As our understanding of network architecture and the emergent properties of nerve networks gets ever more detailed,the need for network models to explain and suggest experiments continues to grow. Dr. Harirngton has adapted a 64electrode planar electrode array to record from the cerebral ganglia of snails. With the multi electrode array, Dr.Harrington and her students are recording the neural activity related to sensory processing in two different snail modelsystems. One of these models is the wolfsnail, a predatory snail that tracks its prey (other snails) by following theirslime trails, detecting the slime with a unique, specialized sense organ. Dr. Harrington and her students have recordedvast amounts of data about the activity of the cerebral ganglia in the snails and changes that occur in that activity inresponse to slime and other stimuli applied to the sensory epithelia. Analyzing the data and understanding itssignificance is a major computational challenge requiring the tools of mathematical biology. The usefulness of the dataas a tool for understanding neural processes will be greatly enhanced by using the data to inform the development ofcomputational models of neural integration and decision-making processes. The data collected from wolfsnail gangliaand those of other snails are being analyzed in three ways: first with a spike sorting program that counts and correlatesneural spike activity across all 64 electrodes to calculate and compare spike frequency and synchronization across theelectrode array, A second approach uses cross-correlation to identify changing patterns of synchronized activity. Athird approach will use an Independent Component Analysis (ICA) algorithm to decompose the activity recorded at the64 electrodes and identify different source signals contributing to the total signal. This approach has been used fordecomposition of evoked field potentials in human EEG and MEG applications, neural recording techniques that areemulated by our invertebrate recordings in many respects. The data collected with the electrode array will be correlatedwith the activity of individual cells recorded electrophysiologically in order to determine the contributions to thenetwork activity attributable to the different types of cells in the ganglia. Combining data about the spatial and temporalpattern of neural activity across the ganglia with information about the biophysical propoerties of individual cells willenable us to develop network and computational models of sensory processing in an invertebrate model system.
| Patel, Kinjal; Shaheen, Nagma; Witherspoon, Jessica et al. (2014) Mucus trail tracking in a predatory snail: olfactory processing retooled to serve a novel sensory modality. Brain Behav 4:83-94 |
| Zhang, Hongmei; Wu, Chia-Yen; Wang, Wenlan et al. (2011) Interneuronal synapses formed by motor neurons appear to be glutamatergic. Neuroreport 22:809-13 |
| Zhang, Hongmei; Robinson, Natallia; Wu, Chiayen et al. (2010) Electrophysiological properties of motor neurons in a mouse model of severe spinal muscular atrophy: in vitro versus in vivo development. PLoS One 5:e11696 |
| Zhang, Hong-Mei; Robinson, Natallia; Gómez-Curet, Ilsa et al. (2009) Neuronal and network activity in networks of cultured spinal motor neurons. Neuroreport 20:849-54 |
