A Selective Membrane-Targeting Repurposed Antibiotic with Activity Against Persistent Methicillin-Resistant Staphylococcus Aureus

A selective membrane-targeting repurposed antibiotic with activity against persistent methicillin-resistant Staphylococcus aureus

Wooseong Kima, Guijin Zoub, Taylor P. A. Haric, Ingrid K. Wiltc, Wenpeng Zhub, Nicolas Galleb, Hammad A. Faizid, Gabriel L. Hendricksa, Katerina Toria, Wen Pana, Xiaowen Huanga,e, Andrew D. Steelec, Erika E. Csataryc, Madeline M. Dekarskec, Jake L. Rosenc, Noelly de Queiroz Ribeirof, Kiho Leea, Jenna Porta, Beth Burgwyn Fuchsa, Petia M. Vlahovskag, William M. Wuestc, Huajian Gaob, Frederick M. Ausubelh,i,1, and Eleftherios Mylonakisa,1

aDivision of Infectious Diseases, Rhode Island Hospital, Warren Alpert Medical School of Brown University, Providence, RI 02903; bSchool of Engineering, Brown University, Providence, RI 02903; cDepartment of Chemistry and Emory Antibiotic Resistance Center, Emory University, Atlanta, GA 30322; dDepartment of Mechanical Engineering, Northwestern University, Evanston, IL 60208; eDepartment of Dermatology, Nanfang Hospital, Southern Medical University, Guangzhou 510515, China; fDepartment of Microbiology, Instituto de Ciências Biológicas, Universidade Federal de Minas Gerais, Belo Horizonte, MG 31270-901, Brazil; gDepartment of Engineering Sciences and Applied Mathematics, Northwestern University, Evanston, IL 60208; hDepartment of Molecular Biology, Massachusetts General Hospital, Boston, MA 02114; and iDepartment of Genetics, Harvard Medical School, Boston, MA 02115

Contributed by Frederick M. Ausubel, June 17, 2019 (sent for review March 22, 2019; reviewed by Dale L. Boger and Ruhong Zhou) Treatment of Staphylococcus aureus infections is complicated by has sparked new interest in membrane-active antimicrobial the development of antibiotic tolerance, a consequence of the therapeutic agents (7). The lipophilic tail of daptomycin, a nat- ability of S. aureus to enter into a nongrowing, dormant state in ural cyclic lipopeptide synthesized by Streptomyces roseosporus, which the organisms are referred to as persisters. We report that inserts into Gram-positive bacterial membranes, forming oligo- the clinically approved anthelmintic agent bithionol kills methicillin- meric pores, causing membrane depolarization, potassium ion S. aureus resistant (MRSA) persister cells, which correlates with its efflux, and rapid cell death (8). Despite strong antimicrobial ability to disrupt the integrity of Gram-positive bacterial membranes. potency against growing bacterial cells, daptomycin has not been Critically, bithionol exhibits significant selectivity for bacterial com- MICROBIOLOGY pared with mammalian cell membranes. All-atom molecular dynam- reported to be effective against persisters (9, 10). ics (MD) simulations demonstrate that the selectivity of bithionol for Recently, our laboratory described the identification of membrane- bacterial membranes correlates with its ability to penetrate and em- active synthetic retinoids that are efficacious in killing MRSA bed in bacterial-mimic lipid bilayers, but not in cholesterol-rich mammalian-mimic lipid bilayers. In addition to causing rapid mem- Significance brane permeabilization, the insertion of bithionol increases membrane fluidity. By using bithionol and nTZDpa (another membrane- There is a critical lack of therapeutic agents to treat infections active antimicrobial agent), as well as analogs of these compounds, caused by nongrowing persister forms of methicillin-resistant we show that the activity of membrane-active compounds against Staphylococcus aureus (MRSA). Although membrane-disrupting MRSA persisters positively correlates with their ability to increase agents can kill persister cells, their therapeutic potential has membrane fluidity, thereby establishing an accurate biophysical in- been mostly overlooked because of low selectivity for bacterial dicator for estimating antipersister potency. Finally, we demonstrate versus mammalian membranes. We report that the clinically that, in combination with gentamicin, bithionol effectively reduces approved anthelmintic drug bithionol kills MRSA persisters by bacterial burdens in a mouse model of chronic deep-seated MRSA disrupting membrane lipid bilayers at concentrations that ex- infection. This work highlights the potential repurposing of bithionol hibit low levels of toxicity to mammalian cells. The selectivity as an antipersister therapeutic agent. of bithionol results from the presence of cholesterol in mammalian but not in bacterial membranes. We also show that the MRSA | bacterial persister | drug repurposing | membrane-active antipersister potency of membrane-active antimicrobial agents antimicrobials | membrane selectivity correlates with their ability to increase membrane fluidity. Our results significantly enhance our understanding of bacterial taphylococcus aureus is a Gram-positive opportunistic human membrane disruption and membrane selectivity. Spathogen carried by approximately one third of the human S. aureus Author contributions: W.K., G.Z., T.P.A.H., I.K.W., W.Z., N.G., H.A.F., G.L.H., W.M.W., H.G., population. Despite antibiotic availability, infections F.M.A., and E.M. designed research; W.K., G.Z., T.P.A.H., I.K.W., W.Z., N.G., H.A.F., G.L.H., are often hard to cure and remain one of the major causes of K.T., W.P., X.H., N.d.Q.R., K.L., J.P., and B.B.F. performed research; T.P.A.H., I.K.W., A.D.S., death (1), in part because of the ability of S. aureus to enter into E.E.C., M.M.D., J.L.R., B.B.F., P.M.V., W.M.W., H.G., F.M.A., and E.M. contributed new a nongrowing antibiotic-tolerant state, in which the organisms reagents/analytic tools; W.K., G.Z., T.P.A.H., I.K.W., W.Z., N.G., H.A.F., G.L.H., P.M.V., W.M.W., H.G., F.M.A., and E.M. analyzed data; and W.K., G.Z., W.M.W., H.G., F.M.A., are referred to as persisters (2). Persisters exhibit significantly and E.M. wrote the paper. reduced biosynthetic processes, which are the major targets for Reviewers: D.L.B., The Scripps Research Institute; and R.Z., Columbia University. most antibiotics (2). Persisters also exist in a metabolically low- Conflict of interest statement: F.M.A. and E.M. have financial interests in Genma Biosci- energy state (3) that prevents the energy-dependent uptake of ences and Octagon Therapeutics, companies that are engaged in developing antimicro- antibiotics such as aminoglycosides (4). S. aureus persisters are bial compounds. E.M.’s and F.M.A.’s interests were reviewed and are managed by Rhode present in high numbers in stationary-phase suspension cultures Island Hospital (E.M.) and Massachusetts General Hospital and Partners HealthCare – (F.M.A.) in accordance with their conflict of interest policies. The remaining authors de- and biofilms (4 6) and are responsible for chronic and relapsing clare no competing financial interests. infections such as endocarditis, osteomyelitis, and prosthetic Published under the PNAS license. implant infections (3). 1To whom correspondence may be addressed. Email: [email protected] Bacterial membranes are attractive antipersister targets be- or [email protected]. cause they can be disrupted independently of growth. Although This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. membrane-active agents are typically toxic to mammals because 1073/pnas.1904700116/-/DCSupplemental. of low membrane selectivity (7), the clinical success of daptomycin Published online July 29, 2019.

