Nuclear and cytoplasmic spatial protein quality control is coordinated by nuclear-vacuolar junctions and perinuclear ESCRT

Abstract

Effective protein quality control (PQC), essential for cellular health, relies on spatial sequestration of misfolded proteins into defined inclusions. Here we reveal the coordination of nuclear and cytoplasmic spatial PQC. Cytoplasmic misfolded proteins concentrate in a cytoplasmic juxtanuclear quality control compartment, while nuclear misfolded proteins sequester into an intranuclear quality control compartment (INQ). Particle tracking reveals that INQ and the juxtanuclear quality control compartment converge to face each other across the nuclear envelope at a site proximal to the nuclear-vacuolar junction marked by perinuclear ESCRT-II/III protein Chm7. Strikingly, convergence at nuclear-vacuolar junction contacts facilitates VPS4-dependent vacuolar clearance of misfolded cytoplasmic and nuclear proteins, the latter entailing extrusion of nuclear INQ into the vacuole. Finding that nuclear-vacuolar contact sites are cellular hubs of spatial PQC to facilitate vacuolar clearance of nuclear and cytoplasmic inclusions highlights the role of cellular architecture in proteostasis maintenance.

Extended Data Fig. 1 |. Spatial sequestration occurs during different types of stress with different client proteins.

(a) Western blot analyses of Gal Shut-off assays showing the clearance of NLS-LuciTs (top) and NES-LuciTs (bottom) with and without proteasome impairment by 50μM Bortezomib. Blot is representative of 3 biologically independent experiments. (b-c) Representative Structured Illumination super-resolution microscopy images taken of cells expressing NLS-VHL (b) or NES-VHL (c) after 120 minutes at 37 °C and treated with 100μM MG132. NLS-LuciTs is shown in green, NES-LuciTs in purple, nuclear pores in gold and Hoechst counterstain in blue. Scale bars are 1μm. (d) Drop test of W303 yeast expressing model proteins without heat shock at 30C (left), with heat shock at 37 °C (middle), and without expression of the plasmids (right). Unprocessed blots are available in source data.

Extended Data Fig. 2 |. The effect of blocking nucleocytoplasmic transport on Ubc9Ts clearance.

(a) Quantitation of the percentage of cells containing nuclear or cytoplasmic inclusions in WT yeast expressing Ubc9Ts-EGFP after 120 minutes at 37 °C with and without treatment with 100μM MG132. A minimum of 500 cells per condition from 3 biologically independent experiments were counted and two-tailed Student’s t-tests were performed comparing the WT yeast without MG132 treatment to WT yeast with MG132 treatment using Prism software. P values were adjusted using two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli with a Q of 5%. Adjusted P value for nuclear no MG132 vs. +MG132 is 0.0035 and cytoplasmic no MG132 vs. +MG132 is 0.0011. Data are shown as mean values ± S.E.M. (b) Representative Structured Illumination super-resolution microscopy images taken of cells expressing EGFP-VHL after 2 hr at 37 °C with DMSO (left) or with 100μM MG132 (right) treatment. VHL is shown in green, nuclear pores in gold, and Hoechst counterstain in blue. Scale bars are 1μm. Numerical source data are available in source data.

Extended Data Fig. 3 |. INQ-JUNQ homing does not occur at the LINC, nucleolus, or involve FG repeats of the nuclear pore central channel.

(a) Graph of the X-Y positions of the INQ and JUNQ compartments by particle tracking of inclusions from cell shown in Figure 2a over the time course of the experiment. (b) Representative confocal image taken of cells co-expressing NLS-EGFP-VHL and NES-DsRed-VHL after 2 hr at 37 °C and treated with 100μM MG132. NLS-fusion proteins are shown in green, NES-fusion proteins in purple, nuclear pores in gold, and Hoechst counterstain in blue. Scale bar is 1μm. (c) Representative confocal fluorescence microscopy images taken of cells co-expressing NLS-EGFP-LuciTs and NES-DsRed-LuciTs (left) after 2 hr at 37 °C and treated with 100μM MG132. NLS-LuciTs is shown in green, NES-LuciTs in purple, nucleolus (Nsr1) in gold and Hoechst counterstain in blue. (right) Line intensity profile showing distance between nucleolus and homed INQ/JUNQ. Scale bars are 1μm. (d) (left) schematic of Mps3 component of LINC complex linking inner and outer nuclear membranes. (right) Representative widefield fluorescence microscopy images taken of cells co-expressing endogenously tagged Mps3-EGFP and NES-DsRed-LuciTs after 120 minutes at 37 °C with and without treatment with 100μM MG132. White arrowheads indicate locations of Mps3 puncta while yellow arrowheads indicate NES-LuciTs puncta. Scale bars are 1μm. (e) WT (top) and nupΔFG (bottom) cells co-expressing NLS-LuciTs and NES-LuciTs were shifted to 37 °C and monitored by live cell time-lapse fluorescence microscopy for the times shown. Scale bars are 1μm.

