![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 18, Issue 11, 4365-4376, November 2007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Center for Cell Signaling, Department of Biochemistry and Molecular Genetics, University of Virginia, Charlottesville, VA 22908
Submitted January 31, 2007;
Revised August 2, 2007;
Accepted August 17, 2007
Monitoring Editor: Karsten Weis
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). In response to these changes, cells rapidly initiate a stress signaling cascade and attempt to restore iso-osmolarity through transport of inorganic ions and initiate an accompanying reorganization of the actin cytoskeleton (Haussinger, 1996; Di Ciano et al., 2002). Short-term recovery of cell volume is facilitated by increasing intracellular concentrations of inorganic ions; however, elevated levels of intracellular salts can be detrimental to protein structure and function (Yancey et al., 1982; Russo et al., 2003). Long-term adaptation to hyperosmotic stress is achieved through signal transduction pathways that communicate with the metabolic and transcriptional machinery, resulting in the production or accumulation of organic osmolytes that increase intracellular osmolarity without adversely affecting protein structure or function (O'Neill, 1999).
One of the most thoroughly studied stress kinases is the mitogen-activated protein kinase (MAPK) p38. p38 and its yeast homologue Hog1p are activated in response to a wide variety of adverse environmental conditions. These include osmotic, UV, and mechanical stress. Cytokines, such as interleukins and tumor necrosis factor 
are also known to activate p38 (Ono and Han, 2000). Hog1p and p38 are both phosphorylated on a threonine, glycine, tyrosine (TGY) motif within the kinase activation loop (Thr 180 and Tyr 182 in humans). Activation in response to hyperosmotic stress occurs in the cytoplasm, but the specific mechanism of activation varies among organisms (Brewster et al., 1993; Han et al., 1994; Raingeaud et al., 1995). MAPK cascades usually involve three kinases, each successively phosphorylating the next kinase in the series, the MAPKKK phosphorylates the MAPKK, which, in turn, phosphorylates the MAPK. In higher eukaryotes, the MAPKKK MEKK3, in association with Rac and OSM, activates the MAPKK MKK3, which then activates p38 MAPK (Uhlik et al., 2003). The osmotic sensor that triggers assembly of the Rac-OSM-MEKK3 complex remains undefined. In yeast, the osmosensors are Sln1 and Sho, which regulate the MAPKKKs Ssk2 and Ste11, respectively. Both Ssk2 and Ste11 activate the MAPKK, Pbs2, which then phosphorylates Hog1 (O'Rourke et al., 2002). In yeast, Hog1 translocates to the nucleus in response to osmotic stress–induced phosphorylation (Ferrigno et al., 1998). The regulation of p38 nuclear transport is not well understood in higher eukaryotes. p38 is found in both the cytosol and nucleus before activation and is responsible for activation of transcription under stress conditions (Ben-Levy et al., 1998).
Nuclear import of Hog1p is mediated by the importin-
family member, NMD5. NMD5 binds to activated Hog1p in the cytoplasm and then translocates through the nuclear pore complex (NPC) to the nucleus where the GTP bound form of Ran (GSP1) binds to NMD5 and releases Hog1p into the nucleoplasm (Ferrigno et al., 1998). Hog1 export is exportin 1–dependent and is predicted to be exported by forming a trimeric complex with exportin 1 and RanGTP, which translocates to the cytoplasm. On reaching the cytoplasm, the Ran GTPase-activating protein, RanGAP, induces Ran catalyzed hydrolysis of GTP to GDP, resulting in disassembly of the export complex (Bischoff et al., 1994). In Saccharomyces cerevisiae, Hog1 is responsible for the activation or repression of 579 genes in response to osmotic stress (O'Rourke and Herskowitz, 2004). Nuclear transport of Hog1p/p38 MAPK is a key aspect of the cellular response to osmotic stress, because multiple nuclear targets of p38 are important for mounting a long-term response to stress (Ko et al., 2002).
An essential feature of most nuclear transport pathways is a dependence on the nucleocytoplasmic gradient of RanGTP (Izaurralde et al., 1997). As mentioned above, RanGTP is required for disassociation of import complexes and is a stoichiometric component of nuclear export complexes. The Ran guanine nucleotide exchange factor (GEF), RCC1 is localized in the nucleus, whereas RanGAP is cytoplasmic. Because of the asymmetric localization of RanGAP and RCC1, Ran in the nucleus is predominantly in the GTP bound state, whereas cytoplasmic Ran is predominantly in the GDP bound state, leading to what is known as the RanGTP gradient. RanGTP is constantly exported to the cytoplasm in association with export complexes. Cytoplasmic RanGDP is imported into the nucleus by its transport receptor, NTF2 (Ribbeck et al., 1998; Smith et al., 1998). Once in the nucleus, RCC1 binds RanGDP and catalyzes exchange of GDP for GTP. With this constant recycling, at steady state, there is both a Ran protein gradient as well as a RanGTP gradient, both of which are concentrated in the nucleus. The Ran-binding protein Mog1 has been linked genetically to the Ran-recycling pathway, between NTF2 import of Ran and the RCC1-catalyzed exchange of nucleotide on Ran (Oki and Nishimoto, 1998; Baker et al., 2001). In S. cerevisiae, deletion of Mog1 leads to temperature-sensitive growth and nuclear transport. This can be overcome by over expressing either NTF2 or Ran (GSP1; Oki and Nishimoto, 1998). A temperature-sensitive mutant of RCC1 in yeast is synthetically lethal when combined with deletion of Mog1 (Baker et al., 2001). Mog1 has also been implicated in a role in the osmo-sensing Sln1 signal transduction pathway (Lu et al., 2004). However, there are no known links between Mog1 function in the Ran gradient and its apparent links to the Sln1 pathway in yeast.
