|Home | About | Journals | Submit | Contact Us | Français|
Three novel aqueous soluble fulgimides, trifluoromethyl carboxylic acid indolylfulgimide 4, dicarboxylic acid indolylfulgimide 5, and carboxylic acid indolylfulgimide 6, were synthesized. Both 4 and 5 can switch back and forth between open and closed forms upon illumination with specific wavelengths of light, while 6 can only switch from the closed form to the open form. In sodium phosphate buffer (pH 7.4) at 37 °C, an unusual hydrolysis of the trifluoromethyl group of the closed form of 4 resulted in 5 which has an additional carboxylic acid group. The closed form of 5 was further decarboxylated to generate 6 which was not photochromic. In buffer, the open form of 4 degraded 20% after 10 days while the closed form of 4 was converted to 5 rapidly. In buffer, both forms of 5 degraded less than 20% after 21 days at 37 °C, and 5 underwent 670 photochemical cycles before degrading by 20%. It is the most robust fulgimide yet reported in aqueous solution.
Photochromic compounds have potential applications in high capacity optical information storage devices, optical molecular switches, and biological sensors.1-3 All these applications depend on the binary nature of photochromic compounds. The interconversion between two key forms upon exposure to specific wavelengths of light is known as photochromism (Scheme 1).
Fulgides and fulgimides, promising photochromic compounds, have been considered as potential optical memory materials due to the readily distinguishable absorption spectrum for each key form, efficient photoreactions, and thermal and photochemical stabilities.1,4 Studies have been conducted to optimize the photochromic properties of fulgides for specific applications.5-7 Optimization has resulted in more thermally and photochemically stable compounds such as fluorinated indolylfulgide 1, which was originally synthesized by Yokoyama and Takahashi.8-10 The most photochemically stable fulgide, a fluorinated indolylfulgide synthesized by Lees et al., undergoes 10,000 photochemical cycles (back and forth conversion between the two key forms) before degrading by 13% in toluene.10 Optimization in almost all cases has been performed in aprotic solvents. However, the properties of fulgides or fulgimides in aqueous solutions have not been thoroughly examined.
Stability in protic environments is an important property of photochromic compounds for their application in optical memory devices and biological optical switches.1,2 Materials used for memory devices are required to maintain stability and function in humid environments. In many biological applications, optical switches need to function in aqueous solvent systems.2,11 Previous studies demonstrated that fulgides were unstable in protic solvents due to the highly reactive succinic anhydride ring in their structure.12,13 Fulgimides, the most important fulgide derivatives, were synthesized to improve stability by replacing the succinic anhydride ring with a succinimide ring.13,14 The closed form of N-phenyl fulgimide 2 displayed three orders of magnitude greater stability in 70/30 ethanol/water relative to the parent fulgide 1 at 25 °C (Scheme 1).13 One of our recent studies indicated that the open form of ethyl ester fulgimide 3 lost 22% of its absorbance at the absorbance maxima while 2 lost 52% after 21 days in 70/30 ethanol/water at 50 °C.14 Furthermore, ethyl ester fulgimide 3 underwent 360 photochemical cycles in 70/30 ethanol/water before degrading by 20% while 2 underwent 170 cycles.14
The photochemical stability of fulgides in ethanol/water was not reported due to their rapid decomposition. Although several studies have determined the photochemical properties of fulgimides in protic solvents,2,11-14 only a few of these studies have reported the properties of fulgimides in aqueous solution.2,11,15 In one particular study fulgimide derivatives were covalently attached to the lysine residues on concanavalin A, where the open form of the fulgimide was shown to be stable in aqueous solution for 48 h at 25 °C.11 This report also indicated that the fulgimide can cycle back and forth between the open and the closed forms at least twice. A recent study in live cells demonstrated that fulgimides can switch back and forth seven times in cellular membranes but not very well in water.2 Therefore, a more systematic study of the photochemical and thermal properties of fulgimides in aqueous solution would accelerate their applications as biological optical switches and sensors.
Herein, we have synthesized and characterized two new photochromic indolylfulgimides 4 and 5 (Scheme 1). Fulgimide 4 was prepared from fluorinated indolylfulgide 1. An unusual hydrolysis of 4 in sodium phosphate buffer resulted in fulgimide 5, while further decarboxylation of 5 yielded a non-photochromic fulgimide 6. Fulgimides 4 and 5 were water soluble at physiological pH because of the hydrophilicity of the carboxylate anion. The absorption spectra, and thermal and photochemical stabilities for 4 and 5 have also been analyzed.
