Microsatellite instability profiling of Lynch syndrome-associated cancers Ferreira, Ana Maria Monteiro

Transcription

1 University of Groningen Microsatellite instability profiling of Lynch syndrome-associated cancers Ferreira, Ana Maria Monteiro IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record Publication date: 2009 Link to publication in University of Groningen/UMCG research database Citation for published version (APA): Ferreira, A. M. M. (2009). Microsatellite instability profiling of Lynch syndrome-associated cancers s.n. Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons). Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Downloaded from the University of Groningen/UMCG research database (Pure): For technical reasons the number of authors shown on this cover page is limited to 10 maximum. Download date:

2 MICROSATELLITE INSTABILITY PROFILING OF LYNCH SYNDROME-ASSOCIATED CANCERS

3 The studies described in this thesis were supported by the Fundação para a Ciência e a Tecnologia (SFRH/BD/18832/2004), Portugal, and by the European Community (FP LIFESCIHEALTH-5, proposal no ). The printing costs of this thesis were supported by: Stichting Nationaal Fonds tegen Kanker voor onderzoek naar reguliere en aanvullende therapieën te Amsterdam; Fundação para a Ciência e a Tecnologia; University of Groningen; University Medical Center Groningen (UMCG); Graduate School for Drug Exploration (GUIDE). Printed by: Grafimedia Facilitair Bedrijf RUG Cover design by: Grafimedia Facilitair Bedrijf RUG 2009, A.M. Monteiro Ferreira. All rights are reserved. No part of this publication may be reproduced or transmitted in any form or by any means without permission of the author. ISBN:

4 MICROSATELLITE INSTABILITY PROFILING OF LYNCH SYNDROME-ASSOCIATED CANCERS Proefschrift ter verkrijging van het doctoraat in de Medische Wetenschappen aan de Rijksuniversiteit Groningen op gezag van de Rector Magnificus, dr. F. Zwarts, in het openbaar te verdedigen op woensdag 13 mei 2009 om uur door Ana Maria Monteiro Ferreira geboren op 4 november 1980 te Amarante, Portugal 3

5 Promotores: Prof. dr. R.M.W. Hofstra Prof. dr. R. Seruca Copromotores: Dr. H. Westers Dr. R.H. Sijmons Beoordelingscommissie: Prof. dr. M.J.E. Mourits Prof. dr. H. Morreau Prof. dr. J. Lubiński 4

6 Recomeça Se puderes Sem angústia E sem pressa. E os passos que deres, Nesse caminho duro Do futuro Dá-os em liberdade. Enquanto não alcances Não descanses. De nenhum fruto queiras só metade. E, nunca saciado, Vai colhendo ilusões sucessivas no pomar. Sempre a sonhar E vendo, Acordado O logro da aventura. És homem, não te esqueças! Só é tua a loucura Onde, com lucidez, te reconheças. Miguel Torga 5

7 Paranimfen: Maria Alves Mateusz Siedliński 6

8 CONTENTS Chapter 1: Introduction 9 General background 10 Aim and outline of the thesis 19 Chapter 2: Mononucleotide precedes dinucleotide instability during colorectal tumour development in Lynch syndrome patients 25 Chapter 3: Do microsatellite instability profiles really differ between colorectal and endometrial tumours? 43 Chapter 4: The hunt for new target genes in endometrial tumors reveals the involvement of the estrogen-receptor pathway in microsatellite unstable cancers 57 Chapter 5: Estrogens, MSI and Lynch syndrome-associated tumors 77 Chapter 6: General discussion, conclusions and future perspectives 97 Chapter 7: Summary 107 Nederlandse samenvatting 111 Resumo 114 Streszczenie 117 Acknowledgments 121 Curriculum Vitae 125 7

9 8

10 CHAPTER 1 Introduction 9

11 GENERAL BACKGROUND 1.1. Lynch syndrome Clinical definition Lynch syndrome is an autosomal dominant inherited cancer-susceptibility syndrome. It is named after Dr. Henry Lynch, whose role was crucial in the clinical and scientific identification of the syndrome as an inherited and relatively frequent cause of colorectal and extra-colonic cancer. Lynch syndrome is also known as hereditary nonpolyposis colorectal cancer (HNPCC), a rather misleading name, as several cancers other than colorectal also belong to the disease spectrum. Lynch syndrome is the most common hereditary cause of colorectal cancer. Long before the genetic mechanism underlying the disease was known, several major clinical features were described, and a first attempt to define a uniform set of minimal criteria for clinical diagnosis of Lynch syndrome based on family history was made in 1990, in a meeting of the International Collaborative Group on HNPCC (ICG-HNPCC), in Amsterdam (Vasen et al., 1991). These became known as the Amsterdam criteria (I). With time, and according to the new findings in the field, especially those related to the genetic basis of the disease, several refinements to this set of criteria were suggested, such as the Japanese, Mount Sinai, and Bethesda criteria (Fujita et al, 1996; Peltomaki et al., 2004; Umar et al., 2004). In 1999, the ICG-HNPCC proposed a new definition for HNPCC/Lynch syndrome and, with it, the revised Amsterdam criteria (II) (Vasen, 1999). (See Box.1) Genetics of Lynch syndrome The start of unravelling the genetic cause of Lynch syndrome was in 1993, when two major findings came together. One was the report of genetic instability associated with replication errors in microsatellite sequences in a large percentage of tumours from Lynch syndrome patients (Aaltonen et al., 1993; Ionov et al., 1993; Peltomaki et al., 1993a; Thibodeau et al., 1993). The other was the identification of two Lynch syndrome loci by linkage analysis, at chromosomes 2p and 3p (Lindblom et al., 1993; Peltomaki et al., 1993b). 10

12 Box 1. Details of the Amsterdam criteria for identifying Lynch syndrome families and the definition of the syndrome, by the 1999 International Collaborative Group on HNPCC (ICG-HNPCC) (Vasen et al., 1999). Amsterdam criteria II At least three relatives should have histologically verified colorectal cancer, cancer of the endometrium, small bowel, ureter, or renal pelvis; One of them should be a first-degree relative of the other two; Familial adenomatous polyposis (FAP) should be excluded; At least two successive generations should be affected; In one of the relatives colorectal cancer should be diagnosed before 50 years of age. ICG-HNPCC definition of HNPCC/Lynch syndrome Familial clustering of colorectal and/or endometrial cancer; Associated extra-colonic cancers: cancer of the stomach, ovary, ureter/renal pelvis, brain, small bowel, hepatobiliary tract, and skin (sebaceous tumours); Development of cancer at an early age; Development of multiple cancers; Features of colorectal cancer: (1) predilection for proximal colon; (2) improved survival; (3) multiple primary (synchronous/metachronous) colorectal cancers; (4) increased proportion of mucinous tumours, poorly differentiated tumours, and tumours with marked host-lymphocytic infiltration and lymphoid aggregation at the tumour margin; Features of colorectal adenoma: (1) the numbers vary from one to a few; (2) increased proportion of adenomas with a villous growth pattern and (3) probably rapid progression from adenoma to carcinoma; High frequency of MSI (MSI-H); Immunohistochemistry: loss of hmlh1, hmsh2, or hmsh6 protein expression; Germline mutation in MMR genes (hmsh2, hmlh1, hmsh3, hmsh6, hpms1,hpms2). During 1994, the first germline mutations were found in two genes identified in those loci (MSH2 and MLH1), both being human homologues of the well-known muts and mutl mismatch repair (MMR) genes of bacteria and yeast. Thus, deficient DNA mismatch repair was identified as the cause of Lynch syndrome. This functional inactivation of the DNA MMR genes is due to germline mutations as the first hit (Fishel et al., 1993; Leach et al.,1993; Bronner et al., 1994; Papadopoulos et al., 1994), followed by somatic inactivation of the second allele as 11

13 the second hit (Hemminki et al., 1994; Lu et al., 1996). This second hit is usually a somatic mutation or loss of heterozygosity (LOH). Germline mutations in the MLH1 and MSH2 genes form the vast majority of mutations found in Lynch syndrome cases (Peltomaki et al., 2004). Two other MMR genes MSH6 and PMS2 were later reported as being involved in the disease as well, since germline mutations are also found in a fraction of Lynch syndrome families (Berends et al., 2002; Hendriks et al., 2006). Germline deletion of the 3' exons of TACSTD1 can cause heritable somatic methylation and inactivation of the neighbouring MSH2 gene and thus Lynch syndrome (Kovacs et al., 2009; Ligtenberg et al., 2009). Also an interstitial deletion at 3p21.3 resulting in the genetic fusion of MLH1 and ITGA9 has been recently reported in a Lynch syndrome family, presumably defining a novel subclass of Lynch syndrome patients (Meyer et al., 2009). Several other genes, such as MLH3 and EXO1, also belong to the MMR pathway and these were therefore screened over the years as well. Germline mutations in MLH3 and EXO1 have been found (Wu et al., 2001a&b), but due to their low frequencies and type, mostly missense, they are not considered to be major players in Lynch syndrome (Hienonen et al., 2003; Jagmohan-Changur et al., 2003; Ou et al., 2008) Mismatch repair and microsatellite instability The MMR system is responsible for correcting errors that escape the activity of the polymerases during DNA replication. The system is able to correct mispaired nucleotides, as well as insertions and deletions loops (IDLs) that typically occur at short DNA tandem repeats - microsatellites. Therefore, when an MMR protein is inactivated, mutations will accumulate in those repeat sequences at a much higher rate (100- to 1000-fold) than that of spontaneous mutations in normal cells (Shibata et al., 1994). This phenomenon is referred to as microsatellite instability (MSI) (Ionov et al., 1993). It is easily recognized by decreased or increased lengths of the microsatellite, and therefore the detection of MSI became a key technique when searching for MMR-deficient tumours. MSI was reported in Lynch syndrome patients in 1993, occurring in over 90% of tumours in those patients, and it was another important piece of the puzzle 12

14 linking MMR deficiency and Lynch syndrome (Aaltonen et al., 1993; Ionov et al., 1993; Peltomaki et al., 1993a; Thibodeau et al., 1993). However, it is also found in a large proportion (15-25%) of sporadic tumours, not only of colorectal origin, but also in gastric and endometrial carcinomas (Boland et al., 1998). The underlying mechanism characterizing the sporadic forms of MSI tumours is also the functional inactivation of an MMR gene, namely MLH1, but in this case the bi-allelic inactivation of the gene is typically due to somatic promoter hypermethylation. MSI is a very early event in the tumorigenic process of tumours with MMR problems, as it has been detected in early lesions such as colorectal adenomas (Giuffrè et al., 2005) Adenoma-carcinoma sequence The adenoma-carcinoma sequence of colorectal cancer represents one of the best-known models of cancer development. Colorectal carcinomas arise through a multistep process, starting from early to high-grade dysplastic adenomas to carcinomas. This process of cancer development is basically caused by the progressive accumulation of genetic alterations in genes involved in cell growth, differentiation, proliferation, and apoptosis (Fearon & Vogelstein, 1990). This accumulation of genetic alterations is thought to be due to genetic instability, in which several distinct forms can be distinguished. Those that are bestdescribed are chromosomal instability (CIN) and microsatellite instability (MIN or MSI) (Royrvik et al., 2007). CIN is characterized by widespread chromosomal abnormalities such as aneuploidy and frequent loss of heterozygosity (LOH). MSI is caused by defects in the DNA mismatch repair (MMR) pathway, and is characterized by the accumulation of mutations in microsatellites (see above). MSI is found in the very early stages of the adenoma carcinoma sequence, although generally in lower frequencies than in carcinomas. It is reported in about 1-2% of sporadic adenomas (Young et al., 1993; Iino et al., 1999; Loukola et al., 1999; Sugai et al., 2003) and in 10-90% of Lynch syndrome-associated adenomas (Aaltonen et al., 1994; Iino et al., 2000; Giuffrè et al., 2005). This wide range of MSI frequencies might be explained by the method of dissection used (laser microdissection vs. manual dissection) and it is related to the multi-clonality of the 13

