Molecular epidemiology of atherosclerosis. FASEBJ. 11, 1021-1031. (1997). Key. Wonis: aging dyslipidosis cigarette smoke hyperten- sion - smooth muscle.
Molecular
epidemiology
of atherosclerosis
SILVIO DE FLORA,*’ ALBERTO IZZOTFI,* DEBRA GIAN LUIGI PETRILLI,1 AND JOF!.I.F.N LEWTAS’2 *Institute of Hygiene and Preventive Medicine, University
WALSH,t
PAOLO
DEGAN,
of Genoa,
Italy; tEpidemiology
and
Biomarker Branch, U.S. Environmental Protection Agency, Research Triangle Park, North Carolina 27711, USA; 1National Institute for Cancer Research, Genoa, Italy; 1Division of Vascular Surgery, Galliera Hospital, Genoa, Italy It has been hypothesized that mutational events may be involved in the atherogenetic process and that at least a portion of atherosclerotic plaques may develop according to an initiation-promotion process of arterial smooth muscle cells, akin to benign tumors. We conducted a study to evaluate the occurrence of oxidative DNA damage and format.ion of DNA adducts in human atherosclerotic lesions and to assess the relationships of these promutagenic alterations with exposure to atherogemc risk factors. Pure DNA was extracted from the tunica media (composed mainly of smooth muscle cells) of abdominal aorta fragments taken at surgery from 85 patients suffering from severe atherosclerotic lesions. DNA adducts were detected by synchronous fluorescence spectrophotometry and 32P postlabeling after enrichment of adducts with either butanol or nuclease P1. 8-Hydroxy-2’-deoxyguanosine (8-OH-dG), a typical indicator of oxidative DNA damage, was measured by HPLC/electrochemical detection. A complete questionnaire reporting genera!, clinical, and laboratory characteristics was available for each patient. All 84 samples tested by P postiabeling were positive by displaying the presence of diagonal radioactive zones and up to 9 individual DNA adducts. Of 52 samples tested, 32 (6 1.5%) yielded typical positive signals at synchronous fluorescence spectrophotometry. All but one of 39 sampies tested had very high levels of 8-OH-dG, thus showing a remarkable oxidative DNA damage. Statistically significant correlations were found between the levels of molecular biomarkers and atherogenic risk factors including age, number of currently smoked cigarettes, ratio of total-to-high density lipoprotein blood cholesterol, blood triglycerides, and blood pressure. The DNA alterations detected in our study may be only one component of the genetic basis of atherogenesis. Moreover, no causal role in the atherogenetic process can be inferred from our results. However, DNA alterations, including oxidative damage and adduction of reactive molecules of either endogenous or exogenous source, were systematically present in the smooth muscle cells of human atherosclerotic lesions and their intensity was significantly correlated with the occurrence of atherogemc risk
ABSTRACT
0892-6638/97/001
1-1021/$02.25
© FASEB
factors in the patients studied.-De Flora, S., Lzzotti, A., Walsh, D., Degan, P., Petrilli, G. L., Lewtas, J. Molecular epidemiology of atherosclerosis. FASEBJ. 11, 1021-1031 (1997) Wonis: aging dyslipidosis cigarette sion - smooth muscle cells DNA adducts damage
smoke hypertenoxidative DNA
Key
IS THE LEADING CAUSE of death in the United States, Europe, andJapan (1) as the main source of cerebral and myocardial infarction, gangrene of the extremities, and loss of function in organs affected by clogging and hardening of arteries (2). Atherogenesis is a degenerative process involving a variety of lesions of the arterial wall. It results from focal intimal thickening formed after endothelial cell injury and uncontrolled proliferation of smooth muscle cells (SMCs) accompanied by the accumulation of extracellular components and by the participation of inflammatory cells (3). According to the responseto-injury hypothesis, atherosclerosis is an excessive inflammatory-fibroproliferative response to various forms of insult to the artery wall (2, 4). Indeed, the structure of atherogenic lesions is quite diversified, which may reflect the composite mechanisms involved in its pathogenesis. Some lines of evidence support the view that at least part of atherogenic lesions may be initiated by mutational events in arterial SMCs and that the plaques may develop according to an initiation-promotion process, akin to a benign tumor (5, 6). It has been suggested that human atherosclerotic plaques often display a
ATHEROSCLEROSIS
,‘
Correspondence:
Institute
and Preventive
of Hygiene
icine, University of Genoa, via A. Pastore Italy. 2 Present address: U.S. Environmental OEA-095, 1200 5Abbreviations:
1, 1-16132 Protection
Med-
Genoa, Agency,
Sixth Ave., Seattle, WA 98101, USA. SMC,
smooth
muscle
cell; SFS, synchronous
fluorescence speetrophotometry; DRZ, diagonal radioactive zones; PAl-I, polyeyelie aromatic hydrocarbons; 8-OH-dG, 8hydroxy-2’-deoxyguanosine; HPLC, high-performance liquid chromatography:
troehemical poproteins;
TLC,
thin-layer
chromatography;
detection; I, indigenous; HDL, LDL, low density lipoproteins.