Trifluoromethyl indolylfulgide 116 was used as the starting material for the synthesis of carboxylic acid indolylfulgimide 4 (Scheme 2). The anhydride ring of 1 was opened via addition of glycine methyl ester. The resulting methyl ester succinamic acid, one of the two possible regioisomers,17 was saponified to generate the corresponding carboxylic acid succinamic acid. Subsequent dehydration of the succinamic acid intermediate with acetyl chloride yielded carboxylic acid indolylfulgimide 4Z. Fulgimide 4C was obtained by irradiating 4Z with 405 nm light (Scheme 2).
During thermal stability measurements of 4C in 50 mM sodium phosphate buffer (pH 7.4) at 37 °C, an unexpected reaction was observed. The reaction was followed by 1H and 19F NMR spectroscopy. The NMR data indicated an unusually high reactivity for 4C. Previously reported C-forms of fluorinated indolylfulgimides have proven to be very stable under various conditions.13,14 In the case of 4C, before incubation at 37 °C, only one resonance at −58 ppm in the 19F NMR spectrum corresponding to the fluorines of the trifluoromethyl group was observed (Figure 1a). After 6 h at 37 °C, the resonance at −58 ppm disappeared, and a new fluorine signal appeared at −122 ppm, which is consistent with the chemical shift of the fluoride anion. The 1H NMR spectrum showed a downfield shift of all hydrogen and methyl resonances (Figure 1b). In order to provide further support for the structure of the resulting product, 13C NMR spectroscopy was performed. The quartet for the carbon of the trifluoromethyl group of 4C at 122 ppm disappeared, and a new singlet at 170 ppm appeared suggesting a carboxylic acid group. Furthermore, the resulting product still maintained photochromic properties and was stable in buffer at 37 °C for several days. Therefore, the most plausible mechanism was the hydrolysis of the trifluoromethyl group to form a carboxylic acid group.18-20 The reaction yielded the photochromic dicarboxylic acid indolylfulgimide 5C (Scheme 3).
Interestingly, when we initially attempted to isolate 5C by EtOAc extraction from an acidified aqueous solution, the organic layer did not contain 5C (Figure 1c). Instead, an extra hydrogen resonance appeared at 6.41 ppm in the 1H NMR spectrum, and the 13C NMR spectrum showed only 20 carbon resonances in comparison with the 21 carbon resonances for 4C and 5C, indicating another compound had been formed during the acidic extraction. The missing carbon resonance occurred in the carboxylic acid region. The lack of any 19F NMR resonance suggested that the fluoride anion was removed in the aqueous layer during extraction. Therefore, we propose that the carboxylic acid group generated from the hydrolysis of the trifluoromethyl group can be decarboxylated to form compound 6C (Scheme 4).21,22
Photochromic studies demonstrated that 6C can be converted to 6E (E-form due to IUPAC priority rules), but the reverse reaction was not observed (Scheme 4). Previously reported indolylfulgides substituted at the 3-position on the indole and having hydrogen at the bridging position were initially obtained in their E-form and also could not be converted to the C-form.23-25 For the first time, we obtained the C-form of such a fulgimide with a hydrogen at the bridging position. We did not investigate the optical properties of 6 further as it was not photochromic.
Syntheses of the dicarboxylic acid indolylfulgimide 5 and the carboxylic acid indolylfulgimide 6 were then carried out as described above. The reaction of 4C to 5C occurred quantitatively and rapidly in buffer (pH 7.4) at 37 °C. Fulgimide 5C was relatively stable in acidic aqueous solution, but CO2 was lost during prolonged extraction with EtOAc.
The UV-vis absorption spectra of 4Z and 4C were measured in toluene. The spectra of 4Z, 5Z, and 5C were obtained in 50 mM sodium phosphate buffer (pH 7.4) (Figure 2). No UV-vis measurements for 4C in buffer and 5 in toluene were performed due to the instability of 4C in buffer (see above) and the poor solubility of 5 in toluene. The wavelengths of maximum absorbance and the extinction coefficients are shown in Table 1. Fulgimide 4Z showed a small bathochromic shift (4 nm) as the solvent was switched from toluene to buffer. In comparison with 4Z, 5Z in buffer demonstrated a 16 nm hypsochromic shift at its absorbance maxima.