The link between Mog1 and the Sln1 pathway is part of the emerging body of evidence that there is cross-talk between the pathways for stress signaling and nuclear transport. In yeast, the arrest of secretion response (ASR), which is brought about through the use of certain Sec mutants, has been shown to result in the delocalization of some nuclear proteins. (Nanduri and Tartakoff, 2001a). Overexpression of Hog1p protects the proper nuclear localization of these proteins during the ASR. In addition, hyper-osmotic stress of yeast has been shown to cause delocalization of nuclear proteins. This occurs in a Pkc1-dependent manner and can be inhibited by overexpression of Hog1 (Nanduri and Tartakoff, 2001b).
In the course of analyzing the effect of stress signaling on nuclear transport, we found that the Ran protein gradient was disrupted in response to osmotic stress. To characterize the mechanism responsible for disruption of the Ran gradient, a battery of assays were used to define how the localization and activity of the nuclear transport machinery is affected by sorbitol stress signaling. This included using assays that measure fluorescence recovery after photobleaching (FRAP) of RCC1 and a RanGTP-sensitive biosensor that registers RanGTP production in the nucleus. We also used high-performance liquid chromatography (HPLC) and an inhibitor that inverts the GTP/GDP ratios, to define the relationship between the Ran gradient and nucleotide levels in the cell. Our results suggest that stress signaling inhibits nuclear transport in mammalian cells by reducing the production of RanGTP in the nucleus.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Small Interfering RNA
HeLa cells were transfected with siGENOME SMARTpool small interfering RNA (siRNA) against human p38 MAPK (cat. no. M-003512-05-0005, Dharmacon, Boulder, CO) using Oligofectamine (Invitrogen, Carlsbad, CA) according to the manufacturer's suggestions. The sequences of the sense strand for the siRNAs are as follows: CAAGGUCUCUGGAGGAAUUUU, GUCAGAAGCUUACAGAUGAUU, CCGCUUAUCUCAUUAACAGUU, and GUCCAUCAUUCAUGCGAAAUU. After 24 h, the cells were plated on glass coverslips, and 36 h after transfection the cells were stressed with sorbitol for the appropriate time and processed for immunofluorescence microscopy.
Immunofluorescence Microscopy
HeLa cells were grown on glass coverslips. After experimental treatment, cells were fixed in 3.75% formaldehyde for 30 min and permeabilized in 0.2% Triton X-100 for 5 min. Blocking and antibody incubations were done using a buffer containing 1x PBS, 2% bovine serum albumin, and 2% newborn calf serum.
The following primary antibodies were used: Ran (BD Transduction Laboratories, Lexington, KY), Mog1 (Steggerda and Paschal, 2001), NTF2 (Steggerda et al., 2000), RCC1 (Santa Cruz Biotechnology, Santa Cruz, CA), importin- (BD Transduction Laboratories), importin- mAb 3e9 (Chi et al., 1995), and RanGAP, pan- and phospho- p38 (Cell Signaling Technology, Beverly, MA). Secondary antibodies used included Cy3-labeled goat anti-rabbit and fluorescein isothiocyanate (FITC)-labeled donkey anti-mouse (Jackson ImmunoResearch, West Grove, PA).
Immunofluorescence images were taken using Openlab (Improvision, Lexington, MA) on a Nikon Eclipse E800 upright microscope (Melville, NY) with a Hamamatsu C4742–95 CCD and a 60x magnification objective with an NA of 1.40 (Bridgewater, NJ).
Fluorescence Resonance Energy Transfer
The YFP IBB CFP (YIC) fluorescence resonance energy transfer (FRET) sensor (Kalab et al., 2002; Li and Zheng, 2004), which contains the first 104 amino acids of Mouse Rch1 was transfected into HeLa cells, which were then stressed with 0.4 M sorbitol for the indicated time. Cells were washed with phosphate-buffered saline and fixed, and acceptor photobleaching was performed on a Zeiss LSM 510 Meta (Thornwood, NY). Fixation of the cells before FRET was found to have an effect on the absolute values of FRET efficiency, but did not change the relative behavior of the FRET sensor in response to changes in RanGTP concentration (Supplementary Figure S3).
Live Cell Analysis by Fluorescence Recovery after Photobleaching and Microinjection
For live cell imaging, HeLa cells were grown on Delta T dishes (Fisher Scientific, Pittsburgh, PA) and mounted on a Bioptechs (Butler, PA) stage warmer. For fluorescence recovery after photobleaching (FRAP) experiments, the cells were transfected with RCC1-GFP (Li et al., 2003) 24 h before the experiment. In microinjection experiments, the cells were microinjected using an Eppendorf Femtojet/Injectman NI 2 microinjector (Fremont, CA) mounted on a Zeiss 510 Meta laser scanning microscope. GST-GFP-NLS (GGNLS; Welch et al., 1999) in injection buffer (10 mM NaPO4, 70 mM KCl, and 1 mM MgCl2) was clarified and injected with an Eppendorf Femtotip. Several cells within a field were injected, the microscope was switched from wide-field mode to laser scanning confocal mode, and image collection was started. Twenty images were taken over the course of 132.7 s.
