Introduction

Chemogenetic technology affords remote and reversible control of neuronal activity by expressing receptors that are designed to be activated by systemically delivered biologically inert actuators. One such tool is the widely used Designer Receptors Exclusively Activated by Designer Drugs (DREADD) system, which is based on mutated muscarinic receptors (Roth, 2016). DREADDs have been widely used in rodent models and their use in nonhuman primates has recently shown some promise, as they could induce significant behavioral changes by simultaneously modulating activity of disparate neuronal populations across large brain regions (Eldridge et al., 2016; Nagai et al., 2016). Whether this promise will translate to reality depends on how long DREADDs can exert their effects. Because the efficacy of chemogenetic manipulation depends on DREADD expression levels (Grayson et al., 2016; Upright et al., 2018), maintaining high and stable expression levels is critical for ensuring consistent control throughout all experiments in a study. While this has proven possible in rodent studies that typically only span several months, whether it is true for studies involving macaque monkeys, which often last multiple years, remains a question. Indeed, despite the growing use of DREADDs in monkey neuroscience research, consistent and reliable data on the temporal dynamics of DREADD expression following viral transduction remain scarce. Specifically, little is known about how quickly DREADD expression is established following viral transduction or how long it is maintained. This gap in knowledge complicates efforts to optimize critical parameters, such as the serotypes, constructs, and titers of viral vectors that all influence expression level and duration.

Positron emission tomography (PET) is an in vivo imaging technique that allows quantification of receptor density in the brain. Several studies have demonstrated the utility of PET for longitudinal measurements of gene expression introduced into the brain via adeno-associated virus (AAV) (Nerella et al., 2023; Vandeputte et al., 2011; Yoon et al., 2014). While these studies have highlighted the potential of using PET imaging to monitor gene expression over time, most have used it as an indirect measure of target gene expression, without exploring the relationship between expression levels and the resultant functional outcomes.

Several PET probes have been developed to visualize DREADD receptor expression. Initial studies in rodents used [¹¹C]clozapine-N-oxide and [¹¹C]clozapine to monitor DREADD expression in the brain (Ji et al., 2016). In nonhuman primates, longitudinal PET studies using [¹¹C]clozapine have successfully tracked virally expressed DREADDs in macaques (Nagai et al., 2016). More recently, DREADD-selective ligands such as [¹⁸F]JHU37107 (Bonaventura et al., 2019) and [¹¹C]deschloroclozapine ([¹¹C]DCZ)(Nagai et al., 2020) have been developed. Among these, [¹¹C]DCZ has become a particularly useful tracer for studying DREADD-mediated functional manipulation in macaques and marmosets. This PET ligand allows researchers to verify the location and density of DREADD expression at the site of AAV injection and at projection sites, as demonstrated in numerous studies (Hirabayashi et al., 2024, 2021; Hori et al., 2021; Mimura et al., 2021; Miyakawa et al., 2023; Miyake et al., 2024; Mueller et al., 2023; Nagai et al., 2020, 2016; Oyama et al., 2024, 2022a, 2022b, 2021; Roseboom et al., 2021). Moreover, unlike post-mortem analysis that relies on the detection of fused or expressed protein tags, [¹¹C]DCZ PET provides a direct, quantitative, and non-invasive measure of DREADD expression level in vivo, using standard reference tissue models (Nagai et al., 2020; Yan et al., 2021). [¹¹C]DCZ PET thus serves as a valuable tool for tracking DREADD expression levels throughout long-term chemogenetics experiments in monkeys.

As the developer of the DREADD-selective agonist DCZ (Nagai et al., 2020), we have consistently performed multiple [¹¹C]DCZ PET scans as part of our DREADD experiments in monkeys, including at baseline before AAV injection, 30–120 days post-injection, and then periodically over the course of the experiments. For the current study, we analyzed these PET data to examine the short- and long-term dynamics of DREADD expression in vivo and assessed how these changes correlated with the ability to manipulate neuronal activity or behavior. Our results provide valuable information with a few caveats and indicate that the promise of using DREADDs in long-term monkey experiments is real.

