Biochimica et Biophysica Acta 1823 (2012) 1958–1966

Contents lists available at SciVerse ScienceDirect

Biochimica et Biophysica Acta

j ourna l homepage: www.e lsev ie r .com/ locate /bbamcr

Heat shock induces a massive but differential inactivation of SUMO-specific proteases

Manuel P. Pinto a,1, Andreia F. Carvalho a,1, Cláudia P. Grou a, José E. Rodríguez-Borges b,Clara Sá-Miranda a, Jorge E. Azevedo a,c,⁎a Instituto de Biologia Molecular e Celular (IBMC), Universidade do Porto, Rua do Campo Alegre, 823, 4150‐180 Porto, Portugalb CIQ-Departamento de Química e Bioquímica, Faculdade de Ciências, Universidade de Porto, Rua do Campo Alegre, 687, 4169‐007 Porto, Portugalc Instituto de Ciências Biomédicas Abel Salazar (ICBAS), Universidade do Porto, Largo Professor Abel Salazar, 2, 4009‐003 Porto, Portugal

Abbreviations: AMC, 7-amino-4-methyl coumarin;affinity chromatography; NEM, N-ethylmaleimide; SEcSENP, SENP catalytic domain; SUMO, small ubiquitinVME, vinyl methyl ester⁎ Corresponding author at: Instituto de BiologiaMolecula

do Porto, Rua do Campo Alegre, 823, 4150‐180 Porto, Pofax: +351 226099157.

E-mail address: jazevedo@ibmc.up.pt (J.E. Azevedo).1 These authors contributed equally to this work.

0167-4889/$ – see front matter © 2012 Elsevier B.V. Alhttp://dx.doi.org/10.1016/j.bbamcr.2012.07.010

a b s t r a c t

a r t i c l e i n f o

Article history:Received 23 March 2012Received in revised form 12 July 2012Accepted 23 July 2012Available online 31 July 2012

Keywords:SUMO-specific proteaseSENPSUMOHeat shockActivity-based probe

Covalent conjugation of the small ubiquitin-like modifier (SUMO) to proteins is a highly dynamic and revers-ible process. Cells maintain a fine-tuned balance between SUMO conjugation and deconjugation. In responseto stress stimuli such as heat shock, this balance is altered resulting in a dramatic increase in the levels ofSUMO conjugates. Whether this reflects an activation of the conjugation cascade, a decrease in the activityof SUMO-specific proteases (SENPs), or both, remains unknown. Here, we show that from the five humanSENPs detected in HeLa cells (SENP1/2/3/6/7) the activities of all but one (SENP6) were largely diminishedafter 30 min of heat shock. The decreased activity is not due to changes in their steady-state levels. Rather,in vitro experiments suggest that these SENPs are intrinsically heat-sensitive, a property most likely emergingfrom their catalytic domains. Heat shock inactivation seems to be a specific property of SENPs because nu-merous members of the related deubiquitinase family of cysteine proteases are not affected by this stresscondition. Overall, our results suggest that SENPs are particularly sensitive to heat shock, a property thatmay be important for the adaptation of cells to this stress condition.

© 2012 Elsevier B.V. All rights reserved.

1. Introduction

Covalent modification of proteins with the small ubiquitin-likemodifier (SUMO) has been implicated in a wide variety of biologicalprocesses such as DNA repair, transcription regulation, chromosomestructure and function, and signal transduction pathways (reviewedin [1–3]). SUMO is ubiquitous in the eukaryotic kingdom [4–8].Lower eukaryotes, such as yeasts and fungi, express only one SUMOwhereas higher organisms can have several paralogs (e.g., eight inArabidopsis thaliana [9]). Mammals express three of these proteins—SUMO1/2/3 [5–7] which can be divided into two structural and func-tional groups. One group comprises SUMO1 alone. The other encom-passes the almost identical and functionally indistinguishable SUMO2and SUMO3 [10,11].

Conjugation of SUMO to proteins involves an enzymatic cascadecomprising a SUMO-activating enzyme (E1) [12], a SUMO-conjugatingenzyme (UBC9) [13] and, at least in some cases, a SUMO ligase (E3;

IMAC, immobilized metal ionNP, SUMO-specific protease;-like modifier; Ub, ubiquitin;

r e Celular (IBMC), Universidadertugal. Tel.: +351 226074900;

l rights reserved.

e.g., PIAS, RanBP2, TOPORS andPc2 [14–17]). Similarly to ubiquitination,proteins can be modified with a single SUMO or with a chain of SUMOmolecules, modifications that probably have different biological out-comes [1,11,18].

Protein SUMOylation is a highly dynamic and reversible process.Reversibility is ensured by a small family of SUMO-specific proteases,the so-called SENPs. SENPs are cysteine proteases containing a typicalHis-Asp-Cys catalytic triad [19]. They are involved in processing precur-sor forms of SUMO (endopeptidase activity) and in removing SUMO orSUMO chains frommodified proteins (isopeptidase activity). Mammalspossess six of these proteases (SENP1–3 and 5–7), which can be phylo-genetically organized into three groups [19–21]. The first group com-prises SENP1 and SENP2, two proteases with a robust endopeptidaseactivity, although they also display some isopeptidase activity [22–27].SENP3 and SENP5 belong to the second group and display mainlyisopeptidase activity [28–30]. The last group comprises SENP6 andSENP7, two proteins that possess a strong isopeptidase activity onpolySUMO chains [31–34]. With the exception of SENP1, all otherSENPs seem to display a high paralog preference toward SUMO2/3over SUMO1 [3,35,36].

Cells maintain a fine-tuned balance between SUMO conjugation anddeconjugation. Under normal physiological conditions most SUMO1 isfound in a conjugated formwhereas SUMO2/3 aremostly unconjugated[10]. Interestingly, the SUMO conjugation/deconjugation balance isaltered in several human diseases including cancer, neurodegenerativedisorders and heart disease, a finding that has fueled intensive research

1959M.P. Pinto et al. / Biochimica et Biophysica Acta 1823 (2012) 1958–1966

in this area (reviewed in [37] and [38]). A similar phenomenon occurswhen cells are exposed to stress stimuli such as heat shock, oxidativestress or alcohol treatment [10,39,40]. Under these conditions there isa rapid shift of most SUMO2/3 from the unconjugated to the conjugatedform [10,41]. The system-wide SUMOylation alterations induced bythese stimuli are particularly well characterized in the case of heatshock. Indeed, hundreds of proteins, many of which are involved instress resistance and damage repair, have been shown to be modifiedby SUMO2/3 in response to this stimulus [42,43]. The ultimate resultof such a modification is probably several-fold. For some proteinspolySUMOylation may promote polyubiquitination by the SUMO-targeted ubiquitin ligase RNF4 leading to their subsequent degradationby the proteasome (e.g., PARP-1 [44]). In other cases, SUMOylationmayregulate protein activity/localization by creating or destroying proteininteraction interfaces [1,3,45,46]. Another possible outcome ofSUMOylation may be just an increase in the solubility of the targetedprotein [47–50].

Themechanisms behind themassive heat shock-induced accumula-tion of SUMOylated proteins remain unknown. In principle, such a phe-nomenon could result from an activation of the SUMO-conjugationcascade, a decrease in the activities of SENPs, or a combination ofboth. In this work we addressed the possible contribution of SENPs forthis phenomenon. We found that the majority of human SENPs areinactivated upon heat shock.

