Reduced Vglut2/Slc17a6 Gene Expression Levels throughout the Mouse Subthalamic Nucleus Cause Cell Loss and Structural Disorganization Followed by Increased Motor Activity and Decreased Sugar Consumption

Strong overlap of Vglut2 and Pitx2 in the STN and reduction of Vglut2 mRNA in cKO mice. A, Representative examples of cell morphology and distribution. Overview of the STN (left) as well as close-up images of the dorsal (right top), ventral (right center), and medial (right bottom). Cells are stained blue; small dots represent silver grains bound to Vglut2 mRNA. B, Double fluorescent in situ hybridization for Vglut2 (red) and Pitx2 (green) as well as nuclear stain with DAPI (blue) in the dorsal (top), ventral (center), and medial (bottom) STN. C, High-resolution in situ hybridization analysis of Vglut2 mRNA-based silver grain intensity shows a generally lowered expression of Vglut2 in the STN of cKO mice.

Decreased consumption of sugar in STN-VGLUT2 cKO mice. A, Schematic self-administration schedule. Black lines indicate the number of days during which each paradigm was carried out. B, Operant self-administration was carried out in modular self-administration boxes with two feeders, one of which delivered pellets upon head entry (south, green = active), and the other did not (north, red = inactive). Yellow: house light. Black bars, head entry holes measuring exploration with no feeder attached. When a mouse made a head entry at the active feeder, a sugar reward was delivered, and simultaneously, light and sound cues were presented to confirm the choice (left); a head entry at the inactive feeder produced neither reward delivery nor a light or sound cue (right). C, During task learning, both controls (white circles) and cKOs (filled circles) decreased time to obtain the maximum of 30 sugar pellets. D, During FR1, the time to consume 30 pellets was significantly higher for cKO (filled circles) compared with control (white circles; left panel). All mice were able to obtain the maximum of 30 pellets during the FR1 protocol (middle panel). The amount of head entries during the inactive time (timeout) decreased in cKO and increased in control (right). E, During FR5, head entries during active time (left), the amount of pellets obtained (middle) and head entries during inactive time (right) were lower for cKO compared with controls. F, During PR, no difference was seen in head entries during the active (left) or the inactive (timeout, right) time between control and cKO mice. G, Cognitive ability testing. During reinstatement, the mice were presented to the original task after an extinction period (left). During reversal, the positions of the active and inactive feeders were switched (right). H, To allow testing of the retention of the task (reinstatement), both cKO and control groups were put through extinction. For five consecutive days, the active feeder delivered both light and sound cues, but no sugar pellets. For both groups, the amount of head entries strongly decreased during both the active time (left) and timeout (right). I, No differences between control and cKO mice were seen during reinstatement or reversal. Statistical analysis of the self-administration data was performed using repeated-measures ANOVA followed by a post hoc test with Bonferroni correction. ^# or *p ≤ 0.05; ^## or **p ≤ 0.01; ^### or ***p ≤ 0.001, ^#### or **** p ≤ 0.0001. Hab., habituation; Rev., reversal; RI, reinstatement.

The average weight of STN-VGLUT2 cKO mice is decreased, whereas home cage food consumption is unaltered. A, Control mice (white circles) had a higher body weight than cKO mice (filled circles) over the course of the experiment. B, Refeeding after self-administration was not significantly different between controls and cKOs. Statistical analysis of weight and food intake data were analyzed using repeated-measures ANOVA followed by a post hoc test with Bonferroni correction. *p ≤ 0.05; **p ≤ 0.001.

Increased seated and free rearing in STN-VGLUT2 cKO mice. Three rearing types were observed in STN-VGLUT2 cKO mice: wall rearing, where the mouse takes support against a vertical surface (A, left); seated rearing, where the mouse supports itself on the tail base (B, left); and free rearing, where the mouse stretches its hind legs and uses no other support (C, left). All rearing types were scored in an open field arena. Both groups were habituated to the environment (# marks the change over time). cKOs show increased seated (B) and free (C) rearing throughout the session (¤ marks the difference between genotypes and * marks the difference at a certain time point between genotypes). Rearing data was analyzed by repeated-measures ANOVA and post hoc test with Bonferroni correction. *p ≤ 0.05; **p ≤ 0.001; ***p ≤ 0.0001.

The availability of dopamine receptor D2 and D3 as well as DAT binding sites in the ventral striatum is regulated by Vglut2 reduction in the STN. A, Representative examples of serial coronal striatal sections analyzed for DAT, D1R, D2R, and D3R (from top to bottom). B, Comparison of specific binding capacity levels expressed as percentage of control for DAT-specific [¹²⁵I]RTI binding, D1R-specific [³H]SCH23390, D2R-specific [¹²⁵I]iodosulpride, and D3R-specific [¹²⁵I]7-OH-PIPAT binding in control and cKO mice. DAT binding was unaltered in the NAcC and upregulated in the NAcSh of cKO mice (left). The amount of D1 receptor binding sites (middle left) in both NAcC and NAcSh remained unaltered, whereas more D2 receptor binding sites were measured in both NAcC and NAcSh (middle right) and fewer D3 receptor binding sites were available in NAcSh (right). Data analyzed by Mann–Whitney U-test. *p ≤ 0.05.

Elevated levels of dynorphin peptides in the NAc of STN-VGLUT2 cKO mice. A, MALDI analysis was performed on cryosections at the level of the DStr and NAc. Each slice was coated with droplets of matrix for ionization of the peptide fragments and analyzed by MALDI time-of-flight. B, A bright-field picture of the cresyl violet–stained section analyzed for aNeo was used to determine the localization of DStr and NAc (black outlines). Representative examples of aNeo (depicted in red) in cKO. Overlays with peptide fragment m/z 1835 and 1393 (middle and right, depicted in green) visualized the fiber bundles used as landmarks for the anatomical localization of the NAc. C, Representative examples of slices used for MALDI imaging in control (left) and cKO (right). aNeo and DynA peptide levels were significantly elevated in cKO and showed a more restricted localization to the NAc. D, Fluorescence intensity of different peptides measured by MALDI imaging, shown as percentage of control (fold change, see table). E, Example peaks of DynA (top) and aNeo (bottom) from MALDI analysis. F, Oligo in situ hybridization analysis of Dyn expression in the forebrain of control (left) and cKO (right) mice. FC: fold change; numbers in brackets indicate which amino acids constitute each peptide fragment. *p ≤ 0.05 and **p ≤ 0.01; error given as SEM.

Optogenetic tracing of Vglut2-Pitx2 coexpressing neurons revealed a projection pattern restricted to the EP and SNr. A–C, Schematic illustrations of injection procedure. A, Mice were stereotactically injected with AAV-ChR2-EYFP into the STN; B, The injection was performed bilaterally but one side at a time to infect the left- and righthand side STN at equal levels; C, Two close coordinates were used for each STN to allow even spread of the virus (–4.25 and –4.75 dorsoventricular). D, Histologic results of injections. AAV-ChR2-EYFP–positive cell bodies in the STN. E, AAV-ChR2-EYFP–positive fibers projecting to EP (left), GP (middle), and SNr (right). Upper row, EYFP expression; bottom row, bright-field photographs of the same area. PSTN: parasubthalamic nucleus.