Root-derived reactive oxygen species spatially pattern bacterial cell-type differentiation during root colonization

Root-derived reactive oxygen species spatially pattern bacterial cell-type differentiation during root colonization | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Root-derived reactive oxygen species spatially pattern bacterial cell-type differentiation during root colonization Christian-Frederic Kaiser, Sabrina Egli, Mirco Keilhammer, M. Carolina Elizondo-Cantú, and 7 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8407390/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Phenotypic heterogeneity is an acclimation strategy in bacteria to cope with fluctuating environments and host immunity through cellular specialization. In the rhizosphere, how root-associated bacteria respond to plant defense signals and whether plants actively shape colonization through bacterial cell-type decisions, remains poorly understood. Cell type-resolved, spatiotemporal mapping of Bacillus subtilis and Bacillus velezensis colonizing Arabidopsis thaliana roots reveals dynamic transitions between flagellated and matrix-producing cell states. We identify an RBOHD-dependent, spatially heterogeneous oxidative landscape generated by the plant as a key regulator of bacterial cell-state balance. We show that exposure to reactive oxygen species (ROS) not only promotes differentiation into matrix-producing Bacillus cells, but ROS acclimation is also a prerequisite for cell type switching and stable association. Our findings demonstrate that bacterial stress evasion and acclimation to the host’s oxidative landscape are key steps in a host-microbe feedback system enabling stable rhizoplane colonization. Biological sciences/Plant sciences/Plant symbiosis Biological sciences/Microbiology/Bacteria/Bacterial host response Biological sciences/Microbiology/Bacteria/Bacterial immune evasion Biological sciences/Microbiology/Bacteria/Bacterial development Biological sciences/Cell biology/Cellular imaging Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Dynamic and diverse environments challenge bacterial populations. In response, bacterial species have evolved phenotypic heterogeneity, variation among isogenic cells, driven by stochastic fluctuations in signaling, genetic noise, microenvironmental variation, and cell–cell interactions 1–4 . The formation of subpopulations affects diverse bacterial traits, such as energy metabolism and lifestyle shifts, and is evident across taxa such as enterobacteria, cyanobacteria, myxobacteria, streptomyces, and bacilli 5–12 . Phenotypic heterogeneity is particularly advantageous during host colonization. In mammalian systems, bacterial subpopulations evade immune detection, tolerate antimicrobials and immune effectors by entering persistent states, dividing labor, or forming biofilms 1,4,13 . Salmonella , Neisseria , Yersinia, and Pseudomonas use such strategies to colonize challenging host niches 14–18 . The physiological importance of structured host-microbe interactions in commensal and mutualistic scenarios is particularly evident in the intestinal biogeography and its impact on human health 19,20 . Phylogenetic heterogeneity in host colonization has drawn major attention in animal and plant systems in recent years 21–25 . However, phenotypic heterogeneity as an adaptive strategy of host-associated microbiota and how host-derived cues are integrated remains largely unexplored. Just as the complex biogeography of the mammalian intestine, the plant rhizosphere offers a similarly complex and compartmentalized environment, with its microbiota structured by gradients in nutrients, oxygen, exudates, and immune activity. These gradients shape microbial niches and affect colonization patterns 26–28 . Plant immune responses are also spatially structured along the root’s axis, generating microenvironments of antimicrobial activity, immune perception, or oxidative stress that can provide positional information for colonizing bacteria 29–34 . Just like the multi-layered mammalian immunity, plants produce antimicrobials 35–38 , modify cell walls 39,40 , and release reactive oxygen species (ROS) 41,42 in response to microbial epitopes 43–45 . These immune constraints have selected adaptive mechanisms in microbes, including effector proteins 46–48 , altered epitopes 49,50 , or cooperative behavior 51 . Together, these observations place host-derived immune activity not only as a shared selective pressure but also as a potential spatial organizer of bacterial heterogeneity in the rhizosphere. Bacillus subtilis ( B. subtilis ), a common gram-positive model strain, demonstrates prominent phenotypic heterogeneity through variation in matrix production, the timing of sporulation, and multi-tasking across subpopulations 52–55 . Functionally distinct cell types arise from defined developmental trajectories, giving rise to flagellated, matrix-producing, or antibiotic-secreting subpopulations 54,56–58 . Division of labor, coordinated migration, structural maturation, and local microenvironmental gradients (e.g., O 2 , nutrients, pH) collectively generate spatial and temporal differentiation patterns within multicellular populations like biofilms 59–62 . Structured biofilm formation is dependent on the lifestyle transition from flagellated cells to matrix-producing cell chains, governed by the bistable SinI–SinR regulatory circuit 63–67,67,68 . This bistability enables a noise-driven individual cell fate decision to a matrix-producing or flagellated state, maintaining cellular heterogeneity 68,69 , which, through the regulation of SinI expression, integrates environmental feedback into the “decision-making” process 70–72 . While such structured differentiation is well established on abiotic substrates, how these developmental programs manifest and are regulated within organismal interactions and dynamically structured host environmentsremains far less understood. Still, phenotypic heterogeneity is rarely explored in rhizobacteria, with recent studies exemplifying the role of phenotypic heterogeneity during plant infection in Pseudomonadaceae 51,73 , while cellular subpopulations in Paenibacillus polymyxa and B. subtilis have been linked to plant-growth promotion and colonization 52,74–78 . Also, phenotypic heterogeneity in B. subtilis is regulated by root-associated factors like sucrose 79 , polysaccharides 59,80,81 , and reactive oxygen species (ROS) 82 , pointing towards possible interactions between immunity and phenotypic heterogeneity. Despite Bacillus phenotypic heterogeneity being observed in the rhizosphere, whether cell differentiation is merely incidental or regulated spatially by plant-derived cues like immunity remains unclear. Studying these interactions in soil is challenging due to opacity and structural complexity. Microfluidic platforms offer a solution by mimicking rhizosphere conditions (reviewed in Kaiser et al. 2023) 83 while enabling high-resolution imaging of root–bacteria interactions 84–88 . In this study, we employ novel microfluidic devices based on the RootChip technology 89,90 and cell type-specific reporters in B. subtilis PS-216 and B. velezensis FZB42 to test whether spatially patterned host-derived ROS regulate bacterial differentiation during colonization. We report that plant-derived ROS shapes the balance between flagellated and matrix-producing cells, identifying ROS as a major regulator of bacterial heterogeneity in the rhizosphere. Our findings support a model that bacterial acclimatization to oxidative stress results in spatially organized plant-microbe interactions by modulating individual cellular decisions, rather than merely limiting bacterial proliferation. Materials and Methods Bacterial Strain Construction All B. subtilis strains are derivatives of PS-216, a natural isolate extracted from Slovenian riverbank soil 91 and B. velezensis FZB42 92 a plant-growth-promoting relative, respectively, and are listed in Table 1. Integration vectors for generating fluorescent promoter fusions were constructed with the pXFP_Star system 93 and are listed in Table 2. For B. velezensis we first engineered a new integration vector pRFP_Star_FZB42 by replacing the B. subtilis amyE sites in pRFP_Star 53 with corresponding sequences from B. velezensis using Gibson assembly 94 . The new vector is available from the Bacillus Genetic Stock Center (http://bgsc.org) under Accession Number ECE796. In general, promoter regions contain 100-300 bp upstream of the ribosome binding site and were amplified from genomic DNA using primers (Table 3) and were inserted into the LIC site of pXFP_Star by ligation-independent cloning (LIC). The small cloning site of pXFP_Star was used to combine different promoter fusions using restriction enzyme ligation cloning, respectively. Plasmids were verified by analytical restriction digestion and sequencing before transforming in B. subtilis via one-step transformation 95 and into B. velezensis by a two-step transformation method 96 , respectively. Correct integration of fluorescent promoter fusions into amyE was verified by PCRs 93 . Knockouts in PS-216 were constructed by replacing the gene of interest with a kanamycin resistance cassette by transformation with a PCR product from the corresponding knockout from the B. subtilis B168 BKK single gene knockout collection 97 Growth Conditions Overnight precultures were grown in LB (Sigma-Aldrich; 10 g/L tryptone, 5 g/L yeast extract, 10 g/L sodium chloride, pH 7.0) at 37°C, 220 rpm, diluted 1:10,000 into fresh LB, and grown to mid-exponential phase (OD₆₀₀ = 0.5–0.8), washed thrice in half-strength Hoagland medium 98 (½HS) (Basal Salt Mixture No. 2, Sigma-Aldrich, 1 g/L MES, pH 5.7 KOH), and resuspended to the desired OD in ½HS. Chloramphenicol (5 µg/mL) or/and kanamycin (25 µg/mL) were added when applicable. Plant Material Arabidopsis thaliana lines used are listed in Table 4. All seeds were filtered to hold a diameter >250 µm, to ensure comparable seed quality across genotypes. Seeds were sterilized using chlorine gas for 2 h, degassed for 1 h, and stratified at 4°C for 48 h submerged in sterile water. For microfluidic experiments seeds, were prepared as described in Guichard et al. 2020 99 , germinated in agar-filled 10 µL pipette tips with ½HS medium (0.8 % plant agar), then grown at constant 22 °C under long-day conditions (16 h light/ 8 h dark, 150 µmol/m²/s) for 4–5 days in an enclosed plant growth cabinet (CLF Plant Climatics GmBH, SED41-CU4C2LT). RootChip and µRhizotron Fabrication and Plant Inoculation RootChip8NF (RC8NF, Fig. S1A) and µRhizotron (Fig. S2) devices with eight microchannels and a soil-mimicking Voronoi pattern with an average 400 µm pore diameter were designed in Fusion 360. Casting molds for PDMS were printed via stereolithography printing (Formlabs, Form3) in clear resin (Formlabs, Clear Resin V4.0), cleaned, UV-cured, and lastly cast with PDMS (DOW Chemicals, Sylgard 184, 10:1). Chips were plasma-bonded to glass slides, sterilized with UV for 10 min for each side, and filled with ½HS medium. Seedlings were transferred pre-germinated from pipette tips into channels and grown at 21°C, 45° inclination, under long-day conditions until 7 days post seeding (dps). Bacterial inoculation was then performed by flushing chips with 1 mL of B. subtilis or B. velezensis FZB42 suspensions (OD₆₀₀ = 0.02) in ½HS. Chips were again incubated in the plant growth chamber and 45° incline until imaging. Microscopy Setup Spatial mapping was performed on a custom-built inverted open-frame epifluorescence microscope (COS; Cairn GmbH, Heidelberg) equipped with a 25× water-immersion objective (NA 1.1, Nikon), motorized stage (ITK Dr. Kassen), laser launch (LDI 89 North), triple-band dichroic mirror (395 nm/ 488 nm/ 561 nm, Chroma), and sCMOS camera (Teledyne Photometrics Kinetix). Fluorescence imaging used 488, 520, and 577 nm excitation (30–50% intensity), 50 ms exposure, and 525+/-50 nm (GFP), 542+/-20 (YFP), and 595+/-26 nm (RFP) emission filters (Semrock). The system was automated utilizing Micromanager (V.2.0) 100,101 . Time-lapse imaging was conducted on a separate custom spinning disk system built on a Nikon Ti-E stand with a 10× S Fluor objective (NA 0.50), CrestOptics X-Light disk (50 µm pinhole), motorized stage (ASI), laser launch (Omicron), filter wheel (Cairn), and Prime 95B sCMOS camera (Teledyne Photometrics). A triple-band dichroic (440 nm/ 514 nm/ 561 nm) and quad-band (405 nm/ 488 nm/ 561 nm/ 640 nm, Chroma) were used with 488 nm, 520 nm, and 561 nm excitation (80 mW), and 525+/-50 nm and 605+/-70 nm emission filters for GFP and RFP, respectively. Images were acquired using Nikon NIS Elements v4.0. Mapping of ROS levels was conducted on an Olympus MVX10 stereo microscope, equipped with a 0.63x objective (MVPLAPO, 0.15 NA, FOV = 55 mm) with magnification of 1.25x, motorized focus (SZX2-FOF-M), CoolLED pE4000 light source, quad-band dichroic mirror 395+/-25, 475+/-28, 555/25, 635+/-18 (Chroma 89402bs), 488 nm excitation with an 519 +/- 26 nm emission filter (Chroma), CellCam Kikker 100MT camera in 2x2 binning mode (14bit, px=6.07µm, Cairn UK). Image acquisition was done through Micro-Manager software version 2.0. Microscopy of Microcolonies LB agar was poured in 4-well cell culture slides (Sarstedt, X-well slide, 94.6140.402) and one half removed. This agar cube was then laterally spotted with B. subtilis PS216 (BIB2352) and incubated overnight at 37 °C. The microcolonies formed at the border of glass, agar and air was then imaged on the Nikon Ti-E system at 20x magnification using the integrated large image function a 9x9 grid was imaged and stitched. Statistical Analysis All statistical analyses were performed in R. As most variables were bounded and non-normally distributed, we did not assume normality for raw data. Followingly, all tests were performed non-parametrically. Pairwise comparisons between genotypes or treatments were conducted using two‑sided Wilcoxon rank‑sum tests, and effect sizes were reported as rank‑biserial correlations (r rb ). Multiple testing was corrected using the Benjamini–Hochberg false discovery rate procedure. All displayed comparisons are calculated with 95 % confidence intervals (CI), effect size, p-value, and adjusted p-value (Tab. 5 – 15). Spatial Mapping of Bacterial Cell Types At 10 days post-seeding (dps) and 3 days post-inoculation (dpi), roots were imaged using epifluorescence microscopy on the Cairn open-frame setup. Each primary root was imaged once at 3 dpi and treated as an independent biological replicate; no root was measured repeatedly. A ~1 mm × 5 mm (width × height) region starting at the root tip was captured using 18 overlapping fields of view (FOVs) (820 × 820 μm each, 0.52 µm/px), spanning the full root diameter in Z-stacks (80–140 μm depth, 5 μm steps) from the cover slide surface. Images were processed in Fiji 102 (v2.17.0), including background correction (Rolling ball, radius = 20 px), stitching, and maximum intensity projected across all Z-slices via a custom script. Generally, all experiments were performed on three independent days, including multiple roots and chip setups. Segmentation of bacterial coverage was performed using a fixed threshold, derived from the average lower threshold of the Col-0 subset (N = 37) using the “Default” auto-thresholding method. Bacterial coverage was segmented based on size and shape to limit the detection of auto-fluorescent plant structures, including single cells and aggregates (Fig. S1C). Each isolated cluster was then identified as a region of interest (ROI), mean fluorescence intensity, size, x and y coordinates isolated. The root midline was manually labelled and extracted as a continuous polyline. The distance between the pixels (0.52 x 0.52 µm) on the root midline and bacterial clusters was determined using nearest-neighbor search via the RANN 103 (v.2.6.2) package, yielding a distance to the root midline. To account for root diameter, distances were corrected by subtracting 100 µm: clusters with corrected values ≤ 0 µm were classified as “rhizoplane”-associated, whereas those > 0 µm were classified as “rhizosphere”-associated, notably, the rhizoplane or root section is a planar Z-projection and does not account for radial depth of the root. The cumulative length of the midline up to the nearest point for each cluster was recorded as its distance from the root tip. Bacterial clusters were binned into radial segments of 50 µm or longitudinal segments of 200 µm. Longitudinal analyses focused on the rhizoplane (≤ 0 µm from the root, e.g. the area of the planar Z-projection converged by the root). Coverage was calculated as the proportion of each bin’s area occupied by fluorescent clusters relative to the total analyzed area. Subpopulation composition was expressed as the proportion of area covered by each bacterial group within the total occupied area. For statistical analyses, these values were first aggregated at the root level, and roots were treated as independent biological replicates. Values were then averaged across roots imaged on different days and in different microfluidic chips. Where applicable, data is displayed as mean value ± standard error (SE) calculated across roots. Continuous two-dimensional kernel density (KDE) estimates were calculated for each root and averaged across replicates using the spatstat 104 package (v3.4.1) in R. KDEs were generated for the area-weighted counts of GFP-expressing, mCherry-expressing, and all fluorescent cells (Total), as well as for the fraction of matrix-producing cells, defined as their coverage relative to the coverage of all fluorescent clusters within an area. Statistical analyses included multiple pairwise comparisons using Two-sided Wilcoxon rank-sum tests with Benjamini–Hochberg correction. Correlations between population fractions were assessed using β‑regression for ratio data bounded between 0 and 1 using the betareg package (v.3.2.4). The same Col‑0 dataset was analyzed in Figures 1 and 5. In these models, the fraction of matrix-producing cells served as the response variable, and longitudinal root position was included as a continuous covariate. Alternatively, the fraction of motile cells in the total population (bounded between 0,1) was plotted against the continuous area coverage by matrix-producing cells. A logit link function was used, and roots were treated as independent biological replicates contributing one proportion value per longitudinal bin. µRhizotron Time-lapse Imaging A single seedling was inoculated with strain BIB2352 (Table 1) at 7 days post-seeding, incubated for 3 additional days, and imaged over 15 hours. Imaging was performed on the Nikon Ti-E system using 1 × 1 mm fields with Z-stacks acquired at 140 µm / 20 μm (7 steps). Image tiles were stitched using the pair-wise stitching plugins 105 , and bacterial (ROIs) were segmented based on fluorescence intensity in ImageJ. This time-lapse experiment represents repeated measurements of the same root over time and therefore constitutes a single biological replicate (n = 1). Bacterial ROIs after intensity-based segmentation contained isolated bacterial aggregates of multiple cells as also single cells, these were considered as “clusters”. ROIs were identified on the last timepoint at 15 h. For each cluster corresponding to the spatial mapping, across timepoints fluorescence intensity in GFP and RFP was measured resulting in the fluorescence trajectory, and after manually labelling the root midline a distance from the root and root tip was calculated based on the XY coordinates of each ROI corresponding to the spatial mapping in the RC8NF using RANN 103 . Each spatially distinguishable bacterial cluster was treated as a biological replicate, and its fluorescence was measured repeatedly over time. Clusters were determined as “adhered” once the fluorescence, primarily in GFP, reached an above background level, switching to matrix producers (P tapA -mCherry) was determined as the first timepoint, using a threshold set as the mean plus one standard deviation of RFP intensity observed in GFP-fluorescent clusters at their final timepoint. Cells were classified as flagellated ( GFP ) below this threshold and matrix-producing ( mcherry ) once this threshold was reached. Averages were built upon each individual bacterial cluster that was captured. A KDE model was used to represent the spatial distribution of all switching events as the location relative to the root and root tip at which clusters first activated mcherry expression. Additionally, a generalized additive model (GAM) was fitted using the mgcv package (v.1.9-4) in R, with the probability that an individual cluster exhibited at least one switching event (binomial family, logit link) as the response variable and longitudinal root position as a continuous covariate modelled with a smooth term. To avoid pseudo replication arising from repeated observations of the same cluster across timepoints, the model was fitted using only the final imaging frame, such that each cluster contributed a single observation corresponding to its final distance from the root tip. Imaging of Plant Immunity Markers Seedlings were inoculated with B. subtilis PS-216 (OD₆₀₀ = 0.02) transcriptional reporter strains (Table 1) or mock-treated, grown 3 days post-inoculation, and imaged in brightfield and YFP channels on the Cairn Openframe microscope in the RootChip8NF accordingly to the previous section. Each primary root was imaged once and treated as an independent biological replicate. A 150 µm-wide midline (288 px) was placed manually utilizing the segmented polyline tool in FiJi (ImageJ), and the YFP fluorescence profile was extracted. Either the average fluorescence across the mean line or the average per position across all imaged roots of each treatment was compared. Statistical analyses included multiple pairwise comparisons using Two-sided Wilcoxon rank-sum tests with Benjamini–Hochberg correction. ROS Detection on Roots At 7 dps, seedlings of A. thaliana (Table 4) grown on nylon meshes (pore diameter 100 µm) were treated with bacterial suspensions of B. subtilis PS216 (P veg -mCherry) or mock-treated with sterile medium. At several timepoints starting before the inoculation up to 3 dpi, ROS was detected using 2 μM H₂DCFDA (stock solution: 20 mM in absolute ethanol) staining for 20 min at room temperature and in the dark and imaged via stereomicroscopy (Olympus MVX10) in the GFP channel. The experiment was repeated across several meshes and days containing multiple roots. Each root was treated as an independent biological replicate and was imaged only once. Meshes were not imaged repeatedly; each mesh was used for a single imaging round and then discarded. Each image was background-corrected (rolling ball, r = 50 µm) and manually labelled with a spline-fitted continuous midline of 25 px (~100 µm) in ImageJ, coordinates and mean fluorescence intensity were extracted. Intensity profiles were averaged across individual roots across all meshes and normalized to the mock-treated average. For time-lapse imaging, untreated meshes were used as 0 dpi “untreated” controls. Multi-well Plate Reader Assays Cultures were prepared in supplemented ½HS (0.5 % w/v L-glutamate, 0.5 % v/v glycerol), diluted to an OD 600 of 0.1, then mixed 1:1 with treatments (final OD 600 0.05) in 200 µL volume on black flat-well 96-well plates (GreinerBIO). Overexpression assays included 1 mM IPTG. Dilution series (1:1) of H₂O₂ were prepared automatically using a pipetting robot (Opentrons, OT-2), and additionally, an unsupplemented medium control was provided. Measurements were made using a CLARIOstar Plus microplate reader (25°C, no shaking), capturing GFP (470 +/- 15 nm, 515 +/- 20 nm, 20 flashes), mCherry (570 +/- 15 nm, 620 +/- 20 nm, 20 flashes), and absorbance at 600 nm (10 flashes). A settling time of 0.5 s was