- Email: [email protected]
Animal models of human behaviour: can they help in understanding pathology and in drug development?
$37
Behavioural Pharmacology
Lectures
F•
Social neuroscience: from genes to social behavior
T. Insel 1. 1National Institute of Mental Health, Room
8235, Bethesda, USA One of the most challenging questions for our understanding of the brain is how this complex organ of a few billion cells manages complex functions such as language, emotion, and consciousness. This presentation reviews a decade of research on the seemingly impossible task of trying to understand the brain mechanisms for complex behavior, specifically social attachment. This task becomes tractable because of two insights. First, there are many mammals which form long-term social attachments. Our studies have focused on prairie voles, a monogamous rodent. The second insight is that a family of neuropeptides, including oxytocin and vasopressin, seems to have an important role in social behavior across evolution. Vasopressin and oxytocin appear to be critical for the development of a long-term selective social bond in prairie voles. Under natural conditions, prairie voles form bonds only after mating. If oxytocin or vasopressin is given in the absence of mating, the voles will bond. More important, if prairie voles mate but are treated with a blocker of oxytocin or vasopressin, they fail to bond. Thus, it appears that these neuropeptides, normally released with mating, are necessary and sufficient for pair bonding. What is intriguing is that these neuropeptides have no such effect in other species of voles that are not monogamous. Prairie voles, as well as other monogamous species, have a unique distribution of brain receptors for oxytocin and vasopressin, such that these neuropeptides can influence reward pathways in the brains of monogamous species. The mechanism for these striking species differences in receptor distribution is not entirely clear, but may be related to a variation in the sequence of the regulatory part of the genes for these receptors. Results from transgenic, viral vector gene transfer, and comparative behavioral studies can now provide a model for species differences in social behavior. One implication of these results involves autism, a neurodevelopmental disorder characterized by a reduction
in social behavior and abnormal social attachment. There are reports of reduced oxytocin in children with autism. Perhaps more important is a recent finding that the same genetic region which discriminates social from non-social voles also appears to show variation in the human genome and that one version of this region is transmitted at a high rate in children with autism. We do not know yet whether the autistic brain has an altered distribution of vasopressin or oxytocin receptors. Nevertheless, a principle elucidated from the vole studies that the confluence of oxytocin or vasopressin circuits with the brain' s reward pathways is critical for social attachment deserves careful study in autism. In addition, the discovery that peptides in this family have "pro-social" effects across diverse taxa recommends an approach for developing treatments to increase social behavior in disorders with social deficits, including autism and schizophrenia.
References [1] Insel, T.R., and Young, L.J., 2001. The neurobiology of attachment. Nat Neurosci Rev 2, 129-36. [2] Insel, T.R., and Femald, R.D., 2004. How the brain processes social information. Ann Rev Neurosci 27, 697W22. [3] Hammock, E.A.D., and Young, L.J., 2005. Microsatellite instability generates diversity in brain and sociobehavioral traits. Science 308, 1630-4.
I•-]
Animal models of human behaviour: can they help in understanding pathology and in drug development?
