FAT10 is degraded by the 20S proteasome in vitro

Several in vitro studies showed that the 20S proteasome actively recognizes and degrades oxidized proteins and proteins with inherent structural disorders. Degradation by the 20S proteasome does not require ubiquitin tagging or the presence of the 19S RP and instead relies on recognition of unstructured regions of the proteins due to aging or oxidation, and of intrinsically unfolded proteins (summarized in the study by Ben-Nissan and Sharon [2014] and Kumar Deshmukh et al [2019]). Because nuclear magnetic resonance studies have shown that the ubiquitin-like modifier FAT10 contains an intrinsically disordered N-terminal tail (Aichem et al, 2018), FAT10 degradation by the 20S proteasome in the absence of the 19S RP was investigated.

Therefore, we used human 20S proteasome purified from LCL721.45 cells (Fig S1) and analyzed the degradation of recombinant FAT10 in vitro for 5 h. Untagged recombinant FAT10 is degraded by the 20S proteasome and, similar to experiments in cellulo (Schmidtke et al, 2019), FAT10 degradation was blocked by the proteasome inhibitor MG-132 (Fig 1A). These results were confirmed with Flag-tagged FAT10 purified from transiently transfected HEK293 cells (Fig 1B). In addition, the degradation of FAT10 conjugates by the 20S proteasome could be detected in this setting. Degradation of both, monomeric FAT10 and FAT10 conjugates, were inhibited by MG-132. Recent studies have shown that replacing the cysteine residues of FAT10 slows down the degradation of FAT10 monomer and conjugates in cellulo (Aichem et al, 2018). A similar stabilization of FAT10 was observed in this study in vitro (Fig 1C), which is in line with the previously raised hypothesis that the weak folding of FAT10 leads to fast proteasomal degradation as it was reported for IDPs (Aichem et al, 2018). Furthermore, it was reported that degradation of FAT10 is independent of VCP, whereas N-terminally truncated FAT10 was dependent on the unfolding activities of VCP (Aichem et al, 2018). Therefore, we compared the in vitro degradation by 20S proteasomes of WT FAT10 with FAT10-ΔAPNASC lacking the N-terminal disordered tail. Interestingly, degradation of FAT10-ΔAPNASC was almost completely arrested (Fig 1D), indicating that the disordered N-terminal tail is required for 20S-mediated degradation.

Taken together, our data show that FAT10 can be degraded by the 20S proteasome in vitro and suggests that this degradation is mediated by its characteristics as an intrinsically disordered protein.

Treatment with PSMC1 siRNA down-regulates mRNA and protein levels of Rpt2 and leads to accumulation of poly-ubiquitylated proteins

20S proteasome–mediated protein degradation has gained attention in the last years, but research in this field is limited by the difficulties of studying 20S proteasome degradation in cellulo (Kumar Deshmukh et al, 2019). To assess whether the results from our in vitro studies using purified 20S proteasomes can be confirmed in a physiological environment, a system to study the 20S proteasome in cellulo was established. An siRNA approach targeting the PSMC1 gene which translates into Rpt2, one of the six AAA+ ATPases at the base of the 19S RP, was used to disturb 26S proteasome function. Because of the proposed key role of Rpt2 in the regulation of the 20S α-subunit gate opening and stabilization of the 26S complex (Köhler et al, 2001; Gillette et al, 2008), we hypothesized that knockdown of Rpt2 would impair the stability of the 19S RP and impair the 26S proteasome activity. We used a HEK293T cell line stably overexpressing Flag-FAT10 and transfected the cells with PSMC1 siRNA. To increase the knockdown efficiency, transfection with siRNA was repeated on day 3. Rpt2 was down-regulated in PSMC1 siRNA–treated cells, whereas Rpt2 levels stayed the same in unspecific control siRNA–treated cells (Fig 2A). Strikingly, poly-ubiquitylated proteins were accumulated on day 3, 5, and 6 post PSMC1 siRNA transfection (Fig 2A), indicating the impairment of 26S proteasome function. This confirms the important role of Rpt2 in the degradation of substrate proteins by the 26S proteasome (Köhler et al, 2001). Analysis of 26S proteasomes in cell lysates by native gel electrophoresis and immunoblotting revealed a decreased amount of 26S proteasomes upon PSMC1 knockdown (Fig 2B).

