JBC Connect with Cosmo for Collagen Detection

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M607954200 on April 12, 2007

J. Biol. Chem., Vol. 282, Issue 28, 20621-20633, July 13, 2007
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
282/28/20621    most recent
M607954200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lin, Q.
Right arrow Articles by Dennery, P. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lin, Q.
Right arrow Articles by Dennery, P. A.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Heme Oxygenase-1 Protein Localizes to the Nucleus and Activates Transcription Factors Important in Oxidative Stress*

Qing Lin{ddagger}§1, Sebastian Weis1, Guang Yang{ddagger}1, Yi-Hao Weng, Rachel Helston||, Kimberly Rish||, Ann Smith||, Jessica Bordner{ddagger}, Tobias Polte, Frank Gaunitz**, and Phyllis A. Dennery{ddagger}§2

From the {ddagger}Children's Hospital of Philadelphia and §Department of Pediatrics, University of Pennsylvania, Philadelphia, Pennsylvania 19104, the Department of Pediatrics, Stanford University, Palo Alto, California 94305, the ||School of Biological Sciences, University of Missouri, Kansas City, Missouri 64110-2499, and **Universität Leipzig, Interdisziplinäres Zentrum, 04107 Leipzig, Germany

Received for publication, August 18, 2006 , and in revised form, March 29, 2007.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Heme oxygenase-1 (HO-1), the rate-limiting enzyme in heme degradation, is an integral membrane protein of the smooth endoplasmic reticulum. However, we detected an HO-1 immunoreactive signal in the nucleus of cultured cells after exposure to hypoxia and heme or heme/hemopexin. Under these conditions, a faster migrating HO-1 immunoreactive band was enriched in nuclear extracts, suggesting that HO-1 was cleaved to allow nuclear entry. This was confirmed by the absence of immunoreactive signal with an antibody against the C terminus and the lack of a C-terminal sequence by gas chromatographymass spectrometry. Incubation with leptomycin B prior to hypoxia abolished nuclear HO-1 and the faster migrating band on Western analysis, suggesting that this process was facilitated by CRM1. Furthermore, preincubation with a cysteine protease inhibitor prevented nuclear entry of green fluorescent protein-labeled HO-1, demonstrating that protease-mediated C-terminal cleavage was also necessary for nuclear transport of HO-1. Nuclear localization was also associated with reduction of HO activity. HO-1 protein, whether it was enzymatically active or not, mediated activation of oxidant-responsive transcription factors, including activator protein-1. Nevertheless, nuclear HO-1 protected cells against hydrogen peroxide-mediated injury equally as well as cytoplasmic HO-1. We speculate that nuclear localization of HO-1 protein may serve to up-regulate genes that promote cytoprotection against oxidative stress.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Heme oxygenase (HO)3 catalyzes the degradation of heme and the formation of biliverdin and carbon monoxide. It is highly inducible in response to various stimuli, including oxidative stress, heavy metals, UV radiation, and inflammation (1-4). Cytoprotective roles for HO have been demonstrated in many models; however, the mechanisms by which this occurs are still under intensive study. Many have speculated that either heme catabolites, such as biliverdin, or its derivative, bilirubin, and carbon monoxide or the degradation of the pro-oxidant heme results in cytoprotection against oxidative stress (5-7). Nevertheless, all of the by-products of the HO reaction, despite being potentially cytoprotective, are also cytotoxic. Bilirubin is a potent neurotoxin (8), as is carbon monoxide (9). Furthermore, the HO reaction releases iron, which could interact with cellular oxidants to generate the hydroxyl radical (10). Transfection with an inactive HO-1 mutant protein results in cytoprotection against chemically induced oxidative stress (11). Because this effect of the mutant HO-1 could not be attributable to changes in heme catabolites, it alludes to a role for the HO-1 protein itself. Furthermore, the inactive form of HO-1 increased catalase and glutathione content (11). This suggests that the HO-1 protein itself may play a role in cellular signaling. If this were true, HO-1 would need to migrate to the nucleus or produce nuclear changes that affect transcription. There are several examples of cytoplasmic enzymes serving in nuclear functions. The steroid regulatory element-binding protein is usually bound to the smooth endoplasmic reticulum (sER) at its C terminus and locates to the nucleus after proteolytic cleavage (12). The amphitrophic protein, CTP:phosphocholine cytidylyltransferase is activated when bound to the sER but is enzymatically inactive in a nuclear reservoir (13). The transcription factor ATF6 is a type II transmembrane (TM) glycoprotein that is usually anchored to the sER, but it can be proteolytically cleaved at its N terminus and migrate to the nucleus in response to ER stress (14).

HO-1 was first identified as an integral type I membrane protein of the sER, oriented such that most of the protein, including the active site, resides in the cytoplasm. A short C terminus resides within the lumen of the sER (15). Translocation of HO-1 to the nucleus would require one or more proteolytic cleavages, on the cytosolic side of the TM region and/or within the TM domain to release a large HO-1 fragment, containing the N terminus, into the cytosol. In vitro experiments document tryptic cleavage of the C terminus of purified HO-1 protein (16). Whether this occurs in intact cells is still unknown. Most proteins that migrate to the nucleus have nuclear localization sequences (NLSs) that are essential for nuclear uptake (reviewed in Ref. 17). Nonetheless, small proteins (less than 50 kDa) that lack an NLS do migrate to the nucleus (18). So far, no NLS has been identified on HO-1. In addition to NLSs, nuclear export sequences (NESs) are also important for the proper localization of some cytoplasmic proteins (19). These are leucine-rich regions, which bind with the export receptor CRM1, forming a complex with Ran-GTP. This allows for passage through the nuclear pore (20). Leptomycin-B (LMB) can inhibit the binding of CRM1 to the NES region (19, 20). This block in nuclear export results in a predominant nuclear expression of a protein. No one has evaluated whether HO-1 has a functional NES; however, there is a highly conserved region on the rat HO-1 protein at amino acids 207-221 (LNIELSEELQALL) with greater than 90% homology to the NES motif (LX1-3LX2-3LXL) on the human immunodeficiency virus type 1 Rev protein (HIV-Rev), which binds to unspliced HIV-1 pre-mRNA and exports it from the nucleus (21). Once HO-1 accesses the nucleus, it is important to understand how it affects cellular functions.

One likely signaling mechanism is modulation of gene transcription. Kravets et al. (22) reported that biliverdin reductase, which catalyzes reduction of the HO activity product, biliverdin, to bilirubin, can induce the expression of HO-1 protein, suggesting that various components of the bilirubin system could regulate the expression of genes in the pathway. This occurred through activation of transcription factors found within the genomic sequence upstream of the HO-1 gene. Distal enhancer 1 (DE1) and DE2 located at -4 and -10 kb upstream of the transcription start site, respectively, are critical for HO-1 induction by most oxidative stimuli, including heme, heavy metals, and hydrogen peroxide (2, 23-25). Both the DE1 and DE2 regions contain multiple stress-responsive elements that represent binding sites of regulatory proteins, such as activator protein (AP)-1, Jun, cAMP-response element-binding protein, Maf, and the Cap'n'collar/basic leucine zipper (CNC-bZIP) transcription factors (26, 27). Unlike biliverdin reductase, HO-1 is not a transcription factor; therefore, it is more likely to have an indirect effect on gene transcription, perhaps through modulation of transcription factor binding.

