Embed Size (px)
Citation preview
2011
David Barker
ALL RIGHTS RESERVED
AFFECT AND ADDICTION: EXAMINING THE ROLE OF AFFECT IN DRUG
SEEKING BEHAVIOR USING ULTRASONIC VOCALIZATIONS IN THE RAT
By
DAVID J. BARKER
A Thesis submitted to the Graduate School-New Brunswick
Rutgers, the State University of New Jersey
In partial fulfillment of the requirements
For the degree of
Master of Science
Graduate Program in Psychology
Written under the direction of
Mark O. West Ph.D.
And approved by
_______________________________________
_______________________________________
_______________________________________
New Brunswick, New Jersey
October 2011
ii
ABSTRACT OF THESIS
AFFECT AND ADDICTION: EXAMINING THE ROLE OF AFFECT IN DRUG
SEEKING BEHAVIOR USING ULTRASONIC VOCALIZATIONS IN THE RAT
By DAVID J. BARKER
Thesis Director:
Professor Mark O. West Ph.D.
Affective processing has a purported role in all aspects of addiction from initiation of
drug use to relapse behavior. While roles for both negative and positive affect have been
reported in humans, animal models of addiction have historically lacked an analogous
measure. Recently, however, it has been proposed that the ultrasonic vocalizations
(USVs) of rodents can provide valuable insight into an animal’s affective state. Rats emit
USVs in two ranges, one from 18-38 kHz that has been correlated with aversive
outcomes, and another from 38-80 kHz that has been linked to appetitive processing. In
order to ascertain the role of affect in an animal model of cocaine addiction, the USVs of
rats were measured across two experiments. First, USVs were recorded from animals in
both a high dose and low dose cocaine self-administration paradigm. The results of this
experiment suggested that animals in the high dose condition emitted significantly more
50-kHz USVs, whereas subjects in the low dose group emitted significantly greater
numbers of 22-kHz USVs. Given the observed dose-dependent, crossover interaction of
22- and 50-kHz USVs, it was predicted that these USVs might be linked to sub-satiety
iii
and circa-satiety drug levels, respectively. In order to explicitly test this hypothesis,
animals were recorded in two clamp conditions during Experiment 2: a High Clamp and a
Low Clamp. During the High Clamp condition, drug levels were fixed at a constant
concentration, just above subjects’ self-determined satiety threshold, for four hours.
During the Low Clamp, subjects’ drug levels were fixed at a constant, sub-satiety
concentration defined as fifty percent of the satiety threshold. The results of this
experiment indicated that 50-kHz USVs were tied to the drug loading period early in the
session, whereas 22-kHz calls were produced when subjects’ drug levels were maintained
at sub-satiety levels. Overall, given their correlation with negative and positive affective
states, the observed results suggest that a duality of affective processing exists during
cocaine self-administration. In fact, the presence of both positive and negative affective
states might suggest that the perseveration of drug-seeking in addicted subjects is due, in
large part, to the additive effects of both negative and positive reinforcement.
iv
Acknowledgments
I would like to that Dr. Mark West for his guidance while conducting this
research as well as his assistance and mentorship during all aspects of my graduate
training. I would also like to thank the members of my thesis committee, Dr. Alex
Kusnecov and Dr. Louis Matzel for their constructive comments on the thesis. Finally, I
would like to thank my peers, both graduate and undergraduate, as well as Jackie
Thomas, Thomas Grace, and Linda King for their excellent assistance. To conclude, it
should be disclosed that Experiment 1 of the present data had been published prior to its
inclusion in the present thesis (Barker et al., 2010). Some Excerpts and Figures have
been reprinted with permission from Springer Science and Business Media.
v
TABLE OF CONTENTS:
Abstract….…………………………………………………………………………….......ii
Acknowledgments……………………………………………………………………......iv
Table of Contents……….……………………………………………………………...….v
List of Tables………………………………………………………………………...….viii
List of Figures……………………………………………………………….………..…..ix
1. Introduction……..…………………………………..…………………………………..1
1.1 Cocaine Addiction……………………………………………….……………1
1.2 Neural Substrates of Cocaine Addiction….…………………………………...2
1.3 Ultrasonic Vocalizations………………………………………………………3
1.4 Neural Substrates of Ultrasonic Vocalizations…………….………………….5
1.5 Addiction, Affect, and USVs………………………………………………….7
2. General Materials and
Methods…………………………………….…………………..10
2.1 Subjects and
Surgery……………..………………………….…..……………10
2.2 USV recordings………………………………………………………………11
2.3 Characterizations of USVs…………………………………………………...11
2.4 Drug Level Calculations……………………………………………………..12
2.5 Experimental Apparatus…………………………………………...…………13
3. Experiment 1 Materials and Methods…………………………………….….………..15
3.1 Shaping………………………………………………………………………15
3.2 Self-Administration…………………………………………………………..16
vi
3.3 Statistical Analysis…………………………………………………..……….17
4. Experiment 1 Results and Discussion………………………………………………....18
4.1 Self-Administration………………………………………………………….18
4.2 Ultrasonic Vocalizations……………………………………………………..19
4.3 Discussion……………………………………………………………………20
5. Experiment 2 Materials and Methods………………………………………………....22
5.1 Baseline Recordings………………………………………………………….22
5.2 Shaping…….……………………………………………………………...…23
5.3 Self-Administration…………………………………………………………..23
5.4 Clamp Test and Recordings………………………………………….………24
5.5 Statistical Analysis…………………………………………………………...25
6. Experiment 2 Results and Discussion…………………………………………………27
6.1 Behavior…………………………………………………………………...…27
6.2 Ultrasonic Vocalizations (Time-Analysis)…………………………………..28
6.2.1 Omnibus Results…………………………………………………...28
6.2.2 Condition x Call Frequency x Session Time: Effect of Time…...…28
6.2.3 Condition x Call Frequency x Session Time: Effect of Frequency..29
6.2.4 Condition x Call Frequency x Session Time: Effect of Condition...30
6.3 Ultrasonic Vocalizations (Call-Type Analysis)……………………………...32
6.3.1 Omnibus Results…………………………………………………...32
6.4 Hour 1 Analysis………………………………………………...……………33
6.5 Discussion……………………………………………………………………35
7. General Discussion…………………………………………………………………....40
vii
8. References……………………………………………………………………………..45
9. Appendices…………………………………………………………………………….49
9.1 Call Type Analysis Post-Hoc Tests…………………………….……………49
9.1.1 Main Effect for Call Type………………………………………….49
9.1.2 Condition x Call Frequency: Within Condition Comparisons……..49
9.1.3 Condition x Call Frequency: Between Condition Comparisons…...50
10. Tables and Figures…………………………………………………………………...51
viii
List of tables
Table 1: Condition x Call Frequency x Call Type: Mean ± SEM………………………..51
Table 2: Condition x Call Frequency x Session Time: Mean ± SEM ………….…..…....52
ix
List of Illustrations
Figure 1: Examples of Ultrasonic Vocalizations (USVs) during Cocaine S-A…………..53
Figure 2: Expt. 1—Cumulative Daily Drug Intake for High and Low Dose Animals…...54
Figure 3: Expt. 1—Frequency by Duration Scatter Plot of All Calls……….…………...55
Figure 4: Expt. 1—Dose-Dependent Crossover Interaction of USVs………………..……..56
Figure 5: Expt. 2—Plexiglas Tube Used for Improving Recording Quality……………..57
Figure 6: Expt. 2—Average Number of Daily Lever Presses During Training.…............58
Figure 7: Expt. 2—Average Drug Consumption (Mg/Kg) Across Training………….….59
Figure 8: Expt. 2—Average Number of Infusions Earned Daily…….…………..….........60
Figure 9: Expt. 2—Percent of Opportunities where an Infusion was Earned…….……..61
Figure 10: Expt. 2—Subjects’ Bodyweight Across Sessions……..............................…....62
Figure 11: Expt. 2—Example Drug Curve for High and Low Clamp Sessions....……….63
Figure 12: Expt. 2—Scatter Plots of All USVs for Each Experimental Condition…........64
Figure 13: Expt. 2—Condition x Call Frequency x Call Type Plot………...…...……….65
Figure 14: Expt. 2—Condition x Call Frequency x Session Time Interaction....….….....66
Figure 15: Expt. 2—USVs in the Pre and Post Drug Period of Hour 1………..….….....67
1
1. Introduction:
1.1—Cocaine Addiction:
Cocaine is a psychomotor stimulant derived from the coca leaf of the Erythroxylon
coca shrub from the mountains of South America. Although the plant has been chewed
for upwards of 1200 years, cocaine in its refined, powder form has only been available
since the mid-to-late 1800’s. Following a mixed reception, cocaine was soon touted by a
number of physicians—including Sigmund Freud—for its therapeutic values. Amongst
the most beneficial of these discoveries were the local anesthetic properties of the drug
when used during surgical preparations. Still, not long after its inception, cocaine was
quickly deplored for its dependence causing qualities, and by 1970 cocaine was made
illegal in the United States altogether (Gay et al 1975; NIDA 2009).
The pharmacological effects of cocaine produce a number of physical and
psychological signs. The physical signs include sympathetic effects such as increased
heart rate and respiration, vascular constriction, pupil dilation, sweating, increased blood
pressure, and decreased appetite (NIDA 1985; Levinthal 2010 pp 90-115). The
psychological effects of the drug include euphoria, alertness, and increased confidence.
These psychological effects are often short-lived and quickly followed by opposing
effects that include dysphoria, irritability, paranoia, insomnia, and depression (Mendelson
& Mello 1996; Williamson et al 1997). The acute transition from positive to negative
psychological symptoms is thought to account for cocaine “binges”, wherein users
typically administer multiple doses with short inter-administration intervals in order to
avoid negative symptoms until their drug supply is exhausted.
2
In the present day, more than 1.5 million people each year meet the criteria for
substance abuse or dependence (Office of Applied Studies 2009). Drug dependence
occurs when a drug is used repeatedly for non-medical purposes in such a manner that a
strong psychological or physiological dependence develops. The key characteristics of
dependence are (1) perseverative drug seeking despite health problems, legal problems,
or interference with personal obligations (2) escalation of drug intake despite attempts to
control drug use, and (3) negative physical or psychological states that coincide with the
cessation of drug use. Individuals who are substance dependent are subject to chronic
episodes of drug relapse (American Psychiatric Association 1994; Koob 2008, 2009,
2009b; Koob & LeMoal 2008).
Although the term “drug dependence” accurately portrays the physiological
adaptations that occur following repeated drug exposure, the term “addiction” will be
used henceforth in order to describe both the physiological effects, as well as the entire
spectrum of drug-seeking and drug-taking behaviors, which may or may not be described
by the use of the term “drug dependence”. Addiction, as defined here, constitutes a
disorder characterized by both longitudinal physiological changes and persistent
behaviors. While the DSM-IV criteria for drug dependence (summarized above)
accurately describe many of the facets of addiction, the term drug dependence inherently
misrepresents the behavioral aspects of the disorder by placing emphasis on the
physiological changes (i.e. the development of a chemical dependence).
1.2—Neural Substrates of Cocaine Addiction:
Cocaine’s reinforcing effects depend on dopamine release within the
mesocorticolimbic system (Ritz et al. 1987; Koob et al 1994). During normal functioning,
3
dopamine is released by neurons projecting from the VTA to the limbic forebrain, where
dopamine binds to receptors on the post-synaptic neuron and is then recycled during the
re-uptake process. Cocaine acts to block the dopamine transporter and consequently, the
re-uptake of dopamine into the pre-synaptic cell, thereby leading to increased levels of
dopamine in the synaptic cleft (NIDA 2009).
The amount of striatal re-uptake blockade, as measured by binding of the dopamine
transporter (DAT), positively correlates with the subjective “high” addicts feel, although
this feeling dissipates faster than DAT occupancy (Volkow et al. 1997). It has been
proposed that, following extended drug use, this modulation of normal dopaminergic
functioning produces lasting changes at the synapse that might account for the transition
between casual drug use and addiction (Kalivas & Volkow 2005).
1.3—Ultrasonic Vocalizations
Rats, much like other mammals, are capable of intraspecies communication. More
specifically, rats are able to produce laryngeal sounds of two distinct types. The first of
these types falls within humans’ audible range (0-18 kHz) and is produced through slow
vibrations of the vocal folds. Sonic vocalizations in the wild are emitted most often in
circa-strike situations (defined as situations in which predatory strike is imminent or in
progress; Fanselow & Lester 1988), but can be elicited by inducing pain or handling
naïve subjects in the laboratory (Brudzynski 2008; Litvin et al. 2007)
The second type of laryngeal sound produced involves constriction and stabilization
of the larynx, such that a small orifice is produced. When air is forced through this
orifice, a whistle-like sound—above the human hearing range (Figure 1; 18-80 kHz)—is
4
produced. These high-frequency emissions are referred to as ultrasonic vocalizations
(USVs; Brudzynski 2008; Johnson et al 2010).
