Molecular Mechanisms of Cationic Fusogenic Liposome Interactions with Bacterial Envelopes

Although fusogenic liposomes offer a promising approach for the delivery of antibiotic payloads across the cell envelope of Gram-negative bacteria, there is still a limited understanding of the individual nanocarrier interactions with the bacterial target. Using super-resolution microscopy, we characterize the interaction dynamics of positively charged fusogenic liposomes with Gram-negative (Escherichia coli) and Gram-positive (Bacillus subtilis) bacteria. The liposomes merge with the outer membrane (OM) of Gram-negative bacteria, while attachment or lipid internalization is observed in Gram-positive cells. Employing total internal reflection fluorescence microscopy, we demonstrated liposome fusion with model supported lipid bilayers. For whole E. coli cells, however, we observed heterogeneous membrane integrations, primarily involving liposome attachment and hemifusion events. With increasing lipopolysaccharide length, the likelihood of full-fusion events was reduced. The integration of artificial lipids into the OM of Gram-negative cells led to membrane destabilization, resulting in decreased bacterial vitality, membrane detachment, and improved codelivery of vancomycin—an effective antibiotic against Gram-positive cells. These findings provide significant insights into the interactions of individual nanocarriers with bacterial envelopes at the single-cell level, uncovering effects that would be missed in bulk measurements. This highlights the importance of conducting single-particle and single-cell investigations to assess the performance of next-generation drug delivery platforms.


2
Mander's colocalization analysis between cytoplasmic DNA and Rh- PE  7   Spearman colocalization analysis between Vancomycin-BODIPY and Rh- PE  7   Degree of coverage analysis between Gram-negative Bacteria and cFLs 7 Modelled SIM data 7 Data visualization and statistical analysis 8

Dynamic light scattering (DLS) and zeta potential measurements
The average hydrodynamic diameter and zeta (ξ) potential of the liposomes were measured using a Zetasizer Nano ZSP (Malvern Panalytical) with an excitation wavelength of 633 nm and with a fixed scattering angle of 173°."Hydrodynamic size distribution measurements were conducted on 1 mL of liposomes with a concentration of 500 μM.Each sample was measured three times, with each measurement consisting of at least 12 runs.The output data included the average particle size distribution (Z average (d.nm)) and the polydispersity index (PDI), which measures the heterogeneity of the particle size distribution.Data are presented as mean ± standard deviation and were calculated from a number of repeats of three independent experiments.
Zeta potential measurements were performed on 1 mL of samples using DTS1070 cells.Liposomes were diluted in 1X PBS to a final concentration of 500 uM.The overnight culture of bacteria was 1:100 diluted in fresh LB media and grown to an OD600 of 0.5 before being washed three times through centrifugation at 6.000 rpm and resuspended in 1X PBS.
Each sample was measured three times, with each measurement consisting of at least 12 runs.Data are presented as mean ± standard deviation and were calculated from a number of repeats of three independent experiments.
Bacterial culture and staining

Supported lipid bilayer formation
Small vesicles for the formation of Supported Lipid Bilayers (SLBs) were prepared using the thin film hydration technique, as previously described. 1E. coli total lipid extract and 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-(7-nitro-2-1,3-benzoxadiazol-4-yl) (ammonium salt) (NBD-PE), reconstituted in chloroform, were purchased from Avanti Polar Lipids and stored at -20ºC.For fluorescent labeling at 488 nm excitation, 0.25 mole % NBD-PE was added to the E. coli Total Lipid Extract.The liposomes were prepared as previously described (see "Liposome preparation" section), resulting in liposomes in 1X PBS with a lipid concentration of 4 mg/mL.The samples were stored at 4°C for up to two weeks.
For the formation of the SLBs, glass coverslips (Academy, 22x40 mm, 0.16-0.19mm thick) that were cleaned with acetone and isopropanol or µ-slide 8 well high glass bottom (Ibidi) were incubated with 200 μL of 2 mg/mL E. coli liposomes for 30 minutes under continuous shaking at room temperature.The samples were then washed three times with 1x PBS.The procedure was repeated once to fill holes in the SLB.The bilayer was allowed to rest for approximately 30 minutes after washing to facilitate healing from potential disruptions during the washing procedure.AFM images were acquired to confirm the formation of the SLB (see Supplementary Figure 1).

