Antibacterial effect of microvesicles released from human neutrophilic granulocytes

Abstract

Cell-derived vesicles represent a recently discovered mechanism for intercellular communication. We investigated their potential role in interaction of microbes with host organisms. We provide evidence that different stimuli induced isolated neutrophilic granulocytes to release microvesicles with different biologic properties. Only opsonized particles initiated the formation of microvesicles that were able to impair bacterial growth. The antibacterial effect of neutrophil-derived microvesicles was independent of production of toxic oxygen metabolites and opsonization or engulfment of the microbes, but depended on β(2) integrin function, continuous actin remodeling, and on the glucose supply. Neutrophil-derived microvesicles were detected in the serum of healthy donors, and their number was significantly increased in the serum of bacteremic patients. We propose a new extracellular mechanism to restrict bacterial growth and dissemination.

Figure 1

Effect of various stimulants on MV formation. Amount of formed MVs was followed on the basis of protein concentration (A-C) or flow cytometry after staining PMNs with anti-CD11b antibody (bacteria were not stained at all). (B-C) Stimulation of 9 × 10⁶/mL PMNs was carried out with 10⁸/mL opsonized S aureus (A-C) or 100nM PMA (B-C) or 1μM fMLF (C) for 20 minutes or 20 ng/mL TNFα for 30 minutes, 100 ng/mL LPS for 120 minutes or 100 ng/mL CXCL-12 for 5 minutes or 10⁸/mL S aureus with or without opsonization (C). In panel A the protein content of bacteria cosedimented with MVs is shown. In panel C and all later figures, this value has been subtracted. Bars show mean ± SEM, n = 4; #P < .05.

Figure 2

Characterization of PMN-derived MVs. (A) Fluorescence microscopy image of separated MVs. CD11b was marked with anti-CD11b R-PE conjugated monoclonal antibody. Original magnification is 630×. (B) Flow cytometry analysis of binding of FITC-conjugated annexin to phosphatidylserine. (C) Representative electron microscopic image of MVs. Original magnification is 10 000×. (D) Analysis of isolated MVs by dynamic light scattering. The x-axis is set to logarithmic scale, a(r_h) denotes the coefficient of autocorrelation function of the scattered electric field. (E) Superoxide production of 10⁶ PMNs or MVs derived from 10⁷ PMNs stimulated by 100nM PMA (dots show mean of measured RFU ± SEM, n = 4).

Figure 3

Effect of PMNs derived MVs on bacterial growth. (A) Effect of PMNs (9 × 10⁶/mL) and of b-MVs (derived from the same amount of PMNs) in time on growth of fresh, opsonized S aureus (9 × 10⁷/mL). Points indicate mean ± SEM, n = 47. (B) MVs were collected from equal number of suspended PMNs incubated with opsonized or nonopsonized S aureus or from adherent PMNs incubated with opsonized S aureus. (C) MVs were collected after incubation of PMNs with the indicated agents (concentration and times are described in Figure 1C) and applied at equal protein concentration. (D) Effect of the same amount of b-MVs on the indicated, opsonized bacteria. (E) Antibacterial capacity of 9 × 10⁶/mL PMNs (■) and of b-MVs (▴; derived from the same amount of PMNs) on opsonized (full symbols) and nonopsonized (empty symbols) S aureus. In panels B through E bacterial growth after 30 minutes incubation is shown; 100% represents the initial bacterial count (9 × 10⁷/mL) that did not change in the absence of nutrients, but increased during the 30 minutes incubation period in the presence of nutrients. In panel B points indicate mean of protein concentration and of bacterial growth rate, ± SEM, n = 4; in panels C through E bars indicate mean ± SEM, n = 4; #P < .05.

Figure 4

Characterization of the antibacterial effect of b-MVs. (A) Effect of various inhibitors (cytochalasin B 10μM; latrunculinA 10μM; jasplakinolide 1μM; wortmannin 300nM; all added 5 minutes before incubation), absence of glucose or presence of blocking CD18-antibodies (clone TS1/18), or 1mM EDTA or 1μM DPI or pretreatment with distilled water or with saponin (1 mg/mL) on antibacterial effect of b-MVs. Change in S aureus growth was measured after 30 minutes incubation with the indicated sample. 100% represents the initial bacterial count that did not change in the absence of nutrients, but increased during the 30 minutes incubation period when inactive or destroyed vesicles were present. It was controlled that inhibitors in the applied concentration and time as well as absence of glucose by themselves did not affect bacterial growth. Bars and points indicate mean, ± SEM, n = 4; #P < .05. (B) Proteomic analysis of the contents of PMN-derived MVs. Equal amount of total protein was analyzed in the different samples. Proteins were ranked by spectral count. (C) Western blot of lactoferrin and myeloperoxidase of s-MVs and b-MVs produced by 2 × 10⁷ PMNs. Actin is shown as loading control. (D) Densitometric analysis of lactoferrin and myeloperoxidase signal related to total protein content of MVs (± SEM; n = 5); #P < .05. See also supplemental Tables 1 and 2.

Figure 5

Potential mechanism of the antibacterial effect of b-MVs. (A-C) Representative transmission fluorescence microscopy images (of 35 similar ones from 4 experiments) about coincubation of b-MVs with opsonized S aureus at time points 0 (A), 10 minutes (B), and 30 minutes (C). Red marks indicate CD11b-positivity and green marks show GFP-expressing S aureus. (D) Confocal microscopy image of a representative clump (35 areas of interest were investigated from 4 experiments). Z-axis was 1 μm. Red and green marks are the same as described. (E) Statistical analysis of images taken from 4 independent experiments. Means were made on each sample by 3 individual investigators, and averaged. On the x-axis the ratio of bacteria and CD11b-positive, at least 1.5-μm wide aggregates to all CD11b-positive events is presented. Concentration of inhibitors is the same as detailed in Figure 4. Error bars show SEM, n = 4; #P < .05. (F) Relation of growth rate to the ratio of clumped bacteria (bacteria within CD11b-positive environment compared with all bacteria). Error bars represent SEM, n = 4. (G) Graphic representation of the proportion of clumped bacteria, large aggregates, and growth rate of bacteria in the presence of 3 different MV preparations. Error bars represent SEM, n = 4.

Figure 6

PMN derived MVs in vivo and ex vivo. MVs were separated from serum, using the same conditions as applied for isolated PMNs. (A) Concentration of PMN derived MVs in serum of healthy donors (n = 6) or patients with clinically verified S aureus bacteremia (n = 12), measured by flow cytometry after staining with anti-CD11b and anti-CD177 antibodies. (B-E) MVs separated from serum were incubated with not opsonized S aureus for 30 minutes at 37°C in HBSS with gentle shaking. Interaction was followed by fluorescent microscopy at the end of incubation. (B) Representative image of MVs from healthy donors (of 35 similar images from 4 independent experiments), whereas panel C is representative of MVs from bacteremic patients (53 similar images from 6 independent experiments). Red shows CD11b positivity, green represents S aureus. Statistical analysis (with same criteria as in Figure 5) of microscopic images shows amount of aggregated MVs to all MVs (D), and amount of clumped bacteria to all bacteria (E). Bars show mean ± SEM; #P < .05.