ECD,
high
elee-
density
Ii-
1021
clonal character, being the progeny of a single SMC (7), and the DNA extracted from both human (8) and cockerel (9) atherosclerotic plaques has transforming ability when transfected into cultured mammalian cells. The proliferation rate of SMCs is normally low (10), but can be stimulated by factors such as platelet-derived growth factors (11) and iipoprotein(a) (12). Moreover, mutations in SMCs could induce the constitutive production of growth factors resulting in autocrine stimulation of growth (3). Mitogens and chemotactic agents can also be released during regenerative repair of injured SMCs (3). The recent discovery of intermediate biomarkers has partially decoded the “black box,” which represents the latency period elapsing between exposure to risk factors and clinical onset of chronic degenerative diseases. The study of these biomarkers has encountered a broad application in cancer research, and warrants interest in the molecular epidemiology of other chronic degenerative diseases as well, including atherosclerosis (13, 14). The formation of DNA adducts, which are products of the covalent reaction of electrophilic molecules with DNA, and the occurrence of oxidative damage to DNA are molecular alterations that may be involved in the pathogenesis of those chronic degenerative diseases in which DNA is the critical target (13). Few studies have investigated the presence of DNA adducts in arteries of experimental animals or exposed humans. DNA adducts were detected by synchronous fluorescence spectrophotometry (SFS) in the thoracic aorta of Sprague-Dawley rats exposed whole body for 40 consecutive days to mainstream cigarette smoke. These molecular lesions were significantly inhibited by oral administration of the chemopreventive agent Nacetylcysteine (15). The DNA extracted from the whole ascending aorta taken at autopsy from two smokers was positive according to postlabeling analysis by showing the presence of two diagonal radioactive zones (DRZ) and of individual DNA adducts (16). This technique detects the presence of various DNA adducts with extremely high sensitivity, in the range of 1 adduct per iO-i0’#{176}nucleotides, depending on the structure of the DNA adduct. Using the same technique, the presence of adducts was investigated in the DNA of cardiopulmonary tissues taken at autopsy in the Silesia region of the Czech Republic, which is characterized by a relatively high exposure to polycyclic aromatic hydrocarbons (PAH) due to the presence of many coke ovens and ironworks. As preliminarily reported (17), DNA adducts were detected in whole aorta, right and left coronary arteries, great cardiac vein, right and left pulmonary arteries, pulmonary vein, pulmonary trunk, myocardium, and endocardium. A pilot study independent of the one described in this paper dealt with the detection of DNA adducts in the tunica me-
1022
Vol.11
October
1997
dia of abdominal aorta, affected by atherosclerotic lesions, taken at surgery from seven patients. DNA adducts were detected by SFS, 32P postlabeling, and high-performance liquid chromatography (HPLC) / fluorescence (13). The present study was undertaken with the aim of evaluating the frequency and patterns of DNA adduct occurrence, as detected by different procedures, in a large group of atherosclerotic subjects. 8-Hydroxy-2’deoxyguanosine (8-OH-clG), the most typical indicator of oxidative DNA damage, was also measured in a subset of samples. In the obvious unavailability of comparable tissue samples from healthy, nonatherosclerotic subjects, the intensity of these promutagenic DNA lesions was related to exposure of the patients under study to known atherogenic risk factors. This led to the discovery that statistically significant correlations exist between these molecular biomarkers and age, smoking habits, blood dyslipidosis, and blood pressure of atherosclerotic patients.
MATERIALS
AND
METHODS
Patients A fragment sions was
Galliera
of abdominal
aorta
affected
by atherosclerotic
le-
removed at surgery for therapeutic purposes at the Hospital in Genoa from 98 atherosclerotic patients
undergoing aortic graft due lesions of the subrenal aorta.
to severe Complete
able for each patient concerning
aneurism information
or stenotic was avail-
age, sex, place of residence,
occupation, family history for atherosclerosis, body weight, mass index, physical activity, stress conditions, diabetes, dietary habits, alcohol consumption, smoking habits, arterial pressure, severity of the atherosclerotic disease, blood cholesterol, triglycerides, and other clinical and laboratory data. Processing
of
aorta
fragments
Immediately after removal, the aorta fragments were immersed in a sterile physiological saline solution, stored at 4#{176}C, and transferred from the Vascular Surgery Department to the University of Genoa laboratory. Within 1-2 h of surgical removal, each fragment was examined and classified according to its macroseopical appearance and carefully cleaned of caleifieations, fatty deposits, and thrombotic materials. After washing with physiological saline to remove debris and blood residues, the three artery layers were dissected and separated as accurately as possible (Fig. I). The adventitial coat was removed, whereas the tunica media and tunica intima were stored separately at -80#{176}C. Due to the severity of alterations of the intima, it was not possible to extract DNA in sufficient amounts for molecular dosimetry analyses.
