UNIVERSITY OF CINCINNATI
______, 20 _____
I,______, hereby submit this as part of the requirements for the degree of:
______in: ______It is entitled: ______
Approved by: ______BLM is a Suppressor of DNA Recombination
A dissertation submitted to the
Division of Research and Advanced Studies Of the University of Cincinnati
In partial fulfillment of the Requirements for the degree of
Doctorate of Philosophy (Ph.D.)
In the Department of Molecular Genetics, Microbiology, and Biochemistry Of the College of Arts and Sciences
2002
by
Joel E. Straughen
B.S., The Ohio State University, 1985 M.D., University of Cincinnati, 2002
Committee Chair: Joanna Groden, Ph.D.
i ABSTRACT
Bloom’s syndrome (BS) is a rare, recessive chromosome breakage disorder characterized by small stature, sun sensitivity, facial erythema, immunodeficiency, female subfertility, male infertility, and a predisposition to a variety of cancers. When this body of work was started, the gene for Bloom’s syndrome (BLM)hadyettobe identified. This work presents characterization of the genomic region at BLM and the identification of BLM. With the cloning of the gene, the answers to a number of questions could be investigated. Additional chapters present data that demonstrate that increased genomic instability and recombination are the result of loss of function of the
Bloom’s syndrome gene product.
Somatic cells from BS individuals are characterized by a high frequency of chromatid exchanges both between and within chromosomes, as well as by a high mutation rate at specific loci. DNAs from BS and normal clonal cell lines were first examined for alterations at microsatellite repeat loci. Alterations in size microsatellite repeats were observed at a 10-fold increase in frequency in BS clones compared to normal clones.
A contiguous representation of 2-Mb region that contains the BLM gene was generated. YAC and P1 clones from the region were identified and ordered using genetic markers in the region along with newly developed sequence tagged sites from radiation- reduced hybrids, polymorphic dinucleotide repeat loci, and end-sequences of YACs and
P1s. The physical map, and DNA markers derived from it, was instrumental in
ii identifying BLM. With the gene identified, the genomic structure was determined and a rapid DNA screening test was developed for the identification of BlmAsh,themost common mutation in BS.
To determine whether BLM can suppress recombination we over-expressed BLM in separate cell lines capable of identifying recombination or frameshift events. However, no significant difference was noted between cells transfected with BLM and those transfected with vector alone.
Finally, we established a mouse model of BS using homologous recombination to disrupt mouse Blm. Genotyping offspring from heterozygous parents did not identify any offspring homozygous for the knockout allele, suggesting embryonic lethal phenotype. In long-term studies, heterozygosity for Blm increases tumor formation compared to wild- type littermates.
iii Acknowledgements
This adventure started in 1992 when I left my previous career as an electrical engineer where I was modifying and testing airborne radar jamming equipment at various military installations around the country. I left this job, intending at the time only to go to medical school. However while waiting for acceptance into medical school, I was fortunate to get a position as a research assistant with a new faculty member at the
University of Cincinnati. Dr. Joanna Groden showed me techniques to manipulate DNA, to alter the course of a cell. She peaked my interest in molecular biology and provided an opportunity for me to contribute to our knowledge of cancer. I appreciate her tutelage, helping me to ask the right questions and giving me the tools to research the answers.
I am grateful for being able to work in Dr. Groden’s laboratory. Post-doctoral fellow Therese Tuohy and later, Kathy Heppner Goss were always willing and able to answer my endless questions concerning laboratory technique and experimental design. I appreciate all the social events we enjoyed as a lab. I will not soon forget the competitiveness of Chris Tzrespacz nor the comic relief so often provided by Chris
Heinen, both former students in the Groden lab. Thanks to all the past and present members of the Groden, I enjoyed it, thoroughly.
During my first year in medical school, I found myself spending more time in the lab than studying. Not that I was disappointed with medical school, but rather excited by molecular genetics and the powerful laboratory techniques to investigate diseases. I was hooked, and after my first year of medical school, I was fortunate to be accepted into the
Physician Scientist Training Program (PSTP) at the University of Cincinnati. This small group of extremely motivated people has been a tremendous treasure and resource. My
iv thanks go to all members of the program, faculty, staff, and students. Thanks for letting me part of something great. Leadership has always been notable in this program. At the time of my matriculation from this program, Dr. Leslie Myatt had the helm. We all appreciated his common sense approach to solving our problems. Special mention goes to Dr. Judith Harmony, who began the PSTP at the University of Cincinnati and encouraged me to apply. Judy is the most enthusiastic supporter that anyone can have.
She always wore the coolest, wildest earrings. Thank you Judy!
Thanks to my thesis committee, Drs. Groden, Menon, Stringer, Stambrook, and
Fagin. Their insight and experience in science and in life was always useful and often entertaining.
Finally, I give my thanks and love to all of my family. To my dad, Dr. William J.
Straughen, who always knew I could do this, and my mom, who really wanted me to be sure that I wanted to do this. Thanks. This quest would have been unlikely to occur or at least far less interesting without my wife, Nancy and my two girls, Tegan and Sloane.
Thanks for patients and support. I owe you one. I will always owe you one.
Joel E. Straughen
v TABLE OF CONTENTS
Abstract. i
Acknowledgements. iv
Table of Contents. 1
List of Figures. 2
List of Tables. 5
List of Abbreviations. 6
Chapter 1. Literature Review. 16
Chapter 2. Thesis Rationale. 46
Chapter 3. Materials and Methods. 48
Chapter 4. Microsatellite Instability in Bloom’s Syndrome. 74
Chapter 5. Physical Mapping of the Bloom’s Syndrome Locus. 91
Chapter 6. Cloning and Identification of BLM. 108
Chapter 7. The Genomic Structure of BLM. 133
Chapter 8. A Rapid Method for Detecting the Predominant Ashkenazi 140
Jewish Mutation in the Bloom’s Syndrome Gene.
Chapter 9. Frameshifting and Recombination in Cell Lines 147
Over-Expressing BLM.
Chapter 10. Creating a Mouse Model of Bloom’s Syndrome. 156
Chapter 11. Discussion. 178
Literature Cited. 192
1 List of Figures
Number Title Page
1 SCE in Bloom’s Syndrome Cells 14
2 Generation of Secondary Cell Lines 74
3 Dinucleotide Repeat Instability in BS Cells 77
4 Trinucleotide Repeat Instability in BS Cells 79
5 Smaller and Larger Novel Alleles in a Single 81
Clone at a One Microsatellite Locus
6 Mismatch Repair in BS Cells 83
7 2-Mb Physical Map of the Bloom’s Syndrome 96
Locus
8 Cross-reference Map of the BS locus 98
9 Fluorescent In Situ Hybridizatin 100
10 Long-range Restriction Map 103
11 Step Taken to Construct the Physical Map 107
12 Radiographic Evidence Supporting a 250-kb Region 109
for the BLM Locus
13 Somatic Cross Point Mapping 112
14 Helicase Motifs in BLM 114
15 RecQ homologues 116
16 Human Homologues to BLM 117
17 Northern Blot Analysis of BLM 119
2 18 SSCP Analysis of Persons with BS 122
19 Mutations Identified in Persons with BS 124
20 Dendorgram of RecQ Family Members 129
21 Genomic Structure of BLM 137
22 Intron/Exon Boundaries of BLM 138
23 Schema for Screening for BLMAsh 144
24 Demonstration of Detecting BLMAsh 145
25 G11 Cell Line Can Measure Frameshift Events 148
26 Photograph of the Colorimetric Assay in G11 Cells 150
27 Results of Overexpressing BLM in G11 Cells 151
28 FSH Cells Can Measure Recombination 152
29 Results of Overexpressing BLM in FSH Cells 154
30 BLM Knock-out Mouse Construct 158
31 PCR Screening of Embryonic Stem Cells 160
32 Gross Examination of Mice Blm-/- fetuses 163
33 Histology of Lymphoma from a Blm+/- mouse 165
34 Micronuclei Formation in Blm+/- Mice Compared 167
to Blm+/+ Mice
35 Mating Strategy for Blm+/- and ApcMin/+ Mice 168
36 Comparison of Tumor Number in Blm+/- and Blm+/+ Mice 169
37 Histology of Intestinal Tumors in Blm+/-, ApcMin/+ Mice 171
38 Strategy and Results of Determination of Loss of 172
Heterozygosity in Tumors from Blm+/-, ApcMin/+ Mice
3 39 Model for the Role of Blm in tumor progression 177
in ApcMin/+ Mice
40 Model for Function of BLM in Suppressing SCE 187
41 Model for Function of BLM in Suppressing DSBs 189
4 List of Tables
Table Title Page
1 Oligonucleotides Used to Amplify Microsatellite Repeat Sequences 76
2 STSs Used in the Physical Mapping of BLM 94
3 Polymorphic Microsatellites Used or Isolated in the Mapping of BLM 101
4 Restriction Fragments in Kilobases Identified by Hybridization of 105
Probes from the 2-Mb YAC contig
5 Mutations in the Candidate Gene in Persons with BS 121
6 Oligonucleotides Used for SSCP Analysis 123
7 Oligonucleotides Used for Determining the Genomic Structure of BLM 135
8 Post-Coital Age of Viable Mouse Embryos 162
5 List of Abbreviations
A Adenosine
BLM Bloom’s Syndrome Gene
BLM Bloom’s Syndrome Protein
BrdU Bromodeoxyuridine
BS Bloom’s Syndrome
C Cytosine cDNA complementary DNA
CEPH Centre d,Etude Polymorphisme Humain
CS Cockayne’s Syndrome
CTP Cytosine Triphosphate dATP deoxyadenosine Triphosphate dCTP deoxycytosine Triphosphate
DEAD Aspartate Glutamate Alanine Aspartate
DExH Aspartate Glutamate any amino acid Aspartate dGTP deoxyadenosine Triphosphate
DMEM Dulbecco’s Modified Eagle Media
DNA Deoxyribonucleic Acid
DSB Double-Stranded Break
6 dsDNA double-stranded DNA
dTTP deoxythymosine Triphosphate
EBV Epstein-Barr Virus
EDTA Ethylenedi tetraacetic Acid
EGFP Enhanced Green Fluorescent Protein
EM Electron Microscopy
FA Fanconi’s Anemia
FISH Fluorenscent In Situ Hybridization
FITC Fluoricein Isothiocyanate
G Guanosine
GTP Guanosine Triphosphate
H & E Hemotoxylin and Eosin
HAT Hypoxanthine and Thymidine
HRPT Hypoxanthine Ribosylphosphoryl Transferase
Ig Immunoglobin
Kda KiloDalton
LB Luria Broth
LCL Lymphoblastoid Cell Line
LOH Loss of Heterozygosity
MLV Murine Leukemia virus mRNA messanger RNA
NBS Nijmegen Breakage Syndrome
NK Natural Killer
7 OD Optical Density
PBS Phosphate Buffered Saline
PCR Polymerase Chain Reaction
PFGE Pulsed Field Gel Electrophoresis
Qr Quadriradials rpm revolutions per minute
RTS Rothmund-Thompson Syndrome
SCE Sister Chromatid Exchange
SDS Sodium Dodecyl Sulfate
SSC Sodium Chloride, Sodium Citrate
SSCP Single-Stranded Conformational Polymorphism ssDNA single-stranded DNA
NTP Nucleotide Triphosphate
STS Sequence Tagged Sites
T Thymosine
TBE Tris-Borate EDTA
TE Tris-EDTA
TTD Trichothiodystrophy
UDS Unscheduled DNA Synthesis
UTP Uridine Triphosphate
UV Ultra-violet
WRN Werner’s Syndrome Protein
WS Werner’s Syndrome
8 XP Xeroderma Pigmentosa
YAC Yeast Artificial Chromosome
9 CHAPTER 1. Literature Review.
This literature review describes pertinent, published information related to
Bloom’s syndrome (BS). The first third of the review will focus on a description of the disease, beginning with the initial case report and clinical presentation of BS patients. A discussion of cytogenetics in BS will be included, as well as a number of biochemical studies of BS cells. The second third of this review will focus on the identification of the
BLM gene, that when mutated is responsible for BS. A review of helicases in general and
RecQ family members in particular will be presented. The final third of this review will focus on the recent work in characterizing the function of the BLM protein.
In 1954, New York City dermatologist David Bloom reported three cases of a unique syndrome of congenital telangiectatic erythema resembling lupus erythematosus in dwarves (Bloom, 1954). Over the next few years, a number of clinical reports referred to other cases of Bloom’s syndrome (BS). Many of these reports highlighted the similarities and differences, as compared to Bloom’s original report, in the clinical presentation (Brunsting, 1957; Lewis, 1957; Fitzpatrick, 1962; Curth, 1964). At this time, however, the patient presentation that characterized BS was not rigidly defined.
Initially, BS was observed to consist of three cardinal features: congenital telangiectatic erythema of the face, sensitivity to sunlight, and stunted growth (Bloom, 1966). Since that time, additional phenotypic data have been included in the clinical description of BS.
Affected persons are characterized by low birth weight and a predisposition to cancer
(Schoen and Shearn, 1967; German, 1968). Many BS patients become diabetics during adolescence or early adulthood (German and Passarge, 1989). Males with BS are sterile
10 with hypogonadism (Kauli et al., 1977) and females are subfertile. There is a case report of woman with BS giving birth to a health baby boy (Chisholm et al., 2001). Although
BS is a very rare disease, some small populations have a much higher rate than others do.
Chief among these is the Ashkenazi Jewish population. The carrier rate among this population is approximately 1 in 104, based on screening for the most common mutation
(described in chapter 8) (Roa et al., 1999). One of the most interesting characteristics of
BS is the enormous predisposition to cancer. BS patients develop a variety of cancers at a young age (German, J. 1997). It is not uncommon for a person with BS to develop one type of cancer, such as a leukemia, only to get a second primary cancer, such as colon cancer, a few years later. Immunodeficiency is also frequently present in BS persons.
The first documented report of immunodeficiency in BS dates to 1967 in a case report of a four-year-old boy with BS who had low levels of IgG, IgA, and IgM (Schoen and Shearn, 1967). In an attempt to better understand the disease and its effect on the immune system, investigators have undertaken a variety of in vitro and in vivo approaches to evaluating BS persons and cells. Abnormal serum concentrations of at least one class of immunoglobin (Ig) were detected in three of four persons with BS
(Hutteroth et al., 1975). Lymphocytes from all four of these persons with BS had an impaired proliferative response. Examination of IgG response in other BS persons showed abnormal kinetics (Weemaes et al., 1984). In a long-term study tracking five BS patients, a Dutch group reported that serum IgG levels showed no age-related increase, that IgA levels were below the 10th percentile in childhood but rose to normal in all but one patient, and that IgM was decreased in four patients at diagnosis and remained so in three of them (Weemaes et al., 1991). Similarly, a Japanese study following 2 BS
11 persons for a 10-year period observed that IgM levels remained low, although both IgG and IgA levels increased with age, as in unaffected persons. Mitogen-induced immunoglobulin production in tissue culture suggested a B-cell dysfunction, although T- cell function appeared normal (Kondo et al., 1992). In addition to the anomalous immunoglobin levels in BS persons, other aspects of the immune system are altered relative to age-matched controls. An evaluation of natural killer (NK) cells from four BS persons found that NK activity against K562 tumor cells was depressed to less than half that of age-matched controls. Treatment of the BS NK cells with interleukin-2 (IL-2) restored their activity to within normal ranges (Ueno et al., 1985). Further work has suggested abnormal mu mRNA alternative splicing from a detailed IgM study on BS lymphoblastoid cell line. IgM was found to be expressed on the surface of BS LCLs, but very little was secreted, as compared to controls. VDJ arrangements appeared normal and total mu mRNA was detected at normal levels. However, the mu protein C-terminal, which contains the secretion signal, was poorly detected compared to normals (Kondo et al., 1992; Ozawa et al., 1994). One of these groups reported that the absence of mutations or deletions in the secretion signal-containing C-terminal coding sequence, the relevant splice sites, or the poly(A) signal sequence (Kondo et al., 1992). Confirmation of normal VDJ recombination in BS was provided by examining signal and coding joint formation from recombination reactions (Hsieh et al., 1993; Petrini et al., 1994).
In 1960, the first cytogenetic analysis of BS lymphocytes in a short-term culture was undertaken using blood from one of Dr. Bloom’s original BS patients (German,
1964). It had been suggested originally that BS might be a result of an abnormal chromosome complement. The chromosomal complement was normal, but increased
12 chromosomal breakage and rearrangement were observed in BS cells (Sawitsky et al.,
1966). Abnormal chromosomal configurations, including quadriradials, ring
chromosomes, and chromosomal fragments were also present in metaphase spreads of BS
cells (German et al., 1965). Cytogenetic studies of BS cells were expanded and
demonstrated abnormal mitotic figures, a high rate of non-disjunction, aneuploidy, and
numerous micronuclei. (Frorath et al., 1984). Differential Giemsa staining after
bromodeoxyuridine (BrdU) incorporation highlighted the most characteristic cytogenetic
feature of BS, an excessive rate of sister-chromatid exchange (SCE) (Chaganti et al.,
1974). SCE rates in BS cells are approximately 12 times higher than those in cells from
unaffected individuals. No other disease exhibits an increased sister-chromatid exchange
rate, thus providing a diagnostic tool for BS (Figure 1). Further cytogenetic analysis
revealed that chromosomal breakpoints and mitotic chiasmata occur at non-random
locations, with an increase of these events at centromeric and telomeric regions of the
chromosomes (Lindgren, 1981). To this point all of the instability data was obtained
from cultured cells. Finally an in vivo study of exfoliated epithelial cells from the oral cavity and urinary tract from persons with BS had a significantly increased number of micronuclei (Rosin and German, 1985). In total, the cytogentic study of cells from persons with BS demonstrated that chromosome breakage and genomic instability is a consistent feature of Bloom’s syndrome.
The truly interesting chromosomal change in BS cells is the quadriradial (Qr).
The typical Qr in BS consists of two homologues in a symmetrical configuration with the
13 A
B
Figure 1. Sister Chromatid Exchange in Normal and BS Cells. Lymphoblastoid cell lines are grown through two rounds of replication in the presence of bromodeoxyuridine and then geimsa stained. A. Cells from a normal individual typically show 5-10 SCE events. B. Cells from a person affected with BS have approximately 10-15 times the rate of SCE as that of normal persons. A few selected SCE events are indicated by arrows.
14 two centromeres in opposite arms and equidistant from a point of exchange between the two chromosomes. These configurations originally were proposed to arise as result of an interchange of DNA segments between homologous chromosomes, and their occurrence was considered cytological evidence that somatic crossing-over could take place in mammalian cells (German, 1964). Qrs are never seen in normal, untreated cells and signify somatic recombination between homologous chromosomes with the potential to generate daughter cells with different genetic components
With these data regarding clinical presentation and cytogenetics, a cadre of investigators began to probe the molecular and biochemical characteristics of BS.
Moving to molecular techniques for examining genomic instability, the analysis of 33 BS clonal cell lines, each originally cloned from the same cell, and 27 normal clonal cell lines, cloned in a similar manner, revealed 20 distinctive band alterations at the highly polymorphic repeat DNA locus D1Z2 (Groden and German, 1992). This genomic instability may be the result of unequal SCE. Conflicting data were reported where no evidence of significant variation of the mutation rates of the two hypermutable CA repeat minisatellites was identified, although the cell lines analyzed were not clonal (Foucault et al., 1996). Lymphoma cells from a person with BS and cell lines derived from this tumor were shown to have microsatellite instability (Kaneko et al., 1996). Analysis of a BS lymphoblastoid cell line displayed nearly 80 times the mutation rate at the HPRT locus compared to a normal cell line. 83.3% of these mutations involved deletions at HPRT, half of which removed the entire gene, 2.5 times the rate seen in the controls (Tachibana et al., 1996). These observations confirm the chromosomal instability phenotype of BS
15 cells, but suggest that much of the instability in BS cell may be due to macroscale DNA alterations.
Based on cytogenetic and molecular genetic observations, it was postulated that abnormal DNA replication in BS cells may be a mechanism by which homologous chromatid interchanges occur. DNA chain growth is significantly slower in BS dermal fibroblasts than in normal cells (Hand and German, 1975). The slow DNA chain growth in semiconservative replication of BS cells could therefore provide sufficient time for unknown mechanisms to increase chromatid exchange. In vitro activities of DNA polymerases alpha, beta, and gamma derived from BS cells were determined to be normal
(Parker and Lieberman, 1977, Bertazzoni et al. 1978). DNA fiber autoradiograms from
BS skin fibroblasts and blood lymphocytes showed a retarded rate of replication fork movement compared to normal adult controls (Hand and German, 1977). A comparison of DNA chain growth in BS cells and normal cells showed that plating density and time in culture affected the results. When high densities of cells were plated and cultured for
48 hours, no significant differences between BS and controls were noted. However, at low density and 24 hour-culture periods, the BS cells demonstrated a reduced chain growth rate. It was concluded that the slow rate of DNA chain growth in BS cells was an artifact introduced by culture conditions (Ockey, 1979). Undeterred, Kapp continued to evaluate DNA replication in BS cells and reported that DNA fork displacement rates in
BS were 55-65% of the normal rate, supporting earlier work done by Hand and German
(Kapp, 1982). Poly(ADP-ribose)polymerase inhibitors increased the frequency of SCE in
BS cells and normal cells, although expression of poly(ADP-ribose)polymerase is normal in BS cells (Shiraishi et al., 1983). Primary cultures and long-term cultures of BS cells
16 have low levels of ATP, suggesting that retarded DNA chain growth and other ATP- dependent functions in BS cells could be caused by a defect in purine biosynthesis or
ATP generation (Taylor et al., 1983). Delayed DNA maturation and resulting breaks in the DNA template were suggested to elevate SCE in BS cells (Ockey and Saffhill, 1986).
Low molecular weight intermediates of DNA replication from BS cells converted to high molecular weight DNA at the same rate as those of normal cells (Hurt and Moses, 1986).
BS cells demonstrated no alteration of DNA-polymerase activity or in the amounts and molecular weights of different DNA-polymerases (Spanos et al., 1986). More advanced methods of measuring DNA chain elongation in S-phase using BrdU incorporation into elongating DNA, and flow cytometry determined that chain elongation in BS cells was slower than that of normal cells in early S-phase (Fujikawa-Yamamot et al., 1987).
Further experiments demonstrated a decrease in topoisomerase II activity and suggested this decrease as a possible explanation for the slow DNA chain elongation and high SCE rates in BS cells. The topoisomerase II activity of BS fibroblasts, as measured by unknotting of phage P4 DNA, is much more strongly inhibited by cell growth in medium containing BrdU than that of normal fibroblasts (Heartlein et al., 1987). This suggests that topoisomerase II may directly or indirectly interact with BLM and require BLM for efficient unknotting of DNA. Deoxyribonuceoside triphosphate pool sizes were normal in BS cells (Kenne and Akerblom, 1990). Finally, biochemical characterization of DNA replication intermediates demonstrated a difference between BS and normal cells (Lonn et al., 1990).
By far the largest collection of published information on BS relates to induced
DNA damage and DNA repair processes. Much of this work focused on the induction of
17 DNA damage by radiation or chemicals and observing cellular sensitivity and/or DNA repair processes. In addition, specific DNA repair-related proteins, including DNA ligases, DNA glycosylases, and topoisomerases were evaluated.
Many investigators have studied sensitivity of BS cells to UV radiation. BS cells were first shown to be sensitive to UV-irradiation by interpreting survival curves of UV- treated BS cells; an excessive low molecular weight component of BS DNA was identified in sucrose gradients, as compared to normal cells following UV irradiation
(Giannelli et al., 1977). Several BS fibroblasts, with the exception of cell line GM1492, displayed the same sensitivity to UV as normal control cells; this was measured by SCE, micronucleus formation, adenovirus reactivation, and colony formation (Krepinsky et al.,
1980). A second group demonstrated similar results with three different lines of BS cells
(Ishizaki et al., 1981). In addition to causing double strand DNA breaks through a direct mechanism, near-UV-irradiation of cells also induces DNA damage by single-strand breaks generated by radical oxygen species. Six of seven BS skin fibroblast cell lines, exhibited near-ultraviolet sensitivity as determined by survival curves (Zbinden and
Cerutti, 1981). In contrast, normal unscheduled DNA synthesis in BS lymphocytes from one affected individual was reported in response to UV-irradiation (Evans et al., 1978).
In other experiments, six of eight BS cell lines demonstrated an increase in DNA breakage relative to normals when treated with near-UV radiation (Hirschi et al., 1981).
Increased sensitivity to UV-irradiation in BS was recorded by measuring an increase in
SCE above the already high spontaneous frequency of SCE, and comparing the results to normal cells (Kurihara et al., 1987b). A similar experiment confirmed these data
18 (Mamada et al., 1989). Most researchers believe that BS cells are more sensitive to UV radiation than normal cells.
Nucleotide excision repair of DNA damage induced by UV radiation in BS cells also has been evaluated. Sensitivity to an ultraviolet-induced endonuclease and measurements of unscheduled DNA synthesis were normal in BS cells (Ahmed and
Setlow, 1978). This contrasts with other data, which show a slightly less efficient repair of UV-induced DNA damage in BS cells (Henson et al., 1981). The rates of incision following UV-irradiation in the presence of DNA synthesis inhibitors are comparable between BS and normal cells, suggesting that the endonuclease function for excision repair processes is intact in BS (Squires et al., 1982). Gene-specific DNA repair of UV-
induced cyclobutane pyrimidine dimers was examined at the DHFR locus, and
preferential repair of this active gene was similar to that of normal cells (Evans and Bohr,
1994).
In addition to excision repair, UV-irradiation induces p53 as a mechanism for
controlling the cell-cycle in the wake of DNA damage and for directly upregulating
transcription of DNA damage repair proteins. p53 accumulation in the nucleus following
UV treatment was completely absent in 2 of 11 BS strains tested (Lu and Lane, 1993).
One of these BS strains expressed no detectable p53, yet was characterized by a normal
G1 cell-cycle arrest after UV exposure (van Laar et al., 1994). These results suggest that
the p53 response in BS cells is normal.
Gamma-irradiation induces single- and double-strand breaks in DNA and has
been used to test both sensitivity and repair in BS cells. Sensitivity of BS cells to γ- irradiation matches that of normal cells using light microscopy to evaluate chromosomal
19 aberrations (Evans et al., 1978). Contrary to this report, BS cells have an increased sensitivity to ionizing radiation compared to normal cells (Aurias et al., 1985). However, gamma-ray-induced DNA nicks are resealed in BS cells with the same efficiency as normal cells (Hirschi et al., 1981, Nocentini, 1995). The same two BS cell lines with no p53 response UV-irradiation displayed no p53 response to X-ray treatment (Lu and Lane,
1993). Although BS cells may be more sensitive to UV- or gamma-radiation, there is no current evidence that the protein deficient in BS cells is involved in radiation-induced repair mechanisms.
In addition to the use of radiation to damage DNA, similar experiments have used chemicals means to evaluate BS cell sensitivity and repair response to DNA damage.
Mitomycin C (MC) is an alkylating agent, covalently bonding DNA, and is often used to induce DNA excision repair. Several researchers have tested the response of BS cells to treatment with MC. BS cells treated with MC experience a 2-fold increase in the SCE rates (Shiraishi and Sandberg, 1978a). These data were supplemented by additional experiments reporting an increase in chiasmata and quadriradials in BS cells after MC treatment (Kuhn, 1978). These results were repeated with comparisons to normal cells in subsequent experiments (Shiraishi and Sandberg, 1979). These reports differ from later studies that confirm BS sensitivity to MC by cell killing, but normal sensitivity to MC as measured by SCE when compared to normal cells (Hook et al., 1984).
Numerous other mutagens have been tested for their affects on BS cells. A few investigators have used n-acetoxy-2-acetyl-aminofluorene (AAAF), a DNA damaging agent, to induce DNA repair. BS cells are proficient in DNA excision repair after comparison to normal cells in the rates of unscheduled DNA synthesis (UDS) and loss of
20 sites sensitive to an ultraviolet endonuclease after AAAF treatment (Ahmed and Setlow,
1978). These results were corroborated in similar UDS experiments (Remsen, 1980). BS cells proved sensitive to ethyl methanesulfonate (EMS) another commonly used DNA damaging agent, which induces more SCE events in lymphocytes from BS individuals than in lymphocytes from controls (Krepinsky et al., 1979). Fibroblast cell lines were tested for lethal effects of alkylation damage produced by N-methyl-N-nitrosourea
(MNU) and N-ethyl-N-nitrosourea (ENU). no excessive sensitivity in BS cells was reported (Teo and Arlett, 1982). One BS cell line was no more sensitive than normal strains to killing by the DNA intercalating and mutagenic agent ethidium bromide (Gupta and Goldstein, 1982). BS cells have also been evaluated for sensitivity, as measured by
SCE, chromosome aberrations, and colony formation (CF), to the carcinogens 4NQO (4- nitroquinoline-N-oxide), MNNG (N-methyl-N'-nitro-N-nitrosoguanidine), AFLG1
(aflatoxin G1), AFLB1 (aflatoxin B1), BNU (butylnitrosourea), and MNU
(methylnitrosourea) and the tumor promoter TPA (12-O-tetradecanoylphorbol-13- acetate). Of the three different BS cell types tested (type I with normal SCE and normal karyotype; type II with high SCE and normal karyotypes; type III with high SCE and abnormal karyotypes), type I had the same sensitivity as normal cells to each chemical.
Type II and III cells had a 2-fold increase in SCE after carcinogen treatment. TPA significantly increased SCE and highly enhanced CF with dose-dependence in type III cells (Shiraishi, 1985). In contrast to earlier studies, BS cells were approximately four times more sensitive than normal cells to the lethal effect of ENU and remarkably hypersensitive to SCE induction by ENU (Kurihara et al., 1987a).
21 Specific proteins have been examined for potential defects in BS cells; DNA ligase I has been the most studied. Consideration of the current data on BS, its sensitivity to DNA-damaging agents, elevated SCE, and chromosomal breakage, slow growth, and slow DNA replication forks, suggested that these features were characteristic of
Escherichia coli and yeast mutants with a defective DNA ligase. In addition, these characteristics suggested that the primary defect in BS may affect replication in S-phase.
DNA ligases and polymerases were considered as candidates for the disease gene product, although previous studies of DNA polymerases have revealed no defect. In contrast, DNA ligase I, one of two known human DNA ligases at the time, was defective in BS cell lines (Willis and Lindahl, 1987, Chan et al., 1987). Willis and Lindahl showed
that DNA ligase Ia from BS cell line GM3403 had a lower enzymatic activity than ligase
Ia from control cell lines and that the GM3403 protein was more heat labile than control
ligase. Chan et al. demonstrated that DNA ligase Ia from BS cell line HG 1270 extracts eluted from a gel-filtration column in a different fraction than seen in normal cell extracts. Work continued on DNA ligase I, in an effort to identify specific mutations in the gene and structural/functional defects in the protein. DNA ligase I was examined in seven BS cell lines, six of which reduced enzymatic activity, the seventh cell line had a dimeric DNA ligase I instead of the normal monomeric form (Willis et al., 1987).
Immunoblotting demonstrated that the size and amount of DNA ligase I (98 kDa) in BS
and normal cells were similar. However, the ligase activity in high SCE BS cells, as
compared to normal cells, was 50-90% reduced in part due to loss of ATP-binding and/or
hydrolytic activity (Chan and Becker, 1988). These data were inconsistent, however,
with a report that a partially purified 130 kDa human DNA ligase from BS cells has
22 higher activity than untransformed or transformed control cells (Mezzina, et al., 1989a).
Indirect evidence of a defect in DNA ligase was provided by determining in vivo DNA joining ability and fidelity of exogenous DNA in BS and normal cells. A 1.3 to 3-fold decrease in joining efficiency and a 1.4 to 5.4-fold increase in mutations at the joining site was observed in BS cell extracts as compared to extracts from normal cells (Runger and Kraemer, 1989). DNA ligase I and II were examined seperately for protein stability and activity in two BS and two normal cell lines. DNA ligase I demonstrated no unusual heat lability in any cell line tested and consistently lower activity in only one of the two
BS cell lines; no differences in DNA ligase II activity between normal and BS cells was observed (Kurihara et al., 1991). The human DNA ligase I gene then was cloned by degenerate oligonucleotide screening of cDNAs from normal and BS cells (Barnes et al.,
1990). The sequences of the DNA ligase I gene in two BS cell lines, one with reduced ligase activity, the other with the dimeric form, were normal, as was the mRNA size by northern blot analysis (Petrini et al., 1991). In addition DNA ligase III also has normal activity in BS cells (Tomkinson et al., 1993).
Uracil-DNA glycosylase received some attention in the study of BS as an enzyme important in the base-excision repair (BER) pathway. The temporal sequence of repair and replication has been examined in BS. Normally, nucleotide-excision repair
(NER) is stimulated prior to base-excision repair, which is upregulated just before replication. In one series of experiments, base-excision repair was monitored by the quantitation of uracil-DNA glycosylase and measurements of unscheduled DNA synthesis (UDS) induced by hydroxyurea. BS cells were characterized by specific alterations in this temporal sequence of uracil-DNA glycosylase regulation and UDS,
23 such that BER was not enhanced prior to the induction of DNA replication, but rather concomitantly with DNA replication (Gupta and Sirover, 1984). Prompted by such results, other investigators began to study temporal induction of uracil-DNA glycosylase.
Maximal expression of uracil-DNA glycosylase was delayed until peak DNA synthesis in cells from one of two BS patients studied (Yamamoto and Fujiwara, 1986). In a separate set of experiments one of three monoclonal antibodies to uracil-DNA glycosylase failed to recognize native uracil-DNA glycosylase from two separate, non-transformed BS cell strains by lack of ELISA reactivity and no inhibition of catalytic enzyme activity.
However, all three antibodies recognized the denatured enzyme by western blot analysis
(Vollberg et al., 1987). This work was later augmented to show this antibody did not recognize uracil-DNA glycosylase in 6 unrelated BS cell lines (Seal et al., 1988, 1990).
These data suggest that BS cells have structurally altered uracil-DNA glycosylase. The importance of the immunological studies was questioned when normal catalytic activity of uracil-DNA glycosylase from BS lymphoblastoid cells was demonstrated (Vilpo and
Vilpo, 1989). This group also purified uracil-DNA glycosylase from normal and BS cells for functional and biochemical testing. Glycosylase activity was demonstrated in uracil-
DNA glycosylase isolated from BS cells, but the isoelectric point, thermostability, and sensitivity to pyrimidine analogues were all significantly altered, suggesting that uracil-
DNA glycosylase was involved in a “BS pathway” (Seal et al., 1991).
In addition to uracil-DNA glycosylase, hypoxanthine-DNA glycosylase was tested for its temporal expression in BS cells. Hypoxanthine-DNA glycosylase was regulated in a similar manner as uracil-DNA glycosylase in BS cells per the results of
Yamamoto and Fujiwara, in that hypoxanthine-DNA glycosylase is delayed by up to nine
24 hours following hydroxyurea administration and does not reach peak expression until
DNA synthesis is at its peak (Dehazya and Sirover, 1986). Other repair proteins and processes have also been evaluated in BS cells. The mismatch repair process in BS is proficient in both G/C and A/C mismatches (Brown et al., 1989, Langland et al., 2001).
The levels of a 42kDa deoxyribnuclease, possibly important for DNA repair, are many times higher in malignantly transformed cell lines and in BS cells compared to control cells (Mezzina et al., 1989b). The human Rad51/RecA homologue gene was sequenced
and shown to be normal in cDNA isolated from three BS cell lines, eliminating this
important repair/recombination protein as a candidate for the mutated gene in BS
(Hellgren et al., 1994).
Numerous investigators examined effects of oxidative stress in BS cells.
Following BrdU treatment, both BS fibroblasts and lymphoblastoid cells displayed cell-
cycle profiles similar to normal cells after treatment with chemicals that induce free
radical formation, such as 4-hydroxy-nonenal for fibroblasts and oxygen or paraquat for
lymphoblastoid cells (Poot et al., 1989, 1990). Such observations suggested that BS cells
may suffer from an excess of endogenous free radicals. Additional support for this
argument came from two different groups that investigated the activity level of
superoxide dismutase (SOD) in BS cells. SOD activity in BS cells was higher than in
normal cells, implicating a high radical oxygen content of BS cells (Lee et al., 1990). In
theory, the elevated radical oxygen content in conjunction with elevated SOD activity
could produce high intracellular levels of hydrogen peroxide, which in turn could
inactivate the enzymes responsible for its elimination, as well as increase SCE and
chromosomal aberrations (Nicotera et al., 1989). These researchers then examined
25 reactive oxygen species levels in BS and control cell lines using a chemiluminescence assay with the calcium ionophore A23187 and chemotactic tripeptide N- formylmethionylleucylphenylalanine. Increases in active oxygen production by 48.6% and 250-314%, respectively, were demonstrated in BS cell lines over controls (Nicotera et al., 1993). Data from the publications reviewed above suggest that several gene products made reasonable candidates for the disease gene product of BS.
From the data collected up until 1995, very few definitive conclusions about the biochemical defect in BS could be made. For every observation made of BS cells by one group, another group made different and often conflicting observations. Without solid biochemical leads, a number of laboratories began on the process of determining the genetic defect responsible for BS by reverse genetics or positional cloning.
A series of papers using somatic cell genetics set the stage for what was to come for the study of BS, describing methods that ultimately would prove useful to defining a genomic region that would contain the disease gene. Experiments using somatic cell hybridization, originally designed to resolve conflicting data regarding co-cultivation experiments between BS cells and normal cells, laid the groundwork for others. One group of researchers suggested a diffusible substance released from BS cells was capable of inducing elevated SCE in normal cells (Tice et al., 1978). However, other investigators noted a decrease in SCE in BS cells when co-cultivated with normal cells
(van Buul et al., 1978; Bartam et al., 1979). By fusing BS cells to normal human fibroblasts, the investigators made euploid somatic cell hybrids. The SCE rate of these hybrid cells matched that of normal cells, indicating that the SCE rate of BS could be corrected by normal cells (Bryant et al., 1979). A similar report using somatic cell
26 hybridization between BS cells and Chinese hamster cells showed complete correction of the elevated SCE rates in BS, indicating that a loss of normal function, rather than a gain of an abnormal function is responsible for the BS SCE rates (Alhadeff et al., 1980).
Another report substantiating Chinese hamster-BS somatic cell hybrids ability to correct fully the high SCE rates of BS cells was published, yet it indicated that the hybrid cells reverted to the high SCE phenotype in long-term cultures (Yoshida and Sekguchi, 1984).
This is most likely due to rodent chromosome loss over time in human-rodent somatic cell hybrids. Somatic cell hybrids were then established to examine the dominance or recessiveness of the low-SCE BS phenotype, as well as to determine whether there were multiple complementation groups in BS. Tetraploid cell fusions between high and low
SCE BS phenotypic cells always resulted in low SCE rates (Shiraishi, 1988). Also, high
SCE LCLs from BS persons of diverse ethnic and geographic backgrounds were used to create BS-BS hybrid cell lines from high SCE cells, none of which exhibited a correction of the high SCE phenotype (Weksberg et al., 1988). These cell-hybridization experiments argued that there was a single genetic locus responsible for BS and that BS occurs as a result of loss of function mutation. In experiments based on microcell- mediated chromosome transfer, a technique similar to somatic cell hybridization, human chromosome 15 complemented the high SCE phenotype of BS cells, first mapping the
BLM locus to chromosome 15 (McDaniel and Schults, 1992).
Eliminating approximately 95 % of the genome from containing BLM was a milestone allowing investigators to focus their search for the gene to a smaller, yet still large region of DNA. The next several steps in the process of cloning the gene for BS required DNA from affected individuals and their kindreds. Fortunately, the Bloom’s
27 syndrome registry maintaining cell lines, familial data, and medical history of nearly all
BS persons has been maintained by Dr. James L. German III (German and Passarge,
1990). The gene for BS, now referred to as BLM, was then localized to 15q26.1 by homozygosity mapping. A polymorphic locus in the FES proto-oncogene, already mapped to 15q26.1, was homozygous in 25 of 26 BS patients whose parents were related, either by consanguinity or descent (German et al., 1994). Additional support for tight linkage between BLM and FES came from studying Ashkenazi Jews, who are more likely to be affected by BS than any other ethnic population. Linkage disequilibrium studies in the Ashkenazim revealed association of particular alleles at FES and the polymorphic locus at D15S127 with BLM (Ellis et al., 1994).