2. Materials and methods

2.1. Plasmids and protein production

The bacterial expression plasmids encoding SUMO2 (UniProt acces-sion number P61956) and the SUMO-conjugation cascade componentsSAE1/SAE2 (E1) and Ubc9were kindly provided by Dr. FraukeMelchior[51,52] and were used to obtain the respective recombinant proteins asdescribed [53].

Plasmids encoding untagged SUMO1or SUMO2 fused to intein and achitin-binding domain was obtained by replacing the hemagglutinin–ubiquitin encoding sequence (HA-Ub) in the pTYB-HAUb plasmid (akind gift from Dr. Hidde Ploegh) [54] with one encoding SUMO1(obtained from Mr. Gene) or SUMO2. A plasmid encoding untaggedubiquitin, pTYB-Ub, was generated by digestion of pTYB-HAUb withNdeI to remove the sequence coding for HA, followed by vectorreligation. The pTYB-SUMO1/SUMO2/Ub constructs were used to pre-pare untagged SUMO1/SUMO2/Ub-aldehyde according to [54,55].

Plasmids pTYB-HA-SUMO1/SUMO2 were obtained by inserting anHA encoding sequence into the NdeI site of pTYB-SUMO1/SUMO2constructs, respectively. The pTYB-HA-SUMO1/SUMO2/Ub plasmidswere used to produce HA-tagged SUMO1/SUMO2/Ub-vinyl methylester (VME), as described [54], except that the ion-exchange purifica-tion step was omitted.

The cDNAs encoding SENP1 and SENP3 were amplified by RT-PCRusing total RNA from human fibroblasts and cloned into the NheI/EcoRIand NdeI/BamHI sites, respectively, of pET-28a vector. The SENP2 cDNAwas amplified using a pET-SENP2 plasmid as a template (a kind giftfrom Dr. Keith Wilkinson) [35] and inserted into the NdeI/EcoRI sites ofthe pET-28a vector. The cDNAs encoding SENP5/6/7 were amplifiedusing as templates pDONR221-SENP5 (PlasmID repository; clone IDHsCD00039689), pENTR223-SENP6 (PlasmID repository; clone IDHsCD00378900) and pCR-BluntII-TOPO-SENP7 (PlasmID repository;clone ID HsCD00347824), respectively. The amplified DNA fragmentswere inserted into the NdeI/XhoI (SENP5 and SENP7) or NheI/BamHI(SENP6) sites of the pET-28a vector. 35S-labeled full-length SENP1–3/5–7 were synthesized using these pET-28a-based plasmids and theTNT® T7 Quick Coupled Transcription/Translation System (Promega) inthe presence of EasyTag™ L-[35S]-Methionine (Perkin Elmer, specificactivity >1000 Ci (37.0 TBq)/mmol), following the manufacturer'sinstructions.

Bacterial expression vectors encoding the catalytic domains ofSENP1/2/5/6/7 (cSENPs) were kindly provided by Dr. Guy Salvesen,and were used to obtain the respective recombinant proteins, asdescribed [24]. For the SYPRO Orange fluorimetric assays (seebelow), cSENPs were purified using a two-step procedure. In thefirst step, cSENP1/2 were subjected to immobilized metal ion affinitychromatography (IMAC) purification using HIS-Select™ nickel affini-ty gel (Sigma) and eluted in 50 mM Tris–HCl, pH 8.0, 150 mM NaCl,10% (w/v) glycerol and 200 mM imidazole. Recombinant cSENP5–7were was purified by IMAC using a TALON metal affinity resin(Clontech) and eluted with 50 mM Tris–HCl, pH 8.0, 350 mM NaCl,10% (w/v) glycerol and 150 mM imidazole. In the second step,cSENP1/2/5 were dialyzed against 50 mM Tris–HCl, pH 8.0, 150 mMNaCl, 10% (w/v) glycerol and 0.5 mM DTT and further purified bysize-exclusion chromatography using a Superose 12 column HR 10/30(GE Healthcare) running with the same buffer. IMAC-purified cSENP6was diluted three-fold in 50 mM Tris–HCl, pH 8.0, 10% (w/v) glyceroland 0.5 mM DTT, applied to a 1-ml HiTrap Q HP (GE Healthcare)anion exchange column, and eluted with a 15-ml linear gradient of150–575 mM NaCl in 50 mM Tris–HCl, pH 8.0, 10% (w/v) glycerol and0.5 mM DTT. IMAC-purified cSENP7 was dialyzed against 50 mM Tris–HCl, pH 8.0, 50 mM NaCl, 10% (w/v) glycerol and 0.5 mM DTT and pu-rified by anion exchange chromatography as above but using a50–525 mM NaCl linear gradient instead. Purified cSENPs werealiquoted, frozen in liquid N2 and stored at−80 °C.

2.2. Cell culture

HeLa cells were grown at 37 °C in the presence of 5% CO2 in a stan-dard Dulbecco's modified Eagle's medium supplemented with 10%heat-inactivated fetal bovine serum, 1% penicillin/streptomycin, 1%L-glutamine, 1% fungyzone and 1% kanamycin. In heat shock experi-ments, cells at 80% confluence were harvested with a cell scraper,resuspended in supplemented medium and incubated at 37 °C or43 °C in awater bath for 30 min, unless otherwise indicated. In recoveryexperiments, heat-shocked cells were transferred to Petri dishes andallowed to recover at 37 °C in a CO2 incubator for the indicated periodsof time, either in the absence or presence of cycloheximide and puromy-cin (30 μg/ml each). To prepare total cell lysates, cells were collected,washed three times with phosphate-buffered saline, and centrifugedat 2000×g for 10 min at 4 °C. Pellets were resuspended in SEM buffer(0.25 M sucrose, 20 mM MOPS-KOH, pH 7.4, 1 mM EDTA-NaOH, pH8.0) supplemented with protease inhibitors cocktail (Sigma) andcontaining either 1 mMDTT or 10 mMN-ethylmaleimide (NEM), as in-dicated. Total cell lysates were prepared by sonication, aliquoted, frozenin liquid N2 and stored at −80 °C.

2.3. 7-Amino-4-methyl coumarin (AMC) activity assays

Determination of SENP activity in total HeLa cell homogenatesusing SUMO1-AMC and SUMO2-AMC substrates (Enzo Life Sciences)were performed essentially as described [35,56]. Briefly, 50 μg of ex-tracts from control or heat-shocked HeLa cells were incubated with500 nM of AMC substrates in 140 μl of assay buffer (50 mM Tris–HCl, pH 7.5, 0.1 mg/ml bovine serum albumin and 10 mM DTT). Hy-drolysis of the substrates was measured by determining the increasein fluorescence (λex=340 nm and λem=440 nm) using a HoribaFluoroMax-4 fluorimeter.

2.4. SENP activity assays using SUMOylated rat liver nuclei proteins

Highly pure nuclei were prepared from rat liver total homoge-nates by the method of Blobel and Potter, essentially as described[57–59]. SUMOylated substrates were prepared by incubating100 μg of purified nuclei with 50 nM SUMO E1, 100 nM Ubc9, 1 μMSUMO2 and 2 mM ATP for 30 min at 43 °C. The SUMO-conjugation

Fig. 1. SENP activity in HeLa cells decreases upon heat shock. A) Total homogenatesfrom HeLa cells grown at 37 °C or heat-shocked at 43 °C for 30 min were incubatedwith either SUMO1-AMC (S1-AMC) or SUMO2-AMC (S2-AMC). Activities were deter-mined fluorimetrically and are presented in the table (n=3). The graphic showsSUMO1/2-AMC hydrolysis catalyzed by control HeLa cell extracts (37 °C; black lines),and by heat-shocked HeLa cell extracts (43 °C; grey lines). B) Rat liver nuclei proteinsSUMO2ylated in vitro (see Section 2. Materials and methods; lane –, 100% input)were incubated at 37 °C with different amounts of total homogenates from HeLa cells(100, 200 or 400 μg) grown at 37 °C or heat-shocked at 43 °C, as indicated, andprocessed for SDS-PAGE/Western blot using an anti-SUMO2/3 antibody. Positive(lane cSP1; containing 250 ng of cSENP1) and negative control reactions (lane S2ald;containing SUMO2-aldehyde and 400 μg of HeLa extracts) were also analyzed. Numbersat the bottom of the upper panel represent relative amounts of SUMO2 conjugatesremaining at the end of the incubation, as determined by densitometric analysis oftotal lanes. The amount of conjugates present in lane – was set to 100%. A section ofthe Ponceau S-stained membrane is presented in the lower panel. Numbers to the leftindicate the molecular masses of protein standards in kilodaltons.

