| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Atherosclerosis and Lipoproteins |
From the Clinical Diabetes Unit, Service of Endocrinology, Diabetes and Nutrition, University Hospital, 1211 Geneva 14, Switzerland.
Correspondence to Prof R.W. James, Clinical Diabetes Unit, Service of Endocrinology, Diabetes and Nutrition, University Hospital, 24, rue Micheli-du-Crest, 1211 Geneva 14, Switzerland. E-mail Richard.James{at}hcuge.ch
| Abstract |
|---|
|
|
|---|
Methods and Results HDL was oxidized with AAPH, copper ions, and hypochlorite. Secretion studies were performed using human paraoxonase-1transfected cells lines and primary rat hepatocytes. Stability studies were performed with recombinant paraoxonase. Conditioned medium had significantly reduced paraoxonase-1 when Cu or AAPH-oxidized HDL was the acceptor complex (P<0.01); reduction was dose-dependent on the degree of oxidation. Oxidized HDL had a reduced capacity to stabilize/improve activity of secreted paraoxonase-1. Reduced secretion could not be attributed to enzyme inactivation by lipoperoxides, reduced binding affinity of HDL, or oxidation of the lipid component alone. Hypochlorite oxidation of HDL did not modify HDL-mediated paraoxonase-1 release, but activity of HDL-associated paraoxonase-1 was particularly sensitive to such treatment.
Conclusions AAPH and copper, but not hypochlorite, oxidation of HDL compromises its ability to promote release of paraoxonase-1 and stabilize enzyme activity. HDL-associated paraoxonase-1 is highly sensitive to hypochlorite. Reducing paraoxonase-1 renders HDL susceptible to oxidation, which may compromise HDL function. It provides a novel example at the HDL level of the detrimental effects of oxidative stress, and underlines the need for further evaluation of the consequences of HDL oxidation.
Oxidation of HDL reduced its ability to promote paraoxonase-1 secretion and to improve the activity of the secreted enzyme. It was not related to modified affinity of HDL for the cell, or lipoperoxide-mediated enzyme inactivation. The degree of HDL oxidation was inversely correlated with the concentration of associated paraoxonase-1.
Key Words: HDL oxidative stress myeloperoxidase hepatocyte lactonase
| Introduction |
|---|
|
|
|---|
Oxidative stress is detrimental to blood lipid metabolism and accorded a major role in the atherosclerotic process.2 The impact is usually considered in terms of oxidative changes to low density lipoproteins (LDL). However, HDL function is also compromised by oxidative stress, as shown over a decade ago for HDL lipid oxidation3 and more recently with myeloperoxidase-mediated oxidation where hypochlorite is implicated.4 Moreover, in comparison to LDL, HDL has a weak complement of antioxidants such as vitamins.5 The lipoprotein can bolster its antioxidant capacity by acquiring peptides capable of such activity. In this context, an important role has been attributed to the enzyme paraoxonase-1 (PON1). Its ability to modulate atherosclerotic lesion development has been demonstrated in animal models and linked to its antioxidant activity.6,7 Conversely, HDL deficient in PON1 is transformed into a prooxidant, proinflammatory complex.8
We recently proposed a mechanism by which HDL can acquire its complement of PON1.9 Association of HDL with the hepatocyte allows transfer of PON1 from a cell membrane location to the lipoprotein. Factors that interfere with this process can influence the PON1 concentration of HDL.9 The present study indicates that the process is sensitive to the oxidative status of HDL. Oxidation of HDL decreases its ability to remove PON1 from cells, modifies its capacity to increase activity of and stabilize desorbed enzyme, and inactivates PON1. It suggests an additional mechanism by which oxidative stress can influence HDL function and impact on the atherosclerotic process.
| Methods |
|---|
|
|
|---|
Release of PON1 From Cells
Studies of PON1 secretion were based on a Chinese hamster ovary (CHO) model stably transfected with human PON1 (CHO-hPON1), which we had previously developed and described.9 Culture medium was harvested and analyzed for enzyme activity (arylesterase [ARE])15 and occasionally for PON1 protein mass (competitive ELISA).15 Lactonase activity was analyzed using a substrate recently developed.16 Studies were also performed using primary rat hepatocytes17 and HepG2 cells, the latter transiently transfected with the coding region of the human PON1 gene.9
Recombinant PON1
Recombinant PON1 (rePON1) variant G3C918 was expressed in E Coli, purified via a His tag and stored in Tris/detergent buffer (Tris 50 mmol/L, pH 7.8; CaCl2, 20 mmol/L; Tergitol, 0.05% [v/v]).
