.—___—_‘_ _‘ 777*”

 

LIBRARY
Michigan State

University
\A A

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative Action/Equal Opportunity Institution
cWWMH.‘

 

 

 

 

 

 

 

 

 

EFFECT OF STRAND-SPECIFIC EXCISION REPAIR ON THE SPECTRA
OF HUTATIONS INDUCED BY BENZOIAIPYRENE-DIOL EPOXIDE
AND ULTRAVIOLET RADIATION IN DIPLOID HUMAN CELLS

By

Ruey-Huu Chen

A DISSERTATION

Sublitted to
Hichigan State University
in partial fulfiIIIent of the require-eats

for the degree of

DOCTOR OF PHILOSOPHY

Depart-eat of Biochelistry

1991

ABSTRACT

EFFECT OF STRAND-SPECIFIC EXCISIDN REPAIR ON THE SPECTRA
OF MUTATIDNS INDUCED BY BENZDlAlPYRENE-DIDL EPOXIDE
AND ULTRAVIOLET RADIATION IN DIPLDID HUMAN CELLS

By

Ruey-Hwa Chen

To study the effect of excision repair on the spectra of mutations
induced in dipioid human ceils by UV and (:)—7B,8a-dihydroxy-9a,loo-epoxy-
7,8,9,10-tetrahydrobenzo[alpyrene (BPDE), I synchronized repair-proficient
ceiis, treated them at the beginning of S phase or in G1 phase severai
hours prior to the onset of S phase, seiected for thioguanine resistant
ceiis, and determined the spectra of mutations in the coding region of the
hypoxanthine(guanine)phosphoribosyi—transferase (HPRT) gene in the
mutants. As a controi, I compared the spectra of mutations similariy
induced in repair-deficient xeroderma pigmentosum (XP) ceiis. There‘was no
difference in the kinds of mutations observed in mutants derived from
either ceii strain treated with a particuiar mutagen either in S or in 6,.
With BPDE, the majority were G.C->T.A transversions; with UV, they were
mainiy G.C->A.T transitions. However, the strand distribution of
premutagenic iesions in mutants from repair-proficient ceiis treated in S
or G1 differed significantiy. In mutants derived from ceiis treated with

BPDE in S, such 1esions were located in either stand, with 25% being in

the transcribed strand; with mutants derived from 6,, all of them were in
the nontranscribed strand. With UV, the strand distribution of such
lesions in the S-derived mutants was 71 transcribed: 29 nontranscribed,
but in the G,-derived mutants, this ratio was 20: 80. In contrast to what
was found in repair-proficient cells, the ratio did not differ in mutants
derived from XP cells treated in S or 6,. This ratio was very similar to
that seen in repair—proficient cells treated in 5. These results strongly
support the hypothesis that human cells preferentially repair UV- and
BPDE-induced lesions from the transcribed strand of the HPRT gene. To test
this, I also measured the rate of repair of BPDE adducts from individual
strands of the HPRT gene, using the UvrABC exinuclease and Southern
hybridizations with strand-specific probes to detect lesions remaining.
BPDE lesions were removed from the transcribed strand at a significantly
faster rate than from the nontranscribed strand, consistent with my
hypothesis. I also found that BPDE adducts were removed faster from either
strand of the HPRT gene than from a transcriptionally inactive locus,
indicating preferential repair of active genes. The results of these
studies provide biochemical and biological evidencetof strand-specific DNA

repair of BPDE adducts in human cells.

This work is dedicated to:

my husband, Yuh-Shan Jou

my mother, Elaine Sha Chen

iv

ACKNOHLEOGNENTS

I would like to express my deep appreciation and gratitude to my
major professor, Dr. Veronica M. Maher, for her never failing support,
advice, and guidance throughout my graduate program. I also would like to
express my deep appreciation to Dr. J. Justin McCormick for his constant
advice and encouragement.

Special thanks go to the other members of my graduate committee,
Drs. Laurie Kaguni, Steven Triezenberg, and Margaret 2. Jones for their
advice and invaluable time.

I would like to acknowledge all the past and present members of the
Carcinogenesis Laboratory for their assistance, encouragement and generous
friendship. In particular, I would like to thank Drs. N. Glenn McGregor
and Lubov Lukash for their collaborations in the study of UV mutagenesis,
and Dr. Jia-Ling Yang for her helpful advice on many technical matters.

Finally, I would like to thank all of those who helped in one way or

another during the course of this work.

TABLE OF CONTENTS

Page
LIST OF TABLES ................................................... ix
LIST OF FIGURES .................................................. xi
ABBREVIATIONS .................................................... xii
INTRODUCTION ..................................................... 1
References ..................................................... 5
CHAPTER I. LITERATURE REVIEW
A. Mechanisms of the DNA repair process responsible
for removal of UV-induced damage and bulky
chemical adducts ........................................ 7
1. Model of nucleotide excision repair ................... 7
2. Current understanding in the nucleotide excision
repair process in mammalian cells ..................... 8
3. Heterogeneity of excision repair in
eukaryotic genome ..................................... 13
3.1 The influence of chromatin organization
on introduction of DNA damage and
efficiency of excision repair ...................... 14
3.2 Preferential repair of actively
transcribed genes .................................. 18
3.3 Effect of transcription on excision repair ......... 22
B. Experimental systems that have been used to
study the mechanisms of mutagenesis ..................... 26
1. Prokaryotic systems .................................. 26
2. Shuttle vector systems in mammalian cells ............ 28
3. Stably integrated exogenous gene system in
mammalian cells ...................................... 31
4. Endogenous mammalian genes system .................... 32
C. The influence of DNA replication and nucleotide
excision repair in mutagenesis .......................... 34
1. General roles of replication and repair
in mutagenesis ....................................... 34
2. Role of heterogeneous repair in mutagenesis .......... 35
REFERENCES .............................................. 39

vi

CHAPTER II. EFFECT OF EXCISIDN REPAIR BY DIPLOID HUMAN
FIBROBLASTS ON THE KINDS AND LOCATIONS OF
MUTATIONS INDUCED BY (i)-75,Ba-DIHYDROXY-
9a,10a-EPOXY-7,8,9,10-TETRAHYDROBENZO[a]PYRENE

IN THE CODING REGION OF THE HPRT GENE ................ 55
Summary ................................................. 56
Introduction ............................................ 58
Materials and methods

Cells and media ...................................... 59

Cell synchronization ................................. 59

Exposure to mutagen and isolation of HPRT mutants.... 60
Synthesis of first-strand cDNA directly

from mRNA in cells ................................. 61
Amplification of HPRT cDNA and DNA sequencing ........ 61
Results ................................................. 63

Cytotoxicity and frequency of mutants
induced by BPDE in cells treated in the

G1 phase or S phase of the cell cycle .............. 63
Characterization of HPRT mutants ..................... 63
Strand distribution of the premutagenic

lesions induced by BPDE ............................ 70
Mutational "hot spot region" obtained from HPRT

mutants in cells exposed in G1 phase .............. 70

Discussion .............................................. 73
Acknowledgments ....................................... 77
References .............................................. 78

CHAPTER III. LACK OF A CELL CYCLE-DEPENDENT STRAND BIAS FDR
MUTATIDNS INDUCED IN THE HPRT GENE BY (i)-7B,8a-
DIHYDRDXY-Qa,10a-EPDXY-7,8,9,10-TETRAHYDRD-
BENZDlalPYRENE IN EXCISIDN REPAIR-DEFICIENT

HUMAN CELLS ......................................... 81
Summary ................................................. 82
Introduction ............................................ 83
Materials and methods

Cells and media ...................................... 85
Cell synchronization ................................. 85
Exposure to mutagen and isolation
of HPRT mutants .................................... 85
Synthesis of first-strand cDNA directly
from mRNA in cells ................................. 86
Amplification of HPRT cDNA and DNA sequencing ........ 86
Results ................................................. 87
Mutant frequency ..................................... 87
Characterization of mutations ........................ 87
Strand distribution of the premutagenic lesions ...... 90
Discussion .............................................. 95
Acknowledgments ......................................... 100
References .............................................. 101

vii

CHAPTER IV. CELL CYCLE-DEPENDENT STRAND BIAS FOR UV-INDUCED

MUTATIONS IN THE TRANSCRIBED STRAND OF EXCISIDN
REPAIR-PROFICIENT HUMAN FIBROBLASTS BUT NOT IN

REPAIR-DEFICIENT CELLS ............................... 104
Summary ................................................. 105
Introduction ............................................ 106
Materials and methods ................................... 110

Cells and media ...................................... 110

Mutagenesis and isolation of HPRT' mutants

(i) Cell synchronization ............................ 110
(ii) Exposure to UV light and
isolation of HPRT“ mutants ..................... 110
Synthesis of first-strand cDNA directly from
mRNA in cells ...................................... 111

Amplification of HPRT cDNA and DNA sequencing ........ 111
Results ................................................. 113

Determining onset of S phase ......................... 113

Survival and mutation frequency ...................... 113

Nucleotide sequence alterations in

xeroderma pigmentosum mutants ...................... 116

Nucleotide sequence alterations in

mutants from repair-proficient cells ............... 118
Discussion .............................................. 128
References .............................................. 133

CHAPTER V. PREFERENTIAL REPAIR AND STRAND-SPECIFIC REPAIR OF

BENZO[A]PYRENE-DIOL EPOXIDE ADDUCTS FROM THE HPRT

OF DIPLOID HUMAN FIBROBLASTS .......................... 136
Summary .................................................... 137
Introduction ............................................... 139
Materials and methods ...................................... 142

Cell culture and cell synchronization ................... 142

BPDE treatment and post-treatment incubation ............ 142

Isolation and purification of DNA ....................... 142

UvrABC excision reaction ................................ 143

DNA denaturation, gel electrophoresis and

Southern blot analysis .................................. 143

Hybridization probes .................................... 144

Quantitation ............................................ 144
Results .................................................... 146

Determining delay in onset of DNA replication

induced by BPDE ......................................... 146

Determining the rate of excision repair of

BPDE adducts ............................................ 146

Evidence of strand-specific repair of BPDE

adducts in human cells .................................. 147

Evidence of preferential repair of BPDE

adducts in human cells .................................. 154
Discussion ................................................. 156
References ................................................. 161

viii

LIST OF TABLES

Table Page
Chapter II

1. Kinds and locations of the mutations induced in the

coding region of the HPRT gene in cells treated with

BPDE in S phase of the cell cycle ............................. 67
2. Kinds and locations of the mutations induced in the

coding region of the HPRT gene in cells treated with

BPDE in G phase of the cell cycle ............................ 68
3. Types of base substitutions observed .......................... 69

Chapter III

1. Kinds and locations of the mutations induced in the coding

region of the HPRT gene in XP128E cells treated with BPDE

in the beginning of the S phase of the cell cycle ............. 88
2. Kinds and locations of the mutations induced in the

coding region of the HPRT gene in XP128E cells treated

with BPDE in early G phase of the cell cycle ................. 89
3. Types of base substitutions observed in the coding

region of the HPRT gene in XP cells treated with BPDE ......... 91
Chapter IV

1. Kinds and locations of mutations induced in the

coding region of the HPRT gene of XP128E cells

irradiated in S phase ......................................... 117
2. Kinds and locations of mutations induced in the

coding region of the HPRT gene of XP12BE cells

irradiated in G phase ........................................ 121
3. Kinds and locations of mutations induced in the

coding region of the HPRT gene of repair-

proficient diploid cells irradiated in S phase ................ 123
4. Kinds and locations of mutations induced in the

coding region of the HPRT gene of repair-

proficient diploid cells irradiated in G phase ............... 126
5. Types of base substitutions induced in the coding

region of the HPRT gene in repair-deficient XP128E

fibroblasts and repair-proficient cells irradiated

in S phase and in G, phase .................................... 127

ix

Chapter V

1. Formation of BPDE adducts in the individual strands

of the 20 kb 8amHI fragment of the HPRT gene and their

rate of removal ............................................... 151
2. Formation of BPDE adducts in the both strands

of the 14 kb EcoRI fragment of the 754 locus and

their rate of removal ......................................... 155

LIST OF FIGURES

Figure

Chapter II

1.

Cytotoxicity and frequency of mutations to
6-thioguanine resistance (TG') induced by BPDE
in human cells treated in early S phase or 12hr

prior to the onset of S phase .............................
. The location of mutations and the strand distribution

of the guanine bases involved in a base change

induced by BPDE in the coding region of the HPRT gene .....

Chapter III

1.

Location of independent HPRT mutations in XP128E
cells treated with BPDE in early S phase or in

early G, phase .............................................

Chapter IV

1.

Incorporation of [3H]thymidine into acid-
insoluble material following a 20-min

pulse-label at the indicated time .........................
. Locations of independent HPRT mutations in XP128E

cells irradiated with UV light (254 nm) at a
fluence of 0.5 J/m in early S phase or 6h before

S phase, in G phase ......................................
. Locations of i

ndependent HPRT mutations in repair-
proficiegt human fibroblasts irradiated with UV light
(6.5 J/m ) in early S phase or in G, phase, 6 to 7h

before S phase ............................................

Chapter V

1.
2.

Molecular organizations of the human HPRT gene

and the 754 locus .........................................

Autoradiograms illustrating extent of repair in the
transcribed (A), nontranscribed (8) strand of the HPRT

gene and in both strands of the 754 locus (C) .............
. Rates of removal of BPDE adducts from the transcribed,

nontranscribed strand of the HPRT gene, and from

both stands of the 754 locus ..............................

xi

Page

64

71

92

.... 114

.... 119

.... 124

148

.... 149

.... 152

AAF
ADA

AFB,
B[c]PHDE

BPDE

BrdUrd
CHO

DHFR

EBV

ENU

HPRT
HSVtk

MNU

MT

N-AAAF
PBS

py-py dimer
6-4 py-py
RT

TetR

TG

ABBREVIATIONS

2-acetylaminofluorene
adenosine deaminase gene
aflatoxin B,
(i)-3a,4B-dihydroxy-la,2a-epoxy—1,2,3,4-tetrahydroxy-
benzo[c]phenanthrene
(i)-7B,8a-dihydroxy-9a,10a-epoxy-7,8,9,10-tetrahydro-
benzo[a]pyrene
bromodeoxyuridine
Chinese hamster ovary cell line
dihydrofolate reductase gene
Epstein-Barr virus
ethylnitrosourea
hypoxanthine (guanine) phosphoribosyltransferase
Herpes simplex virus thymidine kinase gene
methylnitrosourea
metallothionein
N—acetoxy-Z-aminofluorene
phosphate-buffered saline
cyclobutane pyrimidine dimer
6-4 pyrimidine pyrimidone
reverse transcription
tetracycline-resistant gene
6-thioguanine
xii

TGr 6-thioguanine resistant
XP xeroderma pigmentosum

XPRT xanthine-guanine phosphoribosyltransferase

xiii

INTRODUCTION

The relationship of nucleotide excision repair and DNA replication to
mutagenesis in diploid human fibroblasts has been examined by exposing
synchronized cell populations to various carcinogens at different times
across the cell cycle. Such studies showed that the frequency of 6-
thioguanine-resistant (TG') mutants (resulting from the loss of the
functional gene coding for hypoxanthine (guanine) phosphoribosyl-
transferase, HPRT) induced by UV radiation (Maher et al., 1979; Konze-
Thomas et al., 1982) or by (i)-7B,8a-dihydroxy-9a,10a-epoxy-7,8,9,10-
tetrahydrobenzolalpyrene (BPDE) (Yang et'al., 1982; watanabeiet al., 1985)
in repair-proficient cells is highest in cells treated in early S phase
and much lower in cells treated in early G, phase under conditions that
allowed at least 12-15 h. for repair to occur before the DNA replication.
This difference in frequency cannot be explained by the difference in the
physical state of the DNA during the carcinogen treatment, since no such
difference in mutant frequency was found when cells derived from a
xeroderma pigmentosum (XP) patient that are virtually incapable' of
nucleotide excision repair were treated in early S or early G, phase. These
data suggest that nucleotide excision repair prior to the onset of S phase
decreases the mutagenicity of such agents, and that S phase DNA
replication is centrally involved in the conversion of DNA damage into
mutations.

Recently, it has been shown that in cultured rodent and human cells,

2

nucleotide excision repair removes certain lesions, such as cyclobutane
pyrimidine dimers, (py-py dimers) (Bohr et al., 1985; 1986; Madhani et
al., 1986; Mellon et al., 1986), 6-4 pyrimidine-pyrimidone photoproducts
(6-4 py-py lesions) (Thomas et al., 1989), and MNU—induced alkali-labile
lesions (LeDoux et al., 1991), more rapidly from actively transcribed
genes than from bulk DNA, or from nontranscribed genes. Furthermore,
Mellon et al. (1989) showed that in these cells, py-py dimers are removed
from the transcribed strand of an active gene much more rapidly than from
nontranscribed strand. Information on strand-specific excision repair of
BPDE adducts is not yet available. Futhermore, there are no data showing
such repair of UV-induced damage from the human HPRT gene. But if strand
specific repair occurred in that gene in human cells, TGr mutants derived
from the repair-proficient cells treated with BPDE or UV in early S phase
should differ from those taken from cells treated in early G, phase in the
strand distribution of the premutagenic lesions assumed to be responsible
for the mutations. This difference in strand distribution should not occur
in the XP128E cells which are incapable of nucleotide-excision repair.

The extreme complexity of mammalian genome makes it very difficult
to analyze mutations at the DNA sequence level in the endogenous genes of
such cells. However, the recent development of polymerase chain reaction
(PCR) (Saiki et al., 1988) has enabled researchers to selectively amplify
a short segment of DNA more than IDS-fold. This technique, used in
conjunction with reverse transcription (RT) to produce cDNA, provides an
efficient way to determine the sequence alterations in the coding region
of human endogenous genes. Furthermore, Yang et al. (1989) modified this

RT-PCR, so that sequence information can be obtained using a small clone

3
of only a few mutant cells. This makes it possible to study mutation
spectra in endogenous genes of cultured human cells that have a finite
life span.

The objectives of this present work were (1) to investigate the
kinds and locations of mutations induced by BPDE and UV in the HPRT gene
of diploid human fibroblasts treated at the onset of the S phase as well
as in early G, phase; (2) to compare the strand distribution of BPDE- or
UV-induced premutagenic lesions observed from cells treated at S phase
with those from cells treated in G, phase, in order to determine if there
were biological evidenceiof strand-specific repair of UV photoproducts and
BPDE adducts in the human HPRT gene; (3) to determine directly the repair
rates of BPDE adducts from the two DNA strands of the human HPRT gene
using UvrABC exinuclease and Southern blot analysis. The latter study was
undertaken to investigate whether there was biochemical evidence of
strand-specific repair of BPDE adducts.

Chapter I reviews the background literature bearing on these
questions. Chapter 11 consists of a manuscript published in the November
1990 issue of Proceedings of the National Acadely of Science U. S. A., 87,
8680-8684. It describes the research I carried out to determine the kinds
and locations (spectrum) of mutations induced by BPDE in the coding region
of the HPRT gene of repair-proficient human fibroblasts treated in the S
phase or in 6, phase of the cell cycle, in order to investigate the effect
of excision repair on the spectra of such mutations. Chapter III, which
consists of a manuscript published in the May 1991 issue of Cancer
Research, 51, 2587-2592, describes comparable work I carried out with
repair-deficient XP128E cells treated with BPDE in S or 6, phase and

4
compares the results obtained from this study with those I found in the
repair-proficient human cells. Chapter IV is that a manuscript published
in the April 1991 issue of Molecular and Cellular Biology, 11, 1927-1934,
which has been a collaborative work with my colleague Dr. N. Glenn
McGregor. It describes a study of the UV-induced HPRT mutational spectra
in repair-proficient cells and in XP128E cells, as well as the
investigation of the role of excision repair in such mutagenesis. In this
study, the mutagenesis experiments and mutational analysis in repair-
proficient cells were carried out by Dr. McGregor. The comparable study on
the XP128E cells was carried out by me (80% of the work) and by Dr. Lukash
(20% of the work). Chapter V describes the biochemical study on the
comparison of the repair rates of the BPDE adducts on the two strands of
the HPRT gene of human fibroblasts. This chapter is written in the style
of a manuscript for the Proceedings of the National Acadelw of Science U.
S. A. and will be submitted for publication to this journal.

REFERENCES

Bohr, V. A., Dkumoto, D. S., Ho, L., and Hanawalt, P. C. 1986.
Characterization of a DNA repair domain containing the dihydrofolate
reductase gene in Chinese hamster ovary cells. J. Biol. Chem., 261: 16666-
16672.

Bohr, V. A., Smith, C. A., Dkumoto, D. S., and Hanawalt, P. C. 1985. DNA
repair in an active gene: Removal of pyrimidine dimers from the DHFR gene
of CHO cells is much more efficient than in the genome overall. Cell,
40:359-369.

Konze-Thomas, 8., Hazard, R. M., Maher, V. M., and McCormick, J. J. 1982.
Extent of excision repair before DNA synthesis determines the mutagenic
but not the lethal effect of UV radiation. Mutat. Res., 94: 421-434.

LeDoux, S. P., Thangada, M., Bohr, V. A., and Wilson, G. L. 1991.
Heterogeneous repair of methylnitrosourea-induced alkali-labile sites in
different DNA sequences. Cancer Research 51: 775-779.

Madhani, H. D., Bohr, V. A., and Hanawalt, P. C. 1986. Differential DNA
repair in transcriptionally'active'and inactive proto-oncogenes: c-abl and
c-mos. Cell 40: 359-369.

Maher, V. M., Dorney, D. J., Mendrala, A. L., Konze-Thomas, 8., and
McCormick, J. J. 1979. DNA excision repair processes in human cells can
eliminate the cytotoxic and mutagenic consequences of Ultraviolet
irradiation. Mutat. Res., 62: 311-313.

Mellon, 1., Bohr; V. A., Smith, C. .A., and Hanawalt, P. C. 1986.
Preferential DNA repair of an active gene in human cells. Proc. Natl.
Acad. Sci. USA, 83: 8878-8882.

Mellon, I. Spivak, G., and Hanawalt, P. C. 1987. Selective removal of
transcription-blocking DNA damage from the transcribed strand of the
mammalian DHFR gene. Cell, 51: 241-249.

Thomas, D. C., Dkumoto, D. S., Sancar, A., and Bohr, V. A. 1989.
Preferential DNA repair of (6-4) photoproducts in the dihydrofolate
reductase gene of Chinese hamster ovary cells. J. Biol. Chem. 25: 18005-
18010.

Natanabe, M., Maher, V. M., and McCormick, J. J. 1985. Excision repair of
UV- or benzo(a)pyrene diol epoxide-induced lesions in xeroderma
pigmentosum variant cells is "error free". Mutat. Res., 146:285-294.

Yang, L. L., Maher, V. M., and McCormick, J. J. 1982. Relationship between
excision repair and the cytotoxic and mutagenic effect of the "anti" 7,8-
diol-9,10-epoxide of benzo(a)pyrene in human cells. Mutat. Res., 94:435-
447.

6

Yang J.-L., Maher, V. M., and McCormick, J. J. 1989. Amplification and
direct nucleotide sequencing of cDNA from the lysate of low numbers of
diploid human cells. Gene, 83: 347-354.

Saiki, R. K., Gelfand, D. H., Stoffel, S., Scharf, S. J., Higuchi, R.,
Horn, G. T., Mullins, K. 8., and Erich, H. A. 1988. Primer-directed
enzymatic amplification of DNA. with a thermostable DNA polymerase.
Science, 239:487-491.

CHAPTER I

LITERATURE REVIEH

A. Mechanisas of the DNA repair process responsible for renoval of UV-
induced damage and bulky chelical adducts
1. Model of nucleotide excision repair

The DNA repair processes are the cellular responses associated with
the restoration of the normal nucleotide sequence of DNA after damage.
The enzymatic pathways involved in DNA repair vary with the types of
damage introduced, but the most general mechanism is the nucleotide
excision repair pathway, which is responsible for removing UV-induced
cyclobutane pyrimidine dimers (py-py dimers) and 6-4 pyrimidine-pyrimidone
photoproducts (6-4 py-py), as well as adducts produced by benzo(a)pyrene-
diol-epoxide (BPDE), mitomycin C, psoralen, acetylaminofluorene (AAF),
cisplatin, 4-nitroquinoline-l-oxide, and many other carcinogens and/or
chemotherapeutic agents.

During the mid-19605, a series of important biological experiments
were independently reported by Setlow et al. (1964) and by Boyce et al.
(1964) that led to major insights into the nature of nucleotide excision
repair in E. coli. They showed that when bacterial cells were UV-
irradiated and then incubated, thymidine-containing pyrimidine dimers were
lost from the acid-insoluble fraction of the DNA and could be recovered in

the acid-soluble phase. These observations, together with the

8

demonstration of repair synthesis of DNA at about the same time (Pettijohn
& Hanawalt, 1963; 1964) led to a general molecular model for the multi-
step process of nucleotide excision repair. The sequential steps include
(a) preincision recognition of damage; (b) incision of the damaged DNA
strand at or near the site of the defect; (c) excision of a fragment
containing the damaged site or exonucleolytic degradation of the affected
strand; (d) repair replication to replace the excised region with a
corresponding stretch of normal nucleotides; and finally (e) ligation to
join the repair patch at its 3’ end to the contiguous parental DNA strand.

For many years, researchers who studied excision repair in bacteria
had believed that following the damage-specific DNA incising activities,
the actual excision of the damaged nucleotides was performed by an
exonucleolytic degradation of the DNA in the 5’-> 3’ direction (Hanawalt
et al. 1979; Lindahl, 1982). However, it has been demonstrated that E.
coli contains a damage-specific DNA incising enzyme complex, UvrABC, which
cuts the damaged strand both upstream and downstream of the damaged bases
(Sancar & Rupp, 1983; Yeung et al., 1983). This dual incision facilitates
the release of a 12 or 13 base oligonucleotide containing the lesion
(Caron & Grossman, 1988; Kumura et al., 1985; Myles et al., 1987; Sancar
& Rupp, 1983). This observation has led to the suggestion that this type

of enzyme be designated an exinuclease.

2. Current understanding in the nucleotide excision repair process in
Ia-alian cells.
The study of excision repair in mammalian cells was limited for many

years by the lack of available mutant cells defective in their response to

9

DNA damage, until the observation that fibroblasts in culture derived from
the skin of a human patient with the disease xeroderma pigmentosum (XP)
were defective in UV-induced excision repair (Cleaver, 1968). XP is an
autosomal recessive disease characterized clinically by a severe
photosensitivity of the eyes and exposed regions of the skin, a very high
incidence’of skin cancers and frequent neurological abnormalities (Reed et
al., 1969; Robbins et al., 1974; Cleaver & Bootsma, 1975; Robbins, 1978).
Fusions between cells of different patients with this disease in various
pairwise-combinations have established the existence of at least eight
genetic complementation groups, designated A through H (Friedberg et al.,
1979; Cleaver, 1983; Moshell et al., 1983). Members of each of them are
defective or deficient in the excision repair of UV-induced photoproducts
as well as various types of chemical damage to DNA (Setlow, 1978;
Friedberg et al., 1979). It appears that much of the genetic complexity in
XP is concentrated at very early steps in excision repair, specifically
those involved in the incision of damaged DNA (Friedberg et al., 1979).
Besides the eight complementation groups, approximately 10% of all cases
of clinically typical XP do not have an apparent defect in the excision
repair of DNA damage caused by UV radiation. These cases are referred to
XP variants (Robbins & Kraemer, 1972; Cleaver, 1972; Cleaver et al.,
1981).

Excision repair-deficient mutants of rodent cells have’been isolated
from a large population of mutagenized cells in culture (Busch et al.,
1980; Thompson et al., 1980; 1981; Schiomi et al., 1982). Complementation
groups have also been defined for these rodent mutants (the ERCC1-8

series) (Busch et al., 1989; Zdzienicka et al., 1988; Thompson et al.,

10

1988a). These mutants are similar in many respects to the human XP
mutants. However, they differ quantitatively in the degree of deficiency
of excision repair. Most XP groups have low levels of photoproduct
excision, whereas rodent mutants tend to be either completely deficient or
intermediate in repair (Clarkson et al., 1983). The rodent mutants are
also distinctive in being sensitive to cross-linking agents, such as
mitomycin C (Hoy et al., 1985), which is not the characteristic of XP
cells (Fujiwara et al., 1977). Several types of complementation tests have
been performed to establish a link between the members of these two
species, but until now no overlap has been found (Stefanini et al., 1985;
Thompson et al., 1985a). This implies the involvement of a substantial
number of genes in excision repair in mammalian cells and a considerable
complexity of this process at the biochemical level.

Why are so many genes apparently required for the excision repair in
mammalian cells? At least two hypothetical possibilities have been raised.
First, one might imagine that the enzymology of DNA excision repair in
mammalian cells is strictly different from that in E. coli; i.e., a
smaller number of genes is required in bacteria. Alternatively, some of
these mammalian gene products may function specifically to create
accessibility of sites of damage in chromosomes to a catalytically active
endonuclease, since in eukaryotes, DNA in the form of chromatin is
extremely tightly packaged and folded in the nucleus. Several lines of
evidence that support the second possibility'will be described in the next
section.

Many of the details of excision repair in mammalian cells have yet

to be characterized. One approach to the study of mammalian genes required

11

for the excision repair is to attempt the molecular cloning of these
genes, with a view to expressing the proteins they encode in amounts
suitable for direct biochemical study. The rodent mutants have proven to
be more valuable as recipients for gene transfer than XP cell lines
(Hoeijmakers et al., 1987). Cell fusion and DNA-mediated gene transfer
have demonstrated that at least six (ERCC1-6) of these rodent
complementation groups can be corrected by genes from normal human cells
(Thompson et (al., 1985b). Recently, the isolation of several human
excision repair genes, ERCC-l, ERCC-2, ERCC-3, and ERCC-6 that are able to
correct the defect in rodent complementation groups 1, 2, 3, and 6,
respectively, has been reported (Nesterveld et al., 1984; Weber et al.,
1988; Needa et al., 1990a; Troelstra et al., 1990).

The ERCC-l gene was found to be the human equivalent of the S}
cerevisiaee RADIO repair gene (van Duin et al., 1986). In addition, this
gene harbors two domains present in E. coli excision repair gene UvrA and
UvrC (van Duin et al., 1988). However, it fails to correct the defects in
any of the XP complementation groups (van Duin et al., 1989). Since the
rodent mutants were derived in vitro, the absence of certain rodent
complementation groups among XP groups may indicate that some mutations
that remain viable in culture compromise development and viability in the
whole organism (Cleaver, 1990).

The ERCC-Z gene product exhibits strong sequence similarity at the
amino acid level to RA03 of yeast (Weber et al., 1988; 1990). RA03 is
known to have ATPase (Sung et al., 1987a) and helicase activity (Sung et
al., 1987b), suggesting a similar function for ERCC-2 (Weber et al.,
1990). Mutational analysis has shown that the ERCC-Z protein is required

12
for cell viability as well as repair (Naumovshi & Friedberg, 1983; 1986),
and it is also involved in recombination (Montelone et al., 1988). Weber
et al. (1991) recently demonstrated that this gene complements the defect
in XP-D cells.

The ERCC-3 protein harbors putative nucleotide, chromatin, and
helix-turn-helix DNA binding domains and seven consecutive motifs
conserved between two superfamilies of DNA and RNA helicases, suggesting
that it is a DNA repair helicase (Needa et al., 1990b). The gene was found
to correct the excision defect in XP-B (Needa et al., 1990b), which
exhibits a complex phenotype having features of both XP and Cockayne’s
syndrome (Lehmann, 1987).

The ERCC-6 gene has been isolated (Troelstra et al., 1990) by
correcting the excision defect in a chinese hamster cell line, UV61, which
was found to harbor a deficiency in the repair of UV-induced py-py dimers
but permit apparently normal repair of 6-4 py-py lesions (Thompson et al.,
1988b). The open reading frame of this gene encodes a predicted protein of
minimally 1500 amino acids with a putative chromatin binding domain and a
serine phosphorylation site. In addition, it shows an almost perfect match
with 7 domains conserved between the two superfamilies of helicase, which
makes ERCC-6 the third putative helicase involved in excision repair
(Troelstra et al., 1991).

