Plasmids, antibodies and reagents

Plasmids expressing, HA- or myc-tagged synphilin-1, myc-tagged Siah-1 and -2, HA-tagged wild type or mutant ubiquitin, FLAG-tagged wild type or mutant parkin have been described previously [6,11,18]. The myc-tagged CHIP, UbcH6, UbcH7 and UbcH8 constructs were kind gifts from Takahashi R. (Kyoto University, Japan), while FLAG-tagged Dorfin and HA-HHARI/HA-Cbl were provided by G. Sobue (Nagoya University, Japan) and G. Guy (IMCB, Singapore) respectively. Myc-tagged Ubc13 was generated by PCR amplification using His₆-Ubc13 as template (a gift from N. Matsuda, Tokyo Metropolitan Institute of Medical Science) and cloned into pCMV-myc vector via EcoRI and XhoI restriction sites. Untagged full length parkin was generously provided by H. Walden (London Research Institute, UK). YFP-Mitofusin2 was a gift from R. Youle (Addgene plasmid #28010). Control and Ubc13 shRNA (V2LHS_171792 and_220048) were purchased from Thermoscientific. The following mouse monoclonal antibodies were used: anti-c-myc (clone 9E10) (Roche Diagnostics, Indianapolis, IN), anti-FLAG (Sigma, St. Louis, MO, USA), anti-HA (Sigma), anti-β-actin (Sigma), anti-Ubc13 (Zymed, San Francisco, CA), anti-parkin clone PRK8 (Covance) and anti-ubiquitin clone FK1 (BIOMOL, Plymouth Meeting, PA). Linkage-specific K48 and K63 antibodies were purchased from Millipore (Temecula, CA) and BIOMOL (Plymouth Meeting, PA) respectively. Rabbit anti-GFP was purchased from Abcam (Cambridge, UK). All other reagents were purchased from Sigma, except clasto-lactacystin β-lactone (BIOMOL, Plymouth Meeting, PA) and MG-132 (A.G. Scientific, San Diego, CA).

Cell culture, immunoprecipitation, and Western blot analysis

Human embryonic kidney (HEK) 293 cells were grown in DMEM with 10% FBS in a 5% CO₂ atmosphere. Cells were transiently transfected with the desired plasmid(s) using the LipofectAMINE PLUS reagent (Invitrogen, San Diego, CA) according to the manufacturer’s instructions. For proteasome inhibition studies, cells were treated at 24 h post-transfection with 1 (or 2) µM MG132 for 16 h before harvesting. Sequential fractionation of transfected cell lysates into Triton-X-soluble (S) and SDS-soluble (P) fractions was carried out as previously described [18]. Immunoprecipitation from the transfected cell lysates was performed with anti-myc or anti-Ubc13 antibody and protein G PLUS/protein A-agarose (Oncogene Sciences, Uniondale, NY), and then washed five times in lysis buffer. Immunoprecipitates or cell lysates was analyzed by Western blot analysis with ECL detection reagents (Pierce Biotechnology, Rockford, IL). Wild type and Ubc13 null mouse embryonic fibroblasts (MEFs) were kind gifts from Yao T.P. (Duke University, USA) [19]. Primary MEFs from wild type and parkin null mice (exon 7 deletion) (kind gifts from Dawson T.M., Johns Hopkins Medicine, USA) [20] were generated according to published protocol [21].

Lentivirus vector construction and cell transduction

cDNA sequences encoding myc-tagged UbcH7 and Ubc13 were PCR-amplified from their respective pcDNA3.1 plasmid clones using an EcoRV restriction site-containing forward primer (5’ AAGATATCTAGCGTTTAAA- CGGGC 3’) and a MunI restriction site-containing reverse primer (5’ AGCCAATTG- GCGGCCGCTCGAG 3’). Amplified product was digested with EcoRV and MunI and inserted into EcoRV and EcoRI sites of pL6mCWmIRESCherry. The lentivector pL6mCWmIRESCherry was modified from pLenti6/V5-D-TOPO (Invitrogen) by reengineering of the multiple cloning site, insertion of the cPPT and WPRE elements, and insertion of the IRESmCherry reporter cassette. Lentivirus packaging was performed in 293FT cells according to the protocol provided with the ViraPower™ Lentiviral Directional TOPO® Expression Kit (Invitrogen). Lentivirus particles were concentrated from cell culture supernatant according to the protocol of Deiseroth Lab (http://www.stanford.edu/group/dlab/resources/lvprotocol.pdf). Lentivirus carrying the ubiquitin expression constructs was used to transduce wild type or Ubc13 knockout MEFs. Prior to transduction, cells were cultured to ~90% confluence. Concentrated virus particles were added to cell culture medium containing 6 µg/ml of Polybrene. Long term transgene expression was maintained by selecting for resistance to Blasticidin S at a final concentration of 10 µg/ml. Transgene expression was detected by mCherry epifluorescence.

