ABSTRACT

   The increasing resistance of clinically relevant microbes against
   current commercially available antimicrobials underpins the urgent need
   for alternative and novel treatment strategies. Cationic lipidated
   oligomers (CLOs) are innovative alternatives to antimicrobial peptides
   and have reported antimicrobial potential. An understanding of their
   antimicrobial mechanism of action is required to rationally design
   future treatment strategies for CLOs, either in monotherapy or
   synergistic combinations. In the present study, metabolomics was used
   to investigate the potential metabolic pathways involved in the
   mechanisms of antibacterial activity of one CLO, C[12]-o-(BG-D)-10,
   which we have previously shown to be effective against
   methicillin-resistant Staphylococcus aureus (MRSA) ATCC 43300. The
   metabolomes of MRSA ATCC 43300 at 1, 3, and 6 h following treatment
   with C[12]-o-(BG-D)-10 (48 µg/mL, i.e., 3× MIC) were compared to those
   of the untreated controls. Our findings reveal that the studied CLO,
   C[12]-o-(BG-D)-10, disorganized the bacterial membrane as the first
   step toward its antimicrobial effect, as evidenced by marked
   perturbations in the bacterial membrane lipids and peptidoglycan
   biosynthesis observed at early time points, i.e., 1 and 3 h. Central
   carbon metabolism and the biosynthesis of DNA, RNA, and arginine were
   also vigorously perturbed, mainly at early time points. Moreover,
   bacterial cells were under osmotic and oxidative stress across all time
   points, as evident by perturbations of trehalose biosynthesis and
   pentose phosphate shunt. Overall, this metabolomics study has, for the
   first time, revealed that the antimicrobial action of C[12]-o-(BG-D)-10
   may potentially stem from the dysregulation of multiple metabolic
   pathways.

IMPORTANCE

   Antimicrobial resistance poses a significant challenge to healthcare
   systems worldwide. Novel anti-infective therapeutics are urgently
   needed to combat drug-resistant microorganisms. Cationic lipidated
   oligomers (CLOs) show promise as new antibacterial agents against
   Gram-positive pathogens like methicillin-resistant Staphylococcus
   aureus (MRSA). Understanding their molecular mechanism(s) of
   antimicrobial action may help design synergistic CLO treatments along
   with monotherapy. Here, we describe the first metabolomics study to
   investigate the killing mechanism(s) of CLOs against MRSA. The results
   of our study indicate that the CLO, C[12]-o-(BG-D)-10, had a notable
   impact on the biosynthesis and organization of the bacterial cell
   envelope. C[12]-o-(BG-D)-10 also inhibits arginine, histidine, central
   carbon metabolism, and trehalose production, adding to its
   antibacterial characteristics. This work illuminates the unique
   mechanism of action of C[12]-o-(BG-D)-10 and opens an avenue to design
   innovative antibacterial oligomers/polymers for future clinical
   applications.

   KEYWORDS: antimicrobial resistance, peptidoglycan biosynthesis,
   cationic lipidated oligomers, antimicrobial peptides, antimicrobial
   polymer, metabolomics, methicillin-resistant Staphylococcus aureus

INTRODUCTION

   Antimicrobial resistance (AMR) represents a pressing global health
   challenge, posing significant threats to the successful treatment of
   infections and patient outcomes ([42]1). Common mechanisms through
   which microbes develop resistance to antimicrobials include reduced
   drug uptake, drug target modification, and modification of drug and
   drug efflux; these may emerge due to genetic mutations or horizontal
   gene transfer ([43]2). The emergence of AMR is largely attributed to
   the inappropriate and excessive use of various antibacterial agents,
   both within the healthcare sector and the agricultural industry
   ([44]3). The dwindling arsenal of effective treatment options poses a
   substantial challenge, jeopardizing our ability to combat infections
   successfully. As a direct consequence, mortality rates are rising, as
   once-treatable infections become increasingly difficult to treat.
   Beyond the immediate health impacts, AMR causes a substantial economic
   burden through heightened healthcare costs, including due to prolonged
   hospital stays, higher likelihood of readmission, and increased
   expenditure on additional treatments ([45]4). Recognizing AMR as a
   critical issue, global health organizations, such as the Infectious
   Diseases Society of America in 2009, the World Health Organization
   (WHO) in 2017, and the Centers for Disease Control and Prevention
   (US-CDC) of the United States in 2019, have emphasized the need for
   multifaceted strategies, including prudent antimicrobial use,
   surveillance, and the development of novel drugs. They also issued
   lists of critical pathogens on which to focus the research and
   development of new antimicrobials ([46]5[47]–[48]8).

   While conventional antimicrobials are currently used for treating
   infections, the rapid increase in AMR suggests the need for alternative
   treatments ([49]9). Antimicrobial peptides (AMPs) have been
   investigated and have been demonstrated to be very effective in the
   killing of microbes ([50]10[51]–[52]12). However, AMPs have
   limitations, such as instability, toxicity, and high production costs
   ([53]13, [54]14). Recently, synthetic analogs of AMPs, i.e.,
   antimicrobial polymers that mimic the structural features (cationic and
   hydrophobic moieties) of naturally occurring AMPs, have been designed
   and shown to overcome AMP limitations in in vitro studies and in vivo
   mouse models ([55]15). This class of molecules has several advantages,
   such as broad-spectrum antimicrobial activities, a low tendency for
   resistance development, and a rapid bactericidal effect ([56]11,
   [57]16[58]–[59]22). Furthermore, antimicrobial polymers offer extra
   advantages compared to AMPs in terms of stability, durability, and ease
   of large-scale production ([60]17). These antimicrobial polymers have a
   broad range of mechanisms of action from membrane disruption to
   intracellular target inhibition, depending on the structural features
   ([61]17). However, the full mechanism(s) of action of antimicrobial
   polymers remains unclear in the literature as detailed
   structure-property relationships are difficult to elucidate.

   Staphylococcus aureus is a Gram-positive bacterial species categorized
   as a high-priority pathogen by the WHO and a serious threat by the
   US-CDC ([62]23, [63]24). S. aureus is part of the normal flora of
   humans and usually does not cause infection while on the skin. However,
   when the skin barrier is damaged, S. aureus may enter the underlying
   tissue or bloodstream and cause a wide variety of infections ([64]25).
   Currently, methicillin-resistant S. aureus (MRSA) is one of the main
   nosocomial pathogens and is prevalent in hospital settings ([65]26). In
   2019, 473,000 deaths were associated with MRSA infections globally
   ([66]27). In Australia, MRSA infections were associated with increased
   inpatient mortality, as well as greater expense and longer hospital
   length of stay compared with methicillin-susceptible S. aureus
   ([67]28).

