Date of Award

5-2012

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Program

Biomedical Sciences

Track

Neuroscience

Research Advisor

Detlef Heck Ph.D.

Committee

William Armstrong, Ph.D. Joseph Callaway, Ph.D. Robert Foehring, Ph.D. Chrysanthe Preza, Ph.D.

Abstract

The mammalian brain forms neuronal networks and microcircuits with cell-type- and anatomical-specific synaptic connections. Despite great advances in elucidating the cellular physiology of the nervous system, little is known about the computational processes occurring at the level of neuronal microcircuits. Much success has been reported in describing the synaptic input patterns of many brain regions and cell types using photostimulation systems; however, these systems are severely limited in their ability to study the integration of synaptic input from multiple synchronous or temporally correlated presynaptic locations. Here we describe a system that allows the generation of arbitrary 2-D stimulus patterns with thousands of independently controlled sites to manipulate the activity of populations of neurons with high spatial and temporal precision. The PC-controlled Digital-Light-Processing (DLP) based system updates the 780,000 parallel photostimulation beams, or pixels, at a maximum rate of 13 kHz. With the currently used projection objective, the pixel sizes at the plane of focus are 7.3 µm2 . The high-power UV laser source used in this system provides a light flux density sufficient for bins of 8x8 pixels (21.6 µm x 21.6 µm) with dwell times as low 3 ms to reliably induce action potentials in 2.5 mM MNI-caged glutamate. At these settings the effective diameter of a glutamate uncaging site is < 86 µm, which is equivalent to most other UV photostimulation rigs. With DLP photostimulation, sub-threshold responses and action potentials can be synchronously induced at thousands of sites over a 2.76 mm x 2.07 mm area, a capability unmatched by any other current system. This DLP-based system has the unique capability to investigate normal and diseased circuit properties by investigating neuronal responses to spatiotemporally complex activity patterns. This technique was used to investigate the temporal integration of synaptic input in the whisker barrel cortex of mice. The neocortex is organized into layers, with neuronal networks and circuits formed by layer-specific connections. While the anatomical organization of these circuits has been well characterized, the information processing and coding performed by these ensembles is poorly understood. A key component of this investigation concerns the transmission and transformation of the neuronal representation from one neuronal pool to the next. In the rodent somatosensory barrel cortex, histologically-distinguishable “barrels” in layer 4 (L4) receive principal input from a single whisker. L4 projects to layer II/III (L2/3), where the circuit diverges to multiple postsynaptic targets. Using the DLP-photostimulation system, we modulated the synchronicity of action potentials in L4 cells while recording from L2/3 in an acute slice preparation. This data shows that synchronous activity in L4 neurons is highly effective at eliciting strong spiking responses in L2/3 pyramidal cells, while asynchronous L4 activity fails to drive L2/3 to action-potential threshold. Pharmacological manipulation of the slice-bathing solution has suggested that this phenomenon is AMPA-receptor dependent and modulated by NMDA receptor activity. Intracellular pharmacological manipulations suggest that postsynaptic conductances also play a role in the nonlinear L2/3 synaptic integration of L4 activity.

DOI

10.21007/etd.cghs.2012.0153

Comments

Two year embargo expired May 2014

Included in

Neurosciences Commons

Share

COinS