Exploration of potential immune mechanisms in cervical dystonia

Abstract

Background: Although there are many possible causes for cervical dystonia (CD), a specific etiology cannot be identified in most cases. Prior studies have suggested a relationship between autoimmune disease and some cases of CD, pointing to possible immunological mechanisms.

Objective: The goal was to explore the potential role of multiple different immunological mechanisms in CD.

Methods: First, a broad screening test compared neuronal antibodies in controls and CD. Second, unbiased blood plasma proteomics provided a broad screen for potential biologic differences between controls and CD. Third, a multiplex immunoassay compared 37 markers associated with immunological processes in controls and CD. Fourth, relative immune cell frequencies were investigated in blood samples of controls and CD. Finally, sequencing studies investigated the association of HLA DQB1 and DRB1 alleles in controls versus CD.

Results: Screens for anti-neuronal antibodies did not reveal any obvious abnormalities. Plasma proteomics pointed towards certain abnormalities of immune mechanisms, and the multiplex assay pointed more specifically towards abnormalities in T lymphocytes. Abnormal immune cell frequencies were identified for some CD cases, and these cases clustered together as a potential subgroup. Studies of HLA alleles indicated a possible association between CD and DRB1*15:03, which is reported to mediate the penetrance of autoimmune disorders.

Conclusions: Altogether, the association of CD with multiple different blood-based immune measures point to abnormalities in cell-mediated immunity that may play a pathogenic role for a subgroup of individuals with CD.

Keywords: Autoimmune; Cervical dystonia; Dystonia.

Fig. 1.

Shotgun proteomics. Panel A is a table of proteins that were significantly different at two different thresholds (p and FDR both <0.10) for 3 comparisons: controls vs all CD cases combined, controls vs CD without thyroid disease (CD−T), and controls vs CD with thyroid disease (CD + T). Panel B is a Venn diagram showing individual proteins significantly different at p < 0.05, and overlapping proteins in CD + T and CD−T groups. Proteins in bold were increased, while those in italics were decreased (p < 0.05). Since the regression analysis did not show any effect of thyroid disease, these groups were combined for the volcano plot in Panel C. The nominal p-value is shown on the Y-axis and the fold change (FC) is shown on the X-axis as a log2 transform to mitigate the effect of outliers. Proteins that were abnormal relative to controls are shown in red. The three proteins in this group with FDR<0.10 are shown as boxplots in Panel D. Although there was no significant effect of thyroid disease, those with (red) and without (blue) thyroid disease are shown in different colors. Abbreviations: F13B, Coagulation factor XIII B chain; CFHR2, Complement factor H-related protein 2; C2, Complement C2; F12, Coagulation factor XII; HPX, Hemopexin; HPR, Haptoglobin-related protein; HABP2, Hyaluronan-binding protein 2; FETUB, Fetuin-B; VCAM1, Vascular cell adhesion protein 1; PLXDC2, Plexin domain-containing protein 2; DBH, Dopamine-β-hydroxylase; CD44, CD44 antigen; F13A1, Coagulation factor XIII A chain; HGFAC, Hepatocyte growth factor activator; CLEC3B, Tetranectin; F9, Coagulation factor IX; SAA, Serum amyloid A protein; APCS, Serum amyloid P-component; LUM, Lumican; TNXB, Tenascin-X; TLN1, Talin-1; AFM, Afamin; C1QA, Complement C1q subcomponent subunit A; C1QB, Complement C1q subcomponent subunit B; CFP, Properdin; CFH, Complement factor H; PCOLCE, Procollagen C-endopeptidase enhancer 1; AGT, Angiotensinogen; FGA, Fibrinogen alpha chain; SERPINA1, Alpha-1-antitrypsin; CRP, C-reactive protein.

Fig. 2.

Multiplex assay for immunological and inflammatory markers. The boxplots in Panel A show the 6 analytes reaching statistical significance at p < 0.05 before or after considering coincidental thyroid disease, with normal controls in gray, subjects who had CD without thyroid disease in blue, and CD with thyroid disease in red. Panel B is a schematic that shows the major sources (left column) and targets (right column) of these 6 analytes. Analytes in bold were increased, while those in italics were decreased. The right column shows the target of the analytes, and whether the analyte is activating (solid arrow) or inhibitory (hatched arrow).

Fig. 3.

Immune cell frequencies are altered in CD cases. Histograms of B-cells (A), monocytes (B), and the ratio of CD8 to CD4 T-cell frequencies (C) reveal a bimodal distribution in CD (solid line), but a normal distribution in controls (hatched line). Cell counts (gray bars) were normalized to approximate a probable density function. Best fit kernel smooth lines were calculated to illustrate the population distribution for CD (solid line) and control (hatched line) groups in R software. (D) Cluster analysis revealed 40% of cases with abnormal immune cell frequencies clustered together as a potential immune mediated subset of CD (CD-I).