   In recent work, our group has shown that cationic lipidated oligomers
   (CLOs) demonstrate structure-dependent antimicrobial activity against
   both Gram-positive and Gram-negative bacteria and fungi ([68]29).
   Interestingly, while these CLOs have been designed with the same
   structural features (with cationic residues and lipid tail) as the
   clinically used AMP colistin ([69]30), they appear to have wider
   applicability. The synthesized CLOs have repeating cationic residues
   (e.g., tertiary amine, primary amine [mimicking lysine], guanidine
   [mimicking arginine], or imidazole [mimicking histidine]), which help
   these CLOs to electrostatically bind to the negatively charged
   bacterial membranes ([70]17). The lipid tail in each of the CLOs helps
   to disrupt bacterial membranes more effectively ([71]16). One CLO in
   particular, i.e., C[12]-o-(BG-D)-10, with 10 cationic guanidine
   residues and a C[12] hydrocarbon lipid tail, exhibited marked
   antimicrobial activity against MRSA ([72]29). These guanidine groups
   have been reported to form multidentate hydrogen bonds with sulfate and
   phosphate heads on the bacterial anionic membranes, leading to
   efficient bacterial membrane integration ([73]17, [74]31, [75]32).
   However, their precise mechanism(s) of antimicrobial action remains
   largely unknown to date.

   Metabolomics has emerged as a critical tool for the elucidation of the
   mechanisms of action of AMPs such as polymyxins ([76]33, [77]34). In
   this study, we shine a light on the mechanism(s) of bacterial killing
   by the antimicrobial CLO, C[12]-o-(BG-D)-10, against MRSA strain ATCC
   43300 using untargeted metabolomics.

RESULTS AND DISCUSSION

   The antibacterial activity of C[12]-o-(BG-D)-10 was previously assessed
   against S. aureus ATCC 43300 and demonstrated an appreciable activity
   with an MIC of 8–16 µg/mL ([78]29). Additionally, C[12]-o-(BG-D)-10
   exhibited an interesting mechanism of action when comparing
   dose-dependent membrane disruption (via fluorescence propidium iodide
   [PI] assay) and growth inhibition (Fig. S1). Specifically,
   C[12]-o-(BG-D)-10 exhibited only minor interaction with the bacterial
   membrane, with only ~30% membrane damage observed (relative to a
   melittin control) at the highest concentration tested. These results
   suggested that there could be potentially a secondary mechanism
   contributing to the observed antimicrobial activity for
   C[12]-o-(BG-D)-10 ([79]29). To interrogate this complex mode of action,
   a metabolomics study was performed using an initial inoculum of 10^8
   colony-forming units (CFU)/mL with samples at 1, 3, and 6 h. The 48
   µg/mL (3× MIC) C[12]-o-(BG-D)-10 concentration provided maximal
   bacterial killing at 1 h with ~1.5 log[10]CFU/mL decrease compared to
   the control (Fig. S2a). Metabolomics results of different perturbed
   metabolic pathways of MRSA are discussed below.

Multivariate and univariate analysis

   A total of 1,578 putative metabolites were identified under all
   treatment conditions. Out of these, 33% were not mapped to known
   metabolic pathways and 67% were mapped to known metabolic pathways
   according to common databases, e.g., PseudoCyc, MetaCyc, and LipidMaps
   databases. Most of the acquired metabolites belonged to lipid (18%),
   peptide (18%), and amino acid (17%) metabolism, while the minority of
   metabolites belonged to carbohydrate (6%), nucleotide (4%), secondary
   metabolites (2%), energy (1%), and glycan (0.5%) metabolite classes.
   The same databases were then used to designate metabolic classes or map
   the unmapped metabolites. Univariate data analysis was performed using
   two-sample t-tests (log[2] fold change [FC] ≥ 0.59 or ≤−0.59,
   corresponding to a metabolite-level change of approximately 1.5-fold;
   false discovery rate [FDR] adjusted P-value ≤ 0.05) to determine
   significantly perturbed metabolites across all time points (1, 3, and 6
   h) ([80]Table S1). This analysis identified ~476 significantly
   perturbed metabolites (295, 241, and 185 at 1, 3, and 6 h,
   respectively) (Fig. S3). Across all time points, there were 53
   overlapping metabolites with 155, 72, and 57 unique metabolites at 1,
   3, and 6 h, respectively (Fig. S3). The majority of these metabolites
   were diminished in response to treatment with C[12]-o-(BG-D)-10. The
   heatmap showed that the intensities of metabolites varied after
   treatment with C[12]-o-(BG-D)-10 across all time points, especially at
   1 h (Fig. S4). The reproducibility for all sample groups was acceptable
   across all time points (1, 3, and 6 h), where the median relative
   standard deviations (RSDs) across all time points were 14%–16% for
   untreated (control) groups, 17%–22% for treated samples, and 12% for
   the quality control (QC) group, consistent with some baseline
   variability in the dynamics of ordinary bacterial metabolism with and
   without C[12]-o-(BG-D)-10 treatment ([81]Table 1). The well-separated
   treatment and control groups in the principal component analysis (PCA)
   revealed that C[12]-o-(BG-D)-10 treatment altered the metabolomic
   profile of MRSA across all time points (Fig. S5). The classification of
   the significantly impacted metabolites across all time points revealed
   that the lipids, peptides, amino acids, and carbohydrate (including
   glycans) metabolites were largely impacted, while the nucleotide
   metabolites and energy metabolites were less significantly perturbed
   across all the time points (Fig. S6).

TABLE 1.

