results with AfeI-digested plasmid DNA. Following AfeI digestion, a marked increase in copy numbers was observed for amplicons B, C, D, and E, which attained levels comparable to that of amplicon F. For most amplicons, copy numbers remained stable between 40 and 60 cycles, with the exception of amplicon A. The % recovery for amplicons B-E ranged from 94 to 100%. BamHI-digestion between amplicons B and C resulted in increased amplification efficiency for amplicons C, D, and E, whereas amplicons A and B demonstrated no change (Fig. 4d). Furthermore, double digestion with AfeI and NdeI restored amplification efficiency for all amplicons to expected levels (Fig. 4e), confirming uniform amplification efficiency across the rAAV vector genome within the plasmid. Detailed data are provided in Supplementary Tables S6–S8.Fig. 4Full size imageEffect of restriction enzyme digestion on ndPCR amplification of NotI-digested GFP-6.0 Kb vector plasmid DNA. (a) Linear map of the Tn7L-to-Tn7R fragment (7609 bp) showing the positions of six amplicons (A–F, green), the GC-rich region (79% GC, red), restriction enzyme sites (NotI, NdeI, AfeI, BamHI), ITR (orange), and regional GC content (%, grey). A kilobase scale bar is shown below. (b) Representative 1D scatterplots from ndPCR for amplicons A–F under four digestion conditions: control (NotI only), AfeI, BamHI, and AfeI plus NdeI. Positive partitions are shown in blue. (c) Copy number (copies/mL) for amplicons A–F after AfeI digestion at 40 cycles (grey bars) and 60 cycles (black bars). (d) Copy number (copies/mL) for amplicons A–F after BamHI digestion at 40 cycles (grey bars) and 60 cycles (black bars). (e) Copy number (copies/mL) for amplicons A–F after NdeI plus AfeI double digestion at 40 cycles (grey bars) and 60 cycles (black bars). Error bars in (c–e) represent s.e.m. % Recovery is calculated based on a reference concentration of 7.5 × 1013 copies/mL. Detailed data are provided in Supplementary Tables S6–S8. Gel electrophoresis validation of restriction digests is shown in Supplementary Fig. S1.Effect of additive reagents on ndPCR amplificationVarious additives and certain PCR kits have been used to address the challenges posed by GC-rich sequences during PCR amplification. In the GFP-6.0 Kb vector, amplicon B was most significantly affected by proximity to the GC-rich region; therefore, we utilized amplicon B to assess the efficacy of DMSO, betaine, and reagents from both AmpliTaq Gold 360 DNA Polymerase (AmpliTaq) and KOD Xtreme Hot Start DNA Polymerase (KOD) kits on the amplification of NotI-digested GFP-6.0 Kb vector plasmid DNA. Figure 5a shows a gradual increase in copy numbers with increasing DMSO concentration from 0.38% to 3.0%. The % recovery of amplicon B ranged from 11 to 105% and the copy number showed an increase of 1.2 to 11.8-fold over the result with no DMSO added (Supplementary Table S10a). Addition of DMSO between 1.88% to 3.0% enhanced the amplification efficiency of amplicon B to approximately 100%. To test whether similar efficiency gains can be observed with other reagents known to improve the amplification of GC-rich regions, we repeated the experiments with betaine (Fig. 5b). The amplification efficiency was greatly increased in the presence of 0.19 M to 1.5 M betaine. The % recovery ranged from 34 to 102% and the copy number increased 2.6 to 7.9-fold over the results with no betaine added (Supplementary Table S10b). Greater than 90% or higher % recovery was observed for reactions with betaine added from 0.38 M to 1.5 M, indicating this as the optimum concentration range for this assay. We also tested the amplification effects of the reagents from the AmpliTaq and KOD kits on amplicon B (Fig. 5c and d). The AmpliTaq kit contains three separate reagents: AmpliTaq Taq polymerase, 10 × reaction buffer, and enhancer. When all three AmpliTaq reagents were combined with the QIAGEN QIACuity Probe Master Mix, amplification efficiency increased by 10.6-fold over reactions without the AmpliTaq reagents and the % recovery reached 98% (Fig. 5c, Supplementary Table S10c). The addition of AmpliTaq polymerase alone resulted in a 6.1-fold increase in efficiency, with a 56% recovery rate. The use of the Amp enhancer, ranging from 2.5% to 12.5%, led to an increase in efficiency from 3.0-fold to 10.5-fold, with % recovery exceeding 90% at concentrations between 7.5 and 12.5%. However, the copy number was lower than the no additive control when only the AmpliTaq buffer was added. The KOD kit, which contains a polymerase and buffer mixture, showed that the KOD polymerase alone strongly inhibited amplification (ND, not determined), whereas the buffer mixture resulted in a 10.3-fold increase in efficiency, achieving a 95% recovery rate (Fig. 5d, Supplementary Table S10d). The enhanced amplification efficiency observed in ndPCR due to these additives was also associated with a reduction in rain (Supplementary Fig. S2). Reaction mix compositions are provided in Supplementary Table S9; detailed data analyses are provided in Supplementary Table S10.Fig. 5Full size imageEffect of additives on ndPCR amplification efficiency of amplicon B using NotI-digested GFP-6.0 Kb vector plasmid DNA. (a) Copy number (copies/mL) of amplicon B with DMSO at concentrations ranging from 0% to 3.0%. (b) Copy number (copies/mL) of amplicon B with betaine at concentrations ranging from 0 to 1.5 M. (c) Copy number (copies/mL) of amplicon B with AmpliTaq Gold 360 DNA Polymerase components: no additive, all three components combined (All), buffer only, enzyme only, and enhancer at 2.5% to 12.5%. (d) Copy number (copies/mL) of amplicon B with KOD Xtreme Hot Start DNA Polymerase components: no additive, all components combined (All), buffer only, and enzyme only. ND, not determined (no valid data points met acceptance criteria). Error bars represent s.e.m. All reactions used 40-cycle amplification. Reaction mix compositions are provided in Supplementary Table S9; detailed numerical data are provided in Supplementary Table S10.Effect of thermocycling conditions on the ndPCR amplificationQIAcuity Probe Master Mix utilizes QuantiNova DNA Polymerase, which remains inactive at low temperatures due to the presence of QuantiNova Antibody and QuantiNova Guard. During a 2-min pre-activation (hot start) at 95 °C, both the QuantiNova Antibody and Guard are denatured, resulting