Synthesis and Photolytic Assessment of Nitroindolinyl-Caged Calcium Ion Chelators

Neuroactive amino acids derivatised at their carboxylate groups with a photolabile nitroindolinyl group are highly effective reagents for the sub-µs release of neuroactive amino acids in physiological solutions. However, the same does not apply in the case of calcium ion chelators. In this study, nitroindolinyl-caged BAPTA is found to be completely photostable, whereas nitroindolinyl-caged EDTA photolyses only when saturated with calcium ions.


Introduction
Localised fluctuations of free Ca 2+ concentrations play critical roles in essential physiological processes [1][2][3][4][5], such as neurotransmission, skeletal and cardiac muscle contraction, hormone secretion, chemotaxis and blood clotting. In order to achieve sudden jumps of calcium ions in an experimental set up, a number of different photolabile molecules that bind Ca 2+ have been developed over the last 35 years [6][7][8][9][10][11][12][13]. Flash irradiation of these compounds with near-UV light causes either fragmentation or other structural change, which results in a fast change in their calcium affinity, thereby resulting in the rapid release of Ca 2+ ions in solution and triggering a biological response. This concept is illustrated by the first of these reagents, DM-nitrophen, which is a photolabile derivative of EDTA ( Figure 1) [12]. However, despite the canonical scheme of the photolysis shown in Figure 1, it is notable that compounds incorporating an N-(2-nitrobenzyl)glycine moiety, as in DMnitrophen, are susceptible to a side reaction of photo-decarboxylation. For DM-nitrophen, a detailed study showed this comprised approximately 10% of the total amount of the compound photolysed in the absence of Ca 2+ , and significantly more (16.5%) in the presence of saturating Ca 2+ [14].
The advantages and drawbacks of the various reagents for the photorelease of Ca 2+ developed over the years since the introduction of DM-nitrophen have been discussed [15] and the field can be considered to be well matured. In contrast, photochemical means to effect rapid decreases in Ca 2+ concentration have been less well described, yet Ca 2+ signalling is widespread and includes processes (reviewed by Berridge et al. [16] including muscle function, fertilisation, axis formation, cell differentiation, proliferation, transcriptional activation and apoptosis. Although much research has focuses on the initiation of processes due to a rise in the Ca 2+ concentration, there must be a subsequent fall to reset the system. To our knowledge, the only reagents described to reduce Ca 2+ concentrations experimentally are restricted to studies by the Tsien laboratory [17] and to a preliminary communication by Ferenczi et al. [18]. All of the reagents explored in these reports are based on photochemistry

Results and Discussion
Our extended programme for the development of effective photocleavable cages of biologically active compounds included the investigation of the photolysis mechanism(s) of nitroindoline-caged compounds [21,22]. Having established that nitroindolinyl groups covalently linked to neuroactive amino acids can release biologically active compounds on a sub-microsecond time scale upon flash photolysis, we embarked on the synthesis of nitroindoline-caged calcium chelators that could potentially cause rapid calcium uptake upon photorelease. A summary of the results of this work has been disclosed in a review [23] and full details of the synthesis and photochemical evaluation are described here.
We envisaged that it would be optimal to construct the BAPTA framework and covalently attach the photolabile nitroindoline cage towards the end of the synthesis process. We initially planned the synthesis by employing our first developed nitroindoline cage bearing a CH2CO2Me group at the 5-position of the ring [19]. Thus, 2-nitrophenol was easily alkylated to 2-(2-nitrophenoxy)ethanol 1, which was reduced to 2-(2-aminophenoxy)ethanol 2. Double alkylation gave 3, which was tosylated to 4 and converted to communication by Ferenczi et al. [18]. All of the reagents explored in these reports are based on photochemistry in which the calcium affinity of the chelator 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid (BAPTA) is weakened by the incorporation of a photolabile modification. BAPTA itself has Kd ~0.11 µM at pH 7 and, in the strategies previously explored, blocking one of its chelating carboxylate groups weakens Kd by approximately three orders of magnitude [17,18]. A similar reduction is discussed below for the nitroindoline derivative of EDTA 24. We aimed to extend this approach, using the 7-nitroindolinyl caging group previously developed by us for caging L-glutamate [19,20] and other amino acids ( Figure 2) In this event, irradiation of the compounds synthesised showed surprisingly little photosensitivity and the results are described herein.

