TT(N)mGCCTC inhibits archaeal family B DNA polymerases

The proofreading activity of the archaeal family B DNA polymerases enables PCR with high fidelity. However, thermostable proofreading DNA polymerases occasionally failed to amplify target fragment that could be amplified by Taq DNA polymerase. We have previously showed that G-rich sequences, which form G-quadruplex, can bind to and inhibit proofreading DNA polymerases. Here we showed that single-stranded oligonucleotides containing sequences of TT(N)mGCCTC can bind and inhibit archaeal family B DNA polymerases but not Taq DNA polymerase. It is very likely that TT(N)mGCCTC inhibits thermostable DNA polymerases during PCR in a single-stranded form. To the best of our knowledge, this is the first example of DNA sequence that could inhibit DNA polymerase in its single-stranded form.


Results
Inhibitory effect of a primer caused PCR failure using PrimeSTAR GXL DNA polymerase. To clone the promoter region of Olig2 from mouse genomic DNA, we failed to amplify the 2.5 kb fragment using Olig2.5 and OligTSSR as primers (all the oligonucleotides are listed in Table 1) and PrimeSTAR GXL DNA polymerase (GXL DNAP). When we used Olig2. 6 and Olig2F instead of Olig2.5, we can successfully amplify the 2.6 kb and 2 kb fragments, respectively (Fig. 1B). The result suggested that the use of Olig2.5 caused the PCR failure. We then used LA Taq and Taq DNAPs instead of GXL DNAP to amplify these fragments. Both LA Taq and Taq can be used to successfully amplify the 2.6 kb, 2.5 kb and 2 kb fragments (Fig. 1B). The results showed that the PCR failure caused by Olig2.5 is specific to GXL, which also indicated that PCR failure caused by Olig2.5 is not due to inefficient primer-template annealing.
We proposed that Olig2.5 may have inhibitory effect to GXL DNA polymerase. To test this hypothesis, we used the primers Olig2F and OligTSSR to amplify the 2 kb fragment with an additional primer of Olig2.6 or Olig2.5. We found Olig2. 6 GCTCTTACGCGTGCTAGCGAAAAGTATCTCTCGGACCAAGAAG − Bio-Olig2.5 TACCGAGCTCTTACGCGTGCCTCTGACTTGCCTCAGAGC(5′biotin) + Table 1. Information of oligonucleotides. 1 The inhibitory sequence is underlined and bolded. 2 The spaces are represent of deleted bases in the sequence. 3   that adding of Olig2.6 doesn′'t inhibit the amplification of the 2 kb fragment, whereas adding of Olig2.5 blocked the amplification of the 2 kb fragment with GXL but not LA (Fig. 1C), supporting that Olig2.5 has an inhibitory effect to GXL DNAP. To further confirm the inhibitory effect of Olig2.5 to GXL DNA polymerase, we used a pair of primers Ubl35armF and Ubl35armR (Table 1) to amplify a 4.5 kb fragment unrelated to the 2 kb fragment amplified by Olig2F and OligTSSR. We found a similar inhibition of the PCR when Olig2.5 was added into the reaction (Fig. 1D).
To test if the inhibitory effect of Olig2.5 is dose dependent, we used Olig2F and OligTSSR and GXL DNAP to amplify the 2 kb fragment, and various concentrations of Olig2.5 was added into the PCR reactions. We found that the inhibitory effect is decreased when less amount of Olig2.5 is used (Fig. 1E). The inhibitory effect of Olig2.5 required a concentration higher than 0.05 μM (Fig. 1E). We thus raised a question that can we use the Olig2.5 at lower concentrations to decrease its inhibitory effect but keep its priming efficiency high enough? To answer this question, we used 0.4 μM of OligTSSR and various concentration of Olig2.5 to amplify the 2.5 kb fragment. The amplification of the 2.5 kb fragment is inefficient when Olig2.5 was used at the concentration of 0.2 μM. When the concentration of Olig2.5 was reduced to 0.1 μM, the 2.5 kb target was visible, and the amplification efficiency was further increased when the concentration of Olig2.5 was reduced to 0.05 μM (Fig. 1F). The dose-dependent inhibition of Olig2.5 to GXL was further confirmed by real-time PCR (Fig. 1G), which showed that Olig2.5 significantly inhibit PCR at concentrations higher than 0.05 μM.

