A Modular Workflow for Quantitative, Structural and Functional Analysis of Leptospira Biofilms

Biofilms are structured microbial communities in which cells are embedded in a self-secreted extracellular polymeric substance (EPS) matrix composed of polysaccharides, proteins, nucleic acids and lipids¹. This matrix provides mechanical stability, mediates adhesion to surfaces and enhances microbial survival by conferring tolerance to environmental stresses such as desiccation, oxidative stress, and antimicrobials^2,3. In pathogenic bacteria biofilms facilitate persistence, immune evasion, and chronic infection^4,5.

Compared to planktonic cells, which are free-floating and metabolically more homogeneous, biofilm-associated cells exhibit altered gene expression, growth rates, and metabolic activity⁶. These differences confer enhanced tolerance to environmental challenges, antibiotics, and host defenses, while allowing coordinated behaviors such as quorum sensing, nutrient retention, and spatial organization^7,8.

Biofilm formation has been extensively characterized in model organisms such as Pseudomonas aeruginosa, Staphylococcus aureus, and Vibrio cholerae, which have collectively shaped our understanding of the genetic, structural, and physiological bases of the biofilm lifestyle. Studies on Pseudomonas have revealed that exopolysaccharides and quorum sensing work together to shape complex three-dimensional biofilm architectures^9,10. Research on Staphylococcus has emphasized the roles of surface proteins and extracellular DNA in enhancing biofilm cohesion and antibiotic tolerance^11,12. Investigations into Vibrio species further highlighted the remarkable diversity of biofilm morphologies and their ecological significance in aquatic environments¹³. Collectively, these systems have provided a robust conceptual and methodological framework for biofilm research, reinforced by standardized protocols¹⁴ and critical evaluations of existing techniques^15,16, which together highlight the need for harmonized approaches when investigating biofilm formation in less-studied bacteria such as Leptospira.

Members of the spirochete genus Leptospira, the causative agents of leptospirosis, were long presumed to be predominantly planktonic. Recent studies have experimentally demonstrated robust biofilm formation across multiple species¹⁷. While some studies suggest biofilm formation in environmental¹⁸ and host-associated¹⁹ contexts, others have experimentally demonstrated the ability of leptospires to form biofilm in the renal tubules of Rattus norvegicus²⁰ and the vitreous humor of horses²¹, as well as detecting leptospiral species within environmental biofilms^22,23. Deciphering the conditions and mechanisms that govern Leptospira biofilms is therefore critical to understanding environmental transmission, reservoir persistence, and disease pathogenesis^24,25.

Existing Leptospira biofilm assays are fragmented, ranging from qualitative microscopic observations^17,26 to endpoint colorimetric or crystal violet stains^27,28. While these studies have provided valuable insights into biofilm formation, they often lack both the kinetic resolution required to monitor biofilm maturation and dispersal in real time and the ultrastructural context provided by confocal or electron microscopy. Continuous monitoring and 3D structural analysis are considered essential to capture the dynamic nature of biofilm formation and dispersal^14,15. These limitations hamper comparability across studies and limit the systematic evaluation of factors or interventions affecting biofilm formation and dispersal, highlighting the need for a standardized, multi-readout workflow that captures both structural and functional dynamics of Leptospira biofilms in a reproducible and comprehensive manner.

To address these gaps, we developed an integrated, modular workflow that unifies six complementary readouts drawn from the same culture series. First, a high-throughput CV microtiter assay quantifies attached biomass at multiple time points. Second, a time-resolved fractionation assay in 96-well plates partitions total, liquid-phase (non-attached or partially detached), and attached (biofilm) populations by OD₄₀₅ to track their evolution during formation and dispersal. The term liquid-phase is used here to encompass both planktonic-like and detached cells, acknowledging the dynamic continuum between free-living and surface-associated phenotypes^29,30. Third, time-lapse phase-contrast imaging provides non-destructive kinetics of adhesion, aggregate motility, and coalescence. Fourth, confocal laser-scanning microscopy (CLSM) yields full 3D reconstructions with live/dead and matrix-probe readouts, enabling depth-dependent viability and matrix mapping. Fifth, membrane- or coverslip-supported scanning electron microscopy (SEM) resolves extracellular architecture and basal-apical polarity at high resolution. In addition, an in vivo module was incorporated to evaluate virulence using intraperitoneal injection of biofilm aggregates into the susceptible golden Syrian hamster model, a sensitive and well-established model for leptospirosis^31,32,33. This approach links biofilm phenotypes to pathogenic potential, providing functional context that complements in vitro analyses.

Compared with single-modality approaches, this platform offers several advantages: (i) synchronized quantitative (CV and fractionated OD) and structural (CLSM/SEM) readouts; (ii) non-destructive kinetic monitoring of early adhesion, growth, maturation, and dispersal; (iii) 3-D viability and matrix characterization using targeted probes; (iv) ultrastructural visualization of extracellular matrix architecture ; and (v) direct functional assessment of biofilm-associated versus planktonic virulence in a susceptible hamster model, each achievable with standard BSL-2 infrastructure. By leveraging multi-well formats and removable glass or polycarbonate substrates, the workflow supports replicate screens of environmental variables, antimicrobial compounds, or mutant libraries while maintaining sterility, throughput, and cross-validation across modules.

Laboratories focused on ecology, persistence, host-pathogen interactions, or anti-biofilm discovery can adopt individual modules or the full pipeline. Required resources are limited to a plate reader, an inverted microscope with environmental control for time-lapse, access to a confocal microscope and SEM facility, and standard animal facilities for the hamster model, enabling broad adoption and reproducible comparisons across studies.