Image Quantitation
Immunofluorescence images were quantitated using Openlab (Improvision). Mean nuclear fluorescence (N) and mean cytoplasmic fluorescence (C) were measured for each field. After background subtraction, the ratio of mean nuclear fluorescence (N) to mean cytoplasmic fluorescence (C) was calculated for each field and then averages and SDs were calculated for each time point. Experiments were done at least twice, and quantitation represents a minimum of 50 cells per time point.
ImageJ (http://rsb.info.nih.gov/ij/) was used to quantitate GGNLS import, FRET, and FRAP images taken by the Zeiss LSM 510 Meta confocal laser scanning microscope with a 40x oil immersion objective with an NA of 0.55.
Mean nuclear fluorescence of GGNLS was plotted versus time after injection. Initial rate was calculated by determining the slope of the line through the linear region of the concentration versus time plot, and as such, initial rate is in fluorescence units per second. Mean fluorescence of the cytoplasm was used as the initial concentration. Each cell was then plotted as initial rate versus initial concentration. A line was fit to the data from each time point of sorbitol, utilizing all of the cells for that time point. Using the equation of the line for each sorbitol time point and assuming an ideal initial concentration of 150 fluorescence units, an initial rate of import was calculated for each time point of sorbitol stress in order to plot the changes in initial rate of import during sorbitol stress. The data shown is cumulative data from three separate experiments.
FRET efficiency was calculated as (I6 – I5/I6) x 100 in terms of In, where n is the image number. The bleach of the yellow fluorescent protein (YFP) was performed between images 5 and 6, such that FRET efficiency is an expression of the percent dequenching of the cyan fluorescent protein (CFP) acceptor (Karpova et al., 2003). Each time point is representative of 10 cells.
FRAP data were calculated as (Xn/Yn)/(Xprebleach/Yprebleach), where X is the bleached area, Y is an unbleached area in the same nucleus, and n is the image number. By normalizing to the prebleach ratio, loss of fluorescence in the nonbleached area is accounted for, and equilibration is represented by a value of 1. The experiment shown contains data from more than 50 cells.
Statistical significance in the form of p values was determined using the t test: two-sample assuming equal variances in the data analysis package in Excel (Microsoft, Redmond, WA).
Nucleotide Separation by HPLC
Cells were lysed in 0.6 M perchloric acid, neutralized with 3 M KHCO3, and separated by HPLC using a Partisil SAX (strong anion exchanger) column. The areas under the peaks of each nucleotide species were used to compare the ratio of ATP to ADP and the ratio of GTP to GDP.
Nucleotide Exchange and GAP assays
HeLa cells were stressed for 0, 5, and 30 min with 0.4 M sorbitol. A cytoplasmic extract containing RanGAP was made using 0.5% Triton X-100. The remaining nuclear material was then treated with 0.5% Triton X-100 and 500 mM NaCl to release RCC1. The extracts were dialyzed into 1x transport buffer (20 mM HEPES, pH 7.4, 110 mM potassium acetate, 2 mM magnesium acetate, and 0.5 mM EGTA) + 1 mM dithiothreitol + 2 mM sodium vanadate.
The RanGAP assay was carried out using 32P-–labeled GTP loaded on recombinant Ran (Bischoff et al., 1994). The RanGTP was incubated with 100 ng of recombinant RanGAP or 5 µg of cell extract from the 0-, 5-, or 30-min sorbitol treatment. The assay was carried out for 30 min and stopped with SDS, and the nucleotide was resolved using thin-layer chromatography (TLC). TLC was performed with TLC tank buffer (1 M formic acid, 0.5 M LiCl) using cellulose PEI (polyethylenemine) plates from J. T. Baker Chemical (4475-00; Phillipsburg, NJ).
Ran loaded with 32P-–labeled GDP was incubated with RCC1 containing extract in the presence of excess cold GTP. An aliquot was removed at 2-min intervals for 18 min, bound to a nitrocellulose filter, and washed. Radioactivity was measured with a scintillation counter.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
begins to become more nuclear between 5 and 10 min and then gradually accumulates in the nucleus. The N/C ratio of importin- fluctuates for the first 30 min of stress, but maintains a predominantly cytoplasmic localization. From 30 to 60 min, however, importin- also accumulates in the nucleus. Overall, Ran and its regulatory proteins appear to undergo similar changes in N/C ratio and accumulate in the cytoplasm in response to sorbitol. Proteins regulated by Ran, namely the importins, accumulate in the nucleus in response to the sorbitol stress, most likely due to the mislocalization of Ran.
|
|
|
and -. The FRET sensor consists of YFP and CFP separated by the importin-–binding domain (IBB) of importin- and is denoted YIC (Kalab et al., 2002; Li and Zheng, 2004). Importin- binding to the IBB in YIC prevents YFP from being in close proximity to CFP and reduces the FRET signal. In a test tube, binds to YIC but the complex is rapidly dissociated by RanGTP, resulting in a FRET signal increase. Thus, an increase in FRET signal is predicted to occur if importin- becomes limiting in the nucleus. A decrease in RanGTP levels would be predicted to decrease FRET in a single compartment, because importin- and YIC would compete for binding. However, in the context of an intact cell, and cycle between compartments, forming import complexes in the cytoplasm, whereas YIC is only found in the nucleus. If RanGTP production is reduced, import complexes formed in the cytoplasm will fail to dissociate in the nucleus, because there is insufficient RanGTP. Undissociated importin-/ complexes in the nucleus would, therefore, decrease the available importin- and increase FRET.
