Pemphigus vulgaris autoantibody profiling by proteomic technique

Abstract

Pemphigus vulgaris (PV) is a mucocutaneous blistering disease characterized by IgG autoantibodies against the stratified squamous epithelium. Current understanding of PV pathophysiology does not explain the mechanism of acantholysis in patients lacking desmoglein antibodies, which justifies a search for novel targets of pemphigus autoimmunity. We tested 264 pemphigus and 138 normal control sera on the multiplexed protein array platform containing 701 human genes encompassing many known keratinocyte cell-surface molecules and members of protein families targeted by organ-non-specific PV antibodies. The top 10 antigens recognized by the majority of test patients' sera were proteins encoded by the DSC1, DSC3, ATP2C1, PKP3, CHRM3, COL21A1, ANXA8L1, CD88 and CHRNE genes. The most common combinations of target antigens included at least one of the adhesion molecules DSC1, DSC3 or PKP3 and/or the acetylcholine receptor CHRM3 or CHRNE with or without the MHC class II antigen DRA. To identify the PV antibodies most specific to the disease process, we sorted the data based on the ratio of patient to control frequencies of antigen recognition. The frequency of antigen recognition by patients that exceeded that of control by 10 and more times were the molecules encoded by the CD33, GP1BA, CHRND, SLC36A4, CD1B, CD32, CDH8, CDH9, PMP22 and HLA-E genes as well as mitochondrial proteins encoded by the NDUFS1, CYB5B, SOD2, PDHA1 and FH genes. The highest specificity to PV showed combinations of autoantibodies to the calcium pump encoded by ATP2C1 with C5a receptor plus DSC1 or DSC3 or HLA-DRA. The results identified new targets of pemphigus autoimmunity. Novel autoantibody signatures may help explain individual variations in disease severity and treatment response, and serve as sensitive and specific biomarkers for new diagnostic assays in PV patients.

Figure 1. Composition of the multiplexed protein array platform.

Figure 2. Principal component analysis of top antigens.

Unsupervised principal component analysis of the signal intensity for samples from PV patients and healthy controls revealed that these two groups could be segregated on the basis of top 30 antigens with the highest sensitivity (A) and 15 antigens with the highest specificity (B) for PV.

Figure 3. Sensitivity of the microarray in detecting disease-related autoantibodies.

Sorted by percent of positive samples in the group of antigens recognized by 10 percent and more of PV patients. Inset: 25 top antigens.

Figure 4. Specificity of the microarray in revealing disease-specific PV antibodies.

Sorted by ratio of positive patient/control samples in the group of antigens recognized by two and more time frequently by PV patient than control sera. Inset: 25 top antigens.

Figure 5. Combinations of top 10 most common individual antigens targeted by PV antibodies.

A, Top 25 combinations sorted by percent of positive samples. B, Top 25 antigen sorted by the patient/control ratio.