- Email: [email protected]
A critical histidine in the vesicular acetylcholine transporter
Neurochemistry International 36 (2000) 113±117
www.elsevier.com/locate/neuint
A critical histidine in the vesicular acetylcholine transporter James E. Keller, Stanley M. Parsons* Department of Chemistry and The Neuroscience Research Institute, University of California, Santa Barbara, CA 93106, USA Received 15 April 1999; received in revised form 4 July 1999; accepted 25 August 1999
Abstract The role of proton binding sites in the vesicular acetylcholine transporter was investigated by characterization of the pH dependence for the binding of [3H]vesamicol [(ÿ)-trans-2-(4-phenylpiperidino)cyclohexanol] to Torpedo synaptic vesicles. A single proton binds to a site with pKa 7.12 0.1, which is characteristic of histidine, to competitively inhibit vesamicol binding. The histidine-selective reagent diethylpyrocarbonate causes time-dependent inhibition of [3H]vesamicol binding with a rate constant only about 20-fold lower than for reaction with free histidine. Because its pH titration has a simple, ideal shape, this residue probably controls all pH eects in the transporter between pH 6±8. Inhibition of [3H]vesamicol binding by diethylpyrocarbonate was slowed by vesamicol but not acetylcholine, which binds to a separate site. The data suggest that a critical histidine with a pKa of 7.1 is unhindered when reacting with diethylpyrocarbonate. A conformational model for the histidine is proposed to explain why acetylcholine competes with protons but not with diethylpyrocarbonate. A conserved histidine in transmembrane helix VIII possibly is the histidine detected here. # 2000 Published by Elsevier Science Ltd. All rights reserved.
1. Introduction Although not commonly perceived as such, the proton is a type of modi®cation reagent that binds to a relatively limited range of sites in proteins. Protons are particularly important to the vesicular acetylcholine transporter (VAChT), which carries out storage of acetylcholine (ACh) by synaptic vesicles in cholinergic nerve terminals. VAChT exchanges two vesicular protons for one cytoplasmic ACh in each transport cycle (Nguyen et al., 1998), presumably by binding the protons and ACh to speci®c sites. VAChT also binds the compound vesamicol [(ÿ)-trans-2-(4-phenylpiperidino)cyclohexanol], which inhibits transport. Proton binding sites in VAChT were investigated in the current study. An earlier study found that binding of [3H]vesamicol was inhibited by protons with an apparent pKa of 6.26 (Kornreich and Parsons, 1988). This work had not determined the true pKa value for the inhibition and whether protons are competitive, * Corresponding author. Tel.: +1-805-893-2252; fax:+1-805-8934120. E-mail address: [email protected] (S.M. Parsons).
uncompetitive or noncompetitive with respect to vesamicol binding. These features were determined in the current study. Because the true pKa value is consistent with a histidine, the eect of a histidine-selective modi®cation reagent on [3H]vesamicol binding also was determined in the absence and presence of ACh and vesamicol. Taken together, the results suggest the presence of a histidine that controls VAChT conformation.
2. Experimental procedures 2.1. Materials VP1 synaptic vesicles from Torpedo californica electric organ were isolated as described (Gracz and Parsons, 1996). Brie¯y, the isolation involves homogenization of tissue, dierential sedimentation velocity pelleting onto a sucrose/Ficoll cushion, equilibrium buoyant-density banding in sucrose, and size exclusion chromatography on Sephacryl S-1000 in column buer (0.80 M glycine, 1 mM EDTA, 1 mM EGTA and 10 mM HEPES adjusted to pH 7.0 with
0197-0186/00/$ - see front matter # 2000 Published by Elsevier Science Ltd. All rights reserved. PII: S 0 1 9 7 - 0 1 8 6 ( 9 9 ) 0 0 1 1 0 - 2
114
J.E. Keller, S.M. Parsons / Neurochemistry International 36 (2000) 113±117
KOH). Animal use protocols were approved by the local Animal Care Council and meet the guidelines of the US Public Health Service and Department of Agriculture. Puri®ed vesicles were concentrated to about 1 mg protein/ml using a Centricon centrifugal ultra®ltration device (Amicon Corp.). Vesicular protein was determined by the method of Bradford (1976) using bovine serum albumin as standard. [3H]Vesamicol (20 Ci/mmol) was obtained from New England Nuclear, Inc. Diethylpyrocarbonate (DEPC) was obtained from Sigma Chemical Company. All other reagents were from usual commercial sources.
2.2. Proton competition Vesicles were diluted 25-fold into titration buer containing 0.70 M glycine, 1 mM EDTA, 1 mM EGTA, and 100 mM HEPES adjusted to dierent pH values with KOH. In pH 5.00 and 5.50 solutions, 5 mM 2-(N-morpholino)ethanesulfonic acid also was present. The vesicular suspensions (270 ml) were mixed with [3H]vesamicol stock solutions (30 ml) in 0.80 M glycine. Vesicles were incubated 1 h at 238C, and the pH was measured with a glass electrode. Three 95 ml portions of the suspension were ®ltered using vacuum assistance through glass ®ber ®lters (1.2 cm Whatman GF/F) coated with 0.5% polyethylenimine. Unbound vesamicol was washed rapidly through each ®lter using three one ml portions of ice-cold pH 7.8 titration buffer. Filter-bound radioactivity was determined as described below. Total [3H]vesamicol in 5 ml portions of each incubation mixture was determined. Free [3H]vesamicol in solution was calculated by subtracting the amount of bound vesamicol from the total vesamicol present in each sample.
