Application of photoactivatable fluorescent active-site directed probes to serine-containing enzymes

149

Biochimica et Biophysica Acta, 669 (1981) 149-156

Elsevier]North-Holland Biomedical Press BBA 38692

APPLICATION OF PHOTOACTIVATABLE FLUORESCENT ACTIVE-SITE DIRECTED PROBES TO SERINE-CONTAINING ENZYMES KIMON J. ANGEL1DES Department o f Biochemistry, McGill University, Montreal, Quebec, H3G 1 Y6 {CanadaJ

(Received December 4th, 1980) (Revised manuscript received March 20th, 1981)

Key words: Photoactivatable fluorescence; Active site; Protein relaxation; Serine-containing enzyme," Photoaffinity labeling

A photoactivatable fluorescent anthraniloyl group has been directed to the active-site serine group of a-chymotrypsin and trypsin. The acylated derivatives are nonfluorescent until irradiated. When activated by light a highly reactive nitrene is generated which is capable of covalent insertion into the protein matrix. The resultant insertion product of this photolysis is a highly fluorescent reporter group which has little rotational mobility and is cross-linked through the serine to the protein matrix in the active site reqion of the protein. Because of the sensitivity to the polarity of the environment shown by the anthraniloyl chromophore, the dipolar relaxation characteristics of the cross-linked enzyme and deacylated enzyme were determined. These measurements show that little relaxation occurs on the nanosecond time scale for the cross-linked enzyme, but upon deacylation of the serine increased dipolar relaxation of the protein with the attached reporter group is observed. The use of these activesite directed photoactivatable fluorescent probes can be extended to probe the active-site structure of complex enzymes and conformational dynamics of active-site regions in proteins and to serve as potential functional site labels in fluorescence resonance energy transfer measurements.

Introduction Photoaffinity labeling has been a useful tool in elucidating the active-site structure of enzymes, in identifying and locating functional sites of macromolecules, and in aiding in the purification and characterization of receptors and transport proteins [ 1 - 3 ] . Aromatic azido compounds have mostly been used since their stability in the dark is good and because of their relative ease of preparation. Irradiation of these compounds generates a highly reactive nitrene which is capable of nondiscriminate insertion into the functional site of the macromolecule. In order to avoid the application of radioisotopes, labels utilizing fluorescence emission have been used in a few Abbreviations: NPAB,p-nitrophenyl 2-azidobenzoate; TLCK, N-a-p-tosyl-L-lysine chloromethyl ketone HC1; DFP, diisopropyl phosphofluoridate.

instances [4,5]. Fluorescence spectroscopy itself can be a valuable technique for the study of macromolecular conformation and topology [6,7]. During the course of our investigation of the fast sodium channel of excitable membranes, we sought possible photoactivatable covalent fluorescent derivatives of tetrodotoxin and saxitoxin, to provide a potential means towards purification and characterization of the ionchannel, to provide a sensitive marker of the events of covalent bond insertion into the protein, to provide information concerning the conformational dynamics and microenvironment of the toxin binding site and to serve as an energy donor to map the channel structure by fluorescence resonance energy transfer techniques. Towards these ends, a number of photoactivatable fluorescent derivatives of tetrodotoxin and saxitoxin have been prepared [15,16]. The reactive portions were designed to a fulfill a number of chemical and spectroscopic criteria: (1) a ligand

