| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Vascular Biology |
From the Division of Biotechnology Discovery Research, Lilly Research Laboratories, Lilly Corporate Center, Indianapolis, Ind.
Correspondence to Brian W. Grinnell, PhD, Biotechnology Discovery Research, Lilly Research Laboratories, Indianapolis, IN 46285. E-mail bgrinnell{at}lilly.com
| Abstract |
|---|
|
|
|---|
Methods and Results— APC inhibited endothelial TRAIL expression and secretion and its induction by cell activation. To explore the mechanism, we examined factors associated with TRAIL regulation and demonstrated that APC increased the level of EGR-1, a transcriptional factor known to suppress the TRAIL promoter. APC also induced a significant increase in phosphorylation of ERK-1/2, required to activate EGR-1 expression. Activation of ERK-1/2 was dependent on the protease activated receptor-1 (PAR-1), but independent of the endothelial protein C receptor (EPCR). Using siRNA, we found that the effect of APC on the EGR-1/ERK signaling required for TRAIL inhibition was dependent on the S1P1 receptor and S1P1 kinase.
Conclusions— Our data suggest that APC may provide cytoprotective activity by activating the ERK pathway, which upregulates EGR-1 thereby suppressing the expression of TRAIL. Moreover, we provide evidence that APC can induce a cell signaling response through a PAR-1/S1P1-dependent but EPCR-independent mechanism.
We report that APC can suppress the proapoptotic mediator TRAIL by activating the ERK pathway to upregulate EGR-1, a negative regulator of TRAIL expression. The effect of APC was PAR-1– and S1P1-dependent, but independent of the endothelial protein C receptor, suggesting a mechanism to suppress injury in cells not expressing this receptor.
Key Words: endothelial protein C receptor apoptosis protease activated receptor APC TRAIL
| Introduction |
|---|
|
|
|---|
Apoptotic signals induced by TRAIL have been shown to be negatively regulated by the MAPK/ERK pathway,15,16 and cells become susceptible to apoptosis when ERK is dephosphorylated.17,18 Recent studies have also suggested that the early transcription factor, early growth response factor –1 (EGR-1), acts as a negative regulator of TRAIL.19 Of interest, EGR-1 function has been shown to be dependent on ERK1/2 phosphorylation.20
APC is a serine protease with a well characterized anticoagulant activity. Recent studies have demonstrated that APC has antiinflammatory and cytoprotective activities, which are not attributed to the anticoagulant activity of APC.21–23 APC can initiate cell-signaling pathways in cells by activating protease-activated receptor 1 (PAR-1).24 PAR-1 is a member of the 7 transmembrane domain G protein–coupled receptor family, and its activation requires cleavage at a specific site within its extracellular amino terminus.25 This cleavage produces a new amino terminus, which then acts as its own tethered ligand, but optimal cleavage occurs when APC is juxtaposed to PAR-1. It is suggested that APC must bind to endothelial protein C receptor (EPCR) for it to be juxtaposed to PAR-1. Indeed, many studies have demonstrated antiinflammatory and antiapoptotic activities of APC that are dependent on PAR-1 and EPCR; however, previous studies have not elucidated alternative pathways that are activated by APC.24,26,27
The present study was undertaken to determine whether APC could modulate the activity of endothelial TRAIL, an important mediator of vascular injury. We show that APC suppresses the expression of TRAIL mRNA by an EPCR-independent mechanism. We demonstrate that APC increases the phosphorylation and activation of ERK and increases the transcriptional activity of EGR-1, which in turn suppresses TRAIL mRNA expression. We also demonstrate that the activation of EGR-1 by APC is dependent both on PAR-1 and sphingosine 1-phosphate receptor (S1P1)-dependent signaling. These results describe a novel pathway by which APC may protect endothelial cells from mediators such as TRAIL and describe an APC-induced cellular response that is EPCR-independent.
