The scavenger receptor MARCO modulates TLR-induced responses in dendritic cells

Abstract

The scavenger receptor MARCO mediates macrophage recognition and clearance of pathogens and their polyanionic ligands. However, recent studies demonstrate MARCO expression and function in dendritic cells, suggesting MARCO might serve to bridge innate and adaptive immunity. To gain additional insight into the role of MARCO in dendritic cell activation and function, we profiled transcriptomes of mouse splenic dendritic cells obtained from MARCO deficient mice and their wild type counterparts under resting and activating conditions. In silico analysis uncovered major alterations in gene expression in MARCO deficient dendritic cells resulting in dramatic alterations in key dendritic cell-specific pathways and functions. Specifically, changes in CD209, FCGR4 and Complement factors can have major consequences on DC-mediated innate responses. Notably, these perturbations were magnified following activation with the TLR-4 agonist lipopolysaccharide. To validate our in silico data, we challenged DC's with various agonists that recognize all mouse TLRs and assessed expression of a set of immune and inflammatory marker genes. This approach identified a differential contribution of MARCO to TLR activation and validated a major role for MARCO in mounting an inflammatory response. Together, our data demonstrate that MARCO differentially affects TLR-induced DC activation and suggest targeting of MARCO could lead to different outcomes that depend on the inflammatory context encountered by DC.

Figure 1. Expression of MARCO receptor in splenic and bone marrow-derived DC.

(A) MARCO gene expression was determined in BMDC at various time points following treatment with various TLR agonists. Raw data from gene expression dataset GSE17721 were analyzed to extract MARCO expression values. Data were processed for normalization using the RMAexpress tool and gene annotation using the MeV software. (B) MARCO expression as determined by RT-PCR is shown in TLR agonist-activated DC2.4 cell line (left panel), splenic DC from WT and MARCO^−/− DC from 3 individual mice (middle panel), and TLR agonist-activated splenic DC (right panel). GAPDH expression was used for normalization. Data shown as Mean ± SD from triplicates. *P<.05; **P<.01.

Figure 2. Differentially expressed genes in WT and MARCO^−/− DC cells.

High purity DC preparations were isolated from splenocytes from 5–6 mice per group by positive selection with CD11c antibody and incubated overnight in media containing PBS or LPS (100 ng/ml). Total RNA was extracted and subjected to gene expression profiling. (A) Venn diagrams showing the numbers of genes that differ in expression by a factor of at least 2 between WT and MARCO^−/− DC without and with LPS (left diagram), and numbers of genes that are differentially upregulated (middle diagram) or downregulated (right diagram) in WT and MARCO^−/− DC following LPS exposure. (B) Top 15 differentially expressed genes that characterize MARCO vs. WT, WT_LPS vs. WT, MARCO_LPS vs. MARCO, and MARCO_LPS vs. WT_LPS. Data shown represent fold change of gene expression.

Figure 3. Major DC Signaling Pathways and Functions are affected by MARCO.

Sets of differentially expressed genes (fold change of 2 or higher) between different DC conditions were uploaded onto Ingenuity Pathway Analysis and corresponding signaling pathways were predicted. (A) MARCO vs. WT. (B) MARCO_LPS vs. WT_LPS. Statistical significance was set at −log P = 2 (Left Y Axis). The Ratio on the right Y axis represents the fraction of genes that are differentially expressed in our dataset that fall within a specific pathway out of the total number of genes that contribute to that pathway. Similarly, Biofunction analysis was performed for MARCO vs. WT (C) and MARCO_LPS vs. WT_LPS (D). Statistical significance was set at activation z-score = 2. Scores higher than 2 indicate activated functions, whereas scores lower than −2 indicate inhibited functions.

Figure 4. Comparison of Upstream Regulator status between WT and MARCO^−/− DC.

Differentially expressed genes (fold change of 2 or higher) were processed through Ingenuity Pathway Analysis to predict the downstream regulators whose activation status was affected by the absence of MARCO in resting cells (A) or LPS-challenged cells (B). The Venn diagram in (C) shows the transcription factors that respond to LPS in WT (WT_LPS) and MARCO^−/− (MARCO_LPS) DC. Transcription factors that reached the significant activation z-score of −2 or +2 are shown. (D) Shown are representative microRNAs that reached the significant activation z-score of −2 or +2. The IPA tool predicts a microRNA to be activated when enough differentially downregulated genes fall among the targets for this microRNA. The inhibition status is attributed when the opposite occurs.

Figure 5. Involvement of MARCO in the TGF-β signaling pathway.

Expression of differentially expressed genes in our dataset that are known to be regulated through by SMAD transcription factors within the TGF-β signaling pathway was measured by RT-PCR in resting WT and MARCO^−/− DC. *P<.05, **P<.01. Data show the mean ± SD of 3 WT and 3 MARCO^−/− samples where each sample represents a pool of 3 splenocyte preparations.

Figure 6. Validation of highly differentially expressed genes between WT and MARCO^−/− DC in response to TLR challenge.

DC were cultured overnight in the absence and presence of different TLR agonists. RT-PCR was performed to measure gene expression. *P<.05 for MARCO^−/− vs. WT DC. Data show 3 WT and 3 MARCO^−/− samples where each sample represents a pool of 3 splenocyte preparations.

Figure 7. Differential expression of immune and inflammatory marker genes between WT and MARCO^−/− DC in response to TLR agonist challenge.

DC were cultured overnight in the absence and presence of different TLR agonists. RT-PCR was performed to measure gene expression. (A) Basal expression in WT and MARCO^−/− DC in the absence of TLR ligation. *P<.05. Data show the mean ± SD of 3 WT and 3 MARCO^−/− samples where each sample represents a pool of 3 splenocyte preparations. (B) Gene expression ratio for MARCO/WT was calculated to reveal the magnitude of MARCO’s contribution for each individual gene across all TLR agonists.

Figure 8. Impact of MARCO on DC responsiveness to different TLR agonists.

RT-PCR data from Figure 7 was plotted using the Ligand/Ctrl ratio to reveal the contribution of the presence and the impact of the absence of MARCO on TLR-induced inflammatory gene signature in WT and MARCO^−/− DC, respectively. The Ligand/Ctrl Ratio was calculated for each gene to allow comparisons between WT and MARCO^−/− DC across all TLR agonists.

Figure 9. Impact of MARCO on TLR gene expression in DC.

TLR2-9 gene expression was determined in unstimulated splenic WT and MARCO^−/− DC. Data show the mean ± SD of 3 WT and 3 MARCO^−/− samples where each sample represents a pool of 3 splenocyte preparations. *P<.05. GAPDH expression was used for normalization.