Results

PET imaging data using [¹¹C]DCZ were collected from 15 monkeys before (baseline) and after injections of AAV vectors for expressing hM4Di or hM3Dq (see Table 1). To achieve neuron-specific expression, we used AAV vectors with a preference for neuronal infection, such as AAV2 or AAV2.1 (Kimura et al., 2023), or for incorporation with neuron-specific promoters (e.g., human synapsin promoter; hSyn). To assess the specific binding of [¹¹C]DCZ, we calculated the binding potential relative to a non-displaceable radioligand (BP_ND) using a multilinear reference tissue model with the cerebellum as a reference region (Yan et al., 2021). DREADD expression levels were then quantified as the change in BP_ND (ΔBP_ND) from baseline values. This approach allowed us to longitudinally and quantitatively monitor the dynamics of DREADD expression in vivo.

Table 1.

Peak DREADD expression occurred approximately 60 days post-injection

To evaluate the temporal profile of DREADD expression and determine how quickly expression was established following transduction, we analyzed the [¹¹C]DCZ PET data from seven monkeys that received AAV vector injection into the subcortical regions of their brains (Tables 1 and 2). The data were selected according to the following criteria: (1) the titer of injected viral vector was 1.0– 3.0×10¹³ gc/ml, (2) peak ΔBP_ND value was increased by more than 0.5, and (3) PET scans were repeatedly performed until 150 days post-injection. Analysis showed that expression levels of hM4Di and hM3Dq increased rapidly, reaching peak expression at approximately 60 days after viral vector injection. Notably, no clear differences in expression dynamics were observed between hM3Dq and hM4Di (Figure 1).

Figure 1

Table 2.

Longitudinal assessment of DREADD expression levels

We then conducted a longitudinal assessment of post-peak expression levels for hM4Di and hM3Dq using repeated [¹¹C]DCZ PET measurements from 16 injection sites, including cortical and subcortical regions, across 11 monkeys. Figure 2 illustrates the temporal changes in normalized ΔBP_ND, expressed as a percentage of the peak value observed around 60 days post-AAV injection. hM4Di expression levels remained stable at peak levels for approximately 1.5 years before gradually declining after 2 years (Figure 2B). Compared with hM4Di, hM3Dq expression exhibited greater post-peak fluctuations, but nevertheless remained at ∼75% of peak levels after about 1 year and declined to ∼50% by 2 years post-injection (Figure 2C). This post-peak fluctuation was not explained by the cumulative number of times receptors were activated with DCZ (repeated measures two-way ANOVA, main effect of activation times, F_(1,6)=5.745, p=0.054).

Figure 2

DREADDs effectively modulated neuronal activity and behavior for approximately two years

We assessed data from 17 monkeys to evaluate the duration over which DREADD activation produced significant effects on neural activity and behavior. Monkeys expressing hM4Di were assessed through behavioral testing (N = 11) and alterations in neuronal activity using electrophysiology (N = 2). Monkeys expressing hM3Dq were assessed for changes in glucose metabolism via [¹⁸F]FDG-PET (N = 5) or alterations in neuronal activity using electrophysiology (N = 1). Across these assessments, significant chemogenetic effects were observed for up to three years following AAV vector injection (Figure 3). For example, one monkey with bilateral hM4Di expression in the dorsolateral prefrontal cortex (dlPFC; monkey #229) consistently displayed impaired performance on the delayed response task with DCZ administration for up to 2.3 years (867 days) following vector injection (Oyama et al., 2021). Similarly, another monkey with hM3Dq expression in the amygdala (monkey #215) exhibited increased glucose metabolism following DCZ administration for up to 2.5 years (926 days) following vector injection, as measured by FDG-PET (Nagai et al., 2020).