| Methods |
|---|
|
|
|---|
–Induced TRAIL Expression
30 nmol/L Kd for APC interaction with EPCR, although these concentrations are above the therapeutic blood concentration achieved in APC-treated severe sepsis patients. The following morning, 1 ng/mL of TNF-
was added to the cells for 1, 6, or 12 hours, then cells were lysed with M-PER and protein concentrations were determined by BCA assay according to the manufacturers protocol. The quantity of membrane-associated TRAIL was determined by ELISA, and the standard curve was generated using recombinant human TRAIL. Briefly, 96-well plates were coated with anti-TRAIL antibody overnight. The following day, wells were blocked for 1 hour with 4% bovine albumin in PBS. Samples or recombinant human TRAIL were added to the plate for 1 hour, followed by biotinylated anti-TRAIL Ab for 1 hour. Next, 100 µL of HRP streptavidin was added for 30 minutes. Wells were extensively washed with PBS-Tween between each of these steps. The ELISA was developed by adding TMB substrate solution for 20 minutes and stopped by adding H2SO4. Plates were read at A450–A595. The effect of blocking EPCR with an siRNA on APC inhibition of TRAIL was performed in confluent HUVECs treated with a nonspecific control or EPCR-specific siRNA for 48 hours before treatment with 80 nmol/L APC. The level of EPCR expression in control and EPCR siRNA-treated cell was determined using a Western blot with S6 ribosomal protein (S6 RP) as a loading control.
Statistics
All experiments were performed at least in triplicate. Statistics were performed using JMP software (SAS Institute). Error bars indicate SEM. For the analysis of TRAIL response genes, comparison were displayed as a heat map generated using hierarchical clustering in JMP 5.1 (Ward method).
Please see supplemental materials, available online at http://atvb.ahajournals.org for additional details on Materials and Methods.
| Results |
|---|
|
|
|---|
–Induced TRAIL mRNA Expression Is Decreased by APC
was blocked by APC treatment. To determine whether the effect of APC was concentration-dependent, cells were stimulated with various concentrations of APC (0 to 160 nmol/L), and the concentration of TRAIL was measured by ELISA. As Figure 1B demonstrates, APC inhibited the amount of TRAIL secreted from HUVECs in a concentration-dependent manner, with a half-maximal effect at approximately 30 nmol/L. As was observed with the mRNA levels, APC also suppressed the induction of TRAIL secretion induced by TNF-
(Figure 1C).
|
A time-course study was performed to determine how long APC was effective at reducing TRAIL. As shown in Figure 1D, TRAIL was unchanged during the first hour after TNF-
or APC treatment. At 6 hours, TRAIL was increased by 4.8-fold in TNF-
–treated cells; however, TRAIL was reduced by 35% by APC treatment. At 12 hours, TRAIL was further increased by 1.3-fold in TNF-
–stimulated cells, but was unchanged in cells that were costimulated with APC. These results suggest that a single exposure of cells to APC reduces TNF-
–mediated TRAIL protein levels for at least 12 hours. Overall, our results suggest that APC can modulate TRAIL expression alone or after its induction with inflammatory mediators such as TNF-
.
APC Decreases TRAIL Secretion Through an EPCR-Independent Mechanism
Several studies have shown that the antiapoptotic activity of APC can be mediated by the EPCR-dependent activation of PAR-1. To determine whether APC reduced TRAIL secretion via a similar mechanism, cells were treated with APC with or without TNF-
administration in the presence of an antibody that prevents APC interaction with EPCR (JRK1494). As shown in Figure 2A, the ability of APC to inhibit the expression of TRAIL was not affected by blocking the APC-EPCR interaction with anti-EPCR antibody JRK-1494 or with RCR-252 (data not shown). Moreover, blocking the APC-EPCR interaction had no effect on the ability of APC to reduce the TNF-induced secretion of TRAIL (Figure 2B). As a control to show that APC interaction with EPCR had in fact been blocked, in parallel experiments we were able to block the reported EPCR-dependent suppression of ICAM23,25 with the anti-EPCR antibody (supplemental Figure I), and to suppress the known EPCR-dependent effects of APC on staurosporine-induced apoptosis21,44 (supplemental Figure II) and on thrombin-induced changes in permeability determined by the BSA-Evans blue dye method described by Feistritzer and Riewald28 (supplemental Figure III). While these data with blocking antibodies strongly suggested EPCR-independence of APC on TRAIL, we further confirmed this by treating cells with an siRNA, which ablated EPCR expression in the cell (Figure 2C, inset) but had no effect on the ability of APC to suppress the expression of TRAIL.
|
APC Increases the Expression and Functional Activity of EGR-1
Recent studies have suggested that the early transcription factor, EGR-1, acts as a negative regulator of TRAIL.19 To further explore the mechanism for the inhibition of TRAIL by APC, we examined the relationship between the expression of EGR-1 and TRAIL across untreated and TNF-treated cells. As shown in Figure 3A, the level of TRAIL expression was negatively associated with the expression of EGR-1. Moreover, we examined the level of EGR-1 and TRAIL at 6 hours after TNF treatment and observed a 53% decrease in EGR-1 expression and corresponding 61% increase in TRAIL expression (data not shown). These data are consistent with the previous report showing that TRAIL is negatively regulated by EGR-1.19
|
To determine whether APC might suppress TRAIL via EGR-1, we analyzed the level of EGR-1 DNA binding activity by gel shift assay in untreated cells and those treated with APC. We observed significantly higher levels of EGR-1 in cells treated with APC (Figure 3B). As also shown, the binding was specific as cold binding site probe could completely inhibit the gel shift.
