Effects of bombesin and gastrin-releasing peptide on memory processing

Effects of bombesin and gastrin-releasing peptide on memory processing

314 Brain Research. 46i) c1~)88) 314 322 Elscvicl BRE 13927 Effects of bombesin and gastrin-releasing peptide on memory processing James F . F l o...

673KB Sizes 0 Downloads 4 Views

314

Brain Research. 46i) c1~)88) 314 322 Elscvicl

BRE 13927

Effects of bombesin and gastrin-releasing peptide on memory processing James

F . F l o o d 1'2 a n d J o h n E . M o r l e y 1'3

1Geriatric Research, Education and Clinical Center (GRECC) Veterans Administration Hospital, Sepulveda, CA 91343 ( U. S. A. ) and Departments of 2psychiatry and Biobehavioral Sciences and 3Medicine, School of Medicine, University of California, Los Angeles, CA 90024 (U.S.A. ) (Accepted 5 April 1988)

Key words: Bombesin; Gastrin-releasing peptide: Gut peptide; Memory: Mouse; Retention

We have previously shown that feeding mice immediately followingtraining enhances memory retention and that one of the gastrointestinal hormones released during a meal. cholecystokinin, also enhances retention after peripheral administration. In the studies reported here we demonstrate that another gastrointestinal peptide, gastrin-releasing peptide (GRP), enhances retention after peripheral administration, as does its amphibian counterpart, bombesin. GRPt4_27 had the same effect as the intact peptidc, while GRPt_I6 was ineffective at enhancing retention. The dose-response curves showed a characteristic inverted U-shape with high doses of both GRP and bombesin being amnestic. The effect of both peptides was time-dependent and both reversed amnesia induced by the anticholinergic, scopolamine, l.c.v, administration of the peptides required higher doses to produce an effect on memory retention. suggesting that the effect was mediated predominantly through a peripheral mechanism. Doses of the peptides that enhanced memory retention after peripheral administration failed to increase serum glucose, suggesting that glucose modulation was not the mechanism by which GRP and bombesin modulate memory processing. Vagotomy inhibited the memory-enhancing effects of both GRP and bombesin, suggesting that these peptides produced their effect by stimulating ascending vagal pathways. These studies, together with our previous study with cholecystokinin, suggest the existence of a gastrointestinal hormonal system, which is activated by the passage of food through the intestine, that enhances memory retention.

INTRODUCTION Feeding an animal immediately following training enhances retention 17 and cholecystokinin (CCK), one of a family of peptides released as a meal passes through the gut, modulated retention for footshock avoidance training5'7-9A8'23. CCK was found to improve retention in both a time- and dose-dependent manner. CCK prevented amnesia induced by scopolamine, a cholinergic antagonist. C C K ' s ability to enhance retention when administered i.p. was blocked by vagotomy. The purpose of the present experiments was to test the generality of this finding to another gastrointestinal peptide, gastrin-releasing peptide (GRP). G R P was chosen because like CCK it decreases feeding 29'32. The studies also included

the amphibian form of G R P , bombesin (BBS), which is approximately 10 times more potent in its effects on feeding 2'2°. MATERIALS AND METHODS

Subjects After 1 week in the laboratory, CD-1 male mice obtained from Charles River Breeding Laboratories, Wilmington, MA, were individually caged 24-48 h prior to training and remained singly housed until retention was tested 1 week later. A n i m a l rooms were on a 12 h l i g h t - d a r k cycle with lights going on at 06.00 h. The median body weight was 35 g, with a range of 33-38 g. The mice were trained between 07.00 and 15.00 h. Mice were assigned randomly to

Correspondence: J.F. Flood, 151A2, VA Medical Center, Sepulveda, CA 91343, U.S.A. 0006-8993/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)

315 groups of 15 except where otherwise indicated.

Drugs BBS, G R P and G R P fragments were obtained from Peninsula Laboratories, Belmont, CA. Scopolamine hydrobromide trihydrate was obtained from Sigma Chemical Co., St. Louis, MO. Stock solutions were prepared in 0.1% bovine albumin serum and frozen. Test solutions were diluted with saline. Mice received a 0.35 ml i.p. injection of the vehicle or peptide solution within 2 min after training. All solutions were blind-coded to eliminate experimenter bias.

