"Medicine, Faculty of"@en . "DSpace"@en . "UBCV"@en . "Barr, Alasdair McMillan"@en . "2009-10-01T21:38:06Z"@en . "2001"@en . "Doctor of Philosophy - PhD"@en . "University of British Columbia"@en . "Depression is one of the more frequent psychiatric disorders, that continues to exact a tremendous cost to society, in both human and financial terms. Novel therapies for the treatment of this disorder will originate from preclinical research, which is based upon the use of animal models. In the current dissertation, we describe the development of two animal models of depression. The first three experiments describe in detail a series of studies that were conducted to provide further validation of the Chronic Mild Stress (CMS) model of depression. Despite initially promising results in Experiment 1, in which CMS-treated rats exhibited significant decreases in appetitive investigations for a sucrose solution, the finding of Experiments 2 signified that there were fundamental flaws in this model, as CMS-treated rats failed to display reduced motivation to obtain a similar solution under a progressive ratio schedule of reinforcement. In Experiment 3, in vivo microdialysis in the ventral striatum was used to demonstrate that CMS-treated rats failed to exhibit the hypodopaminergia that is hypothesized to lead to depressogenic behaviours in these animals. The results of Experiments 2 and 3 rendered the CMS paradigm unsuitable for further research and questioned its validity as a legitimate model of depression. In Experiments 4-7, we describe the extensive research that we undertook to validate an alternate rodent model of depression, namely the Psychostimulant Withdrawal model. The results of Experiments 4 indicated a reduced motivation of rats to obtain a sucrose solution under a progressive ratio schedule, while Experiment 5 indicated that this model induced sexual deficits in rats comparable to those seen in unipolar depression; Experiment 6 demonstrated protracted negative contrast effects in drug withdrawn animals, hence suggesting this paradigm is amenable to modeling subtle psychological processes. In Experiment 7, we demonstrate that the model responds to an appropriate therapeutic strategy as rats exhibited an earlier recovery of rewarding brain self-stimulation following repeated electroconvulsive shock. The dissertation commences with a general review of depression and animal models of this disorder; it concludes with an evaluation of the Psychostimulant Withdrawal model of depression, integrated with the findings of the current dissertation."@en . "https://circle.library.ubc.ca/rest/handle/2429/13495?expand=metadata"@en . "10400620 bytes"@en . "application/pdf"@en . "T H E D E V E L O P M E N T OF A N A N I M A L M O D E L OF D E P R E S S I O N : A F O C U S O N ANHEDONIA by ALASDAIR M c M I L L A N BARR B . A . , University of British Columbia, 1994 A THESIS SUBMITTED IN P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E OF DOCTOR. OF P H I L O S O P H Y in T H E F A C U L T Y OF G R A D U A T E S T U D I E S Graduate Program i n Neuroscience We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y OF BRITISH C O L U M B I A August, Z001 \u00C2\u00A9 Ala.sda.ir McMil lan . Barr, 2001 UBC Special Collections - Thesis Authorisation Form Page 1 of 1 I n p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f t h e r e q u i r e m e n t s f o r an advanced degree a t the U n i v e r s i t y o f B r i t i s h C o l u m b i a , I agree t h a t t h e L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and s t u d y . I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y p u rposes may be g r a n t e d by th e head o f my department o r by h i s o r her r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l not be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . Department o f /\)\u00E2\u0082\u00ACOtoSCl&JZ$L The U n i v e r s i t y o f B r i t i s h C olumbia Vancouver, Canada Date 23- JA?K< http://vvww.library.ubc.ca/spcoll/thesauth.html 8/23/01 11 Abstract Depression is one of the more frequent psychiatric disorders, that continues to exact a tremendous cost to society, in both human and financial terms. Novel therapies for the treatment of this disorder will originate from preclinical research, which is based upon the use of animal models. In the current dissertation, we describe the development of two animal models of depression. The first three experiments describe in detail a series of studies that were conducted to provide further validation of the Chronic Mild Stress (CMS) model of depression. Despite initially promising results in Experiment 1, in which CMS-treated rats exhibited significant decreases in appetitive investigations for a sucrose solution, the finding of Experiments 2 signified that there were fundamental flaws in this model, as CMS-treated rats failed to display reduced motivation to obtain a similar solution under a progressive ratio schedule of reinforcement. In Experiment 3, in vivo microdialysis in the ventral striatum was used to demonstrate that CMS-treated rats failed to exhibit the hypodopaminergia that is hypothesized to lead to depressogenic behaviours in these animals. The results of Experiments 2 and 3 rendered the CMS paradigm unsuitable for further research and questioned its validity as a legitimate model of depression. In Experiments 4-7, we describe the extensive research that we undertook to validate an alternate rodent model of depression, namely the Psychostimulant Withdrawal model. The results of Experiments 4 indicated a reduced motivation of rats to obtain a sucrose solution under a progressive ratio schedule, while Experiment 5 indicated that this model induced sexual deficits in rats comparable to those seen in unipolar depression; Experiment 6 demonstrated protracted negative contrast effects in I l l drug withdrawn animals, hence suggesting this paradigm is amenable to modeling subtle psychological processes. In Experiment 7, we demonstrate that the model responds to an appropriate therapeutic strategy as rats exhibited an earlier recovery of rewarding brain self-stimulation following repeated electroconvulsive shock. The dissertation commences with a general review of depression and animal models of this disorder; it concludes with an evaluation of the Psychostimulant Withdrawal model of depression, integrated with the findings of the current dissertation. I V Table of Contents Abstract ii List of Tables viii List of Figures ix Acknowledgements x INTRODUCTION 1 Economic cost of depression 2 Epidemiology of depression 4 Genetics of Depression 7 Non-Biological theories of depression 9 Biological theories of depression 11 Animal models of depression 18 Current animal models of depression 25 Animal models of depression - synthesis 32 CHRONIC MILD STRESS 33 EXPERIMENT 1: EFFECTS OF CHRONIC MILD STRESS ON A N APPETITIVELY MOTIVATED TASK Introduction 38 Methods 39 Results 41 Discussion 45 EXPERIMENT 2: CHRONIC MILD STRESS HAS NO EFFECT ON RESPONDING BY RATS FOR SUCROSE UNDER A PROGRESSIVE RATIO SCHEDULE Introduction 49 Methods 50 Results 56 Discussion 62 EXPERIMENT 3: CHRONIC MILD STRESS MODIFIES E X T R A C E L L U L A R L E V E L S OF DOPAMINE AFTER CONSUMPTION OF A SUCROSE SOLUTION Introduction 66 Methods 68 Results 71 Discussion 74 CHRONIC MILD STRESS - SYNTHESIS AMPHETAMINE WITHDRAWAL 77 79 EXPERIMENT 4: WITHDRAWAL FOLLOWING REPEATED EXPOSURE TO D-AMPHETAMINE DECREASES RESPONDING FOR A SUCROSE SOLUTION AS MEASURED BY A PROGRESSIVE RATIO SCHEDULE OF REINFORCEMENT Introduction 83 Methods 85 Results 89 Discussion 98 EXPERIMENT 5: EFFECTS OF WITHDRAWAL FROM A N ESCALATING DOSE SCHEDULE OF D-AMPHETAMINE ON SEXUAL BEHAVIOR IN THE M A L E RAT Introduction 105 Methods 107 Results 109 Discussion 116 EXPERIMENT 6: INCREASED SUCCESSIVE NEGATIVE CONTRAST IN RATS WITHDRAWN FROM A N ESCALATING-DOSE SCHEDULE OF D-AMPHETAMINE vii Introduction 122 Methods 124 Results 127 Discussion 130 EXPERIMENT 7: REPEATED ELECTROCONVULSIVE SHOCK A T T E N U A T E S THE DEPRESSIVE EFFECTS OF 73-AMPHETAMINE WITHDRAWAL ON BRAIN REWARD FUNCTION IN RATS Introduction 135 Methods 137 Results 141 Discussion 146 G E N E R A L DISCUSSION 151 Psychostimulants and psychostimulant withdrawal in humans 155 Treatment of psychostimulant withdrawal in humans 162 Psychostimulant withdrawal as a model of depression in rodents 167 Limitations 184 Future Directions 186 REFERENCES 188 List of Tables Table 2.1 Schedule of events (Protocol for CMS experiment) 55 Table 2.2. Free consumption values of sucrose solution in CMS paradigm 59 Table 4.1 The effect of c/-amphetamine withdrawal upon the Inter-Lick-Intervals taken by rats when consuming a 0.50ml sucrose solution reinforcer (50 licks per Reinforcement) 92 Table 4.2 The effects of drug treatment on both the Post-Reinforcement Pause and the Time to attain each reinforcement, for the first five reinforcements 94 List of Figures Figure 1.1 Effects of CMS upon levels of levels of appetitive and consummatory responding for a 1% sucrose solution 44 Figure 2.1 Effects of changes in sucrose concentration on final ratios (i.e. break points) attained under a progressive ratio schedule by control rats 57 Figure 2.2 Effects of chronic mild stress on final ratios attained under a progressive ratio schedule, as a function of sucrose concentration (7% or 1%) or hours of food and water deprivation (20 Hr or 12 Hr) 60 Figure 2.3 Effects of chronic mild stress on body weight measured weekly 61 Figure 3.1 (A) Effects of exposure to chronic mild stress for 6 weeks on levels of dopamine and metabolites after consumption of a small fixed volume (5mL) of a 1% sucrose solution (B) Probe placements 73 Figure 4.1 The effect of ^-amphetamine withdrawal on responding for a 4% sucrose solution under a progressive ratio schedule of reinforcement, across different test Sessions 91 Figure 4.2 Body weights of subjects 95 Figure 4.3 The effect of ^-amphetamine on free consumption of a 4% sucrose solution from a lickometer, including the four baseline sessions prior to drug administration 97 Figure 5.1 The effect of withdrawal from an escalating dose schedule of J-amphetamine on anticipatory locomotor activity I l l Figure 5.2 The effects of withdrawal from an escalating dose schedule of d-amphetamine on different temporal measures of sexual behavior in male rats 114 Figure 5.3 The effects of withdrawal from an escalating dose schedule of d-amphetamine on different copulatory measures in male rats 115 Figure 6.1 Effects of withdrawal from a 4-day regimen of ^-amphetamine, or vehicle on number of licks for a 4% sucrose solution 129 Figure 7.1 Effects of withdrawal from a 4-day escalating dose schedule of d-amphetamine or vehicle on ICSS responding 143 Figure 7.2 Histological placements of stimulating electrodes located within the lateral hypothalamus 145 X Acknowledgements The better parts of this thesis are dedicated to the memory of my father, Douglas Maxwell Barr: \"By Jove, I am not covetous for gold, Nor care I who doth feed upon my cost; It yearns me not if men my garments wear; Such outward things dwell not in my desires: But if it be a sin to covet honour, I am the most offending soul alive \". Henry V, Act 4, Scene 3. 1 Introduction: As new cures for old diseases are constantly sought, the generation of novel and improved therapies for the treatment of human disorders is a process that is fundamentally based upon the use of animal models. These models have a variety of different functions, sometimes being used to test drugs with putative therapeutic qualities, 'and other times they are used to provide insight into the disease itself. The modeling of psychiatric disorders poses a particular challenge, given the disparate development of the brain between humans and lower mammals, such as rats and mice. Consequently, especial care and attention must be given to the propagation of such models. With a disorder such as depression, in which some of the most \"human\" of experiences are affected, there must be an attempt to span both biological and psychological boundaries. It is not an easy task. The purpose of this dissertation has been the development and validation of an animal model of depression. The dissertation begins with a review of Major Depressive Disorder (MDD), encompassing its prevalence and possible causes. After the importance of this disorder has been described, there is a brief summary of the \"rules\" by which animal models of psychiatric should be developed. Following this is a review of some of the strengths and weaknesses of the more popular animal models of depression that are currently in use. At this point, the dissertation then describes in detail the experiments that were conducted to validate the psychostimulant withdrawal model of depression. 2 Economic Cost of Depression: M D D is one of the leading causes of worldwide disease burden. According to several recent studies of premature mortality and worldwide disability, MDD currently ranks in the top five disease burdens (Murray and Lopez 1997, Holden 2000, Hyman 2000). The World Health Organization (WHO) predicts that MDD will be second only to heart disease by 2020 in terms of economic burden, as mortality from infectious disease declines, and will be the leading cause of worldwide disability in terms of the number of people affected (Holden 2000). The economic impact of MDD is particularly pervasive, due to the chronic nature of the illness and the relatively low rate of mortality when compared to other major diseases, such as cardiovascular disease and cancer. Unlike these other two diseases, MDD afflicts people mostly in the age group 25-44, which often coincides with the most highly productive years of employment (Lane and McDonald 1994). For women, who are diagnosed with MDD approximately three times more often than men, these years represent important childbearing years, suggesting future generational costs to society. M D D is associated with greater morbidity from almost all physical illness, including respiratory problems and cardiovascular disease (Sims 1988). One study showed that medical inpatients with clinically significant depression had more procedures performed, stayed in hospital for 40% longer, and incurred 35% greater hospital costs (Levenson et al. 1990). The majority of the economic burden of depression, however, arises from lost human productivity. The true economic impact of MDD has only become apparent more recently, as emphasis has shifted away from mortality as being the traditional measure of disease burden (Hyman 2000). The WHO, amongst others, has pioneered the use of a measure called the Disability Adjusted 3 Life Year (DALY), which corresponds to healthy years of life lost to disability and premature mortality (Murray and Lopez 1996). Using this measure, MDD ranks second only to cardiovascular disease in terms of lost DALYs, and far ahead of other psychiatric and neurological conditions such as schizophrenia, Alzheimer's disease and alcohol abuse (Hyman 2000). Actual dollar estimates of the financial burden of MDD are difficult to obtain, due to the many \"hidden\" costs of the disorder. Greenberg et al. (1993) reported that the direct costs for the USA resulting from MDD in 1990, including hospital and primary care costs, were US$12.4 billion per year. The indirect costs in the same year, which reflect lost earnings, were at least US$43.7 billion per year. These numbers also fail to account for other indirect costs of MDD, such as higher rates of substance abuse amongst clinically depressed individuals (Markou et al. 1998) that may incur societal costs of many additional billions of dollars. Similar proportional costs have been observed in other countries, such as Australia, where MDD was found to be the top-ranking cause of non-fatal disease burden (Mathers et al. 2000). In Britain, an estimated 80 million working days per year are lost through mental illness (Creed 1993), of which the majority result from MDD. The relative economic impact of MDD may be even higher in some countries, such as in areas of intense conflict; for example, extremely high rates of depression were reported in Lebanon during the early 1990s (Karam et al. 1998). The staggering economic worldwide burden of MDD underscores the importance of understanding this disorder, and generating effective pharmacological compounds for its treatment. 