Molecular basis of aggression

Transcription

1 Molecular basis of aggression 713 Randy J. Nelson and Silvana Chiavegatto Recent pharmacological and genetic studies have dramatically expanded the list of neurotransmitters, hormones, cytokines, enzymes, growth factors, and signaling molecules that influence aggression. In spite of this expansion, serotonin (5-HT) remains the primary molecular determinant of inter-male aggression, whereas other molecules appear to act indirectly through 5-HT signaling. We review evidence of interactions among these molecules and aggressive behavior. Slight modulations in 5-HT levels, turnover, and metabolism, or in receptor subtype activation, density, and binding affinity affect aggression. Activation of specific 5-HT receptors evokes distinct, but highly interacting, second messenger systems and multiple effectors. Understanding the interactions between 5-HT receptor subtypes should lead to novel insights into the molecular mechanisms of aggression. Randy J. Nelson Distinguished Professor of Social and Behavioral Sciences, Depts of Psychology and Neuroscience, The Ohio State University, Columbus, OH 43210, USA. Silvana Chiavegatto Laboratory of Neuroscience-LIM 27, Dept and Institute of Psychiatry, School of Medicine, University of Sao Paulo, São Paulo, Brazil. During the past decade, a remarkable link has been formed among genetics, molecular biology, and behavioral biology. Insights into the molecular mechanisms underlying complex murine behaviors, such as ingestion, mating, maternal care, and other types of social behaviors, are reported with astonishing frequency. Aggression is a complex social behavior with multiple causes 1,2. A primary goal of studies of the molecular mechanisms underlying murine aggression is to determine the neurobehavioral factors that underlie violence in humans. Because human violence is primarily a male proclivity, studies using the most appropriate murine model: testosterone-dependent offensive inter-male aggression which is typically measured in residentintruder or isolation-induced aggression tests will be emphasized here 3. Accordingly, in this review we will highlight various molecules that have been linked to aggression by pharmacological or the latest gene targeting techniques. We conclude that a perturbed central 5-HT system is a primary feature in the expression of aggressive behavior. Limitations of mouse models Is murine aggression a good model for human violence? This question is rarely asked, but is fundamentally important. Although mice and humans share >90% of their genes in common, aggression observed in mice is rarely directly comparable, either in form, muscular output, or social consequences, to the violence displayed by humans. For example, male mice rarely focus their aggression towards females, whereas among humans, women are common targets of male violence. Conversely, the underlying molecular mechanisms found in aggressive mice are also reported in humans displaying aggression (e.g. the 5-HT link). In this sense, the use of mouse models provides an attractive tool to discover new candidate molecules that might mediate human aggression. In virtually all vertebrate species, including humans, males are markedly more aggressive than females. The extent to which human sex differences in violence are the result of hormones (or other physiological differences) or because of differences in socialization remains unspecified. Moderate testosterone concentrations are necessary for expression of male aggression in mice and perhaps also in humans 4. Description of murine aggression Aggression is a primitive, yet highly conserved vertebrate behavior, and it is reasonable to expect that the molecular mechanisms underlying aggression are similar (and possibly ancient) among vertebrates. Species-specific features of aggression might be the result of adaptive co-opting of novel molecules as modulators that have been incorporated into ancestral, pre-existing, neural circuits. Many animal studies of aggression have been conducted in mice; however, most strains of laboratory mice (Mus musculus) have been bred to be quite docile. Consequently, mice must often be put into artificial situations that promote aggression, for example, they are housed in isolation or given electric shocks. Nonhuman aggression has been divided into various types for ease of classification, and the different types of aggression appear to have different physiological causes (reviewed in Refs 5,6) (Table 1) (Fig. 1). Brain regions associated with aggression A neural circuit composed of several regions of the prefrontal cortex, amygdala, hippocampus, medial preoptic area, hypothalamus, anterior cingulated cortex, insular cortex, ventral striatum, and other interconnected structures has been implicated in emotion regulation. Functional or structural abnormalities in one or more of these regions or in the interconnections among them can increase the susceptibility for impulsive aggression and violence 7. Although the brain systems mediating aggression appear to be fairly constant among mammals, many details of the regulatory pathways are speciesspecific. For example, Syrian hamsters exhibit c-fos immunoreactivity in the medial amygdala, bed nucleus of the stria terminalis (BST), ventrolateral hypothalamus and dorsolateral part of periaqueductal gray (PAG) after displaying offensive aggression towards an intruder 8. Because these areas are interconnected with the anterior hypothalamus (AH), the authors suggest an integrated network centered on the AH in the regulation of offensive aggression /01/$ see front matter 2001 Elsevier Science Ltd. All rights reserved. PII: S (00)

