Psychopharmacology (1992) 106:321-329

Psychopharmacology © Springer-Verlag 1992

Fluoxetine decreases brain temperature and REM sleep in Syrian hamsters Bo Gao, Wallace C. Duncan Jr, and Thomas A. Wehr

Clinical PsychobiologyBranch, National Institute of Mental Health, Bethesda, MD 20892, USA Received December 12, 1990 / Final version July 23, 1991 Abstract. The antidepressant drug, fluoxetine (FLX), a selective serotonin reuptake inhibitor, was administered to Syrian hamsters, and its acute and chronic effects on EEG sleep and hypothalamic temperature were recorded. Acute fluoxetine treatment at doses of 5, 10, 20 and 40 mg/kg decreased R E M sleep and hypothalamic temperature in a dose-dependent manner. It increased N R E M sleep, and, at doses of 20 and 40 mg/kg, it increased wakefulness. At 40 mg/kg, it decreased motor activity. During chronic treatment, tolerance developed to FLX's REM sleep-inhibiting effects, but tolerance did not develop to FLX's hypothatamic temperaturedecreasing effects. Chronic FLX treatment produced circadian phase-dependent decreases in temperature beyond those that were observed during acute treatment. The apparent dissociation during chronic treatment between FLX's temperature-lowering effects and its REMdecreasing effects might be related to long-term changes in 5HT receptor function or FLX pharmacokinetics. Key words: Fluoxetine .... REM sleep - Hypothalamic temperature - Syrian hamster - Serotonin

Most antidepressant drugs alter sleep in both human beings and animals. One of the drugs' most consistent effects is suppression of REM sleep. Based on the observation that antidepressants suppress REM sleep and that non-pharmacological suppression of REM sleep improves depression, Vogel (1975, 1983) proposed that the therapeutic mechanism of antidepressant drugs depends on their capacity to suppress R E M sleep. Since many antidepressant drugs alter thermoregulation, it is possible that the REM sleep suppression produced by these drugs is causally related to their effects on thermoregulation. A number of investigators have shown that there are mutual interactions between vigilance states and thermoregulation (Parmeggiani 1987). Offprint requests to." W.C. Duncan

Jr

For example, brain temperature tends to increase during REM sleep and decrease during N R E M sleep (Parmeggiani 1980; Obal et al. 1985). Conversely, REM and N R E M sleep tend to decrease when brain temperature is lowered (Sakaguchi et al. 1979). We have recently reported that chronic treatment with the antidepressant drug clorgyline (CLG), a type A monoamine oxidase inhibitor (MAOI), suppresses REM sleep and decreases peritoneal temperature (Gao et al. 1991) in Syrian hamsters, and that during 3 weeks of treatment, tolerance develops to both effects. Since the time courses for development of tolerance to CLG-induced changes in REM sleep and body temperature are similar (Gao et al. 1991), we suggested that these changes might be causally related. However, some of our data suggested that the changes in REM sleep and body temperature were also partly independent. Serotonergic properties of antidepressant drugs seem to play an important role in mediating both the REM inhibiting and thermoregutatory effects of these compounds. Both the MAOIs which elevate 5HT levels, and 5HT reuptake inhibitors, decrease REM sleep (Reyes et al. 1983; Shipley et al. 1984; Sommerfelt et al. 1987; Nicholson and Pascoe 1988; Ross et al. 1990) and alter body temperature in a variety of species (Clark and Lipton 1986). The role of the serotonergic system in mediating CLG's temperature-decreasing effect in hamsters is uncertain. In one study, 5HT microinjections failed to alter body temperature in hamsters (Reigle et al. 1974). More recent studies conducted with rats indicate that selective activation of 5HT receptors may produce either hypothermic or hypertherrnic temperature responses (Pawlowski 1984; Gudelsky et al. 1986). CLG's temperature decreasing properties in hamsters may similarly be related to selective effects on 5HT receptor subsystems. In order to further examine the effects of antidepressants on thermoregulation and sleep, the interactions between these effects, and the possible role of the serotonergic system in these effects, we measured both acute doseresponse effects and chronic effects of the antidepressant

322

drug fluoxetine (FLX), a selective 5-HT reuptake inhibitor (Wong et al. 1974, 1975; Fuller et al. 1974; Hwang et al. 1980), on hamster EEG sleep and hypothalamic temperature (Thy). Materials and methods Animals and procedures. Male golden hamsters (Mesocrieetus auratus, LGV: lak, Charles River), 12-16 weeks old and weighing approximately 140 g, were used in all studies. Prior to surgery, animals were group-housed in LD 14:10 with access to food and water ad lib±turn. Chronic EEG and E M G electrodes were implanted according to previously described procedures (Gao et al. 1991). Transmitters (Model X M - F H , Mini-Mitter Inc.) for recording hypothalamic temperature (4- 0.05 ° C) and activity (ACT) (Kluger et al. 1990) were implanted at the time of EEG and E M G electrode implantation and were fixed to the skull with the EEG and E M G electrodes by dental cement. Temperature-sensitive probes were positioned with their tips (~460 g diameter) in the hypothalamus at the following stereotaxic coordinates (with bregma and lambda in the same horizontal plane): 0.3mm anterior to bregma, 0.1-0.3 mm lateral to midtine and 8.0 mm below the skull. Surgical procedures for electrodes and transmitter implants were conducted under sodium pentobarbitat anaesthesia (90 mg/kg, IP). After surgery, hamsters were individually housed in transparent polycarbonate jars (diameter x height = 30.5 x 30.5 era) with free access to food and water. The jars were located in a ventilated and sound-attenuated chamber (Napco Model 3810). Animals were maintained under LD 14:10 (L:01:0(~15:00, D: 15:00M?1:00) with an ambient temperature of 22 ± 1° C. Light intensity within the chamber was 10-20 g W c m -2. Hamsters were allowed at least 7 days to recover from surgery and were connected to flexible recording cables for 1 day before EEG sleep recordings were obtained. EEG recordings started at 10:00 and were carried out continuously for 23 h on a Grass model 78B polygraph. When the recording ended, hamsters were disconnected from the recording cables until the next recording was done. In the acute experiment, animals were randomly assigned to one of five experimental conditions. They received either FLX (Eli Lilly