Results and Discussion
Our extended programme for the development of effective photocleavable cages of biologically active compounds included the investigation of the photolysis mechanism(s) of nitroindoline-caged compounds [21,22]. Having established that nitroindolinyl groups covalently linked to neuroactive amino acids can release biologically active compounds on a sub-microsecond time scale upon flash photolysis, we embarked on the synthesis of nitroindoline-caged calcium chelators that could potentially cause rapid calcium uptake upon photorelease. A summary of the results of this work has been disclosed in a review [23] and full details of the synthesis and photochemical evaluation are described here.
We envisaged that it would be optimal to construct the BAPTA framework and covalently attach the photolabile nitroindoline cage towards the end of the synthesis process. We initially planned the synthesis by employing our first developed nitroindoline cage bearing a CH2CO2Me group at the 5-position of the ring [19]. Thus, 2-nitrophenol was easily alkylated to 2-(2-nitrophenoxy)ethanol 1, which was reduced to 2-(2-aminophenoxy)ethanol 2. Double alkylation gave 3, which was tosylated to 4 and converted to

Results and Discussion
Our extended programme for the development of effective photocleavable cages of biologically active compounds included the investigation of the photolysis mechanism(s) of nitroindoline-caged compounds [21,22]. Having established that nitroindolinyl groups covalently linked to neuroactive amino acids can release biologically active compounds on a sub-microsecond time scale upon flash photolysis, we embarked on the synthesis of nitroindoline-caged calcium chelators that could potentially cause rapid calcium uptake upon photorelease. A summary of the results of this work has been disclosed in a review [23] and full details of the synthesis and photochemical evaluation are described here.
We envisaged that it would be optimal to construct the BAPTA framework and covalently attach the photolabile nitroindoline cage towards the end of the synthesis process. We initially planned the synthesis by employing our first developed nitroindoline cage bearing a CH 2 CO 2 Me group at the 5-position of the ring [19]. Thus, 2-nitrophenol was easily alkylated to 2-(2-nitrophenoxy)ethanol 1, which was reduced to 2-(2-aminophenoxy)ethanol 2. Double alkylation gave 3, which was tosylated to 4 and converted to bromide 5. Bromide displacement of 5 with 2-nitrophenol gave the desired compound 6 in a moderate yield (Scheme 1). However, two by-products 6a and 6b were also isolated from this reaction. We speculate that, in competition with the expected intermolecular bromide displacement by the 2-nitrophenoxide, 5 had also undergone an intramolecular cyclisation process to form a quaternary benzoxazonium species, with subsequent dealkylation by 2-nitrophenoxide to give by-products 6a and 6b (Scheme 2). However, two by-products 6a and 6b were also isolated from this reaction. We speculate that, in competition with the expected intermolecular bromide displacement by the 2-nitrophenoxide, 5 had also undergone an intramolecular cyclisation process to form a quaternary benzoxazonium species, with subsequent dealkylation by 2-nitrophenoxide to give by-products 6a and 6b (Scheme 2). However, two by-products 6a and 6b were also isolated from this reaction. We speculate that, in competition with the expected intermolecular bromide displacement by the 2-nitrophenoxide, 5 had also undergone an intramolecular cyclisation process to form a quaternary benzoxazonium species, with subsequent dealkylation by 2-nitrophenoxide to give by-products 6a and 6b (Scheme 2). Compound 6 was then reduced to aniline 7. In straightforward steps, indoline 8 [19] was acylated to give the bromoacetyl indoline 9, which, after nitration, gave the photolabile precursor 10. Alkylation of amine 7 with 10 gave 11 but a number of attempts to achieve a final N-alkylation of 11 with t-butyl bromoacetate gave only a complex mixture of uncharacterised products (Scheme 3).