T-maze footshock avoidance apparatus The T-maze was described in detail previously t°. In brief it consisted of a black plastic start alley with a start box at one end and two goal boxes at the other; a stainless-steel rod floor ran throughout the maze. The start box was separated from the start alley by a plastic guillotine door which prevented the mouse from moving down the alley until the training started. Clear plastic liners were used to remove mice from the goal boxes.

5 avoidance responses in 6 consecutive training trials. Trials to criterion was the measure of retention. A non-parametric measure of retention was derived to better visualize the effects of drug treatments on retention test performance and to correspond with usual reporting practice. For this, the number of trials to the first avoidance response was dichotomized to yield a percent recall score. Those mice making their first avoidance in 3 trials or less were classed as remembering the original training. This criterion was adopted because it has provided optimal separation between the retention test performance of naive mice (with no T-maze training) and well-trained mice ~°. The recall score shows at least a +0.90 correlation with trials to criterion.

Statistical analysis The overall significance of the drug treatment effect was determined by a one-way or two-way analysis of variance ( A N O V A 24'34) run on trials to criterion. Dunnett's t-test was used to make multiple comparisons between each drug group and the control group ~'34. Tukey's t-test was used to make statistical comparisons among experimental groups 4'34.

T-maze training The training parameters were set as follows unless otherwise indicated. The intertrial interval was 30 s with a muffled buzzer (55 db) as the conditioned stimulus and footshock set at 0.30 mA. A training trial started when a mouse was placed into the start box. The guillotine door was raised and the buzzer sounded simultaneously, then 5 s later footshock was applied. At the end of each trial, the mouse was removed from the goal box using the liner and returned to its cage. A new trial began by placing the mouse in the start box, sounding the buzzer and raising the guillotine door, with footshock beginning 5 s later if the mouse did not move into its correct goal box. As training proceeded, a mouse made one of two types of responses. A response latency longer than 5 s was classed as an escape from the footshock. A response latency less than or equal to 5 s was considered an avoidance, since the mouse avoided receiving a footshock.

T-maze retention test One week after training and drug administration, the T-maze training was resumed until the mice made

RESULTS

Experiment 1. Dose-dependent modulation of memory processing by i.p. BBS and GRP The purpose of this experiment was to determine the range of doses over which BBS and G R P improved retention (70% or greater recall score) compared to controls (25% or smaller recall score). The mice were trained in a T-maze as described above and received 1 of 5 doses of either BBS (0.1, 0.5, 1.0, 2.5 or 5.0 ~g/kg, i.p.), G R P (2.5, 5.0, 10.0, 20.0 or 40.0/~g/kg i.p.) or vehicle, within 2 min after training. Most drugs that are recognized as having the ability to improve memory retention do so when administered shortly after training zT. As intended, the retention test performance of the vehicle control group was poor, with only 20% recall score. An analysis by A N O V A of trials to criterion indicated a significant treatment effect by BBS (F5.84 = 3.91, P < 0.01). A n analysis of the means, using Dunnett's t-test, showed that 0.1 and 0.5/~g/kg had no significant effect on retention, but 1.0 and 2.5 /~g/kg significantly reduced the mean trials to crite-

316 100-

rion (Table I) compared to the saline control. The optimal e n h a n c e m e n t of retention test performance was obtained at 1.0,ug/kg with a recall score of 73% (Fig. 1). A similar analysis of test scores for mice treated with G R P indicated a significant effect on re03

tention (E~,x4 = 4.75, P < 0.001). A n analysis of the

I II

~

40"

means, using D u n n e t t ' s t-test, showed that 2.5 and

,?, g~

5.0/~g/kg had no significant effect on retention, but 10 l~g/kg (P < 0.01) and 20/xg/kg (P < 0.05) signifi-

STRONG TRAINING

20"

~o

cantly reduced the mean trials to criterion (Table I) compared to the saline control. The optimal en-

0

o.o

1.o

zb

3.b

DOSE BOMBESIN ( u g / k g

hancement of retention test performance was ob-

4b

s.b

ip)

Fig. 1. Modulation of retention by BBS as a function of different training strengths. Weak training on T-maze footshock avoidance task (see text) resultcd in poor retention in the control group. Retention was optimally improved at 1.0,~g/kg. The poor retention at 2.5 and 5.0 ng/kg could represent an amnestic effect o f BBS on retention. Under conditions ot strong training where the control group showed good retention, BBS administered 1.0 ug/kg or greater of (iRP had significantly poorer recall.

tained at 10#g/kg with a recall score of 80% (Fig. 2). E x p e r i m e n t 2. M o d u l a t i o n o f retention as a f u n c t i o n of'training p a r a m e t e r s

