Scavenger receptor-A functions in phagocytosis of E. coli by bone marrow dendritic cells

Abstract

Class-A scavenger receptors (SR-A) are cellular pattern recognition receptors that bind and traffic a variety of endogenous and microbial ligands. However, despite an emerging role for SR-A as a contributor to the innate immune system, little is known of the regulation or function of SR-A on dendritic cells (DCs). Here we show that SR-A expression is upregulated during murine DC differentiation and that SR-A expression levels correlate with the expression of the murine DC marker CD11c. Using bone marrow-derived DCs (BMDCs) from SR-A knockout (SR-A(-/-)) mice, we investigated the contribution of SR-A to BMDC particulate phagocytosis. Functional analyses demonstrated that SR-A is a critical phagocytic receptor for BMDC internalization of the gram-negative bacteria E. coli. SR-A(-/-) BMDCs were impaired in their ability to phagocytose bacteria, and this deficit varied with the bacteria:BMDC cell ratio. Microscopic and biochemical analyses revealed that SR-A is broadly distributed on the surface of BMDCs and is not physically associated with lipid rafts. However, cholesterol depletion demonstrated dependence of SR-A-mediated phagocytosis upon lipid rafts. These data demonstrate a functional contribution for SR-A in the BMDC phagocytic pathway.

Figure 1

SR-A expression is induced during BMDC differentiation and correlates with CD11c expression levels. (A) BMDCs from SR-A^+/− mice were assayed for CD11c and SR-A expression by FACS analysis. Analysis shows that >90% of Day 6 BMDCs are CDllc⁺SR-A⁺. Cells were gated on viable, differentiated BMDCs by forward and side scatter profile. (B) CB6F1 SR-A^+/− BMDCs were harvested on the indicated days and lysates from these harvests were assayed for SR-A expression by Western Blot (left). Cell lysates from Day 6 BMDCs of SR-A^+/− and SR-A^−/− mice were prepared and assayed by Western blot for total cellular SR-A expression (right). Calreticulin (CRT) was used as a loading control. (C) BMDC cultures were prepared and harvested on indicated days as in (B) and analyzed for CD11c and SR-A expression by FACS analysis. Graph is represented as fold-increase over time, with standard deviation of triplicate samples shown. (D) Anti- SR-A staining specificity was confirmed by FACS analysis using 2F8 mAb to stain Day 6 BMDCs from CB6F1 SR-A^+/−, C57Bl/6, and SR-A^−/− mice. SR-A^+/− BMDCs show robust staining, SR-A staining is not detected in the SR-A^−/− cells, while C57BL/6 -derived SR-A^+/+ BMDCs serve as a negative control since 2F8 does not recognize the C57BL/6-derived SR-A gene product.

Figure 2

SR-A^−/− BMDCs exhibit defective phagocytosis of heat-killed E. coli. (A) Alexa488-labeled and Alexa647-labeled E. coli were incubated at the indicated bacteria:BMDC ratios with SR-A^+/− BMDCs to determine the linear range of BMDC phagocytic activity. The percent of BMDCs that acquired bacteria, as assessed by fluorescence, was used as the analysis parameter. (B) FACS analysis of BMDCs incubated with Alexa488-labeled E. coli at a 3:1 bacteria:BMDC ratio for 20 minutes at 37°C in the absence (None) or presence of the scavenger receptor ligands fucoidan (Fuc, 250 μg/ml) or carrageenan (Car, 75 μg/ml). Chondroitin-B sulfate (ChB, 250 μg/ml) was used as a negative control for the scavenger receptor competitive ligands. (C) SRA^+/− and SRA^−/− BMDCs were incubated with Alexa488-labeled E. coli as in (A) at the indicated bacteria:BMDC ratios, and assessed for relative uptake of E. Coli by FACS analysis. (D) Confocal microscopy of SRA^+/− BMDCs incubated with Alexa647-labeled E. coli (viewed as red) at a 3:1 bacteria:BMDC ratio for 20 minutes at 37°C, followed by counter-staining with DAPI (blue) and FITC-conjugated wheat germ agglutinin (WGA; viewed as green), show that fluorescence conferred to BMDCs by incubation with fluorescently labeled E. coli is due to binding and uptake of the bacteria. (E) SR-A^+/− or SR-A^−/− BMDCs were incubated with the indicated concentrations of Alexa488-labeled IgG (left) or Alexa488-labeled acetylated-LDL (right) for 20 minutes at 37°C, followed by washing and FACS analysis to assess endocytic uptake of each ligand. Standard deviation is shown. (F) FACS analyses of SR-A^+/− (black lines) or SR-A^−/− (grey lines) BMDCs stained for the phagocytic receptors Fc receptors CD16/CD32, DEC205, and Lox-1. The grey-filled histograms represent negative controls.