   Median RSDs of all the metabolites of MRSA before and after treatment
   with C[12]-o-(BG-D)-10 across all the time points, i.e., 1, 3, and 6 h
                               Median RSD %
                               1 h 3 h 6 h QC
   Control (untreated)         14  16  16  12
   Treatment C[12]-o-(BG-D)-10 22  22  17
   [82]Open in a new tab

Pathway enrichment analysis for the significantly perturbed metabolites

   C[12]-o-(BG-D)-10 treatment induced extensive perturbations in the
   metabolomic profile of MRSA ATCC 43300 across all time points.
   Consequently, therefore, we mapped and analyzed the significantly
   impacted metabolic features across all time points (i.e., 1, 3, and 6
   h). The mapping of the significantly perturbed features of MRSA
   revealed that glycerophospholipids and fatty acids (FAs) metabolism,
   peptidoglycan and teichoic acid biosynthesis, DNA and RNA
   biosynthesis/nucleotide biosynthesis, central carbon metabolism,
   arginine biosynthesis, histidine metabolism, and pantothenate and
   co-enzyme A (CoA) biosynthesis were among the most significantly
   perturbed pathways (Table S2).

Glycerophospholipid and fatty acid metabolism

   Following C[12]-o-(BG-D)-10 treatment, lipids were significantly
   perturbed, particularly the glycerophospholipids and fatty acid classes
   compared to other classes, i.e., glycerolipids, sphingolipids, and
   sterol lipids. Among glycerophospholipids, long chain
   glycerophosphoglycerols, including PG(25:0), PG(19:0), and PG(14:0),
   were largely perturbed at 1 h ([83]Fig. 1a). The abundance of a highly
   important metabolite involved in the synthesis of bacterial membrane
   lipids and known precursor of teichoic acid in the Gram-positive
   bacterial cell wall, CDP-glycerol, was significantly decreased
   (log[2]FC = −3.01) at 1 h ([84]Fig. 1a) ([85]35). The levels of two
   more well-known metabolites involved in bacterial membrane lipids
   biosynthesis, sn-glycerol 3-phosphate and sn-glycero-3-phosphocholine,
   were substantially decreased at 1 h post-C[12]-o-(BG-D)-10 treatment
   (log[2]FC = −6.69 and −4.73, respectively) ([86]Fig. 1a)
   ([87]36[88]–[89]38). Also, di-trans,poly-cis-undecaprenyl phosphate
   (also known as lipid-P, bactoprenol, and C55-P) was significantly
   perturbed (log[2]FC = 3.9) at 1 h; lipid-P is involved in transporting
   peptidoglycan subunits from the cytoplasmic face of the cell membrane
   through the periplasmic space to the extracellular surface and plays a
   role in the synthesis of teichoic acid of the bacterial cell envelope
   ([90]35, [91]39[92]–[93]42). Significant perturbations were also
   evident in the level of choline phosphate (log[2]FC = −3.0), which has
   a crucial role in the biosynthesis of wall teichoic acid in bacteria
   ([94]43). The abundance of (R)-3-hydroxybutanoate, an essential
   precursor involved in the synthesis of the bacterial membrane, was
   significantly decreased at 1 h (log[2]FC = −2.6) ([95]Fig. 1a)
   ([96]44).

Fig 1.

   [97]Fig 1
   [98]Open in a new tab

   Significantly impacted lipids in MRSA ATCC 43300 following treatment
   with C[12]-o-(BG-D)10 at 1 h (a), 3 h (b), and 6 h (c). Putative lipid
   names are assigned based on accurate mass (≥1.0-log[2]-FC; P < 0.05).
   PE, phosphoethanolamines; PG, glycerophosphoglycerols; PS,
   glycerophosphoserines; PC, glycerophosphocholines; PA,
   glycerophosphates; PI, glycerophosphoinositols; PGP,
   glycerophosphoglycerophosphates; LysoPE, lysophosphatidylethanolamines;
   LysoPA, lysophosphatidic acid; LysoPC, lysophosphatidylcholines; and
   FA, fatty acids.

   At 3 h, the impact of the C[12]-o-(BG-D)-10 on the glycerophospholipid
   levels continued, as manifested by significant decreases of sn-glycerol
   3-phosphate (log[2]FC = −2.02) and sn-glycero-3-phosphocholine
   (log[2]FC = −2.4), albeit less than at 1 h ([99]Fig. 1b). CDP-glycerol
   was not detected in the later time points, i.e., 3 and 6 h. Choline
   phosphate (log[2]FC = −3.7) was further significantly reduced. Notably,
   abundance of glycerophospholipids and fatty acids, including PG(25:0),
   PG(30:0), FA (17:0), FA (20:0), FA (20:1), FA methyl(14:0), FA
   methyl(18:0), and LysoPC(18:1), was significantly increased
   (≥1-log[2]FC, P < 0.05) except for FA oxo(19:0) (≥−1-log[2]FC, P <
   0.05). Importantly, the abundance of N-butyryl-L-homoserine lactone, a
   quorum-sensing signaling molecule, was markedly decreased (log[2]FC =
   −1.00) in response to C[12]-o-(BG-D)-10 treatment ([100]Fig. 1b)
   ([101]45).

   At 6 h, the level of sn-glycero-3-phosphocholine decreased moderately
   (log[2]FC = −0.68), while sn-glycerol 3-phosphate was completely
   diminished and was not detected. The level of choline phosphate
   (log[2]FC = −1.8) was reduced. Most fatty acids and
   phosphatidylglycerol phosphate levels were further increased, including
   PG (30:0), FA hydroxy (4:0), FA hydroxy (8:0), LysoPC (18:1), and PGP
   (36:5) (≥1-log[2]FC, P < 0.05). The level of (R)-3-hydroxybutanoate was
   further significantly decreased at 6 h (log[2]FC = −1.06). Also, the
   levels of [FA (21:2)] octadecenoic acid and FA methyl (18:0) were
   reduced (≥−1-log[2]FC, P < 0.05) ([102]Fig. 1c).

   Given that a higher number of significant perturbations of the lipid
   bilayer was observed at 1 h compared to 3 and 6 h, this suggests that
   C[12]-o-(BG-D)-10 disrupts the lipid bacterial membrane as the first
   step of the bacterial killing mechanism.

Amino-sugar and sugar-nucleotide metabolism, peptidoglycan and teichoic acid
biosynthesis

   Sugar nucleotides and amino sugars are the building blocks of
   peptidoglycans, involved in the synthesis of bacterial cell walls and
   teichoic acid in Gram-positive bacteria ([103]46). After
   C[12]-o-(BG-D)-10 treatment, sugar nucleotides and amino sugar
   metabolism showed substantial alterations, indicating downregulation of
   peptidoglycan biosynthesis across all time points (1, 3, and 6 h). At 1
   h, the treated bacterial cells displayed significant perturbations of
   metabolites involved in the early stages of peptidoglycan and teichoic
   acid formation. In contrast, modest disturbances were observed at 6 h
   ([104]Fig. 2a).

