Stress and pain responses in rats lacking CCK1 receptors

Stress and pain responses in rats lacking CCK1 receptors

I. Hurwitz a,c, O. Malkesman a,c, Y. Stern a,c, M. Schroeder b,c, Y. Lavi-Avnon b,c,M. Shayit b, Y. Shavit d, G. Wolf d, R. Yirmiya d, A. Weller b,c,*a Interdisciplinary Program in the Brain Sciences, Bar Ilan University, Ramat-Gan, IsraelbDepartment of Psychology, Bar-Ilan University, Ramat-Gan, IsraelcThe Gonda (Goldschmeid) Multidisciplinary Brain Research Center, Bar-Ilan University, Ramat-Gan, IsraeldDepartment of Psychology, The Hebrew University, Jerusalem, Israel

p e p t i d e s 2 7 ( 2 0 0 6 ) 1 4 8 3 – 1 4 8 9

a r t i c l e i n f o

Article history:

Received 12 June 2005

Received in revised form

10 October 2005

Accepted 10 October 2005

Published on line 11 November 2005

Keywords:

USV

Pain

Anxiety

Natural preference

CCK

OLETF rats

a b s t r a c t

CCK involvement in stress- and pain-responsiveness was examined by studying the beha-

vior of infant (11–12-days-old) and adult OLETF rats that do not express CCK1 receptors.

Infant odor- and texture-preferences were also assessed. We hypothesized that OLETF rats

will show behavioral patterns similar to those previously observed after CCK1 antagonist

administration. Rate of separation-induced ultrasonic vocalization was significantly greater

in OLETF compared to controls, in two separate studies. Infant pups of the two strains did

not differ in odor- and texture-preference tests. OLETF rats showed consistently longer hot-

plate paw-lift (as infants, in two separate studies) and paw-lick (as adults) latencies.

Summary: OLETF pups vocalized in isolation more than controls and showed relative

hypoalgesic responses, evident also in adulthood, in concordance with the pharmacological

literature.

# 2005 Elsevier Inc. All rights reserved.

avai lab le at www.sc iencedi rect .com

journal homepage: www.elsev ier .com/ locate /pept ides

CCK is one of the most abundant peptides in the brain [9], in

addition to its role as a gut-hormone in the periphery. CCK

binds to two types of receptors: CCK1 (formerly called CCKA),

localized mainly in the gastrointestinal tract, but also found on

the vagus and in discrete brain areas, and CCK2 (formerly

called CCKB), which are identical to peripheral gastrin

receptors, and are widely distributed throughout the CNS

[25,38–39]. Psychopharmacological research has identified

many physiological and behavioral functions of CCK (e.g.,

[7,8]). In the current study we further examined a few of the

major non-ingestive behavioral effects that might be modu-

lated by CCK: natural preferences and stress- and pain-

reactivity. Instead of the standard pharmacological agonist/

* Corresponding author at: Developmental Psychobiology Laboratory, DGan IL52900, Israel. Tel.: +972 3 5318548; fax: +972 3 535 0267.

E-mail address: [email protected] (A. Weller).

0196-9781/$ – see front matter # 2005 Elsevier Inc. All rights reserveddoi:10.1016/j.peptides.2005.10.009

antagonist approach, we used a rat model in which the CCK1

receptors are completely absent.

Recently, research has demonstrated that the Otsuka Long

Evans Tokushima Fatty (OLETF) rat congenitally lacks a 6 kb

segment in the gene for CCK1 receptors [36], and lacks the

expression of CCK1 receptors [11]. The 6 kb lesion is specific to

the promoter-region and the first and second exons of the

gene for CCK1 receptors, yet it does not play any part in

another gene, the gene for CCK2 receptors. Therefore, it is not

surprising that the expression of CCK2 receptors was shown to

be normal in OLETF rats [11]. In fact, there even may be a

compensatory over-expression of CCK2 receptors, at least at

the age of 14–34 weeks [26].

epartment of Psychology, Bar-Ilan University, Geha Road, Ramat-

.

p e p t i d e s 2 7 ( 2 0 0 6 ) 1 4 8 3 – 1 4 8 91484

OLETF rats exhibit normal behavioral patterns when

diurnal rhythm of body temperature and circadian rhythm

of sleep states are tested [33]. However, their scores are low on

large movements during the dark phase when compared to

control (Long Evans Tokushima Otsuka (LETO) rats [28]. The

hypoactive pattern of OLETF rats was demonstrated also when

OLETF dams were challenged by presenting a pup: OLETF

dams carried pups less frequently than LETO dams regardless

of the pup’s strain [19]. OLETF rats appear to be hyperphagic

from postnatal day 2 [5] and they develop obesity in adulthood

followed by non-insulin-dependent diabetes mellitus [17,24].

Differences between LETO and OLETF strains are additionally

expected in ‘‘anxiety-like responses’’ and ‘‘pain-like

responses’’ for the reasons described in the next paragraphs.

Isolation-induced ultrasonic vocalization (USV) of infant

rodents is often regarded as an animal model of anxiety

[16,29]. USV is mediated by several neurochemical systems

[6,13]. Research with psychopharmacological methods has

repeatedly supported the role of CCK in mediating USV.

