Inhibition of hepatitis C virus replication
by peroxidation of arachidonate and
restoration by vitamin E
Hua Huang, Yan Chen, and Jin Ye*
Department of Molecular Genetics, University of Texas Southwestern Medical Center, Dallas, TX 75390-9046
Edited by Joseph L. Goldstein, University of Texas Southwestern Medical Center, Dallas, TX, and approved October 8, 2007 (received for review
September 5, 2007)
Hepatitis C virus (HCV) is a single-stranded positive-sense RNA
virus of the Flaviviridae family. HCV-infected hepatocytes are
known to produce reactive oxygen species (ROS), which initiate
lipid peroxidation, a reaction that converts polyunsaturated fatty
acids, such as arachidonate, into reactive carbonyls that inactivate
proteins. To study the effect of lipid peroxidation on HCV replica-
tion, we administered arachidonate to Huh7 cells that harbor an
HCV replicon (Huh7-K2040 cells). After incubation in medium sup-
plemented with arachidonate but deprived of lipid-soluble anti-
oxidants, the cellular amount of malondialdehyde (MDA), a prod-
uct of lipid peroxidation, increased markedly in Huh7-K2040 cells
but not in parental Huh7 cells that do not harbor an HCV replicon.
This increase was followed by a sharp reduction (>95%) in HCV
with vitamin E, a lipid-soluble antioxidant. After prolonged incu-
bation of Huh7-K2040 cells with arachidonate in the absence of
lipid-soluble antioxidants, the amount of MDA decreased after the
reduction in the amount of HCV RNA. Thus, in the presence of
arachidonate and in the absence of lipid-soluble antioxidants, HCV
replication induces lipid peroxidation that reduces the amount of
HCV RNA. Our results provide a mechanism for the previous
observation that polyunsaturated fatty acids inhibit HCV replica-
2566], and they suggest that these agents may be effective in
inhibiting HCV replication in vivo.
lipid peroxidation ? polyunsaturated fatty acids
persistently with HCV, and these individuals account for most
cases of chronic liver disease and hepatic failure (1). Current
IFN-based therapies are ineffective in many patients infected by
HCV, and they are plagued with adverse effects (2), underscor-
ing the need for new therapeutic strategies.
HCV is a single-stranded positive-sense RNA virus of the
Flaviviridae family (3). The 9.6-kb HCV genome encodes a
single polyprotein that is posttranslationally processed into at
least 10 structural and nonstructural (NS) proteins. Among the
NS proteins, NS3, NS4A, NS4B, NS5A, and NS5B are sufficient
to support replication of the HCV RNA, as illustrated by the
replication capacity of HCV subgenomic replicons (4). An HCV
subgenomic replicon is an engineered HCV RNA expressing a
selectable marker gene, neo, in place of the structural coding
region. A heterologous viral internal ribosomal entry site is
inserted after the neomycin resistance cassette to direct the
translation of the viral nonstructural proteins NS3–5B. When
and selected with G418, a cell line can be established in which
HCV RNA is constantly replicated (4).
A hallmark of HCV infection is the generation of reactive
oxygen species (ROS) and oxidative stress (5). ROS generation
has been reported in HCV infected patients (6–8), Huh7 cells
infected by an infectious clone of HCV (9), and Huh7 cells that
epatitis C virus (HCV) exacts a heavy toll on public health.
Approximately 170 million people worldwide are infected
contain HCV replicons (10, 11). ROS that are located in
membranes are eradicated by lipid-soluble antioxidants, such as
vitamin E (12). When production of ROS exceeds the scavenger
activity of these lipid-soluble antioxidants, ROS in membranes
converts polyunsaturated fatty acids into toxic reactive carbon-
yls, which inactivate proteins by formation of covalent protein
conjugates (13). Interestingly, Kapadia and Chisari (14) recently
reported that polyunsaturated fatty acids inhibit HCV replica-
tion, but the mechanism of this inhibition is not known.
In the current studies, we used the HCV replicon system to
examine the impact of lipid peroxidation on HCV replication.
Using Huh7 cells harboring an HCV replicon, we observed that
exposure to polyunsaturated fatty acids in the absence of lipid-
soluble antioxidants resulted in an acute increase in the amount
of malondialdehyde (MDA), a product of lipid peroxidation.
This increase was followed by a reduction in HCV RNA.
