Characterization and dynamics of aggresome formation by a cytosolic GFP-chimera

Abstract

Formation of a novel structure, the aggresome, has been proposed to represent a general cellular response to the presence of misfolded proteins (Johnston, J.A., C.L. Ward, and R.R. Kopito. 1998. J. Cell Biol. 143:1883-1898; Wigley, W.C., R.P. Fabunmi, M.G. Lee, C.R. Marino, S. Muallem, G.N. DeMartino, and P.J. Thomas. 1999. J. Cell Biol. 145:481-490). To test the generality of this finding and characterize aspects of aggresome composition and its formation, we investigated the effects of overexpressing a cytosolic protein chimera (GFP-250) in cells. Overexpression of GFP-250 caused formation of aggresomes and was paralleled by the redistribution of the intermediate filament protein vimentin as well as by the recruitment of the proteasome, and the Hsp70 and the chaperonin systems of chaperones. Interestingly, GFP-250 within the aggresome appeared not to be ubiquitinated. In vivo time-lapse analysis of aggresome dynamics showed that small aggregates form within the periphery of the cell and travel on microtubules to the MTOC region where they remain as distinct but closely apposed particulate structures. Overexpression of p50/dynamitin, which causes the dissociation of the dynactin complex, significantly inhibited the formation of aggresomes, suggesting that the minus-end-directed motor activities of cytoplasmic dynein are required for aggresome formation. Perinuclear aggresomes interfered with correct Golgi localization and disrupted the normal astral distribution of microtubules. However, ER-to-Golgi protein transport occurred normally in aggresome containing cells. Our results suggest that aggresomes can be formed by soluble, nonubiquitinated proteins as well as by integral transmembrane ubiquitinated ones, supporting the hypothesis that aggresome formation might be a general cellular response to the presence of misfolded proteins.

Figure 1

Overexpression of a cytosolic protein leads to aggresome formation. COS-7 cells were transfected with chimeras between GFP and various domains of p115. A schematic representation of each chimeric protein used in the transfections is shown above each panel; GFP, green fluorescent protein; H1 and H2, domains within p115 highly homologous to Uso1p; C1-4, predicted p115 coiled-coil regions; AD, p115 acidic domain. 48 h after transfection, cells were processed for immunofluorescence using anti-giantin antibodies. Chimeras containing full-length p115 (GFP-FL; A and B) or lacking a small region of the COOH terminus (GFP-766; C and D) colocalized with giantin in the Golgi region. The chimera containing only the NH₂-terminal H1 and H2 region of p115 (GFP-250; E and F) localized to a perinuclear structure surrounded by displaced Golgi elements. Bar, 10 μm.

Figure 2

Aggresome formation depends on the position of the GFP within the chimera. COS-7 cells were transfected with the indicated GFP chimeras or with GFP alone. The region of p115 within the constructs is the same as in Fig. 1 in GFP-250. 48 h after transfection, cells were imaged for GFP. Only the chimera containing GFP at the NH₂ terminus (A) formed an aggresome, whereas the chimera containing GFP at the COOH terminus (B) or GFP alone (C) did not form an aggresome. Bar, 10 μm.

Figure 3

Ultrastructural analysis of GFP-250–transfected cells. COS-7 cells were transfected with GFP-250, and processed for electron microscopy 48 h later. An aggregate of particulate material is seen close to the nucleus (A). A higher magnification of the material (B) shows that it surrounds a centriole (arrowhead) and is surrounded by filamentous material (dashed rectangle). The filamentous material appears to be composed of 10-nm filaments (asterisk in C). M, mitochondria; N, nucleus. Bars: (A and B) 1 μm; (C) 0.25 μm.

Figure 4

Aggresome formation causes a redistribution of the intermediate filament protein vimentin. COS-7 cells were transfected with GFP-250. 48 h after transfection, cells were processed for immunofluorescence using anti-vimentin antibodies and imaged to localize the GFP-250 and the vimentin. Images in A were acquired using conventional optics. Images in B are deconvoluted 0.4-μm optical sections. Vimentin is recruited into a ring structure around the aggresome. Bars, 10 μm.

Figure 5

Cytosolic chaperones are recruited to the aggresome. COS-7 cells were transfected with GFP-250. (A) 48 h after transfection, cells were processed for immunofluorescence using antibodies to Hdj1, Hdj2, Hsc70, and TCP1, and imaged to localize the GFP chimera and the chaperone. (B) 48 h after transfection, cells were processed for immunofluorescence using antibodies to Hsc70 and vimentin, and imaged to localize the GFP-250, Hsc70, and vimentin. Hdj1, Hdj2, and TCP1 colocalize with the aggresome, while Hsc70 forms a ring around the aggresome. Vimentin colocalizes with the ring of Hsc70. Images in B are deconvoluted 0.4-μm optical sections. Bars, 10 μm.