Administration of a CCK1 receptor agonist (sulfated but not de-

sulfated CCK-octapeptide) significantly reduced USV [42],

while administration of a CCK2 receptor agonist (butylox-

ycarbonyl-CCK-4 (BOC-CCK-4) see Ref. [12], significantly

increased USV [12]. Furthermore, USV reduction produced

by oral ingestion of corn-oil or milk was blocked by the CCK1

receptor antagonist devazepide [4,43]. Recently, it was

demonstrated that 6–9 days old OLETF pups emitted sub-

stantially more ultrasonic vocalizations than LETO pups

during an infant–mother interaction test [19].

Anxiety-like responses have been reported also in adult

OLETF rats. Studies have shown impaired, anxiety-like

performance in elevated radial- and plus-mazes and in the

black and white box test [22,28,47], and extended latencies to

avoid a shock-associated compartment, and food-neophobia

[22], also considered ‘‘anxiety-like responses’’. [Interestingly,

CCK1 receptor knockout mice do not appear to share this

phenotype [23].] However, some of these procedures also

involve a pain sensation and may therefore be attributed to

pain-like responses.

Studies suggest that CCK and related peptides either

attenuate or potentiate morphine-induced analgesia, depend-

ing on the experimental conditions. CCK receptor antagonists,

which do not induce an analgesic response by themselves,

have consistently been shown to enhance the antinociceptive

effects of morphine and endogenous opioids and prevent

morphine tolerance. Most of this research supports the

specificity of CCK2 receptors (see review in Refs. [27 and

studies cited in 30,37,46]). In general, these results suggest that

CCK antagonizes opioid antinociception, mostly via CCK2

receptors. However some findings suggest that CCK1 receptors

may have a role too (e.g. [10,18,40]), depending on several

factors, e.g., the specific opioid ligand inducing the analgesia,

the particular pain assay [1,35], the test context [20] and the

animal species studied [14].

Although the administration of CCK-octapeptide to 11-

days-old infant rats resulted in significantly reduced USV, CCK

was nevertheless ineffective in modulating pain reactivity, as

measured by the paw-lift response from a hot surface [42].

Thus, in contrast to ‘‘anxiety-like responses’’, which were

demonstrated in both infant and adult rats, the involvement of

CCK in thermal-pain sensitivity has been demonstrated to

date only in adult but not in infant rats.

The few findings in OLETF rats are also conflicting, and are

based on different assays. One study reported that these rats

are hypoalgesic. Specifically, hot-plate paw-withdrawal

latency of adult male OLETF rats (N = 6, at 5 months of age)

was significantly longer than in LETO rats [22]. However,

another study of young adult male OLETF and LETO rats

(N = 17 per strain, at about 2 months of age) found virtually

identical mean tail-flick latencies in both strains, in response

to a thermal stimulus [21].

CCK has been implicated in mediating natural as well as

conditioned preferences, in infant rats. The research on

conditioned preferences is described in Weller et al. [44].

The research on natural preferences, more pertinent to the

current study, will be briefly reviewed. We recently focused on

CCK’s involvement in mediating the infant rat’s motivation

and natural attraction towards olfactory and tactile aspects of

the dam and nest. We found that 11–12-days-old rat pups,

administered either with the CCK1 antagonist devazepide or

the CCK2 receptor antagonist L-365,260, significantly increased

their relative preference towards furry rug texture and odor of

maternal feces [34]. Furthermore, in 9-days-old rats, devaze-

pide significantly increased preference levels for a warmer

versus a cooler floor temperature [41]. These sensory stimuli

were chosen as they represent olfactory and tactile aspects of

the dam and nest. The results suggest that activation of CCK

receptors of both types may attenuate the infant’s attraction

towards maternal-related stimuli. For the current study, we

hypothesized that the preference levels of OLETF rats will be

greater than in LETO controls, in a manner similar to the

difference found between devazepide- and control-treated

rats.

Based on the literature reviewed above, we hypothesized

that isolated OLETF rat pups will emit more USV than controls.

This was based on the findings that in situations where CCK is

activated, e.g., by milk or corn-oil, and USV levels are reduced,

an acute blockade of CCK1 receptors by devazepide increases

USV levels. Although the rationale for the prediction regarding

the pain response was less clear, we hypothesized that OLETF

pups and adults would show longer paw-withdrawal latencies

from a hot-plate, following the report by Li et al. [22]. Following

the results of Shayit and Weller [34] our directional hypotheses

regarding OLETF pup preference levels was more tentative.

