Nitric oxide synthase generates nitric oxide locally ... - Semantic Scholar

1 downloads 0 Views 2MB Size Report
Yasuko Iwakiri*†‡, Ayano Satoh§, Suvro Chatterjee¶, Derek K. Toomre§ , Cecile M. ... organelles in cells (10), and low concentrations of NO activate the.
Nitric oxide synthase generates nitric oxide locally to regulate compartmentalized protein S-nitrosylation and protein trafficking Yasuko Iwakiri*†‡, Ayano Satoh§, Suvro Chatterjee¶, Derek K. Toomre§储, Cecile M. Chalouni§, David Fulton**, Roberto J. Groszmann*‡, Vijay H. Shah¶, and William C. Sessa†,†† *Section of Digestive Diseases, Departments of †Pharmacology and §Cell Biology, and 储Institute for Cancer Research, Yale University School of Medicine, New Haven, CT 06510; ‡Hepatic Hemodynamic Laboratory, VA Connecticut Healthcare System, West Haven, CT 06516; ¶Gastroenterology Research Unit, Department of Physiology and Tumor Biology Program, Mayo Clinic, Rochester, MN 55905; and **Vascular Biology Center and Department of Pharmacology, Medical College of Georgia, Augusta, GA 30912

Nitric oxide (NO) is a highly diffusible and short-lived physiological messenger. Despite its diffusible nature, NO modifies thiol groups of specific cysteine residues in target proteins and alters protein function via S-nitrosylation. Although intracellular S-nitrosylation is a specific posttranslational modification, the defined localization of an NO source (nitric oxide synthase, NOS) with protein Snitrosylation has never been directly demonstrated. Endothelial NOS (eNOS) is localized mainly on the Golgi apparatus and in plasma membrane caveolae. Here, we show by using eNOS targeted to either the Golgi or the nucleus that S-nitrosylation is concentrated at the primary site of eNOS localization. Furthermore, localization of eNOS on the Golgi enhances overall Golgi protein S-nitrosylation, the specific S-nitrosylation of N-ethylmaleimidesensitive factor and reduces the speed of protein transport from the endoplasmic reticulum to the plasma membrane in a reversible manner. These data indicate that local NOS action generates organelle-specific protein S-nitrosylation reactions that can regulate intracellular transport processes. endothelial nitric oxide synthase 兩 Golgi 兩 targeting

N

itric oxide (NO), produced by the nitric oxide synthase (NOS) family of proteins, regulates a variety of important physiological responses, including vasodilation, respiration, cell migration, and apoptosis (1–5). NO has been considered to mediate these responses by activating the primary NO effector soluble guanylyl cyclase to produce cGMP (6) or by NO-based chemical modifications of proteins or perhaps lipids. One clear example of cGMPindependent actions of NO is protein thiol group modification by NO known as S-nitrosylation (7). This posttranslational control mechanism regulates important physiological activities of proteins in response to endogenously or exogenously generated NO (8). Thus, S-nitrosylation of proteins is an emerging area of investigation for NO-mediated physiological responses (8). NO is a lipophillic, highly diffusible, and short-lived physiological messenger (9). On the one hand, NO is thought to diffuse over a wide area (100 ␮m), moving freely through membranes of neighboring cells (9). On the other hand, given the apparent promiscuity of NO, the question arises as to how S-nitrosylation might occur in a precisely regulated manner, i.e., protein S-nitrosylation occurs on specific thiol residues in proteins that are targeted to specific organelles in cells (10), and low concentrations of NO activate the ryanodine receptor via thiol modification, whereas higher concentrations inhibit the receptor (11). There are several compelling arguments in favor of the concept that all sources of NO are not bioequivalent. (i) In cardiac myocytes, which express two forms of NOS, gene deletion of either neuronal NOS or endothelial NOS (eNOS) exerts isoform-specific phenotypes, arguing that the source of NO favors local control of different cellular functions (12). Moreover, in many circumstances, the actions of endogenously generated NO from NOS cannot be recapitulated by drugs liberwww.pnas.org兾cgi兾doi兾10.1073兾pnas.0605907103

