© Калуев А.В., Туохимаа П., 2004

PROBLEMS AND PERSPECTIVES

of experimental modeling of anxiety and depression

A.V. KALUEFF, P. TUOHIMAA

Medical School, Tampere University Hospital, University of Tampere, Tampere, Finland

Калуев А.В., Туохимаа П. Проблемы и перспективы экспериментального моделирования тревоги и депрессии // Психофармакол. и биол. наркол. 2004. Т. 4. № 2-3. С. 645-651. Медицинская школа и университетский госпиталь, университет Тампере, Тампере, Финляндия.

Тревога и депрессия являются многофакторными, крайне тяжелыми стрессорными расстройствами. Человек и животные имеют много общих патогенетических механизмов тревоги и депрессии. В настоящей статье рассмотрены экспериментальные подходы к моделированию тревоги и депрессии на животных и сделан подробный анализ существующих проблем и перспектив исследований в данной области биологической психиат-

рии, понимание которых поможет в поиске новых подходов терапии тревожно-депрессивных расстройств.

f.Ключевые слова: тревога, депрессия, биологическая психи-Щ атрия, экспериментальные модели.

Kalueff A.V., Tuohimaa P. Problems and perspectives of experimental modeling of anxiety and depression // Psychopharmacol. Biol. Nar-col. 2004. Vol. 4. № 2-3. P. 645-651. Medical School, Tampere University Hospital, University of Tampere, Tampere, Finland.

Anxiety and depression are mul-tifacetic debilitating stress-related behavioural disorders. Humans and animals share many pathogenic mech-

anisms of these disorders. Here we review the experimental modeling approaches to anxiety and depression in order to contribute to a better understanding of underlying biological aspects of these disorders. The present paper provides a comprehensive overview of the emerging problems and perspectives in the field of experimental biological psychiatry of anxiety and depression. Understanding the existing challenges, limitations and perspectives in the field of experimental modeling of stress-related disorders may assist in the search of novel biologically oriented approaches to the treatment of human anxiety and depression.

VKey words: anxiety; depression; biological psychiatry; experimental models.

INTRODUCTION

In recent years a large body of evidence appear to link stressful life events with an increased vulnerability for anxiety and depression disorders

[1]. Anxiety and depression have long been known to have dramatic impact on behaviour [2]. Numerous clinical data show that patients with anxiety and depression differ in their behavioural manifestations, although in many cases mild anxiety and mild depression are clinically very similar [3]. Despite human and animal depression or anxiety profiles differ in their cognitive-behavioural aspects, animals share the same «behavioural» problem since many findings show similar behavioural effects of animal anxiety and depression [4, 5]. Study of anxiety and depression is a rapidly developing field where all contemporary theories and paradigms are based on cross-disciplinary approaches and data coming from biology, medicine and psychology [5].

During the past decades, a remarkable link has been formed among psychiatry, physiology, molecular biology, and behavioural biology to demonstrate the striking parallel nature of animal and human anxiety/depression behaviours [1, 3]. Specific social behaviours in rodents are very sensitive to anxiety levels [6]. Homologous to human «social anxiety» state, it is a valuable behavioural model for testing putative anxiotropic manipulations [6]. Aggression is a common feature of many forms of anxiety and depression in humans [5], yet it is often seen as a measure of depressiveness in olfactory bulbectomised rodents [7]. While many forms of human anxiety are based on innate fears of natural dangers, exposure to a predator is known to be a useful model of «innate» animal anxiety [8]. Behavioral «despair» and anhedonia models of animal depression are based on similar behavioral symptoms often seen in human depression [2, 5]. There are numerous models of anxiety and depression based on prenatal and neonatal manipulations,

including acute and chronic exposure to various stressors or psychotropic drugs [8]. As expected, such experience result in long-term behavioural and neurobiological changes similar to those seen in humans having early life stress [8].

