DNA Strand Displacement Driven by Host–Guest Interactions

Base-pair-driven toehold-mediated strand displacement (BP-TMSD) is a fundamental concept employed for constructing DNA machines and networks with a gamut of applications—from theranostics to computational devices. To broaden the toolbox of dynamic DNA chemistry, herein, we introduce a synthetic surrogate termed host–guest-driven toehold-mediated strand displacement (HG-TMSD) that utilizes bioorthogonal, cucurbit[7]uril (CB[7]) interactions with guest-linked input sequences. Since control of the strand-displacement process is salient, we demonstrate how HG-TMSD can be finely modulated via changes to the structure of the input sequence (including synthetic guest head-group and/or linker length). Further, for a given input sequence, competing small-molecule guests can serve as effective regulators (with fine and coarse control) of HG-TMSD. To show integration into functional devices, we have incorporated HG-TMSD into machines that control enzyme activity and layered reactions that detect specific microRNA.


S7. References
Here, [S] is the total concentration of the substrate p-NPA (1.0

mM). The Michaelis−Menten constant Km
(the substrate concentration that gives a reaction rate equal to one-half the maximum rate) was first obtained for p-NPA hydrolysis catalyzed by hCA-II in the absence of inhibitors by fitting the data (Supplementary Figure 35) to the Michaelis−Menten equation (2) where V0 is the initial velocity and Vmax is the maximum velocity of the reaction. The Km value was calculated to be 4.4 ± 0.5 mM, and Vmax was 3.97± 0.14 x10 -3  was also 1.58 nmol.

d) Selective detection of miR-182.
To initiate the sequential reaction, 0.1 nmol of S:R duplex (in 100 µL assay buffer) was added to the rinsed beads followed by the addition of 1.0 nmol of miR-182 (in 100 µL assay buffer), and incubated at RT for 1 hour, at which point the MB was magnetically separated, and the fluorescence of the solution analyzed. In order to probe the selectivity of the system, the same protocol was followed using two other miR sequences: miR-183 and miR-381. When S* (control DNA without the host-CB [7] modification) was investigated, 0.1 nmol S*:R duplex was incubated with 1.0 nmol miR-182 and the same protocol was followed.
For the miR-detection system, the fluorescence intensity was monitored against wavelength on a SpectraMax M3 Multi-Mode Microplate Reader using Costar-96 well standard black plates. An excitation wavelength of 470 nm was used, and emission data was collected at each nanometer with a spectral maximum observed at 522 nm. these studies, to the preformed S:R duplex (1 μM) was first incubated with 10 eq. of CGs for 3 hr, followed by incubation with 10 eq. of input C6-Ad for 18 hr. (B) Native PAGE Lanes; 1 = R only, 2 = S:R control, 3 = S:R + CG3 + C6-Ad, 4 = S:R + CG2 + C6-Ad, 5 = S:R + CG1 + C6-Ad, 6 = S:R + C6-Ad, 7 = premade S:C6-Ad product control, 8 = C6-Ad only. PAGE was carried out in 10% TBE buffer at 95 V for 5 hr at RT. This study confirmed that C6-Ad is capable of completely transforming S:R duplex into product duplex in the presence of CG2 at longer incubation period (compared to main text Figure 4B). The sulfonamide headgroup containing DNA sequences, Ei (n=1 and n=2), were synthesized as follows:

S2. Synthesis of DNA Conjugates
Inhibitor small molecule (13.5 µmol) was dissolved in 100 μL 1:1 H2O/DMSO. This solution was added to 0.3 μmol NHS-dT modified oligonucleotide (on CPG beads) followed by the addition of 10 μL DIPEA. The mixture was stirred at room temperature for 3 hours. After centrifugation and removal of the supernatant, the CPG beads were washed with 3 × 1000 μL DMSO and 3 × 1000 μL H2O. The oligonucleotide sequence was then cleaved from the beads and globally deprotected with 1000 μL of ammonium hydroxide (30%) at 55°C for 18 hours. The supernatant was separated and dried with a speedvac concentrator. The resulting residue was dissolved in 100 μL H2O, desalted with a G-25 spin column, purified via RP-HPLC, and characterized with MALDI-TOF. See Supplementary Figure 19 and 20 for their characterization details.

S2.5. RP-HPLC purification of DNA conjugates
All functionalized DNA sequences were purified via sephadex resin microspin G-25 columns (Sigma Aldrich) followed by chromatographic separation using RP-HPLC. The RP-HPLC purification of the synthetic DNA samples were carried out using a Varian Prostar HPLC system, equipped with an Agilent PLRP-S 100 Å 5 μm 4.6 × 250 mm reverse phase column. The column was maintained at 65 °C for all runs. The flow rate was set at 1 mL/min. A gradient composed of two solvents (solvent A is 0.1 M TEAA in 5% acetonitrile and solvent B is 100% acetonitrile) was used and UV-vis absorption was monitored at 260 nm. It is assumed that the difference in extinction coefficients (at 260 nm) of starting material and product species is negligible. The HPLC traces shown in the Supplementary Information S3 were obtained by rerunning the already purified samples. The purity percentages were obtained by using the integrated peak areas of all the peaks in the RP-HPLC trace. The concentrations of purified DNA samples were quantified based on their UV-vis absorption at 260 nm and their molar extinction coefficients obtained by nearest neighbor calculations.