15 tissue, i.e. different areas show different degrees of MSI and different degrees of dysplasia (de Wind et al., 1998; Iino et al., 1999, 2000; Giuffrè et al. 2005; Greenspan et al., 2007). In addition, there may be considerable variation in the methods used to score MSI by different laboratories and between different observers MSI detection An international consensus panel of five microsatellite markers for detecting MSI was proposed in 1997 (Boland et al., 1998) to facilitate the production of easily comparable results, and this has become widely used. The panel includes two mononucleotide markers (BAT-25 and BAT-26) and three dinucleotide markers (D2S123, D5S346 and D17S250). Samples that are unstable for two or more of these markers are designated MSI-high (MSI-H), while samples unstable for one marker are MSI-low (MSI-L); samples that are stable for all the markers are designated microsatellite stable (MSS). If it is necessary to distinguish between MSI-L and MSS, then additional markers should be used (Boland et al., 1998). It is, in fact, common that some labs use a different number of markers. In that case, a sample is MSI-H if it is unstable for more than 30% of the markers used. More recently, a pentaplex PCR assay for 5 mononucleotide markers was proposed (Buhard et al., 2004). It includes the following markers: BAT-25, BAT-26, NR-21, NR-22 and NR-24. The authors claim a sensitivity and specificity of 100%, and suggest that the use of quasi-monomorphic mononucleotide repeats over dinucleotide repeats is advantageous, as the latter are typically polymorphic and more difficult to interpret. It is also believed that there is a greater sensitivity of dinucleotides for MSI-L cases than for MSI-H cases (Hatch et al., 2005). In addition, the use of mononucleotide markers might avoid needing normal tissue for comparison in CRC cases. In fact, it has also been proposed that BAT-26 alone and without normal matching mucosa might be sufficient for detecting MSI-H CRC (Hoang et al., 1997; de la Chapelle, 1999). The above-mentioned pentaplex panel has also been advised for endometrial carcinomas, although normal matching mucosa DNA is in that case still recommended (Wong et al., 2006). 14

16 1.5. Target genes and tissue selection Microsatellites are short DNA tandem repeat sequences spread throughout the genome, including non-coding and coding regions of genes. When MSI occurs in high frequencies in a coding sequence of a gene with important regulatory functions (involved in processes like apoptosis or proliferation for example), it is believed that such a gene, when impaired, contributes to development of cancer. These genes are generally called target genes. This is a rather simplistic definition; however, it has been controversial to agree on the criteria to define a real target gene (Woerner et al., 2001, 2003; Duval & Hamelin, 2002; Perucho, 2003). Due to the general lack of functional studies proving the true involvement of target genes in tumour development, these genes are generally classified as such based on a high mutation frequency. One major problem is the establishment of a valid cut-off value for the mutation frequency to separate real target genes from passengers or bystanders (those having the background mutations expected in an MMR-deficient context but not related with the progression to cancer). In 2002, Duval et al. proposed a cut-off frequency value of 10-15% and this has been used by other groups (Vilkki et al., 2002). A number of target genes have been identified in MMR-deficient tumours, and these are thought to be key players in MSI-H tumorigenesis. Mutations were mostly searched for in MSI-H colorectal tumours, although now many of the genes have been screened in endometrial and gastric tumours as well (Duval et al., 1999; Schwartz et al., 1999; Duval et al., 2001; Vilkki et al., 2002; Royrvik et al., 2007). From several studies it became clear that there are target genes, such as BAX, commonly involved in MSI-H tumours of diverse origin, whereas others show considerable qualitative and quantitative differences between different tumour types, probably due to tissue-specific selection (Myeroff et al., 1995; Duval et al., 1999; Gurin et al., 1999; Schwartz et al., 1999; Semba et al., 2000; Duval et al., 2002a). It is also clear from these studies that more important genes remain to be found, especially in endometrial cancer. These tumours are subjected to less screening than colorectal ones, and only a few genes with a high mutation frequency have been found in them. 15

17 1.6. Endometrium Histology and functional changes The uterus is a hollow muscular, pear-shaped organ weighing gram in a nonpregnant woman. The size of the uterus is highly variable as is demonstrated during pregnancy. There are two parts to the uterus: the main body, known as the corpus, and the lower part, which opens into the vagina, called the cervix. The wall of the uterus consists of three layers: different types of mucosa at the inner side; a thick muscular, highly vascularised part; and a thin layer of serosa covering the intraperitoneal part of the corpus. The cervical canal is covered by a single layer of clindrical mucus secreting cells which extends into the underliying myocervix forming endocervical crypts. The inner lining of the corpus is called endometrium. The endometrium consists of a supportive stroma and an epithelial component the endometrial glands. The thickness and differentiation of the functional layer of the endometrium is highly regulated by the hormonal changes occurring during the menstrual cycle. The endometrial mucosa can be sub-divided in two areas related to those changes: a functional layer, adjacent to the cavity of the uterus, that is sloughed during menstruation and built up afterwards, and a basal inert layer, adjacent to the myometrium, that is not shed during the menstrual cycle and that functions as a regenerative zone for the functional layer. After menopause, in a low estrogenic situation, the endometrium consists of the basal layer only Endometrial cancer Aetiology Endometrial cancer (EC) is one of the most common types of gynaecological cancer in women worldwide. The highest incidence is found in North America, although the highest levels of mortality are in Eastern Europe. The incidence of endometrial cancer increases after menopause; approximately 75% of cases are diagnosed in postmenopausal women (Cancer Research UK website). The major risk factor for endometrial cancer is the high, unopposed exposure to oestrogens (Sherman, 2000; Amant et al., 2005). Therefore, conditions 16

18 increasing the oestrogen levels are considered to increase the disease risk. These include for instance: early menarche, late menopause, nulliparity or low parity, and hormone replacement therapy (HRT) with exogenous oestrogen but without progesterone. Long-term use of tamoxifen, a drug used to treat breast cancer, also increases the risk for endometrial cancer (Polin & Ascher, 2008). Other risk factors for the disease include: a high-fat diet, obesity, hypertension, diabetes, age (more common after age of 50), personal history of breast, colorectal, or ovarian cancer and a family history of endometrial cancer or colon cancer (Lynch syndrome). Endometrial cancer risk has also been suggested to be increased in Cowden syndrome, caused by germline PTEN mutations. The use of oral combined contraceptives, on the other hand, is reported to reduce the risk of EC. Histopathological and molecular types of endometrial carcinomas Endometrial carcinomas are usually divided into two major groups that have different clinical and histological characteristics, as well as molecular differences (Emons et al., 2000; Lax et al., 2004; Ryan et al., 2005). Type I or oestrogen-dependent endometrioid carcinomas (EEC) Representing 80% of sporadic cases, this is the group of oestrogen-related tumours. They occur in both pre- and post-menopausal women, and their architectural features resemble endometrial glands. Tumours of this type are usually well differentiated (low grade) and confined to the uterus (low stage) and therefore the patient generally has a good prognosis. They are frequently preceded by endometrial hyperplasia. Type II or non-oestrogen-dependent ECs The tumours belonging to this group are unrelated to oestrogenic stimulation, and mainly occur in post-menopausal women. They display a more aggressive behaviour and poor prognosis. Frequently, by the time of diagnosis, the tumour has already spread outside the uterus. They are not usually preceded by hyperplasia, but originate from an atrophic endometrium instead. They are high-grade tumours with serous or clear-cell morphology. 17

19 In addition to the histopathological differences referred to above, there are also genetic differences between these two categories of endometrial carcinomas (Doll et al., 2007). In type I EC we can basically find mutations in PTEN (35-50%), K-Ras (15-30%) and β-catenin (25-40%) genes, and MMR defects, detected by high levels of MSI (20-40%). These characteristics are rarely seen in type II EC, which are characterized by high mutation frequencies on P53 gene (90%) and alterations on HER2/NEU and CDH1. The main type of genetic instability in this group is chromosomal instability (CIN), being aneuploidy and loss-of-heterozygosity (LOH) typical of EC type II. MSI (MIN) is extremely rare in these tumours (Emons et al., 2000; Lax et al., 2004; Ryan et al., 2005). Figure 1 shows the progression model of endometrial cancer proposed by Ryan et al. (2005). Hypermethylation PTEN MLH1 MSI Mutations (e.g. KRAS, BAX, MSH2) Endometrial hyperplasia Type I endometrioid adenocarcinoma Normal p53 Endometrial intraepithelial carcinoma Type II serous adenocarcinoma HER2/NEU P53, LOH, HER2/NEU Figure 1. Progression model of endometrial cancers type I and type II progression adapted from Ryan et al. (2005). 18

20 AIM AND OUTLINE OF THE THESIS The main focus of this thesis was to understand the development of tumours that follow the MSI pathway. The study covered Lynch syndrome-associated tumours, with particular emphasis on colorectal and endometrial carcinomas and their sporadic counterparts. Chapter 1 reviews the general background to Lynch syndrome, microsatellite instability, and the hereditary and sporadic cancers associated with this pathway. Chapter 2 addresses how instability evolves along the adenoma-carcinoma sequence of colorectal cancer, and whether we are able to establish different profiles of MSI for hereditary and sporadic adenomas and carcinomas. Knowledge on this process might be helpful in understanding tumour development and in identifying Lynch syndrome patients in an easier and more specific way. In chapter 3, we report on our comparison of the frequencies of instability of different types of microsatellites between colorectal and endometrial MSI-H tumours. In addition, we analyze features such as type (deletions/insertions) and size of microsatellite mutation for possible correlations with tissue specificity. Chapter 4 describes our hunt for new genes involved in MSI-H endometrial tumours. It addresses the instability of mononucleotide repeats occurring in coding sequences. The aim of this work was to identify novel target genes that could explain MSI-H endometrial tumour development, and to unravel molecular pathways related to this type of cancer. We further wanted to determine whether the identified genes were also involved in colorectal and gastric tumours, and we speculate about the functional role of the proteins that are encoded by the genes we found mutated. In chapter 5 we review and try to clarify the mechanisms linking hormones to cancer, and in particular how hormones can play a role in MSI tumorigenesis. Finally, in chapter 6, the major findings of this project are discussed, conclusions are drawn and future perspectives are formulated. 19

21 REFERENCES Aaltonen LA, Peltomäki P, Leach FS, et al. Clues to the pathogenesis of familial colorectal cancer. Science 1993;260(5109): Aaltonen LA, Peltomäki P, Mecklin JP, et al. Replication errors in benign and malignant tumors from hereditary nonpolyposis colorectal cancer patients. Cancer Res 1994;54(7): Amant F, Moerman P, Neven P, et al. Endometrial cancer. Lancet 2005;366: Berends MJ, Wu Y, Sijmons RH, et al. Molecular and clinical characteristics of MSH6 variants: an analysis of 25 index carriers of a germline variant. Am J Hum Genet 2002;70(1): Boland CR, Thibodeau SN, Hamilton SR, et al. A National Cancer Institute Workshop on Microsatellite Instability for Cancer Detection and Familial Predisposition: development of international criteria for the determination of microsatellite instability in colorectal cancer. Cancer Res 1998;58(22): Bronner CE, Baker SM, Morrison PT, et al. Mutation in the DNA mismatch repair gene homologue hmlh1 is associated with hereditary non-polyposis colon cancer. Nature 1994;368(6468): Buhard O, Suraweera N, Lectard A, et al. Quasimonomorphic mononucleotide repeats for high-level microsatellite instability analysis. Dis Markers 2004;20(4-5): de la Chapelle A. Testing tumors for microsatellite instability. Eur J Hum Genet 1999;7(4): de Wind N, Dekker M, van Rossum A, et al. Mouse models for hereditary nonpolyposis colorectal cancer. Cancer Res 1998;58(2): Doll A, Abal M, Rigau M, et al. Novel molecular profiles of endometrial cancer-new light through old windows. J Steroid Biochem Mol Biol 2008;108(3-5): Duval A, Gayet J, Zhou XP, et al. Frequent frameshift mutations of the TCF-4 gene in colorectal cancers with microsatellite instability. Cancer Res 1999;59(17): Duval A, Rolland S, Compoint A, et al. Evolution of instability at coding and non-coding repeat sequences in human MSI-H colorectal cancers. Hum Mol Genet 2001;10(5): Duval A, Reperant M, Compoint A, et al. Target gene mutation profile differs between gastrointestinal and endometrial tumors with mismatch repair deficiency. Cancer Res 2002(a);62(6):