0 005-2795/81/0000-0000/$02.50 © Elsevier/North-Holland Biomedical Press

150 which is reversible in the dark; (2) a rapid half-life for insertion into the receptor site; (3)a large increase in the fluorescence quantum yield upon photolysis which would be very sensitive to the environment; (4) an excited-state lifetime which can be conveniently measured; and (5) a molecule which would be relatively small so that the binding characteristics o f the natural ligand would not be significantly altered. Derivatives of anthranilic acid fulfill many of these criteria and this study reports on the synthesis and chemical and spectroscopic application of these reagents as precursors to photoactivatable fluorescent tetrodotoxin derivatives. We have chosen a-chymotrypsin and trypsin, two serine-containing enzymes, as models in order to characterize the chemical and spectroscopic properties of this probe. Materials and Methods 2-Aminobenzoic acid, carbonyldiimidazole, p-nitrophenol and proflavine hydrochioride were obtained from Aldrich Chemical Company. p-Nitrophenyl anthranilate was obtained from Molecular Probes, Piano, Texas. a-Chymotrypsin (3-times crystallized) and trypsin were from Worthington Biochemicals. N-Benzyloxycarbonyl-L-phenylalaninep-nitrophenyl ester and diisopropylphosphorylchymotrypsin were from Sigma and were used without further purification. All other chemicals and solvents were the best reagent grades commercially available and deionized-distilled water was used in all solutions. Ultraviolet spectroscopy was performed on a Cary 14 spectrophotometer and infrared spectroscopy was done in KBr discs using a Perkin-Elmer Model 257 spectrophotometer. Proton NMR spectra were taken on a Varian XL-200 spectrometer (200 MHz). Photolysis and fluorescence spectroscopy were done with a SLM 4800 subnanosecond scanning fluorimeter equipped with a 450 W Xenon lamp (SLM Instruments, Urbana, IL.) which is interfaced to a MINC 11 computer (Digital Equipment Corporation). Thinlayer chromatography was performed on Eastman 13255 0.16 mm cellulose plates.

function of p-nitrophenyl anthranilate (p-nitrophenyl 2- aminobenzoate) to the azido group. Recrystallized 2-aminobenzoic acid (2.0 retool)was dissolved in 4.0 ml 2 N HC1 and 0.8 ml water was added at 4°C. Sodium nitrite (2 retool) in 1.0 ml water was added and the reaction mixture was stirred for 90 min at 4°C. The reaction mixture was filtered to remove some insoluble material, and then degassed to expel excess nitrous acid. All subsequent steps were carried out in a dark-room equipped with a red light. 2.0 mmol sodium azide in 1.0 ml water were slowly added to the cooled solution and allowed to react with stirring for 45 min at 4°C. At the end of this time a precipitate formed which was collected and subsequently washed with 10 vol. cold water. The material was allowed to dry over P2Os for several days. The azidobenzoic acid was applied to a cellulose TLC plate and developed in chloroform/methanol (3 : 1, v/v). It moved with an R F value of 0.17, and had a melting point of 88-93°C. Esterification of 2-azidobenzoic acid withp-nitrophenol was performed using the carbonyldiimidazole procedure. Carbodiimidazole (1.7 mmol) and 2-azidobenzoic acid (0.6 mmol) were dissolved in 5.0 ml of dry dimethyl formamide (dried over 3-A molecular sieves) and stirred for 30 min prior to introducing p-nitrophenol (0.4 mmol) which was dissolved in 5 ml of dimethyl formamide. The reaction was allowed to proceed overnight in the dark at room temperature. After evaporation of the solvent to dryness, the residue was repeatedly washed with acetone. The sample was recrystallized from absolute ethanol as off-white sheets, m.p. 105-106°C, with an RF of 0.61 in chloroform/methanol ( 3 : 1 , v/v). The second method of preparation of p-nitrophenyl 2-azidobenzoate was similar to the preparation of the parent compound 2-azidobenzoic acid. The starting material was p-nitrophenylanthranilate. However, in order to avoid possible acid-catalyzed hydrolysis of the ester, "the diazotization was done in 0.5 N HC1 at 4°C. Proton NMR ((C 2Ha)2 SO): p-nitrophenylanthranilate: 6~H = 6.48--7.84; 6~NH2 = 8.32; NPAB: CS~H= 6.48--7.84.

Synthesis of p-nitrophenyl 2-azidobenzoate [NPAB) NPAB was prepared by two different routes: either coapling of the imidazole ester of 2-azido-benzoate to p-nitrophenol or by conversion of the 2-amino

Infrared spectroscopy in KBr discs certified the presence of the azido group (Na, 2120 cm-l). The synthesis of p-nitrophenyl-N-methylanthranilate was

151 performed according to the procedure of Haugland and Stryer [8].

