The JOURNAL OF PHARMACOLOGY ANt) EXPERIMENTAL THERAPEUTICS \‘oI. 193, No.

3
Copyright 1975 by The \Viilinrns & Wilkins Co. Printed in U.S.A.

EFFECTS OF CHLORIMIPRAMINE AND LYSERGIC ACID

DIETHYLAMIDE ON EFFLUX OF PRECURSOR-FORMED
3H-SEROTONIN: CORRELATIONS WITH SEROTONERGIC
IMPULSE FLOW’

DOROTHY W. GALLAGER AND GEORGE K. AGHAJANIAN

Departments of Psychiatry and Pharmacology, Yale University School of .lledicinc and 1/iC

Connecticut Mental Health Center, New Haven, Connec1iii1

Accepted for publication February 3, 1975

ABSTRACT
GALLAGER, DOROTHY W. AND GEORGE K. AGHAJANIAN: Effects of chlorimipramimme
and lysergic acid diethylamide on efflux of prescursor-formed :Hserotonifl: Corre-
lations with serotonergic impulse flow. J. Pharmacol. Exp. Timer. 193: 785-795,
1975.

Time effects of d-lysergic acid diethylamide (LSD) and chiorimipramine (CIMI) on the
firing rate of serotonergic (5-HT) neurons and on the in vivo efflux of 5-HT were in-
vestigated in parallel. A cerebroventricular perfusing technique was used to measure the
efflux of 3H-5-HT formed in vivo from 311-tryptophan. Impulse flow in serotonergic
neurons was monitored by single unit recording from raphe (5-HT) neurons. Doses of
LSD and CIMI, which caused a similar degree of inhibition of raphe cell firing, were
foiiimd to affect 5-HT efflux differently. LSD, at both the 75, and 150 pg/kg doses, pro-
duiced a similar decrease in 3H-5-HT efflux. In contrast, CIMI at a low dose (5 mg/kg)
did not reduce 5-HT efflux, despite an inhibition in impulse flow. At a high dose (20
mg/kg), CIMI produced an increase in 3H-5-IIT efflux. We conclude that 1) LSD
decreases 3H-5-HT efflux by directly inhibiting impulse flow in 5-HT neurons and/or
by a local effect oim 5-HT terminals and 2) a low dose of CIMI produces no net change
iim 3H-5-HT efflux because a reduction in impulse flow-depeimdent 5-HT release com-
pensates for blockade by CIMI of 5-HT reuptake.

Time imahlucinogenic ageImt, d-lysergic acid di- at., 1972). The systenmic administratioim of both
etimylamide (LSD), and time tricyclic antidepres- LSD and CIMI uniformly inimibits rapime neu-
saimt, cimlorimipraniine (CIMI), have been ronal firing. Since release of serotonin
simown to have similar effects on time firing rate (5-hmydroxytryptamine; 5-HT) is based at least
of serotonergic neurons located in time midbrain im part on impulse flow (And#{233}n et at., 1966;
rapime area (Agimajaimian et at.. 1968; Sheard et Simeard and Agimajanian, 1968; Holman and
Received for publication September 24, 1974. Vogt, 1970, 1972), botim drugs might be
1 This work was supported in part by Grants expected to decrease time release of 5-liT from
MH-17871 and PHS-NS-05706 and by the State of serotonergic nerve termmals. in vito studies
Coimimecticut.
suggest timat LSD decreases 5-HT release
Send reprint requests to: Dorotlmy W. Gahhager, (HatmdIc and Padjeim, 1971 ; Tilsoim and Sparber,
Department of Psycimiatry, Yale University School
of Medicine, CMHC, 34 Park St., New Haven, 1972). Parallel in vivo studies using CIMI imave
Coimim. 06508. not been done. However, coimtrarv to expecta-

785
786 GALLAGER AND AGHAJANIAN Vol. 193

tions, in iii vitro experiments CIMI imave been in vivo release studies. However, the use of
showim to cause an increased 5-HT “efflux” or precursor-formed indoles has not yet been
release (Farneho and Hamberger, 1971). applied to in vivo 5-HT release studies. This
Despite timese differeimces iim timeir effects on labeling technique provides major advantages:
release, hotim CIMI aimd LSD imave similar 5-HT formed from tryptophan Imas been shown
effects on 5-HT metabolism. In vivo biocimemi- to be specific for serotonergic neurons (Moir
cal stuidies have simown tlmat botim drugs increase aimd Eccleston, 1968; Kuimar et at., 1972; Agima-
endogermous levels of 5-HT and decrease endo- janian aimd Asimer, 1971 : Agimajaimian et at.,
genous levels of 5-hydroxyindoleacetic acid (5- 1973) . In addition, in vitro studies by Snod-
HIAA) (Hosecrans et al., 1967; Freedman et grass and Iverson (1974) imave simown that
at., 1970; Halaris et al., 1973). CIMI and LSD release of 3H-tryptopimatm (3H-TR.P) along with
have also been reported to cause a reduction in 3H-5-HT is associated witlm nonspecific tissue
5-HT synthesis (Lin et at., 1969; Scimubert et damage or to unphysiological effects. Thus
at., 1970a,b; Andn et at., 1968; Corrodi and measuriimg 3H-TRP efflux as well as precursor-
Fuxe, 1969; Shields and Eccleston, 1973; formed 3H-5-HT provides a check on the spec-
Hamoim et at., 1974). ificity of the effects observed.
In the present studies, we imave investigated
in parallel time effects of LSD and CIMI on the Methods
firing rate of 5-HT neurons and on the in vivo Experimental subjects and recording studies.
efflux of 5-HT. While it is imot et. tecimnically Male albino rats (Charles River, 220-280 g) were
possible to measuire directly time syimaptic release used in timese experiments. For purposes of
of a transmitter in vito, it. is possible to measure recording, animals were anesthetized with cimloral
imydrate (400 mg/kg i.p.) and placed in a stereo-
time efflvx of a transmitter, wimicim is time net
taxic apparatus. Animals were kept anesthetized
result of time complex of events: release, reup-
witim additional injections of chloral hydrate as
take and/or metabolism. Previous in vivo stud-
needed timroughout the experiment. In previous
ies on release and metabolism have generally studies, it has been shown timat time dose of LSD
used doses of LSD and CIMI (Rosecrans et at., required to inhibit rapime neurons in chmhorah
1967; Aimd#{233}im
et al., 1968; Lin et at., 1969; imydrate-anesthetized rats (Aghajanian et al., 1970)
Corrodi and Fuxe, 1969; Schubert et at., was virtually identical to time dose in unanestime-
1970a,b; Handle and Padjen, 1971; Tilson and tized rats (Haigler and Agimajanian, 1974). Micro-
Sparber, 1972; Halaris et at., 1973) which are pipettes filled with 2 M NaCl saturated witim fast
muclm higimer timan timose needed to inhibit firing green dye (tip diameters 1-2 ham) were lowered
into the midbrain raphe region witim a hydraulic
(Agimajanian et at., 1968; Simeard et al., 1972)
microdrive system. Electrode potentials were
making time relatioimsimip between firing rate and
passed througim a imigh-impedence amplifier and
efflux unclear. We have given low enough doses
monitored on an oscilloscope. Integrated firing
of botim drugs to permit at least partial recov- rate was recorded by means of a rate meter as
ery from total inhibition of cell firing during described previously (Aghajanian et al., 1970).
the period of perfusion. In addition, by Rat temperature was monitored by Tele-
employing doses of LSD and CIMI wlmich cause Thermometer and maintained between 36-37#{176}C.
equivalent inimibit ion of raphe cell firing, we can Under these conditions, raphe units display a reg-
begin to separate impulse-flow dependent ular rhytimm and a rate of 0.5 to 2.5 spikes/second

