Statistical Analyses
Cenci and Lundblat (2007) indicated that AIM scores should be treated parametrically because the basic scores for LID reflect the proportion of time during which a given behavioral manifestation is present. Also, it has been shown that AIM scores are linearly and closely correlated with parametric quantities (Cenci and Lundblat, 2007).
All statistical analyses were conducted using GraphPad Prism software 6. A p-value of < .05 was considered statistically significant.
The acute administration of doxycycline or COL-3 and vehicle on L-DOPA-induced AIMs scores, and the open field and forepaw adjusting step tests were compared by one-way ANOVA followed by Bonferroni’s test. Data are presented as the mean±SEM with points representing data from individual animals.
A comparison of the chronic administration of doxycycline and vehicle combined with L-DOPA was performed by two-way RM-ANOVA followed by Bonferroni’s test. To analyze the immunohistochemical and fluorescence intensity results, a one‐ or two‐way ANOVA was performed to determine the difference between factors L-DOPA (6-OHDA+vehicle+L-DOPA and doxycycline (6-OHDA+doxycycline+L-DOPA) followed by Bonferroni’s test to determine the statistical difference. For the western blotting analysis, the Students t-test was used to compare the expression of markers in the 6-OHDA+vehicle+L-DOPA vs. 6-OHDA+doxycycline+L-DOPA treated animals. Correlations were estimated using simple regression analyses. Figure legends indicate the test used for each experiment. The data are presented as the mean ± SEM.