Chem. Res. Toxicol. 1994, 7, 281-285
281
Communications Metabolic Studies on Haloperidol and Its Tetrahydropyridine Analog in C57BL/6 Mice Cornelis J. Van der Schyf,? Kay Castagnoli,t Etsuko Usuki,j Hassan G. Fouda,s John M. Rimoldi,$ and Neal Castagnoli, Jr.*J Department of Chemistry, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061 -0212, and Pfizer Inc., Central Research, Groton, Connecticut 08340 Received January 4, 1994"
The neuroleptic agent haloperidol (HP) is biotransformed in humans to a pyridinium metabolite, HPP+, that displays neurotoxic properties resembling those of the l-methyl-4phenyl-1,2,3,6-tetrahydropyridine (MPTP)-derived neurotoxic pyridinium metabolite MPP+. We report here that H P and its tetrahydropyridine dehydration product 4-(4-chlorophenyl)1-[4-(4-fluorophenyl)-4-oxobutyll-1,2,3,6-tetr~ydropyridine (HPTP) are metabolized in vivo by the MPTP-susceptible C57BL/6 mouse to several pyridinium metabolites including HPP+ and the 4-(4-chlorophenyl)-l-[4-(4-fluorophenyl)-4-hydroxybutyllpyridinium species RHPP+, the pyridinium species corresponding to reduced haloperidol (RHP), a major circulating metabolite of HP. Atmospheric pressure ionspray (API) mass spectral data also suggest the formation of fluorophenyl ring-hydroxylated derivatives of these two pyridinium metabolites. Furthermore, HPLC tracings reveal the presence of HPP+,RHPP+, and two phenolic pyridinium metabolites in brain tissue extracts of HPTP, but not HP, treated mice. The neurotoxic potential of MPTP-type pyridinium species suggests that these metabolites may contribute to some of the neurological disorders observed in humans undergoing chronic H P treatment.
Introduction Haloperidol (4-(4-chlorophenyl)-1-[4(4-fluoropheny1)4-oxobutyll-6piperidinol (HP; 1))l is a widely used antipsychotic agent ( 1 )that causes avariety of side effects including reversible parkinsonism and tardive dyskinesia (TD), a movement disorder observed in some individuals following months or, more often, years of drug treatment (2). Although dopamine receptor supersensitivityhas been the most widely accepted pathophysiologic explanation for TD (31, several problems with this theory have been noted ( 4 ) including the persistence of TD in some patients following termination of drug treatment (5). In view of the neurotoxic properties of the l-methyl-4-phenylpyridinium species MPP+ (2), the monoamine oxidase-B (MAO-B)-generated metabolite of the parkinsonianinducing agent l-methyl-4-phenyl-l,2,3,6-tetrahydropyridine (MPTP; 3) (6, 7),the biotransformation of HP or ~
* Address correspondence to this author.
its dehydration product 4-(4-chlorophenyl)-l-[4-(4-fluorophenyl)-4-oxobutyll-l,2,3,6-tetrahydropyridine (HPTP; 4) (8)to MPP+-type pyridinium metabolites could be of neurotoxicological importance especially in the pathogenesis of HP-induced TD. Supporting the proposed relevance of this pathway is evidence that H P is biotransformed in humans to the 4-(Cchlorophenyl)-l-[4-(4fluorophenyl)-4-oxobutyl]pyridiniumspeciesHPP+ (5) (9), a compound that displays MPP+-type neurotoxic properties (10, 11). In an effort to characterize further the possible conversion of HP to potential neurotoxic pyridinium metabolites,we have examined the urine and brain of HP- and HPTP-treated C57BL/6 mice, a speciesknown to be susceptible to the neurotoxic properties of MPTP. R
~
t On sabbatical leave from the Department of Pharmaceutical Chem-
RHP(14) R = N a
MPP'(2)
R1 = H Rp = CH,