Table 1. Baseline hypothalamic temperature (Thy), REM sleep (REM), N R E M sleep (NREM), wakefulness (W) and motor activity (ACT) in hamsters treated with vehicle (VHC) or 5-40 mg/kg fluoxetine"

and Company) at doses of 5, 10, 20 or 40 mg/kg, or a control volume of vehicle (VHC). There were four hamsters in each group. FLX was dissolved in sterile distilled water at a concentration of 5 mg/ml (for 5 and 10 mg/kg) or 10 mg/ml (for 20 and 40 mg/kg). After a 2-day baseline recording, hamsters were injected intraperitoneally at 09:30 a.m. with either freshly prepared FLX or a control volume of VHC (sterile distilled water, 0.14-0.56 ml). In the chronic experiment, animals were injected once daily at 09:30 a.m. with either VHC or 10 mg/kg FLX. This dose was selected since it was the minimum dose that induced significant reductions of both R E M sleep and hypothalamic temperature during the acute dose-response experiment. EEG sleep and Thy data were recorded and analyzed on days 7 and 14 of the chronic treatment. At the end of experiments, hamsters were decapitated and their brains were frozen rapidly in dry ice. Brains were cut into 60 lam sections and examined for placements of transmitter probes. All were later found to be located within the hypothalamus. Data analysis'. Vigilance states were visually scored for wakefulness (W), N R E M sleep and R E M sleep using a 30 s epoch duration, according to criteria previously described (Gao et al. 1991). The percent of each stage was determined for the 5 h of the light phase following the injection (LI:10:00-15:00), the 10h dark phase (D: t5:00-1:00), and the first 8 h of the following light phase (L2:1:00-9:00). The differences between the experimental and baseline days were calculated and expressed as percents of the baseline values. Thy and ACT (counts) were recorded telemetrically every 10 min, and data were stored on a laboratory computer. The differences between the experimental day and baseline day for L1, D and L2 were calculated, and for ACT the difference of ACT was expressed as a percent of the baseline value. For the acute experiment, one-way analysis of variance (ANOVA) was used to analyze the dose effects of FLX and VHC treatments on vigilance states, Thy and ACT. Post hoc unpaired t-tests were used to assess differences between VHC and FLX treatments. For the chronic experiment, two-way ANOVA with repeated measures (treatment x day of treatment) was used to analyze chronic effects of FLX (10 mg/kg) and VHC on these variables during L1, D and L2. Three-way A N O V A with repeated measures was used to analyze the circadian pattern of FLX's efects during chronic treat-

L1 VHC

Thy(°C) REM b NREM b Wb

ACT ~

37.33± 20.81+ 43.76± 35.42± 52.83±

5 mg 0.19 1.5 2.2 3.2 9.5

37.23420.93445.54± 33.66446.154-

10 mg 0.27 1.4 1.1 2.0 14.2

37.54± 18.281 38.76+ 42.97± 65.55±

20 mg 0.14 1.5 4.2 5.5 14.4

37.32± 21.98± 44.70± 33.33± 46.t0±

40 mg 0.21 2.2 2.4 3.7 14.2

37.33± 19.09± 36.71± 44.23± 97.68±

0.08 0.5 3.6 3.2 173

D Thy (°C) REM b NREM b Wb ACT ~

37.33± 0.16 37.74± 5.80± 1.0 7.37± 23.62± 2.0 26.67± 70.58± 3.0 65.91± 192.27±34.5 160.60±

0.20 37.76± 0.12 37.69± 0.12 37.76± 0.14 1.4 7.474- 1.0 6.15± 0.5 6.40± 1.2 26.20± 1.8 22.30± 2.4 2.6 28.43± 3.4 56.53± 7.4 71.34± 3.0 3.9 61.70± 6.4 18.9 144.24±23.3 151.90±30.8 209.78±38.1

L2 Thy (°C) REM b NREM b Wb ACT ~

36.98± 19.90± 47.03i 33.08± 59.60±

0.09 1.5 2.4 3.8 13.5

36.94:t: 19.47± 50.40± 29.93± 39.10±

0.39 1.3 1.9 1.3 13.1

37.23± 19.07± 48.47± 32.21+ 42.76±

0.07 1.6 2.2 3.5 3.8

Results are reported as mean ± SEM b Data are represented as the percentage of recording time ° Data are represented as counts/10 min

37.10± 19.40± 47.05:t: 32.10± 39.90±

0.22 1.1 1.4 2.7 11.3

36.98± 0.10 20.47± 0.5 45.41+ 2.4 34.15± 2.0 73.84±23.9

323 ment. Independent group t-tests were used to assess differences between FLX and VHC. Paired group t-tests were used to assess within group differences.

Results

Effects of acute fluoxetine treatment on Th~ and EEG sleep Table 1 summarizes the baseline data prior to drug administration. Figures 1 and 2 summarize the dose response relative to baseline, of F L X and V H C on hamster Tny, R E M sleep, N R E M sleep, W and ACT. As can be seen in these figures, FLX's effects on T~y, R E M sleep and N R E M were large during the 5 h (L1) following acute injection, and they appeared to increase with increasing doses. V H C injection (Figs. 1 and 2), and the volume of the V H C had no effect on the level of T~y or vigilance state. A N O V A indicated that F L X dose had a statistically significant effect on T~y during L1 ( F = 11.46, d f = 4 , t 5 , P = 0.0002) and D ( F = 8.6614, df= 4,15, P = 0.0008), but not during L2 (F = 2.48, df= 4,15, P = 0.089). Compared with VHC, 5 mg/kg F L X did not significantly decrease T~y, 10 mg/kg F L X decreased T~y during L1, and 20 mg and 40mg/kg F L X decreased Thy during L1 and D (Fig. 2). F L X had a statistically significant effect on R E M sleep during L1 (F = 81.46, df = 4,15, P = 0.0001) and L2 (F=5.709, df=4,15, P=0.0054), but not during D L1

VHC

D

L2

Effects of chronic fluoxetine treatment on Thy and EEG sleep Figure 4 summarizes the effects of F L X or V H C during L1, D, and L2 on days 1, 7 and 14 of treatment, respec-

L1

D

L2

L1

39

30 20.

