A novel C5a-neutralizing mirror-image (l-)aptamer prevents organ failure and improves survival in experimental sepsis

Abstract

Complement factor C5a is a potent proinflammatory mediator that contributes to the pathogenesis of numerous inflammatory diseases. Here, we describe the discovery of NOX-D20, a PEGylated biostable mirror-image mixed (l-)RNA/DNA aptamer (Spiegelmer) that binds to mouse and human C5a with picomolar affinity. In vitro, NOX-D20 inhibited C5a-induced chemotaxis of a CD88-expressing cell line and efficiently antagonized the activation of primary human polymorphonuclear leukocytes (PMN) by C5a. Binding of NOX-D20 to the C5a moiety of human C5 did not interfere with the formation of the terminal membrane attack complex (MAC). In sepsis, for which a specific interventional therapy is currently lacking, complement activation and elevated levels of C5a are suggested to contribute to multiorgan failure and mortality. In the model of polymicrobial sepsis induced by cecal ligation and puncture (CLP), NOX-D20 attenuated inflammation and organ damage, prevented the breakdown of the vascular endothelial barrier, and improved survival. Our study suggests NOX-D20 as a new therapeutic candidate for the treatment of sepsis.

Figure 1

Identification of bio-d-mC5a binding aptamers. (a) Schematic overview of the discovery process. (b) Competitive binding assay for aptamer truncation. [³²P]-labeled aptamer 274-D5 (83 nt) was incubated with bio-d-mC5a in the presence of unlabeled competitor aptamers 274-D5, 274-D5-001 (48 nt), and 274-D5-002 (44 nt) at indicated concentrations. (c) Secondary structure of 274-D5 as predicted by free energy minimization (ViennaRNA). Primer binding sites are in lower case. (d) Competitive binding assay for sequence optimization. [³²P]-labeled aptamer 274-D5-002 was incubated with bio-d-mC5a in the presence of unlabeled competitor aptamers 274-D5-002, 274-C5-002, 274-C8-002, and the composite aptamer 274-C8-002-G14 at indicated concentrations.

Figure 2

Characterization and postselection optimization of C5a-binding Spiegelmer. SPR measurement of NOX-D19 binding to (a) mouse and (b) human C5a. Kinetic rate constants k_a and k_d are shown as mean ± SEM. Data are representative for at least 3 individual measurements. (c) Kinetic rate constants of 2′-deoxyribonucleotide-modified NOX-D19001 variants binding to huC5a were determined by SPR measurement. For unmodified NOX-D19001 (black diamond; dotted lines) mean ± SD of 5 injections is shown. Six modified variants of NOX-D19001 (D09, D16, D17, D30, D32, and D40) with increased overall affinity (K_d = k_d/k_a) were chosen and combined to the six-times modified Spiegelmer NOX-D19001-6xDNA (black square). (d) Inhibition by NOX-D19 (black diamonds) and NOX-D20 (black squares) of CD88⁺ BA/F3 cell chemotaxis stimulated with 0.1 nmol/l huC5a. Mean ± SD of triplicate measurements is shown. Data are representative for four independent experiments. RU, response units.

Figure 3

NOX-D20 binds to C5 but does not inhibit complement-mediated hemolysis. SPR measurement of NOX-D20 binding to human (a) C5a, (b) C5a(desArg), and (c) C5. Kinetic rate constants k_a and k_d are shown as mean ± SEM. Data are representative for at least three individual measurements. (d) Human serum pretreated with NOX-D20 (black squares) or C5-binding aptamer C5C6 (black triangles) was incubated with opsonized sheep erythrocytes. Hemolysis was quantified by photometric measurement of hemoglobin in the supernatant at 405 nm. Normalized data representative for three independent experiments is shown. RU, response units.

Figure 4

NOX-D20 improves survival in CLP-induced polymicrobial sepsis. Mice (n = 9–10 per group) were treated with daily i.p. injections of vehicle (black squares), 1 mg/kg NOX-D20 (black triangles) or 3 mg/kg NOX-D20 (open triangles) for 7 days. One group of mice received a single i.p. dose of 1 mg/kg NOX-D20 after surgery followed by daily vehicle injections (gray circles). Sham operated mice (n = 5) receiving daily vehicle injections were used as controls (black diamonds). (a) Survival was monitored daily and log-rank test was performed for statistical analysis. At day 1 (b) serum creatinine, (c) blood urea nitrogen (BUN), and (d) serum alanine aminotransferase (ALT) levels were determined (n = 9–10 for vehicle and 3 mg/kg NOX-D20; n = 20 for 1 mg/kg NOX-D20; n = 5 for sham). (e) Serum lactate dehydrogenase (LDH) levels were determined 18 hours after CLP surgery and vehicle or NOX-D20 injection (n = 7–10 per group). (f) Mice were i.v. injected with 0.25% w/v Evans blue after CLP surgery and vehicle or NOX-D20 injection (n = 8–10 per group). Ratio of Evans blue in peritoneal lavage (PL) and serum after 18 hours is shown as a measure for capillary leakage. Means ± SEM are shown. For statistical analysis, one-way analysis of variance and Dunnett's comparison was performed (*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001).

Figure 5

NOX-D20 suppresses proinflammatory mediator release in CLP-induced sepsis. Cytokine and chemokine concentrations in (a) peritoneal lavage (PL) and (b) serum were determined 18 hours after CLP surgery and NOX-D20 or vehicle injection. Mean ± SEM for n = 7–10 mice per group is shown. For statistical analysis, one-way analysis of variance and Dunnett's comparison was performed on log-transformed values (*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001).

Figure 6

NOX-D20 suppresses PMN recruitment and activation. Cell numbers of (a) PMN and (b) monocytes were determined in the peritoneal lavage by cytospin 18 hours after CLP surgery and NOX-D20 or vehicle injection. Mean ± SEM of n = 8–10 mice is shown. For statistical analysis, one-way analysis of variance and Dunnett's comparison was performed. (c,d) Primary human PMN were stimulated in vitro with huC5a. Inhibition of (c) PMN chemotaxis and (d) elastase release by NOX-D20 is shown. Mean ± SEM of triplicate measurement of PMN from three individual donors is shown. For statistical analysis, one-way analysis of variance and Dunnett's comparison was performed on log-transformed values (*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001). PMN, polymorphonuclear leukocytes.

Figure 7

Pharmacokinetics study. NMRI mice were i.p. injected with a single dose of 1 mg/kg or 10 mg/kg NOX-D20. Blood was drawn at indicated time points and NOX-D20 concentrations were determined by SPR measurement. Mean ± SEM of four mice per group and time point is shown. Linear regression of log-transformed plasma concentrations at 3–48 hours and 4–28 days is shown. Lower limit of quantification was 0.79 nmol/l NOX-D20. Plasma C5 levels were determined by ELISA (duplicate measurement) in plasma of mice (n = 4) treated with vehicle 10 minutes and 24 hours before blood sampling. Mean C5 plasma concentration is shown as solid line. Gray area indicates 95% CI.