Inclusion formation and autophagic removal

The autophagic clearance of inclusions formed under conditions of proteasomal impairment was investigated using a method originally described by Fortun et al [22]. Cells were first treated with 5 µM lactacystin to facilitate inclusion formation. After 16 h incubation, the treated cells were washed out and allowed to recover in normal media for 24 h. Concurrently, a parallel set of similarly treated cells were incubated with starvation media (1% serum) to stimulate autophagy. Thereafter, cells were processed for immunocytochemical staining for blinded evaluation of inclusions. Statistical significance for all the quantitative data obtained was analyzed using Student’s t-test (*P <0.05, **P < 0.001) unless otherwise stated.

K63-polyubiquitination is enhanced in parkin-expressing cells in the presence of proteasome inhibition

Recently, K63-specific antibodies have become available from commercial sources. Although we have independently confirmed its linkage specificity in the present study (Figure S1A), we found that the sensitivity of commercially available K63 antibodies towards endogenously promoted K63 linkages under normal cell culture conditions (i.e. in the absence of proteasome inhibition) to be rather weak (not shown). To circumvent this problem, we performed our subsequent experiments in cells expressing exogenous HA-tagged wild type ubiquitin. Notably, we observed that exogenously-introduced K63 ubiquitin species (as visualized via anti-UbK63 staining) tend to reside in the pellet fraction of cell lysate (Figure S1B & C), which is consistent with our previous finding that K63-ubiquitination could influence the cellular distribution of proteins [6].

To test our hypothesis that parkin-mediated K63 ubiquitination may be enhanced in cells undergoing proteasomal stress, we next examined the immunoreactivity of anti-UbK63 in Triton-X-100-soluble (S) and -insoluble (P) lysates sequentially prepared from parkin-expressing cells in the presence or absence of proteasome inhibition. We detected a modest but significant increase in the levels of K63-linked polyubiquitination specifically in the P fraction in untreated cells expressing parkin compared to control cells (Figure 1A). Importantly, when these parkin-expressing cells were treated with the proteasome inhibitor, MG132, we observed a dramatic increase in the level of anti-UbK63 immunoreactivity, which again resides predominantly in the P fraction (Figure 1A). The same phenomenon is observed when parkin-expressing cells were treated with PSI and lactacystin, two other proteasome inhibitors but not with DMSO vehicle (Figure S2A-B). Substituting parkin with a truncation mutant deleted of its C-terminal catalytic RING domain (ΔRING) significantly reduces the level of K63 polyubiquitinated proteins in cells treated with MG132, as are substitutions with disease-associated RING mutants, T240R, T415N and G430D (Figure 1B). On the other hand, a parkin mutant carrying the M192L mutation, which resides outside the RING catalytic domain, retains the ability to promote K63-linked polyubiquitination (Figure 1B). Our results thus suggest that proteasome inhibition promotes parkin-mediated K63-linked ubiquitination, an activity that is clearly dependent on the integrity of its RING domain. To extend this finding, we also repeated our experiment with MG132-treated cells expressing other E3 members. Anti-UbK63 immunoblotting of lysates prepared from these variously transfected cells revealed that Siah-1, for which no association with K63-linked polyubiquitination has been reported to date, as well as two other RING-containing E3s, HHARI and Cbl, failed to enhance the levels of K63 polyubiquitinated proteins in the P fraction in response to proteasome inhibition (Figure 1C). However, CHIP (which has been associated with K63-linked ubiquitination) appears capable of mediating the phenomenon (Figure 1C). Taken together, our results demonstrate that in the presence of proteasome inhibition, the K63-linked ubiquitination activity of parkin is significantly enhanced, which results in the accumulation of proteins in detergent-insoluble fractions.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Parkin promotes K63-linked ubiquitination.