R.J. Rodgers 1. 1 University of Leeds, Behavioural Neuroscience Laboratory, Institute of Psychological Sciences, Leeds, United Kingdom The 'translational' question as to whether animal models of human behaviour can help in understanding pathology and in drug development seems straightforward and is one to which the preclinical research scientist would be expected to answer in the affirmative. However, the precise formulation of the question should give pause for thought, not least because use of the word 'can' implies some skepticism on the part of the questioner. Rather than simply respond with righteous indignation, there are
s38
Behavioural Pharmacology
sound reasons why preclinical scientists working in the areas of anxiety and depression should at least empathise with this position. One only has to point to the history of false positives (e.g. 5-HT3 & CCK2 receptor antagonists) and false negatives (e.g. buspirone) in preclinical anxiety research and, more recently, to the discrepancy between the broad-spectrum efficacy of SSRI antidepressants in clinical anxiety disorders versus their general lack of efficacy in animal models of anxiety. Although the therapeutic success of SSRIs in anxiety disorders is consistent with patterns of comorbidity and with the concept of an anxiety depression continuum, it is inconsistent with the clearcut separation of the disorders in DSM classifications. As translational medicine is a two-way street, whereby preclinical research should not only inform but also be informed by clinical research, problems with current animal models clearly have more than one origin. That said, there are a number of reasons to be more optimistic about future prospects. There is little doubt that animal behaviour tests will remain the backbone of research on the basic neurobiology of emotional behaviour and that, in principle, knowledge so gained ought to be relevant both to affective disorders and drug development. General relevance derives from basic evolutionary theory, clear conm~onalities in affective behavioural expression in mammals, and strong evolutionary conservation of underlying neurocircuitry. On the other hand, we must recognise that animal behaviour tests become useful as 'models' only when they satisfy certain criteria. Almost 40 years ago, McKinney and Bunney emphasised that the dependent measure should be reasonably analogous in manifestation to an aspect of the focal human condition (i.e. face, possibly construct, validity), objectively measurable, reversible by clinically-effective compounds (i.e. predictive validity), and reproducible across laboratories. More recently, others have argued that while the first two criteria may be desirable, only the second two are actually essential. The key issue here seems to hinge upon what the model is used for, i.e., as a relatively quick and simple screening tool for novel compounds or as a more time-consuming but more accurate simulation of a particular human condition. I am not at all convinced by this pragmatic dichotomy, not least because of the heavy reliance of screening tests upon predictive validity. Operationally, though somewhat curiously, predictive validity refers to the ability of a given model to detect compounds already known to be clinically-effective and, when used in isolation from other forms of validation, seems to be a recipe for 'groundhog day' (i.e. high probability of identifying only 'me-same' compounds). Thus, a positive profile in a 'predictive only' model can really only be an entrance ticket to rigorous 'proof of concept' testing in more
theoretically-driven tests where the theory element is not so much pharmacological or neurochemical as it is psychological. This approach is already evident in depression research (moving from a positive profile in the reserpine hypothermia screen to assessment in a behavioural despair paradigm and, ultimately, a chronic stress paradigm). But even this strategy begs the question 'what if' , i.e., what if the initial screening test is sensitive only to compounds with a conventional mode of action? Such reasoning implies the need for continual refinement of screening tests as a function of the development of increasingly accurate simulation models. In this context, and recognising the symptomatic diversity of human psychopathologies, it seems clear that a given animal test will 'model' only one or (at best) a small subset of responses relevant to the focal human condition. It is for this reason that many researchers have begun to argue for greater specificity in the use of the term model (e.g. animal model of 'anhedonia' rather than of 'depression' ) or even its abandonment altogether. In the field of anxiety research, and despite a larger range of available procedures, animal models involve dependent measures that are normal, biologically-adaptive responses to threat and not abnormal, maladaptive or pathological responses. However, if one accepts that anxiety disorders can be viewed as the exaggeration and/or inappropriate activation of normally adaptive response patterns, knowledge obtained regarding the neurobiology of the latter may provide crucial insights into the aetiology and treatment of the fom~er. This would be fine were it not for the general insensitivity of existing anxiety models to certain clinically-effective compounds (e.g. SSRIs) and, paradoxically, the plethora of novel therapeutic targets identified by the very same tests. Does this confusion mean that there is something fundamentally wrong with the models themselves or does the problem lie with the conventional use of the tests, namely interchangeably as models of something called 'anxiety' ? Although all animal models of anxiety-like behaviour involve defensive responses to threat, the precise nature of the response (i.e. freezing, fleeing, avoidance, risk assessment, burying, ultrasonic vocalisations) differs from one test to another. As such, a convincing case can be made for redirecting attention to the behavioural differences (rather than similarities) between the various models. These differences could in turn permit a clearer focus on those responses that have most relevance for specific dimensions of the various DSM-IV anxiety disorders. Finally, returning to the 'understanding pathology' component of the question, there is little doubt that certain animal models have led to significant progress in relation both to depression (e.g. psychosocial stress, HPA axis, hippocampal dysfunction) and anxiety disorders
Behavioural Pharmacology (e.g. amygdaloid neurocircuitry, fear learning & memory). However, further progress, especially in anxiety research, requires much greater emphasis on theory-driven (as opposed to almost random observation-driven) research. Examples would include differences in patterns of gene expression between strains or lines selected for high and low incidence of specific defense responses, as well as theoretically-focused studies on the effects of genetic and/or environmental manipulations on a range of clinically-relevant defense responses.