ating NO, suggesting that the compartmentalization of NO equivalents is physiologically important to regulate diverse cellular function. (ii) In red blood cells, S-nitrosohemoglobin binds to and transnitrosylates the plasma membrane anion exchange protein AE1, whereas NO itself does not S-nitrosylate AE1 in the absence of hemoglobin (13, 14). (iii) Inducible NOS (iNOS) can only S-nitrosylate cyclooxygenase after a direct protein–protein interaction; if the interaction between iNOS and cyclooxygenase is prevented (with an inhibitory peptide), iNOS generates NO, but does not S-nitrosylate cyclooxygenase (15). (iv) NOSs are themselves S-nitrosylated, thus NO would appear to be acting ‘‘very locally’’ (16, 17). (v) Privileged sites of S-nitrosylation have been identified in particular, mitochondrial procaspase-3 is constitutively S-nitrosylated whereas the cytosolic form is not (5). (vi) OxyR is S-nitrosylated more readily by S-nitrosoglutathione (GSNO) than S-nitrosocysteine (CysNO) (18), whereas the opposite is true for hemoglobin (19). GSNO (but not CysNO) binds directly to OxyR, whereas it is too large to access the buried Cys in hemoglobin. In comparison to the actions of NO-regulating metal-centered processes (activation of soluble guanylyl cyclase and inhibition of cytochrome oxidase) or its interaction with other radicals, thiol modification via S-nitrosylation is thought to require higher concentrations of NO among the primary biological reactions with NO (20) to sustain protein S-nitrosylation in vitro (21). Therefore, the flux of NO generated by NOS may regulate specific cellular functions despite the diffusible and short-lived properties of NO; however, this has never been directly demonstrated. eNOS is unique among the NOS family members as it is localized mainly to specific intracellular membrane domains, including the cytoplasmic aspect of the Golgi apparatus and plasma membrane caveolae (22–25). Particularly, on the Golgi, it is shown that eNOS is colocalized with well known Golgi proteins, such as Golgi matrix protein 130 (GM130), 53K Golgi protein, and mannosidase II (22–24, 26). Previously, we hypothesized that a NO pool is formed around and within the Golgi and may create a favorable environAuthor contributions: Y.I., C.M.C., and R.J.G. designed research; Y.I., A.S., S.C., and C.M.C. performed research; D.K.T., D.F., and V.H.S. contributed new reagents/analytic tools; Y.I., S.C., D.K.T., C.M.C., V.H.S., and W.C.S. analyzed data; and Y.I., R.J.G., V.H.S., and W.C.S. wrote the paper. The authors declare no conflict of interest. This article is a PNAS direct submission. Abbreviations: NOS, nitric oxide synthase; eNOS, endothelial NOS; GSNO, S-nitrosoglutathione; GSNOR, GSNO reductase; NLS, nuclear localization signal; RFP, red fluorescent protein; DAF, diaminofluorescein; DAF-2T, DAF-2 triazole; DAF-2DA, 4-amino-5-methylamino-2⬘,7⬘difluorescein; SNAP, S-nitroso-N-acetyl-D-L-penicillamine; L-NAME, L-nitroarginine methylester; VSVG, vesicular stomatitis virus glycoprotein; ER, endoplasmic reticulum; NSF, N-ethylmaleimide-sensitive factor. ††To

whom correspondence should be addressed. E-mail: [email protected].

This article contains supporting information online at www.pnas.org/cgi/content/full/ 0605907103/DC1. © 2006 by The National Academy of Sciences of the USA

PNAS 兩 December 26, 2006 兩 vol. 103 兩 no. 52 兩 19777–19782

CELL BIOLOGY

Edited by Louis J. Ignarro, University of California School of Medicine, Los Angeles, CA, and approved October 23, 2006 (received for review July 13, 2006)

Fig. 1. The functional analysis of eNOS constructs. (A) Subcellular localization of eNOS in COS cells. eNOS constructs fused with RFP include: WT-eNOSRFP, which is WT eNOS that primarily localizes at the Golgi apparatus, and RFP-eNOS-NLS to target the nucleus. (Scale bar: 10 ␮m.) (B) eNOS expression of eNOS constructs. (C) NO production in COS-7 cells transfected with eNOS constructs. Forty-eight hours after the transfection of eNOS constructs or control plasmid (RFP only), the media were processed for the measurement of nitrite, a stable breakdown product of NO in aqueous solution by chemiluminescence. (Left) NO release from samples collected from basal (4 h) accumulation of nitrite in serum-containing media is shown. The same cells were then incubated with serum-free DMEM for 6 h and the fresh media containing agonist ATP (100 ␮M) for 30 min. (Right) NO release as a result of ATP stimulation is shown. The cells were lysed to determine the total protein concentration and evaluate equal eNOS expression by Western blot. Data are presented as means ⫾ SEM; n ⫽ 3; *, P ⬍ 0.01; **, P ⬍ 0.05.

ment for local S-nitrosylation of proteins in living cells (1–5). Thus, the purpose of this study is to examine local NO production and function by using organelle-targeted eNOS constructs. Results and Discussion To examine whether Golgi-localized eNOS can generate a local NO pool, we imaged NO production and the source of NO (NOS) simultaneously in living cells and tested whether the localization of NOS correlates with local S-nitrosylation of proteins and whether Golgi-directed NOS regulates cellular exocytosis. To place the source of NO in two distinct subcellular compartments in cells, we generated WT eNOS (WT-eNOS-RFP), which localizes primarily on the Golgi apparatus in transfected cells and a nuclear-localized eNOS [RFP-eNOS-NLS (nuclear localization signal)] as fusion proteins with monomeric red fluorescent protein (RFP) to monitor their subcellular localization in living cells that do not express endogenous NOS. As seen in Fig. 1A, WT-eNOS-RFP was localized in a perinuclear crescent and colocalized with GM130 as described (22–27), whereas RFP-eNOS-NLS was confined to the nucleus. Next, we examined the levels of protein expression and NO release from transfected cells. As seen in Fig. 1B, both constructs expressed well in transfected cells and released NO (measured as NO2⫺ with NO-specific chemiluminescence) (Fig. 1C). Both WTeNOS-RFP and RFP-eNOS-NLS released basal (⬎4 h; Fig. 1C Left) and stimulated (with ATP for 30 min; Fig. 1C Right) NO, with WT-eNOS-RFP releasing two to three times the amount of NO compared with RFP-eNOS-NLS. Although the expression levels of WT-eNOS-RFP were similar to those of RFP-eNOS-NLS, target19778 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0605907103

