Synthesis and characterization of improved chloride-sensitive fluorescent indicators for biological applications

ANALYTICAL

BIOCHEMISTRY

178,355-361(1989)

Synthesis and Characterization of Improved ChlorideSensitive Fluorescent Indicators for Biological Applications A. S. Verkman, M. C. Sellers, A. C. Chao, T. Leung,* and R. Ketcham* Department of Medicine, Cardiovascular University of California, San Francisco,

Received

November

Research Institute, California 94143

and *Department

of Pharmaceutical

Chemistry,

14,198s

A class of N-substituted quinoline compounds has been introduced recently for the fluorescence measurement of Cl concentration in biological preparations. The most Cl-sensitive compound was 6-methoxy-N-[3sulfopropyl] quinolinium with peak excitation and emission wavelengths of 350 and 442 nm and a SternVolmer constant for quenching by Cl of 118 M-‘. Six water-soluble quinoline derivatives were synthesized and characterized for the purposes of increasing Cl sensitivity, adding ester functions for cell trapping, and red-shifting the fluorescence peak wavelengths. Acetic acid ester functions were added at the N-, 2-, and ~-POsitions of the quinoline ring. The best ester compound, N-(6-methoxyquinolyl)acetoethyl ester (MQAE), was water soluble (270 g/liter at 23°C; octanol:HzO partition coefficient of 0.009), had a high Cl sensitivity (Stern-Volmer constant 200 M-l), peak excitation and emission wavelengths of 355 and 460 nm, a fluorescence lifetime of 21.6 ns, and a molar absorbance of 4850 M-’ cm-’ (320 nm). MQAE fluorescence was not altered by the physiological anions HCOS, SOr, and POr , by cations, or by pH. MQAE was used to measure chloride transport in liposome membranes and in cultured LLC-PKl cells in monolayer; MQAE leaked out of cells ~20% in 60 min at 37°C. The physical, optical, and anion quenching properties for the series of ester compounds were determined to establish a set of structure-activity correlates. 0 is89 Academic s-r-, 1~.

The development of Cl-sensitive fluorescent indicators provides a direct approach for measurement of intracellular Cl activity and for analysis of Cl transport mechanisms in biological membranes. The important characteristics of a Cl indicator are high Cl sensitivity and selectivity, rapid response to changes in [Cl], good loading and trapping in cells, low biological toxicity, 0003-2697/89 $3.00 Copyright 0 1989 by Academic Press, All rights of reproduction in any form

Inc. reserved.

high fluorescence quantum yield and molar absorbance, peak fluorescence excitation and emission wavelengths at the red end of the visible spectrum, and a change in fluorescence spectral shape with changes in [Cl] (1). The compound 6-methoxy-N-[3-sulfopropyl] quinolinium (SPQ),’ described first by Wolfbeis and Urban0 (2), has many but not all of the required properties for a Cl-sensitive fluorescent indicator. SPQ fluorescence (excitation 350 nm, emission 442 nm, molar absorbance 5430 M-’ cm-‘, quantum yield 0.6) is strongly quenched by Cl by a collisional mechanism with a Stern-Volmer constant of 118 M-’ and a response time of under 1 ms; the shape of the SPQ fluorescence spectrum is not altered by Cl (3). SPQ fluorescence is not quenched significantly by other physiological ions or by pH. Once loaded by passive diffusion or by direct entrapment, SPQ leaks slowly across biological membranes and has been used to examine successfully chloride transport mechanisms in a variety of biomembrane vesicles (4-7), intact cells (8,9), and liposomes (lo), without apparent biological toxicity. Because there were no established photochemical principles to predict the structural requirements of a chloride-sensitive fluorophore, a series of simple SPQ analogs having a single nitrogen heteroatom quaternized with a sulfoalkyl chain were synthesized and tested (11). Effects of ring structure, length of sulfoalkyl chain, and position and nature of quinoline ring substituents were tested. It was found that maximum chloride sensitivity (Stern-Volmer constant > 50 M-‘) required a quinoline backbone quaternized with a sulfopropyl chain and monosubstituted at ring positions 2 through 6 by an electron-donating group (e.g., methoxy, methyl). On the basis of these initial structure-activity correlations, we have examined further the structural modi‘Abbreviations used: SPQ, 6-methoxy-N-[3-sulfopropyl] ium; MQAE, N-(6-methoxyquinolyl)acetoethyl ester.

quinolin-

365

500

VEXKMAN

ICI‘ AL.

