Multi-drug resistant Staphylococcus aureus, including methicillin-resistant S. aureus (MRSA), has become a worldwide, major health care problem. While initially restricted to clinical settings, drug resistant S. aureus is now one of the key causative agents of community-acquired infections. We have previously demonstrated that copper dependent inhibitors (CDIs), a class of antibiotics that are only active in the presence of copper ions, are effective bactericidal agents against MRSA. A second-generation CDI, APT-6K, exerted bactericidal activity at nanomolar concentrations. At sub-bactericidal concentrations, it effectively synergized with ampicillin to reverse drug resistance in multiple MRSA strains. APT-6K had a favorable therapeutic index when tested on eukaryotic cells (TI: > 30) and, unlike some previously reported CDIs, did not affect mitochondrial activity. These results further establish inhibitors that are activated by the binding of transition metal ions as a promising class of antibiotics, and for the first time, describe their ability to reverse existing drug resistance against clinically relevant antibiotics.
The rapid increase of antibiotic resistance within bacterial populations is associated with longer hospital stays, increased treatment costs, and more patient deaths1,2. An estimated 700,000 individuals die each year as a result of infections with antibiotic resistant bacteria, and the amount of deaths are expected to increase if no alternative, effective therapies are developed3. New antibiotics are essential to avoid a public health crisis. The identification of new antibiotics for Staphylococcus aureus is an especially urgent task, with antibiotic resistance in this bacterium already observed against some of the last line of defense antibiotics such as vancomycin, linezolid, and daptomycin4. An alternative to the discovery of new antibiotics are drugs that restore the efficacy of available antibiotics and overcome bacterial drug resistance mechanisms. Ideal drugs would be ones that are both effective by themselves and that restore the activity of current antibiotics by reversing antibiotic resistance5,6.
Copper dependent inhibitors (CDIs) are a functionally new type of antibiotic gaining increased appreciation due to their ability to inhibit drug resistant bacteria such as S. aureus, Mycobacterium tuberculosis, Mycoplasma spp., and Neisseria gonorrheae7,8,9,10,11,12,13. These compounds utilize copper for their activities and include the FDA approved drug disulfiram and anti-cancer compounds like 8-hydroxyquinoline (8HQ)7,9,13,14,15. Hundreds of new CDIs with antibacterial and antifungal activity have been identified in drug screens against S. aureus, M. tuberculosis, and Cryptococcus neoformans using defined culture medium that contains physiologically relevant concentrations of copper that were previously not identified in these compound libraries when screened under industrial standard conditions (no consideration of transition metal concentrations), demonstrating the untapped potential of CDIs10,12,15,16.
At present, it is unclear whether CDIs target a shared bacterial pathway or whether they target a large array of different functionalities. Some investigations have shown that CDIs have the ability to shut down different ATP generating processes such as oxidative phosphorylation and glycolysis11,17. Studies by others have shown that inhibition of ATP generation can restore sensitivity to different antibiotics in drug resistant bacteria. Examples of this phenomenon include increasing the sensitivity of S. aureus to polymyxins with the ATP synthase inhibitor oligomycin A or improving the efficacy of β-lactam antibiotics against Mycobacterium tuberculosis with the electron transport chain inhibitors 2-aminoimidazoles (2-AIs)18,19. Given reports that some CDIs affect ATP generation, we tested a second-generation CDI called APT-6K and found that it has rapid bactericidal activity in the presence of copper and greatly reduces ATP concentrations prior to cell death. We demonstrate the ability of APT-6K to overcome pre-existing drug resistance in S. aureus and that APT-6K, at concentrations that exert no anti-bacterial effect, restored the activity of ampicillin in resistant MRSA isolates.
Results and Discussion
APT-6K is a potent copper-dependent inhibitor of S. aureus
A previous compound screen identified a group of antibiotics that exerted potent anti S. aureus activity and was characterized by a nitrogen-nitrogen-sulfur-nitrogen (NNSN) motif forming the structural backbone (Fig. 1a, green circle). These compounds only exhibited antibiotic activity in the presence, but not the absence, of copper10. A sub-group of NNSN compounds, which we described as adamantyl-bearing pyrazolyl-thioureas (APT), were further investigated for their activity against S. aureus. These were of particular interest as adamantyl-groups have been reported to convey stability to compounds, a desirable feature for antibiotics20. The inhibitor APT-6K (Fig. 1a) is a second generation copper dependent APT with a minimum inhibitory concentration (MIC) of 150 nM on S. aureus strain Newman in the presence of 50 µM copper, the transition metal concentration that was used in the drug screen (Fig. 1b, blue circles). Of note, copper concentrations in serum range between 10–20 µM and can reach 400 µM within phagolysosomes, where copper ions are part of a physiological anti-bacterial defense mechanisms, but in our experiments 50 µM copper alone is not growth inhibitory21,22,23. The inhibitor APT-6K was found to be highly copper-specific in its anti-bacterial activity, as no other transition metals would activate the compound (Fig. 1b).