2.3. Reaction with DEPC Vesicles were diluted 25-fold into sucrose buer (0.30 M sucrose, 50 mM HEPES, 150 mM NaCl, 0.5 mM EDTA, and 0.5 mM EGTA adjusted to pH 6.7 with NaOH) that in some cases also included ACh or [3H]vesamicol. When ACh was used, ACh esterase was inhibited by prior incubation of vesicles in 0.1 mM paraoxon for 30 min at 238C. DEPC (0.80 M in ethanol) was diluted 250-fold into the vesicle suspension in order to initiate reaction. To terminate the reaction, 20 ml of vesicles was mixed with 180 ml of [3H]vesamicol (440 nM) in pH 7.8 titration buer at 238C at the indicated times and equilibrated for 30 s. Two 90 ml portions were ®ltered separately and washed as described above. Control vesicles were treated similarly without DEPC.
Fig. 1. Eect of protons on vesamicol binding. Puri®ed synaptic vesicles (40 mg/ml) were incubated with dierent concentrations of [3H]vesamicol at pH 5.00 (Q), 5.50 (*), 6.05 (r), 6.70 (q), 7.00 (w), 7.65 (W), 8.25 (r) and 8.80 (R), and the amount of bound [3H]vesamicol was determined. The following competitive model was ®t to the untransformed data by simultaneous nonlinear regression: b Bmax vesf = vesf Kv 1 H =Ka where Bmax refers to total vesamicol binding capacity, b is bound vesamicol in pmol/mg, [ves]f is the concentration of free vesamicol, Kv is the pH-independent dissociation constant for vesamicol, and Ka is the acid dissociation constant, respectively. The ®t for the competitive model is shown in the double reciprocal form. The following parameter values were obtained: pKa=7.120.1, Bmax=25023 pmol/mg, and Kv=7.020.7 nM.
2.4. Determination of bound [3H]vesamicol and regression analysis Wet ®lters were placed in scintillation cocktail (ICN Biomedicals) and tritium decay was measured by liquid scintillation spectrometry at 44% eciency. Nonspeci®c binding of [3H]vesamicol was determined in the presence of 100-fold excess of nonradioactive vesamicol relative to total [3H]vesamicol for every experiment. Nonspeci®c binding (3±11% of total binding) was subtracted from the total binding to yield the speci®c binding presented. Replicate data were averaged. Regression analysis was carried out with the computer program Scientist (Micromath Scienti®c Software, Salt Lake City, UT, USA). Errors quoted are 21 standard deviation.
3. Results 3.1. Inhibition of binding by protons The mechanism by which protons inhibit vesamicol binding was determined by competing dierent concentrations of protons against dierent concentrations of [3H]vesamicol. Torpedo VAChT is stable over the pH range utilized (Kornreich and Parsons, 1988). The data are ®t very well by the competitive model in which a
J.E. Keller, S.M. Parsons / Neurochemistry International 36 (2000) 113±117
Fig. 2. Eect of DEPC on vesamicol binding. Puri®ed synaptic vesicles (30 mg protein per ml) at pH 6.70 were incubated in the absence or presence of 400 nM [3H]vesamicol or 300 mM ACh chloride for 15 min. DEPC was added to 1.6 mM to initiate reaction and [3H]vesamicol binding was assayed periodically. Exponential decay curves were ®t simultaneously to the binding data so as to have common values for vesamicol bound at time 0 and in®nity. The resulting rate constants were 0.12420.014 minÿ1 (*), 0.01520.002 minÿ1 (presence of [3H]vesamicol, Q), and 0.11820.020 minÿ1 (presence of ACh, q).
single bound proton exhibiting a pKa value of 7.1 2 0.1 blocks vesamicol binding that exhibits a Kv value of 7.0 2 0.7 nM (Fig. 1). The uncompetitive model ®t much worse (F-test, p < 0.01), and the noncompetitive model, which contains one more adjustable parameter, did not ®t the data signi®cantly better (F-test, p > 0.95). Because the equation for competitive inhibition contains the ®rst power of [H+], a single bound proton completely blocks the vesamicol binding site. 3.2. Inhibition of binding by DEPC A pKa value of 7.1 suggests a histidine residue. If this is the case, then modifying the histidine with DEPC, which ®rst monocarbethoxylates and then dicarbethoxylates imidazole (Roosemont, 1978), should aect vesamicol binding. DEPC is selective for histidine in proteins between pH 5.5 and 7.5 (Setlow and Mansour, 1970). To test for critical histidine, vesicles were incubated in DEPC at pH 6.70, and vesamicol binding was determined at the times indicated (Fig. 2). DEPC inhibited 98% of the binding with a pseudo ®rst-order rate constant of 0.124 20.014 minÿ1. The rate slowed about 8-fold in nearly saturating vesamicol and was unaected in nearly saturating ACh (Fig. 2). Thus, vesamicol but not ACh either binds close to or induces a conformational change in the DEPC-reactive site. 3.3. Other tests for histidine and cysteine interactions Cysteine residues have been shown to be linked to
115