siRNA-mediated knockdown of Rpt2 affects FAT10 degradation only mildly

This approach was used to investigate the degradation of FAT10 by the 20S proteasome in cellulo. Cells were transfected twice with siRNA, and a cycloheximide (CHX) chase experiment was performed on day 5 post first transfection. This time point was chosen because cells had the lowest Rpt2 levels and later on the viability of the cells was declining, most likely due to dysfunction of the 26S proteasome (data not shown). Rpt2 was down-regulated and poly-ubiquitylated proteins accumulated in PSMC1 siRNA–treated cells (Fig 3A). Similar to in vitro analysis (Fig 1A and B), FAT10 was rapidly degraded in control siRNA–treated samples during 5 h of the CHX chase. In PSMC1 siRNA—treated samples, an accumulation of FAT10 was already seen at time point 0 when starting the CHX chase, suggesting a role of the 26S proteasome in its degradation. However, immunoblotting and quantification of three independent experiments revealed a similar degradation kinetic during the CHX chase like the controls (Fig 3A and C). Nevertheless, degradation of poly-ubiquitylated proteins was markedly reduced in the PSMC1 siRNA samples which show the dysfunction of the 26S proteasome (Fig 3B).

miRNA-mediated knockdown of Rpt2 impairs the 26S proteasome

Even though the siRNA results suggest an impaired degradation of poly-ubiquitylated proteins in the absence of the 26S proteasome, varying knockdown efficiency complicates data interpretation. To overcome this problem, a stable cell line inducibly expressing PMSC1 miRNA was generated. The tetracycline responsive T-REx system (Thermo Fisher Scientific) was used. T-REx-293 cells were stably transfected with the plasmid pT-Rex-Dest30_PSMC1 encoding three different miRNAs directed against PSMC1. All three miRNA sequences are encoded on the same plasmid to ensure common integration, and the concerted GFP expression allowed to sort for positive cells and later to monitor stable expression of the introduced sequences. We found strong GFP expression after induction with tetracycline (Fig 4A) and a strong PSMC1 knockdown (Fig 4B). Time course analysis of the two different cell clones AU and AY revealed strong reduction of Rpt2 3 d post induction which goes along with an accumulation of poly-ubiquitylated proteins (Fig 4C), indicating the impairment of the 26S proteasome. A third clone (AK) did not show differences in accumulation of poly-ubiquitinated proteins and, therefore, was not used for further analyses. Furthermore, native gel analysis showed strongly decreased 26S expression 3 d post induction (Fig 4D). In contrast, the overall chymotrypsin-like activity of the proteasome in cells was not impaired by PSMC1 knockdown as assessed by hydrolysis of the cell-permeable fluorogenic substrate MeOSuc-GLF-AMC (Fig S2) and dose-dependent inhibition by bortezomib, indicating that the 20S proteasome is still functional. Taken together, these data strongly suggest an impaired function of the 26S proteasome due to dissociation of the complex in cells with decreased Rpt2 expression.

miRNA-mediated knockdown of Rpt2 leads to FAT10 accumulation

Because we observed an efficient knockdown of Rpt2 and impaired 26S proteasome function, we performed FAT10 degradation kinetic experiments in these cells. To analyze the degradation of FAT10 independent of the 26S proteasome, we transfected the cells with Flag-FAT10 (pcDNA3.1-3xFLAG-His-FAT10) on day 3 post miRNA induction and performed a CHX chase experiment on day 5. Strong accumulation of poly-ubiquitylated proteins could be observed in PSMC1 knockdown (Tet+) samples which was not markedly reduced during the chase, confirming the impairment of the 26S proteasome (Fig 5). In contrast to the in vitro data, FAT10 degradation was strongly impaired in these cells and showed a similar degradation kinetic as poly-ubiquitin conjugates.

Taken together, our data show that Rpt2 knockdown by miRNA can effectively impair the 26S proteasome, resulting in an accumulation of poly-ubiquitylated proteins. Hence, this system is an attractive tool to investigate in cellulo degradation of proteins by the 20S proteasome, overcoming the limitations of in vitro studies with isolated proteins. Furthermore, we could show that even though FAT10 is degraded by the 20S proteasome in vitro, this does not reflect FAT10 degradation in cellulo.

