The agr Inhibitors Solonamide B and Analogues Alter Immune Responses to Staphylococccus aureus but Do Not Exhibit Adverse Effects on Immune Cell Functions

Abstract

Staphylococcus aureus infections are becoming increasingly difficult to treat due to antibiotic resistance with the community-associated methicillin-resistant S. aureus (CA-MRSA) strains such as USA300 being of particular concern. The inhibition of bacterial virulence has been proposed as an alternative approach to treat multi-drug resistant pathogens. One interesting anti-virulence target is the agr quorum-sensing system, which regulates virulence of CA-MRSA in response to agr-encoded autoinducing peptides. Agr regulation confines exotoxin production to the stationary growth phase with concomitant repression of surface-expressed adhesins. Solonamide B, a non-ribosomal depsipeptide of marine bacterial origin, was recently identified as a putative anti-virulence compound that markedly reduced expression of α-hemolysin and phenol-soluble modulins. To further strengthen solonamide anti-virulence candidacy, we report the chemical synthesis of solonamide analogues, investigation of structure-function relationships, and assessment of their potential to modulate immune cell functions. We found that structural differences between solonamide analogues confer significant differences in interference with agr, while immune cell activity and integrity is generally not affected. Furthermore, treatment of S. aureus with selected solonamides was found to only marginally influence the interaction with fibronectin and biofilm formation, thus addressing the concern that application of compounds inducing an agr-negative state may have adverse interactions with host factors in favor of host colonization.

Fig 1. Compound and AIP Structures.

Structures of autoinducing peptides (AIP-I and III) used in the study as well as tested depsipeptides and their modified lactam analogues.

Fig 2. Biological validation of the synthetic solonamides (a) and lactam analogues (b).

Agar plates containing the hla–lacZ (PC322), the rnaIII-lacZ (SH101F7) or spa–lacZ (PC203) reporter strains of S. aureus were exposed to DMSO (20 μL) containing the test compound (0.5 mg/mL). Vehicle (DMSO), H₂O, AIP-I (autologous) and AIP-III (heterologous) were used as controls. Virulence gene down-regulation is represented by a white zone and up-regulation by a darker than background blue zone.

Fig 3. Solonamides (A) and lactam analogues (B) differentially interfere with AgrC activation as monitored by direct RNAIII expression.

Cultures of RN10829 (P2-agrA:P3-blaZ) containing the pagrC-I-WT, were grown to an OD₆₀₀ of 0.4–0.5 where a 1/10 volume of AIP-I containing supernatant was added and solonamides and analogues in DMSO to a final concentration of 10 μg/mL. Samples obtained at 15 min time intervals after addition of test solutions were analysed for β-lactamase activity. Each bar represents the average of 3 replicates and the error bars represent the standard deviation. Comparisons were made for each individual time point between AIP and AIP + Compound samples. ns (no significance); *, p<0.05; **, p<0.01; ***, p<0.001.

Fig 4. Solonamides marginally enhance adhesion to the extracellular matrix component fibronectin.

Upon exposure of S. aureus strains to 10 μg/mL anti-virulence compounds or vehicle control, attachment to fibronectin was measured in 96-well plates coated with 10 μg/mL fibronectin and staining with 0.1% crystal violet. Absorbance values were corrected for cell density and represented as arbitrary binding units. For statistical significance, comparisons were made between untreated versus vehicle and treated (black bracket), and between vehicle control versus compound treated (red square). ns (no significance);*, p<0.05; **, p<0.01; ***, p<0.001.

Fig 5. Solonamides marginally enhance biofilm formation of WT strain 8325–4.

A dilution series of an 8325–4 culture were inoculated (5 μL) into wells of a 96-well microtiter plate containing 200 μL TSB. SolB (A), ESB (C) or Am16-L (C) was added to final concentrations of 5, 10, 20, 40, and 80μg/mL. Inoculum alone, DMSO and no inoculum were used as controls. Biofilm formation was assessed by crystal violet staining and OD590 nm measurement. Each bar represents the average of 3 experiments, and the error bars represent the standard deviation. For statistical significance, comparisons were made between untreated versus vehicle and treated (black bracket), and between vehicle control versus compound treated (red square). ns (no significance);*, p<0.05; **, p<0.01; ***, p<0.001.

Fig 6. Solonamide-treated S. aureus influence cytokine production by immune cells.

Bone marrow derived murine dendritic cells (A) and human peripheral blood mononuclear cells (B) were stimulated at an MOI of 10 with S. aureus strain 8325–4 pre-treated with SolB (2) ESB (4), Am16-L (9) or DMSO at 10 μg/mL. Protein concentrations of selected cytokines in the supernatants after 20–24 h incubation were measured by enzyme-linked immunosorbent assay for DCs and by cytometric bead array for human PBMCs. The results are based on 3 biological replicates and are normalized to the stimulus controls (LPS for DC; L. plantarum WCFS1 for PBMCs). Comparisons were made between the DMSO compound vehicle and the compounds.

Fig 7. SolB, ESB and Am16-L- reduce interference of T-cell proliferation by S. aureus.

CFDA/SE (25μg/mL) stained human PBMCs (1x10⁶/ml) were co-stimulated with aCD3/aCD28 (0.4ng and 0.024ng/mL respectively) and S. aureus either treated or not with 10μg/mL SolB, ESB or Am16-L at an MOI of 10 for 4 days. Harvested cells were stained with aCD4-PE and analyzed by Flow Cytometry using FACS Diva software. Lymphocytes were gated based on the expression of CD4, and the number of cell divisions was gated according to the CFDA/SE FITC excitation. The data represent the average of 3 donors, and the error bars represent the standard deviation.