Extended Data Fig. 4 |. Detailed representation of the cryo-SXT workflow and interactions between mitochondria and cytoplasmic PQC compartments.

(a) Optical path through the specimen. Key: COL, cryogenic objective lens; SS, specimen stage; SP, specimen port; MG, motorized goniometer; CIM, cryogenic immersion fluid; CCL, low magnification cryogenic objective; CS, cryogenic specimen; CIE, cryogenic imaging environment; AP, adapter port; AW, a heated, angled anti-reflection window. (b) Alignment of fluorescence and soft x-ray tomographic data using fiducial markers. (c) A representative confocal image of the spatial relationship between the INQ and nucleolus. NLS-LuciTs (INQ) is shown in green, nucleolus in gold, and Hoechst counterstain in blue. Scale bar is 1μm. (d) The interaction between mitochondria and cytoplasmic inclusions is also seen by fluorescence confocal microscopy in a representative image of a cell co-expressing mito-GFP and NES-RFP-LuciTs. NES-LuciTs is shown in purple, mitochondria in cyan, and Hoechst counterstain in blue. Scale bar is 1μm. (e) Representative confocal fluorescence microscopy images taken of WT, fission mutants (dnm1Δ and fis1Δ) and fusion mutant (fzo1Δ and ugo1Δ) cells expressing mito-GFP and NES-DsRed-LuciTs after 120 minutes at 37 °C and treated with 100μM MG132. Mito-GFP is shown in cyan, NES-LuciTs in purple, and Hoechst counterstain in blue. Scale bars are 1μm.

Extended Data Fig. 5 |. NVJ-mediated clearance of misfolded proteins.

(a) Endogenously tagged Nvj1-GFP yeast expressing Ubc9Ts-ChFP were shifted to 37 °C and monitored by live cell time-lapse fluorescence microscopy for the times shown. White arrowheads indicate locations of Nvj1 puncta while yellow arrowheads indicate Ubc9Ts-ChFP puncta. Scale bar is 1μm. (b) WT (top) and nvj1Δ (bottom) cells co-expressing NLS-LuciTs and NES-LuciTs were treated with 100μM MG132 and shifted to 37 °C for 30 mins to preform inclusions. Cells were then placed in media containing 50mg/ml cycloheximide (CHX) and 100μM MG132 at 37 °C and monitored by live cell time-lapse fluorescence microscopy for the times shown. Scale bars are 1μm. (c,d) Quantitation of the percentage of cells containing cytoplasmic inclusions in WT, nvj1Δ, and vac8Δ yeast co-expressing NLS-EGFP-LuciTs (c) and NES-DsRed-LuciTs (d) after 2 hr at 37 °C with and without treatment with 100μM MG132. Data are presented as mean values +/− SEM. Numerical source data are available in source data.

Extended Data Fig. 6 |. ESCRT involvement in the clearance of misfolded proteins.

(a) Representative confocal images of WT yeast co-expressing Chm7-EGFP and either NLS-EGFP-LuciTs (left) or NES-DsRed-LuciTs (right) after 120 minutes at 37 °C and treated with 100μM MG132. Chm7 is shown in teal and remains diffuse throughout the cell, NLS-EGFP-LuciTs in green, NES-DsRed-LuciTs in purple, nuclear pores in gold and Hoechst counterstain in blue. Scale bar is 1μm. (b) Representative confocal images of WT and vps23Δ, vps34Δ, and vps15Δ yeast co-expressing NLS-EGFP-LuciTs and NES-DsRed-LuciTs after 2 hr at 37 °C and treated with 100μM MG132. NLS-EGFP-LuciTs is shown in green, NES-DsRed-LuciTs in purple, nuclear pores in gold, and Hoechst counterstain in blue. Insets show the budding INQ encapsulated by nuclear pores. Scale bars are 1μm. Same data as shown in Fig. 6c, but with the green channel separated to clearly detail the colocalization with the cytoplasmic protein.

Extended Data Fig. 7 |. Vacuole-mediated clearance of INQ and JUNQ.