Cell culture

The human lymphoblastoid cell line LCL721.45 (DeMars et al, 1984) was cultured in IMDM supplemented with 10% FBS and 1% penicillin/streptomycin in a humidified 5% CO₂ atmosphere at 37°C. T-REx-293 cells, stably expressing the tetracycline repressor, were originally purchased from Thermo Fisher Scientific. The Flag-FAT10-overexpressing HEK293T cell line was previously described (Roverato et al, 2021). All HEK293-derived cell lines were cultured in DMEM supplemented with 10% FBS and 1% penicillin/streptomycin in a humidified 5% CO₂ atmosphere at 37°C. For T-REx-293 cells tetracycline negative serum from Bio&Sell GmbH was used. 0.05% trypsin–EDTA was used for detachment after washing with PBS. After centrifugation cell number was determined using an automated cell counter (Nexcelom), and cell density was adjusted. All cell culture reagents were purchased from Thermo Fisher Scientific if not stated otherwise.

Cell transfection

Cells were grown in antibiotic-free medium (DMEM with 10% FBS) to a confluency of 80%. Plasmid DNA and TransIT-LT1 transfection reagent (Mirus Bio LLC) were mixed in a ratio of 1:3 (μg DNA:μl reagent) and added dropwise to the cells. Medium was replaced with fresh fully supplemented medium 24 h after transfection. For FAT10 degradation experiments, the plasmid pcDNA3.1-His-3xFLAG-Fat10, encoding FLAG-tagged FAT10 (Chiu et al, 2007), was used for transient transfection.

Purification of Flag-FAT10 from transfected HEK293 cells

HEK293 cells were transfected with pcDNA6-His-3xFlag-FAT10 plasmid and TransIT-LT1 transfection reagent as described above. After 46 h, cells were harvested and lysed with PBS containing 0.5% NP40 and protease inhibitor mix (Roche) for 20 min on ice. Debris were removed by centrifugation (13,000g, 5 min). Nickel–IDA beads (Macherey-Nagel) were added to the supernatant and incubated at 4°C overnight. Samples were eluted from the beads with 300 mM imidazole (200 μl) in three fractions and filtered with an Amicon Ultra-15 3 K anion filter (Merck Millipore Ltd.).

20S proteasome and recombinant FAT10 purification

The 20S proteasome was purified as described previously (Groettrup et al, 1995; Schmidtke et al, 1996; Basler & Groettrup, 2012). In brief, cell pellets from LCL721.45 cells were collected and lysed with 100 mM KCl + 0.1% Triton X-100. The suspension was homogenized with a Dounce homogenizer and subsequently centrifuged for 30 min at 30,600g and 4°C. The supernatant was mixed with the anion exchanger DEAE Sephacel (GE Healthcare) and incubated overnight at 4°C while rotating. The next day, the DEAE mix was column purified and eluted with 500 mM KCl. (NH₄)₂SO₄ was added (35% of total volume), and samples were centrifuged twice for 20 min at 17,211g at 4°C. The resulting pellet was resolved in 100 mM KCl and incubated on ice for 60 min. Afterward, the dissolved pellet was added on a sucrose gradient (15–40%) and centrifuged for 16 h at 274,355g at 4°C.

After centrifugation, the sucrose gradient was separated into fractions of 700 μl and tested for proteasomal activity and further purified via a Resource Q column and fast protein liquid chromatography. Successful isolation of the 20S proteasome was confirmed with activity assay and SDS–PAGE.

Recombinant FAT10 was purified from transfected Escherichia coli BL21(DE3) CodonPlus cells (Stratagene) as described before using Ni-affinity chromatography and size-exclusion gel filtration (Aichem et al, 2018).

In vitro degradation of FAT10

For in vitro degradation analysis, purified 20S proteasome was incubated with different FAT10 forms. 300 ng of purified 20S proteasome was added to 200 ng of recombinant FAT10 or 10 μl eluate (hFLAG-FAT10 from HEK293 cells). As a control for proteasomal-specific degradation, 10 μM proteasome inhibitor MG-132, starting 30 min before FAT10 addition, was used. Samples were incubated at 37°C for the indicated time in proteasome buffer (50 mM Tris, pH 7.5, 25 mM KCl, 10 mM NaCl, 1 mM MgCl₂, 1 mM DTT, 0.1 mM EDTA). After that, samples were boiled with 5× SDS sample buffer supplemented with 5% 2-mercaptoethanol at 95°C for 5 min. Samples were analyzed by SDS–PAGE and Western blotting.