www.pnas.org/cgi/doi/10.1073/pnas.1904700116 PNAS | August 13, 2019 | vol. 116 | no. 33 | 16529–16534 Downloaded by guest on September 29, 2021 persisters (6). They are also relatively nontoxic because they ex- AB Control hibitasignificantamountofselectivityforGram-positivebacterial Bithionol OH OH membranes compared with mammalian membranes (6). Like Cl S Cl daptomycin, the antimicrobial retinoids also appear to insert into Gram-positive bacterial membranes, causing rapid permeabilization Cl Cl and cell death. In contrast to daptomycin, however, a subset of the membrane-active synthetic retinoids also kill nongrowing MRSA persister cells (6). In addition to the synthetic retinoids, C S. aureus MW2 persisters DS. aureus MW2 biofilms 9 9 we have described 2 additional membrane-active compounds with 8 8 8 anti-MRSA activity: NH125, a histidine-kinase inhibitor (11), and 7 7 7 6 6 6 nTZDpa, a nonthiazolidinedione peroxisome proliferator-activated 5 5 4 4 5 Log (CFU/mL) receptor gamma partial agonist (12). Although NH125 and nTZDpa 3 Log3 (CFU/mL) 4 have excellent anti-MRSA persister activity, they both cause 01234 01234 3 Time (h) substantial hemolysis of red blood cells at high concentrations Time (h) Log (CFU/membrane) 32 μg/mL BT 0 μg/mL BT > μ 100 μg/mL Van 100 μg/mL Gm Van Gm ( 32 g/mL) (12, 13), which is a drawback for further devel- 50 μg/mL Cipro 200 μg/mL Lin 16 μg/mL BT 8 μg/mL BT μg/mL μg/mL opment as anti-MRSA therapeutic agents. 100 μg/mL Dap Non-treated Non-treated μg/mL Cipro 100 100 The membrane-active retinoids, NH125, and nTZDpa were E S. aureus VRS1 Persisters 50 ∼ 9 9 8 identified by screening 82,000 synthetic compounds by using a 8 8 7 7 7 – 6 high-throughput Caenorhabditis elegans MRSA infection screen- 6 6 ing assay for compounds that block the ability of MRSA to kill the 5 5 5 nematodes (6, 11, 12). We subsequently screened 185 “hit” com- 4 4 4 Log (CFU/mL) 3 Log3 (CFU/mL) pounds for the additional ability to permeabilize MRSA persisters 3 01234 01234 Log (CFU/membrane) 0 8 16 32 Time (h) Time (h) by using a SYTOX Green membrane permeability assay (11). BT (μg/mL) Another putative membrane-active antimicrobial identified in this 200 μg/mL Lin Non-treated 32 μg/mL BT 16 μg/mL BT 100 μg/mL Dap 8 μg/mL BT 0 μg/mL BT screen was the clinically approved anthelminthic drug bithionol, which was chosen for additional studies because of its well- Fig. 1. Bithionol exhibits bactericidal activity against S. aureus persisters. (A) established pharmacokinetic, toxicity, and safety profiles (14). In Chemical structures of bithionol. (B) TEM micrographs showing the formation this paper, we describe the therapeutic potential of bithionol as an of intracellular mesosome-like structures (red arrows), an abnormal cell di- anti-MRSA persister antibiotic agent and provide a putative mode vision (brown arrow), or cell lysis (blue arrow) in S. aureus MW2 treated with × μ of action that explains the ability of bithionol to selectively disrupt 10 MIC (10 g/mL) bithionol or 0.1% DMSO (control) for 2 h. (Scale bars, 500 nm.) (C and D) Viability of MRSA MW2 stationary-phase (C) or biofilm (D) bacterial compared with mammalian membrane lipid bilayers. persister cells treated with 100× MIC of the conventional antibiotics vanco- Moreover, by studying a panel of bithionol and nTZDpa analogs, mycin (Van), gentamicin (Gm), ciprofloxacin (Cipro), daptomycin (Dap), line- we found a strong correlation between the potency of anti-MRSA zolid (Lin), or the indicated concentrations of bithionol (BT) for 4 h (C)and persister activity and the ability to increase membrane fluidity. 24 h (D), respectively. (E) The viability of S. aureus VRS1 stationary-phase persisters treated with 100× MIC (200 μg/mL) linezolid (Lin), 100× MIC (100 μg/mL) Results daptomycin (dap), or the indicated concentrations of bithionol (BT) as a Bithionol Shows Bactericidal Activity Against both Antibiotic-Resistant function of time. The data points on the x-axis are below the level of de- × 2 × 2 S. aureus and S. aureus Persister Cells. Bithionol (Fig. 1A) was pre- tection (2 10 CFU/mL, or 2 10 CFU per membrane). Individual data points = viously described as an antimicrobial agent with a minimal inhibitory (n 3 biologically independent samples) are shown; error bars represent means ± SD. concentration (MIC) of ∼8to15μg/mL against Gram-positive bacteria, including S. aureus (15). We confirmed that bithionol exhibits antimicrobial activity against a variety of Gram-positive pathogens, within 2 h and 24 h, respectively (Fig. 1 C and D). In contrast, including vancomycin-resistant S. aureus (VRSA) and vancomycin- MW2 persisters displayed a high level of tolerance to 100× MIC and daptomycin-resistant enterococcal strains. In our hands, however, of several conventional antibiotics, including daptomycin and bithionol was more potent than previously reported, with MICs of μ linezolid, which were tested for activity against planktonic per- 0.5 to 2 g/mL, comparable to daptomycin (SI Appendix,Table sisters. Similarly, bithionol killed vancomycin-resistant S. aureus S1). We also confirmed that bithionol exhibits relatively weak (VRSA) strain VRS1 persisters, whereas linezolid and daptomycin antimicrobial activity against Gram-negative pathogens (MICs of exhibited no and nominal activity, respectively (Fig. 1E). 16 to >64 μg/mL; SI Appendix,TableS1). In contrast to a previous study that described bithionol as Bithionol Interacts with and Disrupts the Bacterial Mimetic Lipid bacteriostatic (16), we found that bithionol eradicated ∼107 CFU/mL Bilayer. We performed all-atom molecular dynamics (MD) simu- exponential-phase S. aureus strain MW2 within 3 h at 10 μg/mL lations of bithionol interacting with simulated bacterial mem- (10× MIC; SI Appendix,Fig.S1A). The rate of killing was branes to elucidate a potential mechanism of action. The topology comparable to daptomycin (at 10× MIC) and significantly faster than the killing kinetics of vancomycin (at 10× MIC; SI Appendix, and parameters of bithionol for the GROMOS54a7 force field Fig. S1A). Consistent with its bactericidal activity, 10 μg/mL (17) were generated by Automated Topology Builder (18). We bithionol caused a time-dependent decrease in optical density of used the previously established model of 1,2-dioleoyl-sn-glycero-3- S. aureus cells comparable to the antiseptic detergent benzyldi- phosphocholine (DOPC)/1,2-dioleoyl-sn-glycero-3-phospho- ′ methylhexadecylammonium chloride (16-BAC; SI Appendix,Fig. (1 -rac-glycerol) (DOPG) at a 7:3 ratio (19) to simulate negatively S1B), indicating that bithionol has bacteriolytic activity. Indeed, charged S. aureus membranes (SI Appendix, Materials and Meth- transmission electron micrographs (TEMs) showed that 10× MIC ods). The MD modeling showed that bithionol is initially recruited bithionol disrupts MRSA membranes, causing the intracellular to the membrane surface by the binding of polar hydroxyl and formation of mesosome-like structures, abnormal cell divisions, and chlorine groups to hydrophilic lipid heads (Fig. 2A and Movie S1). cell lysis (Fig. 1B), which we previously observed when S. aureus After several hundred nanoseconds of sustained attachment, cells were treated with membrane-active antimicrobials (12). bithionol penetrates into the membrane interior, maximizing in- We also found that bithionol killed stationary-phase MRSA teractions between nonpolar benzene rings and hydrophobic lipid MW2 planktonic and biofilm persisters in a dose-dependent tails (Movie S1). After penetration, bithionol embeds in the outer manner and completely eradicated them at 32 μg/mL (32× MIC) leaflet of the lipid bilayer (Fig. 2A and Movie S1). The insertion of