FRET was measured in HeLa cells expressing YIC using the acceptor photobleaching method (Karpova et al., 2003). Bleaching of the YFP acceptor causes dequenching of the CFP donor, which allows a measurement of FRET efficiency (Supplementary Figure S1). The FRET efficiency increases slightly from 0 to 10 min of sorbitol stress (Figure 4D). From 10 to 20 min, FRET efficiency decreased to near steady-state levels, followed by a sharp increase in FRET efficiency from 20 to 30 min of stress, which was maintained through 60 min of sorbitol stress. The greatly increased FRET efficiency at later stress time points represents an accumulation of importin-/ complexes in the nucleus, which are unable to bind YIC because endogenous importin- has not been released by RanGTP. These data show that RanGTP production is reduced in response to sorbitol stress.
RCC1 and RanGAP Activities Are Not Inhibited by Sorbitol Stress
We tested whether the mechanism that underlies sorbitol-dependant disruption of the Ran protein gradient involves changes in activity of the components that regulate the nucleotide state of Ran, RanGAP, and RCC1. After exposing HeLa cells to sorbitol for 0, 5, or 30 min, nuclear and cytosolic extracts were prepared as sources of RCC1 and RanGAP, respectively. RCC1 activity in nuclear extract was measured in a guanine nucleotide exchange assay whereby -32P-GDP bound to Ran is exchanged with unlabeled GTP. The nuclear extracts from control and sorbitol-treated cells showed the same levels of RCC1 activity, suggesting that the sorbitol-induced changes in the Ran protein gradient are not due to a biochemical defect in RCC1 activity (Figure 5A). Similarly, we found that RanGAP-mediated conversion of Ran--32P-GTP to Ran--32P-GDP assayed by TLC was comparable in cytosolic extracts from control and sorbitol-treated cells (Figure 5B). Sorbitol stress activates multiple signal transduction pathways (Kultz and Burg, 1998); therefore we also used two-dimensional (2D) gel electrophoresis and immunoblotting to probe for stress-induced posttranslational modifications in RCC1 and RanGAP. The isoform diversity of RCC1 and RanGAP did not appear to change in response to sorbitol stress. Under these conditions, the stress kinase p38 underwent a rightward shift reflective of its phosphorylation (Figure 5C). Though RCC1 and RanGAP are known to undergo multisite phosphorylation (Li and Zheng, 2004; Swaminathan et al., 2004), sorbitol stress affects neither the biochemical activity nor the relative levels of RCC1 and RanGAP isoforms resolved by 2D gel electrophoresis.
|
; Li et al., 2003). Photobleaching studies have shown that the interaction between RCC1 and chromatin is highly dynamic in vivo and is dependent on the ability of the RCC1/Ran complex to bind guanine nucleotide (Li et al., 2003). RCC1 dissociation from chromatin is reduced significantly if RCC1 forms complexes with Ran T24N, a mutant defective for nucleotide binding (Klebe et al., 1995). Therefore, RCC1 mobility measured by FRAP provides a sensitive readout of RCC1 interactions with chromatin, which positively regulate its catalytic function and production of RanGTP. We utilized FRAP to monitor the mobility of transiently transfected RCC1-GFP during sorbitol-induced osmotic stress (Figure 5F). A half time of recovery (T1/2) was calculated for each cell, and values were binned in 10-min intervals (Figure 5D). RCC1 mobility decreased after sorbitol stress, as did the immobile fraction of RCC1 within the nucleus (Figure 5E). This decrease in RCC1 mobility suggests that stress signaling may decrease the rate of chromatin-dependent nucleotide exchange and consequently decrease production of RanGTP. This could be due to a defect upstream of RCC1 and the exchange reaction, or it could be indicative of stress-induced changes in chromatin that lead to a decreased ability for RCC1 to properly dissociate during the exchange reaction. RCC1 mobility is not decreased in response to energy depletion by sodium azide and deoxyglucose, conditions that also lead to a loss of the Ran protein gradient (Supplementary Figure S2). Therefore, the mobility changes in response to sorbitol cannot be accounted for by changes in the nucleotide levels or loss of Ran from the nucleus.
