DNA nanostructure self-assembly in an aqueous ionic liquid solution with enhanced stability and target binding affinity

Abstract

DNA nanostructure-enabled functional constructs have shown potential to improve healthcare outcomes by offering advanced disease diagnostic and therapeutic strategies. Translating this potential of DNA nanostructure-based constructs to real life applications relies on maintaining and enhancing the structural integrity and functions of the surface-anchored moieties. In this study, we explored the possibility of utilizing choline dihydrogen phosphate (CDHP) solution, an aqueous solution of ionic liquid, to assemble DNA nanostructures of different sizes and complexities with enhanced biostability and ligand binding affinity. We show successful formation of the DNA nanostructures in aqueous CDHP solution using gel electrophoresis, atomic force microscopy (AFM), and circular dichroism (CD). Biostability assays reveal that the aqueous CDHP solution may provide passive protection to DNA nanostructures against DNase I and human serum for up to 48 hours. We also demonstrate that this enhanced biostability arises both from the structural conformation imparted during CDHP-mediated folding and from the presence of free CDHP ions in the solution. Notably, removal of free ions reduced the passive protection effect, but did not eliminate it, indicating the contribution of both folding and surrounding free ions. Using flow cytometry and surface plasmon resonance assays, we show that the presence of aqueous CDHP solution can enhance the binding of aptamer-functionalized DNA nanostructures to specific receptors on acute myeloid leukemia (AML) cells. Our strategy of using ionic liquid solution for one-pot preparation with enhanced stability and functionality offers a robust, simpler and faster alternative for DNA nanostructure-based constructs.

Graphical Abstract

INTRODUCTION:

The self-assembly of DNA using tailored sequences has enabled the capability to synthesize nanometer to micrometer scale structures ^1–5. These structures have gained notable attention due to their programmability, monodisperse size distributions, and tunable mechanical properties ^{6, 7}. Moreover, DNA nanostructures have been employed as templates to precisely pattern and direct the assembly of various external moieties including small molecules, metal nanoparticles, proteins, peptides, and aptamers ^8–13. Such DNA nanostructure-templated constructs with specific spatial arrangements of biomolecules have been used to address questions in chemistry, physics, and engineering fields, and have also shown potential to improve healthcare outcomes by providing advanced diagnostic methods, novel vaccine candidates, and effective drug delivery platforms ^{3, 7, 8, 14–26}. Successful formation of DNA nanostructures relies on the presence of metallic cations, with Mg²⁺ being the typically used and well-optimized counter ion for DNA self-assembly. These counter ions help screen the interhelical repulsion and stabilize the stacked form of branched DNA junctions to facilitate proper folding and assembly of DNA strands into desired structures by neutralizing negative charges on the phosphate backbone of DNA ²⁷. Despite their important role, some metallic divalent cations can have adverse effects on the biomedical applications of DNA nanostructures by enhancing nuclease degradation and by promoting structure aggregation when high ionic concentrations are required and used ^{28, 29}. Additional roadblocks to utilizing DNA nanostructure-based constructs in biomedical applications include their poor biostability in physiological fluids and in solutions with low cation concentrations, as well as their susceptibility to hydrolytic cleavage and nuclease degradation under suboptimal pH or ionic conditions, posing challenges for realizing their proper functions and full potential. Several strategies have been developed for improving and tuning the biostability of DNA nanostructures such as surface coating with PEGylated lipids ³⁰, bovine serum albumin ³¹, and PEGylated oligolysines ^{32, 33}, the formation of polyplex micelles using PEG–polylysine block copolymers ³⁴, UV crosslinking of unpaired thymine nucleotides introduced at DNA junctions ³⁵, and design of crossover spacing for intrinsic biostability ³⁶. Some of these approaches may alter the functionality of a DNA nanostructure-carried moiety or cause DNA nanostructures to aggregate, making it challenging to achieve both the protection and desired function of a DNA nanostructure-based construct.

