| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Vascular Biology Research Center and Division of Hematology, Department of Internal Medicine, University of Texas-Houston Medical School, Houston, Texas 77030, USA
1Correspondence: Vascular Biology Research Center and Division of Hematology, Department of Internal Medicine, University of Texas-Houston Medical School, 6431 Fannin MSB 5.016, Houston, TX 77030, USA. E-mail: Kenneth.K.Wu{at}uth.tmc.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
78 kDa salicylate binding protein with sequence identity to immunoglobulin heavy chain binding protein (BiP). The Kd values of SA binding to crude extract and to recombinant BiP were 45.2 and 54.6 µM, respectively. BiP is a chaperone protein containing a polypeptide binding site recognizing specific heptapeptide sequence and an ATP binding site. A heptapeptide with the specific sequence displaced SA binding in a concentration-dependent manner whereas a control heptapeptide did not. Salicylates inhibited ATPase activity stimulated by this specific heptapeptide but did not block ATP binding or induce BiP expression. These results indicate that salicylates bind specifically to the polypeptide binding site of BiP in human cells that may interfere with folding and transport of proteins important in inflammation.—Deng, W.-G., Ruan, K.-H., Du, M., Saunders, M. A., Wu, K. K. Aspirin and salicylate bind to immunoglobulin heavy chain binding protein (BiP) and inhibit its ATPase activity in human fibroblasts.
Key Words: salicylate binding protein • SA • COX-2 • acetylsalicylic acid
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and serves as a messenger molecule to induce the expression of plant defense-related genes (2)
. SA was identified as the active component of willow bark extracts that were shown to have anti-inflammatory properties more than a century ago. Acetylsalicylic acid (ASA or aspirin) was subsequently synthesized and remains one of the most commonly used anti-inflammatory drugs (3)
. Recent studies have shown that SA and aspirin also have anti-neoplastic properties (4)
. The mechanisms by which SA and aspirin exert these pharmacological actions in human are unclear. Several studies suggest that SA suppresses leukocyte activation and adhesion to endothelial cells (5
, 6)
and enhances cancer cell apoptosis (7
, 8)
. These cellular activities have been attributed to inhibition by SA of I
B-kinase activity and IB degradation with the consequent inhibition of nuclear factor B (NF-B) activation (9
, 10)
. Other signaling mechanisms have also been proposed (11
12
13)
. The concentrations of SA (>5 mM) able to inhibit NF-B activation are suprapharmacological, as previously reported (14)
; at such high concentrations, SA inhibits many mammalian cell kinases in a nonspecific manner. It is thus unlikely that the pharmacological actions of SA and aspirin can be explained by suppression of NF-B activation. We have reported that SA and aspirin at pharmacological concentrations (10-5 to 10-3 M) suppress cyclooxygenase-2 (COX-2) transcription induced by proinflammatory cytokines and mitogenic factors (15)
. A recent report has also shown that salicylate at a pharmacological concentration selectively inhibits binding of CCAAT/enhancer binding protein ß (C/EBPß) to COX-2 promoter (16)
. Since COX-2 plays an important role in inflammation (17
, 18)
and tumorigenesis (19)
, this action of aspirin and SA provides a more plausible explanation for their anti-inflammatory and anti-neoplastic effects.
Work from Klessig’s group suggests that the action of SA on plant cells is mediated by SA binding to plant catalase, resulting in an elevation of H2O2, which is a proposed signaling molecule in plant gene expression (20
, 21)
. This proposal was questioned by a report showing that SA bound not only to catalase, but also to other iron-containing enzymes in plant cells (22)
. It was also reported that SA bound to bovine catalase at relatively high concentrations (3 mM and higher). Thus, it remains unclear whether SA at pharmacological concentrations binds to a human cellular protein. Here we report the purification of a protein from crude extract of human fibroblasts that exhibited specific binding to SA with a Kd of 45.2 µM. Amino acid sequence of the NH2-terminal region of this purified SA binding protein was identical to that of human immunoglobulin heavy chain binding protein (BiP). The purified 78 kDa protein was recognized by a human anti-BiP antibody. BiP is a chaperone protein containing a polypeptide site recognizing specific heptapeptide sequences and an ATP binding site. Aspirin and salicylate bound specifically to the polypeptide binding site and inhibited the BiP ATPase activity.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
[14C] SA binding assay
SA binding assays were carried out as described (23)
with some modifications. In brief, 0.5 ml of binding mixture containing 10 mM citrate pH 6.5, 10 mM MgSO4, 1 mM EDTA, 2 mg protein, and increasing concentrations of [14C]SA (12.9 Ci/mol) was incubated at 4°C for 2 h in the presence or absence of a 1000-fold molar excess of unlabeled salicylate. Bound and free [14C]SA were separated by a spin-column exclusion chromatography with a P-6DG desalting gel column (Bio-Rad, Hercules, CA). The radioactivity was measured with a ß-scintillation counter. Specific binding isotherms were constructed by subtracting nonspecific binding from total binding. Kd and Bmax were calculated by a double reciprocal plot.
