Biomarkers of internal exposure/dose Methods to quantify adducts to protein and DNA by LC/MS studied with

benzo[a]pyrene and isocyanates

Emelie Westberg

©Emelie Westberg, Stockholm University 2015

ISBN 978-91-7649-074-7

Printed in Sweden by US-AB, Stockholm 2015

Distributor: Department of Materials and Environmental Chemistry, Stockholm University

Protein image on the front cover; from RCSB PDB (www.rcsb.org) of PDB ID 1AO6

Till min familj.

Table of contents

1 Introduction ......................................................................................... 9 1.1 Background ...................................................................................................... 9

1.1.1 History .................................................................................................. 10 1.1.2 Cancer .................................................................................................. 10

1.2 Measurement of electrophilic compounds ................................................ 11 1.3 Protein adducts ............................................................................................. 12 1.4 DNA adducts .................................................................................................. 13 1.5 Urine metabolites .......................................................................................... 14 1.6 Analytical methods for adduct measurement .......................................... 15 1.7 Micronuclei for measurement of genotoxic damage ............................... 16

2 Polycyclic aromatic hydrocarbons (PAH) ..................................... 17 2.1 Formation ....................................................................................................... 17 2.2 Exposure ......................................................................................................... 17 2.3 Absorption and distribution ......................................................................... 18 2.4 Metabolism ..................................................................................................... 18

2.4.1 Metabolic differences ......................................................................... 21 2.5 Excretion ........................................................................................................ 21 2.6 Toxicity ........................................................................................................... 22 2.7 Analysis of adducts from BP ....................................................................... 23

3 Aim and specific background of the thesis ................................. 24 3.1 Aim of thesis – Paper I–III.......................................................................... 24 3.2 Specific background ..................................................................................... 25

3.2.1 Adduct formation to proteins ........................................................... 25 3.2.2 Adduct formation to DNA .................................................................. 26 3.2.3 The need for a method to quantify protein adducts .................... 26 3.2.4 Model substances used in this thesis.............................................. 27

4 Methods for quantitative analysis of bulky adducts to SA and

DNA .................................................................................................... 28 4.1 Development of a method for measurement of adducts to SA ............ 28

4.1.1 Precipitation of SA for analysis of BPDE-His adducts .................. 28 4.1.2 Hydrolysis of the protein................................................................... 28 4.1.3 Enrichment of the adducts ............................................................... 31 4.1.4 Analysis of DBPDE and DBADE ........................................................ 31 4.1.5 Internal standard and quantification of the adducts ................... 31

4.1.6 LC/MS-MS analysis ............................................................................ 33 4.1.7 The evaluated procedures combined to a method ....................... 33

4.2 Methods for measurement of adducts to DNA ........................................ 35 4.2.1 Qualitative and quantitative standards for analysis of BPDE-dG

adducts in DNA ................................................................................................... 35 4.2.2 The impact of modifications of the method for enzymatic

hydrolysis of DNA ............................................................................................... 36 4.2.3 LC/MS-MS analysis ............................................................................ 39

4.3 Discussion about the methods developed ................................................ 39

5 Investigations of adduct formation to SA and DNA from BPDE

– in vitro and in vivo ................................................................................. 40 5.1 In vitro experiments ..................................................................................... 40

5.1.1 Alkylation of SA from mouse and human with (+)-anti-BPDE,

(±)-anti-BPDE and (±)-syn-BPDE (Paper I–III)........................................... 40 5.1.2 Adduct levels in SA and in DNA after incubation of whole blood

with (±)-anti-BPDE (Paper II) ......................................................................... 41 5.2 In vivo experiments ..................................................................................... 42

5.2.1 BP adducts to SA in mice (Paper II) ............................................... 42 5.2.2 Identification and quantification of specific stereo-isomeric

adducts in SA and DNA from mice (Paper III) ............................................. 43 5.3 Discussion ...................................................................................................... 44

6 Isocyanates ....................................................................................... 46 6.1 Aim of Paper IV and V.................................................................................. 46 6.2 Formation and usage of isocyanates ......................................................... 46 6.3 Exposure and Toxicity .................................................................................. 46 6.4 Biomonitoring of isocyanates ..................................................................... 47 6.5 Analysis of isocyanate adducts to the N-terminal valine of Hb (Paper

IV) ......................................................................................................................... 49 6.6 Evaluation of VH as a biomarker of early renal toxicity (Paper V) ...... 50

6.6.1 Uremia .................................................................................................. 50 6.6.2 Biomarker for adverse effect to kidneys caused by a radioactive

cytostatic drug .................................................................................................... 51 6.6.3 Experiment .......................................................................................... 51

6.7 Summary ........................................................................................................ 52

7 Final discussion ................................................................................. 53

8 Svensk sammanfattning ................................................................. 56

9 Acknowledgements .......................................................................... 59

Abbreviations

177Lu-octreotate

177Lu-[DOTA

0,Tyr

3]-octreotate

ALKP alkaline phosphatase

AUC area under the concentration time curve

b.w. body weight

BP benzo[a]pyrene

BPDE benzo[a]pyrene-7,8-diol-9,10-epoxide

Bq becquerel

CV coefficient of variation

CYP cytochrome P450

dA deoxyadenosine

DBA dibenzo[a,h]anthracene

DBADE dibenzo[a,h]anthracene-3,4-diol-1,2-epoxide

DBP dibenzo[a,l]pyrene

DBPDE dibenzo[a,l]pyrene-11,12-diol-13,14-epoxide

DE diolepoxide

dG deoxyguanosine

dN nucleotide

DNA deoxyribonucleic acid

EH epoxide hydrolase

ELISA enzyme-linked immunosorbent assay

ESI+ positive electrospray ionisation

GC gas chromatography

Hb haemoglobin

His histidine

HPLC high performance liquid chromatography

ITCs isothiocyanates

LC liquid chromatography

LOD limit of detection

LOQ limit of quantification

Lys lysine

MDI methylene diphenyl diisocyanate

MIC methyl isocyanate

MN micronuclei

MRM multiple reaction monitoring

MS mass spectrometry

NCEs normochromatic erythrocytes

PAHDE polycyclic aromatic hydrocarbon diolepoxide

PAH polycyclic aromatic hydrocarbon

PCEs polychromatic erythrocytes

PIC phenyl isocyanate

PIS product ion scan

PREC precursor ion scan

Pro proline

PVH phenyl valine hydantoin

RBCs red blood cells

RBP4 retinol binding protein 4

RNA ribonucleic acid

S/N signal-to-noise ratio

SA (hSA, mSA) serum albumin (human SA, mouse SA)

SPE solid phase extraction

SSTR somastostatin receptors

SVPD snake venom phosphodiesterase

TDI toluene diisocyanate

Trf Refs transferrin-receptor positive reticulocytes

Tyr tyrosine

UV ultra violette

Val valine

WBCs white blood cells

VH valine hydantoin

List of papers

The thesis is based on the following publications and manuscripts, which are

referred to in the text by their roman numerals, Paper I-V. Some unpublished

results are also included in the thesis.

Paper I Westberg E., Hedebrant U., Haglund J., Alsberg T., Eriksson J., Seidel

A., Törnqvist M. (2014) Conditions for sample preparation and quanti-

tative HPLC/MS-MS analysis of bulky adducts to serum albumin with

diolepoxides of polycyclic aromatic hydrocarbons as models. Anal Bio-

anal Chem 406 (5) 1519-1530 DOI:10.1007/s00216-013-7540-7

Paper II Westberg, E., Singh, R., Hedebrant, U., Koukouves, G., Souliotis, V.,

Farmer, P., Segerbäck, D., Kyrtopoulos, S., Törnqvist, M.. (2014) Ad-

duct levels from benzo[a]pyrene diol epoxide: Relative formation to

histidine in serum albumin and to deoxyguanosine in DNA in vitro and

in vivo in mice measured by LC/MS-MS methods. Toxicology Letters

232 (1) 28-36 DOI: 10.1016/j.toxlet.2014.09.019

Paper III Westberg, E., Motwani, H., Lindh, C., Jenssen, D., Abramsson-

Zetterberg, L., Törnqvist, M. Measurement of specific biomarkers of in-

ternal dose and genotoxic effect in benzo[a]pyrene-exposed mice and

their use in relative risk estimation. (manuscript)

Paper IV Davies, R., Rydberg, P., Westberg, E., Motwani, H., Johnston, E.,

Törnqvist, M. (2010) A new general pathway for synthesis of reference

compounds of N-terminal valine-isocyanate adducts. Chemical Re-

search in Toxicology, 23(3), 540-546 DOI:10.1021/tx900278p

Paper V Dalmo J., Westberg E., Barregard L., Svedbom L., Johansson M.,

Törnqvist M., Forssell-Aronsson E. (2014) Evaluation of retinol bind-

ing protein 4 and carbamoylated haemoglobin as potential renal toxicity

biomarkers in adult mice treated with 177

Lu-octreotate. EJNMMI Res

4:1-9 DOI:10.1186/s13550-014-0059-x

Permissions to reproduce the publications were kindly obtained from the

publishers.

9

1 Introduction

1.1 Background

During life all species are surrounded by a variety of chemicals, both natural

and man-made. It is important to remember that all chemicals are not

synthetic and harmful. Most of them are essential for the life, although, some

compounds are toxic, genotoxic or mutagenic and many have unknown toxic

characteristics. Chemicals which are harmful to humans can cause a variety

of effects, e.g. cancer, allergy, or neurological effects.

The risk of toxic effects is dependent on the dose of exposure. Even com-

pounds essential for biochemical mechanisms are often toxic at very high

doses. Fundamentally, there are two major principles for dose-risk relation-

ships: a) with a threshold, this means that there is no risk below a threshold

dose b) linear dose-risk without threshold, which means that even the

lowest dose constitute a risk. The linear relationship is assumed to be valid

for genotoxic chemicals (cancer risk increasing compounds) at the low doses

relevant for human exposure situations. For the Swedish population cancer is

the second most common cause of death (after cardiovascular diseases) (The

National Board of Health and Welfare 2014).

In this thesis focus is on methods for measurement of biomarkers of internal

dose of certain chemicals using liquid chromatography (LC) /mass spec-

trometry (MS). Internal dose reflects the uptake, distribution, metabolism

and excretion of the chemical. Analyses with MS methods (both gas

chromtographic (GC) and with LC) are sensitive, specific and can provide

structural information of the compounds of interest.

In the first part of this thesis, methods for biomarkers for internal dose from

bulky electrophilic compounds are developed, using polycyclic aromatic

hydrocarbons (PAH) as model compounds for bulky adducts. In the second

part of this thesis, an LC/MS method is established for the characterisation

of adducts from isocyanates in Hb. The method was subsequently used for

the investigation of the potency of this adduct as a biomarker for early renal

toxicity.

10

1.1.1 History

In 1775 the English surgeon sir Percivall Pott (1714 –1788) demonstrated an

association between exposure to soot and a high incidence of scrotal cancer,

called chimney sweeps' carcinoma. Pott was one of the first to show that

cancer may be caused by an environmental pollutant and this was the first

occupational cancer reported (Brown et al. 1957).

In the end of the 1950’s adducts to DNA were reported for the first time

(Wheeler et al. 1957). In the late 1960’s Miller and Miller proposed the

concept that ultimate carcinogens (the active form) are electrophiles and that

they exert their biological effects by reacting with cellular nucleophiles

(Miller 1970). In the beginning of the 1970’s Ehrenberg and co-workers

identified the potential of protein adducts as a surrogate for DNA adducts for

measurement of dose in vivo (Ehrenberg et al. 1974). Now, in the the 21st

century the screening for adducts from unknown electrophiles in whole

genomes or proteins (adductomics) is a new approach under fast

development (Balbo et al. 2014).

1.1.2 Cancer

The International Agency for Research on Cancer (IARC) is an organ within

the World Health Organization (WHO). In the past 30 years, IARC has

evaluated the cancer-causing potential of more than 900 candidates and

classified them into five groups (Table 1).

Table 1. Classification made by IARC and number of compounds, mixtures and industrial processes from

their cancer-causing potential.

Group Classification Number of agents

Group 1 Carcinogenic to humans 114

Group 2A Probably carcinogenic to humans 69

Group 2B Possibly carcinogenic to humans 283

Group 3 Unclassifiable as to carcinogenicity in humans 504

Group 4 Probably not carcinogenic to humans 1

The majority of the candidates are classified in the groups of probable,

possible, or unknown risk. About 100 agents are classified into group 1 as

"carcinogenic to humans" but less than 50 of those are classified as individu-

al compounds (e.g. benzo[a]pyrene), the others are mixtures of compounds

or production processes (e.g. coal and aluminium production) (IARC 2014).

Many known carcinogens are alkylating compounds (e.g. ethylene oxide in

Group 1, Table 1). Epidemiological studies have been used to characterize

the carcinogenicity for around 50 compounds (Tomatis et al. 1989).

11

However, in traditional epidemiological studies specific exposures at high

levels, like in certain occupational exposures, are needed to make them use-

ful for detection/estimation of cancer risk factors. The usefulness of epide-

miology for determination if a specific compound cause cancer is limited

because of the low resolution, e.g. due to the individual variation, simultane-

ous exposure to other compounds in addition to high background cancer

incidence (Ehrenberg et al. 1996; Granath et al. 2001).

Cancer is developed in a multistep process where chemical modifications of

DNA may cause mutations that convert a normal cell to a cancer cell.

Mutations can arise from exposure to e.g. radiation or chemicals.

If mutations occur in the DNA at critical sequences oncogenes can be acti-

vated and tumour suppressor genes inactivated, which may lead to the initia-

tion and progression of cancer (Hanahan et al. 2011; La et al. 1996).

For risk assessment of an adverse health effect like cancer, as a consequence

of exposure to electrophilic and genotoxic compounds/metabolites, it is

essential to estimate the internal dose of the compound. The internal dose is

reflecting the concentration over time of the ultimate carcinogen. The rela-

tion between effect and exposure has to be estimated after measurements of

the chemicals in vivo in order to calculate a target- or internal dose

(Figure 1). This thesis will mainly focus on quantification and discussion of

biomarkers of the internal dose of bulky compounds like PAH.

Figure 1. Scheme on events from an exposure of a genotoxic compound to the development of an adverse

health effect and possible biomarkers to detect and measure these events.

1.2 Measurement of electrophilic compounds

Genotoxic compounds, which are electrophilic compounds/metabolites, are

usually too short-lived to be accurately measured in vivo. Instead of analys-

ing the reactive compound per se stable products (adducts, addition products

and also from the latin word adductus meaning “tighten”) can be measured.

Adducts are formed between electrophilic sites in the reactive compound and

nucleophilic sites (usually O, N and S) in biomacromolecules (like DNA,

serum albumin (SA) and haemoglobin (Hb)). Specific MS-analysis of ad-

Exposure Adverse health effects

Internal exposure

Internal dose

Early biological effects

Concentration of the parent chemical and/or metabolites in urine, blood, tissues, exhaled air

Integrated concentration of the parent chemical and/or metabolites; adducts to proteins or DNA

Biological endpoints; DNA adducts Micronuclei Mutations

Target doseDNA adducts

Cancer

12

ducts that contain a tag from the nucleophile, for instance amino acids, re-

duces the risk that compounds which have not been covalently bound are

measured and estimated to originate from adducts.

The stability of the adduct and the efficiency of adduct formation of a com-

pound to a biomacromolecule are important for the usefulness of the adduct

as a biomarker. The stability of the adduct is largely depending on the cell

turnover or the life-span of the biomacromolecule, the characteristics of the

chemical bond formed and whether the adduct is removed by in vivo pro-

cesses (DNA repair). The turnover time for chemically stable adducts

formed in a protein is related to the turnover of the biomacromolecule which

could be specific for different species. Stable adducts make it possible to

calculate the in vivo doses of the precursor electrophile if the rate of adduct

formation is known (theory carefully developed in (Ehrenberg et al. 1974;

Ehrenberg et al. 1983)).

1.3 Protein adducts

Proteins used for adduct measurement are primarily the blood proteins SA

and Hb. These proteins have no repair systems and have known life-spans,

which simplifies calculations of the in vivo dose after chronic exposure of an

electrophilic substance (Ehrenberg et al. 1983).

SA is located free in the blood plasma and consists of three domains I-III,

each domain is divided into two subunits A and B. SA is a single chain pro-

tein consisting of 585 amino acids with a total molecular weight of 66 kDa

(Peters 1995). It is a transport protein, carrying fatty acids, various drugs and

hormones but is also important for the maintenance of colloidal osmotic

blood pressure. Approximately 60% of the total protein content in plasma

constitutes of albumin. Albumin is primarily produced in the liver by the

hepatocytes, in which also most xenobiotics are metabolised (Peters 1995).

Adult human Hb (hHbA hereafter just called hHb) is a blood protein with

four non-covalently bound subunits (two α and two β subunits, which are

similar in both size and structure) (Jain 1993).

The proteins are relatively long-lived in humans (Table 2). Adducts to these

proteins therefore reflects the mean exposure during several weeks/months

instead of the exposure only a few days ago (cf. urine metabolites). Electro-

philic compounds have been shown to bind to several amino acids in the

blood proteins, e.g. histidine (His), cysteine, lysine (Lys) and valine (Val)

(Rubino et al. 2009; Tornqvist et al. 2002).

