Physical responses of bacterial chemoreceptors

Abstract

Chemoreceptors of the bacterium Escherichia coli are thought to form trimers of homodimers that undergo conformational changes upon ligand binding and thereby signal a cytoplasmic kinase. We monitored the physical responses of trimers in living cells lacking other chemotaxis proteins by fluorescently tagging receptors and measuring changes in fluorescence anisotropy. These changes were traced to changes in energy transfer between fluorophores on different dimers of a trimer: attractants move these fluorophores farther apart, and repellents move them closer together. These measurements allowed us to define the responses of bare receptor oligomers to ligand binding and compare them to the corresponding response in kinase activity. Receptor responses could be fit by a simple "two-state" model in which receptor dimers are in either active or inactive conformations, from which energy bias and dissociation constants could be estimated. Comparison with responses in kinase-activity indicated that higher-order interactions are dominant in receptor clusters.

Figure 1

Anisotropy responses to ligand binding. (a) Cartoon of trimer structure, drawn to scale, showing the labelling position. (b) Fluorescence anisotropy traces measured with a population of flhD cells expressing Tar-YFP (green symbols), Tsr-YFP (red symbols), or Tsr^I377P-YFP (blue symbols). Cells attached to a cover slip were exposed to flowing buffer to which serine (1 mM), aspartate (1 mM), or sucrose (100 mM) were added for the periods indicated by the black bars. The delay between the addition and removal of the chemical and the onset or termination of the response is mostly due to the time required for the reagents to reach or be washed away from the cells. (c) The raw fluorescence intensity (in photon counts, integrated over 3 s) in the parallel (open symbols) and perpendicular (closed symbols) channels recorded during the experiment with Tsr-YFP.

Figure 2

The effect of diluting the YFP label. (a) Mixing of different receptor types. Normalized response amplitude, Δr/r₀, obtained from flhD cells co-expressing Tsr-YFP (to serine; blue symbols) or Tar-YFP (to aspartate; red symbols) together with Tar or Tsr, respectively. Tsr-YFP or Tar-YFP were expressed at fixed level, while Tar or Tsr were expressed at different levels by varying the concentration of sodium salicylate, as indicated in the abscissa, corresponding to a fraction of labelled receptors ranging roughly from 0.7 to 0.05. Data were normalized by the response at 0 μM sodium salicylate. The dashed line is a guide to the eye. (b) Bleaching. Anisotropy level measured from flhD cells expressing Tar-YFP. Aspartate (300 μM) was added to the buffer during the intervals indicated by the black bars. Between the first and second trials (left arrow) the intensity of the excitation light was transiently increased by a factor of 16, bleaching some of the fluorophore. Between the second and third trials (right arrow) a new field-of-view was chosen containing cells not previously exposed to the excitation light.

Figure 3

Attractant and repellent responses of receptors in different modification states. (a) Anisotropy responses to addition and removal of aspartate (500 μM) or nickel (100 μM) measured with flhD cells expressing Tar(EEEE)-YFP (green symbols) or Tar(QQQQ)-YFP (red symbols). (b) The anisotropy level measured with various Tar mutants: reading along the arrow, EEEE (green), QEEE (orange), QEQQ (magenta), and QQQQ (red), in experiments in which increasing amounts of nickel (upper plot) or aspartate (lower plot) were added and removed. Data from a few repetitions of each experiment are shown. Anisotropy levels were normalized as follows: lower plot, data were normalized to the value in the presence of 30 μM aspartate; upper plot, data were multiplied by α/r₀, with r₀ the anisotropy value at 0 ligand and α the value read from the corresponding plot in the lower panel at 0 ligand. Lines are fits to the data using Eq. 1, with $K_d^{on}$ and $K_d^{off}$ for aspartate set to 0.2 and 30 μM and for nickel set to 5.5 and 1.4 μM, respectively, and with ΔE_EEEE=1.6 kT, ΔE_QEEE=0.3 kT, ΔE_QEQQ=−1.1 kT, and ΔE_QQQQ=−2.2 kT.

Figure 4

Comparison of receptor and kinase responses. Solid lines: anisotropy responses to varying concentrations of nickel (upper panel) and aspartate (lower panel) normalized to the saturated response, Δr_sat, and color coded as in Fig. 3b: EEEE receptors green, QEEE receptors orange, QQQQ receptors red. These lines are fits to Eq. 1. Dotted lines with open symbols: kinase activities (dose-response plots) for measurements made with cheR⁺cheB⁺ cells expressing only Tar receptors (blue symbols) or cheR cheB cells expressing only modified Tar receptors (using the color code of Fig. 3b).

Figure 5

A cartoon suggesting ways to bridge the gap between the behaviour of bare oligomers and oligomers coupled to the kinase. The kinase can promote direct physical coupling between trimers by supporting tight clustering (top), or promote indirect coupling by mediating conformational spread (bottom).