This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright
Transcript
This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies areencouraged to visit:
Characterization of a novel posttranslational modification in neuronal nitric oxidesynthase by small ubiquitin-related modifier-1
Masatomo Watanabe a, Kouichi Itoh a,b,⁎a Laboratory of Molecular and Cellular Neuroscience, Kagawa School of Pharmaceutical Sciences, Tokushima Bunri University, Sanuki, Kagawa 769-2193, Japanb Laboratory for Brain Science, Kagawa School of Pharmaceutical Sciences, Tokushima Bunri University, Sanuki, Kagawa 769-2193, Japan
a b s t r a c ta r t i c l e i n f o
Article history:Received 18 February 2011Received in revised form 5 April 2011Accepted 12 April 2011Available online 20 April 2011
The multifaceted functions of nitric oxide (NO) in the CNS are defined by the activity of neuronal NOsynathase (nNOS). The activities of nNOS are modulated by posttranslational modifications, such asphosphorylation and ubiquitination, but whether it is modified by small ubiquitin-related modifier (SUMO)remains unknown. The aim of this study was to elucidate whether nNOS is posttranslationally modified bySUMO proteins. Bioinformatic analyses using SUMOplot and SUMOFI predicted that nNOS had potentialSUMO modification sites. When HEK293T cells were transiently co-expressed with nNOS and SUMO-1, twobands corresponding to nNOS-SUMO-1 conjugates were detected. In addition, two nNOS-SUMO-1 conjugateswere confirmed by an in vitro sumoylation assay using recombinant proteins. Furthermore, nNOS-SUMO-1conjugates were identified byMALDI-QIT/TOFmass spectrometry. These findings indicate that nNOS is clearlydefined as a SUMO-1 target protein both in vitro and at the cellular level. We next characterized specificenzymes in the nNOS-SUMO-1 conjugation cycle at the cellular level. SUMO-1 conjugation of nNOS dependedon Ubc9 (E2). The interaction between nNOS and Ubc9 was facilitated by PIASxβ (E3). On the other hand,SUMO-1 was deconjugated from nNOS by SENP1 and SENP2. Overall, this study has newly identified thatnNOS is posttranslationally modified by SUMO-1.
Nitric oxide (NO) has an important role in the regulation ofsynaptic plasticity via the activation of soluble guanyl cyclase [1].Furthermore, NO has paradoxically neuroprotective and neurotoxiceffects in the central nervous system [2]. These paradoxical functionsmediated by NO are due to the activity of neuronal NO synthase(nNOS) in the brain. Ca2+ activated NO production by NMDAreceptors is critically dependent on the activation of nNOS [3,4]. Itwas reported that the activity of nNOS is regulated by post-translational modifications as follows: (1) phosphorylation resultingin the activation of nNOS activity [5], (2) phosphorylation resulting ininhibition of nNOS activity [6,7], and (3) ubiquitination mediated byHsp70/CHIP resulting in degradation of nNOS [8]. These findings
suggest that the generation of NO in the brain is regulated byposttranslational modification-dependent nNOS activity. However,the relevance of small ubiquitin-related modifier modification(sumoylation), which is a common posttranslational modification ofmany proteins, remains unknown.
The three-dimensional structure and the conjugationmechanism ofsmall ubiquitin-related modifier (SUMO) are very similar to those ofubiquitin, but the biological functions of sumoylation are different fromubiquitination [9,10]. Sumoylation has inhibitory effects on polyubiqui-tin-mediateddegradation, anddoesnotprovide a signal forproteasomaldegradation [9–12]. Recently, they have been reports that sumoylationin neurons is closely regulated by the neuronal activities [13–15] andthat the levels of sumoylated proteins were increased in animal modelsof neurological disorders such as brain ischemia [16,17]. Our prelimi-nary study has also shown that the levels of SUMO-1-conjugatedproteins tend to change in the brains of animal models of epilepsy(unpublished data). However, most of these sumoylated proteins notedin previous studies have not yet been identified, and their physiologicalor pathological relevance remains unknown [14–18].
We previously reported that the level of nNOS protein expressionmay be increased by a posttranslational modification in thegeneralized epilepsy model of pentylenetetrazole-induced kindlingrodents, which is caused by abnormal neuronal activity [19]. Based onthese findings, we hypothesized that sumoylation plays an importantrole in the regulation of nNOS protein expression during epileptic
Biochimica et Biophysica Acta 1814 (2011) 900–907
Abbreviations: HEK293T cells, human embryonic kidney 293T cells; MALDI-QIT/TOF, Matrix-assisted laser desorption/ionization quadrupole ion trap time-of-flight;MS, mass spectrometry; NMDA, N-methyl-D-aspartate; NO, nitric oxide; nNOS,neuronal nitric oxide synthase; PIAS, protein inhibitor of activated STAT; SENP,sentrin-specific protease; SUMO, small ubiquitin-related modifier⁎ Corresponding author at: Laboratory for Brain Science and Laboratory of Molecular
and Cellular Neuroscience, Kagawa School of Pharmaceutical Sciences, Tokushima BunriUniversity, Sanuki, Kagawa 769-2193, Japan. Tel.: +81 87 894 5111; fax: +81 87 8940181.
j ourna l homepage: www.e lsev ie r.com/ locate /bbapap
Author's personal copy
neuronal activity. The nNOS protein expression and its activity levelare modulated by various posttranslational modifications, such asphosphorylation and ubiquitination, however, its sumoylation statuswas unknown. The purpose of the present study was to elucidatewhether rat nNOS is posttranslationally modified by SUMO.
