Nature | Vol 578 | 13 February 2020 | 297the course of 3 h. When the cells were treated with the proteasome inhibitor
MG-132 or b-AP15, the number and size of the foci increased, and clearance
was significantly delayed. By contrast, pre-treatment with the ubiqui-
tin E1 inhibitor MLN-7243 almost completely inhibited foci formation
(Fig. 1d, bottom, Extended Data Fig. 3d). Thus, ubiquitylated substrates
are required for the formation of proteasome foci and proteasome activity
is necessary for their clearance. Together, these observations suggested
that hyperosmotic-stress-induced proteasome foci are sites of proteolysis.
Orphan RP is degraded in proteasome foci
A slight increase in the level of ubiquitylated proteins was observed
following hyperosmotic stimulation, and a mass spectrometry-based
ubiquitylome analysis identified changes in several housekeeping pro-
teins, including linker histones, HSP90 and ribosomal proteins (RPs)
(Extended Data Fig. 4a, b). We subsequently focused on RPs. Ribosomes
are constitutively produced at around 7,500 molecules per minute
in cultured human cells via complicated assembly processes within
the nucleus, and orphan RPs that fail to incorporate into ribosomes
are degraded by the ubiquitin–proteasome system^15 –^19. Therefore,
we hypothesized that ribosome biosynthesis might be vulnerable to
hyperosmotic stress. Indeed, electron microscopy analysis showed
the disappearance of the nucleolar dense fibrillar compartment (DFC),
where pre-ribosomal RNA (rRNA) and RPs are assembled, upon hyper-
tonic stress (Fig. 2a). We also observed the emergence of multiple hyper-
dense structures near proteasome foci; these structures disappearedPSMB2–FusionRed
0.2 M sucroseabPSMB2–eGFP0s 20 s 180 s0.2 M sucroseHCT116PSMA1DAPIPSMA1 (α6)0.2 M sucrose
30 minProteasome probeProbePSMB2DAPI30 minPSMB2 K48Ub PSMB2K48Ub DAPIPSMB2–eGFP
0.2 M sucrose, 30 minMLN-7243Control0min 1min 5min 30 min90min 180 minMG-132b-AP150.2 M sucrose012303060Time(h)Averagefoci number per cell012300.51.0Time(h)Median
foci diameter (μm)
Proteasome inhibitorE1 inhibitorNDP = 0.0001
P = 0.0001
P = 0.0043P = 0.0009P = 0.0028dMLN-7243MG-132
b-AP15ControlCryo-ET c024600.51.0Distance(μm)PSMB2 K48UbNormalized intensityFig. 1 | Hyperosmotic stress induces the formation of proteasome foci in the
nucleus. a, HCT116 cells, PSMB2–eGFP or PSMB2–FusionRed cells were
stimulated with 0.2 M sucrose. Endogenous proteasome activity was detected
using a proteasome probe (Me4BodipyFL-Ahx3Leu3VS) and localization of
endogenous proteasomes was determined using PSMA1 (α6) antibody at the
indicated times. Scale bars, 10 μm. b, Cryo-electron tomography (cryo-ET)
image of proteasome foci in the nuclear region following stimulation with
0.2 M sucrose. Scale bar, 0.1 μm. c, Left, PSMB2–eGFP cells were stimulated
with 0.2 M sucrose for 30 min and endogenous K48Ub was detected with K48-
ubiquitin antibody. Right, line profiling of a representative section of the cell,
indicated by a white dashed line. Scale bars, 10 μm. d, Left, time-lapse images of
live PSMB2–eGFP cells. Cells were treated with proteasome inhibitor MG-132
(50 μM, 1 h prior), b-AP15 (1 μM, simultaneously) or ubiquitin-activating
enzyme (E1) inhibitor MLN-7243 (1 μM, 1 h prior), and then stimulated with 0.2 M
sucrose. Scale bars, 10 μm. Representative images from two independent
experiments. Right, changes of the proteasome foci number per cell and foci
diameter in untreated control cells (grey) and MG-132- (magenta), b-AP15-
(green) or MLN-7243- (orange) treated cells. n represents cell numbers
(control: n = 119, 121, 116, 117, 120, 120, 123, 126, 128, 125, 125; MG132: n = 94, 99,
99, 95, 95, 91, 93, 92, 94, 100 and 95; b-AP15: n = 81, 85, 80, 80, 85, 88, 90, 93, 103,
93 and 102; MLN-7243: n = 47, 43, 43, 42, 42, 42, 42, 40, 39, 39 and 39, at 0, 1, 3, 5,
10, 30, 60, 90, 120, 150 and 180 min after sucrose treatment, respectively). Data
are mean ± s.e.m.; one-way ANOVA with Dunnett's test (number) and
median ± s.e.m., Friedman with Dunn’s test (diameter). P values are shown in the
figure. Representative images from four (a, left), three (a, middle and right) or
two (b, d) independent experiments.