Synthesis and use of deuterated 3,4-dihydroxyphenylglycolaldehyde as an internal standard for determination of dopegal in brain tissue by gas chromatography-mass spectrometry

Article abstract of DOI:10.1006/bioo.1996.0015

Trideuterated 3,4-dihydroxyphenylglycolaldehyde (DOPEGAL-d₃) was synthesized from tetradeuterated catechol. After derivatization with N- methyl-N-trimethylsilyltrifluoroacetamide, the identity of DOPEGAL-d₃ was confirmed by comparison with unlabeled DOPEGAL by gas chromatography-mass spectrometry (GC-MS); they have the same pattern of mass fragmentation. The deuterated DOPEGAL was used as an internal standard for the determination of DOPEGAL in human locus ceruleus by GC-MS in the selected ion monitoring mode.

Full text of DOI:10.1006/bioo.1996.0015

BIOORGANIC CHEMISTRY 24, 169–177 (1996)
ARTICLE NO. 0015
Synthesis and Use of Deuterated 3,4-
Dihydroxyphenylglycolaldehyde as an
Internal Standard for Determination
of Dopegal in Brain Tissue by Gas
Chromatography–Mass Spectrometry
S
HU
W
EN
LI,* VINCENT T. SPAZIANO,† WILLIAM H. ELLIOTT,‡ AND
W
ILLIAM J. BURKE
*
Departments of *Neurology, ‡Biochemistry, and †Chemistry, Veterans Administration Medical Center,
and Saint Louis University Health Sciences Center, School of Medicine, St. Louis, Missouri 63110
Received January 11, 1996
Trideuterated 3,4-dihydroxyphenylglycolaldehyde (DOPEGAL-d3) was synthesized from
tetradeuterated catechol. After derivatization with N-methyl-N-trimethylsilyltrifluoro-
acetamide, the identity of DOPEGAL-d3 was confirmed by comparison with unlabeled
DOPEGAL by gas chromatography–mass spectrometry (GC–MS); they have the same pattern
of mass fragmentation. The deuterated DOPEGAL was used as an internal standard for
the determination of DOPEGAL in human locus ceruleus by GC–MS in the selected-ion-
monitoring mode.
1996 Academic Press, Inc.
INTRODUCTION
3-4-Dihydroxyphenylglycolaldehyde (DOPEGAL) is formed by oxidative deami-
nation of noradrenaline (NA) by monamine oxidase (MAO-A) (1, 2). In tissues this
aldehyde is either enzymatically reduced to 3,4-dihydroxyphenylglycol (DOPEG) or
oxidized to 3,4-dihydroxymandelic acid (1, 3). Although estimates of this compound
in tissues have been made (3–6), it has not been previously measured in human
brain using gas chromatography–mass spectrometry (GC–MS). This has been due
to lack of deuterated DOPEGAL internal standard.
We have reported both biological (6) and chemical (7) synthesis of DOPEGAL.
This DOPEGAL standard enabled the development of a method using high-perfor-
mance liquid chromatography with electrochemical detection (HPLC–EC) to mea-
sure tissue levels of DOPEGAL (6). However, GC–MS is a more reliable quantita-
tive method of measuring extremely low levels of bioactive compounds in tissue.
Capillary gas chromatography with selected-ion-monitoring (SIM) mass spectrome-
try affords a combination of great selectivity and sensitivity. The high accuracy can
be obtained with the stable isotope dilution method. Using deuterated analogues
as the corresponding internal standard, SIM provides an ideal basis for specific
quantitative analysis by GC–MS.
We present here the synthesis and use of deuterated DOPEGAL (1) as an
169
0045-2068/96 $18.00
Copyright 1996 by Academic Press, Inc.
All rights of reproduction in any form reserved.

