IN VIVO IMAGING OF ALPHAHERPESVIRUS INFECTION IN THE MOUSE PERIPHERAL NERVOUS SYSTEM

Abstract

Recurrent alphaherpesvirus infections in humans, such as cold sores or shingles, are associated with debilitating peripheral neuropathic pain. The mechanisms that cause such clinical manifestations are still poorly understood. Pseudorabies virus (PRV) is a related alphaherpesvirus for which swine are the natural reservoir, yet this virus is pantropic and infections of non-natural hosts are characterized by a feverish impulse to scratch to the point of self-mutilation, known as the "mad itch" syndrome. Besides its medical burden and disease, PRV is capable of invading the central nervous system by spreading through circuits of synaptically connected neurons. The neuroinvasive attributes of attenuated strains are exploited to understand fundamental aspects of the mammalian nervous system and to trace neural circuits.

The focus of this thesis was to develop an experimental model system for the direct visualization of PRV infection in vivo, and then monitor various aspects of disease progression with either wild type or attenuated PRV strains. We optimized and characterized the salivary circuit, where infection is initiated at the salivary glands, and the peripheral parasympathetic submandibular ganglia (SMG) are imaged. With single-cell resolution, we observed that ganglionic neurons infected with a virulent PRV expressing the calcium sensor GCaMP3 spontaneously generated calcium flashes in synchrony in vivo. Infections with attenuated PRV strains expressing either GCaMP2 or GCaMP3 revealed that the duration and complexity of calcium transients increased over the course of infection but remained asynchronous. Ex vivo imaging of infected salivary tissue showed that the majority of newly made particles during wild type PRV infection of SMG are invested in anterograde transport back to the initial site of infection at the glands, and many intensely bright foci are detected in the infected axon bundles. Finally, we present a model to correlate these results with the known symptomology in vivo.