br Plasticity sensitive periods and alternative developmenta

Plasticity, sensitive periods and alternative developmental trajectories (Question 4) Although there is currently little direct evidence on this issue for autism, examples where transient adaptations have life-long consequences for annexin v development are common. For example, disruption to the GABAergic or glutamergic systems will likely have different effects in early development than in the mature brain. GABA is known to switch from a largely excitatory function in prenatal development toward inhibition in early postnatal development, and this switching may drive the opening and closing of critical periods in sensory cortices (Hensch, 2005; Hensch et al., 1998). Imbalances in excitatory and inhibitory activity are also likely to be significantly moderated by homeostatic cell mechanisms that maintain network stability in the mature brain (Turrigiano and Nelson, 2004; Turrigiano, 2011) in ways that vary across development. These complexities in developmental timing indicate that subtle disturbances in inhibitory/excitatory balance will vary across different phases of development. Finally, evidence from mouse models of neurodevelopmental disorders supports the proposal that some synaptic phenotypes are transient in nature (Kroon et al., 2013). For example, in fmr1 knock-out mice altered plasticity in somatosensory cortex that is observed during the first postnatal week becomes normalized by the third week (Harlow et al., 2010); this may relate to peak expression of the fmr1 gene, which is up-regulated between postnatal days 4 and 14 (Hoerder-Suabedissen et al., 2013). However, these earlier transitory atypicalities subsequently may have knock-on effects on later development stages (“sleeper effects”), which live on as secondary consequences of the initial imbalance. Thus, consideration of the role of disruptions of neurotransmitter systems in autism must take a developmental perspective. Other aspects of human developmental biology show clear evidence for differential developmental pathways depending on early life experience. For example, a variety of related theories stem from “Barker’s hypothesis” (Barker and Osmond, 1986) in which he argued, based on an association between prenatal nutrition and late-onset coronary heart disease, that fetuses adapt to the environment that they expect to enter postnatally. According to this hypothesis the nutrient environment of the fetus usually also represents the expected postnatal environment. When the two match up all is well. However, when they do not, such as in a famine during fetal life followed by a postnatal period of plenty, the earlier adaptation can become harmful as the body is physiologically prepared for an environment that differs markedly from the one that it actually inhabits. Over ensuing years evidence for this view has accumulated, and the account updated to include peri and postnatal influences, and the notion of the “thrifty phenotype” in which fetal glucose conserving adaptation occurs in response to intrauterine hypoglycaemia, creating a significant mis-match when the postnatal nutritional environment is good, and thus increases the risk for several metabolic disorders, including type II diabetes (Hales and Barker, 2001). Similar hypotheses have been advanced for maternal stress during pregnancy and the resulting elevated glucocorticoid exposure of the fetus, resulting in long-term up regulation of the hypothalamo-pituatary-adrenal (HPA) axis after birth. Gluckman and Hanson (2004) classified these phenomena as “predictive adaptive responses”, that result in later disorders only when there is a mis-match between the predicted later environment and reality (for review see De Boo and Harding, 2006). In an intriguing recent study, Filiano et al. (2016) have shown an association between immune system function, brain function and social behavior in mice. Reducing an immune system molecule, interferon gamma, is associated with overactive neurons in prefrontal cortex and poor social behavior. Restoring interferon gamma, returns prefrontal neuron activity to normal (suggesting increased inhibitory neuron activity), and appropriate social behavior. This association between the immune and nervous systems is hypothesized by the authors to be adaptive as infections spread more rapidly when animals are in close contact. Linking this to predictive adaptive responses is evidence that women who develop infections during pregnancy run an increased risk of having a child later diagnosed with autism (Brown et al., 2014; although other results are inconsistent – Zerbo et al., 2016). The intriguing association between prenatal maternal immune response, changes in brain function, and later social behavior, raises the speculative hypothesis that the adaptive brain trajectory we have discussed for autism, could at least partly reflect a predictive adaptive response for an environment rich in disease or environmental toxins.