allowed for each capture. Fluorescence was normalized to OD 600 . Multi-well plate experiments were performed in at least two independent time‑series experiments. Within each experiment, each well containing the same treatment was treated as an independent biological replicate. Wells were measured continuously over time, but liquid samples for spotting onto agarose pads were taken only once per well; each well, therefore, contributed a single biological replicate for a specific time point to the agarose-pad assay. At multiple timepoints during growth in the presence of either 0 or 156 µM hydrogen peroxide (H 2 O 2 ), 1 µL aliquots of B. subtilis PS-216 (BIB2352) or derived mutant cultures from the multi-well plates were spotted onto ½HS 1.5 % agarose pads. Spots were imaged on the Nikon Ti-E using NIS Elements (Large Image Mode), capturing multiple fields of view (5x5, 1x1mm FOVs, 25 % overlap) to generate complete spot images. Fluorescent regions were segmented and background-corrected using untreated agarose pads as a reference. The area covered by bacterial cells was isolated by intensity-based segmentation; such areas were then assigned as ROIs and summarized. This was performed separately for GFP- and RFP-fluorescent areas, which were summed to a total fluorescent area. The cell density was estimated based on the minimal size of distinct fluorescent clusters, which was 4 pixels, e.g. 4 µm 2 . Software Image analysis and processing were performed using ImageJ/Fiji (v1.54r). Statistical analysis, plotting, and data preprocessing were conducted in R (v4.5.2) using the RStudio environment (v2025.09.2+418). Results Root-colonizing Bacillus populations undergo a spatial transition from flagellated to matrix-producing cells along the root’s axis To investigate the cell type differentiation of a root-colonizing rhizobacterium, we first engineered fluorescent reporter systems to monitor cell differentiation in the soil isolate B. subtilis PS-216 91 and the plant-growth-promoting B. velezensis FZB42 92 , respectively. To this end, the expression of gfpmut3 was placed under the control of the hag- promoter that is activated in flagellated cells 57 . mCherry was placed under the tapA promoter that controls expression of the matrix production-associated tapA-sipW-tasA operon 57,107 and is thus activated in matrix producers (Fig. S1A). Stationary-phase cultures grown for 48 h in supplemented ½HS medium (0.5 % w/v glycerol and mono-potassium glutamate), showed a heterogeneous population, composed of green and red fluorescently labelled cells (Fig. S1B, S1C), indicating that motility and matrix production are differentially activated in distinct subpopulations, as expected 64,65 . These bacterial cultures were then used in root colonization experiments with A. thaliana. To track root colonization by fluorescence microscopy, we developed a microfluidic imaging slide termed RootChip8NF based on the RootChip8S 90 scaffold, allowing in situ imaging on living roots of A. thaliana (Fig. S1D). After three days post inoculation (dpi), fluorescence from both fluorescent reporters was visible on the roots colonized by B. subtilis (Fig. 1A, left) and B. velezensis (Fig. 1A, middle), respectively, suggesting that the root environment also supports the coexistence of both flagellated and matrix-producing Bacilli. Notably, the fluorescence distribution varied with distance from the root tip, suggesting that different cell types might be favored in different parts of the root. We estimated the spatial population structure based on the share of matrix producers to the overall bacterial biomass (Fig. 1B, 1C), flagellated cells, as a function of distance from the root and distance from the root tip from the fluorescence distribution (see Methods for details). As expected, bacterial biomass accumulated in the root (rhizoplane, 0 µm from the root) (Fig. 1D, left). Population composition, as estimated from the coverage of green and red fluorescence areas, respectively, showed a pronounced shift with distance from the root tip (Fig. 1B, 1C, 1E). Flagellated cells were preferentially present around the root elongation zone (~500-800 µm from the root tip), and matrix-producing cells increased in abundance and dominated the maturation zone (>850 µm from the root tip) of the root 108 . This shift was consistently observed across multiple roots and chips with both B. subtilis and B. velezensis (Fig.1C, E), respectively. We also confirmed qualitatively, using a non-fluorescent wildtype of PS-216, a constitutively fluorescent (P veg - mcherry ) and fluorophore swapped strain (PS-216, P hag -gfpmut3, P tapA -mCherry ), that observed patterns were not attributed to autofluorescence (Fig. S1E). Secondly, cell type localization and detection of both flagellated and matrix-producing cells remained robust upon fluorophore exchange (Fig. S1E). Together, these data suggested that both Bacilli colonize A. thaliana roots through a spatially organized transition from flagellated to matrix-producing cells. B. velezensis showed overall a higher abundance of matrix producers compared to B. subtilis, and the transition from motility to matrix production occurred closer to the root tip. We also found that matrix-producing cells predominantly covered the rhizoplane (Fig. 1D). To test whether differentiation from flagellated-to-matrix producers is required for rhizoplane association, we performed colonization experiments with PS-216 and a sinI knockout. SinI is known to bind to the SinR transcription factor, the master regulator of biofilm formation 18,71 . Indeed, in the sinI knockout mutant, less bacterial biomass accumulated in the rhizoplane (5.7% estimated coverage) compared to the wildtype (12.9% estimated coverage), while there was no significant difference in biomass in the rhizosphere (Fig. 1G, 1H). This indicates that cell differentiation from flagellated cells towards matrix producers via the SinI-SinR switch is necessary for efficient rhizoplane colonization. B. subtilis cell type differentiation occurs locally confined in the early maturation zone We next aimed to observe cell state transitions from motility to matrix production by monitoring changes in gene expression by time-series imaging. To this end, we performed colonization experiments with the PS-216 reporter strain in a structured microfluidic imaging slide termed the µRhizotron. This device mimics a spatially heterogeneous soil environment by featuring soil-particle-mimicking Voronoi patterns (Fig. S2). We captured continuous cell fate transitions over 14 hours in the actively growing root tip region of a single A. thaliana Col-0 root 3 days post inoculation (Fig. 2A). For a detailed analysis, we focused on a 2 mm segment encompassing the transition from the meristematic to maturation zone 108 (Fig. 2A, upper). We hereby analyzed isolated clusters of fluorescent bacterial cells, which encompassed both single cells and aggregates. Over time, we first observed the presence of green fluorescence signals from P hag, which was then followed by an onset of red fluorescence from P tapA in some clusters (Fig. 2A, lower; white box), indicative of switching events from motility towards matrix production. Based on the fluorescence profile, we designated two subpopulations gfp -expressing clusters (flagellated), and such that displayed a shift from gfp to mCherry expression, termed “switchers”. Averaged across all switching cells, mCherry fluorescence rose steadily from 5 h post-adherence to the root (Fig. 2B), indicating the activation of the P tapA and thus matrix production. The fraction of matrix producers started to rise after 5 hours and increased gradually over time to reach 25 % after 12 hours (Fig. 2C, left). Switching events were confined to a spatial window of 865 +/- 248 (SD) µm from the root tip (Fig. 2D), coinciding with the area of initial root hair formation, indicative of the maturation zone (Fig. 2A). Correspondingly, generalized additive modelling revealed a significant, non-linear dependence of switching outcomes in the final imaging frame (14 h) on distance from the root tip (edf = 8.8, p = 6.8 × 10⁻⁴). Clusters that had undergone at least one switching event during the experiment were enriched at the root cap and at intermediate distances along the root, with a marked increase in switching-associated clusters beyond 500 µm from the root tip and a local maximum around 800 µm (Fig. 2E). In conclusion, these findings suggest cell type switching occurs not homogeneously along the root axis but predominantly localized around the root elongation zone. As a result, the fraction of matrix producers varies as a function of distance from the root tip. Consistent with our previous observations (Fig. 1E), a flagellated-dominant population progressively shifted toward matrix production with increasing distance from the root tip, also in the µRhizotron (Fig. 2C, right). This indicates that the spatial modulation of cell-state transitions in different areas of the root occurs robustly, and even in a spatially heterogeneous environment that mimics aspects of soil. We thus reasoned that host-dependent factors might modulate cell differentiation during root colonization. Root colonization elevates local ROS production in the elongation and early maturation zone We next asked whether spatially patterned oxidative stress could act as an environmental cue shaping bacterial cell-state transitions. As the evasion of host stress through phenotypic heterogeneity is a known strategy in other pathosystems 14–18 , and specifically, since cell type differentiation is environmentally regulated in Bacillus 80,82,109 , we hypothesized that the plant-shaped environment may shape the observed longitudinal patterns of bacterial colonization. Since most bacterial cells reside on the rhizoplane, we focused on oxidative stress due to its spatially confined nature. To assess changes in the oxidative stress environment during bacterial colonization, we employed the oxidative-stress-sensitive dye H 2 DCFDA to map basal ROS production. H 2 DCFDA reports steady-state oxidation but does not distinguish ROS species or identify their cellular source; therefore, our interpretations focus on spatial associations rather than molecular specificity. Fluorescence was captured following a 20-minute dark incubation, enabling comparative estimation of steady-state ROS levels across root zones. This approach provided spatial resolution of ROS accumulation in living A. thaliana roots, offering a functional readout of oxidative stress dynamics in response to colonization. Generally, the peak of ROS levels was found between 1,000 to 1,600 µm from the root tip (Fig. 3A, 3B), with an increased spatial extent after inoculation with PS-216. During colonization over three days with B. subtilis PS-216 (initial OD 600 0.02), we observed a steady increase in the peak and global H 2 DCFDA fluorescence with a root-tip-ward peak in the elongation zone (Fig. 3A, 3B). Compared to mock-treated plants, colonized roots exhibited a significant increase from 0 - 3 dpi (slope = 208.4, τ = 7.5, Kendall’s Tau) while mock-treated plants did not (slope = -0.47, τ = -0.01, Kendall’s Tau) (Fig. 3C). To test whether increased ROS levels would also correlate with immunity activation, transcriptional reporter lines for early pattern-triggered immunity (PTI) were colonized with wild-type PS-216 32 . Furthermore, the early PTI transcriptional reporter pFRK1::3xmVenus~NLS and pPER5::3xmVenus~NLS confirmed an induction of both marker genes in the 1-2 mm from the root tip in response to colonization by PS-216, while the rather general stress-associated marker ZAT12 110 did not show activation (Fig. S3A–C). ROS production in response to biotic interaction is primarily mediated in plants by membrane-bound NADPH oxidases, known as respiratory burst oxidase homologs (RBOHs), with RBOH D and RBOH F playing key roles during PTI in Arabidopsis 111 . To determine whether the observed ROS increase during colonization is dependent on these enzymes, we inoculated rbohD , rbohF , and rbohD/F mutants with B. subtilis PS-216 for 3 days. All genotypes retained a root-tip-ward ROS peak (Fig. 3D); however, H 2 DCFDA fluorescence was notably reduced in rbohD and rbohD/F mutants, particularly in the root tip and maturation zones (Fig. 3E). Quantitatively, both mutants exhibited a significant reduction in mean ROS levels (approximately 21–25%) compared to Col-0 wildtype, while rbohF did not show a significant change (Fig. 3F). These findings demonstrate that B. subtilis colonization elevates ROS levels in a partly RBOH D-dependent manner, with a pronounced maximum in the elongation zone. Notably, this ROS peak spatially coincides with the activation of the PTI markers FRK1 and PER5. Additionally, the activation of immunity and the ROS peak are in proximity to the flagellated cell maximum and the transition point toward matrix-producing cells, indicating that host-derived oxidative stress may provide spatial information for bacterial cell-type transitions. ROS exposure promotes matrix production in Bacillus subtilis To assess whether ROS exposure can induce a shift in population structure, we asked whether hydrogen peroxide (H 2 O 2 ), a main component of root-derived ROS, could elicit cell type-dependent effects in axenic cultures of B. subtilis . To test this under well-controlled conditions, we performed plate-based growth assays of B. subtilis in gradients of hydrogen peroxide, grown at 25 °C and carbon-supplemented Hoagland’s plant medium, to create the closest proxy to our RootChip environment. Estimates for apoplastic ROS under non-stressed conditions indicate nM to low µM concentrations in the apoplastic fluid of A. thaliana , which can rise by several orders of magnitude in response to stress 112,113 . Physiological effects have been reported at ~100 µM H 2 O 2 in stomatal closure 114 , and leaf tissue concentrations of H 2 O 2 can reach ~1 µmol/g fresh weight 115 . We, therefore, decided to test the effect of a wide range of nM to mM concentrations of hydrogen peroxide, applied in parallel to inoculation, on the growth behavior of B. subtilis PS-216 cultures in multi-well plates. Increasing concentrations of H 2 O 2 resulted in a delayed onset of growth, with severe reductions of the final cell density at concentrations >156 µM of H 2 O 2 , and complete inhibition at 10 mM (Fig. 4A). The motility-associated transcriptional reporter P hag - gfpmut3 exhibited a biphasic response at control and low H 2 O 2 concentrations, with an early fluorescence peak during the lag phase followed by a dosage-dependent decline throughout exponential and stationary phases (Fig. 4B). At concentrations exceeding 156 µM, this initial peak was absent. In contrast, the matrix-associated reporter P tapA - mCherry became active upon entry into the exponential phase, maintaining steady levels during the lag phase (Fig. 4C). Peak fluorescence of P tapA - mCherry increased progressively with H 2 O 2 exposure up to 156 µM, suggesting enhanced matrix gene expression under oxidative stress. Consistent with these trends, the ratio of P tapA - mCherry to P hag - gfp fluorescence at the same cell densities increased with rising H 2 O 2 concentrations (Fig. 4D). To assess whether these transcriptional changes reflected shifts in population structure, we imaged 1 µL samples from untreated and 156 µM H 2 O 2 -treated cultures on agarose pads over 48 h (Fig. 4E). Cell density was estimated from the area covered by GFP- and mCherry-expressing cells, estimating a planar cell area of 4 µm 2 116,117 . The initial inoculum (T=0 h) was primarily composed of flagellated cells; both groups displayed an initial reduction of cell density at 12 h, with the ROS-treated group displaying a delayed growth onset and lower final cell density (Fig. 4E, 4F, 4G). While in the untreated group, flagellated cells remained predominant throughout the assessed time, in the ROS-treated group, matrix-producing cells were most abundant (Fig. 4E, 4F, 4G). Followingly, the reporter activity and the share of matrix-producing cells beyond 24 h were found to be higher in the ROS-treated group (Fig. 4H, 4I). These findings raised the question of whether matrix production confers increased ROS tolerance. To test this, we compared wild-type B. subtilis PS-216 with Δ sinI (flagellated-dominant) and Δ sinR (matrix-hyperproducing) mutants in the genomic background. Notably, the Δ sinR strain exhibited robust outgrowth at 156 and 313 µM H 2 O 2 within 24 hpi, in contrast to both wild-type and Δ sinI , supporting the hypothesis that matrix production enhances oxidative stress resilience (Fig. 4J). Together, these results indicate a biphasic effect of ROS with an initial delay in growth and matrix production, which leads to a rise of matrix-producing cells following exposure events, increasing the overall fitness under ROS stress. Indeed, the Δ sinR mutant highlights the stress resilience of matrix-producing cells, indicating acclimation by matrix production. RBOH D activity promoted the flagellated-to-matrix transition on the rhizoplane To investigate whether the plant-derived ROS influences the differentiation into matrix producers during root colonization of A. thaliana (Fig. 1B-E) , we repeated the spatial mapping of bacterial colonization in the ROS-production-impaired respiratory burst oxidase homologue (RBOH) mutants rbohD , rbohD/F, and rboh F (Fig. 3D-F) in direct comparison to the same Col-0 reference group as shown in Fig.1. rbohD and rbohD/F displayed a selective increase of the flagellated cell coverage globally around the primary root segment (1,000 x 5,000 µm, w x h), with the matrix-producing fraction being unaffected (Fig. 5A, 5B, S4A). This is also reflected in the rhizoplane coverage, which transitions from a predominantly matrix producer population in Col-0 to a flagellated-dominant population in the rbohD and rbohDF mutants (Fig. 5D, 5E). Additionally, this effect is seen in the longitudinal axis seen along the longitudinal axis, as rbohD and rbohD/F consistently display a higher fraction of flagellated cells (Fig. 5F, 5G). Beta-regression analysis investigating the relationship of the distance from the root tip and the share of matrix-producing cells revealed that in rbohD and rbohD/F only the elevation and intercept are significantly reduced, though not the slope of the trendline (Fig. 1H,1I), highlighting that the overall tendency towards matrix-producing cells is not affected. Still, a second analysis of the relationship of the share of flagellated cells and matrix producer coverage displays that in rbohD and rbohD/F, a higher degree of matrix producers is required to achieve a similar population structure (Fig. S4B, S4C). As the rbohF mutant neither suppressed ROS production in response to B. subtilis PS-216 (Fig. 3D–F) nor altered bacterial population structure, the observed effects appear specific to rbohD -deficient lines. Disruption of RBOHD activity, therefore shifts bacterial population structure toward flagellated states, consistent with a role for ROS in priming matrix commitment during colonization. Discussion Our investigation reveals that Bacillus subtilis and the related B. velezensis establish a longitudinal cell type pattern along the root axis through cell type differentiation. This transition corresponds to the phase-shift from planktonic growth to robust attachment to the root by biofilm formation 118 - a key lifestyle change governed by cell type differentiation from flagellated-to-matrix-producing cells in Bacillus sp. 54,70 . In accordance, we show that cell-type switching supports effective surface adhesion, as seen in the loss of root-association in the non-switching mutant ΔsinI . This highlights the structural and functional relevance of phenotypic heterogeneity to establish and maintain root association. During association, flagellated-to-matrix producer transitions do not occur ubiquitously, but are tightly confined to a spatial window in the early maturation zone. Indeed, the transition from the elongation to the maturation zone marks a critical physiological boundary in the root, characterized by the emergence of root hairs, reduced immune activity, and shifts in exudation profile 28,119 . This tight spatial confinement of bacterial cell-type differentiation indicates modulation by the host’s responses. Supporting the hypothesis of an active role of the host in bacterial cell type switching, we observe, upon inoculation with B. subtilis , activation of PTI coinciding with an RBOHD-dependent increase in ROS production. Spatial mapping reveals peak ROS levels at the elongation zone, coinciding spatially with flagellated cell accumulation. Surprisingly, the loss of RBOHD reduces ROS levels but also enriches flagellated cells in the colonizing Bacillus population. This indicates that oxidative environments influence the balance of bacterial cell types, rather than directly repressing motility or promoting matrix production. Despite early inhibitory effects, exposure to hydrogen peroxide enriches matrix-producers in B. subtilis. An enhanced stress tolerance of matrix-producing cells lies in known antioxidant traits, including iron siderophore release and elevated catalase expression 71,120,121 . Hence, the ROS-stimulated shift of the bacterial population towards matrix production may be an acclimation mechanism to the oxidative environment that promotes root association. This ROS-induced biofilm formation has also been observed in gram-negative interactors and pathosystems. For example, during root colonization by Salmonella enterica 122 and lung infection due to Staphylococcus aureus 123 , biofilm formation has been reported as a response to oxidative stress. We propose that the spatially polarized ROS gradient along the root axis contributes to a dynamic pattern in which (i) newly recruited bacterial cells at the root apex retain their flagellated state, (ii) subsequently adhere to the root surface, and (iii) transition into matrix production in a stress-exposure-dependent manner, giving rise to a more stress-resilient population through matrix production. These observations point to a causal link between host-derived ROS environments and bacterial cell-state differentiation during colonization. Environmental feedback by the plant host has been well established to shape the phylogenetic heterogeneity of the root microbiota by immunity, ROS production - including RBOHD activity - and exudation 22,28,124,125 . Our findings now highlight phenotypic heterogeneity as an additional layer of complexity. Bacterial cell-by-cell decisions tailor root-colonization outcomes, integrating the diverse micro-environmental landscape of pH, exudates, and immunity, all with potential feedback into phenotypic heterogeneity. Future work will be needed to assess how these gradients jointly shape cell-state dynamics in soil environments or transparent-soil analogs, and to determine how broadly such spatial interactions extend across rhizobacterial taxa. Studying such micro-scaled interactions, which are currently inaccessible by broad-scale omics techniques, is facilitated by using emerging cultivation and imaging approaches involving microfluidics and structured microenvironments 83 . In conclusion, this study highlights the critical need for spatially resolved analyses and the integration of phenotypic heterogeneity to fully capture the complexity of colonization behavior in situ. Declarations Contributions CFK, IBB, and GG designed the study. CFK, SaE, MZ, MK, SeE, and JE performed the experiments. CFK, CE, MK, JE, SaE, SeE and TS generated strains, CK engineered the vector, CFK designed RootChips, and CFK, SaE, and MK produced RootChip devices. CFK, SaE, and MK performed microscopy and image analysis. CFK performed data and statistical analyses. CFK, IBB and GG wrote the manuscript with input from all authors. IBB and GG supervised research. All authors read, revised, and approved the final manuscript. Funding We thank the German Research Foundation (CRC 1535 “Microbial Networking”, project ID 458090666; DFG Heisenberg Professorship, GR4559/4-1; and CEPLAS-EXC-2048/1-project ID 390686111 to GG and SPP2389 project ID 504020040 to IBB) and the Max Planck Society (IBB) for funding. Acknowledgements The authors gratefully acknowledge Michaela Gerads for technical support, Daša Wernerová for support with 3D printing. We thank Rubén Garrido-Oter, Marjorie Guichard, and Julia Frunzke for discussions and for critically reading the manuscript. Declaration of generative AI and AI-assisted technologies in the writing process. During the preparation of this work the author used Copilot (Microsoft) and ChatGPT-5 (OpenAI) in order to aid readability and text editing. After using this tool/service, the author reviewed and edited the content as needed and take full responsibility for the content of the published article. Declaration of Interests The authors declare no competing interests. References Ackermann, M. A functional perspective on phenotypic heterogeneity in microorganisms. Nat Rev Microbiol 13 , 497–508 (2015). Bódi, Z. et al. Phenotypic heterogeneity promotes adaptive evolution. PLoS Biol 15 , e2000644 (2017). Reyes Ruiz, L. M., Williams, C. L. & Tamayo, R. Enhancing bacterial survival through phenotypic heterogeneity. PLoS Pathog 16 , e1008439 (2020). Sherry, J. & Rego, E. H. Phenotypic Heterogeneity in Pathogens. Annual Review of Genetics 58 , 183–209 (2024). Erikson, D. Differentiation of the Vegetative and Sporogenous Phases of the Actinomycetes: 2. Factors affecting the Development of the Aerial Mycelium. Journal of General Microbiology 1 , 45–52 (1947). Wireman, J. W. & Dworkin, M. Morphogenesis and Developmental Interactions in Myxobacteria. Science 189 , 516–523 (1975). Thiel, T. & Pratte, B. Effect on heterocyst differentiation of nitrogen fixation in vegetative cells of the cyanobacterium Anabaena variabilis ATCC 29413. J Bacteriol 183 , 280–286 (2001). Kaiser, D. Coupling cell movement to multicellular development in myxobacteria. Nat Rev Microbiol 1 , 45–54 (2003). Barka, E. A. et al. Taxonomy, Physiology, and Natural Products of Actinobacteria. Microbiol Mol Biol Rev 80 , 1–43 (2016). Govindarajan, S., Albocher, N., Szoke, T., Nussbaum-Shochat, A. & Amster-Choder, O. Phenotypic Heterogeneity in Sugar Utilization by E. coli Is Generated by Stochastic Dispersal of the General PTS Protein EI from Polar Clusters. Front. Microbiol. 