REPORTS
formation. Three aprr7 alleles each lengthen the period of clock-mediated leaf movement by 1.5 to 2 hours without affecting the
phase (Fig. 4, A and B; table S3). Two
aprr5 alleles shorten the period by 1.5 to 2
hours, again without altering the phase
(Fig. 4, C and D; table S3). Similarly, two
appr3 alleles shorten the period but do not
affect phase (Fig. 4, E and F; table S3). In
contrast, one aprr9 allele confers a wildtype period but a phase that lags by 4 to 5
hours (Fig. 4, G and H; table S3). Each
allele affects either the period or phase, but
not both, which is consistent with our observations that the period and phase are not
correlated among the accessions; together,
these observations show that the period and
phase are under different genetic controls.
We present evidence that the period of
the circadian clock in Arabidopsis displays
a great deal of environmentally dependent
natural variation. Our observation of a latitudinal cline in the period of the Arabidopsis circadian clock is consistent with a primary role of the circadian clock in the
synchronization of an organism with its
periodic surroundings. We also demonstrate transgressive segregation of clock parameters in hybrids derived from two commonly studied accessions with very similar
clock parameters, which would facilitate
the exploitation of new ecological niches or
competition in new environments (16 ).
Loci such as the APRR family may act as
primary sources of natural variation, allowing modest complementary positive and
negative effects to modulate the circadian
period and phase to enhance fitness in local environments.
17. J. H. Clarke, R. Mithen, J. K. M. Brown, C. Dean, Mol.
Gen. Genet. 248, 278 (1995).
18. K. Swarup et al., Plant J. 20, 67 (1999).
19. D. E. Somers, A. A. R. Webb, M. Pearson, S. A. Kay,
Development 125, 485 (1998).
20. M. J. Yanovsky, S. A. Kay, Nature 419, 308 (2002).
21. L. C. Roden, H.-R. Song, S. Jackson, K. Morris, I. A.
Carré, Proc. Natl. Acad. Sci. U.S.A. 99, 13313 (2002).
22. K. Shimomura et al., Genome Res. 11, 959 (2001).
23. J. O. Borevitz et al., Genetics 160, 683 (2002).
24. A. Matsushika, S. Makino, M. Kojima, T. Mizuno, Plant
Cell Physiol. 41, 1002 (2000).
25. A. J. Millar, I. A. Carré, C. A. Strayer, N.-H. Chua, S. A.
Kay, Science 267, 1161 (1995).
26. A. Matsushika, A. Imamura, T. Yamashino, T. Mizuno,
Plant Cell Physiol. 43, 833 (2002).
27. S. Makino, A. Matsushika, M. Kojima, T. Yamashino, T.
Mizuno, Plant Cell Physiol. 43, 58 (2002).
28. P. Más, D. Alabadı́, M. J. Yanovsky, T. Oyama, S. A.
Kay, Plant Cell 15, 223 (2003).
29. E. Sato, N. Nakamichi, T. Yamashino, T. Mizuno, Plant
Cell Physiol. 43, 1374 (2002).
30. R. M. Green, E. M. Tobin, Proc. Natl. Acad. Sci. U.S.A.
96, 4176 (1999).
31 January 2003; accepted 15 September 2003
Evidence for Ozone Formation in
Human Atherosclerotic Arteries
Paul Wentworth Jr.,1,7 Jorge Nieva,4 Cindy Takeuchi,2
Roger Galve,1 Anita D. Wentworth,1 Ralph B. Dilley,5
Giacomo A. DeLaria,5 Alan Saven,4 Bernard M. Babior,3
Kim D. Janda,1 Albert Eschenmoser,1,6 Richard A. Lerner1
Here, we report evidence for the production of ozone in human disease. Signature products unique to cholesterol ozonolysis are present within atherosclerotic tissue at the time of carotid endarterectomy, suggesting that ozone
production occurred during lesion development. Furthermore, advanced atherosclerotic plaques generate ozone when the leukocytes within the diseased
arteries are activated in vitro. The steroids produced by cholesterol ozonolysis
cause effects that are thought to be critical to the pathogenesis of atherosclerosis, including cytotoxicity, lipid-loading in macrophages, and deformation
of the apolipoprotein B-100 secondary structure. We propose the trivial designation “atheronals” for this previously unrecognized class of steroids.
