Advertisement

A Rationally Designed Multifunctional Antibiotic for the Treatment of Drug-Resistant Acne

Open ArchivePublished:February 04, 2018DOI:https://doi.org/10.1016/j.jid.2017.11.041
      Acne is a multifactorial skin disease, underpinned by colonization of Propionibacterium acnes and inflammation. The emergence of resistant P. acnes strains has affected the current acne treatment algorithm. This setback served as an impetus for rationally designing a library of next-generation antibiotics that exhibit a bactericidal effect on resistant P. acnes and exert an immunomodulatory function to reduce inflammation. In silico screening showed that one of the molecules, VCD-004, exhibits improved mode of binding to bacterial DNA gyrase. VCD-004 shows high potency against clinical isolates of resistant P. acnes and excellent efficacy in vivo. Furthermore, VCD-004 exhibits a superior mutant prevention index, suggesting that it impedes the development of resistance better than clindamycin. Additionally, it shows optimal skin penetration and has a potent anti-inflammatory effect via reduction of proinflammatory cytokines (IL-6) independent of its antibacterial action. VCD-004 affects P. acnes-induced nuclear accumulation of NF-κB in THP-1 cells. The in vitro viability of human keratinocytes in the presence of VCD-004 indicates a desirable therapeutic window for topical use. Such rationally designed bactericidal and immunomodulatory dual pharmacophore-based lipophilic molecule(s) can emerge as the next-generation topical therapy for acne with underlying resistant P. acnes etiology.

      Abbreviations:

      MIC (minimum inhibitory concentration), MTCC (The Microbial Type Culture Collection and Gene Bank), MPC (mutant prevention concentration), QBP (quinolone binding pocket)

      Introduction

      Acne vulgaris affects almost 9.4% of the world’s population (
      • Tan J.K.
      • Bhate K.
      A global perspective on the eidemeology of acne.
      ). Propionibacterium acnes, a commensal human skin bacterium, is a prime pathogenic factor in the development of acne (
      • Cunliffe W.J.
      • Clayden A.D.
      • Gould D.
      • Simpson N.B.
      Acne vulgaris—its aetiology and treatment. A review.
      ). Topical and oral antibiotics play pivotal role in effective acne management. Clindamycin and erythromycin are commonly applied topical antibiotics, whereas minocycline and doxycycline are administered orally to patients with moderate to severe acne (
      • Aslam I.
      • Fleischer A.
      • Feldman S.
      Emerging drugs for the treatment of acne.
      ). Although current therapy guidelines recommend judicious use of antibiotics in combination with other anti-acne drugs, like retinoids (
      • Nast A.
      • Dréno B.
      • Bettoli V.
      • Degitz K.
      • Erdmann R.
      • Finlay A.Y.
      • et al.
      European Dermatology Forum. European evidence-based (S3) guidelines for the treatment of acne.
      ,
      • Thiboutot D.
      • Gollnick H.
      • Bettoli V.
      • Dréno B.
      • Kang S.
      • Leyden J.J.
      • et al.
      Global Alliance to Improve Outcomes in Acne. New insights into the management of acne: an update from the Global Alliance to Improve Outcomes in Acne group.
      ), extensive use of currently approved antibiotics has led to generation of resistant P. acnes strains, which poses a major setback to the present treatment regimens (
      • Leyden J.J.
      • McGinley K.J.
      • Cavalieri S.
      • Webster G.F.
      • Mills O.H.
      • Kligman A.M.
      Propionibacterium acnes resistance to antibiotics in acne patients.
      ,
      • Simonart T.
      • Dramaix M.
      Treatment of acne with topical antibiotics: lessons from clinical studies.
      ). This necessitates development of drugs that can address the challenge of resistant P. acnes.
      Not all antibiotics that exhibit efficacy against P. acnes are suitable for acne therapy (
      • Sinha M.
      • Sadhasivam S.
      • Bhattacharyya A.
      • Jain S.
      • Ghosh S.
      • Arndt K.A.
      • et al.
      Antibiotic-resistant acne: getting under the skin.
      ). Acne is a multifactorial disease with a significant inflammatory component (
      • Leyden J.J.
      Acne vulgaris is a multifactorial disease.
      ). As a result, we rationalized that the logical design of antibiotics against resistant P. acnes should incorporate pharmacophore(s) that can impart bactericidal activity against resistant strains of P. acnes and further confer an immunomodulatory activity to reduce acne-induced inflammation. We developed a library of molecules, to our knowledge previously unreported, with quinolone backbone and nitro-heterocyclic motif arranged in different spatial orientations to satisfy the desired activities of a next-generation anti-acne drug. The currently used antibiotics, clindamycin and tetracyclines, are mostly bacteriostatic in nature (
      • Tan H.H.
      Topical antibacterial treatments for acne vulgaris: comparative review and guide to selection.
      ). We chose a quinolone scaffold that can interfere in DNA supercoiling in DNA gyrase-DNA complex (
      • Kampranis S.C.
      • Maxwell A.
      The DNA gyrase-quinolone complex. ATP hydrolysis and the mechanism of DNA cleavage.
      ), resulting in a bactericidal effect. Nadifloxacin, a quinolone-based drug, is approved in parts of Europe and Japan for acne treatment (
      • Jacobs M.R.
      • Appelbaum P.C.
      Nadifloxacin: a quinolone for topical treatment of skin infections and potential for systemic use of its active isomer, WCK 771.
      ). Similarly, nitro-heterocyclic molecules have been widely used as antibiotics and anti-inflammatory agents for specific indications (
      • Bannatyne R.M.
      Metronidazole, its bioactive metabolites and acne.
      ,
      • Millikan L.
      The proposed inflammatory pathophysiology of rosacea: implications for treatment.
      ,
      • Shakir L.
      • Javeed A.
      • Ashraf M.
      • Riaz A.
      Metronidazole and the immune system.
      ).
      Using in silico molecular docking, we structured an effective quinolone scaffold with optimum spatial orientation of the nitro-heterocyclic (3-[2-methyl-5-nitro-imidazolyl] propan-2-ol) motif to satisfy efficient binding at the quinolone binding pocket (QBP) of DNA gyrase-DNA complex. We report that one of these molecules, VCD-004, displays a different mode of binding at the QBP compared with fluoroquinolones. VCD-004 exhibits potent antimicrobial activity on clindamycin-susceptible and -resistant P. acnes strains and has a low mutant prevention index compared with clindamycin and nadifloxacin. VCD-004 exerts anti-inflammatory action, distinct from its bactericidal property, against dead P. acnes-induced inflammation both in vitro and in vivo, indicating a dual action on the bacteria and the host inflammation. VCD-004 was found to modulate the P. acnes-induced NF-κB signaling pathway. Furthermore, a topical 1% VCD-004 gel showed excellent skin penetration properties and superior efficacy in a P. acnes murine skin infection model. Finally, a favorable safety profile of VCD-004 in a human keratinocyte cell line connotes the emergence of such bi-pharmacophore molecules as a future therapy against acne, even with an underlying resistant P. acnes etiology.