16530 | www.pnas.org/cgi/doi/10.1073/pnas.1904700116 Kim et al. Downloaded by guest on September 29, 2021 ABindicating that bithionol interacts with and disrupts a bacterial mimetic lipid bilayer. 30 7DOPC/3DOPG 7DOPC/3DOPG 7POPC/3cholesterol 7POPC/3chol 0 ns 314 ns 0 ns 290 ns 117 ns 500 ns 168 ns 500 ns 15 Bithionol Induces Rapid Membrane Permeabilization and an Increase T) B in Membrane Fluidity. (k 0 By using a SYTOX Green permeability assay, we found that, in contrast to daptomycin, bithionol induced dose- -15 1234 dependent membrane permeability in both exponential-phase

Potential of mean force Distance from bilayer center MRSA MW2 cells and stationary-phase MRSA MW2 persisters (nm) CDBithionol DMSO 1 μg/ml 10 μg/ml 0.1% S. aureus MW2 HKC-8 (Fig. 2D and SI Appendix,Fig.S1C). The fluorescence intensity 1000 1000 peaked at 4 μg/mL and then decreased at higher concentrations of 800 800 bithionol (Fig. 2D), most likely as a consequence of nucleic acids 600 600 released form lysed cells (20), consistent with the observation that 400 400 10 μg/mL bithionol lyses MRSA persisters (Fig. 1B). The dose- 200 200 dependent effects of bithionol on membrane permeability corre- Fluorescence intensity Fluorescence intensity 7DOPC/3DOPC 0 0 lated with its dose-dependent killing kinetics (Figs. 1C and 2D and 0204060 0204060 Time (min) Time (min) SI Appendix,Fig.S1C), suggesting that the bactericidal activity of 32 μg/mL 16 μg/mL 8 μg/mL 4 μg/mL 64 μg/mL bithionol results from its membrane-disrupting activity. 2 μg/mL 1 μg/mL 0 μg/mL 32 μg/mL Bithionol DMSO 16 μg/mL 1 μg/ml 10 μg/ml 0.1% Insertion of particular membrane disrupting compounds into E 0.40 0 μg/mL lipid bilayers is known to cause dramatic changes in membrane 0.35 fluidity that disrupts the normal liquid-crystalline phase of the 0.30 ** ** – *** membrane (21 23). This results in passive permeabilization, loss 0.25

Laurdan GP *** of membrane protein functions, leakage of cellular components, 0.20 Increase in fluidity and bacterial death (24). We therefore tested whether bithionol 0.15