Ultrasonic vocalizations can be further divided into three types: fixed frequency calls
(Figure 1A & 1B; FF), frequency modulated calls (Figure 1C; FM), or trills (Figure 1D).
These sub-types are differentiated based upon the presence and number of pitch
modulations during a single call (Ahrens et al. 2009, Wright et al. 2010). FF calls
maintain a stable pitch throughout the entirety of the call, whereas the frequency of FM
calls and trills modulates throughout the call. More specifically, FM calls are
characterized by a single frequency modulation, most often described as a “sweep” or
“jump/step” from one pitch to another, while USVs are characterized as trills when
multiple frequency modulations occur within a single call. These pitch changes can occur
as sweeps, jumps, or some combination of the two (Wright et al. 2010).
Ultrasonic vocalizations in adult rats can be divided into two frequency ranges. Calls
in the first range span approximately 18-33 kHz, and are referred to as “22-kHz” USVs
(Figure 1A & 1C), while calls in the second range span 38-80 kHz, and are referred to as
“50-kHz” USVs (Figure 1B and 1D). Most commonly, 22-kHz call durations range
between 300-3400 ms (here termed “long 22-kHz calls”), although a subset of shorter 22-
kHz calls ranging between 20-300 ms (termed “short 22-kHz calls”) has been described
(Brudzynski et al. 1993; Brudzynski et al. 1991; Barker et al. 2010). On the other hand,
USVs in the 50-kHz range are almost always short, lasting between 20-80 ms (Knutson et
al. 2002).
Long 22-kHz calls occur during aversive events such as social isolation (Francis
1977), predatory odors (Blanchard et al. 1991), or foot-shock (Tonoue et al. 1986).
5
Given the circumstances under which they are emitted, it has been suggested that long
22-kHz calls act as warning or alarm cries for conspecifics (Litvin et al. 2007).
Furthermore, it has been shown that anxiolytic drugs can reduce the production of long
22-kHz USVs, while certain anxiogenic drugs can increase their production (Jelen et al.,
2003). Short 22-kHz USVs have been less studied, but have been shown in concordance
with aversive events as well. For example, short 22-kHz calls have been produced by the
formalin footpad test (Oliveira et al., 2006), handling naïve rats, foot-shock (Brudzynski
et al 1991; Brudzynski et al 1993), or the injection of muscarinic agonists into the
anterior hypothalamus (Brudzynski et al 1991; Brudzynski et al 1993).
Fifty-kHz USVs have been linked with anticipatory and appetitive circumstances
such as social contact (Knutson et al. 1998; Brudzynski & Pniak 2002), amphetamine
conditioned place preference (Knutson et al. 1999), or sexual behaviors (Barfield et al.
1986). Given that 22 kHz USVs occur during negative situational outcomes while 50 kHz
USVs occur during positive situational outcomes, it is generally accepted that USVs
provide insight into the affective state of animals (Knutson et al. 2002) or the semiotic
value of a given situation (Brudzynski 2005).
1.4—Neural Substrates of Ultrasonic Vocalizations
A small number of previous studies have provided limited insight into the neural
systems underlying USV production. Intracerebral glutamate infused into the anterior
hypothalamic-preoptic area has been shown to increase 50-kHz calls in a dose-dependent
manner, as well as to induce USVs in animals that failed to vocalize at lower doses (Fu &
Brudzynski 1994). Interestingly, the dose-dependent increase induced by intracerebral
glutamate produced a ceiling effect on the rate of USVs when administered in
6
combination with systemic amphetamine (which has also been shown to increase 50-kHz
USVs when administered alone; Ahrens et al. 2009). Furthermore, the effects of
glutamate were later attenuated following systemic administration of the dopamine (DA)
antagonist, haloperidol (Wintink & Brudzynski 2001). These data suggest that dopamine
as well as glutamatergic-dopaminergic interactions in the limbic forebrain are important
for the production of 50-kHz USVs.
The involvement of mesolimbic dopamine in the production of 50-kHz USVs has
been further evidenced by Burgdorf and colleagues (2000), who have demonstrated that
the rate of 50-kHz USVs increases when rats anticipate the delivery of electrical brain
stimulation to the ventral tegmental area (VTA)—the main source of mesolimbic
dopamine. This result extends behavioral findings, which have shown that the
anticipation of natural rewards (social interaction, food, etc.) also increases the number of
50-kHz USVs (Knutson et al. 1998). Importantly, these same types of behaviors have
been shown to up-regulate extracellular levels of mesolimbic dopamine (Pfaus et al 1990;
Taber and Fibiger 1997).
While few studies have focused on the neurochemical or neuroanatomical substrates
of 22-kHz USVs, it has been shown that both long and short 22-kHz calls can be induced
by cholinergic stimulation of the anterior hypothalamus (Brudzynski et al 1991;
Brudzynski et al 1993) and surrounding nuclei (which produces irritable and aggressive
behavior in mammals; Brudzynski et al 1995). These cholinergic regions are thought to
be part of a larger “medial cholinoceptive vocalization strip”, for which the cholinergic
afferents originate in the laterodorsal tegmental nucleus (LDT). This hypothesis has been
7
corroborated by the demonstration that glutamatergic stimulation of the LDT will also
stimulate the production of 22-kHz USVs (Brudzynski, 2001)
1.5—Addiction, Affect and USVs
It is often assumed that, during intoxication, cocaine acts solely as a positive
reinforcer (e.g. Bijlsma et al 2010), and that the negative states associated with cocaine
use occur during the withdrawal period that follows (Koob 2008, 2009, 2009b; Covington
& Miczek 2003; Barros & Miczek 1996; Mutschler & Miczek 1998a; Mutschler &
Miczek 1998b). Still, human subjects report that the cocaine high dissipates quickly
(approx. 30-40 minutes; Volkow et al 1997) and it is thought that the acute onset of
negative symptoms that occurs as blood levels of cocaine fall may lead to the typical
“binge” usage of cocaine (Mendelson & Mello 1996).
Furthermore, although negative mood states have been reported as one of the most
common reasons for relapse (especially major relapses; Hodgins et al 1995), emotional
distress, depressed mood, and life dissatisfaction are also major risk factors for the
initiation of drug use (Newcomb & Felix-Ortiz 1992). These results suggest that
emotional dysregulation may play a role in all aspects of stimulant addiction (i.e.
initiation, bingeing, withdrawal and relapse), and that susceptibility to emotional distress
may play a potent role in the transition from casual drug use to drug addiction.
Negative reinforcement has long been considered a possible driving factor for drug
use (Wikler 1948). For example, one hypothesis of returned drug use posits that negative
affect is universal to all drugs of abuse, and that the prepotent driving force behind self-
administration is negative reinforcement (Baker et al 2004). Recent evidence has
supported this hypothesis. That is, taste cues paired with the opportunity to self-
8
administer cocaine in rats selectively produced "aversive" orofacial reactions that
predicted self-administration behavior (Wheeler et al 2008). Moreover, cocaine users
report difficulties regulating their emotions (Fox et al 2007) and retrospectively report a
significant increase in "unpleasant affect" just prior to drug use (McKay et al 1995).
This evidence suggests that cocaine should be considered for its “reinforcing”
properties rather than its “rewarding” effects. That is, cocaine may act as a reward or
incentive (Kelley & Berridge 2002; Robinson & Berridge 1993; Robinson & Berridge
2001), but may also serve as a mechanism through which addicted organisms escape (1)
falling drug levels, (2) acute or extended withdrawal, and (3) inherent neurochemical
imbalances (Mendelson & Mello 1996; Koob & LeMoal 1997; Koob 2008, 2009, 2009b;
Newcomb & Felix-Ortiz 1992).
In recent years a number of laboratories have utilized ultrasonic vocalizations as a
tool for understanding the affective changes induced by drugs of abuse. Studies of rats’
USVs during acute, systemic cocaine withdrawal have demonstrated increased rates of
long 22-kHz USVs in response to a diffuse air-puff relative to controls for both oral
(Barros & Miczek 1996) and intravenous cocaine (Mutschler & Miczek 1998a) as well as
opiates (Covington & Miczek 2003). This increase in aversive long 22 kHz USVs is
exacerbated when subjects endure withdrawal from non-contingent infusions of cocaine,
as compared to controls that had self-administered (Mutschler & Miczek 1998b).
Recent evidence has also indicated that, much like motor activity, 50-kHz USVs can
be sensitized following repeated drug administration (Ahrens et al. 2009). This finding
and others suggest that these calls are intimately tied with the neural circuitry affected by
long-term administration of psychostimulants (Mu et al. 2009). For example, intracranial
9
injections of amphetamine into the nucleus accumbens (NAcc) core and shell caused
increases in the rate of 50-kHz USVs compared to injections localized in the dorsal
striatum, with the most robust increases occurring in response to injections localized to
the NAcc shell (Burgdorf et al. 2001). These results were confirmed by a more recent
study, which noted increases in 50-kHz USVs following amphetamine injections to the
NAcc shell as well as the ventral core (Thompson et al. 2006).
Although many of these studies provide insight into subjects’ affective reactions
during anticipation of drug reward, experimenter administration of stimulants, or
stimulant withdrawal, little emphasis has been placed on understanding the affective
mechanisms that drive the cocaine binge or a subject’s USVs during cocaine self-
administration. In fact, Experiment 1 of the present thesis (Barker et al. 2010) is the first
report, to our knowledge, on USVs during drug self-administration. The present
experiments therefore examine affective changes during a 6-hour cocaine self-
administration binge.
10
2. General Materials and Methods:
2.1—Subjects and Surgery:
Male Long-Evans rats from multiple litters (Charles River, Wilmington, MA)
were singly housed on a 12h: 12h light: dark cycle with dawn at either 1130 h (Expt.1) or
1030 h (Expt. 2). Prior to surgery, subjects were given sufficient food to maintain a pre-
operative weight of 325-335 g. Once subjects reached a stable weight, they were
anesthetized with sodium pentobarbital (50 mg/kg I.P.) and given 10 mg/kg of atropine
methonitrate (I.P.) as well as a 0.25 mg I.M. dose of penicillin (300,000 U/ml) to
maintain respiratory function and prevent infection. Anesthesia was then maintained
throughout the remainder of the surgery using intermittent doses of ketamine
hydrochloride (60 mg/kg I.P.) and sodium pentobarbital (5-10 mg/kg I.P.). During
surgery, animals were implanted with an indwelling catheter in the right jugular vein,
which exited through a j-shaped stainless steel cannula anchored to the skull using dental
cement and three jeweler’s screws. Following surgery, animals were housed in the self-
administration chamber for the remainder of the experiment. All animals were allowed to
recover from surgery for at least 7 days prior to the start of training. While living in this
chamber, a syringe pump delivered a 200 µL infusion of heparinized-saline to subjects
every twenty-five minutes around the clock, as controlled by either a mechanical timer
(Expt. 1) or computer program (Expt. 2), in order to preserve catheter patency. On days
in which animals self-administered cocaine, these infusions were delivered during the 18-
hours in which the experimental contingencies were not in effect. All Protocols were
performed in compliance with the Guide for the Care and Use of Laboratory Animals and
11
have been approved by the Institutional Animal Care and Use Committee, Rutgers
University.
2.2—USV Recordings:
A condenser microphone (CM16/CMPA, Avisoft Bioacoustics, Berlin Germany)
was inserted into the sound-attenuating chamber 15-30 minutes before the
commencement of the self-administration session. The microphone was suspended ~2.5
cm from a set of small holes in the top of the self-administration chamber above the
quadrant proximal to the response manipulandum. Following the start of the session,
recordings were triggered by a 5V TTL pulse sent from a computer program to the
recording hardware (Ultrasound Gate 116H, Avisoft Bioacoustics, Berlin Germany).
Sonorous activity was recorded at either a 192-kHz (Expt. 1) or 250 kHz (Expt. 2)
sampling frequency (16-bits) using Avisoft Recorder software (Avisoft Bioacoustics,
Berlin Germany) and stored for offline analysis. Recorded wave files were then analyzed
using Avisoft SASLab Pro (Avisoft Bioacoustics, Berlin Germany), which enables the
creation of a spectrogram indicating the frequency and duration of individual calls.
