Calix[6]arenes with halogen bond donor groups as selective and efficient anion transporters

Here we present the anion binding and anion transport properties of a series of calix[6]arenes decorated on their small rim with either halogen bond or hydrogen bond donating groups. We show that the halogen bond donating iodotriazole groups enable highly selective transport of chloride and nitrate anions, without transport of protons or hydroxide, at rates similar to those observed with thiourea or squaramide groups.

A dataset for this publication is available at Zenodo (https://doi.org/10.5281/zenodo.6010342) and contains: • NMR spectra for the characterisation of compounds 1a, 1b, 1c, 2, and 3 (Mestrenova files) • NMR spectra for the titration experiments with compounds 1-5 in different solvents (Mestrenova files) • Concentrations of host and guests in the various titration experiments (Excel file) • Transport data in the lucigenin assay (Excel file) • Transport data in the HPTS assay (Excel file) Electronic Supplementary Material (ESI) for ChemComm. This journal is © The Royal Society of Chemistry 2022

General experimental information
All reagents and solvents were obtained from Sigma Aldrich, Fluorochem, Alfa Aesar and VWR, and were used without further purification unless otherwise stated. 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) and cholesterol were purchased from Sigma Aldrich and Acros, respectively. Lipid solutions of POPC and cholesterol were prepared using chloroform that had been deacidified by passage through a column containing basic alumina. POPC solutions were stored at -20 °C and cholesterol solutions were freshly prepared. All aqueous solutions were prepared using deionised water that had been passed through a Millipore filtration system. All liposomes were used within 4h after their preparation.
All NMR spectra for the characterization of compounds ( 1 H, 13 C, COSY, HSQC, and HMBC) were recorded on a Jeol JNM-ECZ400R/S3 spectrometer equipped with a 5 mm NM Royal probe at 298 K. 1 H NMR spectra for titrations were recorded on a Varian VNMRS 400 (9.4 T) spectrometer equipped with an OneNMR probe at 298 K. Solvent signals were used as reference for the chemical shifts. Chemical shifts are expressed in ppm and the coupling constants (J) are expressed in Hertz (Hz). The 1 H residual signals of the solvent were used as reference (CDCl3 7.26 ppm, DMSO-d6 2.50 ppm, and acetone-d6 2.05), as well as the 13 C signals of the solvent (CDCl3 77.16 ppm, DMSO-d6 39.52 ppm, and acetone-d6 206.26).
High-resolution mass spectra were measured on an Agilent QTOF 6520 by electron spray ionisation.
Fluorescence measurements were carried out on a FluoroMax-4 (Horiba) spectrofluorometer equipped with a water-thermostatted cell holder with stirring and an injection port.

Synthesis
Anhydrous THF was distilled over sodium and benzophenone. All the reactions were carried out under inert atmosphere with argon unless stated otherwise. For monitoring the progress of the reaction, thin layer chromatography was performed on silica gel 60 F254 coated aluminum sheets. The compounds were revealed using a UV light source of 254 nm. Whenever required, the purification was performed using column chromatography on silica gel (VWR chemicals, particle size 35-70 m).

Anion binding studies
Binding constants were determined using 1 H NMR spectroscopy and by titrating tetrabutylammonium (TBA) salts into solutions of different calixarenes in deuterated solvents at 298 K. Solutions of receptors (1 mM, 1.5-2 mL) were prepared in deuterated solvent (CDCl3, or acetone-d6, or DMSO-d6). Stock solutions of 0.5 M TBA salts were prepared by dissolving the salt (dried under vacuum) into the different receptor solutions. 500 μL of the solutions of pure receptors were transferred into NMR tubes. Initial spectra were recorded, and aliquots of the guest solutions were then added to the NMR tube and 1 H NMR spectra were recorded after each addition of guest.
The shifts of the signals closest to the binding site were determined and the data were fitted to a 1:1 (host:guest) binding model using the Bindfit v0.5 applet (available as freeware from Supramolecular.org).
Calix[6]arenes with halogen bond donor groups as selective and efficient anion transporters.

Figure S27
Observed changes in chemical shifts and calculated binding curves for the titration of 1a (1 mM) with TBACl at 298K in acetone-d6. The signals of protons 'g', 'i' were used for fitting to a 1:1 binding model, resulting in a Ka of 938924 M -1 (208%), which is above the limit where the Ka can be accurately determined, thus Ka >10 5 M -1 .