APC Increases pERK1/2 Independent of EPCR
EGR-1 binding activity has been shown to be dependent on ERK1/2 phosphorylation,20 so we sought to determine whether APC might be affecting EGR-1 by modulating the level of ERK1/2. As shown in Figure 4A, APC induced a time-dependent increase in the phosphorylation of ERK1/2. There was no change in the level of total ERK1/2 protein in these experiments as determined by Western blot analysis using an antibody to total ERK1/2 (data not shown). If the effect of APC on EGR-1 suppression of TRAIL was dependent on ERK1/2 phosphorylation, we would expect the effect to be EPCR independent as was shown in Figure 2. Therefore, we examined the effect of APC on ERK1/2 phosphorylation in the presence of the anti-EPCR antibody. As shown in Figure 4B, the increase in pERK induced by APC was not suppressed by blocking the EPCR interaction. However, the effect of APC was completely suppressed by the blocking antibody to PAR-1. These data suggest that APC can induce EGR-1/ERK-1/2 cell signaling in a PAR-1–dependent but EPCR-independent manner.
|
Effect of S1P1 Signaling on APC-Dependent EGR-1 Activation
Recent studies have demonstrated that the S1P1 receptor can effect cell signaling by APC.28,29 As the effect of APC on TRAIL and the ERK1/2-EGR pathway appears to be PAR-1–dependent but EPCR-independent, we sought to determine the role of S1P1, especially as recent studies have shown that S1P1 signaling occurs with activation of the ERK1/2 pathway.30–32 As shown in Figure 5A, a siRNA blocking S1P1 (Edg1), but not a scrambled siRNA, significantly decreased the EGR1-DNA interactions induced by APC. We also examined the effect of SPHK1, the kinase required for generating S1P and activating the S1P1 receptor via PAR-1. As shown, the siRNA against SPHK1 also significantly suppressed the transcriptional induction of EGR-1 by APC. Our proposed model for the EPCR-independent activation of signaling by APC to suppress TRAIL expression, based on the model described by Camerer and Coughlin, 33 is shown in Figure 5B.
|
TRAIL-Response Genes and APC Inhibition
We sought to assess the cellular consequence of APC-mediated TRAIL suppression by examining genes known to be suppressed or induced by TRAIL. For this analysis we examined cyclin-dependent kinase 4 (CDK4), previously shown to be suppressed by TRAIL,34 and several genes (SOD2, IFI-15K/G1P2, NK4, HLA-A, HLA-C, TFPI-2) shown to be induced by TRAIL. Using both treated and untreated HUVECs, we first examined the relationship between the expression levels of TRAIL and the expression of these genes. Consistent with their known regulation by TRAIL, we observed a highly significant positive correlation between TRAIL and the set of TRAIL-induced genes, and a highly negative correlation between TRAIL and CDK4 (Figure 6A). We treated cells with APC to determine whether the APC-mediated reduction in TRAIL resulted in a suppression of TRAIL-induced genes. Figure 6B shows the level of TRAIL in this experiment. As shown in Figure 6C, the expression of SOD2, IFI-15K/G1P2, NK4, HLA-A, HLA-C, TFPI-2 were induced as TRAIL levels increased and suppressed by APC coincident with its ability to inhibit TRAIL secretion. Moreover, the TRAIL-suppressed gene CDK4 was increased by APC treatment, coincident with the suppression of TRAIL secretion. These data strongly suggest that the reduction in TRAIL expression by APC alters cellular pathways in HUVECs known to be mediated by TRAIL.
