Functional decline

As expected, the present study confirmed the decline in functional ability over time in SOD1 mice (Gurney et al. 1994). SOD1 mice showed lower body weights than wild- type mice from 14 weeks of age (Figure 2.10), deficits in wire hang duration from 15 weeks of age (Figure 2.12), observable neurological deficits from 16 weeks of age (Figure 2.13), and altered stride pattern from 18 weeks of age (Figure 2.11). Thus, disease onset was detected around 14 to 15 weeks of age in this study. The timing of disease onset is considered a conservative estimate compared to other studies using the SOD1 mouse on a congenic B6 background, which found the earliest appearance of disease symptoms at around 11 weeks of age (Beers et al. 2011a; Beers et al. 2011b). 2.4.5.1 The use of mixed modelling to assess body weight peak

In this study a slightly novel approach to measuring disease onset was used – mixed modelling was used to generate a curve to model the body weight trajectory of SOD1 mice, and the age at which maximum body weight was reached was estimated using calculus. This method was compared with the calculation of weekly averages. Both the average body weight, and the trajectory estimated with mixed modelling, showed differences between SOD1 and WT mice from 14 weeks of age (Figure 2.10). The model-derived estimated maximum body weight, and the model-derived age at which the maximum body weight occurs, were similar to those calculated from the raw data. The mixed model has both advantages and disadvantages over weekly body weight averages. The advantages are that: the mixed model allows all recorded body weight measurements, even taken on a daily basis, to be used instead of creating a weekly average for each mouse; and that the mixed model is able to assign some of the variance in the data to variations over time for individual mice, reducing the residual variance and increasing statistical power to detect differences between treatment or genotype

______________________________________________________________________ 89 groups (Murnane & Willett 2011). The disadvantage of a mixed model is that if the model is not fitted correctly, the estimates generated by the model will not be an accurate reflection of the true data. However, the body weight trajectories of SOD1 mice appear to be well modelled by a cubic function (Figure 2.10), with a cubic function showing superior model fit over a quadratic function (Supplementary Data 1). Thus, mixed modelling may be a novel method of increasing statistical power in studies which examine body weight measurements in the same mice over time.

2.4.5.2 Comparison of tests for functional deficits

Functional decline in SOD1 mice was evident in wire hang duration times prior to becoming evident in stride pattern (Figure 2.11, Figure 2.12). This earlier detection using the wire hang test may be due to the nature of each task – maintaining grip on suspended wire bars requires constant muscle activation, whereas walking along a strip of paper likely requires less muscle strength, and has the advantage that each limb gets a transient rest period while it is in the stationary phase of walking.

The wire hang duration, or paw grip endurance, test has been used previously to assess functional ability and muscle strength in SOD1 mice, with deficits in wire hang duration ability found as early as 12 weeks of age (Weydt et al. 2003). In the present study, wire hang duration was tested to a maximum of 60 seconds, compared with a maximum of 90 seconds in previous studies (Weydt et al. 2003). The use of a longer test time may pick up earlier changes in muscle strength, as increasing the hang time requires an increased amount of strength and endurance.

Computerised treadmill analysis of SOD1 gait pattern indicates that changes between SOD1 and WT mice occur as early as 8 weeks of age (Wooley et al. 2005); however, in the current study differences between SOD1 and WT mice only became apparent from 18 weeks of age onwards (Figure 2.11). The late change in stride lengths observed here are consistent with those observed in other studies using non-treadmill-based footprint analysis (Gurney et al. 1994; Knippenberg et al. 2010). These data indicate that measurements of stance and stride times during treadmill running (Wooley et al. 2005) provide a much more sensitive measure of dysfunction than measurement of stride length on paper. However, the late decline in stride length may correspond to the rapidly-progressing phase of disease, where motor units are lost without compensatory

______________________________________________________________________ 90 re-innervation in the muscle; stride length may be useful in monitoring disease progression during this phase.

In this study we report increasing uniformity measure in SOD1 mice; that is, reduced ability of hindlimb musculature to move the leg in a forward motion during stepping, such that the hindlimb placement falls short of front limb placement on the previous step (Figure 2.11). To our knowledge, this measure has not been previously reported in SOD1 mice and, like the late change in stride length, may be an appropriate measure of measuring the effect of therapeutic interventions on the late phase of disease.

2.4.5.3 Neurological scoring system

The neurological scoring system employed in this study, designed by the ALSTDI, was used to assess neurological deficits present in SOD1 mice through observation. The ALSTDI scoring system allows only five possible neurological scores: no symptoms (NS=0), mild symptoms at disease onset (NS=1), dragging of feet due to loss of muscle strength (NS=2), rigid paralysis or minimal movement (NS=3), and inability to right (NS=4) (Scott et al. 2008). SOD1 mice showed a progressive increase in the neurological score rating throughout disease progression (Figure 2.13); however, of the four neurological scores observed in this study (NS=0-3), there was a substantial time gap between a rating of NS=1 and a rating of NS=2 (see text, Chapter 2.3.2.5). During the period when a given mouse was rated NS=1, alterations in gait could be observed which did not meet the criteria for a rating of NS=2. Additionally, some WT mice were also rated as NS=1 – possibly due to mis-interpretation of the neurological score criteria to include any inwards collapse of the legs upon tail suspension, as sometimes occurred in WT mice – but no WT mice showed gait abnormalities as seen in SOD1 mice. A neurological scoring system with more specific detail around gait abnormalities, such as the Beers/Appel/Simpson/Henkel (BASH) scoring system (Beers et al. 2011b), may be more appropriate for detailed comparison of neurological symptoms present after disease onset in SOD1 mice.

Although the NS=2 rating was reached rather late in disease progression, a rating of NS=2 with dragging feet, or toes curling under while walking, was easily identified and may be a useful tool for measuring the age at which mice develop overt difficulties maintaining normal hindlimb function.

______________________________________________________________________ 91 2.4.5.4 Functional measures of disease progression for evaluating therapeutics Peak body weight, and the inability to maintain wire hang duration for 60 seconds, may be used to evaluate disease onset in future cohorts of SOD1 mice. Changes in stride length, uniformity measure, wire hang duration, body weight, and neurological score rating (foot dragging), may be useful for monitoring disease progression over time in SOD1 mice.

2.4.6 Summary and conclusions