My laboratory studies cellular and synaptic mechanisms of odor information processing in the olfactory system. In the nose, odorants are detected by special G-protein coupled membrane receptors, the olfactory receptors, and are coded into electrical impulses or spikes by millions of olfactory sensory neurons. These signals are relayed through olfactory nerve fibers to an array of ball-shaped structures in the olfactory bulb known as glomeruli. Each glomerulus captures signals from one olfactory receptor out of a large multi-gene family that may include hundreds or thousands of distinct receptors. Glomeruli function as parallel data lines, conveying information about specific molecular features of odorants encoded by different receptors. Different odorants evoke different glomerular activity patterns, creating unique odor maps in the brain. The number of different kinds of receptors determines the diversity of chemicals that can be detected and discriminated. Humans express ~350 functionally distinct receptors, while macrosmatic mammals such as mice and rats may express over 1,000 receptors.
How do neurons in the olfactory bulb respond to and process input signals received by glomeruli from olfactory receptors? How are these signals transformed and reshaped by olfactory bulb neural circuits? How is the processed data packaged by olfactory bulb principal cells, the mitral and tufted cells, for transfer to higher cortical centers? Recent studies suggest that olfactory coding may depend not just on the identities or patterns of activated glomeruli, but also on the timing of their responses. Incoming sensory signals already contain timing information, manifested as varying delays in the activation of presynaptic terminals converging to different glomeruli. A single sniff of the appropriate odorant can stimulate a glomerulus sufficiently to trigger a coordinated cascade of glutamate excitation that culminates in rapid spiking of mitral and tufted cells. Spike synchronization is observed in pairs of mitral cells linked to the same glomerulus, and this is thought to act as a temporal ‘glue’ that binds together signals encoding distinctive molecular features recognized by one or more olfactory receptors.
Figure 1. Optical imaging of glomerular activity patterns.
Different volatile chemicals bind to and activate different sets of olfactory receptors in the nose. In the olfactory bulb each glomerulus receives afferent inputs from one olfactory receptor, creating a spatial representation of receptor identity. Activation maps for various odorant compounds can be visualized by dynamic optical imaging of fluorescent indicator probes. Here we used a transgenic mouse with a green fluorescent protein (GFP)-linked synaptic vesicle associated membrane protein, synaptophluorin (spH), expressed in olfactory nerve terminals to report glomerular activation. The series of images shows specific glomeruli responding to increasing vapor concentrations of methyl-isoeugenol, a derivative of clove oil.
Work in my laboratory has been directed towards elucidating synaptic and dendritic signaling mechanisms that determine the coordination and timing of neuronal activity in the olfactory bulb. By applying patch-clamp and optical stimulation or recording techniques to in vitro slice preparations, we can trace the flow of signals through dendrites and synapses of local circuits. The rules governing this signal traffic will determine the dynamics of excitation and inhibition of principal cells, and the distribution of spike activity among parallel neural assemblies activated by odorants. For example, we find that a specialized class of principal cells, the external tufted cells, can self-amplify their excitation through glutamate autoreceptors, and that these receptors can couple to intracellular calcium signaling pathways. These cells are hypothesized to be pacemakers that coordinate sniff-evoked activity in glomeruli. We are also investigating the synchronization of spikes in another class of principal cells, the middle tufted cells, which are a major conduit for transfer of odor information to anterior regions of olfactory cortex. We find that spike synchrony in both mitral and tufted cells appears to rely on electrical coupling through glomerular gap junctions. Active backpropagation of spikes from cell bodies into glomeruli facilitates spike coupling between cells, and we have demonstrated that this occurs in the branching dendrites of mitral cells of the accessory olfactory bulb. We have found that the long secondary dendrites of mitral cells are major highways for spike traffic across the olfactory bulb. They laterally connect mitral cells of widely separated glomeruli, and in cooperation with local interneurons, the granule cells, they may assist in synchronizing spikes to bind output messages of distant glomeruli. Dendrites of mitral and granule cells are sophisticated signal multiplexing devices, able to conduct action potentials at the same time as they send or receive synaptic signals.
Figure 2. Spike synchronization in mitral and tufted cells.
Using an in vitro olfactory bulb slice, mitral and tufted cells with dendrites projecting to the same glomerulus were identified by loading cells with fluorescent tracer dyes. Simultaneous recordings of spike activity in cell pairs was recorded using two patch-clamp electrodes. For cell pairs connected to the same glomerulus, the number of synchronized spike pairs was significantly higher than expected from random coincidence. This could be seen directly in paired recordings, and was quantified as a sharp peak at zero time lag in the spike cross-correlogram
Another topic of interest in my laboratory is the role of various neuromodulators in regulating and shaping patterns of neuronal activity in the olfactory bulb. We are investigating intrinsic pathways that express the peptide co-transmitter cholecystokinin. The cholecystokinin synthesizing neurons include superficial tufted cells that send intrabulbar projections to glomeruli on opposite sides of the bulb receiving afferent input from the same olfactory receptor. Peptidergic transmission may be involved in coordinating the timing of circuits in mirror symmetric glomerular maps. We are also studying, in collaboration with the Gelperin lab, physiological effects and mechanisms of action of nitric oxide (NO) on bulb circuits. Specific subpopulations of interneurons, including some granule cells and periglomerular cells, stain strongly for the neuronal isoform of nitric oxide synthase. Using custom-fabricated electrochemical micro-sensors, we can detect both tonic and stimulus-evoked production of NO in the bulb. Our efforts are focused on unraveling the complex effects of NO on synaptic coupling and oscillatory dynamics of mitral/ tufted cells and granule cells. Odorant stimulation elicits gamma frequency (> 40 Hz) local field potential oscillations in ensembles of olfactory bulb neurons, a sign of mass synchrony of postsynaptic potentials. As a freely diffusible gaseous transmitter, NO is uniquely qualified to coordinate oscillatory synchrony in large populations of cells. We are testing the hypothesis that it plays a direct role in dynamic temporal coding, as well as in the storage of odor memories through spike timing-dependent synaptic plasticity.
Figure 3. Nitric oxide production in the olfactory bulb.
Neurons in the olfactory bulbs of mice under anesthesia can respond physiologically to inhaled odorants. When we inserted a nitric oxide (NO) sensitive electrochemical microprobe into the granule cell layer of the olfactory bulb, we detected transient elevations in NO immediately following the presentation of odorant stimuli. This indicated that nitric oxide synthase was activated in granule cells by olfactory sensory inputs.
We have also initiated a project to explore the contributions of presynaptic receptor mechanisms to olfactory signal processing. In the laboratory, olfactory model systems have most often been probed with chemically pure odorant stimuli. What happens when the nose is challenged with complex odorant mixtures? Certain volatile compounds have been shown to interact antagonistically at olfactory receptor binding sites, and this has the potential to selectively suppress glomerular output. We are investigating these phenomena in vivo by optical imaging of odor-evoked presynaptic glomerular activity in transgenic mice whose olfactory receptor neurons express the fluorescent probe synaptophluorin.