Attempts to confer UV-resistance on XP cells by transfection with
genomic DNA from normal cells had been unsuccessful, due to the
spontaneous revertants appearing in the recipient cells as well as the
difficulty in stably transfecting high-molecular-weight DNA into human

cells (Lehmann, 1985). However, the mouse gene correcting SV40-transformed

13

XP-A fibroblasts has been cloned by extensive genomic DNA transfection
experiments (Tanaka et al., 1989). Shortly thereafter, a human homolog of
this XP-A-correcting cDNA has been isolated (Tanaka et al., 1990). This
cDNA encodes a hydrophilic protein which contains a distinct zinc-finger
motif, indicating that it interacts directly with DNA. The corresponding
mRNA is greatly reduced in cell lines from several XP-A patients, due to
a G->C transversion at a splice-acceptor site (Satokata et al., 1990).

Attempts to clone other XP genes are now in progress. A mouse DNA
repair gene that partially corrects the defect in the XP-D cell line has
been reported (Arrand et al., 1989). In addition, an apparent loss of a
separate protein that binds to damaged DNA has been reported for the XP-E
cells (Chu & Chang, 1988). This protein has many DNA-binding
characteristics that resemble those of a yeast photolyase. However, it
lacks the function of light-mediated photolysis (Patterson & Chu, 1989).

A soluble cell-free system that can carry out nucleotide excision
repair in vitro has been developed (Wood et al., 1988). This system
allowed the identification of the requirement for the human single-
stranded DNA binding protein in excision repair (Coverley et al. 1991). In
addition, it provides the method of functional assessments of purified XP

gene products by biochemical complementation.

3. Heterogeneity of excision repair in the eukaryotic genole

Eukaryotic DNA is folded in a hierarchy of different structures to
form chromatin. Most of the genome in a cell of a complex organism
contains information unnecessary or inappropriate to its own specialized

function, and is not expressed. It is possible that the requirement for

14
repair systems in such regions is very different from that in active genes
where mutations or blockage of transcription could have direct effects on
cell survival and functions. In addition, the various states of chromatin
may modulate the introduction of certain types of damage into DNA and, as
the proper substrate for the repair system, different chromatin states may
play an important role in the kinetics or efficiency of DNA repair. To
approach the issue of how repair systems might be influenced by different
functional and structural aspects of the mammalian genome, considerable
effort has been devoted in the past few years to investigating repair in
particular genomic components. One approach is to use traditional methods
to analyze repair or adduct frequency in DNA fractionated by procedures
thought to separate it by structure or function, e.g. separating active
and inactive chromatin fractions according to the nuclease accessibility
(Mullenders et al., 1987), or isolating highly repetitive DNA species
(Madhani et al., 1986a). Another approach is to study repair in specific
DNA sequences, such as genes and their flanking regions, whose activity or
chromatin organization can be analyzed and, in some cases, experimentally
manipulated (Bohr, 1987; Hanawalt et al., 1989). With these approaches, a
number of interesting and notable results indicating that excision repair

is not uniform throughout genome have been reported.

3.1 The influence of chrolatin organization on introduction of DNA dalage
and efficiency of excision repair

It is well recognized that DNA in mammalian cells exists in the form
of chromatin which is organized and tightly packaged into higher order
structure. In the first step of DNA packaging, the DNA is bound by histone

15

and non-histone proteins to form the repeating core and linker regions of
the nucleosome structures (Moyne et al., 1981). This structure is further
coiled into a "thick fiber" with 6-9 nucleosomes per turn (Thoma et al.,
1980; Thomas & Butler, 1980). In the interphase nuclei and in metaphase
chromosomes the thick fiber appears to be folded into loops (Cook &
Brazell, 1975; Benyajati & Worcell, 1976) or domains (Paulson & Laemmli,
1977), and these repeating loops are constrained by attachment to a
nuclear skeletal structure termed the nuclear matrix, or scaffold (Paulaon
& Laemmli, 1977; Igo-kemenes & Zachau, 1977). There is increasing evidence
that the nuclear matrix is associated with fundamental processes such as
replication, transcription and repair (Nelson et al., 1986).

Several studies have been done to analyze DNA damage and repair
processes at the level of nucleosomes in chromatin structure using
micrococcal or staphylococcal nuclease, which preferentially' digests
linker regions of chromatin. With this approach, it has been shown that a
number of bulky adducts are formed preferentially at linker regions,
including those induced by the following agents: AAF (Baranyi-Furlong &
Goodman, 1984), N-hydroxy-AAF (Metzger et al., 1977), N-acetoxy-AAF (N-
AAAF) (Kaneko & Cerutti, 1980), dimethyl-N-nitrosamine (Ramanathan et al.,
1976), methylnitrosourea (MNU) (Berkowitz et al., 1981), BPDE (Jahn &
Litman, 1979; Arrand & Murray, 1982), and aflatoxin 8, (Bailey et al.,
1980). Studies have also been performed on the distribution of py-py
dimers between nucleosome core DNA and linker DNA (Williams & Friedberg,
1979; Snapka & Linn, 1981; Niggli & Cerutti, 1982). The results of these
studies indicated that there is essentially no difference between.these

two regions of the nucleosome unit in the distribution of py-py dimers at

16

low to moderate UV doses. However, at a finer level of resolution using
the 3’->5’ exonuclease activity of T4 DNA polymerase, Gale et al. (1987)
showed that the distribution of UV-induced photoproducts (primarily py-py
dimers) is not uniform throughout core DNA, but displays a striking 10.3
(t 0.1) base periodicity. The sites of maximum dimer formation in the core
DNA mapped to positions where the phosphate backbone is farthest from the
core histone surface. This pattern of dimer induction was not obtained
from UV-irradiated core DNA either free in solution or bound tightly to
calcium phosphate crystals, indicating that yield of dimers along the DNA
may be modulated in a Tanner that reflects structural features of the
nucleosome unit.

Evidence of the repair heterogeneity at the level of the nucleosome
has also been reported. In response to UV or chemicals that inflict damage
randomly in chromatin, most of the repair synthesis in mammalian cells
occurs in the staphylococcal nuclease-sensitive regions, but with
increasing repair time the nucleotides inserted during repair synthesis
become progressively more nuclease resistant (Smerdon & Lieberman, 1978;
Lieberman et al., 1979). In addition, it was determined that the 5’ and 3’
end regions of core DNA are markedly enhanced in repair-incorporated
nucleotides (Lan 8: Smerdon, 1985). Recently, Smerdon 8. Thoma (1990)
examined excision repair of UV photoproducts in a small yeast plasmid
(YRp-TRURAP), known to be folded into chromatin with positioned
nucleosomes in viva. Their results showed a correlation between the repair
rate and the relative stability of nucleosomes on the DNA, with a more
stable nucleosome exhibiting less accessibility to the repair enzymes.

Excision repair of a number of adducts has been examined in the

17

heterochromatic alpha DNA which can be physically isolated for analysis.
It was shown that UV damage is repaired with similar efficiency in
repetitive alpha sequences and in bulk DNA (Lieberman & Poirier, 1974),
but that the repair of certain bulky adducts was markedly deficient in
alpha DNA (Zolan et al., 1982; Leadon at al., 1983; Zolan et al., 1984;
Smith, 1987). These studies suggest that the highly condensed chromatin
structure of alpha DNA hinders access of the repair system that acts on
bulky adducts.

The repair of UV-induced damage in human and rodent cells was
investigated at the level of DNA loops attached to the nuclear matrix.
These studies have demonstrated preferential repair in such matrix-
associated DNA as compared to total nuclear DNA, particularly at low
levels of UV exposure (McCready & Cook, 1984; Harless 6 Hewitt, 1987;
Mullenders et al., 1988). This feature of preferential repair has been
examined in some repair-deficient cells. The repair synthesis in XP-C
cells, which carry out most of their residual repair activity in a limited
portion of the genome, appears to be more closely associated with the
nuclear matrix than that in normal cells (Mullenders et al., 1984).
However, in contrast to the observations for XP-C and normal cells, the
repair synthesis in Cockayne’s syndrome fibroblasts is decreased in DNA
associated with the matrix (Mullenders et al., 1988). The relationship
between such heterogeneity in distribution of repair events in chromatin
loops and the preferential repair of transcriptionally active genes will

be discussed later.

18
3.2 Preferential repair of actively transcribed genes.

Repair of py-py dimers has been studied in individual genes and
specific sequences using T4 endonuclease V which cleaves DNA at dimer
sites (Bohr et al., 1985; 1986a; 1988). 8y Southern hybridization and
probing for a sequence of interest, the fraction of molecules without
dimers (zero class) can be quantitated. This value can be used with the
Poisson equation to calculate the actual frequency of dimers in the
fragment, and thereby measure its repair.

With this approach, removal of dimers from a number of genes in
cultured mammalian cells has been reported. It was initially found that
repair of dimers in the essential gene for dihydrofolate reductase (dhfr)
in Chinese hamster ovary (CHO) cells was much more efficient than the
repair in the overall genome (Bohr et al., 1985). This may provide an
explanation for the paradox that rodent cells are deficient in overall
genome DNA repair, but as resistant to UV damage as the repair-proficient
human cells; it is possible that they survive by selectively repairing the
vital sequences. It has been suggested that the determination of repair in
essential genomic regions is an important feature if one is trying to
correlate DNA repair and cell survival (Bohr et al., 1986a).

The fine structure~of the preferentially repaired dhfr region in CHO
cells has been characterized (Bohr' et al., 1986b). Repair is most
efficient in the 5’ portion of the DHFR gene and in the 5’ flanking
sequence, while'a distant upstream sequence>and the 3’ flanking downstream
sequence were each repaired with an efficiency similar to that for the
overall genome. Preferential repair of an active gene has also been found

in human cells. Mellon et al. (1986) showed that dimers in the DHFR gene

19
in a human cell line, 6A3, were removed much more rapidly than those in
the bulk DNA.

Besides the DHFR locus, preferential repair of transcriptionally
active loci has been confirmed in several other genes. In a 21-Kb fragment
encompassing most of the 25-Kb gene encoding 3-hydroxy-3-methylglutaryl
coenzyme A reductase in CHO cells, dimers are removed to about the same
extent as in dhfr (Smith, 1987). In mouse 3T3 cells, Madhani et al.
(1986b) observed efficient repair of the transcriptionally active c-abl
gene in both growing and resting cells. However, very little repair was
observed in 15- and 6-Kb fragments that contained portions of the mouse c-
mos proto-oncogene, the transcription of which has not been detected in
these cultured cells. In human fibroblasts, the preferential repair of the
32-Kb adenosine deaminase (ADA) gene was reported by Mayne et al. (1988).
As a control, they analyzed repair in a 14-Kb fragment that lies in an
unexpressed region of the X chromosome, termed the 754 locus. This
fragment was repaired slowly, with the same kinetics as the overall
genome.

Taken together, these results suggest that the preferential repair
of py-py dimers from active genes is a general phenomenon in mammalian
cells. Several possibilities have been suggested to explain this feature
of repair. The chromatin structure in active genes could have a more
"open" configuration to be easily accessible to repair complex, or
fundamentally different repair systems with different efficiencies could
operate on active or inactive DNA. Finally, the blockage of transcription
by a lesion might make it a more attractive substrate to repair enzymes

(Smith and Mellon, 1990).

20

Using the same approach, removal of dimers from transcriptionally
active and inactive genomic DNA has been examined in repair-deficient
human cells. In spite of the limited overall repair capacity in XP-C cells
(Mansbridge>& Hanawalt, 1983), py-py dimer removal from ADA and DHFR genes
was found to be as efficient as that in normal human cells. However,
although normal cells performed efficient repair from the 754 locus but at
a slower rate than found for the active ADA and DHFR genes, the XP-C cells
showed a very poor repair from the 754 locus (Venema et al., 1990a). In
contrast to XP-C cells, Cockayne’s syndrome cells are able to remove
dimers from DNA but they are unable to repair the DHFR and ADA genes as
rapidly and efficiently as normal cells (Venema et al., 1990b). These
results correlate with the preferential removal of dimers from nuclear
matrix associated DNA. in normal and XP-C cells and its absence in
Cockayne’s syndrome cells. It was suggested that Cockayne’s syndrome cells
have a defect in excision of dimers from active genes located proximal to
the nuclear matrix, whereas XP-C cells lack the ability to render the rest
of the chromatin accessible to repair enzymes (Mullenders et al., 1988).

Little information has been available on the possible gene- or
sequence-specific repair of photoproducts other than py-py dimers
discussed above, or of DNA damage induced by chemical agents. More
recently, techniques have been developed to study other lesions, for
example by the use of the UvrABC excision nuclease to cleave the DNA
strand specifically' at damaged sites (Thomas et al., 1988), or by
formation of strand breaks at sites of alkylation damage (Scicchitano &
Hanawalt, 1989).

The E. coli enzyme complex UvrABC recognizes most DNA adducts that

21

create significant helical distortions (Sancar & Sancar, 1988). Thomas et
al. (1989) measured repair of 6-4 py-py lesions in the dhfr gene of CHO
cells using this enzyme complex in conjunction with another DNA repair
enzyme from E. coli, DNA photolyase. The latter enzyme when bound to a py-
py dimer in DNA and exposed to light causes the loss of the dimer
affecting repair (Brash et al., 1985; Myles et al., 1987). The remaining
photoproducts, which are almost exclusively 6-4 py-py lesions, are then
digested with UvrABC. With Southern blot analysis and probing for specific
sequences, the repair of 6-4 py-py lesions can be quantitated. It was
found that these lesions are repaired more efficiently than py-py dimers
and that there is preferential repair of these lesions in the dhfr coding
region in CHO cells compared with its downstream, noncoding sequences.

With UvrABC treatment, Tang et al., (1989) examined the formation
and repair of N-AAAF induced DNA damage in the DHFR coding and noncoding
sequences. In contrast to that obtained for the repair of py-py dimers,
the rates of removal of N-AAAF-induced damage from both sequences are
similar in cells treated with 10uM N-AAAF. However, it is perhaps worth
noting that the initial lesion frequency (3.1 adducts/14 kb) used in this
study was considerably higher than in most experiments with UV; this might
lessen the apparent differences in rate of repair between transcribed and
nontranscribed sequences.

Evidence on heterogeneous repair of MNU-induced alkali-labile
adducts has recently been provided (LeDoux et al., 1991). MNU
spontaneously decomposes to form a carbonium ion which is capable of
alkylating the nitrogens and oxygens of DNA bases. The predominant adducts

are Nr-methyl-guanine and N3-methyl-adenine which are removed by excision

22
repair (Montesano, 1981). The alkali lability of these two adducts was
exploited to measure their formation and removal in the dhfr and c-fos
loci in CHO cells. Rather than treat the isolated DNA with an
endonuclease, they heated it to produce apurinic sites and then incubated
it in alkali to cleave DNA at these sites. The remaining procedure used
was the same as that developed for assaying the removal of py-py dimers.
It was found that, although repair in the 3’-flanking sequence of dhfr
gene was only slightly less efficient than that seen in its coding region,
repair in the nontranscribed c-fos-containing fragment was markedly
attenuated. This suggests that repair across the entire genome is
heterogenous with different areas repairing at varying rates. However, the
exact mechanisms governing the rate of repair in specific areas remain to

be elucidated.

3.3 Effect of transcription on excision repair

The selective repair of transcriptionally active genes raises some
interesting questions about the relationship between DNA repair and
transcription. Mellon et al. (1987) found a dramatic difference in the
efficiency of removal of dimers from the transcribed and nontranscribed
strands of the DHFR gene in cultured hamster and human cells, with the
transcribed strand repaired preferentially. This finding challenges the
simple model of chromatin accessibility as the sole mediator of
preferential repair, and further strengthens the possibility of the
association between transcription and excision repair.

The relationship between repair and transcription has been studied

in more detail using the approach allowing examination of the same gene

23
under different conditions. Gene activity can be modulated in vitro with
inhibitors or natural effectors, such as Tgene specific inducers or
suppressors. In such cases, it is usually considered that the rate of
transcription of the gene is affected directly by trans-acting molecules,
and that the chromatin structure of the gene is in some permissive state
even when its transcription activity is low.

A number of mammalian genes are inducible after activation by
external inducers and one of the most studied is metallothioneins (MT).
Mammalian genomes encode two closely related metallothionein, MT I and MT
11, which are transiently induced by exposure to heavy metals such as
cadmium, zinc, and copper, and by various circulating factors including
hormones and interferon (Brady, 1982; Kojima & Kaegi, 1978). Repair of
dimers from the MT I gene in a CHO cell line has been examined
comparatively with or without transcriptional activation of the gene by
incubation in the presence of ZnCl2 (Dkumoto & Bohr, 1987). Whereas the
repair efficiency was very low in the uninduced state, it increased about
50% after induction of the gene. Furthermore, no effect of ZnCl2 was
observed for repair in a noninducible gene, dhfr, or on overall repair
synthesis in the cells.

Leadon and Snowden (1988) studied the repair of UV-induced damage in
several members of the large MT gene family in human cells. They used an
immunological method that allows isolation of DNA fragments containing
bromodeoxyuridine (BrdUrd) in repair patches using a monoclonal antibody
that recognizes BrdUrd (Leadon, 1988). Specific 52P-labeled hybridization
probes were used to detect and quantitate the sequences of interest in the

repaired fraction of the DNA. Unlike methods that rely on the loss of an

24

endonuclease-sensitive site from a sequence, this technique allows
comparative estimates of repair, but not actual lesion frequencies. It
also provides a sensitive measurement of repair, since only one repair
event will place an entire fragment into the "repaired" class, regardless
of the lesion frequency on the fragment (Leadon, 1986). With this method,
they showed that repair in fragments carrying the unexpressed MT-IBTand MT-
IIB genes resembled repair in genome overall. About twice as much repair
was found in the MT-IA, MT-IIA, and MT-IE genes which are expressed at a
basal level. Furthermore, inducing transcription by cadmium or steroids
increased repair about another twofold for the three expressed genes which
are responsive to the inducers. The rates of repair in the unexpressed
genes were not altered by these treatment. These results are consistent
with the notion that the repair of these MT genes is correlated with their
rates of transcription.

The findings of strand-specific repair and preferential repair of
induced genes support a direct relationship between nucleotide excision
repair and transcription. The recent demonstration that strand-specific
repair occurs in the E. coli lactose operon (Mellon & Hanawalt, 1989)
further indicates that this feature of transcription-dependent repair may
be a general phenomenon, unrelated to eukaryotic chromatin structure. At
least three models have been suggested to explain the effects of
transcription on DNA repair: (1) the RNA polymerase antenna model, (ii)
the twin supercoiling domain model, and (iii) the RNA-DNA heteroduplex
model (van Houten, 1990).

RNA polymerase antenna model: There is strong evidence that

transcription is blocked by py-py dimers in template DNA in both

25

prokaryotic and eukaryotic cells (Sauerbier & Hercules, 1978). It seems
reasonable to suppose that only damage on the template strand produces
this effect, whereas lesions on the other strand do not. It has been
suggested that this lesion-stalled RNA polymerase acts as a damage antenna
for the repair complex. The configuration of a lesion at the site of
blockage may resemble the one brought about by preincision activities,
thereby bypassing the need for some preincision processing. For example,
DNA unwinding at the site of transcription bubble might circumvent the
need for a repair helicase(s). The other possibility is that some specific
factors in the transcription complex mediate the formation of preincision
complex via direct protein—protein interactions.

Twin supercoiling domain model: It has been shown that passage of
RNA polymerase through a gene leads to two waves of superhelical tension
in the DNA: a positive supercoil wave in front of the transcription
complex and a negative supercoil wave following behind the RNA polymerase
(Tsoa et al., 1989; Wu et al., 1988). It is possible that these localized
supercoiled waves could influence the repair of DNA lesions, although
there is as yet no evidence of such an effect.

RNA-DNA heteroduplex model: Since the conformation of a DNA-RNA
hybrid molecule in the transcription bubble is more A form than a DNA-DNA
duplex, it is possible that a DNA lesion in close proximity to this RNA-
DNA junction would alter the conformation. This might facilitate the

recognition by repair enzymes.

26
B. Experilental systels that have been used to study the lechanisls of
outagenesis

1. Prokaryotic systels

A general strategy for studying mutagenic mechanisms is to determine
the mutational specificity, i.e., the precise nature and locations of
mutations produced by mutagens. The well-developed techniques of bacterial
molecular genetics provide a simple way of obtaining precise mutational
information on prokaryotic systems. Due to the simple organization of
prokaryotic cells, extensive genetic studies have been done using
prokaryotic genes. One of these is the E. coli lacI gene in which a large
number of nonsense sites have been characterized and correlated with
specific codons (Coulondre & Miller, 1977). This system has been widely
used to investigate both spontaneous and induced mutations in bacteria
(Calos & Miller, 1981; Miller, 1982). By comparing the profile of
occurrences of nonsense mutations induced by different mutagens, the
specificity of each mutagen has been determined.

In the 1970’s, most of the methods for analysis of the specificity
of mutagens were based on the production or reversion of nonsense
mutations (Yanofsky, 1971; Prakash & Sherman, 1973; Ames et al., 1973;
Lawrence & Christensen, 1979), and thus were biased to specific sites in
the DNA. Only by analysis of a large number of such sites can one overcome
this limitation (Coulondre & Miller, 1977). To eliminate the need for
mutating specific sites in the DNA, a directed mutagenesis method has been
developed for the analysis of thelmutational specificity of mutagens. With

this approach, Fuchs et al. (1981) analyzed forward mutations induced in

27

the tetracycline-resistance gene (TetR) of plasmid p8R322 by directing the
chemical reaction of carcinogen N-acetoxy-AAF to a restriction fragment
inside in this target gene. The AAF-modified restriction fragment was
reinserted by in vitro ligation into the non-reacted p8R322 restriction
fragment, and then the ligation mixture was used to transform E. coli.
Mutants were selected for ampicillin resistance and tetracycline
sensitivity. The plasmid DNA of such mutants was analyzed for sequence
changes in the fragment where the AAF binding had been directed. Their
results showed that most (90%) of the mutations induced by AAF adducts
were frameshifts and were located within specific sequences, so called
"hot spots" (Fuchs et al., 1981; Koffel-schwartz et al., 1984). This
method is of general applicability to any mutagen or carcinogen that
reacts with DNA. Livneh (1983) used a similar approach to examine UV-
induced mutations in the TetR gene of plasmid pr3. The data revealed that
this systemiallowed the detection of various types of mutations, including
transitions, transversions, frameshifts and deletions.

In addition to plasmid vectors, phage vectors were also used for
analysis of mutational specificity. LeClerc & Isotock (1982) used M13lac
hybrid phage, which contains the M13 single-stranded DNA with an insert of
the regulatory region and the a-peptide of the lad gene of E. coli
(Messing et al., 1977), as a mutation system to determine the UV-induced,
forward mutations. Infection of E. coli JM103lacZ M15 cells with this
hybrid phage allows the production of a functional B-galactosidase by
intracistronic complementation. However, mutations in the lac insert of
phage DNA that inactivate the a-complementing activity give rise to

colorless plagues of infected cells, easily distinguishable from the blue

28
plagues of M13laCI-infected cells. With this system, they have determined
the sequence changes in the 250-nucleotide region of lac Operon from a
large number of UV-induced phage mutants (LeClerc et al., 1984).

The cI repressor gene of lambda phage has also been used for
determining the UV-mutational specificity (Skopek & Hutchinson, 1982; Wood
et al., 1984). Mutants with the clear plaque phenotype were isolated, and
the mutations were sorted into the cI, cII and cIII genes by
complementation tests. Each cI mutation was mapped by genetic crosses. The
appropriate DNA restriction fragment was isolated, and then sequenced. The
spectna of UV-induced mutations observed in these studies are in good
agreement with that obtained by LeClerc & Istock (1982) for the E. coli

lac promoter cloned in M13.

2. Shuttle vector systeus in aauaalian cells.

Although the complexity of higher eukaryotic cells makes it very
difficult to obtain information at the DNA sequence level on mutations in
such cells, the development of recombinant DNA techniques during the last
decade have enabled researchers to devise new approaches to the study of
mutations in mammalian cells. One of these is the use of shuttle vectors
(Calos et al., 1983; Lebkowski et al., 1985; Seidman et al., 1985), which
can replicate in both mammalian and bacterial cells. Therefore, a target
gene carried on this vector can be mutated in mammalian cells and then
transferred to bacteria for rapid analysis of mutations. Although these
SV40-based vectors must be supplied with T antigen for replication, they
have to reJy totally on the host cell DNA repair and DNA replication

machinery, and therefore can be used to investigate the host cell’s

29
mutagenesis activity based on the assumption that mutations are produced
when DNA synthesis occurs on a damaged DNA template (Konze-Thomas et al.,
1982; Grossman et al., 1985).

The lacI shuttle vector, which contains the entire lacI gene, the
SV40 origin of replication and plasmid sequences required for replication
in E. coli, was first used for determination of mutagenic specificity, due
to the elegant identification system of lacI nonsense mutation developed
by Coulondre & Miller (1977). However, this vector acquired mutations in
its target gene at a very high frequency (1%), resulting simply from
passage of the vectors through the mammalian cells (Calos et al., 1983;
Razzaque et al., 1984). This high background frequency presented an
obstacle to studies of mutagens that do not induce a very high mutant
frequency. Later it was found that in human 293 cells, the frequency of
transfection-associated mutations is lower (Lebkowski et al., 1984), and
so these host cells have been used for transfecting the lacI shuttle.
Using these cells, Lebkowski et al. (1985; 1986) found that the UV- and
ethylnitrosourea (ENU)-induced mutant frequencies were significantly
higher than the background frequency, and the mutational specificity of
these two agents was analyzed.

Seidman et al. (1985) showed that the spontaneous mutations obtained
in the SV40-based shuttle vectors can be minimized by judicious design of
the vectors. They constructed the le89 vector, which contains the origin
of replication and B-lactamase gene from p8R327, a supF suppressor tRNA
gene from E. coli as the target for mutational analysis, and the early
region of SV40 DNA to allow replication in mammalian cells. Because of the

small size of the target gene and its strategic location between two

30

essential genes, the B-lactamase gene and the plasmid origin of
replication, the spontaneous mutation frequency in the supF gene was
greatly reduced (~0.04%). This vector, as well as its closely related
plasmid, p3189 (Seidman, 1989), have been widely used for studying the
spontaneous and carcinogen-induced mutations in mammalian cells (Hauser et
al., 1986, Bigger et al., 1989; Kraemer & Seidman, 1989; Maher et al.,
1989; Mah et al., 1990; Bodlt et al., 1991). These studies have
demonstrated that a single base pair change at almost any site in the 85bp
structure of the tRNA results in a mutant phenotype (Kraemer & Seidman,
1989), indicating that this gene is exceptionally responsive to mutagens
and allows for very few silent mutations.

Another way to reduce the background mutation frequency of shuttle
vectors has been the use of Epstein-Barr virus (E8V)-based vectors which
are stably maintained in the cells and replicated synchronously with
genomic DNA. Since most of the spontaneous mutations that occurr in the
transiently replicating, SV40-based vectors are early events and are
induced by transfection process (Razzaque et al., 1983; Lebkowski et al.,
1984), the use of such stably maintained vector allows one to isolate
rare clones of cells that contain only unmutated vector DNA, and thus
circumvent the problem of high vectors’ background frequencies. Two EBV-
shuttles, pHET and pMc, carrying the target genes of Herpes simplex virus
thymidine kinase (HSVtk) and E. coli lacI, respectively, were constructed
for this purpose, and have been used for the analysis of alkylating agent-
induced mutations (DuBridge et al., 1987; Drinkwater & Klinedinst, 1986;
Eckert et al., 1988).

31

3. Stably integrated exogenous gene systel in Ian-alian cells

Shuttle vector systems have provided a rapid way for the study of
sequence alterations in mammalian cells. However, one cannot be certain
that mutations detected using these vectors reflect the normal nuclear
replicative and repair mechanisms of the host cell, because the
extrachromosomal DNA does not exhibit the normal chromosomal context. One
approach to overcoming this problem is to study of the mutations induced
in a small exogenous locus stably integrated into mammalian genome. For
this approach, Tindall et al. (1984; 1986) constructed a cell line, A552,
which carries a single, functional copy of the E. coli gpt gene stably
integrated into a HPRT' CHO cell line. Since the bacterial gpt gene
product, xanthine-guanine phosphoribosyltransferase (XPRT), in the A552
cell line is functionally analogous to HPRT of wild-type CHO cells,
mutants at the gpt locus can be selected by their resistance to 6-
thioguanine (6-TG). This cell line has been used in quantitative and
molecular studies of mutations induced by several mutagens, including X-
ray, UV, ENU, and ICR191 (Tindall et al. 1984; 1986; Stankowski et al.,
1986). However, due to the difficulty to determine the sequence of mutated
target gene, Southern blot analysis was used for their molecular studies
of mutations, which had limited the identification of single base changes
in this system.

Other systems have been devised to sequence mutated genes on a
mammalian chromosome: the HSVtk gene on a shuttle vector that integrates
into a chromosome of tk’ mouse fibroblasts (Zakour et al., 1986), and the
supF gene on the lambda phage DNA stably incorporated in a chromosome of

mouse L cells (Glazer et al., 1986). Both systems allow the rescue of

32
integrated plasmid and phage DNA from nemmalian cells by cfloning and
lambda DNA packaging, respectively, and therefore facilitate the

determination of exact sequence changes on target genes.

4. Endogenous Ian-alian genes systems

The analysis of mutations at the DNA sequence level in the
endogenous genes in mammalian cells requires a) a hemizygous target gene,
b) a selection scheme to isolate mutant cells, and c) a method to retrieve
and amplify the target gene for sequence analysis. The HPRT gene resides
on the X-chromosome in all mammals studied (Thaker, 1985), and therefore,
the locus is hemizygous in males and functionally hemizygous in females.
Furthermore, the selection conditions are available to isolate hprt'
mutants both in vitro (DeMars, 1974) and in vivo (Albertini et al., 1982).
However, until recently the extremely large size of this gene had limited
its application as a target for the generation of mutational spectra
because of the technical difficulties associated with cloning and
sequencing a large gene.

The recent development of polymerase'chain reaction (PCR) (Mullis et
al., 1986; Saiki et al., 1988) has made it possible to selectively amplify
a short segment of DNA more than IDS-fold, so that DNA sequence analysis
can be performed without the need for time-consuming library construction
and screening. This technique, in conjunction‘with the utilization of cDNA
as the source of sequence information, has provided a simple way to
analyze mutations induced in the HPRT gene. Vrieling et al. (1989) have
studied the UV-induced hprt mutations in a Chinese hamster cell line.

Using reverse transcriptase, they synthesized hprt cDNA from total

33

cytoplasmic RNA isolated from 10-20 X 106 cells, amplified hprt cDNA by
PCR, and then cloned it into an M13 vector for sequencing. To minimize the
chance that the sequence alterations were due to nucleotide
misincorporation by the Taq polymerase during amplification, at least two
clones for each mutants were sequenced, and only mutations found in both
clones were considered to be mutagen-induced. The same approach was used
by Carothers et al. (1989) and Simpson et al. (1988) for analyses of
mutations induced in the HPRT gene in human lymphoblasts and in the dhfr
gene of CHO cells, respectively. L

Yang et al. (1989) have developed a method allowing the rapid
detection of sequence alterations in the endogenous genes of mammalian
cells using only a few mutant cells, an improvement that is necessary for
studying mutational spectra in endogenous genes of cultured mammalian
cells that have a finite life span. This method involved synthesis of
total cDNA directly from a lysate of a small number of human cells,
selective amplification of the cDNA of the gene of interest 10‘0- to 10"-
fold, and determination of the nucleotide sequence by direct sequencing of
the amplified product. To avoid detecting the mutations randomly
introduced by reverse transcriptase and Tag polymerase, 100-500 cells are
subjected to reverse transcription and PCR. Since the amplified products
were sequenced directly to obtain the consensus sequence, a random base
substitution, even introduced in the reverse transcription process or in
an early round of PCR, will not be seen on the DNA sequencing gel. This
method is being used widely for studying mutational spectra in endogenous

genes of mammalian cells.