Based on crossover data from the genetic mapping and disequilibrium studies,
BLM was estimated to be approximately 0.8 cM on either side of FES (German et al.,
1994). A 1.8 MB physical map, centering on FES, was constructed in order translate
genetic distances into basepairs and to identify YAC and P1 clones that may contain all
or part of BLM (Straughen et al., 1996). A clever observation concerning high and low
SCE phenotypes in BS LCL’s helped to limit the region from nearly 2 MB centered on
FES to the 400 kb immediately centromeric to FES. It was noted that persons who
inherit a mutate BLM allele from a common ancestor never exhibit the high-SCE/low-
SCE mosaicism. Conversely, the SCE mosaicism predominately arises in cells from
affected individuals who are offspring of parents who do not share a common ancestory.
Comparing polymorphic loci in the BLM region of 11 presumed compound heterozygous
individuals with BS that exhibit the high/low-SCE mosaicism, showed that five of them
had polymorphic loci distal to BLM that were heterozygous in their high SCE cells. The
28 same polymorphic loci in the low-SCE cells from the same individuals had become homozygous, pointing to an intragenic recombination within BLM, allowing a functional
BLM to be expressed in these cells (Ellis et al., 1995). The BS phenotype itself, elevated rates of recombination, allowed BLM to be mapped to a very small region, nearly to the gene itself (Korn and Ramkisson, 1995). Finally, through cDNA selection and mutational anaylsis, BLM was identified as a member of the RecQ helicase family (Ellis et al., 1995).
A variety of processes take place on the DNA substrate, including DNA repair, replication, recombination and transcription. The normally double-stranded (ds) DNA must be transiently and locally unwound into single-stranded DNA to allow these events to take place. Helicases are a family of proteins that unwind DNA and/or RNA, producing stretches of single stranded nucleic acids where other DNA-related metabolic processes can take place. In the late 1960s, researchers began purifying proteins from E. coli that were identified as DNA-dependent ATPases (Debreceni et al., 1970). In 1976, the first paper was published that showed a newly identified DNA-dependent ATPase isolated from E. coli that unwound DNA (Abdell-Monem et el., 1976). Within a few years, several additional proteins from E. coli were isolated with the same enzymatic activity. In 1979, the newly accepted term “helicase” first appeared to describe these unwinding enzymes (Krell et al., 1979). In addition to the DNA-dependent nature of this class of enzymes, RNA-dependent families of these enzymes have also been identified.
Over the past three decades, a large number of helicases have been identified including 12 unique helicases in E. coli,41inS. cerevisiae, and 31 in human (Ellis,
1997). In addition, helicases have been identified in other diverse taxonomical groups
29 including viruses, plants, and insects. Due to the variety of processes that must take place on single-stranded nucleic acid templates, it is not surprising to find a number of unique helicases in any given species.
Helicases are defined by the presence of seven loosely conserved amino acid motifs (Hodgman, 1988). The first (I) and third (II) of these motifs are sometimes referred to as Walker A and B motifs and were originally identified as part of the α and β
subunits of ATP synthase in E. coli (Walker et al., 1982). These motifs are present in
enzymes that do not function as helicases, as well as those that do, and are common
features of enzymes that bind nucleotide triphosphates (Gorbalenya et al., 1989). A conserved lysine residue in motif I is required for ATP-binding (Rozen et al., 1989).
Motifs Ia, III, IV, V, and VI are unique to helicases (Ellis, 1997). DNA helicases contain a DExH box and RNA helicases usually contain a DEAD or DExH box, so named for the single letter amino acid code of the highly conserved sequence within motif II (Schmid and Linder, 1992; Ellis, 1997).
In addition to the amino acid motifs of a helicase, there is also a functional aspect that defines a helicase. The enzyme must release a fragment of ssNTPs in a strand- displacement assay. This functional helicase assay can be carried out with a variety of substrates in order to characterize the enzyme. First, the substrate itself can be dsDNA, dsRNA, or heteroduplexed DNA/RNA, allowing the enzyme to be defined as a DNA,
RNA, or DNA/RNA helicase. Furthermore, the duplex substrate may be blunt-ended, have a 5’ single-stranded extension, and/or a 3’ single-stranded extension. Release of the ssNTP from a 5’ single-stranded extension in the presence of ATP is evidence of a 5’ to
3’ helicase. Likewise, under the same conditions release of ssNTP from a 3’ single-
30 stranded extension defines a 3’ to 5’ helicase. One of the most common substrates for the strand-displacement assay is a partially duplexed plasmid. To generate a partially duplexed plasmid, single-stranded, heat-denatured plasmids can easily be hybridized with small fragments of DNA that are either completely complementary to a short nucleotide sequence on the plasmid or partially complementary, leaving one or both ends of the fragment not bound to the plasmid.
Helicases are classified into a number of families and subfamilies, based on functional assays and/or sequence homology. As described above, helicases can be broadly described as DNA helicases, RNA helicases, or DNA/RNA helicases according to the nucleotide substrate on which they act. RNA helicases are subdivided into DEAD- box, DExH-box, and non-DEAD/DExH-box proteins. A large number of RNA helicases, primarily from yeast, have been identified. These proteins play important roles in RNA metabolism. Several important DEAD-box RNA helicases are briefly described to illustrate some of their putative functional roles. The mammalian translation initiation factor elF-4A possesses RNA-dependent ATPase activity (Grifo et al., 1984). elF-4A
most likely binds the 5’ region of capped mRNA and unwinds potential secondary
structure in the message to allow the translational machinery to locate the initiating AUG
codon. The yeast protein Spb4 has been placed in the RNA helicase family by amino
acid homology to other members. Spb4 mutants show a degraded and/or decreased 60S
and 25S ribosomal subunits, suggesting that this molecule is important in the assembly
and/or maintenance of rRNA, possibly altering local secondary structure and allowing
functional conformations to be reached (Sachs and Davis, 1990). The yeast RNA
helicase Prp28 is involved in mRNA splicing (Strauss and Guthrie, 1991). Prp28 may be
31 necessary for unwinding the U4/U6 snRNAs from the mRNA. Current data suggests that numerous aspects of RNA metabolism are dependent on RNA helicases.
Sequence homology throughout the helicase motifs has allowed DNA helicases to be divided into subfamilies named for the first member identified or characterized. The
RAD25 family contains the yeast RAD25 and human XPB (xeroderma pigmentosa, complementation group B). The RAD3 family of helicases includes the yeast RAD3 and human XPD. The overall homology of these proteins is 73%, with 52% identity (Weber et al., 1990). Lack of functional XPB or XPD leads to xeroderma pigmentosum (XP), complemention group B or D respectively. There are seven complementation groups in
XP, each corresponding to a different protein in a complex that is responsible for nucleotide excision repair (NER). Sun sensitivity, neurological abnormalities, and a varitey of skin cancers characterize XP. The XPB protein unwinds DNA in a 3’ to 5’ manner at the 5’ end of the incision flanking a DNA lesion. In a similar manner, XPD unwinds DNA in the opposite orientation, starting at the 3’ end of the lesion. A functional XPB and XPD are required for NER. The human helicases BLM, WRN, and
RecQL all fit into the RecQ helicase family, as do the bacterial RecQ, S. cerevisiae sgs1, and fission yeast rqh1 and rad12. There is a 36% identity and 56% similarity at the amino acid level within the helicase motifs of BLM and WRN.
A number of DNA-dependent ATPases belong to the large SWI/SNF subfamily.
The yeast proteins SWI2/SNF2 and RAD26 and the human proteins SNF2L1, SNF2L2,
SNF2L3, and SNF2L4, as well as Cockayne’s syndrome B (CSB) and α- thalassemia/mental retardation/X-linked (ATRX) all fit into this family. All SWI/SNF family members contain the seven defining helicase motifs and have varying degrees of
32 homology within the motifs. Interestingly, none of the family members exhibit strand- displacement activity in the standard in vitro assays (Cote et al., 1994). SWI2/SNF2
protein is part of a large decameric complex that in turn is part of the larger complex of
transcriptional regulation machinery (Carlson and Laurent, 1994). SWI/SNF proteins are
important in chromatin remodeling, allowing transcriptional machinery to gain access to
DNA packed in nucleosomes (Cote et al., 1994). Likewise, the CSB protein has a DNA-
dependent ATP hydrolysis activity and helicase motifs, yet fails to show strand
displacement activity (Troelstra et al., 1992; Selby and Sancar, 1997). Individuals with
mutant versions of CSB are defective in transcription-coupled repair (Troelstra et al.,
1992). CS patients suffer from growth deficiency, skeletal abnormalities,
photosensitivity, and mental degeneration (Brumback, 1984).
A general model has been put forward to explain the function of the SWI/SNF
subfamily members that takes into account their helicase motifs and lack of strand-
displacement activity. It may be possible that all or some of the proteins that make up
this family do not have the ability to unwind duplex DNA in vivo. However, they may be
able to displace specific proteins from the DNA substrate, allowing transcriptional and/or
DNA repair proteins, including functional helicases, to have access to the DNA. It has
been proposed that SWI2/SNF2 may displace acetylated histones from DNA that is to be
transcribed (Owen-Hughes et al., 1996). Similarly, it has been suggested that CSB
protein may move RNA polymerase II complexes that have stalled as a result of a DNA
lesion on actively transcribed DNA (Selby and Sancar 1993). However, some
experimental evidence contradicts this proposal, as RNA polymerase II is not removed
from RNA during transcription-coupled repair (Selby and Sancar, 1997).
33 Structural and biochemical data have improved our understanding of helicases and have lent support to mechanistic models. In 1996, the crystal structure of a DNA helicase from B. stearothermophilus was reported (Subramanya et al., 1996). Even
though the report describes a bacterial helicase, it is expected that some general structural
information about helicases can be extrapolated. The crystal structure indicates two large
globular domains with a cleft between them that forms a pocket-like configuration. The
cleft is composed of the helicase domains, the most conserved amino acid sequences of
helicases. The pocket-like structure contains the Walker motifs and binds ADP.
Structural homology to bacterial RecA protein is suggestive of a conformational change
driven by ATP binding and hydrolysis. The crystal structure was determined using
protein monomers in the absence of DNA. Additional structural information has come
from biochemical studies and electron microscopy (EM). The active form of all helicases
tested for protein assembly on DNA is oligomeric (Matson, 1991). EM of a
bacteriophage T7 helicase clearly indicates a hexameric ring structure formed around the
DNA (Egelman et al., 1995). The hexameric helicase ring fluctuates between two
allosteric states: a 3-fold and 6-fold symmetry (Yu et al., 1996b). Both ATP and DNA
affect the conformation of the helicase ring (Yu et al., 1996b). Crosslinking studies show
DNA bound to only one, or possibly two, of the subunits at a time (Bujalowski and
Jezewska, 1995; Yu, et al., 1996b). Functionally, this provides the complex with
multiple ATP-binding and hydrolysis sites necessary for the proposed enzymatic
translocation of helicases (Hill and Tsuchiya, 1981). It has been suggested that the DNA
rotates within the hexameric ring from one subunit to the next as the helicase complex
catalyzes the unwinding and translocates along the DNA. Additionally, there is evidence
34 that some helicases function as dimers. The DNA-dependent ATPase activity of E. coli helicases Rep and helicase II are stimulated by homodimerization (Runyon et al., 1993;
Chao and Lohman, 1991). Additionally, the Rep protein forms homodimers upon binding DNA (Chao and Lohman, 1991). The affinity of several dimeric helicases for ssDNA and dsDNA have been shown to be influenced by nucleotide cofactors (Wong and Lohman, 1992; Arai and Kornberg, 1981; Das et al., 1980). This has lead to a rolling mechanism model for multimeric helicases (Wong and Lohman, 1992). It has been proposed that both subunits are originally bound to ssDNA in this model. The trailing monomer releases the ssDNA upon binding ATP and moves to contact dsDNA. ATP is hydrolyzed and some length of dsDNA is unwound to form ssDNA. Finally, ADP is released from that helicase subunit, as the cycle continues. Helicases, in general, seem to bind DNA in non sequence-specific manner. By binding to the sugar-phosphate backbone of the DNA, it is expected that helicases can exhibit polarity and bind to any available stretch of DNA. The Rep helicase and simian virus (SV40) large T antigen, a helicase- containing protein, bind to the DNA sugar-phosphate backbone (Wong et al., 1992;
SenGupta and Borowiec, 1992). The experimental evidence of functional polarity displayed by many helicases supports the nonspecific binding of helicases to DNA.
An important function of some helicases, and important for the work presented in this thesis, is their ability to manipulate DNA during recombination. A number of different helicases from several species are instrumental in various aspects of DNA recombination. It is likely that the normal execution of homologous recombination relies on one or more helicases. Recombination takes place in several stages: initiation, invasion and strand exchange, translocation, and resolution. Initiation involves a single-
35 or double-strand break (DSB) and the pairing of homologous regions of DNA, usually sister chromatids or homologous chromosomes. It is not clear how DNA strand alignment is achieved. Strand invasion and exchange occurs when a portion of single- stranded DNA is brought close to a region of duplex DNA and begins to displace the homologous strand. A heteroduplexed region of DNA, frequently called a joint molecule or Holliday junction, is formed. Several experiments have shown that the bacterial RecA protein is required for proper strand invasion and exchange and that single-stranded binding protein (SSBP) aids in the reaction (Cunningham et al., 1979; Muniyappa et al.,
1984; Roman and Kowalczykowski, 1989). The yeast Rad51, a RecA homologue, carries out the same reaction in yeast. The efficiency of in vitro strand exchange is augmented by Rad52 and replication protein A, a SSBP homologue (Sung, 1997a). Yeast proteins
Rad54, Rad55, and Rad57 increase the ability of Rad51 to coat dsDNA and induce strand exchange (Petukhova et al., 1999a; Sung, 1997b). In addition, Rad59 binds ssDNA, likely important for efficient strand invasion during homologous recombination
(Petukhova et al., 1999b). During strand invasion, there may be a role for the helicase to separate the duplex DNA at the site of the initial DSB, allowing the free ssDNA end to invade a region of duplex DNA. A helicase may also be involved in assisting Rad 51 to open the dsDNA and allow the ssDNA to invade. Once a heteroduplex molecule has been formed, the Holliday junction may translocate along the four strands of DNA, either increasing or decreasing the amount of recombinant DNA during the process. In bacteria, rec BCD proteins are important in progressing the branch migration of the joint molecule (Taylor and Smith, 1980). Also, rec G acts opposite to rec BCD activity, suppressing genetic recombination (Whitby and Lloyd, 1998). Both rec BC and rec G
36 have helicase activity. It has been proposed that these helicases unwind regions of homologous DNA near the Holliday junction by pulling dsDNA through their oligomeric structure at a rapid rate and releasing it as ssDNA, resulting in branch migration. Finally, there are several ways to resolve the recombination event. First, the substrate may no longer be appropriate. If the translocating junction moves into a region of DNA that has a nick or a DSB, it terminates. Holliday junctions can be resolved by cleavage as well, at least in bacteria. The ruvC protein, as part of the ruvABC complex, binds to and cleaves heteroduplex molecules (Eggleston and West, 2000). The ruvABC complex binds joint molecules in a specific manner to direct the orientation of Holliday resolution (van Gool et al., 1999).
With a general background concerning the global structure and putative functions of helicases, this review will now focus on the RecQ family of helicases, leading to survey of the latest literature on BLM. In 1984, a mutant strain of E. coli was characterized with increased sensitivity to UV light and deficiency in conjugational recombination (Nakayama et al., 1984). The RecBC pathway in wild-type cells requires functional recA and recBC proteins to carry out conjugational recombination. In cells with mutant recB or recC, recombination persists at about 1% of wild-type levels through the RecF pathway (Horii and Clark, 1973). A second mutation in sbcB gene allows the
RecF pathway to carryout nearly wild-type levels of recombination. The new mutation was capable of decreasing RecF pathway recombination in a mutant recBC sbcB background 40- to 80-fold. The mutant gene was mapped back to the bacterial chromosome and provided the name recQ1 (Nakayama et al., 1984). The locus containing recQ1 was cloned and a gene identified with the ability to correct the RecF
37 recombinational deficiency in the mutant strain (Nakayama et al., 1985). A subclone that retained the recombination complementation was then sequenced (Irino et al., 1986).
The RecQ gene was recloned for overexpression and protein analysis. Finally, the RecQ protein product was determined to possess DNA-dependent ATPase and helicase activities (Umezu et al., 1990).
In addition to identifying RecQ as a helicase, a number of other genetic and biochemical tests involving RecQ have provided additional characterization of recombination pathways. When recD, a component of the recBCD complex, is mutated, the recombination rate in E. coli is not reduced, which is quite different from the large decrease in recombination noted with recB or recC mutations. In order to test the
potential use of the RecF pathway in recD mutants, a number of RecF pathway genes
were mutated and tested with mutant recD strains. It has been shown that in E. coli with
mutant recD and recQ demonstrate only a slight decrease in plasmid recombination rates
(Lovett et al., 1988). However, plasmid recombination in recBCD, sbcB mutants is
dependent on RecQ (Berger and Cohen, 1989). RecQ mutations do not affect sensitivity
of E. coli to DNA damage by MMS (Wang and Chang, 1991). However, increased
sensitivity to MMS and deficiency in conjugational recombination have been shown the
triple mutant uvrD, helD,andrecQ, suggesting recombinational DNA repair and
homologous recombination require at least one of these helicases in the mutant recBC
sbcB background. In DSB repair, the gap in the DNA may be filled by gene conversion,
with or without crossover of flanking DNA sequences. Mutations in two genes in the
RecF pathway affect the crossover of flanking DNA in DSB repair. Mutations in RecJ,a
5’ to 3’ exonuclease, severely decreases flanking crossover in DSB repair. Mutations in
38 RecQ moderately decrease flanking crossover and suppress the severe decrease of the
RecJ mutant (Kusano et al., 1994). This has led some investigators to hypothesize that
DSB repair with crossover can be accomplished by one of two routes. One pathway requires RecQ and RecJ with RecQ modifying the DNA prior to RecJ. The second pathway does not involve RecQ or RecJ, but is employed when RecQ is not available.
The ruv proteins and recG protein are also required for processing Holliday junctions.
Double mutant strains have very few conjugational recombinants. However, a RecQ
mutation with ruv and recG mutations allow recombination to increase 6- to 20-fold in a
Rec BCD-dependent manner (Ryder et al., 1995). This provides further evidence that
RecQ is essential for conjugational recombination and that loss of this pathway induces a
second pathway to carry out conjugational recombination.
More recently, the RecQ helicase has been shown to suppress illegitimate
recombination (Hanada et al., 1997). These experiments demonstrated that
recombination events increase in vivo at very short regions of homology (8 to 18
nucleotides) as a result of mutations in recQ. RecQ in concert with RecA and single
stranded binding (SSB) protein can initiate and disrupt DNA recombination (Harmon and
Kowalczykowski, 1998). In this report, RecQ and SSB protein unwind linear dsDNA to
form linear ssDNA; the subsequent addition of RecA and homologous supercoiled
plasmids allows various joint DNA molecules to form. Also, given sufficient time, RecQ
will resolve the joint molecules. Although these are in vitro studies, they provide several
interesting clues to the function of RecQ.
A collection of RecQ homologues has been documented over the past several
years. A homologue of RecQ, Sgs1, has been identified in S. cerevisiae and has been
39 shown to interact with yeast type I topoisomerase, Top3 (Gangloff et al. 1994). Sgs1 is required for the maintenance of genomic stability in S. cerevisiae (Watt et al. 1996).
Yeast with mutant sgs1 demonstrate accelerated aging and nucleolar fragmentation, not unlike Werner’s syndrome cells (Sinclair et al., 1997). Sgs1 is a 3' to 5' DNA and
DNA/RNA helicase. (Bennett et al., 1998). Rqh1, a Schizosaccharomyces pombe RecQ homologue is required for reversible S-phase arrest (Stewart et al., 1997). Additional research suggests that rqh1 functions in S-phase in a recombination pathway that prevents cell death as a result of UV-induced DNA damage that cannot be removed (Murray et al.,
1997). Another S. pombe RecQ homologue, rad12, is also important in cell-cycle checkpoint control (Davey et al., 1998).
In 1994 two groups reported the identification of human DNA helicase with significant homology to RecQ. DNA helicase Q1 or more commonly RecQL, is a 73
KDa protein with 47% homology to the seven helicase motifs in RecQ (Seki et al, 1994;
Puranam and Blackshear, 1994). The mouse homologue to RecQL has also been identified (Puranam et al., 1995). Very little is known about RecQL, although some characaterization has been carried out in XPC and HeLa cells (Tada et al., 1996). There is no known disease state associated with mutations in RecQL.
In 1995, two important human RecQ homologues, BLM and WRN, were identified. There are a number of interesting parallels and contrasts related to these diseases, genes, and proteins. BS has been described previously in this review. Werner’s syndrome (WS) is best described as a premature aging disorder (Epstein et al., 1966;
Goto, 2000). Persons affected with WS tend to show signs of aging in their late 30s and
40s that are similar to what occurs in normal persons in their 70s or 80s. In tissue culture,
40 WS cells have reduced replicative life-span, most similar to cells taken from more elderly individuals (Tollefsbol and Cohen, 1984). At the cellular level, while BS cells display increased SCE, WS display increases in chromosomal breaks and have chromosomes with shortened telomeres, relative to normal cells of the same age (Kruk et al., 1995).
Both syndromes share a predisposition to cancer, although WS cancers simply may be a reflection of premature aging (Goto et al., 1981). Both BLM and WRN were identified as recQ homologues by positional cloning (Ellis et al., 1995; Yu et al., 1996a).
Interestingly, with regard to WRN, a particular polymorphic variant has been associated with myocardial infarction in a Japanese population (Ye et al., 1997). Mismatch repair is comparable to normal levels in extracts of WS cell lines (Bennet et al., 1997). WRN has been expressed in the baculovirus system (Suzuki et al., 1997). and verified as a 3’ to 5’
DNA helicase (Gray et al. 1997; Shen et al., 1998). In addition, the WRN helicase
localizes to the nucleolus in human cells (Marciniak et al., 1998). WRN has two
transcription start sites with most of the message initiated from the downstream site
(Wang et al., 1998). The WRN knockout mouse does not have a dramatic phenotype, but
its cells display a loss of cellular proliferative capacity and sensitivity to topoisomerase
inhibitors (Lebel and Leder, 1998).
Finally, two additional members of the human recQ family, RecQ4 and RecQ5,
were identified simultaneously (Kitao et al., 1998). RecQ4 is a large protein, similar to
BLM and WRN. Like BLM and WRN, mutations in RecQ4 have been identified as the
cause of the rare recessive disorder, in the case of RecQ4, Rothmund-Thomson syndrome
(RTS) (Lindor et al., 2000). Clinically, RTS has many similarities to BS and WS.
Poikiloderma, growth deficiency, aspects of premature aging, and a predisposition to
41 malignancy, especially osteogenic sarcomas, characterize RTS. RecQ5α is a smaller member of the human RecQ family of proteins, containing the core motifs and little else.
Three isoforms, generated by alternate splicing of RecQ5, have been identified
(Shimamoto et al., 2000). The smaller isoforms RecQ5α and RecQ5γ remain cytoplasmic. The largest isoform RecQ5β translocates to the nucleus and interacts with topoisomerase IIIα and topoisomerase IIIβ, making RecQ5β functionally homologous to the yeast SGS1.
Since the identification of BLM in 1995, a small body of work on BLM has been reported. A specific mutation in BLM is associated with a topoisomerase II defect in a subpopulation of BS patients (Foucault et al., 1997). Like WRN, BLM is a 3’ to 5’ helicase (Karow et al., 1997). In an experiment designed to test the affects of WRN and
BLM in sgs1 mutant yeast, both WRN and BLM can suppress the hyperrecombination phenotype of sgs1 mutants (Yamagata et al., 1998). BLM can unwind G4 DNA, the compact form of DNA thought to be present in telomeres (Sun et al., 1998). Finally, several groups have also generated mouse models of BS (Chester et al., 1998, Luo et al.,
2000, Goss et al., manuscript in preparation). Unlike WRN knockout mice and humans with BS, BS knockout mice are embryonic lethals. Mutant embryos are profoundly anemic and appear to undergo hematopoietic apoptosis (Chester et al., 1998, Luo et al.,
2000, Goss et al., manuscript in preparation). The same report demonstrated that cultured murine Blm-/- fibroblasts have approximately five times the numbers of SCEs as
normal murine fibroblasts.
Significant study of Bloom’s syndrome has occurred since Dr. Bloom reported the
first cases in 1954. Prior to the identification of BLM, the majority of experiments were
42 aimed at characterizing the disease clinically, cytogenetically, as well as biochemically.
Much of this work was devoted to measuring the high rate of SCE inherent to BS cells and the effects of irradiation and various chemical mutagens on the SCE rate and other
DNA metabolic events, including repair and replication. Several reports examined specific enzymes involved in these activities, including DNA ligases, DNA glycosylases and DNA polymerases. Without strong evidence of a particular protein defect resulting in BS, several groups, including ours, turned to genetic approaches to identify BLM. This thesis will describe the positional cloning of BLM,as well as genetic studies and physical mapping of the locus. With the identification of BLM, our experiments focused on understanding the function of the BLM protein. To this end, work will be presented here that will compare BLM to other RecQ homologues, identify potential protein-protein interactions and describe an animal model of BS.
43 CHAPTER 2. Thesis Rationale.
Genomic instability is a salient feature of BS, one of a number of genetic diseases that exhibit genomic instability. Ataxia telangiectasia cells are characterized by chromosomal breakage; Werner’s syndrome cells are characterized by chromosomal breakage and shortened telomeres. Other chromosome breakage syndromes include
Fanconi anemia (FA), Nijmegen breakage syndrome (NBS), xeroderma pigementosum
(XP), Cockayne syndrome (CS) and trichothiodystrophy (TTD). Persons affected by these syndromes are predisposed to developing cancers, which are somatic cells that are also characterized by genomic instability and chromosomal abnormalities. Microsatellite repeat instability is a characteristic of some sporatic tumors and tumors from persons with hereditary nonpolyposis colon cancer, a result of mutations that affect DNA mismatch repair.
The cytogenetic hallmark of BS is an increase in SCE compared to normal cells.
PHA-stimulated BS lymphocytes grown in short-term culture typically display SCE of 70 to 120 events per metaphase (Chaganti et al., 1974). Normal lymphocytes usually display 10 or less SCE events per metaphase. In addition to DNA exchanges between sister-chromatids, BS cells also display DNA exchanges between homologous chromosomes and telomeric associations. This work was begun prior to identification of
BLM and is based on these observations of chromosomal and genomic instability in BS cells. Experiments described here also demonstrate an increase in microsatellite instability in BS cell lines as compared to normal cell lines. These observations of
44 chromosomal and genomic instability in BS cells led to the hypothesis that BLM gene product is a protein that maintains genomic stability directly and that BLM is a suppressor of DNA recombination. Physical mapping, positional cloning, and the identification of the BLM gene through mutational analysis of BS DNAs led to the conclusion that BLM encodes a RecQ-like DNA helicase. The establishment of a mouse model of BS by gene targeting technology recapitulated numerous characteristics of BS and suggests that haploinsufficiency of Blm may also lead to genomic instability and cancer predisposition.
45 CHAPTER 3. Materials and Methods.
Polymerase Chain Reaction.
PCR amplifications were carried out in 25 µl reactions with 100ng of template
DNA, 0.4 µM of each oligonucleotide primer, 3% dimethyl sulfoxide (DMSO), 0.8 mM dNTPs (Pharmacia), 1.25 units of Taq polymerase (Boehringer Mannheim) and 1X buffer
(Boehringer Mannheim). Each reaction was denatured for 5 minutes at 94°C, followed
by 35 cycles of 1 minute at 94°C, 1 minute at 55°C- 62°C (depending on the primer) and
1-4 minutes at 72°C (depending on size of amplicon), as well as a final 10 minute
extension at 72°C.
Inverse PCR was carried out on YAC and P1 DNA as described in Groden et al.,
(1991) with primers placed in the BLM exon sequences.
Long range PCR was performed in a Hybaid thermocycler (National Labnet
Company), Perkin-Elmer 4800 thermocycler, or Perkin-Elmer 9600 thermocycler using
either GeneAmp XL PCR Kit (Perkin-Elmer), LA PCR Kit (Takara), or QuickChange
(Stratagene) per the manufacturer’s instructions.
Cloning.
Unless otherwise noted all cloning was done as follows: 2-5 µg of vector DNA
was digested to completion at the recommended temperature with the appropriate
restriction enzyme(s) in 1x buffer and BSA as suggested by manufacturer in a total
reaction volume of 150 µl. Fifteen units of calf intestinal phosphatase were added as
necessary to prevent compatible vector ends from religating. Insert DNA was treated
46 similarly, if plasmid-derived, with exception of phosphatase treatment. If insert DNA was derived from PCR, the PCR product was first ethanol-precipitated prior to digestion in a 150 µl reaction as described. Digested inserts and vectors are electrophoresed in 0.8 to 2.5% gels. Bands of appropriate sizes are excised from the agarose and DNA recovered by Qiaquik DNA gel extraction kit (Qiagen). DNA was quantitated visually by electrophoresis with Lambda Hind III digest ladder (NEB). 50 ng of vector were incubated with equimolar quantities of insert DNA in 1x ligase buffer and 200 units of T4
DNA ligase. All enzymes and buffers are from NEB.
Bacterial and Yeast Culture Techniques.
Bacterial cultures were grown in LB or on LB agar (Fisher) with appropriate selection, induction, and/or indication supplements at 37°C overnight or until reaching desired density. Yeast were grown in YPD or on YPD agar (Fisher) at 30°C overnight or until reaching desired density.
Tissue Culture Techniques.
Mammalian cells were culture by standard techniques. The COS cell-line was maintained in DMEM (Gibco-BRL) supplemented with 10% FBS (Hyclone
Laboratories). Transfections of mammalian cells were performed with Fugene6
(Stratagene) according to manufacturer’s instructions. Stable transfectants were identified and maintained under G418 (Geneticin, Gibco-BRL).
Plasmid DNA isolation from bacterial cultures.
47 Bacterial transformations or frozen stocks were plated LB agar with the appropriate antibiotics. Single colonies were picked and and dissolved in 10 ml of LB with the appropriate antibiotics. This culture was grown overnight at 37°Cinanorbital
shaker. The 10 ml culture was used to inoculate 250 ml of LB with the appropriate
antibiotics, which was allowed to grow for 6 hours in an orbital shaker at 37°C. Plasmid
DNA was isolated using Qiagen per the manufacturers instructions.
DNA Sequencing.
DNA was sequenced using fluorescent labeling and/or an ABI 373A or 377A
automated sequencer or manually using the dideoxy chain termination method by the
dsDNA Cycle Sequencing System (Gibco BRL) according to the manufacturer’s
instructions. Templates derived from PCR and SSCP were purified using Centricon 100
columns (Amicon). Sequencing reactions were electrophoresed through 5% denaturing
polyacrylamide gels. The gels were dried and exposed to Hyperfilm-MP (Amersham)
without intensifying screens.
Microsatellite Instability Study
Creation of secondary LCLs.
Lymphocytes from a normal and BS person were transformed by the Epstein Barr
Virus (EBV), through standard methods to create lymphoblastoid cell lines (LCLs). A
single EBV-transformed cell from the both individuals was isolated and expanded
separately in culture forming a primary LCL. From the normal primary LCL, 32 cell
48 lines, and 30 cell lines from the BS primary LCL, were cloned and successfully expanded in culture to generate secondary LCLs.
Isolation of Genomic DNA from Lymphoblastoid Cell Lines.
Cells from secondary LCL clones of both normal persons and BS patients were pelleted and washed once with PBS. Cell counts were made using a hemacytometer.
Each preparation contained more than 5 x 107 cells. These cells also were pelleted by
centrifugation (1500 rpm; 10 minutes) and stored frozen (-196°C).
Each cell pellet was thawed quickly in a 37°C water bath, and the cells were
resusupended in 6 ml buffer (10 mM Tris-HCl, pH 8.0, 100 mM NaCl, 1 mM EDTA).
An equal volume of buffer + 10% sodium dodecyl sulphate (SDS) was added dropwise,
during vortexing of each tube. Proteinase K concentration of 20 µl/ml and the
suspensions were incubated (37°C; 2 hours). The lysates were extracted once with
phenol, once with phenol choroform:isoamyl alcohol, and twice with chloroform:isoamyl
alcohol. The aqueous phases were incubated at 37°C for 1 hour in RNase A (Sigma
Chemical Co.) at a final concentration of 100 µl/ml, and then were incubated at 37°Cfor
overnight in proteinase K at a final concentration of 20 µl/ml. This was followed by two
phenol extractions, one phenol chloroform:isoamyl alcohol extraction, and three
chloroform: isoamyl alcohol extractions. The aqueous layers then were slowly added to
2.5 volumes of isopropanol. The precipitated DNAs were removed with hooked Pasteur
pipettes and allowed to air-dry. The DNAs then were dissolved overnight at room
temperature in TE, pH 7.0. The DNA concentrations were determined and adjusted to
0.2 µg/ml. The DNAs were stored at 4°C.
49 PCR Amplification of Microsatellite Repeat Sequences.
Mismatch repair assay.
Preparation of cell free extracts and mismatched substrates and procedures for measuring mismatch repair were as described (1). Heteroduplex substrates for repair studies contained the mismatch or unpaired base and a nick in the (–) strand at position –
264 (3´-nicked substrate) or +276 (5´-nicked substrate), where position +1 is the first transcribed nucleotide of the LacZa gene. Repair reactions (25 µl) contained 30 mM 4-
(2-hydroxyethyl)-1-piperazine-ethanesulfonic acid (pH 7.8); 7 mM MgCl2; 200 µM each
CTP,GTP,UTP; 4 mM ATP; 100 µM each dCTP, dATP, dGTP, dTTP; 40 mM creatine phosphate; 100 mg/ml creatine phosphokinase; 15 mM sodium phosphate (pH 7.5); 1 fmol of substrate DNA; and 50 µg of extract proteins. Reactions were incubated for 15 minutes at 37°C. The substrate DNA was recovered and introduced into E. coli NR9162
(mutS) via electroporation and plated to score plaques of mixed and pure phenotypes as described (1). Repair efficiency is expressed in percent as 100 X (1 minus the ratio of of the percentages of mixed bursts obtained from extract-treated and untreated samples).
DNA replication assay.
DNA substrate preparation and procedures for measuring SV40 origin-dependent
DNA replication in extracts were as described (2). Replication reactions (100 µl) contained the same components as for repair except that 10 fmol of wt (i.e., blue)
M13mp2 OriL RFI DNA was used and [a-32P]-dCTP and 4 µg of SV40 T antigen was
50 included. Reactions were incubated for 6 hours at 37°C. Replication was analyzed for total incorporation and products examined by agarose gel electrophoresis as described
(2). To analyze the fidelity of the reaction, replicated products were introduced into
E.coli NR9162 (mutS) via electroporation and plated to score plaques as described (2).
The assay scores errors in the wild-type lacZa gene of M13mp2. Correct replication
products yield dark blue M13 plaques, while errors are lighter blue or colorless plaques.
Since the assay measures loss of a gene function (a-complementation of β-galactosidase
activity) that is not essential for phage production, a wide variety of mutations at many
different sites can be recovered and scored (2).
Physical Mapping and Positional Cloning of BLM
Generation of radiation-reduced hybrids.
Radiation-reduced hybrids were generated by the method of Benham et al., 1989.
Briefly, cells from the somatic cell hybrid GS89K-1 (Warburton et al., 1990) were
enriched for those that carried human chromosome 15 by incubation with antibodies to
human beta-2-microglobulin, encoded by a gene mapped to 15q23, followed by
incubation with FITC-labeled goat anti-mouse IgG and purification of the tagged cells on
a fluoresence activated cell sorter (Becton-Dickson FACStar). The sorted cells were
expanded in culture. Ten million of these cells were irradiated for 20 min with 20
kilorads and mixed with an equal number of HPRT- deficient Chinese hamster A23N
cells, fused using PEG and cultured in HAT medium (D-MEM supplemented with
antibiotics). 218 HAT-resistant clones were picked and seeded in HAT medium;
51 proliferation was observed for 192 clones; 177 were expanded successfully and cryopreserved. DNA was prepared by sodium perchlorate extraction and used in subsequent PCR reactions. Hybrids that had taken up the human DNA including the FES locus were identified by PCR as previously described (German et al., 1994).
DNA probes and STSs.
DNA markers in the region containing FES were identified from the literature.
Sequence-tagged sites or STSs were generated from YAC and P1 end-sequences and from subcloned YAC sequences; previously unrecognized polymorphic repeat sequences within some of the YACs and P1s also were identified.
Alu-Polymerase Chain Reaction.
Alu-PCR was performed using DNA from FES-positive hybrid cell clones in 50
mM KCl, 10 mM Tris-HCl pH 8.3, 0.5-1.5 mM MgCl2, 1 µM Alu AM3 (5'-
GAGCGAGACTCC(G/A)TCTCAAA-3'), 10% DMSO, 0.02% gelatin and 0.025 units
per µl Taq polymerase (Boehringer Mannheim). Temperature cycling was performed on
a Gene Machine II or Sun Biosystems thermalcycler with the following parameters: 15
cycles of 94°C for 1 min, 56°C for 2 min and 72°C for 3 min followed by 15 cycles of
94°C for 1 min, 56°C for 2 min and 72°C for 4 min. Alu-amplified DNAs were analyzed
by 1% agarose gel electrophoresis in 1X TBE buffer followed by blotting onto
nitrocellulose (Schleicher & Schuell). Human DNA was identified in 79% of the cell
clones. Inter-Alu DNA sequences were partially purified from FES-positive hybrids by chromatography through Bio-Spin columns (Bio-Rad) and then subcloned into the T/A
52 cloning vector pT7Blue(R) (Novagen) according to the manufacturer's directions. Inter-
Alu sequences were labeled by random-priming (Boehringer Mannheim) and hybridized to Southern blots of Alu-PCR-amplified DNAs according to standard procedures
(Sambrook et al., 1989). The ends of selected clones were sequenced either by a cycle- sequencing method using an Applied Biosystems Automated Sequencer Model 373A or by manual sequencing as described previously (Ellis et al., 1990).
DNA isolation.
YACs were obtained from three sources: the original CEPH YAC library
generated by Albertsen et al. (1990), the ICRF YAC Library and the YAC libraries
maintained and distributed by Research Genetics, Inc. (Huntsville, AL). Yeast stabs were
streaked on YPD media plates and incubated at 30°C overnight. Single yeast colonies
were grown in YPD liquid cultures and incubated overnight at 30°C. Cultures then were
centrifuged and the pellets resuspended in sorbitol solution (6 mls of 0.9 M sorbitol, 0.1
M Tris-HCl pH 8.0, 0.1 M EDTA plus 1.0 ml of 3 mg/ml zymolase in the sorbitol
solution and 1.0 ml of 0.28 M ß-mercaptoethanol). Mixtures were shaken at 37°C for 1
hour and centrifuged. Pellets were resuspended in 50 mM Tris-HCl pH 8.0, 20 mM
EDTAand1%SDS,andincubatedat65°C. 2.0 ml of 5 M potassium acetate was added
to each; the solutions were placed on ice for 30 minutes, centrifuged and ethanol
precipitated. DNAs were resuspended in TE, incubated with RNase A, extracted by
phenol-chloroform, ethanol precipitated and resuspended in TE for PCR. DNA suitable
for pulsed field gel electrophoresis (PFGE) was prepared by the following method:
cultures were centrifuged, washed and resuspended in 750 ul of 50 mM EDTA. 350 ul of
53 solution made from 5 ml SCE buffer (1 M sorbitol, 0.1 M sodium citrate, 60 mM EDTA),
250 ul ß-mercaptoethanol and 5 mg zymolase was added. 1.5 ml of warm agarose solution (1% type VII agarose (Sigma) in 50 mM EDTA pH 8.0) was added, blocks were prepared, placed in 7.5 ml of solution made from 45 ml of EDTA pH 9.0, 250 µlof2M
Tris pH 8.0, and 3.75 ml ß-mercaptoethanol and rocked overnight at 37°C. Blocks were transferred to 45 ml EDTA pH 9.0, 250 µl of 2 M Tris pH 8.0, 5 ml 10% sarkosyl and 50 mg proteinase K and incubated overnight at 50°C. Blocks were washed and stored in 0.5
M EDTA pH 9.0 at 4°C until used for PFGE.