8 , 2695 (2018). Nakatani, R. J., Itabashi, M., Yamada, T. G., Hiroi, N. F. & Funahashi, A. Intercellular interaction mechanisms promote diversity in intracellular ATP concentration in Escherichia coli populations. Sci Rep 12 , 17946 (2022). Choudhary, D., Lagage, V., Foster, K. R. & Uphoff, S. Phenotypic heterogeneity in the bacterial oxidative stress response is driven by cell-cell interactions. Cell Reports 42 , 112168 (2023). Conlon, B. P. et al. Persister formation in Staphylococcus aureus is associated with ATP depletion. Nat Microbiol 1 , 16051 (2016). Buckling, A. et al. Siderophore-mediated cooperation and virulence in Pseudomonas aeruginosa: Siderophore-mediated cooperation and virulence in P. aeruginosa. FEMS Microbiology Ecology 62 , 135–141 (2007). Chubiz, J. E. C., Golubeva, Y. A., Lin, D., Miller, L. D. & Slauch, J. M. FliZ Regulates Expression of the Salmonella Pathogenicity Island 1 Invasion Locus by Controlling HilD Protein Activity in Salmonella enterica Serovar Typhimurium. J Bacteriol 192 , 6261–6270 (2010). Cahoon, L. A. & Seifert, H. S. Focusing homologous recombination: pilin antigenic variation in the pathogenic Neisseria . Molecular Microbiology 81 , 1136–1143 (2011). Davis, K. M., Mohammadi, S. & Isberg, R. R. Community Behavior and Spatial Regulation within a Bacterial Microcolony in Deep Tissue Sites Serves to Protect against Host Attack. Cell Host & Microbe 17 , 21–31 (2015). Nuss, A. M. et al. A Precise Temperature-Responsive Bistable Switch Controlling Yersinia Virulence. PLoS Pathog 12 , e1006091 (2016). McCallum, G. & Tropini, C. The gut microbiota and its biogeography. Nat Rev Microbiol 22 , 105–118 (2024). Donaldson, G. P., Lee, S. M. & Mazmanian, S. K. Gut biogeography of the bacterial microbiota. Nat Rev Microbiol 14 , 20–32 (2016). Schlaeppi, K., Dombrowski, N., Oter, R. G., Ver Loren Van Themaat, E. & Schulze-Lefert, P. Quantitative divergence of the bacterial root microbiota in Arabidopsis thaliana relatives. Proc. Natl. Acad. Sci. U.S.A. 111 , 585–592 (2014). Bai, Y. et al. Functional overlap of the Arabidopsis leaf and root microbiota. Nature 528 , 364–369 (2015). Bulgarelli, D. et al. Revealing structure and assembly cues for Arabidopsis root-inhabiting bacterial microbiota. Nature 488 , 91–95 (2012). Hou, K. et al. Microbiota in health and diseases. Sig Transduct Target Ther 7 , 135 (2022). Asnicar, F. et al. Gut micro-organisms associated with health, nutrition and dietary interventions. Nature https://doi.org/10.1038/s41586-025-09854-7 (2025) doi:10.1038/s41586-025-09854-7. Chaparro, J. M., Badri, D. V. & Vivanco, J. M. Rhizosphere microbiome assemblage is affected by plant development. The ISME Journal 8 , 790–803 (2014). Zhalnina, K. et al. Dynamic root exudate chemistry and microbial substrate preferences drive patterns in rhizosphere microbial community assembly. Nat Microbiol 3 , 470–480 (2018). Loo, E. et al. Contribution of Sugar Transporters to Spatially Organized Colonization by Microbiota Along the Longitudinal Root Axis of Arabidopsis. https://www.ssrn.com/abstract=4514131 (2023) doi:10.2139/ssrn.4514131. Hawes, M. C. et al. Extracellular DNA: The tip of root defenses? Plant Science 180 , 741–745 (2011). Beck, M. et al. Expression patterns of FLAGELLIN SENSING 2 map to bacterial entry sites in plant shoots and roots. Journal of Experimental Botany 65 , 6487–6498 (2014). Wyrsch, I., Domínguez‐Ferreras, A., Geldner, N. & Boller, T. Tissue‐specific FLAGELLIN ‐ SENSING 2 ( FLS 2) expression in roots restores immune responses in A rabidopsis fls2 mutants. New Phytologist 206 , 774–784 (2015). Zhou, F. et al. Co-incidence of Damage and Microbial Patterns Controls Localized Immune Responses in Roots. Cell 180 , 440-453.e18 (2020). Fortier, M. et al. A fine-tuned defense at the pea root caps: Involvement of border cells and arabinogalactan proteins against soilborne diseases. Front. Plant Sci. 14 , 1132132 (2023). Tsai, H.-H., Wang, J., Geldner, N. & Zhou, F. Spatiotemporal control of root immune responses during microbial colonization. Current Opinion in Plant Biology 74 , 102369 (2023). Campos, M. L., De Souza, C. M., De Oliveira, K. B. S., Dias, S. C. & Franco, O. L. The role of antimicrobial peptides in plant immunity. Journal of Experimental Botany 69 , 4997–5011 (2018). Koprivova, A. et al. Root-specific camalexin biosynthesis controls the plant growth-promoting effects of multiple bacterial strains. Proc. Natl. Acad. Sci. U.S.A. 116 , 15735–15744 (2019). Lopa, F. R., Snigdha, F. N., Lumactud, R. A. & Sikder, M. M. Harnessing Camalexin as a Sustainable and Ecofriendly Strategy to Control Harmful Phytopathogens. Plant Pathology 74 , 2463–2477 (2025). Reichling, J. Plant-Microbe Interactions and Secondary Metabolites with Antibacterial, Antifungal and Antiviral Properties. in Functions and Biotechnology of Plant Secondary Metabolites (ed. Wink, M.) 214–347 (Wiley-Blackwell, Oxford, UK, 2010). doi:10.1002/9781444318876.ch4. Bhandari, D. D., Kim, S.-J. & Brandizzi, F. Fortifying the frontier: cell wall modifications during plant immunity. Current Opinion in Plant Biology 88 , 102816 (2025). Malinovsky, F. G., Fangel, J. U. & Willats, W. G. T. The role of the cell wall in plant immunity. Front. Plant Sci. 5 , (2014). Zipfel, C. & Oldroyd, G. E. D. Plant signalling in symbiosis and immunity. Nature 543 , 328–336 (2017). Baxter, A., Mittler, R. & Suzuki, N. ROS as key players in plant stress signalling. Journal of Experimental Botany 65 , 1229–1240 (2014). Jones, J. D. G. & Dangl, J. L. The plant immune system. Nature 444 , 323–329 (2006). Macho, A. P. & Zipfel, C. Plant PRRs and the Activation of Innate Immune Signaling. Molecular Cell 54 , 263–272 (2014). Lu, Y. & Tsuda, K. Intimate Association of PRR- and NLR-Mediated Signaling in Plant Immunity. MPMI 34 , 3–14 (2021). Asai, S. & Shirasu, K. Plant cells under siege: plant immune system versus pathogen effectors. Current Opinion in Plant Biology 28 , 1–8 (2015). da Cunha, L., Sreerekha, M.-V. & Mackey, D. Defense suppression by virulence effectors of bacterial phytopathogens. Current Opinion in Plant Biology 10 , 349–357 (2007). Giraldo, M. C. & Valent, B. Filamentous plant pathogen effectors in action. Nat Rev Microbiol 11 , 800–814 (2013). Waheed, A. et al. Effector Avr4 in Phytophthora infestans Escapes Host Immunity Mainly Through Early Termination. Front Microbiol 12 , 646062 (2021). Na, R. & Gijzen, M. Escaping Host Immunity: New Tricks for Plant Pathogens. PLoS Pathog 12 , e1005631 (2016). Ruiz-Bedoya, T., Wang, P. W., Desveaux, D. & Guttman, D. S. Cooperative virulence via the collective action of secreted pathogen effectors. Nat Microbiol 8 , 640–650 (2023). Dragoš, A. et al. Division of Labor during Biofilm Matrix Production. Current Biology 28 , 1903-1913.e5 (2018). Mutlu, A. et al. Phenotypic memory in Bacillus subtilis links dormancy entry and exit by a spore quantity-quality tradeoff. Nat Commun 9 , 69 (2018). Yannarell, S. M. et al. Extensive cellular multi-tasking within Bacillus subtilis biofilms. mSystems e00891-22 (2023) doi:10.1128/msystems.00891-22. Otto, S. B. et al. Privatization of Biofilm Matrix in Structurally Heterogeneous Biofilms. mSystems 5 , e00425-20 (2020). Dergham, Y. et al. Direct comparison of spatial transcriptional heterogeneity across diverse Bacillus subtilis biofilm communities. Nat Commun 14 , 7546 (2023). Lopez, D., Vlamakis, H. & Kolter, R. Generation of multiple cell types in Bacillus subtilis . FEMS Microbiol Rev 33 , 152–163 (2009). Kearns, D. B. & Losick, R. Cell population heterogeneity during growth of Bacillus subtilis. Genes Dev 19 , 3083–3094 (2005). Stoodley, P., Sauer, K., Davies, D. G. & Costerton, J. W. Biofilms as Complex Differentiated Communities. Annu. Rev. Microbiol. 56 , 187–209 (2002). Vlamakis, H., Aguilar, C., Losick, R. & Kolter, R. Control of cell fate by the formation of an architecturally complex bacterial community. Genes Dev. 22 , 945–953 (2008). van Gestel, J., Vlamakis, H. & Kolter, R. From Cell Differentiation to Cell Collectives: Bacillus subtilis Uses Division of Labor to Migrate. PLoS Biol 13 , e1002141 (2015). van Gestel, J., Vlamakis, H. & Kolter, R. Division of Labor in Biofilms: the Ecology of Cell Differentiation. Microbiol Spectr 3 , (2015). Lord, N. D. et al. Stochastic antagonism between two proteins governs a bacterial cell fate switch. Science 366 , 116–120 (2019). Bai, U., Mandic-Mulec, I. & Smith, I. SinI modulates the activity of SinR, a developmental switch protein of Bacillus subtilis, by protein-protein interaction. Genes Dev. 7 , 139–148 (1993). Newman, J. A., Rodrigues, C. & Lewis, R. J. Molecular Basis of the Activity of SinR Protein, the Master Regulator of Biofilm Formation in Bacillus subtilis. Journal of Biological Chemistry 288 , 10766–10778 (2013). Kampf, J. et al. Selective Pressure for Biofilm Formation in Bacillus subtilis: Differential Effect of Mutations in the Master Regulator SinR on Bistability. mBio 9 , e01464-18 (2018). Chai, Y., Chu, F., Kolter, R. & Losick, R. Bistability and biofilm formation in Bacillus subtilis: Bistability and biofilm formation in Bacillus subtilis. Molecular Microbiology 67 , 254–263 (2007). Dannenberg, S., Penning, J., Simm, A. & Klumpp, S. The motility-matrix production switch in Bacillus subtilis-a modeling perspective. J Bacteriol 206 , e0004723 (2024). Norman, T. M., Lord, N. D., Paulsson, J. & Losick, R. Memory and modularity in cell-fate decision making. Nature 503 , 481–486 (2013). Arnaouteli, S., Bamford, N. C., Stanley-Wall, N. R. & Kovács, Á. T. Bacillus subtilis biofilm formation and social interactions. Nat Rev Microbiol 19 , 600–614 (2021). Xu, S. et al. The spo0A-sinI-sinR Regulatory Circuit Plays an Essential Role in Biofilm Formation, Nematicidal Activities, and Plant Protection in Bacillus cereus AR156. MPMI 30 , 603–619 (2017). McLoon, A. L., Kolodkin-Gal, I., Rubinstein, S. M., Kolter, R. & Losick, R. Spatial Regulation of Histidine Kinases Governing Biofilm Formation in Bacillus subtilis . J Bacteriol 193 , 679–685 (2011). López-Pagán, N. et al. Pseudomonas syringae subpopulations cooperate by coordinating flagellar and type III secretion spatiotemporal dynamics to facilitate plant infection. Nat Microbiol 10 , 958–972 (2025). Allard-Massicotte, R. et al. Bacillus subtilis Early Colonization of Arabidopsis thaliana Roots Involves Multiple Chemotaxis Receptors. mBio 7 , (2016). Boubsi, F. et al. Pectic homogalacturonan sensed by Bacillus acts as host associated cue to promote establishment and persistence in the rhizosphere. iScience 26 , 107925 (2023). Engelhardt, I. C., Holden, N., Daniell, T. J. & Dupuy, L. X. Mobility and growth in confined spaces are important mechanisms for the establishment of Bacillus subtilis in the rhizosphere. Microbiology 170 , (2024). Lee, Y., Kwon, S., Balaraju, K. & Jeon, Y. Influence of phenotypic variation of Paenibacillus polymyxa E681 on growth promotion in cucumbers. Front. Microbiol. 15 , 1427265 (2024). Dobrange, E. & Van Den Ende, W. Bacterial cell differentiation during plant root colonization: the putative role of fructans. Physiologia Plantarum 177 , e70095 (2025). Tian, T. et al. Sucrose triggers a novel signaling cascade promoting Bacillus subtilis rhizosphere colonization. ISME J 15 , 2723–2737 (2021). Beauregard, P. B., Chai, Y., Vlamakis, H., Losick, R. & Kolter, R. Bacillus subtilis biofilm induction by plant polysaccharides. Proceedings of the National Academy of Sciences 110 , E1621–E1630 (2013). Hoff, G. et al. Surfactin Stimulated by Pectin Molecular Patterns and Root Exudates Acts as a Key Driver of the Bacillus-Plant Mutualistic Interaction. mBio 12 , e01774-21 (2021). Gozzi, K. et al. Bacillus subtilis utilizes the DNA damage response to manage multicellular development. npj Biofilms Microbiomes 3 , 8 (2017). Kaiser, C.-F., Perilli, A., Grossmann, G. & Meroz, Y. Studying root–environment interactions in structured microdevices. Journal of Experimental Botany 74 , 3851–3863 (2023). Stanley, C. E. et al. Dual-flow-RootChip reveals local adaptations of roots towards environmental asymmetry at the physiological and genetic levels. New Phytol 217 , 1357–1369 (2018). Massalha, H., Korenblum, E., Malitsky, S., Shapiro, O. H. & Aharoni, A. Live imaging of root–bacteria interactions in a microfluidics setup. Proc Natl Acad Sci USA 114 , 4549–4554 (2017). Zengler, K. et al. EcoFABs: advancing microbiome science through standardized fabricated ecosystems. Nat Methods 16 , 567–571 (2019). Noirot-Gros, M.-F. et al. Functional Imaging of Microbial Interactions With Tree Roots Using a Microfluidics Setup. Front. Plant Sci. 11 , 408 (2020). Dai, H., Wu, B., Chen, B., Ma, B. & Chu, C. Diel Fluctuation of Extracellular Reactive Oxygen Species Production in the Rhizosphere of Rice. Environ. Sci. Technol. 56 , 9075–9082 (2022). Grossmann, G. et al. The RootChip: An Integrated Microfluidic Chip for Plant Science. The Plant Cell 23 , 4234–4240 (2011). Denninger, P. et al. Distinct RopGEFs Successively Drive Polarization and Outgrowth of Root Hairs. Current Biology 29 , 1854-1865.e5 (2019). Durrett, R. et al. Genome Sequence of the Bacillus subtilis Biofilm-Forming Transformable Strain PS216. Genome Announc 1 , (2013). Fan, B. et al. Bacillus velezensis FZB42 in 2018: The Gram-Positive Model Strain for Plant Growth Promotion and Biocontrol. Front. Microbiol. 9 , 2491 (2018). Trauth, S. & Bischofs, I. B. Ectopic Integration Vectors for Generating Fluorescent Promoter Fusions in Bacillus subtilis with Minimal Dark Noise. PLoS ONE 9 , e98360 (2014). Gibson, D. G. et al. Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat Methods 6 , 343–345 (2009). Hauser, P. M. & Karamata, D. A rapid and simple method for Bacillus subtilis transformation on solid media. Microbiology 140 , 1613–1617 (1994). Harwood, C. R. & Cutting, S. M. Molecular Biological Methods for Bacillus . (J. Wiley & sons, Chichester New York Brisbane [etc.], 1990). Koo, B.-M. et al. Construction and Analysis of Two Genome-Scale Deletion Libraries for Bacillus subtilis. Cell Systems 4 , 291-305.e7 (2017). Hoagland, D. R. & Arnon, D. I. The water-culture method for growing plants without soil. 347 , 32 pp. (1950). Guichard, M., Bertran Garcia de Olalla, E., Stanley, C. E. & Grossmann, G. Microfluidic systems for plant root imaging. in Methods in Cell Biology vol. 160 381–404 (Elsevier, 2020). Edelstein, A., Amodaj, N., Hoover, K., Vale, R. & Stuurman, N. Computer Control of Microscopes Using µManager. CP Molecular Biology 92 , (2010). D. Edelstein, A. et al. Advanced methods of microscope control using μManager software. JBM 1 , 1 (2014). Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat Methods 9 , 676–682 (2012). Jefferis, G., Kemp, S. E., Arya, S. & Mount, D. RANN: Fast Nearest Neighbour Search (Wraps ANN Library) Using L2 Metric. 2.6.2 https://doi.org/10.32614/CRAN.package.RANN (2013). Baddeley, A., Rubak, E. & Turner, R. Spatial Point Patterns: Methodology and Applications with R . (CRC Press, Boca Raton London New York, 2016). doi:10.1201/b19708. Preibisch, S., Saalfeld, S. & Tomancak, P. Globally optimal stitching of tiled 3D microscopic image acquisitions. Bioinformatics 25 , 1463–1465 (2009). Wood, S. mgcv: Mixed GAM Computation Vehicle with Automatic Smoothness Estimation. 1.9-4 https://doi.org/10.32614/CRAN.package.mgcv (2000). Branda, S. S., Chu, F., Kearns, D. B., Losick, R. & Kolter, R. A major protein component of the Bacillus subtilis biofilm matrix. Mol Microbiol 59 , 1229–1238 (2006). Verbelen, J.-P., Cnodder, T. D., Le, J., Vissenberg, K. & Baluška, F. The Root Apex of Arabidopsis thaliana Consists of Four Distinct Zones of Growth Activities: Meristematic Zone, Transition Zone, Fast Elongation Zone and Growth Terminating Zone. Plant Signaling & Behavior 1 , 296–304 (2006). Steinfeld, B. K., Cui, Q., Schmidt, T. & Bischofs, I. B. Communication Determines Population-Level Fitness under Cation Stress by Modulating the Ratio of Motile to Sessile B. Subtilis Cells. http://biorxiv.org/lookup/doi/10.1101/2021.11.30.470380 (2021) doi:10.1101/2021.11.30.470380. Davletova, S., Schlauch, K., Coutu, J. & Mittler, R. The zinc-finger protein Zat12 plays a central role in reactive oxygen and abiotic stress signaling in Arabidopsis. Plant Physiol 139 , 847–856 (2005). Torres, M. A., Dangl, J. L. & Jones, J. D. G. Arabidopsis gp91 phox homologues AtrbohD and AtrbohF are required for accumulation of reactive oxygen intermediates in the plant defense response. Proc. Natl. Acad. Sci. U.S.A. 99 , 517–522 (2002). Daudi, A. et al. The Apoplastic Oxidative Burst Peroxidase in Arabidopsis Is a Major Component of Pattern-Triggered Immunity. The Plant Cell 24 , 275–287 (2012). Podgórska, A., Burian, M. & Szal, B. Extra-Cellular But Extra-Ordinarily Important for Cells: Apoplastic Reactive Oxygen Species Metabolism. Front. Plant Sci. 8 , (2017). Arnaud, D., Deeks, M. J. & Smirnoff, N. RBOHF activates stomatal immunity by modulating both reactive oxygen species and apoplastic pH dynamics in Arabidopsis. The Plant Journal 116 , 404–415 (2023). Cheeseman, J. M. Hydrogen peroxide concentrations in leaves under natural conditions. Journal of Experimental Botany 57 , 2435–2444 (2006). Liu, P. et al. Length-based separation of Bacillus subtilis bacterial populations by viscoelastic microfluidics. Microsyst Nanoeng 8 , 7 (2022). Dion, M. F. et al. Bacillus subtilis cell diameter is determined by the opposing actions of two distinct cell wall synthetic systems. Nat Microbiol 4 , 1294–1305 (2019). Knights, H. E., Jorrin, B., Haskett, T. L. & Poole, P. S. Deciphering bacterial mechanisms of root colonization. Environmental Microbiology Reports 13 , 428–444 (2021). Yeh, Y.-H., Chang, Y.-H., Huang, P.-Y., Huang, J.-B. & Zimmerli, L. Enhanced Arabidopsis pattern-triggered immunity by overexpression of cysteine-rich receptor-like kinases. Front Plant Sci 6 , 322 (2015). Angelini, L. L. et al. Pulcherrimin protects Bacillus subtilis against oxidative stress during biofilm development. npj Biofilms Microbiomes 9 , (2023). Muratov, E., Keilholz, J., Kovács, Á. T. & Moeller, R. The biofilm matrix protects Bacillu subtilis against hydrogen peroxide. Biofilm 9 , 100274 (2025). Aldin, M. S. & Tzipilevich, E. Salmonella exploits the reactive oxygen species generated by the plant immune system to enhance colonization. Preprint at https://doi.org/10.1101/2024.03.04.583411 (2024). Kulkarni, R. et al. Cigarette Smoke Increases Staphylococcus aureus Biofilm Formation via Oxidative Stress. Infect Immun 80 , 3804–3811 (2012). Pfeilmeier, S. et al. The plant NADPH oxidase RBOHD is required for microbiota homeostasis in leaves. Nat Microbiol 6, 852–864 (2021). Ma, K.-W. et al. Coordination of microbe–host homeostasis by crosstalk with plant innate immunity. Nat. Plants 7, 814–825 (2021). Tables Tables are available in the Supplementary Files section. Additional Declarations There is NO Competing Interest. Supplementary Files SupplementalFigureslegendsandTables.docx 20260112S1.png Supplemental Figure 1 20260112S2.png Supplemental Figure 2 20260112S3.png Supplemental Figure 3 20260112S4.png Supplemental Figure 4 Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8407390","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":597732496,"identity":"912966bf-60de-4d6d-a6ed-bff542aa46db","order_by":0,"name":"Christian-Frederic Kaiser","email":"","orcid":"https://orcid.org/0000-0001-9321-1568","institution":"Heinrich Heine University Duesseldorf","correspondingAuthor":false,"prefix":"","firstName":"Christian-Frederic","middleName":"","lastName":"Kaiser","suffix":""},{"id":597732497,"identity":"dc1930af-a2d0-41d4-8530-f939f4909bda","order_by":1,"name":"Sabrina Egli","email":"","orcid":"https://orcid.org/0000-0001-5005-0127","institution":"Heinrich Heine University Duesseldorf","correspondingAuthor":false,"prefix":"","firstName":"Sabrina","middleName":"","lastName":"Egli","suffix":""},{"id":597732498,"identity":"10b558ad-dd7d-4ae6-8d99-4a037bc87e1b","order_by":2,"name":"Mirco Keilhammer","email":"","orcid":"https://orcid.org/0009-0002-4205-670X","institution":"Heinrich Heine University Duesseldorf","correspondingAuthor":false,"prefix":"","firstName":"Mirco","middleName":"","lastName":"Keilhammer","suffix":""},{"id":597732499,"identity":"d2f16a6c-837d-4c35-bc43-97994ea44279","order_by":3,"name":"M. Carolina Elizondo-Cantú","email":"","orcid":"https://orcid.org/0000-0002-9917-4830","institution":"Max Planck Institute for Terrestrial Microbiology","correspondingAuthor":false,"prefix":"","firstName":"M.","middleName":"Carolina","lastName":"Elizondo-Cantú","suffix":""},{"id":597732500,"identity":"ab3a50f1-107b-49c3-826a-5612b1e4ffa0","order_by":4,"name":"Milan Župunski","email":"","orcid":"","institution":"Heinrich Heine University Duesseldorf","correspondingAuthor":false,"prefix":"","firstName":"Milan","middleName":"","lastName":"Župunski","suffix":""},{"id":597732501,"identity":"a128f275-4990-486a-a0f6-754637e39cf0","order_by":5,"name":"Jacqueline Esser","email":"","orcid":"","institution":"Heinrich Heine University Duesseldorf","correspondingAuthor":false,"prefix":"","firstName":"Jacqueline","middleName":"","lastName":"Esser","suffix":""},{"id":597732502,"identity":"fc98b9a7-f119-40c1-a2d4-393e964b9995","order_by":6,"name":"Sebastian Erdrich","email":"","orcid":"https://orcid.org/0000-0001-6820-5473","institution":"Heinrich Heine University Duesseldorf","correspondingAuthor":false,"prefix":"","firstName":"Sebastian","middleName":"","lastName":"Erdrich","suffix":""},{"id":597732503,"identity":"707a1968-85e0-4f14-a446-adab0a3bd16f","order_by":7,"name":"Charlotte Kaspar","email":"","orcid":"","institution":"Max Planck Institute for Terrestrial Microbiology","correspondingAuthor":false,"prefix":"","firstName":"Charlotte","middleName":"","lastName":"Kaspar","suffix":""},{"id":597732504,"identity":"77e20710-3da3-49b1-9a19-0e87b548ef8a","order_by":8,"name":"Tamara Schmidt","email":"","orcid":"","institution":"Heidelberg University","correspondingAuthor":false,"prefix":"","firstName":"Tamara","middleName":"","lastName":"Schmidt","suffix":""},{"id":597732505,"identity":"44fba3b3-bb55-4e8b-9c2f-edb994c7564e","order_by":9,"name":"Ilka Bischofs","email":"","orcid":"https://orcid.org/0000-0001-5905-7106","institution":"Max Planck Institute for Terrestrial Microbiology","correspondingAuthor":false,"prefix":"","firstName":"Ilka","middleName":"","lastName":"Bischofs","suffix":""},{"id":597732495,"identity":"c614d2d7-32f8-417d-b4aa-88a30cccd420","order_by":10,"name":"Guido Grossmann","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA7UlEQVRIiWNgGAWjYFACxgYJZK6cAYjLQ4oWYyK0MDCgaEncQEiLfERy440PDLX5/PzHn274ueNO+nbpHgOGNxW4tRjeSGy2nMFw3HLmjByzm71nnuXunHPGgHHOGTxaZiS2SfMwHDMwuMHDdoO37XDuhhs5Bsy8bcRoOX/82c2/bYfTDcBa/uHxiwRYS42BwYEEs9tAWxIgWhpwazHgeQj0i8EBA0mgX27Lth023HAjreDgnGN4bGlPf3jjQ0WdATDEnt1823ZY3uBG8sYHb2rw2HIATB5GFT2AWwPQFoij6/CpGQWjYBSMgpEOAG9QV7GBOFBWAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0001-7529-9244","institution":"Heinrich Heine University Duesseldorf","correspondingAuthor":true,"prefix":"","firstName":"Guido","middleName":"","lastName":"Grossmann","suffix":""}],"badges":[],"createdAt":"2025-12-19 18:12:02","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8407390/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8407390/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108978023,"identity":"3f57e9b5-5d39-4bb4-b58f-88c16e9fc7f5","added_by":"auto","created_at":"2026-05-11 11:33:45","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":4484436,"visible":true,"origin":"","legend":"\u003cp\u003eFlagellated and matrix-producing cells of \u003cem\u003eBacillus sp.