60

10,

40 20 0

36

0.

39

30.

8O

38

2O,

6O 40 20 0

37

10,

36

m

0

E

FLX l O m g / k g

°o

FLX 2 0 m g / k g

39,

~ c

30

38,

~o

20

37,

o

10

36

'~

0

39

30

38

20

D

L2

8O

38 37

FLX 5 m g / k g

( F = 1.882, df=4,15, P=0.1659). Compared with VHC, all doses of F L X significantly decreased R E M sleep during L1, and 40 mg/kg F L X significantly decreased R E M sleep during L2 (Fig. 2). F L X significantly altered N R E M sleep during L1 (F=3.777, df=4,15, P=0.0256) and D (F=5.91, df=4,15, P=0.0046), but not during L2 (F=2.478, df= 4,15, P = 0.0887). Compared with VHC-treated animals, 10 mg/kg F L X significantly increased N R E M sleep during L1, and 40 mg/kg F L X significantly increased N R E M sleep during D (Fig. 2). F L X significantly altered W during" L1 (F=9.593, df=4,15, P=0.0005). Compared with VHC, 20 and 40 mg/kg F L X significantly increased W during L 1. F L X altered ACT during D (F--5.988, df= 4,15, P--0.004). Compared with VHC, 40 mg/kg F L X significantly decreased ACT during D (Fig. 2). Figure 3 shows the relationship between Thy and R E M sleep during treatment with different doses of FLX during L1. R E M sleep appears to decrease with decreases in Thy that are less than 1° C. Decreases in Tuy greater than 1° C are associated with 100% suppression of R E M sleep.

t~

80

o~

2£

c ~[ 6c oo 4c

37 36

10

35

0

38 FLX 4 0 m g / k g

~;~z~;~i ~i~ii!!ii!!!iiiiiiiii~iiiiiii!i!i!;;!ii!!!iii;

3o so

36

10

34 a

10:00

0 16:00

22:00

4:00

b

10:00

Fig. la-e, This figure shows the daily profiles of hypothalamic temperature (a), REM sleep (b) and NREM sleep (c) following a single vehicle or FLX injection (5--40mg/kg IP). REM sleep and

t6:00

22:00

4:00

C

I0:00

t6:00

22:00

4:00

NREM sleep are plotted as the percentage of the difference from baseline (difference/baseline x 100). Baseline values are shown in Table 1. ~ .... ) FLX; ( ) baseline

324 -2O 2

~3

-40

0 o

"5 v

0

-60 x

u.J

<1

-80 -100 -3.0

-50

**

-2,5

-2.0

•

-1.5

~1.o

-0,5

0

0.5

Fig. 3. FLX-induced decreases in REM sleep are plotted as a

**

function of the FLX-induced decreases in Thy during L1 for individual animals. Symbols indicate different doses. The verticle dotted line represents baseline hypothalamic temperature• Delta scores are calculated as in Fig. 2. (©) 5 mg/kg; (×) 10 mg/kg; (A) 20 mg/kg; (e) 40 mg/kg

1

~ -10 J 100 50 0

~

< -50

~

•

•

AThy (°C)

-100

CD

I •

15o7 100 1

-50 t

-100

VHC

5

10

20

40

Fluoxetine (mg/kg) L1 10:00 15:00

D

L2 1:00 9:00

Fig. 2. Acute effects of fluoxetine (5, 10, 20 and 40 mg/kg) and vehicle (VHC) on hypothalamic temperature (Thy), REM sleep, NREM sleep, wakefulness (W) and Mini-Mitter activity (ACT) after a single injection (n = 4 for each group). Thyfor L1, D and L2 are plotted as the differences from baseline. REM sleep, NREM sleep, W and ACT for L1, D and L2 are plotted as the percentage of the differenceform baseline (difference/baselinex 100). * P < 0.05 compared with VHC-treated animals. ** P< 0.01 compared with VHC-treated animals. The legend indicates the lighting conditions and hours corresponding with Lt, D and L2

tively. During L1, there was a significant F L X treatment effect on Thy (F = 14.01, d f = 1,6, P = 0.0096), and N R E M sleep ( F = 16.21, df= 1,6, P=0.0069). During these first 5 h after the daily F L X injection, Thy decreased relative to its baseline (Figs. 4 and 5), and N R E M sleep increased (Figs. 4 and 7). VHC-injected hamsters exhibited little change from baseline. During L1, there was a significant interaction between treatment and day of treatment on R E M sleep (F = 5.07, df=2,12, P=0.025) and ACT (F=4.16, dr=2,12, P = 0.043). During L1, the greatest reduction in levels of

