Medium sized lactones with hypolipidaemic and antioxidant activity: Synthesis and biological evaluation of promising dual-action anti-atherosclerosis drugs

Article abstract of DOI:10.1016/S0968-0896(98)00252-1

Macrocyclic lactones 1a-b have been synthesized and their potential therapeutic value evaluated. The key structural feature of these active 'chimera' compounds is the 12-membered lactone ring that brings together the well-known polysubstituted hydroquinon

Full text of DOI:10.1016/S0968-0896(98)00252-1

Bioorganic & Medicinal Chemistry 7 (1999) 411±418
Medium Sized Lactones with Hypolipidaemic and Antioxidant
Activity: Synthesis and Biological Evaluation of Promising
Dual-action Anti-atherosclerosis Drugs
Claudia Baldazzi,b Francesca Calderoni, Emanuela Marotta,a Silvano Piani,b
Paolo Righi,a Go€redo Rosini,a,* Stefano Saguatti,b Roberta Tiozzo,c
Sebastiano Calandrac and Francesca Venturelli
a
Á
Dipartimento di Chimica Organica, ``A. Mangini,'' Universita di Bologna, Viale Risorgimento 4, 40136 Bologna, Italy
bDipartimento Chimico e Biologico, Alfa-Wassermann S.p.A., Via Ragazzi del `99, 5, 40133 Bologna, Italy
c
Á
Dipartimento di Scienze Biomediche, Universita di Modena, Via Campi 287, 41100 Modena, Italy
Received 15 July 1998; accepted 20 October 1998
AbstractÐMacrocyclic lactones 1a±b have been synthesized and their potential therapeutic value evaluated. The key structural
feature of these active `chimera' compounds is the 12-membered lactone ring that brings together the well-known polysubstituted
hydroquinone moiety of antioxidants and the a,a-dimethyl substituted acyl residue of gem®brozil. Lactones 1a±b showed better
activity than probucol, a classical phenolic antioxidant, in preventing the Cu++-induced oxidative modi®cation of human LDL.
The hypolipidaemic activity of the new lactones, evaluated as the inhibition of lipids biosynthesis in Hep-G2 cells, was comparable
to that of gem®brozil. These features, added to the lack of cytotoxicity, make this new class of medium sized lactones promising
dual-action drugs useful as anti-atherosclerosis agents. # 1999 Elsevier Science Ltd. All rights reserved.
Introduction
(LDL) have proved to be e€ective in the prevention of
the vascular coronary pathologies and in the treatment
of atheromatous plaques.3,4 The atherogenic risk asso-
ciated with the LDL appears to be related not only to
their plasma concentration but also to their qualitative
characteristics. The possible modi®cations of the struc-
ture and composition of the LDL in the plasma or in
the arterial wall can in fact make these macromolecules
more atherogenic, namely more apt to trigger o€ and to
stimulate the formation of the atheromatous plaque.5
The oxidative processes that occur by action of oxidiz-
ing agents present in the plasma or in the endothelial
cells of the arterial wall are of main importance among
the modi®cations that the LDL can undergo in vivo.
Many in vitro and in vivo experiments, together with
the results of epidemiological investigations, support the
hypothesis that this mechanism represents an important
event in the development of atherosclerosis.6±9 The
LDL are particles not only enriched in cholesterol but
also polyunsaturated fatty acids (PUFA) highly suscep-
tible to undergo the peroxidation from which they are
protected by the presence of many endogenous anti-
oxidizers like tocopherols, carotenes, lycopene, and
ubiquinol-10.
Atherosclerosis is a process of degeneration of the
arterial wall and is associated with many vascular
pathologies like ischaemic cardiopathies (angina pec-
toris and myocardial infarction) and cerebral thrombo-
sis, main causes of death in the industrial countries.
Many e€orts have been carried out to understand the
aetiology of this pathology of wide di€usion and
importance and to ®nd out the possible therapeutic
treatments.
Atherosclerosis is recognized to be a process with com-
posite aetiology that involves in varied measure di€erent
factors and cell types. Endothelial damages represent
the main event in the genesis of the atheroma1 and are
promoters of the progress and the complication of the
atheromatous plaque in the high serum rates of lipids
and cholesterol that characterize the hyperlipidaemias
and that often aggravate the diabetic pathologies.2 The
pharmacological interventions aimed at lowering the
plasma levels of cholesterol and low-density lipoproteins
Key words: Antioxidants; radical scavengers; medium sized lactones;
hypolipidaemics.
The oxidation of the LDL is a chain reaction of lipidic
peroxidation lead by free radicals able to quickly
*Corresponding author. Fax: +39 051 644 3654; e-mail: rosini
@ms.fci.unibo.it
0968-0896/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(98)00252-1

412
C. Baldazzi et al. / Bioorg. Med. Chem. 7 (1999) 411±418
transform the PUFA into lipidic hydroperoxides (pro-
pagation period). The endogenous antioxidizers act as
chain-breaking agents and interrupt the propagation by
means of an e€ective scavenging of the peroxylic radi-
cals and the concentration of hydroperoxides increases
only when the endogenous antioxidizers are depleted
(latency period).10±12
Figure 1.
Chain-breaking antioxidants play a key role in the con-
trol of LDL and can be depicted as anti-atherosclerotic
drugs,13±15 as well as the lipid-soluble antioxidant vita-
min E (tocopherol) is an e€ective inhibitor of LDL oxi-
dation and have great importance in the prevention of
atherosclerosis. Perspective studies on a large scale show
a relation between the taking of considerable amounts
of vitamin E and the lowering of the risk of vascular
pathologies, both in men and in women.16,17
In particular they seems to prevent and/or to delay the
oxidative modi®cation of the human LDL, i.e. they
compete with the chain of propagation of the lipidic
peroxidation through an e€ective scavenging of the
peroxylic radicals.
The oxidative modi®cations of LDL represent, as it has
already been seen, a key event in the pathogenesis of
atherosclerosis and therefore the lactones 1a±b can ®nd
useful therapeutic application in the treatment of the
atherosclerosis and in the prevention of many vascular
pathologies like ischaemic cardiopathies (angina pec-
toris and myocardial infarction), cerebral thrombosis,
and periferal arteriopathies.
