Induction of Anxiety-Like Phenotypes by Knockout of Cannabinoid Type-1 Receptor in Amygdala of Marmosets
1Department of Neurobiology and Department of Neurology of Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou 310058, China
2NHC and CAMS Key Laboratory of Medical Neurobiology, MOE Frontier Science Center for Brain Research and Brain-machine Integration, School of Brain Science and Brian Medicine, Zhejiang University, Hangzhou 310058, China
3Center for Brain Science and Brain-Inspired Intelligence, Research Units for Emotion and Emotion Disorders, Chinese Academy of Medical Sciences, China/Guangdong-Hong Kong-Macao Greater Bay Area, Joint Institute for Genetics and Genome Medicine between Zhejiang University and University of Toronto, Hangzhou 310058, China
4Interdisciplinary Institute of Neuroscience and Technology, School of Medicine, Zhejiang University, Hangzhou 310029, China
*Corresponding Author: Lixia Gao, E-mail: lxgao10@zju.edu.cn Xiao-Ming Li, E-mail: lixm@zju.edu.cnAbstract
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.
Article notes
Competing Interest Statement
The authors have declared no competing interest.
Introduction
The amygdala is critical for the regulation of emotion (Dunsmoor and Paz, 2015; Gothard, 2020; Janak and Tye, 2015; Morrison and Salzman, 2010) and is implicated in various mental disorders such as depression (Hamilton and Gotlib, 2008), social disturbance (Jayakar et al., 2020), and anxiety (Hyde et al., 2011). As such, it is important that we fully understand its functions.
In the central nervous system, the endocannabinoid system (eCB) functions to guard against negative emotions (Lutz et al., 2015) through the cannabinoid type-1 receptor (CB1R), which is encoded by the Cnr1 gene and modulates synaptic transmission by suppressing the release of neurotransmitters (Castillo et al., 2012; Ohno-Shosaku and Kano, 2014). CB1R dysfunction is correlated with psychiatric disorders (Choi et al., 2012; Hungund et al., 2004), and manipulation of CB1R in rodents shows bidirectional regulation of anxiety-like behaviors (Haring et al., 2011; Moreira et al., 2009; Rey et al., 2012). CB1R is highly expressed in the amygdala, and associations between CB1R expression in the amygdala and depression-like behavior have been reported in mice (Shen et al., 2019). However, the function of amygdala CB1R in emotional regulation is far less known in non-human primates (NHPs).
As an NHP model organism, marmosets (Callithrix jacchus) have garnered considerable interest in studies on emotion and cognition due to their highly vocal and emotionally rich social behaviors (Okano, 2021; Okano et al., 2016).
Marmosets are also invaluable in biomedical research due to their genetic and physiological similarities to humans and relative ease of care in captivity (Marx, 2016). Recent advances in adeno-associated virus (AAV)-mediated delivery of CRISPR/Cas9 in adult macaques (Li et al., 2021; Wu et al., 2021) highlight the potential application of genetically modified adult marmosets in neuroscience. Therefore, in the current study, we explored the function of CB1R in the amygdala of marmosets via in vivo gene editing. Through regional delivery of AAV-SaCas9-gRNA, marmosets with CB1R knockout in the amygdala displayed anxiety-like phenotypes. These results revealed the emotion-specific function of the amygdala CB1R in NHPs.
Results
Construction and validation of AAV-mediated CRISPR/Cas9 virus
Based on SaCas9 (Hanna and Doench, 2020), an AAV-mediated CRISPR/Cas9 virus was used to knockout the Cnr1 gene in the amygdala of adult marmosets in vivo. To specifically target the Cnr1 gene, three guide RNAs (gRNAs) were synthesized and inserted into the AAV-hSyn-SaCas9-U6-gRNA vector (Fig. 1A). Considering the deep location and individual variation of the amygdala in NHPs, MRI was applied to localize and target the exact region of the amygdala in marmosets before virus injection (Fig. 1B). To validate knockout efficiency of the virus, one marmoset (Ctrl 1) receiving control virus injection and one marmoset (S) receiving knockout (KO) virus injection were sacrificed via transcardial perfusion, and brain slices were collected for fluorescence staining to evaluate knockout efficiency. Using RNAscope staining, we found much fewer cells expressing CB1R mRNA in the amygdala of KO marmoset than control marmoset (Fig. 1C-E). The numbers of GFAP+ cells and Iba1+ cells (markers of astrocytes and microglial cells, respectively) were the same in control and KO marmosets (Figure 1-figure supplement 1). These results indicate that the AAV-mediated CRISPR/Cas9 in vivo gene editing successfully knocked out the Cnr1 gene in the amygdala of adult marmosets without difference of neuroinflammation effects between two groups.