(A) Left, Representative anti-K63 and anti-FLAG immunoblots of cell extracts sequentially prepared with Triton-X 100 (S) and SDS (P)-containing buffer from untreated or MG132-treated HEK cells transfected with HA-tagged ubiquitin, and vector or FLAG-tagged parkin, as indicated. Right, Bar graph showing the relative densitometric levels of K63-linked ubiqiuitination (normalized to respective actin levels) under different conditions, as indicated (**P < 0.001; ^{# #} P < 0.001). (B–C) Representative anti-K63 immunoblots of S and P fractions of HEK cells expressing (B) wild type or mutant parkin species, (C) parkin and various E3 members, as indicated. The blots above were stripped and reprobed with anti-actin antibody to reflect loading variations. These experiments were repeated 3 times with similar results.

https://doi.org/10.1371/journal.pone.0073235.g001

Proteasome inhibition promotes the recruitment of Ubc13 by parkin

Given that Ubc13 is uniquely associated with K63-linked ubiquitination, and our observation above that parkin-mediated K63-linked ubiquitination is promoted by proteasome inhibition, we surmised that the binding affinity between Ubc13 and parkin may be influenced by the functional status of the proteasome. To address this, we carried out co-immunoprecipation experiments with cells transfected with myc-tagged Ubc13 and FLAG-tagged parkin in the presence or absence of MG132 treatment. As suspected, the amount of parkin that co-immunoprecipitated with Ubc13 is significantly enhanced in the presence of proteasome inhibition (Figures 2A & S2C). Notably, the levels of parkin and Ubc13 are not appreciably affected in MG132-treated cells (Figure 2A), suggesting that their enhanced interaction is likely a result of increased binding affinity for each other. We also observed the same phenomenon with transfected cells treated with two other proteasome inhibitors, PSI and lactacystin, but not with DMSO vehicle (Figure 2B), which correlates with our earlier observation (Figure S2A-B). Further, this phenomenon appears specific to Ubc13 as neither UbcH6 nor H7 mediates such an outcome (Figure 2C). Moreover, unlike parkin, the closely-related E3 member HA-HHARI fails to exhibit increased association with Ubc13 in the presence of proteasome inhibition (Figure 2D), although CHIP1, which enhances K63-linked ubiquitination in the presence of MG132 (Figure 1C), also display increased interaction with Ubc13 under such conditions (Figure 2E). Thus, parkin-Ubc13 interaction appears to be rather specifically enhanced under conditions of proteasome impairment, which provides an explanation for the observed enhancement in K63-linked ubiquitination in MG132-treated parkin-expressing cells. Consistent with this, shRNA-mediated silencing of Ubc13 expression results in a significant reduction in the level of K63-linked ubiquitination in parkin-transfected cells in the presence of MG132 (Figure S2D). Interestingly, we also observed a correlative decrease in parkin expression upon Ubc13 expression silencing, which suggest a role for Ubc13 in stabilizing parkin. Supporting this, we found that the level of parkin is significantly reduced in FLAG-parkin transduced Ubc13-/- MEFs compared to those expressing in wild type MEFs, even in the presence of MG132 treatment (Figure S2E). Notably, MEFs derived from parkin null mice also exhibit a reduction in the level of K63-linked ubiquitination relative to their wild type counterparts, albeit more modestly so, when treated with MG132 (Figure S2F).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. Proteasome inhibition promotes the interaction between parkin and Ubc13.