At present, therapy of disorders of the lipidic metabo-
lism (the familial heterozygotic hypercholesterolaemia,
the combined familial hyperlipidaemia, the dis-b-lipo-
proteinaemia, the familial hypertriglyceridaemia and the
polygenic hypercholesterolaemia, all important causes
of the atherosclerosis), is based on the use of drugs
having hypolipidaemic activity like the ®brates (for
instance gem®brozil (Fig. 1), nicotinic acid, the resins
sequestering the bile acids and the inhibitors of the
enzyme HMGCoA reductase. The metabolic e€ect of
these drugs is that of lowering the plasma concentration
of cholesterol (LDL) and partly of the triglycerides
(VLDL-LDL) through two mechanisms: an increase of
the mediated receptor-removal of the LDL, and/or a
lowering of the synthesis of the VLDL-LDL.18
Results
The synthesis of lactones 1a±b with a 12-membered ring
including the hydroquinone moiety, joined by an ether
bridge, was performed according to Scheme 1.
Compounds 2a±b were prepared following the well-
established procedure of Scott,19 which uses trimethyl-
hydroquinone and alkyl vinyl ketones as inexpensive
starting materials. The reaction of the sodium salts of
2a±b with isopropyl 5-bromo-2,2-dimethylpentanoate
(3) a€orded compounds 4a±b, which upon saponi®cation
The above mentioned considerations point out that
molecules having both hypolipidemic and antioxidant
activity can represent
a therapeutic development
important in the treatment of the disorders of the lipidic
metabolism and in the prevention of the atherosclerosis.
Many hindered phenol-derived antioxidant with both of
these features have been screened since the initial work
on probucol (Fig. 1).
Recently, we have envisioned the possibility to build up
new dual-action drugs by the combination of two
important pharmacophores, the polysubstituted hydro-
quinone structure of classical antioxidants and the 2,2-
gem-dimethyl substituted acyl structure typical of gem-
®brozil, in a novel class of `chimera' compounds with a
12-membered lactone ring as structural feature.
Scheme 1. Synthesis of 14-hydroxy-2,8-dioxabicyclo[10.4.0]hexadec-
12,14,16-trien-7-ones 1a±b. (a) 1. NaH/DMSO, benzene; 2. 2,2-di-
methyl-5-bromopentanoic acid, isopropyl ester (3). (b) KOH/EtOH.
(c) NaBH4/MeOH. (c) 2,20-dipyridyl disul®de, PPh3, CH3CN, AgClO4.
In this paper, we report the synthesis of macrocyclic
lactones 1a±b and some evidences that indicate they are
endowed with antioxidant and hypolipidaemic properties.

C. Baldazzi et al. / Bioorg. Med. Chem. 7 (1999) 411±418
413
Table 1. Latency period evaluation by TBARS method (As reported
in Figure 2(a), the latency period for ethanol as negative control is
38 min)
Table 4. Cytotoxicity e€ect evaluated in vitro on Hep-G2 cells. The
data are expressed as ID50 (mol)
Compd
DMSO
±
EtOH
Compd
Latency periods (min)
3
Probucol
Gem®brozil
1a
1b
>10
±
4
1 mM
5 mM
10 mM
>10
(3.2‹1.5)Â10
5
5
5
(3.2‹1.3)Â10
4
Probucol
1a
1b
±
73
96
±
120
>180
82
>180
>180
(6.3‹1.0)Â10
>10
The electrophoretic mobility of native LDL is taken as a
reference equal to 0, while the migration of the band of
the copper-oxidized LDL at the time of 360 min, only
in presence of ethanol, is taken as reference equal to
100.
Table 2. Relative electrophoretic mobility (REM) on agarose gel of
copper-oxidized LDL
Compd (mM)
REM (%) at 180 min
REM (%) at 360 min
Ethanol (±)
Probucol (10)
1a (10)
60
40
15
15
100
100
30
The inhibition of the oxidative process was evaluated
also by means of the study of the fragmentation of the
ApoB through the electrophoretic analysis on poly-
acrylamide gel in presence of sodium dodecylsulfate
(SDS-PAGE).24,25 The densitometric values related to
the areas of the peaks of the ApoB, expressed as absor-
bance units per millimeter, for the lactones 1a and 1b
and for the probucol, as reference, all used at the con-
centration of 10 mM, are reported in Table 3.
1b (10)
30
gave the corresponding acids 5a±b. The carbonyl group
on the other chain was converted into a hydroxy group
by reduction and the resultant hydroxy acids (6a±b)
were ®nally macrolactonized using the Gerlach mod-
i®cation20 of Corey's `double activation' methodol-
ogy.21 The target molecules 1a±b were obtained in fair
yields and were used for the biological evaluation.
Preliminarily the cytotoxicity was evaluated to establish
the dose to use in lipid synthesis model. The toxicity of
the lactones 1a±b was evaluated in vitro by means of the
cytotoxic e€ect on human hepatocarcinoma (Hep-G2
American Type Culture Collection) cells.26 The cells
growth was evaluated by spectrophotometric determi-
nation of nucleic acids and proteins. The value of the
absorbance is a direct function of the amount of the
macromolecules present and therefore indirectly of the
number of the cells. In Table 4 are reported the values
of ID50, expression of the concentration of compounds
1a±b, in solution both of ethanol and of dimethylsulf-
oxide, able to cause a 50% lowering of the total macro-
molecules.
They were ®rst evaluated by in vitro tests considered
predictive of the antioxidant and hypolipidaemic prop-
erties. In particular, these lactones showed their ability
to prevent and/or to delay the oxidative modi®cation of
LDL, chemically induced by incubation of native LDL
with cupric ions.22 The oxidation of the LDL was
evaluated by monitoring at di€erent time points the
concentration of the aldehydes derived from the degra-
dation of lipid peroxides as well as the modi®cations of
the apolipoprotein B-100 (ApoB). The antioxidant
activity of lactones 1a±b has been expressed in terms of
duration of the latency period which is a parameter of
the resistance to the oxidation of the human LDL. The
results obtained from lactones 1a±b are reported in
Table 1 along with probucol as standard (Fig. 2).
