Rational design of DNA nanostructures for single molecule biosensing

The ability to detect low concentrations of biomarkers in patient samples is one of the cornerstones of modern healthcare. In general, biosensing approaches are based on measuring signals resulting from the interaction of a large ensemble of molecules with the sensor. Here, we report a biosensor platform using DNA origami featuring a central cavity with a target-specific DNA aptamer coupled with a nanopore read-out to enable individual biomarker detection. We show that the modulation of the ion current through the nanopore upon the DNA origami translocation strongly depends on the presence of the biomarker in the cavity. We exploit this to generate a biosensing platform with a limit of detection of 3 nM and capable of the detection of human C-reactive protein (CRP) in clinically relevant fluids. Future development of this approach may enable multiplexed biomarker detection by using ribbons of DNA origami with integrated barcoding.


Supplementary Figures
Concentric squares translocation peak characterisation. DNA aptamer characterisation Supplementary Figure 3: SPR studies for CRP aptamer selection: a) Schematic of DNA aptamer attachment to SPR sensor surface. A mercapto-PEG-COOH was formed on the SPR gold surface for the attachment of amine-modified DNA aptamers via EDC. b) The SPR sensogram for the two DNA aptamers shows that they both bind CRP and hence are suitable for incorporation into DNA origami to be used as carriers. c) Concentration-dependent signals for different concentration of CRP from 500 pg/ml to 10 µg/ml against aptamer 1. d) SPR response plotted against CRP concentration. The solid line is a fit of the Langmuir isotherm to the data with a KD of ~40 nM.
In addition to nanopipette translocation experiments, a gel electrophoresis experiment (2 hours at 80 V at 4C) was carried out to confirm CRP binding to the carrier (figure S4). Carriers (9 nM) were incubated for 30 min with increasing concentrations of CRP (9 nM, 18 nM, 27 nM, and 36 nM; lanes 4-7). For the lowest CRP concentration, two bands can be seen, one at the same position as the band resulting from the unoccupied carriers (lanes 2 and 8), and one slightly retarded, suggesting that CRP-occupied carriers run through the gel slower than unoccupied carriers. The carrier with a nonspecific aptamer (lane 3) only produces one band at the same position as the unoccupied carriers, confirming that no CRP binding takes places without a CRP-specific DNA aptamer.