Preparation of 2,azidobenzoyl-chymotrypsin trypsin

and

The 2-azidobenzoyl-enzyme derivatives were prepared by the addition of 168 gl of final concentration 2.2 • 10 -4 M NPAB (2.2/~nol) in dioxane to 10 ml 1.85" 10-4M a-chymotrypsin (1.87 grnol) in 0.1 M potassium phosphate buffer, pH 6.8, at 4°C. Aliquots were withdrawn at selected times and the concentration of p-nitrophenol determined spectrophotometrically at 410 nm. After 2 h the reaction mixture was fdtered through a cellulose-nitrate filter (0.45 pal) to remove any precipitated material, applied to a PD-10 (Sephadex G-25M)column, and eluted with water. Acylated trypsin was prepared in a similar manner. The acylation kinetics were followed by the release of p-nitrophenol monitored at 410 nm vs. time in 0.1 M potassium phosphate, pH 6.8, at 5°C.

Photolysis and fluorescence spectroscopy High intensity 310 nm irradiation from the 450 W Xenon lamp of the spectrofluorimeter illuminated the sample using 16 run slit widths. Measurements were made in 0.3 × 0.3 cm fluorescence microcuvettes with a capacity of 150 gl to minimize inner filter effects. The sample was kept at 5 °C with the aid of a thermostat. Fluorescence spectra were corrected for wavelength-dependent variation in light source output, phototube response, and monochromator efficiency by comparison with the emission spectrum of quinine bisulfate in 0.1 N H2SO4 [20,21]. The kinetics of the photolytic reaction were monitored by measuring the increase in fluorescence at 425 nm using the time-base mode of the spectrofluorimeter. Steady-state anisotropies were determined in the T format which simultaneously measures the ratio of the vertical and horizontal components of the emitted light with vertically polarized excitation light. Emission anisotropy is given as: 2P 3-P

In+I±

r e = tan q)/27rf •

where r = fluorescence lifetime, f = frequency, ~ = phase shift, and M -- demodulation. Fluorescence lifetimes were measured with a monochromater on the emission channel or with Schott KV filters in cases with low fluorescence intensities. The emission polarizer was set at 55 ° from the vertical position with vertically polarized excitation light to eliminate the effects of rotational motion on the observed lifetimes [9]. For wavelength-resolved lifetimes, the wavelength-dependent time response of the photomultiplier tube was corrected by the use of a phase reference according to Lakowicz et al. [I 1]. The reference compound was 1,4-bis(2-(5-phenyloxazoyl))benzene (POPOP) in ethanol, for which a lifetime of 1.35 ns has been reported [ 11 ]. The quantum yield of the probe bound at the enzyme active-site was determined by comparison with the quantum yield of quinine bisulfate and was corrected for polarized emission [17]. The absorbance of the sample was kept below 0.05 absorbance units to minimize inner filter effects. A quantum yield of 0.7 was used for quinine bisulfate in 0.1 N sulfuric acid at 23°C [18]. Diisopropylphosphoryl-chymotrypsin and TLCKinhibited trypsin were used in control experiments to block the enzyme active-site. Results

(1)

Reaction of p-nitrophenyl a-chymotrypsin and trypsin

(2)

Spectrophotometric measurements indicate that reaction of NPAB with the active-site serine group in both o~-chymotrypsin and trypsin is specific and stoichiometric. Under conditions in which 31.5 /aM

P is the degree of polarization and defined as: P _ I~ - I±

where Itl and 1± refer to the intensity of the parallel and perpendicular components, respectively, of the emitted light. Depolarization reflects the extent of Brownian motion of the fluorophore during the time scale of the fluorescence excited state (10 -9 s). Excited-state lifetimes were determined at frequencies of 30 MHz and 18 MHz by phase-modulation methods and calculated from

