Chem. Res. Toxicol. 1992,5, 411-418
41 1
Inhibition of Acetylcholinesterase by Hemicholiniums, Conformationally Constrained Choline Analogues. Evaluation of Aryl and Alkyl Substituents. Comparisons with Choline and (3-hydroxy phenyl)trimet hylammonium Bong Ho Lee,t*tTerry C. Stelly,$J William J. Colucci,$*lJ. Gabriel Garcia,$?# Richard D. Gandour,*y$and Daniel M. Quinn**t Department of Chemistry, The University of Iowa, Iowa City, Iowa 52242,and Department of Chemistry, Louisiana State University, Baton Rouge, Louisiana 70803-1804 Received December 20, 1991
2-Substituted-2-hydroxy-4,4-dimethylmorpholiniums (hemicholiniums) inhibit acetylcholinesterase (EC 3.1.1.7)-catalyzed hydrolysis of acetylthiocholine (ATCh). The 4-substituted arenes [NH2,NHC(0)CH3, C1, CN, and NO2]have values of inhibition constants (Ki)that range from 220 to 3690 pM, which correlate with Hammett u, p = 0.8. The alkyl compounds, hydrogen, methyl, tert-butyl, and trifluoromethyl, have values of Ki of 550, 560, 1200, and 1200 pM, respectively. These values compare favorably with Ki = 960 pM for choline. The conformation of AChE-bound choline must be gauche to support our suggestion that hemicholiniums are conformationally constrained analogues of choline. (3-Hydroxypheny1)trimethylammonium(5) inhibits most strongly, K i= 0.21 pM, of the compounds examined in this study. The solvent = 0.83 f 0.04) suggests that inhibition by 5 involves hydrogen bonding. isotope effect (H20Ki/DflKi The binding by AChE of the hemicholiniums of various sizes and the strong binding of 5 support an earlier proposal [Schowen, K. B., Smissman, E. E., and Stephen, W. F., Jr. (1975) J. Med. Chem. 18,292-3001 that the active site of AChE has ample space for rotation about the C-C bond in choline. Compound 5, which has one more carbon between the hydroxy and trimethylammonium than does choline, inhibits much more potently than either choline or the hemicholiniums. Compound 5 provides a correct spacer to span the trimethylammonium recognition site and the esteratic site of AChE. This aromatic spacer interacts favorably with the hydrophobic active site, and the phenolic hydroxyl probably hydrogen bonds to the histidine in the esteratic site. Choline in any conformation and the hemicholiniums are too short to make a strong hydrogen bond. Acetylcholinesterase (AChE)l mediates the transmission of nerve impulses across nerve-nerve and neuromuscular synapses in the central and peripheral nervous systems (I). AChE hydrolyzes acetylcholine (ACh) to acetate and choline (2). The arrival of nerve impulses from the pre-
mt!i,!f, Chart I
H3C
CH3NHCO0
EH3
OH
H20
+
(:TcH3-
N-Me Me Me I \
(+ N-Me
+
AcO'
+
H+
/ \
MeMe
synaptic membrane leads to release of ACh into the synaptic cleft. ACh diffuses to the postsynaptic membrane, where it binds to an ACh-specific receptor. Then, AChE hydrolyzes this ACh to restore polarization of the postsynaptic membrane. Many researchers have studied AChE over the past fifty years because of national security, pharmacological, and toxicological reasons. AChE inhibitors serve diverse functions in chemical warfare, control of agricultural and household pests, and amelioration of human diseases. Due to the critical role of AChE in the nervous system, chemists continue to design and characterize AChE inhibitors. The recent report (3)of the atomic structure of 'The University of Iowa. Present address: University of Nebraska, Lincoln, NE. 5 Louisiana State University. 11 Undergraduate Intern. Present address: University of South Alabama Medical School, Mobile, AL. Present address: Ethyl Technical Center, Baton Rouge, LA. #Presentaddress: Lawrence Berkeley Laboratory, Berkeley, CA.
*
0893-228x/92/2705-0411$03.00/0
CH3
1
q,OC(O)NHC*,
+ N(CH3)3
1-
2
[pqoH qoH I \
Me Me
3
N(CH3)2
4
+ N(CH3)3
1-
5
AChE from Torpedo californica will stimulate further activity in this area. Reversible inhibitors that can enter the central nervous system might ameliorate the symptoms of Alzheimer's disease (4-6).As the disease progresses, cholinergic deficiency develops (7). Inhibition of AChE should increase the concentration of ACh in affected regions of the brain. The inhibitors, physostigmine (1; see Chart I) and choline, have improved memory processes in Abbreviations: AChE, acetylcholinesterase;ACh, acetylcholine; ATCh, acetylthiocholine;DMAE, 2-(N,N-dimethylamino)ethano~ DSS, 3-(trimethylsilyl)-l-propanesulfonate; THF, tetrahydrofuran; DTNB, 5,5'-dithiobis(2-nitrobenzoicacid).
0 1992 American Chemical Society
412 Chem. Res. Toxicol., Vol. 5, No.3, 1992
Lee et al. OH, NH), 1645 (m, NHC=O) cm-'. Anal. Calcd for Cl4H,,BrN,O3: C, 48.71; H, 6.13; N, 8.11. Found: C, 48.76; H, 6.22; N, 7.95.
(C) 4,4-Dimethyl-2-hydroxy-2-(4'-ammoniophenyl)-
both animals and humans (IO),though the effect of choline likely arises from more reasons than just inhibition of AChE. Ammonium ions inhibit AChE from various sources. The inhibition probably occurs when ammonium ions bind reversibly to the site on the enzyme where the quaternary ammonium part of ACh binds. (3-Hydroxypheny1)trialkylammonium derivatives, e.g., neostigmine (2), are potent reversible competitive inhibitors of AChE (11). The morpholinium derivative, 2-hydroxy-4,4-dimethyl-2phenylmorpholinium bromide (hemicholinium-15) (3), inhibits both AChE and butyrylcholinesterase (121,but less strongly than 1 and 2 (11). Choline analogues inhibit synaptosomal uptake of choline (12-17). In this report, we describe the inhibition of AChE-catalyzed hydrolysis of acetylthiocholine (ATCh) by 3 - ( N f l dimethy1amino)phenol (a), (3-hydroxypheny1)trimethylammonium iodide (5), and a series of hemicholiniums, 6 1 4 (see Chart 11), using the enzyme from Electrophorus electricus. We compare them with inhibition by choline. Finally, we discuss a correlation between inhibition and the substituent, R, at C2 of the morpholinium ring.
