DNA-only Cascade: A Universal Tool for Signal ... - ACS Publications

Aug 26, 2014 - Copyright © 2014 American Chemical Society. *S. M. Bone. E-mail: [email protected]. Cite this:Anal. Chem. 2014, 86, 18, 9106-9113 ...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/ac

DNA-only Cascade: A Universal Tool for Signal Amplification, Enhancing the Detection of Target Analytes Simon M. Bone,*,†,‡ Nicole J. Hasick,‡ Nicole E. Lima,‡ Simon M. Erskine,‡ Elisa Mokany,‡ and Alison V. Todd†,‡ †

The University of New South Wales, Kensington, New South Wales 2052, Australia SpeeDx Pty Ltd., Eveleigh, New South Wales 2015, Australia



S Supporting Information *

ABSTRACT: Diagnostic tests performed in the field or at the site of patient care would benefit from using a combination of inexpensive, stable chemical reagents and simple instrumentation. Here, we have developed a universal “DNA-only Cascade” (DoC) to quantitatively detect target analytes with increased speed. The DoC utilizes quasicircular structures consisting of temporarily inactivated deoxyribozymes (DNAzymes). The catalytic activity of the DNAzymes is restored in a universal manner in response to a broad range of environmental and biological targets. The present study demonstrates DNAzyme activation in the presence of metal ions (Pb2+), small molecules (deoxyadenosine triphosphate) and nucleic acids homologous to genes from Meningitiscausing bacteria. Furthermore, DoC efficiently discriminates nucleic acid targets differing by a single nucleotide. When detection of analytes is orchestrated by functional nucleic acids, the inclusion of DoC reagents substantially decreases time for detection and allows analyte quantification. The detection of nucleic acids using DoC was further characterized for its capability to be multiplexed and retain its functionality following long-term exposure to ambient temperatures and in a background of complex medium (human serum).

I

The nucleic acid diagnostic potential for DNAzymes that modify nucleic acid substrates has some limitations because the sequences they recognize and bind (the inputs) are also those that they catalytically modify (the outputs). We previously reported the development of Multi-component nucleic acid enzymes (MNAzymes) to circumvent this constraint.6,7 Here, the catalytic cores of candidate 10-23 and 8-17 DNAzymes were split into two parts, and nucleic acid target-sensing arms were added to each half-enzyme molecule (now referred to as a “partzyme”). Thus, only in the presence of the specific target are the partzymes able to bind adjacently to it, reuniting the catalytic core. The newly assembled MNAzyme can then modify a separate substrate to indicate the presence of the target. Similar methods have also been reported that split the E6 DNAzyme.8,9 DNAzymes possess some intrinsic ability to amplify signal due to their multiple turnover capabilities. Amplification beyond this capacity has also been demonstrated via construction of molecular cascades, the first of which was a “cross-catalytic” cascade between two temporarily inactivated 10-23 DNAzymes that were circularized via ligation.10 Although simple in design, the success of the cascade is absolutely dependent on circularization of all components, which can be

n recent years, there has been an increased demand for diagnostic technologies that function outside of a centralized laboratory; for example, in environmental field testing or at the site of patient care. Ideally, in this situation, the instrumentation should be simple and assay components inexpensive and highly stable. Diagnostic technologies displaying both attributes remain relatively scarce. For nucleic acid detection, gold standard target replication strategies such as the polymerase chain reaction (PCR) are not always ideal, because they require expensive, labile protein enzymes and complex equipment to facilitate temperature cycling. Isothermal technologies can be paired with simpler instruments; however, many still use one or more protein enzymes to replicate the gene of interest.1 Consequently, methods of signal amplification have become desirable, particularly those that eliminate the requirement for protein enzymes and/or allow options for detection of disparate types of analytes. An exciting possibility is the potential for protein enzyme-free signal cascades, which solely use the ability of DNA to perform machine-like functions. DNAzymes are ideal candidates for this purpose, as they are enzymes composed of inexpensive and highly stable, synthetic DNA and possess inherent multiple turnover catalytic capabilities.2−4 Several DNAzymes have been evolved via in vitro selection, with a range of nucleic acid-modifying properties such as cleavage or ligation (reviewed by Silverman2) and one unique class can catalyze peroxidasemediated chemical reactions.5 © 2014 American Chemical Society

