Induction of Anxiety-Like Phenotypes by Knockout of Cannabinoid Type-1 Receptor in Amygdala of Marmosets

The amygdala is an important hub for the regulation of emotions, which is crucial for elucidating cellular and molecular mechanisms of many mental diseases. In the central nervous system, the endocannabinoid system plays a key role in the regulation of emotions and mainly functions through the cannabinoid type-1 receptor (CB1R), which is encoded by the Cnr1 gene. Although CB1R is highly expressed in the amygdala of non-human primates, little is known about its function. Here, we investigated the function of CB1R by knocking out the CB1R in the amygdala of adult marmosets through regional delivery of AAV-SaCas9-gRNA. We found that CB1R knockout in the amygdala of marmosets induced anxiety-like behaviors, including disrupted night sleep, agitated psychomotor activity in new environments, and reduced social desire, but had no effect on hedonic state and fear response. Moreover, CB1R-knockout marmosets exhibited up-regulated plasma cortisol levels, suggesting increased stress. These results showed that knockout of CB1R in the amygdala induced anxiety-like phenotypes in marmosets and shed new light on the mechanisms underlying the regulation of anxiety by CB1R in the amygdala of non-human primates.


45
The amygdala is critical for the regulation of emotion (Dunsmoor and Paz, 46 2015;Gothard, 2020;Janak and Tye, 2015;Morrison and Salzman, 2010) and 47 is implicated in various mental disorders such as depression (Hamilton and 48 Gotlib, 2008), social disturbance (Jayakar et al., 2020), and anxiety (Hyde et 49 al., 2011). As such, it is important that we fully understand its functions. 50 In the central nervous system, the endocannabinoid system (eCB) 51 functions to guard against negative emotions (Lutz et al., 2015) through the 52 cannabinoid type-1 receptor (CB1R), which is encoded by the Cnr1 gene and  (Haring et al., 2011;Moreira et al., 2009;58 Rey et al., 2012). CB1R is highly expressed in the amygdala, and associations 59 between CB1R expression in the amygdala and depression-like behavior have 60 been reported in mice (Shen et al., 2019). However, the function of amygdala 61 CB1R in emotional regulation is far less known in non-human primates (NHPs).

62
As an NHP model organism, marmosets (Callithrix jacchus) have garnered 63 considerable interest in studies on emotion and cognition due to their highly 64 vocal and emotionally rich social behaviors (Okano, 2021;Okano et al., 2016).
Marmosets are also invaluable in biomedical research due to their genetic and 66 physiological similarities to humans and relative ease of care in captivity (Marx, 67 2016). Recent advances in adeno-associated virus (AAV)-mediated delivery of 68 CRISPR/Cas9 in adult macaques (Li et al., 2021;Wu et al., 2021) highlight the 69 potential application of genetically modified adult marmosets in neuroscience. 70 Therefore, in the current study, we explored the function of CB1R in the 71 amygdala of marmosets via in vivo gene editing. Through regional delivery of 72 AAV-SaCas9-gRNA, marmosets with CB1R knockout in the amygdala 73 displayed anxiety-like phenotypes. These results revealed the emotion-specific 74 function of the amygdala CB1R in NHPs.  Based on virus injection differences, six marmosets were divided into 96 control and KO groups for behavioral tests, with two males and one female in 97 each group (Table 1). To better investigate the effect of CB1R knockout in the 98 amygdala, we compared behaviors individually before and two months after 99 virus injection (Fig. 1F). Comparisons of changes in behavior between the two 100 groups were also conducted.  Paquet et al., 2007;Ross et al., 2019;Zhou et al., 2019). Results showed that 105 sleep latency to light-off in control marmosets remained the same ( Fig. 2A, B, 106 and D) but was prolonged in KO marmosets with an increasing trend (Fig. 2F, phee calls to communicate with each other (Bezerra and Souto, 2008). We 150 hypothesized that CB1R knockout in the amygdala may induce abnormal social 151 behaviors in marmosets, reflected by a decrease in vocal production. We 152 recorded phee calls of separated marmosets to test social behavior.

153
Interestingly, while phee calls did not change in control marmosets ( Fig. 4A and   154 B), the KO marmosets tended to produce fewer phee calls after virus injection, showed that knockout of CB1R resulted in a decrease in social phee calls ( Fig.   157 4E), implying a reduction in vocal social desire. 158 We also assessed social behavior using the three-chamber social test, as

185
The above results suggest that knockout of CB1R in the amygdala did not 186 change hedonic state or innate fear response in marmosets.

187
Increased plasma cortisol level in CB1R knockout marmosets 188 Our results showed that CB1R knockout produced anxiety-like behaviors in 189 marmosets, which may induce in vivo biochemical abnormalities, such as 190 increased plasma cortisol (Chida and Steptoe, 2009). Thus, we tested whether 191 11 behavioral disruptions were accompanied by abnormal cortisol levels. Morning 192 blood samples were obtained from marmosets, followed by plasma collection 193 and ELISA analysis (Fig. 6A). Interestingly, plasma cortisol levels showed an 194 increasing tendency in the KO group but not in the control group after viral 195 injection ( Fig. 6B and C). Furthermore, intergroup comparison revealed that 196 plasma cortisol levels were higher in the KO marmosets than in the controls 197 (Fig. 6D).

198
These results indicate that CB1R knockout in the amygdala induces   In contrast to previous studies in rodents (Roche et al., 2007;Shen et al., 255 2019), we found that CB1R knockout in marmosets did not result in depression-256 14 like behaviors or altered fear response, which may be explained by species 257 differences. In addition, the previous rodent-based CB1R knockout studies 258 were circuit specific, cell-type specific, or amygdala-subregion specific, 259 whereas we targeted the entire amygdala region regardless of circuit or cell 260 type. Furthermore, as CB1R is mainly distributed in the presynaptic membrane 261 (Lutz et al., 2015), the intra-amygdala infusion of CB1R agonist and antagonist 262 in previous research may have affected CB1R in the amygdala-projecting 263 terminals rather than the amygdala itself, which may complicate the effects. As 264 such, the underlying mechanism still awaits further investigation.

265
There are some limitations in the current study. As mentioned above, we  Table 1.

413
All data generated or analyzed during this study are included in the 414 manuscript and supporting file; Source Data files have been provided.    Density of microglial cells in the amygdala of marmosets. n = 6 slices, two-tailed 12 unpaired t-test, t(10) = 0.7853, P = 0.2799. n.s., not significant; data are mean 13 ± SEM.
14 Source data were provided in figure 1-source data 2.