Figure S29
Observed changes in chemical shifts and calculated binding curves for the titration of 1c (1 mM) with TBACl at 298K in acetone-d6. The signals of protons 'g', 'i' were used for fitting to a 1:1 binding model, resulting in a Ka of 8936 M -1 (10%           For this titration, no significant change in the chemical shift of the proton signals was observed, and we thus conclude that the Ka (1:1) < 10 M -1 .

H NMR titration of calixarene 1a with TBAOAc in CDCl3
Figure S45 1 H NMR spectra (400 MHz) from the titration of calixarene 1a (1 mM) with TBAOAc in CDCl3 at 298 K. The number of equivalents of TBAOAc relative to 1a is shown.
For this titration, no significant change in the chemical shift of the proton signals was observed, and we thus conclude that the Ka (1:1) < 10 M -1 . For this titration, no significant change in the chemical shift of the proton signals was observed, and we thus conclude that the Ka (1:1) < 10 M -1 .

H NMR titration of calixarene 2 with TBACl in CDCl3
Figure S49 1 H NMR spectra (400 MHz) from the titration of calixarene 2 (1 mM) with TBACl in CDCl3 at 298 K. The number of equivalents of TBACl relative to 2 is shown.
For this titration, no significant change in the chemical shift of the proton signals was observed, and we thus conclude that the Ka (1:1) < 10 M -1 .
Calix[6]arenes with halogen bond donor groups as selective and efficient anion transporters.    In acetone-d6, the affinity of 3 for Cl − was found to be slightly lower compared to 4 and 5. However, in DMSO-d6 with 0.5% H2O, receptor 3 had the highest affinity of series 3-5.

Preparation of the vesicles
Liposomes were prepared with sodium salts of various anions (NaA) both inside and outside, where the following aqueous solutions of NaA are used: NaNO3 (225 mM), NaOAc (225 mM), and NaHCO3 (225 mM, adjusted to pH 8 with a small amount of H2SO4).
Phospholipid and cholesterol solutions were combined with solutions of transporters in deacidified chloroform or methanol in a 5 mL round-bottomed flask. Volumes were calculated from the lipid concentrations to obtain a final concentration of 0.4 mM in lipids (POPC + cholesterol), with a 7:3 POPC to cholesterol ratio and a 1:1000, 1:5000, or 1:25'000 transporter to lipid ratio. The solvents from the mixture were evaporated under a flow of dry air and the resulting lipid film was dried in vacuum for at least 1 h. The lipid film was then hydrated with 500 μL of an aqueous solution of 10,10'-dimethyl-9,9'-biacridinium nitrate (lucigenin, 0.8 mM) and NaA (225 mM), sonicated for ca. 30 s and stirred for at least 1 h at room temperature. The heterogeneous multilamellar vesicles were broken down into unilamellar vesicles by 10 freeze-thawing cycles, diluted to 1 mL with NaA solution, and extruded 29 times through polycarbonate membranes (200 nm pore size) at room temperature to give homogeneous large unilamellar vesicles. The external lucigenin was removed by passing the vesicle solution through size exclusion columns (Sephadex G-25) eluted with its respective salt solution.
The resulting vesicles were further diluted with the same salt solution to obtain a final concentration of 0.4 mM in lipids (calculated from the initial quantities of lipids).

General procedure for transport measurements with lucigenin
Fluorescence measurements were performed on 3 mL of the final vesicle solution in a quartz cell with a stir bar, and the fluorescence intensity of lucigenin (excitation at 430 nm and emission at 505 nm) was recorded over time. The temperature of the cuvette holder was controlled with a water bath at 25 °C. 75 μL of aqueous NaCl (1 M, in the same salt solution as the liposomes, to give an external chloride concentration of 25 mM) was added ca. 30 seconds after starting the experiment and the fluorescence was recorded for an additional 10 minutes before lysing of the vesicles with 50 μL of Triton X-100 (5 wt.%) in water.
Each experiment was repeated at least 3 times and the fluorescence data were averaged after removing the initial drop (due to the quenching of remaining external fluorophore) and normalizing each fluorescence value (F) to the initial value (F0). Traces of 500 seconds of transport data are plotted.