the vesamicol binding site in VAChT from Torpedo californica (Kornreich and Parsons, 1988; J.E.K., unpublished observations). Pretreatment of vesicles with the cysteine-speci®c reagents methylmercury chloride or methylmethanethiosulfonate did not aect the inhibition of vesamicol binding by protons or DEPC (data not shown). Conversely, pretreatment of vesicles with the large cysteine-speci®c reagent p-chloromercuriphenylsulfonate did not protect from DEPC, as binding could not be recovered with 2-mercaptoethanol like it is without DEPC modi®cation (data not shown). Finally, hydroxylamine was used in an attempt to reverse inhibition by DEPC. Monocarbethoxylation but not dicarbethoxylation of histidine can be reversed with this reagent (Elodi, 1972; Miles, 1977). Dicarbethoxylated histidine is the normal product of sterically unhindered histidine. Hydroxylamine under conditions (60 mM at pH 6.70) that regenerate monocarbethoxylated histidine did not restore vesamicol binding, suggesting that dicarbethoxylation had occurred. Hydroxylamine had no eect on vesamicol binding to unmodi®ed VAChT (data not shown). 4. Discussion 4.1. Related results and interpretations A single proton binds competitively with respect to vesamicol. The pKa value of 7.1 2 0.1 is consistent with a histidyl side chain. Vesamicol binding previously was reported to be pH-dependent, but a signi®cantly lower apparent pKa was observed. This occurred because a single high concentration of [3H]vesamicol (1 mM) was utilized in the measurement, causing a drop in the observed pKa due to competition. Vesamicol binding also is inhibited by the histidineselective reagent DEPC. To eliminate the possibility that an unprotonated cysteine mediates inhibition by protons and DEPC, cysteines were blocked in several ways. Inhibition by protons was unaltered, indicating that cysteine is not involved in the pH eect. The only other protein groups that sometimes react with diethylpyrocarbonate are the a-amino group, and lysine and tyrosine side chains (Muhlrad et al., 1966; Melchior and Fahrney, 1970; Miles, 1977). The ®rst two groups are not likely involved because trinitrobenzenesulfonate, which reacts with amines, does not inhibit vesamicol binding (Kornreich and Parsons, 1988). Tyrosine also is not likely involved because it requires higher pH to react (Abe and Anan, 1976) and is easily regenerated with hydroxylamine (Osterman-Golkar et al., 1974). Both the pH and DEPC data favor a critical histidine. We assign it pKa 7.1.
116
J.E. Keller, S.M. Parsons / Neurochemistry International 36 (2000) 113±117
The pseudo ®rst-order rate constant for the DEPC reaction can be converted to a second-order constant by dividing by the DEPC concentration. The result is 103 Mÿ1 minÿ1. Reaction of free histidine with DEPC occurs at 1200 Mÿ1 minÿ1 at pH 6.0 and 208 (Wallis and Holbrook, 1973). Because only the unprotonated form of the histidine side chain reacts (Holbrook and Ingram, 1973), the rate constants must be corrected for dierent pKa values to compare them. After accounting for this and the small temperature dierence, we conclude that the critical unprotonated histidine in VAChT reacts about 20-fold more slowly than unprotonated free histidine does. Thus, the critical histidine in VAChT is only slightly hindered sterically in its unprotonated state. This is consistent with dicarbethoxylation of the histidine and its nearly normal pKa value. When a site with a similar pKa is protonated on the outside of VAChT, transport is inhibited (Anderson et al., 1982; Nguyen et al., 1998). Equilibrium binding of a high-anity analogue of ACh also is inhibited by protonation with pKa 7.4 20.3 (Nguyen et al., 1998). Thus, the pH-dependence of vesamicol binding, ACh transport and ACh analogue binding have revealed one or more important protonation events around neutral pH. Do these pH-dependencies arise from protonation of the same site or of dierent sites? Because ACh and vesamicol binding compete with each other (Bahr et al., 1992), inhibitory protonation at one site would be felt at the other site. Complicated pH-dependencies would result if several separate but linked protonation events occurred near neutrality. Because the steepness and shape of the pH titration curves are simple in all cases, a single critical histidine apparently mediates all of the pH-dependence around neutrality. As the pKa value determined here is the most accurate one available, we will use pKa 7.1 2 0.1 for this site in the future. 4.2. Implications lead to a conformational model for the critical histidine Because of the thermodynamic cycle present in an equilibrium, when a proton inhibits binding of ACh, ACh must inhibit binding of the proton. Whether the inhibition is direct or indirect does not matter. In contrast to the inhibitory relationship between ACh and protons, ACh has no eect on reaction with DEPC. The dierence implies that the protonated and unprotonated forms of the critical histidine have dierent orientations. This is because a proton and a DEPC molecule must initially approach the same unprotonated, nucleophilic nitrogen in order to react with it. If the conformation of the site were static, ACh would not likely block proton binding while having no eect on DEPC ``binding''. However, if unprotonated histi-
dine has an orientation dierent from that of protonated histidine, the contrasting eects of ACh can be understood. We propose that a conformational change in VAChT is controlled by the protonation state of this histidine. The proposal is consistent with evidence that the vesamicol and ACh binding sites are not identical to each other because (1) vesamicol and ACh have dierent structure-activity relationships (Rogers and Parsons, 1989; Rogers et al., 1989), (2) recent mutational evidence has dissociated the two sites (Kim et al., 1999; Varoqui and Erickson, 1998), and (3) phosphorylation blocks vesamicol binding but not ACh transport (Barbosa et al., 1997; Clarizia et al., 1998, 1999). A conformational change in VAChT triggered by protonation of this histidine could explain simultaneous disruption of the separate binding