HPP* (5). R1 = CI: Rz I Ma
R
rbF
istry, Potchefstroom University for Christian Higher Education, Potchef15. R1 = CI; R2 = Mb -CHz(CHz)z H stroom 2520, South Africa. 16db R1 = CI: Rz = Nb t Virginia Tech. Na R = H MPTP (3): Rl = H; R2 CH3 1 Pfizer Inc. Nb R = O H HPTP (4) R1 = CI, Rp = Ma e Abstract published in Aduance ACS Abstracts, March 15, 1994. 'Abbreviations: API, atmospheric pressure ionspray; CID, collisioninduced dissociation;GC-EIMS, gas chromatography-electronionization Materials and Methods mass spectrometry; HP, haloperidol (4-(4chlorophenyl)-1-[4-(4-fluorophenyl)-4oxobutyll-4-piperidinol(1)); HPP+,4-(4chlorophenyl)-1-[4chemicals. HP and Tween 80 (SigmaChemical Co.,St.Louis, (4-fluorophenyl)-4-oxobutyllpyridinium species (5); HPTP, 4-(4-chloand HPLC-gradesolvents (Fisher rophenyb-l-[4-(4-fluorophenyl)-4-oxobutyll-1,2,3,6-tetrahydropyridine MO), 2,4-dinitrochIorobenzene Scientific Co., Springfield, NJ), and 4-bromopyridine hydro(4);HRCIMS, high-resolution chemical ionization mass spectrometry; MAO-B, monoamine oxidase-B; MPP+, l-methyl-4-phenylpyridinium chloride, 4-chloro-l-(4-fluorophenyl)-l-butanone, 4-bromochlospecies;MPTP, l-methyl-4-phenyl-l,2,3,6-tetrahydropyridine (3);RHP, 4-~4-chlorophenyl)-1-14-(4-fluorophenyl)-4-hydroxybutyll-4-piperidi- robenzene, NaNs, and LiAlH4 (Aldrich Chemicals, Milwaukee, WI) were obtained from commercialsources. HPTP was prepared no1 (14); RHPP+, 4-(4-chlorophenyl)-l-[4-(4-fluorophenyl)-4-hydroxybutyllpyridinium species (13);TD, tardive dyskinesia. as reported previously (12). I
Qa93-228xI94/2707-Q281$04.5QJQ 0 1994 American Chemical Society
282 Chem. Res. Toricol., Vol. 7,No. 3, 1994 Chemical Syntheses. (A) General Methods. Synthetic reactions were carried out under a nitrogen atmosphere. Tetrahydrofuran (THF) and diethyl ether were distilled from sodium benzophenone ketyl. Proton NMR spectra were recorded on a Bruker WP 270-MHz spectrometer. Chemical shifts (6) are reported in parts per million (ppm)relative to tetramethylsilane or sodium 3-(trimethylsilyl)propionate-2,2,3,4-d4(for D20) as internal standard. Spin multiplicities are given as s (singlet), d (doublet),t (triplet),or m (multiplet). Jvalues are given in hertz (Hz). Gas chromatography-electron ionization mass spectrometry (GC-EIMS) was performed on a Hewlett Packard 5890 GC fitted with an HP-1 capillary column (24 m X 200 pm X 0.33 pm film thickness) which was coupled to a Hewlett Packard 5870 mass-selective detector. Data were acquired using an HP 5970 ChemStation. Normalized peak heights are reported as a percentage of the base peak. High-resolution chemicalionization mass spectrometry (HRCIMS) was performed on a VG 7070 HF instrument using methane as the reagent gas. Melting points were obtained on a Thomas-Hoover melting point apparatus and are uncorrected. Microanalyses were performed by Atlantic Microlab, Inc., Norcross, GA. (B) 4-(4-Chlorophenyl)pyridine (8). A solution of 4-chlorobromobenzene (4.9 g, 25.6 mmol) in 60 mL of anhydrous THF was added over a 30-min period with vigorous stirring to a mixture of Mg turnings (1.23 g, 51.2 mmol) in 10 mL of THF which had been previously exposed to 12 vapor. After stirring for 1 h at ambient temperature, the resulting @-chloropheny1)magnesium bromide (6)was added via syringe to a solutionof 4-bromopyridine (7; 3.00 g, 19.0 mmol) in 50 mL of anhydrous diethyl ether containing 70 mg (0.064 mmol) of tetrakis(tripheny1phosphine)palladium(0). After heating under reflux for 12hand then