Ultra-purification of Lipopolysaccharides reveals species-specific signalling bias of TLR4: importance in macrophage function

Abstract

TLR4 location, and bacterial species-derived lipopolysaccharides, play a significant role in the downstream activation of transcription factors, accessory molecules, and products. Here, this is demonstrated through the use of classically-activated and alternatively-activated macrophages. We show that, when polarized, human macrophages differentially express and localize TLR4, resulting in biased recognition and subsequent signalling of LPS derived from Pseudomonas aeruginosa, Escherichia coli, and Salmonella enterica. Analysis of activation demonstrated that in classically activated macrophages, P. aeruginosa signals from the plasma membrane via TLR4 to p65 dependent on TAK1 and TBK1 signalling. E. coli signals dependent or independent of the endosome, utilizing both TAK1- and TBK1-signalling to induce P65 and IRF3 inducible genes and cytokines. S. enterica however, only induces P65 and IRF3 phosphorylation through signalling via the endosome. This finding outlines clear signalling mechanisms by which innate immune cells, such as macrophages, can distinguish between bacterial species and initiate specialized responses through TLR4.

Figure 1

Differentiated M1 and M2 macrophage expression of TLR4 in vitro. PMA differentiated THP-1 monocytes (Resting M0-Macrophages) were stimulated for 24 h with either (50 ng/ml) recombinant human IFNγ (M1 polarization) or 25 ng/ml each of recombinant human IL-4 and IL-13 (M2). Polarity of matching sets of macrophages were initially confirmed by (A) qPCR analysis for (Ai) M1 markers: CXCL10, CCR7, IL-12 or (Aii) M2 markers: CD206, CD163, CCL17. Cell surface expression of (B) CD14 or (C) TLR4 was determined through flow cytometric analysis. Further analysis of total expression of TLR4 through permeabilization of cells and comparison to cell surface staining reveals (D) Abundance of TLR4 on the surface or total (including internal) of M1 and M2 polarized THP-1 macrophages. (E) Single colour immunofluorescent confocal microscopy images of TLR4 expression within (Ei) M0, (Eii) M1, (Eiii) M2 macrophages show differential patterns of staining within the cells. Images are a representative of 3 independent experiments. qPCR data is expressed as the mean ± SEM of 3 independent differentiations of THP-1 derived macrophages of passages 5–8. (Histograms: Pink-isotype control, Blue—THP-1 monocyte controls, Green and Dark Green—M0 macrophages, Purple and Blue—M1 macrophages, Dark and light grey—M2 macrophages). Scale bar = 20 µm. Statistical analysis performed on qPCR data was a two tailed multiple unpaired students t-test comparing stimulated to control, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 2

Differential M1 and M2 macrophage de novo NF-κB and IRF3 induction and cytokine production in response to LPS species. PMA differentiated THP-1 monocytes (Resting M0-Macrophages) were stimulated for 24hrs with either (50 ng/ml) recombinant human IFNγ (M1 polarization) or 25 ng/ml each of recombinant human IL-4 and IL-13 (M2). Macrophages were challenged with 1 ng/ml LPS in complete media for up to 24 h after which supernatants were collected for (Ai–v) multiplex cytokine (Il-1B, IL-6, IL-9, IL-10, TNFα), (vi) representative MTT viability assay experiment. Data is expressed as the mean ± SEM of 3 independent differentiations of THP-1 derived macrophages of passages 5–8. Mean ± SD is displayed for MTT. Statistical analysis performed on qPCR data was a multiple unpaired students t-tests comparing stimulated to control, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 3

E. coli and S. enterica, but not P. aeruginosa LPS signal partially via the endosome in polarized macrophages resulting in altered inflammatory cytokine production. Differentiated M1 THP-1 macrophages were stimulated with LPS from P. aeruginosa, E. coli or S. enterica with or without pre-incubation with 100 µM Hydroxy-Chloroquine. Induction of (A) LPS induced P65 phosphorylation measured after 30 min with/without Clq (Hydroxychloroquine). (B–E) IP-10, TNFα, IL-6 and IL-8 production was measured in conditioned supernatants. Data is expressed as the mean ± SEM of 3 independent differentiations of THP-1 derived macrophages of passages 5–8. Statistical analysis performed on qPCR data was a one-way ANOVA with Tukeys posthoc test, **P < 0.01, ****P < 0.0001.

Figure 4

LPS-driven phosphorylation of P65 and IRF3 is species specific and TAK1 and/or TBK1 dependent. Cells were then lysed and assayed for (Ai) phospho-P65, total p65, phospho-IRF3, total IRF3 and B-Actin acting as a loading control (Upper panel is a representative of n = 3 experiments). (Aii) Phosphorylated P65 and IRF3 were analysed in comparison to total P65 or IRF3 respectively. Data is represented as mean ± SEM of 3 separate stimulations performed on IFNγ differentiated M1 macrophages Passage 6–9. (B) Graphical representation of proposed signalling dynamics of the individual LPS via TLR4 and targets for TAKINIB or MRT67307 inhibitors. Cells were stimulated with LPS (1 ng/ml) with or without 1 h pre-treatment with TAKINIB (10 µM) or MRT67307 (10 µM) for 30 min in serum starved conditions. Membranes were stripped and re-probed, or cut longitudinally, to probe multiple antibodies within the same membrane, full-uncropped images can be found in Supplemental Fig. S1. The images in figure are from 2 separate gels and membranes. Statistical analysis performed on qPCR data was a one-way ANOVA with Tukeys posthoc test **P < 0.01, ***P < 0.001, ****P < 0.0001. Illustration created using Biorender.com. Figure generated by MS with permissions granted for publication.