Mechanisms Underlying Synergistic Killing of Polymyxin B in Combination with Cannabidiol against Acinetobacter baumannii: A Metabolomic Study

Polymyxins have resurged as the last-resort antibiotics against multidrug-resistant Acinetobacter baumannii. As reports of polymyxin resistance in A. baumannii with monotherapy have become increasingly common, combination therapy is usually the only remaining treatment option. A novel and effective strategy is to combine polymyxins with non-antibiotic drugs. This study aimed to investigate, using untargeted metabolomics, the mechanisms of antibacterial killing synergy of the combination of polymyxin B with a synthetic cannabidiol against A. baumannii ATCC 19606. The antibacterial synergy of the combination against a panel of Gram-negative pathogens (Acinetobacter baumannii, Klebsiella pneumoniae and Pseudomonas aeruginosa) was also explored using checkerboard and static time-kill assays. The polymyxin B–cannabidiol combination showed synergistic antibacterial activity in checkerboard and static time-kill assays against both polymyxin-susceptible and polymyxin-resistant isolates. The metabolomics study at 1 h demonstrated that polymyxin B monotherapy and the combination (to the greatest extent) significantly perturbed the complex interrelated metabolic pathways involved in the bacterial cell envelope biogenesis (amino sugar and nucleotide sugar metabolism, peptidoglycan, and lipopolysaccharide (LPS) biosynthesis), nucleotides (purine and pyrimidine metabolism) and peptide metabolism; notably, these pathways are key regulators of bacterial DNA and RNA biosynthesis. Intriguingly, the combination caused a major perturbation in bacterial membrane lipids (glycerophospholipids and fatty acids) compared to very minimal changes induced by monotherapies. At 4 h, polymyxin B–cannabidiol induced more pronounced effects on the abovementioned pathways compared to the minimal impact of monotherapies. This metabolomics study for the first time showed that in disorganization of the bacterial envelope formation, the DNA and RNA biosynthetic pathways were the most likely molecular mechanisms for the synergy of the combination. The study suggests the possibility of cannabidiol repositioning, in combination with polymyxins, for treatment of MDR polymyxin-resistant Gram-negative infections.


Introduction
Antimicrobial resistance has become a major threat to health and economic wellbeing worldwide [1]. The rapid emergence of MDR Gram-negative pathogens (e.g., carbapenem-

Determination of MIC and FIC
The MICs of polymyxin B and CBD were determined for all bacterial isolates in triplicate on separate days using broth microdilution checkerboard assays. The MIC is defined as the lowest concentration of antimicrobial agent that shows a complete inhibition of visible bacterial growth after overnight incubation. The polymyxin B and CBD stock solutions were prepared immediately before each experiment. Polymyxin B (Betapharma, Shanghai, China) powder was dissolved in MilliQ water and sterilized by membrane syringe filters with a pore size of 0.22 micron (ThermoFisher Scientific, Melbourne, VIC, Australia). Cannabidiol (CBD) (Cayman Chemicals, Ann Arbor, Michigan, USA) was dissolved in dimethyl sulfoxide (DMSO; Sigma-Aldrich, Melbourne, VIC, Australia). Serial concentrations of DMSO (0.25%, 0.5%, 1%, 2.5% v/v) showed no inhibitory effect against all bacteria tested. Serial dilutions (2-fold) of the polymyxin B and CBD were prepared in a cation-adjusted Mueller-Hinton broth (CAMHB; Oxoid, Basingstoke, UK) to obtain solutions with various concentrations. The experiments were performed in 96-well microtiter plates (Techno Plas, Adelaide, SA, Australia) in broth culture containing~10 6 CFU/mL of the bacteria tested. The microdilution plates were incubated at 37 • C for 18-20 h.
The synergy tests for the combination were conducted using the fractional inhibitory concentration index (FICI) analysis. The formula used to calculate the FIC values was as follows: FICI = (MIC of drug A in combination ÷ MIC of drug A alone) + (MIC of drug B in combination ÷ MIC of drug B alone). The FICI results were defined as: synergism FIC < 0.5; addition FIC = 0.5-1.0; indifference FIC = 1-4; antagonism FIC ≥ 4 [22]. An MIC of 128 µg/mL for CBD was used to calculate the FICI scores.

Static Time-Kill Studies
Based on FIC results, static time-kill assays of polymyxin B monotherapy and in combination with CBD were performed against selected strains from different species. Prior to each time-kill assay, a sub-culture of the selected bacterial isolate was prepared from its stock cultures and incubated at 37 • C for 18~24 h. To prepare overnight culture, a single colony was transferred from nutrient agar plate to inoculate 10 mL of fresh CAMHB (Oxoid, UK) in a 50 mL Falcon tube (ThermoFisher Scientific, Australia) and incubated overnight (~16 h) in a water bath shaker at a temperature of 37 • C and 150 rpm shaking speed. Subsequently, the log phase was induced by inoculating 100 µL from the overnight culture into 10 mL of fresh CAMHB broth and incubated at 37 • C in a water bath shaker (shaking speed, 150 rpm) for 2 h. A 200 µL aliquot of bacterial suspension was inoculated into 4 borosilicate culture tubes (capacity 50 mL) containing 20 mL (1:100 dilution) of fresh CAMHB each. Three treatment tubes containing a starting culture of bacteria were spiked with the desired concentrations of polymyxin B and CBD monotherapies or in combination. One culture tube was drug-free to represent the control. At 0, 1, 4, 8 and 24 h, samples were taken from each tube and serially diluted in 0.9% saline, and then automatically plated onto nutrient agar plates using a Whitley Automated Spiral Plater (WASP). After 24 h incubation at 37 • C, the number of colonies on the plates were counted, and the time-kill graphs were plotted as time (hours) vs. log 10 CFU/mL.