Extraction and purification of SA binding protein(s)
SA binding proteins were purified from HFF lysates by binding activity-guided stepwise chromatography similar to the procedure reported (20)
. The cell lysate was centrifuged and the supernatant was brought to 50% of ammonium sulfate saturation. The precipitate was collected after centrifugation and dissolved in a binding buffer (10 mM citrate pH 6.5, 10 mM MgSO4 and 1 mM EDTA). This crude extract was then dialyzed against the same binding buffer overnight and centrifuged to remove undissolved fractions before SA binding assay. The extract was desalted on a Sephadex G-25 column (Pharmacia, Piscataway, NJ) equilibrated with binding buffer, followed by DEAE-Sephacel chromatography on a 2.5 x 15 cm column (Pharmacia) using a KCl gradient from 0 to 1.0 mM at a flow rate of 0.4 ml/min. SA binding activity of the eluted fractions was analyzed and the fractions with SA binding activity were pooled and chromatographed on a blue dextran-agarose column (0.5x15 cm) (Sigma, St. Louis, MO) at a flow rate of 0.2 ml/min with binding buffer containing 0.5 M KCl. The active fractions were combined, concentrated, and injected (0.1/ml per run) into a Superose 6 HR 10/30 column (Pharmacia) connected to an HPLC system (Waters, Millford, MA). Fractions were eluted with binding buffer at a flow rate of 0.5 ml/min. The active fractions were pooled and concentrated. Protein concentration was measured using a protein assay kit (Bio-Rad).
Protein separation and microsequencing
Proteins in various chromatographic preparations were resolved in 12% polyacrylamide gels by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Protein bands were detected by Coomassie blue staining. Molecular weight markers were included to estimate the molecular weight of SA binding proteins. Active fractions obtained from the final Superose 6 HR chromatography step contained two discrete bands. Each band was electroblotted to a PVDF membrane, excised from the gel, and sequenced with Applied Biosystems 473A Sequencer at Baylor College of Medicine.
Preparation of recombinant BiP protein
A full coding sequence of human cDNA of BiP was inserted into a glutathione S-transferase (GST) expression vector pGEX-4T-1 (Pharmacia) and transformed into Escherichia coli strain BL21. Expression of recombinant GST-BiP fusion proteins was induced by isopropyl ß-D-thiogalactopyranoside (1 mM). BiP proteins were purified with a glutathione Sepharose 4B column (Pharmacia) after digesting GST-BiP with thrombin. This recombinant human BiP has been shown to have identical biochemical properties as purified bovine BiP (24)
.
Western blotting analysis
Purified recombinant BiP and purified SA binding protein preparations were separated by SDS-PAGE, electrophoretically transferred to nitrocellulose membranes, and probed with a rabbit polyclonal antibody against human BiP (StressGen, Victoria, B.C., Canada) using an ECL detection kit (Pharmacia).
ATPase measurement
The ATPase was assayed by measuring the free Pi liberated during ATP hydrolysis according to a method previously described (24)
. The standard assay was carried out at 37°C in 20 µl of reaction mixture containing 1 µg BiP, 0.1 µM [
-32P]ATP for 30 min, then 20 µl of 4% SDS was added to stop the reaction. The reaction mixture were extracted with 50 µl fresh phosphate reagent and 200 µl 65:35 xylene:isobutyl alcohol. A 50 µl aliquot was removed from the upper organic phase and the released Pi radioactivity was counted in a liquid scintillation counter.