Representative images of WT cells expressing NES-2xKeima-LuciTs after 2 hr incubation at 37 °C with 100μM MG132. Over time, fluorescence is seen with excitation in the 558 nm channel indicating the NLS-LuciTs has encountered an acidic environment. Insets show the transition from green to red and a structure leaving the inclusion that is fully red. Scale bars on large images are 5 mm. Scale bars on magnifications are 1 μm. Same data shown in Fig. 7d, but with more time points and a larger field of view in the images. Scale bars on large images are 5 μm. Scale bars on magnifications are 1 μm. (b) WT cells expressing NES-2xKeima-LuciTs after 85 min incubation at 37 °C with 100μM MG132. (c) Longer exposure of the blot shown in Fig. 7f to highlight the difference in the number and pattern of the EGFP bands in the WT vs pep4Δ cells. (d) Levels of EGFP at time 0 were measured from Quantitative Western blots such as those shown in Fig. 7e, f (mean ± S.E.M. from three biologically independent experiments). WT and pep4Δ yeast were compared using a two-tailed paired Student’s t-test without reaching statistical significance. (e) WT yeast expressing NLS-EGFP-LuciTs were treated with 8μM of FM4–64 and incubated for 2hr at 37 °C with 100μM MG132. Cells were imaged every 30 sec for 90 mins. Scale bar is 1μm. Same data shown in Fig. 7h, but only WT and with more timepoints during the entry into the vacuole. Source numerical data and unprocessed blots are available in source data.

Fig. 1 |. The INQ and JUNQ are separate nuclear and cytoplasmic PQC compartments.

a, Schematic showing nuclear and cytoplasmic misfolded proteins could be sequestered into separate compartments or the cytoplasmic misfolded proteins could be imported into the nucleus and sequestered into the nuclear PQC compartment INQ. b, Experimental schematic. Temperature-sensitive Luciferase is properly folded at 25 °C. Heat shock at 37 °C leads to unfolding and sequestration of the protein. Degradation of the unfolded and sequestered proteins can be blocked by treatment with the proteasome inhibitor MG132. c,d, Live-cell time-lapse fluorescence microscopy of WT cells expressing NLS-LuciTs (c) or NES-LuciTs (d) at 37 °C, treated with DMSO (top) or 100 μM MG132 (bottom). Representative still frames at the times shown. Scale bars, 1 μm. e,f, Representative SIM images of WT cells expressing NLS-LuciTs (e) or NLS-VHL (f) after 2 h at 37 °C (LuciTS) and 30 °C (VHL) and treated with 100 μM MG132. NLS-LuciTs and NLS-VHL are shown in green, nuclear pores in gold and Hoechst counterstain in blue. Scale bars, 1 μm. Line intensity profiles indicate relative locations of subcellular compartments to Nups and DNA. g,h, Representative SIM images of WT cells expressing NES-LuciTs (g) and NES-VHL (h) after 2 h at 37 °C (LuciTS) and 30 °C (VHL) and treated with 100 μM MG132. NES-LuciTs and NES-VHL are shown in purple, nuclear pores in gold and Hoechst counterstain in blue. Scale bars, 1 μm. Line intensity profiles indicate relative locations of subcellular compartments to Nups and DNA.

Fig. 2 |. Nuclear entry of misfolded proteins is not required for clearance.

a,b, Representative SIM images of WT cells expressing Ubc9Ts-EGFP after 2 h at 37 °C treated with DMSO (a) or 100 μM MG132 (b). Ubc9Ts-EGFP is shown in green, nuclear pores in gold and Hoechst counterstain in blue. Scale bars, 1 μm. Line intensity profiles indicate relative locations of subcellular compartments to Nups and DNA. c, Representative confocal fluorescence microscopy images of cells expressing sGFP with DMSO (−MG132), 100 μM MG132 treatment, co-expression of mHTT97QΔP or 200 nM LMB treatment. Scale bar, 1 μm. d, Ratio of cytoplasmic:nuclear fluorescence. Kruskal–Wallis test with Dunn’s multiple comparisons test was performed using Prism. Adjusted P value of no treatment versus MG132 is 0.0014, no treatment versus 97QΔP is 0.0027 and no treatment versus LMB is <0.0001. Twenty cells per condition from five biologically independent experiments were normalized to the no treatment control, analysed and presented as violin plots with median values shown as dashed lines and quartile values shown as dotted lines. e, Left: schematic illustrating the clearance of Ubc9Ts in WT yeast. Right: timeline of treatments for clearance measurements with shift to 37 °C 60 min before initiation of the measurements. f, Left: schematic illustrating the nup116–5 yeast have sealed nuclear pores at 37 °C, thus blocking nucleocytoplasmic trafficking. Densitometric quantification (middle) of western blot bands (right) measuring the amount of Ubc9Ts-EGFP remaining in shut-off experiment of WT versus nup116-5 cells relative to t = 0 (mean ± s.e.m. from three biologically independent experiments) fitted with a one-phase decay non-linear fit regression line. g, Left: schematic illustrating the sts1–2 yeast do not translocate proteasomes to the nucleus at 37 °C. Densitometric quantification (middle) of western blot bands (right) measuring the amount of Ubc9Ts-EGFP remaining in shut-off experiment of WT versus sts1–2 cells relative to t = 0 (mean ± s.e.m. from three biologically independent experiments) fitted with a one-phase decay non-linear fit regression line. Source numerical data and unprocessed blots are available in source data.

