| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Vascular Biology |
B
From the Cardiovascular Research Center (D.T.B., A.W., S.S., M.E.H., M.A.S., C.C.H.), and the Departments of Pharmacology (C.C.H.) and Microbiology (A.W.O., M.A.S.), University of Virginia, Charlottesville.
Correspondence to Catherine C. Hedrick, Cardiovascular Research Center, University of Virginia, PO Box 801394, 415 Lane Rd; MR5 Rm G123, Charlottesville, VA 22908. E-mail cch6n{at}virginia.edu
| Abstract |
|---|
|
|
|---|
Methods and Results Freshly isolated aortic endothelial cells (EC) from 12/15-LO transgenic mice had significantly greater nuclear factor-
B (NF-
B) activation and ICAM mRNA expression compared with C57BL/6J control. 12/15-LO transgenic EC showed elevated RhoA activity, and inhibition of RhoA using either C3 toxin or the Rho-kinase inhibitor Y-27632 blocked NF-
B activation, ICAM-1 induction, and monocyte adhesion. Furthermore, we show that 12SHETE activates protein kinase C
, which forms a complex with active RhoA and is required for NF-
Bdependent ICAM expression in response to 12SHETE.
Conclusions The 12/15-LO pathway stimulates ICAM-1 expression through the RhoA/protein kinase C
-dependent activation of NF-
B. These findings identify a major signaling pathway in EC through which 12/15-LO contributes to vascular inflammation and atherosclerosis.
12/15-lipoxygenase (12/15-LO) promotes monocyte:endothelial interactions and atherosclerosis. We report that 12/15-LO activation in endothelium activates the small GTPase RhoA which then activates PKC
and NF
B. The activation of this pathway increases endothelial ICAM-1 expression in vivo. ICAM-1 is a primary regulator of monocyte adhesion to endothelium.
Key Words: endothelium 12/15-lipoxygenase RhoA ICAM-1
| Introduction |
|---|
|
|
|---|
12/15-lipoxygenase (12/15-LO) incorporates molecular oxygen in a stereospecific manner into arachidonic and linoleic acids to generate 12- and 15-S-hydroxyeicosatetraenoic acids (12SHETE/15SHETE) and 13-S-hydroxyoctadecaenoic acid (13SHODE), respectively.68 Several vascular cell types produce 12/15-LO eicosanoids, including EC, smooth muscle cells, monocytes, and platelets. Several studies, including ours, have indicated that the 12/15-LO pathway is proinflammatory and increases atherosclerosis development in mice.913 12SHETE, a major product of 12/15-LO, is present in the vasculature in the nanomolar range and activates EC to stimulate monocyte adhesion.14,15 We have previously reported a significant increase in expression of ICAM-1 on aortic endothelium of 12/15-LO transgenic mice (12/15-LOTG), suggesting 12SHETE stimulates monocyte adhesion through enhancing endothelial ICAM-1 expression.13 ICAM-1 and VCAM-1 are both believed to be important for early atherogenesis.16,17 Several studies have indicated that deletion of ICAM-1 in apoE-deficient mice attenuates atherosclerosis progression,1820 although Cybulsky and colleagues reported that VCAM-1, rather than ICAM-1, was the primary regulator of atherosclerosis.21 Kevil et al reported that ICAM-1 was critical for mediating monocyte adhesion to murine EC,22 although Huo and colleagues reported that VCAM-1 played a critical role in mediating this process.23 We have previously reported that both VCAM-1 and ICAM-1 mediate monocyte:endothelial adhesion in mice.13,24 Thus, studies to date suggest that both VCAM-1 and ICAM-1 mediate early inflammatory events in the vessel wall.