2 714 Review Table 1. Types of aggression Type of aggression Refs Anti-predator aggression 83 Defensive aggression (fear-induced) 84 Predatory aggression 83,84 Dominance aggression (inter-male aggression) 83,84 Maternal aggression 84 Sex-related aggression 83,84 Territorial aggression (resident-intruder) 83,84 Irritable aggression (shock-induced) 84 A neural circuit involving the medial hypothalamus and PAG has been identified that subserves defensive rage behavior in cats 9. The hippocampus, amygdala, BST, septal area, cingulate gyrus, and prefrontal cortex project to these structures directly or indirectly and thus can modulate the intensity of attack and rage 9. In rats, attack behavior can be elicited by electrical stimulation of the intermediate hypothalamic area and the ventrolateral pole of the ventromedial hypothalamic nucleus, collectively called the attack area (reviewed in Ref. 10). Afferent and efferent connections to the attack area, including the amygdala, prefrontal cortex, septum, mediodorsal thalamic nucleus, ventral tegmentum, and PAG are also involved in aggressive behavior (reviewed in Ref. 11). Importantly, neurons in these aggressionmediating areas are rich in both steroid hormone receptors, and 5-HT 1A and 5HT 1B receptor subtypes 3. The neural circuitry for other types of aggression remains unspecified. Steroid hormones Androgen receptors Several types of steroid hormones could influence aggression (see Table 2, online supplementary information), but only androgens and estrogens are reviewed here. It is well-documented that castration eliminates aggression. Androgens are important mediators of aggression in several ways: (1) during development, androgens guide the organization of the Fig. 1. Male mice lacking the gene for the neuronal isoform of nitric oxide synthase (nnos / ) are highly aggressive. This aggressiveness has been linked to disturbances in 5-HT metabolism and signaling 24. Photograph courtesy of Jay Van Rennselaer (all rights reserved). brain into a male-like pattern by inducing or preventing neural cell death; (2) later, post-pubertal testosterone (or estrogenic byproducts) stimulates neural circuits that are organized perinatally, presumably by making aggression-inducing stimuli more salient. Importantly, the neuroendocrine mediators of aggressive behavior are primarily androgens or estrogens in different rat and mouse strains 3. For example, testosterone mediates aggression in C57BL/6J mice, whereas estradiol mediates aggression in the CF-1 strain 3. These steroids seem to promote aggressiveness at the level of the lateral septum, medial preoptic area, amygdala, and dorsal raphé nucleus 3. Consistent with the well-documented effects of castration and androgen replacement therapy on male aggression, male mice exhibiting a spontaneous mutation that fails to produce the long form of the androgen receptor (AR) are not aggressive 5,12. In nonhuman animals, strong evidence for a causal link between testosterone and aggression exists, but a similar association in humans remains weak except among abusers of anabolic steroids 4 (but see Ref. 13). Estrogen receptors Male mice with targeted disruption of the gene encoding the α-isoform of the estrogen receptor (ERαKO) display reduced aggression in several testing situations 14,15. Conversely, the β-isoform null mice (ERβKO) exhibit normal or increased aggression depending on social experience 16,17. ERαKO females exhibit increased levels of aggression towards other female mice relative to wild-type females 14,15. Because estrogen is essential for the normal sexual differentiation of the CNS of male (and possibly female) mammals during development 18, studies of adult behavior in ERKO are complicated by the inability to dissociate genetic from ontogenetic causes of behavior. Serotonin (5-HT) Several classical neurotransmitters have been linked to aggression (Table 2, online supplementary information), but the 5-HT data are most convincing. The 5-HT system generally dampens aggression in animals and violent behavior in humans. Impulsivity and high aggressiveness are correlated with low cerebrospinal fluid concentrations of the 5-HT metabolite, 5-HIAA, in humans and nonhuman primates, and reduced 5-HT levels or turnover in the brain of laboratory animals (reviewed in Ref. 19). Pharmacological strategies of increasing 5-HT levels, such as the use of 5-HT precursors, 5-HT reuptake inhibitors, in addition to 5-HT 1A receptor agonists are able to reduce aggressive behavior in rodents Genetic evidence for a role of 5-HT in aggression comes from mice missing specific genes that either directly or indirectly affect 5-HT concentrations or metabolism (Figs 1 and 2). The 5-HT 1B receptor is