release from otimer possible effects of these as described previously (Agimajanian et al., 1968).
After observing and recording a base-line firing
drumgs.
rate for at least 5 minutes, drugs were adminis-
Because endogimeoums 5-HT efflux from seroto-
tered intraperitoneahhy. At time end of each experi-
imergic nerve terimminals is low (Feldberg and
ment, time site of the electrode tip was marked by
\Iyers, 1966) aimd ofteim masked by time pres-
iontophmoretic ejection of fast green (Timomas and
ence of imoim-neuronal 5-lIT (Portig aimd Vogt, Wihson, 1965). Precise location of electrode tip
1969), \ imave use(l precursor-formed ‘Fl-5-HT was hater determined by imistohogical examination.
as a marker for endogenous 5-UT. Chiuehm and Ventricular implantation and perfusion. Hats
)doore (1973) have reported on the use of were anestimetized as above and placed in a ster-
catecimols svntimesized from ‘H-tvrosiime iim their eotaxic apparatus. Two cannuhae were inmplantcd
19;,5 CIMI AND LSD: EFFECTS ON 5-HT EFFLUX 787

in eacim rat : a l)obetim3lene cannula (constructed labeled tryptopiman and 5-HT collected in perfus-
and implanted according to deBahbian Verster et ate samples were quite variable between animals.
a!., 1971) for time intraventricuhar administration Therefore, four 10-minute perfusate samples were
of ‘H-TRP and a Gaddum type push-pull can- collected prior to the administration of drugs so
nuia for perfusion (Gaddum, 1961) . Time lateral timat each animal could serve as his own control.
ventricle was chmosen as the perfusion site for two The data were timen expressed as a percentage of
reasons. First, the possibility of direct tissue time fourths (control) period. Statistical significance
damage i)T time perfusion cannula (Cimase and was calculated by Student’s t test (two-tailed).
Kopin, 1968 : Portig and Vogt, 1969) is decreased Assay of perfusate. The perfusate samples (0.5
i) ventricular placement. Secondly, studies of ml/sanmple) were lyophmilized to dryness and redis-
Ricimards et al. (1973) have demonstrated the solved in 100 /Ll of 80% ethanoL After centrifuga-
presence of serotonergic nerve terminals lining the tion, 50 /Ll of each sample were applied to silica gel
rat cerebral ventricles. By both fluorescence and (LQ6D Linear_QTM, Quantum Industries, Fairfield,
electron microscopy, we imave found that these N.J.) thin-layer plates. Since the radioactivity as-
nerve ternminahs are abolished by midbrain raphe sociated with tryptophan was in great excess, a sol-
lesions establishing this area as the origin of these vent system in wimicim H-TRP remained near the
5-HT ternminahs (Agimajanian and Gaihager, 1975). origin was chosen to prevent streaking and mask-
The pusim-puhi cannula was constructed from ing of the smaller 3H-5-HT peak. However, no
stainless steel tubing time inside tube being 27 single chromatograpimic system was avaihimble in
gauge and time outside 20 gauge. Optimal protru- whicim both 3H-5-HT and 3H-5-HIAA were sepa-
simm of time inner tube was 0.5 mm (Szerb, 1967). rated cleanly from H-TRP. Although attempted,
Two adjacent burr imoles were drilled at 1.5 and 2 the ion exchange method of Scimubert (1974) was
nmimm lateral to time intersection of the coronal and not satisfactory for separating H-TRP from
saggitah sutures (bregma zero) and time push-pull H-5-HIAA and H-5-HT in perfusate samples.
and ventricuihar cannuilae were inserted below the H-TRP was in sucim excess that its 0.1 to 02%
skull surface, 4.2 and 3.5 mm respectively. After overlap with 3H-5-HIAA using multiple excimange
iwrimmanent attachment to the skull, an inner (pro- resins masked time small amounts of time labeled
tection) styhet was inserted into the outer can- acid metabolite. Thus, attempts to quantitate
nuia. After time wound was closed, the animals this indole metabolite were abandoned. The thin-
were allowed to recover from the anesthesia. Six- layer plates were developed in ethyl acetate-n-
teeim to 20 hmouirs after the implantation of timese propanoh-10% ammonia (4:3:1) for 2.0 hours
canimulae, H-TRP (02 mc specific activity: 4.6 and time Hf values for 5-HIAA, TRP and 5-HT
c/mnmoh. Schmwarz-Mann, Orangeburg, N.Y.) was were 024, 027 and 0.65, respectively (Aures et al.,
injected via time polyethelene cannula into the lat- 1968). Time remaining 50 ph from eacim sample were
eral ventricle in a volume of 20 p1. Perfusion with counted in Aquasoh (New England Nuclear Cor-
artificial cerebrospinal fluid (Merlis, 1940) was poration, Boston, Mass.) to determine total
hegun timrougim time adjacent push-pull cannula 1 radioactivity. Since radioactivity was limited to
imour after time adnministration of labeled trypto- positions coinciding with timose of added authmentic
piman iim time nonaimesthetized, unrestrained rats. A 5-HT and TRP (Kuihar et al., 1971), the fluores-
coimstant perfumsion rate of 50 ph/mm was main- cent areas associated with cold carriers of the
tained by a Manostat peristaltic pump (Manostat, respective indoles were visualized under uiltra-
New York, N.Y.). Consecutive 10-minute perfusate violet light, scraped into counting vials containing
samples were collected in test tubes containing 0.1 Aquasoh and quantitated by scintillation counting.
ml of 1% c steine, 200 eg of cold TRP and 100 pg Recoveries from known standards run simultane-
of cold 5-HT. A variation of more than 10% ously were 60% for TRP and 80% for 5-HT.
h(tween time voluimme infused and that recovered Rats were pretreated for 30 minutes witim Ro 4-
was 1 akeim as evidence of a faulty perfusion and 4602, in a dose sufficient to totally block 5-HT
lime data were discarded. At time end of eacim ‘en- synthesis (800 mg/kg i.p.; B#{233}dard et a!., 1971).
tricular perfusion, rats were anethetized and per- After the injection of H-TRP, no radioactivity
fused timrougim time heart with 10% buffered for- in perfusate samples from timese animals was
immahiim. Serial frozen sections (50 pm in thickness) found to he associated with the autlmentic 5-HT
were (ut, mounted and stained to verify cannuha peak. This further indicated timat time radioactivity
piait’tmmeimt. Time area of cannutlation corresponded coinciding with authentic 5-HT was ‘H-5-HT.
to the frontal planes A7470 through A6450 of Determination of endogenous substances.
K#{246}nigand Khippeh (1970). Although four animals Endogenous TRP, 5-HT and 5-HIAA were ann-
were ruim sinmultaneously. the absolute amounts of hyzed in time same tissue sanmples by a coinbina-
 788 GALLAGER AND AGHAJANIAN Vol. 193