\u003c/em\u003e are spatially organized along the longitudinal axis.\u003c/p\u003e\n\u003cp\u003eColonization of 10-day-old Arabidopsis thaliana Col-0 primary roots in the RootChip8NF was imaged 3 days post-inoculation with \u003cem\u003eB. subtilis\u003c/em\u003e PS-216 wild-type (N = 37), \u003cem\u003eΔsinI \u003c/em\u003emutant (N = 13), or \u003cem\u003eB. velezensis\u003c/em\u003e FZB42 (N=19) dual reporter strains (P\u003csub\u003e\u003cem\u003ehag\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e-gfpmut3\u003c/em\u003e, P\u003csub\u003e\u003cem\u003etapA\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e-mCherry\u003c/em\u003e). Multiple fields of view spanning the first 5 mm from the root tip were stitched and maximum-intensity projected (0.5 µm/px; 80–180 Z-planes; 5 µm steps). Bacterial clusters (3,000–30,000 per root), including single cells and aggregates, were segmented by fluorescence intensity and mapped relative to the root, e.g. the planar segment covered by the root, and tip.\u003c/p\u003e\n\u003cp\u003e(A) Composite image showing brightfield (right panel), and a merge of GFP (P\u003csub\u003e\u003cem\u003ehag\u003c/em\u003e\u003c/sub\u003e-\u003cem\u003egfpmut3\u003c/em\u003e, green), and RFP (P\u003csub\u003e\u003cem\u003etapA\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e-mCherry\u003c/em\u003e, magenta) fluorescence along the first 5 mm of the root.\u003c/p\u003e\n\u003cp\u003e(B) Averaged 2D kernel density estimation (KDE) of \u003cem\u003eBacillus \u003c/em\u003ecell type composition (PS-216, N =37; FZB42, N = 19), based on the relative share of matrix-producing cells in the total bacterial coverage (AreaMatrixp./AreaTotal; green = 0%, magenta = 100%). X-axis: distance to the root; Y-axis: distance from root tip.\u003c/p\u003e\n\u003cp\u003e(C) Mean coverage of the rhizoplane, e.g. the root-covered section of the planar Z-projection, by flagellated (green), matrix-producing (magenta), and total fluorescent cells (black), binned every 100 µm along the root axis across root replicates (PS-216, N =37; FZB42, N = 19). Shaded area: standard error (SE). β-regression statistics are displayed in Tab. 5.\u003c/p\u003e\n\u003cp\u003e(D) Mean coverage of flagellated (green) and matrix-producing (magenta) PS-216 cells relative distance to the root, in 50 µm × 5,000 µm bins from the root midline, across root replicates (N = 37). Dotted line marks the end of the root section (X = 0 µm), separating rhizoplane (X ≤ 0 µm) and rhizosphere (X \u0026gt; 0 µm). Shaded area: SE.\u003c/p\u003e\n\u003cp\u003e(E) Fraction of matrix-producing cells on the rhizoplane of PS-216 (red, N = 37) and FZB42 (blue, N = 19), relative to total coverage, in 100 µm bins. Shaded area: SE.\u003c/p\u003e\n\u003cp\u003e(F) Total coverage of fluorescent wildtype PS-216 (black, N = 37)) and \u003cem\u003eΔsinI\u003c/em\u003e (orange, N = 13) cells relative to the root, binned as in (E), across root replicates (Tab. 6). Line and shaded area as in (D). The inserted boxplot displays the average bacterial coverage in rhizoplane and rhizosphere fractions for WT PS-216 (black, N = 37) and \u003cem\u003eΔsinI\u003c/em\u003e (orange, N = 13). Statistical comparison via two-sided Wilcoxon rank-sum test, exact p-values are denoted in the figure.\u003c/p\u003e","description":"","filename":"20260112Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-8407390/v1/4752e8ce32c849d670829b8f.png"},{"id":108973925,"identity":"f997d36f-f27c-44d8-be89-a762556cb9a1","added_by":"auto","created_at":"2026-05-11 10:44:56","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":2720310,"visible":true,"origin":"","legend":"\u003cp\u003eCell type differentiation is spatially confined to the early maturation zone\u003c/p\u003e\n\u003cp\u003eA single 10-day-old \u003cem\u003eA. thaliana\u003c/em\u003e Col-0 primary root (N = 1) was imaged in a structured µRhiztron microfluidic slide for 14 h at 3 dpi after inoculation with the \u003cem\u003eB. subtilis\u003c/em\u003e PS-216 dual reporter (P\u003csub\u003e\u003cem\u003ehag\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e-gfpmut3\u003c/em\u003e, P\u003csub\u003e\u003cem\u003etapA\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e-mCherry\u003c/em\u003e). Images (1 × 1 mm FOV, 1 µm/px) were acquired hourly, stitched, and maximum-intensity projected. Bacterial clusters (N = 992), including single cells and aggregates, were segmented by fluorescence intensity; adherence was defined as the first above-background signal. A fluorescence intensity threshold was set based on the mean and standard deviation of \u003cem\u003emCherry \u003c/em\u003efluorescence of GFP-expressing clusters. Cells were classified as flagellated (\u003cem\u003eGFP\u003c/em\u003e) below this threshold and matrix-producing (\u003cem\u003emCherry\u003c/em\u003e) once this threshold was reached. Cells that transitioned across this threshold were branded as “switchers”.\u003c/p\u003e\n\u003cp\u003e(A) Overview of the first 1.5 - 2.5 mm of the primary root at 0 and 14 h (brightfield = gray, GFP = green, mCherry = magenta); crop for (B) indicated in an orange square. Below are false-color projections of a close-up of root hairs colonized by a microcolony transitioning from flagellated to matrix-producing. The white bar indicates a size of 10 µm. The white square in the upper right panel indicates the isolated region.\u003c/p\u003e\n\u003cp\u003e(B) Average fluorescent trajectories across individual clusters (N = 992) of P\u003csub\u003e\u003cem\u003ehag\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e-gfpmut3 \u003c/em\u003e(green) and P\u003csub\u003e\u003cem\u003etapA\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e-mCherry \u003c/em\u003e(magenta) of flagellated and switching cells following their time of adherence. The shaded area indicates SE. Fluorescence was scaled to the average minimum (0) and maximum (1) of both fluorescent channels, respectively.\u003c/p\u003e\n\u003cp\u003e(C) Average population structure separated into flagellated (green) and matrix-producing cells\u003cem\u003e \u003c/em\u003e(magenta) relative to the growth time of the root and distance from the root tip (N = 224 – 991).\u003c/p\u003e\n\u003cp\u003e(D) 2D density projection of switching events (N = 361, first timepoint above mCherry threshold) plotted by distance from the root tip (Y-axis) and proximity to the root (X-axis). Density is color-coded from purple (low) to yellow (high); individual data points are shown in white (refer to Tab. 7 for summary statistics).\u003c/p\u003e\n\u003cp\u003e(E) Generalized additive model (GAM) of switching outcomes as a function of final distance from the root tip. The curve shows the GAM-predicted probability that a cluster had undergone at least one switching event, with shaded areas indicating 95% confidence intervals. Stacked plots show counts of switched (magenta) and non-switched (green) clusters in the final imaging frame; maximum elevation reflects total cluster count per bin.\u003c/p\u003e","description":"","filename":"20260112Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-8407390/v1/2597564b5ed25c56a1d215e8.png"},{"id":108973926,"identity":"3aa71dd0-ee12-416e-bc83-5ed6366f8973","added_by":"auto","created_at":"2026-05-11 10:44:56","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":2127176,"visible":true,"origin":"","legend":"\u003cp\u003eRoot colonization induces zonal, RBOH-dependent ROS production.\u003c/p\u003e\n\u003cp\u003eA. thaliana seedlings were grown for 7 days on sterile nylon meshes (100 µm pore diameter) over ½HS0 agar and treated by flood inoculation with mock (1/2HS0 medium) or \u003cem\u003eB. subtilis \u003c/em\u003ePS-216 (initial OD\u003csub\u003e600\u003c/sub\u003e 0.02). After 1–3 days post-inoculation (dpi), roots were incubated with 2.5 µM H\u003csub\u003e2\u003c/sub\u003eDCFDA for 20 minutes and imaged in the GFP channel. Midlines spanning 100 µm (25 px) were extracted per root. Each individual root was imaged and measured once and treated as a biological replicate.\u003c/p\u003e\n\u003cp\u003e(A) Representative H\u003csub\u003e2\u003c/sub\u003eDCFDA fluorescence images at 0–3 dpi in mock- and PS-216-inoculated roots (\u003cu\u003eUntreated\u003c/u\u003e: 0 dpi – N = 102; \u003cu\u003eMock\u003c/u\u003e: 1 dpi N = 101, 2 dpi N = 101, 3 dpi N = 85; \u003cu\u003ePS-216\u003c/u\u003e: 1 dpi N = 108, 2 dpi N = 105, 3 dpi N = 93; with 5 meshes per day per condition). Color legend shown to the right. Roots were inoculated with an OD600nm of 0.02.\u003c/p\u003e\n\u003cp\u003e(B) Mean H\u003csub\u003e2\u003c/sub\u003eDCFDA intensity profiles (± SE) over time (0–3 dpi) for mock and B. subtilis PS-216 treatment.\u003c/p\u003e\n\u003cp\u003e(C) Peak H\u003csub\u003e2\u003c/sub\u003eDCFDA fluorescence normalized to 0 dpi untreated controls. Individual roots are shown as violin and dot plots. Coloring identical to (A). The same 0 dpi control group is shown for reference in both panels. The linear regression model is shown as a red line (Tab. 9).\u003c/p\u003e\n\u003cp\u003e(D)\u0026nbsp; Representative H\u003csub\u003e2\u003c/sub\u003eDCFDA fluorescence images at 3 dpi PS-216 inoculated of Col-0 (N = 29), \u003cem\u003erbohD \u003c/em\u003e(N = 39)\u003cem\u003e, rbohD/F \u003c/em\u003e(N = 39)\u003cem\u003e, \u003c/em\u003eand \u003cem\u003erbohF \u003c/em\u003e(N = 19). The color scaling is identical across images and to the scaling shown in Panel A.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e(E) Mean H\u003csub\u003e2\u003c/sub\u003eDCFDA intensity profiles (± SE) over time for Col-0, \u003cem\u003erbohD, rbohD/F, \u003c/em\u003eand \u003cem\u003erbohF\u003c/em\u003e after 3 dpi of \u003cem\u003eB. subtilis \u003c/em\u003ePS-216 treatment.\u003c/p\u003e\n\u003cp\u003e(F) Mean H\u003csub\u003e2\u003c/sub\u003eDCFDA fluorescence intensity compared against the Col-0 reference group. Pairwise Two-sided Wilcoxon test with Benjamini–Hochberg correction against the Col-0 reference group, exact adjusted p-values are denoted in the figure (Tab. 10).\u003c/p\u003e","description":"","filename":"20260112Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-8407390/v1/8c17d8ca80aae2460f6815e9.png"},{"id":108973927,"identity":"2e5578ed-a3f0-4f0d-9350-6df6a99272a5","added_by":"auto","created_at":"2026-05-11 10:44:56","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":2267789,"visible":true,"origin":"","legend":"\u003cp\u003eTransition to matrix-producing cells is promoted by ROS exposure\u003c/p\u003e\n\u003cp\u003eExponentially growing B. subtilis PS-216 dual-reporter cells (P\u003cem\u003ehag-gfpmut3\u003c/em\u003e, P\u003cem\u003etapA-mCherry\u003c/em\u003e) were used to inoculate 96-well plates (200 µL ½HS0 medium supplemented with 0.5% glycerol and glutamate) at an initial OD\u003csub\u003e600\u003c/sub\u003e of 0.05, with hydrogen peroxide concentrations ranging from 38 nM to 1 mM. To minimize edge effects, only inner wells were analyzed. Cultures were incubated at 25 °C without shaking in a plate reader for 48 h, with OD₆₀₀, GFP, and mCherry fluorescence recorded every 5 min (N = 9 per condition).\u003c/p\u003e\n\u003cp\u003e(A–D) Mean OD\u003csub\u003e600\u003c/sub\u003e (A), OD\u003csub\u003e600\u003c/sub\u003e-normalized GFP (\u003cem\u003ehag\u003c/em\u003e) fluorescence (B), and OD\u003csub\u003e600\u003c/sub\u003e-normalized mCherry (tapA) fluorescence (C) over time, and fluorescence ratio of \u003cem\u003emCherry\u003c/em\u003e-to-\u003cem\u003eGFP\u003c/em\u003e over OD₆₀₀, shown with standard error (±SE) indicated by shaded areas. The shared color scale represents the initial H₂O₂ concentrations and is shown in the upper right corner.\u003c/p\u003e\n\u003cp\u003e1 µL of B. subtilis PS-216 cultures was spotted onto 1.5% ½HS0 agarose pads at 12 h intervals (N = 10) during growth in the same multi-well setup (A-D) with either 0 or 156 µM H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2 \u003c/sub\u003esupplementation. Entire spots were imaged across multiple fields of view and stitched using NIS Elements (Large Image Mode). Fluorescent regions were segmented and background-corrected relative to untreated agarose pads. Simultaneously, fluorescence and OD₆₀₀ were monitored in the plate reader.\u003c/p\u003e\n\u003cp\u003e(E) Representative merges stitches of the spotted culture material based on the GFP (green, P\u003csub\u003e\u003cem\u003ehag\u003c/em\u003e\u003c/sub\u003e) and RFP signal (magenta, P\u003csub\u003e\u003cem\u003etapA\u003c/em\u003e\u003c/sub\u003e) channels for untreated and H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e-treated samples.\u003c/p\u003e\n\u003cp\u003e(F) Mean estimated cell density over time in the untreated and H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e-treated sample based on the planar area covered by GFP-expressing (green), mCherry-expressing (magenta) and all fluorescent cells (black, Total) clusters from 1 µL of culture material based on an estimated cell size of 4 µm\u003csup\u003e2\u003c/sup\u003e. The shaded area displays ±SE.\u003c/p\u003e\n\u003cp\u003e(G, I) Mean OD\u003csub\u003e600\u003c/sub\u003e (±SE) with (red) and without H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2 \u003c/sub\u003e(black) supplementation and fluorescence ratio of \u003cem\u003emCherry\u003c/em\u003e-to-\u003cem\u003eGFP\u003c/em\u003e over time, sampling points indicated as dotted vertical lines.\u003c/p\u003e\n\u003cp\u003e(H) Mean (±SE) share of matrix producer covered areas in the total area coverage of 1 µL culture spotted onto agarose with (red) and without H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2 \u003c/sub\u003e(black) supplementation. Standard error is only visible where consistent \u0026gt;0 µm2 coverage was detected in the \u003cem\u003emCherry \u003c/em\u003echannel.\u003c/p\u003e\n\u003cp\u003e(J) Mean OD600 normalized to the OD600 at timepoint 0 over time, for wildtype (black), \u003cem\u003eΔsinI\u003c/em\u003e (green), and \u003cem\u003eΔsinR \u003c/em\u003e(magenta). Similarly to the experimental setup of panels A-D, the knockouts mutants \u003cem\u003eΔsinI \u003c/em\u003eand \u003cem\u003eΔsinR \u003c/em\u003ewere subjected to 0, 15,6, and 313 µM of H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2 \u003c/sub\u003e.\u003c/p\u003e","description":"","filename":"20260112Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-8407390/v1/9a8624a3ad2de71372557417.png"},{"id":108973933,"identity":"9dbbf420-2eb5-4a9c-b9f2-c4c2ee184e75","added_by":"auto","created_at":"2026-05-11 10:44:56","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":2257463,"visible":true,"origin":"","legend":"\u003cp\u003eRBOH D alters the population structure of \u003cem\u003eB. subtilis \u003c/em\u003eroot colonization by promoting the flagellated subpopulation\u003c/p\u003e\n\u003cp\u003eSpatial mapping of \u003cem\u003eB. subtilis\u003c/em\u003e dual-reporter distribution, as in Fig. 1, on 10-day-old \u003cem\u003eA. thaliana\u003c/em\u003e primary roots of Col-0 (N = 37), \u003cem\u003erbohD\u003c/em\u003e(N = 22), \u003cem\u003erbohD/F\u003c/em\u003e (N = 28), and \u003cem\u003erbohF\u003c/em\u003e (N =20) mutants. The utilized Col-0 dataset is identical to Fig. 1. A statistical summary can be found respectively in Tab. 11 – 15.\u003c/p\u003e\n\u003cp\u003e(A) Representative maximum-intensity projections of wildtype and mutant roots colonized with the \u003cem\u003eB. subtilis\u003c/em\u003e PS-216 dual reporter strain (3 dpi) (5 mm root, 0.5 µm/px). The images are false-color merges of the GFP (green), RFP (magenta), and brightfield (grey) images.\u003c/p\u003e\n\u003cp\u003e(B) Total, flagellated, and matrix producer coverage across the microfluidic channel (1,000 x 5,000 µm, w x h) normalized to the assessed area. Pairwise Two-sided Wilcoxon test with Benjamini–Hochberg correction against the. Col-0 reference group, exact adjusted p-values are denoted in the figure (Tab. 11).\u003c/p\u003e\n\u003cp\u003e(C) Averaged 2D KDE of the population structure, shown as the proportion of matrix producers relative to total bacterial coverage (color gradient: green = 0%, magenta = 100%).\u003c/p\u003e\n\u003cp\u003e(D) Mean coverage of flagellated (green) and matrix-producing (magenta) PS-216 cells relative distance to the root, in 50 µm × 5,000 µm bins from the root midline. Dotted line marks the edge of the root segment (X = 0 µm), separating rhizoplane (X ≤ 0 µm) and rhizosphere (X \u0026gt; 0 µm). Shaded area: SE.\u003c/p\u003e\n\u003cp\u003e(E) Flagellated, matrix producer coverage, and share of matrix-producing cells in the rhizoplane-associated population normalized to the root area. Pairwise Two-sided Wilcoxon test with Benjamini–Hochberg correction vs. Col-0, exact adjusted p-values are denoted in the figure (Tab. 12).\u003c/p\u003e\n\u003cp\u003e(F) Mean coverage of the rhizoplane by flagellated (green), matrix-producing (magenta), and total fluorescent cells (black), binned every 100 µm along the root axis. Shaded area: standard error (SE).\u003c/p\u003e\n\u003cp\u003e(G) Matrix producer fraction across three five-sized zones from the root tip (zones as in \u003cstrong\u003eB\u003c/strong\u003e). Points denote biological replicates. Significance versus Col-0 was assessed with a Benjamini-Hochberg-adjusted pair-wise two-sided Wilcoxon test against the Col-0 reference group, exact adjusted p-values are denoted in the figure (Tab. 13).\u003c/p\u003e\n\u003cp\u003e(H) Mean (+/- SE) β-regression models of the matrix producer share on the rhizoplane relative to distance from the root tip, based on averages from 100 × 200 µm bins from the root midline (Tab. 14).\u003c/p\u003e\n\u003cp\u003e(I) Comparative analysis of the β-regression models of the matrix producer shares relative to distance from the root tip across genotypes, assessing elevation, intercept, and slope; significance testing against Col-0 as in (C).\u003c/p\u003e","description":"","filename":"20260112Figures5.png","url":"https://assets-eu.researchsquare.com/files/rs-8407390/v1/44f3e3150570a4ecf2c29d29.png"},{"id":109204637,"identity":"2268658d-a2e3-4b57-ae83-28f250aea35a","added_by":"auto","created_at":"2026-05-13 15:01:40","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":14841126,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8407390/v1/0a5ef57c-3233-4c61-ba25-179e61bce4a0.pdf"},{"id":108973924,"identity":"df694271-ca8e-435d-9f01-72b3f248627b","added_by":"auto","created_at":"2026-05-11 10:44:55","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":58552,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cbr\u003e\u003c/p\u003e","description":"","filename":"SupplementalFigureslegendsandTables.docx","url":"https://assets-eu.researchsquare.com/files/rs-8407390/v1/7358e9f8b34a09bc8f53f154.docx"},{"id":108973928,"identity":"8688d794-6ef2-4af4-a414-de0a50b9d5bd","added_by":"auto","created_at":"2026-05-11 10:44:56","extension":"png","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":2859724,"visible":true,"origin":"","legend":"\u003cp\u003eSupplemental Figure 1\u003c/p\u003e","description":"","filename":"20260112S1.png","url":"https://assets-eu.researchsquare.com/files/rs-8407390/v1/32b312fa0d32ec6a7e3c1073.png"},{"id":108978057,"identity":"11bd5be7-3745-4e86-ba8c-fa5a0e43d1ea","added_by":"auto","created_at":"2026-05-11 11:33:52","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":2499443,"visible":true,"origin":"","legend":"\u003cp\u003eSupplemental Figure 2\u003c/p\u003e","description":"","filename":"20260112S2.png","url":"https://assets-eu.researchsquare.com/files/rs-8407390/v1/2fe952f1cf87a973d6d596ed.png"},{"id":108973931,"identity":"ffd7ef69-c823-49e9-b736-5fbb4c5f26c5","added_by":"auto","created_at":"2026-05-11 10:44:56","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":1526643,"visible":true,"origin":"","legend":"\u003cp\u003eSupplemental Figure 3\u003c/p\u003e","description":"","filename":"20260112S3.png","url":"https://assets-eu.researchsquare.com/files/rs-8407390/v1/1a1295a0b329496b65f7c921.png"},{"id":108973929,"identity":"f494396f-c16d-436f-9641-f37eef0fc90d","added_by":"auto","created_at":"2026-05-11 10:44:56","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":611992,"visible":true,"origin":"","legend":"\u003cp\u003eSupplemental Figure 4\u003c/p\u003e","description":"","filename":"20260112S4.png","url":"https://assets-eu.researchsquare.com/files/rs-8407390/v1/8db223958dfaca86d8c5b074.png"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Root-derived reactive oxygen species spatially pattern bacterial cell-type differentiation during root colonization","fulltext":[{"header":"Introduction","content":"\u003cp\u003eDynamic and diverse environments challenge bacterial populations. In response, bacterial species have evolved phenotypic heterogeneity, variation among isogenic cells, driven by stochastic fluctuations in signaling, genetic noise, microenvironmental variation, and cell\u0026ndash;cell interactions \u003csup\u003e1\u0026ndash;4\u003c/sup\u003e. The formation of subpopulations affects diverse bacterial traits, such as energy metabolism and lifestyle shifts, and is evident across taxa such as enterobacteria, cyanobacteria, myxobacteria, streptomyces, and bacilli \u003csup\u003e5\u0026ndash;12\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003ePhenotypic heterogeneity is particularly advantageous during host colonization. In mammalian systems, bacterial subpopulations evade immune detection, tolerate antimicrobials and immune effectors by entering persistent states, dividing labor, or forming biofilms \u003csup\u003e1,4,13\u003c/sup\u003e. \u003cem\u003eSalmonella\u003c/em\u003e, \u003cem\u003eNeisseria\u003c/em\u003e, \u003cem\u003eYersinia, \u003c/em\u003eand \u003cem\u003ePseudomonas\u003c/em\u003e use such strategies to colonize challenging host niches \u003csup\u003e14\u0026ndash;18\u003c/sup\u003e. The physiological importance of structured host-microbe interactions in commensal and mutualistic scenarios is particularly evident in the intestinal biogeography and its impact on human health \u003csup\u003e19,20\u003c/sup\u003e. Phylogenetic heterogeneity in host colonization has drawn major attention in animal and plant systems in recent years \u003csup\u003e21\u0026ndash;25\u003c/sup\u003e. However, phenotypic heterogeneity as an adaptive strategy of host-associated microbiota and how host-derived cues are integrated remains largely unexplored.