Transcriptional networks in plasmacytoid dendritic cells stimulated with synthetic TLR 7 agonists

Abstract

Background: Plasmacytoid Dendritic Cells (pDC) comprise approximately 0.2 to 0.8% of the blood mononuclear cells and are the primary type 1 interferon (IFN), producing cells, secreting high levels of IFN in response to viral infections. Plasmacytoid dendritic cells express predominantly TLRs 7 & 9, making them responsive to ssRNA and CpG DNA. The objective of this study was to evaluate the molecular and cellular processes altered upon stimulation of pDC with synthetic TLR 7 and TLR 7/8 agonists. To this end, we evaluated changes in global gene expression upon stimulation of 99.9% pure human pDC with the TLR7 selective agonists 3M-852A, and the TLR7/8 agonist 3M-011.

Results: Global gene expression was evaluated using the Affymetrix U133A GeneChip(R) and selected genes were confirmed using real time TaqMan(R) RTPCR. The gene expression profiles of the two agonists were similar indicating that changes in gene expression were solely due to stimulation through TLR7. Type 1 interferons were among the highest induced genes and included IFNB and multiple IFNalpha subtypes, IFNalpha2, alpha5, alpha6, alpha8, alpha1/13, alpha10, alpha14, alpha16, alpha17, alpha21. A large number of chemokines and co-stimulatory molecules as well as the chemokine receptor CCR7 were increased in expression indicating maturation and change in the migratory ability of pDC. Induction of an antiviral state was shown by the expression of several IFN-inducible genes with known anti-viral activity. Further analysis of the data using the pathway analysis tool MetaCore gave insight into molecular and cellular processes impacted. The analysis revealed transcription networks that show increased expression of signaling components in TLR7 and TLR3 pathways, and the cytosolic anti-viral pathway regulated by RIG1 and MDA5, suggestive of optimization of an antiviral state targeted towards RNA viruses. The analysis also revealed increased expression of a network of genes important for protein ISGylation as well as an anti-apoptotic and pro-survival gene expression program.

Conclusion: Thus this study demonstrates that as early as 4 hr post stimulation, synthetic TLR7 agonists induce a complex transcription network responsible for activating pDC for innate anti-viral immune responses with optimized responses towards RNA viruses, increased co-stimulatory capacity, and increased survival.

Figure 1

Purity of pDC preparations as determined by flow cytometry. Representative histograms for partially purified pDC (A) and flow purified pDC (B). pDC were separated from PBMC using anti BDCA4 microbeads. The partially purified cells were then labeled with CD123-APC, BDCA2-FITC, and CD11c-PE antibodies and sorted on a FACSAria flow cytometer to a final purity of > 99%. Numbers in the top, left hand corner of the dot plots indicate the percentage of CD123+, BDCA-2+, CD11c- cells.

Figure 2

Increasing purity of pDC preparations correlate with decrease in the expression of TLR4, TLR5, and TLR8. (A) Change in expression of TLR4 (circles), TLR5 (squares), and TLR8 (triangles), and (B) TLR1 (diamonds), TLR6 (squares), TLR7 (triangles), TLR9 (circles), and TLR10 (stars) with increasing purity of pDC preparation. TLR expression was determined by real time RTPCR analysis. Relative copy number was calculated with respect to 2 ng cDNA after normalization to expression levels of GAPDH in each sample.

Figure 3

Cytokines secreted by flow purified pDC 4 hr post-stimulation with TLR7 agonists. Cell culture supernatants collected after stimulation were evaluated for cytokine protein expression using a Luminex 25-Plex assay. (A) Secreted TNFα, IL8, IFNα, MIP-1α, MIP-1β and IL6. (B) secreted IL1β, IL2R, IL12P70, Rantes, GM-CSF, MCP-1, and IP-10. Dotted bars, vehicle stimulated samples; hatched bars, 3M-852A-stimulated samples and solid bars, 3M-011-stimulated samples. The results are expressed as mean + SD, n = 2 donors.