ATP binding assay
Binding of ATP was determined by using a nitrocellulose filter assay as described previously (25)
. The mixtures for incubation (20 µl total) contained 8 µg of BiP, 20 mM Bis-Tris, pH 7.0, 2 mM MgCl2, 30 mM KCl, 3 µM [3H]ATP, and various concentrations of SA. Trace levels of ADP in mixtures were removed by adding 0.3 µg of pyruvate kinase and 30 µM phosphoenolpyruvate. The reaction was stopped by addition of 3 ml of 10 mM Bis-Tris, pH 7.0, containing 2 mM MgCl2, followed by rapid filtration through a nitrocellulose filter (0.45 µm). The filter was washed three times with 5 ml wash buffer. Radioactivity on the filter was determined with a liquid scintillation spectrometer.
Synthesis of heptapeptides
A heptapeptide with a sequence of KYWWNLL corresponding to a sequence in gp160 of human immune deficiency virus (26)
reported to have a high affinity for human BiP and a control heptapeptide with a sequence of TWFNSTW without binding activity for BiP were synthesized (Sigma-Genosys, Woodlands, TX) and used in evaluating competition with [14C]SA binding to BiP.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A). SA binding was reversed by addition of 1 mM SA at 60 min of binding experiment (Fig. 1B
). The SA binding activity was disrupted by pretreatment of the cell extract with trypsin (20 µg/ml) or 0.5% SDS (data not shown), consistent with binding to a protein in the cell extracts. To purify the SA binding protein, the cell extract was subject to Sephadex G-25 chromatography for desalting, followed by SA binding activity-guided stepwise DEAE-Sephacel chromatography (Fig. 2
A), blue dextran-agarose chromatography (Fig. 2B
), and Superose 6 HR chromatography (Fig. 2C
). Active fractions from each chromatography step were pooled, concentrated, and the proteins in these fractions were separated by SDS-PAGE. After the final chromatography step, a major band at 78 kDa and a minor band at 56 kDa were detected (Fig. 3
A). These two bands were excised and subject to microsequencing. A 15 amino acid sequence of the 78 kDa band was obtained, and sequence comparison shows that it is identical to the first 15 amino acid sequence of mature human BiP (Fig. 3B
). We were unable to obtain the sequence of the minor band because of NH2-terminal block. Table 1
summarized the purification of this 78 kDa SA binding protein. We obtained a 103-fold purification after four steps of chromatography. In view of a 100% sequence identity of the NH2-terminal 15 amino acid residues of our purified protein to that of human BiP, this purified protein is probably BiP. To support this, we performed Western blot analysis of the purified protein preparation using an anti-human BiP antibody. The anti-BiP antibody recognized a single 78 kDa band of our purified protein preparation, which migrated to a position on the gel identical to that of authentic human BiP expressed in E. coli (Fig. 3C
). Thus, our purified protein is biochemically and antigenically identical to human BiP. These results led us to conclude that BiP is a major SA binding protein in human fibroblasts. We therefore carried out subsequent binding experiments with purified human recombinant BiP.
|
|
|
|
The kinetics of SA binding to purified human BiP was determined by using [14C]SA at concentrations from 0–200 µM in the presence and absence of a 1000-fold molar excess of unlabeled SA. SA bound to BiP in a specific, saturable manner and double reciprocal analysis revealed a single binding site with a Kd of 54.6 µM and a Bmax of 263 pmol/mg BiP (Fig. 4
A). The binding parameters initially obtained using crude cell extracts (Fig. 1A
) matched remarkably well with those of purified BiP, confirming that BiP is the SA binding protein in human cells. We next determined the relative potency of aspirin and several SA metabolites or analogs in competing with SA binding by adding a slightly more than 100-fold molar excess of unlabeled aspirin or analogs (5 mM) to [14C] SA (40 µM), which was incubated with purified BiP, and the specific SA binding activity was determined. At this molar excess, unlabeled SA inhibited [14C]SA binding by 90% using vehicle control as the reference. Aspirin has a comparable inhibition as SA. Of several dihydroxybenzoic acid (DHBA) compounds tested, 2,6-DHBA was equivalent to aspirin in inhibiting SA binding whereas 2,5-DHBA, 2,3-DHBA, and 3,5-DHBA showed decreasing inhibiting potency (Fig. 4B
). 3,4-Dimethylbenzoic acid (3,4-DMBA) inhibited SA binding by only 20% whereas 3-hydroxybenzoic acid (3-HBA) did not significantly inhibit SA binding.
