David Alonso Doval and Stefan Matile*
Department of Organic Chemistry, University of Geneva, Geneva, Switzerland. E-mail: stefan.matile@unige.ch; Web: http://www.unige.ch/sciences/chiorg/matile/ Fax: +41 22 379 5123; Tel: +41 22 379 6523
First published on 12th September 2013
A series of systematically deplanarized push–pull oligothiophenes is designed and synthesized to determine the perfect twist for maximal spectroscopic response to their planarization within lipid bilayer membranes. Weak deplanarization naturally gives weak shifts, but strong deplanarization also gives weak shifts because planarization becomes impossible. Intermediate deplanarization turns out to be ideal. The shifts found in response to chromophore planarization are not as dramatic as with lobsters during cooking but sufficient to discriminate solid-ordered and liquid-disordered membranes with the naked eye.
![]() | ||
Fig. 1 Systematic deplanarization of the original [111]-oligothiophene amphiphile 1 leads to [201]-probe 2 and [221]-3. Numbers in brackets indicate the number of methyl groups on β,β′-positions of each thiophene–thiophene bond. |
The new oligothiophenes were synthesized in strict analogy to the previously reported probes.1,6 In brief, the carboxylic acid 4 was subjected to esterification under microwave conditions, and the resulting methyl ester 5 was regioselectively iodinated in the α-position with N-iodosuccinimide (NIS) (Scheme 1). Suzuki–Miyaura cross-coupling between the iodinated thiophene 6 with boronate ester 7 yielded the twisted bithiophene 8. Iodination with NIS gave the halide compound 9 for Suzuki–Miyaura cross-coupling with the regioisomeric boronic ester 10. The obtained terthiophene 11 was regioselectively iodinated, and the resulting product 126 was subjected to Suzuki–Miyaura cross-coupling with the donor terminus 13. The ester group of the quaterthiophene 14 was reduced to the corresponding primary alcohol, which was then oxidized under mild conditions with Dess–Martin periodinane (DMP) to afford aldehyde 15. The strong cyano acceptor terminus was installed by Knoevenagel condensation with the cyanoamide 16, which was readily accessible from ester 17 and amine 18. Deprotection of acetal 19 followed by covalent capture of the guanidinium moiety 20 gave amphiphile 2. The preparation of amphiphiles in situ by oxime or hydrazone formation has been introduced recently to secure synthetic access to membrane-based differential odorant sensors, biosensors, fluorescent membrane probes and siRNA transfection agents.1,7 This dynamic covalent approach is important because the resulting amphiphiles are often problematic to purify and characterize due to self-assembly into micelles and reversed micelles in most solvents at already relatively low concentrations. Nevertheless, oxime 2 has been purified and characterized by RP-HPLC, 1H and 13C NMR spectroscopy, and ESI-MS (Fig. S4†). The same synthetic route was followed to prepare the fully twisted probe 3 (Fig. S5†).
![]() | ||
Scheme 1 Reagents and conditions: (a) MeOH, H2SO4, μW, 100 °C, 15 min, 71%; (b) NIS, CHCl3, AcOH, rt, 16 h, 73%; (c) Pd(PPh3)4, CsF, DMF, 80 °C, 16 h, 77%; (d) NIS, CHCl3, AcOH, rt, 16 h, 51%; (e) Pd(PPh3)4, CsF, DMF, 80 °C, 16 h, 74%; (f) NIS, CHCl3, AcOH, rt, 16 h, quantitative; (g) Pd(PPh3)4, CsF, DMF, 80 °C, 16 h, 71%; (h) 1. DIBAL, CH2Cl2, −78 °C, 2. DMP, CH2Cl2, rt, 20 min, 57%; (i) neat, rt, 48 h, 92%; (j) piperidine, MeCN, 70 °C, 3 h, 82%; (k) 1. TsOH·H2O, CH2Cl2, rt, 10 min, 55%, 2. DMSO, 60 °C, 10 min, 80%. |
In the excitation spectra in DMF, the intermediately twisted [201]-isomer 2 showed a maximum at λex ∼ 390 nm (Fig. 2, black, solid; Table 1). This was Δλex = −71 nm blue-shifted from the weakly twisted [111]-probe 1 at λex = 461 nm (Table 1). The strongly twisted [221]-oligothiophene 3 shifted further to λex = 371 nm (Δλex = −90 nm, Fig. S1,† black, solid; Table 1). Increasing deplanarization in the ground state was thus correctly reflected in the increasing blue shifts of the excitation maxima.
