Digital, three-dimensional average shaped atlas of the heliothis virescens brain with integrated gustatory and olfactory neurons

Pål Kvello

doi:10.3389/neuro.06.014.2009

Abstract

We use the moth Heliothis virescens as model organism for studying the neural network involved in chemosensory coding and learning. The constituent neurons are characterised by intracellular recordings combined with staining, resulting in a single neuron identified in each brain preparation. In order to spatially relate the neurons of different preparations a common brain framework was required. We here present an average shaped atlas of the moth brain. It is based on 11 female brain preparations, each stained with a fluorescent synaptic marker and scanned in confocal laser-scanning microscope. Brain neuropils of each preparation were manually reconstructed in the computer software Amira, followed by generating the atlas using the Iterative Shape Average Procedure. To demonstrate the application of the atlas we have registered two olfactory and two gustatory interneurons, as well as the axonal projections of gustatory receptor neurons into the atlas, visualising their spatial relationships. The olfactory interneurons, showing the typical morphology of inner-tract antennal lobe projection neurons, projected in the calyces of the mushroom body and laterally in the protocerebral lobe. The two gustatory interneurons, responding to sucrose and quinine respectively, projected in different areas of the brain. The wide projections of the quinine responding neuron included a lateral area adjacent to the projections of the olfactory interneurons. The sucrose responding neuron was confined to the suboesophageal ganglion with dendritic arborisations overlapping the axonal projections of the gustatory receptor neurons on the proboscis. By serving as a tool for the integration of neurons, the atlas offers visual access to the spatial relationship between the neurons in three dimensions, and thus facilitates the study of neuronal networks in the Heliothis virescens brain. The moth standard brain is accessible at http://www.ntnu.no/biolog/english/neuroscience/brain

Highlights

Challenged by the need to integrate the rapidly growing data in neuroscience, digital brain atlases have become an important tool serving as a database for neural structures with their three dimensional spatial information
In creating the locust brain atlas two procedures were used for comparison, the Virtual Insect Brain (VIB) procedure initially developed for standardisation of the fruit fly neuroanatomy (Jenett et al, 2006) and the Iterative Shape Averaging (ISA) procedure developed to generate the honeybee standard brain (Rohlfing et al, 2001; Brandt et al, 2005)
This study concluded that the VIB procedure using a global and a local rigid transformation followed by a local nonrigid transformation preserves anatomical variability, whereas the ISA procedure using an affine transformation followed by iterative nonrigid registrations reduces the variability

Summary

Introduction

Challenged by the need to integrate the rapidly growing data in neuroscience, digital brain atlases have become an important tool serving as a database for neural structures with their three dimensional spatial information. Three dimensional digital brain atlases have been generated for four species; the population-based quantitative atlas of the fruit fly Drosophila melanogaster (Rein et al, 2002), the average shaped standard atlas of the honeybee Apis mellifera (Brandt et al, 2005) and the locust Schistocerca gregaria (Kurylas et al, 2008), and the recently made standard brain atlas of the hawkmoth Manduca sexta (Jundi et al, 2009). In creating the locust brain atlas two procedures were used for comparison, the Virtual Insect Brain (VIB) procedure initially developed for standardisation of the fruit fly neuroanatomy (Jenett et al, 2006) and the Iterative Shape Averaging (ISA) procedure developed to generate the honeybee standard brain (Rohlfing et al, 2001; Brandt et al, 2005). This study concluded that the VIB procedure using a global and a local rigid transformation followed by a local nonrigid transformation preserves anatomical variability, whereas the ISA procedure using an affine transformation followed by iterative nonrigid registrations reduces the variability

Methods

Results

Conclusion