(A) Left, A portion of Triton-X-soluble lysates prepared from untreated or MG132-treated HEK293 cells expressing FLAG tagged parkin alone or with myc-tagged Ubc13 were subjected to anti-myc immunoprecipitation followed by anti-FLAG and anti-myc immunoblotting (IPmyc). The remainder lysates prepared from these variously transfected cells (INPUT) were subjected to anti-FLAG and anti-myc immunoblotting to show the expression levels of FLAG-parkin and myc-Ubc13 respectively. Right, Bar graph showing the densitometric levels of parkin that co-immunoprecipitated with Ubc13 from untreated or MG132-treated transfected cells (*P < 0.05). (B-C) Same as above except that experiment included (B) DMSO, PSI and Lactacystin (LC)-treated cells, or (C) myc-tagged UbcH7 or H6 as controls. (D) Same as (A) except that FLAG-tagged parkin was substituted with HA-tagged HHARI and anti-FLAG immunoblotting was replaced by anti-HA immunoblotting. Asterisk denotes non-specific bands. (E) Same as (A) except that FLAG-tagged parkin was substituted with myc-tagged CHIP and anti-FLAG immunoblotting was replaced by anti-myc immunoblotting. Immunoprecipitation was carried out with anti-Ubc13. These experiments were replicated at least three times.

https://doi.org/10.1371/journal.pone.0073235.g002

In view of the recent demonstration by Chaugule and colleagues that parkin activity is normally repressed by its ubiquitin-like (Ubl) domain (which can be disrupted by N-terminal epitope tagging) [23], we examined whether proteasome inhibition can relieve the repression to promote the interaction between untagged full length parkin and Ubc13 and concomitantly enhance K63-linked ubiquitination. Interestingly, we found that untagged parkin also exhibits a greater affinity for Ubc13 under conditions of proteasome inhibition (Figure 3A) and that the levels of K63 polyubiquitin-modified proteins are specifically enhanced in the pellet fraction in untagged parkin-transfected cells in the presence of MG132 treatment (Figure 3B). Thus, proteasome inhibition results in an increased association between Ubc13 and FLAG-tagged or untagged parkin species, both events leading to enhanced K63-linked ubiquitination.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. Proteasome inhibition promotes the interaction between untagged parkin and Ubc13 and concomitantly enhances K63-linked ubiquitination.

(A) A portion of Triton-X-soluble lysates prepared from untreated or MG132-treated HEK293 cells expressing untagged parkin alone or with myc-tagged Ubc13 were subjected to anti-myc immunoprecipitation followed by anti-parkin and anti-myc immunoblotting (IPmyc). The remainder lysates prepared from these variously transfected cells (INPUT) were also subjected to anti-parkin and anti-myc immunoblotting to show the expression levels of parkin and myc-Ubc13 respectively. (B) Representative anti-HA and anti-K63 ubiquitin immunoblots of cell extracts sequentially prepared with Triton-X 100 (S) and SDS (P)-containing buffer from untreated or MG132-treated HEK cells transfected with HA-tagged ubiquitin, and vector or untagged parkin, as indicated.

https://doi.org/10.1371/journal.pone.0073235.g003

Notwithstanding the above, how proteasome inhibition increases the affinity between parkin and Ubc13 remains unclear, although a recent study by Sha and colleagues have demonstrated that parkin phosphorylation by PINK1 promotes its interaction with Ubc13/Uev1a and concomitantly activates its K63-linked ubiquitination activity [24]. We therefore wondered whether parkin is similarly phosphorylated under conditions of proteasome impairment. To address this, we immunoprecipitated parkin from transfected cells in the presence or absence of MG132 and examined its phosphorylation status by immunoblotting with anti-phosphoserine and anti-phosphothreonine antibodies [similar to the ones described by Sha et al [24]]. However, we failed to detect any evidence of parkin phosphorylation in the absence or presence of proteasome inhibition, or even in the presence of PINK1 co-expression, which we have included as a control (Figure S3A & B). As the detection of protein serine/threonine phosphorylation using phospho-specific antibodies is well known to be tricky, we also analyzed the electrophoretic mobility of parkin prepared from untreated or MG132-treated cells but found no alteration to suggest that parkin is phosphorylated in the presence of proteasome inhibition (Figure S3C). Thus, parkin phosphorylation does not appear to be responsible for its increased affinity for Ubc13 under conditions of proteasomal stress.