Compound 6 was then reduced to aniline 7. In straightforward steps, indoline 8 [19] was acylated to give the bromoacetyl indoline 9, which, after nitration, gave the photolabile precursor 10. Alkylation of amine 7 with 10 gave 11 but a number of attempts to achieve a final N-alkylation of 11 with t-butyl bromoacetate gave only a complex mixture of uncharacterised products (Scheme 3). To probe the failure of the required N-alkylation of the secondary amine of 11, we prepared the less congested amine 12. Treatment with a slight excess of t-butyl bromoacetate and either sodium hydride or DIPEA showed that C-alkylation on the 5-substituent of the indoline was the principal product 13, whereas a large excess of the alkylating reagent and a higher reaction temperature gave the doubly alkylated product 14 (Scheme 4).
To avoid the unwanted alkylation, we changed to a 5-methyl substituted indoline, and the synthesis then proceeded smoothly to give 17, which, upon treatment with an excess of t-butyl bromoacetate, gave the protected caged BAPTA 18 in a reasonable yield. Surprisingly, we also isolated a substantial amount of the tetrabutyl ester of BAPTA 18a (Scheme 5). Evidently, hydrolysis by water, which formed from K2CO3 during alkylation, had caused some hydrolysis of the amide, and the released free carboxylate then underwent esterification to give 18a. Deprotection with TFA successfully gave the target nitroindolinylcaged BAPTA 19. (Scheme 5). To probe the failure of the required N-alkylation of the secondary amine of 11, we prepared the less congested amine 12. Treatment with a slight excess of t-butyl bromoacetate and either sodium hydride or DIPEA showed that C-alkylation on the 5-substituent of the indoline was the principal product 13, whereas a large excess of the alkylating reagent and a higher reaction temperature gave the doubly alkylated product 14 (Scheme 4).
To avoid the unwanted alkylation, we changed to a 5-methyl substituted indoline, and the synthesis then proceeded smoothly to give 17, which, upon treatment with an excess of t-butyl bromoacetate, gave the protected caged BAPTA 18 in a reasonable yield. Surprisingly, we also isolated a substantial amount of the tetrabutyl ester of BAPTA 18a (Scheme 5). Evidently, hydrolysis by water, which formed from K 2 CO 3 during alkylation, had caused some hydrolysis of the amide, and the released free carboxylate then underwent esterification to give 18a. Deprotection with TFA successfully gave the target nitroindolinylcaged BAPTA 19 (Scheme 5).
Having prepared our target NI-caged BAPTA chelator, we next examined its near-UV photolysis properties. Surprisingly, sequential irradiations of a solution of 19 monitored via UV spectroscopy, as described in the Experimental Section, showed no change ascribable to the expected nitroindolinyl photocleavage, either with or without Ca 2+ present ( Figure 3). Control irradiation of NI-caged glutamate under identical conditions confirmed the expected photolysis, as previously reported (data not shown) [19]. Having prepared our target NI-caged BAPTA chelator, we next examined its near-UV photolysis properties. Surprisingly, sequential irradiations of a solution of 19 monitored via UV spectroscopy, as described in the Experimental Section, showed no change ascribable to the expected nitroindolinyl photocleavage, either with or without Ca 2+ present ( Figure 3). Control irradiation of NI-caged glutamate under identical conditions confirmed the expected photolysis, as previously reported (data not shown) [19]. It appears that the photostability of 19 is likely to arise from quenching of the excited state of the nitroindolinyl group by the electron-rich aryl ring(s).We have previously observed an apparently similar quenching in the electron-rich 1-acetyl-4-dimethylamino-7nitroindoline [24] and efficient formation of a low-lying triplet state in N,N-dimethylamino-4-nitroaniline has been reported elsewhere [25].