Thermal stability is one of the most important characteristics of fulgides and fulgimides for their applications in optical memory devices or optical switches.10,12,26 Previously, the thermal stability of fulgides and fulgimides was examined in toluene at 80 °C.3,13,14 Therefore, the thermal stability of 4Z and 4C was determined under these conditions. We also examined the stability of 4 and 5 at 37 °C in sodium phosphate buffer (pH 7.4) as this mimics physiological conditions to some extent. The thermal decomposition of 4 and 5 was followed by both 1H NMR and UV-vis spectroscopy. The results are presented in Table 2.
In pure toluene at 80 °C, the decomposition of 4Z was fit to a single exponential decay (Figure 3a). The decomposition rate constants were 0.023 and 0.010 h−1 by UV-vis and 1H NMR spectroscopy, respectively (Table 2). These values are similar to those observed for the parent fulgide 1Z (0.023 h−1) and ethyl ester fulgimide 3Z (0.009 h−1).3,14 The UV-vis spectra also showed a similar pattern, an initial drop in absorbance followed by a red shift and subsequent increase in absorbance.3,8,14 Our previous studies demonstrated that the thermal decomposition pathway for the Z-form of fluorinated indolylfulgides in toluene involves either a reversible Z-to-E-isomerization or the conversion of the Z-form to an intermediate via a 1,5-hydrogen shift from the isopropylidene group.8,10 The intermediate then subsequently rearranges to form a mixture of two isomers. Based on similar spectral kinetics, we postulate that fulgimide 4Z undergoes the same degradation pathway as fulgide 1Z. Therefore, the offset observed in the UV-Vis data in Figure 3a is due to the absorbance of the decomposition products at the λmax of 4Z.8,10 In the case of fulgimide 4C in toluene at 80 °C, a double exponential fit was applied since a relatively rapid decomposition of 3% was observed followed by a slow decomposition (Figure 3a). Fulgimide 4C showed much higher stability than 4Z in toluene, consistent with previously reported fulgides and fulgimides.9,13,14
In 50 mM sodium phosphate buffer (pH 7.4) at 37 °C, single and double exponential fits were applied to the C- and Z-forms, respectively. 4Z and 5Z showed a relatively rapid decline in concentration, which corresponded to Z-to-E-isomerization, followed by a slower decline which corresponds to decomposition (Figures 3b, 3c). According to 1H NMR data, 5Z decomposed three times slower than 4Z in buffer. 4C is unstable and completely converted to 5C in buffer after 3–6 h at 37 °C. 5C showed great thermal durability and very little decomposition was observed after prolonged time in buffer at 37 °C. To account for the difference between the UV-vis and NMR data for 5C in buffer, we measured the decomposition of 5C by UV-vis spectroscopy in both D2O and H2O buffers and determined a solvent isotope effect of 3–4.
The repeatability of the photochemical opening and closing of fulgimides 4 and 5 was measured in toluene and 50 mM sodium phosphate buffer (pH 7.4), respectively. Photochemical stability is required for many applications.26 In toluene, the most stable fulgide reported to date can be switched back and forth over 10,000 times before degrading by 13%10 although for most fulgides the number is less.3,9 In the case of fluorinated indolylfulgmides, a previous study indicated that they can be cycled back and forth between 700 and 3,000 times in toluene before degrading by 20%.13,14
In protic solvent systems, such as methanol, ethanol/water, or water, fulgides are too unstable and/or insoluble to measure their photochemical stability.27 On the other hand, fulgimides previously examined in protic solvents only cycled back and forth a limited number of times.2,12,13 A recent study in our group reported that the ethyl ester fulgimide 3 can be cycled back and forth 360 times before degrading by 20% in 70/30 ethanol/water.14 Several reports about applications of fulgimides in aqueous biological systems have demonstrated that fulgimides can be cycled back and forth several times.2,11
The photochemical stability of fulgimide 4 was initially measured in pure toluene where it degraded by 20% after being cycle back and forth 21 times (Figure 4a), much less stable than its ethyl ester 3.14 We speculate that the rapid photochemical decomposition was affected by the carboxylic acid group and that the addition of base would increase the stability. In the presence of tributylamine (27 mM) in toluene, 4 cycled back and forth 55 times before degrading by 20% (Figure 4a). The cycling times were approximately 35 s (Z- to C-form) and 20 s (C- to Z-form) in both cases, suggesting that the addition of tributylamine slowed down the photochemical decomposition but not the photochemical reaction. Addition of acetic acid (27 mM) did not affect photochemical stability. The photochemical stability of 4 was not measured in buffer due to the instability of 4C. Fulgimide 5 cycled back and forth 670 times before degrading by 20% in buffer (Figure 4) with cycling times of 80 s (Z- to C-form) and 600 s (C- to Z-form). The increased photochemical stability of 5 makes it promising for applications in aqueous solution.