Fig. 2. Kinetics of DAF-2DA (an intracellular NO indicator) in response to exogenous and endogenous NO in living cells. The y axis of each graph is relative fluorescent intensity (RFI) of DAF-2 fluorescence. (A) NO donor, NONOate, increased DAF-2 fluorescence in living cells. Images of DAF-2 taken every minute after COS-7 cells were loaded with DAF-2DA. The fluorescent intensity of DAF-2T was measured after the dye was loaded. The average ⫾ SEM in each time point was obtained from three independent experiments. (B) An inhibitor of NOS (L-NAME) blocked DAF-2 fluorescence in cells that expresses WT-eNOS-RFP. Before being stimulated with ATP (100 ␮M) for NO production, cells were treated with or without L-NAME (100 ␮M). A stock solution of L-NAME (100 mM) was prepared in water and diluted in culture medium (1:1,000) to have a final concentration of 100 ␮M. The treatment of cells with ATP resulted in DAF-2T fluorescence (Upper), which was blocked by L-NAME (Lower). Control RFI values were obtained from COS cells expressing WT-eNOS-RFP treated without a presence of L-NAME. * and **, P ⬍ 0.05. (Scale bars: 100 ␮m.)

ing eNOS to nucleus reduces NOS activation presumably because of insufficient access to calcium/calmodulin and impaired phosphorylation (28). Thus, localization of eNOS in different subcellular organelles influences the amount of NO produced. Because of the instability and low concentrations of NO in biological systems, it had been very difficult to directly visualize NO production in living cells. Direct imaging of NO in biological systems became available with the development of a series of fluorescent indicators diaminofluoresceins (DAFs) in 1998 (29). In the presence of oxygen, NO and NO related reactive nitrogen species nitrosate 4,5-DAF by a second-order reaction to yield the highly fluorescent DAF-2 triazole (DAF-2T). To test whether DAF-2DA (4-amino-5-methylamino-2⬘,7⬘difluorescein) is sensitive to detect intracellular nitrosation (29), COS-7 cells were loaded with DAF-2DA then treated with NONOate/AM, which releases NO intracellularly (Fig. 2A Left for representative images and Right for quantitative data), and time-lapsed images of DAF-2T were taken every minute immediately after the addition of the NO donor by a confocal laser scanning microscope (LSM510 Imaging System; Zeiss, Thornwood, NY). Because COS cells do not contain NOS isoforms, the background fluorescence at time 0 was low. The dye-loaded cells showed an increase in fluorescence intensity immediately after the NO donor was given and the fluorescent signal was diffusely distributed throughout the cell, suggesting the specificity of the dye for detecting NO and that DAF did not concentrate in specific organelles in COS cells, making it an appropriate tool to detect NO derived from eNOS. Similar results were seen when using GSNO and S-nitroso-N-acetyl-D-LIwakiri et al.

CELL BIOLOGY

Fig. 3. NO is confined to the region where eNOS targeted. The x axis of each graph refers to the distance of the laser line scan in the individual cells shown with numbers on a line with an arrow (1, 2, and 3), reflecting the different regions of interest monitored for colabeled DAF-2 fluorescence, and the y axis is the fluorescent intensity for DAF-2T (blue line) and RFP (red line). (A) Regions of nitrosation were formed in and around the Golgi complex where WT-eNOS-RFP is highly localized. Small increases in DAF-2T were also observed on the nuclear membrane and plasma membrane. COS-7 cells were transfected with WT-eNOS-RFP, and nitrosation was monitored at the indicated time after the addition of ATP (100 ␮M). (B) Nitrosation was confined to the nucleus in cells that express RFP-eNOS-NLS. (Scale bars: 10 ␮m.)

penicillamine (SNAP) as NO donors [100 ␮M of each for 30 min; see supporting information (SI) Fig. 6). We also quantified the responsiveness of the dye to detect changes in nitrosation in cells tranfected with WT-eNOS-RFP (Fig. 2B). COS cells were transfected with WT-eNOS-RFP, loaded with DAF-2DA, then stimulated with ATP in the absence or presence of the NOS inhibitor L-nitroarginine methylester (L-NAME). Treatment of cells with ATP resulted in an increase in DAF-2 fluorescence, which was inhibited by L-NAME, showing that we can monitor NO production in single cells expressing transfected eNOS constructs (see quantitative data in Fig. 2B Right). The incomplete inhibition of DAF-2 fluorescence by L-NAME may be caused by the production of other DAF-reactive compounds not derived from eNOS (30–32). Treatment of human endothelial cells with vascular endothelial growth factor increases the detection of DAF-2 nitrosation products in the peri-nuclear region of the cells (33). Thus, we simultaneously monitored the targeted RFP constructs (for eNOS localization) and DAF-2 fluorescence (for nitrosation) in transfected COS-7 cells by confocal microscopy over a 30-min period. In Fig. 3, the x axis of each graph refers to the distance of the laser line scan in the individual cells shown with 1, 2, and 3 reflecting the different regions of interest monitored for colabeled DAF-2 fluorescence, and the y axis is the fluorescent intensity for DAF-2 (blue line) and Iwakiri et al.