Based on virus injection differences, six marmosets were divided into control and KO groups for behavioral tests, with two males and one female in each group (Table 1). To better investigate the effect of CB1R knockout in the amygdala, we compared behaviors individually before and two months after virus injection (Fig. 1F). Comparisons of changes in behavior between the two groups were also conducted.
Disrupted night sleep in CB1R knockout marmosets
As many emotional disorders co-occur with sleep disturbance (Qiu et al., 2019), we monitored the sleep of marmosets using an actigraphy device (Paquet et al., 2007; Ross et al., 2019; Zhou et al., 2019). Results showed that sleep latency to light-off in control marmosets remained the same (Fig. 2A, B, and D) but was prolonged in KO marmosets with an increasing trend (Fig. 2F, G, and I) after virus injection. In addition, night sleep duration in control marmosets did not change (Fig. 2C and E) but was shortened in KO marmosets with a decreasing trend (Fig. 2H and J) after virus injection. By comparing changes in sleep latency and sleep duration between groups, we also found prolonged sleep latency (Fig. 2K) and reduced sleep duration (Fig. 2L) in the KO group animals. We also examined whether decreased night sleep was due to increased daytime sleep but found that daytime nap time was not influenced by virus injection in either the control or KO groups (Figure 2—figure supplement 1). We also explored sleep efficiency (Allison et al., 2016) and fragment index (Zhou et al., 2019), which both contribute to sleep quality, and found that neither were affected in control marmosets (Figure 2—figure supplement 2A and B). However, sleep efficiency showed a decreasing trend (Figure 2—figure supplement 2C and E) and the fragment index showed an increasing trend (Figure 2—figure supplement 2D and F) in KO marmosets, suggesting decreased sleep quality.
Taken together, our results suggest that knockout of CB1R in the amygdala disrupts night sleep quantity and quality in marmosets, implicating emotional dysregulation.
Agitated psychomotor activity in new environment in CB1R knockout marmosets
New environments are stressful and may induce behavioral and emotional abnormalities in animals and humans (Penninx et al., 2021; Spasojevic et al., 2016). We transferred single marmosets to a new environment (cage) (Qiu et al., 2019) to test the effects on behavior. In contrast to the control group, the KO marmoset trajectories were much denser after virus injection, indicating agitation in the new environment (Fig. 3A and E). In addition, while moving distance, average velocity, and stationary time remained the same in the control group (Fig. 3B–D), KO marmosets showed increased moving distance and average velocity (Fig. 3F and G) as well as decreased stationary time (Fig. 3H) after virus injection. Comparisons between groups further confirmed higher agitation in the KO marmosets, with increased moving distance (Fig. 3I) and average velocity (Fig. 3J) and decreased stationary time (Fig. 3K). Marmoset activity in their home cages was also measured by actigraphy (Figure 3—figure supplement 1A). However, no significant changes were found in either group (Figure 3—figure supplement 1B–F), indicating that motor ability was unaffected by CB1R knockout in the amygdala.
The above results suggest that CB1R knockout in the amygdala of marmosets specifically induces agitated psychomotor activity in stressful new environments but not in familiar surroundings.
Decreased social desire for vocal communication in CB1R knockout marmosets
Marmosets are highly vocal and social animals (Miller et al., 2016) and use phee calls to communicate with each other (Bezerra and Souto, 2008). We hypothesized that CB1R knockout in the amygdala may induce abnormal social behaviors in marmosets, reflected by a decrease in vocal production. We recorded phee calls of separated marmosets to test social behavior. Interestingly, while phee calls did not change in control marmosets (Fig. 4A and B), the KO marmosets tended to produce fewer phee calls after virus injection, indicating less vocal social behavior (Fig. 4C and D). Intergroup comparison showed that knockout of CB1R resulted in a decrease in social phee calls (Fig. 4E), implying a reduction in vocal social desire.