Experlmental Procedures Baton Rouge. (A) General Data. T H F was distilled from potassium metal. Hexane was distilled from CaH,, and EhO was distilled from sodium benzophenone ketyl. The alcohols were reagent grade, as was acetone, which was stored over 4-A molecular sieves. Acetone was monobrominated by the procedure of Levene (18),and l,l,l-trifluoroacetone was brominated by the method of Shapiro et al. (19). 2-(Nfl-Dimethylamino)ethanol(DMAE) and chloroethanal dimethyl acetal were used as received. 4'Aminoacetophenone was converted into 4'-acetamidoacetophenone with acetic anhydride in a potassium acetate-buffered solution. The other acetophenones were used as received, and all were brominated by the procedure of Langley (20). Vapor diffusion crystallizations were carried out by dissolving the compound in an appropriate solvent (typically an alcohol) in an oversized beaker or test tube, placing this container in a sealed jar containing an appropriate volatile cosolvent, and allowing the assembly to stand for several days. Alternatively, crystallizationson a larger scale were more conveniently done by placing the solution in a single-neck round-bottomed flask and attaching a sealed cosolvent reservoir. An electrothermal apparatus was used for melting point determinations (uncorrected). 'H and 13C NMR spectra were recorded on either a Bruker WP-200 or an IBM NR-100 spectrometer. Aqueous samples were prepared by dissolving 20 mg of the pure compound in 0.5 mL of D,O containing sodium 3(trimethylsily1)-1-propanesulfonate(DSS) as reference. IR spectra were recorded on either a Perkin Elmer 727B or a 238B spectrophotometer. All elemental analyses were performed by MicAnal Organic Microanalysis of Tucson, AZ. ( B ) 4,4-Dimethyl-2-hydroxy-2-(4'-acetamidophenyl)morpholinium Bromide (6). A solution of 2-bromo4-acetamidoacetophenone (0.528 g, 2.06 mmol) in T H F (40 mL) was stirred at 25 OC while DMAE (0.4 mL, 3.9 "01) in THF (15 mL) was added dropwise. After 90 min, the precipitate was filtered, washed with THF, dried under vacuum, and dissolved in hot MeOHli-PrOH (1:2). Vapor diffusion with T H F at 25 "C gave 0.60 g (84%) of opaque light yellow rhomboidal crystals: mp 190-195 "C dec; 'H NMR (200 MHz, ref to DSS in D20) 6 7.63 (d, 2 H), 7.51 (d, 2 H), 4.67 (m, 1 H), 4.17 (d, 1 H), 3.9-3.3 (m, 4 H), 3.55 (s, 3 H), 3.22 (s, 3 H), 2.18 (s, 3 H); IR (KBr) 3261 (b,
morpholinium Dibromide (7.HBr). A solution of 6 (3.0 g, 6.69 mmol) in 40 mL of 2 M HBr was refluxed for 3 h. After cooling, the solution was concentrated, and the dark yellow solid was dissolved in MeOH. Vapor diffusion with EhO at 25 "C yielded 2.54 g (76%) of light yellow crystals: mp 207 "C dec; 'H NMR (100 MHz, ref to DSS in DzO) 6 7.76 (m, 2 H), 7.48 (m, 2 H), 4.63 (m, 1 H), 4.16 (d, 1 H), 3.8-3.2 (m, 4 H), 3.52 (s, 3 H), 3.20 (8,3 H); IR (KBr) 3320 (b, OH), 3120-2570, 1930 (b, m, NH), 1250, 1089,1070 (e, COCO), 970,931,920 (s, CH3N+)cm-'. Anal. Calcd for C12H&r2Nz02: C, 37.52; H, 5.25; N, 7.29. Found: C, 37.59; H, 5.50; N, 7.25. ( D ) 4,4-Dimethyl-2-hydroxy-2-(4'-chlorophenyl)morpholinium Bromide (8). A solution of 2-bromo-4'-chloroacetophenone (1.5 g, 6.4 mmol) in T H F (25 mL) was stirred at 25 OC while DMAE (0.7 mL, 6.9 m o l ) in THF (25 mL) was added dropwise. After 10 min the white precipitate was fdtered, washed with THF, dried under vacuum, and dissolved in cold EtOH. That solution was vapor diffused with EhO/THF (1:l)at 25 "C to yield 1.73 g (83.8%) of colorless crystals: mp 196-197 "C [lit. (21) mp 215-216 "C]; 'H NMR (200 MHz, ref to DSS in DzO) 6 7.60 (d, 2 H), 7.49 (d, 2 HI, 4.60 (m, 1HI, 4.15 (d, 1 H), 3.6-3.3 (m, 4 H), 3.53 (s, 3 H), 3.20 (s,3 H); IR (KBr) 3258 (b, OH), 2808 (w, ArH), 1033 (w, ArCl) cm-'. Anal. Calcd for Cl2Hl7BrC1NO2:C, 44.67; H, 5.31; N, 4.34. Found: C, 44.84; H, 5.44; N, 4.23.
(E)
4,4-Dimethyl-2-hydroxy-2-(4'-cyanophenyl)-
morpholinium Bromide (9). A solution of DMAE (0.10 mL, 1.0 mmol) in T H F (3 mL) was stirred in an ice bath, while 2bromo-4'-cyanoacetophenone (0.0185 g, 0.084 mmol) in T H F (3 mL) was added dropwise. After 10 min, the white precipitate was filtered, washed with THF, and recrystallized from cold MeOH to yield 22.5 mg (85.6%) of colorless crystals: mp 219-220 "C; 'H NMR (200 MHz, ref to DSS in DzO) 6 7.86 (m, 2 H), 7.78 (m, 2 H), 4.67 (m, 1 H), 4.18 (d, 1H), 3.9-3.5 (m,4 H), 3.54 (8,3 H), 3.20 (s,3 H); IR (KBr) 3260 (b, OH), 2262 (8, C=N), 1090, 1070, 1041 (9.5. s, COCO). 965.930.920 (m. CHIN+)cm-'. Anal. Calcd for C13H17BrN202:C, 49.851 H, 5.47; N, y8.94. Found: C, 4956; H,5.64: N, 8.86. .( F ) 4,4 - D i m e t h y 1- 2 - h y d r o x y - 2 - ( 4 I - n i t r o p h e n y 1 ) morpholinium Bromide (10). A solution of 2-bromo-4'-nitroacetophenone (4.9 g, 20.1 mmol) in T H F (50 mL) was stirred at 25 OC while DMAE (2.05 mL, 20.3 mmol) in T H F (25 mL) was added dropwise. After 10 min, the yellow precipitate was fdtered, washed with THF, dried under vacuum, and dissolved in warm MeOH. That solution was vapor diffused with EhO/THF (1:l) at 25 "C to yield 6.03 g (90.1%) of clear yellow crystals: mp 210 O C dec [lit. (21)mp 213-214 "C]; 'H NMR (200 MHz, ref to DSS in D20) 6 8.33 (d, 2 H), 7.86 (d, 2 H), 4.70 (m, 1H), 4.22 (d, 1H), 3.9-3.3 (m, 4 H), 3.57 (8, 3 H), 3.23 (8,3 H); IR (KBr) 3234 (b, OH), 3049 (w, ArH), 1459, 1319 (e, NOz), 1088,1059, 1014 (8,s, m, COCO), 968, 937, 919 (s, CH3N+) cm-'. Anal. Calcd for C12H17BrN,04:C, 43.26; H, 5.14; N, 8.41. Found: C, 43.40; H, 5.24; N, 8.34. (G) 2-Hydroxy-4,4-dimethylmorpholinium Chloride (1 1). Chloroethanal dimethyl acetal (74 g, 600 mmol) was combined with 100 mL of 18% hydrochloric acid. The mixture was distilled, and the fraction with boiling point 83-90 "C was condensed and dripped through a 15-cm column (13 mm i.d.) packed with 4-A molecular sieves into