B.A. alters MRSA MW2 membrane fluidity by utilizing the membrane μg/ml μg/ml μg/ml μg/ml μg/ml μg/ml 7POPC/3cholesterol 1 2 4 8 16 32 fluidity-sensitive dye Laurdan (25), which exhibits a fluorescence Non-treated emission wavelength shift dependent on the amount of adjacent Fig. 2. Bithionol selectively disrupts bacterial lipid bilayers. (A) Representa- water molecules (25). Because water molecule penetration into

tive configurations of MD simulations of bithionol from left to right: onset, lipid bilayers is in turn determined by lipid packing density and MICROBIOLOGY membrane attachment, membrane penetration, and equilibrium interacting lipid bilayer fluidity, bacterial membrane fluidity can be quanti- with 7DOPC/3DOPG or 7POPC/3cholesterol lipid bilayers. Bithionol and sodium fied by using the Laurdan generalized polarization (GP) metric: ions are depicted as large spheres, phospholipids are represented as chains, GP = (I440 − I490)/(I440 + I490), ranging from −1 (most fluid and and bonds in cholesterols are highlighted by thickened tubes. The atoms in disordered) to +1 (most rigid and ordered) (25). As shown in bithionol, phospholipids, cholesterol, and sodium ions are colored as follows: Fig. 2E, bithionol at concentrations greater than 8 μg/mL in- hydrogen, white; oxygen, red; nitrogen, blue; sulfur, yellow; chlorine, green; carbon, cyan; phosphorus, orange; and sodium, purple. For clarity, water duced a significant decrease in Laurdan GP in S. aureus MW2, molecules are not shown. (B) The free-energy profiles of bithionol penetrating similar to 50 mM benzyl alcohol (BA), a known membrane flu- into the indicated lipid bilayers as a function of the center-of-mass (COM) idizer. These data are consistent with our observation that distance to the bilayer. The dot-dashed blue and red lines mark the surface of bithionol concentrations greater than 16 μg/mL exhibited sig- bacterial and mammalian membranes, respectively, averaged from the COM nificant killing activity against MRSA MW2 persisters (Fig. 1C). locations of phosphate groups in the lipids of the outer leaflets. Error bars represent means ± SD from 3 independent simulations. (C) GUVs consisting of Bithionol Does Not Penetrate Mammalian Membranes. We used MD DOPC/DOPG (7:3) or POPC/cholesterol (7:3) labeled with 0.005% Liss Rhod PE simulations to determine whether bithionol specifically penetrates were treated with the indicated concentrations of bithionol or 0.1% DMSO bacterial compared with mammalian membranes. Mammalian (control) and were monitored over time by using fluorescence microscopy. membranes are comprised of phospholipids, sphingolipids, and (Scale bars, 10 μm.) (D) Uptake of SYTOX Green (Ex = 485 nm, Em = 525 nm) by MRSA MW2 persister cells or human renal proximal cells (HKC-8) treated cholesterol (26), but can be modeled as a simplified bilayer with the indicated concentrations of bithionol. Results are shown as means; composed of the zwitterionic lipid 1,2-palmitoyl-oleoyl-sn-glycero- n = 3 biologically independent samples. Error bars not shown for clarity. (E) S. 3-phosphocholine (POPC) mixed with cholesterol ranging from 20 aureus MW2 membrane fluidity treated with the indicated concentrations of to 50 mol% (27, 28). At a POPC/cholesterol ratio of 7:3, MD bithionol for 1 h was evaluated based on Laurdan generalized polarization simulations using the GROMOS54a7 force field showed that = − + (Laurdan GP). Laurdan GP (I440 I490)/(I440 I490), where I440 and I490 are the bithionol fails to penetrate the simulated mammalian bilayer (Fig. emission intensities at 440 and 490 nm, respectively, when excited at 350 nm. 2 A and B, SI Appendix,Fig.S2B,andMovie S2). Moreover, the The membrane fluidizer benzyl alcohol (50 mM) was used as a positive con- energy barrier and transfer energy for bithionol increases with trol. Individual data points (n = 3 biologically independent samples) are shown; error bars represent means ± SD. Statistical differences between increasing percentages of cholesterol (SI Appendix,Fig.S3A). control and antibiotic treatment groups were analyzed by 1-way ANOVA and These results are consistent with our observation that the presence post hoc Dunnett test (**P = 0.01 and ***P < 0.0001.). of cholesterol results in a more ordered alignment of the membrane lipids (SI Appendix, Fig. S3D), as well as with configura- tional and thermodynamic analyses (29, 30), which demonstrate even a single bithionol molecule causes a substantial local increase that cholesterol condenses the hydrophobic region of the mem- in lipid bilayer disorder (SI Appendix,Fig.S2A and C). These MD brane (SI Appendix,Fig.S3B) and decreases membrane fluidity simulations demonstrate that the polarity of branch groups and (SI Appendix,Fig.S3C). An independent MD simulation using the the hydrophobicity of core rings play important roles in membrane CHARMM force field (31) also showed that bithionol preferen- attachment and penetration. tially penetrates bacterial-mimetic compared with mammalian- We further investigated the effects of bithionol on lipid bila- mimetic lipid bilayers (SI Appendix,Fig.S4). yers by using biomembrane-mimicking giant unilamellar vesicles Consistent with the MD simulations, bithionol did not cause (GUVs) consisting of a single lipid bilayer with a diameter of 10 observable effects on GUVs formed with 7POPC/3cholesterol to 100 μm. GUVs were constructed by using DOPC/DOPG lipids (Fig. 2C and Movies S6–S8). Moreover, bithionol exhibited rel- at a 7:3 ratio as in the MD simulations. Lipid aggregates formed atively little hemolytic activity against human erythrocytes, with on the surfaces of the GUVs exposed to 1 μg/mL bithionol, and, an HC50 >64 μg/mL (SI Appendix, Fig. S5A), and did not induce at 10 μg/mL, the GUVs burst (Fig. 2C and Movies S3–S5), SYTOX Green membrane permeabilization of HKC-8 human

Kim et al. PNAS | August 13, 2019 | vol. 116 | no. 33 | 16531 Downloaded by guest on September 29, 2021 renal proximal cells up to 64 μg/mL (Fig. 2D and SI Appendix, the smaller chlorine and fluorine atoms, although the bromines may Fig. S5 B and C). be slightly disadvantageous for initial binding and penetration.