Exploratory behaviour (walking, approaches, scanning, sniffing, rearing, wall leaning, etc) is most often studied in the laboratory, in forced (non-escapable) or free (free-choice) situations, as stresssensitive behavioural parameter to assess anxiety and/ or depression [9]. Similarly, exploratory drive is known to be severely affected by anxiety and depression in humans [5, 8]. In addition, a number of vetetative behaviours, both in animals and humans, are meaningful parameters that can be sensitive to the level of stress [10, 11, 12]. As such, changes in «additional» behaviours such as self-grooming, defecation, urination, and yawnings [5, 12, 13] may be useful ethological parameters in anxiety/depression research. One can agree that some of these parameters are useful markers of human anxiety or depression. Risk assessment is another important ethological domain in anxiety and depression research [14]. In animals, these «anxiety-defensive postures» include immobility (freezing), stretched attention (orientation) and «flatback» approach to the stimulus [14, 15, 16]. In humans, there is a set of surprisingly similar alarming reactions which, both in animals and humans, are particularly sensitive to stress and can be bidirectionally affected by neurotropic drugs including anxiolytics and, at a lesser extend, antidepressants [5, 11].

However, in experiments it is sometimes difficult to distinguish certain different meaningful behaviours. For example, novelty is often used in stress research as both anxiety-provoking and attractive. Therefore, approach to one stimulus (attraction) will be difficult to distinguish from avoidance of another (aversion) or the lack of motivation to interact (anhedonia) in depression. That is why a detailed analysis and correct terming and interpretation of behaviours are crucial issues for biological psychiatry of anxiety and depression. It must, however, be emphasized that we can not ask animals how they feel in a model or why they react in a certain way. Since the question remains open as to whether we can say that a laboratory animal is «anxious» or «depressed», problems with data interpretations are well-documented in neurobehavioural studies [2, 5]. However, interpretation of the behaviour is critically important since it may be «more in the eye of the researcher than in the brain of the animal» [14] and, as we can add, «in the brain of the patient».

PITFALLS AND PROBLEMS OF ANIMAL MODELS

Considerable attention has been focused on the pitfalls and weaknesses of anxiety and depression

models. As we review these aspects in animals, one can see that many of them are equally important, and perhaps equally ignored, in the clinic. One of the major problems is that «different laboratories commonly employ their own idiosyncratic versions of behavioural tests apparatus and protocols, and any laboratory environment has many unique features» [17]. Can we say that psychiatric clinics too have «own idiosyncratic» views of treating their anxiety and depression patients? Lapin [18] listed other factors often neglected in experimental research of anxiety and depression: vehicle monitoring, handling control, false injection control, body temperature control, general activity control (false positives and negatives), injection effect, time of the day, light conditions, and pre-test manipulations. Ignoring them will almost certainly result in obtaining the data difficult to analyze. As more drugs become available and are used concominantly in different psychopharmacological assays, the potential for drug interaction increases and needs to be taken into consideration. Other significant test problems include: i) strains/species differences, ii) sex and age differences, iii) sensitivity to pretest manipulations (e. g. housing, transportation), iv) construction of apparatus, iv) conditions prior behavioural testing, iiv) scoring method (manual vs. automated), iiiv) acclimation before and after the testing. Variations between laboratories in any of these factors lead to conflicting results [17]. Previous housing regime (individual vs. grouped) is also an important biosocial factor able to affect anxiety and depression-related behaviours [19]. For example, isolated animals tend to be more active in the novel conditions and engage in more exploratory and escape-oriented behaviours [19, 20]. Since animals used in biomedical research are usually reared in small cages that lack key features of their natural habitat, it is now becoming evident that animals need to be reared in enriched conditions in order to be able to produce normal behaviours [20]. Rearing conditions may therefore, and in fact do so, affect behavioural reactions seen in animal models of anxiety and depression, also influencing memory, attention, agonistic behaviours and other important aspects of normal brain-behaviour interaction [17, 19, 20].

Timetable and intensity of the experiments are also key for obtaining correct and reliable experimental data. Since rodents are very sensitive to the rhythm of activity of the researchers and animal house staff (usually much higher from Mondays to Friday), after relatively «easy» weekend days animals will behave differently on Monday than on the rest of the week [21]; interestingly, can the same rule «work» in the clinic, say, to cause different therapeutical sensitivity on Friday and Monday? A significant amount of total variability of the experimental behavioural indexes could also be

attributed to 24h variation [22]. All major behavioural parameters in rodents are largely affected by light/dark phases. Since behavioral scores are much higher during the dark time of the day, this shows the importance of standardization of all experimental procedures [22] especially in their timing aspect. The major challenge here, however, lies in determining which factors need to be standardized [23]. For example, the experimenter identity has extremely important role in animal behaviours and strikingly overweights all other factors (in the order of importance: season, cage density, time of the day, sex, humidity and the order of testing) [23]. As such, we can suggest that the important «therapeutic» role of the doctor-patient relations in the clinic [18] may, in fact, be homologous to these pre-clinical observations.