Although experiment durations varied across monkeys, we consistently observed significant effects throughout the study periods, with only one monkey (#234) being removed from experiments due to a loss of DREADD effectiveness three years after transduction. In all other cases, receptor activation was performed repeatedly several tens of times, resulting in no clear loss of efficacy. After the termination of the chemogenetic studies, DREADD expression was verified by in vitro via postmortem immunohistochemical analysis (Fig. 3C-H). These findings highlight the robust and sustained functionality of DREADD systems in long-term experiments.

Figure 3

As mentioned above, after more than three years, we observed the disappearance of hM4Di expression in one monkey (#234), and this was associated with the loss of chemogenetic behavioral effects. Specifically, activating hM4Di in the functionally defined representation of index finger primary somatosensory cortex (SI_D2) resulted in reversible behavioral deficits for two years, including impaired finger dexterity and hypersensitivity in the foot, confined to the contralateral side (Figure 4A)(Hirabayashi et al., 2021). However, in vivo imaging more than three years after injection indicated diminished hM4Di expression, and subsequent behavioral assessments failed to detect any motor deficits or hypersensitivity (Figure 4B). Post-mortem immunohistochemical examination confirmed the absence hM4Di, with no evidence of neuronal loss at the injection site (Figure 4C).

Figure 4

Factors affecting DREADD expression levels

Finally, we investigated the factors influencing the level of DREADD expression. Despite examining expression across various cortical and subcortical brain regions, including the caudate nucleus, putamen, amygdala, thalamus, orbitofrontal cortex, dorsolateral prefrontal cortex, and SI, we found that peak expression level did not clearly depend on any specific brain area (one-way ANOVA, effect of brain region, P = 0.154). However, comparing peak expression among different viral vector constructs suggested several factors that might have influenced peak expression levels.

To identify these factors, we applied a linear model incorporating multiple variables, including serotype, promoter, titer, tag, and DREADD type. Our analysis revealed that virus titer, along with its interaction with DREADD type, significantly influenced expression levels (Table 3; Figure 5A). In addition, including co-expressed protein tags led to reduced peak expression levels, with the vector encoding the fusion HA-tag sequence at the 5’-terminal site (5’-HA) of the DREADD sequence exerting a stronger suppressive effect compared with the vector with GFP encoding following the internal ribosome entry site (IRES) sequence (Figure 5B).

Figure 5

Table 3.

Discussion

DREADDs have emerged as invaluable tools in nonhuman primate research due to their ability to exert effects on non-spatially restricted brain regions, allowing for simultaneous and discrete targeting of multiple areas. This capability makes DREADDs promising for investigating the causal roles of specific neural pathways and cell types in the highly specialized brain circuits found in primates. However, ethical and logistical constraints on using a large number of monkeys highlight the need of moving beyond a trial-and-error approach. It is essential to aggregate data and provide insights that are more widely applicable, thus enabling the development of a more efficient experimental design for DREADD research in monkeys. In vivo PET visualization of DREADDs offers a non-invasive and quantitative means for real-time monitoring of receptor expression. The present study analyzed PET and functional/behavioral data from 43 injection sites across 20 monkeys and revealed three key findings: (1) after AAV injection, DREADD expression levels peaked after approximately 60 days and remained stable for up to 1.5 years, followed by a gradual decline observed after two years; (2) significant chemogenetic effects on neural activity and behavior persisted for around two years; and (3) virus titer and protein tags significantly influenced peak expression levels, with co-expressed protein tags reducing peak expression. These results offer quantitative insights into the temporal dynamics of DREADD expression, thereby informing strategies for designing long-term experiments. They also confirm the efficacy and reliability of DREADD-mediated interventions.

Technical considerations

In vivo monitoring of gene expression has been conducted using fluorescent reporters observed through a window or fiber, providing a method to verify transducing protein expression and localization (Diester et al., 2011; Ruiz et al., 2013). Although useful for optogenetics or imaging studies, these methods are invasive, less quantitative, and spatially limited, often requiring cranial surgeries and capturing only two-dimensional data. In contrast, DCZ PET imaging provides a direct, non-invasive, volumetric measure of receptor expression, making it particularly suitable for studies in large animals such as macaques and for translational research with potential application to future human therapies.