1. Methods

1.1. Subjects

The 4-month-old adult male rats studied (in Experiment 2)

were shipped by air following weaning (age 24–28 days) from

Otsuka, Japan, and were raised from then on in our colony at

Bar-Ilan University (N = 23 LETO and 24 OLETF). Similarly, the

pups (male and female) studied in the other experiments were

offspring from 10 OLETF and 14 LETO pregnant rats that were

shipped to Israel from Japan, and subsequently produced

litters at Bar-Ilan University. Maternal cages were number

coded and all researchers involved were blind to the strain of

the particular dam and offspring. Each litter was adjusted to

p e p t i d e s 2 7 ( 2 0 0 6 ) 1 4 8 3 – 1 4 8 9 1485

8–10 pups on the day after birth (day 1) and was left

undisturbed until the time of testing, except for food and

water maintenance. Lights in the colony were on from 05:00 to

19:00 h and temperature was maintained between 21 and

25 8C. Separate pools of naive 11–12-days-old pups were

studied in Experiment 1a, Experiment 1b and Experiment 3

(N (LETO and OLETF) = 10 and 9, 24 and 19, 14 and 12,

respectively). Only one pup per litter participated in each

experimental condition. Pups in each experiment were the

siblings of those participating in the other experiments. No

significant strain difference was found in sex-distribution in

Experiments 1a, 3a and 3b (all chi-square analyses were NS). In

Experiment 1b the sex-distribution of the subjects in the two

strains turned out to differ significantly (chi-square = 4.47,

d.f. = 1, p < 0.05). A majority of the LETO pups were male while

a majority of the OLETF pups were female, in this experiment.

1.2. Procedure

All the experiments were performed by researchers that were

‘‘blind’’ to the animals’ strain (OLETF or LETO). In Experiments

1a and 3 the pups received an injection of a control solution,

even though no experimental solution was administered. The

rationale for including this injection was to provide similar

testing conditions as in the previous studies that used

pharmacological agents and included an injection protocol.

1.3. Experiment 1a

The procedure was similar to that described in Weller and

Blass [42]. Briefly, on day 11 or 12, between 15:00 and 18:00 h,

the pup was removed from the nest together with a few

siblings, weighed, and placed with them in a plastic cage in a

warm (33 8C) and humid incubator. Fifteen to thirty minutes

afterwards, it received an IP injection of 2 ml/kg physiological

saline, and returned to the incubator. Fifteen minutes later,

the pup was placed individually in an empty glass beaker

(500 ml) at room temperature (23.8–24.7 8C) and USV were

counted manually for 5 min. USV were made audible

throughout the test period by a QMC mini bat detector set

at 45 � 5 kHz. After the test, the pup’s front left paw was

placed on a hot-plate (48 8C) and the paw-lift latency (PLL) was

assessed manually by a stop-watch (as in Ref. [42]). An upper

limit of 20 s was utilized to avoid tissue damage.

1.4. Experiment 1b

The procedure was similar to that described in Weller and

Gispan [43]. Briefly, the subjects, two pups from each litter,

were treated as in Experiment 1a and housed in an incubator,

awaiting the test similarly, with one exception: they did not

receive an IP injection. The procedure, performed 30–45 min

after separation from the dam, was as follows. Each pup was

isolated in a polycarbonate cage (22.5 cm � 15.5 cm � 14 cm)

for a 10 min test period. This included 2 min alone in an empty

cage for baseline recording of USV and the remaining 8 min in

which the pup was introduced to one of two different

conditions in a test container of identical dimensions: 2 ml

of high fat milk (UHT Long life cream, 10% fat, Tnuva Dairy,

Israel) sweetened by addition of sucrose to make a 10%

solution, warmed to 37 8C and spread equally over heavy

tissue paper (two layers) cut to fit the bottom of the test box

(N = 11 LETO and N = 9 OLETF); or warm tap water spread on

the paper (N = 15 LETO and N = 9 OLETF). USV were counted

during the two baseline and eight test minutes, and expressed

as mean number recorded per min. The session was taped by a

video camera located above the test container. Locomotor

activity was assessed off-line by counting of the number of

lines crossed by the pup over the test session. For this, lines

were marked on the monitor, dividing the test container’s

floor area into six squares of 7.5 cm � 7.5 cm. After the test,

PLL was assessed as described in Section 1.4.

1.5. Experiment 2

Experiment 2 compared thermal-pain responses of 4-

month-old adult male OLETF and LETO rats. The method

was similar to that used by Wolf et al. [45]. The hot-plate

apparatus consisted of a 20 cm � 20 cm brass plate, sur-

rounded by 45 cm-high Plexiglas walls. The hot-plate was

maintained at 53 8C and was covered with cardboard. Each

rat was placed on the cardboard for 30 s of habituation, after

which the cardboard was removed and the latency to hind-

paw lick was measured. The rat was removed from the plate

either immediately after performing the pain response or, in

case of no response, after 60 s (cutoff), to prevent tissue

damage.

1.6. Experiment 3

The procedure was planned to be similar to that used in Shayit

and Weller [34]. Briefly, on day 11 or 12, subjects (N = 14 LETO

and N = 12 OLETF) were removed from the cage in groups of

four littermates (between 13:00 and 17:00 h), weighed and

marked. Next they received an IP injection (2 ml/kg) of the

vehicle solution (9.5% ethyl-alcohol and 4.5% carboxymethyl

cellulose (Sigma) in distilled water) used in Shayit and Weller

[34]. Then they were placed individually in cups lined with 5 g,

of fresh shavings, in an incubator at 33 8C. Twenty-five

minutes later, pups were voided by gently stroking a cotton

swab on their ano-genital area. After waiting five more

minutes in the same conditions, each pup was put in the

center of a test chamber for a 3-min texture preference test

(Experiment 3a). Next, the pup was housed individually again

in the incubator. After 45–55 min it was voided and 5 min later

put into the maternal-odor preference test chamber for 3 min

(Experiment 3b).