Synphilin-1 and mutant DJ1 over-expression further promotes parkin-Ubc13 interaction

Previously, we have demonstrated that synphilin-1 ubiquitination by parkin and Siah-1 occurs via K63 and K48 respectively [11]. Consistent with this, and with our results above, we found that parkin, but not Siah-1 or several other related E3 members, promotes synphilin-1 accumulation in the Triton-X-insoluble P fraction (Figure S4A & B). Not surprisingly, this phenomenon could be mimicked by co-expressing synphilin-1 with K63 ubiquitin mutant, although the lysineless K0 mutant also triggers similar outcome (Figure S4C & D). Given the recent finding by Winklhofer’s group that parkin is capable of mediating linear ubiquitin chain assembly [25], there is a possibility that the K0 mutant could support linear ubiquitination of synphilin-1 in the presence of parkin thereby leading to the stabilization of the protein. As linear and K63 ubiquitin chains are structurally quite similar, we tested to see if the K63 antibody might recognize K0-ubiquitinated proteins but found no evidence of their cross-reactivity (Figure S5). Notwithstanding this, given that synphilin-1 is ubiquitinated by parkin via K63, we reasoned that synphilin-1 over-expression in cells treated with MG132 might further promote the recruitment of Ubc13 by parkin. Indeed, we found that parkin-Ubc13 interaction occurs significantly more strongly in the presence of synphilin-1 over-expression (Figure 4A). This phenomenon is however dependent on proteasome inhibition as the enhancement of parkin and Ubc13 interaction in the presence of synphilin-1 did not take place in untreated transfected cells (Figure 4A). Next, we examined parkin-Ubc13 interaction in MG132-treated cells over-expressing DJ-1 L166P mutant. Based on studies involving ubiquitin mutant over-expression, Olzmann et al has shown that under conditions of proteasome impairment, parkin-mediated ubiquitination of DJ-1 L166P mutant also occurs via K63 [16]. Similar to synphilin-1, we found that over-expression of DJ-1 L166P in cells significantly enhanced the binding between Ubc13 and parkin in the presence but not absence of MG132 (Figure 4B). Our results thus suggest that under conditions of proteasome impairment, parkin preferentially recruits Ubc13 to mediate K63-linked ubiquitination on selected substrates. To support this further, we repeated our experiments with mitofusin 2 (Mfn2), a substrate of parkin whose degradation is accelerated in the presence of the E3 [26], suggesting that parkin-mediated ubiquitination of Mfn2 is unlikely to be K63-linked. In this case, we failed to observe any enhancement in parkin-Ubc13 interaction in the absence or presence of proteasome inhibition (Figure 4C). However, a trivial explanation for this is that the availability of parkin is dramatically reduced in the presence of Mfn2, which is rather curious. Indeed, Mfn2 over-expression appears to promote parkin degradation in our hands, which is mitigated in the presence of MG132 (Figure 4C). Although we remain intrigued by this observation, it is clear that the added enhancement of parkin-Ubc13 interaction that we have observed with synphilin-1 and DJ1L166P in the presence of proteasome inhibition does not apply to all parkin substrates.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 4. Effects of substrate overloading on parkin-Ubc13 association.

(A) A portion of Triton-X-soluble lysates prepared from untreated and MG132-treated HEK293 cells expressing FLAG-tagged parkin alone or with myc-tagged Ubc13 and/or HA-synphilin were subjected to anti-myc immunoprecipitation followed by anti-FLAG, anti-myc and anti-HA immunoblotting (IPmyc). The expression levels of the transfected cDNAs are shown in INPUT blots. (B) As in (A) except that HA-synphilin is replaced by myc-DJ1 mutant and that immunoprecipitation was carried out using anti-Ubc13 antibody. (C) As in (A) except that HA-synphilin is replaced by YFP-Mfn2.

https://doi.org/10.1371/journal.pone.0073235.g004

Parkin facilitates the autophagic clearance of synphilin-1 inclusions formed in the presence of proteasome inhibition

Given the recent finding by our group and others that K63 polyubiquitin may act as a signal to target proteins to the aggresome-autophagy pathway [6,16], it is tempting to speculate that the autophagic clearance of synphilin-1 inclusions formed under conditions of proteasome inhibition may be facilitated by parkin. By means of an inclusion clearance assay described previously [6], we found that exogenously-introduced parkin, but not Siah-1, promotes the autophagic clearance of synphilin-1 inclusions generated in cells treated with the proteasome inhibitor, lactacystin (Figure 5A). Corroborating with this, we further found that whereas synphilin-1 inclusions generated in cells co-expressing Ubc13 and Uev1a are amenable to clearance by autophagy, those formed in cells expressing UbcH7 are apparently resistant to autophagy-mediated clearance (Figure 5B). Thus, the enhanced recruitment of Ubc13 by parkin in the presence of synphilin-1 over-expression and proteasome inhibition appears to be a cellular response in favour of alternate clearance of this parkin substrate via the autophagy route.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 5. Autophagy clearance of synphilin-1-positive inclusions is enhanced in the presence of parkin or Ubc13 overexpression.

