93% loss of DβH immunoreactivity compared with vehicle rats (Fig. 1a, b and Supplementary Data 1). On test days, depletion of noradrenergic neurons induced a significant decrease in the CPA score (Fig. 1c; Supplementary Table 1 and Supplementary Data 1)./p> 0.05./p> 28% and specificity > 95% (Fig. 2d, e and Supplementary Data 1). On test days, optogenetic inhibition of LC neurons, during training or before testing days, dramatically reduced the CPA score (Fig. 2a, f, g, Supplementary Table 2 and Supplementary Data 1). In contrast, optical inhibition of noradrenergic neurons in the absence of CRD has no effect on CPA score (Supplementary Fig. 3a, b and Supplementary Data 1)./p> 0.9999, two-way ANOVA with Bonferroni test). h Representative images of c-Fos expression in EYFP and eNpHR3.0 rats. Scale bar: 50 µm. i Quantification of c-Fos+ cells in the LC and ACC region after optogenetic inhibition (n = 3–5 rats/group, 3 sections from each animal; **p < 0.0083; t5 = 4.224 (LC); ***p < 0.0001, t6 = 9.946 (ACC), unpaired t-test). Results are presented as mean ± SEM. ns = non-significant, p > 0.05./p> 0.05. Propran = propranolol./p>26.5% and specificity >94% (Fig. 4c–e, Supplementary Fig. 1b, and Supplementary Data 1). On test days, optogenetic activation of noradrenergic neurons during training and before testing days significantly increased the CPA score (Fig. 4b, f, Supplementary Fig. 1f, Supplementary Table 4, and Supplementary Data 1). Further, we report that optical stimulation of noradrenergic neurons in the absence of CRD does not affect the CPA score (Supplementary Fig. 3a, c and Supplementary Data 1)./p> 0.05. Propran = propranolol./p>95%) and specificity (>97%; Supplementary Fig. 1c, d and Supplementary Data 1). We found that optogenetic activation of ACC astrocytes during conditioning significantly promoted the CPA score (Supplementary Fig. 1e, Supplementary Table 5, and Supplementary Data 1). Additionally, we report that optical stimulation of ACC astrocytes in the absence of CRD does not change the CPA score (Supplementary Fig. 3a, d and Supplementary Data 1)./p>87% S100β+ cells expressed opto-β2AR+ with a specificity of >97% (Fig. 5a–c and Supplementary Data 1). On test days, optogenetic activation of astrocytic β2ARs during training or before testing days significantly promoted the CPA score (Fig. 5d, e, Supplementary Table 5, and Supplementary Data 1). In addition, we showed opto-activation of β2ARs in ACC astrocytes had no effects on U69593-induced CPA (Fig. 5f, Supplementary Table 5, and Supplementary Data 1). Further, optical stimulation of ACC astrocytic β2ARs in the absence of CRD does not change the CPA score (Supplementary Fig. 3a, e and Supplementary Data 1). In separate group of rats, we also showed that photoactivation of astrocytic β2ARs did not change the VMR to graded pressure of CRD (Supplementary Fig. 2c, d and Supplementary Data 1)./p> 0.05. TD1 = Test day 1. HC = home cage./p>89 % GFAP+ cells in ACC area expressed rβ2AR-mCherry with >96% specificity (Fig. 6a, d, e and Supplementary Data 1). Moreover, co-staining with microglia activation marker Iba1 showed no overlap with β2AR mCherry+ cells (Supplementary Fig. 6a). When co-stained with the neuronal nuclear marker NeuN, it offered approximately 1.84% off-target expression in the ACC neurons (Supplementary Fig. 6b, c and Supplementary Data 1)./p> 0.05./p>88% ACC NeuN+ cells expressed β2AR-mCherry with >98% specificity (Supplementary Fig. 7a, d, e and Supplementary Data 1). The histochemical staining shown that injection of miRNAi into ACC region produced significant reduction in the expression of neuronal β2ARs (Supplementary Fig. 7b, c, f and Supplementary Data 1). The western blot data further confirmed that knockdown effect was more robust in miRNAi(rβ2AR) compared to the negative control rats (Supplementary Fig. 7g and Supplementary Data 1). Notably, on test days, knockdown of ACC neuronal β2ARs has no significant effect on the CPA score (Supplementary Fig. 7h, Supplementary Table 6, and Supplementary Data 1). Intriguingly, taken together, the data demonstrate that ACC astrocytic β2ARs, not the neuronal β2ARs, are required for aversive memory formation./p>87% of GFAP+ cells in the ACC region expressed hM4Di-mCherry with a specificity of >94% (Supplementary Fig. 9a–c and Supplementary Data 1). When co-stained with neuronal nuclear marker, <5.5% of hM4Di-mCherry+ cells overlapped with ACC neurons (Supplementary Figure 6d, e and Supplementary Data 1). On test days, activation of the Gi pathway in ACC astrocytes before training or before testing days substantially blocked the CPA memory (Supplementary Fig. 9d, e, Supplementary Table 7 and Supplementary Data 1). In a separate group of rats, we also showed that the CNO (1 mg/kg b.w) treatment before conditioning itself does not affect the CPA score and c-Fos expression in the ACC region (Supplementary Fig. 10a–d and Supplementary Data 1)./p> 0.05./p>