2-azidobenzoate with

152 NPAB and 22.3 /~M a-chymotrypsin were used, 1.0 equivalent of p-nitrophenol was released per equivalent enzyme as determined by following the absorbance at 410 nm. The acylation rate constant was 2.6 • 10 -2 min -t for hydrolysis by a-chymotrypsin at 5°C in 0.1 M potassium phosphate, pH 6.8. No further release of p-nitrophenol was noted during the next several hours. By comparison, the rate constants for hydrolysis of p-nitrophenyl anthranilate were 3.4" 10 -3 rain -1 and 1.8 • 10 -2 rain -1 for a-chymotrypsin and trypsin, respectively, and for the hydrolysis of p-nitrophenyl N-methylanthranilate 3.1 • 10 -2 rain -s and 1.2" 10 -2 min -t for a-chymotrypsin and trypsin, respectively. Reaction of NPAB with trypsin under conditions of excess substrate yielded an acylation rate of 1.3 • 10 -2 rain-' at 5°C. Under identical reaction conditions release of p-nitrophenol was not observed when DFP-inhibited chymotrypsin or TLCK-inhibited trypsin were used. After gel chromatography of the two acylated enzymes, no further p-nitrophenol was detected by absorption spectro-

photometry, which indicated that the 2-azidobenzoyl group was covalently bound to the enzyme. This was supported by the addition of nucleophilic reagents such as hydroxylamine to the acylated complex, which upon separation of the products yielded active enzyme that could subsequently be acylated by p-nitrophenylanthranilate or by another aliquot of NPAB. The ultraviolet difference spectra of the two 2-azidobenzoyl acylated enzymes are very similar and show absorption maxima at 268 nm with a small shoulder at 308 nm (Fig. 1A). On the other hand, the absorption maxima o f N-methylanthraniloylchymotrypsin and trypsin are centered about 345 nm (Fig. 1B). Upon illimination with monochromatic light at 310 nm the absorption spectra changed (Fig. 1A) and the fluorescence quantum yield of the two 2-azidobenzoyl enzymes increased dramatically. Under these conditions the photolytic reaction was complete in less than 3 min (Fig. 2A). The development of fluorescence by the conversion of the 2-azi-

16,

14

I '7 E u

1c

lO

"7

:E

E u 'T

:E

v

m x

o x

klJ

u)

240 2~,o 2~o 3~o 3~o Wavelength

3b,o 3~,o 3~o 4bo ,2o (nm)

2~o'

280

320

360

400

' 4,b

Wavelength (nm)

Fig. 1. A. Ultraviolet difference absorption spectra of 2-azidobenzoyl-chymotrypsin ( ) and 2-azidobenzoyl-trypsin (. . . . . . ) before and after photolysis at 310 nm for 10 rain. The spectra were taken in 0.I M potassium phosphate buffer, pH 6.8, at 25°(2 and represent dfffelence spectra of 2-azidobenzoyl-chymotrypsin or 2-azidobenzoyl-trypsin vs. a-chymotrypsin or trypsin. B. Difference absorption spectra ofN-methylanthraniloyl-chymotrypsin ( ) and N-methylanthraniloyl-trypsin (. . . . . -).

153 i

i'

i

i

,

,

90

i

i

i

i

,

i

i

i

i

i

//

?

o

5G

2 30

~oo

3bo

'

5~o

z~o

lC

Time (s) 360

400

440

480

520

560

Wavelength (nm) Fig. 2. A. The kinetics of 2-azidobenzoyl-chymotrypsin photolysis. 2-Azidobenzoyl-chymotrypsin was at a concentration of 46 /~M. The increase in quantum yield of the reaction product as a function of time was monitored by measuring the increase in fluorescence at 425 ( ) with excitation at 310 nm or by quenching of the protein tryptophan fluorescence (excitation 290, emission 340 nm) after irradiation at 3 1 0 n m due to the increased energy transfer fron donor tryptophan residues to the acceptor anthraniloyl group. At time = 0, high intensity 310 nm irradiation from the spectrofluorimeter illuminated the sample, using 16 nm slit widths. B. Fluorescence emission spectra (excitation at 340 nm) of the photolyzed products after illumination for 10 min: anthraniloyl-~-chymotrypsin ( ) and trypsin ( . . . . . -). Buffer composition was 0.1 M potassium phpsphate, pH 6.8, and the photolysis was carried out at 5°(3. The fluorescence emission spectrum (excitation at 340 nm) of N-methylanthraniloyl chymotrypsin is shown for comparison ( . . . . ).

