The Blm Helicase Literature Review

Home , Bloom syndrome protein, MLH1, RecF pathway

UNIVERSITY OF CINCINNATI

DATE: 3-22-03

I, Gregory T. Langland , hereby submit this as part of the requirements for the degree of:

DOCTORATE OF PHILOSOPHY (Ph.D.)

in:

The Department of Molecular Genetics, Biochemistry and

Microbiology of the College of Medicine

It is entitled:

INTERACTION BETWEEN THE BLM HELICASE AND THE DNA MISMATCH REPAIR PROTEIN, MLH1

Approved by:

Joanna Groden Ph.D.

Richard Wenstrup M.D.

Jim Stringer Ph.D.

Kathleen Dixon Ph.D.

Peter Stambrook Ph.D.

INTERACTION BETWEEN THE BLM HELICASE AND THE DNA MISMATCH REPAIR PROTEIN, MLH1

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, Biochemistry & Microbiology of the College of Medicine

2002

Gregory T. Langland

B.S., University of Cincinnati, 1992

Committee Chair: Joanna Groden, Ph.D.

iii

`Abstract

Bloom’s syndrome (BS) is a rare autosomal recessive disorder that greatly predisposes affected individuals to cancer. Such individuals also are small in size, sensitive to the sun, have immune dysfunction and gross genomic instability. The cytogenetics of BS cells have been extensively studied and have shown increased levels of homologous recombination, quadriradial formations, telomeric associations and chromosome breakage. The gene responsible for BS has been positionally cloned and and encodes a RecQ helicase family member with strand displacement activity that is dependent on ATP and Mg2+.

In order to have a greater understanding of BLM helicase function in the cell in regards to DNA replication, recombination and repair, we identified protein- partners of BLM. The C-terminus of BLM identified the DNA mismatch repair protein MLH1 from a yeast two-hybrid screen. In vitro and in vivo immunoprecipitations confirmed the interaction between these two proteins.

Using an in vitro mismatch repair assay, BS cell extracts were tested for their ability to correct a single nucleotide mismatch. The BS cell extracts were able to remove the single nucleotide mismatch from the plasmid DNA, demonstrating that the BLM-MLH1 interaction is not necessary to correct a single nucleotide mismatch.

To test the hypothesis that MLH1 may regulate the substrate specificity or helicase activity, two different experiments were performed. Gel-shifts were performed plus or minus the presence of MLH1 with different DNA substrates, however MLH1 had no effect on BLM’s ability to bind the different DNA

substrates. Helicase assays then were performed which demonstrated that

MLH1 or the mutL heterodimer modulates the enzymatic activity of BLM by stimulating BLM’s strand displacement activity on the double-overhang (DO) substrate.

Finally, we performed experiments with the supF20 mutagenesis system and demonstrated that extracts from BS cells are unable to utilize microhomology elements within the supF20 gene to restore supF function following the induction of a double strand break (DSB). Additional experiments with the pUC18 mutagenesis system demonstrate that although the efficiency and fidelity of DSB repair by BS extracts are comparable to those of normal extracts when ligatable ends are present, a significant 5-fold increase in mutation rate with BS extracts is observed when terminal phosphates are removed from the DNA substrate that needs repair. Mutant plasmids recovered following DSB repair by

BS extracts contain smaller deletions within the lacZα gene not commonly recovered from normal extracts. Colorectal cancer cell lineHCT116 extracts lacking MLH1 were also examined although the efficiency and fidelity of end- joining was similar to control extracts. This suggests that the BLM-MLH1 interaction is not necessary for proper end-joining.

In summary this work demonstrates that BS cells lacking the BLM helicase process DSBs differently than normal cells and strongly suggests a role for BLM in aligning micro-homology elements during recombinational events in DSB repair. Disruption of the BLM helicase may lead to replication fork collapses, improper processing of DSBs, genomic instability and ultimately cancer.

Acknowledgments

Someone once told me anything worth having you will have to work hard and fight for. The completion of my doctoral degree has been no different. It’s more than classwork and benchwork. It has been a learning experience, but let’s just keep it at that. I would like to that everybody that supported me through this effort. Many thanks go to Rick for encouraging me to enter graduate school and gave me the support and encouragement needed to further my career in science. I would like to thank Joanna for having me as a student and all of my committee members for their support and their signature, especially Dr. Peter Stambrook, Dr. Sohaib Khan and my mother for the financial support. Chris H. for showing me that you can give a great thirty-minute seminar with virtually no data. Chris T. for teaching how to make money in graduate school via day-trading and/or fantasy football. Kathy, who will always be my favorite post-doc (Oh, sorry now Assistant Professor). Jenn who Chris T. and I will be fighting for when we have our own labs. Will and Amod will always be remembered as DS-1 and DS-2, they made it fun to be at work. Greg B. for being there and conducting triage after a committee meeting or seminar. I definitely have to thank Al for purifying that nasty protein that I never could. I guess that ten years of protein purification definitely means you have skills. Both Tims were entertaining but I preferred the second one. Chelsea, thanks for the movies and the lunches. James for being there through thick and thin. Lisa, Heather, Mike B.,Dirk, Kevin and Mike for entertaining my wife while I was in graduate school. And last but definitely not least I have to thank Rachel for all the love and support.

Now for the questions I hear every day:

When are you are graduating ??

Where are you going ??

Is Rachel moving with you ??

To Be Continued……

vii

Acknowledgements vii

Abstract iv

Table of Contents 1

Abbreviations 3

List of Figures 5

List of Tables 7

Chapter 1 – Literature Review 8 Cancer and Genomic Instability 8 Bloom’s syndrome 11 Patient phenotype 15 Positional cloning of BLM 17 Functional motifs of the BLM helicase 19 BLM and PML bodies 23 BLM homologs and orthologs 24 The BLM protein exists as an oligomeric form 26 BLM and WRN knockout mice 27 DNA replication 29 Sensitivity of BS to DNA damaging agents 31 The BASC complex 35 Protein-partners of the BLM helicase 36

Chapter 2 – Thesis Rationale 39 Examining the role of the BLM helicase in mismatch repair and double-strand break repair.

Chapter 3 – Material and Methods 40 Cell Culture 40 Reagents and Enzymes 40 Nuclear extract preparation 41 Expression construct generation and characterization 42 Yeast-two hybrid screening 42 Isolation and renaturation of BLM-C 43 IVTT immunoprecipitations 44 Mixed lysate immunoprecipitations 45 In vivo immunoprecipitations 46 Mismatch repair assay 46 supF20 double-strand break repair assay 47 pUC18 double-strand break repair assay 48 Expression and purification of yBLM 49 Preparation of helicase substrates 51

Helicase assays 54 Gel-shift assays 55

Chapter 4 – The BLM helicase interacts with MLH1 56 BLM identifies MLH1 in a yeast two-hybrid screen 56 IVTT BLM-C and full-length MLH1 interact 56 Far western assays confirm the interaction between 59 MLH1 and BLM-C BLM and MLH1 interact in vivo 63 DNA mismatch repair activities of BS and control cell 64 extracts are equivalent.

Chapter 5– Stimulation of BLM helicase activity by mismatch repair proteins Mismatch repair in E. coli 66 BLM purification and characterization 69 Gel-shift experiments 73 Helicase assays 75

Chapter 6 – The BLM helicase is necessary for normal 82 double-strand break repair Double-strand break repair in mammalian cells 82 In vitro end-joining assay using the supF20 vector 83 In vitro end-joining assay using the pUC18 vector 83 Sequence analysis of pUC18 mutants 90 Examination of the efficiency and fidelity of 94 End-joining from HCT116 cell extracts.

Chapter 7- Conclusions and Future Directions 97

Chapter 8- References 113

Abbreviations aa- amino acid

AT- ataxia telangiectasia

BLM- Bloom’s syndrome gene

BLM- Bloom’s syndrome protein bp- base pair

BS- Bloom’s syndrome

DNA-PK- DNA protein kinase

DO- double overhang substrate

DSB- double-strand break

ENU- N-ethyl-nitrosurea

HU- hydroxyurea

IDL- insertion/deletion loop

IP- immunoprecipitation

IVTT- in vitro transcription/translation

MLH- mutL homologue

MMC- mitomycin C

NE- nuclear extract

NLS- nuclear localization signal nt.- nucleotide

PML- promyelocytic leukemia

PMS- postmeiotic segregation increased

RPA- replication protein A

SCE- sister-chromatid exchange

SCP- somatic crossover point

SDS-PAGE- sodium dodecyl sulfate- polyacrylamide gel electrophoresis

UDS- unscheduled DNA synthesis

WRN- Werner’s syndrome gene

WRN- Werner’s syndrome protein

WS- Werner’s syndrome

List of Figures Page No.