Activation of nuclear factor-
B (NF-
B) in EC stimulates expression of several proteins involved in inflammatory cell recruitment, including ICAM-1, VCAM-1, and E-selectin.25 Furthermore, NF-
B shows enhanced expression and activation in atherosclerosis-prone sites in vivo, suggesting NF-
Bdependent gene expression is involved in early atherogenesis.26,27 In the current study, we show that 12/15-LO stimulates enhanced ICAM-1 gene transcription through NF-
B. 12SHETE stimulates NF-
B activation and NF-
Bdependent ICAM-1 expression through RhoA and PKC
. Monocyte adhesion to 12/15-LO overexpressing EC and 12SHETE-treated EC is abrogated by inhibition of Rho kinase (ROCK) or protein kinase C
(PKC
). This study is the first to determine the mechanism for the proinflammatory and proatherosclerotic effects of 12/15-LO.
| Materials and Methods |
|---|
|
|
|---|
| Results |
|---|
|
|
|---|
|
We examined the ability of 12SHETE to modulate expression of other molecules that are relevant for mediating monocyte:endothelial interactions. We found that 12SHETE stimulated monocyte chemoattractant protein (MCP-1) mRNA expression in EC by 5.5 fold. E-selectin mRNA and VCAM-1 mRNA expression were not changed by 12SHETE. For the present study, we chose to focus only on the regulation of endothelial ICAM-1 expression by 12SHETE.
12SHETE Stimulates NF-
B Activation
The transcription factor NF-
B regulates inflammatory gene expression in EC. We transfected mouse EC with a NF-
Bluciferase reporter plasmid. We found that EC, freshly isolated from 12/15-LOTG mice, showed dramatically elevated luciferase activity compared with C57BL/6J control (100 000 luciferase units for 12/15-LOTG mouse EC compared with only 2500 luciferase units for B6; P<0.0001). We next tested whether 12SHETE directly stimulated NF-
B activity using HAEC transfected with the NF-
Bluciferase reporter plasmid. Treatment of these HAEC with 100 nM 12SHETE resulted in a 3-fold induction of luciferase activity (please see Figure IA, available online at http://atvb.ahajournals.org). Consistent with previous assays, 12RHETE had no effect.
On activation, the I
B kinase complex phosphorylates I
B resulting in its degradation. This unmasks a nuclear localization signal on the p65 subunit resulting in its translocation to the nucleus. Using immunocytochemistry, we demonstrated that treatment of aortic endothelium with 100 nM 12SHETE rapidly stimulated translocation of NF-
B from the cytosol to the cell nucleus, with &60% of cells showing nuclear p65 staining in response to 1 hour of 12SHETE treatment (Figure 2). In addition to nuclear translocation, p65 undergoes other posttranslational modifications which modulate its activity, including phosphorylation of Ser536 in p65 which stimulates maximal transcriptional activation28. Stimulation with 100 nM 12SHETE resulted in a 3-fold increase in p65 phosphorylation, whereas 12RHETE failed to stimulate p65 Ser536 phosphorylation (see Figure IB).
|
To determine whether NF-
B activation mediates the enhanced ICAM-1 expression in 12SHETE-treated cells, EC were transfected with a nonphosphorylatable form of I
B, termed super-repressor I
B (SR-I
B). Expression of SR-I
B sequesters NF-
B in the cytosol, thereby preventing NF-
Bdependent gene expression. Overexpression of SR-I
B prevented the 12SHETE-induced activation of ICAM-1 expression as assessed by immunocytochemistry (please see Figure IC). Taken together, these data illustrate that elevated 12SHETE production in 12/15-LOTG EC stimulates activation of NF-
B.
RhoA Is Activated in 12/15-LO Transgenic EC
The Rho family of GTPases can activate NF-
Bdependent gene expression in EC.29 Rac stimulates NF-
B through the reduced nicotinamide-adenine dinucleotide phosphate oxidasedependent production of reactive oxygen species.30 In contrast, RhoA stimulates NF-
B through its downstream effector, ROCK.31 To evaluate the role of RhoA in 12SHETE-induced ICAM-1 expression in vivo, we examined RhoA activation in 12/15-LOTG and control B6 mice. RhoA activity was measured by affinity precipitation of active RhoA with the Rho-binding domain of Rhotekin as previously described.32 As shown in Figure 3, 12/15-LOTG murine aortic EC have a 2-fold increase in RhoA activation. 12SHETE (100 nM for 10 minutes) also stimulated RhoA activation in EC within 1 minute and was maximal at 10 minutes (data not shown).
|
12SHETE Activates PKC
Through RhoA
Activated PKC
mobilizes from the cytosol to the membrane. Examination of PKC
cellular localization revealed a 3-fold increase in membrane-associated PKC
in 12/15-LOTG EC compared with C57BL/6J control cells (please see Figure II, available online at http://atvb.ahajournals.org). This increase in membrane-associated PKC
in the 12/15-LOTG EC was blocked by the Rho-kinase (ROCK) inhibitor Y27632 (5 µmol/L for 4 hours). Similarly, we found that 12SHETE (100 nM for 1 hour) stimulated PKC
mobilization to the membrane which was blocked by Y27632 (data not shown).