We also assessed social behavior using the three-chamber social test, as used in rodents (Rein et al., 2020). After habituation, marmosets were tested by the introduction of a stranger followed by the introduction of a cage-mate (Figure 4—figure supplement 1A). However, no significant changes in social behavior were observed in either the control or KO groups (Figure 4—figure supplement 1B–D). A potential explanation may be that rodent behavioral tests are unsuitable for testing sociability in NHPs, given that different species use different modalities to recognize and communicate with conspecifics (Miller et al., 2016).
Taken together, our results suggest that knockout of CB1R in the amygdala decreases social desire in marmosets, implicating anxiety-like social avoidance.
Unaltered hedonic and fear behaviors in CB1R knockout marmosets
We previously showed that disruption of circuit-specific CB1R in the amygdala induces depression-like anhedonia in mice (Shen et al., 2019). In the current study, we tested the hedonic state in marmosets via an adapted sucrose preference test (Alexander et al., 2019) (Fig. 5A). In the test, marmosets consumed sucrose and water gradually over 2 h (Figure 5—figure supplement 1) and sucrose preference was examined in the first 30 and 120 min, respectively. Results showed that sucrose preference was not changed after virus injection in either the control or KO groups (Fig. 5B–D), suggesting no depression-like behavior.
Previous studies have indicated that the amygdala is a pivotal region for fear response (Gothard, 2020; Janak and Tye, 2015). Thus, we measured fear response in marmosets using the snake intruder test based on previous protocols (Melamed et al., 2017) (Fig. 5E). Unexpectedly, knockout of CB1R in the amygdala did not change fear response behavior in marmosets (Fig. 5F– H).
The above results suggest that knockout of CB1R in the amygdala did not change hedonic state or innate fear response in marmosets.
Increased plasma cortisol level in CB1R knockout marmosets
Our results showed that CB1R knockout produced anxiety-like behaviors in marmosets, which may induce in vivo biochemical abnormalities, such as increased plasma cortisol (Chida and Steptoe, 2009). Thus, we tested whether behavioral disruptions were accompanied by abnormal cortisol levels. Morning blood samples were obtained from marmosets, followed by plasma collection and ELISA analysis (Fig. 6A). Interestingly, plasma cortisol levels showed an increasing tendency in the KO group but not in the control group after viral injection (Fig. 6B and C). Furthermore, intergroup comparison revealed that plasma cortisol levels were higher in the KO marmosets than in the controls (Fig. 6D).
These results indicate that CB1R knockout in the amygdala induces anxiety-like behaviors in marmosets accompanied by increasing plasma cortisol levels.
Discussion
Using in vivo gene editing, immunohistochemical analysis, behavioral testing, and biochemical measurement, we studied the function of CB1R in the amygdala of adult marmosets. Results showed that AAV-mediated delivery of the CRISPR/Cas9 system successfully knocked out CB1R expression in the amygdala of adult marmosets and induced anxiety-like phenotypes (Park and Kim, 2020; Penninx et al., 2021), including disrupted night sleep, agitated psychomotor activity in new environments, decreased desire for vocal communication, and increased plasma cortisol levels. Applying gene editing to study the region-specific functions of CB1R in adult marmosets should help improve our understanding of the eCB system in the amygdala of NHPs and advance the clinical application of eCB for the diagnosis and treatment of psychiatric disorders, such as anxiety.
Although rodents are useful models in neuroscience, they differ from humans in multiple aspects, which may hinder our understanding of the mechanisms underlying human emotion, cognition, and mental disorders (Camus et al., 2015; Feng et al., 2020; Izpisua Belmonte et al., 2015). As such, studying the brains of NHPs is critical for translational research due to their homogeneity to humans (Jennings et al., 2016). Electrolytic lesions and pharmacological manipulations via cannulas are commonly used to explore the general function of the amygdala in NHPs (Braesicke et al., 2005; Dal Monte et al., 2015; Wellman et al., 2016), but cause non-negligible side effects and lack molecular marker specificity. At present, gene editing techniques widely used in rodents are not well established in NHPs, which greatly impedes our understanding of the brain mechanisms that underpin mental disorders in humans. Here, for the first time, we established an in vivo gene editing method to study the function of CB1R in the amygdala of marmosets. We successfully demonstrated the feasibility of virus-mediated gene editing in adult NHPs, thus providing a novel strategy to develop transgenic NHP models in a rapid and cost-effective manner (Marx, 2016).