In Experiment 3a, the test chamber was a large poly-

propylene box (38 cm � 21 cm � 18 cm) that was divided into

two equal sections. One half of the arena was covered by furry

rug, while the other was covered by plywood. In Experiment

3b, the chamber was a similar size polypropylene box with an

uncovered floor. A line marked the midline, on which the pup

was placed for the test. On each side of the cage, 4 cm from the

end, a stainless-steel grid blocked access to the cage end. Two

lines marked the area near (3 cm) the grids. Five grams of fresh

wood-shavings were placed behind each grid. Four fresh

maternal feces were placed in a row on top of the shavings

behind one of these two grids, beyond the pup’s reach, just

before the onset of the preference test.

p e p t i d e s 2 7 ( 2 0 0 6 ) 1 4 8 3 – 1 4 8 91486

Fig. 1 – Mean body weight, ultrasonic vocalization (USV)

and pain-response (PLL) of LETO and OLETF pups.

The pup was first placed on the center line of the chamber

and then picked up after 30 s and replaced in the opposite

direction on the midline, in order to control for individual

tendencies to turn in one or the other direction. This 1808

rotation was repeated again every 30 s. The time that the pup

stayed on each side: near the stimulus or the side away from the

stimulus was recorded. The measure of preference used in

Experiment 3a, as in Refs. [36,38] was the total time spent on the

rug-covered side of the test arena. In contrast, the stimulus in

Experiment 3b was located at the far end of the test arena, so

that the pup was exposed to an odor gradient. Accordingly, and

following pilot studies, two measures of odor-preference were

used (as in Ref. [34]). The first was the time spent in proximity

(<3 cm) to the grid, beyond which the feces were located (odor

preferencemeasure1:OPM1).Because itappears thatpupswere

interested in the grid itself when encountering it on any of the

two sides of the arena, a second measure was used: time spent

near (<3 cm) the grid on the odor side minus time spent near the

grid on the opposite side (odor preference measure 2: OPM2).

1.7. Statistical analysis

The results were analyzed by t-test or analysis of variance

(ANOVA). Significant main effects were followed by post hoc

Scheffe tests.

Fig. 2 – Mean rate of vocalizing (USV per minute) during a

baseline 2 min period, and during the presence of milk or

water for the ensuing 8 min, in OLETF compared to LETO

pups.

2. Results

2.1. Experiment 1a

In the first experiment, nine OLETF pups were compared to 10

LETO controls. The results are depicted in Fig. 1. The OLETF

pups weighed significantly more than their LETO controls

(means = 24.6 versus 18.5 g, t(17) = 2.60, p < 0.05). In addition

they vocalized at a significantly higher rate than LETO controls

(means = 108.4 versus 20.7 USV per min, t(17) = 2.91, p < 0.01)

and their latency to remove the front left paw from a hot-plate

was significantly longer than LETO controls (means = 7.2

versus 2.6 s, t(17) = 2.67, p < 0.05). Note that the mean USV

of the OLETF pups was above five-fold greater than controls,

and the mean PLL of the OLETF pups was about three-fold

greater than controls.

2.2. Experiment 1b

OLETF pups weighed significantly more than LETO controls

(means = 24.7 g versus 20.0 g, p < 0.001). Fig. 2 shows the mean

USV produced by infant rats in the isolation test. Rate of USV

(USV/min) was significantly greater (five-fold) in OLETF

compared to LETO pups during the baseline isolation test,

the first 2 min, (t = 3.197, d.f. = 42, p < 0.005). Activity levels did

not differ significantly, in contrast to the hypoactive profile of

adult OLETF rats [22,31]. Similarly, the rate of USV was

significantly greater during the subsequent 8 min in OLETF

compared to LETO pups (one-way ANOVA F = 6.56, d.f. = 3,40,

p < 0.001), with no difference in activity levels. Post hoc tests

revealed that the USV rates were significantly greater when

OLETF and LETO rats were compared during the 8 min interval

in the presence of water (p < 0.004) but not in presence of milk.

It could be suggested that the decrease of USV responses in

the presence of milk results from competition between food

intake and USV responses. However, the pups did not ingest a

significant amount in this study: mean percent weight gain was

not different from zero (one-tailed t-test: p > 0.86). In addition, a

two-way ANOVA failed to find differences in body-weight gain

between OLETFand LETOpups, and between the water and milk

conditions. Furthermore, no strain by condition interaction was

found. Thus, the oral sensation of the milk may most-likely

account for the decreased USV (as in Refs. [5,43]).

The left panel of Fig. 3 shows the mean PLL measured from

these 11–12-days old infant rats immediately after the 5 min

isolation test. Mean PLL was significantly greater (>two-fold)

in OLETF (mean = 7.04) compared to LETO pups (mean = 3.23;

F(1,39) = 4.55, p < 0.05). No significant difference between the

milk and water conditions, or strain by condition interaction

was found.

