The synthetic bacterial lipopeptide Pam3CSK4 modulates respiratory syncytial virus infection independent of TLR activation

Abstract

Respiratory syncytial virus (RSV) is an important cause of acute respiratory disease in infants, immunocompromised subjects and the elderly. However, it is unclear why most primary RSV infections are associated with relatively mild symptoms, whereas some result in severe lower respiratory tract infections and bronchiolitis. Since RSV hospitalization has been associated with respiratory bacterial co-infections, we have tested if bacterial Toll-like receptor (TLR) agonists influence RSV-A2-GFP infection in human primary cells or cell lines. The synthetic bacterial lipopeptide Pam3-Cys-Ser-Lys4 (Pam3CSK4), the prototype ligand for the heterodimeric TLR1/TLR2 complex, enhanced RSV infection in primary epithelial, myeloid and lymphoid cells. Surprisingly, enhancement was optimal when lipopeptides and virus were added simultaneously, whereas addition of Pam3CSK4 immediately after infection had no effect. We have identified two structurally related lipopeptides without TLR-signaling capacity that also modulate RSV infection, whereas Pam3CSK4-reminiscent TLR1/2 agonists did not, and conclude that modulation of infection is independent of TLR activation. A similar TLR-independent enhancement of infection could also be demonstrated for wild-type RSV strains, and for HIV-1, measles virus and human metapneumovirus. We show that the effect of Pam3CSK4 is primarily mediated by enhanced binding of RSV to its target cells. The N-palmitoylated cysteine and the cationic lysines were identified as pivotal for enhanced virus binding. Surprisingly, we observed inhibition of RSV infection in immortalized epithelial cell lines, which was shown to be related to interactions between Pam3CSK4 and negatively charged glycosaminoglycans on these cells, which are known targets for binding of laboratory-adapted but not wild-type RSV. These data suggest a potential role for bacterial lipopeptides in enhanced binding of RSV and other viruses to their target cells, thus affecting viral entry or spread independent of TLR signaling. Moreover, our results also suggest a potential application for these synthetic lipopeptides as adjuvants for live-attenuated viral vaccines.

Figure 1. Pam3CSK4 modulates rgRSV infection in different cell types.

(A–F) Primary cells or immortalized cell-lines were incubated with bacterial TLR ligands using conditions as described previously . In undifferentiated primary normal human bronchial epithelial (NHBE) cells, antigen presenting cells (immature DC, BAL cells and EBV-transformed B-LCL), Pam3CSK4 mediated significant enhancement of rgRSV infection (*p<0.05 in 1-way ANOVA and posthoc Dunnett's correction for multiple comparisons with medium control). In contrast, treatment of A549 or HEp-2 cells with Pam3CSK4 and Pam2CSK4 resulted in reduced rgRSV infection levels (*p<0.05 using Dunnett's correction). Data are presented as percentages GFP-positive cells (means ± SD of triplicates). A representative experiment out of three is shown.

Figure 2. Pam3CSK4 mediates rapid and dose-dependent enhancement of RSV infection in B-LCL.

(A) B-LCL were incubated at 37°C with Pam3CSK4 (10 µg/ml) at 6 h, 3 h, 1.5 h, 0.5 h or 0 h before or 0.5 h, 1.5 h, 3 h, and 6 h after rgRSV infection. Pam3CSK4 treatment resulted in significantly increased rgRSV infection percentages in B-LCL at all time points (*p<0.005 with Student's t-tests with Bonferroni correction for multiple comparisons). (B) B-LCL or virus were pre-incubated with different concentrations of Pam3CSK4 for 30 minutes at 37°C before infection with rgRSV. Pam3CSK4 treatment resulted in significantly increased rgRSV infection percentages (p<0.05 for Pam3CSK4 concentrations ≥3 µg/ml), but no biologically relevant differences were detected between pre-incubation of cells or virus. (C) Cells were treated with different concentrations of Pam3CSK4 either directly before or after rgRSV infection. In all conditions free lipopeptide or free virus was washed away before proceeding to the next step. Pam3CSK4 treatment resulted in significantly higher percentages of infected cells (p<0.01, significant under Bonferroni correction for Pam3CSK4 concentrations ≥3 µg/ml). However, addition of Pam3CSK4 after rgRSV infection and subsequent removal of free virus had no effect. (D) B-LCL were incubated for 30 minutes at 37°C with or without Pam3CSK4 (10 µg/ml) before infection with different RSV strains. Pam3CSK4 treatment resulted in increased RSV infection percentages (p<0.05) for all strains tested. In all panels data are presented as percentages RSV-infected cells (means ± SD of triplicates) 20–24 hours after infection as determined by flow cytometry, using GFP expression in panels A–C and staining with FITC-labeled RSV-specific antibodies in panel D. Representative experiments out of three are shown.

Figure 3. The effect of Pam3CSK4 on B-LCL is independent of TLR signaling, but mediated by enhanced virus binding.