Individual differences contribute significantly to the discrepancies and conflicting findings in the literature [24]. Further important problem arises when it became evident that anxiety and depression are heterogenous and each could be divided in various forms. Wehner et al. [25] note that heterogenity of anxiety reflects its polygenic regulation and many behavioural models are used to study its nature. Discussing possible evolutionary rationale, Marks and Nesse [26] write that for anxiety «subtypes exist because of the benefits of having responses specialized to deal with particular dangers or threats», yet they agree that subtypes of anxiety are not completely distinct as multiple dangers and thereat are common. Various subtypes of anxiety and depression include state (inductive) and trait (consitutive), «male-like» and «female-like», acute and chronic, innate and evoked, mild and severe subtypes [12, 15, 19, 27, 28]. It is now clear that all these forms are not only neuroethologically and physiologically different but also possess differential sensitivity to stressors. Importantly, a common behavioural factor does not emerge from numerous factor analytical studies and common animal models do not measure the same type of phenomena [29], and different strains have different sets of factors controlling their behaviour in the same model, and playing different roles within the model [30], see also Table 1. As estimated, up to 80% of behavioural differences between strains may be due to polygenetic factors.

Clearly, genetic factors may play a key role in producing inconsistent and conflicting results. The availability of an increasing number of transgenic and knockout animals as well as integration of emerging tools and technologies for genetic analysis will provide the groundwork for the genetic dissection of anxiety, depression and related disorders [17]. Numerous data report abnormal social, reproductive, parental, eating behaviours, aggression, learning and memory deficits, anxiety- and depression-like

^ Problems and perspectives

Table 1

Summary of factors controlling behaviours in three different rat lines [30]

Factors Low anxiety line Normal (Wistar) High anxiety line

Locomotor activity +++ ++++ +

Anxiety ++++ ++ +++

Exploration + ++ ++

Risk assessment ++ - ++++

Arousal + + +

alterations as well as abnormal sensitivity to all major classes of psychotropic drugs associated with genetic modifications [25, 31, 32]. However, not only several genes regulate a certain behaviour, but a single gene can control several different behaviours, including those which may be opposite. For example, the same genes regulate activity and defecation in the open field in the mice of certain strains [25]. Together, this makes experimental modeling even more complicated. Variability in the response to anxiolytic/antidepressant drugs may also be related to genetic factors underlying sensitivity to pharmacological treatments [33]. Despite the fact that most of the widely available animal models use state pathological condition, one has to show more logical interest in further study of more pathological (and clinically relevant) «trait» anxiety and depression in animals. For this, genetic approaches, such as knocking out key genes, may be very useful. However, the behavioural effects of a receptor’s gene genetic ablation are not always the same as those of the pharmacological ablation by antagonist administration [34], and many other contributing neurobiological mechanisms have to be taken into consideration before claiming the link between the receptor, the gene and the disorder. Indeed, although common genes regulate general aspects of emotionality [26], a unique set of genes may regulate certain specific emotionality measures, and it is now clear that different groups of genes regulate baseline (trait) and state animal anxiety behaviours [25]. Clearly, «pathological» anxiety is not always an excess of normal «anxiety», and in many cases may represent a new qualitative variation. Thus far, simple increase of normal anxiety in the model may not produce «pathological» anxiety, and behavioural changes observed in the models could be the result of not fully appreciated species-typical normal response. The same basic principle is most likely to be true for different depression subtypes.

Terminology and «labelling» issue is yet another cause for experimental problems and difficulties with data analysis and interpretation — in a way, similar

