Hock cleavage of cholesterol 5α-hydroperoxide: An ozone-free pathway to the cholesterol ozonolysis products identified in arterial plaque and brain tissue

Article abstract of DOI:10.1021/ja804162d

Cholesterol 5α-hydroperoxide, the major product of ¹O₂-oxidation of cholesterol, readily undergoes acid-catalyzed (Hock) cleavage of the C5-C6 bond to yield the 5,6-secosterol ketoaldehyde and the product of its intramolecular aldoli

Full text of DOI:10.1021/ja804162d

Published on Web 08/23/2008
Hock Cleavage of Cholesterol 5r-Hydroperoxide: An Ozone-Free Pathway to the
Cholesterol Ozonolysis Products Identified in Arterial Plaque and Brain Tissue
Johan Brinkhorst, Susheel J. Nara, and Derek A. Pratt*
Department of Chemistry, Queen’s UniVersity, 90 Bader Lane, Ontario, Canada K7L 3N6
Received June 2, 2008; E-mail: pratt@chem.queensu.ca
In a series of fascinating papers in the early 2000s, Wentworth
and co-workers suggested that ozone is produced endogenously by
antibody-catalyzed oxidation of water by singlet dioxygen.1,2 Two
substrates believed to undergo reactions unique to ozone were used
as probes to obtain evidence for ozone formation and its reactivity
to biomolecules: the cleavage of the central bond in indigo carmine
1 to yield isatin sulfonic acid 2 with isotope incorporation from
H218O into the lactam carbonyl of 2 (eq 1) and the cleavage of the
∆
5,6 bond in cholesterol (3) to yield the 5,6-secosterol ketoaldehyde
4 (eq 2), respectively.
Figure 1. 1H NMR spectra (400 MHz; CDCl3) of 6 (top) and 5 (bottom)
formed immediately following addition of 1 uL of trifluoroacetic acid (/ )
H2O). Relevant protons are indicated in structures of 6 and 5 (substituents
on A,B rings are omitted for clarity).
compounds are known to arise from decomposition of lipid
hydroperoxides derived from 3O2- and/or 1O2-mediated lipid
peroxidation,11 and we wondered whether cholesterol hydroper-
oxides could serve as precursors to 4 (and 5). Here, we show that
cholesterol 5R-hydroperoxide (6), the major product of 1O2-
oxidation of cholesterol,12 readily undergoes acid-catalyzed (Hock)
cleavage13 of the C5-C6 bond.
Since the identification of 4 (following derivatization to the 2,4-
dinitrophenylhydrazones, 4-DNPH and 5-DNPH)3 in extracts of
human arterial plaque,2 the proatherogenic effects of 4 (and 5) have
been studied to assess their roles in the pathogenesis of cardiovascular
disease.4 This work has been complemented by the identification of 4
(again, following derivatization to 4-DNPH and 5-DNPH) in human
brain tissue,5 and associated investigations of the role of 4 (and 5) in
the pathogenesis of Alzheimer’s6 and Parkinson’s7 disease.
Many have expressed their reservations about the conclusions
drawn in the foregoing works.8 Pessimism has been fueled further
by Winterbourne and co-workers,9 who demonstrated that the
conversion of 1 to 2 with isotope incorporation from H218O into
the lactam carbonyl of 2 is not unique to ozone, but can also be
carried out by superoxide (hydroperoxyl radical). While this is an
important control experiment for evaluating the validity of the
suggestion that ozone is produced endogenously, it does not provide
an alternative explanation for the appearance of the 2,4-dinitro-
phenylhydrazones derived from 4, the known cholesterol ozonolysis
product,10 in derivatized extracts of arterial plaque and brain tissue.
While we were also fascinated by the suggestion that ozone is
produced endogenously, we wondered whether a more simple
explanation for the formation of 4 (and 5) could exist. Carbonyl
We prepared 6 by photosensitized (singlet) oxidation of choles-
terol14 and subjected it to a variety of protic acids (e.g., HCl,
p-TsOH, TFA) in different solvents (e.g., EtOH, THF, CHCl3),
under all of which Hock cleavage was found to proceed quite readily
as evidenced by 1H NMR (e.g., Figure 1), 13C NMR, and HRMS.
In fact, conversion could be detected in NMR samples of 6,
presumably due to adventitious DCl in the CDCl3.
The treatment of 6 with acid in CHCl3 and THF led to the
quantitative formation of 5 (as shown in Figure 1) which can be
understood by the precedented13 mechanism in Scheme 1 (path a).
9
12224 J. AM. CHEM. SOC. 2008, 130, 12224–12225
10.1021/ja804162d CCC: $40.75
2008 American Chemical Society

