Exploring Antibiotic-Potentiating Effects of Tobramycin–Deferiprone Conjugates in Pseudomonas aeruginosa

Metal ions, including Fe3+, affect the target site binding of some antibiotics and control the porin- and siderophore-mediated uptake of antibiotics. Amphiphilic tobramycins are an emerging class of antibiotic potentiators capable of synergizing with multiple classes of antibiotics against Gram-negative bacteria, including Pseudomonas aeruginosa. To study how the antibiotic-potentiating effect of amphiphilic tobramycins is affected by the presence of intermolecular iron chelators, we conjugated the FDA-approved iron chelator deferiprone (DEF) to tobramycin (TOB). Three TOB-DEF conjugates differing in the length of the carbon tether were prepared and tested for antibacterial activity and synergistic relationships with a panel of antibiotics against clinical isolates of P. aeruginosa. While all TOB-DEF conjugates were inactive against P. aeruginosa, the TOB-DEF conjugates strongly synergized with outer-membrane-impermeable antibiotics, such as novobiocin and rifampicin. Among the three TOB-DEF conjugates, 1c containing a C12 tether showed a remarkable and selective potentiating effect to improve the susceptibility of multidrug-resistant P. aeruginosa isolates to tetracyclines when compared with other antibiotics. However, the antibacterial activity and antibiotic-potentiating effect of the optimized conjugate was not enhanced under iron-depleted conditions, indicating that the function of the antibiotic potentiator is not affected by the Fe3+ concentration.


Introduction
With the rising number of multidrug-resistant (MDR) Gram-negative infections, the Centers for Disease Control and Prevention (CDC), in 2019, categorized Pseudomonas aeruginosa (P.aeruginosa) as a "serious threat" [1][2][3][4].The clinical presentations caused by P. aeruginosa involve severe conditions, such as ventilator-associated pneumonia (VAP), urinary tract infections (UTI), and intra-abdominal infections [3,5,6].Cystic fibrosis (CF) patients are prone to pulmonary infections by the non-mucoid strain of P. aeruginosa, which eventually develops into the mucoid phenotype [7][8][9].In such cases, increased resistance is observed, which can lead to chronic infection with very few therapeutic options [10].Current treatment options to combat MDR P. aeruginosa infections are limited to β-lactams, aminoglycosides, and polymyxins [11].However, resistance to these antibiotics occurs frequently as a result of various resistance mechanisms, including the expression of antibiotic-inactivating enzymes, efflux pumps, modifications of the outer membrane (OM) [12], as well as the extraordinarily high impermeability of the OM.In addition, reduced expression of the OprD porin channel reduces susceptibility to β-lactam antibiotics, such as imipenem (IMI) [12].To bypass the robust OM of Gram-negative bacteria (GNB), β-lactam-based cephalosporins conjugated to siderophores, including the recently approved cefiderocol, have been developed [13].Siderophores are Fe 3+ -chelating molecules (e.g., pyoverdine and pyochelin) that P. aeruginosa releases for essential iron uptake.Siderophores are typically unaffected by the OM barrier and efflux [14,15].Cefiderocol, used against GNB infections, such as complicated UTI and VAP, is a hybrid antibiotic of the cephalosporin ceftazidime linked to catechol 2-chloro-3,4-dihdroxybenzoic acid by a two-carbon-chain linker [13,16].The hybrid structure of cefiderocol results in an increased intracellular concentration in the periplasm of certain GNB and superior stability against serine-and metallo-β-lactamases [16].The catechol moiety in cefiderocol is responsible for extracellular Fe 3+ complexation, mimicking the siderophore released by P. aeruginosa [16].In the periplasmic space, cefiderocol inhibits penicillin-binding proteins involved in the crosslinking of the peptidoglycan chains [17].The unique uptake mechanism of cefiderocol combined with increased stability toward β-lactamases enhances the activity of cefiderocol when compared with ceftazidime and ceftazidime/avibactam against certain MDR, such as P. aeruginosa [18].To overcome the limited therapeutic options to treat MDR P. aeruginosa infections, strategies to potentiate other classes of antibiotics, devoid of potent antipseudomonal activity, are of interest [19][20][21][22].One of these classes is the tetracyclines.Tetracyclines are a class of broad-spectrum antibiotics that interfere with protein translation and synthesis, and are used widely for respiratory infections caused by Mycoplasma pneumoniae, Chlamydia pneumoniae, and Chlamydia psittaci [19].Apart from these species, tetracyclines cover several other Gram-positive and Gram-negative pathogens [23].Tetracyclines enter the periplasm through OM porin channels, such as OmpF and OmpC [24].However, susceptibility to tetracyclines can be decreased due to porin mutation, loss of OmpF porin channels, and reduced OM permeability [25][26][27].In the cytoplasm, the tetracycline-Mg 2+ complex forms a bridge to bind with the 30S bacterial ribosomal unit, thus eliciting a biological response [28].Recent studies have highlighted that tetracycline-Mg 2+ complexation can be hindered by the presence of iron (Fe 3+ ) [29].Several tobramycin (TOB)-based conjugates have been previously studied as tetracycline potentiators to enhance the antipseudomonal activity of tetracyclines.For instance, when the efflux pump inhibitor 1-(1-naphthylmethyl)-piperazine (NMP) was conjugated to the 5-position of TOB, sensitivity to minocycline was improved significantly [30].Mode of action studies indicated that the TOB-NMP conjugate enhances the OM permeability of tetracyclines in P. aeruginosa.In contrast, structure-activity relationship studies have indicated that the amphiphilic nature of the conjugate was critical for the observed antibiotic-potentiating effect [31][32][33].In the present work, we designed amphiphilic hybrid molecules in which the FDA-approved iron (Fe 3+ ) chelator deferiprone (DEF) [34] is conjugated to TOB in the form of TOB-tether-DEF.Three TOB-DEF conjugates (1a-c, Figure 1) containing a variable hydrophobic linker were prepared to study how iron chelation and tether length affect antibiotic potentiation in amphiphilic TOB-DEF conjugates.In addition, we also prepared control compounds 2 (TOB-tether) and 3 (DEF-tether), which are partial fragments of the most effective potentiator TOB-tether-DEF conjugate 1c, to study structure-activity relationships in more detail.