13

Table 2. Information about biomacromolecules in blood used as biomarkers.

SA Hb DNA

Type of blood fraction Free in the

blood plasma1

In the erythrocyte

fraction

Buffy-coat2,

in the leucocytes

% of total blood volume 55% plasma 45% red blood cells

( RBCs)

<1% white blood cells

(WBCs)

Turnover of stable adducts

human ~20 days3 ~124 days4

Complex kinetics in all DNA,

due to repair systems

mouse 1.9 days5 40

Conc. in whole blood 35 mg/mL 150 mg/mL ~0.02 mg/mL6

Molecular weight (MW)

~66 kDa ~64 kDa MW/dN: 309 Da

1 The plasma consists of up to 95% water (by volume) in addition to proteins, fats and sugars.

2 The buffy-coat is composed of leucocytes and thrombocytes. 3 (Sabbioni et al. 1987) 4 (Furne et al. 2003) 5 (Jain 1993) 6 using the procedure described in the “genomic DNA handbook” from Qiagen.

Even adducts to non-blood proteins have been investigated as biomarkers,

e.g. histones and collagen. Histones from liver or brain from a mouse

theoretically could reflect the exposure from one fifth of the lifetime of the

mouse. Collagen could reflect the exposure of the species whole lifetime,

which would be useful in studies of exposures in the past (Skipper et al.

1990).

1.4 DNA adducts

Many mutagens and carcinogens are known to form DNA adducts, by direct

acting or after metabolic activation (e.g. PAH). There are several nucleo-

philic sites in DNA were reactions can occur. The specific sites for adduct

formation depend on the electrophilic species, the nucleophilicity of the site

in the DNA, steric factors and the nucleotide sequence (La et al. 1996). The

most nucleophilic sites in the DNA bases guanine and adenine are two of the

ring nitrogens (N-3 and N-7) (Figure 2). In the helical DNA, deoxy-

guanosine (dG) N-7 is more exposed and therefore more accessible for reac-

tion than e.g. N-3 in deoxyadenosine (dA). This makes N-3 in dA less reac-

tive than N-7 in dG (La et al. 1996).

Non-bulky, low-molecular weight compounds react primarily at the most

accessible and nucleophilic sites in the DNA (e.g. N-7). Adducts originating

from bulky aromatic compounds such as PAH and aromatic amines have

different DNA adduct spectra. The diolepoxide (DE) metabolites from PAH

reacts predominantly with the exocyclic amino groups of dG and dA (N2 and

14

N6, respectively, superscript indicates exocyclic nitrogens). DNA adducts are

measured after enzymatic degradation to deoxynucleotides (dN) or deoxunu-

cleosides.

Both the formation and repair of DNA adducts are depending on the dN-

sequence and may explain some observed mutational hot spots for carcino-

gens and mutagens (La et al. 1996). In order to identify and to correct such

damages there are active repair systems in the DNA. Interindividual varia-

tions of the repair systems for DNA adducts, make the lifetime of DNA ad-

ducts unspecific. Therefore the in vivo exposure of a compound is often

more difficult to calculate from the DNA adduct levels than from adduct

levels in a protein with a known lifetime (e.g. SA) (Ragin et al. 2008). An-

other difficulty with DNA is that the amount is very low compared to the

amount of Hb and SA in blood (Table 2), which is an accessible matrix to

sample in humans.

Figure 2. The four deoxynuclosides in DNA. On dG and dA positions are marked where bulky com-

pounds (e.g. metabolites from PAH) primarily react to form adducts. These positions are not the most

nucleophilic in the respective base but the most accessible positions for large compounds.

1.5 Urine metabolites

Urine is the excretion pathway of water soluble compounds. The hydro-

philicity of xenobiotics could be increased through conjugation with glutha-

tione which after further metabolism is excreted as mercapturic acid or thi-

oethers. Also conjugations with other endogenous compounds, facilitating

excretion in urine, occur. Metabolites or conjugates of electrophilic com-

pounds as well as nucleotides from DNA repair processes in vivo can be

monitored in urine (e.g. (Kotova et al. 2005).

The sampling of urine is a non-invasive method and it is easy to get large

sample volumes. Urine metabolites are excreted from the body within hours

deoxyadenosine

(dA)

deoxyguanosine

(dG)

N

NH

O

O

deoxythymidine

(dT)

deoxycytidine

(dC)

N

NN

N

NH2

dR

NH

N

N

O

NH2N

dR dR

O

HOH

HH

HH

HO Position for the

basedR =

N

N

NH2

O

dR

1

2

3

4

567

8

9 1

2

345

6

1

2

3

4

567

8

9 1

2

34

5

6

PAH

PAH

15

to days after an exposure and reflect a very recent exposure compared to

protein and DNA-adducts. For instance, the half-life of 1-OH-pyrene, the

most common biomarker for PAH exposure in urine, is approximately 4 h

after ingestion of PAH (Buckley et al. 1992).

1.6 Analytical methods for adduct measurement

MS is the most widely used technique for identification and quantitation of

adducts. Many DNA adducts are thermo labile and LC/MS is therefore pre-

ferred instead of GC/MS. By derivatisation of the adducts and a good MS

instrument it is possible to detect protein adducts in the lower range of

femtomoles injected (von Stedingk et al. 2011).

Immunoassays, with specific antibodies, such as radioimunoassays or solid

phase enzyme immunoassays, e.g. enzyme-linked immunosorbent assay

(ELISA), have been extensively applied for the analysis of the modification

of DNA or proteins by carcinogenic compounds. The methods for immuno-

assays were first developed in the 1950’s (Yalow et al. 1959) (awarded the

Nobel price 1977) and are continuously in use and developing to reach

higher specificity and sensitivity. The immunoassay methods can detect

DNA adducts in the range of one adduct per 106 to 10

8 dN. The detection

levels are low but since the methods are relatively unspecific there can be

problems with cross-reactivity to unmodified DNA or adducts with similar

structure. On the other hand, cross-reactivity may be advantageous if the

exposure from a group of substances is aimed to be monitored instead of the

detection of a specific compound. The need to produce antibodies for each

DNA adduct is a major limitation of immunoassays (Phillips et al. 2000;

Plaza et al. 2000).

The 32

P-postlabeling technique has been extensively used for DNA adduct

investigations since its development more than 30 years ago. Briefly,

labelled [γ-32

P]-adenosine triphosphate is enzymatically incorporated to a

modified dN, which are chromatographically separated from other dN and

detected by the decay of the phosphorous by e.g. an on-line radioactivity

flow detector or autoradiography. The 32

P-postlabeling technique is relative-

ly inexpensive to establish and the detection levels are very low. Adducts

can be detected at levels as low as one adduct per 1010

dN and small amounts

(< 10 µg) of DNA is needed (Phillips 1997; Randerath et al. 1981). Yet,

there are limitations with this method, e.g. that the use of the radioactive

phosphorus is a potential health risk as well as, the absence of internal stand-

ards for calculation of recovery and efficiency of the labelling. In addition,

no structural information from the analyte is obtained, which leads to diffi-

culties to identify and quantify the adducts (Balbo et al. 2014).

16

32P-Postlabeling have been used in a large number of studies of bulky com-

pounds such as PAH. In addition to the methods discussed here there are

also other methods (e.g. fluorescence) used for analysis of adducts.

1.7 Micronuclei for measurement of genotoxic damage

Mitosis is the process prior to the cell division which results in two

genetically identical daughter cells. During mitosis the copied chromosomes

are incorporated into the new nuclei. Due to clastogens[1] pieces of

chromosomes or whole chromosomes are sometimes not completely incor-

porated into the nuclei. These pieces/chromosomes can form micronuclei

(MN), which are small DNA-containing bodies, in the cytoplasm outside the

main nucleus.

The red blood cells (RBCs) are developed in the bone marrow, from

erythroblasts (nucleated cells) via very young reticulocytes (Trf Rets) with-

out any nucleus to polychromatic erythrocytes (PCEs). TrfRets have

receptors for transferrin at the surface of the cell membrane. During the ma-

turity process of the RBCs, the Trf Rets turn into PCEs, which still contain

RNA but the nucleus is eliminated. In mice, about one day after the entrance

to the peripheral blood system the PCEs mature to normochromatic

erythrocytes (NCEs) without RNA.

A well-established short term test for genotoxicity of a compound in vivo is

the MN assay. By counting the number of MN, which increases after

exposure to genotoxic compounds, the genotoxic potency can be quantified.

In a relatively new method for measurement of the frequency of MN, the

PCE are counted using a flow cytometry method. Approximately 200 000

PCE are counted from a blood sample. This increases the sensitivity of the

assay compared with the older and still widely used MN assay for rodents

where only around 2000 PCE are counted using an ordinary microscope

(Abramsson-Zetterberg 2003). In addition, since the human spleen efficient-

ly remove all RBCs containing a micronucleus it is not possible to use the

“old” method in human samples, it is necessary to restrict the analysis to the

Trf Ret (Abramsson-Zetterberg et al. 2008).

[1] An agent that cause breaks in chromosomes (e.g. radiation, certain chemicals).

17

2 Polycyclic aromatic hydrocarbons (PAH)

2.1 Formation

PAH refers to a large class of compounds built up of carbons and hydrogens

combined in two (naphthalene) or more (>30) fused aromatic rings

(Figure 3). PAH are released from incomplete combustion of organic mate-

rials (nearly all combustion processes) and are widely spread in the envi-

ronment e.g. from automobile emissions and wood burning (IARC 2010).

PAH are formed in the preparation of smoked or charcoal grilled food. A

variety of PAH are present in high concentrations in tobacco smoke (IARC

2010). The most studied PAH is benzo[a]pyrene (BP), which in 1933 was

the first PAH to be synthesized and isolated from coal (Cook et al. 1933).

Figure 3. Examples of polycyclic aromatic hydrocarbons (PAH). Benzo[a]pyrene, dibenzo[a,l]pyrene

and dibenzo[a,h]anthracene. The pyrene and anthracene parts of the molecules are drawn in orange.

2.2 Exposure

There are occupations where workers are exposed to high levels of PAH, e.g.

coke oven workers (Boysen et al. 2003; Melikian et al. 1999). For the

general population the main PAH exposure is from the diet and by

inhalation, with increased exposure for smokers and during wintertime due

to residential heating (IPCS 1998). The intake of BP from food is estimated

to be 1-2 ng/kg of bodyweight (b.w.) (Abramsson-Zetterberg et al. 2014;

Falco et al. 2003). The European Food Safety Authority (EFSA) and the

United States Environmental Protection Agency (USEPA) have estimated

the daily intake of BP to approximately 5 ng/kg b.w. per day. The predicted

benzo[a]pyrene(BP)

dibenzo[a,l]pyrene(DBP)

dibenzo[a,h]anthracene(DBA)

18

intake may vary depending on the selection of food used for the estimation

of the intake and on the analytical methods used to determine the PAH con-

centrations. The cancer incidence rate due to the exposure of BP in food has

been calculated to about four persons/year in Sweden (the Swedish National

Food Agency 2012). It should be remembered that the exposure is not only

from food. In the ambient air there are variations over several magnitudes in

the concentrations of individual PAH (including BP) (<0.1–100 ng/m3)

(IPCS 1998). Smokers may be exposed for approximately twice as much BP

as non-smokers and high concentrations of several others PAH. The carcino-

genicity caused from BP is estimated to only represent a small part

(0.17–4%) of the carcinogenicity caused from PAH exposure (Boysen et al.

2003). The simultaneous exposure of several PAH and the limitations with

epidemiological studies motivate biomonitoring for the evaluation of internal

PAH exposure as an compliment to environmental monitoring (Chung et al.

2013).

2.3 Absorption and distribution

The size and lipophilicity of an individual PAH play a major role for its ab-

sorption from the diet. The absorption of BP has been estimated to 12–99%

after oral exposure and the highest plasma concentrations were reported 2–6

hours after administration. These figures are from studies on different mam-

mals (rat, goats and pigs) and are obtained with different methods with a

large variation, but consistent is that smaller PAH (3- and 4-rings) are ab-

sorbed to a higher extent than larger PAH (EFSA 2008; Ramesh et al. 2004).

Investigations of the distribution of PAH in the body (mainly studied in

rodents after intravenous or oral administration) show that the parent

compound and its metabolites are found in almost all tissues. Several PAH

and their metabolites have been shown to cross the blood-brain barrier (e.g.

BP (Saunders et al. 2002)) and the placenta in both humans (Madhavan et al.

1995) and pregnant rodents (refs. in (EFSA 2008)).

2.4 Metabolism

PAH are metabolised to a variety of compounds (e.g. epoxides, quinones and

diolepoxides) by cytochrome P450s (CYPs) and other enzymes in both

phase-I and phase-II metabolism systems. BP is oxidized to approximately

twenty metabolites, of which several have been shown to bind to biomacro-

molecules and to induce mutations (Pelkonen et al. 1982). For the metabo-

lism of PAH (e.g. BP) three main pathways are in general described: the

19

diolepoxide pathway, the one-electron oxidation pathway and the ortho-

quinone pathway.

The diolepoxide pathway was the first metabolic pathway described for

PAH. Three enzyme-catalysed steps are involved in the metabolism of the

parent PAH (e.g. BP) to the ultimate carcinogenic metabolite. First, the for-

mation of an epoxide in the aromatic ring system is catalysed by CYPs. The

epoxide is thereafter hydrolysed by epoxide hydrolase (EH) to a trans-diol.

In the third step another epoxide is formed by the action of CYP enzymes,

which oxidases a carbon-carbon double bond adjacent to the diol group

forming a diolepoxide (Xue et al. 2005). The diolepoxide reacts covalently

by cis or trans addition to nucleophiles, which results in several possible

adducts for every nucleophile (Figure 4).

The expression of CYP1A1 and CYP1A2, involved in the metabolism of

PAH, are regulated by the aryl hydrocarbon (Ah) receptor. The Ah-receptor

is induced by PAH and by other ligands such as dioxins and polychlorinated

biphenyls (IPCS 1998). During the metabolic conversion of BP and other

PAH, chiral carbons are created and many of the intermediates have multiple

stereochemical forms (enantiomeric (+/-) and diastereomeric (syn/anti)),

which make the metabolism very complex. The stereochemical isomers of

the PAH diolepoxides (PAHDE) have different reactivity towards biomacro-

molecules. As an example, in SA the (-)-anti-BPDE mainly form adducts to

His, whereas the major metabolite (+)-anti-BPDE primarily forms

carboxylic acid ester adducts (Day et al. 1994). The stereoselectivity towards

specific metabolites is usually very high when PAH are metabolised by

CYPs and EH.

A diolepoxide can be formed at different positions of BP, the most studied is

the BP-7,8-diol-9,10-epoxide (BPDE) in different stereo isomeric forms

(Figure 4) (Peltonen et al. 1995; Xue et al. 2005). After incubation of rat

liver microsomes with BP the first epoxide, essentially formed, was the

7R,8S-epoxide. Mediated by EH, is the allylic 8S-carbon in this epoxide

oxidised and the spatial configuration inversed with the consequence of the

highly stereospecific formation of the 7R,8R-diol. The diols can be further

metabolised, by CYPs or other enzymes, to (+)-anti- or (-)-syn-BP-7,8-diol-

9,10-epoxides. The major and most carcinogenic metabolite of BP is consid-

ered to be the (+)-anti-BPDE (Stowers et al. 1985).

According to Xue et al. (-)-syn is the major metabolite of BP-syn-7,8-diol-

9,10-epoxide (Xue et al. 2005). Although, Peltonen et al. claim that (+)-syn

is the major syn-form (Peltonen et al. 1995). In both reviews are (+)-anti

considered as the major BP-anti-7,8-diol-9,10-epoxide formed. The different

20

conclusions depend on the test system, which starting compound that is used

and the endpoint analysed.

Figure 4. Metabolic scheme via the diolepoxide pathway. The transformation of BP to adducts formed

after metabolism of the BP-7,8-epoxide. The major metabolic pathway is shown with bold arrows.

In the one-electron oxidation pathway relatively stable radical cations are

formed. PAH have relatively low ionization energy, which makes one of the

π-electrons in the aromatic system easy to remove. This kind of reaction can

be mediated by CYP peroxidases. The radical cations formed are electro-

philic and can interact with nucleophilic sites in biomacromolecules (Xue et

al. 2005).

The ortho-quinone pathway involves oxidation of PAH-diols by the enzyme

family aldo-keto reductases 1 (AKR1), e.g. dihydrodiol dehydrogenases.

These can in the presence of nicotinamide adenine dinucleotide phosphate

(NADP+) convert a diol to the corresponding catechol, which may oxidize to

O

BP-7,8-epoxide

OH

HO

(+)-BP-7,8-dihydrodiol

S S

12

3

4567

8

910

1112

trans(-)anti- cis(-)anti- cis(+)syn-trans(+)syn- trans(+)anti- cis(+)anti- trans(-)syn- cis(-)syn-

HO

OH

(-)-BP-7,8-dihydrodiol

R

R

OH

HO

O

(-)anti-BPDE

RR S

S

OH

HO

O

(+)syn-BPDE

R S

S R

OH

HO

O

(-)syn-BPDE

OH

HO

O

(+)anti-BPDE

RS

RS

RSR

S

OH

HO

X

HO

OH

HO

X

HO

OH

HO

X

HO

OH

HO

X

HO

OH

HO

HO

X

OH

HO

HOX

OH

HO

HOX

OH

HO

HOX

S

Benzo[a]pyrene (BP)

R

21

an ortho-quinone. Similar to BPDE, quinones can form both stable and un-

stable adducts to biomacromolecules (European Food Safety Authority (EF-

SA) 2008; IARC 2012). The reactions following this path are competing

with the reactions in the diolepoxide pathway since the substrate is the diols

in both.

