ACTAUNIVERSITATISUPSALIENSISUPPSALA2006

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 113

Analysis of Genetic Alterations inPatients Affected withNeurofibromatosis Type 2 and itsAssociated Tumors

CAISA MARIE HANSSON

ISSN 1651-6206ISBN 91-554-6477-7urn:nbn:se:uu:diva-6511

Till

mamma och pappa

List of publications

I. Caisa M. Hansson, Haider Ali, Carl E.G. Bruder, Ingegerd Fransson, Sindy Kluge, Björn Andersson, Bruce A. Roe, Uwe Menzel and Jan P. Dumanski. Strong conservation of the human NF2 locus based on sequence comparison in five species. Mamm Genome. 2003 Aug; 14(8):526-36.

II. Caisa M. Hansson , Patrick Buckley , Giedre Grigelioniene , Arkadiusz Piotrowski , Caroline Jarbo, Kiran Mantripragada, Anders R. Hellström , Tiit Mathiesen , Jan P. Dumanski. Comprehensive genetic and epigenetic analysis of sporadic meningioma for macro-mutations of 22q and micro-mutations within the NF2 locus. Manuscript.

III. Teresita Díaz de Ståhl, Caisa M. Hansson, Cecilia de Bustos, Kiran K. Mantripragada, Arkadiusz Piotrowski, Magdalena Benetkiewicz, Caroline Jarbo, Leif Wiklund, Tiit Mathiesen, Gunnar Nyberg, V. Peter Collins, D. Gareth Evans, Koichi Ichimura and Jan P. Dumanski. High-resolution array-CGH profiling of germline and tumor- specific copy number alterations on chromosome 22 in patients affected with schwannomas. Human Genetics. Nov; 35-44, 2005

IV. Patrick G. Buckley*, Kiran K. Mantripragada*, Teresita Díaz de Ståhl*, Arkadiusz Piotrowski*, Caisa M. Hansson*, Hajnalka Kiss, Dave Vetrie, Ingemar T. Ernberg, Magnus Nordenskjöld, Lars Bolund, Markku Sainio, Guy A. Rouleau, Michihito Niimura, Andrew J. Wallace, D. Gareth R. Evans, Gintautas Grigelionis, Uwe Menzel, Jan P. Dumanski. Identification of genetic aberrations on chromosome 22 outside the NF2 locus in schwannomatosis and neurofibromatosis type 2. Human Mutation, 26(6): 540-9, 2005.

*These authors should be considered as joint first authors.

Contents

Introduction...................................................................................................11Cancer ......................................................................................................11

Oncogenes ...........................................................................................11Tumor suppressor genes ......................................................................12Cellular mechanisms in cancer ............................................................15

Genetic variation in the human genome...................................................16Chromosomal imbalances....................................................................17Copy number variation ........................................................................19Single nucleotide polymorphism (SNP) ..............................................20

Epigenetics ...............................................................................................21Chromosome 22 .......................................................................................23Neurofibromatosis type 2 .........................................................................23

Clinical aspects of Neurofibromatosis type 2 (NF2) ...........................23The NF2 gene and protein ...................................................................24The genotype and phenotype in NF2...................................................25

Tumors associated with NF2....................................................................27Schwannoma........................................................................................27Meningioma.........................................................................................27Ependymoma .......................................................................................28

Schwannomatosis .....................................................................................28

Methods ........................................................................................................30Evolutionary comparison of DNA sequence............................................30DNA sequencing and mutation analysis ..................................................32Microarray-based comparative genomic hybridization (array-CGH) ......33DNA methylation detection by Pyrosequencing ......................................35

Aims of the study ..........................................................................................38

Results and discussion ..................................................................................39Paper I. Strong conservation of the human NF2 locus based on sequence comparison in five species .......................................................................39Paper II. Comprehensive genetic and epigenetic analysis of sporadic meningioma for macro-mutations of 22q and micro-mutations within the NF2 locus .................................................................................................40

Paper III. High-resolution array-CGH profiling of germline and tumor- specific copy number alterations on chromosome 22 in patients affected with schwannomas ...................................................................................41Paper IV. Identification of genetic aberrations on chromosome 22 outside the NF2 locus in schwannomatosis and neurofibromatosis type 2...........42

Future perspective.........................................................................................44

Acknowledgements.......................................................................................46

References.....................................................................................................48

Abbreviations

ANILFR Average Normalized Inter-Locus Fluorescence Ratio

Array-CGH Array-based comparative genomic hybridization

APC Adenomatous polyposis coli BAC Bacterial artificial chromosome bp base pair CIN Chromosomal instability CML Chronic myeloid leukemia CNV Copy number variation CNP Copy number polymorphism DNA Deoxyribonucleic acid ERM Ezrin, Radixin, Moezin protein family FGF Fibroblast growth factors GLI Glioma-associated oncogene HIF1 Hypoxia-inducible factor-1 LCR Low copy repeat DNA sequences LOH Loss of heterozygozity kb Kilo base pairs MMR Mismatch repair Mb Mega base pairs NAHR Non-homologous recombination NF2 Neurofibromatosis type 2 gene NF2 Neurofibromatosis type 2 syndrome NHEJ Non-homologous end joining PCR Polymerase chain reaction PI3K Phosphatidylinositol 3-kinase RB Retionblasroma RTK Receptor tyrosine kinase SMAD SMA- and MAD-related protein SNP Single nucleotide polymorphism TSG Tumor Suppressor Gene 22q long arm of human chromosome 22 VEGF Vascular endothelial growth factor WHO World health organization

11

Introduction

CancerCancer is caused by alterations in the control and activity of genes that regulates cell proliferation, differentiation and death. Unlike disorders such as cystic fibrosis or Duchenne muscular dystrophy, in which mutations in one gene can cause the disease, there is no single gene defect that causes cancer. Mutations in cancer associated genes may be inherited through germline and result in cancer predisposition or arise by somatic mutation. Mammalian cells have multiple mechanisms that protects against potentially lethal effects caused by mutations in ‘cancer genes’. A census from the literature report 291 genes that are mutated and causally implicated in cancer development 1. Somatic mutations have been shown in 90% of these genes, germline mutations have been identified in 20%, and 10% have displayed mutations both in somatic cells and in germ line 1. Genes that predisposes for cancer are divided into oncogenes and tumor suppressor genes. A subclass of tumor suppressor genes takes care of DNA repair and chromosomal stability (caretaker genes).

Oncogenes Mutated proto-oncogenes are constitutively active or active under conditions where the wild-type allele is silent. Activation of an oncogene can be caused by a variety of mechanisms, such as chromosomal rearrangement, gene amplification or mutations affecting crucial residues that regulate the activity of the gene product. Cancers of the hematopoetic system and connective tissues (such as leukemias, lymphomas and sarcomas) are often characterized by specific chromosomal translocation. These translocations may create a chimeric protein product or juxtapose a gene to the regulatory elements of another gene. The first discovery of such a translocation was the Philadelphia chromosome that is a result of a balanced reciprocal 9;22 translocation commonly found in patients affected by chronic myeloid leukemia 2. The chromosome 9 breakpoint is located in the ABL oncogene which is fused with the BCR gene on chromosome 22, creating a tyrosine kinase with abnormal transforming properties. Oncogenes may also be activated by amplification, either by insertions of an amplified DNA segment in a normal chromosome or as small paired chromatin bodies

12

(double minutes) that are separated from the normal chromosome. An oncogene can also be activated by a point mutation as in the case of the RAS-family of oncogenes. These proteins mediate signaling by G-coupled receptors. Binding of the ligand to its receptor triggers binding of GTP to the RAS protein. The GTP-RAS complex transmits the signal onwards and in the final step activates latent transcription factors. The mutant RAS protein has reduced GTPase activity leading to an excessive response to cellular signaling. An activating mutation in one allele of an oncogene is in general sufficient to give the cell a selective growth advantage. Oncogenes can be roughly divided into five classes: secreted growth factors (e.g. SIS), cell surface receptors (e.g. EGFR), Tyrosine kinases (e.g. SRC), membrane associated G-proteins (e.g. RAS genes) and DNA-binding nuclear proteins/transcription factors (e.g. MYC genes) (Table 1).

Table 1. A selection of oncogenes. Gene Protein function Syndrome/major tumor types

SIS Platelet derived growth factor -chain Glioma, fibrosarcoma

KS3 Member of the FGF family Kaposi's sarcoma

EGFR Protein tyrosine kinase Squamos cell carcinoma

TRK Protein tyrosine kinase Colon carcinoma

RET Protein tyrosine kinase Carcinoma of the thyroid, multiple endocrine neoplasia type 2A/2B

SRC Protein tyrosine kinase Colon carcinoma

BCR/ABL Protein tyrosine kinase Chronic myeloid leukemia

H-RAS GTP-ase Colon, lung, pancreas carcinoma

K-RAS GTP-ase Acute myeloid leukemia, thyroid carcinoma, melanoma

N-RAS GTP-ase Carcinoma, melanoma

N-MYC Transcription factor Neuroblastoma, Lung Carcinoma

L-MYC Transcription factor Lung carcinoma

Tumor suppressor genes Tumor suppressor genes (TSG) are a group of genes that originally were thought to be involved only in regulation of cellular proliferation. However, it has become clear that these genes are involved in a variety of pathways in the cell (Table 2). Mutations in a TSG cause a reduced activity of the gene product. Inactivation of a TSG may arise from missense mutations at residues that are essential for activity, from mutations that truncate the protein, from insertions or deletions that alters the open reading frame or

13

epigenetic silencing. Usually, mutations of both alleles are required for inactivation of a TSG according to Knudsen’s two hit hypothesis 3 (Fig. 1). It was observed that retinoblastoma in its inherited form developed at an earlier age with a higher incidence of bilateral tumors. It was proposed that the predisposing mutation was inherited through germline and that a second, somatic mutation of the other allele was required for oncogenesis. Both inactivating mutations has to occur in somatic cells in the non-hereditary cases and therefore it will take longer time for tumors to develop 3.

However, some TSGs have been described to exert a growth advantage when only one allele is inactivated (haploinsufficiency) 4. For instance, individuals affected by familial platelet disorder are germline carriers of a heterozygous mutation in the CBFA2 gene and have defects both in number and in function of platelets. Approximately 30-50% of these mutation carriers develop acute leukemia without loss of the wild-type CBFA2expression 5. Haploinsufficiency have also been demonstrated in a Smad4(+/-) mouse model.These mice developed gastric cancer through a multi step process, in which gastric polyp’s progress to carcinoma in situ. Loss of the wild-type Smad4 allele was not detected until the late stages of cancer progression 6.

Fig. 1. Mechanisms for inactivation of a tumor suppressor gene. A tumor suppressor gene requires the inactivation of both alleles to become silenced.

Caretaker or stability genes are a type of TSGs that are responsible for repairing mistakes made during normal DNA replication or damage induced

14

by exposure to mutagens. They may be divided into mismatch repair- (MMR), nucleotide-excision repair- (NER) and base-excision repair- genes (BER). These genes also control processes that are responsible for mitotic recombination and chromosomal segregation (e.g. BRCA1, ATM). The task for caretaker genes is to keep genetics alteration to a minimum. Consequently, when these genes become inactivated it allows mutations in other genes to accumulate 7.

Table 2. A selection of tumor suppressor genes. Gene Normal protein

function Familial disorder Major tumor types Pathway

*APC Regulation of cell

adhesion and signaling Adenomatouspolyposis coli

Colorectal cancer APC

ATM DNA damage sensor, cell cycle regulation

Ataxiateleangiectasia

T-cell leukemia CIN

BRCA1 DNA repair Breast cancer Breast and ovarian cancer

CIN

BRCA2 DNA repair Breast cancer Breast cancer CIN CDKN2A Negative regulator of

CDK cyclin complexes Familial melanoma Melanoma pancreatic

cancer RB

MLH1 DNA mismatch repair HNPCC Colon cancer MMR MSH2 DNA mismatch repair HNPCC Colon cancer MMR NF1 GTPase activating

protein for Ras Neurofibromatosis 1 Neurofibroma,

pheochromocytomas RTK

NF2 Cell morphology and membrane signaling

Neurofibromatosis 2 Schwannoma, meningioma,ependymoma

Rac1/JNK

TP53 Transcription factor, checkpoint at the G1/S stage in the cell cycle

Li-Fraumenisyndrome

Sarcoma, breast cancer, other tumors

p53

PTCH Hedgehog signalling Gorlin syndrome Basal cell carcinoma GLI PTEN Lipid phosphatase Cowden syndrome prostatic cancer,

glioblastoma, other tumors

PI3K

RB1 Transcriptional corepression

Retinoblastoma Retinoblastoma, osteosacoma,lymphomas

RB

SMAD4 TGF- signaling (transcription factor)

Juvenile polyposis Pancreatic and colon cancer

SMAD

TSC1,2 Regulates RAB5-GTP activity

Tuberous sclerosis Rhabdomyoma angiomyolipoma, other tumors

PI3K

VHL E3 ligase recognition factor for HIF

VonHippel-Lindau syndrome

Renal cell carcinoma, pancreatic cancer, pheochromocytoma

HIF1

WT1 Transcription factor Wilms tumor Nephroblastom p53 * The gene may be implicated in additional pathways besides the one indicated in the table.