170
LI ET AL.
S
CHEME 1. (a) N-Methylformanilide, POCl3 ; (b) dihydropyran, pyridinum p-toluenesulfonate; (c) 2-
lithio-1,3-dithiane; (d) pyridinum p-toluenesulfonate; and (e) HgCl2 , HgO, acetonitrile.
internal standard to determine the concentration of DOPEGAL in tissue samples
by GC–MS–SIM. We apply this method to the measurement of DOPEGAL in
human brain tissue from the locus ceruleus (LC).
The synthesis of deuterated DOPEGAL (DOPEGAL-d3) (1) was performed as
illustrated in Scheme 1. The deuterated catechol-d4 (2) used as the starting material
was prepared from catechol and deuterated water according to the procedure
described by Clemo and Swan (9). The deuterated 3,4-dihydroxybenzaldehyde-d3
(3) was prepared from 2 by utilizing N-methylformanilide–POCl3 . The synthetic
route from deuterated 3 to the desired compound (1) was based on that of unlabeled
DOPEGAL which we published (7). However, we optimized and modified all of
the reaction conditions and some of the reaction agents were changed, such as
using pyridinum p-toluenesulfonate instead of p-toluenesulfonic acid, and butyl
lithium instead of lithium diisopropylamide. The last two reactions, removal of
the protective group, tetrahydropyranyl, and unmasking of the desired carbonyl
function, could be combined into one step by using mercury chloride and mercuric

DETERMINATION OF DOPEGAL IN BRAIN TISSUE
171
oxide. Therefore, the purity of products was much better and they were easy
to purify.
EXPERIMENTAL
Chemicals and Materials
DOPEGAL was prepared according to our previous paper (7). All other chemi-
cals were purchased either from Aldrich (Milwaukee, WI) or from Sigma (St. Louis,
MO). The brain tissue was obtained from the Saint Louis University Alzheimer
Brain Bank. Melting points were determined on a Fisher–Johns melting point
apparatus and uncorrected. IR spectra were recorded on a Perkin–Elmer 1320
spectrophotometer. Molecular masses were determined via a VG 70-250 SEQ hy-
brid-tandem spectrometer. All HPLC analyses were done using a Waters 6000A
multisolvent delivery system with an Applied Biosystems Model 783 programmable
detector. Column chromatography was performed on silica gel 60 (70–230 mesh)
and TLC analysis on silica gel 60 F254 precoated plates (layer thickness 0.2 mm).
Compounds for which only mass spectral data are given have purities Ͼ98% as
measured by TLC or HPLC.
Synthesis of DOPEGAL-d3 (1)
3,4-Dihydroxybenzaldehyde-d (3). A solution of catechol-d4 (2) (11.4 g, 0.1 mol)
3
in a mixture of N-methylformanilide (12.4 ml, 0.11 mol) and benzene (60 ml) was
treated with phosphorus oxychloride (10 ml, 0.11 mol) at 10–15ЊC. The reaction
mixture was stirred at room temperature overnight and then heated to 65–70ЊC
for 2 h. The supernatant was poured out, and the residue was mixed with ice water,
neutralized with 40% NaOH, and extracted with ethyl acetate. The extract was
washed with saline, dried over anhydrous Na2SO4 , and concentrated. The crude
compound was purified by chromatography over silica gel using 2% CH3OH in
CH2Cl2 as eluting solvent. All the fractions corresponding to product were combined
and cooled to 0ЊC. The separated solid was collected; yield 4.2 g (30%); mp 151–
153ЊC. MS calcd for di-TMS derivative (C13H19D3O3Si2) of 3,4-dihydroxybenzalde-
hyde-d3 : 285.00. Found: 285 (Mϩ).
3,4-Ditetrahydropyranoxybenzaldehyde-d3 (4). Pyridinum p-toluenesulfonate
(500 mg) was added to a solution of 3,4-dihydroxybenzaldehyde-d3 (7.05 g, 0.05
mol) and dihydro-2H-pyran (11 ml, 0.12 mol) in CH2Cl2 (20 ml). The reaction
mixture was stirred for 5 h at room temperature, then washed with 10% NaHCO3
and saline and dried over anhydrous Na2SO4 . Removal of the solvent gave the
crude compound which was purified by chromatography on silica gel column eluting
with 30% CH2Cl2 in hexane. The desired fractions were pooled and concentrated
to afford 13 g (84%): mp 72–74ЊC. MS (FAB)ϩ calcd for (C17H19D3O5): 309.00.
Found: 310 (MHϩ).
2-(3,4-Ditetrahydropyranoxy-Ͱ-hydroxybenzyl)-1,3-dithiane-d3 (5). A solution of
1,3-dithiane (1.19 g, 9.9 mM) in 40 ml of THF was treated with 3.96 ml of 2.5 M
butyl lithium at Ϫ60ЊC under nitrogen. The reaction mixture was stirred at Ϫ40ЊC