2.3— Characterization of USVs:
Each file created by Avisoft Recorder software during self-administration was
opened and visually scanned for patterns resembling USVs. Once a putative, individual
USV had been identified, it was transposed using the Avisoft SASLab Pro software to a
lower frequency (5% of its normal speed) within the audible range of humans for a
secondary auditory confirmation of its “whistle-like” characteristic (Ciucci et al., 2007;
Brudzynski & Holland 2005; Fu & Brudzynski, 1994). Calls that were temporally
proximal to each other were designated as separate calls only if a period of silence greater
12
than 100 ms intervened and no continuity of pitch was observed. USVs were then
designated as fixed-frequency (FF; Figure 1A & 1B), if they maintained a stable mean
frequency (kHz), or frequency-modulated (FM) if the mean frequency showed a 3-kHz or
greater modulation. The mean frequency was calculated for each USV by averaging the
minimum and maximum frequencies. Calls for which the mean frequency modulated by
3-kHz or more were further designated as either “FM” (Figure 1C) calls or “Trills”
(Figure 1D) based on the number of pitch modulations within the call. Simple FM calls
were defined as those with only a single pitch modulation, characterized by a step- or
sweep-, up or down (Wright et al. 2010, Fu & Brudzynski, 1994; Brudzynski & Holland
2005; Knutson et al., 2002), while trills were defined as calls exhibiting two or more
pitch modulations. For analysis, the frequency of each FM call and trill was denoted as a
single value, the arithmetic mean, calculated using the absolute minimum and maximum
frequency values for each call.
2.4 Drug level calculations
Assuming first-order pharmacokinetics, calculated drug levels were derived using
the equation: Bn = (Bn-1 + D) e-KTn
(Yokel & Pickens, 1974);
Where:
Tn = the time since the previous cocaine infusion,
D = infusion dose (mg/kg),
Bn-1 = Calculated drug level (mg/kg) at last infusion
K = decay constant (0.693/t1/2)
13
For all subjects, time (Tn) was calculated on a minute by minute basis. Drug level
was calculated using the either the low (.355mg/kg/infusion), or high (.71mg/kg/infusion)
dose in Experiment 1, or the infusion (.71mg/kg/infusion) or micro-infusion (.0189
mg/kg/infusion) dose for Experiment 2. The rate constant for decay was selected based
on a previous report by Nayak and Colleagues (1976).
2.5 Experimental Apparatus:
All experiments were conducted in a custom Plexiglas chamber measuring 24 cm
x 34.5 cm x 34.5cm and housed inside of a larger wooden sound attenuating chamber
(~76 cm3). The chamber opened at the front by way of two Plexiglas doors and the floors
of the chamber consisted of a series of Plexiglas bars positioned 4.5 cm off the floor over
a litter paper. Animals were attached to an intravenous fluid delivery system consisting
of a single speed syringe pump (Razel Scientific, St. Albans, VT) which connected to a
custom built fluid swivel via 122 cm of tygon tubing. A 35.5 cm spring leash was
connected to the bottom of the fluid swivel and extended down to the head of the animal
through a 5 cm hole in the top of each chamber. Catheters built in-house were contained
inside of the spring leash and continued through the spring leash and a steel cannula on
animals’ heads into the animals’ right jugular vein. The swivel and leash were suspended
from a counterweighted and spring-balanced rod and fulcrum that allowed animals to
move freely about the chamber. For self-administration sessions, glass levers (4.7cm x
2.5 cm) were inserted through a hole 4 cm off the floor of the operant chamber and 6 cm
from the door of the chamber. Levers were set so that 0.049 Newtons of force were
required to activate a micro switch attached to the outer wall of the chamber. Tones (750
Hz), when used, were produced by custom tone generators (M.B. Turnkey Design,
14
Hillsborough, NJ). All experimental apparatus were controlled by a PC running MED-
Associates hardware and software (St. Albans, VT). Ventilation to the chambers was
provided by a series of shaded pole blowers (Grainger Industrial Supply, South Plainfield
NJ) and white noise was produced through a centrally located speaker via a white noise
generator at ~72 dB.
15
3. Experiment 1 Materials and Methods:
The present experiment examined ultrasonic calls emitted by rats during a 6-hour
self-administration binge at two doses of cocaine: 0.3755 (low dose) and 0.71 (high dose)
mg/kg/infusion using a discriminative stimulus (SD) tone to signal cocaine availability
periods on a VI 3-6 minute schedule. The combination of the VI 3-6 minute schedule
and high dose (0.71 mg/kg) was designed to match the average inter-infusion interval of
animals with continuous-access (FR-1) to cocaine (~4-5 minutes). Combining the low,
0.3755 mg/kg dose with the VI 3-6 minute schedule, on the other hand, prevents animals
from achieving “satiety”, i.e., the asymptotic drug levels they would self-administer on an
FR-1 schedule.
3.1—Shaping:
Prior to self-administration training, subjects were assigned to either a low-dose
(n= 10; 0.3755 mg/kg) or high-dose (n=7; 0.71 mg/kg) group. To facilitate the
acquisition of cocaine self-administration, subjects were initially shaped in the presence
of a continually sounding SD
tone. Responding on the lever during shaping terminated the
tone, produced either a 3.75 s (0.355 mg/kg/inf) or 7.5 s (0.71 mg/kg/inf) cocaine
infusion (depending on group assignment), and initiated a thirty-second time out.
Following each TO period, the SD
tone was again presented. Once subjects had
accumulated a sum of 10 self-administered cocaine infusions during a single daily
session, they immediately transitioned to the training contingencies described below. If
subjects did not accumulate 10 infusions in one session, the shaping procedure was
enacted during the subsequent session. All subjects were shaped to self-administer
16
cocaine within three daily sessions. The combination of dose and schedule was such that
low dose animals were unable to attain drug levels above ~4.5 mg/kg while high dose
animals were capable of attaining maximal blood levels of ~9 mg/kg.
3.2—Self-Administration:
During training, at 11:30 hours (immediately following the commencement of the
light cycle), a single non-retractable response lever was mounted on a wall of the operant
conditioning chamber. Following a 3-6 minute timeout (TO), subjects were presented
with an audible tone (3.5 kHz, 70 dB) SD. Responses on the lever in the presence of the
SD terminated the audible stimulus, produced either a 0.355-mg/kg or 0.71 mg/kg
intravenous infusion of cocaine (depending on group assignment), and began the
subsequent TO. If no responses occurred in the presence of the tone for two minutes, the
auditory cue was non-contingently terminated and a new 3-6 min TO began. USV
recordings were triggered by a TTL pulse at the time of tone onset. Each pulse triggered
a recording window consisting of 5 seconds prior to the onset of the tone, the entire tone
duration, and two minutes of the subsequent TO. All sessions lasted for 6 hours or 80
response-contingent infusions, whichever occurred first. Subjects received post-session
feeding in order to maintain a weight of approximately 320-340 grams. Water was
available ad libitum except during self-administration sessions. All animals were run
daily for 21 days (7 days/week) with no breaks between training sessions. USV
recordings occurred, on average, during session 17 ±0.73, with the earliest recording for
any animal occurring on session 11.
17
3.3—Statistical Analysis:
The observed USVs were placed into three bins based on their mean frequency:
18 to 32.99 kHz (22-kHz range), 33 to 37.99 kHz (33-kHz), and 38 to 80 kHz (50-kHz
range). The range from 33 to 37.99 kHz was inserted as a buffer between the well-
defined portions of the 22- and 50-kHz distributions in order to take all USVs in for
analysis. This range will henceforth be referred to as the “33-kHz” range. Given that long
22-kHz USVs constituted only 0.002% of all calls, both long and short 22-kHz USVs
were combined for analysis. The number of USVs (counts) observed in each frequency
bin, for each call type (i.e. FF, FM, and trill) was considered. The main statistical model
consisted of a mixed ANOVA (SAS PROC GLIMMIX, SAS Institute Inc., 2005).
Because graphical pilot analyses determined that the distribution of USV counts was
highly skewed, a gamma distribution, with a log link, was specified for the USV counts
in the mixed ANOVA. In addition, the distribution of the USV counts was transformed
slightly by adding a constant value of 0.001 in order to make it compatible with the
gamma distribution. Post-hoc comparisons were run using a Tukey-Kramer test with α
=0.05.
18
4.—Experiment 1 Results and Discussion:
4.1 Self-Administration:
All animals acquired stable cocaine self-administration prior to the commencement of
USV recordings (Figure 2). Subjects in the low dose group administered 11.78 ± 0.80
mg/kg on their first day of self administration, escalating to 28.10 ± 0.37 mg/kg on day
11, while high dose subjects increased drug intake from 15.80 ± 0.96 to 48.77 ± 1.21
mg/kg. The total number of daily responses also increased in both groups. Low dose
animals increased responding from 292.30 ± 36.73 responses on day 1 to 2112.30 ±
194.97 responses on day 11. High dose animals emitted 198.75 ± 38.69 responses on day
1 and increased to 638.13 ± 60.84 responses on day 11. Finally, the number of missed
opportunities to self-administer decreased between days 1 and 11 for both low (37.50 ±
2.18 to 3.3 ± 0.69) and high (11.88 ± 1.0 to 7.63 ± 0.94) dose subjects.
The mixed ANOVA revealed a significant main effect of self-administration session
and dose, as well as a significant dose session interaction for all three measures of
training proficiency: drug intake (mg/kg) [dose: F(1,160) = 26.32, p < .001, session:
F(10,160) = 24.83, p < .001, dose session: F(10,160) = 4.10, p < .001] missed
opportunities [dose: F(1,160) = 11.25, p = .001, session: F(10,160) = 25.02, p < .001,
dose session: F(10,160) = 14.68, p < .001] and total presses [dose: F(1,160) = 32.44, p
< .001, session: F(10,160) = 7.33, p < .001, dose session: F(10,160) = 4.42, p < .001].
Tukey-Kramer post-hoc comparisons revealed that the total number of presses was
stable by training day 3 [day 3-11- all |t(160)| < 2.70, p > .21], while drug intake (mg/kg)
19
and missed opportunities were stable by day 4 of training for both groups [drug intake
day 4-11- all |t(160)| < 2.70, p > .20; missed opportunities day 4-11- all |t(160)| < 1.53, p
> .91]. These data demonstrate that animals were reliably self-administering cocaine well
before their respective recording days. Finally, post-hoc comparisons of the dose
session interaction for mg/kg drug intake revealed that subjects in the high dose group
(day 11 = 48.77 ± 3.41 mg/kg) administered more cocaine than the low dose group (day
11 = 28.10 ± 1.18 mg/kg) on days 7 [|t(160)| = 4.84, p <.001] through 11 [|t(160)|=5.36, p
< .001] of training. This demonstrates that, following acquisition of the self-
administration task, subjects in the high dose group escalated their intake over that of the
low dose group.
4.2 Ultrasonic Vocalizations:
Individual USVs incorporated into the mixed model are shown in Figure 3 for both
the low (Figure 3A) and high (Figure 3B) dose groups. The mixed ANOVA revealed a
significant three-way interaction of dose frequency bin call type [F(4,831) = 22.66, p
< .001]. Significant two way interactions were observed for frequency bin call type
[F(4,831) = 22.75, p < .001], dose call type [F(2, 831) = 4.36, p < .05], and dose
frequency bin [F(2, 831) = 92.54, p < .001]. The main effects of call type [F (2, 831) =
98.31, p < .001], and frequency bin [F(2, 831) = 76.62, p < .001] were significant. The
main effect of dose [F(1,831) = 3.65, p = .06] was marginally significant.
Tukey-Kramer post-hoc tests revealed that, during the 6-hour self administration
session, subjects in the high dose group made significantly fewer short FF calls [26.57 ±
8.27; |t(831)| = 5.91, p < .001], FM calls [4.57 ± 1.54; |t(831)|= 6.32, p< .001], and trills
[1.14 ± 0.26; |t(831)|= 12.15, p< .001] in the 22-kHz range than subjects in the low dose
20
group [FF- 77.3 ± 22.58; FM- 23.7 ± 8.88; trill- 15.7 ± 7.25]. On the other hand, subjects
in the high dose group made significantly more FF calls [41.29 ± 15.06; |t(831)| = 3.69, p
< .05] and FM calls [23.57 ± 11.03; |t(831)| = 6.10, p < .001] in the 50-kHz range than
subjects in the low dose group [FF- 19.0 ± 4.34; FM- 6.9 ± 2.11]. The number of 50-kHz
trills between the low [5.6 ±1.83] and high [12.0 ± 3.96] dose did not differ [|t(831)| =
2.51, p= 0.52, NS]. Furthermore, no differences were observed for the 33 kHz range,
which separates the 22 and 50 kHz distributions [FF- |t(831)|= 0.93, p = 1.00, NS, FM-
|t(831)| = 0.11, p =1.00 NS, and Trill- |t(831)| = 0.65, p = 1.00, NS]. The observed group
differences between the 22 and 50 kHz ranges are shown in Figure 4 with all call types
combined.