1960 M.P. Pinto et al. / Biochimica et Biophysica Acta 1823 (2012) 1958–1966

machinery as well as endogenous nuclei SENPs was inactivated bytreatment with 5 mM iodoacetamide and the excess of alkylatingagent was then neutralized with 10 mM DTT. The in vitroSUMOylated proteins were then incubated with total extracts fromcontrol or heat-shocked HeLa cells for 30 min at 37 °C in the presenceof 2 mM AMP and 0.15 mM sodium pyrophosphate to prevent any denovo SUMOylation. Where indicated, reactions also received 250 ngof cSENP1 or 0.5 μM SUMO2-aldehyde. All reactions were stoppedwith 10 mM NEM, incubated for 5 min on ice and analyzed bySDS-PAGE/Western blot using an anti-SUMO2/3 antibody [60].

2.5. Covalent labeling with HA-tagged SUMO/Ub-VME probes

Labeling reactions were performed at 25 °C for 15 min with 200 μgof HeLa cell total extracts, or 0.5 μl of 35S-labeled SENPs, or 50 ng of therecombinant catalytic domains in 20 μl of SEM buffer supplementedwith 1 mM DTT, in the presence of 100 ng of HA-SUMO1/SUMO2/Ub-VME. Some reactions were preincubated with 500 ng of the respec-tive aldehyde analogue for 5 min at 25 °C before adding the VMEprobes. Where indicated, samples were treated with 60 ng of cSENP1for 15 min at 25 °C before addition of the VME probe. Reactions wereterminated with Laemmli sample buffer.

2.6. Antibodies

Rabbit polyclonal antibodies against SENPs were kindly provided byDr. Mary Dasso [33,35,61]. The mouse monoclonal anti-SUMO1 (21C7)[7] and anti-SUMO2/3 (8A2) [60] antibodies developed by Michael J.Matunis were obtained from the Developmental Studies HybridomaBank developed under the auspices of the NICHD and maintained bythe University of Iowa, Department of Biology, Iowa City, IA 52242.Mouse monoclonal anti-HA (16B12) and anti-His were purchased fromCovance and GE Healthcare, respectively. Primary antibodies weredetected on Western blots with alkaline phosphatase-conjugatedanti-rabbit or anti-mouse IgGs (Sigma). Densitometric analyses of West-ern blots were performed using the ImageJ software (Rasband, W.S.,ImageJ, U. S. National Institutes of Health, Bethesda, Maryland, USA,http://imagej.nih.gov/ij/, 1997–2011).

2.7. Fluorescence-based denaturation assays

Denaturation assays were performed in 25 μl of 50 mM Tris–HCl,pH 8.0, 150 mM NaCl, 10% (w/v) glycerol and 10 mM DTT, containing1 μM cSENP1/2/5–7 and 5× SYPRO Orange dye (Sigma, stock solution5000×) in a iQ5 Real Time Detection System (Bio-Rad) using 96 wellPCR plates. For differential scanning fluorimetry, the samples wereheated from 20 to 85 °C with a ramping rate of 1 °C/min and fluores-cence monitoring every 0.5 °C. The scan curves were fitted to aBoltzmann sigmoid function using GraphPad Prism 5 software andthe apparent melting temperatures (Tm) were obtained from the tem-perature midpoint of the protein unfolding transition, as described[62,63]. To monitor cSENP denaturation at 43 °C over time, the plateswere first heated for 5 min at 37 °C, followed by 30 min at 43 °C and5 min at 70 ºC, with fluorescence reading every 30 s. All experimentswere performed at least in triplicate.

3. Results and discussion

To address the possible contribution of SENPs to the heatshock-induced accumulation of SUMO conjugates, we determinedthe activity of these proteases in total homogenates from HeLa cellssubjected or not to heat shock treatment. Three different assayswere used for this purpose. In the first, we measured total SENP activ-ity using two fluorogenic substrates, SUMO1-AMC and SUMO2-AMC[35,55]. These SUMO1/2 derivatives carry an AMC moiety at their Cterminus which fluoresces upon hydrolysis by SENPs. As shown in

Fig. 1A, homogenates from cells treated at 43 °C for 30 min display62 and 29% of the activity found in control cells when assayed withSUMO1-AMC and SUMO2-AMC, respectively.

In the second assay, we incubated HeLa cell homogenates with invitro synthesized SUMO2 conjugates (see Section 2 Materials andmethods for details) and monitored the decrease of these conjugatesby SDS-PAGE/Western blot using an anti-SUMO2/3 antibody [60]. Theresult obtained, although of qualitative nature, suggests that SENP ac-tivity is decreased by heat shock treatment (see Fig. 1B).

Finally, the levels of active SENPs in heat-shocked and control cellswere assessed by incubating total cell homogenates with VME-derivatized HA-tagged SUMO1 or SUMO2 [55]. These activity-based

1961M.P. Pinto et al. / Biochimica et Biophysica Acta 1823 (2012) 1958–1966

SUMO probes react with the catalytic cysteines of SENPs yielding co-valently linked SUMO-SENP products. These products can then bedetected by SDS-PAGE/Western blot using an anti-HA antibody. Asshown in Fig. 2A, several protein bands were detected in homoge-nates labeled with the HA-SUMO1/2-VME probes (asterisks in lanes2 and 8, respectively). None of these bands was detected when the la-beling reactions were performed in the presence of a vast excess ofSUMO1-aldehyde (lane 3) or SUMO2-aldehyde (lane 9), two otherpotent inhibitors of SENPs that compete with the VME probes forSENPs catalytic domains [55]. Thus, the species detected in these ex-periments represent active SENPs. In agreement with the results ofthe two assays described above, the intensities of several HA-SUMO1/2-VME-labeled SENPs were decreased in homogenates fromHeLa cells heat-shocked for 30 min (Fig. 2A; compare lanes 2 and8with lanes 5 and 11, respectively). Attempts to recover SENP activityby incubating cell homogenates for 1 h at 25 °C or 37 °C before label-ing reactions with the SUMO-VME probes were not successful (datanot shown), suggesting that heat shock-induced inactivation ofSENPs is an irreversible process (see also below). Importantly, treat-ment of homogenates from heat-shocked cells with the recombinant

Fig. 2. HA-SUMO1/2-VME labeling of SENPs in HeLa cells. A) Total extracts from HeLa cellHA-SUMO2-VME and analyzed by SDS-PAGE/Western blot using the anti-HA antibody toSome samples were pre-incubated with an excess of SUMO1-aldehyde or SUMO2-aldehydconjugates. B) Total homogenates from HeLa cells heat-shocked for 30 min were incubatehalf was treated with HA-SUMO1-VME whereas the other was treated with HA-SUMO2-anti-HA antibody to detect active SENPs (upper panel) or using anti-SUMO1 (anti-S1) or aasterisk indicates the cSENP1-SUMO1/2-VME adducts. C) Same as in A but using HA-Ub-VMmolecular masses of protein standards in kilodaltons.

catalytic domain of SENP1 (cSENP1) prior to incubation withHA-SUMO1/2-VME leads to the disappearance of all SUMO2/3 conju-gates, and nearly all SUMO1 conjugates (Fig. 2B, lower panel), asexpected [64], but does not result in an increase in the labeling inten-sity of SENPs by the VME probes (Fig. 2B, upper panel). Thus, the heatshock-induced decrease in SENPs activity, as assessed with the VMEprobes, is not due to a competition effect exerted by endogenousSUMO protein conjugates. Taken together, these data suggest that a30 min heat shock decreases the activity of several SENPs.