172
LI ET AL.
for 2 h and then cooled to Ϫ70ЊC. A solution of 4 (3.06 g, 9.9 m
M) in THF
(10 ml) was added to the above solution of 2-litho-1,3-dithiane at Ϫ70ЊC. After
that, the reaction mixture was stirred at Ϫ60ЊC for 2.5 h and then allowed to
come to room temperature overnight. Removal of the THF at room temperature
gave a residue which was treated with ice water and extracted with CH2Cl2 .
The extract was washed with water, dried over anhydrous Na2SO4, and concen-
trated. The crude compound was purified by chromatography over silica gel
using 30% hexane in CH2Cl2 as eluting solvent. All the fractions corres-
ponding to the product were pooled and concentrated: yield 2.5 g (58.8%); IR
1
(Nujol) 3450 (OH) cmϪ . MS (FABϩ) calcd for (C21H27D3S2O5): 429.00. Found:
430 (MHϩ).
2-(3,4-Dihydroxy-Ͱ-hydroxybenzyl)-1,3-dithiane-d3 (6). Pyridinium p-toluenesul-
fonate (140 mg) was added to a solution of 5 (2.4 g, 5.6 m ) in 20 ml of 70%
M
CH3OH. The reaction mixture was stirred at room temperature for 5 h and
extracted with ethyl acetate. The extract was washed with 10% NaHCO3 and
water and dried over anhydrous Na2SO4 . Removal of solvent gave the crude
compound which was recrystallized with alcohol to afford 1.29 g (88%); mp
99–100ЊC. MS (FABϪ) calcd for (C11H8D3S2O3): 261.00. Found: 260 (M-H)ϩ.
DOPEGAL-d3. Method A: To a solution of 5 (690 mg, 1.62 m ) in 70 ml
M
of 80% acetronitrile was added mercury chloride (1.4 g) and mercuric oxide
(1.5 g). The reaction mixture was refluxed for 3 h and filtered. The acetonitrile
solution was concentrated and extracted with ethyl acetate. Removal of the
solvent gave the crude product which was purified by chromatography over silica
gel eluting with 1% CH3OH in CH2Cl2 . The desired fractions were pooled and
concentrated to afford 174 mg (62.8%); mp 105–107ЊC; NMR (DMSO-d6) ͳ 9.50
(d, 1H, CH). MS (FABϩ) calcd for (C8H5D3O4): 171.00. Found: 172 (MHϩ).
Method B: Mercury chloride (906 mg) and mercuric oxide (718 mg) were
added to a solution of 6 (321 mg, 0.75 m ) in 35 ml of 80% acetonitrile. The
M
reaction mixture was refluxed for 3 h and extracted with ethyl acetate. The
extract was washed with saline, dried over anhydrous Na2SO4 , and concentrated.
The crude compound was purified by chromatography over silica gel eluting
with 1% CH3OH in CH2Cl2 . The desired fractions were pooled and concentrated
to afford 82 mg (64%); mp 105–107ЊC [Lit. 105–107ЊC (7)]. MS (FABϩ) calcd
for (C8H5D3O4): 171.00. Found: 172 (MHϩ).
Derivatization of DOPEGAL and Calibration Graphs
For direct analysis of standards, the pure compounds were dissolved in
acetonitrile and derivatized under nitrogen with N-methy-N-trimethylsilyltri-
fluoroacetamide (MSTFA) at 70ЊC for 1 h. A standard curve was generated by
adding various amounts of unlabeled DOPEGAL to a fixed concentration of
deuterated DOPEGAL. After derivatization, 1 Ȑl of the mixture of DOPEGAL
and deuterated DOPEGAL was injected into the GC–MS. Calculations were
made on the basis of the DOPEGAL/DOPEGAL-d3 peak-area ratios using
the ions at m/z 355 and 358, respectively. The regression equation is y ϭ
1.057␹ ϩ 0.521 and the correlation coefficient (r2) is 0.9846.