a solution of DMAE (7.13 g, 800 mmol) in 100 mL of T H F at 25 "C. The solution was stirred overnight and yielded a white, gummy precipitate that was washed three times with THF. The white solid, 25.6 g (26%), was dissolved in MeOH, and the solution was vapor diffused with THF/EhO at 5 OC for 2 days, producing deliquescent colorless plates: mp 215 "C dec; [previously reported (22)without physical data] 'H NMR (200 MHz, ref to DSS in DzO) 6 5.38 (d, 1 H), 4.26 (m, 1 H), 4.04 (m, 1H), 3.6-3.2 (m, 4 H), 3.33 (s, 3 H), 3.24 (8, 3 H);13C NMR (25 MHz, ref to CH30H; 49 ppm, in DzO) 6 87.85 (s, Cl), 63.03 [t, C2, 'J(CN) 2.8 Hz], 60.31 [t, C3, 'J(CN) 2.9 Hz],56.14 (s, C4), 54.37 [t, CH3N+, 'J(CN) 3.6 Hz], 53.19 [t,CH3N+, 'J(CN) 3.8 Hz]; IR (KBr) 3337 (b, OH), 1127,1102,1080,1047 (8, COCO), 960, 912,897 (s, CH3N+)cm-'. Anal. Calcd for C&4ClNO2: C, 42.99; H, 8.42; N, 8.36. Found: C, 42.90; H, 8.66; N, 8.18. '
Inhibition of Acetylcholinesterase by Hemicholiniums
(H)2-Hydroxy-2,4,4-trimethylmorpholinium Bromide (12). A solution of 1-bromo-2-propanone(11.0 g, 0.13 mol) in T H F (100
Chem. Res. Toxicol., Vol. 5, No. 3, 1992 413
sodium carbonate was removed by filtration, and the filtrate was evaporated on a rotary evaporator under reduced pressure. The mL) was slowly added (10 min) to a solution of DMAE (11.6 g, resulting residue was heated to reflux with 80 mL of 47% HI for 1 h and again evaporated to dryness. The crude product was 0.13 mol) in T H F (50 mL) at 25 "C. The solution was stirred for 30 min, and a white, gummy precipitate formed. In a glovebag recrystallized from i-PrOH to give 4.8 g of white crystals (10.8% yield): mp 181-183 "C [lit. (24) mp 182-183 "C]; proton NMR under Np, the solvent was decanted, and the precipitate was washed twice with THF, dissolved in 250 mL Of hot i-PrOHIEtOH is consistent with the expected structure. (2:1), and vapor diffused with T H F (dried over Na2S04). After (C) Kinetic Measurements and Data Reduction. Time 24 h at 25 "C, 6.1 g of a white crystalline solid was obtained. An course data (digital absorbance readings vs time) were acquired additional 14.3 g crystallized a t 5 "C, giving opaque white tetand analyzed by using IBM computers that are interfaced to a Beckman DU-40 or a H P 8452A diode-array UV-visible specragonal prisms (total yield 70%): mp 161-162 "C; 'H NMR (200 trophotometer. Reactions were conducted in 1.00 mL of solution MHz, ref to DSS in D,O) 6 4.45 (m, 1 H), 3.93 (d, 1 H), 3.6-3.3 (m, 4 H), 3.41 (s, 3 H), 3.20 (s, 3 H), 1.47 (s, 3 H); 13C NMR (25 in a glass or quartz cuvette in the water-jacketed cell holder of MHz, ref to CH30H; 49 ppm, in DzO) 6 93.49 (8, Cl), 65.22 [t, the spectrophotometer. The temperature was maintained a t 25.0 C2, 'J(CN) 2.5 Hz], 62.81 [t,C3, 'J(CN) 2.8 Hz], 58.01 [t, CH,N+, h 0.1 OC by a Brinkman Lauda RC-3 or Fisher Isotemp refrig'J(CN) 4.0 Hz], 54.72 (s, C4), 50.65 [t, CH3N+,'J(CN) 3.3 Hz], erated circulating water bath on the DU-40 or HP, respectively. 26.76 (s, C5); IR (KBr) 3451 (b, OH), 3071 (m, CH3), 1116, 1071, The pH value of buffer solutions were measured with a Corning Model 125 pH meter equipped with a glass combination electrode. 1051 (s, COCO), 961,901, 874 (s, CH3N+)cm-'. Anal. Calcd for C7H1J3rNO2:C, 37.18; H, 7.13; N, 6.19. Found C, 36.98; H, 7.40; Time courses for the AChE-catalyzed hydrolysis of ATCh were N, 6.07. followed a t 460 nm by using the coupled Ellman's assay (24). (I) 2-tert-Butyl-2-hydroxy-4,4-dimethylmorpholinium When the initial [ATCh] (=0.3 or 0.4 mM) was greater than the Bromide (13). A solution of l-bromo-3,3-dimethyl-2-butanone K , of ATCh (0.14 mM), time courses for total substrate turnover were described by the integrated form of the Michaelis-Menten (5.0 g, 28 mmol) in T H F (50 mL) was added to a solution of equation (25, 26): DMAE (2.5 g, 28 mmol) in THF (25 mL) a t 25 "C. After 30 min of stirring, 6.1 g (87%) of a white solid precipitated. The product was filtered, washed three times with THF, and dissolved in a minimal amount of warm MeOH. After 24 h of vapor diffusion, as described for 11, with T H F at 5 "C, a total of 3.0 g of colorless where K , = the Michaelis constant, [SI, = the initial substrate crystals was collected. Proton NMR showed the presence of concentration, [SI = the substrate concentration after time t , and MeOH. The crystals were powdered, placed under vacuum for V, = the maximal velocity. The following pair of equations 24 h, and then dissolved in i-PrOH. Vapor diffusion of this solution with THF a t 25 "C produced colorless rhomboid crystals: A,-A A , - A0 mp 168-170 "C; 'H NMR (200 MHz, ref to DSS in D20) 6 4.36 [SI = - [SI, = cp - € 8 c p - €8 (m, 1 H), 3.93 (d, 1 H), 3.6-3.3 (m, 4 H), 3.43 (s, 3 H), 3.22 (9, 3 H), 0.97 ( s , 9 H); IR (KBr) 3351 (b, OH), 2957, 2917, 2849 (m, where A , = the absorbance a t infinite time, A, = the initial CH3),1466,1111,1094,(s, COCO), 977,922 (s, CH3N+)cm-'. Anal. absorbance, tp = the absorptivity constant of the product, and Calcd for CloH22BrN02:C, 44.78; H, 8.27; N, 5.22. Found: C, t, = the absorptivity constant of substrate, is used to convert eq 44.40; H, 7.91; N, 5.20. ( J ) 2-Hydroxy-4,4-dimethyl-2-(trifluoromethyl)- 1 into the analogous expression in terms of absorbances: morpholinium Bromide (14). A solution of 1-bromo-3,3,3trifluoro-2-propanone (3.4 g, 17.9 mmol) in T H F (50 mL) was added to a solution of DMAE (1.6 g, 17.9 mmol) in T H F (25 mL) at 25 "C. After the solution was stirred for 30 h, the solvent was Fitting time course data to eq 3 gives V, and K , in A and decanted, and 3.8 g (77%) of a white solid was isolated. The solid A s-' units, respectively, where A is absorbance a t X = 460 nm. was washed three times with T H F and dissolved in 80 mL of Dividing by 4520, the value of tp - t, a t the analytical