sites. The role of such a conformational change is unknown; it could be important in active transport or an undescribed function. In either case, it probably would be conserved. There are ®ve histidines in VAChT from Torpedo californica (numbers 323, 429, 494, 501 and 504; S.M.P. and D. Bravo, unpublished results). Residue 323 in putative transmembrane helix VIII is conserved in all VAChTs (Erickson et al., 1996; Usdin et al., 1995). It is a good candidate for the critical histidine. Acknowledgements We thank Barry Sanchez, Heather Johnson, and Ricardo Garcia for the isolation of synaptic vesicles. This research was supported by grant NS15047 from the National Institute of Neurological Disorders and Stroke. References Abe, K., Anan, F.K., 1976. The chemical modi®cation of beef liver catalase. V. Ethoxyformylation of histidine and tyrosine residues of catalase with diethylpyrocarbonate. J. Biochem. 80, 229±237. Anderson, D.C., King, S.C., Parsons, S.M., 1982. Proton gradient linkage to active uptake of [3H]acetylcholine by Torpedo electric organ synaptic vesicles. Biochemistry 21, 3037±3043. Bahr, B.A., Clarkson, E.D., Rogers, G.A., Noremberg, K., Parsons, S.M., 1992. A kinetic and allosteric model for the acetylcholine transporter-vesamicol receptor in synaptic vesicles. Biochemistry 31, 5752±5762. Barbosa Jr, J., Clarizia, A.D., Gomez, M.V., Romano-Silva, M.A., Prado, V.F., Prado, M.A.M., 1997. Eect of protein kinase C activation on the release of [3H]acetylcholine in the presence of vesamicol. J. Neurochem. 69, 2608±2611. Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248±254. Clarizia, A.D., Romano-Silva, M.A., Prado, V.F., Gomez, M.V., Prado, M.A., 1998. Role of protein kinase C in the release of
J.E. Keller, S.M. Parsons / Neurochemistry International 36 (2000) 113±117 [3H]acetylcholine from myenteric plexus treated with vesamicol. Neurosci. Lett. 244, 115±117. Clarizia, A.D., Gomez, M.V., Romano-Silva, M.A., Parsons, S.M., Prado, V.F., Prado, M.A.M., 1999. Control of the binding of a vesamicol-analogue to the vesicular acetylcholine transporter. NeuroReport 10, 2783±2787. Elodi, P., 1972. Role of histidyl residues in the activity of porcine pancreatic amylase. Acta Biochim. Biophys. Acad. Sci. Hung. 7, 241±245. Erickson, J.D., Weihe, E., Schafer, M.K., Neale, E., Williamson, L., Bonner, T.I., Tao-Cheng, J.H., Eiden, L.E., 1996. The VAChT/ ChAT ``cholinergic gene locus'': new aspects of genetic and vesicular regulation of cholinergic function. Prog. Brain Res. 109, 69±82. Gracz, L.M., Parsons, S.M., 1996. Puri®cation of active synaptic vesicles from the electric organ of Torpedo californica and comparison to reserve vesicles. Biochim. Biophys. Acta 1292, 293±302. Holbrook, J.J., Ingram, V.A., 1973. Ionic properties of an essential histidine residue in pig heart lactate dehydrogenase. Biochem. J. 131, 729±738. Kim, M.-H., Lu, M., Lim, E.-J., Chai, Y.-G., Hersh, L.B., 1999. Mutational analysis of aspartate residues in the transmembrane regions and cytoplasmic loops of rat vesicular acetylcholine transporter. J. Biol. Chem. 274, 673±680. Kornreich, W.D., Parsons, S.M., 1988. Sidedness and chemical and kinetic properties of the vesamicol receptor of cholinergic synaptic vesicles. Biochemistry 27, 5262±5267. Melchior, W.B., Fahrney, D., 1970. Ethoxyformylation of proteins. Reaction of ethoxyformic anhydride with a-chymotrypsin, pepsin and pancreatic ribonuclease at pH 4. Biochemistry 9, 251±258. Miles, E.W., 1977. Modi®cation of histidyl residues in proteins by diethylpyrocarbonate. Methods Enzymol. 47, 431±442. Muhlrad, A., Hegyi, G., Toth, G., 1966. Eect of diethylpyrocarbo-
117
nate on proteins. Acta Biochim. Biophys. Acad. Sci. Hung. 2, 19± 29. Nguyen, M.L., Cox, G.D., Parsons, S.M., 1998. Kinetics parameters for the vesicular acetylcholine transporter: two protons exchange for one acetylcholine. Biochemistry 37, 13400±13410. Osterman-Golkar, S., Ehrenberg, L., Solymosy, F., 1974. Reaction of diethylpyrocarbonate with nucleophiles. Acta Chem. Scand. Ser. B 28, 215±220. Rogers, G.A., Parsons, S.M., 1989. Inhibition of acetylcholine storage by acetylcholine analogs in vitro. Molec. Pharmacol. 36, 333±341. Rogers, G.A., Parsons, S.M., Anderson, D.C., Nilsson, L.M., Bahr, B.A., Kornreich, W.D., Kaufman, R., Jacobs, R.S., Kirtman, B., 1989. Synthesis, in vitro acetylcholine storage-blocking activities, and biological properties of derivatives and analogues of trans-2(4-phenylpiperdino)cyclohexanol (vesamicol). J. Med. Chem. 32, 1217±1230. Roosemont, J.L., 1978. Reaction of histidine residues in proteins with diethylpyrocarbonate: dierential molar absorptivities and reactivities. Anal. Biochem. 88, 314±320. Setlow, B., Mansour, T.E., 1970. Studies on heart phosphofructokinase. Nature of the enzyme desensitized to allosteric control by photo-oxidation and by acylation with ethoxyformic anhydride. J. Biol. Chem. 245, 5524±5533. Usdin, T.B., Eiden, L.E., Bonner, T.I., Erickson, J.D., 1995. Molecular biology of the vesicular ACh transporter. Trends Neurosci. 18, 218±224. Varoqui, H., Erickson, J.D., 1998. Dissociation of the vesicular acetylcholine transporter domains important for high-anity transport recognition, binding of vesamicol and targeting to synaptic vesicles. J. Physiol. (Paris) 92, 141±144. Wallis, R.B., Holbrook, J.J., 1973. The reaction of a histidine residue in glutamate dehydrogenase with diethylpyrocarbonate. Biochem. J. 133, 183±187.