Studies have reported a physical association of RhoA and PKC
in the membrane for maximal activation.3335 Using a coimmunoprecipitation assay, we found that both RhoA and PKC
were part of a protein complex in 12SHETE-treated cells, but not in control cells (Figure 4). The protein complex was disrupted by pharmacological inhibition of either ROCK or PKC
(Figure 4). These data indicate that the elevated PKC
activity in 12/15-LOTG is through a RhoA-dependent mechanism.
|
We speculate that 12SHETE may bind to a guanine nucleotide-binding protein (G protein)coupled receptor (GPCR) on the endothelial surface. Several other GPCRs have been identified for eicosanoids, including LTB4.36 Identifying a specific 12SHETE GPCR is an ongoing process in the laboratory and is beyond the scope of the present study. However, in light of our data that RhoA is part of the signaling cascade in endothelium activated by 12SHETE, we used G12 and G13 minigenes to inhibit G12/13 signaling in EC in response to 12SHETE. These minigenes encode for the COOH termini of heterotrimeric G proteins and, thus, serve as competitive inhibitors of receptor G-protein interactions.37 We found that inhibition of G12 and G13 (but not Gi) reduced 12SHETE induction of ICAM-1 by 60% (please see Figure III, available online at http://atvb.ahajournals.org), suggesting that 12SHETE indeed may activate a GPCR in EC.
Regulation of NF-
B Activation by RhoA and PKC
To verify that RhoA was an activator of NF-
B, we treated cells with the ROCK inhibitor Y-27632 (10 µmol/L for 1 hour) and examined nuclear translocation of the p65 subunit of NF-
B after 12SHETE treatment (100 nM for 30 minutes). ROCK inhibition completely blocked NF-
B translocation to the nucleus, but did not prevent tumor necrosis factor
induced NF-
B activation (Figure 5A). We have found that 12SHETE can also activate Rac in EC (data not shown). To verify that NF-
B is not activated by Rac, we examined p65 translocation in the presence of dominant negative constructs for both Rac and Cdc42. We found that dominant negative constructs of Rac and Cdc42 did not inhibit NF-
B translocation (Figure 5B), suggesting that NF-
B translocation in response to 12SHETE was mediated only by RhoA.
|
To determine whether NF-
B activation is also modulated by PKC
, we examined translocation of the p65 subunit of NF-
B after 12SHETE treatment (100 nM for 30 minutes) with the addition of the PKC
inhibitor Go6976. PKC
inhibition blocked 12SHETE-mediated NF-
B translocation to the nucleus (please see Figure IV, available online at http://atvb.ahajournals.org). These results suggest that 12SHETE stimulates NF-
B activation through a RhoA/PKC
signaling pathway in EC.
Inhibition of RhoA and PKC
Blocks Monocyte:EC Interactions in 12/15-LOTG
Aortic EC from 12/15-LOTG have increased surface expression of ICAM-1.13 We determined whether inhibition of RhoA or PKC
would reduce ICAM-1 expression in 12/15-LOTG murine aortic EC. We found that inhibition of either RhoA or PKC
reduced ICAM-1 mRNA expression in 12/15-LOTG EC (Figure 6).