3 715 generally not influenced aggression 3. It would be instructive to evaluate aggressive behavior systematically in additional 5-HT-receptor subtype knockout mice, and in tissue-specific 5-HT knockout mice, to discriminate between pre- and postsynaptic effects on aggression. With the ever-increasing discovery of 5-HT receptor subtypes, much work remains to be done to clarify the specific role of the various 5-HT receptors and the interactions among 5-HT receptor subtypes that underlie aggression. Fig. 2. Serotonergic innervation in the forebrain of 18-month-old wild-type (a) or brain-derived neurotrophic factor (BDNF) +/ (b) male mice is seen in these darkfield photomicrographs of 5-HTimmunoreactive axons (sagittal sections). 5-HT axon density is normal in neocortex and hippocampus of younger BDNF +/ mice but 5-HT density is reduced in 18-month-old BDNF +/ mice. The BDNF +/ mice develop enhanced inter-male aggressiveness associated with 5-HT dysfunction. Magnification, ~ 100. Reproduced, with permission, from Ref. 23. expressed in a variety of brain regions, including the basal ganglia, PAG, hippocampus, lateral septum, and raphé nuclei, either presynaptically inhibiting 5-HT release or as a heteroreceptor modulating the release of other neurotransmitters. Male mice that lack functional expression of the 5-HT 1B receptor gene (5-HT 1B / ) are more aggressive than wild-type controls 25. Lactating female 5-HT 1B / mice also attack unfamiliar male mice more rapidly and violently 26. Notably, administration of the nonselective 5-HT 1B agonist eltoprazine (one of the so-called serenics ) significantly reduces aggressive behavior in both 5-HT 1B null mice and wild-type mice, presumably by affecting 5-HT 1A receptors 26. Although the 5-HT 1B receptor contributes to aggression, these results suggest that the 5-HT 1B receptor subtype is not the only 5-HT receptor modulating this behavior. In particular, 5-HT 1A receptor activation, which is also induced by eltoprazine, can influence aggressive behavior. Interestingly, mice lacking 5HT 1A receptors are less reactive, and possibly less aggressive, and show more anxiety-related behavior compared with wild-type mice 27, a finding consistent with the observation of increased postsynaptic 5-HT 1A receptor availability in limbic and cortical regions of highly aggressive mice 28. These data do not elucidate, however, the known anti-aggressive effect of 5-HT 1A agonists in rodents 20,21. Although both 5-HT 1A and 5-HT 1B receptors control 5-HT tone, these two receptors probably have different contributions in particular brain areas that modulate the postsynaptic 5-HT inhibitory effects on aggression. Regional brain metabolic alterations in both 5-HT 1A receptor null mice do not provide a clear correlation with their respective phenotypes 29. Documenting the contribution of other 5-HT receptor subtypes in aggression has been precluded by the unavailability of selective ligands, but drugs that target the 5HT 1C, 5-HT 2 or 5-HT 3 sites have Signaling molecules α-ca 2+ calmodulin-dependent kinase II (α-camkii) Evaluation of signaling proteins provides clues about the neural circuits involved in complex behaviors. Activation of specific 5-HT receptors evokes cascades of different signal transduction molecules via distinct, but highly interacting, second messenger systems and multiple effectors. α-camkii knockout mice display reduced aggression in resident-intruder paradigms 30. Heterozygotes, in which only one copy of the α-camkii gene is missing, show normal offensive aggression and elevated defensive aggression 30. α-camkii is a neural-specific signaling molecule found at pre- and postsynaptic regions. It can promote neurotransmitter release in synaptosome preparations 31, and in the CA1 region of the hippocampus 32. CaMK-mediated phosphorylation is involved in activation of tryptophan hydroxylase, the rate-limiting enzyme in 5-HT synthesis 33. Accordingly, 5-HT release is reduced in the dorsal raphé of both α-camkii mutant mice 30. It remains to be tested if the disturbed 5-HT system accounts for aggressive behavior alterations in the α-camkii mutant mice. Importantly, the lack of any signaling protein could have nonspecific effects that might affect aggression or associated behaviors. Thus, careful behavioral analyses and confirmation of the gene knockout data with additional approaches are necessary. Regulator of G-protein signaling-2 (RGS2) RGSs are GTPase-activating proteins (GAP) that attenuate signaling by G-proteins. The RGS-family member RGS2 selectively inhibits G q -mediated activation of phospholipasec in cell membranes 34. The highest RGS2 expression is in the cerebral cortex, and its expression is dynamically responsive to neuronal activity 35. The RGS2 / mice are less aggressive than their heterozygous littermates. These animals also display increased anxiety responses and impaired synaptic development in the hippocampus 36. The results suggest an important link between a molecular regulator of G protein-coupled receptormediated signaling and aggression and anxiety. Further work is needed to clarify which signaling systems are most affected by the absence of RGS2. Breakpoint cluster region (BCR) The product of the bcr gene has been implicated in the regulation of the Rho-family of small GTP-binding