to diagnostics and nozology dillemas that exist in psychiatry. Often the same term is used for different behaviours as well as different terms are used for the same behaviours (see, for example, conflicting definitions of despair-like «immobility» in the forced swim test, discussed in [11]). Here we will illustrate this point by a new, even more meaningful example. Consider the simplest single behaviour — freezing (immobility), a motionless behaviour that can be registered in animals or humans who have just been exposed to an aversive stimulus. Is it a mere temporal «absence of behaviour», inhibition of meaningful ongoing behaviours or, in contrary, activation of a meaningful adaptive behaviour with complex allied stress-sensitive reflexes? Does it mean increased anxiety or, perhaps, depression? In our case, for «immobility» in rats one has to consider many possible behavioural explanations [35-43], including sleep or resting (no body movements and postural support, closed eyes, small movements of the whiskers); catalepsy (lost capacity to move and retained postural support) or learned «pseudocatalepsy» (a specific behavioural response to low doses of neuroleptics); certain forms of horizontal exploration (e. g. rearlike postures); fearful freezing (retained capacity to move and postural support, often seen in anxiety) with or without facial expression of fear, eye-blink reflex, sonic/ultrasonic vocalization; restrained-induced (tonic) immobility or «animal hypnosis» (phenomenon seen in many animal species); heat-loss (sprawling, immobility without tone) or heat-gain postures (immobility with muscle tone); stereotypic freezing (often seen in depression); decision-making behaviour (see above); risk-assessment behaviours (with or without head movements, active breathing, gnawing, licking, sniffing and/or vibrissae micromovements); «stopping» state (a pause which might normally occur between two behavioural bouts); a part of normal self-grooming ritual (interruption or pre-grooming «preparation»); false «immobility» during naturally occuring or stress-evoked vegetative behaviours, such as defecation or urination; hiding or escape behaviours; immobility as a part of agonistic rituals (usually in sub-ordinant animals); illness-induced (seen with bad general appearance) or drug-induced non-specific behavioural inhibition (sedation); «sit» (normal comfort-like behaviour) and predator-induced «tonic-like» immobility (a terminal defense response when there is a physical contact between pray and predator). Needless to say, how different these behaviours are and how hard it is to distinguish between them in the experimental models of anxiety and depression (one can also agree that this comprehensive list of ethological explanations is at least comparable to a variety of motivations or reasons which a psychiatrist would sometimes consider for his patient in the clinic). It appears that each of these behaviours has its own

causes and controlling neural networks, is differentially sensitive to anxiety and depression, and can not be interpreted as a simple non-specific unitary act. In addition, they will be differentially sensitive to psychotropic drugs and manipulations. Perhaps, the most important conclusion from this is that one parameter is not enough to judge and interpret experimental anxiety or depression. As stated by Sarter and Bruno [44], «animal models may not only generate data with very limited validity but also deprive the researcher from achieving the main goal of research using animal models, i.e. to test and extend a theory about the disorder of interest». With this in mind, do we always know that in our experiments we picked the right explanation and interpretation? And, even more importantly, do we actually use the correct model?

APPROACHES AND PERSPECTIVES

The above examples show that the key task in the experimental biological psychiatry of anxiety and depression is to use the right animals in the right model(s) and, following the right protocols, obtain right behavioural data for their correct interpretation within the right theory of the right pathology. In order to reach this goal, we have formulated several essential approaches to a proper experimental modeling of anxiety and depression (see also [39) for discussion). 1. Use a battery of tests vs. a single paradigm. 2. Make sure subjects have no impairments in motor activity and sensory abilities. 3. Study sex-related differences and use male and female animals (vs. one-sex trials). 4. Avoid models based on strong stressors, try to use more «naturalistic» models.

5. Always care about housing conditions to be standard and appropriate. 6. Use ethologically «rich» multi-parameter models vs. single-measure tests (however, more complex test is more difficult to standardize and control fully for variations in test performance). 7. Analyze all behavioural data as a whole integral complex, do not rely on a limited number of «traditional» measures. 8. Avoid test-test interactions or animal re-test, since Trial A may be not only quantitatively but qualitatively different from the Trial B. 9. Consider all factors which may cause conflicting data (transportation, handling, housing). 10. Assess different subtypes of anxiety and/or depression, use both acute and chronic stressors of different strength and origin. 11. Consider hyper/hypo locomotion (false positives and negatives) and baseline (ceiling and floor effects) problems.