22 Duval A, Hamelin R. Mutations at coding repeat sequences in mismatch repair-deficient human cancers: toward a new concept of target genes for instability. Cancer Res 2002(b);62(9): Emons G, Fleckenstein G, Hinney B, et al. Hormonal interactions in endometrial cancer. Endocr Relat Cancer 2000;7(4): Fearon ER, Vogelstein B. A genetic model for colorectal tumorigenesis. Cell 1990;61(5): Fishel R, Lescoe MK, Rao MR, et al. The human mutator gene homolog MSH2 and its association with hereditary nonpolyposis colon cancer. Cell 1993;75(5): Erratum in: Cell. 1994;77(1):167. Fujita S, Moriya Y, Sugihara K, et al. Prognosis of hereditary nonpolyposis colorectal cancer (HNPCC) and the role of Japanese criteria for HNPCC. Jpn J Clin Oncol 1996;26(5): Giuffrè G, Müller A, Brodegger T, et al. Microsatellite analysis of hereditary nonpolyposis colorectal cancer-associated colorectal adenomas by laser-assisted microdissection: correlation with mismatch repair protein expression provides new insights in early steps of tumorigenesis. J Mol Diagn 2005;7(2): Greenspan EJ, Cyr JL, Pleau DC, et al. Microsatellite instability in aberrant crypt foci from patients without concurrent colon cancer. Carcinogenesis 2007;28(4): Gurin CC, Federici MG, Kang L, et al. Causes and consequences of microsatellite instability in endometrial carcinoma. Cancer Res 1999;59(2): Hatch SB, Lightfoot HM Jr, Garwacki CP, et al. Microsatellite instability testing in colorectal carcinoma: choice of markers affects sensitivity of detection of mismatch repair-deficient tumors. Clin Cancer Res 2005;11(6): Hemminki A, Peltomäki P, Mecklin JP, et al. Loss of the wild type MLH1 gene is a feature of hereditary nonpolyposis colorectal cancer. Nat Genet 1994;8(4): Hendriks YM, Jagmohan-Changur S, van der Klift HM, et al. Heterozygous mutations in PMS2 cause hereditary nonpolyposis colorectal carcinoma (Lynch syndrome). Gastroenterology 2006;130(2): Hienonen T, Laiho P, Salovaara R, et al. Little evidence for involvement of MLH3 in colorectal cancer predisposition. Int J Cancer 2003;106(2): Hoang JM, Cottu PH, Thuille B, et al. BAT-26, an indicator of the replication error phenotype in colorectal cancers and cell lines. Cancer Res 1997;57(2):

23 Ionov Y, Peinado MA, Malkhosyan S, et al. Ubiquitous somatic mutations in simple repeated sequences reveal a new mechanism for colonic carcinogenesis. Nature 1993;363(6429): Iino H, Jass JR, Simms LA, et al. DNA microsatellite instability in hyperplastic polyps, serrated adenomas, and mixed polyps: a mild mutator pathway for colorectal cancer? J Clin Pathol 1999;52(1):5-9. Iino H, Simms L, Young J, et al. DNA microsatellite instability and mismatch repair protein loss in adenomas presenting in hereditary non-polyposis colorectal cancer. Gut 2000;47(1): Jagmohan-Changur S, Poikonen T, Vilkki S, et al. EXO1 variants occur commonly in normal population: evidence against a role in hereditary nonpolyposis colorectal cancer. Cancer Res 2003;63(1): Kovacs ME, Papp J, Szentirmay Z, et al. Deletions removing the last exon of TACSTD1 constitute a distinct class of mutations predisposing to Lynch syndrome. Hum Mutat. 2009;30(2): Lax SF. Molecular genetic pathways in various types of endometrial carcinoma: from a phenotypical to a molecular-based classification. Virchows Arch 2004;444(3): Leach FS, Nicolaides NC, Papadopoulos N, et al. Mutations of a muts homolog in hereditary nonpolyposis colorectal cancer. Cell 1993;75(6): Ligtenberg MJ, Kuiper RP, Chan TL, et al. Heritable somatic methylation and inactivation of MSH2 in families with Lynch syndrome due to deletion of the 3' exons of TACSTD1. Nat Genet 2009;41(1): Lindblom A, Tannergård P, Werelius B, et al. Genetic mapping of a second locus predisposing to hereditary non-polyposis colon cancer. Nat Genet 1993;5(3): Loukola A, Salovaara R, Kristo P, et al. Microsatellite instability in adenomas as a marker for hereditary nonpolyposis colorectal cancer. Am J Pathol 1999;155(6): Lu SL, Akiyama Y, Nagasaki H, et al. Loss or somatic mutations of hmsh2 occur in hereditary nonpolyposis colorectal cancers with hmsh2 germline mutations. Jpn J Cancer Res 1996;87(3): Meyer C, Brieger A, Plotz G, et al. An interstitial deletion at 3p21.3 results in the genetic fusion of MLH1 and ITGA9 in a Lynch syndrome family. Clin Cancer Res. 2009;15(3): Myeroff LL, Parsons R, Kim SJ, et al. A transforming growth factor beta receptor type II gene mutation common in colon and gastric but rare in endometrial cancers with microsatellite instability. Cancer Res 1995;55(23):

24 Ou J, Niessen RC, Vonk J, et al. A database to support the interpretation of human mismatch repair gene variants. Hum Mutat 2008 ;29(11): Papadopoulos N, Nicolaides NC, Wei YF, et al. Mutation of a mutl homolog in hereditary colon cancer. Science 1994;263(5153): Peltomäki P, Lothe RA, Aaltonen LA, et al. Microsatellite instability is associated with tumors that characterize the hereditary non-polyposis colorectal carcinoma syndrome. Cancer Res 1993(a);53(24): Peltomäki P, Aaltonen LA, Sistonen P, et al. Genetic mapping of a locus predisposing to human colorectal cancer. Science 1993(b);260(5109): Peltomäki P, Vasen H. Mutations associated with HNPCC predisposition -- Update of ICG- HNPCC/INSiGHT mutation database. Dis Markers 2004;20(4-5): Perucho M. Tumors with microsatellite instability: many mutations, targets and paradoxes. Oncogene 2003;22(15): Polin SA, Ascher SM. The effect of tamoxifen on the genital tract. Cancer Imaging 2008;8: Røyrvik EC, Ahlquist T, Rognes T, et al. Slip slidin' away: a duodecennial review of targeted genes in mismatch repair deficient colorectal cancer. Crit Rev Oncog 2007;13(3): Ryan AJ, Susil B, Jobling TW, et al. Endometrial cancer. Cell Tissue Res 2005 ;322(1): Schwartz S Jr, Yamamoto H, Navarro M, et al. Frameshift mutations at mononucleotide repeats in caspase-5 and other target genes in endometrial and gastrointestinal cancer of the microsatellite mutator phenotype. Cancer Res 1999;59(12): Semba S, Ouyang H, Han SY, et al. Analysis of the candidate target genes for mutation in microsatellite instability-positive cancers of the colorectum, stomach, and endometrium. Int J Oncol 2000;16(4): Sherman ME. Theories of endometrial carcinogenesis: a multidisciplinary approach. Mod Pathol 2000;13: Shibata D, Peinado MA, Ionov Y, et al. Genomic instability in repeated sequences is an early somatic event in colorectal tumorigenesis that persists after transformation. Nat Genet 1994;6(3):

25 Sugai T, Takahashi H, Habano W, et al. Analysis of genetic alterations, classified according to their DNA ploidy pattern, in the progression of colorectal adenomas and early colorectal carcinomas. J Pathol 2003;200(2): Thibodeau SN, Bren G, Schaid D. Microsatellite instability in cancer of the proximal colon. Science 1993;260(5109): Umar A, Boland CR, Terdiman JP, et al. Revised Bethesda Guidelines for hereditary nonpolyposis colorectal cancer (Lynch syndrome) and microsatellite instability. J Natl Cancer Inst 2004;96(4): Vasen HF, Mecklin JP, Khan PM, et al. The International Collaborative Group on Hereditary Non- Polyposis Colorectal Cancer (ICG-HNPCC). Dis Colon Rectum 1991;34(5): Vasen HF, Watson P, Mecklin JP, et al. New clinical criteria for hereditary nonpolyposis colorectal cancer (HNPCC, Lynch syndrome) proposed by the International Collaborative group on HNPCC. Gastroenterology 1999;116(6): Vilkki S, Launonen V, Karhu A, et al. Screening for microsatellite instability target genes in colorectal cancers. J Med Genet 2002;39(11): Woerner SM, Gebert J, Yuan YP, et al. Systematic identification of genes with coding microsatellites mutated in DNA mismatch repair-deficient cancer cells. Int J Cancer 2001;93(1):12-9. Woerner SM, Benner A, Sutter C, et al. Pathogenesis of DNA repair-deficient cancers: a statistical meta-analysis of putative Real Common Target genes. Oncogene 2003;22(15): Wong YF, Cheung TH, Lo KW, et al. Detection of microsatellite instability in endometrial cancer: advantages of a panel of five mononucleotide repeats over the National Cancer Institute panel of markers. Carcinogenesis 2006;27(5): Wu Y, Berends MJ, Post JG, et al. (a) Germline mutations of EXO1 gene in patients with hereditary nonpolyposis colorectal cancer (HNPCC) and atypical HNPCC forms. Gastroenterology 2001;120(7): Wu Y, Berends MJ, Sijmons RH, et al. (b) A role for MLH3 in hereditary nonpolyposis colorectal cancer. Nat Genet 2001;29(2): Young J, Leggett B, Gustafson C, et al. Genomic instability occurs in colorectal carcinomas but not in adenomas. Hum Mutat 1993;2(5): Websites: (January 2009) 24

26 CHAPTER 2 Mononucleotide Precedes Dinucleotide Instability during Colorectal Tumour Development in Lynch Syndrome Patients Ana M. Ferreira 1, 4, Helga Westers 1, Sónia Sousa 4, Ying Wu 1, Renée C. Niessen 1, Maran Olderode-Berends 1, Tineke van der Sluis 2, Peter T.W. Reuvekamp 1, Raquel Seruca 4, Jan H. Kleibeuker 3, Harry Hollema 2, Rolf H. Sijmons 1, Robert M.W. Hofstra 1 Departments of 1 Genetics, 2 Pathology, 3 Gastroenterology, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands. 4 Institute of Molecular Pathology and Immunology of the University of Porto, Porto, Portugal. Under review 25

27 ABSTRACT A progressive accumulation of genetic alterations underlies the adenomacarcinoma sequence of colorectal cancer. This accumulation of mutations is driven by genetic instability, of which there are different types. Microsatellite instability (MSI) is the predominant type present in the tumours of Lynch syndrome patients and in a subset of sporadic tumours. It is generally accepted that MSI can be found in the early stages of tumour progression, such as adenomas; however, the frequencies reported vary widely among studies. Moreover, data on the qualitative differences between adenomas and carcinomas, or between tumours of hereditary and sporadic origin, are scarce. We compared MSI in colorectal adenoma- and colorectal carcinoma samples in order to identify possible differences along the adenomacarcinoma sequence. We compared germline mismatch repair (MMR) gene mutation carriers and non-carriers, to address possible differences of instability patterns between Lynch syndrome patients and patients with sporadic tumours. We found a comparable relative frequency of mono- and dinucleotide instability in sporadic colorectal adenomas and carcinomas, dinucleotide instability being observed most frequently in these sporadic tumours. In MMR gene truncating mutation carriers, the profile is different: colorectal adenomas show predominantly mononucleotide instability and also in colorectal carcinomas more mononucleotide than dinucleotide instability was detected. We conclude that MSI profiles differ between sporadic and Lynch syndrome tumours, and that mononucleotide marker instability precedes dinucleotide marker instability during colorectal tumour development in Lynch syndrome patients. As mononucleotide MSI proves to be highly sensitive for detecting mutation carriers, we propose the use of mononucleotide markers for the identification of possible Lynch syndrome patients. 26

