Fluorogenic In Situ Thioacetalization: Expanding the Chemical Space of Fluorescent Probes, Including Unorthodox, Bifurcated, and Mechanosensitive Chalcogen Bonds

Progress with fluorescent flippers, small-molecule probes to image membrane tension in living systems, has been limited by the effort needed to synthesize the twisted push–pull mechanophore. Here, we move to a higher oxidation level to introduce a new design paradigm that allows the screening of flipper probes rapidly, at best in situ. Late-stage clicking of thioacetals and acetals allows simultaneous attachment of targeting units and interfacers and exploration of the critical chalcogen-bonding donor at the same time. Initial studies focus on plasma membrane targeting and develop the chemical space of acetals and thioacetals, from acyclic amino acids to cyclic 1,3-heterocycles covering dioxanes as well as dithiolanes, dithianes, and dithiepanes, derived also from classics in biology like cysteine, lipoic acid, asparagusic acid, DTT, and epidithiodiketopiperazines. From the functional point of view, the sensitivity of membrane tension imaging in living cells could be doubled, with lifetime differences in FLIM images increasing from 0.55 to 1.11 ns. From a theoretical point of view, the complexity of mechanically coupled chalcogen bonding is explored, revealing, among others, intriguing bifurcated chalcogen bonds.


■ INTRODUCTION
Thio/acetals are popular functional groups in fluorescent probes.Thioacetals have been used for mercury sensing, 1−6 quenching, 7−11 and the detection of reactive oxygen species, 12−15 while acetals have been used to sense pH changes, 16−20 detect metal ions, 21−23 visualize and modulate enzyme activity, 24−27 and for photochromic switching 28−31 and super-resolution microscopy. 32Usually, the conversion of aldehydes into thio/acetals is designed to turn off fluorescence, or vice versa, for different reasons.In this study, thio/ acetalization is designed to turn on, rather than turn off, the fluorescence of flipper probes easily, at best in situ, to simultaneously install variable targeting groups and screen for new chalcogen-bonding donors.
Fluorescent flippers have been introduced as small-molecule fluorescent probes to image membrane tension in living systems (Figure 1a−c). 33For the imaging of biomembrane function, 34−47 membrane tension 48−50 is particularly interesting but also particularly demanding because physical forces are detectable only through the suprastructural changes they cause. 33Inspired by nature, 51 we have considered the concept of planarizable push−pull probes to tackle this challenge. 52he resulting flipper probes are built around two dithienothiophenes, 53,54 one electron rich and one electron poor.In the relaxed ground state I, they are twisted out of coplanarity by chalcogen-bonding repulsion (Figure 1a).Planarization by mechanical compression in the ground state brings donors and acceptors into conjugation.The resulting push−pull system causes a large red-shift of the absorption maximum and increases fluorescence lifetime, intensity, and quantum yield.−47 In lipid bilayer membranes, these changes report on an increasing order from liquid-disordered (L d ) to liquid-ordered (L o ) and solid-ordered (S o ) membranes (Figure 1b).Tension applied to biomembranes increases fluorescence lifetimes because the probe response is dominated by membrane reorganization, particularly tension-induced phase separation (Figure 1c).
Progress with flipper probes has been limited by the synthetic effort required for their preparation.Particularly challenging is the donor terminus.−57 Moreover, donors D are needed in planarized flippers II to generate a strong push−pull system accounting for large red-shifts, but they are incompatible with twisted flippers I because the decoupled donor side oxidizes. 33lippers such as 1 with simple alkoxy donors are not stable for this reason (Figure 1d).Sulfide donors, like in 2, that turn on only when attached to electron-deficient aromatics obtained by the planarization of flippers, did not afford the spectroscopic properties needed for bioimaging.The first functional flipper 3 contains a thenyl ester as a donor and a carboxylate to target plasma membranes.However, flipper 3 was unstable because the thenyl esters and even ethers were easily eliminated (Figure 1g).This problem was not solved but suppressed in flipper 4 by a proximal triazole as a proton scavenger, which works well as a tension reporter (TR) in cells and has been made available for the community as Flipper-TR.
