Figure 1
Dosing for tremor responses among subjects. A. Acquisition of wrist-mounted accelerometry. Left: Subjects had devices for measuring triaxial accelerometry mounted securely to the wrist using vet wrap. Right: Example data recorded from one axis of an accelerometer during Subject 1’s 10 mg/kg harmaline exposure session (time series are shown on the same amplitude scale, and have been bandpass-filtered between 1 and 30 Hz). B. Dosage of harmaline delivered intramuscularly (mg/kg) to a subject during a given harmaline dosing session, arranged in columns according to the time elapsed (days) since the subject’s first harmaline administration of the study (“day 1”).
Figure 2
Self-paced cued reaching task performance under harmaline dosing conditions in S3. A. Representation of the subject’s engagement with the startpad and touchscreen elements of the self-paced cued reaching task. The illustrated colored limb outlines depict the position of the upper limb according to the corresponding stage of each reach-trial. In Standby mode, the subject was not engaged with the task, and the monitor displayed a solid-black screen. Once the task was initiated and the subject was in the hold-phase, the monitor displayed a solid-blue screen until the variable hold period was completed. Finally, the circular target appeared in the center of the screen and the subject could reach and touch the target for a liquid reward. B. Overview of the structure and time course of the self-paced cued reaching task study, consisting of one experimental session per day, 5 consecutive days per week, for 8 weeks. Task duration and preceding washin waiting time after injection is depicted, as well as the correspondence between each week in the study and the harmaline dose delivered during that week.
Figure 3
Spectral content of the forelimb accelerometry recordings following harmaline injection in each subject, at tremorgenic doses. These data are presented in terms of (left) power spectral densities color coded and arranged in time since harmaline injection, (middle) power spectral densities with the same color coding stacked to emphasize spectral peaks, and (right) peak frequency in the 10–14 Hz range plotted over time. Solid lines in the rightmost figures indicate the linear regression fit for peak frequency over time.
Figure 4
Tremor-band motion power ratio over time. Onset and magnitude dynamics of the motion power ratio (MPR) comparing a “tremor-band” of interest with 17–21 Hz content. A: Data points represent the median normalized MPR among all MPR samples within a segment of data (up to but not exceeding 2 minutes in duration with the time value given as the midpoint). Vertical bars indicate 95% confidence intervals. All data recorded prior to the 15 minutes post harmaline mark were treated as baseline data, from which the median is calculated for normalization and statistical-testing purposes. Red data points signify statistically signficant difference above the median baseline. Arrows indicate the earliest occurring segment of tremor-band MPR data which differed significantly from median baseline. Data points are connected with a line representing a piecewise cubic hermite interpolating polynomial (PCHIP) for visualization purposes. B: The aforementioned PCHIP lines from the top figure are plotted together at the same scale to emphasize the differences in magnitude for tremor-band MPR among subjects.
Figure 5
Testing self-paced cued reaching task performance under harmaline dosing conditions. Task performance deteriorated with increasing harmaline dosages compared to the performance during drug-naive sessions, with a dose-dependent trend towards fewer attempts at initiating task trials and an increasing incidence of premature reaches. Data from all drug-naive days were pooled together and displayed with 90% confidence intervals (dotted red lines), while data from harmaline-dose days are displayed separately. The central mark within each violin plot indicates the median, the edges of the vertical gray bar indicate 25th and 75th percentiles, and the entire vertical length of violine plots indicate the full range of data. * indicates harmaline-related performance data that falls outside of the 90% confidence interval of naïve performance.