Kinetics of transcription and translation in single cells. Individual cells were induced with 10 ng/ml aTc at time t = 0. (A) Frag1B cells—wildtype for the efflux pumps genes acrAB. (B) Frag1A cells—acrAB deletion mutant. mRNA concentration (squares) and DsRed concentration (circles) were measured with FCS in single cells. Error bars associated with mRNA measurements were determined as in ref. (9). Horizontal bars at the bottom of the graphs indicate cell division events. Insets: distribution of translation rates per mRNA (monomers/mRNA/min) measured in different individual cells. In Frag1B cells, mean mRNA concentration: 3.7 ± 1.3 transcripts; mean rate of DsRed synthesis (not corrected for growth): 10.6 ± 4.7 monomers/min; mean translation rate per mRNA: 6.5 ± 3.4 monomers/mRNA/min. In Frag1A cells, mean mRNA concentration: 6.9 ± 2.7 transcripts; mean rate of DsRed synthesis (not corrected for growth): 15.5 ± 5 monomers/min; mean translation rate per mRNA: 5.9 ± 2.8 monomers/mRNA/min.

Genetic assay for mRNA/protein measurements in single living cells. (A) Transcriptional fusion of the RFP DsRed coding region fused to two ms2-binding sites. The dsRed-ms2x2 gene was placed under the control of an inducible Tet promoter, pLtetO-1. rbs, ribosomal binding site, taa, in frame stop codon; terminator, transcriptional terminator. (B) A molecule of RNA polymerase transcribes the dsRed-ms2x2 gene. Transcription of the ms2-binding sites occurs after the dsRed gene. The ribosome binds the nascent mRNA transcript (middle cartoon) and the MS2-GFP proteins binds the mRNA transcript once the ms2-binding sites are fully transcribed. This genetic design guarantees that our FCS detection method only measures mRNA transcripts that are free to diffuse and that are not in complex with the RNA polymerase and the DNA.

Experimental FCS setup. A 488 nm laser beam is expanded using a divergent and a convergent lens. The expanded beam feeds a Olympus X71 microscope and a dichroic mirror (DM1) reflects the laser blue light on the bacterium. The emitted green and red light from the fluorescent proteins expressed in the bacterium are transmitted through the dichroic mirror (DM1) and are reflected by a second dichroic mirror (DM2). A third dichroic mirror (DM3) splits the fluorescent signal into a green component and a red component. The light is focused with an achromatic convergent lens onto the cores of two optical fibers, which act as pinholes (ph) and feed two avalanche photodiodes (APD) that produce photon counting time series. An ALV correlator connected to a computer records the time series and computes in real-time the associated autocorrelation functions. A red light (lamp) illuminates the sample from above, and a CCD camera is connected to a monitor (more details in Materials and methods section).

(A) A [mRNA] (nM) as a function of inducer concentration. Each data point represents the average mRNA concentration measured across 30–60 individual Frag1B (circles and solid line) and Frag1A (squares and interrupted line) cells. The number of mRNA transcripts present in the detection volume is converted into a concentration (nM). The farthest point on the left of the graph represents [mRNA] measured in the absence of inducer (0 [aTc]). The inset represents a Hill fit of the normalized data with a Hill coefficient ∼4.5 ± 1.6 for Frag1B and 4.6 ± 1.5 for Frag1A. Error bars represent the combination of the standard error and the systematic error, the latter is determined from the measured [mRNA] in a cell with no pZE31-DsRed-ms2x2 plasmid. This value was 6 nM for Frag1B and 10 nM for Frag1A and was subtracted from all data points. (B) DsRed tetramer concentration (μM) as a function of inducer level. Each point represents the average protein concentrations measured across the same individual cells as in (A). Error bars represent the standard error. The number of DsRed tetramers in the detection volume is converted into concentration (μM) on the left side of the graph. The farthest point on the left of the graph is [DsRed] in the absence of inducer (0 aTc). The inset represents a Hill fit of the normalized data with Hill coefficient of 3 ± 0.2 for Frag1B and 3.1 ± 0.3 for Frag1A. Frag1B, circles and solid line; Frag1A, squares and interrupted line. (C) Protein and mRNA concentrations within the detection volume. Frag1B, circles; Frag1A, squares. Linear fit (forced through 0): dashed lines (Frag1B) with R factor of 0.94, continuous line (Frag1A) with R factor of 0.97.