15

Cellular mechanisms in cancer Cancer cells have defects in regulation of normal cell proliferation and homeostasis. According to a classical review article written by Hanahan and Weinberg are there six essential cell physiology alterations in a cancer cell 8:

(i) Self-sufficiency in growth signaling. Normal cells require mitogenic growth signals to move from a quiescent state to proliferation and will not multiply in the absence of such stimulatory signals. Many of the oncogenes act by mimicking such signaling. Cancer cells frequently acquire the ability to synthesize growth factors that they themselves respond to, creating a positive feedback signaling loop 9. This can be illustrated by the production of PDGFs (platelet-derived growth factors) by gliomas and TGF (tumor growth factor ) by sarcomas 9. Within normal tissue, growth signaling is conducted by neighboring cells (paracrine) or by systemic (endocrine) signals. Consequently, stromal cells within the tumor are probably partly driving the proliferation due to growth signaling 8. The overexpression of growth factor receptors may admit the cancer cell to be hyper responsive to the surrounding levels of growth factor that normally would not trigger proliferation 9.

(ii) Insensitivity to growth-inhibitory signals. Multiple antiproliferative signals maintain cellular quiescence and homeostasis in normal tissue. Cells monitor their external environment and on the basis of sensed signals decide whether to proliferate, to be quiescent, or to enter into a post mitotic state. Antigrowth signaling is associated with the cell cycle, specifically the transit of the cell through the G1 phase. Proliferation can be blocked by forcing cells out of their active proliferative cycle into the quiescent state. Alternatively, cells may be induced to permanently abandon their proliferative potential by being induced into post-mitotic states.

(iii) Acquired resistance toward programmed cell death, apoptosis, is a hallmark of most and perhaps all types of cancer. The apoptotic machinery is activated in response to detection of abnormalities, including DNA damage, signaling imbalance by oncogene action, survival factor insufficiency or hypoxia 10. Resistance to apoptosis is commonly acquired through the p53 tumor suppressor gene. The inactivation of the p53 protein is seen in more than 50% of human cancers and results in the removal of a key component of the DNA damage sensor that can induce apoptosis 11.

(iv) Limitless replicative potential. Senescence is a protective mechanism that can be activated by shortened telomeres or conflicting growth signals that makes cells incapable of proliferation. Telomere maintenance is a key component of the capability for unlimited replication. Avoiding senescence may represent an essential step in tumor progression.

(v) Sustained angiogenesis. Oxygen and nutrients supplied by the vasculature are crucial for cell function and survival. In order to progress,

16

developing neoplasias must acquire angiogenic ability 12, 13. Vascular endothelial growth factor (VEGF) and fibroblast growth factors (FGF1 and 2) are known to initiate angiogenesis 14.

(vi) Tissue invasion and metastasis. Most types of human cancer invade adjacent tissues and travel to distant sites where they may succeed in founding new colonies, metastases. Invasion and metastasis depend upon the same capabilities as the formation of the primary tumor. E-cadherin function, which is a cell-to-cell interaction molecule ubiquitously expressed on epithelial cells, is lost in a majority of forms epithelial cancers.

Fig. 2. Each mutation in a cell population gives the mutated cell a growth advantage and stimulates clonal expansion.

As briefly described above, solid tumors cells must acquire several genetic alterations of cancer genes to achieve a malignant status. These mutations accumulate over time, and every mutation stimulates clonal expansion which results in a larger number of cells that may be subject for subsequent mutations 15 (Fig. 2). Not only alterations at the nucleotide levels are important for tumor formation. Instability at the chromosomal level is also commonly found in cancer. On average 25-30% of the alleles that are present in a normal cell are actually lost in cancer cells, and losses up to 75% of alleles have also been observed 16.

Genetic variation in the human genome Historically, the first genetic differences to be observed in the human genome were changes in quantity and structure of chromosome that were visible under a microscope. These chromosomal abnormalities included aneuploidies 17-19, rearrangements 20-23 and fragile sites 24. Due to development of the sequencing technology, DNA changes on the nucleotide

17

level were later possible to detect. These mainly include single nucleotide polymorphism (SNP), variation of small repetitive elements (mini- and microsatellites); small deletions, insertions, inversions, and duplications.

Gaining knowledge of variation ranging from large chromosomal rearrangements to single nucleotide changes is crucial for understanding the phenotypic differences between individuals in health and disease. A brief description of genetic variation will be given below.

Chromosomal imbalances There are several types of chromosomal imbalances, including an abnormal number of chromosomes (polyploidy or aneuploidy), deletions, duplications, inversions and translocations (Fig.3).

Fig. 3. An overview of chromosomal imbalances in a hypothetical cell containing only two chromosomes. Polyploidy means more than two full sets of homologous chromosomes in a nucleus. Aneuploidy represents presence of an abnormal number of chromosome(s). Interstitial deletion refers to loss of chromosome material on one copy (hemizygous) or both copies (homozygous/biallelic). Reciprocal translocation is an exchange of segments between two chromosomes without any gain or loss of genetic material. Non-reciprocal translocation is a translocation of segments where chromosomal material is lost. Amplification/double minutes are extra-chromosomally amplified chromatin that usually contains a particular chromosome segment. Amplification refers to a chromosome segment that is duplicated. Insertions introduce an extra chromosomal segment. Adapted from Albertson and Pinkel 25.

Chromosome rearrangements may be located throughout the genome, but occur predominately in the pericentromeric and subtelomeric regions 26-28.The breakpoints of rearrangements have been associated with highly homologous genomic segments, in particular low-copy-repeats (LCR) (also called segmental duplications) and AT-rich palindromes 29. LCRs result from

18

chromosomal duplications and may encompass genes, fragments of genes, pseudogenes and gene clusters. Bioinformatic analysis of available human DNA sequence revealed that ~10% of genomic segments are present in at least two copies when applying the threshold of >98% sequence identity over 1 kb 30. The most common mechanism behind the rearrangements that are mediated by LCRs is non-allelic homologous recombination (NAHR) 31

(Fig.4). NAHR between LCRs located on the same chromosome located in direct orientation results in deletions or duplications, whereas NAHR between LCRs located in inverted orientation results in an inversion 31.NAHR between LCRs located on different chromosomes may result in reciprocal translocations.

Not all cases of genomic rearrangements are explained by NAHR. Another mechanism, named non-homologous end joining (NHEJ), have been described for rearrangements that displays scattered breakpoints 29 (Fig.4). These breakpoints are located in apparently unique sequence, or occasionally in Alu elements 32, 33. In NHEJ two non-homologous sequences recombine via a deletion of the intervening segment and additional bases are added in the deletion junction.

NAHR NHEJSubstrates

Misalignment

Crossover

NN...NN

Junction fragmentDeletion with added bases at junction

Recombinationproducts

Cutsite Cutsite Cutsite

Cutsite Cutsite

Fig. 4. Deletion rearrangements mediated by NAHR and NHEJ. NAHR uses LCRs (blue boxes) as substrates for recombination. The LCRs oriented in the same direction, misaligns and the subsequent homologous recombination results in a deletion with a single recombinant LCR. NHEJ uses two non-homologous sequences as substrates for recombination. The sequences are recombined via NHEJ with a deletion of the intervening fragment. Additional bases are added in the deletion junction. Adapted from Shaw et al. 31

19

The association between the location of LCRs and regions of chromosomal instability has been reported in several studies and it has also been implicated as a probable mediator in >25 so called genomic disorders 34.Examples of such disorders are Neurofibromatosis type 1 and DiGeorge syndrome 35, 36.

Copy number variation Copy number variation (CNV) refers to a segment of DNA that is present in variable copies in comparison to a reference genome and may include insertions, deletions and duplications. This type of variation may comprise entire genes, part of genes or their regulatory regions. CNVs may influence the gene dosage, which alone, or in combination with other genetic or environmental factors might have an impact on the phenotype 37.

When the DNA sequence of the human genome was published in 2001 one of the conclusions was that the similarity between the genomic sequence between any two individuals was thought to be 99.9% 38, 39. This was probably not a correct estimation, as it is now reported that SNPs alone constitutes ~0.3% of the human genome 40. Furthermore, copy number variation as a source of genetic diversity is only in its early days of exploration. In 2004 there were two groundbreaking papers published that described the global presence and distribution of copy number variation 41, 42.This type of variation was also reported by our group in the context of analysis of chromosome 22 in patients affected with schwannoma (paper III) and glioma 43. In a recent review article by Feuk et al, available information from literature was complied and it was stated that 563 unique CNVs has been reported 44. Out of these CNVs were 47% (264) copy number losses, 41% (242) copy number gains, and 10% (57) were described both as losses and gains. A high percentage (20-50%) of these CNVs was positioned in the close proximity to low copy repeats 44.

A recent study analyzed SNP data extracted from the HapMap project, using two sets of thirty parent-offspring trios with European and African origin 45.They reported 586 distinct regions of deletion polymorphisms. Interestingly, both size and frequency of the reported deletions differed between the two ethnic groups. The European group displayed 345 deletions at 228 locations with an average size of 10.6 kb (ranging from 0.3 to 404 kb), whereas the African set presented with as much as 590 deletions at 396 locations, with an average size of 8.5 kb (ranging from 0.5 to 1200 kb). Curiously, only 37 deletions were shared between the two populations 45. This likely reflects a considerable diversity commonly found in African populations 46, 47.Furthermore, they estimated that a typical individual have on average 30-50 hemizygous deletions lager than 5 kb, covering in total approximately 550-

20

750 kb 45. The deletions detected in this study spanned 267 known and predicted genes 45. One way of interpreting these data is that deletion polymorphisms might have an important role in the genetics of complex traits.

The discrimination between CNVs coupled to a normal phenotype and these which might be disease causing is still a challenging issue that remains to be addressed. A further complicating fact is that a single CNV might only predispose for a disease in combination with other CNVs and/or other genetic alterations 44. To date, there are 1267 copy number variants reported at the Database of Genomic Variants, which is a catalogue of large-scale variation in the human genome (http://projects.tcag.ca/variation).

Single nucleotide polymorphism (SNP) A single nucleotide polymorphism (SNP) refers to a single base change in DNA. A common definition is that an inherited allelic variation must have >1% population frequency in order to be classified as a SNP. It was estimated that there are > 10 million SNPs in the human genome, on average approximately one SNP every 200-300 nucleotides 40. The impact of single nucleotide differences is variable and elusive, but is clearly dependent upon the location of the polymorphism. The majority of SNPs are likely to be positioned outside of protein encoding regions, but they do not need to be located in the translated sequence to result in a functional change. SNPs positioned in promoters or repressors may influence binding of these sequence elements, resulting in differential regulation of transcription. It has also been demonstrated SNPs may alter mRNA secondary structure 48.

The difference in allele frequency of a SNP is used in the case-control analysis, between e.g. patients and control groups, which is the classical genetic association approach. The disadvantage of this method is that a single SNP has relatively low information content and in a complex disease multiple genes might be involved in its pathogenesis. A large quantity of samples may be required to detect significant associations. SNPs are also used as markers to create high-density maps over the human genome that may be used to map genetic variation between individuals. A haplotype is created by combining multiple adjacent SNPs, which can be used for genetic association studies. Through identification of shared haplotypes, blocks of non-random variable diversity were discovered, indicating that the occurrence of linkage disequilibrium (LD) 49. LD refers to a population phenomenon where two alleles are inherited at greater frequency than would be predicted by random recombination. Although the majority SNPs likely has no direct functional effect, the LD enables correlation of these non-functionally involved SNPs to a disease associated locus. This approach is

21

commonly used in the search for genes causing complex disease 50, 51. The HapMap project is providing us with a map of SNP markers with global coverage. By genotyping approximately one million SNPs in a set of 269 samples with different ethnicity, haplotypes will be defined with the goal of developing a map of non-redundant tag SNPs 52.

