All human CSF samples were collected by lumbar puncture and used under protocols approved by the Institutional Review Boards of the Karolinska Institute and Stanford University.
Label
Not applicable.
Label protocol
Not applicable.
Hybridization protocol
Lipid arrays were produced using a CAMAG automatic TLC sampler robot (ATS4) adapted to print lipids onto PVDF membranes (Invitrogen, CA) taped onto the surface of Gold Seal Microslides microscope slides (Redding, CA) using double-sided scotch tape (3M). The lipids are solubilized in chloroform-methanol-water mixture. 200 nl of each lipid, at a range of 10 to 100 pmol, is sprayed onto each slide under nitrogen gas to form individual features. Printed membranes are stored dry and can retain reactivity for several months. Slides are then blocked at 4°C overnight (1 X phosphate-buffered saline (PBS), 1% Fatty Acid Free Bovine Serum Albumin (FAF-BSA)) and incubated with individual human CSF diluted at 1:10 in PBS/1% FAF-BSA for 2h at 4°C on a 40 rpm shaker. After washing twice with PBS/1% FAF-BSA, 10 min each time on a 45 rpm shaker at RT, reactive antibodies were detected using HRP-conjugated goat-anti-human IgG/IgM secondary antibody (1:8,000 dilution in PBS/1% FAF-BSA, Jackson Immunoresearch). After incubating with the secondary antibody for 60 minutes at 4°C on a 40 rpm shaker, the slides were washed on a 45 rpm shaker at RT, twice with PBS/1% FAF-BSA, 30 min each time; then twice with 1 X PBS, 15 min each time; then rinsed quickly twice with distilled water.
Scan protocol
The slides were incubated for 3 min in chemiluminescence solution (ECL Plus, Amersham) at room temperature before autoradiograph development. Digital images were then produced by scanning array autoradiographs.
Description
Antibody in CSF. We used a CAMAG Automatic TLC Sampler (ATS4) robot to spray 200nl of 10 to 100 pmol of lipids onto PVDF membranes affixed to the surface of microscope slides.
Data processing
GenePix Pro 5.0 software (Molecular Devices, Sunnyvale, CA) was used to extract the net median pixel intensities for individual lipid features from the digital images. The data were not normalized. The mean net digital chemiluminescence units were then generated as the mean of the F450 Median – B450 values from two identical lipid features on each array. A "secondary antibody only" slide was included to subtract out non-specific binding. Data analysis was performed using Significance Analysis for Microarrays (SAM) software, version 1.21, (http://www-stat-class. stanford.edu /SAM/servlet/ SAMServlet) to identify antigen features with statistically significant differences in reactivities between the patient groups. Cluster software was then used to hierarchically group the samples and antigen features on the basis of a pairwise similarity function, and TreeView software was used to display the data as a heat map (http://rana.lbl.gov/EisenSoftware.htm).
