Dynamics of Bacterial Signal Recognition Particle at a Single Molecule Level

Abstract

We have studied the localization and dynamics of bacterial Ffh, part of the SRP complex, its receptor FtsY, and of ribosomes in the Gamma-proteobacterium Shewanella putrefaciens. Using structured illumination microscopy, we show that ribosomes show a pronounced accumulation at the cell poles, whereas SRP and FtsY are distributed at distinct sites along the cell membrane, but they are not accumulated at the poles. Single molecule dynamics can be explained by assuming that all three proteins/complexes move as three distinguishable mobility fractions: a low mobility/static fraction may be engaged in translation, medium-fast diffusing fractions may be transition states, and high mobility populations likely represent freely diffusing molecules/complexes. Diffusion constants suggest that SRP and FtsY move together with slow-mobile ribosomes. Inhibition of transcription leads to loss of static molecules and reduction of medium-mobile fractions, in favor of freely diffusing subunits, while inhibition of translation appears to stall the medium mobile fractions. Depletion of FtsY leads to aggregation of Ffh, but not to loss of the medium mobile fraction, indicating that Ffh/SRP can bind to ribosomes independently from FtsY. Heat maps visualizing the three distinct diffusive populations show that while static molecules are mostly clustered at the cell membrane, diffusive molecules are localized throughout the cytosol. The medium fast populations show an intermediate pattern of preferential localization, suggesting that SRP/FtsY/ribosome transition states may form within the cytosol to finally find a translocon.

Keywords: Shewanella putrefaciens; protein membrane insertion; signal recognition particle; single molecule tracking; structured illumination imaging.

FIGURE 1

Model for the SRP cycle. Lower part from right to left: ribosome subunits binding to mRNA form the 70S particle, to which SRP binds upon detection of the signal sequence at the nascent chain (“signal peptide”). This complex binds to FtsY, either within the cytosol, as indicated by the leftwards arrow, or at the cell membrane. GTP hydrolysis triggers hand over of the ribosome nascent chain complex to the SecYEG translocon (upper part), upon which SRP and FtsY are released. Our data suggest that complex formation of SRP/RNC and FtsY can occur within the cytosol and also at the membrane.

FIGURE 2

Structured illumination microscopy imaging of SRP interaction partners in S. putrefaciens at mid-exponential phase. (A) Static fractions of Ffh are distinctly located in close proximity to the cell membrane. Z-stack projection shows localization densities of static Ffh particles. Higher particle densities and gradients are indicated by multi-color coding with gray (little signal) to red-shifted (high signal). (B) Static fractions of FtsY also show highest densities close to the cell membrane. (C) Different from Ffh and FtsY, L1 shows distinct subcellular localizations representing nucleoid occlusion with high localization density at pole regions and septum. Z-stack projection shows similar patterns throughout the cells. Proteins are fused to sfGFP and are expressed as sole source of the proteins, at physiological levels. Left panels bright field acquisition, second panels SIM acquisition, third panels color coded SIM gradient, right panels color coded Z-stack projections. Scale bars 1 μm.

FIGURE 3

Tracking of SRP, FtsY, and ribosomes in S. putrefaciens. (A) Mean squared displacement (MSD) analyses of Ffh-mVenus, FtsY-mVenus, and L1-mVenus (abbreviated “mV”), showing different overall diffusion constants, corresponding to each molecular weight. MSD assumes an overall diffusion regardless of subdiffusion. (B) Squared displacement (SQD) analyses shows measured data, suggesting three significantly distinct diffusive subpopulations (upper panel). Residual analysis of SQD confirms three population assumption (lower panel): colored curves represent differences between measured data and modeled data (using Brownian motion), which are represented by the “zero” line. Right panel shows a bubble blot, illustrating mean diffusion constants (y-axis) of fractions indicated as circles, with sizes corresponding to fraction size (also written above the circle). (C) Gaussian mixture model (GMM) analyses also identify three significantly distinct diffusive subpopulations at steady-state. Bubbles illustrate fraction sizes, with fractions indicated by the different colored lines. Dashed line indicates sum of the three GMM fits.

FIGURE 4

Gaussian mixture model (GMM) analyses of protein dynamics in response to rifampicin treatment, inhibiting transcription (tracking 30 or 60 min after addition of a sublethal concentration of Rif). Population sizes of mVenus fusions and time of imaging after addition of rifampicin are stated in the panels. Bubbles illustrate fraction sizes, fractions are indicated by the different colored lines, analogous to Figure 3C.

FIGURE 5

GMM analyses after addition of chloramphenicol, inhibiting translation by blocking peptidyl transferase activity. GMM analysis identifies three significantly distinct diffusive subpopulations indicated by the different colored lines, dashed line shows sum of the three fits. Bubbles illustrate fraction sizes.

FIGURE 6

GMM analyses after addition of puromycin, leading to premature translation termination and mRNA release. GMM analysis identifies three significantly distinct diffusive subpopulations indicated by the different colored lines, dashed line shows sum of the three fits. Bubbles illustrate fraction sizes.

FIGURE 7

Changes in Ffh dynamics in response to FtsY depletion. (A) MSD analysis reveals slightly reduced mobility after depletion of FtsY (“- arabinose”). (B) Minor changes in dwell times, determined using a 2 population decay curve, after FtsY depletion (Ffh becomes slightly more static), (C) GMM analyses show a small increase in static Ffh molecules after FtsY depletion.

FIGURE 8

Averaged 3D-SIM image stack projection and confinement maps of Ffh-sfGFP in cells expressing FtsY under control of the arabinose promoter, false-colored in cyan (A), or in cells after depletion of FtsY due to growth in the absence of arabinose, false-colored in magenta (B). Scale bar 1 μm. (C) Confinement map of Ffh-mVenus tracks in the presence of arabinose (FtsY), or (D) in the absence of arabinose/FtsY. Blue track non-confined motion, red tracks confined motion of at least 9 steps within a radius of 120 nm, green tracks transitions between confined and free motion, projected into a standardized cell of 3 × 1 μm size.

FIGURE 9

Heat maps of the preferred localization of the three distinct populations of molecules. Localization heat maps with the localization of the tracks according to their apparent diffusion D* from MSD plots, where MSD(△t) = 4⋅D^∗△t. Linear fit was applied to the first 4 points of the MSD curve. A minimum of Pearson R squared of 0.8 was required. Among conditions, all heat maps have the same number of tracks (in brackets below the protein name) with approximately the same amount of detections (noted as N in the X axis label). Static fraction = D between 0 and 0.05 μm²/s, medium mobile = D between 0.05 and 0.3 μm²/s, high mobile = D 0.3–10 μm²/s. Colors from blue to red indicate increases in occupancy of tracks. Cells have a size of 1 × 3 μm.