www.elsevier.com/locate/neuint
A critical histidine in the vesicular acetylcholine transporter James E. Keller, Stanley M. Parsons* Department of Chemistry and The Neuroscience Research Institute, University of California, Santa Barbara, CA 93106, USA Received 15 April 1999; received in revised form 4 July 1999; accepted 25 August 1999
Abstract The role of proton binding sites in the vesicular acetylcholine transporter was investigated by characterization of the pH dependence for the binding of [3H]vesamicol [(ÿ)-trans-2-(4-phenylpiperidino)cyclohexanol] to Torpedo synaptic vesicles. A single proton binds to a site with pKa 7.12 0.1, which is characteristic of histidine, to competitively inhibit vesamicol binding. The histidine-selective reagent diethylpyrocarbonate causes time-dependent inhibition of [3H]vesamicol binding with a rate constant only about 20-fold lower than for reaction with free histidine. Because its pH titration has a simple, ideal shape, this residue probably controls all pH eects in the transporter between pH 6±8. Inhibition of [3H]vesamicol binding by diethylpyrocarbonate was slowed by vesamicol but not acetylcholine, which binds to a separate site. The data suggest that a critical histidine with a pKa of 7.1 is unhindered when reacting with diethylpyrocarbonate. A conformational model for the histidine is proposed to explain why acetylcholine competes with protons but not with diethylpyrocarbonate. A conserved histidine in transmembrane helix VIII possibly is the histidine detected here. # 2000 Published by Elsevier Science Ltd. All rights reserved.
1. Introduction Although not commonly perceived as such, the proton is a type of modi®cation reagent that binds to a relatively limited range of sites in proteins. Protons are particularly important to the vesicular acetylcholine transporter (VAChT), which carries out storage of acetylcholine (ACh) by synaptic vesicles in cholinergic nerve terminals. VAChT exchanges two vesicular protons for one cytoplasmic ACh in each transport cycle (Nguyen et al., 1998), presumably by binding the protons and ACh to speci®c sites. VAChT also binds the compound vesamicol [(ÿ)-trans-2-(4-phenylpiperidino)cyclohexanol], which inhibits transport. Proton binding sites in VAChT were investigated in the current study. An earlier study found that binding of [3H]vesamicol was inhibited by protons with an apparent pKa of 6.26 (Kornreich and Parsons, 1988). This work had not determined the true pKa value for the inhibition and whether protons are competitive, * Corresponding author. Tel.: +1-805-893-2252; fax:+1-805-8934120. E-mail address: [email protected] (S.M. Parsons).
uncompetitive or noncompetitive with respect to vesamicol binding. These features were determined in the current study. Because the true pKa value is consistent with a histidine, the eect of a histidine-selective modi®cation reagent on [3H]vesamicol binding also was determined in the absence and presence of ACh and vesamicol. Taken together, the results suggest the presence of a histidine that controls VAChT conformation.
2. Experimental procedures 2.1. Materials VP1 synaptic vesicles from Torpedo californica electric organ were isolated as described (Gracz and Parsons, 1996). Brie¯y, the isolation involves homogenization of tissue, dierential sedimentation velocity pelleting onto a sucrose/Ficoll cushion, equilibrium buoyant-density banding in sucrose, and size exclusion chromatography on Sephacryl S-1000 in column buer (0.80 M glycine, 1 mM EDTA, 1 mM EGTA and 10 mM HEPES adjusted to pH 7.0 with
0197-0186/00/$ - see front matter # 2000 Published by Elsevier Science Ltd. All rights reserved. PII: S 0 1 9 7 - 0 1 8 6 ( 9 9 ) 0 0 1 1 0 - 2
114
J.E. Keller, S.M. Parsons / Neurochemistry International 36 (2000) 113±117
KOH). Animal use protocols were approved by the local Animal Care Council and meet the guidelines of the US Public Health Service and Department of Agriculture. Puri®ed vesicles were concentrated to about 1 mg protein/ml using a Centricon centrifugal ultra®ltration device (Amicon Corp.). Vesicular protein was determined by the method of Bradford (1976) using bovine serum albumin as standard. [3H]Vesamicol (20 Ci/mmol) was obtained from New England Nuclear, Inc. Diethylpyrocarbonate (DEPC) was obtained from Sigma Chemical Company. All other reagents were from usual commercial sources.
2.2. Proton competition Vesicles were diluted 25-fold into titration buer containing 0.70 M glycine, 1 mM EDTA, 1 mM EGTA, and 100 mM HEPES adjusted to dierent pH values with KOH. In pH 5.00 and 5.50 solutions, 5 mM 2-(N-morpholino)ethanesulfonic acid also was present. The vesicular suspensions (270 ml) were mixed with [3H]vesamicol stock solutions (30 ml) in 0.80 M glycine. Vesicles were incubated 1 h at 238C, and the pH was measured with a glass electrode. Three 95 ml portions of the suspension were ®ltered using vacuum assistance through glass ®ber ®lters (1.2 cm Whatman GF/F) coated with 0.5% polyethylenimine. Unbound vesamicol was washed rapidly through each ®lter using three one ml portions of ice-cold pH 7.8 titration buffer. Filter-bound radioactivity was determined as described below. Total [3H]vesamicol in 5 ml portions of each incubation mixture was determined. Free [3H]vesamicol in solution was calculated by subtracting the amount of bound vesamicol from the total vesamicol present in each sample.
