Summary of subjects, DREADD used, and functional assessments used in this study

Summary of injection location, type of virus vector, titer, and injected volume used in this study

Time course of DREADD expression levels up to 150 days after injection

(A, B) Parametric coronal images of the increase (relative to baseline) in specific binding of [11C]DCZ (ΔBPND) at 30 (A) and 64 (B) days after AAV injection, overlaid on MR images. Images were obtained from a monkey (#241) that received multiple injections of AAV with different constructs (see Table 2). (C) Time course of in vivo DREADD expression levels (ΔBPND) up to 150 days after the injections, summarized from 10 regions of interest (ROIs) obtained from 7 monkeys. The value at day 0 indicates the baseline (before injection). The black curve is the best-fitted sigmoid curve, which provides a better fit (Bayesian Information Criterion, BIC = 61.1) than does the double logistic model (BIC = 62.9). Lowercase letters correspond to the DREADD-induced regions described in Table 2.

Longitudinal time-course of DREADD expression levels

(A) Parametric coronal images of DREADD expression levels (ΔBPND) overlaid on MR images obtained from monkey #237, measured at 42, 496, 978, and 1,139 days after injection. (B-C) Time course of hM4Di (B) and hM3Dq (C) expression levels normalized to the peak ΔBPND observed between 40- and 80-days post-injection (time between the dotted red vertical lines). Each colored line represents one injection site, and the lowercase letters at the ends of each line correspond to the injection sites listed in Table 2. Each dot represents one [11C]DCZ PET measurement. The dashed black line indicates the group average at six-months intervals. hM4Di expression levels remained stable for approximately 1.5 and those for hM3Dq about 1 year, after which variable patterns of decline were observed.

Duration of effective functional modulation induced by DREADDs

(A-B) Duration of effective functional modulation induced by hM4Di (A) and hM3Dq (B). Blue and red horizontal bars indicate the duration of successful chemogenetic manipulation. Black and white ticks on the horizontal bars mark the timing of agonist administration, with CNO and Compound 21 (C21) explicitly stated; all other ticks correspond to DCZ. For example, repeated DCZ administration consistently induced behavioral change in monkey #229 and increased glucose metabolism in monkey #215. The corresponding functional assessments are summarized in Table 1. The dark blue bar indicates the duration of failed chemogenetic manipulation in monkey #234 (see also Figure 4). Total numbers of agonist administrations (DREADD activation) are shown at the end of each duration bar. Red ticks indicate the timing of DCZ-PET scans and red arrowheads indicate the timing of perfusion. (C) In vivo visualization of hM4Di expression in the dorsolateral prefrontal cortex 377 days after injection (monkey #229). The coronal PET image showing specific binding of [11C]DCZ is overlaid on MR images. (D and E) Nissl-(D) and DAB-stained (E) sections corresponding to C, representing immunoreactivity against reporter protein. White and black arrowheads represent the dorsal and ventral borders of the target regions, respectively. Scale bars, 5 mm. (F) In vivo visualization of hM3Dq expression in the amygdala 980 days after the injection (monkey #215). The coronal PET image showing specific binding of [11C]DCZ is overlaid on MR images. (G and H) Nissl-(G) and DAB-stained (H) sections corresponding to F, representing immunoreactivity against reporter protein.

Disappearance of behavioral effects in an hM4Di-extinguished monkey (#234).

(A) Top, Coronal PET image showing ΔBPND of [11C]DCZ data overlaid on an MR image, obtained 44 days after viral vector injection into the right SID2. Filled and open arrowheads represent the central and ipsilateral sulci, respectively. Middle, Performance of fine grasping using a modified Brinkman board task, assessing the monkey’s ability to pick up small food pellets with its thumb and index fingers. Values indicate the change in total duration to complete the task between pre- and post-DCZ administration sessions. Ipsi (Contra) refers to performance using the hand ipsilateral (contralateral) to the hM4Di-expressing SID2. Gray lines connect performances from the same sessions using different hands. Error bars, s.e.m. *: P < 0.002, paired t-test. +: P < 0.004, paired t-test, adjusted for multiple comparisons. n = 5 sessions. Bottom, DCZ-induced change in the foot withdrawal latency in response to cold (magenta) or control (cyan) stimulation. Negative values indicate faster withdrawal latency following DCZ administration. Gray lines connect performances from the same sessions. Error bars, s.e.m. *: P < 0.001, paired t-test. +: P < 0.001, paired t-test comparing between pre- and post-DCZ administration. n = 7 sessions. (B) Same types of measurements as in A, but following the extinction of hM4Di expression. The PET image was obtained about three years (1,051 days) post-vector injection. n = 5 and n = 4 sessions for Brinkman board task and foot withdrawal test, respectively. (C) A Nissl-stained section demonstrating the absence of neuronal loss at the vector injection sites (left panel) and the contralateral side (right panel). The locations of the filled and open arrowheads correspond to those shown in A and B. Scale bars are 2.5 mm and 250 μm.

Effect of protein tags on peak DREADD expression levels.

Peak expression levels (ΔBPND) for different constructs showing the presence and type of protein tag. Box plots represent the median (central line), interquartile range (box), and the range (whiskers) of the data. Individual data points are labeled with lowercase letters and correspond to the injection sites listed in Table 2. Note that two cases (right putamen of #223 and left putamen of #255) were identified as outliner and excluded from the analysis (see Methods).

Result of the linear model analysis

Effect of promoter and DREADD type on peak expression levels.

Peak expression levels (ΔBPND) are shown for different constructs combining two promoter types and two DREADD types. Box plots represent the median (central line), interquartile range (box), and the range (whiskers) of the data. Individual data points labeled with lowercase letters correspond to the injection sites listed in Table 2.