Figure 1. PBP2 molecules reside in diffusive, immobile, or persistently moving states.
(A) Representative trajectories of PAmCherry-PBP2 molecules (TKL130) obtained by high-frequency imaging (time interval 60 ms) reveals diffusive (blue) and bound (orange) molecules. (B) Probability distribution of single-molecule jump lengths (solid lines, colored) and fit (dashed, black) using a three-state-diffusion model for different time intervals. 78% of PBP2 move diffusively with <D> = 0.042 μm2/s while 22% are immobile. The shaded area indicates standard deviation between six biological replicates. (C–E) Low-frequency imaging (3.6 s with 1 s exposure time) reveals that bound PBP2 molecules are either persistently moving (C) or immobile (D), according to the instantaneous PBP2 velocity. PBP2 molecules show transitions between persistent and immobile states (E). (F–G) Persistently moving PBP2 and MreB filaments show similar speeds (F) and orientations of motion (orientation measured with respect to the cell centerline) (G). (H–I) Average fractions of bound, diffusive, persistently moving, and immobile PAmCherry-PBP2 at native levels (TKL130) (H) or if overexpressed (TKL130/pKC128) (I). Dots show biological replicates.
Figure 1—source data 1. Table containing all data presented in Figure 1 and Figure 1—figure supplements 1–10.
The file Dataset-Fig1.xlsx contains: Figure 1H-I Number of tracks, average diffusion constants, bound and persistent fractions for PAmCherry-PBP2 for native and overexpression conditions, including p-values for bound and persistent fractions. Figure 1F-G Numerical values of the distribution of speeds and orientations of PAmCherry-PBP2 and MreB-msfGFP tracks from low frequency imaging and number of tracks considered. Figure 1—figure supplement 1B-C. Width, length and doubling times obtained from OD600 measurements of MG1655, TKL130, and TKL130/pKC128 cells. Figure 1—figure supplement 1D Bocillin labeling, fluorescence and mass spectrometry measurements of MG1655, TKL130, and TKL130/pKC128 cells. Figure 1—figure supplement 2 Values of bound fractions, average diffusion constants, and sum of residuals of PAmCherry-PBP2 tracks from TKL130/pKC128 strain with high frequency imaging given by the Spot-On method using 2- or 3-state diffusion model. Figure 1—figure supplement 3 Values of bound fractions and diffusion constants of PAmCherry-PBP2 tracks from TKL130/pKC128 strain to compare the results from Spot-On with the Deff based method. Figure 1—figure supplement 5 Raw data for Figure 1F–G. List of velocity, orientation and goodness of fit (R2). Figure 1—figure supplement 7 Number of bound or active PBP2 molecules as a function of the number of PBP2 molecules per cell acquired from mass spectrometry and fluorescence measurements. Figure 1—figure supplement 8A) Shape and OD600 measurements of cells carrying the msfGFP-PBP2 fusion expressed from the native or PLac promoters with different induction levels of IPTG. Figure 1—figure supplement 8B Fluorescence and Western Blot quantifications and mass spectrometry (both DIA and PRM) of cells carrying the msfGFP-PBP2 fusion expressed from the native or PLac promoters with different induction levels of IPTG Figure 1—figure supplement 8C Values of number of tracks, average diffusion constants, and bound fractions for cells carrying the msfGFP-PBP2 fusion, including p-values for bound fractions. Figure 1—figure supplement 9 Shape and fluorescence measurements, number of tracks, average diffusion constants and bound fractions from high frequency imaging of cells carrying msfGFP-PBP2 fusion with induction levels of 5 and 25 uM IPTG after different times of growth, including p-values for bound fractions. Figure 1—figure supplement 10A Values of number of tracks, and persistent fractions for cells carrying msfGFP-PBP2 fusion expressed from the native or PLac promoters with different induction levels of IPTG, including p-values for persistent fractions. Figure 1—figure supplement 10B. List of velocity and goodness of fit (R2) values for cells carrying msfGFP-PBP2 fusion under different induction levels of IPTG.
elife-50629-fig1-data1.xlsx (552.9KB, xlsx)
Figure 1—figure supplement 1. Comparison of PBP2-PAmCherry expressing cells and WT.