Following CB1R knockout in the amygdala, marmosets showed various anxiety-related behaviors, consistent with previous reports indicating that the amygdala and CB1R are involved in the onset and development of anxiety (Janak and Tye, 2015; Lutz et al., 2015; Ramikie and Patel, 2012). Of note, we detected sleep disturbance and agitation in the marmosets with anxiety-like behaviors, which are rarely detected in anxiety model rodents but are key pathological manifestations of anxiety in humans (Park and Kim, 2020). We also detected increased plasma cortisol levels in the KO marmosets, consistent with previous study (Akirav, 2013) and a representative biomarker of anxiety in humans (Chida and Steptoe, 2009), but which is less noted in rodent experiments (Calhoon and Tye, 2015; Tovote et al., 2015). Notwithstanding, the causal relationship between the change in cortisol level and behavior remains unclear and deserves further investigation. However, our results suggest that NHP models are crucial for studying human-relevant diseases and exhibit considerable potential in identifying specific molecular biomarkers for the diagnosis and treatment of psychiatric disorders.
As CB1R exerts its function through synaptic modulation, and not all neurons in the amygdala express CB1R (Shen et al., 2019), a potential explanation for our findings may be that neurons that regulate anxiety state may express higher levels of CB1R. Thus, knockout of CB1R may disrupt the anxiogenic-anxiolytic balance, thereby potentiating anxiety transmission by enhancing amygdala-related anxiogenic circuit activity. However, further circuit-based studies are required to confirm this hypothesis.
In contrast to previous studies in rodents (Roche et al., 2007; Shen et al., 2019), we found that CB1R knockout in marmosets did not result in depression-like behaviors or altered fear response, which may be explained by species differences. In addition, the previous rodent-based CB1R knockout studies were circuit specific, cell-type specific, or amygdala-subregion specific, whereas we targeted the entire amygdala region regardless of circuit or cell type. Furthermore, as CB1R is mainly distributed in the presynaptic membrane (Lutz et al., 2015), the intra-amygdala infusion of CB1R agonist and antagonist in previous research may have affected CB1R in the amygdala-projecting terminals rather than the amygdala itself, which may complicate the effects. As such, the underlying mechanism still awaits further investigation.
There are some limitations in the current study. As mentioned above, we treated the amygdala as a whole and different types of cells and subregions in the amygdala may compensate in function, leading to less dramatic changes in marmoset behavior. Thus, future studies on the different subregions and types of neurons in the NHP amygdala are needed. The use of Cre-expressing marmosets may also be insightful (Okano, 2021). The small sample size in our study due to animal number limitations during the COVID-19 pandemic may explain why some behavioral trends did not reach significance. However, the significant behavioral changes in the KO group were considered prominent and robust.
In conclusion, we established a novel gene editing method in adult marmosets and highlighted the function of CB1R in the marmoset amygdala in the regulation of anxiety-like phenotypes, thus paving the way for further use of NHPs in basic research.
Materials and Methods
Animal ethics
Male and female marmosets (350–450 g, 2–4 years old) were purchased from Johnbio (Jiangsu, China) and kept in the Non-Human Primate Center of Zhejiang University (ZJU) in pairs. All experiment protocols were conducted under the guidelines for the care and use of laboratory animals of ZJU and were approved by the Animal Advisory Committee at ZJU following the National Institutes of Health (NIH) guidelines. Animal details are provided in Table 1.
Viruses
All viruses were constructed and purchased from Taitool Bioscience (Shanghai, China). AAV2/9-hSyn-saCas9-hU6-gRNA(cjCB1R) (1.21 × 1013 viral genomes (vg)/ml) was used to knockout the Cnr1 gene, AAV2/9-hSyn-saCas9-pA-hU6-gRNA(empty) (1.58 × 1013 vg/mL) was used as the control virus, and AAV2/9-hSyn-EGFP-WPRE-pA (1.42 × 1013 vg/mL) was co-injected with the above viruses (1:9) to display injection location.
