Bioconjugate Chem. 2006, 17, 670−676
670
A Blue-Green Absorbing Cross-Linker for Rapid Photoswitching of Peptide Helix Content Lei Chi, Oleg Sadovski, and G. Andrew Woolley* Department of Chemistry, University of Toronto, 80 Saint George Street, Toronto, ON M5S 3H6, Canada . Received December 23, 2005; Revised Manuscript Received March 15, 2006
Azobenzene derivatives can be used to reversibly photoregulate secondary structure when introduced as intramolecular bridges in peptides and proteins. Here we report the design, synthesis, and characterization of a disubstituted N,N-dialkyl azobenzene derivative that absorbs near 480 nm in aqueous solution and relaxes with a half-life of ∼50 ms at room temperature. The wavelength of maximum absorbance and the rate of thermal relaxation are solvent-dependent. An increase in the percentage of organic solvent leads, in general, to a blue shift in the absorbance maximum and a slowing of the relaxation rate. In accordance with the design, the thermal relaxation of the azobenzene cross-linker from cis to trans causes an increase in the helix content of one peptide where the linker is attached via cysteine residues spaced at i, i + 11 positions and a decrease in helix content of another peptide with cysteine residues spaced at i, i + 7. This cross-linker design thus expands the possibilities for fast photocontrol of peptide and protein structure.
INTRODUCTION Reversible optical control of protein structure and function offers the possibility of probing and manipulating individual proteins within the complex environment of a living cell (1, 2). Azobenzene, in particular, has been a popular choice as a photoswitching chromophore due to its robust photophysical properties and relative ease of synthesis (3, 4). However, there is no wavelength at which only one azobenzene isomer absorbs; the absorption spectra of trans and cis forms overlap extensively. Thus, irradiation typically produces photostationary states composed of at most ∼80% cis or ∼95% trans isomers (5, 6). This fact limits the degree of photoswitching that is possible; For example, the fraction of cis isomer can be changed from ∼5% to 80% (16-fold) or trans from ∼20% to 95% (∼5-fold) (6). Thermal isomerization, in contrast, yields >99.99% trans isomer (7). Therefore, if thermal isomerization is used to reset the switch, a much greater fold change in the cis isomer is then possible. The half-life for thermal cis-to-trans isomerization of unmodified azobenzene is on the order of 3-4 days at room temperature (6). If one wished to probe a dynamic process in a living cell using thermal reversion to reset the switch, this half-life would likely be too long. Many cellular signaling processes, such as the actions of kinases and phosphatases, operate on the time frame of seconds to minutes (e.g., refs 8 and 9). Indeed naturally occurring photoswitchable proteins reset in the dark in seconds or less (10, 11). Several groups have shown that the rate of thermal isomerization of azobenzenes can be varied substantially by altering the nature of substituents on the aromatic rings (e.g., refs 1215). Previously we reported the design and synthesis of thiolreactive azobenzene derivatives bearing alkyl, amide, carbamate, and urea substituents that could be used to manipulate peptide secondary structure (7). The fastest of these, the urea-substituted compounds, relaxed with a half-life of ∼10 s at room temperature. While significantly faster than the parent compounds, these photoswitches may still be too slow for certain applications * E-mail:
[email protected]. Telephone/fax: (416) 9780675.
such as probing calcium signaling in myocytes or reversibly modulating neural activity with switchable ion channels (9, 16). In addition to a requirement for rapid thermal relaxation, a critical design feature of such photoswitches is the effective end-to-end distance and how this changes upon isomerization. The urea- and carbamate-substituted compounds, in particular, can accommodate a wide range of end-to-end distances in their cis states, a feature that may limit their effect on the conformation of an attached peptide or protein (17). A smaller number of bonds between the azo moiety and the peptide backbone would be likely to maximize the conformational switching effect. Finally, the wavelength that causes photoisomerization can be of considerable practical importance if one wishes to achieve photocontrol in a biological context. Short wavelength UV light is damaging to proteins and nucleic acids (18). Wavelengths in the 330-370 nm range (where current azobenzene-based photoswitches operate), while compatible with biological systems (1, 2), are less able to penetrate cells and tissues than longer wavelengths (19). Longer wavelength switches would also make possible the incorporation of separately addressable photoswitches into a single protein. With these considerations in mind, we designed the thiol reactive dialkylamino-substituted azobenzene-based photoswitch 1. We report here the synthesis and characterization of this new photoswitch.