Here, we demonstrate that, in response to three different stimuli (hypoxia (3% oxygen) or incubation with hemin or heme-hemopexin (H/HPX)), HO-1 can migrate to the nucleus. Nuclear translocation is associated with truncation of the C terminus of HO-1. In addition, a motif, which we term a putative "nuclear shuttling sequence (NSS)," is important for nuclear import of HO-1. Furthermore, C-terminal truncation of the protein and nuclear migration is also modified by proteolytic cleavage. Nuclear migration resulted in loss of HO activity. We further show that inntracytoplasmic delivery of HO-1 protein activated several transcription factors involved in oxidative stress and that delivery of HO-1 protein or transfection of HO-1 cDNA resulted in activation of a 15-kb oxidant-responsive HO-1 promoter attached to luciferase and protected against hydrogen peroxide-mediated injury, whether the HO-1 protein was active or not.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Cell Lines
The NIH3T3 mouse fibroblast (3T3) cell line was obtained from the American Type Culture Collection (ATCC) (Manassas, VA). The cells were grown in Dulbecco's minimal essential medium (DMEM) (Invitrogen) supplemented with 10% FBS and 1% antibiotic-antimycotic. Mouse hepatoma cells (Hepa) were cultured in DMEM containing 2% FBS and 0.5% gentamicin as described (28). Human HEK293 cells were stably transfected with a FLAG tag at the N terminus of the rat HO-1 cDNA (293 FLAGHO-1) and were similarly maintained in DMEM as with the 3T3 cells. NIH3T3 cells stably transfected with a 15-kb oxidant-responsive HO-1 promoter-luciferase construct (3T3-HO-1/luc cells) were also maintained in DMEM. To obtain mouse embryonic fibroblasts (MEF), mice heterozygous for HO-1 mutant allele were timely mated. Embryonic day 13.5 embryos were collected and minced in PBS. The minced tissue was placed in trypsin/EDTA and incubated at 37 °C for 10 min. The suspension was spun at 1000 rpm for 5 min at 10 °C. The pellet was resuspended in DMEM with 10% FBS, 1% nonessential amino acids, and 1% antibiotic/antimycotic and plated on a cell culture flask. The medium was changed the next day to remove cellular debris. Cells were maintained at subconfluence by passaging every 3-4 days with 0.05% trypsin-EDTA. Approximately 48 h prior to the start of experiments, cells were plated to a confluence of 60-70%.

Cell Culture Conditions
All cells were grown in a 5% CO2 humidified atmosphere at 37 °C in 75-cm2 tissue culture flasks.

Models to Modify HO-1 Expression
Hypoxic Exposure—3T3 and the 293 FLAGHO-1 cells were exposed to hypoxia (3% O2) for 0-48 h.

Incubation with Hemin in Vitro—3T3 cells were incubated with hemin 30 µM for 0-48 h.

Incubation with H/HPX in Vitro—Intact rabbit hemopexin was purified, and stoichiometric 1:1 H/HPX complexes (>90-95% saturation) were characterized and quantitated using heme dissolved in Me2SO, as previously described (28, 29). The extinction coefficients used were 1.1 x 105 M-1 cm-1 at 280 nm for apohemopexin, 1.2 x 105 M-1 cm-1 at 280 nm, and 1.4 x 105 M-1 cm-1 at 405 nm for rabbit mesoheme-hemopexin. Mesoheme was used, because it is more soluble than protoheme and has been shown to have similar biological and regulatory effects in the hemopexin system (30). Heme-hemopexin complexes were dialyzed against PBS at 4 °C before use. While in exponential growth, Hepa cells were rinsed and then incubated for up to3hin serum-free HEPES-buffered DMEM, pH 7.4, supplemented with heme-hemopexin (10 µM), heme (10 µM), or equivalent volumes of their control solvents: PBS or Me2SO, respectively. Heme induction of HO-1 in whole cell extracts was used as a control. For this, cells were incubated in serum-free HEPES-buffered DMEM, pH 7.4, supplemented with heme (10 µM) for 3 h.

Preparation of Nuclear Extracts
This was performed according to published methods (31) with modifications (32).

Identification of Regions of the HO-1 Protein Necessary for Nuclear Translocation
To determine which regions of the protein are responsible for nuclear localization, various rat HO-1 cDNA constructs were inserted into a pEGFP-c1 vector (Clontech, Palo Alto, CA). A mutant expressing an HO-1 protein lacking the C-terminal amino acids 267-289 fused to EGFP protein (EGFPHO-1C{Delta}23) was generated with appropriate primers (see Table 1). The PCR product was cloned into the BglII/BamHI site of the pEGFP-C1 vector.


View this table:
[in this window]
[in a new window]

  		TABLE 1
Primer sequences for HO-1 constructs

 
A mutant expressing HO-1 protein lacking amino acids 207-221 with >90% homology to a known NES (termed NSS; EGF-PHO-1{Delta}NSS) was also generated in a two-step process with PCR primer pairs (see Table 1). The first PCR product was cloned into the BglII/KpnI site, and the second PCR fragment was cloned into the KpnI/BamHI site of the pEGFP-c1 vector. A double deletion mutant lacking the C terminus and the NSS (EGFPHO-1C{Delta}23{Delta}NSS) was generated as above with substitution of the last PCR primer as shown in Table 1.

Full-length rat HO-1 cDNA (1-867 bp) was amplified by PCR (see Table 1). The PCR product was cloned into the BamHI/XhoI site of pcDNA3.1/His (Invitrogen) to generate the HisHO1myc construct. The HisHO1myc fragment was then subcloned into p3XFLAGCMV (Sigma) to generate HisHO1FLAG. A rat HO-1 cDNA (HisHO1mutFLAG) expressing an enzymatically inactive HO-1 with a substitution of histidine 25 to alanine was derived from the HisHO1FLAG by PCR using a site-directed mutagenesis kit from Stratagene (QuikChange II; catalog number 200523-4).

These constructs were verified by DNA sequencing and then transfected into 3T3 cells using Lipofectamine 2000 (Invitrogen). The transfected cells were grown on slides and visualized by conventional fluorescence microscopy or laser confocal microscopy as previously described (33). Fluorescence in the 484 nm excitation and 510 nm emission ranges was used to detect EGFP.

Evaluation of the Putative HO-1 NSS in Nuclear Localization
To test whether the NSS alone mediates compartmentalization of protein into the nucleus, an oligonucleotide containing the NSS sequence corresponding to amino acids 206-222 of rat HO-1 was generated (see Table 1). This sequence was fused to the C terminus of the EGFP protein (NSSEGFP), and the sequence was confirmed by DNA sequencing.

The NSS EGFP or EGFP were transiently transfected into the 3T3 cells. The localization of the EGFP protein was determined using immunocytostaining with the GFP antibody (Molecular Probes, Inc. Eugene, OR). Co-staining with the cytoplasmic protein marker calnexin and nuclear marker 4',6-diamidino-2-phenylindole was performed to demonstrate the subcellular localization using fluorescence microscopy.

Evaluation of the Role of Proteases on HO-1 Subcellular Localization
Cultured 3T3 cells were exposed to hypoxia as described above. The media were changed to DMEM containing 0.5% FBS. Cells were then incubated with a 50 µM concentration of the cysteine protease inhibitor E64d (catalog number E 3132; Sigma) 20 min prior to exposure to hypoxia. Cells were visualized by fluorescent microscopy after immunohistochemical staining with HO-1 antibodies. Cell lysates and nuclear extracts were also subjected to Western analysis.

Detection of Nuclear HO-1 Protein
An antibody specific for the C-terminal amino acids of the HO-1 protein (M19) was obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). An N-terminal antibody to a 30-residue synthetic peptide based on the human HO-1 (SPA-896) was obtained from Stressgen (Assay Designs).

In order to characterize the nuclear HO-1, 293 FLAGHO-1-transfected cells were exposed to hypoxia. The nuclear FLAGHO-1 was isolated and enriched using a FLAG affinity column as per the manufacturer's instruction (anti-FLAG M2; catalog number A 2220; Sigma).