Received: May 15, 2014 Accepted: August 18, 2014 Published: August 26, 2014 9106

dx.doi.org/10.1021/ac501811r | Anal. Chem. 2014, 86, 9106−9113

Analytical Chemistry

Article

The dATP small molecule (Bioline), was provided at a concentration of 5 mM, 500 nM or 300 nM to displace the MNAzyme facilitator so that it was available to assemble MNAzymes formed from 10 nM Partzyme A (ATP(R3)A4/3) and 10 nM Partzyme B (ATP(R3)B5/3). For the left panel, the active MNAzyme cleaved the fluorescently labeled MNAzyme substrate (Sub3-FB). For the right panel, the active MNAzyme cleaved a quasi-circle consisting of a 14 nM BL (C(R34b)) and a 10 nM DNAzyme (Dz77_55(8:9)) with signal produced by cleavage of a 200 nM DNAzyme fluorescent substrate (Sub77_55-FIB). Reactions were performed in 1× immobuffer (Bioline) and initially incubated at 44 °C on the CFX96 realtime PCR detection system for 45 min with no fluorescence monitoring. This was then followed by a 54 °C reaction for 45 min, with fluorescence measured every 30 s. All samples were run in duplicate and results in Figure 3B are the averages from duplicates, which were plotted using Microsoft Excel (Version 14). Detection of the Nucleic Acid Target. The reactions reported in Figure 3C used an MNAzyme comprising 40 nM Partzyme A (cfb2A4/72) and 40 nM Partzyme B (cfb2B5/72), which hybridized to an 80 pM, 20 pM or 5 pM synthetic nucleic acid target. This sequence is homologous to a region of the Streptococcus agalactiae Christie-Atkins-Munch-Petersen (CAMP) gene (GenBank: X72754.1) (AF-cfb2). For the left panel, the active MNAzyme cleaved the fluorescently labeled MNAzyme substrate (Sub72-FIB). For the right panel, the active MNAzyme cleaved a quasi-circle consisting of a 10 nM BL (C(R22h)) and an 8 nM DNAzyme (Dz77_55(8:9)) with signal produced by cleavage of a 200 nM fluorescent DNAzyme substrate (Sub77_55-FIB). All reactions contained 1× PCR buffer II (Life Technologies) and were performed according to the Isothermal Reaction Method with a 53 °C reaction temperature for 2 h. DoC Limit of Detection. The reactions reported in Figure 4A,B contained a quasi-circle consisting of a 10 nM BL (C(R22h)) and an 8 nM DNAzyme (Dz77_55(8:9)) as well as a 200 nM fluorescently labeled DNAzyme substrate (Sub77_55-FIB). The quasi-circle was cleaved by the MNAzyme consisting of 40 nM Partzyme A (cfb2A4/72) and 40 nM Partzyme B (cfb2B5/72) and either no target or a 160 pM, 80 pM, 40 pM, 20 pM, 10 pM, 5 pM or 1 pM synthetic nucleic acid target (AF-cfb2). Each of the reactions were performed in 1× PCR buffer II according to the Isothermal Reaction Method with a 53 °C reaction temperature for 2 h. DoC Single Nucleotide Discrimination Test. The reactions reported in Figure 4C contained a quasi-circle consisting of a 10 nM BL (C(R22h)) and an 8 nM DNAzyme (Dz77_55(8:9)) and a 200 nM fluorescently labeled DNAzyme substrate (Sub77_55-FIB). The quasi-circle was cleaved by the MNAzyme consisting of 50 nM Partzyme A (cfb2A4/72) and 50 nM Partzyme B (cfb2B5/72(R57f2)), a stabilizer which hybridized to the target adjacent to Partzyme B (ST(R57f)) and either no target or 100 pM of the following synthetic nucleic acid targets, which differ by a single nucleotide; the wild-type Target (A) (AF-Lcfb2(R56)), which is fully matched to the partzyme sensor arms, Target (C) (AF-Lcfb2(R56c)), Target (T) (AF-Lcfb2(R56t)) or Target (G) (AF-Lcfb2(R56g)). Reactions were performed in 1× PCR buffer II following the Isothermal Reaction Method with a 54 °C reaction temperature for 2 h. All samples were run in triplicate and the columns in Figure 4C are the averages from triplicates

difficult to achieve by standard ligation and purification methods. An alternative approach reported by Elbaz et al. involved the temporary inactivation of peroxidase-mimicking DNAzymes within a “quasi-circular structure”.11 The use of Watson−Crick hybridization negated the requirement for difficult ligation and purification steps. Despite the broad applications of peroxidase-mimicking DNAzymes, these enzymes cannot catalyze the modification of nucleic acid substrates, thus they provide no mechanism for activation of additional catalytic nucleic acids capable of mediating signal amplification. Here we report the creation of a DNA-only cascade (DoC), which utilizes quasi-circular structures containing inactivated nucleic acid-modifying DNAzymes. The cascades are quantitative and several independent cascades can function simultaneously for multiplex target detection. The broad applicability of the cascades is exemplified here via (1) alternate mechanisms by which DoC can be activated, for example, using catalytic nucleic acids or protein enzymes and (2) the range of stimuli that can initiate DoC including environmental toxins, small molecules and nucleic acid sequences. Further, nucleic acid targets were also efficiently discriminated against that differed by as little as a single nucleotide. The oligonucleotide components of the DoC are inexpensive and their highly stable nature allows the cascade to function in a high background of complex media such as human serum.