Posters
FP~
Explaining birth seasonality in fall-winter depression: idiosyncratic procreational patterns in the patients' parents
E. Pjrek 1, D. Winkler 1, A. Konstantinidis 1, J. Stastny 1, M. Willei0, N. Praschak-Rieder 1, S. Kasper 1. 1Medical
University of Vienna, Department of General Psychiatry, Vienna, Austria Purpose: Recently we have published a study examining the seasonal birth pattern of a cohort of 553 patients suffering from seasonal affective disorder (SAD) [1]. We have found a birth excess in spring and summer as well as a lack of births especially in the first quarter of the year. The most frequently offered interpretation is that season-related factors may have a deleterious effect on the intrauterine development of the central nervous system, thus increasing the risk for certain psychiatric disorders. Some reports have also explained the phenomenon of birth seasonality with an idiosyncratic seasonal conception pattern of the parents of psychiatric patients [2]. The aim of the present study was to test this hypothesis in our sample of SAD patients. Methods: We conducted a telephone interview with the patients to obtain information on the birth months of their siblings. 166 patients could not be reached by phone, 13 patients refused to participate in the study and 2 patients had already died. We employed measures to ensure a high reliability of the sampled data: 34 of 372 cases were eliminated from the evaluation, because patients were not able to remember the full set of birth data of their siblings. Of the remaining patients 267 had at least one sibling. Using the method of chart review to acquire information on the family history of our patients, we were able to exclude 101 siblings (18.8%) with known psychiatric disorders from our evaluation. We first compared the birth months and the quarters of birth of 435 unaffected siblings with that of the general population. Secondly, we compared the birth distribution of our patients with that of their healthy sibs.
$39
Results: There was a significant deviation between the birth distribution of our unaffected siblings and the general population calculated on a monthly basis (X2=20.171, d f = l l , p=0.043). When comparing quarters we found less births than expected in the first (-12.8%) and fourth (-16.1%) quarters of the year and an excess of births in the second (+8.6%) and third (+19.4%; X2= 9.487, dr'=3, p = 0.023). There were no significant differences between the group of SAD patients and their sibs regarding their birth pattern as calculated by months (X2 = 6.364, df = 11, p = 0.848) or quarters (X2 = 3.509, df = 3, p = 0.320). Conclusions: Our results demonstrate that the group of unaffected siblings show a birth pattern much closer to the SAD group than to the general population, which provides further support for the procreational habits hypothesis. However, as other studies have also found some evidence for the importance of seasonally varying environmental factors (such as viral agents) it is likely that seasonal variations of birth frequency are influenced by multiple factors. Further research should be aimed to replicate our findings in non-seasonal depression or bipolar disorder as there is still a lack of studies on birth seasonality in affective disorders.
References [1] Pjrek E, Winkler D, Heiden A, Praschak-Rieder N, Willeit M, Konstantinidis A, Stastny J, Kasper S, 2004. Seasonality of birth in seasonal affective disorder. J Clin Psychiatry 65, 1389-93. [2] Bleuler M, 1991. Seasonality of birth of future schizophrenics and seasonality of depressive episodes. Schizophr Bull 17, 191-2.
F•-•
Long lasting effects of anti-depressants on olfactory bulbectomy induced hyperactivity in the open field
M.E. Breuer 1, L. Groenink 2, R.S. Oosting 2, B. Olivier2.
