The connectomics of brain demyelination: Functional and structural patterns in the cuprizone mouse model
Graphical abstract
The authors demonstrate via non-invasive resting state fMRI that intrinsic functional connectivity is fundamentally reorganized in demyelinated mouse brain, paralleled by widespread damage of structural scaffolding and strong modulations of default mode and hippocampal networks.
Introduction
The elucidation of the brain functional and structural connectome is a major challenge in neuroscience. Network topology and intrinsic neural connectivity is fundamental for understanding the mechanisms underlying the broad range of brain functions like memory, learning and cognition as well as complex behavioral patterns. Capable of coordinating such a variety of actions, the brain combines a regional specialization and segregation of functions with a complex interconnectivity (Reid, 2012). Relying on the structural elements, a hierarchically organized functional communication emerges, particularly suited for efficient local neuronal operations as well as global integration of segregated functions (Park and Friston, 2013). Such a complex network is vulnerable to many sources, like external influences, physiological and psychological changes or immunological events, which all can result in connectivity alterations (Fornito et al., 2015).
One feature of pathological brain function is that a perturbation is rarely restricted to a single area but influences via its connections to other regions. The network organization and dynamics have therefore an impact on the course of a disease, whereas at the same time the disease affects brain connectivity. We thus non-invasively investigated in the mouse this interplay ‘pathology – brain connectome’ in the case of demyelinating pathology, using the cuprizone (CUPR) mouse model (Matsushima and Morell, 2001). Described as the occurrence of myelin damage, from a clinical perspective, the pathological features of demyelination are revealed in pathologies like multiple sclerosis or leukodystrophies (Baumann and Pham-Dinh, 2001). Regardless of its cause, demyelination impairs function: the acute loss of a myelin internode is associated with conduction impairments (Baumann and Pham-Dinh, 2001). Without consistent recovery, at long term myelin breakdown leads to axonal degeneration (Budde et al., 2008), with important consequences on information processing and thus brain function. Therefore, the mutual influence between the neural network changes and the onset, expression and the course of the disease could only be read-out from comprehensive maps of the connectivity architecture, investigated at both structural and functional levels. This goal has driven several recent large-scale efforts that led to generation of increasingly detailed maps of brain connectivity in various species, including mouse (Abe et al., 2012, Oh et al., 2014, Zingg et al., 2014). However, combined non-invasive insight into the structural and functional brain connectome can only be obtained so far via magnetic resonance imaging (MRI) (Deco and Kringelbach, 2014). MRI holds two complementary and non-invasive modalities for probing these features. Diffusion tensor imaging (DTI) is measuring and tracking the diffusion of water molecules along axonal pathways – eventually revealing anatomical connections and therefore structural connectivity (Mori and Zhang, 2006). Resting state functional magnetic resonance imaging (rsfMRI) detects spontaneous, low-frequency fluctuations in the blood oxygen level dependent (BOLD) signal and their temporal correlations – revealing spatial pattern of activation maps referred to as functional connected regions (Fox and Raichle, 2007). The application of these techniques in humans has shed new insight on how diseases affect the brain functioning by altering the neuronal connectivity and has uncovered common features for various pathologies, based on brain networks reorganization (Fox and Greicius, 2010, Lerner et al., 2014).
Additionally, conceptual advances raise hope for not only mere description of pathologic features but for generating hypotheses with prediction value. Such information on network topology could be used to characterize and model vulnerability and resilience to disease and eventually to inform on response to drug therapies (Aung et al., 2013, Fox and Greicius, 2010). For analysis of the mechanisms underlying the development of myelin disorders as well as for the development of new therapeutic strategies, translational studies in mouse models of brain disorders are of great importance (Harsan et al., 2008, Hussain et al., 2013). Exploring with rsfMRI the functional network fluctuations in animal models of demyelination could therefore reveal patterns that might be used as imaging biomarkers of the disease course with high translational value.
The CUPR mouse model of demyelination is a well-established and extensively used model for myelin pathology, with a chronic state of demyelination after twelve weeks of CUPR treatment (Torkildsen et al., 2008) and and it is interesting to look into the connectivity of this model of severe pathology as it is often used for testing remyelinating strategies (Harsan et al., 2008, Kipp et al., 2009). Here, we investigated with rsfMRI the impact of experimentally induced pathology on the functional connectivity (FC) networks of CUPR treated female C57BL/6N mice. In order to clarify how functional interactions arise on the basis of structural connectivity, we combined rsfMRI with DTI and fiber tractography to fully characterize the pathologically induced alterations of mouse brain connectome or eventually identify remodeling features resulting from possible compensatory mechanisms.