To determine the minimal copper concentration required for APT-6K to exert an anti-bacterial activity, copper and APT-6K were titrated against each other (microplate assay) in suspension cultures and following overnight incubation, were transferred to nutrient rich agar plates. No measurable bacterial growth in liquid culture was observed starting at APT-6K concentrations of 1.25 µM or higher in combination with copper concentrations as low as 2.5 µM (Fig. 1c; blue circles). Transfer of these cultures to agar plates enabled bacterial outgrowth, indicating that the effect had been bacteriostatic (Fig. 1d). Bactericidal activity required a minimum copper concentration of 20 µM but was achievable at lower APT-6K concentrations (Fig. 1d; black box). For either bacteriostatic or bactericidal effects, the required copper concentrations remained well within the physiologically relevant range.
Time to death experiments were performed where APT-6K was titrated in the presence of 50 µM copper into bacterial cultures to determine the length of time needed for APT-6K to exert its bactericidal activity. Samples were removed at the indicated time points and transferred to nutrient rich agar plates for bacterial outgrowth. A reduction in bacterial viability could be observed as early as two hours following APT-6K addition, and after five hours of treatment with 300 nM APT-6K, no viable bacteria could be recovered (Fig. 1e). In addition, there was an observable reduction in bacteria at APT-6K concentrations as low as 30 nM.
Previous work had shown that some CDIs can affect different ATP generating processes11,17,24. To understand whether APT-6K would act through this mechanism, we measured intracellular ATP concentrations after APT-6K treatment in the presence of 50 µM copper. ATP-6K was indeed extremely potent at interfering with ATP generation. After one hour of treatment with carbonyl cyanide m-chlorophenyl hydrazone (CCCP; 10 µM), an established inhibitor of oxidative phosphorylation in S. aureus that is frequently used as a positive control, ATP levels had dropped to 40% of the untreated control25. During the same period, treatment with as little as 300 nM APT-6K reduced ATP concentrations to <10% when compared to the ATP concentration in untreated bacteria (Fig. 1f), suggesting that interference with ATP generation is contributing to the antibiotic effect of APT-6K.
APT-6K is well tolerated by eukaryotic cells
An essential requirement for any compound to be considered as a potential antibiotic lead is a relevant therapeutic index indicating the absence of toxicity against eukaryotic cells at concentrations that exert antibacterial activity. To determine APT-6K toxicity against human cells, APT-6K in the absence and presence of copper was titrated on THP-1 cells, a human monocytic cell line, and metabolic activity as a surrogate for cell viability was determined 24 hours post treatment using resazurin. By itself, APT-6K exerted no toxicity up to the maximum tested concentration of 10 µM. In the presence of copper, the toxic concentration (TC90) was determined to be 5 µM (Fig. 2a). Toxicity experiments using peripheral blood mononuclear cells (PBMCs) from six healthy human donors produced an identical TC90 when metabolic activity was used as a readout (Fig. 2b). Flow cytometric analysis of T cells treated with 1.25 µM or less APT-6K plus copper suggested no impairment of protein synthesis based on the expression levels of constitutively expressed T cell markers, such as CD3, CD4, CD8, or CD28 (Supplementary Figures 1–3). Higher concentrations of APT-6K reduced cell surface marker expression in some of the tested donors. APT-6K did not induce the expression of any T cell activation markers (CD25, CD38, CD69, and PD1) or any markers that would indicate pro-inflammatory activity (TNF-α, IFN-γ, and granzyme B) (Supplementary Figures 4 and 5), another important characteristic for lead compounds. In summary, these data suggest an initial therapeutic index of 16–32 relative to the determined MIC90 against S. aureus (0.15 µM), which is well above the therapeutic index threshold of 10 that is commonly used to consider compounds for initial lead optimization26.
Previous work by others has described the eukaryotic toxicity of some CDIs as being, at least in part, due to mitochondrial inhibition. We thus tested for possible effects of APT-6K in the presence of copper on membrane integrity and mitochondrial function as a mechanism of the observed toxicity of APT-6K against eukaryotic cells at high concentrations (>3 µM). For these experiments we used glyoxal-bis(N(4)-methyl-3-thiosemicarbazone (GTSM), a comprehensively characterized copper-dependent antibacterial compound reported to inhibit Complex I within the electron transport chain of both prokaryotes and eukaryotes, as a defined CDI control24. CCCP, a mitochondrial oxidative phosphorylation uncoupler27, and digitonin, a mild nonionic detergent frequently used to permeabilize cell membranes28, served as positive controls for mitochondrial toxicity and membrane integrity effects, respectively. The ToxGlo (Promega) mitochondrial toxicity assay allows for the quantification of both effects. Briefly, a quenched fluorogenic peptide substrate is added (bis-alanyl-alanyl phenylalanyl- rhodamine 110; bis-AAF-R110) that cannot permeate through an intact membrane but following permeation through a compromised cell membrane, it is proteolytically cleaved by a distinct necrosis-associated protease, triggering a fluorescent signal. A second step in the assay determines mitochondrial toxicity by quantifying ATP levels produced under the respective experimental conditions.
As seen in Fig. 2c, 90 minutes after compound addition, membrane integrity was not affected by either GTSM or APT-6K in the presence of copper, or by CCCP. As expected, the membrane integrity was compromised by digitonin, acting as the positive control.