4 Epidemiology of Depression M D D is probably the most common of the psychiatric disorders (Fava and Kendler 2000). Estimates for lifetime prevalence of the disorder, based upon large, community-based surveys, have yielded figures in the range of 15-17% for the United States (Kessler et al 1994). These figures appear to be consistent for most western countries, including Canada (Bland et al. 1988) , New Zealand and Australia (Joyce et al. 1990; Mathers et al. 2000) and Western Europe. Slightly lower figures have been obtained for certain Asian countries, such as Taiwan (Hwu et al. 1989) and South Korea (Lee et al. 1987), although it has been suggested that this may reflect cultural idiosyncrasies in the expression of MDD or the use of different screening devices (Holden 2000). A large, worldwide study conducted by the WHO plans to complete a multinational survey of over 150,000 people in North America, Western Europe, Mexico, Chile, Cuba, Columbia, Ukraine, South Africa, India, China, Japan, Indonesia and New Zealand by mid-2001 (Holden 2000). As the survey seeks to conduct interviews in a local, culturally valid manner, it is hoped that accurate statistics will be obtained for worldwide prevalence of the disorder. The annual prevalence of major depression in North America remains higher than for any other psychiatric disorder, and the National Comorbidity Study noted that nearly 5% of the population reported meeting criteria for MDD in the previous 30 days (Blazer et al. 1998). Of those who are diagnosed with MDD, approximately 15% will eventually commit suicide (APA 1994). M D D is also a chronic, relapsing disorder and it is believed that almost 80% of people who are diagnosed with depression will experience a second or further episode (Mueller and Leon 1996). In addition, evidence from several sources suggests that the average age of onset of 5 M D D has decreased, and the overall risk of suffering from a depressive disorder has increased, over each of the past several generations (Klerman and Weissman 1989; Cross-National Collaborative Group 1992). The epidemiological risk factors associated with depression are extensive (Fava and Kendler 2000). Clearly, one of the most important factors is the gender of the individual; in the National Comorbidity Survey, the lifetime prevalence rate of depression was estimated to be 21.3% in women, and only 12.7%) in men (Blazer et al. 1994). These figures are supported by several other recent North American studies (Regier et al. 1993; Kessler et al. 1994), as well those in other western countries (Piccinelli and Wilkinson 2000). While there are undoubtedly artifactual determinants that may increase the likelihood of gender differences in the measurement of the prevalence of M D D , there is a general consensus that the overall trend of these differences is genuine (Piccinelli and Wilkinson 2000). Aside from gender, there are three other epidemiological risk factors that stand out in the consistency of their association with M D D (Kessler 1997). The first of these, which consists of exposure to stressful life events, is associated with a substantial increase in risk for the onset of M D D . Current data indicate that the nature of the relationship between stress and depression is largely causal (Kendler et al. 1999), as opposed to other interpretations such as simply an increased perception of stress. The most influential stressors include the loss of personal relationships, marital difficulties, job loss and major health problems (Kessler 1997). The second major risk factor for the development of M D D is the experience of childhood adversity (Fava and Kendler 2000). Early childhood experiences, such as family instability, poor mothering, overcrowding and dependence on social welfare are all important factors in determining future vulnerability to M D D (Sadowski et al. 1999; Fergusson et al. 2000). The 6 remaining major risk factor for vulnerability to MDD is certain predisposing personality traits. Psychodynamic theorists, such as Aaron Beck, have long postulated that particular personality types are more vulnerable to the effects of environmental stressors, and current research tends to support this claim. The personality construct \"neuroticism\" has been shown repeatedly to predispose individuals to subsequent episodes of depression (Roberts and Kendler 1999), particularly in women. Several other risk factors have been associated with MDD, such as low levels of social support and urban residence, but the causal association of these factors remains weaker than for the factors described above (Fava and Kendler 2000). 7 Genetics of Depression While environmental factors appear to play a prominent role in the etiology of MDD, it is becoming increasingly clear that genetic factors may be equally as important in the development of the disorder. To date, there have been relatively few studies that have been able to provide accurate estimates of the genetic influence on the development of MDD. The use of large, registry-based twin studies has allowed for the comparison of concordance rates of M D D between monozygotic and dizygotic twins. A large study by Kendler et al. (1993) determined that in women, MDD (when diagnosed using DSM-III-R criteria) was moderately heritable, reflected in a heritability estimate of 0.42. However, the authors of the study concluded that the influence of genetic factors may have been underestimated due to the unreliability of lifetime reports of MDD, and that the true estimate of heritability was closer to a value of 0.70. In support of these data, a more recent study of male twins from the Vietnam Era Twin Registry reported that the additive genetic influence on the development of M D D provided a heritability estimate of 0.47 (Lyons et al. 1998), although this was only for early-onset depression.. It is of interest that the same study determined that late-onset MDD had a much lower heritability estimate (0.10), suggesting the possibility of genetically distinct sub-types of major depression. A recent meta-analysis of twin and adoption studies that included all sub-types of M D D concluded that the heritability of liability to MDD was 33% (Sullivan et al. 2000). This is comparable to the genetic contribution of other important biomedical traits, such as blood pressure and serum cholesterol (Fava and Kendler 2000). The manner of the expression of these genetic factors is likely to be quite complex, though, and unlike that of many other non-psychiatric conditions. For example, it has been hypothesized that genes and environment 8 interact partially to influence the overall risk of illness, but also influence the sensitivity of individuals to the depressive effects of environmental stressors (Kendler et al. 1995). It appears that the manner of inheritance of M D D does not fit models of simple Mendelian inheritance. This suggests that a relatively large number of individual genes are likely to be involved, none of which may themselves have a major impact on the risk of developing M D D (Fava and Kendler 2000). Moreover, these genes could interact with each other in complex interactions, such as during particular developmental phases of the individual. Currently, there is a paucity of data regarding the nature of those genes implicated in the etiology of MDD. This is unlike the situation for Bipolar Depression, in which numerous genetic loci, including 4pl6, 12q23-q24, 16pl3, 21q21 and Xq24-q26 have been indicated as possible regions of linkage (Craddock and Jones 1999). While a recent study provided evidence for a minor genetic contribution to MDD by an allelic variation of the dopamine D3 receptor (Dikeos et al. 1999), most studies have failed to identify regions of genetic linkage to M D D (Balciuniene et al. 1998). Indeed, there has been a surprising inability to provide genetic linkage to genes that would be expected to be involved in the etiology of MDD, such as those involved in the neuroendocrine system (Neiswanger et al. 1998) or the serotonin transporter (Seretti et al. 1999). 9 Non-Biological Theories of Depression While the focus of the current dissertation involves the modeling of various biological attributes of MDD, it is recognized that non-biological theories of the etiology of depression have provided invaluable theoretical frameworks for present views of the disorder. As early as the days of Hippocrates (460-357 BC), references were made to the condition of \"melancholia\", which was described as a state of \"aversion to food, despondency, sleeplessness, irritability and restlessness\" (Freeman 1994). Later Greco-Roman physicians emphasized the balance of the four \"humors\", and the Arab scholar Avicenna speculated that black bile, mixed with phlegm, would lead to illness that was \"coupled with inertia, lack of movement, and quiet\" (Okasha 1999). Galen's temperamental types have often been considered as forerunners of current personality dimensions, which (in certain combinations) can lead to various mood-states (Brink 1979). Late nineteenth and early twentieth century European psychologists, such as Sigmund Freud and Karl Abraham, revolutionized the field of psychiatry by emphasizing the role of psychodynamic, subconscious forces in the individual's mind. Thus, depression was a result of internalized thanatotic impulses that were incompatible with the person's psyche. This inward-turned aggression and ensuing guilt could often be observed as an increased hostility to others, according to these theorists (Wallace 1976). More recent non-biological theories of MDD have tended to focus on cognitive and learning factors as etiological bases for the development of the disorder. In particular, Aaron Beck has emphasized the importance of attributional styles, and how they can lead certain individuals to think with a strongly negative bias (Beck 1971). This cognitive bias leads to a 10 downward spiral of self-reinforcing negativity, and eventually results in a severe mood disturbance (Alloy et al. 1999). Learning theorists, such as Martin Seligman, have preferentially emphasized the role that previous disappointments and failures have on future reward expectancy. The cumulative inability to attain desired goals in certain individuals finally leads to the adoption of a \"learned helplessness,\" whereby action and contingency are no longer associated, and the person simply \"gives up trying\" (Seligman 1972; Abramson et al. 1978). 11 Biological Theories of Depression Monoamine Theory of Depression: As with many other medical disorders, some of the earliest insights into the biological basis of M D D were provided through serendipitous discovery. The observation that sub-groups of depressed patients who were being treated for tuberculosis with the drug iproniazid showed improvement in mood led researchers to discover that the drug had potent properties as a monoamine oxidase inhibitor (MAOI) (Pletscher 1991). At the same time, it was also noted that approximately 15% of patients who were being treated for hypertension with the drug reserpine (Serpasil), which causes depletion of catecholamines in the brain, went on to develop clinical depression (Muller et al. 1955; Goodwin and Bunney 1971). On the basis of these and other data, Joseph Schildkraut proposed the catecholamine hypothesis of M D D in the mid-1960s (Schildkraut 1965). The essence of this theory is that depression arises from abnormally low levels of catecholamines in the brain, with norepinephrine (NE) being the most important of the biogenic amines (Schatzberg and Schildkraut 1994). Almost 40 years later, a strong body of evidence still exists that alterations in the central noradrenergic system play an important, if not always essential, role in the etiology of MDD (Anand and Charney 2000), although it is now less clear whether central NE is deceased or actually increased in unipolar depression. Levels of 3-methoxy-4-hydroxyphenylglycol (MHPG), one of the major metabolites of NE, can be quantified to provide a measure of central NE turnover. However, urinary drug-free measures of MHPG have provided mixed results, with more consistent findings for bipolar disorder, and equivocal results for M D D (Leonard 1997). Nevertheless, plasma measures of N E and its 12 metabolites suggest that this monoamine is reduced in the brain of the majority of unipolar depressives (Grossman and Potter 1999; Lambert et al. 2000); it should be noted, though, that a recent study observed increased in CSF levels of N E in patients with melancholic-type depression (Wong et al. 2000). In support of increased central levels of N E in MDD, several studies have found that patients with MDD exhibit an upregulation of the ci2-adrenoceptor autoreceptor, presumably as a homeostatic compensation to reduce NE neuotransmission (Gonzalez et al. 1994). Additionally, others have reported that ^-adrenoceptors in unipolar depressives exhibit a decreased adenylate cyclase response to specific N E agonists, which would be consistent with a desensitization arising from increased circulating levels of N E (Potter et al. 1993). Given that many antidepressants act to increase levels of NE, primarily through the N E transporter (Frazer 2000; Kent 2000), it appears that the role of NE in MDD is a complex one, and much remains to be resolved in this field of research. In contrast to NE, it is apparent that decreased levels of central serotonin, or 5-hydoxytryptamine, (5-HT) are a major factor in the development of MDD. Multiple independent lines of research support this conclusion. Firstly, imaging and post-mortem studies have indicated a reduced density of the 5-HT transporter (5-HTT) in the brain, consistent with lower levels of central 5-HT (Perry et al. 1983; Malison et al. 1998). Reduced levels of 5-HTTs are also observed in the platelets of unmedicated depressives (Owens and Nemeroff 1994), although recent studies have called into question the relevance of these findings to level of central 5-HTTs (Yatham et al. 2000b). Imaging studies have also reported decreased levels of the 5-HT2 receptor in cortical regions in depressives (Yatham et a l l 999,2000a), but the relationship of this finding to central levels of 5-HT remains undetermined. Levels of 5-hydroxyindoleacetic acid (5-HIAA), the major metabolite of serotonin, are lower in CSF measurements of medication-free 13 unipolar depressives (Csernansky and Sheline 1993), and suggest that central levels of 5HT are lower in MDD. Lower concentrations of 5-HT and 5-HIAA have been reported in postmortem brain tissue of depressive who committed suicide, in agreement with studies that have noted decreased levels of the 5-HT precursor tryptophan in the plasma of depressed patients (Anderson et al. 1990). Hence, tryptophan depletion studies in humans have demonstrated that sub-groups of depressives can be induced to relapse into depression when subjected to acute tryptophan depletion (Leyton et al. 2000), emphasising the importance of physiologically normal levels of 5-HT for stable mood. Finally, many of the most efficacious antidepressant treatments in use today work through increasing serotonin levels in the brain, either directly or indirectly (Mann 1999; Kent 2000). Dopamine (DA), in contrast to N E and 5-HT, has not traditionally been considered an important neurochemical substrate in the etiology of MDD. The reasons for this oversight remain unclear, as ever-increasing data implicate a potentially critical role for D A in unipolar depression (Willner 1995). Given the large volume of evidence obtained from animal studies that emphasises DA's role in reinforcement and reward, as well as the capacity of many direct and indirect D A agonists to increase mood in humans, there is a sound theoretical basis for proposing a major role for DA in the etiology of MDD. Many newer antidepressants, including bupropion and amisulpride, exhibit a pharmacological profile that suggests a primary mode of action through the dopaminergic system. Studies of human depressives have shown conclusively that levels of homovanillic acid (HVA), a metabolite of DA, are decreased in the CSF of these patients (Willner 1983), especially those with psychomotor retardation. 14 Neuropeptide and Neurotrophin Theories of Depression: A number of different neuropeptides have generated recent interest as potential factors in the etiology and treatment of depression. While the entire list is exhaustive, and includes, amongst others, (3-endorphin, neuropeptide Y, cholecystokinin, arginine vasopressin, galanin and somatostatin (Nestler 1998), certain neuropeptides have received particular attention. Primary amongst these neuromodulators is corticotropin-releasing factor (CRF), which is currently the focus of many novel antidepressant drug strategies. CRF, which is secreted by neurons within the hypothalamus, exerts powerful control over the hypothalamic-pituitary adrenal (HPA) axis (Nemeroff 1998), by controlling the release of pituitary neuropeptides. This axis is disrupted in one third to one half of all depressives, and several recent studies have shown that CSF levels of CRF are elevated in sub-groups of patients with MDD, particularly those with a history of early childhood trauma (Plotsky et al. 1998; Heim et al. 2000). Animal studies have shown that central administration of CRF induces depressive-like symptoms, and recently-developed CRF antagonists display an antidepressant-like profile in various pre-clinical paradigms (Macey et al. 2000). Alternate neuropeptides that have recently become a popular focus of research include substance P and a variety of different neurotrophic factors. The discovery that substance P antagonists exhibit antidepressant properties (Kramer et al. 1998) has led to a large number of recent studies that have sought to understand how neurokinin receptors might be involved in MDD. The importance ascribed to various neurotrophic factors in the etiology of M D D is derived largely from animal studies, where factors such as BDNF (and TrkB, its receptor) and NT-3 induce regeneration of 5-HT fibers in the brain (Mamounas et al. 1996), and also display 15 an antidepressant profile in pre-clinical tests (Siuciak 1997). However, there is clearly a need for postmortem studies of the levels of these neurotrophins in the brains of depressed individuals before firm conclusions can be drawn about the role of these factors in the etiology of depression. Neuroimmunological Theories of Depression: Despite being a relatively new field of neuroscience, psychoneuroimmunology has already contributed a large body of evidence that implicates the immune system in the etiology of MDD. Numerous findings have converged to suggest that the excessive release of proinflammatory cytokines in both humans and animals can lead to depressive-like behaviour. In humans, the treatment of cancer patients with cytokines such as interleukin-2 (IL-2) and interferon-alpha (IFNa) can lead to depressive-like symptoms (Meyers 1999), while these same cytokines lead to anhedonia in rodent paradigms of reward-related responding (Anisman and Merali 1999). Depressed humans exhibit numerous signs of immune activation and suppression (Dantzer et al. 1999), and there is consistent evidence for a disturbance in the balance of IL-1 (3 and IL-1 receptor antagonist in MDD (Licinio and Wong 1999). It is also of interest that antidepressants, such as clomipramine and sertraline, are able to inhibit the stimulated production of the inflammatory mediator interferon-gamma (IFN-y), and concurrently increase the production of the anti-inflammatory cytokine IL-10 (Maes et al. 1999). 