Killing of Streptococcus pneumoniae by capsular polysaccharide-specific polymeric IgA, complement, and phagocytes

Abstract

The role of IgA in the control of invasive mucosal pathogens such as Streptococcus pneumoniae is poorly understood. We demonstrate that human pneumococcal capsular polysaccharide-specific IgA initiated dose-dependent killing of S. pneumoniae with complement and phagocytes. The majority of specific IgA in serum was of the polymeric form (pIgA), and the efficiency of pIgA-initiated killing exceeded that of monomeric IgA-initiated killing. In the absence of complement, specific IgA induced minimal bacterial adherence, uptake, and killing. Killing of S. pneumoniae by resting phagocytes with immune IgA required complement, predominantly via the C2-independent alternative pathway, which requires factor B, but not calcium. Both S. pneumoniae-bound IgA and complement were involved, as demonstrated by a 50% decrease in killing with blocking of Fcalpha receptor (CD89) and CR1/CR3 (CD35/CD11b). However, IgA-mediated killing by phagocytes could be reproduced in the absence of opsonic complement by pre-activating phagocytes with the inflammatory products C5a and TNF-alpha. Thus, S. pneumoniae capsule-specific IgA may show distinct roles in effecting clearance of S. pneumoniae in the presence or absence of inflammation. These data suggest mechanisms whereby pIgA may serve to control pneumococcal infections locally and upon the pathogen's entry into the bloodstream.

Figure 1

(a) Distribution of molecular forms of IgA (mIgA and pIgA) in immune serum for type 14 PPS–specific IgA (black circles; solid lines) and total serum IgA (open circles; dotted line) by gel filtration (mean of 2 experiments). Optical density (OD) represents absorbance at 405 nm by ELISA for type 14–specific IgA and absorbance at 280 nm for total serum IgA (0.2 AUFS absorbance units, full scale); the latter represents approximately 1,000-fold more protein than the former (2 μg/mL type 14–specific IgA and 2 mg/mL total IgA). Type 14–specific IgA comprised 1.2% of pIgA and 0.15% of mIgA. Vertical lines represent peaks of pIgA and mIgA standards. (b) Nondenaturing PAGE of purified IgA before and after size fractionation by gel filtration. Lane 1: purified IgA from serum sample before separation; lane 2: IgA standard containing both pIgA and mIgA; lane 3: purified pIga fraction pool; lane 4: purified mIgA pool.

Figure 2

(a) IgA-mediated killing of type 14 S. pneumoniae by differentiated HL-60 cells and baby rabbit complement. Phagocyte/bacteria ratio was 400:1. Tests were run in duplicate; results are shown as the mean ± SEM of 3 experiments. (b) Killing of type 14 S. pneumoniae by human polymorphonuclear leukocytes (PMN) and differentiated HL-60 cells in the presence of 10% baby rabbit complement and 1.2 mg/mL of total IgA (98% IgA, <1% IgG, <2% IgM). Given in Results are means ± SEM of experiments (PMN, n = 4; HL-60 cells, n = 6) with IgA purified from serum of volunteers before and 1 month after immunization with 23-valent pneumococcal vaccine. Levels of type 14–specific IgA were 678 ng/mL before immunization and 2,445 ng/mL after immunization in these pools of purified IgA. *P < 0.0001 for percent kill before immunization vs. 1 month after immunization. (c) The efficacy of molecular forms of IgA on phagocytic killing of type 14 S. pneumoniae. Serial dilutions of purified mIgA and pIgA were tested in the presence of 10% baby rabbit complement for their ability to mediate complement-dependent killing of the organism. Percent kill is given as a function of the concentration of anti–type 14 IgA in purified mIgA and pIgA. Tests were run in duplicate; results are shown as the mean ± SEM of 2 experiments.

Figure 3

Effect of receptor-blocking mAbs and isotype control antibodies on killing of type 14 S. pneumoniae with 2 μg/mL specific IgA and 10% baby rabbit complement. Differentiated HL-60 cells were preincubated for 30 minutes at 25°C with blocking antibodies (40 μg/mL) before addition of IgA-opsonized organisms and complement. Results are shown as the mean ± SEM of 2 experiments. P = 0.08 for percent kill with anti-Fcα receptor antibody compared with killing without blocking antibody.

Figure 4

Transmission electron microscopy of preparations treated with baby rabbit complement and IgA (a–c) or IgG (d). (a) The bacteria in the 3 phagosomes of a neutrophil are surrounded by electron-dense flocculent material. An empty shell is apparently all that remains of 1 bacterium (arrow). ×12,000. (b) Higher magnification of bacteria with coronas, in vacuoles surrounded by primary lysosomes of varying densities (arrowheads). ×25,000. (c) Still higher magnification shows changes in the integrity of a bacterium and membranous material that probably represents bacterial debris (arrow). ×50,000. (d) Three vacuoles are surrounded by electron-dense lysosomes and contain lysosomal constituents, some encasing bacteria, and bacterial debris (arrows). ×16,000.

Figure 5

Effect of cation chelators on phagocytic killing of type 14 S. pneumoniae by IgA in the presence of baby rabbit complement (a) and human complement (b). EDTA, which blocks both alternative and classical pathways, and MgEGTA, which blocks only the classical pathway, were added to killing reactions to determine the complement activation pathway involved in IgA-mediated phagocytic killing of pneumococci by differentiated HL-60 cells. Chelators (EDTA and MgEGTA) were added at 10 mM. All tests were run in triplicate; results are shown as the mean ± SEM of 6 experiments. *P < 0.05 vs. killing without EDTA. Results were the same whether individual experiments used IgA that was 98% pure (<2% IgM) or 99.9% pure (<0.1% IgM).

Figure 6

Effect of depletion and repletion of selected human complement components in human serum on IgA-mediated killing of type 14 S. pneumoniae by IgA. Organisms were incubated with 1 mg/mL of immune IgA, differentiated HL-60 cells, and a complement source consisting of 10% serum from a hypogammaglobulinemic patient, serum depleted of the classical pathway component C2, serum depleted of the alternative pathway component factor B, or serum depleted for factor B and then repleted with a physiologic concentration of factor B (200 μg/mL). All sera were preadsorbed with type 14 pneumococci before the killing assay to remove specific antibody in the complement source. Results are shown as mean ± SEM of 7 experiments. *P < 0.05 vs. normal complement and factor B–repleted complement.

Figure 7

Effect of PMN pretreatment on IgA-mediated killing of type 14 S. pneumoniae. Human PMNs were preincubated with media alone (first 2 columns) or TNF-α (10^–8M), C5a (10^–9M), or both for 30 minutes before addition of log-phase organisms and purified immune IgA (1 mg/mL total IgA). Baby rabbit complement was added to untreated PMNs with IgA as a positive control (first column); IgA with untreated PMNs and no complement served as a negative control (second column). PMNs preincubated with TNF-α and C5a showed no uptake or killing of the organism in the absence of immune IgA (not shown). Results are shown as mean ± SD for 3 experiments.