EpigeneticsA complex and not yet well studied genetic regulatory code is carried in the chromatin structure and in the three dimensional nuclear organizations of chromosomes. This cellular information is referred to as epigenetics, which is commonly defined as hereditable or acquired changes in gene expression that are not associated with changes in DNA sequence. It regulates the accessibility of the primary genetic code and in this way determines the spatial and temporal usage of sequence information. The epigenetic profile of an individual is under environmental influence and changes over time. As a consequence, the expression of genes will also change during the course of life 53. The best known epigenetic modification is DNA methylation. In the mammalian genome methylation takes place at cytosines located 5’ of a guanine in a CpG dinucleotide. This dinucleotide is generally underrepresented in the genome, due to the higher C to T mutation frequency of methylated cytosines. Although there exist stretches with high CG content which are called CpG islands. According to strict criterion, a CpG island is defined as a stretch of DNA sequence longer than 500 bp, with G+C content equal to or greater than 55% and observed CpG/expected CpG of 0.65 54.CpG islands located in the gene promoter and/or the first exon-intron regions are often involved in regulation of gene expression. Generally, unmethylated CpG islands are associated with gene expression and hyper methylated with gene silencing. Methylation imbalances, which result from hypo- or hyper-methylation of CpG islands, may lead to development of cancer. It is likely the sum of disturbances on both genetic and epigenetic levels that is involved in the multi-step transformation from a normal cell into a cancer cell.

Methylation in a normal cell has several important functions, such as: defending the genome from the deleterious effects of parasitic sequence (e.g. retrotransposons) 55; maintenance of conformation and integrity of the chromosome 56; X-chromosome inactivation during development 57 and imprinting of genes to ensure expression in a parent-specific manner 58. The mechanisms by which methylation prevents transcription might be by inhibiting binding of certain transcription factors to their recognition sites that contains methylated CpG dinucleotides 59. Another possible mechanism involve proteins or protein complexes, such as Methyl-CpG-binding proteins

22

MeCP2 or MeCP1, that bind specifically to methylated CpGs and thereby inhibit the binding of transcription factors by blocking the access to these regulatory elements 60, 61. The three known DNA methyltransferases DNMT1, DNMT3A, and DNMT3B are crucial for the methylation machinery in mammalian cells 62, 63. The homozygous elimination of any of these genes in mice has been shown to be lethal, either before birth for Dnmt1 orDnmt3B or four weeks after birth for Dnmt3A 63, 64. Heterozygous 56mutants for any of the DNA methyltransferases appear normal and are fertile 63, 64.All three methyltransferases performs de novo methylation, although with distinct sequences as targets 63, 65. DNMT1 is ubiquitously expressed in somatic tissue and is termed a maintenance methylase due to a greater activity on hemi-methylated DNA after replication62.

There are examples of several genes involved in diverse pathways that are silenced by methylation in cancer, such as: genes involved in cell cycle regulation (RB, TP73), DNA repair (BRCA1, hMLH1, MGMT), drug resistance and detoxification (GSTP, MDR1), differentiation (MYOD, EWT),apoptosis (CASP8, TMS1), angiogenesis (THBS1) as well as metastasis and invasion (CDH1, TIMP-3). Almost half of the genes that are known to harbor germ line mutations causing familial forms of cancer have also been found to be silenced by hypermethylation in its sporadic form. For example germ line mutations in BRCA1 and STK11 genes are known to cause familial forms of breast and colon cancer, respectively. In the sporadic forms, these genes are often epigenetically silenced. It has been suggested that an epigenetic alteration occurs already in stem/progenitor cells and that this would be an early predisposing event 66.

There is global demethylation occurring in genome of cancer cells which is believed to generally affect repetitive intergenic or intronic DNA. This process coexists with both hyper- and hypomethylation of selected genes. The process of global DNA hypomethylation contributes to carcinogenesis by causing chromosomal instability that may lead to loss of heterozygosity 67. In addition, extensive demethylation in centromeric sequences may play a role in aneuploidy. Hypomethylation may also lead to the transcription of previously silent transposons that may move and insert themselves within cellular genes and thereby disrupt normal gene expression 68-70. The expression of imprinted genes is also affected by hypomethylation. This is illustrated by the up-regulation of anti-apoptotic growth factor (IGF-2) as well as the loss of a transformation-suppressing RNA (H19) that causes certain childhood tumors 71.

23

Chromosome 22 This chromosome was the first to be sequenced with in the human genome project and its sequence was published in 199972, two years ahead of the complete sequence of the human genome 38, 39. This small acrocentic chromosome accounts for ~ 1.5% (49.5 Mb) of the genome and contains 701 annotated genes (NCBI, Homo Sapiens build 35.1, http://ncbi.nlm.nih.gov) which makes it very gene rich compared to many other chromosomes 73.Chromosome 22 is one of the first chromosomes to be replicated, with most of the q arm replicating in the first half of the S phase. The early replication can be correlated to its high gene density 74. It is also frequently affected by chromosomal aberrations in a variety of human disorders. These diseases are both cancer related (such as NF2, schwannoma, schwannomatosis, meningioma, ependymoma, pheochromocytoma, chronic myeloid leukemia, Wilms tumor, acral melanoma, glioma, dermatofibrosarcoma protuberans, breast-, ovarian-, colorectal-, and oral- cancer) and non- cancer related (such as schizophrenia, cat-eye syndrome, Emanuel and DiGeorge syndromes).

Neurofibromatosis type 2

Clinical aspects of Neurofibromatosis type 2 (NF2) Neurofibromatosis type 2 (NF2) is an autosomal dominant disorder with a full penetrance at the age of 60. The estimated birth incidence is 1 in 25,000 75. The characteristics of NF2 are bilateral schwannomas of the vestibular branch of the eight cranial nerve. Schwannomas can also arise at other cranial, spinal and cutaneous nerves. Other nervous system tumors in NF2 patients frequently occurring are cranial or spinal meningiomas and spinal ependymomas. Ophthalmologic abnormalities such as juvenile posterior lens opacities (cataract) are also common characteristics of the disease. The most common presenting symptoms are hearing loss, tinnitus, vertigo or imbalance which is usually caused by vestibular schwannomas 76, 77. Seizures or headaches are presenting in 6-8% of patients and are usual symptoms of meningiomas 76, 77. The so far only known genetic predisposing factor in NF2 patients is the inheritance of a germline mutation of the NF2 tumor suppressor gene, although tumor formation requires a second somatic alteration of the remaining allele according to the two hit hypothesis 3. The second hit is commonly a loss of the entire long arm of chromosome 22 or deletions of varying size encompassing the NF2 locus. Approximately half of NF2 patients have no family history of NF2 and thus carry de novomutations 78. The clinical phenotype of NF2 patients can vary between mild

24

(Gardner) to severe (Wishart) form. The mild Gardner type has a late onset (after 25 years) with slowly growing bilateral vestibular schwannomas and rarely additional tumors. The severe Wishart type is manifested by the presence of fast growing bilateral vestibular schwannomas before 25 years of age, and at least two, but often multiple, additional tumors. There are also intermediately affected cases that are not easily placed into either category. Anticipation has also been reported in a few NF2 families 78.The clinical heterogeneity might reflect different genetic mechanisms underlying development of the disease.

Diagnosis criteria for NF2, recommended by the National neurofibromatosis foundation, 2000: Definite NF2

Bilateral vestibular schwannoma (VS) A first degree relative with NF2 plus either unilateral VS or two of the following: meningioma, schwannoma, ependymoma, juvenile posterior subcapsular or cortical cataract

Probable NF2 Unilateral VS before 30 years and one of the following: meningioma, schwannoma, ependymoma, juvenile posterior subcapsular or cortical cataractMultiple meningiomas plus either unilateral VS before 30 years or one of the following: schwannoma or ependymoma.

The NF2 gene and protein The NF2 tumor suppressor gene was mapped to chromosome 22 by genetic linkage analysis in 1987 79, further localized to band 22q12.2 and isolated by positional cloning by two groups in 1993 80, 81. The gene encompasses 17 exons and spans >90 kb of genomic sequence 80. The frequency of mutations in the NF2 gene detected in NF2 patients was 34-70% 82-86. The identified mutations are mainly nonsense-, frame shift-, or splice site- mutations as well as deletions, which are predicted to truncate the protein. Missense mutations are rarely detected. The NF2 gene protein product shows strong sequence similarity with ERM proteins, and was consequently named merlinfrom Moesin, Ezrin, Radixin-Like protein 81. It is also called schwannomin due to its role of preventing schwannoma formation 79. There are two common isoforms of merlin, consisting of 595 (isoform I) and 590 (isoform II) amino acids, respectively. Isoform I comprise exons 1-17, with exon 16 spliced out. Isoform II encompasses exons 1-16, with exon 16 alternatively spliced, replacing 16 terminal residues with 11 new residues. There are at least six additional, less commonly transcribed isoforms of merlin, as a result of alternative splicing, several transcription initiation sites and different polyadenylation signals that terminate the transcript 87.

25

Based on amino acid sequence alignment, merlin exhibits homology to the ERM (Ezrin, Radixin, Moezin) family of membrane-cytoskeleton linking proteins. The ERM proteins belong to the 4.1 superfamily, which in addition to protein 4.1 includes talin and several intra cellular protein-tyrosine phosphatases 88. There are three structural domains in merlin: a conserved N-terminal FERM (band 4.1, ERM) domain, an -helical region and a unique C-terminal domain. The N-terminal of ERM proteins binds directly or indirectly to membrane proteins and the C-terminal binds to actin. This is how ERM proteins may serve as linkers between membrane proteins and the actin cytoskeleton 89. ERM proteins are negatively regulated by intra-molecular interactions between its N- and C-terminal domains. The closed state masks the ligand-binding sites and the protein becomes inactive. The activation of the ERM proteins occurs by Rho-mediated phosphorylation that disrupts the self-association and unmasks the ligand-binding sites 90. Merlin lacks the high-affinity actin binding site that is present in the C-terminus of ERM proteins, although it can interact directly with F-actin through the N-terminus 91-93 or indirectly through II-spectrin 94. Merlin isoform I also adopt the ‘head-to-tail’ closed state in which the N-terminal is associated to the C-terminal, in contrast to merlin isoform II that only exists in its open form. The ‘head-to-tail’ conformation of merlin is less tightly regulated compared to the ERM proteins. Molecular studies suggest that merlin isoform I, in contrast to ERM proteins, become inactive in suppressing growth when the self-associated state is disrupted by phosphorylation 95.Immunocytochemistry studies have shown that merlin colocalized with F-actin in or near the cell membrane, preferentially in filopodia, ruffling membrane and leading edge 96, 97.

Interestingly, merlin expression is more pronounced in neurons and other glia compared to Schwann-, meningothelial- and ependymal- cells 98, even though NF2 associated tumors originate from the latter cell types. Furthermore, NF2 patients do not display any tumors derived from neurons, nor show any other neurological dysfunctions. Thus, merlin may play other roles unrelated to tumor suppressor function in neurons. The expression pattern of merlin compared to the other ERM family proteins indicates that merlin and the ERM proteins might have diverse function in the central nervous system.

The genotype and phenotype in NF2 As mentioned above, NF2 is a clinically heterogeneous disease ranging from very severe, with a rapid course and multiple nervous system tumors at a young age, to relatively mild with a later onset and a more benign course. A number of studies have reported mutations in NF2 gene in patients, however, there is a discrepancy in the mutation frequency between different reports

26

(36-70%), 82, 84-86. Thus, in approximately one third of patients, the genetic mechanism behind the development of NF2 is still unknown.

