Positron Emission Tomography

e-Methods

Positron Emission Tomography

A static PET acquisition was started 40 min after intravenous injection of 250MBq [F-18]-fluorodeoxyglucose (FDG) according to common guidelines (25). Transaxial images were reconstructed by an iterative reconstruction algorithm using the manufacturer default parameters. A low-dose CT acquired immediately before the PET was used for attenuation correction.

First and second follow-up PET images were co-registered to baseline PET using the Coregister-tool of the Statistical Parametric Mapping software package (SPM2) (17). The images were stereotactically normalized to a custom made FDG-PET template in the Montreal Neurological Institute (MNI) space using SPM´s Normalize-tool with standard parameter settings (26). Optimal transformation parameters for normalization were obtained for the baseline PET and applied to baseline and follow-up PETs. Voxel intensities of the normalized images were scaled using the proportional scaling approach implemented in SPM (27).

In preparation of voxel-based testing, stereotactically normalized PET images were smoothed with an isotropic 3-dimensional Gaussian kernel with 10mm full-width-at-half-maximum. Each of the 3 FDG-PET images was compared with corresponding images of a group of 28 normal control subjects on a voxel-by-voxel basis using one-sided two sample t-tests. The significance level was set at  = 0.001 uncorrected for multiple comparisons, which is widely used in single subject analysis of brain FDG-PET. Only effects in clusters of at least 125 voxels corresponding to a volume of 1ml were considered. For direct visualization of changes between consecutive PETs (baseline to first and second follow-up), voxel-based subtraction analysis was performed (Fig. 4E) as described previously (18).

MRI-based volumetry

High-resolution T1-weighted MR images were segmented and stereotactically normalized to the Montreal Neurological Institute (MNI) space using a combined segmentation and registration approach (DARTEL) (SPM8 software package, Wellcome Trust Centre for Neuroimaging, London, UK) (19). Prior tissue probability maps for grey matter (GM), white matter (WM) and cerebrospinal fluid (CSF) generated from 662 healthy elderly subjects aged 63-75 years were used (20). DARTEL provides stereotactically normalized probability maps of GM, WM and CSF for the individual MRI.

Total volume of GM, WM and CSF was obtained by summing over all voxel intensities (multiplied by the voxel volume) in the patient’s GM, WM and CSF probability map. Predefined volumes of interest (VOIs) were used for determination of regional GM volume. VOIs for the frontal, parietal, occipital, and temporal lobe were obtained from a digital lobe atlas in MNI space (20), separately for both hemispheres. VOIs for the ventricles were derived by thresholding the prior CSF probability map. The hippocampus VOI was taken from the SPM Anatomy toolbox (21). GM volume within a given VOI was obtained by summing over all voxel intensities within the given VOI.

Electrophysiological recordings

Coverslips with hippocampal cells exposed to human control serum or anti-NMDAR IgA serum of the index patient (dilution 1:200, 3 days) were transferred to a submerged recording chamber re-perfused with artificial cerebrospinal fluid (ACSF) containing (in mM) NaCl 119, KCl 2.5, MgCl2 1.3, CaCl2 2.5, glucose 10, NaH2PO4 1.0, NaHCO3 26, gassed with carbogen (95 % O2 / 5% CO2; pH 7.4; 290-310 mosmol/l). Recordings were done at 31-32°C using a Multiclamp-700A amplifier (Axon Instruments, Union City, USA). Cells were identified using infrared differential interference-contrast (IR-DIC) video microscopy. Borosilicate glass electrodes (2-5MΩ) were filled with (in mM) K-gluconate 120, KCl 10, Hepes 10, Mg-ATP 3, EGTA 5, MgSO4 2, GTP 1; pH adjusted to 7.2 with KOH. In the whole-cell, current-clamp configuration, de- and hyperpolarizing current steps (100 ms) were applied to characterize the cell’s intrinsic properties. Series resistance Rs was monitored continuously throughout experiments (measured in voltage-clamp configuration); cells were rejected if Rs was >25MΩ or varied >±30% during recordings. No Rs compensation was used. Cellular potentials indicated are liquid-junction potential-corrected (calculated ~14mV). The reversal potential of chloride (-67mV) was determined applying the Nernst equation based on the extra- and intracellular concentrations in our experiments (129.1mM /10mM, respectively). Experiments using laser-induced photolysis of ‘caged’ glutamate to isolate NMDA receptor-mediated currents were performed in the presence of 20µM NBQX (2,3-dioxo-6-nitro-1,2,3,4-tetrahydrobenzo(f)quinoxaline-7-sulfonamide, Sigma Aldrich, Germany) and 1 µM gabazine (6-Imino-3-(4-methoxyphenyl)-1(6H)-pyridazinebutanoic acid hydrobromide, Biotrend, Germany). Cells were voltage-clamped at -50mV. We used 20-40ml of ACSF containing 100μM MNI-caged-l-glutamate (Tocris, Bristol, UK). The maximum time period of recirculation was 4 h. Laser flash duration was 2ms. With a 0.9NA 60×-objective, the optical system achieved an effective light spot diameter of 15μm in the focal plane which was targeted directly on the soma of the cell under investigation.

Signal processing and data analysis

Signals were filtered (4 or 8 kHz, Bessel), digitized with 16-bit resolution (National Instruments, Austin, USA) and sampled at 10 or 20 kHz using Igor Pro (Wavemetrics, Lake Oswego, USA).

Detection of spontaneous EPSCs was done using custom-made code in Matlab (The Mathworks, Natick, MA). Briefly, data were low-pass filtered (2 kHz), offset-corrected and down-sampled to 5 kHz. sEPSC onsets were identified using the detection algorithm described by Kudoh and Taguchi (28). Mean sEPSC frequency was calculated by dividing event number and observation time. To quantify sEPSC amplitudes we determined the negative peak following the onset subtracted by the onset amplitude. For analyzing L-glutamate uncaging-induced NMDAR currents we determined grand average signal including all experiments. The time point at grand average maximum negativity was used as reference time point to quantify amplitudes of NMDAR currents in individual experiments.

All data are reported as means ± SEM. Tests on statistical significance of reported differences were as indicated and involved the Wilcoxon ranksum test for comparison of nonparametric data and two-sample Kolmogorov-Smirnov (K-S) test for comparison of distributions. Significance level α was set to 0.05.