Identification of Nuclear HO-1 Sequence and Molecular Mass
Molecular mass of the nuclear HO-1 was determined using MALDI-TOF mass spectrometry (ABI/PerSeptive Voyager DE-PRO MALDI-TOF; Framingham, MA). Identity of the protein was obtained by trypsin digestion of the colloidal blue-stained protein band followed by liquid chromatography/MS/MS analysis (ThermoElectron LTQ ion trap mass spectrometer; Wistar Proteomic Core Facility, Wistar institute, Philadelphia, PA). The resulting masses and MS/MS spectra were searched against the nonredundant NCBI data base using the TurboSEQUEST browser.

Determination of HO Enzymatic Activity
Heme oxygenase activity was measured by detecting the amount of carbon monoxide generated from cell lysates, as previously described (34).

Intracytoplasmic Delivery of HO-1 Protein
A GST HO-1 fusion construct was expressed in the Escherichia coli strain BL21 (Invitrogen). Bacteria were grown to an OD of 0.6-0.8. Thereafter, the fusion protein was induced with 100 µM isopropyl 1-thio-beta-D-galactopyranoside at 30 °C. After 5 h, bacteria were harvested and sonicated for 5 min. The fusion protein was purified using a GST purification module (Amersham Biosciences) according to the manufacturer's instructions.

Ten µg of GST HO-1 protein was delivered in the 3T3-HO-1/luc or MEF cells using the Pro-JectTM system (Pierce) as per the manufacturer. Briefly, the Pro-JectTM reagent was solubilized in 250 µl of methanol. The methanol was evaporated overnight under a sterile hood, and the dried Pro-JectTM was stored at -20 °C until use. Purified HO-1 protein was diluted in PBS and added to the dried Pro-JectTM reagent. After a 5-min incubation, the mixture was added to the cell culture medium with 5% serum. After 3 h, cells were washed with PBS and maintained in serum-free medium for the first4hofthe incubation. Thereafter, 5% FBS was reincorporated. In other experiments, purified mutant HO-1 protein (H25A) devoid of catalytic activity (gift of Paul Ortiz de Montellano, University of California, San Francisco) was delivered to the cells as described above.

Transient Transfection with HO-1 cDNA
Transient transfection was performed in NIH3T3 cells using Lipofectamine 2000 (Invitrogen). Briefly, 1 day before transfection, 105 cells were seeded in 24-well plates with antibiotic-free growth medium. DNA (0.8 µg) and 2 µl of Lipofectamine 2000 reagent were diluted into 50 µl of DMEM separately. After a 5-min incubation, the DNA and Lipofectamine solution were mixed and incubated for additional 20 min at room temperature. The DNA-Lipofectamine complex was then added to the cells. Cells were grown for 48 h before being subjected to assays.

Evaluation of Transcription Factor Activation
To determine whether HO-1 modulates cellular signaling, lysates from cells receiving GST-purified wild type and mutant HO-1 protein were evaluated for changes in transcription factor activation compared with controls. This was done using a commercially available kit (TransSignalTM Protein/DNA array I, Panomics, Redwood City, CA) according to the manufacturer's instructions. This array did not include Nrf2 or Bach-1.

Verification of DNA Binding to Transcription Factors
Activations observed with the transcription factor array were verified by electrophoretic mobility shift assay (EMSA) using the appropriate transcription factor consensus DNA binding sequence.

After protein delivery, cells were harvested, and the nuclear extract was obtained using a commercially available kit (Pierce) after the addition of a protease inhibitor mixture solution (Sigma). The 32P-labeled probes were incubated with 20 µg of the nuclear protein in a buffer containing 10 mM HEPES (pH 7.9), 1 mM EDTA, 80 mM KCl, 1 µg of poly(dI-dC), and 4% Ficoll. The reaction mixture was incubated at room temperature for 30 min and electrophoresed on 6% polyacrylamide gels. To distinguish nonspecific binding of the nuclear proteins, competition reactions were performed by adding either a 10- or 100-fold excess of nonradiolabeled probe or of a mutated probe to exclude nonspecific binding.

Oligonucleotides with the consensus DNA binding sequences for the given transcription factors were synthesized by the PAN Facility at Stanford University (see Table 2).


View this table:
[in this window]
[in a new window]

  		TABLE 2
Transcription factor consensus sequences used for EMSA

 
Visualization of Oxidant-responsive Promoter Activation
After transfection or protein delivery, the 3T3-HO-1/luc cells were incubated with DMEM containing 1% luciferin for 5 min. The cultures were imaged using the IVIS (In Vivo Imaging System) camera system (Xenogen, Alameda, CA). Pseudoimages of the photon emission were generated, and light intensity was expressed as a ratio to cell counts, as previously described (32).

Statistical Analysis
For comparison between treatment groups, the null hypothesis that there is no difference between treatment means was tested by a single factor analysis of variance for multiple groups or unpaired t test for two groups (Statview 4.02; SAS, Berkeley, CA). Statistical significance (p < 0.05 or p < 0.001) between and within groups was determined by means of the Fischer method of multiple comparisons.


Figure 1
View larger version (58K):
[in this window]
[in a new window]

  		FIGURE 1.
HO-1 localizes to the nucleus after hypoxia exposure. NIH3T3 cells were grown on slides and exposed to 3% oxygen (hypoxia) for 24 h. A, immunocytostaining using antibody specific to the N terminus of HO-1 (SPA 896; green signal) is shown. Co-localization with 4',6-diamidino-2-phenylindole (DAPI) (blue) is indicated by the cyan color. Lack of staining with secondary antibody alone is shown on the upper right (2°Ab). B, nuclear and cytoplasmic fractions were isolated and subjected to Western analysis using N-terminal (N-term; SPA 896) or C-terminal (C-term; M19) antibodies. The marker lane (M) defines the HO-1 immunoreactive protein size. N, normoxia exposure; H, hypoxia exposure. C, nuclear staining of HO-1 (green) with N or C terminus-specific antibodies.

 


Figure 2
View larger version (52K):
[in this window]
[in a new window]

  		FIGURE 2.
HO-1 localizes to the nucleus after incubation with heme or H/HPX. Hepa cells were cultured on Matrigel-coated glass coverslips for 40 h and incubated in serum-free medium containing 10 µM heme for 4 h or 20 µM H/HPX for 5 h. A, immunocytostaining using antibody specific to the N terminus of HO-1 (SPA 896; green signal) is shown. B, representative co-localization (yellow) of HO-1 (green) with propidium iodide (red) using confocal microscopy through stacked Z series. Antibodies specific to the N terminus (SPA 896) and C terminus (M19) of the HO-1 were used to demonstrate the differential nuclear localization. C, nuclear and cytoplasmic fractions were isolated from the Hepa cells incubated with hemin. These were subjected to Western analysis using antibodies against the N and C terminus of HO-1. The small arrow indicates the native HO-1 protein migrating at 32 kDa, and the large arrow indicates the faster migrating (28 kDa) immunoreactive HO-1 signal.

 

   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
HO-1 Protein Localizes to the Nucleus after Hypoxic Exposure—In cultured 3T3 cells, nuclear localization of HO-1 protein was observed after hypoxic exposure (Fig. 1A). The HO-1 signal was enriched in the nucleus as with other known nuclear proteins, such as acetylated histone 3 and CRM1 (data not shown).

In the sER, HO-1 is oriented such that most of the protein, including the active site, resides in the cytoplasm. A short C terminus resides within the lumen of the sER (35). With hypoxic exposure, two HO-1 immunoreactive bands were observed upon Western analysis in the whole cell extract, with one band migrating at ~28 kDa and the other migrating at 32 kDa (Fig. 1B). Using the M19 antibody directed toward the C terminus of HO-1, loss of all HO-1 immunoreactivity was noted in the nuclear extracts, whereas one immunoreactive band migrating at 32 kDa was observed in the whole cell extracts (Fig. 1B). These data suggest that the nuclear HO-1 immunoreactive form does not contain the C terminus. Lack of nuclear localization of the M19 immunoreactive signal by fluorescence microscopy provides further evidence that the nuclear form of HO-1 is devoid of the C terminus (Fig. 1C).