tion of time column method of Schubert et al. firing for approximately 1 hmour (fig. 1, bottom
(1970a) and the fluorimetric o-phthalaldehyde con- trace). As previously reported (Haigler and
densation metimod of Maickel and Miller (1966). Agimajanian, 1973), the inimibition of rapime
Recoveries for tryptophan, 5-HT and 5-HIAA firing by LSD was not found to be associated
were 100, 71 and 90%, respectively. with a decrease in action potential size. This
Drugs and radioactive materials. Chhorimi-
was also observed with CIMI indicating that
pramine hydrochloride (Anafranii, Ciba Pharma-
neitimer drug produces its inhmibitioim by a local
ceutical Company, Summit, N.J.), Ho 4-4602 ([N-
(DL-seryh) - N - (2,3,4- trihydroxybenzyl) hydrazine],
anesthetic effect.
Hoffmaimn-La Roche, Inc., Nutley, N.J.) and d- Perfusion studies. We attempted to deter-
lysergic acid dietimylamide bitartrate were admin- mine how timese drugs, at doses wimich imave sim-
istered i.p., doses of drugs were expressed in ilar effects on raphe firing, would affect 5-HT
terms of their free base. H-tryptophan (4.6 effiux. LSD (75 pg/kg i.p.) administered at the
c/mmol, generally labeled) was obtained from
mmd of time control perfusion period produced a
Schwarz-Mann. Prior to use, 1.0 ml of the ethanol-
significant decrease (P < .025) in the effiux of
water solution was evaporated to dryness and
‘H-5-HT (fig. 2). Higher doses of LSD (150
redissolved in 100 pl of artificial cerebrospinal
flumid. Twenty microhiters of timis solution were and even 300 pg/kg, not shown) were not able
injected intraventricularly in aim approx mate con- to reduce furtimer time efflux of 3H-5-HT (mean
centratmon of 10 mc/pi. of six rats at each dose ± SE.). 3H-5-HT
efflux did not return to base himme but remained
Results depressed and parallel to time coimtrol curve
Recording studies. Time spoimtaimeous rates thmrougimout time perfusion period. In a separate
of neuroimal units tested in time dorsal rapime groump of animals, 75 pg/kg (i.p.) of LSD pro-
ranged from 0.5 to 2.5 spikes/sec. Admimmistra- ducecl a total inimibition of rapime firing for
tion of an intraperitoneal close of 5 mg/kg of approximately 63 minutes (fig. 2, shaded bar).
CIMI produiceci a total inimibitioim of rapime cell LSD did imot alter the efflux of 3H-TRP at any
firing for approximately 1 imour (fig. 1, top time during the perfusion period (table 1).
trace). An i.p. iimjection of LSD (75 pg/kg) In coimtrast, CIMI at a dose of 5 mg/kg i.p.
also produced a total ilmimii)itioim of raphe cell did not., except for oime point, significaimtlv alter

DM1 (5mg/kg)

1501

0 ib5b 515

V)
Ui

0 LSD( 75 jig/kg)
V)
1501

ib 5b s ob o

Time After Injecton(minutes

FIG. 1. A comparison of time response of cells in time (lorSal raPime nucleus to time s’stemic adnministra-
tion of CIMI aimd LSD. The arrow in botim cases marks time point at wimicim drugs were adnministered
i.p. In the top trace, a rapime cell is totally inimibitecl for 55 minutes by the i.p. injection of 5 mg/kg
of CIMI. Jim time lower trace another rapime cell is inlmibited for 60 minutes by time i.p. injection of 75
pg/kg of LSI). Both tracs shown lmere are integrated rate records, with time ordinate indicating time
firing rate in spikes per imminute of imeurons in time dorsal rapime nucleus.
 19;,5 CIMI AND LSD: EFFECTS ON 5-HT EFFLUX 789

LSD

.-.. Control
‘0

75p9/kgLSD
#{149}___. l5OygfkgLSD

‘C

w5C
I
V
I -“ ““ Roph. inhibition

I I I
0 20 4 60 80 100

MINUTES AFTER DRUG

FIG. 2. Time effects of LSD (75 and 150 pg/kg, i.p.) on the spontaneous effiux of H-5-HT into the
ventricular perfusate and on the firing rate of raphe neurons. H-TRP (200 pc) was injected intra-
ventricularly 1 hour prior to the start of perfusion. LSD was administered after a predrug perfusion
period. The mean absolute amount of H-5-HT recovered in the predrug perfusion period was 0.66 ±
0.31 pc/10 mm (75 pg/kg of LSD; N = 6) or 0.49 ± 0.09 nc/10 mm (150 pg/kg of LSD; N = 6).
Time mean absolute amount of H-5-HT recovered from the perfusate of the same period for con-
trol animals was 0.25 ± 0.07 nc/10 mm, N = 11. H-5-HT efflux in subsequent samples was expressed
as a percentage of the respective predrug amounts of H-5-HT. In a separate group of rats, repre-
sented by the shaded bar, the response of raphe cell firing was followed after an i.p. dose of 75 pg/kg
of LSD. Dorsal rapime cells were totally inhibited for 63.3 ± 2.9 minutes (N = 10). *Indicates a
sigimificant difference from control levels (P < .05).

