Skip to Main Content


Gordon Shepherd, MD, DPhil

Professor Emeritus in the Department of Neuroscience

Research Summary

The main research interest of this laboratory is in the neuron as a complex system and in the synaptic organization of neurons into microcircuits in the brain. We focus on the ways that information processing by the neuron takes place through an interplay of the geometry of dendritic branching, the mechanisms of transduction of synaptic or sensory signals, and contributions of passive and active membrane properties.

Research Interests: Olfaction; synaptic organization; brain microcircuits; cortical evolution; computer modeling; neuroinformatics; history of neuroscience

Specialized Terms: Neuronal Dendrites; Dendritic Spines; Synaptic Organization; Olfactory System; Brain Microcircuits; Computational Neuroscience; Neuroinformatics; Functional Connectomes

Extensive Research Description

We use mammalian olfaction as a model system because it has several distinct advantages: the olfactory bulb (OB) receives direct sensory input from olfactory receptors, but is only one synapse from the cortical level of perception; the OB has distinct laminae and cell types which facilitates functional analysis; the OB projects to olfactory cortex (OC), with only a three-layered cortical organization rather than the six layers found in neocortex, which simplifies analysis of activity and connectivitypatterns; and the OB is located on the dorsal surface, which allows access for in vivo imaging and manipulations.

With these advantages we have identified key steps along the way to smell perception. Odor molecules first bind differentially to the olfactory receptors. Activity markers 2-deoxyglucose and fMRI show that the receptor responses are represented by activity patterns in the olfactory glomeruli.

These patterns are processed by feedback and lateral inhibition in microcircuits between mitral and granule cells. The resulting output goes to the three layer olfactory cortex, which we have postulated has a functional organization involving feedback inhibition and excitation. We postulate that this was elaborated in evolution to form six layer neocortex.

Our laboratory contains a computational unit, carrying forward early experiments with Wilfrid Rall, the founder of computational neuroscience and the theoretical foundations of dendritic function. We apply these principles in parallel with experiments to explore mechanisms of information processing in dendritic spines, dendritic trees and cortical microcircuits.

We were among the originators of neuroinformatics, developing the SenseLab Project for the construction of databases for ion currents, receptors, and neurotransmitters in different neuron types, to facilitate the integration of these multidisciplinary data into a new generation of scaled up realistic 3-dimensional computational models of neurons and neuronal systems. Among them is ModelDB, a widely-used curated archive of membrane, neuron and microcircuit models, to enhance the interplay of experiment and modeling for better insight into the cellular basis of brain function.


Research Interests

Anatomy; Dendrites; Information Science; Interneurons; Nervous System; Neurons; Synapses; Pyramidal Cells; Dendritic Spines; Organisms; Phenomena and Processes

Selected Publications