Ethics statement:
This research was approved by the Animal Care and Use Committee of the Institut Pasteur of New Caledonia and conducted according to the guidelines laid out by the Animal Care and Use Committee of the Institut Pasteur of Paris, and European Recommendation 2007/526/EC. Animal Research Experiment Registration Number: IPNC-2018-ARE-001

NOTE: All procedures involving live Leptospira cultures must be conducted in a biosafety level 2 (BSL-2) laboratory, in compliance with institutional biosafety guidelines. All culture manipulations must be performed under a biological safety cabinet to ensure operator safety and prevent environmental contamination.

The Leptospira interrogans serovar Manilae strain L495 used in this study was originally obtained from the Institut Pasteur collection (Paris, France) and is maintained at the Institut Pasteur of New Caledonia. To preserve virulence, the strain is periodically re-isolated from infected hamsters. For all in vitro experiments, cultures were not maintained beyond six subcultures after recovery from the animal host.

All animal procedures must be performed in compliance with institutional guidelines and approved by the appropriate Animal Care and Use Committees. Experiments should adhere to European regulations on animal welfare (EU Directive 2010/63) and to the recommendations of the Public Health Service, ensuring that the use of animals is both justified and ethically regulated.

1. Preparation of the bacterial inoculum

1. Grow Leptospira spp. cells at 30 °C in EMJH media³⁴ under aerobic conditions and without shaking in flat-bottomed, screw-cap glass tubes until the culture reaches mid-log phase. Under these conditions, L. interrogans serovar Manilae strain L495, a low-passage strain regularly maintained in vivo through hamsters, typically reaches the appropriate cell density within 3-5 days. Ensure cultures reach an O.D. of 0.2-0.4 at 405 nm (2 to 5 x 10⁸ cells/mL) before dilution in fresh EMJH media and verify motility and lack of clumping by dark-field microscopy at 20 x magnification.
   NOTE: EMJH has intrinsic absorbance. Blank the spectrophotometer with sterile EMJH and subtract this value from all readings. The inoculum can be standardized either by measuring optical density or by direct cell counting using a Petroff-Hausser chamber. A 1:100 dilution of a verified mid-log culture typically yields OD₄₀₅ = 0.02 (≈ 1 x 10⁶ cells/mL). Mix gently by inversion; do not vortex.
2. Examine a 10 µL aliquot under dark-field microscopy at 20 x magnification. Ensure that ≥ 90% of cells are highly motile and that no visible clumps are present. Dilute the verified mid-log culture 1: 100 in fresh EMJH to obtain OD₄₀₅ = 0.02 (≈ 1 x 10⁶ cells/mL). Mix gently by inversion; do not vortex.
   NOTE: One working inoculum supports all downstream assays detailed in this protocol (Figure 1). Prepare 36 mL to seed a 24-well plate (1 mL/well) or 20 mL to seed a 96-well plate (200 µL/well). Adjust volumes to match the number of plates required.

Figure 1. Global workflow for Leptospira biofilm analysis. Exponentially growing Leptospira cultures are first assessed for growth and standardized by optical density. Cultures are then distributed into glass- or membrane-containing plates, microscopy microdishes, or 96-well plates. The workflow includes seven modules: (1) inoculum preparation; (2) crystal violet staining for biomass quantification; (3) ultrastructural imaging by SEM; (4) dynamic biofilm monitoring by time-lapse phase-contrast microscopy; (5) 3D visualization by CLSM with live/dead staining; (6) time-resolved quantification of attached and non-attached fractions during biofilm formation via optical density ; and (7) investigation of the biofilm functional role during infection in Syrian hamsters. This integrated approach allows multi-modal analysis of biofilm development, structure, and pathogenicity. Please click here to view a larger version of this figure.