2.3. Reaction with DEPC Vesicles were diluted 25-fold into sucrose buer (0.30 M sucrose, 50 mM HEPES, 150 mM NaCl, 0.5 mM EDTA, and 0.5 mM EGTA adjusted to pH 6.7 with NaOH) that in some cases also included ACh or [3H]vesamicol. When ACh was used, ACh esterase was inhibited by prior incubation of vesicles in 0.1 mM paraoxon for 30 min at 238C. DEPC (0.80 M in ethanol) was diluted 250-fold into the vesicle suspension in order to initiate reaction. To terminate the reaction, 20 ml of vesicles was mixed with 180 ml of [3H]vesamicol (440 nM) in pH 7.8 titration buer at 238C at the indicated times and equilibrated for 30 s. Two 90 ml portions were ®ltered separately and washed as described above. Control vesicles were treated similarly without DEPC.
Fig. 1. Eect of protons on vesamicol binding. Puri®ed synaptic vesicles (40 mg/ml) were incubated with dierent concentrations of [3H]vesamicol at pH 5.00 (Q), 5.50 (*), 6.05 (r), 6.70 (q), 7.00 (w), 7.65 (W), 8.25 (r) and 8.80 (R), and the amount of bound [3H]vesamicol was determined. The following competitive model was ®t to the untransformed data by simultaneous nonlinear regression: b Bmax vesf = vesf Kv 1 H =Ka where Bmax refers to total vesamicol binding capacity, b is bound vesamicol in pmol/mg, [ves]f is the concentration of free vesamicol, Kv is the pH-independent dissociation constant for vesamicol, and Ka is the acid dissociation constant, respectively. The ®t for the competitive model is shown in the double reciprocal form. The following parameter values were obtained: pKa=7.120.1, Bmax=25023 pmol/mg, and Kv=7.020.7 nM.
2.4. Determination of bound [3H]vesamicol and regression analysis Wet ®lters were placed in scintillation cocktail (ICN Biomedicals) and tritium decay was measured by liquid scintillation spectrometry at 44% eciency. Nonspeci®c binding of [3H]vesamicol was determined in the presence of 100-fold excess of nonradioactive vesamicol relative to total [3H]vesamicol for every experiment. Nonspeci®c binding (3±11% of total binding) was subtracted from the total binding to yield the speci®c binding presented. Replicate data were averaged. Regression analysis was carried out with the computer program Scientist (Micromath Scienti®c Software, Salt Lake City, UT, USA). Errors quoted are 21 standard deviation.
3. Results 3.1. Inhibition of binding by protons The mechanism by which protons inhibit vesamicol binding was determined by competing dierent concentrations of protons against dierent concentrations of [3H]vesamicol. Torpedo VAChT is stable over the pH range utilized (Kornreich and Parsons, 1988). The data are ®t very well by the competitive model in which a
J.E. Keller, S.M. Parsons / Neurochemistry International 36 (2000) 113±117
Fig. 2. Eect of DEPC on vesamicol binding. Puri®ed synaptic vesicles (30 mg protein per ml) at pH 6.70 were incubated in the absence or presence of 400 nM [3H]vesamicol or 300 mM ACh chloride for 15 min. DEPC was added to 1.6 mM to initiate reaction and [3H]vesamicol binding was assayed periodically. Exponential decay curves were ®t simultaneously to the binding data so as to have common values for vesamicol bound at time 0 and in®nity. The resulting rate constants were 0.12420.014 minÿ1 (*), 0.01520.002 minÿ1 (presence of [3H]vesamicol, Q), and 0.11820.020 minÿ1 (presence of ACh, q).
single bound proton exhibiting a pKa value of 7.1 2 0.1 blocks vesamicol binding that exhibits a Kv value of 7.0 2 0.7 nM (Fig. 1). The uncompetitive model ®t much worse (F-test, p < 0.01), and the noncompetitive model, which contains one more adjustable parameter, did not ®t the data signi®cantly better (F-test, p > 0.95). Because the equation for competitive inhibition contains the ®rst power of [H+], a single bound proton completely blocks the vesamicol binding site. 3.2. Inhibition of binding by DEPC A pKa value of 7.1 suggests a histidine residue. If this is the case, then modifying the histidine with DEPC, which ®rst monocarbethoxylates and then dicarbethoxylates imidazole (Roosemont, 1978), should aect vesamicol binding. DEPC is selective for histidine in proteins between pH 5.5 and 7.5 (Setlow and Mansour, 1970). To test for critical histidine, vesicles were incubated in DEPC at pH 6.70, and vesamicol binding was determined at the times indicated (Fig. 2). DEPC inhibited 98% of the binding with a pseudo ®rst-order rate constant of 0.124 20.014 minÿ1. The rate slowed about 8-fold in nearly saturating vesamicol and was unaected in nearly saturating ACh (Fig. 2). Thus, vesamicol but not ACh either binds close to or induces a conformational change in the DEPC-reactive site. 3.3. Other tests for histidine and cysteine interactions Cysteine residues have been shown to be linked to
115
the vesamicol binding site in VAChT from Torpedo californica (Kornreich and Parsons, 1988; J.E.K., unpublished observations). Pretreatment of vesicles with the cysteine-speci®c reagents methylmercury chloride or methylmethanethiosulfonate did not aect the inhibition of vesamicol binding by protons or DEPC (data not shown). Conversely, pretreatment of vesicles with the large cysteine-speci®c reagent p-chloromercuriphenylsulfonate did not protect from DEPC, as binding could not be recovered with 2-mercaptoethanol like it is without DEPC modi®cation (data not shown). Finally, hydroxylamine was used in an attempt to reverse inhibition by DEPC. Monocarbethoxylation but not dicarbethoxylation of histidine can be reversed with this reagent (Elodi, 1972; Miles, 1977). Dicarbethoxylated histidine is the normal product of sterically unhindered histidine. Hydroxylamine under conditions (60 mM at pH 6.70) that regenerate monocarbethoxylated histidine did not restore vesamicol binding, suggesting that dicarbethoxylation had occurred. Hydroxylamine had no eect on vesamicol binding to unmodi®ed VAChT (data not shown). 4. Discussion 4.1. Related results and interpretations A single proton binds competitively with respect to vesamicol. The pKa value of 7.1 2 0.1 is consistent with a histidyl side chain. Vesamicol binding previously was reported to be pH-dependent, but a signi®cantly lower apparent pKa was observed. This occurred because a single high concentration of [3H]vesamicol (1 mM) was utilized in the measurement, causing a drop in the observed pKa due to competition. Vesamicol binding also is inhibited by the histidineselective reagent DEPC. To eliminate the possibility that an unprotonated cysteine mediates inhibition by protons and DEPC, cysteines were blocked in several ways. Inhibition by protons was unaltered, indicating that cysteine is not involved in the pH eect. The only other protein groups that sometimes react with diethylpyrocarbonate are the a-amino group, and lysine and tyrosine side chains (Muhlrad et al., 1966; Melchior and Fahrney, 1970; Miles, 1977). The ®rst two groups are not likely involved because trinitrobenzenesulfonate, which reacts with amines, does not inhibit vesamicol binding (Kornreich and Parsons, 1988). Tyrosine also is not likely involved because it requires higher pH to react (Abe and Anan, 1976) and is easily regenerated with hydroxylamine (Osterman-Golkar et al., 1974). Both the pH and DEPC data favor a critical histidine. We assign it pKa 7.1.