      Results

      Rational design and synthesis of VCD antibiotics

      We used in silico-based design approach to synthesize five antibiotics, based on a strategy to incorporate a nitro-heterocyclic motif on a quinolone scaffold by varying tethering modes that confer different spatial conformations (Figure 1). These five antibiotics are, to the best of our knowledge, previously unreported. We wanted to investigate the preferred anchoring site between C-7 and C-3 of the quinolone scaffold for anchoring the nitro-heterocyclic motif to obtain better binding interactions toward DNA gyrase-DNA complex compared with other known fluoroquinolones. VCD-002, VCD-004, VCD-006, and VCD-007 were designed by incorporating nitro-imidazole moiety to the piperazine unit present at the C-7 position of the molecule, whereas in VCD-003, the C-3 carboxylic site of the quinolone scaffold was modified with a nitro-imidazole unit to affect the normal metal binding mode of quinolones with an Mg2+ ion present at the QBP.
      Figure 1
      Figure 1Molecular structures of the synthesized antibiotics. (a) VCD-002, (b) VCD-003, (c) VCD-004, (d) VCD 006, and (e) VCD-007.
      Because the crystal structure of P. acnes gyrase is not reported, we performed molecular docking to analyze binding of the putative bi-pharmacophore structures in the QBP of DNA gyrase of a Gram-positive bacteria, Staphylococcus aureus, that has considerable amino acid sequence identity with respect to P. acnes (44.5% for gyrase A and 54.9% for gyrase B subunits). The residues of the two species involved in quinolone binding share significant sequence homology (see Supplementary Figure S1a and b online, highlighted in cyan).
      For molecular docking, we used nadifloxacin and ciprofloxacin as reference ligands. Figure 2 shows energy-minimized binding poses of nadifloxacin and designed antibiotics (VCD-002, VCD-004, and VCD-007) bound to the QBP. Nadifloxacin (Figure 2a) binds like ciprofloxacin (see Supplementary Figure S2a online) near the nick of double-stranded DNA, where the aromatic ring of quinolone gains stacking interactions with nucleotide bases deoxycytosine and deoxyguanosine from top and deoxyadenosine and deoxythiamine from bottom. This intercalation between the bases orients the 3-carboxyl group of nadifloxacin toward Mg2+ and the OH-substituent of piperidine group at the C-7 position toward residues D437, R458, N476, and E477 of gyrase B.
      Figure 2
      Figure 2Binding of antibiotics to bacterial DNA gyrase. Interactions of (a) nadifloxacin, (b) VCD-002, (c) VCD-004, and (d) VCD 007. The binding model used for molecular docking was Staphylococcus aureus DNA gyrase-DNA complex (Protein Data Bank identification: 2XCT). Figures were drawn using PyMol (

      Delano WL. The PyMOL Molecular Graphics System, http://www.pymol.org; 2002 (accessed 12 April 2016).

      ). The residue numbers are based on Staphylococcus aureus DNA numbering of the Protein Data Bank file. Dotted lines denote the hydrogen bond interactions and electrostatic interactions between the ligand (pink) and the protein (green) or DNA (cyan) residues. DA, deoxyadenosine; DG, deoxyguanosine; DT, deoxythiamine; DC, deoxycytosine; Gyr, gyrase.
      In contrast, VCD molecules, depending on the spatial orientation of the nitro-heterocyclic group, prefer a different binding orientation in the QBP, where the position of the 3-carboxyl group and piperazine substituent of the structures are inverted compared with nadifloxacin/ciprofloxacin; thus, the functional moieties interacting with the amino acid residues of subunits gyrase A and gyrase B are switched (Figure 2b–d). Such inverse binding results in a co-ordination complex between Mg2+ (Mg2+ bridge) and the nitro group of VCDs, unlike the 3-carboxyl group in the case of nadifloxacin/ciprofloxacin. Among VCD molecules, VCD-004 showed the best docking pose, presenting maximum stacking interactions with the complex. The rigid tethering mode between piperazine and the nitro-heterocyclic motif in VCD-004 conferred maximum stacking interactions with DNA bases compared with other structures (Figure 2c). Moreover, we observed an increase in noncovalent interactions of VCD-004 in the QBP, including hydrogen bonding of the OH group (present in the linker) to deoxyguanosine and the 3-carboxyl group of the drug with R458, E1088, and D437 of DNA gyrase. In contrast, VCD-002 and VCD-007, with nitro-heterocyclic motifs approximately 6 Å away from C-7 piperazine motifs, show reduced stacking interactions, with DNA and hydrogen bonding interactions with the amino acid residues of the DNA-gyrase complex (Figure 2b and c). Similarly, VCD-003 and VCD-006 showed inferior binding compared with VCD-004 (see Supplementary Figure 2b and c online). The docking energies and the crucial noncovalent interactions such as H-bonding, hydrophobic, electrostatic and stacking interaction between the ligand and the DNA gyrase complex are tabulated (Supplementary Table S1 online).
      The molecules were synthesized (described in the Supplementary Materials, and see Supplementary Figures S3 and S4 online) and characterized by 1H nuclear magnetic resonance and electrospray ionization mass spectroscopy and found to be more than 90% pure on high-performance liquid chromatography analysis.
      The in silico data were validated by exploring the action of VCD-004 on S. aureus DNA gyrase in a cell free assay system. VCD-004 was able to inhibit S. aureus DNA gyrase-mediated DNA supercoiling in a dose-dependent manner, indicative of its role in blocking DNA gyrase activity (see Supplementary Figure S5 online). The degree of supercoiling decreased from 100% (in the absence of the drug) to approximately 41% (in the presence of 1 μmol/L VCD-004). Ciprofloxacin was used as a positive control.