siRNA transfection

Cells were grown in a six-well dish to a confluency of 80% in antibiotic-free medium. PSMC1 siRNA kit (sc-92427; Santa Cruz Biotechnology) was used according to the manufacturer’s instructions to transfect Flag-FAT10-overexpressing Hek293T cells as described previously (Roverato et al, 2021). Transfection was repeated after 3 d to ensure prolonged knockdown. Unspecific scrambled siRNA (sc-37007; Santa Cruz Biotechnology) served as a control. Transfection was performed in FBS-free siRNA transfection medium (sc-36868; Santa Cruz Biotechnology). Medium was changed after 24 h of incubation.

miRNA-mediated knockdown of PSMC1

For generation of a stable cell line with inducible knockdown of PSMC1, we used T-REx-293 cells and the BLOCK-iT Inducible Pol II miR RNAi Expression Vector Kit with EmGFP (Invitrogen) according to the manufacturer’s instructions. In brief, we designed three different miRNAs targeting PSMC1 mRNA using Life Technologies’ RNAi Designer. The following oligonucleotides were cloned into pT-REx-DEST30 Gateway to generate the pT-REx-DEST30_PSMC1 Gateway plasmid: NM_002802.3_658_top (5′-TGCTGATAATATTCAGGATGGGTGAGGTTTTGGCCACTGACTGACCTCACCCACTGAATATTAT-3′), NM_002802.3_658_bottom (5′-CCTGATAATATTCAGTGGGTGAGGTCAGTCAGTGGCCAAAACCTCACCCATCCTGAATATTATC-3′), NM_002802.3_684_top (5′-TGCTGTTAGGAGGCTTTATACCCATCGTTTTGGCCACTGACTGACGATGGGTAAAGCCTCCTAA-3′), NM_002802.3_684_bottom (5′-CCTGTTAGGAGGCTTTACCCATCGTCAGTCAGTGGCCAAAACGATGGGTATAAAGCCTCCTAAC-3′), and NM_002802.3_711_top (5′-TGCTGCCAGGTGGACCATAGAGAATGGTTTTGGCCACTGACTGACCATTCTCTGGTCCACCTGG-3′), NM_002802.3_711_bottom (5′-CCTGCCAGGTGGACCAGAGAATGGTCAGTCAGTGGCCAAAACCATTCTCTATGGTCCACCTGGC-3′). T-REx-293 cells were transfected with the pT-REx-DEST30_PSMC1 Gateway plasmid as described above. After 5 d, cells were treated with 10 μg/ml tetracycline for 6 h, and GFP-positive cells were single-cell sorted into 96-well plates to select for stably transfected single-cell clones. For single-cell sorting and subsequent GFP measurements, cells were harvested and resuspended in PBS containing 2% FBS and 2 mM EDTA. SYTOX Blue (Thermo Fisher Scientific) was included for live/dead cell discrimination. Single cell-sorting was performed with BD FACSAria III, and confirmatory GFP measurements were analyzed with BD Accuri C6 (both BD Biosciences). Generated cell lines were tested negative for mycoplasma.

Real-time RT-PCR

To evaluate knockdown efficiency, RNA was isolated using the RNeasy Mini kit (QIAGEN) according to the manufacturer’s instructions. 1 μg of RNA was reversely transcribed into cDNA using the cDNA Synthesis Kit (Biozym). Real-time RT-PCR was performed using the Blue S’Green qPCR Kit (Biozym) with three specific primers for PSMC1 (PSMC1_F1 5′-AAGAAAAGCAAGAGGAGGAAAG-3′; PSMC1-R1 5′-CACAGATGTAGACACGATGG-3′; PSMC1_F2 5′-AGCAAACCAAACCTCAGCC-3′; PSMC1_R2 5′-TCCCCTAGAATCAAATCCATCC-3′; PSMC1_F3 5′-ACACTACGTCAGCATTCTTTC-3′; PSMC1_R3 5′-ATCCATCAGCACCCCTATC-3′). Expression of PSMC1 was normalized to the expression of HPRT, which served as housekeeping gene (hHPRT-fwd 5′-TGGACAGGACTGAACGTCTTG-3′; hHPRT-rev 5′-CCAGCAGGTCAGCAAAGAATTTA-3′). Results are shown as 2^−ΔΔCt values.