After the failure to record any measurable photolysis of NI-caged BAPTA 19, we changed our focus to the synthesis of MNI-caged EDTA. We chose to employ the more efficient 4-methoxy-7-nitroindolinyl (MNI) cage, previously developed in our laboratory [20,24]. Thus, nitration of N-bromoacetyl indoline 20 gave essentially only the 7-nitro isomer 21, which reacted with secondary amine 22 to give the tri-tert-butyl ester 23. Deprotection with TFA afforded the desired MNI-caged EDTA 24 (Scheme 6). It appears that the photostability of 19 is likely to arise from quenching of the excited state of the nitroindolinyl group by the electron-rich aryl ring(s).We have previously observed an apparently similar quenching in the electron-rich 1-acetyl-4-dimethylamino-7nitroindoline [24] and efficient formation of a low-lying triplet state in N,N-dimethylamino-4-nitroaniline has been reported elsewhere [25].
After the failure to record any measurable photolysis of NI-caged BAPTA 19, we changed our focus to the synthesis of MNI-caged EDTA. We chose to employ the more efficient 4-methoxy-7-nitroindolinyl (MNI) cage, previously developed in our laboratory [20,24]. Thus, nitration of N-bromoacetyl indoline 20 gave essentially only the 7-nitro isomer 21, which reacted with secondary amine 22 to give the tri-tert-butyl ester 23. Deprotection with TFA afforded the desired MNI-caged EDTA 24 (Scheme 6). We then exposed MNI-caged EDTA 24 to sequential near-UV irradiation in the absence (A) and presence (B) of Ca 2+ . Clean photolysis of 24 was only observed in solution (B), producing good isosbestic points and the characteristic appearance of the nitrosoindole photoproduct formation, as evident from the cumulative appearance of the new peak maximum at 413 nm ( Figure 4). Irradiation in the absence of Ca 2+ evidently caused photolytic changes, as shown by the progressive rise in absorption near 340 nm, but the product(s) were not investigated further. It is likely that, in the absence of Ca ions, irradiation of 24 can instigate electron transfer from the aliphatic tertiary amino groups to the excited state of the nitroindoline moiety and thus block the expected photolysis pathway. Such single-electron transfer is likely to trigger decarboxylation of 24, as reported to occur for cation radicals of α-amino acids [14,26]. In solution (B), as the caged chelator has a K d for calcium of~10 µM (cf., the similar K d value for the related pentadentate chelator N-methylethylenediaminetriacetic acid [27]), the large excess of Ca 2+ would cause full saturation. Thus, the lone electron pairs on each of the tertiary amino groups would be unavailable for single-electron transfer and the normal 7-nitroindolinyl photolysis can operate, as shown by the appearance of the 413 nm band upon irradiation. Notably, the 5,7-dinitroindolinyl-caged BAPTA amide reported previously [17]  We then exposed MNI-caged EDTA 24 to sequential near-UV irradiation in the absence (A) and presence (B) of Ca 2+ . Clean photolysis of 24 was only observed in solution (B), producing good isosbestic points and the characteristic appearance of the nitrosoindole photoproduct formation, as evident from the cumulative appearance of the new peak maximum at 413 nm ( Figure 4). Irradiation in the absence of Ca 2+ evidently caused photolytic changes, as shown by the progressive rise in absorption near 340 nm, but the product(s) were not investigated further. It is likely that, in the absence of Ca ions, irradiation of 24 can instigate electron transfer from the aliphatic tertiary amino groups to the excited state of the nitroindoline moiety and thus block the expected photolysis pathway. Such single-electron transfer is likely to trigger decarboxylation of 24, as reported to occur for cation radicals of α-amino acids [14,26]. In solution (B), as the caged chelator has a Kd for calcium of ~10 µM (cf., the similar Kd value for the related pentadentate chelator N-methylethylenediaminetriacetic acid [27]), the large excess of Ca 2+ would cause full saturation. Thus, the lone electron pairs on each of the tertiary amino groups would be unavailable for single-electron transfer and the normal 7-nitroindolinyl photolysis can operate, as shown by the appearance of the 413 nm band upon irradiation. Notably, the 5,7-dinitroindolinyl-caged BAPTA amide reported previously [17] exhibited the same photolytic behaviour as 24 and only photolysed in the presence of excess Ca 2+ .