cooling to room temperature, the reaction mixture was diluted with 100 mL of diethyl ether and then treated with 50 mL of H2O. The product was extracted several times with diethyl ether, and the combined extracts were dried over NazSO4 and concentrated in vacuo to yield a yellow oil. Purification by flash column chromatography (silica gel, 40% ethyl acetate/60% hexanes) followed by recrystallization from hexanes yielded 2.1 g (60%) of 8 as white needles: mp 90-91 "C [lit. (13)mp 70-71 'H NMR (CDC13) 6 7.45 (d, J = 4.5, ArH, 2H), 7.48 (d, J = 8.7, ArH, 2H), 7.57 (d, J = 8.7, ArH, 2H), 8.67 (d, J = 4.5, ArH, 2H); GCEIMS [isothermal at 100 "C for 2 min followed by a ramp of 25 "C for 6 min (tR = 7.4 min)] m/z 189 (M+,loo), 154 (30), 127 (30), 75 (20), 51 (20). (C) 444-Chlorophenyl)-l-(2,4-dinitrophenyl)pyridinium Chloride (9). A solution of 8 (1.0 g, 5.28 mmol) and 2,4dinitrochlorobenzene (1.07 g, 5.28 mmol) in 10 mL of anhydrous acetone was heated under reflux for 6 h. The reaction mixture was cooled,and the precipitated product was collected and washed with acetone. The filtrate was concentrated, the residue in 5 mL of acetone was heated under reflux for 6 h, and the above sequence was repeated. The precipitated solids were combined and recrystallized from methanol/acetonitrile to yield 1.3 g (64%) of the product as pale yellow needles containing 1 mol of CH30H: mp 197-198 "C; 1H NMR (CDsOD) 6 3.39 (CHsOH, s,3H), 7.72 (d, J = 8.7, ArH, 2H), 8.18 (d, J = 8.7, ArH, 2H), 8.34 (d, J = 8.7, ArH, lH), 8.70 (d, J = 7.2, ArH, 2H), 8.91 (dd, J = 8.7, 2.5, ArH, lH), 9.26 (d, J = 2.5, ArH, lH), 9.27 (d, J = 7.2 ArH, 2H). Anal. Calcd for C1,H11N304Cl&H3OH: C, 50.82; H, 3.53; N, 9.88. Found C, 50.88; H, 3.48; N, 9.96. (D)4-Azido-l-(4-fluorophenyl)-l-butanone (11). A solution containing 5.0 g (25 mmol) of 4-chloro-l-(4-fluorophenyl)1-butanone (10)in 30 mL of dimethylformamide/water (255) was treated with NaN3 (2.43 g, 37 mmol) and then heated at 80 OC for 2 h. The reaction mixture was allowed to come to room temperature and then dissolved in 200 mL of CHzClz and the resulting solution washed successivelywith water and brine, dried over Na2S04, and evaporated to dryness. Purification by flash 2The structure for this compound may be the isomeric 2-(4-chloropheny1)pyridinebased on the method of preparation, which would account for the differences in melting points.
Communications column chromatography (silica gel, 15% ethyl acetate/85% hexanes) yielded 4.2 g of the product as a colorless oil: 'H NMR (CDC13) 6 2.04 (m, CHzCHzNs, 2H), 3.06 (t, J = 7.0, C(O)CHz, 2H), 3.42 (t, J = 6.6, CH2N3, 2H), 7.14 (m, ArH, 2H), 8.00 (m, ArH, 2H); GC-MS [isothermal at 60 "C for 2 min followed by a ramp of 25 "C/min for 6 min ( t R = 8.7 min)] m/z [Mt - 281 179 (2), 138 (lo), 123 (100),95 (60), 75 (20), 56 (5). HRCIMS: calcd for C&l83NO, 208.0886; found, 208.0888. (E)[(~S)-4-Amino-l-(4-fluorophenyl)butanol]~ (COOH), (12). Azide 11 (1.0 g, 4.83 mmol) in 25 mL of anhydrous THF was added dropwise over 30 min to a suspension of LiAlH, (0.550 g, 14.5 mmol) in 50 mL of THF. The resulting reaction mixture was stirred for an additional 30 min at room temperature and then treated dropwise with 10 mL of 2 N KzCO3 followed by 5 mL of HzO. This mixture was filtered through Celite, which subsequently was washed several times with CHZC12. The combined filtrates were washed with water and then brine and dried over NazS04. Evaporation of the solvent yielded the crude product as a pale