34
C. The influence of DNA replication and nucleotide excision repair on
lutagenesis

1. General roles of replication and repair in lutagenesis

The effect of nucleotide excision repair on the mutagenicity of
various carcinogens in human diploid cells has been determined in two
ways. One method made use of nucleotide excision repair-deficient
fibroblasts derived from XP patients. Comparisons of these cells with
cells obtained from normal donors showed that the XP cells are much more
sensitive to the nmtagenic effects of several carcinogens, such as UV
(Maher et al., 1979; 1982), BPDE (Yang et al., 1980), 1-nitrosopyrene
(Maher et al., 1988), and metabolic derivatives of several polycyclic
hydrocarbon compounds (Maher et al., 1977). The second method
investigated the frequency of mutant induction in cells treated with
carcinogens and held in confluence. Such cells are able to carry out
excision repair but are unable to replicate. After various periods of time
in confluence, cells were replated at lower density, allowed to replicate,
and after an appropriate expression period assayed for frequency of
induced mutants. The repair-proficient cells released immediately
following treatment exhibited the highest frequency. Furthermore, there
was a gradual reduction of the frequency of induced mutants with time held
in confluence. This was directly correlated with a gradual increase in
survival. However, the repair-deficient cells showed no change in their
survival and frequency of mutants throughout the entire holding period
(Maher et al., 1979; Yang et al., 1980). These results indicate that

excision repair can decrease the cytotoxic and mutagenic consequences of

35
carcinogen treatment, most probably by removing the DNA lesions
responsible for the effects.

IrI order to determine ‘the relationship of DNA replication to
mutagenesis, synchronized human cells were used so that they could be
exposed to carcinogens at various times in the cell cycle and then
examined for the frequency of mutant induction. The results of these
studies showed that the frequencies of mutants induced by UV (Konze-Thomas
et al., 1982), BPDE (Yang et al., 1982), and 1-nitrosopyrene (Maher et
al., 1988) were highest in cells treated in early S phase and much lower
in cells treated in early G, phase, 12-15 hr. before the onset of DNA
synthesis. Furthermore, the frequencies of mutant induction were directly
correlated with the number of unrepaired lesions remaining in the cells’
DNA when the target gene (HPRT) was replicated. These data support the
hypothesis that DNA replication is the cellular event that converts
unexcised DNA lesions into mutations, and that excision repair of such
lesions in human fibroblasts protects them from potentially mutagenic

lesions.

2. Role of heterogeneous repair in mutagenesis

Intragenomic heterogeneity in repair processes may influence
mutagenesis. For instance, the apparent lack of repair of py-py dimers in
unexpressed genes of rodent cells in vitro would be expected to result in
higher mutation frequencies in those genes, compared to actively
transcribed genes. However, since selectable phenotypes result only from
expressed genes, it is not possible to select for mutants in such non-

expressed genes. Therefore, the comparison of mutation frequencies in

36
expressed and silent genes would require development of a specific test
system. One ideal system would be the use of an inducible gene whose loss
of activity could provide cells with a phenotypic change suitable for
selection under defined conditions, which would have to be applied after
mutagen treatment and mutation fixation. So far, no such system has been
available for mutational analysis.

Another important question raised by heterogenous repair is whether
the preferential repair of the transcribed strand of a gene would result
in a strand bias in mutations. Indeed, strand bias was reported for AAF-
(Carothers et al., 1989), BPDE- (Carothers & Grunberger, 1990), and (i)-
3a,4B-dihydroxy-la,2a-epoxy-1,2,3,4,-tetrahydrobenzo[c]phenanthrene
(8[c]PHDE)- (Carothers et al., 1990) induced mutations in the rodent dhfr
gene. However, since the identification mutants depends on the loss of the
functional dhfr gene product, the sites at which a sequence change results
in a non-functional protein may also be biased with respect to the
strands. Notably, the selection of a dhfr" phenotype appears to be very
stringent and tends to yield mostly nonsense and mRNA splicing mutations
(Carothers et al., 1989). Therefore, such data can not readily be used to
make inferences about the repair process, unless comparative studies in
which mutations are induced in cells with different repair ability or with
different lengths of time available for repair prior to mutation fixation
have been performed.

Carothers et al. (1991) designed a nonsense-codon reversion assay,
in which the targets for reversion are two nonsense codons at both ends of
the dhfr gene. The sequence context for 8[c]PHDE-induced base changes was

similar (four tandem purines) on both DNA strands, and all R (purine) ->

37

T transversions primarily induced by 8[c]PHDE at the two nonsense codons
yield dhfr+ phenotype. Thus in this system, mutational strand bias is
independent of the selection for phenotypic change. Since each of the
nonsense codons contains two purines on the nontranscribed strand, a 2:1
ratio of mutations at the purines on the nontranscribed strand versus the
transcribed strand was expected if the DNA strands were mutated with equal
probability. Instead, they found that >90% of the mutations were targeted
to the purines on the nontranscribed strand, consistent with the model of
preferential repair of the transcribed strand. However, preferential
repair may not be the sole explanation for the discrepancy between the
expected and observed ratio. For example, if the frequency of adduct
formation on each of the purine targets were unequal, this would result in
a strand-biased modification. Such data has not been yet determined in
their system.

Another factor that could cause a strand specificity of mutagenesis
is the fidelity of DNA replication of the two DNA strands. Studies on the
replication of SV40 DNA using a reconstituted in vitro system implicated
the requirement of DNA polymerase 6 and polymerase a in the synthesis of
the leading strand and lagging strand, respectively (Prelich & Stillman,
1988). If the ability of the two DNA polymerases to insert correct bases
opposite the carcinogen-modified bases were different, this could result
in a strand-specific mutagenesis. Indeed, it has been shown that during
translesional DNA synthesis on an oligonucleotide template containing an
8-oxo-7-hydrodeoxyguanosine, DNA polymerase a is more error prone than
polymerase 6 (Shibutani et al., 1991). A difference in fidelity of DNA

replication of the two strands has been suggested by Vrieling et al.

38
(1989) as the explanation for the strong strand bias of UV-induced hprt
mutations in a repair-deficient Chinese hamster cell line (V-Hl). Since
the cell cannot repair UV damage, one would expect a random distribution
of premutagenic lesions. However, "90% of the mutations derived from such
cells were targeted to photoproducts in the transcribed strand. Since this
strand is also the leading strand for most part of the hprt gene, they
suggested that polymerase 6 is more error prone than polymerase 0:.
However, this argument is contradictory to the results obtained by
Shibutani et al. (1991). Furthermore, direct evidence of different error
rates of the two polymerases during DNA replication bypassing UV-induced

photoproducts has not been provided.

REFERENCES

Albertini, R. J., Castle, K. L., and Borcherding, W. R. 1982. T-cell
cloning to detect the mutant 6-thioguanine-resistant lymphocytes present
in human peripheral blood. Proc. Natl. Acad. Sci. (USA) 79:6617-6621.

Ames, 8. N., Lee, F. D., and Durston, W. E. 1973. An improved bacterial
test system for the detection and classification of mutagens and
carcinogens. Proc. Natl. Acad. Sci. (USA) 70:782-786.

Arrand, J. E., Bone, N. M., and Johnson, R. T. 1989. Molecular cloning and
characterization of a mammalian excision repair gene that partially
restores UV resistance to xeroderma pigmentosum complementation group 0
cells. Proc. Natl. Acad. Sci. (USA) 86:6997-7001.

Arrand, J. E. and Murray, A. M. 1982. Benzo(a)pyrene groups bind
preferentially to the DNA of active chromatin in human lung cells. Nucleic
Acids Res. 10:1547-1555.

Bailey, G. S., Nixon, J. E., Hendricks, J. 0., Sinnhuber, R. 0., and
Holde, K. E. 1980. Carcinogen aflatoxin 81 is located preferentially
internucleosomal deoxyribonucleic acid following exposure in vivo in
rainbow trout. Biochemistry 19:5836-5842.

Baranyi-Furlong, 8. L. and Goodman, J. I. 1984. Non-random interaction of
N-2-acetylaminofluorene and its N-hydroxy metabolite with DNA of hepatic
parenchymal and non-parenchymal cell nuclei following in vivo
administration of carcinogens. Chem.-Biol. Interact. 48:15-28.

Benyajati, (L. and Worcell, A. 1976. Isolation, characterization, and
structure of the folded interphase genome of Drosophi la melanogaster. Cell
9:393-407.

Berkowitz, E. M. L. and Silk, H. 1981. Methylation of chromosomal DNA by
two alkylating agents differing in carcinogenic potential. Cancer Lett.
12:311-321.

Bigger, C. A. H., Strandberg, J., Yagi, H., Jerina, D. M., and Dipple, A.
1989. Mutagenic specificity of a potent carcinogen, benzo[c]phenanthrene
(4R,3S)-dihydrodiol (25,1R)-epoxide, which reacts with adenine and guanine
in DNA. Proc. Natl. Acad. Sci. (USA) 86:2291-2295.

Bohr, V. A. 1987. Differential DNA repair within the genome. Cancer Rev.
7:28-55.

Bohr, V. A., Dkumoto, D. S., Ho, L., and Hanawalt, P. C. 1986b.
Characterization of a DNA repair domain containing the dihydrofolate
reductase gene in Chinese hamster ovary cells. J. Biol. Chem. 261: 16666-
16672.

Bohr, V. A., and Dkumoto, D. S. 1988. Analysis of pyrimidine dimers in
39

40

defined genes. In: Friedberg E. C., Hanawalt, P. C. (eds) DNA repair: A
Laboratory Manual of Research Procedures. Marcel Dekker, Inc, NY, pp.347-
366.

Bohr, V. A., Dkumoto, D. S., and Hanawalt, P. C. 1986a. Survival of UV-
irradiated mammalian cells correlates with efficient DNA repair in an
essential gene. Proc. Natl. Acad. Sci. (USA) 83: 3830-3833.

Bohr, V. A., Smith, C. A., Dkumoto, D. S., and Hanawalt, P. C. 1985. DNA
repair in an active gene: removal of pyrimidine dimers from the DHFR gene
of CHO cells is nmch were efficient than in the genome overall. Cell
40:359-369.

Boldt, J., Mah, C.-M., Wang, Y.-C., Smith, 8. A., Beland, F. A., Maher, V.
M., and McCormick, J. J. 1991. Kinds of mutations found when a shuttle
vector containing adducts of 1,6-dinitropyrene replicates in human cells.
Carcinogenesis 12:119-126.

Boyce, R. P. and Howard-Flanders, P. 1964. Release of ultraviolet light-
induced thymine dimers from DNA in E. coli. Proc. Natl. Acad. Sci. (USA)
51:293-300.

Brady, F. 0. 1982. The physiological function of metallothionein. Trend
Biochem. Sci. 7:143-145.

Brash, D. E., Franklin, W. A., Sancar, G. 8., Sancar, A., and Haseltine,
W. A. 1985. E. coli DNA photolyase reverses cyclobutane pyrimidine dimers
but not pyrimidine-pyrimidone (6-4) photoproducts. J. Biol. Chem.
260:11438-11441.

Busch, 0. 8., Cleaver, J. E., and Glaser, D. A. 1980. Large-scale
isolation of UV-sensitive clones of CHO cells. Somat. Cell Genet. 6:407-
418.

Busch, D., Greiner, C., Lewis, K., Ford, R., Adair, G., and Thompson, L.
1989. Summary of complementation groups of UV-sensitive CHO cell mutants
isolated by large-scale screening. Mutagenesis 4:349-354.

Calos, M. P., Lebkowski, J. S., and Botchan, M. R. 1983. High mutation
frequency in DNA transfected into mammalian cells. Proc. Natl. Acad. Sci.
(USA) 80:3015-3019.

Calos, M. P., and Miller, J. H. 1981. Genetic and sequence analysis of
frameshift mutations induced by ICR-191. J. Mol. Biol. 153:39-66.

Caron, P. R. and Grossman, L. 1988. Incision of damaged versus nondamaged
DNA by the Escherichia coli UvrABC proteins. Nucleic Acids Res. 16:7855-
7865.

Carothers, A. M. and Grunberger, D. 1990. DNA base changes in
benzo[a]pyrene diol epoxide-induced dihydrofolate reductase mutants of
Chinese hamster ovary cells. Carcinogenesis 11:189-192.

41

Carothers, A. M., Mucha, J., and Grunberger, D. 1991. DNA strand-specific
mutations induced by (:)-3a,4B-dihydroxy-la,2a-epoxy-1,2,3,4-
tetrahydrobenzolc]phenanthrene. Proc. Natl. Acad. Sci. (USA) 88: 5749-
5753.

Carothers, A. M., Steigerwalt, R. W., Urlaub, G., Chasin, L. A., and
Grunberger, D. 1989. DNA base changes and RNA levels in N-acetoxy-Z-
acetylaminofluorene-induced dihydrofolate reductase mutants of CHO cells.
J. Mol. Biol. 208:417-428.

Carothers, A. M., Urlaub, G., Mucha, J., Harvey, R. G., Chasin, L. A., and
Grunberger, D. 1990. Splicing mutations in the CH0 DHFR gene
preferentially induced by (1)-3a,4B-dihydroxy-1a,2a-epoxy-1,2,3,4-
tetrahydrobenzolc]phenanthrene. Proc. Natl. Acad. SCi. (USA) 87:5464-
5468.

Carothers, A. M., Urlaub, G., Mucha, J., Grunberger, 0., and Chasin, L. A.
1989. Point mutation analysis in a mammalian gene: rapid preparation of
total RNA, PCR amplification of cDNA, and Taq sequencing by a novel
method. Biotechniques 7:494-499.

Chu, G. and Chang, E. 1988. Xeroderma pigmentosum group E cells lack a
nuclear factor that binds to damaged DNA. Science 242:564-567.

Clarkson, J. M., Mitchell, 0. L., and Adair, G. M. 1983. The use of an
immunological probe to measure the kinetics of DNA repair in normal and
UV-sensitive mammalian cell lines. Mutat. Res. 112:287-299.

Cleaver, J. E. 1968. Defective repair replication of DNA in xeroderma
pigmentosum. Nature 218:652-656.

Cleaver, J. E. 1972. Xeroderma pigmentosum: variants with normal DNA
repair and normal sensitivity to ultraviolet light. J. Invest. Dermatol.
58:124-128.

Cleaver, J. E. Xeroderma pigmentosum. 1983. In The Metabolic Basis of
Inherited Disease, 5th ed. Stanbury, J. B., Wyngaarden, J. 8.,
Fredrickson, D. S., Goldstein, J. L., and Brown, M. S. eds., p. 1227. New
York: McGraw-Hill.

Cleaver, J. E. 1990. Do we know the cause of xeroderma pigmentosum?
Carcinogenesis 11: 875-882.

Cleaver, J. E. and Bootsma, D. 1975. Xeroderma pigmentosum: biochemical
and genetic characteristics. Ann. Rev. Genet. 9:19-38.

Cleaver, J. E., Greene, A. E., Coriell, L. L., and Mulivor, R. A. 1981.
Xeroderma pigmentosum variants. Cytogenet. Cell Genet. 31:188-192.

Cook, P. R. and Brazell, I. A. 1975. Supercoils in human DNA. J. Cell.
Sci. 19:261-279.

42

Coulondre, C. and Miller, J. H. 1977. Genetic studies of the lac
repressor: mutagenic specificity in the lacI gene of Escherichia coli.
J. Mol. Biol. 117:577-606.

Coverley, 0., Kenny, M. K., Munn, M., Rupp W. 0., Lane, 0. P., and Wood,
R. 0. Requirement for the replication protein SS8 in human DNA excision
repair. Nature 349:538-541.

DeMars, R. 1974. Resistance of cultured human fibroblasts and other cells
to purine and pyrimidine analogues in relation to mutagenesis detection.
Mutat. Res. 24:335-364.

Drinkwater, N. R., and Klinedinst, D. K. 1986. Chemically induced
mutagenesis in a shuttle vector with a low-background mutant frequency.
Proc. Natl. Acad. Sci. (USA) 83:3402-3406.

DuBridge, R. B., Tang, P., Hsia, H. C., Leong, P.-M., Miller, J. H., and
Calos, M. P. 1987. Analysis of mutation in human cells by using an
Epstein-Barr virus shuttle system. Mol. Cell. Biol. 7:379-387.

Eckert, K. A. Ingle. C. A, Klinedinst, D. K, and Drinkwater, N. R.
1988. Molecular analysis of mutations induced in human cells by'N-ethyl- -N-
nitrosourea. Mol. Carcinogenesis 1: 50-56.

Friedberg, E. C., Ehmann, U. K., and Williams, J. I. 1979. Human diseases
associated with defective DNA repair. Adv. Rad. Biol. 8:85.

Fuchs, R. P. P., Schwartz, N., and Daune, M. P. 1981. Hot spots of
frameshift mutations induced by the ultimate carcinogen N-acetoxy-N-Z-
acetylaminofluorene. Nature 294:657-659.

Fujiwara, Y. Tatsumi, M., Sasaki, M. S. 1977. Cross-link repair in human
cells and its possible defect in Fanconi’s anemia cells. J. Mol. Biol.
113:635-649.

Gale, J. M., Nissen, K. A., and Smerdon, M. J. 1987. UV-induced formation
of pyrimidine dimers in nucleosome core DNA is strongly modulated with a
period of 10.3 bases. Proc. Natl. Acad. Sci. (USA) 84:6644-6648.

Glazer, P. M., Sarkar, S. N., and Summers, W. C. 1986. Detection and
analysis of UV-induced mutations in mammalian cell DNA using a i phage
shuttle vector. Proc. Natl. Acad. Sci. (USA) 83:1041-1044.

Grossmann, A., Maher, V. M., and McCormick, J. J. 1985. The frequency of
mutants in human fibroblasts UV-irradiated at various times during S-phase
suggests that genes for thioguanine and diphtheria toxin resistance are
replicated early. Mutat. Res. 152:67-76.

Hanawalt, P. C., Cooper, P. K., Ganesan, A. K., and Smith, C. A. 1979. DNA
repair in bacteria and mammalian cells. Ann. Rev. Biochem. 48:783-836.

Hanawalt, P. C., Mellon, I., Scicchitano, 0., Spivak, G. 1989.

43

Relationships between DNA repair and transcription in defined DNA
sequences in mammalian cells. In: DNA Repair Mechanisms and Their
Biological Implications, Lambert, M. W., ed. pp. 325-337. New York:
Plenum.

Harless, J. and Hewitt, R. R. 1987. Intranuclear localization of UV-
induced DNA repair in human VA-13 cells. Mutat. Res. 183:177-184.

Hauser, J., Seidman, M. M., Sidur, K., and Dixon, K. 1986. Sequence
specificity of point mutations induced during passage of a UV-irradiated
shuttle vector plasmid in monkey cells. Mol. Cell. Biol. 6:277-285.

Hoeijmakers, J. H. J., 0dijk, H., and Westerveld, A. 1987. Differences
between rodent and human cell lines in the amount of integrated DNA after
transfection. Exp. Cell Res. 169:111-119.

Hoy, C. A., Thompson, L. H., Mooney, C. L., and Salazar, E. P. 1985.
Defective cross-link removal in Chinese hamster cell mutants
hypersensitive to bifunctional alkylating agents. Cancer Res. 45:1737-
1743.

Igo-Kemenes, T. and Zachau, H. G. 1977. Domains in chromatin structure.
Cold Spring Harbor symp. Duant. Biol. 42:109-118.

Jahn, C. L. and Litman, G. W. 1979. Accessibility of DNA in chromatin to
the covalent binding of the chemical carcinogen benzo(a)pyrene.
Biochemistry 18:1442-1446.

Kaneko, M. and Cerutti, P. A. 1980. Excision of N-acetoxy-2-acetylamino-
fluorene-induced DNA adducts from chromatin fractions (Mi human
fibroblasts. Cancer Res. 40:4313-4319.

Koffel-Schwartz, N., Verdier, J.-M., Bichara, M., Freund, A.-M., Daune, M.
P., and Fuchs, R. P. P. 1984. Carcinogen-induced mutation spectrum in
wild-type, uvrA and umuC strains of Escherichia coli: strain specificity
and mutation-prone sequences. J. Mol. Biol. 177: 33-51.

Kojima, Y. and Kaegi, J. H. R. 1978. Metallothionein. Trends Biochem. Sci.
3:90-93.

Konze-Thomas, 8., Hazard, R. M., Maher, V. M., McCormick, J. J. 1982.
Extent of excision repair before DNA synthesis determines the mutagenic
but not the lethal effect of UV radiation. Mutat. Res. 94:421-434.

Kraemer, K. H. and Seidman, M. M. 1989. Use of supF, an Escherichia coli
tyrosine suppressor tRNA gene, as a mutagenic target in shuttle-vector
plasmids. Mutat. Res. 220:61-72.

Kumura, K., Sekiguchi, M., Steinum, A.-L., and Seeberg, E. 1985.
Stimulation of the UvrABC enzyme-catalyzed repair reaction by Uer protein
(DNA helicase II). Nucleic Acids Res. 13:1483-1492.

44

Lan, K. S. Y. and Smerdon, M. J. 1985. A nonuniform distribution of
excision repair synthesis in nucleosome core DNA. Biochemistry 24:7771-
7783.

Lawrence, C. W. and Christensen, R. B. 1979. Absence of relationship
between UV-induced reversion frequency and nucleotide sequence'of the CYCl
locus of yeast. Mal. Gen. Genet. 177:31-38.

Leadon, S. A. 1988. Immunological probes for lesions and repair patches in
DNA. In DNA Repair. A Laboratory'Manual of Research Procedures, Friedberg
E. C. and Hanawalt, P. C. eds. pp. 311-326. New York: Marcel Dekker, Inc.

Leadon, S. A. 1986. Differential repair' of DNA damage in specific
nucleotide sequences in monkey cells. Nucleic Acids Res. 14:8979-8995.

Leadon, S. A. Snowden, M. M. 1988. Differential repair of DNA damage in
the human metallothionein gene family. Mal. Cell Biol. 8:5331-5338.

Leadon, S. A., Zolan, M. E., and Hanawalt, P. C. 1983. Restricted repair
of aflatoxin 81 induced damage in a DNA of monkey cells. Nucleic Acids
Res. 11:5675-5689.

Lebkowski, J. S., Clancy, S., Miller, J. H., and Calos, M. P. 1985. The
lacI shuttle: rapid analysis of the mutagenic specificity of ultraviolet
light in human cells. Proc. Natl. Acad. Sci. (USA) 82:8606-8610.

Lebkowski, J. S., DuBridge, R. 8., Antell, E. A., Greisen, K. S., and
Calos, M. P. 1984. Transfected DNA is mutated in monkey, mouse, and human
cells. Mal. Cell. Biol. 4:1951-1960.

Lebkowski, J. S., Miller, J. H., and Calos, M. P. 1986. Determination of
DNA sequence changes induced by ethyl methanesulfanate in human cells,
using a shuttle vector system. Mal. Cell. Biol. 6:1838-1842.

LeClerc, J. E., Istock, N. L., Saran, B. R., and Allen, R. Jr. 1984.
Sequence analysis of ultraviolet-induced mutations in M13lac hybrid phage
DNA. J. Mol. Biol. 180: 217-237.

LeClerc, J. E., and Istock, N. L. 1982. Specificity of UV mutagenesis in
the lac promoter of M13lac hybrid phage DNA. Nature 297:596-598.

LeDoux, S. P., Thangada, M. Bohr, V. A., and Wilson, G. L. 1991.
Heterogeneous repair of methylnitrosourea-induced alkali-labile sites in
different DNA sequences. Cancer Res. 51:775-779.

Lehmann, A. R. 1985. Use of recombinant DNA techniques in cloning DNA
repair genes and in the study of mutagenesis in mammalian cells Mutat.
Res. 150:61-67.

Lehmann, A. R. 1987. Cockayne’s syndrome and trichothiodystrophy:
defective repair without cancer. Cancer Rev. 7:82-103.

45

Lieberman, M. W. and Poirier, M. C. 1974. Intragenomal distribution of
DNA repair synthesis: repair in satellite and mainband DNA in cultured
mouse cells. Proc. Natl. Acad. Sci. (USA) 71:2461-2465.

Lieberman, M. W., Smerdon, M. J., Tisty, T. 0., and Oleson, F. B. 1979. In
Environ. Carcinogenesis, Emelot, P., and Kriek, E. eds., pp. 345-213.
Amsterdam Elsevier/North Holland Biomedical Press, Inc.

Lindahl, T. 1982. DNA repair enzymes. Ann. Rev. Biochem. 51:61-87.

Livneh, Z. 1983. Directed mutagenesis method for analysis of mutagen
specificity: Application to ultraviolet-induced mutagenesis. Proc. Natl.
Acad. Sci. (USA) 80:237-241.

Madhani, H. 0., Bohr, V. A., and Hanawalt, P. C. 1986b. Differential DNA
repair in transcriptionally'active~and inactive prota-oncagenes: c-abl and
c-mos. Cell 45:417-423.

Madhani, H. D., Leadon, S. A., Smith, C. A., and Hanawalt, P. C. 1986a. a
DNA in African green monkey cells is organized into extremely long tandem
arrays. J. Biol. Chem. 261: 2314-2318.

Mah, M. C.-M., Maher, V. M., Thomas, H., Reid, T. M., King, C. M., and
McCormick, J. J. 1989. Mutations induced by aminofluorene-DNA adducts
during replication in human cells. Carcinogenesis 10:2321-2328.

Maher, V. M., Dorney, D. J., Mendrala, A. L., Konze-Thomas, B., and
McCormick, J. J. 1979. DNA excision-repair processes in human cells can
eliminate the cytotoxic and mutagenic consequences of ultraviolet
irradiation. Mutat. Res. 62:311-323.

Maher, V. M., McCormick, J. J., Graver, P. L., and Sims, P. 1977. Effect
of DNA repair an the cytotoxicity and mutagenicity of polycyclic
hydrocarbon derivatives in normal and xeroderma pigmentosum human
fibroblasts. Mutat. Res. 43:117-138.

Maher, V. M., Patton, J. 0., and McCormick, J. J. 1988. Studies on the
metabolism and biological effects of nitrapyrene and related nitro-
polycyclic aromatic compounds in diploid human fibroblasts. Health Effects
Institute Report #17, pp. 1-40.

Maher, V. M., Rowan, L. A., Silinskas, K. C., Kateley, S. A., and
McCormick, J. J. 1982. Frequency of UV-induced neoplastic transformation
of diploid human fibroblasts is higher in xeroderma pigmentosum cells than
in normal cells. Proc. Natl. Acad. Sci. (USA) 79:2613-2617.

Maher, V. M., Yang, J-L., Mah, M. C-M., and McCormick, J. J. 1989.
Comparing the frequency and spectra of mutations induced when an SV-40
shuttle vector containing covalently bound residues of structurally-
related carcinogens replicates in human cells. Mutat. Res. 220:83-92.

Mansbridge, J. N. and Hanawalt, P. C. 1983. Domain-limited repair of DNA

46

in ultraviolet irradiated fibroblasts from xeroderma pigmentosum
complementation group C. In Cellular Responses to DNA Damage. Friedberg,
E. C., and Bridges, B. R., eds. pp. 195-208. New York: Alan R. Liss, Inc.

Mayne, L. V., Mullenders, L. H. F., Van Zeeland, A. A. 1988. Cockayne’s
syndrome: An UV sensitive disorder with a defect in the repair of
transcribing DNA but normal overall excision repair. In: Friedberg, E. C.,
Hanawalt, P. C. (eds) Mechanisms and sansequences of DNA damage
processing. pp.349-353. Alan R. Liss, Inc, NY.

McCready, S. J. and Cook, P. R. 1984. Lesions induced in DNA by
ultraviolet light are repaired at the nuclear cage. J. Cell Sci. 39:53-62.

Mellon, 1., Bohr, V. A., Smith, C. A., and Hanawalt, P. C. 1986.
Preferential DNA repair in transcriptionally active gene in human cells.
Proc. Natl. Acad. Sci. (USA) 83:8878-8882.

Mellon, I. and Hanawalt, P. C. 1989. Induction of the Escherichia coli
lactose>aperon selectively increases repair of its transcribed DNA strand.
Nature 242:95-98.

Mellon, I., Spivak, G., Hanawalt, P. C. 1987. Selective removal of
transcription-blocking DNA damage from transcribed strand of the mammalian
DHFR gene. Cell 51:241-249.

Messing, J., Gronenborn, 8., Muller-Hill, 8., and Hofschneider, P. H.
1977. Filamentous coliphage M13 as a cloning vehicle: Insertion of a
HindII fragment of the lac regulatory region in M13 replicative form in
vitra. Proc. Natl. Acad. Sci. (USA) 74: 3642-3646.

Metzger, G., Wilhem, F. X., and Wilhem, M. L. 1977. Non-random binding of
a chemical carcinogen to the DNA in chromatin. Biochem. Biophys. Res.
Commun. 75:703-706.

Miller, J. H. 1982. Carcinogens induce targeted mutations in Escherichia
coli. Cell 31:5-7.

Montelone, B. A., Hoekstra, M. F., Malone, R. E. 1988. Spontaneous mitotic
recombination in yeast: The hyper-recombinational remI mutations are
alleles of the RA03 gene. Genetics 119:289-301.

Monttesano, R. 1981. Alkylation of DNA and tissue specificity in
nitrosamine carcinogenesis. J. Supramal. Struct. Cell. Biochem. 17:259-
273.

Moshell, A.N., Ganges, M. 8., Lutzner, M. A., Coon, H. G., Barrett, S. F.,
Dupuy, J.-M., and Robbins, J. H. 1983. A new patient with both xeroderma
pigmentosum and Cockayne syndrome establishes the new xeroderma
pigmentosum complementation group H. In Cellular Responses to DNA Damage.
Friedberg, E. C. and Bridges, B. A. eds.. PP. 209-213. New York: Alan R.
Liss.

47

Mayne, G., Freeman, R., Savayosti, S., and Yaniv, M. 1981. A high-
resolution EM study of nucleosomes from SV40 chromatin. J. Mol. Biol.
149:735-744.

Mullenders, L. H. F., Van Kesteren, A. C., Bussmann, C. J., Van Zeeland,
A. A., and Natarajan, A. T. 1984. Preferential repair of nuclear matrix
associated DNA in xeroderma pigmentosum complementation group C. Mutat.
Res. 141:75-82.

Mullenders, L. H. F., Van Kesteren-Van Leeuwen, A. C., Van Zeeland, A. A.,
and Natarajan, A. T. 1988. Nuclear matrix associated DNA is preferentially
repaired in normal human fibroblasts, exposed to a low dose of UV light,
but not in Cockayne’s syndrome fibroblasts. Nucleic Acids Res. 16:10607-
10622.

Mullenders, L. H. F., van Zeeland, A. A., and Natarajan, A. T. 1987. The
localization of ultraviolet-induced excision repair in the nucleus and the
distribution of repair events in higher order chromatin loops in mammalian
cells. J. Cell Sci. (Suppl.) 6:243-262.

Mullis, K., Faloona, F., Scharf, S., Saiki, R., Horn, G., and Erlich, H.
1986. Specific enzymatic amplification of DNA in vitra: the polymerase
chain reaction. Cold Spring Harbor Symp. Duant. Biol. 51:263-273.

Myles, G. M., Van Houten, G., and Sancar, A. 1987. Utilization of DNA
photolyase, pyrimidine dimer endonuclease, and alkali hydrolysis in the
analysis of aberrant ABC exinuclease incisions adjacent to UV-induced DNA
photoproducts. Nucleic Acids Res. 15:1227-1243.

Naumovski, L. and Friedberg, E. C. 1986. Analysis of the essential and
excision repair function of the RA03 gene in Saccharamyces cerevisiae by
mutagenesis. Mal. Cell. Biol. 6:1218-1227.