RFP (magenta line). Cells were observed under baseline conditions and after administration of ATP to evoke NO release. As seen in Fig. 3A, WT-eNOS-RFP was highly localized in region 1, with lesser enzyme in regions 2 and 3. However, upon stimulation of cells with ATP, the highest amount of DAF-2 fluorescence (NO signal; Fig. 3A Right) was detected in a peri-nuclear region overlapping with or adjacent to WT-eNOS-RFP, similar to what was observed in endothelial cells (33). Because NO partitions rapidly into hydrophobic environments such as biological membranes, this result may reflect the membranous nature of protein-NO adducts formed (34). Stated differently, if NO-mediated nitrosation was regulated by diffusion only, the imaged DAF-2 fluorescence changes derived from eNOS should have been similar to that seen with NO donors (see Fig. 2 A). To further confirm that eNOS-derived NO acts locally, we performed identical experiments in cells transfected with RFP-eNOS-NLS. This mutant eNOS highly localized at nucleus (Fig. 3B). NO concentration (imaged by DAF-2) was colocalized in the nucleus (area 1) and peri-nuclear regions (area 2). This pattern of DAF-2 nitrosation in cells expressing RFP-eNOS-NLS was distinct from what has observed in those expressing WT-eNOSRFP. The increase in NO at the peri-nuclear region is probably because NO produced inside of the nucleus could diffuse out of nucleus and form an NO reservoir in the membranous peri-nuclear PNAS 兩 December 26, 2006 兩 vol. 103 兩 no. 52 兩 19779

Fig. 4. Protein S-nitrosylation is restricted to regions of the cell where eNOS is localized. S-nitrosylation of endogenous proteins were detected by using an S-nitrosocysteine mAb. COS cells were transfected with WT-eNOS-RFP or RFPeNOS-NLS, and DAPI was used to detect nuclei (blue). (A) WT-eNOS-RFP was concentrated in a peri-nuclear pattern (red; Left). Labeling with an isotype control antibody for the S-nitrosocysteine mAb showed low background levels (Center). Thus, there was no overlap of red and green (Right). (B) WT-eNOS-RFP colocalized with cellular S-nitrosylated proteins (green) around the peri-nuclear area of cells. (C) RFP-eNOS-NLS (red) was concentrated in nucleus and colocalized with S-nitrosylated proteins (green). (D) The secondary antibody (Alexa 488conjugated goat anti-mouse) for S-nitrosocysteine mouse monoclonal antibody did not cause nonspecific binding to COS-7 cells (Left). The NO donor, SNAP (100 ␮M), increases S-nitrosylated proteins in COS cells, and the pattern of Snitrosylated proteins were distinctive from cells transfected with WT-eNOS-RFP and RFP-eNOS-NLS (Center). Incubation with 0.2%-HgCl2 abolishes S-nitrosylation caused by the treatment with 100 ␮M-SNAP (Right). (Scale bars: 10 ␮m.)

region. Collectively, these data show that NO is preferentially channeled to sites in proximity to the source of eNOS activity. If NO is formed at the primary site of eNOS localization in cells (Fig. 3), then it is tenable that S-nitrosylation of proteins may also be localized as suggested (10, 21). To test this idea, we performed immunofluoresence analysis for protein S-nitrosylation by using an S-nitrosocysteine-specific antibody in COS cells transfected with WT-eNOS-RFP or RFP-eNOS-NLS (Fig. 4). As shown in Fig. 4A Left (red channel), WT-eNOS-RFP is concentrated in a perinuclear pattern. Labeling with an isotype control Ab for the S-nitrocysteine mAb (mouse IgG) demonstrates low background levels (Fig. 4A Center) with little merging of the two patterns. Imaging of cells expressing WT-eNOS-RFP (Fig. 4B Center) or RFP-eNOS-NLS (Fig. 4C Center) shows that the pattern of Snitrosylated proteins coregisters with the targeted eNOS constructs. In contrast, cells treated with the NO donor drug, SNAP, show a markedly different pattern of S-nitrosylated protein immunreactivity (Fig. 4D Center). Moreover, the levels of S-nitrosylated protein detection can be eliminated by preincubation with HgCl2 19780 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0605907103