C O M M U N I C A T I O N S
solvents,14 it is difficult to envision 6 having a sufficient lifetime
to survive surgical tissue removal, homogenization, extraction, and
concentration all under an aerobic atmosphere.
Scheme 1. Proposed mechanism of Hock cleavage of cholesterol
5R-OOH in CHCl3 and THF (path a from I) and EtOH (where path
b from I competes). Substituents on A, B rings are omitted for
clarity.
In summary, we have shown that the cholesterol-derived keto-
aldehyde 4 and its aldolization product 5 can arise from Hock
cleavage of the cholesterol 5R-OOH 6, which has been shown to
occur due to singlet oxygen oxidation of cholesterol (or some
hitherto unknown mechanism) in vivo.18 While our findings do not
unequivocally dismiss the possibility that ozone is produced
endogenously, or that ozone is responsible for the cleavage of the
5,6
∆
bond in cholesterol in vivo, they do clearly demonstrate that
4 and 5 are cholesterol oxidation products not unique to ozonolysis.
Acknowledgment. We are grateful for the support of the Natural
Sciences and Engineering Research Council of Canada and the
Ontario Ministry of Innovation. D.A.P. acknowledges the support
of the Canada Research Chairs program.
Reactions in EtOH, however, led to mixtures of 5 and 4 in a ratio
of ∼5:1. The appearance of 4 in EtOH can be explained on the
basis of the competing capture of oxocarbenium ion I by EtOH to
yield the hemiketal of 4 (path b, Scheme 1), which prevents
cyclization of I to give 5 as in CHCl3 and THF.15 Conversion of 6
to 5 and 4 was followed in ethanolic solutions containing HCl at
concentrations that spanned 4 orders of magnitude (0.1 M - 0.0001
M) to confirm the dependence of the rate of conversion on acid
concentration as well as to confirm that complete conversion was
observed at substoichiometric amounts of acid, supporting acid-
catalysis in the reaction (see Supporting Information).
Supporting Information Available: Complete refs 1a, 2, and 5;
experimental procedures for Hock cleavage reactions, derivatizations
and analyses. This material is available free of charge via the Internet
at http://pubs.acs.org.
References
(1) (a) Wentworth, P., Jr; et al. Science 2002, 298, 2195–2199. (b) Wentworth,
P., Jr.; Wentworth, A. D.; Zhu, X.; Wilson, I. A.; Janda, K. D.;
Eschenmoser, A.; Lerner, R. A. Proc. Natl. Acad. Sci. U.S.A. 2003, 100,
1490–1493. (c) Babior, B. M.; Takeuchi, C.; Ruedi, J.; Gutirrez, A.;
Wentworth, P., Jr Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 3031–3034.
(2) Wentworth, P., Jr.; et al. Science 2003, 302, 1053–1056.
(3) Wang, K.; Bermu´dez, E.; Pryor, W. A. Steroids 1993, 58, 225–229.
(4) Takeuchi, C.; Galve, R.; Nieva, J.; Witter, D. P.; Wentworth, A. D.; Troseth,
R. P.; Lerner, R. A.; Wentworth, P. Biochemistry 2006, 45, 7162–7170.
(5) Zhang, Q.; et al. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 4752–4757.
(6) Bieschke, J.; Zhang, Q.; Powers, E. T.; Lerner, R. A.; Kelly, J. W.
Biochemistry 2005, 44, 4977–4983.
Interestingly, upon DNPH derivatization of both arterial2 and
brain5 tissue extracts, Wentworth and co-workers found 5-DNPH
to be the major product under conditions that they indicated resulted
in ∼20% aldolization of 4-DNPH to 5-DNPH (eq 3). If ozonolysis
was the route by which they had been formed, 4-DNPH should
have been the major product. It was suggested that amino acids or
other entities containing 1° or 2° amino groups were responsible