When using several behavioural models in the multiple «battery» study, one has to remember the timecourse dillema which always exists in neuro-behavioural stress research. Namely, some important behavioural differences may be absent immediately

after the stress but appear later, after the testing. There is no general rule as for when such effects may appear and disappear. Some behavioural effects last several minutes, while others can last for weeks. Some effects appear within 30 minutes after exposure to stress, while others may be visible only 24 hours later. In these cases, the risk of false negative findings is very high if the wrong check-point time is selected. The risk of false positives can be high too if one uses a battery of tests, the most likely strategy of neuroethological research nowadays. Since the behavioural effects induced by earlier tests may overlap with the effects seen after the current stress exposure, all prior test history has to be carefully controlled. Also, the behavioural changes produced in the stressed animals by different stressors may interact with each other and produce long-lasting biological changes which may affect behaviours and make the results complicated and difficult to interpret. Step-by-step and factor-by-factor analysis will eliminate this unwanted overlapping. The refinement of animal models has shown that many traditional experimental paradigms are receptive to further useful modifications and improvements [39]. However, the invention of the new models is becoming necessary to move further, from what Borsini et al. [45] ironically called «models of benzodiazepine pharmacology» to the new generation of models. Several encouraging examples appeared to support this conclusion [46-55]. Here we will briefly outline these principally new modeling approaches which emerge in the field of biological psychiatry of anxiety and depression [5, 11, 46].

These approaches include i) genization, ii) comor-bidization, iii) hybridization and iv) «transitization» of the experimental models of anxiety and depression

[11]. Briefly, genization can be defined as the use of genetically modified animals with specific behavioural disorders and/or principally new behavioral properties, accompanied by a detailed genetic control of behavioural quantitative trait loci linking specific subtypes of behavioural disorders to gene mutations, see [17, 19] for discussion. For example, Gass et al. [53] have developed mutant mice with targeted mutations of glucocorticoid and mineralocorticoid in order to model anxiety and depression. Comorbi-dization approach can be defined as the search for novel animal models of specific comorbidity states, such as anxiety + depression, anxiety + other disorder and depression + other disorder. In practice, this approach is closely related to the genization approach and behavioral neurogenetics [50, 56]. Confirming this, Rezvani et al. [49] have recently shown that the Fawn-Hooded rats can be used as a genetic animal model of comorbid depression and alcoholism.

Hybridization approach can be briefly defined as the invention of qualitatively new models based on the combination of several traditional models, in

order to measure principally new types of disorders in the same animal. One of such models, the elevated T-maze, consists of three elevated arms, one enclosed and two open [47] and is a «hybrid» between the elevated plus maze anxiety test [22, 28] and the Y-maze traditionally used to assess memory [5]. Inhibitory avoidance (representing learned fear) is measured by the time taken to leave the enclosed arm in three consecutive trials. Unconditioned fear is evaluated in the same rat by the time to escape from the open arm. Various behavioral and pharmacological results [47] support the view that inhibitory avoidance in the elevated T-maze may be related to generalized anxiety disorder, while oneway escape may be associated with panic disorder. However, not only different subtypes of anxiety, but also memory per se can be simultaneously tested in this model, as an indirect anxiety marker, see a detailed discussion of the role of activated memory in anxiety in our review [46]. Recently, another hybrid test, the graded anxiety paradigm [48], was introduced as a novel model of murine unconditioned anxiety based on the principles of the elevated plus-maze and light-dark tests [31]. The latter standard tests, based on a dichotomy of avoidance behaviour (walled vs. open arms and dark vs. light compartments) were combined so one walled arm of the maze was made transparent, whereas the other walled arm was opaque-gray [48]. Thus, the graded anxiety test might be useful to dissect up to three distinct anxiety and fear/panic subtypes as well as screen for substances that possess differential anxiolytic profiles [48]. It is possible to expect that other new models can be designed on the base of hybridization approach to target other domains of anxiety and depression (consider, for example, the hybridization of anxiety behavioral models and stress-induced hyperthermia test [48]). The partition test, recently introduced by [54, 55], can also be considered a hybrid of the social interaction test [6] and chronic stress [2, 39]. This interesting anxiety test assesses animals’ behavioural responses to another individual placed in the neighboring sector of the cage, divided by a transparent partition with holes. In this model, the behavioural responses differed depending on the physiological and psychological state of an individual, its social experience, type of partner in the neighboring sector, and line. The test can be informative and productive in experiments designed to study the neurochemical and neurophysiological mechanisms of social and general anxiety, aggression, sexual behavior, and depression [54, 55].