![]() | ||
Fig. 2 Excitation (solid) and emission spectra (dashed) of 2 in DMF (black) and DPPC LUVs at 55 °C (gold) and 25 °C (blue), with an indication of excitation (a–c and e–g) and emission maxima (d) of 1 (a–d) and 3 (e–g) in DPPC LUVs at 55 °C (b, d and f) and 25 °C (c,d and g), and in DMF (a and e). |
Cpdb | λex![]() | Δλex![]() | λem![]() | Δλem![]() |
---|---|---|---|---|
a Compare Fig. 2, S1, measured at 25 °C.b Compounds, see Fig. 1.c Excitation maximum, in wavelength (in nanometers).d Shift of the excitation maximum compared to 1.e Emission maximum.f Shift of the emission maximum compared to 1.g Data from ref. 1. | ||||
1g | 461 | — | 736 | — |
2 | 390 | −71 | 569 | −167 |
3 | 371 | −90 | 572 | −164 |
In comparison to the original maximum of [111]-probe 1, the emission maxima of both [201]-probe 2 and [221]-probe 3 were blue shifted by Δλem ∼ −165 nm to λem ∼ 570 nm (Fig. 2, black, dotted, Table 1). This result demonstrated that strongly deplanarized push–pull probes remain twisted also in the excited state. This finding was very important because it suggested that the planarization of the new probes should be detectable not only in the excitation but also in the emission spectra.
The planarizability of push–pull probes 2 and 3 was tested in DPPC LUVs (large unilamellar vesicles composed of dipalmitoyl phosphatidylcholine). At room temperature, DPPC membranes are in solid-ordered (So) phase. At 41 °C, they undergo a phase transition to liquid-disordered (Ld) phase. Consistent with their planarization by a confining environment, the excitation spectrum of the weakly twisted [111]-probe 1 showed a red shift of +20 nm upon cooling from Ld to So DPPC LUVs (Fig. 2b and c; Table 2).1 The emission maximum at 650 nm was insensitive to membrane fluidity because in the excited state, the fluorophore is already planar (Fig. 2d, Table 2).1
Cpdb | λex![]() | λex![]() | Δλex![]() | λem![]() | λem![]() | Δλem![]() |
---|---|---|---|---|---|---|
a Compare Fig. 2, S1, DPPC = dipalmitoyl phosphatidylcholine, LUVs = large unilamellar vesicles.b Compounds, see Fig. 1.c Excitation maximum at 55 °C, in wavelength (in nanometers).d Excitation maximum at 25 °C.e Shift of excitation maximum upon cooling from 55 °C to 25 °C.f Emission maximum at 55 °C.g Emission maximum at 25 °C.h Shift of the emission maximum upon cooling from 55 °C to 25 °C.i Data from ref. 1. | ||||||
1i | 467 | 487 | +20 | 650 | 650 | 0 |
2 | 449 | 483 | +34 | 562 | 600 | +38 |
3 | 413 | 423 | +10 | 590 | 582 | −8 |
The excitation maximum of the intermediately twisted [201]-probe 2 in Ld DPPC LUVs was at λex = 449 nm (Fig. 2, gold, solid; Table 2). The blue shift Δλex = −18 nm compared to the [111]-probe 1 demonstrated that [201]-probe 2 is more twisted in Ld phase. In So DPPC LUVs at 25 °C, the excitation maxima of both probes were at λex ∼ 485 nm. This suggested that the ground state of both probes is fully planarized in this highly confining solid-ordered environment. More ground-state twisting in Ld phase with identical planarization in So phase naturally meant that the [201]-probe 2 is more sensitive to Ld–So transition than the [111]-probe 1. The bathochromic shift in response to pressure from the surrounding membrane increased from Δλex = +20 nm for 1 to Δλex = +34 nm for 2 (Table 2). Considering the experimental inaccuracy, it can be said that with optimized deplanarization, the sensitivity of the fluorescent membrane probes almost doubled (compare below for DOPC).
The emission maximum of probe 2 in Ld DPPC LUVs was at λem = 562 nm (Fig. 2, gold, dashed; Table 2). The strong hypsochromic shift of Δλem = −88 nm compared to [111]-probe 1 suggested that the excited state of [201]-probe 2 remains partially twisted in Ld DPPC LUVs. Unlike the phase-insensitive emission of probe 1, a bathochromic shift of Δλem = +38 nm was observed with 2 in response to Ld–So transition of DPPC LUVs. These results suggested that also the excited state of [201]-probe 2 is deplanarized in Ld DPPC LUVs and provided corroborative evidence that the Ld–So transition can be sensed not only by ground-state but also by excited-state planarization.