Inasmuch as the generation of MDA relies on ROS produced by
HCV replication, prolonged incubation of cells with polyunsat-
urated fatty acids in the absence of lipid-soluble antioxidants led
to a decline in the amount of MDA after the reduction in HCV
RNA. This decline in lipid peroxidation allows inhibition of
HCV replication without generation of cellular toxicity.
HCV replication generates ROS (10, 11), which have the po-
tential to trigger lipid peroxidation. To monitor the production
of reactive carbonyls generated by lipid peroxidation, we mea-
sured the amount of MDA produced by Huh7-K2040 cells, a line
of Huh7 cells that harbor an HCV replicon (15). MDA is a
product of lipid peroxidation and is commonly used as an index
of this process (16). When incubated in normal culture medium
supplemented with FCS, the Huh7-K2040 cells produced very
little MDA (Fig. 1, lane 1). Addition of arachidonate, a substrate
for lipid peroxidation, raised the amount of MDA by ?8-fold in
to be inhibited by lipid-soluble antioxidants present in FCS, we
next measured the amount of MDA produced in cells cultured
in medium containing FCS that had been depleted of all
lipid-soluble materials by extraction with organic solvents (de-
lipidated FCS) (17). This treatment elicited only a slight increase
in the mount of MDA (Fig. 1, lane 3). However, under these
conditions the addition of arachidonate led to a 48-fold increase
Author contributions: H.H. and J.Y. designed research; H.H. and Y.C. performed research;
H.H., Y.C., and J.Y. analyzed data; and J.Y. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
Freely available online through the PNAS open access option.
*To whom correspondence should be addressed. E-mail: email@example.com.
This article contains supporting information online at www.pnas.org/cgi/content/full/
© 2007 by The National Academy of Sciences of the USA
November 20, 2007 ?
vol. 104 ?
in the amount of MDA (Fig. 1, lane 4). This massive increase was
reversed by addition of vitamin E, a lipid-soluble antioxidant
(Fig. 1, lane 5). Parental Huh7 cells that do not harbor an HCV
replicon produced only modest amounts of MDA when incu-
bated with arachidonate in delipidated FCS (Fig. 1, lane 9). This
result suggests that ROS generated by HCV replication lead to
the generation of MDA (and possibly other reactive carbonyls)
in the presence of a polyunsaturated fatty acid, such as arachid-
onate, when lipid-soluble antioxidants are absent.
The results from Fig. 1, together with an observation that
HCV replication is inhibited by polyunsaturated fatty acids (14),
raise the possibility that products of lipid peroxidation repress
HCV replication. If this is the case, then removal of lipid-soluble
antioxidants from the culture medium should inhibit HCV
replication. To test this hypothesis, we incubated Huh7-K2040
cells in medium supplemented with delipidated FCS that is
resulted in a 70% decrease in the amount of HCV RNA (Fig. 2,
lane 2). Restoration of vitamin E to the culture medium raised
the amount of HCV RNA to the level that was observed in
control cells incubated in medium supplemented with FCS (Fig.
2, lane 3). Addition of vitamin K or coenzyme Q4, two other
lipid-soluble antioxidants (18, 19), also restored replication of
HCV (Fig. 2, lanes 4 and 5). Addition of oleate, a fatty acid that
is abundant in FCS, did not rescue the replication of HCV (Fig.
2, lane 6). Interestingly, addition of a water-soluble antioxidant
vitamin C also did not restore HCV replication (Fig. 2, lane 7),
suggesting a specific requirement of lipid-soluble antioxidants
for HCV replication. Although HCV replication was reduced in
delipidated FCS, the Huh7-K2040 cells did not appear to be
harmed by this treatment, because they continued to grow at the
same rate as observed in the cells cultured in FCS (data not
We next sought to determine whether addition of polyunsat-
in delipidated FCS. In medium containing FCS, the addition of
polyunsaturated fatty acids such as linoleate, linolenate, and
arachidonate led to a 50–70% decrease in the amount of HCV
RNA (Fig. 3A, lanes 3–5). The inhibitory effect was enhanced
when the cells were cultured in medium containing delipidated
FCS (Fig. 3A, lanes 8–10). The combination of delipidated FCS
and arachidonate reduced the amount of HCV RNA by 95%
compared with control cells cultured in FCS without added fatty
acids (Fig. 3A, lane 10). Arachidonate was more potent than
linoleate and linolenate in inhibiting HCV replication, probably
because it contains more double bonds, thus allowing it to
produce more reactive carbonyls after lipid peroxidation (13).