16 Alternate Theories of Depression: Since the discovery of the first effective therapeutic treatments for depression, there have been many theories put forward to explain the biological origin of MDD. A review of all of these theories is certainly beyond the scope of the current dissertation, but it is pertinent to mention a few of them. One of the foremost theories is that alterations in the hypothalamic-pituitary-thyroid axis (HPT) may contribute to unipolar depression in certain individuals, and it is well established that treatment with triiodothyronine (T3) can increase the efficacy of tricyclic antidepressants (Joffe and Levitt 1993). Similarly, there is some convincing evidence for a disruption of the hypothalamic-pituitary-growth hormone (HPG) axis in MDD, which includes a decreased G H response after a challenge with the adrenergic drug clonidine. With the advent of neuroimaging techniques, more recent theories have sought to emphasise that M D D may be expressed as a result of imbalances in regional brain function. For example, it has been hypothesised that a disruption in the balance of the two halves of the frontal cortex may lead to changes in mood and psychiatric symptoms (Goldberg and Podell 1995). PET and fMRI studies indicate that unipolar depression is frequently associated with regionally increased glucose metabolism in the orbitofrontal cortex, amygdala and possibly the anterior cingulate, with corresponding decreases in the dorsolateral prefrontal cortex and basal ganglia (Kennedy et al. 1997). Many of the regional brain alterations in metabolic activity are reversed by treatment with conventional antidepressants or electroconvulsive therapy (ECT), which suggests that such changes may be state-dependent. Molecular theories of MDD have become prominent in the last decade, based largely upon the improvement of various laboratory assay techniques. Instead of focussing upon 17 specific neurotransmitters, some have hypothesised that the primary biological deficit in M D D may lie at the level of signaling pathways, and that antidepressants exert their heterogeneous effects through the transcription factor cyclic adenosine 3', 5'-monophosphate (cAMP) and cAMP-dependent response-element-binding protein (CREB) (Baker and Greenshaw 1989; Duman et al, 1997). Other groups have applied an interdisciplinary approach and sought to combine recent advances in the understanding of chronobiology with the body of knowledge concerning disruption of circadian rhythms in MDD. To date, this novel field of research has failed to identify specific genes or proteins that are implicated in the etiology of MDD, but there are solid theoretical grounds for further research. The recent discovery that specific regions of the adult primate brain continue to produce post-mitotic neurons has led to a novel theory of depression. The essence of this theory is that sustained stress may lead to decreased neurogenesis in the hippocampus, with resulting cognitive deficits and HPA-axis abnormalities (Jacobs et al. 2000); serotonergic antidepressants are able to increase levels of hippocampal neurogenesis and therefore help to reverse these effects. Future theories of the biological basis of MDD are likely to evolve, based upon the large amount of data that are certain to be obtained from genomic and proteomic studies (Flanigan and Leslie 1997). The almost overwhelming quantity of information from these studies will require the use of sophisticated animal models of depression to assist in the interpretation of these data. 18 A n i m a l Models of Depression The term \"animal model of depression\" is frequently a misnomer, for several reasons. The first of these, is that many so-called \"animal models\" of depression are not modeling depression at all, but rather fulfil entirely another function. In addition, if these animal paradigms do attempt to simulate major depression, then they tend only to focus on a specific symptom (and by definition, MDD is recognised as a syndrome that consists of multiple co-presenting symptoms). The list of \"animal models of depression\" has grown to include more than 20 experimental paradigms that are currently in use (Willner 1991). The most common use of animal models of depression is in the pharmaceutical industry, where they are used to screen for novel compounds with antidepressant properties. In these cases, the models may or may not be required to mimic depression, but whether they do so or not is often considered irrelevant, as it is only their capacity to discriminate putative antidepressants that is considered pertinent. For example, two of the most frequently used rodent models of depression are the forced swimming test (FST) and the tail suspension test (TST) (Dalvi and Lucki 1999). The FST, also known as the \"behavioural despair test\" and the \"Porsolt test\", is probably the most well known and ubiquitous antidepressant screen in use (Lucki 1997). The test consists of placing either rats or mice in a cylindrical container that is filled with lukewarm water to a depth that requires the animal to swim to maintain its head above the surface of the water. The animal is placed in the container for approximately ten minutes and its behaviour is monitored and scored, according to certain criteria (Lucki 1997). Preadministration of the antidepressant compound will typically increase the amount of time that the animal engages in \"active behaviors\", compared to vehicle-19 treated subjects. The FST is a highly effective screen for antidepressants (particularly when combined with an open field test), with over 90% accuracy, yet this paradigm provides us with next to no information about the etiology of MDD in humans. Similarly, the tail suspension test has been used successfully in the detection of numerous antidepressant compounds (Dalvi and Lucki 1999). The TST is simple to administer, and is readily reproducible in numerous laboratories. However, the limitations of this model do not extend beyond a capacity to suggest potential antidepressant properties in a drug. While there is clearly an essential need for these types of pharmacological screens, the use of such paradigms is unlikely to increase our knowledge of the biochemical basis of MDD. Pharmacological Validity: It has therefore been suggested that before an animal paradigm can be considered a true \"model\" of a psychiatric disorder, it should be able to meet certain criteria (McKinney and Bunney 1969; Willner 1986; Geyer and Markou 1995). The most frequently applied criterion to an animal model of MDD is that it be able to detect drugs with the relevant therapeutic potential. Typically, a drug is administered to an animal either acutely or sub-chronically, which should lead to quantifiable, objective behavioural alterations in the animal. This is frequently referred to as \"pharmacological validity\" or \"pharmacological isomorphism\" (Matthyse 1986). Ideally, the animal model should be able to detect antidepressant compounds of different chemical classes, such as tricyclics, SSRIs and MAOIs, as well as respond to non-pharmacological therapies for MDD, including ECT, R E M sleep deprivation and transcranial magnetic stimulation (TMS). As approximately 30% of severely depressed humans are unresponsive to antidepressant drugs, the 20 animal model clearly has stringent standards applied to it. In addition, the time course of response to antidepressant therapies should be commensurate with the time course of therapeutic effects that is observed in humans. Thus, animal models in which subjects respond acutely to antidepressant drugs do not closely resemble the human situation, in which therapeutic effects normally take 2-3 weeks to begin (Leon 2000). It has been argued by proponents of acutely-responding models that this may simply reflect differences in metabolism between rats or humans, such as in the rate that neuroactive metabolites are created (Porsolt et al. 1978); alternately, it has also been stated that the dose of drug that rodents receive is almost an order of magnitude greater than humans receive, and that if animals are given lower doses over a longer period, then similar effects are observed (Borsini and Meli 1988). As it is unlikely that experiments with humans will be undertaken to determine the effects of extremely high doses of antidepressant drugs, this issue may remain unresolved for a while. When applying the criterion of pharmacological validity to a model, it is almost as important that the animal does not respond to certain drugs as that it does respond to others. An efficacious animal model of MDD should respond exclusively to antidepressants, and not to other classes of psychoactive drugs, such as anxiolytics or neuroleptics. This type of validity, often referred to as \"discriminant validity\" (Campbell and Fiske 1959), assures that the model is specifically one of depression, and not of other psychiatric disorders. Discriminant validity is also essential for models that are used as pharmacological screens, as it prevents the detection of \"false-positive\" compounds (Willner 1991). However, given that the therapeutic drugs of one class often have therapeutic effects in another disorder, such as the with use of neuroleptics for the treatment of depression (O' Neal et al. 2000) or the use of ECT for schizophrenia (Kupchik et al. 2000), and that a specific symptom (such as anhedonia) may appear in different mental 21 disorders (Sax et al. 1996), a more relaxed attitude towards discriminant pharmacological validity is sometimes allowed, depending upon the behaviour involved in the animal model and the symptom that is being modeled (Willner 1986). Face Validity: A second type of validity that is essential for any animal model of psychiatry is \"face validity\". This refers to the capacity of the animal model to mimic the human condition in a number of important aspects (Mosier 1947). In a seminal paper, McKinney and Bunney (1969) suggested that an animal model of MDD should model depression in etiology, biochemistry, symptomatology and treatment. In practice, though, it is often the case that the animal model is being used to help to understand the etiology and biochemistry of the disorder that it is modeling. Thus, in more recent years, face validity has been applied more as a measure of the capacity of the model to mimic the symptomatology of the human disorder (Geyer and Markou 1995). Without a phenomenological and behavioural isomorphism between the model and the disorder, there can be little confidence that any biochemical changes that may be observed in the animal model bear any relationship to the human condition. It is evident that there are clearly some aspects of M D D that are uniquely human, and could never be modeled in rodents, such as \"feelings of worthlessness\" or \"excessive and inappropriate guilt\". However, the phylogenic development of the human brain has allowed for the conservation of most of the major neuroanatomical regions that are present in rodents, with correlated function (Passingham 1985). The expansion of much of the telencephalon has therefore resulted in differences that are largely quantitative as opposed to qualitative (Kaas 1995). As these more primitive regions of the 22 \"triune\" brain exert a powerful modulatory control over higher cortical centers (MacLean 1985), the behavioural repertoire that is displayed by rodents may be used to model many of the behavioural deficits that are exhibited in depression. In theory, many of the behavioural symptoms of depression may be reflections of physiological changes in subcortical regions of the brain, such as the hypothalamus, basal ganglia or hippocampus (Soares and Mann 1997; Plotsky et al. 1998), that are highly conserved in rodents and humans. Thus, a brief perusal of the criteria that are used to assist in the diagnosis of MDD (see below) indicates that the majority of observable symptoms can be modeled in animals. \u00E2\u0080\u00A2 depressed mood most of the day, nearly every day, as indicated by either subjective report (e.g., feels sad or empty) or observation made by others (e.g., appears tearful). (In children and adolescents, this may be characterized as an irritable mood.) \u00E2\u0080\u00A2 markedly diminished interest or pleasure in all, or almost all, activities most of the day, nearly every day \u00E2\u0080\u00A2 significant weight loss when not dieting or weight gain (e.g., a change of more than 5% of body weight in a month), or decrease or increase in appetite nearly every day. \u00E2\u0080\u00A2 insomnia or hypersomnia nearly every day \u00E2\u0080\u00A2 psychomotor agitation or retardation nearly every day \u00E2\u0080\u00A2 fatigue or loss of energy nearly every day \u00E2\u0080\u00A2 feelings of worthlessness or excessive or inappropriate guilt nearly every day \u00E2\u0080\u00A2 diminished ability to think or concentrate, or indecisiveness, nearly every day \u00E2\u0080\u00A2 recurrent thoughts of death (not just fear of dying), recurrent suicidal ideation without a specific plan, or a suicide attempt or a specific plan for committing suicide Clearly, depressed mood, feelings of guilt and suicidal ideation are beyond the capacity of an animal model to mimic. In contrast, the remaining symptoms can be (and have been) modeled (Willner 1991). Additionally, there are numerous other behavioural alterations that occur in M D D which are amenable to objective simulation in animals, such as decrements in sexual and grooming behaviour. Thus, for an animal model of M D D to have a high degree of face validity, it should be able to induce at least several of the major behavioural disturbances that accompany MDD. 23 Construct Validity: An important remaining criterion for any animal model of psychiatry to fulfil is that of \"construct validity\". The essence of this criterion is that the animal model be based on a sound theoretical rationale, and that it measures what it purports to measure (Willner 1986; Geyer and Markou 1995). In the case of an animal model of MDD, construct validity is often reflected in how accurately the dependent measure represents a cognitive or affective trait. For example, the forced swim test was originally hypothesised to measure \"behavioural despair\" in animals; by the mid-to-late 1980s, however, it was apparent that this interpretation was no longer valid, as it became evident that alternative explanations, such as shifts in ergonomic response strategies, became more plausible (Thierry et al. 1984; Borsini et al. 1986). Hence, the FST, despite its high degree of pharmacological validity and apparent face validity, failed to meet the criteria for an effective animal model of MDD, largely because of its low level of construct validity. Although construct validity is often a difficult aspect of an animal model to determine, it can be increased with the use of well-established animal protocols. For instance, if the animal model is to be used to measure the effects of a depressogenic manipulation on a behavioural trait such as memory or attention, construct validity can be maximised with the use of standardised tasks that have been subject to rigorous testing in their own field, independently of the field of animal models of psychiatry. Reliability and Reproducibility: 24 Even if an animal model of a psychiatric disorder meets the three main criteria for being valid, there are practical considerations of the model that will determine its general usefulness and popularity. The first of these factors is the reliability of the model (Geyer and Markou 1995). The \"reliability\" of the model refers to the consistency and the stability with which the variable of interest that is being observed can be generated. It is suggested that this consistency should be evident in the ability to measure the variable objectively, so that the same behaviour is being measured consistently over repeated experiments. In addition, the model should exhibit a small within-subject and small between-subject variability, so that the power to detect statistical differences is maximised. The \"reproducibility\" of the model refers to the capacity of multiple different research groups to reproduce the behavioural phenomena of the model, under essentially identical conditions (Markou et al. 1993). A model of M D D that meets the three validation criteria, yet cannot be reproduced in different laboratories, must have the generalizability of its results brought into question. Would such a model truly represent an animal model of MDD, or is it merely an artifact of a particular strain of animals that are bred by a particular animal supplier? Although such misgivings may appear trivial and unlikely, there has been widespread concern about exactly this kind of issue in several other animal models of psychiatric disorders. 25 C u r r e n t animal models of depression With more than 20 animal models of MDD that are currently in use (Willner 1991), there would appear to be little need for the development of additional models. However, unlike the situation for many other areas of medicine, such as in the study of diabetes or cancer for which there are satisfactory animal models available, there is a lack of substantive models in psychiatry. Given the vastly greater complexity of an organ such as the brain compared to the somatic organs, it is not surprising that animal models of mental disorders are constantly being reviewed and revised, as more is being learned in the fields of neuroscience and psychiatry. Many of the earlier animal models of MDD are no longer in use: paradigms such as muricidal behaviour (Mann and Enna 1982) and exposure to severe physical stressors (Katz 1982) are no longer considered to be valid models of MDD. At present, there are between 5-10 models that are used widely for research into the etiology and neurobiology of major depression. The following is a brief review of the most popular of these models, with a consideration of their relative strengths and weaknesses. Leaned Helplessness: The learned helplessness phenomenon was originally described by Seligman and his co-workers in dogs, and extended to other species (Seligman 1972). The basis of this model is that exposure to uncontrollable stress leads to behavioural deficits that are not seen in subjects that are exposed to an identical, controllable stressor. With the use of a \"yoked\"-design experiment, Seligman showed that dogs that could control the termination of an electric shock maintained the 26 ability to escape from a subsequent stressor; in contrast, approximately two thirds of yoked animals failed to learn to escape (Overmeier and Seligman 1967). This failure to escape was interpreted as learning by the animal that its earlier attempts to escape were useless, i.e. the animal was \"helpless\". While this model has provided reams of valuable data about the effects of stressor controllability on psychological variables, its contribution to the etiology of M D D is less clear. The learned-helplessness model shows a high degree of pharmacological validity, with positive responses to tricyclic antidepressants (Sherman et al. 1982), ECS (Dorworth and Overmeier 1977), MAOIs and atypical antidepressants (Petty and Sherman 1980). In addition, at first glance the model would also appear to display a high degree of face validity, by mimicking the \"hopelessness\" that is apparent in many depressed humans. However, additional studies have demonstrated that the primary effect of exposure of animals to this paradigm is to generate a high level of anxiety (Willner 1986). There is strong evidence that the learned \"helplessness\" actually reflects cognitive deficits that arise from very high levels of anxiety, impairing the animal's ability to plan executive behaviours effectively (Geyer and Markou 1995). In agreement with this theory, pre-treatment of animals with the anxiolytic drug lorazepam can prevent the induction of learned helplessness (Sherman et al. 1979). The construct