\u003c/p\u003e\n\u003cp\u003eJust as the complex biogeography of the mammalian intestine, the plant rhizosphere offers a similarly complex and compartmentalized environment, with its microbiota structured by gradients in nutrients, oxygen, exudates, and immune activity. These gradients shape microbial niches and affect colonization patterns \u003csup\u003e26\u0026ndash;28\u003c/sup\u003e. Plant immune responses are also spatially structured along the root\u0026rsquo;s axis, generating microenvironments of antimicrobial activity, immune perception, or oxidative stress that can provide positional information for colonizing bacteria\u003csup\u003e29\u0026ndash;34\u003c/sup\u003e. Just like the multi-layered mammalian immunity, plants produce antimicrobials \u003csup\u003e35\u0026ndash;38\u003c/sup\u003e, modify cell walls \u003csup\u003e39,40\u003c/sup\u003e, and release reactive oxygen species (ROS) \u003csup\u003e41,42\u003c/sup\u003e in response to microbial epitopes \u003csup\u003e43\u0026ndash;45\u003c/sup\u003e. These immune constraints have selected adaptive mechanisms in microbes, including effector proteins \u003csup\u003e46\u0026ndash;48\u003c/sup\u003e, altered epitopes \u003csup\u003e49,50\u003c/sup\u003e, or cooperative behavior \u003csup\u003e51\u003c/sup\u003e. Together, these observations place host-derived immune activity not only as a shared selective pressure but also as a potential spatial organizer of bacterial heterogeneity in the rhizosphere.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eBacillus subtilis\u003c/em\u003e (\u003cem\u003eB. subtilis\u003c/em\u003e), a common gram-positive model strain, demonstrates prominent phenotypic heterogeneity through variation in matrix production, the timing of sporulation, and multi-tasking across subpopulations \u003csup\u003e52\u0026ndash;55\u003c/sup\u003e. Functionally distinct cell types arise from defined developmental trajectories, giving rise to flagellated, matrix-producing, or antibiotic-secreting subpopulations \u003csup\u003e54,56\u0026ndash;58\u003c/sup\u003e. Division of labor, coordinated migration, structural maturation, and local microenvironmental gradients (e.g., O\u003csub\u003e2\u003c/sub\u003e, nutrients, pH) collectively generate spatial and temporal differentiation patterns within multicellular populations like biofilms \u003csup\u003e59\u0026ndash;62\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eStructured biofilm formation is dependent on the lifestyle transition from flagellated cells to matrix-producing cell chains, governed by the bistable SinI\u0026ndash;SinR regulatory circuit \u003csup\u003e63\u0026ndash;67,67,68\u003c/sup\u003e. This bistability enables a noise-driven individual cell fate decision to a matrix-producing or flagellated state, maintaining cellular heterogeneity \u003csup\u003e68,69\u003c/sup\u003e, which, through the regulation of SinI expression, integrates environmental feedback into the \u0026ldquo;decision-making\u0026rdquo; process \u003csup\u003e70\u0026ndash;72\u003c/sup\u003e. While such structured differentiation is well established on abiotic substrates, how these developmental programs manifest and are regulated within organismal interactions and dynamically structured host environmentsremains far less understood.\u003c/p\u003e\n\u003cp\u003eStill, phenotypic heterogeneity is rarely explored in rhizobacteria, with recent studies exemplifying the role of phenotypic heterogeneity during plant infection in Pseudomonadaceae \u003csup\u003e51,73\u003c/sup\u003e, while cellular subpopulations in \u003cem\u003ePaenibacillus polymyxa\u003c/em\u003e and \u003cem\u003eB. subtilis\u003c/em\u003e have been linked to plant-growth promotion and colonization \u003csup\u003e52,74\u0026ndash;78\u003c/sup\u003e. Also, phenotypic heterogeneity in \u003cem\u003eB. subtilis\u003c/em\u003e is regulated by root-associated factors like sucrose \u003csup\u003e79\u003c/sup\u003e, polysaccharides \u003csup\u003e59,80,81\u003c/sup\u003e, and reactive oxygen species (ROS) \u003csup\u003e82\u003c/sup\u003e, pointing towards possible interactions between immunity and phenotypic heterogeneity. Despite \u003cem\u003eBacillus \u003c/em\u003ephenotypic heterogeneity being observed in the rhizosphere, whether cell differentiation is merely incidental or regulated spatially by plant-derived cues like immunity remains unclear.\u003c/p\u003e\n\u003cp\u003eStudying these interactions in soil is challenging due to opacity and structural complexity. Microfluidic platforms offer a solution by mimicking rhizosphere conditions (reviewed in Kaiser \u003cem\u003eet al. \u003c/em\u003e2023)\u003csup\u003e83\u003c/sup\u003e while enabling high-resolution imaging of root\u0026ndash;bacteria interactions \u003csup\u003e84\u0026ndash;88\u003c/sup\u003e. In this study, we employ novel microfluidic devices based on the RootChip technology \u003csup\u003e89,90\u003c/sup\u003e and cell type-specific reporters in \u003cem\u003eB. subtilis\u003c/em\u003e PS-216 and \u003cem\u003eB. velezensis \u003c/em\u003eFZB42 to test whether spatially patterned host-derived ROS regulate bacterial differentiation during colonization.\u003c/p\u003e\n\u003cp\u003eWe report that plant-derived ROS shapes the balance between flagellated and matrix-producing cells, identifying ROS as a major regulator of bacterial heterogeneity in the rhizosphere. Our findings support a model that bacterial acclimatization to oxidative stress results in spatially organized plant-microbe interactions by modulating individual cellular decisions, rather than merely limiting bacterial proliferation.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003ch3\u003eBacterial Strain Construction \u003c/h3\u003e\n\u003cp\u003eAll \u003cem\u003eB. subtilis\u003c/em\u003e strains are derivatives of PS-216, a natural isolate extracted from Slovenian riverbank soil\u003csup\u003e91\u003c/sup\u003e and \u003cem\u003eB. velezensis \u003c/em\u003eFZB42\u003csup\u003e92\u003c/sup\u003e a plant-growth-promoting relative, respectively, and are listed in Table 1. Integration vectors for generating fluorescent promoter fusions were constructed with the pXFP_Star system\u003csup\u003e93\u003c/sup\u003e and are listed in Table 2. For \u003cem\u003eB. velezensis\u003c/em\u003e we first engineered a new integration vector pRFP_Star_FZB42 by replacing the \u003cem\u003eB. subtilis amyE\u003c/em\u003e sites in pRFP_Star\u003csup\u003e53\u003c/sup\u003e with corresponding sequences from \u003cem\u003eB. velezensis \u003c/em\u003eusing Gibson assembly\u003csup\u003e94\u003c/sup\u003e. The new vector is available from the Bacillus Genetic Stock Center (http://bgsc.org) under Accession Number ECE796. In general, promoter regions contain 100-300 bp upstream of the ribosome binding site and were amplified from genomic DNA using primers (Table 3) and were inserted into the LIC site of pXFP_Star by ligation-independent cloning (LIC). The small cloning site of pXFP_Star was used to combine different promoter fusions using restriction enzyme ligation cloning, respectively. Plasmids were verified by analytical restriction digestion and sequencing before transforming in \u003cem\u003eB. subtilis \u003c/em\u003evia one-step transformation\u003csup\u003e95\u003c/sup\u003e and into \u003cem\u003eB. velezensis\u003c/em\u003e by a two-step transformation method\u003csup\u003e96\u003c/sup\u003e, respectively. Correct integration of fluorescent promoter fusions into \u003cem\u003eamyE\u003c/em\u003e was verified by PCRs\u003csup\u003e93\u003c/sup\u003e. Knockouts in PS-216 were constructed by replacing the gene of interest with a kanamycin resistance cassette by transformation with a PCR product from the corresponding knockout from the \u003cem\u003eB. subtilis\u003c/em\u003e B168 BKK single gene knockout collection\u003csup\u003e97\u003c/sup\u003e\u003c/p\u003e\n\u003ch3\u003eGrowth Conditions\u003c/h3\u003e\n\u003cp\u003eOvernight precultures were grown in LB (Sigma-Aldrich; 10 g/L tryptone, 5 g/L yeast extract, 10 g/L sodium chloride, pH 7.0) at 37\u0026deg;C, 220 rpm, diluted 1:10,000 into fresh LB, and grown to mid-exponential phase (OD₆₀₀ = 0.5\u0026ndash;0.8), washed thrice in half-strength Hoagland medium\u003csup\u003e98\u003c/sup\u003e (\u0026frac12;HS) (Basal Salt Mixture No. 2, Sigma-Aldrich, 1 g/L MES, pH 5.7 KOH), and resuspended to the desired OD in \u0026frac12;HS. Chloramphenicol (5 \u0026micro;g/mL) or/and kanamycin (25 \u0026micro;g/mL) were added when applicable.\u003c/p\u003e\n\n\u003ch3\u003ePlant Material\u003c/h3\u003e\n\u003cp\u003e\u003cem\u003eArabidopsis thaliana\u003c/em\u003e lines used are listed in Table 4. All seeds were filtered to hold a diameter \u0026gt;250 \u0026micro;m, to ensure comparable seed quality across genotypes. Seeds were sterilized using chlorine gas for 2 h, degassed for 1 h, and stratified at 4\u0026deg;C for 48 h submerged in sterile water. \u003c/p\u003e\n\u003cp\u003eFor microfluidic experiments seeds, were prepared as described in Guichard \u003cem\u003eet al. \u003c/em\u003e2020\u003csup\u003e99\u003c/sup\u003e, germinated in agar-filled 10 \u0026micro;L pipette tips with \u0026frac12;HS medium (0.8 % plant agar), then grown at constant 22 \u0026deg;C under long-day conditions (16 h light/ 8 h dark, 150 \u0026micro;mol/m\u0026sup2;/s) for 4\u0026ndash;5 days in an enclosed plant growth cabinet (CLF Plant Climatics GmBH, SED41-CU4C2LT).\u003c/p\u003e\n\u003ch3\u003eRootChip and \u0026micro;Rhizotron Fabrication and Plant Inoculation\u003c/h3\u003e\n\u003cp\u003eRootChip8NF (RC8NF, Fig. S1A) and \u0026micro;Rhizotron (Fig. S2) devices with eight microchannels and a soil-mimicking Voronoi pattern with an average 400 \u0026micro;m pore diameter were designed in Fusion 360. Casting molds for PDMS were printed via stereolithography printing (Formlabs, Form3) in clear resin (Formlabs, Clear Resin V4.0), cleaned, UV-cured, and lastly cast with PDMS (DOW Chemicals, Sylgard 184, 10:1). Chips were plasma-bonded to glass slides, sterilized with UV for 10 min for each side, and filled with \u0026frac12;HS medium. Seedlings were transferred pre-germinated from pipette tips into channels and grown at 21\u0026deg;C, 45\u0026deg; inclination, under long-day conditions until 7 days post seeding (dps). Bacterial inoculation was then performed by flushing chips with 1 mL of \u003cem\u003eB. subtilis \u003c/em\u003eor \u003cem\u003eB. velezensis \u003c/em\u003eFZB42 suspensions (OD₆₀₀ = 0.02) in \u0026frac12;HS. Chips were again incubated in the plant growth chamber and 45\u0026deg; incline until imaging.\u003c/p\u003e\n\u003ch3\u003eMicroscopy Setup\u003c/h3\u003e\n\u003cp\u003eSpatial mapping was performed on a custom-built inverted open-frame epifluorescence microscope (COS; Cairn GmbH, Heidelberg) equipped with a 25\u0026times; water-immersion objective (NA 1.1, Nikon), motorized stage (ITK Dr. Kassen), laser launch (LDI 89 North), triple-band dichroic mirror (395 nm/ 488 nm/ 561 nm, Chroma), and sCMOS camera (Teledyne Photometrics Kinetix). Fluorescence imaging used 488, 520, and 577 nm excitation (30\u0026ndash;50% intensity), 50 ms exposure, and 525+/-50 nm (GFP), 542+/-20 (YFP), and 595+/-26 nm (RFP) emission filters (Semrock). The system was automated utilizing Micromanager (V.2.0) \u003csup\u003e100,101\u003c/sup\u003e. \u003c/p\u003e\n\u003cp\u003eTime-lapse imaging was conducted on a separate custom spinning disk system built on a Nikon Ti-E stand with a 10\u0026times; S Fluor objective (NA 0.50), CrestOptics X-Light disk (50 \u0026micro;m pinhole), motorized stage (ASI), laser launch (Omicron), filter wheel (Cairn), and Prime 95B sCMOS camera (Teledyne Photometrics). A triple-band dichroic (440 nm/ 514 nm/ 561 nm) and quad-band (405 nm/ 488 nm/ 561 nm/ 640 nm, Chroma) were used with 488 nm, 520 nm, and 561 nm excitation (80 mW), and 525+/-50 nm and 605+/-70 nm emission filters for GFP and RFP, respectively. Images were acquired using Nikon NIS Elements v4.0.\u003c/p\u003e\n\u003cp\u003eMapping of ROS levels was conducted on an Olympus MVX10 stereo microscope, equipped with a 0.63x objective (MVPLAPO, 0.15 NA, FOV = 55 mm) with magnification of 1.25x, motorized focus (SZX2-FOF-M), CoolLED pE4000 light source, quad-band dichroic mirror 395+/-25, 475+/-28, 555/25, 635+/-18 (Chroma 89402bs), 488 nm excitation with an 519 +/- 26 nm emission filter (Chroma), CellCam Kikker 100MT camera in 2x2 binning mode (14bit, px=6.07\u0026micro;m, Cairn UK). Image acquisition was done through Micro-Manager software version 2.0.\u003c/p\u003e\n\u003ch3\u003eMicroscopy of Microcolonies\u003c/h3\u003e\n\u003cp\u003eLB agar was poured in 4-well cell culture slides (Sarstedt, X-well slide, 94.6140.402) and one half removed. This agar cube was then laterally spotted with \u003cem\u003eB. subtilis \u003c/em\u003ePS216 (BIB2352) and incubated overnight at 37 \u0026deg;C. The microcolonies formed at the border of glass, agar and air was then imaged on the Nikon Ti-E system at 20x magnification using the integrated large image function a 9x9 grid was imaged and stitched. \u003c/p\u003e\n\u003ch3\u003eStatistical Analysis\u003c/h3\u003e\n\u003cp\u003eAll statistical analyses were performed in R. As most variables were bounded and non-normally distributed, we did not assume normality for raw data. Followingly, all tests were performed non-parametrically. Pairwise comparisons between genotypes or treatments were conducted using two‑sided Wilcoxon rank‑sum tests, and effect sizes were reported as rank‑biserial correlations (r\u003csub\u003erb\u003c/sub\u003e). Multiple testing was corrected using the Benjamini\u0026ndash;Hochberg false discovery rate procedure. All displayed comparisons are calculated with 95 % confidence intervals (CI), effect size, p-value, and adjusted p-value (Tab. 5 \u0026ndash; 15).\u003c/p\u003e\n\u003ch3\u003eSpatial Mapping of Bacterial Cell Types\u003c/h3\u003e\n\u003cp\u003eAt 10 days post-seeding (dps) and 3 days post-inoculation (dpi), roots were imaged using epifluorescence microscopy on the Cairn open-frame setup. Each primary root was imaged once at 3 dpi and treated as an independent biological replicate; no root was measured repeatedly. A ~1 mm \u0026times; 5 mm (width \u0026times; height) region starting at the root tip was captured using 18 overlapping fields of view (FOVs) (820 \u0026times; 820 \u0026mu;m each, 0.52 \u0026micro;m/px), spanning the full root diameter in Z-stacks (80\u0026ndash;140 \u0026mu;m depth, 5 \u0026mu;m steps) from the cover slide surface. \u003c/p\u003e\n\u003cp\u003eImages were processed in Fiji\u003csup\u003e102\u003c/sup\u003e (v2.17.0), including background correction (Rolling ball, radius = 20 px), stitching, and maximum intensity projected across all Z-slices via a custom script. Generally, all experiments were performed on three independent days, including multiple roots and chip setups. Segmentation of bacterial coverage was performed using a fixed threshold, derived from the average lower threshold of the Col-0 subset (N = 37) using the \u0026ldquo;Default\u0026rdquo; auto-thresholding method. Bacterial coverage was segmented based on size and shape to limit the detection of auto-fluorescent plant structures, including single cells and aggregates (Fig. S1C). Each isolated cluster was then identified as a region of interest (ROI), mean fluorescence intensity, size, x and y coordinates isolated. \u003c/p\u003e\n\u003cp\u003eThe root midline was manually labelled and extracted as a continuous polyline. The distance between the pixels (0.52 x 0.52 \u0026micro;m) on the root midline and bacterial clusters was determined using nearest-neighbor search via the RANN\u003csup\u003e103\u003c/sup\u003e (v.2.6.2) package, yielding a distance to the root midline. To account for root diameter, distances were corrected by subtracting 100 \u0026micro;m: clusters with corrected values \u0026le; 0 \u0026micro;m were classified as \u0026ldquo;rhizoplane\u0026rdquo;-associated, whereas those \u0026gt; 0 \u0026micro;m were classified as \u0026ldquo;rhizosphere\u0026rdquo;-associated, notably, the rhizoplane or root section is a planar Z-projection and does not account for radial depth of the root. The cumulative length of the midline up to the nearest point for each cluster was recorded as its distance from the root tip.\u003c/p\u003e\n\u003cp\u003eBacterial clusters were binned into radial segments of 50 \u0026micro;m or longitudinal segments of 200 \u0026micro;m. Longitudinal analyses focused on the rhizoplane (\u0026le; 0 \u0026micro;m from the root, e.g. the area of the planar Z-projection converged by the root). Coverage was calculated as the proportion of each bin\u0026rsquo;s area occupied by fluorescent clusters relative to the total analyzed area. Subpopulation composition was expressed as the proportion of area covered by each bacterial group within the total occupied area. For statistical analyses, these values were first aggregated at the root level, and roots were treated as independent biological replicates. Values were then averaged across roots imaged on different days and in different microfluidic chips. Where applicable, data is displayed as mean value \u0026plusmn; standard error (SE) calculated across roots.\u003c/p\u003e\n\u003cp\u003eContinuous two-dimensional kernel density (KDE) estimates were calculated for each root and averaged across replicates using the spatstat\u003csup\u003e104\u003c/sup\u003e package (v3.4.1) in R. KDEs were generated for the area-weighted counts of GFP-expressing, mCherry-expressing, and all fluorescent cells (Total), as well as for the fraction of matrix-producing cells, defined as their coverage relative to the coverage of all fluorescent clusters within an area. \u003c/p\u003e\n\u003cp\u003eStatistical analyses included multiple pairwise comparisons using Two-sided Wilcoxon rank-sum tests with Benjamini\u0026ndash;Hochberg correction. Correlations between population fractions were assessed using \u0026beta;‑regression for ratio data bounded between 0 and 1 using the betareg package (v.3.2.4). The same Col‑0 dataset was analyzed in Figures 1 and 5. In these models, the fraction of matrix-producing cells served as the response variable, and longitudinal root position was included as a continuous covariate. Alternatively, the fraction of motile cells in the total population (bounded between 0,1) was plotted against the continuous area coverage by matrix-producing cells. A logit link function was used, and roots were treated as independent biological replicates contributing one proportion value per longitudinal bin.\u003c/p\u003e\n\u003ch3\u003e\u0026micro;Rhizotron Time-lapse Imaging\u003c/h3\u003e\n\u003cp\u003eA single seedling was inoculated with strain BIB2352 (Table 1) at 7 days post-seeding, incubated for 3 additional days, and imaged over 15 hours. Imaging was performed on the Nikon Ti-E system using 1 \u0026times; 1 mm fields with Z-stacks acquired at 140 \u0026micro;m / 20 \u0026mu;m (7 steps). Image tiles were stitched using the pair-wise stitching plugins\u003csup\u003e105\u003c/sup\u003e, and bacterial (ROIs) were segmented based on fluorescence intensity in ImageJ. This time-lapse experiment represents repeated measurements of the same root over time and therefore constitutes a single biological replicate (n = 1). \u003c/p\u003e\n\u003cp\u003eBacterial ROIs after intensity-based segmentation contained isolated bacterial aggregates of multiple cells as also single cells, these were considered as \u0026ldquo;clusters\u0026rdquo;. ROIs were identified on the last timepoint at 15 h. For each cluster corresponding to the spatial mapping, across timepoints fluorescence intensity in GFP and RFP was measured resulting in the fluorescence trajectory, and after manually labelling the root midline a distance from the root and root tip was calculated based on the XY coordinates of each ROI corresponding to the spatial mapping in the RC8NF using RANN\u003csup\u003e103\u003c/sup\u003e. Each spatially distinguishable bacterial cluster was treated as a biological replicate, and its fluorescence was measured repeatedly over time.\u003c/p\u003e\n\u003cp\u003eClusters were determined as \u0026ldquo;adhered\u0026rdquo; once the fluorescence, primarily in GFP, reached an above background level, switching to matrix producers (P\u003cem\u003etapA\u003c/em\u003e-mCherry) was determined as the first timepoint, using a threshold set as the mean plus one standard deviation of RFP intensity observed in GFP-fluorescent clusters at their final timepoint. Cells were classified as flagellated (\u003cem\u003eGFP\u003c/em\u003e) below this threshold and matrix-producing (\u003cem\u003emcherry\u003c/em\u003e) once this threshold was reached. Averages were built upon each individual bacterial cluster that was captured.\u003c/p\u003e\n\u003cp\u003eA KDE model was used to represent the spatial distribution of all switching events as the location relative to the root and root tip at which clusters first activated \u003cem\u003emcherry \u003c/em\u003eexpression. Additionally, a generalized additive model (GAM) was fitted using the mgcv package (v.1.9-4) in R, with the probability that an individual cluster exhibited at least one switching event (binomial family, logit link) as the response variable and longitudinal root position as a continuous covariate modelled with a smooth term. To avoid pseudo replication arising from repeated observations of the same cluster across timepoints, the model was fitted using only the final imaging frame, such that each cluster contributed a single observation corresponding to its final distance from the root tip.\u003c/p\u003e\n\u003ch3\u003eImaging of Plant Immunity Markers\u003c/h3\u003e\n\u003cp\u003eSeedlings were inoculated with \u003cem\u003eB. subtilis\u003c/em\u003e PS-216 (OD₆₀₀ = 0.02) transcriptional reporter strains (Table 1) or mock-treated, grown 3 days post-inoculation, and imaged in brightfield and YFP channels on the Cairn Openframe microscope in the RootChip8NF accordingly to the previous section. Each primary root was imaged once and treated as an independent biological replicate. A 150 \u0026micro;m-wide midline (288 px) was placed manually utilizing the segmented polyline tool in FiJi (ImageJ), and the YFP fluorescence profile was extracted. Either the average fluorescence across the mean line or the average per position across all imaged roots of each treatment was compared. Statistical analyses included multiple pairwise comparisons using Two-sided Wilcoxon rank-sum tests with Benjamini\u0026ndash;Hochberg correction.\u003c/p\u003e\n\u003ch3\u003eROS Detection on Roots\u003c/h3\u003e\n\u003cp\u003eAt 7 dps, seedlings of \u003cem\u003eA. thaliana \u003c/em\u003e(Table 4) grown on nylon meshes (pore diameter 100 \u0026micro;m) were treated with bacterial suspensions of \u003cem\u003eB. subtilis \u003c/em\u003ePS216 (P\u003cem\u003eveg\u003c/em\u003e-mCherry) or mock-treated with sterile medium. At several timepoints starting before the inoculation up to 3 dpi, ROS was detected using 2 \u0026mu;M H₂DCFDA (stock solution: 20 mM in absolute ethanol) staining for 20 min at room temperature and in the dark and imaged via stereomicroscopy (Olympus MVX10) in the GFP channel. The experiment was repeated across several meshes and days containing multiple roots. Each root was treated as an independent biological replicate and was imaged only once. Meshes were not imaged repeatedly; each mesh was used for a single imaging round and then discarded.