Long-term positioning and polar preference of chemoreceptor clusters in E. coli

Abstract

The bacterial chemosensory arrays are a notable model for studying the basic principles of receptor clustering and cellular organization. Here, we provide a new perspective regarding the long-term dynamics of these clusters in growing E. coli cells. We demonstrate that pre-existing lateral clusters tend to avoid translocation to pole regions and, therefore, continually shuttle between the cell poles for many generations while being static relative to the local cell-wall matrix. We also show that the polar preference of clusters results fundamentally from reduced clustering efficiency in the lateral region, rather than a developmental-like progression of clusters. Furthermore, polar preference is surprisingly robust to structural alterations designed to probe preference due to curvature sorting, perturbing the cell envelope physiology affects the cluster-size distribution, and the size-dependent mobility of receptor complexes differs between polar and lateral regions. Thus, distinct envelope physiology in the polar and lateral cell regions may contribute to polar preference.

Fig. 1

Basic characterization of the MG1655/cheA::mYFP (MK4) cells. a Schematic description of core complexes showing the position of the mYFP tag in the core complex. b Colony expansion of the MG1655 (CheA⁺) cells and the cheA::mYFP derivative in soft agar chemotaxis plates after 10 h at 30 °C. Bars are of the same size. c Fluorescence images of cheA::mYFP cells grown to an optical density (OD₆₀₀) of 0.08 or 0.4 in liquid culture. Scale bar corresponds to 2 µm. Histogram of the number of clusters per cell in the two populations (240 and 209 cells, respectively) and the respective polar bias of the clusters (number of polar clusters / total number of clusters) in each subpopulation of cells belonging to each bin of the histogram are also shown

Fig. 2

Typical long-term positional dynamics of clusters in the cheA::mYFP (MK4) cells. a In this example, cells also expressed FtsZ-mCherry (induced by 0.005% arabinose); however, for clarity, the mCherry overlay was added only at the 132 min time point. The central cluster tracked here is labeled throughout the time series by a thick white arrowhead. The new clusters that appeared near the central cluster in the cell pole at 24 and 168 min time points are also labeled. For clarity, few of the high intensity clusters in cells other than those in focus were masked. Fluorescence images were recorded every 4 min, and only sample images are shown. b Similar to a with a focus on a single division event. c Example of a cluster near the cell pole that apparently split into two with one cluster remaining at the pole and the other drifting away. Scale bars corresponds to 1 µm. d Histogram of polar-cluster nucleation times (t_nuc) measured relative to the period between the assembly (t_on) and disassembly (t_off) of the Z-ring during the corresponding cell division (127 clusters). e Using cells without FtsZ-mCherry, clusters within a certain colony at a certain time point were identified and grouped according to their nucleation and final positions, polar or lateral. The data are shown for cells grown in minimal medium containing 20% (dark gray) or 2% (light gray) TB (301 and 133 clusters, respectively). f The relative position of the clusters (α) was defined as the distance between the cluster and a certain cell pole normalized by the length of the cell and is plotted as a function of time (gray circles). Each plot represents the trajectory of a single cluster whose position was quantitatively evaluated once every cell cycle, soon after cell division. The estimated uncertainty in the measured cluster relative position is approximately the size of the symbols. The blue lines were plotted by iterating equation (1). In total, 22 clusters from nine independent experiments were followed, each for several generations (see also Supplementary Fig. 3)

Fig. 3

Initial cluster growth rate. a The main plot presents the growth of a polar cluster, as assessed by following the peak fluorescence intensity. The ‘time-zero’ point was defined by extrapolating from the apparent linear regime to zero intensity (black line). The cluster growth rate (β) was defined as the slope of the growth curve at the linear regime. The inset shows additional examples of polar and lateral clusters. b Histogram of the cluster growth rates (β) of polar (black) and lateral (gray) clusters (105 clusters)