      In vitro antibacterial activity of VCD molecules against P. acnes strains

      The synthesized compounds were evaluated for in vitro bioactivity against P. acnes strains. Widespread prevalence of clindamycin resistance (
      • Ross J.I.
      • Snelling A.M.
      • Eady E.A.
      • Cove J.H.
      • Cunliffe W.J.
      • Leyden J.J.
      • et al.
      Phenotypic and genotypic characterization of antibiotic-resistant Propionibacterium acnes isolated from acne patients attending dermatology clinics in Europe, the USA, Japan and Australia.
      ) prompted us to determine the minimum inhibitory concentrations (MICs) of the molecules against clindamycin- and erythromycin-susceptible and -resistant strains (Table 1). Among all VCD molecules, VCD-004 had the lowest MIC values (∼0.19 μg/ml and 0.1 μg/ml, respectively) in both the strains, which were comparable with nadifloxacin MIC values. The MIC of VCD molecules in S. aureus (MTCC 6908) showed consistent results as per the in silico findings. Concentration-response curves of VCD molecules against the earlier-mentioned P. acnes strains further confirmed that VCD-004 triggered a faster onset of bacterial killing compared with the other molecules in both clindamycin-susceptible and -resistant strains (Figure 3). We therefore selected VCD-004 as the molecule of choice for subsequent studies.
      Table 1In vitro antibacterial activity of different antibiotics against bacterial strains
      AntibioticsMIC (μg/ml)
      Propionibacterium acnesStaphylococcus aureus
      MTCC 1951
      MTCC 1951 and CCARM 9010 are P. acnes clindamycin-susceptible and -resistant strains, respectively. S. aureus MTCC 6908 is used as a positive control.
      CCARM 9010
      MTCC 1951 and CCARM 9010 are P. acnes clindamycin-susceptible and -resistant strains, respectively. S. aureus MTCC 6908 is used as a positive control.
      MTCC 6908
      MTCC 1951 and CCARM 9010 are P. acnes clindamycin-susceptible and -resistant strains, respectively. S. aureus MTCC 6908 is used as a positive control.
      Clindamycin0.0232.001.60
      Erythromycin0.02>200.00
      Nadifloxacin0.130.130.13
      VCD-0021.700.900.90
      VCD-0032.902.902.85
      VCD-0040.190.100.38
      VCD-0060.800.801.60
      VCD-0070.800.801.64
      Abbreviation: MIC, minimum inhibitory concentration; MTCC, The Microbial Type Culture Collection and Gene Bank.
      1 MTCC 1951 and CCARM 9010 are P. acnes clindamycin-susceptible and -resistant strains, respectively. S. aureus MTCC 6908 is used as a positive control.
      Figure 3
      Figure 3In vitro antibacterial activity of VCD antibiotics against Propionibacterium acnes strains. Dose-response curves showing bacterial load in (a) clindamycin-susceptible (MTCC 1951) and (b) clindamycin-resistant (CCARM 9010) P. acnes strains. Data represented as mean ± standard deviation. n = 3.
      Next, we tested VCD-004 using a library of P. acnes clinical isolates with a diverse resistance spectrum to clindamycin and erythromycin (
      • Sadhasivam S.
      • Sinha M.
      • Saini S.
      • Kaur S.P.
      • Gupta T.
      • Sengupta S.
      • et al.
      Heterogeneity and antibiotic resistance in Propionibacterium acnes isolates and its therapeutic implications: blurring the lines between commensal and pathogenic phylotypes.
      ). These isolates had mutations either in A2058G or A2059G (Escherichia coli numbering) in bacterial 23S peptidyl ribosomal RNA that conferred resistance or susceptibility to clindamycin, respectively. VCD-004, like nadifloxacin, showed excellent activity (MIC range = 0.10–0.79 μg/ml) across a wide variety of P. acnes strains having diverse phylotypes and resistance patterns (Table 2).
      Table 2In vitro antibacterial activity of clindamycin and VCD-004 in different clinical isolates of Propionibacterium acnes
      Serial NumberClinical Isolate
      The position of mutation in the nucleotide residue (Escherichia coli numbering) of 23S ribosomal RNA is mentioned in parentheses. A2058G mutation is associated with higher degree of clindamycin resistance, whereas A2059G mutation confers clindamycin susceptibility (Sadhasivam et al., 2016).
      Clindamycin MIC (μg/ml)Nadifloxacin MIC (μg/ml)VCD-004 MIC (μg/ml)
      1V21A1 (A2059G)8<0.130.10
      2V21A2 (A2058G)640.300.38
      3V21A3 (A2058G)320.500.19
      4V21A4 (A2058G)640.300.38
      5V21A5 (A2059G)1<0.130.10
      6V21A6 (A2059G)20.300.19
      7V21A7 (A2059G)20.500.76
      8V21A8 (A2059G)2<0.180.19
      9V21A9 (A2058G)320.300.76
      10V21A10 (A2058G)640.30.19
      Abbreviation: MIC, minimum inhibitory concentration.
      1 The position of mutation in the nucleotide residue (Escherichia coli numbering) of 23S ribosomal RNA is mentioned in parentheses. A2058G mutation is associated with higher degree of clindamycin resistance, whereas A2059G mutation confers clindamycin susceptibility (
      • Sadhasivam S.
      • Sinha M.
      • Saini S.
      • Kaur S.P.
      • Gupta T.
      • Sengupta S.
      • et al.
      Heterogeneity and antibiotic resistance in Propionibacterium acnes isolates and its therapeutic implications: blurring the lines between commensal and pathogenic phylotypes.
      ).