To address some of these challenges in DNA nanotechnology, different solution-based assembly approaches have been explored. While Mg²⁺ remains the dominant choice, it is not universally essential. Notably, DNA nanostructures have been assembled in solutions that substitute Mg²⁺ with other ions or chemicals, demonstrating alternative methods for nanostructure assembly. Such approaches include assembly in low-Mg²⁺ buffers and buffer exchange into other solution conditions ³⁷, assembly in other counter ions ^{29, 38, 39}, in nonmetallic cation substitutes such as polyamines ^40–42, using telechelic polymers carrying positively charged head groups that enable metal-free duplex formation with enhanced thermostability ^{43, 44}, in deep eutectic solvents ⁴⁵, and organic solutions such as acetone ⁴⁶. Together, these examples illustrate the growing interest in expanding beyond conventional metallic and non-metallic environments. Recently, some ionic liquids have gained recognition as solvents to facilitate drug delivery processes ⁴⁷ or improve stability of proteins ⁴⁸. For instance, choline dihydrogen phosphate (CDHP), an ionic liquid (Figure S1), has been shown to actively promote structural folding of short nucleic acids (10–60 bases) with conserved molecular function, and improve resistance to hydrolysis and nuclease degradation ⁴⁹. CDHP has been shown to exhibit excellent biocompatibility as it consists of a choline cation and a phosphate anion, both of which are biologically abundant ^{50, 51}. To date, ionic liquids were only shown to stabilize short nucleic acid structures.

To improve the stability of DNA nanostructures in vitro and in vivo without chemical coating, UV crosslinking, or high concentrations of cations, we explored the use of aqueous CDHP solution for assembling DNA nanostructures of different sizes and complexities in the absence of metallic counter ions. We show successful assembly of tile-based DNA motifs (4-way junction, double crossover motif, 3-helix motif, square and triangle shapes) and origami-based nanostructures (triangle and rectangle shapes) in aqueous CDHP solution. We demonstrate successful formation of these DNA nanostructures in solutions with different CDHP concentrations using gel electrophoresis and further validated the DNA origami assemblies using atomic force microscopy (AFM) and circular dichroism (CD). We show that structures assembled in the aqueous ionic liquid solution exhibit enhanced biostability and moiety-binding affinity, with improved binding to target cell surface biomarkers, making these structures useful in biological applications.

RESULTS:

Self-assembly of tile-based DNA nanostructures in aqueous CDHP solution.

We first started with DNA motifs that are well-characterized by our own groups and others before, including a four-arm junction ⁵², a DNA double crossover (DX) motif ⁵³, and a 3-helix motif ⁵⁴ (Figure 1a and Figure S2). These motifs have been used in the assembly of 2D DNA arrays, periodic positioning of guest particles and drug delivery ^55–57, making them suitable model structures for this study. We assembled the four-arm junction, DX and the 3-helix structures in tris-acetate-EDTA (TAE) buffer containing different concentrations of CDHP using a thermal annealing protocol and validated assembly using non-denaturing polyacrylamide gel electrophoresis. Within the concentration ranges tested for CDHP solution (5–200 mM), we observed an increase in the assembly yields with increasing CDHP concentrations for the 4-way junction and the 3-helix motif, while assembly yields for the DX motif remained similar across this concentration range (Figure 1b–d and Figure S3). Further, the assembly yields for the DX motif were similar to the structure annealed in Mg²⁺-containing buffer in most of the concentration range we tested. The four-arm junction and the 3-helix motif required higher CDHP concentration to result in a yield similar to the Mg²⁺ control. However, CDHP-assembled structures also showed impurities in the form of partially folded structures and higher order aggregates in some CDHP concentrations, resulting in a purity of ~80% compared to the structures assembled in Mg²⁺-containing buffer (Figure S3).

Figure 1: Self-assembly of DNA motifs and tile-based DNA nanostructures in aqueous CDHP solution.

(a) Model DNA motifs used in this study: 4-arm junction, double crossover motif (DX), and 3-helix motif. (b, c &d) Non denaturing PAGE showing assembly of these motifs across different concentrations of CDHP solution. (e) Model DNA tile-based nanostructures used in this study: DNA triangle and DNA square. (f & g) Non denaturing PAGE showing assembly of the tile-based nanostructures across different concentrations of CDHP solution.