SDS–PAGE and Western blotting

Western blotting was performed as described previously (Oliveri et al, 2022). Shortly, cells were harvested and lysed in RIPA (50 mM Tris-buffered HCl, pH 7.5, containing 150 mM NaCl, 1% NP-40, 0.5% SDS) or NP40 lysis buffer (20 mM Tris [pH 7.6], 50 mM NaCl, 10 mM MgCl₂, 1% NP40 [Igepal]) supplemented with 1× protease inhibitor cocktail (Roche) for 30 min on ice. Cell debris was removed via centrifugation for 20 min at 18,000g and 4°C. Protein concentrations were determined using the Pierce BCA protein assay kit (Thermo Fisher Scientific), and equal amounts were subjected to SDS–PAGE. Separated proteins were transferred onto a 0.45-μm nitrocellulose membrane (Whatman Protran) by wet blotting at 110 V for 90 min or by semi-dry blotting using the Trans-Blot Turbo System (Bio-Rad). After transfer, membranes were rinsed with ddH₂O and blocked in Odyssey TBS Blocking Buffer (LI-COR Biosciences) for 60 min at room temperature. Membranes were incubated in primary antibody solutions overnight at 4°C (list of antibodies can be found in Table S1). Anti-GAPDH antibody served as loading control. IRDye800CW goat anti-rabbit or anti-mouse and IRDye680RD goat anti-mouse or anti-rabbit (LI-COR Biosciences) were used as secondary antibodies. Signals were analyzed with the LI-COR Odyssey Imager and the Image Studio Lite Version 5.2.

CHX chase experiment

Cells were grown in six-well plates to a confluency of about 95%. Cells were treated with 50 μg/ml CHX (Merck) to stop protein synthesis. One sample was in addition treated with 10 μM proteasome inhibitor MG-132 (Sigma-Aldrich). Cells were harvested after 0, 1, 2, 3, and 5 h, and cell pellets were stored at −20°C. Samples were analyzed using SDS–PAGE followed by Western blotting.

Native PAGE

Native PAGE was performed as described before (Elsasser et al, 2005; Njomen et al, 2018). In brief, cells were lysed in native lysis buffer (50 mM Tris [pH 8], 0.5 mM EDTA, 10% glycerol, and 4 mM ATP). Cell suspensions were subjected to seven cycles of freezing in liquid nitrogen and thawing at room temperature. BCA assay was performed to determine the protein concentration, and 30 μg were diluted with native lysis buffer to a final volume of 30 μl. Samples were carefully mixed with 7 μl of 5× sample buffer (250 mM Tris, pH 7.4; 50% glycerol, 60 ng/ml xylene cyanol) and loaded onto native 4% polyacrylamide gels (90 mM Tris base/boric acid [pH 8.3], 5 mM MgCl₂, 0.5 mM EDTA, 1 mM ATP, 0.1% APS, 0.1% temed). Electrophoresis was performed at 4°C using 1× Towbin’s basic buffer, starting at 50 V for 30 min followed by 110 V for 3 h. After incubation in SDS running buffer containing 2% SDS and 1.5% 2-mercaptoethanol for 15 min at RT, Western blotting was performed using the Trans-Blot Turbo System (Bio-Rad) for 20 min at 2.5 A. Primary antibodies anti-α1-7 (Table S1) and anti-Rpt5 (TBP1; Table S1) were used to visualize specifically the 20S or 19S/26S complexes, as described previously (Semren et al, 2015).

Fluorogenic substrate assay

To investigate the activity of the 20S proteasome in PSMC1 knockdown cells, the cell permeable substrate MeoSuc-GLF-AMC was used as previously described (Basler et al, 2018). Shortly, cells were treated with different bortezomib concentrations in PBS + 25 mM HEPES for 30 min at 37°C. The cell permeable substrate MeoSuc-GLF-AMC (Bachem) (in DMSO 10 mM) was added at 40 μM to the cells and incubated at 37°C. The fluorescence intensity in the wells containing the cells was measured at an excitation wavelength of 360 nm and emission wavelength of 465 nm (infinite M200 pro; TECAN).

Statistical analysis

Quantification of band intensities of Western blots was performed using Image Studio Lite Version 5.2 (LI-COR Biosciences). Information about sample size is detailed in the figure legends. Statistical analyses were performed with GraphPad Prism Version 9 and are detailed in the figure legends.