S2.6. MALDI-TOF MS analysis of DNA conjugates
All RP-HPLC purified fractions were concentrated with a speedvac concentrator and redissolved in ultrapure water prior to MALDI analysis. MALDI-TOF MS characterization was acquired using a Bruker In an effort to gauge how HG-TMSD compares to BP-TMSD, we prepared an analogous BP-TMSD system (S':R + I'), shown in the inset. Here, the substrate strand S' contains the core DNA domain (c') but incorporates a 6-base toehold (d'). Since in the HG-TMSD system, the substrate S includes a synthetic linker (containing 12 bonds) between the CB7 toehold and the core DNA domain, we also included a C12 linker separating the toehold and the core DNA domains in S'. Similarly, the input sequence I' also contains a C12 linker separating the core domain c and the 6-base toe d.
Since 10 eq. of input sequence was used (compared to substrate duplex), we fit the kinetic data to a pseudofirst order equation (1)  We also compared the displacement kinetics of the HG-TMSD system with conventional Watson-Crick-Franklin base-pair derived strand displacement in the absence of a remote toehold. For this purpose, we used a 6 base toehold and toe domain, on the substrate and input strands, respectively (similar to that of the remote toehold BP-TMSD system described above). Here, the apparent rate constant k was found to be 1.23 ± 0.17 x 10 -2 s -1 , a value that is 20 times faster than the rate constant observed for BP-TMSD with the remote toehold. This result is consistent with the literature because introduction of spacer groups in BP-TMSD is known to slow down the rate of displacement reactions when compared to those that function via contiguous DNA domains. 6 Interestingly, we noted that the HG-TMSD reaction exhibits displacement kinetics that is in between the conventional BP-TMSD and the remote BP-TMSD. Taken together, these studies demonstrate that the CB7-adamantane host-guest interaction is an effective mechanism for initiating DNA strand displacement.

Supplementary Figure 31-ii. Comparison of displacement kinetics decay profiles for BP-TMSD under competitive and
non-competitive conditions. For the competitive conditions: 1 eq. of S":R duplex (1 µM) was first incubated with 10 eq. of competing CS sequence, and upon addition of 10 eq. of I", displacement kinetics were monitored over 1.5x10 3 s.
In order to compare the HG-TMSD and BP-TMSD displacement kinetics under the influence of competing agents (see main text Figure 4 for HG-TMSD displacement kinetics in the presence of competing small molecule guests), competitive conditions were mimicked in the conventional BP-TMSD system by addition of 10 eq. of a 6-nucleotide competitive sequence, CS (with sequence d). Substrate duplex (1 µM) was incubated with 10 eq. of CS for 3 hours, followed by the addition of 10 eq. of input strand (I'') containing sequence c + d. The kinetic decay profile in the presence of the CS (blue) was compared to that in the absence of competitive conditions (orange). BP-TMSD under competitive conditions exhibited an apparent rate constant k of 1.06 x 10 -2 s -1 which did not significantly differ from the conventional BP-TMSD rate constant in the absence of the competitive conditions (1.23 x10 -2 s -1 ). These results suggest that unlike HG-TMSD, the rate of BP-TMSD cannot be readily modulated by using such competitive agents.

S4.2. HG-TMSD mutation studies
In conventional BP-TMSD, the introduction of mutations in the input sequence can lead to decrease in the displacement rate since the base mismatch between the input and substrate strands enhances the rate of The fluorescence-quenching experiments show a significant decrease in the reaction rate when the mutation is in proximity to the guest head-group. Specifically, M3 C6-Ad (purple line), that has a mutation three residues from the 3' end, did not reach the 80% signal level. As the mutation is introduced further away from the guest toe, the lesser the effect on the displacement kinetics. For example, the M14 C6-Ad (green line), with the mutation near the middle of the input sequence, has a t0.8 of 4.4x10 3 s, which is ca. 7 fold slower than the "wild-type" C6-Ad (0.6 ± 0.1 x 10 3 s). Further, when M25 C6-Ad (which has the mutation on the 5' terminuspink line) is used as the input, no appreciable change in the rate is observed compared to wild-type C6-Ad. This latter mutation does not influence the rate of HG-TMSD since the branch migration step is essentially complete (the reporter is already dissociated) when the mismatch base-pairing between S and input M25 C6-Ad is attempted.
Kd for the S:Ei duplex was found to be 7.62 x10 -6 M. Interestingly, the Ei sequence that only has one methylene spacer (n=1) exhibited a Kd of 12.04 x10 -6 M for the single strand state which is ca. ~3 times weaker than for the corresponding single stranded Ei sequence with two methylene spacers.
The Chi  also conducted by using the small molecule inhibitor, 4-(aminomethyl) benzenesulfonamide that resulted in 24% hCA-II activity (the same level of activity observed for Ei (n=1); see main text Figure 5D-red bar).
Taken together, these controls further confirm that the effect of hCA-II inhibition is due to the single stranded Ei binding to hCA-II via the sulfonamide headgroup and that the host-guest interactions are needed to facilitate the strand displacement. Km (the substrate concentration that gives a reaction rate equal to one-half the maximum rate) was first obtained for p-NPA hydrolysis catalyzed by hCA-II in the absence of inhibitors by fitting the data (Supplementary Figure 35) to the Michaelis−Menten equation (2) where V0 is the initial velocity and Vmax is the maximum velocity of the reaction.