EXPERIMENTAL PROCEDURES All chemicals were purchased from Sigma-Aldrich Chemical Co. except if specified otherwise. All 1H and 13C NMR spectra were recorded using a Varian Unity 400 or a Varian Gemini 300 spectrometer. High-resolution mass spectra were obtained either by electron impact (EI) or electrospray (ESI) ionization. Peptide mass spectra were obtained either by ESI or matrixassisted laser desorption (MALDI) ionization. Synthesis of the Cross-Linker. 2-[Ethyl-(4-{4-[ethyl-(2hydroxy-ethyl)-amino]-phenylazo}-phenyl)-amino]-ethanol (3). Ten grams (36 mmol) of 4-(ethyl(2-hydroxyethyl)amino)benzenaminium hydrogensulfate (2) was mixed with 100 mL of ethyl acetate, and a saturated solution of sodium bicarbonate was added (∼15 mL). The reaction mixture turned a red color,
10.1021/bc050363u CCC: $33.50 © 2006 American Chemical Society Published on Web 04/18/2006
Blue-Green Absorbing Cross-Linker
and after the evolution of carbon dioxide was complete, the organic layer was separated. The water layer was extracted with ethyl acetate (3 × 50 mL), and the combined organic layers were dried with Na2SO4. The water layer was then mixed with ethanol (∼ 5 mL) and filtered through a pad of Na2SO4 and Al2O3. The pad was washed with CH2Cl2 (∼100 mL). The CH2Cl2 solution was combined with the ethyl acetate fractions, and the combined solution was evaporated, then dissolved again in CH2Cl2 and dried with Na2SO4. Removal of CH2Cl2 gave 4 g of the free amine of 2 (62% yield). A mixture of 2-((4-aminophenyl)(ethyl)amino)ethanol (2; 4 g, 22 mmol) and manganese dioxide (20 g) was refluxed in dry benzene (200 mL) for 4 h. Removal of residual manganese dioxide and the solvent gave a residue, which was chromatographed on alumina. Elution with benzene gave 2.3 g of 3 (29% yield) as a slurry. 1H NMR (300 MHz, CDCl3) δ ppm: 1.20 (t, J ) 7.3 Hz, 6 H, 2CH3), 3.40-3.60 (m, 8 H, 4 CH2), 3.90 (t, J ) 5.9 Hz, 4 H, 2CH2), 6.80 (d, J ) 9.1 Hz, 4 H, 4CHarom), 7.8 (d, J ) 9.1 Hz, 3 H, 4CHarom). MS (ESI): calcd (MH+) C20H29N4O2 357.47; obsd 357.2. N-(2-Bromoethyl)-4-((4-((2-bromoethyl)(ethyl)amino)phenyl)diazenyl)-N-ethylbenzenamine (4). Methanesulfonyl chloride (0.65 mL, 8.4 mmol) was added in small portions to a stirred solution of 3 (1 g, 2.8 mmol) and TEA (1.56 mL, 11 mmol) in dry THF (50 mL) under N2 maintained at 0 °C. After being stirred for 1 h at 0 °C and 24 h at room temperature, the mixture was poured into cold water and extracted with CH2Cl2. The organic extract was washed with cold water, dried, and concentrated under reduced pressure. LiBr (1.2 g, 14 mmol) was added over the course of 15 min to a solution of the residue so obtained in acetone (100 mL), and the mixture was stirred for 48 h at room temperature. It was then poured into water (400 mL) and extracted with CH2Cl2. The organic extract was washed with water, dried, and purified by silica gel chromatography (CH2Cl2/hexane, 1:1) to yield 4 (0.9 g, 67%). 1H NMR (300 MHz, CDCl3) δ ppm: 1.20 (t, J ) 7.0 Hz, 6 H, 2CH3), 3.40-3.60 (m, 8 H, 4CH2), 3.70-3.80 (m, 4 H, 2CH2), 6.7 (d, J ) 9.3 Hz, 4 H), 7.8 (d, J ) 9.3 Hz, 4 H). 13C NMR (300 MHz, CDCl3) δ ppm: 12.72, 28.24, 45.74, 52.43, 111.39, 124.39, 144.33, 148.15. HRMS (ESI): calcd (MH+) C20H27N4Br2 481.0596; obsd 481.0608. Methanethiosulfonic Acid S-{2-[Ethyl-(4-{4-[ethyl-(2-methanesulfonylsulfanyl-ethyl)-amino]-phenylazo}-phenyl)-amino]ethyl} Ester (1). Compound 4 (0.1 g, 0.2 mmol) and sodium methanethiosulfonate (80 mg, 0.6 mmol) were dissolved in dry DMF (10 mL) under Ar and heated to 60 °C. After 18 h of stirring at this temperature, the DMF was evaporated under reduced pressure. The crude product was dissolved in EtOAc and filtered, and the filtrate was evaporated under reduced pressure. Chromatography on silica (EtOAc/hexanes 1:1, Rf ) 0.36) gave 1 (30 mg, 33%), as a brown-yellow solid. 