Figure 6

The proteasome is recruited to the aggresome and is responsible of the degradation of GFP-250. COS-7 cells were transfected with GFP-250. (A) 48 h after transfection, cells were processed for immunofluorescence using antibodies to the α-subunit of the 20S proteasome, and imaged for GFP-250 and the proteasome. The proteasome is recruited to the GFP-250 aggresome. (B) 36 h after transfection, clasto-lactacystin β-lactone (10 μM) or media were added to the cells for 11 h. Cells were then imaged to localize the GFP chimera. The number of cells containing an aggresome is reported as percentage of transfected cells. 700 cells were counted for each condition. Proteasome inhibition causes an increase in GFP-250 aggresome formation. (C) COS-7 cells transfected for 12 h with GFP-250 were pulse labeled with [³⁵S]methionine for 1 h and then chased with unlabeled medium for the times indicated in the presence or absence of clasto-lactacystin β-lactone. The detergent soluble fractions were then immunoprecipitated with anti-p115 antibodies, separated by SDS-PAGE, and detected by fluorography. The gel was quantitatively evaluated by densitometry and presents percentage of radiolabeled GFP-250 remaining at each chase time point. Proteasome inhibition increases the half-life of GFP-250. Bar, 10 μm.

Figure 6

The proteasome is recruited to the aggresome and is responsible of the degradation of GFP-250. COS-7 cells were transfected with GFP-250. (A) 48 h after transfection, cells were processed for immunofluorescence using antibodies to the α-subunit of the 20S proteasome, and imaged for GFP-250 and the proteasome. The proteasome is recruited to the GFP-250 aggresome. (B) 36 h after transfection, clasto-lactacystin β-lactone (10 μM) or media were added to the cells for 11 h. Cells were then imaged to localize the GFP chimera. The number of cells containing an aggresome is reported as percentage of transfected cells. 700 cells were counted for each condition. Proteasome inhibition causes an increase in GFP-250 aggresome formation. (C) COS-7 cells transfected for 12 h with GFP-250 were pulse labeled with [³⁵S]methionine for 1 h and then chased with unlabeled medium for the times indicated in the presence or absence of clasto-lactacystin β-lactone. The detergent soluble fractions were then immunoprecipitated with anti-p115 antibodies, separated by SDS-PAGE, and detected by fluorography. The gel was quantitatively evaluated by densitometry and presents percentage of radiolabeled GFP-250 remaining at each chase time point. Proteasome inhibition increases the half-life of GFP-250. Bar, 10 μm.

Figure 7

GFP-250 is not ubiquitinated. COS-7 cells were transfected with GFP-250. (A) 48 h after transfection, cells were extracted with Triton X-100, immunoprecipitation buffer (IPB) or RIPA (RIPA) buffer as in Methods, and the soluble (S) and insoluble (I) fractions separated by SDS-PAGE and immunoblotted with antibodies against GFP or tubulin. GFP-250 remains insoluble even under the most stringent extraction conditions. (B) 48 h after transfection, cells were extracted with Triton X-100, and the soluble (S) and insoluble (I) fractions were separated by SDS-PAGE and immunoblotted with antibodies against GFP. The membrane was then stripped and reprobed with anti-ubiquitin antibody. GFP-250 does not appear to be ubiquitinated. (C) 48 h after transfection, cells were processed for immunofluorescence using antibodies against ubiquitin, and imaged to localize GFP-250 and ubiquitin. Ubiquitin does not colocalize with GFP-250 aggresome. COS-7 cells were also transfected with CFTR and incubated with clasto-lactacystin β-lactone (10 μM) for 11 h after 36 h of transfection. Cells were processed for immunofluorescence using antibodies against CFTR and ubiquitin, and imaged to localize CFTR and ubiquitin. Ubiquitin does colocalize with the CFTR aggresome. Bar, 10 μm.

Figure 7

GFP-250 is not ubiquitinated. COS-7 cells were transfected with GFP-250. (A) 48 h after transfection, cells were extracted with Triton X-100, immunoprecipitation buffer (IPB) or RIPA (RIPA) buffer as in Methods, and the soluble (S) and insoluble (I) fractions separated by SDS-PAGE and immunoblotted with antibodies against GFP or tubulin. GFP-250 remains insoluble even under the most stringent extraction conditions. (B) 48 h after transfection, cells were extracted with Triton X-100, and the soluble (S) and insoluble (I) fractions were separated by SDS-PAGE and immunoblotted with antibodies against GFP. The membrane was then stripped and reprobed with anti-ubiquitin antibody. GFP-250 does not appear to be ubiquitinated. (C) 48 h after transfection, cells were processed for immunofluorescence using antibodies against ubiquitin, and imaged to localize GFP-250 and ubiquitin. Ubiquitin does not colocalize with GFP-250 aggresome. COS-7 cells were also transfected with CFTR and incubated with clasto-lactacystin β-lactone (10 μM) for 11 h after 36 h of transfection. Cells were processed for immunofluorescence using antibodies against CFTR and ubiquitin, and imaged to localize CFTR and ubiquitin. Ubiquitin does colocalize with the CFTR aggresome. Bar, 10 μm.