Structure–Activity Relationships. The MD simulations described Bithionol in Combination with Gentamicin Shows Efficacy in a Mouse here earlier predicted that 2 key elements for the membrane ac- Deep-Seated MRSA Infection Model. We previously showed that the tivity of bithionol are the initial binding to the membrane surface aminoglycoside antibiotic gentamicin, which is used to treat se- using phenolic hydroxyl groups and membrane perturbation in- vere chronic MRSA infections despite nephrotoxicity issues (33), duced by chlorinated benzene (Fig. 2A and Movie S1). To further exhibits significant synergism with synthetic retinoids (6). Although μ test this proposed mechanism and to obtain additional insights we found that relatively low concentrations of bithionol (8 g/mL) into the effects of functional groups on the potency of bithionol, do not lead to a significant decrease in the viability of MRSA per- – sisters (Fig. 1 C and D), 8 μg/mL bithionol combined with as little as we conducted structure activity relationship (SAR) studies by μ using a commercially available bithionol analog, Bitin-S, as well as 2 g/mL gentamicin completely eradicated MRSA MW2 stationary- phase persister cells (Fig. 3A). Similarly, 8 μg/mL bithionol + 16 μg/mL with 7 newly synthesized analogs (Table 1 and SI Appendix,Figs. gentamicin eradicated biofilm persisters (Fig. 3B). S6 and S7). The effect of binding affinity on antimicrobial and We further tested the efficacy of bithionol in combination with membrane activity was tested by using Bitin-S, a sulfoxide de- gentamicin in an MRSA mouse thigh infection model, which rivative of bithionol (32), and the bithionol methoxy analog BT- mimics human deep-seated chronic infections (5). Consistent OMe (Table 1). The polar sulfinyl group of Bitin-S provides ad- with a previous study (5), vancomycin, gentamicin, or a combi- ditional hydrophilic interactions with lipid heads. Bitin-S exhibited nation of the 2 did not significantly reduce MRSA CFUs in the decreased antimicrobial activity (MIC 8 μg/mL) and reduced mouse thigh (Fig. 3C), suggesting that the infecting bacterial membrane activity (Table 1 and SI Appendix, Fig. S6). Reducing cells are persisters. Although 30 mg/kg bithionol alone showed polarity by substituting methoxy groups for the 2 hydroxyl groups no effect on the viability of MRSA persisters, 30 mg/kg bithionol (BT-OMe; Table 1 and SI Appendix, Fig. S6)resultedincomplete combined with 30 mg/kg gentamicin killed ∼90% of the MRSA nullification of both antimicrobial and membrane activity (Table 1 persister cells (P < 0.001; Fig. 3C). We evaluated the hepatic and and SI Appendix,Fig.S6), indicating that the phenolic hydroxyl renal toxicity of bithionol in the mice in the experiments de- groups are critical for antimicrobial activity. scribed in Fig. 3C by measuring serum levels of alanine amino- To test whether the size and polarization of the inserted mole- transferase (ALT) and blood urea nitrogen (BUN) (SI Appendix, cules can affect the extent of membrane perturbation (12), we Fig. S8). Although the combination of vancomycin and genta- substituted alternative halogens for chlorine. Replacement of chlo- micin significantly increase BUN levels (P < 0.01), the combi- rine with fluorine resulted in reduced membrane and antimicrobial nation of bithionol and gentamicin increased neither. activities as seen with compounds BT-oF, BT-pF, and BT-opF, whereas substitution with bromine showed similar membrane and antimi- The Killing of MRSA Persisters Is Correlated with Increased Membrane Fluidity. The SAR studies reported here show that some com- crobial activities as bithionol as demonstrated by BT-oBr, BT-pBr, pounds that permeabilize MRSA persister cell membranes to op SI Appendix and BT- Br (Table 1 and ,Fig.S7). It is likely that the SYTOX Green, such as Bitin-S, BT-oF, or BT-pF, do not kill polarity of the C-F bond in the fluoro derivatives may increase the them (Table 1 and SI Appendix, Figs. S6 and S7). We reported a hydrophilicity of the aryl rings, thus decreasing membrane perme- similar result in a previous SAR study on nTZDpa (12), where ability. The bromine derivatives showed increased energy barrier we observed that nTZDpa-analogs 6 and 11 (Fig. 4 B and C) values of 2 to 3 kBT (SI Appendix,TableS2), indicating that the induced rapid SYTOX Green permeabilization, but did not af- larger bromine atoms may cause more membrane perturbation than fect the viability of MRSA persisters (12). Interestingly, the bithionol SAR series showed that only – compounds leading to a significant increase in membrane fluidity Table 1. Structure activity relationships for antibiotic activity killed MRSA persisters (Fig. 4A, Table 1, and SI Appendix,Figs.S6 and membrane activity and S7). To test the hypothesis that there is a general correlation OH OH R1 R1 between an increase in membrane fluidity and antipersister potency, Cl S Cl 2 2 R R R we measured the effect of nTZDpa and 11 nTZDpa derivatives (Fig. 4 B and C) (12) on MRSA membrane fluidity. Consistent with Cl Cl R3 R3 bithionol the bithionol analogs, only nTZDpa analogs that exhibit antipersister potency showed increased membrane fluidity at 32 μg/mL. OH O AlCl (1 equiv) HO O HO OH OH 3 Zn dust (10 equiv) 2 S 2 S 2 2 S 2 R Cl Cl DCM R R AcOH R R Furthermore, we observed substantial correlation between the + (1 equiv) 22 ºC, 8 h 100 ºC, 4 h concentration of compound required for persister killing and the R3 R3 R3 R3 R3 amount of induced membrane fluidity as measured by Laurdan μ 2 Cmpd R R1 R2 R3 MIC PKC MP MFI GP for all 21 tested compounds at 32 g/mL (R 0.64, P of < Bithionol S OH Cl Cl 1 32 Yes Yes slope 0.0001; SI Appendix, Fig. S9). Consistent with these data, Bitin-S SO OH Cl Cl 8 >64 Yes No a recent study reported that the antibiotic agent rhodomyrtone BT-OMe S OMe Cl Cl >64 >64 No No causes increased membrane fluidity in Bacillus subtilis and kills BT-oFSOHF Cl 2 >64 Yes No its persister cells (23). BT-oBr S OH Br Cl 1 32 Yes Yes The data in this section show that the killing of MRSA per- BT-pFSOHClF 2>64YesNo sisters is achieved only when membrane damage by membrane- BT-pBr S OH Cl Br 1 32 Yes Yes active agents is sufficiently severe to show increased membrane BT-opFS OH F F 8 >64 No No fluidity, potentially making membrane fluidity a biophysical in- BT-opBr S OH Br Br 1 32 Yes Yes dicator to identify and measure the potency of antipersister antimicrobial agents. MIC, minimum inhibitory concentration (in micrograms per milliliter); PKC, persister killing concentration (in micrograms per milliliter) required to kill Discussion 5 × 107 CFU/mL MRSA persister cells below the limit of detection; MP, membrane permeabilization (determined based on SYTOX Green fluorescence in- Membrane-active agents have attractive properties as antimi- tensity); MFI, membrane fluidity increase (determined based on Laurdan GP crobial agents, including fast killing rates, antipersister potency, measurement). synergism with other antibiotic agents, and a low probability of