Lactones 1a±b were evaluated in vitro by testing their
e€ect on the synthesis of lipid cells like cholesterol, fatty
acids, triglycerides, phospholipids, and esters of the
cholesterol.27
The inhibition of the oxidative process was also qualita-
tively studied through the evaluation of the modi®cation
of the surface-negative charge (relationship) of the
ApoB by electrophoretic analysis on agarose gel.23 The
relative electrophoretic mobility (REM) of lactones 1a
and 1b and probucol at 10 mM expressed as percent
variation at 180 and 360 min is reported in Table 2.
Lipid synthesis was studied by incubating cells coming
from a human hepatocarcinoma in presence of 14C-
acetate (Amersham) and by measuring28 the incorpora-
tion of the radioactive precursor in the main lipid clas-
ses after chromatographic separation on silica gel.29 The
data reported in Table 5 refer to the incorporation of
Table 3. Oxidative modi®cations on SDS-PAGE of ApoB-100
Compd (mM)
Areaa (AU/mm)
0 min
60 min
90 min
120 min
150 min
180 min
360 min
Ethanol (-)
Probucol (10)
1a (10)
2.3
2.7
2.2
2.5
1.7
1.8
2.0
2.3
1.5
1.6
1.9
2.1
0.5
0.9
1.6
2.0
0.4
0.9
1.5
1.7
0.3
0.6
1.5
1.6
0.2
0.3
1.5
1.6
1b (10)
aAreas of the peaks of the ApoB-100 expressed as absorbance units per millimeter.

414
C. Baldazzi et al. / Bioorg. Med. Chem. 7 (1999) 411±418
14C-acetate in the various lipid classes, expressed as
radioactivity in count per minute (cpm) for culture plate,
for lactones 1a±b, in comparison with the probucol and
the gem®brozil, as standards, and ethanol as negative
control.
Discussion
It should be pointed out that the antioxidant activity of
lactones 1a±b reported in Table 1 is far higher than that
showed by probucol. As a matter of fact, a comparable
Figure 2. Antioxidant activity of (a) ethanol (negative control); (b) probucol (reference compound); and (c) lactone 1b on Cu2+ modi®ed LDL. LDL
(0.5 mg protein/mL) containing ethanol (1.7 mL EtOH/mL LDL); probucol (10 mM) or lactone 1b (1 mM) were incubated at 37 ꢀC in the presence of
5 mM CuSO4. At the indicated times aliquots were removed, EDTA was added (0.24 mM) after cooling at 4 ꢀC and TBARS were determined as
described in Experimental. The intercept on the axis of abscissas of the tangent to the curve at the ¯ex point represents the latency period.
Table 5. In vitro activity on lipid synthesis by incorporation in human hepatocarcinoma cells (Hep-G2) of 14C-acetatea
Compd
Cholesterol (%)
Cholesterol esters (%)
Triglycerides (%)
Fatty acids (%)
Phospholipids (%)
Ethanol
Probucol
16
55
64
34
20
26
40
25
45
64
54
35
76
71
67
73
73
72
+7.0
8.0
3.0
22
11
1.1
4.2
0.9
+1.1
30
36
42
55
44
29
45
42
50
24
31
31
47
34
27
22
35
29
1 mM
10 mM
1 mM
10 mM
1 mM
10 mM
1 mM
10 mM
Gem®brozil
1a
1b
aData are expressed as the variation of 14C-acetate radioactivity (determined as counts per min for culture plate) referred to that of a physiological
solution.

C. Baldazzi et al. / Bioorg. Med. Chem. 7 (1999) 411±418
415
prolongation of the latency period was observed at a
concentration 10 times lower than that of probucol. In
addition, the data reported in Table 2 show that the
lactones 1a and 1b are more e€ective than probucol in
lowering the electrophoretic mobility of LDL, by pre-
venting their oxidative modi®cations. Furthermore, the
data regarding the fragmentation of ApoB, collected in
Table 3, show how lactones 1a±b prevent the dis-
appearance of the ApoB in a more ecient manner than
probucol. This consideration is obvious from the almost
constant maintenance of the area of the peaks of the
ApoB.
Sodium hydride, 5-bromo-2,2-dimethylpentanoic acid
isopropyl ester, 2,20-dipyridyldisul®de, triphenyl phos-
phine, and silver perchlorate hydrate are commercially
available materials and did not require further puri®ca-
tion. All solvents were dried and puri®ed by standard
techniques before use. All reactions were run under
nitrogen. Analytical thin-layer chromatography (TLC)
was performed on silica gel 60 F254 plates (Merck).
Compounds were visualized by illumination with ultra-
violet light (254 nm) or with iodine or with 7% phos-
phomolybdic acid in ethanol. Column chromatography
(¯ash) was performed using silica gel (0.040±0.063 mm,
Merck), with the eluents cited in each case.
The activity on lipid synthesis reported in Table 5 show
that lactones 1a±b interfere with the synthesis of cho-
lesterol in a comparable extent to that of gem®brozil
and lower than probucol. Concerning the synthesis of
the esters of the cholesterol, the compounds show an
activity greater than that of the probucol and compar-
able to that of gem®brozil. The data referred to the
other lipid classes emphasize how the two lactones
practically show the same e€ect on the synthesis of fatty
acids and of phospholipids in comparison to both pro-
bucol and gem®brozil whereas the in¯uence on the
synthesis of the triglycerides, on which only the gem®-
brozil shows a remarkable activity, is practically absent.
The preliminary cytotoxicity data, reported in Table 4,
show how the active concentrations of lactones 1a±b
used in the previous experimental model are far away
from ID50 as well as probucol and gem®brozil.