Nevertheless, PET measurement has certain limitations, including potential variability due to individual differences, underestimation of DREADD expression levels due to the partial volume effect, and the influence of anesthesia on kinetic parameters during imaging sessions. In this study, we have considered these factors and minimized their impact by applying appropriate ROI placements and excluding data collected under inadequate anesthetic conditions (see Methods).

Temporal dynamics of DREADD expression

Our [¹¹C]DCZ PET measurement showed that DREADD expression peaked around 60 days after AAV injection, regardless of whether hM4Di or hM3Dq was used. This pattern was consistent with our previous data of hM4Di expression visualized using [¹¹C]clozapine (Nagai et al., 2016). This gradual rise in AAV-mediated gene expression likely reflects rate-limiting steps following viral infection, including transport to the nucleus, uncoating, and conversion to double-strand DNA (Duan et al., 2000; Ferrari et al., 1996; Li and Samulski, 2020; Wang et al., 2007). Similar temporal patterns of AAV-mediated gene expression have been observed in previous studies using in vivo fluorescent measurement. For example, in rats transfected with an AAV5 encoding fluorescence protein (EYFP) in motor cortex and the hippocampus, in vivo fluorescence signals increased sigmoidally, rapidly until day 35, then slowed down (Diester et al., 2011). Similarly, in macaques, ChR2 expression monitored via fluorescence intensity peaked at around 60 days after AAV9 injection (Nakamichi et al., 2019).

Although our findings are based on a limited dataset targeting neuronal DREADD expression, they offer important insights for designing chemogenetics studies in primates. Specifically, beginning functional experiments at approximately 60 days post-AAV injection will yield reliable results, avoiding unnecessary delays. This timeline may also be applicable to other genetic studies in monkeys, such as those using optogenetics or calcium imaging, which also require proper expression of functional proteins. However, while these tools rely on rapid, light driven responses and are relatively less dependent on stable functional protein expression, DREADDs require sustained expression for consistent long-term effects.

Factors influencing DREADD expression

Our analysis indicated that both virus titer and the presence of protein tags significantly influenced DREADD expression levels. In general, while higher AAV titers result in increased transgene expression in neurons, excessively high titers can trigger immune responses, necessitating caution in their use. In the current study, we used titers that ranged from 1 to 5 × 10¹³ gc/ml, achieving a balance between effective expression and minimal immune activation. Using vectors encoding DREADDs with protein tags, particularly 5’-terminal HA tags, was associated with a reduction of peak expression levels. Thus, although these tags are commonly used for antibody detection in DREADD constructs, they negatively impact expression efficacy and potentially affect protein function. Our findings are consistent with a previous PET study using [¹¹C]clozapine and immunohistochemical analysis, which also found reduced hM4Di expression in constructs with 5’-terminal HA tags (Nagai et al., 2016). Additionally, constructs incorporating IRES-GFP might also result in lower expression levels due to the size and composition of the AAV vector, as has been shown in previous studies (Furler et al., 2001; Kimura et al., 2023; Mizuguchi et al., 2000; Zhou et al., 1998). Nevertheless, these constructs have been frequently used in rodent research and have also been effective in producing long-term chemogenetic effects in monkey studies, including the current one (e.g., monkeys #229 and #238). Thus, while the observed decline in DREADD expression resulting from these constructs is not inherently detrimental to experimental outcomes in monkey studies, it serves as a warning that we must be careful when designing vectors to avoid reductions in experimental efficacy.