Interestingly, treatment of total cell homogenates with HA-taggedubiquitin-VME (HA-Ub-VME), an activity-based probe that reactswith deubiquitinases belonging to the cysteine protease family [54]did not reveal significant differences between control and heat-shocked cells (see Fig. 2C). This finding suggests that the SENP sub-family of cysteine proteases is particularly enriched in heat shock-sensitive members.

Heat shock induced accumulation of SUMO2/3 conjugates is aquite fast event, reaching a maximum after just 5 min of hyperther-mic stress [10,43]. In contrast, the decrease in the levels of these con-jugates in HeLa cells that were heat-shocked at 43 °C for 30 min and

s subjected or not to heat shock for 30 min were incubated with HA-SUMO1-VME ordetect labeled SENPs. A portion of the Ponceau S-stained membrane is also shown.e before adding the VME probe, as indicated. The asterisks mark HA-SUMO1/2-SENPd in the absence (−) or presence (+) of cSENP1. The samples were halved and oneVME. Aliquots of these samples were subjected to SDS-PAGE/Western blot using annti-SUMO2/3 (anti-S2/3) to detect endogenous SUMO conjugates (lower panels). TheE and Ub-aldehyde instead of the SUMO derivatives. Numbers to the left indicate the

1962 M.P. Pinto et al. / Biochimica et Biophysica Acta 1823 (2012) 1958–1966

then allowed to recover at 37 °C is a slow process. As shown recently,only after 2 h of incubation at 37 °C do the levels of SUMO2/3 conju-gates return to their pre-heat shock values [43]. To assess whetherSENP activity is somehow correlated with the kinetics of these twoevents we monitored the levels of SUMO conjugates and active SENPsat different time points during the heat shock and recovery incubations.As shown in Fig. 3A, the massive accumulation of SUMO2/3 conjugatesobserved after just 5 min of heat shock is accompanied by a small, but re-producible, reduction (10–15%, range of three determinations) in the

Fig. 3. Levels of active SENPs and SUMO conjugates during heat shock and recovery at37 °C. A) HeLa cells were subjected to heat shock treatment for the indicated periods oftime. At each time point two aliquots were withdrawn which received or not 10 mMNEM (upper panels and lower panel, respectively). Total cell lysates prepared in thepresence of NEM were processed for SDS-PAGE/Western blot analyses with theanti-SUMO1 or anti-SUMO2/3 antibodies (anti-S1 or anti-S2/3, respectively; upperpanels). Samples prepared in the absence of NEM were incubated with HA-SUMO1-VME or HA-SUMO2-VME and analyzed by SDS-PAGE/Western blot using the anti-HAantibody (lower panel). B) HeLa cells were kept at 37 °C or subjected to heat shockfor 30 min at 43 °C (lanes 37 °C and 43 °C, respectively). Heat-shocked cells werethen allowed to recover at 37 °C for the indicated periods of time, either in the absence(−) or presence of cycloheximide and puromycin (CHX/P), to block protein synthesis.At each time point two aliquots were withdrawn and processed as above. SUMO2/3conjugates are shown in the upper panel. HA-SUMO2-VME-reactive SENPs are shownin the lower panel. The levels of SUMO2/3 conjugates and VME-labeled SENPs at eachtime point were determined by densitometry and are indicated at the bottom of eachpanel. The levels of SUMO2/3 conjugates in lane 43 °C and of labeled SENPs in lane37 °C were set to 100%. Numbers to the left indicate the molecular masses of proteinstandards in kilodaltons.

overall intensity of HA-SUMO1/2-VME-reactive bands. A similar inversecorrelation between the levels of SUMO2/3 conjugates and the intensi-ties of some VME-reactive bands was observed during the 37 °C recov-ery incubation. Indeed, as shown in Fig. 3B, the decrease in the levels ofSUMO2/3 conjugates during this incubation (upper panel) paralleled aslight increase in the intensity of some VME-reactive bands (lowerpanel). This increase is probably the result of protein synthesis becauseit was not observed when cells were incubated at 37 °C in the presenceof protein synthesis inhibitors. Under these conditions, considerableamounts of SUMO2/3 conjugates were still detectable after 6 h of incu-bation. Although the changes in the two parameters monitored in theseexperiments are of different magnitudes (see Section 4. Conclusion), al-together these results suggest that there is an inverse correlation be-tween SENP activity and the levels of SUMO2/3 conjugates.

The decrease in SENP activity observed upon heat shock could re-sult from a decrease in protein levels (e.g., some SENPs might be sim-ply degraded) or from a diminished specific activity. To discriminatebetween these two possibilities, total homogenates from control orheat-shocked (30 min) cells, treated or not with HA-tagged SUMO1/2-VME, were analyzed by Western blot using SENP-specific anti-bodies. We note that many of these antibodies recognize multiple pro-teins in these blots. However, comparison of the band patterns obtained

Fig. 4. Several SENPs are inactivated by heat shock. Total extracts from control (37 °C)or heat-shocked (43 °C) HeLa cells were incubated in the absence (−) or in the pres-ence of HA-tagged SUMO1-VME or SUMO2-VME, as indicated. Samples were analyzedby SDS-PAGE/Western blot using either SENP-specific antibodies (five upper panels) oranti-HA antibody (bottom panel). Unmodified SENPs (SENP) as well as their VME ad-ducts (SENP′) are indicated. Note that SENP2 displays an apparent molecular mass of50 kDa and probably corresponds to the most abundant isoform in HeLa cells, as de-scribed recently [35,68]. Densitometric analyses of Western blots probed with theanti-HA antibody revealed that SENP1 and SENP3 are inactivated upon heat shock by~60% and ~80%, respectively. Numbers to the left indicate the molecular masses of pro-tein standards in kilodaltons.

1963M.P. Pinto et al. / Biochimica et Biophysica Acta 1823 (2012) 1958–1966

with the HA-SUMO1/2-VME-treated and -untreated samples allowedus to unambiguously identify protein bands corresponding to SENP1–3/6–7. Efforts to detect SENP5 in HeLa extracts using three different an-tibodies were not successful thus far. The results of these analyses areshown in Fig. 4. For all analyzed SENPs, no significant differences intheir steady-state levels were detected between control andheat-shocked cells (Fig. 4, compare lanes 1 and 2). Thus, the heatshock-induced decrease in SENP activity probably results from a de-crease in the specific activity of these proteases. The results obtainedwith the HA-SUMO1/2-VME-treated samples corroborate this interpre-tation and allowed us to identify the heat shock-sensitive SENPs. Asshown in Fig. 4 (compare lanes 3 and 5with 4 and 6, respectively), pro-tein bands corresponding to adducts between the SUMO probes andSENP1–3/7 are either undetectable (SENP2/7) or largely decreased(SENP1, ~60% decrease; SENP3, ~80% decrease) in homogenates fromcells subjected to heat shock. The exception is SENP6, whichwas not af-fected by the heat shock treatment at all.