Figure 1. DNA repair genes act as caretakers of the genome. 10

Figure 2. Individual affected by Bloom’s syndrome. 12

Figure 3. BS cells are characterized by high levels of genomic 14 instability.

Figure 4. Functional motifs of the BLM helicase. 21

Figure 5. BLM identifies MLH1 in a yeast two-hybrid screen 57

Figure 6. Immunoprecipitations of in vitro transcribed and 58 translated (IVTT) protein products demonstrate the interaction between the C-terminus of BLM and MLH1.

Figure 7. Mixed lysate immunoprecipitation demonstrates the 61 interaction between full-length BLM and MLH1 or RPA.

Figure 8. Far western assays demonstrate the interaction 62 between the BLM-C terminus and MLH1.

Figure 9. BLM and MLH1 interact in vivo. 63

Figure10. DNA mismatch repair activities of BS and 65 HeLa cell extracts are equivalent.

Figure 11. Co-localization of BLM and hMLH1 in the nucleus 68 of WI-38/VA-13 cells.

Figure 12. The combinatorial specificites of MutS- and 69 MutL-related heterocomplexes.

Figure 13. Silver stain and western blot of BLM fractions 71 purified by nickel chelation affinity chromatography.

Figure 14. The recombinant BLM protein has helicase activity 72 that is ATP and Mg2+dependent.

Figure 15. The BLM helicase has a higher affinity for 4 nt. and 74 12 nt. insertion/deletion loops,

Figure 16. Effect of MLH1 on BLM helicase activity. 77

Figure 17. Effect of MLH1/PMS2 heterodimer on BLM helicase 78 activity.

Figure 18. The MLH1/PMS2 heterodimer stimulates BLM 80 helicase activity in vitro.

Figure 19. Effect of MLH1or MLH1/PMS2 heterodimer on 81 strand displacement activity.

Figure 20. Plasmid substrates used in the DSB repair assays. 84

Figure 21. The efficiency and fidelity of double-strand DNA 88 break repair using normal, corrected BS, BS and AT nuclear extracts.

Figure 22. Deletion mutations in plasmids recovered from 91 the in vitro DSB repair assay using pUC18 EcoRI-CIP as the substrate.

Figure 23. The efficiency of double-strand DNA break repair 95 using normal, corrected BS, BS, AT and HCT116 nuclear extracts.

Figure 24. The fidelity of double-strand DNA break repair 96 using normal, corrected BS, BS, AT and HCT116 nuclear extracts.

Figure25. Deletion mutations in plasmids recovered from 97 the in vitro DSB repair assay using pUC18 EcoRI-CIP as the substrate.

Figure 26. Repeat instability with novel larger and smaller alleles. 103

Figure 27. Proposed pathway for the recombinational 106 DNA repair of replication forks

Figure 28. Proposed model of the BLM helicase 108 and its interaction with mismatch repair proteins correcting a lesion on the leading strand.

Figure 29. Model for the role of BLM in DSB repair 111

List of Tables Page

Table 1- DNA substrates used in gel-shift and helicase assays 53

Table 2- The efficiency and fidelity of DNA DSB repair of 85 linearized supF20 by nuclear extracts from control and BS cells.

Chapter One. Literature Review

Cancer and Genomic Instability

Cancer occurs when normal cellular control mechanisms do not function properly leading to uncontrolled proliferation and subsequent invasion of neighboring tissues. It spreads through the body and sets up secondary growth sites in a process known as metastasis (Hanahan D and Weinberg RA, 2000).

Many different cell types from different organs can develop cancers. Normal cells become malignant by the aberrant expression of proto-oncogenes or the loss of function of tumor suppressor genes. Most oncogenes are derived from normal growth control genes such as growth factors and receptors, signal transduction proteins, transcription factors and cell cycle control proteins (Alberts et al 2002).

Expression patterns of cellular oncogenes and tumor suppressor genes can be altered by local alterations of DNA structure, by chromosomal translocations or by extensive reduplication of DNA regions including an oncogene. Understanding the molecular events responsible for these genetic aberrations is difficult because the processes of eukaryotic DNA replication, repair and recombination are not well understood. Several rare genetic disorders feature both genomic instability and a predisposition to cancer as hallmark features (Figure 1). “Reverse genetics” has proven useful in the identification of the defective genes in these genetic instability syndromes and has shown that

many of these disease genes encoded proteins implicated in DNA metabolism

Figure 1. DNA repair genes act as caretakers of the genome. Defects in DNA repair are associated with genomic instability and gross chromosomal rearrangements, leading to mutations in genes such as p53 and cancer formation. Adapted from Khanna and Jackson 2001.

(Ellis 1997). For example, helicases are mutated in persons with rare inherited human disorders that are characterized by chromosome instability and cancer such as Bloom’s syndrome, Cockayne’s syndrome, trichothiodystrophy, Werner’s syndrome and xeroderma pigmentosum. In order to understand how aberrant

DNA metabolism can lead to cancer formation in a cell, Bloom’s syndrome is an ideal disorder to investigate the relationship between rates of both genomic instability and cancer formation.

Bloom’s syndrome

Bloom’s syndrome (BS) is a rare autosomal recessive disorder first described by David Bloom, a New York dermatologist in the 1950’s (Bloom

1966). The major clinical manifestations are small stature, sun sensitive redness of the face, immunodeficiency, male infertility and the development of cancers of many different types (Figure 2). One of the defining features of the disease is the presence of chromosomal aberrations in BS cells (German 1995). Other genomic instability syndromes such as Fanconi anemia, ataxia telangiectasia and Werner’s syndrome also have elevated levels of chromosome gaps, breaks and rearrangements. Increased rates of both sister-chromatid exchange and quadriradial formations are seen only in BS cells and not in cells derived from any of these other genomic instability syndromes.

Figure 2. The clinical presentation of Bloom’s syndrome (BS). Persons with BS are small in size, sun-sensitive, immunodeficient and predisposed to cancer.

Examination of BS chromosomes gave key insights to the pathological nature of the syndrome (Figure 3). Similar to chromosomes from persons with other diseases that lead to genomic instability, BS chromosomes show increased amounts of gaps, breaks and structurally rearranged chromosomes. The two hallmark cytogenetic features of BS most likely arise due to increased rates of genetic exchange in regions of high homology. These exchanges may be between a chromatid of each of the two homologues of a chromosome pair (e.g., between the two chromosome 4s), where the points of exchange are homologous. Following such an exchange, the lesion detectable microscopically at metaphase is a symmetrical four-armed entity known as a quadriradial (Qr).

These exchanges may also be intrachromosomal, occurring between the two sister chromatids of one chromosome; such exchanges are known as sister chromatid exchange (SCE). These SCEs are detectable in a metaphase spread by appropriate cell staining and microscopy. Normal cells display less than ten

SCEs per metaphase spread, while BS cells typically display 50 to 100 SCEs per metaphase spread.

Increased chromosome gaps and breaks.

Quadriradial formations.

SCE

Increased sister chromatid exchange.

NO SCE

Figure 3. BS cells are characterized by genomic instability.

It has been suggested that the different cytogenetic anomalies seen in cells from BS persons are related to the phase of the cell cycle when the chromosome break occurs. If a break occurs in G1, the broken chromosome end would not unite with the remainder of the chromosome thus forming a gap or break. Breaks in sister chromatids during the S/G2 phase are thought to be responsible for the increased SCE seen in BS cells (Kuhn and Therman 1986).

Two loci have been studied in detail to determine if mutations are also increased in BS cells. The locus encoding HPRT on the X chromosome and the locus that determines the MN blood type on chromosome 4 showed increased numbers of mutations (Tachibana, Tatsumi et al. 1996). Other evidence for genomic instability is the detection of excessive mutations at regions in the genome consisting of highly repetitive sequence (Therman, Otto et al. 1981).