The training conditions described above (weak training) were set so that the saline control group would have poor retention so that enhanced retention test performance could be detected. The lower retention test scores with doses of BBS and G R P beyond the optimal, suggested that both peptides might also cause amnesia under training conditions where

An A N O V A run on trials to criterion indicated a significant drug effect for GRP-treated mice (F5.~4 = 7.17, P < 0.001) and for BBS-treated mice (F5.~4 = 9.26, P < 0.001). A subsequent analysis of mean trials to criterion using D u n n e t t ' s t-test indicated that

the control group shows good retention. To test this, the training conditions (strong training) were altered so that control group would have high recall scores ( 7 0 - 8 0 % ) so that an amnestic effect of BBS or G R P

the group injected with 10/~g/kg of G R P differed from the saline control at P < 0.05 and the groups in-

could be detected. To accomplish this, the nominal footshock level was increased from 0.30 to 0.35 m A , the buzzer intensity was increased from 55 to 65 db and the intertrial interval was 45 s instead of 30 s.

jected with 20 and 40 ~g/kg of G R P differed from the control at P < 0.01; doses of G R P less than t 0 p g / k g did not affect mean trials to criterion significantly

Mice were administered the same doses of BBS and

(Table I1). Groups receiving BBS at 1.0, 2.5 and 5.0 ~g/kg all differed from their control at P < 0.01: doses of BBS lower than 1.0]~g/kg did not significantly affect mean trials to criterion (Table lI).

G R P as in Exp. 1. The n per group was 15. Both BBS and G R P caused significant amnesia at doses equal to or higher than the dose found to be optimal for improving retention in Expt. 1 (Figs. 1,2). TABLE 1

F~'ct o/ BBS and GRP on retention for T-rfiaze footshock avoidance training in weakly trained mice

The values in the table are mean (+ S.E.M.) trials to make 5 avoidances in 6 consecutive trials. The P-values are based on Dunnett's ttest for differences from the control (0l~g/kg) group. Dose (.ug/kg i.p. )

BBS

0

O. 1

9.40_+0.46

8.47+(I.4(I 8.67+0.48

P<

GRP 9.20_+0.43 p<

0.5

1.0

2.5

5.0

10.0

20.0

40.0

7.13+(I.45 8.(10+_(I.32 9.13_+0.43 (1.01

(1.(15

9.07+0.42

8.33+0.48

6.87_+(I.26 7.87_+0.44 8.53_+0.38 0.01 (I.{)5

317

100-

Experiment 3. Time-dependent improvement of memory retention by post-training administration of BBS and GRP Drugs which are recognized as improving memory retention show a decline in effectiveness as the time from the end of training until the drug is administered increases 27. In this experiment, we determined whether BBS and GRP showed time-dependent facilitation of retention for t-maze footshock avoidance training. Mice were trained as in Expt. 1 and were administered either BBS or GRP at the dose yielding the greatest improvement in retention. Separate groups of mice received an injection of BBS (1 ~g/kg i.p.) 0, 30 or 60 rain after training or GRP (10/tg/kg i.p.) at 0, 30, 60 or 90 min after training. Saline was administered 0, 60 or 90 min after training. Retention was tested 1 week after training. The results indicated that both BBS and GRP yielded a time-dependent facilitation of retention typical of drugs which improve retention (Table III).

80-

kkx

tp,` ,, \

60-

d~////'/

/~ll,

/

\ \

WEAK TRAINING "

ii

\',\ G TRAINING

20~

oo

~

~'o

~'5

2b

z'5

3'0

~'5

4'o

DOSE GRP (ug/kg ip)

Fig. 2. Modulation of retention by G R P as a function of different training strengths. Weak training on T-maze footshock avoidance task (see text) resulted in poor retention in the control group thus allowing detection of improved retention as the result of a post-training injection of G R P . Retention was optimally improved at 111.0,ug/kg. The poor retention at 20 and 40 ug/kg could represent an amnestic effect of G R P on retention. To test this, strong training was used which resulted in good retention in the control group. Groups administered 10 ug/kg or more G R P had significantly poorer recall.

T A B L E II

The amnestic effect of BBS and GRP on retention for T-maze footshock avoidance training in well-trained mice The values in the table are m e a n (__+S.E.M.) trials to make 5 avoidances in 6 consecutive trials. The P-values are based on D u n n e t t ' s ttest for differences from the control (0/~g/kg) group.