for catalyzing the aldolization of 4 to 5 in vivo, leading to the greater
than expected amount of 5-DNPH in the derivatized samples. In
light of our results, the predominance of 5-DNPH in the chro-
matograms of the derivatized tissue extracts tempts us to suggest
the that endogenous ozonolysis of cholesterol does not occur to
yield 4 and 5, but instead that Hock fragmentation of 6seither in
vivo or under the acidic conditions of the derivatizationsis
responsible for the predominance of 5.16
(7) (a) Bosco, D. A.; Fowler, D. M.; Zhang, Q. H.; Nieva, J.; Powers, E. T.;
Wentworth, P.; Lerner, R. A.; Kelly, J. W. Nat. Chem. Biol. 2006, 2, 249–
253. (b) Stewart, C. R.; Wilson, L. M.; Zhang, Q.; Pham, C. L. L.;
Waddington, L. J.; Staples, M. K.; Stapleton, D.; Kelly, J. W.; Howlett,
G. J. Biochemistry 2007, 46, 5552–5561.
(8) Most notably: Smith, L. L. Free Radical Biol. Med. 2004, 37, 318–324.
Sies, H. Angew. Chem. Intl. Ed. 2004, 43, 3514–3515.
(9) Kettle, A. J.; Clark, B. M.; Winterbourne, C. C. J. Biol. Chem. 2004, 279,
18521–18525.
(10) Smith, L. L. Chem. Phys. Lipids 1987, 44, 87–125.
(11) Esterbauer, H.; Schaur, R. J.; Zollner, H. Free Radical Biol. Med. 1991,
11, 81–128.
When 6 was subjected to typical derivatization conditions2 (0.2 mM
DNPH, 0.1 M HCl in EtOH, stirred for 2 h at room temperature under
argon), the same dinitrophenylhydrazones were formed as those
obtained following derivatization of the cholesterol ozonolysis product
4, but in ratios reflective of Hock fragmentations, where 5 predomi-
nates. However, under these conditions, the observed ratio of 5-DNPH:
4-DNPH is much larger than the ratio of ∼5:1 observed in the absence
of DNPH (see Supporting Information).17 This suggests that if Hock
fragmentation of 6 is responsible for the appearance of 4 and 5 in the
tissue extracts, it occurs in vivo and not upon derivatization.
The likelihood that Hock fragmentation occurs in vivo is
reinforced by the expected transient nature of cholesterol 5R-OOH,
which is known to rearrange readily to the more thermodynamically
stable cholesterol 7R-OOH in solution (7, eq 4).14 While this
rearrangement, which is believed to occur via an intermediate
allylperoxyl radical, is slowed somewhat in H-bond accepting
(12) Schenck, G. O.; Neumu¨ller, O.-A.; Eisfeld, W. Justus Liebigs Ann. Chem.
1958, 618, 202.
(13) Frimer, A. A. Chem. ReV. 1979, 79, 359–387.
(14) Beckwith, A. L. J.; Davies, A. G.; Davison, I. G. E.; Maccoll, A.; Mruzek,
M. H. J. Chem. Soc., Perkin Trans. II 1989, 815–824.
(15) If the first and second oxocarbenium ion are trapped by EtOH this would
lead to the ketal, which would also prevent cyclization. Regardless, it should
be pointed out that 4 and 5 exist as their diethyl acetals under these
conditions, which prevents cyclization of 4 to 5.
(16) It should be pointed out that Arnold and Staples have shown that allylic
hydroperoxides readily undergo Hock cleavage under standard DNPH
derivatization conditions, see: Arnold, A. R.; Staples, R. Polymer 1992,
33, 1739–1741.
(17) The change in product distribution in the presence of DNPH may be the
result of hydrazone formation on the (hemi)ketal form of 4 generated
directly from the Hock fragmentation; the enamine form would readily
cyclize upon formation of any oxocarbenium intermediate upon intercon-
version of the (hemi)ketal and ketone.
(18) Yamazaki, S.; Ozawa, N.; Hiratsuka, A.; Watabe, T. Free Radical Biol.
Med. 1999, 27, 301–308.
JA804162D
9
J. AM. CHEM. SOC. VOL. 130, NO. 37, 2008 12225