BS phenotype

Persons with BS have small body size with normal proportioning. The face is narrow with prominent ears and nasal features. Hypersensitivity to sunlight to areas of exposed skin during infancy and patchy areas of hyper- and hypo-pigmentation of the skin are common in BS. These pigmentation changes suggest somatic mosaicism in the melanocytes. The voice of a BS patient is high pitched and somewhat coarse and squeaky. There may be a slight excess of minor anatomic anomalies in BS persons, including anomalous digits, wedges

of altered color of the iris, localized areas of white hair and obstructing anomalies of the urethra (German 1995).

Many individuals with BS experience a variable degree of vomiting and diarrhea during infancy, often producing severe dehydration. Immunodeficiency of a generalized type is common in BS and shows a wide range of variation in severity ranging from mild or symptomatic to severe. Viral and bacterial infections are common in the respiratory tract and ear canals of BS patients.

Diabetes mellitus has been diagnosed in twenty of the 165 persons in the BS

Registry, and in all cases is late-onset or noninsulin-dependent type II diabetes.

Reduced fertility is seen in both men and women with BS. In males, the testes are abnormally small and display a complete failure of spermatogenesis.

Women with BS have an abnormally shortened menstrual life with chronic fertility problems. It is interesting to note that three BS females have given birth to normal children.

BS individuals generally have a restricted intellectual ability. Intelligence is generally average to low average, but can vary greatly. The intellectual ability of a person with BS is not predictable as some persons are mentally deficient while others are normal. Some BS individuals have completed college and obtained professional degrees. However even in the BS persons with normal intelligence, there is a noticeable short attention span.

BS persons are much more susceptible to neoplasia development than normal individuals of the same age. This predisposition to cancer is thought to be responsible for the reduced life span of individuals with BS, which is 28 years.

The most impressive aspects of neoplasia in BS are the increased rates at which both benign and malignant neoplasias arise, the wide variety of anatomic sites and cellular types affected, and the exceptionally early age at which time the neoplasias develop. Leukemias and carcinomas are usually responsible for deaths in BS persons. Lymphomas are the most prevalent type of hematopoietic malignancy, while solid tumors are more likely to occur in the colon, skin and breast.

Positional cloning of BLM

Traditional linkage analysis allows scientists to map genes associated with genetic diseases. However, rare autosomal recessive disorders do not often affect large numbers of individuals which hamper these types of statistical analyses. Lander and Botstein first suggested that recessive disorders could be mapped efficiently by studying DNA from consanguineous relationships. This specialized form of gene linkage is known as homozygosity mapping and was used to analyze consanguinitous families with BS. Linkage was found with loci on chromosome 15q (Ellis, Roe et al. 1994). A polymorphic tetranucleotide repeat present in an intron of the FES gene was homozygous in 25 of 26 individuals with BS whose parents were consanguineous. Woodrage et. al. described a patient with features of both BS and Prader-Willi syndrome

(Woodage, et al. 1994). Genetic analysis determined that the chromosome 15’s were isodisomic, suggesting that there had been a failure of chromosome separation in meiosis leading to two copies of a chromosome 15 with a mutation

of BLM. McDaniel and Schultz utilized BS cells as recipients for microcell- mediated chromosome transfer to show that the transfer of human chromosome

15 suppressed the elevated SCE levels in this BLM-deficient cell line (McDaniel and Schultz 1992).

The genetic instability of BS includes increased frequency of somatic recombinational events. Ellis et al. made a key observation in 1995 that helped map the BLM gene in a more timely manner (Ellis, et al. 1995). Studying SCE in

BS cell lines showed high levels of SCE in lymphoblastoid cell lines. However, a small sub-number of these cell lines demonstrated low SCE. Analysis of these high and low cell lines at polymorphic loci proximal and distal to the putative BLM locus showed that loci distal to BLM had become homozygous in low SCE lines compared to high SCE lines from the same person, while loci proximal to BLM remained heterozygous. These observations suggested that a recombinational event had occurred in the low SCE population resulting in a functional BLM gene and low SCE. Ellis et al. used the low SCE cell lines with a reduction to homozygosity to map BLM by a novel approach known as somatic crossover point (SCP) mapping. The gene responsible for BS was mapped by determining the genotype of low SCE from five different BS persons. The strategy identified the most proximal heterozygous locus and the distal locus reduced to homozygosity in the low SCE cell lines.

A strong candidate for BLM was identified by direct selection of a cDNA derived from this minimally defined 250 kB region identified by somatic crossover point mapping. Analysis of this candidate gene revealed a 4437 bp cDNA

putatively encoding a 1417-amino acid(aa) peptide with homology to the RecQ helicases, a subfamily of DExH box-containing DNA and RNA helicases. The presence of chain-terminating mutations in this candidate gene in BS persons permitted the conclusion that this gene was the disease gene for BS. Mutational analysis of the first thirteen unrelated persons with BS identified 7 unique mutations. Four of the seven mutations resulted in premature termination of the protein suggested the cause of BS is loss of enzymatic function of the BLM gene product.

Functional motifs of the BLM helicase

Analysis of the peptide sequence of the BLM gene product shows many different motifs (Figure 3). The center of the protein includes a helicase domain with seven-conserved helicase motifs present in other molecules with defined helicase activities. Very little is know about the function of these helicase motifs, with the exception of the Walker A and B sites necessary for ATP-binding and subsequent hydrolysis. All ATP-dependent helicases possess the

(GXGXGK[T/S]) which forms a P-loop necessary for ATP-binding. The second

Walker motif is also known as the DEAD or DExH box and contains the essential aspartate (D) that interacts with ATP via Mg2+ (Patel and Picha et al. 2000). Site- directed mutagenesis of either the Walker A and B sites demonstrated that these motifs were necessary for ATP-binding, hydrolysis and enzymatic function of many helicases. It should be noted that the use of primary sequence to detect

helicase motifs does not always identify polypeptides that possess strand displacement activity in vitro.

1417 1 ACIDIC MOTIFS ACIDIC nuclear staining. nuclear cells shows punctateGFP-C-terminus of BLM transfected intocos functional nuclearlocalization sequence is C-terminal.. Walker A and B boxes for ATP-binding and hydrolysis.The domain residesinthe center of the protein and contains the Figure 4. Functional motifs of helicase. the BLM POL II POL RNA HELICASE DOMAIN RecQ-Ct GFP-BLM-C HRDC The helicase The helicase NLS

Ian Hickson’s laboratory was the first to show the BLM protein has helicase activity in vitro using a classic in vitro strand displacement assay. Single

-stranded DNA is purified and a complementary radiolabeled oligo is annealled to the M13 template (Karow, et al. 1997). This DNA duplex substrate is mixed the potential helicase with ATP and co-factors crucial for the enzymatic reaction.

Helicase activity is determined by detecting the displaced oligo by native gel electrophoresis and autoradiography. Utilizing this strand displacement assay, purified human BLM protein (hBLM) was shown to possess a 3’ to 5’ DNA-DNA strand displacement activity that was dependent on ATP and Mg2++. The

ATPase activity of this helicase was stimulated by the presence of DNA.

Mutations made in the Walker A site resulted in a loss of helicase activity. This directionality is characteristic of the RecQ subfamily of helicases; E. coli RecQ, yeast Sgs-1 and the human WRN helicase also act on DNA in a 3’ to 5’ dependent manner with respect to the labeled oligo (Chakraverty and Hickson

1999). The BLM helicase also acts upon non-canonical Watson-Crick structures such as G4 DNA which is present in G-rich motifs present in the rDNA gene clusters, the immunoglobulin heavy chain switch regions and telomeric repeats

(Sun, et al. 1998).

Many nuclear proteins contain short amino acid sequences that direct them to the nucleus of a cell. These nuclear localization sequences (NLS) interact with specific nuclear import receptors to permit access to the nucleus.

For a putative NLS to be considered functional, it must pass several tests.

Deletions or mutations of nuclear localization signal in a protein should show that

the protein is excluded from the nucleus or show strictly cytoplasmic staining by immunohistochemistry or immunofluorescence. Attaching a putative NLS to a large non-nuclear protein such as β-glucosidase and showing nuclear import of the chimera suggests the NLS is functional. The bipartite (two stretches of positively charged residues) NLS in the carboxy-terminal domain of BLM was shown to be functional (Figure 4). (Heppner-Goss personal communication) At least two other putative nuclear localization sequences are present in the BLM protein (one in the helicase domain and the other in the amino terminal domain, although there is no experimental evidence that suggests these NLS are functional (Kaneko, Orii et al. 1997).