Fig. 3 |. INQ and JUNQ home to similar location on each side of the nuclear envelope via a cytoplasmic signal linked to nuclear pores.

a, WT cells co-expressing NLS-EGFP-LuciTs and NES-DsRed-LuciTs were shifted to 37 °C, treated with 100 μM MG132 and monitored by live-cell time-lapse fluorescence. Representative still frames at the times shown. Scale bar, 1 μm. b, Graph of the distance between the INQ and JUNQ compartments by particle tracking of inclusions from cell shown in a over the time course of the experiment. The slight variations of tethered distance between 60 min and 90 min is possibly due to the relative migration of one of the inclusions around a subcellular structure. c, Representative SIM image taken of cells co-expressing NLS-EGFP-LuciTs and NES-DsRed-LuciTs after 2 h at 37 °C and treated with 100 μM MG132. NLS-fusion proteins are shown in green, NES-fusion proteins in purple, nuclear pores in gold and Hoechst counterstain in blue. Scale bar, 1 μm. Line intensity profiles indicate relative locations of subcellular compartments and Nups. d, Representative confocal microscopy images taken of WT and three separate nup120Δ yeast cells expressing NES-LuciTs (i), NLS-LuciTs (ii) or co-expressing NLS- and NES-LuciTs (iii) after 2 h at 37 °C. NLS-LuciTs is shown in green, NES-LuciTs in purple, nuclear pores in gold and Hoechst counterstain in blue. Scale bar, 1 μm. Schematics on the right summarize the findings of the data. e, Schematic summarizing the data that the JUNQ localizes to the nuclear pores. A signal is transmitted through or near the nuclear pores to recruit the INQ to the same location, resulting in the homing of the two compartments. Source numerical data are available in source data.

Fig. 4 |. INQ resides near the nucleolus, JUNQ is surrounded by mitochondria and both compartments home into the NVJ.

a, Unfixed, intact yeast cells are frozen in a capillary tube and imaged by fluorescence microscopy and then X-ray tomography. Scale bar, 5 μm. The LACs are measured from the tomography data and used to identify the subcellular compartments and organelles. The fluorescence images are correlated to the X-ray tomography data, and 3D reconstructions of the subcellular components are generated. b, WT cells co-expressing Hsp104-GFP from the endogenous locus and HTT97QP-ChFP after 30 min at 37 °C. Left: X-ray tomograms without and with fluorescence overlay to indicate sites of Hsp104-GFP positive inclusions and HTT97QP-ChFP inclusion. Right: table of LAC values from the X-ray tomograms used to annotate the images and generate a 3D reconstruction shown in c. c, 3D reconstruction of fluorescence-correlated X-ray tomograms showing the JUNQ residing 450 nm outside the barrier of the nucleus (inset, top) and the IPOD marked by mHTT (inset, bottom). d, 3D reconstruction of X-ray tomograms from cell expressing NES-LuciTs after 90 min heat shock at 37 °C and treated with 100 μM MG132. The Q-bodies coalesce into the JUNQ compartment (inset, right). The Q-bodies also interact with vacuoles and the JUNQ compartment is surrounded by a mitochondrial cage (inset, left). e, Q-bodies interact with mitochondria in a separate cell expressing NES-LuciTs. f, Same 3D reconstruction of X-ray tomograms from cell shown in d. Insets are from the same cell rotated 180°. Blue arrows indicate sites where Q-bodies are directly interacting with the vacuoles. Red arrows indicate sites of NVJs. g, 3D reconstruction of X-ray tomograms from cell expressing NLS-LuciTs after 90 min heat shock at 37 °C and treated with 100 μM MG132. The INQ resides 400 nm from the nucleolus (inset). h, Schematic illustrating the findings of the X-ray tomography. The inset highlights two of the component proteins of the NVJ, Nvj1 and Vac8. Scale bars on X-ray 3D reconstructions are 1 μm; insets are 0.5 μm. JUNQ is shown in green, IPOD in orange, Q-bodies in purple, nucleus in yellow, nucleolus in gold, mitochondria in cyan, vacuoles in grey and lipid droplets in white.