The data presented in Fig. 4 also allow us to draw some conclu-sions regarding SENP relative abundances in HeLa cells as well asthe reactivity of these proteases towards the HA-SUMO1/2-VMEprobes. Side-by-side alignment of the blots shown in Fig. 4, togetherwith the behavior of each SENP regarding its reactivity with the elec-trophilic probes and its sensitivity to heat shock, allowed us to corre-late each of the protein bands detected with the anti-HA antibody(bottom panel) with a specific SENP. This correlation suggests the fol-lowing order in SENP abundance: SENP3>SENP1>SENP2~SENP6.No protein bands corresponding to SENP7 and SENP5 were detectedin blots probed with the anti-HA antibody (Fig. 4, bottom panel),suggesting that these two proteins are the least abundant SENPs. Re-garding the reactivity of the different SENPs with HA-SUMO1/2-VMEprobes, and in agreement with previous data [22,26,27,29,34,35], wefound that SENP2, SENP3 and SENP7 display a bias towards reacting

Fig. 5. Heat sensitivity of in vitro synthesized SENPs and their recombinant catalytic domain43 °C. The samples were then treated with HA-SUMO1-VME or HA-SUMO2-VME for 15 mintheir adducts with the VME probes (SENP′) are indicated. Lane I, untreated 35S-labeled SENP43 °C as above and subjected to labeling reactions with HA-SUMO1-VME or HA-SUMO2-VMEcatalytic domains (cSENP) and their adducts with the VME probes (cSENP′) are indicated. Laular masses of protein standards in kilodaltons.

with the HA-SUMO2-VME probe whereas SENP1 reacts equally wellwith both HA-SUMO1/2-VME probes. Similarly to SENP1, SENP6 re-acts with both HA-SUMO1/2-VMEwith no apparent preference for ei-ther probe. This finding is in slight disagreement with previousreports showing that SENP6 reacts better with SUMO2-vinyl sulfonethan with SUMO1-vinyl sulfone [33,35]. The fact that the SUMOprobes used here contain a vinyl methyl ester, a chemical groupthat may be more reactive than the vinyl sulfone moiety [54], may ex-plain this discrepancy.

The decrease in the specific activities of SENP1–3/7 observed uponheat shockmay result from an extrinsic regulatory mechanism (e.g., in-teractionswith unknown regulatory proteins) and/ormay reflect an in-trinsic property of these proteases. Whereas addressing the firsthypothesis would require detailed knowledge on the protein interac-tion network of each SENP, we reasoned that the second hypothesiscould be easily put to the test. For this purpose, we synthesized allhuman SENPs using an in vitro transcription/translation system, andafter subjecting these proteins to a 37 °C or 43 °C incubation, weassessed their activity using the HA-SUMO1/2-VME probes. As shownin Fig. 5A, there is a good correlation between the behavior of eachSENP in this in vitro assay and that of the endogenous proteases inheat-shocked or control cells (Fig. 4). Thus, again with the exceptionof SENP6, all other SENPs, including now also SENP5, are at least partial-ly inactivated by heat treatment.

Identical experiments were performed using bacterial-expressedrecombinant proteins comprising the catalytic domains of SENP1/2/5–7 (cSENP1/2/5–7; the catalytic domain of SENP3 is presently notavailable). As shown in Fig. 5B, and in agreement with previous data[35], cSENP2/5–7 no longer display a preference for the SUMO2-based probe, reacting equally well with both SUMO1/2 probes. Asproposed previously, SENP domains not present in these recombinantproteins may play a role in SUMO selectivity [35]. Importantly, all

s. A) 35S-labeled SENPs were synthesized in vitro and incubated for 30 min at 37 °C orat 25 °C and analyzed by SDS-PAGE/Western blot/autoradiography. SENPs (SENP) and

used in each reaction. B) Recombinant His-tagged cSENPs were preincubated at 37 °C or. Samples were subjected to SDS-PAGE/Western blot using an anti-His antibody. SENPsne I, recombinant cSENPs used in each reaction. Numbers to the left indicate the molec-

1964 M.P. Pinto et al. / Biochimica et Biophysica Acta 1823 (2012) 1958–1966

these recombinant proteins but one, the catalytic domain of SENP6,are extensively (cSENP1) or totally (cSENP2/5/7) inactivated by the43 °C incubation.

Attempts to recover the activity of these recombinant proteins by in-cubating them at 25 °C or 37 °C before performing the labeling reac-tions with the SUMO-VME probes were not successful (data notshown). This finding suggests that inactivation of these proteins byheat treatment is irreversible, probably the result of a denaturationevent. To address this possibility we characterized the thermal stabilityof cSENPs using a fluorescence-based assay [62,63]. The assay exploresthe fact "that folded and unfolded proteins can be distinguishedthrough exposure to a hydrophobic fluoroprobe" [62]. The fluoroprobe,in our case the SYPRO Orange dye, binds to protein hydrophobic do-mains that become exposed during heat-induced unfolding with a con-comitant increase in its fluorescence emission. We first determined theapparentmelting temperatures of cSENPs. As shown in Fig. 6A, cSENP1/2/5/7 present quite similar apparent melting temperatures rangingfrom 44.5 to 48.5 °C. In contrast, a much higher apparent melting tem-perature was determined for cSENP6 (55.6 °C), suggesting that this cat-alytic domain is much more resistant to heat-induced denaturationthan the other cSENPs. A similar conclusion was obtained in a timecourse analysis of denaturation performed at 43 °C (Fig. 6B). In contrastto cSENP6 which was marginally affected by the 43 °C treatment, theother proteins were partially (cSENP1) or completely (cSENP2/5/7) de-natured under these conditions, in good agreement with the data

Fig. 6. Thermal unfolding of the recombinant catalytic domains of SENPs. A) Recombi-nant cSENP1/2/5–7 were analyzed by differential scanning fluorimetry in the presenceof SYPRO Orange dye. The apparent melting temperatures (Tm) obtained from the scancurves are indicated in the table. B) Recombinant cSENPs were incubated in the pres-ence of SYPRO Orange dye for 5 min at 37 °C, followed by 30 min at 43 °C and, finally,5 min at 70 °C. Fluorescence was monitored throughout the incubation time. Fluores-cence intensities were normalized to a maximum of 1 and a minimum of 0. Represen-tative curves for each cSENP are shown.

shown in Fig. 5B. Taken together these data suggest that theheat-induced inactivation of cSENP1/2/5/7 is due to an irreversible de-naturation event.

The sensitivity of SENPs to a 43 °C incubation was also observedwith endogenous HeLa SENPs using a coupled SUMOylation/deSUMOylation in vitro assay. In the experiment shown in Fig. 7A, a ho-mogenate from HeLa cells grown at 37 °C was incubated at 37 °C or43 °C for 15 min in a buffer containing recombinant SUMO2 and ATP.Robust amounts of SUMO2 conjugates were accumulated at 37 °C, butonly when SENPs were inhibited with SUMO2-aldehyde (Fig. 7A, com-pare lanes 2 and 3). Thisfinding suggests that endogenous SENPs hydro-lyze most of the SUMO2 conjugates synthesized in this assay at 37 °C. Adifferent result was obtained at 43 °C. Indeed, whereas the amounts ofSUMO2 conjugates obtained at 37 °C and 43 °C in the presence ofSUMO2-aldehydewere indistinguishable (Fig. 7A, lanes 3 and 5, respec-tively), the amounts of SUMO2 conjugates obtained in the absence ofthe SENP inhibitor were now much larger at 43 °C than at 37 °C (com-pare lanes 2 and 4). Apparently, the activity of the SENP(s) that hydro-lyzes most of the SUMO2 conjugates at 37 °C was largely decreased at43 °C. In agreement with this interpretation, analyses of HeLa cellhomogenates that were pre-treated for 15 min at 37 or 43 °C andthen incubated with the HA-SUMO1/2-VME probes revealed a signifi-cant inactivation of SENPs at 43 °C (Fig. 7B).

