DNAzyme catalytic beacon sensors that resist temperature-dependent variations

Abstract

The temperature-dependent variability of a Pb²⁺-specific 8-17E DNAzyme catalytic beacon sensor has been addressed through the introduction of mismatches in the DNAzyme, and the resulting sensors resist temperature-dependent variations from 4 to 30 °C.

The demand for simple, on-site, real-time detection and quantification of heavy metal ions is increasing every day, since metal ions such as lead and mercury are common environmental contaminants that have significant adverse affects on human health. To meet the demand, a number of highly sensitive and selective metal ion sensors have been developed.¹ However, it has been difficult to rationally design molecules that bind to a specific target metal ion with high selectivity, while applying the same methodology to a wide variety of other metal ion targets. A complementary approach to rational design is combinatorial selection. The combinatorial technique employed for the isolation of deoxyribozymes or DNAzymes is known as in vitro selection, and sequences that bind to metal ions, such as Pb²⁺, Hg²⁺, Zn²⁺, Cu²⁺ and UO ²⁺₂, with high affinity and selectivity have been obtained previously.^2-4 Upon combining the selected DNAzymes with either fluorophore/quencher pairs,⁵ nanoparticles⁶ or electro-active tags,⁷ new and powerful classes of fluorescent, colorimetric and electrochemical sensors have emerged that can be applied for detecting a wide range of metal ions, including both diamagnetic and paramagnetic metal ions. This general applicability, together with their high stability, biocompatibility, biodegradability and cost-effectiveness, makes DNAzymes an ideal choice for designing metal ion sensors.^3,8 Using DNA-based catalytic beacons, we and others have demonstrated fluorescent sensing of many targets with high sensitivity and selectivity.^5,9 A good example is the 8–17E DNAzyme catalytic beacon used for sensing Pb²⁺, through labeling the substrate strand with a fluorophore such as FAM, and the enzyme strand with quenchers such as DABCYL and BHQ-1^s (Fig. 1a). In the presence of Pb²⁺, the substrate strand containing a single ribonucleotide base is cleaved, resulting in the release of the substrate arm containing the fluorophore (Fig. 1b).

Fig. 1.

(a) Predicted secondary structures of the fluorescent DNAzyme lead sensors. F represents the fluorophore, FAM, Q1 is the quencher BHQ-1ρ and Q2 is the DABCYL quencher. The single RNA base on the substrate arm is denoted by rA; (b) schematic representation of the catalytic beacon sensor; (c) kinetics of fluorescence increase of RT (sensor that contains no mismatch), MM1 (sensor containing an A·C mismatch) and MM2 (sensor containing a G·T wobble pair), over the background at 25 and 4 °C, with 2 μM Pb²⁺ in 50 mM HEPES with 100 mM NaCl at pH 7.2. Pb²⁺ is added at the 20 s time point.

Despite the great promise and huge potential of the DNAzyme sensors, there is a major unsolved issue that prevents their wide range applications. Since DNA hybridization and dehybridization play a critical role in almost all DNAzyme sensor designs, the sensor performance is vulnerable to temperature variations, making it difficult for simple on-site and real-time detection, particularly since temperature variations commonly occur between different locations and times of the day. Implementing temperature control units or calibrations increases both the cost and complexity of the sensing operation. Therefore, it is highly desirable to design functional DNA sensors whose performance withstands temperature variations in the practical range. Here, we report a novel strategy of using mismatches in the DNAzyme to tune and control its temperature stability so that the catalytic beacon exhibits similar fluorescence signals at temperatures between 4 °C and 30 °C. In the process, we demonstrated that an A·C mismatch and a G·T wobble pair play different roles in achieving the same goal (Fig. 1a). The strategy demonstrated here can also be applied to some other nucleic acid-based sensors.

The temperature-dependence of the 8-17E DNAzyme catalytic beacon sensor (Fig. 1a) was followed in real-time using fluorescence spectroscopy. The fluorescence increase wasmonitored at 520 nm, the emission wavelength of FAM, as a function of time. At 25 °C, a dramatic increase in fluorescence intensity was observed upon addition of 2·M of Pb²⁺; in contrast, when the same experiment was carried out at 4 °C, a negligible increase was observed (Fig. 1c). The initial rate of substrate release of the sensor was further studied between 4 and 30 °C (Fig. 2a), and while the increase was very slow at 4 °C, it became faster at 15 °C, reaching the maximum at B25 °C, after which the rate started to drop again. Since this sensor exhibited high activity at B25 °C, it was referred to as the RT sensor. The pH variance of 50 mM HEPES with 100 mM NaCl buffer used was <0.4% between 4 and 30 °C, and therefore did not contribute to the temperature dependence of the sensor. For the efficient functioning of the catalytic beacon sensor, a fine balance had to be struck between the high stability of the substrate-enzyme strands for obtaining low background fluorescence in the absence of Pb²⁺, and the low stability of the product–enzyme strands for fast product release and high fluorescence increase in the presence of Pb²⁺. For the RT sensor, the stability of the product-enzyme strand was so high at 4 °C that the product was not released even after Pb²⁺-dependent cleavage. Increasing the temperature provided sufficient energy to facilitate the release of the product strand, and higher rates of fluorescence increase were observed at 15 °C and 25 °C. Above 25 °C, however, the stability of the substrate-enzyme complex gradually started to decrease, causing dehybridization and high fluorescence background before Pb²⁺ addition. Thereby, the rate of fluorescence increase upon Pb²⁺ addition was slightly lowered again at 30 °C.

Fig. 2.

