Carcinogenesis vol.20 no.7 pp.1345–1350, 1999
Detection of 1,N 2-propanodeoxyguanosine adducts of 2-hexenal in
organs of Fischer 344 rats by a 32P-post-labeling technique
Detlef Schuler and Erwin Eder1
Department of Toxicology, University of Würzburg, Versbacher Str. 9,
D-97078 Würzburg, Germany
1To whom correspondence should be addressed
Email: [email protected]
2-Hexenal is an α,β-unsaturated carbonyl compound which
is mutagenic, genotoxic and forms cyclic 1,N2-propanodeoxyguanosine adducts like similar propenals for which
carcinogenicity was shown, e.g. acrolein or crotonaldehyde.
Since humans have a permanent intake of 2-hexenal via
vegetarian food this genotoxic compound is considered to
play a role in human carcinogenicity. The data base is,
however, presently not sufficient for a cancer risk assessment. To date no long term carcinogenicity study on
2-hexenal has been published. Detection of respective DNA
adducts of this substance in animals or humans could allow
cancer risk assessment. Therefore, we have developed a
32P-post-labeling technique based on nuclease P1 enrichment and TLC separation of the labeled adducts. The
respective adducts are stable over a wide pH range from
pH 4 to pH 11 and relatively stable against nuclease P1.
The detection limit was 0.03 adducts per 106 nucleotides
and the recovery was 10%. With this method we have
shown in vivo formation of 1,N 2-propanodeoxyguanosine
adducts of 2-hexenal for the first time and found the
respective DNA adducts in different organs of Fischer 344
rats after gavage of 500, 200 and 50 mg 2-hexenal/kg body
wt. No adducts could be detected in the organs of untreated
rats. There is a clear dependence of the adduct level and
the CBI (covalent binding index) on the dose. The CBI of
2-hexenal calculated on the basis of our adduct levels is
extremely low (0.06). Since intake of 2-hexenal via fruit
and vegetables is very low the cancer risk from 2-hexenal
intake via food must also be considered as very low
according to a first raw estimation on the basis of CBI and
intake. The situation deserves, however, a more precise
risk assessment in the future.
Introduction
2-Hexenal (leaf aldehyde) is a β-propyl substituted propenal
(CH3CH2CH2CH5CHCHO) which can react with both
functional groups, the activated double bond and the carbonyl
group and form exocyclic 1,N2-propanodeoxyguanosine
adducts (1). This type of adduct is considered to be a
promutagenic lesion (2,3) and α,β-unsaturated carbonyl
compounds forming such adducts are mutagenic (4,5), genoAbbreviations: 39-dGp, deoxyguanosine-39-monophosphate; CBI, covalent
binding index; dG, deoxyguanosine; F344 rats, Fischer 344 rats; Hex-dGp,
deoxyguanosine 39-monophosphate adduct of 2-hexenal (for structure see
Figure 1c, for IUPAC nomenclature and structural features see Results);
Hex-pdGp, deoxyguanosine 39-,59-bisphosphate adduct of 2-hexenal; PEI,
polyethylene imine; RP, reverse phase.
© Oxford University Press
toxic (6,7) and some have been shown to be carcinogenic
(8,9). Like other α,β-unsaturated carbonyl compounds, 2hexenal is mutagenic, e.g. in Salmonella typhimurium strain
TA104 (4) and strain TA100 (5), and in Chinese hamster V79
cells (10). It is genotoxic in the unscheduled DNA-synthesis
(UDS) test (11), induces DNA strand breaks in mouse L1210cells (12) and in human Namalva cells (13), leads to formation
of micronuclei and to sister chromatid exchange (14).
2-Hexenal is cytotoxic in cell culture studies; it decreases
the Ca21 deposits in liver cells (15), and inhibits the microsomal glucose-6-phosphatase activity in liver cells of rats (16).
The LD50 is 1.24–2.36 g/kg body wt for CFW mice and 0.49–
1.58 g/kg body wt for CFE rats (17), both after p.o. application.
2-Hexenal is a common component of plants (18–24) and
the main intake for humans is considered to result from
consumption of fruit, vegetables and fruit juices, but also from
beverages like tea (25 p.p.m.) (22). In bananas, for instance,
the 2-hexenal concentration is 40 p.p.m. (20) and in beans it
is 34 p.p.m. (21). 2-Hexenal is considered to act as a natural
fungicide and protects plants not only from penetration by
fungi (19,25) but also by other micro-organisms (26) after
being injured. Evidently, the local 2-hexenal concentration can
be very high at the site of injury and the concentration given
above represents only a mean concentration over all plant
parts. De Lumen et al. (21) and Stone et al. (27) provided
mechanisms for the formation of 2-hexenal from unsaturated
fatty acids, e.g. linolenic acid by lipoxygenase, under air.
Furthermore 2-hexenal is a plant flavouring component
responsible for the typical aroma of some fruits and herbs
(21–27). The utilization of 2-hexenal as ‘natural identical’
flavouring or as ‘natural’ fungicide was taken into consideration
(28). Furthermore, breeding of plants resistant against fungi
or genetic manipulation could also lead to increased 2-hexenal
concentrations in those plants.