Implications for long-term experiments with DREADD expression

One of the most important findings was that AAV-mediated DREADD expression was maintained for up to 2 years for hM4Di and 1 year for hM3Dq. Throughout the duration of the experiments, chemogenetic effects remained consistent, as verified by individual behavioral and functional assessments. However, hM4Di expression diminished after 3 years, coinciding with the loss of chemogenetic behavioral effects in at least one case. This indicates that while chemogenetic effects can be maintained for up to two years, they likely decline gradually beyond this period. Consequently, researchers should anticipate consistent chemogenetic effects for a maximum of two years following AAV injection, and should be mindful that longer-term experiments risk a decline in effectiveness. Specifically, the inevitable failure to detect DREADDs via immunohistochemistry highlights an important caveat for researchers: timely termination of experiments to ensure verification through in vitro examination is necessary.

The gradual decline in expression levels observed beyond two years has several potential causes. First, AAV vectors typically remain episomal in transduced cells, without integrating into the host genome (Weitzman and Linden, 2011). While this poses less of an issue in non-dividing neurons, episomal DNA can still be subject to gradual loss or degradation over time. Second, several mechanisms may contribute the gradual downregulation of transgene expression, including epigenetic modifications, immune responses, and genome stability (Muhuri et al., 2022). For example promoter silencing—often associated with epigenetic modifications—has been reported (Gray et al., 2011). Neuron-specific promoters have been shown to achieve more efficient and prolonged transgene expression than ubiquitous promoters like CMV (Paterna et al., 2000).

Our study found no evidence of cell death at AAV injection sites (Figure 4C), suggesting that neuronal loss was not a primary cause of the eventual decline in expression levels. Repeated receptor activation did not affect DREADD expression levels or chemogenetic efficacy in a previous study (Oyama et al., 2022). Thus, understanding the factors that contribute to the decline in expression remains of critical importance for developing optimal strategies to maintain persistent transgene expression in neurons, both for chemogenetic studies and therapeutic applications. Despite this uncertainty, our data offers important guidelines for chemogenetics studies in nonhuman primates, demonstrating that up to two years of reliable expression is achievable.

Translational implications

This research has significant translational implication, particularly for the development of DREADD-based therapies for treating neurological and psychiatric disorders. DREADD-based approaches have shown promising therapeutic potential, with successful application in monkey models of disease conditions such as epileptiform seizures (Miyakawa et al., 2023) and Parkinson’s disease (Chen et al., 2023). Extending these findings to human clinical applications positions PET imaging as a powerful tool for real-time, non-invasive confirmation of target protein expression—a critical factor for ensuring therapeutic efficacy. PET reporter imaging has already shown its value in human gene therapy for certain movement disorders, where sustained protein expression has been correlated with treatment efficacy (Mittermeyer et al., 2012; Tai et al., 2022). Our study reinforces this utility by demonstrating that when DREADD expression was no longer detectable via PET, the chemogenetic effects were also diminished.

Conclusion

Our study describes short- and long-term dynamics of DREADD expression in nonhuman primates using [¹¹C]DCZ PET imaging, complemented by assessment of functional efficacy. The findings provide critical insights into the practicality of multi-year studies using chemogenetic neuromodulation with DREADDs, including the expression time course and factors affecting expression stability. This work offers valuable guidance for designing future long-term nonhuman primate studies, and underscores the translational potential of using DREADDs combined with PET imaging for non-invasive, gene-targeted therapies for neurological and psychiatric disorders.

Methods

Subjects

A total of 20 macaque monkeys (10 Japanese, Macaca fuscata; 8 Rhesus, Macaca mulatta; 2 Cynomolgus, Macaca fascicularis; 16 males, 4 females; weight, 3.4–9.9 kg; age, 3–13 years at the beginning of experiments) were used (Table 1). The monkeys were kept in individual primate cages in an air-conditioned room. A standard diet, supplementary fruits/vegetables, and a tablet of vitamin C (200 mg) were provided daily. All experimental procedures involving animals were carried out in accordance with the Guide for the Care and Use of Nonhuman Primates in Neuroscience Research (The Japan Neuroscience Society; https://www.jnss.org/en/animal_primates) and were approved by the Animal Ethics Committee of the National Institutes for Quantum Science and Technology (11-1038).