Next, we used triangle- and square-shaped DNA tile nanostructures, which were chosen for their simplicity in design and potential biological applications, as suggested by the RNA version of the structures ⁵⁸. Both structures are composed of three strands (Figure 1e, sequence and design in Figure S4). We first assembled the structures in 1x TAE buffer containing 12.5 mM Mg²⁺ and validated proper assembly using non-denaturing PAGE ⁵⁹. We then substituted different concentrations of CDHP solution (5, 10, 25, 50, 100, and 125 mM) for Mg²⁺ in the TAE buffer to test the impact on proper formation of these structures. We observed that both these nanostructures assembled consistently across all tested CDHP concentrations. The similarity of DNA species mobility across different CDHP concentrations indicates that choline ions from CDHP effectively supported the folding and structural integrity of DNA tiles within the tested range (Figure 1f–g). CDHP likely aids the formation of these nanostructures by forming a network of electrostatic and non-covalent interactions ⁶⁰. Additionally, monovalent ions such as choline and Na⁺ provide an electrostatic screening effect which helps avoid kinetic traps encountered during the high-concentration assembly of tile-based DNA structures ⁶¹.

Self-assembly of origami-based nanostructures in aqueous CDHP solution.

Encouraged by the successful formation of DNA motifs and tile-based nanostructures in aqueous CDHP solution, we further explored the possibility of assembling larger DNA nanostructures as a way to evaluate adaptability of CDHP to DNA nanostructures of varying sizes. We utilized the DNA origami strategy to assemble two larger nanostructures, a triangle and a rectangular DNA origami, to investigate the self-assembly of larger and more complex DNA nanostructures using aqueous CDHP solution (Figure 2a). We first tested the folding of the M13mp18 scaffold DNA (7,249 nt) using staple strands corresponding to a triangle or a rectangle shape in 1x TAE buffer containing different concentrations (5, 10, 25, 50, 100, 125 mM) of CDHP solution using thermal annealing. We characterized the assembled origami nanostructures using non-denaturing agarose gel electrophoresis, atomic force microscopy (AFM) and circular dichroism (CD).

Figure 2: Self-assembly of DNA origami nanostructures in aqueous CDHP solution.

(a) Schematic of self-assembly of triangle and rectangle origami using thermal annealing. (b & e) Agarose gel electrophoresis showing the assembly of the origami nanostructures at different CDHP concentrations, with highest concentration migrating similar to its Mg²⁺ counterpart. (c & f) AFM image confirming the assembly of the origami nanostructures self-assembled at 125 mM CDHP solution. (d & g) Circular dichroism spectra confirming the typical helicity and base stacking of the origami nanostructures when assembled in CDHP solution. Scale bars in inserts and full images are 100 nm.

For triangle-shaped origami, the lower CDHP concentrations (5 to 50 mM) resulted in slower migration bands, indicating loosely folded DNA origami structures. At higher CDHP concentrations (100 mM and 125 mM), the DNA origami triangle migrated similar to that of the structure self-assembled in Mg²⁺-containing buffer, confirming proper structural folding (Figure 2b). AFM imaging revealed the formation of loosely folded origami structures at lower CDHP concentrations (Figure S5), whereas no substantial differences were seen in the overall size or morphology between the DNA origami assembled at higher concentrations of CDHP solution and the Mg²⁺-assembled structures (Figure 2c and Figure S6). However, the CD spectra revealed a slight shift in negative ellipticity band (~1.89 mdeg) at 247 nm, attributed to slight alterations in DNA helicity that may collectively cause slower mobility in agarose gels. The positive ellipticity band at 280 nm, which is attributed to DNA base stacking, remained unchanged (Figure 2d).

For rectangular DNA origami, the agarose gel data showed that a lower CDHP concentration (5 to 25 mM) resulted in two origami species, suggesting an incomplete formation of the structure. On the contrary, higher CDHP concentrations (100 mM and 125 mM) promoted the formation of a single DNA origami species, consistent with the Mg²⁺-assembled nanostructures (Figure 2e). Similar to the triangle origami, AFM imaging revealed no observable differences in overall size or morphology between the CDHP-assembled and Mg²⁺-assembled rectangular DNA origami structures (Figure 2f and Figure S6). CD spectra showed consistent ellipticity bands at 280 nm and 247 nm, indicating proper structural folding at higher ionic liquid concentrations (Figure 2g).