Naumovski, L. and Friedberg, E. C. 1983. A DNA repair gene required for
the incision of damaged DNA is essential for viability in yeast. Proc.
Natl. Acad. Sci. (USA) 80:4818-4821.

Nelson, W. G. Pienta, K. J., Barrack, E, and Coffey, D. S. 1986. The role
of the nuclear matrix in the organization and function of DNA. Annu. Rev.
Biophys. Biophys. Chem. 15:457-475.

Niggli, H. J., and Cerutti, P. A. 1982. Nucleosomal distribution of
thymine photodimers following far- and near-ultraviolet irradiation.
Biochem. Biophys. Res. Commun. 105: 1215-1223.

Dkumoto, D. S. and Bohr, V. A. 1987. DNA repair in the metallothionein
gene increases with transcriptional activation. Nucleic Acids Res.
15:10021-10030.

Patterson, M. and Chu, G. 1989. Evidence that xeroderma pigmentosum cells
from complementation group E are deficient in a homolog of yeast
photolyase. Mol. Cell. Biol. 9:5105-5112.

48

Paulson, J. R. and Laemmli, U. K. 1977. The structure of the histone
depleted metaphase chromosomes. Cell 12:817-828.

Pettijohn, D. and Hanawalt, P. 1964. Evidence for repair-replication of
ultraviolet damaged DNA in bacteria. J. Mal. Bial. 9:395-410.

Pettijohn, D. and Hanawalt, P. 1963. Deoxyribonucleic acid replication in
bacteria following ultraviolet irradiation. BiachinL Biaphy; Acta. 72:127-
129.

Prakash, L. and Sherman, F. 1973. Mutagenic specificity reversion of iso-
1-cytochrome c mutants of yeast. J. Mal. Biol. 79:65-82.

Prelich, G., and Stillman, B. 1988. Coordinated leading and lagging strand
synthesis during SV40 DNA replication in vitra requires PCNA. Cell 53:
117-126.

Ramanathan, R., Rajalakshmi, S., Sarma, D. S. R., and Farber E. 1976.
Nanrandam nature of in viva methylation by dimethylnitrosamine and the
subsequent removal of methylated products from rat liver chromatin DNA.
Cancer Res. 36:2073-2079.

Razzaque, A., Chakrabarti, S., Joffee, S., and Seidman, M. 1984.
Mutagenesis of a shuttle vector plasmid in mammalian cells. Mal. Cell.
Biol. 4:435-441.

Razzaque, A., Mizusawa, H., and Seidman, M. M. 1983. Rearrangement and
mutagenesis of a shuttle vector plasmid after passage in mammalian cells.
Proc. Natl. Acad. Sci. (USA) 80:3010-3014.

Reed, W. B., Landing, 8., Sugarman, G., Cleaver, J. E., and Melnyk, J.
1969. Xeroderma pigmentosum. Clinical and laboratory investigation of its
basic defect. JAMA 207:2073-2079.

Robbins, J. H. 1978. Significance of repair of human DNA: evidence for
studies from xeroderma pigmentosum. J. Natl. Cancer Inst. 61:645.

Robbins, J. H., Kraemer, K. H., Lutzner, M. A., Festoff, B. W., and Coon,
H. G. 1974. Xeroderma pigmentosum. An inherited disease with sun
sensitivity, multiple cutaneous neoplasms and abnormal DNA repair. Ann.
Int. Med. 80:221-248.

Robbins, J. H. and Kraemer, K. H. 1972, Abnormal rate and duration of
ultraviolet-induced thymidine incorporation into lymphocytes from patients
with xeroderma pigmentosum and associated neurological complications.
Mutat. Res. 15:92-97.

Saiki, R. K., Gelfand, D. H., Stoffel, S., Scharf, S. J., Higuchi, R.,
Horn, G. T., Mullis, K. 8., and Erlich, H. A. 1988. Primer-directed
enzymatic amplification of DNA. with a thermostable DNA polymerase.
Science 239:487-491.

49

Sancar, A. and Rupp. N. D. 1983. A novel repair enzyme: UvrABC excision
nuclease of Escherichia coli cuts a DNA strand on both sides of the
damaged region. Cell 33:249-260.

Sancar, A. and Sancar G. B. 1988. DNA repair enzymes. Ann. Rev. Biochem.
57:29-67.

Satokata, 1., Tanaka, K., Miura, N., Miyamoto, I., Satoh, Y., Kondo, S.,
and Okada, Y. 1990. Characterization of a splicing mutation in group A
xeroderma pigmentosum. Proc. Natl. Acad. Sci. (USA) 87:9908-9912.

Sauerbier, W., and Hercules, K. 1978. Gene and transcription unit mapping
by radiation effects. Ann. Rev. Genet. 12:329-363.

Scicchitano, D. S. and Hanawalt, P. C. 1989. Repair of N-methylpurines in
specific DNA sequences in Chinese hamster ovary cells: absence of strand
specificity in the dihydrofolate reductase gene. Proc. Natl. Acad. Sci.
(USA) 86:3050-3054.

Seidman, M. 1989. The development of transient SV 40 based shuttle
vectors for mutagenesis studies: problems and solutions. Mutat. Res.
220:55-60.

Seidman, M. M., Dixon, K., Razzaque, A., Zagursky, R. J., and Berman, M.
L. 1985. A shuttle vector plasmid for studying carcinogen-induced point
mutations in mammalian cells. Gene 389:233-237.

Setlow, R. B. 1978. Repair deficient human disorders and cancer. Nature
271:713-717..

Setlow, R.B. and Carrier, W. L. 1964. The disappearance of thymine dimers
from DNA: an error-correcting mechanism. Proc. Natl. Acad. Sci. (USA)
51:226-231.

Shibutani, S., Takeshita, M., and Grollman, A. P. 1991. Insertion of
specific bases during DNA synthesis past the oxidation-damaged 8-oxodG.
Nature 349: 431-434.

Shiomi, T., Hieda-Shiomi, N., and Sato, K. 1982. Isolation of UV-sensitive
mutants of mouse LSI78Y cells by'a cell suspension spotting method. Somat.
Cell Genet. 8:329-345.

Simpson, 0., Crosby, R. M., and Skopek, T. R. 1988. A method for specific
cloning and sequencing of human HPRT cDNA for mutation analysis. Biochem.
Biaphy. Res. Commun. 151:487-492.

Skopek, T. R., and Hutchinson, F. 1982. DNA base sequence changes induced
by bromauracil mutagenesis of lambda phage. J. Mal. Biol. 159:19-33.

Smerdon, M. J. and Lieberman, M. W. 1978. Nucleosome rearrangements in
human chromatin during UV-induced DNA repair synthesis. Proc. Natl. Acad.
Sci. (USA) 75:4238-4241.

50

Smerdon, M. J. and Thoma, F. 1990. Site-specific DNA repair at the
nucleosome level in a yeast minichromosome. Cell 61:675-684.

Smith, C. A., and Mellon, I. 1990. Clues to the organization of DNA repair
systems gained from studies of intragenomic repair heterogeneity. In Obe,
G. (ed.), Advance in mutagenesis. Vol.1, pp. 153-194.Springer Verlag,
Berlin.

Smith, C. A. 1987. DNA repair in specific sequences in mammalian cells.
J. Cell Sci. (Suppl.) 6:225-241.

Snapka, R. M. and Linn S. 1981. Efficiency of formation of pyrimidine
dimers in SV40 chromatin in vitra. Biochemistry 20:68-72.

Stankowski, L. F., Jr., Tindall, K. R., and Hsie, A. W. 1986. Quantitative
and molecular analyses of ethyl methanesulfanate- and ICR 191-induced
mutation in A552 cells. Mutat. Res. 160:133-147.

Stefanini, M., Keyzer, W., Westerveld, A., and Bootsma, D. 1985.
Interspecies complementation analysis between xeroderma pigmentosum and
UV-sensitive Chinese hamster cells. Exp. cell Res. 161:373-380.

Sung, P., Prakash, L., Matson, S. W., Prakash, S. 1987b. RA03 protein of
Saccharamyces cerevisiae is a DNA helicase. Proc. Natl. Acad. Sci. (USA)
84:8951-8955.

Sung, P., Prakash, L., Weber, S., Prakash, S. 1987a. The RA03 gene of
Saccharamyces cerevisiae encodes a DNA-dependent ATPase. Proc. Natl. Acad.
Sci. (USA) 84:6045-6049.

Tanaka, K., Miura, N., Satokata, I., Miyamoto, I., Yoshida, M. C., Satoh,
Y., Kondo, S., Yasui, A., Okayama, H., and Okada, Y. 1990. Analysis of a
human DNA excision repair gene involved in group A xeroderma pigmentosum
and containing a zinc-finger domain. Nature 348:73-76.

Tanaka, K., Satokata, I., Ogita, Z., Uchida, T., and Okada, Y. 1989.
Molecular cloning of a mouse DNA repair gene that complements the defect
of group-A xeroderma pigmentosum. Proc. Natl. Acad. Sci. (USA) 86:5512-
5516.

Tang, M., Bohr, V. A., Zhang, X., Pierce, J., and Hanawalt, P. C. 1989.
Quantification of aminoflourene’adduct formation and repair in defined DNA
sequences in mammalian cells using the UvrABC nuclease. J. Biol. Chem.
264:14455-14462.

Thaker, J. 1985. The molecular nature of mutations in cultured mammalian
cells: A review. Mutat. Res. 150:431-442.

Thoma, F., Kaller, T. H., and Klug, A. 1980. Involvement of histone H1 in
the organization of the nucleosome and of the salt dependent
superstructures of chromatin. J. Cell Biol. 83:403-427.

51

Thomas, D. C., Morton, A. G., Bohr, V. A., and Sancar A. 1988. General
method for quantifying base adducts in specific mammalian genes. Proc.
Natl. Acad. Sci. (USA) 85:3723-3727.

Thomas, D. C., Dkumoto, D. S., Sancar A., and Bohr, V. A. 1989.
Preferential DNA repair of (6-4) photoproducts in the dihydrofolate
reductase gene of Chinese hamster ovary cells. J. Biol. Chem. 264:18005-
18010.

Thomas, J. O. and Butler, P. J. G. 1980. Size dependence of a stable
higher order structure of chromatin. J. Mol. Biol. 144:89-93.

Thompson, L. H., Busch, 0. 8., Brookman, K., Mooney, C. L., and Glaser, D.
A. 1981. Genetic diversity of UV-sensitive DNA repair mutants of Chinese
hamster ovary cells. Proc. Natl. Acad. Sci. (USA) 78:3734-3737.

Thompson, L. H., Mitchell, 0. L., Regan, J. 0., Bouffler, S. 0., Stewart,
S. A., Carrier, W. L., Nairn, R. S., and Johnson, R. T. 1988b. CHO mutant
UV61 removes (6-4)photoproducts but not cyclobutane dimers. Mutagenesis
4:140-146.

Thompson, L. H., Mooney, C. L., and Brookman, K. W. 1985a. Genetic
complementation between UV-sensitive CHO mutants and xeroderma pigmentosum
fibroblasts. Mutat. Res. 150:423-429.

Thompson, L. H., Mooney, C. L., Burkhart-Schultz, K., Carrano, A. V., and
Sicilliano, M. J. 1985b. Correction of a nucleotide-excision-repair
mutation by human chromosome 19 in hamster-human hybrid cells. Somat. Cell
Mol. Genet. 11:87-92.

Thompson, L. H., Rubin, J. S., Cleaver, J. E., Whitmore, G. F., and
Brookman, K. 1980. A screening method for isolating DNA repair-deficient
mutants of CHO cells. Somat. Cell Genet. 6:391-405.

Thompson, L. H., Shiomi, T., Salazar, E. P., and Stewart, S. A. 1988a. An
eighth complementation group of rodent cells hypersensitive to ultraviolet
radiation. Sbmat. Cell Mol. Genet. 14:605-612.

Tindall, K. R., Stankowski, L. F., Jr., Machanoff, R., and Hsie, A. W.
1986. Analyses of mutation in pSVngt-transformed CHO cells. Mutat. Res.
160:121-131.

Tindall, K. R., Stankowski, L. F., Jr., Machanoff, R., and Hsie, A. W.
1984. Detection of deletion mutations in pSV2gpt-transformed cells. Mal.
Cell. Biol. 4: 1411-1415.

Troelstra, C., 0dijk, H., de Wit, J., Westerveld, A., Thompson, L. H.,
Bootsma, D., and Hoeijmakers, J. H. J. 1990. Molecular cloning of the
human DNA excision repair gene ERCC-6. Mal. Cell. Biol. 10:5806-5813.

Troelstra, C., 0dijk, H., de Wit, J., Westerveld, A., Thompson, L. H.,
Bootsma, 0., and Hoeijmakers, J. H. J. 1991. Cloning and functional

52

properties of the human excision repair gene ERCC-6. In: Narkshap on DNA
Repair with Emphasis on Eukaryotic Systems. p. 18.30, Noordwijkerhout, The
Netherlands.

Tsoa, Y. P., Wu, H.-Y., and Liu, L. F. 1989. Transcription-driven
supercoiling of DNA: direct biochemical evidence from in vitra studies.
Cell 56:111-118.

van Duin, M., de Wit, J., 0dijk, H., Westerveld, A., Yasui, A., Koken, M.
H. M., Hoeijmakers, J. H. J., and Bootsma, D. 1986. Molecular
characterization of the human excision repair gene ERCC-I: cDNA cloning
and amino acid homology with the yeast DNA repair gene RADlO. Cell 44:913-
923.

van Duin, M., van der Tol, J., Warmerdam, P., 0dijk, H., Meijer, 0.,
Westerveld, A., Bootsma, 0., and Hoeijmakers, J. H. J. 1988. Evolution and
mutagenesis of the mammalian excision repair gene ERCC-I. Nucleic Acids
Res. 16:5305-5322.

van Duin, M., Vredeveldt, G., Mayne, L. V., 0dijk, H., Vermeulen, W.,
Klein, 8., Weeda, G., Hoeijmakers, J. H. J., Bootsma, D., and Westerveld,
A. 1989. The cloned human DNA excision repair gene ERCC-l fails to correct
xeroderma pigmentosum complementation groups A through I. Mutat. Res. 217:
83-92.

Van Houten, B. 1990. Nucleotide excision repair in Escherichia coli.
Microbial. Rev. 54:18-51.

Venema, J., Mullenders, L. H. F., Natarajan, A. T., van Zeeland, A. A. and
Mayne, L. V. 1990b. The genetic defect in Cockayne syndrome is associated
with a defect in repair of UV-induced DNA damage in transcriptionally
active DNA. Proc. Natl. Acad. Sci. (USA) 87: 4707-4711.

Venema, J., van Hoffen, A., Natarajan, A. T., van Zeeland, A. A., and
Mullenders, L. H. F. 1990a. The residual repair capacity of xeroderma
pigmentosum complementation group C fibroblasts is highly specific for
transcriptionally active DNA. Nucleic Acid Res. 8: 443-448.

Vrieling, FL, van Roaijen, M. L., Groen, N. A., Zdzienicka, M. 25,
Simons, J. W. I. M., Lohman, P. H. M., and van Zeeland, A. A. 1989. DNA
strand specificity for UV-induced mutations in mammalian cells. Mal.
Cell. Biol. 9:1277-1283.

Weber, C. A., Salazar, E. P., Stewart, S. A., Thompson, L. H. 1988.
Molecular cloning and biological characterization of a human gene, ERCC-
2, that corrects the nucleotide excision repair defect in CHO UV5 cells.
Mal. Cell. Biol. 8:1137-1146.

Weber, C. A., and Salazar, E. P. 1991. Characterization of ERCC-2, a human
nucleotide excision repair gene. In: Workshop on DNA Repair with Emphasis
on Eukaryotic Systems. pp.18.33, Noordwijkerhout, The Netherlands.

53

Weber, C.A., Salazar, E. P., Stewart, 5. A., Thompson, L. H. 1990. ERCC2:
cDNA cloning and molecular characterization of a human nucleotide excision
repair gene with high homology to yeast RA03. EMBO J. 9: 1437-1447.

Weeda, G., van Ham, R. C. A., Vermeulen, W., , Bootsma, D., van der Eb, A.
J., and Hoeijmakers, J. H. J. 1990b. A presumed DNA helicase encoded by
ERCC-3 is involved in the human repair disorders xeroderma pigmentosumnand
Cockayne’s syndrome. Cell 62: 777-791.

Weeda, G., van Ham, R. C. A., Masurel, R., Westerveld, A., 0dijk, H., de
Wit, J., Bootsma, D., van der Eb, A. J., and Hoeijmakers, J. H. J. 1990a.
Molecular cloning and biological characterization of the human excision
repair gene ERCC-3. Mol. Cell. Biol. 10:2570-2581.

Westerveld, A., Hoeijmakers. J. H. J., van Duin, M., de Wit, J., 0dijk,
H., Pastink, A., Wood, R. 0., and Bootsma, D. 1984. Molecular cloning of
a human DNA repair gene. Nature 310: 425-429.

Williams, J. I. and Friedberg, E. C. 1979. DNA excision repair in
chromatin after UV irradiation of human fibroblasts in culture.
Biochemistry 18:3965-3972.

Wood, R. 0., Robbins P., and Lindahl, T. 1988. Complementation of the
xeroderma pigmentosum DNA repair defect in cell-free extracts. Cell 53:
97-106.

Wood, R. D., Skopek, T. R., and Hutchinson, F. 1984. Changes in DNA base
sequence induced by targeted mutagenesis of lambda phage by ultraviolet
light. J. Mal. Biol. 175: 273-291.

Wu, H. Y., ShYY. S., Wang, J. C., and Liu, L. F. 1988. Transcription
generates positively and negatively supercoiled domains in the template.
Cell 53:433-440.

Yang, J.-L., Maher, V. M., and McCormick, J. J. 1989. Amplification and
direct nucleotide sequencing of cDNA from the lysate of low numbers of
diploid human cells. Gene 83:347-354.

Yang, L. L., Maher, V. M., MSCormick, J. J. 1980. Error-free excision of
the cytotoxic, mutagenic N deoxyguanosine DNA adduct formed in human
fibroblasts by (:)-7B,8a-dihydroxy-9a,10a-epoxy-7,8,9,10-tetrahydrobenzo-
[a]pyrene. Proc. Natl. Acad. Sci. (USA) 77:5933-5937.

Yanofsky, C. 1971. In Chemical Mutagens, Hollaender, A. ed., vol. 1, pp.
283-287. New York: Plenum Press.

Yeung, A. T., Mattes, W. 8., 0h, E. Y., and Grossman, L. 1983. Enzymatic
properties of purified Escherichia coli UvrABC proteins. Proc. Natl. Acad.
Sci. (USA) 80:6157-6161.

Zakour, R. A., Schaaper, R. M., and Glickman, B. W. 1986. Introduction,
rescue and expression of plasmid genes in mammalian cells and Escherichia

54
coli. Mutat. Res. 163:3-13.

Zdzienicka, M., van der Schans, I. P., and Simons, J. W. I. M. 1988.
Identification of a new seventh complementation group of UV-sensitive
mutants in Chinese hamster cells. Mutat. Res. 194: 165-170.

Zolan, M. E., Cortopassi, G. A., Smith, C. A., and Hanawalt, P. C. 1982.
Deficient repair of chemical adducts in a DNA of monkey cells. Cell
28:613-619.

Zolan, M. E., Smith, C. A., and Hanawalt, P. C. 1984. Formation and repair
of furocoumarin adducts in a DNA and bulk DNA of monkey cells.
Biochemistry 23:63-69.

CHAPTER II

Effect of excision repair by diploid human fibroblasts on the kinds and
spectra of mutations induced by (t)-7B,8a-dihydroxy—9a,10a-
epoxy-7,8,9,10-tetrahydroben20(a)Pyrene in the coding
region of HPRT gene

Ruey-Hwa Chen, Veronica M. Maher and J. Justin McCormick

Carcinogenesis Laboratory - Fee Hall
Department of Microbiology and Department of Biochemistry

Michigan State University, East Lansing MI 48824-1316 (U.S.A.)

55

SUHHARY

To investigate the effect of nucleotide excision repair on
mutations induced in diploid human fibroblasts by (i)-7B,8a-dihydroxy-
9a,10a-epoxy-7,8,9,10-tetrahydrobenzo(a)pyrene (BPDE), a direct-acting
carcinogen that forms bulky adducts predominantly (>90%) with guanine, we
synchronized cells, exposed them to BPDE at the beginning of S-phase or in
early G,-phase, selected 6-thioguanine-resistant cells, and determined the
nature of the mutations induced in the coding region of the hypoxanthine
(guanine) phosphoribosyl- transferase (HPRT) gene. The mutant frequency
in the populations treated in S was high (30 to 50 times background); for
the G, cells, which had at least 12 hr to excise BPDE adducts before S-
phase DNA replication, the frequency was 3-fold lower. To identify the
mutations, the mRNA in lysates of 100 to 500 cells from each individual
mutant clone was copied into cDNA, and the cDNA of the HPRT gene was
amplified by polymerase chain reaction and sequenced directly from the
product. The majority of the mutations in either population (20/24 and
19/23) consisted of base substitutions, predominantly G-C to T-A. But
their location differed significantly, with one very strong "hot spot"
found only in the G, cells. Assuming that the premutagenic lesion involves
guanine, the strand distribution of the premutagenic lesions also differed
significantly. In the cells treated in S-phase, 21% of such lesions were
located in the transcribed strand, whereas in the G, phase-treated cells,
none were. This suggests that in the HPRT gene of diploid human cells,

excision repair of BPDE adducts occurs preferentially on the transcribed

56

strand.

57

INTRODUCTION

In previous studies, we and our colleagues investigated the

relationship of DNA repair and DNA replication to mutagenesis induced by
carcinogens in diploid human fibroblasts. We found that
the frequency of mutants induced by UV radiation (1, 2), (:)-7B,8a-
dihydroxy-9a,10a-expoxy-,8,9,10-tetrahydrobenzo[a]pyrene (BPDE) (3,
4), or 1-nitroso-pyrene (5) in nucleotide excision repair-proficient cells
is highest in cells treated at early S-phase and much lower in cells
treated in early G,-phase. “The frequency per dose is higher still in cells
from xeroderma pigmentosum (XP) patients that are virtually incapable of
nucleotide excision repair i.e., and mutation induction in these cells is
not cell cycle-dependent (2, 3). These data suggest that repair processes
play a key role in the cell cycle-related sensitivity to mutagenicity
observed in synchronized repair-proficient human fibroblasts, and that DNA
replication is centrally involved in the conversion of DNA damage into
mutations.

The effect of the cell cycle on the frequency of transformation has
also been determined in diploid human cells (6) and rodent cells lines
(7). Populations treated in early S-phase exhibit a higher frequency of
transformed cells than populations exposed in early G,-phase (6, 7). This
difference was not seen with XP cells (6). A higher frequency of
transformants might simply reflect a higher frequency'of mutations induced
in populations treated at the beginning of S-phase. However, it might

result from a difference in the kinds or spectrum of mutations induced in

58

59

the two populations, so that a specific mutation required for
transformation, e.g., leading to the activation of a cellular proto-
oncogene, occurs preferentially' when cells are treated in S-phase.

There are no published studies that have determined whether cells
treated in S-phase and cells treated in G, differ in the type and/or
locaticni of mutations induced. However, Hanawalt, Bohr, and their
colleagues showed that in certain mammalian cell lines the rate of
nucleotide excison repair of pyrimidine dimers in transcriptionally active
genes differs from that occurring in non-actively-transcribed genes (8-
11). Mellon et al. (12) showed that in a Chinese hamster cell line and in
human cells, the rate of excision repair of damage in the transcribed
strand of the gene for dihydrofolate reductase (dhfr) is more rapid than
in the non-transcribed strand. If such differential repair occurs in the
HPRT gene of diploid human cells, and if mutations are introduced by the
DNA replication which occurs in S-phase, significant differences should be
seen between the mutational spectrum of synchronized cells exposed to DNA
damaging agents just as S-phase DNA replication of a target gene is taking
place or exposed in early G,. To test this hypothesis, we treated repair-
proficient human fibroblasts with BPDE in early S-phase or in early G,-
phase, and determined the kinds and location of mutations induced in the
coding region of the hypoxanthine (guanine) phosphoribosyl-transferase
(HPRT) gene. The results are consistent with the hypothesis. The two
populations showed no difference in the kinds of mutations induced by BPDE

but the location of the mutations differed significantly.

MATERIALS AND METHODS

Cells and media. Diploid human cells derived from neonatal
foreskin material (13) were cultured in modified McM medium (14) prepared
with Earle’s salts and containing 10% supplemented calf serum (Hyclone,
Logan UT) (culture medium). For selection of 6-thioguanine-resistant
cells, the same medium, but lacking adenine and containing 5% supplemented
calf serum (Hyclone), 5% fetal calf serum, and 40pM 6-thioguanine was
used.

Cell synchronization. Cells were inoculated at a density 8- to 10-
fold less than that attained at confluence and re-fed culture medium every
other day to stimulate rapid growth to confluence. When the cells reached
confluence, they were fed daily for 3 additional days to allow them to
fill the dish to capacity. The medium was not changed for the next 72 to
96 hr so that the cells would cease proliferation and enter the Go state.
Autoradiography studies showed that in such cultures DNA synthesis was
virtually eliminated (2). To stimulate the cells to enter the cell cycle,
they were plated in culture medium at a density of 10‘ cells/cmz. Previous
studies show that such cells begin DNA synthesis (S-phase) after 16 to 17
hours (4), and that at least 80% of the cells are at S-phase after 4
additional hours (2).

Exposure to mutagen and isolation of HPRT mutants. The details of
the mutagenesis assay have been described (4,5). Briefly, a series of
populations (1.5 x 10‘5 cells each, plated in 150 min-diameter dishes) was
treated 17 hr after release from G0 (at the beginning of S-phase) or 4 hr

after release from G0 (in early G,-phase). The culture medium was replaced

60

61

with serum-free medium. BPDE, freshly dissolved in anhydrous
dimethylsulfoxide, was delivered into the serum-free medium by
micropipette. After 1 hr, the medium was removed, and the cells were
rinsed with phosphate buffered saline (pH 7.4) and re-fed with fresh
culture medium. The cells in one dish were assayed immediately for
survival of colony-forming ability as described (15). The rest were
allowed 8 days for expression of resistance to 6-thioguanine before 0.5
x 10‘ cells from each population were plated in selective medium. When

macroscopic drug resistant clones developed 14 days later, these were

located, isolated, and expanded to 10‘ to 5 x 105 cells.

Synthesis of first-strand cDNA directly from mRNA in cells. Cells
were trypsinized and suspended in culture medium, the cell number was
determined by electronic counting, and the cells were diluted in cold
phosphate buffer saline (pH 7.4). From 100 to 500 cells were transferred
to a 0.5 ml Eppendorf tube and centrifuged for 10 min at 4°C. The
supernatant was removed, and the cell pellet was resuspended in 5 pl of
the cDNA cocktail described in detail by Yang et al (16). The reverse
transcriptase reaction was performed at 37°C for 1 hr to allow the cell
membranes to be lysed by detergent and first-strand cDNA to be synthesized

from total cytoplasmic poly A mRNA.

Amplification of HPRT cDNA and DNA sequencing. The experimental
conditions, optimized for preparing second-strand HPRT cDNA, amplifying
the cDNA 10"-fold, using two 30-cycles of polymerase chain reaction

sequencing the product directly using the three sequencing primers and a

62

modified Sanger dideoxynucleotide procedure have been described (16).

RESULTS

Cytotoxicity and frequency of mutants induced by BPDE in cells treated
in the 6,-phase or S-phase of the cell cycle. Individual populations of
cells were synchronized and exposed to various concentrations of BPDE in
early S-phase or in early G,. The»cells were assayed for survival, and for
the frequency of 6-thioguanine resistant cells (Fig. 1). As expected (3),
there was little difference in the sensitivity of the two populations to
the cytotoxic effects of BPDE indicating that the initial level of DNA
adduct formation in the two populations was comparable (17). In contrast,
the frequency of mutants induced in populations treated in early G,-phase
was 3-fold lower than that of cells treated in early S-phase. These
results support the hypothesis that S-phase DNA replication is responsible
for converting premutagenic BPDE-induced lesions into nmtations since
cells treated in G,-phase will be able to excise a higher percentage of
BPDE-induced adducts than will cells treated in S-phase. It should be
pointed out that in each experiment, the frequency of 6-thioguanine-
resistant cells in the populations treated at the onset of S was 30 to 50
times higher than background; in the cells treated in the G,-phase, it was
10 to 16 times higher than background. Therefore, the chances of
including a mutant containing a spontaneous mutation, rather than a BPDE-
induced mutation, in the samples analyzed at the sequence level was very
low.

Characterizaton of HPRT mutants. To determine the nature of the
BPDE-induced mutations at the DNA sequence level, 20 drug-resistant clones

were isolated from cells treated in S-phase and 20 from cells treated in

63

64

Figure 1. Cytotoxicity (upper panel) and frequency of mutations to 6-
thioguanine resistance (lower panel) induced by BPDE in human cells
treated in early S-phase (0) or 12 hr prior to the onset of S (0). The
frequencies have been corrected for the cloning efficiency (40% to 60%) of
the cells at the time of selection. The background frequencies which
ranges from (5 x 10“ to 15 x 10") have been subtracted. Line drawn by

the method of least squares.

65

 

BPDEULM)
0.2

0.!

 

_

 

 

 

_ _ _

 

 

 

_ _
6 O O

O O
3 2

_ _ _ _:
O O O O O
.fllu 6 4 2 I.

>._._T:m< 022040

J<>_>m3m Hzmomma m.._i_wo L0H owUDQZ

0.2
BPDE (DIV!)

O.|

mjmo 3320.6 80. En.

Figure 1

66

G,. The cells were tested for inability to grow in medium containing
hypoxanthine (10“ M), aminopterin (2 x 10'6 M), and thymidine (3 x 10'5 M),
in order to confirm that they lacked a functional HPRT gene. Then cDNA
was synthesized from the total mRNA of a lysate of 100 to 500 cells from
each clone, and the cDNA of the HPRT gene was amplified Ion-fold and
sequenced directly. The results are shown in Tables 1 and 2. All 20
mutants listed from each phase represent unequivocally independent
mutants, either because their alterations were unique or because they were
derived from independent populations of treated cells.

There was no significant difference between G,- and S-phase-derived
mutants in the kinds of mutations they contained (Table 1 and 2). 16
mutants out of the 20 from each phase contained base substitutions, with
1 mutant from each phase exhibiting a second base substitution at a site
distant from the first, and with 1 mutant from G,-phase cells exhibiting
a tandem base substitution. 0f the remaining 4 mutants from each phase,
3 may also have contained base substitutions in a 5’ or 3’ consensus
splice site, since one of the nine exons was completely deleted from their
cDNA, but no other change was seen. The fourth G,-phase mutant contained
an A-T base pair insertion, and the fourth S-phase mutant exhibited a G-C
base pair deletion.

As shown in Tables 1 and 2, in both populations the BPDE-induced
mutations were not randomly distributed throughout the 9 exons of the
coding region of the gene. The majority (16/24 for G,-phase and 13/23 for
S-phase) were located in exon 2 or the first half of the exon 3. Table 3
shows that there also was no significant difference in the kinds of base

substitutions observed in the mutants obtained from the two populations.