the effiux of 3H-5-HT from that of control rats was negligible. To confirm timis in our system,
(fig. 3), although this dose of CIMI totally we administered time aromatic L-amino acid
inhibited raphe neuronal firing for approxi- decarboxvlase inimii)itor, Ho 4-4602, in a dose
mately time same duration as time 75 pg/kg dose sufficient to block totally 5-HT synthesis (800
of LSD (fig. 3, shaded bar). However, CIMI at mg/kg i.p.; B#{233}dard et at., 1971). Timis synthesis
a (lose of 20 mg/kg produced a significant inhibitor did not significantly alter time rate of
iimerease P < .05) in :IH_5HT efflux (mean of 3H-5-HT effiux as compared to saline-treated
six rats at each dose ± SE.). This increased animals (mean of six rats ± SE., fig. 4). This
efflux of 3H-5-HT returned to control levels indicates that 3H-5-HT synthesis was negligible
approximately 90 minutes after time administra- during the perfusion period suggesting timat the
tion of CIMI. CIMI did not alter the efflux of efflux of preformed 3H-5-HT was the only phe-
3H-TRP from control values at any time nomenoim being measured in these experiments.
durumg time perfusion period (table 1).
Other control experiments were also necessi-
Both LSD and CIMI have been reported to tated by time use of precursor-formed 311-5-IIT.
inhibit 5-HT synthesis (Lin et at., 1969; While the intraventricular administration of
Scimubert et at., 1970a,b; And#{233}n et at., 1968; labeled TRP permitted imigh specific activities
Corrodi and Fuxe, 1969). Since effiux of 5-HT of the precursor and product to be measuired in
was measured by time presence in the perfusate the limited area of brain sampled by perfuision,
of ill_S_UT formed from 3H-TRP, it was nec- it was assumed timat the fate of labeled 5-liT
essmry to eliminate time possibility that 3H-5-HT reflected timat of time endogenous amine. It has
syimthesis was occurring during the perfusion been shown that when TRP is employed as the
immterval. Scimubert and Sedvall (1972) have precursor, the product, 5-HT, is specific for
reported that 1 hour after precursor adminis- serotonergic neurons (Moir and Eccleston,
tration, syntimesis of 3H-5-HT from 3H-TRP 1968; Kuhar et at., 1972; Aghtajanian and
790 GALLAGER AND AGHAJANIAN Vol. 193

TABLI 1 behave similarly. TRP (100 pg) given intraveim-

Spont(z neou.s ‘H-tryptophan (fllu.r in control a tricularly 1 hour prior to sacrifice significantly
(/rug-treate(/
rats” increased endogenous tryptopiman and 5-HIAA
levels in rat brain. Timese data suggest that the
Time Treal iiient b
label did imot represent a loading dose” of
after
injection TRP at time time of perfusion.
Saline LSI) CIMI lb 4-4602
It Imas also been postulated that certain drugs
?7011
could alter TRP availability (Tagliamonte et
0 100) 100 l0() 100
at., 1971; Bruinvels, 1972). Since this could
10 97 ± 10 78 ± 9 82 ± 9 90 ± 4
result in changes of brain 5-HT as well, 3H-
20 81 ±9 87 ± 0) 77 ± 8 76 ± 5
8 79 ± 15 67 ± 4 THP effiux was measured after eacim drug. In
3() 64 ± 6 66 ±
40 55 ±4 66 ± 11 61 ± 8 62 ± 4 addition, Snodgrass and Iverson (1974) have
50 46 ±5 52 ± 9 57 ± 8 53 ± 3 simown thmat release of the precursor, 3H-TRP,
(jo 43 ±4 50 ± 9 52 ± 7 49 ± 6 in addition to 3H-5-HT, is associated witim non-
70 39 ±6 40 ± 8 49 ± 9 42 ± 6 specific, non-neuronal effects. Since 3H-TRP
SO 38 ±3 48 ± 11 49 ± 6 47 ± 4 effiux was not affected by any drug treatment
9() 29 ± 3 40 ± 10 48 ± 12 38 ± 4
(table 1), time data suggest timat 3H-TRP accu-
100 30 ±4 36 ± 10 41 ± 7 35 ± 4
mulation in tissues was not significantly altered
110 32 ±2 33 ± 7 39 ± 8 38 ± 5
by these drugs and that cimanges in 3H-5-HT
efflux were specific drug effects.
a A 1)Ulse dose of ‘H-TRP (200 /2c) was itmjected
intraventricularly 60 miimutes prior to the start of
Discussion
perfusion. One huimdred nmitmutes after ‘H-TRP or
40 immitmutes after the start of perfusiomm, saline, LSD Our results show that LSD and CIMI, at
(75 pg/kg), CIMI (5 ing/kg) or Ro 4-4602 (800 (loses wimicim cause a similar degree of inhibition
tug/kg) was adimministered i.p. of raphe cell firing, affected 5-HT effiux differ-
The nmean absolute aimmounts of 3H-TRP re- ently: a decrease occurred after LSD but there
covere(l in predrug perfusiorm periods were: saline was no cimange in effiux with the low dose of
14.7 ± 3.S ne/lO limits, N = II; LSD 10.4 ± 2.3
CIMI.
nc/10 immin, N = 6; CIMI 10.7 ± 1.5 nc/10 mm,
N = 6; Ru 4-1602 13.5 ± 3.0 ime/lO mm, N = 6. In studies in wimich LSD ‘iv’ls administered to
‘H-TRP efflux in subsequeimt saimmples was expressed perfused rats, at both the 75 and 150 pg/kg
as the immean percetmtage of the respective predrug doses, a similar decrease in 3H-5-HT effiux was
aimmourmts of 3F1-TRP ± SE. observed. While other investigators have sug-
gested that LSD blocks release of 5-HT (Cimase
et ci., 1967; Farnebo and Hamberger, 1971;
Ashmer, 1971; Agimajanian et at., 1973). How- Tilson and Sparber, 1972), thmere is little infor-
ever, several investigators (Asimcroft et at., mation about how the hallucinogenic agent
1965; Moir and Ecciestoim, 1968; Fernstrom accomplisimes timis. Haigler and Agimajanian
and Wurtmaim, 1971) imave denmonstrated timat (1974) have shown that LSD causes a direct
TRP availability can alter brain 5-HT levels inhibition of raphe cell firing. This iimhibition of
aimd tuirtmover. To be sure that steady-state con- firing could lead to a reduction in impulse flow-
ditions were imot affected by our labeling proce- dependent 5-liT release from serotonergic ter-
dure, we evaluated time effects of the 8.9-pg minals as suggested by our data. However,
dose of labeled TRP on endogenous levels of LSD imas also been shown to decrease time for-
indoles iim brain (table 2). One hmour after mation of 5-liT formed from labeled trypto-
intravent ricular administration of this Pulse piman (Lin et at., 1969; Schubert et at., 1970a).
dose of ‘II-TRP, time endogemmous levels of the It is therefore significant that altimough effiux of
precursor (TRP) product (S-FIT) and meta- 3H-5-HT was decreased by LSD, 3H-5-HT
boiite (5-HIAA) were not altered. Nonlabeled synthesis was negligible during the perfusioim
TRP in approximately time sanme dose (10 pg) period. Timus inhibition of synthesis is not
also hmad iso effect. on eimdogeimoums indole levels, sufficient to explain the reduction of 3H-5-HT
suggesting timat labeled aimd autimentic TRP effiux by LSD.
1975 CIMI AND LSD: EFFECTS ON 5-HT EFFLUX 791