DETERMINATION OF DOPEGAL IN BRAIN TISSUE
173
Gas Chromatography–Mass Spectrometry
A Hewlett–Packard GC5890 Series II gas chromatography coupled with a HP
5971 mass selective detector was used. The gas chromatograph was equipped with
Alltech Econo Cap SE-54 capillary column (30-m ϫ 0.25-mm i.d. ϫ 0.25-Ȑm film
thickness). The injector temperature was 250ЊC. Helium was used as carrier gas
(40 kPa). The initial column temperature of 70ЊC was held for 2 min and then
increased at 20ЊC/min to a final temperature of 280ЊC. Data processing was per-
formed using a Hewlett–Packard G1034C MS-ChemStation with Microsoft Win-
dows Version 3.1 software. Data were stored onto a Vectra QS/20 computer. The
mass spectrometer was operated at an electron energy of 70 eV and an EMV 2980.
For the identification of trimethylsilyl derivatives of DOPEGAL spectra were taken
in the mass range 50–500 with a scan speed of 1.5 decade/s.
Tissue Extraction and Derivatization
The LC from a 77-year-old man was dissected at autopsy after a postmortem
interval (PMI) of 3.8 h. The brain tissue was homogenized in 20 volumes of acetone
containing DOPEGAL-d3 as an internal standard and kept in an ice bath for 15
min to allow protein precipitation. The homogenate was centrifuged (11,000g, 4ЊC,
15 min). The supernatant acetone solution was passed through a 0.2-Ȑm filter into
a reaction vial and evaporated under a nitrogen stream at 70ЊC. When the volume
reached 50% of the starting volume, acetonitrile was added and concentration was
continued. Finally, methylene chloride was added to remove all traces of water as
an azeotropic mixture under a nitrogen stream at 70ЊC. MSTFA (25 Ȑl) was added
to the residue. Derivatization proceeded under nitrogen at 70ЊC for 1 h. Protein
concentration was determined on an aliquot of tissue by the method of Lowry (8).
RESULTS
GC–MS Analysis
The mass spectrum of the trimethylsilyl (tri-TMS) derivative of unlabeled
DOPEGAL, shown in Fig. 2, shows the molecular ion at m/z 384 (C17H32O4Si3)ϩ.
Loss of a methyl radical produces the ion at m/z 369. It is generally accepted that
the (M-15)ϩ ion observed in TMS derivatives is formed by loss of a methyl group
from the TMS moiety. Loss of an aldehyde radical (M-CHO) produces the ion at
m/z 355 (C16H31O3Si3)ϩ. It is the strongest ion of DOPEGAL and a characteristic
ion for DOPEGAL and we selected it as a monitor for determination of DOPEGAL.
Loss of a hydrogen and TMS radical from m/z 355 (M-CHO-TMSH) produces the
ion 281 (C13H21O3Si2)ϩ. The ion at m/z 147 (TMS-OuSi(CH3)2)ϩ is observed as
expected for a tri-TMS derivative. The peak at m/z 267 arises from m/z 355 by the
loss of TMSCH3 . The route of formation of the ion at m/z 193 can be explained
by the loss of a methyl, an aldehyde group, and 147 (M-15-29-147). The results of

174
LI ET AL.
TABLE 1
Mass Measurement of Tri-TMS Derivative of DOPEGAL
Formula
Calculated Found
Ion assignments
(M)ϩ
(MUCH3)ϩ
(MUCHO)ϩ
C17H32O4Si3
384.00
369.00
355.00
295.00
281.00
267.00
193.00
384
369
355
295
281
267
193
C16H29O4Si3
C16H31O3Si3
C14H23O3Si2
C13H21O3Si2
C12H19O3Si2
C10H13O2Si
(MUTMSO)ϩ
(M-CHO-TMSH)ϩ
(M-CHO-TMSCH3)ϩ
(M-CHO-TMSOTMS)ϩ
measurement on the molecular ion and major fragments are given in Table 1. The
mass spectrum of the tri-TMS derivative of DOPEGAL-d3 shown in Fig. 1 shows
the same pattern observed from DOPEGAL.
Analysis of DOPEGAL in Brain Tissue
The method described above was used to measure the amount of DOPEGAL
in human brain tissue. Treatment of the tissue and sample preparation was as
previously described, except as noted here. The MSD control program was config-
ured for SIM during the interval of 3–13.5 min postinjection. Under the GC condi-
tion applied, the retention time for TMS derivatives of DOPEGAL was 11.3 min,
for norepinephrine, 7.5 min, for epinephrine, 7.8 min. for 3,4-dihydroxyphenylglycol-
TMS, 13 min, and for dihydroxy mandelic acid-TMS, 13.3 min. Therefore, interfer-
FIG. 1. Mass spectrum of tri-TMS derivative of deuterated DOPEGAL.