2.3. Experiment 2

The right panel of Fig. 3 shows the mean hind-paw-lick latency

of adult rats. As shown, OLETF males (N = 24, mean = 11.6 s)

p e p t i d e s 2 7 ( 2 0 0 6 ) 1 4 8 3 – 1 4 8 9 1487

Fig. 3 – Mean latency to paw-lift in infant (left panel) and to

paw-lick in adult (right panel) OLETF and LETO rats, on the

hot-plate test.

exhibited significantly longer latencies to paw-lick compared

to LETO controls (N = 23, mean = 7.3), (t = 5.073, d.f. = 45,

p < 0.001). The OLETF rats in this experiment weighed

(mean = 513 g, SEM = 9.2) significantly more than the LETO

rats (mean = 433.5, SEM = 4.4), t = 7.85, d.f. = 45, p < 0.001).

2.4. Experiment 3

In both Experiments 3a and 3b, OLETF pups weighed

significantly more than LETO controls (means = 27.9 g versus

19.6 g, p < 0.001, in Experiment 3a; and 26.2 versus 21.6,

p < 0.05, in Experiment 3b, respectively). In Experiment 3a, the

control LETO group, showed a significant preference for the

rug texture (mean = 130.3 s, SEM = 9.07; one-sample t = 4.44,

d.f. = 13, p < 0.001). This replicates previous results in Sprague-

Dawley rats of the same age [36,38], extending them to Long

Evans rats. In the main comparison of interest, OLETF pups

showed virtually identical texture-preference levels

(mean = 131.6, SEM = 12.57; LETO–OLETF t-test: NS).

In Experiment 3b, the control, LETO group, showed a

significant preference for maternal-odor by both measures

(OPM1: mean = 63.7 s, SEM = 9.6; t(13) = 4.78, p < 0.001; OPM2:

mean = 36.9 s, SEM = 12.6; t(13) = 2.93, p < 0.05). In the main

comparison of interest, in contrast to our hypothesis, odor-

preference levels of OLETF pups were not significantly higher

than those of LETO controls on both measures; in fact their

preference levels were significantly lower than controls on one

of the measures (OPM1: mean = 32.5 s, SEM = 6.8; t(24) = 2.57,

p < 0.05; OPM2: mean = 7.73 s, SEM = 11.8).

3. Discussion

In general, the results show that OLETF, compared to LETO

rats, display a pattern of hyper-responsiveness to stress in

the USV measure, and hypo-responsiveness to thermal-

pain. Specifically, OLETF pups emitted much more USV than

controls, and as infants and as adults they exhibited slower

thermal-pain responses than controls. With regard to infant

natural preferences, OLETF pups showed preference levels

similar to those of LETO controls. Overall, this pattern of

results could be the consequence of one or more of the

following possibilities: the lack of a moderating CCK1

mediated effect, a compensatory super-sensitivity of CCK2

receptors, or compensatory changes in other, non-CCK

systems, that modulate stress, pain, and preference beha-

viors.

Regarding USV, the pattern of results confirms, in general,

the findings obtained with the pharmacological approach.

While exogenously administered CCK-8 reduced USV [42],

OLETF pups emitted significantly more USV than controls

(Fig. 1), just as pups treated with the selective antagonist of

CCK1 receptors, devazepide, produced more USV than controls

in two previous studies [4,43]. There are, though, a few

differences in the details. Devazepide has not been shown to

increase USV alone, but rather to block USV-decrease induced

by oral ingestion of milk or corn-oil. In contrast, OLETF rats

were shown here to produce more USV than controls, while

isolated in an empty cup, or in a plastic box containing a paper

towel, damp with water. However, when the towel contained

sweetened milk, the OLETF–LETO difference was not signifi-

cant. This finding is most interesting, because it may suggest

that in the absence of CCK1 receptors, OLETF pups may be

super-sensitized to other neurochemical systems that reduce

USV in response to milk (e.g., opioids; [3]). The possible

development of such adaptive, compensatory mechanisms

should be tested in future studies.

Recently, we have reported, in a study of siblings of the

current subjects, that OLETF pups separated from the dam and

then returned to a foster (OLETF or LETO) dam individually for

10 min, emitted significantly more USV than LETO pups who

underwent an identical procedure [19]. The current findings

extend those results, by showing higher USV levels in OLETF

pups, not only while interacting with a foster dam, but even

when alone, during an isolation test. Taken together, the two

studies strengthen the conclusion that rat pups lacking

functional CCK1 receptors vocalize more than controls.

These findings may be relevant to a study that found low

CCK levels in colicky (excessively crying) human infants. The

authors speculated that colicky infants have impaired CCK

secretion, which contributes to their gallbladder hypocon-

tractility (reported earlier) and excessive crying [15].

Regarding pain, the hypoalgesia on the hot-plate paw-

removal response found in OLETF rats (Fig. 3) was robust,

appearing in two replications in infant rats and in adults. This

finding corroborates a previous report of hypoalgesia in six

adult OLETF rats [22]. The two studies that are not in

accordance with this result reported no differences between

adult OLETF and LETO rats, assessed by a different pain assay

(tail-flick, 21) and no effect of IP administration of CCK-

octapeptide on PLL in infant Sprague-Dawley rats [42].