67

Table 1. Kinds and locations of the mutations induced in the coding region of the HPRT gene in cells treated with
BPDE in S phase of the cell cycle

 

 

Surrounding Strand with
Type of sequence (coding Amino acid the affected
Mutant Position Exon mutation strand)‘ change purine
Base substitution
BPSI6+ 88 2 CC -. T-A GCT CAG GAT Glu - Stop NT
BPSS7 97 2 CC -r T-A TTG QAA AGG Glu -. Stop NT
BPSZB 119 2 CC - T-A CAT GQA CT A Gly —' Val NT
BPSSI 119 2 6C - T-A CAT GQA CTA Gly _. Val NT
BPS-i 130 2 CC - T-A ATG CAC AGG Asp —- Tyr NT
BPS66 134 2 CC - T-A GAC AGG ACT Arg .... Met NT
BPS3 149 3 G-C - T-A C'I'I‘ GCT CGA Ala _. Asp T
BPSI6T I78 3 DC —> T-A GGC CAT CAC His —- Asn T
BPSZI 178 3 6C - T~A GGC CAT CAC I-Iis -. Asn T
BPSZ3 178 3 CrC -r T-A GGC CAT CAC i-Iis -o Asn T
BPS94 212 3 CC ~ T-A AAG GGG GQC Gl}I —. Val NT
BPS47 229 3 0C - T-A GCT CAC CT G Asp -. Tyr NT
BPS34 337 4 GC - T'A GGG CAC ATA Asp - Tyr NT
BPS78 438 6 6C - C~G ACT TIC CTT Leu —. Phe NT
BPSB3 447 6 CC - C-G TCC 'I'I‘C GTC Leu _. Phe NT
BPS75 451 6 AT ~ G-C GTC .566 CAG Arg -- Gly NT
BPSI3 471 6 (3C - T-A AAG ATC GTC Met —- Ile NT
BPS62 478 6 CC —. A-T AAG CTC GCA Val - Ile NT
BPS9S 531 7 0C - T-A CCA GAC 'ITI‘ Asp -. Glu T
BPS41 568 8 CC -o T-A GTA QGA TAT Gly -- Stop NT
Putative splice site mutation
BPS72 28—134 2 Exon 2 missing NA
BPS32 319-384 4 Exon 4 missing NA
BPSG9 486-532 7 Exon 7 missing NA
Other
BPS37 477 or 478 6 Deletion of a G AAQCTC GCA Frameshift NT

 

T. transcribed: NT. nontranscribed: NA. not applicable.
‘The base that was altered is underlined.
iMutant with two base substitutions.

68

Table 2. Kinds and locations of the mutations induced in the coding region of the HPRT gene in cells treated with
BPDE in 01 phase of the cell cycle

 

 

Surrounding Strand with
Type of sequence (coding Amino acid the affected
Mutant Position Exon mutation strand)’ change purine
Base substitution
BPOZZ 88 2 OC -' T'A OCT CAO OAT Olu —» Slap NT
BPO7 97 2 O-C —- T-A ITO CAA AOO Olu —- Stop NT
BPO113 130 2 03C —- T~A ATO CAC AGO Asp -~ Tyr NT
BPG29 134 2 O-C —. T-A OAC ACO ACT Arg —' Met NT
BPGIO9 I39 3 CC —. A'T ACT CAA COT Glu -o Lys NT
BPO107 208 3 OC _. A'T AAO COO OOC Oly _. Arg NT
BPGIOI 211 3 OC _. GO AAO GOO COC Oly -» Arg NT
BPOI9 211 3 O-C —' T-A AAO OOO COC Oly - Cys NT
BPOIZ 212 3 OC —- T’A AAO OOO OCC Oly —- Val NT
BPOI3 212 3 OC —- T-A AAO GOO OCC Oly - Val NT
BPG44 212 3 OC -> A-T AAO OOO OCC Oly —- Asp NT
BPO64 212 3 0C —. T-A AAO OOG GCC Oly - Val NT
BPOTT 229 3 OC -- T~A OCT CAC CTG Asp —. Tyr NT
BPG48 229 3 CC - T'A OCT CAC CTO Asp -- Tyr NT
BPO8T 388 5 CC - C-O AAT CTC ITO Val ~ Leu NT
BPOTG 393 5 OC - GO OTC ITC AIT Leu —- Phe NT
BPOIB 419 6 OC —» T-A ACT OCC AAA Oly -. Val NT
BPO8+ 529 7 OC -- T-A CCA CAC I'IT Asp _. Tyr NT
BPO112 535 8 OC -. T-A ITT CIT GOA Val -' Phe NT
Putative splice site mutation
BPOID 28-134 2 Exon 2 missing NA
BPOllb 319-384 4 Exon 4 missing NA
BPG9 403-485 6 Exon 6 missing NA
Other
BPOI7 98-100 2 insertion of an A TOO GALAOG Frameshit't NT
BPGIOO 535 8 Deletion of a G I'IT CIT OOA Frameshift NT

 

T. transcribed: NT. nontranscribed: NA. not applicable.
‘The base that was altered is underlined.
iMutant with two base substitutions.

69

 

Table 3 Types of base swstitutions in hprt
Iutants cells treated with BPDE in 6,-phase and S-phase

 

 

Type of WW5!
base substitutions 6,-pbase S-phase
Transversions:
60C -v T-A 13 15
(sec 4 as 3 2
A-T -> COG 0 1
ToA _. a.c 1 0
Transitions:

G-c - A-T 2 1
T-A .. 05 1 0

 

Total 20 19

70
All except one for an S-phase-derived mutant and two for Gl-phase mutants,
involved G-C base pairs. Transversion of G-C to T-A occurred
predominantly, viz.,13/20 (65%) for G1-phase and 15/19 (79%) for S-phase.

Strand distribution of the prelutagenic lesions induced by BPDE.
Fig. 2 shows the location (spectrum) of mutations induced by BPDE in cells
treated in G,- and S-phase. BPDE binds DNA predominantly to the N2
position of guanine (18, 19). If'one assumes that the premutagenic lesion
in these cells involves guanine, it is possible to distinguish the DNA
strand in which the premutational lesion was located. In the population
exposed in S-phase, of the 19 mutations which involved G-C base pairs, 4
affected guanine bases (21%) were located on the transcribed strand, and
15 (79%) were located on the nontranscribed strand (Table 2 and Fig. 2).
However, in cells exposed in early Gf-phase, in all 18 cases, the affected
guanine bases were located on the nontranscribed strand. ‘The strand
distribution of guanine bases involved in mutations induced in early G,-
phase differed significantly from that involved in mutations induced at S—
phase mutants (P, < 0.05, by the Chi square test).

Mutational "hot spot” region obtained fraa HPRT lutants in cells
exposed in 6,—phase. One very strong mutational "hot spot" region was
found in the spectrum of mutations induced in cells treated in early G,-
phase (Fig. 2). Eight out of 24 mutations (33%) occurred within positions
208 to 212 (Table 1) in a region containing 6 consecutive guanine residues
flanked by adenine on the 5’ side and cytosine on the 3’ side, a sequence
which appears only once in the entire HPRT coding region. In contrast,
only one out of 23 mutations (4%) induced in cells treated at S-phase

occurred in this region (Table 2).

71

Figure 2. The location of mutations and the strand distribution of the
guanine base involved in a base change induced by BPDE in the coding
region of HPRT gene. The position of the ATG start codon, TAA stop codon,
and the 9 exons are indicated. (Vertical lines, base substitutions;

square; a deleted base' pair; caret, an inserted base pair; shaded

rectangles, deleted exons.)

72

N Guava

mmmIHo I; I;
02595

 

 

 

omeomZ<thOZ
oz<m._.m

wmazaia

 

ommEUmzn‘E.

mmmIHO
073me

 

 

 

owm_mUmZ<m._.ZOZ _ Ga a _ _
oz<mhm

mdedim

 

ommEUmza‘mH _

p.23 __m_ e ._ a _m_ a __
hmm 00w OOm OOV COM

“.1

OO.

_ —O.P< "ZOxm

.rmaa mmam

DISCUSSION

Although the mutant frequency in cells treated 12 hr prior to S-
phase was 3 times lower than that in cells treated in early S-phase, the
kinds of mutations induced were similar. Our finding that BPDE causes
base substitutions mainly at G-C base pairs suggests that mutagenesis is
targeted to adducts. The majority of the base substitutions were G-C to
T-A transversions, which agrees with the results observed in the lac I
gene of E. coli (20), in an extrachromosomal supF gene (21), and in
several endogenous genes of rodent cell lines (22, 23). The predominance
of G-C to T-A substitutions may be the result of the DNA polymerase
preferentially inserting an adenine nucleotide opposite a non-
instructional base containing a bulky BPDE adduct (24). Alternatively,
the presence of a BPDE adduct on guanine may allow purine-purine base
pairing to occur with some frequency and if this mispairing is not
recognized, the result would be a mutation (20). Evidence of stable
purineopurine base pairing in a model oligonucleotide has recently been
provided (25, 26). Also, Norman et al. (27) have provided evidence that
in a double-stranded oligonucleotide, stable base pairing can occur
between adenosine and guanosine carrying a polycyclic aromatic adduct in
the 08 position.

0f the 20 mutants derived from either phase of the cell cycle, 3
produced HPRT mRNA which lacked one of the nine exons completely. These
mutants may be mutated in the 5’ or 3' consensus splice site so that the
splicing machinery cannot recognize the site, and splicing occurs using a

nearest neighbor splice site. Splicing mutants resulting in exon skipping

73

74
have been found in mutants induced in the dhfr gene by UV-radiation or N-
acetoxy-Z-acetylaminofluorene (28, 29), as well as in UV-induced hprt
mutants (30) in CHO cells.

Mellon et al (12) recently reported a significant difference in the
efficiency of removal of UV-induced pyrimidine dimers from the two strands
of theiDHFR gene in cultured hamster and human cells, with the transcribed
strand being selectively repaired in both species. Vrieling et a7 (30)
investigated UV-induced hprt mutants in a repair-proficient and deficient
hamster cell line and proposed preferential repair of the transcribed
strand as one explanation for the strand specificity of the mutations
observed in the repair-proficient cells. Little information has been
available on differential or strand-specific repair of DNA damage induced
by BPDE. However, if one assumes that the mutations we observed that
involved a G-C base pair were targeted to a BPDE guanine adduct, the
strand distribution of the premutagenic lesions induced in two populations
differed significantly. Our data suggest that BPDE adducts are
preferentially removed from the transcribed strand of HPRT gene of these
human cells, so that at the time the unexcised adducts are converted into
mutations, more adducts remain on the nontranscribed strand than on the
transcribed strand. Our data further suggest that DNA replication is
responsible for converting premutagenic lesions into mutations since it is
the cells treated in G1-phase that have the most time to repair before S-
phase DNA replication takes place. It seems highly unlikely, that the
strong strand-preference we observed for the mutations that occurred in
the cells treated in G1-phase could have resulted from preferential binding

of BPDE to the transcribed strand of the HPRT gene in those cells. To

75
distinguish between strand-specific DNA repair and differential DNA strand
binding as an explanation for the observed strand-specificity of
mutations, we will carry out similar experiments in excision repair-
deficient XP cells.

If DNA synthesis is the cellular event converting the BPDE adducts
into mutations, cells treated in S-phase should have little or no time to
excise BPDE adducts before replication. Therefore, in the S-phase cells,
we expected little strand bias for premutational lesions above that caused
by the distribution of guanines in the coding region of the gene, i.e.,
160 guanines in the nontranscribed strand, but only 110 in the transcribed
strand, a 59:41 ratio. However, only 21% of mutations in S-phase cells
corresponded to a guanine located in the transcribed strand. A likely
reason why this value was less than 41% is that there was time for
excision repair. Since treatment of cells with BPDE temporarily stops DNA
replication (31), by the time it resumes, some fraction of the adducts may
have been excised. Furthermore, since our method of synchronizing cells
does not yield 100% cell synchrony, some of cells may not have reached S-
phase at the time they were exposed to BPDE, and so they had some time for
repair.

Mazur and Glickman (22) showed that the "hot spot" region for
mutations induced by BPDE in the adenine phosphoribosyltransferase gene of
CHO cells involved a 5’-AG(G)dA-3’ sequence. In the coding region of the
human HPRT gene, there are 9 such sequences which have not been
interrupted by splicing junctions. However, analysis of our sequence
context data indicates that only one'mutation observed in cells treated in

G,-phase and two in cells treated in S-phase occurred in these sites. In

76
contrast, our study revealed a cell-cycle dependent "hot spot" in the
sequenceiof 5’-A(G)6C-3’ (position 206-213). Eight out of the»24 mutations
induced in G1-phase (33%) were located in this region, compared with 1 out
of 23 from S-phase cells (Fig. 2 and Tables 1 and 2). Although it has
been reported that certain poly(dG) sequences are the preferred sites for
BPDE binding (32, 33), preferential binding cannot explain the cell-cycle
dependent feature of this very prominent "hot spot" which we observed,
since it was not similarly prominent for the S-phase population. It is
more likely that this "hot spot" resulted from inefficient repair of BPDE
adducts in the run of 6 consecutive guanine residues, so that after
excision repair had occurred for a certain period of time, the adducts
remaining in this region represented a higher fraction than originally.
Support for this hypothesis comes from several studies showing that the
helix of poly (dG).poly (dC) prefers an A-like DNA helix geometry, rather
than the traditional 8 form (34, 35). If this run of G’s assumes a
specific local DNA conformation, this could perhaps diminish the
efficiency of excision repair of the BPDE adducts in this region. If
inefficient repair accounts for the "hot spot" seen only in the G1 cells,

it should not be a prominent "hot spot" in XP cells.

ACKNOHLEDGEHENTS

This research was supported in part by DHHS Grant CA21253 from the

National Cancer Institute. The excellent technical assistance of Neil

Davis and Ena Zaccagnini is gratefully acknowledged.

77

10.

11.

12.

13.

14.

15.

16.

REFERENCES

Maher, V. M., Dorney, D. J., Mendrala, A. L., Konze-Thomas, 8., and
McCormick J. J. (1979) Mutat. Res., 62, 311-323.

Konze-Thomas, 8., Hazard, R. M., Maher, V. M., and McCormick, J. J.
(1982) Mutat. Res., 94, 421-434.

Yang, L. L., Maher, V. M., and McCormick, J. M. (1983) Mutat. Res.,
94, 435-447.

Natanabe, M., Maher, V. M., and McCormick, J. J. (1985) Mutat.
Res., 146, 285-294.

Maher, V. M., Patton, J. D., and McCormick, J. J. (1988) in Health
Effects Institute Report (Health Effects Institute, Cambridge, MA),
17.

Maher, V. M., Rowan, L. A., Silinskas, K. C., Kateley, S. A., and
McCormick, J. J. (1982) Proc. Natl. Acad. Sci. USA, 79, 2613-2617.

Grisham, J. N., Greenberg, D. S., Kaufman, D. G., and Smith, G. J.
(1980) Proc. Natl. Acad. Sci., USA, 77, 4813-4817.

Bohr, V. A., Smith, C. A., Dkumoto, D. S., and Hanawalt, P. C.
(1985) Cell, 40, 359-369.

Madhani, H. D., Bohr, V. A., and Hanawalt, P. C. (1986) Cell, 45,
417-423.

Mellon, I., Bohr, V. A., Smith, C. A., and Hanawalt, P. C. (1986)
Proc. Natl. Acad. Sci., USA, 83, 8878-8887.

Bohr, V. A., Phillips, D. H., and Hanawalt, P. C. (1987) Cancer
Res., 47, 6426-6436.

Mellon, I., Spivak, G., and Hanawalt, P. C. (1987) Cell, 51, 241-
249.

McCormick, J. J, and Maher, V. M. (1981) in DNA Repair: A Laboratory
Manual of Research Procedures, eds. Friedberg, E. C. and Hanawalt,
P. C. (Marcel Dekker, New York), 18, 501-521.

Ryan, P. A., Maher, V. M., and McCormick, J. J. (1987) Exp. Cell
Res., 172, 318-328.

Wang, Y., Parks, N. C., Nigle, J. C., Maher, V. M., and McCormick,
J. J. (1986) Mutat. Res., 175, 107-114.

Yang. J.-L., Maher, V. M. and McCormick, J. J. (1989) Gene, 83,
78

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.
33.

79
347-354.

Yang, L. L., Maher, V. M., and McCormick, J. J. (1980) Proc. Natl.
Acad. Sci., USA, 77, 5933-5937.

Osborne, M. R., Beland, F. A., Harvey, R. G., and Brookes, P. (1976)
Int. J. Cancer, 18, 362-368.

Staub, K. M., Meehan, T., Burlingame, A. L., and Calvin, M. (1977)
Proc. Natl. Acad. Sci., USA, 74, 5285-5289.

Eisenstadt, E., Warren, A. J., Porter, A., Atkins, P., and Miller,
J. H. (1982) Proc. Natl. Acad. Sci., USA, 79, 1945-1949.

Yang, J.-L., Maher, V. M., and McCormick, J. J. (1987) Proc. Natl.
Acad. Sci., USA, 3787-3791.

Mazur, M., and Glickman, B. N. (1988) Somat. Cell and Molec.
Genet., 14, 393-400.

Carothers, A. M. and Grunberger, D. (1990) Carcinogenesis, 11, 189-
192.

Strauss, B., Rabkin, S., Sagher, D., and Moore, P. (1982)
Biochimie, 64, 829-838.

Brown, T., Hunter, N. N., Kneale, G., and Kennard, 0. (1986) Proc.
Natl. Acad. Sci., USA, 83, 2402-2406.

Brown, T., Leonard G. A., Booth, E. D., and Chambers, J. (1989) J.
Mol. Biol., 207, 455-457.

Norman, D. Abuaf, P., Hingerty, B. E., Live, 0., Grunberger, D.,
Broyde, S., and Patel, 0. J. (1989) Biochemistry, 28, 7462-7467.

Carothers, A. M., Steigerwalt, R. N., Urlaub, G., Chasin, L. A., and
Grunberger, D. (1989) J. Mol. Biol., 208, 417-428.

Urlaub, 6., Mitchell, P. J., Ciudad, C. 0., and Chasin. L. A. (1989)
Mol. Cell. Biol., 9, 2868-2880.
Vrieling, H., Van Rooijen, M. L., Groen, N. A., Zdzienicka, M. Z.,
Simons, J. N. I. M., Lohman, P. H. M., and van Zeeland, A. A. (1989)
Mol. Cell. Biol. 9, 1277-1283.

Cordeiro-Stone, M., Boyer, J. C., Smith, B. A., and Kaufmann, N. K.
(1986) Carcinogenesis, 7, 1775-1781.

Boles, T. C. and Hogan, M. E., (1986) Biochemistry, 25, 3039-3043.

Kootstra, A., Lew, L. K., Nairn, R. S., and MacLeod, M. C. (1989)
Hol. Carcinogenesis, 1, 239-244.

80
34. Arnott, S., and Selsing, E. (1974) J. Mol. Biol., 88, 551-552.

35. Wang A., H.-J., Fujii, S., van Bloom, J. H., and Rich A. (1982)
Proc. Natl. Acad. Sci. USA, 79, 3968-3972.

CHAPTER III

LACK OF A CELL CYCLE-DEPENDENT STRAND BIAS FOR MUTATIONS INDUCED
IN THE HPRT GENE BY (i)-7B,80-DIHYDROXY-9a,IOa-EPOXY-7,8,9,10-
TETRAHYDROBENZO[A]PYRENE IN EXCISIDN REPAIR-DEFICIENT HUMAN CELLS

Ruey-Hwa Chen, Veronica M. Maher, and J. Justin McCormick

Carcinogenesis Laboratory - Fee Hall
Department of Microbiology and Department of Biochemistry

Michigan State University, East Lansing, MI 48824-1316

81

SUMMARY

He showed previously that in repair-proficient human cells the
location of the premutagenic lesion induced by (:)-78,8a-dihydroxy-9a,10a-
epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene (BPDE), viz. , the guanine in a G-C
base substitution, in mutants derived from cells treated at the beginning
of S phase just when the hypoxanthine (guanine) phosphoribosyltransferase
(HPRT) gene is replicated differs significantly from their location in
cells treated 12 h prior to the beginning of S phase (early G, phase)
(Chen, R.-H. et al., Proc. Natl. Acad. Sci. USA, 87:8680~8684, 1990).
This suggests that the cells preferentially remove BPDE adducts from the
transcribed strand. We have now determined the kinds and location of
independent mutations induced by BPDE in the coding region of the HPRT
gene of synchronized repair-deficient xeroderma pigmentosum (XP) cells
(XP128E, complementation group A), treated at S or in G,. Nineteen out
of 25 mutants derived from S-treated cells and 23 out of 28 mutants from
G1-treated cells contained base substitutions. 89% of these involved a G-C
base pair, primarily G-C -—> T-A transversions. This is similar to the
kinds of mutations we saw in the repair-proficient cells. However, in
contrast to our earlier results, there was no change in strand
distribution of premutagenic BPDE lesions. In both populations, ~ 26% of
the base substitutions involving G-C base pairs had the 6 located in the
transcribed strand, 5/18 in the S phase mutants and 5/21 in the G1 phase
mutants. These results support the hypothesis that the strong strand bias
of induced mutations observed in the repair-proficient cells results from
preferential repair of BPDE-induced DNA damage from the transcribed

strand.

82

INTRODUCTION

Induction of mutations in mammalian cells by carcinogens is considered to
play an important role in the multistep process of carcinogenesis. To
understand the mechanisms by which carcinogens cause malignant
transformation, we and our colleagues are investigating the kinds of
mutations induced in diploid human fibroblasts by various carcinogens, as
well as the role of DNA repair and DNA replication in this process. We
showed that the frequency of mutants induced by UV radiation (1, 2), BPDE
(3, 4), or 1-nitrosopyrene (5) in nucleotide excision repair-proficient
cells is highest in cells treated in early S phase and much lower in cells
treated in early G, phase. This difference cannot be explained by
differences in the physical state of the DNA during treatment, since no
such difference was found when xeroderma pigmentosum cells (XP128E,
complementation group A), which are virtually incapable of nucleotide
excision repair (6), were used in such comparisons (2, 3). These data
suggest that DNA replication is centrally involved in the conversion of
DNA damage into mutations and that excision repair prior to the DNA
replication decreases the frequency of mutants by eliminating DNA damage.

Recently, it has been shown that in cultured hamster and human
cells, UV-induced pyrimidine dimers are excised more rapidly from an
actively transcribed gene, dihydrofolate reductase, than from genome
overall (7, 8), or from the 5’ or 3’ flanking region of the DHFR gene (9,
10). Mellon et al. (11) showed that in these cells such lesions are
removed from the transcribed strand of the gene much more rapidly than

from the nontranscribed strand. If this were also the case with BPDE-

83

84
induced DNA damage in the HPRT gene of diploid human cells, mutants
derived from the repair-proficient cells treated in early S might well
differ from those taken from cells treated in 6, phase in the strand
distribution of the premutagenic lesion assumed to be responsible for the
mutations, i.e., the guanine-BPDE adduct.

We recently tested this hypothesis (12) by analyzing BPDE-induced
mutants from the two populations for the kinds of mutations and their
location in the coding region of the HPRT gene. There was no difference
in the kinds of mutations, but in mutants from cells treated in early S,
the premutagenic lesions were located in either strand, with 25% in the
transcribed strand. In nmtants derived from G,-treated cells, all the
premutagenic lesions were located in the nontranscribed strand. This
difference could reflect preferential repair of BPDE-induced DNA damage
from the transcribed strand of the HPRT gene, but other explanations are
possible. If the observed difference reflects preferential excision
repair, then there should be no such difference in mutants derived from
populations of XP128E cells exposed to BPDE in early S or 12 h prior to
the onset of S phase. To test this hypothesis, as well as to gain
information on the kinds and location of the BPDE-induced mutations in an
endogenous gene of excision repair-deficient human cells, we have now
determined the spectrum of such HPRT mutations in synchronized XP128E
cells. The results showed a transcribed: nontranscribed strand ratio of
premutagenic lesions of 28:72 in S phase mutants and 24:76 in G, phase
mutants, ratios similar to the 25:75 ratio found in repair-proficient
cells exposed to BPDE in early S phase so that there would be little or no

time for repair before the HPRT gene was replicated.

MATERIALS AND METHODS

Cells and Media. XP128E cells obtained from the American Type
culture collection (Rockville, MD) were cultured in modified MCDB-110
medium (13) prepared with Earle’s salts and containing 10% supplemented
calf serum (HyClone, Logan, UT) (culture medium). For selection of 6-
thioguanine-resistant cells, the same medium, but lacking adenine and
containing 5% supplemented calf serum (HyClone). 5% fetal calf serum and
40 uM TG was used.

Cell Synchronization. The method used has been described (4, 12).
In summary, cells were inoculated at a density B-fold less than that
attained at confluence, fed culture medium every other day to stimulate
rapid growth to confluence, and after reaching confluence, fed daily for
3 additional days. The medium was not changed for the next 72 h so that
the cells would cease proliferating (60 state). To stimulate them to re-
enter the cell cycle, the cells were released from confluence, and plated
in fresh culture medium at a density of 10‘ cells/cmz. Such cells begin
DNA synthesis (5 phase) after 16 h (14).

Exposure to Mutagen and Isolation of HPRT’Mutants. The details of
the BPDE treatment and mutagenesis assay has been described (12).
Briefly, a series of independent populations was exposed to BPDE 17 h
after release from G0 (at the beginning of S-phase) or 5 h after release
(in early G, phase). After 1 h, the medium was removed, and the cells were
rinsed with phosphate-buffered saline (pH 7.4), and fed with fresh culture
medium. The cells in one dish were assayed immediately for survival as

described (12, 15). The rest were allowed 8 days for expression of

85

86
resistance to TG before 0.5 x 106 cells from each population were plated
in selective medium. Nhen TG-resistant clones developed 14 days later,
these were located and isolated.

Synthesis of First-Strand cDNA Directly fron IIMA in Cells. Mutant
clones composed of 50 to 400 cells were detached from the dishes by 5 pl
trypsin and suspended in 0.5 ml cold phosphate buffered saline (pH 7.4).
The cell suspension was centrifuged for IO-min at 4°C. The supernatant was
removed, and the cell pellet was resuspended in 5 pl of the cDNA cocktail
described by Yang et al (16). The reverse transcriptase reaction was
performed at 37°C for 1 h to allow the cell membranes to be lysed by
detergent and first-strand cDNA to be synthesized from total poly(A)mRNA
(16).

Amplification of HPRT cDNA and DNA Sequencing. The methods used to
prepare second-strand HPRT cDNA, amplify the cDNA 10"-fold using two 30-
cycles of polymerase chain reaction, and sequence the product directly
using three sequencing primers have been described (16). A fourth
sequencing primer
(5’-5°°CTTTTCACCAGCA‘8°-3’) was also used to sequence three 6, phase mutants
(X8623, XBG26, and XBG48).

RESULTS

Mutant Frequency. XPIZBE cells were synchronized and exposed to
0.013 ~0.022 pH of BPDE in early S- or early G, phase, and assayed for
survival and frequency of TG-resistant cells. Cell survival ranged from
12-34%, and the frequency of TG-resistant cells per 106 clonable cells
ranged from 25 to 94. The frequency observed in untreated populations was
~ 5 mutants per 10‘ clonable cells. There was no significant difference
in survival or mutant frequency between populations exposed in the two
different phases of the cell cycle.

Characterization of Mutations. To decrease the possibility of
including a mutant containing a spontaneous, rather than a BPDE-induced
mutation, we only analyzed the mutational data from mutants derived from
treated populations in which the frequency was at least 8 times higher
than that of the untreated populations. The entire HPRT coding region of
25 such S phase-derived mutants and 28 such G, phase-derived mutants was
amplified by polymerase chain reaction and sequenced directly. The
mutants were recognized as being independent, either because their
alterations were unique or because they had been derived from separate
populations of treated cells. The kinds of mutations and their location
in the coding region of the HPRT gene, and the consequence for the primary
structure of the protein are listed in Tables 1 and 2.

There was no significant difference between S- and G,-derived mutants
in the kinds of mutations they contained. Five mutants from S phase and
4 mutants from the G, phase were missing one or several consecutive exons

from their cDNA. These exon deletions probably resulted from base changes

87

88

 

 

9.: 23.... .53.... o.— .. 63.5 c: :_ 99:253. 9... .533... 39...... :55. 9.; 53.2.32: .9. o. .r. a :_ a. 995:3: 2:. 65...: 32:55.

ii .ll-ll.l| Illtlillllll .III-l!’ll.lll..l'iti‘ll.tllltlillll..

9.3.... a 5...”.—
u:_mm_.: v. :3“.—
..539: r. 5...”.—
9:35. r. :3”.—
ufimfi... n .85—

2:5: o... .5 99:9...57. -

 

535:2... 5.. .<2 62.53....5... .._.Z .........3..:.. .._. a

.......:.9_.....
3.5;.
31:7.
9953'.
9.5.7.
mmmax

2.3.5:...

2: 8:9. 3:2...—

 

 

<z 5:352. 999 90.9 9<9 < s. 22X. N 3 «man...
35:33..
.2 2... l 3. 9<< mt 9<u <9 i 9. o a n8 2......
92 :9 i .2 be o<m to <9 i 9. c a a... Em...
.._z 35 i .5 9<9 <8 <8 9. < i 9.9 a 3. N55.
92 .5. i .5 9<9 <mo <99 <9 i 9.9 a S. 5......
<2 .5 l .5 <<< 6.9... Be 9.9 i <9 .. 2.... 8N5...
9 «Elem 999 w... a: 9.9109 .. 9.... 5....
| 9. .9 2..
...z 5:858... 96 8... 699 < . 9 l U. U o 53.59 NS. $.53.
9 a: .129 6899. 9.6 9.<i9.u e e... 3......
. I <9 i 9.9 3N N
...z 5.. i 3. So 999 o<o 9.9 i <9 N 5859 2N =35.
92 E... l .5 98 com e<< <9 i 9... .. EN 5......
._.z E... i .5 98 8m 9). <9 i 9.5 . EN 3......
9 5< i m... 98 2m 8.. <9 i 9.5 n a: 29.5.
9 =3. l m... o<9 9<m 8.. <9 T 9. .9 .. a... 3.5.x
._ z .3... i =5 69.. o<m o<< ....< l 9... .. e... 2......
.2 321.5 <99 <9. 95 5.9.19... N o: 2.5...
. z 95. i .5 <5 <om :9 < 9 l 9... N 2. <55.
9 .5. i 8.. 9<o 9mo 99< 9.... l 9. 5 N 2. $5...
.2 3v. T .2 Be ":99. <<o < 9 l 9... N N... .55.
92 e... l 3.. <<o wt :6 <9 i 9. .9 N 2. 9%....
«Eggs-2.3m as:
$5.83 993.... 39:33... 5:35: gum .533.— 55:2
2299...: a... 3.... c:_.:< u:=.:..o....w .0 our.
.3... 3.5.x
9.9.0 :8 o...

x... 3...... a. 9...}. u...=.....u..< e... =.. 22.3 5...: no.9... 5.... .333... .... 2...: 92k: 95%. :33. 5...}... o... e. ‘83:... 23:83... 2...? 2.3.99... 3.... 5:2 _ .52....

0‘2 oucthQWr—Chv‘c: oh.z ugsthmp-Gu- 9.—

 

 

 

89

 

 

.o’rrflx’bgl v2 . .‘oocm—bomca
. a... 2. 9...... z......._...._.._. 2.. 2 as... .
. a. . . . filtllilintlo .I.z. .. . . :92}. .. 9‘13!!! 3|.-
9... 999.... 59.9.... 9.... .:.:.:_ E. :. 99:92.99 9... 5999...! 9.9:... __...:m 9.................«................. ..w- .. m .... $35.5 a can... 5.3:!
$5.99.... 9 .n .N 9:52.... 5......“
99.9.9. . .a .93.. .95....
9.5.9.... n .~ 2.9.... 3.9....
9:33.55.
9... 99....» 9.2.2....
<z 59.9.5... 9<9 H919 999 9 a 9.9.9.. 9 2.9 .9 2.9 929....
32.9.90

9 .21.... 999 9<m 9: 9.9195 9 9.9 995....
.2 2... 1.3 9<< mt 9<9 <919.5 9 999 25....
.2 .92 1 9:. 9<9 9m< 9.9 <9 1 9.5 9 999 . 2.5....
.2 :9 1..< 999 99m 9<9 <919.5 9 :9 9.5....
.2 _.> 1 .5 9<9 <99 <99 <9 1 9. 5 9 999 .95....

._ z :5 1 .5 9<9 <m9 <99 9. < 1 9.5 9 999 25:...
._ z .5. 1 .5 9<9 <m... <99 <9 1 9.5 9 999. .95....
._ 2 92 1 .5 9<9 <9m <99 5.9 1 9. 5 9 999 .95....