Fig. 5 |. JUNQ and INQ converge at the NVJ to facilitate clearance.

a, Endogenously tagged Nvj1-GFP yeast expressing NES-DsRed-LuciTs were shifted to 37 °C and monitored by live-cell time-lapse fluorescence microscopy for the times shown. White arrowheads indicate locations of Nvj1 punctum, while yellow arrowheads indicate NES-LuciTs punctum. b,c, Representative SIM images of WT yeast co-expressing NES-DsRed-LuciTs (b) or NLS-DsRed-LuciTs (c) and Nvj1-sfGFP after 2 h incubation at 37 °C with 100 μM MG132. Middle panel of both b and c is an inset to better visualize the relative location of the JUNQ and INQ to the Nvj1. Right panel of both b and c shows additional cells to illustrate other phenotypes seen in the experiment. NES-DsRed-LuciTs (purple), NLS-DsRed-LuciTs (green), Nvj1-sfGFP (yellow) and DNA (blue). d, Representative confocal image of WT yeast expressing Nvj1-sfGFP (yellow) stained with FM4–64 vacuolar dye (cyan). e, Representative SIM image of WT yeast cell co-expressing NLS-DsRed-LuciTs (green) with Nvj1-sfGFP (yellow) after 2 h incubation at 37 °C with 100 μM MG132. The NLS-LuciTs can be seen extruding through the NVJ towards the vacuole. f, WT (top) and nvj1Δ (bottom) cells co-expressing NLS-LuciTs and NES-LuciTs were shifted to 37 °C, treated with 100 μM MG132 and monitored by live-cell time-lapse fluorescence microscopy for the times shown. White arrowheads indicate inclusions of NLS-LuciTs pulled into the cytoplasm. g, Representative confocal images of WT, nvj1Δ and vac8Δ yeast co-expressing NLS-EGFP-LuciTs and NES-DsRed-LuciTs after 2 h at 37 °C and treated with 100 μM MG132. NLS-EGFP-LuciTs (green), NES-DsRed-LuciTs (purple), nuclear pores (gold) and Hoechst counterstain (blue). White arrows indicate cytoplasmic localization of NLS-LuciTs. h, Quantitation of the percentage of cells containing inclusions in WT, nvj1Δ and vac8Δ yeast co-expressing NLS-EGFP-LuciTs (left) and NES-DsRed-LuciTs (right) after 2 h incubation at 37 °C with 100 μM MG132. Data are shown as mean ± s.e.m. i, Drop tests showing serial dilutions of WT, nvj1Δ and vac8Δ yeast at 30 °C and 37 °C with no treatment and 0.5 mg ml⁻¹ AZC treatment after 72 h growth at indicated temperature. All scale bars are 1 μm. Source numerical data are available in source data.

Fig. 6 |. ESCRT-mediated extrusion from the nucleus and clearance.