Synthesis of TOB-DEF Conjugates 1a-c and Control Compounds 2 and 3
TOB-DEF conjugates 1a-c were synthesized as outlined in Scheme 1.Initially, the amino functions of TOB were blocked with tert-butoxycarbonyl (Boc)-protecting groups and all hydroxyl groups, except for those at the C-5 position, were protected as silyl ethers by tertbutyldimethylsilyl chloride (TBDMS-Cl) to produce protected TOB analog 4, as previously described [32].The alkylation of alcohol 4 with various 1,n-dibromoalkanes produced terminal bromo-appended protected TOB 5a-c that differed in the lengths of the aliphatic carbon chains with a 70-78% yield.The corresponding azides obtained in quantitative yields by the nucleophilic displacement of the bromide function using sodium azide in N,N dimethylformamide (DMF) at elevated temperatures were then reduced using Pd(OH) 2 on activated carbon to produce the corresponding amines 6a-c with a 55-60% yield [31].Next, we focused on the preparation of the DEF moiety using commercially available maltol as a starting material.Maltol was subjected to an aminolysis reaction with glycine under basic conditions to produce DEF-modified analog 7 bearing carboxylic acid functionality with a 62% yield following a published procedure [35].The coupling of TOB-tethered amines 6a-c with DEF-1-acetic acid 7 was achieved using hexafluorophosphate azabenzotriazole tetramethyl uronium (HATU) as an amide coupling reagent and triethylamine (Et 3 N) as a base to produce protected TOB-DEF conjugates 8a-c with a 50-60% yield.The global deprotection of Boc and TBDMS groups was achieved using HCl in methanol to produce desired TOB-DEF conjugates 1a-c with a 55-68% yield (Scheme 1).We also prepared control compound 2, a TOB-linker analog of 1c devoid of the DEF moiety, and control compound 3, a DEF-linker analog of 1c devoid of the TOB moiety.For the synthesis of control compound 2, compound 4 was alkylated with 1-iodododecane in the presence of potassium hydroxide (KOH) and tetrabutylammonium hydrogen sulfate (TBAHS) in toluene to yield 9, and was then deprotected using methanolic HCl to obtain compound 2 with a 97% yield.Furthermore, control compound 3 was synthesized by adopting an amide coupling reaction of intermediate 7 with dodecylamine in the presence of HATU and Et 3 N (Scheme 2).