For more details of Methods, please refer to the online supplement at http://atvb.ahajournals.org.
| Results |
|---|
|
|
|---|
PON1 Limits Oxidation of HDL in a Dose Dependent Manner
The impact of variations in PON1 content of HDL on oxidation of the lipoprotein is shown in Figure 1. There was a decrease in the concentration of lipoperoxides that was dose-dependent on the presence of PON1. Lipoperoxides generated in HDL were negatively correlated with the activity of PON1 associated with the lipoprotein (r>0.90 for all time points, P<0.01). In addition, the rate of accumulation of lipoperoxides with time was reduced in HDL preparations with higher concentrations of PON1.
|
Oxidation of HDL Lowers PON1 Activity Released Into Conditioned Medium
HDL has a stimulatory effect on secretion of PON1 into culture medium.9 When HDL, oxidized either with AAPH or copper ions, was incubated with CHO-hPON1 cells, there was a significant reduction in PON1 ARE activity of conditioned medium (P<0.001) compared with nonoxidized HDL (Figure 2A). The reduction of PON1 activity in the culture medium was dose-dependent on the degree of oxidation of HDL as shown in Figure 2B; all activity levels were P<0.01 compared with HDLco (0 LPO concentration). Our previous studies9 revealed that HDL-stimulated release of PON1 was saturable at 10 to 20 µg/mL of HDL protein, corresponding to the concentrations used in Figure 2A and 2B. Concentrations of HDL in the extra vascular compartment bathing cells and tissues are higher.22 Figure 2C shows that even with concentrations up to 300 µg/mL, oxidized HDL demonstrated a significantly lower capacity to increase PON1 activity in culture medium. Comparable results were obtained over short (3 hours) and long (20 hours) periods of incubation of HDL with the cells. The relevance of these observations to PON1 release from hepatocytes was examined in cultures of HepG2 transiently transfected with human PON1 and primary rat hepatocytes. Although releasing a much lower amount of PON1, there remained significant differences in the capacity of nonoxidized and oxidized HDL to promote release of the enzyme (Figure 2D).
|
A possible explanation for the lower levels of PON1 activity in the presence of HDLox may be reduced mass of PON1. This was examined by ELISA quantification of PON1 released into medium in the presence of HDLco or HDLox. As shown in Figure 2E, there was a significant reduction of PON1 mass measured in HDLox medium (P<0.03), corresponding to the reduction in the activity of PON1 released into medium (P<0.001).
Analysis of Affinity Constants for HDL Binding and PON1 Release
Our previous studies9 had established that HDL promoted PON1 release in a high affinity, saturable manner. To examine differing affinities for HDLco and HDLox (oxidized with AAPH) as an explanation for reduction in PON1, enzyme activity was analyzed as a function of HDL concentration. As shown in Figure 3A, for equivalent HDL concentrations, PON1 activity released into conditioned medium was consistently lower with HDLox as the acceptor lipoprotein. However, when apparent affinity constants were calculated based on Lineweaver-Burke plots of released enzyme activity (Figure 3B), there were no significant differences for the two forms of the lipoprotein (HDLco, 2.69±1.1µg HDL protein/mL; HDLox, 1.13±0.31µg HDL protein/mL; n=3 for both; P=0.1). Comparable results were obtained when HDL oxidized with copper ions was analyzed (HDLox, 2.46±1.5 µg protein/mL, mean of 2 analyses).
|
A second series of studies examined the global affinity of the HDL complex for the PON1 secreting cells. Preliminary studies established that iodination of HDL did not modify its capacity to promote PON1 release (data not shown). Binding curves for the association of 125I-HDLco and 125I-HDLox with CHO-hPON1 cells are given in Figure 3C, which is representative of 3 independent determinations with 2 different preparations of HDL. The affinity constant (Figure 3D; 15.5±1.7µg HDL protein/mL; n=3) of HDLco was comparable to that reported in the literature.21,23 HDLox bound to the cells with a higher affinity (6.6±1.9µg HDL protein/mL, n=3; P<0.02).