Deletion of the fragile thenyl ether and direct use of triazole as donor, as in 5, increased stability, but the added aromatic system perturbed spectroscopic properties. 54,57Both thenyl ether and triazole have been proposed to act as noncovalent turn-on donors, forming a chalcogen bond 58−63 as soon as probe planarization deepens the σ hole on the endocyclic sulfur (Figure 1f).Considering the decisive importance of these hypothetical chalcogen-bonding turn-on donors 64 on the one hand and the limitations of thenyl ethers with regard to stability and variability on the other hand, we decided to move one oxidation level higher and explore thio/acetals as clickable donor junctions in fluorescent flipper probes as exemplified in 6.They are shown to provide facile access to rich chemical space, intriguing chalcogen-bonding donor motifs, and more than doubled sensitivity for the imaging of membrane tension in living cells.

Design
Thio/acetal donor junctions could be installed by late-stage modification of common aldehyde precursor 7 with the respective thiols and alcohols III (Figure 1e).Saturated 1,3heterocycles IV (Figure 1e) were particularly inviting because classics like cysteines, lipoic acid, asparagusic acid, 65,66 DTT, 67,68 epidithiodiketopiperazines (ETPs), 69 and so on promise access to rich chemical space.This flipper diversity will allow for the screening for improvements of the original chalcogen-bonding donors V together with an additional, less powerful noncovalent CH•••X bonding donor for a refined push−pull system VI (Figure 1f).Moreover, saturated 1,3heterocycles IV were expected to prevent the thenyl ether elimination VII by intramolecular tethering of the leaving group to enable reverse ring closure (Figure 1g, VIII−X).Overall, donor junctions could thus be expected to secure facile synthetic access to a broad variety of turn-on donors with attached targeting units of free choice, all in one single final step, at best possible in situ.

Molecular Modeling
Computational evaluation of potential chalcogen-bonding donors was performed at the PBE0-D3/def2-TZVP level of theory.The virtual flipper 8 was of interest in silico because the view on the relevant σ hole is not obstructed by chalcogen bonds.Upon planarization, the maximum of the molecular electrostatic potential (MEP) surface on the sulfur increased from E pot = 21.5 to E pot = 23.5 kcal mol −1 , while the C−H donor increased only by E pot = +0.2kcal mol −1 (Figure 2a).The conclusion that chalcogen-bonding donors turn on upon planarization was confirmed by the switching cycle 64 of the pseudo-pull−pull flipper 7 with a misplaced aldehyde acceptor in place of the exocyclic push−pull donor of flipper probes.A macrodipole doubling from μ = 3.2 D for the planar conformer 7a-II to μ = 6.5 D for 7b-II showed how chalcogen bonding in 7b-II but not CH•••O bonding in 7a-II weakens the aldehyde acceptor by reinjection of withdrawn electron density (Figure 2c and Table 1, entries 1, 2).The twice as large push−pull dipole stabilized 7b-II by E rel = −0.36kcal mol −1 compared to 7a-II, while the twisted 7b-I with weaker dipole and chalcogen bonds was only E rel = −0.31kcal mol −1 more stable than 7a-I (Figure 2b).The uphill planarization of the twisted 7b-I (E rel = +1.86kcal mol −1 ) was also facilitated by the doubled push− pull dipole of 7b-II (E rel = +1.82kcal mol −1 ).
Conversion of aldehyde 7 into minimalist thenyl ether 9 and acetal 10 increased the push−pull dipole from μ = 6.5 D over μ = 9.9 to μ = 10.9D, which in turn stabilized planar high-energy conformer II from E rel = +1.82kcal mol −1 over E rel = +1.75kcal mol −1 to E rel = +1.69kcal mol −1 relative to the twisted conformer I (Figure 2c and Table 1, entries 2−4).The origin In the sulfur series, dipole and stabilization of planar conformer II increased from acyclic to cyclic thioacetals.With cyclic thioacetals, dipoles increased with the ring size from μ = 9.4 D for 1,3-dithiolane 12 over μ = 10.5 D for dithiane 13 to record μ = 11.1 D for dithiepane 14 (Figure 2c and Table 1, entries 6−8).This was reflected in the chalcogen bond length decreasing from acyclic 11 with d = 3.23 Å to cyclic 12 and 14 with d ∼ 3.14 Å, all well below vdW radii of d = 3.60 Å (Figure 2c and Table 1, entries 5, 6, 8).The longest chalcogen bond d = 3.38 Å of 13 coincided with the weakest NBO electron transfer contribution of E ET = 0.28 kcal mol −1 , while the push− pull dipole remained very high at μ = 10.5 D and planarization with E rel = +1.69kcal mol −1 was least disfavored (Figure 2c and Table 1, entry 7).This apparent contradiction could be understood with the highest conformational rigidity of the 1,3dithiane chair, which positions both sulfur atoms