BLM and PML-bodies

The BLM helicase localizes to discrete nuclear bodies commonly known as PML (promyelocytic leukemia) bodies, a distinctive class of subnuclear domain (Zhong, et al. 1999). These domains appear by indirect immunofluorescence microscopy as punctate speckles and can be referred to as

Kremer bodies, ND 10 or PODs (PML oncogenic domains). PML was identified by analysis of a chromosome translocation breakpoint in leukemic cells from APL

(acute promyelocytic leukemia) patients. Gene-targeted disruption of PML demonstrated that PML functions as a growth and tumor suppressor in vivo and is important for multiple apoptotic pathways. BLM protein is diffusely localized in cells null for PML. In contrast, PML does form discrete foci without BLM. SCE

rates in cells null for PML displayed a two-fold increase when compared to control cells (Zhong, et al. 1999).

BLM homologs and orthologs

To understand the function(s) of BLM helicase, we can examine the function of RecQ subfamily members in DNA metabolism. The founding member of the RecQ subfamily of DNA/RNA helicases is RecQ, a 70kD polypeptide originally isolated from E. coli mutants in the RecF recombination pathway

(Umezu, et al. 1990). Mutations in the RecF pathway cause a recombination deficiency and increased ultraviolet (UV) sensitivity in a recBC shcB background.

Promoter studies of the E. coli RecQ helicase suggested that the E. coli RecQ promoter is normally repressed by the LexA repressor binding sites ten nucleotides downstream of the transcription start site. RecQ in E. coli was inducible by both UV light and treatment with mitomycin C (MMC) (Bierne,

Seigneur et al. 1997). The E. coli RecQ promoter possesses lexA repressor binding sites and is inducible by both UV light and MMC, suggesting that SOS regulation mediates RecQ gene expression (Mendonca, et al. 1995). The E. coli

RecQ protein possesses an ATP-dependent and 3’ to 5’ helicase activity in vitro.

Its helicase activity was stimulated by several single stranded-binding proteins

(SSB’s), including T4 gene 32 protein and E. coli SSB (Umezu and Nakayama

1993).

RECQL/DNA helicase Q1 is the human homologue to the E.coli RecQ protein (Puranam and Blackshear 1994). These two proteins are equivalent in

size and share homology at the amino acid level. Two isoforms of RECQL have been cloned in mice (Puranam, et al. 1995). The alpha form is expressed in a ubiquitous manner, while the beta form is expressed specifically in the testis. In humans there are at least four different isoforms of RECQL. The chromosomal location and genetic sequence of RECQL have been determined but few expression studies are reported in the literature. It is interesting to note that mutations in RECQL4 have been genetically linked to a rare autosomal recessive disorder associated with genomic instability known as Rothmund-Thomson syndrome (Kitao, et al. 1999).

The BLM helicase is similar in molecular weight to the RecQ-like helicase associated with Werner’s syndrome (WRN). The seven-helicase motifs present in the WRN helicase domain reside roughly in the center of the molecule with other significant functional motives in either the N- or C-terminal domain. The location of the NLS in the C-terminal portion of the polypeptide prevents truncated molecules from nuclear entry. Homozygous mutations of the WRN helicase result in the rare autosomal recessive disorder Werner’s syndrome (WS) associated with premature aging (Epstein and Motulsky 1996). Even though both BS and WS are caused by the loss of function of a RecQ helicase, the phenotype of each rare disorder is very different.

Sgs-1 (suppressor of slow growth phenotype) was originally identified in

Saccharomyces cerevisiae because it rescues the slow growth phenotype of the topIII mutant (Watt, et al. 1996). Yeast lacking functional topoisomerase III display a pleiotropic phenotype including slow growth, G2/M arrest and genomic

instability. Mutations of Sgs-1 suppressed this phenotype in the top3-deficient yeast. Analysis of the region responsible for the suppression of the top3 mutants revealed a gene encoding a RecQ helicase. Further studies revealed that Sgs-1 also interacts with topoisomerase II in yeast and has strand displacement activity in vitro (Watt, et al. 1995).

Rqh-1 is the fission yeast homologue of the BLM helicase. Loss of Rqh-1 function results in yeast that are more sensitive to UV, gamma radiation and the

DNA synthesis inhibitor, hydroxyurea (Stewart, et al. 1997). Overexpression of

Rqh-1 results in cells that are more sensitive to both UV and gamma radiation, and loss of S and G2-phase cell cycle checkpoint control. These results suggest that this helicase is important in both DNA metabolism and cell cycle progression.

The BLM protein exists in an oligomeric form

The quaternary structure of this recombinant protein has been determined using size-exclusion chromatography and electron microscopy (Karow, et al.

1999). Size exclusion chromatography demonstrated that the major ATPase activity eluted at an estimated molecular weight of 800 kD to 1 MegaD.

Antibodies to BLM identified the BLM helicase in these same fractions. Electron microscopy showed that the BLM helicase forms hexameric ring structures with an overall diameter of 13nm with a central hole of 3.5nm diameter. A fourfold square form was also detected, but may represent one of several oligomeric species. The fourfold symmetric images may represent an oligomeric form

distinct from the hexameric ring such as a tetramer or octamer. It was also suggested this image could be the result of the side view of a two-tiered hexameric barrel structure such as the bacteriophage T7 helicase/primase. Most helicases characterized to date act as dimers or hexamers. However, BLM is the first eukaryotic oligomeric helicase. More importantly, this suggests the BLM helicase may form an oligomer and encircle a continuous DNA strand at any internal site without the requirement for a single stranded DNA end. This function may be crucial for the BLM helicase to function in DNA replication to open replication sequences or to aid the transcriptional machinery of the cell by melting promoter sequences.

Blm and Wrn “knockout” mice

Phil Leder’s group was the first to use gene-targetting to target the endogenous Blm gene in mouse (Chester, et al. 1998). A site upstream of the helicase domain was targeted for disruption by homologous recombination in embryonic stem cells. This would disable both the helicase and NLS activity from the protein. Unlike the Bloom syndrome patient, a mouse without a functional BLM gene results in an embryonic lethal phenotype. Specifically, Blm-

/- embryos do not survive past 13.5 days past conception. These embryos also are smaller than wild type controls and exhibit developmental delays and increased rates of apoptosis. As an indication of DNA damage micronuclei were present in the white blood cells of these embryos. Sister chromatid exchange in

BLM-/- embryonic fibroblasts was increased seven fold demonstrating that loss of

helicase function leads to genomic instability, increased rates of apoptosis and embryonic lethality.

Interestingly, this group has also targeted the Werner’s gene in mice (Wu,

He et al. 1998). Homologous recombination generated the mouse helicase domain replaced by a neo-cassette. The targeting strategy employed could potentially give rise to two differently processed mutant WRN transcripts. The first transcript would produce a truncated WRN protein plus the neo cassette.

The second mutant transcript would code for the WRN protein minus helicase motifs three and four. Both of these WRN proteins are presumed to lack helicase activity. The knock out had a mildly increased rate of embryonic lethality and sensitivity to topoisomerase inhibitors. However, mice lacking functional WRN helicase display a relatively normal phenotype during the first year of life.

DNA replication

The unusual cytogenetic anomalies associated with Bloom’s syndrome suggested that there was some inherent defect in DNA metabolism present in cells derived from BS patients. Many early studies involved determining if there are notable differences in DNA regulation between normal and BS cells. Many subtle differences were found in regards to DNA regulation. DNA fork displacement rates were measured in three BS cell lines and compared to normal fibroblasts (Kapp 1982). In each of the BS cell lines tested, the DNA fork displacement rate was less than 65% of the values in the normal fibroblasts.

Nucleotide content of normal and BS cells were measured and the ATP/ADP

ratio was three times lower in the BS cells compared to control (Kenne and

Akerblom 1990). This altered nucleotide pool was hypothesized to be responsible for the genetic abnormalities of BS due to the lack of energy for DNA synthesis/repair enzymes and could be responsible for the phenotypic slow growth.

Size analysis of genomic DNA showed differences in the sizes of the DNA replication intermediates between normal and BS cells suggesting some inherent defect in DNA replication (Ockey and Saffhill 1986; Lonn, Lonn et al. 1990).

Although it should be noted that these differences in DNA replication intermediates were also detected in low SCE populations of BS cells suggesting the molecular mechanism responsible for these intermediates is not the absence of the BLM helicase but of another mutation that was generated in this mutator phenotype.