\u003c/p\u003e\n\u003cp\u003eEach image was background-corrected (rolling ball, r = 50 \u0026micro;m) and manually labelled with a spline-fitted continuous midline of 25 px (~100 \u0026micro;m) in ImageJ, coordinates and mean fluorescence intensity were extracted. Intensity profiles were averaged across individual roots across all meshes and normalized to the mock-treated average. For time-lapse imaging, untreated meshes were used as 0 dpi \u0026ldquo;untreated\u0026rdquo; controls.\u003c/p\u003e\n\u003ch3\u003eMulti-well Plate Reader Assays\u003c/h3\u003e\n\u003cp\u003eCultures were prepared in supplemented \u0026frac12;HS (0.5 % w/v L-glutamate, 0.5 % v/v glycerol), diluted to an OD\u003csub\u003e600\u003c/sub\u003e of 0.1, then mixed 1:1 with treatments (final OD\u003csub\u003e600\u003c/sub\u003e 0.05) in 200 \u0026micro;L volume on black flat-well 96-well plates (GreinerBIO). Overexpression assays included 1 mM IPTG. Dilution series (1:1) of H₂O₂ were prepared automatically using a pipetting robot (Opentrons, OT-2), and additionally, an unsupplemented medium control was provided. Measurements were made using a CLARIOstar Plus microplate reader (25\u0026deg;C, no shaking), capturing GFP (470 +/- 15 nm, 515 +/- 20 nm, 20 flashes), mCherry (570 +/- 15 nm, 620 +/- 20 nm, 20 flashes), and absorbance at 600 nm (10 flashes). A settling time of 0.5 s was allowed for each capture. Fluorescence was normalized to OD\u003csub\u003e600\u003c/sub\u003e. Multi-well plate experiments were performed in at least two independent time‑series experiments. Within each experiment, each well containing the same treatment was treated as an independent biological replicate. Wells were measured continuously over time, but liquid samples for spotting onto agarose pads were taken only once per well; each well, therefore, contributed a single biological replicate for a specific time point to the agarose-pad assay.\u003c/p\u003e\n\u003cp\u003eAt multiple timepoints during growth in the presence of either 0 or 156 \u0026micro;M hydrogen peroxide (H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e), 1 \u0026micro;L aliquots of \u003cem\u003eB. subtilis\u003c/em\u003e PS-216 (BIB2352) or derived mutant cultures from the multi-well plates were spotted onto \u0026frac12;HS 1.5 % agarose pads. Spots were imaged on the Nikon Ti-E using NIS Elements (Large Image Mode), capturing multiple fields of view (5x5, 1x1mm FOVs, 25 % overlap) to generate complete spot images. Fluorescent regions were segmented and background-corrected using untreated agarose pads as a reference. \u003c/p\u003e\n\u003cp\u003eThe area covered by bacterial cells was isolated by intensity-based segmentation; such areas were then assigned as ROIs and summarized. This was performed separately for GFP- and RFP-fluorescent areas, which were summed to a total fluorescent area. The cell density was estimated based on the minimal size of distinct fluorescent clusters, which was 4 pixels, e.g. 4 \u0026micro;m\u003csup\u003e2\u003c/sup\u003e.\u003c/p\u003e\n\u003ch3\u003eSoftware\u003c/h3\u003e\n\u003cp\u003eImage analysis and processing were performed using ImageJ/Fiji (v1.54r). Statistical analysis, plotting, and data preprocessing were conducted in R (v4.5.2) using the RStudio environment (v2025.09.2+418). \u003cbr clear=\"all\"\u003e \u003c/p\u003e"},{"header":"Results","content":"\u003ch2\u003eRoot-colonizing Bacillus populations undergo a spatial transition from flagellated to matrix-producing cells along the root\u0026rsquo;s axis\u003c/h2\u003e\n\u003cp\u003eTo investigate the cell type differentiation of a root-colonizing rhizobacterium, we first engineered fluorescent reporter systems to monitor cell differentiation in the soil isolate \u003cem\u003eB. subtilis \u003c/em\u003ePS-216\u003csup\u003e91\u003c/sup\u003e and the plant-growth-promoting \u003cem\u003eB. velezensis \u003c/em\u003eFZB42\u003csup\u003e92\u003c/sup\u003e, respectively. To this end, the expression of \u003cem\u003egfpmut3 \u003c/em\u003ewas placed under the control of the \u003cem\u003ehag-\u003c/em\u003epromoter that is activated in flagellated cells\u003csup\u003e57\u003c/sup\u003e. \u003cem\u003emCherry \u003c/em\u003ewas placed under the \u003cem\u003etapA \u003c/em\u003epromoter that controls expression of the matrix production-associated \u003cem\u003etapA-sipW-tasA\u003c/em\u003e operon\u003csup\u003e57,107\u003c/sup\u003e and is thus activated in matrix producers (Fig. S1A). Stationary-phase cultures grown for 48 h in supplemented \u0026frac12;HS medium (0.5 % w/v glycerol and mono-potassium glutamate), showed a heterogeneous population, composed of green and red fluorescently labelled cells (Fig. S1B, S1C), indicating that motility and matrix production are differentially activated in distinct subpopulations, as expected\u003csup\u003e64,65\u003c/sup\u003e. These bacterial cultures were then used in root colonization experiments with \u003cem\u003eA. thaliana.\u003c/em\u003e \u003c/p\u003e\n\u003cp\u003eTo track root colonization by fluorescence microscopy, we developed a microfluidic imaging slide termed RootChip8NF based on the RootChip8S\u003csup\u003e90\u003c/sup\u003e scaffold, allowing \u003cem\u003ein situ \u003c/em\u003eimaging on living roots of \u003cem\u003eA. thaliana \u003c/em\u003e(Fig. S1D). After three days post inoculation (dpi), fluorescence from both fluorescent reporters was visible on the roots colonized by \u003cem\u003eB. subtilis\u003c/em\u003e (Fig. 1A, left) and \u003cem\u003eB. velezensis \u003c/em\u003e(Fig. 1A, middle), respectively, suggesting that the root environment also supports the coexistence of both flagellated and matrix-producing Bacilli. Notably, the fluorescence distribution varied with distance from the root tip, suggesting that different cell types might be favored in different parts of the root. We estimated the spatial population structure based on the share of matrix producers to the overall bacterial biomass (Fig. 1B, 1C), flagellated cells, as a function of distance from the root and distance from the root tip from the fluorescence distribution (see Methods for details). \u003c/p\u003e\n\u003cp\u003eAs expected, bacterial biomass accumulated in the root (rhizoplane, \u0026lt;0 \u0026micro;m from the root, e.g., the root-covered section of the planar Z-projection) and declined with lateral distance from the root within a few hundred \u0026micro;m into the root-distal medium (rhizosphere, \u0026gt;0 \u0026micro;m from the root) (Fig. 1D, left). Population composition, as estimated from the coverage of green and red fluorescence areas, respectively, showed a pronounced shift with distance from the root tip (Fig. 1B, 1C, 1E). Flagellated cells were preferentially present around the root elongation zone (~500-800 \u0026micro;m from the root tip), and matrix-producing cells increased in abundance and dominated the maturation zone (\u0026gt;850 \u0026micro;m from the root tip) of the root\u003csup\u003e108\u003c/sup\u003e. \u003c/p\u003e\n\u003cp\u003eThis shift was consistently observed across multiple roots and chips with both \u003cem\u003eB. subtilis\u003c/em\u003e and \u003cem\u003eB. velezensis \u003c/em\u003e(Fig.1C, E), respectively. We also confirmed qualitatively, using a non-fluorescent wildtype of PS-216, a constitutively fluorescent (P\u003cem\u003e\u003csub\u003eveg\u003c/sub\u003e\u003c/em\u003e-\u003cem\u003emcherry\u003c/em\u003e) and fluorophore swapped strain (PS-216, P\u003cem\u003e\u003csub\u003ehag\u003c/sub\u003e-gfpmut3, P\u003csub\u003etapA\u003c/sub\u003e-mCherry\u003c/em\u003e), that observed patterns were not attributed to autofluorescence (Fig. S1E). Secondly, cell type localization and detection of both flagellated and matrix-producing cells remained robust upon fluorophore exchange (Fig. S1E). Together, these data suggested that both \u003cem\u003eBacilli\u003c/em\u003e colonize \u003cem\u003eA. thaliana\u003c/em\u003e roots through a spatially organized transition from flagellated to matrix-producing cells. \u003cem\u003eB. velezensis\u003c/em\u003e showed overall a higher abundance of matrix producers compared to \u003cem\u003eB. subtilis,\u003c/em\u003e and the transition from motility to matrix production occurred closer to the root tip. We also found that matrix-producing cells predominantly covered the rhizoplane (Fig. 1D).\u003c/p\u003e\n\u003cp\u003eTo test whether differentiation from flagellated-to-matrix producers is required for rhizoplane association, we performed colonization experiments with PS-216 and a \u003cem\u003esinI \u003c/em\u003eknockout. SinI is known to bind to the SinR transcription factor, the master regulator of biofilm formation\u003csup\u003e18,71\u003c/sup\u003e. Indeed, in the \u003cem\u003esinI \u003c/em\u003eknockout mutant, less bacterial biomass accumulated in the rhizoplane (5.7% estimated coverage) compared to the wildtype (12.9% estimated coverage), while there was no significant difference in biomass in the rhizosphere (Fig. 1G, 1H). This indicates that cell differentiation from flagellated cells towards matrix producers via the SinI-SinR switch is necessary for efficient rhizoplane colonization. \u003c/p\u003e\n\u003ch2\u003e\u003cem\u003eB. subtilis \u003c/em\u003ecell type differentiation occurs locally confined in the early maturation zone\u003c/h2\u003e\n\u003cp\u003eWe next aimed to observe cell state transitions from motility to matrix production by monitoring changes in gene expression by time-series imaging. To this end, we performed colonization experiments with the PS-216 reporter strain in a structured microfluidic imaging slide termed the \u0026micro;Rhizotron. This device mimics a spatially heterogeneous soil environment by featuring soil-particle-mimicking Voronoi patterns (Fig. S2). We captured continuous cell fate transitions over 14 hours in the actively growing root tip region of a single \u003cem\u003eA. thaliana\u003c/em\u003e Col-0 root 3 days post inoculation (Fig. 2A). For a detailed analysis, we focused on a 2 mm segment encompassing the transition from the meristematic to maturation zone\u003csup\u003e108\u003c/sup\u003e (Fig. 2A, upper). We hereby analyzed isolated clusters of fluorescent bacterial cells, which encompassed both single cells and aggregates.\u003c/p\u003e\n\u003cp\u003eOver time, we first observed the presence of green fluorescence signals from P\u003cem\u003e\u003csub\u003ehag,\u003c/sub\u003e\u003c/em\u003e which was then followed by an onset of red fluorescence from P\u003cem\u003e\u003csub\u003etapA\u003c/sub\u003e\u003c/em\u003e in some clusters (Fig. 2A, lower; white box), indicative of switching events from motility towards matrix production. Based on the fluorescence profile, we designated two subpopulations \u003cem\u003egfp\u003c/em\u003e-expressing clusters (flagellated), and such that displayed a shift from \u003cem\u003egfp \u003c/em\u003eto \u003cem\u003emCherry \u003c/em\u003eexpression, termed \u0026ldquo;switchers\u0026rdquo;. Averaged across all switching cells, \u003cem\u003emCherry \u003c/em\u003efluorescence rose steadily from 5 h post-adherence to the root (Fig. 2B), indicating the activation of the P\u003cem\u003e\u003csub\u003etapA\u003c/sub\u003e\u003c/em\u003e and thus matrix production. The fraction of matrix producers started to rise after 5 hours and increased gradually over time to reach 25 % after 12 hours (Fig. 2C, left). \u003c/p\u003e\n\u003cp\u003eSwitching events were confined to a spatial window of 865 +/- 248 (SD) \u0026micro;m from the root tip (Fig. 2D), coinciding with the area of initial root hair formation, indicative of the maturation zone (Fig. 2A). Correspondingly, generalized additive modelling revealed a significant, non-linear dependence of switching outcomes in the final imaging frame (14 h) on distance from the root tip (edf = 8.8, p = 6.8 \u0026times; 10⁻⁴). Clusters that had undergone at least one switching event during the experiment were enriched at the root cap and at intermediate distances along the root, with a marked increase in switching-associated clusters beyond 500 \u0026micro;m from the root tip and a local maximum around 800 \u0026micro;m (Fig. 2E).\u003c/p\u003e\n\u003cp\u003eIn conclusion, these findings suggest cell type switching occurs not homogeneously along the root axis but predominantly localized around the root elongation zone. As a result, the fraction of matrix producers varies as a function of distance from the root tip. Consistent with our previous observations (Fig. 1E), a flagellated-dominant population progressively shifted toward matrix production with increasing distance from the root tip, also in the \u0026micro;Rhizotron (Fig. 2C, right). This indicates that the spatial modulation of cell-state transitions in different areas of the root occurs robustly, and even in a spatially heterogeneous environment that mimics aspects of soil. We thus reasoned that host-dependent factors might modulate cell differentiation during root colonization.\u003c/p\u003e\n\u003ch2\u003eRoot colonization elevates local ROS production in the elongation and early maturation zone\u003c/h2\u003e\n\u003cp\u003eWe next asked whether spatially patterned oxidative stress could act as an environmental cue shaping bacterial cell-state transitions. As the evasion of host stress through phenotypic heterogeneity is a known strategy in other pathosystems\u003csup\u003e14\u0026ndash;18\u003c/sup\u003e, and specifically, since cell type differentiation is environmentally regulated in \u003cem\u003eBacillus\u003c/em\u003e\u003csup\u003e80,82,109\u003c/sup\u003e, we hypothesized that the plant-shaped environment may shape the observed longitudinal patterns of bacterial colonization. Since most bacterial cells reside on the rhizoplane, we focused on oxidative stress due to its spatially confined nature.\u003c/p\u003e\n\u003cp\u003eTo assess changes in the oxidative stress environment during bacterial colonization, we employed the oxidative-stress-sensitive dye H\u003csub\u003e2\u003c/sub\u003eDCFDA to map basal ROS production. H\u003csub\u003e2\u003c/sub\u003eDCFDA reports steady-state oxidation but does not distinguish ROS species or identify their cellular source; therefore, our interpretations focus on spatial associations rather than molecular specificity. Fluorescence was captured following a 20-minute dark incubation, enabling comparative estimation of steady-state ROS levels across root zones. This approach provided spatial resolution of ROS accumulation in living \u003cem\u003eA. thaliana\u003c/em\u003e roots, offering a functional readout of oxidative stress dynamics in response to colonization.\u003c/p\u003e\n\u003cp\u003eGenerally, the peak of ROS levels was found between 1,000 to 1,600 \u0026micro;m from the root tip (Fig. 3A, 3B), with an increased spatial extent after inoculation with PS-216. During colonization over three days with \u003cem\u003eB. subtilis \u003c/em\u003ePS-216 (initial OD\u003csub\u003e600\u003c/sub\u003e 0.02), we observed a steady increase in the peak and global H\u003csub\u003e2\u003c/sub\u003eDCFDA fluorescence with a root-tip-ward peak in the elongation zone (Fig. 3A, 3B). Compared to mock-treated plants, colonized roots exhibited a significant increase from 0 - 3 dpi (slope = 208.4, \u003cem\u003e\u0026tau; \u003c/em\u003e= 7.5, Kendall\u0026rsquo;s Tau) while mock-treated plants did not (slope = -0.47, \u003cem\u003e\u0026tau;\u003c/em\u003e = -0.01, Kendall\u0026rsquo;s Tau) (Fig. 3C). \u003c/p\u003e\n\u003cp\u003eTo test whether increased ROS levels would also correlate with immunity activation, transcriptional reporter lines for early pattern-triggered immunity (PTI) were colonized with wild-type PS-216\u003csup\u003e32\u003c/sup\u003e. Furthermore, the early PTI transcriptional reporter pFRK1::3xmVenus~NLS and pPER5::3xmVenus~NLS confirmed an induction of both marker genes in the 1-2 mm from the root tip in response to colonization by PS-216, while the rather general stress-associated marker ZAT12\u003csup\u003e110\u003c/sup\u003e did not show activation (Fig. S3A\u0026ndash;C).\u003c/p\u003e\n\u003cp\u003eROS production in response to biotic interaction is primarily mediated in plants by membrane-bound NADPH oxidases, known as respiratory burst oxidase homologs (RBOHs), with RBOH D and RBOH F playing key roles during PTI in \u003cem\u003eArabidopsis\u003c/em\u003e\u003csup\u003e111\u003c/sup\u003e. To determine whether the observed ROS increase during colonization is dependent on these enzymes, we inoculated \u003cem\u003erbohD\u003c/em\u003e, \u003cem\u003erbohF\u003c/em\u003e, and \u003cem\u003erbohD/F\u003c/em\u003e mutants with \u003cem\u003eB. subtilis\u003c/em\u003e PS-216 for 3 days. All genotypes retained a root-tip-ward ROS peak (Fig. 3D); however, H\u003csub\u003e2\u003c/sub\u003eDCFDA fluorescence was notably reduced in \u003cem\u003erbohD\u003c/em\u003e and \u003cem\u003erbohD/F\u003c/em\u003e mutants, particularly in the root tip and maturation zones (Fig. 3E). Quantitatively, both mutants exhibited a significant reduction in mean ROS levels (approximately 21\u0026ndash;25%) compared to Col-0 wildtype, while \u003cem\u003erbohF\u003c/em\u003e did not show a significant change (Fig. 3F).\u003c/p\u003e\n\u003cp\u003eThese findings demonstrate that \u003cem\u003eB. subtilis\u003c/em\u003e colonization elevates ROS levels in a partly RBOH D-dependent manner, with a pronounced maximum in the elongation zone. Notably, this ROS peak spatially coincides with the activation of the PTI markers FRK1 and PER5. Additionally, the activation of immunity and the ROS peak are in proximity to the flagellated cell maximum and the transition point toward matrix-producing cells, indicating that host-derived oxidative stress may provide spatial information for bacterial cell-type transitions.\u003c/p\u003e\n\u003ch2\u003eROS exposure promotes matrix production in\u003cem\u003e Bacillus subtilis\u003c/em\u003e\u003c/h2\u003e\n\u003cp\u003eTo assess whether ROS exposure can induce a shift in population structure, we asked whether hydrogen peroxide (H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e), a main component of root-derived ROS, could elicit cell type-dependent effects in axenic cultures of \u003cem\u003eB. subtilis\u003c/em\u003e. To test this under well-controlled conditions, we performed plate-based growth assays of \u003cem\u003eB. subtilis\u003c/em\u003e in gradients of hydrogen peroxide, grown at 25 \u0026deg;C and carbon-supplemented Hoagland\u0026rsquo;s plant medium, to create the closest proxy to our RootChip environment. Estimates for apoplastic ROS under non-stressed conditions indicate nM to low \u0026micro;M concentrations in the apoplastic fluid of \u003cem\u003eA. thaliana\u003c/em\u003e, which can rise by several orders of magnitude in response to stress\u003csup\u003e112,113\u003c/sup\u003e. Physiological effects have been reported at ~100 \u0026micro;M H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e in stomatal closure \u003csup\u003e114\u003c/sup\u003e, and leaf tissue concentrations of H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e can reach ~1 \u0026micro;mol/g fresh weight \u003csup\u003e115\u003c/sup\u003e. We, therefore, decided to test the effect of a wide range of nM to mM concentrations of hydrogen peroxide, applied in parallel to inoculation, on the growth behavior of \u003cem\u003eB. subtilis \u003c/em\u003ePS-216 cultures in multi-well plates.\u003c/p\u003e\n\u003cp\u003eIncreasing concentrations of H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2 \u003c/sub\u003eresulted in a delayed onset of growth, with severe reductions of the final cell density at concentrations \u0026gt;156 \u0026micro;M of H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e, and complete inhibition at 10 mM (Fig. 4A).\u003cem\u003e \u003c/em\u003eThe motility-associated transcriptional reporter P\u003cem\u003e\u003csub\u003ehag\u003c/sub\u003e\u003c/em\u003e-\u003cem\u003egfpmut3\u003c/em\u003e exhibited a biphasic response at control and low H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2 \u003c/sub\u003econcentrations, with an early fluorescence peak during the lag phase followed by a dosage-dependent decline throughout exponential and stationary phases (Fig. 4B). At concentrations exceeding 156 \u0026micro;M, this initial peak was absent. \u003c/p\u003e\n\u003cp\u003eIn contrast, the matrix-associated reporter P\u003cem\u003e\u003csub\u003etapA\u003c/sub\u003e\u003c/em\u003e-\u003cem\u003emCherry\u003c/em\u003e became active upon entry into the exponential phase, maintaining steady levels during the lag phase (Fig. 4C). Peak fluorescence of P\u003cem\u003e\u003csub\u003etapA\u003c/sub\u003e\u003c/em\u003e-\u003cem\u003emCherry\u003c/em\u003e increased progressively with H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2 \u003c/sub\u003eexposure up to 156 \u0026micro;M, suggesting enhanced matrix gene expression under oxidative stress. Consistent with these trends, the ratio of P\u003cem\u003e\u003csub\u003etapA\u003c/sub\u003e\u003c/em\u003e-\u003cem\u003emCherry\u003c/em\u003e to P\u003cem\u003e\u003csub\u003ehag\u003c/sub\u003e\u003c/em\u003e-\u003cem\u003egfp\u003c/em\u003e fluorescence at the same cell densities increased with rising H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2 \u003c/sub\u003econcentrations (Fig. 4D). \u003c/p\u003e\n\u003cp\u003eTo assess whether these transcriptional changes reflected shifts in population structure, we imaged 1 \u0026micro;L samples from untreated and 156 \u0026micro;M H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e-treated cultures on agarose pads over 48 h (Fig. 4E). Cell density was estimated from the area covered by GFP- and mCherry-expressing cells, estimating a planar cell area of 4 \u0026micro;m\u003csup\u003e2\u003c/sup\u003e \u003csup\u003e116,117\u003c/sup\u003e. The initial inoculum (T=0 h) was primarily composed of flagellated cells; both groups displayed an initial reduction of cell density at 12 h, with the ROS-treated group displaying a delayed growth onset and lower final cell density (Fig. 4E, 4F, 4G). While in the untreated group, flagellated cells remained predominant throughout the assessed time, in the ROS-treated group, matrix-producing cells were most abundant (Fig. 4E, 4F, 4G). Followingly, the reporter activity and the share of matrix-producing cells beyond 24 h were found to be higher in the ROS-treated group (Fig. 4H, 4I).\u003c/p\u003e\n\u003cp\u003eThese findings raised the question of whether matrix production confers increased ROS tolerance. To test this, we compared wild-type \u003cem\u003eB. subtilis \u003c/em\u003ePS-216 with \u0026Delta;\u003cem\u003esinI\u003c/em\u003e (flagellated-dominant) and \u0026Delta;\u003cem\u003esinR\u003c/em\u003e (matrix-hyperproducing) mutants in the genomic background. Notably, the \u0026Delta;\u003cem\u003esinR\u003c/em\u003e strain exhibited robust outgrowth at 156 and 313 \u0026micro;M H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2 \u003c/sub\u003ewithin 24 hpi, in contrast to both wild-type and \u0026Delta;\u003cem\u003esinI\u003c/em\u003e, supporting the hypothesis that matrix production enhances oxidative stress resilience (Fig. 4J).