wavelength, MeOHli-PrOH (21). Vapor diffusion with EhO for 30 h produced converts the parameters, K , and V-, into units of M and M s-', colorless tetragonal prisms: mp 254 "C dec; 'H NMR (200 MHz, respectively. Reactions were initiated by injecting enzyme solution ref to DSS in D20) 6 4.50 (m, 1 H), 4.17 (d, 1 H), 3.8-3.6 (m, 4 into a cuvette containing other components. Reactions were run H), 3.49 (s,3 H), 3.32 (s, 3 H); IR (KBr) 3177 (b, OH), 1397-897 a t least in duplicate. (b, CF3) cm-'. Anal. Calcd for C7H13BrF3N0,: C, 30.02; H, 4.68; N, 5.00. Found: C, 30.17; H, 4.82; N, 5.00. Iowa City. (A) Materials. Inorganic phosphate buffer Results components were reagent grade and were used without further Syntheses and Structure of Hemicholiniums. The purification. Grade V-S AChE (EC 3.1.1.7) from E . electricus 6-14 require 2-(NJV-dimethylamino)ethanol syntheses of was obtained as a lyophilized powder from Sigma Chemical Co. and the appropriate bromomethyl ketone (21). The ex(St. Louis, MO). Prior to use the enzyme was dissolved in 0.1 ception is the synthesis of 11, in which chloroethanal is M sodium phosphate buffer, pH 7.30, that contained 0.1 N NaC1. generated in situ from hydrolysis of chlorcethanal dimethyl Acetylthiocholine (ATCh) chloride and 5,5'-dithiobis(2-nitrobenzoic acid) (DTNB) were obtained from Sigma and used without acetal. The ring closure favors an axial hydroxyl, which further purification. 3-Aminophenol, 3-(N,N-dimethylamino)arises from the anomeric effect (27), the gauche effect, and phenol, methyl iodide (99.5%), anhydrous sodium carbonate, 47% steric effects. The NMR spectra contain separate signals HI, and choline chloride were purchased from Aldrich Chemical for axial and equatorial methyl groups on nitrogen; each Co. (Milwaukee, WI). HPLC-grade MeOH was obtained from methylene signal contains many lines. This suggests to Mallinckrodt (St. Louis, MO). HPLC-grade i-PrOH was obtained us that one conformation predominates. We only detect from Baxter (McGraw Park, IL). Deuterium oxide (99.8%) was the hemiketal form in NMR and IR spectra of 6,8-10, and purchased from Aldrich. Trizma-HC1and Trizma base were from 12-14, but 7 is a mixture of keto and hemiketal forms (28). Sigma. Precoated TLC plates were from Macherey and Nagel, Compound 11 is a hemiacetal and has an NMR spectrum FRG, purchased through Brinkmann Instruments (Westbury, NY). Melting points were determined on a Thomas-Hoover similar to the those of the hemiketals. Single-crystal X-ray capillary melting point apparatus and are uncorrected. Proton analyses2have confirmed nearly all structures except 11, NMR spectra were obtained on a Jeol FX9OQ F T NMR. which has disordered crystals. The structure of 9 is pub(B) (3-Hydroxypheny1)trimethylammoniumIodide (23). lished (29). 3-Aminophenol (17.5 g, 0.16 mol) was heated to reflux overnight with 80 mL of methyl iodide (1.29 mol) and 16 g of anhydrous sodium carbonate (0.15 mol) in 100 mL of methanol in a 500-mL T. C. Stelly, W. J. Colucci, F. R. Fronczek, and R. D. Gandour, round-bottomed flask. After cooling down, the reaction mixture unpublished results.
414 Chem. Res. Toxicol., Vol. 5, No. 3, 1992
Lee et al.
Table I. K , and V,,, for Control Reaction and Inhibitions by 3-(N,N-Dimethylamino)phenol(4)" [41 (mM) 1 0 3 (MI ~ ~ io6vm, (M 9-11 K / V (9) 0 0.135 4.20 32.3 0 0.134 4.20 32.3 0.02 0.174 4.42 40.0 0.02 0.168 4.20 39.6 0.05 0.211 4.42 48.6 0.05 0.203 4.20 49.2 0.10 0.261 4.20 62.6 0.10 0.259 4.20 63.5 0.20 0.369 3.98 93.1 0.20 0.385 3.98 97.3
"The initial ATCh concentration was 0.3 mM, and the DTNB concentration was 0.6 mM. The reaction was initiated by adding 0.2 fig of AChE (3.08 nM) into a cuvette containing ATCh and DTNB. Every reaction was carried out at 25.0 f 0.1 "C in 0.1 M sodium phosphate buffer, pH 7.30, that contained 0.1 N NaCl and was monitored at 460 nm. The inhibition constant, Ki, was calcuM lated from the K / V vs [I] replot. Ki = (1.06 f 0.03) X (12.6%). Table 11. K , and V,,, for Control Reactions and Inhibitions b s (3-Hsdroxs~hen~1)trimethslammonium (5)"
0.139 0.152 0.225 0.219 0.316 0.324 0.480 0.510 0.646 0.786
0 0
0.13 0.13 0.25 0.25 0.51 0.51 1.00 1.00
4.20 4.64 4.20
33.2 32.8 52.8 50.4 70.7 70.7 110.0 110.5 182.9 190.0
4.42
4.42 4.64 4.42 4.64 3.54 4.20
"The initial ATCh concentration was 0.3 mM,and the DTNB concentration was 0.6 mM. The AChE concentration was 3.08 nM. Every reaction was conducted at 25.0 f 0.1 "C in 0.1 M sodium phosphate buffer, pH 7.30, that contained 0.1 N NaCl and was monitored at 460 nm. The inhibition constant, Ki, was calculated from the K / V vs [I] replot. Ki = (2.10 f 0.06) X M (f3.0%). 240
7
160
1
? Y 80
1
I
0 ' 0.00
1
0.40 [I] x
0.80 10'
1.20
(M)
Figure 1. Fit of K / V versus the concentration of 5 to eq 1 (see text). The inhibition constant, Ki, calculated from linear least squares is 210 f 6 nM.
Inhibition of AChE by 4 and 5. Tables I and I1 present kinetic parameters, K , and V-, for inhibition by 4 and 5, respectively. The inhibition constant, Ki, is obtained from the replot of K,/V,, vs inhibitor concentration (Figure l),described by the following equation for competitive inhibition: (4) Because both inhibitors affect only K , by binding only to
Scheme I. Linear Mixed-Inhibition Mechanism for Interaction of AChE with Choline and Hemicholiniums
E + S
__
ES
kcar
- E + S
Kf
E1
-
ESI
aKs Table 111. Ki of Choline and Hemicholiniums comDd inhibitor K; (uM)" Me3N+CH2CH20HC1930 f 80 6 R = C6H4-4-NHC(0)CH3,X = Br660 f 30 7 R = C6H4-4-NH2,X = Br220 f 50 8 R = C6H4-4-Cl,X = Br2590 f 130 9 R = C6H4-4-CN,X = Br3690 f 90 10 R = C6H4-4-NO2,X = Br1630 f 70 11 R = H, X = Cl560 f 50 12 R = CH3, X = Br550 f 20 13 R = CMe,, X = Br1180 f 110 14 R = CF3,X = Br1220 f 120
'Ki is calculated from the K / V vs inhibitor concentration replot.