II: N-(6-Methoxyquinolyl)acetic acid. 6-Methoxyquinoline (1 g; 6.3 mmol) and 0.88 g (1 eq) bromoacetic acid were heated to 75°C under Nz for 10 min. A dark brown precipitate was obtained which was washed with 10 ml ether and 10 ml methanol. A dark yellow solid was recrystallized twice from a 1:l mixture of dilute aqueous alkali and methanol; 0.92 g (67%) II was obtained, mp 153-156°C. FIG. 1. Chemical structures of synthesized compounds. and chemical names are given under Methods.

Syntheses

fications required to maximize Cl sensitivity, to red-shift the fluorescence spectra, and to add ester functions for improved cell entrapment. Six compounds were synthesized and characterized (Fig. 1). The 6-methoxy substituent of SPQ was moved to the 7-position to red-shift the fluorescence spectra (compound I). The nitrogen heteroatom of quinoline was substituted with an acetic acid instead of a sulfoalkyl chain to facilitate coupling reactions (compound II). Four Cl-sensitive esters were prepared (compounds III-VI). Ester compound III had the highest Cl sensitivity of any of the fluorophores prepared to date and was used successfully to measure Cl activity in viable cultured cells. METHODS

Organic Synthesis All compounds were obtained from Aldrich Chemical Co. (Milwaukee, WI). Synthesis of compounds I, IV, and V required the Skraup ring closure reaction to produce the quinoline backbone. All of the product compounds (I-VI) were water soluble and obtained by quaternization of the N-heteroatom of quinoline with 1,3-propane sultone or a carboxy compound containing an (Yhalide. All product compounds were pure as judged by a single spot on reverse-phase thin-layer chromatography (1:35 methanol:chloroform) and a homogeneous fluorescence lifetime. Product structure was confirmed for all compounds by mass spectroscopy and NMR. I: 7- Methoxy - N - (3 - sulfopropyl) quinolinium. mAnisidine (15 g; 0.12 mol), 40.2 g glycerol, 20.5 g sodium 3-nitrobenzenesulfonate, and 23.6 ml concentrated sulfuric acid were refluxed for 3 h. NaOH (12 M) was added to give pH 10. The mixture was extracted with ether; the ether phase was dried with sodium sulfate and rotoevaporated to a small volume. 1,3-Propane sultone (1 eq) was added and the mixture was heated to 110°C for 30 min under N2 giving a yellow precipitate. The precipitate was washed with methanol and recrystallized twice from 1:l water:methanol solution to obtain 5.2 g (15%) I, mp 271-272°C.

III: N-(6-Methoxyquinolyl)acetoethyl ester (MQAE). 6-Methoxyquinoline (1 g; 6.3 mmol) and 1.04 g (1 eq) ethyl bromoacetate were heated to 70°C for 5 min under NS. The red precipitate was collected, pulverized, washed with 10 ml ether, and recrystallized twice from methanol to give 1.05 g (52%) III as light brown crystals, mp 177-179°C. IV: N-(3-Sulfopropyl)-6-quinolylacetoethyl ester. 4Aminophenylacetic acid (8.5 g; 56 mmol), 5.2 g nitrobenzene, 2.16 g ferrous sulfate hydrate, 3.67 g boric acid, 19.1 g glycerol, and 1 ml sulfuric acid were heated under reflux for 5 h. The resultant dark green solution was made alkaline with 12 M NaOH until a black precipitate formed. The solution was filtered and the dark green filtrate was acidified with acetic acid to pH 4.6. The solution was filtered to collect a brown precipitate (quinoline-6-acetic acid). The brown precipitate was washed with water and refluxed for 30 min with 100 ml 0.82 M NaOH and 0.5 g decolorizing carbon. Glacial acetic acid (7.8 ml) was added and the solution was refrigerated; 9.65 g (92%) quinoline-6-acetic acid was obtained. Ethanol (40 ml) was saturated with gaseous HCl by bubbling for 20 min. Quinoline-6-acetic acid (4 g) was added and refluxed for 4 h. The solution as rotoevaporated to a small volume and made alkaline with aqueous Na&03. The basic solution was extracted with ether which was dried and rotoevaporated to a small volume. The resulting liquid was vacuum-distilled at 1.3 mm Hg, and a bath temperature of 24O”C, to give 1.5 g (32%) light yellow oil, bp 162-165°C (quinoline-6-acetoethyl ester). This material was heated to 140°C for 1 h with 1 eq 1,3-propane sultone under Nz. The resultant purple crystals were washed with ether and ethanol. The precipitate was recrystallized twice from ethanol to yield 0.53 g (25%) IV, mp 270-273°C. V: N - (3 - Sulfopropyl) - 2 - quinolylacetoethyl ester. Quinoline N-oxide (29.9 g) and 35.3 ml methyl acetoacetate were heated to 45°C for 9 h under NB. The mixture was diluted in 1 liter ice water, stirred for 10 min, and filtered to give an orange/yellow solid. The solid was dissolved in 500 ml 10% cold HCl, neutralized with aqueous NaHC03, and extracted with chloroform. The chloroform was dried over Na2S04, evaporated to a small volume, and vacuum-distilled at 17O”C, 1.3 mm Hg to give an orange oil, 2-quinolylacetoethyl ester (18%). The oil (1.36 g) was heated with 0.59 ml (1 eq) 1,3-propane sultone to 125°C for 1 h under NP. The mixture was dis-