All data available suggest that humans are permanently
exposed to 2-hexenal and that 2-hexenal plays a role in
human carcinogenicity according to its mutagenic/genotoxic
activities. Unfortunately, no data from cancer studies are
available on 2-hexenal. Therefore, cancer risk assessment for
this compound was hitherto not possible.
In our previous in vitro studies (1) we have investigated the
interaction of this compound with DNA components in detail
and have shown that a pair of diastereomers of the regioisomer
of exocyclic 1,N2-propanodeoxyguanosine adducts is formed
as the major stable adduct in which the OH-group is adjacent
to the N1-atom of the guanine moiety. In general, this type of
DNA adduct is considered as a promutagenic lesion (see
above) and can lead to initiation of cancer cells. Detection and
quantitation of these adducts as markers for the initiation of
cancer could decisively improve the cancer risk assessment.
The in vivo formation of the exocyclic DNA adducts of 2hexenal had not previously been demonstrated (29). We have
developed a new 32P-post-labeling technique based on nuclease
P1 enrichment and TLC separation of the DNA adduct spots
1345
D.Schuler and E.Eder
in order to detect and quantitate the respective DNA adducts
of 2-hexenal in vivo and to allow improved cancer risk
assessment.
Materials and methods
Chemicals and reagents
Caution: 2-Hexenal is irritating and genotoxic. trans-2-Hexenal (99%) was
purchased from Aldrich (Steinheim, Germany). Deoxyguanosine (dG) and
deoxyguanosine-39-monophosphate (39-dGp) were obtained from Sigma
(Deisenhofen, Germany). Nuclease P1 from Penicillium citrinum was purchased from Boehringer (Mannheim, Germany). [γ-32P]ATP (4500 Ci/mmol,
10 µCi/µl) and T4 polynucleotide kinase (30 U/µl) were obtained from USB
(Amersham, Braunschweig, Germany). Polyethylene imine (PEI)–cellulose
TLC plates were purchased from Macherey & Nagel (Düren, Germany). RP18
TLC plates were obtained from Merck (Darmstadt, Germany).
Synthesis of Hex-dGp adducts
Aliquots of 30 mg (0.077 mM) 39-dGp were reacted with 50 µl (0.43 mM)
2-hexenal in 1 ml 0.1 M sodium phosphate buffer (pH 10.7) at room
temperature. Maximum yield was obtained after 8 days. The Hex-39-dGp
adducts obtained were separated by HPLC performed on a Hewlett Packard
1050 pump, a Rheodyne injector and a Hewlett Packard photodiode array
detector. The reaction mixture was separated on a Nucleosil C18 column
(length 250 mm, i.d. 8 mm, particle size 10 µm) with a linear gradient from
10 mM ammonium formate buffer pH 4.7 to methanol with a flow rate of
3.0 ml/min in 80 min (HPLC system 1). The 1H-NMR and mass spectra of
the two diastereomers were practically identical. The HPLC-chromatogram,
the UV-spectrum and the chemical structures of the diastereomers are shown
in Figure 1a, b and c.
1H NMR (Bruker 600 MHz, D O) d 5 0.92 (t, J
2
13,12 5 7.4 Hz, 3H, H-13);
1.35–1.48 (m, 2H, H-12); 1.51–1.69 (m, 3H, H-11, H-7a); 2.24 (pseudo-td,
J7b,7a 5 13.9 Hz, J7b,6 5 J7b,8 5 3.1 Hz, 1H, H-7b); 2.61 (ddd, J29a,29b 5
13.9 Hz, J29a,1’ 5 6.4 Hz, J29a,39 5 3.2 Hz, 1H, H-29a); 2.78 (pseudo-td,
J29b,29a 5 13.9 Hz, J29b,1’ 5 J29b,39 5 7.0 Hz, 1H, H-29b); 3.64–3,72 (m, 1H,
H-6); 3.76 (dd, J59a,59b 5 12.5 Hz, J59a,4’ 5 4.7 Hz, 1H, H-59a); 3.77 (dd,
J59b,59a 5 12.5 Hz, J59a,4’ 5 3.6 Hz, 1H, H-59b); 4.20 (pseudo-q; J4’,39 5
J4’,59a 5 J4’,59b 5 3.8 Hz, 1H, H-4’); 4.78–4.84 (m, 1H, H-39); 6.24 (pseudot, J1’,29a 5 J1’,29b 5 7.0 Hz, 1H, H-1’); 6.29 (pseudo-t, J8,7a 5 J8,7b 5 2.6 Hz,
1H, H-8); 7.92 (s, 1H, H-2).
MS (electrospray 14 kV): m/z 5 446 (M1 1 H).
13C-NMR (150.0 MHz, D O): δ (p.p.m.) 5 15.8 (C-13), 22.5 (C-12), 31.1
2
(C-11), 34.3 (C-7), 40.6 (C-29), 51.8 (C-6), 65.0 (C-59), 72.0 (C-8), 73.2
(C-39), 86.4 (C-4’), 89.1 (C-1’), 111.3 (C-10a), 117.6 (C-2), 151.8 (C-3a),
153.6 (C-4a), 160.5 (C-10).