References and Notes
1. C. R. McClung, P. A. Salomé, T. P. Michael, in The
Arabidopsis Book, C. R. Somerville, E. M. Meyerowitz,
Eds. (American Society of Plant Biologists, Rockville,
MD, 2002) (available online at www.aspb.org/
publications/arabidopsis/, doi 10.1199/tab.0044).
2. H. Daido, J. Theor. Biol. 217, 425 (2002).
3. R. M. Green, S. Tingay, Z.-Y. Wang, E. M. Tobin, Plant
Physiol. 129, 576 (2002).
4. L. M. Beaver et al., Proc. Natl. Acad. Sci. U.S.A. 99,
2134 (2002).
5. P. J. DeCoursey, J. K. Walker, S. A. Smith, J. Comp.
Physiol. A 186, 169 (2000).
6. Y. Ouyang, C. R. Andersson, T. Kondo, S. S. Golden,
C. H. Johnson, Proc. Natl. Acad. Sci. U.S.A. 95, 8660
(1998).
7. L. A. Sawyer et al., Science 278, 2117 (1997).
8. R. Costa, C. P. Kyriacou, Curr. Opin. Neurobiol. 8, 659
(1998).
9. J. D. Plautz et al., J. Biol. Rhythms 12, 204 (1997).
10. J. N. Maloof et al., Nature Genet. 29, 441 (2001).
11. B. Li, J.-I. Suzuki, T. Hara, Oecologia 115, 293 (1998).
12. N. T. Miyashita, A. Kawabe, H. Innan, Genetics 152,
1723 (1999).
13. C. S. Pittendrigh, T. Takamura, J. Biol. Rhythms 4, 217
(1989).
14. C. Weinig, J. R. Stinchcombe, J. Schmitt, Mol. Ecol. 12,
1153 (2003).
15. C. Lister, C. Dean, Plant J. 4, 745 (1993).
16. L. H. Rieseberg, M. A. Archer, R. K. Wayne, Heredity
83, 363 (1999).
31. We thank the Arabidopsis Biological Resource Center
(Ohio State University, Columbus, OH), the Sendai
Arabidopsis Stock Center (Sendai, Japan), and M.
Jakobsson (University of Lund, Sweden) for seed
stocks; M. L. Guerinot and K. Cottingham for helpful
discussions; J. Borevitz for assistance with tests for
epistasis; E. Tobin (University of California, Los Angeles) for providing the cca1-1 mutation introgressed
into the CO1 background; and S. Mishra for plant
maintenance. Supported by grants from NSF to J.R.E.
(MCB-0115103), M.A.M. (IBN-0130021), and C.R.M.
(MCB-9723482 and MCB-0091008). H.J.Y. and T.R.S.
were supported by Richter Undergraduate Research
Fellowships, and E.L.S. was supported by a Women in
Science Project Internship, administered through
Dartmouth College.
Supporting Online Material
www.sciencemag.org/cgi/content/full/302/5647/1049/
DC1
Materials and Methods
Figs. S1 to S3
Tables S1 to S3
References
Ozone is one of the most reactive chemicals
known. Until our studies showed that ozone
may be produced by the immune system as
part of its defense strategy (1–3), this highly
toxic oxygen allotrope had not been considered to be a product of biological systems.
We reported evidence that ozone is generated during the antibody-catalyzed wateroxidation pathway and that this powerful
oxidant could play a role in inflammation
Department of Chemistry, 2Department of Immunology, 3Department of Molecular and Experimental
Medicine, The Scripps Research Institute and The
Skaggs Institute for Chemical Biology, 10550 North
Torrey Pines Road, La Jolla, CA 92037, USA. 4Division
of Hematology and Oncology, 5Division of Cardiothoracic and Vascular Surgery, The Scripps Clinic, 10666
North Torrey Pines Road, La Jolla, CA 92037, USA.