2. Materials and methods
2.1. Plasmid construction
FLAG-tagged rat nNOS (UniProt ID: P29476, FLAG-nNOS) wasgenerated from pME18s-FLAG-tagged nNOS, kindly provided by Y.Watanabe (Showa Pharmaceutical University, Tokyo, Japan) [6], andwas subcloned into the EcoRV and NotI sites of the expression vectorpcDNA3 (Invitrogen Carlsbad, CA, USA). All of the single and multiplelysine-to-arginine substitution mutants [nNOS(K225R), nNOS(K469R), nNOS(K1062R), nNOS(K225/469/1062R)] were generatedby using the QuikChange Site-Directed Mutagenesis Kit (Stratagene,La Jolla, CA, USA). The nNOSmutants encoding the oxygenase domain(nNOSox, 1-747), which contains the calmodulin binding site, andencoding the reductase domain (nNOSred, 748-1429) were generatedfrom pcDNA3-Flag-nNOS using the primers 5′-AATGATATCACCAC-CATGGACTACAAGGACGACGA-3′ (sense, with EcoRV site) and 5′-AATCTCGAGTTACTGCCCCATTAGCTTGGCTGA-3′ (antisense, with XhoIsite) for nNOSox and 5′-AATGAATTCGCCATGGCCAAGAGGGTCAAG-3′(sense, with EcoRI site) and 5′-AATCGCCGGCGTTAGGAGCT-GAAAACCTCATCTGC-3′ (antisense, with NotI site) for nNOSred. Thenucleotide sequence of each nNOS mutant was confirmed by DNAsequencing. pCAG-hemagglutinin (HA)-tagged SUMO-1 (human,UniProt ID: P63165, HA-SUMO-1), HA-SUMO-2 (human, UniProt ID:P61956), HA-SUMO-3 (human, UniProt ID: P55854), pcDNA3-myc-tagged-Ubc9 (human, UniProt ID: P63279, myc-Ubc9), pcDNA-myc-Ubc9 mutant (C93S, replacement of cysteine 93 by serine) and PIASfamily plasmids (mouse) were kindly provided by T. Ohshima(Tokushima Bunri University, Kagawa, Japan) [20]. The HA-SUMO-1mutant (ΔGG, lacking a C-terminal di-glycine motif) was generatedfrom HA-SUMO-1. pEGFP-SENP-1 (mouse, UniProt ID: P59110),pEGFP-SENP-1 mutant (C603A, replacement of cysteine 603 byalanine), pEGFP-SENP-2 (human, UniProt ID: Q9HC62), and pEGFP-SENP-2mutant (C548A, replacement of cysteine 548 by alanine) werekindly provided by T. Nishida (Mie University, Mie, Japan) [21].
For in vitro sumoylation assays, pGEX-6P-1-SUMO-1-GG wasgenerated as described previously [20]. A plasmid expressing His-tagged-Ubc9 (His-Ubc9) was generated by inserting the BamHIfragment of pcDNA3-myc-tagged-Ubc9 into the pTrcHis vector(Invitrogen).
2.2. Immunoprecipitation and immunoblot analyses
HEK293T cells (Cell bank, Riken BioResource Center, Tsukuba,Japan) were grown in 10% FBS/D-MEM and were transfected by thecalcium phosphate-mediated method using 4 μg of a plasmidencoding FLAG-nNOS, 2 μg of a plasmid encoding SUMO and 0–4 μgof the other plasmid DNA per 60-mm-diameter dish. At 36 h aftertransfection, HEK293T cells were collected and homogenized with400 μl of lysis buffer containing 100 mM Tris–HCl (pH 7.4), 150 mMNaCl, 1% NP-40, 0.5% deoxycholate, 0.1% sodium dodecyl sulfate(SDS), 10 mM N-ethylmaleimide (Calbiochem, La Jolla, CA, USA),5 mM EDTA (Dojindo Lab., Kumamoto, Japan), 5 mM EGTA (Sigma-Aldrich Corp., St. Louis, MO, USA) and protease inhibitor cocktail at4 °C using a Dounce homogenizer. The protease inhibitor cocktail wascomposed of 4 mM AEBSF (Roche Diagnostics GmbH, Mannheim,Germany), 1 μg/ml Aprotinin (Roche), 100 μM Leupeptin (Roche),10 ug/ml (microgram) α2-Macroglobulin (Roche), 1 mM Benzamide(Sigma-Aldrich Corp.), 1 μMPepstatin A (Calbiochem), 10 μg/ml Soybentrypsin inhibitor (Sigma-Aldrich Corp.), 10 μg/ml egg trypsin inhibitor
(Sigma-Aldrich Corp.), 10 mM iodoacetamide (Sigma-Aldrich Corp.),10 mM sodium orthovanadate (Sigma-Aldrich Corp.) and 100 μM N-acetyl-leucinyl-leucinyl-norleucinal (Calbiochem). After incubation for20 min on ice, the homogenatewas centrifuged at 18,000×g for 20 min.Until the immunoprecipitation assay, the supernatant (cell lysates) wasstored at−80 °C.
After measuring the protein concentrations of the cell lysates byusing Bradford's Protein Assay Reagent (Bio-Rad Laboratories, Inc.Hercules, CA, USA), equal amounts (1200–1500 μg) of the cell lysateswere incubated with 1/100 (wt/wt) of the monoclonal antibody(mAb)-FLAG M2 (mouse IgG; Sigma-Aldrich Corp.) for 4 h at 4 °C andwere subsequently incubated with Protein G Sepharose 4 Fast Flow(GE Healthcare Bio-Sciences, Uppsala, Sweden) for 1 h at 4 °C.Immunoprecipitates were washed six times with lysis buffer withoutthe protease inhibitor cocktail, eluted in 20 μl of 2× SDS-polyacryl-amide gel electrophoresis (SDS-PAGE) sample buffer at 95 °C andanalyzed by an immunoblot analysis.
The eluted proteins were separated by 6% SDS-PAGE or 3–8%NuPAGE Tris–Acetate gel (Invitrogen), transferred to Immobilon-P(Millipore, Billerica, MA, USA) using 10 mM CAPS-buffer (pH 11.0)containing 5%methanol, and processed for immunoblot analyses withthe anti-HA tag antibody (1:3000, rabbit IgG; MBL, Nagoya, Japan)after blocking with 5% milk in Tris-buffered saline containing 0.05%Tween 20 (TBST) and an HRP-conjugated anti-rabbit IgG antibody(1:10,000; Thermo Fisher Scientific Inc., Rockford, IL, USA), using theSuperSignal WestPico chemiluminescent substrate to visualize theprotein spots according to the manufacturer's instructions (ThermoFisher Scientific Inc.). After blotting with the anti-HA tag antibody, thesame membrane could not be fully reprobed with an anti-FLAG M2antibody (Data not shown). Therefore immunoprecipitated nNOSwasanalyzed by Ponceau S staining. A total of 10 μg of cell lysates wereprocessed individually for the immunobot analyses with the anti-HAtag antibody, anti-FLAG M2 antibody, anti-myc tag antibody (1:1000,mouse IgG; Roche) and anti-GFP antibody (1:1000, mouse IgG;NACALAI TESQUE Inc., Kyoto, Japan).