In the absence of any membrane damage, GTSM in the presence of copper and CCCP resulted in a sharp decrease in ATP production in THP-1 cells. In contrast, APT-6K treatment in the presence of copper did not result in any significant loss of ATP production, indicating the absence of mitochondrial toxicity (Fig. 2d). While this does not exclude secondary mitochondrial toxicity effects of APT-6K at later time points, it is an exciting finding as the observed mitochondrial toxicity of GTSM most likely contributes to its reported in vitro toxicity17. Despite its apparent cell culture toxicity, GTSM is tolerated in mice and has reported anti-cancer and anti-Alzheimer’s activity29,30,31. As such, the results for APT-6K are promising as they suggest that reduced in vitro mitochondrial toxicity may translate into improved therapeutic potential.
Antibiotic effect of APT-6K on multidrug resistant S. aureus
Another criterium for the development of novel antibiotics is their performance against bacterial strains with pre-existing antibiotic resistances. To determine whether the antibiotic activity of APT-6K was maintained against MDR/MRSA strains, we tested a panel of multi-drug resistant clinical S. aureus isolates (Supplementary Table S1). While some of these isolates, represented by MRSA-1 and −2, were found to be sensitive to copper-activated APT-6K in a concentration range similar to what was observed for the Newman laboratory strain (MIC = 300 nM) (Fig. 3a,b), we identified two isolates, MRSA-3 and −4, that naturally had an increased tolerance to APT-6K with MIC90s of 20 µM and 5 µM, respectively (Fig. 3c,d). The inhibitory effect of APT-6K on the two sensitive MRSA isolates could be titrated, but the response to APT-6K on the resistant strains was bi-phasic. We observed an initial inhibitory effect at low concentrations (<1 µM) that never suppressed bacterial growth below the MIC90 cut-off (Phase 1). Additional increases in APT-6K concentrations rendered APT-6K initially less effective (Fig. 3c,d), before the second onset of APT-6K antibacterial activity at even higher concentrations (Phase 2). It is important to emphasize that APT-6K at concentrations below 1 µM still reduced viability of these MRSA isolates by ~75%.
For other CDIs, we have previously measured a small increase in the effective concentration in some MRSA strains, but this is the first time we observed pre-existing, natural tolerance against a CDI in select MRSA strains. Recent publications have highlighted two previously unknown copper resistant proteins in a MRSA isolate (USA-300/JE2) termed CopX and CopL, in addition to the already reported copper resistance pump CopA and the copper chaperone CopZ32,33. CopA and CopZ are found in every S. aureus strain, while the presence of CopX and CopL seems to be limited to MRSA isolates32,34, making them candidate genes for the observed APT-6K tolerance phenotype of MRSA-3 and MRSA-4.
We first tested the four clinical MRSA strains, USA-300/JE2 (the MRSA strain used to describe copX and copL), and the laboratory Newman strain for the presence of copA, copZ, copX and copL genes. PCR revealed that, as expected, all strains/isolates had the copA and copZ locus (Fig. 4a and Supplementary Figure 6). However, Newman and the APT-6K sensitive MRSA-1 and MRSA-2 isolates had neither the copX nor the copL genes. In contrast, USA300/JE2, as reported by others32,34, and the two APT-6K tolerant MRSA isolates had the copX and copL genes, suggesting a possible correlation of APT-6K resistance with the presence of CopX and/or CopL proteins (Fig. 4b and Supplementary Figure 7). Copper titrations on the different S. aureus strains/isolates tracked with the presence of copX and copL. While MRSA-1 and MRSA-2 were only copper tolerant up to a concentration of 125 µM, USA300/JE2 and the copX/copL-positive MRSA-3 and MRSA-4 were copper resistant with a MIC of 500 µM (Supplementary Figures 8).
This prompted us to specifically test the effects of the various copper resistant genes in the JE2 MRSA strain, for which transposon mutants with inactivated copA, copZ, copX or copL are available. As seen in Fig. 4c, individual inactivation of copA, copX, or copL all reduced copper tolerance relative to the parental USA-300/JE2 strain, suggesting that these proteins provide non-redundant activities. Interestingly, inactivation of copZ had no effect on copper tolerance in this experimental system. We next tested the sensitivity of the USA-300/JE2 strains to APT-6K as a function of the presence of copA, copZ, copX or copL. As expected, the parent USA-300/JE2 cells, had a MIC similar to that of the two APT-6K tolerant MRSA isolates (Fig. 4d). Transposon-induced inactivation of copA, copX, or copL rendered JE2 cells APT-6K sensitive, similar to MRSA-1 and MRSA-2 (~60–100-fold reduction in APT-6K tolerance). However, copZ inactivation had no effect on APT-6K tolerance. These data link APT-6K tolerance of selected MRSA to the presence of the recently discovered copX and copL genes.
Given the abundance of literature describing how the occurrence of metal-resistance, including copper-resistance, in bacteria is associated with the occurrence of antibiotic drug resistance, we speculated that the APT-6K tolerance mechanism may overlap with other antibiotic resistance mechanisms35,36,37,38. We thus decided to test whether APT-6K, when co-administered with other antibiotics, could overcome bacterial resistance mechanisms and restore antibiotic sensitivity.