Bacterial Culture Preparation for Metabolomics
Single colonies of A. baumannii ATCC 19606 grown on a nutrient agar plate (Media Preparation Unit, University of Melbourne, Australia) were picked and inoculated into 10 mL of CAMHB (Oxoid, UK) in a 50 mL Falcon tube (ThermoFisher Scientific, Australia) and incubated at 37 • C overnight (~16 h) in a water bath shaker (150 rpm). The next day, a 100-fold dilution of bacterial culture was prepared by transferring a small amount of the overnight culture into 500 mL conical flasks (×4) containing fresh CAMHB. The flasks were immediately incubated in an incubator shaker (Bioline) at 37 • C with shaking at 180 rpm to log phase of~10 8 CFU/mL (OD600~0.5). Stock solutions of polymyxin B, CBD, or the combination were added accordingly to three treatment flasks with concentrations of 2 mg/L for polymyxin B, 4 mg/L for cannabidiol, and the combination of both compounds; the fourth flask served as control (drug-free). Samples were then taken at each time point (1 and 4 h) and immediately quenched to pause further bacterial metabolism. Subsequently, the OD600 reading for all samples was measured and normalized to~0.5 with fresh CAMHB. The bacterial cells were pelleted with centrifugation at 3220× g at 4 • C for 10 min, and the supernatants were removed subsequently. The pellets were stored at −80 • C, awaiting metabolite extraction. Four biological replicates were prepared for each treatment condition to reduce the bias from inherent random variation.

Metabolite Extraction
The bacterial pellets were washed twice in 1 mL of 0.9% saline and centrifuged at 3220× g at 4 • C for 10 min. A cold extraction solvent (chloroform: methanol: water) containing 1 µM each of the internal standards (CHAPS, CAPS, PIPES and TRIS) was used to re-suspend the bacterial pellets. The pellets were then exposed to three freeze-thaw cycles (using liquid nitrogen) to release the intracellular metabolites. After the freezing and thawing process, the supernatants were obtained after centrifugation at 3220× g for 10 min at 4 • C and immediately transferred into an injection vial for liquid chromatography-mass spectrometry (LC-MS) analysis. An equal volume of each sample was combined and used as a quality control (QC) [40].

Data Processing, Bioinformatics, and Statistical Analyses
The LC-MS data processing, bioinformatics and statistical analysis were performed as we have previously described [41].

Synergy Testing of Polymyxin B-Cannabidiol Combinations (FICI Calculation)
The synergistic antibacterial activity of polymyxin B and CBD as monotherapies and in combination was tested against a panel of clinical isolates of A. baumannii, P. aeruginosa, and K. pneumoniae (Table 1). The polymyxin B and CBD combinations demonstrated synergistic antibacterial activity against all of the tested strains of the three species A. baumannii, P. aeruginosa, and K. pneumoniae except for one A. baumannii (the LPS-deficient A. baumannii FADDI-AB065, indifference FICI = 2) and one P. aeruginosa (P. aeruginosa FADDI-PA019m, additive FICI = 0.53) ( Table 1). On the other hand, polymyxin B and CBD monotherapies were found to be ineffective (MICs 8 to >128 mg/L) against the polymyxin B-resistant isolates (except for A. baumannii FADDI-AB065, CBD MIC = 0.5 mg/L) ( Table 1).

