Synopsis

Brain atlases commonly used for characterizing neurodegenerative processes are frequently referenced to a specific imaging modality. Here we describe a combined gray matter and white matter atlas to be used in the study of a rhesus macaque model of Huntington’s disease. We illustrate how this atlas will be used to integrate diffusion tensor imaging and resting-state functional MRI connectivity with measures of cognitive behavior. Prefrontal WM tracts, cortical and basal ganglia regions are labeled in the same space for characterization of WM microstructure changes and cognitive and motor loop connectivity. The results of preliminary study show that these MRI measurements can identify correlations with cognitive behavior measurements.

Abstract

Introduction

Rhesus macaques can be used as animal models of movement and neurodegenerative disorders. Most brain atlas systems have been referenced to a specific MRI modality and are not specific to the functional and anatomical neural systems of focus. We have initiated studies of a rhesus model of Huntington’s disease (HD) in which AAV-mediated expression of HTT85Q in the caudate and putamen is expected to result in the progressive reduction in white matter (WM) microstructure and functional connectivity in pre-defined motor and cognitive and limbic cortico-basal ganglia networks. Diffusion tensor imaging (DTI) and resting-state functional MRI (rs-fMRI) are effective methods to study the WM microstructure and functional connectivity in the brain [1]. In this study, a framework is proposed to integrate DTI and rs-fMRI measurements with measures of cognitive function in the frontal lobe and basal ganglia in rhesus macaque.

Methods

5 adult female and 1 adult male rhesus macaques were scanned using a Siemens Prisma 3T MRI system with a 16-channel “pediatric” radiofrequency (rf) coil under 1% isoflurane sedation. For each scan session, 3D T1-weighted MP-RAGE and T2-weighted SPACE images were acquired with 0.5 mm isotropic resolution; a resting-state functional image with 1.5 mm isotropic resolution and 784 time points was acquired; six diffusion tensor images with 1 mm isotropic resolution and 30 isotropically distributed diffusion weighted directions with b-values of 1000 s/mm2 and six b=0 images, followed by a b=0 diffusion image with the same geometric parameters except reversed phase-encoding direction were acquired. To integrate the DTI and rs-fMRI measurements, templates were constructed using ANTs [2] and registered to a common coordinate frame. The atlas of Calabrese et al. [3] was registered to the T2-weighted template and that of Adluru et al. [4] was registered to the DTI template. Between the two labelmaps, 22 gray matter regions and 6 WM tracts involving cognitive, motor and limbic, and cortical basal ganglia circuits, were identified, as these are areas affected in HD. DTI and rs-fMRI data were processed with FSL [5] and AFNI [6], respectively. As part of an ongoing R01 project, the same animals were trained to perform the Delayed Non-Match to Sample (DNMS) task of working memory [7] to a criterion of 80% correct and their acquisition rates compared to the MRI data.

Results

Fig. 1 illustrates the regions selected for analysis from the T2-weighted and FA templates (DPPFC: dorsolateral prefrontal cortex; OPFC: orbital prefrontal cortex; Cd: caudate; Put: putamen; IC: insular cortex; HIPP: hippocampus; AMY: amygdala; VPPFC: ventrolateral prefrontal cortex; VPMC: ventral premotor cortex; SSC: somatosensory cortex; GPE: globus pallidus external; AT: thalamus anterior; VPF: ventral prefrontal WM; ACR: anterior corona radiate; ALIC: anterior limb of the internal capsule; PLIC: posterior limb of the internal capsule). Fig. 2 indicates regions of functional connectivity that exhibit either negative or positive correlations between z-score and DNMS 1s acquisition (the number of trials required for a monkey to match to a target object following a 1-second delay) and errors (the number of errors in matching to the target object) to criterion, with p<0.05. Negative correlations (p<0.05) between FA in the dorsolateral prefrontal WM tracts with percent errors at 1s were also observed, indicating monkeys who made relatively more errors learning the DNMS rule also had lower FA within a subset of frontal WM tracts.

Discussion and conclusion

Herein, a framework to integrate diffusion and rs-fMRI measurements in pre-defined cognitive, motor and limbic cortico-basal ganglia networks is described for investigating associations between MRI data and cognitive behavior. This preliminary study with 6 subjects will be expanded to a group of 18 animals that will be exposed to AAV-mediated expression of HTT85Q, and followed longitudinally. Preliminary findings using this approach illustrate how correlations between MRI outcomes and behavior measurements will be characterized. This initial analysis supports that frontal cortical regions are involved in acquiring a working memory behavioral task. This framework facilitates study of the relationship between cognitive behavior and brain neural microstructure and connectivity in nonhuman primate research subjects.

Acknowledgements

References

[1] I. Bohanna, N. Georgiou-Karistianis, A.J. Hannan, G.F. Egan, Magnetic resonance imaging as an approach towards identifying neuropathological biomarkers for Huntington's disease, Brain Res Rev, 58 (2008) 209-225.

[2] B.B. Avants, C.L. Epstein, M. Grossman, J.C. Gee, Symmetric diffeomorphic image registration with cross-correlation: evaluating automated labeling of elderly and neurodegenerative brain, Med Image Anal, 12 (2008) 26-41.

[3] E. Calabrese, A. Badea, C.L. Coe, G.R. Lubach, Y. Shi, M.A. Styner, G.A. Johnson, A diffusion tensor MRI atlas of the postmortem rhesus macaque brain, Neuroimage, 117 (2015) 408-416.

[4] N. Adluru, H. Zhang, A.S. Fox, S.E. Shelton, C.M. Ennis, A.M. Bartosic, J.A. Oler, P.M. Tromp do, E. Zakszewski, J.C. Gee, N.H. Kalin, A.L. Alexander, A diffusion tensor brain template for rhesus macaques, Neuroimage, 59 (2012) 306-318.

[5] J.L. Andersson, S.N. Sotiropoulos, Non-parametric representation and prediction of single- and multi-shell diffusion-weighted MRI data using Gaussian processes, Neuroimage, 122 (2015) 166-176.

[6] R.W. Cox, AFNI: software for analysis and visualization of functional magnetic resonance neuroimages, Comput Biomed Res, 29 (1996) 162-173.

[7] S.M. Zola, L.R. Squire, E. Teng, L. Stefanacci, E.A. Buffalo, R.E. Clark, Impaired recognition memory in monkeys after damage limited to the hippocampal region, J Neurosci, 20 (2000) 451-463.