Multiple hypotheses have been put forward regarding cortical circuit dysfunction in autism, including: excitation to
inhibition (E/I) imbalance, hyper- or hypo- connectivity profiles, abnormal processing along the cortical column,
increased variability of sensory encoding, failure of “canonical” cortical computations, and others. Nevertheless, we
currently lack a detailed mechanistic understanding of how cortical circuits malfunction during sensory processing in
autism. This critical knowledge deficit needs to be corrected. Identifying common themes of neural circuit dysfunction
in autism would yield new targets for intervention, potentially opening new windows for restoring function.
Methyl-CpG-binding-protein-2 (MECP2) duplication syndrome is caused by duplication of the MECP2 gene, leading
to progressive intellectual disability and autism in males with 100% penetrance. It is the “mirror image” of Rett
syndrome, caused by loss of MECP2 function. Abnormal synaptic plasticity and E/I imbalance have been causally
implicated in both MECP2-disorders and autism in general and are known to affect cortical function. Notably, we and
others have shown that visual function is abnormal in MECP2-syndromes. Visual cortex, a prototypical “low level”
sensory area, renders itself optimally for studying cortical sensory processing. Here, we mount a comprehensive
effort to understand how sensory processing fails in MECP2-duplication syndrome, by studying the visual cortex.
Recent advances will allow us for the first time to map densely, at single cell resolution, the functional properties and
connectome of a cortical column, one of the fundamental modules of cortical computation, in the MECP2-duplication
mouse model of autism (aim #1). This will be followed by a detailed electro-physiological connectivity analysis of the
cortical circuit to probe underlying mechanisms (aim #2). Data obtained will be used in conjunction with deep learning
techniques to create a sophisticated neuro-realistic model of the cortical circuit, which will allow a detailed, “in
computo,” interrogation of the mechanism of dysfunction and how to ameliorate it (aim #3). Hypotheses formulated
“in computo” can then be tested optogenetically, closing the loop.
Expectations: 1) Obtain the first comprehensive picture of how early sensory processing fails in this model of
autism. 2) Identify specific sub-network connectivity defects reflecting abnormal interneuronal connectivity and
E/I balance, 3) Construct a neuro-realistic model of the cortical microcircuit, which can be used to uncover “in
computo” the mechanism of dysfunction and to probe how one might intervene to repair it. Finally, we will 4)
develop a systematic framework, within which to categorize cortical processing and dysfunction in autism. In the
future, additional models of autism will be brought into this fold to be categorized and systematically studied.
