5-Benzylidene-4-Oxazolidinones are Synergistic with Antibiotics for the Treatment of Staphylococcus aureus Biofilms

Abstract

The failure of frontline antibiotics in the clinic is one of the most serious threats to human health and requires a multitude of novel therapeutics and innovative treatment approaches to curtail the growing crisis. In addition to traditional resistance mechanisms resulting in the lack of efficacy of many antibiotics, most chronic and recurring infections are further made tolerant to antibiotic action by the presence of biofilms. Herein, we report an expanded set of 5-benzylidene-4-oxazolidinones that are able to inhibit the formation of Staphylococcus aureus biofilms, disperse preformed biofilms and in combination with common antibiotics are able to significantly reduce the bacterial load in a robust collagen-matrix model of biofilm infection.

Graphical Abstract

The pervasiveness of multi-drug resistant bacteria is a human health issue that demands continuous development of new antibiotics and unique approaches to combat antibiotic failure in the clinic.^1,2 Few novel classes of antibiotics have been discovered since the golden age of antibiotics, making the options for treatment more and more limited as resistance continues to emerge worldwide. The “ESKAPE” pathogens (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, Enterobacter sp.), which were identified by the CDC as posing the greatest threat to human health, are primary targets for antibiotic development.^3–5 Although significant focus has been directed to gram-negative ESKAPE pathogens in recent years, gram-positive pathogens such as Staphylococcus aureus (S. aureus) and methicillin-resistant S. aureus (MRSA) remain a serious clinical issue with some strains exhibiting resistance to all known classes of antibiotics.⁶ The rapid development of resistance to even recently developed antibiotics and the lack of novel antibiotic scaffolds under development represents an opportunity for not only the identification of new antibiotic scaffolds, but also the exploration of non-traditional approaches to overcome antibiotic failure.⁷

In addition to the common resistance mechanisms bacteria possess (efflux pumps, active-site mutagenesis, enzymatic degradation, etc.), bacteria such as MRSA can also gain tolerance to antibiotics through the formation of biofilms^8,9. A biofilm is a surface-associated community of bacteria encased in an extracellular polymeric matrix of polysaccharides, DNA and proteins, which offers protection from environmental threats, host immune response and antimicrobial agents.^10,11 The phenotypic shift from planktonic bacteria to a surface attached community of bacteria is a unique aspect of biofilms that is heavily regulated genetically and occurs with decreased rates of growth.¹² Inside these biofilms are slow-growing or dormant (persister) cells which are tolerant to antibiotics and are one of the primary causes of chronic and reoccurring infections.^9,13–17 Persister cells are metabolically inactive which makes these cells inherently resistant to antibiotics that target dynamic processes. These persister cell and biofilm properties, as well as other inherent bacterial fitness strategies, allow biofilm formation and bacterial survival on abiotic surfaces such as implanted medical devices, hospital equipment, and clothing which increases rates of secondary infections.¹⁸ As a result, bacterial biofilms have been implicated in many human bacterial infections ranging from dental plaques to cystic fibrosis, and it is likely that the role of biofilms in many infections has been underestimated.^19,20 Given the reduced efficacy of current antibiotics against bacterial biofilms (100–1000× less effective), it is essential to identify novel classes of molecules that are capable of inhibiting and/or eradicating robust biofilms.²¹

There have been several novel classes of small molecules that inhibit, disperse, eradicate, or modulate MRSA and other bacterial biofilms, often based off of natural product leads.^22–25 Representative examples include the 2-aminoimidazoles (2AI),^26–28 meridianin D analogs,²⁹ halogenated quinolines,^30,31 hydroxybenzylidene-indolinones,³² bromophenol-thiohydantoins,³³ and 2-dichloroalkyl-5-benzylidene-4-oxazolidinones,³⁴ among others. During our initial investigation into the 2-dichloroalkyl-4-oxazolidinones,³⁴ and as part of our broader program to develop novel 4-oxazolidinone antimicrobial agents,^34–38 we identified potent biofilm inhibiting and dispersing agents and established an initial SAR for these synoxazolidinone natural product analogs (Figure 1a). Herein, we expand on the SAR of these synoxazolidinone analogs³⁹ and explore their ability to act in concert with common antibiotics against S. aureus biofilms (Figure 1b).

Figure 1.

4-oxazolidione biofilm modulators.

In order to expand upon our previously developed acid-promoted dehydration/cyclization method,³⁷ we set out to determine the scope for the reaction in terms of aldehydes, which was previously optimized for hexanal and various 2-oxa-3-arylpropan- amides (Figure 2a). To our delight, a number of alkyl-aldehydes underwent smooth conversion to the target oxazolidinones with both the dibromomethoxy- and trifluoromethyl-substituted phenyl pyruvic acid amides. In total, 23 novel analogs (1-24, Figure 2b) were prepared incorporating branched alkyl, aryl, alkenyl, and carbocyclic substituents. Unfortunately, under a variety of conditions, aromatic aldehydes, heteroatom-functionalized aldehydes, and α-halogenated aldehydes did not yield the corresponding oxazolidinone product even under forcing conditions. One exception was 4-nitrobenzaldehyde which underwent reaction with 2-oxo-3-(4-(trifluoromethyl)phenyl)propanamide in a pressurized vessel at 100 °C providing oxazolidinone 21. As demonstrated for compound 1, reactions with formaldehyde provided the aminal product, incorporating an additional formaldehyde unit, which was surprisingly stable to purification (SiO₂) and handling. Several additional aminal products were also prepared to explore the role of this group biologically as a second reaction step (22-24, Figure 2b). With these analogs in hand we began to explore their antimicrobial and anti-biofilm properties.