validity of the model rests on three assumptions: (i) animals become helpless after exposure to uncontrollable stress, (ii) a similar state is induced in humans by lack of control over stressors and (iii) this helplessness can lead to depression in humans (Willner 1986). A detailed analysis of each of these assumptions is beyond the scope of the current dissertation, but data suggest that the validity of all are open to question (Willner 1986). Hence, the learned-helplessness model of M D D coalesces into a model with high pharmacological validity, but with uncertain levels of 27 face and construct validity; furthermore, learned-helplessness is often difficult to demonstrate in rats, suggesting a low level of reproducibility (Freda and Kline 1976). Flinders Sensitive and Resistant Lines: As part of an attempt to develop a genetic animal model of depression, the Flinders Sensitive and Resistant Lines (FSL and FRL) were generated. Unlike the development of many other \"genetic\" models in different areas of psychiatry, such as in schizophrenia (Mohn et al. 1999), the FSL and FRL are a product of selective breeding, rather than genetic manipulation through knockout strategies. The FSL are a strain of outbred Sprague-Dawley rats that were originally bred for sensitivity to the anticholinesterase agent diisopropyl flurophopshate (Russell et al. 1982), compared to their controls, the FRLs. An increased sensitivity to cholinergic drugs is frequently observed in depressed humans (Janowsky et al. 1994), and early studies with these strains demonstrated that the FSL rats displayed numerous behavioural deficits compared to the FRL rats that were indicative of depressive symptomatology (Overstreet 1992). Measures of pharmacological validity of this model have generally been positive. The results of studies with tricyclic antidepressants have indicated a selectively greater effect of these drugs on FSL rats compared to FRL rats (Schiller et al. 1992). The SSRI sertraline has also been shown to be effective in this model (Overstreet et al. 1995), and discriminant validity has been demonstrated through the absence of effects of other classes of drugs, such as anxiolytics. In addition, the model exhibits pharmacological isomorphism as it requires chronic treatment with antidepressants before effects are observed. At the level of face validity, the FSL rats exhibit numerous behavioural differences from FRL rats. These effects include sleep deficits, 28 psychomotor retardation and anhedonia (Overstreet 1992), all of which are symptoms of MDD. Also, several physiological differences in FSL rats have been reported that parallel changes noted in MDD, including alterations in the HPA-axis (Owens et al. 1991). The validity of the FSL model of depression is questionable, though, when its construct validity is assessed. Firstly, the pharmacological validity it displays is only apparent when both strains of animals are tested in the forced swim test, rather than a more ethologically valid test of natural behaviour. As the FST is itself criticised for a lack of construct validity, the use of the FST with these strains suggests a degree of circularity. Additionally, the evaluation of other aspects of face validity are often determined with tasks that are sometimes ambiguous. For example, FSL rats were concluded to be \"anhedonic\" in a sucrose preference test after exposure to stress (Pucilowski et al. 1993); the reduced sucrose preference, though, could be a result of alternative factors (Forbes et al. 1996), and in more sophisticated tests of reward related function, such as with rewarding electrical brain stimulation, no effects were observed in this model (Matthews et al. 1996). Practical issues are also of importance in this model, as it requires the development of careful breeding programs, and it may be lengthy and expensive to initiate studies with this model. Neonatal exposure to REM-suppressing drugs: Developmental factors may represent important variables in the etiology of M D D (Kaufman et al. 2000). Hence, several animal models have been generated that emphasise the role of disruption of early development in rodents. The most successful and foremost amongst these is the exposure of neonatal rats to REM-suppressing drugs, such as clomipramine (Vogel et 29 al. 1990). In this model, neonatal rats are injected with drugs that suppress R E M sleep between days postnatal 8 to postnatal 21. No differences are observed in these animals until adulthood (approximately 4 months of age), when affective differences start to emerge (Hartley et al. 1990). The current list of behaviours that has been altered by neonatal exposure to R E M -suppressors includes increased immobility in the FST, changes in R E M sleep patterns, decreased aggressiveness, reward-seeking and sexual activity, as well as increased voluntary alcohol consumption (Dwyer and Rosenwasser 1998). The pharmacological validity of this model remains largely undetermined. Although a wide range of behaviours has been shown to be disrupted by this model, indicating a relatively high degree of face validity, there have been few studies that have attempted to reverse the behavioural sequelae of this model. In one such study, it was demonstrated that the deficits in sexual responding that occurred in male rats after neonatal exposure to clomipramine could be reversed by sub-chronic treatment with imipramine and R E M sleep deprivation (Vogel et al. 1990). A substantially greater body of research is required to demonstrate the ability of different classes of antidepressant therapies to reverse behavioural aberrations before the pharmacological validity of this model can be fully determined. The construct validity of the model appears to have several shortfalls too, as the model may be more representative of an animal model of chronic dysthymia than MDD, per se. Unlike MDD, in which periods of depression are interspersed with periods of remission (Simon 2000), animals in this model are chronically and unrelentingly \"depressed\", which does not accurately simulate the situation with humans. The limitations of this model also include practical considerations, as animals must be raised from birth until at least 4 months of age before depressogenic effects are observable; furthermore, it is 30 impossible to develop a \"within-subjects\" experimental design with this model to measure baseline levels of non-depressed behavior. Olfactory Bulbectomy: The olfactory bulbectomy (OB) model of depression is currently one of the most widely used of all animal models of depression. By removing the olfactory bulbs of adult rats, a syndrome of behavioural deficits develops approximately 1-3 weeks later. Typically, OB rats exhibit hyperlocomotion in a novel environment, and this effect can be reversed by treatment with chronically administered antidepressant treatment (Leonard and Tuite 1981). The OB model is relatively simple and fast to use, and has been widely reproduced with a high degree of reliability in numerous laboratories throughout the world. The widespread popularity of the model derives in large part from its high degree of pharmacological validity (Kelly et al. 1997). All major classes of antidepressant drugs have been shown to be active in this model, including tricyclics (amitryptyline, imipramine), SSRIs (paroxetine, sertraline, fluvoxamine but not fluoxetine), MAOIs (moclobemide) and atypicals (mianserin, nomifensine and trazodone) (Kelly et al. 1997). Most importantly, reversal of OB-altered behaviour requires chronic, as opposed to acute, administration of antidepressant drug treatment. The face validity of the OB model has been moderately well established. The majority of studies have focused on hyperlocomotion, which is of questionable relevance to MDD, but several interesting studies have indicated that cognitive processes are also disturbed. Tasks such as maze learning and impulse control are reduced in OB rats, and some of these effects have 31 been reversed with antidepressant treatment (Redmond et al. 1994). Deficits have also been observed in tasks such as step-down passive avoidance (Joly and Sanger 1986). The construct validity of the OB model is based upon the assumption that lesioning the olfactory bulbs in adult rats can lead to behavioural deficits that reflect an underlying affective disturbance. There is clearly no human correlate for the etiology of this model, and in this respect the model's validity is low. Proponents of the model claim that lesions are secondary to alterations in limbic activity, particularly within the amygdala. However, many of the behavioural effects of the OB model may be attributable in part to increased anxiety, and axiolytics have potent effects in this model. The potential utility of this model is also seriously curtailed by the nature of the inducing manipulation. The removal of the animal's olfactory bulbs precludes measuring responding for most primary rewards, such as for food and sexual partners, because these are primarily olfactory stimuli to rats. Thus, measures of depression based on \"anhedonia\" or motivation are impractical. 3 2 Animal Models of Depression- Synthesis: With numerous animal models of depression available to researchers, there are several key issues facing this field of study. It is worth considering that animal models of depression have many different purposes, and in some regards, these purposes have already been met. For instance, there are several highly effective pharmacological screens for the detection of putatively therapeutic drugs, such as the FST, TST and the OB model. A limitation of these tests, though, is that none of them is clearly selective for the effects of rapidly acting antidepressant treatments, which is a major issue in the field of psychiatry. A second concern in the study of animal models of MDD is the development of a rodent model that simultaneously has both high face validity and high construct validity. For those models that are used to provide insight into the etiology of MDD, it is essential that they have somewhat realistic inducing conditions, and that they produce behavioural phenomena that parallel the symptoms of MDD. In order to maximise the construct validity of the model, the behavioural \"symptoms\" in the animal model should clearly represent the core trait that they are supposed to. Through the use of well-established animal paradigms, this validity of these traits can be determined. Finally, the need for a model with high pharmacological, face and construct validity must be balanced by its reproducibility in other locations. The model must be reliable and provide consistent results, regardless of which particular experimental group performs the study. An ideal model will also take into consideration other factors such as duration, cost and ease of use. 33 C h r o n i c M i l d Stress The link between stress and depression is well-established (Lloyd 1980; Brown and Harris 1988). Less clear is the nature of the link between stress and MDD; for example, do high levels of stress precipitate depression, or do depressed individuals simply have a cognitive and affective bias that makes them recall stressful events more clearly? In addition, what types of stress cause depression, and are there particular symptoms that are brought on by these types of stress? The chronic mild stress (CMS) animal model of depression has been developed, in part, to determine the answers to these questions. Based upon a series of earlier studies by Katz et al. (1981; 1982), in which rats were exposed to a series of severe stressors (such as intense footshock, cold water immersion and 48hr food/water deprivation) with ensuing behavioural alterations, Willner and his colleagues modified the nature of this protocol (Papp et al. 1991; D'Aquila et al. 1994). Instead of using severe stressors, the paradigm was modified to utilise only mild stressors, and rather than measure the effects of stress on locomotor activity, the dependent variable became the hedonic state of the animals. In a typical experiment, animals are exposed to a variety of mild stressors (e.g. cage tilting, change of cage mate, overnight illumination and intermittent food and/or water deprivation), that change every few hours over a period of weeks or months. The hedonic state of the animals is determined by tracking, over repeated weekly tests, a decrease in the consumption of and/or preference for a weak (1-2%) sucrose solution. Alternate measures of anhedonia have also been assessed through the use of place preference conditioning (Papp et al. 1991) and rewarding electrical brain self-stimulation (Moreauetal. 1992). 34 The predictive validity of the model is high, with antidepressant drugs from all major classes, including ECT, having therapeutic effects (Willner 1997a). The model also has a high degree of discriminant pharmacological validity, as anxiolyics (chlordiazepoxide), neuroleptics (haloperidol, risperidone) and psychostimulants (amphetamine) all failed to reverse the CMS-induced anhedonia (Willner 1997a). The most desirable aspect of the CMS model, pharmacologically speaking, is that the onset of therapeutic effects closely mimics the profile of effects in humans. Most antidepressant drugs begin to reverse the CMS-induced consummatory deficits within 2-3 weeks, which reflects the typical duration required to observe effects in patients with MDD. The face validity of the CMS model is high too, as a wide range of normally \"rewarding\" behaviours in rats is attenuated with this model (Willner 1997b), which resembles the \"markedly diminished interest or pleasure in all, or almost all, activities\" that demarcates one of the two core symptoms of depression (American Psychiatric Association 1994) . In addition to decreasing responsiveness to rewards, the CMS model induces the appearance of other symptoms of MDD, including alterations in sexual behaviour and aggression (D'Aquila et al. 1994). Various biological markers of MDD have also been reported in the CMS model, such as advance shifts in diurnal rhythms (Gorka et al. 1996), fractionation of the normal sleep architecture (Cheeta et al. 1996) and hyperresponsiveness of the HPA axis (Ayensu et al. 1995) . The construct validity of the CMS model must be considered amongst the highest of all animal models of depression. Extensive research with the CMS model has determined that the CMS-induced anhedonia is reflected in numerous different measures of \"reward\", most of which generate concordant results. The assertion that reduced consumption of a mildly rewarding sucrose solution is indicative of anhedonia is supported by the following observations: 35 \u00E2\u0080\u00A2 Food and water intake remains unaffected in the CMS model (Papp et al. 1991). This suggests that the reduced consumption of a 1% sucrose solution is not part of a general trait of the animal to consume less fluids or food (e.g. as a result of reduced thirst or hunger), but rather a selective reduction in their consumption of mildly rewarding stimuli. \u00E2\u0080\u00A2 Similar effects are observed in animals consuming calorie-free saccharin solutions (Willner et al. 1987). Again, these data indicate that CMS animals are not showing an aversion to caloric solutions. \u00E2\u0080\u00A2 Effects of CMS are evident in both single-bottle tests and two-bottle (sucrose-water) preference tests (Willner et al. 1987), refuting the possibility that the animals are rejecting fluid consumption in a \"taste-aversion\" manner. However, several recent studies suggested that the decrease in consumption of the 1% sucrose solution may have an alternate explanation. When rats are exposed to CMS, they typically exhibit a decrease in weight compared to control subjects (Willner et al. 1997a). Others have hypothesised that this decrease in weight, of a magnitude between 10-20% of control animals, may explain the decreased consumption of the sucrose and saccharin solutions (Matthews et al 1995; Forbes et al. 1996). These groups have shown that consumption of the sucrose solution is linearly related to body weight, suggesting that lighter, i.e. stressed, animals will consume less. This explanation might also account for the alterations in intracranial self-stimulation (ICSS) with rewarding brain stimulation seen in CMS-exposed rats and mice (Forbes et al. 1996). It is not obvious, however, if this hypothesis provides a valid explanation of why CMS-exposed animals reduce their consumption of a 1 % sucrose solution without affecting their overall consumption of food and water. Clearly, it would therefore be of interest to determine if the rewarding or motivational properties of the 1 % sucrose solution could be shown to be altered in CMS-exposed rats, independent of their actual consumption of the fluid (and therefore independent of potential confound of reduced body weights). 37 E X P E R I M E N T 1: E F F E C T S O F C H R O N I C M I L D STRESS O N A N A P P E T I T I V E L Y M O T I V A T E D T A S K Synopsis: The Chronic M i l d Stress paradigm is an animal model of depression, based upon the exposure of rodents to a series of chronic mild stresses, which results in a decrease of the hedonic capacity of these animals. However, several recent studies have suggested that the C M S model may be confounded by weight loss in stressed animals. We therefore evaluated the effects of exposure of rats to C M S in an appetitively motivated task. Rats were trained to expect the presentation of a 1% sucrose solution after a 10 minute cued anticipatory period. Half of the animals were then exposed to C M S . The number of nose-poke investigations into a niche that subjects made for the sucrose solution during the 10 minute period differed significantly between stressed and non-stressed animals by the second week. In contrast, differences in the consumption of the sucrose solution did not appear until the fourth week of exposure to C M S . These results are discussed with reference to differential effects of stress on appetitive vs consummatory behaviours. 