Temperature-dependent initial rate of substrate release of (a) RT, MM1 and (4: 1) MM1: RT upon addition of 2 μM Pb²⁺, and (b) MM2 upon addition of 750 nM and 500 nM Pb²⁺; RT upon addition of 750 nM Pb²⁺ in 50 mM HEPES with 100 mM NaCl at pH 7.2.

Previous studies have shown that there is a considerable change in the dissociation kinetics and melting temperatures of oligonucleotides that are perfectly matched or contain stable mismatches vs. those containing unstable mismatches. The trend in the stability of internal mismatches in DNA, based on the free energy changes (ΔG° ₃₇ (kcal mol⁻¹)) values at pH 7.0 in 1 M NaCl, calculated using the nearest-neighbor model, follows the order G·T ≥ G·A > A·C.¹⁰

Hence, in order to tune the stability of the system and improve performance at lower temperatures, we propose introducing the least stable A·C mismatch to the middle of the releasing arm of the enzyme strand 17E (T_16.5–C) (called MM1 sensor, see Fig. 1a). Such a mismatch may allow considerable destabilization and thus, release of the product even at low temperatures. Indeed, the MM1 sensor with the mismatch showed a much higher increase of fluorescence intensity at 4 °C in comparison to that of the RT sensor (Fig. 1c). The rates of fluorescence increase were then plotted as a function of temperature (Fig. 2a). While the MM1 sensor showed a high signal response at 4 °C upon Pb²⁺ addition, such a response decreased with increasing temperature from 4 to 30 °C. Since the temperature-dependent fluorescence changes of the RT and MM1 sensors were almost opposite to each other, we investigated the possibility of mixing the two sensors at different ratios to develop a sensor system, whose signal is relatively independent of temperature. Interestingly, at a ratio of 4: 1 for MM1: RT, the fluorescence signal response to Pb²⁺ addition remains almost constant between 4 °C and 30 °C (Fig. 2a).

Having achieved the development of a catalytic beacon sensor system that was relatively independent of temperature using a mixture of two sensors with opposite temperature dependence, we wanted to investigate whether there could be an even simpler system consisting of only a single sensor. Hence, instead of the complete A·C mismatch, as in MM1, we tested 17E (A_16.7–G) (called MM2 sensor, see Fig. 1a), which contained a G·T wobble pair that is less stable than a G–C Watson–Crick pair, but more stable than the A·C mismatch.¹⁰ The wobble pair may be stable enough to allow hybridization of the substrate and the enzyme strand even at higher temperature, while exerting enough destabilization to allow release of the product arm at low temperature. Indeed, this MM2 sensor system displayed a high increase in the fluorescence intensity at both 4 and 25 °C (Fig. 1c), and upon measuring the initial rate, a relatively constant response between 4 and 30 °C was observed (Fig. 2b).

Another sensor containing a complete T·C mismatch on the releasing arm of the enzyme was also studied (see ESI, Fig. S1†) and this sensor exhibited behavior similar to MM1, while a sensor containing an A·C mismatch on the non-releasing arm of the enzyme on the other hand, showed a response similar to the original sensor, RT (see ESI, Fig. S2†). This confirmed that the use of the same mismatch on the non-releasing arm of the enzyme behaved similarly to the original RT sensor, since it was not involved in the release of the substrate arm that dictated the overall sensor behavior. The fluorescence response of sensors containingdual mismatches was also studied (see ESI, Fig. S3†). The first dual mismatch sensor, DMM1 containing the same two mismatches as the MM1 and MM2 sensors (A·G and T·C) on the releasing arm showed a fluorescence response similar to MM1, albeit lower, owing to its considerable destabilization. The second dual mismatch sensor, DMM2 on the other hand, containing one mismatch on the releasing arm (A·G), and one on the non-releasing arm (T·C), showed a response similar to MM2. The overall response therefore, was dictated by the mismatch on the releasing arm of the substrate, which in this case was identical to MM2, further confirming that the sensor behavior was dependent on the nature as well as the position of the mismatch either on the releasing or non-releasing arm.

To ensure that the development of the temperature- independent catalytic beacon sensor did not compromise the sensitivity and selectivity, signal responses of the MM2 sensor with different metal ions were studied. All competing metal ions tested showed a negligible response with the sensor at both 4 and 25 °C, suggesting excellent selectivity (Fig. 3). The sensitivity of the MM2 sensor was also tested and the detection limits for Pb²⁺ at 4 °C and 25 °C were determined to be 50 nM and 20 nM, respectively (see ESI, Fig. S4†). Since the maximum contamination level for Pb²⁺ in drinking water is defined by the US Environment Protection Agency to be 72 nM, and the levels for lead in other media such as paint, dust and soil are more than 72 nM, the temperature-independent DNAzyme catalytic beacons described here are well suited for monitoring Pb²⁺ in the environment.

Fig. 3.

Selectivity of the MM2 sensor. Kinetics of fluorescence increase relative to background at 6 min after incubation with (a) 1 μM of M²⁺ ions at 4 °C and (b) 500 nM of M²⁺ ions at 25 °C, in buffer containing 50 mM HEPES with 100 mM NaCl at pH 7.2.

In summary, we have demonstrated the use of the intrinsic lowered stability of mismatches to develop catalytic beacon Pb²⁺ sensors whose performance is independent of temperatures between 4–30 °C. In the process, we showed that the difference in stability between the A·C mismatch and G·T wobble pair can both be utilized, albeit with different strategies in achieving the same goal. Free of the need for temperature calibration, such a temperature independent sensor system can now find even wider applications for the on-site and real-time detection of lead in the environment. The methodology developed in this study can also be applied for designing other nucleic acid sensors that resist temperature dependent variations.