4. Conclusion

Mammalian cells andmany different organisms respond to severalstress stimuli by accumulating massive amounts of SUMO conjugates[9,10,39,40]. The mechanisms behind these responses are still ill-defined, but in at least two cases data suggesting the involvement ofSENPs in this phenomenon have been provided. One regards the ac-cumulation of SUMO conjugates in mammalian cells exposed to

Fig. 7. Endogenous HeLa SENPs are heat-sensitive in vitro. A) A homogenate from HeLacells grown at 37 °C (200 μg per lane) was incubated for 15 min at 37 °C or 43 °C in abuffer containing recombinant SUMO2 and ATP, either in the presence (+) or absence(−) of SUMO2-aldehyde. Samples were analyzed by SDS-PAGE/Western blot using theanti-SUMO2/3 antibody (anti-S2/3). Lane I, 200 μg of the HeLa cell lysate used in the invitro reactions. B) A homogenate from HeLa cells grown at 37 °C and incubated asabove was treated with either HA-SUMO1-VME or HA-SUMO2-VME. A Western blotusing the anti-HA antibody is presented. SENP-SUMO1/2-VME adducts are indicated(SENP′). Numbers to the left indicate the molecular masses of protein standards inkilodaltons.

1965M.P. Pinto et al. / Biochimica et Biophysica Acta 1823 (2012) 1958–1966

very high concentrations of hydrogen peroxide. Under these condi-tions SENPs are inactivated due to the formation of intra- or inter-molecular disulfide bridges [39,65,66]. The second example regardsSaccharomyces cerevisiae Ulp1, a SUMO-specific protease that is nor-mally found in the nuclear pore complex but that is targeted to andsequestered in the nucleolus in response to alcohol stress [67].

In thisworkwe addressed thepossibility that SENPs could also play arole in the heat shock-induced accumulation of SUMO2/3 protein conju-gates. The specific hypothesis tested here was that the activities of theseproteases could be decreased by this type of stress. The data obtainedindicate that the activities of at least themost abundant SENPs progres-sively decrease during hyperthermic stress, thus suggesting that theseproteases play a role in the accumulation of SUMO2/3 conjugates. Natu-rally, in the absence of data describing how tightly linked is SENP activ-ity to the homeostasis of SUMO2/3 conjugates, the magnitude of theimpact of the decreased SENP activity described in this work on thesteady-state concentration of SUMO2/3 conjugates cannot be presentlyquantified. Furthermore, the steady-state levels of SUMO2/3 conjugatesalso depend on the activity of the SUMO-conjugation cascade andwhether or not its activity changes during heat shock remains to be de-termined. Regardless, the heat shock-induced inactivation of SENPSshould minimize futile SUMO conjugation/deconjugation cycles andthus, at the very least, should provide an advantage in the adaptationof cells to this stress condition.

Acknowledgements

We are indebted to Dr. Mary Dasso (Laboratory of Gene RegulationandDevelopment,National Institute of ChildHealth andHumanDevelop-ment, National Institutes ofHealth, Bethesda,USA) for the generous gift ofanti-SENP antibodies. We would also like to thank Dr. Frauke Melchior(Zentrum für Molekulare Biologie der Universität Heidelberg (ZMBH),DKFZ-ZMBH Alliance, Heidelberg, Germany), Dr. Hidde Ploegh (White-head Institute for Biomedical Research, Massachusetts Institute of Tech-nology, Cambridge, USA), Dr. Guy Salvesen (Sanford–Burnham MedicalResearch Institute, La Jolla, USA), and Dr. Keith Wilkinson (Departmentof Biochemistry, Emory University School of Medicine, Atlanta, USA) forsharing plasmids. This work was supported by grants from Fundaçãopara a Ciência e Tecnologia and Fundo Europeu de Desenvolvimento Re-gional (FEDER) through COMPETE—Programa Operacional Factores deCompetitividade (POFC) in the context of QREN, Portugal (PEst-C/SAU/LA0002/2011).M.P.P. andC.P.G. are supported by Fundação para a Ciênciae Tecnologia, Programa Operacional Potencial Humano (POPH) doQREN and Fundo Social Europeu (FSE). A.F.C. is supported by ProgramaCiência—funded by POPH–QREN–Tipologia 4.2–Promoção do EmpregoCientífico, by Fundo Social Europeu and by national funds fromMCTES.

References

[1] R. Geiss-Friedlander, F. Melchior, Concepts in sumoylation: a decade on, Nat. Rev.Mol. Cell Biol. 8 (2007) 947–956.

[2] Y. Wang, M. Dasso, SUMOylation and deSUMOylation at a glance, J. Cell Sci. 122(2009) 4249–4252.

[3] K.A. Wilkinson, J.M. Henley, Mechanisms, regulation and consequences of proteinSUMOylation, Biochem. J. 428 (2010) 133–145.

[4] M.N. Boddy, K. Howe, L.D. Etkin, E. Solomon, P.S. Freemont, PIC 1, a novelubiquitin-like protein which interacts with the PML component of a multiproteincomplex that is disrupted in acute promyelocytic leukaemia, Oncogene 13 (1996)971–982.

[5] A. Chen, H. Mannen, S.S. Li, Characterization of mouse ubiquitin-like SMT3A andSMT3B cDNAs and gene/pseudogenes, Biochem. Mol. Biol. Int. 46 (1998)1161–1174.

[6] T. Kamitani, K. Kito, H.P. Nguyen, T. Fukuda-Kamitani, E.T. Yeh, Characterizationof a second member of the sentrin family of ubiquitin-like proteins, J. Biol.Chem. 273 (1998) 11349–11353.

[7] M.J. Matunis, E. Coutavas, G. Blobel, A novel ubiquitin-like modification modu-lates the partitioning of the Ran-GTPase-activating protein RanGAP1 betweenthe cytosol and the nuclear pore complex, J. Cell Biol. 135 (1996) 1457–1470.

[8] P.B. Meluh, D. Koshland, Evidence that the MIF2 gene of Saccharomyces cerevisiaeencodes a centromere protein with homology to the mammalian centromereprotein CENP-C, Mol. Biol. Cell 6 (1995) 793–807.

[9] J. Kurepa, J.M. Walker, J. Smalle, M.M. Gosink, S.J. Davis, T.L. Durham, D.Y. Sung,R.D. Vierstra, The small ubiquitin-like modifier (SUMO) protein modification sys-tem in Arabidopsis. Accumulation of SUMO1 and ‐2 conjugates is increased bystress, J. Biol. Chem. 278 (2003) 6862–6872.

[10] H. Saitoh, J. Hinchey, Functional heterogeneity of small ubiquitin-related proteinmodifiers SUMO-1 versus SUMO-2/3, J. Biol. Chem. 275 (2000) 6252–6258.

