Subcellular distribution and dynamics of active proteasome complexes unraveled by a workflow combining in vivo complex cross-linking and quantitative proteomics

Abstract

Through protein degradation, the proteasome plays fundamental roles in different cell compartments. Although the composition of the 20S catalytic core particle (CP) has been well documented, little is known about the composition and dynamics of the regulatory complexes that play a crucial role in its activity, or about how they associate with the CP in different cell compartments, different cell lines, and in response to external stimuli. Because of difficulties performing acceptable cell fractionation while maintaining complex integrity, it has been challenging to characterize proteasome complexes by proteomic approaches. Here, we report an integrated protocol, combining a cross-linking procedure on intact cells with cell fractionation, proteasome immuno-purification, and robust label-free quantitative proteomic analysis by mass spectrometry to determine the distribution and dynamics of cellular proteasome complexes in leukemic cells. Activity profiles of proteasomes were correlated fully with the composition of protein complexes and stoichiometry. Moreover, our results suggest that, at the subcellular level, proteasome function is regulated by dynamic interactions between the 20S CP and its regulatory proteins-which modulate proteasome activity, stability, localization, or substrate uptake-rather than by profound changes in 20S CP composition. Proteasome plasticity was observed both in the 20S CP and in its network of interactions following IFNγ stimulation. The fractionation protocol also revealed specific proteolytic activities and structural features of low-abundance microsomal proteasomes from U937 and KG1a cells. These could be linked to their important roles in the endoplasmic reticulum associated degradation pathway in leukemic cells.

Fig. 1.

Strategy used for the determination of subcellular proteasome complexes distribution in U937 and KG1a cells.

Fig. 2.

Optimization of cross-linking with formaldehyde and validation of the fractionation protocol. A, Effect of in vivo formaldehyde cross-linking on proteasome ChT-like specific activity and purification yield. U937 cells were in vivo treated with different concentrations of formaldehyde (0%, 0.1%, 0.2%, 0.3%, and 0.5%) to stabilize protein–protein interactions. Results were obtained with three biological replicates. Error bars indicate standard deviation. B, Cross-linking with formadehyde at 0.1% is essential to maintain proteasome in the nucleus during the fractionation protocol and is compatible with subcellular fractionation. U937 cells in vivo-treated (0.1 and 0.2%) or not (0%) with formaldehyde were fractionated and each fraction (C: cytosol, M: Microsomes; N: nucleus) was analyzed by Western blot with antibodies directed against GAPDH (cytoplasm marker), Calnexin (ER marker), Histone H1 (nucleus marker 1), XRCC5 (nucleus marker 2), and 20S proteasome. C, Quantitative Mass Spectrometry analysis of proteins from each cellular compartment validates the fractionation protocol with cells cross-linked with 0.1% formaldehyde. Proteins from cytosolic, microsomal, and nuclear fractions obtained from U937 cells cross-linked with 0.1% of formaldehyde were identified and quantified by label-free LC-MS/MS. Two biological replicates were performed. D, Cytoplasmic and nuclear localization of proteasome in U937 cells observed by immunofluorescence analysis. Error bars indicate standard deviation.

Fig. 3.

Quantitation and functional analyses of proteasomes. A, The 20S proteasome and total protein distributions in the cellular compartments. Total proteins (lines) and 20S proteasome contents (bars) from 200 × 10⁶ U937 and KG1a formaldehyde-crosslinked and fractionated cells were measured by Bradford and ELISA assays, respectively. The total cell lysate proteasome contents were calculated by adding the values obtained from the three subcellular compartments. B, Proteasome chymotrypsin-like (ChT-like) specific activities in cytosolic, microsomal, nuclear, and total protein lysates from U937 and KG1a fractionated cells. C, Proteasome ChT-like, trypsin-like (T-like) and Post-Glutamyl-Peptide-Hydrolysis (PGPH) specific activities in total protein lysates from U937 and KG1a nonfractionated cells. In A–C results were obtained from three biological replicates. Error bars indicate standard deviation.

Fig. 4.

Determination of 20S proteasome subunits subcellular distribution and stoichiometry. A, Schematic representation of the four main classes of 20S proteasome with different set of catalytic subunits (11). B, The proportions of all 20S proteasome subtypes were determined by label-free absolute quantification as detailed in the experimental section. C, Kinetic of immunoproteasome formation on IFNγ treatment (20 ng/ml) in the microsomal extract (0 h to 24 h). The proportions of all 20S proteasome subtypes at each time point were determined. In B–C results were obtained from three biological replicates. Error bars indicate standard deviation.

Fig. 5.

Proteasome Interacting Proteins subcellular distribution and stoichiometry of proteasome complexes in two leukemic cell lines. A, B, Proteasome regulators and major proteasome interacting proteins (PIPs) relative subcellular distribution in U937 (A) and KG1a (B) cells. Immuno-purified proteasome complexes were analyzed by label-free quantitative MS. The relative content of proteasome regulators and major PIPs was obtained by plotting their abundance indexes as a function of the cell compartment. This abundance index was normalized with the one of the cytosolic fractions to obtain a fold change in PIP abundance index. C, Chymotrypsin-like activity and stoichiometry of the different proteasome complexes in the three subcellular compartments of U937 and KG1a. The stoichiometry of the different proteasome complexes was estimated by label-free MS as explained in the experimental section. Four types of complexes were considered: the free 20S core complex, and the 20S proteasome associated to 19S, PA28αβ, PA28γ, and PA200. D, Protein quality control in the microsomes of U937 and KG1a. Calculation of the fold enrichment in KG1a versus in U937 of VCP, GRP78, Hsc71, and K48 polyUb chains per 19S RP was performed using the normalized PAIs. In A to D, results were obtained from three biological replicates. Error bars indicate standard deviation. E, Comparison of translation level and activity between U937 and KG1a. Phosphorylation of the small ribosomal subunit S6 and total S6 levels were indicative of translation activity and translation level, respectively. GAPDH and calnexin served as loading controls in the cytosol and in the microsomes, respectively. Detection was performed by Western blot analysis.

Fig. 6.

Dynamics of proteasome complexes under IFNγ stimulation. A, Proteasome regulators and major proteasome interacting proteins (PIPs) kinetics during IFNγ stimulation in U937 microsomes. The relative content of proteasome regulators and major PIPs was obtained by plotting their abundance indexes as a function of the time of IFNγ treatment. This abundance index was normalized with the one of the initial condition to obtain a fold change in PIP abundance index. B, Chymotrypsin-like activity and stoichiometry of the different proteasome complexes in microsomes of U937 during IFNγ stimulation. The stoichiometry of the different proteasome complexes was estimated by label-free MS as explained in the in the experimental section. Four types of complexes were considered: the free 20S core complex, and the 20S proteasome associated to 19S, PA28αβ, PA28γ, and PA200. Results were obtained from three biological replicates. Error bars indicate standard deviation.