1Utrecht University and Rudoph Magnus Institute of Neuroscience, Departments of Psychopharmacology, Pharmacology and Anatomy, Utrecht, The Netherlands; 2 Utrecht University, Department of Psychopharmacology, Utrecht, The Netherlands Purpose: There are currently several animal models in existence that are useful for depression research. However, OBX is one of the few models in which delayed drug effects (as opposed to acute) can be studied. Removal of the olfactory bulbs results in several physiological and behavioral alterations that are favorable in depression paradigms. The operation often results in neuroendocrine and immunological changes reminiscent of those observed
Behavioural Pharmacology
Lectures
F•
Social neuroscience: from genes to social behavior
T. Insel 1. 1National Institute of Mental Health, Room
8235, Bethesda, USA One of the most challenging questions for our understanding of the brain is how this complex organ of a few billion cells manages complex functions such as language, emotion, and consciousness. This presentation reviews a decade of research on the seemingly impossible task of trying to understand the brain mechanisms for complex behavior, specifically social attachment. This task becomes tractable because of two insights. First, there are many mammals which form long-term social attachments. Our studies have focused on prairie voles, a monogamous rodent. The second insight is that a family of neuropeptides, including oxytocin and vasopressin, seems to have an important role in social behavior across evolution. Vasopressin and oxytocin appear to be critical for the development of a long-term selective social bond in prairie voles. Under natural conditions, prairie voles form bonds only after mating. If oxytocin or vasopressin is given in the absence of mating, the voles will bond. More important, if prairie voles mate but are treated with a blocker of oxytocin or vasopressin, they fail to bond. Thus, it appears that these neuropeptides, normally released with mating, are necessary and sufficient for pair bonding. What is intriguing is that these neuropeptides have no such effect in other species of voles that are not monogamous. Prairie voles, as well as other monogamous species, have a unique distribution of brain receptors for oxytocin and vasopressin, such that these neuropeptides can influence reward pathways in the brains of monogamous species. The mechanism for these striking species differences in receptor distribution is not entirely clear, but may be related to a variation in the sequence of the regulatory part of the genes for these receptors. Results from transgenic, viral vector gene transfer, and comparative behavioral studies can now provide a model for species differences in social behavior. One implication of these results involves autism, a neurodevelopmental disorder characterized by a reduction
in social behavior and abnormal social attachment. There are reports of reduced oxytocin in children with autism. Perhaps more important is a recent finding that the same genetic region which discriminates social from non-social voles also appears to show variation in the human genome and that one version of this region is transmitted at a high rate in children with autism. We do not know yet whether the autistic brain has an altered distribution of vasopressin or oxytocin receptors. Nevertheless, a principle elucidated from the vole studies that the confluence of oxytocin or vasopressin circuits with the brain' s reward pathways is critical for social attachment deserves careful study in autism. In addition, the discovery that peptides in this family have "pro-social" effects across diverse taxa recommends an approach for developing treatments to increase social behavior in disorders with social deficits, including autism and schizophrenia.
References [1] Insel, T.R., and Young, L.J., 2001. The neurobiology of attachment. Nat Neurosci Rev 2, 129-36. [2] Insel, T.R., and Femald, R.D., 2004. How the brain processes social information. Ann Rev Neurosci 27, 697W22. [3] Hammock, E.A.D., and Young, L.J., 2005. Microsatellite instability generates diversity in brain and sociobehavioral traits. Science 308, 1630-4.
I•-]
Animal models of human behaviour: can they help in understanding pathology and in drug development?