SYNTHESIS

OF

CHLORIDE

solved in 10 ml methanol and cooled to give an orange oil at the bottom of the flask. The oil was washed twice with ether and twice with ethyl acetate. The oil resisted crystallization. VI: N-(3-Sulfopropyl)-6-quinolyloxyacetoethyl ester. 6-Hydroxyquinoline (1 g; 6.9 mmol) was suspended in 25 ml dry dioxane and 1 eq NaH was added. Ethyl bromoacetate (1.15 g; 1 eq) was added and the solution was heated to 50°C for 4 h. Ten volumes of Hz0 was added and the product was extracted with ether. The ether phase was dried with sodium sulfate and rotoevaporated to a minimum volume. The product was heated with 1 eq 1,3-propane sultone to 150°C for 1 h. The reddishorange solution was dissolved in ethanol and a small amount of ether was added. The solution was refrigerated overnight to give 0.25 g (11%) VI, mp 118120°C. Compound solubilities in water were determined from the optical absorbance of a saturated solution at 23°C. 0ctanol:water partition coefficients were determined by a double extraction procedure as described previously (11). Optical Measurements Absorbance spectra were obtained on a HewlettPackard 8452A diode array spectrophotometer (Palo Alto, CA). Samples were measured in buffer consisting of 5 mM Na phosphate, pH 7.5. Sample concentrations were adjusted to give a peak optical density of under 0.4. Molar extinction coefficients were calculated by the Beer-Lambert Law. Fluorescence spectra and lifetimes were measured on an SLM 48000 fluorimeter (Urbana, IL). Spectra were obtained at 0.1 mM dye concentrations using single excitation and emission monochromators and corrected for the wavelength dependence of lamp intensity profile and detection system sensitivity. Fluorescence intensities of each compound (0.1 mM) were measured relative to SPQ (0.1 mM) using the excitation maxima and the integrated emission spectrum. Fluorescence lifetimes were determined by the phase-modulation technique at 10,30, and 50 MHz using He-Cd laser excitation (Linconix, Sunnyvale, CA; 7.5 mW at 325 nm) and a KV370 cut-on filter. Lifetimes were referenced against a dilute solution of dimethyl-1,4-bis(4methyl-5-phenyloxazol-2-yl)benzene in ethanol (lifetime 1.45 ns). Lifetimes were judged homogeneous by agreement of phase and modulation lifetimes to within 10% at the three frequencies. Rapid kinetic measurements were performed on a Hi-Tech SF51 stopped-flow apparatus (Wiltshire, England). All measurements were performed at 23 “C. Fluorescence

Quenching

Measurements

Quenching of compound fluorescence by a series of anions was tested using a 0.1 mM solution of each compound in 5 mM Na phosphate, pH 7.4. Microliter ali-

357

INDICATORS

quots of a 1 M stock solution of the sodium salt of the quencher were added to the fluorophore solution. Fluorescence intensities measured after quencher addition were corrected for dilution and used to calculate SternVolmer quenching constants from a linear fit to the equation, F,/F = 1 + KJQ], where F, and F are fluorescence intensities in the absence and presence of quencher, [Q] is quencher concentration, and K4 is the Stern-Volmer constant. Biological