2. Crystal Violet - Based Quantification of Biofilms on Coverslips or Membranes

1. Inside a BSL2 biosafety hood, use sterile forceps to lay one 12 mm sterile glass coverslip or one 0.1 µm sterile hydrophilic polycarbonate membrane flat on the bottom of each well in a sterile 24-well plate with lid. Pre-soak the membrane/ glass coverslip for 2 h at 30 °C with 1 mL of sterile EMJH.
   NOTE: Although not essential, pre-soaking the substrate in EMJH improves wetting and slightly enhances bacterial adhesion.
2. Remove the soaking solution and add 1.5 mL of the diluted bacterial suspension (OD₄₀₅ = 0.02 ≈ 1 x 10⁶ cells/mL) to each well. Ensure that the coverslip or membrane remains firmly positioned at the bottom.
3. Incubate the plate at 30 °C under static conditions in a humid atmosphere, maintained by placing a water-filled tray inside the incubator to prevent evaporation of the culture medium. For slow growing strains, a 3-week incubation is recommended to allow the formation of a mature, adherent biofilm.
4. At the desired timepoint, carefully remove as much culture medium as possible from each well without disturbing the biofilm. Gently rinse each well with 1 mL of sterile phosphate-buffered saline (PBS) to remove residual non-adherent cells, keeping the coverslip flat against the well bottom to avoid detaching fragile biofilm cells. Under these conditions, biofilm growth occurs predominantly on the upper surface of the coverslip, with minimal growth on the underside, so rinsing without lifting is sufficient to enrich for surface-attached cells for downstream analyses.
   NOTE: Biofilms are extremely fragile at this stage and can easily be aspirated by mistake. Pipetting should be performed slowly and precisely along the wall of the well to avoid disrupting the biofilm.
5. Fix biofilm samples by adding 1 mL of 4% paraformaldehyde (PFA) in PBS at 37 °C for 30 min. Remove the fixative and carefully rinse twice with 1 mL PBS.
   NOTE: For fixation, a 4% formaldehyde solution was freshly prepared (16% methanol-free stock diluted 1:4 in PBS) and discarded after use, since polymerization into paraformaldehyde reduces cross-linking activity.
6. Add 1 mL of 0.1% (w/v) Crystal Violet (CV) solution to each well and incubate at room temperature for 15 min, ensuring that the membrane or coverslip is fully covered.
7. Discard the dye and rinse twice with 1 mL PBS.
   NOTE: Crystal violet and paraformaldehyde (PFA) are toxic and can irritate the skin and eyes. Always wear gloves and handle both chemicals with care, preferably in a fume hood. Dispose of waste in designated hazardous-dye or chemical-waste containers according to institutional safety guidelines.
8. Tilt the plate and drain residual liquid. Air-dry the wells at room temperature until the substrate appears completely dry (≥ 4 h, preferably overnight).
   NOTE: At this stage, dot-like or reticulated structures may be visible on the coverslip or membrane surface (Figure 2A).
9. Add 500 µL elution buffer (50% ethanol, 50% glacial acetic acid (vol/vol)) to each well. Incubate the plate for 15 min. Pipette up and down to fully dissolve the crystal violet bound to the biofilm.
10. Transfer 200 µL of each sample into an optically clear 96-well microplate and measure absorbance at 570 nm. Subtract a substrate-only blank prepared by processing an uninoculated coverslip or membrane through fixation, CV staining, washes, and elution to remove intrinsic binding/background. Record the mean ± standard deviation for at least three technical replicates. Dilute samples with elution buffer if the absorbance exceeds the linear range of the spectrophotometer.

3. Biofilm visualization using Scanning Electron microscopy (SEM)

1. Inside a BSL2 biosafety hood, use sterile forceps to lay one 12 mm sterile glass coverslip or one 0.1 µm sterile hydrophilic polycarbonate membrane flat on the bottom of each well in a sterile 24-well plate with lid. Pre-soak the membrane/ glass coverslip for 2 h at 30 °C with 1 mL of sterile EMJH.
2. Remove the soaking solution and add 1.5 mL of the diluted bacterial suspension (OD₄₀₅ = 0.02 ≈ 1 x 10⁶ cells/mL) to each well. Ensure that the coverslip or membrane remains firmly positioned at the bottom.
3. Incubate the plate at 30 °C under static conditions in a humidity control incubator to prevent evaporation of the culture medium. For slow growing strains, a 3-week incubation is recommended to allow the formation of a mature, adherent biofilm
4. At the desired timepoint, carefully remove as much culture medium as possible from each well without disturbing the biofilm.
5. Then gently rinse each well with 1 mL of sterile PBS to remove residual non-adherent cells, keeping the coverslip flat against the well bottom to avoid detaching fragile biofilm cells. Under these conditions, biofilm growth occurs predominantly on the upper surface of the coverslip, with minimal growth on the underside, so rinsing without lifting is sufficient to enrich for surface-attached cells for downstream analyses.
   NOTE: Biofilms are extremely fragile at this stage and can easily be aspirated by mistake. Pipetting should be performed slowly and precisely along the wall of the well to avoid disrupting the biofilm.
6. Add a solution of 4% paraformaldehyde (PFA) and 1% glutaraldehyde in sodium cacodylate buffer (0.2 M, pH 7.4) directly into the well to fix the biofilm.
7. Incubate for 30 min at 37 °C then remove the fixative and rinse the coverslip or membrane twice with PBS. This approach preserves the surface-attached biofilm while minimizing detachment, as most biofilm growth occurs on the upper surface of the coverslip under our conditions.
   NOTE: For fixation, a 4% formaldehyde and 1% glutaraldehyde solution was freshly prepared (16% methanol-free stock diluted 1:4 in sodium cacodylate buffer) and discarded after use, since polymerization into paraformaldehyde reduces cross-linking activity.
8. Immerse the substrate in 1% osmium tetroxide (OsO₄) diluted in PBS for 1 hour to enhance SEM contrast. Rinse twice with PBS.
9. Dehydrate the samples by immersing them sequentially in graded ethanol series of increasing concentration: 25%, 50%, 70%, 90%, and 100% (v/v), each for 10 min.
10. Add 500 µL of hexamethyldisilazane and incubate for 5 min, then replace with fresh HMDS and incubate an additional 5 min. Remove excess HMDS and allow the sample to air-dry completely under the fume hood.
    NOTE: Glutaraldehyde, PFA, sodium cacodylate, osmium tetroxide and hexamethyldisilazane are toxic and must be handled under a chemical fume hood with appropriate personal protective equipment.
11. Mount the dried samples on SEM stubs using double-sided conductive carbon tape. Sputter-coat the samples with a thin layer (~10 nm) of gold or platinum, providing sufficient electron contrast for SEM imaging. This allows high-resolution visualization of biofilm architecture, including cell-cell contacts, extracellular matrix morphology, density, and heterogeneity, without the need for additional dyes or stains.
12. Load the stubs into the SEM using the appropriate sample holder. Evacuate the chamber overnight, when possible, to improve imaging quality, then acquire secondary electron images at 5 - 15 kV and suitable magnifications to reveal biofilm ultrastructure (Figure 2).