(Fig. 4D Right), which hydrolyzes S–NO bonds, before immunostaining with the S-nitrocysteine mAb. Collectively, these data indicate that NOS compartmentalization can govern the subcellular distribution of S-nitrosylated proteins in cells. These data suggest that physiological and pathophysiological stimuli can stimulate endothelial cells to regulate eNOS-mediated NO binding to target proteins, a result that may explain different cellular responses induced by NO donors and activation of endogenous NO synthesis. The above results suggest that the localization of NOS can determine local protein S-nitrosylation. To test whether eNOS localization influences the function of NO, we investigated the transport of a temperature-sensitive vesicular stomatitis virus glycoprotein (VSVG) tagged with GFP (tsO45-VSVG-GFP labeled as VSVG-GFP) from the endoplasmic reticulum (ER) to the cell surface. This glycoprotein is retained in the ER at 40°C by its misfolding, but upon temperature reduction to 32°C, folds properly and moves out of ER and into the Golgi complex before being transported to the plasma membrane (35). BSC-1 cells were used for this study because they do not express NOS isoforms and are easier to microinject with plasmid DNA constructs than COS cells. BSC-1 cells were microinjected with VSVG-GFP and a control plasmid, WT-eNOS-RFP or RFP-eNOS-NLS, and incubated at 40°C for 2–3 h to trap VSVG-GFP in the ER and then shifted to 32°C in the presence of cycloheximide to block further protein synthesis. As shown in Fig. 5A, in cells expressing VSVG-GFP (after 40 min release from the ER) and injected with control plasmid, there was surface targeting of VSVG relative to the total amount of VSVG in the cells. In contrast, in cells injected with WT-eNOSRFP, there was a marked delay in targeting of VSVG-GFP to the cell surface, whereas cells injected with RFP-eNOS-NLS exhibited similar transport properties to control-injected cells. To further determine whether the reduced rate of transport observed in cells with WT-eNOS-RFP depends on NO, cells expressing WT-eNOSRFP were treated with L-NAME to abrogate NO release and the VSVG exocytosis was examined (Fig. 5B). Cells treated with L-NAME restored the rate of VSVG-GFP transport to that observed in control or RFP-eNOS-NLS-expressing cells, demonstrating control of basal membrane transport by NO, perhaps via S-nitrosylation of proteins important for secretion. One protein implicated in secretion and a target for Snitrosylation is the N-ethylmaleimide-sensitive factor (NSF). NSF was first identified as a cytosolic protein necessary for in vitro reconstitution of intercisternal Golgi transport, and subsequently shown to regulate vesicle trafficking and exocytosis by mediating membrane fusion through its ATPase activity (36). Indeed, the S-nitrosylation of NSF by eNOS inhibits the stimulated exocytosis of Von Willebrand’s factor in endothelial cells (37). Moreover, S-nitrosylation of NSF has been shown to inhibit NSF disassembly of soluble NSF attachment protein receptor (SNARE) complexes, resulting in the inhibition of exocytosis in endothelial cells (37). Based on these findings, we tested whether the delayed transport observed in cells expressing WT-eNOS-RFP is associated with S-nitrosylation of NSF by using a biotin-switch method to label S-nitrosylated proteins as described (Fig. 5C). As seen in Fig. 5C, transfection of both WT-eNOS-RFP or RFP-eNOS-NLS results in the S-nitrosylation of NSF via a HgCl2-sensitive bond. However, the level of NSF S-nitrosylation is markedly greater in cells expressing WT-eNOS-RFP compared with cells expressing RFP-eNOS-NLS, suggesting that the decreased VSVG-GFP trafficking in cells expressing WT-eNOS-RFP may be caused, in part, by increased NSF S-nitrosylation and inhibition of NSF activity as described (37). These observations clearly indicate that localization of eNOS influences protein S-nitrosylation and subsequent NO-dependent cellular functions. Our data support the idea that NOS generates NO locally to regulate compartmentalized protein S-nitrosylation and signaling akin to calcium-calmodulin or protein phosphorylation being regulated by subcellular targeting of a calmodulin target or a kinase Iwakiri et al.

Methods Fig. 5. NO confined near the Golgi apparatus by WT-eNOS-RFP delays the exocytotic pathway of VSVG trafficking. (A) VSVG transport to the plasma membrane was impaired in cells expressing WT-eNOS-RFP. BSC-1 cells, grown on 12-mm coverslips in a 12-well dish, were microinjected with a mixture of tsO45VSVG-GFP plasmid DNA and WT-eNOS-RFP, RFP-eNOS-NLS, or RFP alone (control). After the microinjection, cells were incubated at 40°C for 2–3 h for the expression, then incubated at 4°C for 15 min for the protein folding in the presence of cyclohexamide (100 ␮g/ml). At this point, tsO45-VSVG-GFP protein is exclusively localized at ER. Then, incubation of cells at 32°C was started to chase the transport of tsO45-VSVG-GFP from ER to Golgi and to the surface of plasma membrane (40 min). After the incubation at 32°C for 40 min, cells were fixed in 4% paraformaldehyde and processed for the immunofluorescence of cell surface VSVG protein, using a mAb to the luminal/cell surface VSVG that was labeled with Alexa 647. The quantification of the surface VSVG protein was performed by measuring the fluorescent intensity of Alexa 647, which was normalized by total expression of VSVG in cells by measuring fluorescent intensity of GFP. *, P ⬍ 0.01. (B) The impaired VSVG transport in WT-eNOS-RFP-expressing cells was NOdependent. To determine whether the decreased VSVG transport in cells expressing WT-eNOS-RFP is NO-dependent, cells microinjected with WT-eNOS-RFP and tsO45-VSVG-GFP were incubated at 40°C for 3 h in the presence or absence of L-NAME (100 ␮M). Temperature was then shifted to 32°C for 40 min to chase VSVG transport to the cell surface. *, P ⬍ 0.01. (C) S-nitrosylation of NSF was increased in cells expressing WT-eNOS-RFP. COS-7 cells, transfected with myc-NSF and WT-eNOS-RFP or RFP-eNOS-NLS by transient transfection, were used 48 h after transfection. As a control, cells were transfected with myc-NSF alone to check for endogenous S-nitrosylation. Cells were incubated in DMEM without serum for 6 h, followed by the incubation with DMEM containing ATP (100 ␮M) with or without 0.2% HgCl2 for 1 h. To detect S-nitrosylation of NSF, the biotin-switch method developed by Jaffrey and colleagues (46, 47) was used. (Top) Purified S-nitrosylated protein was subjected to 10% SDS/PAGE and Western blot analysis by using myc antibody to detect NSF. (Middle and Bottom) eNOS expression and NSF expression were also determined in the starting lysate before the isolation of S-nitrosylated proteins. Below the blot is the densitometric ratio of nitrosylated NSF to total NSF in the starting lysates.

substrate. The accumulation of a local NO pool and S-nitrosylation may regulate normal physiological processes or exert nitrosative stress (21). It is probable that the turnover of the NO pool is Iwakiri et al.