Influence of HDL on PON1 Activity
Previous studies have established that the lipid and apolipoprotein components of HDL play distinct but complementary roles in promoting PON1 release.9,24,25 To determine whether the reduced release of PON1 could be attributed to the lipid component alone of HDL, a total lipid extract of the lipoprotein was made and used in secretion studies as protein-free vesicles. PON1 release into conditioned medium containing oxidized and nonoxidized total HDL lipid extracts is shown in Figure 4A (representative graph of 4 independent analyses). The dose-response curves were largely superimposable, suggesting that oxidation had not modified the release process mediated by HDL lipids. Of particular relevance are the equivalents levels of PON1 activity found with both nonoxidized and oxidized HDL lipids. The presence of oxidized lipids does not appear to have an inactivating effect on secreted PON1. The lipid extracts were, however, less efficient than the native HDL complex, which promoted release of PON1 into conditioned medium at lower concentration (Figure 4A).
|
The latter point was further analyzed in conjunction with the ability of HDL to improve PON1 activity.9,24,26 Serum-free HAMS medium was conditioned (16 hours) with CHO-hPON1 cells and harvested. HDLco and HDLox (50 µg/mL) were added to the conditioned medium and assayed immediately for activity and after incubation (3 hour, 37°C). At t0, equivalent activities were found in both incubation mixes, despite the presence of oxidized lipids with HDLox. After incubation there was a significant rise in activity in both preparations (Figure 4B; P<0.001), but activity rose to a lesser extent with HDLox (+12%) than with HDLco (+42%, P<0.001 v HDLox). Studies were repeated with preparations of recombinant PON1. HDL-free rePON1 preparations were diluted into HDLco or HDLox samples (50 µg/mL) and activities measured immediately and after incubation (3 hours, 37°C). Dilution of rePON1 into HDL caused an immediate increase in activity, which was equivalent for HDLco and HDLox (Figure 4C). Subsequent incubation was associated with a decrease in rePON1 activity in the absence of HDL, probably caused by dilution of the Tergitol detergent stabilizing purified rePON1. In contrast, there were further increases in activity (P<0.01) after 3 hours for both forms of HDL, although to a lesser extent with HDLox (+9%) than with HDLco (+47%, P<0.001 v HDLox).
A final analysis compared the influence of HDLco and HDLox on PON1 with respect to its lactonase activity. Lactonase activity in CHO-hPON1 conditioned medium was significantly lower when oxidized HDL at 50 µg/mL (13.8% v HDLco, P<0.001) or 300 µg/mL (14.6%, P<0.001) (supplemental Figure IIIA) was the acceptor complex. When HDL was added to serum-free medium previously conditioned with the same cell line, there was an increase in lactonase activity for both HDLco (+72.8±8.8%) and HDLox (+48.0±4.8%), the increase being significantly greater (P<0.0001) for nonoxidized HDL (supplemental Figure IIIB).
Impact of Hypochlorite Treatment on HDL-PON1 Association
Hypochlorite is an oxidising agent produced by the activity of myeloperoxidase in vivo. Pretreatment of PON1-free HDL with hypochlorite (HDL-OCl) before incubation with CHO-hPON1 did not cause a significant decrease in the level of enzyme activity secreted by CHO-hPON1 cells (Figure 5A). In addition, HDL-OCl showed similar kinetics to control HDL with respect to stimulation of PON1 release from cells (Figure 5B). Thus, hypochlorite treatment of HDL did not appear to influence its ability to promote PON1 release and bind the enzyme. However, when HDL isolated from serum and containing active PON1 was treated with hypochlorite, a different picture emerged. As shown in Figure 5C, there was a dose-dependent decrease in PON1 activity on treatment with hypochlorite, with highest concentrations causing complete inactivation of the enzyme. This was paralleled by a dose-dependent decrease in monomeric PON1 as analyzed by Western blotting (Figure 5C and 5D). Hypochlorite-induced changes to PON1 activity and peptide occurred at concentrations lower than those that led to apoAI aggregation (supplemental Figure IB).
|
| Discussion |
|---|
|
|
|---|
Numerous studies have shown an inverse correlation between the activity of HDL-associated PON1 and oxidation of the lipoprotein13,31,32 as confirmed in this study. PON1 deficiency can thus have pathophysiological consequences for HDL. Our data indicate that several factors may contribute to reduced PON1 activity in oxidized HDL. One is a reduced capacity to stimulate PON1 release from the cell. It does not result from altered affinity of HDLox for the cell in that binding affinities measured as a function of PON1 activity9 did not differ between HDLco and HDLox. The latter was confirmed by complementary studies with radiolabeled HDL to measure affinity of the HDL complex, where HDLox tended to have a higher affinity than HDLco for the cell surface. This has been previously reported for HDL binding.21 The reasons for the reduced ability to remove PON1 from the cell are not clear, although it does mirror the reduced capacity to efflux cholesterol to HDLox.3,27 Modifications to the peptide components of HDL do not appear to be a major determinant as suggested by the studies with HDL-OCl.
The second factor appears to be a reduced capacity of HDL to further activate enzyme activity (an increase in PON1 specific activity). A positive impact on PON1 activity is a feature of the native HDL complex and has been independently reported by several groups.9,2426 The mechanism is presently unclear, although correct orientation of the active site into an optimal configuration may be a feature. Irrespective of the mechanism, there is clear evidence in the present report from 2 distinct preparations of PON1 that HDLox is deficient in its ability to further activate the enzyme.