Studies

A series of biological studies were performed using ester compound III. Membrane permeability was determined across 0.2-pm-diameter liposomes (30 mol% cholesterol in phosphatidylcholine, Sigma Chemical Co., St. Louis, MO). SPQ (5 mM) or compound III was entrapped in liposomes by extrusion through 0.2-pm Nuclepore filters using 400 psi (12). External dye was removed by passage down a Sephadex G-25M exclusion column (10). The quantity of entrapped fluorophore was determined from the amplitude of the fluorescence signal in response to addition of 50 mM external Cl (see Results and Discussion). The permeability and Cl sensitivity of compound III in intact cells were determined using the renal epithelial cell line LLC-PKl (ATC No. CL-101) grown to confluency on 18-mm round cover slips in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum. Compound III (5 mM) was loaded into cells by a 3-min incubation in hypotonic (150 mOsm) solution at 37°C (9). Cells were placed in a chamber in which fluids bathing the cells could be changed rapidly. Cell fluorescence (excitation 360 f 5 nm, emission > 410 nm) was monitored through a 40X quartz objective (Leitz, N.A. 0.65, glycerol immersion) using a Nikon inverted epifluorescence microscope. RESULTS

AND

DISCUSSION

As outlined in the introduction, specific structural modifications of the SPQ molecule were made to improve Cl sensitivity, to red-shift the fluorescence spectra, and to add ester functions without compromising Cl sensitivity or optical properties. The sensitivities of the six new compounds and SPQ to a series of anions are summarized in Table 1. The fluorescence of these compounds was quenched in a dose-dependent manner by the halides Cl, Br, and I, by SCN, and by organic anions such as citrate; there was no effect of the anions NOg, S04, and HC03, the cations Na, K, and Ca, and pH in the range 4-8 on fluorescence. Stern-Volmer plots for quenching of each compound by the anions were linear in the anion concentration range O-50 mM (see Fig. 2); anion sensitivities were described by the Stern-Volmer constant (M-l). Lifetime and spectral studies showed that the fluorescence was quenched by a collisional

358

VERKMAN TABLE Stern-Volmer

Constants

1 for Anion

Stern-Volmer Compound

SPQ I II III IV V VI

Quenching

constant

(M-l)

Cl

Br

I

SCN

Citrate

118 9 113 200 75 105 80

175 50 205 293 89 135 90

276 85 325 456 115 176 144

211 69 255 410 106 142 112

15 2 31 41 22 44 11

Note. Stern-Volmer constants were the fluorescent compounds (0.1 mM in by a series of anions as described under were linear to anion concentrations of generally >0.995).

determined for quenching of 5 mM Na phosphate, pH 7.5) Methods. Stern-Volmer plots 50 mM in all cases (r Values

mechanism without a change in the shape of the excitation or emission spectra (not shown), similar to results obtained for SPQ (2). Stopped-flow kinetic studies showed that fluorescence was quenched in SCN > Br > Cl > citrate % (SO,, NOa, HC03). Compound I has the same chemical structure as SPQ except that the methoxy substituent is at the 7instead of the 6-position. On the basis of our previous observation that substitution at the g-position of quinoline results in a red-shift in the fluorescence spectrum (ll), compound I was synthesized to have similar physical and chemical properties to SPQ with a red-shifted fluorescence spectrum. Surprisingly, the ability of Cl to quench fluorescence was almost abolished for this compound. Similarly, the compound 8-ethoxy-N-(3-sulfopropyl) quinolinium, prepared by quaternization of 8ethoxyquinoline with 1,3-propane sultone, had a marked red-shifted fluorescence spectrum (excitation 370 nm, emission 505 nm) with a Stern-Volmer constant for quenching by Cl of <2 M-‘. These results indicate that the position of the alkoxy group strongly influences Cl sensitivity; substitution at positions 7 and 8 gives a fluorophore with little Cl sensitivity. Compounds II and III are the first Cl-sensitive fluorophores where the N-heteroatom was quaternized with a group other than a sulfoalkyl chain. The N-substituted acetic acid compound (II) is highly Cl-sensitive and could be of utility in coupling the quinoline chromophore with other Cl-insensitive chromophores, polymeric dextrans, or porous glass beads. In addition, compound II would be produced upon enzymatic ester hydrolysis of selected ester derivatives of the N-acetic acid (e.g., acetoxymethyl or pivolic esters) that are substrates for intracellular esterases. Compound III, the ethyl ester of

ET

AL.

compound II, has the highest Cl sensitivity of any compound synthesized to date. Interestingly, compound III is the first Cl-sensitive indicator having a net positive charge, suggesting that the negative charge of the carboxy (compound II) or the sulfonate (compounds IVVI) groups hinders the quenching interaction between the anions and the positive nitrogen. Compounds IV and V have ester functions at the 2- and g-position on the quinoline ring. Compound VI is very similar to SPQ except that an acetic acid ethyl ester function has been added to the 6-methoxy group. Each of these compounds is quenched strongly by Cl. Further studies were performed to support the finding that deletion of the negative charge on the alkyl chain at the N-position increases Cl sensitivity. Stern-Volmer constants for quenching of compound II by Cl were obtained as a function of pH to examine the effects of carboxy group protonation. At pH values of 7.5,6.5,5.5,4.5, 3.5, 2.5, and 1.5, Stern-Volmer constants were (in M-l) 99, 98, 96, 98, 96, 137, and 150, indicating that deletion of the negative charge by carboxy group protonation results in increased Cl sensitivity. In addition, a compound having no negative charge, 6-methoxy-[N-ethyl]quinolinium, was synthesized by reaction of 6-methoxyquinoline with iodoethane. This compound had a SternVolmer constant for quenching by Cl of 146 M-‘, higher than any of the compounds having a negative charge on the N-substituent. The physical and optical properties of the synthesized compounds are summarized in Table 2. Each of the compounds was water soluble. Compound III had the greatest water solubility and lowest octanol:water partition coefficient, probably because it is a cationic salt and hence dissolves with its anion, Br. Interestingly, compound I, which is structurally similar to SPQ, had a relatively low water solubility and high octanol:water parti-

FIG. 2. Stern-Volmer plots for quenching of compounds I, II, and III by Cl. Aliquots of a 2 M NaCl solution were added to a stirred suspension of each compound (0.1 mM) in 5 mM Na phosphate, pH 7.5. F,,/F is the fluorescence in the presence of Cl divided by that in the absence of Cl. Data were fitted by a linear regression; Stern-Volmer constants are given in Table 1.

SYNTHESIS

OF

CHLORIDE TABLE

359

INDICATORS 2

Physical and Optical Properties of Synthesized Compounds

Solubility Compound

SPQ I II

hM)

109

6

OctanokH20 partition coefficient

Fluorescence Peak

ex/em

Absorbance 7

x

(X0.01)

(nm)

b-4

I

1.04

350/442

25.3

1.0

(M-l

1.12

385/436 350/445

8.2 23.3

1.7 0.4

III

1144

0.89

355/460

21.6

1.0

IV V VI

19

1.25 1.82 1.51

3221420 3351405 3471436

11.0 9.5 20.7

0.9

Em-‘)

5430 3470 8090 4530 2890 4850 2800 6660 4760 5110

318

117

83

2.92

bun) 350 352 318 346 320 350 322 318 316

Note. Compounds are given by number; see Fig. 1 and Methods for chemical names and structures. Data are given for each absorbance peak if more than one peak is present. Abbreviations: 7, fluorescence lifetime; I, relative fluorescence intensity, X, wavelength, t, molar absorbance.

tion coefficient. The octanol:water partition coefficient correlates inversly with water solubility and provides a semiquantitative measure of the passive bilayer permeability. The octanol:water partition coefficient for the new compounds provides an estimate of their half-time for leakage across artificial and biological membranes when compared to data reported for SPQ. The fluorescence excitation and emission spectra for compounds I, III, and VI are given in Fig. 3. For biological applications, red-shifted fluorescence spectra are desirable to minimize cell and instrument autofluorescence and photodynamic cell injury. As predicted, the fluorescence excitation spectrum of compound I was redshifted compared to that for SPQ. Compound III had red-shifted excitation and emission spectra, which together with its high Cl sensitivity, makes it a desirable compound for use in biological systems. The fluorescence spectra for compounds VI and V, having a -CH2attached to the quinoline backbone at the 2- and ~-POsitions, are blue-shifted and similar to the spectral properties of a series of methyl-substituted quinolinium dyes reported previously (11). For this reason, compound VI was synthesized to contain an -0-CH2- moiety substituted on the quinoline ring, similar to the 6methoxy group on SPQ. The fluorescence excitation and emission spectra of compound VI were similar to those for SPQ. The nanosecond fluorescence lifetimes of Cl indicators have been shown to correlate roughly with Cl sensitivity (11). The fluorescence lifetime of compound III, the most Cl-sensitive compound, was one of the highest of the six compounds, 21.6 ns. The long lifetime of this compound makes it a useful pH-insensitive lifetime standard in the ultraviolet. The brightness of a fluorophore is a function of molar absorbance at the fluores-