Stability of the adducts
A sample of 7.5 nmol of the Hex-dGp adducts was incubated in 0.1 M
phosphate buffer at 37°C for 5 days. The pH value was varied from pH 2 to
pH 13. Aliquots were analyzed by the same HPLC setup as described
above using a Nucleosil C18 column (length 250 mm, i.d. 4 mm, particle size
5 µm) with a linear gradient from 10 mM ammonium formate buffer pH 4.7
to methanol with a flow rate of 1.0 ml/min in 20 min (HPLC-System 2).
Studies on the nuclease P1 stability of the nucleoside-39-monophosphates
Aliquots of 8 nmol of the different nucleoside 39-phosphates (equivalent to
10 µg DNA) were incubated with 6 µl nuclease P1 mixture in a total volume
of 18 µl for 10 min at 37°C. The nuclease P1 mixture consisted of 1.2 µl
(0.6 µg, 10% of the regular amount used for post-labeling) nuclease P1
solution, 1.8 µl 0.3 mM ZnCl2 and 3.0 µl buffer (0.25 M sodium acetate,
40 mM sodium succinate, 20 mM CaCl2). After 10 min the reaction was
stopped by adding 2.4 µl of 0.5 M Tris base. An aliquot (10 µl) of the mixture
was analyzed by HPLC System 2. The pH value of the sodium acetate buffer
was varied for the determination of optimum pH.
In order to determine the Michaelis–Menten kinetics, different amounts of
the substrate were used at pH 4.0 as described above. The evaluation was
performed according to Lineweaver–Burk using the equation: 1/v 5 1/vmax 1
(KM/vmax)31/S, where KM is the Michaelis–Menten constant.
Using Hex-dGp, the incubation was performed with 6 µg nuclease P1 for
180 min.
Treatment of animals
The animals were fed with standard diet from Altromin (Lage, Germany).
Oral doses of 500, 200 or 50 mg 2-hexenal/kg body wt dissolved in corn oil
were administered to 8-week-old male Fischer 344 (F344) rats (190–210 g)
(four rats per group) by gavage. Control animals were treated with corn oil
only. The rats were sacrificed at different times ranging from 8 h to 4
1346
Fig. 1. (a) HPLC chromatogram of the reaction of 2-hexenal with 39-dGp at
pH 10.7 and room temperature after 8 days. (b) UV spectra of Hex-dGp 1 and
2 and of 39-dGp. (c) Chemical structures of the diastereomers Hex-dGp 1 and
Hex-dGp 2.
days. The following organs were taken: duodenum, esophagus, forestomach,
glandular stomach, kidney, liver, lung, rectum, urinary bladder. The liver was
perfused with 0.9% sodium chloride solution and the other organ tissues were
washed with sodium chloride solution. The organs were either worked up
directly or frozen with liquid nitrogen and stored at –70°C. DNA was isolated
according to the phenol extraction method as described by Gupta (30). The
concentration of the DNA was quantitated using the absorption at 260 nm.
DNA hydrolysis and nuclease P1 treatment
DNA (10 µg) was incubated for 4 h at 37°C with 3 µl enzyme mixture
containing 0.2 U/µl (1 µg/µl) micrococcus nuclease, 0.002 U/µl (1 µg/µl)
spleen phosphodiesterase and 4 µl DNA digestion buffer (25 mM CaCl2,
50 mM sodium succinate pH 6.0) in a total volume of 20 µl. Next, 6 µl
nuclease P1 mixture was added which consisted of 1.2 µl (6 µg) nuclease P1
solution, 1.8 µl 0.3 mM ZnCl2 and 3 µl 250 mM sodium acetate pH 4.0. The
reactant mixture was incubated for 1 h at 37°C and the reaction stopped by
adding 2.4 µl 0.5 M Tris base. The solution was desiccated to dryness and
redissolved in 10 µl water.
32P-post-labeling
To the hydrolyzed and nuclease P1-treated solution 6.5 µl labeling mixture
was added. It was made as a mixture of 1 µl kinase buffer (100 mM
dithiothreitol, 100 mM MgCl2, 10 mM spermidine, 200 mM bicine/NaOH,
pH 9.5), 0.5 µl 800 mM bicine, 5 µl 2.2 µmol [γ-32P]ATP (167 TBq/mmol,
4500 Ci/mmol; 1.9 MBq, 50 µCi) and 0.25 µl (7.5 U) T4 polynucleotide
kinase. The sample was incubated for 45 min at 37°C. Finally, 4 µl (40 µU)
apyrase was added and the solution further incubated for 30 min at the same
temperature.