Dose (/lg/kg i.p. )

BBS P< GRP P<

0

0.1

0.5

1.0

2.5

5.0

7.07+0.34

7.00_+0.23

7.87+0.49

9.20+0.37 0.01

9.40_+0.40 0.01 7.27+0.33

9.47_+0.48 0.01 7.87-+0.52

7.40_+0.36

10.0

20.0

40.0

8.80+0.42 0.05

9.53+0.32 0.01

9.40+0.33 0.01

T A B L E Ill

Time-dependent/~wilitation ~f T-maze,/~otshock avoidance training by BBS and GRP Experimental values are mcan _+ S.E.M. The P-values are based on D u n n e t t ' s t-test for differences from the control injected 0 min after training.

Time of administration after training (min) 0 Saline Recall score BBS P< Recall score Saline Recall score GRP P< Recall score

(%)

(%) (%)

(%)

9.00 20 7.27 0.01 67 9.53 20 7.27 0.01 73

30 + 0.36 + 0.34

7.60 + 0.38 0.05 53

60 9.20 _+ 0.32 20 8.67 _+ 0.37

n.s. 27

+ 0.32 + 0.28

7.67 _+ 0.36 0.01 60

90

8.40 _+ 0.38 0.05 40

9.60 + 0.41 13 9.00 + 0.47

n.s.

20

318 Since there was no significant difference across control groups all statistics were run using scores from the saline group injected 0 rain after training. An A N O V A run on trials to criterion indicated that BBS showed a time-dependent facilitation of retention (&,5~, = 5.58, P < 0.005) with significantly fewer trials required to reach criterion than the saline control when BBS was administered 0 min (P < 0.01) or 30 min (P < 0.05) after training: BBS administered 6(} rain after training did not significantly reduce trials to criterion compared to the saline control. G R P yielded a similar time-dependent effect on trials to criterion (&,:0 = 6.9l, P < 0.001) in which G R P administered either 0 min (P < 0.01), 30 min (P < 0.01) or 60 min (P < 0.05) but not 90 min after training required significantly fewer trials to reach criterion than the saline control. Experiment 4. Anti-amnestic effect o f B B S and G R P We and others have reported that amnesia caused by inhibitors of brain protein synthesis can be blocked by a variety of neurotransmitter agonists and hormones <11-141<1931. Recently, we found that adrenocorticotropic hormone~_10 (ACTHa_>) and drugs affecting 6 neurotransmitter systems blocked scopolamine-induced amnesia 15-t'). The purpose of Expt. 4 was to determine if BBS and G R P would block amnesia caused by the [email protected] receptor antagonist, scopolamine (SCO). The training conditions in this experiment were as in Expt. 2 where the parameters were altered to assure that control mice would have high recall scores (70-80%) in order to detect amnesia caused by scopolamine. Retention was tested one week after training. The n per group was 15. SCO (1 mg/kg, s.c.) or saline (S) was injected immediately after training and BBS (1 ,ug/kg i.p.), G R P (10 /~g/kg i.p.) or saline was administered 30 rain

later. Four groups were used: S(S), SCO(S), SCO(BBS) and SCO(GRP). The results indrcated that the S(S) group remembered well, with an 80% recall score. SCO(S) induced amnesia (2(Vf recall score). The group that received scopolamine plus BBS (SCO(BBS)) had a recall score of 8(V~, and the group receiving scopolamine plus G R P (SCO(GRP)) had a recall score of 70%, indicating that both BBS and G R P counteracted SCO-induced amnesia. An A N O V A run on trials to criterion indicated a significant drug effect (E~.> = 4.23, P < (I.025). Mice treated only with scopolamine took significantly more trials to reach criterion than saline-treated controls (P < 0.01). A further analysis of mean trials to criterion showed that SCO groups receiving BBS or G R P required significantly fewer trials to reach criterion than the group receiving SCO plus S (t' < 0.()1, Tukey's t-test). The mean trials to criterion (+_ S.E.M.) were; S(S), 6.90 +_ 0.43; SCO(S), 8.90 __ 0.55: SCO(BBS), 7.10 _+ 0.29; S C O ( G R P ) , 7.5 +_ (I.53. Experiment 5. Effect of" G R P peptide ]?agments on memory retention BBS was first isolated from toad skin t. More recently, its mammalian form G R P was isolated e". Table IV shows the sequences of amino acids for both peptides. The G R P peptide 14-27 and BBS share a common sequence of amino acids except for a Gin/His replacement at position 20 of the G R P peptide. The purpose of this experiment was to determine if G R P peptide fragments showed differential ability to enhance retention. The peptide fragments were administered on an equimolar basis as follows: GRPL_,v (10 ,ug/kg i.p.), GRPI_I6 (6.2 ug/kg i,p.), GRPI4_27 (5.6 ~tg/kg i.p.). Mice were trained as in Expt. 1 and immediately after training received an injection of either saline or G R P peptides. Retention