4. Time-lapse phase-contrast imaging of growing biofilm

1. Pipette 500 µL of the working inoculum (OD₄₀₅ = 0.02 ≈ 1 x 10⁶ cells/mL; see Section 1) into a sterile 35 mm glass bottom Hi-Q4 dish equipped with a plane-parallel lid to eliminate meniscus distortion. Avoid introducing bubbles and close the dish immediately.
   NOTE: 35 mm glass bottom Hi-Q4 dish contains 4 distinct culture areas that can be used to simultaneously image 4 different conditions. Allocate one compartment to sterile EMJH medium as a negative control and the 3 others for technical replicates or different strains/mutants.
2. Place the dish on the environmental stage of an inverted microscope fitted with phase-contrast optics, a motorized focus drive (Biostation IMQ), and a humidified enclosure set to 30 °C and 95% relative humidity. These conditions help minimize evaporation during extended imaging (up to 8 days).
3. Launch the BioStation IMQ application, and in the main interface, open the New Time-lapse Setting window to access the configuration panel. Before starting the experiment, check the environmental parameters on the Status Display, ensuring that both the Chamber and Water temperature indicators show Stable to prevent frame drift during acquisition.
4. On the New Time-lapse Setting screen, select the Live tab, and configure the imaging parameters in the Observation Condition panel by choosing Phase Contrast (Ph) as the filter and setting the Magnification to 20 X.
5. Under Manual Mode, adjust the Light Intensity and Exposure Time to optimize contrast while avoiding saturation. Verify this using the Saturation Check button. Define four acquisition points corresponding to the centers of the four compartments.
6. Position the stage sequentially over each compartment using the Jog Dial or by clicking directly within the Live Observation Image Display.
7. Refine the focus with the Focus Buttons, and record each position by clicking the Time-lapse Experiment Point Registration button. The registered points appear as numbered blue frames in the Observation Point Verification Display and are confirmed in the Points Tab.
8. Program the acquisition schedule by opening the Time Tab and clicking New to access the Timelapse Dialog Box, where the Acquisition Cycle is set to 30 min and the Total Time to 7 days.
9. After clicking Apply, the software automatically calculates the number of acquisition rounds. Activate Autofocus by right-clicking the title bar, selecting Preferences, and enabling AF Mode to ensure refocusing at each stage position. Verify the consistency of Exposure Settings, Gain, and Light Intensity across all points, then save the entire configuration using the Save button.
10. Start the acquisition by clicking Start Time-lapse. The software automatically switches to the Time-lapse Images, allowing continuous monitoring of the acquisition progress and chamber stability throughout the 7-day experiment. All images are automatically saved in TIFF format within the experiment folder created by the software.
11. Process and quantify the image series by importing the image stacks into FIJI (Figure 3A, B, C).
12. Open the image stacks corresponding to each position via File > Import > Image Sequence, ensuring that frames are loaded in chronological order.
13. Convert the aligned stacks to 8-bit grayscale using Image > Type > 8-bit to standardize pixel intensity.
14. Apply thresholding using Image > Adjust > Threshold to isolate bacterial biofilm regions. Apply consistent threshold settings across all frames by selecting Apply, then save the result as a binary mask using Process > Binary > Make Binary.
15. Perform quantification with Analyze > Analyze Particles, recording parameters such as Area, Percent Area, and Mean Intensity.
16. Export the results from each frame automatically to a spreadsheet via the Results window (File > Save As > .csv) for kinetic analysis. Express all measurements as the proportion of the field area covered by biofilm (surface coverage).

5. Biofilm visualization using Confocal Scanning Laser microscopy (CLSM)

1. Inoculate 1.5 mL of the working inoculum (OD₄₀₅ = 0.02 ≈ 1 x 10⁶ cells/mL; see Section 1) into a sterile 35 mm glass-bottom dish. Incubate statically in a humidified box containing a water reservoir in an incubator at 30 °C until the desired developmental stage is reached (up to 21 days).
2. At the chosen timepoint, carefully aspirate the medium slowly along the dish wall without disturbing the biofilm. Rinse once with 2 mL of sterile PBS to remove planktonic cells taking extreme caution to not detach or disrupt the biofilm.
3. Prepare staining solutions that require incubation with unfixed (live) biofilms (perform before fixation).
   1. For live/dead staining, add 3 µL SYTO9 green fluorescent nucleic acid stain (1.67 mM) and 3 µL propidium iodide (18.3 mM) to 10 mL PBS to create staining solution. Add 200 µL of the staining solution directly onto the biofilm forming bacteria, then incubate for 30 min at room temperature in the dark, and rinse once in PBS.
   2. Live cell tracer can also be used to stain the cells prior fixation. Incubate live cells with a 10 µM CFDA/SE tracer for 30 min. Alternatively label the membranes with FM4-64 at a concentration of 5 µg/mL can also be used. Rinse once in PBS.
4. Fix the biofilm by adding 2 mL of a 4% paraformaldehyde solution in PBS and incubate for 30 min at 37 °C. Remove the fixative and rinse twice with PBS.
5. Add a drop of antifade mounting medium to cover the biofilm and avoid bubbles.
6. Acquire Z-stacks on an inverted confocal microscope equipped with a tunable white-light laser (WLL) and a HC PL APO CS2 63×/1.4 NA oil-immersion objective. For CFDA/SE, set the excitation to 488 nm and collect emission between 500-550 nm, using a pinhole of 1 Airy unit (AU). For FM4-64, set the excitation to 561 nm and detect emission between 630-700 nm, also with a 1 AU pinhole.
7. Acquire Z-stacks at 0.5-1.0 µm intervals throughout the entire biofilm thickness. Adjust laser power and detector gain to maximize the dynamic range while avoiding saturation (<1% saturated pixels).
8. Use line averaging (2-4×) to improve the signal-to-noise ratio, and maintain all imaging parameters (laser power, detector bandwidth, pinhole size, scan speed) constant across samples.
9. Save image stacks as 12-bit files and export them in .lif format for quantitative analysis in FIJI (ImageJ).
   NOTE: Before imaging, ensure that microscope settings (e.g., laser power, detector gain, pinhole size) are standardized using control samples stained with the same dyes. This calibration step is essential to minimize variability between acquisitions and to allow reliable comparison across experiments.
10. Process and quantify confocal image stacks using FIJI equipped with the COMSTAT plugin (or equivalent 3D analysis software)¹⁰. Measure biovolume, mean and maximum thickness, surface coverage, and live/dead ratios (or other predefined metrics). Export representative orthogonal views and maximum-intensity projections for illustrative purposes (Figure 3D, E).