Energy Levels Are Affected by Sorbitol Stress
RCC1 is not specific in its catalysis of nucleotide exchange; it is capable of loading either GTP or GDP on to Ran with equal efficiency (Bischoff and Ponstingl, 1991). It is believed that the loading of GTP onto Ran by RCC1 is determined by the ratio of GTP to GDP present in the cell and that GTP is loaded because it is the more abundant guanine nucleotide. Additionally, proper localization of Ran to the nucleus is dependent on the ability of RCC1 to load Ran with GTP (Ren et al., 1993). Given these observations, it is predicted that if sorbitol stress caused a decrease in the ratio of GTP to GDP, it could cause Ran to be loaded with GDP, thereby diminishing the Ran protein gradient. To investigate this possibility, we analyzed the nucleotide levels in sorbitol-stressed cells by HPLC separation of ATP, ADP, GTP, and GDP. We found that the triphosphate forms of both adenine nucleotides and guanine nucleotides decreased in relation to the diphosphate form in response to sorbitol stress (Figure 6). Interestingly, the ATP/ADP and GTP/GDP ratios show clear evidence of reduction and recovery through the sorbitol stress time course, reminiscent of the Ran N/C ratios. Immediately upon stress, Ran N/C, Mog1 N/C, ATP/ADP, and GTP/GDP ratios all decline. From 10 to 20 min of stress, the ratios of all but Mog1 undergo a partial recovery. However, from 20 to 30 min, whereas the Ran N/C ratio remains relatively constant, the ATP/ADP and GTP/GDP ratios decrease again, reaching levels similar to the 10-min time point. From 30 to 60 min Ran N/C remains relatively unchanged, whereas ATP/ADP and GTP/GDP ratios again recover toward steady-state levels. The initial drop in Ran N/C may be a result of the decrease in nucleotide triphosphate to diphosphate ratios, or available GTP, but the second drop in GTP/GDP ratios obviously does not correlate with changes in Ran nucleocytoplasmic localization.
|
). HeLa cells were treated with ribavirin (0, 10, 100, and 250 µM) overnight and subsequently analyzed by HPLC to assess changes in nucleotide levels. In parallel, HeLa cells were treated with ribavirin and processed for immunofluorescence microscopy. Ribavirin had a strong effect on GTP levels in cells, as 250 µM ribavirin produced a 75.9% reduction in cellular GTP (Figure 7A). Ribavirin had no effect on Ran and Mog1 localization at 10 µM; however, higher concentration of ribavirin actually increased N/C ratio slightly from 2.22–2.57 (p < 0.005) and 2.49–2.86 (p < 0.05) for Ran and Mog1, respectively, at 100 µM ribavirin (Figure 7, B and C). Because ribavirin had a stronger effect on the levels of GTP than on GDP, there was an inversion of the GTP/GDP ratio. In untreated cells we observed a GTP/GDP ratio of 1.88, whereas in cells treated with 250 µM ribavirin, the ratio was 0.15. These data show that neither reducing the level of GTP, nor altering the GTP/GDP ratio is sufficient to disrupt the Ran protein gradient in cells.
|
|
isoform by 
80% as assessed by immunoblotting with a pan-p38 antibody (data not shown). Because the pan-p38 antibody did not work well for immunofluorescence microscopy, phospho-p38 antibody was used in immunofluorescence experiments. We used the phospho-p38 antibody to quantify the levels of p38 and N/C ratios of Ran in the same cells over a 60-min time course of 0.4 M sorbitol treatment. Sample images from the 0- and 10-min time points indicated that cells expressing higher levels of p38 contained higher nuclear concentrations of Ran (Figure 9A). We measured the phospho-p38 and Ran levels in three sets of cell: untransfected cells (labeled Untransfected), p38 siRNA transfected cells in which p38 staining was reduced (labeled Reduced p38), and cells from the p38 siRNA transfection in which p38 staining was comparable to the untransfected culture (labeled Normal p38). Similar levels of p38 phosphorylation were induced by sorbitol in the untransfected and normal p38 cells, whereas the phospho-p38 levels in detected cells subject to knockdown remained at near-background levels throughout the time course (Figure 9B). Reducing the p38 levels in HeLa cells resulted in a reduction of the p38-dependent protection and/or recovery of the Ran protein gradient (Figure 9C). The SB203580 results and the p38 knockdown data suggest that p38 plays a role in regulating the Ran protein gradient in response to hyperosmotic stress.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Nuclear Transport and the Stress Response
Nuclear transport is critical for the cellular response to stress, which requires both MAP kinases and transcription factors. The nuclear transport and transcriptional activity of NFAT5/TonEBP is regulated by osmotic stress. Under isotonic conditions TonEBP is predominantly cytoplasmic (Tong et al., 2006). On hypertonic stress, it accumulates in the nucleus where it up-regulates the expression of several genes required for the accumulation of organic osmolytes: aldose reductase, the enzyme responsible for production of sorbitol; the sodium/myo-inositol transporter; and the betaine 
-butyric acid transporter (Ho, 2003). In hypotonic solutions, TonEBP is exported and has a predominantly cytoplasmic localization (Woo et al., 2000). In addition to TonEBP, p38 MAPK, Erk 1/2, Erk 5, and Jnk are activated and must also import into the nucleus upon sorbitol stress (Itoh et al., 1994; Rosette and Karin, 1996; Kato et al., 1997; Ferrigno et al., 1998; Yan et al., 1999; Cyert, 2001). Thus, changes that result in diminished capacity for nuclear import will diminish the cellular capacity to deal with stress.
