August 2006 - August 2012
Our lab is using electrophysiology, optogenetics, and psychophysics to understand the principles of the sensory information processing. Specifically we are focused on two questions: 1) how is odor information coded in the brain of the awake, behaving mouse? And 2) how is information relevant to animal behavior extracted by the brain? In short, we want to know what the mouse’s nose tells its brain.
Recently, our laboratory has been focused on temporal aspects of olfactory coding. We discovered that a) olfactory neuronal code at the level of olfactory bulb is temporally very precise (~10 ms) [Shusterman-2011], and b) the mammalian olfactory system can read and interpret temporal patterns at this time scales [Smear-2011]. Our efforts are directed towards establishing causal connection between neuronal coding and animal behavior.
Our lab moved to the Neuroscience Institute at NYU Medical School in September 2012.
Until recently, most of our knowledge of the function and organization of olfactory sensory systems has been obtained from preparations from anesthetized animals. However, the recording of neuronal activity in awake animals is of critical importance, because sensory information processing, even at the early stages, is modulated by fully active feedback from several areas of the brain. Learning-related changes in neural representation of a stimulus can be recorded in real time, and specific quantifiable behaviors can be used to interpret the outcome of information processing by alert animals.
The power of combining psychophysical and neurophysiological methods has been demonstrated in the study of visual information processing, especially in primates. As a model organism, mice enable us to use modern genetic tools to monitor, modify, and control brain circuits. Using a mouse to study sensory information processing leads naturally to focusing on olfaction because of its high relevance to rodent behavior.
Odor representation in the awake mouse
Mammals sense odors through a large number of olfactory receptor neurons in olfactory epithelium. Each sensory neuron expresses one and only one gene out of a large family of olfactory receptor genes (~1200 genes in mice), but all axons of the receptor neurons, which express the same gene, converge into one or two small areas in the olfactory bulb, called glomerulus. Mitral/tufted cells, the first recipient of odor information after receptors, take their inputs from the glomerulus and send an axon to the olfactory cortices and other brain areas via the lateral olfactory tract. We are interested in the coding of olfactory information at the level of mitral/tufted cells.
What does a nose tell the brain? In our previous work, we found a striking difference in odor responses between the awake and anesthetized state (Rinberg, et.al., 2006). The spontaneous activity of mitral cells in the awake mouse is significantly higher compared to that in anesthetized mice, and odor responses are significantly smaller on the background of spontaneous activity. How are odors represented by activity of mitral/tufted cells in the olfactory bulb of awake mouse? To answer this question we need to precisely control the stimulus and monitor the behavior that follows. Our recent experiments demonstrated that neuronal response of mitral/tufted cells are locked to the sniffing/breathing pattern with very high temporal precision of ~10 ms. (Shusterman, et.al., 2011). The olfactory code at the level of the mitral/tufted cells is temporally and spatially very diverse. One odor is represented by activity of many cells and each cell exhibits temporally diverse patterns of excitation and inhibition in response to odors.
Temporal locking of olfactory responses to the phase of the sniffing cycle constitutes coding invariance in respect to the frequency of animal breathing/sniffing. Humans can identify odors independently of how fast they inhale an odor, which may be explained by observed coding invariance. The current work in the lab is directed towards understanding concentration invariance: an odor identity is independent of its concentration.
From sensory coding to behavior
What are the features of this code that are used to produce behavior related to olfactory input? To approach this question we developed behavioral/electrophysiological paradigms to confine the features important for behavior temporally and spatially. In collaboration with Thomas Bozza (Northwestern University), we developed transgenic mice, which express Channel Rhodopsin in all olfactory receptors and in one class of olfactory receptor neurons. The opto-genetic approach allows us to control the stimulus precisely in time, by using temporally well-defined light stimulation, and in space by exciting a single glomerulus in mouse olfactory bulb.
Using opto-genetics allowed us to decouple olfactory stimulation and mouse sniffing. By illuminating epithelium in mouse nose via optical fiber, we are able to present spatially identical and temporally confined stimulus to the mouse olfactory system. This allows us to study the perception of a single stimulus cue, the timing relative to the sniff cycle, sniff phase. We found that mice can discriminate light-evoked input that is shifted in the sniff cycle by as little as 10 ms [Smear-2011].
Experiments with single glomerulus excitation are directed toward understanding the role of a single channel of information, one receptor type, in the distributed coding of olfactory stimulus.
Combining optogentic stimulation with electrophysiology and behavior allows us to dissect complex problem of olfactory perception into smaller features, and gain understanding of how these features are represented and processed in the brain.
- mouse olfactory genetics
- data analysis
- human psychophysics data analysis
- human psychophysics
- mouse/rat behavior
- modeling of air flow dynamics in the mouse nose