Subsequent analyses of USVs within the high dose group revealed significantly
more 50-kHz FF calls [|t(831)| = 7.93, p <.001], FM calls [|t(831)| = 7.69, p < .001] and
trills [|t(831)|= 7.48, p < .001] relative to short 22-kHz FF calls, FM calls, and trills,
respectively. In contrast, within the low dose group, the number of short 22-kHz calls
greatly outweighed the number of 50-kHz calls for all three types [FF- |t(831)| = 17.19, p
< .001; FM- |t(831)| = 12.58, p < .001; trills- |t(831)| = 9.76, p < .001].
4.3—Discussion:
Ultrasonic vocalizations emitted by high dose subjects were consistent with those
observed in subjects receiving different doses of systemic cocaine (Mu et al. 2009), or
intravenous (Ahrens et al. 2009) or systemic amphetamine (Wintink & Brudzynski 2001),
all of which have shown that 50-kHz USVs are produced in response to drug
administration. On the other hand, the substantial number of short 22-kHz calls observed
for both the high and low dose groups in the present study is an important difference from
21
other studies that have recorded ultrasonic vocalizations during the administration of
psychomotor stimulants.
The results demonstrate a dose dependent, crossover interaction (Figure 4) of
short 22 and 50 kHz USVs that remained stable between call types (FF, FM, trills).
Animals that maintained self-administered drug levels below 4.2 mg/kg (the maximum
drug level achieved by any animal in the low dose group) exhibited predominantly short
22 kHz USVs and fewer 50 kHz USVs. In contrast, animals that maintained self-
administered drug levels as high as 6.98 mg/kg (the maximum drug level achieved by any
animal in the high dose group) exhibited predominantly 50 kHz USVs and fewer short 22
kHz USVs. Furthermore, the peak drug level of high dose subjects (all subjects >5.1
mg/kg) was in every case above the peak level observed for low dose subjects (4.2 mg/kg
or below).
An important factor that may have influenced the predominance of short 22-kHz
vocalizations in the low-dose group is that the 3-6 minute VI schedule of reinforcement
during the present self-administration paradigm precluded rats from attaining 'satiety' at
this dose (Ghitza et al. 2003; Root et al. 2009). Previous studies would suggest that
changes in USVs are directly tied to rising and falling levels of extracellular DA within
the mesolimbic DA system (Burgdorf et al. 2000, 2001; Thompson et al. 2006; Wintink
& Brudzynski 2001), which are critical for self-administration (Ritz et al. 1987).
Therefore, a subject’s transient drug levels, which are modulated by both dose and
temporal opportunities for self-administration, might play a critical role in the modulation
of USVs during cocaine self-administration.
22
5. Experiment 2 Materials and Methods:
The results of Experiment 1 suggest that blood levels of cocaine played a key role
in modulating ultrasonic vocalizations. Still, a number of other factors might explain the
observed differences (e.g. differences in rates of responding). To directly test the
hypothesis that blood levels of cocaine modulate the production of ultrasonic
vocalizations and that, more specifically, blood-levels of cocaine that approximate a
subject’s “preferred” drug level (satiety) induce a positive affective state, while those
further below the subjects’ preferred level (sub-satiety) produce a negative affective state,
subjects were tested using a series of drug “clamps”. During drug clamps subjects’ drug
levels were fixed at a given drug level using cocaine micro-infusions. All subjects USVs
were recorded during two baseline conditions as well as both a “High Clamp” (HC),
where drug levels were held at supra-satiety levels as well as a “Low Clamp” (LC),
where subjects’ blood levels were held sub-satiety.
5.1—Baseline Recordings:
Prior to surgery, subjects (n=8) were allowed to live in the self-administration
chamber for 3 days. On the fourth day, USVs were recorded for all subjects during two
6-hour baseline periods. The first of these recordings occurred from 10:30 AM to 16:30
PM (Day). The second recording occurred from 22:30 PM to 04:30AM (Night) of the
subsequent day. These recordings will henceforth be referred to as the Pre-Surgery Day
(PSD) and Pre-Surgery Night (PSN) recordings. The PSD recording acted as a time of
day control for self-administration sessions, while the PSN recording acted as control for
23
activity levels, as subjects are more active nocturnally when not self-administering
cocaine.
5.2—Shaping:
On the first day of self-administration training, subjects were initially shaped
using a 40 s TO period and 120 s availability period. Responding during the availability
period produced a 7.5 s (0.71 mg/kg/inf) cocaine infusion and initiated a forty-second
time out. Once subjects had accumulated a sum of 10 self-administered cocaine infusions
during a single daily session, they immediately transitioned to the training contingencies
described below. If subjects did not accumulate 10 infusions in one session, the shaping
procedure was enacted during the subsequent session. All subjects were shaped to self-
administer cocaine within the first three daily sessions.
5.3—Self –Administration:
Self-administration training began daily at 10:30 hours, immediately following
the commencement of the light cycle. At the start of the session a single non-retractable
response lever was mounted on the side of the operant conditioning chamber. For the first
10 infusions subjects were allowed to rapidly “load-up” on a schedule that included a
120 s drug availability period followed by a 40 s TO. Following the 10th
infusion,
availability periods (120 s.) were presented for the remainder of the session after a 1-6
minute variable TO. Responses during the availability period produced a 0.71mg/kg
infusion and initiated the next TO. If no responses occurred for two minutes, the
availability period was terminated and the next TO was initiated. All sessions lasted for 6
hours or 80 response-contingent infusions, whichever occurred first. Subjects received
post-session feeding in order to maintain a weight of approximately 320-340 grams.
24
Water was available ad libitum except during self-administration sessions. All animals
were trained for 14 days (7 days/week) with no breaks between training sessions.
5.4—Clamp Test and Recordings:
Following 14 days of training the role of drug level in USV production was
tested by administering two drug clamps: a Low (sub-satiety) Clamp (LC) and High
(supra-satiety) Clamp (HC). For clamp conditions, drug satiety was defined specifically
for each subject as the subject’s maximum peak drug level during the second week of
training, as recent work in our lab suggests that this correlates highly with response
cessation (Root et al., 2011), and it has been suggested by others that response cessation
is a reliable measure of drug satiety (Norman & Tsibulsky 2006). For the HC condition,
subjects’ drug levels were fixed at .1 mg/kg above each subject’s peak drug level (supra-
satiety threshold), while “sub-satiety” during the LC condition was defined as 50% of the
subject’s supra-satiety threshold. During each clamp, subjects were maintained at a fixed
drug level by delivering computer-controlled intravenous microinjections of cocaine with
a constant inter-infusion interval that, once adjusted, maintained the animal’s drug levels
within ± .02 mg/kg of the target level. For all subjects, microinjections consisted of
recurring 200ms infusions (.0189 mg/kg/infusion) with a custom inter-infusion interval
designed to maintain each subject’s clamp dose. Subjects were counterbalanced such that
half the subjects (n=4) received a sub-satiety clamp first, while the other half received a
supra-satiety clamp first (n=4). All subjects were then tested in the opposite clamp
condition following a 1-3 day return to regular training contingencies. On clamp testing
days, subjects were allowed to self-administer under their training contingencies for 90
minutes. Subsequently, the response manipulandum was removed and subjects began a
25
transition period for 30 minutes. During the transition period, subjects’ drug levels were
slowly increased or decreased in order to approach their clamp level. Upon stabilizing at
the clamp level, subjects’ drug levels were fixed for the remaining 240 minutes of the six
hour session. USV recordings were triggered by a TTL pulse at the start (time zero) of
the 6-hour clamp session, and were recorded continuously for the entire 6 hours. For
each recording a custom Plexiglas tube was used to house the microphone and increase
the signal:noise in recordings (Figure 5).
5.5—Statistical Analysis:
The main statistical analysis consisted of two repeated measures ANOVAs. The
repeated measures ANOVAs were run using SAS PROC GLIMMIX (SAS Institute Inc.,
2005). The first model consisted of a 4 x 6 x 3 design examining four levels of
Condition, six levels of Session Time (hours) and three levels of Call Frequency. The
four levels of Condition were PSD, PSN, HC and LC. The six levels of Session Time
consisted of hours one through six of each session. For Call Frequency the observed
USVs were placed into three bins based on their mean frequency: 18 to 32.99 kHz (22-
kHz range), 33 to 37.99 kHz (33-kHz), and 38 to 80 kHz (50-kHz range) (see sections 2.3
& 3.3 for further details). The dependent measure for the model consisted of the number
of USVs (counts) observed in each Condition, for each time bin, at each Call Frequency.
After determining that the data were highly-skewed, a gamma distribution, with a log
link, was specified for the model. To accommodate the gamma distribution, a constant of
1 count was added to every cell of the model. Post-hoc comparisons were run using a
Tukey-Kramer test with α =0.05.
26
To examine differences that occurred as a function of Call Type a second model
was included. The inclusion of a second model was necessary, as pilot analyses where
Call Type was included as a fourth factor in the above mentioned model created a highly-
irregular zero inflated distribution (i.e. by dividing the data into more categories, the
number of bins where zero counts are observed increases exponentially).The second
model consisted of a 4 x 3 x 3 design examining four levels of Condition, three levels of
Call Frequency and three levels of Call Type. The four levels of Condition were PSD,
PSN, HC and LC. The three levels of call type were FF, FM, and Trill. Call Frequency
was designated in the same manner as the previous model. The dependent measure for
the model consisted of the number of USVs (counts) observed in each Condition, for
each Call Frequency bin, at each Call Type. After determining that the data were highly-
skewed, a gamma distribution, with a log link, was specified for the model. To
accommodate the gamma distribution, a constant of 1 count was added to every cell of
the model, thereby removing instances of zero counts. Post-hoc comparisons were run
using a Tukey-Kramer test with α =0.05. Given that long 22-kHz USVs accounted for
only 2.6% of the observed vocalizations, both long and short 22-kHz USVs were
combined for analysis in both models.
27
6. Experiment 2 Results and Discussion:
6.1 Behavior:
Animals were shaped to respond within the first 1-3 days of training. For all
subjects the number of responses increased from 26.63 ± 3.95 lever presses during the
first session to 202.43 ± 38.37 lever presses in session fourteen (Figure 6). Similarly,
subjects increased their cocaine intake from 11.79 ± 1.37 mg/kg in session one to 34.94 ±
3.26 mg/kg during session fourteen (Figure 7), which coincided with an increase in the
number of earned infusions from 14.88 ± 1.90 in the first session to 48.00 ± 4.37
infusions in session fourteen (Figure 8) and an increase in the percentage of opportunities
wherein subjects received an infusion from 12.55 ± 1.21% to 50.92 ± 5.02% (Figure 9).
In contrast, subjects’ missed opportunities decreased from 102.75 ± 6.94 missed
opportunities during session one to 47.29 ± 6.12 missed opportunities during session
fourteen. Finally, subjects’ weights decreased from 333.75 ± 5.60 grams on session one
to 319.43 ± 2.85 grams during session fourteen (Figure 10). Importantly, this profile of
self-administration is consistent with observations from Experiment 1.
Subjects weighed an average of 322.00 ± 2.27 grams on the day of the high clamp
condition and 321.00 ± 3.36 grams during the low clamp condition. Moreover, drug
levels in the high clamp condition were fixed at 6.41 ± 0.37mg/kg, while the average
drug level for the low clamp was 3.20 ± 0.18mg/kg. Prior to lever removal, subjects
earned an average of 17.38 ± 2.03 infusions in the high clamp condition and 17.00 ± 2.90
infusions in the low clamp condition. A drug curve from a representative animal for both
the low and high clamp condition is shown in figure 11.
28
6.2 Ultrasonic Vocalizations (Time Analysis):
6.2.1—Omnibus Results: The repeated measures ANOVA examining Condition
x Call Frequency x Session Time (binned by hour) revealed a significant three way
interaction [F(30,497)=2.67, p<0.0001], significant two-way interactions of Condition x
Call Frequency[F(6,497)=17.59, p<0.0001], Condition x Session Time[F(15,497)=8.69,
p<0.0001] and Frequency x Session Time[F(10,497)=3.74, p<0.0001], and significant
main effects for Condition [F(3,497)=50.86, p<0.0001], Call Frequency [F(2,497)=75.91,
p<0.0001] and Session Time [F(5,497)=19.76, p<0.0001]. The mean ± SEM for all cells
in the Condition x Call Frequency x Session time model can be found in Table 2.
6.2.2—Condition x Call Frequency x Session Time: Effect of time. A series of
Tukey-Kramer post-hoc comparisons was used to examine the effects of session time on
USV production. Each of these comparisons was made within-condition while holding
the level of Call Frequency constant, thereby allowing the effects of time within each
condition to be extracted for each individual frequency range (Figure 14).