Biostability of DNA nanostructures in aqueous CDHP solution

Biostability remains a significant challenge to apply DNA nanostructure-based platforms in disease diagnostics and therapeutics due to the susceptibility of DNA nanostructures to digestion by nucleases and degradation in low cation concentration-containing matrices ⁶². Therefore, we were motivated to understand the effect of CDHP on the biostability of DNA nanostructures. For this purpose, we chose three representative nanostructures of different sizes and complexities: a DNA double crossover motif, a DNA square tile and a DNA triangle origami and evaluated nuclease resistance using a gel electrophoresis-based protocol we established before ^{29, 63}. More specifically, we utilized DNase I, the most abundant nuclease in blood and plasma, which nonspecifically cleaves both strands of double-stranded DNA and exhibits optimal activity at the physiological temperature of 37 °C ^{36, 62} (Figure 3a). We investigated the resistance to DNase I degradation by incubating the DNA nanostructures with specific concentrations of DNase I above physiologically relevant level (0.35 units/mL) at 37 °C for various time intervals ⁶⁴. The DNase I-treated samples were analyzed on a non-denaturing gel and the band intensities corresponding to the intact nanostructures were quantified at each time point to generate nuclease degradation profiles. For the double crossover motif assembled in 250 mM CDHP and maintained in CDHP-rich environment, we observed no degradation with 0.1 U DNase I even after an hour (top panel of Figure 3b), whereas the same structure assembled and maintained in Mg²⁺-containing buffer degraded rapidly within a few minutes (bottom panel of Figure 3b), with ~90% degraded in an hour (Figure 3e and Figure S7). The DNA square tile and DNA origami nanostructures assembled in 125 mM CDHP and maintained in CDHP-rich environment also showed minimal to no degradation after 48-hour incubation with 1.0 unit/mL of DNase I (top panel of Figure 3c–d). However, the nanostructures assembled and maintained in conventional Mg²⁺-containing buffer showed ~95% degradation within a few hours under the same conditions (Figure 3c–d, bottom panels, Figure 3f–g, and Figure S8). To compare the biostability in CDHP to solution conditions other than magnesium-containing buffers, we evaluated the nuclease resistance of DNA triangle origami structures assembled in Mg²⁺ and buffer exchanged to 125 mM NaCl, 125 mM HEPES, and 125 mM Tris buffers. Under identical DNase I treatment conditions, DNA triangle origami was almost completely degraded in both HEPES and Tris buffers. However, in the presence of 125 mM NaCl, we observed a partial inhibition of DNase I activity, resulting in slower degradation of the DNA nanostructures (Figure S9). This observation is consistent with previous reports showing that ions such as Na⁺ and K⁺ can modulate DNase I activity by interfering with its catalytic efficiency ⁶⁵ and our prior work showing enhanced biostability of DNA nanostructures in monovalent ions such as Na⁺, K⁺ and Li⁺ compared to divalent ions such as Mg²⁺, Ca²⁺ and Ba²⁺. However, aqueous CDHP solution provides much higher biostability compared to all the other solution conditions tested herein.

Figure 3: Biostability of the nanostructures in aqueous CDHP solution.

Schematic of model DNA nanostructure treatment with (a) DNase I and (h) DNase I-spiked 10% human serum. Non-denaturing gels of nanostructures in either CDHP or Mg²⁺ treated with (b-d) DNase I and (i-j) DNase I-spiked 10% human serum. Degradation profiles showing that nanostructures assembled and maintained in CDHP solution exhibit resistance to (e-g) DNase I and (k-l) DNase I-spiked 10 % human serum digestions. In panels (e-g, k-l), data are presented as the mean ± standard error (SE), n = 3 independent samples.