P1 clones were obtained following PCR-based screening of DNA pools from
Genome Systems Inc. (St Louis, MO). P1s were originally carried in the NS3529 strain
of E. coli and then transferred to DH10B by an F’ episome. P1s were streaked on LB
plates containing kanamycin (25 µg/ml). Single colonies were picked and grown in LB
media containing kanamycin. This culture was added to an overnight F3' episome culture
grown from a single colony taken from an LB plate containing chloramphenicol (10
µg/ml). This mating mixture was incubated and plated. A single colony then was grown
overnight in LB containing kanamycin and chloramphenicol at 37°C and then added to an
overnight DH10B culture of LB containing streptomycin (100 µg/ml). This second
mating was carried out under the same conditions as the first. The mating mixture was
plated on an LB plate containing kanamycin, chloramphenicol and streptomycin. Single
colonies from these plates were then cultured overnight in 500 ml of LB and DNA
prepared using Qiagen Maxi Kits.
Isolation of DNA from SSCP conformers was performed as described previously
by Groden et al. (1991, 1993).
54 Isolation of Polymorphic Microsatellites.
Polymorphic sequences from selected YACs were isolated from shotgun libraries prepared from YAC-containing yeast DNAs digested with Sau3A, mixed with BamHI- digested Bluescript DNA and ligated by standard procedures (Sambrook et al., 1989).
The ligated DNAs were transformed into DH5α cells (Gibco-BRL) and the transformants
were bound to nylon membranes (Amersham). The membranes were hybridized
(Sambrook et al. 1989) with a labeled polyCA-polyGT DNA probe (Pharmacia).
Positive transformants were purified, the DNAs were prepared and sequencing was
performed. Sequences were compared to known sequences in the DNA database at the
National Library of Medicine. Any sequence that was not previously identified and that
contained a CA-repeat element with 12 or more repeat units was used to develop a PCR
assay to check the chromosomal origin and to test for polymorphism in a set of human
samples.
Fluorescent in situ hybridization (FISH) analysis of YAC DNAs.
FISH was performed according to Pinkel et al. (1986) with some modifications.
Metaphase chromosomes were obtained from phytohemagglutinin-stimulated human
blood lymphocytes using standard procedures. Slides pretreated with RNase were
denatured in 70% formamide and 2X SSC at 70°C for 2 min. Probes were biotin-labeled
with the Bionick Kit (Gibco-BRL) according to the manufacturer's instructions and
purified on Bio-Spin chromatography columns (Bio-Rad). A quantity of Cot-1 DNA
(Gibco-BRL) equal to half the probe DNA quantity was added to the probe before
55 ethanol precipitation. Biotin-labeled probes were denatured and repeat sequences supressed by pre-hybridization for 1-2 hrs at 37°C. To identify the chromosome 15s, a digoxigenin-labeled chromosome 15 centromeric satellite probe (D15Z1) (Oncor) was denatured separately for 5 min at 70°C and added to the suppressed probe. Then,
hybridization was performed at 37°C overnight. Post-hybridization washes were at 40°C
for 15 min in 50% formamide and 2X SSC, and then at 37°C for 8 min in 2X SSC. Dual-
detection, amplification and DAPI counterstain procedures were performed as specified
by the manufacturer (Oncor). Banded chromosomes and signals were evaluated under a
Zeiss Axioskop fluorescence microscope using the appropriate filters and photographed
on Kodak 400 ASA Ektachrome color slide film.
Pulsed Field Gel Electrophoresis (PFGE).
Lymphoblastoid cells were cultured under standard conditions, washed in
phosphate-buffered saline and resuspended in 1% low-melt agarose in buffered saline at a
concentration of 1.25 x 107 cells per ml. Agarose blocks were prepared and restriction
enzyme digestions performed as described (Birren and Lai, 1993). Blocks were inserted
and sealed into the wells of 1% agarose gels in 0.5 X TBE. Electrophoresis using a
CHEF-DRII system (Bio-Rad) was carried out in 0.5 X TBE at 14°C and 6V/cm with a
pulse-time of 60 seconds for 15 hours followed by a pulse-time of 90 sec for 9 hours.
Saccharomyces cerevisiae chromosomes and/or a lambda ladder were used as size
markers.
Southern blotting.
56 After electrophoresis, gels were stained for 15 min with 0.5 ug/ml of ethidium bromide, photographed with UV illumination, and destained in 0.5 X TBE for 30 min.
Gels were blotted to Hybond-N+ according to the manufacturer's instructions and UV
crosslinked. Hybridizations of PFGE blots were performed as described (German et al.,
1994). Membranes were hybridized with either radiolabeled (kinased) oligonucleotides
or fragments of DNA amplified by random hexanucleotide priming. Membranes were
washed in 0.1% SDS and 0.5X SSC and placed on XAR5 film (Kodak) overnight with or
without intensifying screens.
Subjects and Samples.
Persons with BS in whom low SCE lymphocytes have arisen have been described
previously (German et al., 1996). Epstein-Barr virus-transformed LCLs were developed
from these and other persons with BS by standard culture methods using material
obtained through the Bloom’s Syndrome Registry (German and Passarge, 1989). The
recombinant low SCE LCLs in which reduction to homozygosity had been detected and
the cells used to determine the constitutional genotypes of the five persons from whom
these recombinant low SCE LCLs were developed also have been described previously
(Ellis et al., 1995). The polymorphic loci typed included some previously reported
(Beckmann et al., 1993; Gyapay et al., 1994) and others that were identified during the
physical mapping of the BLM region of chromosome 15 (Straughen et al., 1996). The
methods of preparation of DNA samples, oligonucleotide primers, and conditons for PCR
amplification of microsatellite polymorphisms on chromosome 15 have been described
previously (German et al, 1994; Ellis et al., 1994; Straughen et al., 1996).
57 Direct cDNA Selection.
Direct cDNA selection was carried out as described by Parimoo et al. (1991). In brief, DNAs (15 ng) from commercial λ cDNA libraries prepared from cultured foreskin fibroblasts (Clontech) and Jurkat cells (Stratagene) were amplified by PCR using primer set A for the foreskin cDNA library and the universal M13 forward and reverse primers for the Jurkat cDNA library under standard conditions with Taq polymerase (Boehringer
Mannheim). EcoRI-digested cosmid (c905) or P1 (1958) DNAs (100 ng) bound to hybond-N membrane in 10X SSC were denatured in 0.5 M NaCl and fixed by UV crosslinking. Hybridization of the PCR-amplified cDNAs to repetitive sequences on the cosmid and P1 clones was blocked by prehybridizing the membranes with Cot1 DNA (25 ng/µl; Gibco BRL); poly(dl)-poly(dC) (20 ng/µl; Pharmacia), vector DNA (pWE15 or pAD10SacBII at 25 ng/µl) in 5X SSPE, 5X Denhardt’s solution, and 0.5% SDS at 65°C overnight. Hybridization of the PCR-amplified cDNAs (25ng/µl) was at 65°C for 2 days in the same solution without poly(dl)-poly(dC). The membranes were washed and without elution the bound cDNAs were amplified by PCR with primer set A followed by nested PCR with primer set B for the fibroblast library and the T3 and T7 primers for the
Jurkat library. A sample of the PCR product after each amplification was analyzed by agarose gel electrophoresis, and another was cloned into Bluescript. Independent clones were picked at random, plasmid DNAs were prepared, and insert sizes were determined by restriction enzyme digestion and agarose gel electrophoresis. Inserts from selected clones were purified and used as hybridization probes against all of the other clones as well as against selected genomic DNAs to determine the chromosomal origin of the
58 sequences (see below). The enrichment procedure was repeated and the selected cDNA clones analyzed again. The fibroblast cDNA clone 905-28 was obtained after two rounds of selection (250,000-fold enriched) and was sequenced by the dideoxy chain termination technique (Sanger et al., 1977; Tabor and Richardson, 1987).
The genomic origin of clones isolated by direct selection was verified by hybridization of inserts to Southern blots of DNAs from the following: clones in the contig; human cells; and two human x hamster somatic cell hybrids, one of which contains an intact chromosome 15 as the only human chromosome present (GS89K-1;
Warburon et al., 1990) and one in which the only chromosome material present had, through a translocation, lost all the sequences distal to band 15q25 (GM10664); obtained from the NIGMS Human Genetic Mutant Cell Repository at the Cornell Institute for
Medical Research).
cDNA Cloning, 5’ RACE, and cDNA sequencing.
The selected cDNA 905-28 was hybridized to 106 clones from HeLa cDNA library (Stratagene) according to standard procedures (Sambrook et al., 1989). We isolated 28 λ clones and converted them to Bluescript plasmids by superinfection with
ExAssist helper phage (Stratagene). DNA was prepared, and 15 independent size classes of clones were identified. The 5’ end of a clone from each class was sequenced with
Bluescript SK primer. To extend the sequence, we synthesized two oligonucleotides from the beginning and the end of each of the 5’ sequences, and sequencing was performed on the largest cDNA clone obtained by hybridization (clone H1). This
59 procedure provided sequences from both DNA strands for most of the H1 cDNA.
Ambiguous segments were determined by sequencing with specific oligonucleotides.
Because the reading frame was open at the 5’ end of the H1 clones, additional upstream sequences were obtained by PCR methods. PCR was carried out on DNA prepared from the HeLa cDNA library using the 5’ end of H1 and the T3 sequencing primer. The PCR products were cloned into pT7Blue (Novagen), 18 clones were isolated, and the eight largest inserts were sequenced. The three largest of these clones (5’-5, 5’-
15, and 5’-17) extended the sequence 289 bp 5’ of the H1 cDNA. Database searches then were carried out according to the method of Altschul et al. (1990) using sequence as
queries against the collected amino acid sequence databases that are accessible through
the National Library of Medicine.
A full-length clone referred to as B3 was constructed by performing PCR of a
HeLa library DNA using an oligonucleotide (Y180) from the 5’ end of the H1-5’
sequence and an internal oligonucleotide (BC13) that permitted amplification of a 739 bp
product. EagIandSmaI sites were used to clone the product into NotI–SmaI-digested
H1 DNA.
The 461 bp EagI–SmaI reagment of B3 was isolated and used to probe 8 x 106
clones of a pREP4-clones unidirectional cDNA library from DEB-treated lymphoblastoid
cells (Strathdee et al., 1992). We identified 12 clones, and the 5’ ends of 11 were
sequenced. Of these, eight are apparently full-length cDNAs. By restriction enzyme
analysis, one of the 12 clones contained a deletion 3’ of nucleotide 2897 and an insertion
of about 250 bp.
60 5’ RACE was performed to characterize the 5’ sequences of the candidate gene using a Clontech Marathon cDNA amplification kit according to the specifications of the manufacturer. In brief, first-strand synthesis was carried out with MMLV reverse transcriptase using poly(T)-primed RNAs prepared from cultured fibroblast, lymphoblastoid, and HeLa cells and poly(a)+ RNA from placenta (provided in the kit).
Then, second-strand synthesis was performed with RNase H, E. coli polymerase I, and E.
coli DNA ligase. The DNA ends were made blunt with T7 DNA polymerase, and
adapters with overhanging ends were ligated to the cDNA. Nested PCRs then were
carried out using 5’ oligonucleotides from the adaptor (AP1 and AP2) and internal 3’
oligonucleotides from the H1-5’ sequence (BC5 and BC11). Bands derived from the H1-
5’ sequences were identified in all four of the cDNA samples. PCR products from the 5’
RACE-amplified fibroblasts cDNA were coned into Bluescript, and 5’ends of 12 clones
were sequenced.
Northern Blot Analysis.
RNAs were prepared from cultured cells using TRIzol reagent (Gibco BRL)
according to the instructions of the manufacturer. Total RNAs (30 µg) were size
separated by electrophoresis through 6.3% formaldehyde, 1.2% agarose gels in 0.02 M
MOPS, 0.05 M sodium acetate (pH 7.0), and 0.0001 M EDTA. The RNAs were
transferred to Hybond-N (Amersham) in 20X SSPE and fixed to the membranes by UV
crosslinking. Hybridizations were performed as described previously (Ellis et al.,
1994b).
61 SSCP Analysis.
After first-strand synthesis, PCR was carried out with 200 ng cDNA, 5.2 pmol of each oligonucleotide primer, 3% DMSO, 0.2 mM dNTPs (Pharmacia), 1X reaction buffer
(Boehringer Mannheim), 0.25 units of Taq polymerase (Boehringer Mannheim), and 1.0
Ci of [α-32P]dCTP in a total volume of 10 µl. Each reaction was overlaid with mineral
oil and initially denatured for 5 minutes at 94°C followed by 35 cycles of 94°C for 1
minute, 60°C for 1 minute, and 72°C for 1 minute. The last cycle was extended at 72°C
for 5 minutes. PCR products were diluted in 25 µl of 0.1% SDS, 10 mM EDTA and 25
µl of 95% formamide, 20 mM EDTA, 0.5% bromophenol blue, and 0.5% xylene cyanol.
Two conditions for electophoresis were carried out for each set of reactions. In one, 90
mM Tris borate, 2 mM EDTA (pH 7.5) (Gibco BRL), 35% MDE (AT Biochem) 10%
glycerol gels were used for electrophoresis at room temperature, cooled by fans; in the
other, 90 mM Tris borate, 2 mM EDTA (pH 7.5) (Gibco BRL), 25% MDE (AT
Biochem) gels were used for electrophoresis at 4°C. Electrophoresis was cared out for
both conditions at 40 W constant power in 0.6X TBE running buffer. After
electrophoresis, gels were transferred to 3M paper and dried on a vacuum slab dryer.
Autoradiography overnight with Kodak XAR5film without intensifying screens was
sufficient to detect bands.
Over-Expression and Mutational Studies
Constructs for Over-expression of BLM.
62 Two constructs were used to over-express BLM in mammalian cells. Full-length
BLM cDNA was originally cloned into the pOP-RSVCAT vector (Stratagene) at the NotI
sites, by replacing the CAT gene with BLM. Full-length BLM was also cloned in-frame into the pEGFP-C2 vector by a BglII/BamHI fusion at the 5’ end and Pst I at the 3’ end.
This construct was generated by digesting and ligating fragments from a previously created BLM construct, pBS:BLM. Both pOP:BLM and pEGFP-C2:BLM were used in the tissue culture recombination and frameshifting experiments.
Full-length BLM from pOPRSV:BLM was also cloned into the baculovirus transfer vector pFASTBAC-HTb (Stratagene) at the NotI sites creating pFASTBAC-
HTb:BLM. A triple FLAG (Sigma) epitope, generated by PCR was cloned into the 3’ end of this construct immediately before the stop using the AflII site in BLM and the KpnI site in the vector to make pFASTBAC-HTb:BLM-3FLAG. The insert from this construct was excised and cloned in-frame into pFASTBAC-HTa at the NotIandKpnI sites to generate pFASTBAC-HTa:BLM-3FLAG. This construct was successfully used to express BLM in
SF-21 insect cells. In addition, standard mutagenesis techniques (Stratagene) were used to create point mutants of pFASTBAC-HTa:BLM-3FLAG at highly conserved helicase domain residues. pFASTBAC-HTa:BLM-3FLAG-K695E and pFASTBAC-HTa:BLM-
3FLAG-D795A expressed recombinant BLM that was non-functional and used as negative controls in the in vitro assays. All insert sequences of all constructs were verified by sequencing.
Histochemical Assays.
63 G11 cells (Dr. James Stringer, University of Cincinnati) were rinsed once in PBS
(Gibco BRL) and then fixed for 7 minutes at 4°C in 4% paraformaldehyde (Electron
Microscopy Sciences) in phosphate buffered saline (PBS). The cells were then rinsed once in PBS to remove the fixative, covered in PBS, and incubated at 65°C for 45
minutes to denature endogenous alkaline phosphatases. PBS was removed from the cells
and replaced by a substrate solution (1.0 mg/ml nitroblue tetrazolium and 1.0 mg/ml 5-
bromo-4-chloro-3-indoylphosphate in 100 mM Tris, pH 10.0) that stains the cells purple
to indicate enzymatic activity of placental alkaline phosphatase. After incubating at 4°C
overnight, the staining solution was replaced by PBS and stained foci were counted.
FSH cells (Dr. James Stringer, University of Cincinnati) were rinsed once in PBS
and then fixed for 5 minutes at 4°C in 4% paraformaldehyde in PBS. The cells were then
rinsed once in PBS and incubated overnight at 37°Cinatβ-galactosidase stain solution
(0.5 mg/ml X-gal, 44 mM HEPES, 3 mM potassium ferrocyanide, 3 mM potassium
ferricyanide, and 0.7 mM MgCl2). The cells were then rinsed and stored under PBS and
stained foci counted.
Mouse Model
Lambda-Phage Library Screening.
A λ-phage genomic library derived from 129B6 mouse strain was titered and
plated to a density of three-genomes on 50 15 cm plates. Each plate consisted of a 80 ml
of a 15% agar (Diffco) bottom layer with a top agarose (LE Seakem, FMC) layer (9.0 ml)
composed of 0.6 ml of an OD600 = 0.6 overnight culture of E. coli strain DP50 in LB
64 containing 10 mM MgSO4 and 0.2% maltose, infected for 20 minutes with approximately
10,000 pfu and then mixed into 8.0 ml of 0.7% agarose LB at 45°C. Infection was
permitted for up to 2 days or when clear plaques could be identified on the opaque
bacterial agar. Plates were uniquely marked with India ink and double (set A and B)
membrane (ICN) lifts made of the plaques. Lifts were denatured in 1.5 M NaCl and 0.5
N NaOH, neutralized in 1.5 M NaCl and 0.5 M Tris (pH 8.0), rinsed in 2X SSC, air dried,
and UV-crosslinked.
The lifts were prehybridized with 10 µg salmon sperm DNA for 8 hours in
Denhardt’s solution. After boiling for 5 minutes, a purified, randomly primed
radiolabeled probe created from mouse genomic DNA, originally PCR amplified with
human BLM primers C1-2F and C1-3R was added to the hybridization solution with a
final activity of 106 cpm/ml. After hybridizing for 48 hours, the lifts were washed for 2
hours with increasing stringency every 30 minutes to a final wash of 0.1% SDS and 0.5X
SSC. The lifts were placed on XAR5 film (Kodak) with intensifying screens for 48
hours. Seven positives were identified on set A lifts and confirmed with set B lifts. Six
of these seven positives were successfully amplified and plaque purified through one or
two more rounds of screening.
λ-phage DNA Isolation.
Purified λ-phage clones were grown (10,000 pfu/15 cm plate) independently in a
layer of top agarose over bottom agarose until plaques overlapped. Each plate was
overlaid with 5 ml of Sinsheimer’s medium (SM) (50 mM Tris-HCl, pH 7.5, 100 mM
NaCl,8mMMgSO4, 0.01% gelatin). The top agarose and SM were collected,
65 emulsified, and centrifuged at 10000 x g for 10 minutes at 4°C. The supernatant was decanted and chloroform added to 0.3%. To each 30 ml lysate was added 1.5 g NaCl and
3.0 g PEG-8000, and allowed to incubate overnight at 37°C. The treated lysates were centrifuged at 10,000 x g for 10 minutes. The phage pellet was resuspended in SM and incubated overnight at 37°C with 100 µg of proteinase K. DNA was extracted through
standard phenol and chloroform extraction procedures followed by ethanol precipitation.
Approximately 2 x 107 cells from tissue culture or 3 mm of mouse tail tissue was incubated in 0.5 ml of digestion buffer ( 50 mM Tris-HCl, pH 8.0, 100 mM EDTA, 100 mM NaCl, 1.0% SDS, and 300 µg of proteinase K) at 50°C overnight. After brief mixing
to ensure digestion, samples were phenol-chloroform extracted, chloroform extracted,
and then ethanol precipitated. DNA was spooled out on a pipette tip, dissolved in TE,
and quantitated.
Generation of the Targeting Construct.
Following analysis of phage clones by digestion and Southern blotting a 9.6 kb
SacI fragment was generated by PCR from one of the phage clones and cloned into
Bluescript (Stratagene). This genomic mouse fragment contained sequences homologous
to human exons 10, 11, and 12. These three exons and the intervening sequences were
excised by digestion 5’ with BclI and 3’ with BamHI after preparing the plasmid from the dam methylase deficient E. coli strain GM1963, and replaced by the HPRT selection marker. The negative viral selection gene thimidine kinase was added outside the region of homology
66 ES Cell Targeting.
DNA from the Blm knockout construct was ethanol precipitated and resuspended in sterile water to a final concentration of 80 nM and delivered to the Gene Targeting
Facility at the University of Cincinnati. The construct was targeted in 129B6 ES cells by standard techniques (Thomas and Capecchi, 1987). ES clones were selected and screened by PCR and Southern blotting. Two ES cell clones were identified.
Blastocyst Injection.
Two ES cell clones were given to the Blastocyst Injection Facility at the
University of Cincinnati and injected into blastocysts using standard procedures (Gossler, et al, 1986).
Establishment of Mouse Colony.
One correctly targeted cell line was microinjected into 129/SvEv blastocysts.
Resulting chimeric males were mated to Black Swiss (Taconic Laboratories) outbred mice (to test for germline transmission and for aging experiments) and backcrossed to
129/SvEv (Taconic) to produce a co-isogenic strain. All mice were housed in micro- isolator cages; animal husbandry was provided by the University of Cincinnati
Laboratory Animal Medical Services. Logistic regression was used for the survival curve and risk assessment for spontaneous lymphoma development in Blm+/- compared to
Blm+/+ littermates.
MLV-induced tumorigenesis in Blm+/+ and Blm+/- mice.
67 Three litters of neonatal pups from a Blm+/- female and wild-type male (Black
Swiss in the F7 generation) were injected with 100 µl Murine Leukemia Virus (MLV)
intraperitoneally at a concentration of 1x105 pfu/ml (Haupt et al., 1991; van Lohuizen et
al., 1991). Mice were monitored for signs of tumor formation and subjected to necropsy
at the time of death. Samples of thymic lymphoma were frozen in OCT embedding
medium for histological analysis. 5-µm sections were stained with anti-CD4 (Pharmigen,
1:100) and anti-CD8 (Pharmigen, 1:250) biotinylated rat anti-mouse antibodies.
Immunoreactivity was detected with AEC colorometric substrate (Zymed). Tissue
sections were counterstained in Mayer’s hematoxylin; serial sections were stained with
hematoxylin and eosin. Statistical analysis of age at the time of death was performed
with a Student T-test.
Establishment and Analysis of Blm+/- primary fibroblasts.
Fibroblast cultures were derived from explants of lung tissue from Blm+/+ and
Blm+/- mice and maintained in DMEM (GIBCO) supplemented with 10% fetal bovine
serum (Hyclone) at 37ºC and 5% CO2. Wild-type and Blm heterozygous cells were harvested at the 10th and 8th passages respectively. The differential staining of sister chromatids in cells that were cultured in 10uM bromodeoxyuridine utilized Hoescht
33258, exposure to light, followed by staining with Giemsa as described previously in
German et al. (1977). The SCEs in intact metaphases were enumerated by light microscopy with a 100X objective. The numbers of SCEs per 40 chromosomes were determined for both diploid, near-diploid and aneuploid cell lines. Fifteen cells were scored from the Blm+/+ and Blm+/- cultures. Micronuclei were quantified in these
68 cultures. A minimum of 1000 intact cells was examined for the presence of micronuclei or smaller purple staining bodies located outside the nucleus. Variations in size and number of micronuclei were observed in preparations from both mice strains. Two observers that were uninformed of the genotype of the three cell lines performed analyses independently. Statistical analysis of the percentage of cells with micronuclei was performed with a Chi-squared test with significance established at p≤0.05.
ApcMin/+and Blm+/- crosses and tumor analysis.
C57BL/6J-ApcMin male mice (The Jackson Laboratory) were mated to Blm+/-
female mice (backcrossed to 129/SvEv to the N3 generation) to obtain progeny that were
ApcMin/+,Blm+/+ or ApcMin/+,Blm+/-. Mice were genotyped at Blm as described above and
for the ApcMin mutation as previously described (Wilson et al., 1997), and sacrificed at
approximately 120 days of age. Tissues were harvested and fixed in 4%
paraformaldehyde overnight at 4°C and then transferred to 70% ethanol for analysis and
storage. Tail DNA was re-genotyped at the time of sacrifice. The gastrointestinal tract
was opened longitudinally, intestinal contents removed, and tumors evaluated using a
Stemi SV6 dissecting microscope (Zeiss) at 8X. Statistical analyses were performed with
the non-parametric Wilcoxon Rank Sum Test using Prism Version 2.0 statistical software
(GraphPad Software, Inc). For histological analysis, tissues were paraffin-embedded and
4-µm sections stained with hematoxylin and eosin (H&E). The gastrointestinal tract was
divided into four equal segments and representative sections analyzed using a BH-2 light
microscope (Olympus) at 1-100X magnification. Representative histological sections of
all tissues, including brain, mammary gland, testis, thymus and blood, were similarly
69 assessed. Gross tissue specimens and H&E sections were photographed with a DXC930 color video camera (Sony) using Scion Image 1.60 and NIH Image 1.60 software configured to the dissecting microscope or upright light microscope.
Molecular analysis of lymphomas and adenomas.
Tissues from moribund animals were harvested and fixed in 4% paraformaldehyde. Representative paraffin-embedded sections were stained with H&E and examined for pathology. Paraffin-embedded lymphoma or adenoma, and adjacent normal tissue were micro-dissected and DNA was extracted by xylene extraction, proteinase K digestion and ethanol precipitation. Five adenomas from ApcMin/+,Blm+/+
and 20 adenomas from ApcMin/+,Blm+/- mice were analyzed. The status of the Blm gene in tumor and normal tissue was characterized by PCR and Southern blotting. Primers 1F,
HPRTF (5’-GCCAGCTCATTCCTCCGACTCA-3’) and MI-22R (5’-
GTATTAGCGCACATAATATGAGAACA-3’) were used to amplify both the targeted allele (780 bp) and the wild-type allele (400 bp). The PCR conditions were: 94°C for 30 sec.; 94°C for 30 sec., 60°C for 30 sec., and 72°C for 30 sec. (30 cycles); 72°C for 5 min.
The reaction mixture contained: 100 ng DNA, 0.4 µM primer MI-22R, 0.2 µM primer
1F, 0.2 µM primer HPRT4F, 0.4 mM dNTPs, 3% DMSO, 0.8 units of Taq polymerase,
10 mM Tris-HCl, 1.5 mM MgCl2 and 50 mM KCl pH 8.3. PCR products were
electrophoresed through a 2% agarose gel and probed with a 32P-labeled oligonucleotide
(5’-ATATTCCTCACTACAAGTGGCAT-3’) that hybridizes to both the targeted and
wild-type PCR products. PCR amplification was quantified with a Molecular Dynamics
phosphorimager and analysis software.
70 The Apc locus was analyzed by a PCR-based quantification assay described by
Luongo et al. (1994). Both Apc+ and ApcMin alleles were amplified by PCR and digested with HindIII, resulting in 144-bp (ApcMin/+ allele) and 123-bp (Apc+ allele) products that
were separated by electrophoresis in a 6% denaturing polyacrylamide gel. Additionally,
chromosome 18 microsatellite markers D18Mit19, D18Mit17 and D18Mit123 were
analyzed by PCR. 32P-labeled primers (D18Mit19, 5’-ATTGGGTGTTCAGGTGCAG-3’
and 5’-ATGCACAATAGCTCATAGCTTCT-3’; D18Mit17,5’-
TCAGGCAGATTCCAAGCAG-3’ and 5’-CTGTGGGTAGCCCAAGTCAT-3’;
D18Mit123, 5'-GGAATATATTACAGAAGAAAGCACAGG-3' and 5'-
TCTGACACTGACTGGAACTACACA-3') were used to amplify the 129/SvEv and
C57BL/6 alleles of chromosome 18. PCR conditions were: 5 min at 94°C; 94°C for 1 min., 58°C for 1 min., and 72°C for 1 min. (30 cycles). The reaction mixture contained:
100 ng DNA, 0.4 µM forward primer, 0.2 µM reverse primer, 0.2 µM kinased reverse primer, 0.4 mM dNTPs, 0.8 units Taq polymerase, 3% DMSO, 10 mM Tris-HCl, 1.5 mM
MgCl2 and 50 mM KCl pH 8.3. All products were resolved on 6% denaturing
polyacrylamide gels and quantified as described above.
71 Chapter 4. Microsatellite Instability in Bloom’s Syndrome Cells
Cytogenetic studies of cells from persons with BS have shown that a higher than normal frequency of sister chromatid exchanges (SCE) is present in BS cells (Chaganti et
al., 1974). SCEs are visualized by growth of cells for two cell cycles in the thymidine
analogue 5-bromodeoxyuridine followed by a staining procedure that differentiates the
two pairs of sister chromatids. In normal human lymphocytes, only a few SCEs are
present per metaphase spread; in BS, the number is eight-to-ten-fold higher (Figure 1).
Whether these exchanges are equal or unequal is unknown, although some data suggest
that, at least at some frequency, they can be unequal (Groden and German, 1992). A high
incidence of chromosome breakage and homologous chromosome associations known as
quadriradials (Qrs) also have been observed in cultured cells from BS individuals
(German et al., 1965). These Qrs represent exchanges of genetic information between
homologues as proven by the high incidence of loss of heterozygosity that affects
clonally derived BS cell lines as compared to the parental cell line (Groden, Nakamura
and German. 1990). Terminal associations (TAs) of homologous chromosomes also have
been observed at a higher frequency in BS than in normal cells (German, unpublished
observations). These chromsome configurations are believed to represent the exchange
of genetic information between homologues in a similar manner to the Qr, with the point
of exchange located distally on the chromosome arm. Alternatively, these associations
may imply a dysfunction with the mechanism(s) by which telomeres are maintained.
72 This study used DNA from clonally derived lymphoblastiod cell lines (LCLs) that had been cultured non-clonally for over one year (Figure 2). These two cell lines originally were derived from lymphocytes taken from one BS person and one normal control subject. During this year, it was assumed that recombination events or mutations would occur at a characteristic frequency in each cell line, and that once single cells were isolated and expanded as clones, these alterations could be assessed precisely. The analysis of dinucleotide repeat loci revealed the frequent generation of novel alleles at many of these loci throughout the genome in BS cells, suggesting that the types of chromosomal events that are elevated in BS include at least one or more that are associated with the generation of these highly polymorphic alleles. In addition, we demonstrated that disease-associated trinucleotide repeat sequences also undergo expansion and contraction in BS cells. In total, these observations suggest that the mechanism through which these repeats are altered may be recombination-associated unequal sister chromatid exchange as they occur in both normal and BS cells at a frequency reminiscent of the SCEs that occur in these two cell types.
Lastly, the observation of microsatellite instability in BS cells suggested that the disruption of the BLM gene, in a manner similar to mutations in the human homologues of the bacterial DNA mismatch repair genes (hMLH1, hMSH2, PMS1,andPMS2))in
hereditary non-polyposis colon cancer patients and tumors characterized by genomic
instability, may alter normal mechanism of DNA mismatch repair. Studies using lysates
from BS and non-BS cells, however, did not identify any differences in the abilities of
these lysates to single base pair mismatches or to support DNA replication.
73 lymphocytes from whole EBV blood
EBV transformed lymphocytes
Primary
Secondary
Figure 2. Generation of secondary lymphoblastoid cell-lines (LCLs). Peripheral blood was collected from a person with BS and a normal person and fractionated by centrifugation. Lymphocytes were then placed in culture media with Epstein-Barr Virus (EBV) to transform B-lymphocytes into lymphoblastoid cell-lines. After a year in culture cells from each cell line were isolated, diluted to one cell, and expanded in culture to create the primary LCLs. The secondary LCLs were generated by isolating individual cells from one of the primary LCLs and allowing them to expand in culture. DNAs from the secondary LCLs were prepared and used in further experiments.
74 BS cells show an increased frequency of dinucleotide repeat instability.
As part of our goal to characterize the genomic instability of BS cells, analysis of eight telomeric dinucleotide repeat loci was performed using DNA from BS and control cell lines. Each dinucleotide repeat locus mapped to a different chromosomal arm. For each locus, the same set of 30 BS clonal cell lines was examined for alterations. These experiment relied on PCR, with primers flanking microsatellite repeat sequences (Table
1). Our results indicate that four of the eight dinucleotide repeat loci have novel alleles in at least one of the BS clones. The novel alleles were present in different clones. We found examples of both smaller (Figure 3) and larger novel alleles. There is one instance where several clones have a novel allele at the same locus, likely the result of a single
DNA altering event that occurred prior to isolation of cells for the secondary clones. The normal clonal cell lines were also tested in the same manner as the BS cell lines. Only one of the dinucleotide repeat markers indicated a novel allele in normal clonal cell lines
(data not shown). SCE is a normal event and is expected to occur in normal cells, but at a very attenuated rate as compared to BS cells.
BS cells show an increased frequency of trinucleotide repeat instability.
Loci for four trinucleotide repeat diseases were also evaluated in both the normal and BS cell lines. The loci examined were those for spinal and bulbar muscular atrophy, myotonic dystrophy, Huntington's disease, and fragile-X. The normal clones were all heterozygous and carried identically sized alleles at each locus. However, one BS cell
75 Table 1. Oligonucleotides Used to Amplify Microsatellite Repeat Sequences
Locus Primer Numbers Forward Primer Reverse Primer
4p16.3 HD3/ HD5 CCTTCGAGTCCCTCAAGTCCTTC CGGCTGAGGCAGCAGCGGCTGT
D5S392 AFMO28xb12a/m GCTATTCCCACAAAGGCA GGCGGATCATTGAGTGC
D7S972 MFD172A/MFD172B GTCAAAAAGAGTTATGCTTG TGTTTAATTCCATAAGCCCA
D8S201 MFD199A/MFD199B TGGCTAACACGGTGAAACCA ATCAGACCAATAACCCCAGG
D11S922 AFM217yb10a/m GGGGCATCTTTGGCTA TCCGGTTTGGTTCAGG
D15S49 ACTC-A /ACTC-B CTCCCCCACACAAAGAAG TCGTTCCCAGGTATGGAA
D17S261 MFD41A/MFD41B CAGGTTCTGTCATAGGACTA TTCTGGAAACCTACTCCTGA
D18S40 CU18-001A/CU18-001B CAAGATAGATGCATTTTCCAGT CATCCAAAGGGTGAATGTGT
D18S50 CU18-009A/CU18-009B TACTCAGGGCAACCCCTAAA CAATTCTATTACTGTCTTCTTTTGCAT
D19S95 DM-101/DM-102 CTTCCCAGGCCTGCAGTTTGCCCATC GAACGGGGCTCGAAGGGTCCTTGTAGC
D19S180 MFD195A/MFD195B CTAAATATCCATCAAGGAATG TACTCCATTTTCATTCAGGT
20q13 UT242A/UT242B TTCCTGATCAGTCACCATGTA CAGCCTGGGTGACAGAGAA
Xq11-12 SBMA-3/ SBMA-5 CTGGTGAAGGTTGCTGTTCCTC GCCTGTTGAACTCTTCTGAGC
76 D5S47 D18S5 18 19 20 21 22 23 24 25 27 28 29 30 31 32
UT24 LNSC 56789101112 17 18 19 20 21 22 23 24
Figure 3. Dinucleotide repeat instability in BS secondary clones. DNA from clones of BS cells, as described in Figure 4-1 and in Materials and Methods, was used as a template in a PCR assay. Primers that flank dinucleotide repeat sequences were used, with one of the primers 32P-radiolabeled at its 5’ end. The resulting PCR products were electrophoresed through nondenaturing polyacrylamide gels. The gels were dried and autoradiographed. Specific primers are noted above each image, and lanes labeled with the numerical designation of the clone.
77 line showed alleles of novel sizes at two of the four trinucleotide repeat loci. The spinal and bulbar muscular atrophy and the Huntington's disease (Figure 4) both generated smaller than expected alleles in one BS clone.
BS cells with novel alleles at di- and trinucleotide repeats are non-recombinant for distal markers on the same chromosome.
To support our hypothesis that the novel alleles observed in the BS subclones were the result of recombination between sister chromatids on the same chromosome and not those on homologous chromosomes, ie. intra-chromosomal events, we examined some of these subclones for the retention of heterozygosity at more distal regions of the chromsome. At the microsatellite repeat loci we evaluated on chromosome 4p
(Huntington’s disease gene) and 5q, we identified clones that contained uniquely sized repeats. We then examined additional microsatellite repeat loci more distal on these chromosomes for loss of one allele. The clones that had alterations in the original loci evaluated on chromosome 4p and 5q did not have loss of either allele at more distal microsatellite repeat loci (data not shown). These data support the argument that novel microsatellite repeat alleles in BS cell lines is the result of SCE, rather than recombination between homologous chromosomes,
Furthermore, the appearance of novel alleles in the two BS clones at the androgen receptor gene associated with X-linked spinal and bulbar muscular atrophy must have been generated by an intrachromosomal mechanism because the disease gene is X-linked and the BS patient from whom the cells were derived was a male.
78 Huntington’s Disease
Figure 4. Trinucleotiderepeat instability in BS secondary clones.DNA from secondary clones of BS cells, created as described in Figure 4-1 and in Materials and Methods, were used as templates in a PCR assay. Primers that flank thetrinucleotiderepeat sequence in the Huntington disease gene were used, with one of the primers 32P-radiolabeled at its 5’end. The resulting PCR reaction was run on a nondenaturing polyacrylamidegel. The gels were dried andautoradiographed. Lanes are labeled to indicate the secondary clones analyzed.
79 A BS clone with two novel alleles at a single locus.
A single BS clone at the dinucleotide repeat loci D8S201 has two novel alleles
(Figure 5). The original alleles in this clone are still evident, indicating the microsatellite altering event took place after isolation of the secondary clone. The intensity of the novel alleles appears to be similar, but the sizes were not determined. Equal intensity implies that the event that generated these new alleles took place in the same cell at the same time. The most likely explanation for a single clone with novel smaller and larger alleles of equal intensity is a single recombination event occurring at this microsatellite locus.
Analysis of BS protein extracts for proficiency in DNA mismatch repair.
To determine if 1525 BS cells are deficient in mismatch repair, cell free extracts were prepared and tested for their ability to repair a variety of mispaired substrates in vitro.WeusedanM13mp2 DNA substrate containing a covalently-closed (+) strand and a (–) strand with a nick (to direct repair to this strand) located several hundred base pairs away from the mispair located in the lacZ a-complementation gene. The (+) strand encodes one plaque phenotype (either colorless or blue) and the (–) strand encodes another plaque phenotype. If the unrepaired heteroduplex is introduced into an E. coli strain deficient in methyl-directed mismatch repair, plaques will have a mixed phenotype due to expression of both strands. However, repair in a repair-proficient human cell extract will reduce the percentage of mixed plaques and increase the ratio of the (+) strand phenotype relative to that of the (–) strand phenotype, because the nick directs repair to the (–) strand.
80 15 16 17 18 19 20 21 22 23
Figure 5. Repeat instability with novel larger and smaller alleles. DNA from secondary clones of BS cells, created as described in Figure 4-1 and in Materials and Methods, were used as templates in a PCR assay. Primers that flank the dinucleotide repeat sequence D18S50 were used, with one of the primers 32P-radiolabeled at its 5’ end. The resulting PCR reaction was run on a nondenaturing polyacrylamide gel. The gels were dried and autoradiographed.
81 An extract of 1525 BS cells repairs a G•G mispair as efficiently as a HeLa cell extract (Figure 6). Repair is observed regardless of whether the nick is 3´ or 5´ to the mismatch, consistent with the bi-directional repair capability of the human mismatch repair system. The change in the ratio of blue to colorless plaques indicates that repair is specific for the minus-strand, as directed by the nick in that strand. The 1525 extract also repairs substrates containing an A•C mismatch or either of two different unpaired nucleotides. In contrast to these results, extracts of cell lines exhibiting microsatellite instability and having mutations in any of four mismatch repair genes (hMSH2, GTBP,
hMLH1,orhPMS2) are uniformly deficient in strand-specific mismatch repair (Boyer et
al., 1995).