Static Time-Kill Studies
The time-kill assessment of polymyxin B and CBD as monotherapies or in combination was further carried out against polymyxin B-susceptible A.  Polymyxin B monotherapy displayed significant killing activity against A. baumannii ATCC 19606, as manifested by a >2 log10 CFU/mL decline in bacterial counts compared with the control (untreated) at early time points (1 and 4 h); however, the killing curve of polymyxin B was comparable with the control at 24 h. Insignificant killing effect against the same strain was observed following CBD monotherapy across all time points ( Figure  1). On the other hand, polymyxin B-CBD combination showed a higher killing curve compared to the control at all time exposures (>−2 log10 CFU/mL, 1 h; ≥−4 log10 CFU/mL, 4 and 8 h; >−3.5 log10 CFU/mL, 24 h). No significant decrease in the bacterial burden was noticed following polymyxin B (except at 4 h, −2 log10 CFU/mL) and CBD monotherapy against FADDI-AB144 at all time points, whilst a more than 4 log10 CFU/mL decrease in bacterial Polymyxin B monotherapy displayed significant killing activity against A. baumannii ATCC 19606, as manifested by a >2 log 10 CFU/mL decline in bacterial counts compared with the control (untreated) at early time points (1 and 4 h); however, the killing curve of polymyxin B was comparable with the control at 24 h. Insignificant killing effect against the same strain was observed following CBD monotherapy across all time points (Figure 1). On the other hand, polymyxin B-CBD combination showed a higher killing curve compared to the control at all time exposures (>−2 log 10 CFU/mL, 1 h; ≥−4 log 10 CFU/mL, 4 and 8 h; >−3.5 log 10 CFU/mL, 24 h). No significant decrease in the bacterial burden was noticed following polymyxin B (except at 4 h, −2 log 10 CFU/mL) and CBD monotherapy against FADDI-AB144 at all time points, whilst a more than 4 log 10 CFU/mL decrease in bacterial count compared with control was seen after the combination at 1 and 4 h ( Figure 1). The combination also continued to reduce the bacterial burden by ≥2 log 10 CFU/mL at later time points (8 and 24 h). Against K. pneumoniae strains, although a higher killing curve was observed after polymyxin B monotherapy (−3.0 log 10 CFU/mL) against K. pneumoniae FADDI-KP002 at early time points (1 and 4 h), inconsistent regrowth occurred, with very minor log 10 CFU/mL difference compared to the control after 24 h. Notably, the combination therapy was more effective in bacterial killing, as manifested by a ≥4.0-6.0 log 10 CFU/mL decrease in bacterial counts compared with untreated controls across all time points (except 24 h). However, a completely different scene was observed against K. pneumoniae FADDI-KP012, wherein polymyxin B monotherapy did not display significant bacterial killing at all time points, as manifested by time-kill curves that resembled untreated controls ( Figure 1). Nevertheless, polymyxin B and CBD combination therapy displayed potent bacterial killing effect, peaking at 1 and 4 h with a 3.0-6.0 log 10 CFU/mL decline in bacterial counts compared to the control, which was sustained with a 3.0-4.0 log 10 CFU/mL decline after 8 and 24 h, respectively, compared to the control ( Figure 1). Regarding P. aeruginosa strains, only the combination therapy was effective in causing significant bacterial killing activity, as manifested by a >1.5-3 log 10 CFU/mL (FADDI-PA006) and >5-6 log 10 CFU/mL (FADDI-PA070) decrease in the bacterial burden across all time exposures (Figure 1). CBD monotherapy displayed no antimicrobial activity against any of the strains assessed during 24 h treatment, which was indicated by a time-kill curve pattern similar to the untreated control ( Figure 1).
The current findings suggest that the synergy between polymyxin B and the CBD most likely arises from bioavailability synergy [42], which means that polymyxins permeabilize the Gram-negative OM and thereby allow the entry of CBD into the intracellular compartment to subsequently exert its action on its intracellular targets [36,38].
The significant metabolites were determined by using one-way analysis of variance (ANOVA) (≥0.58-log 2 -fold; p ≤ 0.05). The data precision of all sample groups was agreeable at both time points (1 and 4 h), where the median RSDs were (17-23%) and (19-26%) for untreated (control) and treated samples, respectively, at all time points, consistent with some baseline dynamic variations in bacterial metabolism (Supplementary Table S1). The PCA plots demonstrated that the combination treatment samples were significantly distinguished from the untreated control samples, moderately at 1 h and largely at 4 h (Supplementary Figure S2A). Notably, the polymyxin B and combination treatments were nearly overlapped at 1 h, suggesting that polymyxin B initially governed the synergy of the combination treatment, whereas polymyxin B alone closely resembled the untreated control samples at the later time point, 4 h (Supplementary Figure S2A). CBD treated samples were undistinguished from the untreated control groups across all time exposures (1 and 4 h) (Supplementary Figure S2A). The same pattern of differences between the treated and untreated (control) groups was observed with the heatmaps (Supplementary Figure S2B).
A higher perturbation was induced by the combination treatment, as indicated by a perturbation of 267 (204 decreased and 63 increased) and 152 (139 decreased and 13 increased) significant metabolites at 1 and 4 h, respectively, whereas polymyxin B monotherapy perturbed 187 (174 decreased and 13 increased) significant metabolites at 1 h and a far smaller number at 4 h, with only 50 significantly decreased metabolites. CBD showed less effect, significantly perturbing only 29 (15 decreased and 14 increased) and 10 (6 decreased and 4 increased) significant metabolites at 1 and 4 h, respectively (Supplementary Figure S3A).
The metabolic profiling analysis of the significantly altered metabolites indicated that the combination treatment, at both time points, largely perturbed (mainly decreased) lipids, amino acids and peptides compared to the slightly impacted metabolite carbohydrates, nucleotides and cofactors and vitamin metabolism (Supplementary Figure S3B). Likewise, but to a lesser extent, polymyxin B monotherapy induced a similar pattern, wherein lipids, amino acids and peptides were the impacted metabolite classes (Supplementary Figure S3B).