Fig 2.

   [105]Fig 2
   [106]Open in a new tab

   (a) Diagrammatic representation of all significantly impacted
   (increased: blue, decreased: red) amino-sugar and sugar-nucleotide
   metabolites in MRSA ATCC 43300 following treatment with
   C[12]-o-(BG-D)-10. (b) Significantly impacted amino-sugar and
   sugar-nucleotides in MRSA ATCC 43300 following treatment with
   C[12]-o-(BG-D)-10 at 1 h (red), 3 h (green), and 6 h (blue). Putative
   metabolite names are assigned based on accurate mass (≥1.0-log[2]-FC; P
   < 0.05).

   At 1 h after C[12]-o-(BG-D)-10 treatment, the levels of seven crucial
   precursors involved in the biosynthesis of bacterial cell walls
   exhibited a significant decline. These included
   N-acetyl-D-glucosaminate (GlcNAc), N-acetylmuramate (MurNAc),
   N-acetyl-D-glucosamine 6-phosphate (GlcNAc-6P),
   1,6-anhydro-N-acetylmuramate (anhMurNAc), UDP-N-acetylmuramate
   (UDP-MurNAc), and UDP-N-acetyl-D-glucosamine (UDP-GlcNAc)
   (≥−2-log[2]FC, P < 0.05) ([107]Fig. 2b). Likewise, the precursors of
   the downstream pathway, peptidoglycan, including D-alanyl-D-alanine,
   UDP-N-acetylmuramoyl-L-alanine (UDP-MurNAc-L-Ala),
   UDP-MurAc(oyl-L-Ala-D-gamma-Glu-L-Lys-D-Ala-D-Ala),
   UDP-N-acetyl-2-amino-2-deoxy-D-glucuronate, and L-alanine were also
   markedly reduced (≥−1-log[2]-FC, P < 0.05). Interestingly,
   C[12]-o-(BG-D)-10 caused a significant impact on the crucial elements
   of wall teichoic acid. Notably, it decreased the levels of key
   constituents namely MurNAc and GlcNAc, which also play integral roles
   in amino sugar metabolism, as well as the level of CDP-ribitol, a
   pivotal molecule of wall teichoic acid biogenesis (≥−1.0-log[2]FC, P <
   0.05) ([108]47). The level of choline involved in the synthesis of
   crucial intermediates of teichoic acid was also significantly reduced
   (log[2]FC = −3.32) ([109]48). UDP-glucose, a nucleotide sugar that
   serves as a key intracellular intermediary in the biosynthesis of the
   bacterial cell envelope, was significantly lowered (log[2]FC = −2.1) at
   1 h ([110]Fig. 2b) ([111]49).

   At 3 h, further perturbations in the building blocks of peptidoglycan
   were observed, where the levels of amino and nucleotide sugar
   intermediates involved in the generation of peptidoglycan building
   blocks were significantly reduced. These included D-glucosamine
   6-phosphate, N-acetyl-D-glucosamine, and GlcNAc, which were reduced
   (≥−2-log[2]FC, P < 0.05) ([112]Fig. 2b). Additionally, the level of
   UDP-MurNAc-L-Ala and
   UDP-MurNAc(oyl-L-Ala-D-gamma-Glu-L-Lys-D-Ala-D-Ala) was increased
   (≥1-log[2]FC, P < 0.05), while the level of D-glucosamine and L-alanine
   was decreased (≥−1-log[2]FC, P < 0.05). The level of D-ribitol
   5-phosphate was significantly increased (log[2]FC = 1.7), which has a
   crucial role in the synthesis of wall teichoic acid in the bacterial
   cell envelope ([113]Fig. 2b) ([114]40).

   Contrary to the pronounced effects observed at earlier time points, 1
   and 3 h, C[12]-o-(BG-D)-10 treatment caused a minor effect on sugar
   nucleotides and amino sugars and its downstream pathways (peptidoglycan
   and wall teichoic acid) at 6 h. Notably, there was a significant
   increase in the levels of UDP-glucose, N-acetylneuraminate,
   UDP-MurNAc(oyl-L-Ala-D-gamma-Glu-L-Lys-D-Ala-D-Ala), D-ribitol
   5-phosphate, and L-alanine (≥1-log[2]FC, P < 0.05) ([115]Fig. 2b).

   Similar to the response patterns of glycerophospholipids and fatty
   acids, the amino and nucleotide sugars and metabolites involved in the
   biogenesis of peptidoglycan were more frequently perturbed at 1 h than
   at later time points.

Histidine metabolism

   The histidine metabolic pathway plays a crucial role in fundamental
   regulatory processes in bacteria, including amino acids, purines, and
   thiamine biosynthesis ([116]50). The main metabolic intermediate
   generated by the histidine pathway is
   5′-phosphoribosyl-4-carboxamide-5-aminoimidazole (AICAR). This
   intermediate is at the crossroads between purine-histidine cross talk
   ([117]51). Therefore, the histidine pathway has been extensively
   investigated as a promising therapeutic target for novel antibiotics to
   treat infections caused by Staphylococcus ([118]50, [119]52). Treatment
   with C[12]-o-(BG-D)-10 led to perturbation in several essential
   metabolites within the histidine biosynthetic pathway, notably
   including intermediates like L-glutamine and L-glutamate, both
   recognized as indicators of bacterial stress response ([120]Fig. 3a)
   ([121]53).

Fig 3.

   [122]Fig 3
   [123]Open in a new tab

   (a) Diagrammatic representation of all significantly impacted
   (increased: blue, decreased: red) histidine metabolites in MRSA ATCC
   43300 following treatment with C[12]-o-(BG-D)-10. (b) Significantly
   impacted histidine metabolites in MRSA ATCC 43300 following treatment
   with C[12]-o-(BG-D)-10 at 1 h (red), 3 h (green), and 6 h (blue).
   Putative metabolite names are assigned based on accurate mass
   (≥1.0-log[2]-FC; P < 0.05).