28 INTRODUCTION The adenoma-carcinoma sequence of colorectal cancer is the best-known model of cancer development. Colorectal carcinomas arise through a multi-step process, starting from early adenomas to high-grade dysplastic adenomas to carcinomas. This process of cancer development is basically caused by the progressive accumulation of genetic alterations in genes involved in cell growth, differentiation, proliferation, and apoptosis (Fearon et al., 1990). This accumulation of genetic alterations is thought to be driven by genetic instability, of which several distinct forms can be distinguished. Those that are best-described are chromosomal instability (CIN) and microsatellite instability (MIN) (Komarova et al., 2002). CIN is characterized by widespread chromosomal abnormalities, such as aneuploidy and frequent loss of-heterozygosity (LOH). MIN is characterized by the accumulation of mutations in short repetitive sequences, known as microsatellites. The underlying mechanism of microsatellite instability is a defect in the DNA mismatch repair (MMR) pathway. The MMR pathway corrects replication errors, such as mispaired nucleotides, as well as small insertions and deletions resulting from slippage of the polymerases during replication of microsatellites. MMR deficiency therefore leads to an accumulation of mutations in microsatellites and it is recognized by decreased or (less often) increased microsatellite lengths. This phenomenon is referred to as microsatellite instability (MSI) (Ionov et al., 1993); it was first described in Lynch syndrome patients (Aaltonen et al., 1993; Ionov et al., 1993; Peltomaki et al., 1993; Thibodeau et al., 1993) and detected in over 90% of tumours in those patients. It is also found in a large proportion (15-25%) of sporadic colorectal (CRC) and endometrial (EC) carcinomas (Boland et al., 1998). It was shown that the functional inactivation of the MMR pathway by a germline mutation in one of the MMR genes and, in addition, somatic inactivation of the corresponding wild-type allele causes Lynch syndrome (Fishel et al., 1993; Leach et al., 1993; Bronner et al. 1994; Papadopoulos et al., 1994). The occurrence of MSI is not only reported in colorectal carcinomas but also in colorectal adenomas (AD) although generally in lower frequencies, in approximately 1-2% of sporadic AD (Young et al., 1993; Iino et al., 1999; Loukola 27

29 et al., 1999; Sugai et al., 2003), and in 10-90% of Lynch syndrome-associated AD (Aaltonen et al., 1994; Iino et al., 2000; Giuffrè et al., 2005). Previous studies comparing patterns of MSI in different tumour types and stages suggest that tumours with different tissue origin or in different stages of tumorigenesis show different levels of instability (Furlan et al., 2002; Kuismanen et al., 2002). However, these studies generally refer to quantitative differences in (mononucleotide) instability between tumour types. Data on qualitative differences are scarce. In this study we have analyzed mononucleotide and dinucleotide markers to define specific qualitative profiles of MSI in colorectal adenomas and carcinomas of Lynch syndrome patients and of sporadic cases. MATERIALS AND METHODS Samples The participants in this study came from two sources: either known MMR gene mutation carriers under surveillance at our Hospital, or participants from previous studies in our group who had been diagnosed with CRC under the age of 50 years or had two or more Lynch syndrome-related cancers, including at least one CRC, irrespective of age and family history. This study was approved by the local medical ethical committee. All patients had been analyzed for germline mutations in the MLH1, MSH2, and MSH6 genes (for details see Rijcken et al., 2002; Niessen et al., 2006). In total we included paraffin-embedded tumour tissue sections and the respective normal tissue/blood from 67 colon adenomas (AD) and 213 colon carcinomas (CRC). Twenty-two of the AD were classified as low-grade dysplasia and 20 as high-grade dysplasia; information on the grade of dysplasia was not available for the other 25 AD. Thirty-two out of the 67 AD and 20 out of the 213 CRC were from patients who carried a pathogenic germline mutation in one of the three MMR genes (MLH1, MSH2, and MSH6). These patients are referred to as truncating mutation carriers. This group includes 7 pairs of adenoma and carcinoma samples of the same patient. In the CRC group, in addition to the 20 pathogenic mutation carriers, 12 patients carry missense mutations, all of unknown pathological significance. 28

30 MSI Analysis MSI analysis was performed using a panel of three mononucleotide markers (BAT25, BAT26, BAT40) and three dinucleotide markers (D2S123, D5S346, D17S250) as described previously (Berends et al., 2002). For the AD and the AD/CRC pairs from the same patient, 3 additional mononucleotide markers were analyzed (NR27, NR21, NR24). DNA was extracted from formalin-fixed paraffinembedded tumour sections and compared with DNA isolated from normal tissue from paraffin-embedded sections (when available) or peripheral blood lymphocytes from the same patient, as described previously (Berends et al., 2002). The samples were classified as MSI-High when more than 30% of the markers analyzed were unstable. Statistical Analysis The statistical analyses were performed using the 2 test or Fisher s exact test. P values <0.05 were considered to be significant. RESULTS Frequencies of instability truncating mutation carriers show the highest MSI frequency Frequencies of MSI-H, MSI-L and MSS samples distributed by tumour type and presence/absence of MMR mutations are shown in Table 1. As expected, the samples from the truncating mutation carriers show higher frequencies of MSI-H and lower frequencies of MSS/MSI-L than the samples from the non-carriers. Among the AD a significantly smaller proportion exhibit MSI-H (6% of the noncarriers; 56% of the truncating mutation carriers) compared to the CRC (36% of the non-carriers; 90% of the truncating mutation carriers) (P< 0.05). The missense mutation carriers show instability frequencies very similar to the non-carriers. 29

31 Table 1. Distribution of the samples by presence or absence of germline MMR mutations, tumour tissue type (adenoma/carcinoma) and MSI status. Non-carriers Mutation carriers Total (N) Truncating Missense AD CRC AD CRC CRC (35) (181) (32) (20) (12) (280) MSI-H 6% (2) 36% (65) 56% (18) 90% (18) 33% (4) 107 MSI-L 46% (16) 28% (51) 9% ( 3) 10% ( 2) 33% (4) 76 MSS 49% (17) 36% (65) 34% (11) 0% (0) 33% (4) 97 AD colon adenomas; CRC colorectal cancer; MSI-H microsatellite instability high; MSI-L microsatellite instability low; MSS microsatellite stable. Figure 1. Frequencies of instable mono- and dinucleotide markers in MSI-L and MSI-H samples from CA and CRC. * significant. Higher mono- and dinucleotide instability in MSI-H tumours compared to MSI-L tumours Figure 1 shows the observed frequencies of unstable mono- and dinucleotide markers in AD and CRC, in both MSI-L and MSI-H cases, without stratification of samples into mutation carriers or non-carriers. For the carcinomas, both mono- and 30

32 dinucleotide markers are more unstable in the MSI-H group than in the MSI-L group, and instability at the dinucleotide level was significantly more frequently observed than mononucleotide instability. In AD, in MSI-L samples dinucleotide instability was also significantly higher than mononucleotide (36% vs. 5%, p<0.05), however in the MSI-H group mononucleotide was markedly increased and significantly more frequent than dinucleotide instability (85% vs. 43%, p value <0.05). Repeat instability depends on MMR mutations Stratifying the samples for the presence/absence of germline mutations in one of the MMR genes MLH1, MSH2 and MSH6, the profiles of instability obtained were in the CRC set quite different from those described above (compare Figure 1 and 2). Figure 2. Frequencies of instable mono- and dinucleotide markers in MSI-L and MSI-H of non-carriers, missense mutation carriers, and truncating mutation carriers, for colorectal carcinoma samples. NS not significant; * significant. When comparing truncating mutation carriers and non-carriers in the MSI-H CRC group (Figure 2) several significant differences were observed: the noncarriers show significantly more dinucleotide instability (66%) than mononucleotide 31

33 instability (49%); truncating mutation carriers had more mononucleotide instability than non-carriers (78% vs. 49%, p<0.05), but similar frequencies of dinucleotide instability (67% vs. 66%); CRC MSI-L samples from non-carriers showed significantly more dinucleotide instability, while the MSI-L samples of truncating mutation carriers had similar frequencies of mono- and dinucleotide instability. The MSI-H cancers of carriers of MMR missense mutations retained preferential dinucleotide instability over mononucleotide instability, resembling more the pattern seen for the non-carriers, but the sample size was too small to make a clear statement. In AD, the following observations were made: dinucleotide instability is seen in both non-carriers and mutation carriers groups at similar frequencies, whereas mononucleotide instability was by far more frequently present in mutation carriers (p<0.05) (Table 2). When comparing low-grade (LD) with high-grade dysplastic (HD) adenomas from mutations carriers, we detected MSI-H in 38% (5/13) of the LD adenomas and in 67% (6/9) of the HD adenomas. In LD samples of mutation carriers, instability was found only in adenomas from MLH1 mutation carriers, whereas in HD samples, instability was observed in adenomas from all three types of mutation carriers (MLH1, MSH2, MSH6). Moreover, LD adenomas from mutation carriers showed mainly mononucleotide instability, whereas HD adenomas showed both mono- and dinucleotide instability (Table 2). Pairs adenoma and carcinoma from the same patient We also had 7 patients from whom we could obtain both an AD and a CRC. We observed both mono- and dinucleotide instability at high frequency. Mononucleotide instability was seen more frequent than dinucleotide instability (but not significantly different) (Table 3). 32

34 Table 2. Instability results for colon adenomas. A) Mononucleotide markers Dinucleotide markers Patient ID Grade MMR mutation MSI Status NR27 NR21 NR24 BAT25 BAT26 BAT40 D2S123 D5S346 D17S250 Y313 ND Y239 LD Y72 ND Y65 LD Y71 ND Y190 HD Y264 ND Y200 HD Y81 HD Y210 HD Y234 HD Y89 ND Y123.1 LD Y192 LD Y123.2 LD Y79 HD Y301 ND Y176-2 ND Y13 LD Y39 ND Y56 LD Y167 LD Y170 HD Y225 HD Y257 ND Y267 LD Y202 HD Y223 HD Y268 HD Y309 ND Y311 ND Y317 ND Y327 ND Y331 ND Y294 ND NON-CARRIERS MSI- H MSI-L MSS A) adenomas from non-carriers; B) adenomas from mutation carriers LD low-grade dysplasia; HD high-grade dysplasia; ND no data available. Results in black mean unstable; grey mean stable; white no result available. 33

35 B) Mononucleotide markers Dinucleotide markers Patient ID Grade MMR mutation MSI Status NR27 NR21 NR24 BAT25 BAT26 BAT40 D2S123 D5S346 D17S250 2T LD 8T LD 9T LD 17T HD 13T LD 16T HD 3T HD 18T ND 10T LD MLH1 1 ND MSI-H 5T LD 29T LD MSI-L 21T LD 25T ND 5 ND 6 ND MSS 14T HD 6T HD Y21 ND 2 ND 3 ND 4 ND MSH2 7 ND MSI-H Y112 HD 4T LD 20T LD MSS Y241 HD MSI-H 26T HD MSI-L 7T LD 12T LD Y86 HD 1T LD MSH6 MSS 34

36 Table 3. MSI results for adenoma and carcinoma of the same patient. Mononucleotide markers Dinucleotide markers NR27 NR21 NR24 BAT25 BAT26 D2S123 D5S346 D17S250 1 AD MLH1 CRC MLH1 2 AD MSH2 CRC MSH2 3 AD MSH2 CRC MSH2 4 AD MSH2 CRC MSH2 5 AD MLH1 CRC MLH1 6 AD MLH1 CRC MLH1 7 AD MSH2 CRC MSH2 AD colorectal adenoma; CRC colorectal carcinoma; black means unstable; grey means stable; white no result available. DISCUSSION In the present study, we compared MSI in colorectal adenomas (AD) and colorectal carcinomas (CRC) in order to identify possible differences along the adenomacarcinoma sequence. We compared germline MMR mutation carriers and noncarriers, to address possible differences of instability patterns between Lynch syndrome patients and patients with sporadic tumours. The CA in our study showed a significantly lower proportion of MSI-H cases than CRC, both in non-carriers (6% vs. 36%, p<0.05) and in truncating mutation carriers (56% vs. 90%, p<0.05) (no adenomas were available from missense mutation carriers). Our study confirms the reported difference in MSI frequencies during the transition from adenoma to carcinoma (Shibata et al., 1994; Grady et al., 1998; Loukola et al., 1999; Iino et al., 2000; Sugai et al., 2003; Giuffrè et al., 2005). We further analyzed how instability is distributed amongst mononucleotide versus dinucleotide markers, in both MSI-L and MSI-H groups of AD and CRC 35