On the other hand, CBD monotherapy caused a suppression in the levels of lipid intermediates (mainly at 1 h) and an elevation in the levels of amino acids (mainly at 4 h) (Supplementary Figure S3B). Notably, the combination treatment induced 148 and 103 unique significant metabolites at 1 and 4 h, respectively (Supplementary Figure S3C). There were only 10 and 4 intermediates in common between all treatment conditions at 1 and 4 h, respectively (Supplementary Figure S3C). The combination treatment shared~30-35% of the intermediates with polymyxin B monotherapy at 1 and 4 h (Supplementary Figure S3C).

Analyses of Dysregulated Glycerophospholipid and Fatty Acid Metabolism
The major lipid components of the cellular envelope of Gram-negative bacteria include phosphatidylglycerol (PG), phosphatidylethanolamine (PE) and cardiolipin [43]. These lipid precursors are crucial to maintaining the integrity of the bacterial cell envelope to adapt and survive under ever-changing environmental stress [44].
On the other hand, at 4 h, although the impact of combination therapy on lipid metabolism declined, a remarkable perturbation pattern of lipid intermediates particularly against glycerophospholipids was observed ( Figure 2C). The levels of two essential lipids (sn-glycerol 3-phosphate and sn-glycero-3-phosphoethanolamine) involved in the rate-limiting step of the biosynthesis of bacterial membrane phosphatidic acid declined in response to the combination therapy (≥−1.0-log 2 -fold; p ≤ 0.05) ( Figure 2C) [45]. Furthermore, the effect on fatty acid elongation was slightly sustained following the combination therapy, as manifested by a significant decline in the abundance of oleoyl-CoA (log 2 FC = −2.8) ( Figure 2C).  In comparison, the impact from polymyxin B and cannabidiol monotherapies at 4 h started to disappear; polymyxin B decreased the abundance of oleoyl-CoA, sn-glycerol 3-phosphate and sn-glycero-3-phosphoethanolamine (≥−1.0-log 2 -fold; p ≤ 0.05), whereas the level of only one fatty acid precursor significantly declined after cannabidiol treatment ( Figure 2C).

Analyses of Dysregulated
Amino-Sugar and Nucleotide-Sugar Metabolism, the Pentose Phosphate Pathway, and Downstream Peptidoglycan and Lipopolysaccharide Biosynthesis Gram-negative bacteria are surrounded by a complex multilayered structure (OM, a thin peptidoglycan cell wall, and inner cell membrane) that protects them from the harsh environment, but selectively enables the passage of nutrients from the outside and waste products from the inside [49]. Notably, polymyxin B-cannabidiol treatment caused a significant perturbation (slightly at 1 h and largely at 4 h) in the abundance of key metabolites involved in amino-sugar and nucleotide-sugar metabolism, the pentose phosphate pathway (PPP), and the direct interrelated metabolic pathways of cell envelope biogenesis, peptidoglycan, and lipopolysaccharide (LPS) formation ( Figure 3; Supplementary Figure S4). At 1 h, the abundance of one essential component (UDP-Nacetyl-D-glucosamine (UDP-GlcNAc)) of amino-sugar and nucleotide-sugar metabolism and two fundamental precursors of PPP (D-sedoheptulose 7-phosphate and D-ribulose 5-phosphate (Ru5P)) underwent a significant decline after the combination treatment (≥−2.0-log 2 -fold; p ≤ 0.05) (Supplementary Figure S4). As a downstream consequence, the combination treatment also caused a marked depletion in the abundance of three peptidoglycan building blocks, namely UDP-N-acetylmuramoyl-L-alanyl-D-glutamyl-meso-2,6diaminopimeloyl-D-alanyl-D-alanine (UDP-MurNAc-L-Ala- Abstract: Polymyxins have resurged as the last-resort antibiotics against m netobacter baumannii. As reports of polymyxin resistance in A. baumannii w become increasingly common, combination therapy is usually the only rema A novel and effective strategy is to combine polymyxins with non-antibi aimed to investigate, using untargeted metabolomics, the mechanisms of a ergy of the combination of polymyxin B with a synthetic cannabidiol again 19606. The antibacterial synergy of the combination against a panel of Gra (Acinetobacter baumannii, Klebsiella pneumoniae and Pseudomonas aeruginosa) w checkerboard and static time-kill assays. The polymyxin B-cannabidiol com ergistic antibacterial activity in checkerboard and static time-kill assays ag susceptible and polymyxin-resistant isolates. The metabolomics study at 1 h ymyxin B monotherapy and the combination (to the greatest extent) sign complex interrelated metabolic pathways involved in the bacterial cell enve sugar and nucleotide sugar metabolism, peptidoglycan, and lipopolysacch sis), nucleotides (purine and pyrimidine metabolism) and peptide metabolis ways are key regulators of bacterial DNA and RNA biosynthesis. Intrigu caused a major perturbation in bacterial membrane lipids (glycerophospho compared to very minimal changes induced by monotherapies. At 4 h, po induced more pronounced effects on the abovementioned pathways compa pact of monotherapies. This metabolomics study for the first time showed of the bacterial envelope formation, the DNA and RNA biosynthetic pathwa molecular mechanisms for the synergy of the combination. The study sug cannabidiol repositioning, in combination with polymyxins, for treatment sistant Gram-negative infections. ƴ   Figure S4). Consequently, a dramatic decrease in the levels of two common intermediates (3-deoxy-D-manno-octulosonate (KDO), log 2 FC = −1.2; D-glycero-D-manno-heptose 7-phosphate, log 2 FC = −2.5) of LPS formation was observed following the combination treatment at 1 h (Supplementary Figure S4).