in the activation of QuantiNova DNA Polymerase and the initiation of PCR amplification. Available data do not indicate that extended pre-activation times negatively impact performance. The original multiplex ndPCR method was adopted from a ddPCR protocol and optimized for quantification of rBV DNA impurities in rAAV products by in-partition lysis using multiple amplicons. To ensure comprehensive viral particle lysis, a pre-activation phase at 95 °C for 10 min was employed, followed by three-step cycling (denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 30 s) for 40 cycles, with an additional 20 cycles in certain protocols. Three-step cycling with separate annealing and extension temperatures was used because the 43 amplicons in the original multiplex assay27 have different melting temperatures, and a single combined annealing/extension temperature may not be optimal for all amplicons. In contrast, QIAGEN’s recommended protocol consists of a hot start at 95 °C for 2 min, followed by 40 cycles of two-step cycling (95 °C for 15 s and 60 °C for 30 s). Because our original protocol differed from QIAGEN’s recommendation in both pre-activation duration (10 vs. 2 min) and cycling format (three-step vs. two-step), we systematically evaluated four thermocycling conditions to determine which parameter was associated with the reduced amplification efficiency (Fig. 6). An annealing/extension temperature of 56 °C was used in these comparison studies for the two-step conditions, as preliminary study showed that both 56 °C and the recommended 60 °C yielded similar amplification efficiency (Supplementary Fig. S3). The following four conditions were performed: Condition I (10-min pre-activation, three-step cycling; Fig. 6a) resulted in low % recovery for all amplicons except amplicon F, with amplicon B showing the lowest recovery. Condition II (2-min pre-activation, two-step cycling; Fig. 6b) yielded high % recovery for all amplicons. To distinguish between the effects of pre-activation duration and cycling format, two additional conditions were tested: Condition III (2-min pre-activation, three-step cycling; Fig. 6c) and Condition IV (10-min pre-activation, two-step cycling; Fig. 6d). Condition III yielded % recovery of 97–107% for all amplicons, similar to Condition II. Condition IV showed reduced amplification efficiency for amplicons A–E, similar to Condition I. Amplicon F showed consistent amplification efficiency across all four conditions. These results indicate that the pre-activation duration, rather than the cycling format, is the primary factor associated with the reduced amplification efficiency observed in the presence of the GC-rich region. Under Condition II, an edge effect was observed at high template concentrations, where positive partitions were not evenly distributed within wells for amplicons A–E (Supplementary Figs. S4 and S5). The reference dye was uniformly distributed under all conditions, indicating even distribution of reagents among partitions. Under Condition IV, positive partitions exhibited uniform distribution but with lower density for amplicons A–E compared to amplicon F.Fig. 6Full size imageEffect of thermocycling conditions on ndPCR amplification of NotI-digested GFP-6.0 Kb vector plasmid DNA. % Recovery for amplicons A–F at 40 cycles (grey bars) and 60 cycles (black bars) under four thermocycling conditions. (a) Condition I: 10-min pre-activation at 95 °C, three-step cycling (94 °C for 30 s, 55 °C for 30 s, 72 °C for 30 s). (b) Condition II: 2-min pre-activation at 95 °C, two-step cycling (95 °C for 15 s, 56 °C for 30 s). (c) Condition III: 2-min pre-activation at 95 °C, three-step cycling (94 °C for 30 s, 55 °C for 30 s, 72 °C for 30 s). (d) Condition IV: 10-min pre-activation at 95 °C, two-step cycling (95 °C for 15 s, 56 °C for 30 s). Error bars represent s.e.m. ndPCR signal maps for Conditions II and IV are shown in Supplementary Fig. S4; representative well images are shown in Supplementary Fig. S5.Effect of GC-rich region on qPCR amplificationTo evaluate the effect of GC-rich regions on qPCR, amplicons A to F were tested using NotI-digested plasmid DNA as the standard curve to quantify NotI-digested and NotI plus AfeI-digested plasmid DNA. The 7.5 × 1013 copies/mL concentration as described previously was used for the calculation of the % recovery in the qPCR experiments. When NotI-digested plasmid DNA was used as both the standard curve and sample (Fig. 7a), plasmid DNA sample had a recovery rate of 79% to 106% (Supplementary Table S11a). The % recovery of NotI plus AfeI-digested plasmid DNA quantified against the NotI-digested standard curve was 105% and 116% with amplicons A and F, respectively. However, recovery rates for amplicons B, C, D and E varied significantly, ranging from 298 to 651% (Fig. 7b, Supplementary Table S11b). A 2544 bp gBlock DNA fragment containing amplicons A to E was synthesized to use as a standard curve to quantify NotI-digested and NotI plus AfeI-digested plasmid DNA. Figure 7c shows that amplicons A to D had very low recovery with NotI-digested plasmid DNA, while amplicon E recovered 129% (Supplementary Table S12a). When plasmid DNA was digested by NotI plus AfeI, amplicon A still had low recovery, but recovery rates for amplicons B to E were much higher compared to NotI-digested plasmid DNA, especially amplicons B and C, which reached expected levels (Fig. 7d, Supplementary Table S12b). Detailed data including mean Ct values are provided in Supplementary Tables S11 and S12.Fig. 7Full size imageEffect of GC-rich regions on qPCR quantitation using different standard curve materials. % Recovery of amplicons quantified by qPCR using two standard curve sources and two sample conditions. % Recovery is calculated as (mean copies/mL ÷ 7.5 × 1013) × 100. (a) NotI-digested plasmid DNA standard curve, NotI-digested plasmid DNA sample (amplicons A–F). (b) NotI-digested plasmid DNA standard curve, NotI plus AfeI-digested plasmid DNA sample (amplicons A–F). (c) gBlock DNA standard curve, NotI-digested plasmid DNA sample (amplicons A–E). (d) gBlock DNA standard curve, NotI plus AfeI-digested plasmid DNA sample (amplicons A–E). Amplicon F is not included in (c) and (d) as it is located outside the rAAV region and is not represented in the gBlock DNA. Error bars represent s.e.m. (n = 2). Detailed data including mean Ct values and copies/mL are provided in Supplementary Tables S11 and S12.Effect of additives on qPCR amplificationAmplicon B was utilized to assess the effect of additives —DMSO, betaine, and 7-d-dGTP—on qPCR quantitation of NotI-digested plasmid DNA. When NotI-digested plasmid DNA served as the standard curve (Fig. 8a), the addition of DMSO at concentrations ranging from 2 to 6% and betaine from 0.5 M to 1.5 M resulted in increased % recovery, by 6.7 to 8.7-fold and 1.9 to 7.9-fold, respectively. 