HO-1 Protein Localizes to the Nucleus after Hemin or H/HPX Incubation—To address whether the generation of a nuclear form of HO-1 was a general phenomenon, we also investigated whether other known inducers of HO-1 had a similar effect on HO-1 nuclear migration as hypoxia. In cultured Hepa cells, increased signal intensity and nuclear localization of HO-1 immunoreactive protein was observed after incubation with 10 µM heme for 4 h or 20 µM H/HPX for 5 h (Fig. 2A). Using confocal microscopy and co-staining with propidium iodide, the M19 immunoreactive signal was never localized to the nucleus but rather remained in the perinuclear area (Fig. 2B). Additionally, nuclear extracts were enriched with a faster migrating band (~28 kDa) and were devoid of the native 32-kDa HO-1 immunoreactive band upon Western analysis, corroborating that the nuclear form of HO-1 is truncated at the C terminus (Fig. 2C).

The C-terminal Region of HO-1 Inhibits Nuclear Import—To further analyze the role of the C terminus of HO-1 protein using immunocytostaining, we evaluated nuclear localization of EGFP protein fused to HO-1 fragments lacking the C-terminal 23 amino acids. Transfection with the full-length EGFPHO-1 cDNA resulted in cytoplasmic fluorescence, whereas nuclear fluorescent signal was associated with transfection of the C-terminal mutant lacking the last 23 amino acids (C{Delta}23) (Fig. 3A).

There is a highly conserved leucine-rich region with strong homology for the HIV-Rev NLS (21) near the C terminus of HO-1 protein, referred to as NSS (see the crystal structure in Fig. 3). Transfection of a mutant with double deletion lacking the NSS and C{Delta}23 (C{Delta}23{Delta}NSS) yielded cytoplasmic fluorescence (Fig. 3A), indicating that the NSS functions to facilitate nuclear import. Paradoxically, when the NSS fused to GFP it behaved more as a putative NES (Fig. 3B), suggesting that other elements of the HO-1 protein are also required to mediate the NSS function.


Figure 3
View larger version (38K):
[in this window]
[in a new window]

  		FIGURE 3.
HO-1 nuclear localization is mediated by the C terminus and a putative NSS. A, various mutants with or without the NSS or the C-terminal 23 amino acids were fused to the pEGFPc1 vector. The plasmids were transfected into 3T3 cells. The upper left panel demonstrates the GFP signal in live cells after transfection. Locations of the deletion and truncation (arrows) are indicated in relation to the crystal structure of the rat HO-1 protein (middle panel) (accession number NM_012580.2). The C- and N-terminal regions on the crystal structure are indicated (C and N, respectively). Note that none of the deletions or truncation involves the heme binding site (H). The right panel shows a schematic representation of the HO-1 cDNA deletion and truncation. Hatched box, EGFP; solid box, HO-1 cDNA mutants. FL, full-length HO-1 cDNA; C{Delta}23, HO-1 cDNA without the C-terminal 23 amino acids; C{Delta}23{Delta}NSS, HO-1 cDNA without the C-terminal 23 amino acids and the NSS sequence. B, cultured 3T3 cells were grown on slides and transiently transfected with NSSpEGFPc-1 or pEGFPc-1 alone. Cells were fixed with methanol and stained with GFP (green), calnexin (red), and DAPI (blue) to demonstrate localization of NSS EGFP and EGFP.

 
Typically, NES-like leucine rich regions bind to CRM1. We also demonstrated interaction of HO-1 protein with CRM1 by recovering 35S label in an in vitro translated HO-1 protein after co-incubation with GST-CRM1 using a pull-down assay (Fig. 4A). However, deletion of the putative NSS did not completely eliminate HO-1 binding to CRM1, indicating that there are probably other CRM1 binding regions on the protein. CRM1-mediated nuclear export can be effectively blocked by LMB. We therefore tested whether preincubation of cultured cells with LMB, and exposure to hypoxia could result in the exclusion of the HO-1 immunoreactive signal from the nucleus. This was in fact observed (Fig. 4B) as well as loss of the more rapidly migrating (28-kDa) immunoreactive HO-1 band on Western analysis after both hypoxia and hemin incubation (Fig. 4C).

The Nuclear Form of HO-1 Lacks 52 Amino Acids from the C Terminus—To verify loss of the C terminus and to determine the actual cleavage site of the HO-1, the nuclear fraction of the 293 FLAGHO-1 cells was isolated using a stringent method. The nuclear fraction showed no cytoplasmic contamination, as verified by exclusion of calnexin (data not shown). After purification with a FLAG column, the eluted contents showed a predominant band (arrow) between 30 and 45 kDa on a colloidal blue-stained SDS-polyacrylamide gel (Fig. 5). This was further verified to be immunoreactive with both FLAG and HO-1 antibodies (data not shown). MALDI-TOF mass spectroscopy showed a peak with a molecular mass of 34,071.64 Da. Trypsin-digested peptide sequences from this band revealed an intact N terminus and lack of C terminus of the HO-1 using liquid chromatography/MS/MS. After deduction of the FLAG molecular weight, the truncated HO-1 mass was 27,185 Da, corresponding to the amino acids 1-237 of rat HO-1 protein.

Protease Inhibition Reduces Nuclear Localization of HO-1—Calpains are ubiquitous calcium-dependent cysteine proteases that regulate several proteins involved in cell cycle progression, including p27Kip1 (19) as well as c-Fos and c-Jun (36). Calpain-mediated cleavage of proteins, such as amyloid precursor protein found in Alzheimer disease, results in the formation of fragments that are targeted to the nucleus and regulate gene expression (37). To understand whether truncation of HO-1 at the C terminus could be due to proteolytic cleavage by proteases, such as calpains, 3T3 cells were preincubated with a cysteine protease inhibitor (E64d) prior to hypoxic exposure. The E64d compound is esterified and diffuses readily into the cytoplasm to inhibit calpain activity (38). This obviated nuclear translocation of HO-1 immunoreactive signal (Fig. 6A). In addition, preincubation with E64d visibly reduced the intensity of the faster migrating band on Western analysis, further documenting that this band was the result of proteolytic cleavage of the HO-1 protein (Fig. 6B). In contrast to these observations, nuclear translocation of HO-1 after hemin could not be obviated by incubation with catalase (0.3 µM) or superoxide dismutase (0.6 µM) (data not shown), suggesting that the cleavage was not a result of direct oxidative damage to the HO-1 protein.


Figure 4
View larger version (44K):
[in this window]
[in a new window]

  		FIGURE 4.
The NSS regulates HO-1 nuclear localization. A, the interaction of HO-1 with CRM-1 is demonstrated by GST pull-down assays. A plasmid expressing a GST CRM1 protein in a pGEX 4T1 vector was used to generate the GST CRM1 fusion protein. This fusion protein yielded a single band on SDS-PAGE (data not shown) after E. coli amplification. Equal amounts of the cell lysates were loaded onto various GST minicolumns to allow for binding. The deletion and truncation mutants of HO-1 as well as a random deletion construct were in vitro translated in the presence of [35S]methionine. After the binding reaction, HO-1 (arrow) was eluted with GST and equally loaded to a gradient gel. A representative radiograph is shown in the upper left panel. On the middle panel, equal loading of the in vitro translated HO-1 products was verified by radiography prior to the binding reaction. The right panel shows a schematic representation of the HO-1 cDNA deletion and truncation. FL, full-length HO-1 cDNA; FL{Delta}NSS, full-length HO-1 cDNA without NSS sequence; FL{Delta}Random, full-length HO-1 cDNA without a sequence corresponding to 444-462 bp on rat HO-1; C{Delta}23{Delta}NSS, HO-1 cDNA without the C-terminal 23 amino acids and NSS sequence. B, cultured 3T3 cells were incubated with LMB prior to hypoxic exposure. HO-1 immunocytostaining is shown. A representative Western analysis demonstrates HO-1 immunoreactive signal after preincubation of LMB to block CRM1-mediated nuclear shuttling in 3T3 cells exposed to hypoxia. C, representative Western blot showing HO-1 immunoreactive protein in NIH3T3 cells exposed to hypoxia for 4 h (left) and in Hepa cells incubated with hemin for 4 h (right). The 32-kDa HO-1 immunoreactive signal is indicated by the small arrow, and the more rapidly migrating immunoreactive signal (~28 kDa) is shown by the large arrow. In both examples, {alpha}-actin staining is shown to demonstrate equal loading.