Figure S56
Representative transport curves for the lucigenin assay. a) No transport of Cl − in the absence of transporter, b) Transport of Cl − in the presence of 1a (1:5k transporter to lipid ratio), as monitored by the lucigenin assay in 225 mM NaNO3, upon addition of 25 mM NaCl. The green box indicates the part of the curves that is normalized, averaged, and plotted in Figure 3a and b.
All data recorded in NaCl solution are provided in the main text, while data for Cl − /HCO3 − and Cl − /AcO − antiport are provided in Figures S57-58.
Calix[6]arenes with halogen bond donor groups as selective and efficient anion transporters.

Quantification of transport rates
The quantification of the transport rates (see Table 1) was performed as described previously. vi According to the Stern-Volmer equation, the inverse of the normalized fluorescence intensity (F0/F) is directly proportional to the concentration of chloride inside the vesicles.
The obtained curve for F0/F (0-500 s) is fitted to the double exponential function: Differentiating the function with respect to t at t=0 gives Further, the specific initial rate [I] is obtained by dividing the initial rate I by transporter to lipid ratio.

Figure S57
Transport of Cl − by compounds 1a and 3 (preincorporated in the vesicles at 1:1k transporter to lipid ratio) as monitored by the lucigenin assay in 225 mM NaHCO3, upon addition of 25 mM NaCl. 4.2 HPTS assay with a NMDG base pulse A 1.0 M solution of N-methyl-D-glucamine (NMDG) was prepared and 50 mL of this solution was combined with the appropriate amounts of HCl and HEPES to prepare 500 mL aqueous solution of N methyl-D-glucamine hydrochloride (NMDGH + Cl − , 100 mM) and HEPES (10 mM) at pH 6.8. Similarly, a solution of NMDGH + NO3 − (100 mM) and HEPEs (10 mM) at pH 6.8 was prepared by combining and diluting solutions of NMDG, HNO3, and HEPES. HPTS was dissolved in these NMDGH + A − solutions at a concentration of 0.1 mM.  in deacidified chloroform were combined with a solution of test receptor (1 mM in chloroform) in a 5 mL round bottom flask to obtain a POPC:cholesterol ratio of 7:3 and ratios of receptors to the total amount of lipids of 1:1000, 1:5000 and 1:25000. The solvents were evaporated under a flow of air and the resulting lipid film was dried under high vacuum for 1 h. The lipid film was then hydrated with 500 μL of a solution of 0.1 mM HPTS in the solution of NMDGH + A − at pH 6.8. The resulting mixture was sonicated for 30 s and stirred for 1 h to give heterogeneous vesicles. Multilamellar vesicles were disrupted by 10 freeze-thaw cycles. The mixture was diluted to 1 mL (by adding 0.5 mL of NMDGH + A − solution) and carefully extruded 29 times through a polycarbonate membrane with 200 nm pores in a mini-extruder (Avestin LiposoFast-Basic). The external HPTS was removed by passing the liposomes though a pre-packed size exclusion column (containing 8.3 mL Sephadex G-25 medium), eluted with NMDGH + A − solution. The collected large unilamellar vesicles were further diluted with NMDGH + A − solution to obtain a solution with 0.1 mM total lipid concentration at pH 6.8.

General procedure for transport measurements with HPTS
3.00 mL of this liposome solution was placed in a quartz cuvette with a small stir bar and the temperature was allowed to stabilize at 25 ˚C for 3-5 minutes inside the sample compartment of a Fluoromax-4 spectrometer. The fluorescence intensities (excitation at 403 and 455 nm, emission at 511 nm) were measured over time at 25 ˚C while stirring the solution, and 30 μL of NMDG (0.5 M in water) was added 30 seconds after the start of the measurement to increase the external pH to 7.7. The dissipation of this pH gradient was monitored by following the fluorescence for 200 seconds, after which the liposomes were lysed by addition of 60 μL of Triton X-100 (5% w/w in water). These experiments were repeated with carbonyl cyanide 3-chlorophenylhydrazone (CCCP, 5 μL of a 60 μM solution in methanol; 1:1k CCCP to lipid ratio) added to the liposomes (3 minutes before the start of the measurement), serving as proton transporter. The fluorescence intensity with excitation at 455 nm (for deprotonated HPTS) was divided by the fluorescence intensity with excitation at 403 nm (for protonated HPTS). The ratios of intensities over time were normalized from 0 (before addition of base) to 1 (plateau upon lysing of the liposomes) according to the formula below.