Cells and Materials. We obtained COS-7 and BSC-1 cells from ATCC (Manassas, VA). All cells were grown in DMEM supplemented with 2 mM L-glutamine, 100 units/ml penicillin, 100 ␮g/ml streptomycin, and 10% (vol/vol) FBS. DAF-2DA and DAF-2FM were from Calbiochem (San Diego, CA). Monoclonal antibodies against eNOS and Hsp90 were from BD Biosciences (San Jose, CA). S-nitrosocysteine was from A.G. Scientific, Inc. (San Diego, CA), and Myc was from Cell Signaling Technology, (Beverly, MA). Polyclonal antibody against NSF was from Santa Cruz Biotechnology (Santa Cruz, CA). Alexa Fluor 488 antimouse IgG and Alexa Fluor 680 anti-mouse IgG secondary antibodies were from Invitrogen/Molecular Probes (Carlsbad, CA). IRDye800-conjugated anti-rabbit IgG secondary antibody was from Rockland Immunochemical Inc. (Gilbertsville, PA). The cDNA of monomeric RFP was a gift from Roger Tsien at the University of California at San Diego (La Jolla, CA). eNOS Constructs. WT eNOS and eNOS-NLS fused with monomeric RFP in pCDNA3 were made to monitor the subcellular localization of eNOS, WT-eNOS-RFP, and RFP-eNOS-NLS, respectively. For WT-eNOS-RFP, RFP was inserted into the bovine eNOS cDNA in pCDNA3 at the C terminus. eNOS construct that targets nucleus was made uby sing a modified eNOS with a mutation on the myristoylation site on eNOS (eNOS G2A, thereby called ‘‘mutant’’), which therefore stays in cytosol by preventing Nmyristoylation and cysteine palmitoylation of eNOS, modifications that are required for targeting eNOS to the Golgi and plasma membrane (23, 26). Three repeats of NLSs (PKKKRKVD) was fused at the C terminus of the cDNA of eNOS (G2A). RFP was fused with eNOS (G2A)-NLS at the N terminus. Transfection. For the real-time confocal imaging of NO in living cells, COS-7 cells were seeded on coverslips and transfected the next day with 1–2 ␮g of RFP-eNOS constructs by using LipofectAmine 2000 (Invitrogen) in OptiMEM media (Invitrogen), according to the manufacturer’s instructions. Cells were used 48–72 h after the transfection. PNAS 兩 December 26, 2006 兩 vol. 103 兩 no. 52 兩 19781

CELL BIOLOGY

regulated by the balance between the protein S-nitrosylation initiated by NOS and the reduction in S-nitrosylated proteins by GSNO reductase (GSNOR) or an equivalent enzyme (38, 39). GSNOR uses GSNO as a substrate (40). As GSNO exists in equilibrium with S-nitrosylated proteins, GSNOR indirectly regulates the cellular level of S-nitrosylated proteins. In fact, mice lacking the GSNOR gene have increased basal levels of total S-nitrosylated proteins (41). It is tempting to speculate that eNOS and GSNOR may colocalize and regulate the turnover of the local NO pool and thus NO-mediated signaling and cellular functions. As mentioned before, a high NO concentration (in the presence of oxygen to form N2O3) is thought to be required for the S-nitrosylation of proteins (20, 21). We demonstrate in this study that the compartmentalization of NOS, which forms a relatively high local concentration of NO equivalents, creates a favorable environment for S-nitrosylation of proteins. However, a high concentration of NO may not be a strict requirement if S-nitrosylation involves reactions between NO and thiyl radicals or that catalyzed by transition metals reactions (15, 42). In conclusion, we directly demonstrate that eNOS channels NO locally despite the highly diffusible nature of NO. Local NO at the Golgi apparatus created by eNOS is associated with the formation of regionally confined protein S-nitrosylation and restricted actions of NO on the rate of secretion of a model protein traveling through the Golgi apparatus. Considering that the Golgi apparatus is constantly being remodeled during the cell cycle and is the site of several important posttranslational modifications, the local regulation of protein function via S-nitrosylation of proteins within the Golgi is an exciting avenue to explore.

NO Release. COS-7 cells were plated in six-well plates. Forty-eight

hours after the transfection of eNOS constructs or control plasmid, the media were processed for the measurement of nitrite, a stable breakdown product of NO in aqueous solution by chemiluminescence (33). Nitrite accumulation was measured by using a NO analyzer (Sievers Instruments, Boulder, CO). For more details see SI Text. Fluorescence Imaging. For real-time detection of NO production in living cells, the membrane-permeable fluorescent indicator DAF2DA was used (29). Once inside cells, it is deacetylated by intracellular esterases to become DAF-2 and can be detected with excitation/emission maxima of 495/515 nm, respectively. The dyeloaded cells on coverslips were incubated at 37°C for 10 min in a Hepes-buffered solution (concentrations: 130 mmol/liter NaCl, 5 mmol/liter KCl, 1.25 mmol/liter CaCl2, 1.2 mmol/liter KH2PO4, 1 mmol/liter MgSO4, 19.7 mmol/liter Hepes, and 5 mmol/liter glucose; pH 7.4) or PBS followed by an incubation with 10 ␮M DAF-2DA. Dye-loaded cells were excited with the 488 nm of a krypton/argon laser for DAF-2 and the 543 nm of a helium/neon laser for RFP, and increases in DAF-2 (for NO) and RFP (for eNOS localization) fluorescence were monitored simultaneously. The Zeiss LSM 510 Confocal Imaging System was used for the detection. For more details see SI Text. VSVG–GFP Transport Assay. BSC-1 cells, grown on 12-mm coverslips,