dobenzoyl group can also be followed by the protein tryptophan fluorescence after the sample has been irradiated at 310 nm. With excitation at 290 nm the reduction in the emission peak of the protein centered at 340 nm due to increased energy-transfer from tryptophan to anthraniloyl yielded a similar rate constant for the photolytic reaction and development of fluorescent product (Fig. 2A, - . . . . . ). The rate of photolysis, however, may be varied by altering the lamp intensity or wavelength of irradiation. Fluorescence emission spectra of the a-chymotrypsin ( . . . . ) and trypsin ( . . . . . . ) NPAB reaction products taken after 10 min irradiation show maxima centered around 415 nm (Fig. 2B). The absolute quantum yields of the photolyzed fluorescent products are 0.43 for a-chymotrypsin and 0.49 for trypsin. Further addition of p-nitrophenylanthranilate to the photolyzed enzymes did not result in the appearance of p-nitrophenol. No fluorescence increase was observed when NPAB and DFP-inhibited cx-chymotrypsin or TLCK-inhibited trypsin were mixed and then

photolyzed. These results lend further support to the idea that the photolysed fluorescence product is bound to the protein and localized to the active-site region. After photolysis, addition of 0.1 M hydroxylamine to deacylate the enzyme did not eliminate the fluorescent signal. Desalting of the photolabeled deacylated enzyme on Sephadex G-25 gave a derivative which had 82% of its original catalytic activity towards p-nitrophenylantranilate or N-benzyloxycarbonyl-L-phenylalanine p-nitrophenyl ester. Analysis of the kinetic behaviour of the modified enzyme showed that the dissociation constant at pH 5.3 (0.1 M potassium phosphate) of proflavine [19] was 2.1.10-SM compared to 2.3" 1 0 ~ M for native o~-chymotrypsin. Photolysis of the parent compound, 2-azidobenzoic acid, was also carried out in a variety of solvents and with a few nucleophiles. The results show that in methanol and ethanol the conversion of fluorescent compound is slow (/Cobs= 6.1 • 10 -3 s -t) with emission

154

maxima of 402 and 395 nm, respectively. In Fig. 2B, it can be seen that the fluorescence spectra as a result of photolysis are slightly blue-shifted and exhibit a small shoulder on the red side of the emission maximum. It appeared that the photolytic reaction product was heterogenous and it was of interest to characterize the nature of the covalent bond formed between the azido group and the protein. The basic emission resembles that of the alkylamino product (e.g., acylation product of p-nitrophenyl N-methylanthranilate with a-chymotrypsin). When the reaction product of 2-azidobenzoic acid in imidazole buffer (0.1 M) was examined, a similar shoulder about 430 nm was observed, suggesting that one possible reaction product could be between the azido group and a nucleophilic protein imidazole group. Reaction of 2-azidobenzoic acid in methionine or aspartate-containing buffers gave spectra similar to that of unbuffered water. The 450 nm steady-state emission anisotropies with excitation at 340 nm of the photolyzed-fluorescent NPAB-enzyme conjugates were 0.312 and 0.333 for a-chymotrypsin and trypsin, respectively, with excited-state lifetimes of 7.3 -+0.4 ns and 7.6 -+ 0.3 ns for a-chymotrypsin and trypsin. Since methylanthranilate, which serves as a model for anthraniloylchymotrypsin, exhibits a substantial Stoke's shift [8], the dipolar relaxation characteristics of the covalent fluorescent insertion product were examined by wavelength-resolved average lifetimes. Fig. 3 shows the average lifetimes across the emission spectrum of methylanthranilate in ethanol and in glycerol, photolyzed anthraniloyl-chymotrypsin and photolyT:ed anthraniloyl-chymotrypsin to which 0.1 M hydroxylamine was added and repurified by gel chromatography. The fluorescence lifetime of methylanthranilate in ethanol is almost unchanged across the emission spectrum. When the solvent becomes more viscous, the lifetime of methylanthranilate becomes more wavelength-dependent, which suggests that spectral relaxation is occurring on the time scale of the fluorescence emission [ 10,11 ]. The lifetime of photolyzed anthraniloyl-chymotrypsin does not change with increasing wavelength. However, when hydroxylamine is added to deacylate the enzyme-substrate complex, the fluorescent probe remaining firmly attached to the protein matrix through the photolyzed azido group, an increase of 3.8 ns occurs from

12 ~, 10 o

8 .~

6

u-

2 •

3~o '

i

i

i

42o

i

i

i

i

4;,o

1

i

i

5oo

Wavelength, (nm)

Fig. 3. Fluorescence lifetimes of methylanthranilate in ethanol (e) at 25°(2 and in glycerol (o) at 15°C, photolyzed anthraniloyl-chymotrypsin (z~) at 25°C, and photolyzed anthraniloyl-chymotrypsin at 25°C which was deacylated with hydroxylamine (A). Apparent phase shift lifetimes were obtained at 30 MHz.

