Synthesis of vancomycin fluorescent probes that retain antimicrobial activity, identify Gram-positive bacteria, and detect Gram-negative outer membrane damage

The study addresses the need for mechanism-informed fluorescent probes that retain antimicrobial activity to study bacteria and antibiotic interactions, amid rising antimicrobial resistance. Vancomycin, a key glycopeptide antibiotic for Gram-positive infections, binds the D-Ala-D-Ala terminus of Lipid II. Resistance in Enterococci (vanA/vanB-mediated D-Ala-D-Lac substitution) and glycopeptide-intermediate S. aureus (GISA) threatens efficacy, with VRSA cases also reported. Existing membrane dyes (e.g., NPN, SYTOX Green, propidium iodide) are widely used to assess membrane damage but lack mechanism specificity. Previously reported vancomycin-fluorophore conjugates (e.g., BODIPY, fluorescein, rhodamine, Alexa Fluor 532, IRDye 800CW) have been used for localization and diagnostic studies; however, many show reduced antimicrobial activity, particularly with bulky or negatively charged fluorophores. The authors hypothesize that small, minimally charged fluorophores attached via a modular azide intermediate at the vancomycin C-terminus can yield fluorescent probes that retain antimicrobial activity and enable selective visualization of Gram-positive bacteria and detection of Gram-negative outer membrane permeabilization.

The paper surveys prior work on fluorescent dyes for membrane integrity (NPN, SYTOX Green, propidium iodide) and antibiotic-derived probes. Reported vancomycin conjugates include BODIPY FL, fluorescein, rhodamine, Alexa Fluor 532, and IRDye 800CW, used for studies of localization, biofilm penetration, resistance mechanisms, and diagnostics. Activity losses were noted for some derivatives (e.g., fluorescein–vancomycin MIC 20 µg/mL; BODIPY–vancomycin MIC 2.5 µg/mL vs vancomycin 0.13 µg/mL against B. subtilis), likely due to charge and size effects interacting with negatively charged teichoic acids. Click-chemistry (CuAAC) has been previously applied to glycopeptides for modified derivatives, dimers, and probes. These insights motivate developing a modular, small-fluorophore probe set retaining activity and offering different colors for co-labelling.

- Probe design and synthesis: Vancomycin was modified at the C-terminal carboxyl group via PyBOP-mediated amidation to introduce azido linkers: 3-azidopropyl (Van-C3-N3, 2), 8-azidooctyl (Van-C8-N3, 3), and azido-PEG3 (Van-PEG3-N3, 4). Cu-catalyzed azide–alkyne cycloaddition (CuAAC) was used to conjugate small, minimally charged fluorophores: 7-nitrobenzofurazan (NBD) and 7-(dimethylamino)coumarin-4-acetic acid (DMACA), yielding probes 7 (Van-8C-Tz-NBD), 8 (Van-8C-Tz-DMACA), and 9 (Van-PEG3-Tz-NBD). CuAAC conditions were optimized (DMF/t-BuOH/H2O, CuSO4·5H2O with Na ascorbate, HOAc, 100 °C microwave, 15 min) with fluorophore-specific catalyst loadings due to vancomycin’s Cu-chelation tendencies. - Antimicrobial activity (MIC): Broth microdilution in non-binding surface 96-well plates against Gram-positive strains (S. aureus including MRSA/GISA/VRSA, S. pneumoniae, E. faecalis, E. faecium) and Gram-negative E. coli strains (wild-type and mutants). Stocks of probes were in water; commercial probes (Van-FITC, Van-BODIPY) in DMSO as per manuals. - Flow cytometry: Bacteria incubated with probes (typically 16–32 µg/mL, 37 °C), fluorescence measured (FITC and blue channels), gating to obtain mean fluorescence intensity and total fluorescence intensity. Co-staining with SYTO 60 to assess Gram selectivity. - High-resolution microscopy: SR-SIM and 3D-SIM for Gram-positive labeling (co-stains: FM4-64FX membrane dye; Hoechst 33342 or SYTO 21 nucleic acid). Airyscan super-resolution confocal used for comparison with commercial probes and for E. coli temperature experiments. - OM permeabilization assays (Gram-negative): SR-SIM imaging of E. coli strains (wild-type; mutants DC2, lpxC, waaL) co-stained with FM4-64FX to assess probe 9 access to PG under genetic perturbations. Temperature-dependent OM integrity assessed at 15 °C vs 37 °C. - Quantification of OM damage: Plate reader and flow cytometry workflows developed using E. coli treated with membrane-active antibiotics (tachyplesin-1, arenicin-3, polymyxin B, colistin, citropin, octapeptin C4) vs non-membrane-active controls (gentamicin, trimethoprim, erythromycin), followed by probe 9 labeling. - Single-cell microfluidics: Mother-machine device used to monitor accumulation of probe 9 in hundreds of individual E. coli over time, with/without polymyxin B. Image analysis extracted fluorescence normalization by cell size. Mathematical modeling inferred kinetic parameters: time of onset (t0), uptake rate constant (k1), and saturation (Fmax).