EXPERIMENTAL DETAILS All synthetic oligonucleotides were purchased from either Integrated DNA Technologies (Coralville, IA, USA) or Biosearch Technologies (Petaluma, CA, USA). The oligonucleotide sequences are listed in Table S-1, Supporting Information. Isothermal Reaction Method. Reactions were initiated by the addition of 45 mM MgCl2 (Ambion) and the total volume of each reaction was 25 μL. Reactions were incubated at a constant temperature (specified for each individual method) in a CFX96 real-time PCR detection system (Bio-Rad) with fluorescence measured every 30 s. All samples were performed in duplicate or triplicate. Unless presented otherwise, results in figures are the averages from replicates that were plotted using Microsoft Excel (Version 14). Detection of the Metal Ion Target, Pb2+. The reactions reported in Figure 3A contained 50 nM of the Pb2+-dependent DNAzyme (Dz-GR5(13:9)) in the presence of 15 nM, 10 nM or 5 nM lead acetate trihydrate (Pb(CH3CO2)2·3H2O) (SigmaAldrich). All reactions contained buffer consisting of 50 mM 4(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) pH 8.2 (Sigma-Aldrich), 100 mM NaCl and 20 mM MgCl2 (both from Ambion). For the left panel, the active Pb2+-dependent DNAzyme cleaved 200 nM of its fluorescently labeled substrate (Sub1(8:19)-FIB). For the right panel, the active Pb2+dependent DNAzyme cleaved a quasi-circle consisting of a 14 nM blocking oligonucleotide (BL) (C(R53a)) and a 10 nM 10:23 DNAzyme (Dz74_55(8:8)) with signal produced by cleavage of a 200 nM fluorescent DNAzyme substrate (Sub74_55-FIB). All reactions were performed according to the Isothermal reaction method, with a 50 °C reaction temperature for 2 h. Detection of the Small Molecule Target, dATP. The reactions reported in Figure 3B contained an MNAzyme facilitator (AF-ATP(R4a)) partially hybridized to an oligonucleotide comprising the aptamer for dATP (AP-ATP(R4e)). 9107

dx.doi.org/10.1021/ac501811r | Anal. Chem. 2014, 86, 9106−9113

Analytical Chemistry

Article

Figure 1. Creation of the DNA-only Cascade (DoC). (A) A quasi-circular DNAzyme structure is composed of (1) a nucleic acid-modifying DNAzyme and (2) a blocking oligonucleotide (BL). The inhibitory regions of the BL are complementary to the substrate-binding arms of the DNAzyme and a portion of the catalytic core. When the two oligonucleotides are hybridized together, the 5′ end of the DNAzyme is paired with the 5′ end of the BL (and vice versa for the 3′ ends), and a quasi-circular structure is formed. In this conformation, the blocked DNAzyme is temporarily inactive. (B) A DNA-only cascade occurs when the intermediate region of the BL includes nucleic acid substrates for both the blocked DNAzyme and for an initiating enzyme responsive to a specific target. In the presence of target, the initiating enzyme cleaves the initiating enzyme substrate, which releases the DNAzyme from the quasi-circular structure, restoring its catalytic activity. The DoC is then triggered when the active DNAzyme cleaves its substrate within additional quasi-circles, resulting in the release of additional DNAzymes. The active DNAzyme can also cleave an independent, fluorescently labeled DNAzyme substrate, responsible for the production of fluorescent signal.

the fluorophores HEX, TR and FAM, respectively, and were each provided at a concentration of 200 nM. Reactions were performed according to the Isothermal Reaction Method with a 54 °C reaction temperature for 2.5 h, either as a singleplex containing only one quasi-circle, MNAzyme and fluorescent substrate or as a multiplex containing all components together.

with the standard deviation of triplicates presented as error bars on each column. DoC Multiplex Test. The multiplex reaction reported in Figure 5 consisted of three quasi-circles; Circle A, Circle B and Circle C. Circle A consisted of a 10 nM BL (C(R39a)-P) and 8 nM DNAzyme 1 (Dz77_55(8:9)). Circle B consisted of a 16 nM BL (C(R28b)) and 10 nM DNAzyme 2 (Dz72(8:9)). Circle C consisted of a 15 nM BL (C(R26j)) and 10 nM DNAzyme 3 (Dz6(9:10)). Each cascade was initiated by an MNAzyme with partzymes at 50 nM that hybridized to a target present at a concentration of 100 pM. Control reactions lacked target. MNAzyme 1 was designed to initiate the Circle A reaction and it consisted of Partzyme A (cfb2A4/45-P) and Partzyme B (cfb2B5/45-P) and Target 1 (AF-cfb2). MNAzyme 2 was designed to initiate the Circle B reaction and it consisted of Partzyme A (cpsA8A4/56-P), Partzyme B (cpsA8B5/56-P) and Target 2, which is homologous to a region within the Streptococcus pneumoniae capsular polysaccharide biosynthesis (cpsA) gene (Genbank: NC_011072.1) (AF-cpsA8). MNAzyme 3 was designed to initiate the Circle C reaction and it consisted of Partzyme A (ctr_A4/80-P), Partzyme B (ctr_B5/ 80-P) and Target 3, which is homologous to a region of the Neisseria meningitidis capsular transport (ctrA) gene (Genbank: HQ156899.1) (AF-ctrA5). The fluorescent DNAzyme 1 substrate (Sub77_55-HB), DNAzyme 2 substrate (Sub72TRB2) and DNAzyme 3 substrate (Sub6-FB) were labeled with



