Gloria Choi

For her graduate studies at Caltech, Choi joined the lab of David Anderson, who at the time was just switching his research focus from neural crest cells to neural circuits underlying innate behaviors.[4] Early into her graduate career, Choi explored the fate specification and differentiation of cells of the central nervous system in development. Her first paper in which she was co-first author highlighted the finding that the transcription factor, Olig2, sequentially defines motor neuron and oligodendrocyte fate specification.[5] However, Choi and her colleagues were curious about the differentiation and specification of astrocytes in the central nervous system. Accordingly, they probed the roles of Olig1 and Olig2 transcription factors in astroglial specification and found that just as Olig1/2 promote motor neuron specification and suppress interneuron specification in the neurogenic phase, Olig1/2 promote oligodendrocyte specification and suppress astroglial specification in the later phase.[6] This work has changed the way scientists think about the “logic” or “decision tree” that governs the differentiation of multipotent neural stems cells into neurons, astrocytes, and oligodendrocytes as they showed that cell subtype is fate restricted prior to restricting the overall neuronal or glial fate.[7]

Choi then led some of the first projects in Anderson's lab exploring the neural circuits underlying innate behaviours.[8] All animals possess the innate ability to detect social stimuli of another animal and exhibit either defensive or reproductive behaviours in response.[8] Choi sought to find the developmental substrate that enabled hardwiring of this type of sensory perception but divergent behavioural outcomes.[8] She found two parallel circuits, both delineated from the Lhx transcription factor family, that project from the posterior medial amygdala to the ventromedial hypothalamus, yet they possessed opposite neurotransmitter types.[8] Even though these projections both receive the same information regarding olfactory cues in the environment, one circuit is implicated in reproductive behaviours and the other is implicated in defensive behaviours.[8] Her work suggests a potential neural mechanism enabling divergent behaviours from the same sensory stimulus that is genetically hardwired in development.[8]

After defending her PhD in 2005, Choi left the West Coast to join the lab of Nobel Laureate, Richard Axel for her postdoctoral training.[9] Continuing to study olfaction, Choi explored how the piriform cortex encodes olfactory stimuli and drives behaviour. She discovered in 2011, that the piriform cortex does not spatially encode odors.[10] To elucidate this finding, Choi stimulated random ensembles of piriform neurons, using optogenetics, and paired stimulation with an unconditioned stimulus, either shock or reward.[10]  She found that after learning, stimulating these same ensembles in the absence of the unconditioned stimulus elicited the corresponding behavioural response.[10] These novel findings suggest that the piriform cortex is able to elicit robust behaviours in the absence of sensory cues.[10]

Oxytocin has been shown to play a critical role in many social behaviours and olfaction appears to be a very important driver of socially behaviours.[14] Due to this knowledge, Choi and her team explored the role of oxytocin in social behaviours driven by olfactory cues.[14] The first publication from Choi's lab at MIT highlighted their discovery that oxytocin signalling is specifically required to learn associations between olfactory stimuli and social cues but not nonsocial cues.[14] Further, they found that activating oxytocin positive neurons enabled social learning and that oxytocin mediated signalling in the piriform cortex is necessary to mediate social learning to both appetitive and aversive olfactory cues.[14]

Choi's interests soon shifted towards exploring a deeper understanding of why mothers with viral infections during pregnancy have offspring with an increased frequency of autism spectrum disorder (ASD).[15] The experimental model for this clinical observation is known as Maternal Immune Activation (or MIA) which Choi used to examine which immune cell population is implicated in driving ASD-like behaviours.[15] Choi and her team found that a specific type of T cell (RORyT-dependent effector T lymphocytes) along with the release of a specific signalling molecule, cytokine IL-17a, are required for MIA driven ASD-like behaviours in offspring.[15] The IL-17a released by T cells in the mother leads to abnormal cortical phenotypes in the offspring suggesting that targeting these specific T cells in the mother might be a means to prevent the development of inflammation induced ASD-like behaviours in offspring.[15]

Choi and her colleagues then explored the underlying mechanisms by which T cells are activated in the immunocompromised mother to promote release of IL-17a and development of ASD-like behaviours in offspring.[16] They found that maternal intestinal bacteria promote T cell differentiation in T helper 17 cells which release IL-17a and mediate the development of cortical abnormalities.[16] Shortly after, Choi and her team identified the main brain region affected in MIA driven models of ASD-like behaviour.[17] They found that the dysgranular zone of the primary somatosensory cortex was the site of abnormalities and that inhibiting this region reduced ASD-like behavioural abnormalities in offspring of mothers with MIA.[17]

Choi and her colleagues recently established the role of IL-17 in mitigating the aberrant social behaviour in both MIA and monogenic models for ASD-like behaviour.[18] The administration of LPS, which induces IL-17a, rescued the behavioural phenotypes in MIA models while direct administration of IL-17a into the primary somatosensory cortex was required to reduce behavioural abnormalities in monogenic ASD models.[18] Their findings highlight a neuroimmune mechanism underlying developmental disorders and that administration of IL-17a during inflammation might provide a possible therapy to reduce neuronal activity on the primary somatosensory cortex and prevent social behaviour deficits.[18]