- Synthesis: Efficient modular route to azido-vancomycin intermediates (2–4) and fluorescent probes (7–9) via optimized CuAAC (completed in 15 min at 100 °C; good yields). NBD probe 7 (Ex/Em 475/535 nm) and DMACA probe 8 (375/480 nm) provide complementary colors; additional NBD probe 9 via PEG3 linker. - Antimicrobial activity (Gram-positive): Azido-vancomycin 3 (C8 linker) showed improved potency vs vancomycin, typically 2–4× better MICs (e.g., MRSA 0.5 µg/mL; VRE 32 µg/mL; GISA 2 µg/mL vs vancomycin 1–2, ≥32, and 8 µg/mL respectively). Fluorescent probes 7 and 8 (C8 linkers) retained or improved activity ~2–4× vs vancomycin across multiple strains. Probe 9 (PEG3) most closely matched parent vancomycin MICs. Compared to commercial probes: Van-BODIPY stained strongly but showed non-specific E. coli labeling; Van-FITC had poor antimicrobial activity and staining. - Selectivity and staining: Flow cytometry showed strong Gram-positive selectivity. Selectivity ratios (S. aureus over E. coli, fluorescence uptake): 7 = 101:1; 8 = 25:1; 9 = 52:1. Probe 9 uptake varied with resistance phenotype: increased fluorescence in GISA and daptomycin-resistant strains (≈4–18× at 64 µg/mL vs ATCC 25923), and reduced in VRSA (≈6–50× lower). - Microscopy of Gram-positive bacteria: SR-SIM/3D-SIM revealed intense septal labeling consistent with Lipid II localization and nascent PG synthesis. Probes 7 and 9 showed strong septal enrichment relative to lateral walls; Van-BODIPY labeled more uniformly with more intracellular signal. Cross-section profiles showed PG probe peak outside membrane dye peak, allowing spatial discrimination of PG vs membrane. Co-labelling with TMP- and linezolid-derived probes confirmed intracellular penetration of those probes relative to PG localization. - Gram-negative OM permeability: Probe 9 generally did not label wild-type E. coli, but strongly labeled OM-impaired mutants: DC2 (periplasmic osmoregulation defect) and lpxC (Lipid A biosynthesis), with MICs of vancomycin and probe 9 at 64 µg/mL for lpxC mutant. O-antigen mutant (waaL) showed no increased labeling vs parent. Cold stress (15 °C) induced substantial labeling at the septum and PG vs minimal at 37 °C, indicating increased OM permeability. - Quantitative OM damage assays: Both flow cytometry and plate reader methods detected concentration-dependent increases in probe 9 labeling after exposure to membrane-active agents (tachyplesin-1, arenicin-3, polymyxin B, colistin, citropin, octapeptin C4), with no increase for non-membrane-active antibiotics. - Single-cell microfluidics and modeling: Polymyxin B markedly increased the fraction of E. coli accumulating probe 9. Kinetic parameters showed significant changes with polymyxin B: earlier accumulation (mean t0 6750 s vs 9374 s), higher uptake rate (k1 0.19 vs 0.01 a.u./s^2), and higher saturation (Fmax 149 vs 67 a.u.), all with p < 0.0001.

The probes address the need for mechanism-specific, activity-retaining fluorescent tools. By attaching small, minimally charged fluorophores via a modular azide at the vancomycin C-terminus, the constructs preserved or enhanced antibacterial potency compared to the parent antibiotic and outperformed some commercial probes in selectivity. Their strong septal enrichment confirms specific binding to nascent PG and provides a spatial reference layer that complements intracellular antibiotic probes for co-labelling. The probes’ inability to label intact Gram-negative bacteria but robust labeling upon OM compromise (genetic mutations, cold stress, or membrane-active agents) directly links fluorescence to OM permeability, enabling practical assays (flow cytometry, plate reader) and single-cell microfluidic quantification of OM damage. These findings demonstrate that vancomycin-based probes can both visualize Gram-positive PG dynamics and quantify Gram-negative OM perturbation, thereby aiding antibiotic mode-of-action studies and potentially diagnostics.

Modular azido-vancomycin intermediates enabled rapid synthesis of small-fluorophore probes (NBD, DMACA) that retain antimicrobial activity and selectively label Gram-positive bacteria. The probes clearly visualize nascent PG at the division septum with super-resolution microscopy and serve as spatial references for co-labelling with intracellular antibiotic probes. In Gram-negative bacteria, the probes report OM integrity: they label cells only when the OM is compromised (by genetic mutations, temperature stress, or membrane-active compounds), with quantification feasible via flow cytometry, plate reader, and single-cell microfluidics with kinetic modeling. Future work should extend validation across broader bacterial species and resistance phenotypes and expand the fluorophore palette to facilitate multiplexed imaging.

Selectivity and performance were assessed on a limited set of Gram-positive and Gram-negative species and strains; broader validation is needed. The DMACA probe (8) exhibited lower selectivity than NBD-based probes. Some comparisons with commercial probes involved different solvents (water vs DMSO), which may influence staining. Gram-negative labeling depends on OM compromise, so results reflect permeability under specific experimental conditions rather than general uptake.