at almost equal distance for a formal bifurcated 70−72 three-center chalcogen bond, 73 where long distances are compensated by doubled interactions (d 1 = 3.39 Å and d 2 = 3.46 Å, Figure 2f,g).The existence and attractive nature of a bifurcated chalcogen bond were confirmed by NCIplot analysis of 13, which showed the presence of two reduced density gradient disk shape green isosurfaces (Figure 2g).This unorthodox bifurcated chalcogen bond occurred with dithiane 13 but not with the ring-contracted dithiolane 12, the ring-expanded dithiepane 14, and dioxane 10 with oxygen instead of sulfur donors in the chair.In dithiolane 12, the second heteroatom was oriented toward the CH•••X bond acceptor, allowing the chalcogen-bonding heteroatom to position best for minimal bond length (d = 3.13 Å) and maximal NBO electron transfer (E ET = 3.31 kcal mol −1 , Figure 2c−e and Table 1, entry 6).The same was true in the oxygen series with dioxane 10, with the NBO electron transfer being naturally smaller (Figure 2c and Table 1, entry 4).Increasing conformational flexibility in dithiepane 14 presumably accounted for the best balance of all parameters, resulting in a record dipole together with a short, nonbifurcated chalcogen bond and substantial NBO electron transfer (E ET = 2.33 kcal mol −1 , Figure 2c and Table 1, entry 8).
Almost equal stabilization (E rel = +1.69kcal mol −1 ) and polarization (μ = 10.5 D) of planar high-energy conformer 13 compared to ring-contracted 12 implied that unorthodox bifurcated chalcogen bonds can be at least as powerful as  optimized conventional chalcogen single bonds.The significance of these bifurcated chalcogen bonds could only be identified and appreciated in the context of the coupled processes in planarizable push−pull probes.They will be of interest in future design strategies in general.The overall trends identified by molecular modeling reflect the complexity of the system.Also, without inclusion of charge transfer and excited-state structures, they should thus be considered with due caution in interpreting flipper performance.
In Situ Thioacetalization Aldehyde 7 was obtained from previously reported ether 15 by simultaneous deprotection and oxidation with DDQ (Figure 3a and Scheme S1).Flipper 15 was synthesized in 12 steps from commercially available starting materials following reported procedures. 33Weakening of the push−pull system in 15 by the aldehyde acceptor in 7 was correctly reflected by the emission changing from orange to green.As initial targets for donor junctions, 1,3-dioxolane 16 and thioacetals 17−20 were selected to probe for accessibility of the central motifs from popular starting materials like reduced asparagusic acid 65,66 or dithiothreitol 67,68 (DTT, Figures 3 and 4a).Thio/acetalization of the pseudo-pull−pull fluorophore 7 was visible by the naked eye by a change from green back to orange fluorescence (Figure 3a).As expected from the installation of a push−pull system, this red-shift coincided with a de facto turnon increase in fluorescence (Figure 3c).
In the twisted form I of substrate and product in solution, changes in the absorption and excitation spectra were naturally less spectacular.The spectrum of substrate 7 in chloroform showed two maxima at 367 and 406 nm (Figure 3b, blue).Upon thioacetalization, the two maxima moved apart to longer and shorter wavelengths (Figure 3b,d, red).Direct detection of thioacetalization was realized in the absorption spectra (Figure 3d).The formation of the acyclic thioacetal 17 occurred with a t 50 ∼ 6 min (Figure S10).The in situ synthesis of 1,3-dithiolane 18 showed similar kinetics, while the formation of dithiane 19 was faster (t 50 ∼ 2 min) and dithiepane 20 formed instantaneously (t 50 < 2 min).Thioacetal formation was confirmed by LC-MS (Figures S11−S16) and NMR analyses, performed after purification of some compounds.Added to HeLa Kyoto (HK) cells with or without short work-up, in situ produced thioacetals were not toxic and afforded FLIM images that were indistinguishable from images obtained from isolated and purified flippers (Figure 4b).
Thioacetal flipper 19 was stable in buffer at pH = 7 and pH = 5 for more than 15 h (Figure S22a,b).The complementary acetal flipper 16 showed identical stability at pH = 7 but hydrolyzed at pH = 5 with t 50 = 4 h (Figure S22c,d).Usually, the replacement of oxygen by sulfur in functional groups decreases stability and enables dynamic covalent chemistry because the larger size gives weaker bonds, stronger acids, and better leaving groups.The exception with thioacetals, long known, originates from the different mechanism at work, with C�S + intermediates that are harder to access than C�O + intermediates, for the same reason.