Exocyclic 1,N 2-propanoadducts of 2-hexenal in vivo
Chromatography
For the detection of the Hex-pdGp adducts in vivo a contact transfer method
was used. The different samples were applied in portions of ~5 µl to an RP18
TLC sheet (length: 10 cm) to which a wick was attached (7 cm, Whatman
#17). The chromatography was performed with 0.4 M ammonium formate/
formic acid pH 6.0 overnight. The origins were cut out (131 cm2, no washing)
and attached with clothes pegs to PEI–cellulose sheets (10310 cm2). The
transfer to the PEI–cellulose material was carried out by developing the sheets
with n-propanol:water:nonidet-P40 (50:50:1, v/v/v) at 50°C. The RP18 chips
were removed and the TLC plates were washed thoroughly with water. After
drying, the chromatography was performed in the first direction (D1) with
1.2 M ammonium formate/formic acid pH 3.5 after attaching wicks (4 cm,
Whatman #1). The wicks were removed and the TLC sheets were washed in
water and air-dried. Another wick (4 cm, Whatman #1) was attached for the
chromatography in the second direction (D2) with 0.3 M ammonium sulphate
(adjusted to pH 7.5 with disodium hydrogenphosphate). The chromatograms
were analyzed by a PhosphorImager™ (Molecular Dynamics).
Quantitation of DNA adducts
Quantitation of the adduct levels was based on the recovery of the amount of
adducts for the reference compound Hex-dGp in an extra sample with DNA
of a control animal. The number of nucleotides was calculated via the UV
absorption of the DNA solution. The recovery of the adducts was determined
in the same manner: the respective amounts of the Hex-dGp standards were
added to DNA, the whole post-labeling procedure was performed and
quantitated as described above, and the mean values and standard deviations
of 3–5 independent determinations of the recoveries were calculated.
Calculation of the covalent binding index (CBI)
The CBI was calculated according to Lutz (31), dividing the adduct levels
measured in adducts/106 nucleotides by the dose applied in mmol substance/
kg body wt.
Results
Synthesis and characterization of standards
The Hex-dGp adducts were synthesized as standards by
reaction of 2-hexenal and 39-dGp as described above. The
respective HPLC chromatogram and the UV-spectra of the
formed adducts are shown in Figure 1a and b. The two resulting
adducts were found to be a pair of diastereomers of the
regioisomer shown in Figure 1c in which the OH-group is
adjacent to the N1 of the purine ring. The 1H-NMR spectra
demonstrated that only the trans-isomer was formed with no
detectable amounts of the cis-isomer. No detectable amounts of
other adducts could be found. Instead of the correct IUPAC
nomenclature (6R,8R)-3-(29-deoxy, 39-phosphate-β-D-erythropentafuranosyl)-6-propyl-5,6,7,8-tetrahydro-8-hydroxy-pyrimido[1,2-a]purine-10(3H)-one and (6S,8S)-3-(29-deoxy-39phosphate-β-D-erythro-penta-furanosyl)-6-propyl-5-6-7-8tetrahydro-8-hydropyrimido[1,2-a]purine-10(3H)-one, the common names 8-hydroxy-6-propyl-1,N 2-29-deoxyguanosine-39monophosphate 1 and 2 or 2-hexenal 1,N2-propanodeoxyguanosine adducts (Hex-dGp) are generally used. The structure,
configuration and conformation of Hex-dGp were assigned by
spectroscopic methods, in particular by 1H-NMR. The 1H-NMR
spectra of the two diastereomers of Hex-dGp are practically
identical and the 1H-NMR spectra (as well as 13C NMR, UV, IR
and mass spectra) of Hex-dG (32) and Hex-dGp are very similar.
32P-post-labeling technique
The detection of the respective adducts Hex-dGp 1 and 2 is
based on a 32P-post-labeling nuclease P1 technique and on the
separation of the labeled adducts with PEI–cellulose TLC. In
addition, the contact transfer method with RP18 TLC as
described in Materials and methods was used in our new
method in order to minimize the radioactive background on
the TLC sheet, in particular, in the region of Hex-dGp spots.
With this TLC method, the diastereomers Hex-dGp 1 and 2
appear in one spot and are clearly separated from normal
Fig. 2. Dependence of the chemical stability of Hex-dGp on the pH value after
either 48 h (s) or 120 h (d) incubation time.
nucleotides and other radioactive impurities (see also Figure
3b).
The adducts are stable within a wide pH range from pH 4
to pH 11 (Figure 2). This means that the most suitable pH
conditions can be chosen for each single post-labeling step.
The Hex-dGp adducts are also quite stable against nuclease
P1. In our enzyme kinetic studies we found a maximum
reaction rate of vmax 5 6.5 pmol/min/mg enzyme and a
Michaelis constant, KM, of 3.9 mM. For comparison the vmax
of, e.g. 39-dTp (deoxythymidine-39-monophosphate), the most
stable nucleotide is 7400 pmol/min/mg enzyme. Since the
kinetics are approximately first order with very low substrate
amounts as used in the post-labeling technique we can calculate
a rate constant of k 5 0.00056 min–1 for Hex-dGp using the
equation:
k 5 [vmax 3 m(enzyme)]/(KM 3 V)
Under our conditions the amount of enzyme (m) is 6 µg
nuclease P1 and the reaction volume (V) is 18 µl.
With this method we received a labeling efficiency of 35%
and a recovery of 10%. The limit of quantification is in the
range of 0.03 adducts/106 nucleotides. The relative standard
deviation at the limit of quantification is ~40%; in comparison,
the relative standard deviation at a level of 1 adduct/106
nucleotides is 27% with our method (see Materials and
methods, ‘Quantification of DNA adducts’).