yellow oil. The residue in ethyl acetate was treated with a solution of oxalic acid in ethyl acetate to yield a solid that was recrystallized from ethanol/water/acetonitrileto yield 800 mg (37%) of the bisamine oxalate as fluffy white crystals: mp 206-207 "C; lH NMR (DzO) 6 1.63 (m, CHzCHpNH2, 2H), 1.82 (m, CHOHCHz, 2H), 3.00 (t, J = 7.1, NHzCH2, lH), 4.74 (t, J = 6.5, CHOH, lH), 7.16 (m, ArH, 2H), 7.40 (m, ArH, 2H); GC-MS [isothermal at 100 "C for 2 min followed by a ramp of 25 "C for 6 min (tR = 6.6 min)] m/z 183 (MHt, 5), 166 (51, 123 (30), 97 (60), 77 (30), 59 (100). Anal. Calcd for C22Hd2N206: C, 57.89; H, 6.62; N, 6.14. Found: C, 57.90; H, 6.63; N 6.14. (F)(R,S)-4-(4-Chloropheny1)- 1-[ 4-(4-fluorophenyl)-4-hydroxybutyllpyridinium Chloride [RHPP+.Cl- (13)]. To a solution of the (dinitropheny1)pyridinium intermediate 9 (490 mg, 1.25 mmol) in 25 mL of 1-butanol was added a solution of amine 12 (free base; 300 mg, 1.6 mmol) in 5 mL of 1-butanol. The resulting blood-red mixture was heated under reflux for 9 h, during which time the color gradually turned to bright yellow. The solvent was evaporated under reduced pressure and the resulting residue dissolvedin 20 mL of CH2C12. Precipitation of the product was accomplished by the addition of diethyl ether. Following two recrystallizations from CHsOH/CH3CN, 375 mg (76%) of pale yellow needles was obtained: mp 205-206 "C; lH NMR (DMSO-&) 6 1.59 (m, CH2CH2NH2, 2H), 1.97 (m, CHOHCH2, 2H), 4.62 (m, CHOH and NH2CH2, 3H), 5.47 (d, J = 4.4, OH, lH), 7.11 (m, ArH, 2H), 7.32 (m, ArH, 2H), 7.71 (d, J = 8.6, ArH, 2H), 8.12 (d, J = 8.6, ArH, 2H), 8.53 (d, J = 6.9, ArH, 2H), 9.16 308 (e = 25 300). Anal. (d,J = 6.9, ArH, 2H); UV (CHsOH) A, Calcd for CzIH&lzFNO: C, 64.30; H, 5.14; N, 3.57. Found: C, 64.30; H, 5.14; N, 3.50. Animal Studies. HP (10mg/kg) and HPTP (50 mg/kg) were administered intraperitoneally in 0.2 mL of sterile distilled water containing 20% Tween 80 to groups of 12 retired breeder male C57BL/6 mice (Harlan-Sprague Dawley, Dublin, VA) weighing approximately 30 g each. A control group of 12 vehicle-treated mice was included in the study. For urine collections, animals were housed in individual metabolic cages designed to separate urine and feces. Food and drinking water were supplied ad libitum. Urine samples were collected during the 48-h period following the injection of drug or vehicle. For the LC/MS/MS analyses, urine collections from three HP- or HPTP-treated animals were pooled. At the termination of the urine collections, animals were sacrificed by cervical dislocation and the whole brains (from which the striata had been removed) were isolated and weighed. The samples were stored at -70 O C prior to analysis. Analytical Studies. The individual and pooled urine samples were processed by a method slightly modified from that reported previously (14). In brief, the samples were treated with 0.5 M KZHPO4to adjust the pH to 8.5 and then were loaded (up to 400 pL dependent upon urine output) onto a Millipore Sep-pak (Millipore-Waters, Bedford, MA). The loaded Sep-paks were washed with 2 mL of Milli-Q UV Plus purified water (MilliporeWaters), following which the analyte was eluted using 1 mL of
Chem. Res. Toxicol., Vol. 7, No.3,1994 283
Communications HP Treated
HPTP Treated Urine
Table 1. API LC/MS/MD CID Fragmentation Data for HP and HPTP Metabolites Present in CS7BL/6 Brain and Urine
HPTP Treated Brain
~
~~
CID Fragment Ions F R
s t
P"
R
tH2
; M1M3M4 I ,
R
.._.. ....-._._.
0 4
.