6
Laboratorium für organische Chemie, Eidgenössische
Technische Hochschule Hönggerberg HCl-H309, Universitaetstrasse 16 CH-8093 Zürich, Switzerland. 7Department of Biochemistry, Oxford Glycobiology Institute University of Oxford, South Parks Road, Oxford
OX1 3QU, UK.
1
(1–3). One postulate of this work was that
ozone might be generated wherever singlet
(1⌬g) oxygen (1O2*) and antibodies are
juxtaposed, as occurs in most inflammatory
responses.
Current concepts concerning the pathogenesis of atherosclerosis have undergone a
paradigm shift: previously, the process was
thought to be linked to mechanical stress in
large bore arteries; now, inflammation associated with the presence of leukocytes and
immunoglobulins is thought to play a central
role (4–8). Given that all the components
necessary to activate the antibody-catalyzed
water-oxidation pathway are thought to be
present in atherosclerotic arteries, we investigated whether ozone is produced during the
evolution of human atherosclerosis and
whether excised advanced plaque material
could be induced to produce ozone in vitro.
We have previously demonstrated that
among the oxidants thought to be associated
with inflammation, only ozone oxidizes indi-
www.sciencemag.org SCIENCE VOL 302 7 NOVEMBER 2003
1053
REPORTS
go carmine 1, a chemical trap for ozone (9,
10), to give isatin sulfonic acid 2 with isotope
incorporation from H218O into the lactam
carbonyl of 2 (Fig. 1) (1, 2). Thus, using 1 as
an analytical tool, we investigated whether
atherosclerotic tissue, obtained by carotid
endarterectomy (n ⫽ 15) and treated with
Fig. 1. Oxidation of indigo carmine 1 to isatin
sulfonic acid 2 by PMA-treated human atherosclerotic plaque (11). (A) Endpoint observation
of the bleaching of 1 by a PMA-activated atherosclerotic lesion. Each vial contains an equal
amount of a dispersion of atherosclerotic
plaque (⬃50 mg wet weight) in a solution of 1
(200 ␮M) and bovine catalase (50 ␮g) in PBS
(10 mM sodium phosphate, 150 mM NaCl) with
pH 7.4. The photograph was taken 30 min after
the addition of a solution of PMA (10 ␮L, 40
␮g/mL) in DMSO to the vial on the right
(DMSO of the same volume was added to the
vial on the left). Total volume of reaction was 1
mL. In all cases for which only DMSO was
added, no bleaching of the visible absorbance of
1 occurred. Molecular structures of 1 (left) and
2 (right) are shown above (A). (B) Reversedphase HPLC analysis of the supernatant of the
two vials shown in (A). Isatin sulfonic acid 2 has
a RT ⬃ 9.71 min. (C) Negative ion electrospray
mass spectrometry (MS) of the supernatant,
after centrifugation, obtained from activation
of representative human atherosclerotic plaque
material with PMA (40 ␮g) in PBS ( pH 7.4) in
H218O (⬎95%) in the presence of bovine catalase (100 ␮g), as described in (A) (11). M/Z,
mass-charge ratio.
1054
phorbol myristate acetate (PMA), generated
an oxidant with the chemical signature of
ozone (Fig. 1). Plaque samples in phosphatebuffered saline (PBS) ( pH 7.4) were split into
two equal portions and either dimethyl sulfoxide (DMSO) or a solution of PMA in
DMSO was added (11). Bleaching of the
visible absorbance of 1 was observed in 14 of
the 15 plaque samples upon PMA addition
(Fig. 1A) (12). Bleaching was accompanied
by the formation of isatin sulfonic acid 2 in
amounts that varied among the plaques (1.0
to 262.1 nmol/mg) (Fig. 1B, table S1). When
the PMA treatment of atherosclerotic plaque
material occurred in a solution of 1 in H218Ocontaining PBS (⬎95% 18O) (n ⫽ 2), about
40% of the lactam carbonyl oxygen of 1
incorporated 18O, as shown by the relative
intensities of the [mass – H]– 228 and 230
mass fragment peaks in the mass spectrum of
the isolated cleaved product 2 (Fig. 1C).