APT-6K restores ampicillin activity against of MRSA isolates
A particularly interesting antibiotic for these experiments is ampicillin, as recent studies have provided proof of concept that β-lactam resistance in S. aureus can be reversed. For example, daptomycin has been demonstrated to restore β-lactam sensitivity in MRSA strains, and vice versa, β-lactams restored daptomycin sensitivity39,40. More importantly, ampicillin is on the World Health Organization’s Model List of Essential Medicines due to its low toxicity, low cost of production, and efficaciousness41. However, resistance to methicillin, which implies ampicillin resistance, is widespread throughout the world with over 50% of S. aureus clinical isolates in the United States and parts of Europe and Asia testing positive for methicillin resistance42.
To test whether ampicillin would regain antibacterial activity in the presence of APT-6K, we first established the inhibitory concentrations of ampicillin and APT-6K in the presence of copper for each of the four clinical MRSA isolates (Supplementary Table S2). With MICs of 128, 32, and 64 µg/ml, three of the isolates, MRSA-1, -3, and -4, displayed resistance to ampicillin, while MRSA-2 was sensitive to ampicillin under the chosen experimental conditions, although the clinical resistance profile had labeled it as ampicillin resistant.
Guided by these concentration data, APT-6K and ampicillin were then titrated against each other in microplate assays and S. aureus growth was measured using optical density after overnight growth. For each of the MRSA isolates, the results provided evidence that the two drugs interacted and APT-6K restored the antibacterial effect of ampicillin in the otherwise resistant strains. The specific nature of the APT-6K/ampicillin interaction effects varied with observed differences in the ampicillin resistance and APT-6K tolerance profiles of the clinical isolates (Fig. 5).
For the APT-6K sensitive/ampicillin resistant MRSA-1 isolate, this is immediately visible by the ampicillin induced shift of the APT-6K IC50/IC90 values to lower concentrations (Fig. 5a). IC50/IC90 value reductions could be accomplished by as little as 8 µg/ml ampicillin which is within physiologically relevant ampicillin concentrations that can be readily achieved in patients. Similar observations were made with the APT-6K sensitive/ampicillin sensitive MRSA-2 isolate. Interestingly, addition of ampicillin, at much lower concentrations improved the APT-6K IC50 value, but not the IC90 value to lower concentrations (Fig. 5b). The most interesting results came from the two APT-6K tolerant/ampicillin resistant MRSA isolates (MRSA-3 and MRSA-4). For both, MRSA-3 and MRSA-4, which exhibit a bi-phasic response to APT-6K and for which APT-6K alone does not suppress bacterial growth below the IC90 threshold at a therapeutically relevant concentration, the addition of physiologically relevant ampicillin concentrations overcame APT-6K tolerance (Fig. 5c,d). For either isolate, ampicillin concentrations between 4–8 µg/ml in combination with 0.3–0.6 µM APT-6K were sufficient to suppress bacterial growth below the IC90 threshold. The ability of ampicillin to either lower the APT-6K IC90 (Fig. 5a) or to overcome bacterial APT-6K tolerance (Fig. 5c,d) in MRSA isolates at concentrations it would be inactive by itself (1–8 µg/ml) clearly demonstrates positive drug interactions between the two compounds/drugs, and this can only occur if APT-6K at least partially restores the antibacterial activity of ampicillin against these clinical MRSA isolates.
Drug resistant infections in the US increased from 5% to 11% between 2002 and 2014, resulting in an annual ~2.2 billion dollar burden for the United States health care system43. To combat these bacteria, both new antibiotics as well as strategies to reduce resistance rates are urgently needed. In S. aureus, drug combinations to overcome pre-existing drug resistance with ß-lactam antibiotics have shown some promise in in vitro studies and in animal models, with human efficacy studies underway39,44,45,46,47. Our findings add APT-6K to the currently small list of compounds that can exert a direct antibacterial effect in the sub-micromolar range and can reverse ampicillin resistance at even lower concentrations. Also, APT-6K restored ampicillin activity to a concentration range (1–8 µg/ml) that is readily achievable in the serum of patients taking ampicillin48,49. Given the large body of literature that links not only copper-resistance, but more generally metal-resistance in bacteria with antibiotic resistance, the finding that some copper-drugs can overcome antibiotic resistance, exemplified by ampicillin resistance in this study, may actually not be surprising and should spur investigations into this phenomenon beyond MRSA (reviewed in: Poole 201738). The exact mechanism of APT-6K/ampicillin synergy is subject of ongoing study. For APT-6K, the relatively favorable therapeutic index in the absence of mitochondrial toxicities is also a marked improvement over earlier CDIs such as GTSM. This low toxicity extends to primary human PBMCs, which after treatment do not exhibit noticeable changes in surface markers, activation markers, or inflammatory cytokines. As such, our findings add APT-6K to the growing list of effective metal-dependent compounds that form a new class of antibiotics and provide the first example of a CDI that can reverse antibiotic resistance to the classic ß-lactam antibiotic, ampicillin.