 
Nuclear Translocation of HO-1 Reduces HO Activity—To understand the effect of nuclear localization on HO-1 activity, cell lysates from 3T3 cells exposed to hypoxia were evaluated for CO formation in the presence of excess heme. Total HO activity was reduced after hypoxia as compared with cell lysates from controls in normal cell culture conditions, despite the increased HO-1 protein observed on Western analysis (Fig. 7A). Because a substantial portion of the HO-1 protein is localized to the nucleus after hypoxia, this implies that the nuclear HO-1 is less active than the form found in the sER. Preincubation with LMB prior to hypoxic exposure restored HO activity to nearly basal levels (Fig. 7A). Since hypoxia and LMB do not substantially increase HO-1 protein content (Fig. 4C), the lack of HO activity in hypoxia and the reversal effect by LMB are probably due to a shift in unclear localization.

To further understand whether C-terminal truncation results in altered HO activity, transient transfection with FL or with C{Delta}23 was performed and evaluated for HO activity. Although transfection with FL resulted in a 6-fold increase in the HO activity, transfection with C{Delta}23 resulted in only a 1.2-fold increase as compared with untransfected, Lipofectamine-treated, or vector-transfected controls (p = 0.35; Fig. 7B). Western analysis documented that there was no loss of HO-1 immunoreactive protein after the C{Delta}23 transfection (data not shown). Overall, these data demonstrate that nuclear translocation of HO-1 is associated with loss of HO activity despite abundance of HO-1 protein.

HO-1 Protein Delivery Mediates Activation of Various Transcription Factors—Nuclear localization of HO-1 protein suggests a signaling role. Because HO-1 is not a traditional transcription factor, we hypothesized that HO-1 protein could modulate oxidant-responsive transcription factor binding, thus modulating gene expression. A lipid-based protein delivery system (Pro-jectTM; Pierce) was used to deliver HO-1 protein intracytoplasmically. A GST-HO-1 fusion protein was expressed and purified with a GST column followed by thrombin digestion to cleave the GST tag. To determine the efficiency of the protein delivery system and to optimize transfection conditions, we first delivered exogenous HO-1 protein into HO-1-deficient MEF cells. Immunoreactive signal of HO-1 was detected within 2 h after transfection in both the whole cell and nuclear extracts, demonstrating the effectiveness of the protein delivery system (Fig. 8A). In this model, the HO-1 protein levels peaked at 2 h and returned to base line by 8 h, suggesting rapid degradation of the HO-1 protein.

To determine which transcription factors are regulated by HO-1 protein, a trans-signal array technique was employed to assess differential DNA binding activities of 54 transcription factors after intracytoplasmic delivery of HO-1 protein. Activation was only considered positive if there was a >3-fold increase compared with controls treated with Pro-JectTM alone (Table 3). To further determine whether transcription factor activation was due to the by-products of the HO reaction or the HO-1 protein itself, the assay was also conducted in cells where the inactive HO-1 mutant H25A protein was delivered. After HO-1 or H25A HO-1 protein delivery, transcription factors AP-1, AP-2, and Brn-3 showed 6-, 6-, and 9-fold increased activation, respectively, whereas core-binding factor (CBF) was activated 3-fold by the active HO-1 but was activated 9-fold by the mutant protein, suggesting an inhibitory effect of the HO by-products on the activation of this transcription factor. In the case of CCAT displacement protein (CDP), Ets/PEA3, STAT1, STAT3, and STAT4, these were activated only with delivery of the native enzyme but not with the mutant protein.


View this table:
[in this window]
[in a new window]

  		TABLE 3
Transcription factor activation after HO-1 protein delivery

 


Figure 5
View larger version (63K):
[in this window]
[in a new window]

  		FIGURE 5.
Detection and analysis of nuclear FLAGHO-1 after hypoxia. HEK293 cells were stably transfected with FLAGHO-1 cDNA. After exposure to 3% oxygen for 24 h, intact nuclei were isolated, and the protein was extracted. Purification and enrichment of the FLAGHO-1 were achieved with a FLAG affinity column. An aliquot of the eluent was denatured and loaded onto a 4-12% gradient acrylamide gel followed by electrophoresis. Colloidal blue staining of the gel revealed a predominant band between 30 and 45 kDa (arrow). The sample was used for determination of molecular mass and peptide sequencing as described under "Results." M, molecular weight marker.

 


Figure 6
View larger version (38K):
[in this window]
[in a new window]

  		FIGURE 6.
Cysteine protease inhibition prevents HO-1 cleavage and nuclear migration. Cultured 3T3 cells were incubated with E64d, a cysteine protease inhibitor that is cell-permeable, prior to exposure to hypoxia. A, immunocytostaining with HO-1 antibody (SPA 896). B, representative Western analysis of HO-1 protein in cell lysates after hypoxia and preincubation with E64d.

 


Figure 7
View larger version (13K):
[in this window]
[in a new window]

  		FIGURE 7.
Nuclear HO-1 protein has reduced enzymatic activity. Cultured 3T3 cells were exposed to hypoxia for 24 h. A, hypoxia alone reduced HO activity by 40%, whereas a preincubation with LMB recovered HO activity. Values represent the mean ± S.E. of three separate experiments. *, p < 0.05 versus all controls. B, in other experiments, cultured 3T3 cells were transfected with EGFPHO-1C{Delta}23 (C{Delta}23) or EGFPFL HO-1 (FL) cDNA and then evaluated for HO activity as described under "Materials and Methods." Values represent the mean ± S.E. of three separate experiments. *, p < 0.05 versus all controls. Controls were untransfected cells (C), cells treated with Lipofectamine (L), and cells transfected with the EGFP vector alone (V).

 


Figure 8
View larger version (41K):
[in this window]
[in a new window]

  		FIGURE 8.
Delivery of HO-1 protein activates an oxidant-responsive promoter. A, verification of HO-1 proteins delivered into HO-1-deficient MEF cells. Representative Western blot demonstrating increased HO-1 immunoreactive protein 0-24 h after HO-1 protein delivery. No HO-1 immunoreactive signal is seen in untransfected (UT) cells. Note the rapid disappearance of HO-1 protein after 4 h. The inset demonstrates delivery of the HO-1 protein in whole cell as well as nuclear extracts. Bottom, densitometric evaluation of HO-1 immunoreactive protein after HO-1 protein delivery in HO-1 -/-MEF cells. Values are the means ± S.E. of four independent determinations expressed as a percentage of the 2 h values to allow for normalization. *, p < 0.05 versus 2h. Right, detection of HO-1 protein in cytosolic and nuclear extracts before and after HO-1 delivery. Calnexin was used as marker for cytosolic protein. Note the presence of HO-1 protein in both the cytosolic and nuclear fractions after transfection. B, a pseudoimage of the photon emission from 3T3-HO-1/luc cells after delivery of HO-1 protein over 0-24 h. Negative controls are cells that did not receive any protein (untransfected) or cells where other proteins, such as galactosidase (Gal), a fluorescein isothiocyanate-labeled immunoglobulin (Ab), or Pro-JectTM alone (P) were delivered. Quantitative luminescence intensity is shown below. Values are the means ± S.E. of four independent measurements. *, p < 0.05 versus untransfected cells.