      Mutant prevention concentration of VCD-004 against P. acnes (MTCC 1951)

      It is essential to estimate the probability that antibiotics under development will restrict the generation of resistant strains. Mutant prevention concentration (MPC) signifies the lowest drug concentration above which the selective proliferation of resistant mutants is expected to occur only rarely (
      • Blondeau J.M.
      New concepts in antimicrobial susceptibility testing: the mutant prevention concentration and mutant selection window approach.
      ). The MPC of VCD-004 was compared with other anti-acne antibiotics, nadifloxacin and clindamycin. VCD-004 showed an MPC index (MPC:MIC ratio) of 1.5, compared with 9.2 and 6.5 for nadifloxacin and clindamycin, respectively (Table 3), indicating that P. acnes has less probability of developing resistance against VCD-004 than against clindamycin and nadifloxacin.
      Table 3MPC of VCD-004 in Propionibacterium acnes (MTCC 1951) compared with currently used antibiotics clindamycin and nadifloxacin
      DrugMIC (μg/ml)MPC (μg/ml)MPC:MIC
      VCD-0040.200.301.5
      Clindamycin0.020.136.5
      Nadifloxacin0.131.209.2
      Abbreviations: MIC, minimum inhibitory concentration; MPC, mutant prevention concentration.

      VCD-004 exerts anti-inflammatory effects against P. acnes-induced inflammation

      Because acne has an underlying inflammatory etiology, VCD-004 was designed to exhibit anti-inflammatory properties in addition to its antibacterial effects. We studied in vitro anti-inflammatory potency of VCD-004 in human monocytic THP-1 cells, a cell line well established for study of immune modulation (
      • Chanput W.
      • Mes J.J.
      • Wichers H.J.
      THP-1 cell line: an in vitro cell model for immune modulation approach.
      ). The cells were induced using heat-killed P. acnes to selectively assess the anti-inflammatory effects of VCD-004, followed by measurement of acne-relevant cytokines (IL-1α, IL-1β, IL-6, and IL-8) in the culture supernatant. At the tested concentration (25 μg/ml), VCD-004 caused a significant reduction (nearly 60%) in P. acnes-induced IL-6 levels (Figure 4a) and was nontoxic to the cells (see Supplementary Figure S6a online). Dexamethasone, a steroidal anti-inflammatory agent used as a positive control, showed nearly 100% reduction in IL-6 levels and was nontoxic to the cells at the tested concentration (see Supplementary Figure S6b). VCD-004 did not have a significant effect on IL-8 levels (Figure 4b); however, there were detectable inhibitory effects of VCD-004 (∼20–25% reduction) on P. acnes-induced IL-1α and IL-1β levels (Figures 4c and 4d).
      Figure 4
      Figure 4Anti-inflammatory effects of VCD-004 against Propionibacterium acnes. In vitro effect of VCD-004 on P. acnes induced cytokines (a) IL-6, (b) IL-8, (c) IL-1α, and (d) IL-1β release in THP-1 cells. Dexamethasone (Dexameth) was used as positive control. The data are represented as mean ± standard deviation. n = 3. (e, f) In vivo effect of VCD-004, 1% topical formulation on dead P. acnes-induced paw edema. (e) Change in paw size (%) relative to baseline, in rats, after induction of inflammation by dead P. acnes injection and treatment with VCD-004 formulation/marketed formulations at two different time points, 8 hours and 32 hours after induction of inflammation. The data are represented as mean ± standard deviation. n = 4. Treatment with the drug was started at 8 hours after infection. IC group received P. acnes injection but no treatment. (f) Percentage inhibition of paw swelling by different formulations at 32 hours after induction of inflammation compared with IC. IC, induction control.
      Literature reports suggest the role of NF-κB in the cytokine response to P. acnes induction (
      • Chen Q.
      • Koga T.
      • Uchi H.
      • Hara H.
      • Terao H.
      • Moroi Y.
      • Urabe K.
      • Furue M.
      Propionibacterium acnes-induced IL-8 production may be mediated by NF-kappaB activation in human monocytes.
      ,
      • Grange P.A.
      • Raingeaud J.
      • Calvez V.
      • Dupin N.
      Nicotinamide inhibits Propionibacterium acnes-induced IL-8 production in keratinocytes through the NF-kappaB and MAPK pathways.
      ,
      • Huang W.C.
      • Tsai T.H.
      • Chuang L.T.
      • Li Y.Y.
      • Zouboulis C.C.
      • Tsai P.J.
      Anti-bacterial and anti-inflammatory properties of capric acid against Propionibacterium acnes: a comparative study with lauric acid.
      ,
      • Tsai H.H.
      • Lee W.R.
      • Wang P.H.
      • Cheng K.T.
      • Chen Y.C.
      • Shen S.C.
      Propionibacterium acnes-induced iNOS and COX-2 protein expression via ROS-dependent NF-κB and AP-1 activation in macrophages.
      ). We investigated the effect of VCD-004 in NF-κB activation by looking at the levels of nuclear NF-κB in VCD-004–treated or untreated THP-1 cells induced with dead P. acnes. Our results show a clear increase in the abundance of NF-κB in the nucleus 30 minutes after induction with dead bacteria (see Supplementary Figure S7 online, lane 5). Pre-treatment with VCD-004 (25 μg/ml) prevents this NF-κB accumulation in the nucleus 30 minutes after induction with dead P. acnes (see Supplementary Figure S7, lane 6). These data suggest that VCD-004 may act through the NF-κB pathway to influence cytokine gene expression in THP-1 cells in response to dead P. acnes exposure.
      We further extended the study of VCD-004 anti-inflammatory effects in a dead P. acnes-induced rat paw edema model. Paw sizes were measured before (time T0) and 8 hours after injection of dead P. acnes in the plantar surface of the hind paw. Test formulations (1% topical gel of VCD-004) were applied topically at the site of inflammation, and paw sizes were measured 24 hours after treatment (32 hours after induction). Measurable paw swelling was achieved 8 hours after induction, and we observed significant differences in paw sizes between treated and untreated animals at 24 hours after treatment (Figure 4e). VCD-004 mediated 74.01 ± 13.04% inhibition of paw swelling versus untreated infection controls. This reduction was equivalent to that induced by dapsone (67.43 ± 17.98%), a known anti-inflammatory agent approved for acne therapy (Figure 4f). The placebo formulation of VCD-004 did not elicit any significant change compared with infection control 24 hours after treatment.
      We further explored the effect of VCD-004 on different cytokine expressions in the same model by evaluating mRNA expression of paw tissue cytokines using quantitative real time PCR. IL-1α, IL-1β, and IL-6 expressions were assessed in rat paw tissues of infection control, VCD-004–treated, and dapsone-treated animals 32 hours after induction. VCD-004 treatment resulted in significant reduction of IL-1α gene expression compared with infection control. Mild reduction was apparent for IL-6 expression, and IL-1β did not show any difference from the infection control (see Supplementary Figure S8 online). Dapsone had no effect on any of the tested cytokines.

