Locomotor analysis identifies early compensatory changes during disease progression and subgroup classification in a mouse model of amyotrophic lateral sclerosis

Abstract

Amyotrophic lateral sclerosis is a motoneuron degenerative disease that is challenging to diagnose and presents with considerable variability in survival. Early identification and enhanced understanding of symptomatic patterns could aid in diagnosis and provide an avenue for monitoring disease progression. Use of the mSOD1G93A mouse model provides control of the confounding environmental factors and genetic heterogeneity seen in amyotrophic lateral sclerosis patients, while investigating underlying disease-induced changes. In the present study, we performed a longitudinal behavioral assessment paradigm and identified an early hindlimb symptom, resembling the common gait abnormality foot drop, along with an accompanying forelimb compensatory mechanism in the mSOD1G93A mouse. Following these initial changes, mSOD1 mice displayed a temporary hindlimb compensatory mechanism resembling an exaggerated steppage gait. As the disease progressed, these compensatory mechanisms were not sufficient to sustain fundamental locomotor parameters and more severe deficits appeared. We next applied these initial findings to investigate the inherent variability in B6SJL mSOD1G93A survival. We identified four behavioral variables that, when combined in a cluster analysis, identified two subpopulations with different disease progression rates: a fast progression group and a slow progression group. This behavioral assessment paradigm, with its analytical approaches, provides a method for monitoring disease progression and detecting mSOD1 subgroups with different disease severities. This affords researchers an opportunity to search for genetic modifiers or other factors that likely enhance or slow disease progression. Such factors are possible therapeutic targets with the potential to slow disease progression and provide insight into the underlying pathology and disease mechanisms.

Keywords: SOD1 mouse; amyotrophic lateral sclerosis; disease progression; disease variability; locomotor; motoneuron degenerative disease; nerve regeneration; neural regeneration.

Figure 1

Depiction of the rodent step cycle. The rodent step cycle (s) of an individual paw/limb consists of a swing phase (s) and a stand phase (s). Right hind step cycle includes the swing phase, where the paw is elevated (A) and the stand phase (B, C), where the paw is placed on the ground. The right front step cycle is identical to the hind step cycle, with respect to both the stand phase (A, B) and swing phase (C). Adapted from Scholle et al. (2010).

Figure 2

mSOD1^G93A hindlimb foot drop and forelimb compensatory mechanism. Initial biomechanical changes in mSOD1 mice (n = 24), compared to WT mice (n = 16), began at 61 days of age and resemble hindlimb foot drop and a forward-shift in the center-of-gravity, likely a forelimb compensatory mechanism. (A) Increased hind paw print length in mSOD1 vs. WT, was identified as early as 61 days of age and persisted until 129 days of age. (B) Front swing phase duration was decreased in mSOD1 mice between 61–147 days of age and was likely due to an increase in mSOD1 front swing speed (C) from 61–147 days of age. (D) Front duty cycle was also increased in mSOD1 mice between 61–91 and 101–149 days of age. (E) Increased front paw print length was detected in mSOD1 mice at 61–149 days of age, compared to WT mice. (F) Max intensity at hind was decreased in mSOD1 mice between 65–149 days of age, relative to WT mice, while max contact at front (G) was reduced in mSOD1 mice between 61–149 days of age, compared to WT mice. Data was generated using the CatWalk and expressed as mean ± SEM; *Significant differences between WT and mSOD1 groups were determined using Linear Mixed Model with Repeated Measures (see methods; p ≤ 0.05). Data after 150 days of age not shown. WT: Wild-type; mSOD1^G93A: superoxide dismutase-1, glycine 93 to alanine mutation.

Figure 3

Temporary hindlimb steppage gait in mSOD1^G93A mice. mSOD1 mice display temporary, hindlimb compensatory changes from 69–85 days of age. Starting at 51 days of age, mSOD1 (n = 24) and WT mice (n = 16) were assessed using the CatWalk. (A) Hind swing phase duration was temporarily decreased in mSOD1 vs. WT mice from 69–87 days of age. (B) Hind swing speed was temporarily increased in mSOD1 mice between 69–85 days of age, then decreased compared to WT mice from 115–149 days of age. (C) Hind stand phase duration was temporarily increased in mSOD1 mice (69–85 days of age), then displayed a decrease in mSOD1 vs. WT from 117–149 days of age. (D) Hind duty cycle was temporarily increased in mSOD1 mice (69–85 days of age), followed by a decrease in mSOD1 vs. WT from 127–149 days of age. Data expressed as mean ± SEM; *Significant differences between WT and mSOD1 groups determined using Linear Mixed Model with Repeated Measures (see methods; p ≤ 0.05). Data after 150 days of age not shown. WT: Wild-type; mSOD1^G93A: superoxide dismutase-1, glycine 93 to alanine mutation.

