Figure 1.
Aha incorporation. HeLa cells were starved of Met (15 and 30 min for 2-min and 5-s pulses, respectively), pulsed ±2 mM Aha, and chased (0–5 min; 0.2 mM Met without Aha).
After fixation and ‘clicking’ on Alexa555, DNA was counterstained with DAPI, images collected using a wide-field microscope, and fluorescence intensities (± SD) in the cytoplasm (cyto) and nucleus (nuc) normalized relative to values in the untreated cytoplasm. *: P<0.0001 (Student's two-tailed t test, n = 20 cells). Bars: 10 µm. (A) 2-min Aha pulse. (i,ii) Aha labels both nucleus and cytoplasm, with the nucleus being the brightest. (iii) Pretreatment with anisomycin (aniso; 100 µg/ml; 2 h) reduces signals in both nucleus and cytoplasm. (iv) Slightly more signal is found in the larger area of the cytoplasm. (v) Signals in both compartments disappear during a chase. (B) After a 2-min pulse, signals in both compartments disappear quickly during a chase. Best fits of simple exponential curves to the data are included, but note that we do not know how many kinetic populations there might be (Materials and Methods S1). (C) MG132 (100 µg/ml; 2 h) increases signal given by a 2-min pulse. (D) 5-s Aha pulse. (i-iii) Signal is sensitive to anisomycin (aniso; 100 µg/ml; 15 min).
Figure 2.
Puromycin incorporation. HeLa cells were pre-treated with cycloheximide (100 µg/ml; 15 min) to slow ribosomes, pulsed with puromycin (puro; 91 µM; 0–60 s), fixed, puromycylated peptides immuno-labeled with Cy3, DNA stained with DAPI, and images collected.
(A-D) Typical confocal sections through the centre of nuclei. A 5-s pulse gives bright nuclear foci. After 30 s, cytoplasmic and nuclear signals are more similar and diffuse. After 60 s, the peri-nuclear region is the brightest. Bar: 10 µm. (E) Using wide-field images, Cy3 intensities (±SD; n = 20 cells) are expressed relative to the cytoplasmic value after a 60-s pulse.
Figure 3.
Incorporation of heavy amino acids detected using NanoSIMS. HeLa were starved of relevant amino acid(s), grown (10, 120 s) in heavy amino acids, fixed, sectioned, images collected, and distributions of light (12C14N−), heavy (12C15N−), or heavy-heavy ions (13C15N−) measured.
In some cases, inhibitors were present during starvation and the pulse, or chases followed the pulse. (A) An example after 30-min starvation of Lys + Arg, and 2-min growth in 8 mM [15N]Lys +4 mM [15N]Arg. (i,ii) Images show distributions of light and heavy ions in one section (regions of interest in the nucleus, N, and cytoplasm, C, indicated). Bar: 10 µm. (iii) Signals due to heavy ions are expressed as ratios (± SD) relative to total numbers of ions (i.e., 12C15N−/[12C15N−+12C14N−]) normalized to the natural abundance of 15N. *: P<0.0001 (Student's two-tailed t test; n = 19–26 cells). Row 1: cells unexposed to the heavy labels give a ratio characteristic of the natural abundance. Rows 2–3: a 10-s pulse in heavy medium increases ratios, and anisomycin (aniso; 100 µg/ml; 30 min) reduces this. Rows 4–6: a 120-s pulse yields higher ratios, which are reduced by a chase at 37°C (but not 4°C). (B) After 30-min starvation in the absence of either Met (or Lys + Arg), and a 2-min pulse in 2 mM [15N]Met (or 8 mM [15N]Lys +4 mM [15N]Arg), signals in both compartments decline quickly during a chase. Best fits of simple exponential curves to the data are included, but note that we do not know how many kinetic populations there might be (Materials and Methods S1). (i) With [15N]Met (which is incorporated into the N-terminus and internally), essentially all label disappears by 5 min. (ii) After labeling with [15N]Lys + [15N]Arg (which are never incorporated into the N-terminus), the decline is slower (which is consistent with these internal labels being preferentially incorporated into a minor, longer-lived, fraction). For comparison, curves (dotted lines) are reproduced from panel (i). (C) Signal is due to the formation of peptide bonds. Cells were starved (15 min) of all amino acids, and grown (2 min) in [13C]amino acids + [15N]amino acids ±100 µg/ml anisomycin or 100 µg/ml cycloheximide. (i) Some peptide bonds are then formed by a ribosome covalently linking a carboxyl 13C atom in one [13C]amino acid to an amino 15N atom in a [15N]amino acid; the NanoSIMS bombardment generates (directly or indirectly) 13C15N−. R1 and R2: different residues. (ii) After collecting images, distributions of 13C15N− and 12C14N− ions in regions of interest in the nucleus (nuc) and cytoplasm (cyto) were measured; signals due to (heavy-heavy) 13C15N− are expressed as ratios (± SD) relative to those due to (light-light) 12C14N−. *: P<0.0001 (Student's two-tailed t test; n = 11–18 cells). Row 1: cells unexposed to heavy labels give ratios characteristic of the natural abundance. Row 2: growth in both heavy labels increases ratios. Rows 3,4: 100 µg/ml anisomycin (aniso) or cycloheximide (chx) reduces ratios. Rows 5,6: a chase at 37°C also reduces ratios, whilst one at 4°C gives less reduction.
Figure 4.
Effects of PTCs on nascent Cd2 RNA. Cos-7 cells (which encode the SV40 T antigen) were co-transfected with constructs encoding the ‘Tet’ repressor and a test vector (with the ‘Tet’ promoter driving Cd2 ± a PTC).
By 24(which also encodes the SV40 ori) replicates to give ∼8,000 mini-chromosomes/cell; the ‘Tet’ promoter is silent and no Cd2 RNA is detected. Now doxycycline (10 µM; 45 min) is added, cells fixed 26 h after transfection, intronic Cd2 RNA detected by RNA FISH, DNA stained with DAPI, and images collected using a wide-field microscope. (A,B) Two views of one field after co-transfecting the PTC– vector (only the cell at the top was transfected and expresses intronic Cd2 RNA in nuclear foci). Bar: 10 µm. (C) After subtracting background, intensities (± SD) seen in nuclei are expressed relative to the value found in untreated cells transfected with the PTC¯ vector. *: P<0.0004 (Student's two-tailed t test, n = 20 cells). (i) A PTC reduces levels of intronic RNA, but cycloheximide (chx; 100 µg/ml; 2 h) more than reverses the effect. (ii) After transfecting the PTC¯ vector, SSA (100 ng/ml; 45 min) reduces levels of intronic RNA, and anisomycin (aniso; 100 µg/ml; 2 h) reverses this effect.
Figure 5.
A model illustrating how ‘dark-matter’ peptides (green lines i-iii) and the ‘mature’ proteome (iv) arise.
Most initiated ribosomes terminate prematurely (giving i and iii), and some translate to the end of an uORF (giving ii); the resulting peptides are rapidly degraded (half-life <1 min), to give rise to the astonishing turnover seen using short pulses. A minority of ribosomes translate the whole ORF (giving iv); such peptides are the ones detected conventionally using long pulses (they are generally stable and constitute the ‘mature’ proteome). During long pulses, most peptides i-iii are degraded and so are not detected.