38 Introduction: The CMS animal model of depression focuses upon and successfully emulates the core depressive symptom of anhedonia (a reduced interest and/or pleasure in normally rewarding activities) (Willner et al. 1987). By subjecting animals to a prolonged series of mildly uncomfortable and unpredictable stressors, a significant reduction is observed in subjects' consumption of rewarding stimuli, such as a weak sucrose solution or ICSS of natural reward sites in the brain (Papp et al. 1991; Moreau et al. 1992). While there is a strong body of evidence that supports the hypothesis that CMS leads to a reduced hedonic capacity in rodents, several recent studies have provided evidence which suggests that these \"consummatory\" changes may be more closely linked to changes in body weight (Matthews et al 1995; Forbes et al. 1996). The capacity of the CMS model to affect an animals' interest in obtaining the sucrose solution, independent of its free consumption of the solution, may provide an opportunity to assess anhedonia without the confound of stress-induced weight loss. The majority of studies to date have investigated the effect of CMS on \"consummatory\" behaviours. However, it has been well established that \"appetitive\" or \"preparatory\" behaviours (i.e. those instrumental behaviours that lead to the availability of the rewarding stimulus) are more susceptible to behavioural disruption by various experimental manipulations. In particular, dopamine (DA) antagonists and 6-OHDA lesions of the mesoaccumbens D A pathway have been shown to disrupt preferentially conditioned or instrumental responses before affecting consummatory behaviours (Fibiger 1993; Salamone 1994). 39 Given the large body of evidence implicating the role of DA in anhedonia and the CMS procedure (Muscat et al. 1990; Papp et al. 1994), the current experiment was undertaken to determine if an appetitive task in the CMS model of depression would be affected preferentially compared to the standard consummatory measure. Furthermore, it was predicted that CMS-induced changes in an appetitive task could provide an operational measure of anhedonia that would not be confounded by changes in the weights of stressed animals. Methods: Subjects Animals were male Sprague-Dawley rats (2 groups; n=9 per group), weighing 300-350g at the start of the experiment. The animals were housed individually in a temperature and light controlled colony (12 hr light/dark cycle; lights on at 06:30) with food and water available ad libitum, except when specified as otherwise. All procedures were performed under the guidelines of the Canadian Council on Animal Care and the University of British Columbia Committee on Animal Care. Training The testing chambers consisted of Plexiglas boxes (26 x 25 x 29 cm) with a hole in one wall (4.5 cm above the floor) through which the spout of a Richter drinking tube could be placed. Inset 2.5 cm from this wall was an opaque, \" false\" aluminium wall, which in turn had a 7.5 x 9 cm section removed in the center to allow the animal free access to the Richter tube. Another 40 aluminium panel (10 x 13 cm) was placed over the smaller cut-out section, firmly supported between two parallel guide strips, which could be manually lowered or raised to deny or allow the animal access to the Richter tube. This guillotine door was clearly painted white with a large black \"+\" sign to make it highly visible to the subject. By using a chamber with this design, it was possible to control access by the rat to the Richter tube, and to observe when the subject poked its head through the \" false\" aluminium wall to investigate the presence the Richter tube and/or drink from it. Subjects were given two 60 minute habituation sessions in the test chambers on consecutive days, during which the Richter tube was not presented and the aluminium panel remained lowered. All animals were then given two 1 hour presentations of the Richter tube in their home cages, also on consecutive days, when the tubes were filled with a 1% sucrose solution. Following these habituation sessions to the test chambers and sucrose solution, subjects were given a twelve session training schedule. The animals were placed individually in the test chambers at approximately 16:00 , and deprived of both food and water. On the ensuing day, at approximately 12:00 (always 20 hours after food and water deprivation), the animals were exposed to white noise (\u00C2\u00AB 65 dB) and the aluminium panel was raised to provide access to the Richter tube. During the first two training sessions, the Richter tube (filled with a 1% sucrose solution) was presented immediately upon the initiation of the two above cues; the radio was then turned off after 60 seconds, but the aluminium panel remained open for 1 hour, and subjects were able to drink freely during this period. At the end of the session, the panel was lowered and 10 minutes later subjects were returned to their home cages, where they were given free access to both food and water for 48 hrs. Training continued according to this schedule. 41 For the first two training sessions, Richter tubes were presented at the same time as the auditory cue and the raising of the panel, while for the next two sessions, the tube was presented 2 minutes after cue onset. In these and subsequent training sessions, the white noise was always terminated exactly 60 seconds after presentation of the Richter tube, regardless of the time between the initiation of the cues and the presentation of the sucrose. Two training sessions with a 5 minute interval, two sessions with 8 minute intervals, and four sessions with 10 minute intervals completed the twelve session training schedule. During the last two 10 minute interval sessions, four separate behavioural measures were manually recorded. The first of these measured the time from when the auditory cue began and the aluminium panel was raised to when the subject first placed his head or nose past the \" false\" aluminium wall to investigate the presence of the Richter tube (this was termed as \" latency to door\" ). The second measure recorded the \" number of investigations\" that the rat made during the 10 minute period between cue initiation and sucrose presentation - defined as the number of times that the subject placed any part of his head past the \" false\" aluminium wall. Thirdly, the latency to consume sucrose was measured from when the Richter tube was presented, after the 10 minute cued anticipatory period. Finally, the total amount of sucrose solution that the rat consumed was measured by weighing the preweighed Richter tube at the end of the 60 minute consummatory period. Testing Subjects were divided into two groups of nine rats each, matched on their sucrose consumption from the final two training sessions. One group served as controls, while the other was subjected to the chronic mild stress regimen. The testing procedure was similar to the final four training sessions: briefly, subjects were food and water deprived at 16:00h and placed into 42 the test chambers, and at 12:00 the next day, the auditory cue was turned on and the aluminium panel raised. A 10 minute anticipatory period followed, during which the rat typically made a number of investigations of the niche that would contain the spout of the Richter tube. After 10 minutes the sucrose solution was presented, and the rat was allowed to consume freely for 60 minutes. During the test sessions, the four separate behavioural measures were recorded as described above; 10 minutes after testing, subjects were returned to their home cages and given food and water ad libitum. All subjects were tested once per week, on the same day. The results were analysed by analysis of variance (ANOVA) and supplemented with tests of simple main effects. Stress The weekly stress regime used was similar to that used by Willner et al. (1987), and consisted of one period of overnight illumination, two periods of 45\u00C2\u00B0 cage tilt (7 and 8 hours), two periods of soiled housing (18 hours) and two periods of paired housing (22 hours). Results: The latencies for subjects to investigate the niche at the start of each test session did not differ significantly [F(l,16) = 3.476,p < 0.081], although there was a trend towards longer latencies in the experimental group. Nor were there significant differences between groups in the latencies the groups took to consume from the Richter tube upon its presentation, which were normally within the range of 3 - 7 seconds. The CMS procedure, however, produced a dramatic effect on the number of investigations made in the 10 minute period prior to the availability of sucrose [F(l,16) = 18.84, p < 0.001], as well as a significant group x time effect [F(9,144) = 2.03,/? < 0.04]. A strongly significant effect of CMS was observed, in accordance with previous studies, on the total mount of 1% sucrose solution that was consumed [F(l,16) = 13.04, p < 0.002] [Figure 1.1]. Post-hoc tests revealed that CMS-exposed animals exhibited significantly fewer niche investigations after only three weeks of stress, whereas decreases in sucrose solution consumption did not appear until after the fifth week of exposure to CMS. 44 o CO Q I ! ! ! ! ! L 1 2 3 4 5 6 W E E K S Fig. 1.1. Effects of CMS upon levels of levels of appetitive and consummatory responding for a 1 % sucrose solution. The duration of the appetitive and consummatory phases are 10 and 60 min respectively. * Denotes significantly different at p < 0.05 level, o Control, \u00E2\u0080\u00A2 CMS 45 Discussion: The data from this experiment provide further new evidence that the CMS animal model of depression accurately resembles a number of different symptoms of human endogenous depression. The major finding of this study was that an appetitive behaviour could provide a more sensitive measure of the effects of CMS than a consummatory one, in that the appetitive behaviour was disrupted earlier and more significantly than the consummatory behaviour. While few studies have measured the effects of chronic mild stress on concurrent appetitive and consummatory behaviours, the results of the present experiment are consistent with the observation of decreased sexual mounting activity (considered an appetitive behaviour) by stressed as compared to non-stressed male rats (D'Aquila et al. 1994). In this study, measures of sexual consummatory behaviour (i.e. intromission, ejaculations) were not recorded. A number of studies have demonstrated place- preference conditioning deficits to rewarding stimuli in stressed animals (Papp et al. 1991; Muscat et al. 1992), which may be seen as a disruption of anticipatory/ appetitive behaviour; however, these studies did not evaluate the longitudinal development of behavioural effects of chronic exposure to CMS, unlike the present study. Evidence that appetitive behaviours may be disrupted more easily than consummatory ones comes from numerous studies which have observed the effects of a wide range of dopamine antagonists on such behaviours (for reviews see Fibiger 1993; Salamone 1994). One earlier study performed in the author's laboratory used the D2 antagonist pimozide on unstressed rats and measured its consequences on a number of different appetitive and consummatory behaviours, similar to those used in this study (Blackburn et al. 1987). The results from that study demonstrated a significant effect on the number of investigations rats made into a niche in 46 anticipation of a cued meal, which resembles the effect of CMS in this experiment. A significant effect was also seen on the latencies taken for subjects to enter the niche for the first time, typically being 2-3 times longer in drugged rats than controls; a result of comparable magnitude was found in this study too, but the large variance in individual latencies meant that this effect barely missed significance. This latter result may emphasise the importance of the role that individual differences in response to stress play when using the CMS procedure. The second appetitive latency measured in this study was not significantly different between groups, as both groups of rats took equal time to drink the sucrose solution upon its presentation. Thus, conditional or instrumental responding may be more severely disrupted when the associative link between CS and UCS is weaker. It would therefore be of interest to investigate the effects of CMS on other types of conditioned responding when compared to consummatory behaviour, such as in the acquisition of a new response in the absence of the primary reward, or on second-order schedules of reinforcement (Robbins et al. 1989). Given the large literature demonstrating the ability of stress to affect D A release in various limbic structures, including the nucleus accumbens (Carlson et al. 1993; Kalivas and Duffy 1995), it is likely that the similarity between the effects of CMS and low doses of D A antagonists on appetitive behaviours is not coincidental. Previous studies have established that CMS causes subsensitivity of postsynaptic D2 receptors in the nucleus accumbens (Willner et al. 1991; Papp et al. 1994), and that this in turn may be partly responsible for rats' decreased sensitivity to reward. As the nucleus accumbens is also critical for many appetitive behaviours (Salamone 1994), it is possible that these subsensitive D2 receptors also disrupt appetitive behaviours. Typically, appetitive responses produce a smaller in vivo D A efflux in the nucleus accumbens than consummatory behaviours (Fiorino et al. 1997), and this may explain why, in a 47 subsensitive system, appetitive behaviours are disrupted more easily than consummatory ones (Fibiger 1993. The use of an appetitive paradigm that measures an animal's interest in obtaining a natural reward should theoretically be independent of stress-induced changes in body weight. Although it has been hypothesised that decreased consumption of a weak sucrose solution may be related more to changes in body weight than a state of anhedonia (Forbes et al. 1996), the appetitive measures used in the present study suggest that anhedonia (i.e. a reduced interest in normally rewarding stimuli) may be apparent by the third week of exposure to CMS. In conclusion, anhedonia - the core symptom of depression that the CMS model replicates - is often defined as a \" loss of interest or pleasure\" (American Psychiatric Association 1994) and currently nearly all studies involving the CMS procedure have focused on the loss of \"pleasure\" rather than loss of \"interest\". By using the ethologist's distinction of dividing goal-directed behaviours into both appetitive and consummatory behaviours (i.e. interest versus pleasure), results from this experiment demonstrate that these two types of behaviour are differentially sensitive to the CMS. This observation may help to provide further insight into the progressive development of the anhedonia which the CMS animal model of depression duplicates, and even help in the creation of antidepressant drugs designed to be more effective in the alleviation of appetitive than consummatory behaviours in humans. 48 EXPERIMENT 2: CHRONIC MILD STRESS HAS NO EFFECT ON RESPONDING B Y RATS FOR SUCROSE UNDER A PROGRESSIVE RATIO SCHEDULE Synopsis: Exposure of rats to chronic unpredictable mild stress (CMS) has been shown to produce a syndrome in which a wide range of consummatory behaviours are attenuated, resembling a state of anhedonia, which may be reversed by treatment with antidepressant drugs. The aim of the present study was to determine if CMS would also affect a rat's motivation to respond for a sucrose solution, as assessed by its performance under a progressive ratio (PR) schedule of reinforcement. Control studies demonstrated that break points in non-stressed rats were sensitive to both the concentration of sucrose solution used, as well as the period of food and water deprivation used prior to testing. Exposure of rats to CMS had no effect upon break points when responding under a PR schedule for either a 1% or 7% sucrose solution, although subjects did display the typical reduction in consumption of a freely consumed 1% sucrose solution. These results are not readily understood within the theoretical framework of the CMS model of anhedonia, and imply instead that both the neural and psychological correlates of motivation may be less susceptible to modulation by the effects of CMS than the free consumption of sweet solutions. 