Fig 1.
Optogenetic inhibition of D2R-neurons selectively promoted WM maintenance and retrieval under low cognitive load.
A Left: Schematic representation of virus injection and optic fiber implantation in DMS of Adora2a-Cre (+) mice. Right: Representative images showing ArchT (green) and DAPI (blue) in DMS, as well as projections to GPe in green. Scale bar, 100 μm. B Representative images of c-Fos induction following optogenetic inhibition of DMS D2R-neurons. C Quantitative analysis revealed a significant increase in c-Fos expression following photoinhibition of D2R-neurons, primarily non-merged with the virus (Independent-Sample t test, **p < 0.01). D Schematic of the experimental design for light stimulation during the “sample” phase. Photoinhibition of DMS D2R-neurons during the “sample” phase did not significantly affect WM under both low (E) and high cognitive loads (F, repeated measures [RM] two-way ANOVA, ns P > 0.05; n = 11 for both groups). G Schematic of the experimental design for light stimulation during the “delay” phase. Photoinhibition of DMS D2R-neurons during the 10 s “delay” phase resulted in a significant improvement in WM performance under low cognitive load (H, RM two-way ANOVA, main effect, F1, 16 = 4.914, #p < 0.05; interaction, F4, 64 = 0.2553, p > 0.05; GFP, n = 9; ArchT, n = 11), but not under high cognitive load (I, RM 2-way ANOVA, ns p > 0.05). J Schematic of the experimental design for light stimulation during the “choice” phase. Photoinhibition of DMS D2R-neurons during the “choice” phase significantly improved WM performance under low cognitive load (K, RM two-way ANOVA, main effect, F1, 18 = 4.853, #p < 0.05; interaction, F5, 90 = 0.4557, p > 0.05; Fisher’s LSD post-hoc comparisons, Day 3, *p < 0.05, Day 4, *p < 0.05; n = 10 for both groups), but not under high cognitive load (L, RM 2-way ANOVA, ns p > 0.05). The data underlying panel C can be found in S1 Data. Data are represented as mean ± SEM.
Fig 2.
Optogenetic activation of anterior but not posterior D2R-neurons impaired WM under low cognitive load.
A Left: Schematic illustrating virus injection and optic fiber implantation in DMS of Adora2a-Cre (+) mice. Right: Representative images showing ChR2 (red) and DAPI (blue) expression in DMS. B Photo-stimulation of ChR2 led to notable c-Fos induction in DMS (green). Scale bar, 100 μm. C Quantitative analysis revealed a significant increase in c-Fos expression due to photoactivation of D2R-neurons, predominantly merged with the virus (Independent Samples t test, **p < 0.01). D Schematic of the experimental design for light stimulation during the “delay” phase. E During the 10 s “delay” phase, photoactivation of anterior DMS (AP: +0.98 mm) D2R-neurons significantly impaired WM performance under low cognitive load (RM two-way ANOVA, main effect, F1, 22 = 8.79, ###p < 0.001; interaction, F5, 110 = 1.635, p > 0.05; Bonferroni’s post-hoc comparisons, Day 3, **p < 0.01; mCherry, n = 13; ChR2, n = 11). F Under high cognitive loads (delay 20, 30, 60 s), photoactivation of anterior DMS (AP: +0.98 mm) D2R-neurons didn’t significantly affect WM performance (RM two-way ANOVA, ns p > 0.05; mCherry, n = 13, ChR2, n = 11). Photoactivation of relatively posterior DMS (AP: +0.50 mm) D2R-neurons during the “delay” phase didn’t affect WM performance under both low (H, two-way RM ANOVA, ns P > 0.05) and high cognitive loads (I, two-way RM ANOVA, ns P > 0.05; n = 10 for both groups). The data underlying panel C can be found in S1 Data. Data are represented as mean ± SEM.
Fig 3.
Optogenetic inhibition of D1R-neurons selectively improved WM maintenance and retrieval under higher cognitive loads.