Viral vector production

We used a total of 11 adeno-associated virus (AAV) vectors. AAV2-Syn-HA-hM3Dq, AAV2-Syn-HA-hM4Di, and AAV5-Syn-HA-hM4Di were purchased from Addgene (MA, USA). The other viral vectors were produced by a helper-free triple transfection procedure, which was purified by affinity chromatography (GE Healthcare, Chicago, USA). Viral titer was determined by quantitative polymerase chain reaction using Taq-Man technology (Life Technologies, Waltham, USA).

Surgical procedures and viral vector injections

Surgical procedures have been described in detail in the previous reports (Hirabayashi et al., 2021; Nagai et al., 2020, 2016; Oyama et al., 2023, 2022a, 2021). The AAV vectors and brain regions into which they were injected are summarized in Table 2. Before surgery, structural scans of the head were acquired via magnetic resonance (MR) imaging (7 tesla 400 mm/SS system, Bruker, Billerica, MA, USA) and X-ray computed tomography (CT) (Accuitomo170, J. MORITA CO., Kyoto, Japan) under anesthesia (continuous intravenous infusion of propofol 0.2–0.6 mg/kg/min). Overlay MR and CT images were created using PMOD^® image analysis software (PMOD 4.4, PMOD Technologies Ltd, Zurich, Switzerland) to estimate the stereotaxic coordinates of target brain structures.

Surgeries were performed under aseptic conditions in a fully equipped operating suite. We monitored body temperature, heart rate, peripheral oxygen saturation (SpO₂), and endo-tidal CO₂ throughout all surgical procedures. Monkeys were immobilized by intramuscular (i.m.) injection of ketamine (5–10 mg/kg) and xylazine (0.2–0.5 mg/kg) after i.m. injection of atropine sulfate (0.02–0.05 mg/kg) and then intubated with an endotracheal tube. Anesthesia was maintained with isoflurane (1%–3%, to effect). After surgery, prophylactic antibiotics and analgesics (cefmetazole, 25–50 mg/kg/day; ketoprofen, 1–2 mg/kg/day) were administered for 7 days.

Viral vectors were injected with a stereotaxic apparatus or a hand-held syringe. For stereotaxic injection, burr holes (∼8 mm in diameter) were created for inserting the injection needle. Viruses were pressure-injected using a 10-μl Hamilton syringe (model 1701 RN, Hamilton, Reno, UV) mounted on a motorized microinjector (UMP3T-2, WPI, Sarasota, FL) held by a manipulator (model 1460, David Kopf, Tujunga, CA) on the stereotaxic frame. After making an approximately 3 mm cut in the dura mater, the injection needle was inserted into the brain and slowly moved down 1 to 2 mm beyond the target and then kept stationary for 5 min, after which it was pulled back up to the target location. Injection speed was 0.1 to 0.5 μl/min. After each injection, the needle remained in situ for 15 min to minimize backflow along the needle and then was pulled out very slowly.

For handheld injection, the target cortex of was exposed by removing a bone flap and reflecting the dura mater. Injections were made under visual guidance through an operating microscope (Leica M220, Leica Microsystems GmbH, Wetzlar, Germany) at an oblique angle to the brain surface. Each hemisphere received nine tracks of injections: one at the caudal tip, four along the dorsal bank, and four along the ventral bank of the principal sulcus, with 3 to 5 µL per track, depending on the depth (Oyama et al., 2023).

PET imaging

PET imaging was conducted as previously reported (Nagai et al., 2020). Briefly, PET scans were performed using a microPET Focus 220 scanner (Siemens Medical Solutions USA, Malvern, USA). Monkeys were initially immobilized with ketamine (5–10 mg/kg) and xylazine (0.2–0.5 mg/kg) and maintained under anesthesia with isoflurane (1%–3%) throughout the PET procedures. The isoflurane level was carefully adjusted to ensure the stability of continuously monitoring physiological parameters, such as body temperature, heart rate, SpO₂, and endo-tidal CO₂. Endo-tidal CO₂ levels were maintained between 35 and 45 mmHg to prevent substantial changes in cerebral blood flow (Markwalder et al., 1984), which could impact tracer kinetics. Transmission scans were performed for about 20 min with a Ge-68 source. Emission scans were acquired in 3D list mode with an energy window of 350–750 keV after intravenous bolus injection of [¹¹C]DCZ (∼350 MBq) and [¹⁸F]FDG (∼200 MBq). Data acquisition lasted for 90 min.