On the other hand, at 1 h, although polymyxin B monotherapy caused a significant alteration in the main pathways involved in cell envelope assembly, its effect was less potent compared to that of the combination therapy (Supplementary Figure S4). The levels of one amino-sugar and nucleotide-sugar intermediate, UDP-GlcNAc, and two intermediates of PPP (D-sedoheptulose 7-phosphate and D-ribulose 5-phosphate) experienced a substantial decline after polymyxin B monotherapy (≥−0.59-log 2 -fold; p ≤ 0.05) (Supplementary Figure S4). Simultaneously, polymyxin B monotherapy also disrupted peptidoglycan biosynthesis, as manifested by an alteration in the levels of UDP-MurNAc-L-Ala-     At 4 h, different patterns of perturbation were observed wherein the combination treatment induced a profound inhibitory impact on amino-sugar and nucleotide-sugar metabolism and PPP, peptidoglycan and LPS biosynthesis ( Figure 3A,B). The levels of two important building blocks of amino-sugar and nucleotide-sugar metabolism, namely UDP-GlcNAc and UDP-N-acetylmuramate (UDP-MurNAc), were depleted after the combination treatment at 4 h (≥−2.0-log 2 -fold; p ≤ 0.05) ( Figure 3A). Similarly to the combination impact at 1 h, the abundance of two crucial PPP precursors (D-sedoheptulose 7-phosphate and D-ribulose 5-phosphate) was significantly decreased following the combination therapy (≥−1.5-log 2 -fold; p ≤ 0.05) ( Figure 3B). A greater impact on the downstream pathway for peptidoglycan biosynthesis was seen after the combination therapy. This was shown by the fact that the combination markedly decreased the abundance of five peptidoglycan biosynthesis intermediates, namely UDP-MurNAc-L-Ala-    Figure 3A). Importantly, the combination treatment caused a profound decrease (log 2 FC = −1.1) in the abundance of undecaprenyl phosphate (C55-P) at 4 h ( Figure 3A). The combination treatment also reduced the abundance of two essential components of LPS biogenesis, KDO and D-glycero-D-manno-heptose 7-phosphate (log 2 FC = −2.5 and 1.8, respectively) ( Figure 3B).
Although polymyxin B alone perturbed a similar number (except for undecaprenyl phosphate) of intermediates involved in the cell envelope biosynthetic interconnected pathways, it was relatively less potent than the combination therapy at 4 h ( Figure 3A,B). This was manifested by a lower intensity reduction in the levels of two amino-sugar and nucleotide-sugar metabolism intermediates (UDP-GlcNAc and UDP-MurNAc) and an equal number of PPP components (D-sedoheptulose 7-phosphate and Ru5P) due to polymyxin B monotherapy (≥−0.59-log 2 -fold; p ≤ 0.05) ( Figure 3A,B). Consequently, the abundance of five essential peptidoglycan intermediates was diminished after polymyxin B monotherapy, including UDP-MurNAc-L-Ala- Abstract: Polymyxins have resurged as the last-resort antibiotics against multidrug-resista netobacter baumannii. As reports of polymyxin resistance in A. baumannii with monotherap become increasingly common, combination therapy is usually the only remaining treatment o A novel and effective strategy is to combine polymyxins with non-antibiotic drugs. This aimed to investigate, using untargeted metabolomics, the mechanisms of antibacterial killin ergy of the combination of polymyxin B with a synthetic cannabidiol against A. baumannii 19606. The antibacterial synergy of the combination against a panel of Gram-negative path (Acinetobacter baumannii, Klebsiella pneumoniae and Pseudomonas aeruginosa) was also explored checkerboard and static time-kill assays. The polymyxin B-cannabidiol combination showe ergistic antibacterial activity in checkerboard and static time-kill assays against both polym susceptible and polymyxin-resistant isolates. The metabolomics study at 1 h demonstrated th ymyxin B monotherapy and the combination (to the greatest extent) significantly perturb complex interrelated metabolic pathways involved in the bacterial cell envelope biogenesis ( sugar and nucleotide sugar metabolism, peptidoglycan, and lipopolysaccharide (LPS) bios sis), nucleotides (purine and pyrimidine metabolism) and peptide metabolism; notably, these ways are key regulators of bacterial DNA and RNA biosynthesis. Intriguingly, the combi caused a major perturbation in bacterial membrane lipids (glycerophospholipids and fatty compared to very minimal changes induced by monotherapies. At 4 h, polymyxin B-canna induced more pronounced effects on the abovementioned pathways compared to the minim pact of monotherapies. This metabolomics study for the first time showed that in disorgan of the bacterial envelope formation, the DNA and RNA biosynthetic pathways were the mos molecular mechanisms for the synergy of the combination. The study suggests the possib cannabidiol repositioning, in combination with polymyxins, for treatment of MDR polymy sistant Gram-negative infections. ƴ   Figure 3A). The impact of polymyxin B monotherapy on the LPS biosynthetic pathway was sustained at 4 h, although it was less potent compared to the combination therapy, causing a less intense reduction in the abundance of KDO and D-glycero-D-manno-heptose 7-phosphate (log 2 FC = −1.6, −1.7, respectively) ( Figure 3B). Intriguingly, at 4 h, the inhibitory impact of cannabidiol monotherapy on the amino sugar donor, UDP-GlcNAc, continued, and moreover, it was also effective in decreasing the level of a key inner-core LPS precursor, namely KDO (≥−0.59-log 2 -fold; p ≤ 0.05) ( Figure 3A,B).