Unlike the results with polyunsaturated fatty acids, treatment of
Huh7-K2040 cells with oleate, a monounsaturated fatty acid that
is not subject to lipid peroxidation (13), did not significantly
affect replication of HCV in cells that were cultured in medium
supplemented with either FCS (Fig. 3A, lane 2) or delipidated
FCS (Fig. 3A, lane 7).
Fig. 3B shows an experiment in which we analyzed the amount
of arachidonate required to inhibit HCV replication. When
Huh7-K2040 cells were cultured in medium supplemented with
FCS, the IC50for arachidonate was ?60 ?M (Fig. 3B). The IC50
was decreased to 10 ?M when these cells were incubated in
medium supplemented with delipidated FCS (Fig. 3B).
If polyunsaturated fatty acids inhibit HCV replication through
lipid peroxidation, then this effect should be countered by
lipid-soluble antioxidants. To test this hypothesis, Huh7-K2040
cells cultured in medium containing delipidated FCS with or
amounts of vitamin E. As expected, when cells were cultured in
the absence of arachidonate, vitamin E restored the amount of
HCV RNA to the level observed in cells cultured in FCS, which
is set at 1 in Fig. 4A. In the presence of arachidonate, vitamin E
increased the amount of HCV RNA from 5% to 190% of that
in control cells cultured in FCS (Fig. 4A). The concentration of
vitamin E that produced a half-maximal effect in both culture
conditions was 40 nM (Fig. 4A). Fig. 4B shows that addition of
vitamin E to cells cultured in FCS did not further increase the
amount of HCV RNA, indicating that vitamin E became limiting
only in delipidated FCS but not in FCS.
medium containing delipidated FCS and arachidonate. On day 0, Huh7 (blue)
or Huh7-K2040 (red) cells were set up at 7 ? 105cells per 60-mm dish. On day
1, cells were switched to medium supplemented with 10% of FCS or delipi-
that is conjugated with BSA as indicated. Sixteen hours later, on day 2, cells
in Materials and Methods. Values (mean ? SD) from three independent
experiments are presented.
Increased production of MDA in Huh7-K2040 cells incubated in
of indicated serum and treated with 0.3 ?M indicated lipids or vitamins.
Seventy-two hours later, on day 5, cells were harvested, and the amount of
HCV RNA was determined by quantitative real-time PCR analysis. Values
in control cells cultured in medium supplemented with FCS, which is set at 1.
Lipid-soluble antioxidants are required for efficient HCV replication
Huang et al.
November 20, 2007 ?
vol. 104 ?
no. 47 ?
The inhibition of HCV replication by replication-dependent
lipid peroxidation creates a paradox whereby HCV should
inhibit its own replication, which in turn should cause a drop in
reactive carbonyls produced by lipid peroxidation. Eventually, a
new steady state should be reached in which replication of HCV
continues at a low level that does not produce enough ROS to
inhibit HCV replication. To test this hypothesis, we performed
chronic experiments in which we measured the amount of MDA
in Huh7 and in replicon-bearing Huh7-K2040 cells that were
cultured in medium containing delipidated FCS supplemented
accumulated during 72 h of treatment (Fig. 5A, filled triangles).
In Huh7-K2040 cells, MDA increased markedly in the first 12–24
h and then decreased to a level that was even lower than that in
Huh7 cells after 48 h of treatment (Fig. 5A, open triangles). To
determine whether the decrease in the amount of MDA corre-
of MDA and HCV RNA in Huh7-K2040 cells that were cultured
in medium containing delipidated FCS supplemented with ara-
chidonate for various amounts of time. As shown in Fig. 5B,
MDA production reached a maximum after Huh7-K2040 cells
were switched into this medium for 12 h, at which time HCV
RNA was decreased by only 40%. After that, the decrease in the
amount of MDA paralleled the reduction in HCV RNA (Fig.