Chemistry
5-[4-Acetoxy-2-(3-oxobutyl)-3,5,6-trimethylphenoxy]-2,2-
dimethylpentanoic acid, isopropyl ester (4a). A solution
of 2a (14.0 g, 0.053 mol) in benzene (240 mL) was slowly
added under stirring to a suspension of 97% sodium
hydride (1.82 g, 0.074 mol) in dimethylsulfoxide
(100 mL). The temperature of the mixture was main-
tained at 30±40 ꢀC with an oil bath. After 20±40 min the
hydrogen evolution ceased and the solution turned
brown-red. A second ¯ask was charged with 5-bromo-
2,2-dimethylpentanoic acid isopropyl ester (13.3 g,
0.052 mol) in benzene (60 mL) and the solution of the
®rst ¯ask was transferred into the second while stirring
by using a 100 mL syringe under a continuous ¯ushing
of nitrogen through both ¯asks. The mixture was stirred
and heated at 65ꢀC for 19 h. The reaction mixture was
cooled to room temperature, diluted with ethyl ether,
and the water layer acidi®ed to pH 5 by means of a 10%
aqueous solution of hydrochloric acid. The layers were
then separated, the aqueous phase was extracted again
with ether, and the extracts were washed with water and
dried over Na2SO4. The evaporation of the solvent
under reduced pressure furnished a residue that was
puri®ed by ¯ash-chromatography on silica gel column
(9/1 methylene chloride/ethyl ether). The evaporation of
the solvent gave compound 4a (13.35 g, 58% yield) pure
as an oil: IR (®lm) n 1757, 1719 (CˆO) cm 1; 1H NMR
(CDCl3) d 1.20 (6H, s, geminal CH3), 1.23 (6H, d,
J=6.2 Hz, 2 isopropyl CH3), 1.71 (4H, m, -CH2CH2C),
2.02 (3H, s, arom. CH3), 2.04 (3H, s, arom. CH3), 2.15
(3H, s, CH3-CO-), 2.16 (3H, s, arom. CH3), 2.34 (3H, s,
CH3COO), 2.54±2.68 (2H, m, -CH2-), 2.78±2.90 (2H, m,
-CH2-), 3.60±3.70 (2H, m, CH2O), 5.00 (1H, m,
J=6.2 Hz, isopropyl CH); 13C NMR (CDCl3) d 12.6
(arom. CH3), 13.1 (arom. CH3), 13.2 (arom. CH3), 20.6
(CH3COO-), 21.7 (-COCH3), 21.9 (OCH(CH3)2), 25.2
((CH3)2CCO), 26.1 (CH2), 29.9 (CH2), 37.1 (CH2), 42.1
(CH2CCO-), 44.0 (CH2CH2CO), 67.6 (isopropyl CH),
73.8 (OCH2), 126.8, 128.2, 128.7, 131.3, 144.9, 153.9
(arom. 6 C), 169.6 (CH3COO-), 177.7 (CCOO-), 208.6
(-CO-); MS m/z 434 (M+), 222, 171, 129, 43.
Conclusions
New 12-membered lactones 1a±b, structurally related to
phenol-induced antioxidant and gem®brozil, have been
prepared by a practical and ecient synthetic sequence.
Compounds 1a and 1b showed good antioxidant prop-
erties in vitro along with a favorable activity on lipid
biosynthesis. These features make this new class of
medium-sized lactones promising dual-action drugs for
anti-atherosclerosis treatments.30
Experimental
Melting points were measured on a Buchi apparatus
1
and were uncorrected. H NMR and 13C NMR spectra
were recorded on a Varian Gemini 200 spectrometer or
a Varian Gemini 300 NMR spectrometer. Data were
reported as follows: chemical shifts in ppm using tetra-
methylsilane (TMS) as internal standard on the d scale,
multiplicity
(s=singlet,
d=doublet,
t=triplet,
q=quartet, m=multiplet), coupling constant(s) in
hertz, and integration (for 1H NMR). Mass spectra
were measured on a VG 7070E mass spectrometer, at a
70 eV ionization voltage and with a 6 kV acceleration
voltage and the data are reported with the base peak
underlined. The IR spectra were obtained on a Perkin±
Elmer model 983/G spectrophotometer and generally,
when not otherwise speci®ed, preparing the specimen in
potassium bromide (KBr) and recording the spectrum
between 4000 and 600 nm.
5-[4-Acetoxy-2-(3-oxopentyl)-3,5,6-trimethylphenoxy]-2,2-
dimethyl pentanoic acid, isopropyl ester (4b). This com-
pound was obtained in the same way as described for
4a, starting from 2b (20.0 g, 0.072 mol in 500 mL of
benzene), 97% sodium hydride (2.07 g, 0.086 mol in
150 mL of DMSO) and 5-bromo-2,2-dimethyl-pentanoic

416
C. Baldazzi et al. / Bioorg. Med. Chem. 7 (1999) 411±418
acid isopropyl ester (21.6 g, 0.086 mol in 100 mL of
benzene). Column ¯ash-chromatography (6/4 petroleum
ether/ethyl ether) gave compound 4b (16.5 g, 51% yield)
as an oil: IR (®lm) n 1758 (CˆO), 1716 (CˆO) cm 1; 1H
NMR (CDCl3) d 0.95 (3H, t, J=7.3Hz, COCH2CH3),
1.12 (6H, s, C(CH3)2), 1.18 (6H, d, J=6.2 Hz, OCH(CH3)2),
1.63 (4H, m, CH2CH2), 1.91 (3H, s, arom. CH3), 1.98
(3H, s, arom. CH3), 2.10 (3H, s, arom. CH3), 2.22 (3H,
s, COCH3), 2.35 (2H, q, J=7.4 Hz, COCH2CH3), 2.51
(2H, m, CH2), 2.80 (2H, m, CH2), 3.58 (2H, m, CH2O),
4.90 (1H, m, J=6.5 Hz, OCH(CH3)2); 13C NMR
(CDCl3) d 8.5 (COCH2CH3), 13.1 (arom. CH3), 13.6
(arom. CH3), 13.7 (arom. CH3), 21.1 (COCH3), 22.35
(OCH(CH3)2), 25.7 (CH3)2CCO), 26.6 (CH2CH2O),
36.5 (CH2), 37.6 (CH2), 42.5 (C(CH3)2), 43.0 (CH2),
67.9 (CH(CH3)2), 74.1 (OCH2), 126.9, 128.3, 128.8,
131.65, 144.9, 154.0 (6 arom. C), 169.7 (CH3COO),
177.7 (COO), 211.3 (CO). MS m/z 448 (M+), 279, 236,
171, 129, 43.