The strongly twisted, weakly fluorescent [221]-probe 3 showed little difference in excitation and emission spectra in response to changes in membrane structure. The maximum in the excitation spectrum remained strongly blue-shifted compared to the maxima of probes 1 and 2, suggesting that the ground state of the probe 3 is too twisted to be planarized in either Ld or So DPPC membranes (Fig. 2f and g, Table 2, Fig. S1†). The same is true for the excited state (Table 2, Fig. S1†).
Control experiments were performed in dioleoyl phosphatidylcholine (DOPC) LUVs. These membranes remain in Ld phase at 25 °C and at 55 °C. In DOPC LUVs at 55 °C, the fluorescence of the original [111]-probe 1 was almost the same as that in organic solvents or Ld DPPC LUVs and showed very little red shift upon cooling (λex = 466 nm, Δλex = +2 nm, Table 3).1 The excitation maximum of the hypersensitive [201]-probe 2 was at λex = 436 nm under identical conditions, that is, Δλex = −30 nm compared to 1 and Δλex = −13 nm compared to 2 in Ld DPPC LUVs (Tables 2 and 3, Fig. S2†). These overall small shifts suggested that the ground state of 2 remains partially twisted also in “hot” DOPC LUVs.
Cpdb | λex![]() | λex![]() | Δλex![]() | λem![]() | λem![]() | Δλem![]() |
---|---|---|---|---|---|---|
a Compare Fig. S2 and S3, DOPC = dioleoyl phosphatidylcholine, LUVs = large unilamellar vesicles.b Compounds, see Fig. 1.c Excitation maximum at 55 °C, in wavelength (in nanometers).d Excitation maximum at 25 °C.e Shift of excitation maximum upon cooling from 55 °C to 25 °C.f Emission maximum at 55 °C.g Emission maximum at 25 °C.h Shift of the emission maximum upon cooling from 55 °C to 25 °C.i Data from ref. 1. | ||||||
1i | 466 | 468 | +2 | 650 | 650 | 0 |
2 | 436 | 446 | +10 | 557 | 557 | 0 |
3 | 367 | 367 | 0 | 607 | 595 | −12 |
The Δλex = +10 nm found upon cooling placed the excitation of 2 in “cold” DOPC LUVs together with that in Ld DPPC LUVs. This small shift could originate from partial ground-state planarization, although thermochromic effects cannot be excluded. In any case, the overall hypsochromic position and poor sensitivity of the excitation spectrum of 2 in DOPC LUVs provided excellent general support that the large Δλex = +34 nm found for Ld–So transition in DPPC LUVs originates indeed from planarization of 2 in the ground state. Comparison of DOPC-LUV and DPPC-LUV samples at 25 °C demonstrated that the Δλex = +37 nm for planarization of [201]-probe 2 in So lipid bilayer membranes is sufficient to be seen with the “naked eye” (compare TOC graphic). This colorimetric detection of the nature of lipid bilayer membranes with the naked eye was not observed with the poorly twisted original [111]-probe 11 and is thus unique for the almost perfect twist in the new [201]-probe 2.
The sensitivity of the best probe 2 to liquid-ordered (Lo) phase was explored in the biologically relevant SM–CL LUVs (SM: sphingomyelin, CL: cholesterol; SM–CL = 3:
1). Excitation was found at λex = 488 nm, which is slightly red-shifted from that in So DPPC LUVs (Δλex = +5 nm, Fig. 3, blue, solid). The same trend was observed for emission, which maximized at λem = 619 nm (Δλem = +19 nm, Fig. 3, blue, dotted). This suggested that further planarization of probe 2 might occur in Lo phase in both ground and excited states compared to that in So phase, although the effect is rather small.
![]() | ||
Fig. 3 Excitation (solid) and emission spectra (dashed) of 2 in DOPC (gold), SM–CL (blue), and DOPC–SM–CL LUVs (grey) at 25 °C (blue). Maxima are indicated in nanometers. |
Fluorescent properties of probe 2 were also measured in DOPC–SM–CL LUVs under conditions where Lo and Ld phases coexist in the same vesicle (DOPC–SM–CL = 1:
1
:
1). The excitation maxima moved to λex = 466 nm, which is nearly halfway between excitation in pure Lo and Ld phases (Fig. 3, grey, solid). This finding suggested that probe 2 partitions into both phases and can, in principle, report their coexistence in heterogeneous vesicles simultaneously. The emission in DOPC–SM–CL LUVs was roughly the same as in Lo phase (Fig. 3, grey, dashed). This difference in the behavior of the excitation spectra could possibly be explained by fluorescence resonance energy transfer from the twisted fluorophore in Ld to the flattened fluorophore in Lo phase.
Footnote |
† Electronic supplementary information (ESI) available: Detailed procedures and results for all reported experiments. See DOI: 10.1039/c3ob41706a |
This journal is © The Royal Society of Chemistry 2013 |