Magnetic resonance imaging (MRI)-guided amygdala location
MRI data were acquired on a 7-T research scanner (Siemens Healthcare, Erlangen, Germany) with a single loop coil (RAPID MR International, Columbus, OH, USA) for signal reception and transmission. Structural images were obtained at an isotropic voxel size of 0.3 mm. Briefly, animals were anesthetized with alfaxalone (intramuscular (i.m.), 0.1 mg/kg, Jurox, North Kansas City, USA) and maintained by sustained isoflurane (0.5%–1.2%, Yipin Pharmacology, Hebei, China). Two glass tubes (OD: 1.0 mm, ID: 0.58 mm; Sutter Instrument, USA) filled with gel were placed interaurally as markers. After the animal was fixed in a home-made MRI compatible stereotaxic apparatus, the scan procedure was started. The MRI data were analyzed using MRIcron v1.0 (NITRC, USA) to calculate coordinates of the amygdala to guide virus injection.
Virus injection
The surgery and virus injection were performed under aseptic conditions. The animals were initially anesthetized with alfaxalone (i.m., 0.1 mg/kg, Jurox, North Kansas City, USA) and maintained by isoflurane (0.5%–1.2%, Yipin Pharmacology, Hebei, China). The head was fixed in a stereotaxic frame (RWD, Shenzhen, China) and a craniotomy was carried out using a dental drill according to the coordinates. A cocktail of viruses was infused through a Hamilton syringe (Hamilton, USA) placed in a syringe pump (KD Scientific, USA) at a speed of 80 nL/min for a total of 1 μL at each injection site. After infusion, the syringe was left in place for 5 min. The same operation was repeated until all coordinates were injected. After injection, the wound was carefully cleaned and sutured. The animals were given special care for two weeks with additional nutritional supplements and physiological examination. Post-injection measurements were performed two months after injection.
Histological analysis and imaging
Animals were first euthanized by administering an overdose of sodium pentobarbitone (100 mg/kg) and then transcardially perfused with phosphate-buffered saline (PBS), followed by 4% paraformaldehyde. Brains were removed and post-fixed in 4% paraformaldehyde overnight at 4 °C, and then immersed in 30% sucrose in PBS for 72 h. The brains were embedded in Optimal Cutting Temperature compound (Sakura Finetek, USA) and 20-μm cryosections were cut using a cryostat (Leica Microsystems, Germany), then washed with PBS (5 min/time).
RNAscope in-situ hybridization was performed using a RNAscope Multiplex Fluorescent Reagent Kit v2 (ACDbio, USA) and RNAscope Probe-cj-Cnr1 (Cat No. 565041) (ACDbio, USA) according to the manufacturer’s standard protocols.
For immunofluorescence staining, sections first underwent target retrieval pretreatment, and were then blocked with 3% bovine serum albumin in PBST (0.3% Triton X-100 in PBS) for 1 h and incubated with primary antibodies overnight at 4 °C. The sections were then incubated with fluorescent secondary antibodies (1:400, Invitrogen, USA) for 2 h at room temperature. The primary antibodies were anti-GFAP (1:800, BioLegend, SMI-21R, USA) and anti-Iba1 (1:800, Wako, 019-19741, Japan). Anti-GFP was used to better show EGFP expression in the amygdala. Sections were mounted after 4′,6-diamidino-2-phenylindole (DAPI) staining (1:5 000, Sigma-Aldrich, USA). Confocal images were captured under 20× objective using A-1R Confocal Microscope (Nikon, Japan). Cell counts were performed with ImageJ v1.52 (NIH, USA).
