Detection of Single-Nucleotide Polymorphism on uidA Gene of Escherichia coli by a Multiplexed Electrochemical DNA Biosensor with Oligonucleotide-Incorporated Nonfouling Surface

1. Introduction

Single nucleotide polymorphism (SNP) is a single base mutation in a DNA sequence, with a usual alternative of two possible nucleotides at one position. In all sequence variations [1] SNPs are most prevalent and stably inherited types, which are regarded as potent genetic markers and promising indicators for biomedical research, drug development, clinical diagnosis, and disease therapy [2–5].

Therefore, it has great value to develop SNP analysis methods with high sensitivity, specificity, and cost-effectivity to meet the demand of fast, practical analysis for genetic analysis of diseases and subtle genetic risk factors. So far several methods for genotyping SNPs have been reported based on a wide variety of specific probes and enzymes, including allele-specific oligonucleotide (ASO) hybridization [6], oligonucleotide ligation assays (OLA) [7–9] and primer extension assays [10], connected with the signal reading out on different physical platforms covering fluorescence [8,11], mass spectroscopy [12,13], electrochemistry [14,15], surface plasmon resonance [16] and so on.

Among all the available read-out mechanisms, the electrochemical DNA biosensor [17–19] is attracting considerable interest due to its high sensitivity, low cost, and use of a miniaturized and portable device [20]. However electrochemical DNA biosensors are still often hampered by unspecific adsorption [21] in complex matrix and real-world samples. Many efforts have been taken to improve the performance of the electrode surface as a key interface where the specific biological response and electron transferring happens [22,23].

Oligo(ethylene glycol) (OEG) is popularly employed to prepare nonfouling surfaces for its effect of protein-resistance. Inspired by OEG, our lab developed a highly selective oligonucleotide-incorporated nonfouling electrode surface (ONS) and successfully applied it to electrochemical DNA sensors. The ONS was constructed by self-assembling of OEG-terminated thiols (SH-OEG) and thiolated DNA probes (SH-DNA) together on to the electrode. When used for hybridization detection, the electrochemical biosensor was demonstrated to have nonfouling property and repel nonspecific adsorption of proteins, leading to an electrochemical DNA sensor that can effectively identify DNA targets with higher sequence specificity even in the presence of complicated biological fluids (e.g., human sera) [24].

More recently, we developed a ligase-based strategy [25] for SNP detection with this ONS electrochemical DNA sensor on a 16-electrode array as a proof-of-concept experiment. Benefited from the excellent specificity of the ONS electrode [26], the sensor was demonstrated to have higher differentiation ability toward SNPs on synthesis DNA than that with other electrode surface.

In this work, we aimed to optimize the biosensing principle and demonstrate the practicality of the ONS-based electrochemical DNA biosensor. We analyzed a SNP on β-glucuronidase gene (uidA) of Escherichia coli (E. coli) (T93G) which is well-known to be of great significance to distinguish the most harmful E. coli O157:H7 to prevent a possible life-threatening outbreak and to treat the patient pertinently.

2. Experimental Section

3. Results and Discussion

Our SNP detection strategy was based on the DNA-templated ligation and the outstanding property of the ONS electrode (Scheme 1). Gold electrodes were first covered with the capture probes which combined to the gold surface through strong Au-S covalent bonds, and then SH-OEG was added to form a SH-DNA (capture probes)/SH-OEG self-assembled monolayer (SAM). Two capture probes (capture wild/mutant, CP-W/M) were employed in this work, which differed only at the last base at 3′-terminal as listed in Table 1. A “sandwich” like structure would be formed in the presence of a target DNA and a signal probe. Depending on the last base, each capture probe could only hybridize to one target DNA with perfect matching (CP-W to target W, CP-M to T-M), but had a missmatch point adjacent to the signal probe with the other DNA (CP-W to T-M, CP-M to T-W). When the specific hybridization was fully matched, a DNA templated ligation would be performed by the function of the ligase, as a result, the signal probe with a biotin label covalently bound to the capture probe. In contrast, the ligation would be stopped due to the mismatched adjacent point, and then a stringent wash step would denature the sandwich structure and remove the unbound signal probe. After the ligation and the following wash step, avidin-HRP is incubated with the surface, which would catalyze the reduction reaction of hydrogen peroxide and provide amperometric readout for the SNP detection. In order to realize simultaneous detection of two possible alleles, we employed a 16-electrode sensor array to perform multiplexed SNP genotyping.