116
J.E. Keller, S.M. Parsons / Neurochemistry International 36 (2000) 113±117
The pseudo ®rst-order rate constant for the DEPC reaction can be converted to a second-order constant by dividing by the DEPC concentration. The result is 103 Mÿ1 minÿ1. Reaction of free histidine with DEPC occurs at 1200 Mÿ1 minÿ1 at pH 6.0 and 208 (Wallis and Holbrook, 1973). Because only the unprotonated form of the histidine side chain reacts (Holbrook and Ingram, 1973), the rate constants must be corrected for dierent pKa values to compare them. After accounting for this and the small temperature dierence, we conclude that the critical unprotonated histidine in VAChT reacts about 20-fold more slowly than unprotonated free histidine does. Thus, the critical histidine in VAChT is only slightly hindered sterically in its unprotonated state. This is consistent with dicarbethoxylation of the histidine and its nearly normal pKa value. When a site with a similar pKa is protonated on the outside of VAChT, transport is inhibited (Anderson et al., 1982; Nguyen et al., 1998). Equilibrium binding of a high-anity analogue of ACh also is inhibited by protonation with pKa 7.4 20.3 (Nguyen et al., 1998). Thus, the pH-dependence of vesamicol binding, ACh transport and ACh analogue binding have revealed one or more important protonation events around neutral pH. Do these pH-dependencies arise from protonation of the same site or of dierent sites? Because ACh and vesamicol binding compete with each other (Bahr et al., 1992), inhibitory protonation at one site would be felt at the other site. Complicated pH-dependencies would result if several separate but linked protonation events occurred near neutrality. Because the steepness and shape of the pH titration curves are simple in all cases, a single critical histidine apparently mediates all of the pH-dependence around neutrality. As the pKa value determined here is the most accurate one available, we will use pKa 7.1 2 0.1 for this site in the future. 4.2. Implications lead to a conformational model for the critical histidine Because of the thermodynamic cycle present in an equilibrium, when a proton inhibits binding of ACh, ACh must inhibit binding of the proton. Whether the inhibition is direct or indirect does not matter. In contrast to the inhibitory relationship between ACh and protons, ACh has no eect on reaction with DEPC. The dierence implies that the protonated and unprotonated forms of the critical histidine have dierent orientations. This is because a proton and a DEPC molecule must initially approach the same unprotonated, nucleophilic nitrogen in order to react with it. If the conformation of the site were static, ACh would not likely block proton binding while having no eect on DEPC ``binding''. However, if unprotonated histi-
dine has an orientation dierent from that of protonated histidine, the contrasting eects of ACh can be understood. We propose that a conformational change in VAChT is controlled by the protonation state of this histidine. The proposal is consistent with evidence that the vesamicol and ACh binding sites are not identical to each other because (1) vesamicol and ACh have dierent structure-activity relationships (Rogers and Parsons, 1989; Rogers et al., 1989), (2) recent mutational evidence has dissociated the two sites (Kim et al., 1999; Varoqui and Erickson, 1998), and (3) phosphorylation blocks vesamicol binding but not ACh transport (Barbosa et al., 1997; Clarizia et al., 1998, 1999). A conformational change in VAChT triggered by protonation of this histidine could explain simultaneous disruption of the separate binding sites. The role of such a conformational change is unknown; it could be important in active transport or an undescribed function. In either case, it probably would be conserved. There are ®ve histidines in VAChT from Torpedo californica (numbers 323, 429, 494, 501 and 504; S.M.P. and D. Bravo, unpublished results). Residue 323 in putative transmembrane helix VIII is conserved in all VAChTs (Erickson et al., 1996; Usdin et al., 1995). It is a good candidate for the critical histidine. Acknowledgements We thank Barry Sanchez, Heather Johnson, and Ricardo Garcia for the isolation of synaptic vesicles. This research was supported by grant NS15047 from the National Institute of Neurological Disorders and Stroke. References Abe, K., Anan, F.K., 1976. The chemical modi®cation of beef liver catalase. V. Ethoxyformylation of histidine and tyrosine residues of catalase with diethylpyrocarbonate. J. Biochem. 