DNA extraction from atherosclerotic
lesions
The tunica media, which is composed mainly of SMCs (as confirmed by histological analyses) (Fig. 1), was used for DNA extraction. Within 1 month of storage at -80#{176}C,the tuniea media was homogenized in a Polytron apparatus at 4#{176}C in 250 mM sucrose, 50 mM Tris HC1 (pH 7.4). DNA was isolated by solvent extraction with an automatic DNA extractor (Gene-
The FASEB Journal
DE FLORA ET AL.
ABDOMINAL
TUNICA MEDIA
AORTA
‘I, PURIFIED
DNA
DNA ADDUCTS
32p POSTLABELING
Butanol
Figure
1. Flow
diagram
indicates sponding
extraction
OXIDATIVE DAMAGE
?,(nm)
of the experimental
dosimetry
techniques
part
are also reported.
341, Applied Biosystems, Foster dure for DNA extraction was basically
City, Calif.).
tractions
minimize
pure
of the study.
Elutlon time (mm)
of positive results obtained by testing aorta DNA 2P postlabeling (butanol enrichment) showing the pattern of autoradiographic spots (see Fig. 2); SFS or synchronous fluorescence speetrophotometry (the X-axis the emission wavelength); and separation by HPLC and detection of the 8-OH-dG peak by ECD and of peaks correto normal nueleotides by UV detection.
samples by molecular distribution
SFS
The proce-
that described by Gupta (18), except that the homogenized tissue was sequentially treated first with a mixture of RNase A and TI for 1 h and then with proteinase K for 5 h at 55#{176}C, followed by two extractions with a phenol-water-chloroform mixture and two ex-
with
formation
chloroform
only.
To
of 8-OH-dG during extraction,
the
possible
the phenol used was
of the highest available quality and was never used alone. Moreover, all steps were carried out in a helium atmosphere. The material was further washed with water-saturated isoamyl alcohol in order to remove noneovalently DNA-bound PAH residues (19) and was precipitated by ethyl alcohol. The extracted DNA was quantified by spectrophotometric analysis. In 85 of the 98 tunica media samples, the amount of recovered DNA ranged between 0.04 and 1.72 i.tg/lOO mg wet tissue (mean ± SD: 0.33±0.31), as detected by measuring the absorbanee at 260 nm. The ranges of 230/260 and 260/280 absorbanee ratios were 0.22-0.54 and 1.62-1.97, respectively, and the means ± SD were 0.43 ± 0.07 and 1.79 ± 0.67, respectively. In the remaining 13 samples, the amount of DNA and/or its
MOLECULAR
EPIDEMIOLOGY
OF ATHEROSCLEROSIS
Typical
examples
They include
purity was not stifficient to be reliably ters under study. The 85 DNA samples
for the paramelyophilized and, depending on the amount of DNA available, the material was tested for at least one of the molecular dosimetry analyses according to the general scheme depicted in Fig. 1.
Evaluation
tested were
of DNA adducts by ‘2P postlabeling
analyses were conducted in the Research Triangle Park (NC.) laboratory, as described previously (20). Briefly, 3.0 ig of aorta DNA were enzymatically digested and enriched with either nuelease P1 or butanol. The digests enriched for addueted nucleotides were then labeled for 30 mm with approximately 50 IlCi [‘2PjATP at 3000 Ci/mmol (Amersham, Ai-lington Heights, Ill.). The radiolableled digest was applied to polyethyleneimine-cellulose plates prepared according to the method of Gupta et al. (21) for multidirectional thin-layer chromatography (TLC). The solvents were as follows: DI, 1 M sodium phosphate (pH 6.8); D3, 4 M lithium formate, 7 M urea (pH 3.45), with predevelopment to 1 cm 2P postlabeling
1023
TABLE 1. Occurrence of DNA adducts and oxidative at surgery from atherosclerotic patients 32P postlabeled (adducts/lO’
Aorta sample (code)
Nuclease
1 2 3 4 5 6 7
6.7 9.7 8.4 8.5 5.0 7.4 7.7
8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 35 37 38 39 40
P1
damage to the DNA of smooth muscle cells from 85 samples of abdominal
aorta taken
DNA adducts nucleotides) Butanol
SFS-positive DNA adducts (fluorescence units)
Oxidative DNA damage (mol 8-OH-dG/105 mol dG)
NT#{176} NT NT NT 5.7 11.4 10.3
48.3 33.3 16.7 26.7 23.3 20.0 17.0
3.1 NT 1000.0 21.2 163.4 12.0 50.3 501.0 65.5 98.0 NT 900.0 72.1 NT 16.7
64 65 68 70 71 75 77 78 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94
17.1 4.4 68.4 21.8 12.0 15.3 60.3 38.8 43.5 17.7 47.1 15.2 15.8 20.7 6.0 10.7 42.1 27.7 NT 16.0 8.7 16.3 5.9
97
81.6
76.2
NT
98
42.4
31.7
NT
‘NT,
NT >1000.0
not tested.