were microinjected with a mixture of tsO45-VSVG-GFP plasmid DNA (43) and WT-eNOS-RFP, RFP-eNOS-NLS, or RFP alone. After the microinjection, cells were incubated at 40°C for 2–3 h for the expression, then incubated at 4°C for 15 min for protein folding in the presence of cyclohexamide (100 ␮g/ml). At this point, tsO45-VSVG-GFP protein was exclusively localized at the ER. Then, incubation of cells at 32°C was started to chase the transport of tsO45-VSVG-GFP from the ER to the Golgi and the surface of plasma membrane (40 min). After incubation at 32°C for 40 min, cells were fixed in 4% paraformaldehyde for 15 min and processed for the immunofluorescence of cell surface VSVG protein, using a mAb VG (44), a kind gift from Ira Mellman (Yale University), which was labeled with Alexa 647, according to the manufacturer’s instructions (Invitrogen). The quantification of the surface VSVG protein was performed by measuring the fluorescent intensity of 1. Ignarro LJ, Buga GM, Wood KS, Byrns RE, Chaudhuri G (1987) Proc Natl Acad Sci USA 84:9265–9269. 2. Clementi E, Brown GC, Feelisch M, Moncada S (1998) Proc Natl Acad Sci USA 95:7631– 7636. 3. Kawasaki K, Smith RS, Jr, Hsieh CM, Sun J, Chao J, Liao JK (2003) Mol Cell Biol 23:5726–5737. 4. Mannick JB, Schonhoff C, Papeta N, Ghafourifar P, Szibor M, Fang K, Gaston B (2001) J Cell Biol 154:1111–1116. 5. Mannick JB, Hausladen A, Liu L, Hess DT, Zeng M, Miao QX, Kane LS, Gow AJ, Stamler JS (1999) Science 284:651–654. 6. Arnold WP, Mittal CK, Katsuki S, Murad F (1977) Proc Natl Acad Sci USA 74:3203–3207. 7. Hess DT, Matsumoto A, Kim SO, Marshall HE, Stamler JS (2005) Nat Rev Mol Cell Biol 6:150–166. 8. Stamler JS, Simon DI, Osborne JA, Mullins ME, Jaraki O, Michel T, Singel DJ, Loscalzo J (1992) Proc Natl Acad Sci USA 89:444–448. 9. Lancaster JR, Jr (1997) Nitric Oxide 1:18–30. 10. Gow AJ, Chen Q, Hess DT, Day BJ, Ischiropoulos H, Stamler JS (2002) J Biol Chem 277:9637–9640. 11. Eu JP, Sun J, Xu L, Stamler JS, Meissner G (2000) Cell 102:499–509. 12. Barouch LA, Harrison RW, Skaf MW, Rosas GO, Cappola TP, Kobeissi ZA, Hobai IA, Lemmon CA, Burnett AL, O’Rourke B, et al. (2002) Nature 416:337–339. 13. Pawloski JR, Hess DT, Stamler JS (2001) Nature 409:622–626. 14. Pawloski JR, Hess DT, Stamler JS (2005) Proc Natl Acad Sci USA 102:2531–2536. 15. Kim SF, Huri DA, Snyder SH (2005) Science 310:1966–1970. 16. Erwin PA, Lin AJ, Golan DE, Michel T (2005) J Biol Chem 280:19888–19894. 17. Erwin PA, Mitchell DA, Sartoretto J, Marletta MA, Michel T (2006) J Biol Chem 281:151–157. 18. Kim SO, Merchant K, Nudelman R, Beyer WF, Jr, Keng T, DeAngelo J, Hausladen A, Stamler JS (2002) Cell 109:383–396. 19. Jia L, Bonaventura C, Bonaventura J, Stamler JS (1996) Nature 380:221–226. 20. Griffiths C, Garthwaite J (2001) J Physiol (London) 536:855–862. 21. Gow AJ, Ischiropoulos H (2001) J Cell Physiol 187:277–282. 22. Garcia-Cardena G, Fan R, Stern DF, Liu J, Sessa WC (1996) J Biol Chem 271:27237–27240. 23. Sessa WC, Garcia-Cardena G, Liu J, Keh A, Pollock JS, Bradley J, Thiru S, Braverman IM, Desai KM (1995) J Biol Chem 270:17641–17644.