To assess the biostability of DNA nanostructures in physiologically relevant conditions, we tested the degradation of DNA nanostructures in human serum (Figure 3h). Commercially available human serum is typically heat-inactivated during processing, which leads to the denaturation and loss of activity of endogenous DNase I ^{66, 67}. As a result, DNA nanostructures remained stable in human serum, showing no significant degradation in CDHP and Mg²⁺ containing buffers (Figure S10). Therefore, to create a more representative model of nuclease-rich physiological conditions, we spiked the serum with exogenous DNase I. Again, the DNA nanostructures assembled and maintained in CDHP solution showed minimal to no degradation for up to 48 hours when incubated with human serum containing DNase I (Figure 3i–j, top panels), whereas the same nanostructures assembled and maintained in a Mg²⁺ containing buffer showed ~95% degradation within a few hours (Figure 3i–j,bottom panels, Figure 3k–l and Figure S11). We attribute the observed DNA band retardation in the agarose gel to some non-specific interaction between DNA nanostructures and serum-containing proteins. Overall, these results show the ability to enhance the biostability of DNA nanostructures through assembly and maintenance conditions such as the use of CDHP.

To further evaluate the contribution of CDHP-based assembly to nuclease resistance, we examined the degradation profile of DNA square tile and triangle origami structures in the absence of free CDHP ions. We assembled the two nanostructures in TAE buffer containing either 12.5 mM Mg²⁺ or 125 mM CDHP, followed by buffer exchange into 1× DNase buffer (10 mM Tris-HCl, 2.5 mM MgCl₂, 0.5 mM CaCl₂, pH 7.6). CDHP-assembled DNA square tile and triangle origami structures exhibit delayed degradation profiles compared to their Mg²⁺-assembled counterparts following DNase I incubation (Figure 4a–b). The CDHP-assembled DNA square tile shows ~50% degradation at 60 min, whereas the same structure assembled in Mg²⁺ buffer is ~70% degraded by only 15 min. A similar trend was observed for the origami triangle, with CDHP-assembled structures showing ~1.5–2-fold improved nuclease resistance relative to Mg²⁺ assembled nanostructures. This improved biostability possibly arises from structural effects imparted by CDHP ions present on nanostructures during the assembly, which may result in delayed onset of nuclease cleavage. While the presence of free CDHP ions in the solution further improves biostability (Figure 3), these results indicate that CDHP assembly alone confers measurable resistance to nuclease degradation relative to assembly in conventional Mg²⁺ buffers.

Figure 4: Biostability of CDHP assembled nanostructures without surrounding free ions.

DNA nanostructures (a) DNA square tile and (b) DNA triangle origami buffer exchanged into DNase I reaction buffer and treated with DNase I. Non denaturing gels and degradation profiles of nanostructures assembled in CDHP show improved nuclease resistance to DNase I compared to nanostructures assembled in Mg²⁺. Data are presented as the mean ± standard error (SE), n = 3 independent samples.

To understand how CDHP could affect the native confirmation of DNase I, we performed native polyacrylamide gel electrophoresis (PAGE) of DNase I incubated with different concentrations of CDHP solution under the same conditions as our biostability assays (Figure S12). The results showed that the native conformation of DNase I is preserved in the presence of CDHP. However, at higher concentrations of CDHP solution, a mobility shift was observed. To further validate enzyme integrity, DNase I was incubated in CDHP (125 mM) for 3 hours and buffer-exchanged into Mg²⁺ solution, where it migrated identically to the Mg²⁺-only control and retained full nuclease activity (Figure S13). Together, these findings confirm that CDHP does not structurally denature DNase I but renders it functionally inactive, with activity restored upon buffer exchange. Given these observations, we next investigated the influence of CDHP on DNase I activity through molecular docking and molecular dynamics simulations. We used an experimentally determined structure of DNase I in complex with a double-stranded DNA (dsDNA) and docked the 3D structure of choline into the active site of DNase I. Prior to docking, the free energy of the DNase I-dsDNA complex was calculated to be −21.38×10³ kJ/mol. Following the docking of choline to this complex, the free energy changed to −21.04×10³ kJ/mol. This increase in free energy suggests that the introduction of choline destabilized the DNase I and dsDNA complex. To better understand and visualize this destabilization, we simulated the DNase I-dsDNA-choline complex for 50 ns and observed that the free energy of the complex further increased to −19.68×10³ kJ/mol. To understand the origin of this destabilization, we superimposed 3 frames of the structure derived from the simulation (frame 1, frame 600 and frame 1000) (Figure S14a). The superimposition revealed that strand-2 of the dsDNA (the strand interacting with DNase I active site) moves away and out of the enzyme active site during the simulation (Movie S1). Simultaneously, choline, which initially shows hydrogen-bonds with strand 2, gradually moves deeper into the DNase I active site forming hydrogen bonds with residues from DNase I (Figure S14b&c). Analysis of the RMSD shows a similar trend for the complex as well as the individual components (DNase I, strand 1 and strand 2 of the dsDNA) (Figure S14d). The molecules show higher deviations initially, gradually stabilizing over the duration of the simulation. Notably, strand 2 of the dsDNA which move out the DNase I active site exhibits maximum deviations. Taken together, these observations hint at the ability of CDHP to interfere in the enzymatic activity of DNase I in cleaving DNA, thus conferring passive protection to DNA.

