Plasmacytoid dendritic cells and C1q differentially regulate inflammatory gene induction by lupus immune complexes

Abstract

Immune complexes (ICs) play a pivotal role in causing inflammation in systemic lupus erythematosus (SLE). Yet, it remains unclear what the dominant blood cell type(s) and inflammation-related gene programs stimulated by lupus ICs are. To address these questions, we exposed normal human PBMCs or CD14(+) isolated monocytes to SLE ICs in the presence or absence of C1q and performed microarray analysis and other tests for cell activation. By microarray analysis, we identified genes and pathways regulated by SLE ICs that are both type I IFN dependent and independent. We also found that C1q-containing ICs markedly reduced expression of the majority of IFN-response genes and also influenced the expression of multiple other genes induced by SLE ICs. Surprisingly, IC activation of isolated CD14(+) monocytes did not upregulate CD40 and CD86 and only modestly stimulated inflammatory gene expression. However, when monocyte subsets were purified and analyzed separately, the low-abundance CD14(dim) ("patrolling") subpopulation was more responsive to ICs. These observations demonstrate the importance of plasmacytoid dendritic cells, CD14(dim) monocytes, and C1q as key regulators of inflammatory properties of ICs and identify many pathways through which they act.

Figure 1. Differential expression of genes and pathways in PBMCs after exposure to SLE ICs

A–C, SLE ICs were incubated with unprimed normal donor PBMCs for 5 h. Following RNA isolation, gene expression was quantified by microarray as described in Materials and Methods. A, Summary of genes regulated more than 1.5-fold compared to unstimulated controls with the intersection representing overlapping genes in two different donors (D1 and D2). B and C, The top 5 canonical pathways (B) and networks (C) are shown following PBMC IC stimulation. D, Five upregulated and 1 downregulated gene from the microarray data were quantified by qRT-PCR with 4 independent normal PBMC donor stimulations. E, The fold changes from the microarray and qRT-PCR for the same samples were plotted and the pearson correlation coefficient was calculated (r = 0.8952, P<0.0001). FC = fold change relative to unstimulated controls.

Figure 2. Differential expression of genes and pathways in PBMCs after exposure to C1q containing ICs (C1q-ICs) compared to ICs alone

A–B, SLE ICs were added to the same normal donor PBMCs as in Fig. 1 in the presence or absence of purified C1q (50 μg/ml). A and B, The top 5 canonical pathways (A) and networks (B) are shown for the comparison of C1q-ICs to ICs alone. C, 3 downregulated genes from the microarray data were quantified by qRT-PCR with a total of 4 independent normal PBMC donor stimulations. Results are shown as fold change compared to unstimulated controls with error bars representing the SEM. D, The fold changes from the microarray and qRT-PCR for IC alone, C1q-IC and C1q alone conditions were plotted and the pearson correlation coefficient was calculated (r = 0.9576, P<0.0001). *, P<0.05 with one-way ANOVA and Tukey’s multiple comparison post-test.

Figure 3. Differential expression of genes and pathways in purified CD14+ monocytes after exposure to SLE ICs

A and B, SLE ICs were added to normal donor purified monocytes for 5 h and RNA isolated for microarray analysis as described in Materials and Methods. A, Summary of genes differentially regulated more than 1.5-fold is shown with the intersection representing overlapping genes with 2 different donors (D1 and D2). B, The top 5 canonical pathways for each donor is shown with only one canonical pathway shared between them (in bold with genes listed in parentheses). C, Four upregulated and 1 downregulated gene from the microarray data were quantified by qRT-PCR with a total of 4 independent monocyte stimulations. Results are shown as fold up- or down-regulated comparatively to unstimulated controls. D, The fold changes from the microarray and qRT-PCR for the same samples were plotted and the pearson correlation coefficient was calculated (r = 0.9753, P<0.0001).

Figure 4. SLE ICs upregulate CD86 only on CD14^dim monocytes

A and B, SLE ICs were incubated with total CD14⁺ purified monocytes (>96% pure) in medium alone (A) or with or without 4 h priming with 500 U/ml Type I IFN (B). After 2 days in culture, CD86 and CD40 expression was quantified by flow cytometry. Results in (A) are expressed as mean fluorescent intensities (MFI) from 3 independent experiments with 2–3 different SLE patient ICs per experiment. Results in (B) are representative of 3 independent experiments with 3 different SLE patient ICs added per experiment. C, The 3 different monocyte subsets, CD14^dimCD16⁺, CD14⁺CD16⁺, and CD14⁺CD16⁻ were isolated by flow sorting and exposed to SLE ICs as in (A). CD86 expression was quantified and expressed as the ratio of MFI following IC stimulation/medium (No IC) control. D and E, SLE ICs were formed using fluorescently labeled SmRNP antigen in the absence (D) or presence of C1q (20 μg/ml) (E) and allowed to bind to PBMCs for 30 min on ice. D, Left panel, Representative histogram with the MFI of ICs bound to each monocyte subset shown in the legend below. Right panel, Compiled MFI data of SLE IC binding to each monocyte subset with 1–6 SLE ICs per experiment for a total of 3 independent experiments. E, Left panel, Representative experiment with 4 SLE ICs and the binding to monocyte subsets within PBMCs with or without C1q. Right panel, Compiled data for SLE IC binding to monocyte subsets expressed as an MFI ratio of C1q-IC to IC binding in the absence of C1q (IC alone) from 3 experiments with a total of 11 SLE patient ICs. Horizontal lines are mean +/− SEM). n.s., not significant; *, P<0.05; **, P<0.01 with unpaired t-test (A), one-way ANOVA with Tukey’s multiple comparison post-test (C–E).

Figure 5. SLE ICs upregulate monocyte CD86 and CD40 expression in PBMC cultures but increased expression requires the presence of pDCs

SLE ICs were incubated with total PBMC cultures as in Fig. 1. A, After 2 d, CD14⁺ monocytes were analyzed for CD86 and CD40 expression by flow cytometry. Results from 4 independent experiments with 3–5 different SLE patient ICs per experiment are expressed as MFI where each line is a different SLE IC stimulation. B, Monocyte expression of CD86 and CD40 following SLE IC or loxoribine (200 μM) stimulation was analyzed by flow cytometry as in (A) except that PBMCs were either mock depleted or depleted of pDCs (pDC dep) with BDCA-4 magnetic beads (representative depletion shown in upper panel). MFI results from 3 independent experiments with at least 2 different SLE patient ICs per experiment are shown in the lower panel. Loxo, loxoribine. MED, medium. *, P<0.05; **, P<0.01; ***, P<0.001 with unpaired t-test (A), paired t-test (B).

Figure 6. pDCs are required for maximal induction IL-1RA, IP-10, MCP-1 and MCP-3, but not IL-8, in response to SLE ICs

A and B, Cytokines/chemokines were measured by multiplex analysis as described in the Materials and Methods after mock depleted (Mock) or pDC depleted (pDC dep) PBMCs (A) or purified monocytes (B) were exposed to SLE ICs for 2 days of stimulation. Raw data in pg/ml are plotted from 2 independent experiments with 2 different donors and 3 (A) or 2 (B) different SLE ICs per experiment. No IC indicates the presence of freeze-thawed antigen alone.