Several links have been established between stress and nuclear transport. Oxidative stress, UV stress, heat shock, and osmotic stress have all been linked to defects in transport, predominantly through the mislocalization of nuclear transport factors or the cytoplasmic localization of nuclear proteins. UV stress and oxidative stress induced by hydrogen peroxide have been shown to cause Ran relocalization to the cytoplasm and importin- accumulation in the nucleus (Czubryt et al., 2000; Miyamoto et al., 2004). In yeast, oxidative stress has been shown to reduce NLS-dependent import (Stochaj et al., 2000). Also, osmotic stress and certain Sec mutants, which induce the ASR, are capable of delocalizing nuclear and nucleolar proteins (Nanduri and Tartakoff, 2001a,b). However, recovery of nuclear transport or nuclear transport factor localization has not been previously shown, either due to experimental design or the model system used.
The stress kinase p38 MAPK is activated by sorbitol-induced osmotic stress (Brewster et al., 1993; Han et al., 1994). When yeast experience hyperosmotic stress, the nucleolar protein Fpr3p relocalizes to the cytoplasm; however, over expression of Hog1 in this setting has a protective effect on Fpr3p nuclear localization (Nanduri and Tartakoff, 2001b). We examined the role of p38 MAPK in the nuclear transport factor response to osmotic stress using the p38 MAPK inhibitor, SB203580. In the presence of SB203580, after the addition of sorbitol, there is a drop in Ran N/C ratio, followed by a steady recovery to 60 min, bypassing the initial recovery usually seen at 20 min, but maintaining the second recovery. In yeast, Hog1 has been shown to have protective effects on nuclear transport pathways during both osmotic stress and the ASR, which is a stress brought about through the use of certain Sec mutants (Nanduri et al., 1999; Nanduri and Tartakoff, 2001a). Recovery of the Ran protein gradient in response to sorbitol stress was slowed by SB203580 and by p38 knockdown, whereas the initial delocalization of Ran was unaffected under these conditions. We therefore conclude that p38 MAPK does not have a protective role in HeLa cells, but does play a role in recovery from sorbitol stress.
The Ran Protein Gradient Is Not Strictly Dependent on Guanine Nucleotide
It is widely held that the maintenance of the Ran protein gradient is dependent on energy levels in the cell (Schwoebel et al., 2002). RCC1 will catalyze exchange of either GDP or GTP on Ran with equal efficiency (Bischoff and Ponstingl, 1991). Therefore RanGTP generation is dependent on GTP being the predominant species of guanine nucleotide present. In tsBN2 cells, RCC1 is temperature sensitive, and when grown at the nonpermissive temperature, these cells lose the Ran protein gradient, presumably due to a requirement for Ran to be in the GTP-bound form in order to concentrate in the nucleus under steady-state conditions (Ren et al., 1993). When cells are treated with sodium azide and deoxyglucose, Ran rapidly delocalizes, again presumably because of the loss of ATP and due to the nucleotide diphosphate kinase GTP levels (Schwoebel et al., 2002).
Oxidative and UV stress have both been shown to cause Ran delocalization (Czubryt et al., 2000; Kodiha et al., 2004; Miyamoto et al., 2004). It has recently been reported that the Ran delocalization in response to oxidative stress induced by hydrogen peroxide is due to a stress-induced drop in ATP levels (Yasuda et al., 2006) However, our measurements of nucleotide levels during osmotic stress showed that Ran relocalization does not strictly correlate with changes in energy levels in the cell. Although sorbitol stress caused a decrease in both ATP/ADP and GTP/GDP ratios within 10 min, mimicking the changes in Ran N/C ratios, later time points showed little correlation between Ran localization and ATP/ADP or GTP/GDP levels. Use of ribavirin to deplete guanine nucleotide levels resulted in a large drop in GTP and GDP, as well as an inversion of the GTP/GDP ratio. However, there was no significant decrease in Ran N/C ratio, contrary to what would be predicted based on tsBN2 cells. Treatment of cells with SB203580 to inhibit p38 MAPK decreased prestress nucleotide ratios without changing the N/C ratio of Ran. On sorbitol stress, the control cells showed normal decrease in Ran N/C, GTP/GDP, and ATP/ADP ratios. However, in the presence of SB203580, the Ran N/C ratio decreased similarly to the control cells, whereas the GTP/GDP and ATP/ADP ratios both increased from their lower than normal starting levels, demonstrating that the delocalization of Ran is not due to the sudden drop in nucleotide levels, because the drop in Ran nuclear levels occurs whether ATP/ADP and GTP/GDP ratios are decreasing or increasing. Collectively, these data argue for a more complex mechanism of Ran delocalization upon sorbitol stress, not explained simply by changes in energy levels.
Stress Signaling to the Nucleus
Biochemical, genetic, and cell biological data provide compelling evidence for the involvement of Mog1, NTF2, and RCC1 in the maintenance of the Ran gradient (Bischoff and Ponstingl, 1991; Tachibana et al., 1994; Paschal et al., 1997; Oki and Nishimoto, 1998; Ribbeck et al., 1998; Smith et al., 1998; Steggerda and Paschal, 2000; Baker et al., 2001). Our data support an interdependence of the localization and function of these proteins during sorbitol stress–induced disruption and recovery of the Ran protein gradient as well. The kinetics of delocalization of Mog1, NTF2, and Ran as well as the concomitant decrease in RCC1 mobility suggests that all four of these proteins are targets of stress signaling (Figure 11). Reduced function of all of these proteins in response to sorbitol could contribute to loss of the Ran protein gradient, in which case a single initiating event for disruption of the Ran protein gradient might not be apparent.
|
and - in the form of undissociated complexes, causing unoccupied YIC reporter to display FRET (Figure 4). Disruption of the Ran protein gradient would also be predicted to occur under conditions where GTP is limiting, but surprisingly, this is not the case. Cells treated with the drug ribavirin have an inverted ratio of GTP/GDP; however, this condition does not disrupt the Ran gradient. Thus, the changes in nucleotide levels that occur during stress signaling are predicted to be insufficient to cause a breakdown in the Ran protein gradient.