   At 1 h post-C[12]-o-(BG-D)-10 treatment, 10 essential intermediates of
   histidine metabolism were significantly perturbed. Specifically,
   4-imidazolone-5-acetate, N-carbamyl-L-glutamate,
   hydantoin-5-propionate, L-glutamine, and hercynine displayed more
   significant alteration (≥−2-log[2]FC, P < 0.05) compared to
   4-imidazolone-5-propanoate, N-formyl-L-glutamate,
   N-formimino-L-glutamate, L-aspartate, and 2-oxoglutarate (≥−1-log[2]FC,
   P < 0.05) ([124]Fig. 3b).

   At 3 h, alterations were more pronounced than at 1 h. This was evident
   in decreased abundance of L-glutamate, L-glutamine, N-formimino
   L-glutamate, and 4-imidazolone-5-propanoate (≥−1-log[2]FC, P < 0.05),
   accompanied by an increased level of 4-oxoglutaramate (log[2]FC = 3.1).
   Importantly, the level of 5-phosphoribosyl-N-formylglycinamide (FGAR)
   (log[2]FC = 2.4) significantly increased, a key element in AICAR
   synthesis ([125]Fig. 3b).

   At 6 h, further significant perturbations were observed in the
   histidine biosynthetic pathway. FGAR, N-formimino-L-glutamate, and
   L-histidinal levels increased (≥1-log[2]FC, P < 0.05), while the levels
   of hercynine, N-formyl-L-methionine, and N-carbamoyl-L-aspartate
   decreased (≥−1-log[2]FC, P < 0.05) ([126]Fig. 3b). These time-sensitive
   changes underscore a dynamic response of the histidine pathway to the
   treatment.

Nucleotide (purine and pyrimidine) metabolism

   C[12]-o-(BG-D)-10 treatment induced a marked dysregulation in the
   histidine interconnected pathway and nucleotide (purine and pyrimidine)
   metabolism, both are crucial for DNA and RNA formation ([127]Fig. 4a)
   ([128]54). This impact was more pronounced at 1 and 3 h compared to the
   later time point, 6 h ([129]Fig. 4b).

Fig 4.

   [130]Fig 4
   [131]Open in a new tab

   (a) Diagrammatic representation of all significantly impacted
   (increased: blue, decreased: red) pyrimidine and purine metabolites in
   MRSA ATCC 43300 following treatment with C[12]-o-(BG-D)-10. (b)
   Significantly impacted nucleotide and peptide metabolites in MRSA ATCC
   43300 following treatment with C[12]-o-(BG-D)-10 at 1 h (red), 3 h
   (green), and 6 h (blue). Putative metabolite names are assigned based
   on accurate mass (≥1.0-log[2]-FC; P < 0.05).

   At 1 h, a marked decline in the levels of 14 nucleotides was observed
   following C[12]-o-(BG-D)-10 treatment, including ATP, UTP, AMP, and
   dAMP (≥−1.5-log[2]FC, P < 0.05), which ultimately impacted DNA and RNA
   synthesis ([132]Fig. 4b) ([133]55, [134]56). Among nucleotide bases,
   adenine (log[2]FC = −1.2) was significantly reduced ([135]Fig. 4b). An
   important observation at 1 h was that the level of 3,5 cyclic AMP
   required for cyclic di-AMP synthesis was profoundly reduced (log[2]FC =
   −2.7) ([136]Fig. 4b). Cyclic di-AMP is an essential regulator for
   several bacterial biochemical processes, including metabolic pathways
   governing bacterial growth (fatty acid, carbohydrates, and nucleotide
   metabolism), virulence, biofilm formation, and cell cycle progression
   ([137]57). It is also involved in bacterial cell survival mechanisms
   ([138]58). The level of 5,6-dihydrothymine was reduced (log[2]FC =
   −1.6). Furthermore, the level of N-carbamoyl-L-aspartate, another key
   intermediate in the metabolism of purines and pyrimidines, was also
   decreased (log[2]FC = −1.91) ([139]Fig. 4b) ([140]59).

   At 3 h, significant perturbations were observed in the levels of two
   pivotal intermediates of purine and pyrimidine metabolism.
   Specifically, the level of 5,6-dihydrothymine (log[2]FC = −1.2)
   exhibited a decrease, while the level of
   5-phosphoribosyl-N-formylglycinamide (log[2]FC = 2.4) showed an
   increase ([141]59, [142]60). Moreover, the levels of
   N-carbamoyl-L-aspartate and N-formiminoglycine, thymine, and guanine
   demonstrated increases (≥1.0-log[2]FC, P < 0.05) ([143]Fig. 4b).
   However, the level of ATP, which is the major source of energy in the
   bacteria, experienced a further significant reduction (log[2]FC = −1.9)
   ([144]61).

   At 6 h, a notable increase was observed in the level of
   5′-phosphoribosyl-N-formylglycinamide (log[2]FC = 3.7). Conversely, the
   N-carbamoyl-L-aspartate level experienced a further decline (log[2]FC =
   −2.2), while the level of N-formiminoglycine increased (log[2]FC = 2.6)
   ([145]Fig. 4b). Orotate (log[2]FC = −1.9), an essential metabolite in
   the biosynthesis of pyrimidine, was significantly decreased at 6 h
   ([146]Fig. 4b) ([147]62). Among nucleotide bases, the level of guanine
   (log[2]FC = 1.4) was significantly increased ([148]Fig. 4b).

Central carbon metabolism

   Central carbon metabolism, which includes glycolysis, pentose phosphate
   pathway (PPP), and tricarboxylic acid cycle (TCA), is required for
   bacteria to generate energy in the form of ATP. This process provides
   precursors for all the biosynthetic reactions that are required for
   cell survival ([149]63, [150]64). This makes central carbon metabolism,
   i.e., carbon uptake and carbon utilization, an attractive antimicrobial
   target because these processes are typically essential for microbial
   survival ([151]65[152]–[153]67). Innate immunity, which frequently
   deprives bacteria of iron, amino acids, and other essential nutrients,
   serves as an example of the efficacy of this strategy
   ([154]68[155]–[156]70). C[12]-o-(BG-D)-10 treatment caused
   perturbations in all three pathways of central carbon metabolism more
   significant at 1 h compared to later time points ([157]Fig. 5a).

Fig 5.

   [158]Fig 5
   [159]Open in a new tab

   (a) Diagrammatic representation of all significantly impacted
   (increased: blue, decreased: red) central carbon metabolism metabolites
   in MRSA ATCC 43300 following treatment with C[12]-o-(BG-D)-10. (b)
   Significantly impacted central carbon metabolism metabolites in MRSA
   ATCC 43300 following treatment with C[12]-o-(BG-D)-10 at 1 h (red), 3 h
   (green), and 6 h (blue). Putative metabolite names are assigned based
   on accurate mass (≥1.0-log[2]-FC; P < 0.05).