free enzyme, both inhibitors are classified as reversible competitive inhibitors. Inhibition by 5 shows a solvent isotope effect, H2°Ki/DZoKi, of 0.83 f 0.04. The pKa of 5 determined from spectrophotometric titration agrees closely with that (8.15) determined by titration. The absorbance at 294 nm of a 0.5 mM solution of 5 increases sigmoidally with pH in the range of 7.6-9.0; a nonlinear least-squares analysis of the data gives pK, = 8.2 f 0.3 (30). Inhibition of AChE by Choline. Choline changes both K, and V,, of AChE-catalyzed hydrolysis of ATCh. With 1.9 mM choline in the reaction mixture, V,, decreases by 30%, while Km increases by 100% compared to the control reaction. Therefore, choline binds to both the free enzyme and, albeit less tightly, to ES complexes, as outlined in Scheme I. The ES complex to which choline binds i3 probably the acetyl-AChE intermediate, which accumulates during catalytic turnover (2,31). The value of Ki for the competitive component of this mechanism can be determined by constructing a replot according to eq 4. The replot gives a value of Ki for choline of 930 pM (Table 111). To avoid the complexity of interpreting the many enzyme forms that contribute to aKi, we only consider Ki, which involves inhibiting the free enzyme. Inhibition by Hemicholiniums. The hemicholiniums also inhibit AChE by the mechanism of Scheme I, and accordingly, values of Ki were calculated by fitting K / V versus [I] data to eq 4. Table I11 gives the values of Ki for the hemicholiniums, 6-14. The values of Ki for these compounds resemble that for choline inhibition. Compound 11, a choline analogue,and 12, an isomeric analogue of acetylcholine, bind quite strongly. The aryl series contains the strongest and weakest inhibitors. Discussion 3-Hydroxyphenylamines, 4 and 5. Compound 5, the principal metabolite of neostigmine, which is further metabolized to a glucuronide conjugate (32),potently inhibits AChE (Ki of 0.21 pM). Compound 4, which is monodemethylated 5, reversibly inhibits AChE with a value of Ki 500-fold larger than that of 5. Our pKa and Ki of 5 compare favorably with Wilson and Quan's (33),8.1 and 0.31 pM, respectively. Compound 5 binds 1000-foldand
Chem. Res. Toxicol., Vol. 5, No. 3, 1992 415
Inhibition of Acetylcholinesterase by Hemicholiniums 120-fold more tightly than the saturated analogue, (3hydroxycyclohexyl)trimethylammonium,and the deshydroxy analogue, (trimethylphenyl)ammonium, respectively (11). These comparisons suggest that a benzoid ring and a more acidic (phenol vs cyclohexanol) hydrogen bond donor favor binding. Compound 5 (pK, = 8.2) is a better donor than hemicholiniums (pK, 11)and choline (pK, 14). Hydrogen bonding of the phenolic hydroxyl with the imidazole (pK, = 6.3) in the esteratic site of the AChE active site (34) can explain the enhanced binding of 5. The small ApK, between the imidazole and the phenol predicts a strong hydrogen bond. Crystallographic studies (35) show that when the donor and acceptor are chemically similar, short, strong hydrogen bonds form. The phenolate zwitterion of ethyl(3-hydroxypheny1)dimethyla”onium (tensilon)inhibits AChE less than the undissociated phenol (36). The solvent isotope effect on Kifor competitive inhibition by 5 of AChE-catalyzed hydrolysis of acetylthiocholine is 0.83. Julin and Kirsch (37) measured solvent isotope effects of 1.13-1.42 for competitive inhibition by various ligands of chymotrypsin catalysis. Enzyme reactions that involve protolytic catalysis typically have solvent isotope effects on V of 2-4 (38,39). The solvent isotope effect for AChE-catalyzed hydrolysis of ATCh, H2°Vmax/ D”V,, = 2, suggests that a proton transfers in the catalytic step (40). The inverse solvent isotope effect on Ki means that 5 binds more tightly in HzO than in D20. The inverse solvent isotope effect on Ki can be explained by following equation:
-
-
Hz0Ki/D20Ki =
n@I/nqjF+I 1
I
(5)
In eq 5, the product in the numerator is for sites in the E1 complex that have values of 4, isotopic fractionation factors, # 1. The product in the denominator is for isotopic fractionation at sites on free AChE and free 5. Because phenols and the amino acid side chains of the AChE active site (3) (i,e., Serm, Hism) have fractionation factors near unity (38,39),one is left with fractionation in E1 to describe the solvent isotope effect on Ki.The formation of a short, strong hydrogen bond between the phenolic hydroxy on 5 and His440could give 4E1= 0.83 and thus account for the isotope effect. Kreevoy and co-workers (41, 42) have shown that in organic solvents strong hydrogen bonds between donors and acceptors that have small ApK,‘s give values of 4 < 1. The hydrogen bond that we propose is not as strong as those seen by Kreevoy et al. We cannot dismiss that the isotope effect arises from the multiplication of numerous small isotope effects (43) that could be created by desolvation of both inhibitor and active site. Julin and Kirsch (37) see effects of this type for small normal isotope effects. Because our effect is inverse, we favor the strong hydrogen bond formation. Choline and Hemicholiniums 6-14. Our value of Ki for choline in AChE-catalyzed hydrolysis of ATCh (930 pM)agrees with that (1000 p M ) for AChE-catalyzed hydrolysis of ACh (44). The values of Ki of the hemicholiniums using AChE from E. electricus closely resemble that of choline, but are lower than those for phenylhemicholinium evaluated with bovine, horse, and frog AChE’s (12). This similarity suggests that the conformation of hemicholiniums (gauche with respect to N+CCO), which we have from crystal structures (29),is the conformation of choline at the choline recognition site on AChE. Accordingly, we propose that choline is in the gauche conformation when it binds to AChE. Ring-Chain Tautomerism. The hemicholiniums presumably bind to the enzymes as cyclic structures.
-2.00
-2.50 .
-4.00
t y‘ ‘
-0.80
I
0.80
0.00 Sigma
Figure 2. A linear plot of log Ki vs substituent constant for 4-substituted phenyl hemicholiniums. Data (substituent, u, log Ki): NHz, -0.66, -3.66; NHC(O)CH,, 0.00, -3.18; C1, 0.23, -2.59;
CN, 0.66, -2.43; NOz, 0.78, -2.79.
Depending on the substituent, the cycle dominates the equilibrium (eq 6). Equilibrium constants for hemiU
Me‘
he
choliniums, Ar = 4-HzN-C6H4,4-H3C(0)-C6H4,Ph, and 4-H3N+-C6H4, are 1.49,9.2, 221, and 187, respectively (28). The values of pK, for the hemiketal hydroxyl are -12.0, and 11.3, and =10.8 for Ar = 4-H2N-C6H4,4-H3C(0)-C6H4, Ph, respectively. Of the five compounds studied here, only compound 7 (Ar = 4-HzN-C6H4)exists as nearly equal amounts of open and closed forms. Aryl Substituents. Figure 2 shows a plot of log Ki vs substituent constant (a) (45). Except for 10 (Ar = 402N-C6H,), log Ki correlates fairly well with u. If one fits all five points to a straight line, as shown in Figure 2, p equals 0.8 f 0.2 (r = 0.875). When the point for the Ki of 10 is excluded, p equals 1.0 f 0.2 (r = 0.965). The value of p (=l)suggests that the electron-withdrawing groups favor release of the inhibitor from the enzyme. Inhibition increases with a higher pK, of the hemiketal hydroxyl and increasing chain form in ring-chain tautomerism (28). These correlations, however, might be coincidental; and specific interactions, e.g., hydrogen bonding, dipole-dipole, charge transfer, etc., between enzyme and substituent may determine the magnitude of Ki. Compound 10 may deviate because of just such interactions. Hydrogen bonding with the substituents, NHAc, NH2, and NOz, may contribute to the binding of 6, 7, and 10, respectively. In summary, we can only guess at why log Ki correlates so well with u (vide infra). Alkyl Substituents. The values of Ki for the alkylsubstituted hemicholiniums, 11-14, indicate that AChE shows modest selectivity for size and polarity of the group on C2. Compounds 11 (R = H) and 12 (R = CH,) have identical values of Ki, which are smaller than that of choline. Compounds 13 (R = CF,) and 14 [R = C(CH,),] also have identical values of Ki, which are larger than that of choline, but only twice as large as those for 11 and 12. The identical values for 13 and 14 argue against any selectivity for substituents at C2 on the basis of alkyl polarity. Combined with the range in the values of Ki for 6-10, AChE shows a modest steric selectivity for substituents at C2. Models for AChE Binding of Hemicholiniums. The crystal structures of all hemicholiniums show an axial
416 Chem. Res. Toxicol., Vol. 5, No. 3, 1992
I
Me3N
+ 15
+148
O-C-C-N+ , deg
hydroxy group at C2. The hemicholiniums have only gauche conformations for the three O-C-C-N+ torsion angles: (H)O-C2-C3-N+, +72O; O-C2-C3-N+, - 5 O O ; and O-C5-C6-N+, +61°. The N-O(H) distance is 3.02 A. H
Lee et al.