It has been debated if there is a correlation between the type of NF2mutation and the phenotype. In general, there is genotype-phenotype correlation with patients affected by truncating mutations. These patients mainly develop a severe phenotype 99. This is in accordance with Ruttledge et al. which reported that patients with protein-truncating mutations presents with severe NF2, while missense mutations only is found in the mild phenotype 85. An additional study connected the type of NF2 gene mutation to the phenotype and concluded that constitutional missense mutations, splice-site mutations, large deletions, or somatic mosaicism had significantly fewer tumors than patients with constitutional nonsense or frameshift mutations 100. It has also been suggested that the location of splice site mutations is correlated to severity of NF2, using age at onset of symptoms and number of intracranial meningiomas as indicators. Individuals with splice site mutations in exons 1 to 5 had more severe disease than those with splice site mutations in exons 11 to 15 83. However, there are also many contradicting cases; patients with constitutional splice site mutations have been reported with variable phenotypes, ranging from severe to asymptomatic 101. Furthermore, truncating NF2 mutations that are usually associated with the severe phenotype have also been identified in patients with a mild phenotype 82. The analysis of constitutionally transmitted mutations showed clinical variability within the same family 101-103.

Regarding cases presenting deletions of NF2, the size of chromosome 22 deletions might also be correlated to phenotype. A complete NF2 gene deletion has been reported in NF2 families, and in a number of cases it has been associated with the mild phenotype 104, 105. The presence of mild phenotype has also been correlated to somatic mosaicism, which is caused by post-zygotic mutations in the early stage of embryo development 99, 106.Thus, only a subpopulation of the normal cells would be carrying the constitutional mutation. Large deletions that extended beyond the NF2 locus were reported to be associated with the moderate or severe phenotype 107.These large deletions might, in addition to inactivate one copy of the NF2gene, also affect other loci that have an influence on the NF2 phenotype. Furthermore, a linkage analysis performed in families affected by a clinically very closely related disease called schwannomatosis (see below), excluded alterations in the NF2 gene as a germline event in these patients, and proposed a schwannomatosis candidate locus centromeric of the NF2 gene 108. This not yet characterized schwannomatosis predisposition gene seems to be a candidate for a NF2 phenotype modifier gene, possibly contributing to the severe Wishart form of NF2.

27

Tumors associated with NF2NF2 patients may display a number of other tumors in addition to bilateral vestibular schwannomas, i.e., schwannoma at other cranial, spinal and cutaneous nerves, meningioma, and ependymoma.

Schwannoma Schwannomas are benign tumors arising from Schwann cells that form the myelin sheaths surrounding nerve axons. They may develop at various locations in the body, sporadically or in association with NF2 and schwannomatosis. A large group of schwannomas affect the VIIIth cranial nerve, but can also occur on other cranial, spinal or peripheral nerves. Ninety three percent of VIIIth cranial nerve schwannomas develop as solitary sporadic tumors which occur unilaterally, the remaining 7% occur in the context of NF2 75. Inactivation of both NF2 alleles is the fundamental genetic aberration detected in these tumors. The inactivation occurs either somatically in sporadic or schwannomatosis-associated schwannomas or by a combination of germline and somatic NF2 gene mutations in NF2-associated schwannomas. Previously reported frequencies of 22q-associated deletions in schwannomas display large discrepancies, ranging from 30-80% 109-113 and the NF2 gene is so far the only characterized tumor suppressor gene involved in the pathogenesis of these tumors.

Meningioma Meningioma is the most frequently occurring intracranial tumor and has an annual incidence of 7.8 per 100 000 individuals 114. Clinically treated meningiomas constitute approximately 13-26% of all primary brain tumors 115. Epidemiological studies indicate that ~90% of all meningiomas are asymptomatic 116. Meningioma is thought to be derived from the arachnoid cap cells of the leptomeninges, which covers the brain and the spinal cord 117. About 90% of meningiomas occur in the cranial and 10% in the spinal meningies. They typically arise after the fifth decade of life, twice as often in the female as in the male population. The majority of meningiomas are benign slowly growing solitary, sporadic tumors but they may progresses into anaplastic tumors of World Health Organization (WHO) grade II or III. More than 25% of meningiomas cannot be cured because of their inaccessible anatomical location or their invasiveness. Approximately half of NF2 patients develop meningiomas, making it the second most common tumor type associated with the disease. Furthermore, multiple and familial meningiomas are rare, the majority of cases are associated with the NF2 syndrome 118-120. Biallelic inactivation of the NF2 gene is found in meningiomas of all grades and is thought to be an early event in

28

tumorgenesis 121. Previous studies have reported a mutation frequency of the NF2 gene in sporadic meningiomas between 20 and 60% 122-126. However, loss of chromosome 22 outside the NF2 locus and in the absence of NF2mutations has also been reported 122 (paper II), suggesting additional genes located on 22q to be involved in meningioma formation. There are also reports of aberrations on other chromosomes in the context of meningioma formation, the second most common is LOH of the short arm of chromosome 1 127-130.

Ependymoma Ependymoma is a subtype of glial cell-derived tumors that originates from the cerebral ventricle or from the spinal canal. Ependymoma is classified as WHO grade II, and is the third most common brain tumor diagnosed in children 131. The vast majority of pediatric ependymomas are intracranial, while the adult counterpart more commonly arises intraspinally 131. Spinal ependymomas can be seen in approximately 6% of patients with severe type of NF2 76. Compared to other glioma subtypes, relatively little is known regarding the molecular pathogenesis of ependymomas. The most frequently reported abnormality is deletion of chromosome 22 132-135. LOH involving the long arm of chromosome 22, together with NF2 mutations have been reported in several ependymomas 136.

SchwannomatosisSchwannomatosis is a recently defined disease entity that is closely related to NF2. Schwannomatosis patients display multiple schwannomas at various locations, but never presents with bilateral vestibular schwannomas, which is the hallmark of NF2. Furthermore, schwannomatosis patient do not develop other types of nervous system tumors that are commonly found in NF2 patients 137. Analyses of the NF2 gene in tumors from sporadic schwannomatosis cases reveal that these somatic cells carry inactivating mutations in a similar spectrum as NF2-derived schwannomas. Analysis of multiple tumors from the same patients revealed somatic mosaicism for different NF2 gene mutations 111. Furthermore, the incidence of sporadic schwannomatosis seems to be as frequent as for NF2 120, 138. However, familial clustering of schwannomatosis cases is rare. When familial schwannomatosis were examined, no cases of NF2 gene mutations were found in constitutional DNA of these patients 139, 140. In addition, linkage analysis performed in eight schwannomatosis families excluded germline inactivation of NF2 gene, and proposed a schwannomatosis candidate locus centromeric of the NF2 gene 108.

29

Diagnosis criteria for schwannomatosis 141

Definite schwannomatosis At least two non-intradermal schwannomas and Absence of 8th cranial nerve schwannomas on MRI and Age older than 30 years OR one schwannoma plus a first degree relative who meets the above criteria

Possible schwannomatosis All the above criteria except age, younger than 30 years

Segmental schwannomatosis The above criteria for definite or possible schwannomatosis and Restriction to one limb OR less than five contiguous segments of the spine andAbsence of a first degree relative who meets the criteria

30

Methods

Evolutionary comparison of DNA sequenceIt was previously thought that most human DNA did not have any function (‘junk DNA’) and that the functional DNA was only contained within protein coding regions. However, this assumption has been revised and a challenging goal is to identify the functional elements that reside in the human genome. One method of identifying such sequence elements is by the alignment and comparison of DNA sequence between different species in order to identify sequence stretches that remained conserved through evolution. Both previously not identified protein coding sequences as well as evolutionarily conserved non-coding sequences could be identified by this approach. This is based on the assumption that functional sequences evolve slower then non-functional sequence.

Currently there are two major web based DNA sequence alignment tools available for such comparisons; PipMaker/MultiPipMaker 142, 143 and Vista 144. During the course of this work I have mainly utilized the PIPMaker tools. PipMaker and MultiPipMaker are used for alignment and visualization to compare long genomic sequences and identify conserved segments, using BLASTZ as the underlying alignment program. The level of sequence conservation between the reference sequence and one (or several) additional sequences is visualized in a percentage identity plot (PIP). Stretches of sequence identity in a pair wise alignment is plotted corresponding to their length and the level of percentage identity according to position in the reference sequence. Furthermore, PipMaker also utilizes RepeatMasker program for the identification of interspersed repeats and low complexity DNA sequences (http://www.repeatmasker.org). PIPMaker is useful not only to identify sequence conservation between species, but also to visualize the distribution of different classes of repetitive sequence and CpG islands in the reference sequence.

Sequence alignment between the whole human and mouse genomes has been performed. From these comparisons were 327 000 conserved non-genic sequences (CNGs) were identified. The applied cut-off criteria were: >100 bp of gap-free alignment with at least 70% sequence identity 145. Out of the identified CNGs were the majority, 65%, located intergenic and 35% were

31

positioned within genes. Interestingly, the distribution of CNGs in the genome seemed to be negatively correlated to the distribution of genes (Fig. 5). It might be expected that CNGs require proximity to genes in order to regulate gene function. Although the higher density of CNGs in gene-poor regions may indicate that also ‘gene-deserts’ are under strong selective constraint.

Fig. 5. Correlation of CNGs and exons. There is a negative correlation between the density of CNGs and exons in 1 Mb windows of the human genome. The regression line is indicative and does not imply a linear relationship between the two variables. Adapted from Dermitzakis et al 145.

When comparing the human and mouse DNA sequence, the above mentioned threshold (>70% identity, >100 bp of gap-free alignment) is commonly used 145. By applying this criterion in human-mouse sequence comparisons, Loots and colleagues discovered the coordinate regulator of Interleukins 4, 13 and 5 146. Although a pair-wise sequence comparison between human and mouse is powerful, the sequence alignment of additional species increases the ability to distinguish non-coding functional elements 147. The power of comparisons may increase considerable if an evolutionarily distant species e.g. chicken and pufferfish is included. This may be exemplified by the identification of a CNG that were highly conserved between human and pufferfish and was proved to regulated expression of the DACH gene 148. CNGs possessing functional importance may also be targets for pathogenic mutations. A CNG displaying conservation between human and pufferfish was mapped to intron 5 of the LMBR1 gene. This CNG, called ZRS, was proven to regulate the expression of the sonic hedgehog gene. A single nucleotide mutation in ZRS was reported to be causative of preaxial polydactyly (presence of extra fingers or toes) 149.

32

DNA sequencing and mutation analysis Dideoxy nucleotide sequencing, or Sanger sequencing after its inventor Frederick Sanger, is the most common method used for DNA sequencing 150.In this method the DNA may be cloned into a sequencing vector or amplified by PCR, to be further used as a template in a sequencing reaction. Synthesis is initiated from an oligonucleotide that is annealed to the sequencing template. DNA polymerase is used for building a new single stranded DNA copy by incorporating nucleotides. While DNA chains are normally made up of deoxynucleotides (dNTPs), the Sanger method uses a mixture of deoxynucleotides and dideoxynucleotides. Dideoxynucleotides (ddNTPs) are missing a hydroxy (OH) group at the 3' position (Fig. 6), at which new nucleotide attaches to form a chain. A missing OH group in the 3' position interrupts chain elongation. dNTPs and differentially fluorescently labeled ddNTPs are randomly inserted during the elongation in the sequencing reaction. The result of the amplification is a large number of DNA fragments of different size, terminated by a fluorescently labeled ddNTP.

Fig. 6. Dideoxynucleotides (ddNTPs) are missing a hydroxy (OH) group at the 3' position.

The amplified fragments are separated according to their length in a capillary gel eletrophoresis system. The identity of the terminating ddNTPs at the end of each fragment is detected by a laser together with a CCD (charge coupled device) –camera. The emitted light at different wave lengths is presented as peaks and interpreted as a chromatogram by computer software. DNA sequencing is a commonly used high throughput method for mutation detection at the nucleotide level (Fig. 7). This method is highly suitable for detection of base substitutions, small insertions and deletions. Although, large deletions sometimes may not be detected. This method is also dependent on careful primer design. If a polymorphic site is present within the annealing site of a primer, preferential amplification of one allele may occur. Tumors are often composed of a heterogeneous cell population.

33

Consequently, mutations present in only a small fraction of cells may pass undetected. If the ‘contaminating’ cells are of stromal origin, the problem may be circumvented by micro dissection of tumor cells.