TABLE IV Amino acid sequences /i:wBBS and (;Rt" BBS

pGlu-Gln-Arg-Leu-Gly-Asn-Gln-Trp-Ala-Val-Gly-His-Leu-Met-NH~

GRP

Ala-Pro-Val-Scr-VaI-Gly-Gly-Gly-Thr-VaI-Leu-Ala-Lys1

-Mct-Tyr-Pro-Arg-Gly-Asn-His-'lrp-Ala-Val-Gly-His-Leu-Met-NH_~ 14 1~ 2O 27

319 TABLE V Effect of i. c. v. administration of B BS or GRP on retention

The values in the table are mean (+ S.E.M.) trials to make 5 avoidances in 6 consecutive trials. The P-values are based on Dunnett's ttest for differences from the control (0 ug/kg) group. Dose (~zg/brain)

BBS P< GRP P<

0

0.05

0.10

0.50

9.40 + 0.40

8.73 + 0.33

8.33 + 0.44 0.05

6.73 + 0.28 0.01 8.87 +_0.40

9.60 + (I.38

was tested 1 week later. The saline control showed poor retention with 13% recall score as did GRPI_I6 with 20% recall score. Mice receiving GRPl_27 or GRPI4_27 had 73% recall scores. A n A N O V A run on trials to criterion indicated a significant treatment effect (F3.56 = 11.08, P < 0.001). A further analysis of the means indicated that groups receiving GRPI_27 or GRPla_27 took significantly fewer trials to reach criterion than the saline-treated group (P < 0.01, Dunnett's t-test). The group receiving GRPI_16 did not differ significantly from the saline control. The mean trials to criterion (_+ S.E.M.) were as follows: saline 8.73 + 0.38; GRP~_27 7.13 + 0.33; GRPI_~6 9.60 + 0.47; GRPI4_277.13 + 0.32. Experiment 6. Effect o f i. c. v. administration o f B B S and G R P on memory retention

We tested if BBS and G R P affected memory processing by acting directly on the central nervous system. Mice were trained as in Expt. 1. At 48-24 h before training a single hole was drilled over the third ventricle (-0.5 mm relative to bregma, 0.5 right of the central suture) while mice were in a stereotaxic instrument under methoxyflurane (Metofane) anesthesia. Within 3 min after training, mice were anesthetized with enflurane, returned to the stereotaxic apparatus and received BBS or G R P in 2/~1 of saline or saline alone into the third ventricle. The doses of BBS were: 0, 0.05, 0.10, 0.50fig. The doses of G R P were: 0, 0.05, 0.10, 0.25fig. The saline control groups showed poor retention (13% recall score). A n A N O V A run on trials to criterion yielded a significant drug effect (~,56 = 10.30, P < 0.001). A further analysis of the means for mice receiving BBS by Dunnett's t-test showed that the

1.0

2.5

7.87 + 0.36 0.01

7.33 + 0.33 0.01

group receiving 0.1 ktg differed significantly from the saline control at P < 0.05 and the group receiving 0.5 ~g at P < 0.01. The trials to criterion for groups given G R P similarly yielded a signifcant treatment effect (F3,56 = 8.10, P < 0.001). Both groups receiving G R P at 1.0 and 2.5/~g took significantly (P < 0.01) fewer trials to reach criterion than the saline control (Table V). Experiment 7. Effect o f vagotomy on memory enhancement by B B S and G R P