S4.4.1. Determination of Anchor duplex loading on MB
In order to determine the concentration of anchor duplex A:C6-Ad that was loaded onto the magnetic beads, we first prepared a duplex analog that was fluorescent. Specifically, a 5'-biotin and 3'-fluorescein into the magnetic bead displacement assay (see Figure 6A, main text). A LOD was determined using 3 SD below the averaged blank signal. (B) The linear portion of the graph was used to calculate the LOD value (using the linear fit Y= mX + C). The LOD was found to be 6.3 pmol. Buffer used in these studies: 20 mM Tris, 10 mM NaCl, 5 mM MgCl2, pH = 7.5. Results are the mean ± SD of 3 independent experiments.

S5. Supplementary Tables
Note: S* and I* are the core, unmodified, DNA sequences lacking modifications for host and guest attachments (as well as linkers), respectively. S' and I' are core DNA sequences that have an internal 12 carbon spacer (X = spacer C12 CE phosphoramidite) within the sequence. S" and I" are core DNA sequences without any modifications.  in 5% acetonitrile and solvent B is 100% acetonitrile.

S6.1. Synthesis and characterization of DNA-small molecule conjugates
DNA sequences used in this set of experiments are mentioned in the Supplementary Table 5.
Synthesis of all CB [7] conjugated DNA sequences (S; 25mer/ 8mer/ 7mer/ 5mer/ 4mer) followed the synthetic scheme shown in Supplementary Figure 1. Synthesis of all adamantane conjugated DNA sequences followed the synthetic scheme shown in the Supplementary Figure 2. All DNA conjugates were

S6.2. General experimental of thermal denaturation studies
UV-Vis spectroscopy followed melting curves of the duplexes were collected in buffer containing 20 mM Tris, 10 mM NaCl, and 5 mM MgCl2 at pH = 7.5. Spectra were collected in a closed capped quartz cuvette, which had a pathlength 1 cm. Absorbance was monitored at 260 nm while ramping the sample from 20°C to 90°C and back to 20°C at a rate of 2 degrees per minute. All duplexes were denatured at a concentration of 1 uM and denaturation experiments for each duplex were performed in triplicate. The melting temperature was determined using a sigmoidal fit. Predicted denaturation temperatures for unmodified (i.e., no host or guest) duplexes were calculated using the OligoAnalyzer tool (Integrated DNA Technologies) software.

S6.3. Results and discussion of thermal denaturation studies of host-guest stabilized DNA duplexes
Thermal denaturation experiments were performed on the 25mer duplex S:I (where I is C6-Ad) to determine the stabilizing effects of the host-guest interactions. the 25-mer duplex. The melting temperature (Tm) was found to be 80.9°C. This is significantly enhanced (by 15.2°C) relative to the predicted transition for the corresponding unmodified 25-mer (see Supplementary   Table 7). We also performed renaturation studies (Supplementary Figure 38(B) to make sure that the duplex is fully dissociated (and an upper baseline is achieved) at 90 °C. The Tm of 80.9°C is remarkably stable and corresponds, for e.g., to a conventional unmodified duplex consisting of over 69 base pairs with GC content of 52.2%.

Supplementary
Thermal denaturation experiments were performed on host-guest stabilized shorter duplexes. In order to determine whether host-guest interactions can be used to stabilize otherwise unstable duplexes at ambient temperature, thermal denaturation experiments were also carried out on CB [7]-adamantane containing duplexes consisting of short sequences (Supplementary Table 7). Interestingly, the stability of each of these duplexes is enhanced significantly in the presence of the host-guest interaction, with shorter duplexes showing increasing ΔTm. Further, for very short duplexes, consisting of 5 or 4 base pairs (see Supplementary Figure 39 C & D) that are conventionally completely unstable at room temperature (are melted at 0 °C), we still observe an appreciable melting profile upon addition of the host-guest interaction.

Supplementary
The Tms for these duplexes coalesce at ~ 53°C. This finding suggests that these duplexes are predominantly stabilized by the host-guest interaction. Note: the data for the 4 and 5 base pair containing duplexes display scattering, especially at temperatures above ~ 70°C. This is likely due to noise arising from the fact that there is low absolute change in absorption going from the stacked to un-stacked state for such very short duplexes.