19782 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0605907103

Alexa 647, which was normalized by total expression of VSVG in cells by measuring the fluorescent intensity of GFP. S-Nitrosocysteine Immunocytochemistry. Immunostaining of Snitrosylated proteins was performed as described by Gow et al. (45) with some modifications. In brief, cells were transfected with WT-eNOS-RFP or RFP-eNOS-NLS as described. Twentyfour hours after transfection, cells were trypsinized and plated on 12-mm cover glasses with ⬇30% confluency. After the attachment, cells were incubated in Hepes buffer containing 10 ␮M oxy-hemoglobin for 2 h to block the NO released to the neighboring cells. Cells were fixed with 4% paraformaldehyde for 10 min at room temperature. After permeabilization in goat serum containing 0.3% Triton X-100, 0.1 mM neocuproine, and 1 mM EDTA, cells were incubated with a mAb against Snitrosocysteine (2 ␮g/ml) overnight at 4°C. After washing, cells were incubated with secondary antibody, Alexa 488-conjugated goat anti-mouse (1:100), in 100% goat serum at room temperature in the dark for 1 h. In a separate experiment, cells, not transfected, were incubated with 100 ␮M SNAP for 30 min, and immunostaining for S-nitrosocysteine was performed. As a negative control, cells treated with SNAP (100 ␮M) were further incubated with 0.8% HgCl2 for 1 h at 37°C, which selectively displaced NO from S–NO bonds on cysteine residue. Snitrosylated proteins were visualized by confocal microscopy. Determination of S-Nitrosylation of NSF. COS-7 cells, transfected with myc-NSF and WT-eNOS-RFP or RFP-eNOS-NLS by transient transfection as described, were used 48 h after the transfection. Cells were incubated in DMEM without serum for 6 h, followed by the incubation with DMEM containing ATP (100 ␮M) with or without 0.2% HgCl2 for 1 h. To detect S-nitrosylation of proteins, the ‘‘biotin-switch’’ method developed by Jaffrey and colleagues was used (46, 47). Purified S-nitrosylated proteins were subjected to 10% SDS/PAGE (20 ␮g protein per lane) and Western blot analysis. NSF was detected by using both anti-myc and anti-NSF antibodies. For more details see SI Text. We thank Dr. Graham Warren for helpful comments. This work was supported by National Institutes of Health Grants R01 HL64793, R01 HL 61371, R01 HL 57665, and P01 HL 70295 (to W.C.S.), National Institutes of Health Award K01 DK067933-01 (to Y.I.), and New Investigator Award P30 DK34989 from the Yale Liver Center (to Y.I.). 24. Garcia-Cardena G, Oh P, Liu J, Schnitzer JE, Sessa WC (1996) Proc Natl Acad Sci USA 93:6448–6453. 25. Garcia-Cardena G, Martasek P, Masters BS, Skidd PM, Couet J, Li S, Lisanti MP, Sessa WC (1997) J Biol Chem 272:25437–25440. 26. Liu J, Hughes TE, Sessa WC (1997) J Cell Biol 137:1525–1535. 27. Govers R, van der Sluijs P, van Donselaar E, Slot JW, Rabelink TJ (2002) J Histochem Cytochem 50:779–788. 28. Jagnandan D, Sessa WC, Fulton D (2005) Am J Physiol 289:C1024–C1033. 29. Kojima H, Nakatsubo N, Kikuchi K, Kawahara S, Kirino Y, Nagoshi H, Hirata Y, Nagano T (1998) Anal Chem 70:2446–2453. 30. Balcerczyk A, Soszynski M, Bartosz G (2005) Free Radical Biol Med 39:327–335. 31. Planchet E, Kaiser WM (2006) J Exp Bot 57:3043–3055. 32. Jourd’heuil D (2002) Free Radical Biol Med 33:676–684. 33. Fulton D, Fontana J, Sowa G, Gratton JP, Lin M, Li KX, Michell B, Kemp BE, Rodman D, Sessa WC (2002) J Biol Chem 277:4277–4284. 34. Liu X, Miller MJ, Joshi MS, Thomas DD, Lancaster JR, Jr (1998) Proc Natl Acad Sci USA 95:2175–2179. 35. Hirschberg K, Miller CM, Ellenberg J, Presley JF, Siggia ED, Phair RD, Lippincott-Schwartz J (1998) J Cell Biol 143:1485–1503. 36. Morgan A, Burgoyne RD (2004) Curr Biol 14:R968–R970. 37. Matsushita K, Morrell CN, Cambien B, Yang SX, Yamakuchi M, Bao C, Hara MR, Quick RA, Cao W, O’Rourke B, et al. (2003) Cell 115:139–150. 38. Liu L, Hausladen A, Zeng M, Que L, Heitman J, Stamler JS, Steverding D (2001) Redox Rep 6:209–210. 39. Liu L, Hausladen A, Zeng M, Que L, Heitman J, Stamler JS (2001) Nature 410:490–494. 40. Jensen DE, Belka GK, Du Bois GC (1998) Biochem J 331:659–668. 41. Liu L, Yan Y, Zeng M, Zhang J, Hanes MA, Ahearn G, McMahon TJ, Dickfeld T, Marshall HE, Que LG, Stamler JS (2004) Cell 116:617–628. 42. Stamler JS, Lamas S, Fang FC (2001) Cell 106:675–683. 43. Keller P, Toomre D, Diaz E, White J, Simons K (2001) Nat Cell Biol 3:140–149. 44. Diao A, Rahman D, Pappin DJ, Lucocq J, Lowe M (2003) J Cell Biol 160:201–212. 45. Gow AJ, Davis CW, Munson D, Ischiropoulos H (2004) Methods Mol Biol 279:167–172. 46. Jaffrey SR, Erdjument-Bromage H, Ferris CD, Tempst P, Snyder SH (2001) Nat Cell Biol 3:193–197. 47. Jaffrey SR, Snyder SH (2001) Sci STKE 2001:PL1.

Iwakiri et al.