Cholesterol is a major lipid component of
atherosclerotic plaques (13) and is present at
such high concentrations that, in certain cases, it can form a crystalline phase within the
lipid core of the diseased artery. Previous
studies have shown that among a panel of
oxidants such as triplet oxygen (3O2 ), 1O2*,
superoxide anion, hydroxyl radical (from the
reaction between hydrogen peroxide and
Fe2⫹), and ozone, only ozone cleaves the ⌬5,6
double bond of cholesterol 3 to yield the
5,6-secosterol 4a as the principle product
(14–17) (Fig. 2A). Thus, we analyzed excised atherosclerotic plaque material (n ⫽ 14)
for the presence of 4a, before and after PMA
treatment, using a modification of the proce-
dure developed in a chemical study by Pryor
and colleagues (18). In brief, the method
involved extraction of a suspension of homogenized plaque material into an organic
solvent, and then treatment of the organic
extract with 2,4-dinitrophenylhydrazine hydrochloride (DNPH HCl) (11). The derivatized plaque extract (DPE) was then analyzed
by high-performance liquid chromatography
(HPLC) and mass spectroscopy for the presence of 4b, the 2,4-dinitrophenylhydrazone
derivative of 4a (Fig. 3C) (11). Hydrazone 4b
was detected in 11 of 14 unactivated DPEs
(between 6.8 and 61.3 pmol/mg of plaque)
(Fig. 3A) and in all activated DPEs (between
1.4 and 200.6 pmol/mg) (Fig. 3B). The
amounts of 4a, as judged by the mean
amounts of 4b, significantly increased after
PMA treatment (Fig. 3F).
In addition to 4b, two hydrazones were
observed during the HPLC analysis of
DPEs; peak A had a retention time (RT) ⬃
20.5 min and [M-H]– ⫽ 597, and peak B
had a RT ⬃ 18.0 min and [M-H]– ⫽ 579
(Fig. 3, A and B). By comparative analyses
and co-injection with authentic samples,
peak A was assigned as the hydrazone derivative 5b of the aldol condensation product 5a (Fig. 3C) and peak B was assigned as
the hydrazone derivative 6b of the A-ring
dehydration product 6a (Fig. 3D).
Molecules that contain primary or secondary amino groups are known to catalyze
crossed-aldolization reactions between ketones and aldehydes (19), so we studied
whether such components, which are present
in plaques and blood, facilitate the conver-
Fig. 2. (A) Chemical formulas showing cholesterol (3), its ozonolysis product 5,6-secosterol (4a),
the aldolization product (5a), and their respective hydrazone derivatives 4b and 5b. Under
conditions of hydrazone formation, the amount of 5b formed from 4a during the derivatization
process is found to be ⬃20% (determined by HPLC) (11). The configurational assignments of 5a
and 5b are those made within the literature (18). (B) Synthesized dehydration product 6a and its
DNPH derivative 6b. (RT ⬃ 18 min and [M-H]– 579 in Fig. 3, A, B, and D).
7 NOVEMBER 2003 VOL 302 SCIENCE www.sciencemag.org
REPORTS
sion of 4a into 5a. The following components
accelerate this reaction (11): L-Pro (2 hours,
complete conversion), Gly (24 hours, complete conversion), L-Lys䡠HCl (24 hours, complete conversion), L-Lys(ethyl ester)䡠2HCl
(100 hours, 62% conversion), and extracts
from atheromatous arteries (22 hours, complete conversion), whole blood (15 hours,
complete conversion), plasma (15 hours,
complete conversion), and serum (15 hours,
complete conversion).
As described above, the amount of ketoaldehyde 4a within the plaques significantly
increases upon activation with PMA. However, in the case of aldol 5a, the overall effect
of plaque activation is less clear. Cases are
observed for which the levels of 5a increase
(Fig. 3G, patients F and H) or decrease (Fig.
3G, patients C, G, and N) upon plaque activation. At this time, we have no experimental
evidence to account for this plaque-to-plaque
variability, and further studies are underway
to answer this question.
