As part of our ongoing coverage of the 2009 International Congress on Schizophrenia Research (ICOSR), 28 March to 1 April 2009, in San Diego, California, we bring you session summaries from some of the Young Investigator travel award winners. We are indebted to Daniel Wolf of the University of Pennsylvania for this report.
7 April 2009. Human brain development continues into early adulthood, with critical changes in structure and function occurring during the transition from adolescence to adulthood. Schizophrenia is now generally understood to be a neurodevelopmental brain disorder, but neurodevelopmental research on schizophrenia has focused primarily on early brain development. Since the onset of schizophrenia typically occurs during adolescence or early adulthood, it will be critical to learn more about the normal developmental changes occurring during this period, and identify ways in which schizophrenia and the emergence of psychosis may be linked to abnormalities in this developmental transition. This goal inspired an exciting ICOSR meeting session entitled “The relevance of adolescent brain development for the pathophysiology of schizophrenia,” chaired by Peter Uhlhaas of the Max-Planck Institute for Brain Research, Frankfurt. Four speakers described various aspects of cortical development during adolescence, and discussed implications of their data for understanding schizophrenia pathophysiology and for developing novel therapeutic approaches.
Patricio O’Donnell of the University of Maryland described his group’s electrophysiological slice-preparation studies on the development of prefrontal cortex (PFC) microcircuits in the rat in the first talk, “Periadolescent maturation of dopamine actions in the prefrontal cortex.” He focused on the ability of dopamine to regulate the balance of excitation and inhibition in PFC in ways that enhance strong signals and suppress weak signals, thereby increasing signal-to-noise and attention to salient stimuli during cognitive processes such as working memory. In prepubertal rats (one month old), which are normally used for electrophysiological studies, dopamine inputs from the ventral tegmental area act through D1 dopamine receptors to excite PFC pyramidal neurons through synergistic interactions with glutamate acting on NMDA receptors. During the transition from adolescence (one to two months old) into adulthood, the sensitivity of excitation to dopamine and NMDA increases. This change is both quantitative, in terms of left-shifted dose response curves, and qualitative, with the appearance in adult PFC slices of spontaneous prolonged depolarizations and superimposed spikes when NMDA and a D1 agonist are present.
While the transition to adulthood increased dopamine-induced excitation through this D1 mechanism, it also increased inhibition via D2 effects on fast-spiking PFC interneurons. In the prepubertal rat, dopamine exerted mixed effects on interneurons, with D1 activating but D2 inhibiting them. In contrast, in slices from adult rats, both D1 and D2 agonists increased activity of interneurons, thereby indirectly suppressing pyramidal cell firing. Thus, compared with the preadolescent rat, adult rat PFC function was marked by an increase in dopamine’s ability to generate both direct excitation and indirect inhibition of pyramidal neurons, effects expected to contribute to increased fine-tuning and enhanced signal-to-noise in prefrontal function. In closing, O’Donnell noted that rats with neonatal ventral hippocampal lesions, a neurodevelopmental model of schizophrenia with adolescent-onset behavioral abnormalities, did not show the above changes in PFC responses during adolescence but instead remained in the prepubertal mode of function into adulthood. Those findings would be described in detail in a later session at the meeting.
David Lewis of the University of Pittsburgh spoke next on “Developmental refinements in primate prefrontal cortical circuitry,” reviewing his group’s work on dorsolateral prefrontal cortex (DLPFC) development in monkeys. Providing background, he noted that structural and functional abnormalities in DLPFC are strongly linked to schizophrenia. As in humans, working memory performance continues to improve into early adulthood. In monkeys, the dependence of working memory on DLPFC increases with age; reversible cooling lesions of DLPFC begin to affect working memory at ~16 months of age but only cause substantial impairment after puberty (36 months old). Lewis then reviewed three examples of DLPFC circuit components which underwent important developmental changes during adolescence in monkey and were also known to be abnormal in schizophrenia. First, he described developmental changes in excitatory axospinous synapses of cortical layer 3, which are selectively reduced in DLPFC in schizophrenia. In monkeys, spine density on layer 3 pyramidal cells rises in the early postnatal period, then plateaus, then declines (synaptic pruning) from 15-42 months, and then is stable in adulthood. Examination of markers of axospinous synapse functional maturity (low glutamate release probability with both AMPA and NMDA postsynaptic activity) showed that these synapses became fully functionally mature by 15 months, prior to onset of pruning. This contradicted prior assumptions that pruned synapses were immature, indicated that an “exuberant” excess of functional synapses during childhood could theoretically compensate for impaired excitatory signaling in those at risk for schizophrenia.
