Background Although diffuse alveolar hemorrhage (DAH) is an uncommon complication of SLE, over half of patients die. B6 mice with pristane-induced lupus and DAH are an animal model of this disorder. ANCA-negative small vessel vasculitis with hemorrhage and hemosiderin-laden macrophages are seen in lung tissue from both SLE patients and mice with DAH. DAH in pristane-lupus is prevented by depleting macrophages whereas neutrophil depletion has no effect. IL-10 deficiency exacerbates DAH while macrophage repolarization induced by liver X receptors (LXR) prevents it. The myxomavirus-encoded serpin Serp-1 impairs macrophage activation and plasminogen activation and blocks DAH caused by MHV68 infection. We asked whether it also could block DAH in pristane-induced lupus.

Methods B6 mice were treated with pristane ± daily injection of recombinant Serp-1. Severity of DAH was assessed at day-14 by gross pathology, H&E, and Prussian blue staining. LXR activation was assessed by measuring Nr1h3 mRNA (encoding LXRα) and flow cytometry for the LXRregulated ATP Binding Cassette Subfamily A Member 1 (ABCA1) protein. The effect of Serp-1 on macrophage polarization was investigated in lung tissue and in RAW264.7 cells.

Results DAH was prevented by recombinant Serp-1. Serp-1 treatment repolarized macrophages toward an anti-inflammatory M2-like phenotype, increased expression of the transcription factor.

Kruppel-like factor-4 (Klf4), which regulates IL-10 production, and increased expression of Nr1h3, which along with Klf4 regulates M2 polarization. Serp-1 also corrected a lupusassociated deficit of LXR-regulated reverse cholesterol transporter protein ABCA1. In RAW264.7 cells, Serp-1 increased Klf4 mRNA levels and LPS-stimulated IL-10 secretion while reducing TNFα. Although Serp-1 affects both thrombotic and thrombolytic pathways, the induction of DAH by pristane was unaffected by the absence of plasminogen activator inhibitor1 or tissue plasminogen activator, suggesting that protection is related to the action of Serp-1 on macrophage function.

Conclusions Serp-1 blocks pristane-induced lung hemorrhage by enhancing LXR-regulated M2 macrophage polarization and Klf4-regulated IL-10 production. In view of the similarities between DAH in pristane-treated mice and SLE patients, clinical trials of Serp-1 for DAH in SLE may be warranted. Since Serp-1 treatment increases expression of the reverse cholesterol transporter ABCA1, it also may have beneficial effects on atherosclerosis in SLE patients. Supporting the feasibility of future clinical studies in SLE, Serp-1 treatment reduced myocardial damage in patients with acute coronary syndrome.

Acknowledgments This work was supported by NIH grant R01-AR44731. We are grateful to Dr. Alexandra Lucas (Arizona State University) for providing recombinant Serp-1 protein.