Analyses of Perturbations in Nucleotide and Peptide Metabolism
Bacterial RNA synthesis requires ATP, guanosine-triphosphate (GTP), cytidine-triphosphate (CTP) and uridine-triphosphate (UTP) as the building blocks, and deoxynucleotide triphosphates (such as dATP, dCTP and dGTP) are the substrates for bacterial DNA polymerases, which are responsible for both replication and repair of cellular DNA [50,51].
Polymyxin B monotherapy caused a marked depletion in the levels of nucleotides predominantly at 1 h and to a lesser extent at 4 h, and this effect was even more pronounced following combination treatment particularly at 4 h (≥0.59-log 2 -fold; p ≤ 0.05) ( Figure 4A,B). At 1 h, a marked decline in the levels of 14 nucleotides was observed following polymyxin B monotherapy, including dAMP, dTMP, uridine and xanthine (≥−1.0-log 2 -fold; p ≤ 0.05) ( Figure 4B), whereas cannabidiol monotherapy had modest effects, significantly decreasing the abundance of five fundamental nucleotides at 1 h (≥−0.59-log 2 -fold; p ≤ 0.05) ( Figure 4B). Fifteen nucleotides underwent a remarkable decrease in their levels after combination treatment (except for an increment in adenosine and deoxyguanosine levels) ( Figure 4B).  At 4 h, the combination profoundly perturbed the abundance of 24 nucleotide intermediates (≥1.0-log 2 -fold; p ≤ 0.05) ( Figure 4A,B). Among the nucleotides, IMP, inosine, adenosine, and deoxyguanosine were uniquely perturbed due to the combination therapy at both time points (1 and 4 h) (≥1.0-log 2 -fold; p ≤ 0.05) ( Figure 4B). Similarly to the 1 h time point, albeit to a lesser extent, polymyxin B monotherapy caused a reduction of 10 nucleotides at 4 h (≥−0.59-log 2 -fold; p ≤ 0.05) ( Figure 4B), whereas cannabidiol monotherapy displayed a negligible effect at 4 h by depleting only the CDP level (log 2 FC = −0.94) ( Figure 4B).
Notably, a large number of metabolites involved in peptide metabolism were significantly altered following combination treatment at 1 h (41 peptides) and to a limited degree at 4 h (17 peptides) (≥0.59-log 2 -fold, p ≤ 0.05) (Supplementary Figure S5). Likewise, but to a lesser extent, polymyxin B monotherapy caused a significant inhibitory effect on 34 precursors of peptide metabolism at 1 h, while only two peptide intermediates showed a significant decrease following polymyxin B monotherapy at 4 h (Supplementary Figure S5). No significant effect was observed in the samples treated with cannabidiol monotherapy at both time exposures (1 and 4 h).

Discussion
To prolong and maximize the clinical efficacy of polymyxins and prevent the emergence of antibiotic resistance, novel polymyxin combination therapy against MDR Gramnegative bacteria such as A. baumannii is desperately needed. Although researchers have shown that polymyxins in combination with cannabidiol displayed synergistic killing activity against Gram-negative bacteria, including A. baumannii, they did not perform timekill assays or investigate the precise mechanisms behind the synergistic killing activity of polymyxins in combination with the cannabidiol against A. baumannii [33,35,36]. To the best of our knowledge, this is the first study to decipher the mechanisms of synergistic killing activity of polymyxins and synthetic cannabidiol combination therapy against A. baumannii using untargeted metabolomics.
The checkerboard and time-kill assays demonstrated synergistic bacterial killing by polymyxins plus cannabidiol treatment against a panel of MDR Gram-negative pathogens (A. baumannii, K. pneumoniae and P. aeruginosa) (Figure 1; Table 1). The metabolomics study showed that polymyxin B-cannabidiol treatment caused a greater perturbation in A. baumannii ATCC 19606 metabolome compared to monotherapies, particularly at the later time exposure (4 h). The most significant findings for integrating pathway enrichment analysis include (i) dysregulation in bacterial cell envelope biogenesis, (ii) inhibition of DNA and RNA metabolism, as reflected by the major perturbation of purine and pyrimidine metabolism, and (iii) perturbation of peptide metabolism.