15’

0
CIMI

0
U-. Control
‘0 o_____o 5mg/kg CIMI

100 .___. 2Omg/kgClMI

‘C
D

5c

P.riod tntnl Rook. inhibition

*

0 20 40 60 80 100

MINUTES AFTER DRUG

FIG. 3. Effects of chlorimipramine (CIMI, 5 and 20 mg/kg i.p.) on the spontaneous efflux of H-5-HT
measured in ventricular perfusate and on the firing rate of raphe neurons. CIMI was administered
i.p. (5 or 20 mg/kg) after a predrug perfusion period (N - 6). The mean absolute amount of H-
5-HT recovered in the predrug perfusion period is 0.18 ± 0.05 nc/10 mm (5 mg/kg of CIMI) or 0.14
± 0.03 nc/10 mm (20 mg/kg of CIMI). The mean absolute amount of H-5-HT recovered from the
perfusate of the sanme period for control animals was 025 ± 0.07 nc/10 mm, N = 11. H-5-HT efflux
in subsequent samples is expressed as a percentage of the respective predrug amounts of H-5-HT.
In a separate group of rats, represented by the shaded bar, the response of raphe cell firing was
nmeasured after an i.p. dose of 5 nmg/kg of CIMI. Dorsal raphe cells were totally inhibited for 56.2 ±
3.6 minutes (N = 8).

0
R04- 4602

‘0
B #{149}Control
0--- --o R04-4602

‘C

50
I

I I I I I
0 20 40 60 80 100
MINUTES AFTER DRUG

FIG. 4. A comparison of the spontaneous efflux of H-5-HT in control rats to the effiux of H-5-HT
after Ho 4-4602. H-TRP (200 pc) was injected intraventricularly 1 hour prior to the start of per-
fusion. At the end of a predrug perfusion period, Ro 4-4602 (800 mg/kg i.p.) was administered to rats
and ventricular perfusion was continued for 11 additional collection periods of 10 minutes. The mean
absolute amount of H-5-HT recovered in the predrug perfusion period was 0.28 ± 0.11 nc/l0 mm;
N = 6. The mean absolute amouint of H-5-HT recovered from the perfusate of same period for con-
trol aninmals was 0.25± 0.07 nc/10 mm; N = 11. H-5-HT efflux in subsequent samples is expressed
as a percentage of this predrug H-5-HT.
792 GALLAGER AND AGHAJANIAN Vol. 193

TABLE 2 the diminished efflux of 3H-5-liT during the

(‘oncent rations of tryptophan, .5-lIT and #{244}-HIAAin period of recovery from LSD.
whole rat brain tissue following the intraventricvlar CIMI at a dose (5 mg/kg) that totally
administration of tryptophan inhibited raphe neuronal firing for approxi-
mately time same duration as LSD did not
Tryptopliana Tryptophan 511T 511 fAA
decrease the effiux of 3H-5-HT. Furthermore,
, suit, ± S.D g/g ± S.D. g/g ± S.D. CIMI at a dose of 20 mg/kg produced a signif-
(I 4.59 ± 0.47 0.02 ± 0.14 0.27 ± 0.04
icant increase in 3H-5-HT efflux. Thus, it is
8.9b 5.OG ± 0.40 0.66 ± 0.08 0.28 ± 0.04
1(1 5.0(1± 0.18 0.68 ± 0(4 0.28 ± 0.03 clear timat changes in firing rate cannot alone
its) 7.40 ± 0.63 0.66 ± 0.07 0.44 ± 0.23a explain the dose-dependent effects of CIMI on
effiux.
‘ Tryt)tophan was administered in a constant volume of 20 tl
1(1 l)olethe13’rle cannuia implanted according to the method of
It. has been postulated that CIMI inhibits
deBalbian \‘erstez et al. (1971). Rat.s were implanted 36 hours raphe neuronal firing through an indirect mech-
prior to treatment and sacrificed 1 hour after the intraventricu-
anism by first blocking 5-liT uptake: the
lar injection. Values are the titeati of 5 rats ± S.D.
b ‘[‘his dose of ‘ll-TRP (200 4() wa conlpute(l to represent resulting excess of 5-liT in the synaptic cleft
8.9 c of the amino acid. could then lead to a negative feedback inhibi-
P <.01.
I’ < .05. tion of raphe neurons (Corrodi and Fuxe,
1969) . The indirect action of CIMI on raphe
cell firing is supported by the observation that
Aimotimer possible explanation for the ability
time depression of rapime firing by CIMI can be
of LSI) to inimibit 5-liT effiux is based on the
blocked by pretreatment with the 5-liT syn-
potetmt presynaptic effect. of time hallucinogen
thesis inhibitor, p-chloropimenylalanine (Sheard
(Ilaigler and Aghajanian, 1974). Presynaptic
et at., 1972). In terms of the negative feedback
receptor timeory suggests that, in some systems,
model, the reduction in impulse flow-dependent
timere are receptors on the presynaptic mem-
5-liT release may entirely compensate for time
brane wimicim are capable of modulating trans-
block in 5-liT uptake by low doses of CIMI.
mitter release (Kirpekar and Puig, 1971;
Starke, 1971; Eimero et at., 1972). Thus LSD Presynaptic receptor theory imas been used to
could inimibit transmitter release by activating explain wimy potent inlmibitors of norepinephrine
resyhmtpt ic receptors located on serotonergic (NE) neuronal uptake are not very effective in
ternmiimals. in vitro studies by Chase et at. increasing “transmitter overflow” or effiux
(196fl, Katz and Kopin (1969) and Farnebo (Enero et at., 1972). When neuronal uptake is
atmd Flamberger (1971) Imave simown that LSD impaired, presynaptic inimibition of transmitter
causes a reduction in stimulation-induced release is enimanced because a imigimer synaptic
release of 5-liT in brain tissue slices. Since it is gap concentration of NE is available for activa-
unlikely timat iimtaet serotonergic patimways exist tion of presynaptic receptors (Langer, 1974). A
in braiim slices, suds presynaptic receptors could parallel serotonergic presynaptic receptor could
explain time effects produced by LSD iim thmese also account for time lack of effect of CIMI on
expernnents where imerve terminal areas are 5-HT effltmx as seen in our system. Since evi-
selmarated from timeir soma. Because the inhibi- dence is accumulating timat 5-liT is inhibitory
tioim of inmpulse flow by LSD and its possible on the rapime (presynaptic) system (Agimajanian