16532 | www.pnas.org/cgi/doi/10.1073/pnas.1904700116 Kim et al. Downloaded by guest on September 29, 2021 AB C Appendix,Fig.S7). Replacement of the chlorines with bromines 8 9 10 decreased initial binding to the lipid bilayer (SI Appendix,Table 8 7 S2), but increased destabilization of the membrane compared with 7 6 9 6 5 *** chlorine and fluorine after penetration. Collectively, the antimicrobial 5 8 4 4 activity of bithionol can be modulated by binding affinity to lipid head

Log (CFU/mL) 3 3 Log (CFU/g) 7 groups, penetration depth, and molecule size. 01234 Time (h) Log (CFU/membrane) 6 In contrast to bithionol and other membrane-active compounds Van Gm BT 8 μg/mL BT + 2 μg/mL Gm μg/mLμg/mL Gmμg/mL Gm Gm we have studied, daptomycin does not induce rapid SYTOX 8 μg/mL BT + 4 μg/mL Gm Non-treated8 μg/mL BT Control 8 μg/mL BT + 8 μg/mL Gm 16 μg/mL Gm Van+Gm BT + Gm Green membrane permeabilization or kill MRSA persisters 8 μg/mL BT + 16 μg/mL Gm 8 μg/mL BT (Fig. 1C and SI Appendix, Fig. S1C). Importantly, we discovered 16 μg/mL Gm that bithionol, nTZDpa, and their analogs that exhibit anti- Non-treated 8 μg/mL8 μg/mL BT8 μ +g/mL8 BT 2μ g/mL + BT 4 +BT 8 + 16 μg/mL Gm MRSA persister potency induce SYTOX Green membrane per- Fig. 3. Bithionol shows synergism with gentamicin against MRSA persisters meabilization and an increase in membrane fluidity (Fig. 4, Table 1, in vitro and in vivo. (A) Stationary-phase or (B) biofilm MRSA MW2 persisters and SI Appendix,Figs.S6andS7). These data are consistent with were treated with the indicated concentrations of bithionol (BT) combined previous reports that show that insertion of compounds into with gentamicin (Gm). Colony forming units (CFUs) were measured by serial membrane bilayers can increase membrane disorder and fluidity, dilution and plating on TSA plates. The data points on the x-axis are below the level of detection (2 × 102 CFU/mL, or 2 × 102 CFU per membrane). In- which subsequently causes passive membrane permeabilization (21, dividual data points (n = 3 biologically independent samples) are shown; 22). Interestingly, we found that the initiation of SYTOX Green error bars represent means ± SD. (C) Ten infected mice per group (n = 10 membrane permeabilization occurs at a lower concentration of biologically independent animals) were treated with control (5% Killophor + the membrane-active compounds than is required to increase 5% ethanol, i.p.), vancomycin (30 mg/kg, i.p.), gentamicin (30 mg/kg, s.c.), membrane fluidity (Fig. 2 D and E and SI Appendix, Figs. S6 and bithionol (30 mg/kg, i.p.), or vancomycin (30 mg/kg, i.p.) or bithionol (30 mg/kg, S7). Furthermore, some membrane-active agents, such as bitin-S, i.p.) combined with gentamicin (30 mg/kg, s.c.) every 12 h for 3 d at 24 h bithionol fluorine analogs, and nTZDpa-analogs 6 and 11, that postinfection. At 12 h after the last treatment, mice were euthanized. induce SYTOX Green membrane permeabilization do not cause Their thighs were excised and homogenized. CFUs from each mouse thigh are plotted as individual points, and error bars represent the SD in each a change in membrane fluidity and do not kill persisters (Fig. 4 B experimental group. Statistical differences between control and antibiotic and C and SI Appendix, Figs. S6 and S7). Bacterial membranes treatment groups were analyzed by 1-way ANOVA and post hoc Tukey test consist of lipid rafts organized into microdomains having different (***P < 0.001). lipid compositions (35). Depending on lipid compositions, one MICROBIOLOGY domain may be less ordered and more fluid, whereas another domain may be more ordered and rigid (35). Because ordered and resistance selection (7). Unfortunately, most of these agents also rigid domains show more resistance to membrane active agents indiscriminately disrupt mammalian membranes. However, evolu- (35) up to a certain threshold concentration, only less ordered tion has taken advantage of differences in bacterial and eukaryotic domains would be affected by membrane-active compounds, thus membranes, as reflected in the production of cationic antimicro- making them SYTOX Green-permeable. However, this type of bial peptides by animals and plants that specifically target bacterial cells. Gram-positive bacterial membrane lipid bilayers contain ∼25% anionic phospholipids, whereas mammalian membranes are composed of zwitterionic (neutral) phospholipids and 20 to 50% A C 4 0.4 R2 cholesterol (27, 34). Cationic antimicrobial peptides bind prefer- 3 X entially to negatively charged bacterial membranes, and cholesterol 0.3 *** *** Cl O in animal membranes creates a condensing effect that confers *** *** N 0.2 OH