Although inflammation is widely accepted as a factor in the pathogenesis of
atherosclerosis (4, 5), no specific noninvasive
method exists that can distinguish inflammatory artery disease from other inflammatory
processes. We have analyzed plasma samples
from two cohorts of patients for the presence
of either 4a or 5a. Cohort A consisted of
patients (n ⫽ 8) that had an atherosclerosis
disease state that was sufficiently advanced to
warrant endarterectomy. Cohort B (n ⫽ 15)
consisted of random patients that were attending a general medical clinic. In six of
eight of the patients in cohort A, aldol 5a was
detected, in amounts ranging from 70 to 1690
nM (Fig. 3H) (⬃1 to 10 nM is the detection
limit of our assay) (11). In only 1 of the 15
plasma samples from cohort B is there detectable 5a. No ketoaldehyde 4a was detected in
any patient’s blood sample (⬃1 to 10 nM is
the detection limit of our assay). We surmise
that either 4a is converted into 5a by catalysts
contained in the blood or components within
the plasma have differential affinities for 4a
and 5a. In the past, serum analysis of “oxysterols” has been fraught with difficulty because of problems with cholesterol auto-oxidation (20). However, as described above,
among all of the oxidation products of cholesterol by biologically relevant oxidation of
cholesterol 3, steroid derivatives 4a and 5a
are unique to ozone. Although an investigation of the clinical significance of these results is beyond the scope of the present study,
the implication is that the presence of the
aldolization product 5a in plasma, detected as
its DNP hydrazone derivative 5b, could be a
marker for advanced arterial inflammation in
atherosclerosis.
Many cholesterol oxidation products possess biological activities such as cytotoxicity
(21), atherogenicity (22), and mutagenicity
Fig. 3. Liquid chromatography–mass spectrometry analysis of DPEs and synthesized authentic
samples of hydrazones 4b, 5b, and 6b. Conditions: Adsorbosphere-HS reverse-phase C18 column, 75% acetonitrile, 20% water, 5% methanol, 0.5 mL/min flow rate, 360-nm detection,
inline negative ion electrospray MS (Hitachi
M8000 machine) of a plaque extract after derivatization with DNPH. 4b (RT ⬃ 14.1 min), 5b
(RT ⬃ 20.5 min), and 6b (RT ⬃ 18 min) in an
atherosclerotic lesion before activation with
PMA (40 ␮g/mL). (A) Plaque material before
activation with PMA. (B) HPLC and MS analysis
[under the same conditions as in (A)] of plaque
material after activation with PMA (40 ␮g/mL),
extraction, and derivatization with an ethanolic
solution of DNPH hydrochloride (2 mM) for 2
hours. (C to E) HPLC and MS (insets) analysis of
synthetic 4b (C), synthetic 6b (D), and synthetic
5b (E). The mass fragment label above each
peak in (C), (D), and (E) is the observed [M-H]–
fragment. All peak assignments were made by comparison to co-injection with synthetic materials.
(F) Bar chart showing the measured concentration of 4b (11) after extraction and derivatization of
4a from atherosclerotic lesions of patients, pre- and postactivation with PMA. Numbers of amounts
before and after activation determined by a Student’s t test (two-tail) (P ⬍ 0.05, n ⫽ 14) analysis
using GraphPad Prism V3 for Macintosh. (G) Bar chart showing the measured concentration of 5b
after extraction and derivatization of 5a from atherosclerotic lesions of patients, pre- and
postactivation with PMA (n ⫽ 14). (H) Bar chart showing measured concentrations of 5b after
extraction and derivatization of 5a from plasma samples taken from patients. Cohort A (n ⫽ 8)
patients were to undergo a carotid endarterectomy procedure within 24 hours (plasma analysis
was performed 3 days after sample collection). Cohort B (n ⫽ 15) consisted of random patients
attending a general medical clinic (plasma analysis was performed 7 days after sample collection).
Incubation of 5a in plasma has revealed that in vitro plasma levels of 5a fall by ⬃5% per day (at
room temperature); typically the efficiency with which 5a is extracted from plasma is ⬎90%, and
the values of replicate extractions are within 5%. Under the conditions of our assay, the detection limit
of 4b and 5b is 1 to 10 nM; in cases for which no value of 4b or 5b is given, the level is ⬍10 nM.
www.sciencemag.org SCIENCE VOL 302 7 NOVEMBER 2003
1055
REPORTS
exclude O3 generation by an alternative
chemical or biochemical route. If such a route
exists, it would have to be triggered by PMA.
