A synthetic ion channel with anisotropic ligand response

Biological membranes play pivotal roles in the cellular activities. Transmembrane proteins are the central molecules that conduct membrane-mediated biochemical functions such as signal transduction and substance transportation. Not only the molecular functions but also the supramolecular properties of the transmembrane proteins such as self-assembly, delocalization, orientation and signal response are essential for controlling cellular activities. Here we report anisotropic ligand responses of a synthetic multipass transmembrane ion channel. An unsymmetrical molecular structure allows for oriented insertion of the synthetic amphiphile to a bilayer by addition to a pre-formed membrane. Complexation with a ligand prompts ion transportation by forming a supramolecular channel, and removal of the ligand deactivates the transportation function. Biomimetic regulation of the synthetic channel by agonistic and antagonistic ligands is also demonstrated not only in an artificial membrane but also in a biological membrane of a living cell.

Synthesis of (R)-16. Compound (R)-16 was synthesized from 6 by following the reported procedures. 1,2 Optical resolution of 13. To racemic 13 (3.15 g, 8.74 mmol) was added quinine (2.84 g, 8.74 mmol) and acetone (40 mL) at 25 °C, and the mixture was heated to reflux. After being cooled to 20 °C, colorless crystals 13•quinine precipitated out, which were collected by filtration and dried under reduced pressure. The crystals were dissolved in a mixture of Et2O (30 mL) and 4M HCl aq. (30 mL), and the organic extract was dried over Na2SO4 followed by filtration from the insoluble substances through a filter paper. The filtrate was evaporated to dryness under reduced pressure, and the residual yellow solid was reprecipitated with acetone and hexane to isolate (S)-13 (0.779 g, 2.16 mmol) as yellow crystals.
The colorless crystals of 13•quinine were recrystallized in CHCl3 and hexane to prepare single crystals for X-ray crystallographic analysis, which allowed for determination of the absolute configuration of the enantiomer of 13 to be S ( Supplementary Fig. 1).
Recrystallization of the racemic 13 by following the similar procedure using quinidine in place of quinine allowed for the isolation of (R)-13 in the enantiomeric excess of >99% evaluated by chromatography. CH2Cl2/MeOH (20/1) as an eluent to afford 19 (720 mg) as yellowish oil as a crude product.
Due to the instability, the product was utilized for the next reaction immediately. have prepared DOPC•2mer LUVs through pre-loading and post-loading methods, and compared their zeta potential. As shown in Supplementary Fig. 22, DOPC•2mer pre LUVs, DOPC•2mer post LUVs, and DOPC LUVs without 2mer showed zeta potential of -12.6 ± 0.37 mV, -18.4 ± 0.41 mV, and -6.2 ± 0.13 mV, respectively. We then estimated the orientationselectivity of 2mer using these values. For DOPC•2mer pre LUVs, 50% of 2mer is supposedly oriented to the extravesicular medium and the other 50% is oriented to the intravesicular medium. 4 Therefore, when 2mer is exclusively oriented to the extravesicular medium, its zeta potential should be the double of DOPC•2mer pre LUVs. Based on this assumption, the orientation-selectivity OS (%) for DOPC•2mer post LUVs was calculated as follows: where ζ0, ζpre, and ζpost represent zeta potential of DOPC LUVs without 2mer, DOPC•2mer pre LUVs, and DOPC•2mer post LUVs, respectively. The OS for DOPC•2mer post LUVs was therefore calculated to be 95%.  Fig. 29c) was located at each 2mer. In order to survey structural features of the complex with different ligands, the structure of the complex was equilibrated in DOPC bilayer membrane and water molecules ( Supplementary Fig. 29d), and a structural model was obtained after a 500 ps simulation.
The complexes with ligands kept a three-2mer complex ( Supplementary Fig. 29e,h), and however, the pore size among three 2mer differed for the PA and PPN bound complexes.
The pore size of the PA bound complex ( Supplementary Fig. 29f) was larger than the PPN bound complex ( Supplementary Fig. 29i), and PPN filled the pore among the three-2mer complex. This structural feature is consistent with the experimental suggestions of the ion permeability, depending on PA or PPN binding.
The structural difference of the complex with different ligands seems to be originated from the different interaction patterns between ligands and 2mer: PA mainly interacted with the phosphate group of 2mer through the electrostatic interactions. In contrast, PPN mainly formed hydrophobic packings with aromatic groups of 2mer, and the binding region of 2mer with PPN was much larger than that of PA, reflecting that the molecular size of PPN is larger than that of PA. In addition, PA were bound at the rim of the 2mer complex, and PPN was located at the deeper position of the pore than that of PA, despite the fact that PA and PPN were set to the same depth at the initial structure. These structural insights are consistent with the experimental findings by NMR studies (Supplementary Fig. 28).