l)resimaptic effects were not separated in ouir in et at., 1972; Haigler and Aghajanian, 1974)
viva experimeimts, we cannot distinguish perimaps time CIMI effects are suggestive of a
between timese possibilities. It is interesting to presynaptic 5-liT receptor. However, effiux of
note timat altlmougim raplme cell firing was inhib- 3H-5-HT is increased, not decreased, by a high
iteci for approximately 1 imour and complete dose (20 mg/kg) of CIMI. Thus presynaptic
recovery of neuronal firing to control rates inhibition of 5-liT release cannot explain this
required aim additioimal 20 to 30 miimutes (Agha- effect of CIMI. The increase in 3li-5-liT effiux
janiaim et at., 196S), effiux of 3H-5-HT did not at a high dose of CIMI (20 mg/kg) may be
returim to control levels at aimy time during this due to a more complete blockade of 5-liT
interval. Timese data suggest that inhibition of reuptake. Carlsson (1970) has reported that a
firing may not he a complete explanation for dose of 7.5 mg/kg of CIMI (intraperitoneal
1975 CIMI AND LSD: EFFECTS ON 5-HT EFFLUX 793

administration in mice) is the ED5O for the Acknowledgments. We would like to timank
chlorimipramine-induced blockade of the mem- Dr. W. Scott of Hoffmann-La Roche, Inc.,
brane pump in 5-lIT neurons. Thus, uptake Nutley, N.J. for supplying time Ho 4-4602;
blockade of 5-liT is a dose-dependent effect of Ciba Pharmaceutical Co., Summit, N.J. for
CIMI wimicim may be correlated with the effects supplying the chlorimipramine HC1 (Anaf-
of this drug on 5-liT efflux. Increased efflux of ranil) ; and the National Insitute of Mental
3H-5-HT returns to control levels 90 minutes Health Center for Studies of Narcotics and
after time administration of 20 mg/kg of CIMI. Drug Abuse for providing time LSD used in
While time metabolic fate of CIMI has not been these experiments.
determined, a closely related compound, imi-
pramine, imas a half-life of 0.6 hour in brain References
tissue (Bickel and Weder, 1968) . In addition,
AGHAJANLAN, G. K. AND ASHER, I. : Histocimemical
liver microsomal studies indicate that CIMI fluorescence of raphe neurons : Selective enhance-
and inmipramine are metabolized at the same ment by tryptophan. Science (Washington)
172: 1159-1161, 1971.
rate (.J. Sclmeimkman and H. Nybhck, personal
AGHAJANIAN, G. K., FOOTE, W. AND SHEARD, M.:
communication). Thus, the time required for Lysergic acid diethylamide : Sensitive neuronal
3H-5-HT efflux to return to control levels units in the midbrain raphe. Science (Washing-
ton) 161: 706-708, 1968.
could be related to the rate of degradation of AGHAJANIAN, G. K., Foers, W. E. AND SHEARD, M.
CIMI. H. : Action of psychotogenic drugs on single
mmdbrnin raphe neurons. J. Pharmacol. Exp.
Time increase in 5-HT effiux caused by the Ther. 171: 178-187, 1970.
higim dose of CIMI could conceivably result AGHAJANIAN, G. K. AND GALLAGER, D. W. : Raphe
origin of serotonergic nerves terminating in the
from a direct release of 5-liT. However, data cerebral ventricles. Brain Res., in prcss, 1975.
of Ilalaris et at. (1973) show that the endogen- AGHAJANIAN, G. K., HAIGLER, ti. AND BLOOM, F.:
ous levels of 5-liT are increased and levels of Lysergic acid diethylamide and serotonin:
Direct actions on serotonin-conta at ing neurons.
5-HIAA are decreased by this treatment. Since Life Sci. Part I Physiol. Ph:rtimcol. 11:
release of 5-HT is associated with a lowering of 615-622, 1972.
AGHAJANIAN, G. K., KUHAR, M. J. AN!) 1O1H, R.
5-FIT as well as an increase in 5-HIAA (Sheard
H.: Serotonin-containing neuronal i rik tirya
and Aghajanian, 1968; Gumulka et cii., 1969; and terminals: Differential effects of p-chloro-
Kostowski et at., 1969), these data are not con- phenylalanine. Brain Res. 53: 85-101, 1973.
ANDEN, N.-E., CORRODI, H., FUXE, k. AN’J I:OK-
sistent with a direct release of 5-HT by CIMI. FELT, T.: Evidence for a central 5-imydroxyti Vp-
tamine receptor stimulation by lysergic’ cid
hm conclusion, these data suggest that LSD
diethylamide. Brit. J. Pharmacol. 34: 1-7, 168.
and CIMI, at doses which cause a similar ANDN, N.-E., FTJXE, K. AND H#{246}KFELT. T.: ‘l’he
degree of inhibition of raphe cell firing, affect importance of nervous impulse flow for the
depletion of the monoamines from central neu-
5-liT effiux differently and by different mecha- rons by some drugs. J. Pharm. Pimarmacol. 18:
nisms. It appears that LSD, by directly inimibit- 630-632, 1966.