812 0 4 8 1 2 0 4 812 Ret. time (min)
Figure1. HPLC/fluorescencetracingsof urine extractsobtained from C57BL/6 mice treated with vehicle only or HP (left panel, single animal), or HPTP (center panel, pooled urine also used in API-LC mass spectral analysis), and a brain extract of a C57BL/6 mouse treated with HPTP (right panel). methanol followed by 2 mL of 0.5% acetic acid/methanol. The combined eluenta were evaporatedto dryness under a stream of nitrogen, and the residue was dissolved in the HPLC mobile phase (see below) using 2 times the loaded sample volume. For the HPLC/fluorescence studies, these solutionswere diluted (up to 1in 500)to give concentrationsof analytesthat would produce peak heightswithinthe recording range ofthe integratorfollowing injection of a 50-pL aliquot. HPLC analyses were performed using a Waters Wisp 717 autosampler, pBondapak C18 10 pm, 150 mm X 3.9 mm column, a Perkin Elmer LS-40 fluorescence detector, LCI-100 integrator, and a Beckman llOA pump set at a flow rate of 1 mL/min. The mobile phase was prepared as follows: to 400 mL of CH&N and 600 mL of 10 mM NHdOAc were added 2.2 mL of triethylamine and 0.8 mL of glacial acetic acid. In a separate analysis,solid NaBH, (0.3 mg) was added to 200-pL aliquots of the diluted analytical urine samples. These mixtures were vortexed and then were allowed to stand at room temperature for 30 min, following which they were reanalyzed using the above assay. The pooled urine extractswere analyzed by atmosphericpressure ionization (API)LC/MS/MS on a SCIEX API I11triquadrupolemass spectrometeras described previously (9). The samples obtained from the Sep-pakswere dissolved in 500 pL of the mobile phase, and 50-pL aliquota were introduced via a Rheodyne 7413 injectionvalve onto the same HPLC column using essentiallythe same HPLC conditionsas described above. The ionspray interface was set at 4000 V and the nebulizer pressure at 40 psi air. The collision energy was -30 V for collisioninduced dissociation (CID) experiments. The extent of autoxidation of HP and HPTP was examined in human urine samples spiked with 1 mM drug, which were allowed to stand at room temperature for 48 h and then analyzed for HPP+. Brain samples were processed by adding to each sample 4 mL of 1.15%aqueous KCl/g of brain tissue followed by homogenization using a glass PotterElvehjem homogenizer. An equal volume of 2% glacial CH&OzH/98% CHsCN was added, and the homogenates were vortexed and allowed to stand at room temperature for 10 min. The samples were vortexed again and then centrifuged in an Eppendorf microcentrifuge at 14 000 rpm for 4 min. A 1-mL aliquot was removed from the resulting supernatant and evaporatedunder nitrogen. The residue in 200 pL of the mobile phase described above was vortexed and again centrifuged at 14000 rpm for 4 min. A 50-pL aliquot of the supernatant was analyzed by the HPLC/fluorescence assay.
Results and Discussion The HPLC/fluorescence tracings of the urine extracts obtained from HP- and HPTP-treated animals (Figure 1) show a number of fluorescent peaks, designated as M1M5, which are not present in the tracings of the corresponding vehicle-treated mouse urine extracts. Consistent with pyridinium-type structures, treatment of the urine extracts from drug-injected animals with NaBH4
Metabolite M112 (16ajh) M3 (15) M4 (RHPF? 13) M5 (HPP*, 5)
M+ 372/374 3701372 3561356 3541356
1901192 (I) NP 1901192 (i) 161 (R = OH: v) 1901192 (I) NP 190/192 (I) 165 (R I H: I/)
NP 165 (R = O H vilwb) 139 (R = OH' vi) NP NP 149 (R = H: iv) 123 (R = H: H i ) NP
caused these peaks to disappear, presumably due to conversion to the corresponding tetrahydropyridine derivatives. On the basis of its coelution with the synthetic standard ( t =~12.7 min), M5 was identified tentatively as HPP+ (see Table 1for mass spectral data). Confirmation of this assignment was obtained by API LC/MS and LC/MS/MS analysis of pooled urine extracts from animals receiving HP. The peak corresponding to M5 displayed a pair ofions at mlz 354 and 356, the 36Cl-and 37C1-containingparent ions, respectively, expected for HPP+. The CID daughter ion spectrum of the mlz 356 ion contained a weak fragment ion at mlz 192 ( 5 % ),assigned to the ([37C11chlorophenyl)pyridinium species i, and strong fragment ions a t mlz 165 (100%)and 123(45%)whichpreviouslyhadbeenassigned to structures ii and iii (9). The possibility that the HPP+ detected in these urine samples was due to autoxidation of H P or HPTP was ruled out by chemical stability studies using human urine samples spiked with 1mM H P or 1 mM HPTP and stored in air for 