49 Introduction: Currently, few animal models are available which provide adequate face and construct validity for the symptoms of depression in humans. One paradigm that addresses this issue is the Chronic Mild Stress (CMS) model as developed by Willner et al. (1987; 1991), which is purported to model the core human depressive symptom of anhedonia (i.e., a loss of interest in or pleasure from normally rewarding activities). An impressive body of evidence has been gathered to support the hypothesis that animals subjected to a daily series of relatively innocuous stressors begin to show anhedonic-like behaviour after two to three weeks of CMS. Typically, rats exposed to sufficient periods of CMS demonstrate a decrease in their consumption of a weak sucrose solution (Willner et al. 1987), and also an increase in the stimulation frequency required to support threshold intracranial self-stimulation (ICSS) of the ventral tegmental area (Moreau et al. 1992). Both of these effects imply that there is a dysregulation of the neural substrates involved in normal hedonic behaviour. A substantial body of data also indicates that one of the neural substrates affected by CMS is the mesolimbic dopamine (DA) system (Papp et al. 1994; Willner et al. 1991); alterations in the stimulated release of D A in the nucleus accumbens (Stamford et al. 1991) have been reported, as well as decreased numbers of D2 receptors in the same region of the brain (Papp et al. 1994). Numerous experiments have shown that either 6-OHDA lesions of dopaminergic neurons or systemic doses of dopaminergic antagonists tend to disrupt preferentially behaviours linked to incentive motivation (Fibiger 1993; Salamone 1994). Appetitive/preparatory and complex instrumental responses are particularly sensitive to 50 modulation of the mesolimbic DA system, and doses of neuroleptics that inhibit lever-pressing for either food or water leave consummatory responses intact (Salamone 1996). In recent years, the Progressive Ratio (PR) procedure has seen increased use as a means to measure an animal's motivation to respond for rewarding stimuli (Richardson and Roberts 1996). Specifically, by training rats to lever-press for a fixed reward, and progressively increasing the number of presses required to obtain each subsequent reward, a measure can be obtained of the total amount of effort that animals will expend, with the final ratio achieved (i.e. the \"break point\") providing an objective evaluation of the subject's motivation to work for a particular reward. Reduced motivation is an important symptom of melancholic-type depression, and therefore, the present experiment was undertaken to determine if a state of anhedonia induced by CMS model of anhedonia could be inferred from a reduction in an animal's motivation to work for a sucrose reward, as measured by break point values obtained with the PR procedure. Furthermore, it was also of interest to determine whether performance on a PR schedule would provide a more sensitive measure of anhedonia than the traditional consummatory measures (Willner et al. 1987; 1991). Methods: Subjects Twenty-four male Long-Evans rats (Charles River, Quebec) weighing 250-300g at the start of the experiment, were housed individually in a temperature controlled colony (21+1 \u00C2\u00B0C) under a 12-hr light-dark cycle (lights on at 07:00 hrs). One subject had to be sacrificed during 51 the experiment, and its data were not included. All experiments were conducted in accordance with the Canadian Council on Animal Care guidelines for work with laboratory animals. Training Subjects were maintained on a constant 20-hr food and water deprivation schedule (in which food and water were available for only 4 hrs per day), whereby their weight was reduced to approximately 90% of their free feeding weight and maintained at this level by carefully controlling food intake (subjects were limited to 21 grams of food per day). Pilot studies had demonstrated that this procedure generated more stable patterns of responding on the PR schedule than intermittent deprivation. All subjects were tested daily between 09:00 and 13:00 hours, individually, in four Plexiglas test cages (25 x 25 x 25 cm) enclosed within sound attenuating chambers, and fitted with a removable response lever and lickometer. Based upon previously published experiments (Montgomery and Suri 1996) which reported that consumption of different concentrations of a sucrose solution is described by an inverted u-shaped curve, animals in the present study were trained with a 7% sucrose solution, the most preferred concentration. Initially, subjects were given a 48-hr exposure period to the 7% sucrose in their home cages, and were then allowed two 30 min habituation sessions in the test chambers, during which both the response lever and fluid spout were withdrawn. For the next four days, subjects were given daily 1-hr sessions during which they had free access to the sucrose solution in the test chambers, which they could obtain by licking a metal spout connected to a lickometer (all subjects learned this readily). Response levers were then introduced into the chambers and subjects were placed on a 1-hr session fixed ratio (FR) schedule of responding as 52 follows: 3 days at FR1/0.05 ml of 7% sucrose solution, 2 days at FR2/0.10 ml sucrose, 1 day at FR3/0.15 ml sucrose and 2 days at FR10/0.40 ml sucrose per reinforcement. All rats learned to barpress on an FR schedule and were subsequently placed on a PR schedule of reinforcement. Response requirements were based on the following schedule, whereby successive reinforcements could be earned according to the following number of barpresses: 1, 3, 6, 10, 15, 20, 25, 32, 40, 50, 62, 77, 95, 118, 145. The final ratio achieved represented the \"break point\" value, and the session was terminated when subjects failed to reach the next barpress criterion within 1 hour. The size of the reward was set at 0.5 ml of 7% sucrose solution per reinforcement, as this was shown in a pilot study to sustain high levels of responding, without allowing rats to consume quantities of sucrose near to their free consumption values, thus minimizing the effect of satiety on the PR data. Animals received six PR sessions, and levels of responding were stable by the fourth session (mean performance over the last three sessions varied by approximately \u00C2\u00B1 1 break point per subject). To verify that rats responding on a PR schedule were responsive to alterations in the hedonic value of the sucrose solution, subjects were split into two groups (n = 11; 12) which were matched on their break points over the last three training sessions. Each group then received two 1-hr free consumption sessions in the test chambers over two days, while the response levers were removed: one group received 0.7% sucrose while the other was given 30% sucrose. Subjects were returned to their home cages for 24-hrs and on the following day animals were tested on a PR schedule with one of the two different concentrations of sucrose solution. After this test, all rats were rehabituated to the 7%> solution with two 1-hr free consumption sessions, followed by one further PR session. 53 Chronic Mild Stress Following rehabituation to the 7% sucrose solution, subjects were again divided into two groups, matched on break point values from their last three training sessions (it should be noted that these two groups did not correspond to the groups formed previously to examine hedonic shifts with the 0.7% and 30% sucrose solutions). One group (n = 12) remained as control subjects, while the other group (n = 11) was transferred to a separate colony and exposed to the CMS procedure as described by Willner et al. (1987; 1991) The basic CMS protocol was as follows: two 18-hr periods of paired housing; two 20-hr periods of 40\u00C2\u00B0 cage tilt; two 18-hr periods of damp bedding; one period of overnight illumination and four 2-hr sessions of white noise (> 75 dB). All stressors were applied pseudo-randomly throughout the week to make them less predictable, and all stressors were terminated at least two hours before any behavioural testing occurred. It is important to note that the stressor of acute intermittent food and water deprivation used routinely by Willner and colleagues was not employed, as all subjects were already on a 20-hr deprivation schedule with limited access to food. Behavioural Testing Following CMS Behavioural testing commenced four days after the initiation of CMS (see Table 2.1), and thereafter animals were tested on the PR schedule for 7% sucrose every four days. This procedure continued for seven test sessions (i.e. until 28 days after CMS began), by which time animals submitted to CMS are reported to exhibit significant changes in their hedonic responses relative to control subjects (Willner et al. 1991). As no effects on break point values were evident at this time, it was hypothesized that the magnitude of the reward may have been 54 overshadowing any differences in motivation between both groups. Therefore, the 7% sucrose solution was replaced with a 1% sucrose solution and the experiment was repeated as follows. Four days after the last PR session (i.e. day 32), the response levers were removed and subjects were allowed to consume 7% sucrose freely for one hour. On days 35 and 36, all subject were habituated to the 1% solution with 1-hr free consumption sessions, and on day 40 both groups were tested for their free consumption of 1% sucrose solution during a 1-hr session (the normal consummatory measure used in CMS studies). Testing continued with the PR schedule on a weekly basis for the next three weeks, while standard 1-hr free consumption tests with 1% sucrose were conducted every three days after a previous PR session. The final phase of this study examined the effect of minimal food and water deprivation on the rewarding value of the sucrose solution. Animals were now given ad lib access to food and water for 12-hr a day and tested after a 12-hr deprivation period (in comparison with the previous 20-hr restriction). Subjects were given one week for their weight to stabilize under this new regimen, and were then tested on days 65 and 72 for their free consumption of 1% sucrose. They were tested on days 69 and 76 with the PR schedule. After the final PR test on day 76, rats were rehabituated to the 7% sucrose solution used previously with two 1-hr free consumption session on days 79 and 80, and were finally tested for their free consumption of 7% sucrose on day 83. The weights of all animals were recorded weekly, after subjects had been fed and watered, and at least four hrs after subjects had been exposed to any type of stressor. 55 T A B L E 2.1 SCHEDULE OF EVENTS Day(s) Event(s) 0 Subjects matched into 2 groups 1-83 One group exposed to CMS regimen 4,8,12,16,20,24,28 All subjects tested under PR schedule for 7% sucrose 32 1 Hr test of free consumption of 7% sucrose, 20 Hr dep. 35,36 Habituation to 1% sucrose in test chambers 40,47,54 1 Hr test of free consumption of 1% sucrose, 20 Hr dep. 44,51,57 Tested under PR schedule for 1% sucrose, 20 Hr dep. 58-64 Ad lib access to food and water 65,72 1 Hr test of free consumption of 1% sucrose, 12 Hr dep. 69,76 Tested under PR schedule for 1% sucrose, 12 Hr dep. 79,80 Rehabituation to 7% sucrose solution in test chambers 83 1 Hr test of free consumption of 7% sucrose, 12 Hr dep. 56 Results: The effects of testing animals under the PR procedure with three different concentrations of sucrose solution are shown in Fig. 2.1. Analysis with paired t-tests revealed that when the concentration of the sucrose solution was increased from 7% to 30%, there was a small increase in the value of the break point which was only marginally significant, [t(l,21) = 3.86, 0.05 < p < 0.10]. However, a reduction in the sucrose concentration from 7% to 0.7% was accompanied by a significant decrease in the break point, [t(l,21) = 88.0,/? < 0.001]. This result demonstrated that a 7%> sucrose solution may represent an optimal concentration for investigating decreases in a rat's motivation to respond for a sweet reward. This responsiveness to alterations in the reward value of the sucrose solution was seen again during the main testing phase of the experiment, as shown in Fig.2.2. A two-factor ANOVA demonstrated a significant effect of testing condition, [F(2,42) = 108.136,/? O.001], but no interactive effects. Initially, subjects in both CMS and control groups were deprived of food and water for 20 hrs and responded for 7% sucrose; when this solution was replaced with a 1% solution, a significant reduction in break points was observed in both groups. Additionally, when the concentration of the sucrose solution (and hence its hedonic value) remained at 1%, and drive level was decreased by reducing deprivation from 20-hr to 12-hr, a further significant drop in break points was observed. In contrast to these changes in break points in both the experimental and control groups to alterations in hedonic value and drive level, CMS had no significant effect on break points with any of the three different conditions, [F(l,21) < 1, NS]. 57 0.7% 7% 30% Sucrose Concentration Fig. 2.1 Effects of changes in sucrose concentration on final ratios (i.e. break points) attained under a progressive ratio schedule by control rats. *** denotes p<0.001 compared to 0.7% sucrose; + denotes p<0.10 compared to 7% sucrose solution. 58 When subjects were tested on the standard measure of hedonia (free consumption of a 1% sucrose solution for 1-hr) after 20-hr of food and water deprivation on days 40, 48 and 54, a one factor A N O V A showed that animals in the CMS condition consumed significantly less sucrose than control subjects, [F(l,21) = 8.31, p <0.01], (Table 2.2). There was no significant interaction. Similarly, when animals were tested with the 1% sucrose but after only 12-hr of deprivation, on days 65 and 72, CMS clearly reduced consumption again, [F(l ,21) = 11.88, p <0.005]. The same subjects displayed no effect of CMS when given a 1-hr free consumption test of 7% sucrose, after 20-hrs of deprivation on day 32, [F(l,21) = 0.92, NS], or after 12-hrs of deprivation on day 83, [F(l,21) = 0.43, NS]. A one factor A N O V A of body weights measured weekly (Fig. 2.3) demonstrated a significant effect of CMS on this measure, [F(l ,21) = 36.27, p O.001], as well as a significant effect of test day, [F(l 1,231) = 142.50, p <0.001]; and an interactive effect of group x test day, [F(l 1,231) = 10.43, p <0.001]. A simple main effects post-hoc comparison between groups demonstrated that the body weights of rats in the CMS and control groups differed significantly by the second week of testing, which lasted until the end of the experiment. 59 T A B L E 2.2 FREE SUCROSE CONSUMPTION V A L U E S (g) FOR 1HR TEST SESSIONS Control CMS 1% SUCROSE, 20 HR DEP. 13.88 \u00C2\u00B1 1.03 9.58 \u00C2\u00B11.08 * 1% SUCROSE, 12 HR DEP. 11.19 \u00C2\u00B1 0 . 7 0 7.68 \u00C2\u00B1 0 . 7 4 * * 7% SUCROSE, 20 HR DEP. 34.33 \u00C2\u00B1 2.24 36.46 \u00C2\u00B1 2.34 7% SUCROSE, 12 H R D E P 29.33 \u00C2\u00B1 1.73 31.73 \u00C2\u00B1 1.80 Values are mean \u00C2\u00B1 SEM * CMS and control groups differ significantly, p<0.01 ** CMS and control groups differ significantly, p<0.005 60 Fig. 2.2 Effects of chronic mild stress on final ratios attained under a progressive ratio schedule, as a function of sucrose concentration (7% or 1%) or hours of food and water deprivation (20 Hr or 12 Hr) for both CMS (T) (n=l 1) and control (o) (n=12) groups. A: Final ratio (i.e. break point) values and corresponding cumulative total lever presses for individual daily test sessions. No significant difference was found between CMS and control groups. B: Final ratio values averaged over the test sessions for different test conditions. Data are presented as (mean \u00C2\u00B1 SEM). *** within brackets denotes significantly different (p<0.001) between testing condition, but with no between group difference. 61 1 2 3 4 5 6 7 8 9 10 11 12 Test Week Fig.2.3 Effects of chronic mild stress on body weight, measured weekly. Control subjects were maintained at 90% of free feeding body weight from weeks 1-8, after which subjects were deprived of food and water for 12 Hr per day. * denotes a significant difference (p<0.01) from controls, ** denotes a significant difference (p<0.005) from controls, *** denotes a significant difference (p<0.001) from controls for all weeks enclosed by brackets. 62 Discussion: The present results confirm one of the main findings of the CMS model of depression, namely that rats subjected to this treatment consume smaller quantities of 1% sucrose solution than control subjects (Willner et al. 1987; 1991). Paradoxically, the same rats do not differ with respect to the amount of work they are willing to perform to receive a fixed quantity (0.5 ml) of a 1% sucrose solution, as measured by the break point achieved under a PR schedule of reinforcement. Control experiments confirmed that this procedure was sensitive to the effect of reducing the concentration of sucrose from 7% to 0.7%, as reflected in a significant reduction in the value of the break point. This effect was replicated within the main experiment of this study following a reduction from 7% to 1% sucrose solution. Furthermore, when drive level was reduced by subjecting the rats to only 12 hrs as compared to 20 hrs of deprivation, we again observed a significant reduction in the break point in both the CMS and control groups. Two conclusions may be drawn from the pattern of results: first that the PR procedure is sensitive to changes in motivation induced by a reduction in the value of the sucrose reinforcer or drive level; and second, that rats subjected to CMS are as motivated as control subjects to work for different concentrations of sucrose solution. This pattern of results would appear to run counter to an important prediction of the CMS model of depression, namely that this treatment reduces the reward value of sucrose solutions because of anhedonia (Papp et al. 1994). One explanation for the differential effects of CMS on free-consumption of sucrose solution as compared with operant responding for this reinforcer under a PR schedule, may be related to selective effects of CMS on the hedonic property of a sweet solution, as distinct from the desire or willingness to work for the same stimulus. Berridge and his colleagues refer to the 63 psychological correlates of these two conditions, as \"liking\" versus \"wanting\" (Berridge et al. 1989; Berridge 1996). These researchers draw a further distinction between these two constructs, in terms of the underlying neural substrates. The motivation to respond for a sweet solution is linked to activity in the mesotelencephalic dopamine system. Berridge et al. have shown that 6-OHDA lesions of the nucleus accumbens cause aphagia without affecting hedonic reactions to a sweet solution (Berridge et al. 1989). More recently, Ikemoto and Panksepp (1996) demonstrated that D A antagonists can reduce \"wanting\" a sucrose solution by rats as assessed by various appetitive measures, whereas measures of consumption of the solution were unaltered. Complementary to this work, Agmo et al. (1995) have established that specific doses of neuroleptics can block the formation of a place preference while leaving sucrose consumption unchanged; conversely, certain doses of the opioid antagonist naloxone reduce consumption but do not affect place preference conditioning. Therefore it appears that \"liking\" may be partially subserved by separate neural systems from those involved in \"wanting\". The neural systems involved in \"wanting\" probably include the mesolimbic DA system and regulate most appetitive behaviours, whereas those circuits concerned with \"liking\" may regulate consummatory behaviours (Berridge 1996). If this explanation is correct, then it follows that the present finding of no change in break points after CMS poses a challenge to the role of the mesotelencephalic dopamine system in depression as inferred from studies with the CMS procedure (Willner et al. 1991). The question remains as to whether the break point value is a valid measure of the anhedonia characteristic of human depression. One study has addressed this issue, in which depressed human subjects were tested for their performance on a PR schedule, using a monetary reinforcer (Hughes et al. 1985). While the results showed that depression caused a significant 64 reduction in break points, a number of methodological concerns, such as the small sample size and absence of a control group, emphasize the need for a well-controlled study applying the PR paradigm to depressed human volunteers, in order to test its efficacy as an instrument for measuring melancholic-type depression. At present, there are no other reported studies with animal models of depression in which the motivation of animals to obtain primary rewards has been determined. Recently, the CMS model of anhedonia has been criticized for failing to take significant reductions in body weight into account as a factor in reduced sucrose intake. When sucrose consumption was adjusted for the difference in body weights between stressed and control animals, consumption deficits between these groups were no longer significant (Forbes et al. 1996). In the present experiment, differences in body weight between CMS-exposed rats and controls were apparent by the second week of behavioural testing. When sucrose consumption was recalculated according to the method of Matthews et al. (1995; Forbes et al. 1996), and consumption per gram of body weight was determined, the sucrose intake did not differ significantly between the two groups. However, the issue is complicated by the fact that despite the notable difference in weight between stressed and control animals (approximately 10-15%), similar volumes of a 7% sucrose solution were consumed when subjects of each group were allowed to drink freely for 1 hr: these data indicate that sucrose consumption is not rigidly determined by body weight. Nevertheless, these data raise another note of caution in interpreting the effects of CMS on hedonic processes. 