SfN 2011—Sex, Stress, Immunity, and Neurodevelopment
Elise Malavasi, a graduate student at the University of Edinburgh, was kind enough to volunteer as a meeting correspondent for this year's Society for Neuroscience meeting, 12-16 November 2011 in Washington, DC.
7 December 2011. This symposium, chaired by Geert J. De Vries, University of Massachusetts, Amherst, explored the effects of pre- and postnatal environmental stressors on the vulnerability to behavioral disorders. Many of these effects are known to be sexually dimorphic, which leads to the question of whether stress exposure can at least partially explain the sometimes striking gender differences in incidence of central nervous system disorders.
Alan Brown, from Columbia University, New York, added to the session with a life course perspective on schizophrenia: different environmental exposures are known to act at different stages of brain development to increase the risk of disease, but what proportion of this additional risk is due to immediate effects of such exposures, and what proportion is instead a consequence of long-term cascades of events triggered by the environment? Brown pointed out that many of the past studies that explored the contribution of the environment on the risk of schizophrenia were affected by methodological limitations, including misclassification of exposures and selection bias. However, there is good reason to be optimistic now that several large birth cohorts with thorough databases on pre- and perinatal exposures as well as available biological specimens have passed adolescence. One such cohort is the Child Health and Development Study (CHDS), which comprises over 19,000 individuals born in California between 1959 and 1967, with mental health records available for over 12,000 of them. Brown and his teamís past and current efforts have been focused on analyzing data from this large cohort to extract important information on the environmental determinants of mood disorders.
Brown has focused part of his studies on the effect of prenatal infections such as maternal toxoplasmosis, influenza, and genital infections on the risk of schizophrenia. Exposure to influenza during the first trimester of pregnancy was observed to increase the risk of schizophrenia sevenfold, whereas exposure during the second half of pregnancy had no influence on the risk of schizophrenia (Brown et al., 2004). Interestingly, the figures were reversed for bipolar disorder, with a seven- or sixfold increased risk of disease for exposure during the second and third trimester, respectively, and a much milder increase in risk for earlier exposures, suggesting a correlation between time of exposure and severity of the outcome. One of the proposed mechanisms by which prenatal infections could increase the risk of mental illness is by stimulating the production of pro-inflammatory cytokines. Brown observed that in the CHDS, maternal levels of IL-8 were increased in the second trimester in mothers of children who later developed schizophrenia (Brown et al., 2004). Brown remarked that IL-8 is known to mediate neutrophil attraction and discharge of lysosomal enzymes, producing free radicals that may damage the developing brain. In the same birth cohort, Brown and his team also observed a correlation between prenatal infection and neurobiological phenotypes in schizophrenics. In particular, patients who had been exposed to toxoplasma or influenza prenatally performed worse in the Wisconsin card sorting test, and had more severe verbal and working memory as well as motor coordination deficits than the non-exposed patients (Brown et al., 2009). In the light of relatively modest effect sizes of individual risk factors and susceptibility genes, Brown emphasized the need for more studies designed to dissect the interplay between genetic and environmental risk factors. Finally, he pointed out that the proportion of the incidence of schizophrenia that is attributable to prenatal infection as a whole is as high as 33 percent, highlighting the importance of better infection prevention to reduce the incidence of schizophrenia.
Marian JoŽls, from the University of Utrecht, The Netherlands, shifted the focus from epidemiological to animal studies. JoŽls started by explaining how natural variation in maternal care, represented by the motherís licking and grooming (LG) behavior, correlates with brain morphology and function in the offspring. In the hippocampal CA1 cells of the offspring of low LG mothers, the dendritic tree is less elaborated and the spine density is lower. Furthermore, maternal LG has profound effects on long-term potentiation that are modulated by exposure to the stress hormone corticosterone. In fact, while LTP is impaired in offspring of low LG mothers under basal conditions, it is significantly enhanced in the presence of corticosterone (Champagne et al., 2008). Enhanced plasticity in low LG offspring under stressful conditions was also detected in vivo. Low LG offspring had impaired spatial memory, but stronger fear-conditioned memory compared to high LG offspring (Champagne et al., 2008). Similar findings were obtained following maternal deprivation, a more severe form of early life stress (Oomen et al., 2010). Collectively, JoŽls' observations indicate that cognitive performance is optimal when early and late life environments match, suggesting that early life stress may program the brain so that it responds optimally to stressful conditions later in life. Given the large intra-litter variability in the amount of maternal care received by littermates, JoŽls decided to study the effect of maternal LG on individual animals rather than whole litters. In the elevated plus maze, the amount of LG received by the animals positively correlated with the amount of time spent in the open arms, denoting reduced anxiety in high LG offspring. Other parameters, like social play behavior in adolescence, were positively correlated with maternal care in males only. Maternal LG also influenced decision-making in the offspring, as well as the ability to induce long-term potentiation in the hippocampus.
The next speaker, Quentin Pittman, from the University of Calgary, Canada, talked about the relationship between neonatal sickness and development of autonomic and behavioral control. To investigate the effect of early life contact with pathogens on immunological and behavioral parameters in the adult, Pittman and coworkers injected two-week-old rats with LPS or saline and assessed the animals once they had reached adulthood (Galic et al., 2008). They noted no differences in maternal behavior towards LPS- or saline-treated pups, but injection of LPS in adult rats that had been exposed to LPS neonatally (LPS neo) elicited milder immune activation compared to unexposed littermates. Additionally, LPS neo resulted in hyperalgesia later in life. On the other hand, no major behavioral differences were noted between LPS neo animals and littermates, both under basal and stressful conditions. Pittman and colleagues hypothesized that the hyperalgesia observed in LPS neo rats could be caused by the release of pro-inflammatory cytokines. Pittman argued that, although cytokines do not cross the blood-brain barrier, they activate receptors on endothelial cells of brain vessels, which in turn produce new cytokines and release them in the brain. Endothelial-derived cytokines, together with prostaglandins generated by COX-2 activation, can affect synaptic activity. Consistent with this hypothesis, LPS neo rats were more susceptible to seizures induced by convulsants, and to excitotoxic damage and neuronal death in the CA1 and CA3 regions of the hippocampus, and they also displayed enhanced synaptic excitability.