Experimental Section
1 H NMR spectra were determined on Varian Unityplus 500 or JEOL FX90Q spectrometers in CDCl3 solution with TMS as an internal reference, unless otherwise specified. Elemental analyses were carried out by MEDAC Ltd., Surrey, UK. Electrospray mass spectra were recorded at the School of Pharmacy, University of London. Merck 9385 silica gel was used for flash chromatography. Analytical HPLC was performed on a 250 mm × 4 mm Merck Lichrospher RP8 column or a 125 mm × 4 mm Whatman Partisphere SAX column. The flow rate was 1.5 mL min −1 with either column. Preparative HPLC was carried out on a 2 cm × 30 cm column (Waters C18 packing, Cat. No. 20594) at a 2 mL min −1 flow rate. Details of mobile phases are given at relevant points in the text. Triethylammonium bicarbonate (TEAB) solution was prepared by bubbling CO2 into an ice-cold aqueous solution of 1 M triethylamine until the pH stabilised (pH ~7.4). Preparative anion-exchange chromatography used a column of DEAE-cellulose (2 cm × 20 cm). Detection for all analytical and preparative chromatography was at 254 nm. Organic solvents were dried over anhydrous Na2SO4 and evaporated under reduced pressure. Hexanes (bp 40 °C-60 °C) were redistilled before use. Photolysis experiments were performed in a Rayonet RPR-100 photochemical reactor fitted with 16 nm × 350 nm lamps.

Experimental Section
1 H NMR spectra were determined on Varian Unityplus 500 or JEOL FX90Q spectrometers in CDCl 3 solution with TMS as an internal reference, unless otherwise specified. Elemental analyses were carried out by MEDAC Ltd., Surrey, UK. Electrospray mass spectra were recorded at the School of Pharmacy, University of London. Merck 9385 silica gel was used for flash chromatography. Analytical HPLC was performed on a 250 mm × 4 mm Merck Lichrospher RP8 column or a 125 mm × 4 mm Whatman Partisphere SAX column. The flow rate was 1.5 mL min −1 with either column. Preparative HPLC was carried out on a 2 cm × 30 cm column (Waters C 18 packing, Cat. No. 20594) at a 2 mL min −1 flow rate. Details of mobile phases are given at relevant points in the text. Triethylammonium bicarbonate (TEAB) solution was prepared by bubbling CO 2 into an ice-cold aqueous solution of 1 M triethylamine until the pH stabilised (pH~7.4). Preparative anion-exchange chromatography used a column of DEAE-cellulose (2 cm × 20 cm). Detection for all analytical and preparative chromatography was at 254 nm. Organic solvents were dried over anhydrous Na 2 SO 4 and evaporated under reduced pressure. Hexanes (bp 40-60 • C) were redistilled before use. Photolysis experiments were performed in a Rayonet RPR-100 photochemical reactor fitted with 16 nm × 350 nm lamps.