2.4.1 Metabolic differences

Despite numerous studies, the literature gives no answer if there are large

quantitative differences between species in the metabolism of BP to the

genotoxic metabolites. Knowledge about metabolic differences between

species is important when interspecies extrapolation should be performed.

The stereoselective metabolism of PAH may differ between various organs

and species, partly depending on expression of various CYPs (Xue et al.

2005). From two studies in which cells were exposed to BP it was observed

that human cells form more BPDE-dG adducts than mouse and rat at the

same concentration (summarised in Table 3) (Kulkarni et al. 1986; Moore et

al. 1987). This indicates that humans metabolise BP to a larger extent than

the rodents tested.

Table 3. Formation of anti- and syn-BPDE-dG adducts in different species and cells.

Adduct formation (pmol BPDE-dG/mg DNA)

(-)-anti (+)-anti (±)-syn BP

(µM)

cells from tissue Time

(h)

human1 0.014 0.284 0.038 1 endometrium 18

rat1 0.013 0.004 1 endometrium 18

mouse1 0.0732 1 endometrium 18

human2

13.6

2 mammary glands 24

rat2

0.49

2 mammary glands 24

1 (Kulkarni et al. 1986) 2 (Moore et al. 1987)

2.5 Excretion

PAH do not bioaccumulate or biomagnify since most organisms have the

capacity to metabolise these compounds (IPCS 1998). From several studies

it is evident that parent PAH, and their metabolites, with few rings (3- and 4-

ring PAH like pyrene and anthracene) are excreted in the urine or milk from

lactating animals, while larger PAH are excreted in the faeces (summarised

by (EFSA 2008)).

22

A minor fraction (4–13%) of the radioactivity from BP was detected in the

urine in a study of 14

C-labelled BP intravenously injected into rats (Kotin et

al. 1959). From another study where a mixture of BP, anthracene and pyrene

was given to rats, less than 0.4% of the individual and unmetabolised PAH

was excreted in the faeces. These results indicate a substantial metabolism of

BP.

Xenobiotics may also be excreted by lactation, although for BP it appears to

be a minor excretion pathway. From lactating goats orally administrated with

a single dose of 14

C-labelled BP it has been shown that 0.2% of the dose was

excreted (summed during the five days after exposure) in the milk (Grova et

al. 2002).

2.6 Toxicity

Several of the PAH are classified by IARC as carcinogenic in experimental

animals. BP has been shown to be genotoxic in studies with different end-

points and is classified in Group 1 as a human carcinogen (cf. Table 1)

(IARC 2012).

The PAH can be classified as “bay” or “fjord” compounds, depending on the

organization of the rings. A bay is surrounded by three of the aromatic rings

whereas the fjord is surrounded by four of the rings (Figure 5). A DE with

the epoxide close to the bay region of the PAH (e.g. BP-7,8-diol-9,10-

epoxides) is due to steric hindrance more resistant to hydrolysis by the en-

zyme EH than a DE formed somewhere else on the PAH (Casarett and

Doull’s 1995). A PAHDE with the DE close the fjord (e.g. diben-

zo[a,l]pyrene-11,12-diol-13,14-epoxide (DBPDE)) are even more protected

than a bay region DE and may therefore be in the reactive state for an even

longer period of time. Diolepoxides situated in the bay- or fjord area can be

stable enough to be trapped by surrounding nucleophiles to form adducts,

which have been observed in vivo in DNA (Arif et al. 1999; Stowers et al.

1984) and in vitro in proteins (by fjord and bay DEs (Brunmark et al. 1997).

23

Figure 5. Benzo[a]pyrene-7,8-diol-9,10-epoxide (BPDE) and the dibenzo[a,l]pyrene-11,12-diol-13,14-

epoxide (DBPDE) with the bay region and the fjord region marked.

2.7 Analysis of adducts from BP

Over the years, DNA adducts from PAH have primarily been biomonitored

by 32

P-postlabeling. Protein adducts have relied on ELISA methods or

chemical-specific methods (GC/MS, LC/MS or fluorescence) of tetrols

released from a protein by hydrolysis of carboxylic ester adducts. Although,

the yield of these tetrols varies substantially between different reports

(Kaefferlein et al. 2010). One cause may be the measurement of tetrols that

have not been covalently bound as ester adducts. For an accurate measure-

ment of protein bound BP it is important to extract all non-covalently bound

tetrols. Tests with various extraction procedures of in vitro alkylated Hb

have been performed, with 0-35% of non-bound tetrols left after extraction

(Naylor et al. 1990). Another cause to the large variation in the measure-

ments of tetrols from BP is probably the already mentioned cross-reactivity

with other PAH.

The ELISA method, used in a large number of studies of protein adducts

from PAH, has been unable to distinguish between exposed and nonexposed

persons in the majority of studies of PAH exposure (Kaefferlein et al. 2010).

These methods are sensitive but not sufficiently specific to be a tool for envi-

ronmental monitoring of specific PAH (Roche et al. 1985). Furthermore, the

large variation between studies indicates that it is difficult to

compare/extrapolate results.

24

3 Aim and specific background of the thesis

3.1 Aim of thesis – Paper I–III

The general aim of the first part of this thesis was to broaden the methodolo-

gy of protein adducts to applications for in vivo dosimetry of high-molecular

weight compounds to be used as a tool in cancer risk assessment.

Specific aims have been stated for Paper I–III:

In Paper I

- Develop a LC/MS-MS method for analysis of stable adducts from

bulky compounds to His in SA, aiming at sensitivity of the method

in the same range as similar methods for low molecular weight

adducts. The method was aimed to be general to allow further

development for applications to different bulky adducts. Three

diolepoxides of PAH were used as model compounds.

Prerequisites for the method

- The amount of SA needed per analysis must be <10 mg.

- Practical procedures should be developed using standard laboratory

equipment and instruments.

In Paper II

- Clarify if adducts from (+)-anti-BPDE, assumed to contribute to the

highest genotoxic risk from BPDE-isomers, is formed to His in SA

from mice (tested in vitro).

- Detect and quantify adducts formed to His in SA from individual

BP-exposed mice using the method developed in Paper I (in vivo).

- Quantify the relationship between adduct levels from BPDE to His

in SA and to dG in DNA from the same animals.

- Quantify the adduct levels in SA and DNA isolated from whole

blood incubated with BPDE (in vitro).

In Paper III

- Characterise the isomeric BPDE-His adducts obtained in the analy-

sis of the in vivo mice samples using SA in vitro alkylated with

racemic or enantiospecific BPDEs.

- Compare the SA-, the DNA-adduct levels and the MN frequency

obtained after administration of BP to mice.

25

3.2 Specific background

3.2.1 Adduct formation to proteins

It is known that BPDE form adducts to both carboxylic groups (forming

esters) and to His146

in SA. The covalent bond to His was shown to be

formed between the epoxide and the Nτ in the imidazole structure of His

(Day et al. 1991; Zaia et al. 1994) (Figure 6a). Carboxylic ester adducts,

particularly in mice, are not well suited for in vivo dosimetry, since they are

susceptible to hydrolysis both in vitro and in vivo (Naylor et al. 1990). The

sites in SA where PAHDE binds are located in subdomain IB, whereas low-

molecular weight electrophiles preferably bind to IIA and IIIA (Brunmark et

al. 1997).

So far, measurements of tetrols (Figure 6b) from BP are the most common

analytes for BPDE adducts in blood proteins. The tetrols are formed after

mild acid hydrolysis of the protein and have been analysed by sensitive

chemical-specific methods (based on detection with MS or fluorescence).

A disadvantage of carboxylic acid ester adducts, released and analysed as

tetrols, is that the “tag” showing their origin as a covalently bound adduct is

lost. This means that non-covalently bound tetrols from hydrolysed BPDE

must be removed completely before the hydrolysis of the protein for an

accurate quantification (cf. (Naylor et al. 1990)).

The formation of stable BPDE adducts to His has been shown in vitro in

both hHb (Helleberg et al. 2000) and human SA (hSA) (Day et al. 1991c).

The stable adducts facilitate their use as biomarkers. A rather large fraction

(up to 75%) of the total covalent binding of (±)-BPDE to SA has been esti-

mated to be adducts to His (Brunmark et al. 1997; Helleberg et al. 2000).

In human Hb the fraction of adducts to His after in vitro incubation with

BPDE were reported to be lower (10%) than to hSA (Helleberg et al. 2000).

The BPDE adduct to His in SA has been detected and quantified at low

levels in vivo (Ozbal et al. 2000). The method was based on immuno affinity

for enrichment of the adduct and LC with laser-induced fluorescence (LIF)

detection. The method had a very low LOD enabling adducts to be detected

in human samples but the study has never been reproduced.

26

Figure 6. (A) benzo[a]pyrenediolepoxide-histidine (BPDE-His) adduct with the covalent bond at the Nτ

of the histidine (B) tetrol formed after hydrolysis of BPDE.

Adducts formed at the reactive sulphur in the side chain of cysteine are

commonly observed when SA is alkylated with low molecular weight

agents. However, adducts to cysteine after alkylation with BPDE have

neither been observed in SA (Day et al. 1991; Westberg et al. 2014) nor in

Hb (Naylor et al. 1990).

3.2.2 Adduct formation to DNA

BPDE form adducts to dN in the DNA. The covalent bond is mainly formed

between the C-10 in BPDE and (cf. section 1.4) the exocyclic N2 group of

dG similar to other bulky compounds. To a minor degree, adducts between

the BPDE and the exocyclic amino groups N6 of dA and N

4 of deoxycytidine

are formed (Figure 2) (Cheng et al. 1989; Singh et al. 2006; Straub et al.

1977). Adducts from BPDE to the phosphate group in nucleotides have also

been identified (Koreeda et al. 1978).

3.2.3 The need for a method to quantify protein adducts

A review comprising a large number of human studies of exposure of PAH

(predominantly with analysis of PAH-tetrols) concluded that a new sensitive

and specific method for the quantification of protein adducts is required for

reliable biomonitoring (Kaefferlein et al. 2010). There are certain

characteristics of a protein adduct method desired/needed to enable quantifi-

cation of in vivo doses of an electrophile: the adducts analysed should be

stable in vivo and contain a residue of the protein proving that it has been

bound to the protein and that it is not only e.g. a hydrolysis product of the

electrophile, the method for detection should also have high sensitivity and

give structure-specific information about the analytes (i.e. use of MS for

analysis). Furthermore, to be able to perform dose-risk assessments and

interspecies extrapolations, it is desirable that the adducts could be quanti-

OH

OH

NH2O

N

N

HO

t

p

OH

OH

OH

OH

OH

A B

27

fied for in vivo dose assessment in both experimental animals and exposed

humans. No available method for protein adducts from PAH fulfills these

criteria. The method described in Paper I was developed with the aim to

meet these requirements.

3.2.4 Model substances used in this thesis

In the first part of the thesis a method to measure adducts to His in SA as

biomarkers is developed. When analysing DNA, adducts to dG are the

biomarkers primary used. The PAHDE primarily studied with regard to ad-

ducts in this thesis is (±)-anti-benzo[a]pyrene-7,8-diol-9,10-epoxide. Two

other PAHDE were also used, (±)-anti-dibenzo[a,h]anthracene-3,4-diol-1,2-

epoxide and dibenzo[a,l]pyrene-11,12-diol-13,14-epoxide1. In addition, (+)-

anti-BPDE and (±)-syn-BPDE were used in some experiments.

1also designated as dibenzo[def,p]chrysene – the most systematic IUPAC name is naphtho[1,2,3,4-pqr]tetraphen

28

4 Methods for quantitative analysis of bulky adducts to SA and DNA

4.1 Development of a method for measurement of adducts to SA

The development of a method for measurement of bulky adducts to SA was

performed on in vitro alkylated hSA, using BPDE as a model compound.

The parameters for precipitation and hydrolysis of the protein, and

enrichment and analysis of the adducts were evaluated. The procedures are

summarised and shortly commented in Table 4, and some procedures are

discussed in more detail in the following text.

4.1.1 Precipitation of SA for analysis of BPDE-His adducts

In the method developed in Paper I commercial hSA (99% purity) was used

and in Paper II and III the SA was precipitated from plasma from mice.

When using samples prepared from commercial SA, high concentrations of

ammonium sulphate ((NH4)2SO4) and acid to decrease the pH can be added

simultaneously, since the aim is only to precipitate the SA. To obtain pure

SA when starting from plasma the globulins need to be precipitated and

removed first, which occur at a lower salt concentration at neutral pH. If the

concentration of (NH4)2SO4 is too high or the pH too low the SA may

co-precipitate with the globulins (Augustinsson et al. 1966; Peters 1995).

4.1.2 Hydrolysis of the protein

The reactive His in SA is situated in a hydrophobic cavity in the interior of

the protein, which thus has to be degraded before analysis of adducts to His.

The degradation of a protein can be performed by acidic or alkaline

hydrolysis (Landry et al. 1996), with hydrazine (Akabori et al. 1952) or by

enzymatic hydrolysis (Walker, 2002).

A method using hydrazine for the degradation of SA was earlier initiated in

our laboratory for the studies of adducts to His in SA from BPDE (Helleberg

et al. 2000). The results were promising, BPDE-His was obtained as one

analyte and could be analysed by LC/MS with good response. Despite this,

29

further development of the method was discontinued due to the risk of

handling and restricted availability of hydrazine.

In Paper I, degradation with enzymatic and alkaline hydrolysis of SA was

evaluated. With the alkaline degradation BPDE-His and BPDE-Lys were

obtained. With the enzymatic hydrolysis the analytes obtained were BPDE-

His-Proline (-Pro), -His-Pro-Tyrosine (-Tyr) and -Lys. BPDE bound to His

was found as the major stable adduct with both methods (Figure 2 in Paper

I). From both procedures, hydrolysis products were formed, which were

probably generated both from hydrolysis of BPDE and of ester adducts. The

procedures for hydrolysis were optimised to reach a low LOD, for instance

by trying to obtain BPDE-His as one analyte from enzymatic degradation.

Conditions for alkaline hydrolysis were altered by using NaOH at different

concentrations, times and temperatures (Table 4). However, BPDE-His,

which appeared to be relatively stable even in strong alkaline conditions,

was to some extent degraded. It was shown that the enzymatic method re-

sulted in higher measured adduct levels even though two analytes (the di-

and tripeptide) of the BPDE-His were formed. To obtain one analyte from

enzymatic hydrolysis, pronase degradation was combined with the addition

of the enzyme papain and the detergent RapiGest™ (Table 4). However,

neither the use of RapiGest™ nor papain resulted in a single analyte after

enzymatic hydrolysis.

Finally, the procedure using pronase for protein degradation was chosen

since the LOD from the enzymatic degradation was about 12 times lower

than the LOD from the alkaline hydrolysis, using LC/MS-MS for analysis.

After the publication of Paper I, a suggestion of using prolidase to degrade

the BPDE-His-Pro and BPDE-His-Pro-Tyr was conveyed. The problem to

hydrolyse the peptide bond around His neighbouring a bulky hydrophobic

amino acid (e.g. Tyr) has been discussed (Walker, 2002). A bulky PAH-

moiety attached to the His will probably decrease the efficiency of the en-

zymatic cleavage even further. An additional experiment was performed

with prolidase. However, the BPDE-His was still not obtained with the en-

zymatic method containing pronase and prolidase (not published).

30

Table 4. Procedures and parameters evaluated in the development of a method for analysis of bulky adducts to serum albumin, using BPDE as the model compound. In addition,

comments on the experimental results are summarised in this table.

Procedure studied Evaluated parameters Conditions tested Optimised results Comments on the evaluated parameters

Adhesion to material Tube material Eppendorf LoBind®; polypropylene; unsilanised glass; silanised glass;

Eppendorf LoBind® Increased yield with more than 40% with Eppendorf LoBind® tubes, no large difference between the other materials.

Precipitation of SA pH pH 4 and 7 pH 4 The adducts were stable and the combined yield of BPDE-His-Pro and His-Pro-Tyr was the same at both pH studied, although a shift of the ratio was seen.

Time in acidic solution 0.3, 20 and 40 h 0.3 h Decreasing amount of SA/mg solid was observed with increasing time but the same amount adduct/mg SA. Probably more salt is formed during time.

Solvent for hydrolysis Conc. of NH4HCO3 to dissolve SA 50 and 200 mM 50 mM An increase in yield of ca 25% the yield with 50 mM.

Enzymatic hydrolysis Type of enzyme addition pronase; pronase + papain; pronase + RapiGest

Enzymatic with pronase Pronase alone gave lowest LOD (12 times lower compared to alkaline hydrolysis) even though two analytes had to be considered.

Enzyme/substrate 0.1 and 0.4 mg pronase/mg SA 0.1 mg pronase/mg SA No obvious effect was observed on the yield by the increased enzyme-substrate ratio tested.

Concentration of SA 0.2 and 20 mg SA/mL 20 mg SA/mL No obvious effect was observed on the yield by the decreased concentration of SA.

Alkaline hydrolysis Final concentration of NaOH 1, 2 and 4 M 2 M Stronger concentration of NaOH resulted in degradation of the adduct.

Temperature; time 23, 100 and 130°C; 0.5 - 46 h 100°C, 5-24 h Lower temperature gave low yields irrespective of time. The time for hydrolysis was not crucial.

Enrichment of adducts Method Solid phase extraction (SPE); LC/UV; Butanol extraction at pH 3 and 7

SPE or HPLC SPE and HPLC were comparable methods regarding the yield but SPE preferable due to less time and solvent consumption. Butanol extraction gave <50% of the yield compared to SPE or HPLC.

LC/MS-MS parameters Mobile phases Methanol (MeOH) and acetonitrile(AcN); formic acid and trifluoroacetic acid

MeOH with formic acid With MeOH and formic acid in the eluents the response was about the double compared to AcN and trifluoroacetic acid.

Electrospray polarity Positive and negative Positive mode

Matrix effect in the LC/MS-MS

Amount SA/sample 10, 25 and 40 mg SA 10 mg SA The sample amount was critical for yield/response in the analysis. Method optimised for 10 mg SA.

31

4.1.3 Enrichment of the adducts

Enrichment of the analytes was aimed to be performed by the use of standard

laboratory procedures and equipment. Butanol extraction and isolation of the

adducts using LC and solid phase extraction (SPE) were compared. LC and

SPE gave similar yields, in contrast to butanol extraction for which the yield

was significantly lower (cf. Table 4). Different test tube materials were also

evaluated. The use of Eppendorf Protein LoBind® tubes, compared to other

tubes of plastic or glass, was shown to increase the yield with approximately

40%. This shows that these analytes are apt to stick to the surface of

different materials, with a consequence for the yield of measured adducts.

4.1.4 Analysis of DBPDE and DBADE

In addition to the analysis of hSA alkylated with BPDE the method was

tested on hSA alkylated with the diolepoxides of DBP and DBA (Figure 3).

The analytes and the fragments obtained in the MS analysis after the

enzymatic hydrolysis were the corresponding analytes/fragments as achieved

from the BPDE alkylation. This was determined by analyses using LC/UV

and LC/MS-MS. Although, after the alkylation with DBADE the major

adduct was formed to Lys (Figure 7).

Figure 7. LC/UV chromatogram (λ 291 nm) from analysis of SA alkylated with diben-

zo[a,h]anthracenediolepoxide (DBADE). From LC/MS-MS analysis, the peaks have been identified as

DBADE-lysine (-Lys), DBADE-histidine-proline (-His-Pro) and DBADE-His-Pro-Tyrosine (-Tyr).

4.1.5 Internal standard and quantification of the adducts

During the development of a method and the process to decide suitable

biomarkers/analytes a proper internal standard (IS) is in general not available

or too expensive to synthesise. In Paper I, a simple internal standard

approach was applied using SA alkylated in vitro with another PAHDE,

which compensates for variations in the enzymatic hydrolysis and to some

extent for the condition of the LC/MS instrument, although not for possible

ion suppression in the analysis.

Inte

nsi

ty

DBADE-Lys

DBADE-His-Pro

DBADE-His-Pro-Tyr

Time

32

With an equation established in Paper I (Eq. 1) the concentration of the

injected sample was calculated from the extinction coefficient (ε) for the

adduct and the peak area obtained in LC/UV analysis.

The equation enabled the quantification of low amounts of adducts in a

crude mixture of the enzymatically digested SA. In the investigations for

Paper I and II, the ε of 29,000 M-1

cm-1

(λ345 nm) for BPDE-dG was used for

the calculations of the adduct levels of BPDE-His-Pro and BPDE-His-Pro-

Tyr in SA. The aromatic ring system in the adduct from PAH is assumed to

be the major contributor to the absorbance of the analyte. Neither these

amino acids nor the dG absorb light at 345 nm and were therefore considered

negligible to influence the ε. This is supported by earlier studies where the ε

of BPDE-dN were not affected by the addition of an oligonucleotide

(Weinstein et al. 1976).

The matrix-based calibration curves established by LC/MS-MS using this

internal standard approach were linear and indicated good reproducibility in

the results (coefficient of variation (CV) 11%) (shown for DBADE adducts

in Figure 8). The areas from BPDE-His-Pro and BPDE-His-Pro-Tyr were

summed since that resulted in the most linear calibration curves. The LOD of

the method was ca. 1 fmol BPDE-His adducts/mg SA, after injection of 5 mg

SA (a calibration curve for BPDE-adducts is shown in Figure 4, Paper I) and

a limit of quantification (LOQ) of around 50 fmol on-column.

Approximately the same LOD and LOQ was achieved for the adducts from

DBPDE and DBADE as well.

inj

injVlεQ

fAC

=

inj

injVlεQ

fAC

=

Cinj=concentration of the injected sample (mol/L); A=peak area (µV·second);

f=flow rate (Litre/second); Q=output voltage/absorption unit (µV/AU);

ε=extinction coefficient (M-1·cm-1); l=cuvette length (cm); Vinj=injected volume (Litre)

(Eq. 1)

33

Figure 8. A calibration curve (0.86-208 fmol adduct/mg SA, n=6) from LC/MS-MS (ESI+) analysis for

the ratio of the summed analyte areas (DBADE-His-Pro and DBADE-His-Pro-Tyr), and the summed

areas of the internal standard (BPDE-His-Pro and BPDE-His-Pro-Tyr).

4.1.6 LC/MS-MS analysis

In the LC/MS-MS analysis of the adducts to SA the best mobile phase tested

was shown to constitute of methanol:water with formic acid as an additive.

The analyses were performed with electrospray ionisation in the positive ion

mode (ESI+). When the product ion scan mode2 (PIS) was used, several

fragments were obtained that were common for the different PAHDE

(Figure 3 and Table 2 in Paper I). For the analyses using multiple reaction

monitoring (MRM) two or three of the fragments obtained in the PIS were

chosen.

In order to reach a low LOD, different LC/MS instruments were tested. For

the method development in Paper I, a column-switch set-up in the LC/MS-

MS-system was introduced. This set-up enabled larger volumes to be

injected and pre-concentration of the analytes resulting in better sensitivity.

However, with a new MS-instrument and without the knowledge about the

low adduct levels in SA from mice the column-switch was considered un-

necessary. For the analysis of in vivo samples two different models of

Qtrap/Triple quadropoles from AB Sciex was used (Paper II and III). In

future work, the use of a column-switch, enabling injection of the whole

sample (10 mg) should be taken into consideration.

4.1.7 The evaluated procedures combined to a method

An important purpose with the method developed in Paper I was to be able

to measure bulky adducts to SA from individual mice exposed to BP. In the

22 In PIS the first mass analyser has a fixed mass while the second perform a scan.

DB

AD

E-H

is-P

ro a

nd

DB

PD

E-H

is-P

ro-T

yr/

IS (B

PD

E-H

is-P

ro a

nd

BP

DE

-His

-Pro

-Tyr

)

0

0.5

1.0

1.5

2.0

50 100 150 200 250

y= 0.0095x-0.032R2= 0.9978

fmol DBADE-His-Pro and DBPDE-His-Pro-Tyr/ mg SA

34

final method samples of 10 mg SA were processed, which was sufficient for

two analyses from one mouse.

When the optimised procedures were combined, a robust and easy-to-use

method was achieved for analysis of bulky adducts from SA. The SA was

enzymatically hydrolysed using pronase and the PAHDE-alkylated peptides

were enriched by SPE or LC/UV, followed by analysis and quantification

using LC/MS-MS (ESI+) (Figure 9). During the development of the method

and when analysing non-characterised adducts LC/UV was considered

preferable. When the properties and the identity of the adducts are known the

use of SPE is preferable since it is less time and solvent consuming.

Figure 9. The analytical method established for analysis of bulky adducts to serum albumin. The steps in

the procedure were evaluated and the best options combined into this method.

Enzymatic hydrolysis(20 mg SA/ml, 0.09 mg pronase/mg SA,

37 C, over night)

PAHDE-SA 10 mg Addition of IS of another PAHDE-SA

(50 mM NH4HCO3)

Enrichment with SPE or HPLC (H2O:methanol)

LC/MS-MS-analysis (ESI+, H2O:methanol, 0.1% formic acid)

Precipitation of SA from plasmausing Eppendorf protein LoBind® tubes1. addition of saturated (NH4)2SO4 (1 vol.)

globulins precipitate 2. centrifugation and decrease of pH (pH 4)

SA precipitate

Wash of PAHDE-SA(methanol, ethyl acetate and pentane)

35

4.2 Methods for measurement of adducts to DNA

4.2.1 Qualitative and quantitative standards for analysis of

BPDE-dG adducts in DNA

A method was established for LC/MS-MS analysis of DNA adducts from

BPDE. A standard of BPDE-dG was needed for qualitative and quantitative

analyses for the investigation of BPDE adducts to DNA. Qualitative

reference adducts of BPDE-dG and BPDE-dA were generated in vitro by

incubation of DNA with (±)-anti-BPDE followed by enzymatic hydrolysis of

the DNA.

A quantitative BPDE-dG standard was not commercially available. To

determine the amount of BPDE-dG in the qualitative standard, the adduct

level was intended to be measured from the peak area of the adduct with

LC/UV analysis of the BPDE-alkylated DNA after enzymatic degradation

(cf. the use of Eq. 1 for the determination of the adduct levels). However, the

BPDE-dG adduct co-eluted with hydrolysis products (triols or tetrols) of the

BPDE, as also observed by others (cf. (Singh et al. 2006)). The alkylated

BPDE-DNA was extensively extracted and the organic solvent was analysed

with a spectrophotometer (scan λ250-400 nm). After 10 extractions, with ethyl

acetate and diethyl ether, hydrolysis products from the BPDE were still

present in the extract from the reaction solution. It was concluded that a

complete extraction of the non-covalent and hydrolysed products was not

possible (cf. extraction of BP-alkylated Hb, in section 2.7, Naylor et al.

1990).

The generation of a quantitative standard for BPDE-dG was also

investigated by the alkylation of an oligonucleotide expected to have a

different retention time in LC/UV than the hydrolysis products of BPDE.

However, several peaks were achieved and it was difficult to conclude which

fraction(s) to collect and digest and problematic to get reliable/repeatable

results. Another attempt was to use a BPDE alkylated DNA standard earlier

used for 32

P-labeling (provided by Dan Segerbäck, KI, Huddinge). However,

the amount of the standard and degree of alkylation (1.1 BPDE-dG/106 dN)

were found to be too low to be useful in the LC/MS analysis. Finally,

through cooperation with R. Singh in Leicester, UK, a stable

isotope-substituted internal standard (15

N5)BPDE-dG previously synthesised

(Singh et al. 2006) could be obtained. This standard was used to quantify the

BPDE-dG adducts formed in the in vitro experiment with whole blood.

36

4.2.2 The impact of modifications of the method for enzymatic

hydrolysis of DNA

DNA can be degraded with nucleases (e.g. nuclease P1 (NP1) and deoxy-

ribonuclease I (DNase I)) to oligo- or mononucleotides (containing the base,

sugar and phosphate group). The nucleotides can thereafter be degraded to

nucleosides (base and sugar). The enzymes used in this study were different

kinds of nucleases and alkaline phosphatase (Table 5).

Table 5. A short description of the enzymes used in the enzymatic degradation of the DNA in Paper II.

The non-repeatable and inconclusive results from the experiment with the

alkylated oligonucleotide (cf. section 4.2.1) raised the question; was some-

thing wrong with the enzymatic hydrolysis of DNA? A test was performed

(unpublished) to study if modifications of the enzymatic degradation method

of DNA (briefly described in Paper II) could have impact on the yield of the

hydrolysed DNA. Five methods for enzymatic hydrolysis (with minor

modifications) were compared (Table 6). The hydrolysed DNA samples

were analysed with regard to the yield of dG using LC/MS-MS.

1. Endonuclease Hydrolyses the phosphodiester bonds from either the 3’ or the 5’ side

within (endo) a polynucleotide chain.

Nuclease P1 (NP1) A nonspecific endonuclease resulting in nucleoside 5′-phosphates and 5′-phosphooligonucleotides. The suppliers states that NP1 does not “appre-

ciably” attack double-stranded nucleic acids.

Deoxyribonuclease I (DNase I)

A nonspecific endonuclease resulting in nucleoside 5′-phosphates and 5′-phosphooligonucleotides.

Acts on both single- and double-stranded DNA.

2. Exonuclease Hydrolyses the phosphodiester bonds from either the 3’ or the 5’ end

(exo) of a polynucleotide chain.

Snake venom

phosphodiesterase I

(SVPD I)

A nonspecific exonuclease that attacks free 3'-hydroxyl terminus, which

results in 5' mononucleotides.

3. Phosphatase Dephosphorylates nucleotides by hydrolysis of the phosphate monoester.

Alkaline phosphatase

(ALKP)

Hydrolyses the phosphate group in nucleotides yielding nucleosides.

Most effective in alkaline conditions.

37

Table 6. Modifications of the procedures tested for enzymatic degradation of DNA to nucleosides.

Sample Procedures tested for hydrolysis of DNA

1a, 1b The original used method without denaturation (Westberg et al. 2015).

2 Denaturation of the DNA at 100°C for 3 min, followed by the original method used for

samples 1a and 1b.

3a, 3b Denaturation of the DNA at 100°C for 3 min, followed by simultaneous addition of SVPD

and alkaline phosphatase.

4 Dilution of the DNA solution (10 times) followed by the procedure used for sample 2,

including denaturation.

5 The use of DNase I (incubation 6 h, 37°C) instead of NP1, without prior denaturation. Less

amount of ALKP used in this method (method modified from (Singh et al. 2010)).

Information sheets from the suppliers of the endonuclease nuclease P1 (NP1)

imply that the DNA must be single-stranded to exert optimum hydrolysis.

Denaturation at 100°C to obtain single-stranded DNA was therefore

investigated (sample 2, 3a, 3b, and 4 in Figure 10). Serial and simultaneous

addition of the exonuclease snake venom phosphodiesterase (SVPD) and

alkaline phosphatase (ALKP) were also tested. Inconsiderable variations of

the yield of dG were observed after these modifications, which therefore

were judged to be non-critical for the DNA hydrolysis (Figure 10). Thus, for

the final experiments, in Paper II, the original method without denaturation

of the DNA and with serial enzymatic incubations was used.

Figure 10. The relative amount of dG analysed using LC/MS-MS to compare modifications in the enzy-

matic hydrolysis method of DNA. The amount dG in the samples are normalised to sample 1a.

It is important to notice that one sample appeared to be hydrolysed to a very

low extent (3b, Figure 10). This sample was a duplicate of sample 3a and no

deviation during the sample preparation was noticed. With regard to the

previous inconclusive results, which led to this enzymatic test and the result

of sample 3b it was found important to measure the efficiency of hydrolysis

as the amount of dN injected in the analysis of each sample. Therefore, for

0%

20%

40%

60%

80%

100%

120%

1a 1b 2 3a 3b 4 5

% d

G

Sample number

38

the work in Paper III a calibration curve for dG was established and used for

measurement of the dG injected in the analysis of adducts to dG

(a calibration curve for dA was also used as a complement). The amount of

DNA was quantified with two different approaches a) with a spectrophoto-

meter before the degradation of the DNA and b) by LC/MS using a

calibration curve for dG. About 30% (with a range between 60 and 100%)

lower amounts of dG were measured from MS analysis after the hydrolysis

of DNA than calculated from DNA measurements before hydrolysis.

The enzymes used for the degradation of the DNA are expensive and the

storage time for dissolved enzymes is often relatively short. For the work in

in Paper III an investigation was performed where only half the amount of

the enzymes was used to degrade the DNA (Figure 11), which appeared to

be as effective as the method used in Paper II.

Figure 11. The final procedure used in Paper III for the enzymatic hydrolysis of DNA.

100 µg DNA in 1 mM Tris-HCl(1 mg/mL)

using Eppendorf DNA LoBind® tubes

NH4OAc (0.1 vol, 0.1 M, pH 5.3)Addition of ZnCl to a final conc. of 0.35 mM

Nuclease P1 40 u/mg DNA45°C, 2.5 h

→ oligonucleotides

NH4HCO3 (0.1 v., 1 M)Exonuclease SVPD 0.04 u/mg DNA

37°C, 2.5 h→ oligonucleotides

Alkaline phosphatase 10 u/mg DNA37°C, 1.5 h

→ nucleosides

Isolation of DNA using Qiagen kit

LC/MS-MS analysis(ESI+, H2O:acetonitrile, 0.1% formic acid

39

A possibility for future development of the enzymatic method for DNA

hydrolysis could instead of NP1 employ Benzonase® (Quinlivan et al.

2008), which is said to act on both single- and double-stranded DNA and has

a pH and temperature preferences similar to the other enzymes. Although,

not evaluated in this work since NP1 was shown to degrade double-stranded

DNA as well.

4.2.3 LC/MS-MS analysis

In the LC/MS-MS analysis of the adducts to DNA the best mobile phase

tested was a mixture of acetonitrile and water with formic acid as an

additive. Two different instruments were used, both of them Qtrap/Triple

quadropoles from AB Sciex. The analyses were always performed with

electrospray ionisation in the positive ion mode (ESI+). Full scan and PIS

were applied for the characterisation of the adducts (Paper II). MRM was

used in all quantitative analyses or for detection of known adducts in

samples (Paper II and III).

4.3 Discussion about the methods developed

When studying adducts formed from PAHDEs in SA the most abundant

adducts are to His (Brunmark et al. 1997). In analysis of protein adducts

from low-molecular weight compounds are adducts to cysteine common.

Although, in incubations of SA with BP cysteine adducts are rare, which

may depend on the relatively large size of BP, which may cause steric

hindrance and low reactivity to the cysteine sites. In Paper I, no adduct to

cysteine was detected when the adducts from three PAHDEs were character-

ised.

The method for quantification of adducts from PAHDEs in SA fulfils the

major part of the characteristics stated for a protein adduct method (cf.

section 1.2). Stable adducts from the studied PAHDEs with a residue from

SA are specifically analysed using MS detection. Good reproducibility and

high sensitivity of the method were achieved.

A method for the LC/MS-MS analysis of BPDE adducts to DNA was

established and from a collaboration a quantitative (15

N5)BPDE-dG standard

was obtained. It was shown that neither minor changes of the method for

enzymatic hydrolysis of DNA nor the use of half the amount of enzymes

affected the yield of dG to any appreciable extent.

40

5 Investigations of adduct formation to SA and DNA from BPDE – in vitro and in vivo

5.1 In vitro experiments

5.1.1 Alkylation of SA from mouse and human with (+)-anti-

BPDE, (±)-anti-BPDE and (±)-syn-BPDE (Paper I–III)

The most genotoxic BP-metabolite, (+)-anti-BPDE, has been shown to be

the major enantiomer metabolically formed from BPDE in mice exposed to

BP (Stowers et al. 1985). Though, according to literature the (+)-enantiomer

does not form His-adducts in human SA (Day et al. 1994). Due to the simi-

larity of the the amino acid sequence surrounding His146

the same would be

expected for mSA. It had to be clarified whether the (+)-anti-BPDE form

adducts to His in SA from mice or not, in order to elucidate if the adduct

could be used as a measure of internal dose of the most genotoxic BP-

metabolite in mice.

In Paper I, hSA was in vitro alkylated with (±)-anti-BPDE. In Paper II and

III, the (+)-anti-BPDE was shown to form adducts to His with mSA in vitro

and in vivo, respectively. The LOD of our method was probably lower than

the detection level of the fluorescence line narrowing (FLN) method used by

Day et al. 1994, which explains that we detected the adducts to His in

difference to the earlier observation.

SA from mouse was also alkylated in vitro with (±)-anti-BPDE to compare

the reactivity towards mSA and hSA. The results indicate that there is no

major difference in the reactivity towards the reactive His146

in SA from

mice or humans (Paper II). However, alkylation with (±)-anti-BPDE showed

higher adduct level than a corresponding alkylation with the (+)-enantiomer,

which indicate that the (-)-enantiomer formed ca. 10–15 times higher level

of adducts after alkylation with the same initial concentration (Paper II and

III).

For the investigations performed in Paper III it was possible to incubate hSA

with (±)-syn-BPDE, which had not been accessible earlier. The LC-elution

order of the racemic syn-BPDE in relation to (+)- and (-)-anti-BPDE was

41

studied. The (±)-syn-BPDE eluted as one single peak in between the

separated (+)-anti-BPDE and the (-)-anti-BPDE under the LC conditions

used (cf. Figure 2, in Paper III).

5.1.2 Adduct levels in SA and in DNA after incubation of

whole blood with (±)-anti-BPDE (Paper II)

An in vitro study was performed to determine the relationship of adduct

formation from BPDE in SA and in DNA. Incubations of human whole

blood with (±)-anti-BPDE were performed, which means that the SA and

DNA were incubated at equivalent and physiological conditions, with

minimal influence of DNA repair and without need for metabolic activation

to the electrophile. One advantage of using whole blood from human instead

of mice was that sufficient amount of DNA could be obtained. If mouse

blood should be used for this experiment a large number of animals would

be required. DNA from the buffy-coat and SA from the plasma were isolated

and the BPDE adducts formed to His in SA and dG in DNA were analysed

(Figure 12).

Figure 12. Outline of the experiment where samples of human whole blood were incubated with (±)-anti-

BPDE. The DNA from the buffy-coat and the SA from the plasma were isolated and the respective ad-

ducts analysed with LC/MS-MS.

It was a perfect linear association between the adduct level in SA and the

increased amount of BPDE added. The increase of the adduct level was 1.1

pmol BPDE-His adducts/mg of SA for each added µg of BPDE/mL blood.

The adduct level measured in DNA increased with 2400 pmol BPDE-dG/mg

of DNA for each added µg of BPDE/mL blood (Figure 13).

OH

OH

OHNu

Isolation of the SA from the plasma.

Incubation with BPDE of human whole blood(37°C, 3 h).

Enzymatic degradation of the DNA and LC/MS analysis of the BPDE-dG.

Isolation of the DNA from the buffy-coat.

Enzymatic degradation of the SA, followed by enrichment and LC/MS analysis of the BPDE-His adducts.

42

Figure 13. Adduct levels of BPDE-His from SA (n=6) and BPDE-dG from DNA (n=3) after in vitro

alkylation of human blood and analysis by LC/MS-MS (ESI+). The adduct levels were plotted against the

amount of (±)-anti-BPDE added to the samples of human whole blood.

5.2 In vivo experiments

5.2.1 BP adducts to SA in mice (Paper II)

In the scope of this thesis two experiments were performed with mice ex-

posed to BP. In the first of these experiments (Paper II), mice (wild-type

CD2F1) were treated with a single intraperitoneal injection of BP

(100 mg/kg b.w.) and sacrificed at different days after exposure. Blood

samples were collected at 1, 3, 7, and 28 days after exposure (3 mice per

day) and the BPDE-His adducts in SA were analysed using the method de-

veloped in Paper I.

In Paper II, adducts to His in SA from BP-exposed mice were detected and

quantified using LC/MS-MS for the first time, although at levels close to the

LOD of the method. The BPDE-His adducts were detected as two peaks of

BPDE-His-Pro, most probably stereoisomeric forms. The adduct levels were

highest at day 3, measured as 13 fmol BPDE-His-Pro/mg SA (Table 7) and

could not be detected in the samples from 28 days after exposure

(chromatograms from days 1, 3 and 7 are shown in Figure 3 in Paper II).

The BPDE-His adduct level in SA was compared with the BPDE-dG adduct

level in DNA (previously measured by (Singh et al. 2006)) from the livers in

the same individual mice. The SA adduct levels in mice were shown to be

very low compared to the DNA adduct levels (in pmol/mg; 1:400,

SA:DNA).

0

10

20

30

40

50

60

0 10 20 30 40 50

BPDE (µg) added/ mL whole human blood

pm

ol B

PD

E-H

is-P

ro +

BP

DE

-His

-Pro

-Tyr

adduct/ m

g S

A

0

20000

40000

60000

80000

100000

120000

140000

pm

ol B

PD

E-d

G/ m

g d

NBPDE-His-Pro+BPDE-His-Pro-Tyr

y = 1.0981x + 0.5342

r2= 0.9989

BPDE-dG

y = 2400x

43

To elucidate what influences the adduct levels, the ratios of the adduct levels

to SA and to DNA from the in vitro experiment and the in vivo study were

compared. The ratio after the in vitro alkylation of whole blood (1:1850,

SA:DNA at 10 µg/ml) was higher than the ratio from the in vivo samples).

Table 7. BPDE-His-Pro-adduct levels in SA from mice intraperitoneal treated with BP (100 mg/kg b.w.).

Days after administration of

100 mg BP/kg b.w.

Number of

mice

Mean adduct level (fmol BPDE-

His-Pro/mg) (range min-max)

Signal to noise

(S/N)

1 3 3.4 (3.4-3.5) 2-4

3 3 13.1 (12.2-14.5) 14-22

71 3 4.0 (2.8-4.9) 2-4 1

At day 28 after exposure were no adducts detected.

In the analysis of mSA from in vivo experiments, peaks with the MRM tran-

sitions corresponding to BPDE-His-Pro were detected that had small shifts

in the retention times compared to the peaks in hSA in vitro alkylated with

(±)-anti-BPDE. These shifts were assumed to depend on the formation of

adducts from other BPDE-isomers formed in vivo, e.g. syn-BPDE, these

were not identified in Paper II.

SA samples from mice exposed to DBA and DBP were also analysed.

The adduct levels were <LOQ (data not shown).

5.2.2 Identification and quantification of specific stereo-

isomeric adducts in SA and DNA from mice (Paper III)

In the second in vivo experiment (Paper III), mice (BalbC) were treated with

a single intraperitoneal injection of BP (25, 50 and 100 mg/kg b.w., 2 mice

per dose) and sacrificed 48 hours after injection. Blood samples and the liver

were collected. The adducts of BPDE-His in SA and BPDE-dG in DNA

were prepared and analysed essentially using the same procedures as in Pa-

per II. The frequency of MN was also measured.

The amount of SA obtained was only sufficient for one analysis from each

mouse. The BPDE-adduct levels to His in SA were quantified even though

the intensities of the peaks were close to LOQ. The adduct levels at 100 mg

BP/kg b.w. were about the same as the levels reported in Paper II.

An improvement compared to the in vivo experiments presented in Paper II

was that a separation of the adducts from the different stereoisomeric BPDEs

was achieved. The separation could probably be assigned to the use of an

ultra-performance LC/MS instrumentation.

44

In the LC/MS-MS chromatograms from the exposed mice three peaks are

considered to reflect BPDE-His adducts. The first peak eluting was

identified as (+)-anti-BPDE-His-Pro. The other peaks observed probably

correspond to (+)-, (-)-syn-BPDE-His-Pro and (-)-anti-BPDE-His-Pro (with

a possible co-elution of the stereoisomers) or another BPDE-His-Pro isomer

(see Figure 1, in Paper III).

The BPDE-adducts in the DNA from the exposed mice were detected as

three peaks to dA and as four peaks to dG (Figure 4, in Paper III). The peak

corresponding to (+)-anti-BPDE-dG was identified in the samples of hydro-

lysed DNA from the mice.

The three biomarkers studied in this experiment (adducts to SA, DNA and

the frequency of MN) showed significant increase with increasing dose of

BP. The data on SA adducts were used to make a preliminary estimate of the

in vivo dose of (+)-anti-BPDE and used for an estimation of its relative

genotoxic potence.

5.3 Discussion

Suggestions have been proposed to why the in vivo adduct levels in SA are

relatively low compared to the adduct levels in DNA. The difference may

depend on the degree of formation of the studied BPDE adducts or other

processes in vivo, e.g., distribution or metabolism of BP. From our in vitro

results we conclude that the rate of adduct formation to His and dG is the

major factor to the difference, possibly due to more steric hindrance in the

SA than in the DNA. The lower adduct level in SA may be a result of only

one major reaction site in SA, which is located in a hydrophobic pocket in

the SA structure. In Paper II it was estimated that each DNA strand has

approximately 50 times more dG sites (per gram), although some of the dG

are more reactive than others (e.g. a GG-sequence) (Rill et al. 1990).

Higher levels of adducts are formed when double-stranded DNA react with

BPDE compared to reactions with single-stranded DNA. This is probably

due to intercalation of the BPDE to the double-stranded DNA, which may

prolong the lifetime of the DE. This may be another reason for the lower

level of adducts formed in SA, which tertiary structure resembles more of

single-stranded than double-stranded DNA.

From a published study of exposed mice it could be roughly estimated that

less than 1% of the BP is metabolically transformed to BPDE measured as

DNA adducts (Ginsberg et al. 1990). Despite the relatively lower efficiency

of adduct formation to SA it is indicated that BPDE metabolites are efficient

45

to form the protein adduct; compared to other studied compounds (e.g.

simple epoxides). In Table 3, in section 2.4.1, DNA-adduct levels from

BPDE in cells from human, mouse and rat are listed. From this table it is

indicated that human metabolise BP to a higher degree (4–28 times) than the

rodent species tested. Although, the difference in metabolism between

mouse and human was relatively small. If humans metabolise BP like

laboratory exposed mice do, the adduct level in humans is expected to be 104

times lower than in the exposed mice (based on daily intake of BP and the

adduct levels measured in exposed mice, Paper III).

The BPDE-His adduct in SA detected with LC/MS is a potentially good

biomarker for in vivo dose of BP exposure, because it is specifically

measured as a stable adduct and His in the analyte proves that the BPDE has

been attached to a protein. The BPDE adduct levels (both to DNA and to

SA) in mice administered with 25 mg/kg of bodyweight are measurable

(Paper III) but not several orders of magnitude above LOD. The method will

require considerable improvement with regard to LOD in order to measure

human exposure of BP. The LOD may be improved e.g. by the use of

nano-LC and column-switch techniques.

46

6 Isocyanates

6.1 Aim of Paper IV and V

- Establishment of an LC/MS-MS method for detection and quantifi-

cation of adducts from isocyanates to N-terminal valine in Hb.

- Evaluate if adducts from isocyanic acid measured by this method

could be appropriate as a biomarker for kidney failure.

6.2 Formation and usage of isocyanates

Isocyanates are common industrial chemicals and can be divided in two

groups by their structure and their usage; a) monoisocyanates (Figure 14)

which are used in the production of pesticides, e.g. methyl isocyanate (MIC).

MIC is a starting agent in the production of the pesticide carbaryl. In a

tragical accident in India 1984, MIC was spread over the whole town of

Bhopal. Several thousands of people were chronically affected or dead in the

after-effects of the poisonous gas (Sriramachari 2004).; b) diisocyanates,

used in the polymer industry (e.g. in upholstered furniture). Toluene di-

isocyanate (TDI) and methylene diphenyl diisocyanate (MDI) are the

isocyanates produced in highest amounts globaly. These are used for the

production of polyurethanes (Verschoor et al. 2014). Isocyanates are also

formed from incomplete combustion of nitrogen-containing materials

(Hertzberg et al. 2003) and MIC has been detected in tobacco smoke

(Philippe et al. 1965).

Figure 14. A schematic figure of monoisocyanate where R can be various alkyl groups.

6.3 Exposure and Toxicity

Exposure to isocyanates occurs mainly from industrial processes and

constitutes an occupational health problem. For the non-occupational

exposed population the use of certain consumer products (glues, spray foam

R=H, isocyanic acic R=CH3, methyl isocyanate (MIC)R=C6H5, phenyl isocyanate (PIC)

47

insulation etc.) is the major exposure (Verschoor et al. 2014). Beside these

exogenous exposures there is also an endogenous exposure to the simplest

analouge isocyanic acid, which is non-enzymatically formed when urea

breaks down.

In contrast to PAH, metabolic activation is not necessary for isocyanates for

reaction with biomacromolecules, and adduct formation, to occur. Even at

very low concentrations of isocyanates induce irritation to eyes and throat,

higher concentrations cause severe damage to the eyes and lungs (e.g.

asthmatic problems). Exposure to isocyanates is one major cause of occupa-

tional related asthma (Kumar et al. 2009).

Isocyanates include compounds that are known to cause cancer in animals

and are classified as possible human carcinogens (in group 2B and 3 in

Table 1) but the major adverse effects are severe irritation and sensitation to

these compounds.

A group of compounds closely related to the isocyanates are the isothio-

cyanates (ITCs), in which the oxygen in the isocyanate is replaced with

sulfur. ITCs have been shown to be chemopreventive (suppress or prevent

the risk of developing cancer). Precursors to the ITCs occurs naturally in

cruciferous vegetables e.g. broccoli and cabbage and are transformed to ITC

from enzymes released in the chewing process (Herr et al. 2010).

6.4 Biomonitoring of isocyanates

Isocyanates are low-molecular weight compounds with a highly electrophilic

carbon, which make them reactive and useful in the industry. Due to the

electrophilicity they are also reactive to biological molecules and form ad-

ducts.

Immunological methods have been performed in order to monitor the expo-

sure of isocyanates. However, considerable variations in interlaboratory

comparisons are observed and it is difficult to produce proper antigens

(Kumar et al. 2009) (Ott et al. 2007). Biomonitoring of isocyanates in hu-

mans has been conducted by hydrolysis and derivatisation of amines, which

are released and isolated from acidified plasma or urine followed by detec-

tion with GC/MS (Skarping et al. 1995). However, this methodology is gen-

eral for analysis of amines and not specific for detection of isocyanates.

Isocyanates have also been analysed as adducts to DNA by LC/MS-MS from

samples prepared in vitro (Beyerbach et al. 2006). The benefits with LC/MS-

MS for determination of isocyanate adducts are the same as for other ad-

48

ducts, it is a specific determination of the analytes and it is often possible to

detect the adducts at low levels.

In parallel with the publication of Paper IV, a study of SA adducts to Lys in

rats exposed to MDI, and analysed using LC/MS-MS, was published (Kumar

et al. 2009). The authors claim that their method can be used to monitor

workers exposed to MDI. This method is methodologically relatively similar

to the method evaluated for bulky adducts in Paper I. The method detects

specific isocyanates and may be well-suited for measurements of aromatic

isocyanates but not for low-molecular weight isocyanates like isocyanic acid

due to difficulties of the enrichment.

In a carbamoylation3 reaction, an isocyanate can form an irreversible adduct

to the N-terminal valine in Hb. In valine, the nucleophilic primary amine

group is a common site for adduct formation. This adduct can be detached

from the protein by mild acid hydrolysis and analysed for biomonitoring of

exposure to the isocyanate (Figure 15). The most simple isocyanate is

isocyanic acid with only a hydrogen in the R-position. The analyte formed

from isocyanic acid is the lipophilic valine hydantoin (VH). By this method

a liphophilic analyte is formed from a hydrophilic reactive compound.

Figure 15. The formation of valine hydantoin from isocyanate, which in vivo form a stable adduct with

the nitrogen in e.g. the N-terminal valine in haemoglobin. The adduct can be detached in vitro by acid

hydrolysis, using LC/MS-MS for analysis.

A relatively crude method for LC/UV-analysis of VH has been developed

and used in research about uremia. The carbamoylated adduct was directly

detached from the protein by acid hydrolysis of the blood samples. The VH

was thereafter extracted and analysed (Kwan et al. 1990; Wynckel et al.

2000). A refined version of this method using precipitated globin was

described by Mraz et al. (1999) and was the method, which with only minor

modifications was used for the work-up of the samples analysed in Paper IV

and V.

3 The use of the terms carbamoylation and carbamylation are discussed by different authors,

some claim that the meaning is the same and some that it differs (Jelkmann 2008). Interna-

tional Union of Pure and Applied Chemistry (IUPAC) recommend carbamoylation, which

therefore is used in this thesis.

valine hydantoin

the detached adduct,

LC/MS (ESI+) m/z 143

H3C CH3

NH2

O

NH

Hb

N

C

O

3CH

NH

O

NH

Hb

O

HN

R

N-terminal

valine in Hb

isocyanate

carbamoylation

carbamoylated

N-terminal valine

(hydantoin adduct)

hydrolysis

R

NH

R

CH3

H+

in vivo in vitroCH3

CH3

N

49

6.5 Analysis of isocyanate adducts to the N-terminal

valine of Hb (Paper IV)

The use of the N-terminal valines in Hb for biomonitoring of isocyanates has

been hampered by the difficulty to obtain reference compounds due to side

reactions and polymerisation of the isocyanate used as starting material. In

Paper IV, a new general method for the synthesis of reference compounds

was published. This method for easier production of reference standards

opens for possibilities of new applications in studies of isocyanates. In the

new method a common precursor is used for the synthesis of various mono-

and diisocyanates-valine hydantoins. In Paper IV, the reference compounds

of isocyanic acid, MIC, phenyl isocyanate (PIC) and 2,6-toluenediisocyanate

as the valine hydantoins were synthesised using the general method

developed.

The contribution to Paper IV discussed in this thesis, was the establishment

of an LC/MS-MS (ESI+) method for the characterisation and analysis of the

reference standards synthesised. The different valine hydantoins yields some

common fragments in the LC/MS analyses. Using PIS, a fragment

originating from the N-terminal valine was determined to be 72. Another

fragment specific for all adducts, is from the loss of CO (-28) (Figure 16).

The limit of quantification (LOQ, S/N >10) was less than 1 pmol VH inject-

ed.

Figure 16. A LC/MS product ion scan of valine hydantoin (VH). Some fragments are typical for the

analysis of the adducts formed from different isocyanates. For VH the m/z for the molecular ion [M+H]+

is 143. In the product ion scan mode (PIS) the molecule gives two major fragments, which both were

found typical for the valin hydantoins of the studied isocyanates.

The LC/MS method was thereafter applied for analysis of a few samples of

human control blood for determination of the background level of the valine

hydantoin (Figure 2, in Paper IV). Globin was precipitated from the erythro-

115.2

143.2

50 60 70 80 90 100 110 120 130 140 150

Inte

nsity

97.2 102.1

72.2

-28

m/z, Da

28

m/z 72

HN

HN

O

O

[M+H]+

50

cytes in the blood. A mixture of concentrated acids was added and the acid

hydrolysis was performed at 100°C. The pH of the mixture was increased

and the VH extracted with ethyl acetate. The measured adduct level from

isocyanic acid in the samples was quantified to approximately 200 nmol

VH/g globin, which was in the range of previously observed levels

(Wynckel et al. 2000). The S/N for analysis of VH in the human blood

samples was about 1000, consequently no large efforts was performed to

achieve a lower LOD of the method.

For detection of unknown adducts, samples prepared from human blood

were analysed using LC/MS-MS in precursor ion scan (PREC)4 mode with

m/z 72. One possible adduct was observed, but not identified.

6.6 Evaluation of VH as a biomarker of early renal toxicity (Paper V)

6.6.1 Uremia

Urea and creatinine are end products in the amino acid and protein

metabolism. Urea is transported from the blood to the kidneys and is

normally excreted in the urine. Urea spontaneously dissociates to ammonium

and cyanate ions and isocyanic acid is formed (Figure 17).

Figure 17. The endogenous formation of isocyanic acid from spontaneous dissociation of urea.

For persons with uremia, the kidneys are not working properly and the

endogenous concentration of urea increases leading to increased

concentrations of isocyanic acid as well (e.g. (Kraus et al. 2001)).

4 PREC is the opposite of PIS. In PREC the second mass analyser is set to a fixed mass and a

scan is performed in the first mass analyser.

C

O

H2N NH2urea

NH4+ -N C O+

cyanate

HN C O

isocyanic acid

51

6.6.2 Biomarker for adverse effect to kidneys caused by a

radioactive cytostatic drug

Neuroendocrine tumours (tumours that arise from cells in the endocrine and

nervous system) can be treated with radioactive 177

Lu-[DOTA0,Tyr

3]-

octreotate (177

Lu-octreotate). The tumours express somastostatin receptors

(SSTR) and since the 177

Lu-octreotate is a somastostatin analogue it is a

target seeker with the tumours as the target. The kidneys and the bone

marrow are the main dose-limiting organs after exposure of 177

Lu-octreotate.

The administrated dose of 177

Lu-octreotate should be as high as possible to

exterminate the cancer cells but at a level were the dose-limiting organs are

unaffected. The uptake and retention of 177

Lu-octreotate have large inter-

individual variations (Larsson et al. 2012) and biomarkers of early risks of

renal injury is needed to enable individual treatment for this type of cancer.

As a part of Paper V, the detached N-terminal valine isocyanate adduct was

investigated as a biomarker for early nephrotoxic effect (poisonous effect on

the kidneys) in mice administrated the radioactive cytostatic.

6.6.3 Experiment

Mice (n=6 per dose) were injected with 0, 60 or 120 MBq of 177

Lu-

octreotate, corresponding to ca 0, 21 and 42 gray of absorbed dose in the

kidneys. Samples of urine were collected before administration and at days

14, 30, 60 and 90. The urine samples were used for measurement of retinol

binding protein 4 (RBP4) and creatinine levels. One kidney for histological

evaluation and blood samples for analysis of VH were collected at the time

for sacrifice after 90 days.

VH was evaluated as a possible biomarker of changes in the level of urea.

The blood samples were prepared as briefly described in section 6.5. Phenyl

valine hydantoin (PVH) was used as a volumetric standard in the analysis of

VH. In the LC/MS analyses of the samples only a small non-significant in-

crease was observed in the highest dose-group compared to the control

group. Either, the renal failure was not severe enough to affect the level of

urea leading to elevated VH levels compared to the natural background level

(see Figure 3, Paper V), or the LC/MS method was not sensitive enough to

differentiate between the small changes in the VH concentrations that might

have arisen. The peaks corresponding to VH in the analyses were always

more than 10 times S/N.

The LC/MS analyses of the mice samples had a CV of 2-21% in the samples,

which were injected three times. To control the repeatability of the method,

triplicate Hb samples from an untreated mouse were processed and analysed

52

(CV 11%). In this study there was no isotope-substituted VH available as

internal standard, which would have compensated for differences in the MS-

analysis of the same sample.

The histological analyses of the kidney tissue did not indicate any sign of

renal injury associated with the exposure of 177

Lu-octreotate. However, the

mean value of the RBP4/creatinine level ratio at the different sampling days

in both dose groups was statistically significant different from the mean

value of the respective dose collected at day 0. At the highest dose (120

MBq) the RBP4 levels increased at an earlier time point and reached a high-

er level compared to the other doses.

6.7 Summary

A LC/MS-MS method for the analysis of valine hydantoin adducts from

different isocyanates was established. The LC/MS-MS method was applied

to control human blood samples and the amount of VH was quantified to

approximately 200 nmol VH/g globin. The reference adducts were made

with a new general method for the synthesis of N-terminal valine isocyanate

adducts, which was the main work of paper IV.

VH is a result of adduct formation to N-teminal valine from endogenously

formed isocyanic acid after dissociation of urea. Irradiation used for cancer

therapy has negative effect on the kidneys. VH was therefore suggested as a

possible biomarker of kidney failure, which was evaluated in Paper V. In the

study, mice were administrated with 177

Lu-octreotate. The blood was

processed and the LC/MS-MS method from Paper IV was used for the anal-

ysis of VH with minor modifications. The conclusion from the conditions in

this experiment was that VH is not an applicable biomarker for early detec-

tion of renal toxicity caused by the cytostatic drug. However, protein RBP4

was a promising biomarker in urine for early detection of this toxicity.

53

7 Final discussion

A general aim for this thesis was to broaden the approach to use protein ad-

ducts for in vivo dose measurement to bulky adducts, particularly studying

the most genotoxic metabolite (+)-anti-BPDE of BP. A method for the

detection and quantification by LC/MS of adducts to SA from bulky adducts

was developed in Paper I. The LOD achieved was similar to the LOD re-

ported for the most sensitive MS methods for measurement of derivatised Hb

adducts from low-molecular weight compounds by GC/MS-MS and LC/MS-

MS. Yet, the LOD of this method in its present version is most probably not

sufficiently low for measurement of PAH adducts in exposed humans.

The method was applied to samples from exposed mice. A specific adduct

could be attributed to (+)-anti-BPDE in vivo. The BPDE-adduct levels in SA

in BP-exposed mice was shown to be very low, although in vitro experi-

ments showed an efficient adduct formation from BPDE to His in SA. Due

to the low levels of BPDE-adducts in SA from BP-exposed mice encoun-

tered in the analysis by this method, an LC/MS method for DNA adduct

analysis was established. The aim was to compare efficiency of adduct for-

mation to SA and DNA from adducts formed with BPDE in vitro. This com-

parison showed that the adduct formation to DNA was 103 times more effec-

tive than to SA. For the LC/MS-MS method established for analysis of DNA

adducts in Paper II, the LOD was in the same range as achieved by others

(Beland et al. 2005; Singh et al. 2006).

For both the SA- and the DNA-adduct methods used in this thesis the LOD

highly depends on the MS-instrument used for detection. Repeated injections

of the same samples showed large variations (up to 30%) in the LC/MS-MS

analysis. This was through-out the major contribution to the total method

variation in the quantification of the adducts. With a proper IS, this problem

would have been reduced. The conclusion drawn after studying the literature

of this field is that quantification of adducts often has inherent methodologi-

cal uncertainties of different types.

DNA adducts are critical for the initiation of cancer and analysis of DNA

adducts are important for measurement of target dose of the genotoxic agent

to improve cancer risk assessments. However, in human bio-monitoring

protein adducts have several advantages. During the lifetime of a protein,

54

adducts formed from chronic exposure accumulate. Protein adducts are good

markers of dose in blood and can be used as a substitute for target dose. Alt-

hough, in studies when samples from individual mice are used as in the pre-

sent thesis, the short half-life (~2 days) and the low amount of mSA availa-

ble, are limitations in analysis of SA adducts. After acute exposure to BP in

mice, DNA is a good alternative biomacromolecule to SA for analysis of

adducts from BPDE due to its high efficiency of adduct formation and since

a large amount of DNA can be isolated from the liver of one mouse.

To be able to estimate the risk of a toxic effect from a compound the dose in

the meaning “area under the concentration time curve” (AUC) of the geno-

toxic compound/metabolite should be measureable. In the general human

population, BP-metabolites analysed as specific adducts will possibly never

be reliably quantified due to the low adduct levels formed. Animal data may

be possible to use for extrapolation of adduct levels to exposures relevant to

humans. However, the dose measurement by adducts in humans would make

it possible to adjust for metabolic differences. One possibility to estimate

doses of reactive metabolites is by combined extrapolation of in vivo data

(mice, rat) and in vitro metabolism systems using different species (Motwani

et al. 2014). From an ethical point of view, the set-up of a scientific study

should strive to reduce animal experiments and the number of animals used,

as performed in this thesis.

In animals exposed to high doses of BP the measured BPDE-His levels in

SA in vivo are low. This is due to the low doses of the BPDE-metabolites

formed. The adducts formed by the diolepoxides of BP to SA (Helleberg

2001) and DNA (Beland et al. 2005) are probably only a fraction of all ad-

ducts formed after the exposure of BP. Around 10% of the total radioactivity

in DNA was ascribed BPDE-dG adducts measured by LC/MS-MS after ad-

ministration of radiolabelled BP to mice (Beland et al. 2005). Studies of

blood protein adducts in mice administered with radiolabelled BP have also

indicated that the majority of the adducts formed originated from other reac-

tive metabolite(s) than (±)-anti-BPDE (Helleberg 2001; Naylor et al. 1990).

A hypothesis is that they originate from metabolically formed quinones.

Further research is needed to elucidate this.

Despite the methodological difficulties it was possible to make a preliminary

estimation of the dose in vivo of the most genotoxic metabolite (+)-anti-

BPDE in exposed mice from the corresponding SA adduct. Such data made

it also possible to make a preliminary comparison of the genotoxic potency

of BPDE using γ-radiation as a standard, with regard to fMNPCE in vivo in

mice. Further studies are required to strengthen the estimation of genotoxic

potency and make more accurate comparisons of genotoxic risk with regard

to exposure to BP and e.g. to acrylamide from food. The key for such com-

55

parisons is the AUC in vivo and in vitro, which in vivo could be measured by

adducts. The estimation of risk at low doses of BP and PAH is a more com-

plicated issue than only measuring in vivo dose of (+)-anti-BPDE, e.g. what

happens at very low doses (Madureira et al., 2014) and in combination of

exposures (Beland et al., 2005). Though improved risk estimations have to

rely on quantitative and specific methods. The studies in this thesis could be

considered as basic methodological work in this direction.

56

8 Svensk sammanfattning

För att beräkna risken för toxisk effekt från olika substanser behöver man

kunna mäta hur mycket av ämnet som finns i kroppen (intern exponering).

Långlivade ämnen kan man mäta i sin grundform eller som metaboliter in

vivo (i kroppen) medan reaktiva ämnen ofta är för kortlivade för att kunna

mätas. Istället för att analysera det reaktiva ämnet i sig kan stabila reaktions-

produkter (addukter) med biomakromolekyler analyseras. Addukter kan

bildas när ett elektrofilt reaktivt ämne (elektrofil) reagerar med en nukleofil

atom, t.ex. O, N eller S i exempelvis DNA eller blodproteinerna serum al-

bumin (SA) och hemoglobin (Hb). Addukten kan användas som en biomar-

kör för intern exponering/interndos av det reaktiva ämnet.

Avhandlingen är främst inriktad på utveckling av metoder för kvantifiering

med masspektrometri (MS) av biomarkörer för intern dos som kan användas

som verktyg i proceduren för att uppskatta risken att utveckla cancer till

följd av exponering för elektrofila ämnen, som polycykliska aromatiska

kolväten (PAH) (Paper I-III). I avhandlingen ingår också arbeten där en me-

tod för MS har utvecklats för att analysera isocyanter in vivo. Metoden an-

vändes sedan för att undersöka om isocyansyra kan användas som biomarkör

av njurskada (Paper IV-V).

PAH är grupp av föreningar som bildas vid ofullständig förbränning och

finns spridda överallt i vår omgivning. För allmänheten sker den största ex-

poneringen för PAH via mat och luft. Benso[a]pyren (BP), di-

benso[a,h]antracen (DBA) och dibenso[a,l]pyren (DBP) är exempel på PAH

(Figur 3). När PAH metaboliseras bildas elektrofilt reaktiva ämnen, t.ex.

diolepoxider (DE) (Figur 5). Från BP kan olika isomera och stereoisomera

DE bildas (Figur 4) med olika reaktivitet och toxicitet.

I Paper I utvecklades en metod för att analysera stabila addukter till SA från elektrofila ämnen såsom diolepoxider från PAH. Metoden bygger på att fälla ut proteinet, bryta ner det med enzymer och anrika de modifierade aminosy-rorna med fastfasextraktion eller vätskekromatografi (LC). Därefter sker analys med LC/MS-MS, vilket ger både strukturinformation och hög selekti-vitet. Vid inkubering av SA från människa med (±)-anti-BPDE bildas främst addukt till ett specifikt histidin. Den utvecklade metoden var enkel att utföra, samt visade sig vara robust och ha hög reproducerbarhet. Metodens detekt-

57

ionsnivå är densamma för de tre PAHDE som studerades, ca 1 fmol ad-dukt/mg SA (vid injektion av 5 mg SA).

I Paper II användes den metod som utvecklades i Paper I. Den mest gentox-

iska enantiomeren av BPDE, (+)-anti-BPDE, visades bilda addukt med

histidin i SA från möss efter reaktion in vitro (i provrörsförsök). Därefter

detekterades och kvantifierades, för första gången med MS-analys, specifika

addukter från BPDE i SA från möss som exponerats för BP (100 mg BP/kg

kroppsvikt). Nivån av addukter från BPDE i SA visade sig vara väsentligt

lägre än i DNA (isolerat från levern) som mätts tidigare i samma möss

(Singh et al. 2006). För att ytterligare studera sambandet mellan DNA- och

SA-addukter sattes en analysmetod upp för detektion och kvantifiering av

addukter mellan BPDE och deoxyguanosin (dG) i DNA. Därefter gjordes en

inkubering av humant helblod med (±)-anti-BPDE in vitro. Linjäriteten mel-

lan adduktnivån i SA och den tillsatta mängden BPDE var mycket hög

(r2=0,998), vilket ytterligare påvisade att den använda metoden var robust.

Skillnaden mellan de analyserade addukterna till histidin i SA och till dG i

DNA var i det här försöket större än i försöket med exponerade möss. Slut-

satsen som drogs var att reaktiviteten av BPDE mot histidin är mycket lägre

än mot dG.

I Paper III identifierades den addukt som bildats från (+)-anti-BPDE i BP-

exponerade möss (25, 50 och 100 mg BP/kg kroppsvikt). Identifieringen

gjordes genom att analysera SA alkylerat in vitro med olika stereoisomerer

av BPDE och jämförelse med de i musproven erhållna addukttopparna. Ni-

vån av bildade addukter till SA och DNA samt frekvensen av mikrokärnor

(ett mått på kromosombrott som kan orsakas av genotoxiska föreningar) i de

exponerade mössen jämfördes också. Dessa data användes även för att göra

en preliminär uppskattning av relativ genotoxisk potens av BP-exponering.

Isocyanater är en klass av ämnen som är vanliga inom kemisk industri för

produktion av t.ex. bekämpningsmedel och polyuretanskum (används som

tätningsmedel och i möbelstoppning). Exponering för isocyanater är den

mest bidragande orsaken till yrkesrelaterad astma. Redan mycket låga halter

kan medföra besvär i ögon och luftvägar. Vi har en endogen (intern) expone-

ring för den enklaste isocyanaten (isocyansyra), vilken bildas från urea som

är en restprodukt från kroppens proteinnedbrytning. Personer med nedsatt

njurfunktion kan få högre koncentration av urea och därmed också högre

koncentration av isocyansyra i kroppen. Isocyanater är starkt elektrofila och

kan bilda addukter med t.ex. det N-terminala valinet i Hb.

Eftersom isocyanater är så reaktiva är det svårt att tillverka referens-

substanser av dessa föreningar, vilket är ett tema i Paper IV. Ett delarbete i

Paper IV var att utveckla en LC/MS-MS-metod för att karaktärisera och

58

analysera referenssubstanser av addukter från isocyanater som bundit till

valin i Hb och därefter klyvts av från proteinet med sur hydrolys. Den av-

klyvda addukten är en hydantoin bestående av isocyanaten och det N-

terminala valinet (valinhydantoin, VH) (Figur 15). Blodprover från männi-

ska upparbetades och bakgrundsnivån av VH beräknades till cirka 200 nmol

VH/g globin, vilket var i samma nivå som tidigare mätningar (Wynckel et al.

2000).

Neuroendokrina cancertumörer har celler som kan uttrycka receptorer för

somastotin. För behandling av dessa tumörer används radioaktivt 177

Lu-

[DOTA0,Tyr

3]-octreotate (

177Lu-octreotate) eftersom det kan binda till so-

mastotinreceptorer. 177

Lu-octreotate är på grund av detta en ”målsökande”

cytostatika och strålar tumören mer specifikt än konventionell strålbehand-

ling av cancer. Den interindividuella variationen av 177

Lu-octreotate-upptag

är stor och njurar och benmärg är de organ som är mest känsliga. För att få

en effektiv cancerbehandling vill man ha en så hög dos av cytostatikan som

möjligt utan att skada organen. En biomarkör för tidiga förändringar i de

mest känsliga organens kapacitet behövs för att kunna dosjustera behand-

lingen individuellt. Ett delarbete i Paper V bestod i att undersöka om VH kan

användas som biomarkör för tidig njurskada. Möss injicerades med 177

Lu-

octreotate (0, 60 eller 120 MBq) och blodprov togs 90 dagar efter expone-

ring. Man kunde se en liten ökning av VH-nivån men skillnaden var inte

signifikant mellan kontroll- och doserade möss. VH kan därför inte anses

som någon bra biomarkör under de rådande experimentella betingelserna.

Detta beror troligen på att koncentrationen av urea inte förändrades nämn-

värt eller på att metoden inte var tillräckligt känslig för att kunna upptäcka

förändringar i nivån av VH över en kort tid jämfört med bakgrundsnivån av

addukter som ackumuleras över Hb:s livslängd i mössen.

59

9 Acknowledgements

Jag har ju varit på miljökemi ganska länge och många personer har kommit

och gått under arbetets gång och årens lopp. Bara en ena söndag har jag känt

av det andra pratar om som söndagsångest. Det är bara delvis för att det är så

roligt att forska främst är det tack vare ”miljökemiandan”.

Av alla er som jag vill tacka är det främst Margareta för många år av spän-

nande forskning! Trots att det så ofta har gått trögt och känts som ett steg

fram och två steg bak har Du lyckats få mig att tycka att det är spännande så

fort vi diskuterar. Det är imponerande hur du orkar!

Tack också min, på senare tid, biträdande handledare Dan som alltid svarar

snabbt på mina lustiga mailfrågor. Såklart också ett stort tack till mina med-

författare – tack alla ni!

Rum C308 har en bra atmosfär. Tack mina ”roomisar”, både gamla och nya.

Anna Malmvärn – världens bästa exjobbshandledare, rumskompis och nu en

av mina närmaste vänner. Jana – en härlig roomie som delat med sig av sin

gästvänlighet över hela Europa. Vår tejprulle fick heta ”AJE” och det gör

den än men nu står det för de ”nya” och lika underbara roomisarna Anna-

Karin och Jessica – för sällskap på kvällar och helger, för behövliga diskuss-

ioner i stort som smått och vilka fina blommor vi har (eller i alla fall gröna

växter för blommorna är ju inte så mångfaldiga)!

Åke – du sa att jag hade en glöd i ögonen att vilja fortsätta, den har jag tänkt

på när jag ibland tyckt att det varit tungt och så har jag stretat vidare. Carl

Axel, Sören – inga blommor att tala om men fin är hibiskusen i alla fall.

Riskgruppen – Ulla – jag har saknat dig som vapendragare många gånger,

Hans – som alltid ställer upp, Hitesh – det har varit extra kul att ha jobbat

tillsammans nu på slutet synd att vi inte började tidigare, Henrik – MKs nya

mediala kändis, tidningar, utbildningsfilmer etc., Cissi – det är bara du som

har haft en coati i väskan, Birgit – tack för alla extra omtankar du ger, Jenny

– en frisk fläkt med otaliga trådar till miljökemi, Anna V – vi stod i val och

kval om att doktorera men nu står vi (till slut) båda med boken i vår hand,

Lucia och Lisa – perfect with two baking girls when I had a busy autumn.

Charlotta – värdefull som praktiskt och beslutstagande bollplank, Johanna –

60

goda tårtrecept och allt är möjligt, Per – nja, nu på slutet har det varit dåligt

med skräp-tv men rätt länge höll jag i, Jeanette, Ronnie, Virginia, Cristina,

Linda Ö.

Lillemor – exjobbshandledare som gav mig ett positivt första intryck av mil-

jökemi, Anita – en skön klippa att prata med, Maria S, Lisa, Göran, Sara,

Anna C, Tati, Ulrika, Andreas – gänget på syntes, jag saknar er! Mats – på

gott och ont, kinonerna var problematiska även för dig, Karin N, Hrönn,

Lotta, Maria, Johan E – ovärderlig hjälp vid maskintrubbel, Ge, Yihui, Vi-

vian, Johan F – vi klarar nästan att hålla tiderna för fika och lunch men vi

saknar dig ändå, Karin L – Tack för att du vågade sitta kvar på skotern, Anna

S – jag saknar dig i ostklubben, Linda L – många resor tillsammans Sval-

bard, Amsterdam, Geilo, Island och Angera, Ioannis – det är trevligt när du

kommer in och ”stör” lite i det upptagna rummet ”tack för samtalen”, Dennis

– som alltid har den värsta story att berätta.

Min familj – Patrik, Myra och Hilde – ni är de bästa ♥♥♥ och nu ska jag se

och lyssna, inte bara jobba. Vi är en 4-klöver mitt på en blomsteräng av

andra kära och viktiga personer – Mamma och Pappa ni är underbara som

kommer till undsättning i alla (o-)möjliga situationer och resten av vår släkt

och våra vänner. Ni är alla OVÄRDERLIGT viktiga! Jag kan inte nämna er

alla – ni för många (underbart lyxproblem) men ta åt er av er högst betydel-

sefulla viktighet! Jag har er i mitt hjärta! Kram.

61

Reference List

Abramsson-Zetterberg, L., Darnerud, P. O., and Wretling, S. (2014) Low intake of polycyclic aromatic

hydrocarbons in Sweden: Results based on market basket data and a barbecue study. Food

Chem. Toxicol., 74107-111

Abramsson-Zetterberg, L. (2003) The dose-response relationship at very low doses of acrylamide is

linear in the flow cytometer-based mouse micronucleus assay. Mutat. Res. , Genet. Toxicol. En-

viron. Mutagen., 5352: 215-222

Abramsson-Zetterberg, L., Vikström, A. C., Törnqvist, M., and Hellenäs, K. E. (2008) Differences in the

frequency of micronucleated erythrocytes in humans in relation to consumption of fried carbo-

hydrate-rich food. Mutat. Res. , Genet. Toxicol. Environ. Mutagen., 6531-2: 50-56

Akabori, S., Ohno, K., and Narita, K. (1952) Hydrazinolysis of proteins and peptides: method for the

characterization of carboxyl terminal amino acids in proteins. Bull. Chem. Soc. Jpn., 25214-218

Arif, J. M., Smith, W. A., and Gupta, R. C. (1999) DNA adduct formation and persistence in rat tissues

following exposure to the mammary carcinogen dibenzo[a,l]pyrene. Carcinogenesis, 206: 1147-

1150

Augustinsson, K.-B., Experimentell biokemi (Bonniers, Danmark, 1966)

Balbo, S., Turesky, R. J., and Villalta, P. W. (2014) DNA Adductomics. Chem. Res. Toxicol., 273: 356-

366

Beland, F. A., Churchwell, M. I., Von Tungeln, L. S., Chen, S., Fu, P. P., Culp, S. J., Schoket, B.,

Gyorffy, E., Minarovits, J., Poirier, M. C., Bowman, E. D., Weston, A., and Doerge, D. R.

(2005) High-Performance Liquid Chromatography Electrospray Ionization Tandem Mass Spec-

trometry for the Detection and Quantitation of Benzo[a]pyrene-DNA Adducts. Chemical Re-

search in Toxicology, 188: 1306-1315

Beyerbach, A., Farmer, P. B., and Sabbioni, G. (2006) Biomarkers for Isocyanate Exposure: Synthesis of

Isocyanate DNA Adducts. Chem. Res. Toxicol., 1912: 1611-1618

Boysen, G. and Hecht, S. S. (2003) Analysis of DNA and protein adducts of benzo[a]pyrene in human

tissues using structure-specific methods. Mutation Research, 5431: 17-30

Brown, J. R. and Thornton, J. L. (1957) Percivall Pott (1714-1788) and chimney sweepers' cancer of the

scrotum. Br J Ind Med, 141: 68-70

Brunmark, P., Harriman, S., Skipper, P. L., Wishnok, J. S., Amin, S., and Tannenbaum, S. R. (1997)

Identification of Subdomain IB in Human Serum Albumin as a Major Binding Site for Polycyclic

Aromatic Hydrocarbon Epoxides. Chemical Research in Toxicology, 108: 880-886

Buckley, T. J. and Lioy, P. J. (1992) An examination of the time course from human dietary exposure to

polycyclic aromatic hydrocarbons to urinary elimination of 1-hydroxypyrene. Br. J. Ind. Med.,

492: 113-124

Cheng, S. C., Hilton, B. D., Roman, J. M., and Dipple, A. (1989) DNA adducts from carcinogenic and

noncarcinogenic enantiomers of benzo[a]pyrene dihydrodiol epoxide. Chem Res Toxicol, 25:

334-340

Chung, M. K., Regazzoni, L. G., McClean, M., Herrick, R., and Rappaport, S. M. (2013) A sandwich

ELISA for measuring benzo[a]pyrene-albumin adducts in human plasma. Anal Biochem, 4352:

140-149

Cook, J. W., Hewett, C. L., and Hieger, I. (1933) Isolation of a cancer-producing hydrocarbon from coal

tar. I. Concentration of the active substance. J. Chem. Soc.395-405

Day, B. W., Skipper, P. L., Zaia, J., Singh, K., and Tannenbaum, S. R. (1994) Enantiospecificity of Cova-

lent Adduct Formation by Benzo[a]pyrene anti-Diol Epoxide with Human Serum Albumin.

Chemical Research in Toxicology, 76: 829-835

62

Day, B. W., Skipper, P. L., Zaia, J., and Tannenbaum, S. R. (1991) Benzo[a]pyrene anti-diol epoxide

covalently modifies human serum albumin carboxylate side chains and imidazole side chain of

histidine(146). Journal of the American Chemical Society, 11322: 8505-8509

EFSA (2008) Polycyclic aromatic hydrocarbons in food Scientific Opinion of the Panel on Contaminants

in the Food Chain. The EFSA journal, 72466-114

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. Mu-

tat. Res., 242: 83-103

Ehrenberg, L., Moustacchi, E., and Osterman-Golkar, S. (1983) International Commission for Protection

Against Environmental Mutagens and Carcinogens. Dosimetry of genotoxic agents and dose-

response relationships of their effects. Mutat Res, 1232: 121-182

Ehrenberg, L., Granath, F., and Törnqvist, M. (1996) Macromolecule adducts as biomarkers of exposure

to environmental mutagens in human populations. Environ. Health Perspect. Suppl., 104Suppl.

3: 423-428

European Food Safety Authority (EFSA) (2008) Polycyclic aromatic hydrocarbons in food Scientific

Opinion of the Panel on Contaminants in the Food Chain. The EFSA journal, 72466-114

Falco, G., Domingo, J. L., Llobet, J. M., Teixido, A., Casas, C., and Mueller, L. (2003) Polycyclic aro-

matic hydrocarbons in foods: human exposure through the diet in Catalonia, Spain. J. Food

Prot., 6612: 2325-2331

Furne, J. K., Springfield, J. R., Ho, S. B., and Levitt, M. D. (2003) Simplification of the end-alveolar

carbon monoxide technique to assess erythrocyte survival. J. Lab. Clin. Med., 1421: 52-57

Ginsberg, G. L. and Atherholt, T. B. (1990) DNA adduct formation in mouse tissues in relation to serum

levels of benzo(a)pyrene-diol-epoxide after injection of benzo(a)pyrene or the diol-epoxide. Can-

cer Res, 504: 1189-1194

Granath, F., Ehrenberg, L., Paulsson, B., and Törnqvist, M. (2001) Cancer risk from exposure to occupa-

tional acrylamide. Occup Environ Med, 589: 608-609

Grova, N., Feidt, C., Laurent, C., and Rychen, G. (2002) [14C] Milk, urine and feces excretion kinetics in

lactating goats after an oral administration of [14C]polycyclic aromatic hydrocarbons. Int.

Dairy J., 1212: 1025-1031

Hanahan, D. and Weinberg, R. A. (2011) Hallmarks of cancer: the next generation. Cell (Cambridge,

MA, U. S. ), 1445: 646-674

Helleberg, H., Biomarkers aiming at cancer risk estimation of polycyclic aromatic hydrocarbons. (Thesis,

Nykopia tryck AB, Stockholm, 2001)

Helleberg, H. and Törnqvist, M. (2000) A new approach for measuring protein adducts from ben-

zo[a]pyrene diolepoxide by high performance liquid chromatography/tandem mass spectrome-

try. Rapid Communications in Mass Spectrometry, 1418: 1644-1653

Herr, I. and Buechler, M. W. (2010) Dietary constituents of broccoli and other cruciferous vegetables:

implications for prevention and therapy of cancer. Cancer Treat. Rev., 365: 377-383

Hertzberg, T., Blomqvist, P., Dalene, M., and Skarping, G. (2003) Particles and isocyanates from fires.