A Top: Schematic illustrating virus injection and optic fiber implantation in DMS of Drd1-Cre (+) mice. Bottom: Representative images showing ArchT expression (green), DAPI staining (blue) in DMS, and its projections to the GPi and SNR. B Photoinhibition of ArchT resulted in c-Fos induction in DMS (red). Scale bar, 100 μm. C Quantitative analysis demonstrated a significant increase in c-Fos expression due to photoinhibition of D1R-neurons, primarily in cells not co-localized with the virus (Independent Samples t test, **p < 0.01). D Schematic of the experimental design for light stimulation during the “sample” phase. Photoinhibition of DMS D1R-neurons during the “sample” phase did not significantly affect WM under both low (E) and high cognitive loads (F, RM two-way ANOVA, ns P > 0.05; GFP = 10, ArchT = 11). G Schematic of the experimental design for light stimulation during the “delay” phase. Photoinhibition of DMS D1R-neurons during the 10 s “delay” phase did not significantly impact WM performance under low cognitive load (H, RM two-way ANOVA, ns p > 0.05; GFP = 12, ArchT = 13), but significantly improved WM performance under high cognitive load (I, RM two-way ANOVA, main effect, F1, 30 = 14.95, ###p < 0.001; interaction, F3, 69 = 3.690, p < 0.05; Bonferroni’s post-hoc comparisons, Delay 60s, ****p < 0.0001). J Schematic of the experimental design for light stimulation during the “choice” phase. Photoinhibition of DMS D1R-neurons during the “choice” phase did not significantly impact WM performance under low cognitive load (K, RM two-way ANOVA, ns p > 0.05; n = 9 for both groups), but significantly improved WM performance under high cognitive load (L, RM two-way ANOVA, main effect, F1, 16 = 9.066, ##p < 0.01; interaction, F3, 48 = 1.536, p > 0.05; Bonferroni’s post-hoc comparisons, Delay 60s, **p < 0.01). The data underlying panel C can be found in S1 Data. Data are represented as mean ± SEM.
Fig 4.
Optogenetic activation of D1R-neurons selectively impaired WM maintenance and retrieval under higher cognitive loads.
A Top: Schematic of virus injection and optic fiber implantation in DMS of Drd1-Cre (+) mice. Bottom: Representative images of ChR2 expression (red) in DMS, with DAPI staining (blue) and projections in SNR. B Photo-stimulation of ChR2 induced c-Fos increase in DMS (green). Scale bar, 100 μm. C Quantitative analysis showed significant c-Fos increase due to photoactivation of DMS D1R-neurons, which was mostly merged with the virus (Independent Samples T test, **p < 0.01, ***p < 0.001). D Schematic of the experimental design for light stimulation during the “delay” phase. E During the 10 s “delay” phase, photoactivation of DMS D1R-neurons didn’t significantly affect WM performance under low cognitive load (RM two-way ANOVA, ns p > 0.05; n = 10 for both groups). F As cognitive loads increased, photoactivation of DMS D1R-neurons during the “delay” phase significantly impaired WM performance (RM two-way ANOVA, ### p < 0.001; Bonferroni’s post-hoc comparisons, delay 30 s, *p < 0.05, delay 60 s, ****p < 0.0001). G Without stimulation, there were no significant differences in WM performance (Independent Samples T test, ns p > 0.05). H Top: Schematic of optic fiber implantation in DMS of Drd1a-Cre/Ai32 mice. Bottom: Representative images of ChR2 expression (green) in DMS, with DAPI staining (blue) and projections into GPi and SNR. I Photo-stimulation of ChR2 induced c-Fos expression in DMS (red). Scale bar, 100 μm. J Quantitative analysis revealed a significant increase in c-Fos expression due to photoactivation of DMS D1R-neurons (Independent Samples T test, *p < 0.05). K Schematic of the experimental design for light stimulation during the “sample”, “delay” or “choice” phase. L No light was delivered during the task acquisition stage (Day 1-6). M Under the 10 s delay condition (low cognitive load), mild photoactivation (~1 mW) of DMS D1R-neurons during the “sample”, “delay” or “choice” phases did not significantly impact WM performance. N Under the 60 s delay condition (high cognitive load), mild photoactivation during the “delay” and “choice” phases markedly impaired WM performance, whereas activation during the “sample” phase had no effect (Independent Samples t test, “Delay light”: *p < 0.05, “Choice light”: **p < 0.01; ChR2 (−) = 9, ChR2 (+) = 8). The data underlying panels C, G, J, M, and N can be found in S1 Data. Data are represented as mean ± SEM.
Fig 5.
Optogenetic inhibition of D2R-, D1R-neurons enhanced signal detection sensitivity under low and high cognitive load, respectively, without affecting motivational and motor states in the operant DNMTP task.