Microarray-based comparative genomic hybridization (array-CGH) A recent development of an advanced technique to detect genomic DNA copy number imbalances was first described by in 1997 by Solinas-Toldo and collegues and referred to as matrix-CGH 151 or array-CGH 152. This method is based on the same principles as metaphase-CGH, in which DNA are labeled with two different fluorochromes and competitively hybridized to metaphase chromosomes on a glass slide. The ratio of fluorescence intensities detected is indicative of the relative DNA copy number in test versus reference DNA. Array-CGH is performed in a similar way, except that the targets are mapped genomic clones immobilized in spots on a glass slide instead of metaphase chromosomes (Fig. 8). The genomic clones used for this purpose are generally BAC clones, with an average insert of human DNA sequence of ~100-200 kb.

Fig. 7. A sequence chromatogram displaying a heterozygous C/T position indicated by an arrow.

34

Fig. 8. Array based comparative genomic hybridization (array CGH). Test and reference DNA is differentially labeled, mixed and coprecipitated with unlabelled Cot-1 DNA for blocking of repetitive sequences. The probe is deposited on the slide with array targets and hybridized. The slides are washed and fluorescence signals are detected. Using image-processing software, chromosomal regions with gains or loss of DNA sequence can be detected by an abnormal ratio between test and reference.

This novel array-based approach introduced a new dimension in medical genetics in terms of higher resolution, specificity and its potential to study a multitude of loci in a single experiment. One of the major advantages of the technique is that it allows a high resolution comparison of test (e.g. tumor-derived or constitutional DNA) against a reference DNA from an unrelated individual. This permits discrimination between DNA copy number variations present in the constitutional DNA of the patient and aberrations present only in tumor cells. An important fact to take into consideration when interpreting results is that the method relies on the comparison to a reference genome. There is no commonly used standardized reference genome available, although samples are sometimes pooled and utilized as an average reference genome. This fact makes the comparisons between studies from different laboratories and the standardization of databases for copy number variations complicated.

Chromosome rearrangements have been associated with the genomic architecture, in particular low-copy-repeats and AT-rich palindromes 31. The current approach for the construction of genomic microarrays is mainly using genomic clones as targets. This approach has some limitations in analysis and applicability to regions that are rich in common and/or low-

35

copy-repeats. One way to circumvent these problems is to generate PCR-product based (alternatively oligonucleotide based) repeat-free microarray targets. Briefly, this includes the amplification of repeat-free and non-redundant sequences from genomic clones and to spot them either in pools or as single PCR-products on a glass slide (paper IV). This approach gives the advantage of analysis of any region regardless the content of redundant sequence as well as an increased resolution of the analysis. In addition to the detection of CNVs array-CGH can be applied in: i) discovery and precise mapping of a multitude of unknown transcripts; ii) studies of transcription factor binding sites (via so called ChiP-to-chip assays; Chromatin Immuno-Precipitation) hybridized to DNA arrays; iii) quantitative and qualitative assays of methylation patterns across the genome; iv) analysis of replication timing; and v) evolutionary comparisons.

DNA methylation detection by Pyrosequencing Previous protocols for analysis of methylation status of multiple CpGs relied on bisulfite treatment, followed by PCR, cloning of DNA into plasmid vectors and sequencing 153. It is a labor intensive procedure that requires sequencing of numerous clones for each sample in order to reach statistical significance. Pyrosequencing (Biotage, Uppsala, Sweden) is a novel method for methylation detection that has overcome some of the main disadvantages in previously reported protocols for analysis of methylation status 154.

Fig.9. Methylation detection of bisulfite treated DNA. Unmethylated C (red) and methylated C (green) are differentiated by bisulfite treatment and PCR. C not followed by G acts as control for the bisulfite step (blue column). Theratio C/mC at each CpG site (peaks in orange column) is measured in sequence context. Figure adapted from Biotage

36

This method relies on bisulfite treatment of DNA to convert unmethylated cytosines into uracils, whereas methylated cytosines remain unchanged (Fig. 9). Uracils are amplified as thymidines in the subsequent PCR reaction of the converted DNA. Cytosines that are not followed by a guanine are not methylated and should always be fully converted to thymidines. This might be used as a control for the bisulphate treatment reaction. In vitromethylation may be performed using the CpG methylase M.Sss I and used as a positive control for methylation detection. One PCR primer in each pair has to be biotinylated, which allows for separation of the double stranded PCR-products into a single stranded template. Pyrosequencing analyzes single-stranded DNA templates by synthesizing the complementary strand (Fig. 10). The nucleotides are dispensed by the Pyrosequencing instrument during the sequencing reaction. For every successful incorporation of a nucleotide, pyrophosphate (PPi) is released. PPi is converted in enzyme-catalyzed reactions to drive light emission in a quantity that is proportional to the number of incorporations.

Fig. 10. The principle of Pyrosequencing. Nucleotides are added sequentially by the Pyrosequencing instrument. For every successful incorporation, pyrophosphate (PPi) is released. PPi is converted in enzyme-catalyzed reactions to drive light emission in a quantity that is proportional to the number of incorporations. The light produced is detected by a charge coupled device (CCD) camera and seen as a peak in a pyrogram. Figure adapted from Biotage (Uppsala).

The major advantages in this method is that it allows methylation detection of multiple (10-15) CpG sites in the same reaction in a quantitative manner and it is reasonable high throughput as it is performed in 96-well format. Although, a disadvantage is that Pyrosequencing does not provide us with

37

the information on the methylation haplotype and produces relatively short stretches of sequence.

38

Aims of the study

To identify unknown conserved non-genic (CNG) sequences that may act as regulatory elements within the NF2 locus, by comparison of orthologous DNA sequence derived from evolutionarily distant species.

To determine common regions of loss or gain on chromosome 22 in tumor- and constitutional derived DNA that could indicate the position of putative gene(s) involved in the development of meningioma. The aim was also to analyze point mutations in exons and CNG sequences in the NF2 locus as well as determine the methylation profile of the NF2 promoter region.

To determine common regions of loss or gain on chromosome 22, in constitutional and tumor-derived tissue from sporadic schwannomas, that could indicate the position of putative gene(s) involved in the tumorigenesis of sporadic schwannoma and schwannomatosis.

To study disease-associated genetic aberrations outside the NF2 locus on chromosome 22 in constitutional- and tumor-derived samples from schwannomatosis and NF2 patients, aiming at the identification of a schwannomatosis candidate gene.

39

Results and discussion

Paper I. Strong conservation of the human NF2 locus based on sequence comparison in five species The DNA sequence covering the human NF2 locus and orthologous loci from four additional species (baboon, mouse, rat and pufferfish) were compared with the purpose of identifying conserved non-genic sequences (CNGs) that could act as regulatory elements of the NF2 gene. The aim was to further analyze the identified CNGs for mutations in DNA derived from patients affected with NF2 associated tumors. A commonly used criterion in human-mouse sequence comparisons for a sequence to be qualified as functionally conserved is at least 70% sequence identity over at least 100 bp of gap-free alignment. The percentage sequence that was considered as conserved (using the above mentioned criteria) in the non-coding part of the NF2 gene was higher (3.8%) compared to the protein coding part of the gene (2.7 %). This extra-exonic sequence conservation could to some extent be explained by the presence of additional transcriptional unit(s) within the NF2locus.

Pufferfish was the evolutionarily most distant species used in the sequence comparisons. The analysis of pufferfish sequence revealed a duplication of neurofibromin 2. In silico translation identified two putative proteins with 74% and 76% identity to the human protein. The DNA sequence conservation between pufferfish and human was mainly restricted to exonic sequence and no conservation within the promoter regions immediately upstream exon 1 could be identified. There was, however, a CNG between promoters of the NIPSNAP1 and NF2 genes that displayed conservation between all five species. This CNG, called Inter 1, might represent a novel regulatory element, important for the function of this locus. Analysis of available ESTs provided no evidence for inter1 to constitute an additional transcriptional unit. There is strong support that non-genic sequences which display high conservation between human and pufferfish may possess regulatory functions and even be targeted for pathogenic mutations.

40

Paper II. Comprehensive genetic and epigenetic analysis of sporadic meningioma for macro-mutations of 22q and micro-mutations within the NF2 locus

Meningioma is a very frequently occurring intracranial tumor that is both clinically and histopathologically heterogeneous. Meningioma is also a part of the complex tumor syndrome NF2. In this paper, we studied a comprehensive series of sporadic meningioma on multiple levels. The presence of macro-mutations, involving large scale aberrations of 22q were investigated using array-CGH. The occurrence of micro-mutations within the NF2 locus was analyzed by searching for point mutations within the mRNA encoding sequence and within CNGs that were defined in paper I. We also thoroughly studied the methylation status of the promoter region of the NF2gene.

When macro-mutations are considered, the predominant profile was monosomy 22 detected in 48% of meningiomas. The second category named as “other aberrations” accounted for 13% of samples and comprised terminal and, interstitial deletions as well as gains. At least two minimal overlapping candidate regions outside the NF2 locus were identifies and should be studied further. These regions; R1 (~550 kb) and R2 (~250 kb), are small enough to select a limited number of candidate genes for further analysis. The third category of tumors (39%) did not display any in gene dosage changes on 22q. Mutation analysis within the mRNA encoding part and within the selected CNGs which were either exon flanking or not contiguous with exons, involved the sequencing of 17.5 kb in totally100 tumors. Thirty nine alterations were identified among the 100 samples, 38 of the mutations were positioned in the coding sequence or in splice sites and one was identified in the CNG inter1. Overall, we could detect mutations consistent with biallelic inactivation in 36% of meningiomas. Finally, analysis of methylation at 40 CpG sites distributed within 750 bp of the promoter region of the NF2 gene suggests that aberrant methylation of this locus does not play a major role in meningioma development.

We observed pronounced differences of macro- and micro-mutation frequency between different histopathological tumor subtypes. We found a higher frequency of bi-allelic NF2 inactivation in fibroblastic (52%), compared to meningothelial (18%) tumors. Moreover, the presence of macro-mutations on 22q also shows marked differences between fibroblastic (86%) and meningothelial (39%) subtypes. This genetic heterogeneity is likely to be the underlying factor which determines the histological diversity observed in these tumors. Application of high-resolution genome-wide

41

arrays would most probably contribute in the identification of define specific factors which contribute to their genesis.

Paper III. High-resolution array-CGH profiling of germline and tumor- specific copy number alterations on chromosome 22 in patients affected with schwannomasThe rationale of paper III was to determine whether common regions of loss or gain on chromosome 22 could indicate the position of putative gene(s), in addition to the NF2 gene, which could be involved in the development of schwannoma and/or schwannomatosis. We analyzed schwannoma and constitutional derived DNA from patients to distinguish between tumor specific changes and aberrations transmitted through germ line. In the 88 tumors analyzed, we detected heterozygous deletions in 53% of the schwannomas. Several of the alterations detected had common breakpoints, indicating the presence of recombination–prone, unstable genomic structures. The most commonly detected aberration was monosomy 22. The second largest aberration was terminal deletions affecting almost the entire 22q. We also identified overlapping interstitial deletions of various sizes encompassing the NF2 gene. This confirms the importance of the NF2 gene in schwannoma development. However, it cannot be ruled out that the large deletions detected also affect other genes which may be involved in schwannoma development.

Copy number variation analysis in constitutional DNA from schwannoma patients identified eight different loci across chromosome 22, which did not overlap the NF2 locus. Four of the discovered alterations occurred in only one individual, called Rare Copy Number Variation (RCNV). The remaining four variations were observed in more than one individual, called Common Copy Number Variation (CCNV). To test if these CNVs may be involved in schwannoma predisposition we further analyzed peripheral blood derived DNA from 18 phenotypically normal individuals. We could confirm the presence of all CCNVs, but none of the RCNVs among the normal controls. We conclude that the RCNVs identified in this work may involve additional genes important for schwannoma predisposition, and should be further investigated. Although, an assessment of larger number of normal individuals is necessary in order to prove these results.