Detection of Hex-dGp adducts in male F344 rats
Figure 3a represents the post-labeling TLC of liver DNA of
untreated rats as control. Figure 3b shows the 32P-post-labeling
TLC analysis of liver DNA of male F344 rats 2 days after
p.o. application of 500 mg 2-hexenal/kg body wt (right hand
side). No adducts could be found in the liver DNA or in other
organs of untreated F344 rats used throughout our experiments,
who were kept under our standard conditions and fed with
our standard diet (see Materials and methods). Figure 4
demonstrates the time dependence of adduct formation with
2-hexenal. Highest adduct levels were obtained 2 days after
the gavage. Four days after gavage the level is even higher
than after 1 day. No adducts could be detected 8 h after gavage.
Figure 5a shows the levels of the Hex-dGp adducts in
different organs 2 days after gavage of 500 mg/kg body wt
and Figure 5b the levels in the same organs 2 days after
gavage of 200 mg/kg body wt. Highest adduct levels were
found in the organs which came directly into contact with
2-hexenal after gavage, i.e. forestomach, liver and esophagus
(first site effect). Furthermore, the adduct concentrations in
the urinary bladder and the lung were too low for quantitation.
1347
D.Schuler and E.Eder
Fig. 3. (a) Post-labeling TLC analysis of liver DNA in untreated male F344 rats (control). (b) Post-labeling TLC analysis of liver DNA 2 days after administering
an oral dose of 500 mg 2-hexenal/kg body wt.
Fig. 4. Dependence of Hex-dG levels in the livers of male F344 rats on the
time after application of an oral dose of 500 mg 2-hexenal/kg body wt. (Mean
values and standard deviations of four post-labeling determinations.)
After gavage of 50 mg/kg body wt no adducts or only traces
of adducts could be seen in most of the organs investigated.
The only organ in which we could quantify adducts after
application of the 50 mg dose was the esophagus where we
found a level of 0.08 adducts/106 nucleotides. The dose
dependence of the adduct formation is presented in Figure 6a
for the organs in which highest adduct levels were measured,
i.e. forestomach, liver and esophagus. As already mentioned,
with a dose of 50 mg/kg body wt, quantifiable adduct levels
were only found in the esophagus. At this dose, the adduct
levels in the other two organs were just below the quantitation
limit and were estimated to be in the range of 0.01–0.02
adducts/106 nucleotides. It is remarkable that the graph of the
adduct-level/dose-dependence shows a convex shape (Figure
6a). This means that at higher doses there is a disproportionate
increase in adduct levels if compared with the lower doses.
The CBI is also dose-dependent. According to our results,
the CBI is not higher than 0.06 at low doses, whereas at the
higher doses of 200 and 500 mg/kg body wt it is in the range
of 0.22–0.62.
Discussion
As stated in detail in the Introduction, 2-hexenal is mutagenic,
genotoxic and is permanently present in our food as a natural
component of fruit and vegetables. Since dietary habits and
natural constituents of food are a major factor in human
1348
Fig. 5. (a) Hex-dGp adduct levels in the DNA of different organs of F344
rats 2 days after gavage of 500 mg 2-hexenal/kg body wt. (b) Hex-dGp
adduct levels in different organs of F344 rats 2 days after gavage of 200 mg
2-hexenal/kg body wt. (Mean values and standard deviations, four rats in each
case.)
carcinogenesis (33) the role of intake of 2-hexenal via food
for carcinogenicity has to be examined. To date, no long
term carcinogenicity studies on 2-hexenal are available, however, as described in the Results, 2-hexenal forms the analogous
DNA adducts as other α,β-unsaturated carbonyl compounds
known as carcinogens, e.g. acrolein or crotonaldehyde
(1,34,35,36) and is mutagenic and genotoxic like these compounds (see Introduction). Recently, the influence of this type
of adduct on DNA damage and mutation was investigated
(2,3,37,38) and these adducts are considered to be promutagenic
and to play a role in the initiation, but probably also in promotion
and progression, of cancer. Nevertheless, it remains unclear, in
Exocyclic 1,N 2-propanoadducts of 2-hexenal in vivo
Fig. 6. (a) Dependence of adduct levels on dose in three different organs of F344 rats. (b) Dependence of the CBI of 2-hexenal on dose in organs of F344 rats. d,
Forestomach; j, liver; e, esophagus.
part, whether the different structural features of the adducts, e.g.
regioisomerism, cis- and trans-isomerism or conformations have
special influence on cancer induction. In summary, there are
some significant indications that 2-hexenal plays a role in human
carcinogenicity and that detection of 2-hexenal DNA adducts
in vivo can improve the assessment of the cancer risk arising
from intake of 2-hexenal via vegetarian food.