4 716 Review proteins. BCR has GAP activity towards Rac1/2 and Cdc42Hs (Refs 37,38). BCR has been extensively studied in cancer research because of its localization in hematopoietic tissues and its involvement in leukemia; however, in spite of its localization in hippocampus, piriform cortex, and the olfactory nuclei 39, no role in the CNS was suggested until recently. Male BCR null mutant mice display increased fighting 40 that is abolished in BCR / mice expressing a human BCR transgene construct. These data point to a physiological role for BCR and therefore the Rho-family of small GTPases proteins in the expression of aggressive behavior. Interactions between 5-HT and other molecules in aggression Steroid hormones Androgens, either acting directly or via estrogenic metabolites, tend to facilitate aggression, whereas 5-HT tends to inhibit aggression. Exposure to androgens early in life affects the expression and distribution of 5-HT receptor subtypes 3, Both testosterone and estradiol increase mrna encoding 5-HT 2A receptor and binding site densities in the brains of male rats 41. Importantly, both androgens and estrogens modulate 5-HT 1A receptor agonist effects on murine aggression 42. Thus, sex steroid hormones and 5-HT interact on several levels to influence the likelihood of aggression. Vasopressin Arginine vasopressin (AVP) is another hormone that plays a crucial role in aggression and other social behaviors 45,46 (Table 2, supplementary online information). The effects of AVP on aggression, centered in the AH, appear to be mediated by 5-HT. 5-HT 1A, 5-HT 1B,and AVP V 1A receptors have all been identified in the AH using autoradioagraphy 47. Microinjections of AVP into the AH in combination with 5-HT 1A or 5-HT 1B receptor agonists revealed that only the 5-HT 1A receptor activation inhibited AVP-facilitated aggression HT neurons project into the AH, and 5-HT appears to inhibit AVP-facilitated offensive aggression by activating 5-HT 1A receptors 47. Histamine Other neurotransmitters also influence aggression via the 5-HT system. Both pharmacological and genetic evidence indicate a facilitatory role for central histamine (HA) via H 1 -receptors in aggression 48,49. Consistent with these findings, intracerebroventricular HA administration decreases 5-HT levels in the hypothalamus of rats 50 ; H 1 -receptor null mice exhibit less aggression and increased 5-HT turnover in several brain areas 49. Substance P Although most pharmacological studies point to an inhibitory action of substance P (SP) in aggression 51,52, the genetic approach in which the SP-preferred receptor neurokinin-1 (NK1) is knocked out, indicates the contrary. The NK1 / mice are less aggressive 53 and exhibit an increase in 5-HT function accompanied by a selective desensitization of 5-HT 1A inhibitory autoreceptors 54. Nitric oxide Nitric oxide (NO) also serves as an aggressionmodulating neurotransmitter 55. Male neuronal NO synthase (nnos) null mice and wild-type mice in which nnos is pharmacologically suppressed are highly aggressive 56,57 (Fig. 1). Castration and testosterone replacement studies in both nnos / and wild-type mice exclude an activational role for gonadal steroids in the elevated aggression 56,58. NO also appears to affect aggressive behavior via 5-HT. Excessive aggressiveness and impulsiveness of nnos knockout mice are caused by selective decrements in 5-HT turnover and deficient 5-HT 1A receptor function in brain regions regulating emotion 24. Although precisely how NO interacts with the 5-HT system in vivo remains unspecified, these results indicate an important role for NO in normal brain 5-HT function and might have significant implications for the treatment of psychiatric disorders characterized by aggressiveness and impulsivity. It was possible that NO from endothelial tissue could also contribute to aggression; enos / mice, however, display virtually no aggression even after pharmacological normalization of blood pressure 59. Interestingly, these mutant mice exhibit an accelerated 5-HT turnover in the frontal cortex 60 strengthening the link between NO, 5-HT, and aggression. Therefore, the absence of NO from either neuronal or endothelial sources modulates the 5-HT system in opposite ways, leading to contrary effects in aggression. Monoamine oxidase A Metabolic enzymes such as monoamine oxidase A (MAOA) also contribute to aggression because they function to alter neurotransmitter levels. MAOA is predominantly found in catecholaminergic neurons in the brain, but MAOA catalyzes with high affinity the oxidative deamination of 5-HT, noradrenaline (NA), and dopamine (DA) 61. Inhibition of MAOA correlates with reduced aggression in isolated male mice 62 and foot shock-induced aggression 63, probably as a result of increased 5-HT levels. However, humans treated with pharmacological inhibitors of the MAO enzymes for depression display no change in impulsivity or aggression. Interestingly, MAOA deficiency, caused by a point mutation in its coding gene, correlates with impulsive aggression in several males from a Dutch family 64. Ablation of the gene encoding MAOA in mice leads to high levels of offensive aggression, in spite of elevated 5-HT concentrations 65. However, the metabolic disturbance caused by chronic MAOA deficiency induces several alterations in these mutant mice 66,67, including upregulation of adenosine A 2A receptors 68, and abnormalities of 5-HT receptor