ing impulse flow produces a reduction in 3H-5- ASHCROFr, G. W., ECCLESTON, D. AND CRAWFORD,
T. B.: 5-Hydroxyindole metabolism in rat
HT efflux. However, a possible local brain. A study of intermediate metabolism
presyimaptic inhibition of 5-liT release by LSD using the technique of tryptophan loading-i
must also be considered. In contrast, a low dose Methods. J. Neurochem. 12: 483-492, 1965.
AURES, D., FLEMING, R. AND HAKANSON. R.: Sep-
of CIMI does not cause a decrease in 3H-5-HT aration and detection of biogenic amines by
effiux despite an inhibition of raphe firing. thin-layer chromatography. J. Chromatogr. 33:
480-493, 1968.
Although a block in 5-liT reuptake by low
BfnARD, P., CARLSSON, A., FUXE, K. AND LINDQV-
doses of CIMI might be expected to increase :sr, M.: Origin of 5-hydroxytryptophan and L-
effiux, timis is apparently counterbalanced by dopa accumulating in brain following decarbox-
ylase inhibition. Naunyn-Schniedebergs Arch.
the reduction in impulse flow-related 5-HT
Pharmakol. 269: 1-6, 1971.
release, resulting in no net change in 3H-5-liT BICKEL, M. H. AND WEDER, H. J.: Time total fate
effiux. Timis is consistent with the view that the of a drug: Kinetics of distribution, excretion
and formation of 14 metabolites in rats treated
primary action of CIMI is to block 5-HT reup- with imipramine. Arch. Int. Pharmacodyn. Timer.
take and that the inimibition of firing is a com- 173: 433-463, 1968.
pensatory response to this diminished neuronal BRUINVELS, J.: Inhibition of the biosynthesis of
5-hydroxytryptamine in rat brain by imipram-
reuptake. inc. Eur. J. Pharmacol. 20: 231-237, 1972.
794 GALLAGER AND AGHAJANIAN Vol. 193

C.LssoN. A. : Effect of drugs on amine uptake of 5HT in rat brain slices. Brain Res. 69:
mecimanisms in the brain. In Bayer-Symposium 265-280, 1974.
:11, p. 223-233, Springer-Verlag, Heidelberg, HOLMAN, R. AND VOGT, M. : Release of 5-hydroxy-
1970. tryptarn.ine (SilT) from septum and caudate
CHASE, T., BREESE, G. AND KOPIN, I. : Serotonin nucleus. J. Physiol. (London) 210: 163P, 1970.
release from brain slices by electrical stimula- HOLMAN, R. AND VOGT, M. : Release of 5-hydroxy-
tion : Regional differences and effect of LSD. tryptamine from caudate nucleus and septum.
Science (Waslmington) 157: 1461-1463, 1967. J. Physiol. (London) 223 : 243-254, 1972.
CHASE, T. AND K0PIN, I. : Stimulus-induced KATZ, R. I. AND KOPIN, I. J. : Effect of d-LSD
release of substances from olfactory bulb using and related compounds on release of
time pusim-pull cannula. Nature (London) 217: norepinephrine-H3 and serotonin-H3 evoked
466-467, 1968. from brain slices by electrical stimulation. Phar-
CHIUEH, C. C. AND MOORE, K. E. : Release of macol. Res. Commun. 1 : 54-62, 1969.
endogenously synthesized catechols from the KIRPEKAR, S. M. AND PUIG, M. : Effect of flow-stop
caudate nucleus by stimulation of the nigros- on noradrenaline release from normal spleens
triatal pathway and by administration of d-am- and spleens treated with cocaine, phentolanmine
pimetanmine. Brain Res. 50: 221-225, 1973. or phenoxybenzamine. Brit. J. Pharmacol. 43:
CORRODI, H. AND Fuxz, K. : Decreased turnover in 359-369, 1971.
central 5HT nerve terminals induced by antide- K#{246}NIG, J. AND KLIPPEL, R. : The Rat Brain : A
pressant drugs of the imipramine type. Eur. J. Stereotaxic Atlas, R. E. Krieger Publishing Co.,
Pharmacol. 7: 56-59, 1969. Huntington, N.Y., 1970.
i)EBALBI.AN VERSTER, F., ROBINSON, C. A., HENGEV- Kosi’owsKl, W., GIACALONE, E., GARATrINI, S. AND
ELD. C. A. AND SANDERS-BUSH, E. : Freelmand VALZELLI, L. : Electrical stimulation of midbrain
(erebroventricular mi ection technique for itnan- raphe . Biochemical, behavioral and bioelectrical
esthetized rats. Life Sci. Part I Plmysiol. Pimar- effects. Eur. J. Pharmacol. 7: 170-175, 1969.
nmacol. 10: 1395-1402, 1971. KUHAR, M., AGHAJANIAN, G. AND ROTH, R. H.:
ENERO, M. A., LANGER, S. Z., ROTHLIN, 11. P. AND Tryptophan hydroxylase activity and synaptoso-
STEFANO, F. J. E. : Role of time alpha-adrenocep- mal uptake of serotonin in discrete brain
tor in regulating nonadrenaline overflow by regions after midbrain raphe lesions : Correla-
imerve stimulation. Brit. J. Pimarmacol. 44: tions with serotonin levels and histochemical
672-688, 1972. fluorescence. Brain Res. 44: 165-176, 1972.
FARNEBO, L.-O. and HAMBERGER, B.: Drug-induced KTJHAR, M. J., ROTH, R. H. AND AGHAJANIAN, G.
changes in time release of H-monoamines from K.: Selective reduction of tryptophan imydroxy-
field stimulated rat brain slices. Acta Plmysiol. lase activity in rat forebrain after midbrain
Scand. 371: 35-44, 1971. raphe lesions. Brain Res. 35: 167-176, 1971.
FELDBERG, W. AND MYERS, R. D.: Appearance of LANGER, S. Z.: Presynaptic regulation of catecho-
5-imydroxytryptamine and an unidentified phar- lamine release. Biochem. Pharmacol. 23:
macologically active lipmd acid in effluent from 1793-1800, 1974.
perfused ventricles. J. Plmysiol. (London) 184: LIN, R., NGAI, S. AND COSTA, E.: Lysergic acid
837-855, 1966. diethylamide: Role in conversion of plasma
FERNSTROM, J. AND WURTMAN, R.: Brain sero- tryptophan to brain serotonin (5-hydroxytryp-
tonin content: Physiological dependence on tamine). Science (Washington) 166: 237-239,
plasma tryptophan levels. Science (Washington) 1969.
173: 149-152, 1971. MAICKEL, R. P. AND MILLER, F. P.: Fluorescent
FREEDMAN, D.. GOTTLIEB, R. AND LOVELL, R.: Psy- products formed by reaction of indole deriva-
chotomimetic drugs aimd brain 5-hydroxytryp- tives and o-phthaldialdehyde. Anal. Chem. 38:
tanmine metabolism. Biochem. Pimarmacol. 19: 1937-1938, 1966.
1181-1188, 1970. MERLIS, J. K.: Time effect of cimanges in the cal-
G.ADIWM, J. H.: Pusim-pull cannulae. J. Plmysioi. cium content of the cerebrospinal fluid on
(London) 155: 1-20, 1961. spinal reflex activity in time dog. Amer. J. Phy-
GUMULK.A, W., SAMANIN, R., GARATTINI, S. AND siol. 131: 67-72, 1940.
VALZELLI, L.: Effect of stimulation of midbrain MOIR, A. AND ECCLESTON, D.: The effects of pre-
raphe on serotonin (SilT) level and turnover in cursor loading in the cerebral metabolism of 5-
differeimt areas of rat brain. Eur. J. Pharmacol. hydroxyindoles. J. Neurochem. 15: 1093-1108,
8: 380-384. 1969. 1968.
HAICLER. H. J. AND AGHAJANIAN, G. K.: Mescaline P0R’rIG, P. AND VOGT, M.: Release into the cere-
and LSD: Direct and indirect effects on sero- bral ventricles of substances witim possible trans-
tonin-contaiing imeurons in brain. Eur. J. Phar- mitter function in the caudate nucleus. J. Phy-
macoi. 21: 53-60. 1973. siol. (London) 204: 687-715, 1969.
HAIGLER, H. ANt) AGHAJANIAN, 0. K.: Lysergic RANDI, M. AND PADJEN, A.: Effect of
acid dietimylanmide and serotonin: A comparison N,N-dimethyltryptamine and d-lysergic acid
of effects on serotonergic neurons and neurons diethylamide on the release of 5-hydroxyindoles
receiving a serotonergic input. J. Pimarmacol. in rat forebrain. Nature (London) 230:
Exp. Timer. 188: 688-699, 1974. 532-533, 1971.
HALARI5, A.. LOVELL, R. AND FREEDMAN. D.: Effect RICHARDS, J., LOREZ, H. AND TRANZER, J.: Indo-
of cimloriimmipramiime on the Inetabolism of lealkylamine nerve terminals in cerebral ventri-
5-imydroxytryptamine in time rat brain. Biochem. cles: Identification by electron microscopy and
Plmarimmacol. 22: 2200-2202, 1973. fluorescence histochemistry. Brain Res. 57:
HAMON. M.. B0UIIG0IN, S., JAGGER, J. AND GL0wIN- 277-288. 1973.
SKI. J.: Effects of LSD on syntimesis and release ROSECRANS, J., LOVELL, R. AND FREEDMAN, D.:
1975 CIMI AND LSD: EFFECTS ON 5-HT EFFLIJX 795