2.3. In vitro sumoylation analysis
The recombinant rat nNOS expressed in the Escherichia coli strainBL21 was affinity purified with adenosine 2′,5′-diphosphate-agarose(Sigma-Aldrich Corp.), as described previously [22]. Mature SUMO-1(exposing its C-terminal diglycine motif, SUMO-1-GG) expressed as aGST-fusion protein in the E. coli strain BL21 (DE3) was affinity purifiedwith glutathione S-sepharose beads [20] and cleaved with PreScissionProtease (GE Healthcare Bio-Sciences) to remove the GST-tag. His-Ubc9was expressed in the BL21 bacteria, and was affinity purified withHisLink™ Protein Purification Resin (Promega, Madison, WI, USA).SUMO activating enzyme E1 heterocomplex (human, SAE1/SAE2)purchased from Boston Biochem (Cambridge, MA, USA). The in vitrosumoylation assay was performed according to the method publishedbyOhshima and Shimotohno [20], withminormodifications. Two pmolof recombinant rat nNOSwas incubatedwith 2 pmol of E1 (SAE1/SAE2),20 pmol of E2 (His-Ubc9), and 20 pmol of SUMO-1-GG in a reactionbuffer consisting of 50 mM Tris–HCl (pH 7.5), 10 mM ATP, 10 U/mlcreatine kinase, 10 mM phosphocreatine and 3 mM MgCl2 for 1 h at30 °C. After incubation, nNOS-SUMO-1 conjugates that were separatedby 6% SDS-PAGE were transferred to Immobilon-P and were processedfor an immunoblot analysis with anti-nNOS (1:15,000, rabbit IgG;Sigma-Aldrich Corp.) after blocking with 5% milk in TBST and HRP-conjugated anti-rabbit IgG antibody (1:10,000) using the SuperSignalWestPico chemiluminescent substrate.
2.4. Identification of nNOS-SUMO-1 conjugates by MALDI-QIT/TOF MS
After the in vitro sumoylation reaction, the nNOS-SUMO-1conjugates separated by 6% SDS-PAGE were analyzed by SYPRO
901M. Watanabe, K. Itoh / Biochimica et Biophysica Acta 1814 (2011) 900–907
Author's personal copy
Ruby fluorescence staining (Bio-Rad) with a laser scanner (MolecularImager FX Pro; Bio-Rad). The nNOS-SUMO-1 conjugates weresubjected to in-gel trypsin digestion followed by MS and MS/MSanalyses as described previously [23]. Briefly, protein bands wereexcised from the SDS-PAGE gels by using a FluoroPhoreStar 3000(Anatech, Tokyo, Japan) and then were incubated with 50 ng trypsin(Sequencing-Grade Modified Trypsin; Promega) overnight at 30 °C.After tryptic digestion, the MS and MS/MS spectra of peptidefragments were measured by an AXIMA-QIT/TOF (Shimadzu Biotech,Kyoto, Japan) mass spectrometer. To identify the internal amino acidsequences, the resultant masses were searched against SwissProtdatabases by using the MS/MS ion search software program (MASCOTMS/MS ion search; Matrix Science Ltd., London, United Kingdom).
3. Results
3.1. Bioinformatic prediction of sumoylation sites in nNOS
To determine whether nNOS has potential SUMO modificationsites, we performed a bioinformatic screening for high-probabilitysumoylation sites using a SUMOplot™ (http://www.abgent.com/tools/sumoplot_login) analysis. As shown in Table 1, three high-probability sumoylation sites at Lysine 1062, Lysine 225 and Lysine469 in rat and mouse nNOS were predicted. In order to facilitate theidentification of the negatively charged amino acid-dependent
sumoylation motif (NDSM) [24] and phosphorylation-dependentsumoylation motif (PDSM) [25] at these putative SUMO sites, weutilized another software package, the SUMOFI (SUMO motif finder,http://csbi.ltdk.helsinki.fi/sumofi/). Based on the SUMOFI prediction,nNOS had one NDSM sequence at Lysine 1062, but no consensusPDSM sequences (Table 1). These bioinformatic analyses predictedthat nNOS has potential SUMO modification sites.
3.2. nNOS is a substrate for SUMO-1 modification in HEK293T cells
We first examined whether nNOS is modified by the SUMO familyin HEK293T cells transiently expressing both FLAG-nNOS and HA-SUMOs. To detect SUMO modification of nNOS, the cell lysates wereimmunoprecipitated with an anti-FLAG antibody, followed by animmunoblot analysis using an anti-HA antibody. Two major bands(approximately 180 and 200 kDa) were detected in the immunopre-cipitates from cell lysates of FLAG-nNOS and HA-SUMO-1 co-transfectants (Fig. 1A arrows). Mature human SUMO-1 is an 11 kDaprotein, but one SUMO-1 conjugate appeared to be approximately20 kDa larger than the molecular weight of most substrates on SDS-PAGE [26]. These findings suggest that the 180 and 200 kDaimmunoreactive bands might be correspond to nNOS (160 kDa)conjugated to one and two SUMO-1 molecules, respectively.
To confirm whether the 180 and 200 kDa bands were represen-tative of covalently conjugated nNOS and SUMO-1, we constructed aSUMO-1 ΔGG mutant that lacks the C-terminal di-glycine motifessential for covalent modification. The 180 and 200 kDa bandscompletely disappeared in the HEK293T cells transfected with FLAG-nNOS and the HA-SUMO-1 ΔGGmutant (Fig. 1B). This result indicatesthat the two major immunoreactive bands correspond to nNOS-SUMO-1 conjugates linked by an isopeptide bond via a C-terminal di-glycine motif. On the other hand, nNOS-SUMO-2 and -SUMO-3
Table 1The high probability SUMO-1 acceptor sites in rat nNOS.
Position Sequence Score by SUMOplot NDSM by SUMOFI PDSM by SUMOFI
1062 VKVE 0.93 Yes No225 AKAE 0.79 No No469 GKHD 0.67 No No
Fig. 1. nNOS is posttranslationally modified by SUMO-1 in HEK293T cells. A. nNOS is a substrate for sumoylation. Immunoprecipitates (IP) obtained with a monoclonal anti-FLAGM2antibody (mAb-FLAG) were analyzed by immunoblotting (IB) with a polyclonal anti-HA antibody (pAb-HA). Two nNOS-SUMO conjugates were detected (arrows, top panel).Ponceau S staining showed the nNOS proteins in the IP (2nd panel). The third and 4th panels show the total levels of SUMO proteins obtained by IB with pAb-HA and nNOS proteinsby IB with a mAb-FLAG, respectively. *, non specific signals. B. nNOS is covalently modified by SUMO-1. Immunoprecipitates (IP) obtained with the monoclonal anti-FLAG M2antibody (mAb-FLAG) were analyzed by immunoblotting (IB) with a polyclonal anti-HA antibody (pAb-HA). The two nNOS-SUMO-1 conjugates (arrows, top panel) completelydisappeared in cells expressing the SUMO-1 ΔGG mutant. Ponceau S staining showed the presence of the nNOS proteins in the IP (2nd panel). The two lower panels show the totallevels of SUMO-1 proteins obtained by IB with pAb-HA and nNOS proteins obtained by IB with mAb-FLAG, respectively. *, non specific signals.