with
2.5 M ammonium formate (pH 3.5); D4, 0.8 M lithium 7 M urea, 0.5 M Tris-HC1 (pH 8.0), with predevelto 1 cm with 0.5 M Tris-HCI (pH 8.0); D5, same direction as D4 using I M sodium phosphate (pH 6.8). Intensitring screen-enhanced autoradiography at -80#{176}Cfor 24-72 h was used to detect the presence of radiolabeled adduets on the TLC plates. An example of a positive aorta DNA sample is shown in Fig. 1. The individual adduets were carefully excised, covered in 10 ml of ethanol, and counted for radioactivity in a beta counter (model CA1900, Packard, Downers Grove, Ill.). DNA adduct standards were analyzed in each experiment to ensure proper enhancement and labeling ef-
chloride, opment
ficiencies.
A benzo(a)pyrene
DNA adduct
was used for all nu-
clease P1 samples and a 4-aminobiphenyl DNA adduet was used for butanol-extracted samples. Calf thymus DNA was used as well as reagent blank for the negative controls in each experiment. Individual normal nucleotides were determined for each sample in each experiment. The relative adduet levels were calculated by using the representative average normal nueleotide for all samples. Evaluation
of DNA adducts by synchronous
fluorescence
spectrophotometry SFS analyses were carried out in the University of Genoa laboratory on 52 DNA samples, as described in previous studies (22, 23). In brief, a standardized amount of 50 tg DNA for each sample was hydrolyzed in 0.1 N HCI at 90#{176}C for 4 h in sealed glass vials. Synchronous scanning was performed by using a Hitachi F-3000 fluorescence speetrophotometer with a fixed A) of 34 nm between excitation and emission. The intensity of the fluorescence peak was expressed in fluorescence units, which indicate the difference between the intensity of
MOLECULAR
EPIDEMIOLOGY
OF ATHEROSCLEROSIS
the signal corresponding to the peak and the baseline. This was further subtracted by a fixed value of 10, which was assumed as an arbitrary threshold of sensitivity of the method, in order to distinguish specific fluorescence peaks from aspecifie baseline fluctuations. A sample of 50 tg calf thymus DNA (Sigma Chemical Co., St. Louis, Mo.), with or without the addition of 300 fmol benzo(a)pyrene tetrol (National Cancer Institute, Chemical Carcinogen Repository, Kansas City, Mo.), was also analyzed as a control. A typical fluorescence spectrum for a positive aorta DNA sample is shown in Fig. 1.
Evaluation 8-OH-dG Institute
of oxidative
to DNA
was quantified in 39 DNA samples at the for Cancer Research laboratory in Genoa.
of 20 .Lg DNA 1.3 U nuclease
phatase, tubes
damage
described
(24).
Aliquots
from each aorta sample were hydrolyzed P1 for I h at 37#{176}C and with 2 U alkaline
filtered
(SpinX,
National
through
Costar),
0.22 im membrane and
Unmodified
analyzed
nucleoudes
by HPLC
were
with phos-
in microfuge as previously
quantified
by
data from UV detection using calf thymus DNA as a reference, and 8-OH-dG was quantified by an eleetroehemical detector (ECD) calibrated with an authentic standard (25). A typical chromatogram obtained with a standard DNA sample showing the UV as well as the ECD elution patterns is shown in Fig. 1.
RESULTS Sufficient at least
amounts of pure DNA to be analyzed one molecular parameter could be
for ex-
1025
32PPOSTLABELED DNA
ADDUCrS/102
NUCLEOTIDES
MIN
ADDUCr 2 ADDUCT 3 ADDUCr 4 ADDUCF 5
ADoucr 6 ADDUCT 7 Aonucr 8 ADDUCt 9
2.60 4.80
-
BUTANOL
(%)
MIN
-
100.0
5.70
-
MAx
- 81.60 - 51.30
100.0
5.40
3.70
96.3