65 EXPERIMENT 3: CHRONIC MILD STRESS MODIFIES E X T R A C E L L U L A R L E V E L S OF DOPAMINE AFTER CONSUMPTION OF A SUCROSE SOLUTION Synopsis: A large body of evidence suggests that alterations in neurotransmission within the mesocorticolimbic dopamine system may contribute towards the development and expression of MDD. The chronic mild stress (CMS) animal model of depression represents a well-tested rodent paradigm that emulates anhedonia, one of the two core symptoms of unipolar depression. Indirect evidence exists for changes in dopaminergic neurotransmission in the CMS model. The purpose of the present study was therefore to determine if these hypothesised changes could be detected in vivo, with the use of in vivo microdialysis. After 6 weeks of exposure to CMS (or control conditions), rats were implanted with microdialysis probes directed at the nucleus accumbens. Subjects were then allowed to consume a fixed volume of a 1% sucrose solution over 1 hr, and dopamine and its metabolites were measured. The results indicated an unexpectedly greater relative increase of dopamine release in CMS compared to control animals. These results call into question the validity of CMS as an animal model of hypodopaminergic-related melancholic-type depression. 66 Introduction'. The monoamine theory of depression is perhaps one of the most established theories in the field of psychiatry (Hyman 2000). For over thirty years, there has been an ever-increasing body of evidence which suggests that alterations in monoamine neurotransmission can contribute to the expression of depressive disorders (Schildkraut 1965; Delgado 2000). Although the majority of research has focused upon norepinephrine and serotonin as neural substrates of MDD, there remains a convincing argument that changes in dopamine neurotransmission may be an important factor in the etiology of MDD. A wide range of data have provided support for the hypothesis that reduced neurotransmission within the mesocorticolimbic dopamine system may contribute towards depressive symptomatology (Willner 1995); in particular, it is posited that lower levels of endogenous dopamine release may lead to anhedonia and vegetative symptoms. Furthermore, recent preclinical studies with rodents have indicated that different classes of antidepressant drugs may act through a common dopaminergic pathway to generate their therapeutic effects (Santiago et al. 1998; Zhang et al. 2000) The CMS model of depression is an animal paradigm of MDD that seeks to simulate the core depressive symptom of depression (Willner et al. 1987). An impressive body of data indicate that after rodents are exposure to chronic levels of mild, unpredictable stress, they begin to exhibit a decrease in their hedonic capacity for rewarding stimuli. This effect is typically measured by the reduced consumption of a weak sucrose solution in CMS-exposed animals (Papp et al. 1991). In a related manner, a number of studies using the CMS model have found evidence for changes in dopaminergic neurotransmission. Both behavioural and receptor binding studies implicate changes in dopaminergic neurotransmission specifically within the nucleus 67 accumbens (Papp et al. 1993; 1994). Given that mesoaccumbens dopamine is hypothesised to play an important role in motivated behaviour (Fibiger and Phillips 1988; Ikemoto and Panksepp 1999; Schultz et al. 2000), it is clearly of interest to determine whether or not there are in vivo differences in dopaminergic neurotransmission in CMS-exposed rats. The development and use of in vivo microdialysis in recent years has provided an invaluable research tool for the measurement of levels of neurotransmission in awake, freely-moving animals (Tossman et al. 1986; Stamford and Justice 1996). When the technique is combined with sample analysis by high performance liquid chromatography with electrochemical detection (HPLC-ED), it allows for the specific and sensitive measurement of levels of monoamines and their metabolites in discrete brain regions (Di Ciano et al. 1995). The purpose of the present experiment was thus to determine if rats that are exposed to CMS for a chronic duration exhibit changes in basal levels of dopaminergic transmission, as well as in response to a rewarding stimulus. As a caveat, however, several recent studies have shown that dopamine release within the nucleus accumbens is linked not only to the incentive qualities of a consumed stimulus, but also the amount of the stimulus that is consumed (Wilson et al. 1995; Martel and Fantino 1996). Because we are interested in the former rather than the latter index of dopamine function, all animals were allowed to consume only a small, fixed quantity (5ml) of 1% sucrose, to avoid a potential confound of reduced fluid consumption in CMS-exposed animals. 68 Methods: Subjects Twenty male Long-Evans rats (Charles River, Quebec) weighing 250-3OOg at the start of the experiment, were housed individually in a temperature controlled colony (21\u00C2\u00B11\u00C2\u00B0C) under a 12-hr light-dark cycle (lights on at 07:00 hrs). All experiments were conducted in accordance with the Canadian Council on Animal Care guidelines for work with laboratory animals. Surgery Rats were anaesthetised with ketamine hydrochloride (100 mg/kg, i.p.) and xylazine (10 mg/kg, i.p.), and placed in a standard stereotaxic apparatus. The dorsal surface of the skull was exposed and holes were drilled bilaterally over the nucleus accumbens (AP = +1.7mm from bregma, M L = \u00C2\u00B1 1.2mm from midline), and two guide cannulae (19 guage, 15mm) were implanted bilaterally to a DV depth of-1.0mm from dura. The guide cannulae were secured to the skull with surgical stainless steel screws and dental acrylic. The guide cannulae were permanently obdurated until the insertion of the microdialysis probes. Animals were allowed to recover from surgery for at least one week before behavioural manipulation began. Chronic Mild Stress Animals were randomly divided into two groups (n=10 per group). One group remained as control subjects, while the other group was transferred to a separate colony and exposed to the CMS procedure as described by Willner et al. (1987; 1991). Three animals from the CMS-exposed group eventually had to be excluded from the study, as the guide cannulae of these 69 subjects became irreparably damaged by the daily exposure to CMS. The basic CMS protocol was as follows: two 18-hr periods of paired housing; two 20-hr periods of 40\u00C2\u00B0 cage tilt; two 18-hr periods of damp bedding; one period of overnight illumination and four 2-hr sessions of white noise (> 75 dB). All stressors were applied pseudo-randomly throughout the week to make them less predictable, and all stressors were terminated at least two hours before any behavioural testing occurred. Behavioural Training and Testing Following the second week of exposure to CMS, both groups of animals were given a 24 hr exposure period to a 1% sucrose solution in their home cages. The following day, animals were then allowed a 30 min habituation session in the microdialysis test chambers, during which the fluid spout was withdrawn. From this time hence, animals were given once-weekly exposure to a small (5 ml), fixed volume of 1% sucrose solution in the microdialysis test chambers, which they were allowed to consume over a 1 hr period. This volume represents an amount of liquid which is less than that normally drunk by either CMS-exposed or control animals, as determined by results obtained from either our or others' laboratories (Willner 1987; Phillips and Barr 1997; Barr and Phillips 1998). The purpose of restricting the animals' consumption to a small fixed volume was to ensure that any observable differences in neurotransmission would not simply reflect different volumes consumed by the two groups (Wilson et al. 1995; Martel and Fantino 1996). After the sixth week of exposure to CMS, both groups of animals were implanted with microdialysis probes and left overnight in the microdialysis test chambers. The following day, after stable levels of dopamine had been obtained, rats were allowed to consume a fixed (5 ml) volume of 1% sucrose solution over a 1 hr period. Dialysate samples were collected for nine 70 successive samples (each sample was collected every 15 minutes) subsequent to the presentation of the sucrose. Microdialysis Procedure The microdialysis probes used in this procedure were constructed according to the design of Fiorino et al. (1993). In brief, the probes were of a concentric style, consisting of a stainless-steel cannula (24 gauge, 34mm), polyethylene tubing (PE 50, Intramendic, Boston MA), fused silica tubing (PolymicroTechnologies Inc., Phoenix AZ) and a semi-permeable hollow-fiber membrane at the distal end (65kDa cutoff weight, 340 pm o.d., 2mm exposed membrane, Filtral 12, Germany). Probes were perfused at 1.0 ul/min with a modified Ringer's solution (0.01 M sodium phosphate buffer, pH 7.4, 1.3 mM CaCl 2 , 3.0 mM KC1, 1.0 mM MgCl 2 , 147 mM NaCl) using a gastight syringe (Hamilton, Reno NV) and a syringe pump (Harvard Apparatus, South Natick MA). Probes were secured by a guide collar to the guide cannula, and a steel coil that was attached to a liquid swivel (Instech, PA) was used to protect the probe. Microdialysis analytes, which typically consisted of a volume of 15ul, were separated by reverse-phase chromatography (Ultrasphere column, Beackman, Fullerton CA, ODS 5pm, 15cm, 4.6mm, inner diameter). The mobile phase consisted of 6g/L of sodium acetate, 70 mg/L of sodium octyl sulfate, 10 mg/L EDTA, 35mL/L of glacial acetic acid, which was pH adjusted to 3.5. 10% degassed methanol was added to the mobile phase prior to use. Analyte concentrations were quantified by electrochemical detection, using a Coulochem II EC detector (ESA, Bedford MA); the working potentials were +450mV (electrode 1), -300mV (electrode 2) and +450mV (guard cell). Chromatograms were recorded on a dual-pen chart recorder (Kipp and Zonen, 71 Bohemia NY). The typical probe recoveries were 22% for dopamine and 18% for DOPAC and H V A . Histology After the completion of the microdialysis procedure, each animal was given an overdose of chloral hydrate and perfused intracardially with saline and 4% formalin. The animals' brains were removed and stored in formalin, until they were frozen and cut into coronal sections (50um) for verification of probe placements. Only those probe placements that were within the nucleus accumbens (either core or shell) were used in further analysis. Results: The concentration of analyte dopamine, its metabolites DOPAC and H V A , and the 5-HT metabolite 5-HIAA were determined by measuring the peak heights on chromatographs and interpolating them on a standardised calibration curve. The mean values (\u00C2\u00B1 SEM) for basal levels of dopamine in the nucleus accumbens in CMS-exposed and control animals were 2.71 \u00C2\u00B1 0.40 vs 1.66 \u00C2\u00B1 0.40 nM, with no significant difference between the two groups [t(l,15) = 1.311, NS]. Similarly, the basal concentrations of both metabolites were not significantly different between the groups, with a mean value for DOPAC of 0.729 \u00C2\u00B1 0.075 uM for CMS-exposed rats vs 0.594 \u00C2\u00B1 0.075 uM for controls [t(l,15) = 0.881, NS], and for H V A the values were 0.221 \u00C2\u00B1 0.018 uM (CMS) vs 0.177 \u00C2\u00B1 0.018 uM (controls) [t( 1,15) = 1.217, NS]. The basal levels of the 72 serotonin metabolite 5-HIAA in the two groups were 0.219 \u00C2\u00B1 0.011 uM (CMS) vs 0.183 \u00C2\u00B1 0.011 uM (controls), which were not significantly different from each other [t(l,15) = 1.644, NS]. As basal levels of dopamine and the three metabolites were not significantly different between the groups, the levels of these compounds after presentation of the sucrose solution were expressed as percentages of the average basal values for each individual animal. These data were analysed by a one-factor repeated-measures A N O V A , with stress condition as the factor, measured across time [Figure 3.1]. The results of the A N O V A indicated that there was a significant group effect on levels of dopamine after presentation of the 1% sucrose solution [F(l,15) = 5.92,/? < 0.05], as well as a significant effect of time [F(8,120) = 3.07,/? < 0.05]. Post-hoc analysis using tests of simple main effects revealed that CMS-exposed animals exhibited significantly higher levels of dopamine release at different time points for up to 2Vz hours after the presentation of the 1% sucrose solution [Figure 3.1 A]. Analysis of the levels of the dopamine metabolite DOPAC with A N O V A failed to observe a significant group effect [F(l,15) = 0.771, NS] [Figure 3.IB]. Similarly, there was no significant group effect upon levels of the dopamine metabolite H V A either [F(l,15) = 0.331, NS] [Figure 3.1C]. The serotonin metabolite 5-HIAA was lower in CMS-exposed animals than in controls after presentation of the 1% sucrose solution, an effect that was marginally significant [F(l,15) = 3.18, 0.05

3 sees) in consumption of the sucrose solution: these results do not resemble the typical pattern of a uniform increase in inter-lick intervals, without pauses, seen in rats that are administered doses of neuroleptics sufficient to generate motor impairments (Fowler and Mortell 1992). The decrease in break points observed in d-amphetamine-treated subjects in withdrawal may represent a reduction in energy allocation and operant response maintenance (anergia) as is observed following decreased function in the mesoaccumbens and mesostriatal dopamine systems (Salamone 1992). This explanation is unlikely for several reasons. Previous studies which have investigated the impact of different variables in PR paradigms have found that break points, in untreated rats, are far more sensitive to alterations in the reinforcing value of the food reward (such as reward size) than changes in effort requirements (such as the height of the response lever) (Skjoldager et al. 1993). Additionally, a recent study by Sokolowski and Salamone (1998) has demonstrated that dopamine depletions in the nucleus accumbens only induce deficits in operant behaviour on schedules which generate high levels of responding, such as an FR5 schedule, but not in schedules producing moderate levels of responding, including a VI 30 schedule. Subjects in the present study had 1 hour to attain each reinforcement, and break points in withdrawn rats were relatively low, suggesting that the current operant schedule placed only a low energy demand on subjects. Evidence from the present study also implies that rats were not anergic, because the d-amphetamine-treated rats did not show an increase in the post-reinforcement pause or a decrease in their rate of responding, as measured by latency to attain each successive reinforcement. Therefore these data indicate that the d-amphetamine-treated and 100 control subjects responded with equal vigor. Other recent data from our laboratory have also shown that J-amphetamine withdrawal impaired certain motivational components of male rat sexual behaviour, whereas these rats displayed high levels of physically demanding copulatory activity for a 25 min period, consistent with an absence of either motoric or anergic deficits (Barr etal. 1998). The PR procedure has been used previously in our laboratory to show that reductions of the concentration of the sucrose solution, or a decrease in the level of food and water deprivation, both produced corresponding decreases in break points (Barr and Phillips 1998). This procedure therefore provides a reliable technique for assessing changes in motivation to respond for a natural reward, and hence the present data strongly imply that following J-amphetamine withdrawal rats experience significant reductions in their motivation to obtain a previously preferred reward. This reduced motivation may correspond to what Berridge and his colleagues (Berridge and Valenstein 1991, Robinson and Berridge 1993, Berridge 1996) refer to as \"wanting\", as distinct from \"liking\", which is more closely related to alternative hedonic processes. Indeed, in a recent study (Potts et al. 1997) depressed patients with anhedonic symptoms showed the same ability to discriminate the sensory qualities of sweet solutions as non-depressed control subjects. In the present study, high levels of motivation were not required to maintain the reflexive lick response necessary to consume the freely available sucrose solution, whereas in the PR paradigm the motivation to attain the next reinforcement had to be maintained for up to one hour. We interpret the lack of effect of drug withdrawal on the free consumption of sucrose as evidence for normal hedonic processes and attribute a motivational deficit to the finding that these rats were unable to maintain the level of responding required to maintain higher ratio reinforcements. Two other behavioural measures recorded during the 101 experiment provide additional support for this hypothesis. The increase in latencies taken by subjects to begin responding once the test sessions had begun, as well as the pauses between bouts of licking, are consistent with a decrease in motivation to obtain the sucrose reward. Previous experiments have examined the effects of a similar regimen of drug administration on ICSS responding and observed results consistent with those from this study. In an early study, Leith and Barrett (1976) demonstrated that amphetamine withdrawal depressed the facilitation of ICSS responding normally seen by low doses of d-amphetamine, and that this effect lasted for approximately four days. Similarly, Cassens et al (1981) observed an increase in the current intensities required to sustain ICSS responding, with a return to baseline within 120 hrs after the final drug injection. Thus, there appears to be little difference between the effects of amphetamine withdrawal on ICSS responding