The new post-labeling technique we have developed offers
the opportunity to detect 2-hexenal adducts in vivo at a relatively
low detection limit. Optimum conditions for the single steps
of the labeling procedures could be worked out and it was
demonstrated that the respective Hex-dGp adducts are chemically stable over a wide pH range and against the enzyme nuclease
P1. Both properties are prerequisites for a sensitive 32P-postlabeling detection of these adducts because no enrichment
method other than the nuclease P1 method proved to be effective
for the adducts in our experiments and because the optimum pH
values for the various steps of the labeling procedure differed
over a wide range. With this 32P-post-labeling technique we
could demonstrate in vivo DNA adduct formation of 2-hexenal
for the first time. Recently Gölzer et al. (29) reported that they
could not find 2-hexenal adducts in the stomach of rats treated
with 320 mg/kg body wt 2-hexenal by gavage. The authors
looked, however, for adducts as early as 12 h after gavage. As
we have shown in Figure 4, we could not find adducts 8 h after
gavage and think it likely that adducts would still not have been
found 12 h after gavage. Thus, the data of Gölzer et al. are in
line with ours in that 2-hexenal adducts appear relatively late
after gavage. This effect is surprising at first glance since 2hexenal is a directly alkylating compound which requires no
bioactivation for adduct formation. Recently, Witz (39) pointed
out that conjugation of α,β-unsaturated carbonyl compounds
with glutathione must not necessarily be a detoxication
mechanism because the conjugation is reversible in equilibrium
and that the conjugate may dissociate in cell compartments of
low glutathione content, e.g. in the cell nucleus where DNA is
located. Glutathione addition to the double bond of 2-hexenal
(Michael addition) is considered to be one of the major and
fastest biological reactions which these compounds undergo
(7,39). On the other hand, 2-hexenal does not possess high
chemical reactivity and the reaction rate of 2-hexenal with DNA
components is relatively low in vitro (32) (see also ‘Synthesis
of Hex-dGp adducts’ in Material and methods). Evidently it also
takes some time until detectable amounts of adducts are formed
in vivo. Figure 4 also indicates that the Hex-dGp adducts might
be repaired to a certain extent since the adduct level 4 days after
gavage is only one third that detected 2 days after gavage,
however, the decrease in the adduct level can also be explained,
at least in part, by cell turnover.
The organ distribution of the adducts (Figure 5a and b) demonstrates a tendency for site of first contact binding, despite the
fact that adducts are formed rather late. Evidently, 2-hexenal is
directly absorbed after gavage by the tissues of the forestomach, esophagus and duodenum but to a lesser degree in the
glandular stomach. Systemic distribution evidently also
occurs to a certain extent because a high adduct level is also
found in the kidney. A first pass effect should be expected in the
liver and, most likely, the systemic distribution of 2-hexenal
runs partly via the stabilized transport form of the 2-hexenal–
glutathione conjugate which may prevent 2-hexenal from
further biotransformation such as epoxidation of the double
bond. Glutathione conjugation and Michael addition of other
SH-group-containing biological molecules such as certain amino
acids is probably also responsible for the dependence of the
adduct levels on the doses. High doses of 2-hexenal lead to a
glutathione depletion and also to a decrease of other SH-groupcontaining compounds and therefore a higher amount of the
dose is available for DNA binding. Thus, the binding at higher
doses is disproportionate if compared with lower doses (see
Figure 6a). This effect is observed, in particular, if considering
the relationship between the CBI and the dose (Figure 6b).
According to its definition (see Materials and methods, ‘Calculation of the CBI’) the CBI is, in general, independent of the dose.
The CBI can be utilized for a cancer risk estimation (31). The
adduct levels, or the CBI as found in the case of higher 2-hexenal
doses, may be considered as irrelevant for cancer risk assessment
since such high doses of 2-hexenal do not occur in the normal
human diet, however, glutathione depletion may also occur in
humans under certain circumstances, e.g. intake of certain medication, independently of the 2-hexenal intake and glutathione
depletion results in increased CBIs as pointed out above. Therefore, higher CBIs than that calculated for the lowest dose should
also be taken into consideration if performing a cancer risk
assessment. In every case the CBI of 2-hexenal is extremely
low and according to Lutz (31) the carcinogenic potency of
substances with such low CBIs is also extremely low. If considering that the intake of 2-hexenal via food is very
low, according to the literature data shown in the Introduction,
one can estimate that the cancer risk arising from 2-hexenal
intake via consumption of fruit and vegetables is also very
low. The dose versus DNA adduct formation relationship
shown here demonstrates that relatively high single doses are
1349
D.Schuler and E.Eder
required to detect DNA adducts with the detection limit of our
method. According to refs 19–28 a daily 2-hexenal intake of
~0.15 mg/kg body wt/day can be estimated. As a single dose
this intake would lead to very low and unmeasurable DNA
adduct levels considering our dose versus DNA adduct level
relationship. Thus, we can estimate, on the basis of both the
expected low DNA adduct levels and the low CBI, that the
benefit from eating fruit and vegetables is much higher than the
low, unavoidable risk from 2-hexenal intake. However, nothing
can be said about a steady state DNA adduct level after a
permanent low intake of 2-hexenal via food and beverages. The
situation deserves a somewhat more precise assessment on the
basis of the TD50, which can be derived from the CBIs according
to the correlation of Lutz (31), and on the basis of a more exact
evaluation of the daily 2-hexenal intake. This more precise
assessment will be performed and discussed in a forthcoming
paper.
