Bacterial strains used in this study are listed in Table 1. All strains were cultured in 1:20 diluted Trypticase Soy Broth (1/20 TSB) at 37°C shaking at 200 rpm. For solid medium, TSB was supplemented with 1.5% agar. Human THP-1 cell lines were cultivated in RPMI-1640 medium containing 10% fetal calf serum at 37°C with 5% CO₂. The C. elegans AU37 strain [glp-4(bn2); sek-1(km4)] was obtained from the Caenorhabditis Genetics Center (CGC) and maintained according to standards (Stiernagle, 2006). The following antibacterials were used: ofloxacin, ciprofloxacin, rifampicin, polymyxin B, vancomycin (Sigma–Aldrich), and 1-[(2,4-dichlorophenethyl)amino]-3-phenoxypropan-2-ol (SPI009; CD3) with concentrations indicated throughout the text.

Antibacterial assays were performed on different clinically relevant pathogens as previously described (Liebens et al., 2017). Briefly, stationary phase cultures were treated for 5 h with 17 or 34 μg/mL of SPI009 alone or in combination with an appropriate antibiotic to assess anti-bacterial and anti-persister effects, respectively. To evaluate activity against resistant strains, stationary phase cultures were treated for 5 h with 1x, 4x, and 8x MIC concentrations of the respective antibiotic; 17 or 34 μg/mL SPI009 or the combination of both. After treatment, cells were washed and viability was assessed via plating.

Overnight cultures of P. aeruginosa PA14 WT or S. aureus ATCC 33591 were diluted 1:100 in 1/20 TSB medium supplemented with 2% DMSO (carrier control) or increasing concentrations of SPI009 (4.25–68 μg/mL). Biofilms were grown for 24 h at 37°C on the bottom of a polystyrene 96-well plate, non-shaking. Medium and free-living cells were removed and the biofilms were washed, scraped off and passed five times through a syringe (0.5 mm × 1.6 mm) to disrupt any cell clumps and obtain single cells (Hermans et al., 2011). Appropriate dilutions made in 1x PBS were plated on solid TSB agar plates to assess biofilm growth under different conditions.

To explore the biofilm eradicating effects of SPI009, overnight cultures of P. aeruginosa PA14 WT or S. aureus ATCC 33591 were diluted 1:100 in 1/20 TSB medium and incubated for 24 h at 37°C (non-shaking). Mature biofilms were treated for 5 h with 2% DMSO and increasing concentrations of SPI009 (8.5–136 μg/mL) at 37°C, non-shaking, after which the remaining biofilms were processed as described above.

A three-dimensional wound biofilm model was used, as previously described (Townsend et al., 2016). P. aeruginosa coated cellulose matrices, obtained after 2 h of adhesion (1 × 10⁶ cells/mL), were placed onto the hydrogels after which 3D biofilm development was allowed for 24 h at 37°C. Mature biofilms were treated for 24 h with DMSO (1%), 10 μg/mL ofloxacin, 34 and 69 μg/mL of SPI009 or the combination of ofloxacin and SPI009. Any non-adherent cells were removed by rinsing after which biomass was removed by sonication at 35 kHz for 10 min and DNA was extracted. Samples were prepared as previously described and viability-based qPCR using P. aeruginosa specific primers F- GGGCGAAGAAGGAAATGGTC and R- CAGGTGGCGTAGGTGGAGAA was used to determine live and total fractions of biofilm cells under different treatment conditions (Smith et al., 2016). Standard curves were used to convert the obtained qPCR values to colony forming estimates (CFEs), after which log₁₀-transformed values were used for statistical analysis, as described below. All experiments were carried out in triplicate, each containing three technical repeats.

