Subdiffusion of loci and cytoplasmic particles are different in compressed Escherichia coli cells

Abstract

The complex physical nature of the bacterial intracellular environment remains largely unknown, and has relevance for key biochemical and biological processes of the cell. Although recent work has addressed the role of non-equilibrium sources of activity and crowding, the consequences of mechanical perturbations are relatively less explored. Here we use a microfabricated valve system to track both fluorescently labeled chromosomal loci and cytoplasmic particles in Escherichia coli cells shortly after applying a compressive force, observing the response on time scales that are too sudden to allow for biochemical response from the cell. Cytoplasmic diffusion slows markedly on compression but the exponent governing the growth of the ensemble-averaged mean-squared displacement of cytoplasmic particles is unaffected. In contrast, the corresponding exponent for DNA loci changes significantly. These results suggest that DNA elasticity and nucleoid organization play a more important role in loci subdiffusion than cytoplasmic viscoelasticity under such short time scales.

Fig. 1

Experimental setup. a E. coli cells were loaded into the PDMS chip by a syringe. The control channel (blue) was pressurized and controlled by a manual pressure regulator with an airtight pressure tube to apply pressure to the PDMS valve. b Cartoon depicting the compression of E. coli cells using the microfabricated valve. PDMS device dimensions: ~20 mm (L) × 20 mm (W) × 10 mm (H). c Scheme of the loci analyzed in this work. The origin of replication is labeled by the OriC label. The chromosomal coordinates of the Ori2, Ter1, and NSL1 loci are 3,928,826, 1,056,444, and 3,739,123, respectively, at a distance of 4,943, 1,772,236, and 184,760 bp from the origin, respectively. d Example of tagged loci (Ori2) in E. coli cells under a compression of 15 psi. The bright dots in the image are the tagged loci. The outlines of the cell walls are distinguishable in the image. The scale bar corresponds to 10 μm

Fig. 2

Applying pressure affects the diffusivity of cytoplasmic particles and DNA loci differently. a Ensemble-averaged mean-squared displacement for cytoplasmic particles confined in slits (blue multiplication symbols) and under 10 psi compression (gold circles). Linear regression to these data in Fig. S2a yields α = 0.75 ± 0.02 for the slits and α = 0.72 ± 0.005 for 10 psi compression, where the uncertainties represent 95% confidence intervals for the linear regression. b Equivalent data for the GFP-parB labeled NLS1 locus in slow growth media. The corresponding exponents from the linear regression in Fig. S2b are α = 0.51 ± 0.003 for slits and α = 0.45 ± 0.002 for 10 psi compression, again using 95% confidence intervals

Fig. 3

Exponents characterizing the growth of the ensemble-averaged mean-squared displacement for cytoplasmic particles and DNA loci are different under compression. The individual panels provide the exponents α obtained for cytoplasmic particles (Fig. 2a) as well as three different GFP-parB tagged loci (NSL1, Ori2, and Ter1; see Fig. 1c for chromosomal coordinates) under slow growth conditions (M9 + Glu) and fast growth conditions (M9 + Glu + CAA) in slits (uncompressed) and under compressions of 10 psi, 15 psi, and 20 psi. The error bars indicate 95% confidence intervals for the linear regression to the ensemble-averaged MSD data in Fig. S2; most error bars are smaller than the symbol size. Table S2 tabulates the values of D_app and α for each growth conditions and the uncertainty using 95% confidence intervals

Fig. 4

The probability density functions (PDF) of displacement distributions of cytoplasmic particles over 1 s both with and without compression have long tails. Displacement distributions over δ = 1 s in compressed cells at 10 psi (gold circles) and in cells packed in slits (blue multiplication symbols), scaled by the standard deviation of the displacements (σ_δ), for a cytoplasmic particles and b NSL1 loci under slow growth conditions (M9 + Glu). The solid lines are a fit with Laplace distributions. The cells trapped by microfabricated valves or slits are randomly oriented. To expedite the analysis, we selected a fixed x-direction to be aligned with the camera sensor (i.e., not necessarily oriented along a pole of a cell) and the distribution of 1D displacement along that direction is plotted. The displacement distributions are rescaled by the standard deviation of the displacement size, σ_δ