CD16+ monocytes give rise to CD103+RALDH2+TCF4+ dendritic cells with unique transcriptional and immunological features

Abstract

Classical CD16^- vs intermediate/nonclassical CD16⁺ monocytes differ in their homing potential and biological functions, but whether they differentiate into dendritic cells (DCs) with distinct contributions to immunity against bacterial/viral pathogens remains poorly investigated. Here, we employed a systems biology approach to identify clinically relevant differences between CD16⁺ and CD16^- monocyte-derived DCs (MDDCs). Although both CD16⁺ and CD16^- MDDCs acquire classical immature/mature DC markers in vitro, genome-wide transcriptional profiling revealed unique molecular signatures for CD16⁺ MDDCs, including adhesion molecules (ITGAE/CD103), transcription factors (TCF7L2/TCF4), and enzymes (ALDH1A2/RALDH2), whereas CD16^- MDDCs exhibit a CDH1/E-cadherin⁺ phenotype. Of note, lipopolysaccharides (LPS) upregulated distinct transcripts in CD16⁺ (eg, CCL8, SIGLEC1, MIR4439, SCIN, interleukin [IL]-7R, PLTP, tumor necrosis factor [TNF]) and CD16^- MDDCs (eg, MMP10, MMP1, TGM2, IL-1A, TNFRSF11A, lysosomal-associated membrane protein 1, MMP8). Also, unique sets of HIV-modulated genes were identified in the 2 subsets. Further gene set enrichment analysis identified canonical pathways that pointed to "inflammation" as the major feature of CD16⁺ MDDCs at immature stage and on LPS/HIV exposure. Finally, functional validations and meta-analysis comparing the transcriptome of monocyte and MDDC subsets revealed that CD16⁺ vs CD16^- monocytes preserved their superior ability to produce TNF-α and CCL22, as well as other sets of transcripts (eg, TCF4), during differentiation into DC. These results provide evidence that monocyte subsets are transcriptionally imprinted/programmed with specific differentiation fates, with intermediate/nonclassical CD16⁺ monocytes being precursors for pro-inflammatory CD103⁺RALDH2⁺TCF4⁺ DCs that may play key roles in mucosal immunity homeostasis/pathogenesis. Thus, alterations in the CD16⁺ /CD16^- monocyte ratios during pathological conditions may dramatically influence the quality of MDDC-mediated immunity.

Graphical abstract

Figure 1.

CD16⁺and CD16⁻ monocytes differentiate into DCs with distinct transcriptional profiles. (A) Shown is the experimental flowchart. Briefly, total monocytes were purified by negative selection using magnetic beads. Highly pure CD16⁺ and CD16⁻ monocytes were subsequently sorted by FACS on staining with CD16 Abs and a cocktail of FITC-conjugated nonmonocyte lineage-specific Abs (CD1c, CD3, CD8, CD19, and CD56; supplemental Figure 1). Immature MDDCs were generated by culturing monocyte subsets in the presence of GM-CSF and IL-4 for 6 days. Total RNA was extracted from matched CD16⁺ and CD16⁻ MDDCs (n = 5) that were exposed to media, LPS, or HIV for 24 hours. Genome-wide transcriptional profiling was performed using the Affymetrix HG Plus 2.0 microarrays. (B-C) Shown are the expressions of CD14 and CD16 on total monocytes before sort (B) and the expression of CD14, CD16, CD1c, and HLA-DR on immature CD16⁺ and CD16⁻ MDDCs on differentiation in vitro (C). Results in B-C are from 1 donor representative of results generated with cells from more than 10 donors. (D-H) Shown are Venn diagrams of differentially expressed genes in CD16⁺ vs CD16⁻ MDDCs in response to media (D), LPS (E), and HIV (G), as well as the representation of the number of commonly and differentially expressed genes on exposure to LPS (F) and HIV (H). n.s., not significant.

Figure 2.

Gene ontology classification of differentially expressed genes in immature CD16⁺and CD16⁻ MDDCs. Transcriptional profiling was performed as described in Figure 1A,D. Differentially expressed genes in CD16⁺ (green) vs CD16⁻ (blue) MDDCs exposed to media (immature; P < .05; FC cutoff, 1.3) were classified using gene ontology in (A) transcription factors, (B) cytokines, (C) adhesion molecules, (D) chemotaxis, and (E) cell projections. Each heat map column represents data from a distinct donor for matched immature CD16⁺ vs CD16⁻ MDDCs (n = 5).