The spectroscopic properties in L d LUVs (large unilamellar vesicles) of all new flippers were roughly the same as for the original Flipper-TR 4, characterized by a broad excitation maximum at λ ex ∼ 440 nm (Figure S17).In L o LUVs, the excitation maxima red-shifted around +80 nm to a 0−0 transition at λ ex ∼ 520 nm, with distinct differences between the different flippers, particularly with regard to vibrational fine structure (vide infra, Figures 3b and S18−S21).These results indicated that, remarkably, all new donor junctions produced operational, mechanosensitive flippers.

Fluorescence Lifetime Imaging Microscopy
The full flipper collection was analyzed by FLIM (fluorescence lifetime imaging microscopy) of GUVs (giant unilamellar vesicles) and HK cells (Figures 4 and S23−S29, Table 2).Lifetimes τ av or τ 1 for the longer component were extracted from FLIM images by biexponential fitting of decay curves.Reported from GUVs are τ 1 in L o membranes, termed τ Lo (SM/CL 7:3, Figure 4c, top), their difference Δτ GUV to τ 1 in L d membranes (DOPC), and τ Lo m , τ 1 in L o domains of mixed membranes containing L o and L d domains (DOPC/SM/CL 58:25:17, Figure 4c, bottom, and 4a, Table 2).From HK cells, reported are τ iso for τ 1 of plasma membranes under isoosmotic conditions (Figures 4a and 4e,f, top) and their difference Δτ TR to τ 1 under hyperosmotic conditions (Figures 4e,f, bottom, and 4a, Table 2).
Within the series 16−20, selected to implement computational guidelines with readily accessible reagents, fluorescence lifetimes were generally higher for S−O chalcogen bonds, that is, ether 4 and acetal 16, than for thioacetals 17−20 (Figure 4a and Table 2, entries 1−6).In the thioacetal series, lifetimes generally decreased with an increasing flipper macrodipole.This trend was presumably caused by increasing flipper mispositioning rather than differences in chalcogen bonding.Under isoosmotic conditions, the selectivity of plasma membrane labeling was best for the most hydrophilic, dianionic acyclic thioacetal 17 (Figure 4d, bottom).The worst selectivity coincided with the shortest lifetimes for flipper 20 with a DTT dithiepane junction that misses the charge for plasma membrane targeting (Figure 4d, top).Plasma membrane labeling with the intermediate anionic dithiolane 18 (Figure 4e, top) and dithiane 19 (Figure 4b) was good under isoosmotic conditions.However, under hyperosmotic conditions, all thioacetal flippers were internalized rapidly (Figures 4e, bottom, and S29), demonstrating that there is much room for improvement for membrane targeting, which should improve fluorescent lifetimes at the same time.
To image changes in membrane tension, large differences in lifetime under iso-and hyperosmotic conditions in HK cells are most important.Together with differences in intensity, this Δτ TR defines the responsiveness to changes in membrane tension.Under the present conditions, the original Flipper-TR 4 had Δτ TR = −0.5 ns (Figure 4a and Table 2, entry 1).With acetal 16, this did not change, but thioacetals 17 and 18 had higher sensitivity up to Δτ TR = −0.7 ns, while the poor interfacing of dithiepane 20 was not only reflected by internalization and lowest τ iso = 4.3 ns but also in a drop of sensitivity down to less relevant Δτ TR = −0.2ns (Figure 4a and Table 2, entries 2−4, 6).Within the structurally comparable dithiolane 18 and dithiane 19, stronger conventional chalcogen bonding (12, 13, Figure 2c) increased the sensitivity of tension imaging from Δτ TR = −0.5 to Δτ TR = −0.7 ns (Figure 4a and Table 2, entries 4, 5).