902 M. Watanabe, K. Itoh / Biochimica et Biophysica Acta 1814 (2011) 900–907
Author's personal copy
conjugates were only minimally detected in the HEK293T cellstransfected with FLAG-nNOS and HA-SUMO-2 or HA-SUMO-3(Fig. 1A). Thus, nNOS is covalently conjugated to one or two SUMO-1 molecules, rather than to SUMO-2 and SUMO-3.
3.3. nNOS is a substrate for SUMO-1 modification in vitro
To further examine whether nNOS is indeed a substrate for SUMO-1 modification, we performed an in vitro sumoylation assay usingrecombinant nNOS proteins. The nNOS proteins were incubated withSUMO-1 reaction components in the presence of ATP, because SUMOis activated in an ATP-dependent manner by an E1 enzyme consistingof the SAE1/UBA2 heterodimer [9,12,26]. The analysis of the reactionproducts by immunoblotting with an anti-nNOS antibody revealedtwo slower migrating forms (approximately 180 and 200 kDa bands)of nNOS that were generated in an ATP-dependent and SUMO-1-dependent manner (Fig. 2A). These slower migrating bands corre-sponded to the nNOS-SUMO-1 conjugates that were shown in thelysates from HEK293T cells transfected with nNOS and SUMO-1(Fig. 1A). These results indicate that nNOS is a substrate forsumoylation in vitro.
3.4. Identification of nNOS-SUMO-1 conjugates by MALDI-QIT/TOF massspectrometry
To confirm that the proteins within the two slower migratingbands (180 and 200 kDa) obtained during the in vitro sumoylationassay were nNOS-SUMO-1 conjugates, each band was extracted fromthe gel and analyzed byMALDI-QIT/TOFmass-spectrometry (MS). Themolecular masses of tryptic fragments of each target band ‘[a]–[c]’shown in Fig. 2A were measured using MALDI-QIT/TOF MS, and eachtarget was identified by a MASCOT PMF search against Swiss-Protdatabases. Band [a] and band [c] were identified as the same rat nNOSprotein (GenBank ID: NP_434686; Fig. 2B, black dots). To confirm theresults of the PMF search, bands [a] and [c] were sequenced byMS/MS
Fig. 2. Identification of SUMO-1modifications of nNOS in vitro. A. SUMO-1 is conjugatedto nNOS in vitro. After an in vitro sumoylation reaction, the nNOS-SUMO-1 conjugateswere analyzed by immunoblotting (IB) with a polyclonal anti-nNOS antibody (pAb-nNOS) (upper panel) and by SYPRO Ruby staining (lower panel). The nNOS-SUMO-1conjugates were detected as shown in bands [b] and [c]. B. The MALDI-TOF MSspectrum of a tryptic digest of the nNOS-SUMO-1 conjugates. The masses of trypticpeptides from bands [a] and [c] shown in A were analyzed by using AXIMA-QIT/TOF MSand were searched against the Swiss-Prot databases using the MASCOT PMF searchsoftware program. Rat nNOS was identified, and its assigned peaks are shown as blackdots. C. The MS/MS spectrum of nNOS-SUMO-1 conjugates determined using MALDI-QIT/TOF MS. The tryptic peptides from bands [a] and [c] shown in A were sequenced byan MS/MS analysis. The common precursor peptide shown in the MS spectrum(1993.07 m/z, asterisks shown in B) and a specific precursor peptide derived fromnNOS-SUMO1 conjugates (896.48 m/z, arrow head shown in B) were fragmented byusing AXIMA-QIT/TOF MS and were searched against the Swiss-Prot databases usingthe MASCOT MS/MS ion search software program. Rat nNOS and SUMO-1 wereidentified by a MS/MS ion search of the 1993.07 m/z (upper panel) and 896.48 m/z(lower panel) peaks, respectively. Their amino acid sequences were identified by adatabase search. Each peak is annotated with the name of the ion type derived from thestructure.
Fig. 3. SUMO-1 modification of nNOS depends on the E2 enzyme (Ubc9) activity.Immunoprecipitates (IP) obtained with the monoclonal anti-FLAG M2 antibody (mAb-FLAG) were analyzed by immunoblotting (IB) with a polyclonal anti-HA antibody(pAb-HA). The nNOS-SUMO-1 conjugates were increased in a Ubc9-dependent manner(arrows, top panel). Ponceau S staining showed the nNOS proteins in the IP (2nd panel).The three lower panels show the total levels of SUMO-1 proteins obtained by IB withpAb-HA, nNOS proteins obtained by IB with mAb-FLAG, and Ubc9 proteins obtained by IBwith a monoclonal anti-myc antibody (mAb-myc), respectively. *, non specific signals.
903M. Watanabe, K. Itoh / Biochimica et Biophysica Acta 1814 (2011) 900–907
Author's personal copy
analysis using MALDI-QIT/TOF (Fig. 2C). Based on a MASCOT MS/MSion search against the Swiss-Prot database, the internal amino acidsequence of the 1993.07 m/z peak (Fig. 2B asterisks) was identified as1239 NPQVPCILVGPGTGIAPFR1257 in rat nNOS (Fig. 2C, top panel).Next, to identify the specific peak observed in the MS spectrum ofband [c] but not in band [a], the tryptic fragment with a peak at896.48 m/z (Fig. 2B, arrowhead) was chosen for MS/MS sequencing.By a MASCOT MS/MS ion search against the Swiss-Prot database, weidentified the internal amino acid sequence of the 896.48 m/z peak as74FLFEGQR80 in SUMO-1 (Fig. 2C, bottom panel). These resultsindicate that the protein within the slower migrating band [c](200 kDa) in Fig. 2A is a nNOS-SUMO-1 conjugate. In addition, theother slower band [b] (180 kDa) was also identified as a nNOS-SUMO-1 conjugate (data not shown). Overall, these findings confirm thatnNOS is a SUMO-1 target protein both in vitro and at the cellular level.