350 to 500 nm, indicating significant dipolar relaxation of the protein matrix on the nanosecond time scale. Discussion

A highly fluorescent anthraniloyl group which can be photoactivated has been inserted specifically at the active-site of a-chymotrypsin and trypsin. Acylation of the active-site serine residue of these enzymes occurs with the concomitant release ofp-nitrophenol, and subsequent photolysis irreversibly links the reporter group to the proteins. In the case of chymotrypsin one possible insertion product may be the imidazole of histidine 57, part of the catalytic triad of this enzyme [12]. The azidoanthraniloyl enzyme derivatives are essentially non-fluorescent until illuminated, which can be predicted by the presence of the azido group which can undergo an n ~ rr* transition, thus quenching the fluorescence [4,12]. Upon photolysis the quantum yield increases dramatically as the nitrene generated reacts to produce substituted amines or a guanidinium residue in the case of c~-chymotrypsin. The results presented here have fulftUed many of the design criteria: (1) the reagent reacts specifically with

155 the catalytic groups which are located at the active site of these enzymes; (2) the reporter group is relatively small; (3) it has a substantial quantum yield and a measurable excited-state lifetime and (4) it is also able to report on the environmental characteristics of the active site of these enzymes. The emission maxima are consistent with the highly polar nature of the environment of the anthraniloyl group of chymotrypsin [8]. Measurement of the fluorescence polarization concludes that the active-site structure of the acylenzyme is rigid. Furthermore, from the steadystate anisotropy and the dipolar relaxation characteristics of the anthraniloyl moiety, it is deduced that the photolyzed chromophore has little rotational mobility. Some evidence for the nature of the covalently inserted product in a-chymotrypsin indicates that the imidazole may be cross-linked to the acylated serine. Analysis of the active-site structure determined by X-ray crystallography shows that the active site can accomodate the anthraniloyl group (dimensions 6 ,~ wide + 3.5 A thick) and that cross-linking between the O"r of serine and the N ~2 of histidine (approx. 3.1 X) is possible [12]. The loss in 18% catalytic activity after deacylation of the modified enzyme may result from partial labeling of the histidine. The fact that proflavine binds to the modified enzyme active-site with an affinity equal to that of the native enzyme suggests that the introduction of the fluorescent label leads to a relatively minor perturbation and that no gross change in the active-site structure is involved. Certainly one distinct advantage of these probes is shown in Fig. 3, where relaxation of the protein matrix on the nanosecond time scale can be observed using this covalently inserted fluorescent probe. The average lifetime increases as the wavelength of observation increases and becomes most pronounced on the red edge of the fluorescence emission spectrum. Similar phenomena have b'een reported by Grinvald and Steinberg [14] and Lakowicz and coworkers [10,11] for dipolar relaxation around the tryptophan residues of selected proteins. The environmental sensitivity of the anthraniloyl group facilitates separation of the relaxed and unrelaxed states of the chromophore. A single average lifetime has been used to describe the decay process, although the actual rate law may be more complex. Inversion of the apparent fluorescence phase shift and demodu-