The relatively high doses of BBS and G R P needed to enhance retention, after central administration, suggested a peripheral mechanism of action of the gut peptides on memory processing. CCK's enhancing effect on retention was blocked by gastric vagotomy 17. The purpose of this experiment was to determine if the memory-enhancing effects of BBS and G R P are also mediated by the vagus nerve. In this experiment T A C ( S W ) male mice 8 weeks of age were obtained from Taconic Farms Inc., Germantown, NY. Pilot studies had indicated that the memory-enhancing doses of BBS and G R P enhanced retention in T A C ( S W ) mice as they did in CD-1 mice. Saline, BBS (1/~g/kg i.p.) or G R P (10/~g/kg i.p.) were administered to vagotomized, sham-operated or non-operated mice (9 groups). The success of the vagotomies was determined by the ratio of stomach to body weight 17. BBS and G R P enhanced retention in sham and non-operated mice but not in vagotomized mice (Fig. 3). A two-way A N O V A (operation by drug treatment) indicated a significant experimental effect of the operation (F2.99 = 8.32, P < 0.001) and drug treatment (F2,99 = 8.32, P < 0.001). The interaction of operation and drug treatment was not significant.

glucose compared to saline injected control (220 ± 52 mg/dl).

lll~

~ S^LIN~

~

80m[SlM (! ug/kg tp)

,m/

t~

8tt}

VAGOTOMY

SHAH TREATMENT GROUPS

Fig. 3. Effect of vagotomy on retention improved by BBS or GRP administered immediately after T-maze footshock avoidance training. BBS and GRP failed to enhance retention in vagotomized mice but did enhance retention significantly in sham-operated and non-operated controls. A partitioning of the sum of squares indicated that among the drug conditions a significant effect was obtained for BBS (F2.99 = 4.01, P < 0.025) and for G R P (F2m = 8.74, P < 0.001). Among the operations a significant effect was obtained with sham-operated mice (~,99 = 4.98, P < 0.025) and with non-operated mice (F~.99 = 7.30, P < 0.005). A further analysis of mean differences in trials to criterion between the saline-injected controls and peptide-treated groups for each operated condition indicated that BBS and GRP enhanced retention only in sham and non-operated mice (P < 0.05).

Experiment 8. Effect of BBS and GRP administration on blood serum glucose Post-training administration of glucose was reported to enhance retention for passive avoidance training =. BBS and GRP elevate blood serum glucose levels when administered to rats 3'>. Thus, another possible mechanism for peripheral mediation of BBS's and GRP's enhancement of retention may be increasing circulating levels of glucose shortly after training. To determine if BBS and GRP significantly increased blood serum glucose, mice received either saline, BBS (1 ~g/kg), or GRP (10 flg/kg) and were sacrificed 15 min later to obtain blood plasma. The n per group was 15. Serum glucose was determined by an enzymatic (hexolunase) assay; the reagents were obtained from Sigma Chemical Co., St. Louis, MO. Neither BBS (226 +- 9 mg/dl) nor GRP (217 +_ 6 mg/dl) increased mean (+ S.E.M.) serum

DISCUSSION Previously, we demonstrated that feeding mice immediately following training enhanced memory retention and that one of the gastrointestinal hormones released during a meal, CCK, also enhanced retention after peripheral administration 17. In the studies reported here, we show that BBS and GRP enhanced retention test performance when administered immediately after training and one week before testing retention. Based on the dose yielding the greatest improvement in retention test scores BBS (1 ug/kg) was found to be 10 times more potent than G R P (10 ug/kg) which is close to the difference in potency found for inhibiting feeding in food deprived animals 2"2° and inducing hyperglycemia in rats 3. At doses higher than optimal for improving retention, BBS and GRP show reduced effectiveness at improving retention (Expt. 1). This inverted U-shaped doseresponse function is typical of compounds which modulate retention. Under conditions of training where the saline control group remembered, BBS and GRP caused amnesia, starting at the optimal dose for enhancing retention in under trained mice. Thus the right side of the inverted U-shaped curve actually represents an amnestic effect due to overdose. Both BBS and GRP showed time-dependent facilitation of retention such that the longer after training they were administered the less likely they were to facilitate retention. This also indicates that BBS and GRP did not act proactively to have a direct effect on retention test performance per se. BBS and GRP blocked the amnestic effect of scopolamine which is an effect shared by a diverse group of compounds (e.g. ACTH4_m, D-amphetamine, arecoline, clonidine, CCK octapeptide, fluoxetine, hydergine, naloxone, nalmefene, neuropeptide Y, picrotoxin, piribedal, strychnine, tacrine) all of which modulate memory retention ts w. GRPI~_z7 administered on an equimolar basis with GRP facilitated retention but GRP~_I6 did not. Since GRP~a_27 is structurally very similar to BBS, one might have expected this fragment to have greater potency and thus should have been administered on