3.5. Ubc9 is required for the SUMO-1 modification of nNOS
Because the SUMO conjugating pathway depends on the activity ofthe E2 enzyme Ubc9 [9,12,27], we investigated whether the SUMO-1conjugation of nNOS also depends on Ubc9 (Fig. 3). Ubc9 significantlyincreased the two nNOS-SUMO-1 conjugates in HEK293T cellstransfected with a combination of FLAG-nNOS, HA-SUMO-1 andmyc-Ubc9. A Ubc9 (C/S) mutant, which completely abolished theSUMO-1 conjugating activity, did not increase the two nNOS-SUMO-1conjugates, even though nNOS was sumoylated by endogenous Ubc9in Ubc9 (C/S) mutant-transfected HEK293T cells (lanes 2 and 7 inFig. 3). These results indicate that Ubc9 is required for the SUMO-1conjugation of nNOS, thus further confirming that it is a substrate forconventional SUMO-1 modification.
3.6. SUMO-1 modification of nNOS is regulated by the PIASxβ E3 ligase
It is well known that E3 ligases in the SUMO conjugation pathwayfunction as adaptors that stabilize the interaction between the SUMO-
Ubc9 thioester and the acceptor proteins in vivo. The PIAS familycontaining the SIZ/PIAS-RING domain mediates the SUMO conjuga-tion to a wide variety of substrates [28]. Therefore, we investigatedwhether the stability of the nNOS and SUMO-1-Ubc9 thioester isincreased by PIAS family members (PIASxβ and PIAS1). SUMO-1conjugation to nNOS was analyzed in HEK293T cells transientlyexpressing either PIASxβ or PIAS1. The intensities of the nNOS-SUMO-1 signals were increased 3-fold (299%) and 2-fold (187%) by transientoverexpression of PIASxβ and PIAS1, respectively (Fig. 4, IP, anti-HA).Because the expression level of PIASxβ was lower than PIAS1 (Fig. 4,input, anti-myc), it appears that the interaction between nNOS andSUMO-1-Ubc9 is stabilized more strongly by PIASxβ than PIAS1. Tofurther clarify the importance of PIASxβ activity on nNOS sumoyla-tion, we transfected cells with a PIASxβ (C/S) mutant whichcompletely lacks E3 ligase activity. The PIASxβ (C/S) mutant inhibitedthe accumulation of nNOS-SUMO-1 conjugates. Taken together, thesefindings clearly indicate that PIASxβ facilitates the sumoylation ofnNOS by Ubc9.
3.7. Sumoylated nNOS is desumoylated by the SUMO-specific proteasesSENP1and SENP2
Generally, sumoylated proteins can be readily and rapidlydeconjugated by the isopepidase activity of the SENPs. In mammals,six SENPs have been shown to function as SUMO-specific proteases. Inparticular, SUMO-1 is deconjugated from substrates by SENP1 andSENP2 [29]. Therefore, we examined whether SENP1 and SENP2desumoylate nNOS. As shown in Fig. 5, SENP1 and SENP2 completelyabrogated the nNOS-SUMO-1 modification. To clarify the importanceof SENP isopeptidase activity on nNOS-SUMO-1 desumoylation, wetransfected HKE293T cells with SENP1 (C/A) and SENP2 (C/A)mutants, in which the conserved cysteine residue within the catalyticdomain had been changed to an alanine. These mutants completelylacked SUMO-1-specific isopeptidase activity, and their transfectiondid not result in an increase in nNOS desumoylation (Fig. 5). Taken
Fig. 4. PIASxβ functions as an E3-ligase for nNOS. Immunoprecipitates (IP) obtained with a monoclonal anti-FLAG M2 antibody (mAb-FLAG) were analyzed by immunoblotting (IB)with a polyclonal anti-HA antibody (pAb-HA). Co-expression of PIASxβ significantly increased the nNOS-SUMO-1 conjugation (arrows, top panel). The PIASxβ (C/S) mutant did notincrease nNOS-SUMO-1 conjugation. Ponceau S staining showed the nNOS proteins in IP (2nd panel). The three lower panels show the total levels of SUMO-1 proteins obtained by IBwith pAb-HA, PIAS proteins obtained by IB with a monoclonal anti-myc antibody (mAb-myc) and nNOS proteins obtained by IB with mAb-FLAG, respectively. *, non specific signals.
904 M. Watanabe, K. Itoh / Biochimica et Biophysica Acta 1814 (2011) 900–907
Author's personal copy
together, these findings clearly indicate that SENP1 and SENP2mediate the desumoylation of nNOS.
3.8. The identification of sumoylation sites in nNOS
Finally, we examined the potential sumoylation sites in nNOS.Based on the SUMOplot analysis (Table 1), we focused on threepotential sites (K224, K469, and K1062), and generated nNOSmutants in which these lysine residues were individually (K224R,K469R, or K1062R) or collectively (3KR; K224R/K469R/K1062R)changed to arginine residues. When each mutant was co-expressedwith HA-SUMO-1, two nNOS-SUMO-1 conjugates (180 and 200 kDa)were detected in all transfectants of these mutants (Fig. 6A). Theseresults indicate that the three sumoylation candidate sites indicatedby the SUMOplot analysis are not the actual sites in nNOS in HEK293Tcells.
Next we performed deletion experiments to characterize thesumoylation domain in nNOS. We generated an oxygenase domain(nNOSox, rat nNOS 1-747) and a reductase domain (nNOSred, ratnNOS 748-1429) mutant (Fig. 6B). Many ladder SUMO-conjugatedbands were detected when the HEK293T cells were co-transfectedwith HA-SUMO-1 and FLAG-nNOSox or FLAG-nNOSred (Fig. 6C). Sinceonly two major bands were detected in the immunoprecipitates fromcell lysates of cells expressing the full-length nNOS, the sumoylationsites of nNOSox and nNOSred may be different from the original sites
on the full-length rat nNOS protein. We were not able to successfullyidentify the exact sumoylation sites at this time. However, our dataindicate that nNOS is definitely sumoylated by Ubc9/PIASxβ and isdesumoylated by SENP1/2 at the cellular level. Further experimentswill be necessary to identify the sumoylation sites and functions ofsumoylated-nNOS.
4. Discussion
Posttranslational protein modifications are versatile mechanismsthat cells use to control the function of proteins by regulating theiractivity, subcellular localization, stability, as well as their interactionwith other proteins. The posttranslational modifications of nNOS suchas phosphorylation [5–7] and ubiquitination [8] play an importantrole in the regulation of nNOS enzymatic activity. In addition, othersumoylated proteins have recently been found to be important forvarious cellular functions [9]. In the current study, we determinedwhether nNOS was posttanslationally sumoylated in cells.