The response of the FRET sensor under conditions of osmotic stress may appear to be contradictory to data generated with the same construct by the Zheng lab, as well as the work done with a similar construct by the Weis lab (Kalab et al., 2002, 2006; Li and Zheng, 2004). The YIC sensor detects free importin-, and, as such does not measure RanGTP directly. Nonetheless, through a variety of controls these groups showed that a high FRET signal is correlated with high RanGTP levels and a low FRET signal is correlated with low RanGTP levels. Our data, in contrast, shows there is an increase in FRET signal under conditions where we predict there is a decrease in RanGTP levels. So why is there a discrepancy? Previous applications of the YIC sensor have been primarily in mitotic cells. In the absence of a nuclear envelope, the YIC sensor and importin- compete equally for binding to importin-. In the presence of a nuclear envelope, however, there is an unequal competition for binding to the YIC sensor. Why? 1) The YIC sensor is nuclear in interphase cells. 2) Importin- and - are shuttling, form complexes in the cytoplasm, and enter the nucleus. 3) Sorbitol stress increases the nuclear concentration of importin- and -. The reason why the YIC FRET signal increases in response to sorbitol relates to a decrease in the size of the pool of importin- available for binding YIC. The explanation that we have offered, that there is less importin- available to bind YIC because reduced RanGTP production is conducive to importin-/ complex formation (a high-affinity complex), is plausible. The alternative explanation is that sorbitol actually increases RanGTP production. However, it is difficult to reconcile such an interpretation with the fact that sodium azide and deoxyglucose addition to cells, a condition that reduces RanGTP levels, increases the YIC FRET signal (Supplementary Figure S3). It does deserve mention that the YIC signal decreases in tsBN2 cells after several hours of temperature shift (Li and Zheng, 2004), but it is possible that a gradual loss of RanGTP in that setting is not directly comparable to the acute stress signaling in our system.
Mog1 remains a mysterious player in the Ran cycle. The synthetic lethality of 
mog1 with a temperature-sensitive RCC1 allele together with the fact that Mog1 binds directly to Ran both point to a function proximal to nucleotide exchange (Oki and Nishimoto, 1998; Baker et al., 2001). We found that Mog1 undergoes delocalization from the nucleus in response to sorbitol-induced stress. This may reflect a stress-induced uncoupling of Mog1 from the Ran cycle, possibly in response to the changes in binding to and dissociating from chromatin. Defining how sorbitol-induced signaling alters the interaction of RCC1 with chromatin may hold the key to understanding how signal transduction is used to disrupt and re-establish the Ran protein gradient in cells.
|
ACKNOWLEDGMENTS |
|---|
|
Footnotes |
|---|
The online version of this article contains supplemental material at MBC Online (http://www.molbiolcell.org). 
Address correspondence to: Bryce M. Paschal (paschal{at}virginia.edu)
Abbreviations used: FRAP, fluorescence recovery after photobleaching; FRET, fluorescence resonance energy transfer; YIC, yellow fluorescent protein, importin-–binding domain, cyan fluorescent protein; GGNLS, GST-GFP-NLS; YFP, yellow fluorescent protein; CFP, cyan fluorescent protein; IBB, importin-–binding domain; N/C, ratio of mean nuclear fluorescence to mean cytoplasmic fluorescence.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Ben-Levy, R., Hooper, S., Wilson, R., Paterson, H. F., and Marshall, C. J. (1998). Nuclear export of the stress-activated protein kinase p38 mediated by its substrate MAPKAP kinase-2. Curr. Biol 8, 1049–1057.[CrossRef][Medline]
Bischoff, F. R., Klebe, C., Kretschmer, J., Wittinghofer, A., and Ponstingl, H. (1994). RanGAP1 induces GTPase activity of nuclear Ras-related Ran. Proc. Natl. Acad. Sci. USA 91, 2587–2591.
Bischoff, F. R., and Ponstingl, H. (1991). Catalysis of guanine nucleotide exchange on Ran by the mitotic regulator RCC1. Nature 354, 80–82.[CrossRef][Medline]
Brewster, J. L., de Valoir, T., Dwyer, N. D., Winter, E., and Gustin, M. C. (1993). An osmosensing signal transduction pathway in yeast. Science 259, 1760–1763.
Chi, N. C., Adam, E. J., and Adam, S. A. (1995). Sequence and characterization of cytoplasmic nuclear protein import factor p97. J. Cell Biol 130, 265–274.
Cyert, M. S. (2001). Regulation of nuclear localization during signaling. J. Biol. Chem 276, 20805–20808.
Czubryt, M. P., Austria, J. A., and Pierce, G. N. (2000). Hydrogen peroxide inhibition of nuclear protein import is mediated by the mitogen-activated protein kinase, ERK2. J. Cell Biol 148, 7–16.