321 an equimolar basis with BBS. This suggests that the 20th amino acid which differs in BBS and G R P may prevent its enzymatic destruction and account for its greater potency in mammals. It is clear from the greater quantity of BBS and G R P required to enhance retention by i.c.v, administration (0.5 ~g and 2.5 ~g, respectively) compared to i.p. administration (0.035 Bg and 0.35 ~g respectively) that BBS and G R P probably modulate m e m o r y processing by a peripheral mechanism even when administered centrally. However, we cannot rule out the possibility that substantially lower doses would improve retention with localized injections. Two possible peripheral mechanisms of action were studied: elevation of blood serum glucose level and vagal stimulation. Vagotomy blocked the enhancement of retention by BBS and G R P indicating that vagal stimulation is one mechanism by which they improved m e m o r y processing. This is similar to the ability of vagotomy to inhibit the memory-enhancing effect of CCK octapeptide 17. The ability of CCK to inhibit feeding is also blocked by vagotomy 31. Vagotomy alone is insufficient to block the inhibition of feeding by BBS 28. However, a combination of vagotomy plus high thoracic cord section does block the feeding inhibition produced by BBS 33. Thus the mechanism by which BBS modulates feeding appears to be different from that through which it enhances memory.

REFERENCES 1 Anastasi, A., Erspamer, V. and Bucci, M., Isolation and structure of bombesin and alytesis, two analogous active peptides from the skin of the European amphibians Bomina and Alytes, Experientia, 27 (1971) 166-167. 2 Billington, C.J., Levine, A.S. and Morley, J.E., Are peptides truly satiety agents? A method of testing for neurohumoral satiety effects, Am. J. Physiol., 265 (1983) R920-R929. 3 Brown, M., Marki, M. and Rivier, J., Is gastrin releasing peptide mammalian bombesin?, Life Sci., 27 (1980) 125-128. 4 Bruning, J.L. and Kintz, B.L., Computational Handbook of Statistics, second edition, Scott, Foresman, Glenview, IL, 1987, pp. 18-30, 122-124, 128-130. 5 Cohen, S.L., Knight, M., Tamminga, C.A. and Chase, T.N., Cholecystokinin-Octapeptide effects on conditionedavoidance behavior, Stereotypy and catalepsy, Eur. J. Pharmacol., 83 (1982) 213-222. 6 Davis, H.P. and Squire, L.R., Protein synthesis and mere-

Since administration of glucose also improves retention 22, increased serum glucose levels may be another mechanism of action by which BBS and G R P enhanced memory processing. However, at the doses used to enhance memory, no effect on blood glucose level was detected suggesting that it is not a major mechanism by which G R P or BBS enhance memory. In conclusion, we have demonstrated that both BBS and G R P enhance memory by a peripheral mechanism which most probably involves activation of the ascending vagal fibers. These studies, together with our previous study on CCK TM suggest the existence of a gastrointestinal hormonal system, which is activated by the passage of food through the intestine, that enhances memory retention. From an evolutionary perspective such a system could play an important role in the successful foraging for food.

ACKNOWLEDGEMENTS The research was supported by Medical Research Service of the Veterans Administration and by the Sepulveda Geriatric Research, Education and Clinical Center ( G R E C C ) . We express our appreciation to Ernesto N. Hernandez and Bridget S. Mercy for technical assistance. Vagotomies were performed under the supervision of L. Gunther, Taconic Farms Inc.

ory: a review, Psych. Bull., 96 (1984) 518-559. 7 Deupree, D. and Hsiao, S., Cholecystokinin octapeptide, proglumide, and passive avoidance in rats, Peptides, 8 (1987) 25-28. 8 Fekete, M., Szabo, A., Balazs, M., Penke, B. and Telgedy, G., Effects of intraventricular administration of cholecystokinin octapeptide sulfate ester and unsulfated cholecystokinin octapeptide on active avoidance and conditioned feeding behavior of rats, Acta Physiol. Acad. Sci. Hung., 58 (1981) 39-45. 9 Fekete, M., Kadas, T., Penke, B. and Telgedy, G., Modulation of passive avoidance behavior by cholecystokinin octapeptide in the rat, Neuropeptides, 1 (1981) 301-307. 10 Flood, J.F., Bennett, E.L., Rosenzweig, M.R. and Orme, A.E., The influence of duration of protein synthesis inhibition on memory, Physiol. Behav., 10 (1975) 555-562. 11 Flood, J.F., Jarvik, M.E., Bennett, E.L., Orme, A.E. and Rosenzweig, M.R., The effect of stimulants, depressants, and protein synthesis inhibition on retention, Behav. Biol., 20 (1977) 168-183. 12 Flood, J.F., Bennett, E.L., Orme, A.E., Rosenzweig,