RESULTS Creation of the DNA-only Cascade (DoC). The basic component of the DoC is a quasi-circular DNAzyme structure (Figure 1A). The quasi-circle is composed of two partially complementary oligonucleotides; the first is a DNAzyme capable of cleaving a nucleic acid substrate and the second is designated here as a blocking oligonucleotide (BL). The BL consists of three distinct regions. Inhibitory region 1 is located at the 5′ terminus and inhibitory region 2 is at the 3′ terminus, with each being complementary to a portion of the DNAzyme. The third is the intermediate region, which is universal and can, if desired, consist of any sequence that can be recognized and/ or be modified by another molecule such as an enzyme. When the DNAzyme and BL hybridize to one another, the 5′ and 3′ termini of the DNAzyme are paired with the corresponding 5′ and 3′ termini of the BL, such that in order to maintain correct polarity of Watson−Crick hybridization, a quasi-circular structure is formed. In this conformation, the blocked 9108

dx.doi.org/10.1021/ac501811r | Anal. Chem. 2014, 86, 9106−9113

Analytical Chemistry

Article

DNAzyme is temporarily inactive and unable to modify any substrates. The DoC is pictured in Figure 1B. In this scheme, the BL of a quasi-circle contains two adjacent nucleic acid substrates within its intermediate region, namely, the initiating enzyme substrate and a DNAzyme substrate. Hybridization between the DNAzyme and terminal BL regions is initially favored over that of the substrate within the intermediate region due to a greater number of complementary base pairs that can form. The sequence of events for the DoC is as follows. First, in the presence of a target analyte, the target-responsive initiating enzyme (e.g., MNAzyme) can modify the initiating enzyme substrate resulting in the release of the DNAzyme from the quasi-circular structure and concomitant restoration of its catalytic activity. Cleavage of this substrate region of the BL produces two shorter oligonucleotide fragments, each of which have reduced thermal stability of binding to the DNAzyme alone, than they do when contained within a single molecule. Second, the active DNAzyme can then modify its substrate within additional quasi-circles resulting in additional DNAzyme activation thus causing a feedback cascade reaction. Finally, the active DNAzymes can also cleave a second, separate substrate that is fluorescently labeled providing a means for generating signal in conjunction with the cascade. Target-Dependent Methods of DoC Initiation. To demonstrate the broad applicability of the DOC, it was used to indicate the presence of different environmental and biological targets. The GR-5 DNAzyme was used for the detection of the metal ion target, Pb2+ (Figure 2A) because this possesses specific RNA-cleaving activity in the presence of Pb2+ ions.12,13 The small molecule target (dATP) was detected via an aptamer that binds specifically to adenosine triphosphate (ATP)14 (Figure 2B). The aptamer was initially hybridized and blocked by a complementary oligonucleotide capable of functioning as a MNAzyme facilitator. The binding of dATP to the aptamer displaced the facilitator, which then assembled an active MNAzyme. In addition, a nucleic acid target was directly detected by an MNAzyme where the target itself functioned as the facilitator for MNAzyme assembly (Figure 2C). The specific target was a synthetic DNA target homologous to the cyclic AMP gene of Streptococcus agalactiae (meningitis-causing bacterium). To initiate the DoC in response to each of these targets, the intermediate region of the BL was modified accordingly, so that the initiating enzyme substrate was capable of cleavage by either the Pb2+-dependent DNAzyme or an MNAzyme. For the detection of dATP, an initial incubation at a lower temperature (44 °C) was required for the aptamer to specifically bind to dATP and mediate the release of the MNAzyme facilitator. This was then followed by signal amplification at an elevated temperature (54 °C) for optimal functioning of both the MNAzyme and the DoC. The remaining target detection strategies were performed isothermally. Signal amplification from the DoC was demonstrated with each target detection strategy by performing reactions that either lacked or contained the quasi-circular structures required for the DoC (Figure 3, left and right panels, respectively). For reactions lacking the DoC (direct target detection only), signal was generated via the cleavage of an independent, fluorescently labeled substrate tailored for the initiating enzymes (Pb2+dependent DNAzyme or MNAzyme) (schematic depicted in Figure S-1, Supporting Information). For reactions containing both target detection and DoC signal amplification, signal was

Figure 2. Initiating nucleic acid enzymes can trigger the DoC in a target-dependent manner. (A) The Pb2+-dependent DNAzyme becomes active in the presence of Pb2+ ions. (B) An aptamer sequence is initially hybridized to a partially complementary oligonucleotide, which has the potential to become an assembly facilitator for a multicomponent nucleic acid enzyme (MNAzyme). In the presence of the small molecule target (dATP), the aptamer preferentially binds its target resulting in the release of the MNAzyme facilitator. (C) The nucleic acid target directly functions as the facilitator for the assembly of an MNAzyme. The MNAzyme consists of two partzymes each comprising a substrate-binding arm, a partial catalytic core and a sensor arm complementary to the facilitator. Each method can initiate the DoC via the incorporation of the substrate for either the Pb2+-dependent DNAzyme (panel A) or an MNAzyme (panels B and C) within the intermediate region of a quasi-circular structure.

