Functional and multi-modal characterization of cell types in selected circuits

Tolias, Andreas

Abstract

Recentadvancesinsingle?celltranscriptomicsindissociatedcellshavepermittedtheunbiasedclassification ofunique,molecularly?definedcelltypesinmanybrainregions.However,linkingthesemolecularly?defined celltypestotheircorrespondingmorphological,physiological,andfunctionalphenotypesremainsamajorchal? lengeinthefield.Wehaveassembledaninterdisciplinaryteamcapableofcombiningandfurtheroptimizing cutting?edgetechnologiesincludingPatch?seq(amethodwedevelopedthatcombineswhole?cellpatch?clamp recordingsandsingle?cellRNAsequencing),multi?photoncalciumimaging,multiplexedfluorescentinsituhy? bridization(MERFISH),andstate?of?artmachinelearningtoaddressthisgapinknowledge.
In aim1, duringthe firsttwoyears,wewillprovideacomprehensivecensusofthecelltypesthatcomprisethemouseprimaryvisual cortexcircuitrybylinkingsinglecellRNA?seq,morphological,andinvitroelectrophysiologicaldata.Duringthe last three years, we will define the cell types of higher order visual areas. In addition, we will characterize a distributednetworkoffivesubcorticalregionsinvolvedinthecontrolofinnatesocialbehaviors,includingthe ventromedialhypothalamicnucleus,medialpreopticnucleus,anteriorhypothalamicnucleus,posteriorbednu? cleus of the stria terminalis, and posterior medial amygdala. By studying both cortical and subcortical brain regions,wewillbeabletocomparesimilaritiesanddifferencesacrossthoseregions,aswellascompareprinci? plesofcelltypeorganizationbetweenevolutionarilyancientsubcorticalandmorerecentlyevolvedcorticalre? gionsofthebrain.
In aim2, wewillstudythefunctionalpropertiesoftranscriptomically?definedcelltypesin thevisualcortexofthemouse(areasV1,LM,PMandAM).Wewillperformmulti?photoncalciumimagingin behavingmiceduringthepresentationofavarietyofvisualstimulitocharacterizeindetailthereceptivefield propertiesoftheseneuronsfollowedbyMERFISHtoidentifythegeneticprofileoftherecordedneurons.We willemploythistobothCrelinesthatlabelknownbroadclassesofneuronsaswellasdenseimagingofcortical populationstoprovideacomplete,specific(i.e.inthesameanimal)characterizationofbothcelltypefunction andmolecularprofiling.Thecombineddatafromthesetwospecificaimspromisetoprovidethemostcomplete understandingofcelltypestodate,includingexpressionprofiles(e.g.ionchannelandreceptorlevels)morphol? ogy,single?cellelectrophysiology,andinvivofunctionalproperties.Themethodsandpipelinesthatwillbeop? timizedinthissectionoftheproposalwillalsolaythefoundationtofurtherapplythesemethodsindifferent partsofthebrainaswellasstudyanimalmodelsofdiseases.

Public Health Relevance

Recent advances in single-cell transcriptomics in dissociated cells have permitted the unbiased classifica- tion of unique, molecularly-defined cell types in many brain regions, however, linking these molecularly- defined cell types to their corresponding morphological, physiological, and functional phenotypes remains a major challenge in the field. We will use cutting-edge technologies including Patch-seq, multi-photon calcium imaging, multiplexed fluorescent in situ hybridization, and state-of-art machine learning to address this gap in knowledge. The combined data promise to provide the most complete understanding of cell types at molecu- lar, structural and functional levels to date, which will lay the foundation to further apply these methods in different parts of the brain as well as study animal models of diseases.