Comparisons Within-Condition for the Condition x Call Frequency x Session time
interaction revealed that no differences existed across hours in either the PSD or PSN
conditions for the 50-kHz ( all p >0.99, N.S.), 33-kHz ( all p >0.99, N.S.), or 22-kHz (all
p >0.98, N.S.) frequency ranges.
Post-hoc tests on the High Clamp condition revealed that there were more 50-
kHz calls during hour 1 than all subsequent hours [1 vs. 2: t(497)=7.88, p<0.0001; 1 vs.
3: t(497)=8.00, p<0.0001; 1 vs. 4: t(497)=8.72, p<0.0001; 1 vs. 5: t(497)=8.32, p<0.0001;
1 vs. 6: t(497)=8.95, p<0.0001]. On the other hand, there were no differences in 50-kHz
calls for all of the remaining pairwise combinations of hours 2 through 6 (all p>0.99,
29
N.S.). Similarly, there were significantly more 22-kHz calls in hour 1 of the HC condition
than hours 3 through 6 [1 vs. 3: t(497)=4.35, p<0.05; 1 vs. 4: t(497)=6.53, p<0.0001; 1
vs. 5: t(497)=5.77, p<0.0001; 1 vs. 6: t(497)=6.53, p<0.0001]. The difference between
hours 1 and 2 was not significant (p=0.99, N.S.). With the exception of a marginally
significant difference in the number of 22-kHz calls between hours 2 and 4 (p=0.054) and
2 and 6 (p=0.054), none of the other pairwise comparisons of Session Time were
significant. No differences were observed for any of the pairwise comparisons of Session
Time for calls in the 33-kHz range (all p>0.99, N.S.).
In the LC condition, there were more 50-kHz calls in hour 1 than all other hours
[1 vs. 2: t(497)=8.01, p<0.0001; 1 vs. 3: t(497)=6.39, p<0.0001; 1 vs. 4: t(497)=8.19,
p<0.0001; 1 vs. 5: t(497)=8.66, p<0.0001; 1 vs. 6: t(497)=8.62, p<0.0001], but none of
the other pairwise comparisons between hours 2-6 were significant (all p>0.99, N.S.).
Similarly, none of the pairwise comparisons were significant across hours for either the
33-kHz (all p>0.73,N.S.) or 22-kHz (all p>0.99) frequency ranges.
6.2.3—Condition x Call Frequency x Session Time: Effect of Call Frequency.
A second series of Tukey-Kramer post-hoc comparisons was used to examine differences
in call frequency. Each of these comparisons was made within-condition while holding
levels of session time constant. These comparisons were used to illuminate differences in
the numbers of calls between different categories of frequency during a given
combination of condition and hour (Figure 14).
Post-hoc comparisons suggested that there were no differences between the
number of USVs in the 50- and 33-kHz, 50- and 22-kHz, or 22- and 33-kHz ranges
30
during any hour of the experiment for both the PSD (all p>0.41, N.S.) or PSN (all p
>0.99, N.S.) conditions.
In the HC condition, it was observed that there were significantly more 50- than
33-kHz USVs during hour 1 [t(497) = 6.39,p<0.0001]. Conversely, there were no
differences between the numbers of 50- and 22-kHz or 22- and 33-kHz during the first
hour of the HC condition (p =.99, N.S. and p=.10, N.S. respectively). Furthermore, no
differences existed between any of the aforementioned frequency ranges during the
subsequent 5 hours (all p >0.54, N.S.).
In the LC condition there was a significant difference between the number of 50-
and 33-kHz USVs during hour 1[t(497)=6.11, P<0.0001], but no differences between 50-
and 22-kHz or 22- and 33-kHz USVs. Importantly, this demonstrates that there were
similar patterns of USV production in the first hour (before the clamp was imposed) by
subjects in both the LC and HC condition. In contrast, subjects in the LC condition
emitted significantly greater numbers of 22-kHz USVs than 33-kHz or 50-kHz USVs
during hours 2-6 [22- vs. 33-kHz: hour 2 t(497)= 6.20, p<0.0001; hour 3 t(497)=6.54,
p<0.0001; hour 4 t(497)= 5.77, p<0.0001; hour 5 t(497)=6.43, p<0.0001; hour 6
t(497)=6.53, p<0.0001 || 22- vs. 50-kHz: hour 2 t(497)=5.82, p<0.0001; hour 3
t(497)=5.04, p<0.01; hour 4 t(497)=5.25, p<0.0001; hour 5 t(497)=5.89, p<0.0001; hour
6 t(497)=5.92, p<0.0001]
6.2.4—Condition x Call Frequency x Session Time: Effect of Condition. A
third series of Tukey-Kramer post-hoc comparisons was used to examine the effects of
Condition on USV production. These comparisons were made by holding the levels of
Session Time and Call Frequency constant. This comparison allowed for the
31
examination of differences between conditions at equivalent time-points and frequency
bins (Figure 14).
Comparisons between conditions revealed no differences in the number of 50-,
33-, or 22-kHz USVs between the PSD and PSN conditions during any hour of the
experiment (all p>0.89, N.S.). It was also shown that there were more 50-kHz USVs in
the HC condition than both the PSD [t(497)=7.74, p<0.0001] and PSN [t(497)=8.19,
p<0.0001] conditions during the first hour. However, there were no differences between
the HC and PSD or PSN conditions in hours 2 through 6 (all p>0.99, N.S.). Similarly,
there were no differences in the number of 33-kHz USVs between the HC and PSD or
PSN conditions during any hour of the experiment (all p>0.98). Interestingly, there were
more 22-kHz USVs in the HC condition during hour 1 than in the PSN condition
[t(497)=5.92, p<0.0001], but not the PSD condition (p=.73, N.S.). Nevertheless, there
were no differences in 22-kHz USVs between the HC and PSD or PSN conditions for the
remaining hours of the experiment (all p >0.57, N.S.). In line with the HC condition, it
was observed that more 50-kHz USVs were observed in hour 1 of the LC condition, than
in hour 1 of either the PSD[t(497)=8.21, p<0.0001] or PSN[t(497)=8.66, p<0.0001]
condition, but that no differences existed during the remaining hours (all p>0.99). In
addition, no differences in 30-kHz USVs were observed between the LC and PSD or PSN
conditions (all p>0.53, N.S.). In Contrast, there were more 22-kHz USVs in all hours
of the LC condition than in the PSD and PSN conditions with the exception of the
comparisons between LC and PSD during hours 1 and 4 (p=0.34, N.S. and p=0.14, N.S.
respectively)[LC vs. PSD: hour 2 t(497)=6.33, p<0.0001; hour 3 t(497)=6.80, p<0.0001;
hour 5 t(497)=5.21, p<0.001; hour 6 t(497)=5.55, p<0.0001;LC vs. PSN: hour 1
32
t(497)=6.37, p<0.0001; hour 2 t(497)=4.93, p<0.01; hour 3 t(497)=5.83, p<0.0001; hour
4 t(497)=4.84, p<0.01; hour 5 t(497)=4.72, p<0.01; hour 6 t(497)=4.89, p<0.01]
Comparisons between the HC and LC conditions suggested that no differences
existed in the number of 50- or 33-kHz calls during any hour of the session (all p>0.99,
N.S.). In addition, there were no differences in the number of 22-kHz USVs emitted
between the HC and LC conditions during the first two hours of the experiment (all
p>0.76, N.S.). On the other hand, subjects in the LC condition emitted significantly
greater numbers of 22-kHz USVs during the last four hours of the session than subjects in
the HC condition [hour 3 t(497)=5.89, p<0.0001; hour 4 t(497)=6.48, p<0.0001; hour 5
t(497)=5.90, p<0.0001; hour 6 t(497)=6.72, p<0.0001].
6.3 Ultrasonic Vocalizations (Call-Type Analysis):
6.3.1—Omnibus results. The individual USVs for all conditions are shown in
figure 12. The mean number of counts ± SEM for each cell of the Condition x Call
Frequency x Call Type model can be found in Table 1. Omnibus results for the repeated
measures ANOVA did not reveal a significant three-way interaction [F(12,245)= 1.08,
p=0.38, N.S.] or a significant two-way interaction of Condition x Call Type
[F(6,245)=.71, p=0.64, N.S.]. Moreover, the interaction of Call Frequency and Call Type
was only marginally significant [F(4,245)=2.26,p=.064]. On the other hand, there was a
significant interaction of Condition x Call Frequency [F(6,245)=8.66, p<0.0001] as well
as significant main effects for Condition [F(3,245)=58.87, p<0.0001], Call Frequency
[F(2,245)=78.60, p<0.0001], and Call Type [F(2,245)=6.90, p<0.01]. The omnibus
results for the model suggest that the interaction between Condition and Call Frequency
had the greatest effect on USV production, while Call Type is weakly (if at all)
33
modulated as a function of Condition, Call Frequency or both. Given that the interaction
between Condition and Call Frequency is more adequately described in the previous
model’s three-way interaction and that Call Type seems unaffected by the main variable
of interest (Condition), the post-hoc results for the present model are presented in
Appendix A.
6.4. Hour 1 Analysis:
The results of the Condition x Call Frequency x Session Time analysis suggested
that substantial numbers of 50- and 22-kHz USVs occurred during the first hour of the
session in both the Low and High Clamp conditions. While hour 1 consisted primarily of
the loading period, animals exhibited varying latencies to initiate responding ranging
from 2 to 25 minutes prior to the first infusion (mean 7.5 ±1.35 minutes). Therefore,
given that animals tend to emit only one call frequency at a given time point and given
that a number of studies have demonstrated that animals emit anticipatory USVs (Ma et
al 2010; Maier et al 2010), an analysis of the first hour was conducted in order to discern
the differences in subjects’ USV emissions that occurred before and after animals self-
administered their first infusion of cocaine. The period of time prior to the first infusion
will henceforth be referred to as the Pre Drug Period and the time period after the first
infusion will be referred to as the Post Drug Period.
Contingencies for Hour 1 of the LC and HC conditions were equivalent, and it
was determined that the condition and the order of condition presentation had no effect
on USV production. Therefore, USVs from both the HC and LC conditions were
combined for the Hour 1 analysis. From the combined data it was determined that
subjects emitted 2.04 ± 1.18 22-kHz USVs/min (corresponding to 45.3±9.8% of the total
34
22-kHz USVs in hour 1) in the Pre Drug Period and 0.15 ± 0.06 22-kHz USVs in the Post
Drug Period. In contrast, animals emitted only 0.35± 0.28 50-kHz USVs/min
(14.7±8.5% of the total 50-kHz USVs in hour 1) in the Pre Drug Period but emitted 1.19
± 0.32 50-kHz USVs/min in the Post Drug Period. Finally, 33-kHz USVs remained low
across both drug periods with approximately 50% of the 33-kHz calls occurring in each
drug period (0.08± 0.04 and 0.05 ± 0.02 for Pre and Post respectively). Data for the Hour
1 Analysis are shown in Figure 15.
The Repeated Measures ANOVA for the Hour 1 analysis consisted of a 3 x 2
design examining three levels of Call Frequency (22-, 33-, and 50-kHz frequency bins)
and two levels of Drug Period (Pre and Post). The model revealed a significant main
effect of Call Frequency [F (2,83)=11.20, p <0.0001] and a significant interaction of Call
Frequency and Drug Period [F(2,83)=11.38, p<0.0001], however, no significant main
effect of Drug Period was observed [F(1,83)=1.54, p=0.22, N.S.].
Tukey-Kramer post hoc comparisons for the Call Frequency x Drug Period interaction
revealed that 22-kHz USVs were emitted at a greater rate during the Pre Drug Period than
50- [t(83)=3,26, p <0.05] or 33-kHz [t(83)=4.68, p<0.001] USVs. However, no difference
was found between rates of 33- and 50-kHz USVs during the Pre Drug Period (p=0.72,
N.S.). On the other hand, subjects emitted 50-kHz USVs at a greater rate during the Post
Drug Period than 22- [t(83)= 3.48, P<0.01] or 33-kHz [t(83)=4.49, p<0.001] USVs.
However, no differences were found between 22- and 33-kHz USVs during the Post Drug
Period (p= 0.91, N.S.). Finally, subjects emitted 22-kHz USVs at a higher rate during the
Pre Drug Period than the Post Drug Period [t(83)=4.19, p<0.001]. Comparisons between
35
the Pre and Post Drug Periods for the 33- and 50-kHz ranges did not indicate any
differences in rates of calling (p=0.99 and 0.12, respectively).
6.5. Discussion:
An examination of the Condition x Call Frequency x Session Time interaction
revealed no differences between all hours for the baseline conditions; however, there
were significantly more 50-kHz USVs and 22-kHz USVs in the HC condition during
hour 1 than during the remaining hours of the HC session. In addition, animals emitted
more 50-kHz USVs during the first hour of the LC condition than any other hour.