Analysis of DNA replication. Since an extract of 1525 cells is mismatch repair
proficient, we next examined whether the microsatellite instability might result from
reduced replication fidelity. The proteins present in human cell extracts can completely
replicate double-stranded DNA substrates containing the SV40 origin, with only the
addition of SV40 T antigen required to initiate replication at the origin. Although
replication activity is readily observed in a HeLa cell extract using 50 µg of extract
protein (Thomas et al., 1991), an initial replication reaction with an equivalent amount of an extract of 1525 cells using the standard amount of T antigen (1 µg) yielded no replication. We then added four times this amount of T antigen and did obtain replication, but at a slightly lower efficiency than with a HeLa cell extract under identical conditions (65 versus 98 picomoles total dNTP incorporated for extracts of 1525 and
HeLa
82 G:G (3' nick) G:G (5' nick) 80 80 60 60 40 40 20
% Repair 20
% Repair 0 0 HeLa BS HeLa BS
G:G (3' nick) G:G (5' nick) 2.5 4 2.0 3 1.5 2 1.0 0.5 1 Blue/white Blue/white 0 0 Mock HeLa BS Mock HeLa BS
Figure 6. Bloom’s syndrome cells are as efficient as HeLa cells at mismatch repair. A M13mp2 duplex DNA substrate was used for the MMR assay. The substrate contained a circular (+) strand and a nicked (-) strand intended to direct repair to the (-) strand. The (+) strand contains the β- galactosidase gene. The substrate has a G/G mismatch the error in the (-) strand near the sequence complementary to the 5’ end of the β-galactosidase gene. If the (-) strand is repaired in the cell free system and the reaction products transformed into bacteria, blue colonies will result. However, a stop codon is coded in the β-galactosidase gene if the (+) strand is repaired, and white colonies will be seen. A mixture of blue and white colonies will be seen if neither strand is repaired in the cell free system, since the bacterial system will repair either strand with equal efficiency.
83 cells, respectively). No replication was observed in reactions incubated in the absence of
T antigen.
When replication products from both reactions were analyzed by electrophoresis in an agarose gel in the presence of ethidium bromide, the product distributions for 1525 and HeLa reactions were similar. The majority of the products migrated coincident with supercoiled or relaxed monomer-length, circular DNA, with some high molecular weight
DNA products also observed (Roberts and Kunkel, 1994). Diagnostic restriction endonuclease digestions indicated that the radiolabeled DNA products resulted from semiconservative replication (Roberts and Kunkel, 1994).
To assess the fidelity of replication in extracts of 1525 cells, the DpnI-resistant replication products were introduced into E. coli to score LacZ a-complementation mutants. The mutant fraction was only slightly above the background mutant frequency of unreplicated DNA, indicating that few if any errors were generated during replication in vitro. This is similar to previous observations with a HeLa cell extract that yielded replication error rates varying from 6.2 x 10-6 to 0.1 x 10-6, depending on the base
substitution or single-base frame shift error considered (Thomas et al., 1991). Thus, no
obvious replication error phenotype is detectable in this forward mutation assay.
Discussion.
The molecular analysis of DNA from BS cells previously has revealed the
occurrence of homologous recombination and alterations in midi-repetitive DNA
sequences. The work presented here augments previous observations with the addition of
microsatellite repeat mutations to the variety of genomic instabilities in BS cells.
84 Instability at microsatellite loci is a characteristic of a large number of sporadic tumors as well as of the tumors found in individuals affected by hereditary non-polyposis colon cancer (HNPCC) (Young et al., 1993). Six human DNA mismatch repair genes, homologues of DNA mismatch repair genes of E. coli have been identified (Bronner et
al., 1994; Fishel et al., 1994; Horii et al., 1994; Paquis-Flucklinger et al., 1997; Lipkin et
al., 2000). Mutations in these genes are present in the germline of HNPCC patients and
in sporadic tumors characterized by genomic instability (Fishel et al., 1994). The
mutation of both copies of any one of these genes in a somatic cell is thought to render
the genomes of this cell and its progeny unstable. Consequently, the likelihood that a
coding region, perhaps one at a locus involved in growth control, will become mutated
may also be high.
The dinucleotide repeat instability found in the BS cell lines is reminiscent of the
genomic instability found in the colorectal tumors first characterized by Perucho et al.
(1992). It may be, in part, this type of genomic instability that ultimately contributes to
the high predisposition to neoplasia in BS patients. Interestingly, many of the solid
tumors that have been documented in BS individuals are those that are associated with
genomic instability, especially tumors of the colon, rectum, endometrium and bladder.
Recently, Lowy et al. reported a person with BS who developed multiple adenomas in the
colon, primarily at the cecum, with a clinical presentation reminiscent of familial
adenomatous polyposis coli. The Bloom’s Syndrome Registry also has documented a
high incidence of gastrointestinal tumors in persons with BS (Lowy et al., 2001).
Although BS cells lines exhibit microsatellite instability, no mismatch repair
defect was detected in our assays. It is possible that BLM is involved in a more
85 specialized mismatch repair pathway that was not tested here, such as the repair of small or large loops. Alternatively, a more sensitive reversion assay may identify more subtle errors in the mismatch repair process as a result of BLM mutations. It is interesting to note that microsatellite repeat instability of BS cells is lower than that observed in known mismatch repair deficient cell lines. These observations support the argument that BLM is not part of the mismatch repair system, and leaves open the possibility that unequal
SCE recombination results in microsatellite repeat instability.
Interestingly, the BLM helicase interacts directly with hMLH1, a mismatch repair protein (Langland et al., 2001). Interaction between the C-terminus of BLM and hMLH1 was identified by a yeast two-hybrid screen and confirmed by far Western and co- immunoprecipitation techniques. It is not clear how or whether BLM is involved in mismatch repair. One possible explanation for the interaction is that the replication complex stalls after a DNA mismatch is incorporated by the DNA polymerase. BLM and the mismatch repair machinery then repair the defect. BLM unwinds the DNA, allowing
RPA to maintain strand separation. BLM then detaches from the DNA and MLH1, and the repair complex completes the mismatch repair. As noted above this may only occur in particular mismatches or loops not tested in our mismatch repair assay. Other possibilities include a repair complex that contains proteins that can initiate mismatch repair as well as double-strand break repair. This complex could be tethered to the replication complex or present in sufficient concentration to identify and initiate the specific repair necessary. Once the repair, either mismatch or double-strand break, is initiated, the MLH1-BLM interaction ceases and the appropriate protein continues to
86 interact with the DNA, directing the DNA repair. The MLH1-BLM complex may serve as a common entry and decision point for immediate post replicative repair mechanisms.
The observation of the three normal subclones with alterations at two dinucleotide repeat loci also is note-worthy. SCE is a characteristic of all cells examined so far with few exceptions; baseline SCE levels in untreated cells are low but they do occur. If dinucleotide repeat alleles are generated at a particular frequency in the germline and somatic cells of normal individuals, as evidenced by the high degree of polymorphism at these loci in the normal population, it is reasonable to predict a low level of instability in the normal subclones. The detection of two clones with the identically sized alteration is most likely a cell culture artifact, with two cells of the same genotype cloned from the unselected population. In support of this argument the intensity of the novel bands is approximately that of the other parental band in these two clones. These clones, therefore, may represent only one altering event that occurred earlier prior to the secondary cloning.
The high level of SCE detected cytogenetically in BS cells and the occurrence of a low level of exchanges in normal cells, supports the hypothesis that SCE is responsible for the gains and losses of repeats observed in the di- and tri-nucleotide repeat sequences.
However, another hypothesis that can be considered is that exchanges and recombination may occur after an error is introduced into a DNA sequence and it is this error-prone replication process that is increased in BS cells. The increase in SCE frequency and recombination would then be a consequence of this increased mutation frequency.
Furthermore, this work provides evidence that these types of polymorphic alleles, comprised of short repeat sequences, may be generated by unequal sister chromatid
87 exchange. Most work in the literature suggests that polymerase slippage during replication, gene conversion, or even copy choice is responsible for the generation of these novel repeat lengths. DNA polymerase α and β are normal in BS cells. Therefore, it is unlikely that abnormal function of either enzyme is generating the high level of chromatid exchange and microsatellite instability observed here. Experiments performed as part of this work indicated that DNA replication in BS cell extracts is as proficient, but at reduced efficiency as compared to HeLa cell extracts. The reduced efficiency of BS cell extract replication is interesting in light of the fact that BLM is a helicase, yet fails to explain genomic instability. The addition of increased concentrations of SV40 large T antigen, a helicase activity-containing enzyme, was required to allow replication in BS extracts comparable to HeLa extracts. Wolff et al. reported that a new allele was generated at a VNTR locus without exchange of flanking DNA sequences; their work also supports the idea that these errors of replication can be intrachromosomal, but does not argue one way or the other between unequal SCE or slippage. Our previous analysis of a midi-satellite tandem-repeat locus (D1Z2) in these same BS and normal cell clones revealed high rates of mutations at this locus in the BS cells. Unequal sister-chromatid exchange giving rise to intra-locus mutation was concluded to be the most plausible explanation for the accumulation of the changes detected. In this example, the size of the repeat was 120 base pairs, significantly larger than microsatellites.
The potential relationship between SCE and microsatellite repeat instability in our experiments can be examined. With the assumption that there are approximately 100
SCE events in each cell cycle in BS cells and the primary clone experiences 100 cell cycles prior to secondary clonal selection (the primary clone was grown for 18 months
88 prior to DNA isolation), then any particular cell selected as a secondary clone would have the opportunity to have at least 10,000 SCE events. If SCE events occur randomly in the genome (3 x 109 bp) and the genome is parsed into 100 bp fragments (the approximate size of the PCR products examined in these instability experiments), there would be 3 x
107 100-bp fragments. One would expect to see one incidence of instability in every 300
of these fragments examined, if every SCE that occurred resulted in a novel allele. This
is far less than the 12% of loci examined that have microsatellite instability in our assay.
If it is assumed that SCE events occur primarily at repetitive DNA, which is
conservatively estimated to make up approximately 5% of the genome, then only 5 x 105
100-bp fragments need be considered. This would result in an instability rate of nearly
20%, if every SCE that occurred resulted in a novel allele.
Persons with BS are affected by a number of other unusual clinical
manifestations, in addition to the high cancer incidence. These include immuno-
deficiency, sun sensitivity, small stature, microcephaly, learning disabilities, diabetes and
reduced fertility. Perhaps some of these manifestations of BS can be correlated with
expansions of gene-associated repeats in somatic tissues. For example, patients affected
by myotonic dystrophy frequently are found to develop abnormal glucose responses in
addition to the muscle abnormalities. The alteration in the disease-associated triplet
repeat sequences in the BS cells suggests that these types of somatic recombinations may
occur in vivo and could generate mutations in specific tissues of persons with BS. In
addition, the generation of these novel repeat sequences in BS cells may provide an
opportunity to isolate genes that contain these types of repeats and that are characterized
by expansions.
89 Chapter 5. Physical Mapping of the Bloom’s Syndrome Locus.
This chapter describes a 2-Mb contiguous map of P1s and YACs that contains the BLM gene. This contig was assembled initially using known genetic
markers and STSs from this region of chromosome 15 and completed with the
use of YAC and P1 end-sequencing and with STSs developed from Alu-PCR markers from FES-positive radiation-reduced hybrids. Although the minimal region containing the BLM gene was determined by genetic methods, this chapter reports the identification of reagents that were useful in the development not only of more polymorphic DNA markers with which to define further the BLM locus, but also of reagents that were used to identify expressed sequences from theregiontowhichBLM has been assigned. The 2-Mb YAC and P1 map was generated to define the BLM region, identify additional polymorhic markers and reduce the region in which the BLM gene was known to reside. The purpose of this work was to aid in the cloning and identification of BLM.
Complementation analyses established that a single locus, designated
BLM, appears to be mutated in BS (Weksberg et al., 1988). McDaniel and
Schultz (1992) localized BLM to chromosome 15 by microcell-mediated chromosome transfer; this localization was confirmed and refined to band q26.1 by demonstrating linkage of BLM to the FES proto-oncogene in 21
consanguineous BS families (German et al., 1994). FES has been mapped
previously by in situ hybridization (Jhanwar et al., 1984). Additional polymorphic
markers have been identified in the region around FES; the use of homozygosity
90 mapping has been extended to include these markers. In addition, the map position of BLM has been refined by using two other genetic approaches: linkage-disequilibrium mapping in Ashkenazi Jews with BS (Ellis et al., 1994) and analysis of cell lines in which intragenic recombination has occurred within BLM
(Ellis et al., 1995a). This second and novel approach relies upon the detection of reduction to homozygosity of all loci distal to BLM. BLM was localized within a
very small region immediately proximal to FES and more recently has been
isolated from this same region (Ellis et al., 1995b).
The construction of a physical map of the BLM-containing region on 15q26
initially was undertaken to facilitate the identification of new genetic markers in
this region of the genome and ultimately to clone the BLM gene. We report the
identification of YACs and P1 clones that form a contiguous representation of a
2-Mb region flanking the BLM gene, as well as the construction of a long-range
restriction map of this region. Radiation-reduced hybrid cell lines were
developed from a somatic cell hybrid that carried chromosome 15 as its only
human chromosome. These hybrid cell lines were used to generate new
markers for the region of the chromosome that included the FES gene and
presumably the tightly linked BLM gene. A set of six polymorphic microsatellites was identified within the YAC and P1 contig to localize precisely the position of
BLM (Ellis et al., 1995b). The clones and sequence-tagged sites (STSs) that compose this contig were a valuable resource from which to isolate the BLM
gene and will be so for other genes that are localized to this region of distal 15q.
91 Isolation of distal 15q DNA Sequences from Radiation Hybrids.
A panel of DNAs was constructed from 173 radiation hybrids to isolate sequences in the BLM region. The donor cell line was the somatic cell hybrid
GS89K1-S, which carries a chromosome 15 as its only human contribution
(Warburton et al., 1990); the recipient was the HPRT-deficient Chinese hamster
(CHO) cell line A23N. The DNA panel was screened by PCR to determine which
hybrids contained the human FES locus, one known to be closely linked to BLM
(German et al., 1994; Ellis et al., 1994). Some of the FES-positive hybrids were
predicted to carry BLM as well as other regions syntenic with either FES or BLM.
Eight human FES-positive hybrids were identified.
Six of the eight hybrids were chosen at random as sources for the
isolation of sequences in the BLM region. Human-specific DNA sequences were
amplified by Alu-PCR, and 209 DNA fragments were subcloned; from these, 69
clones that contained uniquely sized inter-Alu sequences were chosen for further
study. It was possible that the FES-positive hybrids contained DNA sequences from sections of chromosome 15 other than the BLM region. Therefore, these 69 clones were also sublocalized on chromosome 15 by hybridization to DNA from two somatic cell hybrids: GS89K1-S, which contains human chromosome 15, and GM10664, which contains a structurally rearranged chromosome 15, the distal portion of which, 15q25-qter, is missing as a result of an X:15 translocation.
The subcloned DNA fragments were used as hybridization probes for DNA amplified by Alu-PCR. Genomic DNAs from human and hamster were included as positive and negative controls, respectively. Twenty-one probes that
92 hybridized to Alu-PCR-amplified DNA from GS89K1-S but not from GM10664 were identified. DNA sequencing was performed, and STSs were developed from 19 of the 21 clones. These 19 were screened again by PCR; 13 proved by
PCR analysis to be situated within 15q25-15qter.
Construction of a YAC and P1 Contig of the BLM Region.
Linkage analyses have shown that BLM is tightly linked to FES and
D15S127, two loci that are separated by 30 kb, and that BLM lies between
D15S116 and these two loci (German et al., 1994; Ellis et al., 1994, 1995a).
Therefore, STSs and cloned DNA sequences from the region around FES were
used to construct a physical map (Table 2). YAC pools from the CEPH YAC
libraries A and B were screened by PCR with primers that amplify the previously
reported markers D15S127, FES, D15S158, and IP15M9; all these markers have
failed to segregate from each other in genotype analyses of CEPH families.
Primers that amplify loci D15S116 and D15202 were also used for YAC screening, even though these loci are about 1.3 cM proximal to D15S127
(Beckman et al., 1993). In addition, STSs generated from the radiation-reduced hybrids and from genes known to map to band 15q26 (FUR, PEPN, and IGFR) were also used for screening. Last, YACs from the ICRF Reference Library
Database (now at the Max Planck Institute) were screened using DNA probes for
FES and the clones that mapped to 15q25-15qter from the radiation-reduced hybrids, with the exception of repetitive probe 7.38. In general, screening was performed by PCR of YAC and P1 maintained human DNA libraries. Pooled
93 Table 2 Sequence Tagged Sites Used to Construct the Physical Map of the BLM Locus
Primer size Locus Name Forward Sequence Reverse Sequence Source (bp) YAC and P1 end-clone-derived sequences D15S1153 963-3HF TGGAGCTGGGTGATGATGAA GGCAACTGTCACCTAATCTC Y963C3 122 D15S1159 4431-5 TGCCCTAGCAGCCTATGCTG CCTTAGCACACTAGGAGTCAC P1-4431 217 D15S1141 4431-3 AGCAGCTTCCTGGCATGAAAG TCAGTCCCACATTGGCATGAG P1-4431 135 D15S1142 4429-5-3 ATTACAGNTGTGAGCCACTGC CGAGCCACTGTGCCCGGC P1-4429 76 D15S1143 4428-3 GCTGGGATTAGGCATGAGCC TCGGTGAGCTGAGATCACGC P1-44328 153 D15S1144 4430-3 AGGTAGAGCAGCCCTTCAGG ATGGTTCTAGGCACCTGCCC P1-4430 180 D15S1145 1958-3 TCTAAAATTTTCTGATTTTCCCA TGAATTTCAGCAAAACAACTAAT P1-1958 137 D15S1146 4429-3 TTTTTGTGGGTACATAGTAGGTG ATGCATTGGATTGTTTGCAACTC P1-4429 130 D15S1147 4428-3 AGCGGCCCAGGAAAACATGG AGTGCAACGTGCACACTCCC P1-4428 144 D15S1148 1957-3 TTGGGTTGAGAAGTGCAGCT CTGATAATCAACTCCGCCCA P1-1957 141 D15S1149 1959-3-2 CACATGTAAGTCCTGAATCCA TATGACAGAGAAGCATCACCA P1-1959 150 D15S1150 1958-5-2 TGGGTAAGTGGGGGTGTAAA GACTGGTGATTTAGGACTATG P1-1958 168 D15S1151 1957-5 CAGCGTGACCAGCTACCC TACTGGAACTCCTGGACCTT P1-1957 164 D15S1152 1959-5 CAGGGAATTTCTCCAGAAGG CTTCTTATGCTATGAGAGCTC P1-1959 170 D15S1154 932-5HR AAATTCACTGGGTCCTTTCACC CCAGAGAGTAAGAATCAAGTCAT Y932F12 101 D15S1155 884-5HR CCTGGGGTCAGCAACAATGG ACTACAAGAGACGCTGCTCAG Y884D4 81 D15S1156 868-5TR AACTTAATTTGGGGTGGCCG CTACAGGTGTCTGCCACTAT Y868E8 132 D15S1157 852-5TR ACCCACTGCGGATTTCTACTC TCCAAGAGCCAGTACGTGAAG Y852D7 138 D15S1158 963-5TR TTCATACCCTGCAGCATTTC AAGGTAGACTTGAAGGGATG Y963C3 129 D15S1132 Y158/159 ATGCTCACCTCCCAGAAGAC AATCTAGAGGTTAGGAGGAGG y9000B0459 308 D15S1131 Y160/161 TACCAACCAGGACCACTGAG ATGGATAAACCTTCTTCTCC y9000B0459 375 D15S1129 Y162/163 TGCTGACTGATTGCATTCAC CATGTTGCATAAATGGGTGG y9000G01100 282 D15S1130 Y164/165 CTTCTGTTTTCTCTGCAAGG TGTCTGTCCTCACTTTCCAT y9000G01100 263
Other YAC- and P1-derived sequences D15S1120 Y90/91 AGGATGGTTGTATCAGTATTG AACTGAAATTATAAACTCCCTGT Y963C3 ~1300 D15S1133 963-6 TGTCTGTCCTTCCAGCCCCCAT CAATGAAACCAGGCAGAAGC D15S1138 963-3 CTGCAGGTAAGTAGGAATGG CTAAACCTATAGGCTAGGG Y963C3 ~185 D15S1137 Y96/97 GGGTCTCTCCAGCCAATAGT CTAAGAGTATCCAAGCCTCA Y963C3 ~1300
Radiation-reduced hybrid-derived inter-Alu sequences D15S1124 7.38 GGGAAAGCCAAGTGAATTAC CCCACACAAAGAGTGGTTTC NE7 286 D15S1125 7.73 CATCCTATGACTGTTATGAC CACTGTGCTTGGCTCTAAGG NE7 385 D15S1126 7.74 ATTCCTATACCAGGGCTGGG CCTTTAGGGGAAATCAAACC NE7 365 D15S1127 7.91 GTGGCTCCACAGGGAATC ATTGGAAGATAGAGGAAGGG NE7 145 D15S1115 18.20 TGGAGGCGTAAGGAGAATAAAGA GCTGGCAGGAGGCAAAGTAG NE18 ~1100 D15S1128 18.37 GCCTCCCAATACTCAGCATTTC TGGGAGTGGAACTGCATACAGG NE18 147 D15S1119 18.38 TGCCCCCGACCCTAAGTGAAT CAGATAGATGGCAGCGTAGTAAAT NE18 ~1000 D15S1134 18.39 GCTCACAAGGTCATACAACG TGAAATGTCCAGAAAAG(G/T)CG NE18 168 D15S1106 18.44 ATTACCAAACCTGATACCAACA AAACATTACTAAGAGCATTCATC NE18 ~550 D15S1123 67.26 CTCACTCTCCCAGCTCATTG CCAGCCACAGTACCTTGCCT NE67 103 D15S1122 69.1 CTGTGTACCCACTATGTGCT CACTCTGAACAGTGGCAGCC NE69 105 D15S1135 69.4 ATCCAGTCTGGTCTCTGGTG GGTACCCTTCCAATCAGGGA NE69 467 D15S1136 69.16 GCGACTAGGATGTGCATACT CCCTCGGACCATCCACCCTG NE69 109
Previously published sequences fUR FUR-5/3 GGCGCGGCTCACCCTGTC TAACCATCTGCGGAGTAGTCAT P1-1957 ~550 PEPN GCTTCTATATTTCCAAGTCCC CGACATTGACAAGACTGAGC 400 IGFR GAGACAGCTTCTCTGCAGTA TCCGGACACGAGGAATCAGC
94 DNAs were used as template material and primers were designed to either previously published sequences or to sequences identified from fragments of human DNA from radiation-reduced hybrids that mapped to the region by hybridization.
Initially, two YACs that contained FES were identified from the original
CEPH A YAC library. However, both YACs (179B12 and 93A4) were found to be chimeric by FISH analysis and end-sequencing of the genomic inserts. PCR screening of both the CEPH A and B libraries (Research Genetics, Inc.) was completed for all the loci in Figure 7. End sequencing of these YACs and subsequent PCR with primers based upon these sequences were used to reveal overlaps and to extend the contig. All YACs were also assessed for chimerism by FISH analysis. YACs positive for the loci IP15M9 and D15S158 overlapped
with a 1-Mb nonchimeric YAC, designated 963C3, which in turn overlapped with
YACs positive for D15S1108 and D15S996. Although YAC 963C3 was not chimeric, it did not contain sequences from FES or D15S127. Most likely, this is due to an internal deletion. Further screening of the YAC libraries with end clone primers provided several other YACs that were contained completely within YAC
963C3 or, although redundant to sections of YAC 963C3, also extended the contig distally from the FES locus. A contig of YACs that encompassed a region of approximately 2-Mb of genomic DNA including the BLM gene, was assembled
(Figure 7).
The internal deletion in YAC 963C3 was closed by the isolation of three unique P1 clones from a P1 library (Genome Systems, Inc.) that was PCR-
95 50 Kb
D15S1131 D15S996 D15S1153 D15S1159 D15S1108D15S1141 D15S1142D15S1129D15S1143D15S1144 D15S1145D15S1146D15S1147D15S1148D15S1149D15S127D15S1150FURFESD15S1151D15S1152 D15S158D15S1120 DS15S1139D15S1133D15S1154 D15S1155D15S1138 D15S1156D15S1157IP15M9 D15S1158
93A4 79B12
963C3
Y900B0459 852D7
Y900G01100 8484E8 1957 932F12
Y900D01110 1958 884D4 878F7
Y900H0342 1959
4431/4432 4428 4144
4429
5914
5913
4430
Figure 7. The physical map of the BLM region. The contig of approximeately 2.0 Mb is bounded by D15S113 proximally and D15S1158 distally. Solid vertical lines at the end of YACs/P1s represent the identification of an STS from end sequence that falls onto the physical map. Solid vertical lines within YACs/P1s represent inclusion of that STS within the clone. YACs 963C3 and 852D7 have internal deletions, as determined by PCR and restriction mapping; this is indicated by the downward angled lines within these clones.
96 screened with primers for locus D15S127; these P1 clones were designated
1957, 1958, and 1959. P1s 1957 and 1959 were also positive for FES.The ends of the genomic insert for each of these P1s were sequenced, and primers that amplified each end sequence were designed. It was determined that the deletion in YAC 963C3 was contained within the three P1s and that the contig immediately surrounding the FES/D15S127 loci was complete (Figure 8). To refine the map further, walking via P1 end-sequencing and PCR screening continued in the proximal direction, using the proximal end of 1958 as the starting point. This screen provided three additional P1s 4428, 4429, and 4430.
Additional P1 clones in the region flanking FES and D15S127 were identified with primers to D15S1108: two identical P1 clones, 4431 and 4432, were identified.
Further P1 screens with primers from the distal end of 4431 were negative, which raises the possibility that a gap exists in the genomic coverage of the P1 library.
A final attempt to link the P1s from D15S1108 to those from the FES/D15S127
P1 contig was made by hybridization screening the P1 library with a fragment from a candidate cDNA located near the proximal end of P1 4429. This screen yielded two P1 clones, but both extended distally and did not link P1 4431 with
P1 4429 or with distal P1s 4428, 4429, or 4430.
A YAC designated 883C4 was identified with sequences from D15S116 and D15S202, two loci proximal to FES, although it failed to overlap with any of the YACs from the 2-Mb contig. Additionally, YACs identified by PCR of loci
D15S1119, IGFR, D15S1115, D15S1123, and D15S1122 failedtooverlapwith
97 Locus FES FUR IP15M9 D15S996 D15S127 D15S158 Size (Kb) YAC/P1 Source D15S1131 D15S1153 D15S1159 D15S1108 D15S1141 D15S1142 D15S1129 D15S1143 D15S1144 D15S1145 D15S1146 D15S1147 D15S1148 D15S1149 D15S1150 D15S1151 D15S1152 D15S1120 D15S1139 D15S1133 D15S1154 D15S1155 D15S1138 D15S1156 D15S1157 D15S1158
ND Y900H0459 ICRF 600 Y900G01100 ICRF 200 Y900D01110 ICRF 450 Y900B0342 ICRF ND P4431/4432 P1 90 P4429 P1 ND P6913 P1 ND P6914 P1 80 P4428 P1 70 P4430 P1 95 P1958 P1 1070 Y963C3 CEPH(B) 75 P1957 P1 ND Y93A4 CEPH(A) ND Y179B12 CEPH(A) 80 P1959 P1 ND P41449 P1 600 Y852D7 CEPH(B) 600 Y932F12 CEPH(B) 300 Y884D4 CEPH(B) 590 Y868E8 CEPH(B) 1720 Y878F7 CEPH(B)
Figure 8. A cross reference of the YACs and P1s versus loci identified by sequence-tagged sites (STSs) in this study. The shaded blocks indicate that an STS from a particular locus is located on that vector insert. Blocks that are not shaded represent loci not included within a particular clone. The first column shows size (in Kb) of each YAC and P1; ND indicates that the size is unknown. The second column shows the name of the YAC or P1. The third column shows the library from which each clone was isolated.
98 YACs in the 2-Mb contig, but were found to map outside the region defined proximally by locus D15S116 and distally by locus D15S127.
Fluorescence in Situ Hybridization Analysis of the YACs from the BLM
Region.
YAC DNAs, some isolated from pulsed-field gels, were labeled and hybridized. These results show that YACs 963C3, 852D7, y900D1170, y900F1044, and y900D0967 hybridize only to 15q25-ter (Figure 9). YACs
179B12, 932F12, 884D4, and 939A10 were chimeric, composed of DNA segments from 15q25-ter and chromosomes 1p, 5q, 5qter, and 5q, respectively.
YAC 93A4 identified sequences on chromosome 21q only, although it was ascertained originally by PCR with primers to FES.
Isolation of Polymorphic Markers from the BLM Region.
Previously unidentified DNA polymorphisms were developed from selected YACs to help define the smallest region that contained BLM. Yeast
DNAs containing YACs 963C3, 883C4, y900D1170, or 919A1 were subcloned, plated, and colony-lifted. Clones were hybridized to a labeled poly(CA)-poly(GT) probe (Table 3). Forty-seven clones were purified; 26 of these were unique.
DNA sequence analysis showed that eight were derived from the yeast genome and that four contained polymorphic sequences that had been identified previously. Sequences from some of the remaining clones were used to design
99 Figure 9. Fluorescence in situ hybridization of YAC 963C3 to human metaphase chromosomes. YAC 963C3 spans the minimal region that contains BLM and hybridizes to 15q25-qter.
100 Table 3 Polymorphic Microsatellites Used or Isolated in Positional Cloning of BLM
Locus Primer Name Forward Sequence Reverse Sequence Source
D15S1140 CA919-12D/L CAAGGATGTTCACTGTAGCTG GTTAACAGTTGGGCTGTATGC Y919A1 D15S116 AFM078zf7a/m AGCTTCCAACTNCGCCCTCC AGGGGTGTTACATCGCGGGT Beckmann et al., 1993 D15S202 AFM282wg5a/m AACTGGGTGGCACAGTGAG CAGGACCTTTGCACAGGC Gyapay et al., 1994 D15S996a AFM6027wd5F GAAGGATGGTTTGAGCCC ACTTAGGAATAATCATTACTGGCAT Weissenbach (per. com., 1995) D15S1108a Y66/153 CTTACTATATGCTAGGTAGG GTGTGAAGATATAGTATGAGG Y963C3 D15S127a AFMxe11a/m CCAACCACACTGGGAA AACAGTTGCCCACGGT Beckmann et al., 1993 FESa FES1.1/1.2 CTCCAGCCTGGCGAAAGAAT GGAAGATGGAGTGGCTGTTA Polymeropoulos et al., 1990 D15S158a AFM234vf12a/m CAGGAGACCTCCAAACACA TTTCAGCCAAGAAGCACG Gyapay et al., 1994 D15S1139b 963-2D/L CAATCACACGTGCATGCATCC GGTAGGACAGAAGCCAGATCAT Y963C3 IP15M9a IP15M9F/R CACCTCTGCTTGAAGGAAGG AGCTCTGGATGCTTCTGTAC Beckmann et al., 1993 D15S1116 16-12D/L CTCCTGTATGCACATGTATACTACC CCACTACTCCAATCTCTTGTACCTG y900D1170
a Polymorphism located within the 2-Mb contig. b The same locus is identified by the primer AFMa113wh5 (J. Weissenbach, Paris, personal communication, 1995).
101 PCR primers that flanked the repeat sequences to identify polymorphisms in the
DNA of ten parents of BS patients. Four of the loci identified by these sequences were nonpolymorphic, but three proved to be polymorphic. The sequences from the remaining seven clones were not characterized further. In addition, the polymorphic CA-repeat D15S1108 was identified in sequences contained in a cosmid isolated with probe 18.44.
The polymorphic loci reported here, supplemented by published polymorphic loci, were genotyped in the collection of consanguineous BS families; recombinant chromosomes were identified. By this analysis, the sequences contained in the YACs 919A1 and 883C4 were proximal to BLM, whereas sequences contained in the YAC y900D1170 were distal. Seven polymorphic loci are present in the 2-Mb contig but only six were selected for genotyping in the BS families, since the seventh, D15S1139, was known to map distally to D15S158 and therefore to BLM,withD15S996 being proximal and
IP15M9 being distal. This localization has been confirmed by linkage- disequilibrium mapping and by the use of cell lines in which intragenic recombination has occurred (Ellis et al., 1995a). Therefore, the 2-Mb YAC contig that includes both D15S996 and IP15M9 also contains BLM.
Long-Range Restriction Mapping of the BLM Region by Pulsed-Field Gel
Electrophoresis.
Long-range restriction mapping (Figure 10) was performed by analysis of
Southern blots of gels generated by PFGE. DNAs from normal controls were
102 D15S1131 D15S996 D15S1153 D15S1159 D15S1108D15S1141 D15S1142D15S1129D15S1143D15S1144 D15S1145D15S1146D15S1147D15S1148D15S1149D15S127D15S1150FURFESD15S1151D15S1152 D15S158D15S1120 DS15S1139D15S1133D15S1154 D15S1155D15S1138 D15S1156D15S1157IP15M9 D15S1158
963C3 Y900B0459 852D7 8484E8 Y900G01100 1957 932F12 Y900D01110 1958 884D4 878F7 Y900H0342 1959 4431/4432 4428 4144 4429 5914 5913 4430 ED22 ED1-4 18.44 ED24 FES 59-6 AP9 963-2 963-3
175 125 125 25 125 50 25 2525 25 50 100 150 75 150
B B Na B B B Na B Ne B M N B Ne M S M Ne Ne M S Na Ne M No Ne S Mr Ne S Ne Na No S S Ne Na No No Nr No Ne S S S
Figure 10. Long-range restriction map of the BLM region and representation of YACs and P1s within it. The map is composed of restriction enzyme sites generated by hybridizations with nine DNA probes (solid boxes) and seven restriction enzymes (B, BssHII; M, MluI; Na, NarI; Ne, NaeI; No, NotI; Nr, NruI; S, SfiI). The genomic sequences are depicted by the thick horizontal line, on which restriction sites are marked (vertical lines). The distances between restriction enzyme sites are shown in kilobases. Two nonoverlapping restriction maps have been joined together at the thick vertical line, as the exact length of sequences between the two maps is unknown. Genomic sequences in both mapped regions are contained in the 1-Mb YAC, 963C3; therefore, the length between the two maps is very short (<200 kb). Horizontal lines of intermediate thickness represent YACs and P1s. The array of ordered STSs (vertical intersecting lines on the upper horizontal line) is shown above the YACs and P1s.
103 digested with BssHII, NruI, MluI, SfiI, NotI, NaeI, NarI, and various combinations of these enzymes. Blots were hybridized serially to a panel of probes derived from the 2-Mb YAC contig as well as to probes outside this region (Table 4). A
575-kb NruI fragment was detected by the DNA probes ED1-4, 18.44, ED24,
FES, and 59-6. This region is contained in YAC 963C3, in the YACs identified using 18.44, and in three P1 clones (1957, 1958, and 1959) representing the proximal end of the 2-Mb contig. The probes AP9, 963-2, and 963-3 detected an
800-kb BssHII fragment and an 800-kb NotI fragment. This region is also contained in 963C3 and the YACs identified with primers that by PCR amplify
IP15M9 and D15S158, loci positioned in the distal end of the 2-Mb contig.
Despite extensive efforts, the maps of two regions have not yet been joined experimentally; thus, in Figure 10, the two regions have been linked arbitrarily at a potentially shared MluI site. The two regions must be close together because the 963C3 YAC is only 1-Mb yet contains most of the sequences identified in the
575-kb NruI and the 800-kb BssHII fragments. The MluI site puts the two regions as close together as possible.
The sequences defined by the probes 18.20, 18.38, and 69.1 all hybridize to the same 1400-kb BssHII fragment. This result is consistent with an Alu-PCR analysis that indicated that the YACs isolated with 18.20 and 18.38 overlap. The
YAC isolated with 69.1 has yet to be connected with the other YACs. Long- range restriction mapping has been performed with several other DNA probes derived from the radiation hybrids (18.37, 18.39, 67.26, and 69.16) as well as
104 Table 4 Restriction Fragment in Kilobases Identified by Hybridization of Probes from the 2-Mb YAC Contig
Probe BssHII MluI NruI SfiI NarI NotI NaeI B/M B/Nr B/S B/Na B/No B/Ne M/Nr M/S (B) (M) (Nr) (S) (Na) (No) (Ne)
ED22 150 550 325 50 300 325 150 150 50 150 150 50 ED1-4 250 425 575 25 275 400 400 250 250 25 125 250 250 250 25 18.44 250 425 575 125 275 400 400 250 250 125 125 250 250 250 125 ED24 150 200 575 125 50 400 400 150 150 125 150 150 150 200 125 FES 50 200 575 25 25 75 50 50 50 25 50 50 50 200 25 59-6 50 125 575 75 LM 25 25 50 50 50 50 25 25 125 75 AP9 800 325 LM 150 LM 800 225 225 800 150 800 800 250 325 150 963-2 800 325 LM 225 LM 800 425 225 800 225 800 800 425 325 75 963-3 800 575 LM 350 LM 800 425 575 800 350 800 800 425 575 350
LM indicates tht the DNA fragment migrated at the position of limiting mobility on the gel.
105 with probes from the D15S116-positive YAC 883C4, but no common fragments were identified.
Conclusions.
The gene responsible for BS, BLM, is known to map to human
chromosome band 15q26.1 (German et al., 1994). Linkage analysis and
homozygosity mapping in consanguineous BS families were used to narrow the
position of BLM to all small regions of chromosome 15 around FES.Linkage
disequilibrium between specific alleles at FES and D15S127 and the BLM
mutation in the Ashkenazi Jewish population also was invaluable in initially
refining the limits of BLM locus (Ellis et al., 1994). Figure 11 summarizes the
steps necessary to generate the physical map at the BLM locus.
106 Complementation Analysis: single locus mutation;Weksberget al., 1988.
Microcell-Mediated Chromosome Transfer: localization to chromosome 15;cDaniel M and Schultz, 1992.
Linkage Analysis: localization to 15q26.1; Germanet al., 1994.
Linkage-disequilibrium Mapping: refinement of localization to 15q26.1; Elliset al., 1994.
Radiation-reduced Hybrid Cell Lines: identification of polymorphic markers linked to FES; Elliset al., 1995.
Construction of a 2-Mb physical map about the FES identificationl of YAC and P1contigsandSTSs; Straughenet al., 1996.
Fluorescent In Situ YAC colony Long-range Restriction Hybridization: hybridization: Mapping verification that contig identification of by Southern Blot and elements additional PFGE: are localize to 15q26.1 polymorphic markers; refinement of the Straughenet al., 1996. Straughenet al., 1996. distances between contigSTSs; Straughenet al., 1996.
Intragenic Recombination of BS cells: limiting of the region that contains BLM to 250 Kb; Elliset al., 1995.
Figure 11. Summary of Steps Taken to Construct the Physical Map at the BLM Locus. The text within the bold boxes indicats techniques and results presented in this work. The underlined text is work published by our collaborators that assisted in the mapping and localization of BLM.
107 Chapter 6. The Bloom’s Syndrome Gene Product is homologous to RecQ
Helicases.
The hypermutable phenotype of BS cells is due, at least in part, to a great increase in somatic recombination. Although cells from all persons with BS exhibit the diagnostic high SCE rate, in some persons a minor population of low
SCE lymphocytes exists in the blood (Ellis et al., 1995). LCLs with low SCE rates, consistent with unaffected individuals (approximately 5-10
SCEs/metaphase), can be developed from these low SCE lymphocytes. In multiple low SCE LCLs examined from 11 persons with BS, polymorphic loci distal to BLM on 15q had become homozygous in the low SCE LCLs (Figure 12).
These observations supported the hypothesis that low SCE lymphocytes arose through recombination within BLM in persons with BS who were compound heterozygotes or who had inherited unique paternal and maternal mutant BLM alleles. A recombination event in a precursor stem cell in these compound heterozygotes gave rise to a cell whose progeny had a functionally wild-type gene and phenotypically low SCE rate (Ellis et al., 1995).
The low SCE LCLs in which reduction to homozygosity had occurred were used for localizing BLM by an approach referred to as somatic crossover point
(SCP) mapping. The most precise map position of BLM was determined by comparing the genotypes of the recombinant low SCE LCLs from the five persons mentioned above with their constitutional genotypes at loci in the region
108 Figure 12. Autoradiographic Evidence to Support the 250-kb Region of SCP Mapping. The four persons of five from whom low SCE LCLs had been established that were informative at D15S1108 or D15S127 are shown. To determine both the constitutional and the recombinant cell line genotypes, we carried out PCRs using DNA samples prepared from high SCE cells (Ph) and low SCE LCLs (Pl) of persons with BS as well as samples from their fathers (PF) and their mothers (PM). These persons are identified by their Bloom’s Syndrome Registry designations (German and Passarge, 1989). Arrows point to DNA fragments amplified from the heterozygous alleles of the constitutional genotypes, pat (for paternal) and mat (for maternal). Asterisks mark alleles in the low SCE LCLs that are lost through somatic crossing over. Lines mark DNA fragments amplified from alleles of the parents but that were not transmitted to the offspring with BS. From one of the four persons with BS, 11 different clonal LCLs were examined; three of the 11 had undergone reduction to homozygosity at loci distal to BLM, as explained elsewhere (Ellis et al., 1995). Autoradiographic patterns are shown from two of the 11 low SCE LCLs from (IaTh), one representative of cell lines in which allele losses were detected (Pl sample on right) and another of cell lines in which they were not (Pl sample on left).
