Specific detection of DNA and RNA targets using a novel isothermal nucleic acid amplification assay based on the formation of a three-way junction structure

Abstract

The formation of DNA three-way junction (3WJ) structures has been utilised to develop a novel isothermal nucleic acid amplification assay (SMART) for the detection of specific DNA or RNA targets. The assay consists of two oligonucleotide probes that hybridise to a specific target sequence and, only then, to each other forming a 3WJ structure. One probe (template for the RNA signal) contains a non-functional single-stranded T7 RNA polymerase promoter sequence. This promoter sequence is made double-stranded (hence functional) by DNA polymerase, allowing T7 RNA polymerase to generate a target-dependent RNA signal which is measured by an enzyme-linked oligosorbent assay (ELOSA). The sequence of the RNA signal is always the same, regardless of the original target sequence. The SMART assay was successfully tested in model systems with several single-stranded synthetic targets, both DNA and RNA. The assay could also detect specific target sequences in both genomic DNA and total RNA from Escherichia coli. It was also possible to generate signal from E.coli samples without prior extraction of nucleic acid, showing that for some targets, sample purification may not be required. The assay is simple to perform and easily adaptable to different targets.

Figure 1

The SMART assay. (A) Formation of a 3WJ. Extension and template probes anneal to the target, and only then to each other (overlap between the two probes is only 8 bp). The short extension probe has a free 3′-OH to allow extension. The template probe includes a single-stranded (non-functional) T7 RNA polymerase promoter (Pr) and sequences to allow the capture and detection of the RNA signal. The 3′ end of the template probe is blocked (x) by phosphorylation to prevent extension. (B) Extension and transcription generate an RNA signal. Bst DNA polymerase extension of the extension probe generates a double-stranded (ds), hence functional, T7 RNA polymerase promoter (Pr), allowing transcription of multiple copies of an RNA signal (RNA1) by T7 RNA polymerase. If required, RNA1 anneals to a second template (probe for RNA amplification), leading to further extension and transcription by the DNA and RNA polymerases to generate increased amounts of a second RNA signal (RNA2). (C) Detection and quantification of the RNA signal by ELOSA. Specific sequences included in RNA signals 1 and 2 allow capture, via a biotinylated probe, onto the streptavidin-coated well of a microtitre plate and detection and quantification via an alkaline phosphatase (AP)-linked probe. Wash steps remove unbound probe and the colour change of AP substrate (4-nitrophenyl phosphate) is followed for 30 min (37°C) at 405 nm. By comparing the AP activities of different samples with a standard curve, the relative amounts of signal may be calculated.

Figure 2

(A) Comparison of signals from different synthetic targets. Synthetic target sequences were designed for: CFTR, human cystic fibrosis transmembrane conductance regulator gene (GenBank accession no. 6995995); Mtb, the IS6110 sequence from M.tuberculosis (GenBank accession no. 6523392); and HBV, the small surface antigen of Hepatitis B virus, isolate rbo11 (GenBank accession no. 6684102). Extension and template probes were identical, except that the target-hybridising regions were altered, making them specific for the different targets. Probes were added at 60 fmol extension and 50 fmol template for 50 fmol target. Negative controls contained no target. The amount of RNA signal produced was determined by ELOSA (detecting RNA1). (B) Signal generated from 1 fmol target. CFTR probes were added at 50 fmol extension and 10 fmol template for 1 fmol target. Other reaction conditions were unaltered. Negative controls contained no target.

Figure 3

Comparison of the signals generated from DNA or RNA targets. Both targets (single-stranded, synthetic oligonucleotides) had the CFTR sequence (GenBank accession no. 6995995), identical except for the T or U residues and the dNTPs or NTPs. The same probes were used for both targets, added at 60 fmol CFTR-ext and 50 fmol CFTR-tem for 50 fmol target. Negative controls contained no target oligonucleotide. The amount of RNA signal produced was determined by ELOSA (detecting RNA1).

Figure 4

Effect of a linker molecule in the template probe. 23S-tem probes were used with or without a hexaethylene glycol linker at the 3WJ site. All other probes were identical in the different reaction samples. Negative controls contained no target oligonucleotide. All 3WJ reaction samples, including the no target controls, contained 100 ng non-target genomic DNA (M.lysodeikticus). The RNA signal produced was amplified further before signal detection and quantification by ELOSA (detecting RNA2). Probe concentrations were 5 fmol 23S-ext and 1 fmol 23S-tem with 20 fmol probe for RNA amplification for 50 amol 23S synthetic target.

Figure 5

Generation of signals from complex targets. (A) Detection of the gene encoding 23S rRNA in genomic DNA from E.coli K12 MG1655. Genomic DNA extracted from mid log phase cells was quantified and used in SMART reactions. All 3WJ reaction samples, including those containing no E.coli DNA, were made up to a total of 50 ng genomic DNA with non-target (Micrococcus sp.) DNA. The RNA signal produced was amplified further before signal detection and quantification by ELOSA (detecting RNA2). Results of triplicate reactions are shown. (B) Detection of 23S rRNA from E.coli K12 MG1655 total RNA. Total RNA extracted from mid log phase cells was quantified and used in SMART reactions. All 3WJ reaction samples, including those containing no E.coli RNA, contained 100 ng non-target genomic DNA (Micrococcus sp.). The RNA signal produced was amplified further before signal detection and quantification by ELOSA (detecting RNA2). Results of triplicate reactions are shown. The same probes (23S-ext and 23S-tem) were used for the DNA and RNA targets: and were used at 5 fmol extension and 1 fmol template (containing linker molecule), with 20 fmol probe for RNA amplification.

Figure 6

Detection of 23S target in a crude E.coli culture. Escherichia coli K12 MG1655 was grown to mid log phase in tryptone broth. SMART reactions were set up, using dilutions of a fresh E.coli culture (2 or 1 µl neat, or 1 µl of various dilutions in sterile tryptone broth) as the target. Cell counts were performed on the culture used. One microlitre of neat culture contained 3 × 10⁵ cells. The no target control sample (NT) contained 1 µl sterile tryptone broth. All 3WJ reaction samples contained 100 ng non-target nucleic acid (Micrococcus sp. genomic DNA). Standard reaction conditions were used, amplifying the RNA signal produced from the 3WJ before signal detection and quantification by ELOSA (detecting RNA2). Results of duplicate reactions are shown. Probe concentrations were 5 fmol 23S-ext, 1 fmol 23S-tem (containing linker molecule) and 20 fmol probe for RNA amplification.