H
We cannot, for the moment, prove how hemicholiniums bind to AChE. Does AChE bind the closed or open form? If closed, then there are two enantiomers, which cannot be resolved because of the lability of the hemiketal group. Which enantiomer binds better, or do both bind the same? We can, however, speculate on the nature of the binding site, by comparing these data with our recent data on substrate activity (40). The respective relative values of V / K for ATCh, propanoylthiocholine, and butanoylthiocholine are 1, 0.3, and 0,001, which points to a highly circumscribed steric tolerance for the esteratic site. The relative values of Ki for the arylhemicholiniums range from 1to 0.06, which depends on electronic effects, not size. The range for the smallest (R= H)to the largest [R = C(CH,),] is only 2-fold. The R group on the hemicholiniums must be directed away from the esteratic site, which is located at the bottom of a deep gorge in the enzyme. The walls of the gorge, lined with aromatic and other hydrophobic side chains, provides a likely site for charge-transfer interactions with the aryl group of the arylhemicholiniums. AChE binding of arylhemicholiniums improves with electron-donating substituents, which implies an electron-poor site in the gorge. This proposed charge-transfer interaction might be independent of ring-chain tautomerism. Comparison with Other Conformationally Constrained Analogues. Smissman and co-workers’ conformationally rigid analogues, 15-17, of acetylcholine probed torsion-angle requirements for AChE-catalyzed hydrolysis (see Chart 111) (46-49). They found that AChE hydrolyzes 15 and 16 quickly, but 17 very slowly. They suggested (49) that AChE-catalyzed hydrolysis requires a large positive anticlinal O-C-C-N+ torsion angle and proposed that AChE has flexibility in binding the quaternary ammonium group. Given the strong conformational preference for gauche over anti in choline and ACh (50-53), Smissman’s proposal suggests that the transition structure of AChE-catalyzed hydrolysis of ACh arises from a less stable ground-state conformation. This proposal is another example of the Curtin-Hammett principle (54, 55), which states that a reaction rate depends only on the lowest energy transition state and not on the energies of ground-state conformations. The conformation of bound choline, the product, must be gauche to support our suggestion that hemicholiniums are conformationally constrained analogues of choline. Given the various substituents at C2 in the hemicholiniums and the strong binding of the (3-
16 +169
17
+74
hydroxyphenyl)trimethylammoniums, especially 5, we agree with Schowen et al. (49) that there is ample space for free rotation about the C-C bond in choline. Once choline departs the putative tetrahedral intermediate, it rotates from anti to gauche. Model for AChE Binding of 5. Compound 5, which has one more carbon between the hydroxy and trimethylammonium than does choline, inhibits much more potently than either choline or the hemicholiniums. Compound 5 provides both a correct spacer to span the trimethylammonium recognition site and the esteratic site of AChE and an aromatic moiety for interaction with the hydrophobic walls of the active site. In addition, the tight binding of 5 might result from the conformational rigidity of the aromatic ring, which would position the phenolic hydroxy near the space that the carbonyl carbon of ACh occupies in the anti conformation, which Smissman et al. (46-49) suggest is the reactive conformation. Active Site of AChE. Sussman et al. (3)have used the anti conformation to model the tetrahedral intermediate in AChE catalysis. Their model shows (a) interactions between TrpMand the quaternary ammonium ion, (b) a covalent bond between Ser200and the carbonyl carbon of ACh, (c) hydrogen bonding between Hisu0 and the leaving-group oxygen, and (d) hydrogen bonding between the carbonyl oxy anion and the NH groups of Gly118,Gly119, the putative “oxyanion hole”). The strong and Alam1(viz., complex of AChE and 5 might result from the quaternary ammonium group on 5 interacting with TrpM,while the phenolic hydroxy hydrogen bonds with Hisw (and perhaps Ser2”’). Calculations with the SYBYL molecular modeling program (56) support this idea: the N to carbonyl distance in the anti conformation of ACh is 4.94 A, a close match to the Ne.-0 distance, 4.75 A, calculated for 5. Choline and the hemicholiniums are too short to make a strong hydrogen bond. SYBYL calculations on choline support this idea, because the N--O distance in the gauche conformation is 3.03 A and in the anti conformation, 3.82 A. A comparison of the values of Ki of choline and the hemicholiniums, coupled with the finding of Smissman et al. (46-49) that their gauche analogues inhibit AChE, strongly suggests that AChE binds the gauche conformation of choline.
Summary The hemicholiniums inhibit AChE with a potency equal to choline, the product inhibitor, but not as well as 5. If closed, they do provide a molecular skeleton on which to anchor other groups to explore the topography of the active site. With the crystal structure of AChE as a guide, we hope to design analogues that will test our suggestions of how hemicholiniums bind to AChE. Acknowledgment. D.M.Q. and R.D.G. thank the NIH for support of this work through Grants NS21334 and GM42016, respectively. Registry No. 4, 99-07-0; 5, 2498-27-3; 6, 140361-25-7; 7, 136294-54-7; 8, 4303-91-7; 9, 140361-26-8; 10, 4303-90-6; 11,
Inhibition of Acetylcholinesterase by Hemicholiniums 140361-27-9; 12, 140361-28-0; 13, 140361-29-1; 14, 140361-30-4;
D W ,108-01-0; AChE, 9000-81-1; 2-bromo-4’-acetamidoacetophenone, 21675-02-5; 2-bromo-4’-chloroacetophenone, 536-38-9; 2-bromo-4’-cyanoacetophenone,20099-89-2; 2-bromo-4’-nitroacetophenone, 99-81-0; l-bromo-2-propanone, 598-31-2; 1bromo-3,3-dimethyl-2-butanone, 5469-26-1; 1-bromo-3,3,3-trifluoro-2-propanone, 431-35-6; choline, 62-49-7.