TOB-DEF Conjugates 1a-c Potentiate Multiple Classes of Antibiotics against P. aeruginosa PAO1
Following the assessment of antibacterial activity, checkerboard assays [36] were performed to evaluate the antibiotic-potentiating effect of conjugates 1a-c.The synergy of the conjugates with a diverse panel of 11 antibiotics, including OM-impermeable and OMpermeable antibiotics, was initially tested against P. aeruginosa PAO1 (Figure 2).P. aeruginosa PAO1 was selected as previous TOB-based conjugates exhibited the greatest potentiating effect against this organism [30][31][32][33].The high molecular weight (MW >600 Da) antibiotics novobiocin (NOV) and rifampicin (RIF) were selected because of the poor OM permeability of these antibiotics against P. aeruginosa [30][31][32][33].Similarly, tetracyclines were selected, as these molecules typically possess poor antibacterial activity against P. aeruginosa as a result of low permeability [30][31][32][33].On the other hand, β-lactams like ceftazidime (CAZ) and meropenem (MER) are inactivated by β-lactamases and have reduced periplasmic uptake from the loss of porin channels [13,14].Synergistic interactions corresponded to a ≥4-fold reduction in the MICs of both agents.Conjugates 1a-c synergized with OM-impermeable NOV and RIF, with the exception of conjugate 1b, which did not synergize with RIF against P. aeruginosa PAO1.None of the conjugates were able to potentiate levofloxacin (LEV) and the β-lactams MER, IMI, and CAZ, with the exception of conjugate 1c, which reduced the MIC of aztreonam (ATM) 4-fold.Interestingly, all conjugates were able to synergize with the tetracyclines minocycline (MIN), doxycycline (DOX), tigecycline (TIG), and eravacycline (ERV), except for conjugate 1a, which was unable to synergize with ERV.It was noted that conjugate 1c was superior in bringing down the MICs of NOV, RIF, MIN, DOX, TIG, and ATM when compared with the shorter conjugates 1a-b (Figure 2).The interaction of conjugates 1a-c with the selected antibiotics was also evaluated by calculating the fractional inhibitory concentration index (FICI) (Table S2).FICI values of >0.5 but ≤4, ≤0.5, and >4 indicate no interaction, synergism, and antagonism, respectively [36].As the combination of conjugate 1c with various antibiotics resulted in lower FICI values, conjugate 1c showed higher synergy with the antibiotics than conjugates 1a and 1b.In addition, the potentiation effect of the most potent TOB-DEF conjugate 1c with NOV, RIF, MIN, DOX, CAZ, and ATM in PAO1 was also examined in comparison with the gold-standard OM permeabilizer polymyxin B nonapeptide (PMBN).Our studies revealed that PMBN displayed 4-to 8-fold higher potentiation of MIN, DOX, CAZ, and ATM than compound 1c.Moreover, PMBN was more effective than compound 1c in the potentiation of NOV (256-fold) and RIF (32-fold) (Table S2).Following the results against the reference strain P. aeruginosa PAO1, the most effective potentiator analog 1c was selected for further study with clinical isolates of P. aeruginosa (Figure 3).Three P. aeruginosa MDR strains, PA259, PA262, and PA264, resistant to fluoroquinolones, β-lactams.and tetracyclines (Table S3), were used, along with the same panel of antibiotics.At first, the MIC of TOB-DEF conjugate 1c was determined to be >128 µg/mL against all three selected isolates (Table S1).When limiting the concentration of conjugate 1c to 8.5 µM (8 µg/mL), we observed strain-dependent antibiotic potentiation.For instance, conjugate 1c reduced the MIC of NOV 4-to 64-fold and RIF 16-to 64-fold (except PA262), indicating that compound 1c enhanced the permeability of RIF and NOV.Interestingly, conjugate 1c retained excellent tetracycline potentiation.For instance, conjugate 1c consistently reduced the MIC 16-to 64-fold for MIN, 16-to 128-fold for DOX, and 4-to 64-fold for TIG against the three MDR strains.In contrast, the MIC of ERV was only lowered 4-to 8-fold in PA259 and PA264, but not in PA262.While conjugate 1c attained interpretative susceptibility breakpoints of MIN (≤4 µg/mL, Acinetobacter spp.), DOX (≤4 µg/mL, Acinetobacter spp.), and TIG (≤1 µg/mL, Staphylococcus spp.) against PA259 and PA264, the interpretative susceptibility breakpoint (≤0.5 µg/mL, Enterobacter spp.) of ERV was not achieved against the three MDR strains (Figure 3 and Table S4) [37].Collectively, these results suggest that conjugate 1c is a powerful potentiator of tetracyclines against P. aeruginosa.In addition, we also studied tetracycline potentiation in a non-mucoid CF isolate PA095, which was resistant to MIN, DOX, TIG, and ERV (Table S3).While conjugate 1c exhibited very poor antibacterial activity (>128 µg/mL) against PA095 (Table S1), the compound effectively synergized with tetracyclines.For instance, conjugate 1c (8.5 µM) reduced the MIC of MIN, DOX, TIG, and ERV by 8-128-fold (Table S4).The observed synergistic interaction resulted in absolute MIC values of MIN (0.5 µg/mL), DOX (0.125 µg/mL), TIG (2 µg/mL), and ERV (0.5 µg/mL), indicating that the interpretative tetracycline susceptibility breakpoints can be reached in this isolate (Table S4).