Finally, transitization approach is based on the use of «transitory» models of anxiety and depression pathogenesis [5, 11] in order to assess the transition from one stress-related pathology (usually, anxiety) to another (depression). These models are generally based

on chronic use of relatively mild stressors, such as social or novelty stress. For example, rodents with peripheral ZnSO4-induced anosmia develop anxiety-like profile on the week 1 but produce depressive-like behaviours on the week 3 or 4, accompanied by a robust disregulation of the CNS genes expression [51, 52]. Since olfaction plays an important role in animal behaviour, this model is a good example of «transitory» models and can therefore be used to study transition from uncertainty-induced anxiety to anhedonia-like depression. Importantly, this model seems to be homologous to many human olfactory-related anxiety and depression disorders [51]. In line with our anosmia model, the partition test (see above) can also be used as a «transitory» model. While in humans social stress has long been known to provoke anxiety and depression [3, 8), chronic social stress in the partition test — long experience of defeat in daily social inter-male confrontations and permanent living in olfactory contact with aggressive males — has been shown to produce depression in submissive animals [54, 55]. After 20 social defeats these animals display weight loss, decreased locomotion in the open-field test and increased immobility in the Porsolt’s swim test, while treatment with imipramine prevent animal depression [55]. Based on clear homology between animal and human stress-related behaviours, this and other similar tests may be therefore used to study anxiety-depressive pathogenesis in its «online» dynamics, including progression from «early» anxiety to depression. Since anxiety and depression have been linked to each other in numerous studies both in humans and animals [1, 3, 8], and even suggested to have common pathogenic mechanisms [57], the need in «transitory» tests, as a novel class of animal anxiety and depression models, is becoming increasingly important.

In conclusion, we suggest that the extensive use of the existing traditional experimental models, their revalidation and sophistication, as well as the invention of the novel classes of models, such as genetic, comorbidity, hybrid or «transitory» tests of anxiety and depression, will bring further important insights in the field of biological psychiatry of anxiety and depression. Together, this will not only allow to search for novel anxiolytic and antidepressant drugs but also discover principally novel classes of psychotropic drugs differentially targeting the whole spectrum of animal and human stress-related behavioural disorders.

REFERENCES

1. Arborelius L, Owens M.J., Plotsky P.M., Nemeroff C.B. The role of corticotropin-releasing factor in depression and anxiety disorders // J. Endocrinol. 1999. Vol. 160. P. 1-12.

2. Willner P. Validity, reliability and utility of chronic mild stress model of depression: a 10-year review and

evaluation // Psychopharmacol. 1997. Vol. 134. P.319-329.

3. Sandford J.J., Argyropoulos S.V., Nutt D.J. The psychobiology of anxiolytic drugs. Part 1: Basic neurobiology // Pharmacol. Ther. 2000. Vol. 88. P. 197-212.

4. Paterson A., Whitting P.J., Gray JA. et al. Lack of consistent behavioural effects of Maudsley reactive and non-reactive rats in a number of animal tests of anxiety and activity // Psychopharmacol. 2001. Vol. 154. P.336-342.

5. Kalueff A.V. Problems of the study of stress-related behaviour. Kiev: KSF, 1999.

6. File S.E., Seth P. A review of 25 years of the social interaction test // Eur. J. Pharmacol. 2003. Vol. 463. P. 35-53.

7. Kelly J.P., Wrynn A.S., Leonard B.E. The olfactory bulbectomized rat as a model of depression: an update // Pharmacol. Ther. 1997. Vol. 74. P. 299-316.

8. Newport D.J., Stowe Z.N., Nemeroff C.B. Parental depression: animal models of an adverse life event // Am. J. Psychiatry 2002. Vol. 159. P. 1265-1283.

9. Cruzio W.E. Genetic dissection of mouse exploratory behaviour // Behav. Brain. Res. 2001. Vol. 125. P. 127-132.

10. Berridge K.C., Aldridge J.W. Super-stereotypy I: Enchancement of a complex movement sequence by systemic dopamine D1 agonists // Synapse. 2000. Vol. 37. P. 194-204.

11. Kalueff A.V. Today and tomorrow of anxiety research // Stress Behav. 2003. Vol. 8. P. 145-147.

12. File S.E. Factors controlling measures of anxiety and responses to novelty in the mouse // Behav. Brain. Res. 2001. Vol. 125. P. 151-157.

13. Eguibar J.R., Romero-Carbente J.C., Moyaho A. Behavioural differences between selectively bred rats: D1 versus D2 receptors in yawning and grooming // Pharmacol. Biochem. Behav. 2003. Vol. 74. P. 827-832.