 
In contrast to the above, DNA binding activity of transcription factor NF-{kappa}B was down-regulated to 65% after delivery of the mutant HO-1 protein but was unchanged after delivery of the native HO-1. Activity of transcription factor SP-1 was down-regulated by delivery of both wild type and mutant HO-1 proteins.

Transcriptional factor activation was further verified using EMSA to demonstrate binding of nuclear proteins to the consensus DNA sequence. Increased binding of nuclear proteins to the AP-1 consensus sequence after delivery of native or mutant HO-1 protein was observed. To further substantiate AP-1 activation, a 5-fold increase in nuclear phosphorylated c-Jun was observed after HO-1 protein delivery (Fig. 9, inset). Binding to Brn-3, CDP, and CBF and STAT3 binding sites was also enhanced (Fig. 9A). Neither STAT1 nor Ets/PEA3 showed altered DNA binding by EMSA, suggesting nonspecific activation on the transcription factor array (data not shown). Binding of nuclear proteins to NF-{kappa}B and SP-1 sites was visibly diminished after delivery of either the native or the inactive HO-1 protein in contrast to the transcription factor array data (Fig. 9B). Because Nrf2 is an important mediator of oxidative stress signaling, binding to this consensus sequence was also investigated. No differences in Nrf2 binding were observed after delivery of native or mutant HO-1 protein (data not shown).


Figure 9
View larger version (36K):
[in this window]
[in a new window]

  		FIGURE 9.
HO-1 protein delivery modulates DNA binding activities of transcription factors. A, representative EMSA using oligonucleotide consensus sequences from transcription factors activated by HO-1 in a transcription factor array (AP-2, CBF, and Brn-3, respectively). F, free probe; U, untreated cells; P, cells receiving Pro-JectTM vehicle alone; H, cells receiving HO-1 protein; mH, cells receiving the inactive mutant (H25A) HO-1 protein. C, cold competition with unlabeled probe. M, cold competition with a mutant probe (see Table 2 for consensus sequences). B, binding of AP-1 or NF-{kappa}B sites to nuclear proteins after HO-1 delivery. Binding of NIH 3T3 nuclear proteins to an AP-1 (A) or NF-{kappa}B consensus sequence after HO-1 protein delivery is shown. Inset, immunoreactive phosphorylated c-Jun levels after HO-1 protein delivery.

 
HO-1 Overexpression Activates an Oxidant-responsive HO-1 Promoter—To determine the role of HO-1 protein in the regulation of an oxidative stress-responsive promoter, a lipid-based protein delivery system (Pro-JectTM; Pierce) was used to incorporate HO-1 protein intracytoplasmically into NIH3T3 cells stably transfected with a luciferase reporter driven by a 15-kb HO-1 promoter (3T3-HO-1/luc). This reporter contains multiple oxidant-responsive transcription factors, as previously described (39). Luciferase activity in these cells directly correlates with promoter activation and can be measured by the photon emission with IVIS. Next, we delivered the pure HO-1 protein into the 3T3-HO-1/luc cells. Delivery of HO-1 protein was associated with increased HO-1 promoter activation in these cells. Compared with untreated cells, the maximum HO-1 promoter activation (1.7-fold) was observed 12 h after the HO-1 protein delivery (Fig. 8B). Delivery of other proteins, such as beta-galactosidase or fluorescein isothiocyanate-labeled IgG, did not increase HO-1 promoter activation; nor did delivery of the vehicle (Pro-JectTM) alone (Fig. 8B).

Using another model of HO-1 overexpression, similar results were obtained. Transfection of the HO-1/luc cells with the full-length His-tagged HO-1 cDNA and the mutant HO-1 cDNA construct in which histidine 25 was changed to alanine (HisHO-1mutFLAG), yielding an enzymatically inactive HO-1, was verified by Western blot analysis using anti-HO-1 and anti-His antibodies (Fig. 10A). Cells transfected with HisHO-1FLAG and HisHO-1mutFLAG had similar levels of HO-1 and His signals, suggesting similar transfection and expression efficiency (Fig. 10A). The 32 kDa band detected using an anti-HO-1 antibody represents the endogenous HO-1 protein, whereas the 40 kDa band in the same blot represents the transfected HisHO-1FLAG fusion proteins. Anti-His antibody could only detect the 40-kDa HisHO-1FLAG fusion proteins (Fig. 10A).

Twenty-four h after transient transfection into the HO-1/luc cells, luciferin was added into the culture medium, and luciferase activity was measured with IVIS. As shown in Fig. 10B, transfection with HO-1 cDNA was associated with a 3-5-fold increase in HO-1 promoter activation compared with cells transfected with transfection reagent or p3XFLAG vector, as demonstrated by increased photon emission (Fig. 10B). We also observed that the HO-1{Delta}23 C-terminally truncated HO-1 increases the promoter activity to 1.8-fold compared with vector control; however, the activity is only 52% compared with the full-length HO-1, which activated the promoter to 3.58-fold compared with vector control. Deletions from either the N or the C terminus (data not shown) indicate that the intact HO-1 protein is required for full HO-1 promoter activation; however, the C-terminal truncated isoform can also facilitate the promoter activation.

The HO-1 promoter has several antioxidant response elements within the distal enhancer regions, which have been identified. Preliminary evaluations revealed synergistic activation of both the DE1 and DE2 enhancer regions by HO-1 transfection (data not shown). This demonstrates that HO-1 protein modulates oxidant-responsive signaling.

Nuclear Translocation of HO-1 Enhances Resistance to H2O2 Toxicity but Does Not Protect against Hemin-mediated Cell Death—In other models, nuclear translocation of a cytoplasmic protein alters its function (13). In our case, the nuclear form of HO-1 was inactive and resulted in alterations in transcription factors involved in oxidative stress. This suggested that nuclear HO-1 may enhance the protection against oxidative stress. To verify this, cells transfected with EGFPHO-1FL and EGFPHO-1C{Delta}23 were incubated with 200 µM H2O2 for 4 h. Transfection with both the FL and the C{Delta}23 constructs was associated with improved cell viability as compared with vector-transfected controls (Fig. 11A). None of the other HO-1 fragments (i.e. 201, 202-, 456, and 465-), regardless of their nuclear localization pattern, resulted in a significant cytoprotection against H2O2 toxicity (data not shown). In contrast, only cells transfected with the FL construct that is enzymatically active were protected against hemin (100 µM) toxicity, presumably by allowing for the degradation of the toxic pro-oxidant and the production of the protective heme catabolites (Fig. 11B).


Figure 10
View larger version (44K):
[in this window]
[in a new window]

  		FIGURE 10.
HO-1 cDNA activates an oxidant-responsive promoter. NIH3T3 cells were stably transfected with a luciferase reporter controlled by the 15-kb HO-1 promoter (3T3-HO-1/luc). Different amounts of cDNAs encoding His-tagged HO-1 proteins were transiently transfected into 3T3-HO-1/luc cells. A, Western blot analysis of HO-1 and His expression. 20 µg of protein from whole cell lysates were separated by 4-12% SDS gel and probed with anti HO-1 and anti-His antibodies. The 40 kDa band represents the HO-1-His fusion protein, and the 32 kDa band represents the endogenous HO-1 protein. H, wild type HO-1 cDNA; M, mutant HO-1 cDNA with a His25 to Ala substitution. B, 24 h after transfection, luciferase expressions were analyzed with IVIS for photon emissions. C, control 3T3-HO-1/luc cells without cDNA transfection. V, cells transfected with the pCMV expression vector. Quantitative data are presented as mean ± S.E. from three independent experiments with duplicate samples. *, p < 0.05 versus control.