Determination of the EC50 value
As concentrations higher than 1:1000 transporters per lipid could not be pre-incorporated into the LUVs reliably, no plateau of transport could be reached, and the data could not be fitted to the Hill equation. Therefore, the EC50 value for the transport response by 1a (in presence of CCCP, at 220s) was estimated from a logarithmic fit as 50% of the response from the normalized intensity ratio in absence of transporter (at 220 s, 0%) to the lysis level (100%). These studies on the rate of Cl − and NO3 − uniport in HPTS-based assays show higher transport rates for 1a

HPTS assay with a TBAOH base pulse
A 100 mM solution of sodium gluconate was buffered at pH 7.0 with 10 mM HEPES. HPTS was dissolved in this 100 mM sodium gluconate solution at a concentration of 0.1 mM.
POPC and cholesterol solutions (15-20 mM) in deacidified chloroform were combined with a solution of receptor (1 mM in chloroform) in a 5 mL round bottom flask to obtain a POPC:cholesterol ratio of 7:3 and ratios of receptors to the total amount of lipids of 1:5k. The solvents were evaporated under a flow of air and the resulting lipid film was dried under high vacuum for 1 h. The lipid film was then hydrated with 500 μL of a solution of 0.1 mM HPTS in the solution of 100 mM sodium gluconate buffered at pH 7.0 with 10 mM HEPES. The resulting mixture was sonicated for 30 s and stirred for 1 h to give heterogeneous vesicles. Multilamellar vesicles were disrupted by 10 freeze-thaw cycles. The mixture was diluted to 1 mL (by adding 0.5 mL of buffered sodium gluconate solution) and carefully extruded 29 times through a polycarbonate membrane with 200 nm pores in a mini-extruder (Avestin LiposoFast-Basic). The external HPTS was removed by passing the liposomes though a pre-packed size exclusion column (containing 8.3 mL Sephadex G-25 medium), eluted with buffered sodium gluconate solution. The collected large unilamellar vesicles were further diluted with sodium gluconate solution to obtain a solution with 0.1 mM total lipid concentration at pH 7.0.
The transport experiments were performed in the same way as described for those with the HPTS assay and adding a NMDG base pulse. The only difference is that here 30 μL of 0.5 M tetrabutylammonium hydroxide (TBAOH, in the solution of 100 mM sodium gluconate) was added 30 seconds after the start of the measurement to increase the external pH to 7.6-7.7.
In this assay, the movement of charge associated with the transport of protons or hydroxide by the receptors can be balanced by the free diffusion of the tetrabutylammonium cation through the membrane, while gluconate anions are too polar to allow their transport. vii Also in this assay, 1a does not show any transport activity, indicating that this compound cannot transport H + nor OH − .

Single-crystal X-ray diffraction analysis
Crystals suitable for single-crystal X-ray diffraction analysis were obtained by the slow evaporation of 1a in dichloromethane at room temperature. Diffraction data were collected using the Oxford Diffraction Gemini R Ultra diffractometer (Cu Kα, multilayer mirror, Ruby CCD area detector) at 120(2) K. Intensity measurements were performed on a rapidly cooled crystal (0.35 x 0.30 x 0.22 mm 3 ) in the range 2.55º ≤ θ ≤ 67.14º. Data collection, unit cells determination and data reduction were carried out using CrysAlis PRO software package viii using Olex2 ix and shelXle xa , the structure was solved with the SHELXT 2015 xb structure solution program by Intrinsic Phasing methods and refined by full-matrix least squares on |F| 2 using SHELXL-2018/3 xb Non-hydrogen atoms were refined anisotropically, while C-H hydrogen atoms were placed in geometrically calculated positions using a riding model. In the asymmetric unit, the 2-{4-[3,5-bis(trifluoromethyl)phenyl]-5iodo-1H-1,2,3-triazol-1-yl}ethan-1-oxy-group is disordered over three positions in such way that in two position the iodine atom orientated towards the cavity and one orientated in the opposite direction). When the iodine orientated towards the cavity it is coordinated to water molecule. Occupancies of two first positions are equal to occupancy of the water molecule. Disordered hydrogen atoms of water were omitted from the refinement. Besides, the asymmetric unit also contains one molecule of dichloromethane disordered over 3-fold axis. CCDC 2125602 contains the supplementary crystallographic data for this paper. Copies of the data can be obtained free of charge via http:/www.ccdc.cam.ac.uk/const/retrieving.html or from the Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: (+44)1223-336-033; e-mail: deposit@ccdc.cam.ac.uk). The most relevant crystallographic data are summarized in Table S3.