In summary, we have synthesized three novel aqueous soluble indolylfulgimides, 4, 5, and 6. Hydrolysis of the trifluoromethyl group of 4C was observed in a fluorinated indolylfulgimide for the first time. Hydrolysis of 4C resulted in 5C which was further decarboxylated to 6C upon extraction. 6C lacked any photochromic properties. The absorbance maxima of 4Z varied only slightly between toluene and buffer. A notable blue shift in the absorbance maxima of 5Z compared to 4Z was observed in buffer due to the additional carboxylic acid group on the bridging carbon. Fulgimide 5 displayed great thermal and photochemical stabilities in sodium phosphate buffer (pH 7.4). 5Z and 5C degraded less than 20% after 500 h at 37 °C, and 5 underwent 670 photochemical cycles before degrading by 20%. Fulgimide 5 is the most robust fulgimide yet reported in aqueous solution.
All commercially available materials were used without further purification. 1H and 13C NMR spectra were internally referenced to TMS (0.00 ppm) or solvent (7.26 and 77.00 ppm for CDCl3; 3.31 and 49.00 ppm for CD3OD; 4.79 ppm for D2O).
N,N-Diisopropylethylamine (2.7 g, 20.8 mmol) was added slowly with stirring to a mixture of the HCl salt of glycine methyl ester (1.28 g, 10 mmol) and trifluoromethyl indolylfulgide 1Z (1.84 g, 5.2 mmol) in 100 mL of acetonitrile at 0 °C. After stirring overnight, the solvent was removed in vacuo. The residue was added to 0.5 M HCl (100 mL) and extracted with ethyl acetate (3 × 100 mL). The combined organic layers were dried over MgSO4, filtered, and concentrated in vacuo. Purification of the residue by silica gel chromatography (60:40:2 hexane/EtOAc/AcOH) provided 1.94 g of the crude amide acid ester (orange solid). NaOH (1.1 g, 26.5 mmol) was added to the crude amide acid ester in 250 mL of methanol, and the reaction mixture was stirred for 2 h at room temperature. The solvent was then removed in vacuo. The residue was added to 100 mL of Na2CO3 (0.19 M) and extracted with ethyl acetate (2 × 75 mL). The aqueous solution was acidified with 8.0 mL of concd HCl and extracted with ethyl acetate (3 × 75 mL). The combined organic layers were dried over MgSO4 and filtered. The solvent was concentrated in vacuo to provide 1.75 g of the crude amide acid. AcCl (6.1 g, 78 mmol) was added to the crude amide acid in 100 mL of dichloromethane at reflux, and the reaction mixture was refluxed under Ar for 48 h. The solution was cooled down to room temperature and stirred for 7 d under Ar. The solvent was then removed in vacuo. The residue was added to 100 mL of H2O and extracted with ethyl acetate (3 × 75 mL). The combined organic layers were dried over MgSO4, filtered, and concentrated in vacuo. The residue was purified by silica gel chromatography (70:30:2 hexane/EtOAc/AcOH) and recrystallized from methanol to provide 0.73 g (35% from 1) of the carboxylic acid indolylfulgimide 4. Z-form: 1H NMR (CD3OD, 400 MHz) δ 7.39 (d, J = 8.1 Hz, 1H), 7.24 (d, J = 8.0 Hz, 1H), 7.19 (td, J = 7.4, 1.1 Hz, 1H), 7.09 (td, J = 7.5, 0.9 Hz, 1H), 4.36 (s, 2H), 3.73 (s, 3H), 2.26 (s, 3H), 2.11 (s, 3H), 0.96 (s, 3H). 