      Topical application of VCD-004 (1% topical gel) decreases P. acnes infection load in vivo

      To test the efficacy of the 1% VCD-004 topical gel formulation, we used a murine P. acnes ear infection model. Live P. acnes (strain number CCARM 9010) (1 × 107 colony-forming units) were injected intradermally into mouse ears, and bacterial load achieved was measured at 8 hours after injection. Topical treatment with VCD-004 1% gel, vehicle, or reference clindamycin 1% gel was performed every 12 hours starting at 8 hours after infection. VCD-004 imparted a significant reduction in P. acnes count (greater than 1 log) after 40 hours from the initiation of treatment, which was significantly greater than the 0.5 log reduction in P. acnes load achieved with clindamycin gel (1%) at the same time point under similar experimental conditions. The vehicle control showed no appreciable decrease in infection at 40 hours after treatment (Figure 5a–e).
      Figure 5
      Figure 5In vivo efficacy of VCD-004 topical formulation in Propionibacterium acnes-induced mouse ear infection model. (a) Bacterial load (log10 CFU/ear) was determined at designated time points during the course of the efficacy experiment after treatment with drugs, VCD-004 1% topical formulation/marketed 1% clindamycin gel as per prescribed dose. Infected vehicle control (VC) did not receive any treatment. Treatment time points were 8 hours, 20 hours, 32 hours, and 44 hours after infection (shown by arrows). The data are represented as mean ± standard deviation. n = 4. (b–e) Pictures of mouse ear at 40 hours after treatment with drug: (b) uninfected control, (c) infected vehicle control, (d) VCD-004 1% topical gel treated, and (e) marketed 1% clindamycin gel treated. CFU, colony-forming unit; h, hour.

      In vitro cytotoxicity in human keratinocyte cells

      A potential candidate for topical application, we tested the cytotoxicity of VCD-004 in a human keratinocyte cell line (HaCaT) at different concentrations (between 25 μg/ml and 400 μg/ml). The half maximal inhibitory concentration of the molecule could not be achieved at the tested range (see Supplementary Figure S9 online).