The ability of BS cells to join double strand breaks (DSB) in vivo was measured by Runger and Kraemer in 1989 (Runger and Kraemer 1989).

Linearized plasmid DNA was transfected in either BS cells or normal cells as a control. BS cells have a 3 fold lower rate of ligation efficiency when compared to the control cells. Mutation frequency was 2 to 21 fold higher in the BS cells as compared to control, and analysis of these mutations revealed point mutations to be predominating factor causing loss of transgene function with the occurrence of insertions and deletions as well.

DSB plays a crucial role in normal immunoglobulin gene processing.

Combinatorial joining occurs between the V,D, and J segments and variability in

the position of joining and nucleotide addition in heavy chain genes present putative molecular mechanisms in DNA regulation that could be altered in BS cells. These altered processes help explain the genomic instability and the impaired immune function of patients with BS.

Finally, it has been demonstrated that cells exfoliated from BS patients are ten times more likely to be micronucleated (Rosin and German 1985). Micronuclei arise when some chromosome fragment lacking a kinetochore necessarily lags at anaphase to be encompassed with its own private nuclear membrane at the time that the two main groups of telophase chromosomes are surrounded by the major nuclear membrane of each of the daughter cells.

These data suggests the BLM helicase have functional roles in DNA replication and DSB repair. Differences in DNA fork displacement rates and DSB repair functions point to specific pathways where the BLM helicase could play a functional role. Since this helicase is implicated to be important in DNA replication and repair and lack of this helicase leads to cancer, many studies have focussed on response to DNA damaging agents and subsequent cell-cycle progression.

Sensitivity of BS cells to DNA damaging agents

Many studies have utilized the BS cell lines to examine their sensitivities to different DNA damaging agents and measuring sister chromatid-exchange, cell cycle-progression and the expression patterns of proteins shown to be important in cell cycle progression and genomic stability such as p53. Giannelli

and others decided to study the effects of both UV irradiation and specific DNA polymerase inhibitors on unscheduled DNA synthesis (UDS) in 1982 (Giannelli,

Botcherby et al. 1982). This study helped lead the way to study DNA damaging agents and cell cycle progression in BS cells. UV irradiation induced more UDS in ten BS cell lines compared to control. Aphidocolin could suppress this UDS in nine of the ten BS cell lines (Krepinsky, Rainbow et al. 1980). This suggests one possible function of the BLM helicase is to somehow suppress DNA replication at extraneous sites, possibly where DNA damage has occurred. The results of the study provided the field with two directions. The first was to examine the sensitivity of BS cells to other DNA damaging agents. Secondly, the cells could be examined to determine if cell cycle-checkpoints and other proteins important in monitoring cell cycle-progression and genomic stability are functioning correctly.

Several studies have been conducted to measure the sensitivity of BS cells to the DNA damaging agent mitomycin C (MMC). MMC is an alkylating agent, which produces bulky monoadducts and interstrand crosslinks. Several studies have been performed with BS cells to measure their sensitivity to MMC by monitoring their viability and rates of SCE. Shiraishi and Sandberg were the first group to show BS lymphocytes were more sensitive to MMC by measuring viability and SCE (Shiraishi and Sandberg 1978). BS lymphocytes were much more sensitive to MMC and SCE increased at least two fold compared to normal lymphocytes. However, another group had performed similar experiments and

BS fibroblasts were much more sensitive to MMC in the cell viability assays but

did not present the increased rates of SCE (Hook, Kwok et al. 1984). It was suggested that the differences reported could be due to the different BS cells used. The first group used lymphocytes while the second group used fibroblasts.

Another possibility is that is considerable heterogeneity between the different BS patients, which is present in the varied phenotype of this disorder.

BS cells have also been shown to be sensitive to ethylating agents such as N-ethyl-nitrosourea (ENU) (Kurihara, Inoue et al. 1987). Unlike the modest increase of SCE due to MMC exposure, ENU shows a greatly enhanced effect on the BS cell lines compared to control. BS cells have been shown to be four times as sensitive to the lethal effects of ENU and SCE increased 60 fold in BS cell lines.

The first study to address specific role of BLM in a cell cycle-specific manner utilized the fact that gamma radiation of mammalian cells in G2 phase just prior to mitosis produces chromatid aberrations such as breaks and gaps (Aurias,

Antoine et al. 1985). It was hypothesized that there would be more breaks and gaps in skin fibroblasts from individuals with different types of familial cancer and from fibroblasts derived from individuals with genetic diseases that predispose them to cancer. Gamma radiation was used to induce breaks in the DNA. Cells were trapped in mitosis one and a half hours later using colcemid (this ensured that the cells counted for the study were in the G2 phase when exposed to the radiation), and mitotic cells were stained and chromatid breaks and gaps were counted. There were increased rates of chromatid breaks and gaps in the fibroblasts derived from familial cancer as well as the cancer prone syndromes

compared to control fibroblasts. Specifically, this suggests that the BLM helicase may play an active role in chromosome segregation in the G2 phase of the cell cycle.

The relationship between ultraviolet (UV) light and its effects on BS cells has been intimately studied due to sun sensitivity present in the affected individual (Evans, Adams et al. 1978). BS cells were ten times more sensitive to

UV light as measured by cell viability assays and SCE increased ten fold when exposed (Krepinsky, Rainbow et al. 1980). To help explain the increased genomic instability following exposure to UV light, the incision step utilized by the cell to remove the photoadduct was measured in BS cells compared to normal fibroblasts. Incision products were detected less frequently in BS cells suggesting a defect in either the detection of the photoadduct or the actual incision step itself is somehow inhibited or delayed due to the lack of the BLM helicase.

Absence of functional p53 or other proteins important to genome integrity and cell cycle progression could help explain the sensitivity of BS cells to DNA damaging agents. It has been well established that p53 is activated at both transcriptional and post-transcriptional levels by a variety of DNA damaging agents. Activation of p53 by DNA damage arrests the cells in the G1 phase of the cell cycle; preventing DNA replication from occurring on damaged templates.

Cells mutant in p53 function will progress normally through the cell cycle entering

S-phase even in the presence of damaged DNA. Van Laar et. Al. ascertained the p53 status in cell lines derived from hereditary disorders associated with UV

sensitivity which included BS (van Laar, Steegenga et al. 1994). First, this study gives a FACS analysis profile of BS cells compared to controls. BS cell lines show similar percentages of cells in different phases of the cell cycle. Nine out of ten BS cell lines arrest in G1 when exposed to UV light suggesting functional p53 is present. Thus the G1 checkpoint of the cell cycle is functional in most BS cell lines. It is interesting to note that rqh1 in fission yeast regulates cell cycle checkpoint control and is homologous to the BLM helicase.

More recently Judith Campisi’s group has studied the BLM helicase and its response to DNA damage (Bischof, Kim et al. 2001). The BLM helicase resides in PML-bodies during late S and G2 phases of the cell cycle. BLM colocalized with RAD51 and RPA after cells were treated with ionizing radiation.

Finally, they observed a G2 delay where BLM protein levels accumulated that was both ATM and p53 independent. Other groups have reported that BLM protein levels decrease when exposed to UVC and observe BLM phosphorylation when cells are treated with HU or UVC (Ababou, Dumaire et al. 2002). Pichierri and Franchitto showed that functional BLM is required for the localization of the

RAD50-MRE11-NBS1 complex to sites of replication arrest but is not essential in the activation of BRCA1 either after stalled replication forks or gamma irradatioin.

This group reported that the BLM protein is phosphorylated after replication arrest in an Ataxia and ATR-dependent manner, which is contrary to the Campisi report (Franchitto and Pichierri 2002).

The BRCA1-associated genome surveillance complex

The BLM helicase exists in an oligomeric form and is known to exist in a large 2-mega Dalton with other proteins intimately involved in DNA repair. This complex has been named BASC (BRCA1-associated genome surveillance complex) (Wang, Cortez et al. 2000). Members of this complex include BRCA1,

RAD50-MRE11-NBS1, MSH2 and 6, MLH1, ATM and BLM. Members of this complex can be co-immunoprecipitated from HeLa nuclear extracts and co- localize when cells are treated with specific DNA damaging agents. Both HU and ionizing radiation greatly increased the number of foci that contained both

BRCA1 and BLM compared to untreated cells. The BLM helicase also colocalized with the RAD50-MRE11-NBS1 complex when cells were exposed to

HU strongly suggesting that all these proteins work in unison to resolve these aberrant DNA structures.

It is interesting to note that many of the proteins associated with BASC possess the ability to bind abnormal DNA structures, such as double-strand breaks, base-pair mismatches, Holliday junctions, cruciform DNA and telomere repeat sequences. Therefore it is possible that these proteins may act as sensors for aberrant DNA structures that arise during DNA replication, recombination and repair. RAD50-MRE11-NBS1 and ATM may detect double- strand breaks and the mismatch repair proteins may detect distortions in the helix due to the chemical modifications of bases or the presence of IDLs

(Insertion/deletion loops). The MSH2-MSH6 heterodimer not only binds to mismatched DNA but also has affinity for Holliday junctions thereby acting as

sensors of recombination and replication fork damage. BLM has affinity for forked DNA, synthetic Holliday junctions and telomeric G4 DNA.

BASC also has signal transducers such as the ATM kinase and ATR.

This would allow for a rapid response to activate cell-cycle checkpoints through phosphorylation and activation of p53 and Chk proteins.

Protein partnering

Protein partners can provide essential clues to the function of a specific gene in the cell. The previous section of this review discussed the proteins that

BLM existed in a complex. This section of the review is dedicated to proteins that directly interact with BLM or possess a functional interaction.

The first protein shown to co-localize and co-immunoprecipitate with the

BLM protein was human topoisomerase IIIα (Johnson, Lombard et al. 2000; Wu,

Davies et al. 2000). Topoisomerases are enzymes known to relieve torsional stress introduced into DNA when complemetary strands are separated by helicases. This interaction was expected, since SGS1 and yeast TOP3α have been shown to interact both physically and genetically. BLM and human topoisomerase IIIα are able to bind to each other in vitro showing that the interaction is direct. Topo IIIα has two different binding regions in the BLM helicase. The N-terminal binding region is aa 1-212 and the C-terminal binding region is aa 1266-1417. BLM and yeast TOPIII were shown to interact genetically similar to the SGS1 TOPIII interaction. The slow growth phenotype seen in the top3 background can be generated by a BLM contruct with the

topoisomerase interaction domains deleted in a sgs1 backgound. These data strongly suggest that topoisomerases are required to interact with helicases directly to resolve the topological substrates that the helicases generate during unwinding duplex DNA.

Replication protein A (RPA) is the eukaryotic version of bacterial single- stranded binding protein and is intimately associated with many different aspects of DNA metabolism (Brosh, Li et al. 2000). Far western analysis showed a direct interaction between the BLM helicase and 70-kDa subunit of RPA. RPA stimulates BLM helicase activity and is necessary for BLM to be processive on long tracts of DNA (> 250bp) RPA and BLM colocalize in meiotic prophase of mouse spermatocytes further suggesting an in vivo role for this interaction

(Walpita, Plug et al. 1999).

The RAD51 protein is key to homologous recombination in eukaryotes and

Hickson’s group was the first to detect the interaction between the BLM helicase and RAD51 via a yeast two-hybrid screen (Wu, Davies et al. 2001). The C- terminal domain of BLM (aa 966-1417) interacted with human RAD51 protein.

Far western blots determined that RAD51 interacted with both the N-terminus of

BLM (aa 1-212) and C-terminus (aa 1317-1417). BLM and RAD51 co-localize to discrete foci in response to gamma-irradiation. BLM and RAD51 exist as a complex in cells arrested with aphidicolin. SGS1 and yeast RAD51 were shown to interact demonstrating that this interaction is evolutionarily conserved.

p53 and BLM interact both in vitro and in vivo and p53 mediated apoptosis is reduced in BS cells (Wang, Tseng et al. 2001). Restoring BLM expression to

BS cells allows p53-mediated apoptosis to occur. p53 also attenuates BLM’s ability to unwind Holliday junctions in vitro. p53 also has been shown to associate with recombinative repair complexes during S phase with MSH2, however the presence of BLM has not been examined (Zink, Mayr et al. 2002).

Helicases are important tools for the cell to regulate DNA replication, repair, recombination and gene transcription. Understanding the function of these enzymes and how the cell regulates their activities will provide key insights into how mutations can spontaneously arise due to errors in nucleic acid metabolism, which in turn may lead to cancer formation. Many helicases have been identified by either traditional biochemical purification or by positional cloning. Many of these helicases unwind duplex nucleic acid molecules in vitro, but very little is known about their specific functions or activities in the cell. It is currently hypothesized that different families of helicases function by a common molecular mechanism and different cell types will utilize the mechanistic features of the particular helicase. An example of this would be the use of different types of helicases in the immune system to regulate and generate immunoglobulin diversity through genetic means. Therefore understanding the regulation of the

BLM helicase will provide clues about how the cell regulates the function of BLM and possibly other eukaryotic RecQ helicases, that when mutated lead to rare autosomal recessive disorders associated with increased incidents of cancer formation and genomic instability.

Chapter Two

Thesis Rationale: Examining the Role of the BLM helicase in Mismatch repair and Double-Strand Break Repair.

The aim of this work is to determine the role of the BLM helicase in regards to DNA repair. Many other helicases play active roles in DNA repair.

RecQ helicase is responsible for regulating recombination in E. coli via the RecF repair pathway. Another example is the UvrD helicase which is necessary for strand displacement during mismatch repair in E. coli . Eukaryotic helicases also play functional roles in DNA repair.

Absence of RecQ helicases in mammalian cells results in increased levels of recombination and increased mutation frequency. WS cells are sensitive to the

DNA damaging agent 4-nitroquinone and BS cells are mildly sensitive to a variety of DNA-damaging agents. The BLM helicase exists in a 2-megadalton complex with other proteins implicated in mismatch repair and double-strand break repair.

Therefore we hypothesized that the BLM helicase may play crucial roles in mismatch and double-strand break repair. To further understand the role of the

BLM helicase in DNA repair, we performed yeast-two hybrid screens to identify direct protein partners of the BLM helicase that could directly implicate this RecQ helicase in specific pathways of DNA repair and we performed helicase an in vitro double-strand break and mismatch repair assays to further understand the role of the BLM helicase in DNA repair.

Chapter Three

Materials and Methods

Cell Culture

HCT116 cells were purchased from the American Type Culture Collection

(ATCC). HCT116 cells were cultured in McCoy’s 5a Medium (Life Technologies,

Inc.). The SV40-transformed BS fibroblast cell line, GM08505C, was purchased from Coriell Laboratories (Camden, NJ). This cell line was derived from an

Ashkenazi Jewish female, homozygous for the BLMASH allele (a 6 bp deletion/7 bp insertion at nucleotide 2281). Cells were grown in Dulbecco’s Modified

Minimal Essential Medium (Life Technologies), supplemented with 10% fetal bovine serum (FBS, Hyclone). The SV40-transformed ataxia telangiectasia (AT) fibroblast cell line, AT5BIVA, was purchased from the Human Genetics Mutant

Cell Repository (Camden, NJ). The SV40-transformed normal human fetal lung fibroblast cell line (W138VA13) was purchased from American Type Culture

Collection (Rockville, MD) and used as a control. Both AT and normal cells were grown in Eagle’s Minimal Essential Medium (Life Technologies), supplemented with 10% FBS, 100 U/ml penicillin,100µg/ml streptomycin.

Reagents and Enzymes

Restriction endonucleases and T4 DNA ligase were purchased from New

England Biolabs. PCR was performed using Pwo polymerase from Roche

Molecular Biochemicals. Oligonucleotides were synthesized on an Applied

Biosystems synthesizer and DNA sequencing was performed at the University of

Cincinnati DNA Core Laboratory. Plasmids were purified using Qiagen purification kits. In vitro transcription and translation (IVTT) reactions were performed using the Promega TNT-coupled rabbit reticulocyte lysates system.

Radiolabeled [35S]methionine from PerkinElmer Life Sciences was used to label proteins. To detect BLM by western blot, a polyclonal antibody to BLM was purchased from Novus. To immunoprecipitate BLM, goat polyclonal antibodies from Santa Cruz, Both BLM-1 and 2, were used, while MLH1 western analyses and immunoprecipitations were performed with monoclonal antibodies from

PharMingen. An α-replication protein A 70-kDa subunit antibody (Ab-1) and protein A/G-agarose were purchased from Calbiochem.