membrane rigidity and prevents the penetration of antimicrobial Laurdan GP 2 peptides (34). 3 0.1 4 1 We found that bithionol penetration into negatively charged F Br F Br F Br R o o p p B.A. op op BT- BT- 1 2 Bitin-S BT- BT- BT- Cmpd R R X bacterial mimetic lipid bilayers (7DOPC/3DOPG) is energeti- Bithionol BT-OMe BT- cally favorable, whereas bithionol penetration into cholesterol- Non-treated nTZDpa 4-Cl HS B >64 >64 4 4-Cl 4-Cl S 0.4 ns ns rich mammalian mimetic lipid bilayers (7POPC/3cholesterol) is 16 32 64 5 4-Cl 4-tBu S 32 32 * 16 64 ** 8 *** *** 6 4-Cl HO energetically unfavorable (Fig. 2 A and B). Further, as the pro- 0.3 *** *** ** *** portion of cholesterol increases from 0 to 30%, the penetration *** 8 10 3,4-Cl H O 16 *** 11 4-Cl 4-Cl O of bithionol becomes increasingly unfavorable (SI Appendix, Fig. 0.2 *** 12 4-Cl 3,4-Cl O Laurdan GP S3A), indicating that cholesterol plays a key role in bithionol’s 13 4-Cl 4-Br O 0.1 14 4-Cl 4-I O membrane selectivity. 4 5 6 10 11 12 13 14 S21 2,4-Cl HO MD modeling suggests that the 2 polarized phenolic hydroxyl B.A. S21 S24 S26 S24 4-Cl 4-tBu O nTZDpa S26 4-Cl 4-CF O groups of bithionol play a major role in the presumed initial binding Non-treated 3 to phospholipid headgroups via hydrogen bonding. SAR studies showed that a methoxy analog of bithionol nullified bioactivity, Fig. 4. Relationship between antipersister activity and alteration in membrane fluidity. Membrane fluidity of (A) bithionol and its analogs or (B) supporting this proposed mechanism of interaction (Table 1 and SI nTZDpa and its analogs at 32 μg/mL was evaluated by Laurdan GP. The Appendix, Table S2 and Fig. S6). In addition to providing presumed membrane fluidizer benzyl alcohol (B.A.; 50 mM) was used as positive con- polar interactions with lipid headgroups, the bithionol chlorine trol. (B) The blue-colored numbers above each bar indicate persister killing moieties also appear to play a key role in lipid bilayer perturbation. concentration (PKC, in micrograms per milliliter) required to kill 5 × 107 CFU/mL Briefly, after attachment, the binding affinity of bithionol is domi- MRSA persister below the limit of detection (2 × 102 CFU/mL). (A and B) nated by hydrophobic interactions between its aromatic rings and Individual data points (n = 3 biologically independent experiments) are ± the hydrophobic tails of the membrane lipids, which drives the shown; error bars represent means SD. Statistical differences between control and antibiotic treatment groups were analyzed by 1-way ANOVA penetration of the chlorinated benzene into the outer leaflet of the and post hoc Dunnett test; ns, no significance (P > 0.05), *P = 0.05, **P = lipid bilayer, causing lipid bilayer perturbation (SI Appendix,Fig.S2 0.01, and ***P < 0.001. Individual data points (n = 3 biologically in- A and C). Accordingly, the replacement of chlorine with fluorine dependent experiments) are shown; error bars represent means ± SD. (C) resulted in a decrease in antimicrobial potency (Table 1 and SI The structures of nTZDpa and its analogs.

Kim et al. PNAS | August 13, 2019 | vol. 116 | no. 33 | 16533 Downloaded by guest on September 29, 2021 localized membrane damage may not be sufficient to cause an mammalian membranes and support the conclusion that overall increase in membrane fluidity. Over the threshold concen- membrane-active antimicrobial agents have promising potential tration, most membrane domains would be disrupted, and, sub- to be used for treating chronic infections caused by bacterial sequently, overall membrane fluidity would increase. It is also persisters. possible that some compounds may attack only less ordered and more fluid areas of membranes, which causes SYTOX Green Materials and Methods membrane permeability, but not an overall increase in membrane SI Appendix, Materials and Methods describes in detail the materials and fluidity. In any case, the killing of MRSA persisters is apparently procedures used in this study, including bacterial strains and growth con- achieved only when the bacterial membrane is sufficiently dam- ditions, antimicrobial agents and chemicals, transmission electron micros- aged to show increased membrane fluidity as detected by copy, a minimal inhibitory concentration (MIC) assay, a killing kinetics assay, Laurdan GP. a persister cell killing assay, a biofilm persister killing assay, a bacterial In conclusion, we report that the clinically approved anthel- membrane permeability assay, a mammalian membrane permeability assay, a membrane fluidity assay, a human blood hemolysis assay, a giant uni- mintic bithionol is an effective antimicrobial agent against both lamellar vesicles (GUVs) assay, all-atom molecular dynamics (MD) simula- multidrug-resistant and -persistent Gram-positive pathogens. tions, a deep-seated mouse thigh infection model, and general procedures Bithionol kills Gram-positive (but not Gram-negative) bacterial for the synthesis of bithionol analogs. cells by disrupting lipid bilayers while maintaining high selectivity for bacterial compared with mammalian membranes, a conse- ACKNOWLEDGMENTS. This study was supported by National Institutes of quence of the presence of cholesterol in mammalian membranes. Health Grants P01 AI083214 (to F.M.A. and E.M.), P20 GM121344 (to Further, bithionol in combination with gentamicin effectively B.B.F.), and R35 GM119426 (to W.M.W.) and by National Science Foundation Grant CMMI-1562904 (to H.G.). We thank the Institute of Chemistry and Cell eradicates S. aureus persisters in vitro and significantly reduces Biology (ICCB)–Longwood at Harvard Medical School for providing the bacterial burden in a mouse model of chronic deep-seated chemical libraries used in this study. We thank Dr. L. Rice for generously MRSA infection. We also demonstrate that increased mem- providing the E. faecium strains. The simulations reported were performed brane fluidity is a biophysical indicator to identify potent anti- on resources provided by the Extreme Science and Engineering Discovery Environment (XSEDE) through Grant MSS090046 and the Center for Compu- persister compounds. Our studies provide further understanding of tation and Visualization (CCV) at Brown University. The NMR instruments the molecular mechanisms by which membrane-active small used in this work were supported by National Science Foundation Grant molecules selectively disrupt Gram-positive bacterial over CHE-1531620.

