Experimental design and data analysis issues contribute to inconsistent results of C-bouton changes in amyotrophic lateral sclerosis

Abstract

The possible presence of pathological changes in cholinergic synaptic inputs (C-boutons) is a contentious topic within the ALS field. Conflicting data reported on this issue makes it difficult to assess the role(s) of these synaptic inputs in ALS. Our objective was to determine whether the reported changes are truly statistically and biologically significant and why replication is problematic. This is an urgent question, as C-boutons are an important regulator of spinal motoneuron excitability, and pathological changes in motoneuron excitability are present throughout disease progression. Using male mice of the SOD1-G93A high-expresser transgenic (G93A) mouse model of ALS, we examined C-boutons on spinal motoneurons. We performed histological analysis at high statistical power, which showed no difference in C-bouton size in G93A vs. wild-type motoneurons throughout disease progression. In an attempt to examine the underlying reasons for our failure to replicate reported changes, we performed further histological analyses using several variations on experimental design and data analysis that were reported in the ALS literature. This analysis showed that factors related to experimental design, such as grouping unit, sampling strategy, and blinding status, potentially contribute to the discrepancy in published data on C-bouton size changes. Next, we systematically analyzed the impact of study design variability and potential bias on reported results from experimental and pre-clinical studies of ALS. Strikingly, we found that practices such as blinding and power analysis are not systematically reported in the ALS field. Protocols to standardize experimental design and minimize bias are thus critical to advancing the ALS field.

Significance Statement: C-boutons are cholinergic synaptic inputs that have been implicated as an important regulator of motoneuron excitability. Pathological changes in the size of C-boutons have been previously characterized in disease models of ALS, with conflicting results reported. This issue is critical, as the dysregulation of motoneuron excitability is present throughout disease progression. We show here that reported changes in C-bouton size, number, and density – thought to be either disease-related mechanisms or compensatory processes to maintain cell excitability – cannot be replicated. Furthermore, we provide evidence suggesting possible reasons for failure of replication due to variability in experimental design and analysis practices. Additionally, we show such variability is widespread in other animal studies of ALS, and propose practices for improved consistency.