|
The BiP molecule comprises an ATPase domain at the NH2-terminal region that binds ATP and hydrolyzes it to ADP and a peptide binding domain at the carboxyl-terminal region that binds polypeptide with specific heptapeptide sequences (27
, 28)
. SA did not inhibit ATP binding to recombinant BiP (Fig. 5
A). To determine whether SA bound to the peptide binding site, we synthesized a heptapeptide (KYWWNLL) based on a sequence in HIV gp160 that was reported to bind BiP with a Kapp of 60 µM and a control sequence (TWFNSTW) also based on the HIV gp160 sequence, which was reported to bind poorly to BiP (Kapp>500 µM) (26)
. The active peptide competed with [14C]SA binding in a concentration-dependent manner whereas the control peptide had no effect on SA binding (Fig. 5B
). These results suggest that SA binds to the peptide binding site of BiP. Peptide binding to BiP stimulates the BiP ATPase activity that converts ATP to ADP, thereby closing the peptide binding pocket (28)
. We determine whether SA inhibits ATPase activity stimulated by the active peptide. Both SA and aspirin inhibited the basal ATPase activity in a concentration-dependent manner whereas the inactive benzoate 3,4-DMBA did not (Fig. 6
A). SA at 100 µM also significantly inhibited ATPase activity stimulated by 1 mM active peptide whereas 3,4-DMBA had no effect on ATPase activity (Fig. 6B
). These results further support the binding of SA to the peptide binding site of BiP.
|
|
It is suggested that reduced availability of free BiP as a result of BiP complexing with misfolded proteins signals the induction of BiP expression (28)
. To test whether SA binding to BiP transmitted a signal for BiP overexpression, we treated HFF with high concentrations of SA or its analogs and determined BiP levels. BiP proteins were constitutively expressed in control HFF and neither SA nor aspirin altered the BiP protein levels (Fig. 7
). Similarly, neither 2,6-DHBA nor 2,5-DHBA altered the BiP protein level (Fig. 7)
. Endoplasmic reticulum (ER) stress has been reported to induce the expression of C/EBP homologous protein (CHOP) (29)
. CHOP forms heterodimers with C/EBP and suppresses transcriptional activation by C/EBP. Since C/EBPß is a key transactivator of COX-2 promoter and salicylate suppresses COX-2 promoter activity by inhibiting C/EBPß binding to a proximal regulatory element on COX-2 promoter (16)
, we determined whether salicylate binding to BiP induces CHOP expression. CHOP protein was undetected in HFF as determined by Western blot analysis using a CHOP-specific antibody and salicylate at 10-5 M to 10-3 M did not induce CHOP (data not shown).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
50 µM) is therapeutically relevant, as aspirin or other salicylates achieve their therapeutic effect at a concentration close to the dissociation constant of SA binding to BiP. SA could bind to an isoform of BiP protein whose NH2-terminal sequence and molecular weight are close to the authentic BiP. However, we consider this unlikely because BiP is encoded by a single gene copy and there are no reported data or functional implication of the existence of BiP isoforms. In our final purification step, a minor 56 kDa protein was also detected to have SA binding activity. Its identity is unclear at this time. As binding kinetics of crude extracts fits a single class of binding site for SA, it is unlikely that this 56 kDa protein plays a major part in SA binding.