R E M sleep were recorded on the first day of treatment (D1), followed by a return towards baseline during D7 and D14 (Figs. 4 and 6). Compared with VHC, FLX significantly decreased R E M sleep on day 1 (P=0.01) and day 7 (P=0.018), but not on day t4 (P=0.127). Compared with D1 of F L X treatment, chronic F L X treatment produced a significant increase in R E M sleep on D7 (P<0.02, t = - 5.03, d f = 3) and on D14 ( P < 0.05, t = - 3.25, df= 3). Activity levels increased during the first day of F L X treatment, but decreased during D7 and D 14 (Fig. 4). During D, there was a significant effect of day of treatment on R E M sleep ( F = 4.09, d f = 2,12, P = 0.0043), N R E M sleep (F=4.65, df=2,12, P=0.032), and W (F = 5.90, d f = 2,12, P=0.016). During L2, there was a significant effect of F L X treatment on Thy ( F = 9.44, df= 1,6, P=0.0219) (Fig. 4). Figures 5-7 show the daily profiles of Thy, R E M and N R E M sleep, during baseline and D1, D7, and D14 of FLX treatment. In order to closely examine the chronic effects of FLX over the course of a day, hourly values of Thy and vigilance states were analyzed using a three factor ANOVA (TRT, HOUR, DAY) with repeated measures on two factors (HOUR and DAY). The results are summarized in Table 2. A highly significant circadian variation (P<0.0001) was present throughout chronic F L X treatment for Thy (F=12.90; df=22,138), R E M sleep (F =27.37; d f =22,138), N R E M sleep (F =20.65; df=22,138), W (F=27.37; df=22,138) and ACT (F = t6.07; d f = 22,138). FLX's temperature-lowering effects were greatest between 1000 and 1200 hours and during D1, D7 and D14 (Fig. 5). The effects observed on D1 between 1000 and 1200 hours continued to be present on D7 and D14. In contrast, FLX's temperature-towering effects between 0100 and 0900 hours were enhanced on D7 and D14 compared with D1. During darkness, both temperatureelevating and temperature-lowering effects of FLX were observed. Two temperature peaks were present in the dark phase, one at lights off (1500) and another one about 5 h later. During F L X treatment, temperature between these two peaks decreased relative to baseline. ANOVA indicated a significant three factor interaction

325 1

DI

D7

D14

D1

D7

L1

D14

D

L2

39.0 38.5 38.0 FLX-treated Animals n=4

>-

37.5 37.0

-so]

36.5 39.0

-lOO

1

**

1" 38.5

"

100]

"5

501

O" 38.0 VHC-treated %. Animals n = 4 ~- 37.5 37.0

*50 -100

36,5 10:00

,oo 1

18:00

22:00

2:00

6:00

is a significant interaction between treatment, day of treatment, and hour, on the level of Thy. Compared with VHC, FLX significantly (P<0.05) decreased Thy at t100 on D1, at 1000, 1900, and 0500 on D7, and at 1000, 0300 and 0400 on D14. Compared with D1 of FLX treatment, chronic FLX treatment decreased Thy at 0200 and 0500 during D7, and at 0200 during D14. ( - - - - ) Baseline; (..... ) D1 ; (--)D7;( )914

-lO0

15o]

o=,001 (.3 <<~

14:00

Fig, 5. Daily profiles of hourly hypothalamic temperature means (Thy) during baseline (BL) and day 1 (D1), day 7 (97), and day 14 (D14) of 10 mg/kg FLX (n = 4) and vehicle (n = 4) treatments. There

q

¢,¢.

-50 ] -100 VHC

FLX 10 mg/kg L1

D

L2

10:00 15:00 1:00 9:00

Fig. 4. Chronic effects of 10mg/kg fluoxetine (FLJO ( n = 4 ) and VHC (n=4) on hypothalamic temperature (Thy), REM sleep, NREM sleep, W and ACT during baseline, and during day 1 (D1), day 7 (D7), and day 14 (D14). Data are calculated as described in Fig. 2. During L1, there was a significant FLX treatment effect on Thy (P= 0.0096), and NREM sleep (P = 0.0069), and a significant interaction between treatment and day of treatment on REM sleep (P = 0.025) and ACT (P = 0.043). During D, there was a significant effect of day of treatment on REM sleep (P = 0.0043), NREM sleep (P=0.032), and W (P=0.016). During L2, there was a significant effect of FLX treatment on Thy (P= 0.0219). * P< 0.05 compared with VHC-treated animals. **P<0.01 compared with VHCtreated animals. 1"P<0.05 compared with D1 FLX between treatment (TRT), day o f treatment (DAY), and hour ( H O U R ) , on the level of Thy (F = 11.57; df = 66,396; P = 0.0223; Table 2). Analysis o f hourly celt means indicated that F L X , c o m p a r e d with V H C , significantly ( P < 0.05) decreased Thy at 1100 ( t = 2.884, df= 6) on D1,

at 1000 (t = 3.37, df= 6), 1900 (t = 3.319, df= 6) and 0500 (t = 3.372, df= 6) on D7, and at 1000 (t = 2.541, df= 6), 0300 (t = 2.908, dr= 6) and 0400 (t = 2.551, df = 6) on D14 (Fig. 5). C o m p a r e d with D 1 o f F L X treatment, chronic F L X treatment decreased Thy at 0200 ( P < 0.029, t = 3.98, d f = 3 ) and 0500 ( P < 0 . 0 1 8 , t=4.81, d f = 3 ) during D7, and at 0200 ( P < 0 . 0 2 6 , t = 4 . 1 3 , d f = 3 ) during D14 (Fig. 5). F L X ' s R E M sleep-decreasing properties were greatest between 1000 and 1500 hours. M a x i m u m R E M sleep suppression was observed within 1-2 h post-injection during D1. C o m p a r e d with D1, the level o f R E M sleep gradually increased between 1000 and 1500 hours during D7 and D14. C o m p a r e d with the baseline, F L X did not suppress R E M sleep between 1500 and 0900 hours. A N O V A o f R E M sleep indicated significant two factor interactions for treatment (TRT) and day o f treatment (DAY) ( F = 4 . 4 8 ; d f = 3 , 1 8 ; P=0.0162), and for treatment (TRT) and hour ( H O U R ) ( F = 2 . 9 2 ; df=23,138; P = 0.0001) (Table 2 and Fig. 6). F L X produced a small, non-significant increase in N R E M sleep between 1000 and 1500 hours during D1 and D7, but not during D14 (Fig. 7). Between 1500 and 0900 hours F L X did not consistently alter N R E M sleep. Chronic F L X treatment did not significantly affect W or A C T (Table 2 and Fig. 7). There were no significant main

326 L1

L2

8o

30

L2

L1

%E

%" E

60 20

~S FkX-treated Animalsn=4

o FLX-treated Animals n = 4 "~

o ~¢ 40

10

20 orZ

c~

80.

30 %` E

%E VHC-treated

Animals n = 4

~S

60. c

20

2 VHO-treated °o Animals n = 4 ~ 4 0 "b

o

~6 10

20 G2 7

cr

0

0

,

10:00

14:00

18:00

22:00

2:00

10:00

6:00

14:00

18:00

22:00

2:00

6:00

Fig. 6. Daily profiles of mean hourly REM sleep during baseline

Fig. 7. Daily profiles of mean hourly NREM sleep during baseline

(BL), day 1 (Dt), day 7 (D7), and day 14 (D14) of 10 mg/kg FLX (n = 4) and vehicle (n = 4) treatments. Data are calculated as the percentage of hourly recording time. There are significant interactions between a) treatment and day of treatment, and b) treatment and hour. For symbols see legend of Fig. 5

(BL), day 1 (DI), day 7 (DT), and day 14 (D14) of 10 mg&g FLX (n= 4) and vehicle (n=4) treatments. Data are calculated as the percentage of hourly recording time. There are no significant interactions between treatment, day of treatment, or hour for NREM sleep. For symbols see legend of Fig. 5

Table 2. ANOVA summary of chronic fluoxetine effects on circadian variation ofhypothalamic temperature (Th), REM sleep, NREM Sleep, Wakefulness (W), and Activity (ACT) Th

TRT DAY HR TRT x DAY TRT x HR TRT x DAY x HR

REM Sleep

NREM Sleep

W

ACT

df

F

df

F

df

F

df

F

df

F

1,6 3,18 22,138 3,18 23,138 66,396

0.0t 1.69 12.90a 3.17" 0.37 1.57b

1,6 3,18 22,138 3,18 23,138 66,396

0.98 6.59 ° 24.34 a 4.48 b 2.92 a 1.28

1,6 3,18 22,138 3,18 23,138 66,396

0.04 2.85 20.65 't 0.96 1.18 1.28

1,6 3,18 22,138 3,18 23,138 66,396

0.01 1.28 27.37 a 0.44 1.54 1.22

1,6 1,18 22,138 3,18 23,138 66,396

0.95 0.81 16.07a 0.89 1.03 1.I 3

a P<0.05

b P < 0.025 c P<0.01 d p < 0.0001

effects o r i n t e r a c t i o n s b e t w e e n T R T , H O U R , a n d D A Y for N R E M , W o r A C T ( T a b l e 2).

Discussion

T o o u r k n o w l e d g e , this is the first s t u d y in w h i c h a c u t e a n d c h r o n i c effects o f F L X o n E E G sleep a n d b r a i n t e m p e r a t u r e were m o n i t o r e d s i m u l t a n e o u s l y . F o l l o w i n g a c u t e t r e a t m e n t , F L X 1) d e c r e a s e d R E M sleep a n d Thy in a d o s e - d e p e n d e n t m a n n e r ; 2) i n c r e a s e d N R E M sleep d u r i n g L1 at low d o s e a n d d u r i n g D at high d o s e ; 3) in-

c r e a s e d W d u r i n g L1 (20 a n d 40 m g / k g ) , a n d 4) d e c r e a s e d A C T d u r i n g D (40 m g / k g ) . D u r i n g c h r o n i c t r e a t m e n t , p a r t i a l t o l e r a n c e d e v e l o p e d to the F L X - i n d u c e d decrease in R E M sleep (Fig. 4). I n c o n t r a s t , the m a g n i t u d e o f the F L X - i n d u c e d d r o p in Thy a c t u a l l y i n c r e a s e d d u r i n g t h e h o u r s c o r r e s p o n d i n g to L2 (0300-0500; Fig. 5).

Acute effects of fluoxetine on thermoregulation and vigilance states Thermoregulation. F L X ( L i n 1978; M c C l e a r y a n d L e a n d e r 1982; K u l a k o w s k i 1984) a n d o t h e r s e r o t o n i n r e u p -

327 take inhibitors, such as CGP 6085 A [4-(5,6-dimethyl-2benzofuranyl) piperidine HCL] (Kulakowski 1984) and citalopram (Kulakowski 1984), have been reported to decrease peritoneal and rectal temperature in rats during acute studies. The current experiments with Syrian hamsters are consistent with these previous results and suggest that 5HT reuptake inhibition decreases body temperature in hamsters, although this mechanism remains to be directly tested. The FLX-induced decrease in body temperature may be mediated by 5-HT1A receptor stimulation following 5HT reuptake inhibition, since intraventricular injection of 5-HT (Feldberg and Lotti 1967; Myers and Yaksh 1968), and systemic injection of the 5-HT1A agonists 8-OH-DPAT (Gudelsky et al. 1986; Wozniak et al. 1988), buspirone (Edgar 1990) or low dose 5-MeODMT (Gudelsky et al. 1986) have been reported to decrease body temperature in rats. A selective 5-HT1A receptor antagonist could be used in conjunction with FLX to assess the role of 5-HT1A receptor activation in FLX's temperature-decreasing effect. In contrast to the foregoing results with agents specific for 5-HT1A receptors, systemic injections of the less specific 5HT receptor agonists m-CPP (Pawlowski 1984) or MK-212 (Gudelsky et al. 1986) increase body temperature in rats. The fact that 5HT can exert opposite effects on the serotonergic subsystems involved in thermoregulation might help to explain a previous report that hypothalamic 5-HT microinjections failed to alter rectal temperature in hamsters (Reigle et al. 1974). Vigilance states. The acute effects of FLX on REM sleep in hamsters are similar to those observed in humans (Bardeleben et al. 1988; Nicholson and Pascoe 1988) and rats (Slater et al. 1978; Fornal and Radulovacki 1980; Pastel and Fernstrom 1987), and are similar to the effects of other selective serotonin reuptake inhibitors in both humans and animals, such as zimelidine (Reyes et al. 1983, 1986; Shipley et al. 1984; Kleinlogel and Burki 1987; Sommerfelt et al. 1987), alaproclate (Sommerfelt et al. 1987), sertraline (Ross et al. 1990), paroxetine (Kleinlogel and Burki 1987) and indalpine (Hilaire et al. 1984). In contrast to our finding that FLX increased NREM sleep in hamsters, other investigators have not observed this effect in humans (Bardeleben et al. 1988; Nicholson and Pascoe 1988) or in rats (Fornal and Radulovacki 1980). The effects of other serotonin reuptake inhibitors on NREM sleep are also inconsistent. In summary, suppression of REM sleep appears to be the only effect on vigilance state that is common to various serotonin reuptake inhibitors and various species. Interaction between therrnoregulation and R E M sleep. The fact that REM sleep and Thy decreased in parallel during acute FLX treatments at doses ranging from 5 to 20 mg/kg (Figs. 2 and 3) suggests that these two changes might be causally related to each other. For example, the small decrease in T~y observed at 10 mg/kg might have been secondary to the concomitant decrease in REM sleep. REM sleep is usually accompanied by increased brain temperature (Parmeggiani 1987); thus, inhibition