37 samples. When no distinction is made between mutation carriers and non-carriers, the distribution of instability is similar between MSI-L and MSI-H samples, in CRC: dinucleotide markers are more frequently unstable than mononucleotide markers. In AD, unstable dinucleotide markers are also more frequent than mononucleotide markers in MSI-L tumours, but in MSI-H the opposite situation is found. The results are, however, different when we split the CRC samples into different groups: truncating mutation carriers, missense mutation carriers and noncarriers. We observed that the high frequency of dinucleotide instability in MSI-H tumours is due to the inclusion of non-carriers and missense mutation carriers, and that the mononucleotide instability is mainly due to the inclusion of mutation carriers. In AD from non-carriers mainly the dinucleotide markers were unstable; a very low frequency of mononucleotide instability was seen, resembling the MSI-L CRC of non-carriers. In the AD of truncating mutation carriers, mononucleotide instability was generally predominant. As more mono- to dinucleotide instability is observed, our data suggest that mononucleotide instability is a very early event in the carcinogenic process of tumours having mismatch repair mutations, and that mononucleotide instability precedes that of dinucleotide repeats. We also included 7 adenoma/carcinoma pair from the same patient. Again we observe a difference between mononucleotide instability and dinucleotide instability however this was not significant due to the small number of cases analysed. This idea is further supported by our results in low- and high-grade dysplastic adenomas from mutation carriers. Low-grade adenomas have less MSI and mainly mononucleotide instability, whereas 67% of the high-grade adenomas were MSI-H, with the instability found in both mono- and dinucleotide markers. As far as the observed preference of dinucleotide instability in early lesions (AD) of non-mutation carriers is concerned, we hypothesize that the dinucleotide instability in these cases represents a kind of background as seen in MSI-L tumours, which is not a sign of an underlying MMR deficiency. Part of the dinucleotide instability seen in MSI-H CRC of carriers and non-carriers might therefore likely also occur independently of MMR deficiency. In the case of Lynch syndrome tumours, with proven MMR deficiency, mononucleotide instability can be considered a true result of the underlying MMR deficiency. 36

38 Our hypothesis is in line with the finding of only dinucleotide instability in a study of MSI-L CRC using a large number of MSI markers (Laiho et al., 2002). The same authors suggest that all CRCs would display a MSI-L phenotype if a large number of markers is used (Laiho et al., 2002). A possible explanation for our findings might be that the normal MMR system more easily corrects mismatches in mononucleotides than dinucleotides; this would mean that part of the dinucleotide instability is background noise in tumours. However, to our knowledge, such a difference in repair outcome has not been demonstrated. Interestingly, the number of mononucleotide repeats in the entire genome is higher than the number of dinucleotide ones (Borstnik et al., 2004), and as more instability is seen in these less frequent dinucleotide repeats in MMRproficient tumors, this suggests either a higher vulnerability of dinucleotide repeats to the occurrence of mismatches and/or a lower capacity of the normal MMR system to repair them. It is important to keep in mind that, although the number of mononucleotide repeats in the genome is higher compared to the number of dinucleotide repeats, they are not always expected to be more unstable than dinucleotide repeats; the size and base composition of the repeat can have a strong influence on the degree of instability (Boyer et al., 2002). In addition, one can also speculate that, depending on the rate of replication of tumour cells, dinucleotides might be more prone to acquire mutations than mononucleotides. Another explanation why mononucleotide instability is seen earlier and more frequently in AD compared to dinucleotide instability in mutation carriers might be the fact that the instability seen in mononucleotide repeats is almost always due to deletions of a certain length, and this has consequences for the chance of detection. As new mutations in an MMR-deficient tumour happen often and in multiple cells, the tumours should be considered multi-clonal. The different clones will not obscure detection of mononucleotide instability because of their similar mutation size. For dinucleotide repeats this is different. The mutations in these repeats are often different and because of this, the different clones, with differently sized alleles might make it less easy to detect instability in these repeats. This, however, does not explain the finding of mostly dinucleotide instability in non mutations carriers. 37

39 Implications for diagnostics of HNPCC colorectal adenomas Our results suggest an advantage to using mononucleotide markers for identifying colorectal adenomas and carcinomas associated with Lynch syndrome, since we observed that the MSI-H adenomas from mutation carriers predominantly show mononucleotide instability. Moreover, assuming that mononucleotide instability precedes dinucleotide instability in adenomas of MMR-truncating mutation carriers, analyzing mononucleotide markers would make it possible to detect MSI in very early lesions. To our knowledge, our data are the first to show that the use of a panel of only mononucleotide markers, as previously recommended for the detection of MSI-H hereditary CRC (Buhard et al., 2004), should also be used for the identification of Lynch syndrome patients through the testing of colon adenomas. Another practical advantage of using a mononucleotide marker panel is the fact that DNA from corresponding normal tissue is not always necessary (de la Chapelle, 1999; Buhard et al., 2004). CONCLUSIONS We show that mononucleotide instability is a very early event in the development of MSI tumours with MMR truncating mutations and that in Lynch syndrome associated tumours mononucleotide instability precedes dinucleotide instability. We therefore recommend using mononucleotide markers to identify possible Lynch syndrome patients. ACKNOWLEDGEMENTS The authors wish to thank Dr. Hermien de Walle for assistance with the statistical analyses, Dr. Richard Hamelin for helpful comments, and Jackie Senior for editing the manuscript. This work was supported by Fundação para a Ciência e a Tecnologia, Portugal (SFRH/BD/18832/2004) and by the European Community (FP LIFESCIHEALTH-5, proposal no ). 38

40 REFERENCES Aaltonen LA, Peltomäki P, Leach FS, et al. Clues to the pathogenesis of familial colorectal cancer. Science 1993;260(5109): Aaltonen LA, Peltomäki P, Mecklin JP, et al. Replication errors in benign and malignant tumors from hereditary nonpolyposis colorectal cancer patients. Cancer Res 1994;54(7): Berends MJ, Wu Y, Sijmons RH, et al. Molecular and clinical characteristics of MSH6 variants: an analysis of 25 index carriers of a germline variant. Am J Hum Genet 2002;70(1): Boland CR, Thibodeau SN, Hamilton SR, et al. A National Cancer Institute Workshop on Microsatellite Instability for Cancer Detection and Familial Predisposition: development of international criteria for the determination of microsatellite instability in colorectal cancer. Cancer Res 1998;58(22): Borstnik B, Pumpernik D. Mutational dynamics of short tandem repeats in human genome. Europhys Lett 2004;65(2):290. Boyer JC, Yamada NA, Roques CN, et al. Sequence dependent instability of mononucleotide microsatellites in cultured mismatch repair proficient and deficient mammalian cells. Hum Mol Genet 2002;11(6): Bronner CE, Baker SM, Morrison PT, et al. Mutation in the DNA mismatch repair gene homologue hmlh1 is associated with hereditary non-polyposis colon cancer. Nature 1994;368(6468): Buhard O, Suraweera N, Lectard A, et al. Quasimonomorphic mononucleotide repeats for high-level microsatellite instability analysis. Dis Markers 2004;20(4-5): De la Chapelle A. Testing tumors for microsatellite instability. Eur J Hum Genet 1999;7(4): Fearon ER, Vogelstein B. A genetic model for colorectal tumorigenesis. Cell 1990;61(5): Fishel R, Lescoe MK, Rao MR, et al. The human mutator gene homolog MSH2 and its association with hereditary nonpolyposis colon cancer. Cell 1993;75(5): Erratum in: Cell 1994;77(1):167. Furlan D, Casati B, Cerutti R, et al. Genetic progression in sporadic endometrial and gastrointestinal cancers with high microsatellite instability. J Pathol 2002;197(5): Giuffrè G, Müller A, Brodegger T, et al. Microsatellite analysis of hereditary nonpolyposis colorectal cancer-associated colorectal adenomas by laser-assisted microdissection: correlation with mismatch 39

41 repair protein expression provides new insights in early steps of tumorigenesis. J Mol Diagn 2005;7(2): Grady WM, Rajput A, Myeroff L, et al. Mutation of the type II transforming growth factor-beta receptor is coincident with the transformation of human colon adenomas to malignant carcinomas. Cancer Res 1998;58(14): Iino H, Jass JR, Simms LA, et al. DNA microsatellite instability in hyperplastic polyps, serrated adenomas, and mixed polyps: a mild mutator pathway for colorectal cancer? J Clin Pathol 1999;52(1):5-9. Iino H, Simms L, Young J, et al. DNA microsatellite instability and mismatch repair protein loss in adenomas presenting in hereditary non-polyposis colorectal cancer. Gut 2000;47(1): Ionov Y, Peinado MA, Malkhosyan S, et al. Ubiquitous somatic mutations in simple repeated sequences reveal a new mechanism for colonic carcinogenesis. Nature 1993;363(6429): Komarova NL, Lengauer C, Vogelstein B, et al. Dynamics of genetic instability in sporadic and familial colorectal cancer. Cancer Biol Ther 2002;1(6): Kuismanen SA, Moisio AL, Schweizer P, et al. Endometrial and colorectal tumors from patients with hereditary nonpolyposis colon cancer display different patterns of microsatellite instability. Am J Pathol 2002;160(6): Laiho P, Launonen V, Lahermo P, et al. Low-level microsatellite instability in most colorectal carcinomas. Cancer Res 2002;62(4): Leach FS, Nicolaides NC, Papadopoulos N, et al. Mutations of a muts homolog in hereditary nonpolyposis colorectal cancer. Cell 1993;75(6): Loukola A, Salovaara R, Kristo P, et al. Microsatellite instability in adenomas as a marker for hereditary nonpolyposis colorectal cancer. Am J Pathol 1999;155(6): Niessen RC, Berends MJ, Wu Y, et al. Identification of mismatch repair gene mutations in young patients with colorectal cancer and in patients with multiple tumours associated with hereditary nonpolyposis colorectal cancer. Gut 2006;55(12): Papadopoulos N, Nicolaides NC, Wei YF, et al. Mutation of a mutl homolog in hereditary colon cancer. Science 1994;263(5153):

42 Peltomäki P, Lothe RA, Aaltonen LA, et al. Microsatellite instability is associated with tumors that characterize the hereditary non-polyposis colorectal carcinoma syndrome. Cancer Res 1993;53(24): Rijcken FE, Hollema H, Kleibeuker JH. Proximal adenomas in hereditary non-polyposis colorectal cancer are prone to rapid malignant transformation. Gut 2002;50(3): Shibata D, Peinado MA, Ionov Y, et al. Genomic instability in repeated sequences is an early somatic event in colorectal tumorigenesis that persists after transformation. Nat Genet 1994;6(3): Sugai T, Takahashi H, Habano W, et al. Analysis of genetic alterations, classified according to their DNA ploidy pattern, in the progression of colorectal adenomas and early colorectal carcinomas. J Pathol 2003;200(2): Thibodeau SN, Bren G, Schaid D. Microsatellite instability in cancer of the proximal colon. Science 1993;260(5109): Young J, Leggett B, Gustafson C, et al. Genomic instability occurs in colorectal carcinomas but not in adenomas. Hum Mutat 1993;2(5):

43 42

44 CHAPTER 3 Do Microsatellite Instability Profiles Really Differ Between Colorectal and Endometrial Tumours? Ana M. Ferreira 1, Helga Westers 1, Ying Wu 1, Renée C. Niessen 1, Maran Olderode- Berends 1, Tineke van der Sluis 2, Ate G. van der Zee 3, Harry Hollema 2, Jan H. Kleibeuker 4, Rolf H. Sijmons 1, Robert M.W. Hofstra 1 Departments of 1 Genetics, 2 Pathology, 3 Gynecology, and 4 Gastroenterology, University Medical Center Groningen, University of Groningen, Groningen, the Netherlands. Genes Chromosomes and Cancer, in press. 43