New records in sensitivity despite unoptimized interfacing implied much room to improve on thioacetal flippers.Dithiane 19 from asparagusic acid was selected as the starting point, also to assess the potential of the unusual bifurcated chalcogen  bonds (Figure 2f).Dithiane flipper 22 obtained from reduced lipoic acid did not improve Δτ TR = −0.5 ns, presumably because the longer linker is too hydrophobic to match the membrane interface well (Figure 4a and Table 2, entry 7).The specifically designed dithiane flippers 23 and particularly 6 with negative and positive charges placed at a distance as in the original Flipper-TR 4 and with better matching linkers increased sensitivity to tension changes up to Δτ TR = −0.8,counted exclusively for the plasma membrane (Figure 4a and Table 2, entries 8, 9).These maximized Δτ TR sensitivity coincided with nearly suppressed internalization also under hyperosmotic conditions (Figure 4f vs 4e, bottom).
Determination of tension sensitivity Δτ TR from alternative τ av gave even larger differences (Figure 4g,h).Compared to the original Flipper-TR 4 at Δτ av = −0.55 ± 0.04 ns, already anionic dithiane flipper 23 increased to Δτ av = −0.87 ± 0.03 ns.With the cationic dithiane flipper 6, the sensitivity of membrane tension imaging in living cells doubled beyond 1 ns, i.e., Δτ av = −1.11± 0.11 ns (Figure 4h).The significant increases in responsiveness from 19 over 23 to 6 were obtained with the same mechanophore.Therefore, they originated from interactions with the surrounding membranes.An ongoing systematic study suggests that in deconvoluted vibrational fine structures of excitation spectra, an intensity ratio of the second, formally 0−1 transition divided by the third transition of I 1/2 > 1 is indicative of excellent matching and partitioning into ordered domains (unpublished).In agreement with this emerging understanding, I 1/2 values did indeed increase significantly with increasing responsiveness from dithianes 19 to 23 and 6 (Figure S20; to some extent, also S o /L d intensity ratios, Figure S19).
Increased sensitivity to image membrane tension with dithianes was particularly impressive because this is perhaps the most intriguing but presumably not the most promising chalcogen-bonding donor junction, and only three analogues were tested to improve (Figure 4a).To illustrate the vast chemical space accessible, ETPs 66,68,69 were considered as clickable donor junctions.This natural-product-derived motif yields disulfides at maximal tension that are of interest to penetrate cells. 66,68,69The parent ETP was reduced prior to in situ thioacetalization with aldehyde 7 (Scheme S9).The resulting thioacetal junction 24 features a seven-membered ring like the dysfunctional dithiepane 20 but performed much better, more like dithiolane 17, including high Δτ TR = −0.6 at similarly impressive τ Lo , τ Lo m , τ iso , and Δτ GUV (Figure 4a and Table 2).

■ CONCLUSIONS
This study introduces a new strategy to facilitate synthetic access to fluorescent flipper probes.The solution of longstanding problems is found by moving one oxidation level higher, from ethers to thio/acetals.Clickable chalcogenbonding donor junctions enable late-stage modifications to screen for noncovalent donors and targeting units in one step, at best in situ.Computational exploration of the opened structural space is attractive to dissect different modes of coupled chalcogen bonding including intriguing bifurcated chalcogen bonds.All of the explored acetal and thioacetal junctions provided operational tension probes.Fluorescence properties in cells are overall dominated by interfacing with the cellular environment rather than the nature of the chalcogenbonding donor.Already introductory examples for late-stage screening of donor junctions suffice to provide fluorescent probes that double the sensitivity to changes in membrane tension in living cells.
Based on these results, in situ thioacetalization will be of practical use to easily access intracellular targeting in the broadest sense.However, the most promising is the disclosed access to a rich structural space on a new oxidation level.Latestage clicking is not limited to thio/acetals, which promises access to intriguing chalcogen-bonding donor motifs and more (Figure 1e).While nitrogen or selenium may be less attractive in this context (quenching, acidity, etc.), perspectives with sulfur and oxygen beyond ring contraction and expansion, interfacing, and targeting include catechols and thiocatechols, for instance, or the integration into larger systems, from oligosaccharide to peptide chemistry, including α-helix stapling.These perspectives are valid and inspiring for fluorescent probes in general.