lation lifetimes was not observed, probably due in part to the multi-exponential decay of the fluorescence [11]. This work provides a basis for further work on the dynamics of protein fluctuations on the nanosecond time scale using extrinsic probes which are covalently inserted into the protein matrix. By these methods it is possible to direct a probe of selected spectral properties to an enzyme active-site or ligand-binding site of a receptor using the natural affinity of a ligand, to covalently link the probe to the active site and cleave off the ligand, leaving the probe attached to the macromolecular matrix. One would then use the attached probe to observe the rapid conformational fluctuations of the macromolecule on the nanosecond time scale in the presence and absence of ligands. In conclusion we have shown that 2-azidobenzoic acid can serve as a highly fluorescent photoactivatable label which can be easily made functional and conjugated to an enzyme substrate or to a ligand that, while still retaining selective and high affinity of binding and catalysis for the site studies, also contains an activatable reactive and environmentally sensitive reporter group. Furthermore, this probe has introduced only minor changes in the acylation kinetics, most likely due to its small size. One could thus expect to utilize these systems to attach to specific ligands which are exquisitely sensitive to alterations in structures (e.g., tetrodotoxin) [15,16]. This is necessary in order to probe the dynamics of the macromolecular functional sites. In addition, the excited-state decay kinetics are sufficiently long so that F6rster singlet-singlet energy transfer measurements can be carried out. Preliminary investigations of the tetrodotoxin binding site on the fast sodium channel in excitable membranes reveal that the anthraniloyl and azido-anthraniloyl tetrodotoxin derivatives retain their very high affinity (nM) for the toxin site (KD = 3.2 nM for tetrodotoxin, KD = 16.8 nM for azido-anthraniloyl tetrodotoxin) [15,16]. These molecules will provide the necessary handles in order to characterize and purify the proteins involved at the ion-channel functional sites, to serve in distance measurements between functional sites of the ion channel and to describe the dynamic motions of the ion channel in the absence and presence of neurotoxins.

156 Acknowledgement This work was supported by grants from the Research Corporation and the Medical Research Council of Canada.

References 1 Knowles, J.R. (1972) Acc. Chem. Res. 5,155-160 2 Bayley, H. and Knowles, J.R. (1977) Methods Enzymol. 46, 69-114 3 Das, M. and Fox, C.F. (1979) Annu. Rev. Biophys. Bioeng. 8,165-193 4 Dreyfuss, G., Schwartz, K., Blout, E.R., Barrio, J.R., Liu, F. and Leonard, N.J. (1978) Proc. Natl. Acad. Sci. U.S.A. 75, 1199-1203 5 Dockter, M.E. (1979) J. Biol. Chem. 254, 2161-2164 6 Biochemical Fluorescence Concepts, (1975 and 1976) Vols. 1 and 2, ((;hen, R.F. and Edelhoch, H., eds.), Marcel Dekker, Inc., New York, NY 7 Hammes, G.G. (1980) in Protein-Protein Interactions, (Frieden, C. and Nichol, L.W., eds.), Wiley-Interscience, New York, NY

8 Haugland, R.P. and Stryer, L. (1967) Conformation of Biopolymers 1, pp. 321-335, Academic Press, New York 9 Spencer, R.D. and Weber, G. (1970) J. Chem. Phys. 52, 1654-1663 10 Lakowicz, J.R. and Cherek, H. (1980) J. Biol. Chem. 255,831-834 11 Lakowicz, J.R., Cherek, H. and Bevan, D.R. (1980) J. Biol. Chem. 255,4403-4406 12 Kraut, J. (1977) Annu. Rev. Biochem. 46,331-358 13 Barrio, J.R., Sattsangi, P.D., Gruber, B.A., Dammann, L.G. and Leonard, N.J. (1976) J. Am. Chem. Soc. 98, 7408-7414 14 Grinvald, A. and Steinberg, I.Z. (1974) Biochemistry 13, 5107-5177 15 Ross, N. and Angelides, K.J. (1980) Am. Chem. Soc. Div. Biol. Chem. 95A 16 Angelides, K.J. (1981) Biochemistry 20, in the press 17 Shinitsky, M. (1972)J. Chem. Phys. 56, 5979-5984 18 Scott, T.H., Spencer, R.D., Leonard, N.J. and Weber, G. (1970) J. Am. Chem. Soc. 92,687-692 19 Bernhard, S.A., Lee, B.F. and Tashijian, Z.H. (1966) J. Mol. Biol. 18,405-418 20 Chen, R.F. (1967) Anal. Biochem. 19,374-380 21 Melhuish, W.H. (1962) J. Opt. Soc. Am. 52, 1256-1261