Effects of lysergic acid diethylamide on the SHIELDS, P. J. AND ECCLESTON, D.: Evidence for
metabolism of brain 5-hydroxytryptamine. the synthesis and storage of 5-hydroxytryp-
Biochem. Plmarmacol. 16: 2011-2021, 1967. tamine in two separate pools in the brain. J.
SCHUBERT, J.: Labelled 5-hydroxytryptamine and Neurochem. 20: 881-888, 1973.
5-imydroxyindoleacetic acid formed in vivo from SNODGRASS, S. R. AND IvEnsoN, L. L.: Formation
3H-tryptophan in rat brain; effect of probene- and release of 3H-tryptamine fronm H-
cid. Acta Physiol. Scand. 90: 401-408, 1974. tryptophan in rat spinal cord slices. Advan.
SCHUBERT, J., NYBACK, H. AND SEDVALL, G.: Accu- Biochem. Pharmacol. 10: 141-150, 1974.
mulation and disappearance of H-5-
STutx.i, K.: Influence of a-receptor stimulants on
hydroxytryptamine formed from 3H-tryptophan
noradrenaline release. Naturwissenschaften 58:
in mouse brain: Effect of LSD-25. Eur. J. Phar-
420, 1971.
macol. 10: 215-224, 1970a.
SCHUBERT, J., NYBACK. H. AND SEDVALL, G.: Effect SZERB, J. C.: Model experiments with Gaddum’s
push-pull cannulae. Can. J. Physiol. Pharmacol.
of antidepressant drugs on accumulation and
disappearance of monoamines formed in vivo 45: 613-620, 1967.
from labeled precursors in mouse brain. J. TAGLIAMONTE, A., TAGLIAMONTE, P., PEREZ-CRUET,
Pharm. Pharmacol. 22: 136-139, 1970b. J., STEIN, S. AND GESSA, G.: Effect of psycho-
SCHUBERT, J. AND SEDVALL, G.: Accumulation and tropic drugs on tryptophan concentration in the
disappearance of H-5-hydroxytryptamine rat brain. J. Pharmacol. Exp. Timer. 177:
formed in vivo from 3H-tryptophan in various 475-480, 1971.
regions of the rat brain. Eur. J. Pharmacol. 17: THOMAS, R. C. AND WILSON, V. J.: Precise locali-
75-80, 1972. zation of Renshaw cells with a new marking
SHEARD, M. H. AND AGHAJANIAN, G. K.: Stimula- tecimnique. Nature (London) 206: 211-213,
tion of time midbrain raphe: Effect on serotonin 1965.
metabolism. J. Pimarmacol. Exp. Ther. 163: TILSON, H. A. AND SPARBER, S. S.: Studies on the
425-430, 1968. concurrent behavioral and neurochemical effects
SHEARD, M., ZOLOVICK, A. AND AGHAJANLAN, G. K.: of psychoactive drugs using the push-pull can-
Raphe neurons: Effect of tricycic antidepres- nula. J. Pharmacol. Exp. Ther. 181: 387-398,
sant drugs. Brain Res. 43: 690-694, 1972. 1972.