The results presented here suggest that the
endogenous production of ozone may be a
contributory factor in atherosclerosis and
may link the otherwise seemingly independent factors of cholesterol accumulation
and oxidation, inflammation, and cellular
damage that contribute to the pathogenesis
of this disease.
Until now, dioxygen has been the sole
form of oxygen thought to be biologically
relevant. The discovery that trioxygen species occur in biology changes the landscape
of oxidation chemistry in vivo and could lead
to the discovery of new reaction pathways
and products that are of critical importance in
normal and pathological situations.
References and Notes
Fig. 4. LDL incubated with atheronal-A 4a and
atheronal-B 5a induces lipid-loading of macrophages to produce foam cells. J774.1 macrophage cells were incubated in media plus 2,6di-tert-butyl-4-methylphenol and with (A) LDL
(100 ␮g/mL) or (B) LDL and atheronal-A 4a (20
␮M). Cells were fixed with 4% formaldehyde
and stained with hematoxylin and oil red O
(visualized at ⫻100). The effect of atheronal-B
5a on lipid-loading with the J774.1 macrophages was indistinguishable from 4a.
(23). Given that the atheronals 4a and 5a
have never before been considered to occur in
man, we have investigated their effects on
key aspects of atherogenesis. These molecules are cytotoxic to a range of cell types
present within atheromatous arteries (fig.
S4); upon coadministration with low-density
lipoprotein (LDL), they trigger foam cell formation in tissue macrophages (Fig. 4), and
upon binding to LDL, they cause a significant
and time-dependent loss in the secondary
structure of apolipoprotein B-100 (fig. S5).
Further studies are in progress to fully elucidate the role of 4a and 5a in atherosclerosis.
The experimental evidence shows (i) that
PMA-treated plaque material generates a molecular species that reacts with the double
bond of indigo carmine 1 by a process that
has the signature of ozonolysis and (ii) that
the ⌬5,6– double bond of cholesterol is oxidized to yield the atheronals, a process unique
to ozone (24). This is compelling evidence
that atherosclerotic plaques can generate
ozone. Furthermore, because the unique
ozone oxidation products of cholesterol, 4a
and 5a, are present in excised atherosclerotic
plaques before PMA activation, it seems likely that ozone is generated throughout the
evolution of the disease. However, given that
atherosclerotic material from the artery is a
far more complex system than we have previously studied (1–3, 25, 26), we do not
1056
1. P. Wentworth Jr. et al., Science 298, 2195 (2002);
published online 14 November 2002 (10.1126/
science.1077642).
2. B. M. Babior, C. Takeuchi, J. Ruedi, A. Guitierrez, P.
Wentworth Jr., Proc. Natl. Acad. Sci. U.S.A. 100,
3920 (2003).
3. P. Wentworth Jr. et al., Proc. Natl. Acad. Sci. U.S.A.
100, 1490 (2003).
4. R. Ross, New Engl. J. Med. 340, 115 (1999).
5. G. K. Hansson, P. Libby, U. Schönbeck, Z.-Q. Yan, Circ.
Res. 91, 281 (2002).
6. D. Steinberg, J. Biol. Chem. 272, 20963 (1997).
7. D. Steinberg, S. Parthasarathy, T. E. Carew, J. C. Khoo,
J. L. Witztum, New Engl. J. Med. 320, 915 (1989).
8. U. P. Steinbrecher, S. Parthasarathy, D. S. Leake, J. L.
Witzum, D. Steinberg, Proc. Natl. Acad. Sci. U.S.A.
81, 3883 (1984).
9. K. Takeuchi, S. Kutsuna, T. Ibusuki, Anal. Chim. Acta
230, 183 (1990).
10. K. Takeuchi, I. Takeuchi, Anal. Chem. 61, 619 (1989).
11. Materials and methods are available as supporting
material on Science Online.