Characterisation of carriers
Supplementary Figure 6a: Scatter plot of peaks for (i) unoccupied carriers and (ii) carriers incubated with a 10x excess of CRP (90 nM). Ion current events which are double peaks are plotted orange, and the ones that fall inside the 95% confidence ellipse are plotted as circles, the others as triangles. Only events that fall within the 95% confidence ellipse are considered double peaks. The same analysis was carried out for single peaks (plotted in blue). Ion current events which resembled neither a double nor a single peak are shown as black triangles and are excluded from the analysis. . Legend abbreviations, dp (orange circles) -double peaks, sp (blue circles)single peaks, ncp (black triangles) -non-classified peaks, ncsp (blue triangles) -non-classified single peaks, ncdp (orange triangles) -non-classified double peaks, dp ellipse (orange ellipse) -95% confidence ellipse for dp events, sp ellipse (blue ellipse) -95% confidence ellipse for sp events. (iii) Scatter plot of peak amplitude versus dwell time, with the 95% confidence ellipses overlaid. All observed peaks (black cross) fall outside the overlaid confidence ellipses. Legend abbreviations, black cross -observed peaks (events), dp ellipse (orange ellipse) -95% confidence ellipse for dp events, sp ellipse (blue ellipse) -95% confidence ellipse for sp events.
Supplementary Figure 6c: The DNA nanostructures were incubated in translocation buffer at room temperature for 4 hours to assess the impact of the 2 mM Ca 2+ on the stability of the DNA origami structure. (i) AFM micrograph showing a mixture of intact and degraded carriers (indicated by white arrows). (ii) Peak amplitude and dwell time histograms of the translocation ion current peaks. A range of different peak shapes were observed, including single(blue triangles -sp) and double peaks(orange triangles -dp) and single peaks with shoulders(pink triangles -sp with shoulder). (iii). Scatter plot of peak amplitude versus dwell time, with the 95% confidence ellipses overlaid. Only around 10% of peaks which resemble single peaks also fall into the single peak 95% confidence ellipse and hence would lead to false positives. Legend abbreviations, dp (orange triangles)double peaks, sp (blue triangles) -single peaks, sp with shoulder (pink triangles) -single peaks with shoulder, ncsp (black triangles) -non-classified single peaks, dp ellipse (orange ellipse) -95% confidence ellipse for dp events, sp ellipse (blue ellipse) -95% confidence ellipse for sp events. The individual observed ion current peaks are classified as described in the main text and colour-coded (double peaks orange and single peaks blue, triangles represent unclassified events which are not considered for analysis). The orange and blue ellipses represent the 95% confidence ellipses for double and single peaks and are taken from the analysis described in figure S6a. The panels for 9 nM and 90 nM are the same as presented in figure 3b. Legend abbreviations, dp (orange circles) -double peaks, sp (blue circles) -single peaks, ncp (black triangles) -non-classified peaks, ncsp (blue triangles) -non-classified single peaks, ncdp (orange triangles) -non-classified double peaks, dp ellipse (orange ellipse) -95% confidence ellipse for dp events, sp ellipse (blue ellipse) -95% confidence ellipse for sp events.
Supplementary Figure 8: a) Ion current peak analysis for carrier made with non-specific aptamer challenged with 90 nM CRP. Experimental conditions are the same as used for specific sensing at the highest CRP concentration, i.e. 9 nM carrier and 90 nM CRP. A peak amplitude and dwell time of 69±5 pA and 0.3±0.13 ms, respectively, were found, which is in line with those for an empty carrier. The scatter plot analysis which shows the classification of the ion current events as described in the main text shows that no occupied carriers were detected, demonstrating the high specificity of the sensing approach. b) Ion current peak analysis for CRPspecific carriers challenged with 90 nM of a different protein (MupB). Experimental conditions are the same as used for specific sensing at the highest CRP concentration, i.e. 9 nM carrier and 90 nM MupB. A peak amplitude and dwell time of 60±7 pA and 0.25±0.1 ms, respectively, were found, which is again similar to those for an empty carrier. c) The scatter plot analysis which shows the classification of the ion current events as described in the main text shows that no occupied carriers were detected, demonstrating the high specificity of the sensing approach. Legend abbreviations, dp (orange circles) -double peaks, ncp (black triangles) -non-classified peaks, dp ellipse (orange ellipse) -95% confidence ellipse for dp events.
Supplementary Figure 9: Scatter plot of peak amplitude versus dwell time for carriers (9 nM) subjected to four different concentrations of CRP (0 nM, 3 nM, 9 nM and 36 nM) in plasma. For 0 nM, no single peaks were observed, and the cluster of double peaks was used to define the 95% confidence interval for the unoccupied carrier. Similarly, the 95% confidence ellipse for the single peaks was generated from the data of the highest CRP concentration. Legend abbreviations, dp (orange circles) -double peaks, sp (blue circles) -single peaks, ncp (black triangles) -non-classified peaks, ncdp (orange triangles) -non-classified double peaks, dp ellipse (orange ellipse) -95% confidence ellipse for dp events, sp ellipse (blue ellipse) -95% confidence ellipse for sp events. Single-stranded DNA scaffold design and routing map for concentric squares origamis For this study, a modified phagemid (plasmid with m13 region) was used to create the three scaffolds of varying lengths to create the DNA origamis with similar outer but different cavity dimensions. The scaffolds required for the three DNA origamis (concentric squares (Con) A, B and C) were stitched together from two separate fragments derived from lambda phage DNA along with pBluescript II SK(+) phagemid (PBS). The PBS(+) are plasmids containing the phage origin of replication (f1 ori) and an antibiotic resistance gene which allows for sense strand rescue by a helper phage. The 9 kb custom scaffold required for the DNA origami ConA was created by inserting a 6.1kb sequence, derived as two separate fragments from lambda phage, into the 2.9 kb phagemid. These two fragments (approximately 1.8 kb and 4.2 kb) were selected devoid of protein coding regions to eliminate any interference in the phage growth and downstream process. The map for scaffold assembly is provided in Figure S11b and S11c.