Second, he described developmental changes in the molecular makeup of pyramidal neuron axon initial segments (AIS) that receive input from chandelier cell interneurons. Ankyrin G, an AIS protein needed for ion channel functioning, is reduced selectively in cortical layers 2 and 3 in schizophrenia relative to both healthy subjects and patients with major depression. In monkeys ankyrin G levels decline during early development and then stabilize, while other AIS proteins such as gephyrin do not decline until adolescence. AIS components, and hence chandelier cell synaptic neurotransmission, therefore show complex developmental dynamics, with periods of rapid change during the early postnatal period and again during adolescence. These dynamics could potentially contribute to protracted maturation of DLPFC-dependent behaviors and to schizophrenia pathophysiology.
Third, Lewis reviewed reciprocal alterations in α1 and α2 subunits of GABA receptors. In monkeys, α1 mRNA and protein expression increases while α2 levels decrease over the course of development even into early adulthood. As α1-containing GABA receptors induce inhibitory postsynaptic currents with faster decay kinetics than those containing α2, IPSC kinetics should speed up over development. In schizophrenia, cortical α1 mRNA is decreased, while α2 mRNA is increased, suggesting the normal developmental change in this ratio may not proceed to completion in schizophrenia, potentially impairing GABA-dependent synaptic plasticity. Lewis concluded with the hopeful suggestion that preventative interventions for individuals at high risk for developing schizophrenia might be profitably directed at enhancing the normal developmental trajectories of the specific synapse types altered in the illness.
Peter Uhlhaas, the session chair, then presented his work on local and long-range synchronizing oscillations in cortical neural networks in his talk, "Maturation of task-related neural synchrony and its relevance for the development and pathophysiology of schizophrenia." Such oscillations are crucial for high-level brain functions, are important in guiding normal brain development, and are impaired in schizophrenia. He reviewed a developmental EEG study of oscillations in healthy human subjects ranging from age 6-21, performing a gestalt face perception task. Accuracy in this task improved throughout that age range; reaction time also sped up with increasing age, except for a transient anomalous slowing in the late adolescent age group. Several types of oscillations, including γ frequency oscillations reflecting local cortical synchronization, phase-locked β oscillations reflecting long-range synchronization, and θ oscillations all increased with age, except for transient reductions in a “destabilization” period of late adolescence, followed by a rapid increase (“restabilization”) into early adulthood. Uhlhaas then described related studies in schizophrenia patients using EEG and MEG, which demonstrated strong reductions in γ, β, and θ synchronizing oscillations, so that schizophrenia oscillatory patterns could be seen as having failed to mature to the adult state.
Uhlhaas concluded with a developmental model of schizophrenia pathophysiology incorporating the above findings, in which an initial insult occurs in the pre-/perinatal period, then additional vulnerability arises during reorganization of cortical networks during early adolescence, with frank psychosis emerging during the transient “destabilization” phase in late adolescence.
Beatriz Luna of the University of Pittsburgh, the final presenter, reported on “Brain system changes underlying the development of working memory through adolescence: neuroimaging studies.” She used an oculomotor delayed response task, requiring working memory to guide saccades. The precision of memory-guided saccades increased with age, with maximal performance not achieved until ~20 years of age. fMRI data showed that adolescent activation patterns were intermediate between children and adults. While regional patterns varied depending on task design, increasing age led, in general, to more efficient use of key regions such as DLPFC, consistent with known synaptic pruning, and enhanced integration of prefrontal activity with a broader network of other regions, which in turn is consistent with progressive myelination of long-range cortical white matter tracts. She concluded that while core aspects of working memory are already available in adolescence, processing becomes more efficient and more integrated during the transition to adulthood. In speculating about connections to schizophrenia, she suggested that a transition to more distributed functional networks in adults might reveal abnormalities in systems that were not previously relied upon; these abnormalities in turn could impede further development toward mature operations and ultimately cause frank psychopathology to emerge.—Daniel H. Wolf.