Functional Roles for Short-term Synaptic Plasticity in Neuronal Networks

Bose, Booth

The primary goal of this project is to derive general

principles that underlie how short-term synaptic plasticity

(STSP) is utilized by neuronal networks in three different

computational and architectural settings. These settings are

motivated by concrete biological examples arising in crustacean

stomatogastric ganglion (STG), rat hippocampus, and cortex.

Specific aims include determining the effect on phase maintenance

of multiple depressing synapses and intrinsic cell properties in

pacemaker-driven networks, determining how multiple depressing

synapses can introduce multiple, co-existent stable firing

patterns in reciprocally-connected networks, and determining how

depressing excitatory and inhibitory synaptic inputs in an

afferent-driven network can generate frequency-selective, steady

state, and transient network responses. The investigators

develop and use techniques of geometric singular perturbation

theory to project and analyze the dynamics of these complicated,

high-dimensional neuronal networks onto lower-dimensional slow

manifolds. These techniques allow them to understand how

different synaptic and intrinsic parameters contribute to and

modulate network behavior. They work closely with

experimentalists who, in a parallel research program, are

investigating the effects of synaptic depression in the

crustacean STG.

Synaptic plasticity refers to the ability of a synapse to

change its strength as a function of its usage. It is widely

found in neuronal circuits across the brain. While experimental

studies of short=term synaptic plasticity (STSP) are necessarily

focused on the particulars of the neural system under

investigation, modeling of the type proposed here can provide

insights into the more general properties of STSP. The

investigators study the possibility that seemingly independent

roles for STSP can be grouped together based on how the network

architecture constrains STSP to operate. Elucidating the general

principles behind these operations provides a framework for

understanding how STSP participates in very diverse neuronal

computations across brain regions. Due to its interdisciplinary

nature, this project is expected to be of interest to members of

the experimental, computational and analytic neuroscience

communities. Additionally, the investigators continue to teach

computational neuroscience and mathematical biology courses that

they recently developed. Graduate and undergraduate students

have the opportunity to work directly with their experimental

collaborators. Thus students become well positioned to continue

pursuits in either experimental or theoretical fields, or in an

area that combines the two. This enhances the development of a

trained workforce at the critical intersection of mathematics and

biology.