\u003c/p\u003e\n\u003cp\u003eTogether, these results indicate a biphasic effect of ROS with an initial delay in growth and matrix production, which leads to a rise of matrix-producing cells following exposure events, increasing the overall fitness under ROS stress. Indeed, the \u0026Delta;\u003cem\u003esinR\u003c/em\u003e mutant highlights the stress resilience of matrix-producing cells, indicating acclimation by matrix production. \u003c/p\u003e\n\u003ch2\u003eRBOH D activity promoted the flagellated-to-matrix transition on the rhizoplane\u003c/h2\u003e\n\u003cp\u003eTo investigate whether the plant-derived ROS influences the differentiation into matrix producers during root colonization of \u003cem\u003eA. thaliana\u003c/em\u003e (Fig. 1B-E)\u003cem\u003e,\u003c/em\u003e we repeated the spatial mapping of bacterial colonization in the ROS-production-impaired respiratory burst oxidase homologue (RBOH) mutants \u003cem\u003erbohD\u003c/em\u003e, \u003cem\u003erbohD/F,\u003c/em\u003e and \u003cem\u003erboh\u003c/em\u003eF (Fig. 3D-F) in direct comparison to the same Col-0 reference group as shown in Fig.1. \u003c/p\u003e\n\u003cp\u003e\u003cem\u003erbohD \u003c/em\u003eand \u003cem\u003erbohD/F \u003c/em\u003edisplayed a selective increase of the flagellated cell coverage globally around the primary root segment (1,000 x 5,000 \u0026micro;m, w x h), with the matrix-producing fraction being unaffected (Fig. 5A, 5B, S4A). This is also reflected in the rhizoplane coverage, which transitions from a predominantly matrix producer population in Col-0 to a flagellated-dominant population in the \u003cem\u003erbohD \u003c/em\u003eand \u003cem\u003erbohDF \u003c/em\u003emutants (Fig. 5D, 5E). Additionally, this effect is seen in the longitudinal axis seen along the longitudinal axis, as \u003cem\u003erbohD \u003c/em\u003eand \u003cem\u003erbohD/F\u003c/em\u003e consistently display a higher fraction of flagellated cells (Fig. 5F, 5G).\u003c/p\u003e\n\u003cp\u003eBeta-regression analysis investigating the relationship of the distance from the root tip and the share of matrix-producing cells revealed that in \u003cem\u003erbohD\u003c/em\u003e and \u003cem\u003erbohD/F\u003c/em\u003e only the elevation and intercept are significantly reduced, though not the slope of the trendline (Fig. 1H,1I), highlighting that the overall tendency towards matrix-producing cells is not affected. Still, a second analysis of the relationship of the share of flagellated cells and matrix producer coverage displays that in \u003cem\u003erbohD\u003c/em\u003e and \u003cem\u003erbohD/F,\u003c/em\u003e a higher degree of matrix producers is required to achieve a similar population structure (Fig. S4B, S4C). \u003c/p\u003e\n\u003cp\u003eAs the \u003cem\u003erbohF\u003c/em\u003e mutant neither suppressed ROS production in response to \u003cem\u003eB. subtilis\u003c/em\u003e PS-216 (Fig. 3D\u0026ndash;F) nor altered bacterial population structure, the observed effects appear specific to \u003cem\u003erbohD\u003c/em\u003e-deficient lines. Disruption of RBOHD activity, therefore shifts bacterial population structure toward flagellated states, consistent with a role for ROS in priming matrix commitment during colonization.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eOur investigation reveals that \u003cem\u003eBacillus subtilis\u003c/em\u003e and the related \u003cem\u003eB. velezensis \u003c/em\u003eestablish a longitudinal cell type pattern along the root axis through cell type differentiation. This transition corresponds to the phase-shift from planktonic growth to robust attachment to the root by biofilm formation\u003csup\u003e118\u003c/sup\u003e - a key lifestyle change governed by cell type differentiation from flagellated-to-matrix-producing cells in \u003cem\u003eBacillus sp.\u003c/em\u003e\u003csup\u003e54,70\u003c/sup\u003e. In accordance, we show that cell-type switching supports effective surface adhesion, as seen in the loss of root-association in the non-switching mutant \u003cem\u003e\u0026Delta;sinI\u003c/em\u003e. This highlights the structural and functional relevance of phenotypic heterogeneity to establish and maintain root association.\u003c/p\u003e\n\u003cp\u003eDuring association, flagellated-to-matrix producer transitions do not occur ubiquitously, but are tightly confined to a spatial window in the early maturation zone. Indeed, the transition from the elongation to the maturation zone marks a critical physiological boundary in the root, characterized by the emergence of root hairs, reduced immune activity, and shifts in exudation profile\u003csup\u003e28,119\u003c/sup\u003e. This tight spatial confinement of bacterial cell-type differentiation indicates modulation by the host\u0026rsquo;s responses.\u003c/p\u003e\n\u003cp\u003eSupporting the hypothesis of an active role of the host in bacterial cell type switching, we observe, upon inoculation with \u003cem\u003eB. subtilis\u003c/em\u003e, activation of PTI coinciding with an RBOHD-dependent increase in ROS production. Spatial mapping reveals peak ROS levels at the elongation zone, coinciding spatially with flagellated cell accumulation. Surprisingly, the loss of RBOHD reduces ROS levels but also enriches flagellated cells in the colonizing \u003cem\u003eBacillus \u003c/em\u003epopulation. This indicates that oxidative environments influence the balance of bacterial cell types, rather than directly repressing motility or promoting matrix production. Despite early inhibitory effects, exposure to hydrogen peroxide enriches matrix-producers in \u003cem\u003eB. subtilis.\u003c/em\u003e An enhanced stress tolerance of matrix-producing cells lies in known antioxidant traits, including iron siderophore release and elevated catalase expression \u003csup\u003e71,120,121\u003c/sup\u003e. Hence, the ROS-stimulated shift of the bacterial population towards matrix production may be an acclimation mechanism to the oxidative environment that promotes root association. This ROS-induced biofilm formation has also been observed in gram-negative interactors and pathosystems. For example, during root colonization by \u003cem\u003eSalmonella enterica \u003c/em\u003e\u003csup\u003e122\u003c/sup\u003e\u003cem\u003e \u003c/em\u003eand lung infection due to \u003cem\u003eStaphylococcus aureus \u003c/em\u003e\u003csup\u003e123\u003c/sup\u003e, biofilm formation has been reported as a response to oxidative stress.\u003c/p\u003e\n\u003cp\u003eWe propose that the spatially polarized ROS gradient along the root axis contributes to a dynamic pattern in which (i) newly recruited bacterial cells at the root apex retain their flagellated state, (ii) subsequently adhere to the root surface, and (iii) transition into matrix production in a stress-exposure-dependent manner, giving rise to a more stress-resilient population through matrix production. These observations point to a causal link between host-derived ROS environments and bacterial cell-state differentiation during colonization.\u003c/p\u003e\n\u003cp\u003eEnvironmental feedback by the plant host has been well established to shape the phylogenetic heterogeneity of the root microbiota by immunity, ROS production - including RBOHD activity - and exudation \u003csup\u003e22,28,124,125\u003c/sup\u003e. Our findings now highlight phenotypic heterogeneity as an additional layer of complexity. Bacterial cell-by-cell decisions tailor root-colonization outcomes, integrating the diverse micro-environmental landscape of pH, exudates, and immunity, all with potential feedback into phenotypic heterogeneity. \u003c/p\u003e\n\u003cp\u003eFuture work will be needed to assess how these gradients jointly shape cell-state dynamics in soil environments or transparent-soil analogs, and to determine how broadly such spatial interactions extend across rhizobacterial taxa. Studying such micro-scaled interactions, which are currently inaccessible by broad-scale omics techniques, is facilitated by using emerging cultivation and imaging approaches involving microfluidics and structured microenvironments\u003csup\u003e83\u003c/sup\u003e. In conclusion, this study highlights the critical need for spatially resolved analyses and the integration of phenotypic heterogeneity to fully capture the complexity of colonization behavior \u003cem\u003ein situ. \u003c/em\u003e\u003cbr clear=\"all\"\u003e \u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eContributions\u003c/p\u003e\n\u003cp\u003eCFK, IBB, and GG designed the study. CFK, SaE, MZ, MK, SeE, and JE performed the experiments. CFK, CE, MK, JE, SaE, SeE and TS generated strains, CK engineered the vector, CFK designed RootChips, and CFK, SaE, and MK produced RootChip devices. CFK, SaE, and MK performed microscopy and image analysis. CFK performed data and statistical analyses. CFK, IBB and GG wrote the manuscript with input from all authors. IBB and GG supervised research. All authors read, revised, and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003eFunding\u003c/p\u003e\n\u003cp\u003eWe thank the German Research Foundation (CRC 1535 \u0026ldquo;Microbial Networking\u0026rdquo;, project ID 458090666; DFG Heisenberg Professorship, GR4559/4-1; and CEPLAS-EXC-2048/1-project ID 390686111 to GG and SPP2389 project ID 504020040 to IBB) and the Max Planck Society (IBB) for funding.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAcknowledgements\u003c/p\u003e\n\u003cp\u003eThe authors gratefully acknowledge Michaela Gerads for technical support, Da\u0026scaron;a Wernerov\u0026aacute; for support with 3D printing. We thank Rub\u0026eacute;n Garrido-Oter, Marjorie Guichard, and Julia Frunzke for discussions and for critically reading the manuscript.\u003c/p\u003e\n\u003cp\u003eDeclaration of generative AI and AI-assisted technologies in the writing process.\u003c/p\u003e\n\u003cp\u003eDuring the preparation of this work the author used Copilot (Microsoft) and ChatGPT-5 (OpenAI) in order to aid readability and text editing. After using this tool/service, the author reviewed and edited the content as needed and take full responsibility for the content of the published article.\u003c/p\u003e\n\u003cp\u003eDeclaration of Interests\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAckermann, M. A functional perspective on phenotypic heterogeneity in microorganisms. \u003cem\u003eNat Rev Microbiol \u003c/em\u003e\u003cstrong\u003e13\u003c/strong\u003e, 497\u0026ndash;508 (2015). \u003c/li\u003e\n\u003cli\u003eB\u0026oacute;di, Z. \u003cem\u003eet al.\u003c/em\u003e Phenotypic heterogeneity promotes adaptive evolution. \u003cem\u003ePLoS Biol \u003c/em\u003e\u003cstrong\u003e15\u003c/strong\u003e, e2000644 (2017). \u003c/li\u003e\n\u003cli\u003eReyes Ruiz, L. M., Williams, C. L. \u0026amp; Tamayo, R. Enhancing bacterial survival through phenotypic heterogeneity. \u003cem\u003ePLoS Pathog \u003c/em\u003e\u003cstrong\u003e16\u003c/strong\u003e, e1008439 (2020). \u003c/li\u003e\n\u003cli\u003eSherry, J. \u0026amp; Rego, E. H. Phenotypic Heterogeneity in Pathogens. \u003cem\u003eAnnual Review of Genetics \u003c/em\u003e\u003cstrong\u003e58\u003c/strong\u003e, 183\u0026ndash;209 (2024). \u003c/li\u003e\n\u003cli\u003eErikson, D. Differentiation of the Vegetative and Sporogenous Phases of the Actinomycetes: 2. Factors affecting the Development of the Aerial Mycelium. \u003cem\u003eJournal of General Microbiology \u003c/em\u003e\u003cstrong\u003e1\u003c/strong\u003e, 45\u0026ndash;52 (1947). \u003c/li\u003e\n\u003cli\u003eWireman, J. W. \u0026amp; Dworkin, M. Morphogenesis and Developmental Interactions in Myxobacteria. \u003cem\u003eScience \u003c/em\u003e\u003cstrong\u003e189\u003c/strong\u003e, 516\u0026ndash;523 (1975). \u003c/li\u003e\n\u003cli\u003eThiel, T. \u0026amp; Pratte, B. Effect on heterocyst differentiation of nitrogen fixation in vegetative cells of the cyanobacterium Anabaena variabilis ATCC 29413. \u003cem\u003eJ Bacteriol \u003c/em\u003e\u003cstrong\u003e183\u003c/strong\u003e, 280\u0026ndash;286 (2001). \u003c/li\u003e\n\u003cli\u003eKaiser, D. Coupling cell movement to multicellular development in myxobacteria. \u003cem\u003eNat Rev Microbiol \u003c/em\u003e\u003cstrong\u003e1\u003c/strong\u003e, 45\u0026ndash;54 (2003). \u003c/li\u003e\n\u003cli\u003eBarka, E. A. \u003cem\u003eet al.\u003c/em\u003e Taxonomy, Physiology, and Natural Products of Actinobacteria. \u003cem\u003eMicrobiol Mol Biol Rev \u003c/em\u003e\u003cstrong\u003e80\u003c/strong\u003e, 1\u0026ndash;43 (2016). \u003c/li\u003e\n\u003cli\u003eGovindarajan, S., Albocher, N., Szoke, T., Nussbaum-Shochat, A. \u0026amp; Amster-Choder, O. Phenotypic Heterogeneity in Sugar Utilization by E. coli Is Generated by Stochastic Dispersal of the General PTS Protein EI from Polar Clusters. \u003cem\u003eFront. Microbiol. \u003c/em\u003e\u003cstrong\u003e8\u003c/strong\u003e, 2695 (2018). \u003c/li\u003e\n\u003cli\u003eNakatani, R. J., Itabashi, M., Yamada, T. G., Hiroi, N. F. \u0026amp; Funahashi, A. Intercellular interaction mechanisms promote diversity in intracellular ATP concentration in Escherichia coli populations. \u003cem\u003eSci Rep \u003c/em\u003e\u003cstrong\u003e12\u003c/strong\u003e, 17946 (2022). \u003c/li\u003e\n\u003cli\u003eChoudhary, D., Lagage, V., Foster, K. R. \u0026amp; Uphoff, S. Phenotypic heterogeneity in the bacterial oxidative stress response is driven by cell-cell interactions. \u003cem\u003eCell Reports \u003c/em\u003e\u003cstrong\u003e42\u003c/strong\u003e, 112168 (2023). \u003c/li\u003e\n\u003cli\u003eConlon, B. P. \u003cem\u003eet al.\u003c/em\u003e Persister formation in Staphylococcus aureus is associated with ATP depletion. \u003cem\u003eNat Microbiol \u003c/em\u003e\u003cstrong\u003e1\u003c/strong\u003e, 16051 (2016). \u003c/li\u003e\n\u003cli\u003eBuckling, A. \u003cem\u003eet al.\u003c/em\u003e Siderophore-mediated cooperation and virulence in Pseudomonas aeruginosa: Siderophore-mediated cooperation and virulence in P. aeruginosa. \u003cem\u003eFEMS Microbiology Ecology \u003c/em\u003e\u003cstrong\u003e62\u003c/strong\u003e, 135\u0026ndash;141 (2007). \u003c/li\u003e\n\u003cli\u003eChubiz, J. E. C., Golubeva, Y. A., Lin, D., Miller, L. D. \u0026amp; Slauch, J. M. FliZ Regulates Expression of the \u003cem\u003eSalmonella\u003c/em\u003e Pathogenicity Island 1 Invasion Locus by Controlling HilD Protein Activity in \u003cem\u003eSalmonella enterica\u003c/em\u003e Serovar Typhimurium. \u003cem\u003eJ Bacteriol \u003c/em\u003e\u003cstrong\u003e192\u003c/strong\u003e, 6261\u0026ndash;6270 (2010). \u003c/li\u003e\n\u003cli\u003eCahoon, L. A. \u0026amp; Seifert, H. S. Focusing homologous recombination: pilin antigenic variation in the pathogenic \u003cem\u003eNeisseria\u003c/em\u003e. \u003cem\u003eMolecular Microbiology \u003c/em\u003e\u003cstrong\u003e81\u003c/strong\u003e, 1136\u0026ndash;1143 (2011). \u003c/li\u003e\n\u003cli\u003eDavis, K. M., Mohammadi, S. \u0026amp; Isberg, R. R. Community Behavior and Spatial Regulation within a Bacterial Microcolony in Deep Tissue Sites Serves to Protect against Host Attack. \u003cem\u003eCell Host \u0026amp; Microbe \u003c/em\u003e\u003cstrong\u003e17\u003c/strong\u003e, 21\u0026ndash;31 (2015). \u003c/li\u003e\n\u003cli\u003eNuss, A. M. \u003cem\u003eet al.\u003c/em\u003e A Precise Temperature-Responsive Bistable Switch Controlling Yersinia Virulence. \u003cem\u003ePLoS Pathog \u003c/em\u003e\u003cstrong\u003e12\u003c/strong\u003e, e1006091 (2016). \u003c/li\u003e\n\u003cli\u003eMcCallum, G. \u0026amp; Tropini, C. The gut microbiota and its biogeography. \u003cem\u003eNat Rev Microbiol \u003c/em\u003e\u003cstrong\u003e22\u003c/strong\u003e, 105\u0026ndash;118 (2024). \u003c/li\u003e\n\u003cli\u003eDonaldson, G. P., Lee, S. M. \u0026amp; Mazmanian, S. K. Gut biogeography of the bacterial microbiota. \u003cem\u003eNat Rev Microbiol \u003c/em\u003e\u003cstrong\u003e14\u003c/strong\u003e, 20\u0026ndash;32 (2016). \u003c/li\u003e\n\u003cli\u003eSchlaeppi, K., Dombrowski, N., Oter, R. G., Ver Loren Van Themaat, E. \u0026amp; Schulze-Lefert, P. Quantitative divergence of the bacterial root microbiota in \u003cem\u003eArabidopsis thaliana\u003c/em\u003e relatives. \u003cem\u003eProc. Natl. Acad. Sci. U.S.A. \u003c/em\u003e\u003cstrong\u003e111\u003c/strong\u003e, 585\u0026ndash;592 (2014). \u003c/li\u003e\n\u003cli\u003eBai, Y. \u003cem\u003eet al.\u003c/em\u003e Functional overlap of the Arabidopsis leaf and root microbiota. \u003cem\u003eNature \u003c/em\u003e\u003cstrong\u003e528\u003c/strong\u003e, 364\u0026ndash;369 (2015). \u003c/li\u003e\n\u003cli\u003eBulgarelli, D. \u003cem\u003eet al.\u003c/em\u003e Revealing structure and assembly cues for Arabidopsis root-inhabiting bacterial microbiota. \u003cem\u003eNature \u003c/em\u003e\u003cstrong\u003e488\u003c/strong\u003e, 91\u0026ndash;95 (2012). \u003c/li\u003e\n\u003cli\u003eHou, K. \u003cem\u003eet al.\u003c/em\u003e Microbiota in health and diseases. \u003cem\u003eSig Transduct Target Ther \u003c/em\u003e\u003cstrong\u003e7\u003c/strong\u003e, 135 (2022). \u003c/li\u003e\n\u003cli\u003eAsnicar, F. \u003cem\u003eet al.\u003c/em\u003e Gut micro-organisms associated with health, nutrition and dietary interventions. \u003cem\u003eNature \u003c/em\u003ehttps://doi.org/10.1038/s41586-025-09854-7 (2025) doi:10.1038/s41586-025-09854-7. \u003c/li\u003e\n\u003cli\u003eChaparro, J. M., Badri, D. V. \u0026amp; Vivanco, J. M. Rhizosphere microbiome assemblage is affected by plant development. \u003cem\u003eThe ISME Journal \u003c/em\u003e\u003cstrong\u003e8\u003c/strong\u003e, 790\u0026ndash;803 (2014). \u003c/li\u003e\n\u003cli\u003eZhalnina, K. \u003cem\u003eet al.\u003c/em\u003e Dynamic root exudate chemistry and microbial substrate preferences drive patterns in rhizosphere microbial community assembly. \u003cem\u003eNat Microbiol \u003c/em\u003e\u003cstrong\u003e3\u003c/strong\u003e, 470\u0026ndash;480 (2018). \u003c/li\u003e\n\u003cli\u003eLoo, E. et al. Contribution of Sugar Transporters to Spatially Organized Colonization by Microbiota Along the Longitudinal Root Axis of Arabidopsis. https://www.ssrn.com/abstract=4514131 (2023) doi:10.2139/ssrn.4514131. \u003c/li\u003e\n\u003cli\u003eHawes, M. C. \u003cem\u003eet al.\u003c/em\u003e Extracellular DNA: The tip of root defenses? \u003cem\u003ePlant Science \u003c/em\u003e\u003cstrong\u003e180\u003c/strong\u003e, 741\u0026ndash;745 (2011). \u003c/li\u003e\n\u003cli\u003eBeck, M. \u003cem\u003eet al.\u003c/em\u003e Expression patterns of FLAGELLIN SENSING 2 map to bacterial entry sites in plant shoots and roots. \u003cem\u003eJournal of Experimental Botany \u003c/em\u003e\u003cstrong\u003e65\u003c/strong\u003e, 6487\u0026ndash;6498 (2014). \u003c/li\u003e\n\u003cli\u003eWyrsch, I., Dom\u0026iacute;nguez‐Ferreras, A., Geldner, N. \u0026amp; Boller, T. Tissue‐specific FLAGELLIN ‐ SENSING 2 ( FLS 2) expression in roots restores immune responses in A rabidopsis \u003cem\u003efls2\u003c/em\u003e mutants. \u003cem\u003eNew Phytologist \u003c/em\u003e\u003cstrong\u003e206\u003c/strong\u003e, 774\u0026ndash;784 (2015). \u003c/li\u003e\n\u003cli\u003eZhou, F. \u003cem\u003eet al.\u003c/em\u003e Co-incidence of Damage and Microbial Patterns Controls Localized Immune Responses in Roots. \u003cem\u003eCell \u003c/em\u003e\u003cstrong\u003e180\u003c/strong\u003e, 440-453.e18 (2020). \u003c/li\u003e\n\u003cli\u003eFortier, M. \u003cem\u003eet al.\u003c/em\u003e A fine-tuned defense at the pea root caps: Involvement of border cells and arabinogalactan proteins against soilborne diseases. \u003cem\u003eFront. Plant Sci. \u003c/em\u003e\u003cstrong\u003e14\u003c/strong\u003e, 1132132 (2023). \u003c/li\u003e\n\u003cli\u003eTsai, H.-H., Wang, J., Geldner, N. \u0026amp; Zhou, F. Spatiotemporal control of root immune responses during microbial colonization. \u003cem\u003eCurrent Opinion in Plant Biology \u003c/em\u003e\u003cstrong\u003e74\u003c/strong\u003e, 102369 (2023). \u003c/li\u003e\n\u003cli\u003eCampos, M. L., De Souza, C. M., De Oliveira, K. B. S., Dias, S. C. \u0026amp; Franco, O. L. The role of antimicrobial peptides in plant immunity. \u003cem\u003eJournal of Experimental Botany \u003c/em\u003e\u003cstrong\u003e69\u003c/strong\u003e, 4997\u0026ndash;5011 (2018). \u003c/li\u003e\n\u003cli\u003eKoprivova, A. \u003cem\u003eet al.\u003c/em\u003e Root-specific camalexin biosynthesis controls the plant growth-promoting effects of multiple bacterial strains. \u003cem\u003eProc. Natl. Acad. Sci. U.S.A. \u003c/em\u003e\u003cstrong\u003e116\u003c/strong\u003e, 15735\u0026ndash;15744 (2019). \u003c/li\u003e\n\u003cli\u003eLopa, F. R., Snigdha, F. N., Lumactud, R. A. \u0026amp; Sikder, M. M. Harnessing Camalexin as a Sustainable and Ecofriendly Strategy to Control Harmful Phytopathogens. \u003cem\u003ePlant Pathology \u003c/em\u003e\u003cstrong\u003e74\u003c/strong\u003e, 2463\u0026ndash;2477 (2025). \u003c/li\u003e\n\u003cli\u003eReichling, J. Plant-Microbe Interactions and Secondary Metabolites with Antibacterial, Antifungal and Antiviral Properties. in \u003cem\u003eFunctions and Biotechnology of Plant Secondary Metabolites\u003c/em\u003e (ed. Wink, M.) 214\u0026ndash;347 (Wiley-Blackwell, Oxford, UK, 2010). doi:10.1002/9781444318876.ch4. \u003c/li\u003e\n\u003cli\u003eBhandari, D. D., Kim, S.-J. \u0026amp; Brandizzi, F. Fortifying the frontier: cell wall modifications during plant immunity. \u003cem\u003eCurrent Opinion in Plant Biology \u003c/em\u003e\u003cstrong\u003e88\u003c/strong\u003e, 102816 (2025). \u003c/li\u003e\n\u003cli\u003eMalinovsky, F. G., Fangel, J. U. \u0026amp; Willats, W. G. T. The role of the cell wall in plant immunity. \u003cem\u003eFront. Plant Sci. \u003c/em\u003e\u003cstrong\u003e5\u003c/strong\u003e, (2014). \u003c/li\u003e\n\u003cli\u003eZipfel, C. \u0026amp; Oldroyd, G. E. D. Plant signalling in symbiosis and immunity. \u003cem\u003eNature \u003c/em\u003e\u003cstrong\u003e543\u003c/strong\u003e, 328\u0026ndash;336 (2017). \u003c/li\u003e\n\u003cli\u003eBaxter, A., Mittler, R. \u0026amp; Suzuki, N. ROS as key players in plant stress signalling. \u003cem\u003eJournal of Experimental Botany \u003c/em\u003e\u003cstrong\u003e65\u003c/strong\u003e, 1229\u0026ndash;1240 (2014). \u003c/li\u003e\n\u003cli\u003eJones, J. D. G. \u0026amp; Dangl, J. L. The plant immune system. \u003cem\u003eNature \u003c/em\u003e\u003cstrong\u003e444\u003c/strong\u003e, 323\u0026ndash;329 (2006). \u003c/li\u003e\n\u003cli\u003eMacho, A. P. \u0026amp; Zipfel, C. Plant PRRs and the Activation of Innate Immune Signaling. \u003cem\u003eMolecular Cell \u003c/em\u003e\u003cstrong\u003e54\u003c/strong\u003e, 263\u0026ndash;272 (2014). \u003c/li\u003e\n\u003cli\u003eLu, Y. \u0026amp; Tsuda, K. Intimate Association of PRR- and NLR-Mediated Signaling in Plant Immunity. \u003cem\u003eMPMI \u003c/em\u003e\u003cstrong\u003e34\u003c/strong\u003e, 3\u0026ndash;14 (2021). \u003c/li\u003e\n\u003cli\u003eAsai, S. \u0026amp; Shirasu, K. Plant cells under siege: plant immune system versus pathogen effectors. \u003cem\u003eCurrent Opinion in Plant Biology \u003c/em\u003e\u003cstrong\u003e28\u003c/strong\u003e, 1\u0026ndash;8 (2015). \u003c/li\u003e\n\u003cli\u003eda Cunha, L., Sreerekha, M.-V. \u0026amp; Mackey, D. Defense suppression by virulence effectors of bacterial phytopathogens. \u003cem\u003eCurrent Opinion in Plant Biology \u003c/em\u003e\u003cstrong\u003e10\u003c/strong\u003e, 349\u0026ndash;357 (2007). \u003c/li\u003e\n\u003cli\u003eGiraldo, M. C. \u0026amp; Valent, B. Filamentous plant pathogen effectors in action. \u003cem\u003eNat Rev Microbiol \u003c/em\u003e\u003cstrong\u003e11\u003c/strong\u003e, 800\u0026ndash;814 (2013). \u003c/li\u003e\n\u003cli\u003eWaheed, A. \u003cem\u003eet al.\u003c/em\u003e Effector Avr4 in Phytophthora infestans Escapes Host Immunity Mainly Through Early Termination. \u003cem\u003eFront Microbiol \u003c/em\u003e\u003cstrong\u003e12\u003c/strong\u003e, 646062 (2021). \u003c/li\u003e\n\u003cli\u003eNa, R. \u0026amp; Gijzen, M. Escaping Host Immunity: New Tricks for Plant Pathogens. \u003cem\u003ePLoS Pathog \u003c/em\u003e\u003cstrong\u003e12\u003c/strong\u003e, e1005631 (2016). \u003c/li\u003e\n\u003cli\u003eRuiz-Bedoya, T., Wang, P. W., Desveaux, D. \u0026amp; Guttman, D. S. Cooperative virulence via the collective action of secreted pathogen effectors. \u003cem\u003eNat Microbiol \u003c/em\u003e\u003cstrong\u003e8\u003c/strong\u003e, 640\u0026ndash;650 (2023). \u003c/li\u003e\n\u003cli\u003eDrago\u0026scaron;, A. \u003cem\u003eet al.