Brandforsk project 324-021. 2003:05

IARC (International Agency for Research on Cancer) Some Non-heterocyclic Polycyclic Aromatic Hy-

drocarbons and Some Related Exposures (2010) Monographs on the Evaluation of Carcinogenic

Risks to Humans, volume 92 (Lyon, France: WHO).

Retrieved from: http://monographs.iarc.fr/ENG/Monographs/vol92/mono92.pdf

IARC A review of human carcinogens: chemical agents and related occupations. (2012). Monographs on

the Evaluation of Carcinogenic Risks to Humans, volume 100F (Lyon, France: WHO).

Retrieved from: http://monographs.iarc.fr/ENG/Monographs/vol100F/mono100F.pdf

63

IARC Agents Classified by the IARC Monographs, (2014) Volumes 1–11

Retrieved from: http://monographs.iarc.fr/ENG/Classification/ClassificationsGroupOrder.pdf

IPCS (International Programme on Chemical Safety) Selected non-heterocyclic polycyclic aromatic

hydrocarbons. (1998) Environmental Health Criteria 202 (Geneva, Switzerland: WHO). ISBN

92-4-157202-7

Jain, N. C., Essentials of veterinary hematology. (Lea & Febiger, USA, 1993) ISBN 0-8121-1437-X

Jelkmann, W. (2008) 'O', erythropoietin carbamoylation versus carbamylation. Nephrol Dial Transplant,

239: 3033

Kaefferlein, H. U., Marczynski, B., Mensing, T., and Bruening, T. (2010) Albumin and hemoglobin

adducts of benzo[a]pyrene in humans-Analytical methods, exposure assessment, and recommen-

dations for future directions. Crit. Rev. Toxicol., 402: 126-150

Klaassen, C.D., ed., Casarett and Doull’s toxicology: the basic science of poisons, 5th ed. (McGraw-Hill,

New York, NY, 1995) ISBN 0-07-105476-6

Koreeda, M., Moore, P. D., Wislocki, P. G., Levin, W., Yagi, H., and Jerina, D. M. (1978) Binding of

benzo[a]pyrene 7,8-diol-9,10-epoxides to DNA, RNA, and protein of mouse skin occurs with

high stereoselectivity. Science, 1994330: 778-781

Kotin, P., Falk, H. L., and Buser, R. (1959) Distribution, retention, and elimination of benzo[a]-pyrene-

C14 after administration to mice and rats. J. Natl. Cancer Inst. (1940-1978), 23541-555

Kotova, N., Hemminki, K., and Segerbäck, D. (2005) Urinary Thymidine Dimer as a Marker of Total

Body Burden of UV-Inflicted DNA Damage in Humans. Cancer Epidemiol. , Biomarkers Prev.,

1412: 2868-2872

Kraus, L. M., Gaber, L., Handorf, C. R., Marti, H. P., and Kraus, A. P., Jr. (2001) Carbamoylation of

glomerular and tubular proteins in patients with kidney failure: A potential mechanism of ongo-

ing renal damage. Swiss Med. Wkly., 13111/12: 139-145

Kulkarni, M. S., Calloway, K., Irigaray, M. F., and Kaufman, D. G. (1986) Species differences in the

formation of benzo(a)pyrene-DNA adducts in rodent and human endometrium. Cancer Res., 466:

2888-2891

Kumar, A., Dongari, N., and Sabbioni, G. (2009) New Isocyanate-Specific Albumin Adducts of 4,4'-

Methylenediphenyl Diisocyanate (MDI) in Rats. Chem. Res. Toxicol., 2212: 1975-1983

Kwan, J. T. C., Carr, E. C., Bending, M. R., and Barron, J. L. (1990) Determination of carbamylated

hemoglobin by high-performance liquid chromatography. Clin. Chem. (Winston-Salem, N. C. ),

364: 607-610

La, D. K. and Swenberg, J. A. (1996) DNA adducts: biological markers of exposure and potential appli-

cations to risk assessment. Mutat. Res. , Rev. Genet. Toxicol., 3651-3: 129-146

Landry, J. and Delhaye, S. (1996) A simple and rapid procedure for hydrolyzing minute amounts of

proteins with alkali. Anal. Biochem., 2431: 191-194

Larsson, M., Bernhardt, P., Svensson, J. B., Waengberg, B., Ahlman, H., and Forssell-Aronsson, E.

(2012) Estimation of absorbed dose to the kidneys in patients after treatment with 177Lu-

octreotate: comparison between methods based on planar scintigraphy. EJNMMI Res., 21: 49,

13

Madhavan, N. D. and Naidu, K. A. (1995) Polycyclic aromatic hydrocarbons in placenta, maternal

blood, umbilical cord blood and milk of Indian women. Hum Exp Toxicol, 146: 503-506

Madureira, D. J., Weiss, F. T., Van Midwoud, P., Helbling, D. E., Sturla, S. J., and Schirmer, K. (2014)

Systems Toxicology Approach to Understand the Kinetics of Benzo(a)pyrene Uptake, Biotrans-

formation, and DNA Adduct Formation in a Liver Cell Model. Chem. Res. Toxicol., 273: 443-

453

64

Melikian, A. A., Malpure, S., John, A., Meng, M., Schoket, B., Mayer, G., Vincze, I., Kolozsi-

Ringelhann, A., and Hecht, S. S. (1999) Determination of hemoglobin and serum albumin ad-

ducts of benzo[a]pyrene by gas chromatography-mass spectrometry in humans and their relation

to exposure and to other biological markers. Polycyclic Aromat. Compd., 171-4: 125-134

Miller, J. A. (1970) Carcinogenesis by chemicals: an overview--G. H. A. Clowes memorial lecture.

Cancer Res, 303: 559-576

Moore, C. J., Pruess, S., Mauthe, R. J., Gould, M. N., and Baird, W. M. (1987) Interspecies differences in

the major DNA adducts formed from benzo[a]pyrene but not 7,12-dimethylbenz[a]anthracene in

rat and human mammary cell cultures. Cancer Res., 4716: 4402-4406

Motwani, H. V. and Törnqvist, M. (2014) In vivo doses of butadiene epoxides as estimated from in vitro

enzyme kinetics by using cob(I)alamin and measured hemoglobin adducts: An inter-species ex-

trapolation approach. Toxicol. Appl. Pharmacol., 2813: 276-284

National Food Agency, Sweden. (No 7, 2012) Market Basket 2010 – chemical analysis, exposure estima-

tion and health-related assessment of nutrients and toxic compounds in Swedish food baskets. Retrieved from: http://www.slv.se/sv/grupp3/Rapporter/Rapportserien-1994-2009/

Naylor, S., Gan, L. S., Day, B. W., Pastorelli, R., Skipper, P. L., and Tannenbaum, S. R. (1990) Ben-

zo[a]pyrene diol epoxide adduct formation in mouse and human hemoglobin: physicochemical

basis for dosimetry. Chemical Research in Toxicology, 32: 111-117

Ott, M. G., Jolly, A. T., Burkert, A. L., and Brown, W. E. (2007) Issues in diisocyanate antibody testing.

Crit. Rev. Toxicol., 377: 567-585

Ozbal, C. C., Skipper, P. L., Yu, M. C., London, S. J., Dasari, R. R., and Tannenbaum, S. R. (2000)

Quantification of (7S,8R)-dihydroxy-(9 R,10S)-epoxy-7,8,9,10- tetrahydrobenzo[a ]pyrene ad-

ducts in human serum albumin by laser-induced fluorescence: Implications for the in Vivo Me-

tabolism of benzo[a]pyrene. Cancer Epidemiology, Biomarkers & Prevention, 97: 733-734

Pelkonen, O. and Nebert, D. W. (1982) Metabolism of polycyclic aromatic hydrocarbons: etiologic role

in carcinogenesis. Pharmacol Rev, 342: 189-222

Peltonen, K. and Dipple, A. (1995) Polycyclic aromatic hydrocarbons: Chemistry of DNA adduct for-

mation. J. Occup. Environ. Med., 371: 52-58

Peters, T., All about albumin – Biochemistry, Genetics, and Medical Applications,(Academic Press, New

York, NY, 1995) ISBN 978-0-12-552110-9

Philippe, R. J. and Honeycutt, R. G. (1965) Methyl isocyanate in cigaret smoke and its retention by an

adsorption-type filter. Tob. Int., 933-37

Phillips, D. H., Farmer, P. B., Beland, F. A., Nath, R. G., Poirier, M. C., Reddy, M. V., and Turteltaub, K.

W. (2000) Methods of DNA adduct determination and their application to testing compounds for

genotoxicity. Environ. Mol. Mutagen., 353: 222-233

Phillips, D. H. (1997) Detection of DNA modifications by the 32P-postlabelling assay. Mutat. Res. ,

Fundam. Mol. Mech. Mutagen., 3781,2: 1-12

Plaza, G., Ulfig, K., and Tien, A. J. (2000) Immunoassays and environmental studies. Pol. J. Environ.

Stud., 94: 231-236

Quinlivan, E. P. and Gregory, J. F., III (2008) DNA digestion to deoxyribonucleoside: A simplified one-

step procedure. Anal. Biochem., 3732: 383-385

Ragin, A. D., Crawford, K. E., Etheredge, A. A., Grainger, J., and Patterson, D. G., Jr. (2008) A Gas

Chromatography-Isotope Dilution High-Resolution Mass Spectrometry Method for Quantifica-

tion of Isomeric Benzo[a]pyrene Diol Epoxide Hemoglobin Adducts in Humans. J. Anal. Toxi-

col., 329: 728-736

65

Ramesh, A., Walker, S., Hood, D., Guillen, M., Schneider, K., and Weyand, E. (2004) Bioavailability and

Risk Assessment of Orally Ingested Polycyclic Aromatic Hydrocarbons. Int. J. Toxicol., 235:

301-333

Randerath, K., Reddy, M. V., and Gupta, R. C. (1981) Phosphorus-32 labeling test for DNA damage.

Proc. Natl. Acad. Sci. U. S. A., 7810: 6126-6129

Rill, R. L. and Marsch, G. A. (1990) Sequence preferences of covalent DNA binding by anti-(+)- and

anti-(-)-benzo[a]pyrenediol epoxides. Biochemistry, 2925: 6050-6058

Roche, C. J., Zinger, D., Geacintov, N. E., and Harvey, R. G. (1985) Enhancement of stability of 7β,8α-

dihydroxy-9α-epoxybenzo[a]pyrene by complex formation with serum albumin. Cancer Bio-

chem. Biophys., 81: 35-40

Rubino, F. M., Pitton, M., Di Fabio, D., and Colombi, A. (2009) Toward an "omic" physiopathology of

reactive chemicals: thirty years of mass spectrometric study of the protein adducts with endoge-

nous and xenobiotic compounds. Mass Spectrom. Rev., 285: 725-784

Sabbioni, G., Skipper, P. L., Buchi, G., and Tannenbaum, S. R. (1987) Isolation and characterization of

the major serum albumin adduct formed by aflatoxin B1 in vivo in rats. Carcinogenesis (Lon-

don), 86: 819-824

Saunders, C. R., Ramesh, A., and Shockley, D. C. (2002) Modulation of neurotoxic behavior in F-344

rats by temporal disposition of benzo(a)pyrene. Toxicol. Lett., 1291-2: 33-45

Singh, R., Gaskell, M., Le Pla, R. C., Kaur, B., zim-Araghi, A., Roach, J., Koukouves, G., Souliotis, V.

L., Kyrtopoulos, S. A., and Farmer, P. B. (2006) Detection and Quantitation of Benzo[a]pyrene-

Derived DNA Adducts in Mouse Liver by Liquid Chromatography-Tandem Mass Spectrometry:

Comparison with 32P-Postlabeling. Chemical Research in Toxicology, 196: 868-878

Singh, R., Teichert, F., Seidel, A., Roach, J., Cordell, R., Cheng, M. K., Frank, H., Steward, W. P., Man-

son, M. M., and Farmer, P. B. (2010) Development of a targeted adductomic method for the de-

termination of polycyclic aromatic hydrocarbon DNA adducts using online column-switching

liquid chromatography/tandem mass spectrometry. Rapid Commun. Mass Spectrom., 2416:

2329-2340

Skarping, G., Dalene, M., and Littorin, M. (1995) 4,4'-Methylenedianiline in hydrolyzed serum and urine

from a worker exposed to thermal degradation products of methylene diphenyl diisocyanate elas-

tomers. Int. Arch. Occup. Environ. Health, 672: 73-77

Skipper, P. L. and Tannenbaum, S. R. (1990) Protein adducts in the molecular dosimetry of chemical

carcinogens. Carcinogenesis (London), 114: 507-518

Sriramachari, S. (2004) The Bhopal gas tragedy: an environmental disaster. Curr. Sci., 867: 905-920

Stowers, S. J. and Anderson, M. W. (1984) Ubiquitous binding of benzo[a]pyrene metabolites to DNA

and protein in tissues of the mouse and rabbit. Chem. -Biol. Interact., 512: 151-166

Stowers, S. J. and Anderson, M. W. (1985) Formation and persistence of benzo[a]pyrene metabolite-

DNA adducts. EHP, Environ. Health Perspect., 6231-39

Straub, K. M., Meehan, T., Burlingame, A. L., and Calvin, M. (1977) Identification of the major adducts

formed by reaction of benzo[a]pyrene diol epoxide with DNA in vitro. Proc. Natl. Acad. Sci. U.

S. A., 7412: 5285-5289

The National Board of Health and Welfare, S. (2014) Official statistics of Sweden, Causes of death 2014-

8-5. ISBN 978-91-7555-202-6

Retrieved from: http://www.socialstyrelsen.se/publikationer2014/2014-8-5/Sidor/default.aspx

Tomatis, L., Aitio, A., Wilbourn, J., and Shuker, L. (1989) Human carcinogens so far identified. Jpn. J.

Cancer Res., 809: 795-807

Törnqvist, M., Fred, C., Haglund, J., Helleberg, H., Paulsson, B., and Rydberg, P. (2002) Protein ad-

ducts: quantitative and qualitative aspects of their formation, analysis and applications. Journal

66

of Chromatography, B: Analytical Technologies in the Biomedical and Life Sciences, 7781-2:

279-308

Verschoor, L. and Verschoor, A. H. (2014) Nonoccupational and occupational exposure to isocyanates.

Curr. Opin. Pulm. Med., 202: 199-204

von Stedingk, H., Vikström, A. C., Rydberg, P., Pedersen, M., Nielsen, J. K. S., Segerbäck, D., Knudsen,

L. E., and Tornqvist, M. (2011) Analysis of Hemoglobin Adducts from Acrylamide, Glycidamide,

and Ethylene Oxide in Paired Mother/Cord Blood Samples from Denmark. Chem. Res. Toxicol.,

2411: 1957-1965

Walker, J.M., ed., The protein protocols handbook, 2nd ed. (Humana Press, Totowa, NJ, 2002)

ISBN 0-89603-940-4

Weinstein, I. B., Jeffrey, A. M., Jennette, K. W., Blobstein, S. H., Harvey, R. G., Harris, C., Autrup, H.,

Kasai, H., and Nakanishi, K. (1976) Benzo[a]pyrene diol epoxides as intermediates in nucleic

acid binding in vitro and in vivo. Science, 1934253: 592-595

Westberg, E., Hedebrant, U., Haglund, J., Alsberg, T., Eriksson, J., Seidel, A., and Törnqvist, M. (2014)

Conditions for sample preparation and quantitative HPLC/MS-MS analysis of bulky adducts to

serum albumin with diolepoxides of polycyclic aromatic hydrocarbons as models. Anal. Bioanal.

Chem., 4065: 1519-1530

Westberg, E. A. C., Singh, R., Hedebrant, U., Koukouves, G., Souliotis, V. L., Farmer, P. B., Segerbäck,

D., Kyrtopoulos, S., and Törnqvist, M. A. (2015) Adduct levels from benzo[a]pyrenediol epox-

ide: Relative formation to histidine in serum albumin and to deoxyguanosine in DNA in vitro and

in vivo in mice measured by LC/MS-MS methods. Toxicol. Lett., 2321: 28-36

Wheeler, G. P. and Skipper, H. E. (1957) Mustards. III. In vivo fixation of carbon-14 from nitrogen

mustard-C14H3 in nucleic acid fractions of animal tissues. Arch. Biochem. Biophys., 72465-475

Wynckel, A., Randoux, C., Millart, H., Desroches, C., Gillery, P., Canivet, E., and Chanard, J. (2000)

Kinetics of carbamylated hemoglobin in acute renal failure. Nephrol. , Dial. , Transplant., 158:

1183-1188

Xue, W. and Warshawsky, D. (2005) Metabolic activation of polycyclic and heterocyclic aromatic hydro-

carbons and DNA damage: A review. Toxicol. Appl. Pharmacol., 2061: 73-93

Yalow, R. S. and Berson, S. A. (1959) Assay of plasma insulin in human subjects by immunological

methods. Nature (London, U. K. ), 184Suppl. No. 21: 1648-1649

Zaia, J. and Biemann, K. (1994) Proteinase K Activity Inhibitor Near Amino Acids Carrying Large Sub-

stituents: Three PAH Diolepoxides Covalently Modify His-146 Human Serum Albumin Residue.

Journal of the American Chemical Society, 11616: 7407-7408