45 ABSTRACT Microsatellite instability (MSI) occurs in more than 90% of the tumours of Lynch syndrome patients, and in 15-25% of sporadic colorectal (CRC) and endometrial carcinomas (EC). Previous studies comparing EC and CRC using BAT markers showed that the frequency of unstable markers is lower in EC, and that the size of the mutations is smaller in EC. In the present study we analyzed the type (insertions/deletions), size and frequency of mutations occurring at three BAT and three dinucleotide markers in CRC and EC, in order to elucidate whether it is possible to establish different MSI profiles in carcinomas of different tissue origin. We show that mononucleotide markers nearly always become shorter whereas dinucleotide markers can become shorter or longer, in both CRC and EC. We therefore conclude that the type of mutation is a marker-dependent feature rather than tissue-dependent. However, we observed that the size of the deletions/insertions differs between CRC and EC, with EC having shorter alterations. The frequency of mono- and dinucleotide instability found in both tissues is comparable, with mononucleotide and dinucleotide markers being affected at similar rates. We conclude that it is not possible to define clearly different MSI profiles that could distinguish MSI-H in CRC and EC. We propose that the differences observed might indicate different durations of tumour development and/or differences in tissue turnover between colorectal and endometrial epithelium, rather than reflecting truly different MSI profiles. We therefore suggest that the same MSI tests can be used for both tumour types. 44

46 INTRODUCTION Microsatellite instability (MSI) is a form of genetic instability caused by defects in the DNA mismatch repair (MMR) pathway. MSI was first associated with Lynch syndrome patients in 1993 (Aaltonen et al., 1993; Ionov et al., 1993; Peltomäki et al., 1993; Thibodeau et al., 1993) and occurs in over 90% of the tumours of these patients. It is also found in a large proportion (15-25%) of sporadic colorectal (CRC) and endometrial (EC) carcinomas (Boland et al., 1998). The MMR pathway corrects mispaired nucleotides as well as small insertions and deletions, this last group resulting from slippage of the polymerases during replication of short DNA repeat sequences (microsatellites). When inactivating mutations occur within the MMR genes, such as MLH1, MSH2, and MSH6, the MMR pathway becomes deficient and an accumulation of insertions or deletions is observed in microsatellite sequences (Ionov et al., 1993). This phenomenon is referred to as microsatellite instability. MSI is therefore easily recognized by increased or decreased microsatellite lengths. An international consensus panel of five microsatellite markers was established to facilitate the detection and analysis of MSI and this panel is widely used (Boland et al., 1998). Samples that are unstable for two or more of these markers are considered MSI-high (MSI-H); samples unstable for one marker are called MSI-low (MSI-L); samples stable for all markers are called microsatellite stable (MSS). When a different number of markers are used, a sample is MSI-H when it shows instability in more than 30% of the markers used. Microsatellite mutations occur both at coding and non-coding repeats. Genes frequently found mutated in MSI-H tumours (also called target genes) play important roles in tumour development pathways and show mutations in their coding microsatellite sequences. The profile of target genes is thought to be different in CRC and EC, both in quantitative and qualitative ways (Duval et al., 2002). Previous studies also compared patterns of MSI in sporadic and Lynch syndrome-associated CRC and EC at the non-coding level, in particular by analyzing the mononucleotide BAT markers. It was shown that in both sporadic and Lynch syndrome-associated tumours, the proportion of unstable markers is 45

47 lower in EC than in CRC, and that the size of the allelic variations is smaller in EC than in CRC (Furlan et al., 2002; Kuismanen et al., 2002). These studies thus show quantitative differences in non-coding mononucleotide instability between the two types of tumours. Data on MSI of different types of markers and on qualitative differences are, however, scarce. Do CRC and EC show different ratios of insertions/deletions? Do EC display smaller insertions/deletions than CRC also in dinucleotide markers? Do CRC and EC have different preferences for specific types of MSI markers, as they have for target genes? In this study we addressed these questions in order to elucidate whether it is possible to define different profiles of MSI for tumours with different tissue origin, such as colorectal and endometrial cancers. MATERIALS AND METHODS Samples The patients participating in this study were all suspected of having Lynch syndrome and were selected from other research studies being conducted by our group (Berends et al., 2002; Niessen et al., 2006). All patients gave their informed consent for the study. The patients had either been diagnosed with CRC or EC under the age of 50 years, or had two or more Lynch syndrome-related cancers, including at least one CRC, irrespective of age and family history. The cases had been analyzed for germline mutations in the MLH1, MSH2, and MSH6 genes (Berends et al., 2001; Berends et al., 2003; Niessen et al., 2006). In total we included paraffin-embedded tumour tissue sections and normal tissue/blood samples from 194 colon carcinomas (CRC) and 68 endometrial carcinomas (EC). Thirteen out of the 194 CRC and 7 out of the 68 EC were from patients who carried a pathogenic germline mutation in one of the three MMR genes (MLH1, MSH2, and MSH6). These patients are referred to as mutation carriers. Average age of tumour onset for the different groups of patients is presented in table 2. 46

48 MSI Analysis MSI analysis was performed by fragment analysis using a panel of three mononucleotide markers (BAT25, BAT26, BAT40) and three dinucleotide markers (D2S123, D5S346, D17S250) as described previously (Berends et al., 2002). DNA was extracted from formalin-fixed, paraffin-embedded tumour sections and compared with DNA isolated from normal tissue from paraffin-embedded sections (when available) or peripheral blood lymphocytes from the same patient, as described previously (Berends et al., 2002). The samples were classified as MSI-H if more than 30% of the markers analyzed were unstable. Only samples with informative results for four or more of the six MSI markers (independent of the type of marker) were included in this study. For the MSI-H tumours, type of microsatellite mutation (deletion/insertion) and the size of these mutations were analyzed for every marker, and frequencies of instability were calculated. The size of mutations was measured as the difference between the highest peak in the normal tumour and the farthest peak in the unstable tumour. Statistical Analysis For the differences in type of mutations and frequencies of instability, the 2 test or Fisher`s Exact test were used. P values <0.05 were considered to be significant. Two-way factorial ANOVA was used for the differences in mutation size between markers and tumour tissues. RESULTS Frequency of microsatellite instability is highest in mutation carriers One-hundred and five tumors (40%) were classified as MSI-H and selected for further analysis. The frequencies of instability were overall as we expected from the literature, with mutation carriers showing significantly higher frequencies of MSI-H than non-carriers, both in colorectal tumors and in endometrial tumors (table 1). MSI-L was found in mutation carriers in the CRC group only, in two cases, both harboring an MSH6 mutation. In the EC mutation carriers, two MSS cases were detected; one carried an MSH6 mutation and the other an MSH2 mutation. 47

49 Table 1. Distribution of the samples by MSI status, presence or absence of germline mismatch repair mutations, and tumour tissue type Non-carriers Mutation carriers Total CRC EC CRC EC N N MSI-H 36% 40% 85% 71% 105 MSI-L 28% 29% 15% 0% 71 MSS 36% 31% 0% 29% 86 CRC colorectal carcinomas; EC endometrial carcinomas; MSI-H microsatellite instability high; MSI-L microsatellite instability low; MSS microsatellite stable; N, absolute number. Table 2. Average age of tumor onset of the different groups of patients Non-carriers Mutation carriers CRC EC CRC EC MSI-H MSI-L+MSS CRC colorectal carcinomas; EC endometrial carcinomas; MSI-H microsatellite instability high; MSI-L microsatellite instability low; MSS microsatellite stable. Type of microsatellite mutations (insertions/deletions) depends on type of repeat Frequencies of deletions and/or insertions occurring at each microsatellite marker were calculated for the MSI-H CRC and EC samples. No association between the prevalence of deletions or insertions and tumour type was found. A strong correlation between type of markers (mononucleotide vs. dinucleotide markers) and type of microsatellite mutation (insertion/deletion) was, however, observed. Mononucleotide markers were almost exclusively targets of deletions (98% in CRC and 100% in EC), whereas dinucleotide loci showed both deletions and insertions (Fig. 1). Simultaneous insertions and deletions were also detected in all dinucleotide markers (Fig. 1). 48

50 Figure 1. Frequencies of deletions, insertions, and simultaneous deletions and insertions in mononucleotide (BAT25, BAT26, BAT40) and dinucleotide (D2S123, D5S346, D17S250) MSI markers in MSI-H colorectal (upper panel) and endometrial (lower panel) carcinomas. Size of mutations in EC is smaller than in CRC The size of insertions/deletions was analyzed for each unstable locus of MSI-H tumours. The difference in size as defined in this study corresponded to the allele with the maximum length difference from the normal allele (observed in normal tissue/blood from the same patient). The mutation sizes were determined for both tumour types (Fig. 2). The mutations were significantly smaller in EC (average 6.02±0.45bp) than in CRC (average 7.67±0.34bp) (ANOVA: F (1,203) =11.25, 49

51 P<0.001). However, these differences are not statistically significant when analyzing each marker separately, except for D2S123. Different markers showed different mutation sizes (ANOVA: F (5,203) =14.67, P<0.001), but the relative differences between them remained similar in both tissues, as there is no interaction between the two variables (ANOVA: F (5,203) =1.3, P>0.1). Figure 2. Size of the insertions/deletions (in bp) for each MSI marker, in MSI-H colorectal (CRC) and endometrial (EC) carcinomas. * Statistically significant differences between the CRC/EC pair for each marker. Distribution of microsatellite instability Frequencies of instability were calculated for each MSI marker in both types of tumours. Neither of the two tissues showed a statistically significant preference for a specific type of marker. Both mononucleotide and dinucleotide markers are equally affected in the two tumour types and none of the markers was differently affected when we compared colorectal and endometrial tumours (Fig. 3). 50

52 Figure 3. Frequencies of instability observed for the mononucleotide (BAT25, BAT26, BAT40) and dinucleotide (D2S123, D5S346, D17S250) microsatellite markers, in MSI-H colorectal and endometrial tumours. The upper panel shows the frequencies for each marker; the lower panel shows the total of mononucleotide and dinucleotide instability. CRC, colorectal carcinomas; EC, endometrial carcinomas. DISCUSSION We report an analysis of mononucleotide and dinucleotide MSI markers, with regard to type, size and frequency of the mutations in MSI-H tumours with different tissue origins, namely colorectal and endometrial carcinomas. Looking at these features we found no significant differences between the EC and CRC MSI profiles, or at least not great enough to justify applying different MSI tests for the two tumour types. No statistically significant differences between mutation carriers and noncarriers were found. For this reason we were able to group all MSI-H cases 51

53 together for the analyses of the different MSI features. Nevertheless, we should keep in mind that the number of mutation carriers used in this study is much smaller than the number of non-carriers. Considering the age of the patients, mutation carriers developed CRC earlier than non-mutation carriers for MSI-H cases (table 2). Overall the average age of onset did not differ significantly among the mutation carriers and the non-mutation carriers. Our data show that the ratio of insertions and deletions is a markerdependent feature rather than a tissue-dependent one, as all the mononucleotide markers we studied showed almost exclusively deletions, while dinucleotide markers showed deletions, insertions, and simultaneous deletions and insertions, both in CRC and EC. The occurrence of mainly deletions in the mononucleotide markers was what we expected from the literature, since mutations in these markers are commonly referred to as shortenings. In the first reports on the involvement of microsatellite mutations in colon carcinogenesis mediated by a mutation in the MMR system ( mutator mutation ) (Ionov et al., 1993), a striking imbalance of deletions over insertions in Poly (A) sequences in CRC cell lines, with various degrees of microsatellite instability, was described. It was also known that, in Saccharomyces cerevisae, frameshifts on single base pair tracts tend to be deletions (Kunkel et al., 1989; Henderson and Petes, 1992). With respect to dinucleotide instability, if the three dinucleotide markers are taken as a whole, a tendency for only insertions over only deletions was observed in both CRC (44% vs. 37%) and EC (49% vs. 32%). Simultaneous deletions and insertions were found in 19% of dinucleotide loci in CRC and in 20% in EC. These results are in agreement with previous studies suggesting that insertions are more common than deletions among dinucleotide repeats (Twerdi et al., 1999; Ellegren, 2000; Yamada et al., 2002). This close association of the occurrence of insertions or deletions with the type of MSI marker suggests that characteristics of the repeats, such as repeat length, have more influence on the type of mutation occurring at a microsatellite repeat than the tissue origin of the tumour in which those mutations arise. Repeat length, together with base composition and number of repeat units per tract, are some of the features known to influence the mechanism of slipped-strand mispairing (Boyer et al, 2002), the main mechanism generating insertions or 52