generated via the cleavage of an independent, fluorescently labeled substrate for the DNAzyme that was released from the quasi-circles during the cascade. For each type of analyte, three concentrations are shown as well as a no target control reaction. In general, there was either a slow, linear increase in fluorescence signal or in some cases, no detectable increase in signal above the no target control for reactions lacking the DoC (Figure 3, left-hand panels). In contrast, reactions containing the DoC produced signal that proceeded at a faster rate and reached a plateau much sooner than the corresponding reactions lacking the DoC (Figure 3, right-hand panels). In the case of the Pb2+ and nucleic acid targets (Figure 3A,C, respectively) the 5 nM and 5 pM concentrations, respectively, only became detectable upon the addition of the DoC quasicircle. Additional experiments were performed to confirm the specificity of Pb2+ detection with the DoC over that of other divalent metal ions (Figure S-2, Supporting Information), and also for the specificity of dATP detection over that of other deoxynucleotide-triphosphates (Figure S-3, Supporting Information). An alternative method for nucleic acid target detection was also demonstrated that used a protein enzyme rather than a catalytic nucleic acid as the initiating enzyme (Figure S-4, Supporting Information). In this protocol, the target directly hybridized to the intermediate region of a quasi-circle, which 9109

dx.doi.org/10.1021/ac501811r | Anal. Chem. 2014, 86, 9106−9113

Analytical Chemistry

Article

Figure 3. Raw fluorescence data for the strategies depicted in Figure 2 either alone or in combination with DoC signal amplification. (A) Detection of the metal ion target (Pb2+), (B) the small molecule target (dATP) and (C) the nucleic acid target. For each target, the graphs pictured on the left demonstrate the signal achieved from the initiating enzymes only, modifying their fluorescently labeled substrates. The graphs on the right demonstrate the signal achieved when the initiating enzymes trigger the DNA-only cascade.

resulted in the completion of a double-stranded recognition site for a nicking restriction enzyme. The enzyme then selectively nicked the BL to initiate the cascade while leaving the target intact, which could then be recycled. Further Characterization of the DoC. The DoC was further characterized for the detection of the nucleic acid target. When the concentration of this target was titrated down, the fluorescent signal continued to display a sigmoidal shape, indicative of exponential signal amplification (Figure 4A). Point of inflection (POI) values (time corresponding to maximum slope of the curve) were calculated for the six highest standards (160 pM to 5 pM). These values were then plotted against the log of the starting quantity to produce a standard curve. The standard curve produced a correlation coefficient (R2) of 0.99, indicating the DoC is quantitative within this target range (Figure 4B). The limit of detection was calculated to be 2.3 pM, according to the method described by Wang and Zhang.15 Briefly, the detection limit was derived from the difference between the POI value of the no target control and 3 times the standard deviation of the POI value of the no target control. In comparison, when the specific MNAzyme that had been used to initiate DoC was tested in the absence of the DoC reagents, a signal was only apparent above background in the presence of the 20 pM target. Because the signal was linear, the limit of detection was calculated by an alternate method utilizing the same principle (Li et al.16) to be 17.7 pM (Figure S-5, Supporting Information). Single-nucleotide discrimination was also demonstrated with the nucleic acid target and variants, which differed by a single nucleotide. The wild-type sequence, which contained an Adenine (A) at a central position within the sequence, was replaced with a Cytosine (C), Thymine (T) or Guanosine (G) to create the variant targets. The POI value

for each target was subtracted from that of the no target control, with the wild-type target (A) producing a much greater difference over that of the other three variant targets (Figure 4C). To demonstrate stability of reagents, the DoC reaction components (except for the Mg2+ necessary for DNAzyme and MNAzyme catalytic activity) were incubated at 40 °C for 48 h. Aliquots were taken at the start of the incubation period (0 h) and reactions were initiated by the addition of MgCl2 and then monitored by the production of fluorescent signal. This was repeated for aliquots sampled at the end of the incubation period (48 h). There was minimal difference in fluorescence signal observed between replicates of the same concentration at the different time points (Figure S-6, Supporting Information). To demonstrate performance in complex media, the DoC was tested in a background of 50% human serum that had been heat treated (70 °C for 5 min) to denature RNases. Detectable increases in fluorescence signal were evident for 100 pM and 20 pM target concentrations with very little increase in background (no target control) fluorescence (Figure S-7, Supporting Information). Multiplex Target Detection. Multiplex target detection was performed with three independently functioning cascades placed together in the same reaction chamber (Figure 5A). Three nucleic acid targets, Target 1, Target 2 and Target 3, each functioned to assemble MNAzyme 1, MNAzyme 2 and MNAzyme 3, respectively. Target 1 was the nucleic acid target used for the initial experiments reported in Figures 3 and 4 and is homologous to the cyclic AMP gene of Streptococcus agalactiae. Target 2 is homologous to a region of the Neisseria meningitidis ctrA gene and Target 3 is homologous to a region of the Streptococcus pneumoniae cpsA gene. Each of these targets 9110