To investigate whether these changes could be caused by inflammation in the brain, Pittman and colleagues measured plasma cytokine levels in neonate rats, and found a marked increase in pro-inflammatory cytokines such as IL-1β, TNFα, and IL-6 immediately after the injection. Interestingly, antagonizing the post-injection surge of TNFα, but not IL-1β, rescued the increased excitability in the adult brain, and administration of TNFα in neonate rats reproduced the effects of LPS neo in the adult. The observation that minocycline, which is known to inhibit microglial activation, prevented the TNFα-induced enhancement of neuronal excitability strongly implicated microglia in the mechanism of action of TNFα. Although there were no changes in basal cytokine levels or microglial activity in adult rats exposed to LPS perinatally, Pittman observed increased mRNA levels of the NMDAR2 subunit in adults exposed to neonatal LPS, which may contribute to the increased neuronal excitability in these animals. Interestingly, the mRNA levels of the ion channel NKCC1, but not NKCC2, were increased in adult LPS neo rats. Replacement of NKCC1 channels with NKCC2 channels during development is responsible for the functional switch of GABA from depolarizing in neonates to hyperpolarizing in adults. Consequently, the GABA reversal potential was shifted in LPS neo adult rats. The increased excitability in these rats could be reversed by a NKCC1 blocker, indicating that the neonatal transporter is still functional in adult LPS neo rats. Pittman concluded by suggesting that neonatal LPS treatment may somehow—possibly by affecting DNA methylation, ďclampĒ the neuronal phenotype, so that production of certain neonate-specific proteins is maintained into adulthood, resulting in functional anomalies.
Maria-Paz Viveros, from the Universidad Complutense in Madrid, Spain, presented her work on sex-dependent behavioral, physiological, and immunological effects of developmental stress and drug use. Viveros initially focused on the effects of drug use during adolescence, a developmental period during which several psychiatric conditions first manifest themselves, and that frequently coincides with the initial exposure to addictive drugs such as tobacco and cannabis. Viveros pointed out that some of the changes that occur in the brain during adolescence are shared between species, one example being the late maturation of the prefrontal cortex, which correlates with an increased inclination towards risk-taking behavior and drug abuse in adolescence. The partial conservation of the developmental changes in the adolescent brain means that the effects of drug abuse and stress during this time window can be modeled in rodents. Cannabis, one of the substances that are most commonly abused in adolescence, interferes with the endocannabinoid system, which plays important roles in emotional homeostasis, reward systems, and brain development. Elements of the endocannabinoid system are present in the brain from very early life, but the system only completes its development during adolescence, suggesting that cannabis consumption during this period may have critical and long-term effects. To test this, Viveros and her team exposed adolescent rats to a cannabinoid agonist and/or nicotine, and then subjected the animals to cognitive and behavioral testing once they had reached adulthood (Mateos et al., 2010). Interestingly, they found sex-dependent impairments in cognitive abilities and mood, as well as lasting changes in the nicotinic and cannabinoid receptors in the brain. They also observed clear gender-dependent differences in the ability of pre-exposure to nicotine to modify the effect of acute administration of a cannabinoid agonist on anxiety-related behaviors in adolescent rats (Marco et al., 2006). Furthermore, THC (the active principle of cannabis) and MDMA (ecstasy) also interacted in adolescent rats, with the most severe cognitive effects observed in males treated with both drugs. Viveros also observed differences between male and female rats in the acquisition, maintenance, and relapse of nicotine dependence, whereby females became addicted more rapidly and maintained the addiction longer, reflecting the fact that quitting smoking is harder for women than for men.
Viveros and colleagues also tested the effect of cannabis use in adolescence on the use of other, ďharderĒ drugs later in life, and found that pre-exposure to cannabis increased morphine self-administration in adult male rats only, which is possibly linked to the observation that adolescent cannabis altered the functionality of the μ-opioid receptor in males but not females (Biscaia et al., 2008).
Finally, Viveros presented some data on the effects of acute perinatal stress, modeled in rats by maternal deprivation (MD) for 24 hours at postnatal day 9. MD in rats is known to produce behavioral abnormalities resembling schizophrenia in the adults. When analyzing the effects of MD, Viveros and coworkers observed marked sex-dependent changes affecting different systems (Viveros et al., 2009). While the HPA axis was activated in both sexes, serum leptin levels were decreased in males and increased in females, which may be related to the behavioral abnormalities. MD also affected the endocannabinoid system in a sex-dimorphic way, with significantly increased levels of 2-arachidonylglycerol (2-AG), a ligand for CB1 and CB2 cannabinoid receptors, and increased brain CB1 immunoreactivity in male rats but not females. In addition, they reported sex-dependent changes in the development of neurons and glia in the hippocampus and cerebellum after MD. In the light of these findings, Viveros concluded by stressing the importance of addressing sex dimorphism in psychiatric disorders and addiction, and better understanding the mechanisms driving the observed gender-dependent changes.—Elise Malavasi.