[11] M.H. Tatham, E. Jaffray, O.A. Vaughan, J.M. Desterro, C.H. Botting, J.H. Naismith,R.T. Hay, Polymeric chains of SUMO-2 and SUMO-3 are conjugated to protein sub-strates by SAE1/SAE2 and Ubc9, J. Biol. Chem. 276 (2001) 35368–35374.

[12] J.M. Desterro, M.S. Rodriguez, G.D. Kemp, R.T. Hay, Identification of the enzymerequired for activation of the small ubiquitin-like protein SUMO-1, J. Biol. Chem.274 (1999) 10618–10624.

[13] S.E. Schwarz, K. Matuschewski, D. Liakopoulos, M. Scheffner, S. Jentsch, Theubiquitin-like proteins SMT3 and SUMO-1 are conjugated by the UBC9 E2 en-zyme, Proc. Natl. Acad. Sci. U. S. A. 95 (1998) 560–564.

[14] E.S. Johnson, A.A. Gupta, An E3-like factor that promotes SUMO conjugation to theyeast septins, Cell 106 (2001) 735–744.

[15] M.H. Kagey, T.A. Melhuish, D. Wotton, The polycomb protein Pc2 is a SUMO E3,Cell 113 (2003) 127–137.

[16] O. Kirsh, J.S. Seeler, A. Pichler, A. Gast, S. Muller, E. Miska, M. Mathieu, A.Harel-Bellan, T. Kouzarides, F. Melchior, A. Dejean, The SUMO E3 ligase RanBP2promotes modification of the HDAC4 deacetylase, EMBO J. 21 (2002) 2682–2691.

[17] S. Weger, E. Hammer, R. Heilbronn, Topors acts as a SUMO-1 E3 ligase for p53 invitro and in vivo, FEBS Lett. 579 (2005) 5007–5012.

[18] R.T. Hay, SUMO: a history of modification, Mol. Cell 18 (2005) 1–12.[19] M. Drag, G.S. Salvesen, DeSUMOylating enzymes—SENPs, IUBMB Life 60 (2008)

734–742.[20] R.T. Hay, SUMO-specific proteases: a twist in the tail, Trends Cell Biol. 17 (2007)

370–376.[21] D. Mukhopadhyay, M. Dasso, Modification in reverse: the SUMO proteases,

Trends Biochem. Sci. 32 (2007) 286–295.[22] L. Gong, S. Millas, G.G. Maul, E.T. Yeh, Differential regulation of sentrinized pro-

teins by a novel sentrin-specific protease, J. Biol. Chem. 275 (2000) 3355–3359.[23] J. Hang, M. Dasso, Association of the human SUMO-1 protease SENP2 with the nu-

clear pore, J. Biol. Chem. 277 (2002) 19961–19966.[24] J. Mikolajczyk, M. Drag, M. Bekes, J.T. Cao, Z. Ronai, G.S. Salvesen, Small ubiquitin-

related modifier (SUMO)-specific proteases: profiling the specificities and activitiesof human SENPs, J. Biol. Chem. 282 (2007) 26217–26224.

[25] D. Reverter, C.D. Lima, A basis for SUMO protease specificity provided by analysisof human Senp2 and a Senp2-SUMO complex, Structure 12 (2004) 1519–1531.

[26] L.N. Shen, C. Dong, H. Liu, J.H. Naismith, R.T. Hay, The structure of SENP1-SUMO-2complex suggests a structural basis for discrimination between SUMO paraloguesduring processing, Biochem. J. 397 (2006) 279–288.

[27] Z. Xu, S.W. Au, Mapping residues of SUMO precursors essential in differentialmaturation by SUMO-specific protease, SENP1, Biochem. J. 386 (2005) 325–330.

[28] A. Di Bacco, J. Ouyang, H.Y. Lee, A. Catic, H. Ploegh, G. Gill, The SUMO-specific pro-tease SENP5 is required for cell division, Mol. Cell. Biol. 26 (2006) 4489–4498.

[29] L. Gong, E.T. Yeh, Characterization of a family of nucleolar SUMO-specific proteaseswith preference for SUMO-2 or SUMO-3, J. Biol. Chem. 281 (2006) 15869–15877.

[30] T. Nishida, H. Tanaka, H. Yasuda, A novel mammalian Smt3-specific isopeptidase1 (SMT3IP1) localized in the nucleolus at interphase, Eur. J. Biochem. 267 (2000)6423–6427.

[31] K.I. Kim, S.H. Baek, Y.J. Jeon, S. Nishimori, T. Suzuki, S. Uchida, N. Shimbara, H.Saitoh, K. Tanaka, C.H. Chung, A new SUMO-1-specific protease, SUSP1, that ishighly expressed in reproductive organs, J. Biol. Chem. 275 (2000) 14102–14106.

[32] C.D. Lima, D. Reverter, Structure of the human SENP7 catalytic domain andpoly-SUMO deconjugation activities for SENP6 and SENP7, J. Biol. Chem. 283(2008) 32045–32055.

[33] D. Mukhopadhyay, F. Ayaydin, N. Kolli, S.H. Tan, T. Anan, A. Kametaka, Y. Azuma,K.D. Wilkinson, M. Dasso, SUSP1 antagonizes formation of highly SUMO2/3-conjugated species, J. Cell Biol. 174 (2006) 939–949.

[34] L.N. Shen, M.C. Geoffroy, E.G. Jaffray, R.T. Hay, Characterization of SENP7, aSUMO-2/3-specific isopeptidase, Biochem. J. 421 (2009) 223–230.

[35] N. Kolli, J. Mikolajczyk, M. Drag, D. Mukhopadhyay, N. Moffatt, M. Dasso, G.Salvesen, K.D. Wilkinson, Distribution and paralogue specificity of mammaliandeSUMOylating enzymes, Biochem. J. 430 (2010) 335–344.

[36] E.T. Yeh, SUMOylation and de-SUMOylation: wrestling with life's processes,J. Biol. Chem. 284 (2009) 8223–8227.

[37] T. Bawa-Khalfe, E.T. Yeh, SUMO losing balance: SUMO proteases disrupt SUMOhomeostasis to facilitate cancer development and progression, Genes Cancer 1(2010) 748–752.

[38] K.D. Sarge, O.K. Park-Sarge, SUMO and its role in human diseases, Int. Rev. CellMol. Biol. 288 (2011) 167–183.

[39] D. Tempe, M. Piechaczyk, G. Bossis, SUMO under stress, Biochem. Soc. Trans. 36(2008) 874–878.

[40] W. Zhou, J.J. Ryan, H. Zhou, Global analyses of sumoylated proteins in Saccharomycescerevisiae. Induction of protein sumoylation by cellular stresses, J. Biol. Chem. 279(2004) 32262–32268.

[41] F. Ayaydin, M. Dasso, Distinct in vivo dynamics of vertebrate SUMO paralogues,Mol. Biol. Cell 15 (2004) 5208–5218.

[42] H.A. Blomster, V. Hietakangas, J. Wu, P. Kouvonen, S. Hautaniemi, L. Sistonen,Novel proteomics strategy brings insight into the prevalence of SUMO-2 targetsites, Mol. Cell. Proteomics 8 (2009) 1382–1390.

[43] F. Golebiowski, I. Matic, M.H. Tatham, C. Cole, Y. Yin, A. Nakamura, J. Cox, G.J.Barton, M. Mann, R.T. Hay, System-wide changes to SUMO modifications in re-sponse to heat shock, Sci. Signal. 2 (2009) ra24.

1966 M.P. Pinto et al. / Biochimica et Biophysica Acta 1823 (2012) 1958–1966

[44] N. Martin, K. Schwamborn, V. Schreiber, A. Werner, C. Guillier, X.D. Zhang, O.Bischof, J.S. Seeler, A. Dejean, PARP-1 transcriptional activity is regulated bysumoylation upon heat shock, EMBO J. 28 (2009) 3534–3548.