Material and Methods
Bacterial growth conditions, cell culture, compounds, and metals
S. aureus strain Newman was verified by whole genome sequencing. The MRSA strains used in this study were obtained in a deidentified manner from UAB Laboratory Medicine after confirmation of their drug resistance. The transposon mutants and the parent JE2 were obtained from BEI as part of the Nebraska Transposon Mutant Library. Bacteria were inoculated from stocks stored at −80 °C into Mueller Hinton media (Oxoid LTD) and incubated overnight at 37 °C with shaking at 180 rpm. THP-1 monocytes and PBMCs were cultured in Roswell Park Memorial Institute (RPMI) media (Corning) supplemented with 10% fetal bovine serum (FBS, Atlanta Biologicals, Inc.), glutamine, penicillin and streptomycin, and incubated at 37 °C in the presence of 5% CO2. APT-6K was purchased from ChemBridge Corp., reconstituted and aliquoted in DMSO at a 10 mM stock concentration, and stored at −80 °C. Ampicillin (Fisher Scientific) was reconstituted and aliquoted in ddH2O at 10 mg/mL and stored at −80 °C. Copper sulfate (Acros Organics), zinc sulfate (Fisher Scientific), manganese sulfate (Alfa Aesar), cobalt sulfate (Alfa Aesar), nickel sulfate (Acros Organics), and iron chloride (Acros Organics) were reconstituted and aliquoted in ddH2O at 100 mM and stored at 4 °C.
Minimum inhibitory concentration (MIC) and metal specificity testing
Compound testing was performed in sterile 96-well plates. Compounds and the indicated metal salts were mixed in RPMI-1640 media without phenol red (Corning) that had a trace metal supplementation (RPMI-TM) to promote growth (3 µM EDTA, 50 µM MgCl2, 0.7 µM CaCl2, 80 µM NaMoO4, 168 µM CoCl2, 0.55 µM MnCl2, 0.7 µM ZnSO4, 2 µM FeSO4). Throughout the manuscript, metal concentrations refer to metal concentrations added in excess of the metals contained in this basic culture medium. Compound and antibiotic concentrations were titrated 2-fold from well to well. Bacteria were harvested from exponentially growing cultures that had been reinoculated from overnight cultures. The harvested bacteria were washed twice in RPMI-TM and then distributed to each well for a final OD of 0.005 (~5 × 106 bacteria/mL). Within each plate, there were sterility controls and no compound controls that were used as blanks or as 100% growth, respectively. Plates were sealed with parafilm to reduce evaporation and kept at 37 °C with 5% CO2 for 18–24 hours, after which each well was resuspended and OD600 readings were taken on a plate reader (Cytation 3 or Synergy 2, BioTek). MIC was defined as the lowest concentration at which growth was reduced by at least 90% in comparison to the untreated controls.
Time to death assay and bactericidal activity
Time to death and bactericidal activity assays were setup identically to that of the MIC determination assays. At the indicated time points or at the end of incubation (18–20 hours), the wells were resuspended and 5 µL of their contents were directly spotted onto MH agar plates and incubated overnight at 37 °C.
Quantification of intracellular ATP concentrations
To determine the effect of the tested compounds on energy production, we determined ATP concentrations as a surrogate marker. APT-6K in the presence of 50 µM copper and CCCP were titrated on S. aureus strain Newman seeded at an OD of 0.005 in RPMI-TM medium and then incubated for 1 hour at 37 °C. At this time, the BacTiter-Glo assay (Promega) was used to measure the ATP content of treated bacteria. Luminescence was read on a Cytation 3 spectrophotometer.
Testing of drug synergies
Microplate assays were performed in 96-well plates to determine the effects of copper or antibiotics when combined with APT-6K. The titrations of each test compound or metal ion were prepared in separate plates. The first plate had the test compound titrated across the plate (from column to column), while the second plate had the test compound or metal titrated 2-fold down the plate (from row to row). Volume from plate two was then transferred into plate one. Bacteria were washed twice and added to each well for a final assay OD of 0.005 (~5 × 106 bacteria/mL). Each microplate assay was done in triplicate, and each plate had sterility controls for blanking and controls without compound for normalization. Plates were incubated for 18–24 hours at 37 °C with 5% CO2. After incubation, plates were resuspended and OD600 readings were taken on a plate reader.
Toxicity of APT-6K on human cells
Toxicity of APT-6K on the monocytic THP-1 cells or peripheral blood mononuclear cells (PBMCs) isolated from buffy coats obtained from the Red Cross was assayed in the presence and absence of copper. Compounds were titrated and diluted in similar manner as described for S. aureus, but RPMI 1640 complemented with 10% FBS was used. After 24 hours of treatment, resazurin (Sigma-Aldrich) was added for a final concentration of 2.5 µg/mL. The viability status of the culture was analyzed by resazurin conversion resulting in fluorescence that was measured at excitation/emission: 530 nm/590 nm.
Flow cytometric analysis and reagents
LIVE/DEAD Fixable Aqua Stain (Invitrogen), excited by violet 405 nm laser and detected by 512 nm emission channel, was used to determine the percentage of dead cells. The T cell phenotype was determined by staining with mAbs: CD3-APC efluor780 (SK7, eBiosciences), CD4-BV786 (SK3, Biolegend), CD8-V500 (RPA-T8, BD Biosciences), and CD28-BV711 (CD28.2, BD Biosciences). Activation status was determined using mAbs: PD-1-PE-eFlour610 (J105, eBioscience), CD38-BUV737 (HB7, BD Biosciences), CD69-FITC (FN50, BD Biosciences), and CD25-BUV395 (2A3, BD Biosciences). Cytolytic profile was determined using mAbs: TNF-α-APC (6401.1111, BD Biosciences), Granzyme B-BV421 (GB11; BD Biosciences), and IFN-γ-PE-cy7 (B27, BD Biosciences). Monoclonal Ab CD14 BUV805 (M5E2, BD Biosciences, Franklin Lake, NJ) was used to exclude monocytes from analysis. Intracellular staining (ICS) was performed using the BD Cytofix/Cytoperm kit according to the manufacturer’s protocols. All antibody-stained cells were fixed in 1% formaldehyde (Sigma) prior to sample acquisition on a Symphony flow cytometer (BD Biosciences). Gates for flow cytometric acquisition and analyses were based on “fluorescence-minus-one” (FMO) controls and single stain compensation controls.