of REM sleep might cause Thy to be lower. Conversely, the decrease in REM and NREM sleep observed at 40 mg/kg might have resulted from the large FLXinduced decrease in Thy, since non-pharmacological cooling of the hypothalamus to a similar temperature has been reported to decrease REM sleep and NREM sleep (Sakaguchi et al. 1979). Without further experiments, it is not possible to decide among these possible causal relationships. FLX may also affect each of these variables independently. This latter possibility has to be considered, particularly at the low dose (5 mg/kg), which significantly decreased REM sleep during L1 without significantly decreasing Thy. However, a type II error could explain our failure to observe a FLX-induced decrease in T~y, since a) the group sizes are small and b) FLX (5 mg/kg) did produce a small, significant decrease in Yhyduring the first hour after injection. From the current data, one cannot determine whether 5HT uptake inhibition is specifically responsible for the observed changes in hypothalamic temperature and vigilance states in hamsters. In rats, acute administration of FLX (10 mg/kg, IP), produces increased levels of 5HT, but not dopamine, within the nucleus accumbens, as measured by push-pull perfusion (Guan and McBride 1988). It also increases the 5HT signal within the hippocampus as measured by in vivo voltametry (Marsden et al. 1979). Changes produced by this dose of FLX appear to result from 5HT uptake inhibition rather than 5HT release since a) this dose decreases levels of 5HIAA (Guan and McBride 1988; Schmidt et al. 1988) and b) between 1 and 10 mg/kg, FLX antagonizes the p-chloroamphetamine-induced depletion of 5HT (Fuller et al. 1975). All of these studies suggest that at doses of about 10 mg/kg IP, FLX selectively inhibits 5HT uptake, at least in rats. At doses above 10 mg/kg, FLX's effects could be due to additional properties such as 5HT release or dopamine uptake inhibition (Garattini et al. 1989). It is necessary to do similar experiments, or experiments utilizing 5HT receptor antagonists or the serotonin precursor 5-hydroxytryptophan to potentiate FLX's effects in hamsters, to assess the likelihood that 5HT uptake inhibition mediates FLX's effects on hamster thermoregulation and vigilance state. Chronic effects of fluoxetine on thermoregulation and vigilance states This is the first study to describe the chronic effects of a selective 5-HT reuptake inhibitor on the daily patterns of vigilance state and body temperature. Our results indicate that during chronic FLX treatment, partial tolerance developed to the FLX-induced decrease in REM sleep during L1 (Figs 4 and 6). In contrast, no tolerance developed to the FLX-induced lowering of Thy during L1 (Figs 4 and 5). During some of the hours of L2, chronic FLX significantly decreased Tby compared with FLX treatment (D7: 0500; D14: 0300, 0400), or compared with D 1 of FLX treatment (D7: 2400, 0200, 0500; D14: 0200) (Fig. 5), but did not significantly effect

328 R E M sleep. During acute treatment FLX's temperaturelowering effects might be related to FLX's R E M sleep decreasing effects. However, during chronic treatment the two effects appear to be dissociated. Most studies are in agreement that acute treatment with serotonin reuptake inhibitors suppresses R E M sleep; however, the literature is not consistent regarding the development o f tolerance to this effect during chronic treatment. Ross et al. (1990) reported that in cats, tolerance to R E M sleep-suppression developed when sertraline was administered for 17 days. On the other hand, Reyes et al. (1986) found no tolerance to R E M sleep-suppression when zimelidine was administered for 2 weeks. Our results with F L X more closely resemble those of Ross et al. with sertraline (1990). Although chronic treatment with MAOIs (Gudelsky et al. 1986; Wozniak et al. 1988) and 5HT reuptake inhibitors (Wozniak et al. 1988) fails to decrease the level of body temperature in rats, these antidepressants both decrease the level of Thy in hamsters. It has not been determined whether these results reflect differences in species or methodology. The fact that FLX's temperature-lowering effect increased while its REM-sleep-suppressing effect decreased over time suggests that the mechanism(s) responsible for the drug's effects on vigilance and thermoregulation during chronic treatment are complex. The situation is different during acute treatment, when R E M sleep and Thy decreased in parallel (Fig. 3). The changing picture that emerges during chronic treatment may reflect the pharmacokinetics o f FLX. F L X has a half-life of 1-3 days in humans (Bergstrom et al. 1988), and steady-state plasma levels are not reached until 30 days of treatment (Farid et al. 1986). Alternatively, the complex picture that emerges during chronic treatment might be due to changes in receptor function that result from chronic 5HT reuptake inhibition. In rats, chronic treatment with F L X (10 mg/kg/day) produces subsensitivity of 5 - H T receptors (Wong and Bymaster 1981 ; Wong et al. 1985) and decreases the function of the terminal 5HT autoreceptor (Blier et al. 1988). In the chronic study, FLX's Thy-lowering effect was greatest during the light portion o f the L D cycle. The effect of F L X treatment (10 mg/kg) on Thy was initially confined to L1 on day 1 of treatment, then extended to L2 by day 14. We have recently reported that CLG's lowering of peritoneal (Gao et al. 1991) and hypothalamic temperature (Gao and Duncan 1989) in Syrian hamsters is similarly confined to the light phase of the lightdark cycle. It remains to be determined whether these phase-dependent effects are secondary to drug-induced changes in behavior or physiological state, to the interaction of drug effects with the light dark cycle, or to both. R E M sleep suppression (Vogel 1975, 1983) and lowering of body temperature (Wehr 1990) have both been hypothesized to be related to the antidepressant mechanism of clinical treatments of depression. In humans, F L X doses of about 1 mg/kg which are associated with clinical remission (Bergstrom et al. 1988; Montgomery 1989) decrease R E M sleep (Bardeleben et al. 1988; Nicholson and Pascoe 1988). To our knowledge, there have been no