Atomic force microscopy
Atomic force microscopy (AFM) images of cFLs and E. coli SLBs were acquired using a BioScope Resolve microscope (Bruker) in scanasyst mode, employing Peakforce HIRS-FA probes (Bruker) with a nominal spring constant of 0.35 N m−1 and a resonant frequency of 165 kHz.cFLs at a concentration of 10 nM were deposited onto freshly cleaved mica substrates and incubated for 10 minutes before being washed with 3 mL of 1X PBS.AFM images were recorded at scan speeds of 1 Hz, with tip-sample interaction forces ranging between 100 and 300 pN.The recorded topography data were first-order flattened using Nanoscope analysis software 2.0 (Bruker) before height measurements on the bilayers were conducted by capturing cross-sections across various regions of interest.To quantify the liposome spreading area, AFM images were converted to 8-bit images using Fiji (ImageJ) and then thresholded to select the bilayer patches.The liposome spreading area was determined using the built-in particle analyzer.The image processing pipeline is illustrated in Supplementary Fig. 7.The microscopy experiments were performed in a temperature controlled room at 20 °C.

Structured Illumination Microscopy
1 mL of a bacterial suspension (OD600 of 0.5) was incubated with liposomes at varying concentrations, as indicated in the main text, in a shaking incubator at 37ºC for 15 minutes.For co-delivery experiments involving Vancomycin and BODIPY™ FL Conjugate (Invitrogen), bacteria were incubated with 10 µM cFLs and 10 µg/mL Vancomycin in 1X PBS.
After the incubation, the cell suspensions were vortexed, centrifuged (6000 rpm, 2 minutes), and then resuspended in 1x PBS.The bacteria were deposited (1 µL) onto a glass slide (Academy, 22x40 mm, 0.16-0.19mm thick) and immobilized beneath a custom-made 5 mm x 5 mm agarose pad (1% w/v).The agarose pad was covered with a glass coverslip to prevent drying before imaging.
Fluorescence microscopy images were acquired using a 3-colour structured illumination microscopy (SIM) technique. 260x 1.2 NA water immersion lens (UPLSAPO 60XW, Olympus) focused the structured illumination pattern onto the sample and captured the emitted fluorescence light, which was then projected onto an sCMOS camera (Orca-flash 4.0, Hamamatsu).The excitation wavelengths used were 488 nm (iBEAM-SMART-488, Toptica) for bacteria, 561 nm (OBIS 561, Coherent) for Rh-PE labeled liposomes, and 640 nm (MLD 640, Cobolt) for membrane-stained bacteria.Image acquisition was performed using previously described custom SIM software.3 SIM reconstructions were carried out using the open-source software FairSIM, 4 following best practices for parameter selection.5 Each pixel measured 107 nm during acquisition, corresponding to 53.5 nm after reconstruction.The microscopy experiments were performed in a temperature controlled room at 20 °C.

Total internal reflection fluorescence Microscopy
Time-lapse Total Internal Reflection Fluorescence (TIRF) microscopy images were acquired for E. coli Supported Lipid Bilayers (SLBs) and whole bacteria exposed to cFLs.The SLBs were prepared in Ibidi wells, as previously reported in "Supported Lipid Bilayer formation".The E. coli cells were immobilized on 0.1% (w/v) poly-L-lysine (PLL) coated Ibidi wells.The Ibidi wells were subjected to O2 plasma treatment before the PLL suspension was incubated for 15 minutes.
The surface was then washed with deionized water (DI H2O) and dried using continuous nitrogen flow.GFP-labeled bacteria were prepared in 1X PBS at an optical density (OD) of 0.5 and incubated on the PLL-coated glass surface for 10 minutes before being washed with 1X PBS.Subsequently, 200 µL of 1X PBS was added to either the E. coli SLB-coated wells or BL21 cell-immobilized wells.Initially, the 488 nm excitation was used to focus the SLB or bacteria into the imaging plane, following which the 561 nm excitation channel was used to record the addition of 100 nM cFLs.
Time-lapse TIRF-M experiments involving liposome fusion were conducted using a custom-built microscope based on a microscope frame (IX-73, Olympus), equipped with a 100x 1.49 NA oil objective lens (UAPON100XOTIRF, Olympus).
The excitation wavelengths employed were 488 nm (Sapphire LDP, Coherent) for SLBs or bacteria and 561 nm (Jive 05-01, Cobolt) for Rh-PE labeled liposomes.The samples were imaged in TIRF mode, and the fluorescence was captured using an electron-multiplying CCD (iXon Ultra, Andor).The pixel size was measured to be 117 nm.TIRF recordings were acquired at an exposure time of 34 ms for SLB experiments using an EM gain of 200 over a 512x512 pixel region.
TIRF experiments involving bacteria were conducted with exposure times of 10 ms over a 256x256 pixel region or 22 ms over a 512x512 pixel region.The microscopy experiments were performed in a temperature controlled room at 20 °C.