cence excitation wavelength and quantum yield. The data in Table 2 show similar molar absorbance values for the new compounds. These values are similar to the molar absorbance of the calcium indicator fura-2, but are an order of magnitude lower than those of fluorescein and rhodamine compounds. The relative intensities listed in Table 2 provide a semiquantitative measure of

EXCITATION

400

WAVELENGTH

450

EMISSION

FIG. 3. compounds centrations

Corrected

fluorescence

I, III, and VI. Spectra in 5 mM Na phosphate,

500

WAVELENGTH excitation and emission spectra of were obtained at 0.1 mM dye conpH 7.5.

360

VERKMAN

compound brightness when viewed by fluorescence microscopy. The relative intensity was greatest for compound I, probably because of its relatively high molar absorbance, and similar for SPQ and compounds II, III, and VI. These results show that ester functions can be added to the quinolinium class of Cl-sensitive fluorescent indicators without loss of Cl sensitivity. Although it was not the purpose of the present study to place multiple acetoxymethyl ester functions on the quinoline backbone for intracellular cleavage, these results establish a strategy for the addition of such ester functions at several positions. It was found that Cl sensitivity can be increased greatly by substitution with an acetoethyl ester function at the N-position (compound III). The resultant compound is very hydrophilic and has a higher Cl sensitivity and a red-shifted fluorescence spectra compared to SPQ. Further experiments were performed to show that compound III (MQAE) is suitable for studies of Cl transport in liposomes and in intact cells. MQAE was entrapped into 90% phosphatidylcholine/ 10% cholesterol liposomes by extrusion through 0.2-pm filters; external MQAE was removed by exclusion chromatography. Figure 4, top shows the time course of MQAE fluorescence in response to a 50 IIIM inward Cl gradient. There was a time course of decreasing fluorescence corresponding to Cl influx and MQAE quenching. The initial rate of Cl influx was 0.18 mM/s. Similar studies performed by stopped-flow fluorimetry in the presence of the ionophore tributyltin showed that Cl influx into liposomes is measurable with a half-time of under 100 ms (not shown). The curve labeled “24 h” represents the same Cl influx measurement repeated after incubating the liposomes containing entrapped MQAE for 24 h at 4°C. There was approximately 30% dye leakage. At 23°C MQAE leaked from liposomes ~10% in 4 h. These results are consistent with the high water solubility and low octanol:water partition coefficient of MQAE. Experiments were performed using intact kidney epithelial cells (LLC-PKl) grown in monolayer culture on glass cover slips. To assess toxicity, effects of MQAE on cell growth and viability were determined. MQAE (5 mM) was added to subconfluent cells in culture. There was no morphological evidence of cell toxicity after 48 h; cells reached confluency at the same time as cells not exposed to MQAE. For fluorescence microscopy studies, MQAE was loaded into cells made transiently permeable to small molecules by a brief exposure to a hypotonic solution containing MQAE. The fluorescence of individual cells was monitored by epifluorescence microscopy (Fig. 4, bottom). There was no detectable photobleaching when cells were illuminated with a low, but adequate light intensity for accurate measurement of intracellular Cl activity. At 23”C, MQAE leaked out of cells lo-20% in 60 min, as judged from the fractional decrease in fluorescence (relative to the baseline signal after SCN addi-

ET

AL.

0.6

-

TBT loos

2 2

15

TIME

I.0 0

KSCN

0.8

0.6

TIME

FIG.