studies of the effects of F L X on body temperature in humans. A report that chronic treatment with the 5HT reuptake blocker clomipramine attenuates the hyperthermic response to mCPP (Zohar et at. 1988) suggests that F L X might also affect body temperature in humans. Clinical experiments that combine thermal challenges with drug treatments might help to determine whether the thermoregulatory effects o f F L X contribute to its psychoactive properties.

Acknowledgements. We would like to thank the Eli Lilly Company for providing fluoxetine, and Drs. Dennis L. Murphy and William Z. Potter for their valuable comments on a previous version of the manuscript. References Bardeleben UV, Steiger A, Gerken A, Holsboer F (1988) Effects of fluoxetine upon pharmacoendocrine and sleep-EEG parameters in normal controls. Int Clin Psychopharmacol 4 [suppl 1]: 1-5 Bergstrom RF, Lemberger L, Farid NA, Wolen RL (t988) Clinical pharmacology and pharmacokinetics of fluoxetine: a review. Br J Psychiatry 153 [suppl 3]: 47-50 Blier P, Chaput Y, Montigny CD (1988) Long-term 5-HT reuptake blockade, but not monoamine oxidase inhibition, decreases the function of terminal 5-HT autoreceptors: an electrophysiological study in the rat brain. Naunyn-Schmiedeberg's Arch Pharmacol 337:236254 Clark WG, Lipton JM (1986) Changes in body temperature after administration of adrenergic and serotonergic agents and related drugs including antidepressants: II. Neurosci Biobehav Rev 10:153-220 Edgar DM, Seidet W F, Dement WC (1990) Effects of buspirone on sleep-wake and body temperature circadian rhythms in the rat. Sleep Res 19:59 Farid NA, Bergstrom RF, Lemberger L, Ziege EA, Tenbarge J, Wolen RL, Dhahir P (1986) Studies on the disposition of fluoxetine and radioactive isotopes. 15th CINP Congress, Puerto Rico, Collegium Internationale Neuropsychopharmacologium Feldberg W, Lotti VJ (1967) Temperature response to monoamines and inhibitor of MAO injected into the cerebral ventricles of rats. Br J Pharmacol Chemother 31:152-161 Fornal C, Radulovacki M (1980) Fenfluramine, fluoxetine and quipazine suppress sleep in rats. Soc Neurosci Abstr 6:51 Fuller RW, Perry KW, Molloy BB (1974) Effect of an uptake inhibitor on serotonin metabolism in rat brain: studies with 3-(p-trifluoromethylphenoxy)-n-methyl-3-phenylpropylamine (Lilly 110140). Life Sci 15: 1161-1171 Fuller RW, Perry KW, Molloy BB (1975) Effect of 3-(p-trifluoromethylphenoxy)-n-methyl-3-phenylpropylamine on the depletion of brain serotonin by 4-chloroamphetamine. J Pharmacol Exp Ther 193 : 796-803 Gao B, Duncan WC (1989) Antidepressant drug treatment with an MAOI decreased hypothalamic temperature in Syrian hamsters. 19th Annual Meeting Society for Neuroscience Abstracts 15:852 Gao B, Duncan WC, Wehr TA (1991) Clorgyline-induced reduction in body temperature and its relationship to vigilance states in Syrian hamsters. Neuropsychopharmacology 4:187-197 Garattini S, Mennini T, Samanin R (1989) Reduction of food intake by manipulation of central serotonin. Br J Psychiatry 155 [suppl 8] :41-51 Guan XM, McBride WJ (1988) Fluoxetine increases the extracellular levels of serotonin in the nucleus accumbens. Brain Res Bull 21 : 43-46 Gudelsky GA, Koenig JI, Mettzer HY (t986) Thermoregulatory responses to serotonin (5 HT) receptor stimulation in the rat. Neuropharmacology 25 : 1307-1313