5 717 Acknowledgements R.J.N. was supported by NIH grants MH and MH S.C. thanks Barasch Sylmar Ind. Metal. LTDA (Brazil) for financial support. subtypes 69. Thus, it is probable that the role attributed to genetic MAOA deficiency in aggressive behavior is actually a consequence of secondary effects in several systems. Neural cell adhesion molecule Neural cell adhesion molecule (NCAM) is important during development and in adult neural plasticity 70,71. Both NCAM / and NCAM +/ mice display elevated anxiety and aggression 72,73. Lower doses of 5-HT 1A agonists are necessary to reduce anxiety (and presumably aggressiveness) in the NCAM / or NCAM +/ mice compared with wild-type mice, suggesting a functional change in the 5-HT 1A receptor 73. Surprisingly, 5-HT 1A binding and 5-HT and 5-HIAA tissue content were unaltered among the NCAM genotypes 73. Re-expression of NCAM180, an isoform expressed relatively late in development 74, as a transgene in the brain of male NCAM / mice rescues normal anxiety and aggressive behaviors, as well as 5-HT 1A receptor function 75. Taken together, these results suggest an involvement of endogenous NCAM and the 5-HT system through 5-HT 1A receptors, but the specific molecular mechanism or its role in aggression remains to be determined. Interleukins Because 5-HT is a widely distributed neurotransmitter system, and because at least 14 different 5-HT receptor subtypes exist, the potential for subtle interactions among receptor subtypes and their signal transduction pathways is high. For example, individuals that are ill often display a constellation of behavioral changes collectively known as sickness behaviors that are mediated by interleukins (IL) and other cytokines 76. Sick animals are generally less social and less aggressive compared with uninfected individuals. Aggressive behavior is reduced by IL-1β administration in a dose-dependent manner 77. IL-1β appears to increase 5-HT and NA neuronal functioning in some brain areas related to emotional behaviors 78. IL-6 knockout mice show increased aggression 79. Transgenic male mice that overexpress the gene encoding human transforming growth factor α (TGFα) exhibit enhanced aggressive behavior 80 accompanied by increased plasma 17-β-estradiol concentrations and reduced 5-HT turnover in the brain. Interestingly, the heightened aggressiveness in these mice is reversed with either 5-HT uptake inhibitors 81 or by castration 82. Conclusions Although many other molecules can affect aggressive behavior (Table 2, online supplementary information), most appear to influence aggression by affecting the signaling properties of 5-HT. Androgens, or their metabolic byproducts, interact with 5-HT receptors to facilitate aggression. Early exposure to androgens influences expression and binding affinity of specific 5-HT receptor subtypes; postpubertal androgens also modulate 5-HT and its receptors. Slight modulations in 5-HT levels, turnover, and metabolism, or in receptor subtype activation, density, and binding affinity each affect aggression in different ways. Importantly, manipulations of signaling proteins can also dramatically affect aggression. Activation of specific 5-HT receptors evokes cascades of different signal transduction molecules via distinct, but highly interacting, second messenger systems, and via multiple effectors. The integrity of this complex pathway seems necessary for normal expression and termination of aggressive behavior. Understanding the interactions of 5-HT receptor subtypes should lead to novel insights into the molecular mechanisms underlying aggression. References 1 Grisolia, J.S. et al., eds (1997) Violence: From Biology to Society. Elsevier, Amsterdam 2 Filley, C.M. et al. (2001) Toward an understanding of violence: neurobehavioral aspects of unwarranted physical aggression: Aspen Neurobehavioral Conference consensus statement. Neuropsychiatry Neuropsychol. Behav. Neurol. 14, Simon, N.G. et al. (1998) Testosterone and its metabolites modulate 5HT 1A and 5HT 1B agonist effects on intermale aggression. Neurosci. Biobehav. Rev. 23, Zitzmann, M. and Nieschlag, E. (2001) Testosterone levels in healthy men and the relation to behavioural and physical characteristics: facts and constructs. Eur. J. Endocrinol. 