Figure 8

Dynamics of aggresome formation. COS-7 cells were transfected with GFP-250 and 48 h later imaged every 60 s for 120 min. (A) Image series showing aggresome formation. Accompanying QuickTime™ movie shows aggresome progression between 60 and 120 min (video available at http://www.jcb.org/cgi/content/full/146/6/1239/F8/DC1). (B) Path taken by a representative particle en route to the aggresome. One particle was followed over a period of 45 min until it disappeared into the aggresome. (C) Fluorescent intensities associated with the aggresome region of interest (ROI) for the cell shown in A are plotted at 60-s intervals. ROI is marked in B by a gray square. (D) Particle velocity in control (closed circles) and in nocodazole-treated (open circles) cells. Transfected cells were incubated with 10 μg/μl of nocodazole for 1 h at 4°C, then returned to 37°C and imaged every 10 s. (E) Population analysis of motile particles in control (black bars) and nocodazole-treated (gray bars) cells. Velocities of 10 different particles were measured, grouped into 0.025-μm/s intervals, and plotted. Representative traces of one particle for each condition are shown. Bar, 10 μm.

Figure 8

Dynamics of aggresome formation. COS-7 cells were transfected with GFP-250 and 48 h later imaged every 60 s for 120 min. (A) Image series showing aggresome formation. Accompanying QuickTime™ movie shows aggresome progression between 60 and 120 min (video available at http://www.jcb.org/cgi/content/full/146/6/1239/F8/DC1). (B) Path taken by a representative particle en route to the aggresome. One particle was followed over a period of 45 min until it disappeared into the aggresome. (C) Fluorescent intensities associated with the aggresome region of interest (ROI) for the cell shown in A are plotted at 60-s intervals. ROI is marked in B by a gray square. (D) Particle velocity in control (closed circles) and in nocodazole-treated (open circles) cells. Transfected cells were incubated with 10 μg/μl of nocodazole for 1 h at 4°C, then returned to 37°C and imaged every 10 s. (E) Population analysis of motile particles in control (black bars) and nocodazole-treated (gray bars) cells. Velocities of 10 different particles were measured, grouped into 0.025-μm/s intervals, and plotted. Representative traces of one particle for each condition are shown. Bar, 10 μm.

Figure 9

p50/dynamitin overexpression inhibits aggresome formation. COS-7 cells were either transfected with GFP-250 or cotransfected with GFP-250 and p50/dynamitin in a 1:2 ratio. 48 h after transfection, cells were processed for immunofluorescence using antibodies against p50/dynamitin, and imaged to localize the GFP-250 and p50/dynamitin. The number of cells containing an aggresome in cells transfected only with GFP-250 (302 cells counted) and in cotransfected cells (143 cells counted) were counted and are expressed as percentage of transfected cells. Cells expressing p50/dynamitin show significant reduction in aggresome formation. B shows two representative examples of cells expressing high levels of p50/dynamitin, in which only small peripheral particulates can be seen. The results are from three independent experiments. Bar, 10 μm.

Figure 9

p50/dynamitin overexpression inhibits aggresome formation. COS-7 cells were either transfected with GFP-250 or cotransfected with GFP-250 and p50/dynamitin in a 1:2 ratio. 48 h after transfection, cells were processed for immunofluorescence using antibodies against p50/dynamitin, and imaged to localize the GFP-250 and p50/dynamitin. The number of cells containing an aggresome in cells transfected only with GFP-250 (302 cells counted) and in cotransfected cells (143 cells counted) were counted and are expressed as percentage of transfected cells. Cells expressing p50/dynamitin show significant reduction in aggresome formation. B shows two representative examples of cells expressing high levels of p50/dynamitin, in which only small peripheral particulates can be seen. The results are from three independent experiments. Bar, 10 μm.

Figure 10

Disruption of the Golgi complex and microtubules does not overtly inhibit ER to Golgi transport. COS-7 cells were transfected with GFP-250. (A and B) 48 h after transfection, cells were processed for immunofluorescence using antibodies against giantin (A) or tubulin (B), and imaged to localize the GFP-250 and giantin or tubulin. Giantin is progressively displaced by the forming GFP-250 aggresome. Microtubule structure is perturbed and microtubules surround the aggresome. (C) 48 h after transfection, cells were infected with VSVtsO45 at the nonpermissive temperature of 42°C. Cells were either maintained at 42°C or shifted to the permissive temperature of 32°C for 1 h. Cells were then processed for immunofluorescence using antibodies against VSV-G protein, and imaged to localize the GFP chimera and the VSV-G protein. In untransfected cells as well as in cells containing an aggresome, VSV-G protein was delivered to the Golgi complex. Bars, 10 μm.