The lectin pathway of complement activation is a critical component of the innate immune response to pneumococcal infection

Abstract

The complement system plays a key role in host defense against pneumococcal infection. Three different pathways, the classical, alternative and lectin pathways, mediate complement activation. While there is limited information available on the roles of the classical and the alternative activation pathways of complement in fighting streptococcal infection, little is known about the role of the lectin pathway, mainly due to the lack of appropriate experimental models of lectin pathway deficiency. We have recently established a mouse strain deficient of the lectin pathway effector enzyme mannan-binding lectin associated serine protease-2 (MASP-2) and shown that this mouse strain is unable to form the lectin pathway specific C3 and C5 convertases. Here we report that MASP-2 deficient mice (which can still activate complement via the classical pathway and the alternative pathway) are highly susceptible to pneumococcal infection and fail to opsonize Streptococcus pneumoniae in the none-immune host. This defect in complement opsonisation severely compromises pathogen clearance in the lectin pathway deficient host. Using sera from mice and humans with defined complement deficiencies, we demonstrate that mouse ficolin A, human L-ficolin, and collectin 11 in both species, but not mannan-binding lectin (MBL), are the pattern recognition molecules that drive lectin pathway activation on the surface of S. pneumoniae. We further show that pneumococcal opsonisation via the lectin pathway can proceed in the absence of C4. This study corroborates the essential function of MASP-2 in the lectin pathway and highlights the importance of MBL-independent lectin pathway activation in the host defense against pneumococci.

Figure 1. MASP-2 is essential for C3 deposition on S pneumoniae.

Serial dilutions of sera were incubated in microtiter plates coated with S. pneumoniae D39 or N-acetylated BSA (as a control), and C3 deposition determined by ELISA. A. Shows the raw data from one experiment with WT and MASP-2 deficient murine serum (means ±SEM). In B, the experiment was extended to include murine sera deficient in other complement components. Results are duplicates (±SD) and are normalised to the C3 deposition observed in the WT control. C. Time course of C3 activation on S. pneumoniae. 1∶40 diluted murine sera were incubated in microtiter plates coated with S. pneumoniae for the times indicated then C3 deposition assayed. Results are means of duplicates and are representative of three independent experiments. D. FACS analysis of C3 deposition on S. pneumoniae opsonised with MASP-2 −/− serum (black line), WT serum (red) and C4 −/− serum (green). The blue trace shows non-opsonised bacteria. E. Results from 3 independent FACS analyses of C3 deposition (mean fluorescent intensity ±SEM). F. Inhibition of MASP-2 activity with mAb AbD04211 abolished C3b deposition on S. pneumoniae opsonised with C4 deficient serum. (means of triplicates ±SEM; p value from Student's t-test). G. In human serum, MBL deficiency had no effect on C3 deposition, while C4 deficiency had no significant effect on the EC₅₀, but reduces absolute C3b deposition by about 50% (means ±SEM; n = 3 for NHS and MBL −/− serum, 1 for C4 deficient serum). H. Correlation between L-ficolin serum concentration and C3b deposition on S. pneumoniae immobilised on microtiter plates for 47 samples of NHS (solid line shows Fisher transformation of Pearson's correlation coefficient; dashed lines, 95% CI thereof).

Figure 2. Deposition of C4 breakdown products on S. pneumoniae.

ELSIA plates were coated with S. pneumoniae D39, E. coli or mannan (as positive controls) and incubated with NHS for 1 h at 37°C. Bound C4 was detected using an antibody against C4c (which also detects C4b) (A) or an antibody against C4dg, the last breakdown product of C4 that remains covalently attached to the activating surface (B). C and D: Detection of C4c (C) and C4dg (D) on live S. pneumoniae D39 opsonised with NHS. Green trace shows anti-C4 antibodies; the purple shading, an isotype control Ab.

Figure 3. Binding of lectin pathway recognition molecules to S. pneumoniae.

Microtiter plates coated with formalin-fixed S. pneumoniae or control substrates were used to capture lectin pathway recognition complexes from WT mouse serum or NHS. Murine MBL-A and MBL-C (A), murine ficolin A (B), murine CL-11 (C) human MBL (D), FCN1 & 3 (E) and FCN2 and CL-K1 (F) were assayed by ELISA, as described in materials and methods. Only ficolin A, CL-11 and FCN2 (L-ficolin) bound to the bacteria. Results are means of duplicates and are representative of three independent experiments.

Figure 4. MASP-2 deficiency impairs phagocytosis of S. pneumoniae by polymorphonuclear leukocytes.

S. pneumoniae were pre-incubated with murine serum (20% v/v), mixed with freshly isolated human PMN and incubated for 2 hr before being immobilised on microscope slides and stained with eosin Y and azur II (REASTAIN), as described in materials and methods. Bacteria opsonised with WT serum (A) are internalized by PMNs, whereas bacteria opsonised with MASP-2 −/− serum (B) are excluded (arrows). C. Electron micrograph showing pneumococci opsonised with WT serum inside a PMN. D. Samples were removed from the S. pneumoniae/PMN mix at the times indicated and viable bacteria determined. WT serum (crosses) facilitated killing by PMN, whereas MASP-2 −/− serum (open squares) is severely compromised in its ability to opsonise S. pneumoniae. Controls were run in parallel containing non-opsonised bacteria (triangles) and bacteria opsonised with WT serum, but without PMN (circles). Results are means (±SEM) of triplicates.

Figure 5. Lectin pathway deficiency significantly increases the severity of S. pneumoniae infection.

Lectin pathway deficient mice (open squares) and WT littermates (crosses) were infected intranasally with 1×10⁶ cfu S. pneumoniae D39. C57/BL6 ^{Masp2 −/−} (A) and C57/BL6 ^{Fcna −/−} (B) mice are impaired in their survival of S. pneumoniae infection (Mantel-cox log-rank test). C–F. Viable S. pneumoniae counted in lung homogenates and peripheral blood at the indicated time points after infection. Results in C–E are means (±SEM) of five animals sacrificed at each time point. *p<0.05; **p<0.01 (ANOVA).

Figure 6. Inhibition of MASP-2 increases the severity of S. pneumoniae infection.

WT C57/BL6 mice were dosed i.p. with 1.0 mg/kg body weight of anti-MASP-2 mAb AbD04211 (open triangles) or an isotype control mAb (crosses) 12 h before being infected with 1×10⁶ cfu S. pneumoniae D39. MASP-2 inhibition significantly worsened survival (A) and there was a corresponding increase in viable S. pneumoniae counted in peripheral blood after infection (B). Treatment with 20 mg/kg body weight ceftriaxone 12 hr before infection and every 12 hr thereafter afforded complete protection from S. pneumoniae in animals injected with AbD04211 (filled triangles in A). Results in B are means (±SEM) of twelve animals from which blood was taken at each time point. *p<0.05 (ANOVA).