and operant responding for naturally rewarding stimuli such as a sucrose solution. The present findings therefore provide additional support for the use of ICSS paradigms to investigate changes in the responsiveness of neural reward systems in models of drug withdrawal and depression (Geyer and Markou 1995, Koob 1995). In a related study, Markou and Koob (1991) employed a drug self-administration procedure to establish post-cocaine anhedonia measured by a significant increase in ICSS current threshold. Current thresholds were elevated for 5 days post-drug treatment. Pretreatment with the tricyclic antidepressant desmethylimipramine returned ICSS thresholds to pre-drug values 12 hours after cessation of cocaine self-administration. The post-drug depression of ICSS responding associated with ^-amphetamine withdrawal has also been shown to be responsive to the tricyclic antidepressants imipramine and amitriptyline. Kokkinidis et al. (1980) demonstrated a mitigation of the effects of a ten day amphetamine regimen by these tricyclics, when ICSS responding at sites in the substantia nigra 102 was measured. It is not known how effective other classes of antidepressant drugs, such as the selective serotonin reuptake inhibitors, or other antidepressant therapies (such as electro-convulsive therapy (White and Barrett 1981) or R E M sleep deprivation) might be in animal models of psychostimulant withdrawal-induced anhedonia. It would therefore be of interest to examine the effects of antidepressant treatment on the suppressed responding for a sucrose solution, on a PR schedule, following the escalating dose of d-amphetamine protocol used in the present experiment. One problem facing such research is the short duration of the anhedonic effects which are typically observed in animals. With most drug regimens producing observable effects for only a few days, and at most a couple of weeks, this may not allow for the development of a theoretically viable model with which to examine the effects of many antidepressant treatments, which require a minimum of two to three weeks before changes in mood are seen in humans (Post et al. 1987) and certain animal models of depression (Willner et al. 1987). Much important research has focused on individual differences in susceptibility to drugs of abuse (Piazza and LeMoal 1996), however to-date little has been done to identify factors which might predispose certain subjects to display more prolonged and severe psychostimulant withdrawal symptoms. Identification of such factors might lead to the development of an animal model of anhedonia in which the observable effects last for notably longer than in current models. Alternatively, repeated bingeing schedules such as those engaged in by human cocaine addicts (Gawin and Kleber 1986; 1988), or dosing with other types of psychostimulants (such as with M D M A , which has produced long-lasting deficits in monkeys on PR performance for natural rewards (Frederick et al. 1995)) may also provide opportunities for the generation of a longer duration animal model of anhedonia. 103 In conclusion, the present study has provided additional support for animal models of psychostimulant withdrawal-induced anhedonia by observing reductions in motivation for a natural reward by rats, as assessed by performance on a PR schedule of reinforcement. The duration of the effects, including measures of latency and inter-lick intervals, lasted approximately the same duration as previously reported alterations of ICSS responding, and so provide further support for ICSS techniques as a tool for the measurement of reinforcement. Further developments in the use of the PR paradigm as an instrument for assessing anhedonia in psychostimulant withdrawal may be related to its capacity to measure reversals of drug induced anhedonia by antidepressant treatments. 104 EXPERIMENT 5: EFFECTS OF WITHDRAWAL FROM A N ESCALATING DOSE SCHEDULE OF D-AMPHETAMINE ON SEXUAL BEHAVIOR IN T H E M A L E RAT Synposis: The present study sought to determine the effect of withdrawal from an escalating dose schedule of ^-amphetamine on sexual behavior in male rats. Rats were introduced into testing chambers and, after a 5-min period during which locomotor activity was monitored, receptive females were presented. Copulation was observed for 25 min. Tests were conducted every 5 days until stable levels of sexual behavior were obtained. Upon repeated testing, male rats displayed an increase in the amount of activity associated with the anticipation of an estrous female. Half of the male rats were then subjected to a four day, escalating dose schedule of /^-amphetamine administration, while half received vehicle. 12 hours after the final drug injection, subjects were tested for sexual behavior. Withdrawal from the drug was associated with decrements in several motivational components of sexual behavior, including decreased anticipatory locomotor activity and increased post-ejaculatory intervals, while consummatory measured remained largely unaffected. This pattern of sexual deficits resembles those seen in human depressive disorders, and therefore provides additional support for the use of psychostimulant withdrawal as a rodent model of depression. 105 Introduction: Withdrawal from binge-like doses of various psychostimulant drugs, including cocaine and J-amphetamine, reliably induces a state of dysphoria in humans (Satel et al. 1991; Volkow et al. 1991; Pathiraja et al. 1995). This state is characterized by symptoms which include anhedonia, anergia and anxiety, and may last from weeks to months (Gawin and Kleber 1986; 1988). The symptoms associated with psychostimulant withdrawal bear a strong resemblance to those of major depressive disorder (MDD), to such an extent that the DSM-IV contains a specific category for the diagnosis of substance-induced mood disorders to differentiate them from isomorphic endogenous depression (American Psychiatric Association 1994). Administration of binge-like doses of psychostimulants to animals, followed by a period of drug withdrawal, produces many of the same symptoms that are observed in humans who experience drug withdrawal (Markou and Koob 1991). Extensive research has demonstrated that withdrawal from either cocaine or J-amphetamine in rodents generates a period of anhedonia, which is most commonly assessed by alterations in responding for reinforcing electrical brain self-stimulation (Leith and Barrett 1980; Wise and Munn 1995). Decreased locomotor activity is another commonly observed effect of withdrawal from psychostimulants in rats (Hitzemann et al. 1977), and may model anergia, while withdrawal-induced anxiety has also been demonstrated through the use of sophisticated behavioral paradigms (Mutschler and Miczek 1998). This combination of withdrawal-associated behavioral sequelae and their resemblance to the symptoms of MDD in humans has led to the utilization of psychostimulant withdrawal as a rodent model of depression. 106 Most animal paradigms of depression are developed with the hope that they will accurately model many of the major symptoms of MDD. One widely reported human symptom of M D D is a decrease in sexual behavior, noted as either a loss of libido or problems in sexual functioning (Casper et al. 1985; American Psychiatric Association 1994; Baldwin 1996). Despite the high incidence with which this symptom occurs in human depressives, there have been relatively few studies in which animal models of depression have been used to examine alterations in the sexual behavior of sexually experienced subjects. Several deficits in both the appetitive and consummatory components of sexual behavior have been reported in rats with limited or no copulatory experience using other models of depression (Edwards et al. 1990; Neill et al. 1990; D'Aquila et al. 1994; Vogel et al. 1996;. In the present study, the psychostimulant-withdrawal model of depression was evaluated by using rats that were well-trained and exhibited high levels of sexual activity before the drug treatment conditions were implemented. Although the use of sexually naive rats is valid, the study of sexual behavior in experienced rats in this context has several advantages. First, it increases the face validity of a drug induced model of depression, because many depressed patients have had sexual experience prior to the time of their period of sexual dysfunction -emphasis should be placed on studying the disturbance of established sexual behavior, rather than its acquisition. Second, the activity associated with the anticipation of a receptive female in well-trained rats can provide a valuable measure of sexual motivation. A number of reports have demonstrated the utility of this approach by using bilevel chambers for the study of male sexual behavior (Mendelson and Pfaus 1989). An increase or decrease in the number of level changes made prior to the presentation of a receptive mate in these chambers is hypothesized to reflect an increase or decrease in motivation, respectively. 107 The present study assessed the psychostimulant-withdrawal model of depression by examining the effect of d-amphetamine withdrawal on sexual behavior in experienced male rats. A modified unilevel testing apparatus, analogous to the bilevel chamber, was employed in order to provide a more detailed evaluation of male rat sexual motivation. Methods. Male Long-Evans rats (225-250g; Charles River, QU) were group housed in wire cages (18 X 25 X 65 cm) and entrained on a reverse light-dark cycle (light off 7 a.m. - 7 p.m.). One week after arrival, male rats were housed individually in plastic cages with bedding (Carefresh, Absorption Corp., Bellingham, WA) for the remainder of the experiment. The colony room temperature was approximately 20\u00C2\u00B0C and rats had unlimited access to food (Purina Rat Chow) and water. Training and testing occurred during the middle third of the dark phase. Female rats, housed in a separate colony, were anesthetized with ketamine hydrochloride (80 mg/kg, i.p.) and xylazine (10 mg/kg, i.p.) and bilaterally ovariectomised at least four weeks prior to testing. Sexual receptivity in the stimulus females was induced by subcutaneous injections of estradiol benzoate (10 u.g) and progesterone (500 p.g), 48 and 4 h, respectively, before each test session. Throughout the experiment, male rats were tested every five days for sexual behavior in Plexiglas unilevel testing chambers. The chambers (24 x 32 x 48 cm) were fitted with a central Plexiglas partition (32 cm) creating a perimeter \"racetrack\" area around which the female could pace sexual behavior. Like the bilevel chambers, these chambers encourage the exposure of the 108 rats' flanks to the experimenter during copulation thereby facilitating scoring of sexual behavior (Mendelson and Pfaus 1989). Another advantage to this simple design is the relative ease of introducing and removing rats from the chamber. The chambers were equipped with infrared photobeam emitter/detector pairs (one beam across its length on each side of the partition; 3 beams across its width; 12 cm apart) to monitor activity. The number of side changes (SCs), counted as the complete movement from one side of the chamber to other side of the partition, were recorded automatically by computer (2 Hz scan rate). At the beginning of each test session, male rats were introduced into the chambers and activity was recorded for 5 min. Following the introduction of a sexually receptive female, copulatory behavior was monitored for 25 minutes. Each session was videotaped and subsequently scored for standard measures of sexual behavior using a computer and appropriate software (courtesy of Sonoko Ogawa). Testing chambers were washed thoroughly between tests to remove residual odors/pheromones. Once consistent measures of anticipatory and copulatory behavior were obtained (i.e. anticipatory SC (< 10% variation), intromission latencies (IL) < 5 min, and ejaculation latencies (EL) < 10 min for 3 consecutive days), male rats were assigned to either the sucrose solution. Analysis of the data during the 84 hr period following drug termination with the repeated-measures A N O V A indicated a significant main effect of drug treatment, [F(l,28) = 7.31, < 0.05], as animals that were exposed to the escalating dose regimen of d-amphetamine exhibited reduced consumption of the 4% sucrose solution, compared to vehicle-treated subjects. There was also a significant main effect of pre-shift sucrose solution [F(l,28) = 4.38, p < 0.05], whereby both groups of rats that were allowed to consume the 32% sucrose solution displayed dramatically reduced consumption of the 4%> solution after the animals were downshifted to this new reward, confirming a negative contrast effect. The A N O V A also 128 indicated a significant interaction of drug treatment x pre-shift solution x time [F(4,l 12) = 6.43, p< 0.001]. The significant interaction was analyzed further with the use of post-hoc tests [Figure 6.1]. These tests revealed that both of the downshifted groups (32%\u00E2\u0080\u0094>4%) displayed reduced consumption of the novel 4%> sucrose solution compared to the unshifted (4%%\u00E2\u0080\u0094>4%) groups during their first two exposures to the 4% solution. On the third exposure to the 4%> solution, at 60 hr after drug termination, the vehicle-treated group displayed an unexpected increase in-consumption of the 4%> solution, compared to both of the unshifted groups. This effect had diminished by the fourth test session, when there was no longer any difference between these groups. In comparison, the downshifted group that had been exposed to (/-amphetamine displayed reduced levels of consumption of the 4% solution for 3, as opposed to 2, test sessions, and returned to control levels of consumption by the fourth test session. When the two downshifted groups were compared to each other, the (/-amphetamine-treated group exhibited lower levels of consumption across the first three test sessions, although this effect was only marginally significant on the first test day. 129 Successive Negative Contrast 4-Day Drug Administration T B-Line ih \u00E2\u0080\u00A2 32%-4% (Amp) \u00E2\u0080\u00A2 32%-4% (Veh) 4%-4% (Veh) 4%-4% (Amp) 12 36 60 Time after drug administration (hr) 84 F i g 6.1 Effects of withdrawal from a 4-day regimen of ^-amphetamine, or vehicle, on number of licks for a 4% sucrose solution. Animals were given 5 min fluid consumption tests at different time points before (B-Line) and after withdrawal from drug administration. * = significantly different from 4%-4% (VEH) group, p < 0.05 = = significantly different from 32%-4% (VEH) group, p < 0.10 # = significantly different from 32%-4% (VEH) group, p < 0.05 130 Discussion: In the present experiment, we have demonstrated that rats exhibit a greater consumatory negative contrast compared to control subjects when they are tested after withdrawal from a binge-like regimen of J-amphetamine. This effect was manifested as a marginally significant increase in the size of the contrast effect on the first day of exposure to the devalued sucrose solution. By the second day of exposure to the devalued stimulus, the magnitude of the contrast effect was substantially greater between the downshifted groups, as vehicle treated animals showed a more rapid recovery from the exposure to the devalued stimulus. On the third day of exposure to the devalued sucrose solution, only the ^-amphetamine withdrawn animals continued to exhibit a contrast effect, indicating that withdrawal from a psychostimulant drug can perpetuate negative contrast effects in rodents. A number of different hypotheses have been proposed to account for the phenomenon of negative contrast effects. One of the most influential of such theories postulates that contrast effects arise from associative generalization decrements (Spear and Spitzer 1966; Capaldi 1972), whereby changes in either the rewarding environment or the rewarding stimulus lead to a reduced association between the two, with a commensurate decrease in consumption of the reward. Although generalization decrements fail to account for several important aspects of contrast effects, such as the existence of positive contrast effects (Flaherty 1996), it may be argued than in this case animals in the novel state of drug withdrawal would exhibit potent generalization decrements when exposed to the devalued stimulus for the first time. This explanation is unlikely, however, as the animals in the group that received 30 presses per min) indicated a significant main effect of drug treatment [F(l,33) = 4.58, p < 0.05], a significant effect of test session [F(3,99) = 10.17,/? < 0.001], and a significant interaction of drug treatment x test session [F(3,99) = 12.18,/? < 0.001]. Further analysis of these results with post-hoc tests revealed that 12 hr after drug termination, both the (/-amphetamine ECS and sham-ECS groups displayed significantly higher threshold current values than either of the vehicle-treated groups [Figure 7.1]. By 36 hr after drug termination, threshold currents for the d-amphetamine sham-ECS group displayed threshold currents that were significantly greater than currents for both of the vehicle-treated groups, and marginally greater than the (/-amphetamine ECS group (p = 0.081). At this time point, mean threshold currents for the (/-amphetamine ECS group did not differ from the vehicle-treated groups. On the final test session, at 60 hr after drug termination, threshold currents were significantly greater for the (/-amphetamine sham-ECS group than the other groups. 143 1200 E 900 g 600 co co 0 \u00C2\u00A3 300 \u00E2\u0080\u0094 \u00E2\u0080\u00A2 \u00E2\u0080\u0094 A M P - E C S \u00E2\u0080\u0094 \u00E2\u0080\u00A2 \u00E2\u0080\u0094 A M P - S H A M - - o - - V E H - E C S V E H - S H A M 8 10 12 14 16 18 20 22 24 26 28 1200 900 600 300 0 12 Hr Withdrawal - o- - o- _ , '' -7 > I S E M ,', -\u00C2\u00B0 /o7a THRESHOLD J ! ! L 8 10 12 14 16 18 20 22 24 26 28 1200 | C E 900 \u00E2\u0080\u00A2' THRESHOLD . 8 10 12 14 16 18 20 22 24 26 28 Current [\iA] 1200 900 600 300 0 60 Hr Withdrawal 'J^/a THRESHOLD i l i i i I S E M 8 10 12 14 16 18 20 22 24 26 28 Current (MA) Fig 7.1 Effects of withdrawal from a 4-day escalating dose schedule of "Thesis/Dissertation"@en . "2001-11"@en . "10.14288/1.0090835"@en . "eng"@en . "Neuroscience"@en . "Vancouver : University of British Columbia Library"@en . "University of British Columbia"@en . "For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use."@en . "Graduate"@en . "The development of an animal model of depression : a focus on anhedonia"@en . "Text"@en . "http://hdl.handle.net/2429/13495"@en .