|
Finally, we examined whether inhibition of either RhoA or PKC
in aortic EC from 12/15-LOTG would block monocyte adhesion. 12/15-LOTG EC had a significant 3-fold increase in WEHI monocyte cell adhesion (please see Figure VA, available online at http://atvb.ahajournals.org). WEHI monocyte adhesion to 12/15-LO transgenic EC was significantly reduced by 50% on pharmacological inhibition of either RhoA (using C3 toxin or Y27632) or PKC
(using Go6976) (Figure VA). Use of siRNA to PKC
also reduced WEHI monocyte adhesion by 50% (Figure VA). Interestingly, simultaneous addition of both RhoA and PKC
inhibitors to 12/15-LOTG EC completely blocked WEHI monocyte adhesion (100% reduction) (Figure VA). Similar results of inhibition of RhoA and PKC
on monocyte adhesion were obtained with 12SHETE. Inhibition of PKC
using Go6976 or a specific PKC
siRNA reduced 12SHETE-mediated MonoMac6 monocyte adhesion by 70% (please see Figure VB). Inhibition of RhoA activation using either C3 toxin or Y27632 reduced 12SHETE-mediated monocyte adhesion by 60% to 70% (Figure VB). Using a combination of G06976 and Y27632 to simultaneously inhibit both molecules blocked WEHI adhesion by 100% (Figure VB). These data suggest that a RhoA/PKC
pathway regulates monocyte: endothelial adhesion similarly in response to 12/15LO upregulation or addition of 12SHETE.
| Discussion |
|---|
|
|
|---|
, which induced EC ICAM-1 expression through NF-
B. This is the first report to indicate that the 12/15-LO pathway in EC can directly modulate monocyte adhesion through RhoA-dependent and NF-
Bdependent regulation of endothelial ICAM-1 expression. Thus, 12SHETE acts as a second messenger within endothelium to stimulate proinflammatory signaling contributing to vascular inflammation and atherosclerosis.
In the current study, we found significant upregulation of endothelial ICAM-1 expression at both the mRNA and protein levels by 12SHETE. Furthermore, we describe that the pathway by which 12SHETE stimulates ICAM-1 expression is through RhoA and NF-
B. Sultana and colleagues reported that 12SHETE activated NF-
B in human umbilical vein EC.38 However, these investigators did not examine RhoA activation or ICAM-1 regulation by 12SHETE. In a previous study by our group, we treated HAEC with 1 nM 12SHETE and did not observe an increase in ICAM-1.15 We did not examine 100 nM 12SHETE in this previous study. We have reported that diabetic db/db mice, which show a 3-fold increase in 12/15-LO activity, have little or no increase in endothelial ICAM-1.24 These mice have a 3-fold increase in 12SHETE production in vivo, but we saw no change in endothelial ICAM-1 expression in these mice. Therefore, we were surprised to observe regulation of ICAM-1 expression in response to 12SHETE. We acknowledge the discrepancy between the results of the db/db mouse study and this present study. One explanation for the different findings is that ICAM-1 expression is modulated by multiple factors, especially in the setting of diabetes. There are many factors that can regulate ICAM-1 expression in diabetes, including glucose, nitric oxide, Th2 cytokines, and bioactive lipids.3942 Sphingosine-1-phosphate, oxidized phospholipids, and lipoxins are bioactive lipids that may oppose the proinflammatory effects of 12SHETE. Indeed, Kieran and colleagues have shown that lipoxin A4 levels reduce ICAM-1 expression in a kidney IRI model.43 We have shown that sphingosine-1-phosphate prevents endothelial activation in response to tumor necrosis factor
,44 and we observe similar effects in diabetic db/db mice (data not shown). Bochkov and colleagues reported that specific oxidized phospholipids can prevent adhesion molecule expression and protect mice from lipopolysaccharide-mediated sepsis.45 In a 12/15-LOTG, which had a 4-fold expression of 12/15-LO gene, we did observe a significant increase in endothelial ICAM-1 expression,13 which was the initial basis for the current study. We focused on 12SHETE in this study because it is a primary product of the 12/15-LO pathway. However, other eicosanoids produced by this pathway, including 13SHODE and 12-hydroperoxyeicosatetraenoic acid, also stimulate monocyte adhesion to endothelium.13 Thus, although this current study is focused on 12SHETE, additional eicosanoids could play a role in mediating ICAM-1 gene expression, and may explain, in part, the observed increase in ICAM-1 expression in 12/15-LOTG.13 The 12/15-LOTG also developed spontaneous fatty streak lesions on a chow diet, suggesting that ICAM-1 contributes to atherogenesis. Increased monocyte adhesion to endothelium in 12/15-LOTG was attributable in large part to elevations in ICAM-1 expression on the endothelium.13 Although we found a significant increase in ICAM-1 expression in 12/15-LOTG that showed increased atherosclerosis, we did not determine that ICAM-1 regulated atherosclerosis development in these mice.