Cell vitality assay
E. coli cells resuspended in PBS were prepared as previously described.200 µL of the bacteria was transferred into flatbottomed 96-well plates and exposed to increasing concentrations of cFLs (0 µM, 1 µM, 10 µM, 100 µM) for 30 minutes at 37 °C with continuous shaking.Subsequently, 20 µL of the 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) reagent (Abcam) was added to each well and incubated at 37 °C for 3 hours, according to the manufacturer's recommendation.The spectrophotometric absorbance of the samples was collected at 490 nm with a FLUOstar Omega plate reader (BMG Labtech).Data are presented as mean ± standard deviation and were calculated from a number of three repeats of three independent experiments.

TIRF-M analysis of Fusion Events
Fusion events were manually detected in the raw video files as appearances of high-intensity signals in the liposome fluorescence channel that colocalized with a bacterium or the SLB plane.By selecting a 3x3 pixel area around the maximum intensity pixel (i.e.supposedly the attachment side), the normalised maximum fluorescence profiles over time have been extracted.The time-point of attachment was set to T=0 and the following 6 seconds were extracted for further analysis.For each fusion event, the following equation was fitted to the data: where I describes maximum fluorescence intensity and t is time, in order to extract the diffusion exponent λ and fluorescence plateau .For further distinction between full-fusion and hemifusion events, we used a threshold of 0.5 for  (hemifusion events being characterised as  ≥ 0.5).

Colocalization analysis Mander's colocalization analysis between cytoplasmic DNA and Rh-PE
To quantify and visualise colocalization between the cFL and Gram-positive bacteria, we use an adjusted Mander's colocalization coefficient (MOC) of the cFL channel with the Gram-positive bacteria channel.First, both channels are independently otsu thresholded.As we are only interested in positions where we find localised cFL, we only filter for those positions in the Gram-negative bacteria channel (i.e.  =  for  with   = 0 and   =   for  with   > 0) where   denotes the intensity of the Gram-negative bacteria channel at pixel  and   the intensity of the cFL channel at pixel , respectively.With this adjusted channel data, we then calculate the well established MOC as

Spearman colocalization analysis between Vancomycin-BODIPY and Rh-PE
To quantify and visualise colocalization between the Vancomycin-BODIPY and Rh-PE, we use the Spearman rank correlation coefficient (SPCC).After otsu thresholding each channel, we then calculate the SPCC as where  is the number of pixels of the image and   describes the difference in rank (ranked regarding pixel intensity) of the two channels for each pixel .

Degree of coverage analysis between Gram-negative Bacteria and cFLs
To analyze the extent of liposome coverage on Gram-negative BL21 cells, we determined the overlap between the bacterial outline and the liposomal Rh-PE signal.To achieve this, the reconstructed two-channel SIM images of the cytoplasmic GFP and liposomal Rh-PE were auto thresholded using Fiji 6 .Next, the outline of the cytoplasmic GFP signal was determined using the particle analyzer in Fiji on the thresholded GFP signal.The percentage of overlap between each outline and the thresholded Rh-PE signal was calculated using a custom-written Python code.The image processing pipeline is displayed in Supplementary Fig. 8.

Modelled SIM data
For comparison of the experimentally acquired SIM data to a capsule model of the bacteria, artificial SIM images were generated from the approximate dimensions of the bacteria being imaged.To achieve this, SIM images of the cytoplasmic dye were first binarized and segmented to extract the length and width of the cross section.From these dimensions, an ideal stadium shape was created and rotated around the major axis to form the capsule model of the bacteria.The surface of this 3D volume was used as a model for the membrane and the interior of the volume was used as a model for the cytosol.
From the modelled 3D structures, artificial SIM images were generated by convolving the volume with a model point spread function (PSF) for the microscope used.This PSF was itself approximated from a basic model of the detection PSF and the SIM reconstruction parameters as reported from the SIM reconstruction software used. 7Modelling the SIM process in this way is important as the reconstruction process imparts optical sectioning to the images which is not accounted for in a simplistic widefield model of the imaging system. 8Line profiles across the minor axis of the resulting SIM images were then compared to the corresponding line profiles of the experimentally acquired SIM images.All image analysis and modelling was performed in MATLAB.The code underlying this work can be found on the GitHub repository: https://github.com/edward-n-ward/bacteria-modelling.