Programming a topologically constrained DNA nanostructure into a sensor

Abstract

Many rationally engineered DNA nanostructures use mechanically interlocked topologies to connect individual DNA components, and their physical connectivity is achieved through the formation of a strong linking duplex. The existence of such a structural element also poses a significant topological constraint on functions of component rings. Herein, we hypothesize and confirm that DNA catenanes with a strong linking duplex prevent component rings from acting as the template for rolling circle amplification (RCA). However, by using an RNA-containing DNA [2] catenane with a strong linking duplex, we show that a stimuli-responsive RNA-cleaving DNAzyme can linearize one component ring, and thus enable RCA, producing an ultra-sensitive biosensing system. As an example, a DNA catenane biosensor is engineered to detect the model bacterial pathogen Escherichia coli through binding of a secreted protein, with a detection limit of 10 cells ml(-1), thus establishing a new platform for further applications of mechanically interlocked DNA nanostructures.

Figure 1. Inability of a D2C to undergo RCA.

(a) Schematic illustration of a D2C made of ^CDNA_i and r^CDNA_ii, with a linking duplex of 24 bp. (b) Sequences of rD2C1, DP1 and DP2. Boxed nucleotides represent the 24-bp linking duplex. F: fluorescein-dT; R, adenosine ribonucleotide; Q, dabcyl-dT. (c) Synthesis of rD2C1 by circularizing linear DNA_ii over ^CDNA_i as the template. Lane M, markers made of ^LDNA_ii, ^CDNA_i and r^CDNA_ii. Lane R: circularization mixture. (d) RCA reactions with gel-purified ^CDNA_i, r^CDNA_ii and rD2C1 using DP1 and DP2 as primers. Lane L, DNA ladders ranging from 1 to 10 kbp; RP, RCA product.

Figure 2. Cleavage of an RNA-containing D2C by an RCD.

(a) Restoration of RCA compatibility of an rD2C using an RCD. (b) Cleavage of rD2C1 by EC1, an E. coli-responsive DNAzyme. Concentration of E. coli: 10⁵ cells ml⁻¹. Reaction mixtures were analyzed by 10% denaturing PAGE. EC1M: a mutant EC1 that cannot be activated by E. coli. Both r^CDNA_ii and ^CDNA_i in rD2C1 were radioactively labelled with ³²P to facilitate DNA visualization on the gel. Clv%: per cent cleavage.

Figure 3. 3′–5′ exonucleolytic activity of ϕ29DP on r^CDNA_ii and rD2C1.

Degradation of EC1-mediated cleavage product of r^CDNA_ii (a,b) and rD2C1 (c) by ϕ29DP and PNK. Concentration of E. coli: 10⁵ cells ml⁻¹. Reaction mixtures were analyzed by 20% denaturing PAGE. LF, large DNA fragment; SF, small DNA fragment. M lanes contain various DNA markers as indicated. r^CDNA_ii, both r^CDNA_ii and ^CDNA_i in rD2C1 were radioactively labelled with ³²P to facilitate DNA visualization on the gel.

Figure 4. E. coli-dependent RCA reaction.

(a) RCA reactions of rD2C1 in the presence of E. coli (10⁵ cells ml⁻¹) analyzed using 0.6% agarose gel electrophoresis. Note every reaction also contained PNK and dNTPs. L, DNA ladders ranging from 1 to 10 kbp; RP, RCA product. (b) Determination of detection sensitivity through analysis of RP using 0.6% agarose gel electrophoresis. (c) Determination of detection sensitivity via the colourimetric assay enabled by PW17 peroxidase DNAzyme. (d) Analysis of assay specificity using the colourimetric assay. The gram-negative bacteria used were Serratia fonticola (SF), Achromobacter xylosoxidans (AX), Yersinia ruckeri (YR) and Hafnia alvei (HA). The gram-positive bacteria used were Leuconostoc mesenteroides (LM) and Pediococcus acidilactici (PA).

Figure 5. E. coli-dependent HRCA reaction.

(a) Schematic illustration of HRCA. FP1, forward primer; RP1, reverse primer. (b) Denaturing PAGE analysis of HRCA products. RP, RCA products; SA, secondary amplicons produced from the initial RCA products. (c) Real-time monitoring of HRCA reactions at various E. coli concentrations (cells ml⁻¹).