144, Maxson, S.C. (2000) Genetic influences on aggressive behavior. In Genetic Influences on Neural and Behavioral Functions (Pfaff, D.W. et al., eds), pp , CRC Press 6 Nelson, R.J. and Chiavegatto, S. (2000) Aggression in knockout mice. Inst. Lab. Anim. Res. J. 41, Davidson, R.J. (2000) Dysfunction in the neural circuitry of emotion regulation a possible prelude to violence. Science 289, Delville, Y. et al. (2000) Neural connections of the anterior hypothalamus and agonistic behavior in golden hamsters. Brain Behav. Evol. 55, Gregg, T.R. and Siegel, A. (2001) Brain structures and neurotransmitters regulating aggression in cats: implications for human aggression. Prog. Neuro-Psychopharmacol. Biol. Psychiatry 25, Kruk, M.R. (1991) Ethology and pharmacology of hypothalamic aggression in the rat. Neurosci. Biobehav. Rev. 15, Siegel, A. et al. (1999) Neuropharmacology of brain-stimulation-evoked aggression. Neurosci. Biobehav. Rev. 23, Olsen, K. (1983) Genetic determinants of sexual differentiation. In Hormones and Behaviour of Higher Vertebrates. (Balthazart, J. et al. eds), p. 138, Springer-Verlag 13 Pope, H.G. Jr et al. (2000) Effects of supraphysiologic doses of testosterone on mood and aggression in normal men: a randomized controlled trial. Arch. Gen. Psychiatry 57, Ogawa, S. et al. (1997) Behavioral effects of estrogen receptor gene disruption in male mice. Proc. Natl. Acad. Sci. U. S. A. 94, Ogawa, S. et al. (1998) Roles of estrogen receptorα gene expression in reproduction-related behaviors in female mice. Endocrinology 139, Ogawa, S. et al. (1999) Survival of reproductive behaviors in estrogen receptor β gene-deficient (βerko) male and female mice. Proc. Natl. Acad. Sci. U. S. A. 96, Ogawa, S. et al. (2000) Abolition of male sexual behaviors in mice lacking estrogen receptors α and β (αβerko). Proc. Natl. Acad. Sci. U. S. A. 97, Arnold, A.P. (1996) Genetically triggered sexual differentiation of brain and behavior. Horm. Behav. 30, Lesch, K.P. and Merschdorf, U. (2000) Impulsivity, aggression, and serotonin: a molecular psychobiological perspective. Behav. Sci. Law 18, Olivier, B. et al. (1995) Serotonin receptors and animal models of aggressive behavior. Pharmacopsychiatry, 28, Miczek, K.A. et al. (1998) Alcohol-heightened aggression in mice: attenuation by 5-HT1A receptor agonists. Psychopharmacology (Berl.) 139,

6 718 Review 22 Fish, E.W. et al. (1999) Aggression heightened by alcohol or social instigation in mice: reduction by the 5-HT(1B) receptor agonist CP-94,253. Psychopharmacology (Berl) 146, Lyons, W.E. et al. (1999) Brain-derived neurotrophic factor-deficient mice develop aggressiveness and hyperphagia in conjunction with brain serotonergic abnormalities. Proc. Natl. Acad. Sci. U. S. A. 96, Chiavegatto, S. et al. (2001) Brain serotonin dysfunction accounts for aggression in male mice lacking neuronal nitric oxide synthase. Proc. Natl. Acad. Sci. U. S. A. 98, Saudou, F. et al. (1994) Enhanced aggressive behavior in mice lacking 5-HT1B receptor. Science 265, Ramboz, S. et al. (1996) 5-HT1B receptor knock out behavioral consequences. Behav. Brain Res. 73, Zhuang, X. et al. (1999) Altered emotional states in knockout mice lacking 5-HT1A or 5-HT1B receptors. Neuropsychopharmacology, 21(2 Suppl), 52S 60S 28 Korte, S.M. et al. (1996) Enhanced 5-HT1A receptor expression in forebrain regions of aggressive house mice. Brain Res. 736, Ase, A.R. et al. (2000) Altered serotonin and dopamine metabolism in the CNS of serotonin 5-HT(1A) or 5-HT(1B) receptor knockout mice. J. Neurochem. 75, Chen, C. et al. (1994) Abnormal fear response and aggressive behavior in mutant mice deficient for α-calcium-calmodulin-kinase II. Science, 266, Nichols, R.A. et al. (1990) Calcium/calmodulindependent protein kinase II increases glutamate and noradrenaline release from synaptosomes. Nature 343, Chapman, P.F. et al. (1995) The alpha- Ca 2+ /calmodulin kinase II: a bidirectional modulator of presynaptic plasticity. Neuron 14, Ehret, M. et al. (1989) Formal demonstration of the phosphorylation of rat brain tryptophan hydroxylase by Ca 2+ /calmodulin-dependent protein kinase. J. Neurochem. 52, Heximer, S.P. et al. (1997) RGS2/G0S8 is a selective inhibitor of Gqα function. Proc. Natl. Acad. Sci. U. S. A. 94, Ingi, T. et al. (1998) Dynamic regulation of RGS2 suggests a novel mechanism in G-protein signaling and neuronal plasticity. J. Neurosci.18, Oliveira-dos-Santos, A.J. et al. (2000) Regulation of T cell activation, anxiety, and male aggression by RGS2. Proc. Natl. Acad. Sci. U. S. A. 297, Diekmann, D. et al. (1991) Bcr encodes a GTPaseactivating protein for p21rac. Nature 351, Hart, M.J. et al. (1992) A GDP dissociation inhibitor that serves as a GTPase inhibitor for the Ras-like protein CDC42Hs. Science 258, Fioretos, T. et al. (1995) Regional localization and developmental expression of the BCR gene in rodent brain. Cell Mol. Biol. Res. 41, Voncken, J.W. et al. (1998) Abnormal stress response and increased fighting behavior in mice lacking the bcr gene product. Int. J. Mol. Med. 2, Sumner, B.E. and Fink, G. (1998) Testosterone as well as estrogen increases serotonin 2A receptor mrna and binding site densities in the male rat brain. Brain Res. Mol. Brain Res. 59, Clifford, A. et al. (1999) Androgens and estrogens modulate 5-HT 1A agonist effects on aggression. Physiol. Behav. 65, Ferrari, P.F. et al. (1999) The influence of gender and age on neonatal rat hypothalamic 5-HT 1A and 5-HT 2A receptors. Cell Mol. Neurobiol. 