25.2 (C(CH3)2), 26.2 (CH2), 36.2 (CH2), 37.2 (CH2),
42.2 (C(CH3)2), 43.0 (CH2), 73.9 (OCH2), 120.2, 121.6,
128.0, 130.7, 148.6, 149.6 (6 arom. C), 184.3 (COOH),
211.9 (CO); MS m/z 364 (M+), 236, 164, 57.
2,2-Dimethyl-5-[4-hydroxy-2-(3-hydroxybutyl)-3,5,6-tri-
methylphenoxy]pentanoic acid (6a). Sodium borohy-
dride (4.9 g, 0.129 mol) was added portionwise to a
solution of compound 5a (4.2 g, 0.012 mol) in methanol
(125 mL). After stirring 4 h at room temperature, the
reaction mixture was concentrated under reduced pres-
sure, diluted with ethyl ether (250 mL) and a 6 M aqu-
eous solution of hydrochloric acid was slowly added to
destroy the unreacted hydride. The ethereal layer was
separated, washed with water, and dried over Na2SO4.
The evaporation of the solvent at reduced pressure gave
compound 6a (3.93 g, 93% yield) as pure product: mp
44±46 ꢀC; IR (KBr) n 3213 (OH), 1702 (CˆO) cm 1; 1H
NMR (CDCl3) d 1.15 (3H, d, J=6.8 Hz, CH3CH), 1.30
(6H, s, geminal CH3), 1.50±1.90 (6H, m, 3 CH2), 2.10
(3H, s, arom. CH3), 2.18 (6H, s, 2 arom. CH3), 2.65±
2.88 (2H, m, CH2), 3.52±3.72 (3H, m, CH2 and OH),
5.48 (2H, broad s, 2 OH); MS m/z 352 (M+), 224, 206,
164.
2,2-Dimethyl-5-[4-hydroxy-2-(3-oxobutyl)-3,5,6-trimethyl-
phenoxy]pentanoic acid (5a). Compound 4a (20.0 g,
0.046 mol) dissolved in 95% ethanol (100 mL) was
added to potassium hydroxide (22.6 g, 0.40 mol) in 95%
ethanol (300 mL). The reaction mixture was re¯uxed for
18 h. After cooling, it was diluted by addition of water
and acidi®ed to pH 5 with a 10% aqueous solution of
hydrochloric acid. This solution was concentrated under
reduced pressure and extracted with diethyl ether. The
crude product obtained by evaporation of the solvent at
reduced pressure was puri®ed by ®ltration on a short
column of silica gel eluting with diethyl ether. Solvent
removal under reduced pressure gave 5a (10.6 g, 65%
yield) as pure product: mp 108±110 ꢀC (from n-hexane);
2,2-Dimethyl-5-[4-hydroxy-2-(3-hydroxypentyl)-3,5,6-tri-
methylphenoxy]pentanoic acid (6b). This compound was
obtained in the same way as described for 6a, starting
from 5b (8.0 g, 0.022 mol) dissolved in methanol
(240 mL) by treatment with sodium borohydride (9.04 g,
0.24 mol). Work up of the reaction mixture was per-
formed as depicted before after stirring at room tem-
perature for 1 h to give pure 5b (7.49 g, 93% yield): mp
38±42 ꢀC; IR (KBr) n 3499 (OH), 1710 (CˆO) cm 1; 1H
NMR (CDCl3) d 0.9 (3H, t, J=8 Hz, CH2CH3), 1.25
(6H, s, C(CH3)2), 1.4±1.65 (m, 4H, CH2CH2), 1.7±1.9
(m, 4H, CHCH2CH2), 2.15 (s, 9H, 3 arom. CH3), 2.7±
2.85 (2H, m, CH2CH3), 3.3±3.45 (m, 1H, CH), 3.6±3.75
(m, 2H, OCH2), 6.1±7.0 (bs, 3H, 3 OH); 13C NMR
(CDCl3) d 10.4 (CH3), 12.25 (arom. CH3), 12.7 (arom.
CH3), 13.5 (arom. CH3), 23.5 (CH2), 25.45 (CH3), 26.3
(CH2), 30.2 (CH2), 37.2 (CH2), 42.35 (q, C), 72.8 (CH),
74.5 (CH2), 120.5, 121.6, 127.75, 131.3, 149.0, 149.7 (6
arom. C), 183.3 (CO); MS m/z 366 (M+), 348, 322, 238,
220, 165.
1
IR (KBr) n 3499 (OH), 1715 (CˆO) cm 1; H NMR
(CDCl3) d 1.05 (3H, t, J=7.3 Hz, COCH2CH3), 1.25
(6H, s, C(CH3)2), 1.75 (4H, s, 2 CH2), 2.15 (3H, s, arom.
CH3), 2.2 (6H, s, 2 arom. CH3), 2.45 (2H, q, J=7.0 Hz,
COCH2CH3), 2.6 (2H, m, CH2), 2.85 (2H, m, CH2), 3.7
(2H, s, OCH2), 4.9 (2H, broad s, 2 OH); 13C NMR
(CDCl3) d 8.2 (COCH2CH3), 12.1 (arom. CH3), 12.6
(arom. CH3), 13.3 (arom. CH3), 22.0 (COCH2CH3),
25.2 (C(CH3)2), 26.2 (CH2), 36.2 (CH2), 37.2 (CH2),
42.2 (C(CH3)2), 43.0 (CH2), 73.9 (OCH2), 120.2, 121.6,
128.0, 130.7, 148.6, 149.6 (6 arom. C), 184.3 (COOH),
211.9 (CO); MS m/z 364 (M+), 236, 164, 57.