80, 229±237. Anderson, D.C., King, S.C., Parsons, S.M., 1982. Proton gradient linkage to active uptake of [3H]acetylcholine by Torpedo electric organ synaptic vesicles. Biochemistry 21, 3037±3043. Bahr, B.A., Clarkson, E.D., Rogers, G.A., Noremberg, K., Parsons, S.M., 1992. A kinetic and allosteric model for the acetylcholine transporter-vesamicol receptor in synaptic vesicles. Biochemistry 31, 5752±5762. Barbosa Jr, J., Clarizia, A.D., Gomez, M.V., Romano-Silva, M.A., Prado, V.F., Prado, M.A.M., 1997. Eect of protein kinase C activation on the release of [3H]acetylcholine in the presence of vesamicol. J. Neurochem. 69, 2608±2611. Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248±254. Clarizia, A.D., Romano-Silva, M.A., Prado, V.F., Gomez, M.V., Prado, M.A., 1998. Role of protein kinase C in the release of
J.E. Keller, S.M. Parsons / Neurochemistry International 36 (2000) 113±117 [3H]acetylcholine from myenteric plexus treated with vesamicol. Neurosci. Lett. 244, 115±117. Clarizia, A.D., Gomez, M.V., Romano-Silva, M.A., Parsons, S.M., Prado, V.F., Prado, M.A.M., 1999. Control of the binding of a vesamicol-analogue to the vesicular acetylcholine transporter. NeuroReport 10, 2783±2787. Elodi, P., 1972. Role of histidyl residues in the activity of porcine pancreatic amylase. Acta Biochim. Biophys. Acad. Sci. Hung. 7, 241±245. Erickson, J.D., Weihe, E., Schafer, M.K., Neale, E., Williamson, L., Bonner, T.I., Tao-Cheng, J.H., Eiden, L.E., 1996. The VAChT/ ChAT ``cholinergic gene locus'': new aspects of genetic and vesicular regulation of cholinergic function. Prog. Brain Res. 109, 69±82. Gracz, L.M., Parsons, S.M., 1996. Puri®cation of active synaptic vesicles from the electric organ of Torpedo californica and comparison to reserve vesicles. Biochim. Biophys. Acta 1292, 293±302. Holbrook, J.J., Ingram, V.A., 1973. Ionic properties of an essential histidine residue in pig heart lactate dehydrogenase. Biochem. J. 131, 729±738. Kim, M.-H., Lu, M., Lim, E.-J., Chai, Y.-G., Hersh, L.B., 1999. Mutational analysis of aspartate residues in the transmembrane regions and cytoplasmic loops of rat vesicular acetylcholine transporter. J. Biol. Chem. 274, 673±680. Kornreich, W.D., Parsons, S.M., 1988. Sidedness and chemical and kinetic properties of the vesamicol receptor of cholinergic synaptic vesicles. Biochemistry 27, 5262±5267. Melchior, W.B., Fahrney, D., 1970. Ethoxyformylation of proteins. Reaction of ethoxyformic anhydride with a-chymotrypsin, pepsin and pancreatic ribonuclease at pH 4. Biochemistry 9, 251±258. Miles, E.W., 1977. Modi®cation of histidyl residues in proteins by diethylpyrocarbonate. Methods Enzymol. 47, 431±442. Muhlrad, A., Hegyi, G., Toth, G., 1966. Eect of diethylpyrocarbo-
117
nate on proteins. Acta Biochim. Biophys. Acad. Sci. Hung. 2, 19± 29. Nguyen, M.L., Cox, G.D., Parsons, S.M., 1998. Kinetics parameters for the vesicular acetylcholine transporter: two protons exchange for one acetylcholine. Biochemistry 37, 13400±13410. Osterman-Golkar, S., Ehrenberg, L., Solymosy, F., 1974. Reaction of diethylpyrocarbonate with nucleophiles. Acta Chem. Scand. Ser. B 28, 215±220. Rogers, G.A., Parsons, S.M., 1989. Inhibition of acetylcholine storage by acetylcholine analogs in vitro. Molec. Pharmacol. 36, 333±341. Rogers, G.A., Parsons, S.M., Anderson, D.C., Nilsson, L.M., Bahr, B.A., Kornreich, W.D., Kaufman, R., Jacobs, R.S., Kirtman, B., 1989. Synthesis, in vitro acetylcholine storage-blocking activities, and biological properties of derivatives and analogues of trans-2(4-phenylpiperdino)cyclohexanol (vesamicol). J. Med. Chem. 32, 1217±1230. Roosemont, J.L., 1978. Reaction of histidine residues in proteins with diethylpyrocarbonate: dierential molar absorptivities and reactivities. Anal. Biochem. 88, 314±320. Setlow, B., Mansour, T.E., 1970. Studies on heart phosphofructokinase. Nature of the enzyme desensitized to allosteric control by photo-oxidation and by acylation with ethoxyformic anhydride. J. Biol. Chem. 245, 5524±5533. Usdin, T.B., Eiden, L.E., Bonner, T.I., Erickson, J.D., 1995. Molecular biology of the vesicular ACh transporter. Trends Neurosci. 18, 218±224. Varoqui, H., Erickson, J.D., 1998. Dissociation of the vesicular acetylcholine transporter domains important for high-anity transport recognition, binding of vesamicol and targeting to synaptic vesicles. J. Physiol. (Paris) 92, 141±144. Wallis, R.B., Holbrook, J.J., 1973. The reaction of a histidine residue in glutamate dehydrogenase with diethylpyrocarbonate. Biochem. J. 133, 183±187.