Although our current finding of hypoalgesia in OLETF rats

may be due to an alteration in the number or sensitivity of

CCK2 receptors, it is more plausible to suggest that CCK1

receptors normally play a hyperalgesic modulatory role, at

least in the context of the mildly stressful hot-plate test, and

p e p t i d e s 2 7 ( 2 0 0 6 ) 1 4 8 3 – 1 4 8 91488

therefore the absence of these receptors results in a relative

hypoalgesia in the OLETF rat.

Regarding natural preferences, as expected, control LETO

pups demonstrated relative texture- and maternal-odor pre-

ferences, thus replicating our results in Sprague-Dawley rats.

However, the preferences of OLETF rats did not differ, in marked

contrast to our previous results with CCK receptor antagonists.

Note that these ‘‘negative’’ results appear to be highly reliable.

First, they were conceptually replicated in two separate studies

of siblings from the same litters. Second, the experimenters

were blind to the pups’ strain. Third, while the groups did not

differ in their preference levels, significant OLETF–LETO

differences were indeed observed when comparing pups from

the same litters on other behavioral measures (USV and PLL,

Experiment 1, above). Fourth, from our experience [32,34] the

preference levels demonstrated in the current study were

moderate, allowing for further increases in preference levels.

This argues against the possibility of a ‘‘ceiling effect’’. Taken

together, the results show that OLETF pups preferred the test

stimuli to a similar degree as LETO controls. They did not show a

greater preference (as was the case in Sprague-Dawley rats

treated with CCK-receptor antagonists). It is possible that the

neurochemical systems relevant for these preferences have

undergone compensatory processes in OLETF rats during the

prenatal and postnatal periods preceding our experiment. We

note that the previous findings in which selective antagonists of

both 1- and 2-type receptors produced similar effects did not

allow for a clear prediction for the pattern of preference to be

expected in OLETF pups.

Experiments 1 and 3 were performed in infants, before the

age in which OLETF rats become diabetic and obese [17,24]. To

verify this issue, we examined glucose levels from a different

study, in progress. No significant OLETF–LETO differences

were found on postnatal day 7 (overall mean glucose

levels = 101.9 mg/dl, N = 21) and on postnatal day 15

(123.8 mg/dl, N = 17). A potential limitation of this study is

the lack of body temperature measurement. USV are affected

by ambient temperature [18], which was measured and was

stable during Experiment 1. Nevertheless, it is conceivable that

a portion of the USV differences reported above, obtained at

standard room temperature, could be explained by reactivity

to differential ambient-body temperature differentials, if

OLETF and LETO pups differed in body temperature. To

examine this we measured, in a follow-up study, axillary

temperatures of 25 OLETF and 20 LETO pups from 11 litters

at the ages of PND 10–12. After 2 h group housing in an

incubator, temperature was measured. They were then placed

individually in a clean cage with bedding at room temperature

(24–25 8C) for 10–15 min, and then temperature was measured

again. Two-way ANOVA showed that there was no significant

temperature difference between the strains (mean LETO =

35.6 8C, OLETF = 36.2 8C). Temperature decreased overall sig-

nificantly over time (average from 36.4 to 35.4 8C), with no

interaction between strain and time. Thus, the concern

regarding body temperature differences is not supported.

One final limitation of this study should be mentioned. The

adult OLETF rat has been characterized as lacking CCK1

receptors. OLETF rats did not differ from controls (when

hyperphagia and increased body weight were controlled by

pair-feeding) in hypothalamic POMC and NPY in the arcuate

nucleus and in the leptin receptor (long form, Ob–Rb).

However, they are characterized by overexpression of NPY

in the dorsal medial hypothalamus [2]. Nevertheless, this

model derives from selective breeding, not single-gene-

mutation technology. Therefore, it is still possible that some

other, as yet not identified genes may also be malfunctioning

in the OLETF rat, a gene that may affect stress and pain

reactivity by a non-CCK pathway, ultimately accounting for

our current results by an alternative mechanism.

Acknowledgements

The rats were a generous gift of Dr. Kawano of Tokushima

Research Institute and Otsuka Pharmaceutical, Tokushima,

Japan. The authors thank Ofra Schwartz for animal care and

careful maintenance of ‘‘blind’’ coding of the litters’ geno-

types. The authors thank Adi David, Itay Peleg, Danielle

Schwartz and Daniel Markovitz for help in data collection. OM

and YLA were supported by the President’s doctoral fellow-

ship, Bar-Ilan University. OM was supported beforehand by a

fellowship from Bar-Ilan University’s Interdisciplinary Studies

Committee. This work was supported by a grant from the

Israel Science Foundation to AW.

r e f e r e n c e s

[1] Barbaz BS, Hall NR, Liebman JM. Antagonism of morphineanalgesia by CCK-8-S does not extend to all assays nor allopiate analgesics. Peptides 1989;9:1295–300.

[2] Bi S, Ladenheim EE, Schwartz GJ, Moran TH. A role for NPYoverexpression in the dorsomedial hypothalamus inhyperphagia and obesity of OLETF rats. Am J Physiol2001;281:R254–60.