Fig. 5. SENP1 and SENP2 function as SUMO-deconjugating enzymes for nNOS-SUMO-1conjugates. Immunoprecipitates (IP) obtained with the monoclonal anti-FLAG M2antibody (mAb-FLAG) were analyzed by immunoblotting (IB) with a polyclonal anti-HA antibody (pAb-HA). SUMO-1 was completely deconjugated from nNOS by SENP1and SENP2 (arrows, top panel). SENP (C/A) mutants completely lacked SUMO-1-specific isopeptidase activity. Ponceau S staining showed the nNOS proteins in IP (2ndpanel). The three lower panels show the total levels of SUMO-1 proteins obtained by IBwith pAb-HA, SENP proteins by IBwith amonoclonal anti-GFP antibody (mAb-GFP) andnNOS proteins by IB with a mAb-FLAG, respectively. *, non specific signals.
Fig. 6. Sumoylation sites in nNOS. A. SUMO-1 modification of nNOS mutants with Lyssubstitutions. Immunoprecipitates (IP) obtained with the monoclonal anti-FLAG M2antibody (mAb-FLAG) were analyzed by immunoblotting (IB) with a polyclonal anti-HA antibody (pAb-HA). Two nNOS-SUMO-1 conjugates were detected for single KRmutant- and multiple KR mutant-nNOS (arrows, upper panel). Therefore, it appearsthat nNOS is sumoylated at lysine residues other than Lys-225, Lys-469 and Lys-1062.Ponceau S staining showed the nNOS proteins in IP (lower panel). B. A schematicrepresentation of rat nNOS (full-length) and its truncation mutants. The N-terminaloxygenase domain (Ox), C-terminal reductase domain (Red) and calmodulin bindingdomain (CaM) of nNOS are shown. C. SUMO-1 modification of truncation mutants ofnNOS. Immunoprecipitates obtained with a monoclonal anti-FLAG M2 antibody (mAb-FLAG) were analyzed by immunoblotting (IB) with a polyclonal anti-HA antibody (pAb-HA). Multiple SUMO-1 conjugates were detected for each truncation mutant of nNOS(arrows, left panel). The right panel shows the expression levels of nNOS proteinsobtained by IB with mAb-FLAG.
905M. Watanabe, K. Itoh / Biochimica et Biophysica Acta 1814 (2011) 900–907
Author's personal copy
We observed that the nNOS protein was predominantly sumoy-lated by SUMO-1, and that it was covalently modified by at least oneSUMO-1 molecule in HEK293T cells transiently cotransfected withnNOS and SUMO-1 (Fig. 1). To confirm the nNOS sumoylation in cells,in vitro sumoylation was examined using recombinant proteins, andthe partial sequences of the nNOS-SUMO-1 conjugates were analyzedby MALDI-QIT/TOF MS. These analyses demonstrated that the SUMOpeptide was contained in the tryptic fragments from the immunore-active bands of nNOS (Fig. 2C). Therefore, these results indicate thatnNOS is clearly modified by SUMO-1 at both the in vitro and cellularlevels.
Sumoylation appears to be a highly selective process both withrespect to the choice of substrates as well as to the timing of theirmodification. The enzymes required for reversible sumoylation, suchas the SUMO-activating enzyme (E1), the single SUMO-conjugatingenzyme (E2) and SUMO ligases (E3) have sequences with similaritiesto their counterparts in the ubiquitin system, and are conserved fromyeast to humans [12]. The mammalian SUMO system includes E1 (aheterodimer of SAE1 and SAE2), E2 (Ubc9), multiple E3s [PIAS(protein inhibitor of activated STAT) family etc.] and deconjugatingenzymes (SUMO-specific proteases; SENPs) [9,12]. Newly translatedSUMO proteins are inactive precursors that are initially matured bythe SENPs. The mature form of SUMO is then activated in an ATP-dependent manner with the E1 enzyme SAE1/SAE2. These two stepsproceed independently of the sumoylation targets.
Given that these steps are substrate-independent, we character-ized the specific enzymes directly mediating the SUMO cycle of nNOSas the target protein. The Ubc9 E2 enzyme catalyzed the transfer ofSUMO-1 to the nNOS protein which was localized in the cytoplasm(Fig. 3). Studies on Ubc9 inmammals have shown that it is localized inthe cytoplasm [30,31] and synaptosome [13,14], although Ubc9 is apredominantly nuclear protein [32]. The prototypes of the PIAS classof E3 ligases were initially discovered in yeast [33], and mammaliancells contain at least five PIAS proteins including splice variants, PIAS1,xα, xβ, 3 and y [12,28]. PIAS proteins facilitate the sumoylationprocess by acting as adaptors that stabilize the interaction betweenSUMO-1 thioester-loaded Ubc9 and a wide variety of target sub-strates. In the nNOS SUMO cycle, PIASxβmediated nNOS sumoylationas the E3 ligase (Fig. 4). SENPs play important roles in the SUMO cycle,because they have a dual role as processing proteases for SUMOprecursors and as isopeptidases for substrate de-sumoylation [29].The SUMO-1 conjugation of nNOS by Ubc9 and PIASxβ is subsequentlyremoved by SENP-1 and SENP-2. Taken together, our results identifythe specific enzymatic system mediating the SUMO cycle of nNOS.
Next, a search was conducted for the sumolyation sites in thenNOS protein by a SUMOplot analysis. Three high-probabilitysumoylation sites at K1062, K225 and K469 in rat nNOS werepredicted (Table 1). K1062 is located at the protruding-fingercontaining the CD2A regulatory element [34], K225 is located at thePIN (the protein inhibitor of nNOS) binding domain [35], and K469 islocated between the heme and arginine binding sites [36]. It isconceivable that a conformational change induced by the sumoylationat these sites could play an important role inmodifying the function ofnNOS by sumoylation. However, unexpectedly, nNOS mutants with alysine (K) to arginine (R) substitution (K1062R, K224R, and K469R) aswell as the triple KR mutant were still sumoylated (Fig. 6A). Thissuggests that other consensus [27,30] or non-consensus [37] lysineresidues other than the three predicted sites in nNOS are the target(s)for SUMO-1 modification. Since the predicted sumoylation sites werenot characterized as SUMO-1 modification sites in nNOS, weperformed domain deletion experiments to determine potentialregions affected by sumoylation in the nNOS protein. nNOS consistsof a N-terminal catalytic oxigenase domain (nNOSox) and a C-terminal reductase domain (nNOSred) [38,39]. Both nNOSox andnNOSred mutants were non-specifically sumoylated by SUMO-1 intransient transfectants (Fig. 6C). This finding suggests that the
conformation of the full-length nNOS protein, rather than localconsensus sequences, may play an important role in its selectiverecognition by Ubc9. Further studies will be required to characterizethe sumoylation sites and the functions of sumoylated nNOS.