12. Quantification of the oxidation of 1 was performed
by HPLC analysis of 2, as described previously (1).
13. D. M. Small, Arteriosclerosis 8, 103 (1988).
14. J. Gumulka, L. L. Smith, J. Am. Chem. Soc. 105, 1972
(1983).
15. K. Jaworski, L. L. Smith, J. Org. Chem. 53, 545 (1988).
16. Z. Paryzek, J. Martynow, W. Swoboda, J. Chem. Soc.
Perkin Trans. 1, 1222 (1990).
17. J. W. Cornforth, G. D. Hunter, G. Popják, Biochem. J.
54, 590 (1953).
18. K. Wang, E. Bermúdez, W. A. Pryor, Steroids 58, 225
(1993).
19. J.-L. Reymond, Y. Chen, J. Org. Chem. 60, 6970
(1995).
20. E. Lund, I. Björkhem, Acc. Chem. Res. 28, 241 (1995).
21. H. Hietter, P. Bischoff, J. P. Beck, G. Ourisson, B. Luu,
Cancer Biochem. Biophys. 9, 75 (1986).
22. J. L. Lorenso, M. Allorio, F. Bernini, A. Corsini, R.
Fumagalli, FEBS Lett. 218, 77 (1987).
23. A. Sevanian, A. R. Peterson, Proc. Natl. Acad. Sci.
U.S.A. 81, 4198 (1984).
24. J. Gumulka, J. St-Pyrek, L. L. Smith, Lipids 17, 197
(1982).
25. A. D. Wentworth, L. H. Jones, P. J. Wentworth, K. D.
Janda, R. A. Lerner, Proc. Natl. Acad. Sci. U.S.A. 97,
10930 (2000).
26. P. Wentworth Jr. et al., Science 293, 1806 (2001).
27. We thank members of the TSRI mass spectroscopy
facility, especially G. Suizdak and B. Webb for assistance with the 18O isotope analysis; members of the
Scripps Hybridoma laboratory, especially D. Kubiak,
C. Bautista, and L. Kerwin for tissue culture and
cytotoxicity assay assistance. Supported by NIH
PO1CA27489 (Program Project Grant to R.A.L.), GM
43858 (K.D.J.), and The Skaggs Institute for Chemical
Biology. C.T. is supported in part by NIH training
grant (5T32AI07606); R.G. is supported by La Secretaria de Estado de Educacion y Universidades and El
Fondo Social Europeo.
Supporting Online Material
www.sciencemag.org/cgi/content/full/302/5647/1053/
DC1
Materials and Methods
Figs. S1 to S4
Tables S1 and S2
References
23 July 2003; accepted 15 September 2003
Induction of APOBEC3G
Ubiquitination and Degradation by
an HIV-1 Vif-Cul5-SCF Complex
Xianghui Yu,1,2* Yunkai Yu,1* Bindong Liu,1* Kun Luo,1
Wei Kong,2 Panyong Mao,1 Xiao-Fang Yu1,3†
Human immunodeficiency virus–1 (HIV-1) Vif is essential for viral evasion of
host antiviral factor CEM15/APOBEC3G. We report that Vif interacts with
cellular proteins Cul5, elongins B and C, and Rbx1 to form an Skp1-cullin-F-box
(SCF)–like complex. The ability of Vif to suppress antiviral activity of APOBEC3G
was specifically dependent on Cul5-SCF function, allowing Vif to interact with
APOBEC3G and induce its ubiquitination and degradation. A Vif mutant that
interacted with APOBEC3G but not with Cul5-SCF was functionally inactive. The
Cul5-SCF was also required for Vif function in distantly related simian immunodeficiency virus mac. These results indicate that the conserved Cul5-SCF
pathway used by Vif is a potential target for antiviral development.
The vif open reading frame is present in all
lentiviruses except equine infectious anemia virus and is required for viral replication and
pathogenicity in vivo (1–12). In the absence of
Vif, HIV-1 virions that are produced from nonpermissive primary T lymphocytes and certain
T cell lines are defective and cannot initiate
productive infection. Recently, CEM15/
7 NOVEMBER 2003 VOL 302 SCIENCE www.sciencemag.org