[3] Blass EM, Fitzgerald E. Milk-induced analgesia andcomforting in 10-days-old rats: opioid mediation.Pharmacol Biochem Behav 1988;29:9–13.

[4] Blass EM, Shide DJ. Endogenous cholecystokinin reducesvocalization in isolated 10-days-old rats. Behav Neurosci1993;107:488–92.

[5] Blumberg S, Haba D, Schroeder M, Smith GP, Weller A.Independent ingestion and microstructure of feedingpatterns in infant rats lacking CCK1 receptors. Am J Physiol2005, in press.

[6] Brunelli SAB, Hofer MA. Selective breeding for an infantilephenotype (isolation calling): a window on developmentalforces. In: Blass EM, editor. Handbook of behavioralneurobiology, vol. 13. New York: Plenum Press; 2001. p. 433–

82.

[7] Crawley JN, Corwin RL. Biological actions ofcholecystokinin. Peptides 1994;15:731–5.

[8] Dauge V, Bohme GA, Crawley JN, Durieux C, Stutzmann JM,Feger J, et al. Investigation of behavioral andelectrophysiological responses induced by selectivestimulation of CCK-B receptor by using a new highly potentCCK analog, BC264. Synapse 1990;6:73–80.

[9] Dockray GJ. Immunochemical evidence ofchoolecystokinin-like peptides in brain. Nature1976;264:568–70.

[10] Dourish CT, Hawley D, Iversen SD. Enhancement ofmorphine analgesia and prevention of morphine tolerancein the rat by the cholecystokinin antagonist L-364,718. Eur JPharmacol 1988;147:469–72.

p e p t i d e s 2 7 ( 2 0 0 6 ) 1 4 8 3 – 1 4 8 9 1489

[11] Funakoshi A, Miyassaka K, Shinozaki H, Masuda M,Kawanami T, Takata Y, et al. An animal model ofcongenital defect of gene expression of cholecystokinin(CCK)-A receptor. Biochem Biophys Res Commun1995;210:787–96.

[12] Harhammer R, Schafer U, Henklein P, Ott T, Repke H.CCK-8-related C-terminal tetrapeptides: affinities forcentral CCKB and peripheral CCKA receptors. Eur JPharmacol 1991;209:263–6.

[13] Hofer MA. Multiple regulators of ultrasonic vocalization inthe infant rat. Psychoneuroendocrinology 1996;21:203–17.

[14] Hokfelt T, Holmberg K, Shi TJ, Broberger C. CCK-ergicmechanisms in sensory systems. Scand J Clin Lab InvestSuppl 2001;234:69–74.

[15] Huhtala V, Lehtonen L, Uvnas-Moberg K, Korvenranta H.Low plasma cholecystokinin levels in colicky infants. JPediatr Gastroenterol Nutr 2003;37:42–6.

[16] Insel TR, Winslow JT. Rat pup ultrasonic vocalizations: anethologically relevant behavior responsive to anxiolytics.In: Olivier B, Mos J, Slangen JL, editors. Animal modelsin psychopharmacology. Basel: Birkhauser-Verlag; 1991. p.15–36.

[17] Kawano KT, Hirashima S, Mori S, Saitoh Y, Kurosumi M,Natori T. Spontaneous long-term hyperglycemic rat withdiabetic complications: Otsuka Long Evans TokushimaFatty (OLETF) strain. Diabetes 1992;41:1422–8.

[18] Kraebel KS, Brasser SM, Campbell JO, Spear LP, Spear NE.Developmental differences in temporal patterns andpotentiation of isolation-induced ultrasonic vocalizations:Influence of temperature variables. Dev Psychobiol2002;40:147–59.

[19] Lavi-Avnon Y, Malkesman O, Hurwitz I, Weller A. Mother–infant interactions in rats lacking CCKA receptors. BehavNeurosci 2004;118:282–9.

[20] Lavigne GJ, Millington WR, Mueller GP. The CCK-A andCCK-B receptor antagonists, devazepide and L-365,260,enhance morphine antinociception only in non-acclimatedrats exposed to a novel environment. Neuropeptides1992;21:119–29.

[21] Lee GS, Han JB, Shin MK, Hong MC, Kim SW, Min BI, et al.Enhancement of electroacupuncture-induced analgesiceffect in cholecystokinin-A receptor deficient rats. BrainRes Bull 2003;62:161–4.

[22] Li X-L, Aou S, Hori T, Oomura Y. Spatial memory deficit andemotional abnormality in OLETF rats. Physiol Behav2002;75:15–23.

[23] Miyasaka K, Kobayashi S, Ohta M, Kanai S, Yoshida Y,Nagata A, et al. Anxiety-related behaviors incholecystokinin-A, B, and AB receptor gene knockout micein the plus-maze. Neurosci Lett 2002;335:115–8.

[24] Moran TH, Katz LF, Plata-Salaman CR, Schwartz GJ.Disordered food intake and obesity in rats lackingcholecystokinin A receptors. Am J Physiol 1998;274:R618–25.