Acknowledgements
We wish to thank Mrs Elisabeth Weinfurtner for excellent technical assistance.
This work was supported by Deutsche Forschungsgemeinschaft, SFB 172.
References
1. Eder,E., Hoffman,C., Sporer,S. and Scheckenbach,S. (1993) Biomonitoring
studies and susceptibility markers for acrolein congeners and allylic and
benzylic compounds. Environ. Health Perspect., 99, 245–247.
2. Plum,G.E., Grollman,A.P., Johnson,F. and Breslauer,K.J. (1992) Influence
of an exocyclic guanine adduct on the thermal stability, conformation and
melting thermodynamics of a DNA duplex. Biochemistry, 31, 12096–12102.
3. Czerny,C., Eder,E. and Rünger,T.M. (1998) Genotoxicity and mutagenicity
of the α,β-unsaturated carbonyl compound crotonaldehyde (butenal) on a
plasmid shuttle vector. Mutat. Res., 407, 125–134.
4. Marnett,L.J., Hurd,H.K., Hollstein,M.C., Levin,D.E., Esterbauer,H. and
Ames,B.N. (1985) Naturally occurring carbonyl compounds are mutagens
in Salmonella-tester-strain TA 104. Mutat. Res., 148, 25–34.
5. Eder,E., Deininger,C., Neudecker,T. and Deininger,D. (1992) Mutagenicity
of β-alkyl substituted acrolein congeners in the Salmonella typhimurium
strain TA100 and genotoxicity-testing in the SOS chromotest. Environ. Mol.
Mutag., 19, 337–345.
6. Eder,E., Hoffman,C., Bastian,H., Deininger,C. and Scheckenbach,S. (1990)
Molecular mechanisms of DNA damage initiated by α,β-unsaturated
carbonyl compounds as criteria for genotoxicity and mutagenicity. Environ.
Health Perspect., 88, 99–106.
7. Schaur,R.J., Zollner,H. and Esterbauer,H. (1991) Biological effects of
aldehydes with particular attention to 4-hydroxynonenal and malonaldehyde.
In Vigo-Pelfrey,C. (ed.) Membrane Lipid Oxidation. CRC Press, Ann Arbor,
MI, pp. 141–156.
8. Chung,F.-L., Tanaka,T. and Hecht,S.S. (1989) Induction of liver tumors in
F344 rats by crotonaldehyde. Cancer Res., 46, 1285–1289.
9. Cohen,S.M., Garland,E.M., St. John,M. Okamura,T. and Smith,R.A. (1992)
Acrolein initiates rat urinary bladder carcinogenesis. Cancer Res., 52,
3577–3581.
10. Canonero,R., Martelli,A., Marinari,U.M. and Brambilla,G. (1990) Mutation
induction in Chinese hamster lung V79 cells by five alk-2-enals produced
by lipid peroxidation. Mutat. Res., 244, 153–156.
11. Griffin,D.S. and Segall,H.J. (1986) Genotoxicity and cytotoxicity of selected
pyrrolizidine alkaloids, a possible alkenal metabolite of the alkaloids and
related alkenals. Toxicol. Appl. Pharmacol., 86, 227–234.
12. Eder,E., Scheckenbach,S., Deininger,C. and Hoffman,C. (1993) The possible
role of α,β -unsaturated carbonyl compounds in mutagenesis and
carcinogenesis. Toxicol. Lett., 67, 87–103.
13. Eisenbrand,G., Schuhmacher,J. and Gölzer,P. (1995) The influence of
glutathione and detoxifiying enzymes on DNA damage induced by 2-alkenals
in primary rat hepatocytes and human lymphoblastoid cells. Chem. Res.
Toxicol., 8, 40–46.
14. Dittberner,U., Eisenbrand,G. and Zankl,H. (1995) Genotoxic effects of the
α,β-unsaturated aldehydes 2-trans-butenal, 2-trans-hexenal and 2-trans, 6cis-nonadienal. Mutat. Res., 335, 259–265.
15. Griffin,D.S. and Segall,H.J. (1989) Effects of the pyrrolizidine alkaloid
1350
senecionine and the alkenals trans-4-OH-hexenal and trans-2-hexenal on
intracellular calcium compartmentation in isolated hepatocytes. Biochem.
Pharmacol., 37, 391–397.
16. Jørgensen,B.M., Agerbo,P., Jensen,B., Børresen,T. and Hølmer,G. (1992)
Inhibition of microsomal glucose 6-phosphatase by unsaturated aliphatic
aldehydes and ketones. Chem. Biol. Interact., 81, 209–218.
17. Gaunt,I.F., Colley,J., Wright,M., Creasey,M., Grasso,P. and Gangolli,S.D.
(1971) Acute and short-term toxicity studies on trans-2-hexenal. Food
Cosmet. Toxicol., 9, 775–786.
18. Major,R.T., Collins,O.D., Marchini,P. and Schnabel,H.W. (1972) Formation
of 2-hexenal by leaves. Phytochemistry, 11, 607–610.