(A) Bocillin-binding assay to compare expression levels of PBP2 in the wild-type strain (MG1655), the strain expressing PBP2-PAmCherry from the native locus (TKL130), and the strain overexpressing PBP2-PAmCherry (TKL130/pKC128). Quantification in (D). (B) Average cell dimensions obtained by phase-contrast microscopy and computational image segmentation. (C) Average doubling times during steady-state exponential growth in batch culture (from OD600). (D) Different methods to compare PBP2 expression levels in different strains (from left to right): Bocillin labeling (from A), single-cell fluorescence levels measured in epi-fluorescence mode, mass spectrometry [Data Independent Acquisitions (DIA) and Parallel Reaction Monitoring (PRM)]. For the first three methods, PBP2 levels are normalized by the corresponding value in TKL130. For PRM, we obtained absolute numbers of proteins per cell by comparing to reference peptides and colony counting. With both mass spectrometry methods, we observe a higher fold-change than through the other methods. Dots represent biological replicates.
Figure 1—figure supplement 2. Comparing 2- and 3-state-diffusion models to fit experimental data through Spot-On.
(A) Probability distributions of single-molecule jump lengths (solid lines, colored) and fit (dashed, black) using a two-state (left) or three-state (right) diffusion model for different time intervals for native levels (TKL130) and for over-expression (TKL130/pKC128) of PBP2-PAmCherry. Shaded regions show standard deviations between biological replicates. (B) Comparison of bound fractions and average diffusion constants acquired by fitting two-state and three-state diffusion models shown in (A). Dots represent biological replicates. (C) Normalized sum of residuals found by using multi-state models with Spot-On. Error bars show standard error between biological replicates.
Figure 1—figure supplement 3. An alternative approach to fit a two-state diffusion model, based on the distribution of effective diffusion constants.
(A) Heat map of sum of squared differences (RSS) between the Deff distributions of single-track effective diffusion constants Deff obtained from experimental data or computational simulations of a two-state model, using different model parameters D (diffusion constant of the diffusive fraction) and σ (localization precision). Parameter sets giving the lowest 5 RSS values are shown with green diamonds. Best fit is given by D = 0.04 μm2/s and σ = 20 nm. (B) We verified that the non-diffusive population was indeed not diffusing, with Dbound = 0 μm2/s (left), while a finite diffusion constant Dbound > 0.002 um2/s gives poor agreement between simulation and experiment. Here, the experimental Deff distribution is the mean of 6 biological replicates. (C) We compared the results of our method with the Spot-On code (2-state model) in TKL130 (native levels) and TKL130/pKC128 (overexpression), respectively.
Figure 1—figure supplement 4. Transitions between immobile and persistent states.
Example tracks and velocity as a function of time for example tracks that show transitions between persistent and immobile states.
Figure 1—figure supplement 5. Analysis of bound PAmCherry-PBP2 molecules.
(A) Velocity distribution of all PBP2 tracks measured with 1 s intervals. The velocity of individual tracks was determined by fitting a quadratic function to the MSD. (B) Velocity distributions for directed trajectories of PBP2 and MreB as found by selecting for an increased goodness of fit measured by R2 of a quadratic function to the MSD. (C) The stricter the goodness of fit criterion (minimal R2) the less trajectories contribute to the mean track velocity. (D) The mean velocity increases with increasing minimum R2. The dashed line indicates the value chose for the distributions in Figure 1. (E–F) The same analysis applied on 4-step segments of trajectories measured with 3.6 s intervals delivers smaller mean velocities, likely because fast trajectories reside for a shorter amount of time in the field of view.
Figure 1—figure supplement 6. Quantitative analysis of persistent and immobile states based on computational simulations.
(A) Distribution of measured single-step displacements in one dimension. A fit of a normal distribution to the data delivers a standard deviation of 36 nm, which corresponds to a localization error of single localization events of 25 nm. (B) We computationally simulated trajectories such that the length distribution of the simulated trajectories resembles the one from measured trajectories. (C) Fraction of immobile segments measured in simulations of immobile (blue) or persistent (red) molecules and in experimentally measured tracks (yellow) as a function of the moving-average window size and for different velocity thresholds. The red horizontal lines signify 5% and 95% probability thresholds, respectively. Error bars are from bootstrapping. For a window size of 4 steps and a velocity threshold of 8 nm/s the rate of wrong annotation is smaller than 1% both in simulations of purely persistent or immobile molecules. For pairs of w and vthr that lead to high accuracy of the determination of immobile and persistent segments the immobile fraction of the experimental data shows similar results. (D–E) MSD’s of single-track segments (gray lines) classified as (D) immobile or (E) persistent compared to the MSD of all respective segments (blue line). For simulated trajectories that can switch between the immobile and the persistent state (simulated with v = 12 nm/s, kip = 0.015 s−1, kpi = 0.021 s−1) we find a similar behavior of the MSD curves (red line). (F) Distribution of MSD’s of immobile and persistent segments for different numbers of steps N.