[25] Moran TH, Robinson PH, Goldrich MS, McHugh PR. Twobrain cholecystokinin receptors: implications forbehavioral actions. Brain Res 1986;362:175–9.

[26] Nakamura H, Kihara Y, Tashiro M, Kanagawa K, ShiroharaH, Yamamoto M, et al. Defects of cholecystokinin (CCK)-Areceptor gene expression and CCK-A receptor-mediatedbiological functions in Otsuka Long Evans Tokushima Fatty(OLETF) rats. J Gastroenterol 1998;33:702–9.

[27] Noble F, Roques BP. The role of CCK2 receptors in thehomeostasis of the opioid system. Drugs Today (Barc)2003;39:897–908.

[28] Nomoto S, Miyake M, Ohta M, Funakoshi A, Miyasaka K.Impaired learning and memory in OLETF rats withoutcholecystokinin (CCK)-A receptor. Physiol Behav1999;66:869–72.

[29] Olivier B, Molewijk HE, van der Heyden JAM, van OorschotR, Mos J, Miczek KA. Ultrasonic vocalizations in rat pups:Effects of serotonergic ligands. Neurosci Biobehav Rev1998;23:215–27.

[30] Rastegar H, Homayoun H, Afifi M, Rezayat M, Dehpour AR.Modulation of cholestasis-induced antinociception by CCKreceptor agonists and antagonists. EurNeuropsychopharmacol 2002;12:111–8.

[31] Rex A, Barth T, Voigt J-P, Domeney AM, Fink H. Effects ofcholecystokinin tetrapeptide and sulfated cholecystokininoctapeptide in rat models of anxiety. Neurosci Lett1994;172:139–42.

[32] Rakover-Atar S, Weller A. The influence of naturalpreference for tactile stimuli on appetitive learning in ratpups. Dev Psychobiol 1997;30:29–39.

[33] Sei M, Sei H, Shima K. Spontaneous activity, sleep, andbody temperature in rats lacking the CCK-A receptor.Physiol Behav 1999;68:25–9.

[34] Shayit M, Weller A. Cholecystokinin receptor antagonistsincrease the rat pup’s preference towards maternal-odorand rug texture. Dev Psychobiol 2001;38:164–73.

[35] Suh HW, Kim YH, Choi YS, Song DK. Involvement ofdifferent subtypes of cholecystokinin receptors in opioidantinociception in the mouse. Peptides 1995;16:1229–34.

[36] Takiguchi S, Takata Y, Funakoshi A, Miyasaka K, Kataoka K,Fujimura Y, et al. Disrupted cholecystokinin type-Areceptor (CCKAR) gene in OLETF rats. Gene 1997;197:169–75.

[37] Tortorici V, Nogueira L, Salas R, Vanegas H. Involvement oflocal cholecystokinin in the tolerance induced by morphinemicroinjections into the periaqueductal gray of rats. Pain2003;102:9–16.

[38] Wank SA. G protein-coupled receptors in gastrointestinalphysiology I. CCK receptors: an exemplary family. Am JPhysiol 1998;274:G607–13.

[39] Wank SA, Pisegna JR, De Weerth A. Brain andgastrointestinal cholecystokinin receptor family: structureand functional expression. Proc Natl Acad Sci USA1992;89:3125–9.

[40] Watkins LR, Kinscheneck IB, Kaufman EFC, Miller J, FrenkH, Mayer DJ. Cholecystokinin antagonists selectivelypotentiate analgesia induced by endogenous opiates. BrainRes 1985;327:181–90.

[41] Weller A. The ontogeny of motivation: Hedonic preferencesand their biological basis in developing rats. In: Blass EM,editor. Handbook of Behavioral Neurobiology, vol. 13. NewYork: Plenum Press; 2001. p. 483–516.

[42] Weller A, Blass EM. Behavioral evidence forcholecystokinin–opiate interactions in neonatal rats. Am JPhysiol 1988;255:R901–7.

[43] Weller A, Gispan IH. A cholecystokinin receptor antagonistblocks milk-induced but not maternal-contact-induceddecrease in ultrasonic vocalization in rat pups. DevPsychobiol 2000;37:35–43.

[44] Weller A, Tsitolovskya L, Gispan IH, Rabinowitz S.Examining the role of cholecystokinin in appetitivelearning in the infant rat. Peptides 2001;22:1317–23.

[45] Wolf G, Yirmiya R, Goshen I, Iverfeldt K, Holmlund L,Takeda K, et al. Impairment of interleukin-1 (IL-1) signalingreduces basal pain sensitivity in mice: genetic,pharmacological and developmental aspects. Pain2003;104:471–80.

[46] Xie JY, Herman DS, Stiller CO, Gardell LR, Ossipov MH, Lai J,et al. Cholecystokinin in the rostral ventromedial medullamediates opioid-induced hyperalgesia and antinociceptivetolerance. J Neurosci 2005;25:409–16.

[47] Yamamoto Y, Akiyoshi J, Kiyota A, Katsuragi S, Tsutsumi T,Isogawa K, et al. Increased anxiety behavior in OLETF ratswithout cholecystokinin-A receptor. Brain Res Bull2000;53:789–92.