Supplementary Figure 11a: Schematic of concentric square scaffold design. a) Scaffold for ConA DNA origami tile indicated with the upper (ConB_U, ConC_U) and lower (ConB_L, ConC_L) primers for deriving b) ConB and c) ConC scaffolds respectively.
The three fragments that make up the scaffold -referred to as f1, f2 and f3 -were PCR amplified from lambda phage DNA and the PBS plasmid, respectively, using appropriate primers. The primers used in the amplification procedure were designed to incorporate the restriction sites EcoRI and NspI flanking f2, and NcoI and NspI flanking f3. The two fragments were then restriction digested with the respective enzymes to form compatible overhangs, and then ligated to yield one long fragment (f23). Fragment f1 was then combined with fragment f23 using HiFi DNA assembly cloning (NEB,UK), resulting in the double stranded circular f123 DNA.
Subsequently the circular dsDNA (f123) was chemically transformed in competent E.Coli cells to produce the custom scaffold. The transformed cells with the f123 plasmid were selected using their antibiotic resistance gene and recovered with a miniprep to be cotransformed with helper plasmids (m13cp cells) to produce ssDNA f123. The final ssDNA f123 product constituted 9073 bp and was used as the scaffold for the largest DNA construct ConA.
The scaffolds for the two smaller DNA constructs, ConB and ConC, were produced by PCR amplification of the double stranded f123 as template. As before the primers were designed to incorporate the restriction site KasI for the product ConB and BmtI for the product ConC. Subsequent restriction and ligation followed by similar co-transformation protocol with helper plasmids as above resulted in 8515 bp and 6307 bp ssDNA scaffolds for Con B and C, respectively. In the first instance, the scaffolds were produced in 5 ml cultures and subsequently followed by large-scale scaffold harvesting by scaling up to 1-litre cultures. The cultures were harvested by separating the bacteria from the phage and subsequent phenol chloroform DNA extraction. Biosensor with CRP-specific aptamer 2 Supplementary figure 12: Scatter plot of peaks for (i) unoccupied carriers and (ii) carriers (6 nM) incubated with a 17x excess of CRP (100 nM). Ion current events which are double peaks are plotted orange, and the ones that fall inside the 95% confidence ellipse are plotted as circles, the others as triangles. Only events that fall within the 95% confidence ellipse are considered double peaks. The same analyses were carried out for single peaks (plotted in blue). Ion current events which resembled neither a double nor a single peak are shown as black triangles and are excluded from the analysis. Legend abbreviations, dp (orange circles) -double peaks, sp (blue circles) -single peaks, ncp (black triangles) -non-classified peaks, ncsp (blue triangles) -non-classified single peaks, dp ellipse (orange ellipse) -95% confidence ellipse for dp events, sp ellipse (blue ellipse) -95% confidence ellipse for sp events.
Supplementary Figure 13b: Scatter plot of ion current events plotted as peak amplitude versus dwell time for carriers incubated with different CRP concentration (6 to 100 nM). The ion peak signatures of double peak and single peaks observed for the samples are colour-coded orange and blue respectively and indicated by a 95% confidence eclipse obtained from figure S12, unclassified events are represented as triangles. Legend abbreviations, dp (orange circles) -double peaks, sp (blue circles) -single peaks, ncp (black triangles) -nonclassified peaks, ncsp (blue triangles) -non-classified single peaks, ncdp (orange triangles) -non-classified double peaks, dp ellipse (orange ellipse) -95% confidence ellipse for dp events, sp ellipse (blue ellipse) -95% confidence ellipse for sp events.
Supplementary figure 13c: Normalised single peak count, i.e. ratio of single peaks vs total classified peaks against CRP concentration (table S5). The data were fitted with Langmuir isotherm (solid line) and revealed a Kd of 12±5 nM. The dashed lines represent the confidence boundaries of the fit. The error bars denote standard deviation of ion current translocation experiments conducted on different days using three different nanopipettes.

Schematic of DNA origami carrier design
Supplementary figure 14: CaDNAno scaffold routing of the DNA origami frame used as the carrier. The scaffold strand is coloured blue whereas the numerous short staple stranded are presented in pink.