Decreased number of parvalbumin and cholinergic interneurons in the striatum of individuals with Tourette syndrome

Abstract

Corticobasal ganglia neuronal ensembles bring automatic motor skills into voluntary control and integrate them into ongoing motor behavior. A 5% decrease in caudate (Cd) nucleus volume is the most consistent structural finding in the brain of patients with Tourette syndrome (TS), but the cellular abnormalities that underlie this decrease in volume are unclear. In this study the density of different types of interneurons and medium spiny neurons (MSNs) in the striatum was assessed in the postmortem brains of 5 TS subjects as compared with normal controls (NC) by unbiased stereological analyses. TS patients demonstrated a 50%-60% decrease of both parvalbumin (PV)+ and choline acetyltransferase (ChAT)+ cholinergic interneurons in the Cd and the putamen (Pt). Cholinergic interneurons were decreased in TS patients in the associative and sensorimotor regions but not in the limbic regions of the striatum, such that the normal gradient in density of cholinergic cells (highest in associative regions, intermediate in sensorimotor and lowest in limbic regions) was abolished. No significant difference was present in the densities of medium-sized calretinin (CR)+ interneurons, MSNs, and total neurons. The selective deficit of PV+ and cholinergic striatal interneurons in TS subjects may result in an impaired cortico/thalamic control of striatal neuron firing in TS.

Figure 1

Anatomical representation of striatal regions sampled by stereological analyses. A-F, Representative images of CR-immunostained sections taken from case 4454. C’ and F’ are larger magnifications of C and F, respectively. Contours for the caudate (Cd), putamen (Pt) and nucleus accumbens (NA) hand-drawn using StereoInvestigator are superimposed on each image. These are shown for illustrative purpose only, as sampling of sections was much more frequent than shown here. The approximate level of each image in the anteroposterior axis is indicated by a dashed line. The head of the Cd and the rostral Pt were defined as the region between their anteriormost border and the anterior boundary of the globus pallidus pars externa (GPe) (shown in C, C’). The Cd body was defined as the Cd region between the anterior edge of the GPe and the posterior edge of the globus pallidus pars interna (GPi) (shown in F, F’). Scale bar = 5mm.

Figure 2

Unbiased stereological estimates of parvalbumin (PV)+ neuron densities in the caudate (Cd) (A) and the putamen (Pt) (B) of Tourette syndrome (TS) and normal control (NC) brains. Each small symbol represents a single subject. N = number of subjects. Across all regions (Cd, Pt), there was an overall strong significant decrease in PV+ neurons in TS (F (1, 8) = 14.085, p = 0.006, ANOVA), but no statistically significant difference between the regions and subregions (anterior, posterior).

Figure 3

Representative images of PV immunostaining in the head of the caudate of normal control (A, C) and Tourette syndrome (B, D) brains. Panels A and B are composites of several smaller panels. C and D are high magnifications of A and B, respectively. Scale bar in A is 100 μm for A and B; scale bar in C is 50 μm for C and D.

Figure 4

Representative images of CR immunostaining in the head of the caudate of normal control (A,C) and Tourette syndrome (B,D) brains. C and D are high magnifications of A and B, respectively. Arrows point to large-size CR+ interneurons and arrowheads point to the more abundant medium-size CR+ interneurons. Scale bar in A is 100 μm for A and B; scale bar in C is 50 μm for C and D.

Figure 5

CR+ neuron densities in the caudate (Cd) and putamen (Pt) of Tourette syndrome (TS) and normal control (NC) brains using stereological analyses. (A,C) Caudate; (B) Putamen. Each small symbol in represents a single subject. N = number of subjects. A,B: Across all regions (Cd, Pt), there was an overall strong significant decrease in large-size CR+ interneurons in TS (F (1, 8) = 36.994, p = 0.001, ANOVA), but no statistically significant difference among the regions and subregions (anterior, posterior). C: The density of medium-sized CR+ interneurons was not changed in the Cd of TS subjects. Note the different Y scales in C and A, B.

Figure 6

Representative images of ChAT immunostaining in the head of the caudate of normal control (A,C) and Tourette syndrome (B,D) brains. C and D are high magnifications of A and B, respectively. Arrows point to ChAT+ neurons. Scale bar in A is 100 μm for A and B; scale bar in C is 50 μm for C and D.

Figure 7

Unbiased stereological estimates of ChAT+ neuron densities in the striatum Tourette syndrome (TS) and normal control (NC) brains. Each small symbol represents a single subject. (A) Caudate; (B) Putamen. N = number of subjects. Across all regions (Cd, Pt), there was an overall strong significant decrease in ChAT+ interneurons in TS (F (1, 8) = 10.641, p = 0.011, ANOVA), but no statistically significant difference among the regions and subregions (anterior, posterior).

Figure 8

A: Functional subdivision of the basal ganglia in the coronal plane, based on previous studies (Bernácer et al., 2007; Morel et al., 2002; Parent and Hazrati, 1995). a-d: Drawings are shown in a rostrocaudal order. AS, associative; SM, sensorimotor; LI, limbic. B: The distribution of cholinergic interneurons density, as assessed by counts of large sized CR+ cells, in the functional territories of the striatum. Each small symbol represents a single subject. N = number of subjects. ANOVA showed an overall significant effect of diagnosis (F (1, 8) = 25.852; p= 0.001) and region (F (2, 16) = 5.335; p= 0.017) and a significant interaction of diagnosis versus region (F (2, 16) = 5.631; p= 0.014). *p<0.05; **p<0.025, post-hoc tests with Sidak adjustment.