Methyl 2-(1-(2-bromoacetyl)-7-nitroindolin-5-yl)acetate 10
To a stirred solution of NaNO 3 (221 mg, 2.6 mmol) in TFA (12 mL) was added 9 (749 mg, 2.4 mmol) and the mixture was stirred at rt for 4 h. The red/brown solution was poured into ice-cold water and extracted with EtOAc. The combined organic phases were washed with saturated aq. NaHCO 3 , brine, dried and evaporated to give a red viscous oil, which, after trituration with Et 2 O, afforded 10 (679 mg, 79%) as yellow crystals, mp 102-103 • C (EtOAc-hexanes); ν max /cm −1 (Nujol) 1730, 1670, 1530, 1395, 1375  To a solution of 12 (80 mg, 0.2 mmol) in dry DMF (5 mL) was added sodium hydride (60% dispersion in mineral oil; 8 mg, 0.2 mmol) and t-butyl bromoacetate (43 mg, 0.22 mmol) and the mixture was refluxed under nitrogen. The progress of the reaction was followed by TLC (EtOAc-hexanes (3:2)). After 3 h more t-butyl bromoacetate was added (43 mg) and the heating continued for further 1 h. The solution was diluted with water (50 mL) and washed with EtOAc. The combined organic phases were washed with saturated aq. NaHCO 3 , brine, dried, and evaporated to give brown oil. Flash chromatography (EtOAc-hexanes (1:1)) gave 13 (51 mg, 50%) as a pale oil; 1 H NMR (500 MHz)  To a solution of 12 (100 mg, 0.25 mmol) in dry MeCN (10 mL) was added anhydrous K 2 CO 3 (138 mg, 1 mmol) and t-butyl bromoacetate (488 mg, 2.5 mmol) and the mixture was heated under reflux. After 2 h, more t-butyl bromoacetate was added (488 mg, 2.5 mmol) and the heating continued for a further 16 h. The solid was filtered off, washed with MeCN and the filtrate evaporated. The residue was dissolved in EtOAc and washed with saturated aq. NaHCO 3 , brine, dried, and evaporated to give a brown oil. Flash chromatography (EtOAc-hexanes (2:3)) gave 14 (70 mg, 44%) as a pale oil; 1 H NMR (500 MHz) δ 7.57 (1H, br s, H-6), 7.36 (1H, br s, H-4 3.14. 2-Bromo-1-(5-methylindolin-1-yl)ethan-1-one 15 To an ice-cold solution of 5-methylindole (1.31 g, 10 mmol) in acetic acid (50 mL) NaCNBH 3 (2.48 g, 30 mmol) was added portionwise over 10 min (exothermic reaction) and the mixture was then stirred at rt for 0.5 h. Water (2-3 mL) was added and the solvent removed in vacuo. The residue was dissolved in EtOAc and washed with saturated aq. NaHCO 3 , brine, dried, and evaporated to give 5-methylindoline (1.33 g, 100%) as a light brown oil, which was immediately used in the next step; 1 H NMR (90 MHz) δ 7.08-6. To a solution of 17 (345 mg, 0.5 mmol) in dry MeCN (15 mL) was added anhydrous K 2 CO 3 (276 mg, 2 mmol) and t-butyl bromoacetate (975 mg, 5 mmol) and the mixture was heated under reflux. After 2 h more t-butyl bromoacetate was added (975 mg) and the heating continued for a further 16 h. The solid was filtered off, washed with MeCN and the filtrate evaporated. The residue was dissolved in EtOAc and washed with saturated aq. NaHCO 3 and brine, dried, and evaporated to give a brown oil. Flash chromatography (EtOAc-hexanes (2:3)) gave two products. The first eluted material, a viscous oil which crystallised after trituration with Et 2 O-hexanes, identified as tetra-tert-butyl 1,2bis (2-aminophenoxy)  A solution of the crude brown film 18 (163 mg, 0.202 mmol) in TFA (10 mL) was stirred at rt for 4 h. The red-brown solution was concentrated in vacuo and the residue dissolved in water (80 mL). The pH was raised from 1.92 to 7.2, via the careful addition of 1 M aq. NaOH, and extracted with ether (3 × 80 mL). The aqueous solution (89 mL) was first concentrated in vacuo and then filtered through a 0.2 mm cellulose membrane and analysed via reversephase HPLC (mobile phase 25 mM Na phosphate, pH 6.0 + 30% MeCN at 1.5 mL/min). A major peak eluted at t R 5.4 min and a minor peak at t R 6.4 min. The solution was made up to 25 mM Na phosphate, pH 6.0, and loaded onto a preparative HPLC column. The column was eluted first with 25 mM Na phosphate, pH 6.0 for 1 h (all flow rates 1.5 mL/min) and then with 25 mM Na