1H NMR (400 MHz, CDCl3) δ ppm: 1.25 (t, J ) 7 Hz, 6.H, 2CH3), 3.30-3.45 (m, 10 H, 2CH2, 2CH3), 3.5 (q, J ) 7.3 Hz, 4 H, 2 CH2), 3.80 (t, J ) 7.Hz, 4 H 2CH2), 6.8 (d, J ) 9 Hz, 4 H), 7.8 (d, J ) 9 Hz, 4 H). 13C NMR (300 MHz, CDCl3) δ ppm: 12.53, 33.39, 45.78, 50.55, 50.72, 111.78, 124.45, 144.45, 148.20. HRMS (ESI): calcd (MH+) C22H33N4S4O4 545.1379; obsd 545.1390. Peptide Cross-Linking. Peptides FK-11 (acetyl-EACAREAAAREAACRQ-amide) and JRK-7 (acetyl-EACARVAibAACEAAARQ-amide) were prepared using standard Fmoc-based solid-phase synthetic methods as described previously (20). Intramolecular cross-linking of cysteine residues was performed for FK-11 as follows: 2.0 mg (1.2 µmol) of FK-11 was dissolved in a total volume of 3.8 mL of a solvent consisting of 1.0 mL of 10 mM Tris‚Cl, 1 mM EDTA buffer (pH 8.0) and 2.8 mL of DMSO. This solution was stirred, and 200 µL of a
Bioconjugate Chem., Vol. 17, No. 3, 2006 671
14.5 mM solution of the linker 1 in DMSO was added dropwise. Then the mixture was left to stir for 2 h protected from light. The modified peptide was purified by Bio-gel P4 gel-filtration chromatography, followed by HPLC [Zorbax 300SB-C3 column; 20-80% acetonitrile/H2O (+0.1% trifluoroacetic acid) linear gradient over the course of 30 min; elution at 54% acetonitrile]. The peptide molecular composition was confirmed by ESI-MS (calculated for C8513C2H139N31O24S4 ) 2133.5 Da; observed ) 2133.4 Da). For JRK-7, 0.5 mg (0.3 µmol) of peptide was dissolved in a total volume of 1.8 mL of a solvent consisting of 0.5 mL of 10 mM Tris‚Cl, 1 mM EDTA buffer (pH 8.0) and 1.3 mL of THF. This solution was stirred, and 200 µL of a 3.65 mM solution of the linker 1 in THF was added dropwise. Then the mixture was left to stir for 2 h protected from light. The modified peptide was purified by HPLC [Zorbax 300SB-C3 column; 20-70% acetonitrile/H2O (+0.1% trifluoroacetic acid) linear gradient over the course of 25 min; elution at 65% acetonitrile]. The peptide molecular composition was confirmed by electrospray ionization MS (calculated for C85H136N28O22S4 ) 2030.5 Da; observed ) 2030.4 Da). The glutathione derivative of 1 was prepared as follows: 2.8 mg (9.1 µmol) of glutathione was dissolved in a total volume of 4.5 mL of a solvent consisting of 1.5 mL of 10 mM Tris‚Cl, 1 mM EDTA buffer (pH 8.0) and 3.0 mL of THF. This solution was stirred, and 200 µL of a 9.0 mM solution of the linker 1 in THF was added dropwise. Then the mixture was left to stir overnight protected from light. The modified peptide was purified by Bio-gel P4 gel-filtration chromatography, followed by HPLC [Zorbax 300SB-C18 column; 20-70% acetonitrile/H2O (+0.1% trifluoroacetic acid) linear gradient over the course of 30 min; elution at 48% acetonitrile]. The peptide molecular composition was confirmed by electrospray ionization MS (calculated for C40H58N10O12S4 ) 999.2 Da; observed ) 999.1 Da). UV/Vis Spectra and Photoisomerization. Ultraviolet absorbance spectra were obtained using either a Perkin-Elmer Lambda 2 spectrophotometer or a diode array UV-vis spectrophotometer (Ocean Optics Inc., USB2000) coupled to a temperature-controlled cuvette holder (Quantum Northwest, Inc.). The latter arrangement was used to determine thermal relaxation rates of cross-linked species. Irradiation of the sample (at 90° to the light source and detector used for the absorbance measurements) was carried out using a xenon lamp (450 W) coupled to a double monochromator with slits at 16 nm and 16 nm. Rates of thermal cis-to-trans isomerization were measured for a series of temperatures by monitoring absorbance at 480 nm after irradiation to convert a percentage of the solution to the cis isomer. All curves could be fit well by single-exponential decay kinetics. Solution conditions are described in the figure legends. For rapid (