Conjugate 1c Exhibited Superior Potentiation of Tetracyclines When Compared with Control Compounds 2 and 3
To understand the structural implications of linking DEF to TOB, the potentiating effects of control compound 2 (conjugate 1c without DEF) and compound 3 (conjugate 1c without TOB) were tested in combination with tetracyclines against P. aeruginosa PAO1 and MDR P. aeruginosa PA259.In P. aeruginosa PAO1, compound 2 potentiated MIN and DOX 16-fold and 4-fold, respectively, while additive interactions were observed with compound 3.Moreover, both control compounds 2 and 3 failed to synergize with TIG and ERV against PAO1.Similarly, control compound 2 potentiated all four tetracyclines in MDR PA259, but no synergy was observed with compound 3. Overall, hybrid 1c induced greatly enhanced tetracycline activity in comparison with control compound 2 in PAO1, indicating that both DEF and TOB were required for optimal potentiation (Figure 4 and Tables S5 and S6).

Tetracycline Potentiation of Conjugate 1c Is Reduced under Iron-Depleted (ID) Conditions
The antibiotic-potentiating effect of conjugate 1c was also assessed in iron-depleted cation-adjusted Mueller Hinton broth (ID-CAMHB) to probe if the molecule exhibited a mechanism similar to that of the siderophore-mediated uptake of cefiderocol [16].Under iron-depleted (ID) conditions, the susceptibility of P. aeruginosa PAO1 to cefiderocol improved 8-fold, indicating that ID conditions facilitate the siderophore-dependent uptake of cefiderocol.However, no decrease in the MIC of cefiderocol was observed in ID-CAMHB with MDR P. aeruginosa strains PA259, PA262, and PA264 (Table S7).Interestingly, the exposure of conjugate 1c in combination with tetracyclines (MIN, DOX, TIG, and ERV) to ID-CAMHB against the three MDR P. aeruginosa strains resulted in lower potentiation in the majority of cases when compared with CAMHB.This indicates that ID conditions do not enhance the tetracycline-potentiating effect of conjugate 1c (Table 1).Synergistic combinations are highlighted in green.

Compound 1c Disrupts the Outer Membrane of P. aeruginosa Isolates
Our preliminary checkerboard screening with TOB-DEF (1c) and hydrophobic antibiotics, such as NOV and RIF, against wild-type and MDR-P.aeruginosa isolates revealed a significant synergistic effect with NOV and RIF (Figures 2 and 3).These findings suggested that compound 1c functioned as an OM permeabilizer.To further ascertain the effect of compound 1c on the disruption of OM integrity, we performed the checkerboard assay with compound 1c and tetracyclines in the presence of Mg 2+ -enriched CAMHB media against wild-type PAO1 and MDR PA259.It is believed that bivalent cations reduce the negative charges of adjacent LPS molecules in the OM, thereby stabilizing the crosslinking of LPS [36][37][38].These studies revealed that high Mg 2+ (20 mM) concentrations abolished the MIN-and DOX-potentiating effects of 1c in comparison with standard CAMHB media (Table 2).It is also worth noting that, in Mg 2+ -supplemented CAMHB, the MICs of MIN and DOX were also elevated 8-to 16-fold against PAO1 and PA259.Collectively, the data suggest that conjugate 1c potentiated tetracyclines in P. aeruginosa by enhancing their OM uptake.These results are consistent with the tetracycline-potentiating effect of related TOB-based OM permeabilizers [30][31][32][33][34]39]. Synergistic combinations are highlighted in green.