With the purpose of verification and optimization of the anlaysis strategy, we synthesized two 26-base DNA targets (T-W and T-M), that had sequences corresponding to a fragment of the wild and mutant type uidA gene respectively, which means the T-W fully match with CP-W, but had a single mismatch point with CP-M; the T-M did just the opposite.

When the target DNA was added, it hybridized to the capture probe and combined a signal probe to form a “sandwich” structure. If fully complementary to the capture probe, the target DNA would template a ligation between two probes with the effect of a ligase, leading to a significantly enhanced HRP-catalyzed reduction current. When there was a mismatched site, the reduction current signal was decreased due to the absence of ligation and wash-off of the signal probe. The amperometric I-t curves provide quantitative measurements, which reveal that the differentiation ratio between the fully complementary target and the mutant was as high as 4-fold (Figure 1). We thus demonstrated that, by using appropriate probes and ligase, unambiguous electrochemical signals could be achieved for the fully matched DNA while not the mismatched sequence harboring the SNP.

For the analysis of a mixed target solution, the concentration of the total DNA target is critical for the performance of the biosensor. As shown in Figure 2, DNA samples containing 10% mutant DNA (T-M) and 90% wild DNA (T-W) was analyzed by two different electrodes, the current signals were all increased on both electrodes when the total concentration was increased from 5 nM to 20 nM, but the signal difference ratio R (Ratio of current signal from electrode with CP-W to that from electrode with CP-M) was decreased (from 1.5 to 2.1). It’s easy to understand, the fully matched signal would obviously increase with the increase of the total concentration; on the other side, more mismatched target would also generate higher unspecific adsorption, more inaccurate ligation and more survival in denaturation step which would hamper the discrimination of the biosensor. So we choose 5 nM as an optimized total target concentration for the analysis of a mixed DNA sample.

We then employed a 16-electrode sensor array to perform multiplexed SNP detection of a series of samples with different concentration ratio of T-W and T-M. Firstly two capture probes (CP-W/M) were assembled onto the electrodes. When the mixed sample was added, only the perfectly matched targets would template the ligation and consequently cause a catalyzed current signal while mismatched ones could not. As shown in Figure 3, the average signals of I-t curves from different electrodes were changed highly depended on the concentration ratio of mismatched target: When the percentage of the T-M increased from 0% to 100%, the current signal from electrode with CP-W decreased while that from electrode with CP-M increased. When we calculated the signal difference ratio R (Figure 3(B)), it was more than 10 times higher at 100% mutation ratio than at 0%, and 10% mutant DNA (T-M) was clearly discriminated based on the signal ratio.

Finally, to challenge the practical applicability of our multiplexed electrochemical biosensor, we analyzed a PCR amplicons from mixed sample containing both the wild and mutant uidA gene of E. coli. By using an asymmetric PCR protocol [27,28], we obtained ssDNA targets of 250 bp that could be directly sensed. As shown by the result of electrophoretic analysis (Figure 4), PCR amplification got anticipative products from all samples including wild type uidA gene, mutant type uidA gene, and their mixture.

Amplicons from different templates were analyzed by our biosensor of electrodes with capture-M, the result was shown in Figure 5. Interestingly, our sensor performed equally well despite the fact that this target is much longer than the model. The negative control (NC) and mismatched sample (100% wild uidA amplicon) got only negligible signal. But when the percentage of mutant uidA template increased, the amperometric signal accordingly got higher. We found that, by coupling asymmetric PCR, this sensor could get as high as about five-fold signal distinction between 100% mutant and 100% wild wild uidA PCR amplicons. By this multiplexed electrochemical biosensor 10% mutant uidA genomic DNA could be selectively identified.

4. Conclusions

In summary, we have optimized the performance of the ONS- and ligation-based electrochemical DNA sensor for its practicability, and utilized the method for the detection of a mutant site (T93G) in uidA gene of E. coli which is known to be an important marker for the bacteria hazard analysis. We demonstrate that the 16-electrode sensor array provides a multiplexed platform for simultaneous detection of allelic genes, and 10% mismatched DNA is successfully distinguished. Finally, a series of PCR amplicons from templates of increasing percentage of mutant DNA is analyzed. Results showed that the signal distinction was as good even when the target sequence is lengthened to 250 bp in a complex PCR product. Therefore, we proved the ONS electrochemical DNA biosensor to be a promising tool for the selective and inexpensive analysis of SNPs in practical application.