Photoswitchable Bis(amidopyrroles): Modulating Anion Transport Activity Independent of Binding Affinity

Toward photocontrol of anion transport across the bilayer membrane, stiff-stilbene, which has dimethyl substituents in the five-membered rings, is functionalized with amidopyrrole units. UV–vis and 1H NMR studies show high photostability and photoconversion yields. Where the photoaddressable (E)- and (Z)-isomers exhibit comparable binding affinities, as determined by 1H NMR titrations, fluorescence-based transport assays reveal significantly higher transport activity for the (Z)-isomers. Changing the binding affinity is thus not a necessity for modulating transport. Additionally, transport can be triggered in situ by light.


UV-vis photoisomerization studies
Vesicles with transporter pre-incorporated were prepared from a stock solution of 10 mM POPC and 1 mM transporter (10 mol%) in CHCl3, which was diluted with a 10 mM POPC solution to the appropriate molar ratio. Using the above protocol, with the exception that now an Illustra NAP™-25 Sephadex® G-25 columns was used, stock solutions with a concentration of 1.6 mM in lipid and with various concentrations of the transporter were obtained.
For each measurement, the stock solution was diluted with buffer to a standard volume (2.5 mL) in a PS cuvette to obtain a solution with a concentration of 0.1 mM in lipid. The sample was stirred at rt in the fluorimeter, and the fluorescence ratio of HPTS (λex = 454 nm, λem = 511 nm, base form, divided by λex = 403 nm, λem = 511 nm, acid form) was measured over time. At t = 30 s, the compounds were added as a DMSO solution (5.0 µL, varying concentrations by dilution from a 5 mM stock solution). To initiate transport, a pulse of NaOH (25 µL, 0.5 M) was given at t = 60s to generate a pH gradient of pH 7 inside and pH 8 outside.
Vesicles were lysed with Triton X-100 (50 µL, 11 wt% in H2O/DMSO 7:1 v/v) at t = 370s, and a final reading was taken at t = 430s. The fractional fluorescent intensity (If) was then calculated using the following formula: Where Rt is the fluorescent ratio at time t, R0 is the initial fluorescent ratio measured just prior to the base-pulse at t = 59s, and Rd is the final ratio measured at t = 430s.

Hill analysis
The transport activity measured at t = 360s was plotted as a function of transporter concentration. Using Origin 2022 this data was fitted to the Hill equation: Where y is the If at t = 360s and x is the transporter concentration (in mol% with respect to lipid). The parameters to be fitted are: y0 which is the If when no transporter is added, ymax is the maximum If that is obtained by the receptor, k is the concentration needed to get 50% of this observed maximum If and n is the Hill coefficient.
The EC50 value is defined as the concentration needed to reach 50% of the maximum possible chloride efflux. The value of k will only correspond to EC50 if the receptor is able to induce 100% of efflux. In the case that 100% efflux is not reached even at the highest loadings, e.g.

In situ irradiation
A vesicle solution (0.1 mM in lipids, 2.5 mL) with 0.1 mol% of compound (E)-2 preincorporated, was prepared as stated in the above protocol for the HPTS assay. The fluorescence ratio of HPTS was followed over time, and subsequently converted to the fractional fluorescent intensity (If) using equation S1. At t = 60s, the NaOH base-pulse (25 µL, 0.5 M) was added to initiate transport. During the experiment samples were irradiated between t = 120s -150s using a 340 nm or 365 nm LED mounted on the lid of the fluorimeter (see Figure S54). At t = 370s the vesicles were lysed with Triton X-100 (50 µL, 11 wt% in H2O/DMSO 7:1 v/v), and a final reading was taken at t = 430 s. The traces of the irradiated samples were compared to traces that were not irradiated. and mounting of the LED for irradiation of the sample (right).

Cationophore coupled assay
For the cationophore-coupled assays a solution of 37.5 mM POPC and 0.375 mM transporter (1 mol%) in CHCl3 was prepared in a round bottom flask. The solvent was evaporated to create a lipid film which was dried under vacuum for at least 12 h. The film was hydrated by vortexing with the internal solution consisting of KCl (300 mM) buffered to pH 7.20 with HEPES (10 mM). Next, the suspension was subjected to 9 freeze and thaw cycles, using liquid nitrogen and a water bath at 45°C. The solution was left standing at rt for 30 min, and subsequently extruded 25 times through a 200 nm polycarbonate membrane to obtain unilamellar vesicles. The external solution was exchanged for one containing KGlu (300 mM) buffered to pH 7.20 with HEPES (10 mM) using an Illustra NAP™-10 Sephadex® G-25 column, which afforded stock solutions with a concentration between 15-18 mM in lipid.
During the assays the chloride concentration was measured using an Accumet chlorideselective electrode. For each measurement, the stock solution was diluted with buffer to a standard volume (5.0 mL) in a glass vial to obtain a solution with a concentration of 1.0 mM in lipid. The sample was stirred at rt and the potential was measured over time using the electrode. At t = 60s the cationophore (valinomycin or monensin) was added as a DMSO solution (10 μL, 0.5 mM, 0.1 mol%) to initiate transport. Vesicles were lysed with Triton X-100 (50 µL, 11 wt% in H2O/DMSO 7:1 v/v) at t = 360s and a final reading corresponding to 100% efflux was taken at t = 480s.
The electrode was calibrated prior to each experiment according to the supplier's manual, by recording the potential of solutions with a known chloride concentration. A calibration curve was generated by fitting to the modified Nernst equation (Equation S4).

= + ln ( )
Where y is the potential (mV), x is the chloride concentration and a and b are the parameters to be fitted. Using the calibration curve, readings (mV) of the experiment were converted to chloride concentrations, and subsequently to percentages of efflux using equation S5.
Where Ct is the concentration at time t, C0 is the initial concentration measured just prior to addition of the cationophore at t = 57s, and Cd is the final ratio measured at t = 480s. DMSO was used as control (each measurement done in duplicate).

Geometry optimization by DFT
Input geometries were generated using ArgusLab. [3] The Gaussian 09 program [4] was used for geometry optimization: First, energy minimization was performed at the semi-empirical PM3 level of theory, and subsequently at the DFT B3LYP/6-31++G(d,p) level of theory using an IEF-PCM DMSO solvation model. The DFT optimized geometries were found to have zero imaginary frequencies. Table S1. Cartesian coordinates of (Z)-1⊂Cl − .   Table S2. Cartesian coordinates of (Z)-1⊂OAc − .