54 deletions in microsatellites during DNA replication (Levinson and Gutman, 1987; Henderson and Petes, 1992). While analyzing the size of the deletions/insertions, we observed significant differences between CRC and EC, with EC showing smaller mutations than CRC for all markers, as previously described in the literature for the BAT markers. The differences are, however, not statistically significant in our study when comparing each marker alone. For instance Kuismanen et al. (2002) reported a mean deviation (bp) of 6.7 in CRC and 4.1 in EC for BAT25; and a mean deviation (bp) of 13.5 in CRC, and 8.5 in EC for BAT26. For the same markers we observed the same tendency to larger mutations in CRC: a mean deviation (bp) of 6.08 ± 0.41 in CRC, and 5.25 ± 0.62 in EC for BAT25; and 7.82 ± 0.65 in CRC and 6.00 ± 0.89 in EC for BAT26. The apparent differences found between our data and the mentioned study (Kuismanen et al., 2002) might be explained by differences in the classification of MSI by different observers. The number of samples included might also play a role in these differences, namely the inclusion of tumours with MMR mutations, typically more unstable than those not carrying MMR mutations. The stage of the tumour more specifically the rounds of replication that a given tumour has undergone might, in our opinion, also influence the size of microsatellite mutations. Furthermore, although the different markers showed different mutation sizes, the relative differences between them remained similar in both tissue types, leading to comparable patterns of instability, as observed for the type of mutation. Analyzing each marker alone, the frequencies of instability were not significantly different between CRC and EC for any of the markers. If we consider two groups, one of the three mononucleotide markers and one of the three dinucleotide markers, the pattern was again similar, with mono- and dinucleotide markers being affected in equal amounts and both of them similarly affected in CRC and EC. In conclusion, our results suggest that it is not possible to define specific profiles of MSI marker instability to distinguish tumours of different tissue origins. The features analyzed in our study type, size and frequency of instability of MSI markers seem to be representative of common patterns of MSI in CRC and EC. 53

55 A possible explanation for the quantitative differences described between CRC and EC, with EC having usually less unstable markers and smaller deletions/insertions, might be that they indicate different durations of tumour development, rather than reflecting real differences in profiles of the two tumour types. Tissue specificities, such as differences in tissue turnover between colorectal and endometrial epithelium might lead to different timings of tumour development and, in practice, result in different levels of instability. This would be in agreement with the tumour clock model of Shibata et al. (1996), who proposed that microsatellite alterations could be seen as a proxy for the number of cell divisions. Mutations accumulate with the number of replications, serving as a molecular clock to define the time of tumorigenesis and tracing the history of the tumour. ACKNOWLEDGEMENTS The authors thank Pedro Lourenço for the statistical analyses, and Jackie Senior for editing the manuscript. This work was supported by Fundação para a Ciência e a Tecnologia, Portugal (SFRH/BD/18832/2004) and by the European Community (FP LIFESCIHEALTH-5, proposal no ). 54

56 REFERENCES Aaltonen LA, Peltomäki P, Leach FS, et al Clues to the pathogenesis of familial colorectal cancer. Science 260: Berends MJ, Hollema H, Wu Y, et al MLH1 and MSH2 protein expression as a pre-screening marker in hereditary and non-hereditary endometrial hyperplasia and cancer. Int J Cancer 92: Berends MJ, Wu Y, Sijmons RH, et al Molecular and clinical characteristics of MSH6 variants: an analysis of 25 index carriers of a germline variant. Am J Hum Genet 70: Berends MJ, Wu Y, Sijmons RH, et al Toward new strategies to select young endometrial cancer patients for mismatch repair gene mutation analysis. J Clin Oncol 21: Boland CR, Thibodeau SN, Hamilton SR, et al A National Cancer Institute Workshop on Microsatellite Instability for cancer detection and familial predisposition: development of international criteria for the determination of microsatellite instability in colorectal cancer. Cancer Res 58: Boyer JC, Yamada NA, Roques CN, et al Sequence dependent instability of mononucleotide microsatellites in cultured mismatch repair proficient and deficient mammalian cells. Hum Mol Genet 11: Duval A, Reperant M, Compoint A, et al Target gene mutation profile differs between gastrointestinal and endometrial tumors with mismatch repair deficiency. Cancer Res 62: Ellegren H Heterogeneous mutation processes in human microsatellite DNA sequences. Nat Genet 24: Furlan D, Casati B, Cerutti R, et al Genetic progression in sporadic endometrial and gastrointestinal cancers with high microsatellite instability. J Pathol 197: Henderson ST, Petes TD Instability of simple sequence DNA in Saccharomyces cerevisiae. Mol Cell Biol 12: Ionov Y, Peinado MA, Malkhosyan S, et al Ubiquitous somatic mutations in simple repeated sequences reveal a new mechanism for colonic carcinogenesis. Nature 363: Kuismanen SA, Moisio AL, Schweizer P, et al Endometrial and colorectal tumors from patients with hereditary nonpolyposis colon cancer display different patterns of microsatellite instability. Am J Pathol 160:

57 Kunkel TA, Hamatake RK, Motto-Fox J, et al Fidelity of DNA polymerase I and the DNA polymerase I-DNA primase complex from Saccharomyces cerevisiae. Mol Cell Biol 9: Levinson G, Gutman GA Slipped-strand mispairing: a major mechanism for DNA sequence evolution. Mol Biol Evol 4: Niessen RC, Berends MJ, Wu Y, et al Identification of mismatch repair gene mutations in young patients with colorectal cancer and in patients with multiple tumours associated with hereditary nonpolyposis colorectal cancer. Gut 55: Peltomäki P, Lothe RA, Aaltonen LA, et al Microsatellite instability is associated with tumors that characterize the hereditary non-polyposis colorectal carcinoma syndrome. Cancer Res 53: Shibata D, Navidi W, Salovaara R, et al Somatic microsatellite mutations as molecular tumor clocks. Nat Med 2: Thibodeau SN, Bren G, Schaid D Microsatellite instability in cancer of the proximal colon. Science 260: Twerdi CD, Boyer JC, Farber RA Relative rates of insertion and deletion mutations in a microsatellite sequence in cultured cells. Proc Natl Acad Sci U S A 96: Yamada NA, Smith GA, Castro A, et al Relative rates of insertion and deletion mutations in dinucleotide repeats of various lengths in mismatch repair proficient mouse and mismatch repair deficient human cells. Mutat Res 499:

58 CHAPTER 4 The Hunt for New Target Genes in Endometrial Tumors Reveals the Involvement of the Estrogen-Receptor Pathway in Microsatellite Unstable Cancers Ana M. Ferreira 1,7, Iina Niittymäki 5, Sónia Sousa 7, Frans Gerbens 1, Krista Bos 1, Krista A. Kooi 1, Chris Esendam 1, Peter Terpstra 4, Menno Hardonk 4, Tineke van der Sluis 2, Monika Zazula 6, Jerzy Stachura 6, Ate G. van der Zee 3, Harry Hollema 2, Rolf H. Sijmons 1, Lauri A. Aaltonen 5, Helga Westers 1, Raquel Seruca 7, Robert M. W. Hofstra 1 Departments of 1 Genetics, 2 Pathology, 3 Gynecology, 4 Medical Biology, University Medical Center Groningen, University of Groningen, Groningen, the Netherlands. 5 Department of Medical Genetics, University of Helsinki, Helsinki, Finland. 6 Department of Patomorfology, Medical College, Jagiellonian University, Krakow, Poland. 7 Institute of Molecular Pathology and Immunology of the University of Porto, Porto, Portugal. Manuscript in preparation 57

59 ABSTRACT Microsatellite instability (MSI) in tumors results among others in an accumulation of mutations in (target) genes. Previous studies suggest that the profile of highly mutated target genes differs by tumor-type and that in particular for endometrial tumors the frequently mutated target genes remain to be identified. In our search for such highly mutated target genes in mismatch repair deficient endometrial cancers we identified 44 possible target genes of which 7 are highly mutated (>15%). Besides a high mutation frequency, 5 of these 7 could, by function, be linked to cancer development. Two genes encode proteins involved in chromatin remodeling (MBD6 and JMJD1C), one protein (JAK1) is involved in the JAK/STAT pathway, a pathway known to be implicated in cancer, one protein (KIAA1009) is essential in chromosome segregation and mitotic spindle assembly, and finally the most frequently mutated gene, NRIP1, encodes a co-repressor of the estrogen receptor (ER) pathway. Furthermore, we analyzed colorectal and gastric MMR deficient (MSI-H) tumors for mutations in ten of the identified target genes. Our data show that some of these newly identified target genes are tissue specificity, while others seem to play a more common role in MSI- H tumors, independently of the origin of the tissue. We therefore present a new profile of target genes, genes likely involved in endometrial cancer development. The most promising one is NRIP1, a gene influencing the ER pathway, the pathway with proven association with endometrial cancer development. These findings might prove relevant to look for additional target genes and it might give new insights for the design of novel therapeutic treatments. 58

60 INTRODUCTION Endometrial carcinoma (EC) is one of the most common forms of cancer among women in Western countries. High exposure to estrogens, obesity and family history are considered the main risk factors for the disease. Moreover, EC is the most common extra-colonic cancer in Lynch syndrome patients. This syndrome, also known as hereditary nonpolyposis colorectal cancer (HNPCC), is caused by germline mutations in the DNA mismatch repair (MMR) genes (Aaltonen et al., 1993; Ionov et al, 1993; Thibodeau et al., 1993). The MMR system repairs DNA replication errors that are not immediately corrected by DNA polymerase. The MMR system, therefore, plays a crucial role in DNA replication accuracy. Functional inactivation of MMR genes by mutations or epigenetic changes leads, among others, to the accumulation of insertions/deletions. These are easily identified in short DNA tandem repeat sequences (microsatellites); this phenotype is called microsatellite instability (MSI). MSI can be detected in tumors from Lynch syndrome patients (Aaltonen et al., 1993; Ionov et al., 1993; Peltomäki et al., 1993; Thibodeau et al., 1993), but is also present in a fraction (~15-25%) of sporadic cases of endometrial, colorectal, and gastric cancer (Boland et al., 1998). Genes containing repeat sequences are vulnerable to replication errors in MMR-deficient tumors. MSI can occur at non-coding but also at coding repeat sequences of regulatory genes which might play a role in tumor development. Such genes are generally called target genes and thought to be the key players during MSI-H carcinogenesis. Over 160 target genes have been identified to date in MMR-deficient tumors (Vilkki et al., 2002; Røyrvik et al., 2007). Mutations were mostly searched for in MSI-H colorectal tumors (CRC). Endometrial and gastric (GC) tumors have been analyzed to a lesser extent and were mainly studied for target genes previously reported in CRC. From previous studies it becomes however clear that although there are target genes commonly involved in MSI-H tumors of diverse origin, e.g. BAX, others show considerable tissue specificity (Duval et al., 1999; Schwartz et al., 1999; Semba et al., 2000). It has been shown that the profile of target genes differs between EC and gastrointestinal tumors with MMR deficiency, in both qualitative and quantitative manners (Myeroff et al., 1995; Gurin et al., 1999; Duval et al., 2002). In fact in EC a small number of (highly 59