Mitochondrial toxicity was measured as a comparison of ATP levels and membrane integrity, using the Mitochondrial ToxGlo system (Promega). The assay was carried out as described by the manufacturer in a white 384 well-plate. THP-1 cells were grown in standard RPMI 1640 medium, then passaged for one day prior to the experiment in glucose-free RPMI supplemented with 0.2% galactose to better visualize mitochondrial inhibition. The assay itself was conducted using FBS-free and galactose supplemented RPMI. Digitonin and CCCP, as recommended by the manufacturer, were used as cytotoxicity and mitochondrial toxicity controls, respectively. Cytotoxicity was normalized using the highest digitonin concentration as 0% viability, and the no treatment control as 100% viability.
PCR and gel electrophoresis
Colony PCR was used to amplify the genes of interest. Briefly, a colony of the test strain was diluted in water and then added as the DNA source for the PCR using a Taq polymerase master mix (NEB). The PCR conditions used were based on the manufacture’s recommendations with an annealing temperature of 62 °C for 30 seconds and an extension time of 3 mins for 30 cycles.
Primers were designed to amplify the full gene and the primer sequences are provided in Supplementary Table 3. The PCR products were visualized after separation on a 1% agarose gel. The expected product sizes were 2791 bp for copA, 566 bp for copZ, 2729 bp for copX, and 1263 bp for copL.
Data analysis and chemical structures
Each experiment is representative of at least two independent experiments and contains at least 3 technical replicates. Error bars represent standard deviation of technical replicates unless stated otherwise. Data were analyzed and graphed using Excel (Microsoft), GraphPad Prism 8 (GraphPad), FlowJo (Treestar), and PowerPoint (Microsoft). Primers were designed using SnapGene (GSL Biotech LLC). Marvin Sketch was used for drawing and displaying chemical structures (MarvinSketch 6.1.4, 2013, ChemAxon (http://www.chemaxon.com)).
Aslam, B. et al. Antibiotic resistance: a rundown of a global crisis. Infect Drug Resist 11, 1645–1658, https://doi.org/10.2147/IDR.S173867 (2018).
Furuya, E. Y. & Lowy, F. D. Antimicrobial-resistant bacteria in the community setting. Nature reviews. Microbiology 4, 36–45, https://doi.org/10.1038/nrmicro1325 (2006).
O’Neill, J. UK review on antimicrobial resistance: tackling a crisis for the health and wealth of nations. https://amr-review.org/sites/default/files/AMR%20Review%20Paper%20-%20Tackling%20a%20crisis%20for%20the%20health%20and%20wealth%20of%20nations_1.pdf (2014).
Foster, T. J. Antibiotic resistance in Staphylococcus aureus. Current status and future prospects. FEMS Microbiology Reviews 41, 430–449, https://doi.org/10.1093/femsre/fux007 (2017).
Doern, C. D. When Does 2 Plus 2 Equal 5? A Review of Antimicrobial Synergy Testing. Journal of Clinical Microbiology 52, 4124, https://doi.org/10.1128/JCM.01121-14 (2014).
Acar, J. F. Antibiotic synergy and antagonism. Med Clin North Am 84, 1391–1406, https://doi.org/10.1016/s0025-7125(05)70294-7 (2000).
Dalecki, A. G. et al. Disulfiram and Copper Ions Kill Mycobacterium tuberculosis in a Synergistic Manner. Antimicrob Agents Chemother 59, 4835–4844, https://doi.org/10.1128/AAC.00692-15 (2015).
Haeili, M. et al. Copper complexation screen reveals compounds with potent antibiotic properties against methicillin-resistant Staphylococcus aureus. Antimicrob Agents Chemother 58, 3727–3736, https://doi.org/10.1128/aac.02316-13 (2014).
Shah, S. et al. 8-Hydroxyquinolines Are Boosting Agents of Copper-Related Toxicity in Mycobacterium tuberculosis. Antimicrob Agents Chemother 60, 5765–5776, https://doi.org/10.1128/aac.00325-16 (2016).
Dalecki, A. G. et al. Combinatorial phenotypic screen uncovers unrecognized family of extended thiourea inhibitors with copper-dependent anti-staphylococcal activity. Metallomics 8, 412–421, https://doi.org/10.1039/c6mt00003g (2016).
Djoko, K. Y., Paterson, B. M., Donnelly, P. S. & McEwan, A. G. Antimicrobial effects of copper(II) bis(thiosemicarbazonato) complexes provide new insight into their biochemical mode of action. Metallomics 6, 854–863, https://doi.org/10.1039/c3mt00348e (2014).
Salina, E. G. et al. Copper-related toxicity in replicating and dormant Mycobacterium tuberculosis caused by 1-hydroxy-5-R-pyridine-2(1H)-thiones. Metallomics 10, 992–1002, https://doi.org/10.1039/C8MT00067K (2018).