References (1) Rosenberry, T. L. (1975) Acetylcholinesterase. Adu. Enzymol. Relat. Areas Mol. Biol. 43, 103-218. (2) Quinn, D. M. (1987) Acetylcholinesterase. Enzyme structure, reaction dynamics and virtual transition states. Chem. Rev. 87, 955-979. (3) Sussman, J. L., Harel, M., Frolow, F., Oefner, C., Goldman, A., Toker, L., and Silman, I. (1991) Atomic structure of acetylcholinesterase from Torpedo californica: a prototypic acetylcholine-binding protein. Science (Washington, D.C.) 253, 872-879. (4) Taylor, P. (1985) Anticholinesterase agents. In Pharmacological Basis of Therapeutics (Gilman, A. G., Goodman, L. S., and Murad, F., Eds.) pp 110-129, McMillan, New York. (5) Summers, W. K., Majowki, L. V., Marsh, G. M., Tachiki, K., and Kling, A. (1986) Oral tetrahydroaminoacridine in long-term treatment of senile dementia, Alzheimer type. N. Engl. J. Med. 315, 1241-1245. (6) Thal, L. J., Fuld, P. A., Masur, D. M., and Sharpless, N. S. (1983) Oral physostigmine and lecithin improve memory in Alzheimer disease. Ann. Neurol. 13,491-496. (7) Marchbanks, R. M. (1982) Biochemistry of Alzheimer’s dementia. J. Neurochem. 39, 9-15. (81 Davis. K. L.. Mohs. R. C.. Tinklenbere. J. R.. and Pfferbaum, A. (1978) Physostigmine: ‘improvement of long-term memory processes in normal humans. Science (Washington, D.C.) 201, 272-274. (9) Sitaram, N., Weingartner, H., and Gillin, J. C. (1978) Human serial learning: enhancement with arecoline and choline and impairment with scopolamine. Science (Washington, D.C.) 201, 274-276. (10) Triggle, D. J. (1971) Conformational change and regulatory processes in proteins. In Neurotransmitter-Receptor Znteractions (Triggle, D. J., Ed.) pp 63-71, Academic Press, London. (11) Bakry, N. M., Eldefrawi, A. T., Eldefrawi, M. E., and Riker, W. F., Jr. (1982) Interactions of quaternary ammonium drugs with acetylcholinesterase and acetylcholine receptor of Torpedo electric organ. Mol. Pharmacol. 22,63-71. (12) Hemsworth, B. A. (1971) Effect of hemicholinium analogs, HC-15, on neuromuscular transmission. Eur. J.Pharmacol. 15, 91-100. (13) Sterling, G. H., Doukas, P. H., Ricciardi, F. J., Jr., Biedrzycka, D. W., and O’Neill, J. J. (1986) Inhibition of high-affinity choline uptake and acetylcholine synthesis by quinuclidinyl and hemicholinium derivatives. J. Neurochem. 46, 1170-1175. (14) Hemsworth, B. A., Darmer, K. I., Jr., and Bosmann, H. B. (1971) The incorporation of choline into isolated synaptosomal and synaptic vesicle fraction in the presence of quaternary ammonium compounds. Neuropharmacology 10,109-120. (15) Simon, J. R., Mittag, T. W., and Kuhar, M. J. (1975) Inhibition of synaptosomal uptake of choline by various choline analogs. Biochem. Pharmacol. 24,1139-1142. (16) Diamond, I., and Milfay, D. (1972) Uptake of [3H-methyl]choline by microsomal synaptosomal, mitochondrial and synaptic vesicle fractions of rate brain. The effect of hemicholinium. J. Neurochem. 19, 1899-1909. (17) Bhattacharyya, B., Flynn, J. R., Cannon, J. G., and Long, J. P. (1986) Anticholinesterase activity and structure activity relationships of a new series of hemicholinium-3 analogs. Eur. J. Pharmacol. 132, 107-114. (18) Levene, P. A. (1943) Bromoacetone. Organic Syntheses (Blatt, A. H., Ed.) Collect. Vol. 11,2nd ed.,pp 88-89, John Wiley & Sons, New York. (19) Shapiro, B. L., Lin, H.-L., and Johnston, M. D., Jr. (1973) NMR studies and rotational isomerism in fluoroacetones. I. Trifluoroacetones. J. Magn. Reson. 9, 305-325. (20) Langley, W. D. (1941) p-Bromophenacyl bromide. Organic Syntheses (Gilman, H., and Blatt, A. H., Eds.)Collect. Vol. I, 2nd ed., pp 127-128, John Wiley & Sons, New York. (21) Anderson, E. L., Casey, J. E., Jr., Force, E. E., Jensen, E. M., Matz, R. S., and Rivard, D. E. (1966) Antiviral agents. 11. Sub..I
~
I
I
Chem. Res. Toxicol., Vol. 5, No. 3, 1992 413 stituted morpholinium quaternary salts. J. Med. Chem. 9, 211-213. (22) Yokoo, A. (1948) Preparation of some amino acids and amino aldehydes. Bull. Tokyo Znst. Technol. 13, 13-17; (1951) Chem. Abstr. 45, 553h. (23) Cocolas, G. H., Granford, G., and Choi, H. S. Y. (1974) Studies on the conformational requirements of substrate and inhibitor on acetylcholinesterase. J. Med. Chem. 17, 938-943. (24) Ellman, G. L., Coutney, K. D., Andres, V., Jr., and Featherstone, R. M. (1961) A new and rapid colorimetric determination of acetylcholinesterase activity. Biochem. Pharmacol. 7,88-95. (25) Stout, J. S., Sutton, L. D., and Quinn, D. M. (1985) Acylenzyme mechanism and solvent isotope effects for cholesterol esterasecatalyzed hydrolysis of p-nitrophenyl butyrate. Biochim. Biophys. Acta 837, 6-12. (26) Segel, I. H. (1975) Kinetics of unireactant enzymes. Enzyme Kinetics, pp 54-64, John Wiley & Sons, New York. (27) Kirby, A. J. (1983) Stereoelectronic effects on structure. The Anomeric Effect and Related Stereoelectronic Effects at Oxygen, pp 3-32, Springer-Verlag, Berlin. (28) Soresen, P. E., McClelland, R. A., and Gandour, R. D. (1991) Intramolecular hemiacetals. The acid-base-catalyzed ring-chain interconversion of 2-substituted 2-hydroxy-4,4-dimethylmorpholinium cations in aqueous solution. Acta Chem. Scand. 45, 558-566. (29) Altbach, M. I., Fronczek, F. R., Gandour, R. D., Garcia-Guajardo, J. G., Lin, Y. M., and Watkins, S. F. (1988) 2-(4-cyanophenyl)-2-hydroxy-4,4-dimethylmorpholiniumbromide. Acta Crystallogr., Sect C C44,1689-1690. (30) Skoog, D. A. (1985) Potentiometric methods. Principles of Instrumental Analysis, 3rd ed., pp 632-640, Saunders College Publishing, Philadelphia. (31) Froede, H. C., and Wilson, I. B. (1984) Direct determination of acetyl-enzyme intermediate in the acetylcholinesterase-catalyzed hydrolysis of acetylcholine and acetylthiocholine. J. Biol.Chem. 259, 11010-11013. (32) Hussain, M. A,, Roberts, J. B., Thomas, B. H., and Wilson, A. (1969) Metabolism and excretion of 3-hydroxyphenyltrimethylammonium and neostigmine. Br. J. Pharmacol. 35, 344-350. (33) Wilson, I. B., and Quan, C. (1958) Acetylcholinesterase studies on molecular complementariness. Arch. Biochem. Biophys. 73, 131-143. (34) Haworth, R. D., Lamberton, A. H., and Woodcock, D. (1947) Influence of chemical constitution upon toxicity. I. Compounds related to ‘doryl”. J. Chem. SOC., 176-182. (35) Olovsson, I., and Jonsson, P.-G. (1976) X-ray and neutron diffraction studies of hydrogen bonded systems. In The Hydrogen Bond (Schuster, P., Zundel, G., and Sandorfy, C., Eds.) Vol. 2, pp 426-433, North Holland, Amsterdam. (36) Wilson, I. B. (1955)The interaction of tensilon and neostigmine with acetylcholinesterase. Arch. Znt. Pharmacodyn. Ther. 104, 204-21 1. (37) Julin, D. A,, and Kirsch, J. F. (1981) Deuterium oxide solvent isotope effects on chymotrypsin-inhibitor dissociation constants. Fed. Prod., Fed. Am. SOC. Exp. Biol. 40, 1657-1668. (38) Schowen, K. B. J. (1978) Solvent hydrogen isotope effects. In Transition States ofBiochemica1 Processes (Gandour, R. D., and Schowen, R. L., Eds.) pp 225-283, Plenum Press, New York. (39) Schowen, K. B., and Schowen, R. L. (1982) Solvent isotope effects on enzyme systems. Methods Enzymol. 87, 551-556. (40) Pryor, A. N., Selwood, T., Leu, L.-S., Andracki, M. A,, Lee, B. H., Rao, M., Rosenberry, T. L., Doctor, B. P., Silman, I., and Quinn, D. M. (1992) Simple general acid-base catalysis of physiological acetylcholinesterase reactions. J. Am. Chem. SOC. (in press). (41) Kreevoy, M. M., Liang, T.-M., and Chang, K.-C. (1977) Structures and isotopic fractionation factors of complexes AHA-’. J. Am. Chem. SOC. 99, 5207-5209. (42) Kreevoy, M. M., and Liang, T.-M. (1980) Structures and isotopic fractionation factors of complexes, A1HA2?. J. Am. Chem. SOC. 102, 3315-3322. (43) Kresge, A. J. (1975) Solvent isotope effects and the mechanism 95, 3065-3067. of chymotrypsin action. J. Am. Chem. SOC. (44)Hasan, F. B., Elkind, J. L., Cohen, S. G., and Cohen, J. B. (1981) Cationic and uncharged substrates and reversible inhibitors in hydrolysis by acetylcholinesterase. J.Biol.Chem. 256,7781-7785. (45) McDaniel, D. H., and Brown, H. C. (1958) Extended table of Hammett substituent constants based on the ionization of substituted benzoic acids. J. Org. Chem. 23, 420-427.