dx.doi.org/10.1021/ac501811r | Anal. Chem. 2014, 86, 9106−9113

Analytical Chemistry



Article

DISCUSSION

This paper outlines the construction of a DoC consisting of nucleic acid-modifying, quasi-circular DNAzyme structures and then describes its application as a universal tool for signal amplification. The quasi-circles consisted of DNAzymes that were temporarily inactivated via Watson−Crick hybridization to two complementary regions of blocking oligonucleotides. Restoration of catalytic activity was then achieved via modification of an intermediate region between these complementary regions in response to the presence of specific analytes. We propose that the intermediate region of the quasicircle is universal and that a multitude of methods may be utilized to modify this region and initiate the cascade in response to a variety of targets. For proof of concept, we demonstrated this using catalytic nucleic acids (DNAzymes or MNAzymes) or protein enzymes as the initiating enzymes, which were activated in the presence of the environmental toxin Pb2+, the biological molecule dATP and nucleic acid targets homologous to genes from meningitis-causative agents. However, initiation of the DoC may not be limited to only enzymatic methods. For example, direct hybridization of a target to a BL molecule containing an aptamer specific to that target, may also be sufficient to initiate the cascade. For small molecule detection, we used an aptamer with specificity for a particular target, dATP. Aptamers have been generated for a plethora of diverse targets, including proteins, peptides, small molecules, nucleic acids and even whole cells using the SELEX (Systematic Evolution of Ligands by Exponential Enrichment) technology.17−19 The ability of aptamers to initiate DoC, means that as the catalogue of aptamers continues to expand this should allow for the detection of a very broad range of target analytes using the DoC. MNAzymes, which were used to directly detect nucleic acid targets, are also easily amenable to detect different sequences, with only the sequence of the sensor arms needing to be changed for each new target, further expanding the diagnostic repertoire available. MNAzymes are exquisitely specific and require that sensor arms be well-matched to the target sequence. Even minor sequence variations are not well tolerated by the enzyme. When combined with the DoC, this feature is amplified such that single nucleotide differences were efficiently discriminated between. In humans, such differences can include single nucleotide polymorphisms (SNPs), which can influence an individual’s susceptibility to disease and response to drugs, or they may be acquired point mutations, which can also influence response to drugs, for example, cancer therapies. For metal ion detection, we utilized the Pb2+-dependent DNAzyme GR-5. Lead sensing DNAzymes are commonly reported in the literature and have been linked with fluorescent,20,21 colorimetric21−23 and electrochemical24 outputs. The majority of studies used the 8-17 DNAzyme, which, unfortunately, is also active in the presence of a number of other ions such as zinc, manganese, cobalt and cadmium.13 In our protocols, we used the GR-5 DNAzyme, which was recently characterized as being highly specific in its requirement for Pb2+.13 This remained the case when combined with our DoC technology (Figure S-2, Supporting Information). DNAzymes generally are highly dependent on the presence of divalent metal ions and many nucleic acid-modifying DNAzymes display the same selectivity for a particular ion as the GR-5 does for

Figure 4. Further characterization of the DNA-only cascade. A) Raw fluorescent data outlining the limit of detection for the nucleic acid target. (B) Standard curve comparing point of inflection (POI) time (y-axis) vs the log of the starting concentration (x-axis) for the six highest target concentrations in panel A. (C) Comparison of the change in POI between nucleic acid targets that differ by a single nucleotide and that of the no target control. Target (A) is the wildtype sequence and was fully matched to the MNAzyme sensor arm. Error bars indicate the standard deviation of three replicates.

are derived from infectious agents responsible for causing bacterial meningitis, a disease that can progress rapidly following initial, vague “flu-like” symptoms. The simultaneous detection of these targets would therefore be highly valuable for a point of care diagnostic assay. Each active MNAzyme then initiated its own DoC via cleavage of the respective MNAzyme substrate within the intermediate region of a quasi-circular structure. Circle A contained the MNAzyme 1 substrate, Circle B contained the MNAzyme 2 substrate and Circle C contained the MNAzyme 3 substrate. To test for multiplexing capability, two different types of reactions were performed with all conditions kept constant apart from the presence of the specific quasi-circular structures. The first were singleplex reactions consisting of each individual quasi-circle only and the second were multiplex reactions containing all three quasi-circles together. Minimal difference in fluorescence signal was observed between singleplex and multiplex reactions indicating the cascades can perform independently to amplify signal following the detection of different targets (Figure 5B). 9111

dx.doi.org/10.1021/ac501811r | Anal. Chem. 2014, 86, 9106−9113

Analytical Chemistry

Article

Figure 5. Simultaneous detection of three nucleic acid targets and the amplification of signal via three, independently functioning DNA-only cascades. (A) The three quasi-circles are denoted Circle A, Circle B and Circle C. Each circle is cleaved by an MNAzyme (MNAzyme 1, 2 or 3, respectively) that assembles in the presence of a nucleic acid target (Target 1, 2 or 3, respectively). Different fluorescent dyes were utilized for each independent DNAzyme substrate; HEX for the Circle A cascade, Texas Red (TR) for the Circle B cascade and FAM for the Circle C cascade. (B) Raw fluorescent data individually shown from the HEX, TR and FAM channels indicates there are only minor differences in fluorescent signal when each cascade is present separately (solid lines) vs when multiplexed together in the same reaction tube (dashed lines).