42

Paper IV. Identification of genetic aberrations on chromosome 22 outside the NF2 locus in schwannomatosis and neurofibromatosis type 2 The genetic factors underlying the differences between schwannomatosis and NF2 are poorly understood, although available evidence implicates chromosome 22 as the primary locus of interest. Analyses of the NF2 gene in schwannoma tumors from schwannomatosis patients reveal the presence of inactivating mutations in a similar spectrum as in NF2-derived schwannomas. However, when analyzing familial schwannomatosis cases, no NF2 gene mutations were found in constitutional DNA of these patients. In addition, linkage analysis excluded NF2 gene alterations to be the predisposing germline event, and proposed that a schwannomatosis gene exists centromeric of the NF2 gene 108. In this study we profiled DNA samples from sporadic and familial schwannomatosis, NF2 and a large cohort of normal controls. Using a tiling-path chromosome 22 genomic array, we identified two candidate regions of copy number variation (R1 and R2), which were further characterized by a PCR-product based array with higher resolution. The latter approach allows detection of minute alterations in total genomic DNA regardless of the content of repetitive sequence. We detected rearrangements of the immunoglobulin lambda (IGL) locus (candidate region R1) in peripheral blood-derived DNA from one schwannomatosis patient, which cannot be exclusively due to B-cell specific somatic recombination of IGL. Analysis of sporadic schwannomas provided further evidence for the specificity of this genomic imbalance in samples related to the schwannomatosis and NF2. This IGL locus rearrangement detected in sporadic schwannoma S-473, was similar to the above described in the schwannomatosis samples. Analysis of normal controls indicated that these IGL rearrangements were restricted to schwannomatosis/schwannoma patients.

We also constructed a high-resolution array across the second region of interest (R2) that maps in the region of the maximum LOD score of 6.6 determined by the above described linkage study 108. We detected DNA copy number polymorphisms affecting the glutathione S-transferase theta 1 (GSTT1) locus. GSTT1 DNA copy number polymorphism has been previously associated with numerous cancer types. As the GSTT1 gene is located immediately upstream of the calcineurin binding protein 1 (CABIN1)gene promoter, one could envisage that this polymorphic variation might, in some individuals, affect the normal function of the CABIN1 gene promoter and/or other upstream gene regulatory elements. We further describe missense mutations in the CABIN1 gene that are specific to samples from

43

schwannomatosis, NF2 and schwannoma. This makes CABIN1 a plausible candidate gene, which may contribute to the pathogenesis of these disorders.

44

Future perspective

During the course of this work I focused mainly on studying genetic aberrations in both constitutional and tumor derived DNA from patients that are affected by NF2, or by tumors that are associated with NF2. The purpose was to identify mechanism(s) that, in addition to inactivation of the NF2tumor suppressor gene, contributes to the development of these tumors. One of the main tools used for this purpose was a chromosome 22 tiling path array that was constructed within our research group. The continued work on chromosome 22 should mainly involve even more comprehensive characterization of the schwannomatosis candidate locus with a PCR-based genomic array. In this context it should be mentioned that the candidate region 1 (R1) that were identified in sporadic meningiomas (paper II) falls within the region of 22q which was indicated to contain the schwannomatosis predisposition locus 108. The focus could be on the integration of gene dosage analysis together with changes of methylation status of CpG island and gene expression analysis. Methylation and expression profiling would be performed for all known and predicted genes located within the schwannomatosis candidate predisposition region. Methodology for construction of genomic arrays allowing integration of the above three levels of analysis exists today. Furthermore, resolution of such an array can be down to below 1 kb per data point and at the same time cover large genomic regions, such as a single chromosome.

As we identified the CABIN1 gene as a plausible candidate for schwannomatosis and NF2 phenotype modification (paper IV), this gene should be further studied. An increased number of both familial and sporadic cases of NF2 and schwannomatosis, as well as tumors associated with NF2 could be analyzed for mutations. A larger series of normal controls should also be tested to confirm that mutations are specific to the patient groups tested. This would provide a more complete picture of the distribution and frequency of CABIN1 mutations in DNA from this cohort of patients. The methylation status of the CABIN1 promoter CpG island could also be a subject for investigation. The above line of investigation should also take into account other candidate loci, for instance the locus in the telomeric vicinity of CABIN1 gene identified in the course of analysis of ependymomas 155, 156. The identified conserved non-genic sequence (inter1)positioned approximately 15 kb 5’ of the NF2 gene, displayed conservation

45

between all five species which is likely reflecting its functional importance (paper I). Therefore, inter 1 could be an interesting candidate for further functional studies.

Our findings related to chromosome 22, regarding analyses of constitutional and tumor-derived DNA, should also be extended towards the entire human genome. We have recently established in our research group a genome wide array comprised of ~32,000 BAC clones, providing 99% coverage of the sequenced part of the human genome, with an average resolution of below 100 kb 157 (Bruder et al, manuscript in preparation). This array represents a high-resolution artificial genome printed on a single microscope slide, offering the possibility to globally analyze the human genome for genetic changes as well as epigenetic aberrations. Before copy number variation at a specific locus could be associated to any type of disease predisposition, a large cohort of phenotypically normal individuals has to be analyzed, in order to establish a consensus of normal variation. This whole genome array could be applied in the analysis of series of patients that were studied in this thesis: (i) to investigate differences in copy number between patients affected by the mild and severe NF2 phenotype, (ii) to profile DNA derived from sporadic schwannomatosis, and if DNA samples are available, also cases of familial schwannomatosis, to identify the predisposing germline event, (iii) to analyze patients affected with sporadic schwannoma which could provide with clues in the search for the schwannomatosis candidate locus, (iv) to study sporadic meningiomas to determine the differences in the gene dosage between different histopathological phenotypes.

46

Acknowledgements

This work was performed at the Department of Genetics and Pathology, Uppsala University. It was supported by grants from the U.S. Army Medical Research and Materiel Command, the Swedish Cancer Foundation, the Swedish Research Council and Uppsala University.

I would like to express my sincere gratitude to a lot of people that made this thesis come true.

First of all, I would like to thank my supervisor Jan Dumanski for accepting me as a PhD student, for teaching me about science, about the painful procedure of writing about science and encouraging me to become an independent scientist. I have also appreciated the open atmosphere in the lab, always welcoming new people to join our group.

There are/were a bunch of really great people in the ‘Dumanski lab’ and it feels a little bit sad that we now will be spread all over the world. Good luck to you all!

THANKS to:

I would like to start with the people that I probably spent the most time with; Magda, Cecilia, Patrick, Kiran, Caroline, for being such great room mates, good collaborators in the lab and fun party-people! And of course, our new room mates Johanna and Helena.

The girls in the lab: Tere, I have very much enjoyed working with you. A special thanks for always taking your time to read and correct my writing, giving me a shoulder to cry on and for haircuts in the lab; Brigitta, thanks for being fun and a good friend (and for all mp3-books); Ann-Mari and Lotta Thyr for good times in the lab; Lotta Thuresson for being such a sparkling and friendly person and Elisabeth for helping out with paperwork.

The boys in the boys room: Ricardo, Uwe, and Gintas mainly for helping out when the damn computer don't do what I tell it to do; and Arek, for being such a friendly, helpful and beer drinking guy.

47

Calle, for being such a happy ‘prick’ and for keeping track of all my lab-equipment…

Magda and Cecilia, for us being the girls hanging in until the end…

I would also like to thank some past group members: Giedre (for fighting the MEGAbajs with me), John D, Anders H, Lotta WV, Carina, Ingegerd, Isabel, Hajni, Haider, and Patrik Boman.

I got to know a bunch of really wonderful people during the years at Rudbeck. I will not even begin counting names….but you know who you are!

Jag vill också nämna alla härliga ‘hästmänniskor’ som jag har lärt känna i lilla Djurgården och i Norby stallet. Livet i stallet har varit mitt andnigshål. Särskilt tack till Madde och Zorba, tack vare er började jag rida igen; Tesa och Biggen, ni är verkligen två positiva själar. Sist och mest vill jag tacka Ami och Tim för allt roligt vi gjort tillsammans, det har betytt massor för mig!

Min kära familj, jag älskar er. Mamma och pappa, för att ni alltid tror på mig och för att ni stöttar mig i alla väder. Ni har skämt bort mig, och det upskattar jag! Lilly och Sixten, min underbara farmor och farfar. Samt resten av min stora familj, ni är många…

Marcus

Caisa

48

References

1. Futreal, P. A. et al. A census of human cancer genes. Nat Rev Cancer 4, 177-83 (2004).

2. Nowell, P. C. & Hungerford, D. A. Chromosome studies on normal and leukemic human leukocytes. J Natl Cancer Inst 25, 85-109 (1960).

3. Knudson, A. G., Jr. Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci U S A 68, 820-3 (1971).

4. Santarosa, M. & Ashworth, A. Haploinsufficiency for tumour suppressor genes: when you don't need to go all the way. Biochim Biophys Acta 1654, 105-22 (2004).

5. Song, W. J. et al. Haploinsufficiency of CBFA2 causes familial thrombocytopenia with propensity to develop acute myelogenous leukaemia. Nat Genet 23, 166-75 (1999).

6. Xu, X. et al. Haploid loss of the tumor suppressor Smad4/Dpc4 initiates gastric polyposis and cancer in mice. Oncogene 19, 1868-74 (2000).

7. Friedberg, E. C. DNA damage and repair. Nature 421, 436-40 (2003).

8. Hanahan, D. & Weinberg, R. A. The hallmarks of cancer. Cell 100, 57-70 (2000).

9. P. Fedi, S. R. T. a. S. A. A. Growth factors (ed. .F. Holland, R. C. B., D.L. Morton, E. Frei, D.W. Kufe and R.R. Weichselbaum) (Williams and Wilkins, Baltimore, 1997).

10. Evan, G. & Littlewood, T. A matter of life and cell death. Science 281, 1317-22 (1998).

11. Harris, C. C. p53 tumor suppressor gene: from the basic research laboratory to the clinic--an abridged historical perspective. Carcinogenesis 17, 1187-98 (1996).

12. Volpert, O. V., Dameron, K. M. & Bouck, N. Sequential development of an angiogenic phenotype by human fibroblasts progressing to tumorigenicity. Oncogene 14, 1495-502 (1997).

13. Hanahan, D. & Folkman, J. Patterns and emerging mechanisms of the angiogenic switch during tumorigenesis. Cell 86, 353-64 (1996).

14. Veikkola, T. & Alitalo, K. VEGFs, receptors and angiogenesis. Semin Cancer Biol 9, 211-20 (1999).

15. Nowell, P. C. Tumor progression: a brief historical perspective. Semin Cancer Biol 12, 261-6 (2002).

49

16. Lengauer, C., Kinzler, K. W. & Vogelstein, B. Genetic instabilities in human cancers. Nature 396, 643-9 (1998).

17. Jacobs, P. A., Baikie, A. G., Court Brown, W. M. & Strong, J. A. The somatic chromosomes in mongolism. Lancet 1, 710 (1959).

18. Edwards, J. H., Harnden, D. G., Cameron, A. H., Crosse, V. M. & Wolff, O. H. A new trisomic syndrome. Lancet 1, 787-90 (1960).

19. Patau, K., Smith, D. W., Therman, E., Inhorn, S. L. & Wagner, H. P. Multiple congenital anomaly caused by an extra autosome. Lancet 1, 790-3 (1960).

20. Bobrow, M., Joness, L. F. & Clarke, G. A complex chromosomal rearrangement with formation of a ring 4. J Med Genet 8, 235-9 (1971).

21. Jacobs, P. A., Matsuura, J. S., Mayer, M. & Newlands, I. M. A cytogenetic survey of an institution for the mentally retarded: I. Chromosome abnormalities. Clin Genet 13, 37-60 (1978).

22. Warburton, D. De novo balanced chromosome rearrangements and extra marker chromosomes identified at prenatal diagnosis: clinical significance and distribution of breakpoints. Am J Hum Genet 49, 995-1013 (1991).

23. Jacobs, P. A., Browne, C., Gregson, N., Joyce, C. & White, H. Estimates of the frequency of chromosome abnormalities detectable in unselected newborns using moderate levels of banding. J Med Genet 29, 103-8 (1992).

24. Lubs, H. A. A marker X chromosome. Am J Hum Genet 21, 231-44 (1969).

25. Albertson, D. G. & Pinkel, D. Genomic microarrays in human genetic disease and cancer. Hum Mol Genet 12 Spec No 2, R145-52 (2003).

26. Amann, J., Valentine, M., Kidd, V. J. & Lahti, J. M. Localization of chi1-related helicase genes to human chromosome regions 12p11 and 12p13: similarity between parts of these genes and conserved human telomeric-associated DNA. Genomics 32, 260-5 (1996).

27. Trask, B. J. et al. Large multi-chromosomal duplications encompass many members of the olfactory receptor gene family in the human genome. Hum Mol Genet 7, 2007-20 (1998).