14. Overall K.L. Natural animal models of human psychiatry conditions: assessment of mechanisms and validity // Prog. Neuropsychopharm. Biol. Psychiatry 2000. Vol. 24. P. 727-776.

15. Blanchard R.J., Nikulina J.N., Sakai R.R. et al. Behavioural and endocrine change following chronic predatory stress // Physiol. Behav. 1998. Vol. 63. P. 561-569.

16. Kemble E.D., Bolwahnn B.L. Immediate and long-term effects of novel odors on risk assessment in mice // Physiol. Behav. 1997. Vol. 61. P. 543-549.

17. Finn DA.., Rutledge-Gorman M.T., Crabbe J.C. Genetic animal models of anxiety // Neurogenet. 2003. Vol. 4. P. 109-135.

18. Lapin I.P. Models of anxiety in mice: experimental verification and critical methodology // Exp. Clin. Pharm. 2000. Vol. 63. P. 58-62.

19. Palanza P. Animal models of anxiety and depression: how are females different? // Neurosci. Biobehav. Rev. 2001. Vol. 5. P. 219-233.

20. Clement E.Y., Calatayd F., Belzumg C. Genetic basis of anxiety-like behaviour: a critical review // Brain Res. Bull. 2002. Vol. 57. P. 57-71.

21. Belzung C. Handbook of Molecular genetics for brain and behaviour research // Measuring rodent exploratory behaviour / W.E. Cruzio, R.T. Gerlai. New York, Elsevier, 1999.

22. Andrade M.M., Tome M.F., Santiago E.S. et al. Longitudinal study of daily variation of rats’ behaviour in the elevated plus maze // Physiol. Behav. 2003. Vol. 78. P. 125-133.

23. Chesler E.J., Wilson S.G., Lariviere W.R. et al. Influences of laboratory environment on behaviour // Nature Neurosci. 2002. Vol. 5. P. 1101-1102.

24. Ho Y.J., Eichendorf J., Schwarting R.K. Individual response profiles of male Wistar rats in animal models for anxiety and depression // Behav. Brain Res. 2002. Vol. 136. P. 1-12.

25. Wehner J.M., Radcliffe RA., Bowers B.J. Quantitative genetics and mouse behaviour // Ann. Revs Neurosci. 2001. Vol. 24. P. 845-867.

26. Marks I., Nesse R. The evolutionary analysis of anxiety disorders // Understanding Mental health. 2001. Depression. org. uk.

27. Calatayud F., Belzung C. Emotional reactivity in mice, a case of nongenetic heredity? // Physiol. Behav. 2001. Vol. 74. P. 355-362.

28. Belzung C, Griebel G. Measuring normal and pathological anxiety-like behaviour in mice: a review // Behav. Brain Res. 2001. Vol. 125. P. 141-149.

29. Aguilar R., Gil L., Flint J. et al. Learned fear, emotional reactivity and fear of heights: a factor analytic map from a large F2 intercross of Roman rat strains // Brain Res. Bull. 2002. Vol. 57. P. 17-26.

30. Ohl F., Toschi N., Wigger A. et al. Dimensions of emotionality in a rat model of innate anxiety. // Behav. Neurosci. 2001. Vol. 115. P. 429-436.

31. Crawley J.N. Behavioural phenotyping of transgenic and knockout mice: experimental design and evaluation of general health, sensory functions, motor abilities, and specific behavioural tests. // Brain Res. 1999. Vol. 835. P. 18-26.

32. Bilkei-Gorzo A., Racz I., Michel K., Zimmer A. Diminished anxiety- and depression-related behaviours in mice with selective deletion of the Tac1 gene // J. Neurosci. 2002. Vol. 22. P. 10046-10052.

33. Belzung C. The genetic basis of the pharmacological effects

of anxiolytics: a review based on rodent models // Behav. Pharmacol. 2001. Vol. 12. P. 451-460.

34. Rodgers R.J. Anxious genes, emerging themes // Behav. Pharmacol. 2001. Vol. 12. P. 471-476.

35. Whishaw I.Q., Haun F., Kolb B. Modern Techniques in Neuroscience // Analysis of behaviour in laboratory rodents / Windhorst U., Johansson H., Berlin, Springer-Verlag, 1999.

36. Choleris E., Thomas A.W., Kavaliers M., Prato F.S. A detailed ethological analysis of the mouse open field test: effects of diazepam, chlordiazepoxide and an extremely low frequently pulsed magnetic field // Neurosci. Biobehav. Revs 2001. Vol. 25. P. 235-260.