6. Time-Resolved Quantification of Attached and Non-Attached Fractions during Biofilm Formation in 96-Well Microtiter Assays

1. Inoculate 200 µL of the working inoculum (OD₄₀₅ = 0.02 ≈ 1 x 10⁶ cells/mL; see Section 1) into the wells of a 96-well plate. Incubate statically at 30 °C until the desired time point (days 3, 5, 7, 10, 12, 14, 17, or 21).
   NOTE: Border effects and evaporation are common when using 96-well plates, especially if the incubator lacks humidity control. To minimize these effects, add 200 µL of sterile distilled water to all peripheral wells. In addition, plates were placed inside a humidified box containing a water reservoir, which is then positioned within the incubator to further reduce evaporation and maintain consistent humidity across wells.
2. At each time point, prepare three sample types:
   1. Total suspension - Homogenize the entire content of one well containing the biofilm by gently pipetting up and down several times, avoiding bubble formation. This fraction represents the total bacterial population, including both non-attached leptospires and those within the semi-adherent biofilm. This homogenization step helps to disperse cell aggregates that could interfere with subsequent absorbance measurements.
   2. Liquid phase - From a second well, carefully remove 180 µL of the supernatant without disturbing the biofilm layer. Transfer to an empty well and add 20 µL of fresh EMJH medium. Gently pipette up and down several times to disperse cell aggregates. This fraction represents the non-attached cells present in the liquid phase, which may include both freely suspended leptospires and detached biofilm aggregates.
   3. Biofilm - In the same well used for step 2.2, resuspend the remaining biofilm by adding 180 µL of fresh EMJH medium and gently pipetting up and down several times. This fraction corresponds to leptospires within the biofilm.
3. Measure the absorbance of each suspension at 405 nm using a spectrophotometer, after ensuring that each well has been thoroughly homogenized, and plot the values to generate a representative graph (Figure 4A).

7. Investigating the functional role of biofilm during host infection

1. Inoculate 200 µL of the working inoculum (OD₄₀₅ = 0.02 ≈ 1 x 10⁶ cells/mL; see Section 1) into the wells of a 96-well plate. Incubate statically at 30 °C until the desired developmental stage is reached (up to 21 days).
2. To determine the inoculum concentration, dedicate specific "biofilm control" wells that will be used only for growth monitoring and not for animal injection. Once the biofilm is macroscopically visible, gently remove 180 µL of supernatant from one control well without disturbing the biofilm. Replace with 180 µL of fresh EMJH medium, resuspend the biofilm aggregates carefully without generating bubbles, and measure the OD₄₀₅ to estimate bacterial concentration.
   NOTE: In principle, washing the wells prior to quantification could help remove non-adherent cells. However, because Leptospira biofilms exhibit weak adhesion in 96-well plates, washing would lead to uncontrolled loss of biofilm material. For this reason, the resuspension step was performed without prior washing to ensure reproducibility between wells and consistent bacterial concentrations.
   Biofilm control wells typically contain ~2 x 10⁸ leptospires, which was chosen as the standard inoculum for hamster infection experiments. This concentration also defines the target number of planktonic leptospires used as controls. When included, planktonic controls are prepared by freshly subculturing Leptospira in EMJH medium under shaking conditions at 30 °C for up to 5 days, corresponding to the mid- to late-exponential growth phase that yields a similar cell density.
3. For animal inoculation, use separate biofilm wells (not the control wells). Carefully remove 180 µL of supernatant from the wells to eliminate most liquid-phase leptospires.
4. Collect the remaining 20 µL containing biofilm aggregates using a 200 µL pipette tip to avoid disrupting the structure.
   NOTE: The biofilm is fragile. Ensure aggregates are visible before and after collection. Prepare extra wells in case repetition is required.
5. Transfer the aggregates into a syringe prefilled with 300 µL of fresh EMJH medium. Allow the aggregates to settle by gravity.
6. Attach a 21 G needle and gently expel excess EMJH until a final volume of 200 µL is reached. Ensure no air bubbles remain and that biofilm aggregates remain visible.
   NOTE: Waiting for aggregates to settle minimizes loss during volume adjustment. Prior to in vivo use, verify that aggregates remain intact after passage through a 21 G needle (Figure 4C, D).
7. Perform intraperitoneal injection of 2 x 10⁸ leptospires in 200 µL EMJH medium.
   NOTE: Use 7-8-week-old golden Syrian hamsters (both sexes), maintained under standard housing conditions. For negative controls, inject 200 µL EMJH medium only (one animal per replicate).
8. Monitor animals twice daily for up to 21 days for clinical signs of leptospirosis (ruffled fur, prostration, and reduced response to stimulation). Euthanize immediately via carbon dioxide inhalation if suffering is observed and record the euthanasia time as the time of death.
9. At day 21, euthanize all surviving animals.
   NOTE: Perform three independent biological replicates with cultures prepared on different dates. Each replicate includes two animals per condition and one negative-control animal.

When the inoculum is prepared correctly, cultures enter mid-log phase within 3-5 days and exhibit OD405 values of ~ 0.2-0.4 with bright, highly motile fields under dark-field microscopy (≥ 90% of cells motile) and no visible clumps. Suboptimal preparations present as sluggish bacteria, and heterogeneous motility across the field; such cultures frequently yield weak biofilms and should be discarded. In practice, confirming motility and OD side-by-side immediately before seeding minimizes failed runs. Establishing these quality control gates at the inoculum stage is the single best predictor of success across downstream applications. Although the methodology was optimized for L. interrogans serovar Manilae L495, the same procedures were also applied to Leptospira biflexa Patoc strain to demonstrate cross-species applicability. L. biflexa generally forms less cohesive and thinner biofilms compared to L. interrogans, yet the characteristic developmental sequence and structural features remain detectable with appropriate parameter adjustments. Including both species thus underscores the adaptability of the workflow to pathogenic and saprophytic Leptospira alike.

After 21 days of static incubation at 30 °C in a humidified chamber (or the appropriate duration for the Leptospira species used), biofilms become visible to the naked eye on both glass coverslips and hydrophilic polycarbonate membranes. Successful growth produces characteristic CV patterns after 2-3 weeks such as dot-like, branching, or reticulated footprints attached to the surface (Figure 2A).

In a typical run, wild-type L. interrogans Manilae L495 reaches ~ 50% surface coverage by week 3, while low-biofilm mutants' plateau near ~ 20% and high-biofilm phenotypes approach ~ 70-80%, establishing a practical dynamic range for screening. The extent of biofilm formation may also vary depending on the Leptospira species and strain; for instance, L. biflexa Patoc strain develops visible biofilms more rapidly, with previous studies considering structures as early as 120 h post-inoculation to be mature biofilms35.

Crystal violet staining provides a reliable and reproducible method for quantifying biofilm biomass. In well-developed biofilms, staining results in deep purple coloration localized to the coverslip or membrane area, indicating high levels of attached biomass (Figure 2A). Visual cues during the staining and solubilization steps also provide useful indicators of protocol success. For instance, uneven CV distribution or pale staining may be due to under-seeding, evaporation, or aggressive washing that detaches early biofilms.

Absorbance readings at 570 nm reflect the amount of retained stain, and therefore, the relative biofilm density (Figure 2B). In representative experiments, L. interrogans cultures grown under optimal conditions show consistent and reproducible OD₅₇₀ readings across replicates, reflecting stable biofilm formation. In contrast, large variations between replicates are often observed when samples undergo excessive pipetting during medium changes, indicating poor adhesion or partial biofilm detachment. Such variability should be considered a sign of technical issues, and the affected samples should be excluded or the protocol carefully reviewed. Notably, L. biflexa Patoc forms more extensive biofilms on both glass coverslips and polycarbonate membranes under similar conditions, which is reflected in higher CV absorbance values, demonstrating that the protocol is adaptable and effective across Leptospira species.

Successful SEM preparation is able to reveal extracellular matrix deposits as early as day 3, followed by a strikingly polarized architecture in mature biofilms: a rough, channeled basal face (often with > 5 µm channels) that anchors a porous inner structure, and a smoother apical face where spirochetes lie enmeshed in dense matrix (Figure 2C, D, E, F). High-magnification fields frequently capture branching extracellular filaments and occasional mushroom-like protrusions; features that correspond to the coalescence dynamics seen by time-lapse (Supplemental video 1). In suboptimal preparations, collapsed matrices, charging, and indistinct cell outlines may be observed, usually reflecting inadequate post-fixation, insufficient conductive coating, or poor drying. When basal-apical polarity and pervasive channels are evident, SEM readouts align closely with confocal estimates of thickness and porosity, confirming successful execution of both preparation and imaging.

In effective time-lapse series, isolated puncta appear within 24-72 h and progressively coalesce into larger aggregates that sweep across the surface before space limitations slow their motion (Figure 3A, B, C). Quantitative segmentation yields a monotonic increase in total covered area, while aggregate counts rise, peak, and then decline as collisions and mergers dominate between ~12 and 216 h. These kinetics (area up, aggregates number down), indicate active accretion rather than simple sedimentation. Failed or borderline runs lack early puncta, show flat area-over-time curves, or suffer focus drift linked to unstable temperature/humidity. Maintaining a stabilized 30 °C environment with 95% humidity and using autofocus at each time point typically restores clear trajectories suitable for comparing mutants or treatments.

Representative Z-stacks from mature biofilms show foam-like, multilayered architecture typically exceeding 50 µm in thickness, with most cells staining live (SYTO 9-positive) and occasional central voids indicative of collective rearrangements during growth (Figure 3D). Matrix probes can be used to investigate matrix composition: WGA highlights polysaccharide epitopes, BOBO-3 labels abundant extracellular DNA, and protein-selective stains contribute little outside cell-associated signal (Figure 3E). Suboptimal results include thin or discontinuous stacks dominated by propidium iodide, strong photobleaching, or inconsistent gain settings, each of which undermines quantitative comparability. Interpreting thickness, biovolume, and live/dead ratios together (while holding laser power, detector gain, and pinhole constant across conditions) confirms maturation status and supports direct comparisons with SEM ultrastructure and CV biomass.

The biofilm formation process of Leptospira typically follows distinct phases, though exact timings and values may vary depending on the species or strain. Using L. interrogans strain Manilae L495 as a reference, the expected progression over 21 days includes an initial phase (days 0-3) where bacteria remain mostly planktonic with minimal biofilm (Figure 4A). This is followed by an exponential growth phase (days 3-7) during which both planktonic and biofilm-associated bacteria increase, reaching around 9 x 10⁸ cells/mL, accompanied by the formation and expansion of biofilm aggregates. Between days 7 and 12, planktonic bacteria decline significantly, while biofilm-associated cells peak, representing roughly 80% of the population. Finally, during the maturation phase (days 12-21), biofilm bacterial numbers decrease without a rise in planktonic cells, yet biofilm size and complexity continue to grow. Observing this sequence of changes indicates that the protocol effectively captures the dynamic development and maturation of Leptospira biofilms.

When properly performed, the infection protocol uses inocula containing ~2 x 10⁸ leptospires in 200 µL EMJH, prepared from well-defined planktonic (5-day) or biofilm (21-day) cultures. Although these two culture types represent distinct physiological states, this difference is intentional, as the experiment aims to determine whether Leptospira biofilms -- characterized by reduced metabolic activity and structural differentiation -- retain the ability to initiate infection. This contrast is supported by transcriptomic studies showing major shifts in gene expression between planktonic and biofilm Leptospira35,36. Biofilm aggregates are carefully harvested to preserve their structure and remain intact after passage through a 21 G needle, as confirmed before injection (Figure 4C, D). Following intraperitoneal injection, golden Syrian hamsters typically display clinical signs of leptospirosis within 3 to 5 days (e.g., lethargy, ruffled fur, prostration). Negative controls injected with EMJH medium alone show no signs of disease. The progression and severity of clinical signs, as well as time-to-euthanasia, vary in accordance with the inoculum: planktonic bacteria often induce earlier and more acute symptoms, whereas biofilm-derived aggregates may cause a delayed but persistent infection (Figure 4B). Monitoring twice daily for up to 21 days allows capturing the full disease course.

Figure 2. Crystal violet staining, quantification, and ultrastructural imaging of Leptospira biofilms. (A) Crystal violet (CV) staining of biofilms grown on different substrates. CV staining allows visualization of biofilm architecture and initial attachment patterns for different species and substrates. i. L. interrogans on polycarbonate filter; ii. L. biflexa on polycarbonate filter; iii. L. interrogans on glass coverslip; iv. L. biflexa on glass coverslip. (B) Example of quantitative biofilm formation assessed by CV absorbance (OD570 nm) over time for L. interrogans and L. biflexa. This kinetic readout captures both early adhesion and biomass accumulation dynamics, highlighting differences in biofilm growth between pathogenic and saprophytic species. This figure has been modified from27. (C-E) Scanning electron microscopy (SEM) of L. interrogans biofilms: (c) 3-day-old biofilm showing early microcolony formation and initial extracellular matrix deposition;(d) 14-day-old biofilm displaying mature architecture with extensive matrix and three-dimensional organization;(e-f) 3-week-old biofilm illustrating structural consolidation and matrix maturation over prolonged culture. This combination of CV staining, quantitative OD measurements, and SEM imaging provides a comprehensive overview of biofilm development, from early attachment to mature ultrastructural organization, enabling direct comparison across species and culture durations. Please click here to view a larger version of this figure.

Figure 3. Visualization of Leptospira biofilm formation. Phase-contrast images acquired with a BioStation show biofilm development at (A) 48 h, (B) 96 h, and (C) 144 h. (D) CLSM reconstruction of a Leptospira biofilm displaying total biovolume with orthogonal slices for 3D visualization. (E) Confocal staining of a mature biofilm with DAPI (green) and WGA (red), highlighting bacterial cells and extracellular matrix components. This figure has been modified from37. Please click here to view a larger version of this figure.

Figure 4. Time-Resolved Quantification of Planktonic and Biofilm Fractions and functional assessment of Leptospira biofilms. (A) Kinetics of biofilm formation measured by absorbance at 405 nm. For each time point, readings were taken from the total well, the biofilm fraction, and the supernatant containing planktonic leptospires, allowing distinction between attached and free-swimming bacteria. (B) Example survival curves of hamsters injected with biofilm aggregates, planktonic leptospires, or EMJH control. Biofilm-injected animals show reduced virulence, with some surviving 21 days, whereas planktonic leptospires cause rapid mortality. (C-D) Preliminary validation of biofilm integrity: aggregates are visible in the syringe prior to injection (C) and remain intact after passage through a 21 G needle (D, white arrows), confirming that biofilm structure is preserved during handling. This figure has been modified from36. Please click here to view a larger version of this figure.

