Subject selection and clinical assessments
The study was approved by the regional ethics committee (Yorkshire & The Humber, UK). Written informed consent was obtained from all patients. Patients were recruited from the Ataxia and Hepatology outpatient clinics at the Royal Hallamshire Hospital, Sheffield, UK. Patients included were grouped as ‘alcohol ataxia (AA)’ and had a history of chronic ‘alcohol dependence’, defined by the 4th edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) Diagnostic Criteria for Alcohol Abuse and Dependence [13]. A subgroup of patients with alcohol-related ‘chronic liver disease’ (defined as alcohol related liver disease diagnosed by a hepatologist and disease present for more than 1 year) and ataxia (CLDA) was identified from this patient group.
Detailed neurological history and any background history of autoimmune disorders were recorded, as were age at onset, duration of ataxia, requirement for mobility aids, and current alcohol intake or abstinence at the time of recruitment.
Detailed neurological examination was performed. Ataxia was classified as gait, limb or both and severity was assessed as mild (mobilising independently or with one walking aid), moderate (mobilising with 2 walking aids or walking frame) or severe (wheelchair-dependent). The severity assessment was adapted and modified from previously published data [14]. Objective measurement of the severity of ataxia was rated using the Scale for the Assessment and Rating of Ataxia (SARA) [15, 16] (see Additional file 1). All patients were investigated for other causes of ataxia and were excluded if an alternative cause was found. Tests depending on clinical indications included blood cell counts, biochemistry, thyroid function, B12, folate and genetic testing for inherited spinocerebellar ataxias (SCA 1, 2, 3, 6, 7) and Friedreich’s ataxia (FA).
Brain imaging
Volumetric 3T MR imaging and single-voxel H1 MR spectroscopy of the cerebellum were undertaken in patients with clinical evidence of ataxia. This imaging protocol is in clinical use as part of the investigations of all patients with cerebellar ataxia who attend the National Ataxia Centre in the Royal Hallamshire Hospital, Sheffield, UK. The brain imaging protocol for structural, volumetric and spectroscopy studies have been previously reported [12].
The vermis was assessed as a whole for volumetric analysis. A manual approach was adopted for volumetric analysis of the vermis to avoid partial volume averaging effects that may also be affected by the degree of atrophy potentially confounding VBM results (unpublished data). For vermian volumetric studies, the midline vermis was identified on sagittal T1 volume images with reference to the midsagittal plane of the cerebellum [17]. Based on this reference, manual measurement of cross sectional areas of the whole vermis was possible. Vermian volume (V) in mm3 was measured as a sum of vermis cross sectional areas (mm2) multiplied by 0.9 mm thickness. Vermian volume was expressed as a percentage of the total intracranial volume (%V:TIV).
Healthy controls recruited, were age and gender matched with patient subjects and had undergone the same MR imaging protocol. The details of the healthy controls have also been previously reported [18].
Blood collection and serological tests
Blood samples were collected at recruitment and the serum was aliquoted to avoid repeat freeze-thawing and stored at −20 ° C in sealed tubes.
All patients had total immunoglobulin levels, IgA and IgG anti-gliadin antibodies (AGA), anti-endomysial antibodies (EMA) and IgA anti-transglutaminase 2 (TG2) antibodies, measured by standard laboratory protocols at the Immunology Department, Northern General Hospital, Sheffield. Briefly, the measurement for AGA and TG2 were performed by enzyme linked immunosorbent assay (ELISA) (Aesku. Diagnostics-Grifols, Germany). EMA was tested by indirect immunofluorescence on slides containing monkey oesophagus tissue (Inova-Instrumentation Laboratories, USA). Patient sera were also used for the detection of IgA and IgG to transglutaminase 6 (TG6) by ELISA as previously described [19].
Human Leukocyte Antigen (HLA) typing was performed at the National Blood Service, Sheffield, UK.
Immunohistochemistry
Rat cerebellar tissue was acquired from adult Sprague-Dawley rats (Biological Services, Faculty of Medicine, Dentistry & Health, University of Sheffield). The cerebellar tissue was mounted using Cryo-M-Bed tissue embedding polymer (Bright, UK) and snap-frozen with isopentane/cooled on liquid nitrogen. The cerebellar tissue was cryo-sectioned into 10 μm sagittal sections that were collected on polysine-coated slides (Thermo Scientific, UK), stored at −80 ° C in an airtight container.
Mouse anti-calbindin-D-28K monoclonal antibody (Sigma, UK) diluted 1:200 in PBS was used to visualize Purkinje cells. Patient sera were diluted 1:200 or 1:600 in PBS before incubation with sections for 1 h. 1:600 was identified as the optimum dilution in determining the reactivity of patient sera on rat cerebellar tissue in agreement with our previous study [20]. Negative controls included sections incubated without patient sera. A horseradish peroxidase-conjugated goat anti-human or goat anti-mouse IgG antibody (Jackson ImmunoResearch Laboratories, USA) was used as secondary antibodies.
Images were captured at X100 and X200 magnification. Two blinded observers performed the evaluation of the Purkinje cell staining intensity on rat cerebellar sections independently. Staining was classed as ‘weak’ or ‘strong’ if Purkinje cell staining was above background levels. ‘Negative’ was recorded if staining did not exceed background levels. Staining of other neuronal cell populations (granular layer) was also noted. Concordance rate between assessors was 79 %.
Statistical analysis
Statistical analysis was performed using PRISM 6 software package (GraphPad Software Inc.). Demographic, clinical and imaging characteristics are presented as means with standard deviations (mean ± SD). The Independent-Samples Mann-Whitney U Test was used to determine any difference between mean % cerebellar volume (CBV) : total intracranial volume (TIV) and mean % vermian volume (V) : total intracranial volume (TIV) between patients and controls. The χ2 test was used for comparing the prevalence of anti-gliadin antibodies and anti-transglutaminase 6 antibodies in the study group with that of the healthy population; and between the subgroups. Results were considered statistically significant for p < 0.05.