The Ghrelin/GHS-R1a Pathway is Involved in the Mechanism of the CTA-regulated Neuronal Loop

Objective: We aimed to investigate the effects of ghrelin and its receptor GHS-R1a on the regulation of taste aversion-associated emotional memory. Methods: We studied the effect of selective and reversible inactivation of emotion-associated neurons on emotions processed by the local neuronal loop of the lateral amygdala and insular cortex. We used intraperitoneal clozapine-N oxide (CNO) injection, local brain microinjection, and taste aversion test. Results: Microinjection of ghrelin into the lateral amygdala blocked the acquisition of taste aversion memory in mice. However, microinjection of ghrelin did not affect memory consolidation. Adeno-associated virus (AAV) microinjection into the lateral amygdaloid nucleus blocked the acquisition of taste aversion memory in mice. Moreover, intraperitoneal injection of CNO inhibited taste aversion memory formation, altering the aversion index. Conclusion: Microinjection of ghrelin and AAV-CaMKII-hM4Di-2A-GHS-R1a-GFP into the lateral amygdala inhibited the acquisition of conditioned taste aversion in mice. The inhibition of the memory formation was achieved through the activation of growth hormone secretagogue receptor 1a (GHS-R1a).

The amygdala plays an important role in the acquisition, consolidation, and extraction of emotional memory [7] . It has been shown that the amygdala receives projections from ghrelin neurons.  CaMKII-hM4Di-2A-GHS-R1a-GFP into a   specific brain region, infected neurons co-express   GHS-R1a,  hM4Di, and GFP. Subsequent intraperitoneal injection of clozapine-N-oxide (CNO) [8][9][10][11] specifically activates hM4Di after 45 min and results in neuronal inactivation due to hyperpolarization. Conditioned taste aversion (CTA) is an experimental model commonly used in studies of memory and learning. In addition, CTA is widely used in studies on acquisition, consolidation, and subsidence processes and mechanisms of emotional memory [8] . The lateral amygdala and insular cortex (IC) are important brain regions involved in the acquisition and storage of CTA-associated emotion and memory [12][13] .
In this study, we explored the effects of ghrelin and its receptor GHS-R1a on the regulation and the underlying molecular mechanism of taste aversion-associated emotional memory by using the DREADD system [14] . The DREADD system allows selective and reversible functional regulation of specific neuron types. We studied the effect of selective and reversible inactivation of emotion-associated neurons on emotions associated with the neuronal circuit of the lateral amygdala and IC using intraperitoneal injection of CNO, microinjection, and taste aversion behavior tests.

Animals and reagents
Adult male Jac mice (2-3 months of age, weighing 25-30 g) were obtained from Beijing Vital River Laboratory Animal Technology Co., Ltd. Ghrelin was purchased from Tocris Bioscience (USA), and was prepared as 1 mmol/L stock solution by using sterile saline. Ghrelin stocks were divided into aliquots and stored at -80°C. Sterile saline was used to dilute ghrelin stocks.

Construction of HSV and AAV viral plasmids
The hM4Di plasmid was purchased from Addgene.
The GHS-R1a plasmid was a gift from Helen Wise After the injection, the needle remained at the injection site for another 5 min before being slowly removed to prevent spilling. After surgery, mice recovered for 4-6 weeks before undergoing behavioral tests.

Histological examination
mices were anesthetized, and the tissue was fixed by perfusing 4% paraformaldehyde into the heart. After perfusion, the brain was removed, and fixed with paraformaldehyde for another 12 h. Subsequently, the brains were transferred to 30% sucrose for 2 days.
The brains were frozen and sectioned into 40-µm sections. After staining with crystal violet, locations where the syringe and needles had been placed were examined by using LEICA microscope.

CTA experiment
LiCl consumption test: Mice were water-deprived for 24 h and were allowed to adapt to the following

Successful transfection of AAV viruses and neuronal overexpression of GHS-R1a and hM4Di
After  (Figure 1).

Microinjection of ghrelin into the amygdala blocked the acquisition, but not the formation, of aversion memory
Twenty min prior to CTA training, 12 ng of ghrelin (0.5 µl/side) were microinjected into the lateral amygdala. After 24 h, the AI of the ghrelin-treated group (n = 10) was significantly lower than the AI of the control group (n = 10; Figure 2A). The taste AI of the ghrelin-treated group did not differ significantly from chance (50% cutoff, p>0.05, Figure 2A). Low doses of (12 µg/0.5 µl) ghrelin microinjected into the lateral amygdala inhibited the acquisition of CTA memory in mice. In contrast, microinjection of ghrelin into the amygdala did not affect the liquid uptake of mice during CTA training and memory tests (Figures 2B and 2C).

Microinjection of AAV into the IC blocked the acquisition of CTA memory in mice
Prior to the start of CTA training, AAV-hM4DGi-GHS-R1a and the corresponding viral AAV-CON were microinjected into the IC. After 24 h, the AI of the mice injected with AAV-hM4DGi-GHS-R1a (n = 10) was smaller than the AI of the AAV-CON group (n = 10) ( Figure 3C).
However, virus microinjection into the IC did not significantly affect liquid uptake of mice during CTA training and memory tests (Figures 3A and 3B).

Microinjection of AAV into the IC blocked the acquisition of CTA memory in mice; concurrent intraperitoneal injection of CNO altered the formation of CTA memory
Microinjection of AAV into the IC blocked the acquisition of CTA memory in mice (Figure 4 C).
However, liquid uptake during CTA training and memory tests was not significantly affected ( Figures   4A and 4B). Intraperitoneal injection of CNO prior to the start of the test revealed that the specific activation of hM4DGi receptors by CNO led to hyperpolarization and neuronal inactivation. The AI of the AAV-hM4DGi-GHS-R1a group (n = 10) was reduced fter CNO injection (Figure 4 D). After 24 h, when CNO was fully metabolized, hM4DGi receptors were no longer activated, and the AI recovered.

Microinjection of AAV into the amygdala blocked the acquisition of CTA memory in mice; concurrent injection of CNO inhibited the formation of CTA memory
Microinjection of the AAV-hM4DGi-GHS-R1a into the amygdala blocked the acquisition of CTA memory in mice (Figure 5 C, test 1), and did not significantly affect liquid uptake during CTA training and memory tests (Figures 5A and 5B). However, intraperitoneal injection of CNO prior to the test revealed that the AAV-CON and AAV-HA/hM4DGi were activated and triggered the silencing of neuronal inactivation. Thus, AI decreased and memory formation was inhibited (Figure 5 C, test 2). The AI of the AAV-hM4DGi-GHS-R1a group (n = 10) was significantly lower than the AI of the AAV-CON group (n = 10; Figure 5 C test 2). After 24 h, when CNO was fully metabolized, the neuronal activity and the acquisition of memory recovered (n = 10, **p ＜ 0.01, *p ＜ 0.05). Accordingly, we observed increased AI (Figure 5 C test 3).

Discussion
In the present study, we demonstrated that Moreover, AAV injection led to hM4Di expression.
Interestingly, intraperitoneal injection of CNO prior