Figure 4

Cluster analysis of 680 genes regulated in expression by 3M-852A and 3M-011. Gene expression analysis was performed at 4 hr post-stimulation of flow-purified pDC with 3M-852A and 3M-011 using the Affymetrix GeneChip U133A as described in Methods. D1 and D2, designate two different donors for pDC preparation in which the pDC purity was > 99%. Global gene expression was determined using the Affymetrix U133A GeneChip. Two-way hierarchical clustering was performed as described in the Methods section, using the Unweighted Pair-Group Method with Arithmetic mean (UPGMA) and the Euclidean similarity measure. The log2 fold change values were used for the analysis. Insert bar chart shows the expression change scale with red, green, and white signifying increased, decreased, and unchanged expression, respectively. Expression changes were evaluated with respect to vehicle stimulated pDC from the same donor. Expression changes for the 680 genes are documented in Additional File 1.

Figure 5

Transcription regulatory network indicates utilization of various transcription factors following activation of TLR7. Protein-interaction network was generated for expression changes in pDC stimulated with 3M-852A and 3M-011 4 hr post-stimulation. The network shows genes altered in expression in the data set which were transcriptionally regulated via IRF7, NFkB, cJun, IRF5, IRF8 through potential interaction with the TLR7 pathway and through ISGF3 via activation of the type 1 interferon receptor. Network was generated using the Analyze network (receptor) and the direct interactions algorithms using the whole expression data. Major hubs have been highlighted in bold. Remaining details of network are as described in Methods. Legend describing symbols in the network is found in Additional file 2.

Figure 6

Stimulation of pDC with 3M-852A and 3M-011 results in increased expression of a large number of chemokines. Protein-interaction network was generated for expression changes in pDC 4 hr post stimulation with 3M-852A and 3M-011. The network was generated using a list of genes in the GO function classification of chemokines and the shortest path algorithm. The network summarizes interactions between chemokines increased in expression and their respective receptors. Green arrows indicate binding interactions or covalent modifications that result in activation or increased transcription. Red arrows indicate binding interactions or covalent modifications that result in inhibition of activity or suppression of transcription.

Figure 7

Potential transcription regulation of genes through pathways secondary to stimulation of TLR7. Protein-interaction network was generated for expression changes in pDC stimulated with 3M-852A and 3M-011 4 hr post stimulation. Network was generated using the Analyze network (receptor) and the direct interactions algorithms using the whole expression data. The network shows genes altered in expression in the data set that are transcriptionally regulated through NFkB, c-Myc, and CREB1. Potential activation of these transcription factors through the chemokine receptor CCR1, CCR3 and the IL8 receptors is depicted. Bold green lines highlight the canonical pathways for activation of the three transcription factors. Fine lines going out of the transcription factors designate transcription regulation. Remaining network details are as described in Figure 6.

Figure 8

Stimulation of pDC with 3M-852A and 3M-011 results in the induction of an anti-apoptotic gene expression program. Protein-interaction network was generated for expression changes in pDC 4 hr post stimulation with 3M-852A and 3M-011. The network was generated from a list of genes with GO process classification of anti-apoptosis, using the shortest path algorithm. The network highlights the large number of anti-apoptotic genes that are increased in expression with a concomitant decrease in expression of key pro-apoptotic genes including several that are transscriptionally regulated by p53. Remaining network details are as described in Figure 6.

Figure 9

Protein regulatory network indicates importance of ISGylation pathway in TLR7-mediated activation of pDC. Protein-interaction network was generated for expression changes in pDC stimulated with 3M-852A and 3M-011 4 hr post-stimulation. The network was generated from a group of genes with the GO ontology classification of ubiquitin cycle using the shortest path algorithm. The network summarizes interactions between the ubiquitin and the ISG15 pathways for protein modification and the genes that are altered in expression upon stimulation of pDC with TLR7 agonists. Remaining network details are as described in Figure 6.