48 h. The peak heights corresponding to HPP+ observed in the HPLC/fluorescence tracings were less than 1% of those observed in the tracings obtained with the experimental urine samples. An additional peak was observed in the mlz 356 ion chromatogram which corresponds to the HPLC peak with t~ = 8.2 min (M4, Figure 1). The CID daughter ion spectrum of this ion displayed fragment ions a t mlz 190 (75%) and 149 (100%). The mlz 190 fragment ion may be assigned to the ([3sC11chloropheny1)pyridinium species i while the fragment ion a t mlz 149 is likely to be derived from the fluorophenyl-containing portion of the molecule. One plausible structure (iv) for this fragment ion would result from cleavage of the (chloropheny1)pyridyl moiety and dehydration of a carbinol group formed from metabolic reduction of the carbonyl group present in HP. These considerations led us to the possibility that M4 is the pyridinium species 13 (RHPP+, [WlIM+ = 356) which could be formed by metabolic oxidation of reduced haloperidol (4-(4-chlorophenyl)-1-[4-(4-fluorophenyl)-4hydroxybutyl] -4-piperidinol (RHP; 1411, a well-documented metabolite of H P (151,and/or by reduction of the carbonyl group present in HPP+. Further clarification of the structure of M4 was sought by synthesis of RHPP+ (Scheme 1). The key intermediates for the preparation of 13 were the 1-(2,4-dinitrophenyl)pyridinium species 9 and 4-amino-l-(4-fluorophenyl)butanol (12). The amino alcohol 12 was obtained by reduction of the corresponding azidoketone 11,which in turn was prepared from the commercially available chloro compound 10 and NaN3. The (dinitropheny1)pyridinium intermediate 9 was obtained by reaction of 4-(4-chloropheny1)pyridine(8)with 2,4-dinitrochlorobenzene.Compound 8 was synthesized by a palladium-catalyzed cross-
284 Chem. Res. Toxicol., Vol. 7, No. 3, 1994
Communications
Scheme 1. Synthetic Pathway Leading to the Reduced Haloperidol Pyridinium Species RHPP+ (13)
10 R.CI 11 R = N 3
12
F'
12
F
6
7
8
S
13
coupling reaction (16)between (4chlorophenyl)magneaium bromide (6) and 4-bromopyridine (7). Coupling of 9 and 12 according to a procedure first described by Zincke (17) gave RHPP+as its chloride salt. The HPLC t R of synthetic 13 was identical to that of M4. Full confirmation of the structure of M4 was obtained by API LC/MS which displayed the expected parent ions for the W P C 1isotopes at masses 356/358. The CID spectra of these parent ions displayed fragment ions at mlz 190/192 and 149,to which we now assign the structures i and iv, respectively. Since the parent ions of both HPP+ and RHPP+ fragment under API LC/MS/MS conditions to generate the 4- (4-chloropheny1)pyridiniumfragment ion i at mlz 190, an mlz 190 selected-ion chromatogram was reconstructed from the spectra of the pooled urine extracts obtained from HP- and HPTP-treated mice to search for additional pyridinium metabolites. The tracing from the H P urine sample showed a peak ( t R = 7.6 min) corresponding to M3 which eluted at the leading edge of the M4 peak, and an earlier eluting peak ( t R = 4.0 min) corresponding to M1. The API LC/MS spectrum of M3 displayed parent ions at mlz 370/372 (36C1/37Cl), i.e., 16 mass units higher than those observed for HPP+, while M1 displayed similar ions a t m/z 372/374 (Table l ) , Le., 16 mass units higher than those observed for RHPP+. Analysis of these results led us to propose hydroxylated derivatives of HPP+andRHPP+ for these new metabolites of HP. In addition to the weak ion a t mlz 190 (5%),the CID spectrum of mlz 370 from metabolite M3 showed fragment ions at mlz 181 (100%) and 139 (75%), i.e., 16 mass units higher than fragment ions ii and iii observed in the corresponding CID spectrum of the [36ClIHPP+ parent ion. These data suggested that M3 is a fluorophenyl-hydroxylated derivative 15 of HPP+, in which case the CID fragment ions at mlz 181and 139may be assigned to structures v and vi. Similar analysis of the mlz 372 ion for M1 showed, in addition to the strong fragment ion at mlz 190 (70%;i), a fragment ion a t mlz 165 (loo%),i.e., 16 mass units higher than the RHPP+-derived base peak fragment ion iv. On the basis of this analysis we suggest 16a and viia as tentative structures of M1 and the mlz 165 fragment ion, respectively. The reconstructed mlz 190 ion chromatogram observed in the HPTP experiment also showed peaks for M1 and M3. An additional peak was observed in this tracing which corresponded to M2 ( t =~5.9 min). The CID spectrum of the ion a t mlz 372 displayed fragment ions at mlz 190 (70%)and 165 (100%; viib) consistent with a regioisomer (16b) of M1.