Pb2+. This also creates the potential for additional metal ion detection that can be coupled with the DoC technology. Signal amplification technologies that are completely proteinenzyme free are currently under-represented within the field that is composed mainly of methods using one or more protein enzymes. There are many advantages to our DoC reaction, which requires only synthetic oligonucleotides. We calculated that the approximate reagent cost is around $US 0.03 per assay, which was approximately 30 times less expensive than the cheapest fluorescent probe-based quantitative PCR assay and approximately 140 times less expensive than a commercially available isothermal target amplification assay requiring polymerases and additional DNA-modifying protein enzymes. We found that performance was virtually unchanged following the storage of oligonucleotide components for 2 days at 40 °C, potentially eliminating the requirement for cold chain storage and transportation. By heat-treating human serum samples, we were able to denature RNases and allow the DoC to function in as high as 50% background serum. Other isothermal technologies have reported similar performances in concentrations of around 1−10% complex media (e.g., human serum, cell lysate) compared to their preferred buffer,25−29 but to our knowledge, do not report functionality in concentrations higher than this. The nucleic acid target detection limit of the DoC reported here is around 2 pM synthetic DNA target within a 2 h time frame. This provides a nearly 10-fold improvement over the limit we calculated using the initiating MNAzyme alone in the absence of the DoC (Figure S-5, Supporting Information).

Further, the speed of detection was reduced with the initiating MNAzyme alone requiring 2 h to detect a 20 pM target compared to only 1 h with the addition of DoC reagents. Although we have previously reported observable signal above background in the presence of a 5 pM target for a specific, highly optimized MNAzyme,6 the initiating MNAzyme used in this study was less efficient at detecting low target concentrations. Further, an advantage of the DoC comes from the production of an exponential versus linear signal, which simplifies extraction of quantitative data. A limitation of the DoC, however, is an increase in background (no target) signal compared to that of the MNAzyme alone. We propose that future improvements in sensitivity may be improved with further optimization of the quasi-circle design and oligonucleotide concentrations. Despite the current limitations of DoC, many of the existing protein-free technologies have similar detection limits but are often reported to require much longer time frames for detection of up to 12 h11,30−32 (Table S-2, Supporting Information). In addition, many may be restricted to nucleic acid detection and are difficult to multiplex and/or require multistep protocols. For the alternate circularized DNAzyme technologies, the longer time-frames required for detection may be due to the complementarity between the substrate-binding arms of the DNAzymes and the initiating nucleic acid enzyme.10,11 With our quasi-circles, the unique substrate sequences we provide allows for a distinction between the substrate-binding arms of different DNAzymes and the 9112

dx.doi.org/10.1021/ac501811r | Anal. Chem. 2014, 86, 9106−9113

Analytical Chemistry

Article

Author Contributions

initiating enzyme, thereby preventing unwanted complementarity. The fluorescent signal produced by the DoC also increased at an exponential rate and there was a strong correlation between the known concentrations of target and the time at which they would reach a threshold fluorescence level, indicating the DoC is capable of quantification. This feature is lacking from many of the previously reported signal amplification technologies. This predictability can allow for the determination of target quantities from unknown samples and is particularly important for some diagnostic applications (for example, detecting pathogen load), where this information can direct treatment, evaluate therapy and/or predict patient outcomes.1 Further, we demonstrated that several independent autocatalytic cascades could be developed and multiplexed together. This could allow for the incorporation of additional controls to validate experimental results, potentially eliminating false negatives due to failed amplification or extraction. The simultaneous detection of multiple targets is also useful for increasing throughput, which is particularly important, as biological sample volumes are often low and may not be sufficient to split among multiple reactions. Finally, the ability to multiplex is essential for the detection of diseases that are attributed to the presence of more than one target. One such example is bacterial Meningitis, which we exemplified here via the simultaneous detection of targets homologous to genes from S. agalactiae, N. meningitidis and S. pneumoniae. To our knowledge, this is the first report demonstrating multiple protein enzyme-free signal amplification cascades functioning simultaneously.

The paper was written through contributions of all authors. All authors have given approval to the final version of the paper. Notes

The authors declare the following competing financial interest(s). The following authors own stock within SpeeDx Pty Ltd: S. Bone, E. Mokany and A. Todd.

■ ■

ACKNOWLEDGMENTS This work was supported by SpeeDx Pty Ltd.



CONCLUSION The DoC can be used as a universal tool for the multiplexed detection of a broad range of input targets producing amplified, quantitative output signals in real-time. The protein enzymefree composition and isothermal functionality of the cascades confers significant potential for point-of-care diagnostic applications where simple, inexpensive and highly stable components are highly desired.



