Role of the property of C-reactive protein to activate the classical pathway of complement in protecting mice from pneumococcal infection

Abstract

C-reactive protein (CRP) is not an acute-phase protein in mice, and therefore, mice are widely used to investigate the functions of human CRP. It has been shown that CRP protects mice from pneumococcal infection, and an active complement system is required for full protection. In this study, we assessed the contribution of CRP's ability of activating the classical pathway of complement in the protection of mice from lethal infection with virulent Streptococcus pneumoniae type 3. We used two CRP mutants, Y175A and K114A. The Y175A CRP does not bind C1q and does not activate complement in human serum. The K114A CRP binds C1q and activates complement more efficiently than wild-type CRP. Passively administered, both CRP mutants and the wild-type CRP protected mice from infection equally. Infected mice injected with wild-type or mutant CRP had reduced bacteremia, resulting in lower mortality and increased longevity compared with mice that did not receive CRP. Thus, the protection of mice was independent of CRP-mediated activation of the classical pathway of complement. To confirm that human CRP does not differentiate between human and mouse complement, we analyzed the binding of human CRP to mouse C1q. Surprisingly, CRP did not react with mouse C1q, although both mutant and wild-type CRP activated mouse C3, indicating species specificity of CRP-C1q interaction. We conclude that the mouse is an unfit animal for exploring CRP-mediated activation of the classical complement pathway, and that the characteristic of CRP to activate the classical complement pathway has no role in protecting mice from infection.

FIGURE 1

Survival curves of mice infected with S. pneumoniae. Four groups of mice, 12–18 mice in each group, were injected with 10⁸ CFU bacteria, with or without 150 μg of WT, Y175A, or K114A CRP. CRP was injected first; bacteria were injected 30 min later. Deaths were recorded twice per day for 10 days. The data are combined from three separate experiments. Statistical analyses were performed using one-way ANOVA and Tukey's multiple comparison test. The p values for the differences among groups A/B, A/C, and A/D are <0.05. The p values for the differences among groups B/C, B/D, and C/D are >0.05. The 95% confidence intervals are: A/B, 0.17–0.94; A/C, 0.17–1.05; A/D, 0.09–0.96; B/C, 0.38– 0.49; B/D, 0.46–0.41; and C/D, 0.56–0.39.

FIGURE 2

Bacteremia in mice treated with WT, Y175A, and K114A CRP. Blood samples were collected from each surviving mouse shown in Fig. 1 for the first 5 days postinfection. Bacteremia was determined by plating. Each dot represents one mouse. The horizontal line in each group of mice represents the median value of bacteremia in that group. The p values (Mann-Whitney U test) for the differences among groups A/B, A/C, and A/D, on days 2, 3, and 4 are <0.05. The p values for the differences among groups B/C, B/D, and C/D, on days 2, 3, and 4 are >0.05.

FIGURE 3

Binding of CRP by live virulent S. pneumoniae. A fixed number of live bacteria were grown in the presence of increasing amount of CRP, as shown on the x-axis. After 4 h, CRP that bound to bacteria was eluted and measured by ELISA. The amount of bound CRP is shown on the y-axis.

FIGURE 4

Complement activation by CRP complexed to PnC as determined by the C3 deposition assay. A representative experiment is shown. A, Dose-response curves of the binding of human C3 fragments to complexes of PnC with WT and mutant CRP. Microtiter wells were coated with PnC. Increasing concentrations of purified CRP were then added to the wells. Normal human serum was added to the wells coated with CRP-PnC. Deposited C3 was detected by using anti-human C3d mAb. Values on the y-axis represent binding of anti-C3d to C3 deposited on PnC-CRP complexes. B, Dose-response curves of the binding of mouse C3 fragments to the complexes of PnC with WT and mutant CRP. Normal mouse serum was added to the wells coated with CRP-PnC. Deposited C3 was detected by using anti-mouse C3 Ab. Values on the y-axis represent binding of anti-C3 to C3 deposited on PnC-CRP complexes. C, Dose-response curves of the binding of mouse C3 fragments to the complexes of PnC with WT CRP. Assay as in B.

FIGURE 5

Binding of mouse C1q to CRP-PnC and to Ag-Ab complexes. A representative experiment is shown. A, Purified mouse C1q was added to microtiter wells coated with PnC and preincubated with CRP. Bound C1q was then detected by using an anti-mouse C1q Ab. B, As in A, except that the wells were coated with CRP and preincubated with anti-CRP to generate Ag-Ab complexes on the wells. C, As in A, except that mouse serum was used instead of purified mouse C1q. D, As in B, except that mouse serum was used instead of purified mouse C1q.