Cytotoxicity Study of Conjugates 1a-c
One of the concerns of membrane-active agents is the risk of inducing nonselective cytotoxicity [30][31][32].Therefore, conjugates 1a-c, positive control doxorubicin, and negative control polymyxin B were tested for cytotoxicity against the HEK293 and HepG2 cell lines.Our results indicate that longer tethers enhanced the cytotoxicity of the conjugates.The best potentiator 1c, at its active concentration of 8.5 µM, reduced the cell viability of HEK293 and HepG2 cells to 75% and 70%, respectively, relative to those of the controls with vehicles.Moreover, increased concentrations of 1c resulted in further reduced cell viabilities in both cell lines, which were absent in polymyxin B (Figure S1).

Discussion
Bifunctional amphiphilic TOB conjugates in which the TOB moiety is linked to antibiotics (ciprofloxacin, moxifloxacin, and RIF), efflux pump inhibitors (NMP), and metal chelators (cyclam) via a hydrophobic spacer are effective antibiotic potentiators that enhance the OM permeability of multiple classes of antibiotics [30][31][32][33].TOB-based bifunctional conjugates are believed to destabilize the OM by displacing bivalent cations (Mg 2+ or Ca 2+ ), which are stabilizing counterions for the phosphate groups of lipid A and the phosphorylated core sugars that prevent repulsion between and among individual LPS molecules.This leads to a localized disruption of LPS in the OM, allowing non-porin-mediated passage of antibiotics into the periplasm [40][41][42][43].The structure-activity relationships determined for bifunctional amphiphilic TOB conjugates have revealed that the nature of the spacer and its amphiphilicity are critical for optimal antibiotic potentiation [30][31][32][33].In this study, we extended the design of bifunctional amphiphilic TOB conjugates to TOB-DEF conjugates.DEF is an FDA-approved Fe 3+ -chelator that could serve as a siderophore mimic to shuttle Fe 3+ ions through the OM of GNB, including P. aeruginosa.We were interested in studying how the Fe 3+ chelating properties affect the antibiotic-potentiating effect of amphiphilic TOB.Three TOB-DEF conjugates 1a-c differing in the length of the hydrophobic spacer were prepared and lacked standalone antibacterial activity (MIC >128 µg/mL) against standard GNB reference strains, which was consistent with previous findings [30][31][32][33].A lack of standalone antibacterial activity is desired for an antibiotic potentiator to reduce the risk of rapid resistance development [30][31][32][33].Subsequently, conjugates 1a-c were screened in combination with a panel of 11 antibiotics against the reference P. aeruginosa PAO1 strain.
The screening indicated that TOB-DEF conjugate 1c with the most extended spacer (C 12 ) displayed the highest antibiotic-potentiating effect at a fixed concentration of 8.5 µM.Conjugate 1c was able to synergize with OM-impermeable RIF and NOV (8-fold), consistent with a destabilizing effect of 1c on the OM.In addition, conjugate 1c showed high selectivity for potentiating tetracyclines against P. aeruginosa PAO1.For instance, we observed reductions in the MICs of DOX (64-fold) and TIG (32-fold).In contrast, conjugate 1c did not potentiate (≤2-fold) β-lactams, such as CAZ, MER, and IMI, while a 4-fold potentiation was observed with ATM.The encouraging potentiating effects of 1c in P. aeruginosa PAO1 prompted us to extend the study to MDR P. aeruginosa isolates.When compared with P. aeruginosa PAO1, equal, but mostly greater, antibiotic potentiation was observed in the MDR PA259 and PA264 strains, indicating that MDR P. aeruginosa strains were more susceptible to the antibiotic potentiation of 1c.However, a reduced antibiotic-potentiating effect was noted in MDR PA262.When compared with all three MDR P. aeruginosa strains, conjugate 1c consistently showed the greatest synergy with tetracyclines MIN, DOX, and TIG.To further understand the structural requirements of conjugate 1c to induce selective tetracycline potentiation in P. aeruginosa, we explored the tetracycline-potentiating effect of control compounds 2 and 3.These studies confirm that optimal tetracycline potentiation requires the presence of TOB, DEF, and a C 12 spacer.We also explored the synergistic relationship of conjugate 1c with tetracyclines in P. aeruginosa strains using CAMHB and ID-CAMHB to study how low Fe 3+ concentrations affect antibiotic potentiation.These studies show that, under ID conditions, slightly reduced tetracycline potentiation and slightly reduced antibacterial activity of 1c were observed in most MDR P. aeruginosa strains.Therefore, ID conditions did not improve the tetracycline potentiation of conjugate 1c, which was unexpected and remains poorly understood.Furthermore, in order to understand the mode of action of why conjugate 1c potentiates tetracycline antibiotics, we studied the tetracycline-potentiating effects in the presence of elevated Mg 2+ concentrations.Traditional OM permeabilizers like PMBN compete with bivalent metal ions for the LPS binding site in the OM which typically results in greatly reduced antibacterial activity or antibiotic potentiation [40][41][42][43].Increasing the concentration of [Mg 2+ ] to 20 mM resulted in a complete loss of tetracycline potentiation, which confirms that conjugate 1c targeted the OM of P. aeruginosa.Unfortunately, optimized conjugate 1c with an amphiphilic C 12 tether displayed greatly increased cytotoxicity against two cell lines at relevant concentrations (8.5 µM) when compared with the less amphiphilic conjugates 1a and 1b.Therefore, further structural optimization is required to reduce the cytotoxicity of hit compound 1c.

