Figure S1 The breaking point for the cocaine was tested under the progressive ratio (PR) procedure on WD1 and WD45. The breaking point was increased on WD45 compared with WD1. Data are expressed as means ± SEM, n = 4. * p < 0.05, WD1 vs. WD45.

Figure S2 Intraperitoneal or intracerebroventricular injection of LTCCs antagonist verapamil inhibited seeking behavior after long-term withdrawal. (A) Intraperitoneal injection (10 mg/kg, i.p.) of LTCCs attenuated seeking behavior. The data are expressed as means ± SEM, n = 4. ** p < 0.05, verapamil treatment vs. no treatment. (B) Intracerebroventricular injection (100 ug, i.v.c) of LTCCs attenuated seeking behavior. The data are expressed as means ± SEM, n = 3. * p < 0.05, verapamil treatment vs. no treatment.

Figure S3 The effect of knockdown Cav1.3 in the nucleus accumbens (NAc) on cocaine seeking behavior after a long-term withdrawal. (A) The Cav1.3 shRNA sequences used to target CACNA1D and the sequence was as follows: GTTGTCCTGAAC TCCATTATTCTCGAGAATAATGGAGTTCAGGACAAC. shRNA Cav1.3 significantly decreased the expression of Cav1.3 compared with scrambled shRNA. n = 3, ** p < 0.01, scrambled shRNA vs. Cav1.3 shRNA. (B) Knockdown of Cav1.3 in the NAc had no effect on the cocaine seeking behavior on WD45, n = 4. Data are expressed as means ± SEM.

Figure S4 Administration of shRNA Cav1.2 does not affect the locomotor activity of animals. Either shRNA Cav1.2 (n = 5) or scrambled shRNA (n = 5) was injected in the NAc of tree shrews. Seven days later, the locomotor activity was detected in an operant chamber. There was no difference in locomotor activity between two groups (unpaired t-test, t = 0.47, p = 0.65). Data are expressed as means ± SEM.

Videos information. We uploaded two videos showing two tree shrews in the extinction test after 45 days of withdrawal. One tree shrew was injected with LTCCs antagonist verapamil in the NAc 10 min before the test, and another tree shrew was injected with saline in the NAc 10 min before the test. In these videos, the tree shrew injected with saline responded to the active nose-poke twice in 30 sec. In contrast, the tree shrew injected with verapamil did not respond to the active nose-poke, but the locomotor activity was not affected by the verapamil injection.