Figure 1—figure supplement 7. Number of bound PBP2 molecules increases with increasing PBP2 levels.
(A–B) Number of bound PBP2 molecules () (A) and active PBP2 molecules () (B) as a function of the number of PBP2 molecules per cell (NPBP2). b and p are the bound and persistent fractions of molecules, respectively. Since DIA, fluorescence, and Western Blot results only gave relative changes of PBP2 numbers, we used PRM values for TKL130 (for PAmCherry-PBP2) or for TU230(attLHC943) with 5 μM IPTG induction (for msfGFP-PBP2), respectively. (C) Density of tracks obtained by slow tracking for TKL130 and TKL130/pKC128, using same photo-activation and imaging conditions. Dots represent single fields of view (40 × 40 um). Despite variations between different fields of view, the fold-change of the median is of the same order as the relative change of bound molecules obtained in (A).
Figure 1—figure supplement 8. Comparison of AV127, msfGFP-PBP2 (TU230(attLHC943)), msfGFP-PBP2(L61R) (TU230(attLHC943)), and WT strains.
(A) Length (top left), width (bottom left), and growth curves (right) of the strains carrying msfGFP-PBP2 (AV127 or IPTG-inducible) and msfGFP-PBP2(L61R) (labeled 'Mut') for different induction levels in comparison to MG1655. Gray and blue bars show cell dimensions after 6 and 10 hr of growth, respectively (see also Figure 1—figure supplement 9). Doubling times are obtained from exponential fits (dashed lines) to three biological replicates (different colors). (B) PBP2 fold changes acquired from epi-fluorescence images, GFP-Western Blotting, and mass spectrometry measurements (DIA and PRM). The values are normalized by the value acquired from 5 μM IPTG induction except for PRM counts. PRM measurements combined with colony counting yield absolute numbers of proteins per cell. (C) Average diffusion constants and bound fractions. Gray bars show data after 6 hr of growth. Dots represent biological replicates.
Figure 1—figure supplement 9. Time-dependent effect of low msfGFP-PBP2 expression.
(A) Cell length, width, and GFP intensity as a function of time of TU230(attLHC943) cells for two different induction levels of 5 μM IPTG (black) and 25 μM IPTG (red). (B) Bound fractions and average diffusion constants. Red lines in (B) indicate the values measured for 25 μM induction during steady-state growth. Shaded areas and error bars show standard deviation between at least three technical replicates.
Figure 1—figure supplement 10. Low-frequency tracking of msfGFP-PBP2 and msfGFP-PBP2(L61R) cells under different induction levels.
(A) Persistent fractions for different expression levels of msfGFP-PBP2 and msfGFP-PBP2(L61R). (B) Left. Mean velocity as a function of minimal R2, which are obtained from a quadratic fit to single-track MSD’s of the form y = a + bx2. Right. Mean velocity of tracks, which satisfy R2 > 0.9. Dots represent biological replicates.
Figure 1—video 1. High-frequency imaging of PAmCherry-PBP2.
Download video file (4.6MB, mp4)
Left. Denoised bright-field image taken at the beginning of the video. Right. Raw images from high-frequency imaging (imaging interval 60 ms) of PAmCherry–PBP2 show diffusive and bound molecules. Blue circles and lines represent peaks and corresponding tracks considered for analysis.
Figure 1—video 2. Low-frequency imaging of PAmCherry-PBP2.
Download video file (2MB, mp4)
Left. Denoised bright field image Right. Raw images from low frequency imaging (imaging interval 3.5 s) of PAmCherry–PBP2 show immobile and persistently moving molecules. Blue circles and lines represent peaks and corresponding tracks considered for analysis.
Figure 1—video 3. High-frequency imaging of msfGFP-PBP2.
Download video file (16.5MB, mp4)
Left. Denoised bright field image taken at the beginning of the video. Right. Raw images from high frequency imaging (imaging interval 60 ms) of msfGFP–PBP2 (AV127) show diffusive and bound molecules. Blue circles and lines represent peaks and corresponding tracks considered for analysis.
Figure 1—video 4. Low-frequency imaging of msfGFP-PBP2.
Download video file (3.3MB, mp4)
Left. Denoised bright field image Right. Raw images from low frequency imaging (imaging interval 3.5 s) of msfGFP–PBP2 (AV127) show immobile and persistently moving molecules. Blue circles and lines represent peaks and corresponding tracks considered for analysis.