A translational approach to capture gait signatures of neurological disorders in mice and humans

Abstract

A method for capturing gait signatures in neurological conditions that allows comparison of human gait with animal models would be of great value in translational research. However, the velocity dependence of gait parameters and differences between quadruped and biped gait have made this comparison challenging. Here we present an approach that accounts for changes in velocity during walking and allows for translation across species. In mice, we represented spatial and temporal gait parameters as a function of velocity and established regression models that reproducibly capture the signatures of these relationships during walking. In experimental parkinsonism models, regression curves representing these relationships shifted from baseline, implicating changes in gait signatures, but with marked differences between models. Gait parameters in healthy human subjects followed similar strict velocity dependent relationships which were altered in Parkinson's patients in ways that resemble some but not all mouse models. This novel approach is suitable to quantify qualitative walking abnormalities related to CNS circuit dysfunction across species, identify appropriate animal models, and it provides important translational opportunities.Walking is a complex behavior that requires not only control of initiation and termination of locomotion, but also ongoing adjustments of speed, stride length, cadence, direction, and posture in response to dynamic internal and external cues. Walking can be affected in many different ways secondary to musculoskeletal or cardiovascular problems, as well as dysfunction of the peripheral or the central nervous system (CNS). Gait disorders due to specific CNS circuit pathologies can be clinically recognized based upon phenomenology, such as shuffling gait or festination in Parkinson's disease or ataxic gait in cerebellar disorders 1, 2 . The wide variety of causal conditions and subtypes of gait disorders illustrate that abnormal function or pathology in multiple CNS networks differentially affect gait.Human gait can be objectively measured using direct or indirect kinematic, spatial or temporal measures via video-analysis, footfall studies, or wearable devices. Advances in imaging technology have allowed visualization of putative supratentorial circuitries involved in gait control [3][4][5][6][7][8][9] . However, the functional-anatomical basis for the various gait abnormalities so readily recognized in clinical settings remains poorly understood as approaches to isolate relevant networks during life are relatively limited. This gap in knowledge stands in the way of the development of specific, circuit based treatment strategies targeted to the various specific gait abnormalities.Animal models offer a means to circumvent current barriers of studying gait in humans. In mice, highly selective approaches are now available to identify components of CNS circuits from cortex to spinal cord that mediate or modulate locomotion, including walking 10-13 . Effects of interventions can then be measured u...

Profiling walking dysfunction in multiple sclerosis: characterisation, classification and progression over time

“…v max50 was higher in healthy controls (1.08 ± 0.12 m/s) than in patients (0.64 ± 0.17 m/s; p < 0.0001, two-tailed, unpaired t-test; Supplementary Table S1 ). As previously reported by others 18 , 20 , 26 , we found that most kinematic gait parameters changed significantly with walking speed (Supplementary Table S1 ) demonstrating that differential walking speeds in patients and healthy controls confound the precise characterisation of walking deficits.…”

“…In the aforementioned studies, however, patients and healthy controls walked at their respective self-selected speed, resulting in significant speed differences between groups. As the vast majority of locomotor parameters is strongly influenced by gait velocity 18 , slower walking speed in patients relative to controls is a major confounder, limiting the characterisation of MS-related gait pathophysiology 9 , 19 , 20 .…”

Proposal of a new conceptual gait model for patients with Parkinson’s disease based on factor analysis

“…Many authors reported that this variable is also one of the strongest predictors of future falls both in patients with PD [1,40] and in elderly people [41]. Of note, both in animals and humans, gait parameters change as a function of speed even under "normal" conditions [42]; for this reason, it was challenging to decide which variables to retain in our model. We believe that one of the causes for the non-convergence of previous models lies in their having included a plethora of highly correlated variables.…”

A multimodal approach to identify clinically relevant parameters to monitor disease progression in a preclinical model of neuropediatric disease

“…This method is widely employed in the clinic to objectively describe gait abnormalities in patients. While there are large differences in ambulation mode between bipedal humans and quadrupedal mice, recent developments have demonstrated correlations between parameters of human and mouse gait, and the translational utility of gait analysis in mouse models ( 25 ). Similarly, in terms of noninvasive brain imaging, there are many anatomical similarities between the human and murine brain, and studies have shown that noninvasive imaging often detects similar changes in human patients and mouse models of neurodegenerative disease ( 26 ).…”

“…Cohorts of wild type and Cln6 nclf mice of both sexes were monitored longitudinally using noninvasive imaging, including T2-weighted magnetic resonance imaging (T2-MRI), diffusion tensor MRI (DTI), 1 H magnetic resonance spectroscopy (MRS), and positron emission tomography (PET) was performed periodically between 3-12 months of age while kinematic gait analysis (KGA), which captures a large number of metrics describing gait, was assayed from 6-12 months of age. Variations of all of these techniques are widely used in the clinic and have shown correlations between mouse models and human subjects ( 25, 26 ). Once we had characterized these parameters in the CLN6 disease mouse model, we used a recently developed form of principal component analysis (PCA), contrastive PCA (cPCA), to cluster the four imaging modalities and gait analysis in order to derive new variables that best capture and define the progressive nature of the disease.…”