Target binding function of aptamers patterned on DNA origami nanostructures assembled in aqueous CDHP solution

After observing the biostability enhancing effects of CDHP, we evaluated whether choline ions possibly interfere with the molecular recognition capabilities of DNA nanostructures by validating the binding capabilities of aptamers patterned on a DNA nanostructure to target cells. We first confirmed that aqueous CDHP solution did not interfere with cell viability after 24 hours of treatment (Figure S15). Next, we utilized a previously reported DNA origami-aptamer construct to target CD-117 protein (c-kit), a surface marker overexpressed on HEL 92.1.7, an acute myeloid leukemia (AML) cell line ¹³ (Figure 5a). We modified a CD117-specific aptamer at one end with a FAM label (for binding analysis) and at the other end with an overhang complementary to single-stranded extensions on a rectangular DNA origami. Single stranded extensions on the DNA origami provided 88 anchor sites for precise aptamer positioning (Figure 5a and Figure S16). To analyze aptamer binding to target cells, we incubated the free FAM-aptamer or DNA origami-bound FAM-aptamers (Figure 5b) with HEL 92.1.7 cells (that naturally express CD117) for one hour, followed by flow cytometry assays. We performed these experiments with aptamers and DNA origami-aptamer constructs prepared in Mg²⁺ or CDHP-containing buffers. We observed increased fluorescence for aptamers and aptamer-coated origami samples in CDHP compared to those in Mg²⁺ containing buffer, confirmed by the shift in the histogram peaks and median fluorescence intensity values (Figure 5c–d and Figure S17). While a small population of cells remained unlabeled by the aptamers assembled in the Mg²⁺ containing buffer (marked by black boxes in Figure 5c and Figure S17), the aptamers in CDHP demonstrate better ability to label target cells, indicating effective performance in recognizing and targeting receptors in a cellular environment. Further, CDHP-mediated binding enhancement was more prominent for structures assembled in CDHP (125 mM), with a ~2.76-fold increase in binding for aptamers at 880 nM and a ~3.27-fold increase for origami-aptamer constructs at 10 nM (Figure 5c–d). This increase in aptamer binding efficiency may be attributed to the dehydration effect surrounding the proteins and aptamers induced by ionic liquids such as CDHP as reported previously ^{60, 68}. Additionally, biostability assays conducted in fluorescence-activated cell sorting (FACS) buffer containing either CDHP or Mg²⁺ confirm that the DNA nanostructures remain intact during the incubation, validating that binding differences reflect aptamer performance rather than structural degradation (Figure S18). Subsequently, we incubated CD-117 aptamer and aptamer-origami constructs in HL-60 cell line (CD117⁻, CD123⁺) and observed significantly lower fluorescence intensity compared to positive target origami-aptamer constructs (same rectangle origami decorated with CD123 targeting aptamers), confirming the specificity of aptamer-mediated binding of DNA nanostructures to target cell types (Figure S19).