Thioacetal Flippers Made In Situ
Flippers 17, 18, 19, and 20 were prepared in situ by adding HCl (24 μL of 4 M in dioxane, 96 μmol) and a solution of the respective dithiol/thiol in DMF (6 μL of 1 M, 6 μmol, N-acetyl-L-Cys 21, 2,3dimercaptopropanoic acid, reduced asparagusic acid, DTT) to a solution of 7 (300 μL of 0.2 mM, 0.06 μmol) in CH 2 Cl 2 at 25 °C (Scheme S8).Absorption spectra (l = 0.1 cm) of the reaction mixture were recorded every 1−5 min, and the corresponding mixture without 7 was used as the background (Figures 3d and S9).Based on the time-dependent absorption spectra, the kinetics of conversion from 7 to 17, 18, 19, and 20 were determined (Figure S10).The complete consumption of 7 and the generation of the desired products were confirmed by LC-MS of the reaction mixtures (Figures S11−S14).For direct use in FLIM imaging of GUVs and living cells, the obtained product mixtures were extracted with brine and CH 2 Cl 2 , and the organic phase was dried over Na 2 SO 4 and filtered through a cotton plug in a pipet.The filtrate was concentrated, and the residue was dissolved in DMSO (0.5 mL for 17, 19, and 20 and 1.0 mL for 18).
For flipper 24, solutions of the parent ETP 69 (1 M, 30 μL) and tris(hydroxypropyl)phosphine (THPP, 1 M, 60 μL) in DMF were combined and stirred for 10 min (Scheme S9).The resulting solution of reduced ETP and HCl (4 M, 120 μL in dioxane) was added to a solution of 7 (1.0 mM, 300 μL) in CH 2 Cl 2 .LC-MS confirmed the formation of 24 in situ (Figure S15) and showed the presence of residual 7 (Figure S16).For direct use in the FLIM imaging of GUVs and living cells, the reaction mixture was stirred for 48 h.Then, the product mixture was extracted with brine and CH 2 Cl 2 , dried over Na 2 SO 4 , filtered through a cotton plug in a pipet, concentrated in vacuo, and dissolved in DMSO (0.5 mL).

FLIM Imaging of Thioacetal Flippers Made In Situ
For FLIM imaging of GUVs, 10 μL of stock solutions of GUVs and 0.2 to 0.4 μL of stock solutions of flippers in DMSO (0.6 mM 17, 19, 20, 0.3 mM 18, 0.5 mM 24) were added to 190 μL of Tris buffer (10 mM Tris/Tris•HCl, 100 mM NaCl, pH 7.4).The obtained suspensions were placed on a 35 mm glass bottom dish (Mattek Corporation, P35G-1.5-14-C) and left for 15 min at room temperature before imaging (Figure S25).Leica Application Suite Software LASX FLIM 4.5.0Stellaris or SymPhoTime 64 software from PicoQuant was used to analyze the FLIM images.The lifetimes τ 1 were calculated from a biexponential fit of the signal coming from GUVs (selected as the ROI by "painting").
For FLIM measurements in HeLa Kyoto cells, the cells (8 × 10 4 cells mL −1 ) were seeded in FluoroBrite DMEM (high D-glucose, without phenol red) medium containing 10% fetal bovine serum (FBS), 1% penicillin/streptomycin (PS), and 1% glutamine and kept at 37 °C at 5% CO 2 overnight.Then, the cells were washed (3 × 1 mL) with PBS buffer and incubated with DMEM medium containing the corresponding probe (0.6 μM for 18, 1 μM for other probes, 1 mL) for 10 min at 37 °C at 5% CO 2 .The images were acquired without exchanging the incubation medium or additional washing.The hypertonic shock was achieved by adding 1 mL of 1 M sucrose medium containing the corresponding probe (0.6 μM for 18, 1 μM for other probes) in the dish containing 1 mL of isotonic medium for 30 min (Figures 4e and S29).FLIM data were analyzed by using SymPhoTime 64 software (PicoQuant) that fit fluorescent decay data (at least 6 cells per picture, 4 cells for 23) to a biexponential deconvolution model from the plasma membrane only.

Table 1 .
Computational Data for the Flipper Probes entry Cps a E rel (kcal mol −1 ) b μ (D) c d (Å) d E ET (kcal mol −1 ) e a Compounds, see Figure2.b Relative energy cost for planarization from conformer I (90°torsion angle) to conformer II (180°, Figure1a).c Macrodipole of planar conformer II.d Chalcogen-bonding distance S−X (X = O, S) in planar conformer II.e NBO electrontransfer component of chalcogen bond in conformer II.
a Compounds, see Figures 1 and 4. b d Fluorescence lifetime τ 1 in L o domains of mixed L o + L d DOPC/SM/CL 58:25:17 GUVs.e Fluorescence lifetime τ 1 in plasma membrane of isoosmotic HK cells.f Difference in fluorescence lifetime τ 1 between isoosmotic and hyperosmotic HK cells.