A Schematic representation of the operant DNMTP task. B Schematic of the experimental design for light stimulation during the operant DNMTP task. C–E During the “sample” phase, photoinhibition of DMS D2R-neurons didn’t significantly impact the percentages of correct responses (C), SI (D) and RI (E, RM two-way ANOVA, p > 0.05). F–H During the “delay” phase, photoinhibition of DMS D2R-neurons significantly improved the percentage of correct responses (F) and SI (G), particularly at 2 and 5 s delays (F: RM two-way ANOVA, main effect, F1, 54 = 4.255, #p < 0.05; interaction, F2, 54 = 3.024, p = 0.0569; Fisher’s LSD post-hoc comparisons, Delay 2 s, *p < 0.05, Delay 5 s, *p < 0.05; G: RM two-way ANOVA, main effect, F1, 54 = 5.091, #p < 0.05; interaction, F2, 54 = 1.434, p > 0.05; Fisher’s LSD post-hoc comparisons, Delay 2 s, *p < 0.05), but didn’t significantly affect RI (H). I–K During the “choice” phase, photoinhibition of DMS D2R-neurons significantly enhanced the percentage of correct responses (I, RM two-way ANOVA, main effect, F1, 54 = 10.72, ##p < 0.01; interaction, F2, 54 = 0.4727, p > 0.05; Fisher’s LSD post-hoc comparisons, Delay 2 s, *p = 0.0504, Delay 5 s, *p < 0.05) and SI (J, RM two-way ANOVA, main effect, F1, 54 = 8.527, ##p < 0.01; interaction, F2, 54 = 0.7768, p > 0.05; Fisher’s LSD post-hoc comparisons, Delay 5 s, *p < 0.05), particularly notable at 2 and 5 s delays. However, there were no significant changes in RI (K). Sample sizes: GFP, n = 8, ArchT, n = 9. L–N During the “sample” phase, photoinhibition of DMS D1R-neurons had no significant effects on the percentage of correct responses (L), SI (M), and RI (N). O–Q During the “delay” phase, photoinhibition of DMS D1R-neurons significantly improved the percentage of correct responses (O, Independent T test, delay 10 s, *p < 0.05) and SI (P, Independent T test, Delay 10 s, *p < 0.05) of WM during the 10 s delay. However, photoinhibition didn’t affect RI (Q). R–T During the “choice” phase, photoinhibition of DMS D1R-neurons significantly improved the percentage of correct responses (R, Independent T test, delay 10 s, *p < 0.05) and SI (S, Independent T test, Delay 10 s, *p < 0.05) under the 10 s delay condition. Photoinhibition of DMS D1R-neurons during the “choice” phase didn’t impact RI (T). Sample sizes: ArchT(+), n = 9, ArchT(−), n = 9. Data are presented as mean ± SEM.
Fig 6.
Optogenetic inhibition of DMS dopamine signaling rescue WM impairment induced by D1R-neuron activation.
A Schematic illustrating the viral injection of DIO-ChR2-EYFP or DIO-EYFP, along with the implantation of optic fiber in DMS, and the injection of mTH-ArchT-mCherry or mTH-mCherry, with optic fiber implantation in VTA/SNc of Drd1-Cre (+) mice. B Left: Representative images displaying ChR2 expression (green), ArchT fibers from VTA/SNc (red), alongside TH staining (magenta) and DAPI (blue) in DMS; Right: Representative images in VTA/SNc showing ArchT expression (red), ChR2 projections in SNR (green), TH staining (magenta) and DAPI (blue) in VTA/SNc. ArchT was selectively expressed in TH-positive dopaminergic neurons. C During the 60 s “delay” phase, photoactivation of DMS D1R-neurons with 470 nm laser severely impaired WM performance (Group 1 ‘DMS EYFP + VTA/SNc mCherry’ vs. Group 3 ‘DMS ChR2 + VTA/SNc ArchT’: RM two-way ANOVA, main effect, F1, 14 = 15.92, ##p < 0.01; interaction, F3, 24 = 0.9179, p > 0.05; Fisher’s LSD post-hoc comparisons, Day 2, *p < 0.05, Day 3, **p < 0.01; Group 1 ‘DMS EYFP + VTA/SNc mCherry’ vs. Group 2 ‘DMS ChR2 + VTA/SNc mCherry’: RM two-way ANOVA, main effect, F1, 16 = 20.22, ###p < 0.001; interaction, F3, 48 = 0.5375, p > 0.05; Fisher’s LSD post-hoc comparisons, Day 2, *p < 0.05, Day 