 


Figure 11
View larger version (12K):
[in this window]
[in a new window]

  		FIGURE 11.
Effect of nuclear localization of HO-1 on cell viability. Cultured 3T3 cells were transfected with EGFPHO-1C{Delta}23 (C{Delta}23) or EGFPFL (FL) HO-1 cDNA and incubated with either H2O2 (A) or hemin (B), and cell viability was evaluated. Values represent the mean ± S.E. of three separate experiments.*,p<0.05versus cells transfected with EGFP (V) vector alone.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
In the initial purification of HO-1, a 32-kDa microsomal protein was found in the ER (35). However, HO-1 has been documented in various subcellular fractions using immunohistochemical techniques (40, 41). Here, we document nuclear localization of HO-1 using three separate conditions known to induce HO-1 and demonstrate that the truncated HO-1 protein in the nucleus has reduced enzymatic activity.

After hypoxia or incubation with hemin and H/HPX, HO-1 is clearly observed in the nucleus and co-localizes with known nuclear proteins. Because HO-1 is an integral membrane protein of the sER, nuclear translocation would require cleavage of the membrane-bound domain. In an earlier publication (16), a 28-kDa protein lacking the C-terminal 23 amino acids of HO-1 was recovered after trypsin incubation. This HO-1 was no longer membrane-bound but water-soluble. In partial agreement with this finding, our mass spectroscopy results identified a nuclear form of HO-1 that was 27 kDa in size and was lacking the C terminus. This could allow for translocation to the nucleus. Although nuclear import of small proteins less than 50 kDa can occur by diffusion (18), nuclear localization is facilitated by the presence of an NLS (17). No sequence matching an NLS was detected in the HO-1 protein.

Nuclear-cytoplasmic shuttling of various proteins is facilitated by the presence of other sequences. On the rat HO-1, amino acids 207-221 have greater than 90% homology to a published NES (19, 42) (referred to as NSS). In other examples, the NES sequence facilitates nuclear export. In fact, when the putative NSS sequence was fused to a GFP, it localized to the cytoplasm (Fig. 3B). Surprisingly, modification of the NSS sequence on HO-1 was required for nuclear import of HO-1 protein, because its absence impeded nuclear localization despite truncation of the C terminus (Fig. 3A). Although this is the opposite of what we had expected, others show that in some cases NES motifs can function as NLS within a protein as with UL84 of the human cytomegalovirus (43). Additionally, oxidative modification can modify the function of an NES. In the case of Nrf2, when this protein containing an NLS forms a complex with Keap1, a protein with a dominant NES, it is retained in the cytoplasm. After oxidative modification of the NES, Nrf2 dissociates from Keap1 and migrates to the nucleus (44). Perhaps HO-1 forms a complex with an NLS-containing protein to facilitate its nuclear translocation. Although it is known that HO-1 can interact with other proteins (45), it remains to be determined whether such an interaction could enhance nuclear shuttling.

Nuclear migration of HO-1 was limited by LMB, an antimicrobial that abolishes CRM1-mediated nuclear-cytoplasmic transport. In most instances, CRM1 serves as an exportin, which binds with RanGTP to allow for passage through the nuclear pore and export to the cytoplasm (19). Here we clearly document binding of HO-1 to CRM1; therefore, we suspect that in this case, CRM1-mediated binding to HO-1 facilitates nuclear import rather than export. In one report, nuclear migration of the beta-subunit of the nuclear pore targeting complex did not require RanGTP and only required the nuclear pore complex-binding domain, which is also involved in NLS-mediated import (46). This suggests that CRM1 could be involved in bidirectional shuttling across the nuclear pore.

Hypoxia and/or hemin-enhanced HO-1 nuclear entry could be secondary to toxic effects leading to enhanced nuclear permeability or due to oxidative modification of the nuclear shuttling sequence (44). Although preliminary evidence suggests that the HO-1 protein can be oxidatively modified4 and that oxidative modification is known to modify NES-CRM1 binding (44), we have not yet elucidated a specific site of oxidation of the HO-1 protein. Furthermore, the effects of hemin on HO-1 nuclear localization are not strictly due to oxidative degradation, because they cannot be prevented by antioxidants, such as catalase and superoxide dismutase.

Previous reports document cleavage of purified HO-1 protein with trypsin in vitro (16). Here we show that inhibition of a cysteine protease obviates the nuclear translocation of HO-1 and that the precise site of cleavage of HO-1 observed after hypoxia involves the C-terminal 52 amino acids. This suggests that hypoxia-mediated cellular stress activates proteolytic cleavage of the HO-1 protein. Nonetheless, PESTfind analysis detected a C-terminal PEST domain on the mouse HO-1 protein at amino acids 239-254 (RPASLVQDTTSAETPR), as was previously suggested (47), where the PEST score can be significantly lowered from +3.8 to -20.8 with alanine substitution of the serines and threonines, akin to the PEST domain of I{kappa}B{alpha} (48). Importantly, calpain-mediated cleavage at this domain would also yield a 27-kDa protein as was detected in our experiments.

When HO-1 localizes to the nucleus after hypoxia, there is an associated loss of total HO activity (the sum of HO-1- and HO-2-derived activities). Because cellular HO-2 protein content is low in the cell lines tested and HO-2 is not induced or inhibited by hypoxia or hemin, this implies that HO-1 becomes less enzymatically active when it is C-terminally truncated and imported to the nucleus. In previous experiments, a 28-kDa HO-1 protein fragment in the presence of detergent-solubilized NADPH cytochrome c P450 reductase failed to transfer the second electron from NADPH needed for degradation of the heme (16, 35). However, more recent experiments document that truncation from the C terminus of human HO-1 by 23 amino acids did not reduce HO specific activity, but further truncation by 56-68 amino acids did reduce HO activity by 50% (29). In our experiments, transfection with a construct expressing an HO-1 mutant predominantly expressed in the nucleus with truncation of the 23 terminal amino acids reduced HO activity 5-fold. This form of HO-1 was predominantly expressed in the nucleus. Furthermore, the exact site of cleavage of nuclear HO-1 was identified using gas chromatography/mass spectrometry and MALDI-TOF. This demonstrated loss of the terminal 53 amino acids of the HO-1 protein, yielding a 27-kDa fragment. This degree of truncation would predictably reduce HO-1 activity.

It is possible that the loss of HO activity could result from additional post-translational modification of the HO-1 protein, although this has not been defined yet. Nonetheless, using modeling, it is clear that C-terminal truncation does not alter the heme catalytic pocket and therefore should not prevent substrate binding and limit HO activity (see Fig. 3A).

To explain the role of nuclear HO-1, we also document the activation of key transcription factors involved in oxidative stress and cellular proliferation after HO-1 delivery. Transcription factor regulation by HO-1 could be mediated by direct protein-DNA interaction akin to the action of other transcription factors. However, the structure of HO-1 does not reveal traditional DNA binding motifs, a characteristic of most transcription factors. Direct HO-1-mediated transcriptional activation is less likely without a DNA binding motif, as seen with transcription factors (reviewed in Ref. 49). Another potential mechanism could involve binding of HO-1 to a transcription factor or a protein complex. Although HO-1 can bind to its isoenzyme HO-2 (45), no reports demonstrate binding of HO-1 to known transcription factors. However, HO-1 could bind to other proteins, serving as the non-DNA-binding protein of a transcriptional factor complex as with the non-DNA-binding CBFbeta subunit of the core binding factor, which increases the affinity of the DNA-binding domain of CBF{alpha} for DNA while making no direct contacts to the DNA (50). This remains to be determined.