Fig. 4

Positioning of modified receptor clusters. a Tar-mYFP expressed in addition to the native receptors in cells lacking CheA and CheW (UU1607). b Tar(head)-mYFP or Tsr(I377P)-mYFP expressed as the sole chemoreceptor in cells lacking CheA and CheW (UU2806). c cheA::mYFP cheW-X2 cells (MK9) grown under standard conditions (left) or slow-growth conditions (5% TB, right). d Image: cheA::mYFP cells (MK4) expressing additional CheW-X2 proteins (pAV305, induced by 1 µM NaSal). Bar graph: The fraction of lateral clusters in MK4 cells with or without the addition of CheW-X2 (dark gray; 0 or 1 µM NaSal) or MK4 cells with additional CheW expressed from an analogous plasmid (light gray; 1 µM NaSal). See also Supplementary Fig. 4. e A schematic summary (see text). Scale bar corresponds to 2 µm, throughout

Fig. 5

Positioning of modified receptors. A schematic description of the receptor variants is provided in the upper panel (see also Supplementary Fig. 5). a–d and f Fluorescence images of Δ(cheA cheW MCPs) (UU2806) cells expressing various receptor mutants from an inducible plasmid (pRR48, 100–200 µM IPTG) together with CheA::mYFP and CheW (pAV295,  0.3 µM NaSal). e Images of TorS-mYFP expressed in MG1655 cells (pES42, induced with 10 µM IPTG). g The fraction of polar clusters found in cells expressing the various receptor variants (a total of 208, 148, 88, 206, 169, and 191 clusters analyzed in parts a–f, respectively). Notably, with the exception of part d, the overall polar bias of these receptor variants is significantly larger since the clusters were generically larger in the pole regions than in the lateral region. Additional images are presented in Supplementary Fig. 6. Scale bar corresponds to 2 µm, throughout

Fig. 6

The effect of altered membrane environments on receptor clusters. a Effect of TolA. Distribution of cluster intensity in cheA::mYFP cells with (MK4; dark gray) or without (MK13; light gray) TolA. A total of 472 clusters (in 264 cells) or 475 clusters (in 82 cells) were sampled from each strain, respectively. Fluorescence images of typical TolA⁻ cell is also shown. The fluorescence image on the right demonstrate polar bias in TolA⁻ cells in which CheW-X2 was also expressed from a plasmid using 1 µM NaSal (see also Supplementary Fig. 7). b Left: Δ(cheA cheW MCPs) (UU2806) cells expressing Tsr(I377P)-mYFP receptors and imaged at various times after exposure to Cm (20 µg/ml). Right: UU2806 cells expressing H-NS-mYFP (pAV357, induced by 2 µM IPTG) and imaged at various times after exposure to Cm (20 µg/ml). Corresponding intensity profiles measured along the long axis of the cell (between the marks) are also shown. See also Supplementary Fig. 8. Scale bars correspond to 2 µm. c Mobility of receptor complexes in the polar (dark gray) and lateral (light gray) cell regions measured using LPA-SPT (with frame repetition time of 320 ms). The mean (±s.e.m.) mobility (apparent diffusivity, D) of clusters in each track-length bin (bin size = 15 frames) is shown. Track length serves as a proxy for cluster size (see text)

Fig. 7

Positioning of receptor clusters. a The long-term positional dynamics of clusters. Clusters that nucleate in pole regions remain polar. Clusters that nucleate in the lateral region remain static relative to their local cell-wall matrix (green arrowhead) for many generations, at least in the direction of the long axis of the cell, and thus effectively shuttle between the cell poles. b Mid-cell lateral clusters avoid becoming polar during cell division by remaining static relative to their original cell-wall environment and avoid translocation to the new pole region (blue). c To the extent that the cell envelope physiology affects the local environment of the receptors, it might be expected that variations are created between the pole and lateral regions. Such variations can potentially affect the basic ‘diffusion and capture’ clustering dynamics by affecting the local mobility and clustering efficiency of receptor complexes and, thus, possibly promote polar preference