Figure 2. Synthesis and biological activity.

a) Synthetic scheme to access 4-oxazolidinones; b) analogs prepared; c) MIC and biofilm data. S. aureus ATCC: 29213, MRSA ATCC: BAA-44, A. baumannii ATCC: 19606. Biofilm studies conducted on BAA-44 and 25923.

To begin evaluating this new panel of analogs we sought to define the MICs for these compounds against S. aureus, MRSA and A. baumannii (Figure 2c). From our previous work on synoxazolidinone A and analogs we did not expect to achieve particularly potent compounds, but instead were looking to explore their anti-biofilm activities relative to their MICs. Indeed, a majority of the compounds prepared lacked activity across all three pathogens, with only compounds 15 and 24 providing promising levels (4 μg/mL) of antimicrobial activity, albeit only against the gram-positive pathogens. While these levels of activity are not yet at a clinically useful level, it is important to highlight that these compounds are not halogenated as the compounds in our previous work and represent simplified starting points for further optimization. Compound 24, bearing the aminal functionality, performed the best in this initial screen, although the lifetime of this functional group in the biological assays is currently unknown. Of the compounds tested against A. baumannii, only compound 22 displayed activity at 128 μg/mL.

We next tested the biofilm inhibition and dispersion activities of select compounds, looking to evaluate how these analogs compared with already potent dichlorinated oxazolidinone derivatives prepared previously.³⁴ Most of the alkyl derivatives tested possessed low levels of inhibition and dispersion (5, 7, 9, 10, 12, 14, 18, 22 and 23), with none of these new analogs reaching 50% activity at 40 μM (Figure 2c). Importantly, 15 and 24, the two most active compounds in our MIC assays, were able to inhibit and disperse biofilms in a dose-dependent manner. As these compounds have improved MICs, we rationalized that at least part of this activity was due to the antimicrobial action of these analogs; however, this alone does not account for the observed dispersion activity as the minimum biofilm eradication concentration (MBEC) for both 15 and 24 is 256 μg/mL (Figure 2c). Based on our previous observations, and the activities presented herein, we postulated that the active oxazolidinones could work synergistically with existing antibiotics thereby significantly lowering the MBEC. To this end we explored the MBEC of doxycycline (MBEC = 256 μg/mL) upon treatment with 15 and 24 and found that the oxazolidinones reduced the MBEC to 16 μg/mL – a 16-fold decrease. Following these MBEC experiments with a checkerboard assay to determine synergy, we calculated a FIC of 0.16 for 15 and 0.19 for 24, highlighting the synergistic nature of this compound combination (FIC < 0.5 = synergistic).⁴⁰

Having demonstrated that 15 and 24 work synergistically with doxycycline we aimed to explore the efficacy of these compounds in a more robust biofilm model that better reflects the biofilm composition in wounds (48 h aged biofilms in collagen coated plates)⁴¹. We first tested the activity of 24 in this biofilm model and imaged the living (Syto9, green) and dead (PI, red) cells across a concentration range of 0.25 μM to 160 uM (Figure 3a); as observed, there is little activity up to 10 μM and subsequent increasing activity up to 160 μM. As the MIC is in this 10–20 μM range this data compares well with that collected earlier and also is consistent with the MBEC data as there is still viable bacteria in the biofilm even at 160 μM. We then evaluated the log reduction of viable bacteria by the treatment of these biofilms with a series of commonly employed antibiotics and 24 at 40 μM (Figure 3b). Antibiotics alone did not reduce the bacterial load >2 log CFU/mL at 10× MIC (within the pharmacokinetic range of the chosen antibiotics). However, when these antibiotics were combined with 24 bacterial load was decreased > 2 log CFU/mL. These results were confirmed by imaging the biofilms using a live/dead stain and imaging with a fluorescence microscope. We found that combining 24 with any antibiotic tested herein resulting in increased staining of dead cells (red), with the doxycycline combination being the most effective. These results are significant as the biofilms employed in these more robust assays are less metabolically active and therefore less subject to the action of common antibiotics than those in the MBEC assay discussed previously.

Figure 3. Effects of 24 on mature biofilms.

a) Treatment across a dose range and subsequent imaging; b) log reduction of viable bacteria after treatment for 24 hours; c) Imaging of biofilms with antibiotic alone (10× MIC) or in combination with 24 (40 μM). Bars are means and standard deviations of three technical replicates and significant differences (p<0.05) as determined by ANOVA with Tukey post-hoc are indicated by differing letters.

In conclusion, we have demonstrated that the 4-oxazolidinone scaffold present in the synoxazolidinone natural products is capable of working synergistically with commonly employed antibiotics to significantly reduce the viable bacteria in robust S. aureus biofilms. The compounds reported herein represent starting points for the development of novel agents to combat gram positive biofilms, currently lacking effective treatment options. Current efforts are focused at determining the mechanism of action of this class of heterocycles and further improving their potency.