Citation for Published Item: Publisher's Copyright Statement: Use Policy

The full-text may be used and/or reproduced, and given to third parties in any format or medium, without prior permission or charge, for personal research or study, educational, or not-for-prot purposes provided that: • a full bibliographic reference is made to the original source • a link is made to the metadata record in DRO • the full-text is not changed in any way The full-text must not be sold in any format or medium without the formal permission of the copyright holders. Please consult the full DRO policy for further details.

for the 3 age/groups, respectively) did not reveal any significant difference between the two genotypes analysed (Two-Way ANOVA, F (1,42) =0.016, p=0.9006 main effect of genotype). Values are means ± S.E.M. (b) Basal activity was measured in 16M old control (n=6) and Trkb PENK-KO mice (n=4) by monitoring the mice for 3 days after being individually confined in their home-cages. Basal activity was detected by sensing the body-heat image in the infrared radiation; this analysis showed no differences between the two genotypes. Values are means ±S.E.M. Figure S3. Catwalk automated gait analysis parameters related to single paws and couplings in mice at 5M of age. (a-d) No differences between genotypes were found in dynamic paw parameters including (a) swing, (b) swing speed, (c) stand and (d) stand index. (e) Smaller print width for the left hind paw (LH) in Trkb PENK-KO mice compared to controls. (f-h) Trkb PENK-KO mice showed slightly decreased timely coordination between (f) ipsilateral (RHRF) and (g) girdle paws (RHLH), and better coordination between (h) diagonal paws (RFLH) as measured by couplings. Values are means ± S.E.M.; n=5 per group, * p<0.05, two-tailed unpaired Student's ttest; RFright forepaw, RHright hindpaw, LFleft forepaw, LHleft hindpaw.  Figure S7. Stereological estimation of cell density. Image of typical sagittal section of mouse brain used for the stereological estimation of DARPP32 positive cells as an example. Cells were counted with an unbiased counting frame size 50x30 µm for DARPP32. The frames were randomly positioned by the "Unbiased counting frame" macro of ImageJ (v1.46r, NIH) with a regular distance between frames of x=300, y=280 µm. Cells in each frame were counted complying with the optical fractionator counting rules. Striatal boundaries were defined using the following morphological references: the corpus callosum, the external capsule, the lateral ventricle and the anterior commissure. Scale bar, 300 µm. CP, caudate putamen; VL, lateral ventricle; cc, corpus callosum.

Supplementary Figure S8. Representative original gels.
(a-c) Blots showed in the figures (3e, 7a and 7b, respectively) come from multiple probing of the same membrane. In order to do so, the membranes obtained upon transfer of a long mini gel were horizontally cut around the expected molecular weight of the target protein (this was determined by probing an independent whole membrane with the antibody of interest). Each membrane was cut into 3 to 4 pieces and incubated with the appropriate antibody. In the case of figure 7, pospho-specific antibodies were probed first in order to avoid any alteration of the signal that may result from the stripping procedure. Once the appropriate band was detected and analyzed membranes were stripped by low-pH stripping buffer, re-exposed to control for successful stripping, and then probed again with the non-pospho antibody. Each gel was repeated 2 to 3 times to ensure consistency of the data. Dashed lines indicate the region where the membrane was cut.

Supplementary Table S1
General gait parameters are normal in Trkb PENK-KO mice compared to controls.

Parameters related to the interlimb coordination:
Step sequence distributions: the order in which the paws are placed on the glass plate described in Supplementary Table S3. Data analysis was performed by automatically labelling all areas containing one or more pixels above a 25 pixel threshold and then assigning those areas to one of the paws. A typical run/crossing contained at least six step cycles and averaged data from all step cycles in a run were used in the analysis.

The regularity index
In-situ hybridization: RNA probes were prepared by PCR (Trkb: forward 5' TCAGCATATCAAGAGACAC, reverse 5' CTGTACACATCTCGGGACAT; Enkephalin: Allen Mouse Brain Atlas probes, Riboprobe ID: RP_060315_01_A07, http://mouse.brainmap.org). Cryosections were washed with PBS, treated with 10 μg ml −1 Proteinase K (Roche) in PBS for 6 min at 37°C, fixed in 4%(wt/vol) PFA for 10 min and washed with PBS. Acetylation was performed for 10 min in 0.25% acetic anhydride, prehybridized and then hybridized with biotinylated RNA probes (800 ng ml −1 ) at 58°C for 24-40 hours. Posthybridization washes consisted of 30 min 2x SSC at room temperature, 2 times 30 min 2x SSC at 65°C, 2 times 30 min 0.1 SSC. After blocking, sections were incubated with an alkaline phosphatase-conjugated DIG antibody (Roche) and staining was developed with BCIP/NBT. When immunofluorescence was performed, sections were refixed 30 min in 4%(wt/vol) PFA after the last PBS rinse, and staining was performed as described above. was then used to reconstruct the single neuron morphology by a semi-automated procedure.
Specifically, the method relies on an automated procedure exclusively for the reconstruction of the cell soma volume and dendritic arborization but spines were selected manually from confocal z-stacks and considered such when fully visible with clearly identifiable head. A threshold for the minimum spine head diameter was set at ≥0.143 μm in order to avoid the identification of false positive. Moreover, the branch length was >0.5 μm. The Imaris software calculated then spine density, dendritic thickness and complexity of the dendritic arborization. These parameters were chosen according to previously published procedure based on the same software 58,59 . For the analysis of punctae images of double immunofluorescent staining from 3 Trkb PENK-WT and 3 Trkb PENK-KO mice obtained with specific antibodies were analysed by confocal microscopy. Punctae on each parvalbumin stained cell were then selected when present at the cell surface with a minimum diameter of 0.2μm for gephyrin and 0.5 μm for GAD67. For each animal 8 to 10 cells were acquired.
The punctae density was then calculated by Imaris software according to previously published work 58,59 .

cFos-ir and punctate labelling quantification: Counts of cFos-ir cells in the CPu and LGP
were performed on 5 to 6 non-consecutive medial to lateral sagittal sections (spanning around 600 μm).
LGP location was determined based on the expression of the marker parvalbumin that is first of all expressed within the LGP and secondly, strongly highlights the reticular nucleus of the thalamus that dorso-laterally defines the edge of the globus pallidus. A surface of approximately 1 to 2 mm 2 for the CPu and of 0.1 to 0.4 mm 2 for the LGP was acquired by tile scan imaging from each section. The density of cFos-ir cells in the specified area was then determined by counting the total number of neurons in the area using the software Imaris (Bitplane). Adjacent sections were then used to determine the density of GAD67-and gephyrin-containing synapses on pallidosubthalamic LGP neurons. Cells were identified within the LGP by parvalbumin labeling. The parvalbumin staining was also used to determine the volume of the soma and the first tract of the primary dendrites (when visible). Punctate labeling corresponding to the surface of the parvalbumin staining was analyzed by the use of the software Imaris. The density of sphere-shaped positive area for GAD67 and gephyrin was determined. Minimum of 5 to 8 cells per mouse were analyzed (3 animals per genotype).
Membrane current was filtered at 0.1 kHz and digitized at 100-200 μs using Clampex 10.2 (gap-free mode). Cells were voltage-clamped at −70 mV to determine basic membrane properties and to examine spontaneous EPSCs. Picrotoxin (50μm) was applied to block GABA A receptor-mediated currents. Baseline activity was recorded for 3-5 min after 10-12 min in picrotoxin. For extracellular recordings borosilicate pipettes were filled with 2 M NaCl (resistance 10-15 MΩ). Amplitude of input-output responses were plotted and fit by a