6,6,9,13,15,16-Hexamethyl-14-hydroxy-2,8-dioxabicyclo-
[10.4.0]hexadec-12,14,16-trien-7-one (1a). A solution of
compound 6a (2.0 g, 0.005 mol), 2,20-dipyridyldisul®de
(2.5 g, 0.011 mol) and triphenyl phosphine (3.0 g,
0.011 mol) in anhydrous acetonitrile (60 mL) was
prepared. The resulting mixture was kept at 25 ꢀC for
4 h after which it was diluted with p-xylene (70 mL) and
transferred into a pressure-equalizing dropping funnel
equipped on a three-necked ¯ask in which silver per-
chlorate hydrate (2.74 g, 0.012 mol) in p-xylene (150 mL)
was heated to re¯ux. The solution containing com-
pound 6a was slowly added (1 h) to the re¯uxing per-
chlorate solution. The mixture was re¯uxed for an
additional 18 h, the reaction mixture was then cooled
and the re¯ux condenser was changed with a Claisen
head for distillation. The solution was concentrated by
distilling the solvents at atmospheric pressure. The residue
2,2-Dimethyl-5-[4-hydroxy-2-(3-oxopentyl)-3,5,6-trimethyl-
phenoxy]pentanoic acid (5b). This compound was
obtained in the same way as described for 5a, starting
from 4b (20.0 g, 0.045 mol in 100 mL of 95% of ethanol)
when treated with potassium hydroxide (22.6 g, 0.40 mol
in 300 mL of 95% ethanol) in re¯uxing conditions for
18 h. Solvent removal at reduced pressure gave pure 5b
(11.1 g, 67% yield): mp 108±110 ꢀC (from n-hexane); IR
1
(KBr) n 3499 (OH), 1715 (CˆO) cm
;
1H NMR
(CDCl3) d 1.05 (3H, t, J=7.3 Hz, COCH2CH3), 1.25
(6H, s, C(CH3)2), 1.75 (4H, s, 2 CH2), 2.15 (3H, s, arom.
CH3), 2.2 (6H, s, 2 arom. CH3), 2.45 (2H, q, J=7.0 Hz,
COCH2CH3), 2.6 (2H, m, CH2), 2.85 (2H, m, CH2), 3.7
(2H, s, OCH2), 4.9 (2H, broad s, 2 OH); 13C NMR
(CDCl3) d 8.2 (COCH2CH3), 12.1 (arom. CH3), 12.6
(arom. CH3), 13.3 (arom. CH3), 22.0 (COCH2CH3),

C. Baldazzi et al. / Bioorg. Med. Chem. 7 (1999) 411±418
417
was added to silica gel for chromatography (20 g), col-
lected into a round-bottom ¯ask and the residual sol-
vent of this mixture was evaporated at reduced pressure.
The powder so obtained was added to the head of a
silica gel column previously prepared with 1/1 diethyl
ether/petroleum ether. This mixture was used as eluent
and the chromatographic fractions were monitored by
TLC. The evaporation at reduced pressure of the sol-
vent of the collected fractions gave pure lactone 1a
(0.77 g, 46% yield): mp 127±130 ꢀC; IR (KBr) n 3452
(OH), 1709 (CˆO) cm 1; 1H NMR (CDCl3) d 1.18 (3H,
s, CH3C), 1.22 (3H, d, J=6.4 Hz, CH3CH), 1.27 (3H, s,
CH3C), 1.50±1.90 (5H, m, 2 CH2 and CH2), 2.12 (3H, s,
arom. CH3), 2.14 (3H, s, arom. CH3), 2.16 (3H, s, arom.
CH3), 2.35 (1H, m, CH2), 2.55 (1H, m, CH2), 3.05 (1H,
m, CH2), 3.65 (2H, m, CH2O), 4.50 (1H, s, OH), 4.85
(1H, m, CHO); 13C NMR (CDCl3) d 12.3 (arom. CH3),
12.6 (arom. CH3), 13.6 (arom. CH3), 19.2 (CH3CH),
21.8 (CH2), 23.5 (CH3C), 26.1 (CH2), 28.6 (CH2), 36.0
(CH2), 38.1 (CH2), 42.7 (C), 69.6 (CH), 73.1 (CH2),
120.0, 121.2, 128.1, 131.5, 148.5, 150.2 (6 arom. C),
177.8 (COO); MS m/z 334 (M+) 206, 164, 119, 91.
through a 0.45 mm membrane (Millex HV ®lters) and
the protein concentration determined, according to
Lowry's method.32 The solution of LDL was diluted with
PBS-EDTA to a concentration of 500 mg protein/mL.
Antioxidant activity determination: LDL oxidation. The
LDL solution so obtained (1 mL) was incubated at
42 ꢀC for 45 min in presence either of 1.7 mL of a solu-
tion of lactones at various concentration in ethanol to a
®nal concentration of lactones as reported in Table 1
and ethanol and probucol as standards. At the end of
the incubation the solutions were placed on ice for
45 min and dialysed at 4 ꢀC for 36 h in PBS EDTA-free.
The oxidation of LDL was carried out by the addition
of a cupric sulfate solution at a ®nal concentration of
5 mM, at 37 ꢀC either in the presence of the testing com-
pounds or of ethanol alone. At de®ned time intervals
the reaction was stopped by placing the test tubes at
4 ꢀC after the addition of 0.24 mM EDTA.
TBARS method. The degree of oxidation of LDL was
checked by spectrophotometric evaluation (l=535 nm)
of the condensation products between thiobarbituric
acid and aldehydes formed from oxidative breakdown
of LDL, using malondialdehyde (MDA) as standard,
according to the TBARS method.33 The results of the
spectrophotometric reading, expressed as ng of MDA/
mg LDL protein, were reported as a function of the
oxidation time. The latency interval was determined as
the intercept on abscissas of the tangent to the curve at
the ¯ex point. Whenever the oxidation of LDL was
inhibited over the experimental time we conventionally
indicated a latency period more than 180 min (see
Table 1).