.. 99.9. 1 .2 99< <9m <99 9. < 1 9. 5 9 999 . .95....
<2 99.9 1 .3 999 9H9 9... .9. < 1 <.... .. 9.... 995....
.2 9:13... 9<9 m9< <9< <9 1 9. 5 9 9N9 995....
.2 2.. 1 99,. 99< m9 999 <9 1 9. 5 9 2.9 .95....
.2 a.< 1 .5 999 9.1.9 9<< 5.9 1 9 5 9 SN .95....
.2 9.9 1 .5 999 99m 92 <9 1 9. 5 . SN 85....
92 9.... 1 9.9 999 B9 999 < .9 .I 9. 5 9 99. 995....
.2 2... 19.9 999 99 999 <919.5 .. 99. 25....

.. 951.... 99< m<9 9<9 <919.9 .. 2: 2.5....

.- 5< 1 n... 9<9 9<m 999 <9 195 n 9: 995....

I <9 1 9. 5 . .2
92 5< 1 .92 <99 99< 9<9 9. < 1 <9 . .5959 9: 995....
.2 3.192 9<9 51.9999 <919.5 .. .2 8.5....
.2 9.21 9.< 2: 9m< 9<9 .9. < 19.5 N .2 N55.
<2 .9... 1 3.. <<9 9H9 9<9 5.9 1 <.... N 99 .95....
.9 .9... 1 ”.9 <9< m99 :9 <9 19.5 N 99 9N5....
9:523:33 one:
99:52.9 959:9 9995.58 .5222: gnu 5.5.3.. 59......
59.99.... 9... 5.99 ::_:.< 5.95.225 .9 9....P

9...; 992.9.

 

9.9.0 =99 9... \9 99...... 5 ......:9 3. m9: 5...: V9.99... 9:99 Mafia». :. 9...... .22: 95.5 5.99. «.599. 9c. .... V925... namesafi 955 2.9.3.2.. 3.... 22...! N 939..

90

at splice consensus sequences, but the change in the HPRT gene has not
been evaluated, since only the coding region of the gene was sequenced in
this study. Besides these exon deletion mutants, all but one mutants from
either group contained base substitutions. Most had only a single base
substitution. One mutant from each phase exhibited tandem base
substitutions and one S phase mutant carried a complex mutation (a base
substitution in conjunction with a -1 frameshift). One mutant from each
phase had a single base deletion. Table 3 shows that 89% of the base
substitutions derived from either phase involved G-C base pairs. Among
these, transversions of G-C to T-A predominated: 60% (12/20) for S phase
and 60% (14/24) for G1 phase. In both phases, the base alterations were
distributed throughout the coding region, but a substantial fraction were
located in exons 2, 3, and 8. The number of mutations analyzed was not
large enough to tell if their locations varied with the phase of the cell
cycle.

Strand Distribution of the Prenatagenic Lesions. In analysis of the
data for evidence of differences in location of premutagenic lesions that
gave rise to the observed mutation, we used only the mutations involving
G-C base pairs. This is because studies in mammalian cell lines in
culture reveal that BPDE binds only to purines, with > 95% of the DNA
adducts involving guanine (17, 18). As noted in Tables 1 and 2, and
diagrammed in Figure 1, in mutants derived from S phase, 5 out of 18 such
mutations (28%) had the guanine in the transcribed strand. A similar
result was found with mutants derived from G1 phase, i.e. in five out of
21 (24%) of the mutations, the guanine was in the transcribed strand.

This result differs significantly from that found in repair—proficient

91

 

Table 3 Types of base substitutions observed in the
coding region of the HPRT gene in XP cells treated

 

 

with BPDE
Types of base N9, of substitutions observed
substitutions S phase G1 phase
Transversions:
g.c --> T-A 12 14
G-C -—> C-G 2 3
T-A --> G-C 1 0
T-A --> A-T 0 2
Transitions:
G-C --> A-T 4 4
T-A --> C-G 1 1

 

Total 20 24

92

Figure 1. Location of independent HPRT mutations in XP128E cells treated
with BPDE in early S phase (27 mutations) or in early G, phase (29
mutations). Exon deletions are shown as open rectangles; base
substitutions are indicated by vertical bars; tandem mutations by vertical

arrows; and single base deletions by squares.

3
9

mmuzbo

oz<mkm
cum-mumza‘mhzoz

oz<mhm
ommEumz<mh

mmehO

oz<mhm
omm_mumz<mhzoz

oz<mhm
omeumzaEh

a 9.30.:

 

 

 

 

 

 

 

 

 

 

 

 

 

DE‘ n M l“
= _ _ 1 = _ u— _ _
_ _ uquala
_ _
D
n
- l—IU
n“ nu AJ :0
. _ q
_ a. . wmdzanm
_ _ _
I q - A d d u . _ ll
how 00w 000 00c. DOM DON OO— — "Eda wm<m

m.._|_mu mmNEX Z. dmkomam woam

94

cells treated in G, phase, in which all the premutagenic lesions were
located in the nontranscribed strand (12).

Seven out of 47 base pair changes identified did not involve a 600
base pair. However, among these, one from each phase was carried by a
mutant containing a tandem base substitution occurring at a 5’-TG sequence
(X8530 and X8650). It is very likely that each of these tandem base
substitutions was caused by the presencerof a single BPDE adduct formed at
the 3’ guanine. Therefore, only 5 nmtations out of the 47 analyzed

(<11%) did not involve GvC base pairs.

DISCUSSION

Our finding that the kinds and location of BPDE-induced mutations in
the coding region of the HPRT gene of excision repair-deficient XP128E
cells treated in early G, phase or in early S phase were similar indicates
that the physical state of the DNA in the cell during treatment did not
affect the spectrum. The data support the conclusion from our previous
study (12) that the difference in the strand distribution of premutagenic
lesions observed between repair-proficient human cells treated in S phase
and such cells treated in G, phase is caused by preferential repair of
BPDE-induced DNA damage from the transcribed strand of the HPRT gene.
This strand-specific repair of a bulky BPDE-DNA adduct is consistent with
the finding of Mellon et al. (11) who showed that UV-induced pyrimidine
dimers are preferentially removed from the transcribed strand of actively
transcribed genes, such as the DHFR gene of human cells.

The kinds of mutations induced by BPDE in XP128E cells are similar
to those we found in repair-proficient cells (12), those induced in an
extrachromosomal gene, supF, replicating in repair-proficient human cells
(19), as well as those found in two other endogenous genes of repair-
proficient mammalian cells (20, 21). In all the studies described above,
60C to T-A transversions predominate. This transversion would result if
the DNA polymerase(s) preferentially inserted a deoxyadenosine
triphosphate across from a non-instructional base containing a bulky BPDE-
[JNA adduct (22), or if the presence of a BPDE adduct on guanine allowed a
stable purine-purine pairing to occur during replication, and this

nnispairing were not recognized and removed. There is evidence from model

95

96

oligonucleotides that unmodified guanine'can pair stably'with adenine (23,
24). Also, Norman et al. (25) showed that in a double-stranded
oligonucleotide containing a single aminofluorene residue bound to guanine
in the 08 position, adenine opposite the modified guanine produces a
stable structure which places the aminofluorene in the B—DNA minor grove,
with the guanine syn. Whether this occurs with BPDE-guanine adducts is
not yet known.

The kinds of mutations we observed in XPIZBE cells differ from the
kinds found by Bernelot-Moens et al. (26) in the lacI gene of excision
repair-deficient E. coli with BPDE. In addition to 6°C to T-A
transversions, A-T to T-A transversions and deletions of a G-C base pair
(-1 frame shift) were frequently observed. This may reflect a difference
in mutational mechanisms between procaryotic and eucaryotic cells.

We observed only five mutations involving A-T base pairs. Although
BPDE binds to adenine at very low frequencies (17, 18), some of these
mutations may represent pre-existing background mutations. Since the
majority (89%) of the base substitutions analyzed in the coding region of
the HPRT gene of mutants derived from untreated diploid human fibroblasts
involved A-T pairs (27), and since the frequencies we obtained with BPDE
in these XP cells were only 8 to 15-fold above background, we could have
included one or two background mutants in our analysis.

Two tandem base substitutions and one complex mutation were
recovered in the present study. These events have not been reported in
other mammalian endogenous genes after BPDE treatment (20, 21), nor in
BPDE-induced HPRT mutants derived from repair-proficient cells (12, 27).

However, three mutants out of 86 generated when a shuttle vector carrying

97
BPDE-guanine adducts replicated in human cells contained such tandem base
substitutions (19). The tandem base substitutions may be the result of a
reduction of the fidelity of the DNA polymerase at the time of replication
bypassing bulky lesions. Reduction of fidelity, followed by a strand
slippage induced by BPDE adducts, could also account for the complex
mutation observed in mutant X8536.

The distribution of guanine nucleotides between the transcribed and
nontranscribed strands in the coding region of the human HPRT gene is
41:59. The ratio becomes 38:62 if base substitutions that will not result
in amino acid changes are excluded. Since XP128E cells do not remove
BPDE-guanine adducts, if binding of BPDE to the HPRT gene is random, and
every amino acid change caused by a BPDE-induced mutation were to affect
the activity of the HPRT enzyme, then one would expect 38% of the observed
mutations to correspond to a BPDE-guanine adduct in the transcribed
strand. Instead, only 26% did (28% for S phase and 24% for G, phase).
This is very similar to the distribution we found in populations of
repair-proficient cells exposed to BPDE in early S phase so that they had
no time repair before replication of HPRT occurred (12). A 26:74 ratio is
not significantly different from the theoretical 38:62 ratio (P = 0.11).
However, we suggest that the 26:74 ratio reflects the strand distribution
of guanines that cannot be mutated without affecting the activity of the
protein. If this is the case, then studies in which the premutagenic
lesions involved cytosine ought to show just the opposite strand
distribution, i.e. 75% transcribed , 25% nontranscribed. No study of HPRT
mutations induced in the absence of excision repair by a nmtagen that

causes only cytosine lesions has been reported. However, we and our

98
colleagues (14) recently carried out a study using UV radiation, an agent
which induces the mutations at dipyrimidine sites and for which the
affected pyrimidine is predominately cytosine. The results showed that
the distribution of premutagenic lesions in repair-proficient cells
treated in early S phase, as well as XP128E cells irradiated in S or G,.
was 75% transcribed strand : 25% nontranscribed strand.

Vrieling et al. (28) attributed the strong strand bias of UV-induced
hprt mutations observed in UV—sensitive Chinese hamster cells (V-Hl), in
which 90% of the mutations resulted from photoproducts in the transcribed
strand, to a difference in fidelity of DNA replication between the leading
and the lagging strand, with a more error-prone polymerase (e.g.,
polymerase delta) replicating the transcribed strand. Our finding that
fewer mutations were targeted by BPDE-modified guanine in the transcribed
strand of the HPRT gene than is predicted by the distribution of guanines
in that strand argues against this explanation. Direct evidence that
polymerase delta is not necessarily more error-prone than polymerase alpha
comes from a study by Shibutani et al. (29) who showed that during
translesional DNA synthesis on an oligonucleotide template containing an
8-oxo-7-hydrodeoxyguanosine, the correct nucleotide is less frequently
inserted by polymerase alpha than by polymerase delta.

In our previous study (12), a cell cycle-dependent "hot spot region"
was observed in HPRT mutants derived from repair-proficient cells treated
with BPDE in G, phase and allowed 12 h for excision repair before 5.
Twenty-nine percent of the mutations induced in cells treated in G, phase
were located in the sequence of 5’A(G)6C-3’ (positions 206-213), compared
to 4% in the cells treated in S phase. If, as we suggested, this "hot

99

spot region" resulted from inefficient repair of BPDE adducts in the run
of 6 consecutive guanine bases, so that after excision repair had
occurred, the adducts remaining in this region represented a higher
fraction than originally present, there should not be any such difference
in the XP128E cells. In the present study, only two mutations induced in
each phase were located in this region, a significantly smaller percentage
(8% for S and 7% for 6,) than the 29% seen in repair-proficient cells
treated in G, phase. These results support the hypothesis that inefficient
excision repair can account for the prominent "hot spot region" in the
latter population. Several studies showed that the helix of poly(dG)-
poly(dC) resembles an A-like helix geometry, rather than the traditional
8 form (30, 31). If a BPDE-adduct formed in the run of 6 G’s were to
produce a specific conformation at the site of the adduct, this could
affect the recognition by repair enzymes. If localized unwinding is
necessary for the formation of an excision repair complex at the site of
the BPDE-DNA adduct, and the repair complex melts an A+T-rich region more
easily than a G+C-rich region (32), this could also explain why repair is .
less efficient in this G+C~rich region of the HPRT gene.

In summary, we have shown that there is no significant difference in
the kinds of mutations induced by BPDE in repair-deficient XP128E cells,
their location in the HPRT gene, or in the strand distribution of the
premutagenic lesion that targeted the mutations. These results support
the hypothesis that the strong strand bias of induced mutations observed
in the repair-proficient cells (12) results from preferential repair of

BPDE-induced DNA damage from the transcribed strand.

ACKNOHLEDGHENT

We gratefully thank Andy Mendians and Ena Zaccagnini for their excellent

technical assistance.

.100

10.

REFERENCES

Maher, V. M., Dorney, D. J., Mendrala, A. L, Konze-Thomas, 8., and
McCormick, J. J. DNA excision-repair processes in human cells can
eliminate the cytotoxic and mutagenic consequences of Ultraviolet
irradiation. Mutat. Res., 62: 311-313, 1979.

Konze-Thomas, 8., Hazard, R. M., Maher, V. M., and McCormick, J. J.
Extent of excision repair before DNA synthesis determines the
mutagenic but not the lethal effect of UV radiation. Mutat. Res.,
94: 421-434, 1982.

Yang, L. L., Maher, V. M., and McCormick, J. J. Relationship
between excision repair and the cytotoxic and mutagenic effect of
the "anti" 7,8-diol-9,10-epoxide of benzo[a]pyrene in human cells.
Mutat. Res., 94: 435-447, 1982.

Natanabe, M., Maher, V. M., and McCormick, J. J. Excision repair of
UV-or benzo[a]pyrene diol epoxide-induced lesions in xeroderma
pigmentosum variant cells is "error free". Mutat. Res., 146: 285-
294, 1985.

Maher, V. M., Patton, J. 0., and McCormick, J. J. Studies on the
metabolism and biological effects of nitropyrene and related nitro-
polycyclic aromatic compounds in diploid human fibroblasts. In:
Health Effects Report, Vol. 17, pp. 1-40. Cambridge, MA: Health
Effects Institute, 1988.

Petinga, R. A., Andrews, A. D, Tarone, R. E., and Robbins, J. H.
Typical xeroderma pigmentosum complementary group A fibroblasts have
detectable ultraviolet light-induced unscheduled DNA. synthesis.
Biochim. Biophs. Acta., 479: 400-410, 1977.

Bohr, V. A., Smith C. A., Dkumoto, D. S., and Hanawalt, P. C. DNA
repair in an active gene: Removal of pyrimidine dimers from the
DHFR gene of CHO cells is much more efficient than in the genome
overall. Cell, 40: 359-369, 1985.

Mellon, I., Bohr; V. A., Smith, C. A., and Hanawalt, P. C.
Preferential DNA repair of an active gene in human cells. Proc.
Natl. Acad. Sci. USA, 83: 8878-8882, 1986.

Bohr, V. A., Dkumoto, D. S., and Hanawalt, P. C. Survival of UV-
irradiated mammalian cells correlates with efficient DNA repair in
an essential gene. Proc. Natl. Acad. Sci. USA, 83: 3830-3833,
1986.

Bohr, V. A., Dkumoto, D. S. Ho, L., and Hanawalt, P. C.
Characterization of a DNA repair domain containing the dihydrofolate
reductase gene in Chinese hamster Ovary cells. J. Biol. Chem., 261:
16666-16672, 1986.

101

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

£21.

102

Mellon, I. Spivak, G., and Hanawalt, P. C. Selective removal of
transcription-blocking DNA damage from the transcribed strand of the
mammalian DHFR gene. Cell, 51: 241-249, 1987.

Chen, R.-H., Maher, V. M., and McCormick, J. J. Effect of excision
repair by diploid human fibroblasts on the kinds and locations of
mutations induced by (1)-7B,8a-dihydroxy-9a,10a-epoxy-7,8,9,10-
tetrahydrobenzo[a]-pyrene in the coding region of HPRT gene. Proc.
Natl. Acad. Sci. USA, 83: 8680-8684, 1990.

Ryan P. A., McCormick, J. J., and Maher, V. M. Modification of
MCDB-llO medium to support prolonged growth and consistent high
cloning efficiency of diploid human fibroblasts. Exp. Cell Res.,
172: 313-328, 1987.

McGregor, H. G., Chen, R.-H., Lukash, L., Maher, V. M., and
McCormick, J. J. Cell cycle-dependent strand bias for UV-induced
mutations in the transcribed strand of excision repair-proficient
human fibroblasts, but not in repair-deficient cells. Mol. Cell.
Biol., in press, 1991.

Wang, Y., Park, H. C., Higle, J. C., Maher, V. M., and McCormick, J.
J. Fibroblasts from patients with inherited predisposition to
retinoblastoma exhibit normal sensitivity to the mutagenic effects
of ionizing radiation. Mutat. Res., 175: 107-114, 1986.

Yang, J.-L., Maher, V. M., and McCormick, J. J. Amplification and
direct nucleotide sequencing of cDNA from the lysate of low numbers
of diploid human cells. Gene, 83: 347-354, 1989.

Straub, K. M., Meeham, T., Burlingame, A. L., and Calvin, M.
Identification of the major adducts formed by reaction of
benzo[a]pyrene diol epoxide with DNA in vitro. Proc. Natl. Acad.
Sci. USA, 74: 5285-5289, 1977.

Osborne, M. R., Jacobs, S., Harvey, R. G., and Brookes, P. (1981)
Honor products from the reaction of (+) and (-) benzo[a]pyrene-anti-
diolepoxide with DNA. Carcinogenesis, 2:553 558, 1981.

Yang, J.-L., Maher, V. M., and McCormick, J. J. Kinds of mutations
formed when a shuttle vector containing adducts of (:)-78,8a-
dihydroxy-9a,10a-epoxy-7,8,9,lO-tetrahydrobenzo[a]pyrenereplicates
in human cells. Proc. Natl. Acad. Sci. USA, 84: 3787-3791, 1987.

Mazur, M., and Glickman, B. N. Sequence specificity of mutations
induced by benzo[a]pyrene-7,8-diol-9,10-epoxide at endogenous aprt
gene in CHO cells. Somat. Cell Mol. Genet., 14: 393-400, 1988.

Carothers, A. M., and Grunberger, 0. DNA base changes in
benzo[a]pyrene diol epoxide-induced dihydrofolate reductase mutants
of Chinese hamster Ovary cells. Carcinogenesis, 11: 189-192, 1990.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

103

Strauss, B., Rabkin, S., Sagher, 0., and Moore, P. The role of DNA
polymerase in base substitution mutagenesis on non-instructional
templates. Biochimie, 64: 829-838, 1982.

Brown, T., Hunter, w. N., Kneale, G., and Kennard, 0. Molecular
structure of the G-A base pair in DNA and its implication for the
mechanism of transversion mutations. Proc. Natl. Acad. Sci. USA,
83: 2402-2406, 1986.

Brown, T., Leonard, G. A., Booth, E. 0., and Chambers, J. Crystal
structure and stability of a DNA duplex containing A(anti)-G(syn)
base-pairs. J. Mol. Biol. 207: 455-457, 1989.

Norman, 0., Abuaf, P., Hingerty, B. E., Live, 0., Grunberger, 0.,
Broyde, S., and Patel, 0. J. NMR and computational characterization
of the N-(deoxyguanosin-B-yl)aminofluorene adduct [(AF)G] opposite
adenosine in DNA: (AF)G[syn]-A[anti] pair formation and its pH
dependence. Biochemistry, 28: 7462-7467, 1989.

Bernelot-Moens, C., Glickman, B. N., and Gordon, A. J. E., Induction
of specific frameshift and base substitution events by
benzo[a]pyrene diol epoxide in excision-repair-deficient Escherichia
coli. Carcinogenesis, 11: 781-785, 1990.

Yang, J.-L., Chen, R.-H., Maher, V. M., and McCormick, J. J. Kinds
and location of nmtations induced by (:)-78,8a-dihydroxy-9a,10a-
epoxy-7,8,9,10-tetrahydrobenzola]pyrene in the coding region of the
hypoxanthine (guanine) phosphoribosyltransferase gene ‘hi diploid
human fibroblasts. Carcinogenesis, 12: 71-75, 1991.

Vrieling, H., Van Rooijen, M. L., Groen, N. A., Zdzienicka, M. Z.,
Simons, J. H. I. M., Lohman, P. H. M., and van Zeeland, A. A. DNA
strand specificity for UV-induced mutations in mammalian cells.
Mol. Cell. Biol., 9: 1277-1283, 1989.

Shibutani, S., Takeshita, M., and Grollman, A. Pu Insertion of
specific bases during DNA synthesis past the oxidation damaged base
8-oxodG. Nature, 349: 431-434, 1991.

Wang, A. H.-J., Fujii, S., van Bloom, J. H., and Rich, A. Molecular
structure of the octamer d(G-G-C-C-G-G-C-C):modified A-DNA. Proc.
Natl. Acad. Sci. USA, 79: 3968-3972, 1982.

Drew, H. R., and Travers, A. A., DNA structural variations in the E.
coli ter promoter. Cell, 37: 491-502, 1984.

Van Houten, 8. Nucleotide excision repair in ESCherichia coli.
Microbiol. Rev., 54: 18-51, 1990.

CHAPTER IV

Cell Cycle-Dependent Strand Bias for UV-Induced Mutations in the
Transcribed
Strand of Excision Repair-Proficient Hulan Fibroblasts, But Not in
Repair-Deficient Cells

H. Glenn McGregor, Ruey-Hwa Chen, Lubov Lukash, Veronica M. Maher, and
J. Justin McCorlick

Carcinogenesis Laboratory - Fee Hall
Department of Microbiology and Department of Biochemistry
Michigan State University, Fee Hall, East Lansing, MI 48824-1316
Telphone: (517) 353-7785: FAX (517) 353-9004

104

To study the effect of nucleotide excision repair on the spectrum of
mutations induced in diploid human fibroblasts by UV254m’ we synchronized
repair-proficient cells and irradiated them when the HPRT gene was about
to be replicated (early S phase) so there would be no time for repair in
that gene before replication, or in G, phase 6 h prior to S, and determined
the kinds and location of mutations in that gene. As a control, we also
compared the spectra of mutations induced in synchronized populations of
xeroderma pigmentosum (XP) cells (XP128E cells, which are unable to excise
UV-induced DNA damage). Among the 84 mutants sequenced, base
substitutions predominated. 0f the XP mutants from S or G, and the repair-
proficient mutants from S, ~62% of these were G-C --> A-T. In the repair-
proficient mutants from G,, 47% were. In mutants from the repair-
proficient cells irradiated in S, 71% of the premutagenic lesions (10/14)
were located in the transcribed strand; with mutants from such cells
irradiated in G,, only 20% were (3/15). In contrast, there was no
statistically significant difference in the fraction of premutagenic
lesions located in the transcribed strand of the XP128E cells; ~ 75% of
the premutagenic lesions (24/32) were located in that strand, i.e., 15/19
(79%) in the S phase cells and 9/13 (69%) in the G, cells. The switch in
strand bias supports preferential nucleotide excision repair of UV-induced

damage in the transcribed strand of the HPRT gene.

105

INTRODUCTION

Populations of diploid human fibroblasts, synchronized by release
from the density-inhibited Go state and plated at 10‘tcells/cm2, begin DNA
synthesis after approximately 16 h (26). If populations of nucleotide
excision repair-proficient cells are irradiated with ”Vanna light 17 h
after release from confluence, the frequency of 6-thioguanine-resistant
(TG') mutants (resulting from loss of function of the gene coding for
hypoxanthine (guanine) phosphoribosyltrans-
ferase, HPRT) is significantly higher than in cells irradiated in early G,
phase (9). This difference in frequency cannot be attributed to
differences in the physical state of the DNA since no such difference in
mutation frequency is found when diploid xeroderma pigmentosum (XP) cells
(XP128E, 'complementation group A), which are virtually devoid of
nucleotide excision repair (18), are irradiated in early S phase or in
early G, phase (9). These data suggest that S phase replication is
centrally involved in the conversion of potentially mutagenic DNA damage
into mutations ("fixation"), and that excision repair prior to the onset

of S phase decreases the frequency of mutants by eliminating such lesions.

0f the two major classes of potentially mutagenic UV photoproducts,
i.e., cyclobutane pyrimidine dimers (Py-Py dimers) and 6-4 pyrimidine-
pyrimidone (6-4 Py-Py) lesions, the latter have been shown to be removed
from human genomic DNA more rapidly than the former (15). It has also
been shown that Py-Py dimers are excised more rapidly from an actively

‘transcribed gene of human cells, dihydrofolate reductase (DHFR), than from

106

107

bulk DNA (1, 12). What is more, Mellon et al. (14) demonstrated that
there is a strand bias in the rate of removal of Py-Py dimers from the
DHFR gene in human cells. The majority of such lesions are removed within
24 h, but dimers in the transcribed strand are excised much faster than
those in the nontranscribed strand. Information on strand-specific
excision repair of UV-induced damage from the HPRT gene of human cells is
not yet available. But if such repair occurs, HPRT mutations induced by
UV in excision repair-proficient human cells irradiated at the onset of S
phase should arise from 6-4 Py-Py lesions and Py-Py dimers located in
either strand of DNA. In contrast, the lesions responsible for the
mutations induced in such cells allowed 6 h for repair before the onset of
S phase should arise mainly from Py-Py dimers, and these should be located
primarily in the nontranscribed strand. This switch in strands should not
occur in XP128E cells.

There are no published data on the kinds and locations (i.e, the
spectrum) of UV-induced mutations in an endogenous gene of human cells,
nor data showing ‘whether excision repair in such cells alters the
spectrum. The present study was designed to examine the spectrum of
mutations induced by UV in the HPRT gene of diploid human cells and to
determine, at the DNA sequence level, the effect of nucleotide excision
repair of UV-induced photoproducts on the induction of such mutations. Ne
were particularly interested to see if there were biological evidence of
strand-specific repair of UV photoproducts in diploid human fibroblasts
and to compare the results with what was found in V79 Chinese hamster cell
lines by Vrieling et al. (25).

Although the time of HPRT replication has not been measured directly,

108

studies of the mutation frequency of TG-resistant mutants induced by UV
irradiation at different times across the S phase of synchronized
papulations of human fibroblasts showed that the highest frequency
occurred when cells were irradiated in the first 1/4 of S phase (7). A
similar finding was reported for Syrian hamster cells by Tsutsui et al.
(23). To investigate the effect of nucleotide excision repair on the
spectruni of UV-induced lesions, we irradiated repair-proficient and
repair-deficient diploid human cells in early S phase, just before the
HPRT gene is replicated, or in G, phase, 6 h before replication and
determined the kinds and location (spectrum) of the nmtations in the
coding region of the gene. We then analyzed the data to determine in
which strand the premutagenic dipyrimidine photoproduct was located. The
distribution of such sites in the coding region of the human HPRT gene is:
59% transcribed strand, 41% nontranscribed. If their mutagenic potential
is comparable and their distribution in HPRT sequences coding for critical
parts of the protein is random, then the mutations induced in repair-
proficient cells irradiated in early S phase and in XP128E cells ought to
be 59% arising from lesions in the transcribed strand, and 41% from
lesions in the nontranscribed strand. Mutations in repair-proficient
cells irradiated in G, phase might not show this pattern.

The XP128E cells showed a transcribed:nontranscribed strand ratio of
premutagenic lesions of 79:21 in S phase and 69:31 in G, phase mutants.
The repair-proficient cells also showed a 71:29 ratio in the S phase
mutants, but the G, mutants showed a 20:80 ratio. These data indicate bias
toward UV-induced mutations resulting from lesions in the transcribed

strand of the HPRT gene in the absence of repair, and preferential repair

109

of such lesions in that strand.

MATERIALS AND m5

Cells and ledia. Finite life span diploid human fibroblasts,
designated NFSL89, were explanted from the foreskin of a normal newborn
(11). These cells have normal nucleotide excision repair capacity.
XP128E cells, were obtained from the American Type Culture Collection
(Rockville, MD). Early-passage cells were used and were routinely
cultured in a modified MCDB-IIO medium (20) containing 10% supplemented
bovine calf serum (Hy Clone, Logan, Utah) (culture medium). For selection
of TGr cells, the same medium, but lacking adenine and containing 5% fetal
bovine serum, 5% supplemented calf serum and 40 DM T6 was used. For
thymidine incorporation experiments, the medium was changed to Eagle’s

minimal essential medium (MEM) to eliminate thymidine.

Mutagenesis and Isolation of HPRT' Mutants.

Cell Synchronization. Cells were driven into the G,, state as
described (26). To stimulate the cells to enter the cell cycle, they were
plated in culture medium at a density of 10‘ cells/cmz. The time of onset
of S phase following release from G0 was determined by the incorporation

of 3H-thymidine into acid-insoluble material as described (26).

Exposure to UV Light and Isolation HPRT: of Mutants. A series of
independent populations (1.5 x 106 cells each, plated in 150 mm-diameter
plastic dishes) was irradiated 17 h after release from G0 (at the beginning
of S phase) or 11 h after release from G,, (in G, phase). The culture

medium was aspirated, and the cells were washed with sterile phosphate

110

111
buffer saline (PBS, pH 7.4). The cells were irradiated as described (17)
using a UV fluence of 6.5 J/h? for repair-proficient cells and 0.5 J/nF for
repair-deficient cells and re-fed with culture medium. Cells plated at
cloning density were similarly exposed and used to determine survival of
colony-forming ability. ‘The target cells were allowed an 8 day expression
period before selection of 2.5 X 10'5 cells from each population with TG was
begun as described (11). When macroscopic drug-resistant clones had
developed 14 days later, these were located, isolated, and expanded into
cultures composed of ~ 2.0 X 10“ cells, or the cDNA was amplified directly

from the original clone.

Synthesis of first—strand cDNA directly fron IRMA in cells. Cells
were trypsinized and suspended in culture medium and if the cells in a
colony had been expanded, the cell number was determined by electronic
counting, and the cells were diluted in cold PBS (pH 7.4). From 100 to 500
cells in PBS were transferred to a 0.5 ml Eppendorf tube and centrifuged
for 10 min at 4°C. The supernatant was removed, and the cell pellet was
resuspended in 5 ul of the cDNA cocktail described in detail by Yang et
al. (28). The reverse transcriptase reaction was performed at 37°C for 1
h to allow the cell membranes to be lysed by detergent and first-strand

cDNA to be synthesized from total cytoplasmic poly(A)mRNA (28).

Amlification of HPRT cDNA and DNA sequencing. The experimental
conditions, optimized for preparing second-strand HPRT cDNA, amplifying
the cDNA 10"-fold using two Polymerase Chain Reaction (PCR) stages of 30-

cycles each, and sequencing the product directly using three sequencing

112
primers and a modified Sanger dideoxynucleotide procedure have been

described (28).

RESULTS

Deter-hing Onset of S Phase. Repair-proficient NFSLB9 fibroblasts
and repair-deficient XP128E fibroblasts were synchronized by release from
confluence (density-induced G0 state) and plated into a series of culture
dishes in Eagle’s MEM at 10‘ cells/cmz. At various times post-release, the
cells were pulse-labeled with tritiated thymidine (5 uCi/ml medium) for 20
min and the acid-precipitable counts were determined. As shown in Fig. 1,
DNA synthesis in both cell lines began 15-16 h following release. By 17

h both cell populations were actively incorporating tritiated thymidine.