5B). After 72 h, the amount of MDA and HCV RNA were both
reduced by ?90% (Fig. 5B). Incubation of Huh7-K2040 cells in
delipidated FCS supplemented with arachidonate did not gen-
erate cellular toxicity, because the rate of cell growth measured
by the increase in total cellular protein in this culture condition
was not different from that in cells cultured under normal
condition in medium supplemented with FCS (Fig. 5C). The lack
of cellular toxicity is most likely attributable to the transient
Huh7-K2040 cells cultured in delipidated FCS than in FCS. On day 0, Huh7-
switched to medium supplemented with 10% of FCS (red) or delipidated FCS
two hours later, on day 5, cells were harvested, and the amount of HCV RNA
was determined by quantitative real-time PCR analysis. Values (mean ? SD of
three independent experiments) are presented relative to the control cells
which is set at 1. The scale for relative HCV RNA in cells cultured in FCS is
different from that in cells cultured in delipidated FCS because incubation in
delipidated FCS alone without any treatment of fatty acids resulted in a 60%
reduction in the amount of HCV RNA. (B) On day 2, cells were switched to
medium supplemented with 10% of FCS (red) or delipidated FCS (blue) and
hours later, on day 5, cells were harvested, and the amount of HCV RNA was
determined by quantitative real-time PCR analysis. For cells incubated with
either FCS or delipidated FCS, values (mean ? SD of three independent
experiments) are plotted relative to that in control cells that were not treated
with arachidonate, which is set at 1.
Polyunsaturated fatty acids inhibit HCV replication more potently in
arachidonate. On day 0, Huh7-K2040 cells were set up at 4 ? 105cells per
60-mm dish. (A) On day 2, except for control cells that were continued to be
maintained in medium containing 10% FCS, cells were switched to medium
supplemented with 10% delipidated FCS in the absence (blue) or presence
(red) of 0.1 mM BSA-conjugated arachidonate and treated with indicated
amount of vitamin E. Seventy-two hours later, on day 5, cells were harvested,
and the amount of HCV RNA was determined by quantitative real-time PCR
analysis. Values are plotted relative to the control cells cultured in medium
supplemented with FCS without any treatment, which is set at 1. Results from
dot. (B) On day 2, cells were switched to medium supplemented with 10% of
onate. These cells were then treated with or without 0.3 ?M vitamin E.
Seventy-two hours later, on day 5, cells were harvested, and the amount of
HCV RNA was determined and presented as described in A. Results (mean ?
SD) from three independent experiments are shown.
Vitamin E restores HCV replication in Huh7-K2040 cells treated with
www.pnas.org?cgi?doi?10.1073?pnas.0708423104 Huang et al.
rather than persistent activation of lipid peroxidation (Fig. 5 A
and B). Long-term treatment with reactive carbonyls generated
by lipid peroxidation is known to be toxic to cells (13). Similar
results shown in Fig. 5 were observed in one other independent
experiment and were shown in supporting information (SI)
Studies have reported that HCV replication produces ROS (10,
11), which are normally detoxified by lipid-soluble antioxidants,
such as vitamin E (12). In the current study, we show that HCV
replication is inhibited by lipid peroxidation that can be blocked
by lipid-soluble antioxidants such as vitamin E (Figs. 2–4). When
cultured in medium deprived of lipid-soluble antioxidants and
containing arachidonate, Huh7-K2040 cells exhibited markedly
elevated lipid peroxidation as measured by the production of
MDA (Fig. 1). The high rate of lipid peroxidation led to reduced
HCV replication (Figs. 2–4). Consequently, the rate of MDA
production declined (Fig. 5 A and B). Eventually, a new steady
state was reached in which HCV replicated at a low level that did
not produce sufficient ROS to further inhibit HCV replication.
Consistent with this notion, addition of arachidonate in the
absence of lipid-soluble antioxidants to Huh7-K2040 cells inhib-
ited HCV replication by ?95% during the first three days of the
treatment (Fig. 5B), but longer treatment (up to 6 days) did not
further change the amount of HCV RNA (data not shown).
We also examined whether polyunsaturated fatty acids in the
absence of lipid-soluble antioxidants also inhibited HCV repli-
cation in Huh7-derived cells infected by the JFH1 strain of HCV
(data not shown). Unfortunately, such treatment was toxic to
these cells, which made the results difficult to be interpreted.
Exactly how lipid peroxidation inhibits HCV replication re-
mains unclear. Several products of lipid peroxidation such as
MDA and 4-hydroxy-2-nonenal are known to inactivate proteins
by formation of covalent protein conjugates (13). Thus, these
products of lipid peroxidation may inactivate HCV NS proteins
or host proteins required for HCV replication. Moreover, MDA
is able to bind RNA covalently (20), a reaction that might allow
MDA to inactive HCV RNA directly.