Checkerboard Assay
The checkerboard assay was carried out in 96-well plates.The adjuvants under study were diluted along the ordinate, while the antibiotics were diluted along the abscissa.The plates were incubated with equal volumes of inoculum, prepared similarly as discussed above, at 37 • C for 18 h.The growth pattern was then observed using an EMax Plus microplate reader (Molecular Devices, Sunnyvale, CA, USA).The MICs were determined as the lowest concentrations of drugs responsible for inhibition of growth [30].Successively, FICIs were calculated for the antibiotic combinations using the formula: FICIs of ≤0.5, 0.5 < x ≤4, and >4 correlated to synergistic, additive, and antagonistic interactions of the antibiotic combination, respectively [36].ID-CAMHB was prepared from CAMHB as reported previously [46].The determination of the MICs was performed in biological duplicates.If the values were not within 2-fold in agreement, then the assay was repeated.

Cell Viability Assay
Toxicity against HEK293 and HepG2 Cells Human embryonic kidney (HEK293) and human hepatoma (HepG2) cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) in a humidified 5% atmospheric CO 2 incubator at 37 • C. Equal numbers of cells in 50 µL media (5000-HEK293 and 8000-HepG2) were plated in designated wells in a 96-well plate.The wells with only media and no cells served as blanks.After 24 h of incubation, experimental wells with cells and corresponding blank wells were incubated with drugs in 50 µL to the desired concentration (0-200 µM) for 48 h.To assess the cell viability, PrestoBlue reagent from Invitrogen was added to the wells to a final concentration of 10% (v/v), and plates were incubated in the CO 2 incubator for an additional 1 h.Subsequently, the fluorescence was measured with excitation and emission wavelengths of 560 and 590 nm, respectively, using the SpectraMax M2 plate reader from Molecular Devices.The cell viability was interpreted as previously stated [47].The values from blank wells were subtracted from those from the corresponding wells with cells.Finally, the cell viability was calculated relative to that of the controls with vehicle.The data were plotted as line graphs, and the plots indicated the means ± standard deviations of two individual experiments, with five wells with cells dedicated to each concentration.

Conclusions
TOB-DEF conjugates form a new class of selective tetracycline antibiotic potentiators against P. aeruginosa.The potentiating effect of conjugate 1c required a hydrophobic tether and the presence of the Fe 3+ chelator DEF in order to achieve optimal tetracycline potentiation.Similar tether effects were previously observed with related amphiphilic tobramycin conjugates that display considerable cytotoxicity once the carbon tether reaches a length of C 8 or higher [30][31][32][33].In addition, the reasons why DEF was required remain unclear, as Fe 3+ -depleted conditions in the bacterial media had little consequence and mostly showed no improvement in the antibacterial activity or antibiotic potentiation.Conjugate 1c (8.5 µM) is believed to perturb the LPS layer of P. aeruginosa's OM, leading to a transient destabilization of the OM that increases the permeability of tetracycline antibiotics, as seen for other polybasic amphiphiles [30,41,42].This conclusion is supported by the fact that tetracycline potentiation of 1c was abolished in the presence of elevated Mg 2+ concentrations (20 mM), as Mg 2+ competes with the binding of 1c to LPS.Collectively, our data suggest that the observed synergy of conjugate 1c with tetracyclines is independent of the Fe 3+ -complexing properties of the molecule and is the result of enhanced OM permeability in P. aeruginosa.In addition, the reasons for the observed selectivity of tetracycline potentiation of 1c when compared with PMBN remain unclear.