Di Ciano, C., Nie, Z., Szaszi, K., Lewis, A., Uruno, T., Zhan, X., Rotstein, O. D., Mak, A., and Kapus, A. (2002). Osmotic stress-induced remodeling of the cortical cytoskeleton. Am J. Physiol. Cell Physiol 283, C850–C865.
Ferrigno, P., Posas, F., Koepp, D., Saito, H., and Silver, P. A. (1998). Regulated nucleo/cytoplasmic exchange of HOG1 MAPK requires the importin beta homologs NMD5 and XPO1. EMBO J 17, 5606–5614.[CrossRef][Medline]
Han, J., Lee, J. D., Bibbs, L., and Ulevitch, R. J. (1994). A MAP kinase targeted by endotoxin and hyperosmolarity in mammalian cells. Science 265, 808–811.
Haussinger, D. (1996). The role of cellular hydration in the regulation of cell function. Biochem. J 313, (Pt 3), 697–710.[Medline]
Ho, S. N. (2003). The role of NFAT5/TonEBP in establishing an optimal intracellular environment. Arch. Biochem. Biophys 413, 151–157.[CrossRef][Medline]
Itoh, T., Yamauchi, A., Miyai, A., Yokoyama, K., Kamada, T., Ueda, N., and Fujiwara, Y. (1994). Mitogen-activated protein kinase and its activator are regulated by hypertonic stress in Madin-Darby canine kidney cells. J. Clin. Invest 93, 2387–2392.[Medline]
Izaurralde, E., Kutay, U., von Kobbe, C., Mattaj, I. W., and Gorlich, D. (1997). The asymmetric distribution of the constituents of the Ran system is essential for transport into and out of the nucleus. EMBO J 16, 6535–6547.[CrossRef][Medline]
Kalab, P., Pralle, A., Isacoff, E. Y., Heald, R., and Weis, K. (2006). Analysis of a RanGTP-regulated gradient in mitotic somatic cells. Nature 440, 697–701.[CrossRef][Medline]
Kalab, P., Weis, K., and Heald, R. (2002). Visualization of a Ran-GTP gradient in interphase and mitotic Xenopus egg extracts. Science 295, 2452–2456.
Karpova, T. S., Baumann, C. T., He, L., Wu, X., Grammer, A., Lipsky, P., Hager, G. L., and McNally, J. G. (2003). Fluorescence resonance energy transfer from cyan to yellow fluorescent protein detected by acceptor photobleaching using confocal microscopy and a single laser. J. Microsc 209, 56–70.[Medline]
Kato, Y., Kravchenko, V. V., Tapping, R. I., Han, J., Ulevitch, R. J., and Lee, J. D. (1997). BMK1/ERK5 regulates serum-induced early gene expression through transcription factor MEF2C. EMBO J 16, 7054–7066.[CrossRef][Medline]
Klebe, C., Bischoff, F. R., Ponstingl, H., and Wittinghofer, A. (1995). Interaction of the nuclear GTP-binding protein Ran with its regulatory proteins RCC1 and RanGAP1. Biochemistry 34, 639–647.[CrossRef][Medline]
Ko, B. C., Lam, A. K., Kapus, A., Fan, L., Chung, S. K., and Chung, S. S. (2002). Fyn and p38 signaling are both required for maximal hypertonic activation of the osmotic response element-binding protein/tonicity-responsive enhancer-binding protein (OREBP/TonEBP). J. Biol. Chem 277, 46085–46092.
Kodiha, M., Chu, A., Matusiewicz, N., and Stochaj, U. (2004). Multiple mechanisms promote the inhibition of classical nuclear import upon exposure to severe oxidative stress. Cell Death Differ 11, 862–874.[CrossRef][Medline]
Kultz, D., and Burg, M. (1998). Evolution of osmotic stress signaling via MAP kinase cascades. J. Exp. Biol 201, 3015–3021.[Abstract]
Lang, F., Busch, G. L., Ritter, M., Volkl, H., Waldegger, S., Gulbins, E., and Haussinger, D. (1998). Functional significance of cell volume regulatory mechanisms. Physiol. Rev 78, 247–306.
Leyssen, P., Balzarini, J., De Clercq, E., and Neyts, J. (2005). The predominant mechanism by which ribavirin exerts its antiviral activity in vitro against flaviviruses and paramyxoviruses is mediated by inhibition of IMP dehydrogenase. J. Virol 79, 1943–1947.
Li, H. Y., Wirtz, D., and Zheng, Y. (2003). A mechanism of coupling RCC1 mobility to RanGTP production on the chromatin in vivo. J. Cell Biol 160, 635–644.
Li, H. Y., and Zheng, Y. (2004). Phosphorylation of RCC1 in mitosis is essential for producing a high RanGTP concentration on chromosomes and for spindle assembly in mammalian cells. Genes. Dev 18, 512–527.
). Coverslips from cultures transfected with p38 siRNA were used to select cells with reduced and normal p38 levels. The plot labeled reduced p38 (
) refers to cells transfected with p38 siRNA that showed reduced phospho-p38 staining. The plot labeled normal p38 (
) refers to cells transfected with p38 siRNA that showed approximately the same level of phospho-p38 staining as untransfected cells. Error bars, SD. (C) N/C ratios of Ran in the same cells in which phospho-p38 levels were determined. Error bars, SEM.