|
| Discussion |
|---|
|
|
|---|
, which is upregulated during sepsis. These observations provide insight into a potential mechanism by which APC might decrease apoptosis and improve organ function in severe sepsis patients. Moreover, the ability of APC to suppress TRAIL-regulated genes associated with proinflammatory and immunomodulatory activity (Figure 6), ie, beyond the apoptosis pathway, suggests additional potential protective mechanisms. The initial suggestion that APC could block apoptosis and promote cell survival came from a transcriptional profiling experiment, 23 and these results were confirmed in cell-based assays and animal models.23,25,36 These studies suggested that APC mediates the suppression of apoptosis in a PAR-1/EPCR dependent mechanism by decreasing proapoptotic protein such as Bax, increasing antiapoptotic protein Bcl-2, and inhibiting the activation of caspase 3 and 8. In the current study, we show that APC can decrease the mRNA levels and secretion of the potent apoptosis factor TRAIL, via activation of the EGR-1/ERK-1/2 pathway. In contrast to previous reports, the mechanism of this effect appears to be PAR-1 dependent, but EPCR-independent. To the best of our knowledge, this is the first study to demonstrate that APC may activate cellular pathways in a PAR-1–dependent but EPCR-independent mechanism. Our data open the possibility of another cofactor involved in APC–PAR-1 interaction. A logical candidate would be thrombomodulin (TM); however, we were unable to demonstrate that APC-TM interactions are involved with the cell signaling we have described (data not shown).
Recent studies have suggested that APC signaling via PAR-1 can be transactivated by the S1P-S1P1 receptor interaction.28,29 PAR-1 is coupled to several G proteins including G
-β
. The β
subunit ultimately leads to the activation of the MAPK and ERK pathways, which can activate several transcription factors that transcribe genes containing a serum response element (SRE).37 The transcription factor EGR-1 has 5 SREs, and studies have demonstrated that ERK can increase functional EGR-1.38 Recent studies have also shown that S1P1 signaling activates the ERK1/2 pathway.30–32 In the current study, we demonstrate that APC induces ERK-1/2 phosphorylation and increases the level of EGR-1 mRNA and functionally active protein. However, unlike previous studies,28,29 the effect of APC signaling though PAR-1 and S1P1 signaling was independent of EPCR. EPCR has been shown to be expressed primarily on large vessels, but not in the normal microvasculature, and several studies have demonstrated that EPCR is differentially regulated during tissue injury.39–41 Thus, protective signaling mechanisms independent of EPCR may be important under conditions of little or no EPCR expression.
APC plays an important role in the modulation of vascular function not only through inhibition of thrombin generation, but also by receptor-mediated effects via PAR-1 activation, which results in activation of cytoprotective and antiinflammatory pathways. The requirement of EPCR for APC to signal via PAR-1 has been documented, but its relevance and physiological role have been questioned based on kinetics and tissue distribution.42 Our data suggest that the activation of protective signaling pathways, such as the ability to activate EGR-1/ERK-1/2 signaling to suppress TRAIL, can occur independent of EPCR. This suggests that conclusions on the role of signaling in the efficacy of APC may not be solely dependent on the tissue expression of this receptor. Recent studies have shown that the components required for the activation of PC and for APC signaling are colocalized to lipid rafts,43 indicating the complex nature of the microenvironment on the membrane surface involved in the generation of APC and in PAR-1 signaling. Clearly, further studies will be needed to dissect the relative importance and balance of the EPCR-independent versus EPCR-dependent signaling in the in vivo function of APC, and to determine the importance of EPCR-independent signaling during endogenous APC generation versus exogenous APC exposure during therapy. Overall, the results described in this report provide a new mechanistic understanding for the antiapoptotic and antiinflammatory functions of APC and define a novel signaling pathway that is not dependent on EPCR.
| Acknowledgments |
|---|
Disclosures
The authors are employees of Eli Lilly and Co, who produces human activated protein C for treatment of severe sepsis.
| Footnotes |
|---|
| References |
|---|
|
|
|---|
Related Article:
This article has been cited by other articles:
![]() |
L. J. Toltl, S. Beaudin, P. C. Liaw, and and the Canadian Critical Care Translational Biolo Activated Protein C Up-Regulates IL-10 and Inhibits Tissue Factor in Blood Monocytes J. Immunol., August 1, 2008; 181(3): 2165 - 2173. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Secchiero and G. Zauli The Puzzling Role of TRAIL in Endothelial Cell Biology Arterioscler. Thromb. Vasc. Biol., February 1, 2008; 28(2): e4 - e4. [Full Text] [PDF] |
||||
![]() |
L. A. O'Brien, M. A. Richardson, D. T. Berg, B. Gerlitz, A. Gupta, and B. W. Grinnell The Puzzling Role of TRAIL in Endothelial Cell Biology Arterioscler. Thromb. Vasc. Biol., February 1, 2008; 28(2): e5 - e6. [Full Text] [PDF] |
||||
![]() |
M. Riewald and R. A. Schuepbach Protective Signaling Pathways of Activated Protein C in Endothelial Cells Arterioscler. Thromb. Vasc. Biol., January 1, 2008; 28(1): 1 - 3. [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
ATVB Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2007 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |