FUNDAMENTAL AND APPLIED TOXICOLOGY 2 9 , 9 4 - 1 0 1 (1996)
Article No. 0010
Chromatographic Characterization of Hemoglobin Benzo[a]pyrene7,8-diol-9,10-epoxide Adducts
STEVEN R. MYERS, JOSEPH A. SPINNATO, AND MARIA T. PINORINI
Department of Pharmacology and Toxicology, School of Medicine, University of Louisville, Louisville, Kentucky 40292
Received February 9, 1995; accepted June 30, 1995
Exposure to ethylene oxide present in occupational enviChromatographic Characterization of Hemoglobin Benzo[a]- ronments and in tobacco smoke results in the formation of
pyrene-7,8-dio!-9,10-epoxide Adducts. MYERS, S. R., SPINNATO, hydroxyethyl-hemoglobin adducts via reaction at the NJ. A., and PINORINI, M. T. (1996). Fundam. Appl. Toxicol. 29, 94terminal a amino group (vaJine), the imidazole nitrogens of
101.
histidine, and the /393 cysteine sulfhydryl group (Fanner et
The formation of hemoglobin-carcinogen adducts has been de- al., 1986; Tornqvist et al., 1986). Cigarette smoking also
tected in carcinogen-treated animals and in human populations. results in the formation of an adduct between the hemoglobin
Although polynuclear aromatic hydrocarbons are ubiquitous in
/?93 cysteine SH group and 4-nitrosobiphenyl arising from
the human environment and DNA-aromatic hydrocarbon adducts have been detected in human tissue, the occurrence of hemo- the metabolism of 4-aminobiphenyl (Bryant et al., 1987).
globin-polynuclear aromatic hydrocarbon adducts in humans has Adducts with hemoglobin and other aryl amines have also
not been thoroughly described. In this study we examined the been detected and reported (Skipper et al., 1988). Studies
effects of reaction conditions on the extent of in vitro reaction with rats have shown that metabolites of nitrosamines presof human hemoglobin and (+) [3H]benzo[a]pyrene-7,8-diol-9,10- ent in tobacco smoke also form adducts with hemoglobin
epoxide (antiXBPDE), a metabolite thought to be largely responsi- (Carmella and Hecht, 1987).
ble for the carcinogenic effect of benzo[a]pyrene. The chromatoAdministration of carcinogenic polynuclear aromatic hygraphic properties of the resulting hemoglobin-BPDE adducts
drocarbons (PAH) to rodents results in the formation of hewere examined by conventional DEAE-cellulose ion exchange liquid chromatography and by reversed phase high performance liq- moglobin-PAH adducts (Pereira and Chang, 1981; Shugart,
uid chromatography. Several adducts were formed which were 1985; Wallin et al., 1987) for which little structural informachromatographically resolved from hemoglobin and from the indi- tion is currently available. Acid treatment (0.1 A' HC1, 80°C)
vidual globins. Some adducts were basic and some acidic relative of hemoglobin obtained from benzo[a]pyrene-treated mice
to unaltered hemoglobin, suggesting adduct formation by reaction released chromatographically detected benzo[a]pyrene-teat carboxyl and basic nitrogen groups, respectively. Alteration of trols, indicating that the adducts were at least in part formed
the ion - chromatographic properties of the adducts by an ionic by reaction of benzo[a]pyrene diol-epoxides (BPDE) with
sulfhydryl reagent, together with only a moderate effect of pH on hemoglobin (Shugart, 1985), metabolites largely thought to
the extent of adduct formation, indicated that the adducts were
be responsible for the carcinogenic and mutagenic activity of
not formed via reaction with the 091 cysteine sulfhydryl group.
The chromatographic techniques employed may be applicable for benzo[fl]pyrene (Dipple et al., 1984). However, comparative
the characterization and analysis of other hemoglobin-carcinogen radiometric and immunochemical measurements (Wallin et
al., 1987) recently showed that benzo[a]pyrene tetrols readdUCtS. C 1996 SodMy of Toxicology
leased by acid treatment of hemoglobin from benzo[a]pyrene-treated mice accounted for only a minor portion of the
Since covalent adducts formed by reaction of electrophilic covalently bound benzo[a]pyrene.
Epoxides and diol-epoxides are presumably the ultimate
carcinogen metabolites with hemoglobin was proposed as a
carcinogenic
forms of many polynuclear aromatic hydrocarmeasure-of carcinogen exposure (Ehrenberg et al., 1974),
bons
(Dipple
et al., 1984). These reactive epoxides have
formation of hemoglobin adducts with several types of carbeen
shown
to
react with DNA via nucleophilic attack of
cinogens has been described (Pereira and Chang, 1981). The
DNA
bases.
Similar
reactions of alkylating compounds can
structures of these adducts with several types of carcinogens
occur
with
the
nucleophilic
sites of hemoglobin such as the
has been determined (Pereira et al., 1984; Segerback, 1983;
sulfhydryl,
amino,
carboxyl,
and hydroxyl groups, and it has
Green et al., 1984) and the dose response and kinetic relabeen
postulated
that
hemoglobin
adducts of various types
tionships for their formation and disappearance in animals
may
be
useful
in
the
measurement
of exposure and metabohas been explored (Pereira and Chang, 1981; Shugart, 1985;
lism
of
various
compounds.
Segerback, 1983; Carmella and Hecht, 1987).
0272-0590/96 $12.00
Copyright O 1996 by the Society of Toxicology
All rights of reproduction in any form reserved
HEMOGLOBIN-BENZO[a]PYRENE ADDUCTS
Recently, Day et al. (1990) demonstrated a molecular dosimetry of polynuclear aromatic hydrocarbon epoxides and
diol epoxide adducts with hemoglobin. The results from these
studies indicated that alkylation of carboxylic acids to form
esters is a general reaction for both epoxide and diol epoxide
metabolites of PAH. Furthermore, the formation of the ester
adducts was found to give rise to alcoholic products (tetrols)
of the polynuclear aromatic hydrocarbons upon hydrolysis of
the adduct, which upon derivatization are suitable substrates
for gas chromatographic and mass spectral analysis.
In additional studies, Weston et al. (1989) demonstrated
fluorescence and mass spectral evidence for the formation
of benzo[a]pyrene anti-diol epoxide DNA and hemoglobin
adducts in humans. In these studies, high performance liquid
chromatography and fluorescence spectroscopy were used to
detect and identify residues of benzo[a]pyrene diol-epoxide
which were released upon mild acid hydrolysis of human
hemoglobin or DNA. Levels of adducts detected by these
methods ranged from 1.5 fmol benzo[a]pyrene diol-epoxide
adduct per microgram of DNA to approximately 1.0 ng of
the diol-epoxide per gram of hemoglobin (Weston et al.,
1989). Recently, Day and colleagues (Day et al, 1991) identified a stable ester adduct formed by reaction of benzo[a]pyrene diol-epoxide with human hemoglobin. This adduct
was characterized as occurring in the a chain of hemoglobin
at an aspartate residue (amino acid 47). Additional studies
(Hutchins et al., 1988) have shown the isolation and characterization of major fluoranthene hemoglobin adducts formed
in vivo in rats. In this series of studies, the carcinogen fluoranthene was administered in the diet. After several weeks,
the animals were sacrificed and hemoglobin analyzed for the
presence of fluoranthene adducts. The results demonstrated
that fluoranthene bound covalently to rat hemoglobin
through the formation of fluoranthene 2,3-dihydrodiol-l,
lOb-epoxides at the Pl2s residue (Hutchins et al., 1988).
These results suggest that reactive sulfhydryl groups of hemoglobin may be targets for the formation of covalent hemoglobin adducts. In this report we describe the effects of in
vitro reaction conditions on the extent of adduct formation
between [3H]BPDE and human hemoglobin, as well as the
liquid chromatographic resolution of the intact hemoglobin
BPDE or globin BPDE adducts from the unmodified peptides. We report information about the hemoglobin sites with
which BPDE reacts, and about chromatographic techniques
that may be generally applicable for isolation and structural
analysis of hemoglobin-carcinogen adducts.
MATERIALS AND METHODS
Chemicals. [1,3-3H]BPDE (1,249 mCi/mmol) was purchased from the
NCI Radiochemical Carcinogen Repository, Chemsyn Science Laboratories
(Lenexa, KS). Nonlabeled BPDE was purchased from the NCI Chemical
Carcinogen Reference Standard Repository, Midwest Research Institute
(Kansas City, MO). 4-(Iodoacetamido)salicylic acid (ISAL) was purchased
95
from Sigma Chemical Co. (St. Louis, MO) and was recrystallized from
ethanol. All other chemicals and reagents were of the highest grade commercially available.
Isolation of hemoglobin.
Human blood samples were obtained from
volunteers from the University of Louisville Ambulatory Care Center. Blood
samples were obtained only from nonsmokers, to minimize contamination
of blood samples from tobacco smoke carcinogens. Saline-washed, packed
human erythrocytes were lysed with 2 vol of water. NaCl was added to 0.5
M and the stroma free lysate was obtained by centrifugation at 40,000g for
30 min. The hemoglobin solution was dialyzed overnight (4°C) against 0.15
M NaCl to remove 2,3-diphosphoglycerate. Portions of the dialyzed solution
were frozen by dropwise addition to liquid nitrogen and stored at -20°C.
Analysis of hemoglobin and globin concentrations.
Hemoglobin concentrations were determined as cyanmethemoglobin (Tentori and Salvati,
1981) €no = 110 mM"1. Identical absorbance values were obtained in the
presence and absence of 1% Triton X-100. Chromatographic fractions were
monitored by measurement of Atli (oxyhemoglobin, e = 125 mM"1) or A,l5
(carboxyhemoglobin, e = 190 mM"1). Globin was determined from e^m =
12.8 mM"1. All concentrations are expressed on a heme or globin monomer
basis. Hemoglobin preparations contained less than 1% methemoglobin.
Reaction of benzo[a\pyrent diol-epoxide with hemoglobin.
Hemoglobin solutions were dialyzed (4°C) against 5 mM Tris-KCl buffer, pH 7.7,
and concentrated to 10 to 15 mM by surrounding the dialysis bag with dry
Sephadex G-100. A solution of [3H]BPDE (0.01 vol, 1 nun) in tetrahydrofuran was added to 0.99 vol of hemoglobin in buffer at 37°C. pH values were
determined at 37°C.
In control reactions, the BPDE was hydrolyzed to benzo[a]pyrene tetrols
prior to reaction with hemoglobin. For these controls, a mixture of 0.01
vol of [3H]BPDE solution and 0.89 vol of buffer was incubated at 37°C for
30 min to hydrolyze the [3H]BPDE. Hemoglobin (0.1 vol) was then added
and portions were subsequently removed for analysis after incubation. Reaction times corresponded to those chosen for the complete reaction in which
the BPDE was not hydrolyzed prior to addition of the hemoglobin.
The concentration of the [3H]BPDE in each reaction mixture was determined by radiometric measurement, using a specific activity obtained by
the spectrophotometric determination of BPDE concentration (logeju =
4.69; ethanol).
Extraction of benzo[a']pyrene diol-epoxide reaction mixtures.
Extractions were performed at 4°C. Portions of the reaction mixtures typically
containing 50 nmol Hb were added to either 4 ml 15 mM HC1 in acetone
(with vigorous stirring) or to 4 ml of acetone (stirred after 3 - 5 min) followed by centrifugation at -5°C. Globin precipitates (HCl-acetone) were
washed with 2 ml each of HCI acetone, 50 mM imidazole in acetone,
and acetonitrile (to avoid interference by residual acetone). Hemoglobin
precipitates (neutral acetone) were washed three times with 2 ml acetone
and then once with 2 ml acetonitrile. The precipitates were dried under
nitrogen at room temperature.
Dried globin was dissolved in 1.0 ml water (final pH 5-6), A^ was
measured, and 0.2 ml was used for liquid scintillation counting (5 ml of
Aquasol II; Amersham). Dried hemoglobin was dissolved in 0.5 ml of 1%
aqueous Triton X-100. Individual 0.2-ml portions were used for cyanmethemoglobin analyses and for liquid scintillation counting (heme was decolorized by the addition of ethanolic m-chloroperoxybenzoic acid). The globin
and hemoglobin concentrations were used to caJculate the specific contents
of the adduct preparations.
Chromatographic analysis of hemoglobin benzo[a]pyrene diol-epoxide
adducts. [3H]BPDE hemoglobin reaction mixtures were applied to columns of Sephadex G-25 (Pharmacia) equilibrated and eluted with water at
4°C. DEAE cellulose (DE-52, Whatman) chromatography was performed
as previously described (Schroeder and Huisman, 1980), except that a linear
NaCl gradient was used and the buffers were saturated with carbon monoxide to prevent the formation of methemoglobin in fractions containing low
concentrations of hemoglobin. A 1 x 25-cm column was equilibrated with
96
MYERS, SPINNATO, AND PINORINI
glycine-KCN buffer (0.2 M glycine, 0.01% KCN, pH 7.7). The upper 1 2 cm of bed was composed of a 3:1 mixture (v/v, hydrated materials) of
Sephadex G-25 and DEAE-cellulose to allow the hemoglobin to bind less
densely and more uniformly. Three forms of hemoglobin are separated by
this technique, hemoglobin A,, (the primary adult hemoglobin), hemoglobin
Aic (glycosylated hemoglobin), and hemoglobin A2 (minor hemoglobin
variant).
BPDE-treated hemoglobin (20-30 mg) obtained by gel filtration (elution
with water) was diluted with an equal volume of glycine-KCN buffer,
applied to the DEAE-cellulose column, and the hemoglobins were eluted
at 36 rruVhr with a gradient composed of 600 ml each of 5 and 60 nun NaCl
in glycine-KCN buffer. NaCl concentrations in the fractions were measured
conductometrically. Fractions were collected and counted for radioactivity
and hemoglobin determined spectrophotometrically.
Reversed phase HPLC analysis was conducted at 30°C using a 4.6 mm
x 25 cm Vydac C-4 column. A Waters HPLC system equipped with a
model M600E solvent delivery system delivered a gradient of 38 to 45%
acetonitrile, containing 0.1% trifluoroacetic acid, at a flow rate of 1.0 ml/
min in 52 min. Absorbance was monitored using a Waters model 996
photodiode array detector. Chromatographic data was analyzed using a
NEC 486/33 computer containing the Waters Millennium chromatographic
manager (version 2.0).
RESULTS
Effect of pH and Extraction Procedure on Extent of
Adduct Formation
Results of initial experiments conducted using phosphate
buffers were complicated by the formation of the BPDE
phosphate esters (Whalen et al., 1979) which were removed
by extraction of the hemoglobin with HCl-acetone, but not
with neutral acetone. Thus, results of experiments presented
here were obtained using either 20 mM Tris-imidazole-KCl
or 20 mM Tris-HCl buffers supplemented with 0.1 M KC1.
The low concentration of the buffer was selected to minimize
Tris-dependent hydrolysis of BPDE (Whalen et al., 1979).
The levels of the BPDE incorporation depended only moderately on pH, between pH 6 and 8, and were similar or
identical when measured using the two extraction procedures
(Fig. 1). In control reactions, in which the BPDE was hydrolyzed to benzo[a]pyrene tetrols prior to addition of the
hemoglobin, the level of protein associated radioactivity was
10-15% of the radioactivity associated with the complete
reactions, in which the BPDE was not hydrolyzed prior to
addition of the hemoglobin.
Effect of Hemoglobin and Benzo[a]pyrene Diol-epoxide
Concentrations on Extent of Adduct Formation
The level of incorporation was directly dependent on the
BPDE concentration for hemoglobin concentrations ranging
from 0.1 to 4 mM (Fig. 2A), while for fixed BPDE concentrations, the extent of adduct formation approached maximal
values as the hemoglobin concentration was increased (Fig.
2B). From 3 to 27% of the BPDE was incorporated into
protein (for 0.1 and 4 mM hemoglobin, respectively). Thus
i
Q
Q.
m
1.
Y
acidic acetone extractor!
neutral acetone extraction
acidic acetone extraction
neutral acetone extraction
PH
FIG. 1. Effect of pH and extraction procedure on extent of adduct
formation. Reaction mixtures (0.3 ml; 1.0 mM Hb; 10 /JM BPDE) were
incubated for 30 min in 20 mM Tris-imidazole-KCl buffer. Portions (0.05
ml) were extracted with either acidic or neutral acetone. The mean values
(n = 4) and standard deviations for complete reactions (two upper curves)
are corrected for the level of binding for the corresponding control reactions
(n = 4) (two lower curves).
at the greater hemoglobin concentrations, reaction of BPDE
with protein effectively competed with BPDE hydrolysis.
Kinetics of Adduct Formation
The pseudo first order formation of hemoglobin BPDE
adduct (1 mM hemoglobin) was complete in 15-20 min (Fig.
3). The apparent first order rate constant was 0.33 min"1
(inset). When hemoglobin was omitted from the reaction
and nonradioactive BPDE was used, the first order rate constant for BPDE hydrolysis was found to be 0.89 min"1 as
determined spectrophotometrically. Thus, the presence of
hemoglobin may have decreased the rate of hydrolysis of the
BPDE in a similar manner as observed for albumin (Roche et
al., 1985), liver microsomes (Dock et al., 1987), and various
cellular systems (Macleod et al., 1987).
Effect of Carbon Monoxide or Deoxygenation on the
Extent of Adduct Formation
The presence of carbon monoxide had no appreciable effect on the adduct level of BPDE incorporation either in
complete or control reactions. Thus because the presence of
carbon monoxide essentially prevents oxygen-heme interaction, BPDE incorporation was not appreciably due to hemedependent oxidation of BPDE or its products by hydrolysis.
Under air, hemoglobin is nearly saturated with oxygen.
The reactivities of some electrophilic sites in hemoglobin
97
HEMOGLOBIN-BENZO[a]PYRENE ADDUCTS
A 1.0
B
1.0
o.o
1
BPDE,
2
3
Hemoglobin, mM
4
FIG. 2. Effect of BPDE and hemoglobin concentrations on extent of adduct formation. After incubation of hemoglobin/BPDE mixtures at 37°C in
20 mM Tris-imidazole-KCl buffer, pH 7.4, for 30 min, a 10 to 50 nmol portion of hemoglobin was extracted with neutral acetone. Control reactions
were conducted only for 1 mM Hb. Identical data is shown in A and B, respectively, illustrating the effects of BPDE and Hb concentrations. Data
reported represents averaged data from duplicate experiments. The concentrations of hemoglobin (mM, Fig. 2A) and BPDE (/JM, Fig. 2B) are shown at
the ends of the curves
are substantially altered (sterically or electrostatically) by
the conformational changes associated with oxygen binding
(Perutz, 1970). Table 1 shows the level of adduct formation
for deoxyhemoglobin was 88% relative to the level for oxyhemoglobin. Thus, it is not likely that the sites of adduct
formation include any whose reactivity depends on the conformation of the protein.
Gel Filtration of Benzo[a]pyrene Diol-epoxide-Treated
Hemoglobin
The level of radioactivity in hemoglobin obtained by gel
filtration of control reaction mixtures (30-min treatment of
hemoglobin in preincubated buffer-BPDE mixture) (Table
2, Control A) was 10-15% relative to complete reaction (1
mM Hb, 7 //M BPDE, 30 min). This level was diminished
to about 2% if the control reaction mixture was applied to
the column immediately after addition of hemoglobin rather
than after a 30-min incubation (Control B). Most of the
radioactivity not associated with hemoglobin was retained
on Sephadex G-25 during extensive elution with water or the
glycine-KCN buffer, but was eluted with water—methanol
mixtures.
Treatment with Acid or Base
Ethyl acetate extraction of the hemoglobin fraction obtained from gel filtration removed about 10% of the radioacTABLE 1
Effect of Deoxygenation on the Extent of [3H]BPDE-hemoglobin
Adduct Formation"
Reaction No.
1
2
3
4
20
30
40
Reaction time, min
60
FIG. 3. Kinetics of adduct formation. Reaction mixtures (0.7 ml) contained 20 mM Tris-imidazole-KCl buffer, pH 7.4, 1.0 mM Hb, and 10 /IM
BPDE. Reactions were stopped by the addition of a 0.05-ml portion to HC1acetone (to rapidly hydrolyze BPDE). The complete reaction (upper curve
and inset) was initiated by the addition of BPDE to the hemoglobin solution.
For the control reaction (lower curve) BPDE was preincubated for 30 min
before addition of Hb at / = 0.
Oxygen
BPDE
Cone, (JJM)
No
No
Yes
Yes
9.30
9.55
9.38
9.46
Specific content
(pmol BPDE/nmol globin)
2.59
2.57
2.98
2.87
± 0.14
± 0.04
±0.13
±0.12
"The gas phases above the hemoglobin solutions (0.6 ml; 1.0 mM Hb in
40 mM Tris-HCl, 100 mM KC1, pH 7.4) sealed with a septum were purged
with humidified nitrogen at I5-2O°C with stirring. Conversion to deoxyhemoglobin was complete after 80-90 min as indicated by periodic measurement of changes in A73,_s4j. After warming the solutions to 37°C, BPDE
was added to the reaction vials that had either remained sealed (reactions
I and 2) or had been opened to allow reconversion to oxyhemoglobin
(reactions 3 and 4). After 20-25 min, the absence of oxyhemoglobin in
the sealed vials was confirmed, and three 0.05-ml portions of each reaction
mixture were extracted with neutral acetone. The values for the specific
content are mean ± SD (n = 5).
98
MYERS, SPINNATO, AND PINORINI
TABLE 2
Gel Filtration and Treatment with Acid and Base (n = 5)
Ethyl acetate"
Treatment
EXPERIMENT A
Gel filtration
Complete
Control A
Control B
EXPERIMENT Bc
0.2 M K phosphate,
pH 7.4, room
temperature
Complete
Control A
100° C, 10 minutes
Complete
Control A
0.1 N NaOH, 37*0,
2 hours
Complete
Control A
0.1 N HC1, 80°C,
3 hours
Complete
Control A
Specific* content
(pmol BPDE/mmol Hb)
extractable
radioactivity
(%)
As
obtained
Acetone
precipitation
11.3 ± 0.2
52.3 ± 2.8
70.7 ± 7.4
1.03
0.12
0.02
1.09 ± 0 . 0 3
0.079 ± 0.0004
0.008 ± 0.0004
10(0)
54
18(8)
30
65 (55)
39
57 (47)
38
"Aqueous solutions were extracted three times with two volumes of
ethyl acetate. The same complete and control A fractions were used for
experiments A and B.
'Specific contents of the gel filtration fractions were determined before
(as obtained) and after acetone precipitation.
c
The treatments with HO, NaOH, and phosphate buffer were as previously described. Values are for analyses in triplicate. Values in parentheses are corrected for noncovalently bound radioactivity, 10%.
tivity (Table 2). The data in Table 2 also show that of the
remaining radioactivity, about 50% was removed by heating
in 0.1 N NaOH or HC1, and about 10% by heating at neutral
pH. Storage for 3 hr at 30°C in 50% aqueous acetonitrile
containing 0.1% trifluoroacetic acid (conditions used for reversed phase HPLC) did not increase the amount of ethyl
acetate extractable radioactivity.
DEAE-Cellulose Ion-Exchange Chromatography of
Benzo[a]pyrene Diol-epoxide-Treated Hemoglobin
Some carcinogens and mutagens react with the electrophilic hemoglobin 09i cysteine sulfhydryl group (Pereira et
al., 1984; Segerback, 1983; Green etai, 1984). BPDE reacts
nonenzymatically with 2-mercaptoethanol, glutathione, and
cysteine (Michaud et al., 1983; Jerstrom et al., 1984). To
investigate the role of the 0n cysteine sulfhydryl groups in
the formation of BPDE-hemoglobin adducts, we examined
the effect of ISAL treatment on their ion chromatographic
behavior. BPDE hemoglobin (1 ITIM) obtained from gel filtration was treated for 6 hr at 5°C in a mixture containing
0.1 M sodium phosphate buffer, pH 7.0, 0.2 M KC1, and 10
rriM ISAL, followed by the addition of dithiothreitol (20 ITIM)
and gel filtration on a column equilibrated with glycineKCN buffer. ISAL reacted stoichiometrically with the 0n
cysteine sulfhydryl groups as determined by colorometric
analysis of SH content (Grassetti and Murray, 1967) and
spectrophotometric analysis (Rosen et al., 1973) of the covalently bound salicylic moiety in globin and hemoglobin.
Comparison of Figs. 4A and 4B shows that treatment with
ISAL, an anionic sulfhydrl reagent, increased the retention
volumes for all of the major and minor optically detected
hemoglobins as well as for all of the [3H]BPDE hemoglobin
adducts. Thus, because the adducts were subject to modification by ISAL, adduct formation does not appear to occur
via reaction of the BPDE with the 093 cysteine sulfhydryl
group. ISAL modified adduct fractions 1 and 2 were better
resolved after ISAL treatment (Fig. 4B), presumably because
of the greater elution volume.
Resolution of Globins by Reversed Phase High Pressure
Liquid Chromatography (HPLC)
In the ion-exchange DEAE cellulose chromatography system described above, hemoglobin migrates as a-^2 tetramers
in equilibrium with a0 dimers (Guidotti et al., 1963). In
contrast, a reversed phase HPLC system (Fig. 5) allows resolution of the individual a and 0 globins.
The results in Fig. 5 show that at least seven BPDE adducts were formed during the course of the reaction of BPDE
and hemoglobin. The widths for peaks 5 and 6 suggest that
they represent predominant single components, but unresolved components may have also been present. Their increased hydrophobicity (greater retention times relative to
native globin) is consistent with incorporation of the aromatic moiety of the benzo[a]pyrene nucleus with proteins.
Some of the less abundant adducts with retention times similar to or less than that for a globin may have arisen from
reaction of [3H]BPDE with either a globin, 0 globin variants,
or glycated a and 0 globins, while other adducts of BPDE
and hemoglobin may be structurally unique.
Figure 4A shows that two principal types of [3H]BPDEhemoglobin adducts were resolved by DEAE cellulose ion
exchange chromatography, one more basic than hemoglobin
Ao (fractions 1 and 2; 28 and 14% of recovered radioactivity,
respectively) and one more acidic (fractions 3 and 4; 18 and
DISCUSSION
8% of the recovered radioactivity, respectively). The specific
content of fraction 1 (the most highly purified) was at least
Metabolic activation of polynuclear aromatic hydrocar100 pmol/nmol hemoglobin (initially 3 pmol/nmol).
bons in human populations is evident from the detection of
HEMOGLOBIN-BENZO[a]PYRENE ADDUCTS
-BAL
40
i
30
"5
•o
c
m
3 20
o
I
60
B 40
45
30
30
g 20
u
E
99
-J
10
K
o.
o
Q.
U
i
0
i m
i<^»faL^—i
i
i
i
i
i
i B^»J|i
WO ZOO 300 400 600 MO 700 600 000
Elution Volume, mL
"0
100 200 300 400 600 600 700 800 BoS
Elution Volume, mL
FIG. 4. DEAE-cellulose ion-exchange chromatography and the effect of treatment with ISAL. Hemoglobin (2.0 ml, 2.0 DIM) in 20 mM Tris-HCl
(pH 7.4V100 mM KCl was treated with [JH]BPDE (13 JJM) and equal portions (20 mg) obtained by gel filtration were analyzed before (A) and after (B)
treatment with ISAL. The chromatographic procedure allows the separation of three forms of hemoglobin, hemoglobin Ao, hemoglobin A 2 , and hemoglobin
A| C . Fractions were collected and counted for BPDE associated radioactivity. Reaction of hemoglobin with ISAL was found to cause a corresponding
increase in the retention volumes of all three hemoglobins, indicating the availability of the 0n cysteine to react with ISAL after reaction of hemoglobin
with BPDE.
DNA adducts with benzo[a]pyrene metabolites in human
lung (Perera et al., 1982) and white blood cells (Shamsuddin
et al., 1985). Benzo[a]pyrene-hemoglobin and benzo[a]pyrene-DNA adducts are formed in benzo[a]pyrene-treated rats
and mice (Pereira and Chang, 1981; Shugart, 1985; Wallin et
al., 1987). Thus, it is likely that benzo[a]pyrene hemoglobin
adducts are present in benzo[a]pyrene exposed humans.
1600
1000 -
a
o
600 -
FIG. 5. Reversed-phase HPLC of BPDE-treated hemoglobin (54 fig in
10 ft\). Hemoglobin obtained by gel filtration was analyzed and 0.4-min
fractions were collected for radiochemical determination of BPDE-hemoglobin adducts. The chromatogram shows the separation of the a and 0
chains of human hemoglobin. At least seven adducts were detected, suggesting adduct formation at both a and fi chains.
The ISAL-dependent alteration of the ion chromatographic behavior of all [3H]BPDE hemoglobin adducts demonstrated that adduct formation does not involve the /3 93
cysteine sulfhydryl group. This conclusion is consistent with
the relatively small effect of pH on adduct formation since
the reactivity of the (39i cysteine sulfhydryl group of oxyhemoglobin are strongly dependent on pH (Hallaway et al.,
1980).
Ion exchange chromatography of BPDE-treated hemoglobin (Fig. 4A) revealed the presence of at least two types of
adducts: those more basic than hemoglobin AQ (fractions
1 and 2) and those distinctly or slightly more acidic than
hemoglobin Ao (fractions 3 and 4). Alteration of the ion
chromatographic mobility or the isoelectric point of hemoglobin resulting from reaction with a neutral reagent can
occur both because of reaction at an ionizable site or less
directly by alteration of local electrostatic interactions (Garel
et al., 1982). If we assume that the BPDE-dependent alterations of chromatographic behavior were due primarily to
reactions at ionizable sites, then the predominant basic adducts (fractions 1 and 2, Fig. 4) may have arisen via reaction
of BPDE at carboxyl groups, yielding the corresponding
esters. If this were true, the level of adduct formation would
depend relatively little on pH, as we observed, because the
concentration of nucleophilic carboxylate anion would not
vary significantly between pH 6 and 8. The effect of pH
must be interpreted cautiously because the rate of BPDE
hydrolysis increases with decreasing pH (Whalen et al.,
1977).
100
MYERS, SPINNATO, AND P1NORINI
Reaction of BPDE with a globin carboxyl group would
also be consistent with its reaction with acetate observed by
Yagi (1977) and also with the release during mild
acid hydrolysis (0.12 N HC1, 80°C) of 7,8,9,10-tetrahydroxybenzo[a]pyrene from globin of benzo[a]pyrene-treated
mice (Shugart, 1985) and BPDE-treated mouse globin (Wallin et at, 1987). Evidence has been reported for esterification
of hemoglobin by metabolites of dimethylnitrosamine (Kim
et ai, 1981) and a tobacco-specific nitrosamine (Carmella
and Hecht, 1987).
Glucose-dependent modification of N-terminal valine in
P globin is readily detected by conventional ion exchange
chromatography while modification of N-terminal valine in
a a globin has only a minor effect on ion chromatographic
properties (Shapiro et ai, 1980). Therefore, the acidic
BPDE-hemoglobin adducts (fractions 3 and 4, Fig. 4) may
have arisen via reaction of BPDE with basic groups (e.g.,
histidine or N-terminal valine) presumably decreasing their
pA" (as for benzyl amine). Alkene oxides and metabolites
of alkyl nitrosamines form adducts with basic nitrogens in
hemoglobin (Farmer et ai, 1986). Reaction of BPDE at sites
other than carboxyl, amino, or imidazole groups in both a
and f3 chains of hemoglobin cannot therefore be ruled out,
as shown by the number of adducts found under reversed
phase HPLC conditions (Fig. 5).
The assignment of two classes of adducts according to
their ion-chromatographic properties may be related to our
observation of differential sensitivity to hydrolysis, and the
observation of Wallin et al. (1987) that exposure of
[3H]BPDE-treated mouse globin to 0.1 N HCL (80°C) released only about 40% of the radioactivity, primarily as
tetrols.
The small difference we observed between the levels of
adduct formation for oxy and deoxyhemoglobin suggests
that sites of adduct formation do not include those whose
reactivity is substantially different for oxy and deoxy hemoglobin (including some carboxyl and amino groups) (Perutz,
1970). We cannot rule out that qualitatively different adducts
are formed by reaction with oxyhemoglobin and deoxyhemoglobin. An understanding of the significance of the results
will require further structural analysis.
In addition to providing information specifically about the
reaction of BPDE with hemoglobin and the isolation of the
adducts, our results suggest development of alternative approaches for characterization and analysis of hemoglobin
carcinogen adducts. The chromatographic approach we used
for evaluating the role of the /3 93 cysteine SH group in adduct
formation is applicable to the study of adducts with other
electrophiles. Analysis of the relative levels of adduct formation for oxy and deoxyhemoglobin can also provide indirect
information about sites of adduct formation.
Furthermore, the results suggest that it may be generally
possible, either with or without prechromatographic chemi-
cal modification of the native hemoglobin, to use ion chromatography or electrophoretic techniques to isolate fractions
enriched in hemoglobin—or globin carcinogen adducts
formed by reaction either with the 093 cysteine SH group,
basic nitrogen groups, or carboxyl groups.
REFERENCES
Bryant, M. S., Skipper, P. L., Tannenbaum, S. R., and Maclure, M. (1987).
Hemoglobin adducts of 4-aminobiphenyl in smokers and nonsmokers.
Cancer Res. 47(2), 602-608.
Carmella, S. G., and Hecht, S. S. (1987). Formation of hemoglobin adducts
upon treatment of F344 rats with the tobacco-specific nitrosamines 4(methylnitrosamino)-l-{3-pyridyl)-l-butanone and N'-nitrosonornicotine. Cancer Res. 47(10), 2626-2630.
Day, B. W., Naylor, S., Gan, L. S., Sahah, Y., Nguyen, T. T., Skipper,
P. L., Wishnok, J. S., and Tannenbaum, S. R. (1990). Molecular dosimetry
of polycyclic aromatic hydrocarbon epoxides and diol epoxides via hemoglobin adducts. Cancer Res. 50(15), 4611-4618.
Day, B. W., Skipper, P L., Rich, R. H., Naylor, S., and Tannenbaum,
S. R. (1991). Conversion of a hemoglobin alpha chain aspartate(47) ester
to N-(2,3-dihydroxypropyl)asparagine as a method for identification of
the principal binding site for benzo[a]pyrene anti-diol epoxide. Chem.
Res. Toxicol. 4(3), 359-363.
Dipple, A., Moschel, R. C , and Bigger, C. A. H. (1984) Polynuclear aromatic hydrocarbons. In Chemical Carcinogens (C. E. Searle, Ed.), Vol.
1, pp. 41-163. ACS, Washington, DC.
Dock, L., Martinez, M., and Jemstrom, B. (1987). Increased stability of
(+I-Y1 beta, 8 alpha-dihydroxy 9 alpha, 10 alpha epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene through interaction with subcellular fractions of rat
liver. Chem. Biol. Interact. 61, 31-44.
Ehrenberg, L., Hiesche, K. D., Osterman-Golkar, S., and Wennberg, I.
(1974). Evaluation of genetic risks of alkylating agents: Tissue doses in
the mouse from air contaminated with ethylene oxide. Mutat. Res. 24,
83-103.
Farmer, P. B., Shuker, E. G., and Bird, I. (1986). DNA and protein adducts
as indicators of in vivo methylation by nitrosatable drugs. Carcinogenesis
7(1), 49-52.
Garel, M. C, Beuzard, Y., Thillet, J., Domenget, C , Martin, J., Galacteros,
F.,and Rosa, J. (1982). Binding of 21 thiol reagents to human hemoglobin
in solution and in intact cells. Eur J Biochem. 123, 513-519.
Grassetti, D. R., and Murray, J. F. (1967). Determination of sulfhydryl
groups with 2,2'- or 4,4'-dithiodipyndine. Arch. Biochem. Biophys. 119,
41-49.
Green, L C , Skipper, P. L., Turesky, R. J., Bryant, M. S , and Tannenbaum,
S. R. (1984). In vivo dosimetry of 4-aminobiphenyl in rats via a cysteine
adduct in hemoglobin Cancer Res. 44(10), 4254-4259.
Guidotti, G., Konigsberg, W., and Craig, L. C (1963). On the dissociation
of normal adult human hemoglobin. Proc. Natl. Acad. Sci. USA 50, 774.
Hallaway, B. E., Hedlund, B., and Bensen, E. S. (1980). Studies on the
effect of reagent and protein charges on reactivity of the beta 93 sulfhydryl group of human hemoglobin using selected mutations. Arch. Biochem. Bwphys 203, 332-342.
Hutchins, D. A., Skipper, P. L , Naylor, S., and Tannenbaum, S. R. (1988).
Isolation and characterization of the major ftuoranthene-hemoglobin adducts formed in vivo in the rat. Cancer Res. 48(17), 4756-4761.
Jerstrom, B., Martinez, M., Svensson, S. A., and Dock, L. (1984). Metabolism of benzo[a]pyrene-7,8-dihydrodiol and benzo[a]pyrene-7,8-dihydrodiol-9,10-epoxide to protein-binding products and glutathione conjugates in isolated rat hepatocytes. Carcinogenesis 5, 1079-1085.
HEMOGLOBIN-BENZO[a]PYRENE ADDUCTS
Kim, S., Paik, W. K., Choi, J., Lothkar, P. D., and Magee, P. N. (1981).
Microsome-dependent methylation of erythrocyte proteins by dimethylnitrosamine. Carcinogenesis 2, 179-182.
MacLeod, M. C , Adair, G., Dickson-Black, D., Pevny, T., and Humphrey,
R. M. (1987). Stabilization of a reactive, electrophilic carcinogen, benzo[a]pyrene diol epoxide, by mammalian cells. Chem. Biol. Interact. 63,
279-289.
Michaud, D. P., Gupta, S. C , Whalen, D. L., Sayer, J. M., and Jerina,
D. M. (1983). Effects of pH and salt concentration on the hydrolysis of
a benz[alpha]pyrene 7,8-diol-9,10-epoxide catalyzed by DNA and polyadenylic acid. Chem. Biol. Interact. 44, 41-52.
Pereira, M. A., and Chang, L. W. (1981). Binding of chemical carcinogens
and mutagens to rat hemoglobin. Chem. Biol. Interact. 33(2-3), 3 0 1 305.
Pereira, M. A., Chang, J. L., Ferguson, J., and Couri, D. (1984). Binding
of chloroform to the cysteine of hemoglobin. Chem. Biol. Interact. 51,
115-124.
Perera, F. P., Poirier, M. C , Yupsa, S. H., Nakayama, J., Jaretzki, A.,
Curnen, M. M., Knowles, D. M., and Weinstein, I. B., (1982). A pilot
project in molecular camcer epidemiology: Determination of benzo[a]pyrene-DNA adducts in animal and human tissues by immunoassays.
Carcinogenesis 3, 1405.
Perutz, M. F. (1970). Stereochemistry of cooperative effects in haemoglobin. Nature 228, 726.
Roche, C. J., Zinger, D., Geacintov, N. E., and Harvey, R. G. (1985).
Enhancement of stability of anti-BaPDE by complex formation with serum albumin. Cancer Biochem. Biophys. 8, 35.
Rosen, N. L., Bishop, L., Burnett, J. B., Bishop, M., and Colman, R. F.
(1973). Methionyl residue critical for activity and regulation of bovine
liver glutamate dehydrogenase. J. Biol. Chem. 248, 7359-7369.
Schroeder, W. A., and Huisman, T. (1980). The Chromatography of Hemoglobin, pp. 7 4 - 8 1 . Marcel Dekker, New York.
Segerback, D. (1983). Alkylation of DNA and hemoglobin in the mouse
following exposure to ethene and ethene oxide Chem. Biol. Interact. 45,
139-151.
101
Shamsuddin, A., Sinopoli, N. T., Hemminki, K., Boesch, R. R., and Harris,
C. C. (1985). Detection of benzo[a]pyrene:DNA adducts in human white
blood cells. Cancer Res. 45, 66.
Shapiro, R., McManus, M. J., Zalut, C , and Bunn, H. F. (1980). Sites of
nonenzymatic glycosylation of human hemoglobin. J. Biol. Chem. 255,
3120.
Shugart, L. (1985). Quantitating exposure to chemical carcinogens: In vivo
alkylation of hemoglobin by benzo[a]pyrene. Toxicology 34, 211.
Skipper, P. L., Bryant, M. S., and Tannenbaum, S. R. (1988). Determination
of human exposure to carcinogenic aromatic amines from hemoglobin
adducts in selected population groups. In Carcinogenic and Mutagemc
Responses to Aromatic Amines and Nitroarenes (C. M. King, L. J. Romano, and D. Schuetzle, Eds.), pp. 6 5 - 7 1 . Elsevier, New York.
Tentori, L., and Salvati, A. M. (1981). Hemoglobinometry in human blood.
Methods Enzymol. 76, 707.
T6mqvist, M., Mowrer, J., Jensen, S., and Ehrenberg, L. (1986). Monitoring
of environmental cancer initiators through hemoglobin adducts by a modified Edman degradation method. Anal. Biochem. 154, 255.
Wallin, H., Jeffrey, A. M., and Santella, R. M. (1987). Investigation of
benzo[a]pyrene-globin adducts. Cancer Lett. 35, 139.
Weston, A., Rowe, M. L., Manchester, D. K., Farmer, P. B., Mann, D. L.,
and Harris, C. C. (1989). Fluorescence and mass spectral evidence for
the formation of benzo[a]pyrene anti-diol-epoxide-DNA and hemoglobin
adducts in humans. Carcinogenesis 10(2), 251-257.
Whalen, D. L., Montemarano, J. A., Thakker, D. R., Yagi, H., and Jerina,
D. M. (1977). Changes of mechanisms and product distributions in the
hydrolysis of benzo[a]pyrene-7,8-diol-9,10-epoxide metabolites induced
by changes in pH. J. Am. Chem. Soc. 99, 5522.
Whalen, D. L., Ross, A. M., Montemarano, J. A., Thakker, D. R., Yagi,
H., and Jerina, D. M. (1979). General acid catalysis in the hydrolysis of
benzo[a]pyrene 7,8-diol-9,10-epoxides. J. Am Chem Soc. 101, 5086.
Yagi, H., Thakker, D. R., Hernandez, O., Koreeda, M., and Jerina, D. M.,
(1977). Synthesis and reactions of the hightly mutagenic 7,8-diol-9,10epoxides of the carcinogen benzo[a]pyrene. J. Am, Chem. Soc. 99, 1604.