DETERMINATION OF DOPEGAL IN BRAIN TISSUE
175
F
IG. 2. Mass spectrum of tri-TMS derivative of unlabeled DOPEGAL.
ence of the other compounds bearing the catechol/benzylhydroxy structure is ex-
cluded. The chromatographic peak areas at retention time 11.3 min for ions at
m/z 355 and 358 were integrated by computer and ratios of those areas calculated.
After determination of appropriate area ratios and comparison to the calibration
graph, the level of DOPEGAL in human brain tissue was found to be 2.2 Ϯ 0.1
ng/mg wet tissue or 119 ng Ϯ 5 ng/mg protein.
CONCLUSION
The synthesis of deuterated DOPEGAL is important for several reasons. We
report here for the first time the quantitation of DOPEGAL in human locus ceruleus
using GC–MS techniques. The human brain in a number of disorders, such as
Parkinson’s disease and Alzheimer’s disease, is studied primarily using postmortem
tissue. Deficits in catecholamines, including NA, are thought to play a role in these
disorders (12–14). DOPEGAL is the precursor of the other major brain metabolites
of NA, DOPEG, and 3-methoxy-4-hydroxyphenylglycol (MHPG) (15). Carlsson
and Winblad have suggested that in postmortem brain the sum of a catecholamine
plus its major metabolite is a better index of the level of catecholamine in brain at
the time of death than the catecholamine alone (16). This is supported by the
findings of Warsh et al. (17) that catecholamines are metabolized in brain after
death. The GC–MS measurement of DOPEGAL will therefore contribute to a
more accurate estimation of premortem NA levels in postmortem human brain. In
addition, because the enzymes which metabolize DOPEGAL (aldehyde reductase,
aldehyde dehydrogenase, and catechol-O-methyltransferase) either are in very low
amounts (18) or are extraneuronal (18, 20, 21). DOPEGAL may be an index of

176
LI ET AL.
the intraneuronal metabolism of newly formed NA (22). The concentration of
DOPEGAL reported here is comparable to that reported for MHPG, the major
NA metabolite in rat LC (11).
The amount of LC DOPEGAL is 14 times that reported for NA and 57 times the
reported value for MHPG in postmortem human LC (10). Several methodological
differences may have led to this variation. The earlier report (8) used HPLC–EC
instead of GC–MS to measure LC catechols. The mean delay between death and
autopsy was 4.5 times as long as the PMI used here. In addition, there was a delay
of up to 2 h prior to freezing at Ϫ70ЊC (10). In our study tissue was frozen immedi-
ately at autopsy and kept at Ϫ140ЊC. This difference is especially important in view
of the report of further metabolism of catechols after death (17). Finally, because
DOPEGAL is synthesized in neurons, it should be found primarily in neurons and
not glial tissue. However, most of the tissue dissected is nonneuronal. Therefore,
some variation in amounts of DOPEGAL expressed as ng/mg tissue could be due
to differences in dissection technique. Regardless of these differences, our results
indicate that DOPEGAL is a major NA metabolite in postmortem human brain.
DOPEGAL is the initial MAO-A metabolite of NA. Blashko (23) suggested
that aldehyde metabolites of MAO action on amines may be toxic to cells in which
they are formed. Markey et al. (24) have shown that the exogenous toxin 1-methyl-
4-phenyl-1,2,4,6-tetrahydropyridine, which causes a form of Parkinson’s disease,
becomes toxic to neurons only after its oxidation to the pyridinium formed by
MAO-B. We have shown that DOPEGAL is toxic to neurons both in vivo and in
vitro (25). The concentration of DOPEGAL we report here in locus ceruleus tissue
is only 2.3-fold lower than the concentration of 5.0 Ȑg/ml which we found toxic to
differentiated PC-12 cells (25). We have also reported a 2.2-fold increase in an
enzyme in the biosynthetic pathway of DOPEGAL in C-1 epinephrine neurons
(26) in Alzheimer’s disease (AD). The synthesis of deuterated DOPEGAL thus
allows us to determine if DOPEGAL has reached toxic levels in adrenergic neurons
in AD.
ACKNOWLEDGMENTS
This work was supported by the Veterans Affairs Research Program. The authors thank the Alzheimer
Brain Bank of Saint Louis University for supplying brain tissue. We express our thanks to Dr. F. A.
Bencsach, University of South Alabama, for aid in molecular masses and to Dr. J. Shoemaker, Depart-
ment of Biochemistry, Saint Louis University, for his very valuable advice. We are also indebted to
Mrs. Cathy Smith for excellent technical assistance.
REFERENCES
1. DUNCAN, R J. S., AND
SOURKES, T. L. (1974) J. Neurochem. 122, 663–669.
2. DUNCAN, R. J. S. (1975) Can. J. Biochem. 53, 920–922.
3. DAVIS, V. E., CASHAW, J. L., M
CL
AUGHLIN, B. R., AND
HAMLIN, T. A. (1974) Biochem. Pharmacol.
23, 1877–1889.
4. DAVIS, V. E., WALSH, M. J., AND
YAMANAK, Y. (1970) J. Pharmacol. Exp. Ther. 174, 401–411.