Interestingly, the analysis of 22- and 50-kHz USVs during the first hour of the Clamp
sessions revealed that animals tend to emit predominantly 22-kHz USVs prior to self-
administering their first infusion of drug, but emit mostly 50-kHz USVs after their first
infusion. Importantly, this suggests that 50-kHz USVs are correlated with rising levels of
cocaine during the loading period. Moreover, the presence of 22-kHz USVs during the
Pre Drug Period might reflect a negative anticipatory reaction to cocaine (as has been
shown by others; Wheeler et al 2008) or the effects of acute withdrawal from cocaine
administered during the previous session.
In Contrast to the HC condition, no differences in 22-kHz USVs were observed
across hours for the LC condition. An examination of the mean number of 22-kHz USVs
emitted across hours (Table 2) and comparisons between Call Frequency ranges during
each hour of the LC condition indicated that this was due to the consistent emission of
22-kHz USVs by subjects in the LC condition across the session. Consistent with the
hypothesis for Experiment 2, this suggests that negative affective USVs were correlated
with the sub-satiety drug levels that animals experienced during the final hours of the LC
36
session. Similar to the HC condition, 22-kHZ USVs during the first hour occurred at their
highest rates prior to the first infusion.
Appropriately, the differences observed between hours of the session also
coincide with a number of differences between frequencies. First, during hour 1 of the
clamp session, subjects in the HC condition showed significantly more 50-kHz USVs
than 33-kHz, but not 22-kHz USVs. Moreover, no differences were observed during the
remaining hours. Second, subjects in the LC condition showed significantly more 50-kHz
USVs than 33-kHz (but not 22-kHz) USVs during the first hour (load up). Also, the
number of 22-kHz USVs was significantly greater than the numbers of both 50- and 33-
kHz calls for the remaining hours of the session. Finally, subjects in the PSD and PSN
conditions showed no differences when comparing between call frequencies. Thus, while
the results for 50-kHz USVs do not support the notion that static circa-satiety drug levels
produced positive affective USVs, the persistence of short 22-kHz USVs in the LC
condition supports the role of sub-satiety drug levels in the induction of negative affect.
Lastly, when comparing between conditions, it was shown that no differences
were observed between the PSD and PSN conditions at any frequency range, during any
hour of the experiment. This point is of some importance, as it has been suggested that
rats might demonstrate negative affect during diurnal testing. While the nocturnal
patterns of rats might suggest an innate dispreference for light, it would appear that
subjects habituated to the chamber lighting after only three days, and that only the
brightest levels of light induce a consistent aversive reaction across time (Barker et al.
2010b). In fact, subjects produced low numbers of USVs in all frequency ranges, with no
discernable differences between frequencies or between the different times of day.
37
Another important observation was that no differences existed between any conditions
within the 33-kHz range. Given that this range is operationally meant to act as a
statistical buffer between well-defined portions of the negative and positive affective
frequency distributions, it is essential that there is congruence in this range amongst the
various conditions. This result was consistent across both Experiment 1 and Experiment 2
suggesting that the 33-37.99 kHz range acts as an appropriate buffer for future studies.
Between condition comparisons also demonstrated that more 50-kHz USVs were
emitted by subjects in greater numbers during the first hour of the HC and LC conditions
than in the PSD and PSN conditions, even though the baseline conditions and
experimental conditions did not differ from one another with respect to 50-kHz USVs.
Again, this implicates cocaine, and specifically the loading period, in the production of
50-kHz USVs. Furthermore, during hour 1, the number of 22-kHz USVs was greater in
the LC condition than both baseline conditions, and the number of 22-kHz USVs in the
HC condition was greater than the PSN condition. However, upon the completion of
hour 1, subjects in the HC group reduced their rates of calling to levels that were similar
to the baseline conditions, whereas subjects in the LC group continued to emit 22-kHz
USVs at greater rates than each of the other conditions for nearly all comparisons during
the latter 5 hours of the experiment, suggesting that negative affect, indexed by short 22-
kHz USVs, is in fact produced as a result of the contingencies of the low clamp (i.e. sub-
satiety drug levels).
Consistent with the results from Experiment 1, the majority of 22-kHz USVs
(97.4%) observed were short. In fact, none of the observed calls in the LC and HC
conditions met the generally accepted standards for long 22-kHz calls (i.e. a duration
38
>300ms). Instead, the long 22-kHz USVs that were observed occurred in the baseline
conditions. Also consistent with Experiment 1, subjects in the LC condition emitted more
22-kHz USVs than subjects in all other conditions. Most importantly, this result
demonstrates that while 50-kHz USVs were produced in similar numbers during both the
HC and LC conditions, there was in fact a difference in the production of 22-kHz USVs
when blood levels of cocaine were held below animals’ preferred levels. This finding
adds to growing evidence in support of the view that short 22-kHz calls reflect negative
affect (Brudzynski et al. 1991, 1993, 2001; Oliveira et al., 2006; Takahashi et al 2010;
Barker et al, 2010).
While inconsistent with the hypotheses derived from Experiment 1, the observed
similarity in 50-kHz call production between the experimental groups suggests that the
process involved in the induction of 50-kHz USVs was consistent between the self-
administration conditions. Indeed, given the high levels of variability in individual rates
of calling, the within-subjects design in Experiment 2 should likely be considered the
more accurate of the two experiments for this type of comparison. Comparisons of 50-
kHz USVs also revealed that more positive affective calls were emitted in the
experimental conditions than either of the baseline recordings (which did not differ from
one another with respect to 50-kHz calls), substantiating the notion that the emission of
50-kHz USVs was in fact induced by the experimental contingencies (i.e. the
administration of cocaine).
The results of Experiment 2 also extend the results of Experiment 1 by
demonstrating that animals emit low levels of USVs in all frequency ranges under drug-
free circumstances. Moreover, Experiment 2 demonstrates that differences in animals’
39
ultrasonic emissions during Experiment 1 were not related to differential rates of lever
responding between dose groups. Specifically, the levers were removed prior to clamp
testing for animals in Experiment 2. While the observed difference in effort remained a
plausible contributor to the production of USVs (especially short 22-kHz calls) during
Experiment 1, this seems unlikely given the lever-free replication during the second
experiment.
40
7. General Discussion:
The present results demonstrate a dual process of cocaine self-administration
which may have implications for the treatment of substance abuse. While cocaine is
often presumed to act solely as a positive reinforcer, the present experiments have
demonstrated that vocalizations recognized as reflecting both positive and negative
affective processing occur during self-administration. Interestingly, subjects emitted
primarily negative affective (22-kHz) calls prior to self-administration but then emitted
primarily positive affective (50-kHz) USVs during the loading period. These calls
subsided when animals approached circa-satiety drug levels. On the other hand, when
animals’ blood levels of cocaine were statically held at sub-satiety levels, animals
initiated and maintained the emission of negative affective calls.
An important difference between the present experiments and others examining
ultrasonic vocalizations during the administration of psychomotor stimulants was the
presence of short 22-kHz calls. Although peculiar, this difference might be accounted for
by selective sampling of the 50-kHz range (Mu et al. 2009), by operational definitions
which exclude 22-kHz USVs less than 300 ms in duration, or by experimenter-
administered versus self-administered stimulants (Ahrens et al. 2009). In support of the
last possibility, the present animals likely self-administered blood drug levels below those
reached via single experimenter-administered bolus injections greater than 15 mg/kg
(Nicolaysen et al. 1988; Pettit and Justice 1991; Weiss et al. 1992; Stuber et al. 2005).
Indeed, these USVs would also be less prevalent in experiments where subjects are able
to maintain circa-satiety drug levels.
41
While the procedures of some published studies have excluded short 22-kHz
USVs, it is also true that none of the aforementioned studies combining USVs and the
administration of psychomotor stimulants have reported the presence of long 22-kHz
USVs (Ahrens et al 2009; Mu et al 2009; Burgdorf et al 2001; Thompson et al 2006;
Wright t al 2010). Still, the inclusion of short 22-kHz USVs may be of particular
importance during cocaine self administration. As seen during experiment 2, subjects in
the baseline conditions emitted both long and short 22-kHz USVs. On the other hand,
long calls were absent during cocaine self-administration conditions. This result is
consistent with the actions of psychomotor stimulants, which decrease the likelihood of
long-patterned movements and increase the number of short, stereotyped movements
(Flagel et al 2007; Kuczenski et al 1991). Importantly, there is no empirical rationale for
the separation of long and short 22-kHz USVs. In fact, a recent report by Takahashi and
colleagues (2010) suggested that the long and short 22-kHz USVs belong to a single
statistical “cluster” and found that both long and short 22-kHz USVs were observed
during fighting behavior by concomitants, whereas these calls were not observed during
feeding or general locomotion. A final reason for not excluding short 22-kHz USVs from
study is that their durations are no different from those of 50-kHz calls, which are not
challenged on the basis of their short durations.
In light of the pharmacological effects of cocaine, it is important to consider the
possibility that short 22-kHz USVs might be related to its sympathomimetic effects (e.g.
increased heart rate, respiration etc.). While the present experiments cannot rule out the
possibility that such calls are produced as a byproduct of these effects, evidence from our
laboratory does suggest that motivation exists to escape the state in which these calls are
42
produced (Root et al., 2011). That is, when animals’ drug levels are sub-satiety, animals
will continually respond in order to receive an infusion (present Experiment 1). In
contrast, when drug levels are supra-satiety, animals exhibit a complete cessation of
responding (Root et al., 2011; Norman & Tsibulsky 2006). Combined with the
prevalence of negative affective calls at these drug levels (present study), these findings
suggest that responses during a cocaine binge may act as a form of escape behavior,
wherein subjects perseverate responding in order to escape or avoid drug levels that
produce the most aversive effects, e.g., sympathomimetic effects or an irritant drive state
analogous to hunger (Wise et al, 1995). The cessation of responding and absence of these
calls and at circa-satiety drug levels suggests that satiety acts as a goal-state during self-
administration, providing negative reinforcement by replacing the aversive sub-satiety
state.
While previous theories of 50-kHz production have suggested a dopaminergic
mechanism of action (Fu & Brudzynski 1994; Ahrens et al. 2009; Wintink & Brudzynski
2001; Brudzynski 2008), it is unclear whether the present data support this assertion.
Indeed, 50-kHz USVs were observed during the drug loading period, presumably when
extracellular levels of dopamine were experiencing their most dynamic changes.
However, rates of 50-kHz USVs were low prior to the first infusion of cocaine (although
pre-session calls can be conditioned: Ma et al 2010; Maier et al 2010). Thus, while
extracellular dopamine remained elevated for the entire 6 hour session, rates of 50-kHz
USV production declined significantly after the first hour of the experiment in both the
high and low clamp groups.
43
This curious result might suggest that 1) dopamine plays a role in modulating the
production of 50-kHz USVs but that rates of USV production are not directly linked to
levels of mesolimbic dopamine 2) Subjects rapidly habituate to increased levels of
extracellular dopamine after which calling subsides 3) 50-kHz USVs are produced as a
result of changes in extracellular dopamine rather than static levels of dopamine
(including high static levels) or 4) The effects of increased extracellular dopamine are
counteracted by an opposing neural process. Some theories of USV production postulate
that 22-kHz USVs are modulated by cholinergic projections from the laterodorsal
tegmental nucleus, while 50-kHz USVs are controlled by dopaminergic projections from
the VTA (Brudzynski 2008). Appropriate to these theories, cocaine has an effect on both
of these neurotransmitter systems. Given the anatomical overlap between these two
systems (e.g. within the nucleus accumbens, ventral pallidum, and other mesolimbic
regions; Satoh & Fibiger 1986; Cornwall et al. 1990;Roberts & Koob 1982), it stands to
reason that different doses or fluctuating levels of cocaine alter the cholinergic-
dopaminergic interaction in such a way as to modulate USV production. Certainly,
evidence has suggested differential roles for these two neurotransmitters in the nucleus
accumbens when examining approach and avoidance behavior (Rada & Hoebel 2001;
Hoebel, Avena, & Rada 2007).