a, Left: the ESCRT-II/III protein Chm7 has been shown to play a role in clearance of defective nuclear pores and nuclear membrane quality control; therefore, it may be involved in the homing of INQ and JUNQ to the NVJ. Right: representative confocal images of WT yeast co-expressing Chm7_OPEN-EGFP and either NLS-DsRed-LuciTs (top) or NES-DsRed-LuciTs (bottom) after 2 h at 37 °C and treated with 100 μM MG132. Arrows indicate locations of puncta for each protein. Scale bar, 1 μm. b, Representative confocal images of WT and chm7Δ yeast co-expressing NLS-EGFP-LuciTs and NES-DsRed-LuciTs after 120 min at 37 °C and treated with 100 μM MG132. Arrows indicate locations of puncta for each protein. c, ESCRT protein Vps23, and PI3K complex Vps34–Vps15 may be involved in clearance of INQ and JUNQ. Representative confocal images of WT and vps23Δ, vps34Δ and vps15Δ yeast co-expressing NLS-EGFP-LuciTs and NES-DsRed-LuciTs after 2 h at 37 °C and treated with 100 μM MG132. Insets show the budding INQ encapsulated by nuclear pores. d, ESCRT-III proteins and the ATPase Vps4 are known to remodel membranes and could be involved in vacuolar import of PQC compartments. e, Quantitation of the percentage of cells containing inclusions of NLS- or NES-LuciTs in WT and vps4Δ yeast co-expressing NLS-EGFP-LuciTs and NES-DsRed-LuciTs after 2 h incubation at 37 °C with 100 μM MG132. Data are shown as mean ± s.e.m. f, Representative confocal images of WT and vps4Δ yeast co-expressing NLS-EGFP-LuciTs and NES-DsRed-LuciTs after 2 h at 37 °C and treated with 100 μM MG132. Green arrows indicate cytoplasmic localization of NLS-LuciTs. All scale bars are 1 μm. Chm7_OPEN is shown in teal, NLS-DsRed-LuciTs in green, NES-DsRed-LuciTs in purple, nuclear pores in gold and Hoechst counterstain in blue. Source numerical data are available in source data.

Fig. 7 |. Vacuolar clearance of the INQ and JUNQ.

a, Possible routes of entry into the vacuole for the INQ and JUNQ. b, Schematic of the LuciTs-Keima experiments. c, WT cells expressing NLS-2xKeima-LuciTs after 2 h incubation at 37 °C with 100 μM MG132. Over time, fluorescence is seen with excitation in the 558 nm channel indicating the NLS-LuciTs has encountered an acidic environment. Inset shows the transition from green to red and a structure leaving the inclusion that is fully red. Scale bars on large images are 5 μm. Scale bars on magnifications are 1 μm. Images are representative of two biologically independent experiments. d, Representative images of WT cells expressing NES-2xKeima-LuciTs after 2 h incubation at 37 °C with 100 μM MG132. Over time, fluorescence is seen with excitation in the 589 nm channel indicating the NLS-LuciTs has encountered an acidic environment. Insets show the transition from green to red and a structure leaving the inclusion that is fully red. Scale bars on large images are 5 μm. Scale bars on magnifications are 1 μm. Images are representative of two biologically independent experiments. e,f, Representative western blots of NLS-EGFP-LuciTs (e) and NES-EGFP-LuciTs (f) in WT and pep4Δ yeast. pep4Δ yeast were also treated with 1 mM PMSF to completely inhibit vacuolar proteases. A decrease in fragments can be seen in both the NLS (e) and NES (f) blots. g, Densitometric quantification of western blot bands measuring the amount of full-length LuciTs-EGFP and the fragments seen at T₀. The ratio of fragment intensity to full-length protein is shown in the graph (mean ± s.e.m. from three biologically independent experiments). One-way analysis of variance was performed using Prism software followed by Dunnett’s multiple comparisons test. Adjusted P value for WT versus pep4Δ NLS-LuciTs is 0.0018, and WT versus pep4Δ NES-LuciTs is 0.0220. Source numerical data and unprocessed blots are available in source data.

Fig. 8 |. Nuclear and cytoplasmic spatial PQC.

a, WT, nvj1Δ, and vps4Δ yeast expressing NLS-GFP-LuciTs were treated with 8 μM of FM4–64 and incubated for 2 h at 37 °C with 100 μM MG132. Cells were imaged every 5 min for 90 min. Scale bar is 1 μm. b, Model for vacuolar targeting and clearance of nuclear and cytoplasmic protein inclusions in yeast. (1) Upon heat shock, cytoplasmic and nuclear LuciTs locally misfold, and recruit chaperones and other PQC factors that facilitate the subcellular spatial sequestration on distinct protein quality control inclusions: (2) The Intranuclear Quality Control compartment (INQ) in the nucleus and Q-bodies in the cytoplasm. (3) Q-bodies then coalesce to form the JUNQ, and misfolded proteins located in the INQ and the JUNQ can be degraded by the UPS system in the nucleus and the cytoplasm, respectively (3, ii). (4) Alternatively, under conditions of limited proteasomal activity, INQ and JUNQ can converge on the periphery of the nuclear envelope, a mechanism mediated by the nuclear pores (homing). (5) The homing mechanism could represent a coordinated way to target both inclusions for clearance at the vacuole. (5, ii) Vacuolar-mediated clearance of both INQ and JUNQ. ESCRT-family proteins are involved in organizing the INQ and JUNQ for vacuolar degradation.