13C NMR (CD3OD, 100 MHz) δ 169.4, 167.9, 165.7, 156.1, 139.1, 138.5, 133.8, 130.5 (q, J = 36 Hz), 126.8, 124.0 (q, J = 272 Hz), 123.8, 122.9, 121.7, 120.2, 110.5, 108.3, 39.7, 30.2, 26.8, 22.3, 12.0. Anal. Calcd for C21H19F3N2O4: C, 60.00; H, 4.56; N, 6.66. Found: C, 60.28; H, 4.89; N, 6.39. C-form: 1H NMR (CD3OD, 400 MHz) δ 7.67 (d, J = 8.3 Hz, 1H), 7.38 (td, J = 7.8, 1.1 Hz, 1H), 6.77 (t, J = 7.9 Hz, 2H), 4.24 (d, J = 17.7 Hz, 1H), 4.20 (d, J = 17.7 Hz, 1H), 2.96 (s, 3H), 1.81 (s, 3H), 1.37 (s, 3H), 1.23 (s, 3H). 13C NMR (CD3OD, 100 MHz) δ 171.2, 170.0, 167.1, 161.7, 161.6, 141.4, 137.1, 136.3, 129.1 (q, J = 7 Hz), 124.2 (q, J = 272 Hz), 120.5, 119.9, 110.9, 106.8 (q, J = 37 Hz), 77.3, 40.2, 39.6, 33.0, 19.9, 19.6, 14.8. HRMS (ESI+) calcd for C21H19F N + 3 2O4 (M + Na)+ 443.1195, obsd 443.1195.
Carboxylic acid indolylfulgimide 4Z (0.19 g, 0.45 mmol) in 250 mL of toluene was irradiated with 405 nm light to obtain the photostationary state. Purification of the resulting 4C was performed via silica gel chromatography (70:30:2 hexane/EtOAc/AcOH) followed by recrystallization from CH2Cl2/hexane to provide 0.14 g (74%) of 4C. Fulgimide 4C (0.10 g, 0.24 mmol) in 50 mL of 50 mM sodium phosphate buffer (pH 7.4) was incubated at 37 °C for 12 h. The solution was then acidified with dilute HCl (1 M) to pH 5 and extracted with ethyl acetate (3 × 25 mL). The combined organic layers were dried over MgSO4, filtered, and concentrated in vacuo. Purification was performed via recrystallization from CH2Cl2/toluene to provide 40 mg (42%) of the dicarboxylic acid indolylfulgimide 5. C-form: 1H NMR (CD3OD, 400 MHz) δ 7.72 (d, J = 8.3 Hz, 1H), 7.30 (td, J = 8.1, 1.2 Hz, 1H), 6.67 - 6.71 (m, 2H), 4.22 (d, J = 17.3 Hz, 1H), 4.17 (d, J = 17.8 Hz, 1H), 2.98 (s, 3H), 1.81 (s, 3H), 1.40 (s, 3H), 1.20 (s, 3H); 13C NMR (CD3OD, 100 MHz) δ 171.3, 170.5, 169.9, 168.2, 159.7, 156.6, 137.8, 137.6, 135.4, 126.8, 122.1, 119.5, 110.9, 110.2, 74.3, 41.4, 39.3, 32.1, 20.4, 19.2, 15.8. Anal. Calcd for C21H20N2O6: C, 63.63; H, 5.09; N, 7.07. Found: C, 63.63; H, 5.34; N, 6.89. Z-form: 1H NMR (CD3OD, 400 MHz) δ 7.46 (d, J = 8.0 Hz, 1H), 7.41 (d, J = 8.3 Hz, 1H), 7.20 (t, J = 7.5, 1H), 7.07 (t, J = 7.7, 1H), 4.34 (s, 2H), 3.74 (s, 3H), 2.29 (s, 3H), 2.26 (s, 3H), 1.07 (s, 3H); 13C NMR (CD3OD, 100 MHz) δ 172.6, 170.7, 169.0, 168.1, 154.1, 140.2, 138.5, 136.8, 126.8, 123.6, 123.4, 123.2, 121.6, 120.3, 110.6, 110.1, 39.7, 30.2, 26.6, 22.6, 12.0. HRMS (ESI+) calcd for C20H20N2O6 (M + Na)+ 419.1246, obsd 419.1233.