19, Zhang, L. et al. (1999) Sex differences in expression of serotonin receptors (subtypes 1A and 2A) in rat brain: A possible role of testosterone. Neuroscience, 94, Albers, H.E. and Bamshad, M. (1998) Role of vasopressin and oxytocin in the control of social behavior in Syrian hamsters (Mesocricetus auratus). Prog. Brain Res. 119, Ferris, C.F. (2000) Adolescent stress and neural plasticity in hamsters: a vasopressin serotonin model of inappropriate aggressive behaviour. Exp. Physiol. 85, 85S 90S 47 Ferris, C.F. et al. (1999) Serotonin regulation of aggressive behavior in male golden hamsters (Mesocricetus auratus). Behav. Neurosci. 113, Noguchi, S. et al. (1992) The suppression of olfactory bulbectomy-induced muricide by antidepressants and antihistamines via histamine H1 receptor blocking. Physiol. Behav. 51, Yanai, K. et al. (1998) Behavioural characterization and amounts of brain monoamines and their metabolites in mice lacking histamine H1 receptors. Neuroscience 87, Chiavegatto, S. et al. (1998) Histamine and spontaneous motor activity: biphasic changes, receptors involved and participation of the striatal dopamine system. Life Sci. 62, Hall, M.E. and Stewart, J.M. (1984) Modulation of isolation-induced fighting by N- and C-terminal analogs of substance P: evidence for multiple recognition sites. Peptides 5, File, S.E. (2000) NKP609, an NK1 receptor antagonist, has an anxiolytic action in the social interaction test in rats. Psychopharmacology 152, De Felipe, C. et al. (1998) Altered nociception, analgesia and aggression in mice lacking the receptor for substance P. Nature 392, Santarelli, L. et al. (2001) Genetic and pharmacological disruption of neurokinin 1 receptor function decreases anxiety-related behaviors and increases serotonergic function. Proc. Natl. Acad. Sci. U. S. A. 98, Nelson, R.J. et al. (1997) Role of nitric oxide in neuroendocrine regulation of physiology and behavior. Front. Neuroendocrinol. 18, Nelson, R.J. et al. (1995) Behavioural abnormalities in male mice lacking neuronal nitric oxide synthase. Nature 378, Demas, G.E. et al. (1997) Inhibition of neuronal nitric oxide synthase increases aggressive behavior in mice. Mol. Med. 3, Kriegsfeld, L.J. et al. (1997) Aggressive behavior in male mice lacking the gene for nnos is testosterone-dependent. Brain Res. 769, Demas, G.E. et al. (1999) Elimination of aggressive behavior in male mice lacking endothelial nitric oxide synthase. J. Neurosci. 19, RC30 60 Frisch, C. et al. (2000) Superior water maze performance and increase in fear-related behavior in the endothelial nitric oxide synthase-deficient mouse together with monoamine changes in cerebellum and ventral striatum. J. Neurosci. 20, Shih, J.C. et al. (1999) Monoamine oxidase: From genes to behavior. Annu. Rev. Neurosci. 22, Florvall, L. et al. (1978) Selective monoamine oxidase inhibitors. 1. Compounds related to 4-aminophenethylamine. J. Med.Chem. 21, Datla, K.P. and Bhattacharya, S.K. (1990) Effect of selective monoamine oxidase A and B inhibitors on footshock induced aggression in paired rats. Indian J. Exp. Biol. 28, Brunner, H.G. et al. (1993) Abnormal behavior associated with a point mutation in the structural gene for monoamine oxidase A. Science 262, Cases, O. et al. (1995) Aggressive behavior and altered amounts of brain serotonin and norepinephrine in mice lacking MAOA. Science 268, Cases, O. et al. (1996) Lack of barrels in the somatosensory cortex of monoamine oxidase A-deficient mice: role of a serotonin excess during the critical period. Neuron 16, Holschneider, D.P. et al. (2000) Regional cerebral cortical activation in monoamine oxidase A-deficient mice: differential effects of chronic versus acute elevations in serotonin and norepinephrine. Neuroscience 101, Mossner, R. et al. (2000) Differential regulation of adenosine A(1) and A(2A) receptors in serotonin transporter and monoamine oxidase A-deficient mice. Eur. Neuropsychopharmacol. 10, Bou-Flores, C. and Hilaire, G. (2000) 5-Hydroxytryptamine(2A) and 5-hydroxytryptamine(1B) receptors are differently affected by the monoamine oxidase A-deficiency in the Tg8 transgenic mouse. Neurosci. Lett. 296, Goridis, C. and Brunet, J.F. (1992) NCAM: structural diversity, function, and regulation of expression. Semin. Cell Biol. 3, Scholey, A.B. et al. (1993) A role for the neural cell adhesion molecule in a late, consolidating phase of glycoprotein synthesis six hours following passive avoidance training of the young chick. Neuroscience 55, Stork, O. et al. (1997) Increased intermale aggression and neuroendocrine response in mice deficient for the neural cell adhesion molecule (NCAM). Eur. J. Neurosci. 9, Stork, O. et al. (1999) Anxiety and increased 5-HT1A receptor response in NCAM null mutant mice. J. Neurobiol. 40, Persohn, E. and Schachner, M. (1987) Immunoelectron microscopic localization of the neural cell adhesion molecules L1 and N-CAM during postnatal development of the mouse cerebellum. J. Cell Biol. 105, Stork, O. et al. (2000) Recovery of emotional behaviour in neural cell adhesion molecule (NCAM) null mutant mice through transgenic expression of NCAM180. Eur. J. Neurosci. 12, Hart, B.L. (1988) Biological basis of the behavior of sick animals. Neurosci. Biobehav. Rev. 12,