Daily activity and sleep measurement in home cages
Daily activity and sleep patterns were assessed using an ActiWatch Mini (CamNtech, UK) worn on the neck of each animal. Animals were allowed to habituate to the device for 2 days followed by 7-day data collection. Daytime was defined as 07:00 am to 19:00 pm and night-time was defined as 19:00 pm to 07:00 am according to light/dark cycle (lights on at 07:00 am and off at 19:00 pm). Activity data were analyzed in 1-min epochs using Sleep Analysis v7 (CamNtech, UK) off-line. Sleep latency was defined by the timepoint of the animal falling asleep relative to 19:00 pm. Sleep duration was defined as the time difference between light off and the time the animal fell asleep. Sleep efficiency was defined as the percentage of sleep time minus waking time. Sleep fragment index was determined by the percentage of waking time plus the percentage of immobile time. Calculation details are provided in the Sleep Analysis v7 software guidelines.
New environment activity examination
The marmosets were transferred to a new cage (1 × 1 × 1 m) and their activity was recorded with a camera for 10 min. Trajectories and moving distances were analyzed using a custom Matlab 2017b (MathWorks, USA) program (provided by Prof. Xinjian Li). Average velocity was defined as moving distance divided by recording time. Stationary time was defined as total duration of immobile phases of more than 1 s.
Vocal test
The animals were separated from their home cage-mates and colony and transferred to the testing room. A microphone was placed in front of the transfer cage for vocal recording for 20 min. Phee calls were manually analyzed using Adobe Audition CS6 v5.0 (Adobe, USA).
Sucrose preference test
Sucrose preference tests were performed in the home cages in the morning. Before testing, cage-mates were transferred to an adjacent cage, after which, each marmoset was provided with two bottles containing 100 mL of 10% sucrose solution and 100 mL of water, respectively. The sucrose preference test lasted for 2 h. The weights of the sucrose and water were measured and the positions of the two bottles were exchanged every 30 min. Sucrose preference index was Vsucrose consumed / (Vsucrose consumed + Vwater consumed).
Snake intruder test
Animals were first habituated to the test apparatus for 10 min. A rubber snake was then introduced in the corner and monkey behavior was recorded by video. Snake-paired corner time was determined as the time spent in the same corner as the snake. Snake inspection time was determined as the time spent staring at the snake. Immobile time was determined as the time spent immobile for more than 1 s.
Blood sample collection and cortisol test
Blood samples (0.5 mL) were collected intravenously in EDTA-coated tubes at 10:00 am for two continuous days. The blood samples were immediately centrifuged at 1 000 ×g for 10 min at room temperature. The plasma was collected and stored at −80 °C until use. Plasma cortisol levels were measured using an ELISA toolkit (Enzo, ADI-901-071, USA) according to the standard protocols provided.
Quantification and statistical analysis
Sample size for statistical comparisons were referred to previous research (Li et al., 2021; Wu et al., 2021). All data were analyzed with GraphPad Prism v6.01 (GraphPad Software, USA). Differences in individuals before and after virus injection were tested using two-tailed paired t-test and differences between groups were tested using two-tailed unpaired t-test. Repeated two-way analysis of variance (ANOVA) was used to test differences in the three-chamber test. Data are shown as mean ± standard error of the mean (SEM). Differences were considered statistically significant at * P < 0.05, ** P < 0.01, *** P < 0.001, and **** P < 0.0001.
All data generated or analyzed during this study are included in the manuscript and supporting file; Source Data files have been provided.
Acknowledgements
We thank Dr. Zilong Qiu for providing the AAV-SaCas9 vector. We are grateful to Research Assistant Shuangshuang Liu from the Core Facilities of Zhejiang University School of Medicine, as well as Dr. Sanhua Fang and Research Assistant Li Liu from the Core Facilities of Zhejiang University Institute of Neuroscience. This work was supported by Zhejiang Province Natural Science Foundation of China (LD22H090003), Key-Area Research and Development Program of Guangdong Province (2019B030335001 and 2018B030334001), National Natural Science Foundation of China (31871070, 82090031, 32071097, 31871056 and 32170991), Key R&D Program of Zhejiang Province (2020C03009), Fundamental Research Funds for the Central Universities (2021FZZX001-37), Non-Profit Central Research Institute Fund of the Chinese Academy of Medical Sciences (2019PT310023), and CAMS Innovation Fund for Medical Sciences (2019-I2M-5-057).
Conflict of Interests
The authors report no biomedical financial interests or potential conflicts of interests.