To estimate the specific binding of [¹¹C]DCZ, regional binding potential relative to a non-displaceable radioligand (BP_ND) was calculated by PMOD^® with an original multilinear reference tissue model (MRTMo) (Yan et al., 2021). The cerebellum was used as a reference region, and t* was 15 min. The DREADD expression level was defined as ΔBP_ND, calculated by subtracting the pre-injection value from the post-injection value because the BP_ND value included off-target binding. This minimized individual differences because off-target binding varied among subjects. The volumes of interest (VOIs) reflecting DREADD expression were identified by drawing contours at half of the maximum ΔBP_ND value in the ΔBP_ND parametric image, which was smoothed with a 2 mm Gaussian filter, and overlaid onto the corresponding region in each individual MR image.

For the FDG study, DCZ (1 μg/kg) or vehicle was administered intravenously 1 min before FDG injection. Data were converted to Standardized Uptake Value (SUV) images, averaged between 30- and 60-min, and normalized to SUV ratio (SUVR) images using the mean whole-brain value. Metabolic changes induced by DCZ were calculated as ΔSUVR and defined by the different DCZ and vehicle conditions, with VOIs for FDG analysis matching those used in the [¹¹C]DCZ-PET study.

Administration of DREADD agonists

DCZ (HY-42110, MedChemExpress) and CNO (Toronto Research) were dissolved in dimethyl sulfoxide (DMSO FUJIFILM Wako Pure Chemical Co.), aliquoted, and stored at −30 °C. These stock solutions were then diluted in saline to a final volume of 0.1 ml (2.5% DMSO in saline), achieving a dose of 0.1 ml/kg and 3 mg/kg for DCZ and CNO, respectively. For behavioral testing, an agonist was injected 15 min before experiments began, unless otherwise noted. Fresh solutions were prepared on each day of usage.

Behavioral tasks

The behavioral experiments are summarized in Table 1. Complete descriptions of the detailed procedures for each behavioral experiment are available in the previous reports (Hirabayashi et al., 2024, 2021; Hori et al., 2021; Oyama et al., 2024, 2022b, 2021). Here we focus on a modified Brinkman board task and a sensitivity to cold stimulation task, which were used to analyze the behavior of a monkey (#234) that had hM4Di expressed in its primary somatosensory cortex (Fig. 4). In the modified Brinkman board task, monkeys picked food pellets out of 10 slots in a board with one of their hands (i.e., contralateral or ipsilateral to the DREADD-expressing hemisphere), and the amount of time needed to pick up all ten items was recorded. The time to complete each trial was normalized to the number of pellets successfully picked up to obtain a value that reflected manual dexterity. In a single experimental session, the monkey performed five to ten trials with each hand before and 10 min after intravenous systemic administration of DCZ. In the sensitivity to cold task, monkeys were trained to place the sole of their foot on a cold metal plate that was placed in front of their monkey chair. The latency with which they withdrew their foot from the plate was recorded and used as an index of sensitivity to a cutaneous (cold) stimulus. If a monkey did not withdraw its foot within 60 s, the trial was terminated and a withdraw latency of 60 s was assigned for that trial. The monkeys performed five to ten trials for each of two temperature conditions (30°C for control and 10°C for cold stimulation) both before and 10 min after DCZ administration. Chemogenetic effects were assessed by comparing performance on the tasks pre- and post-injection and between contralateral and ipsilateral hands and feet (only the contralateral side should be affected by DCZ).