37. Fundarro A. Pinch-induced catalepsy in mice: a useful model to investigate antidepressant or anxiolytic drugs // Progr. Neuropsychopharmacol. Biol. Psychiatry 1998. Vol. 22. P. 147-158.

38. Bouwknecht JA., Hijzen T.H., van der Gugten J. et al. Stress-induced hyperthermia in mice: effects of

flesinoxan on heart rate and body temperature // Eur. J. Pharmacol. 2000. Vol. 400. P. 59-66.

39. Cryan J.F., Markou A., Lucki I. Assessing antidepressant activity in rodents: recent developments and future needs // Tr. Pharmacol. Sci. 2002. Vol. 23. P. 238-245.

40. Davis M., Whalen PJ. The amygdala: vigilance and emotion // Mol. Psychiatry 2001. Vol. 6. P. 13-34.

41. Drai D., Kafkali N., Behjami-ni Y., Elmer G., Golani I. Rats and mice share common ethologically relevant parameters of exploratory behaviour // Behav. Brain Res. 2001. Vol. 125. P. 133-140.

42. Ferre S., Guix T., Prat G. et al. Is experimental catalepsy properly measured? // Pharmacol. Biochem. Behav. 1990. Vol. 35. P. 753-757.

43. Olsen C.K., Hogg S., Lapiz M.D. Tonic immobility in guinea pigs: a behavioural response for detecting an anxiolytic-like effects? // Behav. Pharmacol. 2002. Vol. 13. P. 261-269.

44. Sarter M., Bruno J.P. Biological psychiatry // Animal models in biological psychiatry / D’haenen H., den Boer J.A., Willner P., New York, John Willey and Sons, 2002.

45. Borsini F., Podhorna J., Marazziti D. Do animal models of anxiety predict anxiolytic-like effects of antidepressants? // Psychopharmacol. 2002. Vol. 163. P. 121-131.

46. Kalueff A.V., Nutt D.J. Role of GABA in memory and anxiety // Depress. Anxiety 1996. Vol. 4. P. 100-110.

47. Graeff F.G., Netto C.F., Zangrossi H. The elevated T-maze as an experimental model of anxiety // Neurosci. Biobehav. Revs 1998. Vol. 23. P. 237-246.

48. Dere E., Topic B., De Souza M.A. The graded anxiety test: a novel test of murine unconditioned anxiety based on the principles of the elevated plus-maze and light-dark test // J. Neurosci. Meth. 2002. Vol. 122. P. 65-73.

49. Rezvani A.H., Parsian A., Overstreet D.H. The Fawn-Hooded (FH/Wjd) rat: a genetic animal model of comorbid depression and alcoholism // Psychiatr. Genet. 2002. Vol. 12. P. 1-16.

50. Clement E.Y., Calatayd F., Belzumg C. Genetic basis of anxiety-like behaviour: a critical review // Brain Res. Bull. 2002. Vol. 57. P. 57-71.

51. Makarchuk N.E., Kalueff A.V. Olfaction and behaviour. Kiev: KSF, 2000.

52. Kalueff A.V., Maisky VA., Pilyavskii A.I., Makarchuk N.E. Persistent c-fos expression and NADPH-d reactivity in the medulla and the lumbar spinal cord in rat with short-term peripheral anosmia // Neurosci. Lett. 2001. Vol. 301. P. 131-134.

53. Gass P., Reichardt H.M., Strekalova T. et al. Mice with

targeted mutations of glucocorticoid and

mineralocorticoid receptors: models for depression and anxiety? // Physiol. Behav. 2001. Vol. 73. P. 811-825.

54. Kudryavtseva N.N. Use of the «partition» test in behavioral and pharmacological experiments. // Neurosci. Behav. Physiol. 2003. Vol. 33. P. 461-471.

55. Kudryavtseva N.N., Bakshtanovskaya I.V., Koryakina LA. // Pharmacol. Biochem. Behav. 1991. Vol. 38. P. 315-320.

56. Flint J. Animal models of anxiety and their molecular dissection. // Sem. Cell Devel. Biol. 2003. Vol. 14. P.37-42.

57. Boyer P. Do anxiety and depression have a common pathophysiological mechanism? // Acta Psychiatr. Scand. Suppl. 2000. Vol. 406. P. 24-29.