Fulgimide 4C (0.142 g, 0.32 mmol) in 250 mL of 50 mM sodium phosphate buffer (pH 7.4) was incubated at 37 °C for 12 h. The solution was then acidified with concd HCl to pH 1 and extracted with ethyl acetate (3 × 75 mL). The combined organic layers were left overnight, dried over MgSO4, filtered, and concentrated in vacuo. Purification was performed via recrystallization from CH2Cl2/hexane to provide 76 mg (64%) of the carboxylic acid indolylfulgimide 6. C-form: 1H NMR (D2O, 400 MHz) δ 7.58 (d, J = 7.6 Hz, 1H), 7.34 (td, J = 7.7, 1.2 Hz, 1H), 6.79 (td, J = 7.5, 0.8 Hz, 1H), 6.74 (d, J = 8.3, 1H), 6.52 (s, 1H), 4.00 (s, 2H), 3.93 (s, 3H), 1.74 (s, 3H), 1.33 (s, 3H), 1.14 (s, 3H); 13C NMR (CDCl3, 100 MHz) δ 173.0, 170.0, 168.5, 157.8, 157.4, 138.9, 135.2, 133.4, 123.5, 123.4, 118.8, 109.1, 100.3, 72.8, 41.3, 38.6, 31.9, 20.2, 19.3, 16.1. HRMS (ESI+) calcd for C20H20N2O4 (M + Na)+ 375.1321, obsd 375.1323. E-form: 1H NMR (D2O, 400 MHz) δ 7.87 (s, 1H), 7.51 (d, J = 8.3 Hz, 1H), 7.45 (d, J = 7.9 Hz, 1H), 7.30 (td, J = 7.4, 1.0 Hz, 1H), 7.17 (td, J = 7.6, 1.0 Hz, 1H), 4.17 (s, 2H), 3.75 (s, 3H), 2.45 (s, 3H), 2.31 (s, 3H), 1.19 (s, 3H); 13C NMR (CDCl3, 100 MHz) δ 172.2, 169.6, 168.2, 152.1, 141.2, 136.9, 127.5, 126.4, 123.4, 122.1, 121.3, 121.0, 120.1, 110.4, 109.4, 38.9, 30.2, 26.7, 22.2, 11.6. HRMS (ESI+) calcd for C20H20N2O4 (M + Na)+ 375.1321, obsd 375.1305.
Concentrated, air-saturated stock solutions of 4Z in toluene and 50 mM sodium phosphate buffer (pH 7.4) were prepared in duplicate or triplicate. From each stock solution, five samples ranging in concentration from 0.25 to 0.05 mM were then prepared by dilution with toluene or buffer. A UV-vis spectrum was then acquired for each sample. Extinction coefficients and λmax were determined. According to 1H NMR data, Z/E-isomerization in D2O with sodium phosphate buffer at room temperature in 1 h was insignificant (1 – 2%). No isomerization was observed in toluene.
A concentrated, air-saturated stock solution of 4C (see synthesis) in toluene was diluted to four or five different concentrations, and their UV–vis spectra obtained. Each 4C solution was then quantitatively converted to 4Z solution by illumination with 515 nm light, and the concentration of fulgimide present was ascertained using the predetermined extinction coefficient of 4Z. Since the original concentration of 4C will be equivalent to the final concentration of 4Z, the original concentration of 4C was determined. The extinction coefficient and λmax for 4C were then determined from the initial spectra.
The extinction coefficient and λmax for 5C in 50 mM sodium phosphate buffer (pH 7.4) were determined in the same manner as for 4Z. To obtain these values for 5Z, four or five diluted 5C solutions in buffer were then quantitatively converted to 5Z solutions by irradiation with 515 nm light. UV-vis spectra of freshly prepared 5Z solutions were measured, and the extinction coefficient was obtained using the previously determined extinction coefficient of 5C. Typical error was 3%.
A solution of 4Z in toluene-d8 was illuminated with 405 nm light, and the Z/E/C- ratio was monitored via 1H NMR spectroscopy until PSS405nm was achieved. To measure the PSS405nm of 5 in D2O with 50 mM sodium phosphate buffer (pD 7.4), a solution of 5C, which was initially obtained from 4C, was converted to 5Z using 515 nm light. PSS was then achieved by irradiation of 5Z with 405 nm light. Z/E/C- ratio was monitored by 1H NMR spectroscopy.