Statistics

All data and statistical analyses were performed using the R statistical computing environment (R Development Core Team, 2004). One way ANOVA was used to examine the relationship between vector injection sites (caudate, putamen, thalamus, amygdala, OFC, and SI) and peak expression level (peak ΔBP_ND value). Repeated measures two-way ANOVA was used to examine the effect of cumulative number of activation times × days after injection on post-peak fluctuation. A stepwise model selection using the MASS::stepAIC() function was performed to identify factors influencing the peak expression level. Starting with a full model that included five factors—virus titer, serotype, promoter, DREADD type, and tag—along with all their interactions, the best-fitting linear model was determined based on Akaike information criterion (AIC). The resulting best model included four variables, excluding serotype, and three interactions (Titer:DREADD type, Titer:Promoter, and DREADD type:Promoter). The significance of their contributions to the peak expression level, as well as the effect sizes (generalized eta squared), were assessed using Type II analysis of variance with the rstatix::anova_test() function.

Histology and immunostaining

For histological inspection, monkeys were deeply anesthetized with sodium pentobarbital (50 mg/kg, i.v.) or sodium thiopental (50 mg/kg, i.v.) after administration of ketamine hydrochloride (5‒10 mg/kg, i.m.) and then transcardially perfused with saline at 4°C, followed by 4% paraformaldehyde in 0.1 M phosphate buffered saline (PBS, pH 7.4) at 4°C. The brain was removed from the skull, postfixed in the same fresh fixative overnight, saturated with 30% sucrose in phosphate buffer (PB) at 4°C, and then cut serially into 50-μm-thick sections on a freezing microtome. For visualizing immunoreactive GFP (co-expressed with hM4Di) signals, a series of every 6th section was immersed in 1% skim milk for 1 h at room temperature and incubated overnight at 4°C with rabbit anti-GFP monoclonal antibody (1:500, G10362, Thermo Fisher Scientific, Waltham, MA, USA) in PBS containing 0.1% Triton X-100 and 1% normal goat serum for 2 days at 4°C. The sections were then incubated in the same fresh medium containing biotinylated goat anti-rabbit IgG antibody (1:1,000; Jackson ImmunoResearch, West Grove, PA, USA) for 2 h at room temperature, followed by avidin-biotin-peroxidase complex (ABC Elite, Vector Laboratories, Burlingame, CA, USA) for 2 h at room temperature. For visualizing the antigen, the sections were reacted in 0.05 M Tris-HCl buffer (pH 7.6) containing 0.04% diaminobenzidine (DAB), 0.04% NiCl₂, and 0.003% H₂O₂. The sections were mounted on gelatin-coated glass slides, air-dried, and cover-slipped. Parts of other sections were Nissl-stained with 1% Cresyl violet. Images of sections were digitally captured using an optical microscope equipped with a high-grade charge-coupled device (CCD) camera (Biorevo, Keyence, Osaka, Japan).

Acknowledgements

This study was supported by MEXT/JSPS KAKENHI Grant Numbers JP19K08138, JP23K27098 (to YN), 24H00734 (to TH), 22H05157 (to KI), and JP20H05955 (to TM), by JST PRESTO Grant Numbers JPMJPR22S3 (to KO), by AMED Grant Numbers JP23wm0625001 (to TM) and JP24wm0625307 (to TH), and by the Moonshot Research & Development Program (Millennia Program) from JST Grant Number JPMJMS2295 (to TM and KI). Japanese monkeys were provided by the Japan MEXT National Bio-Resource Project “Japanese Monkeys”. We thank Jun Kamei, Ryuji Yamaguchi, Yuichi Matsuda, Yoshio Sugii, Takashi Okauchi, Risa Suma, Tomomi Kokufuta, Rie Yoshida, and Akari Ueshiba for their technical assistance.

Additional information

Author contributions

Conceptualization: YN and TM; Resources: KI and KK; Investigation: YN, Yukiko H, TH, KO, and NM; Visualization: YN, KM, and TH; Formal analysis: YN, KM, and TH; Writing—original draft: YN, TH, and TM; Writing—review & editing: All authors.