   At 1 h, 18 important intermediates of central carbon metabolism,
   interlinking glycolysis, TCA, and PPP were significantly perturbed
   ([160]Fig. 5b). This included, glycerone phosphate, phosphoenolpyruvate
   (PEP), succinyl-CoA, D-ribose 5-phosphate, 3-phospho-D-glycerate,
   L-cysteine, D-glucose, and sedoheptulose-7-phosphate, 2-oxoglutarate,
   L-alanine, succinate, L-aspartate, citrate, D-gluconic acid, and
   O-phospho-L-serine (≥−1-log[2]FC, P < 0.05) ([161]Fig. 5b).

   The effect of C[12]-o-(BG-D)-10 continued at 3 h, when D-glucose,
   D-gluconic acid, L-glutamate, L-alanine, succinate, PEP, (S)-malate,
   3-phospho-D-glycerate, and succinyl-CoA (≥−1-log[2]FC, P < 0.05) were
   significantly perturbed. O-phospho-L-serine and glycerone phosphate
   (≥1-log[2]FC, P < 0.05) were also perturbed ([162]Fig. 5b).

   At 6 h, most of the metabolites of the central carbon metabolism were
   diminished and were not detected except for L-alanine, which was
   further significantly perturbed (log[2]FC = 1.15) ([163]Fig. 5b).

   Taken together, the crucial central carbon metabolism was largely and
   significantly perturbed at 1 and 3 h, while only minor perturbations
   were observed at 6 h.

Arginine metabolism

   The disruption of arginine (one of the most versatile and
   inter-convertible) metabolism has recently emerged as a powerful
   approach to control and subvert bacterial pathogenesis ([164]71).
   C[12]-o-(BG-D)-10 treatment significantly impacted arginine metabolism,
   particularly at 1 and 3 h, with minimal changes observed after 6 h of
   treatment. ([165]Fig. 6a).

Fig 6.

   [166]Fig 6
   [167]Open in a new tab

   (a) Diagrammatic representation of all significantly impacted
   (increased: blue, decreased: red) arginine metabolism and interrelated
   TCA cycle metabolites in MRSA ATCC 43300 following treatment with
   C[12]-o-(BG-D)-10. (b) Significantly impacted metabolites in the
   arginine biosynthesis and metabolism in MRSA ATCC 43300 following
   treatment with C[12]-o-(BG-D)-10 at 1 h (red), 3 h (green), and 6 h
   (blue). Putative metabolite names are assigned based on accurate mass
   (≥1.0-log[2]-FC; P < 0.05).

   After 1 h of treatment, nine metabolites of arginine biosynthesis,
   including L-aspartate, L-glutamine, L-citrulline, L-ornithine,
   N-acetyl-L-glutamate, N-(L-arginino) succinate, N-acetyl-L-glutamate
   5-semialdehyde, 2-oxoglutarate, and N-acetylornithine, were
   significantly perturbed (≥−1-log[2]-FC, P < 0.05) ([168]Fig. 6b).

   At 3 h, the inhibitory impact of C[12]-o-(BG-D)-10 on arginine
   biosynthesis continued, although at an intensity wherein the levels of
   six important intermediates, including L-glutamine, L-ornithine,
   L-citrulline, N-(L-arginino) succinate, N-acetyl-L-citrulline, and
   N-acetylornithine, were significantly reduced (≥−1-log[2]-FC, P < 0.05)
   ([169]Fig. 6b).

   By 6 h, the level of N-acetyl-L-glutamate was increased (log[2]FC =
   1.0), signifying a bacterial stress response. The levels of
   L-citrulline, N-acetylornithine, and N-acetyl-L-citrulline were
   decreased (≥−1-log[2]FC, P < 0.05) ([170]Fig. 6b).

Coenzyme A biosynthesis

   Pantothenic acid, also known as vitamin B5, is essentially required by
   bacteria to synthesize coenzyme A (CoA). CoA, in turn, is crucial for
   the generation of fatty acids, carbohydrates, proteins, and even some
   intermediates within the TCA cycle ([171]72). C[12]-o-(BG-D)-10
   treatment resulted in marked disruption in the pantothenate and CoA
   biosynthesis (Fig. S7a).

   At 1 h, C[12]-o-(BG-D)-10 treatment caused significant perturbations in
   the levels of five metabolites related to the pantothenate and CoA
   biosynthesis pathways (Fig. S7b). In particular, the levels of
   L-cysteine, CoA, L-aspartate, and 5,6-dihydrouracil significantly
   decreased (≥−1.0-log[2]FC, P < 0.05) (Fig. S7b). In contrast, uracil
   level increased (log[2]FC = 1.04). At 3 h, there was a notable decrease
   (≥−2-log2FC, P < 0.05) in the N-((R)-pantothenoyl)-L-cysteine and
   (R)−2,3-dihydroxy-3-methylbutanoate. Additionally, the level of
   pantothenate decreased (log[2]FC = −1.5), whereas
   3-methyl-2-oxobutanoic acid increased (log[2]FC = 1.4) (Fig. S7b). The
   effect of C[12]-o-(BG-D)-10 persisted at 6 h, when the levels of
   pantetheine and pantothenate increased (≥1-log[2]FC, P < 0.05), while
   N-((R)-pantothenoyl)-L-cysteine demonstrated decreases (log[2]FC =
   −2.6) (Fig. S7b).

Homeostasis and stress metabolism

   C[12]-o-(BG-D)-10 treatment induced a significant perturbation in the
   metabolites responsible for bacterial homeostasis.
   D-glucose-6-phosphate, a substrate for trehalose-6-phosphate and
   UDP-glucose synthesis, experienced marked perturbations across all time
   points: at 1 h (log[2]FC = −1.8), 3 h (log[2]FC = −0.7), and 6 h
   (log[2]FC = 1.1) (Fig. S8) ([172]73, [173]74). Trehalose, in its
   dephosphorylated form, serves as a recognized osmoprotectant in
   bacteria during osmotic stress ([174]75). Significant alteration in the
   level of trehalose-6-phosphate at 3 h (log[2]FC = −1.71) indicated that
   the C[12]-o-(BG-D)-10 treatment induced osmotic stress. Choline, a
   precursor to glycine betaine, which acts as a potent osmoprotectant
   underwent significant reduction at 1 h (log[2]FC = −3.3) ([175]76).
   However, choline was not detected at the later time points. Likewise,
   PPP intermediates involved in the maintenance of redox homeostasis in
   the cell were markedly perturbed, which was indicated by a significant
   decrease in the levels of ribose-5-phosphate, NADP+, and CMP
   (≥−1.5-log[2]FC, P < 0.05) at 1 h (Fig. S8) ([176]77). However, these
   PPP intermediates were not detected at later time points.