109 around BLM. The strategy was to identify the most proximal polymorphic locus
possible that was constitutionally heterozygous and that had been reduced to
homozygosity in the low SCE LCLs, and to identify the most distal polymorphic
locus possible that had remained constitutionally heterozygous. BLM would map
to the shortest interval defined by the reduced (distal) and the unreduced
(proximal) heterozygous markers. The power of the approach was limited only
by the density of polymorphic loci available in the immediate vicinity of BLM.In this chapter we describe the SCP mapping, isolation, and identification of BLM.
Localization of BLM to a 250-kb Interval.
BLM previously was localized by SCP mapping to a 1.3 cM interval bounded proximally by D15S116 and distally by four tightly linked loci (D15S127,
FES, D15S158, and IP15M9) (Ellis et al., 1995). The four loci are present in a 1-
2 cM interval on chromosome 15 (Bechmann et al., 1993; Gyapay et al., 1994).
The order of these four loci was determined by PCR analysis of clones in a 2-Mb
YAC and P1 contig that encompasses BLM (Straughen et al., 1996). The four loci were oriented with respect to the telomere by finding a recombinant chromosome in a BS family in which crossing over had occurred between BLM and IP15M9,placingIP15M9 on the distal end of the contig. Because D15S127 was the most proximal locus that was reduced to homozygosity in low SCE
LCLs, polymorphic loci in the region proximal to it were sought. There, a polymorphic locus, D15S1108, was identified that remained constitutionally heterozygous in the recombinant low SCE LCLs, in contrast with locus D15S127,
110 which had become homozygous (Figure 13). This shift from heterozygosity to homozygosity of markers indicated that BLM is situated in the 250-kb region between D15S1108 and D15S127.
Two genes, FES and FUR, mapped distal to D15S127 in this region of chromosome 15. SCP mapping thereby eliminated them as candidate genes for
BLM. Consistent with this conclusion, an earlier mutation search in six BS LCLs had failed to uncover mutations in FUR.
Isolation of a Candidate Gene for BLM.
cDNAs were isolated from the 250-kb region between D15S1108 and
D15S127 by direct cDNA selection using cDNA libraries from cultured fibroblasts and the T-lymphocyte tumor cell line Jurkat. Libraries from these cell lines were chosen because fibroblasts and T-lymphocytes from persons with BS exhibit the high SCE phenotype, suggesting that BLM wouldbeexpressedinthesecell types. Direct selection experiments using cosmid c905 identified an 847 bp cDNA designated 905-28 after two rounds of direct selection. It was found in fewerthan1in106 clones screened in the fibroblast library but was present in 6 of 28 selected cDNA clones, a 250,000-fold enrichment. The six cDNAs represented by 905-28 were the only selected cDNAs that mapped to the BLM region by Southern blot analysis and that identified nonrepetitive sequences in the human genome. The 905-28 cDNA identified single-copy sequences that were situated approximately 55-kb proximal to FUR (Figure 13).
111
Heterozygosity Homozygosity
D15S996 D15S1108 D15S127 D15S158
250250 375
BLMFUR FES
12550 125 50
Y963C3 P4429 Y852D7 P4428 P1959 P4430 P1957 P1958 c905 c701 c812 c1005 c702 c1013
Figure 13. Somatic Cross-Point (SCP) Mapping of BLM. The upper horizontal line is the genetic map of the BLM region on chromosome 15q, which indicates the order and distances (kilobases) between the polymorphic microsatellite loci that were estimated by long-range restriction mapping. The distance between D15S127 and FES (not indicated) was determined to be 30-kb by restriction enzyme mapping of a cosmid contig. Vertical lines indicate the position of the marker loci; the circle represents the centromere. The interval between loci D15S1108 and D15S127 is expanded below the map. Vertical lines intersecting mark the unmethylated CpG-rich regions identified by long-range restriction mapping, and arrows indicate the direction of transcription of three genes in the region. Certain YACs (Y), P1s (P), and cosmids (c) from the contig are depicted by horizontal lines underneath the map. Dashes of the YAC lines indicate internal deletions. At the top, the horizontal crosshatched bars indicate regions proximal to BLM that remained heterozygous in the low SCE LCLs and regions distal to BLM that had become homozygous. The solid box represents the minimal region to which BLM was thus assigned by SCP mapping.
112 The 905-28 cDNA then was used to screen a HeLa cDNA library; 28 cDNAs were isolated, representing at least 15 distinct classes of overlapping clones. Each of these classes had the same sequence as the 905-28 cDNA at the 3’ends but a different length of 5’ sequence. In the longest cDNA isolated, clone H1, a long reading frame was found that was open to the 5’ end.
Additional sequences upstream of the start of the H1 cDNA were identified by a
PCR-cloning method. Clones extending 5’ of the H1 cDNA were isolated from the HeLa library, permitting the identification of 4437 bp of sequence, which we refer to as the H1-5’ sequence (GenBank Accession U39817).
Translating the sequence from the first in-frame ATG, 75 bp from the 5’ end of the clone, the H1-5’ sequence encodes a 1417 amino acid peptide with a predicted molecular mass of 159 kDa. No in-frame stop codons were present between this ATG and the 3’ end of the H1 sequences. An extensive cDNA analysis was carried out to map the 5’ end of the candidate gene. We screened
8x106 LCL cDNA clones by hybridization with a 5’ probe, isolated 11 clones, and sequenced their 5’ ends. In addition, 12 fibroblast clones prepared by a 5’ rapid amplification of cDNA ends (RACE) technique were sequenced. Both analyses indicate that the H1-5’ sequence is full length.
The predicted peptide encoded in the H1-5’ sequence was used to carry out a BLASTP search of amino acid sequence databases. The searches identified significant homologies to three known peptides in the RecQ subfamily of DExH box-containing helicases (Figure 14). The amino acid homologies were concentrated in the region (residues 649 – 1041) containing the seven conserved
113 1 64 104 141 NH COOH Helicase Domain
DEAH
I Ia II III IV V VI
Figure 14. Helicase Motifs in BLM. A. The linear structure of the BLM protein is shown with an expanded view of the center of the protein that constitutes the seven helicase motifs. The helicase motifs are drawn to indicate approximate relative size and position.
114 helicase domains of the human RECQL (44% homology), Saccharomyces
cerevisiae SGS1 (43% homology), and Escherichia coli recQ (42% homology) genes. Since the identification of BLM, numerous homologues of BLM have been identified both within human and across species. Each homologue is characterized by a helicase motifs and a high degree of homology within the helicase domain (Figures 15 and 16). Thus, the product of the candidate gene contains motifs homologous to DNA helicases, suggesting that the protein is an enzyme engaged in DNA manipulation.
The seven helicase domains identified by their homology to RecQ constitute only the middle third of the predicted peptide. Between residues 588 and 661, amino acid identities were discovered with three short motifs present in a broad phylogenetic spectrum of RNA polymerase II largest subunits. The function of these motifs is unknown. No other significant homologies were identified to amino acid sequences in databases.
The amino acid composition of the nonhelicase regions of the predicted peptide is unusual. The amino-terminal 648 residues of the peptide are rich in acidic (17%), basic (12%), and polar (34%) amino acids; 13% of the residues are serines. Similarly, the carboxy-terminal 376 residues also are rich in acidic
(11%), basic (16%), and polar (30%) amino acids; and, again, 14% of the residues are serines. The function of these highly charged regions is unknown.
RNA Expression of the Candidate Gene in Cultured Cells.
115 NH2 COOH 1 649 1041 1417 Hs
1 657 1049 1416 Mm 96% 1 601 993 1364 Xl 86% 1 379 771 1142 Gg 84% 1 719 1110 1487 Dm 55%
Figure 15. RecQ helicases from other organisms with homology to BLM. Homologues to human BLM (Hs) have been identified in several species, including mouse (Mm), African green frog (Xl), chicken, (Gg), and fruit fly (Dm). Linear representations of these proteins are shown, aligned according to their relatively conserved helicase regions (green shading). Each of these proteins contains the seven helicase motifs (gray shading) within the helicase region and the percent identity to the human BLM is indicated. Within the helicase region mouse BLM is the most homologous to human BLM with 96% identity, followed by frog, chicken, and fruit fly, with 86%, 84%, and 55% homology, respectively.
116 NH2 COOH 1 649 1041 1417 BLM
1 657 1049 1416 WRN 34% 17 458 659 RecQL
144% 410 RecQ4
146231% 866 1208 RecQ5 40%
Figure 16. Human RecQ helicases with homology to BLM. The five recQ homologues that have been identified in human are shown aligned by the seven helicase motifs (gray shading). The percent of identity within the helicase regions (green shading) of these proteins to BLM is noted. Each protein contains the seven motifs within helicase regions of nearly identical amino acid lengths. Otherwise, there is little similarity between the human homologues of recQ. RecQL is the closest to BLM with 44% amino acid identity within the helicase region, followed by recQ5, WRN, and recQ4 with 40%, 34%, and 31%, respectively. Deficiencies in three of these proteins lead to known disease states in the human: BLM – Bloom’s syndrome, WRN – Werner’s syndrome, and recQ4 – Rothmund-Thompson syndrome.
117 Northern blot analysis was used to determine the size of the full-length
transcript from the candidate gene. The H1 cDNA was hybridized to total RNAs
prepared from HeLa cells, normal diploid cultured fibroblasts, and non-BS LCLs.
Two cDNA bands at approximately 4.5-kb were visualized on the autoradiogram
(Figure 17). This size is consistent with the length of the longest cDNAs
sequenced.
In addition, Northern blot analysis was performed using total RNAs
prepared from LCLs from seven unrelated persons with BS (Figure 17). In three
BS LCLs, the quantity of RNAs identified by hybridization to the H1 cDNA was
decreased in comparison to that of the control LCLs. In four BS LCLs, the
pattern of the two RNA bands was abnormal compared to that in the normal
cells; in one BS LCL, the lower band was absent; in another, the upper band is
absent; and in two others, the intensity of the lower of the two bands was
increased and upper decreased. The RNA loading was equal in all the lanes, as
evidenced by hybridization with a probe for glyceraldehydes-3-phosphate
dehydrogenase (G3PD) gene. These observations suggest that RNAs identified by the H1 cDNA might be destabilized in BS LCLs as result of mutations in the candidate gene. The derivation of the two RNA bands from the gene is unknown.
Mutation in the Candidate Gene in Persons with BS.
To determine whether the candidate gene is BLM, we prepared RNAs from LCLs from 13 unrelated persons with BS and from cell lines from four
118 Figure 17. Northern Blot Analysis of the H1-5’ Sequences Expressed in Cultured Cells. A. RNA preparations were analyzed from HG2162, a normal LCL; HG2635, a normal diploid fibroblast cell line; and HeLa cells. B. RNA preparations were analyzed from HG1943 and HG2162, which are normal LCLs, and from HG2703, HG1584, HG1987, HG1972, HG2231, HG1626, and HG2820, which are BS LCLs. Total RNA (30 g)from each cell line was loaded in each lane. Labeled probes (the H1 cDNA shown in the upper panels and a cDNA for G3PD showninthelowerpanels) were hybridized to membranes of he blotted gels and, after washing, the membranes were exposed for 1-3 days (H1 cDNA) or 15 minutes (G3PD cDNA). On a 7 day exposure, faint bands resembling the hybridization pattern in normal cells were detected at the 4.5-kb position in HG2703, HG1584, and HG2820.
119 unaffected controls. These RNAs were used to generate cDNAs for mutational analysis of the expressed sequences of the candidate gene. Sequences in these
13 BS and four control non-BS cDNAs were amplified in approximately 200-bp segments using PCR primers designed from the open reading frame in the H1-5’ sequence (Table 5). The amplified segments were analyzed by single strand conformation polymorphism (SSCP) analysis using two conditions for electrophoresis. Novel SSCP conformers (Figure 18) were identified from BS
DNA, and the genetic changes underlying them were ascertained by sequencing.
We identified seven unique mutations in ten persons with BS (Table 6 and
Figure 19), as well as four polymorphic base pairs. Of the mutations, four introduced premature nonsense codons into the coding sequence and three introduced amino acid substitutions; one of the four chain-terminating mutation arose by a 3-bp deletion, one by a nucleotide substitution, one by a 1-bp insertion that caused a frameshift, and one by a 6-bp deletion accompanied by a
7-bp insertion that also caused a frameshift. This last mutation was detected in all four persons with Ashkenazi Jewish ancestry. The potential products encoded in these four mutant alleles are 185, 271, 515, and 739 amino acids in length, respectively, and none contains a complete set of the seven helicase domains. Of these mutant alleles, three were detected in the homozygous state.
These observations are evidence that the H1-sequence is mutated in persons with BS, thereby proving that this candidate gene is BLM.
Finally, two putative missense mutations were identified in two persons with BS that introduced amino acid substitutions at residues conserved in RecQ
120 Table 5. Mutations Identified in the Candidate Gene in Persons with BS Mutation Nucleotide Amino Acid Predicted Patient1 Ancestry Cell Line Zygosity Position2 Mutation Mutation Length
97(AsOk) Japanese HG1926 Homo 631 3-bp deletion Nonsense/S→stop 185 112(NaSch) American/European HG2510 Hetero 888 A→T Nonsense/K→stop 271 93(YoYa) Japanese HG1626 Homo 1610 1-bp insertion Frameshift/→stop 515 139(ViKre) American/European HG2231 Hetero 2089 A→G Missense/Q→R 1417 15(MaRo) Ashkenazi Jewish HG1514 Homo 2281 6-bp deletion/ Frameshift/→stop 739 7-bp insertion3 42(RaFr) Ashkenazi Jewish HG2522 Homo 2281 6-bp deletion/ Frameshift/→stop 739 7-bp insertion3 107(MyAsa) Ashkenazi Jewish HG2654 Homo 2281 6-bp deletion/ Frameshift/→stop 739 7-bp insertion3 Nr2(CrSe) Ashkenazi Jewish HG2727 Homo 2281 6-bp deletion/ Frameshift/→stop 739 7-bp insertion3 92(VaBi) Italian HG1584 Homo 2596 T→C Missense/I→T 1417 113(DaDem) Italian HG1624 Homo 3238 G→C Missense/C→S 1417
1 Bloom’s Syndrome Registery designations. 2 Nucleotide positions based on cDNA sequence provided to GenBank. 3 Deletion ATCTGA and insertion of TAGATTC causes the insertion of the novel codons LDSRstop after amino acid 736.
121 Figure 18. Novel SSCP Conformers Detected in cDNA Samples Isolated from BS LCLs. Each panel includes five lanes of cDNAs from five unrelated persons with BS amplified with oligonucleotides designed from a unique regin of the Candidate gene. The novel conformers in which mutations were detected are shown in the center lanes of each panel: BS LCL HG1514 from 15(MaRo) in (a), BS LCL HG1624 from 113(DaDem) in (b), BS LCL HG1926 from 97(AsOk) in (c), BS LCL HG2231 from 139(ViKre) in (d), and BS LCL HG1626 from 93(YoYa) in (e). Not shown are the novel conformers detected in 92(VaBi) and 112(NaSch).
122 Table 6 Pairs of Primer Sequences Used for SSCP Analysis of BLM Product Name Forward Sequence1 Reverse Sequence1 Length (bp)
C1-B GGATCCTGGTTCCGTCCGC GAGGTTCACTGAAGGAAAAGTC 269 C1-A CAACTAGAACGTCACTCAGCC GAAGTCCTTGACCCTTTGCTG 233 C1-1 GACTTTTCCTTCAGTGAACCTC GGGATTTCTTTACAGTTGGTGTG 186 C1-2 CCAGATTTCTTGCAGACTCCG CTCTTACAAAGTGACTTTGGGG 213 C1-3 CTTTAAGTACCATCAATGATTGGG CCTCAGTCAAATCTATTTGCTCG 227 C1-4 GAGTAAGCACTGCTCAGAAATC GCTTAACCATTCTGAGTCATCC 160 C1-5 CGAGCAAATAGATTTGACTGAGG CAATACATGGAACTTTCTCAGTTG 223 C1-6 GAAGATGCTCAGGAAAGTGAC CGTACTAAGGCATTTTGAAGAGG 215 C1-7 CAACTGAGAAAGTTCCATGTATTG CACAGTCTGTGCTGGTTTCTG 239 C1-9 CTATTCCTGATGATAAACTGAAAC CCTTCATAGAATTCCCTGTAGG 200 C1-10 GTGGAGATACAGGCCTGATTC GTGTTTCAGCCCAGTTGCTAC 244 C1-11 CAGGATTCTCTGCCACCAGG GCAGTATGTTTATTCTGATCTTTC 183 C1-12 CAGGAAATGTTCTCACAAGCAC CCTTGATGGGTTGATAGGCAG 203 C1-13 CAGCCAGCAAATCTTCCACAG CGCTCATGTTTCAGATTTCTGG 204 C1-14 GAATTATACTGACAAGTCAGCAC GATCTACGATAAGTGATCTCAAG 295 C1-15 CTCCTGGGGTCACTGTTGTC GAGTCTGTTACTTGCACAGATC 211 C1-16 CAATCATAAAACTTCTATATGTCAC GCCATCACCGGAACAGAAGG 207 C1-17 GTGGGGACATGATTTTCGTCAAG GATTATGTCTGTTAAAGCTCATG 175 C1-18 GACATCCTGACTCAGCTGAAG CGTGTCAGCCATGGTGTCAC 203 C1-19 GCACCACCCATATGATTCAGG CAGATAACCTGACAGCCATCC 179 C1-20 GATGAAGTGCAGCAGAAGTGG CAGTCTGGTCACATCATGATAG 221 C1-21 GCAGAGCTGGAAGAGATGGG GCTGTATTCTCCTGCATTCCG 188 C1-22 GTATAGCATGGTACATTACTGTG CCTTGTGATGAACTATGTTCTTG 228 C1-23 GACTGACGATGTGAAAAGTATTG CCAAAATCTTGTCAAGTATCAGC 235 C1-24 CCAGTCAGGTATATTTGGAAAAG GGAATTTTCTGTTTCCATAAAGTC 206 C1-25 CGATCGCTTATGTGATGCTCG CAAGCTTCTTGAGAGTGACGG 248 C1-26 GAACTTACAGAAGTCTGCAAATC GATGTCCATTCAGAGTATTTCTG 208 C1-27 GGTGTTACTGAAGACAAACTGG GGGTATTTCCTCGTCAAGCTC 168 C1-28 GGATAAGCCTGTCCAGCAGC CCTAGATATCTTTCTACATGTGG 214 C1-29 GCTTCCAGTGGTTCCAAGGC GTTATGAGAATGCATATGAAGGC 204 C1-30 CTCAAGCGACATCAGGAGCC CAAGAATAACAGCTTTATAGTCAC 178 1 All sequence in thr 5’ to 3’ direction
123 N = Nonsense F = Frameshift M = Missense NN FMMM F(4) cDNA 75 nt 4417 nt
hgDNA 1
Ashkenazi Mutation
Figure 19. Mutations identified in BLM. A graphic representation of the mutations in the cDNA, listed in Table is shown. N = nonsense mutations, F = frameshift mutations, and M = missense mutations. The paranthetic 4 indicates the identical, homozygous framshift mutations that were identified at this location in four different individuals.
124 helicases, and one was identified that introduced an amino acid substitution of
cysteine to serine in the carboxy-terminal region of the peptide. Because the
three genetic alterations could be polymorphisms and the actual BS-associated
mutations could have gone undetected, analyses of the BLM gene product in
vitro will be required to demonstrate whether these substitutions cause the
mutant phenotype.
Discussion.
In this study, BLM was isolated by a positional cloning strategy. BLM was
localized by homozygosity mapping to a 2-cM interval flanking FES (German et
al., 1994), a gene already mapped to chromosome band 15q26.1. A 2-Mb YAC
and P1 contig encompassing FES was constructed, and closely spaced
polymorphic DNA markers in the contig were identified (Straughen et al., 1996).
BLM then was assigned by SCP mapping to a 250-kb interval in the contig, one
bounded by the polymorphic loci D15S1108 and D15S127 (Figure 13). A cDNA clone was isolated by direct cDNA selection using a cosmid clone from the interval, and cDNA analysis identified the 4437-bp sequence. This sequence encodes a peptide homologous to the RecQ helicases. RNA transcripts 4.5-kb long were identified by Northern blot analysis, and electrophoretic abnormalities in RNAs were detected in cells from seven unrelated persons with BS, suggesting that these RNAs are derived from mutant BLM genes. Finally, reverse transcription PCR and SSCP analyses disclosed seven unique mutations in ten persons with BS (Table 6 and Figure 19), four that are chain terminating
125 and three that are putative missense substitutions, two of the three affecting amino acid residues conserved in RecQ helicases and the third changing a cysteine to a serine.
The candidate gene for BLM isolated from the interval identified by SCP mapping encodes a 1417 amino acid peptide, homologous to RecQ helicases.
Mutational analysis of the first 13 unrelated persons with BS examined permitted the identification of seven unique mutations in ten of them (Table 6). The fact that four of the seven mutations characterized result in premature termination of translation indicates that the course of most BS is the loss enzymatic activity of the BLM gene product. Identification of loss-of-function mutations in BLM is consistent with the autosomal recessive transmission of BS, and the homology of
BLM and RecQ suggests that BLM has enzymatic activity. Thus, we predict that most BS mutations result in loss of function of BLM.
This loss of enzymatic activity is not lethal in cells because three of the chain-terminating mutations were detected in a homozygous state. The nonlethality could result from the existence of some residual enzymatic activity in the truncated peptides; however, this seems unlikely because one of the homozygous chain-terminating mutations results in chain termination after only
185 amino acids. Apparently, the function of BLM is not essential for cell survival. Other factors, including WRN, RecQ4, or RecQ5, may be able to substitute for BLM, albeit inefficiently. It is also possible that other factors are not required to substitute for BLM, but the processes in which BLM functions continue to occur at a reduced or inefficient rate.
126 In the four persons with Jewish ancestry, a 6-bp deletion and 7-bp
insertion at nucleotide 2281 were identified, and each of the four persons was
homozygous for the mutation. Homozygosity was predictable because linkage
disequilibrium had been detected in Ashkenazi Jews with BS between BLM,
D15S127, and FES (Ellis et al., 1994a). Thus, a person who carried this 6-bp
deletion and 7-bp insertion was a founder of the Ashkenazi Jewish population,
and nearly all Ashkenazi Jews will BS inherit the mutation identical by descent
from this common ancestor. Identification of the mutation now permits the
screening of carriers in the Ashkenazim by a PCR test (Straughen et al., 1997).
BS is an autosomal recessive trait with high penetrance and expressivity.
The observation of loss-of-function mutations in BLM helps to explain these genetic characteristics. The short stature, characteristic facies, facial sun sensitivity, hyper- and hypopigmented patches on the skin, immunodeficiency, male infertility, female subfertility, premature menopause, and the predispositions to late onset diabetes and to neoplasia exist in virtually all groups of persons with the syndrome. The BS phenotype is similar in Ashkenazi Jews, the Dutch, the
Flemish, Germans, Italians, Greeks, Turks, and Japanese; i.e., wherever it has been diagnosed. In addition, the elevated chromatid exchange and the hypermutability are constant cellular manifestations. No more variability in the expression of the mutations has been detected in persons with BS who inherit an identical mutation by descent from a common ancestor, as happens in Ashkenazi
Jews with BS and in non-Ashkenazi Jewish persons with BS whose parents are cousins, than has been detected in persons who are compound heterozygotes
127 (German et al., 1996). This can be rationalized by the argument that all
mutations leading to BS render the protein product nonfunctional or nonexistent.
Nevertheless, with BLM cloned, it is possible to identify the mutations in any person with BS, and subtler genotype-phenotype correlations that may have been overlooked, can be characterized.
The BLM gene product contains amino acid motifs that are homologous to motifs in the RecQ helicases, a subfamily of DExH box-containing DNA and RNA helicases. A dendrogram relating numerous RecQ family members by amino acid identity in each protein’s helicase region is presented in Figure 20. recQ is an E. coli gene that is a member of the RecF recombination pathway (Nakayama et al., 1984), a pathway of genes in which mutations abolish the conjugational recombination proficiency and ultraviolet (UV) resistance of a mutant strain lacking both RecBCD (part of exonuclease V) and SbcB (exonuclease I) activities
(Horii and Clark, 1973). RecQ has DNA-dependent ATPase and DNA helicase activities and can translocate on single-stranded DNA in a 3’-5’ direction (Umezu et al., 1990). 3’-5’ helicase activity has been demonstrated in other RecQ family members, including SGS1, WRN, and BLM (Bennett et al., 1998; Shen, et al.,
1998; karow, et al., 1997).
Besides BLM, several other RecQ family members are known. SGS1 is
a yeast gene in which mutations suppress the slow growth of cells carrying
mutations in the TOP3 topoisomerase gene (Gangloff et al., 1994). RECQL is a
human gene isolated from HeLa cells. rqh1 from S. pombe is required for
reversible S-phase arrest (Stewart et
128 NH2 COOH 1 649 1041 1417 Hs
1472 863 1231 Ce
1661 1052 1447 Sc
1501 894 1328 Sp
173 458 659 Hs
1163 580 809 Ce
1413 470 Dm
10 410 Hs
9 387 601 Ec
1807 1218 1530 Dm
1462 866 1208 Hs
1530 914 1432 Hs
Figure 20. Dendrogram of recQ family members: A linear schematic of many recQ family members aligned by the helicase domains (gray shading) within the helicase region (green shading) are shown. Several recQ homologues from human (Hs), fruit fly (Dm), round worm (Ce), fission yeast (Sp), and brewer’s yeast (Sc) are included. Dendrographic associations are based on helicase region identity at the amino acid level. BLM is closest to the predicted C. elegans protein T04A11.6. BLM is related closely to the well-characterized Sgs1 and rqh1 proteins from S. cerevasaea and S. pombe, respectively. The dendrogram was constructed by comparing the amino acid sequence of the helicase region of each protein with that of every other protein listed and linking the proteins by relative identities.
129 al., 1997). Two other human members of the RecQ family of helicases include RECQL4, the causitive agent of Rothmund-Thompson syndrome (RTS), and RECQL5 (Kitao et al.,
1999). In addition to helicase domains, BLM contains amino-terminal and carboxy- terminal regions that are composed predominantly of charged and polar amino acid residues. The presence of nonhelicase regions in BLM raises the possibility of additional enzymatic activities. The nonhelicase regions could provide functional specificity to
BLM, ie., promoting interactions with other proteins, or to provide substrates for phosphorylation that might regulate BLM activity with relation to the cell cycle.
In a recent tabulation of 42 inherited disease-associated genes isolated by positional cloning (Collins, 1995), 19 were transmitted as autosomal dominants and 17 as X-linked recessives; however, only five were autosomal recessives.
The reasons for the relative paucity of positionally cloned autosomal recessive disease-associated genes are at least twofold. First, the cloning of over half of the genes (26 of the 42 tabulated) was aided by chromosome breakpoints within or near the disease associated gene; however, only one of these was in an autosomal recessive. Second, and of greater importance, the number of families transmitting rare autosomal recessive disease-associated genes generally is small, and the number of persons in sibships who would be informative in recombinational analysis also is small. Because a single investigator usually cannot obtain the numbers of families required for linkage analysis, the localization and subsequent positional cloning of rare autosomal recessive genes have lagged behind that of dominant and X-linked recessive genes.
130 Even when samples from numerous families have been collected and
analyzed, usually the amount of positional information obtained is limited. In the
case of BS, the Bloom’s Syndrome Registry (German and Passarge, 1989), a
research resource that has provided the material for all our recent genetical
studies, made possible an extensive recombinational analysis of BLM by homozygosity mapping. This analysis permitted a minimum regional assignment of BLM to approximately 1.3-Mb (unpublished data). This size of minimum interval is typical of recombinational analysis of genes from a 1.3-Mb region would have been laborious.
The problem of too little positional information in available families can be mitigated in exceptional situations in which linkage disequilibrium between the disease-associated gene and tightly linked polymorphisms can be detected in a genetic isolate. In these cases, localization of a gene to a short interval in the genome by haplotype analysis can be more exact than is possible using standard linkage analysis of family data (Kerem et al., 1989; Sirugo et al., 1992; Lehesjoki
et al., 1993). Linkage disequilibrium in fact was a strategy available in BS (Ellis
et al., 1994a), and it permitted a minimum regional assignment of BLM to the same 250-kb interval reported here (Ellis et al., 1996). This approach could have allowedustocloneBLM. Instead, we first carried out SCP mapping.
In the SCP mapping strategy, we took advantage of the recombinant cell lines established from BS somatic cells in which crossing over within BLM had taken place, resulting in the correction of the mutant phenotype in their progenies
(Ellis et al., 1995). After a segregational event, all polymorphic loci distal to BLM
131 were reduced to homozygostity in half of the cases of intragenic recombination.
This mapping method was preferred to linkage disequilibrium mapping because
the crossovers that permitted localization of BLM had occurred within the gene itself and fewer genotypes were required for the analysis. By genotyping polymorphic loci that flank BLM in high SCE and low SCE samples from only five persons with BS and their parents, we delimited the position of BLM to the short interval bounded by the marker loci D15S1108 and D15S127 (Figure 13). With
BLM assigned to such a short interval, the cloning of BLM became
straightforward. The first candidate gene isolated from the interval proved to be
BLM.
Helicases unwind double-stranded nucleic acid molecules. Helicases are
believed to be tightly regulated to allow spatial and temporal enzymatic
specificity. Helicases can be involved in DNA repair, replication, or
recombination. A universal function for the RecQ helicases has yet to be
established. SGS1 has been shown to suppress homologous recombination and
genomic instability (Myung et al., 2001). Some evidence has been provided that
implicates rqh1 in resolving Holliday junctions, a DNA recombination intermediate
(Doe et al., 2000). WRN has been shown to suppress hyperrecombination in
SGS1 mutant yeast (Yamagata et al., 1998). Abnormal mitotic recombinations
have also been identified in Werners syndrome cell lines (Prince et al., 2001).
There is data suggestive that BLM is also involved in a recombination pathway.
First, BS cells have a elevated rate of SCE, a readily detectable result of a
specific recombinational event. Second, as detailed in this chapter, SCP
132 mapping relies on the localization of a recombination that occurs within the gene itself. BLM’s ability to suppress recombination is supported by the facts that SCP couldbeusedtolocalizeBLM and recombination occurred in several unique BS cell lines. In addition, like WRN, BLM can suppress hyperrecombination in SGS1 mutant yeast (Yamagata et al., 1998). Recently, BLM has been shown to interact with RAD51, which catalyzes recombination by binding to DNA at regions of double-strand breaks (Wu et al., 2001).
133 Chapter 7. The Genomic Structure of BLM.
The genomic structure and intron-exon boundaries of BLM have been determined
in order to facilitate mutational analysis of BLM. This information will be useful in
confirming the diagnosis of BS, to identify carriers of the disease and to characterize
tumors that carry mutant BLM alleles. PCR, long-range PCR, Southern blot analysis,
subcloning, and DNA sequencing have been employed to accomplish this
characterization. BLM is composed of 22 exons, which range in size from 70 to 701 bp.
The gene covers a genomic distance of approximately 100-kb, excluding 5’ and 3’ regulatory regions. Exons one and two are separated by more than 29-kb.
Determination of the Intron-Exon Structure of BLM.
Primers designed from the BLM cDNA sequence for SSCP mutational analysis of
BS patients (Ellis et al., 1995) were used to amplify normal genomic DNA (Table 7).
Sizes of these amplification products were compared to the sizes of amplification
products from cDNA using the same primer pairs. This provided an initial map for intron
locations and sizes. Several of these PCRs, using genomic template failed to amplify,
suggesting two possibilities: one, that the intronic sequence is too large for conventional
PCR,; or two, that the PCR primers straddle an exon-exon boundary in the cDNA and do
not have enough consecutive homology to anneal to the genomic template. These
explanations were addressed by synthesizing long-range PCR primers in the vicinity of
the original primer and by repeating the PCR. The long-range PCR proved successful in
obtaining exon distances that were not obtained by regular PCR, in all but one case, exon
134 Table 7 PCR Primers Used to Determine the Intron Sizes of BLM
Intron Name Forward Sequence Reverse Sequence
2 AF/BR CAACTAGAACGTCACTCAGCC GAGGTTCACTGAAGGAAAAGTC 3 6F/R GAAGATGCTCAGGAAAGTGAC CGTACTAAGGCATTTTGAAGAGG 4 7F/R-2 CAACTGAGAAGTTCCATGTATTG CACAGTCTGTGCTGGTTTCTG 6 9F/R CTATTCCTGATGATAAACTGAAAC GGAAGTATCTTAAGGGACATCC 7 13F/R CAGCCAGCAATCTTCCACAG CGCTCATGTTCAGATTTCTGG 8 14-3F/14R CCAGAAATCTGAAACATGACGC GATCTACGATAAGTGATCTCAAG 9 15F/FG3R CTCCTGGGGTCACTGTTGTC ACCTGTCAGATATGTAGCTGGA 10 16F/Gen16R CAATCATAAAACTTCTATATGTCAC GTTACCATTTGGGGTTTCTGG 12 18F/17R GACATCCTGACTCAGCTGAAG CTAATACAGACAATTTCGAGTAC 13 18F-B/R GGCATTTGATTGCCTAGAATGG CGTGTCAGCCATGGTGTCAC 14 XL20F/R GTGATTCTGCCAGAGATGAAG... GGTATAGAAAAGCAGGCAGTG... TGCAGCAGAAG AGATATTTCCCC 15 21F/R GCAGAGCTGGAAGAGATGGG GCTGTATTCTCTGCATTCCG 16 22F/R GTATAGCATGGTACATTACTGTG CCTTGTGATGAACTATGTTCTTG 17 23F/R GACTGACGATGTGAAAAGTATTG CCAAAATCTTGTCAAGTATCATC 18 25F/R CGATCGCTTATGTGATGCTCG CAAGCTTCTTGAGAGTGACGG 19 XL26F/R GAGGAACTTACATAAGTCTGC... CCTGGCGATGTCCATTCAGAG... AAATCTCTGGGG TATTTCTG 20 27F/R GGTGTTACTGAAGACAAACTGG GGGTATTTCCTCGTCAAGCTC 21 28F/R GGATAAGCCTGTCCAGCAGC CCTAGATATCTTTCTACATGTGG
135 1. Lastly, DNA sequence information in Genbank (Evans et al., 1997) was accessed and used to complete the intron-exon maps assembled here, in particular, exon 1, which failed to amplify in regular or long-range PCR.
BLM is composed of 22 exons, ranging from 70 to 701 bp (Figure 21). Intron sizes were estimated from the PCR products spanning from one exon to another and in some cases were confirmed by sequencing; the introns range from 206-bp to 29.5-kb in size. Exon 1 is the smallest exon, consisting only of 5’ untranslated sequence, based on the best Kozak sequence and ATG present in exon 2. The intron between exons 1 and 2 is extremely large, 29.5 kb (Evans et al., 1997) and is one of the largest introns reported in literature.
Determination of the Intron Sequences of BLM.
The PCR products used to position exon boundaries and intron sizes also were used as templates to determine the sequence of intronic regions immediately adjacent to the BLM exons. Primers used to generate the PCR product or nested primers were used to obtain intronic sequences from the 5’ and 3’ sides of each exon boundary (Figure 22).
All intronic splice junction sites matched known consensus sequences (Mount, 1982 and
Figure 22).
Mapping BLM onto the YAC/P1 Contig.
YAC and P1 clones first used to construct a representation of the BLM locus were further characterized by mapping of the BLM exon-specific clones (Straughen et al.,
1996). PCR primers (Table 7) were used to amplify specific regions of the gene from
136 ntified 21 22 20 c905 c904 I sites and distances were determined through is shown across the top with thick vertical lines EcoR BLM I sites. The P1 1958 4.3 1.9 4.8 3.4 13.0 0.9 9.0 2.0 7.5 EcoR I restriction enzyme digestion sites. Intronic distances in kilobases are shown D15S1144 D15S1145 EcoR The genomic representation of P1 4430 2.1 1.5 2.1 810 1.7 13.2 2.0 5.6 3.2 2.0 5.4 15.0 2.0 3.5 7 9 12 13 14 15 16 17 18 19 I sites are shown in kilobases below the 1.6 5.2 0.3 0.2 0.5 3611 EcoR 245 and relative location of each exon. P1 4428 BLM region. The mapping allows association of specific exons with particular elements of the contig. BLM 29.9 1.8 2.9 9.2 13.0 6.0 10.0 3.2 0.9 4.3 5.0 5.0 0.2 4.3 P1 4429 1 BAC 1 extending up indicating exons andabove thin the vertical lines exons extending and downsequencing distances indicating and between Southern blot the analysis (see Figure). At the bottom a contig that includes the entire gene is shown composed of P1s, BACs, and cosmids ide Figure 21. Contig containing during the mapping of the
137 . aacagctctgcct ctacatgttagaatc cttgctagggaactcc tctcactgaattagaag tatgtatttactgaataaaa agctctttaatagaaataaa acatgagatatcttctcttta tcctctggataacctttttatt agtctactcctgctattgtgg ggaattatataagaaaaac ccacaataagatatataaaa caacgtaacacaaagattgtg tagttcttctttttaaaataaaca aagaatttattacaagtacatag gtttaatttagtttataatatcattc tatttttaagattttaacaaaatttt aagtgaaaaatggacacgggca aaccaaacatgcacatgtagaat atattattgtgaagtatagtgtcttt tttaaaaatgtagatacaaattagattg
gtgagtaccgcgcgcgtaactacgggtcggtccgcattgatctagccctgctctggcggcccggcccgg gtaacaattattttatcttcattttagtatgttcattgtacttttttattcaaagctagccattgggaatagtcatga
70
172 2479 1294 1956 2381 2897 3093 3825 1033 1161 2148 2267 2630 2736 3284 3433 3632 3948 4148 4417 873 is composed of 22 exons which vary in length from 70 nt (exon 1) to 701 nt (exon 3).
Exon 1 70 nt.
Exon 2 102Exon nt. 3 701Exon nt. 4 160Exon nt. 5 128Exon nt. 6 133Exon nt. 7 662Exon nt. 8 192Exon nt. 9 119 nt. Exon 11 99 nt. Exon 10 114 nt. Exon 12 149 nt. Exon 13 107 nt. Exon 14 161 nt. Exon 15 196 nt. Exon 16 191 nt. Exon 17 148 nt. Exon 18 200 nt. Exon 19 193 nt. Exon 20 123 nt. Exon 21 202 nt. Exon 22 266 nt.