R.J. Rodgers 1. 1 University of Leeds, Behavioural Neuroscience Laboratory, Institute of Psychological Sciences, Leeds, United Kingdom The 'translational' question as to whether animal models of human behaviour can help in understanding pathology and in drug development seems straightforward and is one to which the preclinical research scientist would be expected to answer in the affirmative. However, the precise formulation of the question should give pause for thought, not least because use of the word 'can' implies some skepticism on the part of the questioner. Rather than simply respond with righteous indignation, there are
s38
Behavioural Pharmacology
sound reasons why preclinical scientists working in the areas of anxiety and depression should at least empathise with this position. One only has to point to the history of false positives (e.g. 5-HT3 & CCK2 receptor antagonists) and false negatives (e.g. buspirone) in preclinical anxiety research and, more recently, to the discrepancy between the broad-spectrum efficacy of SSRI antidepressants in clinical anxiety disorders versus their general lack of efficacy in animal models of anxiety. Although the therapeutic success of SSRIs in anxiety disorders is consistent with patterns of comorbidity and with the concept of an anxiety depression continuum, it is inconsistent with the clearcut separation of the disorders in DSM classifications. As translational medicine is a two-way street, whereby preclinical research should not only inform but also be informed by clinical research, problems with current animal models clearly have more than one origin. That said, there are a number of reasons to be more optimistic about future prospects. There is little doubt that animal behaviour tests will remain the backbone of research on the basic neurobiology of emotional behaviour and that, in principle, knowledge so gained ought to be relevant both to affective disorders and drug development. General relevance derives from basic evolutionary theory, clear conm~onalities in affective behavioural expression in mammals, and strong evolutionary conservation of underlying neurocircuitry. On the other hand, we must recognise that animal behaviour tests become useful as 'models' only when they satisfy certain criteria. Almost 40 years ago, McKinney and Bunney emphasised that the dependent measure should be reasonably analogous in manifestation to an aspect of the focal human condition (i.e. face, possibly construct, validity), objectively measurable, reversible by clinically-effective compounds (i.e. predictive validity), and reproducible across laboratories. More recently, others have argued that while the first two criteria may be desirable, only the second two are actually essential. The key issue here seems to hinge upon what the model is used for, i.e., as a relatively quick and simple screening tool for novel compounds or as a more time-consuming but more accurate simulation of a particular human condition. I am not at all convinced by this pragmatic dichotomy, not least because of the heavy reliance of screening tests upon predictive validity. Operationally, though somewhat curiously, predictive validity refers to the ability of a given model to detect compounds already known to be clinically-effective and, when used in isolation from other forms of validation, seems to be a recipe for 'groundhog day' (i.e. high probability of identifying only 'me-same' compounds). Thus, a positive profile in a 'predictive only' model can really only be an entrance ticket to rigorous 'proof of concept' testing in more
theoretically-driven tests where the theory element is not so much pharmacological or neurochemical as it is psychological. This approach is already evident in depression research (moving from a positive profile in the reserpine hypothermia screen to assessment in a behavioural despair paradigm and, ultimately, a chronic stress paradigm). But even this strategy begs the question 'what if' , i.e., what if the initial screening test is sensitive only to compounds with a conventional mode of action? Such reasoning implies the need for continual refinement of screening tests as a function of the development of increasingly accurate simulation models. In this context, and recognising the symptomatic diversity of human psychopathologies, it seems clear that a given animal test will 'model' only one or (at best) a small subset of responses relevant to the focal human condition. It is for this reason that many researchers have begun to argue for greater specificity in the use of the term model (e.g. animal model of 'anhedonia' rather than of 'depression' ) or even its abandonment altogether. In the field of anxiety research, and despite a larger range of available procedures, animal models involve dependent measures that are normal, biologically-adaptive responses to threat and not abnormal, maladaptive or pathological responses. However, if one accepts that anxiety disorders can be viewed as the exaggeration and/or inappropriate activation of normally adaptive response patterns, knowledge obtained regarding the neurobiology of the latter may provide crucial insights into the aetiology and treatment of the fom~er. This would be fine were it not for the general insensitivity of existing anxiety models to certain clinically-effective compounds (e.g. SSRIs) and, paradoxically, the plethora of novel therapeutic targets identified by the very same tests. Does this confusion mean that there is something fundamentally wrong with the models themselves or does the problem lie with the conventional use of the tests, namely interchangeably as models of something called 'anxiety' ? Although all animal models of anxiety-like behaviour involve defensive responses to threat, the precise nature of the response (i.e. freezing, fleeing, avoidance, risk assessment, burying, ultrasonic vocalisations) differs from one test to another. As such, a convincing case can be made for redirecting attention to the behavioural differences (rather than similarities) between the various models. These differences could in turn permit a clearer focus on those responses that have most relevance for specific dimensions of the various DSM-IV anxiety disorders. Finally, returning to the 'understanding pathology' component of the question, there is little doubt that certain animal models have led to significant progress in relation both to depression (e.g. psychosocial stress, HPA axis, hippocampal dysfunction) and anxiety disorders
Behavioural Pharmacology (e.g. amygdaloid neurocircuitry, fear learning & memory). However, further progress, especially in anxiety research, requires much greater emphasis on theory-driven (as opposed to almost random observation-driven) research. Examples would include differences in patterns of gene expression between strains or lines selected for high and low incidence of specific defense responses, as well as theoretically-focused studies on the effects of genetic and/or environmental manipulations on a range of clinically-relevant defense responses.