Oida and co-workers (18)have published evidence for the presence of three conjugated metabolites of H P in human urine. In addition to the well-known piperidinol O-glucuronide, phenolic O-sulfate and O-glucuronide conjugates of a fluorophenyl ring-hydroxylated metabolite derived from RHP were well characterized by field desorption mass spectral and 'H NMR analysis. Therefore, it may be reasonable to speculate that the pyridinium metabolites observed in our studies are generated from the corresponding piperidinols and/or tetrahydropyridines. We also have attempted to determine if any of these pyridinium metabolites may be present in brain tissue of HP- or HPTP-treated mice. The HPLC/fluorescence tracings of mouse brain homogenate extracts from HPtreated mice were essentially identical to those from control animals. On the other hand, all of the corresponding tracings of the HPTP-treated mouse brain extracts showed peaks eluting at the retention times of HPP+ (M5),RHPP+ (M4), M2, and M1 (Figure 1). The relative peak heights of these metabolites are different from those seen in the chromatograms of the HPTP-treated mouse urine extracts, which raises the possibility that these pyridinium metabolites may accumulate selectively in the brain or may be formed in the brain. The recent characterization of cytochrome P450 2D1 in canine striatal homogenates (19) opens the possibility that monooxygenases capable of catalyzing the oxidation of tetrahydropyridines may be present in the brain. In summary, the results presented in this communication document that C57BL/6 mice metabolize H P and HPTP to HPP+, RHPP+, and what appear to be phenolic pyridinium species. Furthermore, HPLC/fluorescence analysis provides evidence for the presence of HPP+, RHPP+ and two phenolic pyridinium metabolites in the brain tissues of HPTP-treated animals. The presence of these metabolites in the HPTP- but not HP-treated mouse brain homogenates might argue that the tetrahydropyridines may be intermediates in the formation of the correspondingpyridinium metabolites, at least in the brain. Although more detailed information regarding the metabolic pathways leading to these compounds is required, these preliminary results suggest that oxidative aromatization of partially oxidized piperidine ring systems such as H P and HPTP may represent a more common metabolic route than previously recognized. Acknowledgment. This study was supported by the National Institute of Neurological and Communicative Disorders and Stroke (NS 28792) and the Harvey W. Peters Center for the Study of Parkinson's Disease. C.J.V.d.S. acknowledgessupport from the South African Foundation for Research Development (FRD), the Medical Research Council (MRC),and Potchefstroom University. We thank Mr. Kim Harich for the HRCI mass spectral data and the staff of the Virginia Tech Laboratory Animal Resources Facility for technical support.
References (1) Baldessarini, R.J. (1991) Drugs and the treatment of psychiatric disorders. In The PharmacologicalBasis of Therapeutics,pp 383435, Pergamon Press, New York. (2) Gerlach,J., and Casey, D. E. (1988)Tardive dyskinesia.Acta Psych. Scand. 77,369-378. (3) Casey, D. E. (1991) Neuroleptic drug-induced extrapyramidal syndromes and tardive dyskinesia.Curr. Opin. Psychiatry4,8649.