\u003c/em\u003e Division of Labor during Biofilm Matrix Production. \u003cem\u003eCurrent Biology \u003c/em\u003e\u003cstrong\u003e28\u003c/strong\u003e, 1903-1913.e5 (2018). \u003c/li\u003e\n\u003cli\u003eMutlu, A. \u003cem\u003eet al.\u003c/em\u003e Phenotypic memory in Bacillus subtilis links dormancy entry and exit by a spore quantity-quality tradeoff. \u003cem\u003eNat Commun \u003c/em\u003e\u003cstrong\u003e9\u003c/strong\u003e, 69 (2018). \u003c/li\u003e\n\u003cli\u003eYannarell, S. M. \u003cem\u003eet al.\u003c/em\u003e Extensive cellular multi-tasking within \u003cem\u003eBacillus subtilis\u003c/em\u003e biofilms. \u003cem\u003emSystems\u003c/em\u003e e00891-22 (2023) doi:10.1128/msystems.00891-22. \u003c/li\u003e\n\u003cli\u003eOtto, S. B. \u003cem\u003eet al.\u003c/em\u003e Privatization of Biofilm Matrix in Structurally Heterogeneous Biofilms. \u003cem\u003emSystems \u003c/em\u003e\u003cstrong\u003e5\u003c/strong\u003e, e00425-20 (2020). \u003c/li\u003e\n\u003cli\u003eDergham, Y. \u003cem\u003eet al.\u003c/em\u003e Direct comparison of spatial transcriptional heterogeneity across diverse Bacillus subtilis biofilm communities. \u003cem\u003eNat Commun \u003c/em\u003e\u003cstrong\u003e14\u003c/strong\u003e, 7546 (2023). \u003c/li\u003e\n\u003cli\u003eLopez, D., Vlamakis, H. \u0026amp; Kolter, R. Generation of multiple cell types in \u003cem\u003eBacillus subtilis\u003c/em\u003e. \u003cem\u003eFEMS Microbiol Rev \u003c/em\u003e\u003cstrong\u003e33\u003c/strong\u003e, 152\u0026ndash;163 (2009). \u003c/li\u003e\n\u003cli\u003eKearns, D. B. \u0026amp; Losick, R. Cell population heterogeneity during growth of Bacillus subtilis. \u003cem\u003eGenes Dev \u003c/em\u003e\u003cstrong\u003e19\u003c/strong\u003e, 3083\u0026ndash;3094 (2005). \u003c/li\u003e\n\u003cli\u003eStoodley, P., Sauer, K., Davies, D. G. \u0026amp; Costerton, J. W. Biofilms as Complex Differentiated Communities. \u003cem\u003eAnnu. Rev. Microbiol. \u003c/em\u003e\u003cstrong\u003e56\u003c/strong\u003e, 187\u0026ndash;209 (2002). \u003c/li\u003e\n\u003cli\u003eVlamakis, H., Aguilar, C., Losick, R. \u0026amp; Kolter, R. Control of cell fate by the formation of an architecturally complex bacterial community. \u003cem\u003eGenes Dev. \u003c/em\u003e\u003cstrong\u003e22\u003c/strong\u003e, 945\u0026ndash;953 (2008). \u003c/li\u003e\n\u003cli\u003evan Gestel, J., Vlamakis, H. \u0026amp; Kolter, R. From Cell Differentiation to Cell Collectives: Bacillus subtilis Uses Division of Labor to Migrate. \u003cem\u003ePLoS Biol \u003c/em\u003e\u003cstrong\u003e13\u003c/strong\u003e, e1002141 (2015). \u003c/li\u003e\n\u003cli\u003evan Gestel, J., Vlamakis, H. \u0026amp; Kolter, R. Division of Labor in Biofilms: the Ecology of Cell Differentiation. \u003cem\u003eMicrobiol Spectr \u003c/em\u003e\u003cstrong\u003e3\u003c/strong\u003e, (2015). \u003c/li\u003e\n\u003cli\u003eLord, N. D. \u003cem\u003eet al.\u003c/em\u003e Stochastic antagonism between two proteins governs a bacterial cell fate switch. \u003cem\u003eScience \u003c/em\u003e\u003cstrong\u003e366\u003c/strong\u003e, 116\u0026ndash;120 (2019). \u003c/li\u003e\n\u003cli\u003eBai, U., Mandic-Mulec, I. \u0026amp; Smith, I. SinI modulates the activity of SinR, a developmental switch protein of Bacillus subtilis, by protein-protein interaction. \u003cem\u003eGenes Dev. \u003c/em\u003e\u003cstrong\u003e7\u003c/strong\u003e, 139\u0026ndash;148 (1993). \u003c/li\u003e\n\u003cli\u003eNewman, J. A., Rodrigues, C. \u0026amp; Lewis, R. J. Molecular Basis of the Activity of SinR Protein, the Master Regulator of Biofilm Formation in Bacillus subtilis. \u003cem\u003eJournal of Biological Chemistry \u003c/em\u003e\u003cstrong\u003e288\u003c/strong\u003e, 10766\u0026ndash;10778 (2013). \u003c/li\u003e\n\u003cli\u003eKampf, J. \u003cem\u003eet al.\u003c/em\u003e Selective Pressure for Biofilm Formation in Bacillus subtilis: Differential Effect of Mutations in the Master Regulator SinR on Bistability. \u003cem\u003emBio \u003c/em\u003e\u003cstrong\u003e9\u003c/strong\u003e, e01464-18 (2018). \u003c/li\u003e\n\u003cli\u003eChai, Y., Chu, F., Kolter, R. \u0026amp; Losick, R. Bistability and biofilm formation in Bacillus subtilis: Bistability and biofilm formation in Bacillus subtilis. \u003cem\u003eMolecular Microbiology \u003c/em\u003e\u003cstrong\u003e67\u003c/strong\u003e, 254\u0026ndash;263 (2007). \u003c/li\u003e\n\u003cli\u003eDannenberg, S., Penning, J., Simm, A. \u0026amp; Klumpp, S. The motility-matrix production switch in Bacillus subtilis-a modeling perspective. \u003cem\u003eJ Bacteriol \u003c/em\u003e\u003cstrong\u003e206\u003c/strong\u003e, e0004723 (2024). \u003c/li\u003e\n\u003cli\u003eNorman, T. M., Lord, N. D., Paulsson, J. \u0026amp; Losick, R. Memory and modularity in cell-fate decision making. \u003cem\u003eNature \u003c/em\u003e\u003cstrong\u003e503\u003c/strong\u003e, 481\u0026ndash;486 (2013). \u003c/li\u003e\n\u003cli\u003eArnaouteli, S., Bamford, N. C., Stanley-Wall, N. R. \u0026amp; Kov\u0026aacute;cs, \u0026Aacute;. T. Bacillus subtilis biofilm formation and social interactions. \u003cem\u003eNat Rev Microbiol \u003c/em\u003e\u003cstrong\u003e19\u003c/strong\u003e, 600\u0026ndash;614 (2021). \u003c/li\u003e\n\u003cli\u003eXu, S. \u003cem\u003eet al.\u003c/em\u003e The \u003cem\u003espo0A-sinI-sinR\u003c/em\u003e Regulatory Circuit Plays an Essential Role in Biofilm Formation, Nematicidal Activities, and Plant Protection in \u003cem\u003eBacillus cereus\u003c/em\u003e AR156. \u003cem\u003eMPMI \u003c/em\u003e\u003cstrong\u003e30\u003c/strong\u003e, 603\u0026ndash;619 (2017). \u003c/li\u003e\n\u003cli\u003eMcLoon, A. L., Kolodkin-Gal, I., Rubinstein, S. M., Kolter, R. \u0026amp; Losick, R. Spatial Regulation of Histidine Kinases Governing Biofilm Formation in \u003cem\u003eBacillus subtilis\u003c/em\u003e. \u003cem\u003eJ Bacteriol \u003c/em\u003e\u003cstrong\u003e193\u003c/strong\u003e, 679\u0026ndash;685 (2011). \u003c/li\u003e\n\u003cli\u003eL\u0026oacute;pez-Pag\u0026aacute;n, N. \u003cem\u003eet al.\u003c/em\u003e Pseudomonas syringae subpopulations cooperate by coordinating flagellar and type III secretion spatiotemporal dynamics to facilitate plant infection. \u003cem\u003eNat Microbiol \u003c/em\u003e\u003cstrong\u003e10\u003c/strong\u003e, 958\u0026ndash;972 (2025). \u003c/li\u003e\n\u003cli\u003eAllard-Massicotte, R. \u003cem\u003eet al.\u003c/em\u003e Bacillus subtilis Early Colonization of Arabidopsis thaliana Roots Involves Multiple Chemotaxis Receptors. \u003cem\u003emBio \u003c/em\u003e\u003cstrong\u003e7\u003c/strong\u003e, (2016). \u003c/li\u003e\n\u003cli\u003eBoubsi, F. \u003cem\u003eet al.\u003c/em\u003e Pectic homogalacturonan sensed by Bacillus acts as host associated cue to promote establishment and persistence in the rhizosphere. \u003cem\u003eiScience \u003c/em\u003e\u003cstrong\u003e26\u003c/strong\u003e, 107925 (2023). \u003c/li\u003e\n\u003cli\u003eEngelhardt, I. C., Holden, N., Daniell, T. J. \u0026amp; Dupuy, L. X. Mobility and growth in confined spaces are important mechanisms for the establishment of Bacillus subtilis in the rhizosphere. \u003cem\u003eMicrobiology \u003c/em\u003e\u003cstrong\u003e170\u003c/strong\u003e, (2024). \u003c/li\u003e\n\u003cli\u003eLee, Y., Kwon, S., Balaraju, K. \u0026amp; Jeon, Y. Influence of phenotypic variation of Paenibacillus polymyxa E681 on growth promotion in cucumbers. \u003cem\u003eFront. Microbiol. \u003c/em\u003e\u003cstrong\u003e15\u003c/strong\u003e, 1427265 (2024). \u003c/li\u003e\n\u003cli\u003eDobrange, E. \u0026amp; Van Den Ende, W. Bacterial cell differentiation during plant root colonization: the putative role of fructans. \u003cem\u003ePhysiologia Plantarum \u003c/em\u003e\u003cstrong\u003e177\u003c/strong\u003e, e70095 (2025). \u003c/li\u003e\n\u003cli\u003eTian, T. \u003cem\u003eet al.\u003c/em\u003e Sucrose triggers a novel signaling cascade promoting Bacillus subtilis rhizosphere colonization. \u003cem\u003eISME J \u003c/em\u003e\u003cstrong\u003e15\u003c/strong\u003e, 2723\u0026ndash;2737 (2021). \u003c/li\u003e\n\u003cli\u003eBeauregard, P. B., Chai, Y., Vlamakis, H., Losick, R. \u0026amp; Kolter, R. Bacillus subtilis biofilm induction by plant polysaccharides. \u003cem\u003eProceedings of the National Academy of Sciences \u003c/em\u003e\u003cstrong\u003e110\u003c/strong\u003e, E1621\u0026ndash;E1630 (2013). \u003c/li\u003e\n\u003cli\u003eHoff, G. \u003cem\u003eet al.\u003c/em\u003e Surfactin Stimulated by Pectin Molecular Patterns and Root Exudates Acts as a Key Driver of the Bacillus-Plant Mutualistic Interaction. \u003cem\u003emBio \u003c/em\u003e\u003cstrong\u003e12\u003c/strong\u003e, e01774-21 (2021). \u003c/li\u003e\n\u003cli\u003eGozzi, K. \u003cem\u003eet al.\u003c/em\u003e Bacillus subtilis utilizes the DNA damage response to manage multicellular development. \u003cem\u003enpj Biofilms Microbiomes \u003c/em\u003e\u003cstrong\u003e3\u003c/strong\u003e, 8 (2017). \u003c/li\u003e\n\u003cli\u003eKaiser, C.-F., Perilli, A., Grossmann, G. \u0026amp; Meroz, Y. Studying root\u0026ndash;environment interactions in structured microdevices. \u003cem\u003eJournal of Experimental Botany \u003c/em\u003e\u003cstrong\u003e74\u003c/strong\u003e, 3851\u0026ndash;3863 (2023). \u003c/li\u003e\n\u003cli\u003eStanley, C. E. \u003cem\u003eet al.\u003c/em\u003e Dual-flow-RootChip reveals local adaptations of roots towards environmental asymmetry at the physiological and genetic levels. \u003cem\u003eNew Phytol \u003c/em\u003e\u003cstrong\u003e217\u003c/strong\u003e, 1357\u0026ndash;1369 (2018). \u003c/li\u003e\n\u003cli\u003eMassalha, H., Korenblum, E., Malitsky, S., Shapiro, O. H. \u0026amp; Aharoni, A. Live imaging of root\u0026ndash;bacteria interactions in a microfluidics setup. \u003cem\u003eProc Natl Acad Sci USA \u003c/em\u003e\u003cstrong\u003e114\u003c/strong\u003e, 4549\u0026ndash;4554 (2017). \u003c/li\u003e\n\u003cli\u003eZengler, K. \u003cem\u003eet al.\u003c/em\u003e EcoFABs: advancing microbiome science through standardized fabricated ecosystems. \u003cem\u003eNat Methods \u003c/em\u003e\u003cstrong\u003e16\u003c/strong\u003e, 567\u0026ndash;571 (2019). \u003c/li\u003e\n\u003cli\u003eNoirot-Gros, M.-F. \u003cem\u003eet al.\u003c/em\u003e Functional Imaging of Microbial Interactions With Tree Roots Using a Microfluidics Setup. \u003cem\u003eFront. Plant Sci. \u003c/em\u003e\u003cstrong\u003e11\u003c/strong\u003e, 408 (2020). \u003c/li\u003e\n\u003cli\u003eDai, H., Wu, B., Chen, B., Ma, B. \u0026amp; Chu, C. Diel Fluctuation of Extracellular Reactive Oxygen Species Production in the Rhizosphere of Rice. \u003cem\u003eEnviron. Sci. Technol. \u003c/em\u003e\u003cstrong\u003e56\u003c/strong\u003e, 9075\u0026ndash;9082 (2022). \u003c/li\u003e\n\u003cli\u003eGrossmann, G. \u003cem\u003eet al.\u003c/em\u003e The RootChip: An Integrated Microfluidic Chip for Plant Science. \u003cem\u003eThe Plant Cell \u003c/em\u003e\u003cstrong\u003e23\u003c/strong\u003e, 4234\u0026ndash;4240 (2011). \u003c/li\u003e\n\u003cli\u003eDenninger, P. \u003cem\u003eet al.\u003c/em\u003e Distinct RopGEFs Successively Drive Polarization and Outgrowth of Root Hairs. \u003cem\u003eCurrent Biology \u003c/em\u003e\u003cstrong\u003e29\u003c/strong\u003e, 1854-1865.e5 (2019). \u003c/li\u003e\n\u003cli\u003eDurrett, R. \u003cem\u003eet al.\u003c/em\u003e Genome Sequence of the Bacillus subtilis Biofilm-Forming Transformable Strain PS216. \u003cem\u003eGenome Announc \u003c/em\u003e\u003cstrong\u003e1\u003c/strong\u003e, (2013). \u003c/li\u003e\n\u003cli\u003eFan, B. \u003cem\u003eet al.\u003c/em\u003e Bacillus velezensis FZB42 in 2018: The Gram-Positive Model Strain for Plant Growth Promotion and Biocontrol. \u003cem\u003eFront. Microbiol. \u003c/em\u003e\u003cstrong\u003e9\u003c/strong\u003e, 2491 (2018). \u003c/li\u003e\n\u003cli\u003eTrauth, S. \u0026amp; Bischofs, I. B. Ectopic Integration Vectors for Generating Fluorescent Promoter Fusions in Bacillus subtilis with Minimal Dark Noise. \u003cem\u003ePLoS ONE \u003c/em\u003e\u003cstrong\u003e9\u003c/strong\u003e, e98360 (2014). \u003c/li\u003e\n\u003cli\u003eGibson, D. G. \u003cem\u003eet al.\u003c/em\u003e Enzymatic assembly of DNA molecules up to several hundred kilobases. \u003cem\u003eNat Methods \u003c/em\u003e\u003cstrong\u003e6\u003c/strong\u003e, 343\u0026ndash;345 (2009). \u003c/li\u003e\n\u003cli\u003eHauser, P. M. \u0026amp; Karamata, D. A rapid and simple method for Bacillus subtilis transformation on solid media. \u003cem\u003eMicrobiology \u003c/em\u003e\u003cstrong\u003e140\u003c/strong\u003e, 1613\u0026ndash;1617 (1994). \u003c/li\u003e\n\u003cli\u003eHarwood, C. R. \u0026amp; Cutting, S. M. \u003cem\u003eMolecular Biological Methods for Bacillus\u003c/em\u003e. (J. Wiley \u0026amp; sons, Chichester New York Brisbane [etc.], 1990). \u003c/li\u003e\n\u003cli\u003eKoo, B.-M. \u003cem\u003eet al.\u003c/em\u003e Construction and Analysis of Two Genome-Scale Deletion Libraries for Bacillus subtilis. \u003cem\u003eCell Systems \u003c/em\u003e\u003cstrong\u003e4\u003c/strong\u003e, 291-305.e7 (2017). \u003c/li\u003e\n\u003cli\u003eHoagland, D. R. \u0026amp; Arnon, D. I. The water-culture method for growing plants without soil. \u003cstrong\u003e347\u003c/strong\u003e, 32 pp. (1950). \u003c/li\u003e\n\u003cli\u003eGuichard, M., Bertran Garcia de Olalla, E., Stanley, C. E. \u0026amp; Grossmann, G. Microfluidic systems for plant root imaging. in \u003cem\u003eMethods in Cell Biology\u003c/em\u003e vol. 160 381\u0026ndash;404 (Elsevier, 2020). \u003c/li\u003e\n\u003cli\u003eEdelstein, A., Amodaj, N., Hoover, K., Vale, R. \u0026amp; Stuurman, N. Computer Control of Microscopes Using \u0026micro;Manager. \u003cem\u003eCP Molecular Biology \u003c/em\u003e\u003cstrong\u003e92\u003c/strong\u003e, (2010). \u003c/li\u003e\n\u003cli\u003eD. Edelstein, A. \u003cem\u003eet al.\u003c/em\u003e Advanced methods of microscope control using \u0026mu;Manager software. \u003cem\u003eJBM \u003c/em\u003e\u003cstrong\u003e1\u003c/strong\u003e, 1 (2014). \u003c/li\u003e\n\u003cli\u003eSchindelin, J. \u003cem\u003eet al.\u003c/em\u003e Fiji: an open-source platform for biological-image analysis. \u003cem\u003eNat Methods \u003c/em\u003e\u003cstrong\u003e9\u003c/strong\u003e, 676\u0026ndash;682 (2012). \u003c/li\u003e\n\u003cli\u003eJefferis, G., Kemp, S. E., Arya, S. \u0026amp; Mount, D. RANN: Fast Nearest Neighbour Search (Wraps ANN Library) Using L2 Metric. 2.6.2 https://doi.org/10.32614/CRAN.package.RANN (2013). \u003c/li\u003e\n\u003cli\u003eBaddeley, A., Rubak, E. \u0026amp; Turner, R. \u003cem\u003eSpatial Point Patterns: Methodology and Applications with R\u003c/em\u003e. (CRC Press, Boca Raton London New York, 2016). doi:10.1201/b19708. \u003c/li\u003e\n\u003cli\u003ePreibisch, S., Saalfeld, S. \u0026amp; Tomancak, P. Globally optimal stitching of tiled 3D microscopic image acquisitions. \u003cem\u003eBioinformatics \u003c/em\u003e\u003cstrong\u003e25\u003c/strong\u003e, 1463\u0026ndash;1465 (2009). \u003c/li\u003e\n\u003cli\u003eWood, S. mgcv: Mixed GAM Computation Vehicle with Automatic Smoothness Estimation. 1.9-4 https://doi.org/10.32614/CRAN.package.mgcv (2000). \u003c/li\u003e\n\u003cli\u003eBranda, S. S., Chu, F., Kearns, D. B., Losick, R. \u0026amp; Kolter, R. A major protein component of the Bacillus subtilis biofilm matrix. \u003cem\u003eMol Microbiol \u003c/em\u003e\u003cstrong\u003e59\u003c/strong\u003e, 1229\u0026ndash;1238 (2006). \u003c/li\u003e\n\u003cli\u003eVerbelen, J.-P., Cnodder, T. D., Le, J., Vissenberg, K. \u0026amp; Balu\u0026scaron;ka, F. The Root Apex of\u003cem\u003eArabidopsis thaliana\u003c/em\u003eConsists of Four Distinct Zones of Growth Activities: Meristematic Zone, Transition Zone, Fast Elongation Zone and Growth Terminating Zone. \u003cem\u003ePlant Signaling \u0026amp; Behavior \u003c/em\u003e\u003cstrong\u003e1\u003c/strong\u003e, 296\u0026ndash;304 (2006). \u003c/li\u003e\n\u003cli\u003eSteinfeld, B. K., Cui, Q., Schmidt, T. \u0026amp; Bischofs, I. B. Communication Determines Population-Level Fitness under Cation Stress by Modulating the Ratio of Motile to Sessile B. Subtilis Cells. http://biorxiv.org/lookup/doi/10.1101/2021.11.30.470380 (2021) doi:10.1101/2021.11.30.470380. \u003c/li\u003e\n\u003cli\u003eDavletova, S., Schlauch, K., Coutu, J. \u0026amp; Mittler, R. The zinc-finger protein Zat12 plays a central role in reactive oxygen and abiotic stress signaling in Arabidopsis. \u003cem\u003ePlant Physiol \u003c/em\u003e\u003cstrong\u003e139\u003c/strong\u003e, 847\u0026ndash;856 (2005). \u003c/li\u003e\n\u003cli\u003eTorres, M. A., Dangl, J. L. \u0026amp; Jones, J. D. G. \u003cem\u003eArabidopsis\u003c/em\u003e gp91 \u003csup\u003ephox\u003c/sup\u003e homologues \u003cem\u003eAtrbohD\u003c/em\u003e and \u003cem\u003eAtrbohF\u003c/em\u003e are required for accumulation of reactive oxygen intermediates in the plant defense response. \u003cem\u003eProc. Natl. Acad. Sci. U.S.A. \u003c/em\u003e\u003cstrong\u003e99\u003c/strong\u003e, 517\u0026ndash;522 (2002). \u003c/li\u003e\n\u003cli\u003eDaudi, A. \u003cem\u003eet al.\u003c/em\u003e The Apoplastic Oxidative Burst Peroxidase in \u003cem\u003eArabidopsis\u003c/em\u003e Is a Major Component of Pattern-Triggered Immunity. \u003cem\u003eThe Plant Cell \u003c/em\u003e\u003cstrong\u003e24\u003c/strong\u003e, 275\u0026ndash;287 (2012). \u003c/li\u003e\n\u003cli\u003ePodg\u0026oacute;rska, A., Burian, M. \u0026amp; Szal, B. Extra-Cellular But Extra-Ordinarily Important for Cells: Apoplastic Reactive Oxygen Species Metabolism. \u003cem\u003eFront. Plant Sci. \u003c/em\u003e\u003cstrong\u003e8\u003c/strong\u003e, (2017). \u003c/li\u003e\n\u003cli\u003eArnaud, D., Deeks, M. J. \u0026amp; Smirnoff, N. RBOHF activates stomatal immunity by modulating both reactive oxygen species and apoplastic pH dynamics in Arabidopsis. \u003cem\u003eThe Plant Journal \u003c/em\u003e\u003cstrong\u003e116\u003c/strong\u003e, 404\u0026ndash;415 (2023). \u003c/li\u003e\n\u003cli\u003eCheeseman, J. M. Hydrogen peroxide concentrations in leaves under natural conditions. \u003cem\u003eJournal of Experimental Botany \u003c/em\u003e\u003cstrong\u003e57\u003c/strong\u003e, 2435\u0026ndash;2444 (2006). \u003c/li\u003e\n\u003cli\u003eLiu, P. \u003cem\u003eet al.\u003c/em\u003e Length-based separation of Bacillus subtilis bacterial populations by viscoelastic microfluidics. \u003cem\u003eMicrosyst Nanoeng \u003c/em\u003e\u003cstrong\u003e8\u003c/strong\u003e, 7 (2022). \u003c/li\u003e\n\u003cli\u003eDion, M. F. \u003cem\u003eet al.\u003c/em\u003e Bacillus subtilis cell diameter is determined by the opposing actions of two distinct cell wall synthetic systems. \u003cem\u003eNat Microbiol \u003c/em\u003e\u003cstrong\u003e4\u003c/strong\u003e, 1294\u0026ndash;1305 (2019). \u003c/li\u003e\n\u003cli\u003eKnights, H. E., Jorrin, B., Haskett, T. L. \u0026amp; Poole, P. S. Deciphering bacterial mechanisms of root colonization. \u003cem\u003eEnvironmental Microbiology Reports \u003c/em\u003e\u003cstrong\u003e13\u003c/strong\u003e, 428\u0026ndash;444 (2021). \u003c/li\u003e\n\u003cli\u003eYeh, Y.-H., Chang, Y.-H., Huang, P.-Y., Huang, J.-B. \u0026amp; Zimmerli, L. Enhanced Arabidopsis pattern-triggered immunity by overexpression of cysteine-rich receptor-like kinases. \u003cem\u003eFront Plant Sci \u003c/em\u003e\u003cstrong\u003e6\u003c/strong\u003e, 322 (2015). \u003c/li\u003e\n\u003cli\u003eAngelini, L. L. \u003cem\u003eet al.\u003c/em\u003e Pulcherrimin protects Bacillus subtilis against oxidative stress during biofilm development. \u003cem\u003enpj Biofilms Microbiomes \u003c/em\u003e\u003cstrong\u003e9\u003c/strong\u003e, (2023). \u003c/li\u003e\n\u003cli\u003eMuratov, E., Keilholz, J., Kov\u0026aacute;cs, \u0026Aacute;. T. \u0026amp; Moeller, R. The biofilm matrix protects Bacillu subtilis against hydrogen peroxide. \u003cem\u003eBiofilm \u003c/em\u003e\u003cstrong\u003e9\u003c/strong\u003e, 100274 (2025). \u003c/li\u003e\n\u003cli\u003eAldin, M. S. \u0026amp; Tzipilevich, E. Salmonella exploits the reactive oxygen species generated by the plant immune system to enhance colonization. Preprint at https://doi.org/10.1101/2024.03.04.583411 (2024). \u003c/li\u003e\n\u003cli\u003eKulkarni, R. \u003cem\u003eet al.\u003c/em\u003e Cigarette Smoke Increases Staphylococcus aureus Biofilm Formation via Oxidative Stress. \u003cem\u003eInfect Immun \u003c/em\u003e\u003cstrong\u003e80\u003c/strong\u003e, 3804\u0026ndash;3811 (2012). \u003c/li\u003e\n\u003cli\u003ePfeilmeier, S. et al. The plant NADPH oxidase RBOHD is required for microbiota homeostasis in leaves. Nat Microbiol 6, 852\u0026ndash;864 (2021). \u003c/li\u003e\n\u003cli\u003eMa, K.-W. et al. Coordination of microbe\u0026ndash;host homeostasis by crosstalk with plant innate immunity. Nat. Plants 7, 814\u0026ndash;825 (2021).\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTables are available in the Supplementary Files section.\u003c/p\u003e\n"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-8407390/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8407390/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Phenotypic heterogeneity is an acclimation strategy in bacteria to cope with fluctuating environments and host immunity through cellular specialization. In the rhizosphere, how root-associated bacteria respond to plant defense signals and whether plants actively shape colonization through bacterial cell-type decisions, remains poorly understood. Cell type-resolved, spatiotemporal mapping of Bacillus subtilis and Bacillus velezensis colonizing Arabidopsis thaliana roots reveals dynamic transitions between flagellated and matrix-producing cell states. We identify an RBOHD-dependent, spatially heterogeneous oxidative landscape generated by the plant as a key regulator of bacterial cell-state balance. We show that exposure to reactive oxygen species (ROS) not only promotes differentiation into matrix-producing Bacillus cells, but ROS acclimation is also a prerequisite for cell type switching and stable association. Our findings demonstrate that bacterial stress evasion and acclimation to the host’s oxidative landscape are key steps in a host-microbe feedback system enabling stable rhizoplane colonization.","manuscriptTitle":"Root-derived reactive oxygen species spatially pattern bacterial cell-type differentiation during root colonization","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-11 10:44:48","doi":"10.21203/rs.3.rs-8407390/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-microbiology","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"nmicrobiol","sideBox":"Learn more about [Nature Microbiology](http://www.nature.com/nmicrobiol/)","snPcode":"","submissionUrl":"","title":"Nature Microbiology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Research","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"5751819d-602b-4d5a-93b9-20cf5348fe82","owner":[],"postedDate":"May 11th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":63610532,"name":"Biological sciences/Plant sciences/Plant symbiosis"},{"id":63610533,"name":"Biological sciences/Microbiology/Bacteria/Bacterial host response"},{"id":63610534,"name":"Biological sciences/Microbiology/Bacteria/Bacterial immune evasion"},{"id":63610535,"name":"Biological sciences/Microbiology/Bacteria/Bacterial development"},{"id":63610536,"name":"Biological sciences/Cell biology/Cellular imaging"}],"tags":[],"updatedAt":"2026-05-11T10:44:48+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-11 10:44:48","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8407390","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8407390","identity":"rs-8407390","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}