Recently, several studies showed that the brain [16–18] andsynaptosome [13,14] contain many SUMO-1-modified proteins,which have not yet been identified except for a few plasmamembraneproteins (GluR6, mGluR8, Kv 1.5), a transcription factor (MEF2) andneurological disease-associated proteins (β amyloid, Huntingtin) [9].We previously examined whether nNOS-SUMO-1 conjugates existedin the brain. However, no nNOS-SUMO-1 conjugates were detected bythe immunopurification methods used for that study [23] (data notshown). Although the brain has components of the sumoylationmachinery, nNOS-SUMO-1 conjugates may not have been detected inbrain during our previous study for the following reasons: (1) the poolof endogenous free SUMO-1 is very low [40], (2) sumoylation isreadily reversible by the SENP proteases [9,12,26], (3) our detectionmethods may not have been sufficient for brain tissue specimens.Interestingly SUMO-1, Ubc9 [41] and PIASxβ [42] are all expressed incerebellar granule cells, which highly express nNOS proteins [43].These findings suggest that nNOSmay be posttranslationally modifiedby SUMO-1 in the brain, although the colocalization patterns of nNOSand SUMO-1 and their modifiers in the brain remain unknown.
In conclusion, this investigation has demonstrated that nNOS isposttranslationally modified by specific enzymes mediating theSUMO-1 cycle of nNOS. This third posttranslational modificationmight be important for regulating nNOS function in the brain.Although the sumoylation sites have not yet been identified, theidentification of the functional consequences of sumoylated nNOSprotein presents an important challenge for future research.
Acknowledgements
We are grateful to Dr. Takayuki Ohshima (Kagawa School ofPharmaceutical Sciences, Tokushima Bunri University) for his adviceand for providing plasmids for SUMO, Ubc9 and PIAS family members.We also thank Dr. Yasuo Watanabe (Department of Pharmacology,Showa Pharmaceutical University) for providing the plasmid forpME18s-FLAG-tagged nNOS and the E. coli for recombinant nNOS andDr. Tamotsu Nishida (Life Science Research Center, Mie University) forproviding the plasmids for the SENPs. We also thank A. Miyai, M.Komori and K. Okubo for assistance.
This work was supported by Grants-in Aid for Scientific Research(C) (to M. W. and K. I.) from the JSPS, a Grant from the Japan EpilepsyResearch Foundation (to K.I.) and a Grant from Tokushima BunriUniversity for Educational Reform and Collaborative Research (to M.W. and K.I.).
References
[1] J. Garthwaite, Concepts of neural nitric oxide-mediated transmission, Eur. J.Neurosci. 27 (2008) 2783–2802.
[2] V. Calabrese, C. Mancuso, M. Calvani, E. Rizzarelli, D.A. Butterfield, A.M. Stella,Nitric oxide in the central nervous system: neuroprotection versus neurotoxicity,Nat. Rev. Neurosci. 8 (2007) 766–775.
[3] J. Garthwaite, S.L. Charles, R. Chess-Williams, Endothelium-derived relaxing factorrelease on activation of NMDA receptors suggests role as intercellular messengerin the brain, Nature 336 (1988) 385–388.
[4] J. Garthwaite, C.L. Boulton, Nitric oxide signaling in the central nervous system,Annu. Rev. Physiol. 57 (1995) 683–706.
[5] G.A. Rameau, D.S. Tukey, E.D. Garcin-Hosfield, R.F. Titcombe, C. Misra, L. Khatri, E.D. Getzoff, E.B. Ziff, Biphasic coupling of neuronal nitric oxide synthasephosphorylation to the NMDA receptor regulates AMPA receptor trafficking andneuronal cell death, J. Neurosci. 27 (2007) 3445–3455.
[6] K. Komeima, Y. Hayashi, Y. Naito, Y. Watanabe, Inhibition of neuronal nitric-oxidesynthase by calcium/calmodulin-dependent protein kinase IIalpha throughSer847 phosphorylation in NG108-15 neuronal cells, J. Biol. Chem. 275 (2000)28139–28143.
[7] G.A. Rameau, L.Y. Chiu, E.B. Ziff, Bidirectional regulation of neuronal nitric-oxidesynthase phosphorylation at serine 847 by the N-methyl-D-aspartate receptor, J.Biol. Chem. 279 (2004) 14307–14314.
906 M. Watanabe, K. Itoh / Biochimica et Biophysica Acta 1814 (2011) 900–907
Author's personal copy
[8] H.M. Peng, Y. Morishima, G.J. Jenkins, A.Y. Dunbar, M. Lau, C. Patterson,W.B. Pratt, Y.Osawa, Ubiquitylation of neuronal nitric-oxide synthase by CHIP, a chaperone-dependent E3 ligase, J. Biol. Chem. 279 (2004) 52970–52977.
[9] S. Martin, K.A. Wilkinson, A. Nishimune, J.M. Henley, Emerging extranuclear rolesof protein SUMOylation in neuronal function and dysfunction, Nat. Rev. Neurosci.8 (2007) 948–959.
[10] H.D. Ulrich, Mutual interactions between the SUMO and ubiquitin systems: a pleaof no contest, Trends Cell Biol. 15 (2005) 525–532.
[11] M.C. Geoffroy, R.T. Hay, An additional role for SUMO in ubiquitin-mediatedproteolysis, Nat. Rev. Mol. Cell Biol. 10 (2009) 564–568.
[12] K.A. Wilkinson, J.M. Henley, Mechanisms, regulation and consequences of proteinSUMOylation, Biochem. J. 428 (2010) 133–145.
[13] S. Martin, A. Nishimune, J.R. Mellor, J.M. Henley, SUMOylation regulates kainate-receptor-mediated synaptic transmission, Nature 447 (2007) 321–325.
[14] M. Feligioni, A. Nishimune, J.M. Henley, Protein SUMOylation modulates calciuminflux and glutamate release from presynaptic terminals, Eur. J. Neurosci. 29(2009) 1348–1356.
[15] H. Lu, B. Liu, S. You, Q. Xue, F. Zhang, J. Cheng, B. Yu, The activity-dependent stimuliincrease SUMO modification in SHSY5Y cells, Biochem. Biophys. Res. Commun.390 (2009) 872–876.
[16] H. Cimarosti, C. Lindberg, S.F. Bomholt, L.C. Rønn, J.M. Henley, Increased proteinSUMOylation following focal cerebral ischemia, Neuropharmacology 54 (2008)280–289.