Figure 5: Conservation of aptamer function in aqueous CDHP solution.

Schematic of (a) rectangle origami-aptamer construct and (b) incubating aptamers or origami-aptamer construct with acute myeloid leukemia cells (HEL 92.1.7). Illustration from NIAID NIH BIOART Source bioart.niaid.nih.gov/bioart/173. (c) Fluorescence histograms of HEL 92.1.7 cells incubated with aptamer or origami-aptamer constructs reveal shift in intensities for samples assembled in CDHP solution. Black boxes indicate a small population of cells that remained unlabeled in samples assembled in the Mg²⁺ solution. (d) Median fluorescence intensity (MFI) analysis of DNA samples in CDHP shows a 2.76-fold increase in the binding for aptamers and 3.27-fold increase for origami-aptamer construct compared to those in Mg²⁺ containing buffer. Data in panel (d) are presented as the mean ± standard error (SE), n = 4 biologically independent samples.

The enhanced binding of CDHP-assembled aptamers prompted us to further investigate the binding strength of aptamer under these conditions. We performed surface plasmon resonance (SPR) analysis to quantitatively assess the aptamer-cell binding in either CDHP or Mg²⁺ containing buffer (Figure 6a–b). We observed that the binding strength of the free aptamer against c-kit (CD117) improved in CDHP (K_D = 218.78 ± 17.57 nM) than that in Mg²⁺ containing buffer (K_D = 416.27 ± 29.84 nM), indicating a ~2-fold increase in affinity (Figure 6c–d and Figure S20). Aptamers with higher K_D values (weaker binding) exhibited faster dissociation kinetics (Figure 6c) than those with lower K_D values (stronger binding), resulting in slower signal decay during the dissociation phase (Figure 6d). This increase in binding affinity was comparable to the ~2-fold increase observed in cell binding by flow cytometry assay. Interestingly, we observed unconventional negative signals in the SPR data for CDHP-assembled aptamers. This phenomenon can be attributed to a reduction in the hydration layer around the aptamer when folded in CDHP, leading to a deviation in the mass refractive index and the observed SPR response. Our data aligns with the previous findings that showed that negative signals in SPR sensorgrams could arise from a negative complex refractive index increment deviation which counterbalances the expected increase in signal associated with target recognition and aptamer folding ⁶⁹. The improved aptamer-cell binding affinity suggests that the folding and stabilization of the aptamer in CDHP provides structural integrity and target recognition.

Figure 6: SPR analysis of aptamer binding to CD117.

Illustrations depict the proposed binding of DNA aptamers (orange) folded in (a) Mg²⁺-containing buffer and (b) CDHP-containing buffer to immobilized CD117 (blue) on the sensor chip (grey). Representative SPR sensorgrams comparing aptamer binding behavior in (c) Mg²⁺ and (d) CDHP buffers. Aptamers were injected at various concentrations over immobilized CD117 (Flow Cell 2) at a flow rate of 5 μL/min. Each injection consisted of a 120-second association phase followed by a 240-second dissociation phase. The sensor surface was regenerated between runs using 10 mM glycine (pH 2.0) for 30 seconds at 30 μL/min). Illustration from NIAID NIH BIOART Source bioart.niaid.nih.gov/bioart/436.

DISCUSSION

DNA nanostructure folding typically relies on the binding of metal cations to the major and minor grooves of DNA for effective charge screening and stabilization ⁷⁰, with their interactions being primarily ionic and susceptible to thermal fluctuations. In this work, we demonstrated that aqueous CDHP solution enables the successful assembly of several DNA motifs and DNA origami. Their bulky nature, combined with their hydroxyl group and positively charged amino group, enable the formation of multiple hydrogen bonds and a network of stabilizing interactions with DNA ⁷¹. In addition to supporting successful assembly in ionic liquid solution, we show two key enhancements resulting from this study: (1) enhanced biostability for structures assembled and maintained in aqueous CDHP solution and (2) improved aptamer binding efficiency.