Nuclear Extracts

Nuclear extracts were prepared from cell lines based on the modified method of Lopez and Coppey as described in Li et al. with all the steps performed at 4°C. 1 x 108 cells were harvested, washed in PBS, and resuspended in 4 mL of hypotonic buffer [20 mM Hepes (pH 7.5), 5 mM KCl, 1.5 mM MgCl2 and 1mM DTT]. Cells were gently lysed using a Dounce homogenizer and lysis was monitored by trypan blue staining. The released intact nuclei were recovered by centrifugation at 2000 x g for 1 min. The nuclei were extensively washed before the nuclear envelope was broken by three cycles of freezing and thawing. Debris was pelleted and soluble nuclear protein was recovered from the supernatant by ammonium sulfate precipitation. The

precipitate was resuspended in 50 mM Tris (pH 7.5), 0.1 mM EDTA, 10 mM 2- mercaptoethanol, 0.1 mM phenylmethylsulfonyl fluoride, and 10% glycerol and dialyzed overnight against the same buffer. Nuclear extracts were snap frozen and stored in small aliquots at –80°C.

Expression construct generation and characterization

BLM-pET constructs for protein expression were derived by PCR using the pOPRSVI-BLM construct as a template (Ellis, Proytcheva et al. 1999). The pET30A-BLM-C construct was generated by PCR amplification with Pwo polymerase and cloning of BLM nucleotides 2931-3995 directionally with the

EcoRI and BamHI sites present in the polylinker. The pET24D-BLM-C construct was generated by PCR amplification and cloning of BLM nucleotides 3063-4325 into the NcoI site in the polylinker. The pET24D-BLM-N (nucleotides 75-1838) and the pET24D-BLM-H (nucleotides1839-3077) constructs were generated by

PCR amplification of BLM and cloning into the NheI site. Full-length MLH1 was generated by PCR from the plasmid identified in the yeast two-hybrid screen and subsequently was cloned into the pRSETB plasmid. IVTT reactions were analyzed by western analysis with antibodies specific for either BLM or MLH1.

Products from IVTT reactions were also verified by immunoprecipitation.

Yeast-two hybrid screening

A yeast two-hybrid screen was performed using the Matchmaker Two-

Hybrid System from Clontech. DNA encoding the C-terminus of BLM

(nucleotides 3108-4319) was cloned into the pAS2 vector by PCR. The C- terminus of BLM was used as bait to screen a human B-lymphocyte cDNA library also purchased from Clontech. Transformants were plated onto a trp-, leu-, his- plus 50 mM 3-AT plates. Plates were incubated at 30°C for 10 days and the growth of the yeast monitored. β-galactosidase assays for potential interacting clones were performed according to the instructions provided by Clontech

Matchmaker Two-Hybrid system.

Isolation and renaturation of BLM-C

BLM-C protein was expressed from pET30A-BLM-C in E. coli BL21

(DE3) pLysS induced with 0.5 mM isopropyl-1-thiogalactopyranoside, collected by centrifugation, rapidly frozen and stored at –80°C. All subsequent procedures were performed at 4°C. Cells were thawed and resuspended in ice cold 10 mM

Tris-HCl pH 7.5, 150 mM NaCl, 1 mM KCl supplemented with a bacterial protease inhibitor cocktail (Sigma). Cells were passed through a french press once at 20,000 pounds per square inch. Cell lysates were centrifuged at 15000 x g for one hour. Since the majority of the BLM-C protein resided in the insoluable fraction, the pellet was resuspended in his-binding buffer (5 mM imidazole, 500 mM NaCl with 8 M urea, centrifuged as before and passed through a 0.4 micron filter before application to a Ni(II)-agarose column (His-Bind Resin, Novagen).

Protein was eluted with his-binding buffer containing 8 M urea and 250 mM imidazole. Fractions were separated by 15% sodium dodecyl sulfate- polyacrylamide gel electrophoresis (SDS-PAGE) and stained with Coomassie

blue to determine yield and purity. Fractions that were more than 80% pure were dialyzed in a stepwise fashion in 10 mM Tris-HCl pH 7.5, 150 mM NaCl, 10% glycerol with decreasing amounts of urea to obtain soluable protein.

Approximately 30% of the total protein remained in solution after removal of urea.

Far western analysis

Twenty-five µg of purified BLM-C terminus protein was resolved by 15%

SDS-PAGE. The protein was transferred to PVDF membrane (Millipore) in transfer buffer at 90V for 1 hour. The membrane was stained with Ponceau S to detect the protein and this portion of the membrane was excised. The membrane and bound protein were blocked in 5% non-fat dry milk (NFDM) in tris-buffered saline (TBS) for 1 hour and incubated with 25 µg of nuclear extracts for 1 hour on ice. The membrane was washed four times with TBS-Tween 20

(0.1%) and the bound proteins eluted by boiling in 1X SDS-PAGE loading buffer.

These samples were separated by SDS-PAGE, transferred to PVDF membrane and probed with an α-MLH1 antibody (Pharmigen) to detect MLH1.

IVTT immunoprecipitations

IVTT (Promega) reactions with [35S]-methionine were performed with pET24D-BLM-C or pRSETB-MLH1 using purified DNA (Qiagen) according to the manufacturer’s recommendations. Thirty µl of each IVTT reaction was added to a total volume of 500 µl of binding buffer (20 mM Tris pH 7.5, 10% glycerol, 150 mM NaCl, 5 mM EDTA, 1 mM dithiothreitol, 0.1% Tween-20, 1 mM

phenylmethylsulfonyl fluoride and mammalian protease inhibitors (Sigma)).

Proteins were incubated for 1 hour to allow binding before they were precleared with 25 µl of protein A/G beads. Five µg of primary antibody and 25 µl of protein

A/G were used in each reaction. The beads were washed four times with binding buffer and then boiled. Proteins were separated by 10% SDS-PAGE. The [35S] signal was intensified with Enhance (NEN-DuPont) following the instructions of the manufacturer.

Mixed lysate immunoprecipitations

Full-length human BLM cDNA with a hexahistidine tag at the 5’ end was cloned into the pFASTBAC-HTc vector purchased from Gibco-BRL to create the plasmid pFASTBAC-HTc-BLM . Virus production for insect cell infection was performed according to the protocol provided by the Bac-To-Bac Baculovirus

Expression Systems (Gibco-BRL). Insect cells (Sf21) were infected with the viral stock pFASTBAC-HTc-hBLM with a multiplicity of infection of 2. Infected insect cells were incubated at 30° C for 48 hours and lysed in 50 mM Tris-HCl pH 8.5, 5 mM 2-mercaptoethanol, 100 mM KCl, 1mM PMSF, 1% Nonidet P-40 at 4°C.

Insoluable material was removed by centrifugation at 10,000g for 30 minutes at

4° C. Insect cell lysates were mixed with K562 human cell nuclear extracts (NE), which contain MLH1. Equal volumes of insect cell and K562 NE were precleared with pansorbin A (Calbiochem). Either 10 µg of α-MLH1 (Pharmigen) or 10 µg of

α-BLM-1 (Santa Cruz) antibody was used to immunoprecipitate MLH1 or BLM respectively. Fifty µl of protein A/G agarose was added, incubated for 1 hour and

washed extensively with IP wash buffer (20 mM Tris-HCl pH 8.0, 0.1 M NaCl, 1 mM EDTA, 1% NP-40). The precipitated proteins were eluted into 1X SDS-

PAGE buffer and separated by 10% SDS-PAGE. Western blot analysis was performed to detect the presence of BLM, MLH1 or replication protein A (70kD subunit) in the immunoprecipitates.

In vivo immunoprecipitations

Five mg of K562 NE was precleared with 50 µl of protein A/G beads.

Fifty µg of BLM affinity-purified antibodies (Santa Cruz BLM-1 and -2) were added to the precleared K562 NE and rotated for 2 hours at 4° C. One hundred fifty µl of protein A/G agarose beads was added to the mixture and rotated for an additional 2 hours. The immunoprecipitates were washed with 1 ml of IP wash buffer four times. The precipitated proteins were eluted into 1X SDS-PAGE buffer and separated by 8% SDS-PAGE. Western blot analysis was performed to detect BLM and MLH1.