The current study explains the previous observation that
that harbor HCV replicons (14). If such inhibition occurs in
livers of infected patients, the possibility exists that the balance
between lipid peroxidation and HCV replication may help to
determine the level of HCV RNA. It will be interesting to
examine whether HCV viral load is inversely correlated with the
amount of serum MDA in HCV-infected patients. This data also
raise the possibility that dietary supplement or pharmacological
preparation of polyunsaturated fatty acids (21) may help to
suppress HCV replication in patients. Based on our in vitro data,
peroxidation of polyunsaturated fatty acids only occurs in cells
in which HCV is actively replicating (Figs. 1 and 5B). If this
selectivity also occurs in vivo, then such treatment may be able
to inhibit HCV replication without intolerable toxicity.
Materials and Methods
Materials. We obtained vitamin E (?-tocopherol), coenzyme Q4,
vitamin K, oleate, linoleate, linolenate, and arachidonate from
Sigma and defatted BSA from Roche Molecular Biochemicals.
Delipidated FCS was prepared exactly as described in ref. 17.
Briefly, 500 ml of FCS was mixed with 400 ml of 1-butanol and
600 ml of isopropyl ether at room temperature for 20 min,
followed by a 20-min incubation on ice. After centrifugation, the
aqueous phase was reextracted with 200 ml of isopropyl ether,
recentrifuged, subjected to evaporation under a stream of ni-
trogen gas, lyophilized, reconstituted with 200 ml of distilled
water, and dialyzed against PBS. Multiple aliquots were stored
Cell Culture. Huh7 cells were maintained in medium A (Dulbec-
co’s modified Eagle medium with 4.5 g/liter glucose, 100 units/ml
penicillin, 100 ?g/ml streptomycin sulfate, and 10% FCS).
Huh7-K2040 cells are Huh7 cells that harbor a genotype 1b HCV
subgenomic replicon (15). They were maintained in medium A
supplemented with 200 ?g/ml G418. Both cells were maintained
in monolayer culture at 37°C in 5% CO2. Huh7-K2040 cells were
a gift from M. Gale (University of Washington, Seattle, WA).
Quantitative Real-Time PCR. The measurement of HCV or cellular
RNA was performed by real-time PCR analysis with a protocol
described in ref. 22. Briefly, RNA was extracted from cells
pooled from two 60-mm dishes, using the RNeasy Mini Kit
(Qiagen, Valencia, CA). First-strand cDNA was synthesized
from the DNA-free RNA, using TaqMan reverse-transcription
reagents (Applied Biosystems). Triplicate samples of first strand
cDNA were subjected to real-time PCR quantification, using
forward and reverse primers for the indicated RNA with human
36B4 as an invariant control. Relative amounts of mRNA were
calculated by using the comparative CTmethod.
Lipid Peroxidation Analysis. The extent of lipid peroxidation was
measured by the amount of MDA, which was assayed by the
up at 7 ? 105cells per 60-mm dish. On day 1, cells were switched to medium supplemented with 10% of FCS or 10% delipidated FCS containing 0.1 mM
BSA-conjugated arachidonate. After incubation in these medium for the indicated amount of time, cells were harvested. (A) The amount of MDA in Huh7 and
Huh7-K2040 cells cultured in medium containing delipidated FCS and arachidonate was quantified as described in Fig. 1. (B) The amount of MDA and HCV RNA
cultured in either medium was determined. (B and C) Values shown are relative to the amount in cells harvested on day 1 immediately before medium were
and were shown in SI Fig. 6.
Prolonged incubation of Huh7-K2040 cells with arachidonate in the absence of lipid-soluble antioxidants. On day 0, Huh7 or Huh7-K2040 cells were set
Huang et al.
November 20, 2007 ?
vol. 104 ?
no. 47 ?
thiobarbituric acid reactive substances (TBARS) kit (Zepto- Download full-text
Metrix). Cells pooled from two 60-mm dishes were resuspended
in 300 ?l of buffer A (1.5 mM KH2PO4, 8.1 mM Na2HPO4, 2.7
mM KCl, and 137 mM NaCl, pH 7.4) and homogenized by
passing through 22-gauge needles 10 times followed by sonica-
tion at 100% amplitude for 5 min in a microprocessor-controlled
ultrasonic water bath (Lab-Line Instruments). The amount of
MDA in 40 ?l of cell homogenates was measured by a spectro-
scopic method described in manufacturer’s protocol. The absor-
bance was read by a SAFIRE plate reader, using Xfluor4
software (Tecan) at 532 nm. The activity of lipid peroxidation
was expressed as the amount of MDA normalized by the amount
of cellular protein.