2382 1 71 1162 1295 2268 2737 2898 3633 1034 1957 2149 2480 2631 3094 3285 3434 3826 3949 4151 173 874 BLM Flanking intronic sequences areExons, shown shown as in gray lowernumber boxes, case. in are each exon numbered All is and also introns contain included contain diagonally the at the number the standard of lateral margins nucleotides consensus of (nt.) each splice in exon sequences, box. that gt exon. at the The 5’-end corresponding first and and ag last at cDNA the nucleotide 3’-end Figure 22. Exon size and flanking intronic sequence. ggcttccccaggaagcagccaatcggaataggcaagcttccggcgggaagtgagccagggcttg aaaaattagttttgtagagttggggggtttcttaaaatggatccatctaatctagtttttccattatttttcag taatcgctcatgccctgttctttctgtctcattagtggttaacaaatctatgtttatcaactgttttactgtagttaagggtttcaaaattatacatttattgagtctagcctatagtatgattggcttaacatttttttatttgcagagccaccatgcctagcccaagacttttttttttttccctcaaagaaaaatattaacaacataattattttatagctacagatttgcttttgtggcctaccagagtaaactacttatatttaatacgttgttctcttttctctcttcaggcaagggaaatgctaaagctgtactttcactgtattcatgtacgatttttcttaacgttgattattttcctaggatttcttttgtaacttttacattcatgctctgaagacagaacctgacagatatttttcattgttctctttcagtaaaacctaaggacaaatgtaattttgtcaggttaatgtataaaattgaaattgtttacttcttttatacttagaatttgaagaccacagaatcatgaggtgatgtgtttcagtgtttttacatgtctaatgtatttctggcctagtgagcagtgttggctttttatagaaggaagctccaagtagtctgaaaagcagtatttttttttccaactag gtaaactagctaaataattagcattattatttgtttctgggatactttaaatt acatgtagtctataacaatacctaaagtcatattttctcataataactaaattttatgtttgggacttttttag gtacaagcaatattttagacataccatgtatttcaactacttacttttgaaaa ctgaaaatgtagtgtaaattgtgtttttgtttatgttaaaaattcttgtttctcagtactcttggtttcttggcag gtatcttaattttcccccttctggaatatatctgattatatttctaccactct gaagcatctattatgaaaatgttccttcaagtctgtgccattatgaatctaataagcttttgcttttatatcag gtaagtttaaaataaattgaatgcttatatgaaaacaaaactgtcccaaaata tattctaaatatttctatcatatgctctatttttcccctataagtaatgtcttactatagtgttcatgtgtttttag gtatgtatttttagaagtgaattggcaggaatccattggcagatgttaaatgaa gttaatattaaaccctagtaatctaggcattgttaccttaattatagcagaaagtattctctttttattcatag gtaagttataaaaatactaataaaaacacgccttagaaacaattaaatttcag ctttttagcctcttctatttgagggtgatgatatacgtacatttactcatcttacttcctgtatcttcttatcag gtttgtatttatatcattattttaaaatatattaaagaccactagaatacata ttaaataaagcccctgtatgggtacaagtgcacatatacccactcctatgatttgtttctctctcataaag gtaaatactgttttttatatccggaaataccgataaatacatactaccaacaata gtgaatgagcctgaattcagtgggttttctataggtgataatttaaattcctaattttatgcctttgcacag gtaagttgttgcacgtcacgtatttgagaaccctggggcagtgactgccaaagctg ctcatatacactaaaaacacgtggaccagtgcgacatcacctgtaaacatctgcattttccatttgtag gtgagtacagccatgtgattagctgtctagaagtaacaaatgtctttttagta caggttgagaggaagaaggtcattcatttttggtttcatttaacattttgatttttttctttgtcacatttcag gtaacatttttaaagataaacaaataatagaaataatcttttatagcatat gtaagctgggctccattgtagagacattctgtcatcttcagcctcatgatagt gtaaaaaaagaagttttaaaattctttataattaaatttttttctcttac gtaagtcatctgttttgaatgtttgagttacttcaattgaaattgaacatcta gtatagtatttttcatgtttattttattatctcacaatgagtgaaccaaaat gtgggtacacatgtatcctttgttacgtggcacagattaataggccgaaagttaat gttagtacacagccatgtgtgttctctaaaagcctgtttaatgtgaagcgacgcg gtatgttttgtgacatctttttcaatatagggaacaagggaagaaaggacaaaagtgc atttgaagtttttactcgtctctattaatatttaaataaatgctggggggtga cctcccctcaaaaaacattgtgattaatgcaaagtacctaactccactgatttctttttccctcactttttag gtaagtgttttgactggtttgctgtcacataggcactaacttaccacattgtac
138 YAC and P1 DNAs, as well as to probe Southern blots of YAC and P1 DNAs. The presence or absence of a PCR product of the expected size was used to determine whether a region was contained within a particular clone. Southern blot analysis confirmed that a particular exon was contained within a specific clone on the map contig
(data not shown).
Discussion.
In this chapter we present the genomic structure of BLM. The approximate number and position of intron-exon boundaries were initially determined by PCR.
Sequencing PCR amplified genomic DNA provided refinements of our estimations.
Sequencing not only allowed the exact location of intron-exon boundaries, but also provided intronic sequence that flanked those boundaries. Each exon was mapped back to the contig created in chapter 5 of this work by a combination PCR and Southern blot analysis. The use of specific genomic DNA fragments (P1 or BAC) as the template in
PCR amplifications provided much of the mapping data. Southern blot analysis was used to confirm PCR results.
Exon 1 and its flanking genomic DNA proved to be a difficult fragment to locate.
Being the smallest exon at 75 bp and the farthest away from a neighboring exon at 29.5 kb, it is not surprising that PCR amplification was an inadequate technique.
Approximately one year after we initiated the genomic analysis of BLM, Evans et al.
(1997) submitted to Genbank the sequence of a large genomic fragment that contained the first 15 exons of BLM, making the exact size and sequence of the first intron available. This also served as an independent mean by which our experimental
139 determination of boundaries and flanking sequences could be verified for the exons two
through15ofBLM.
The RNA transcripts that require splicing to remove introns contain specific
sequences in the intron that signal the splicing machinery within the nucleus to excise the
intron. The majority of introns within transcripts are flanked by GU at the 5’ end and AG
at the 3’ end. The splicing machinery identifies these dinucleotides as the terminal ends
of the intron to be removed. All introns within BLM are identified by the 5’ GT and 3’
AG sequences (Figure 22).
Identification of mutations is one of several potential application of the genomic structure of BLM. Sequences immediately flanking the exons can be used to design PCR oligonucleotide primers that can amplify entire exons. These amplified products can be used directly for mutational analysis or as templates for exon specific DNA probes. Not all mutations affecting the expression, length or content of a message are found in the coding region. Mutations found in introns, promoter and enhancer sequences, and polyadenylation sites can also be responsible for mutant phenotypes. Mutations located in introns may affect the transcript splice sites or the formation of the lariat structure.
The consequences of such a mutation would result in missplicing and/or lack of splicing.
Mutations in the promoter or enhancer elements are likely to alter the expression of the gene, potentially decreasing the expression enough to induce a phenotypic response.
Mutations in the polyadenylation site result could result in a shorted or missing polyadenosine tail, decreasing message stability. The determination of the genomic structure of BLM provides sequences necessary for analysis of noncoding regions.
140 Patient DNAs, particularly those in which mutations could not be identified by SSCP analysis of cDNA during the cloning of the gene, can be evaluated.
141 Chapter 8. A Rapid Method for Detecting the Predominant Ashkenazi Jewish
Mutation in the Bloom’s Syndrome Gene.
Bloom’s syndrome, an autosomal recessively transmitted form of growth
deficiency, is rare in all populations but relatively common in Ashkenazi Jews (German,
1993). In the Bloom’s Syndrome Registry, a population composed of most persons
diagnosed with BS, one or both parents of 52 of the 168 persons registered are Ashkenazi
Jews (German and Passarge, 1987). The relatively high frequency of mutation at the
Bloom’s syndrome locus (BLM) in the Ashkenazim had been predicted to be on the basis
of founder effect rather than due to selection (German, 1979). This hypothesis was
recently confirmed when the same mutation in BLM was identified in each of the first
seven Ashkenazi Jewish persons with BS who were examined. Furthermore, each of
these seven patients was homozygous for that mutation (Ellis et al., 1995). Other
mutations in BLM were transmitted in families of other ethnic backgrounds (Ellis et al.,
1995). Thus, identification of blmAsh, the name for the mutant Ashkenazi Jewish allele of
BLM, has unique diagnostic value in the Ashkenazim.
The mutation blmAsh is a 6-bp deletion/7-bp insertion at nucleotide 2,281 of the
open reading frame of BLM that results in a frame shift and a stop codon at nucleotide
2,292 (Ellis et al., 1995). blmAsh introduces a BstNI restriction enzyme site into BLM.
Therefore, digestion of DNA with this altered sequence produces uniquely sized bands
that permit differentiation of unaffected noncarriers (blm+/blm+), unaffected heterozygous carriers (blm+/blmAsh), and homozygous affected individuals (blmAsh/blmAsh). A screening
142 test for use with either cDNA or genomic DNA is presented. This screening method is
not intended to identify non-blmAsh mutations in BS patients or carriers.
Genomic DNA amplification was accomplished using a forward primer that is
complementary to intron sequences upstream of the mutation, between exons 9 and 10.
The forward primer for the cDNA amplification includes sequences 2,191-2,210 in the
BLM open reading frame (Ellis et al., 1995). Both forward primers employed include a
BstNI site as an internal control for the activity of the restriction enzyme. The cDNA
forward primer includes a site found in the BLM cDNA 86-bp upstream of the mutation.
A BstNI site was added to the genomic DNA primer because no naturally occurring
BstNI site is present in this intron sequence. In addition, the forward primers include
M13-UP sequences that permit adequate separation of the undigested and the digested
PCR products and sequencing by means of a standard primer, should it prove necessary.
Reverse primers were designed from sequences overlapping exons 10 and 11 for cDNA
or intron sequences between exons 10 and 11 for genomic DNA amplification. The
primer design, BstNI sites, and BLM sequence in the region of blmAsh are shown (Figure
23).
The total length of the PCR product amplified from cDNA is 231-bp; that of
genomic DNA is 337-bp. cDNA from persons lacking blmAsh when digested with will generate DNA fragments of 209-bp and 22-bp, although only a single band is seen because the 22-bp fragment is not visible by agarose gel electrophoresis (Figure 24). cDNA from persons with BS who are homozygous for blmAsh will contain two BstNI
restriction sites within the amplified PCR product. Restriction enzyme digestion of this
type of PCR product will generate two bands: 115-bp and 94-bp. DNA from a person
143 DNA without BLMAsh : …CATATCTGACAGGTGA
Genomic DNA with BLMAsh :
5’Primer tgtaaaacgacggccagtccaggacaaatgtaattttgtcagg…
…CATTAGATTCCAGGTGA… …ccagaaaccccaaatggtaac
Complement of the 3’Primer Control BstN I Site Ashkenazi Ash Mutational cDNA with BLM : BstN I Site 5’Primer TGTAAAACGACGGCCAGTCTCCTGGGGTCACTGTTGTC…
…CATTAGATTCCAGGTGA… ...GATCTGTGCAAGTAACAGACTC Complement of the 3’Primer
Figure 23. PCR-based Schema for Detecting BlmAsh. DNA sequences of the BstNI restriction sites in the PCR products from cDNA and genomic DNA for the detection of BlmAsh.PCRprimers, diagrammatic representation of BstNI sites, and fragment lengths after PCR and BstNI digestion are shown. Nucleotides altered by BlmAsh are shown in large bold type. BstNI sites are underlined with the thick line. Specific primer sequence is underlined with the solid arrow. The M13-universal primer sequence used as a primer tail is shown with a dotted underline. Upper case represents exon sequence. Lower case represents intron sequence.
144 Figure 24. Genotyping of individuals for the BlmAsh mutation in BLM. (a) Generated using genomic DNA as the template and the appropriate genomic PCR primers. (b) Generated using cDNA as the template and the appropriate cDNA primers. The order of the samples in the gel is the same for both the genomic DNA and the cDNA: lanes 1 and 9, 100-bp ladders; lanes 2 and 8, uncut PCR products from control DNA extracted from a cell line derived from a normal non-Jewish person; lanes 3 and 7, BstNI degested PCR products from control DNA extracted from a cell line derived from a normal non-Jewish person; lane 4, PCR products from DNA from cell line HG1963, derived from the father of 142 (MaMatu); lane 5 PCR products from DNA derived from cell line HG2811, which is from 142 (MaMatu), who has BS; lane 6, PCR products from DNA from cell line HG1962, derived from the mother of 142 (MaMatu). Size markers are shown on the left in 200-bp increments. Arrow size of bands generated by PCR indicated on the right. U, undigested DNA; c digested DNA. The pedigree of 142 (MaMatu), shown at the top, illustrates the persons from whom DNA was amplified; squares, males; circles, females; open and filled halves of symbols, nonmutated and mutated BLM alleles, respectively.
145 who is heterozygous for blmAsh, both compound heterozygotes with BS carrying a single
Ashkenazi allele or an unaffected carrier, will generate three bands: 209, 115, and 94-bp.
It is interesting to note that in blm+/blmAsh heterozygotes, representation of the mutant
allele as ascertained by nonquantitative RT-PCR method is much less than that of the
normal allele. This suggests that the mutant RNA is either less stable or is expressed at a
lower level than the normal allele. Evidence for such instability aslo comes from
Northern blot analysis of RNA from persons with BS noted in the previous chapter (Ellis
et al., 1995). Digestion of the genomic PCR product gives the following results: DNA
from unaffected noncarriers of blmAsh will generate a 317-bp band because of the forward
primer restriction site; DNA from persons who are heterozygous for blmAsh will generate
bands of 317, 231, and 86-bp; DNA from persons with BS who are homozygous for
blmAsh will generate two bands, of 231-bp and 86-bp.
Thus, the mutation transmitted in the Ashkenazim, clearly responsible for the vast
majority, if not all instances, of BS in that population, can be readily identified by
conventional molecular techniques. There are several uses for such a technique: (1)
determination of the heterozygote frequency of the blmAsh in the Ashkenazim; (2) identification of blmAsh carriers in Ashkenazi Jewish families ascertained through BS; (3)
genotyping of embryos and fetuses known to be at high risk of being homozygous
blmAsh/blmAsh; and (4) use in the experimental laboratory.
146 Chapter 9. Frameshifting and Recombination in Cell Lines Over-Expressing BLM.
BS cells have an inherent genomic instability as observed by the high SCE rates, telomeric associations, and quadriradial formations previously reported and microsatellite repeat instability as discussed in chapter 4. The specific biochemical defect that results from a lack of the BLM helicase, leading to genomic instability is not clear. Helicases are important enzymes capable of unwinding double-stranded nucleic acids involved in such metabolic processes as DNA replication, DNA repair, DNA recombination, and transcription. Based on the hypothesis that BLM is a suppressor of DNA recombination and evidence of genomic instability in BS cells, we asked the question: Can BLM suppress genomic instability in tissue culture? We used two mammalian cell lines, each with a unique transgenic construct intended to measure specific DNA metabolic events.
One, the G11 cell line assesses frame shifts, while the other, FSH indicates DNA recombination at a specific locus.
Over Expression of BLM in the G11 cell line.
The G11 cell line (Provided by Dr. Jim Stringer) was originally derived from mouse NIH 3T3 fibroblasts and carries a transgenic construct of a modified placental alkaline phosphatase (PLAP) gene driven by the Rous sarcoma virus (RSV) promoter
(Figure 25). The wild type PLAP gene has four consecutive guanine residues starting at the sixth nucleotide within the coding region. The modified construct contains an additional seven guanines, altering the open reading frame of PLAP so that no functional
147 wild-type PLAP exon 1 PRSV ATG - CTG - GGG - CCC - TGC -
GGG-GGG-G frameshift-1 (11Gs) PRSV ATG - CTG - GGG - CCC - TGC - PLAP exon 1
PRSV ATG - CTG-GGG-GGG-GGG-GCC-CTG-C
Figure 25. G11 cell line contains a marker gene for measuring frameshifts. G11 cells, originally derived from green monkey kidney cells, contain a modified placental alkaline phosphatase (PLAP) transgene. The wild-type gene has four consecutive guanines near the 5’ end of the coding sequence. G11 cells contain 11 consecutive guanines at that 5’ location, altering the reading. A frameshift event can restore the reading frame, which can be detected by a colorimetric assay.
148 protein is produced (Deprimo et al., 1998). A frame shift mutation reduces the 11
guanines to ten, resulting in a corrected open reading frame and a functional PLAP.
Purple stain from the colorimetric assay described in materials and methods indicates the
presence of functional PLAP protein (Figure 26).
To address the contribution of BLM to nucleotide repeat stability, BLM was over
expressed in G11 cells using the pOPRSV:BLM construct, which was generated by
replacing the CAT gene of the pOPRSVICAT (Stratagene) vector with wild type BLM.
Data indicated a subtle decrease in the number of frame shift events per 104 cells
transiently transfected with pOPRSV:BLM, compared to vector alone. In order to get a
more accurate comparison stable transfectants were isolated and scored for frame shift
events. Unfortunately, the data from the stable transfectants does not indicate any
significant difference between cell lines expressing BLM and those containing vector
alone (Figure 27).
Over Expression of BLM in the FSH cell line.
The FSH cell line was made from green monkey kidney cells that contain a
modified β-galactosidase transgenic construct. This construct consists of 2 β- galactosidase genes in tandem, neither of which can produce a functional protein in their current state. The 5’ β-galactosidase gene has no promoter, but a wild type coding
sequence, and the 3’ β-galactosidase gene has a promoter, yet the ATG has been deleted
at 5’ end of the coding sequence (Figure 28). An intrachromosomal recombination or an
SCE event between the 5’ and 3’ β-galactosidase genes will allow functional β-
149 C
D
Figure 26. G11 Cells that Undergo a Frameshift Event can be Stained Purple. G11 cells were plated and incubated for four days, at which time they were fixed in 4% gluteraldehyde. incubated at 60°Cforan hour to denature endogenous placental alkaline phosphatase (PLAP), and stained according to procedures in Materials and Methods.
150 Frameshifts in G11 Cells
0.000035
0.00003
0.000025
0.00002 Without BLM 0.000015 with BLM
0.00001
0.000005
0 C (-) Mock 12345 Transfected Clone
Figure 27. Result of Frameshifting Experiments in G11 Cells. Cell lines were derived from G11 cells previously transfected with either the pOPRSV:BLM construct or the pOPRSV empty vector. Equal concentrations of stable-transfected G11 cells were plated and grown for six days prior to fixing and staining.
151 FSHcelltransgenicconstruct
∆ PSV40 FRT LacZ hgh FRT LacZ Poly A
recombination event
sister chromatid exchange
∆ PSV40 FRT LacZ hgh FRT LacZ Poly A
∆ PSV40 FRT LacZ hgh FRT LacZ Poly A
or
Intrastrand exchange
∆ PSV40 FRT LacZ hgh FRT LacZ Poly A
∆ PSV40 FRT LacZ hgh FRT LacZ Poly A
Cells stain blue with X-gal
PSV40 FRT LacZ hgh
Figure 28. FSH cell line contains a marker gene for measuring recombination. FSH cells were originally derived from mouse NIH 3T3 fibroblasts and contain two copies of the β-galactosidase gene. The upstream copy is lacking an initiating ATG and the downstream copy has no promoter. A homologous recombination between the two β-galactosidase genes, either through sister chromatids or intrastrand exchange, will allow the promoter to drive the functional gene, which can be detected by a
colorimetric assay. PSV40 is the simian virus early promoter. FRT is the restriction site used to create the construct. LacZ is the β-galactosidase gene. hgh is the hygromycin resistance gene.
152 galactosidase to be produced. A colorimetric assay, using X-gal, stains cells with
functional β-galactosidase protein blue.
BLM was over expressed in the FSH cell using the same construct and methods
used in the G11 experiments in order to test the hypothesis that BLM affects genomic
stability by suppressing DNA recombination. Initially, results from transient
transfections indicated a marginal decrease in recombinational events in cells that
contained the BLM construct. However, subsequent stable cell lines expressing BLM did
not yield a decrease in the number recombination events (Figure 29).
At the time these experiments were performed, there were no commercially
available antibodies to BLM, subsequently, no way of directly examining that our
construct was over expressing BLM in these cell lines. However, the BLM expression
construct used in these experiments was also used in previous experiments by to reduce
the SCE rates in BS cells (Ellis et al., 1999). In addition, Southern blots were made from
the stable FSH and G11 cell lines that over expressed BLM to verify the presence of the
construct (data not shown). Finally, to visualize that BLM could be expressed in these
cells in a transient transfection assay, a second expression construct was created, EGFP-
C2:BLM, in which an enhanced green fluorescent protein - wild type BLM fusion protein
was driven by the cytomegalovirus (CMV) promoter. Expression of EGFP-BLM could
be seen under UV microscopy with the appropriate filter (data not shown). The
distribution pattern of EGFP-BLM is consistent with that seen in immunofluorescence
experiments since published with the use of anti-BLM antibodies (Zhong et al., 1999). In these experiments there was no significant change in the number of frame shift or recombination events in the respective cell lines.
153 Recombination in FSH Cells
0.00006
0.00005
0.00004 without BLM
0.00003 with BLM
0.00002
0.00001
0 C (-) Mock 1 2 3 4 5 Transfected Clone
Figure 29. Result of Recombination Experiments in FSH Cells. Cell lines were derived from FSH cells previously transfected with either the pOPRSV:BLM construct or the pOPRSV empty vector. Equal concentrations of stable-transfected FSH cells were plated and grown for six days prior to fixing and staining.
154 Discussion.
Helicases are enzymatic proteins that transiently open double-stranded nucleic acids to allow other metabolic processes such as DNA recombination, repair, and replication take place. These series of experiments were designed to test the involvement of BLM in mononucleotide repeat stability and recombination. Nucleotide repeat instability has often been theorized to be the result of DNA polymerase slippage during replication or repair. It does not appear, from these data, that BLM is responsible for maintaining polymerase fidelity or in recombination. Of course, there are alternative explanations for our results. In the G11 and FSH cells that we used, there is presumably an endogenous BLM homologue. This homologue could be affecting frame shift and recombination events to a maximum extent, prior to the addition of our BLM expression construct. In other words, there is a pre-existing saturating amount of BLM or BLM is not the limiting factor, making further expression of BLM ineffectual. In addition, there are several different types of recombination. From experiments done in the bacterial system we know that different types of recombination and different aspects within recombination require different helicases. Some helicases may function in the classical recombination initiated through double-strand breaks; others may be employed to allow recombination between highly homologous regions of DNA; still other helicases may be used in recombinations involving very short homologous segments. Considering these possibilities, BLM may be able to suppress DNA recombination, but only under the appropriate circumstances. Our tissue culture experiments were limited to poly-guanine stability and DNA recombination at highly homologous sequences by the cell lines made available to us.
155 Chapter 10. A Mouse Model of Bloom’s Syndrome.
With the identification of BLM, the generation of a Blm knockout mouse was pursued. The goal was to make a mouse model of BS, characterize its features, and more importantly, to develop a genetic tool to manipulate genomic stability and tumor predisposition in mice. Arguably, the predisposition to a variety of cancers at a young age is the most notable phenotype of BS and predicted trait of our BS mouse model. By breeding our Blm knockout mouse to other strains of mice that develop tumors, we hope to accelerate the process of tumorigenesis. The Blm knockout mouse also will serve as a facilitator to identify new cancer-causing genes through loss of heterozygosity (LOH) studies of tumors identified in these mice. BS cells exhibit excessive recombination, which increases the likelihood of LOH. The loss of normal tumor suppressor allele located in the genomic regions affected by LOH may be associated with specific cancer phenotype. Mapping areas of LOH within tumors samples from Blm knock-out mice may provide candidate tumor suppressor genes, or at least genomic regions, for further investigation. These ideas have suggested the hypothesis that the Blm knock-out mouse can be used as a genetic tool to increase genomic instability and hence the rate and onset of cancer in mice in order to identify new genes involved in tumorigenesis.
Design of the Blm Knock Out Construct and Generation of Mice Carrying the
Targeted Allele.
Six λ-phage clones from a Black Swiss mouse genomic library were identified through three rounds of hybridiaztion screening with a 400 bp probe. The probe was
156 created by amplifying mouse (C57) genomic DNA using human-specific primers C1-2F
and C1-3R, followed by gel purification and a second round of amplification with random
primers and the human-specific primers C1-2F and C1-3R with [α-32P]-dCTP (Table 6).
DNA was prepared from all 6 clones and one clone was chosen for further study by
testing positive by PCR with primers originally designed for SSCP analysis of BLM.
This clone contained regions homologous to human BLM, exons 10, 11, and 12. This λ- phage clone was digested separately with a variety of restriction enzymes and shotgun- cloned into pBluescript. Colonies were screened by hybridization with a mouse DNA probe homologous to human exon 10 and a 9.6 kb SacI fragment containing DNA
homologous to human exons 10, 11, and 12. The mouse genomic DNA then served as
the backbone for the knockout construct. In an effort to make our mouse model similar
to the human disease, we choose the genomic region in which the most common mutation
in BS persons had been identified as the basis for our knockout. The 6-bp deletion/7-bp
insertion at nucleotide 2281 is referred to as the Ashkenazi mutation or BLMAsh, reflecting the fact that all Ashkenazi Jews with BS have this mutation. This mutation is located near the 5’ end of the human exon 10. Since exon 10 is 114 nucleotides, there is a potential for an in-frame read-through if just exon 10 is removed. The same is true for exon 11. Therefore, the construct was designed so that exons 10, 11, and 12 were removed, changing the reading frame downstream of exon 9 in the knockout construct.
In the final construct, exons 10, 11, and 12 were replaced by the selection gene HPRT within the regions of homology, and the viral TK gene, a negative selection marker, was included outside the region of homology (Figure 30). In the presence of HAT media,
157 TK HPRT ClaI Sac Sac BclI BamH
M
5.1 1.3 Blm targeting HPRT TK M
genomic
knocked-out Blm HPRT
Figure 30. The Blm knock-out mouse construct. Genomic mouse DNA fragments were isolated that contained regions homologous to human exon 10 through radiolabeled probe hybridization with a λ–phage genomic mouse DNA library. Appropriate λ–phage inserts were restriction enzyme digested and characterized by Southern blot. A 9.6 kb SacI–SacI genomic mouse fragment (black line) containing DNA homologous to human exons 10, 11, and 12 (black boxes) was cloned into the plasmid Bluescript (Stratagene) multiple cloning site (MCS,light gray). The resulting plasmid was further characterized by Southern blot techniques and sequencing. Sequences homologous to human exons 10, 11, and 12 were excised by BclI(5’)andBamHI (3’) and replaced with the human HPRT gene for positive selection. The thimidine kinase (TK) gene from the herpes simplex virus was cloned outside the region of homology for negative selection. The resulting construct, with homogous arms of 5.1-kb and 1.3-kb, was used for homologous recombination in mouse embryonic stem (ES) cells.
158 clones expressing HPRT will survive and those expressing TK will not survive in the presence of 6-GT.
The construct was given to the Gene Targeting Core at the University of
Cincinnati. Approximately six weeks after targeting, 235 ES cell clones were isolated.
One half of the ES cells (Bl6) from each surviving clone were cyropreserved and the other half used to prepare DNA. From the 235 clones isolated at the targeting core, PCR
(Figure 31) and Southern blot analysis identified two that contained homologously integrated clones. ES cells from these clones were thawed and expanded in culture by
Core personnel and given to the Blastocyst Injection Core. The Core injected the targeted ES cells into 129/SvEv blastocysts. The injected blastocysts were then placed in pseudopregnant females. Pups born from these females were scored for coat color chimerism (black on agouti coat color indicating successful integration of injected ES cells into the blastocysts). Chimeric mice were bred to several strains of mice (Black
Swiss, 129/SvEv) in an effort to increase the chances of successful transmission of the knock-out construct to viable mice. DNAs from pups of these matings were screened for the presence of the targeted allele. The injection core repeated their procedures three times in an effort to increase the number of founder mice. Germline transmission was confirmed from a single chimeric mouse with the identification of four heterozygous mice in two litters.
With two female and two male heterozygous mice, matings were set up to breed the mutation to homozygosity. A formal study was initiated, examining the genotypes of pups born to matings of heterozygous mice. The expected ratio of wild-type to
159 non-targeted allele 10 11 12 end of construct targeted allele HPRT PCR Primers targeted allele 123456789C-
Figure 31. Embryonic stem cell screening for targeted alleles. ES cells were screened by PCR and Southern blot techniques for the homologous recombination of the knockout mouse construct. DNAs from ES cells that survived selection were prepared by standard techniques. For PCR screening, 200 ng of DNA was amplified using either a 5’ primer homologous to DNA in the HPRT gene or to DNA in intron 11 and a 3’ primer that was 3’ to the region of homology of the construct. All screening PCRs were repeated at least one time. Resulting ethydium bromide-stained bands on agarose gels indicated the presence of wild-type and knock-out alleles. Clones that were considered positive for homologous recombination of the construct by PCR were further screened by Southern blot. DNAs were restriction enzyme digested by PstI, resulting in identification either 4.1-kb bands in the wild-type allele or 10.4-kb bands in the targeted allele when the blot is probed with a 700 nt probe just 3’ to the region of homology.
160 heterozygous to homozygous offspring is 1:2:1, yet, a ratio of 1:2.3:0 after 270 pups was
recorded. The Blm-/- mouse was determined to be an embryonic lethal. Although the
embryonic lethality is disappointing, several studies on the Blm+/- mice have continued
with unprecedented results. A number of investigations have been completed, including
characterization of the embryos and primary cultures, and of the adult heterozygous mice.
In addition, experiments demonstrated an increased frequency of heterozygous mice to
develop spontaneous tumors, an increased sensitivity to MLV-induced tumors and an
accelerated onset and severity of tumors in a line of mice predisposed to intestinal tumors
due to a mutation in the murine Apc gene.
Embryo characterization.
In order to understand the role of BLM in utero, a study was initiated to identify
Blm-/- mouse embryos, and compare them to heterozygote and wild-type embryos.
Embryos were collected from heterozygous matings from day 7.5 to day 14.5 post-coitus.
Size, appearance, and viability were recorded for each embryo. Table 8 summarizes the data. Day 0.5 was considered the time when the cervical mucus plug first appeared in the female. In general, no difference between wild-type and heterozygous mice embryos was noted. However, Blm-/- mouse embryos were measurably smaller, paler, and more friable than litter-matched controls (Figure 32). In addition, no Blm-/- mouse embryos were
found viable after day 13.5 in utero, as defined by the presence of a heart beat, although
three embryos were identified at day E14.5.
Blm heterozygous Mice have an increased incidence of lymphoma.
161 Table 8. Percent genotype by Post-coital Age of Viable Embryos.
day1 10.5 11.5 12.52 13.52 14.52 15.52 genotype
Blm+/+ 28% 21% 31% 32% 30% 29%
Blm+/- 56% 60% 56% 54% 67% 71%
Blm-/- 16% 19% 13% 14% 4% 0
total3 77 100 90 98 88 52
1 day post coital 2 p<0.05 3 total number of mice examined at each particular day post-coital
162 10.5 dpc 12.5 dpc 14.5 dpc 1cm 1CM Blm+/-
Blm-/-
Figure 32. Blm-/- embryos are small and anemic. Embryos were harvested from heterozygous matings on the indicated day post coitus (dpc) for analysis. After phenotypic characterization, the embryos were genotyped by PCR and Southern blot. Blm-/- embryos were consistently smaller, paler, and more likely to be found non-viable at the time of harvest, than Blm+/- litter matched controls. No viable Blm-/- embryos were identified after 13.5 dpc.
163 Six of 38 heterozygous Blm mice spontaneously developed lymphoma during the
first six months of life. None of 29 wild-type littermates developed tumors over the same
period. Histologically, the lymphomas (Figure 33) were thymic in origin with other
organ metastases present. These tumors demonstrate CD8 expression by
immunohistochemisty, consistent with T-cell lymphoma. During this study it was noted
that Blm+/+ mice in the same background (Black Swiss) also developed lymphomas, resulting in decreased survival at one year. However, Blm heterozygosity is associated with more tumors at a younger age. There is a 1.8-fold increase in lymphoma incidence in Blm+/- mice as compared to Blm+/+ littermates in mice less than six months of age. At the time of this writing it is unknown whether or not the incidence of lymphoma changes over time in Blm+/- mice. The study is ongoing. In the six tumors examined for LOH of
Blm, the wild type allele remained. These results suggest a decrease in the expression of
Blm, is sufficient to promote tumor formation. Interestingly, this is not mechanism of
increased incidence of tumors noted in human heterozygotes.
Primary cultures from Blm heterozygous mice demonstrate chromosome instability.
We reasoned that if haploinsufficiency is resulting in increased tumor formation
in our mice, then genomic instability is a likely mechanism by which the tumors are
produced. Micronuclei and sister chromatid exchange are two methods of measuring
chromosome instability that have been applied to human BS cell lines. BS cells have
increased rates of SCE and more micronuclei as compared to normal control cell lines
(Chaganti et al., 1974 ; Rosin and German, 1985). Attempts were made to examine these
164 F4 +/- +/-
F5
+/+ +/-
H&E H&E α-CD8
+/- +/- +/+ Figure 33. Lymphoma in Blm mice. (A) F4 generation Blm mice were crossed to generate Blm and Blm+/- mice for comparison of tumor formation. (B) Blm+/- mice developed lymphoma during the study, where as none of the wild-type subjects developed tumors.
165 parameters in our Blm+/- mice. SCE analysis of Blm+/- mice and Blm+/+ mice were
inconclusive (data not shown), as a result of relatively high levels of SCE in all cells
examined. Micronuclei from 1,000 interphase nuclei from one wild-type primary lung
fibroblast culture and more than 1,000 interphase nuclei from each of two Blm+/- cultures
(two different mice) were evaluated (Figure 34). Micronuclei rates were 4.0 % for wild- type and 8.3 % for heterozygotes. These data suggest that these cells have increased genomic instability in response to a reduction in BLM expression.
Blm heterozygosity promotes ApcMin/+ intestinal tumorigenesis
To test the hypothesis, that Blm can be used as genetic tool to increase the rates and onset of tumor formation by increasing genomic instability, and consequently, LOH at key tumor suppressor loci, heterozygous Blm mice were mated with Min mice that
carry a mutation in the murine Apc gene (Figure 35). ApcMin/+ mice carry a chain-
terminating mutation in one allele of the Apc tumor suppressor gene (ApcMin), develop
multiple intestinal adenomas and are a mouse model of familial adenomatous polyposis
coli (Moser et al., 1990; Su et al., 1992). ApcMin mice are well characterized in terms of
mouse strain variations, tumors, and molecular mechanisms of tumorigenisis, making
them an appropriate choice for a background on which to evaluate Blm
haploinsufficiency. Blm+/- mice were crossed with ApcMin/+ mice and progeny were
sacrificed at 4 months of age. The entire gastrointestinal tract was opened longitudinally
and examined for polyps. Tumors were evaluated quantitatively by number (Figure 36)
and location, and qualitatively by size and histology.
166 Blm+/+
Blm+/-
Figure 34. Primary lung fibroblasts from Blm+/- mice have a two-fold increase in the percentage of cells with micronuclei. Representative cells from Blm+/+ (A) and Blm+/- (B) primary lung fibroblast cultures demonstrate characteristic micronuclei (arrows), although more are present in the Blm+/- cells. Micronuclei are visible microscopically as small, darkly stained bodies outside the nucleus in cells cultured in BrdU, incubated in Hoescht 33258, exposed to light, and stained with Giemsa (1000x). Arrowheads indicate a forming micronucleus (B, top right) and an abnormal chromosome protruding from a nucleus (B, bottom center). Quantification of these data is summarized in Table 1.
167 Blm+/- ApcMin/+
Blm+/+ Blm+/- Blm+/+ Blm+/- ApcMin/+ ApcMin/+ Apc+/+ Apc+/+
Figure 35. Mice lines used for studies of tumor induction. Blm+/- mice were mated to ApcMin/+ mice. ApcMin/+ mice are a well-characterized tumorigenic strain. The effect of Blm haploinsufficiency on the Apc background was tested by examining tumor number and grade in Blm+/-,ApcMin/+ offspring and comparing to those in Blm+/+,ApcMin/+.
168 70
60
50
40 x = 31.4 30 n=8 20
tumor number/mouse x = 14.2 10 n=14 0
ApcMin/+; ApcMin/+; Blm+/+ Blm+/-
Figure 36. Blm+/- mice develop more ApcMin/+-induced gastrointestinal tumors than Blm+/+ littermates. ApcMin/+ mice (C57BL6/J) were crossed with Blm+/- mice (backcrossed 129/SvEv to N3 generation). Offspring were sacrificed at 4 months of age for blinded analysis of tumors throughout the entire gastrointestinal tract. Means of the number of tumors per mouse are shown. Statistical analysis was performed with the non-parametric Wilcoxon Rank Sum test (p<0.05).
169 ApcMin/+,Blm+/+ mice developed an average 14.2 +/- 10.2 tumors per mouse (n =
14). A two-fold increase in gastrointestinal tumors was observed in ApcMin/+,Blm+/- mice
(31.4 +/- 19.1; n = 8) compared to ApcMin/+,Blm+/+ mice (p<0.05). As expected, the
majority of tumors were in the small intestine, with a twice as many tumors in the
ApcMin/+,Blm+/- (30.9 +/- 18.9) mice small intestine as compared to ApcMin/+,Blm+/+ (14.5
+/- 10.0) mice. In addition to small intestine polyps, colonic tumors were identified in four of the eight ApcMin/+, Blm+/- mice, whereas only 1 of the ApcMin/+, Blm+/+ mice
developed a colonic tumor. No tumors were identifed in mice with an Apc+/+ genotype.
All tumors from ApcMin/+, Blm+/+ and ApcMin/+, Blm+/- miceweregraded
histologically as low- or high-grade adenomas. Both genotypes developed intraepithelial
and low-grade adenomas in the small intestine (Figure 37). Only ApcMin/+; Blm+/- animals
developed high-grade neoplasms (5/223 tumors examined) in the small intestine.
Histologic evaluation of colonic tumors (one ApcMin/+, Blm+/+ and four ApcMin/+, Blm+/- ) indicated that all exhibited high-grade dysplasia.
To provide evidence of LOH in our mice, genomic regions both distal and proximal to the murine Apc gene were evaluated. The Apc gene in the mouse maps to
15cM on chromosome 18 (Justice et al., 1992; Luongo et al., 1993). PCR-based single-
stranded length polymorphism (SSLP) techniques were used to quantitate LOH at
D18Mit19 (proximal to Apc at 2.0 cM) and D18Mit17 (distal to Apc at 20.0 cM) (Figure
38A). Both normal and tumor tissue were assayed in each mouse from which the 18 tumors arose, so that the ratio of the 129/SvEv allele to the C57BL/6 allele could be calculated and used as a baseline. These data show that tumor formation in ApcMin/+ mice
are characterized by the loss of Apc and all other markers, both proximal and distal, on
170 A BC
ApcMin/+; T Blm+/+ N
D EF
ApcMin/+; T Blm+/-
N
Figure 37. Gross and histological analyses of ApcMin/+-induced tumors from Blm+/- and Blm+/+ littermates demonstrate that some Blm+/ - tumorsaremoredysplastic. (A-F) Small intestinal tumor from ApcMin/+;Blm+/+ (A-C)andApcMin/+;Blm+/- (D-F) mice. Arrows indicate individual adenomas in gross specimens. Sections are stained with hemotoxylin and eosin. Higher magnification (1000X) views of the same tumors are shown in C and F. A and D, 10X; B and E, 100X.
171 A B Blm+/+ Blm+/- B6 129 TN TNTN
C57BL/6 D18Mit19 129/SvEv Apc (15.0 cM) D18Mit17 (20.0 cM) ApcMin- C57BL/6 Apc Apc+- 129/SvEv
C57BL/6
D18Mit17 129/SvEv
Figure 38. Molecular analysis of adenomas from ApcMin/+;Blm+/- and ApcMin/+;Blm+/+ mice. A. Schematic diagram of mouse chromosome 18 with the genetic location of three microsatellite markers and the Apc gene. Distances from the centromere in cM (black circle) are shown in parentheses.B. Representative genotyping analysis of Apc and chromosome 18 microsatellite markers D18Mit17 and D18Mit123 in DNA from microdissected adenoma (T) and adjacent normal tissue (N). The pair of samples on the left is from an ApcMin/+;Blm+/+ mouse; the pairs in the middle and on the left are from two different ApcMin/+;Blm+/- mice. Tail DNA from C57BL/6 (B6) and 129/SvEv (129) mice are included as controls. Bands corresponding to the C57BL/6 and 129/SvEv alleles are highlighted with arrows. C. PCR/Southern analysis of Blm in DNA microdissected adenoma (T) and adjacent normal tissue (N) samples. Bands representing the wild-type (400 bp) and targeted (780 bp) alleles are shown. All three tumor/normal tissue pairs are from one ApcMin/+;Blm+/- mouse.
172 chromosome 18 (Figure 38B). Tumor development in ApcMin/+, Blm+/- mice was similar
with 13 of 14 tumors showing loss of the wild-type Apc allele and proximal and distal
markers. One of the ApcMin/+, Blm+/- tumors exhibited LOH at Apc and at the distal marker, D18Mit17, yet remained heterozygous at the proximal marker, D18Mit19.A second tumor remained heterozygous proximal and distal to Apc. The mechanism of Apc loss in ApcMin/+, Blm+/+ mice is consistent with isodisomy or chromosomal loss, but
somatic recombination between Apc and D18Mit19 occurred in one tumor sample with
Blm haploinsufficiency.