The thermal/hydrolytic stability of the Z- and C-forms of fulgimides 4 and 5 was measured using UV-vis and 1H NMR spectroscopy. Solutions of 4Z were prepared in toluene or 50 mM sodium phosphate buffer (pH 7.4) and transferred into several ampoules. NMR samples of 4Z were prepared in toluene-d8 or D2O with 50 mM sodium phosphate buffer (pD 7.4). UV-vis and 1H NMR spectra of these initial samples were then acquired. Ampoules and NMR tubes were sealed and incubated in water baths which were maintained at 80 °C (toluene) or at 37 °C (buffer). At predetermined times, ampoules and NMR tubes were removed, and their contents were analyzed by UVvis and 1H NMR spectroscopy, respectively. UV-vis and 1H NMR spectra were then compared to the initial spectra. The thermal stability of 4C in toluene was measured using a PSS405nm solution, and evaluated as described for 4Z. In the case of 4C in buffer, the UV-vis spectra of 4C and its decomposition product were almost identical, thus UV-vis spectroscopy was not used to follow the decomposition of 4C; only the 1H NMR experiment was performed. Several pure 4C solutions were prepared in D2O with buffer (pD 7.4) and transferred into several ampoules which were then placed in a water bath maintained at 37 °C. At prescribed times, solutions were transferred into NMR tubes, and their spectra were taken immediately. To determine the stability of 5C and 5Z in buffer, pure 5C solutions were used while 5Z solutions were prepared by irradiation of a 5C solution with 515 nm light. These 5C and 5Z solutions were then analyzed in the same manner as 4Z. In addition, decomposition of 5C in D2O with buffer (pD 7.4) was also followed by UV-vis spectroscopy. Typical error was 5 × 10−5 h−1 with the exception of 4Z in toluene which was 0.003 h−1.
For 1H NMR spectroscopy, the residual toluene resonance (toluene) or added DMSO resonance (buffer) were utilized as internal standards, and signals corresponding to the individual species were integrated relative to the internal standards. To confirm the solvent isotope effect for 5C in buffer, two experiments in D2O and H2O buffer solutions were performed simultaneously and followed by UV-vis spectroscopy.
Air-saturated solutions of 4Z in toluene and in toluene in the presence of an excess of tributylamine (27 mM) or acetic acid (27 mM) were prepared with initial absorbencies of 0.6 – 0.8 at the absorption maxima. Samples were irradiated to PSS405nm with 405 nm light, and the absorbencies at λmax were measured. Then, in three cases (toluene, toluene/tributylamine, toluene/acetic acid), fresh 4Z solutions were irradiated to 90% of PSS405nm. The time taken to achieve 90% of the absorbance at PSS405nm was then recorded (coloration reaction Z to C). The 90% PSS mixture was then irradiated with 515 nm light using a separate filter. The time taken for the absorbance at λmax of the C-form to reach < 1% was recorded (decolorization reaction C to Z). Once the duration of irradiation was established for both the 90% PSS405nm coloration and < 1% C-form decolorization reactions, the system was automated through the use of a filter switch. All solutions were capped and stirred. After a designated number of irradiation cycles (coloration followed by decolorization), the samples were fully converted to PSS405nm, and their UV-vis spectra scanned. The photochemical stability was then determined by comparison with the initial PSS405nm (PSS at zero irradiation cycles) absorption spectra.
To measure the photochemical stability of 5 in 50 mM sodium phosphate buffer (pH 7.4) the freshly obtained 5C solutions were quantitatively converted to 5Z solutions by irradiating with 515 nm light. The same procedure as described for 4 was then applied. A control experiment to investigate the thermal decomposition of 5Z at room temperature after 120 h was also performed. After 120 h, besides Z to E isomerization, the thermal decomposition was determined by 1H NMR to be 1%. Z to E isomerization will not affect the photochemical decomposition results as these two forms are interconverted photochemically under our conditions. Typical error was 20%.
Financial support from the NIH/NIGMS programs (S06GM008205 and SC3GM084752) is gratefully acknowledged. We thank Dr. Kevin O'Shea from Florida International University for his helpful comments.