With transcription factor signal arrays and EMSAs, increased activation of the AP-1 site was observed after transfection with active or inactive HO-1 protein. However, we did not observe increased Nrf2 binding activity on electrophoretic mobility assay. This is an important mechanism by which oxidative regulation occurs in many models (51, 52). In the case of HO-1 gene regulation, the AP-1 subunit protein c-Jun does not appear to interact with Nrf2 but rather acts as a transactivator of the HO-1 promoter (27). We observed increased phosphorylation of c-Jun, a subunit of AP-1, suggesting that HO-1 protein delivery can lead to phosphorylation of the AP-1 complex proteins with resultant nuclear translocation and transcriptional activation. Whether phosphorylation of HO-1 is required for this signaling is not known. Whether HO-1 or other proteins complex with c-jun to modulate AP-1 binding is also not yet defined. Because HO-1 regulates both heme and iron levels in the cell, alterations in heme and/or iron and subsequent changes in cellular redox could have caused the change in AP-1 activation, as demonstrated by others (53). In a previous report, transfection of mutant HO-1 protein into HO-1-deficient cells resulted in altered catalase gene expression (11). This effect was attributed to accumulation of heme. In our model, cells with endogenous HO-1 protein were transfected with the mutant HO-1. This should not alter heme levels, because the native HO-1 should maintain the heme equilibrium. Furthermore, delivery of both the native and mutant HO-1 protein had the same effect on AP-1 activation, and the delivered HO-1 protein was relatively short lived compared with the native form. Therefore, any effect on heme or iron levels would probably revert to normal within a few hours. In addition, if the changes in transcription factor activation were due to changes in heme and/or iron, we would have anticipated an increase in AP-1 and NF-{kappa}B activation rather than an increase in AP-1 and a decrease in NF-{kappa}B activation as shown, because heme and iron positively regulate NF-{kappa}B activation as they do AP-1 (53).

Both enzymatically active and inactive HO-1 protein activated the transcription factors Brn-3, CBF, and AP-2. The class IV POU domain transcription factor Brn-3 family members alter differentiation, migration, or survival of specific central neuronal populations (54). The CBFs are required at various stages of hematopoiesis, and the CBF transcription factor complex is essential for cell proliferation and viability and for initiating hematopoiesis in the embryo (55). The transcription factor AP-2 plays an important role in embryonic development by influencing cell differentiation, proliferation, and survival. It is also implicated in carcinogenesis and abnormal cell proliferation (56). Other transcription factors, such as CDP and STAT3, were highly expressed with delivery of the active but not with the mutant inactive HO-1 protein. The differential activation of HO-1 protein delivery on CDP and STAT3, depending on HO activity, suggests that one of the HO by-products might modulate that expression of these proteins. In fact, CO can modulate STAT3 activation, and this is responsible for the antiapoptotic effect of CO in endothelial cells (57). No evidence exists regarding the effect of HO by-products on CDP activation so far.

Overall, increased activation of transcription factors involved in cell proliferation, such as Brn-3, CBF, and AP-2, was observed after intracytoplasmic delivery would suggest that increased activation of these factors by HO-1 protein or its by-products would probably increase cellular proliferation. Several reports suggest an antiproliferative effect of the HO by-product CO (58-60). The role of HO-1 protein devoid of activity on cell proliferation needs to be systematically addressed. Nuclear localization of suggests that HO-1 protein may serve in a signaling capacity, perhaps to regulate cellular proliferation and up-regulate antioxidant defenses.

The regulation of HO-1 gene expression by oxidative stress has been well characterized. The proximal enhancer region as well as two 5' distal enhancer regions, DE1 and DE2, mediates transcriptional activation of the HO-1 gene in response to various stimuli. Each enhancer region contains multiple copies of the cis-acting stress-responsive element, similar to the Maf response element, and binding sites of the AP-1 class of transcription factors. Key regulators through DE1 and DE2 include the Nrf2 transcription factor, which can heterodimerize to Maf and activate HO-1 gene transcription, and Bach-1, a heme-regulated and hypoxia-inducible basic leucine zipper protein, to repress HO-1 gene expression by binding to the multiple Maf recognition elements on the HO-1 gene (61). Therefore, the HO-1 promoter is a useful model to understand the effect of HO-1 on oxidative response elements. We demonstrate activation of the HO-1 promoter by the HO-1 protein, whether it is active or not. We also show that full activation requires the C terminus but that C-terminal truncation still mediates this activation. This reinforces the notion that the HO-1 protein modulates oxidant-mediated cellular responses, probably through modulation of oxidant-responsive transcription factors.

Last, despite reduced activity, the truncated nuclear form of HO-1 protein retained its cytoprotective role against H2O2-mediated injury, further suggesting that the HO-1 protein itself is important to cellular cytoprotection. A previous report showed that transfection with an inactive mutant HO-1 increased catalase gene expression and glutathione content (11). We also demonstrate here that transfection with HO-1 cDNA or delivery of HO-1 protein results in activation of various transcription factors, whether the enzyme is active or not. This suggests that the HO-1 protein itself may play a role in cellular signaling. In contrast, no protection was provided against hemin-mediated injury. This is logical, because HO activity is vital to the degradation of the pro-oxidant heme to antioxidant bile pigments. It will be important to understand what downstream effects result from HO-1 nuclear entry and how these precisely mediate cellular cytoprotection.

In summary, we have demonstrated that hypoxia and two other inducers of HO-1, heme and heme-hemopexin, result in nuclear localization of HO-1 and that this process results from protease-mediated, C-terminal truncation of HO-1 protein linked to CRM1-mediated nuclear-cytoplasmic shuttling. Nuclear localization also reduces HO activity. Additionally, HO-1 protein altered binding of transcription factors involved in oxidative stress and allowed for protection against H2O2.We speculate that nuclear localization of HO-1 is an important signaling pathway to provide cytoprotection in the presence of oxidative stress.


   FOOTNOTES
 
* This work was funded by National Institutes of Health Grant HL-58752 (to P. A. D.), the Research Incentive Funds of the University of Missouri-Kansas City (to A. S.), and the Hess and Mary L. Johnson funds from Stanford University (to P. A. D.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

1 These three authors contributed equally to this work. Back

2 To whom correspondence should be addressed: Division of Neonatology, Children's Hospital of Philadelphia 34th and Civic Center Blvd., Philadelphia, PA, 19104. Tel.: 215-590-1653; Fax: 267-426-5632; E-mail: dennery{at}email.chop.edu.

3 The abbreviations used are: HO, heme oxygenase; sER, smooth endoplasmic reticulum; TM, transmembrane; NLS, nuclear localization sequence; NES, nuclear export sequence; LMB, Leptomycin-B; DE1 and -2, distal enhancer 1 and 2, respectively; NSS, nuclear shuttling sequence; DMEM, Dulbecco's modified Eagle's medium; FBS, fetal bovine serum; MEF, mouse embryo fibroblast; PBS, phosphate-buffered saline; GFP, green fluorescent protein; EGFP, enhanced green fluorescent protein; MALDI, matrix-assisted laser desorption ionization; TOF, time-of-flight; MS, mass spectrometry; EMSA, electrophoretic mobility shift assay; HIV, human immunodeficiency virus; AP, activator protein; CBF, core-binding factor; CDP, CCAT displacement protein; H/HPX, heme-hemopexin; 3T3, NIH3T3. Back

4 G. Yang, A. Hu, J. Bordner, and P. A. Dennery, unpublished observations. Back


   ACKNOWLEDGMENTS
 
We thank Dr. Bryan R. Cullen (Department of Molecular Genetics and Microbiology, Duke University) for providing the pGEX 4T1 CRM1 vector. We thank Jessica Bordner for invaluable technical assistance. We are grateful to Peggy McDonald, Jennifer McIntyre, and Cheryn Javis for administrative assistance.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Alam, J., Shibahara, S., and Smith, A. (1989) J.