28. Eichler, E. E., Archidiacono, N. & Rocchi, M. CAGGG repeats and the pericentromeric duplication of the hominoid genome. Genome Res 9, 1048-58 (1999).

29. Roth, D. B. & Wilson, J. H. Nonhomologous recombination in mammalian cells: role for short sequence homologies in the joining reaction. Mol Cell Biol 6, 4295-304 (1986).

30. Bailey, J. A., Yavor, A. M., Massa, H. F., Trask, B. J. & Eichler, E. E. Segmental duplications: organization and impact within the current human genome project assembly. Genome Res 11, 1005-17 (2001).

50

31. Shaw, C. J. & Lupski, J. R. Implications of human genome architecture for rearrangement-based disorders: the genomic basis of disease. Hum Mol Genet 13 Spec No 1, R57-64 (2004).

32. Nobile, C. et al. Analysis of 22 deletion breakpoints in dystrophin intron 49. Hum Genet 110, 418-21 (2002).

33. Stankiewicz, P. et al. Genome architecture catalyzes nonrecurrent chromosomal rearrangements. Am J Hum Genet 72, 1101-16 (2003).

34. Stankiewicz, P. & Lupski, J. R. Genome architecture, rearrangements and genomic disorders. Trends Genet 18, 74-82 (2002).

35. Mantripragada, K. K. et al. DNA copy-number analysis of the 22q11 deletion-syndrome region using array-CGH with genomic and PCR-based targets. Int J Mol Med 13, 273-9 (2004).

36. Mantripragada, K. K. et al. Identification of novel deletion breakpoints bordered by segmental duplications in the NF1 locus using high resolution array-CGH. J Med Genet 43, 28-38 (2006).

37. Inoue, K. & Lupski, J. R. Molecular mechanisms for genomic disorders. Annu Rev Genomics Hum Genet 3, 199-242 (2002).

38. Lander, E. S. et al. Initial sequencing and analysis of the human genome. Nature 409, 860-921 (2001).

39. Venter, J. C. et al. The sequence of the human genome. Science 291, 1304-51 (2001).

40. Beaudet, L., Bedard, J., Breton, B., Mercuri, R. J. & Budarf, M. L. Homogeneous assays for single-nucleotide polymorphism typing using AlphaScreen. Genome Res 11, 600-8 (2001).

41. Iafrate, A. J. et al. Detection of large-scale variation in the human genome. Nat Genet 36, 949-51 (2004).

42. Sebat, J. et al. Large-scale copy number polymorphism in the human genome. Science 305, 525-8 (2004).

43. de Stahl, T. D. et al. Chromosome 22 tiling-path array-CGH analysis identifies germ-line- and tumor-specific aberrations in patients with glioblastoma multiforme. Genes Chromosomes Cancer 44, 161-9 (2005).

44. Feuk, L., Carson, A. R. & Scherer, S. W. Structural variation in the human genome. Nat Rev Genet 7, 85-97 (2006).

45. Conrad, D. F., Andrews, T. D., Carter, N. P., Hurles, M. E. & Pritchard, J. K. A high-resolution survey of deletion polymorphism in the human genome. Nat Genet 38, 75-81 (2006).

46. Weber, J. L. et al. Human diallelic insertion/deletion polymorphisms. Am J Hum Genet 71, 854-62 (2002).

47. Bhangale, T. R., Rieder, M. J., Livingston, R. J. & Nickerson, D. A. Comprehensive identification and characterization of diallelic insertion-deletion polymorphisms in 330 human candidate genes. Hum Mol Genet 14, 59-69 (2005).

51

48. Shen, L. X., Basilion, J. P. & Stanton, V. P., Jr. Single-nucleotide polymorphisms can cause different structural folds of mRNA. Proc Natl Acad Sci U S A 96, 7871-6 (1999).

49. Wiltshire, T. et al. Genome-wide single-nucleotide polymorphism analysis defines haplotype patterns in mouse. Proc Natl Acad Sci U S A 100, 3380-5 (2003).

50. Cuppens, H. et al. Polyvariant mutant cystic fibrosis transmembrane conductance regulator genes. The polymorphic (Tg)m locus explains the partial penetrance of the T5 polymorphism as a disease mutation. J Clin Invest 101, 487-96 (1998).

51. Gambetti, P., Parchi, P. & Chen, S. G. Hereditary Creutzfeldt-Jakob disease and fatal familial insomnia. Clin Lab Med 23, 43-64 (2003).

52. Altshuler, D. et al. A haplotype map of the human genome. Nature 437, 1299-320 (2005).

53. Fraga, M. F. et al. Epigenetic differences arise during the lifetime of monozygotic twins. Proc Natl Acad Sci U S A 102, 10604-9 (2005).

54. Takai, D. & Jones, P. A. Comprehensive analysis of CpG islands in human chromosomes 21 and 22. Proc Natl Acad Sci U S A 99, 3740-5 (2002).

55. O'Neill, R. J., O'Neill, M. J. & Graves, J. A. Undermethylation associated with retroelement activation and chromosome remodelling in an interspecific mammalian hybrid. Nature 393, 68-72 (1998).

56. Ballestar, E. & Esteller, M. The impact of chromatin in human cancer: linking DNA methylation to gene silencing. Carcinogenesis 23, 1103-9 (2002).

57. Goto, T. & Monk, M. Regulation of X-chromosome inactivation in development in mice and humans. Microbiol Mol Biol Rev 62, 362-78 (1998).

58. Tycko, B. DNA methylation in genomic imprinting. Mutat Res 386, 131-40 (1997).

59. Iguchi-Ariga, S. M. & Schaffner, W. CpG methylation of the cAMP-responsive enhancer/promoter sequence TGACGTCA abolishes specific factor binding as well as transcriptional activation. Genes Dev 3, 612-9 (1989).

60. Nan, X., Cross, S. & Bird, A. Gene silencing by methyl-CpG-binding proteins. Novartis Found Symp 214, 6-16; discussion 16-21, 46-50 (1998).

61. Bird, A. P. & Wolffe, A. P. Methylation-induced repression--belts, braces, and chromatin. Cell 99, 451-4 (1999).

62. Bestor, T., Laudano, A., Mattaliano, R. & Ingram, V. Cloning and sequencing of a cDNA encoding DNA methyltransferase of mouse cells. The carboxyl-terminal domain of the mammalian enzymes is related to bacterial restriction methyltransferases. J Mol Biol 203, 971-83 (1988).

52

63. Okano, M., Bell, D. W., Haber, D. A. & Li, E. DNA methyltransferases Dnmt3a and Dnmt3b are essential for de novo methylation and mammalian development. Cell 99, 247-57 (1999).

64. Li, E., Bestor, T. H. & Jaenisch, R. Targeted mutation of the DNA methyltransferase gene results in embryonic lethality. Cell 69, 915-26 (1992).

65. Hsieh, C. L. In vivo activity of murine de novo methyltransferases, Dnmt3a and Dnmt3b. Mol Cell Biol 19, 8211-8 (1999).

66. Feinberg, A. P., Ohlsson, R. & Henikoff, S. The epigenetic progenitor origin of human cancer. Nat Rev Genet 7, 21-33 (2006).

67. Esteller, M. & Herman, J. G. Cancer as an epigenetic disease: DNA methylation and chromatin alterations in human tumours. J Pathol 196, 1-7 (2002).

68. Thayer, R. E., Singer, M. F. & Fanning, T. G. Undermethylation of specific LINE-1 sequences in human cells producing a LINE-1-encoded protein. Gene 133, 273-7 (1993).

69. Alves, G., Tatro, A. & Fanning, T. Differential methylation of human LINE-1 retrotransposons in malignant cells. Gene 176, 39-44 (1996).

70. Yoder, J. A., Walsh, C. P. & Bestor, T. H. Cytosine methylation and the ecology of intragenomic parasites. Trends Genet 13, 335-40 (1997).

71. Feinberg, A. P. Imprinting of a genomic domain of 11p15 and loss of imprinting in cancer: an introduction. Cancer Res 59, 1743s-1746s (1999).

72. Dunham, I. et al. The DNA sequence of human chromosome 22. Nature 402, 489-95 (1999).

73. Deloukas, P. et al. A physical map of 30,000 human genes. Science 282, 744-6 (1998).

74. Woodfine, K. et al. Replication timing of the human genome. Hum Mol Genet 13, 191-202 (2004).

75. Evans, D. G. et al. Incidence of vestibular schwannoma and neurofibromatosis 2 in the North West of England over a 10-year period: higher incidence than previously thought. Otol Neurotol 26, 93-7 (2005).

76. Evans, D. G. et al. A clinical study of type 2 neurofibromatosis. Q J Med 84, 603-18 (1992).

77. Parry, D. M. et al. Neurofibromatosis 2 (NF2): clinical characteristics of 63 affected individuals and clinical evidence for heterogeneity. Am J Med Genet 52, 450-61 (1994).

78. Evans, D. G. et al. A genetic study of type 2 neurofibromatosis in the United Kingdom. I. Prevalence, mutation rate, fitness, and confirmation of maternal transmission effect on severity. J Med Genet 29, 841-6 (1992).

53

79. Rouleau, G. A. et al. Genetic linkage of bilateral acoustic neurofibromatosis to a DNA marker on chromosome 22. Nature 329, 246-8 (1987).

80. Pulst, S. M., Rouleau, G. A., Marineau, C., Fain, P. & Sieb, J. P. Familial meningioma is not allelic to neurofibromatosis 2. Neurology 43, 2096-8 (1993).

81. Trofatter, J. A. et al. A novel moesin-, ezrin-, radixin-like gene is a candidate for the neurofibromatosis 2 tumor suppressor. Cell 75, 826 (1993).

82. Kluwe, L. et al. Identification of NF2 germ-line mutations and comparison with neurofibromatosis 2 phenotypes. Hum Genet 98, 534-8 (1996).

83. Baser, M. E. et al. The location of constitutional neurofibromatosis 2 (NF2) splice site mutations is associated with the severity of NF2. J Med Genet 42, 540-6 (2005).

84. Parry, D. M. et al. Germ-line mutations in the neurofibromatosis 2 gene: correlations with disease severity and retinal abnormalities. Am J Hum Genet 59, 529-39 (1996).

85. Ruttledge, M. H. et al. Type of mutation in the neurofibromatosis type 2 gene (NF2) frequently determines severity of disease. Am J Hum Genet 59, 331-42 (1996).

86. Kluwe, L. et al. Screening for large mutations of the NF2 gene. Genes Chromosomes Cancer 42, 384-91 (2005).

87. Chang, L. S., Akhmametyeva, E. M., Wu, Y., Zhu, L. & Welling, D. B. Multiple transcription initiation sites, alternative splicing, and differential polyadenylation contribute to the complexity of human neurofibromatosis 2 transcripts. Genomics 79, 63-76 (2002).

88. Tsukita, S. & Yonemura, S. ERM proteins: head-to-tail regulation of actin-plasma membrane interaction. Trends Biochem Sci 22, 53-8 (1997).

89. Kondo, T. et al. ERM (ezrin/radixin/moesin)-based molecular mechanism of microvillar breakdown at an early stage of apoptosis. J Cell Biol 139, 749-58 (1997).

90. Matsui, T. et al. Rho-kinase phosphorylates COOH-terminal threonines of ezrin/radixin/moesin (ERM) proteins and regulates their head-to-tail association. J Cell Biol 140, 647-57 (1998).

91. Xu, H. M. & Gutmann, D. H. Merlin differentially associates with the microtubule and actin cytoskeleton. J Neurosci Res 51, 403-15 (1998).

92. James, M. F., Manchanda, N., Gonzalez-Agosti, C., Hartwig, J. H. & Ramesh, V. The neurofibromatosis 2 protein product merlin selectively binds F-actin but not G-actin, and stabilizes the filaments through a lateral association. Biochem J 356, 377-86 (2001).

54

93. Brault, E. et al. Normal membrane localization and actin association of the NF2 tumor suppressor protein are dependent on folding of its N-terminal domain. J Cell Sci 114, 1901-12 (2001).

94. Scoles, D. R. et al. Neurofibromatosis 2 tumour suppressor schwannomin interacts with betaII-spectrin. Nat Genet 18, 354-9 (1998).

95. Shaw, R. J. et al. The Nf2 tumor suppressor, merlin, functions in Rac-dependent signaling. Dev Cell 1, 63-72 (2001).

96. Gonzalez-Agosti, C. et al. The merlin tumor suppressor localizes preferentially in membrane ruffles. Oncogene 13, 1239-47 (1996).