Survival and lutation frequency; Synchronized populations of XP128E
cells were irradiated with 0.5- 0.6 J/m2 of UV light 17 h post-release
from confluence at the beginning of S phase, or in G, phase 6 h prior to
S, and assayed for survival and the frequency of TGr cells. Eight
replicate cultures in either S or G, phase of the cell cycle were
irradiated in each experiment, together with non-irradiated controls, and
three experiments were conducted for a total of 24 independent populations
of cells irradiated at the beginning of S phase and 24 p0pulations
irradiated 6 h before the onset of S phase. Cell survival ranged from 10-
34%. The frequency of TGr mutants per 106 clonable cells in populations
irradiated in early S phase ranged from 223 to 324; that in populations
irradiated 6 h prior to the onset of S ranged from 253 to 323. The
background frequency (in the absence of UV treatment) was less than 10 TGr
mutants per 10‘5 clonable cells. The cloning efficiency of the cells at the

time of selection was 15-22%.

113

114

Figure 1. Incorporation of 3H-thymidine into acid-insoluble material
following a 20 min pulse-label at the indicated times. Counts have been
normalized to 105 cells, and the background (300 cpm/IO5 cells) has been

subtracted. Populations were released from density-induced Goiat T = 0 h.

CPM x Io‘3(H)

N

115

 

 

o—o NFSL89
O—O XPIZBE

l l L l l

 

OO
4 8 l2 I6 20 24

HOURS AFTER RELEASE
FROM CONFLUENCE

Figure 1

28

 

(o—o>,__ou x wao

116

Repair-proficient diploid human fibroblasts were synchronized and
irradiated as described above, but using a fluence of 6.5 J/mz. Their
survival, which was the same whether irradiation was in early S phase or
in G, phase, ranged from 18-22%. The frequency of TGr mutants per 106
clonable cells in populations irradiated in early S phase ranged from 145
to 162; that in populations irradiated 6 h prior to the onset of S phase,
ranged from 67 to 76. The frequency observed in unirradiated cells was 5-
10 mutants per 10‘S clonable cells. The cloning efficiency of the repair-

proficient cells ranged from 50-60%.

Nucleotide sequence alterations in XP lutants. The coding region of
the HPRT gene of 42 independent TGr mutants was amplified by PCR and the
product was directly sequenced. Of this number, 22 were derived from
populations of XP128E cells irradiated at S phase and 20 from cells in G,
phase. The mutants could be distinguished as independent either because
the sequence alterations were unique, or they had been generated in
separate populations. The kinds of mutations seen in mutants derived from
cells irradiated in S phase, their location in the coding region, and the
consequence for the primary structure of the protein are tabulated in
Table 1. The majority were base substitutions. Among these, transitions
predominated, accounting for 68% (15/22) of the total. One mutant had an
insertion of 62 bp from intron 5 which very probably resulted from
inactivation of the 5’ splice site of intron 5 by a G~C --> A-T
transition and use of a downstream cryptic splice donor site (16). Three
mutants had an exon completely deleted. These deletions probably resulted

from base substitutions in splice sites, but because*we sequenced only the

1..

63:03.... .._. 63:85:52 .._.2 a

 

 

a 2.3..
n 33..
a .82..
.5323... 25 8:3 9.25...—
n :93...
.. =2:3... 3-8 .3333 .. . <10 . o .2... 88.. 8.3m 8....
50:32.. 0.... 3.3m
e 5.0 1...... 8.. S... 9:. o . 01. .< a a... «.83.
e =m< 1“? 2.... ...<u uo< .. . <10 . 0 a .88 ~83.
5< 1 < .5... 8.. 2.... .. . <10 . 0 .8
e 0.523 oz .55 8.. 9.... .. . <10 . 0 a 52...... 88 ”max
.1 m... 1...< ...<c .09. 2.... .. . <10 . 0 a an 2.3..
h 5< 1:2 ...<... 03 ......0 ._. . <10 . O a can cmmax
e a... 1:0 .2... <8 <8 ._ H «kuw a .52.... WNW :2...
e 50 120 ......< 5... ......... 0 . 010 . 0 a 3.. 2me
.. 2.0 1.5 ......... <8 .5... .. . <10 . o m 9H .32..
+2 =3 1.2... ......o ......H 35 0 . 01< . .. a 2n 2me
... 2... 1 _a> <8 9.... .92 < . e10 . 0 e at. GE...
.. 2.1.9: 9...... a: as. .. . <10 . 0 e E. 3......
.. S... 1am< 5.... S... .20 .. . <10 . .0 e 84 2.9.x
.. =5 1 20 coo can 92 e . <10 . 0 m o8 Nvmnx
.. am 192 .8... an: 68 < ...10 . 0 n n: 2%...
...z .62 10.. 9.... a: «.8 0 . 01< . .. N 8. 3me
..z :2 16.. 9.... 5.. S5 .. . <1< . ... N 2. «.me
e m... 120 <5 .au .35 %.«Hw “m N 52...... "Mn .83.
ez 3. 12.. ...<< .3 E... < . e10 . 0 N z. 33..
act—.3395 omam
coc=._=._..E_v once
so 3:263. 5.2.5:. :98 :35... E...
5“...“me Eon oEE< SEE—5.5m ._o 2:3. 90.0.5 :95 .553.— oaeazo ._o 09:.

 

8...... m 5 00.3.5... 3.3 QQNEX .o 2.0m :3: u... no .530. 95.3 o... E nous—0... 80:82.. so 32.32 can v.15! A MEM—4:.

118
coding region of the gene, the nature of the presumed intron mutations
which led to such exon deletions has not been evaluated.

As shown in Table 1 and diagrammed in Fig. 2, the base substitutions
were distributed throughout the exons, but with more located in exon 8
than would be expected on a random basis. The premutagenic dipyrimidine
photoproducts presumed to be responsible for the observed base
substitutions were primarily located in the transcribed strand (79%,
15/19).

Table 2 lists the sequence alterations found in 20 mutants isolated
from XP128E cells irradiated 6 h before the onset of S phase. The kinds
of mutations did not differ significantly from those induced at S phase,
nor did their location in the gene (Fig. 2). The majority contained base
substitutions, predominantly G-C --> A-T, i.e., 56% (9/16), and the
corresponding dipyrimidine photoproducts were primarily located in the
transcribed strand, 69% (9/13). There were 6 exon deletions resulting
from putative splice site mutations. One mutant (XUG20) contained a
mutation that did not occur at a dipyrimidine site, but involved a base
substitution and a deletion at an ACA site. This sequence has been
reported to form a complex photoproduct following UV radiation (2), but we
cannot be certain that this unusual photoproduct was the premutagenic
lesion, and therefore this mutation has been excluded from analysis of the

strand location of the premutagenic lesions.

Nucleotide sequence alterations in Iutants fron repair-proficient
cells. Our hypothesis predicts that the mutations induced in the repair-

proficient cells irradiated at the onset of‘S will resemble those found in

119

Figure 2. Location of independent HPRT mutations in XP128E cells
irradiated with 0.5 J/m2 UV light (254 nm) in early S phase (22 mutants)
or 6 hr before S phase, in G, phase (19 mutants). Mutation XUG20, which
did not occur opposite a dipyrimidine sequence, has been excluded. Exon
deletions are shown as open rectangles; base substitutions are indicated
by vertical lines; and tandem mutations are indicated by vertical arrows.
A base substitution in intron 5 of the transcribed strand is indicated by
an asterisk. Strand assignments were made on the basis of which strand
contained the dipyrimidine sequence which presumably targeted the
mutation. There was no statistically significant difference between cells
irradiated in S phase or in G, phase in the kinds of mutations induced or

in the strand distribution of the potentially mutagenic photoproducts.

120

. a 0.5.5.:

 

 

 

 

 

895345202 .. _ 02:16
ozqmem
895345 .u _ a . _ a
_
$01.5 nU
0095345202 _ = mmazaim
ozamem .
89533: _ .ilfilal. : : _ _ . .7
m _ m _ N _ mu —m_ V _ M _ _— "ZOxw
lll— _ _ _ _ — Ill
.3 00... 08 00v 00m 08 om. .. "ES. 09...

m44m0 HZMGEMD lm_<n.wm
>3

121

.0033... 000.. 0...... 0.000 ...x. 0...... .0....0 00:... 03.50.00 0.. .0530 0...- 00:03.2...»

0...... 0... .0 0.3.... 2.... .00...E.0.00 0.. .0530 03.0.0... 0... c. 0032.... 0......» 0... 6.0.0.0... .<b< 0029.... 00.00. 0.009.550... 0... .0... 0.8.00 0.. .0058 03 a
000.32.83.00 ....2 .00....0000... .... u

 

 

0 0.0....
0 08:...
m 200x
n «NODX
< .00....
< :03.
02.0.0... 0.... 00...... 02.5....
e =.0 1 ...0 ..<.. <00 <00 0 . < 1 0.0 a 00. 0.0....
e 0.... 1 ...0 ...<... <00 :0 0 . < 1 0. . 0 .. .0. :03.
p 02 1 ...0 ...<.. <00 :0 .0. < 10. 0 .. .0. 20....
.. 02 1 ...0 ...<... <00 <...0 .. . < 1 0.0 0 00. .00..
h 2... 1 00.. ....< 0...... 0.0 < . .. 1 0 .0 m an 0.0....
I 0.088.... 0...< 0<... ...<0 < . e 1 ... H < m SN 80....
a O U hmvN
.. .. .. . < 1 0 . 0 SN
.0 m .. 1 .0 000 000 0<< .0.< 10. 0 n .505. .08 :03.
..z 5< 1 0.. 0...< ...H< <...0 .. . < 1 < ... N N. :03.
..z E. 1 8.. 0...< <00 <00 0. 0 1 < . .. N NN. N53.
0 20 1 ..0 <8 <00 ...<0 0 . < 1 0 . 0 N 0.. 20:0.
.. .20 1 20 ...<0 00 ...00 < . e 1 0 . 0 N 0.. :03.
..z &< 1.... ...00 ...<0 ...<0 0. 0 1 < ... N .N0 ..Noax
.. .. . < 1 < . .. N0
.2 .. 0 1 8.. ........ <0. ...<0 0. < 1 < . .. N 505.. ..0 N83.
2.01“? 00< 000 0...< e.< 10.0 ...
.. h.__ 1.0.2 00< 000 0.2 e . < 1 0. 0 . .5052 .n 3.0....
00.5.33... 00....
u0...0......>...0 00......0 00000—000 00.3.0... :30 ..o. .00 .530... 0....
00.00.... 5.3 0......m 0.00 o......< 2.0.58.5 .o 0...... 00.0.09 :95 ... m 09.2.0 ..0 2......

 

00...... .0 :. 00.0.00... 2.00 mafia—x ..o 0:00 Ext: 0... ..0 00.00. 00.000 0... 0. 000:0... 0.5.3.0... ..0 2.0.30... 0:0 0.00.0. .N mama...

122

the XP128E cells treated in S or G1 phase. As shown in Table 3, which
presents the HPRT sequence alterations found in 22 mutants derived from
such populations, this is the case. The majority of the mutants (IS/22)
contained single or tandem base substitution, for a total of 18 base
substitutions in the 15 mutants; of this number 11(61%) were G‘C --> A-T
transitions. With the exception of the A-T --> C-G base substitution in
mutant NUSlB, all base pair substitutions occurred at dipyrimidine sites
where photoproducts could form, and 71% of these photoproducts (10/14)
were located in the transcribed strand. The substitution in mutant NU513
may represent a pre-existing background mutation since 8 of the 9 base
substitutions we found in unirradiated background mutants involved A-T
base pairs (27).

One S phase mutant had a 10 bp deletion in exon 8; the premutagenic
lesion resulting in this deletion could not be determined. Seven putative
splice site nmtations which resulted in complete exon deletions were
found. The strand distribution of these mutations, which is shown in Fig.
3, resembled that found in mutants from either population of XP128E cells.

Table 4 lists the spectrum of mutations analyzed from 20 mutants
isolated from populations of repair-proficient fibroblasts irradiated in
G1 phase, 6 h before the onset of scheduled DNA synthesis. The majority
were base substitutions, mainly G-C --> A-T (47%), but the fraction of the
premutagenic photoproducts located in the transcribed strand was greatly
reduced, i.e., 20%, rather than 71%. As shown in Fig. 3, there were few
mutations resulting from lesions in the transcribed strand. For
convenience, Table 5 compares the frequency of the various kinds of base

substitutions observed in the 4 different populations.

123

.27. 2:25:36 0 .a ...qu 8.. E: .555... z... 3.... 3.520.. 0.. 3....8 :83— u_..oua.::.o... o... 9.3.2.30 9...... 0:... a
63:03.... .._. €3t3...2...oz ..—.2 a

 

 

.37.:

52.2% .3... 8m?“
.0 .O

.. ~.m:2

.. e.m:z

. ..m:z

.. -m:z

. :m:z

n ...m:z

v 8:2
.8223... 0.7. 03...... 22.3...—

.r 5< T a... PE. .25 8.. .. . < T o . o a .8 8:2

s .8 T .5. ...<0 8.. 9:. o . u T o . o a 8.. 8:2

9 asses". So 8.. 9:. b . < T w ”w a mum 8:2
+2 :5 T of. 8.. 0H... 9:. o . u T < 4 a can 232
.5 3m T 2.. <8 <8 8.. < . .T T o . u .. 3. 5:2

b .3 T .5. 8.. a“... :2 %H«Hw ”w h 5.53 an” 5:2

..oT a... 84 52 0.... u . o T .F . < 3%
.. 09.29 oz 8.. 32 2b 0 . c T e . < .. :32... :3. «8:2
.r o: T .92 95 a: a: e . < T u . o e ..3 «8:2
< . ,F T o . u so 3..

.5 :3 Ten. .22 <3 .52 < . ._. T o . u o .53 :3. .232
b 2... T .6: 98 8.... 8... To . u T u . o o ..3. «5:2
9 m3 1.5 8... ca 92 %H«Hw .w . 50...... WWW :32
e «E T ..o .5... <8 2... e . < T u . o a S.

.r 0.32.. oz .5... 4.8 coo o . o T b . < m 2. 5:2
92 S. T :3 ...2 can <8 0 . u T < . ... N «N. 3:2
.I 9% Te? ...8 H5. ...<0 0 . u T .r . < N 3 ..m:z

333:9...” 8.5
$523.35 0935 3.8.63 3:32.. ..98 2.3:... 3:.
$23.... 22. o.._...< 9.6.59.5 3.2.3. 3.0.00 :95 .553.— o9.....u no 2.3.
5.3 2.25

 

03.... m 5 3.22.... £8 22...:
...o_oc:...-._a..u. ..o 2.0m h$~= o... ..o :23. 9.58 o... E 532:: 2.2.8:... ..o 30:32 0.3 £2.5— .m mamtfi.

124

Figure 3. Location of independent HPRT mutations in repair-proficient
human fibrobiasts irradiated with 6.5 .J/m2 uv iight in eariy 5 phase (21
mutants) or in G1 phase, 6-7 hr before S phase (20 mutants). Mutation
NUSl3, which did not occur opposite a dipyrimidine sequence, has been
exciuded. Mutations are indicated as in Fig. 1. In 71% of the mutants
from ceiis irradiated in S phase, the premutagenic iesion was Tocated in
the transcribed strand; in oniy 20% of the mutants from ceiis irradiated
in G1 phase was the premutagenic Tesion Tocated there. This switch in
strand bias indicates preferentiai repair of the transcribed strand of the

HPRT gene.

 

 

125

. n 0.26.:

 

 

 

 

 

 

 

nHIIIIIhIIIIIU
mKUIHO “U “HUM“.
ozqmcm
ommaomquzoz a _ _ . _ _ slfilall mquaia
oz<mpm
ommiomzaE =
oz<Em u
BmEomzaEzoz _ _ = mmaxaim
ozqmpm
895?qu a“ _ “a _ _ e
m m k. w n. v m _ "zoxu
Iii. __ _ d r _q _ F _ _ _ .iil
Em 08 0ko 00¢ 00m cow 8. _ "can. “mam

mun—mu .rzmzoloma I gamma
>3

126

63528. .._. ”cantons-=52 .._.z .

 

 

a 20:2

a 2032

v «.032

v «032

v 2032

.8252: 2.... 3:3 252:.—
..z 5:858.— .52 SH 58 0 . o T« ”.2... a 52...... ”% 8:2
#2 soi— T 2E .50 99H 03 0 . U T < .._. h «an 3032
.rz :3 T 2.. 9: .35 a: < .._. T 0 . U a «.9. 20:2
.rz com T sui— .E.< was 0.5 0 . U T < . h n Nam £022
...z o: T .2... SE. .5: <2. < . ._. T 0 . 0 v :2 oODZ
2.2 3: Tea. 0...< 3H So 0. 0 T < . ._. m 3 :0:2
._. :5 T .20 .5... ago wow < . ._. T 0 . 0 m 2N 3:2
e 20 T 20 one 8o 22 2. . < T 0 . 0 N SN :82
b _a> T o: a: a: 2.0 0 . 0 T h . < N «2 5:2
e2 .2..— T 2.: ...<0 ea .22 «MM w “w N .523 ”Mm 30:2
2.2 :3 T 2... a: 5:. use ._. . < T < ..r N a: 5:2
ez 2... T 8.. .52 Su «.2 « H w M w ”w N .522 ”MM N0:z
ez 3m T :0 52 8H 2.... h . < T < . ._. N 8 20:2
2.2 :3 T 2.— .Su <3 5.6 < . h T 0. 0 N 3. 20:2
.5 «2 T 2.. you <8 53 0 . 0 T 0. 0 N 9. 8:2
5:333...“ 095
cofiwtucotwfic owe—Eu 85:03 .5223: :93 2.2::— 0::
...w3._.=m..m :8 8:5 usesga co 25. 8.0.0: cam 5:3.— oasza .a 25.

 

82.. :0 5 022.55 £3 223:
Eotcoaéaao. ..o 2.0» .532 on. ..o .532 9:03 2: :_ .0325. 20:22:. .3 25:30. :5 225. .v mam—<2.

127
TABLE 5. Types of base substitutions induced in the coding region of the
HPRT gene in repair-deficient XP128E fibroblasts and repair-proficient

ceiis irradiated in S phase and in G, phase

 

Number of substitutions observed

 

 

Type of base XP128E ceiis Repair-proficient geiis
substitution S phase G, phase 5 phase G, phase
Transitions:

G-C --> A-T 15 9 11 8
A-T --> 6°C 1 1 5 5
Transversions:

G-C --> C-G 1 0 2 1

G-C --> T-A 2 2 o 1

A-T --> C-G 2 1 0 0
A-T --> T-A 1 3 o 2

TOTAL 22 16 18 17

 

DISCUSSION

The mutation and survival data for the repair-proficient fibroblasts
and the XP128E cells are consistent with previously published reports from
this laboratory (9). As discussed by those investigators, the higher
survival of the colony-forming ability'in the repair-proficient cells than
in XP128E cells indicates that following irradiation, there is time for
DNA repair before "reproductive death" occurs. But the lack of variation
in survival with the cell cycle indicates that the time available for such
repair is not time prior to S phase. Instead, it probably reflects time
before the cells experience a need for certain critical proteins that
cannot be produced because of blocks in transcription.

The present data show that in human cells that have no time (or no
ability) to excise UV photoproducts, the strand distribution of
potentially mutagenic lesions averages 75% transcribed:25% nontranscribed,
but in repair-proficient cells given 6 h for repair before S, the
distribution is reversed to 20:80. We argue that such a switch in strand
distribution is best explained by preferential repair of lesions from the
transcribed strand of the HPRT gene. There was no statistically
significant difference in the frequency or kinds of mutations induced, or
in the strand distribution of the premutagenic lesions in the mutants
derived from populations of XP128E cells irradiated in either phase of the
cell cycle. The data in Fig. 2 suggest that in S phase cells, mutations
induced by UV are more likely to occur in the 3’ half of the HPRT gene
(72%) and in G, phase cells, they are more evenly distributed (47% in the

3’ end). This apparent change cannot be due to DNA excision repair since

128

129

XP128E cells are virtually devoid of such repair. This difference was not
seen in XP128E cells following exposure to benzo(a)pyrene diol epoxide
(BPDE) (3) and so it might reflect differential accessibility of the gene
to UV damage during S and G, phase. Nevertheless, it is important to note
that there was no cell cycle-related change in the sttggg distributign of
UV-induced premutagenic lesions in these>XP128E cells. Furthermore, there
was no statistically significant difference between the XP128E cells in S
or G, and the NFSL89 cells in S, in the strand distribution of premutagenic
lesions. Therefore, the switch in strand bias observed in mutants derived
from the repair-proficient cells irradiated in S or G, phase can be
attributed to preferential repair of the transcribed strand.

This conclusion, which is based on a comparison of repair-proficient
diploid human fibroblasts irradiated in two different phases of the cell
cycle, agrees with that of Vrieling et al. (25). These investigators
based their conclusion on a comparison of mutants from asynchronous
cultures of a repair—proficient and a repair-deficient Chinese hamster
cell line. Both studies support the work of Mellon et al. (14),
indicating that UV-induced dipyrimidine dimers are preferentially removed
from the transcribed strand of an actively transcribed gene in a human
cell line. Although it is not yet clear how such targeting of the
excision repair complex to a particular strand occurs, preferential repair
of the transcribed strand is abolished in mammalian cells by inhibition of
RNA polymerase II with a-amanitin (4). Conversely, in E. coli, induction
of a gene has been shown to induce preferential repair of the transcribed
strand (13). These studies suggest that there is a coupling of

transcription with nucleotide excision repair of UV-induced lesions. As

130
suggested by Mellon et al. (14), it is possible, for instance, that the
stalled RNA polymerase acts as a signal to target the repair complex to
the site of the transcription-blocking lesion on the transcribed strand.

Glickman and colleagues (5, 6) have suggested from data in E. coli
and in CHO rodent cells that UV-induced G-C --> A-T transitions result
predominantly from 6-4 Py-Py lesions. However, Vrieling et al. (25)
suggest that such transitions result primarily from Py-Py dimers because
they found that 100% (15/15) of the UV-induced base substitutions were G-C
—-> A-T in a Chinese hamster cell line (VH-l) that is incapable of
repairing Py-Py dimers, but repairs 6-4 Py-Py lesions at a rate 50% that
of their repair-proficient V-79 hamster cell line. Since in human cells
virtually all 6-4 Py-Py lesions are repaired within 6 h (15), if 6-4 Py-Py
lesions uniquely cause a particular kind of base substitution, and if
these substitutions constitute a substantial fraction of the UV—induced
mutations seen in these cells, we should have observed a significant
difference in the kinds of base substitutions in mutants from repair-
proficient cells irradiated in S and those in G,. This was not observed
in our study. Therefore, we conclude that in human cells G-C --> A-T
transitions result from both unexcised Py-Py dimers and 6-4 photoproducts.
This could explain why such transitions are the most common mutations
induced by UV radiation in a variety of organisms (5, 8, 10, 21, 24).

As noted above, the distribution of dipyrimidine sites between the
transcribed and the nontranscribed strands in the coding region of the
human HPRT gene is 59:41. If all lesions were equally mutagenic, and if
each mutation resulted in a nonfunctional protein, 59% of the mutations in

XP128E cells or repair-proficient cells irradiated in S should have arisen

131
from photoproducts in the transcribed strand, and 41% from lesions in the
complementary strand. Instead, we found a distribution of 75:25.
Vrieling et al. (25) found a still larger shift, 90:10, and suggested that
the bias reflected a more error-prone polymerase (i.e., polymerase delta)
replicating the transcribed strand of the gene. As discussed by these
investigators, the origin of replication of the HPRT gene is putatively
located in the first intron (22), implying that the transcribed strand of
most of this gene is also the leading strand during DNA replication and,
therefore, is synthesized by DNA polymerase delta (19). If polymerase
delta were more error-prone than polymerase alpha, lesions in the
transcribed strand would be more likely to result in mutations than
lesions in the complementary strand. However, there is at least one other
possible explanation. If photoproducts involving cytosine were more
mutagenic than those involving thymine, and if in the coding region of the
HPRT gene, mutations involving a C-G base pair with the C in the
transcribed strand were particularly likely to result in a nonfunctional
enzyme, this would explain the observed bias. We found that 71% of the
UV-induced base substitutions (52/73) involved C-G base pairs, and among
mutations targeted to lesions in the transcribed strand, 89% (40/45)
involved C-G base pairs. In contrast, only 43% (12/28) of the mutations
targeted to lesions in the nontranscribed strand involved C-G base pairs.
Furthermore, Chen et al. (3), using the same system, recently showed that
when guanine substituted with a bulky residue~of BPDE was the premutagenic
lesion in the transcribed strand of HPRT, the distribution of premutagenic
lesions was lower than predicted, i.e., 25% instead of 40%. This argues

against an error-prone delta polymerase causing the increase above the

132
expected strand distribution observed in the UV studies. Rather, the
increase seen in the XP128E cells and the repair-proficient cells in S
more likely reflects selection for nonfunctional HPRT protein.

In summary, we have shown that if diploid human fibroblasts are able
to excise UV-induced photoproducts before DNA replication occurs, they
exhibit a significant switch in the strand distribution of premutagenic
lesions. The ratio of lesions in the transcribed to nontranscribed strand
is reversed from 75:25 to 20:80, consistent with preferential excision

repair of UV photoproducts from the transcribed strand of the HPRT gene.

10.

11.

REFERENCES

Bohr, V. A., D. S. Dkulato, and P. C. Hanawalt. 1986. Survival of
UV-irradiated mammalian cells correlates with efficient DNA repair
in an essential gene. Proc. Natl. Acad. Sci. USA 83:3830-3833.

Bourré, F., G. Renault, and A. Sarasin. 1987. Sequence effect on
alkali-sensitive sites in UV-irradiated SV-40 DNA. Nucleic Acids
Res. 15:8861-8875.

Chen. R.-H., V. M. Maher, and J. J. McCorIick. 1990. Effect of
excision repair by diploid human fibroblasts on the spectra of
mutations induced by (:)-7B,8a-dihydroxy-9a,10a-epoxy-7,8,9,10-
tetrahydrobenzola]pyrene in the coding region of the HPRT gene.
Proc. Natl. Acad. Sci. USA 87:8680-8684.

Christians, F. C., and P. C. Hanawalt. 1990. Reduced strand-
specific repair of pyrimidine dimers in a-amanitin treated mammalian
cells. Proc. Amer. Ass. Cancer Res. 31:2.

Drobetsky,. E. A., A. J. Grosovsky, and B. M. Glickman. 1987. The
specificity of UV-induced mutations at an endogenous locus in
mammalian cells. Proc Natl. Acad. Sci. USA 84:9103-9107.

Slick-an, B. H., R. M. Schaaper, M. A. Haseltine, R. L. Dunn, and D.
E. Brash. 1986. The C-C (6-4) photoproduct is mutagenic in E.
coli. Proc. Natl. Acad. Sci. USA 83:6945-6949.

Gross-am, A., V. M. Maher, and J. J. McCorIick. 1985. The
frequency of mutants in human fibaroblasts UV-irradiated at various
times during S-phase suggests that genes for thioguanine and
diphtheria toxin resistance are replicated early. Mutat. Res.
52:67-76.

Mauser, J., M. M. Said-an, K. Sidur, and K. Dixon. 1986. Sequence

specificity of point mutations induced during passage of a UV-
irradiated shuttle vector plasmid in monkey cells. Mol. Cell Biol.
:277—285.

Konze-Thonas, 8.. R. M. Hazard, V. M. Maher, and J. J. McCorIick.

1982. Extent of excision repair before DNA synthesis determines the

mutagenic but not the lethal effect of UV radiation. Mutat. Res.
:421-434.

Kunz, D. A., M. K. Pierce, J. R. A. His, and C. M. Giroux. 1987.
DNA sequence analysis of the mutational specificity of UV light in
the Sup4-o gene of yeast. Mutagenesis 2:445-453.

McCorIick, J. J., and V. M. Maher. 1981. Measurement of colony-
forming ability and mutagenesis in diploid human cells. In E. C.
Freidberg and P. C. Hanawalt (eds.), Techniques in DNA repair, vol.

133

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

1. Marcel Dekker, New York, pp. 501-521.

Mellon, I., V. A. Bohr, C. A. Slith, and P. C. Hanawalt. 1986.
Preferential DNA repair of an active gene in human cells. Proc
Natl. Acad. Sci. USA 83:8878-8882.

Mellon, I., and P. C. Hanawalt. 1989. Induction of the Escherichia
coli lactose operon selectively increases repair of its transcribed
DNA strand. Nature 342:95-98.

Mellon, I., G. Spivak, and P. C. Hanawalt. 1987. Selective removal
of transcription-blocking DNA damage from the transcribed strand of
the mammalian DHFR gene. Cell 51:241-249.

Mitchell, D. L., C. A. Maipek, and J. M. Clarkson. 1985. (6-4)
Photoproducts are removed from the DNA of UV-irradiated mammalian
cells more efficiently than cyclobutane pyrimidine dimers. Mutat.
Res. 143:109-112.

Nelson, K. K., and M. R. Green. 1990. Mechanism for cryptic splice
site activation during pre-mRNA splicing. Proc. Natl Acad. Sci.
USA. 87:6253-6257.

Patton. J. D., L. A. Rowan, A. L. Mendrala, J. M. Howell, V. M.
Maher, and J. J. McCorlick. 1984. Xeroderma pigmentosum
fibroblasts including cells from XP variants are abnormally
sensitive to the mutagenic and cytotoxic action of broad spectrum
simulated sunlight. Photochem. Photobiol. 39:37-42.

Petinga, R. A., A. D. Andrews. R. E. Tarone, and J. M. Robbins.
1977. Typical xeroderma pigmentosum complementation group A
fibroblasts have detectable ultraviolet light-induced unscheduled
DNA synthesis, Biochim. Biophys. Acta. 479:400-410.

Prelich, G., and B. Still-an. 1988. Coordinated leading and
lagging strand synthesis during SV40 DNA replication in vitro
requires PCNA. Cell 53:117-126.

Ryan, P. A., J. J. McCorIick, and V. M. Maher. 1987. Modification

of MCDB-llO medium to support prolonged growth and consistent high
cloning efficiency of diploid human fibroblasts. Exp. Cell Res.
72:313-328.

Schaaper, R. M., R. L. Dunn, and B. M. Slick-an. 1987. Mechanisms
of UV-induced mutations: Mutational spectra in the E. coli lacI
gene for a wild type and an excision repair deficient strain. J.
Mol. Biol. 198:187-202.

Sykes, It. C., IL. Lin, SL .1. Huang, It. E. Fro-son, qul A. (L
Chinault. 1988. Yeast ARS function and nuclear matrix association
coincide in a short sequence from the human HPRT locus. Mol. Gen.
Genet. 212:301-309.

134

23.

24.

25.

26.

27.

28.

135

Tsutsui, T, J. C. Barrett, and P. D. P. ‘l’s’o. 1978. Induction of
6-thioguanine- and ouabain-resistant mutations in synchronized
Syrian hamster cell cultures during different periods of the S-
phase. Mutat. Res. 52:255-264.

Urlaub, P. J., P. J. Mitchell, C. J. Ciudad, and L. A. Chasin.
1989. Nonsense mutations in the dihydrofolate reductase gene affect
RNA processing. Mol. Cell. Biol. 9:2868-2880.

Vrieling. M., M. L. Van Roaijen, M. A. Groen, M. Z. Zdzienicka, J.
M. I. M. Simons, P. H. M. Lohman, and A. A. van Zeeland. 1989. DNA
strand specificity for UV-induced mutations in mammalian cells.
Mol. Cell Biol. 9:1277-1283.

Matanabe. M., V. M. Maher, and J. J. McCorlick. 1985. Excision
repair of UV- or benzo[a]pyrene diol epoxide-induced lesions in
xeroderma pigmentosum variant cells is "error-free". Mutat. Res.
146:285-294.

Yang, J.-L., R.-M. Chen, V. M. Maher, and J. J. McCorIick. 1991.
Kinds and location of mutations induced by (:)-7B,8a-dihydroxy-
9a,10a-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene in the coding region
of the hypoxanthine (guanine) phosphoribosyltransferase gene in
diploid human fibroblasts. Carcinogenesis (in press).