DETERMINATION OF DOPEGAL IN BRAIN TISSUE
177
5. TURNER, A. J., ILLINGWORTH, J. A., AND
T
6. BURKE, W. J., MATTAMMAL, M. B., MARSHALL, G. L., AND
180, 79–84.
IPTON, K. F. (1974) Biochem. J. 144, 353–360.
C
HUNG, H. D. (1980) Anal. Biochem.
7. LI, S. W., ELLIOTT, W. H., AND BURKE, W. J. (1994) Bioorg. Chem. 22, 337–343.
8. LOWRY, O. H., ROSEBROUGH, N. J., FARR, A. L., AND
RANDALL, R. J. (1951) J. Biol. Chem.
193, 265–275.
9. CLEMO, G. R., AND SWAN, G. A. (1942) J. Chem. Soc. 395.
10. CASH, R., DENNIS, T., L’HEUREUX, R., RAISMAN, R., JAVOY-AGID, F., AND
SCATTON, B. (1987)
Neurology 37, 42–47.
11. EMOTO, H., KOGA, K., ISHII, H., YOKOO, H., YOSHIDA, M., AND
TANAKA, M. (1993) Brain Res.
627, 171–176.
12. RIEDERER, P., BIRKMAYER, W., SEEMAN, D., AND
13. ADOLFSSON, R., GOTTFIES, C. G., OVERLAND, L., ROOS, B. E., AND
W
UKETICH, S. (1977) J. Neural Transm. 41, 241–251.
INBLAD, B. (1978) in Alzheimer’s
W
Disease: Senile Dementia and Related Disorders (Katzman, K., Terry, R. D., and Bick, K. L., Eds.),
pp. 441–451, Raven Press, New York.
14. BURKE, W. J., CHUNG, H. D., NAKRA, B. R. S., GROSSBERG, G. T., AND
Neurol. 22, 278–280.
JOH, T. H. (1987) Ann.
15. SHARMAN, P. F. (1973) Br. Med. Bull. 29, 110–115.
16. CARLSSON, A., AND
17. WARSH, J. J., GODSE, D. D., L
18. RIS, M. M., AND VON ARTBURG, J. P. (1973) Eur. J. Biochem. 37, 69–77.
W
INBLAD, B. (1976) J. Neural Transm. 38, 271–276.
I
, P. P., AND HEUNG, S. W. (1981) J. Neurochem. 36, 902–907.
S
W
19. GROTE, S. S., MOSES, S. G., ROBINS, E., HUDGENS, R. W., AND CRONINGER, A. B. (1974) J. Neurochem.
23, 791–802.
20. DUNCAN, R. J. S., SOURKES, T. L., DUBROVSKY, B. O., AND
21. MARSDEN, C. A., BROCH, O. J., AND ULDBERG, H. C. (1972) Eur. J. Pharmacol. 19, 35–42.
22. RUTLEDGE, C. O., AND ONASON, J. (1967) J. Pharmacol. Exp. Ther. 157, 493–502.
QUIK, M. (1975) J. Neurochem. 24, 143–147.
G
J
23. BLASCHKO, H. (1952) Pharmacol. Rev. 4, 415–453.
24. MARKEY et al. (1984) Nature (London) 311, 464–467.
25. BURKE, W. J., L
I, S. W., ELLIOT, W. H., SCHMITT, C. A., CHUNG, H. D., AND
HARING, J. (1994)
Soc. Neurosci. 20, 248.
26. BURKE, W. J., CHUNG, H. D., MARSHALL, G. L., GILLESPIE, K. N., AND
Geriatr. Soc. 38, 1275–1282.
JOH, T. H. (1990) J. Am.