phosphate, pH 6.0 + 30% MeCN. Fractions containing pure product were analysed, combined, and quantified via UV spectroscopy: λ max (25 mM Na phosphate, pH 7.0/nm 345 (ε/M −1 cm −1 2700)) to give 19 (NI-caged BAPTA) (48 µmol). The contaminated fractions were re-analysed via standard anion exchange HPLC (mobile phase 50 mM ammonium phosphate, pH 6.0 + 10% MeCN at 1.5 mL/min), showing a minor peak at t R 2.8 min and a major peak at t R 6.0 min. The solution was then concentrated, and the residue dissolved in water (87 mL) and the pH was adjusted to 7.42. The aqueous solution was then purified via anion-exchange chromatography using a linear gradient formed from 10 to 500 mM TEAB (each 250 mL). Fractions containing the product, which were eluted at~240 mM TEAB were analysed as above. Pure fractions were combined, concentrated, and re-evaporated from MeOH. The residue was dissolved in water (1 mL) and quantified via UV (11.2 mM, 11 µmol). The total yield of isolated pure product 19 was 59 µmol (29%). A portion of the product was exchanged to sodium salt with Dowex-50. Two separate solutions of 19 (0.5 mM)-one in 25 mM MOPS, pH 7.09 containing 5 mM dithiothreitol plus 2.5 mM EDTA and the other in 25 mM MOPS, pH 7.09 containing 5 mM dithiothreitol plus 2.5 mM Ca 2+ -were irradiated for increasing lengths of time (0, 0.5, 1, and 2 min) in 1 mm path-length cells in a Rayonet Photochemical Reactor (16 × 350 nm lamps). No photolysis was detected via UV spectroscopy in either solution ( Figure 3). As control experiments, two separate solutions of methyl 1-[S-(4-amino-4-carboxybutanoyl)]-7-nitroindoline-5-acetate, NI-caged glutamate (0.5 mM) were irradiated under the same conditions described above. The expected 7-nitrosoindole (λ max 413 nm) was observed, as previously reported [18] (data not shown).

Estimated Extent of Photolysis of MNI-Caged EDTA 24 by HPLC
Two separate solutions of 24 (0.5 mM)-one in 25 mM MOPS, pH 7.02 plus 2.5 mM EDTA and the other in 25 mM MOPS, pH 7.09 plus 2.5 mM Ca 2+ -were irradiated for 0.25 min in 1 mm path-length cells in a Rayonet Photochemical Reactor (16 × 350 nm lamps). The extent of photolysis was determined via reverse-phase HPLC (mobile phase 15 mM Na phosphate, pH 6.0 + 10% MeCN, 1.5 mL/min). Quantification was based on peak heights compared to those of unphotolysed controls. The extent of photolysis of the solution containing 2.5 mM EDTA was 30.1% and that of the one containing 2.5 mM Ca 2+ was 49.1%.

Conclusions
Our aim in exploring 7-nitroindolinyl-caged calcium chelators as tools for rapidly lowering Ca 2+ concentrations in physiological media was to exploit the sub-microsecond time scale of photolysis of this caging technology. Previously reported examples [17,18] of caged calcium chelators exhibited photolysis rates that were significantly slower, at best in the~250 µs range.
Synthesis of the nitroindolinyl-caged BAPTA 19 was relatively straightforward but lengthy, and ultimately fruitless, as it was not susceptible to photolysis either with or without the presence of calcium. In contrast, nitroindolinyl-caged EDTA 24 was readily accessed via a short synthetic route but clean photolysis to release free EDTA only occurred in the presence of saturating Ca 2+ , thus enabling no capacity to bind additional Ca 2+ . It appears that in both the BAPTA and EDTA cases, strong electron-donating centres are capable of quenching the excited state of the nitroindoline. Our experiments, and those of previous investigators, seem to have exhausted the options for conventional caging of calcium chelators. An intriguing possibility would be a photoisomerisable scaffold that could bring together two spatially separated iminoacetate groups, thereby assembling a complete hexadentate ligand. If this could be achieved, it would have an additional advantage of minimal Ca 2+ affinity in the pre-irradiated form. Although attractive as a concept, present resources place this beyond our capability.