7 Cirulli, F. et al. (1998) Behavioral effects of peripheral interleukin-1 administration in adult CD-1 mice: specific inhibition of the offensive components of intermale agonistic behavior. Brain Res. 791, Brebner, K. et al. (2000) Synergistic effects of interleukin-1beta, interleukin-6, and tumor necrosis factor-alpha: central monoamine, corticosterone, and behavioral variations. Neuropsychopharmacology 22, Alleva, E. et al. (1998) Behavioural characterization of interleukin-6 overexpressing or deficient mice during agonistic encounters. Eur. J. Neurosci. 10, Hilakivi-Clarke, L.A. et al. (1992) Alterations in behavior, steroid hormones and natural killer cell activity in male transgenic TGF alpha mice. Brain Res. 588, Hilakivi-Clarke, L.A. and Goldberg, R. (1993) Effects of tryptophan and serotonin uptake inhibitors on behavior in male transgenic transforming growth factor alpha mice. Eur. J. Pharmacol. 237, Hilakivi-Clarke, L. (1994) Overexpression of transforming growth factor alpha in transgenic mice alters nonreproductive, sex-related behavioral differences: interaction with gonadal hormones. Behav. Neurosci. 108, Wilson, E.O. (1975) Sociobiology. Harvard University Press, Cambridge 84 Moyer, K.E. (1971) The Physiology of Hostility, Markham, Chicago Molecular physiology and pathophysiology of tight junctions in the blood brain barrier Jason D. Huber, Richard D. Egleton and Thomas P. Davis Disruption of the tight junctions (TJs) of the blood brain barrier (BBB) is a hallmark of many CNS pathologies, including stroke, HIV encephalitis, Alzheimer s disease, multiple sclerosis and bacterial meningitis. Furthermore, systemic-derived inflammation has recently been shown to cause BBB tight junctional disruption and increased paracellular permeability. The BBB is capable of rapid modulation in response to physiological stimuli at the cytoskeletal level, which enables it to protect the brain parenchyma and maintain a homeostatic environment. By allowing the loosening of TJs and an increase in paracellular permeability, the BBB is able to bend without breaking ; thereby, maintaining structural integrity. Jason D. Huber Richard D. Egleton Thomas P. Davis* Department of Pharmacology, The University of Arizona College of Medicine, Tucson, AZ 85724, USA. * davistp@ u.arizona.edu The blood brain barrier (BBB), once thought to be a static, rigid wall between the CNS and the periphery, is actually a dynamic, complex structure that is capable of rapid modulation, even under adverse conditions. Tight junctions (TJs) and the lack of fenestrations enable the BBB to regulate brain parenchyma composition. Growing evidence indicates that TJs are important in BBB regulation during many pathological insults. This article investigates the composition of TJs in the BBB and the signaling mechanisms involved in both the physiological and pathophysiological regulation of BBB function. Blood brain barrier The BBB is a physical and metabolic barrier between the CNS and the systemic circulation, which serves to regulate and protect the microenvironment of the brain. The BBB is characterized by the presence of TJs, which result in high transendothelial electrical resistance ( Ω*cm 2 ) and decreased paracellular permeability 1. The cerebral microvasculature is ensheathed by astrocytic end feet, which play an essential role in maintaining BBB phenotype 2. Astrocytes confer a protective role on the BBB against hypoxia and aglycemia 3. Interestingly, astrocytes isolated from stroke-prone spontaneously hypertensive rats do not have the ability to induce BBB properties as well as do normal astrocytes 4. Consequently, upon ischemic insult, astrocyte swelling and BBB lesions in the stroke-prone rats compromise the BBB at a much greater rate than seen in control rats 4. Failure to maintain BBB integrity can have profound effects on the CNS (Box 1). Changes in BBB function have been described in several neurological disorders, including stroke, multiple sclerosis and Alzheimer s disease. Recent evidence shows that inflammation, induced peripherally by hindpaw injection, leads to altered expression of TJ proteins in the BBB and increased paracellular permeability 5. Inflammatory diseases such as chronic relapsing multiple sclerosis show persistent BBB dysfunction 6, which, in experimental models, precedes any neurological deficits 7. In Alzheimer s disease, β-amyloid induces cytokine release and monocyte migration across an in vitro BBB (Refs 8,9) and in vivo β-amyloid deposition leads to degeneration of the microvascular basement membrane and alterations in BBB permeability 10. Many of these changes have been linked to alterations in BBB TJs. Tight junctions TJs of the BBB create a rate-limiting barrier to paracellular diffusion of solutes between endothelial cells. They are the most apical element of the junctional complex, which includes both tight and adherens junctions. Structurally, TJs form a continuous network of parallel, interconnected, intramembrane strands of protein arranged as a /01/$ see front matter 2001 Elsevier Science Ltd. All rights reserved. PII: S (00)02004-X