Primer Olig2.5 inhibits family B DNA polymerases. Taq and GXL belong to family A and family B
DNA polymerases, respectively. LA is a Taq based enzyme mixture supplemented with a low level of a family B DNA polymerase. We asked if Olig2.5 could inhibit other family B DNAPs. Using bacterial culture harbouring the cloned 2.6 kb Olig2 promoter fragment as template DNA, Taq and LA can amplify the specific product and a non-specific product. The additional primer Olig2.5 doesn't inhibit the amplification of the specific and non-specific products by Taq and LA (Fig. 2). Without Olig2.5, the specific product was successfully amplified by Q5, PS, GXL, Phusion, Cobuddy and KOD. Different from Taq and LA, when Olig2.5 was added to the PCR reaction, the amplification of the specific and non-specific products by Q5, PS, GXL, Phusion, Cobuddy and KOD were inhibited (Fig. 2). The results showed that the Olig2.5 is inhibitory to at least four family B DNA polymerases, Q5, PS (GXL is derived from PS by proprietary point mutation and PCNA enhancement), Cobuddy and KOD, suggesting that other family B DNA polymerases may be inhibited as well.

TT(N)mGCCTC sequences caused inhibitory effect.
To narrow down the sequence that causing the inhibitory effect to the family B DNAPs, we tested two oligonucleotides truncated from Olig2.5 to see if they can inhibit PCR (please see Table 1 for sequence information and PCR inhibitory effect of the tested oligonucleotides). The result showed that deletion of 8 nucleotides from the 5′ end of Olig2.5 (Olig2d5) still inhibits PCR, whereas deletion of 8 nucleotides from the 3′ end of Olig2.5 (Olig2d3) eliminated its inhibition (Fig. 3A). This means that a few nucleotides that is essential for PCR inhibition of Olig2.5 was deleted in Olig2d3. We then test several oligonucleotides derived from Olig2.5 with various deletion at the ends. We found that Olig2d51, deletion of 5 nucleotides from Olig2d5, didn't inhibit PCR (Fig. 3B), suggesting inhibition sequence is initiated in the 5 nucleotides. Olig2d31 has 3 more nucleotides than Olig2d32, the different effect between Olig2d31 and Olig2d32 in PCR indicated that inhibition sequence is ended in the 3 nucleotides (Fig. 3B). To see if the inhibition sequence is a continuous one or not, we used 3 oligonucleotides derived from Olig2d31 by various deletion at the middle region. Although Olig2d312 and Olig2d313 didn't showed PCR inhibition effect, Olig2d311 retains PCR inhibition effect (Fig. 3C), indicating that the inhibition sequence is a discontinuous one.
Based on Olig2d31, Olig2d32 and Olig2d311, we further test d311R1 and d311R2 to determine the 3′ end nucleotide required for PCR inhibition. Similarly, Olig2d5, Olig2d51 and Olig2d311, 4 oligonucleotides were used to determine the 5′ end nucleotide required for PCR inhibition. The results showed that d311R2 and d311L2 can inhibit the PCR (Fig. 3D). Together with the inability of Olig2d32 and d311L3 to inhibit PCR ( Fig. 3B and D), the inhibitory sequence was shortened as d311LR1 (Table 1). Considering the differences of PCR inhibition between Olig2d311 and Olig2d312, we used d311LR1 and its derivatives to test their PCR inhibitory effect. All the derivatives by deletion of 1-4 nucleotides from d311LR1 could inhibit PCR (Fig. 3E). We further used a series of oligonucleotides with single nucleotide mutation to determine the minimum inhibitory sequence. When added at a final concentration of 0.4 μM, sequences with mutation of TT at the 5′ end and GCCTC at the 3′ end into other nucleotides did not cause PCR inhibition (Fig. 3F top). When used at 0.6 μM, only mutations at GCCTC didn't cause PCR inhibition (Fig. 3F bottom). Finally, we found that triplicate repeats of GCCTC failed to inhibit PCR. However, TTcGCCTC, a complement sequence of d311LR1 between TT and GCCTC, still inhibited PCR (Fig. 3G). This indicated that both TT and GCCTC are required for PCR inhibition. We then used real-time PCR to determine the number of nucleotides required between TT and GCCTC to inhibit DNAP. We found that N10 inhibit PCR more severely than N9, and there is no inhibitory effect of oligonucleotide N8 (Fig. 3H). Taken together, these results suggested that TT(N)mGCCTC (Fig. 3I), where (N)m means any sequence equal or longer than 9 nucleotides, can inhibit PCR using archaeal family B DNAPs. It is interesting that the frequency of TT(N) 9-20 GCCTC in archaeal is higher than the estimated GCCTC frequency, whereas the frequency of TT(N) 9-20 GCCTC in bacterial is lower than the estimated GCCTC frequency ( Table 2).