7-d-dGTP at concentrations from 0.6 mM to 1.4 mM led to a change of 0.9 to 3.8-fold (Supplementary Table S13a).Fig. 8Full size imageEffect of additives on qPCR quantitation of amplicon B using NotI-digested GFP-6.0 Kb vector plasmid DNA. % Recovery of amplicon B with DMSO (2–6%), betaine (0.50–1.50 M), and 7-deaza-dGTP (7-d-dGTP; 0.60–1.40 mM) quantified against three different standard curves. % Recovery is calculated as (mean copies/mL ÷ 7.5 × 1013) × 100. (a) NotI-digested plasmid DNA standard curve. (b) NotI plus AfeI-digested plasmid DNA standard curve. (c) gBlock DNA standard curve. Dashed vertical lines separate the three additive groups. Error bars represent s.e.m. (n = 3). Detailed data including mean Ct values, copies/mL, and fold changes are provided in Supplementary Table S13.When NotI plus AfeI-digested plasmid DNA was used as the standard curve to quantify NotI-digested plasmid DNA (Fig. 8b, Supplementary Table S13b), amplicon B demonstrated a recovery rate of only 9% without additives. The inclusion of DMSO (2% to 6%) raised copy numbers by approximately 6.5 to 8.4-fold, while betaine (0.5 M to 1.5 M) increased the copy number by 1.9 to 7.6-fold, elevating % recovery from 9% to ranges of 59–76% and 17–69%, respectively. The presence of 7-d-dGTP (0.6 mM to 1.4 mM) modified copy numbers by 0.9 to 3.7-fold, resulting in % recovery between 8 and 34%.When a gBlock DNA fragment was employed as the standard curve to quantify NotI-digested GFP-6.0 Kb vector plasmid DNA (Fig. 8c, Supplementary Table S13c), the % recovery without additives was 12%. DMSO (2 to 6%) increased copy numbers by 5.2 to 6.6-fold, achieving recovery rates of 63–81%. Betaine (0.5 M to 1.5 M) enhanced copy numbers by 1.7 to sixfold, corresponding to recovery rates of 21–73%. The use of 7-d-dGTP (0.6 mM to 1.4 mM) resulted in a change of 0.8 to 3.1-fold. Detailed data including mean Ct values, copies/mL, and fold changes are provided in Supplementary Table S13.Effect of GC-rich region on droplet digital PCR (ddPCR) amplificationThe effect of the GC-rich region on ddPCR amplification was evaluated using amplicons A–F with NotI-digested GFP-6.0 Kb vector plasmid DNA. The copy numbers did not differ substantially among amplicons A–F (Supplementary Fig. S7 and Table S14). The GC-rich proximity effect observed in ndPCR was not observed in ddPCR under the conditions tested.DiscussionQuantitative PCR, including qPCR and digital PCR, is commonly used for the quantitative analysis of DNA and RNA samples31. Multiple factors, such as GC-rich regions5, secondary structures6,7, and specific reagents present in samples8, can affect their quantitation. During the development of ndPCR for quantitating rAAV products, we used IDT’s online PrimerQuest tool to design multiple primer/probe pairs (amplicons) optimized for probe-based digital PCR. The GC content of these amplicons ranges from 38 to 61% (Fig. 1b), well below the 65% threshold commonly used to define GC-rich sequences9. However, when we quantitated NotI-linearized plasmid DNA with these amplicons in ndPCR, the copy numbers varied substantially among different amplicons despite using the same amount of plasmid DNA. Amplicon B yielded the lowest copy number, only 9% of the value obtained with amplicon F (Fig. 2c, Supplementary Table S5). We hypothesized that the approximately 300 bp region surrounding the CBA promoter (Figs. 1c and 4a) with 79% GC content may contribute to observed variation in amplification efficiency among amplicons located at different distances from this region. Restriction enzyme digestion experiments supported this hypothesis. AfeI whose recognition site is located between the GC-rich region and amplicon B, restored amplification efficiency to 94–100% recovery for all amplicons on the 3’ side of the AfeI site (amplicons B through E, Supplementary Table S6), while the amplification efficiency of amplicon A, located on the 5′ side of the AfeI site and thus still physically linked to the GC-rich region, remained unchanged (Fig. 4c). Additional digestion with BamHI (Fig. 4d) and double digestion with AfeI plus NdeI (Fig. 4e) further supported the association between the GC-rich CBA promoter region and reduced amplification efficiency of neighboring amplicons, even though the amplicons themselves do not have high GC content. The amplification efficiency decreased progressively from amplicon E to amplicon B. Even amplicon E, which is separated by 4.7 kb from the GC-rich CBA promoter region showed only 78% amplification efficiency relative to amplicon F (Supplementary Table S5). These data suggest that the GC-rich region is associated with reduced PCR amplification of neighboring amplicons in a manner that may be related to distance, although only six spatial positions were tested and formal regression analysis was not performed. To our knowledge, this is the first report describing the proximity effect of GC-rich regions on PCR amplification. When assessing the effect of GC-rich regions on PCR amplification, it is important not only to consider GC content within the amplicon—which can potentially be optimized during primer and probe design—but also to evaluate the presence of GC-rich sequences adjacent to the amplicon. Restriction enzyme digestion of samples prior to PCR analysis has been used to improve PCR amplification efficiency23,32. Our data suggest that the greatest benefit is observed with a restriction enzyme recognition site between the GC-rich region and the amplicon. Sonication of samples has also been shown to enhance PCR amplification efficiency33, which is consistent with our previous findings27. Sonication can be employed to fragment high-molecular-weight DNA into smaller fragments to mitigate issues associated with GC-rich regions during PCR. The DNA extraction strategy should also be considered as a variable, as different