97. Pelton, P. D. et al. Ruffling membrane, stress fiber, cell spreading and proliferation abnormalities in human Schwannoma cells. Oncogene 17, 2195-209 (1998).

98. Stemmer-Rachamimov, A. O. et al. Expression of NF2-encoded merlin and related ERM family proteins in the human central nervous system. J Neuropathol Exp Neurol 56, 735-42 (1997).

99. Evans, D. G., Trueman, L., Wallace, A., Collins, S. & Strachan, T. Genotype/phenotype correlations in type 2 neurofibromatosis (NF2): evidence for more severe disease associated with truncating mutations. J Med Genet 35, 450-5 (1998).

100. Baser, M. E. et al. Genotype-phenotype correlations for nervous system tumors in neurofibromatosis 2: a population-based study. Am J Hum Genet 75, 231-9 (2004).

101. Kluwe, L. et al. Phenotypic variability associated with 14 splice-site mutations in the NF2 gene. Am J Med Genet 77, 228-33 (1998).

102. Mautner, V. F., Baser, M. E. & Kluwe, L. Phenotypic variability in two families with novel splice-site and frameshift NF2 mutations. Hum Genet 98, 203-6 (1996).

103. Scoles, D. R., Baser, M. E. & Pulst, S. M. A missense mutation in the neurofibromatosis 2 gene occurs in patients with mild and severe phenotypes. Neurology 47, 544-6 (1996).

104. Zucman-Rossi, J. et al. NF2 gene in neurofibromatosis type 2 patients. Hum Mol Genet 7, 2095-101 (1998).

105. Lopez-Correa, C., Zucman-Rossi, J., Brems, H., Thomas, G. & Legius, E. NF2 gene deletion in a family with a mild phenotype. J Med Genet 37, 75-7 (2000).

106. Kluwe, L. & Mautner, V. F. Mosaicism in sporadic neurofibromatosis 2 patients. Hum Mol Genet 7, 2051-5 (1998).

107. Bruder, C. E. et al. High resolution deletion analysis of constitutional DNA from neurofibromatosis type 2 (NF2) patients using microarray-CGH. Hum Mol Genet 10, 271-82 (2001).

108. MacCollin, M. et al. Familial schwannomatosis: exclusion of the NF2 locus as the germline event. Neurology 60, 1968-74 (2003).

55

109. Seizinger, B. R., Martuza, R. L. & Gusella, J. F. Loss of genes on chromosome 22 in tumorigenesis of human acoustic neuroma. Nature 322, 644-7 (1986).

110. Bijlsma, E. K., Brouwer-Mladin, R., Bosch, D. A., Westerveld, A. & Hulsebos, T. J. Molecular characterization of chromosome 22 deletions in schwannomas. Genes Chromosomes Cancer 5, 201-5 (1992).

111. Gusella, J. F., Ramesh, V., MacCollin, M. & Jacoby, L. B. Neurofibromatosis 2: loss of merlin's protective spell. Curr Opin Genet Dev 6, 87-92 (1996).

112. Bruder, C. E. et al. A group of schwannomas with interstitial deletions on 22q located outside the NF2 locus shows no detectable mutations in the NF2 gene. Hum Genet 104, 418-24 (1999).

113. Antinheimo, J. et al. Genetic aberrations in sporadic and neurofibromatosis 2 (NF2)-associated schwannomas studied by comparative genomic hybridization (CGH). Acta Neurochir (Wien) 142, 1099-104; discussion 1104-5 (2000).

114. Longstreth, W. T., Jr., Dennis, L. K., McGuire, V. M., Drangsholt, M. T. & Koepsell, T. D. Epidemiology of intracranial meningioma. Cancer 72, 639-48 (1993).

115. Russel, D. S. & Rubinstein, L. J. Pathology of the tumors of the nervous system (Edward Arnold, London, 1989).

116. Rausing, A., Ybo, W. & Stenflo, J. Intracranial meningioma--a population study of ten years. Acta Neurol Scand 46, 102-10 (1970).

117. Tohma, Y., Yamashima, T. & Yamashita, J. Immunohistochemical localization of cell adhesion molecule epithelial cadherin in human arachnoid villi and meningiomas. Cancer Res 52, 1981-7 (1992).

118. Nakasu, S., Hirano, A., Shimura, T. & Llena, J. F. Incidental meningiomas in autopsy study. Surg Neurol 27, 319-22 (1987).

119. Turgut, M., Palaoglu, S., Ozcan, O. E., Gurcay, O. & Eryilmaz, M. Multiple meningiomas of the central nervous system without the stigmata of neurofibromatosis. Clinical and therapeutic study. Neurosurg Rev 20, 117-23 (1997).

120. Antinheimo, J. et al. Population-based analysis of sporadic and type 2 neurofibromatosis-associated meningiomas and schwannomas. Neurology 54, 71-6 (2000).

121. Perry, A. et al. Merlin, DAL-1, and progesterone receptor expression in clinicopathologic subsets of meningioma: a correlative immunohistochemical study of 175 cases. J Neuropathol Exp Neurol 59, 872-9 (2000).

122. Ruttledge, M. H. et al. Evidence for the complete inactivation of the NF2 gene in the majority of sporadic meningiomas. Nat Genet 6, 180-4 (1994).

123. Papi, L. et al. Somatic mutations in the neurofibromatosis type 2 gene in sporadic meningiomas. Hum Genet 95, 347-51 (1995).

56

124. Simon, M. et al. Allelic losses on chromosomes 14, 10, and 1 in atypical and malignant meningiomas: a genetic model of meningioma progression. Cancer Res 55, 4696-701 (1995).

125. Wellenreuther, R. et al. Analysis of the neurofibromatosis 2 gene reveals molecular variants of meningioma. Am J Pathol 146, 827-32 (1995).

126. De Vitis, L. R. et al. Screening for mutations in the neurofibromatosis type 2 (NF2) gene in sporadic meningiomas. Hum Genet 97, 632-7 (1996).

127. Sulman, E. P. et al. Identification of a consistent region of allelic loss on 1p32 in meningiomas: correlation with increased morbidity. Cancer Res 58, 3226-30 (1998).

128. Ishino, S. et al. Loss of material from chromosome arm 1p during malignant progression of meningioma revealed by fluorescent in situ hybridization. Cancer 83, 360-6 (1998).

129. Bello, M. J. et al. High-resolution analysis of chromosome arm 1p alterations in meningioma. Cancer Genet Cytogenet 120, 30-6 (2000).

130. Buckley, P. G. et al. Comprehensive DNA copy number profiling of meningioma using a chromosome 1 tiling path microarray identifies novel candidate tumor suppressor loci. Cancer Res 65, 2653-61 (2005).

131. Hamilton, R. L. & Pollack, I. F. The molecular biology of ependymomas. Brain Pathol 7, 807-22 (1997).

132. Weremowicz, S., Kupsky, W. J., Morton, C. C. & Fletcher, J. A. Cytogenetic evidence for a chromosome 22 tumor suppressor gene in ependymoma. Cancer Genet Cytogenet 61, 193-6 (1992).

133. Bijlsma, E. K. et al. Molecular analysis of genetic changes in ependymomas. Genes Chromosomes Cancer 13, 272-7 (1995).

134. Wernicke, C. et al. Involvement of chromosome 22 in ependymomas. Cancer Genet Cytogenet 79, 173-6 (1995).

135. Lamszus, K. et al. Molecular genetic alterations on chromosomes 11 and 22 in ependymomas. Int J Cancer 91, 803-8 (2001).

136. Ebert, C. et al. Molecular genetic analysis of ependymal tumors. NF2 mutations and chromosome 22q loss occur preferentially in intramedullary spinal ependymomas. Am J Pathol 155, 627-32 (1999).

137. MacCollin, M., Woodfin, W., Kronn, D. & Short, M. P. Schwannomatosis: a clinical and pathologic study. Neurology 46, 1072-9 (1996).

138. Seppala, M. T. et al. Multiple schwannomas: schwannomatosis or neurofibromatosis type 2? J Neurosurg 89, 36-41 (1998).

139. Jacoby, L. B. et al. Molecular analysis of the NF2 tumor-suppressor gene in schwannomatosis. Am J Hum Genet 61, 1293-302 (1997).

57

140. Kaufman, D. L. et al. Somatic instability of the NF2 gene in schwannomatosis. Arch Neurol 60, 1317-20 (2003).

141. MacCollin, M. et al. Diagnostic criteria for schwannomatosis. Neurology 64, 1838-45 (2005).

142. Schwartz, S. et al. PipMaker--a web server for aligning two genomic DNA sequences. Genome Res 10, 577-86. (2000).

143. Schwartz, S. et al. MultiPipMaker and supporting tools: Alignments and analysis of multiple genomic DNA sequences. Nucleic Acids Res 31, 3518-24 (2003).

144. Mayor, C. et al. VISTA : visualizing global DNA sequence alignments of arbitrary length. Bioinformatics 16, 1046-7 (2000).

145. Dermitzakis, E. T., Reymond, A. & Antonarakis, S. E. Conserved non-genic sequences - an unexpected feature of mammalian genomes. Nat Rev Genet 6, 151-7 (2005).

146. Loots, G. G. et al. Identification of a coordinate regulator of interleukins 4, 13, and 5 by cross-species sequence comparisons. Science 288, 136-40 (2000).

147. Thomas, J. W. et al. Comparative analyses of multi-species sequences from targeted genomic regions. Nature 424, 788-93 (2003).

148. Nobrega, M. A., Ovcharenko, I., Afzal, V. & Rubin, E. M. Scanning human gene deserts for long-range enhancers. Science 302, 413 (2003).

149. Lettice, L. A. et al. A long-range Shh enhancer regulates expression in the developing limb and fin and is associated with preaxial polydactyly. Hum Mol Genet 12, 1725-35 (2003).

150. Sanger, F., Nicklen, S. & Coulson, A. R. DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci U S A 74, 5463-7. (1977).

151. Solinas-Toldo, S. et al. Matrix-based comparative genomic hybridization: biochips to screen for genomic imbalances. Genes Chromosomes Cancer 20, 399-407 (1997).

152. Pinkel, D. et al. High resolution analysis of DNA copy number variation using comparative genomic hybridization to microarrays. Nat Genet 20, 207-11 (1998).

153. Frommer, M. et al. A genomic sequencing protocol that yields a positive display of 5-methylcytosine residues in individual DNA strands. Proc Natl Acad Sci U S A 89, 1827-31 (1992).

154. Uhlmann, K., Brinckmann, A., Toliat, M. R., Ritter, H. & Nurnberg, P. Evaluation of a potential epigenetic biomarker by quantitative methyl-single nucleotide polymorphism analysis. Electrophoresis 23, 4072-9 (2002).

155. Ammerlaan, A. C. et al. Localization of a putative low-penetrance ependymoma susceptibility locus to 22q11 using a chromosome 22

58

tiling-path genomic microarray. Genes Chromosomes Cancer 43, 329-38 (2005).

156. Cecilia de Bustos, T. D. d. S., Arkadiusz Piotrowski, Kiran K. Mantripragada,, Patrick G. Buckley, E. D., Caisa Hansson, Gintautas Grigelionis, Uwe Menzel, & Dumanski, J. P. Analysis of Copy Number Variation in Normal Human Population within a Region Containing Complex Segmental Duplications on 22q11 Using High Resolution Array-CGH. Genomics. Accepted for publication (2006).

157. Ishkanian, A. S. et al. A tiling resolution DNA microarray with complete coverage of the human genome. Nat Genet 36, 299-303 (2004).

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 113

Editor: The Dean of the Faculty of Medicine

A doctoral dissertation from the Faculty of Medicine, UppsalaUniversity, is usually a summary of a number of papers. A fewcopies of the complete dissertation are kept at major Swedishresearch libraries, while the summary alone is distributedinternationally through the series Digital ComprehensiveSummaries of Uppsala Dissertations from the Faculty ofMedicine. (Prior to January, 2005, the series was publishedunder the title “Comprehensive Summaries of UppsalaDissertations from the Faculty of Medicine”.)

Distribution: publications.uu.seurn:nbn:se:uu:diva-6511

ACTAUNIVERSITATISUPSALIENSISUPPSALA2006