(167am) Reversible Hybridization of Sequence-Defined Oligocarbamates

The programmable self-assembly of materials with predictable properties is a long-standing challenge in material science^1–4. Overcoming this challenge is necessary for making materials whose properties can be tuned for a desired application. Producing such self-assembled materials requires specific and directional noncovalent interactions to consistently guide building blocks into one out of many possible structures.

DNA is one of the most effective ligands for directed self-assembly because its directional hydrogen-bonding base pairs allow for reversible, sequence-specific hybridization of complementary strands. Scientists can fully leverage these properties because DNA can be synthesized with precise control over length, composition, and monomer sequence. These handles can all be tuned to attain excellent specificity and high affinity between complementary strands. This has led to DNA's use as a structure-directing ligand for programmable nanoparticle assembly. DNA-functionalized gold nanoparticles (DNA-AuNPs), in particular, have frequently been used for biomedical applications because they combine the directional, specific, and reversible binding of oligonucleotides with the unique optoelectronic properties of gold nanoparticles. For example, color changes resulting from the aggregation of gold nanoparticles have enabled DNA-AuNPs to serve as efficient biosensors for proteins and cancer cells⁵. Unfortunately, DNA cannot be synthesized at a scale sufficient for non-biological materials science applications. Furthermore, DNA is only soluble in aqueous solutions and cannot direct the assembly of non-water-soluble nanoparticles, many of which have interesting catalytic, photonic, and magnetic properties.

To address this unmet need, we present the scalable synthesis of sequence-defined oligocarbamates (SeDOCs) equipped with thymine and diaminotriazine moieties that can undergo reversible and directional hybridization in the organic phase. SeDOCs are synthesized via iterative reductive amination and carbamation reactions and contain a moderately rigid backbone thanks to aromatic units derived from the bio-renewable monomer, vanillin. The rigid backbone combines with the directional and specific triple hydrogen bonds between thymine and diaminotriazine to limit SeDOCs’ conformational flexibility.

In this work, the effect of monovalent monomer sequence on the binding strength (K_a) of complementary SeDOC 3-mers was evaluated. SeDOCs are named according to their pendant group sequence, where m, D, and T stand for methyl, diaminotriazine, and thymine respectively. Oligomers synthesized for this study include mDm, mmD, mTm, and mmT (Fig. 1A). A methyl group is used as a blank unit in locations where D and T moieties are omitted. Oligomers were characterized by liquid chromatography-mass spectrometry (LC-MS) and ¹H-NMR.

Hybridization between SeDOCs with complementary pendant groups was assessed prior to measuring the strength of their interactions. Hybridization was detected via diffusion ordered spectroscopy (DOSY), a ¹H-NMR technique that measures the diffusion coefficient for each species in solution. A decrease in a SeDOC's diffusion coefficient when mixed with a complementary ligand is indicative of association because the larger hybrid should diffuse more slowly than the smaller, pure oligomer. The diffusion coefficients of hybridized SeDOCs were measured by combining an oligomer with a large excess of its complement in deuterochloroform. A D ligand (i.e., mDm or mmD) was added to a T ligand (i.e., mTm or mmT) at a molar ratio of approximately 3.5 D:1 T. The diffusion coefficient of the hybrid was measured using the resonance of the methyl protons on thymine. The diffusion coefficient of pure mTm was 3.96x10^-6 cm²/s. When mTm was combined with mDm or mmD, the diffusion coefficient decreased to 3.75 x10^-6 cm²/s and 3.78x10^-6 cm²/s, respectively (Fig. 1B). The same resonance on pure mmT had a diffusion coefficient of 4.01x10^-6 0.011x10^-6 cm²/s. This decreased to 3.67x10^-6 0.017x10^-6 cm²/s when mmT was combined with mDm, and to 3.69x10^-6 cm²/s when mixed with mmD. Experiments are ongoing to determine whether these changes are statistically significant. The current data, which shows a decrease in mTm and mmT's diffusion coefficients in the presence of excess mDm or mmD, indicates that SeDOCs with complementary pendant groups form hybrids.

A ¹H-NMR titration was performed with mDm and mmT to determine whether the association was driven by hydrogen bonding between D and T. In this experiment, mDm was incrementally added to mmT, and a change in the chemical shifts of the hydrogen bonding protons was observed. A downfield shift upon titration is a classic indicator of hydrogen bonding. H-bonding reduces the electron density around the donor proton, resulting in an increased chemical shift. The H-bonding proton on the thymine's Watson-Crick face (Fig. 1D) was explicitly monitored in this experiment. The chemical shift of the proton of interest increased from 8.40 ppm in pure mmT to 11.99 ppm in a 5:1 mDm:mmT mixture (Fig. 1C). This large downfield shift confirms that complementary SeDOCs form hydrogen bonds.

¹H-NMR experiments confirmed that complementary SeDOCs form hybrids and that association is driven, at least in part, by hydrogen bonding between D and T. Isothermal titration calorimetry (ITC) was then used to measure the strength of these interactions. ITC directly measures binding stoichiometry (n), K_a, and ΔH, which enables the calculation of ΔG and ΔS. ITC experiments were performed for every pair of complementary ligands (i.e., mDm/mTm, mDm/mmT, mmD/mTm, mmD/mmT). In each titration, a 100 mM solution of D ligand in chloroform was incrementally injected into a 10 mM solution of T ligand in chloroform at 25 °C. Injecting D into T led to binding events that generated heat. This data was fit to an independent-sites binding model to obtain n, K_a, and ΔH. The values of K_a and ΔH were applied to thermodynamic equations to calculate ΔG and ΔS. The data showed that monovalent SeDOC 3-mers form 1:1 dimers and that K_a, ΔH, and ΔS do not depend on the sequence of the monovalent ligands (Fig. 1E, F, and G). As a result, monomer sequence cannot be used as a handle for tuning the binding strength of monovalent SeDOCs. These experiments also revealed that SeDOC hybridization is both enthalpically- and entropically-driven (Fig. 1F and G). The increase in entropy upon binding suggests that a solvophobic effect could contribute to the association.

Additional studies are underway to evaluate the effect of valency (i.e., the number of D-T interactions between a given pair of ligands) on K_a. It is expected that increasing the valency from mono, to di, to trivalent should give an increase in K_a of one or two orders of magnitude. However, increased steric repulsion caused by having multiple bulky pendant groups could limit the expected increase in binding strength. Divalent (DmD, mDD, TmT, and mTT) and trivalent (DDD and TTT) oligomers are being synthesized to determine the relative contributions of these effects. ¹H-NMR will be used to detect hybridization of the multivalent SeDOCs, and binding thermodynamics will be quantified by ITC.

References

1. Jones, R. Why nanotechnology needs better polymer chemistry. Nat. Nanotechnol. 3, 699–700 (2008).
2. Szczypiński, F. T., Gabrielli, L. & Hunter, C. A. Emergent supramolecular assembly properties of a recognition-encoded oligoester. Chem. Sci. 10, 5397–5404 (2019).
3. Hebel, M. et al. Sequence Programming with Dynamic Boronic Acid/Catechol Binary Codes. J. Am. Chem. Soc. 141, 14026–14031 (2019).
4. Szczypinski, F. T. & Hunter, C. A. Building blocks for recognition-encoded oligoesters that form H-bonded duplexes. Chem. Sci. 10, 2444 (2019).
5. Medley, C. D. et al. Gold Nanoparticle-Based Colorimetric Assay for the Direct Detection of Cancerous Cells. Anal. Chem. 80, 1067–1072 (2008).