Finally, a quantitative analysis, similar to that done on the Apc allele, was
conducted for the Blm allele. There was no significant difference between normal and
tumor sample in the 14 tumors assayed from ApcMin/+, Blm+/- mice. These results indicate
that loss of Blm is not required for the increase in tumor formation in ApcMin/+, Blm+/-
mice and haploinsufficiency in Blm is adequate to promote Apc-mediated tumorigenesis.
Discussion.
We have created a mouse model of the human disease Bloom’s syndrome by replacing exons 10, 11, and 12 of the murine Blm gene with the HPRT gene. Genotyping of pups born to matings of heterozygous females to heterozygous males revealed that the homozygous Blm knockout state is not viable. While heterozygous embryos are grossly similar to their Blm+/+ littermates, Blm-/- littermates are significantly smaller, paler, and more friable. Although the heterozygous mice originally appeared similar to the normal littermates, differences have been identified. Blm+/- mice have a lower long-term survival
and an increase incidence of lymphoma. We also demonstrated that primary cultured
173 cells from heterozygous Blm mice have a two-fold increase in genomic instability over their littermates, based on a comparison of micronuclei formation. Finally, through a series of experiments where Blm heterozygosity was bred onto the ApcMin/+ background,
we demonstrated that Blm haploinsufficiency could promote tumor formation.
In human BS cells, there is ample evidence of genomic instability, including elevated rates of SCE and micronuclei. Our haploinsufficient Blm mouse model recapitulates the genomic instability, as determined by micronuclei of primary cell cultures. Micronuclei are significantly elevated in Blm+/- mice. The SCE data were not
significant, due to a broad range of SCE levels in all cells. This may be attributable, at
least in part, to the relatively high baseline rates of SCE in the C57BL/6 mouse strain
(Nishi. et al. 1993). The possibility exists that haploinsufficiency leads to a level of Blm
that is adequate for maintaining normal or near normal rates of SCE, yet allows other
mechanisms of genomic instability to progress, resulting in increases in micronuclei.
These considerations may provide some insight to the functional pathways in which BLM
is involved and which have priority within the cell.
Blm haploinsufficiency results in decreased long-term survival. Heterozygous animals have a higher incidence of spontaneous T-cell lymphoma than wild-type mice.
Many of these tumors develop at a younger age than wild-type mice. Lymphoma in wild- type mice greater than one year of age has been well documented (Frith et al., 1993).
The incidence of lymphoma can be increased by the loss of any of several tumor
suppressors, including p53, Atm,andMlh1 (French et al., 2001; Liao, M.J. et al., 1999;
Kawate et al. 2000). The loss of these genes results in an increase in chromosomal
instability, an important consideration in recombination-sensitive tissue, including
174 lymphocytes and germ cells. Blm can be considered in this group of tumor suppressors,
that when mutated, even by disruption of one allele in mice, increases the incidence of
lymphoma. Furthermore, lymphoma is a common malignancy in BS patients. A
decrease in the quantity of BLM may negatively affect the genomic stability in a cell type
that is dependent on a complex series of recombinations.
In addition to an increased incidence of spontaneous lymphoma, we report that
thepresenceoftheBlm+/- genotype is associated with increased number and higher
histologic grade of tumors in mice predisposed to develop tumors. Mice with the Blm+/-,
ApcMin/+ genotypes have both an increased number and more severe histologic grade of
intestinal adenomas than ApcMin/+ mice. Further affects of Blm haploinsufficiency have
been noted with half of the Blm+/-, ApcMin/+ mice developing colonic adenomas, compared
to only 7.1% of the Blm+/+; ApcMin/+ mice. Most of the tumors in both genotypes of animals were located in the small intestine. Tumors were analyzed for LOH at both the
Blm and Apc loci. LOH of Apc is seen in tumors from ApcMin/+ mice. As one may
suspect, Blm+/-, ApcMin/+ tumors showed LOH at Apc in all adenomas examined. None of
the tumors exhibited LOH at Blm. To characterize the mechanism leading to the LOH at
the Apc locus, proximal and distal polymorphic markers were analyzed. All 14 adenomas except one lost heterozygosity distal to Apc. Interestingly, two of the 14 adenomas
remained heterozygous proximal to Apc. These samples provide evidence of somatic
recombination between Apc and the proximal marker in the presence of Blm
haploinsufficiency. It has been documented that there is an increase in somatic
recombinations in lymphoblastiod cell lines of BS patients (Groden et al., 1990). With
an increase in tumor number in Blm+/-, ApcMin/+ mice, we may speculate that some of the
175 other Blm+/-, ApcMin/+ samples may have undergone multiple recombination events leading to the LOH both proximal and distal to Apc, our techniques were not sufficiently sensitive to provide those data. Based on data presented here, Figure 39 describes a model of a mechanism by which Blm haploinsufficiency promotes tumor formation.
When this mouse model for BS was conceived, we had envisioned that it could
“induce” recombination events, increase LOH, and increase tumor formation through loss of yet to be identified tumor suppressor genes. Although the evaluation of our model is not complete, we have made significant progress toward our original goal. The heterozygous Blm mouse has many interesting characteristics, spontaneous lymphoma formation, an increase in intestinal adenoma number and histiologic grade on the Min mouse background, and evidence of increased genomic instability and somatic recombination. Additional studies will be necessary to isolate other loci and genes that are involved in the formation of specific tumor types through the expected induction of
LOH. Although other mechanisms of tumor formation are possible, LOH is arguably an important potential mechanism when high rates of recombination are induced. Further studies to characterize the molecular mechanism of our results, comparing haploinsufficiency to dominant negative effects are required. Lastly, the Blm-/- animals were not viable, leaving one to speculate that our model, although interesting in its own right, may not properly mimic the disorder in humans. A tissue specific knockout of Blm is underway.
176 A Loss of Apc
Apc+/- -/- Blm+/+ Apc Blm+/+
B Mutation Additional Additional Recombination Mutations Mutations Repair
Loss of Apc +/- Apc Apc-/- Apc-/- +/- Blm Blm+/- Blm+/- Normal Epithelium Low-grade Adenoma High-grade Adenoma Carcinoma
Figure 39. Model of the mechanism by which Blm haploinsufficiency promotes tumorigenesis. A. The rate-limiting step in intestinal tumor initiation in ApcMin/+ mice is loss of the wild-type Apc allele through chromosomal loss of the Apc+ chromosome. Acquired mutations at other loci, including Dpc4 (Smad4), p53, K-ras, may be necessary to promote these tumors to malignancy. B. Haploinsufficiency of Blm in ApcMin/+ mice creates an environment of genomic instability, presumably by subtly affecting somatic recombination and mutation rates, which increases the likelihood of loss of critical tumor suppressors, such as Apc. Additionally, decreases in DNA repair capacity or increases in point mutation frequency may promote tumor progression. Development of spontaneous lymphoma in Blm heterozygous mice would presumably occur through a similar mechanism, although the mutational targets are different.
177 Chapter 11. Discussion.
BS is a rare autosomal recessive disease characterized by facial erythema, sun
sensitivity, small stature, immunodeficiency, male sterility, female subfertility, and
predisposition to cancer. At the cellular level various aspects of genomic instability are
apparent. BS cells exhibit an elevated rate of SCE. BS cells also have been reported to
have an increase LOH in somatic cells (Groden et al., 1990). These observations served
as preliminary evidence and a starting point for the research present here. One plausible
explanation for these observations is an increase recombination in BS cells, which
directly lead to the hypothesis for this body of work that BLM is a suppressor of
recombination.
The information presented in this dissertation is presented chronologically and
includes microsatellite repeat instability in BS cells, the mapping of the BS region, the
identification of BLM, its genomic structure, and a mouse model of BS. While each of these areas discussed is intended to provide evidence that supports the hypothesis, some of the data is suggestive of the molecular mechanisms in which BLM functions to suppress recombination.
Genomic instability has been demonstrated in BS cells by SCE, LOH, and somatic hypermutability. Microsatellite repeat instability can now be added to this list.
Microsatellite instability in BS secondary cell lines is increased as compared to secondary cell lines derived from a normal individual. Several primary EBV-transformed lymphocytes from a BS and a normal individual were isolated and grown clonally in
178 tissue culture as primary cell lines. Several secondary cell lines from both normal and BS cells were created by isolating and clonally expanding individual cells from a single primary clone. Secondary cell lines were cultured for approximately 12 months before being collected and DNAs prepared. DNAs were analyzed by electrophoresis of PCR- amplified microsatellite repeat loci. The results indicate that microsatellite repeat instability is 12-fold more frequent in BS cell lines than normal cell lines. One or more molecular mechanisms may be responsible for the increase in microsatellite repeat instability. Mutations in the mismatch repair proteins increase microsatellite repeat instability and formation of colonic tumors. Unfortunately, no association between BLM and single base mismatch repair has been identified in several experiments. Unrepaired polymerase slippage during DNA replication is another mechanism leading to alteration in the number of repeats at microsatellite repeat loci. Another is that recombination events are permitted to take place at regions of microhomology, such as microsatellite repeat loci, which may lead to unequal exhanges in the absence of BLM. It has been reported that recombinations in microhomologous regions do occur in E. coli in the absence of RecQ, but are suppressed when RecQ is present (Hanada et al., 1997). These data argue that BLM functions as a suppressor of recombination.
Identification of BLM
To identify the causative defect in BS, candidate gene approaches were undertaken by numerous groups of researchers. Many conflicting results were obtained with no candidate gene consistently supported by the data. Ultimately, with access to a large patient database through the BS Registry, our group decided positional cloning
179 techniques were an appropriate method for identifying BLM. One of the early steps in
positional cloning was to create a physical map of the genomic region identified by
linkage analysis. Linkage analysis showed that the gene for BLM was tightly linked to
the protoncogene FES at 15q26.1. A physical map composed of YACs, P1s, and cosmids
created a 2-Mb contig.
Since recombination events occur frequently in BS cells relative to normal cells, it
was postulated that the hyperrecombination could be used to our advantage in the search
for BLM. Individuals that inherit a unique mutant copy of BLM from each parent, ie.
compound heterozygotes at BLM, may undergo homologous recombination between the
two BLM alleles in some somatic cells. Cells that undergo this type of recombination
would be predicted to express BLM and have a normal phenotype. Some cell lines
derived from BS patients do indeed have a subpopulation of normal cell, based on SCE
levels. By examining polymorphic repeat loci that mapped to the 15q26.1 region in cell
lines that had undergone a BLM intragenic recombination (ie. low SCE) and comparing the results to cell lines from the same individual that had not undergone that recombination event (ie. high SCE), it was possible to determine a 250-kb region in which BLM resided. Components of the physical map that contained the 250-kb region were screened for candidate genes by hybridization to cDNA libraries. The first candidate gene, a 4437 nt cDNA, had mutations in the coding sequence in seven of the first 10 BS patients examined. With the identification of BLM, additional characterizations of the gene and protein were possible. The amino acid sequence indicates that BLM has homology to DNA helicases and is a member of the RecQ helicase family.
180 Genomic Structure of BLM
To continue the characterization of BLM, we first defined its genomic structure.
BLM is composed of 22 exons and includes a first exon that is composed of 70
nucleotides. of untranslated sequence located more than 29.5-kb upstream of exon 2. The
sequence of the introns were determined and demonstrated that every intron begins and
ends with the standard dinucleotide sequence GT and AG, respectively. In total, the BLM
locus is 105.7-kb. The genomic structure and intron sequence are important tools for
mutational analyses, while the cDNA will enable studies of the function of the BLM
protein to be performed.
Mouse Model of BS
As part of the functional investigation of the protein, a mouse model of BS was
created. This model was intended as resource for the study of tumor formation, to
identify novel tumor suppressors or modifiers. The gene-targeting construct was
designed to approximate the most common mutation in BS patients, a frame shift at
nucleotide 2281 in the cDNA, known as the BLMAsh. This mutation causes a stop codon in exon 10, due to a 6-bp deletion and 7-bp insertion. The construct did not contain exons 10, 11, and 12, which were replaced with the marker gene HPRT. Initial studies of the resulting mouse model, demonstrated that homozygosity for this mutant allele resulted in embryonic lethality. Initially, no phenotypic differences were identified between normal and heterozygous mice. While establishing lines of mice with homogenous genetic background for further characterization, several heterozygous mice
181 developed lymphoma. Several studies using the heterozygotes were subsequently undertaken to evaluate the potential of Blm+/- mice to develop tumors.
During the first six months of life, Blm+/- mice demonstrated a significant increase in lymphoma as compared to wild type mice. In addition, when Blm+/- mice are bred to
ApcMin/+ mice, the resultant ApcMin/+,Blm+/- mice develop twice as many tumors as the
ApcMin/+,Blm+/+ mice. Most of the tumors are located in the small intestine, characteristic of ApcMin/+ mice. However, four colonic tumors developed in ApcMin/+,Blm+/- mice, while, only one was identified in the ApcMin/+,Blm+/+ mice. The histologic grading of all tumors revealed some with high-grade dysplasia in the small intestine in the double heterozygotes and no tumors with high-grade dysplasia in the ApcMin/+ mice. All colonic tumors in both genotypes of mice displayed high-grade dysplasia.
One explanation for the increase in tumor number and grade is haploinsufficiency of Blm in ApcMin/+,Blm+/- mice. All tumors examined in these mice maintained at least one normal allele of Blm. The heterozygous mice may be producing inadequate amounts of Blm protein which in turn increases genomic instability just enough to lead to a tumorigenic phenotype. The prediction that an increase in recombination and LOH at tumor suppressor loci will be associated with an increase of tumor numbers in haploinsufficient mice was supported by our experimental results. This mouse model of
BS provides the research community with a tool for identifying novel genes that influence tumor formation.
Our current BS mouse model has provided some interesting results and is still worthy of further investigations, since there is an effect on tumor predisposition in the heterozygotes. More dramatic results, potentially leading to faster and more complete
182 understanding of some of the molecular pathways of tumorigenesis, can be approached
through tissue-specific or inducible mouse models. A tissue-specific knock-out, using
the Cre-LoxP system is one possible solution to working with heterozygous mice. In this
system, short specific directional sequences are used to flank the gene to be knocked out.
A transgenic animal with a tissue-specific promoter expresses the viral LoxP sequence
specific recombinase, Cre. Breeding the transgenic Cre mouse to the heterozygous LoxP
mouse would then provide offspring that are Blm-/- in specific tissues. A lymphoid-
specific promoter driving Cre is one example of this experiment, since persons with BS
are immunodeficient and frequently develop lymphomas. In addition, the formation of
lymphoma noted in the heterozygous mice, may be accelerated to provide more tumor
samples.
Another alternative to the tissue-specific mouse model is one that employs a
molecular switch to turn the promoter of Blm on or off as desired. One such system uses tetracycline, or more commonly now, doxycylcine, to turn the gene of interest on or off, depending on system design (Furth et al., 1994). I would expect it would be more appropriate to design this model so that adding doxycylcine to the diet of mice would turn off the expression of Blm. By choosing this option, the mice are expected to be viable without any dietary additive. The doxycylcine dosing of the mice can be varied in time and amount to induce a phenotype. The tet-inducible system has been characteristically leaky, so that in the system described here even with high levels of doxycylcine there is likely to be some minimal expression of Blm. The question then becomes how much Blm is necessary to prevent a phenotype. I would argue that the tet-
183 off system may still be useful, particularly since the heterozygous mice are viable, yet
develop tumors.
Tissue Culture Recombination and Frame-shift Experiments
The research presented here leaves numerous open questions and topics for future
investigation. The role of BLM in mammalian cell lines was initially investigated with
indicator cell lines designed to identify recombination and frame-shifting events. The
first cell line contained the two nonfunctional LacZ genes separated by a few kilobases of
DNA. One LacZ gene was normal in sequence, but lacked a promoter; while the second included a promoter but contained a nonsense mutation near the 5’ end. A homologous recombination event between the two LacZ genes, either serially or via SCE, produces a functional LacZ gene. The presence of a functional LacZ gene was evaluated by a simple colorimetric assay. The second cell line contained an altered transgenic alkaline phosphatase gene driven by the RSV promoter. The normal alkaline phosphatase gene contains four consecutive guanines near the 5’ ATG. This was mutated to 11 guanines,
changing the reading frame and preventing functional protein. However, a frame-shift
event by –1 or +2 would restore the reading frame of the gene, producing functional
protein. A colorimeteric assay can be used to score cells that contain functional alkaline
phosphatase. Over expression of BLM did not affect frame-shifting rates. Several
possible conclusions can be drawn from these data. Human BLM may not be able to
function in green monkey kidney cells. The conditions in which BLM functions may not
occur with single nucleotide repeat sequences. Another possible conclusion is that BLM
is not involved in frame-shifting, although dinucleotide and trinucleotide microsatellite
184 repeat studies indicate otherwise. Although our methods were not successful, a tissue culture approach to evaluating the effect of BLM on recombination continues to be worthwhile. The approach could be altered slightly to make it more relevant by starting with a human BS cell line and transfecting in the marker gene so cells in which BLM is expressed can be compared to those without BLM. A gene that would indicate recombination in living cells would also be useful. An approach with the enhanced green fluorescent protein (EGFP) in place of LacZ, would allow evaluation of these cells in real-time. In addition, such cells could be counted and sorted on a FACS machine for further assays in tissue culture, based on whether or not they have exhibited recombination.
BLM Protein Interactions
Numerous experiments can be done with purified BLM protein. These include evaluation of protein-protein interactions with novel and candidate proteins and determining how BLM influences the function of proteins with which it is already known to interact, including p53, MLH1, and RAD51. MLH1 interacts directly with BLM through the carboxy-terminal segment of BLM. MLH1 is a well-characterized protein in
DNA mismatch repair, although we have demonstrated that mismatch repair functions normally in BS cells. MLH1 also plays a key role in meiosis at synaptonemal complexes.
It remains to be seen whether BLM plays any part in meiotic pathways, although, this could be predicted by the clinical phenotypes of BS that include male sterility and female subfertility.
185 Another experiment to test the role of BLM in recombination is an in vitro strand invasion and exchange assay. It has recently been shown that RAD51 interacts with
BLM (Wu et al., 2001). RAD51 is a key component in homologous recombination.
Using an in vitro system, it has been demonstrated that RAD51 forms a nucleoprotein filament on single stranded DNA during the initiation of the strand invasion process
(Sung and Robberson, 1995). The use of a similar system would be an appropriate first step to investigating the role of BLM in the initial phases of recombination. The quantitative and temporal efficiencies of strand invasion and exchange could be evaluated with and without BLM. A number of other proteins involved in recombination, RPA,
RAD52, RAD55, RAD57, and RAD59 could also be added to this assay.
Models of BLM Function.
BLM is a helicase in the RecQ family that is capable of unwinding double- stranded DNA. The precise cellular function of BLM has yet to be clarified. Recent research, however, is suggestive. It has been noted that BS cells have an elevated rate of
SCE. BLM directly interacts with RAD51, RPA, topoisomerase IIIα, MLH1, and p53.
BLM is organized in nuclear foci and present in multimeric complexes that contain numerous proteins involved in recombination and repair, including BRCA1, MLH1,
MSH2, MSH6, ATM, RAD50, MRE11, and Nbs. I would like to propose a model of
BLM function (Figure 40) to organize the data on the association of BLM with these proteins and the phenotype of persons, cells, and mice lacking BLM.
DNA damage such as cyclobutane pyrimidine dimers, abasic sites, and backbone nicks are repaired through the base excision repair (BER) pathway. During G1-S-phase
186 DNA damage
BER
Figure 40. Model of BLM function. During S-phase leading and lagging daughter strands are synthesized. Damage affecting DNA a sufficient distance ahead of the replication complex (gray circle) is appropriately repaired through the base excision repair (BER) pathway. DNA damage incurred on the lagging strand that is too close to the replication complex, whether it is ahead of the replication complex in a region of dsDNA or after strand separation in the loop of ssDNA, is not repaired by BER. The replication complex may stall, back track, and allow the BER to repair the lesion ahead of the replication fork, or it may leap over the lesion, installing the next Okazaki fragment (blue bars). Similarly, a lesion on the ssDNA lagging strand loop is bypassed as the next Okazaki fragment is placed. As the replication complex synthesized the lagging strand, it comes upon a gap in the ssDNA. Since there is no way for the replication machinery to know whether this is damage to the ssDNA is a nick or a gap, fidelity of the repair requires the use of a DSB repair mechanism. The sister chromatid serves as a convenient and appropriate template for DSB repair. The replication complex releases the DSB and continues synthesis at the replication fork.
187 of the cell cycle, this repair continues, induced by DNA damage and the p53 response.
However, under certain circumstances, such as DNA damage that is too close to the replication complex, unusual DNA conformations, lack of BER resources, or lagging strand nick after strand separation, the BER machinery may fail to repair the defect. The replication fork jumps over the damage on the lagging DNA strand in order to initiate the next Okazaki fragment. As the nucleotides are filled in complementary to the lagging strand, the replication machinery comes upon the nick and stalls on the current lagging strand before initiating another Okazaki fragment. At this point, one sister chromatid has a nick on one strand and a gap (from the complementary nick to the preceding Okazaki fragment) on the other strand, essentially a double stranded break (DSB) in the DNA. An accurate and efficient mechanism for the cell to repair this DSB is to use the intact sister chromatid as a template for the deficient one, using repair by homologous recombination
(Johnson and Jasin, 2000). In normal cells, the DSB is repair through gene conversion, without DNA recombination. BLM prevents movement of the recombination intermediate, the Holliday junction, downstream from the site of repair (Figure 41). If
BLM is absent, the Holliday junction is free to migrate. In this case, the Holliday junction may translocate until it reaches an impassible DNA conformation or protein bound to the DNA. This translocation may be passive, seeking the lowest energy or active, receiving assistance from helicases that function opposite of BLM. After the translocation, the DNA intermediate is resolved, resulting in SCE.
This model would predict that BLM associates with other proteins involved in recombination, as well as those proteins involved in replication and some involved in repair. The association of BLM with mismatch repair proteins would not be
188 A
B
Sister Chromatide Exchange
C
Suppression of Recombination
Figure 41. BLM functions to suppress recombination during DSB repair. A. DSBs are processed with the potential for recombination through the formation of a Holliday junction, a recombination intermediate. B. In the absence of BLM, Holliday junctions may translocate along the DNA, leading to recombination with the resolution of the Holliday junction. C. BLM functions to suppress recombination by unwinding dsDNA interactions between recombinant strands of DNA during the DSB repair. Resolution of the Holliday junction in the presence of BLM results in no recombination.
189 inappropriate based on this model. The mismatch repair system follows DNA replication to correct any base mismatch errors incurred by the replicating DNA polymerase. BLM and the recombinational repair pathway could operate in a similar temporal and spacial manner, forming a complementary repair mechanism for various damage incurred or missed by the DNA polymerase.
The model also provides some explanation for the BS phenotypes. As noted above, it addresses the elevated SCE characteristic of BS cells. Extensions of this model support other phenotypes associated with BS, including microsatellite repeat instability and cancer. If the homologous recombinational repair pathway is active at other times during the cell cycle, G0-G1 for example, there is no available sister chromatid for recombination. I propose that the homologous chromosome becomes the template for homologous recombination during DSB repair. However, BLM maintains the same function, limiting the interaction of homologous chromosomes to repair only, suppressing recombination.
Why is it important to suppress recombination in somatic cells? If all recombination events were perfectly homologous, one could argue that suppression of this type of DNA manipulation is unnecessary. The fact that BLM exists and BS patients are highly predisposed to cancer supports arguments that many homologous recombination events are either not resolved equally or not initiated with sufficient homology to ensure accurate repair of DSBs. If rather short regions of DNA homology are sufficient to initiate DSB repair by homologous recombination, the template may not actually be the homologous region to the DSB. If recombinational repair is allowed to proceed, there is a potential for errors in repair. If this occurs in regions of
190 microhomology such as microsatellites, then the resulting repair would result in novel number of repeats at this locus. A similar event within a gene could bring about a mutation. In the expansion of this model, BLM operates in G0 and G1,aswellasinS- phase. There is an assumption that recombination has some degree of inherent illegitimacy and that BLM is required to suppress recombination to prevent accumulation of mutations.
This model suggests numerous experimental directions. One area is the role of
BLM during the cell cycle. Some of this work has been started by others and has shown that BLM is modified throughout the cell cycle. It remains to be seen if BLM is exclusively active during DNA synthesis, or is there a low level of function throughout the cell cycle. Additional experiments would focus on the proposed correlation between
DNA damage on the lagging strand and the initiation of homologous recombination post- replication during S phase.
New discoveries of protein partners of BLM have been made recently. The details of these interactions, as well as probing for novel interactions, will help in our understanding of the functions of BLM. BLM may interact with various members of the replication complex, as has been demonstrated with members of the mismatch repair system. Finally, this work would provide a better understanding of homologous recombination in mammalian cells.
This model of BLM function has provided a way of integrating my thoughts about
BLM. I had focused much of my research and reading on recombination, but in an attempt to create a comprehensive model for BLM, I believe its function is more involved than in mere suppression of recombination. This model proposes that BLM functions
191 during replication as part of a mechanism to repair DNA damage through recombination.
Therefore, it is not that BLM functions independently in repair, recombination, or replication, but it likely functions simultaneously in repair, recombination, and replication.
192 Literature Cited
Abdel-Monem, M., Chanal, M.C., and Hoffmann-Berling, H. 1976. DNA unwinding enzyme II of Escherichia coli. Purification and characterization of the ATPase activity.
Eur. J. Biochem. 79:33-8.
Ahmed, F.E. and Setlow, R.B. 1978. Excision repair in ataxia telangiectasia, Fanconi's anemia, Cockayne syndrome, and Bloom's syndrome after treatment with ultraviolet radiation and N-acetoxy-2-acetylaminofluorene. Biochim. Biophys. Acta. 521:805-17.
Arai, K and Kornberg, A. 1981. Mechanism of dnaB protein action. III. Allosteric role of
ATP in the alteration of DNA structure by dnaB protein in priming replication.
J. Biol. Chem. 256:5260-6.
Alhadeff, B., Velivasakis, M., Pagan-Charry, I., Wright, W.C., and Siniscalco, M. 1980.
High rate of sister chromatid exchanges of Bloom's syndrome chromosomes is corrected in rodent human somatic cell hybrids. Cytogenet. Cell Genet. 27:8-23.
Aurias, A., Antoine, J.L., Assathiany, R., Odievre, M., and Dutrillaux, B. 1985. Radiation sensitivity of Bloom's syndrome lymphocytes during S and G2 phases. Cancer Genet.
Cytogenet. 16:131-6.
193 Bamezai, R. and Shiraishi, Y. 1987. Three-way differentiation of sister chromatids in
endoreduplicated (M3) chromosomes of Bloom syndrome B-lymphoid cell line. Hum.
Genet. 75:239-43.
Barnes, D.E., Johnston, L.H., Kodama, K, Tomkinson, A.E., Lasko, D.D., and Lindahl,
T. 1990. Human DNA ligase I cDNA: cloning and functional expression in
Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. 87:6679-83.
Bartram, C.R., Rudiger, H.W., and Passarge, E. Frequency of sister chromatid exchanges in Bloom syndrome fibroblasts reduced by cocultivation with normal cells. Hum. Genet.
46:331-4.
Bennett, R.J., Sharp, J.A., and Wang, J.C. 1998. Purification and characterization of the
Sgs1 DNA helicase activity of Saccharomyces cerevisiae. J. Biol. Chem. 273:9644-50.
Bennett, S.E., Umar, A., Oshima, J., Monnat, R.J. Jr., and Kunkel, T.A. 1997. Mismatch repair in extracts of Werner syndrome cell lines. Cancer Res. 57:2956-60.
Berger, I. and Cohen, A. 1989. Suppression of recA deficiency in plasmid recombination by bacteriophage lambda beta protein in RecBCD- ExoI- Escherichia coli cells. J.
Bacteriol. 171:3523-9.
194 Bertazzoni, U., Scovassi, A.I., Stefanini, M., Giulotto, E., Spadari, S., Pedrini, A.M.
1978. DNA polymerases alpha beta and gamma in inherited diseases affecting DNA repair. Nucleic Acids Res. 5:2189-96.
Bloom, D. 1954. Congenital telangiectatic erythema resembling lupus erythematosus in dwarfs. Amer. J. Dis. Chil. 88:754-8.
Bloom, D. 1966. The syndrome of congenital telangiectatic erythema and stunted growth:
Observations and studies. J. Pediat. 68:103-13.
Bloom, A.D. 1975. Genetics and cytogenetics of ataxia telangiectasia and Bloom's syndrome. Mod. Probl. Paediatr. 17:6-10.
Bourgeois, C.A., Calverley, M.H., Forman, L., and Polani, P.E. 1975. Bloom's syndrome: A probable new case with cytogenetic findings. J. Med. Genet. 12:423-5.
Boyer, J.C., Umar, A. Risinger, J.I., Lipford, R., Kane, M., Yin, S., Barrett, J.C.,
Kolodner, R.D., and Kunkel, T.A. 1995. Microsatellite instability, mismatch repair deficiency and genetic defects in human cancer cell lines. Cancer Res. 55:6063-70.
Bronner, C.E., Baker, S.M., Morrison, P.T., Warren, G., Smith, L.G., Lescoe, M.K.,
Kane, M., Earabino, C., Lipford, J., and Lindblom, A., et al. 1994. Mutation in the DNA
195 mismatch repair gene homologue hMLH1 is associated with hereditary non-polyposis colon cancer. Nature. 368:258-61.
Brothman, A.R., Cram, L.S., Brothman, L.J., and Kraemer, P.M. 1987. Cultured Bloom's syndrome substrains: A relationship between growth in low serum and the expression of double minute chromosomes. Cancer Genet. Cytogenet. 26:287-97.
Brown, T.C., Zbinden, I., Cerutti, P.A., and Jiricny, J. 1989. Modified SV40 for analysis of mismatch repair in simian and human cells. Mutat. Res. 220:115-23.
Brumback, R.A. 1984. Cockayne's syndrome diagnosis. Neurology 34:842-3.
Brunsting, L.A. 1957. Congenital telangiectasia of the face; ichthyosiform erythroderma?; mild pituitary dwarfism (Levi-Loraine type). Arch. Derm. 75:771-3.
Bryant, E.M., Hoehn, H., and Martin, G.M. 1979. Normalisation of sister chromatid exchange frequencies in Bloom's syndrome by euploid cell hybridisation. Nature
279:795-6.
Carlson, M. and Laurent, B.C. 1994. The SNF/SWI family of global transcriptional activators. Curr. Opin. Cell Biol. 6:396-402.
196 Chaganti, R.S., Schonberg, S., and German, J. 1974. A manyfold increase in sister chromatid exchanges in Bloom's syndrome lymphocytes. Proc. Natl. Acad. Sci. 71:4508-
12.
Chan, J.Y., Becker, F.F. 1988. Defective DNA ligase I in Bloom's syndrome cells.
Simultaneous analysis using immunoblotting and the ligase-[32P]AMP adduct assay. J.
Biol. Chem. 263:18231-5.
Chan, J.Y., Becker, F.F., German, J., and Ray, J.H. 1987. Altered DNA ligase I activity in Bloom's syndrome cells. Nature 325:357-9.
Chao, K.L. and Lohman, T.M. 1991. DNA-induced dimerization of the Escherichia coli
Rep helicase. J. Mol. Biol. 221:1165-81.
Chester, N., Kuo, F., Kozak, C., O'Hara, C.D., and Leder, P. 1998. Stage-specific apoptosis, developmental delay, and embryonic lethality in mice homozygous for a targeted disruption in the murine Bloom's syndrome gene. Genes Dev. 12:3382-93.
Chisholm, C.A., Bray, M.J., and Karns L.B. 2001. Successful pregnancy in a woman with
Bloom syndrome. Am. J. Med. Genet. 102:136-8.
Cote, J., Quinn, J., Workman, J.L., and Peterson, C.L. 1994. Stimulation of GAL4 derivative binding to nucleosomal DNA by the yeast SWI/SNF complex. Science 265:53-
197 60.
Cunningham, R.P., Shibata, T., DasGupta, C., and Radding, C.M. 1979. Single strands induce recA protein to unwind duplex DNA for homologous pairing. Nature 28:191-5.
Curth, H.O. 1964. Blooom’s syndrome (congenital telangiectatic erythema resembling lupus erythematosus in Loraine-Levi dwarfs): case presentation. Arch. Derm. 90:349-50.
Dallapiccola, B. 1987. Cytogenetics of Mendelian mutations associated with cancer proneness. Cancer Genet. Cytogenet. 26:85-94.
Das, R.H., Yarranton, G.T., and Gefter, M.L. 1980. Enzyme-catalyzed DNA unwinding.
The role of ATP in helicase III activity. J. Biol. Chem. 255:8069-73.
Davey, S., Han, C.S., Ramer, S.A., Klassen, J.C., Jacobson, A., Eisenberger, A., Hopkins,
K.M., Lieberman, H.B., and Freyer, G.A. 1998. Fission yeast rad12+ regulates cell cycle checkpoint control and is homologous to the Bloom's syndrome disease gene. Mol. Cell
Biol. 5:2721-8.
Debreceni N, Behme MT, Ebisuzaki K. 1970. A DNA-dependent ATPase from E. coli infected with bacteriophaget4. Biochem. Biophys. Res. Commun. 41:115-21.
198 Dehazya, P. and Sirover, M.A. 1986. Regulation of hypoxanthine DNA glycosylase in
normal human and Bloom's syndrome fibroblasts. Cancer Res. 46:3756-61.
DePrimo SE, Cao J, Hersh MN, Stringer JR. Use of human placental alkaline
phosphatase transgenes to detect somatic mutation in mice in situ. Methods. 1998
6:49-61.
Egelman, H.H., Yu, X., Wild, R., Hingorani, M.M., and Patel, S.S. 1995. Bacteriophage
T7 helicase/primase proteins form rings around single-stranded DNA that suggest a
general structure for hexameric helicases. Proc. Natl. Acad. Sci. 92:3869-73.
Eggleston, A.K. and West, S.C. 2000. Cleavage of Holliday junctions by the Escherichia
coli RuvABC complex. J. Biol Chem. 275:26467-76.
Ellis, N.A. 1997. DNA helicases in inherited human disorders. Curr. Opin. Genet. Dev.
7:354-63.
Ellis, N.A., Lennon, D.J., Proytcheva, M., Alhadeff, B., Henderson, E.E., and German, J.
1995. Somatic intragenic recombination within the mutated locus BLM can correct the high sister-chromatid exchange phenotype of Bloom syndrome cells.
Am. J. Hum. Genet. 57:1019-27.
199 Ellis, N.A., Groden, J., Ye, T.Z., Straughen, J., Lennon, D.J., Ciocci, S., Proytcheva, M., and German, J. 1995. The Bloom's syndrome gene product is homologous to RecQ helicases. Cell 83:655-66.
Ellis, N.A., Roe, A.M., Kozloski, J., Proytcheva, M., Falk, C., and German, J. 1994.
Linkage disequilibrium between the FES, D15S127,andBLM loci in Ashkenazi Jews with Bloom syndrome. Am. J. Hum. Genet. 55:453-60.
Ellis, N.A., Proytcheva, M., Sanz, M.M., Ye, T.Z., and German, J.1999. Transfection of
BLM into cultured bloom syndrome cells reduces the sister-chromatid exchange rate toward normal. Am. J. Hum. Genet. 65:1368-74.
Epstein, C.J., Martin, G.M., Schultz, A.L., and Motulsky, A.G. 1966. Werner's syndrome:
A review of its symptomatology, natural history, pathologic features, genetics and relationship to the natural aging process. Medicine (Baltimore). 45:177-221.
Evans, H.J., Adams, A.C., Clarkson, J.M., and German, J. 1978. Chromosome aberrations and unscheduled DNA synthesis in X- and UV-irradiated lymphocytes from a boy with Bloom's syndrome and a man with xeroderma pigmentosum. Cytogenet. Cell
Genet. 20:124-40.
200 Evans, M.K. and Bohr, V.A. 1994. Gene-specific DNA repair of UV-induced cyclobutane pyrimidine dimers in some cancer-prone and premature-aging human syndromes. Mutat. Res. 314:221-31.
Fishel, R., Lescoe, M.K., Rao, M.R., Copeland, N.G., Jenkins, N.A., Garber, J., Kane,
M., and Kolodner, R. 1993. The human mutator gene homolog MSH2 and its association with hereditary nonpolyposis colon cancer. Cell.75:1027-38.
Fitzpatrick, T.B. 1962. Congenital telangiectatic erythema resembling lupus erythematosus in dwarfs: Case presentation. Arch. Derm. 86:127-8.
Frorath, B., Schmidt-Preuss, U., Siemers, U., Zollner, M., and Rudiger, H.W. 1984.
Heterozygous carriers for Bloom syndrome exhibit a spontaneously increased micronucleus formation in cultured fibroblasts. Hum Genet. 67:52-5.
Foucault, F., Buard, J., Praz, F., Jaulin, C., Stoppa-Lyonnet, D., Vergnaud, G., and Amor-
Gueret, M. 1996. Stability of microsatellites and minisatellites in Bloom syndrome, a human syndrome of genetic instability. Mutat. Res. 362:227-36.
Foucault, F., Vaury, C., Barakat, A., Thibout, D., Planchon, P., Jaulin, C., Praz, F., and
Amor-Gueret, M. 1997. Characterization of a new BLM mutation associated with a topoisomerase II alpha defect in a patient with Bloom's syndrome. Hum. Mol. Genet.
201 6:1427-34.
Fujikawa-Yamamoto, K., Odashima, S., Kurihara, T., and Murakami, F. 1987. DNA synthesis in Bloom's syndrome fibroblasts. Cell Tissue Kinet. 20:69-76.
Fundia, A.F., Gorla, N.B., Bonduel, M.M., Azpilicueta, O., Lejarraga, H., Muriel, F.S., and Larripa, I.B. 1992. Increased expression of 5q31 fragile site in a Bloom syndrome family. Hum. Genet. 89:569-72.
Gangloff, S., McDonald, J.P., Bendixen, C., Arthur, L., and Rothstein, R. 1994. The yeast type I topoisomerase Top3 interacts with Sgs1, a DNA helicase homolog: A potential eukaryotic reverse gyrase. Mol. Cell. Biol. 14:8391-8.
German, J. 1964. Cytological evidence for crossing-over in vitro in human lymphoid cells. Science 144:298-301.
German J. 1969. Bloom's syndrome. I. Genetical and clinical observations in the first twenty-seven patients. Am. J. Hum. Genet. 21:196-227.
German J. 1997. Bloom's syndrome. XX. The first 100 cancers. Cancer Genet.
Cytogenet. 93:100-6.
202 German, J. and Passarge, E. 1989. Bloom's syndrome. XII. Report from the Registry for
1987. Clin Genet. 35:57-69.
German, J.L. and Passarge, E. 1990. Bloom's Syndrome Registry. Int. J. Dermatol.
29:233-4.
German, J., Ellis, N.A., and Proytcheva, M. 1996. Bloom's syndrome. XIX. Cytogenetic and population evidence for genetic heterogeneity. Clin. Genet. 49:223-31.
German, J., Roe, A.M., Leppert, M.F., and Ellis, N.A. 1994. Bloom syndrome: An analysis of consanguineous families assigns the locus mutated to chromosome band
15q26.1. Proc. Natl. Acad. Sci. 91:6669-73.
German, J., Schonberg, S., Louie, E., and Chaganti, R.S. 1977. Bloom's syndrome. IV.
Sister-chromatid exchanges in lymphocytes. Am. J. Hum. Genet. 29:248-55.
German, J., Crippa, L.P., and Bloom, D. 1974. Bloom's syndrome. III. Analysis of the chromosome aberration characteristic of this disorder. Chromosoma 48:361-6.
Gianneli, F., Benson, P.F., Pawsey, S.A., and Polani, P.E. 1977. Ultraviolet light sensitivity and delayed DNA-chain maturation in Bloom's syndrome fibroblasts. Nature
265:466-9.
203 Giesler, T., Baker ,K., Zhang, B., McDaniel, L.D., and Schultz, R.A. Correction of the
Bloom syndrome cellular phenotypes. 1997. Somat. Cell Mol. Genet 5:303-12.
Gorbalenya AE, Blinov VM, Donchenko AP, Koonin EV. 1989. An NTP-binding motif is the most conserved sequence in a highly diverged monophyletic group of proteins involved in positive strand RNA viral replication. J. Mol. Evol. 28:256-68.
Goss et al., 2001. Manuscript in preparation.
Goto, M. 2000. Werner's syndrome: From clinics to genetics. Clin. Exp. Rheumatol.