(A) Bar graph showing the relative fold difference in the number of synphilin-1-positive inclusions formed in lactcystin-treated cells co-expressing (a) parkin or Siah-1, or (b) Ubc13/Uev1a or UbcH7, that were allowed to recover in normal (no autophagy induction, -) or low serum (autophagy induced, +) medium (*P < 0.05, **P < 0.001 vs control group).

https://doi.org/10.1371/journal.pone.0073235.g005

K63-polyubiquitination protects cells against proteasome inhibition-induced cytotoxicity

We surmised that the upregulation of K63-linked ubiquitination in the presence of proteasome inhibition may be a cellular protective response. To address this, we subjected wild type and Ubc13-/- MEFs to MG132 treatment to examine their relative susceptibility to proteasome inhibition-induced cell death. As per our speculated protective role of K63-linked ubiquitination, we observed that a significant population of the MG132-treated Ubc13-/- MEFs became rounded and reflective, i.e. indicative of dying cells whereas wild type MEFs treated with MG132 were relatively spared of these features (Figure 6A). Quantitative measurement of cell death at this time point by means of flow cytometry correlates well with our morphological observations, which we further showed is related to proteasome inhibition (Figure 6B). Moreover, immunoblotting of lysates prepared from these cells revealed a dose-dependent enhancement in the levels of two cell death markers, i.e. cleaved caspase 3 and cleaved PARP in Ubc13-/- MEFs in the presence of MG132-mediated proteasome inhibition relative to their untreated or wild type MEF counterparts (Figure 6C). Together, these results suggest that the inability of Ubc13-/- MEFs to promote K63-linked ubiquitination under conditions of proteasome impairment renders them vulnerable to cell death. Consistent with this, we detected significantly reduced levels (although curiously not complete absence) of K63-linked ubiquitination in Ubc13-/- MEFs in the presence of MG132 treatment, the levels of which comparatively increased robustly in a dose-dependent manner in wild type MEFs in response to MG132-mediated proteasome inhibition (Figure 6D). Importantly, we further demonstrated that MG132-induced cell death in Ubc13-/- MEFs can be rescued by the ectopic expression of exogenous Ubc13, as evident by the reduced levels of cleaved caspase 3 and cleaved PARP in these cells compared to un-transduced Ubc13-/- MEFs (Figure 6E). Conversely, the introduction of UbcH7 into Ubc13-/- MEFs appears to aggravate the extent of MG132-induced cell death in these cells instead of rescuing them (Figure 6E). Thus, the upregulation of Ubc13-mediated K63-linked ubiquitination under conditions of proteasome impairment appears to fulfil a protective role.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 6. Ubc13 null MEFs are vulnerable to proteasome inhibition-induced cytotoxicity.

(A) Representative phase-contrast images showing an obvious increase in the population of rounded, refractile cells in MG132-treated Ubc13 -/- MEFs relative to their untreated or MG132-treated wild type counterparts. (B) The cell viability (left) (Fold difference between MG132-treated vs. untreated cells) and proteasome activity (right) of wild type and Ubc13 -/- in the absence or presence of MG132 treatment (16h) were measured and plotted as a bar graph respectively. (C) Immunoblots showing the levels of cleaved caspase 3 and cleaved PARP in lysates prepared from wild type and Ubc13 -/- MEFs in the absence or presence of MG132 treatment (1 or 2 µM, as indicated). (D) Same as (C) as anti-K63 antibodies were used. (E) Immunoblots showing the levels of cleaved caspase 3 and cleaved PARP in lysates prepared from Ubc13 -/- MEFs transduced with lenti-Ubc13 or UbcH7, as indicated, in the absence or presence of 2 µM MG132.

https://doi.org/10.1371/journal.pone.0073235.g006