Conclusion

   MRSA is a serious threat to global health and contributes to nearly
   half a million annual deaths. As further resistance emerges against
   current antimicrobials in clinical use, there is an urgent need for new
   treatment options. Our group has synthesized CLOs modeled on the
   structure of antimicrobial peptides. C[12]-o-(BG-D)-10 was previously
   found to inhibit the growth of MRSA in vitro; however, this activity
   was only partially attributable to membrane disruption as evident via
   fluorescence assay. The full mechanism(s) of its antimicrobial activity
   was not fully understood. This is the first study to investigate the
   mechanism(s) of antimicrobial action of a CLO, i.e., C[12]-o-(BG-D)-10
   using metabolomics. C[12]-o-(BG-D)-10 antimicrobial action commences
   with disrupting the bacterial cell envelope. Early at 1 h, significant
   perturbations were observed in cell membrane lipids and
   glycerophospholipids, along with sugar nucleotides and amino-sugar
   metabolites linked to peptidoglycan and teichoic acid biosynthesis. The
   polymer also affected RNA and DNA biosynthesis and led to pronounced
   perturbations in histidine metabolism (linked to the synthesis of
   purines and pyrimidines), energy metabolism (i.e., arginine and TCA
   cycle), pantothenate biosynthesis, and CoA biogenesis (essentially
   required by cells for survival and normal growth). C[12]-o-(BG-D)-10
   also perturbed central carbon metabolism and the stress pathway in the
   bacteria more prominently at the initial time points, i.e., 1 and 3 h.
   These insights on the mechanisms of action of C[12]-o-(BG-D)-10 will
   enable the rational design of antimicrobial combinations of clinically
   available antimicrobials with C[12]-o-(BG-D)-10 in future in vitro and
   in vivo studies in an approach to achieve synergistic and effective
   bacterial killing.

MATERIALS AND METHODS

CLO and antibiotic stock solution, media, and bacterial isolates

   C[12]-o-(BG-D)-10 was synthesized by Cu(0)-mediated reversible
   deactivation radical polymerization using the protocol by Grace et al.
   ([177]29). The CLO stock solutions were prepared by dissolving the CLO
   in DMSO first, diluting with MilliQ water (to 20% DMSO), and then
   vortexing until clear. DMSO was filtered through 0.22 µm sterile nylon
   filters before use. Methicillin-resistant S. aureus ATCC 43300 was used
   in the study. All susceptibility and time-kill studies were performed
   in cation-adjusted Mueller-Hinton broth (CAMHB; containing 20–25 mg/L
   Ca^2+ and 10–12.5 mg/L Mg^2+; BD, Sparks, MD, USA). Viable counting was
   performed on cation-adjusted Mueller–Hinton agar (CAMHA; containing 25
   mg/L Ca^2+ and 12.5 mg/L Mg^2+; BD, Sparks, MD, USA).

Antibacterial killing kinetics of CLO

   A static concentration time-kill assay of C[12]-o-(BG-D)-10 against
   MRSA ATCC 43300 was performed (Fig. S2). Before the time-kill assay,
   MRSA ATCC 43300 was sub-cultured on a CAMHA plate and then incubated at
   37°C for ~18–24 h. Three colonies were transferred from the CAMHA plate
   to inoculate 10 mL of sterile CAMHB in a 50 mL Falcon tube and
   incubated overnight in a shaking water bath (37°C, 150 rpm, ~16 h). The
   optical density of the bacterial suspension was measured using a
   spectrophotometer, and the suspension was appropriately diluted to
   achieve the targeted initial inoculum of ∼10^6 CFU/mL (Fig. S2b). The
   inoculated tubes were dosed with C[12]-o-(BG-D)-10 (in 20% DMSO) to
   achieve concentrations of 8, 16, and 64 µg/mL. The DMSO concentrations
   in the tubes were ≤0.125%. One culture tube was drug-free as a control.
   At 0, 1.5, 5, 24, 48, and 72 h, 1 mL samples were collected from each
   tube, centrifuged, and washed twice with 0.9% normal saline. The
   samples were then serially diluted in saline plates and plated onto
   CAMHA plates. After 24-h incubation at 37°C, the CFU were counted, and
   the time-kill curves were graphed as log[10] CFU/mL vs time (hours).

Metabolomics sample preparation

   An untargeted metabolomics study was carried out to explore the
   mechanism(s) of action of C[12]-o-(BG-D)-10 against MRSA ATCC 43300
   using a concentration of 48 µg/mL (i.e., 3× MIC). Samples were taken
   and analyzed at the 1-, 3-, and 6-h time points in four biological
   replicates. An overnight culture was prepared by inoculating a single
   colony into 100 mL CAMHB in 250 mL conical flasks (Pyrex) and
   incubating the suspension in a shaker at 37°C and 180 rpm for ~16 h.
   After overnight incubation, log-phase cells were prepared in fresh MHB
   and then incubated for 2 h at 37°C at 180 rpm to the log phase with a
   starting bacterial inoculum of 10^8 CFU/mL. Then, C[12]-o-(BG-D)-10 was
   added to obtain the desired concentration of 48 µg/mL (3× MIC), in
   parallel to a CLO-free control for each replicate. The flasks were then
   incubated at 37°C with shaking at 180 rpm. At each time point (0, 1, 3,
   and 6 h), 15 mL samples were transferred to 50 mL Falcon tubes for
   quenching, and the optical density reading at 600 nm (OD[600]) was then
   measured and normalized to the pre-treatment level of approximately
   ~0.5 with fresh CAMHB. Samples were then centrifuged at 3,220 × g and
   4°C for 10 min, and the supernatants were removed. The pellets were
   stored at −80°C until metabolite extraction. The experiment was
   performed in four biological replicates to reduce the bias from
   inherent random variation.