Our work shows that DNA nanostructures assembled and maintained in aqueous CDHP solution show minimal to no degradation against a nuclease and in serum containing nucleases compared to the almost-immediate degradation of structures assembled and maintained in Mg²⁺-containing buffers. Previous reports suggest that the nuclease resistance provided by CDHP comes from choline’s ability to destabilize nucleases at concentrations as high as 4 M ⁴⁹. However, it is interesting to note the nuclease inhibitory effect of CDHP at a lower concentration (125 mM) against the DNase I concentrations tested in this study. The resistance to higher DNase I concentrations can be enhanced by increasing CDHP levels after initial assembly at lower concentrations. This could be due to the ability of choline ion to bind within the minor grooves of dsDNA, shielding the nitrogenous bases from rapid degradation of DNA ⁷². Our data shows the ability of aqueous CDHP solution to inhibit nuclease-based DNA digestion at concentrations above the physiological relevant levels of DNase I ⁶⁴ for up to 48 hours, while simultaneously allowing the assembly and maintaining the functionality of biological moieties (proteins and nucleic acids) attached to these structures. In contrast, the DNA nanostructures assembled in Mg²⁺, a co-factor known to catalyze nuclease activity ⁷³, remain susceptible to rapid nuclease degradation under the same conditions.

Additionally, while the complete removal of free CDHP ions leads to a reduction in nuclease resistance of the DNA nanostructures; the structures still exhibit measurable resistance compared to those assembled in conventional Mg²⁺ buffers. Compared to previous stabilization strategies for DNA nanostructures, our approach offers an alternative to provide nuclease resistance without requiring additional coating. This CDHP-based strategy is attractive due to its simplicity and compatibility with downstream applications. In in vitro assays, the use of CDHP-containing buffers can directly improve nanostructure stability in nuclease rich environments, including biological fluids. Moreover, since the protective effect from CDHP assembly persists even after the removal of free ions, this method may be extended to in vivo applications such as cellular delivery and immunotherapy.

In addition to structural biostability, our study also demonstrates that aptamers and aptamer-functionalized DNA nanostructures assembled in CDHP exhibit improved binding to cells of up to 3.27-fold. The presence of choline ions in the CDHP could be beneficial in target recognition as they can overcome the charge-charge repulsions during the formation of aptamer’s secondary and tertiary structure ^{49, 74}. Moreover, the molecular environment created by CDHP appears to support protein stability on the surface of the cell, potentially preserving the native conformation of aptamer-binding sites and enhancing the overall recognition process in a cellular environment ⁷⁵. This dual effect, stabilizing both the aptamer and its target, creates a more favorable interaction landscape for aptamer-based targeting strategies.

It is worth noting that the use of any DNA nanostructure self-assembly matrix including ionic liquids in biological applications needs to include the potential cytotoxicity and compatibility with living systems. Some ionic liquids (such as imidazolium- or pyridinium-based) contain bulky aromatic groups that disrupt lipid bilayers or denature proteins through strong hydrophobic or chaotropic interactions, thus causing cytotoxicity ^{76, 77}. In contrast, choline-based ionic liquids are more hydrophilic and biologically compatible ⁷⁸. Specifically, choline dihydrogen phosphate (CDHP) is composed of choline, a quaternary ammonium cation that is an essential nutrient for health ⁵⁰, and dihydrogen phosphate, a biologically prevalent anion ⁷⁹. This combination results in a hydrophilic ionic liquid with low membrane permeability and minimal hydrophobic interactions, which reduces its potential for cellular toxicity arising from any non-specific outcomes ⁸⁰. Furthermore, CDHP has been previously reported to stabilize protein secondary structures and preserve enzyme activity ^81–84. Our findings further support this by demonstrating that aqueous CDHP solution not only preserves DNA nanostructure integrity but also improves the functional activity of biological moieties such as aptamers and proteins. This suggests that DNA nanostructures assembled and maintained in CDHP could be integrated into therapeutic and diagnostic platforms without biocompatibility concerns. Overall, our strategy offers a robust, simple, and economical process to prepare DNA nanostructure-based functional constructs for biomedical applications.

Supplementary Material

The Supporting Information is available free of charge at https://pubs.acs.org. Materials and methods, additional experimental results and DNA sequences used.