418
Chem. Res. Toxicol. 1992,5, 418-425
(46) Smiesman, E. E., Nelson, W. L., LaPidus, J. B., and Day, J. L. (1966)Conformational aspects of acetylcholine receptor sites. The halides isomeric 3-trimethylammonium-2-acetoxy-trans-decalin and the isomeric a,@-dimethylacetylcholine halides. J. Med. Chem. 9,458-465. (47) Stephen, W. F., Smissman, E. E., Schowen, K. B., and Self, G. W. (1972) Conformational aspects of systems related to acetylcholine. 3. Base-catalyzed and acetylcholinesterase-catalyzed hydrolysis of the isomeric dl-3-trimethylammonium-2-acetoxytrans-decalin halides and the isomeric dl-1-methyl-3-acetoxytrans-decahydroquinoline methiodides. J. Med. Chem. 15, 241-243. (48) Smissman, E. E., and Parker, G. R. (1973) Conformational aspects of systems related to acetylcholine. 5. Syntheses of the dl-a-(e)-methyl-, dl-3(e)-methyl-, and dl-2(e), 3(e)-dimethyl-3(a)-trimethylammonium-2(a)-acetoxy-trans-decalin halides. J. Med. Chem. 16, 23-27. (49) Schowen, K. B., Smissman, E. E., and Stephen, W. F., Jr. (1975) Base-catalyzed and cholinesterase-catalyzed hydrolysis of acetylcholine and optically active analogs. J. Med. Chem. 18, 292-300.
(50) Sundaralingam, M. (1968) Conformational relationship between the 0-C-C-N+system of phospholipids and substrates of muscarinic and cholinergic systems. Nature (London) 217, 35-37. (51) Cushley, R. J., and Mautner, H. G. (1970) NMR studies on the conformation of acetylcholine isologues. Tetrahedron 26, 2151-2159. (52) Partington, P., Feeney, J., and Burgen, A. S. V. (1972) The conformation of acetylcholine and related compounds in aqueous solution as studied by nuclear magnetic resonance spectroscopy. Mol. Pharmacol. 8, 269-277. (53) For a summary of crystallographic data see: Datta, N., Mondal, P., and Pauling, P. (1980) Structure of acetylcholine tetraphenylborate. Acta Crystallogr., Sect. B. B36,906-909. (54) Curtin, D. Y. (1954)Stereochemicalcontrol of organic reactions. Differences in the behavior of diastereomers. I. Ethane derivatives. The cis effect. Rec. Chem. Prog. 15, 111-128. (55) Hammett, L. P. Transition-state theory. Physical Organic Chemistry: Reaction Rates, Equilibria, and Mechanisms, 2nd ed., pp 119-120, McGraw-Hill, New York. (56) SYBYL Molecular Modeling Software, Version 5.4, Tripos Associates, Inc., 1699 s.Hanley Rd., Suite 103, St. Louis, MO 63144.
Toxicology of Drinking Water Disinfection Byproducts from Nutrients. Rate Studies of Destruction of Polyunsaturated Fatty Adds in Vitro by Chlorine-Based Disinfectants J. Peter Berm Environmental Monitoring Systems Laboratory, US.Environmental Protection Agency, 26 West Martin Luther King Drive, Cincinnati, Ohio 45268 Received February 12, 1992 As model reactions between unsaturated fats and water disinfectants in the GI tract, relative rates of destruction of seven polyunsaturated fatty acids (L, aLn, yLn, Ara, EPA, DH, and DT) by OC1- and NH&l were investigated in vitro. Using millimolar solutions of seven PUFAs combined with various OC1- mole ratios, disappearance of PUFAs was followed by UV spectrophotometry a t pH = 9.5 and a t 35 "C via conjugated hydroperoxydienes a t 234 nm. While OC1- rapidly destroyed all PUFAs, NH2C1 was inert. Overall second-order rate constants computed for L a t increasing times disclosed that the attack on the cis-CH=CHCH,CH=CH moiety by OC1- does not follow simple second-order kinetics. Using a logit-log transform and second-order polynomial regression analysis of L's disappearance in a stoichiometric ([L] = 1.2 mM,[ClO-] = 2.4 mM) mix, data were analyzed by the time ratio method of Schwemer and Frost. These agreed with a sequential system of a t least two irreversible second-order reactions having lzl = 15.6 L-mol-'4 and k2 = 2.6 L.mol-'d. Preliminary GC/MS analysis indicated that the initial product is a mix of chlorohydrin isomers. These undergo second addition of HOC1 and/or lose halogens and polymerize. Additional minor products were also C W 9 mono- and bifupctional carboxylates and mixed acid aldehydes. Studies with mol equiv of C1-'- free 36C10- allowed estimation of covalent binding of C1 by L a t various times, supporting the kinetic findings. For other PUFAs of higher degree unsaturation, the complexity of feasible reactions precluded an analogous approach. As a substitute, reactions of each PUFA at identical molarities and conditions were followed. Using logit-log transforms again, relative apparent rates were computed a t t l 2, showing that increasing unsaturation and chain length decreases reactivity with OC1-. Binding of 36Clshowed that PUFAs of increasing unsaturation are principally destroyed via oxidation.
I ntroductlon The chemical fate of chlorine-based disinfectants and byproducts has been of public health interest. Exogenous formation of chlorinated byproducts of the process, e.g., generation of mutagenic and possibly carcinogenic chlorinated organics from naturally occurring solutes and OC1have been documented ( I ) . As preventive means to control the occurrence of such genotoxic agents, ammonia is added to the treatment
stream postchlorination, converting OC1- to monochloramine (H,NCl). The NC1 moiety of the latter species is a poor chlorinating agent, while the molecule retains somewhat diminished, albeit adequate microbicidal efficacy (2). A more complex issue arises in the endogenous formation of byproducts ensuing from the ingestion of residual chlorine. The significance of this phenomenon is implicit in that over 100 million consumers ingest via drinking
This article not subject to U S . Copyright. Published 1992 by the American Chemical Society