The role of ICAM-1 in mediating atherosclerosis development is unclear. ICAM-1 is increased in atherosclerosis, but is not exclusively localized to lesion-prone sites, whereas VCAM-1 appears to be primarily expressed at aortic lesion areas in mice.16 Several groups have reported that atherosclerosis-susceptible mice that were deficient in ICAM-1 showed decreased atherosclerosis development compared with apoE-deficient mice alone.1820 However, in a separate mouse study, Cybulsky and colleagues concluded that VCAM-1, rather than ICAM-1, was important for atherogenesis.21 We have previously reported that ICAM-1 is critical for monocyte adhesion to diabetic endothelium.24 The role of ICAM-1 in diabetic atherosclerosis is completely unknown. Typically, mouse models of Type 2 diabetes do not develop atherosclerosis. This is a limiting factor in mouse studies of Type 2 diabetes, and the fact that we did not observe increased ICAM-1 expression in the diabetic db/db mouse does not indicate that ICAM-1 is not important in atherosclerosis. Several reports have indicated that ICAM-1 levels are increased in patients with Type 2 diabetes, and these patients also have increased incidence of atherosclerosis.46,47 However, studies to delineate the role of ICAM-1 in diabetic atherosclerosis are currently underway in our laboratory.
We found that 12SHETE also regulates MCP-1 mRNA levels in EC. MCP-1 is an important regulator of monocyte recruitment to endothelium.48,49 Previous studies have shown that MCP-1 and interleukin-8 are regulated, in part, by oxidized phospholipid components of minimally modified low-density lipoprotein cholesterol.50 NF-
B has been shown to regulate expression of MCP-1 in EC.51 For purposes of this present study, we focused our efforts entirely on ICAM-1. However, it is clear that MCP-1 is also important for monocyte adhesion to endothelium, and expression of endothelial MCP-1 is also regulated by 12SHETE.
RhoA and PKC
appear to work synergistically to activate NF-
B in EC in response to 12/15-LO products. RhoA and PKC
activity were both significantly increased in EC isolated from 12/15-LO TG mice and are colocalized within a protein complex in 12SHETE-stimulated EC (Figure 4). There are reports that RhoA and PKC
work in concert in EC to regulate each others activity.3335 Furthermore, inhibition of RhoA has been shown to block PKC
translocation and activation in EC.33 Our data indicate that RhoA is upstream of PKC
, and activation of RhoA stimulates PKC
mobilization to the endothelial plasma membrane. Our data also suggest that the RhoA-dependent translocation of PKC
by 12SHETE may further enhance RhoA activity, stimulating a positive feedback loop to amplify the signaling through this pathway.
We hypothesize that 12SHETE is secreted by activated endothelium and monocytes and binds to a specific cell surface receptor on vascular cells, thus amplifying and prolonging the inflammatory response.15,5254 A specific 12SHETE receptor has not yet been identified; however, Szekeres and colleagues have reported the presence of putative low affinity and high affinity 12SHETE receptors in carcinoma cells.55 We hypothesize that the 12SHETE receptor may be a GPCR. Our criteria for making this assumption are based on (1) the activation of the small GTPase RhoA by 1 2SHETE, (2) the low nanomolar concentrations of 12SHETE required to activate endothelial signaling pathways, (3) the rapid activation of the RhoA/PKC
/NF-
B signaling pathway in EC by 12SHETE, and (4) the reduction in 12SHETE-mediated NF-
B activation in response to blocking the G proteins G12 and G13 in endothelium. Studies to identify this receptor(s) are currently ongoing in the laboratory.