This protocol presents a modular and integrative workflow for studying Leptospira biofilms, integrating quantitative biomass (crystal violet)^9,27, dynamics (time-lapse phase contrast), three-dimensional composition (confocal laser-scanning microscopy with targeted probes)³⁸, ultrastructure (scanning electron microscopy), and functional assessment (hamster infection). By using the same culture series across modules, batch effects are minimized and within-experiment cross-validation is enabled, which is especially important for slow-growing, fragile spirochetes. Each module complements the others: CV quantifies "how much," time-lapse shows "how fast," CLSM reveals "what's inside," SEM resolves "how it's built," and in vivo testing provides "proof-of-function." Inclusion of L. biflexa demonstrates adaptability across species, highlighting faster and denser biofilm formation and emphasizing methodological flexibility³⁵.

The success of each module depends on inoculum quality and physiological state. Mid-log planktonic cultures (3-5 days), highly motile and free of aggregates, yield reproducible adhesion and biofilm development. Only low-passage isolates should be used to ensure robust biofilm formation and reproducibility.

The CV assay anchors the workflow through its speed, scalability, and throughput^9,27. It enables rapid screening of mutants, media, or anti-biofilm compounds. However, because Leptospira biofilms are fragile and heterogeneous, gentle manipulation and replicate validation are essential. CV cannot distinguish compact vs porous architectures or detect matrix composition -- hence its role as a screening gate for deeper structural analyses.

Time-lapse imaging bridges that gap by revealing biofilm kinetics in real time. It captures the emergence, coalescence, and immobilization of motile aggregates, exposing transient phenomena that endpoint assays miss. Environmental stability (temperature, humidity, autofocus) is critical. These recordings have shown that even when CV biomass appears stable, dynamics can diverge -- indicating structural rearrangements or viability loss in deeper strata.

CLSM provides volumetric and chemical insight without dehydration artefacts³⁹. Live/dead staining reveals cell viability gradients within the biofilm, consistent with reports of limited oxygen diffusion and metabolic stratification^6,40. Lectins and nucleic-acid dyes expose abundant extracellular DNA and specific glycan motifs, allowing quantification and characterization of matrix components^41,42. CLSM data frequently reveal internal voids, echoing the collective rearrangements seen in time-lapse³⁸. Still, limitations include probe specificity, photobleaching, and diffraction-limited axial resolution; therefore, microscope settings and dye performance should be standardized and tested beforehand to ensure reliable, comparable results across experiments.

SEM offers complementary ultrastructural resolution⁴³. At early stages, it resolves nascent aggregates; at maturity (≈ 15-21 days), it reveals a polarized architecture: a rough, channeled basal layer for anchorage and exchange, and a smoother apical matrix-rich surface³⁷. Careful fixation, HMDS drying, and correlation with confocal data mitigate artefacts due to dehydration or collapse⁴⁴.

Together, these four modules enable internal cross-validation: when CV, CLSM, and SEM converge, conclusions are robust; when they diverge, the discrepancies themselves are informative -- revealing, for example, increased biomass with reduced viability or unchanged biomass with altered matrix composition.

Several intrinsic limitations remain. Leptospira biofilms are heterogeneous and delicate, making them sensitive to handling and dehydration. CV integrates total biomass but not viability¹⁶; CLSM cannot exceed optical limits³⁸; SEM risks preparation artefacts. Mortality in deeper strata is expected due to oxygen gradients⁶, but this bias is systematic and comparable across conditions. Biofilm maturation reaches a plateau around 15-21 days, associated with transcriptional signatures of nutrient limitation³⁶. Later stages remain poorly characterized.

In vivo assays provide functional context. Injecting aggregates intraperitoneally tests whether biofilm-derived cells differ in virulence or persistence from planktonic counterparts. Although intraperitoneal inoculation is not a natural route, it offers a controlled comparative model³⁶. Aggregate heterogeneity introduces some variance, but consistent handling and matched inoculum size mitigate this. Importantly, biofilm persistence traits (e.g., ECM-mediated stress tolerance) do not necessarily translate into enhanced acute virulence, highlighting the ecological rather than pathogenic role of biofilm formation²⁴.

The modular design enables scalable use. For rapid screening, CV and time-lapse suffice. For mechanistic insight, CLSM and SEM add compositional and architectural resolution. Modules can integrate enzymatic or chemical perturbations (DNase, glycosidase), alternative probes for glycans or nucleic acids, or microfluidic systems to test flow responses. The same inoculum can feed transcriptomic or proteomic analyses, directly linking biofilm phenotype to regulation (e.g., c-di-GMP signaling, starvation response). This framework also accommodates mutant screening, environmental perturbation tests, and anti-biofilm or anti-virulence compound evaluation.

This integrated workflow transforms isolated observations into a coherent multidimensional understanding of Leptospira biofilms. By combining throughput (CV), dynamics (time-lapse), chemistry and depth (CLSM), and ultrastructure (SEM), the approach captures the mobile, porous, and polarized nature of Leptospira communities with fidelity. Extending earlier studies^17,35,36, it demonstrates that coordinated, modular experimentation reveals phenomena invisible to single-method studies -- such as rising biomass despite falling viability, or divergent kinetics among species. Ultimately, this approach supports comparative, reproducible, and functionally relevant analyses across species, mutants, and conditions, providing a foundation for mechanistic microbiology, ecological resilience, and anti-biofilm strategy design.