Measurement of Protein Concentration. Protein concentration was
measured by a BCA kit (Pierce) following the manufacturer’s
Preparation of BSA-Conjugated Fatty Acids. A 10 mM stock solution
of each fatty acid was prepared by diluting the free fatty acid in
ethanol and precipitating it with 0.25 M NaOH. The precipitated
sodium salt was then evaporated under nitrogen gas, reconsti-
tuted with 0.15 M NaCl, and stirred at room temperature for 10
min with defatted BSA [final concentration at 10% (wt/vol) in
0.15 M NaCl]. Each solution was stored in multiple aliquots at
?20°C and protected from light in tubes evacuated under
We thank Michael S. Brown and Joseph L. Goldstein for their constant
support and helpful comments; Michael Gale for his generous gift of
Huh7-K2040 cells; Saada Abdalla for excellent technical assistance; Lisa
Beatty, Marissa Hodgin, and Ijeoma Onwuneme for invaluable assis-
tance in tissue culture; and Jeff Cormier for real-time PCR analysis. This
work was supported by National Institutes of Health Grant HL-20948
and the Perot Family Foundation.
1. Chisari FV (2005) Nature 436:930–932.
2. Feld JJ, Hoofnagle JH (2005) Nature 436:967–972.
4. Lohmann V, Ko ¨rner F, Koch JO, Herian U, Theilmann L, Bartenschlager R
(1999) Science 285:110–113.
5. Seronello S, Sheikh MY, Choi J (2007) Free Rad Biol Med 43:869–882.
6. Kageyama F, Kobayashi Y, Kawasaki T, Toyokuni S, Uchida K, Nakamura H
(2000) Am J Gastroenterol 95:1041–1050.
7. Sumida Y, Nakashima T, Yoh T, Nakajima Y, Ishikawa H, Mitsuyoshi H,
Sakamoto Y, Okanoue T, Kashima K, Nakamura H, et al. (2000) J Hepatol
8. Shimoda R, Nagashima M, Sakamoto M, Yamaguchi N, Hirohashi S, Yokota
J, Kasai H (1994) Cancer Res 54:3171–3172.
9. Machida K, Cheng KTH, Lai CK, Jeng KS, Sung VMH, Lai MMC (2006)
J Virol 80:7199–7207.
10. Gong G, Waris G, Tanveer R, Siddiqui A (2001) Proc Natl Acad Sci USA
11. Qadri I, Iwahashi M, Capasso JM, Hopken MW, Flores S, Schaack J, Simon
FR (2004) Biochem J 378:919–928.
12. Wang X, Quinn PJ (1999) Prog Lipid Res 38:309–336.
13. Dalle - Donne I, Aldini G, Carini M, Colombo R, Rossi R, Milzani A (2006)
J Cell Mol Med 10:389–406.
14. Kapadia SB, Chisari FV (2005) Proc Natl Acad Sci USA 102:2561–2566.
15. Ye J, Wang C, Sumpter R, Jr, Brown MS, Goldstein JL, Gale M, Jr (2003) Proc
Natl Acad Sci USA 100:15865–15870.
16. Janero DR (1990) Free Radic Biol Med 9:515–540.
17. Hannah VC, Ou J, Luong A, Goldstein JL, Brown MS (2001) J Biol Chem
18. Canfield LM, Davy LA, Thomas GL (1985) Biochem Biophys Res Commun
19. Genova ML, Pich MM, Biondi A, Bernacchia A, Falasca A, Bovina C,
Formiggini G, Castelli GP, Lenaz G (2003) Exp Biol Med 228:506–513.
20. Sevilla CL, Mahle NH, Eliezer N, Uzieblo A, O’Hara SM, Nokubo M, Miller
R, Rouzer CA, Marnett LJ (1997) Chem Res Toxicol 10:172–180.
21. Bays H (2006) Am J Cardiol 98:71–76.
22. Liang G, Yang J, Horton JD, Hammer RE, Goldstein JL, Brown MS (2002)
J Biol Chem 277:9520–9528.
www.pnas.org?cgi?doi?10.1073?pnas.0708423104 Huang et al.