Infection of human THP-1 cells was performed as described previously, with minor modifications (Buyck et al., 2013). Since a newly synthesized batch of SPI009 was used for this experiment, cytotoxicity assessment via an LDH enzyme assay was repeated for the THP-1 cell line, as previously described (Liebens et al., 2017). After THP-1 infection with P. aeruginosa PAO1 and subsequent removal of any non-phagocytozed or adherent bacteria, ciprofloxacin and SPI009 were added in final concentrations of, respectively, 0–20 μg/mL and 6.8 or 10.2 μg/mL. After 5 h of treatment, eukaryotic cells were collected in three consecutive centrifugation steps and complete cell lysis was obtained by sonication (10 s). Lysates were used for bacterial CFU counting and determination of protein content by Lowry’s assay (Bio-Rad DC protein assay kit; Bio-Rad laboratories, Hercules, CA, United States). For analysis of surviving bacterial cells, CFU data were divided by corresponding protein content for normalization.

AU37 nematodes were synchronized as previously described (Porta-de-la-Riva et al., 2012) to obtain L4 worms suitable for toxicity and infection assays (Briers et al., 2014). Larvae obtained after bleaching were plated onto solid NGM-OP50 agar plates and incubated at 25°C during 2 days to allow development of the worms to the L4 stage. Worms were transferred to fresh NGM agar plates containing OP50 (toxicity testing and uninfected control) or PA14 (infection) for an additional 24 h at 25°C.

To evaluate toxicity of SPI009 L4 nematodes grown on OP50 were transferred to 12-well plates (20–30 worms/well) containing different concentrations of SPI009 (8.5–136 μg/mL) in 1.5 mL NGM:M9 (1:4). Controls consisted of untreated worms and DMSO (2% and 20%). For the infection assay, adult worms were allowed to feed on NGM-PA14 plates for 24 h, after which residual bacteria were removed and nematodes were divided over a 12-well plate (20–30 worms/well). Different treatments were prepared in 1.5 mL NGM:M9 (1:4) and consisted of an untreated control, 1.56 μg/mL ciprofloxacin (5x MIC), 8.5 μg/mL of SPI009 and the combination of ciprofloxacin and SPI009. As an additional control, uninfected worms were included. For both assays, worms were incubated at 25°C and survival was scored visually for 6 days.

Unless mentioned otherwise, all statistical analyses were performed on log₁₀-transformed data using GraphPad Prism software (version 6.01). Bacterial survival after different treatments was compared to the untreated or antibiotic control using a one-way ANOVA (α = 0.05), with Dunnett’s correction for multiple comparisons. Statistical comparison of mono- and combination treatment in resistant strains was done using a two-way ANOVA (α = 0.05) with Tukey correction for multiple comparisons. Statistical analysis of the in vivo C. elegans data was done by means of a log-rank test using GraphPad Prism.

The activity of SPI009 was previously assessed in P. aeruginosa PA14 and several clinical isolates where combination with ofloxacin significantly decreased the persister fraction in all strains tested (Liebens et al., 2017). In the present study, we challenged a panel of clinically relevant species, including the ESKAPE pathogens (Figure 1A), B. cenocepacia and E. coli (Figure 1B). For each species appropriate concentrations of a conventional antibiotic used in the clinic were selected to allow only persister cells to survive (Supplementary Figure S1). Combination of the antibiotic with 17 μg/mL SPI009 significantly decreased the number of surviving bacteria for five of the eight species with reductions in CFU ranging between 1.5 ± 0.1 and 6.0 ± 0.2 log units and complete eradication of K. pneumoniae. Addition of 34 μg/mL completely eradicated the bacterial cultures of five of the eight species tested and resulted in significant 6.6 ± 0.5 log, 6.2 ± 1.3 log, and 5.4 ± 0.5 log reductions in bacterial survival for S. aureus, E. faecium, and B. cenocepacia, respectively. No reduction in survival is observed after treatment with 17 μg/mL for either of the Gram-positive species, E. faecium and S. aureus. These results suggest that the latter two species, and the Gram-negative B. cenocepacia, are slightly less sensitive toward the combination therapy. K. pneumoniae proved the most susceptible species toward SPI009. Overall, the obtained results further support the antibacterial effect of SPI009 and reveal a broad-spectrum activity.