      Discussion

      With the emergence of resistance to the currently approved antibiotics in clinical isolates of P. acnes, there is an urgent need to look beyond the existing repertoire of therapeutic agents for targeting P acnes. Although antimicrobials, such as benzoyl peroxides, have emerged as an alternate choice, clinical evidence shows that such antimicrobials are maximally effective only when combined with antibiotics. Therefore, there is an unmet need for next-generation antibiotics that, however, need to meet the benchmark of being able to retard the development of resistance in addition to being effective against resistant strains of P. acnes. Here, we report the rational design of an antibiotic, VCD-004, to the best of our knowledge previously unreported, that exerts a bactericidal activity against P. acnes strains resistant to existing first-line antibiotics and exhibits excellent resistance-retarding properties. Additionally, VCD-004 exerts an anti-inflammatory effect independent of its bactericidal effect.
      Although our approach for designing the described library of antibiotics was based on the integration of a nitro-heterocyclic moiety into a quinolone scaffold, the diversity of the structure-activity relationship encoded in the library is evident from the fact that the molecules exhibited different degrees of efficacy against P. acnes. This suggests that just the presence of two pharmacophores is not sufficient for the bactericidal effect; the conformation of the final structure and the spatial orientation of both pharmacophores together contribute to the binding with the DNA-gyrase complex, resulting in differential antibacterial efficacy. The sequence homology between active site residues of S. aureus and P. acnes DNA gyrase justifies the correlation between in silico screening of the molecules using S. aureus DNA gyrase with antibacterial efficacy against P. acnes. VCD-004 outperformed all other synthesized bi-pharmacophore structures in terms of binding to the bacterial target, forming crucial interactions with the enzyme and nucleotide residues in its docked pose at the inhibitor site compared with other molecules. Additionally, quinolone resistance is reported to arise because of point mutations in DNA gyrase, primarily in residues S1084 and E477 (
      • Hooper D.C.
      Emerging mechanisms of fluoroquinolone resistance.
      ). VCD-004 was not involved in direct hydrogen bonding with these two commonly mutating residues. This improved mode of binding of VCD-004 could thus explain the narrow window between MPC and MIC, indicating a greater ability to prevent the emergence of mutants. Our data indicate that low dose of VCD-004 is sufficient to severely restrict the selection of resistant mutants compared with currently available drugs. This further emphasizes the advantages of a carefully designed entity with distinct interactions over the repurposing of old systemic antibiotics.
      Apart from antibacterial effect, the ability to reduce host-mediated inflammation can have a critical impact on the clinical outcome of acne therapy.
      • Kim J.
      • Ochoa M.T.
      • Krutzik S.R.
      • Takeuchi O.
      • Uematsu S.
      • Legaspi A.J.
      • et al.
      Activation of toll-like receptor 2 in acne triggers inflammatory cytokine responses.
      reported that P. acnes triggers an inflammatory cytokine response including TNF-α, IL-6, IL-8, and IL-12 in macrophages by the activation of TLR-2. The IL-1 family of cytokines, especially IL-1α, has been implicated as an initiator and a key player associated with tissue damage in acne patients (
      • Ingham E.
      • Eady E.A.
      • Goodwin C.E.
      • Cove J.H.
      • Cunliffe W.J.
      Pro-inflammatory levels of interleukin-1 alpha-like bioactivity are present in the majority of open comedones in acne vulgaris.
      ). In this study, dead P. acnes was used to induce inflammation in both in vitro and in vivo models to evaluate the direct anti-inflammatory activity of VCD-004, independent of its bactericidal action. Use of relevant bacteria as an inducing agent was an intuitive choice to mimic the inflammatory cascades in acne as closely as possible. In vitro, VCD-004 significantly reduced IL-6 levels. Multiple studies have implicated P. acnes in the induction of IL-6 in different cell types relevant to skin, including macrophages, keratinocytes, and sebocytes (
      • Huang W.C.
      • Tsai T.H.
      • Chuang L.T.
      • Li Y.Y.
      • Zouboulis C.C.
      • Tsai P.J.
      Anti-bacterial and anti-inflammatory properties of capric acid against Propionibacterium acnes: a comparative study with lauric acid.
      ,
      • Kim J.
      • Ochoa M.T.
      • Krutzik S.R.
      • Takeuchi O.
      • Uematsu S.
      • Legaspi A.J.
      • et al.
      Activation of toll-like receptor 2 in acne triggers inflammatory cytokine responses.
      ,
      • Lee W.J.
      • Jung H.D.
      • Chi S.G.
      • Kim B.S.
      • Lee S.J.
      • Kim D.W.
      • et al.
      Effect of dihydrotestosterone on the upregulation of inflammatory cytokines in cultured sebocytes.
      ). IL-6 expression has been found to be enhanced in acne-involved skin (
      • Alestas T.
      • Ganceviciene R.
      • Fimmel S.
      • Müller-Decker K.
      • Zouboulis C.C.
      Enzymes involved in the biosynthesis of leukotriene B4 and prostaglandin E2 are active in sebaceous glands.
      ). The expression of cytokines via NF-κB signaling in response to P. acnes exposure has been reported in various cell types including human monocytes (
      • Chen Q.
      • Koga T.
      • Uchi H.
      • Hara H.
      • Terao H.
      • Moroi Y.
      • Urabe K.
      • Furue M.
      Propionibacterium acnes-induced IL-8 production may be mediated by NF-kappaB activation in human monocytes.
      ), keratinocytes (
      • Grange P.A.
      • Raingeaud J.
      • Calvez V.
      • Dupin N.
      Nicotinamide inhibits Propionibacterium acnes-induced IL-8 production in keratinocytes through the NF-kappaB and MAPK pathways.
      ), and sebocytes (
      • Huang Y.C.
      • Yang C.H.
      • Li T.T.
      • Zouboulis C.C.
      • Hsu H.C.
      Cell-free extracts of Propionibacterium acnes stimulate cytokine production through activation of p38 MAPK and Toll-like receptor in SZ95 sebocytes.
      ). Our observation of decreased nuclear NF-κB levels in VCD-004–treated samples of THP-1 cells induced with dead P. acnes suggests that VCD-004 may mediate its anti-inflammatory properties, at least in part, through modulation of the NF-κB pathway of cytokine expression.
      Furthermore, the comparable efficacy of a topical formulation of VCD-004 and dapsone, in a rat paw edema model, indicates that VCD-004 could emerge as an exciting bi-functional molecule for acne treatment, with the reduction in inflammation contributing to improved therapeutic efficacy. A significant decrease in IL-1α gene expression in VCD-004–treated paw tissues indicates that the anti-inflammatory role of the drug in vivo may partly act via reduction of P. acnes-induced cytokines. Dapsone did not mediate a decrease in any of the tested cytokine expressions in the rat paw edema model, because it may recruit alternative anti-inflammatory mediators (
      • Booth S.A.
      • Moody C.E.
      • Dahl M.V.
      • Herron M.J.
      • Nelson R.D.
      Dapsone suppresses integrin-mediated neutrophil adherence function.
      ). We believe that the anti-inflammatory action of VCD-004 is greater than that which can be potentially achieved through its bactericidal activity, because inflammation was induced using dead bacteria in these experiments. Although there are no good models to study acne in mice, the in vivo efficacy of 1% VCD-004 gel in the P. acnes ear infection model and in the rat paw edema model indicates that the drug can indeed modulate two key components in acne pathogenesis.
      For delivering and retaining sufficient amount of the drug in the relevant skin layers, the n-octanol/water partition coefficient (P) of the drug molecules becomes an important determinant. The theoretical log P value of 1.23 for VCD-004 (ChemAxon, version 16.9.12; ChemAxon, Budapest, Hungary) is consistent with the range of log P values that facilitates drug penetration (
      • Liu X.
      • Testa B.
      • Fahr A.
      Lipophilicity and its relationship with passive drug permeation.
      ). An ex vivo skin penetration of 1% VCD-004 formulation showed a high amount of skin retention, surpassing the required MIC level of VCD-004 against P. acnes by several fold (see Supplementary Materials).
      Prolonged use of the current repertoire of antibiotics has led to the prevalence of resistant P. acnes strains in the skin microbiome (
      • Leyden J.J.
      • McGinley K.J.
      • Cavalieri S.
      • Webster G.F.
      • Mills O.H.
      • Kligman A.M.
      Propionibacterium acnes resistance to antibiotics in acne patients.
      ), and the incidence of drug-resistant acne is growing rapidly (
      • Sinha M.
      • Sadhasivam S.
      • Bhattacharyya A.
      • Jain S.
      • Ghosh S.
      • Arndt K.A.
      • et al.
      Antibiotic-resistant acne: getting under the skin.
      ). There is need for an antibiotic that inhibits a target distinct from the ribosomal targets inhibited by the currently approved antibiotics for acne. Several aspects of VCD-004 can facilitate future acne therapy. The ability of VCD-004 to bind to gyrase and induce P. acnes cell death at sub-micromolar concentrations can therefore address the emerging challenge of drug-resistant acne. The faster onset of bactericidal effect could potentially have an implication in the clinic. Second, the ability of VCD-004 to bind to the QBP via a unique mechanism means that it can potentially retard development of resistance, as validated by its low MPC/MIC ratio compared with currently used drugs. Furthermore, the lipophilicity, good skin penetration, and safety profile (human keratinocyte cell line) of VCD-004 may allow the drug to maintain bactericidal activity in a sebaceous environment to obtain maximum therapeutic outcomes and, hence, lead to the use of the molecule as a topical agent. Finally, a significant decrease in acne-associated inflammation by VCD-004 would provide significant therapeutic relief to patients suffering from acne. Such rationally designed multifunctional antibiotics can emerge as the unique paradigm in the management of multifactorial diseases like acne.

      Materials and Methods

      Materials

      All materials procured for experimental procedures are included in the Supplementary Materials.