[45] M. Brunet Simioni, A. De Thonel, A. Hammann, A.L. Joly, G. Bossis, E. Fourmaux, A.Bouchot, J. Landry, M. Piechaczyk, C. Garrido, Heat shock protein 27 is involved inSUMO-2/3 modification of heat shock factor 1 and thereby modulates the tran-scription factor activity, Oncogene 28 (2009) 3332–3344.

[46] V. Hietakangas, J.K. Ahlskog, A.M. Jakobsson, M. Hellesuo, N.M. Sahlberg, C.I.Holmberg, A. Mikhailov, J.J. Palvimo, L. Pirkkala, L. Sistonen, Phosphorylation ofserine 303 is a prerequisite for the stress-inducible SUMO modification of heatshock factor 1, Mol. Cell. Biol. 23 (2003) 2953–2968.

[47] A. Janer, A. Werner, J. Takahashi-Fujigasaki, A. Daret, H. Fujigasaki, K. Takada,C. Duyckaerts, A. Brice, A. Dejean, A. Sittler, SUMOylation attenuates the ag-gregation propensity and cellular toxicity of the polyglutamine expandedataxin-7, Hum. Mol. Genet. 19 (2010) 181–195.

[48] P. Krumova, E. Meulmeester, M. Garrido, M. Tirard, H.H. Hsiao, G. Bossis, H. Urlaub,M. Zweckstetter, S. Kugler, F. Melchior, M. Bahr, J.H. Weishaupt, Sumoylation in-hibits alpha-synuclein aggregation and toxicity, J. Cell Biol. 194 (2011) 49–60.

[49] S. Mukherjee, M. Thomas, N. Dadgar, A.P. Lieberman, J.A. Iniguez-Lluhi, Smallubiquitin-like modifier (SUMO) modification of the androgen receptor attenuatespolyglutamine-mediated aggregation, J. Biol. Chem. 284 (2009) 21296–21306.

[50] J.S. Steffan, N. Agrawal, J. Pallos, E. Rockabrand, L.C. Trotman, N. Slepko, K. Illes, T.Lukacsovich, Y.Z. Zhu, E. Cattaneo, P.P. Pandolfi, L.M. Thompson, J.L. Marsh, SUMOmodification of Huntingtin and Huntington's disease pathology, Science 304(2004) 100–104.

[51] E. Meulmeester, M. Kunze, H.H. Hsiao, H. Urlaub, F. Melchior, Mechanism andconsequences for paralog-specific sumoylation of ubiquitin-specific protease 25,Mol. Cell 30 (2008) 610–619.

[52] A. Pichler, A. Gast, J.S. Seeler, A. Dejean, F. Melchior, The nucleoporin RanBP2 hasSUMO1 E3 ligase activity, Cell 108 (2002) 109–120.

[53] A. Werner, M.C. Moutty, U. Moller, F. Melchior, Performing in vitro sumoylationreactions using recombinant enzymes, Methods Mol. Biol. 497 (2009) 187–199.

[54] A. Borodovsky, H. Ovaa, N. Kolli, T. Gan-Erdene, K.D. Wilkinson, H.L. Ploegh, B.M.Kessler, Chemistry-based functional proteomics reveals novel members of thedeubiquitinating enzyme family, Chem. Biol. 9 (2002) 1149–1159.

[55] K.D. Wilkinson, T. Gan-Erdene, N. Kolli, Derivitization of the C-terminus ofubiquitin and ubiquitin-like proteins using intein chemistry: methods and uses,Methods Enzymol. 399 (2005) 37–51.

[56] L.C. Dang, F.D. Melandri, R.L. Stein, Kinetic and mechanistic studies on the hydro-lysis of ubiquitin C-terminal 7-amido-4-methylcoumarin by deubiquitinating en-zymes, Biochemistry 37 (1998) 1868–1879.

[57] G. Blobel, V.R. Potter, Nuclei from rat liver: isolation method that combines puritywith high yield, Science 154 (1966) 1662–1665.

[58] J.M. Graham, Isolation of nuclei and nuclear membranes from animal tissues, in:J.S. Bonifacino, M. Dasso, J.B. Harford, J. Lippincott-Schwartz, K.M. Yamada(Eds.), Curr. Protoc. Cell Biol, John Wiley & Sons, Inc., New Jersey, 2001,pp. 3.10.1–3.10.19.

[59] M.J. Matunis, Isolation and fractionation of rat liver nuclear envelopes and nucle-ar pore complexes, Methods 39 (2006) 277–283.

[60] X.D. Zhang, J. Goeres, H. Zhang, T.J. Yen, A.C. Porter, M.J. Matunis, SUMO-2/3 mod-ification and binding regulate the association of CENP-E with kinetochores andprogression through mitosis, Mol. Cell. 29 (2008) 729–741.

[61] C. Yun, Y. Wang, D. Mukhopadhyay, P. Backlund, N. Kolli, A. Yergey, K.D.Wilkinson, M. Dasso, Nucleolar protein B23/nucleophosmin regulates the verte-brate SUMO pathway through SENP3 and SENP5 proteases, J. Cell Biol. 183(2008) 589–595.

[62] U.B. Ericsson, B.M. Hallberg, G.T. Detitta, N. Dekker, P. Nordlund, Thermofluor-basedhigh-throughput stability optimization of proteins for structural studies, Anal.Biochem. 357 (2006) 289–298.

[63] M. Vedadi, F.H. Niesen, A. Allali-Hassani, O.Y. Fedorov, P.J. Finerty Jr., G.A.Wasney, R. Yeung, C. Arrowsmith, L.J. Ball, H. Berglund, R. Hui, B.D. Marsden,P. Nordlund, M. Sundstrom, J. Weigelt, A.M. Edwards, Chemical screeningmethods to identify ligands that promote protein stability, protein crystalliza-tion, and structure determination, Proc. Natl. Acad. Sci. U. S. A. 103 (2006)15835–15840.

[64] M. Bekes, J. Prudden, T. Srikumar, B. Raught, M.N. Boddy, G.S. Salvesen, The dy-namics and mechanism of SUMO chain deconjugation by SUMO-specific prote-ases, J. Biol. Chem. 286 (2011) 10238–10247.

[65] G. Bossis, F. Melchior, Regulation of SUMOylation by reversible oxidation ofSUMO conjugating enzymes, Mol. Cell. 21 (2006) 349–357.

[66] Z. Xu, L.S. Lam, L.H. Lam, S.F. Chau, T.B. Ng, S.W. Au, Molecular basis of the redoxregulation of SUMO proteases: a protective mechanism of intermolecular disul-fide linkage against irreversible sulfhydryl oxidation, FASEB J. 22 (2008)127–137.

[67] Y. Sydorskyy, T. Srikumar, S.M. Jeram, S. Wheaton, F.J. Vizeacoumar, T.Makhnevych, Y.T. Chong, A.C. Gingras, B. Raught, A novel mechanism for SUMOsystem control: regulated Ulp1 nucleolar sequestration, Mol. Cell. Biol. 30(2010) 4452–4462.

[68] J. Goeres, P.K. Chan, D. Mukhopadhyay, H. Zhang, B. Raught, M.J. Matunis, TheSUMO-specific isopeptidase SENP2 associates dynamically with nuclear porecomplexes through interactions with karyopherins and the Nup107-160nucleoporin subcomplex, Mol. Biol. Cell. 22 (2011) 4868–4882.