Background/Purpose: We recently described the phenomenon in systemic lupus erythematosus (SLE) in which mitochondria are extruded into the extracellular space during formation of neutrophil extracellular traps (NETs). NET formation has been demonstrated to be inflammatory and immunogenic, possibly including mitochondria. While poorly characterized, it is known that in both SLE and rheumatoid arthritis (RA) a small subset of patients are positive for anti-mitochondrial antibodies (AMAs). Frustratingly, current mitochondrial isolation techniques produce mitochondria that are wrapped in nuclear DNA (nuDNA), blurring the distinction of AMAs and anti-dsDNA antibodies, commonly seen in SLE. Here, we aim to perfect mitochondrial isolation, characterize the targets of AMAs, and determine their clinical implications in RA and SLE.

Methods: Mitochondria from HepG2 cells isolated by Dounce homogenization and subsequent DNase treatment were tested for nuDNA contamination by quantitative PCR (qPCR) of 16S rRNA and 18S rRNA genes and organelle purity by fluorescence-activated cell sorting (FACS). Isolated mitochondria were used to test sera from healthy individuals (HC, n=17), SLE patients (n=44), and RA patients (n=101) for anti-mitochondrial IgG antibodies via FACS and western blot (WB) methods. Finally, sera were tested for presence of AMA by immunocytochemistry using a HepG2 cell line with GFP-tagged mitochondria.

Results: Our mitochondrial isolation technique yielded over a 1,000-fold reduction in nuDNA over standard techniques, reducing the possibility of anti-dsDNA antibodies binding to residual nuDNA on the mitochondrial surface. Further, we did not observe any binding of propidium iodide, an impermeable DNA-binding dye, or anti-dsDNA antibodies, to the isolated mitochondria. Similarly, this technique produced particles that were 99.1% positive for mitochondria markers MitoView and MitoTracker. Using this ultrapure isolate, we found that SLE and RA patients had significantly elevated reactivity against isolated mitochondria compared to HC (p<0.0001, 727 vs 333 MFI and 680 vs 400 MFI, respectively). Levels of AMAs were particularly elevated in patients with severe disease, including erosive disease (RA, p=0.01) and nephritis (SLE, p=0.02). AMA reactivity was confirmed with ICC using HepG2 cells with GFP-tagged mitochondria. Lastly, we found mitochondrial proteins near band sizes 15, 30, 35, 45, and 75kDa that were consistently labeled by SLE sera, indicating an unequal antigenicity between mitochondrial proteins.

Conclusion: Here, we introduce a highly effective method of mitochondrial isolation in which both nuDNA and non-mitochondrial organelle contamination are reduced to insignificant levels compared to previous methods. Subsets of SLE and RA patients have been reported to have AMAs, but they have not been well-characterized. In this study, we show a tremendous increase in AMAs in SLE and RA patients and we illuminate that these AMAs target only a subset of mitochondrial proteins. Little is known about AMA targets, and it is expected that their identification will open new avenues for SLE and RA monitoring, stratification, treatment and prevention.

« Back to 2018 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/highly-elevated-levels-of-anti-mitochondrial-antibodies-in-systemic-lupus-erythematosus-and-rheumatoid-arthritis/