Totten, A. H. et al. Differential Susceptibility of Mycoplasma and Ureaplasma Species to Compound-Enhanced Copper Toxicity. Frontiers in Microbiology 10, https://doi.org/10.3389/fmicb.2019.01720 (2019).
Festa, R. A., Helsel, M. E., Franz, K. J. & Thiele, D. J. Exploiting innate immune cell activation of a copper-dependent antimicrobial agent during infection. Chem Biol 21, 977–987, https://doi.org/10.1016/j.chembiol.2014.06.009 (2014).
Helsel, M. E., White, E. J., Razvi, S. Z., Alies, B. & Franz, K. J. Chemical and functional properties of metal chelators that mobilize copper to elicit fungal killing of Cryptococcus neoformans. Metallomics 9, 69–81, https://doi.org/10.1039/c6mt00172f (2017).
Dalecki, A. G. et al. High-throughput screening and Bayesian machine learning for copper-dependent inhibitors of Staphylococcus aureus. Metallomics 11, 696–706, https://doi.org/10.1039/C8MT00342D (2019).
Crawford, C. L. et al. Pyrazolopyrimidinones, a novel class of copper-dependent bactericidal antibiotics against multi-drug resistant S. aureus. Metallomics 11, 784–798, https://doi.org/10.1039/C8MT00316E (2019).
Vestergaard, M. et al. Inhibition of the ATP Synthase Eliminates the Intrinsic Resistance of Staphylococcus aureus towards Polymyxins. mBio 8, e01114–01117, https://doi.org/10.1128/mBio.01114-17 (2017).
Jeon, A. B. et al. 2-aminoimidazoles collapse mycobacterial proton motive force and block the electron transport chain. Scientific reports 9, 1513–1513, https://doi.org/10.1038/s41598-018-38064-7 (2019).
Stimac, A., Sekutor, M., Mlinaric-Majerski, K., Frkanec, L. & Frkanec, R. Adamantane in Drug Delivery Systems and Surface Recognition. Molecules 22, https://doi.org/10.3390/molecules22020297 (2017).
Subashchandrabose, S. et al. Host-specific induction of Escherichia coli fitness genes during human urinary tract infection. Proceedings of the National Academy of Sciences 111, 18327–18332, https://doi.org/10.1073/pnas.1415959112 (2014).
Wagner, D. et al. Elemental analysis of Mycobacterium avium-, Mycobacterium tuberculosis-, and Mycobacterium smegmatis-containing phagosomes indicates pathogen-induced microenvironments within the host cell’s endosomal system. J Immunol 174, 1491–1500 (2005).
White, C., Lee, J., Kambe, T., Fritsche, K. & Petris, M. J. A role for the ATP7A copper-transporting ATPase in macrophage bactericidal activity. J Biol Chem 284, 33949–33956 (2009).
Djoko, K. Y., Donnelly, P. S. & McEwan, A. G. Inhibition of respiratory Complex I by copper(ii)-bis(thiosemicarbazonato) complexes. Metallomics 6, 2250–2259, https://doi.org/10.1039/C4MT00226A (2014).
Novo, D., Perlmutter, N. G., Hunt, R. H. & Shapiro, H. M. Accurate flow cytometric membrane potential measurement in bacteria using diethyloxacarbocyanine and a ratiometric technique. Cytometry 35, 55–63 (1999).
Silver, L. L. Challenges of Antibacterial Discovery. Clinical microbiology reviews 24, 71–109, https://doi.org/10.1128/cmr.00030-10 (2011).
Lou, P.-H. et al. Mitochondrial uncouplers with an extraordinary dynamic range. The Biochemical journal 407, 129–140, https://doi.org/10.1042/BJ20070606 (2007).
Schulz, I. Permeabilizing cells: some methods and applications for the study of intracellular processes. Methods Enzymol 192, 280–300, https://doi.org/10.1016/0076-6879(90)92077-q (1990).
Paterson, B. M. & Donnelly, P. S. Copper complexes of bis(thiosemicarbazones): from chemotherapeutics to diagnostic and therapeutic radiopharmaceuticals. Chem Soc Rev 40, 3005–3018, https://doi.org/10.1039/c0cs00215a (2011).
Crouch, P. J. et al. Increasing Cu bioavailability inhibits Abeta oligomers and tau phosphorylation. Proceedings of the National Academy of Sciences of the United States of America 106, 381–386, https://doi.org/10.1073/pnas.0809057106 (2009).
Cater, M. A. et al. Increasing intracellular bioavailable copper selectively targets prostate cancer cells. ACS Chem Biol 8, 1621–1631, https://doi.org/10.1021/cb400198p (2013).
Purves, J. et al. A horizontally gene transferred copper resistance locus confers hyper-resistance to antibacterial copper toxicity and enables survival of community acquired methicillin resistant Staphylococcus aureus USA300 in macrophages. Environ Microbiol, https://doi.org/10.1111/1462-2920.14088 (2018).
Rosario-Cruz, Z. et al. The copBL operon protects Staphylococcus aureus from copper toxicity: CopL is an extracellular membrane-associated copper-binding protein. https://doi.org/10.1074/jbc.RA118.004723 (2019).