Mismatch repair assay

Preparation of cell free extracts and mismatched substrates and procedures for measuring mismatch repair were as described (Umar, Boyer et al.

1994). 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 lacZα 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. Substrate DNA was recovered and introduced into E. coli NR9162 (mutS) via electroporation and plated to score plaques as described [20]. Repair efficiency is expressed in percent as 100 X (1 minus the ratio of the percentages of mixed bursts obtained from extract-treated and untreated samples).

supF20 DSB Repair Assay

SupF20 was a generous gift from Michael Seidman and Michael Lin

(National Institute on Aging, National Institutes of Health, Baltimore, MD).

SupF20 is a derivative of the pZ189 plasmid and carries an ampicillin resistance gene and a modified Escherichia coli supF suppressor tRNA gene that serves as a mutagenesis marker. An insertion, including a BssHII restriction endonuclease site, in the 5’ region of the tRNA sequence destroys supF function. Function can only be restored when a recombinational event across a duplicated microhomology patch results in a specific 11bp deletion. In this study, a DSB was generated in purified supF20 DNA at the BssHII restriction endonuclease site. Linear DNA was separated from circular plasmid by agarose gel electrophoresis and purified by organic extraction. To assess the DSB repair efficiency of nuclear extracts, 50 µg of protein from each extract were incubated

for one hour at 30°C with 1µg of linear supF20 in 50 µL of DSB repair buffer [65.5 mM Tris (ph 7.5), 10 mM MgSO4, 1 mM ATP, 91 nM EDTA and 9.1% glycerol].

The reaction was stopped by the addition of EDTA to a final concentration of 20 mM and by incubation in RNAse A (Life Technologies, Inc.) and proteinase K

(Roche). DNA was purified by organic extraction and 1-10ng were used to transform E. coli MBM7070, a strain carrying an amber mutation in the lacZα gene. This mutation can be overcome by functional supF resulting in β- galactosidase synthesis. Transformants were selected on LB plates containing

100 µg/mL ampicillin and 40 µg/mL X-gal. Transformation efficiency, reflecting

DSB repair efficiency, was expressed as the total number of colonies per ug of

DNA. Fidelity of repair was determined by calculating the frequency of supF20 mutants over the total number of transformed colonies. SupF20 mutants with restored supF function were identified as blue colonies on X-gal-containing plates.

pUC18 DSB Repair Assay

pUC18 carries an ampicillin resistance gene and the lacZα gene. A DSB was generated in purified pUC18 DNA at the EcoRI restriction endonuclease site, disrupting the lacZα gene. Linear DNA was separated from circular plasmid by electrophoresis and purified from agarose by organic extraction. For experiments removing the terminal phosphate at the DSB site, DNA was incubated with 1U alkaline calf intestinal phosphatase (Life Technologies, Inc.) for one hour at 37°C. To assess the DSB repair efficiency of the nuclear

extracts, 50 µg of protein from each extract were incubated for one hour at 30°C with 1 µg of linear pUC18 in a 50 µL reaction mixture of DSB repair buffer. The reaction was stopped and DNA purified as in the supF20 DSB repair assay above. DNA (1-10 ng) was used to transform E. coli DH5α. Transformants were selected on LB plates containing 100 µg/ml of ampicillin and 40 µg/mL of X-gal.

Transformation efficiency, reflecting DSB repair efficiency, was expressed as the total number of colonies per µg of DNA. Fidelity of repair was determined by calculating the frequency of lacZα mutants over the total number of transformed colonies. LacZα mutants were identified as white or light blue colonies on X-gal containing plates. Statistical significance was determined using the students’ t test. Mutant colonies were picked and restreaked onto fresh LB plates containing 100 µg/mL ampicillin and 40 µg/mL X-gal. DNA was purified from overnight cultures using a Qiagen miniprep kit and sequenced using an 18 nt primer (TGAGAGTGCACCATATGC) located at nucleotide 172 of the pUC18c plasmid. Sequencing was performed by the Univeristy DNA Core Laboratory at the College of Medicine (Cincinnati, OH) using an Applied Biosystem Inc. 377 automated sequencer and dye terminator sequencing chemistry.

yBLM expression and purification

The yeast construct pJK1 (to express hBLM in yeast) was a gift from Ian

Hickson (Karow, Chakraverty et al. 1997) and yeast strain JEL1 (MATα leu2 trp1 ura3-52 prb1-1122 pep4-3 dhis3::PGAL10-GAL4) was provided by Jim Wang.

(Austin, Marsh et al. 1995) Briefly, the human BLM cDNA was cloned into the

pYES2 (Invitrogen). The N-terminus of BLM was modified to add a yeast Kozak consensus sequence and the first five codons of the TOP2 open reading frame.

The C-terminus was modified to add a hexahistidine tag to aid in purification.

JEL1 yeast transformed with pJK1 were grown in –Ura yeast media containing 2% glucose at 30°C to an OD600 of 0.6. Yeast were centrifuged at

3000 x g for six minutes, resuspended in growth media –Ura plus 2% galactose of equal volume and grown for 24 hours at 30°C. The yeast cells were pelleted as before. All subsequent steps were performed at 4°C. The yeast pellet was resuspended in Buffer A: 50 mM KPO4 pH=7.0 500 mM KCl 10% glycerol.

Mammalian protease inhibitor cocktail was added at 1/500 (Sigma P-8340).

Typically, 35 mL of buffer A was used to resuspend a one liter culture of yeast.

To lyse the yeast cells a French press was employed. Yeast were loaded into cold French press and 20,000 psi was applied. The French press was opened and the lysate was collected in a dropwise fashion. The yeast lysate was centrifuged at 20,000 x g for 30 minutes to remove insoluable material. While the sample was spinning, NTA resin (Qiagen) was prepared. The column was charged with five bed volumes of 1X charge buffer (50 mM NiSO4). The column was equilibrated with 3 bed volumes of 1X binding buffer (Buffer A plus 50 mM imidizole). The supernatent was mixed with the equilibrated resin and allowed to bind for 2-5 hours. The sample was returned to the column and the flow through was collected. All fractions were saved including flow through for western analysis to determine if BLM protein was present. The column was washed with

1X binding buffer until baseline OD280 is achieved. BLM was eluted as a single

step with Buffer A plus 500 mM imidizole. One mL fractions of this single step elution were collected. The fractions containing BLM protein were dialyzed into

Buffer Z: 60 mM Tris-HCl pH=7.5, 100 mM KCl, 1mM EDTA, 1mM DTT and 10% glycerol. The protein sample was then concentrated on a YM30 spin column

(Millipore), aliquoted, snap frozen in liquid nitrogen and stored at –80°C.

Preparation of helicase substrates

Single-stranded DNA oligomers were 5'-end-labeled with 32P according to

Sambrook et al. To prepare the partial DNA duplex for helicase assays, the labeled oligomer was mixed with a 2-fold molar excess of a complementary

unlabeled DNA oligomer in 10 mM Tris-HCl buffer, pH 8.0, 10 mM MgCl2; 1mM

EDTA with a mineral oil overlay to prevent evaporation; following denaturation at

100 C for 15 min, the DNA was allowed to anneal by slow cooling to room temperature. The annealed substrate was resolved from the unannealed radioactive oligo by agarose gel electrophoresis (4% non-denaturing 1X TBE (90 mM Tris base, 90 mM boric acid, 1.0 mM EDTA)). After adequate separation, the annealed substrate was excised from the gel and extracted from the gel by the

‘freeze-squeeze’ method. The gel slice was sliced into small pieces with a razor blade and smashed up in between parafilm. The gel material is placed in a microfuge tube with 500 µl of phenol/chloroform and 300 µl of TE. This tube is vortexed and placed at –80°C until completely frozen. The tube was spun at

10,000 x g for 20 minutes at room temperature. The aqueous phase was

removed and extracted again with chloroform/isoamyl alcohol 49:1. The DNA was precipitated with sodium acetate/ethanol and resuspended in TE.

Table I- Synthetic substrates used in gel-shift and helicase assays

DO substrate -also referred to as double overhang substrate or WRN substrate has both 5’ and 3’ single stranded regions

20 mer 5'-d(CGCTAGCAATATTCTGCAGC)-3' 5’ 3’ 3’ * 5’ 46 mer 5'-d(GCGCGGAAGCTTGGCTGCAGAATATTGCTAGCGGGAATTCGGCGCG)-3'

Blunt end duplex- 50mer