To investigate the possible use of SPI009 as an adjuvant in antibacterial combination therapies, several (multi)drug-resistant strains were treated with 1x, 4x, and 8x MIC concentrations of the antibiotic, alone and in combination with SPI009. While SPI009 alone did not cause a significant decrease in survival of the ofloxacin resistant P. aeruginosa PA62, addition of 17 or 34 μg/mL of SPI009 significantly reduced the number of surviving cells by 5.3 ± 0.9 and 7.8 ± 0.9 log units at 4x MIC of ofloxacin while combination with 8x MIC completely eradicated the bacterial culture (Figure 2A). In comparison, treatment with ofloxacin alone caused 0.8 ± 0.9 log and 2.8 ± 0.9 log decreases in surviving cells at concentrations of 4x MIC and 8x MIC, respectively.

A similar trend was observed in the polymyxin B resistant P. aeruginosa 9BR (Figure 2B). Here, addition of the antibiotic alone had a slightly greater effect but combination with SPI009 still significantly improved the treatment and 17 μg/mL of SPI009 successfully eradicated the entire bacterial culture in combination with 4x MIC of polymyxin B. A somewhat smaller effect was observed in the polymyxin B resistant B. cenocepacia strain K56-2, for which addition of 17 μg/mL and 34 μg/mL SPI009 to 4x MIC polymyxin B resulted in significant 4.9 ± 0.5 and 5.2 ± 0.5 log decreases in survival. Combinations with higher concentrations of polymyxin B (8x MIC) did not further decrease the number of surviving cells (Figure 2C). The obtained results clearly demonstrate the effective use of SPI009 as an adjuvant for antibacterial therapy thereby facilitating the treatment of different antibiotic-resistant strains. Furthermore, SPI009 retains activity in multidrug-resistant strains, revealing the lack of cross-resistance. Importantly, resensitization of resistant strains could restore the effectiveness of established antibiotics.

To assess biofilm inhibiting properties of SPI009 in P. aeruginosa and S. aureus, biofilm growth was allowed in the presence of increasing concentrations of SPI009 (Figure 3A). Analysis of the obtained results clearly show an effective inhibition of biofilm growth in both P. aeruginosa and S. aureus. For P. aeruginosa, a steep increase in inhibitory activity was observed at concentrations above 8.5 μg/mL, resulting in 1.8 ± 0.5 log and 2.4 ± 0.4 log decreases at 17 or 34 μg/mL SPI009, respectively, and complete inhibition of biofilm growth at 68 μg/mL. S. aureus showed a more gradual decrease in biofilm formation with 34 μg/mL and 68 μg/mL resulting in significant 6.2 ± 0.6 log and 6.4 ± 0.6 log decreases in biofilm formation, respectively. These results clearly demonstrate the potent biofilm inhibiting activity of SPI009 for both Gram-negative and Gram-positive model pathogens.

To explore biofilm eradication, SPI009 was added to mature biofilms and survival was assessed after 5 h of treatment. For P. aeruginosa the lower concentrations (8.5 and 17 μg/mL) caused a decrease in biofilm survival of about 0.8 log units (Figure 3B). Doses of 34 μg/mL or higher significantly decreased the number of surviving biofilms cells, resulting in 4.2 ± 0.6; 6.2 ± 0.6; and 6.6 ± 0.6 log reductions. In comparison, 10 μg/mL of the conventional antibiotic ofloxacin caused a significant 4.5 ± 1 log decrease in the number of surviving biofilm cells (Supplementary Figure S2A). For S. aureus, the treatment of mature biofilms with lower concentrations of SPI009 proved slightly less effective than for P. aeruginosa. Treatment with higher concentrations did cause extensive damage, resulting in significant decreases in biofilm survival ranging between 2.5 ± 0.7 and 5.4 ± 0.6 log. For the 96-well biofilm models used in this study, the combination of SPI009 with a conventional antibiotic did not further decrease the number of surviving cells as compared to mono-treatment with SPI009 (Supplementary Figure S2). Overall, SPI009 shows potent activity in biofilms of both Gram-negative and Gram-positive species and is capable of significantly inhibiting biofilm formation and decreasing survival of mature biofilms.