In summary, the 12/15-LO pathway generates proinflammatory eicosanoids that signal within endothelium to promote monocyte:endothelial adhesion. Specifically, we have found that 12/15-LO transgenic EC had increased RhoA activity and membrane-bound PKC
, resulting in NF-
B activation and increased ICAM-1 production. We also found that the primary eicosanoid product of the 12/15-LO pathway that regulated this signaling cascade was 12SHETE. Thus, 12SHETE acts as a second messenger in endothelium to promote monocyte:EC interactions, thereby creating a proinflammatory environment important for atherogenesis.
| Acknowledgments |
|---|
B construct. Received April 11, 2005; accepted August 29, 2005.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
R. A. Gubitosi-Klug, R. Talahalli, Y. Du, J. L. Nadler, and T. S. Kern 5-Lipoxygenase, but Not 12/15-Lipoxygenase, Contributes to Degeneration of Retinal Capillaries in a Mouse Model of Diabetic Retinopathy Diabetes, May 1, 2008; 57(5): 1387 - 1393. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Harkewicz, K. Hartvigsen, F. Almazan, E. A. Dennis, J. L. Witztum, and Y. I. Miller Cholesteryl Ester Hydroperoxides Are Biologically Active Components of Minimally Oxidized Low Density Lipoprotein J. Biol. Chem., April 18, 2008; 283(16): 10241 - 10251. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. W. Tew, E. L. Lorimer, T. J. Berg, H. Zhi, R. Li, and C. L. Williams SmgGDS Regulates Cell Proliferation, Migration, and NF-{kappa}B Transcriptional Activity in Non-small Cell Lung Carcinoma J. Biol. Chem., January 11, 2008; 283(2): 963 - 976. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. M. Nikolic, M. C. Gong, J. Turk, and S. R. Post Class A Scavenger Receptor-mediated Macrophage Adhesion Requires Coupling of Calcium-independent Phospholipase A2 and 12/15-Lipoxygenase to Rac and Cdc42 Activation J. Biol. Chem., November 16, 2007; 282(46): 33405 - 33411. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. A. Stapleton, Adam. G. Goodwill, M. E. James, and J. C. Frisbee Altered mechanisms of endothelium-dependent dilation in skeletal muscle arterioles with genetic hypercholesterolemia Am J Physiol Regulatory Integrative Comp Physiol, September 1, 2007; 293(3): R1110 - R1119. [Abstract] [Full Text] [PDF] |
||||
![]() |
X.-Y. Zhu, E. Daghini, A. R. Chade, C. Napoli, E. L. Ritman, A. Lerman, and L. O. Lerman Simvastatin Prevents Coronary Microvascular Remodeling in Renovascular Hypertensive Pigs J. Am. Soc. Nephrol., April 1, 2007; 18(4): 1209 - 1217. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. T. Bolick, A. M. Whetzel, M. Skaflen, T. L. Deem, J. Lee, and C. C. Hedrick Absence of the G Protein-Coupled Receptor G2A in Mice Promotes Monocyte/Endothelial Interactions in Aorta Circ. Res., March 2, 2007; 100(4): 572 - 580. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Kimura, H. Tomura, C. Mogi, A. Kuwabara, A. Damirin, T. Ishizuka, A. Sekiguchi, M. Ishiwara, D.-S. Im, K. Sato, et al. Role of Scavenger Receptor Class B Type I and Sphingosine 1-Phosphate Receptors in High Density Lipoprotein-induced Inhibition of Adhesion Molecule Expression in Endothelial Cells J. Biol. Chem., December 8, 2006; 281(49): 37457 - 37467. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Pang, A. Lin, M. Holford, B. E. Enerson, B. Lu, M. P. Lawton, E. Floyd, and H. Zhao Pathway analysis using random forests classification and regression Bioinformatics, August 15, 2006; 22(16): 2028 - 2036. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. D. Funk Lipoxygenase Pathways as Mediators of Early Inflammatory Events in Atherosclerosis. Arterioscler. Thromb. Vasc. Biol., June 1, 2006; 26(6): 1204 - 1206. [Full Text] [PDF] |
||||
![]() |
D. T. Bolick, S. Srinivasan, A. Whetzel, L. C. Fuller, and C. C. Hedrick 12/15 Lipoxygenase Mediates Monocyte Adhesion to Aortic Endothelium in Apolipoprotein E-Deficient Mice Through Activation of RhoA and NF-{kappa}B Arterioscler. Thromb. Vasc. Biol., June 1, 2006; 26(6): 1260 - 1266. [Abstract] [Full Text] [PDF] |