extraction methods can produce templates with varying degrees of fragmentation, which may affect whether GC-rich regions remain physically linked to target amplicons.While the restriction enzyme strategy demonstrated in this study required prior knowledge of restriction sites located between the GC-rich region and target amplicons, a more generalizable approach may be possible using frequent-cutting restriction enzymes whose recognition sequences are enriched in GC-rich DNA. For example, HinP1I recognizes the tetranucleotide sequence GCGC, which preferentially occurs in GC-rich regions. In silico analysis of the Tn7L–Tn7R fragment revealed that HinP1I sites are disproportionately concentrated within the CBA/CMV GC-rich region (79% GC), where 11 of the 32 total sites are located, corresponding to an average fragment size of approximately 43 bp upon digestion (Supplementary Table S15). In contrast, the other GC-rich regions (66–68% GC), which were not associated with reduced amplification efficiency, contained only 1–4 HinP1I sites each, and no HinP1I sites were present within amplicons A–E (38–61% GC). Because GCGC motifs are inherently enriched in highly GC-dense sequences, this concept may be applicable to other high-GC regions, including mammalian CpG islands and gene therapy vector promoters. However, because HinP1I is sensitive to CpG methylation, its utility for CpG-methylated mammalian genomic DNA may be limited, and amplicons would need to be screened to ensure the absence of internal recognition sites. Nevertheless, for the purpose of mitigating GC-rich proximity effects through disruption of long-range DNA structure, complete digestion is not required, and partial cleavage at a subset of sites may remain effective even when some CpG sites are methylated. This computationally derived concept warrants future experimental evaluation and represents a potential direction toward developing a more generalizable enzymatic strategy to mitigate GC-rich proximity effects.Rain formation must be taken into account during the development of ndPCR34. Our data showed that reduced amplification efficiency near GC-rich regions produced significant rain in ndPCR. Comparing amplicon E with amplicon F (Fig. 2b and c) illustrates that the effect on amplification efficiency may not be apparent without a reference amplicon for comparison. When amplification efficiency is optimal, copy numbers remain stable between 40 and 60 cycles; an increase in copy number with additional cycles may indicate suboptimal amplification (Fig. 2). This pattern was reproducible across experiments using different restriction enzyme digestions (Figs. 3a, 4c,d) and thermocycling conditions (Fig. 6). During the development of ndPCR, testing at least two cycle sets (for example, 40 cycles and 60 cycles) is recommended based on our observations. This approach can be employed to assess amplification efficiency and address rain-related concerns in ndPCR.The effect of GC-rich sequences on qPCR is complex because the outcome depends on the relationship between the standard curve material and the sample, when a standard curve is utilized in quantitation. If the standard material and samples share similar GC-rich characteristics, the measured copy number tends to approximate the expected value (Fig. 7a, Supplementary Table S11a). However, when neighboring GC-rich regions differ between standard and sample, quantification becomes inaccurate. For example, using NotI-digested plasmid DNA as the standard to quantify NotI plus AfeI-digested samples led to substantial overestimation (298–651% recovery; Fig. 7b, Supplementary Table S11b), because AfeI digestion mitigated the GC-rich effect in the sample but not in the standard. A gBlock DNA fragment can serve as an alternative standard curve source; however, because the GC-rich region was removed during gBlock synthesis, the gBlock may not accurately represent the original sample. When gBlock DNA was used as the standard curve, amplicons A to D showed low % recovery with NotI-digested plasmid DNA (Fig. 7c, Supplementary Table S12a), while amplicon E showed 129% recovery. Digestion of the plasmid DNA with AfeI increased the copy numbers for amplicons B and C to expected values (Fig. 7d, Supplementary Table S12b), as this digestion reduced the GC-rich effect for these amplicons. Amplicon D showed low amplification efficiency even after plasmid DNA was digested with AfeI. This may be due to intrinsic sequence characteristics of amplicon D, which showed a higher Ct value than other amplicons in qPCR (Supplementary Table S11a) and lower positive droplet amplitude in ddPCR (Supplementary Fig. S8). However, this did not affect quantitation in ndPCR or ddPCR but may have a greater effect on qPCR accuracy. The reason for the high % recovery of amplicon E (129%) with the gBlock standard and NotI-digested plasmid DNA (Fig. 7c) remains unclear. Our data quantitatively illustrate the effect of GC-rich region on the qPCR quantitation. The effect is dependent on the relationship between standard material and sample.Multiple reagents and kits have been reported or developed to address the GC-rich effect. DMSO reduced the GC-rich effect in ndPCR in a dose-dependent manner. In ndPCR, the amplification efficiency reached 99 to 105% recovery at DMSO at 1.88% to 3.0% (Fig. 5a, Supplementary Table S10a) compared to only 9% recovery without additives. Betaine at 0.75 M to 1.5 M achieved % recovery of 96% to 102% (Fig. 5b, Supplementary Table S10b). DMSO and betaine are known to destabilize secondary structures and reduce the melting temperature of GC-rich sequences5,12,13. They may also disrupt the interaction between GC-rich regions and neighboring amplicons. As certain additives can inhibit PCR amplification at elevated concentrations, it is advisable to use DMSO within the range of 1.88–3.0% and betaine between 0.75 and 1.5 M, or to optimize concentrations for specific assays, to effectively manage GC-rich templates in ndPCR. The enhancer from the AmpliTaq kit or the buffer from the KOD kit also reduced the GC-rich effect (Fig. 5c and d). Additionally, the Taq polymerase in the AmpliTaq kit appeared to partially reduce the GC-rich effect, suggesting that certain Taq polymerases may perform differently with respect to GC-rich templates. In qPCR, the addition of 3% to 5% DMSO increased recovery rates, achieving over 70% recovery (Fig. 8, Supplementary Table S13b and c). The use of betaine at concentrations of 0.75 to 1.0 M resulted in recovery improvements exceeding 60%. In contrast, 7-d-dGTP provided only a modest increase in amplification efficiency.Hot start Taq polymerases are widely employed in PCR applications, utilizing various activation mechanisms and hot start incubation periods that typically range from 30 s to approximately 15 min35,36,37,38,39. In our study, we developed assays for the direct quantification of AAV samples without pre-lysis, necessitating a preheating step (pre-activation) at 95 °C for 10 min to ensure complete lysis of viral particles in the partitions. A pronounced GC-rich effect was observed under these conditions (Fig. 6a). However, when the hot start time was reduced to 2 min instead of 10 min, the GC-rich effect was not observed (Fig. 6b and c). A pronounced GC-rich effect was also observed with a 5-min pre-activation (Supplementary Fig. S6). This change could not be attributed solely to loss of overall enzyme activity during longer pre-activation, as restriction enzyme digestion corrected the effect. Extended pre-activation may result in complete denaturation of double-stranded DNA, potentially facilitating interactions between GC-rich regions and neighboring amplicon sequences within the denatured single-stranded template. This could subsequently affect the amplification efficiency of Taq polymerase. Under a brief 2-min pre-activation, GC-rich sequences still affected the spatial distribution of positive partitions at high template concentration, as a pronounced edge effect was observed (Supplementary Figs. S4 and S5). In this study, the use of six three-fold serial dilutions combined with predefined acceptance and linearity criteria enabled identification and exclusion of affected data points. However, in routine digital PCR applications where samples are often analyzed at one or two dilutions, edge effects at high template concentrations can cause data points to fail acceptance criteria such as minimum partition counts, requiring reruns, or produce large variability between dilutions. Nevertheless, the underlying cause of the edge effect warrants further investigation. In separate studies, amplicons with GC content of 66–72% exhibited lower-than-expected amplification efficiency even with a 2-min pre-activation, and also produced pronounced edge effects at high sample concentrations (unpublished data, Wang et al.). This suggests that a shorter pre-activation duration does not eliminate the GC-rich effect but may mask it. When the GC-rich sequence is located within the amplicon itself rather than in a neighboring region, the 2-min pre-activation alone is insufficient to resolve the amplification challenge. Furthermore, the impact of GC-rich sequences within an amplicon may differ mechanistically from their proximity effect on neighboring amplicons. In contrast, additives such as DMSO and betaine, as well as physical separation of GC-rich regions by restriction enzyme digestion or sonication, mitigated the effect regardless of whether the GC-rich sequence was within or adjacent to the amplicon. These approaches may therefore represent more broadly applicable solutions for managing GC-rich templates in PCR-based quantitation.The GC-rich proximity effect observed in ndPCR was not observed in ddPCR under the conditions tested (Supplementary Fig. S7 and Table S14). The Bio-Rad ddPCR master mix contains proprietary components whose composition is not fully disclosed. Given that certain additives and polymerases substantially reduced the GC-rich effect in our ndPCR experiments (Fig. 5), it is plausible that one or more proprietary components in the ddPCR master mix may similarly mitigate the effect. This observation does not necessarily indicate that ddPCR is inherently immune to the GC-rich proximity effect; rather, the master mix formulation may compensate for it.Figure 1c reveals six GC-rich regions (> 65%) across the Tn7L–Tn7R fragment. Despite the presence of multiple GC-rich regions linked to amplicons A–F, our data consistently showed that only the CBA/CMV promoter region, which peaks at 79% GC over approximately 300 bp, was associated with reduced amplification efficiency of neighboring amplicons. The remaining GC-rich regions, including the 5′ ITR, enhanced green fluorescent protein (EGFP), Woodchuck Hepatitis Virus Posttranscriptional Regulatory Element (WPRE), and 3′ ITR, all have peak GC content in the range of 66–68% and did not measurably affect the amplification of adjacent amplicons. This observation suggests that the GC-rich proximity effect may require GC content substantially above the 65% threshold—potentially exceeding 70%—a level that is commonly reached in mammalian promoter regions used in gene therapy vectors and in CpG islands associated with endogenous genes. Notably, amplicon F is located approximately 300 bp from the 3′ ITR GC-rich region yet showed no reduction in amplification efficiency, whereas amplicon B, at a similar distance from the CBA/CMV region, yielded only 9% recovery. This contrast suggests that the proximity effect is dependent not only on distance but also on the characteristics of the GC-rich region. The contrasting behavior of the 3′ ITR and CBA/CMV GC-rich regions may reflect differences in both GC content and structural conformation. Whether stable secondary structures such as ITR hairpins limit the GC-rich effect to immediately adjacent sequences, while unstructured linear GC-rich regions exert a longer-range effect, remains to be investigated.This study has several limitations. The findings are based on a single plasmid model (GFP-6.0 Kb vector), and generalization to other sequence contexts should be made with caution. The data demonstrate an association between proximity to a GC-rich region and reduced amplification efficiency; however, the contribution of other potential confounders—including secondary structures, palindromic sequences, hairpins, and local melting constraints—cannot be excluded. An in silico secondary structure analysis was not performed and is acknowledged as a limitation. The generation of derivative constructs with varying GC content inserted into the gBlock may provide stronger evidence for causality and represents an important direction for future research. Additionally, the effect of cycling conditions on qPCR was not investigated in this study and remains an area for further exploration.In summary, this study provides a quantitative assessment of the association between neighboring GC-rich regions and reduced amplification efficiency in both qPCR and ndPCR. The quantitation was affected not only by the GC content of the amplicon itself, but also by its proximity to a GC-rich region. While the effect of GC-rich sequences within an amplicon is readily identified during primer and probe design, the influence of neighboring GC-rich regions on amplification efficiency is not apparent and may go unrecognized. Notably, while the GC-rich proximity effect in ndPCR produces visible indicators such as rain that may alert the user to suboptimal amplification, the same effect in qPCR may go entirely undetected, as the standard curve and amplification plots may appear normal despite producing inaccurate quantitation. This is particularly important when synthetic DNA fragments such as gBlock are used as standard curve materials, as the absence of neighboring GC-rich regions in the synthetic standard may lead to inaccurate quantitation of samples containing such sequences. We identified practical approaches to mitigate this effect, including additives (DMSO at 1.88–3.0% or betaine at 0.75–1.5 M, or optimizing concentrations for specific assays), restriction enzyme digestion, and sonication to physically separate GC-rich regions from target amplicons. Among these, DMSO and betaine may offer the most broadly applicable solution, as they do not require prior knowledge of neighboring sequences and can be readily incorporated into existing qPCR and ndPCR workflows. Although this study used an rAAV vector plasmid as the model system, GC-rich regions are prevalent in many mammalian promoters, enhancers, and other gene regulatory elements40, and the proximity effect described here may be relevant to a wide range of PCR-based quantitation applications beyond gene therapy vector characterization. Validation in additional sequence contexts and template systems will be important to determine the generalizability of these observations.Materials and methodsReagents and instrumentsAfeI (New England BioLabs, R0632S); FastDigest BamHI (Thermo Fisher FD0054); FastDigest NdeI (Thermo Fisher, FD0584); DNA Suspension buffer pH 8.0 (Teknova, T0223); Poly (A) (Sigma/Roche, 10108626001) QIAcuity Probe Master Mix (QIAGEN, 250102); QIAcuity Nanoplate 8.5 96 well (QIAGEN, 250021); KOD Xtreme Hot Start DNA polymerase kit (EMD Millipore Corp, 71975-3); AmpliTaq Gold 360 DNA polymerase kit (Applied Biosystems, 4398820); QIAcuity Eight (QIAGEN). Primers and probes were synthesized by Integrated DNA Technologies (IDT). The locations of all amplicons are shown in Fig. 1. All sequences are listed in Supplementary Table S1 and S2. Primer stocks were prepared at 200 µM, and probe stocks were prepared at 100 µM in TE buffer. A 20 × primer and probe mixture was made by mixing primers at a concentration of 16 µM and probes at 8 µM (final concentration of the stock mixture).ndPCRRestriction enzyme digestion was performed following the vendors’ instructions. Plasmid DNA (designed internally and manufactured by Aldevron) was diluted in PCR dilution buffer (DNA suspension buffer pH 8.0, 0.1% Pluronic F80, and 100 ng/mL Poly (A)). The lowest final dilution factor for the plasmid DNA at around 1 mg/mL was 12,000,000 (final) following a threefold serial dilution with a total of six dilution points and two negative controls (NTC). The ndPCR mixture included 50 µL of QIAcuity Probe Master Mix, 90 µL H2O, and 10 µL primer/probe mixture, sufficient for 8 wells with 12 µL each. The ndPCR mixture setups of the additive treatments are listed in the Supplementary Table S9. Transfers of 4 µL of diluted plasmid DNA and 12 µL of ndPCR mixture were made to a new 96-well low DNA-binding plate, mixed well, and then 13 µL was transferred to a 96-well QIAcuity Nanoplate and sealed. The ndPCR was performed on a QIAcuity Eight instrument with the program: 95 °C for 10 min (1 cycle), followed by 40 cycles of 94 °C for 30 s, 55 °C for 30 s, and 72 °C for 30 s. The ndPCR plate was imaged with the default settings (Green channel, exposure at 500 ms and gain at 6). For some ndPCR plates, thermocycling was performed with additional 20 cycles (total of 60 cycles) of 94 °C for 30 s, 55 °C for 30 s, and 72 °C for 30 s and imaged again. For the thermocycling comparison studies, the conditions are described in Fig. 6.ndPCR data analysisData were analyzed using a standardized Excel template with predefined acceptance criteria. Thresholds were set above negative partitions and below positive partitions using NTCs as references to determine copies/µL for each well. The following acceptance criteria were applied: (1) total partitions per well ≥ 6,000; (2) positive partitions ≥ 50; (3) negative partitions ≥ 500; and (4) NTC ≤ 5 copies per well. For each amplicon, dilution-corrected concentrations were calculated for all wells meeting the above criteria. To assess linearity within each dilution series, the dilution-corrected concentration (copies/mL) of each qualifying well was compared to a reference value. The reference was defined as the highest dilution-corrected concentration among wells meeting all three partition criteria. If the reference point caused the majority of remaining qualifying data points to fall below 70% of the reference value, the reference point was considered an outlier and the next highest qualifying data point was used instead. This process was repeated until a suitable reference was identified. Data points with dilution-corrected concentrations ≥ 70% of the reference value were included in the final calculation. The reported concentration for each amplicon was determined by averaging the dilution-corrected values of all data points meeting both the partition and linearity criteria. Descriptive statistics reported include the mean, standard deviation (SD), coefficient of variation (%CV), and the number of accepted dilution points (n). Error bars in all figures represent the standard error of the mean (s.e.m.). % Amplification efficiency was calculated as (amplicon copy number ÷ amplicon F copy number) × 100 for the initial observation experiment (Fig. 2c, Supplementary Table S5). % Recovery was calculated as (mean copies/mL ÷ 7.5 × 1013 copies/mL) × 100 for all restriction enzyme digestion and additive experiments.qPCRReagents and Instruments: AfeI (New England BioLabs, R0632S), 10X TE Buffer (Teknova, T3457), TaqMan Fast Advanced Master Mix (Applied Biosystems, 4444557), MicroAmp Fast Optical Reaction Plate (Applied Biosystems, 4346907), 7500 Fast Real-Time PCR System (Applied Biosystems), gBlock (IDT, Supplementary Table S4). Probe and primer stocks for qPCR were identical to those described above for ndPCR. A 20 × probe and primer mixture was prepared by mixing primers to 500 nM and probes to 250 nM (concentration in final qPCR reaction).Restriction enzyme digestion on plasmid DNA was performed following the vendor’s instructions. After digestion, plasmid DNA and gBlock were diluted in a 10 × series in 1X TE buffer to prepare a 4-point standard curve from 7.5 × 106 copies/µL to 7.5 × 103 copies/µL along with no template controls (NTC). The qPCR mix contained 15 µL of 2 × master mix, 1.5 µL of 20 × probe/primer mix, 3.5 µL nuclease-free water, and 10 µL of template DNA per reaction. The final concentrations of additive used in the additive treatment experiments are shown in Fig. 8. Total volume of 30 µL per reaction was loaded into duplicate wells of a reaction plate and run with the following thermocycling conditions: 95⁰C for 20 s (activation), then 40 cycles of 95 °C for 3 s (denature) and 60 °C for 30 s (extend/anneal). Data were collected on the FAM channel for all wells.Data acceptance criteria included PCR efficiency of 90–110%, R2 ≥ 0.98, and NTC wells must produce no amplification or a Ct value greater than the lowest standard. For data analysis, sample results were obtained by interpolation from the standard curve, then dilution corrected to obtain the final concentration value used for calculations.ddPCRDroplet digital PCR (ddPCR) was performed using the QX200 Droplet Digital PCR System (Bio-Rad Laboratories). Plasmid DNA was diluted in PCR dilution buffer (DNA suspension buffer pH 8.0, 0.1% Pluronic F80, and 100 ng/mL Poly (A)) to approximately 4 × 106-fold, then further serially diluted threefold for a total of seven dilution points and one negative control (NC). The ddPCR reaction mix (22 µL total volume) consisted of ddPCR Supermix for Probes (No dUTP) (Bio-Rad, 1863024), primers at a final concentration of 900 nM and probe at 250 nM, and template DNA. Droplets were generated using an Automated Droplet Generator (Bio-Rad). Following droplet generation, samples were amplified on a C1000 Touch Thermal Cycler (Bio-Rad) with the following thermocycling conditions: 95 °C for 10 min, then 40 cycles of 94 °C for 30 s and 60 °C for 60 s, followed by 98 °C for 10 min and a final 30-min hold at 12 °C. After amplification, droplet analysis was performed with a QX200 droplet reader and QX Manager software. Thresholds were set manually for each primer/probe set based on the amplitude of the negative droplet population. The following acceptance criteria were applied: (1) Total accepted droplet counts per well ≥ 10,000; (2) Positive Droplets ≥ 100; (3) Negative Droplets ≥ 100; (4) Negative Control wells have a concentration of ≤ 5 copies/µL.Use of AI assistanceA large language model (Claude, Anthropic) was used to assist with manuscript revision, including editing and refining text, and drafting figure legends. All AI-generated content was reviewed, verified, and approved by the authors, who take full responsibility for the accuracy and integrity of the work.Data availabilityAny data not shown or detailed in the manuscript can be requested from the corresponding authors. Any non-proprietary questions will be addressed if possible.ReferencesSaiki, R. K. et al. Enzymatic amplification of beta-globin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science 230(4732), 1350–1354 (1985).Article ADS PubMed Google Scholar Zhu, H. et al. PCR past, present and future. Biotechniques 69(4), 317–325 (2020).Article PubMed PubMed Central Google Scholar Laura, S. L. & Rukmanidevi, D. A review of PCR, principle, and its applications. Int. J. Pharm. Res. Appl. 8(3), 3502–3513 (2023).Google Scholar Bustin, S. & Huggett, J. qPCR primer design revisited. Biomol. Detect. Quantif. 14, 19–28 (2017).Article PubMed PubMed Central Google Scholar Mamedov, T. G. et al. A fundamental study of the PCR amplification of GC-rich DNA templates. Comput. Biol. Chem. 32(6), 452–457 (2008).Article PubMed PubMed Central Google Scholar Liu, Z. et al. Disruptors, a new class of oligonucleotide reagents, significantly improved PCR performance on templates containing stable intramolecular secondary structures. Anal. Biochem. 624, 114169 (2021).Article PubMed Google Scholar Liu, Z. et al. Transient stem-loop structure of nucleic acid template may interfere with polymerase chain reaction through endonuclease activity of Taq DNA polymerase. Gene 764, 145095 (2021).Article PubMed Google Scholar Uchii, K. et al. Comparison of inhibition resistance among PCR reagents for detection and quantification of environmental DNA. Environ. DNA 1(4), 359–367 (2019).Article Google Scholar Hube, F., Reverdiau, P., Iochmann, S. & Gruel, Y. Improved PCR method for amplification of GC-rich DNA sequences. Mol. Biotechnol. 31(1), 81–84 (2005).Article PubMed Google Scholar Yang, Z. et al. Enhancement effects and mechanism studies of two Bismuth-based materials assisted by DMSO and Glycerol in GC-rich PCR. Molecules https://doi.org/10.3390/molecules28114515 (2023).Article PubMed PubMed Central Google Scholar Elias, M. et al. Evolution of organic solvent-resistant DNA polymerases. ACS Synth. Biol. 12(11), 3170–3188 (2023).Article PubMed Google Scholar Karunanathie, H. et al. PCR enhancers: Types, mechanisms, and applications in long-range PCR. Biochimie 197, 130–143 (2022).Article PubMed Google Scholar Musso, M. et al. Betaine, dimethyl sulfoxide, and 7-deaza-dGTP, a powerful mixture for amplification of GC-rich DNA sequences. J. Mol. Diagn. 8(5), 544–550 (2006).Article PubMed PubMed Central Google Scholar Varadharajan, B. & Parani, M. DMSO and betaine significantly enhance the PCR amplification of ITS2 DNA barcodes from plants. Genome 64(3), 165–171 (2021).Article PubMed Google Scholar Xie, B. et al. Sweet enhancers of polymerase chain reaction. PLoS ONE 19(10), e0311939 (2024).Article PubMed PubMed Central Google Scholar Harshitha, R. & Arunraj, D. R. Real-time quantitative PCR: A tool for absolute and relative quantification. Biochem. Mol. Biol. Educ. 49(5), 800–812 (2021).Article PubMed Google Scholar VanGuilder, H. D., Vrana, K. E. & Freeman, W. M. Twenty-five years of quantitative PCR for gene expression analysis. Biotechniques 44(5), 619–626 (2008).Article PubMed Google Scholar Quan, P. L., Sauzade, M. & Brouzes, E. dPCR: A technology review. Sensors https://doi.org/10.3390/s18041271 (2018).Article PubMed PubMed Central Google Scholar Tan, L. L. et al. Current commercial dPCR platforms: Technology and market review. Crit. Rev. Biotechnol. 43(3), 433–464 (2023).Article MathSciNet PubMed Google Scholar Kojabad, A. A. et al. Droplet digital PCR of viral DNA/RNA, current progress, challenges, and future perspectives. J. Med. Virol. 93(7), 4182–4197 (2021).Article PubMed PubMed Central Google Scholar Lindner, L. et al. Reliable and robust droplet digital PCR (ddPCR) and RT-ddPCR protocols for mouse studies. Methods 191, 95–106 (2021).Article PubMed Google Scholar Ganova, M. et al. Multiplexed digital polymerase chain reaction as a powerful diagnostic tool. Biosens. Bioelectron. 181, 113155 (2021).Article PubMed Google Scholar Gross, M. et al. Comparing the precision of two digital PCR applications for copy number comparisons in protists. Sci. Rep. 15(1), 27095 (2025).Article ADS PubMed PubMed Central Google Scholar Hou, Y. et al. Droplet-based digital PCR (ddPCR) and its applications. TrAC Trends Anal. Chem. https://doi.org/10.1016/j.trac.2022.116897 (2023).Article Google Scholar Doganturk, Y. E., Dag-Guzel, A. & Kuskucu, M. A. Development of a nanoplate-based digital PCR test method for quantitative detection of Human Adenovirus DNA. Infect. Dis. Clin. Microbiol. 5(4), 353–366 (2023).Article PubMed PubMed Central Google Scholar Villa, C., Costa, J. & Mafra, I. First nanoplate digital PCR method to trace allergenic foods: Improved sensitivity for the detection of sesame. Food Chem. 444, 138650 (2024).Article PubMed Google Scholar Wang, X. et al. Characterization and quantitation of baculoviral DNA in rAAV vectors produced in Sf9 cells. Mol. Ther. Methods Clin. Dev. 33(4), 101569 (2025).Article PubMed PubMed Central Google Scholar Wang, X. et al. Development of high-throughput analytical methods for characterization and quantitation of rAAV genome integrity. Mol. Ther. Methods Clin. Dev. 33(4), 101579 (2025).Article PubMed PubMed Central Google Scholar Datta, S. et al. Enhancement of PCR detection limit by single-tube restriction endonuclease-PCR (RE-PCR). Mol. Diagn. Ther. 20(3), 297–305 (2016).Article PubMed Google Scholar Duong, T. et al. Three-dimensional linkage analysis with digital PCR for genome integrity and identity of recombinant adeno-associated virus. Sci. Rep. 15(1), 2154 (2025).Article ADS PubMed PubMed Central Google Scholar Zhang, H. et al. Quantitative or digital PCR? A comparative analysis for choosing the optimal one for biosensing applications. TrAC Trends Anal. Chem. https://doi.org/10.1016/j.trac.2024.117676 (2024).Article Google Scholar Kashyap, S. et al. Brief restriction endonuclease digestion of genomic DNA improves PCR amplifiability and reproducibility of SSR loci: Augmenting exploration of hidden genetic variability in an endemic carp of South-East Asia. Gene Rep. https://doi.org/10.1016/j.genrep.2023.101796 (2023).Article Google Scholar Veal, C. D. et al. A mechanistic basis for amplification differences between samples and between genome regions. BMC Genomics 13, 455 (2012).Article PubMed PubMed Central Google Scholar Berden, P. et al. Amplification efficiency and template accessibility as distinct causes of rain in digital PCR: Monte Carlo modeling and experimental validation. Anal. Chem. 94(45), 15781–15789 (2022).Article ADS PubMed Google Scholar Lu, S. et al. An alternative hot start PCR method using a nuclease-deficient ExoIII from Escherichia coli. Mol. Biotechnol. 61(12), 938–944 (2019).Article PubMed Google Scholar Rutschke, N. et al. Hot start reverse transcriptase: An approach for improved real-time RT-PCR performance. J. Anal. Sci. Technol. 6(1), 20 (2015).Article PubMed PubMed Central Google Scholar Birch, D. E. Simplified hot start PCR. Nature 381(6581), 445–446 (1996).Article ADS PubMed Google Scholar Scalice, E. R., Sharkey, D. J. & Daiss, J. L. Monoclonal antibodies prepared against the DNA polymerase from thermus aquaticus are potent inhibitors of enzyme activity. J. Immunol. Methods. 172(2), 147–163 (1994).Article PubMed Google Scholar Dang, C. & Jayasena, S. D. Oligonucleotide inhibitors of Taq DNA polymerase facilitate detection of low copy number targets by PCR. J. Mol. Biol. 264(2), 268–278 (1996).Article PubMed Google Scholar Hapgood, J. P., Riedemann, J. & Scherer, S. D. Regulation of gene expression by GC-rich DNA cis-elements. Cell Biol. Int. 25(1), 17–31 (2001).Article PubMed Google Scholar Download referencesAcknowledgementsWe thank Miles Armijos for technical support on ndPCR.Author informationAuthors and AffiliationsLilly Research Laboratories, Eli Lilly & Co., Lilly Technology Center, Indianapolis, IN, 46221, USAXushan Wang, Jeff Kruchkow, Anthony Rodríguez-Vargas, Sabine Wenzel & Sarah M. RicherAuthorsXushan WangView author publicationsSearch author on:PubMed Google ScholarJeff KruchkowView author publicationsSearch author on:PubMed Google ScholarAnthony Rodríguez-VargasView author publicationsSearch author on:PubMed Google ScholarSabine WenzelView author publicationsSearch author on:PubMed Google ScholarSarah M. RicherView author publicationsSearch author on:PubMed Google ScholarContributionsConceptualization: X.W. and S.M.R., Data curation: X.W., J.K. and S.W., Formal analysis: X.W., J.K., A.R.-V. and S.W., Investigation: X.W., J.K. and S.W., Methodology: X.W., Supervision: S.M.R., Visualization: X.W., J.K., A.R.-V. and S.W., Writing –original draft: X.W., Writing-review & editing: X.W., J.K., A.R.-V., S.W. and S.M.R.Corresponding authorsCorrespondence to Xushan Wang or Sarah M. Richer.Ethics declarationsCompeting interestsThe authors declare no competing interests.Additional informationPublisher’s noteSpringer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.Supplementary InformationRights and permissionsOpen Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.Reprints and permissionsAbout this article