Section snippets
Ethics statement
All procedures were conducted in compliance with the national guidelines of the German animal protective law for the use of laboratory animals and with permission of the responsible local authorities for the University Medical Center Freiburg (Regierungspräsidium Freiburg, permit numbers: G-08/15).
Animal setup
C57BL/6N female mice, purchased from Charles River Laboratories (Sulzfeld, Germany) at 7 weeks of age, were kept for the whole duration of the experimental procedures under standard animal room
Cuprizone induced microstructural changes reflected by DTI
To non-invasively verify if CUPR treatment had demyelination-related effects in the mouse brain, we used the T2-weighted MRI contrast and DTI and fiber tractography derived parameters (fiber density - FD and fractional anisotropy – FA) as measures for such structural modifications (Fig. 1). The high content of lipids and proteins of myelin generated low signal intensity and therefore dark contrast along WM tracts of CTRL animals in T2-weighted imaging (Fig. 1A, upper panel). CUPR administration
Discussion
Connectomics research has gained significant importance to elucidate features of human brain network in health and disease. Repertoires of neural circuits mediating specific functional processes have been generated via analysis of spontaneous BOLD fluctuations in the brain at rest (rsfMRI) (Fox and Raichle, 2007), as well as from diffusion based tractography (DTI and fiber tracking) (Mori and Zhang, 2006). Their dynamics and distinct patterns of alterations may manifest across an array of
Author Contributions
N.S.H. and L.-A.H. conceived and designed experiments/interpreted results/wrote paper; N.S.H. and A.E.M. performed experiments/collected data; N.S.H. and L.-A.H. analyzed data; H.-L.L., M.R. and T.B. contributed analysis tools; J.H. and D.v.E. contributed reagents/materials/equipment.
Conflict of Interest
The authors declare no conflict of interest.
Abbreviations
Abbreviations according to the Allen Mouse Brain Atlas (http://mouse.brain-map.org/static/atlas)GRAY MATTER ACA Anterior Cingulate Area ACB Nucleus Accumbens AI Agranular Insular Area AUD Auditory Areas BST Bed Nuclei of Stria Terminalis CA3 Ammon's Horn Field CA3 COA Cortical Amygdalar Area CP Caudoputamen CTXsp Cortical Subplate DG Dentate Gyrus DP Dorsal peduncular area GU Gustatory Areas HPF Hippocampal Formation HY Hypothalamus ILA Infralimbic Area LSX Lateral Septal Complex LZ Hypothalamic Lateral Zone MB Midbrain MO
Acknowledgments
This work was supported by the Brain Links Brain Tools (BLBT) cluster of excellence from Freiburg (MouseNet 31 project), ERANET-Neuron AF12-NEUR0008-01 -WM2NA and Humboldt Foundation postdoctoral fellowship awarded to L.-A.H. We thank Dr. M. Said Ghandour (Institut des Sciences Biologiques - INSB, Faculté de Médecine de l'Université de Strasbourg, France) for providing MBP and CAII antibodies.
References (86)
- et al.
Opto-fMRI analysis for exploring the neuronal connectivity of the hippocampal formation in rats
Neurosci. Res.
(2012) - et al.
Super-resolution track-density imaging studies of mouse brain: comparison to histology
NeuroImage Neuroergonomics: Hum. brain action work
(2012) - et al.
Graph theory findings in the pathophysiology of temporal lobe epilepsy
Clin. Neurophysiol.
(2014) - et al.
Great expectations: using whole-brain computational connectomics for understanding neuropsychiatric disorders
Neuron
(2014) - et al.
Quantitative evaluation of 10 tractography algorithms on a realistic diffusion MR phantom
NeuroImage
(2011) - et al.
Diffusion kurtosis imaging probes cortical alterations and white matter pathology following cuprizone induced demyelination and spontaneous remyelination
NeuroImage
(2016) - et al.
Functional connectivity fMRI in mouse brain at 7 T using isoflurane
J. Neurosci. Methods
(2013) - et al.
Validating the independent components of neuroimaging time series via clustering and visualization
NeuroImage
(2004) - et al.
Epileptic seizures and hippocampal damage after cuprizone-induced demyelination in C57BL/6 mice
Exp. Neurol.
(2008) - et al.
Tracking dynamic resting-state networks at higher frequencies using MR-encephalography
NeuroImage
(2013)