Article Figures & Data
Figures
- Figure 1.
Reduced cerebellothalamocortical connectivity in dystonia gene carriers. The four seed masks (below and above the decussation on each side, a–d) were placed within the major outflow pathway of the cerebellum. PCTs were generated from each seed mask and compared across groups. Comparison of PCTs originating from the seed masks below the decussation of the superior cerebellar peduncle (a and b) revealed significant reductions in connectivity in the mutation carriers in the proximal segment of the CbTC pathway, near the dentate nucleus (clusters a1 and b1). PCTs generated from seed mask c above the decussation also exhibited reduced connectivity in gene carriers in this region (c1). Comparison of PCTs originating from the seed masks in the dorsal pons (c and d) revealed abnormal reductions in connectivity in the nonmanifesting carriers (clusters c2 and d1) in the subgyral white matter of the sensorimotor cortex. The size, location, and significance level of these clusters are provided in Table 1.
- Figure 2.
Reduced cerebellar outflow in dystonia gene carriers. A, Two-dimensional projections of the group mean PCTs (see Materials and Methods) generated from seed masks a and b (Fig. 1) in NL subjects (n = 8), and in NM (n = 8) and MAN (n = 12) dystonia mutation carriers. Reduced fiber tract integrity is evident in the proximal portion of the cerebellar outflow pathway in the NM and MAN groups. B, Box-and-whisker plot showing connectivity values for cluster a1 (Fig. 1) in the proximal portion of the CbTC pathway. Connectivity in this region, corresponding to dentatothalamic projections, was abnormally reduced in both MAN and NM gene carriers (p < 0.001 and p < 0.01, Mann–Whitney U tests). A significant decreasing trend (NL > NM > MAN, bold lines) was detected across the three groups (p < 0.001, J–T test of trends).
- Figure 3.
Correlations of cerebellar outflow pathway connectivity with brain activation. Voxelwise searches of the whole-brain volume (left) revealed several discrete clusters in which cerebellar connectivity correlated significantly with rCBF recorded during task performance (see Table 2). The displayed correlational maps were thresholded at p = 0.001 at peak voxel with a cluster cutoff of 30 voxels. For each significant cluster, rCBF for each subject (averaged across trial runs) was plotted (middle) against the corresponding connectivity measures. Correlations with rCBF recorded during the movement (MOVE, blue) and rest (REST, orange) conditions (see Materials and Methods) were displayed separately for each region. We also plotted (right) motor activation responses [Δ = (MOVE − REST) differences in rCBF, black] for each cluster. For both sets of plots, significant correlations (p < 0.01, Pearson product moment correlations) were indicated by bold lines and larger dots. A, Connectivity in the cerebellum correlated positively with movement-related activation in the same region (MOVE: r = 0.78, p = 0.008; REST: r = 0.55, p = 0.098; MOVE − REST: r = 0.79, p = 0.006). B, Reduced cerebellar outflow tract integrity was associated with lower resting rCBF in the VL nucleus of the thalamus (REST: r = 0.73, p = 0.017; MOVE: r = 0.54, p = 0.11; MOVE − REST: r = −0.60, p = 0.069). There were no significant correlations in the neighboring thalamic nuclei (supplemental Fig. 3, available at www.jneurosci.org as supplemental material). C, D, Cerebellar connectivity correlated negatively with motor cortical activation. The correlation was movement specific (MOVE: r = −0.85, p = 0.002; REST: r = −0.27, p = 0.44; MOVE − REST: r = −0.63, p = 0.052) in the SMC and task-independent (MOVE: r = −0.87, p = 0.001; REST: r = −0.76, p = 0.011) in the SMA.
- Figure 4.
Reduced thalamocortical connectivity in nonmanifesting dystonia gene carriers. A, Three-dimensional displays of the group mean PCTs generated from seed masks c and d (Fig. 1) in NL subjects (n = 8), and in NM (n = 8) and MAN (n = 12) dystonia mutation carriers. The size of this fiber tract is reduced distally, most pronounced in the nonmanifesting carriers. B, Box-and-whisker plots showing connectivity values for cluster c2 (Fig. 1) in the distal portion of the CbTC pathway. Connectivity in this region, corresponding to thalamocortical projections, was reduced in NM carriers relative to controls (p = 0.016) and MAN carriers (p = 0.025, Mann–Whitney U tests). A significant trend (NL > MAN > NM, bold lines) was detected across the three groups (p = 0.001, J–T test of trends).
- Figure 5.
Penetrance is determined by connectivity changes at proximal and distal sites. A, PLS-DA was used to classify individual mutation carriers as penetrant or nonpenetrant based on their measured connectivity at the four loci described above (Fig. 1, clusters a1, b1, c2, and d1). Two significant discriminant functions accurately categorized the subjects. Component 1 (x-axis) separated the dystonia gene carriers (DYT, n = 20) from NL subjects (n = 8) (p < 0.00001; Mann–Whitney U test). Component 2 discriminated MAN (n = 12) from NM (n = 8) (p = 0.002, Mann–Whitney U test). The distribution of subject scores on the two discriminant functions was similar for both genotypic subgroups. [See supplemental Fig. 4, available at www.jneurosci.org as supplemental material, for validation of the PLS-DA results and for the regional loadings of the two discriminant functions.] B, Box-and-whisker plots showing the connectivity values for the cerebellar (CER) and SMC clusters from the individual gene carriers. For each subject, z-transformed connectivity values for each region were right-left averaged and connected by lines (DYT1, orange; DYT6, blue). Subject differences in connectivity between regions (δ = CER − SMC; inset) were compared for the MAN and NM carriers. Differences in δ were found to be significant (p = 0.001, Mann–Whitney U test), with positive values for the NM group and negative values for the MAN group. C, Monte Carlo simulations predicting the penetrance rate in dystonia gene carriers based on randomly chosen connectivity values for the proximal and distal CbTC pathway segments. Different values were assumed (x-axis) for the correlation coefficient relating the two connectivity measures. (The details of this model and its assumptions are presented in Materials and Methods). A correlation coefficient of 0.56, based on the regression analysis of DYT1 carrier and control data (SMC = a · CER + b: a = 0.52, b = 0.12; r = 0.56, p = 0.01, n = 19), predicted a penetrance rate of ∼36% (see Results).
Tables
- Table 2.
Regions with significant correlations between cerebral blood flow and connectivity
Correlation Region Coordinatesa Sizeb Tmax p valuec x y z Positive Cerebellum 28 −56 −40 186 5.16 0.041* Negative SMA 6 4 58 641 6.49 <0.001* Precuneus 2 −38 52 359 5.67 0.002* DLPFC 26 28 44 227 5.48 0.018* mPFC 6 54 20 708 5.18 <0.001* SMC −30 −34 48 30 4.51 <0.001† SMC 44 −30 68 40 4.18 <0.001†
Supplemental Data
Files in this Data Supplement:
- supplemental material - Supplemental Material