4. Biological studies of compound III (MQAE). Top, liposomes (-10 pM PC, 1 pM cholesterol) containing 5 mM MQAE, 100 mM K gluconate, 5 mM Na phosphate, 1 pg/ml valinomycin, pH 7.5, were suspended in a buffer containing 50 mM K gluconate, 50 mM KCl, 5 mM Na phosphate, pH 7.5. The time course of MQAE fluorescence (excitation 355 nm, emission > 400 nm) was measured immediately and at 24 h after removal of external MQAE by column chromatography (see Methods). Bottom, LLC-PKl cells were loaded with MQAE and total cell fluorescence was quantitated by epifluorescence microscopy (excitation 360 + 5 nm, emission > 410 nm). Cells were bathed with a series of calibration solutions of varying chloride activities (shown), high K (120 mM) and the ionophores nigericin (5 pM) and tributyltin (10 wM; Ref. (8)). Gluconate was used to replace Cl. At the end of the experiment, 150 mM KSCN containing 5 pM valinomycin was added to quench >99% of MQAE fluorescence.

tion) in the absence of external Cl. At 37°C there was 20-30% MQAE leakage at 60 min. It was reported that the sensitivity of SPQ to Cl was an order of magnitude lower within cells than with that in aqueous solution, probably because of a combination of factors including the presence of cytoplasmic organic anions and proteins, and the increased cytoplasmic viscosity (8). It is thus important to design Cl indicators with maximal Cl sensitivity in aqueous solutions to maximize intracellular sensitivity. In LLC-PKl cells, as shown in Fig. 4, bottom, MQAE fluorescence decreased by -10% for an increase in intracellular chloride activity from 0 to 12 mM; 50% of MQAE fluorescence was quenched at -60 mM Cl. This represents a -50% higher Cl sensitivity than that reported for SPQ in cells of the intact kidney proximal tubule (8). These results indicate that MQAE has suitable physical, optical, and biological properties for measurement of Cl transport in intact cultured cells. The very high Cl sensitivity of MQAE and its deesterified acid (compound II) should facilitate the development of specialized ap-

SYNTHESIS

plications of Cl-sensitive indicators including Cl sensors and dextran-bound Cl indicators.

OF

CHLORIDE

fiberoptic

361

INDICATORS

3. Illsley, 1219.

N. P., and Verkman,

A. S. (1987)

Bio&mistry

4. Chen, P.-Y., Illsley, N. P., and Verkman, Physiol. 254, F114-F120.

ACKNOWLEDGMENTS

26,1215-

A. S. (1988)

Amer.

J.

5. Chen, P.-Y., and Verkman, A.

This work was supported by Grants DK39354, DK35124 and HL42368 from the National Institutes of Health, by a grant from the National Cystic Fibrosis Foundation, and by a grant-in-aid from the American Heart Association with funds from the Long Beach Chapter. Dr. Chao was supported bv traininn Grant HL07185. Dr. Verkman -is an established investigator of the American Heart Association. REFERENCES

S. (1988) Biochemistry 27, 655fxn ---. 6. Fong, P., Illsley, N. P., Widdicombe, J. H., and Verkman, A. S. (1988) J. Membr. Biol. 104,233-239. 7. Pearce, D., and Verkman, A. S. (1989) Biophys. J., in press.

8. Krapf, R., Berry, 955-962. 9. Verkman, and Chao,

C. A., andverkman,

A. S. (1988)

A. S., Sellers, M. C., Ketchum, A. C. (1989) Kidney Znt. 35,492.

1. Verkman, A. S., Chen, P.-Y., Davis, B., Fong, P., Illsley, N. P., and Krapf, R. (1988) in Cellular and Molecular Basis of Cystic Fibrosis (Mastella, G., and Quinton, P. M. Eds.), pp. 471-478, San Francisco Press, San Francisco.

11. Krapf,

2. Wolfbeis, O., and Urbano, 314,577-581.

12. Mayer, L. D., Hope, M. J., and Cullis, phys. Acta 858,161-168.

E. (1983)

Fresenius’

2. Anal.

Chem.

R., Widdicombe,

10. Verkman, A. S., Takla, R., Sefton, B., Basbaum, combe, J. H. (1989) Biochemistry, in press. R., Illsley,

N. P., Tseng,

Biophys.

H. C., and Verkman, P. R. (1986)

J. 53, J. H.,

C., and WiddiA. S. (1988) Biochim.

Bio-