      Bacterial strains

      P. acnes MTCC 1951 and S. aureus MTCC 6908 were obtained from MTCC (Institute of Microbial Technology, Chandigarh, India). P. acnes CCARM 9010 was procured from CCARM (Seoul, Korea). Clinical isolates were established and characterized as described earlier (
      • Sadhasivam S.
      • Sinha M.
      • Saini S.
      • Kaur S.P.
      • Gupta T.
      • Sengupta S.
      • et al.
      Heterogeneity and antibiotic resistance in Propionibacterium acnes isolates and its therapeutic implications: blurring the lines between commensal and pathogenic phylotypes.
      ). P. acnes ATCC 6919 was procured from Anthem Biosciences (Bangalore, India) and from American Type Culture Collection (Manassas, VA).

      Cell lines

      The human monocyte cell line THP-1 was procured from Anthem Biosciences and from American Type Culture Collection. The human keratinocyte cell line HaCaT was obtained from NCCS (Pune, India) by TheraIndx Lifesciences (Bangalore, India).

      Animals

      Animal studies were performed at TheraIndx Lifesciences (Bangalore, India) by following all ethical practices as laid down in the guidelines for animal care (as per Government of India, Committee for the Purpose of Control and Supervision of Experiments on Animals). The animal studies had been performed by certified personnel after receiving Institute ethical clearances from TheraIndx Lifesciences (Bangalore, India). Female BALB/c mice (6–8 weeks old) and male Sprague Dawley rats (6–8 weeks old) were procured from Vivo Biotech (Hyderabad, India).

      Design and synthesis of bi-pharmacophore molecules

      Complete synthetic schemes of all molecules and the experimental procedures for synthesis are described in Supplementary Figure S1a and discussed in the Supplementary Materials and Methods.

      Synthesis of 1% VCD-004 topical gel

      Preparation of topical formulation of 1% VCD-004 is described in the Supplementary Materials.

      Molecular docking of VCD molecules in the bacterial DNA gyrase-DNA complex

      The details of the docking procedure are provided in the Supplementary Materials.

      Antibacterial potency of the library of VCD molecules against P. acnes strains

      MIC of drugs against different P. acnes and S. aureus strains was determined as per Clinical & Laboratory Standards Institute 2012 guidelines M11-A8 and M100S-S22, respectively, (see Supplementary Materials).

      Determination of mutant prevention concentration

      MPC of VCD-004 was determined against P. acnes MTCC 1951 as described by
      • Metzler K.
      • Drlica K.
      • Blondeau J.M.
      Minimal inhibitory and mutant prevention concentrations of azithromycin, clarithromycin and erythromycin for clinical isolates of Streptococcus pneumoniae.
      . The details are depicted in the Supplementary Materials.

      In vitro anti-inflammatory activities of VCD-004 in the THP-1 cell line

      The study details are available in the Supplementary Materials. The primers used for the cytokines and control gene (GAPDH) are listed in Supplementary Table S2 online.

      P. acnes-induced rat paw edema studies

      Sprague Dawley rats were used to study paw edema induced by P. acnes CCARM 9010. The animal studies were performed by certified personnel after receiving Institute ethical clearances (see Supplementary Materials).

      Western blot analysis of nuclear NF-κB p65 in THP-1 cells

      The details are provided in the Supplementary Materials.

      In vivo efficacy studies in a P. acnes skin infection model

      The study is described in the Supplementary Materials. The animal studies were performed following all ethical practices as laid down in the guidelines for animal care by the Government of India, Committee for the Purpose of Control and Supervision of Experiments on Animals.

      MTT assay of VCD-004 in human keratinocyte (HaCaT) cells

      The study is described in the Supplementary Materials.

      Statistical analysis

      All the results are expressed as the mean ± standard deviation for the number of separate experiments indicated in each case (n ≥ 3). Student t test was used for comparison between two groups.

      Conflict of Interest

      Sumana Ghosh, S Sadhasivam, S Saini, and Shamik Ghosh are employees of Vyome Biosciences. S Sengupta is a co-founder of Vyome Biosciences and member of the Board of Directors. He holds equity in and acts as a consultant for Vyome Biosciences. SR, SPK, and DK are former employees of Vyome Biosciences. JM and DU state no conflict of interest. Vyome Biosciences provided the financial support for the study.

      Supplementary Material

      References

        • Alestas T.
        • Ganceviciene R.
        • Fimmel S.
        • Müller-Decker K.
        • Zouboulis C.C.
        Enzymes involved in the biosynthesis of leukotriene B4 and prostaglandin E2 are active in sebaceous glands.
        J Mol Med (Berl). 2006; 84: 75-87
        • Aslam I.
        • Fleischer A.
        • Feldman S.
        Emerging drugs for the treatment of acne.
        Expert Opin Emerg Drugs. 2015; 20: 91-101
        • Bannatyne R.M.
        Metronidazole, its bioactive metabolites and acne.
        Curr Med Res Opin. 1999; 15: 298-299
        • Blondeau J.M.
        New concepts in antimicrobial susceptibility testing: the mutant prevention concentration and mutant selection window approach.
        Vet Dermatol. 2009; 20: 383-396
        • Booth S.A.
        • Moody C.E.
        • Dahl M.V.
        • Herron M.J.
        • Nelson R.D.
        Dapsone suppresses integrin-mediated neutrophil adherence function.
        J Invest Dermatol. 1992; 98: 135-140
        • Chen Q.
        • Koga T.
        • Uchi H.
        • Hara H.
        • Terao H.
        • Moroi Y.
        • Urabe K.
        • Furue M.
        Propionibacterium acnes-induced IL-8 production may be mediated by NF-kappaB activation in human monocytes.
        J Dermatol Sci. 2002; 29: 97-103
        • Chanput W.
        • Mes J.J.
        • Wichers H.J.
        THP-1 cell line: an in vitro cell model for immune modulation approach.
        Int Immunopharmacol. 2014; 23: 37-45
        • Cunliffe W.J.
        • Clayden A.D.
        • Gould D.
        • Simpson N.B.
        Acne vulgaris—its aetiology and treatment. A review.
        Clin Exp Dermatol. 1981; 6: 461-469
      1. Delano WL. The PyMOL Molecular Graphics System, http://www.pymol.org; 2002 (accessed 12 April 2016).

