Autoimmunity Profiling of Vaccine-Induced Childhood Narcolepsy

By: Anna Häggmark1, Ida Hossar1, Mohsen Khademi2, Tomas Olsson2, Markku Partinen3, Mathias Uhlén3, Jochen Schwenk1 and Peter Nilsson1, Poster - The Human Protein Atlas
1Science for Life Laboratory, Dept. of Proteomics, KTH - Royal Institute of Technology, Stockholm, Sweden
2Neuroimmunology Unit, Department of Clinical Neuroscience, Karolinska Hospital, Stockholm Sweden
3Department of Vaccines and Immune Protection, National Institute for Health and Welfare, Helsinki, Finland

Download the poster (2.1 Mb PDF)

The incidence of childhood narcolepsy drastically increased in Sweden and Finland following the H1N1 influenza vaccination in 2009. Narcolepsy is regarded as a rare sleeping disorder with yet unknown cause and the death of a specific type of neuron has lead to the hypothesis that autoimmunity mechanisms could be involved.

Here we describe a method for autoimmunity profiling of serum samples in an undirected manner. In total, 235 serum samples collected from narcoleptic patients and controls were analyzed on planar antigen arrays. This was conducted first as an initial screening with few samples and a large number of antigens followed by an extended screening of fewer antigens but on a larger number of samples.

Workflow for autoimmunity profiling on planar arrays

Figure 1. Workflow for autoimmunity profiling on planar arrays. Antigens were printed on planar arrays in blocks of 384. Serum samples were diluted and applied to the array and detection of reactivities facilitated by a fluorescently labelled anti-human IgG antibody. Measuring of the fluorescence intensities generated sample reactivity profilies for which sample specific cut-off values based on sample median + 10xMAD were applied to identify reactive antigens.


Serum samples from both Swedish and Finnish patients were analyzed in this study. The Finnish cohort consisted of serum from 40 narcolepsy patients and 18 controls while the Swedish cohort contained plasma from 59 narcolepsy patients and 118 controls. The control samples included both patients with other neurological diseases (OND) as well as healthy controls.

Antigen arrays

The antigens used in this study were protein fragments of approximately 100 amino acids, selected based on the sequence having low homolgy to other human proteins. The arrays are routinely produced within the Human Protein Atlas project as a tool for verification of antibody specificities.

Study design

Figure 2. Study design. An initial screening was performed of 14 narcolepsy samples on 3840 antigens followed by a verification of 221 samples on 768 antigens. A bead-based platform enabling a higher throughput of samples will be used to verify the interesting findings from the screening.


Individual heterogeniety in extended screening

Figure 3. Individual heterogeniety in extended screening. An overview of the results showed that (A) the majority of reactive antigens were found in single samples while very few were common between more than 10 samples and (B) that for the majority of samples, 5-10 out of 768 antigens were identified as reactive but the samples ranged from 1 to 90 reactive antigens.

Overview of reactivities in the extended screening

Figure 4. Overview of reactivities in the extended screening. The figure shows a heatmap of identified reactivities in all samples and for antigens found to be reactive in 10 or more samples. At this early stage of data analysis, no clear disease-specific pattern was apparent in the comparison of narcolepsy patients (red) and controls (green). Even though the individual heterogeneity was very large, a few antigens showed reactivity in more than 40 samples.


  • An autoimmunity screening was performed of narcolepsy and control serum samples on planar antigen arrays
  • Sample specific cut-off values were applied to identify reactive antigens
  • The results indicate a large individual heterogeniety in the autoantibody repertoire with the majority of antigens showing reactivity in only single samples
  • Comparing reactivities among narcolepsy patients and controls so far revealed no clear disease-specific pattern
  • Five antigens showed reactivity in >40 samples, these targets will be selected for further analysis together with those that provide a differentiating pattern between sample groups
  • More extensive data analysis is planned based on available clinical information
  • A bead based assay enabling analysis of a larger number of samples will be applied for verification of interesting findings
Related Links