322

I3

14

15

16 17

18

19

20 21

22

23

M.R. and Jarvik, M.E., Memory: modification of anisomycin-induced amnesia by stimulants and depress~mts, Science, 199 11978) 324-326. Flood. J.F., Vidal, D., Bennett, E i . , Orme, A.E., Vasquez, S. and Jarvik, M.E., Memory facilitating and antiamnestic effects of cortico steroids, Pharrnacol. Biochem. Behav., 8 (1978) 81-87. Flood, J.F., Smith. G.E. and Cherkin, A., Hydergine enhances memory in mice, J. Pharmacol.. 16, Suppl. 111 (1985) 47-59. Flood, J.F. and Cherkin, A., Scopolamine effects on memory retention in mice: a model of dementia'?, Behav. Neurol. Biol.. 45 (1986) 169-184. Flood, J.F. and Cherkin, A., Fluoxetine enhances memory processing in mice, Psychopharmacology, 93 (1987) 36-44. Flood, J.F., Smith, G.E. and Morley, J.E., Modulation of memory processing by cholecystokinin: dependence on the vagus nerve, Science. 234 (1987) 832-834. Flood, J.F., Hernandez, E.N. and Morley, J.E., Modulation of memory processing by neuropeptide Y. Brain Research, 421 (1987)280-290. Flood, J.F., Cherkin, A. and Morley, J.E., Antagonism of endogenous opioids modulates memory processing, Brain Research. 422 (1987) 218-234. Gibbs, J., Effect of bombesin on feeding behavior, Lll/~, Sci., 38 (1985) 147-153. Gibbs, J., Fauser, D.J., Rowe, E.A., Rolls, B.J. and Maddison, S.P., Bombesin suppresses feeding in rats, Nature (Lond.). 282 (1979) 208-210. Gold, P.E., Vogt, J. and Hall, J.L., Glucose effects on memory: behavioral and pharmacological characteristics, Behav. Neural Biol., 46 (1986) 145-155. Katsuura, G. and Itoh, S., Prevention of experimental amnesia by peripherally administered cholecystokinin oct~peptide in the rat, Drug Dev. Res., 7 (1986) 269-276.

24 Keppcl, G., Design and Analysis: A Researcher's Handhook, Prentice-Hall, Englewood Cliffs, N,I. 1973, pp. 556 559. 25 Levine, A.S. and Morley, J.E., Peptidergic control of insulin-induced feeding, Peptides, 2 (1981) 261-264. 26 McDonald, T.J., Jorvnall, H., Ni[sson, G.. Vagnc, M.. Ghatei, M., Bloom, S.R. and Mutt, V., Characterization of a gastrin releasing peptide from porcine non-antral gastric tissue, Biochern. Biophys. Res. Comrnu,., 911 11979) 227-233. 27 McG~lugh, J.L.. Time-dependent proecsscs in memory storage. Science. 153 (1966) 1351-1358. 28 Morley, J.E., Levine, A.S.. Kneip, J. and Grace, M., The effect of vagotomy on the satiety effects of neuropeptides and naloxone. Life Sci., 30 (1982) 1943-1947. 29 Morley, J.E., Levine, A.S., Kneip, J., Grace, M. and Billington, C.J., The effect of peripherally administered satiety substances on feeding induced by butorphanol tartratc, Pharmacol. Biochem. Behav.. 19 (1983) 577-582. 30 Nakajima, S., Amnesic effect of cycloheximidc in the mouse mediated by adrenocortical hormones, J. Comp. Physiol. Psychol., 88 (1975) 378-385. 31 Smith, G.P., Jerome, C., Cushin, B.J., Etermo, K. and Simansky, J., Abdominal w~gotomy blocks the satiety effect of cholecystokinin in the rat, Science. 212 (1981) 1036-1037. 32 Stein, L.J. and Woods, S.C., Gastrin releasing peptide reduces meal size in rats, Peptides, 3 (1982) 833-835. 33 Stuckey, J.A., Gibbs, J. and Smith, C.P., Neural disconnection of gut from brain blocks bombesin-induced satiety, Peptides. 6 11985) 1249-1251. 34 Wirier, B.J., Statistical Principles in Experimental Design, 2rid edn., McGraw-Hill, New York, 197l, pp. 196-2111, 397-4O2.