Communications (4) Kornhubner, J., Riederer, P. I., Reynolds, G. P., B e c k ” , H., Jellinger, K., and Gabriel, E. (1989) SH-Spiperone binding sites in post-mortem brains from schizophrenic patients: relationship to neuroleptic drug treatment, abnormal movements, and positive symptoms. J. Neural Transm. 75,l-10. (5) Glazer, W. M., Moore, D. C., Schooler, N. R. Brenner, L.. M., and Morgenstern, H. (1984) Tardive dyskinesia. Arch. Gen. Psychiatry 42, 623-627. (6) Markey, S. P., and Schmuff, N. R. (1986) The pharmacology of the Parkinsonian syndrome producing neurotoxin MPTP (l-methyl4-phenyl-1,2,3,6-tetr&ydropyridine)and structurally related compounds. Med. Res. Rev. 6,389-429. (7) Langston, W. J. (1989) Mechanisms underlying neuronal degeneration in Parksinson’s Disease: an experimental and theoretical treatise. Movement Disorders Suppl. 1, 4, S15-525. (8) Fang, J., and Gorrod, J. W. (1991) Dehydration is the first step in the bioactivationof haloperidolto its pyridinium metabolite. Tozicol. Lett. 59, 117-123. (9) Subramanyam, B., Pond, S. M., Eyles, D. W., Whiteford, H. A., Fouda, H. G., and Castagnoli, N., Jr. (1991) Identification of a potentially neurotoxic pyridinium metabolite in the urine of schizophrenic patients treated with haloperidol. Biochem. Biophys. Res. Commun. 181, 573-578. (10) Rollema, H., Subramanyam, B., Skolnik, M., d‘Engelbronner, J., and Castagnoli, N., Jr. (1991) MPP+-like neurotoxicity of HPP+, a pyridinium metabolite of haloperidol: a microdialysis study. In Monitoring Molecules in Neuroscience, pp 367-372, University Center for Pharmacy, Groningen, The Netherlands. (11) Rollema, H., Skolnik, M., d’Engelbronner, J., Igarashi, K., Usuki, E., and Castagnoli, N., Jr. (1993) MPP+ like neurotoxicity of a
Chem. Res. Toxicol., Vol. 7, No. 3, 1994 286 pyridinium metabolite derived from haloperidol: in vivo microdialvsis and in vitro mitochondrial studies. J.Pharmacol. EXD.Ther. 268 (I), 380-387. (121 Subramanvam. B.. Woolf. T.. and Castaenoli. N.. Jr. (1991)Studies on the in vkro conversion‘ofhaloperidofto a b k n t i d l y neurotoxic pyridinium metabolite. Chem. Res. Toxicol. 4,123-128. (13) Butterworth, E. C., Heilbron, I. M., and Hey, D. H. (1940) Arylpyridines. Part 11. Some substituted phenylpyridines. J. Chem Soc., 355-358. (14) Igarashi, K., and Castagnoli, N., Jr. (1992) Determination of the pyridinium metabolite derived from haloperidol in brain tissue, plasma and urine by high-performance liquid chromatography with fluorescence detection. J. Chromatogr. Biomed. Appl. 579, 277283. (15) Forsman,A.,andLarseon,M. (1978)Metaboismofhaloperidol. Curr. Ther. Res. 24, 567-568. (16) Negishi, E., Luo, F., Frisbee, R., and Matsushita, J. (1981) A regiospecific synthesis of carbosubstituted heteroaromatic derrivatives via Pd-catalyzed cross coupling. Heterocycles 18, 117-123. (17) Genieson, Y.,Marazano, C., Mehmandoust, M., Gnecco, D., and Das,B. C. (1992) Zincke’s reaction with chiral primary amines: A practical entry to pyridinium salts of interest in aeymmetric syntheses. Synlett, 431-434. (18) Oida, T., Terauchi, Y.,Yoshida, K., Kagemoto, A.,and Sekine, Y. (1989) Use of antisera in the isolation of human specific conjugates of haloperidol. Xenobiotica 19, 781-793. (19) Tyndale, R. F., Sunahara, R., Inaba, T., Kalow, W.,Gonzalez, F. J., and Niznik, H. B. (1991) Neuronal cytochrome-P450IIDI (debriaoquine/sparteine-type)-potent inhibition of activity by (-)-cocaine and nucleotide sequence identity to human hepatic P45O-gene CYP2D6. Mol. Pharmacol. 40, 63-68.