19. Major,R.T., Marchini,P. and Sproston,T. (1960) Isolation from Ginkgo biloba
L. of an inhibitor of fungus growth. J. Biol. Chem., 235, 3298–3299.
20. Drawert,F., Heimann,W., Emberger,R. and Tressl,R. (1966) Enzymatische
Bildung von Hexen-(2)-al-(1), Hexanal und deren Vorstufen. Z. Naturforsch.,
694, 200–208.
21. de Lumen,B.O., Stone,E.J., Kazeniac,S.J. and Forsythe,R.H. (1978)
Formation of volatile flavor compounds in green beans from linoleic and
linolenic acids. J. Food Sci., 43, 698–708.
22. Skobeleva,N.I., Bezzubov,A.A., Petrova,T.A. and Bokuchava,M.A. (1979)
Aroma-forming substances in black tea. Appl. Biochem. Microbiol., 15,
682–687.
23. Schildknecht,H.
and
Rauch,G.
(1961)
Über
Insekten-und
Pflanzenabwehrstoffe, ihre Isolierung und Strukturaufklärung. Angew.
Chemie, 73, 629.
24. Buttery,R.G., Teranishi,R. and Ling,L.C. (1987) Fresh tomato aroma
volatiles: a quantitative study. J. Agric. Food Chem., 35, 540–544.
25. Casperson,G., Banasiak,L., Lyr,H. and Sunkel,M. (1986) Wirkung von 2Ethyl-2-en-1-al und verwandter Verbindungen auf Wachstum und
Ultrastruktur von Pilzen. J. Basic Microbiol., 26, 259–269.
26. Schildknecht,H. and Rauch,G. (1961) Die chemische Natur der
‘Luftphytoncide’ von Blattpflanzen insbesondere von Robinia pseudacacia.
Z. Naturforschg., 16b, 422–429.
27. Stone,E.J., Hall,R.M. and Kazeniac,S.J. (1975) Formation of aldehydes and
alcohols in tomato fruit from U-14C-labeled linolenic and linoleic acids. J.
Food Sci., 40, 1137–1141.
28. Eiter,K. (1970) Insekten Sexuallockstoffe. In Wegler,R. (ed.) Chemie der
Pflanzen und Schädlingsbekämpfungsmittel. Springer, Berlin, Germany, pp.
497–522.
29. Gölzer,P., Janzowski,C., Pool-Zobel,B.L. and Eisenbrand,G. (1996) (E)-2Hexenal-induced DNA damage and formation of cyclic 1,N2-( 1,3-propano)29-deoxyguanosine adducts in mammalian cells. Chem. Res. Toxicol., 9,
1207–1213.
30. Gupta,R. (1996) 32P-postlabeling for detection of DNA adducts. In Pfeifer,G.
(ed.) Technologies for Detection of DNA Damage and Mutations. Plenum
Press, New York, pp. 45–61.
31. Lutz,W.K. (1986) Quantitative evaluation of DNA binding data for risk
estimation and for classification of direct binding carcinogens. J. Cancer
Res. Clin. Oncol., 112, 85–91.
32. Eder,E. and Hoffman,C. (1993) Identification and characterization of
deoxyguanosine adducts of mutagenic β-alkyl-substituted acrolein
congeners. Chem. Res. Toxicol., 6, 486–494.
33. Doll,P. and Peto,R. (1981) The causes of cancer: quantitative estimates of
avoidable risks of cancer in the United States today. J. Natl Cancer Inst., 66,
1192–1308.
34. Maccubin,A.E., Caballes,L., Scapaticci,F., Struck,R.F. and Gurtoo,H.L.
(1990) 32P-postlabeling analysis of binding of the cyclophosphamide
metabolite, acrolein, to DNA. Cancer Commun., 2, 207–211.
35. Chung,F.-L., Young,R. and Hecht,S.S. (1989) Detection of cyclic 1,N2propanodeoxyguanosine adducts in DNA of rats treated with Nnitrosopyrrolidine and mice treated with crotonaldehyde. Carcinogenesis,
10, 1291–1297.
36. Eder,E. and Hoffman,C. (1992) Identification and characterization of
deoxyguanosine crotonaldehyde adducts. Formation of 1,N2,7,8-bis-cyclic
adducts. Chem. Res. Toxicol., 5, 802–808.
37. Benamira,M., Singh,U. and Marnett,L.J. (1992) Site-specific frameshift
mutagenesis by a propanodeoxyguanosine adduct positioned in the (CpG)4
hot-spot of Salmonella typhimurium his D 3052 carried on an M13 vector.
J. Biol. Chem., 267, 392–400.
38. Moriya,M., Zhang,W., Johnson,F. and Grollman,A.P. (1994) Mutagenic
potency of exocyclic DNA adducts: marked differences between Escherichia
coli and simian kidney cells. Proc. Natl Acad. Sci. USA, 92, 11899–11903.
39. Witz,G.C. (1989) Biological interactions of α,β-unsaturated aldehydes. Free
Radical Biol. Med., 7, 333–349.
Received December 30, 1998; revised March 8, 1999; accepted April 1, 1999