TT(N)mGCCTC-containing sequence binds to family B DNA polymerase but not Taq DNA polymerase. To investigate if TT(N)mGCCTC-containing sequences can bind to and thus inhibit the family B
DNA polymerases, we performed electrophoresis mobility shift assay (EMSA). Taq didn't cause any band shift of the Bio-Olig2.5 bands (Fig. 4A), indicating that without Mg 2+ the affinity between Taq and single-stranded DNA is too low to be detected. However, a shifted band of Bio-Olig2.5 was detected when there is a family B DNA polymerase such as GXL, Q5 and KOD (Fig. 4A). The shifted bands disappeared when excess amount of N10 was added as unlabeled specific competitor (Fig. 4A). However, the shifted bands persisted even there was 1000 excess amount of MutL2-16 as non-specific competitor, indicating that proofreading DNA polymerases specifically bind to TT(N)mGCCTC-containing sequences. We finally used purified Taq and KOD to measure the equilibrium dissociation constant (K D ) between DNAPs and Olig2.5. Our results showed that approximately 50% of the Olig2.5 was in a complex with KOD when the latter is at 40 nM (Fig. 4B), suggesting a K D value of about 40 nM. In contrast, there's no detectable interaction between Olig2.5 and Taq even Taq was used at a concentration as high as 1000 nM (Fig. 4B).

Discussion
We here showed that TT(N)mGCCTC-containing oligonucleotides can specifically inhibit PCR using proofreading family B DNAPs. A few DNA aptamers that can inhibit DNAPs have been identified [17][18][19] . These aptamers bind to and inhibit DNAPs at relatively low temperature through their secondary structures or G-quadruplex structures. As these structures are normally disrupted during PCR, these aptamers can be used to increase PCR specificity. As TT(N)mGCCTC cannot form a stable secondary structure, the inhibition effect of TT(N)mGCCTC to proofreading DNAPs relies on its primary structure. To the best of our knowledge, this is the first report describing of archaeal family B DNAPs inhibited by the single-stranded form of oligonucleotides.
PCR may fail if a TT(N)mGCCTC-containing primer was used together with a proofreading DNA polymerase, as the primers are normally used at a final concentration higher than 0.1 μM. Researchers tend to add more primers to increase the amplification efficiency by enhancing the template-primer binding, which will further decrease the efficiency of PCR if proofreading DNA polymerases and TT(N)mGCCTC-containing primers are used. In such cases, use the inhibitory primers at lower concentrations such as 0.1 μM and 0.05 μM will greatly improve the PCR yield, as the inhibition effect of TT(N)mGCCTC-containing primers are not significant when they are used below 0.1 μM (Fig. 1D&E). The user manual of KOD-Plus-Neo (http://www.toyobo-global.com/ seihin/xr/lifescience/support/manual/KOD-401.pdf) showed that higher concentrations of primers may inhibit PCR amplification of long targets. The long target bands was already weakened before the concentration of primers are high enough to introduce non-specific amplification of shorter fragments, indicating that the inhibition is not due to the non-specific amplifican. Possibly there are sequences other than the TT(N)mGCCTC described here, and the previously described GGGGG and GGGGHGG that can inhibit proofreading DNA polymerases 16 . The use of lower concentration of primers could be a general rule if a proofreading DNA polymerase failed to amplify a target.
It is well known that the sequence 5′-GAGGC-C-GAGGC-C-GCCTC-G-GCCTC-3′, consisting of four GAGGC/GCCTC repeats, are essential for the large-T-antigen binding and DNA replication of simian virus 40 (SV40) 20,21 . Here, the GAGGC sequence is reverse complement to GCCTC. The four GAGGC/GCCTC repeats are also present in the origin of DNA replication of polyoma virus JC and polyoma virus BK 22 . The large-T-antigen can bind to GAGGC/GCCTC repeats and unwind the duplex DNA 20,21 . Protein-protein interaction between the  large-T-antigen and DNA polymerase α, a eukaryotic family B DNAP, directly stimulate the replication of the SV40 genome 23 . We don't know if DNA polymerase α utilizes a distantly located TT from GCCTC for its DNA binding and replication. Although the frequency of TT(N)mGCCTC is higher than expected in archaeal genome (  China). His tag was cloned into pTrc-Taq to generate pHis-Taq plasmid using T4 DNA polymerase 24 . Briefly, 50 ng of NcoI digested pTrc-Taq plasmid treated with 1 unit of T4 DNA polymerase at 20 °C for 3 min, followed by 75 °C for 5 min to inactivate T4 DNA polymerase, then oligonucleotides HisInTaqF and HisInTaqR (Table 1) were added into the reaction during the 75 °C incubation, the reaction was incubated at 50 °C for 10 min before transformation into competent cells. pHis-Taq and pET22b-KOD were transformed into Rosetta (DE3) for expression, and His-tagged Taq and KOD were purified by Ni-NTA-Sefinose Column (Sangon Biotech). Purified Taq and KOD were used for measuring DNA binding constant. PCR conditions. PCR primers are listed in Table 1. PCR programs for Q5, Cobuddy, PS, GXL and KOD is: initial denaturation at 98 °C for 30 s; followed by 25-35 cycles of denaturation at 98 °C for 10 s, and annealing and extension at 66 °C for 30 s/kilobase (kb); and a final extension at 66 °C for 5 min. The same program except that annealing and extension for 50 s/kb was used for LA. For Taq, the initial denaturation was at 94 °C for 3 min,