After confirming the biofilm eradication capacity of SPI009 in a standard biofilm set-up, a more clinically relevant model was used to assess the clinical potential of SPI009 as a topical antibacterial treatment. Using a porous cellulose matrix placed upon a moist hydrogel allowed the growth of a complex, three-dimensional hydrated structure, effectively mimicking biofilms in a chronic wound environment (Townsend et al., 2016; Kean et al., 2017). Assessment of viability was performed by means of live/dead quantitative PCR (Figure 4). For the viable cells, treatment with increasing concentrations of SPI009 alone resulted in significant 1.6 ± 0.5 log (34 μg/mL) and 2.0 ± 0.5 log (68 μg/mL) decreases in the number of surviving cells. The obtained results confirm the biofilm eradication capacity of SPI009, both as an antimicrobial and as part of a combination therapy, and this in a more complex, realistic biofilm environment.

Next, the anti-persister and antibacterial activities of SPI009 were verified in a recently developed P. aeruginosa intracellular infection model (Buyck et al., 2013). Human THP-1 cells were infected with PAO1 cells (MOI 10) and treated for 5 h with different concentrations of ciprofloxacin, alone or in combination with 6.8 or 10.2 μg/mL of SPI009. Concentrations of SPI009 were chosen to be well below the determined IC₅₀ value of 24.5 ± 1.36 μg/mL. After treatment, both the number of surviving PAO1 cells and the amount of eukaryotic proteins present was assessed, as this can provide information about the possible toxic effect of the different treatments and the infecting bacteria. While treatment with SPI009 alone caused non-significant decreases of 0.78 ± 0.7 and 0.89 ± 0.7 log units in surviving bacteria, addition of SPI009 to ciprofloxacin greatly improved the antibacterial effect for all concentrations tested and this in a dose-dependent manner (Figure 5). Maximal antibacterial activity for the combination therapy with 10.2 μg/mL of SPI009 occurs at ciprofloxacin concentrations of 10 μg/mL, resulting in complete eradication of the bacterial culture. Moreover, all combinations tested significantly reduced the bacterial load as compared to ciprofloxacin alone. Combination treatment with 6.8 μg/mL SPI009 showed a maximal 0.78 ± 0.6 log decrease as compared to antibiotic alone at a ciprofloxacin concentration of 20 μg/mL. These results clearly show that SPI009 can effectively penetrate the eukaryotic cell membrane, without causing extensive damage, to eradicate the intracellular P. aeruginosa infection.

Since the antibacterial effect of SPI009 was demonstrated extensively in vitro, a next step was to assess the effect of this new compound in an in vivo C. elegans gut infection model. Toxicity testing of SPI009 in C. elegans revealed minor levels of toxicity at 68 μg/mL and >80% killing at 136 μg/mL (Supplementary Figure S3), excluding these concentrations from further experiments. Analysis of the different DMSO concentrations suggests that the observed toxicity is mainly caused by increasing concentrations of the solvent.

Infection of nematodes with PA14 resulted in 91.5% killing within 6 days after the start of infection, confirming the highly virulent nature of the PA14 strain in this model (Figure 6). Addition of 8.5 μg/mL of SPI009 alone slightly improved survival but not as good as 5x MIC of ciprofloxacin, resulting in survival rates of 19.0% (P = 0.045) and 46.6% (P < 0.0001), respectively. However, addition of 8.5 μg/mL of SPI009 to ciprofloxacin greatly increased survival, resulting in 73.8% nematode survival after 6 days. These results show a significant improvement in antibacterial effect of the combination therapy compared to the untreated (P < 0.0001) and ciprofloxacin-treated (P = 0.0001) controls (Supplementary Table S1). Since low doses of SPI009 can greatly enhance the effect of conventional antibiotic treatment, resulting in more than 73% survival, these results indicate the highly efficient antibacterial and potentiating effect of SPI009 as part of a combination therapy.