3, **p < 0.01, Day 4, *p < 0.05). D During the “delay” phase, photoinhibition of dopaminergic neurons in VTA/SNc (red prismatic) with 560 nm laser effectively rescued D1R-neuron activation-induced WM impairment, yielding performance comparable to the control group (black circle) (Group 2 ‘DMS ChR2 + VTA/SNc mCherry’ vs. Group 3 ‘DMS ChR2 + VTA/SNc ArchT’: RM two-way ANOVA, main effect, F1, 14 = 5.024, #p < 0.05; interaction, F2, 48 = 1.485, p > 0.05; Fisher’s LSD post-hoc comparisons, Day 7, **p < 0.01). E In Group 3, dual stimulation significantly improved WM performance compared to single stimulation (red column). Additionally, it demonstrated notable enhancements over Group 2, with no significant difference with Group 1 (One-way ANOVA for comparisons among the three groups; Paired-samples t test for group 3’s comparisons between single and dual stimulation; Group 1, n = 9, Group 2, n = 9, Group 3, n = 7). F Without interventions, WM performance among the three groups was comparable, showing no significant differences. G The viral injection strategy was similar to that described in panel A. Twin-core optic fibers were implanted into DMS to activate D1R-neurons and inhibit DMS dopaminergic terminals, either respectively or simultaneously. H Left: Representative images showing ChR2 expression (green), ArchT fibers from VTA/SNc (red), TH staining (magenta), and DAPI (blue) in DMS; Right: Representative images VTA/SNc showing ArchT expression (red), ChR2 fibers in SNR from DMS (green), TH staining (magenta), and DAPI (blue). ArchT was selectively expressed in TH-positive dopaminergic neurons VTA/SNc. I WM performances of the three groups were comparable without manipulations during the learning stage (Day 1–4). J During the 60 s “delay” phase, photoactivation of DMS D1R-neurons with 470 nm laser severely impaired WM performance (Group 1 ‘DMS EYFP + VTA/SNc mCherry’ vs. Group 3 ‘DMS ChR2 + VTA/SNc ArchT’: RM two-way ANOVA, main effect, F1, 56 = 14.20, ## p < 0.01; interaction, F3, 56 = 1.123, p > 0.05; Fisher’s LSD post-hoc comparisons, Day 6, **p < 0.01, Day 7, *p < 0.05). K During the “delay” phase, photoinhibition of DMS dopaminergic terminals with 560 nm laser (red prismatic) successfully rescued WM impairment induced by D1R-neuron activation, yielding performance comparable to the control group (black circle) (Group 3 vs. Group 1: RM two-way ANOVA, main effect, F1, 14 = 4.751, p > 0.05, interaction, F2, 28 = 0.2166, p > 0.05; Group 3 vs. Group 2 ‘DMS ChR2 + VTA/SNc mCherry’: RM two-way ANOVA, main effect, F1, 42 = 11.41, ## p < 0.01, interaction, F2, 42 = 2.926, p > 0.05, Fisher’s LSD post-hoc comparisons, Day 10, **p < 0.01, Day 11, *p < 0.05; Group 1 vs. Group 2: RM two-way ANOVA, main effect, F1, 36 = 32.49, #### p < 0.0001, interaction, F2, 36 = 3.513, p < 0.05, Bonferroni’s post-hoc comparisons, Day 10, **** p < 0.0001, Day 11, ** p < 0.01). L In Group 3, dual stimulation resulted in a significant enhancement of WM performance compared to single stimulation (red column). Group 3 demonstrated notable improvement over Group 2, with no significant differences compared to Group 1 (One-way ANOVA for the comparisons of the three groups; Paired-samples t test for Group 3’s comparisons between single stimulation and dual stimulation; Group 1, n = 7, Group 2, n = 7, Group 3, n = 9). M Without manipulations, no significant differences in performance were observed among the three groups (One-way ANOVA, ns, p > 0.05). N Photoinhibition of dopaminergic terminals in DMS alone had no significant impact on WM performance (Independent Samples t test, p > 0.05). The data underlying panels E, F, L, M, and N can be found in S1 Data. Data are presented as mean ± SEM.