9-Ethyl-14-hydroxy-6,6,13,15,16-pentamethyl-2,8-dioxabi-
cyclo[10.4.0]hexadec-12,14,16-trien-7-one (1b). A solution
of compound 6b (4.00 g, 11.4 mmol), 2,20-dipyr-
idyldisul®de (4.76 g, 21.6 mmol), and triphenylpho-
sphine (5.68 g, 21.6 mmol) in anhydrous acetonitrile
(110 mL) was diluted with p-xylene (120 mL) and added
over 1 h to a boiling solution of silver perchlorate
hydrate (5.2 g, 23.0 mmol) of p-xylene (300 mL). After
12 h of re¯uxing, the crude product was recovered by
evaporation of the solvent as previously described and
puri®ed by column chromatography (1/1 diethyl ether/
petroleum ether). Evaporation at reduced pressure of
the solvent of the collected fractions gave pure lactone
1b (1.80 g, 46% yield): mp 126±128 ꢀC (from ethyl ether
Agarose gel electrophoresis. The electrophoretic mobi-
lity of ApoB was measured applying LDL to a 0.6%
agarose gel in barbital/sodium barbital bu€er pH 8.4±
8.8 (Helena Laboratories kit `Titan Gel Multi-slot lipo-
17 electrophoresis system') and using the Bio-Phoresis
Horizontal Electrophoresis Cell (BIO-RAD). The native
LDL solutions (2 mL) and samples of copper-oxidated
LDL at incubation times reported were applied on the
agarose gel plate and 7 min were waited before running.
The electrophoretic conditions were: the application
point was on the cathodic side (Ð), voltage 60 V, elec-
trophoresis time 40 min, temperature 15 ꢀC. At the end
of the electrophoresis period, the plate was dried with
air at 60±70 ꢀC for 10 min and placed in the lipoprotein
staining solution (0.1% (w/v) Fat Red 7B in 95%
methanol) for 2 min. The plate was destained in two
consecutive washes of destaining solution (75% (v/v)
methanol in water) for 15 s and dried at 60±70 ꢀC for
5 min. The migration of native LDL was taken as equal
to 0 while the migration of copper-oxidized LDL at
360 min of oxidation in the presence of ethanol was
taken as 100. The electrophoretic mobility of copper-
oxidized LDL, in the presence of testing compounds,
was expressed as relative mobility (REM) at 180 and
360 min.
and n-hexane); IR (KBr) n 3448 (OH), 1699 (CˆO)
1
cm
;
1H NMR (CDCl3) d 0.9 (t, J=8 Hz, 3H,
CH2CH3), 1.2 (s, 3H, CH3), 1.3 (s, 3H, CH3), 1.4±1.9
(m, 7H, CH2CH3, CH2CH2 1H CH2CH2CH), 2.15 (s,
9H, 3 arom. CH3), 2.35±2.6 (m, 2H, CH2CH), 2.9±3.1
(m, 1H, 1H CH2CH2CH), 3.5±3.8 (m, 2H, CH2), 4.55 (s,
1H, OH), 4.75±4.9 (m, 1H, CH); 13C NMR (CDCl3) d
10.4 (CH3), 12.3 (arom. CH3), 12.7 (arom. CH3), 13.7
(arom. CH3), 21.5 (CH2), 23.5 (CH3), 26.6 (2 CH2), 28.6
(CH3), 33.7 (CH2), 37.3 (CH2), 42.7 (q, C), 72.8
(CH2), 74.9 (CH), 120.0, 121.2, 128.1, 131.9, 148.4,
150.3 (6 arom. C), 178.2 (CO); MS m/z 348 (M+), 220,
165.
Biological assays
Native LDL isolation from human plasma. Human
plasma was obtained from blood collected, in presence
of EDTA, according to the Havey's method,31 from
healthy volunteers fasting for 8 h. LDL were isolated
from the plasma by sequential ultracentrifugation at
densities between 1.006 and 1.063 g/mL in a Sorval
OTD 65 B ultracentrifuge for 20 h at 55,000 rpm and at
4 ꢀC. LDL were dialysed at 4 ꢀC against 0.01 M phos-
phate bu€er (pH 7.4) and 0.15 M NaCl containing
EDTA (0.24 mM) (PBS-EDTA solution), ®ltered
SDS-polyacrylamide gel electrophoresis. LDL were dis-
solved in sample bu€er containing 48% 0.7 M SDS,

418
C. Baldazzi et al. / Bioorg. Med. Chem. 7 (1999) 411±418
20% 2-mercaptoethanol, 12% 1 M TRIS pH 6.8, 20%
glycerol by incubating in boiling water for 4 min. Ver-
tical gel electrophoresis was performed according to the
method of Laemmli24 using a 5% acryilamide gel at a
constant voltage of 30 V for 4.5 h in a BioRad Protean
II cell with tap water cooling. Gels were ®xed in 10%
acetic acid in methanol/water 1/1, stained with Coomassie
Brilliant Blue G250 and subjected to densitometric
scanning by LKB Mod. 2222-020 Ultrascan XL with
integrators LKB 2220 and LKB 2190 GelScan software
package.
4. Steinberg, D. Arteriosclerosis 1983, 3, 283.
5. Steinberg, D.; Parthasarathy, S.; Carew, T. E.; Khoo, J. C.;
Witztum, J. L. N. Engl. J. Med. 1989, 320, 915.
6. Esterbauer, H.; Waeg, G.; Puhl. H. Br. Med. Bull. 1993, 49,
566.
7. D. Steinberg showed the presence of antibodies anti-OX-
LDL and of OX-LDL, oxidized low density lipoproteins, in
the atheromatous lesions of patients su€ering from advanced
atherosclerosis.5 Moreover, OX-LDL appeared toxic towards
the endothelial cells and a pro-in¯ammatory activity towards
monocytes and lymphocytes and an increased susceptibility to
the taking from the macrophages and the smooth muscle cells
with formation of foam cells was showed.8 Such properties can
have an important part in the processes of formation and
progression of the atheromatous lesion.9
In vitro cytotoxic activity
8. Witztum, J. L.; Steinberg. D. J. Clin. Invest. 1991, 88, 1785.
9. Ross, R. Nature 1993, 362, 801.
10. Burton, G. W.; Ingold, K. U. Acc. Chem. Res. 1989, 19,
194.
