(Self) assembled news: recent highlights from the supramolecular chemistry literature (quarter 2, 2024)

KEYWORDS:

• Highlight
• news
• review
• supramolecular chemistry

Foldamers take single-stranded DNA across lipid membranes

Over the past two decades, the Huc group has pioneered the synthesis of foldamers based on quinoline derivatives, and have shown that they can mimic [1], or interact with [2] DNA. Now Dong and co-workers have used Huc’s foldamers to transport single-stranded DNA across membranes [3]. They used two different foldamers, each of which is ~3 nm long: one has cationic ammonium groups arranged linearly, while the other has the charges in a helical arrangement. Fluorescence titration experiments revealed that both foldamers could bind to DNA with association constants > 10⁶ M⁻¹, with the strong binding attributed to complementary electrostatic interactions between the cationic foldamers and anion DNA. Interestingly, the foldamer with a helical arrangement of positive charges preferentially binds single-stranded DNA, while that with the linearly arranged cationic groups preferentially binds double-stranded DNA. Despite these differences in binding strength, both foldamers were able to transport single-stranded DNA across membranes, with the foldamer containing linearly-aligned charges showing slightly higher transport activity. Studies in human cell lines demonstrated that this transport was viable in vivo, with an mRNA strand being transported effectively into cells.

Automating imine cage discovery

Cages formed through imine self – assembly have been well-studied for numerous applications, including as porous materials. Given the wide accessibility of amine and aldehyde building blocks, and the synthetic ease of the imine condensation reaction, a near-endless variety of cage architectures could be envisaged. However, testing whether these structures form, and for those that do whether the resulting cages are shape-persistent and porous, is a laborious process. Previous work led by Jelfs and Cooper used a robot to study cage synthesis from a total of 78 precursor combinations (three tri-amines and 26 di-aldehydes) [4], but robots are currently expensive and thus of limited accessibility. A team led by Jelfs and Greenaway now reports a study of 366 precursor combinations using a relatively inexpensive liquid handling platform (cost < US$10,000, see Figure 1 for example precursor combinations and analysis) [5].

Figure 1. Example output of automated cage screening for three precursor pairings used by Jelfs, Greenaway and co-workers [5]. The outer circle represents turbidity studies (where a blue “pass” indicates all material was soluble and a yellow “fail” indicates formation of an insoluble precipitate), the middle circle represents whether complete disappearance of the aldehyde resonance was observed by¹H NMR spectroscopy and the inner circle represents whether a single cage product was formed (a “fail” in this case would indicate the presence of more than one different cage).

The various combinations were screened in 1 mL volume reactions in 48-well plates and then automated processes were used to obtain and analyse turbidity, ¹H NMR and mass spectrometry data to indicate if cage formation was successful. Of the 366 precursor combinations, 54 gave clean formation of a single soluble cage topology, while formation of a single cage topology but with incomplete conversion or an insoluble precipitate was the most common outcome (83 precursor combinations). Automated computational modelling was used to predict which of the four most common cage types was likely for any given combination ([2 + 3], [4 + 6], [6 + 9] or [8 + 12] arrangements of tritopic and ditopic components), and to determine whether a cage was likely to have intrinsic porosity. The results are an impressive demonstration of how relatively inexpensive equipment can be used to survey a wide area of chemical space, and fit into a growing trend (highlighted in a recent editorial in this journal [6] for increasing use of automation in supramolecular chemistry.

Weak interactions give strong anion binding

The use of highly polarised C – H hydrogen bonds as anion recognition motifs has become firmly established in the literature [7], notwithstanding that as recently as the late 90s there was still debate as to whether these donors could form ‘real’ hydrogen bonds [8,9]. Highly polarised C – H donors such as glycoluril, triazole, triazolium and imidazolium groups are typically used. However, a recent paper from the groups of Chmielewski and Szumna demonstrates that simply adding one nitro group to each phenyl ring of otherwise electron-rich resorcin[4]arenes gives macrocycles that demonstrate remarkable halide anion recognition properties, even in aqueous media [10]. Indeed, binding of halides in 9:1 THF:water was achieved, with iodide binding to one receptor with an association constant of 11,500 M⁻¹ in this solvent.The high binding strength and an unusual solvent dependence was attributed to the anion being attracted to the macrocycle by the high dipole moment created by the nitro substituents, but binding occurring at a more hydrophobic region of the host containing multiple preorganised C–H donors. The authors demonstrated the efficacy of their receptors by transporting Cl^– anions across lipid membranes, with the compound active in transport assays at very low loadings.

In brief

Mechanochemical route to mixed-ligand cages

Bloch and colleagues have found that post-synthetic ligand exchange within cuboctahedral coordination cages can be achieved via solvent-free grinding [11]. Ligand exchange can occur between either a homoleptic cage and a second ligand or two homoleptic cages. The conversion to statistical heteroleptic cage mixtures is complete in a matter of minutes, compared to corresponding solution-phase conversions which take more than a week. This solvent-free approach is not limited by the solubility of the starting cages or ligands, which could broaden the scope of mixed-ligand porous materials synthesis.

ChemFETs to study binding between aqueous anions and hydrophobic receptors

As part of our upcoming ‘Method Articles’ collection, Johnson and co-workers have described how they make use of chemically sensitive field effect transistor (ChemFET) technology to study anion binding [12]. Drop-casting a hydrophobic anion receptor blended with a polymer onto a ChemFET electrode surface produces an anion-responsive sensor which can be used to quantitatively detect analyte anion concentrations in aqueous solutions. Parameters such as the detection limit, selectivity and sensitivity can then be readily derived, with case studies provided to demonstrate typical analyses and troubleshooting. A major selling point of the ChemFET approach is the ability to evaluate the aqueous anion affinity of hydrophobic receptors that are not soluble in aqueous media and therefore cannot be studied using typical solution-based titration experiments.

Making the “improbable” probable

Typical rotaxane syntheses rely on complimentary non-covalent interactions between the linear and cyclic precursors to direct threading. Interlocking these components in the absence of specific molecular recognition seems improbable, yet Cong and co-workers have reported a synthesis of remarkable all-benzene rotaxanes, in which there are no obvious interactions between interlocked components [13]. The key innovation is the isolation of a pre-rotaxane module which covalently links the macrocycle with the beginnings of the axle via a removable azo-linker. Aryl bromide groups on the axle serve as handles for extension and stoppering via palladium coupling chemistry, and the azo group can be tracelessly cleaved to remove the covalent attachment and produce truly interlocked structures.

Disclosure statement

No potential conflict of interest was reported by the author(s).