Data Availability Statement:
The data presented in this study are available within the article and the Supplementary Material.

Figure 2 .
Figure 2. Fold potentiation of select antibiotics against wild-type P. aeruginosa PAO1 in the presence of 8.5 µM of TOB-DEF conjugates 1a-c; a ≥4-fold potentiation indicates synergism.

2. 4 .
Conjugate 1c Synergizes with a Panel of Antibiotics against MDR Isolates of P. aeruginosa

Figure 3 .
Figure 3. Fold potentiation of select antibiotics against MDR clinical isolates of P. aeruginosa in the presence of 8.5 µM of compound 1c; a ≥4-fold potentiation indicates synergism.

Figure 4 .
Figure 4. Comparison of the fold potentiation of tetracyclines against wild-type P. aeruginosa PAO1 and clinical isolate P. aeruginosa PA259 in the presence of 8.5 µM of compound 1c, 2, or 3; a ≥4-fold potentiation indicates synergism.

Figure S1 :
Cytotoxicity data for compounds 1a-c; Figure S2: 1 H and 13 C NMR spectra of compound 1a in D 2 O; Figure S3: COSY and HSQC NMR spectra of compound 1a in D 2 O; Figure S4: HMBC NMR spectrum of compound 1a in D 2 O; Figure S5: 1 H and 13 C NMR spectra of compound 1b in D 2 O; Figure S6: COSY and HSQC NMR spectra of compound 1b in D 2 O; Figure S7: HMBC NMR spectrum of compound 1b in D 2 O; Figure S8: 1 H and 13 C NMR spectra of compound 1c in D 2 O; Figure S9: COSY and HSQC NMR spectra of compound 1c in D 2 O; Figure S10: HMBC NMR spectrum of compound 1c in D 2 O; Figure S11: 1 H and 13 C NMR spectra of compound 2 in D 2 O; Figure S12: COSY and HSQC NMR spectra of compound 2 in D 2 O; Figure S13: HMBC NMR spectrum of compound 2 in D 2 O; Figure S14: 1 H and 13 C NMR spectra of compound 3 in D 2 O; Figure S15: 1 H NMR spectra of compound 5a and 5b in CDCl 3 ; Figure S16: 1 H NMR spectra of compound 5c and 6a in CDCl 3 ; Figure S17: 1 H NMR spectra of compound 6b and 6c in CDCl 3 ; Figure S18: 1 H and 13 C NMR spectra of compound 7 in D 2 O; Figure S19: 1 H and 13 C NMR spectra of compound 8a in CDCl 3 ; Figure S20: 1 H and 13 C NMR spectra of compound 8b in CDCl 3 ; Figure S21: 1 H and 13 C NMR spectra of compound 8c in CDCl 3 ; Figure S22: 1 H NMR spectrum of compound 9 in CDCl 3 ; HPLC chromatograms of compounds 1a-c.Author Contributions: Conceptualization, K.G., S.D. and F.S.; Formal analysis, K.G., S.D. and F.S.; Validation analysis, K.G., S.D. and F.S.; Writing, K.G., S.D. and F.S.; Biological data acquisition, K.G., R.A., D.M.R. and D.R.; Student supervision, S.D., G.A. and F.S.; Project administration, S.D., F.S. and G.A.; Funding acquisition, F.S. All authors have read and agreed to the published version of the manuscript.Funding: This research was funded by the Natural Sciences and Engineering Research Council of Canada (NSERC) in the form of a discovery grant (2018-06047) and by the Canadian Institutes of Health Research (CIHR) in the form of a pilot project (162159) and project grant (169664).Institutional Review Board Statement: Not applicable.Informed Consent Statement: Not applicable.