Yang, J.-L., V. M. Maher, and J. J. McCorlick. 1989. Amplification
and direct nucleotide sequencing of cDNA from the lysate of low
numbers of diploid human cells. Gene 83:347-354.

CHAPTER V

Preferential repair and strand-specific repair of benzo[a]pyrene diol

epoxide adducts from the HPRT gene of diploid human fibroblasts

Ruey-Hwa Chen', Veronica M. Maher', Jaap Brouwerz,

and J. Justin McCormick1

1Carcinogenesis Laboratory - Fee Hall

Department of Microbiology and Department of Biochemistry

Michigan State University, East Lansing, MI 48824-1316 (U.S.A.)

2Laboratory of Molecular Genetics
Department of Biochemistry
Gorlaeus Laboratories

Leiden University

P. 0. Box 9502-2300

RA Leiden, THE NETHERLANDS

136

SUMMARY

He showed previously that if excision repair-proficient human cells
are allowed time for repair before onset of S phase, the premutagenic
lesions formed by (:)-7B,8a-dihydroxy-9a,10a-epoxy-7,8,9,10-
tetrahydrobenzo[a]pyrene (BPDE) are lost from the transcribed strand of
the hypoxanthine(guanine) phosphoribosyltransferase (HPRT) gene faster
than from the nontranscribed strand. No change in strand distribution was
seen with repair-deficient cells. These results suggest strand-specific
repair of BPDE-induced DNA damage in human cells. To test this hypothesis,
we measured the initial number of BPDE adducts formed in each strand of
the actively transcribed HPRT gene and their rate of repair, using the
UvrABC exinuclease in conjunction with Southern hybridization and strand-
specific probes. We also investigated the possibility of preferential
repair of such damage by comparing these rates with the rate of loss of
damage from the inactive 754 locus located in the same X-chromosome.
Following exposure to 1.0 or 1.2 uM BPDE, the frequency of adducts formed
in either strand of a 20 kb fragment that lies entirely within the
transcription unit of the HPRT gene was similar. The frequency of adducts
formed in the 14 kb 754 fragment was slightly lower ("20%) than in the
HPRT gene. In contrast to the frequency of adduct formation, the rate of
repair in the two strands of the HPRT fragment differed significantly.
Within 7 hr after treatment with 1.2 uM BPDE, 55% of the adducts had been
removed from the transcribed strand, while only 26% had been removed from
the other strand; after 20 hr of repair, these values were 87% and 58%,
respectively, confirming the prediction of our mutagenesis studies. In

contrast to the rapid rate of repair in the HPRT gene, only ~14% of the

137

138
BPDE adducts were lost from the 754 locus, a value even slower than the
rate of loss from the overall genome (i.e., 38%). Our results demonstrate
strand-specific and preferential repair of chemically-induced BPDE bulky
adducts in human cells. They suggest that the heterogeneous repair of BPDE
adducts cannot be accounted for merely by the greatly increased rate of
the repair specific to the transcribed strand of the active genes, and

point to a role for the chromatin structure.

IMTRDDUCTIOM

Recent studies have demonstrated that UV-induced cyclobutane dimers
are repaired more rapidly (i.e., preferentially) from thei actively
transcribed genes of rodent and human cell lines in culture than from
inactive genes or from nontranscribed segments of DNA or from the overall
genome of the cells (1-4). Furthermore, Mellon et al. (5) showed that in
these cells such lesions are excised faster from the transcribed strand of
theiactively transcribed dihydrofolate reductase (dhfr) gene than from the
nontranscribed strand of that gene. These results led to the investigators
to hypothesize that nucleotide excision repair is coupled to gene
transcription (5). Indeed, a factor responsible for such coupling has
recently been partially purified from E. coli cell extract (6).

Although the existence of such heterogeneous repair of cyclobutane
dimers from the DNA of mammalian cells is well documented, the
applicability of this model to bulky chemical residues covalently bound to
DNA (adducts) is unclear and controversial. The results of our studies(7,
8) of the effect of excision repair in diploid human cells on the strand
distribution of the principal premutagenic lesions induced by (:)-7B,8a-
dihydroxy-9a,10a-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene (BPDE) showed
that in the absence of repair, such lesions are located in both strands of
the coding region of the hypoxanthine (guanine) phosphribosyltransferase
(HPRT) gene, but after 12 hr of excision repair, no premutagenic lesions
remains in the transcribed strand. These results provide strong biological
evidence of strand-specific repair of BPDE adducts from the human HPRT
gene.

Such strand-specific repair could account for the strand bias for

139

140

BPDE-induced premutagenic lesions observed by Mazur and Glickman (9) in
the adenine phOphoribosyltransferase gene and by Carothers and Grunberger
(10) in the dhfr gene of excision repair-proficient Chinese hamster cell
line (CHO cells), although these investigators did not demonstrate that in
the absence of excision repair premutagenic BPDE lesions were distributed
in both strands of their target genes. However, using the technique
originally developed by Bohr et al. (1) to detect lesions initially
present or remaining in specific genes and using the E. coli UvrABC
exinuclease to recognize and incise bulky DNA adducts, Tang et al. (11)
demonstrated that there is no difference in the rate of repair of DNA
damage induced by a structurally related compound, N-acetoxy-Z-
acetylaminofluorene, from the active and nonactive (nontranscribed)
regions of the dhfr gene of CHO cells. More recently, Tang and his
colleagues (12) reported that BPDE induced DNA adducts are also not
preferentially removed from such regions of the dhfr gene of CHO cells.

The present study was designed to directly measure the ability of
human cells to remove BPDE adducts from either strand of the HPRT gene to
see if strand-specific repair occurred. We also wanted to test for
preferential excision repair of BPDE adducts by comparing the rate of
their removal from an active gene (HPRT) and an inactive locus (13) on the
same X-chromosome (the 754 locus). To investigate these questions, we
synchronized normal human fibroblasts and treated them with BPDE in early
G, phase so that there could be a long period for excision repair before
the onset of semi-conservative DNA replication. The cells were harvested
immediately or after various times after treated, and the DNA was

extracted and assayed for lesions in the HPRT gene. It has been

141

extracted and assayed for lesions in the HPRT gene. It has been
demonstrated that the UvrABC exinuclease is able to specifically and
quantitatively incise at least 80% of BPDE-DNA adducts in human genomic
DNA (14). Using this enzyme, in conjunction with the Southern blotting and
hybridization with probes specific for the individual strands of the HPRT
gene, we quantified the initial numbers of BPDE-DNA adducts formed in the
two strands and the numbers remaining at various times post treatment.

The results showed that there was no significant difference in the
initial frequencies of BPDE-DNA adducts formed in the two strands of the
HPRT fragment, but the rate of removal of adducts from the two strands was
markedly different. The transcribed strand was repaired significantly
faster than the complementary strand. Furthermore, the rate of excision

repair of BPDE adducts from the 754 locus was slower still.

MATERIALS AND METHODS
ll ul ur n ll n hr ni i n Diploid human male
fibroblasts, derived from foreskin material of a neonate (15), were
cultured in Eagle’s medium containing 10% supplemented bovine calf serum
(HyClone, Logan, UT). Such cells were driven into the Go state by density
inhibition and nutrient depletion as described (16). To stimulate the
cells to re-enter the cell cycle, they were released from confluence, and
plated in fresh culture medium at a density of 10‘ cells/cmz. The time of
onset of S phase following release from G,, was determined by the
incorporation of 3H-thymidine into acid-insoluble material as described
(16).
BPDE t - i 1 Cells were treated
5 hr after release from G,, (in early G, phase) to allow them to have a long
repair period before DNA replication began. They' were rinsed with
phosphate-buffered saline (PBS) and then treated with BPDE as described
(7), using a dose of 1.0 or 1.2 uM. After treatment, cells were either
lysed immediately or incubated for up to 24 hr in fresh culture medium.
W After a specific repair period,
cells were washed with PBS and lysed in lysis buffer containing 50 mM
Tris.HCl (pH 8.0), 10 mM EDTA, 100 mM NaCl, 0.5% sodium lauryl sarkosine,
and 100 ug/ml proteinase K. The cell lysates were incubated at 50°C for 8-
16 hr. The DNA solution was extracted with a equal volume of phenol,
phenol/chloroform (1:1), and chloroform. DNA was then precipitated with
ethanol, resuspended in TE buffer (10mM Tris.HCl, pH7.5 and lmM EDTA), and
treated with RNase A (50 ug/ml) for 2 hr at 37°C. The DNA was re-extracted

as above and precipitated with ethanol. After centrifugation, the DNA

142

143

pellet was dissolved in TE buffer at a concentration of 0.2 ug/ul. The
purified DNA was digested with restriction enzyme BamHI or EcoRI (5
units/ug DNA) at 37°C for 3 to 5 hr. The completion of digestion was
verified by electrophoresis of the DNA samples on agarose minigels. The
digested DNA was purified by extraction and precipitation as above.

u!:Aߤ_ggg1stgg_tgggt1ggs The UvrA, UvrB and UvrC proteins were
prepared as described (17). 10 ug of DNA derived from cells which had been
allowed various lengths of time to repair damage was used for repair
analysis. To serve as an internal standard, 5 pg of linearized plasmid DNA
containing a fragment of the gene to be probed was added into each DNA
sample. This mixture was then separated into two equal parts; one was
exposed to UvrABC exinuclease, the other to the exinuclease buffer alone.
The exinuclease reaction mixture was composed of 5 ug restricted genomic
DNA plus 17 pmoles of each of the three subunits of UvrABC in a final
volume of 250 ul buffer containing 50 mM Tris.HCl (pH7.5), 10mM MgClZ, 75
mM KCl, ZmM ATP, 1 mM dithiothreitol, and 1 ug/ul bovine serum albumin
(BSA). At the end of 1 hr incubation at 37°C, the reaction was stopped by
adding 0.1 ug/ul proteinase K and 0.1% Sodium dodecyl sulphate ($05) to
each sample and incubating for 1 hr at 42°C. The DNA was purified by
ultrafiltration using Centricon-30 tubes and precipitated with ethanol.

 

The DNA samples were dissolved in 90% formamide containing 0.025%
bromophenol blue and incubated at 37°C for 30 min to denature the DNA.
Immediately after incubation, the DNA samples were electrophoresed at 2.5
V/cm for 16 hr in a 0.6% agarose gel. After electrophoresis, the DNA in

the gel was stained and depurinated with acid using standard procedures

144

(18). The DNA was then transferred to a Zeta-probe GT membrane (Bio-Rad,
Richmond, CA) using the conditions recommended by the manufacturer.
Hybridizations were performed in 20 ml of solution containing 50%
formamide, 5X SSPE (10mM NaCl, 10mM NaPO‘ (pH 7.8), lmM EDTA), 0.5% 505,
0.02% Ficoll, 0.02% polyvinylpyrolidone, 0.02% BSA, denatured salmon sperm
DNA (50 ug/ml), E. coli t-RNA (100 ug/ml) and 2-3 x 10’ cpm 32P-iaheied
probe at 42°C for 20-24 hr. The riboprobes used for strand-specific
hybridizations to the BamHI fragment covering the 5’-half of the HPRT
genomic DNA were generated using the method described by Melton et al.
(19), with the modifications suggested by Boeringer-Mannheim Biochemicals.
DNA probes were labeled with 3ZP-dNTPs by random primer extension (20).
After hybridizations, the membranes were washed at 60°C with a final wash
step at 0.1x SSPE, 0.1% SDS and exposed to Kodak XAR-S X-ray films using
intensifying screens.

WW Plasmid pGZPa (21), which was constructed by
subcloning a 1.4 kb EcoRI-XhoI fragment containing sequences from the
first intron of the human HPRT gene into the vector pGEMZ, was kindly
provided by Dr. A. C. Chinault (Baylor college, Houston, TX). This plasmid
contains the SP6 and T7 promoters. RNA transcripts synthesized by the SP6
and T7 polymerases hybridize with the transcribed and nontranscribed
strands of HPRT" genomic fragment, respectively. The .2.0 kb HindIII
fragment of the 754 locus (22) was a gift from Dr. L. H. F. Mullenders
(Leiden University, The Netherlands).

nggtttgtjgns The intensities of bands were quantified using the
Bio-Image Visage 110 system (Millipore). The intensity of the full-length

fragment band was normalized with the internal standard band referred to

145
above. The average number of UvrABC sensitive sites per fragment was
calculated by the Poisson distribution equation as described by Bohr et
al. (23). These calculations took into consideration the nonspecific
incisions produced by the UvrABC enzyme complex, which ranged from 0.05 to

0.15 incision per 10 kb in the present experiments.

RESULTS

Determining delay in enset ef DNA replieatjen indueed by DPDE,
Repair-proficient human fibroblasts were synchronized by release from
confluence and plated into a series of culture dishes at 10‘ cells/cmz. The
density was used because it corresponded to the experimental conditions we
had used previously in our study of the "effect of excision repair on the
strand distribution of premutagenic BPDE adducts in the human HPRT gene"
(8, 9). Before beginning biochemical measurement of the rate of removal of
such adducts from cellular DNA, we determined the earliest time of onset
of semiconservative DNA synthesis in cells treated with a 1.0 or 1.2 uM
dose of BPDE. For this measurement, cells were plated into a series of 60
mm-diam dishes and assayed as described for the incorporation of tritiated
thymidine into acid-insoluble material at various times after treatment.
Although in untreated populations plated at this density DNA replication
begins 16 hr following release from confluence (16), the cells exposed to
1.0 or 1.2 uM BPDE showed no evidence>of incorporating tritiated thymidine
even when assayed 24 hr after treatment, i.e., 29 hr after release from
confluence (data not shown). These results indicated that it would be
possibleito harvest DNA from cells after allowing various lengths of times
for excision repair up to at least 24 hr posttreatment without the need to
separate newly-synthesized DNA from parental DNA.

Determining the rate ef exejsien [egair ef DEDE adduets, For these
assays, cells were plated at 10‘ cells/cm2 into a series of 150 mm-diam
dishes, 10-15 dishes per repair time period to be assayed. After 5 hr, the
cells were treated with BPDE and populations were harvested immediately or

after various repair periods. For each time point determination, DNA was

146

147

extracted from treated or control populations, digested with an
appropriate restriction enzyme, and then treated with UvrABC exinuclease
or with incubation buffer. The DNA samples were then denatured under
neutral conditions as described by ‘Tang et al. (11) to avoid the
possibility of DNA strand breaks formed at the sites of the alkali-labile
N7-guanine adducts through a nonenzymatic mechanism (24). They were then
separated on agarose gels and analyzed by Southern hybridizations using
gene- or strand-specific probes.

Evidenee ef strand-spegific repair ef BPDE adducts in human eells,
Using this method, we measured the initial number of BPDE-induced lesions
formed in the individual strands of the 20 kb BamHI fragment of the human
HPRT gene, as well as their rates of removal. This fragment is located in
the 5’-half of the gene and is entirely within the transcription unit
(Fig. 1). To avoid detection of homologous pseudogenes containing HPRT
exon sequences (25), we used riboprobes complementary to intron sequences
of the gene. Representative autoradiograms of such repair studies are
shown in Fig. 2A and 28. The intensities of the full length fragments were
determined by densitometric scanning, and the numbers of incisions and the
percentage of adducts removed were calculated as described. The data
obtained from two separate experiments are given in Table 1, and plotted
in Fig. 3. The level of adduct formation induced by BPDE was found to
increase with dose, and to be similar in each strand of the HPRT fragment
at the two doses used. However, the rates of their removal from the
transcribed and nontranscribed strand were significantly different. Within
7 hr following the treatment, 55% of the adducts had been removed from the

transcribed strand, compared to a level of 26% removed from the

148

 

 

 

 

flIGENE
1 23 4 5 6 789
1 ll 1 l l "L
‘ mu 2
Ilkbprobe
754me
HRHR HH RR
ler l I 11
T Essex

- 2 kb probe

Figure 1. Molecular organizations of the human HPRT gene and the 754
locus. The BamHI and EcoRI restriction fragments analyzed for repair are
indicated. Vertical bars in the HPRT and 754 maps indicate exons and
restriction sites, respectively. ‘The locations of the probes are shown as

rectangles. H, HindIII; R, EcoRI.

149

Figure 2. Autoradiograms illustrating extent of repair in the transcribed
(A), nontranscribed (B) strand of the HPRT gene and in both strands of the
754 locus (C). DNA was isolated from untreated human fibroblasts (first
two lanes of each panel), or from cells incubated for the indicated repair
period posttreatment with BPDE at a dose of 1.2 uM. DNA samples were
restricted with the appropriate enzyme. DNA fragment containing the
sequences to be probed was included in each DNA sample as an internal
marker. DNA samples were then treated or not treated with UvrABC and
subjected to electrophoresis and Southern hybridizations with 32P-labeled
probes as described. The upper panel shows the bands corresponding to the
20 kb fragment of the HPRT gene or the 14 kb fragment of the 754 locus.
The bottom panel contains bands corresponding to the DNA fragments serving

as internal markers.

150

N 0.52”.

0.0.9... i .. 0.... ....: ODD

+i+l+|+l +I+I+I+i +|+i+l+|om<§=
++++++I| ++++++II ++++++I|moam
3 N 8 3 N. 8 3 N. o 256:

151

.cmuumeunnm :mmn m>ec m.: o: mo.o soc» cause; sows:

mcowmwucw owmwumamcoc one .cmnweummu we acwccwum oweymsouwmcmu mg» soc» umumpaupeo mew «began: one.

 

 

 

 

mm mo.ofim:.o oo: o em
om mo.oflmm.o mm No.0Hmm.o m
o No.owvm.2 o No.oflom.H o o.:
mm No.oflmu.o mm N~.o om
mm No.0HHm.H mm mu.o N
o H.oflm~.~ o No.2 0 ~.H
eweqme a aucmEmwem\mcomeucH ewwgme x aucmsme»u\mcopmru:2 Aggy 2:
we?» ewenmm mmoo
cceeum umnweumcmLHCOz ccmgum cmnwcumcwep

 

Pc>osme we mean; gems»

use pamEaace HIEaN a: ON as: ea mucaaum _a=u.>wuc2 as: :. Joanne. Noam Lo cowueELOa .2 apaap

152

Figure 3. Rates of removal of BPDE adducts from the transcribed (closed
symbols), nontranscribed (open symbols) strand of the HPRT gene, and from
both strands of the 754 locus (half closed symbols). Circles and squares
indicate data obtained from cells treated with 1.0 8. 1.2 uM BPDE,

respectively.

153

 

 

 

IOO

 

24

INCUBATION TIME (hr)

 

 

om>0§mm 9.03004 moam m0 Fzmomma

 

Figure 3

154
nontranscribed strand. By 20 hr, the vast majority (87%) of the adducts
had been removed from the transcribed strand, whereas approximately half
(58%) of them still remained in the nontranscribed strand of the HPRT
gene.

Evidenee pf preferential repair pf BPDE adduets in human eellsII To
investigate whether there was a difference in the rate of excision repair
of BPDE adducts from a nontranscribed (inactive) genomic sequence compared
to the actively transcribed HPRT gene, the formation of such adducts and
their rate of repair were determined in a 14 kb EcoRI fragment (Fig. 1) of
the transcriptionally inactive 754 locus, using a 2.0 kb probe labeled by
the random priming method. The autoradiogram of that study is shown in
Fig. 2C. The analyzed data are given in Table 2, and plotted in Fig. 3.
The initial frequency of adducts formed by treatment of 1.2 uM BPDE was
0.92/14 kb fragment or 1.31/20 kb, which is slightly (~20%) less than that
observed in either strand of the HPRT fragment. However, the efficiency of
repair in the 754 locus was markedly reduced. Few adducts (12%) were
removed from this region in the 20 hr period after treatment, while 90%
and 59% of them were repaired in the transcribed and nontranscribed

strands of the HPRT gene at this time point, respectively.

155

Table 2. Formation of BPDE adducts in both strands of the 14 kb EcoRI

fragment of the 754 locus and their rate of removal

 

 

Dose Repair time Incisions/fragmenta % repair
11M (hr)
1.2 0 0.92 0
7 0.77 16
20 0.81 12

 

3The numbers are calculated from the densitometric scanning as described.

The nonspecific incision (0.05/fragment) has been subtracted.

Discussion

We have determined the formation and rate of repair of DNA damage
induced by a chemical carcinogen, BPDE, in the individual strands of the
single copy, single allele human HPRT gene. The results of this study show
biochemical evidence of strand-specific repair of BPDE-induced bulky
adducts, with the transcribed strand being repaired faster than the other
strand. This is consistent with the hypothesis derived from our
mutagenesis studies, namely, that strand-specific repair is the
explanation for the difference in strand distribution of premutagenic
lesions observed in repair-proficient cells treated in S phase and such
cells treated in G, phase. There is biochemical evidence of strand-specific
repair of DNA damage induced by UV (5, 26), psoralen plus near UV light
(6), and cisplatinum (6) in E. coli and/or mammalian cells. Such repair
has been suggested to account for the fact that essentially all of the
premutagenic lesions induced by (:)-3a,4B-dihydroxy-1a,Za-epoxy-l,2,3,4,-
tetrahydrobenzo[clphenanthrene in the dhfr gene of CHO cells were located
in the nontranscribed strand (27). Such strand-specific repair, which was
first reported by Hanawalt and his colleagues (5), has suggested a
specific coupling between transcription and repair. Recently, an in vitro
system carrying out strand-specific repair has been developed using cell-
free extract derived from E. coli, and indeed a candidate "transcription-
repair coupling factor" has been partially purified from this extract (6).
There are at least two ways that this coupling factor could lead to a
faster rate of repair in the transcribed strand than in the opposite
strand. This factor may associate with component(s) of the transcriptional

complex via protein-protein interactions. The presence of this factor in

156

157

the transcriptional complex may facilitate the assembly of the repair
complex in the vicinity of the transcriptional complex, so that repair
enzymes can scan the strand that is being transcribed. Another
possibility is that this factor directly or indirectly recognizes a unique
DNA structure or other signal(s) generated by the stalled transcriptional
complex at the site of DNA damage, and targets the repair enzymes to the
template strand.

In the present study, the removal of BPDE adducts from the
transcriptionally inactive 754 locus was found to be much less efficient
than removal from either strand of the HPRT gene. This indicates that
there are at least three different rates of repair of BPDE adducts within
the genome of human cells. The slow repair rate observed in the 754 locus
agrees with the hypothesis suggested by Hanawalt and his colleagues (1, 3)
that DNA damage located in the condensed, inactive chromatin is less
accessible to repair enzymes, compared to that in the active chromatin.
Our finding that repair of BPDE adducts in both strands of the inactive
gene is even slower than that observed in the nontranscribed strand of
the active gene is consistent with what was recently found for the repair
of aflatoxin B,-induced adducts in the active and inactive human
metallothionein genes (28). However, in rodent and human cell lines the
removal of cyclobutane pyrimidine dimers from the nontranscribed strand of
the DHFR gene occurs at a rate equal to the rate of removal from both
strands of the nonactively transcribed region (5). Little is known about
the mechanism(s) of repair in the compact chromatin where the inactive
genes are located. However, the lack of repair of cyclobutane dimers in

the inactive genes of xeroderma pigmentosumicells of complementation group

158
C (22, 29), which have been shown to repair transcriptionally active genes
at a normal rate, is consistent with a requirement for an additional
factor to enable the repair enzymes to reach DNA damage in such chromatin.
If so, our results with BPDE suggest that this factor functions less
efficiently for processing some bulky adducts, such as the BPDE-adducts,
than it does for cyclobutane dimers.

In contrast to the results we obtained in the present study, Tang
and Zhang (12) reported that the rate of removal of BPDE adducts from the
transcriptionally active and the 3’ nonactive regions of the dhfr gene of
CHO cells was equal. The reason for the discrepancy between these two
studies is not known, but the cell types used and the genes examined are
not the same, and the concentration of BPDE they used was higher than
ours. Furthermore, we compared the rate of repair in an active gene to
that '01 an inactive locus, and they assayed the actively transcribed
segment of the dhfr gene and its 3’ nontranscribed region. It has been
reported that in certain genes the region in‘which the preferential repair
occurs is larger than the entire transcription unit (30). It is therefore
possible that, unlike repair of cyclobutane dimers (31), repair of BPDE
adducts extended beyond the transcribed section of the dhfr gene into its
nontranscribed flanking regions. Another possible reason for the
discrepancy is that double-stranded probes were used in their analysis.
This gives information on the average rate of repair in the two DNA
strands, and that would lessen the difference detected in their study.

In order to use the Southern hybridization technique for studies of
repair rates, one needs to introduce an average of at least one UvrABC

efficiently recognizable lesion per fragment analyzed and not more than

amass-21'2”!

159

two such lesions. Under our experiment conditions, i.e., diploid human
cells plated at a density of 10‘ cells/cmz, this was accomplished using
1.0uM BPDE, a dose that yielded a mean of 1.4 UvrABC sites per 20 kb
fragment. This level of BPDE-induced damage reduced the colony forming
ability of the p0pulation to less than 0.1% of the untreated control.
However, during the 20 to 24 hr following BPDE treatment, the cells still
remained attached to the dishes. It is important to point out that such
cells were capable of removing almost all the UvrABC sites from the HPRT
gene within the 24 hr following BPDE treatment, even though the majority
of the cells in the population would eventually fail to replicate.

The level of BPDE adduct formation we observed in each strand of the
HPRT gene was very similar, indicating that there was no strand-specific
modification by this agent. However, the frequency of adduct formation in
the 754 locus was only 80% of that in the HPRT gene. Certain chemical
carcinogens including BPDE have been found to react preferentially with
regions of DNA that are sensitive to DNase I (32, 33). Sucrose gradient
sedimentation analysis has also demonstrated a preferential binding of
various polycyclic aromatic hydrocarbons to the slowly-migrating
euchromatin fractions (34, 35). These studies suggest a binding preference
of such chemicals to actively transcribed genes. But, it has also been
reported that the microsomally activated parental compound,
Benzola]pyrene, binds preferentially to the DNase I-resistant regions of
chromatin (36). Additionally, use of Southern transfer and hybridization
with gene-specific probes indicated that a transcriptionally active gene
and an inactive gene were similarly modified by a serious of polycyclic

aromatic compounds (37). Nevertheless, chromatin accessibility may not be

160
the only explanation for the slightly lower binding frequency we observed
in the 754 locus. It should be pointed out that the G.C content of a
fragment could also affect this frequency, since BPDE binds predominantly
to guanine (38). Information on the G.C content of the 754 fragment used
in this study is not yet available.

10.
11.

12.

13.

14.

15.

16.

17.
18.

REFERENCES

Bohr, V. A., Smith, C. A., Dkumoto, D. S. 8 Hanawalt, P. C. (1985)
Cell 40, 359-369.

Madhani, H. 0., Bohr, V. A. 8 Hanawalt, P. C. (1986) Cell 45, 417-
422.

Mellon, I., Bohr, V. A., Smith, C. A. 8 Hanawalt, P. C. (1986) Proc.
Natl. Acad. Sci. USA 83, 8878-8882.

Bohr, V. A., Dkumoto, D. S. 8 Hanawalt, P. C. (1986) Proc. Natl.
Acad. Sci. USA 83, 3830-3833.

Mellon, I. Spivak, G. 8 Hanawalt, P. C. (1987) Cell 51, 241-249.

Selby, C. P. 8 Sancar, A. (1991) Proc. Natl. Acad. Sci. USA 88,
8232-8236.

Chen, R.-H., Maher, V. M. 8 McCormick, J. J. (1990) Proc. Natl.
Acad. Sci. USA 87, 8680-8684.

Chen, R.-H., Maher, V. M. 8 McCormick, J. J. (1991) Cancer Res. 51,
2587-2592.

Mazur, M. 8 Glickman, B. N. (1988) Somat. Cell Mol. Genet. 14, 393-
400.

Carothers, A. M. 8 Grunberger, D. (1990) Carcinogenesis 11, 189-192.

Tang, M.-S., Bohr, V. A., Zhang, X.-S., Pierce, J. 8 Hanawalt, P. C.
(1989) J. Biol. Chem. 264, 14455-14462.

Tang, M.-S. 8 Zhang, X.-S. (1991) Proc. Amer. Assoc. Cancer Res. 32,
6. ,

Hofker, M. H., van 0mmen, G. J. 8., Bakker, E., Burmeister, M. 8
Pearson, P. L. (1986) Hum. Genet. 74, 270-274.

Van Houten, B., Masker, N. E., Carrier, H. L. 8 Regan, J. D. (1986)
Carcinogenesis 7, 83-87.

McCormick, J. J. 8 Maher, V. M. (1981) in Techniques in DNA Repair,
eds. Freidberg, E. C. 8 Hanawalt, P. C. (Marcel Dekker, Inc., New
York) Vol. 1. pp. 501-521.

Natanabe, M., Maher, V. M. 8 McCormick, J. J. (1985) Mutat. Res.
146, 285-294.

Brouwer, J. J. Biol. Chem. (in press)
Maniatis, T., Fritsch, E. F. 8 Sambrook, J. (1982) in Molecular
161

19.

20.
21.

22.

23.

24.

25.

26.
27.

28.
29.

30.

31.

32.

33.
34.

35.

36.

162

Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold
Spring Harbor, New York.

Melton, D. A., Krieg, P. A., Rebagliati, M. R., Maniatis, T., Zinn,
K. 8 Green, M. D. (1984) Nucleic Acids Res. 12, 7035-7056.

Feinberg, A. P. 8 Vogelstein, B. (1983) Anal. Biochem. 132, 6-13.‘

Lin, 0. 8 Chinault, A. C. (1988) Somat. Cell Mol. Genet. 14, 261-
272.

Venema, J., van Hoffen, A., Natarajan, A. T., van Zeeland, A. A. 8
Mullenders, L. H. F. (1990) Nucleic Acids Res. 18, 443-448.

Bohr, V. A. 8 Dkumoto, D. S. (1988) in DNA Repair: A Laboratory
Manual of Research Procedures, eds. Friedberg, E. C. 8 Hanawalt, P.
C. (Marcel Dekker, Inc., New York) Vol. 3, pp. 347-366.

Osborne, M. 8 Merrifield, K. (1985) Chem.-Biol. Interactions, 53,
183-195.

Chinault, A. C. 8 Caskey, C. T. (1984) Progress in Nucleic Acid Res.
8 Male. Biol. 31, 295-313.

Mellon, I. 8 Hanawalt, P. C. (1989) Nature 342, 95-98.

Carothers, A. M., Mucha, J. 8 Grunberger, D. (1991) Proc. Natl.
Acad. Sci. USA 88, 5749-5753.

Leadon, S. A. 8 Lawrence, P. A. (1991) Mutat. Res. 255, 67-78.

Kantor, G. J., Barsalou, L. S. 8 Hanawalt, P. C. (1990) Mutat. Res.
235, 171-180.

Dkumoto, D. S. 8 Bohr, V. A. (1987) Nucleic Acids Res. 15, 10021-
10030.

Bohr, V. A., Dkumoto, D. S., Ho, L. 8 Hanawalt, P. C. (1986) J.
Biol. Chem. 261, 1666-16672.

Arrand, J. E. 8 Murray, A. M. (1982) Nucleic Acids Res. 10, 1547-
1555.

Cox, R. (1979) Cancer Res. 39, 2675-2678.

Moyer, G. H., Gumbiner, B. 8 Austin, G. E. (1977) Cancer Letters 2,
259-266.

Moses, H. L., Webster, R. A., Martin, G. D. 8 Spelsberg, T. C.
(1976) Cancer Res. 36, 2905-2910.

Jahn, C. L. 8 Litman, G. W. (1979) Biochemistry 18, 1442-1449.

163
37. Nose, K. 8 Nikaido, 0. (1984) Biochim. Biophys.Acta 781, 273-278.

38. Osborne, M. R., Jacobs, S., Harvey, R. G. 8 Brookes, P. (1981)
Carcinogenesis 2, 553-558.

 

"TilillfilljfllillgiiflizllMilli“

4799