[17] Y.J. Lee, P. Castri, J. Bembry, D. Maric, S. Auh, J.M. Hallenbeck, SUMOylationparticipates in induction of ischemic tolerance, J. Neurochem. 109 (2009) 257–267.
[18] W. Yang, H. Sheng, H.M. Homi, D.S. Warner, W. Paschen, Cerebral ischemia/strokeand small ubiquitin-like modifier (SUMO) conjugation — a new target fortherapeutic intervention? J. Neurochem. 106 (2008) 989–999.
[19] K. Itoh, M. Watanabe, K. Yoshikawa, Y. Kanaho, L.J. Berliner, H. Fujii, Magneticresonance and biochemical studies during pentylenetetrazole-kindling develop-ment: the relationship between nitric oxide, neuronal nitric oxide synthase andseizures, Neuroscience 129 (2004) 757–766.
[20] T. Ohshima, K. Shimotohno, Transforming growth factor-beta-mediated signalingvia the p38MAP kinase pathway activates Smad-dependent transcription throughSUMO-1 modification of Smad4, J. Biol. Chem. 278 (2003) 50833–50842.
[21] T. Nishida, F. Kaneko, M. Kitagawa, H. Yasuda, Characterization of a novelmammalian SUMO-1/Smt3-specific isopeptidase, a homologue of rat Axam,which is an Axin-binding protein promoting β-catenin degradation, J. Biol. Chem.276 (2001) 39060–39066.
[22] L.J. Roman, E.A. Sheta, P. Martasek, S.S. Gross, Q. Liu, B.S. Masters, High-levelexpression of functional rat neuronal nitric oxide synthase in Escherichia coli, Proc.Natl. Acad. Sci. U. S. A. 92 (1995) 8428–8432.
[23] M. Watanabe, Y. Sakurai, T. Ichinose, Y. Aikawa, M. Kotani, K. Itoh, Monoclonalantibody Rip specifically recognizes 2′,3′-cyclic nucleotide 3'-phosphodiesterasein oligodendrocytes, J. Neurosci. Res. 84 (2006) 525–533.
[24] S.H. Yang, A. Galanis, J. Witty, A.D. Sharrocks, An extended consensus motifenhances the specificity of substrate modification by SUMO, EMBO J. 25 (2006)5083–5093.
[25] V. Hietakangas, J. Anckar, H.A. Blomster, M. Fujimoto, J.J. Palvimo, A. Nakai, L.Sistonen, PDSM, a motif for phosphorylation-dependent SUMO modification,Proc. Natl. Acad. Sci. U. S. A. 103 (2006) 45–50.
[26] E.S. Johnson, Protein modification by SUMO, Annu. Rev. Biochem. 73 (2004)355–382.
[27] D.A. Sampson, M. Wang, M.J. Matunis, The small ubiquitin-like modifier-1(SUMO-1) consensus sequence mediates Ubc9 binding and is essential for SUMO-1 modification, J. Biol. Chem. 276 (2001) 21664–21669.
[28] M.M. Rytinki, S. Kaikkonen, P. Pehkonen, T. Jääskeläinen, J.J. Palvimo, PIASproteins: pleiotropic interactors associated with SUMO, Cell. Mol. Life Sci. 66(2009) 3029–3041.
[29] D. Mukhopadhyay, M. Dasso, Modification in reverse: the SUMO proteases, TrendsBiochem. Sci. 32 (2007) 286–295.
[30] M.S. Rodriguez, C. Dargemont, R.T. Hay, SUMO-1 conjugation in vivo requires botha consensus modification motif and nuclear targeting, J. Biol. Chem. 276 (2001)12654–12659.
[31] H. Zhang, H. Saitoh, M.J. Matunis, Enzymes of the SUMO modification pathwaylocalize to filaments of the nuclear pore complex, Mol. Cell. Biol. 22 (2002)6498–6508.
[32] W. Seufert, B. Futcher, S. Jentsch, Role of a ubiquitin-conjugating enzyme indegradation of S- and M-phase cyclins, Nature 373 (1995) 78–81.
[33] E.S. Johnson, A.A. Gupta, An E3-like factor that promotes SUMO conjugation to theyeast septins, Cell 106 (2001) 735–744.
[34] G.M. Knudsen, C.R. Nishida, S.D. Mooney, P.R. Ortiz de Montellano, Nitric-oxidesynthase (NOS) reductase domain models suggest a new control element inendothelial NOS that attenuates calmodulin-dependent activity, J. Biol. Chem. 278(2003) 31814–31824.
[35] S.R. Jaffrey, S.H. Snyder, PIN: an associated protein inhibitor of neuronal nitricoxide synthase, Science 274 (1996) 774–777.
[37] H.A. Blomster, S.Y. Imanishi, J. Siimes, J. Kastu, N.A. Morrice, J.E. Eriksson, L.Sistonen, In vivo identification of sumoylation sites by a signature tag andcysteine-targeted affinity purification, J. Biol. Chem. 285 (2010) 19324–19329.
[38] E.D. Garcin, C.M. Bruns, S.J. Lloyd, D.J. Hosfield, M. Tiso, R. Gachhui, D.J. Stuehr, J.A. Tainer, E.D. Getzoff, Structural basis for isozyme-specific regulation ofelectron transfer in nitric-oxide synthase, J. Biol. Chem. 279 (2004)37918–37927.
[39] L. Zhou, D.Y. Zhu, Neuronal nitric oxide synthase: structure, subcellularlocalization, regulation, and clinical implications, Nitric Oxide 20 (2009)223–230.
[40] H. Saitoh, J. Hinchey, Functional heterogeneity of small ubiquitin-relatedprotein modifiers SUMO-1 versus SUMO-2/3, J. Biol. Chem. 275 (2000)6252–6258.
[41] M. Watanabe, K. Takahashi, K. Tomizawa, H. Mizusawa, H. Takahashi, Develop-mental regulation of Ubc9 in the rat nervous system, Acta Biochim. Pol. 55 (2008)681–686.
[42] A. Shalizi, P.M. Bilimoria, J. Stegmüller, B. Gaudillière, Y. Yang, K. Shuai, A. Bonni,PIASx is a MEF2 SUMO E3 ligase that promotes postsynaptic dendriticmorphogenesis, J. Neurosci. 27 (2007) 10037–10046.
[43] D.S. Bredt, P.M. Hwang, S.H. Snyder, Localization of nitric oxide synthaseindicating a neural role for nitric oxide, Nature 347 (1990) 768–770.
907M. Watanabe, K. Itoh / Biochimica et Biophysica Acta 1814 (2011) 900–907