Metabolomics metabolite extraction

   The bacterial pellets were washed twice in 1 mL of 0.9% saline and then
   centrifuged at 3,220 × g and 4°C for 5 min to remove residual
   extracellular metabolites and medium components. The washed pellets
   were resuspended in a cold extraction solvent
   (chloroform-methanol-water at 1:3:1, vol/vol) containing 1 µM each of
   the internal standards
   3-[(3-cholamidopropyl)-dimethylammonio]−1-propanesulfonate,
   N-cyclohexyl-3-aminopropanesulfonic acid, piperazine-N, N-bis
   (2-ethanesulfonic acid), and Tris. The samples were then frozen in
   liquid nitrogen, thawed on ice, and vortexed to release the
   intracellular metabolites (three times). Next, the samples were
   transferred to 1.5 mL Eppendorf tubes and centrifuged at 14,000 g at
   4°C for 10 min to remove any particulate matter. Finally, 200 µL of the
   supernatant was transferred into injection vials for liquid
   chromatography-mass spectrometry (LC-MS) analysis. An equal volume of
   each sample was combined and used as a QC sample.

   To ensure that our metabolite extraction specifically targeted
   live-stressed bacterial cells, we performed a series of optimization
   steps, including a high inoculum time-kill assay, as illustrated in
   Fig. S2a. By this method, we determined that there was at least 10^6.5
   (i.e., 3 million) CFU/mL present in the samples, which equates to >45
   million viable CFU per harvested sample. Furthermore, it is crucial to
   highlight that the mechanism of action of the polymer, as previously
   tested and confirmed in this study, demonstrates a membrane-damaging
   effect. This characteristic supports the assertion that the percentage
   of dead but intact cells in the pelleted samples is expected to be low.

LC-MS analysis of metabolites

   Both hydrophilic interaction liquid chromatography (HILIC) and
   reversed-phase liquid chromatography (RPLC) coupled with
   high-resolution mass spectrometry were employed to ensure the detection
   of both hydrophilic and hydrophobic metabolites. Samples were analyzed
   on a Dionex U3000 high-performance liquid chromatography system in
   tandem with a Q-Exactive Orbitrap mass spectrometer (Thermo Fisher) in
   both positive and negative ion modes with a resolution at 35,000. The
   HILIC method was described previously in detail ([178]78). Briefly,
   samples maintained at 4°C were eluted through a ZIC-pHILIC column (5
   µm, polymeric, 150 × 4.6 mm; SeQuant, Merck) by mobile phase A (20 mM
   ammonium carbonate) and mobile phase B (acetonitrile). The gradient
   started with 80% mobile phase B at a flow rate of 0.3 mL/min and was
   followed by a linear gradient to 50% mobile phase B over 15 min. The
   Ascentis Express C8 column (5 cm × 2.1 mm, 2.7 µm) (catalog no.
   53,831-U; Sigma-Aldrich) was applied in the RPLC method. The samples
   were controlled at 4°C and eluted by mobile phase A (40% of isopropanol
   and 60% of Milli-Q water with 8 mM ammonium formate and 2 mM formic
   acid) and mobile phase B (98% of isopropanol and 2% of Milli-Q water
   with 8 mM ammonium formate and 2 mM formic acid). The linear gradient
   started from 100% mobile phase A to a final composition of 35% mobile
   phase A and 65% mobile phase B over 24 min at 0.2 mL/min. All samples
   were analyzed within a single LC-MS batch to avoid variations. The
   pooled quality control samples, internal standards, and total ion
   chromatograms were assessed to evaluate the chromatographic peaks,
   signal reproducibility, and stability of the analytes. To assist in the
   identification of metabolites, a mixture of ∼500 metabolite standards
   was analyzed within the same batch.

Data processing, bioinformatics, and statistical analyses

   IDEOM (Identification and Evaluation of Metabolites) was used to
   convert raw data obtained by LC-MS to annotated and hyperlinked
   metabolites ([179]79). ProteoWizard, a freely available software
   library for LC-MS data analysis, was first used to extract mzXML files
   from LC-MS raw sheets. These files were then processed using XCMS, a
   graphical user interface, for peak picking and generating peakML files
   ([180]80). Using MZmatch.R tool, peaks were aligned and filtered with
   minimum detectible intensity of 100,000 and RSD of <0.5 and peak shape
   (coda dw) of >0.8. Using the same MZmatch, the missing peaks were also
   retrieved and annotated. Common sources of noise (contaminant signals,
   peak shoulders, and irreproducible peaks) were removed using the IDEOM
   interface. The gain and loss of protons were corrected in positive and
   negative electrospray ionization mode and then a data-dependent mass
   recalibration (2 ppm) step for putative metabolites was performed.
   Metabolites confirmed with authentic standards were assigned with MSI
   level 1 identification. Putative metabolites (with MSI level 2
   identification) were identified by comparing their accurate masses and
   retention times with the standards in the databases including KEGG
   (Kyoto Encyclopedia of Genes and Genomics), LipidMaps, MetaCyc, and
   preferably EcoCyc. Peak height intensities were used for the
   quantification of the metabolites. Statistical analysis was performed
   using MetaboAnalyst 5.0 ([181]81), a freely available online
   statistical tool. Briefly, putative metabolites (with median RSD ≤ 20%
   and confidence interval ≥ 5) were extracted from IDEOM and tabled per
   time point, then uploaded on MetaboAnalyst 5.0. Data were filtered
   using the interquartile range, normalized by the median, log[2]
   transformed, and autoscaled. Fold change was calculated relative to the
   control from the corresponding time point. Univariate analysis was
   performed using two-sample t-tests (log[2]FC ≥ 0.59 or ≤−0.59,
   corresponding to a metabolite level change of approximately 1.5-fold;
   FDR adjusted P-value ≤ 0.05) to determine significantly perturbed
   metabolites for each time point. Multivariate analysis was performed
   and included the generation of heat maps and PCA plots. Finally, the
   KEGG IDs of metabolites were uploaded to KEGG Mapper ([182]82), and
   pathways were constructed. The individual value plots for the
   significantly perturbed metabolites after treatment with
   C[12]-o-(BG-D)-10 across all the time points can be found in the
   supplemental material (Fig. S9 to S14).

ACKNOWLEDGMENTS