        • Grange P.A.
        • Raingeaud J.
        • Calvez V.
        • Dupin N.
        Nicotinamide inhibits Propionibacterium acnes-induced IL-8 production in keratinocytes through the NF-kappaB and MAPK pathways.
        J Dermatol Sci. 2009; 56: 106-112
        • Hooper D.C.
        Emerging mechanisms of fluoroquinolone resistance.
        Emerg Infect Dis. 2001; 7: 337-341
        • Huang W.C.
        • Tsai T.H.
        • Chuang L.T.
        • Li Y.Y.
        • Zouboulis C.C.
        • Tsai P.J.
        Anti-bacterial and anti-inflammatory properties of capric acid against Propionibacterium acnes: a comparative study with lauric acid.
        J Dermatol Sci. 2014; 73: 232-240
        • Huang Y.C.
        • Yang C.H.
        • Li T.T.
        • Zouboulis C.C.
        • Hsu H.C.
        Cell-free extracts of Propionibacterium acnes stimulate cytokine production through activation of p38 MAPK and Toll-like receptor in SZ95 sebocytes.
        Life Sci. 2015; 139: 123-131
        • Ingham E.
        • Eady E.A.
        • Goodwin C.E.
        • Cove J.H.
        • Cunliffe W.J.
        Pro-inflammatory levels of interleukin-1 alpha-like bioactivity are present in the majority of open comedones in acne vulgaris.
        J Invest Dermatol. 1992; 98: 895-901
        • Jacobs M.R.
        • Appelbaum P.C.
        Nadifloxacin: a quinolone for topical treatment of skin infections and potential for systemic use of its active isomer, WCK 771.
        Expert Opin Pharmacother. 2006; 7: 1957-1966
        • Kampranis S.C.
        • Maxwell A.
        The DNA gyrase-quinolone complex. ATP hydrolysis and the mechanism of DNA cleavage.
        J Biol Chem. 1998; 273: 22615-22626
        • Kim J.
        • Ochoa M.T.
        • Krutzik S.R.
        • Takeuchi O.
        • Uematsu S.
        • Legaspi A.J.
        • et al.
        Activation of toll-like receptor 2 in acne triggers inflammatory cytokine responses.
        J Immunol. 2002; 169: 1535-1541
        • Lee W.J.
        • Jung H.D.
        • Chi S.G.
        • Kim B.S.
        • Lee S.J.
        • Kim D.W.
        • et al.
        Effect of dihydrotestosterone on the upregulation of inflammatory cytokines in cultured sebocytes.
        Arch Dermatol Res. 2010; 302: 429-433
        • Leyden J.J.
        Acne vulgaris is a multifactorial disease.
        J Am Acad Dermatol. 2003; 49: S199
        • Leyden J.J.
        • McGinley K.J.
        • Cavalieri S.
        • Webster G.F.
        • Mills O.H.
        • Kligman A.M.
        Propionibacterium acnes resistance to antibiotics in acne patients.
        J Am Acad Dermatol. 1983; 8: 41-45
        • Liu X.
        • Testa B.
        • Fahr A.
        Lipophilicity and its relationship with passive drug permeation.
        Pharm Res. 2011; 28: 962-977
        • Metzler K.
        • Drlica K.
        • Blondeau J.M.
        Minimal inhibitory and mutant prevention concentrations of azithromycin, clarithromycin and erythromycin for clinical isolates of Streptococcus pneumoniae.
        J Antimicrob Chemother. 2013; 68: 631-635
        • Millikan L.
        The proposed inflammatory pathophysiology of rosacea: implications for treatment.
        Skinmed. 2003; 2: 43-47
        • Nast A.
        • Dréno B.
        • Bettoli V.
        • Degitz K.
        • Erdmann R.
        • Finlay A.Y.
        • et al.
        European Dermatology Forum. European evidence-based (S3) guidelines for the treatment of acne.
        J Eur Acad Dermatol Venereol. 2012; 26: 1-29
        • Ross J.I.
        • Snelling A.M.
        • Eady E.A.
        • Cove J.H.
        • Cunliffe W.J.
        • Leyden J.J.
        • et al.
        Phenotypic and genotypic characterization of antibiotic-resistant Propionibacterium acnes isolated from acne patients attending dermatology clinics in Europe, the USA, Japan and Australia.
        Br J Dermatol. 2001; 144: 339-346
        • Sadhasivam S.
        • Sinha M.
        • Saini S.
        • Kaur S.P.
        • Gupta T.
        • Sengupta S.
        • et al.
        Heterogeneity and antibiotic resistance in Propionibacterium acnes isolates and its therapeutic implications: blurring the lines between commensal and pathogenic phylotypes.
        Dermatol Ther. 2016; 29: 451-454
        • Shakir L.
        • Javeed A.
        • Ashraf M.
        • Riaz A.
        Metronidazole and the immune system.
        Pharmazie. 2011; 66: 393-398
        • Simonart T.
        • Dramaix M.
        Treatment of acne with topical antibiotics: lessons from clinical studies.
        Br J Dermatol. 2005; 153: 395-403
        • Sinha M.
        • Sadhasivam S.
        • Bhattacharyya A.
        • Jain S.
        • Ghosh S.
        • Arndt K.A.
        • et al.
        Antibiotic-resistant acne: getting under the skin.
        Semin Cutan Med Surg. 2016; 35: 62-67
        • Tan H.H.
        Topical antibacterial treatments for acne vulgaris: comparative review and guide to selection.
        Am J Clin Dermatol. 2004; 5: 79-84
        • Tan J.K.
        • Bhate K.
        A global perspective on the eidemeology of acne.
        Br J Dermatol. 2015; 172: 3-12
        • Thiboutot D.
        • Gollnick H.
        • Bettoli V.
        • Dréno B.
        • Kang S.
        • Leyden J.J.
        • et al.
        Global Alliance to Improve Outcomes in Acne. New insights into the management of acne: an update from the Global Alliance to Improve Outcomes in Acne group.
        J Am Acad Dermatol. 2009; 60: S1-S50
        • Tsai H.H.
        • Lee W.R.
        • Wang P.H.
        • Cheng K.T.
        • Chen Y.C.
        • Shen S.C.
        Propionibacterium acnes-induced iNOS and COX-2 protein expression via ROS-dependent NF-κB and AP-1 activation in macrophages.
        J Dermatol Sci. 2013; 69: 122-131