Figure 3.

Differential gene expression in CD16⁺and CD16⁻ MDDCs in response to LPS. Transcriptional profiling was performed as described in Figure 1A,E-F with matched MDDC subsets exposed to media (immature) or LPS (mature) for 24 hours. Shown are top-regulated genes in immature and mature CD16⁺ compared with CD16⁻ MDDCs (A) and differentially expressed genes in CD16⁺ vs CD16⁻ MDDC subsets exposed to LPS (P < .05; FC cutoff, 1.3) linked to the gene ontology terms chemotaxis (B) and cytokines (C). Heat map cells are scaled by the expression level z-scores for each probe individually. Results were generated with cells from 5 different donors identified with different color codes.

Figure 4.

Differential gene expression in CD16⁺and CD16⁻ MDDCs in response to HIV. Transcriptional profiling was performed as described in Figure 1A,G-H, with matched MDDCs subsets exposed to media (immature) or HIV for 24 hours. Shown are top differentially regulated genes in immature and HIV-exposed CD16⁺ (green) and CD16⁻ MDDCs (pink) (A) and differentially expressed genes in CD16⁺ (green) vs CD16⁻ (blue) MDDC subsets exposed to HIV (P < .05; FC cutoff, 1.3) linked to the gene ontology terms exosome (B). Heat map cells are scaled by the expression-level z-scores for each probe individually. Results were generated with cells from 5 different donors identified with different color codes.

Figure 5.

CD16⁺and CD16⁻ MDDCs exposed to media, LPS, or HIV exhibit unique molecular signatures. Genome-wide transcriptional profiles were generated as described in Figure 1A,D-H for MDDC subsets exposed to media, LPS, or HIV for 24 hours. (A) Gene set enrichment analysis allowed the identification of top-regulated canonical pathways (C2), commonly and differentially expressed between CD16⁺ vs CD16⁻ MDDCs exposed to media, LPS, or HIV-1. (B) Shown are top-regulated micro-RNAs (MIR) differentially expressed between CD16⁺ and CD16⁻ MDDCs exposed to media, LPS, or infectious HIV virions. Results were generated with matched MDDC subsets from 5 different donors.

Figure 6.

Novel functional markers for CD16⁺and CD16⁻ MDDCs. MDDC subsets were generated and exposed to media (immature), LPS (mature), or HIV for 24 hours, as described in Figure 1A. Shown are real-time RT-PCR validation of relative ITGAE/CD103, CDH1/E-cadherin, ALDH1A2/RALDH2, and TCF7L2/TCF4 mRNA expression in immature MDDC subsets (A), as well as IGSF2/CD101 and SIGLEC1/CD169 in HIV-exposed MDDC subsets (B). (A-B) Results were generated with matched MDDC subsets from 3 to 5 different donors. Each symbol represents the median value of 2 to 3 RT-PCR replicates, with values for CD16⁻ MDDCs being considered 1. Paired Student t test P values are indicated on the graphs. (C) Shown are levels of TNF-α, CCL22, and CCL18 quantified by ELISA in cell culture supernatants from matched CD16⁺ and CD16⁻ MDDCs on exposure to media, LPS, or HIV (n = 4-5, mean ± SEM). Paired Student t test P values for log₁₀ cytokine levels are indicated on the graphs.

Figure 7.

Meta-analysis identifies a transcriptional signature conserved by CD16⁺and CD16⁻ monocytes during differentiation into DCs. Genes differentially expressed in CD16⁺ vs CD16⁻ monocytes (GSE16836) and CD16⁺ vs CD16⁻ MDDCs (GSE111474, current manuscript) were subject to a meta-analysis that led to the identification of a transcriptional signature preserved during monocyte differentiation into DCs.

Figure 8.

Unique transcriptional signatures discriminate CD16⁺from CD16⁻ MDDCs. Shown are top-differentially expressed genes in CD16⁺ vs CD16⁻ MDDCs at immature stage or on exposure to LPS and HIV. Selected top-modulated genes encode for transcriptional factors, surface markers, adhesion molecules, chemotaxis, cytokines, functional markers, HIV interactors, LPS response, and HIV response. Transcripts in bold were validated at RNA and/or protein level in Figure 6.