Population coding of olfactory information in the moth antennal lobe (AL)

Unlike pheromonal molecules, which are specific ligands for neurons innervating the MGC, an area in the male AL devoted for processing conspecific female sex pheromones, many plant-derived odorants can activate a good portion of the entire AL. To study the population response of these neurons, I use multiunit recording technique (figure panels on the left) to monitor the activities of multiple neurons simultaneously. Simply speaking, some neurons show excitatory responses and other show inhibitory responses, and yet many recorded neurons are not responsive to any given odor stimulus. Using multivariate statistics I analyze how these neurons respond as a group and how individuals contribute to the population response. Some observations lead to another long term project - mapping the spatial distribution of odor-evoked activity in the antennal lobe, and comparing the map across animals. To do so, I combine the electrical recording methods with neuron staining, tissue marking and Amira 3-D reconstruction (Panel A in another figure) in my research.

Back to top

Home

Synchrony generation, detection and behavioral implication

Functional roles of synchronized neural activities have long been proposed in sensory systems, olfactory system included. To test this hypothesis, I conduct dual recordings either using two intracellular or extracellular electrodes from  the macroglomerular complex (MGC) in the AL of male Manduca sexta (figure panels to the right). Stimulated with their natural odor molecules at behaviorally relevant concentrations, the output neurons or projection neurons (PNs) of MGC are more or less synchronized depending on whether they belong to the same or different glomeruli. For example, two “Cumulus” output neurons are better synchronized than two neurons that innervate “Cumulus” and “Toroid” respectively (panel a, b). These uniglomerular PNs (panel B in the linked figure) are excited by the specific input to their associated glomeruli, but also inhibited by the input to their neighboring glomeruli. Such inhibition is termed lateral inhibition (panel C and D in the linked figure). More interestingly, on average the highest degree of synchrony is found to locate at the beginning part of the response, thus called “onset synchrony” (panel d, plus sign). Moreover, synchrony between PNs is positively correlated with the intensity of lateral inhibition (panel b, c), which can be modulated by the input to neighboring glomeruli.  

Synchronized neural activity makes little sense if the down-stream neurons don’t detect it. To study this question, I simultaneously record PNs from AL and their postsynaptic targets in the lateral protocerebrum. If the lateral protocerebral neurons function as coincidence detectors, their response is expected to be correlated with the synchronized activity of PNs in the AL. To manipulate the synchrony between antennal lobe output neurons, GABA receptor antagonists can be surgically injected into the antennal lobe and later assess the moth’s flight behavior in wind tunnel.         

Back to top

Home

Multitasking of Lateral inhibition

• Filtering  cross-excitation

Even in the highly specific sex pheromone system of Manduca moths, we have seen examples that the pheromone sensitive sensilla are activated by some plant odorants at high concentrations, but the PNs are not necessarily affected by the same plant odorants at the same concentrations. This observation leads to a hypothesis that the cross-excitation occurred at the periphery is corrected in the glomerular network (left figure). One possible mechanism that may allow central correction (or filtering) is lateral inhibition.   

Manipulation of information processing to modify behavior

(in collaboration with Dr. Jeff Riffell)

The principles of information processing that we have learnt from physiological experiments need to be tested in behavioral assays, but a common difficulty is to identify the clear relationship between one mode of information processing and a specific behavior. The MGC of male ALs in Manduca sexta provides clear advantage in that aspect. On one hand, the MGC exclusively processes information about conspecific female sex pheromones; on the other hand, pheromones reliably induce zigzag upwind flight behavior. I am interested in correlating the response patterns of MGC projection neurons with the upwind flight behavior of male moths, involving  pharmacologically manipulation of the response patterns of these neurons.     

Back to top

Home

Population response integrated from 12 antennal lobe neurons to two chemically similar odorants, linalool (red dots) and nerol (blue dots). The trajectories describe the response dynamics in a 3-D coordinates composed by the first three principle axes.

A. Three dimensional reconstruction of a male antennal lobe. Cumulus (green) and Toroid (red) are the two major glomeruli comprising the macroglomerular complex (MGC) that is devoted to process conspecific female sex pheromones. Little is known about the 2nd Toroid (yellow). B. Examples of two projection neurons innervating the MGC. The green and red outlines represent Cumulus and Toroid, respectively. C. Intracellular traces showing odor-evoked responses recorded from a Toroid projection neuron. The neuron was inhibited by the ligand (C15) to the Cumulus and excited by its own specific ligand Bal as well as the blend of C15 and Bal. D. Similarly, intracellular recordings showing the response patterns of a Cumulus projection neuron.

Pseudo-colored neuronal firing rate of 50 Toroid and 38 Cumulus projection neurons in response to Bal, ligand of Toroid cells, and C15, a chemically more stable mimic of the natural ligand of Cumulus cells, showing response specificity of these two populations of neurons. Stimulus marks are indicated at the bottom of the figure window. Response magnitude is normalized to the maximum.

Home

Conditional probability plot, based on the data collected from multiunit recordings, shows what is the likely hood of a cell responding to other stimuli (comparison stimuli) under the condition that it responded to a given stimulus (root stimuli). Cells responding to plant odorants (hot color labels) have much higher probability to respond to other plant odorants. The same is true for pheromone (black labels) responsive cells. Therefore, cells from antennal lobe are readily subdivided into two populations - plant odor responsive population and sex pheromone responsive population.  

Home

A bird-view video clip showing a male moth flying upwind to a female pheromone source. The wind blew from right to left. The bright spots on the bottom half of the clip are the reflection of red light bulbs placed above the wind tunnel.