Polymyxins exert primary antibacterial killing activity by disorganizing the bacterial OM [9,52]. Unsurprisingly, at 1 h, polymyxin B caused a marked suppression, albeit far less compared to that achieved with the combination, in fatty acids, glycerophospholipids metabolism as well as key lipids of OM such as trans-hexadec-2-enoyl-CoA, sn-glycero-3phosphocholine and sn-glycero-3-phosphoethanolamine (Figure 2A,B). Our findings agreed with previous metabolomics and transcriptomics studies that showed that A. baumannii metabolites and genes involved primarily in bacterial OM biosynthesis and phospholipid trafficking experienced a significant alteration at 1 h polymyxin posttreatment [19,53]. However, the impact from polymyxin B monotherapy on levels of essential bacterial membrane lipids was minimal at 4 h, suggesting that polymyxin resistance may occur within 4 h after polymyxin B monotherapy. The antimicrobial effect of cannabidiol has recently gained great attention; however, the precise mode of action is still unclear [34][35][36][37][38]. It was found that cannabidiol inhibited membrane lipid synthesis of S. aureus [38]. This evidence is in line with our findings that showed the inhibitory impact of cannabidiol on levels of essential bacterial membrane lipids at 1 h, including phosphatidylethanolamine, snglycero-3-phosphoethanolamine and 1-hexadecanoyl-2-sn-glycero-3-phosphate ( Figure 2A). Nevertheless, this effect disappeared at the later time exposure of 4 h. Promisingly, the combination caused extensive perturbations (largely at 1 h and to a lesser extent at 4 h) in fatty acids and glycerophospholipids (Figure 2A,B). There has been growing interest in fatty acid biosynthesis as a major target for antibiotic development, for instance, the newly discovered fatty acid elongation inhibitors, platensimycin and platencin [54]. A greater reduction in the levels of key bacterial membrane lipids (e.g., phosphatidylethanolamine, 1,2-diacyl-sn-glycerol 3-phosphate (phosphatidic acid), sn-glycero-3-phosphoethanolamine and sn-glycerol 3-phosphate) was observed after the combination treatment at both time points (1 and 4 h) (Figure 2A,B). In Gram-negative bacteria, phosphatidylethanolamine accounts for 70-80% of total membrane lipids, which are synthesized on the cytoplasmic side of the inner membrane [55,56]. It has been found that an alteration in the levels of phosphatidylethanolamine leads to critically compromised cell integrity and ultimately cell death [44]. Sn-glycerol 3-phosphate is an essential precursor required for the formation of membrane phospholipids; it undergoes acylation at the 1-position to form lysophosphatidic acid (LPA), and then a second acylation step produces phosphatidic acid (PA), the principal component in the synthesis of bacterial membrane glycerolipids [57]. Targeting phosphatidic acid synthesis in Gram-negative bacteria might offer a new and effective strategy to develop novel antibacterial therapeutics [46]. Importantly, our study is the first to report that combining polymyxin B with cannabidiol caused a marked suppression in the components of peptidoglycan and LPS biogenesis at 1 h, including UDP-MurNAc-L-Ala- Abstract: Polymyxins have resurged as the last-resort antibiotics against multidrug-resistant Acinetobacter baumannii. As reports of polymyxin resistance in A. baumannii with monotherapy have become increasingly common, combination therapy is usually the only remaining treatment option. A novel and effective strategy is to combine polymyxins with non-antibiotic drugs. This study aimed to investigate, using untargeted metabolomics, the mechanisms of antibacterial killing synergy of the combination of polymyxin B with a synthetic cannabidiol against A. baumannii ATCC 19606. The antibacterial synergy of the combination against a panel of Gram-negative pathogens (Acinetobacter baumannii, Klebsiella pneumoniae and Pseudomonas aeruginosa) was also explored using checkerboard and static time-kill assays. The polymyxin B-cannabidiol combination showed synergistic antibacterial activity in checkerboard and static time-kill assays against both polymyxinsusceptible and polymyxin-resistant isolates. The metabolomics study at 1 h demonstrated that polymyxin B monotherapy and the combination (to the greatest extent) significantly perturbed the complex interrelated metabolic pathways involved in the bacterial cell envelope biogenesis (amino sugar and nucleotide sugar metabolism, peptidoglycan, and lipopolysaccharide (LPS) biosynthesis), nucleotides (purine and pyrimidine metabolism) and peptide metabolism; notably, these pathways are key regulators of bacterial DNA and RNA biosynthesis. Intriguingly, the combination caused a major perturbation in bacterial membrane lipids (glycerophospholipids and fatty acids) compared to very minimal changes induced by monotherapies. At 4 h, polymyxin B-cannabidiol induced more pronounced effects on the abovementioned pathways compared to the minimal impact of monotherapies. This metabolomics study for the first time showed that in disorganization of the bacterial envelope formation, the DNA and RNA biosynthetic pathways were the most likely molecular mechanisms for the synergy of the combination.   