BiP (also known as GRP78) is an ER resident protein. It belongs to the heat shock protein 70 (HSP70) family and the glucose-regulated protein family with important chaperone functions (28)
. A major activity of BiP is binding of newly synthesized polypeptides, allowing appropriate protein folding and transport across the membrane (27
, 30)
. These complex actions are executed by coordinated biochemical reactions including rapid binding of polypeptides by ATP-bound BiP, activation of ATPase, which converts ATP to ADP, slow dissociation of polypeptide in ADP-bound BiP to allow for proper protein folding, and an eventual replacement of ADP by ATP resulting in rapid dissociation of the properly folded proteins from BiP (30)
. Thus, polypeptide binding to BiP is a pivotal step in BiP chaperone function. Based on a computer scoring system verified by experimental data, it has been demonstrated that BiP binds to proteins with heptameric motifs consisting of hydrophobic and aromatic amino acid residue (31
32
33)
. Several such motifs were reported in the HIV gp160 protein (26)
. To determine whether SA binds to the peptide binding site, we chose an active heptapeptide (KYWWNLL) and a control heptapeptide (TWFNSTW) from the reported gp160 sequence (26)
. We confirmed that the active peptide increased ATPase activity, an indication of active peptide binding, whereas the control peptide did not have a significant effect on ATPase activity. That the active peptide competed with [14C]SA binding in a concentration-dependent manner is consistent with binding of SA to the peptide binding site of BiP. This is further confirmed by inhibition of peptide-stimulated ATPase activity by SA but not an inactive benzoate derivative.
Binding of SA to the peptide binding site of BiP may interfere with appropriate protein folding, which may in turn cause a condition resembling ER distress. A consequence of ER stress is the feedback up-regulation of BiP expression (28)
. However, we did not find increased BiP protein levels in HFF treated with sodium salicylate even at high concentrations. Nor did we observe the expression of CHOP, another molecule induced by ER stress (29)
, in HFF after salicylate treatment. These data suggest that binding of SA to BiP does not cause global ER stress as seen in glucose starvation. Salicylate binding to BiP could cause inhibition of protein synthesis, as SA has been shown to inhibit protein synthesis (34)
. However, the concentrations of sodium salicylate that suppress global proteins synthesis were much higher than the Kd value for salicylate binding to BiP. SA at pharmacological concentrations could interfere with synthesis of a selective group of proteins that are highly dependent on BiP for folding and transport. There is little information about the proteins whose synthesis and function are significantly altered by salicylate binding to BiP. A recent report that salicylate at therapeutic concentration inhibits p70/p85 S6 kinase activity (35)
may provide a clue. p70/p85 S6 kinase targets cellular proteins, including C/EBPß, for phosphorylation (36)
. Since salicylate suppresses C/EBPß binding to COX-2 promoter region, thereby inhibiting COX-2 promoter activity (15
, 16)
, it may be speculated that through its binding to BiP, salicylate may perturb a protein involved in activation of p70/p85 S6 kinase. Our results show that aspirin is as potent as sodium salicylate in inhibiting [14C]SA binding to BiP. This is consistent with our earlier finding that aspirin and sodium salicylate inhibit COX-2 transcription at equivalent concentrations (15)
. Evaluation of a series of SA metabolites and analogs suggests that the OH group at position 2 of benzoic acid appears to be crucial for BiP binding and inhibition of C/EBPß-dependent transcription of genes such as COX-2.
|
ACKNOWLEDGMENTS |
|---|
Received for publication April 12, 2001.
Revision received June 29, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. Baumeister, D. Dong, Y. Fu, and A. S. Lee Transcriptional induction of GRP78/BiP by histone deacetylase inhibitors and resistance to histone deacetylase inhibitor-induced apoptosis Mol. Cancer Ther., May 1, 2009; 8(5): 1086 - 1094. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Pyrko, A. H. Schonthal, F. M. Hofman, T. C. Chen, and A. S. Lee The Unfolded Protein Response Regulator GRP78/BiP as a Novel Target for Increasing Chemosensitivity in Malignant Gliomas Cancer Res., October 15, 2007; 67(20): 9809 - 9816. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. S. Lee GRP78 Induction in Cancer: Therapeutic and Prognostic Implications Cancer Res., April 15, 2007; 67(8): 3496 - 3499. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Silva, D. Wang, A. A. Komar, B. A. Castilho, and B. R. G. Williams Salicylates Trigger Protein Synthesis Inhibition in a Protein Kinase R-like Endoplasmic Reticulum Kinase-dependent Manner J. Biol. Chem., April 6, 2007; 282(14): 10164 - 10171. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. A. Cieslik, Y. Zhu, M. Shtivelband, and K. K. Wu Inhibition of p90 Ribosomal S6 Kinase-mediated CCAAT/Enhancer-binding Protein {beta} Activation and Cyclooxygenase-2 Expression by Salicylate J. Biol. Chem., May 6, 2005; 280(18): 18411 - 18417. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