1. A. S. Lee et al., Methicillin-resistant Staphylococcus aureus. Nat. Rev. Dis. Primers 4, 19. Y.-F. Chen, T.-L. Sun, Y. Sun, H. W. Huang, Interaction of daptomycin with lipid bi- 18033 (2018). layers: A lipid extracting effect. Biochemistry 53, 5384–5392 (2014). 2. K. Lewis, Persister cells. Annu. Rev. Microbiol. 64, 357–372 (2010). 20. B. L. Roth, M. Poot, S. T. Yue, P. J. Millard, Bacterial viability and antibiotic susceptibility 3. B. P. Conlon et al., Persister formation in Staphylococcus aureus is associated with ATP testing with SYTOX green nucleic acid stain. Appl. Environ. Microbiol. 63,2421–2431 (1997). depletion. Nat. Microbiol. 1, 16051 (2016). 21. M. L. Cartron et al., Bactericidal activity of the human skin fatty acid cis-6- 4. K. R. Allison, M. P. Brynildsen, J. J. Collins, Metabolite-enabled eradication of bacterial hexadecanoic acid on Staphylococcus aureus. Antimicrob. Agents Chemother. 58, persisters by aminoglycosides. Nature 473, 216–220 (2011). 3599–3609 (2014). 5. B. P. Conlon et al., Activated ClpP kills persisters and eradicates a chronic biofilm in- 22. Y. Wu et al., Antibacterial activity and membrane-disruptive mechanism of 3-p-trans- fection. Nature 503, 365–370 (2013). coumaroyl-2-hydroxyquinic acid, a novel phenolic compound from pine needles of 6. W. Kim et al., A new class of synthetic retinoid antibiotics effective against bacterial Cedrus deodara, against Staphylococcus aureus. Molecules 21, E1084 (2016). persisters. Nature 556, 103–107 (2018). 23. D. Saeloh et al., The novel antibiotic rhodomyrtone traps membrane proteins in 7. J. G. Hurdle, A. J. O’Neill, I. Chopra, R. E. Lee, Targeting bacterial membrane function: vesicles with increased fluidity. PLoS Pathog. 14, e1006876 (2018). An underexploited mechanism for treating persistent infections. Nat. Rev. Microbiol. 24. R. M. Epand, C. Walker, R. F. Epand, N. A. Magarvey, Molecular mechanisms of 9,62–75 (2011). membrane targeting antibiotics. Biochim. Biophys. Acta 1858, 980–987 (2016). 8. W. R. Miller, A. S. Bayer, C. A. Arias, Mechanism of action and resistance to dapto- 25. T. Parasassi, E. Gratton, Membrane lipid domains and dynamics as detected by mycin in Staphylococcus aureus and Enterococci. Cold Spring Harb. Perspect. Med. 6, Laurdan fluorescence. J. Fluoresc. 5,59–69 (1995). a026997 (2016). 26. T. Harayama, H. Riezman, Understanding the diversity of membrane lipid composi- 9. B. Sharma, A. V. Brown, N. E. Matluck, L. T. Hu, K. Lewis, Borrelia burgdorferi,the tion. Nat. Rev. Mol. Cell Biol. 19, 281–296 (2018). causative agent of lyme disease, forms drug-tolerant persister cells. Antimicrob. 27. M. Deleu, J.-M. Crowet, M. N. Nasir, L. Lins, Complementary biophysical tools to in- Agents Chemother. 59, 4616–4624 (2015). vestigate lipid specificity in the interaction between bioactive molecules and the 10. G. S. Sahukhal, S. Pandey, M. O. Elasri, msaABCR operon is involved in persister cell plasma membrane: A review. Biochim. Biophys. Acta 1838, 3171–3190 (2014). formation in Staphylococcus aureus. BMC Microbiol. 17, 218 (2017). 28. C. Hong, D. P. Tieleman, Y. Wang, Microsecond molecular dynamics simulations of 11. W. Kim et al., Identification of an antimicrobial agent effective against methicillin- lipid mixing. Langmuir 30, 11993–12001 (2014). resistant Staphylococcus aureus persisters using a fluorescence-based screening 29. F. de Meyer, B. Smit, Effect of cholesterol on the structure of a phospholipid bilayer. strategy. PLoS One 10, e0127640 (2015). Proc. Natl. Acad. Sci. U.S.A. 106, 3654–3658 (2009). 12. W. Kim et al., Discovery and optimization of nTZDpa as an antibiotic effective against 30. T. Róg, M. Pasenkiewicz-Gierula, I. Vattulainen, M. Karttunen, Ordering effects of bacterial persisters. ACS Infect. Dis. 4, 1540–1545 (2018). cholesterol and its analogues. Biochim. Biophys. Acta 1788,97–121 (2009). 13. W. Kim et al., NH125 kills methicillin-resistant Staphylococcus aureus persisters by 31. J. B. Klauda et al., Update of the CHARMM all-atom additive force field for lipids: lipid bilayer disruption. Future Med. Chem. 8, 257–269 (2016). Validation on six lipid types. J. Phys. Chem. B 114, 7830–7843 (2010). 14. J. Keiser, J. Utzinger, Food-borne trematodiasis: Current chemotherapy and advances 32. T. Meshi, M. Yoshikawa, Y. Sato, Metabolic fate of bis(3,5-dichloro-2-hydroxyphenyl)- with artemisinins and synthetic trioxolanes. Trends Parasitol. 23, 555–562 (2007). sulfoxide (bithionol sulfoxide). Biochem. Pharmacol. 19, 1351–1361 (1970). 15. F. S. Barr, G. F. Collins, L. G. Wyatt, Potentiation of the antimicrobial activity of bi- 33. C. Liu et al.; Infectious Diseases Society of America, Clinical practice guidelines by the thionol. J. Pharm. Sci. 54, 801–802 (1965). infectious diseases society of America for the treatment of methicillin-resistant 16. A. L. Erlandson, Jr, C. A. Lawrence, Inactivating medium for hexachlorophene (G11) types Staphylococcus aureus infections in adults and children. Clin. Infect. Dis. 52, e18– of compounds and some substituted phenolic disinfectants. Science 118,274–276 (1953). e55 (2011). 17. N. Schmid et al., Definition and testing of the GROMOS force-field versions 54A7 and 34. J. R. Brender, A. J. McHenry, A. Ramamoorthy, Does cholesterol play a role in the 54B7. Eur. Biophys. J. 40, 843–856 (2011). bacterial selectivity of antimicrobial peptides? Front. Immunol. 3, 195 (2012). 18. A. K. Malde et al., An automated force field topology builder (ATB) and repository: 35. M. Bramkamp, D. López, Exploring the existence of lipid rafts in bacteria. Microbiol. Version 1.0. J. Chem. Theory Comput. 7, 4026–4037 (2011). Mol. Biol. Rev. 79,81–100 (2015).

16534 | www.pnas.org/cgi/doi/10.1073/pnas.1904700116 Kim et al. Downloaded by guest on September 29, 2021