Sitthisak, S., Knutsson, L., Webb, J. W. & Jayaswal, R. K. Molecular characterization of the copper transport system in Staphylococcus aureus. Microbiology 153, 4274–4283 (2007).
Gullberg, E., Albrecht, L. M., Karlsson, C., Sandegren, L. & Andersson, D. I. Selection of a multidrug resistance plasmid by sublethal levels of antibiotics and heavy metals. mBio 5, e01918, https://doi.org/10.1128/mBio.01918-14 (2014).
Gómez-Sanz, E. et al. Novel erm(T)-carrying multiresistance plasmids from porcine and human isolates of methicillin-resistant Staphylococcus aureus ST398 that also harbor cadmium and copper resistance determinants. Antimicrobial agents and chemotherapy 57, 3275–3282, https://doi.org/10.1128/AAC.00171-13 (2013).
Cavaco, L. M. et al. Cloning and occurrence of czrC, a gene conferring cadmium and zinc resistance in methicillin-resistant Staphylococcus aureus CC398 isolates. Antimicrobial agents and chemotherapy 54, 3605–3608, https://doi.org/10.1128/AAC.00058-10 (2010).
Poole, K. At the Nexus of Antibiotics and Metals: The Impact of Cu and Zn on Antibiotic Activity and Resistance. Trends in microbiology, https://doi.org/10.1016/j.tim.2017.04.010 (2017).
Mehta, S. et al. β-Lactams increase the antibacterial activity of daptomycin against clinical methicillin-resistant Staphylococcus aureus strains and prevent selection of daptomycin-resistant derivatives. Antimicrobial agents and chemotherapy 56, 6192–6200, https://doi.org/10.1128/AAC.01525-12 (2012).
Renzoni, A. et al. Molecular Bases Determining Daptomycin Resistance-Mediated Resensitization to β-Lactams (Seesaw Effect) in Methicillin-Resistant Staphylococcus aureus. Antimicrobial Agents and Chemotherapy 61, e01634–01616, https://doi.org/10.1128/aac.01634-16 (2017).
Organization, W. H. World Health Organization Model List of Essential Medicines, 21st List, 2019. (2019).
Lee, A. S. et al. Methicillin-resistant Staphylococcus aureus. Nature Reviews Disease Primers 4, 18033, https://doi.org/10.1038/nrdp.2018.33 (2018).
Thorpe, K. E., Joski, P. & Johnston, K. J. Antibiotic-Resistant Infection Treatment Costs Have Doubled Since 2002, Now Exceeding $2 Billion Annually. Health Aff (Millwood) 37, 662–669, https://doi.org/10.1377/hlthaff.2017.1153 (2018).
Foster, T. J. Can beta-Lactam Antibiotics Be Resurrected to Combat MRSA? Trends in microbiology 27, 26–38, https://doi.org/10.1016/j.tim.2018.06.005 (2019).
Rand, K. H. & Houck, H. J. Synergy of Daptomycin with Oxacillin and Other β-Lactams against Methicillin-Resistant Staphylococcus aureus. Antimicrobial Agents and Chemotherapy 48, 2871–2875, https://doi.org/10.1128/aac.48.8.2871-2875.2004 (2004).
Geriak, M. et al. Clinical Data on Daptomycin plus Ceftaroline versus Standard of Care Monotherapy in the Treatment of Methicillin-Resistant Staphylococcus aureus Bacteremia. Antimicrob Agents Chemother 63, https://doi.org/10.1128/aac.02483-18 (2019).
Tong, S. Y. C. et al. CAMERA2 – combination antibiotic therapy for methicillin-resistant Staphylococcus aureus infection: study protocol for a randomised controlled trial. Trials 17, 170, https://doi.org/10.1186/s13063-016-1295-3 (2016).
Giachetto, G. et al. Ampicillin and penicillin concentration in serum and pleural fluid of hospitalized children with community-acquired pneumonia. Pediatr Infect Dis J 23, 625–629, https://doi.org/10.1097/01.inf.0000128783.11218.c9 (2004).
Davies, B. & Maesen, F. Serum and sputum antibiotic levels after ampicillin, amoxycillin and bacampicillin chronic bronchitis patients. Infection 7(Suppl 5), S465–468, https://doi.org/10.1007/bf01659773 (1979).
We thank Dr. Terje Dokland for providing strain S. aureus Newman and Dr. William Benjamin for providing and characterizing clinical isolate MRSA strains. Kaitlyn Schaaf is now affiliated with Vanderbilt University, USA and Frank Wolschendorf is at HTK Hygiene Technologie Kompetenzzentrum GmbH, Bamberg, Germany. Transposon mutants were provided by the Network on Antimicrobial Resistance in Staphylococcus aureus (NARSA) for distribution by BEI Resources, NIAID, NIH: Nebraska Transposon Mutant Library (NTML) Genetic Toolbox, NR-49947.This study was supported by National Institute of Health (NIH) grant R01AI121364 awarded to O.K.
The authors declare no competing interests.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Crawford, C.L., Dalecki, A.G., Perez, M.D. et al. A copper-dependent compound restores ampicillin sensitivity in multidrug-resistant Staphylococcus aureus. Sci Rep 10, 8955 (2020). https://doi.org/10.1038/s41598-020-65978-y