A third possibility is that, once associated with HDL, PON1 may be particularly susceptible to inhibition by hypochlorite. Hypochlorite, which in vivo is derived from myeloperoxidase activity,4 was used in the present study as it principally targets peptide components. We could find no strong indication of lipid oxidation in treated HDL, as noted previously.21 One notable feature, however, was the extreme sensitivity of PON1 activity to hypochlorite treatment, despite a large excess of other peptides within HDL. Inhibition of PON1 activity was observed at HOCl:HDL protein ratios much lower than those previously used to study the effects of the oxidant on HDL function.21 A major effect on an HDL function (reverse cholesterol transport) provoked by an apparently mild, hypochlorite treatment regime has previously been reported.21 Myeloperoxidase is suggested to be one of the principal oxidizing agents giving rise to dysfunctional HDL in vivo.4 It exerts its effect on HDL predominantly within the artery wall14 where any buffering, antioxidant capacity of serum protecting PON1 would be diminished. In this scenario, ambient serum activities of PON1 would not reflect the effective activity or efficiency of the enzyme within the artery wall.
Finally, reduced, secreted PON1 activity in the presence of HDLox could also arise from inactivation of the enzyme by lipid peroxides.33 Our data suggest that it is not a major cause of reduced PON1 associated with HDL in the present studies. Firstly, there was an accompanying reduction of PON1 protein. Secondly, addition of HDLox to presecreted or recombinant PON1 did not decrease activity. Finally, protein-free oxidized HDL lipids did not show a reduction in PON1 activity in conditioned medium in comparison to protein-free nonoxidized lipids.
Phenylacetate (arylesterase activity) and paraoxon (paraoxonase activity) are the two substrates most commonly used to monitor PON1 activity. The present report focused on phenylacetate hydrolytic activity as it was recently shown to be a more accurate marker for the antioxidant activity of PON1 than is paraoxon.28 However, the antioxidant activity has been most tightly linked to lactonase activity of PON1.26,28 To ensure that our observations also apply to the latter activity, it was analyzed using a recently developed substrate.16,26 HDL oxidation was shown to affect lactonase activity in the same manner as it affected arylesterase activity; a reduced capacity to promote secretion and to increase enzyme activity.
The present study was based largely on the use of a CHO-hPON1 cell model of PON1 secretion, which we validated in a previous report.9 Its notable advantage is to minimize potentially confounding influences that could arise from cosecretion of lipids and apo AI by hepatocytes. Confirmation was thus sought with hepatocytes that HDLox had a reduced ability to promote PON1 release from this cell type. This was shown to be the case using the human hepatocyte cell line, HepG2 transiently transfected with human PON1, and primary hepatocytes obtained from rat.
In conclusion, we demonstrate several effects of HDL oxidation on the PON1 content and activity of the lipoprotein. The clinical consequences of the observations await further investigation. However, it has been established by a number of independent studies that states of increased oxidative stress (diabetes, cardiovascular disease, smoking) are associated with reduced serum PON1.3437 It is suggested to reflect in part PON1 inactivation by lipid peroxides. The present study indicates that HDL oxidation has a more fundamental impact on PON1 metabolism. It may have multiple consequences, as other HDL functions can be compromised by oxidation. The data underline the pathophysiological implications of HDL oxidation, which merits further investigation in the context of HDL as a primary therapeutic target and growing awareness of dysfunctional HDL.
| Acknowledgments |
|---|
Sources of Funding
The study was supported by a grant from the Swiss National Research Foundation (to R.W.J.).
Disclosures
None.
| Footnotes |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
P. Jaichander, K. Selvarajan, M. Garelnabi, and S. Parthasarathy Induction of paraoxonase 1 and apolipoprotein A-I gene expression by aspirin J. Lipid Res., October 1, 2008; 49(10): 2142 - 2148. [Abstract] [Full Text] [PDF] |
||||
![]() |
X. Moren, S. Deakin, M.-L. Liu, M.-R. Taskinen, and R. W. James HDL subfraction distribution of paraoxonase-1 and its relevance to enzyme activity and resistance to oxidative stress J. Lipid Res., June 1, 2008; 49(6): 1246 - 1253. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Valiyaveettil, N. Kar, M. Z. Ashraf, T. V. Byzova, M. Febbraio, and E. A. Podrez Oxidized high-density lipoprotein inhibits platelet activation and aggregation via scavenger receptor BI Blood, February 15, 2008; 111(4): 1962 - 1971. [Abstract] [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. |