Figure 3

The phagocytosis defect in SRA^−/− BMDCs is a function of multiple parameters. (A) SRA^+/− and SRA^−/− BMDCs were untreated or were incubated with Alexa488-E. coli at a 3:1 bacteria:BMDC ratio. Total mean fluorescence intensity (MFI) was assayed by FACS, with SRA^−/− BMDCs exhibiting impairment in bacterial uptake. (B) FACS histogram of SRA^+/− BMDCs incubated with Alexa488-E. coli at a 3:1 bacteria:BMDC ratio. Gate (M1) indicates BMDCs that accumulated Alexa488-E. coli. The phagocytic impairment of SR-A^−/− BMDCs was then analyzed by two different parameters, percent of dendritic cells that phagocytose bacteria (C) and the relative number of bacteria phagocytosed per dendritic cell (D). (C) The percent of BMDCs that accumulate E. coli was calculated by the number of cells within gate M1 (as in (B)) divided by the total number of cells analyzed. Twice the percentage of SR-A -expressing BMDCs acquired bacteria in comparison to SR-A –deficient BMDCs. (D) The relative number of bacteria phagocytosed per dendritic cell within gate M1 was assessed by mean Alexa488 fluorescence. Amongst the dendritic cells that accumulated bacteria, SR-A^−/− BMDCs acquired fewer bacteria per cell than SR-A –expressing BMDCs. Standard deviation is shown.

Figure 4

SRA^−/− dendritic cells are impaired in their ability to phagocytose live E. coli. SRA^+/+, SRA^+/− and SRA^−/− BMDCs were incubated with live E. coli bacteria at the indicated multiplicity of infection (MOI) for 45 minutes either on ice, or at 37°C. Gentamycin was then used to kill non-internalized bacteria. Bacteria phagocytosed (protected from antibiotic) was quantified as the number of colony-forming units (CFU) following gentamycin treatment. Standard deviation is shown.

Figure 5

SR-A –mediated phagocytosis is dependent on lipid rafts although SR-A does not physically associate with lipid rafts at the cell membrane. (A) Scanning Electron Microscopy of CB6F1 BMDCs shows the distinct morphological features of dendritic cells (top panel). Magnification of the section outlined in the white box (shown in the bottom panel) reveals gold-decorated SR-A on both dendritic extensions (small arrows) and lamellapodia (large arrows). 18 nM gold particles were used to decorate the 2F8-stained BMDCs. (B) Confocal microscopy of SR-A^+/− BMDCs stained for SR-A (green) shows broad surface expression of SR-A. Cells were counter-stained with DAPI (blue). (C) Detergent-insoluble BMDC membranes (lipid rafts) were separated from detergent-soluble membrane components using a discontinuous sucrose gradient as described in Methods. 1 ml fractions from the gradient were collected and assessed by Western blot for SR-A expression. SR-A was observed in the detergent-soluble membrane fractions (fractions 1-2) but not in the detergent-insoluble lipid raft fractions (fractions 5-6) of the gradient, while raft-associated cholera toxin-B subunit (CT-B, FITC-conjugated) migrated to fractions 5 and 6 of the gradient. (D) SRA^+/− and SRA^−/− BMDCs were incubated in the presence or absence of 5 mM β-Cyclodextrin (β-CD) for 30 minutes at 37°C, followed by washing. Cells were then incubated with live E. coli bacteria at an MOI of 10:1, followed by incubation in gentamycin to kill non-internalized bacteria. Phagocytosis of bacteria was evaluated as the number of colony-forming units (CFU) per treatment. Standard deviation is shown. (E) SRA^+/− and SRA^−/− BMDCs treated with or without 5 mM β-CD were subsequently incubated with a 10:1 MOI of live E. coli for 45 minutes at 37°C or 4°C as indicated, followed by gentamycin treatment to kill non-internalized bacteria. Phagocytosis of bacteria is represented by the number of colony-forming units (CFU) per treatment. Standard deviation is shown. (F) β-CD treated and untreated SRA^+/− BMDCs were comparatively assessed for cell viability by FACS analysis of propidium iodide staining. With the described experimental conditions β-CD treatment did not affect the viability of BMDCs.