11. Hall, E. D.; Braughler, J. M. Free Radical Biol. Med.
1989, 6, 303.
12. Van Acker, S. A. B. E.; Koymans, L. M. H.; Bast, A. Free
Radical Biol. Med. 1993, 15, 311.
13. Carew, T. E.; Schwenke, D. C.; Steinberg, D. Proc. Natl.
Acad. Sci. USA 1987, 84, 7725.
14. Esterbauer, H.; Dieber-Rotheneder, M.; Striegl, G.;
Waeg, G. Am. J. Clin. Nutr. 1991, 53, 314S.
15. Kita, T.; Nagano, Y.; Yokode, M.; Ishii, K.; Kume, N.;
Ooshima, A.; Yoshida, H.; Kawai, C. Proc. Natl. Acad. Sci.
USA 1987, 84, 5928.
16. Rimm, E. B.; Stampfer, M. J.; Ascherio, A.; Giovannucci,
E.; Colditz, G. A.; Willet, W. C. New Engl. J. Med. 1993, 328,
1450.
The con¯uent Hep-G2 cells were exposed to the pro-
ducts tested or reference compounds, dissolved in
DMSO or ethanol (0.01% v/v versus incubation med-
ium volume) as reported in Table 4, for 24 h in Dulbec-
co's Modi®ed Eagle Medium (DMEM), containing 5%
of fetal calf serum. At the end of incubation, the med-
ium was removed and the monolayer was washed twice
with PBS-phosphate bu€ered saline and ®xed twice with
60% methanol at 4 ꢀC for 30 min. The dried monolayer
was solubilized with NaOH 0.3 M for 12 h at 25 ꢀC and
the optical density of the lysate at 260 nm was mea-
sured. Calibration curve was obtained with suspended
Hep-G2 cells culture containing di€erent amount of
cells. DMSO and ethanol at experimental concentration
did not express any toxicity.
17. Stampfer, M. J.; Hennekens, C. H.; Manson, J. E.; Colditz,
G. A.; Rosner, B.; Willet, W. C. New Engl. J. Med. 1993, 328, 1444.
18. Generally, the best reductions of the plasma LDL have
been obtained by means of the resins sequestering the bile
acids and with the statins, inhibitors of the HMGCoA reduc-
tase, while the plasma levels of the triglycerides are mostly
a€ected by the treatment with nicotinic acid and ®brates. Bet-
teridge, D. J. Curr. Opin. Lipidol. 1993, 4, 49.
19. Scott, J. W.; Bizzarro, F. T.; Parrish, D. R.; Saucy, G.
Helv. Chim. Acta 1976, 59, 290.
20. Gerlach, H.; Thalmann, A. Helv. Chim. Acta 1974, 57, 2661.
21. Corey, E. J.; Nicolaou, K. C. J. Am. Chem. Soc. 1974, 96,
5614.
22. Steinbrecher, U. P. J. Biol. Chem. 1987, 262, 3603.
23. Vanderyse, L.; Dvreese, A. M.; Baert, J.; Vanloo, B.; Lins, L.;
Ruysschaert, J. M.; Rosseneu, M. Atherosclerosis 1992, 97,
187.
E€ect on lipid synthesis in Hep-G2 cells
Hep-G2 cells monolayers were incubated for 24 h
according to the method of Barnhard27 in the presence
of tested products and reference compounds at the con-
centrations reported in Table 5. Hep-G2 cells were
incubated for 3 h in the presence of 14C-acetate. After
removing medium and washing, the lipids were extracted
twice with heptane±isopropyl alcohol mixture and the
solutions evaporated. The radioactivity incorporated in
the main lipid classes was measured after chromato-
graphic separation29 on silica gel (Merck F254) and the
quantifying radioactivity by liquid scintillation counting.28
Acknowledgements
24. Laemmli, U. K. Nature 1970, 227, 680.
25. Fong, L. G.; Parthasarathy, S.; Witsum, J. L.; Steinberg,
D. J. Lipid Res. 1987, 28, 1466.
26. Cingi, R.; De Angelis, I.; Fortunati, E.; Reggiani, D.;
Bianchi, V.; Tiozzo, R.; Zucco, F. Toxicol. in vitro 1991, 5,
119.
27. Barnhard, J. W.; Li, D. L. Am. J. Cardiol. 1988, 62, 52B.
28. Gherardi, E.; Calandra, S. Lipids 1980, 15, 108.
29. Skipski, V. P.; Smolowe, A. F.; Sullivan, M.; Barklay, J.
Biochem. Biophys. Acta 1965, 106, 386.
30. Rosini, G.; Baldazzi, C.; Romagnoli, E.; Saguatti, S.;
Piani, S. Eur. Patent Appl. 1996, EP 718295 A1; US Patent
5,608,084, 1997; Chem Abs. 1996, 125, 114726.
31. Havel, R. J.; Eder, H. A.; Bragdon, J. H. J. Clin. Invest.
1955, 34, 1345.
This research was supported by two Alfa Wassermann
S.p.A. Postdoctoral Fellowships to F.C. and F.V. and
by research grants from the Ministero dell'Universita e
della Ricerca Scienti®ca (MURST), Italy, the Consiglio
Nazionale delle Ricerche (CNR), Italy, and by the Uni-
versity of Bologna (Progetto d'Ateneo: Biomodulatori
Organici: Sintesi, Proprieta ed Applicazioni). We wish
to thank Cristina Carrara and Dr. Eleonora Romagnoli
for technical assistance and Dr. Miriam Barbanti for
useful discussions.
References and Notes
32. Lowry, O. H.; Rosebrough, N. J.; Farr, A. L.; Randall, R.
J. J. Biol. Chem. 1951, 193, 265.
33. Fleisher, S. In Methods in Enzymology; Packer, L., Ed.;
New York Press: New York, 1978; Vol. 52.
1. Ross, R. N. Engl. J. Med. 1986, 314, 488.
2. Ross, R. Am. J. Pathol. 1993, 143, 987.
3. Steinberg, D. JAMA 1985, 253, 2080.