Figure 4

Fundamental locomotor parameters are impaired in mSOD1^G93A mice as the earlier hindlimb compensatory changes fade. At 83 days of age, mSOD1 mice (n = 24) display an increase in cadence (mean steps/s of all paws; A) and a decrease in hind stride length (B) compared to WT mice (n = 16). (C) The most commonly used coupling pattern, right hind (RH)→left front (LF), was significantly reduced, as a percentage during the step cycle, in mSOD1 mice from 85–149 days of age. Locomotor parameters were measured using the CatWalk. Data expressed as mean ± SEM; *Significant differences between WT and mSOD1 groups determined using Linear Mixed Model with Repeated Measures (see methods; p ≤ 0.05). Data after 150 days of age not shown. WT: Wild-type; mSOD1^G93A: superoxide dismutase-1, glycine 93 to alanine mutation.

Figure 5

Gross motor impairments and muscular deficits apparent in mSOD1^G93A mice between 88–110 days of age. (A) Motor score in mSOD1 mice (n = 24) decreased compared to WT mice (n = 16) as early as 88 days of age. (B) Total distance traveled in open field was reduced in mSOD1 mice from 92–150 days of age, compared to WT mice. (C) Grip Strength of the front limbs, as measured by the bar attachment, was assessed starting at 67–158 days of age. A decrease in Grip Strength of mSOD1 mice (n = 23), relative to WT mice, was observed from 95–158 days of age. (D) Max speed achieved in open field was reduced in mSOD1 mice from 102–150 days of age. (E) Relative placement of the right paws (print positions right) was significantly altered in mSOD1 mice compared to WT from 103–149 days of age. (F) Mean body weight of mSOD1 mice was statistically and consistently lower than WT weight starting as early as 110 days of age. Only every other data point is shown for body weight. Data expressed as mean ± SEM; *Significant differences between WT and mSOD1 groups determined using Linear Mixed Model with Repeated Measures (see methods; p ≤ 0.05). Data after 150 days of age not shown in panels A, B, D–F. Parameters measured using the CatWalk (E), ANY-Maze (B, D), and Grip Strength-bar test (C). WT: Wild-type; mSOD1^G93A: superoxide dismutase-1, glycine 93 to alanine mutation.

Figure 6

Survival analysis of mSOD1^G93A subgroups. K-Means for Joint Longitudinal Data (KML3D) cluster analysis was performed by combining the trajectories of four locomotor parameters (motor score, body weight, print positions right, and max speed) to identify two mSOD1 subgroups with significantly different disease progression rates. Survival analysis reveals a significant difference in the lifespan of the fast disease progression group (FPG; red; n = 9) and the slow disease progression group (SPG; blue; n = 15). Logrank (Mantel-Cox) test. mSOD1^G93A: superoxide dismutase-1, glycine 93 to alanine mutation.

Figure 7

Fundamental locomotor parameters revealed significant differences in disease progression between mSOD1^G93A subgroups: the fast progression group (FPG) and the slow progression group (SPG). Locomotor parameters were measured using the CatWalk. A total of 34 testing days were performed before all FPG mice were retired from CatWalk, at 121 days of age. (A) Front stride length was significantly decreased in the FPG (n = 9) vs. SPG (n = 15) across 29 testing days, starting at 55 days of age. Hind stride length (B) was also reduced in the FPG vs. SPG across 19 testing days, beginning at 55 days of age. Both front step cycle (C) and hind step cycle (D) were also decreased in the FPG, starting at 57 days of age, compared to the SPG. Over the 34 CatWalk testing days, front step cycle significantly differed 14 days, while hind step cycle differed eight days. (E) Front swing phase duration was significantly reduced in the FPG vs. SPG staring at 51 days of age, for a total of 22 days. (F) Cadence (mean steps/s) was increased in the FPG vs. SPG, for a total of 14 testing days, beginning at 53 days of age. (G) Hind single stance phase duration revealed a delayed difference in between FPG vs. SPG, beginning at 97 days of age (SPG slope = –0.000416; FPG slope = –0.000757). (H) Similarly, hind stand phase duration also revealed a significant difference between FPG (slope = –0.00117) and SPG (slope = –0.000430) starting at 97 days of age. Data expressed as mean ± SEM. *Significant differences between WT and mSOD1 groups determined using Linear Mixed Model with Repeated Measures (see methods; p ≤ 0.05). Data after 150 days of age not shown. WT data is shown as a single regression line for comparison purposes only (models: cubic for A; quadratic for B–H). WT: Wild-type; mSOD1^G93A: superoxide dismutase-1, glycine 93 to alanine mutation.