Layer specific tracing of corticocortical and thalamocortical connectivity in the rodent using manganese enhanced MRI

Abstract

Information about layer specific connections in the brain comes mainly from classical neuronal tracers that rely on histology. Manganese Enhanced MRI (MEMRI) has mapped connectivity along a number of brain pathways in several animal models. It is not clear at what level of specificity neuronal connectivity measured using MEMRI tracing can resolve. The goal of this work was to determine if neural tracing using MEMRI could distinguish layer inputs of major pathways of the cortex. To accomplish this, tracing was performed between hemispheres of the somatosensory (S1) cortex and between the thalamus and S1 cortex. T₁ mapping and T₁ weighted pulse sequences detected layer specific tracing after local injection of MnCl₂. Approximately 12 h following injections into S1 cortex, maximal T₁ reductions were observed at 0.6 ± 0.07 and 1.1 ± 0.12 mm from the brain surface in the contralateral S1. These distances correspond to the positions of layer 3 and 5 consistent with the known callosal inputs along this pathway. Four to six hours following injection of MnCl₂ into the thalamus there were maximal T₁ reductions between 0.7 ± 0.08 and 0.8 ± 0.08 mm from the surface of the brain, which corresponds to layer 4. This is consistent with terminations of the known thalamocortical projections. In order to observe the first synapse projection, it was critical to perform MRI at the right time after injections to detect layer specificity with MEMRI. At later time points, tracing through the cortical network led to more uniform contrast throughout the cortex due to its complex neuronal connections. These results are consistent with well established neuronal pathways within the somatosensory cortex and demonstrate that layer specific somatosensory connections can be detected in vivo using MEMRI.

Introduction

The anatomical and functional architecture of the somatosensory cortex has been studied extensively in the past using neuronal tracing methods. From the pioneering work of Woolsey over fifty years ago, the primary somatosensory cortex was found to be separated into regional boundaries defining areas of the brain where sensory information is processed from inputs from the periphery (Woolsey and Fairmen, 1946). The use of anterograde and retrograde tracing techniques have been critical for working out the neuronal pathways responsible for the organization of the brain into functional regions and for understanding the specific inputs/outputs that connect brain regions (for review see Kobbert et al., 2000). Despite the tremendous knowledge gathered over fifty years from these architectural studies, a shortcoming of these techniques is that they require histology to visualize the tracer. Moreover, unless specific types of retroviruses are used, then traditional neural tracers can detect connections only one synapse away. Until recently it was not possible to gather data in vivo making it difficult to make repeated measurements within individual subjects over time in order to study changes in connections that might occur due to learning, plasticity, and pathological conditions.

The great success of neuroimaging techniques such as positron emission tomography (PET), magnetic resonance imaging (MRI) and optical imaging at measuring anatomical and functional events related to neuronal activity within living subjects has led to great interest in expanding the information about neural systems that can be obtained with these techniques (Fox et al., 1986, Kwong et al., 1992, Ogawa et al., 1993, Denk et al., 1990). In particular, MRI has proven to be an excellent method for examining the complex structure and function of brain tissue. The advent of functional MRI (for review see Mathews and Jezzard, 2004), and diffusion MRI (Le Bihan, 2003) has, in combination with more traditional anatomical MRI, given an unprecedented view of the human brain. Indeed, the development of high field MRI for animal and human work is leading to image resolution beginning to approach that seen in traditional histological techniques (Aoki et al., 2004a, Aoki et al., 2004b, Duyn et al., 2007).

There has been growing interest in the use of Manganese Enhanced MRI (MEMRI) to enable MRI to detect neuronal activity (Lin and Koretsky, 1997, Aoki et al., 2002, Yu et al., 2005), to enable MRI to measure a number of aspects of neuroarchitecture (Watanabe et al., 2002, Aoki et al., 2004a, Aoki et al., 2004b, Silva et al., 2004), and to trace specific neuronal pathways (Pautler et al., 1998). Indeed, the ability of MEMRI to obtain information about neuronal connectivity in analogy to classical anterograde track tracers has led to its widespread use to study connectivity in rodents, birds, and monkeys (Saleem et al., 2002, Van der Linden et al., 2002, Leergaard et al., 2003, Pautler, 2004, Bilgen et al., 2006, Chuang and Koretsky, 2006, Murayama et al., 2006, Bearer et al., 2007, Lee et al., 2007, Pelled et al., 2007, Van der Zijden et al., 2007, Minoshima and Cross, 2008, Silva et al., 2008). The fact that Mn²⁺ moves in an anterograde direction along the appropriate neuronal pathway and can cross synapses has led to the ability of using MRI to detect functional neural circuits (Pautler et al., 1998, Pautler and Koretsky, 2002, Saleem et al., 2002, Pautler, 2004).

Despite the growing body of work using MEMRI track tracing it is not yet clear at what level of resolution connectivity can be measured. There is growing evidence that when delivered to the whole brain, Mn²⁺ leads to enhancement of MRI in a manner that can be sensitive to detection of layers in the olfactory bulb, hippocampus, and cerebral cortex (Aoki et al., 2004a, Aoki et al., 2004b, Watanabe et al., 2002, Silva et al., 2004). One report has shown preliminary data that indicated when MnCl₂ was injected into the thalamus layer specific inputs into layer 4 of the cortex could be detected (Silva et al., 2008). Evidence has been presented for layer specific tracing from the olfactory bulb with MEMRI as well (Pautler and Koretsky, 2002). The present work was designed to build on this initial finding to determine if MEMRI based track tracing could resolve layer inputs into the cortex along two major pathways. MEMRI tracing was performed between hemispheres of the somatosensory (S1) cortex, and between the thalamus and the S1 cortex. Results show that the known layer specific inputs of these pathways into somatosensory cortex could be detected by MEMRI. Critical to detecting layer specific inputs was performing MRI at the specific time after local injection of MnCl₂. This opens the possibility to assess layer specific changes in patterns of connectivity in individual animals associated with learning, plasticity and pathological states such as stroke or peripheral nerve injury.