Data from the present experiments are consistent with data from human subjects,
suggesting that both negative and positive affective processing play a role in cocaine
addiction (Volkow et al 1997; Fox et al 2007; Newcomb & Felix-Ortiz 1992; Hodgins et
al 1995; McKay et al 1995). Furthermore, the data suggest a possible mechanism by
which drug seeking is perpetuated: 1) Either positive (Ma et al 2010; Maier et al 2010) or
44
negative anticipation (present study) initiate bingeing, which induces a positive affective
state (present study); however, consistent with data from humans (Volkow et al. 1997),
subjects’ affective states fall more rapidly than levels of extracellular dopamine. 2)
Bingeing behavior is perpetuated by negative reinforcement as subjects continue to self-
administer over time in order to relieve the onset of negative affective symptoms (e.g.
dysphoria; Mendelson & Mello 1996). 3) After learning that drug use is an effective
method for relieving negative affect, subjects return to drug use during the dysphoric
effects of psychological withdrawal (Covington & Miczek 2003; Barros & Miczek 1996;
Mutschler & Miczek 1998a; Mutschler & Miczek 1998b). 4) The learning that occurs
during drug abuse (i.e. positive anticipation, initial positive effects of drug, and relief of
negative affect through drug use) become generalized during abstinence such that A) cues
that evoke anticipation of drug elicit drug-seeking behaviors (Ghitza et al. 2003) or B)
situations that evoke a negative affective state may trigger an escape response that leads
to drug use (Hodgins et al. 1995;Baker et al. 2004; Fox et al. 2007; Koob 1997; Koob
2008; Koob 2009; Koob 2009b; McKay et al. 1995) .
45
8. References
Ahrens AM, Ma ST, Maier EY, Duvauchelle CL, Shallert T (2009) Repeated intravenous
amphetamine exposure: rapid and persistent of 50-kHz. Behav Brain Res
197:205-209
American Psychiatric Association: Diagnostic and Statistical Manual of Mental
Disorders, Fourth Edition. Washington, DC, American Psychiatric Association,
1994.
Baker TB, Piper ME, McCarthy DE, Majeskie MR, Fiore MC (2004) Addiction
motivation reformulated: an affective processing model of negative
reinforcement. Psychol Rev 111:33-51
Barfield RJ, Thomas, DA (1986) The role of ultrasonic vocalizations in the regulation of
reproduction in rats. Ann N Y Acad Sci 474:33-43
Barker DJ, Root DH, Ma S, Jha S, Megehee L, Pawlak AP, West MO (2010) Dose-
dependent differences in short ultrasonic vocalizations emitted by rats during
cocaine self-administration. Psychopharmacology 211:435-442
Barros HMT, Miczek, KA (1996) Withdrawal from oral cocaine in rats:
ultrasonic vocalizations and tactile startle. Psychopharmacology 125:379-384
Bijlsma EY, Olivier B, Groenink L (2010) Cocaine induced changes in affective state
modulate the light-enhanced startle response. Behav Brain Res 213:117-120
Blanchard RJ, Blanchard DC, Agullana R, Weiss SM (1991) Twenty-two kHz alarm
cries to presentation of a predator, by laboratory rats living in visible burrow
systems. Physiol Behav 50:967-972
Brudzynski SM, Ociepa D, Bihari F (1991) Comparison between cholinergically and
naturally induced ultrasonic vocalization in the rat. J Psychiatr Neurosci 16:221-
226
Brudzynski SM, Bihari F, Ociepa D, & Fu X (1993) Analysis of 22 kHz ultrasonic
vocalizations in laboratory rats: long and short calls. Physiol Behav 54:215-221
Brudzynski SM, Eckersdorf B, Golebiewski H, (1995) Regional Specificity of the
emotional-aversive response by carbachol in the cat brain: A quantitative mapping
study. J Psychiatr Neurosci 20(2): 119-132
Brudzynski SM, Barnabi F (1996) Contribution of the ascending cholinergic pathways in
the production of ultrasonic vocalization in the rat. Behav Brain Res 80:145-152.
Brudzynski SM (2005) Principles of rat communication: quantitative parameters of
ultrasonic calls in rats. Behav Genet 35:85-92
Brudzynski SM, Holland G (2005) Acoustic characteristics of air-puff induced 22-kHz
alarm calls in direct recordings. Neurosci Biobehav Rev 29:1169-1180
Brudzynski SM (2008) Communication of adult rats by ultrasonic vocalization:
Biological, sociobiological, and neuroscience approaches. ILAR J 50: 43-50.
Brudzynski SM, Pniak A (2002) Social contacts and production of 50-kHz
short ultrasonic calls in adult rats. J Comp Psychol 116:73-82
Burgdorf J, Knutson B, Panksepp J (2000) Anticipation of rewarding electrical brain
stimulation evokes ultrasonic vocalization in rats. Behav Neurosci 114:320-327
Burgdorf J, Knutson B, Panksepp J, Ikemoto S (2001) Nucleus accumbens amphetamine
microinjections unconditionally elicit 50-kHz ultrasonic vocalizations in rats.
Behav Neurosci 115: 940-944
46
Burgdorf J, Kroes RA, Moskal JR, Pfaus JG, Brudzynski SM, Panksepp SM (2008).
Ultrasonic vocalizations of rats (Rattus norvegicus) during mating, play, and
aggression: Behavioral concomitants, relationship to reward, and self-
administration of playback. J Comp Psychol 122(4): 357-67.
Childress AR, Ehrman RN, Wang Z, Li Y, Sciortino N, Hakun J et al. (2008) Prelude to
passion: limbic activation by “unseen” drug and sexual cues. PLoS One 3: e1506
Ciucci MR, Ma ST, Fox C, Kane JR, Ramig LO, Schallert T (2007) Qualitative changes
in ultrasonic vocalization in rats after unilateral dopamine depletion or
haloperidol: a preliminary study. Behav Brain Res 182:284-289
Covington HE III, Miczek KA (2003) Vocalizations during withdrawal from opiates and
cocaine: possible expressions of affective distress. Eur J Pharmacol 467:1-13
Fanselow MS, Lester LS (1988) A functional behavioristic approach to aversively
motivated behavior: predatory imminence as a determinant of the topography of
defensive behavior. In: Evolution and learning (Bolles RC, Beecher MD, eds).
Hillsdale, NJ: Erlbaum.
Fox HC, Axelrod SR, Paliwal P, Sleeper J, Sinha R (2007) Difficulties in emotion
regulation and impulse control during cocaine abstinence. Drug Alcohol Depend
89:298-301
Francis RL (1977) 22-kHz calls by isolated rats. Nature 265:236-238
Fu X, Brudzynski SM (1994) High-frequency ultrasonic vocalization induced by
intracerebral glutamate in rats. Pharmacol Biochem Behav 49:835-841
Gay GR, Inaba DS, Sheppard CS, Newmeyer JA, Rappolt RT (1975). Cocaine: History,
epidemiology, human pharmacology, and treatment. A perspective on a new
debut for an old girl. Clinical toxicology 8: 149-178.
Hodgins DC, el-Guebaly N, Armstrong S (1995) Prospective and retrospective reports of
mood states before relapse to substance use. Journal of Consulting and Clinical
Psychology 63(3): 400-407
Kalivas PW, Volkow ND (2005) The neural basis of addiction: A pathology of
motivation and choice. Am J Psychiatry 162: 1403:1413.
Kelley AE, Berridge KC (2002) The neuroscience of natural rewards: Relevance to
addictive drugs. J Neurosci 22(9): 3306-3311.
Knutson B, Burgdorf J, Panksepp J (2002) Ultrasonic vocalizations as indices of affective
states in rats. Psychol Bull, 128:961-977
Knutson B, Burgdorf J, Panksepp J (1998) Anticipation of play elicits high-frequency
ultrasonic vocalizations in young rats. J Comp Psychol 112:65-73
Knutson B, Burgdorf J, Panksepp J (1999) High frequency ultrasonic vocalizations index
conditioned pharmacological reward in rats. Physiology & Behavior 66: 639-643
Koob GF, Caine B, Markou A, Pulvirenti L, Weiss F (1994) Role for the mesocortical
dopamine system in the motivating effects of cocaine. NIDA Research
Monograph 145:1-18
Koob GF, LeMoal M (1997) Drug Abuse: Hedonic homeostatic dysregulation. Science
278:52-58.
Koob GF, LeMoal M (2008) Addiction and the brain Anti-reward system. Annu Rev
Psychol 59:29-53.
Koob GF (2008) Hedonic homeostatic dysregulation as a driver of drug seeking behavior.
Drug Discov Today Dis Models 5(4):207-215.
47
Koob GF (2009) Dynamics of neuronal circuits in addiction: Reward, antireward and
emotional memory. Pharmacopsychiatry 42:S32-S41.
Koob GF (2009b) Neurobiological substrates for the dark side of compulsivity in
addiction. Neuropharmacology 56:18-31.
Levinthal CF (2010) Drugs Behavior and Modern Society. Boston, MA: Allyn & Bacon
Litvin Y, Blanchard DC, Blanchard RJ (2007) Rat 22 kHz vocalizations as alarm cries.
Behav Brain Res 182:166-172.
McKay JR, Rutheford MJ, Alterman AI, Cacciola, JS, Kaplan MR (1995) An
examination of the cocaine relapse process. Drug Alcohol Depend 38:35-43
Mendelson JH, Mello NK (1996) Management of cocaine abuse and dependence. Drug
Therapy 334: 965-972
Mu P, Fuchs T, Saal DB, Sorg BA, Dong Y, Panksepp J (2009) Repeated cocaine
exposure induces sensitization of ultrasonic vocalizations in rats. Neurosci Lett
453:31-35
Mutschler NH, Miczek KA (1998a) Withdrawal from IV cocaine “binges” in rats:
ultrasonic distress calls and startle. Psychopharmacology 135:161-168
Mutschler NH, Miczek KA (1998b) Withdrawal from a self-administered or non-
contingent cocaine binge: differences in ultrasonic distress vocalizations in rats.
Psychopharmacology 136:402-408
Newcomb MD, Felix-Ortiz M (1992) Multiple protective and risk factors for drug use
and abuse: Cross-sectional and prospective findings. Journal of personality and
social psychology 63(2): 280-296
[NIDA] National Institute on Drug Abuse (US). 1985. Cocaine use in America:
Epidemiologic and clinical perspectives. Research monograph series. Rockville
(MD): Department of health and human services (US). DHHS Pub. No.
(ADM)90-1414.
[NIDA] National Institute on Drug Abuse (US). 2009. Cocaine: Abuse and addiction.
Research report series. Rockville (MD): National institutes of health (US). NIH
Pub. No. 09-4166.
Norman AB, Tsibulsky VL (2006) The compulsion zone: a pharmacological theory of
acquired cocaine self-administration. Brain Res 1116:143-152
Office of Applied Studies (2009). Results from the 2008 National Survey on Drug Use
and Health: National findings (DHHS Publication No. SMA 09-4434, NSDUH
Series H-36). Rockville, MD: Substance Abuse and Mental Health Services
Administration.
Ritz MC, Lamb RJ, Goldberg SR, Kuhar MJ. (1987). Cocaine receptors on dopamine
transporters are related to self-administration of cocaine. Science. 237:1219-23.
Robinson TE, Berridge KC (1993) The neural basis of drug craving: an incentive
sensitization theory of addiction. Brain Res Brain Res Rev 18(3): 247-291
Robinson TE, Berridge KC (2001) Incentive sensitization and addiction. Addiction 96(1):
103-114
Takahashi N, Kashino M, Hironaka N (2010). Structure of Rat Ultrasonic Vocalizations
and its relevance to behavior. PLoS One 5(11):e14115. Doi:10.1371/journal.pone.
0014115
48
Thompson B, Leonard KC, Brudzynski SM (2006) Amphetamine-induced 50 kHz calls
from rat nucleus accumbens: a quantitative mapping study and acoustic analysis.
Behav Brain Res 168:64-73
Tonoue T, Ashida Y, Makino H, Hata H (1986) Inhibition of shock-elicited ultrasonic
vocalizations by opioid peptides in the rat: a psychotropic effect.
Psychoneuroendocrinology 11: 177-184
Volkow ND, Wang GJ, Fischman MW, Foltin RW, Fowler JS Abumrad NN et al. (1997)
Relationship between subjective effects of cocaine and dopamine transporter
occupancy. Nature 386:827-830.
Wheeler RA, Twining RC, Jones JL, Slater JM, Grigson PS, Carelli RM (2008)
Behavioral and electrophysiological indices of negative affect predict cocaine
self-administration. Neuron 57:774-785
Wintink AJ, Brudzynski SM (2001) The related roles of dopamine and glutamate in the
initiation of 50-kHz ultrasonic calls in adult rats. Pharmacol Biochem Behav
70:317-323
Wikler, A (1948) Recent progress in research on the neurophysiologic basis of morphine
addiction. Am J Psychiatry 105:329-338.
Williamson S, Gossop M, Powis B, Griffiths P, Fountain J, Strang J (1997) Adverse
effects of stimulant drugs in a community sample of drug users. Drug and Alcohol
Dependence 44: 87-94.