Posters
FP~
Explaining birth seasonality in fall-winter depression: idiosyncratic procreational patterns in the patients' parents
E. Pjrek 1, D. Winkler 1, A. Konstantinidis 1, J. Stastny 1, M. Willei0, N. Praschak-Rieder 1, S. Kasper 1. 1Medical
University of Vienna, Department of General Psychiatry, Vienna, Austria Purpose: Recently we have published a study examining the seasonal birth pattern of a cohort of 553 patients suffering from seasonal affective disorder (SAD) [1]. We have found a birth excess in spring and summer as well as a lack of births especially in the first quarter of the year. The most frequently offered interpretation is that season-related factors may have a deleterious effect on the intrauterine development of the central nervous system, thus increasing the risk for certain psychiatric disorders. Some reports have also explained the phenomenon of birth seasonality with an idiosyncratic seasonal conception pattern of the parents of psychiatric patients [2]. The aim of the present study was to test this hypothesis in our sample of SAD patients. Methods: We conducted a telephone interview with the patients to obtain information on the birth months of their siblings. 166 patients could not be reached by phone, 13 patients refused to participate in the study and 2 patients had already died. We employed measures to ensure a high reliability of the sampled data: 34 of 372 cases were eliminated from the evaluation, because patients were not able to remember the full set of birth data of their siblings. Of the remaining patients 267 had at least one sibling. Using the method of chart review to acquire information on the family history of our patients, we were able to exclude 101 siblings (18.8%) with known psychiatric disorders from our evaluation. We first compared the birth months and the quarters of birth of 435 unaffected siblings with that of the general population. Secondly, we compared the birth distribution of our patients with that of their healthy sibs.
$39
Results: There was a significant deviation between the birth distribution of our unaffected siblings and the general population calculated on a monthly basis (X2=20.171, d f = l l , p=0.043). When comparing quarters we found less births than expected in the first (-12.8%) and fourth (-16.1%) quarters of the year and an excess of births in the second (+8.6%) and third (+19.4%; X2= 9.487, dr'=3, p = 0.023). There were no significant differences between the group of SAD patients and their sibs regarding their birth pattern as calculated by months (X2 = 6.364, df = 11, p = 0.848) or quarters (X2 = 3.509, df = 3, p = 0.320). Conclusions: Our results demonstrate that the group of unaffected siblings show a birth pattern much closer to the SAD group than to the general population, which provides further support for the procreational habits hypothesis. However, as other studies have also found some evidence for the importance of seasonally varying environmental factors (such as viral agents) it is likely that seasonal variations of birth frequency are influenced by multiple factors. Further research should be aimed to replicate our findings in non-seasonal depression or bipolar disorder as there is still a lack of studies on birth seasonality in affective disorders.
References [1] Pjrek E, Winkler D, Heiden A, Praschak-Rieder N, Willeit M, Konstantinidis A, Stastny J, Kasper S, 2004. Seasonality of birth in seasonal affective disorder. J Clin Psychiatry 65, 1389-93. [2] Bleuler M, 1991. Seasonality of birth of future schizophrenics and seasonality of depressive episodes. Schizophr Bull 17, 191-2.
F•-•
Long lasting effects of anti-depressants on olfactory bulbectomy induced hyperactivity in the open field
M.E. Breuer 1, L. Groenink 2, R.S. Oosting 2, B. Olivier2.
1Utrecht University and Rudoph Magnus Institute of Neuroscience, Departments of Psychopharmacology, Pharmacology and Anatomy, Utrecht, The Netherlands; 2 Utrecht University, Department of Psychopharmacology, Utrecht, The Netherlands Purpose: There are currently several animal models in existence that are useful for depression research. However, OBX is one of the few models in which delayed drug effects (as opposed to acute) can be studied. Removal of the olfactory bulbs results in several physiological and behavioral alterations that are favorable in depression paradigms. The operation often results in neuroendocrine and immunological changes reminiscent of those observed