Abstract Text: Objective To explore genetic association between anti-NMDA-receptor encephalitis (anti-NMDARE) and human leukocyte antigen (HLA), killer cell immunoglobulin-like receptors (KIR) and other loci.
Methods GWAS in 511 (non-post herpetic) cases and 2,827 PCA-matched controls and allelic KIR and HLA sequencing in 449 cases and 1,577 ethnically matched controls.
Results GWAS significant loci include ACP2/NR1H3 and DMXL2 genes. ACP2/NR1H3 locus is in high LD with LRP4 and MADD (rs11039155; p=10-13; OR=2.1). The lead variant is a strong expression quantitative trait (eQTL) for all of these genes (p < 10-10) across various tissues. The second locus is in the vicinity of DMXL2 on chromosome 15 (lead variant rs76144151; p=1.1x10-8; OR=1.58) and regulates expression of DMXL2 in various tissues. Of note, LRRK1 is not significant. These results suggest involvement of the rab3-associated synaptic release or lysosomal acidification pathways. Strikingly, there is a weak, but significant signal in HLA class II after controlling for population stratification, suggesting predisposing effects of DQA1*01:01~DQB1*05:01 (OR=2.2). Using KIR sequencing, we found independent associations with specific KIR2DL4 and KIR3DL3 alleles (OR=1.98 and OR=4.29, respectively), loci interacting with potential checkpoint inhibitor genes HLA-G and HHLA respectively. Primary expression of these KIR genes in CD56bright NK-cells and decreased CD56bright NK-cell numbers in patient vs control suggest involvement of NK cells in anti-NMDARE.
Conclusion This study is the, so far, largest genetic study on anti-NMDARE. Our results suggest involvement of innate immunity, HLA and KIR in this disease. Follow-up on the functional roles of these loci and validation in a larger cohort is needed.