Wise RA, Newton P, Leeb K, Burnette B, Pocock D, Justice JB Jr (1995) Fluctuations in
nucleus accumbens dopamine concentration during intravenous cocaine self-
administration in rats. Psychopharmacology (Berl) 120(1): 10-20.
Wöhr M, Houx B, Schwarting RK, Spruijt B (2008). Effects of experience and context on
50-kHz vocalizations in rats. Physiol Behav 93:766-76.
Wright JM, Gourdon JC, Clarke PBS (2010) Identification of multiple call categories
within the rich repertoire of adult rat 50-kHz ultrasonic vocalizations: effects of
amphetamine and social context.
49
9. Appendices
9.1 Call Type Analysis Post-Hoc Tests:
9.1.1—Main effect for Call type. Tukey-Kramer post-hoc tests for the main
effect of Call Type revealed that there was a significantly greater number of FF calls than
both FMs [t(245)=2.49, p<.05] and Trills [t(245)=3.63, p<0.001]. However, there were
no differences in the numbers of FMs or trills (p=0.49, N.S.). The main effects of
Condition and Call Frequency were not analyzed individually due to their involvement in
the significant Condition x Call Frequency interaction.
9.1.2—Condition x Call Frequency: Within condition comparisons. A number
of within-condition comparisons were made in the post-hoc analysis of the Condition x
Call Frequency interaction. The purpose of these comparisons was to examine
differences in the number of calls emitted in various frequency ranges, while holding
levels of condition constant (Figure 13). The results of these comparisons demonstrated
that animals in the PSD condition exhibited no differences between 22- and 50-kHz calls
(p=0.99, N.S.) or between 50- and 33-kHz calls (p=0.38, N.S.). However, there was a
significant difference between the number of 22- and 33-kHz calls [t(245)=3.64, p<0.05].
For subjects in the PSN condition, there were no differences in the numbers of 50- and
22-kHz calls (p=1.00, N.S.), 50- and 33-kHz calls (p=0.28, N.S.) or 22- and 33-kHz calls
(p=0.22 N.S.). For both the low and high clamp conditions, there were no differences in
the amount of 50- and 22-kHz calls emitted (LC: p=0.65, N.S.; HC: p=0.30, N.S.), but in
both conditions there were significantly more 50-kHz[LC: t(245)=8.26, p<0.0001; HC:
t(245)=8.00, p<0.0001] and 22-kHz calls[LC: t(245)=10.32, p<0.0001; HC: t(245)=5.43,
p<0.0001] than 33-kHz calls.
50
9.1.3—Condition x Call Frequency: Between Condition Comparisons. A set
of “between-condition” comparisons were made using the post-hoc comparisons for the
Condition x Call Frequency interaction. These comparisons were intended to examine
differences between conditions using equivalent levels of Call Frequency. The results
showed that animals emitted more 50-kHz USVs during the LC and HC conditions than
during the PSD [LC vs. PSD: t(245)=8.05, p<0.0001; HC vs. PSD: t(245)=6.88,
p<0.0001] and PSN [LC vs. PSN: t(245)=7.65, p<0.0001; HC vs. PSN: t(245)=6.48,
p<0.0001] conditions (Figure 13C). Conversely, there were no observable differences in
the number of 50-kHz USVs emitted when comparing the HC and LC conditions
(p=0.99, N.S.) or the PSN and PSD conditions (p=1.00, N.S.). Comparisons within the
33-kHz frequency range suggested there were no observable differences between
conditions (all p>0.52, N.S.; Figure 13B). Finally, animals in the LC condition emitted
more 22-kHz USVs than subjects in any other condition [PSD: t(245)=8.92, p<0.0001;
PSN: t(245)=9.60, p<0.0001; HC: t(245)=5.81, p<0.0001]. Also, subjects emitted more
22-kHz USVs during the HC condition than during the PSN condition [t(245)=3.79,
p<0.01]. However, no differences in 22-kHz calls were observed between the HC and
PSD (p=0.84) or PSD and PSN (p=0.99) conditions (Figure 13A).
51
10. Tables and Figures
Table 1 Condition x Call Frequency x Call Type: Mean ± SEM
PSD PSN HC LC
22-kHz USVs
FF 5.00 ± 1.95 6.43 ± 2.05 12.00 ± 3.67 54.5 ± 26.12
FM 2.25 ± 1.44 2.57 ± 0.78 6.88 ± 3.54 61.00 ± 34.22
TRILL 1.63 ± 0.98 0.29 ± 0.18 6.5 ± 3.96 44.75 ± 23.71
33-kHz USVs
FF 0.25 ± 0.25 1.00 ± 0.69 3.75 ± 2.06 4.25 ± 2.15
FM 0.25 ± 0.16 0.29 ± 0.18 0.25 ± 0.25 1.13 ± 0.58
TRILL 0.5 ± 0.27 0.29 ± 0.18 0.13 ± 0.13 0.5 ± 0.27
50-kHz USVs
FF 1.25 ± 0.49 1.43 ± 0.65 27.25 ± 12.94 29.13 ± 11.69
FM 2.25 ± 0.67 2.00 ± 0.69 19.13 ± 8.49 26.88 ± 9.77
TRILL 1.5 ± 0.60 3.43 ± 1.13 17.5 ± 15.10 19.88 ± 8.53
*p<0.05
**P<0.01
52
Table 2 Condition x Call Frequency x Session Time: Mean ± SEM
Hour
PSD PSN HC LC
22-kHz USVs
1 4.69 ± 1.52 1.34 ± 0.43 18.32 ± 5.94 22.33 ± 7.24
2 1.52 ± 0.49 2.82 ± 0.92 6.44 ± 2.09 24.93 ± 8.09
3 1.80 ± 0.58 2.76 ± 0.90 2.69 ± 0.87 36.24 ± 11.76
4 3.22 ± 1.05 2.11 ± 0.69 1.02 ± 0.33 17.93 ± 5.82
5 1.94 ± 0.63 2.41 ± 0.78 1.43 ± 0.46 19.35 ± 6.28
6 1.71 ± 0.56 2.30 ± 0.75 1.02 ± 0.33 19.90 ± 6.46
33-kHz USVs
1 1.02 ± 0.33 1.16 ± 0.38 3.16 ± 1.03 4.42 ± 1.43
2 1.02 ± 0.32 1.02 ± 0.33 1.49 ± 0.48 1.61 ± 0.52
3 1.16 ± 0.38 1.24 ± 0.40 2.10 ± 0.68 2.01 ± 0.65
4 1.02 ± 0.33 1.45 ± 0.47 1.24 ± 0.40 1.4 ± 0.45
5 1.46 ± 0.47 1.02 ± 0.33 1.13 ± 0.37 1.13 ± 0.37
6 1.35 ± 0.44 1.41 ± 0.46 1.13 ± 0.36 1.11 ± 0.36
50-kHz USVs
1 1.74 ± 0.57 1.43 ± 0.46 53.26 ± 17.27 65.60 ± 21.28
2 1.47 ± 0.48 1.43 ± 0.47 1.64 ± 0.53 1.91 ± 0.62
3 2.39 ± 0.78 2.16 ± 0.70 1.55 ± 0.50 3.90 ± 1.27
4 1.54 ± 0.50 2.15 ± 0.70 1.33 ± 0.37 1.76 ± 0.57
5 2.11 ± 0.68 2.29 ± 0.74 1.35 ± 0.44 1.43 ± 0.47
6 1.51 ± 0.49 2.40 ±0.78 0.94 ± 0.33 1.46 ± 0.47
*p<0.05
**P<0.01
53
Figure 1
Figure 1. Examples of ultrasonic vocalizations during cocaine self-administration. The
X-axis is time (ms) and the Y-axis is frequency (kHz). The visible noise-band in the
lower portion of each figure is the audible white-noise present during all sessions. A) A
short fixed-frequency USV in the frequency range of 22-kHz. B) A fixed frequency USV
in the range of 50-kHz. C) A frequency modulated USV in the range of 22-kHz D) A
USV with multiple pitch modulations (trill) in the range of 50-kHz.
54
Figure 2
Figure 2. Cumulative daily drug intake for both the high dose (dark circles) and low dose
(open circles) cocaine self-administration groups. The X-axis shows all sessions from the
start of training until session 11, when the earliest USV recordings occurred. The average
amount (mg/kg) of cocaine consumed daily ± SEM is shown on the Y-axis.
55
Figure 3
Figure 3. Distribution of ultrasonic vocalizations as a function of duration for both the
low (A) and high (B) dose groups. Median frequency (kHz) is represented on the X-axis,
USV duration is shown on the Y-axis (non linear scale reduces overlap of data points).
Open circles represent fixed frequency (FF) calls, while the dark circles represent both
frequency modulated (FM) calls and trills.
56
Figure 4
Figure 4. The average number of USVs ± SEM for both the low (open bars) and high
(dark bars) dose groups during a 6-hour cocaine self-administration binge. Frequencies
(kHz) on the X-axis are separated into the 22-kHz range (18.0-32.9 kHz), the 50-kHz
range (38.1-80.0) and the range between the 22 and 50-kHz distributions (33.0-38.0
kHz). The total number (counts) of USVs is shown on the Y-axis. All call types (FF, FM,
and Trill) are combined in the present figure.
57
Figure 5
Figure 5. A custom built Plexiglas apparatus for increasing the signal : noise in USV
recordings. The acrylic tube is placed over a series of holes on the roof of the operant
chamber, proximal to the response manipulandum.
58
Figure 6
Figure 6. The average number of lever presses emitted by subjects across training
sessions. The number of level presses per session is shown on the Y-axis, while
individual sessions are depicted on the X-axis. All subjects increased the number of
responses over weeks with the most pronounced changes in learning and response rate
occurring during the first week.
59
Figure 7
Figure 7. Total drug consumed (mg/kg/session). The amount of drug consumed per
session is shown on the Y-axis, while sessions are shown on the X-axis. Typical of long-
access models of cocaine-self administration, subjects escalated their intake of cocaine
over sessions.
60
Figure 8
Figure 8. The mean number of infusions earned per session. The number of infusions
per session is depicted on the Y-Axis, while individual sessions are represented on the X-
axis. Coincident with the increases in drug intake across sessions, the number of earned
infusions also increased across weeks.
61
Figure 9
Figure 9. The percent of drug opportunities periods where an infusion was earned. The
percentage of opportunities obtained is depicted on the Y-axis, while individual sessions
are depicted on the X-axis. In correspondence with the increased number of infusions
and increased drug intake, the percent of opportunities where an infusion was obtained
also increased over sessions.
62
Figure 10
Figure 10. Subjects Bodyweight (grams) over sessions. Bodyweight in grams is
depicted on the Y-axis, while session is depicted on the X-Axis. Bodyweights for the
high-clamp (HC) and low-clamp (LC) sessions are shown on the right. Consistent with
the effects of psychomotor stimulants, subjects exhibited decreases in bodyweight after
repeated drug use.
63
Figure 11.
Figure 11. Drug curves for both the High and Low Clamp conditions from a
representative animal. The calculated blood level of cocaine is represented on the Y-axis,
while session time is represented on the X-axis. Subjects were allowed to load normally
on cocaine for the first 90 minutes before being transitioned to their High or Low Clamp
level over a 30 minute period. Drug levels were then fixed for the final 240 minutes of
the session (120-360 minutes).
64
Figure 12
Figure 12. Scatter plots of the four USV conditions PSD(A), PSN(B), HC(C) and
LC(D). Frequency (kHz) is represented on the X-axis, while duration (seconds) is
represented on the Y axis. The Y-Axis is shown on a non-linear scale to reduce the
overlap of points.
65
Figure 13
Figure 13. Graphical results of the Condition x Call Frequency x Call Type repeated
measures ANOVA. Call Type is represented on the X-Axes, while the number of USVs
(Counts) is represented on each Y-axis. Separate bars are used to represent each
condition, while separate plots represent the (A) 22-kHz, (B) 33-kHz, and (C) 50-kHz
frequency ranges respectively.
66
Figure 14
Figure 14. Graphical results of the Condition X Call Frequency X Session Time repeated
measured ANOVA. Session time (hours) is represented on the X-axes, while the number
of USVs (counts) is represented on each Y-axis for the (A) 22-kHz, (B) 33-kHz, or (C)
50-kHz ranges. Separate conditions are represented by different bars in each plot.
67
Figure 15
Figure 15. USVs in the Pre and Post Drug period in Hour 1. The number of USVs per
minute (Y-Axis) that occurred prior to the first infusion (Pre-Drug) or after the first
infusion (Post-Drug) are reported for the 22- 33- and 50-kHz ranges (X-Axis).
Significantly greater rates of 22-kHz USVs than 33- or 50-kHz USVs were observed in
the Pre-Drug period, while significantly more 50-kHz USVs were observed during the
Post-Drug period than 22- or 33-kHz USVs.