18:760-6.
Goto, M., Tanimoto, K., Horiuchi, Y., and Sasazuki, T. 1981. Family analysis of
Werner's syndrome: A survey of 42 Japanese families with a review of the literature.
Clin. Genet. 19:8-15.
Gray, M.D., Shen, J.C., Kamath-Loeb, A.S., Blank, A., Sopher, B.L., Martin, G.M.,
Oshima, J., and Loeb, L.A. 1997. The Werner syndrome protein is a DNA helicase. Nat.
Genet. 17:100-3.
Gretzula, J.C., Hevia, O., and Weber, P.J. 1987. Bloom's syndrome. J. Am. Acad.
Dermatol. 17:479-88.
204 Grifo, J.A., Abramson, R.D., Satler, C.A., and Merrick, W.C. 1984. RNA-stimulated.
ATPase activity of eukaryotic initiation factors. J. Biol. Chem. 259:8648-54.
Groden, J., Nakamura, Y and German, J. Molecular evidence that homologous
recombination occurs in proliferating human somatic cells. 1990. Proc. Natl. Acad. Sci.
87:4315-9.
Groden, J. and German, J. 1992. Bloom's syndrome. XVIII. Hypermutability at a tandem-
repeat locus. Hum. Genet. 90:360-7.
Gupta, R.S. and Goldstein, S. 1982. Human fibroblast strains showing increased
sensitivity or resistance to ethidium bromide. Mutat. Res. 1982 Sep;105(3):183-8.
Gupta, P.K. and Sirover, M.A. 1984. Altered temporal expression of DNA repair in
hypermutable Bloom's syndrome cells. Proc. Natl. Acad. Sci. 81:757-61.
Hanada, K., Ukita, T., Kohno, Y., Saito, K., Kato, J., and Ikeda, H. 1997. RecQ DNA
helicase is a suppressor of illegitimate recombination in Escherichia coli. Proc. Natl.
Acad. Sci. 94:3860-5.
Hand, R. and German, J. 1975. A retarded rate of DNA chain growth in Bloom's syndrome. Proc. Natl. Acad. Sci. 72:758-62.
205 Harmon, F.G. and Kowalczykowski, S.C. 1998. RecQ helicase, in concert with RecA and
SSB proteins, initiates and disrupts DNA recombination. Genes Dev. 12:1134-44.
Heartlein, M.W., Tsuji, H., and Latt, S.A. 1987. 5-Bromodeoxyuridine-dependent increase in sister chromatid exchange formation in Bloom's syndrome is associated with reduction in topoisomerase II activity. Exp. Cell Res. 169:245-54.
Hellgren, D., Sahlen, S., Ljungqvist, S., and Kenne, K. 1994. The human RAD51/RecA homologue gene is not a candidate gene for Bloom's syndrome. Hum. Genet. 94:705-7.
Henson, P., Selsky, C.A., and Little, J.B. 1981. Excision of ultraviolet damage and the effect of irradiation on DNA synthesis in a strain of Bloom's syndrome fibroblasts.Cancer
Res. 41:760-6.
Hill, T.L. and Tsuchiya, T. 1981. Theoretical aspects of translocation on DNA:
Adenosine triphosphatases and treadmilling binding proteins. Proc. Natl. Acad. Sci.
78:4796-800.
Hirschi, M., Netrawali, M.S., Remsen, J.F., and Cerutti P.A. 1981. Formation of DNA single-strand breaks by near-ultraviolet and gamma-rays in normal and Bloom's syndrome skin fibroblasts. Cancer Res. 41:2003-7.
Hodgman, T.C. 1988. A new superfamily of replicative proteins. Nature 333:22-3.
206 Hojo, E.T., van Diemen, P.C., Darroudi, F., and Natarajan, A.T. 1995. Spontaneous chromosomal aberrations in Fanconi anaemia and ataxia telangiectasia fibroblast and
Bloom's syndrome lymphoblastoid cell lines as detected by conventional cytogenetic analysis and fluorescence in situ hybridisation (FISH) technique. Mutat. Res. 334:59-69.
Hook, G.J., Kwok, E., and Heddle, J.A. 1984. Sensitivity of Bloom syndrome fibroblasts to mitomycin C. Mutat. Res. 131:223-30.
Horii, Z. and Clark, A.J. 1973. Genetic analysis of the recF pathway to genetic recombination in Escherichia coli K12: Isolation and characterization of mutants. J. Mol.
Biol. 80:327-44.
Horii, A., Han, H.J., Sasaki, S., Shimada, M., and Nakamura, Y. 1994. Cloning, characterization and chromosomal assignment of the human genes homologous to yeast
PMS1, a member of mismatch repair genes.Biochem. Biophys. Res. Commun. 204:1257-
64.
Hsieh, C.L., Arlett, C.F., Lieber, M.R. 1993. V(D)J recombination in ataxia telangiectasia, Bloom's syndrome, and a DNA ligase I-associated immunodeficiency disorder. J. Biol. Chem. 268:20105-9.
207 Hurt, M.M. and Moses, R.E. 1986. Conversion of replicative intermediates in human
DNA-repair defective cells. Exp. Cell Res. 163:396-404.
Hutteroth, T.H., Litwin, S.D., and German, J. 1975. Abnormal immune responses of
Bloom's syndrome lymphocytes in vitro. J. Clin. Invest. 56:1-7.
Irino, N., Nakayama, K., and Nakayama, H. 1986. The recQ gene of Escherichia coli
K12: Primary structure and evidence for SOS regulation. Mol. Gen. Genet. 205:298-304.
Ishizaki, K., Yagi, T., Inoue, M., Nikaido, O., and Takebe, H. 1981. DNA repair in
Bloom's syndrome fibroblasts after UV irradiation or treatment with mitomycin C. Mutat.
Res. 80:213-9.
Kapp, L.N. 1982. DNA fork displacement rates in Bloom's syndrome fibroblasts.
Biochim. Biophys. Acta. 696:226-7.
Kaneko, H., Inoue, R., Yamada, Y., Sukegawa, K., Fukao, T., Tashita, H., Teramoto, T.,
Kasahara, K., Takami, T., and Kondo, N. 1996. Microsatellite instability in B-cell lymphoma originating from Bloom syndrome. Int. J. Cancer. 1996 69:480-3.
Karow, J.K., Chakraverty, R.K., and Hickson, I.D. 1997. The Bloom's syndrome gene product is a 3'-5' DNA helicase. J. Biol. Chem. 272:30611-4.
208 Kenne, K, and Akerblom, L. 1990. Deoxyribonucleoside triphosphate pool levels in three cell strains of human chromosome instability syndromes: Ataxia telangiectasia
(GM2052), Bloom's syndrome (GM1492), and Fanconi's anemia (GM368). Cancer
Biochem. Biophys. 11:69-77.
Kim, S., Vollberg, T.M., Ro, J.Y., Kim, M., Sirover, M.A. 1986. O6-methylguanine methyltransferase increases before S-phase in normal human cells but does not increase in hypermutable Bloom's syndrome cells. Mutat. Res. 173:141-5.
Kitao, S., Ohsugi, I., Ichikawa, K., Goto, M., Furuichi, Y., and Shimamoto, A. 1998.
Cloning of two new human helicase genes of the RecQ family: Biological significance of multiple species in higher eukaryotes. Genomics. 54:443-52.
Kondo, N., Motoyoshi, F., Mori, S., Kuwabara, N., Orii, T., and German, J. 1992. Long- term study of the immunodeficiency of Bloom's syndrome. Acta Paediatr. 81:86-90.
Korn, R. and Ramkisson, Y. 1995. Bloom's syndrome. Mapping to the point. Nature
378:557-80.
Krell, H., Durwald, H., Hoffmann-Berling, H. 1979. A DNA-unwinding enzyme induced in bacteriophage-T4-infected Escherichia coli cells. Eur. J. Biochem. 93:387-95.
209 Krepinsky, A.B., Heddle, J.A., and German, J. 1979. Sensitivity of Bloom's syndrome lymphocytes to ethyl methanesulfonate. Hum. Genet. 50:151-6.
Krepinsky, A.B., Rainbow, A.J., and Heddle, J.A. 1980. Studies on the ultraviolet light sensitivity of Bloom's syndrome fibroblasts. Mutat. Res. 69:357-68.
Kruk, P.A., Rampino, N.J., and Bohr, V.A. 1995. DNA damage and repair in telomeres:
Relation to aging. Proc. Natl. Acad. Sci. 92:258-62.
Kauli R., Prager-lewin, R., Kaufman, H., and Laron, Z. 1977. Gonadal function in
Bloom's syndrome. Clin Endocrinol. 6:285-9.
Kuhn, E.M. 1978. Mitotic chiasmata and other quadriradials in mitomycin C-treated
Bloom's syndrome lymphocytes. Chromosoma 66:287-97.
Kuhn, E.M. 1980. Effects of X-irradiation in G1 and G2 on Bloom's syndrome and normal chromosomes. Hum. Genet. 54:335-41.
Kuhn, E.M. and Therman, E. 1979. No increased chromosome breakage in three Bloom's syndrome heterozygotes. J. Med. Genet. 16:219-22.
Kuhn, E.M. and Therman, E. 1986. Cytogenetics of Bloom's syndrome.
Cancer Genet. Cytogenet. 22(1):1-18.
210 Kurihara, T., Inoue, M., and Tatsumi, K. 1987a. Hypersensitivity of Bloom's syndrome fibroblasts to N-ethyl-N-nitrosourea. Mutat. Res. 184:147-51.
Kurihara, T., Tatsumi, K., Takahashi, H., and Inoue, M. 1987b. Sister-chromatid exchanges induced by ultraviolet light in Bloom's syndrome fibroblasts. Mutat. Res.
183:197-202.
Kurihara, T., Teraoka, H., Inoue, M., Takebe, H., and Tatsumi, K. 1991. Two types of
DNA ligase I activity in lymphoblastoid cells from patients with Bloom's syndrome. Jpn.
J. Cancer Res. 82:51-7.
Kusano, K., Sunohara, Y., Takahashi, N., Yoshikura, H., and Kobayashi, I. 1994. DNA double-strand break repair: Genetic determinants of flanking crossing-over. Proc. Natl.
Acad. Sci. 91:1173-7.
Langland et al., 2001.
Lebel, M. and Leder, P. 1998 A deletion within the murine Werner syndrome helicase induces sensitivity to inhibitors of topoisomerase and loss of cellular proliferative capacity. Proc. Natl. Acad. Sci. 95:13097-102.
211 Lechner, J.F., Kaighn, M.E., Jetten, A.M., Groden, J., and German, J. 1983. Bloom's syndrome cells have an abnormal serum growth response. Exp. Cell Res. 145:381-8.
Lee, K.H., Abe, S., Yanabe, Y., Matsuda, I., and Yoshida, M.C. 1990. Superoxide dismutase activity and chromosome damage in cultured chromosome instability syndrome cells. Mutat. Res. 244:251-6.
Lewis, B.L. 1957. Discussion and case presentation. Arch. Derm. 75:772.
Lindgren, V. 1981. The location of chromosome breaks in Bloom's syndrome.
Cytogenet. Cell Genet. 29:99-106.
Lindor, N.M., Furuichi, Y., Kitao, S., Shimamoto, A., Arndt, C., Jalal, S. Rothmund-
Thomson syndrome due to RECQ4 helicase mutations: Report and clinical and molecular comparisons with Bloom syndrome and Werner syndrome. Am. J. Med .Genet. 90:223-8.
Lonn, U., Lonn, S., Nylen, U., and Winblad, G. 1990. Increased levels of 5-fluorouracil- induced DNA lesions in Bloom's syndrome. Int. J. Cancer. 45:494-9.
Lovett, S.T., Luisi-DeLuca, C., and Kolodner, R.D. 1988. The genetic dependence of recombination in recD mutants of Escherichia coli. Genetics. 120:37-45.
212 Lipkin, S.M., Wang, V., Jacoby, R., Banerjee-Basu, S., Baxevanis, A.D., Lynch, H.T.,
Elliott, R.M., and Collins, F.S. 2000. MLH3: a DNA mismatch repair gene associated with mammalian microsatellite instability.Nat. Genet. 24:27-35.
Lowy, A.M., Kordich, J.J., Gismondi, V., Varesco, L., Blough, R.I., and Groden, J. 2001.
Numerous Colonic Adenomas in an Individual with Bloom's Syndrome.Gastroenterology
121:435-9.
Lu, X. and Lane, D.P. 1993. Differential induction of transcriptionally active p53 following UV or ionizing radiation: Defects in chromosome instability syndromes? Cell
75:765-78.
Luo, G. Santoro, I.M, McDaniel, L.D., Nishijima, I., Mills, M., Youssoufian, H. Vogel,
H., Schultz, R.A., and Bradley, A. 2000. Cancer predisposition caused by elevated mitotic recombination in Bloom mice. Nat Genet. 26:424-9.
Mamada, A., Kondo, S., and Satoh, Y. 1989. Different sensitivities to ultraviolet light- induced cytotoxicity and sister chromatid exchanges in xeroderma pigmentosum and
Bloom's syndrome fibroblasts. Photodermatol. 6:124-30.
Marciniak, R.A., Lombard, D.B., Johnson, F.B., and Guarente, L. 1998. Nucleolar localization of the Werner syndrome protein in human cells. Proc. Natl. Acad. Sci.
213 95:6887-92.
Martin, R.H., Rademaker, A., and German, J. 1994. Chromosomal breakage in human spermatozoa, a heterozygous effect of the Bloom syndrome mutation. Am. J. Hum.
Genet. 55:1242-6.
McDaniel, L.D. and Schultz, R.A. 1992. Elevated sister chromatid exchange phenotype of Bloom syndrome cells is complemented by human chromosome 15. Proc. Natl. Acad.
Sci. 89:7968-72.
Matson, S.W. 1991. DNA helicases of Escherichia coli. Prog. Nucleic Acid Res. Mol.
Biol. 40:289-326.
Mezzina, M., Nardelli, J., Nocentini, S., Remault, G., and Sarasin, A. 1989a. DNA ligase activity in human cell lines from normal donors and Bloom's syndrome patients. Nucleic
Acids Res. 17:3091-106.
Mezzina, M., Nocentini, S., Nardelli, J., Renault, G., Moustacchi, E., and Sarasin, A.
1989b. Enhanced deoxyribonuclease activity in human transformed cells and in Bloom's syndrome cells. Mol. Carcinog. 2:179-83.
214 Muniyappa, K., Shaner, S.L., Tsang, S.S., and Radding, C.M. 1984. Mechanism of the concerted action of recA protein and helix-destabilizing proteins in homologous recombination. Proc. Natl. Acad. Sci. 81:2757-61.
Murray, J.M., Lindsay, H.D., Munday, C.A., and Carr, A.M. 1997. Role of
Schizosaccharomyces pombe RecQ homolog, recombination, and checkpoint genes in
UV damage tolerance. Mol. Cell Biol. 17:6868-75.
Nakayama, H., Nakayama, K., Nakayama, R., Irino, N., Nakayama, Y., Hanawalt, P.C.
1984. Isolation and genetic characterization of a thymineless death-resistant mutant of
Escherichia coli K12: Identification of a new mutation (recQ1) that blocks the RecF recombination pathway. Mol. Gen. Genet. 195:474-80.
Nakayama K, Irino N, Nakayama H. 1985. The recQ gene of Escherichia coli K12:
Molecular cloning and isolation of insertion mutants. Mol. Gen. Genet. 200:266-71.
Nocentini, S. 1995. Comet assay analysis of repair of DNA strand breaks in normal and deficient human cells exposed to radiations and chemicals. Evidence for a repair pathway specificity of DNA ligation. Radiat. Res. 144:170-80.
Nicotera, T.M. 1991. Molecular and biochemical aspects of Bloom's syndrome. Cancer
Genet. Cytogenet. 53:1-13.
215 Nicotera, T.M., Notaro, J., Notaro, S., Schumer, J., and Sandberg, A.A. 1989. Elevated superoxide dismutase in Bloom's syndrome: A genetic condition of oxidative stress.
Cancer Res. 49:5239-43.
Nicotera, T., Thusu, K., and Dandona, P. 1993. Elevated production of active oxygen in
Bloom's syndrome cell lines. Cancer Res. 53:5104-7.
Ockey, C.H. 1979. Quantitative replicon analysis of DNA synthesis in cancer-prone conditions and the defects in Bloom's syndrome. J. Cell Sci. 40:125-44.
Ockey, C.H. and Saffhill, R. 1986. Delayed DNA maturation, a possible cause of the elevated sister-chromatid exchange in Bloom's syndrome. Carcinogenesis 7:53-7.
Owen-Hughes, T., Utley, R.T., Cote, J., Peterson, C.L., and Workman, J.L. 1996.
Persistent site-specific remodeling of a nucleosome array by transient action of the
SWI/SNF complex. Science 273:513-6.
Ozawa, T., Kondo, N., Kato, Y., Motoyoshi, F., Suzuki, Y., Shimozawa, N., Kasahara,
K., and Orii, T. 1994. A wild-type mu s C-terminal gene is expressed in Bloom's syndrome cells. Eur. J. Immunogenet. 21:133-9.
216 Paquis-Flucklinger, V., Santucci-Darmanin, S., Paul, R., Saunieres, A., Turc-Carel, C., and Desnuelle, C. 1997. Cloning and expression analysis of a meiosis-specific MutS homolog: the human MSH4 gene. Genomics 44:188-94.
Parker, V.P. and Lieberman, M.W 1977. Levels of DNA polymerases alpha, beta, and gamma in control and repair-deficient human diploid fibroblasts 1. Nucleic Acids Res.
4:2029-37.
Petrini, J.H., Donovan, J.W., Dimare, C., and Weaver, D.T. 1994. Normal V(D)J coding junction formation in DNA ligase I deficiency syndromes. J. Immunol. 152:176-83.
Petrini, J.H., Huwiler, K.G., and Weaver, D.T. 1991. A wild-type DNA ligase I gene is expressed in Bloom's syndrome cells. Proc. Natl. Acad. Sci. 88:7615-9.
Petukhova, G., Stratton, S.A., Sung, P. Single strand DNA binding and annealing activities in the yeast recombination factor Rad59. J. Biol. Chem. 274:33839-42.
Petukhova, G., Van Komen, S., Vergano, S., Klein, H., and Sung, P. 1999. Yeast Rad54 promotes Rad51-dependent homologous DNA pairing via ATP hydrolysis-driven change in DNA double helix conformation. J. Biol. Chem. 274:29453-62.
217 Poot, M., Hoehn, H., Nicotera, T.M., and Rudiger, H.W. 1989. Cell kinetic evidence suggests elevated oxidative stress in cultured cells of Bloom's syndrome. Free Radic. Res.
Commun. 7:179-87.
Poot, M., Rudiger, H.W., and Hoehn, H. 1990. Detection of free radical-induced DNA damage with bromodeoxyuridine/Hoechst flow cytometry: Implications for Bloom's syndrome. Mutat. Res. 238:203-7.
Puranam, K.L. and Blackshear, P.J. 1994. Cloning and characterization of RECQL, a potential human homologue of the Escherichia coli DNA helicase RecQ. J. Biol. Chem.
269:29838-45.
Puranam, K.L., Kennington, E., Sait, S.N., Shows, T.B., Rochelle, J.M., Seldin, M.F., and Blackshear, P.J. 1995. Chromosomal localization of the gene encoding the human
DNA helicase RECQL and its mouse homologue. Genomics 26:595-8.
Rauh, J.L. and Soukup, S.W. 1968. Bloom's syndrome. Am. J. Dis. Child. 116:409-13.
Remsen, J.F. 1980. Repair of damage by N-acetoxy-2-acetylaminofluorene in Bloom's syndrome. Mutat. Res. 72:151-4.
Roa, B.B., Savino, C.V., and Richards, C.S. 1999. Ashkenazi Jewish population frequency of the Bloom syndrome gene 2281 delta 6ins7 mutation. Genet. Test. 3:219-21.
218 Roberts, J.D. and Kunkel, T.A. 1994. Determining the fidelity of bidirectional DNA replication in human cell extracts. Chromosomes and Gene Analysis, Meth. Molec.
Genetics 2:295-313.
Roman, L.J. and Kowalczykowski, S.C. 1989. Formation of heteroduplex DNA promoted by the combined activities of Escherichia coli recA and recBCD proteins. J. Biol. Chem.
264:18340-8.
Rosin, M.P. and German, J. 1985. Evidence for chromosome instability in vivo in Bloom syndrome: increased numbers of micronuclei in exfoliated cells. Hum. Genet. 71:187-91.
Rozen, F., Pelletier, J., Trachsel, H., and Sonenberg, N. 1989. A lysine substitution in the
ATP-binding site of eucaryotic initiation factor 4A abrogates nucleotide-binding activity.
Mol. Cell Biol. 9:4061-3.
Rudiger, H.W. 1982. Cytogenetic demonstration of a corrective factor in Bloom's syndrome. IARC Sci. Publ. 39:137-45.
Runger, T.M. and Kraemer, K.H. 1989. Joining of linear plasmid DNA is reduced and error-prone in Bloom's syndrome cells. EMBO J. 8:1419-25.
219 Runyon, G.T., Wong, I., Lohman, T.M. 1993. Overexpression, purification, DNA
binding, and dimerization of the Escherichia coli uvrD gene product (helicase II).
Biochemistry 32:602-12.
Ryder, L., Whitby, M.C., and Lloyd, R.G. 1994. Mutation of recF, recJ, recO, recQ,or recR improves Hfr recombination in resolvase-deficient ruv recG strains of Escherichia coli. J. Bacteriol. 176:1570-7.
Sachs, A.B. and Davis, R.W. 1990. Translation initiation and ribosomal biogenesis:
Involvement of a putative rRNA helicase and RPL46. Science 247:1077-9.
Sawitsky, A., Bloom, D., and German, J. 1966. Chromosomal breakage and acute leukemia in congenital telangiectatic erythema and stunted growth. Ann. Intern. Med.
65:487-95.
Schoen, E.J. and Shearn, M.A. 1967. Immunoglobulin deficiency in Bloom’s syndrome.
Amer. J. Dis. Child. 113:594-6.
Schroeder, T.M. and German, J. 1974. Bloom's syndrome and Fanconi's anemia:
Demonstration of two distinctive patterns of chromosome disruption and rearrangement.
Humangenetik 25:299-306.
220 Schmid, S.R. and Linder, P. 1992. D-E-A-D protein family of putative RNA helicases.
Mol. Microbiol. 6:283-91.
Seal, G., Brech, K., Karp, S.J., Cool, B.L., and Sirover, M.A. 1988. Immunological lesions in human uracil DNA glycosylase: Association with Bloom syndrome. Proc. Natl.
Acad. Sci. 85:2339-43.
Seal, G., Henderson, E.E., and Sirover, M.A. 1990. Immunological alteration of the
Bloom's syndrome uracil DNA glycosylase in Epstein-Barr virus-transformed human lymphoblastoid cells. Mutat. Res. 243:241-8.
Seal, G., Tallarida, R.J., and Sirover, M.A. 1991. Purification and properties of the uracil
DNA glycosylase from Bloom's syndrome. Biochim. Biophys. Acta. 1097:299-308.
Seki, M., Miyazawa, H., Tada, S., Yanagisawa, J., Yamaoka, T., Hoshino, S., Ozawa, K.,
Eki, T., Nogami, M., and Okumura, K. 1994. Molecular cloning of cDNA encoding human DNA helicase Q1 which has homology to Escherichia coli Rec Q helicase and localization of the gene at chromosome 12p12. Nucleic Acids Res. 22:4566-73.
Selby, C.P. and Sancar, A. 1993. Molecular mechanism of transcription-repair coupling.
Science 260:53-8.
221 Selby, C.P. and Sancar, A. 1997. Human transcription-repair coupling factor
CSB/ERCC6 is a DNA-stimulated ATPase but is not a helicase and does not disrupt the ternary transcription complex of stalled RNA polymerase II. J. Biol. Chem. 272:1885-90.
SenGupta, D.J. and Borowiec, J.A. Strand-specific recognition of a synthetic DNA replication fork by the SV40 large tumor antigen. Science 256:1656-61.
Shen, J.C., Gray, M.D., Oshima, J., and Loeb, L.A. 1998. Characterization of Werner syndrome protein DNA helicase activity: Directionality, substrate dependence and stimulation by replication protein A. Nucleic Acids Res. 26:2879-85.
Shimamoto, A., Nishikawa, K., Kitao, S., and Furuichi, Y. 2000. Human RecQ5beta, a large isomer of RecQ5 DNA helicase, localizes in the nucleoplasm and interacts with topoisomerases 3alpha and 3beta. Nucleic Acids Res. 28:1647-55.
Shiraishi Y. 1985. Bloom syndrome B-lymphoblastoid cells are hypersensitive towards carcinogen and tumor promoter-induced chromosomal alterations and growth in agar.
EMBO J. 4:2553-60.
Shiraishi Y. 1988. Effects of cell fusion and deoxynucleosides on sister-chromatid exchanges in B-lymphoblastoid cell lines from Five Bloom syndrome patients. Mutat.
Res. 199:75-83.
222 Shiraishi, Y. 1990. Nature and role of high sister chromatid exchanges in Bloom syndrome cells. Some cytogenetic and immunological aspects. Cancer Genet. Cytogenet.
50:175-87.
Shiraishi, Y. and Li, M.G. 1990. Bromodeoxyuridine (BrdU) template and thymidine pool effects on high frequencies of sister-chromatid exchange (SCE) in Bloom syndrome cells and a mutant cell line (AsHa) originated from ataxia telangiectasia. Mutat. Res.
230:177-86.
Shiraishi, Y. 1988. Effects of cell fusion and deoxynucleosides on sister-chromatid exchanges in B-lymphoblastoid cell lines from 5 Bloom syndrome patients. Mutat Res.
199:75-83.
Shiraishi, Y., Miyagi, T., Miwa, M., and Sandberg, A.A. 1982. Effects of temperature on sister chromatid exchange in Bloom syndrome cells. Gann. 73:514-6.
Shiraishi, Y., Ogata, T., and Soma, H. 1992. Enhanced expression of stomach cancer antigen derived from malignantly transformed bloom syndrome cells previously labeled with bromodeoxyuridine. Cancer Genet. Cytogenet. 62:1-8.
Shiraishi Y. and Sandberg, A.A. 1978a. Effects of mitomycin C on sister chromatid exchange in normal and Bloom's syndrome cells. Mutat. Res. 49:233-8.
223 Shiraishi Y. and Sandberg, A.A. 1978b. Evaluation of sister chromatid exchanges in
Bloom's syndrome. Cytobios. 21:175-84.
Shiraishi Y. and Sandberg, A.A. 1979. Effects of various chemical agents on sister chromatid exchanges, chromosome aberrations, and DNA repair in normal and abnormal human lymphoid cell lines. J. Natl. Cancer Inst. 62:27-35.
Shiraishi, Y., Tanaka, Y., Kato, M., Miwa, M., and Sugimura, T. 1983. Effect of poly(ADP-ribose)polymerase inhibitors on the frequency of sister-chromatid exchanges in Bloom syndrome cells. Mutat. Res. 122:223-8.
Sinclair, D.A., Mills, K., and Guarente, L. 1997. Accelerated aging and nucleolar fragmentation in yeast sgs1 mutants. Science 277:1313-6.
Spanos, A., Holliday, R., and German, J. 1986. Bloom's syndrome. XIII. DNA- polymerase activity of cultured lymphoblastoid cells. Hum. Genet. 73:119-22.
Squires, S., Johnson, R.T., and Collins, A.R. 1982. Initial rates of DNA incision in UV- irradiated human cells: Differences between normal, xeroderma pigmentosum and tumour cells. Mutat. Res. 95:389-404.
224 Stewart, E., Chapman, C.R., Al-Khodairy, F., Carr, A.M., and Enoch, T. 1997. rqh1+, a fission yeast gene related to the Bloom's and Werner's syndrome genes, is required for reversible S-phase arrest. EMBO J. 16:2682-92.
Straughen, J., Ciocci, S., Ye, T.Z., Lennon, D.N., Proytcheva, M., Alhadeff, B.,
Goodfellow, P., German, J., Ellis, N.A., and Groden, J. 1996. Physical mapping of the
Bloom syndrome region by the identification of YAC and P1 clones from human chromosome 15 band q26.1. Genomics 35:118-28.
Straughen, J.E., Johnson, J., McLaren, D., Proytcheva, M., Ellis, N., German, J., and
Groden, J. 1998. A Rapid Method for Detecting the Predominant Ashkenazi Jewish
Mutation in the Bloom’s Syndrome Gene. Hum. Mutation 11:175-8.
Straughen, J.E., et al., 2001. Genomic and Microsatellite Repeat Instability in BS.
Manuscript in Preparation.
Straughen, J.E., et al., 2001. Genomic Structure of BLM. Manuscript in Preparation.
Strauss, E.J. and Guthrie, C. 1991. A cold-sensitive mRNA splicing mutant is a member of the RNA helicase gene family. Genes Dev. 5:629-41.
Subramanya, H.S., Bird, L.E., Brannigan, J.A., and Wigley, D.B. 1996. Crystal structure of a DExx box DNA helicase. Nature 384:379-83.
225 Sun, H., Karow, J.K., Hickson, I.D., and Maizels, N. 1998. The Bloom's syndrome helicase unwinds G4 DNA. J. Biol. Chem. 273:27587-92.
Sung, P. 1997a. Function of yeast Rad52 protein as a mediator between replication protein A and the Rad51 recombinase. J. Biol. Chem. 272:28194-7.
Sung, P. 1997b. Yeast Rad55 and Rad57 proteins form a heterodimer that functions with replication protein A to promote DNA strand exchange by Rad51 recombinase. Genes
Dev. 11:1111-21.
Suzuki, N., Shimamoto, A., Imamura, O., Kuromitsu, J., Kitao, S., Goto, M., and
Furuichi, Y. 1997. DNA helicase activity in Werner's syndrome gene product synthesized in a baculovirus system. Nucleic Acids Res. 25:2973-8.
Tachibana, A., Tatsumi, K., Masui, T., and Kato, T. 1996. Large deletions at the HPRT locus associated with the mutator phenotype in a Bloom's syndrome lymphoblastoid cell line. Mol. Carcin. 17:41-7.
Tada, S., Yanagisawa, J., Sonoyama, T., Miyajima, A., Seki, M., Ui, M., Enomoto, T.
1996. Characterization of the properties of a human homologue of Escherichia coli RecQ from xeroderma pigmentosum group C and from HeLa cells. Cell Struct. Funct. 21:123-
226 32.
Taylor, A. and Smith, G.R. 1980. Unwinding and rewinding of DNA by the RecBC enzyme. Cell 22:447-57.
Taylor, M.W., Kothari, R.M., Holland, G.D., Martinez-Valdez, H., and Zeige, G. 1983. A comparison of purine and pyrimidine pools in Bloom's syndrome and normal cells.
Cancer Biochem. Biophys. 7:19-25.
Teo, I.A. and Arlett, C.F. 1982. The response of a variety of human fibroblast cell strains to the lethal effects of alkylating agents. Carcinogenesis 3:33-7.
Thomas, D.C., Roberts, J.D., Sabatino, R.D., Myers, T.W., Downey, K.M., So, A.G.,
Bambara, R.A. and Kunkel, T.A. 1991. Fidelity of mammalian DNA replication and replicative DNA polymerases. Biochemistry 30:11751-9.
Tice, R., Windler, G., Bary, J.M. Effect of cocultivation on sister chromatid exchange frequencies in Bloom's syndrome and normal fibroblast cells. Nature 273:538-40.
Tollefsbol, T.O. and Cohen, H.J. 1984. The effect of age on the accumulation of labile triosephosphate isomerase and thymidine incorporation in pokeweed mitogen stimulated human lymphocytes. J. Gerontol. 39:398-405.
227 Tomkinson, A.E., Lasko, D.D., and Lindahl, T. 1990. Biochemical properties of mammalian DNA ligase I and the molecular defect in Bloom's syndrome. Prog. Clin.
Biol. Res. 340A:283-94.
Tomkinson, A.E., Starr, R., and Schultz, R.A.. 1993. DNA ligase III is the major high molecular weight DNA joining activity in SV40-transformed human fibroblasts: Normal levels of DNA ligase III activity in Bloom syndrome cells. Nucleic Acids Res. 21:5425-
30.
Troelstra, C., van Gool, A., de Wit, J., Vermeulen, W., Bootsma, D., and Hoeijmakers,
J.H. 1992. ERCC6, a member of a subfamily of putative helicases, is involved in
Cockayne's syndrome and preferential repair of active genes. Cell 71:939-53.
Tsuji, H. and Kojima, T. 1985. Presence of abnormally high incidences of sister chromatid exchanges in three successive cell cycles in Bloom's syndrome lymphocytes.
Chromosoma 93:87-93.
Ueno, Y., Miyawaki, T., Seki, H., Hara, K., Sato, T., Taniguchi, N., Takahashi, H.,
Kondo, N. 1985. Impaired natural killer cell activity in Bloom's syndrome could be restored by human recombinant IL-2 in vitro. Clin. Immunol. Immunopathol. 35:226-33.
Umezu, K., Nakayama, K., Nakayama, H. 1990. Escherichia coli RecQ protein is a DNA helicase. Proc. Natl. Acad. Sci. 87:5363-7.
228 van Buul, P.P., Natarajan, A.T., and Verdegaal-Immerzeel, E.A. 1978. Suppression of the frequencies of sister chromatid exchanges in Bloom's syndrome fibroblasts by co- cultivation with Chinese hamster cells. Hum. Genet. 44:187-9.
van Gool, A.J., Hajibagheri, N.M., Stasiak, A., and West, S.C. 1999. Assembly of the
Escherichia coli RuvABC resolvasome directs the orientation of holliday junction resolution. Genes Dev. 13:1861-70.
van Laar, T., Steegenga, W.T., Jochemsen, A.G., Terleth, C., and van der Eb, A.J. 1994.
Bloom's syndrome cells GM1492 lack detectable p53 protein but exhibit normal G1 cell- cycle arrest after UV irradiation. Oncogene 9:981-3.
van Laar, T., Steegenga, W.T., Jochemsen, A.G., Terleth, C., and van der Eb, A.J. 1995.
GM1492 human diploid skin fibroblasts lack the p53-dependent G1 cell-cycle checkpoint. Biochem. Biophys. Res. Commun. 217:769-76.
Vilpo, J.A. and Vilpo, L.M. 1989. Normal uracil-DNA glycosylase activity in Bloom's syndrome cells. Mutat. Res. 210:59-62.
Vollberg, T.M., Seal, G., and Sirover, M.A. 1987. Monoclonal antibodies detect conformational abnormality of uracil DNA glycosylase in Bloom's syndrome cells.
Carcinogenesis 8:1725-9.
229 Walker, J.E., Saraste, M., Runswick, M.J., Gay, N.J. 1982. Distantly related sequences in the alpha- and beta-subunits of ATP synthase, myosin, kinases and other ATP-requiring enzymes and a common nucleotide binding fold. EMBO J. 1:945-51
Wang, L., Hunt, K.E., Martin, G.M., and Oshima, J. 1998. Structure and function of the human Werner syndrome gene promoter: Evidence for transcriptional modulation.
Nucleic Acids Res. 26:3480-5.
Wang, T.C. and Chang, H.Y. 1991. Effect of rec mutations on viability and processing of
DNA damaged by methylmethane sulfonate in xth, nth, nfo cells of Escherichia coli.
Biochem. Biophys. Res. Commun. 180:774-81.
Watt, P.M., Hickson, I.D., Borts, R.H., and Louis, E.J. 1996. SGS1, a homologue of the
Bloom's and Werner's syndrome genes, is required for maintenance of genome stability in
Saccharomyces cerevisiae. Genetics 144:935-45.
Weber, C.A., Salazar, E.P., Stewart, S.A., and Thompson, L.H. 1990. ERCC2: cDNA cloning and molecular characterization of a human nucleotide excision repair gene with high homology to yeast RAD3. EMBO J. 9:1437-47.
Weemaes, C.M., Bakkeren, J.A., Haraldsson, A., and Smeets, D.F. 1991. Immunological studies in Bloom's syndrome. A follow-up report. Ann. Genet. 34:201-5.
230 Weemaes, C.M., The, T.H., van Munster, P.J., and Bakkeren, J.A. 1984. Antibody
responses in vivo in chromosome instability syndromes with immunodeficiency. Clin.
Exp. Immunol. 57:529-34.
Weksberg, R. 1995. Low-sister-chromatid-exchange Bloom syndrome cell lines: an
important new tool for mapping the basic genetic defect in Bloom syndrome and for
unraveling the biology of human tumor development. Am. J. Hum. Genet. 57:994-7.
Weksberg, R., Smith, C., Anson-Cartwright, L., and Maloney, K. 1988. Bloom
syndrome: A single complementation group defines patients of diverse ethnic origin. Am.
J. Hum. Genet. 42:816-24.
West, M.J., Sullivan, N.F., and Willis, A.E. 1995. Translational upregulation of the c-myc
oncogene in Bloom's syndrome cell lines. Oncogene 11:2515-24.
Whitby, M.C. and Lloyd, R. G. 1998. Targeting Holliday junctions by the RecG branch
migration protein of Escherichia coli. J. Biol. Chem. 273:19729-39.
Willis, A.E. and Lindahl, T. 1987. DNA ligase I deficiency in Bloom's syndrome. Nature
325:355-7.
231 Willis, A.E., Weksberg, R., Tomlinson, S., and Lindahl, T. 1987. Structural alterations of
DNA ligase I in Bloom syndrome. Proc. Natl. Acad. Sci. 84:8016-20.
Wolff RK, Nakamura Y, White R. Molecular characterization of a spontaneously generated new allele at a VNTR locus: no exchange of flanking DNA sequence.
Genomics. 1988;3:347-51.
Wong, I., Chao, K.L., Bujalowski, W., and Lohman, T.M. 1992. DNA-induced dimerization of the Escherichia coli rep helicase. Allosteric effects of single-stranded and duplex DNA. J. Biol. Chem. 267:7596-610.
Wong, I., Lohman, T.M. 1992. Allosteric effects of nucleotide cofactors on Escherichia coli Rep helicase-DNA binding. Scienc. 256:350-5.
Yamagata, K., Kato, J., Shimamoto, A., Goto, M., Furuichi, Y., and Ikeda, H. 1998.
Bloom's and Werner's syndrome genes suppress hyperrecombination in yeast sgs1 mutant: Implication for genomic instability in human diseases. Proc. Natl. Acad. Sci.
958733-8.
Yamamoto and Fujiwara, 1986. Abnormal regulation of uracil-DNA glycosylase induction during cell cycle and cell passage in Bloom's syndrome fibroblasts.
Carcinogenesis 7:305-10.
232 Ye, L., Miki, T., Nakura, J., Oshima, J., Kamino, K., Rakugi, H., Ikegami, H., Higaki, J.,
Edland, S.D., Martin, G.M., and Ogihara, T. 1997. Association of a polymorphic variant of the Werner helicase gene with myocardial infarction in a Japanese population. Am. J.
Med. Genet. 68:494-8.
Yoshida, M.C. and Sekguchi, T. 1984. Correction of sister chromatid exchanges in
Bloom's syndrome fibroblasts after fusion with cytoplasts from a Chinese hamster cell line. Exp. Cell Res. 155:315-9.
Young, J., Leggett, B., Gustafson, C., Ward, M., Searle, J., Thomas, L., Buttenshaw, R., and Chenevix-Trench, G. 1993. Genomic instability occurs in colorectal carcinomas but not in adenomas. Hum. Mutat. 2:351-4.
Yu, C.E., Oshima, J., Fu, Y.H., Wijsman, E.M., Hisama, F., Alisch, R., Matthews, S.,
Nakura, J., Miki, T., Ouais, S., Martin, G.M., Mulligan, J., and Schellenberg, G.D. 1996a.
Positional cloning of the Werner's syndrome gene. Science 272:258-62.
Yu, X., Hingorani, M.M., Patel, S.S., and Egelman, E.H. 1996b. DNA is bound within the central hole to one or two of the six subunits of the T7 DNA helicase. Nat. Struct.
Biol. 3:740-3.
Zbinden, I. and Cerutti, P. 1981. Near-ultraviolet sensitivity of skin fibroblasts of patients with Bloom's syndrome. Biochem. Biophys. Res. Commun. 98:579-87.
233 Zhong S, Hu P, Ye TZ, Stan R, Ellis NA, Pandolfi PP. 1999. A role for PML and the nuclear body in genomic stability. Oncogene 18:7941-7.
234