Figure S4). This influence may have arisen from inhibition of amino-sugar and nucleotide-sugar metabolism and the pentose phosphate pathway, which are key sources of precursors for peptidoglycan and LPS synthesis. This was manifested by a significant decline in the levels of UDP-GlcNAc, D-sedoheptulose 7-phosphate and Ru5P after combination treatment (Supplementary Figure S4). UDP-GlcNAc plays a pivotal role as an amino sugar donor in several transferase reactions in the biosynthesis of peptidoglycan, the core lipid A components of the LPS, and certain O-antigens of Gram-negative bacteria [58,59]. The synthesis of KDO, an essential residue of the LPS inner core, is initiated by the enzyme D-arabinose-5-phosphate isomerase (API), which regulates the reversible isomerization of D-ribulose 5-phosphate (Ru5P) to D-arabinose-5-phosphate, a precursor of 3-deoxy-D-manno-octulosonate [60]. It is important to note that the combination treatment remained effective in disorganizing the bacterial envelope formation at 4 h, in which more perturbations in peptidoglycan and LPS biosynthesis were most prominent ( Figure 3A,B). Intriguingly, the combination treatment caused a profound decrease in the abundance of undecaprenyl phosphate (C55-P), also known as bactoprenol, a sugar carrier lipid that mediates the biosynthesis of bacterial extracellular polysaccharides such as cell wall peptidoglycan and LPS [61,62], at 4 h ( Figure 3A,B). It has been proposed that bactoprenol is a potential novel target for new antibacterial agents [63,64].
Consistent with our previous results, it was not unusual that polymyxin B monotherapy caused a significant alteration, though less intense compared to the combination, in the main pathways involved in the cell envelope assembly, including peptidoglycan and LPS biosynthesis at both time points (1 and 4 h) ( Figure 3A,B) [17][18][19]. Remarkably, pathway analysis also showed that despite being a non-antibiotic, cannabidiol reduced the KDO and UDP-GlcNAc levels at 4 h ( Figure 3B). Hence, taking data from all of the abovementioned studies together, a possible mechanism of synergistic killing by the polymyxin B-cannabidiol combination is strongly related to the suppression of amino sugar and nucleotide sugar metabolism and PPP and subsequent peptidoglycan and LPS biosynthesis, which eventually leads to bacterial membrane structure deformity.
Apart from its disruptive impact on the bacterial membrane, the combination therapy produced profound changes in the nucleotide pool of A. baumannii ATCC 19606 at both time exposures (1 and 4 h); however, the maximum effect on the nucleotide pool was at 4 h ( Figure 4A,B). The perturbed nucleotides were mainly those related to purine and pyrimidine metabolism (e.g., ATP, ADP, UDP and GMP), which ultimately impacted DNA and RNA synthesis. Nucleotide metabolism such as purine metabolism is actively involved in antibiotic efficacy and lethality [65,66]. It has been found that dysregulation in purine metabolism can alter antibiotic lethality; this has been manifested by an increase in the bacterial killing activity of gentamicin following genetic deletion of enzymes involved in purine metabolism [65]. Given the significant alteration of the majority of the nucleotides involved in purine and pyrimidine, it is highly likely that the combination treatment compromised DNA and RNA biosynthesis in A. baumannii. Notably, it has been previously reported that polymyxins were able to dysregulate the nucleotide pool of different Gramnegative pathogens such as A. baumannii, K. pneumoniae and P. aeruginosa [50,64,65]. A recent study using radiolabeled macromolecular synthesis assays revealed the ability of cannabidiol to inhibit DNA, RNA and protein synthesis in S. aureus [38]. In line with these findings, polymyxin B and cannabidiol monotherapies demonstrated significant perturbations, albeit to a lesser extent compared to their combination, in the levels of nucleotides at 1 h, which then faded away at 4 h ( Figure 4A,B). In combination, our findings showed polymyxin B and cannabidiol additionally affected peptide metabolism, which most likely led to inhibition of protein synthesis in A. baumannii.
Overall, the novel findings from this study are that polymyxin B monotherapy produces significant changes in multiple cellular metabolic pathways of cell envelope biogenesis, all of which were further disorganized by adding cannabidiol. Furthermore, the combination therapy greatly perturbed the nucleotide and peptide pools, suggesting a unique mechanism (non-bacterial membrane involvement) of synergistic killing for the polymyxin B-cannabidiol treatment. The current findings also showed that cannabidiol affects LPS, DNA, and lipid biosynthesis. The study highlights the importance of elucidating the complex and dynamic interactions of multiple cellular metabolic pathways due to antibiotic-nonantibiotic treatment. Such findings will aid in the understanding of the most likely mechanisms behind the killing synergy and ultimately facilitate future repurposing of cannabidiol as an antimicrobial agent in combination with polymyxins.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/pharmaceutics14040786/s1, Figure S1: Total acquired metabolites and the proportion of each metabolite classes; Table S1