(A) B-LCL were pre-incubated with the myeloid differentiating factor 88 (MyD88) homodimerization inhibitory peptide, a control peptide, or polyclonal blocking antibodies to TLR1, TLR2 and/or TLR4, followed by Pam3CSK4 incubation (10 µg/ml, 30 minutes, 37°C) and subsequent infection with rgRSV. In all conditions Pam3CSK4 treatment resulted in enhancement of rgRSV infection (p<0.05). (B) B-LCL were incubated with different lipopeptides (for molecular structures see Figure S1) with or without TLR signaling capacities. In addition to TLR agonist Pam3CSK4 two non-TLR2 activating lipopeptides Pam-Cys-SK4 and PHCSK4 enhanced rgRSV infection, whereas the TLR agonist Pam3CSP4 did not (p<0.05 with Dunnett's correction for multiple comparisons for Pam3CSK4, Pam-Cys-SK4, and PHCSK4 concentrations 3 and 10 µg/ml). (C) RSV binding to B-LCL was determined by pre-incubating rgRSV with the different lipopeptides, followed by assessment of binding to cells. RSV binding was quantified by staining with FITC-labeled antibodies to RSV or to influenza-B as a control. Addition of the lipopeptides Pam3CSK4, Pam-Cys-SK4 and PHCSK4 but not Pam3CSP4 significantly increased binding of RSV to the cells (p<0.05 using Dunnett's correction for multiple testing). (D) Direct RSV binding to lipopeptides was examined in an ELISA format. Lipopeptides were coated on high-binding ELISA plates and incubated with RSV, or with MV strain Edmonston as a control. RSV binding was measured using an RSV-specific monoclonal antibody, followed by detection using a goat-anti-mouse peroxidase and TMB as a substrate. Coating with Pam3CSK4, Pam-Cys-SK4 and PHCSK4 but not with Pam3CSP4 resulted in increased RSV binding (p<0.05 compared with Dunnett's correction). Data are shown as extinctions at 450 nm (means ± SD of duplicates). (E) B-LCL were incubated with Pam3CSK4-CF or CSK4-CF at 4°C or 37°C to measure direct (lipo)peptide binding to the cells. At both temperatures Pam3CSK4-CF showed significantly stronger binding than CSK4-CF (p<0.01 with Bonferroni correction at concentrations 10 and 30 µ/ml). In panels A-C results are shown as percentages GFP- or FITC-positive cells (means ± SD of duplicates) and in Panel E data are shown as geometric mean fluorescence (means ± SD of triplicates). Representative examples of at least three experiments are shown.

Figure 4. Importance of lipopeptide charge and structure.

(A, C, D) B-LCL were incubated with different lipopeptides for 30 minutes at 37°C before infection with rgRSV. Only Pam3CSK4, Pam-Cys-SK4, Pam3CSK2 and Pam3CSK5 resulted in enhancement of rgRSV infection, whereas Pam3CSE4 resulted in inhibition of rgRSV infection compared to medium control (p<0.05 in 1-way ANOVA and posthoc Dunnett's correction). (B) Natural and chemically-derived polyanionic compounds were assessed to block the Pam3CSK4-mediated enhancement of RSV infection. The compounds were added to cells that had been pre-incubated with Pam3CSK4 or Pam3CSP4 as a control, followed by infection with RSV. The next day percentages GFP-expressing cells were determined by flow cytometry. The strongly polyanionic compounds heparin, dextran sulfate, over-sulfated heparin and over-sulfated chondroitin sulfate were able to abrogate the Pam3CSK4-mediated enhancement of rgRSV infection (p<0.05 using posthoc Dunnett's correction), whereas the control compounds chondroitin sulfate and de-sulfated heparin had only a minimal effect or no effect, respectively. Error bars represent the standard deviation of triplicates. A representative experiment out of three is shown.

Figure 5. Enhancement of infection with other enveloped viruses.

(A–D) Jurkat-CCR5 cells (A) or B-LCL (B–D) were incubated with different lipopeptides (10 µg/ml, Pam3CSK4, Pam-Cys-SK4, PHCSK4, or Pam3CSP4) before infection with HIV-RgBaL-GFP (A), HMPV-EGFP (B), rMV-Edmonston-EGFP (C) or rMV-IC323-EGFP (D), respectively. Two (HIV-1) or one (HMPV, MV) day later the percentage of GFP-expressing cells was determined by flow cytometry. In all experiments Pam3CSK4, Pam-Cys-SK4 and PHCSK4 treatment resulted in significantly higher infection percentages (p<0.05 using posthoc Dunnett's correction for multiple comparisons). Error bars represent the standard deviation of triplicates. A representative experiment out of three is shown.

Figure 6. Mechanism of reduced rgRSV infection of A549 cells in relation to GAG expression.

A549, CHO K-1, CHO pgsA-745 or CHO pgsD-677 cells were infected with rgRSV after pre-incubation of the cells (black symbols) or the virus (white symbols) with Pam3CSK4 (10 µg/ml) for 30 minutes at 37°C. In the GAG-expressing cell lines A549 and CHO-K1 pre-incubation of cells with Pam3CSK4 resulted in reduction of RSV infection, but in the GAG-deficient cell lines CHO-pgsA745 and CHO-pgsD677 this same condition resulted in enhancement of infection (p<0.01 in multiple Student's t-test with Bonferroni correction). Data are presented as percentages GFP-positive cells (means ± SD of triplicates). A representative experiment out of three is shown.