329 Hilaire SK-DS, Merica H, Gaillard JM (I 984) The effects of indalpine - a selective inhibitor of 5-HT uptake - on rat paradoxical sleep. Eur J Pharmacol 98: 413-418 Hwang EC, Magnussen I, Van Woert MH (1980) Effects of chronic fluoxetine administration on serotonin metabolism. Res Commun Chem Pathol Pharmacol 29:79 98 Kleinlogel H, Burki HR (1987) Effects of the selective 5-hydroxytryptamine uptake inhibitors paroxetine and zimeldine on EEG sleep and waking stages in the rat. Neuropsychobiology 17:20(~212 Kluger MJ, Corm CA, Franklin B, Freter R, Abrams GD (1990) Effect of gastrointestinal flora on body temperature of rats and mice. Am J Physiol 258:R552-557 Kulakowski EC, Hurst JH, Miller T, Linnoila M, Lovenberg W (1984) The hypothermic effect of 4-(5,6-dimethyl-2-benzofuranyl) piperidine HGL (CGP 6085 A) in Wistar kyoto rats Life Sci 35:1329-1335 Lin MT (1978) Effects of specific inhibitors of 5-hydroxytryptamine uptake on thermoregulation in rats. J Physiol 284:147-154 Marsden CA, Conti J, Strope E, Curzon G, Adams RN (1979) Monitoring 5-hydroxytryptamine release in the brain of the freely moving unanaesthetized rat using in vivo voltametry. Brain Res 171 : 85-99 McCleary P, Leander JD (1982) Meperidine-induced hypothermia in the rat. Neuropharmacology 21 : 1337-1341 Montgomery SA (1989) The efficacy of ttuoxetine as an antidepressant in the short and long term. Int Clin Psychopharmacol 4 [suppl 1]: 11-39 Myers RM, Yaksh TL (1968) Feeding and temperature responses in the unrestrained rat after injections of cholinergic and aminergic substances into the cerebral ventricales. Physiol Behav 3: 917-928 Nicholson AN, Pascoe PA (1988) Studies on the modulation of the sleep-wakefulness continuum in man by flnoxetine, a 5~HT uptake inhibitor. Neuropharmacology 27 : 597-602 Obal Jr F, Rubicsek G, Alfoldi P, Sary G, Obal F (1985) Changes in the brain and core temperatures in relation to various arousal states in rats in the light and dark periods of the day. Pflugers Arch 404:73-79 Parmeggiani PL (1980) Temperature regulation during sleep: a study in homeostasis. In: Orem J, Barnes CD (eds) Physiology in sleep (Research topics in physiology). Academic Press, New York, pp 97-143 Parmeggiani PL (1987) Interaction between sleep and thermoregulation: an aspect of the control of behavioral states. Sleep 10:426-435 Pastel RH, Fernstrom JD (1987) Short-term effects of fluoxetine and triflouromethylphenylpiperazine on electroencephalographic sleep in the rat. Brain Res 436:92-102 Pawlowski L (1984) Amitriptyline and femoxetine, but not clomipramine or citalopram, antagonize hyperthermia induced by directly acting 5-hydroxytryptamine-like drugs in heat adapted rats. J Pharm Pharmacol 36:197-199 Reigle TG, Wolf HH (1974) Potential neurotransmitters and receptor mechanisms involved in the central control of body temperature in golden hamsters. J Pharmacol Exp Ther 189:97-109

Reyes RB, Hill SY, Kupfer DJ (1983) Effects of acute doses of zimelidine on REM sleep in rats. Psychopharmacology 80:214-216 Reyes RB, Hill SY, Kupfer DJ (1986) Effects of repeated zimelidine administration on sleep parameters in the rat. Psychopharmacotogy 88: 54-57 Ross RJ, Ball WA, Gresch PJ, Morrison AR (1990) REM sleep suppression by monoamine reuptake blockade: development of tolerance with repeated drug administration. Biol Psychiatry 28: 231-239 Sakaguchi S, Glotzbach SG, Heller HC (1979) Influence of hypothalamic and ambient temperatures on sleep in kangaroo rats. Am J Physiol 237:R80-R88 Schmidt MJ, Fuller RW, Wong DT (1988) Fluoxetine, a highly selective serotonin reptake inhibitor: a review of preclinical studies. Br J Psychiatry 153 [suppl 3]:40-46 Shipley JE, Kupfer DK, Deaty RS, Griffin SJ, Coble PA, McEachran AB, Grochocinski VJ (1984) Differential effects of amitriptyline and of zimelidine on the sleep electroencephalogram of depressed patients. Clin Pharmacol Ther 36:251-259 Slater IH, Jones GT, Moore RA (1978) Inhibition of REM sleep by fluoxetine, a specific inhibitor of serotonin uptake. Neuropharmacology 17:383-389 Sommerfelt L, Hauge ER, Ursin R (1987) Similar effect on REM sleep but differential effect on slow wave sleep of the two 5-HT uptake inhibitors zimeldine and alaproclate in cats and rats. J Neural Transm 68:127-144 Vogel GW (t975) A review of REM sleep deprivation. Arch Gen Psychiatry 32:749-761 Vogel GW (1983) Evidence for REM sleep deprivation as the mechanism of action of antidepressant drugs. Prog Neuro Psychopharmacol Biol Psychiatry 7:343-349 Wehr TA (1990) Sleep loss: a preventable cause of mania and other excited states. J Clin Psychiatry 50 [12, suppl]:8-16 Wong DT, Bymaster FP (1981) Subsensitivity of serotonin receptors after long-term treatment of rats with fluoxetine. Res Comm Chem Path Pharmacol 32: 41-51 Wong DT, Horng JS, Bymaster FP, Hauser KL, Molloy BB (1974) A selective inhibitor of serotonin uptake: Lilly 110140, 3-(ptrifluoromethylphenoxy)-n-methyl-3-phenylpropylamine. Life Sci 15:471-479 Wong DT, Bymaster FP, Horng JS, Molloy BB (1975) A new selective inhibitor for uptake of serotonin into synaptosomes of rat brain: 3-(p-trifluoromethylphenoxy)-n-methyl-3-phenylpropylamine. J Pharmacol Exp Ther 193:804--811 Wong DT, Reid LR, Bymaster FP, Threlkeld PG (1985) Chronic effects of fluoxetine, a selective inhibitor of serotonin uptake, on neurotransmitter receptors. J Neural Transm 64:251-269 Wozniak KM, Aulakh CS, Hill JL, Murphy DL (1988) The effect of 8-OH-DPAT on temperature in the rat and its modification by chronic antidepressant treatments. Pharmacol Biochem Behav 30: 451456 Zohar J, Insel TR, Zohar-Kadouch RC, Hill JL, Murphy DL (1988) Serotonergic responsivity in obsessive-compulsive disorder. Arch Gen Psyhiatry 45:167-172