ASSOCIATED CONTENT

S Supporting Information *

Oligonucleotide sequences used for experiments, comparison of the DoC with previously published cascades which also do not require protein enzymes, schematic depicting direct target detection methods in the absence of the DoC, raw fluorescence data from an experiment demonstrating the specificity of Pb2+ detection with the DoC, raw fluorescence data from an experiment demonstrating the specificity of dATP detection with the DoC, initiation of the DoC via nucleic acid targetdirected restriction enzyme (RE) cleavage, raw fluorescence data from an experiment outlining the limit of detection for an MNAzyme detecting a nucleic acid target in the absence of the DoC, comparison of the DoC before and after incubation of oligonucleotide components at 40 °C for 48 h and raw fluorescence data derived from the DoC performed in buffer only or a background of 50% preheated human serum. This material is available free of charge via the Internet at http:// pubs.acs.org.



REFERENCES

(1) Craw, P.; Balachandran, W. Lab Chip 2012, 12, 2469. (2) Silverman, S. K. Nucleic Acids Res. 2005, 33, 6151. (3) Silverman, S. K. Acc. Chem. Res. 2009, 42, 1521. (4) Silverman, S. K.; Begley, T. P. In Wiley Encyclopedia of Chemical Biology; John Wiley & Sons, Inc.: New York, 2007. (5) Travascio, P.; Li, Y.; Sen, D. Chem. Biol. 1998, 5, 505. (6) Mokany, E.; Bone, S. M.; Young, P. E.; Doan, T. B.; Todd, A. V. J. Am. Chem. Soc. 2010, 132, 1051. (7) Mokany, E.; Tan, Y. L.; Bone, S. M.; Fuery, C. J.; Todd, A. V. Clin. Chem. 2013, 59, 419. (8) Kolpashchikov, D. M. ChemBioChem 2007, 8, 2039. (9) Elbaz, J.; Lioubashevski, O.; Wang, F.; Remacle, F.; Levine, R. D.; Willner, I. Nat. Nanotechnol. 2010, 5, 417. (10) Levy, M.; Ellington, A. D. Proc. Natl. Acad. Sci. U. S. A. 2003, 100, 6416. (11) Elbaz, J.; Moshe, M.; Shlyahovsky, B.; Willner, I. Chemistry 2009, 15, 3411. (12) Breaker, R. R.; Joyce, G. F. Chem. Biol. 1994, 1, 223. (13) Lan, T.; Furuya, K.; Lu, Y. Chem. Commun. (Cambridge, U. K.) 2010, 46, 3896. (14) Huizenga, D. E.; Szostak, J. W. Biochemistry 1995, 34, 656. (15) Wang, G. L.; Zhang, C. Y. Anal. Chem. 2012, 84, 7037. (16) Li, J. J.; Chu, Y.; Lee, B. Y.; Xie, X. S. Nucleic Acids Res. 2008, 36, e36. (17) Khan, H. Gene Ther. Mol. Biol. 2008, 12, 111. (18) Sampson, T. World Pat. Inf. 2003, 25, 123. (19) Sassolas, A.; Blum, L. J.; Leca-Bouvier, B. D. Analyst 2011, 136, 257. (20) Zhang, L.; Han, B.; Li, T.; Wang, E. Chem. Commun. (Cambridge, U. K.) 2011, 47, 3099. (21) Shimron, S.; Elbaz, J.; Henning, A.; Willner, I. Chem. Commun. (Cambridge, U. K.) 2010, 46, 3250. (22) Bi, S.; Yan, Y.; Hao, S.; Zhang, S. Angew. Chem., Int. Ed. 2010, 49, 4438. (23) Miao, X.; Ling, L.; Shuai, X. Chem. Commun. (Cambridge, U. K.) 2011, 47, 4192. (24) Yang, X.; Xu, J.; Tang, X.; Liu, H.; Tian, D. Chem. Commun. (Cambridge, U. K.) 2010, 46, 3107. (25) Hu, R.; Fu, T.; Zhang, X. B.; Kong, R. M.; Qiu, L. P.; Liu, Y. R.; Liang, X. T.; Tan, W.; Shen, G. L.; Yu, R. Q. Chem. Commun. (Cambridge, U. K.) 2012, 48, 9507. (26) Huang, Y.; Chen, J.; Zhao, S.; Shi, M.; Chen, Z. F.; Liang, H. Anal. Chem. 2013, 85, 4423. (27) Yan, C.; Jiang, C.; Jiang, J.; Yu, R. Anal. Sci. 2013, 29, 605. (28) Abe, T.; Segawa, Y.; Watanabe, H.; Yotoriyama, T.; Kai, S.; Yasuda, A.; Shimizu, N.; Tojo, N. Lab Chip 2011, 11, 1166. (29) Song, P.; Xiang, Y.; Xing, H.; Zhou, Z.; Tong, A.; Lu, Y. Anal. Chem. 2012, 84, 2916. (30) Wang, F.; Elbaz, J.; Orbach, R.; Magen, N.; Willner, I. J. Am. Chem. Soc. 2011, 133, 17149. (31) Wang, F.; Elbaz, J.; Teller, C.; Willner, I. Angew. Chem., Int. Ed. 2011, 50, 295. (32) Wang, F.; Lu, C. H.; Willner, I. Chem. Rev. 2014, 114, 2881.

AUTHOR INFORMATION

Corresponding Author

*S. M. Bone. E-mail: [email protected]. 9113

dx.doi.org/10.1021/ac501811r | Anal. Chem. 2014, 86, 9106−9113