61 mutated) target have been identified so far, suggesting that other not yet identified target genes remain to be found. This study aims at identifying these target genes for MSI-H EC. MATERIALS AND METHODS Samples -Six fresh-frozen normal endometrial tissue samples were obtained from six women undergoing surgery at the University Medical Center Groningen (UMCG, Groningen, the Netherlands) for other reasons than uterine cancer. These samples were used for the expression arrays experiments. -Forty-two paraffin-embedded tissue sections from MSI-H endometrioid endometrial carcinomas were obtained from the Department of Pathology, University Medical Center Groningen (Groningen, the Netherlands) and from the Department of Pathomorphology, Jagiellonian University (Cracow, Poland). Freshfrozen tumor tissues were available for 10 of these samples. -Forty MSI-H colorectal tumors were obtained from the Department of Medical Genetics, University of Helsinki (Helsinki, Finland), and from the Department of Pathology, University Medical Center Groningen (Groningen, the Netherlands). -Fifteen MSI-H gastric tumors were obtained from The Institute of Molecular Pathology and Immunology of the University of Porto (Porto, Portugal) All the patients participating on this study have given their written consent. DNA isolation Genomic DNA was isolated from fresh-frozen tissue and formalin-fixed, paraffinembedded tumor tissue using the Qiagen DNA Mini Kit (Qiagen, Venlo, the Netherlands), using a standard protocol (protocol available on request). Micro-array experiments RNA isolation RNA was isolated using the RNeasy mini kit (Qiagen, Valencia, CA) and was treated with RNase-free DNase I (Qiagen) as described by the manufacturer. 60

62 mrna Amplification and Cy-dye coupling Linear amplification of mrna was performed essentially according to a protocol of the Dutch Cancer Institute ( Briefly, amplification started with first strand cdna synthesis from 2 µg of total RNA, using Superscript II RT-polymerase (GIBCO - Invitrogen) and a specific oligo(dt) primer containing a 17bp T7 polymerase recognition site (5'- ggccagtgaattgtaatacgactcactatagggaggcggt-24-3') (Eurogentec, Seraing, Belgium). After second strand synthesis, double-stranded cdna was purified with the Qiaquick PCR purification kit (Qiagen). In vitro transcription was performed with the T7 Megascript kit (Ambion, Huntingdon-Cambridgeshire, UK) as described by the manufacturer, but using instead of UTP, a 1:1 mixture of aminoallyl-utp (Ambion) and UTP with a final concentration of 7.5 mm for all NTPs ('t Hoen et al., 2004). Amplified RNA (arna) was purified with the RNA clean up protocol (Qiagen). Five µg of arna was labeled by coupling monoreactive Cyanine 3 (2.5 nmol per reaction) or Cyanine 5 (2.5 nmol per reaction) fluorophores (Amersham Biosciences, Little Chalfont, Buckinghamshire, UK) to the aminoallyl-modified nucleotides. Labelled arna was separated from unincorporated Cyanine 3 or Cyanine 5 molecules with Microspin G50 columns (Millipore Corp, Amsterdam, The Netherlands) following the recommendations of the manufacturer. Experimental design For the identification of genes expressed in normal endometrium a randomized design was applied for micro-array hybridization. Each of all six normal cdna endometrium tissue samples was labeled with Cyanine 3 and Cyanine 5 separately and subsequently assigned at random to a sample labeled with the opposite dye for hybridization. Micro-array hybridization In-house manufactured human oligonucleotide arrays were used containing the Qiagen/operon 21, mer human gene specific oligonucleotide set version 2.1 extended with 4,000 negative and positive control features. The oligonucleotides were printed in a concentration of 10 pm on Ultra-GAPS amino-silane coated slides (Corning BV. Life Sciences, New York, USA) using BioRobotics 10K quill pins with 61

63 the MicroGrid spotter (Isogen). Blocking, prehybridization and hybridization were performed as described by Hegde et al. (2000), with some modifications (detailed protocol available on request). Hybridization was performed in hybridization chambers (Telechem International Inc, Sunnyvale, CA, USA) in a water bath at 52 C in the dark for approximately 48 h. Subsequently, slides were washed, dried by centrifugation at 800 rpm during 3 min and scanned with an Affymetrix GMS428 TM array scanner. Micro-array data analysis Fluorescent signal intensity data for each spot and for each fluorophore were extracted from the scanned images of each micro-array slide using ImaGene version 5.6 (BioDiscovery, El Segundo, California, USA). Signal intensity data were log transformed and for each spot the Cyanine 5 signal intensity/cyanine 3 signal intensity ratio was determined and subjected to print-tip lowess intensity dependent normalization using the Limma package from the Bioconductor project in R ( Since no dependency exists between both samples during hybridization ('t Hoen et al., 2004), normalized log-ratios were back transformed to log intensities. Further data analysis was performed using BRB ArrayTools v3.2 developed by Dr. Richard Simon and Amy Peng Lam ( Basically, data was vigorously filtered to exclude control spots, empty spots, spots with high between-pixel-intensity variability and spots designated as bad by eye. Genes that had more than 25% missing data across all observations were excluded from the analysis. Genes with expression 10 times higher than the background were identified. Selection of mononucleotide repeats A computer program (Repeat Finder) was created for the purpose of finding repetitive tracts in DNA sequences. A data file containing the coding sequences (CDS) of all the genes selected by microarrays in normal endometrium was uploaded to the program. Genes with mononucleotide tracts of (A)7, (T)7, (C)7, (G)7, (A)8, and (T)8 in there coding sequence were identified and selected. 62

64 Mutation screening The sequences encompassing the repeats of interest were extracted from the Ensembl database ( Primers were designed using the Primer3 program ( Amplicons were amplified and subsequently PCR products purified using ExoSAP-IT enzymatic reagent (US Biochemical Corporation) according to the manufacturer s instructions. Mutation analysis was performed by direct sequencing using big dye Terminator Kit version 3.1 (Applied Biosystems) and ABI 3730 Automatic DNA sequencer (Applied Biosystems) following the recommendations of the manufacturer. The list of genes screened and the primer sequences and PCR conditions are available on request. The mutation analysis in the gastric tumor samples was performed by size separation using multiplex PCR. The products were read in a ABI 3100 sequence analyzer using Peack scanner Software v1,0 with a 500 liz size standard electropherogram. RESULTS Expression profiling of normal human endometrium Six normal endometrial tissue samples were used for expression profiling with 21K oligonucleotide micro-arrays. A total of 2338 genes showed expression values 10X higher than the background signals. These genes were considered as being clearly expressed in normal endometrial tissue and therefore selected for the next step of finding repeat sequences. Genes containing coding mononucleotide repeats (A/T)7, (C/G)7, (A/T)8 The analysis of the coding sequences of the 2338 selected genes with the computer program Repeat Finder revealed 573 repeats of interest identified in 432 out of the 2338 genes. The number of repeats of each type was the following: 244 (A)7, 74 (T)7, 114 (C)7, 57 (G)7, 72 (A)8, and 12 (T)8 repeats. The program Repeat Finder is home made. 63

65 Mutation screening of the mononucleotide repeats in MSI-H EC Four hundred and seventy six (476) primer sets were successfully designed (for the others we did not succeeded designing good primer pairs) encompassing a total of 496 repeats (in 382 genes): 214 A(7), 61 T(7), 104 C(7), 46 G(7), 62 A(8), and 9 T(8) repeats. The primers were first used for mutation analysis in 10 MSI-H EC for which frozen material was available. Heterozygous frameshift mutations (figure 1) resulting either from insertions or deletions (all +/-1bp) were found in 49 repeats (for at least 1 tumor DNA sample), in 44 candidate genes (some genes host more than one repeat of interest). The screening of those repeats was then extended to 32 additional MSI-H EC samples (results are given in table 1). Seven genes were found to be mutated in 15% or more of the samples: NRIP1, SRPR, MBD6, JAK1, KIAA1009, JMJD1C, ADD3, with mutation frequencies of 34%, 26%, 24%, 20%, 19%, 15% and 15%, respectively (Table I). The other genes had mutation frequencies below the cut-off of 15%. Figure 1. Example of frameshift mutations at coding mononucleotide repeats. In the upper panel, the normal (reverse) sequence of the (A)8 repeat in NRIP1 gene is depicted; the lower panel shows a deletion of an (A) base found for that repeat in a tumor sample. 64

66 Table I. Results of the mutational screening for the forty-four candidate target genes for which a mutation was found in at least one tumor. Gene Description Repeat Mutat Freq. (%) SEC16 SEC16 homolog A (S. cerevisiae) C7 5,7 (2/35) SPG20 spastic paraplegia 20 (Troyer syndrome) T8 5,9 (2/34) CAMSAP1L1 calmodulin regulated spectrin-associated protein 1-like 1 A8 7,1 (2/28) ZMIZ1 zinc finger, MIZ-type containing 1 C7 4,3 (1/23) ZMIZ1 zinc finger, MIZ-type containing 1 C7 9,1 (2/22) INTU inturned planar cell polarity effector homolog (Drosophila) A8 4,2 (1/24) IGSF9 immunoglobulin superfamily, member 9 C7 4,3 (1/23) KIAA1370 KIAA1370 A8 12,9 (4/31) JMJD1C jumonji domain containing 1C A8 15,1 (5/33) FAM135A family with sequence similarity 135, member A A8 3,3 (1/30) KIAA1797 KIAA1797 A7 6,7 (1/15) MBD6 methyl-cpg binding domain protein 6 3XC7 24,1 (7/29) DACH1 dachshund homolog 1 (Drosophila) A7 3,1 (1/32) POLE3 polymerase (DNA directed), epsilon 3 (p17 subunit) A7 4,8 (1/21) JAM3 junctional adhesion molecule 3 G7 8,3 (3/36) HEXDC hexosaminidase (glycosyl hydrolase family 20, catalytic domain) containing C7 + G7 11,4 (4/35) CHD4 chromodomain helicase DNA binding protein 4 A7 8,1 (3/37) LYN v-yes-1 Yamaguchi sarcoma viral related oncogene homolog A7 6,7 (2/30) PPP1R10 protein phosphatase 1, regulatory (inhibitor) subunit 10 C7 3,7 (1/27) SRPR signal recognition particle receptor ('docking protein') A8 25,8 (8/31) TTC3 tetratricopeptide repeat domain 3 A8 3,3 (1/30) NRIP1 nuclear receptor interacting protein 1 A8 34,3 (12/35) AP3B1 adaptor-related protein complex 3, beta 1 subun A8 8,1 (3/37) INPPL1 inositol polyphosphate phosphatase-like 1 C7 14,3 (5/35) FLNB filamin B, beta (actin binding protein 278) G7 9,7 (3/31) JAK1 Janus kinase 1 (a protein tyrosine kinase) A8 20 (7/35) BAT1 HLA-B associated transcript 1 T8 2,9 (1/35) MTA1 metastasis associated 1 G7 3,3 (1/30) FXR1 fragile X mental retardation, autosomal homolog 1 A8 12,5 (2/16) IFNGR2 interferon gamma receptor 2 (interferon gamma transducer 1) T7 2,8 (1/36) RBM6 RNA binding motif protein 6 G7 3,4 (1/29) SF3B2 splicing factor 3b, subunit 2, 145kDa A8 10,7 (3/28) RABGAP1 RAB GTPase activating protein 1 A8 6,25 (2/32) TNPO3 transportin 3 C7 2,9 (1/34) TRPM5 transient receptor potential cation channel, subfamily M, member 5 C7 5,3 (1/19) ITM2B integral membrane protein 2B C7 2,8 (1/36) SVIL supervillin G7 10,8 (4/37) INTS12 integrator complex subunit 12 T7 5,9 (2/34) C17orf63 chromosome 17 open reading frame 63 C7 3,7 (1/27) KIAA1009 KIAA1009 T8 + A7 18,5 (5/27) CPEB3 cytoplasmic polyadenylation element binding protein 3 C7 2,9 (1/34) NOL7 nucleolar protein 7, 27kDa A8 7,1 (2/28) ADD3 adducin 3 (gamma) A8 14,7 (5/34) PHKB phosphorylase kinase, beta A7 4,3 (1/23) TFPI tissue factor pathway inhibitor (lipoprotein-associated coagulation inhibitor) A7 3,1 (1/320 Mutation screening of 10 target genes in CRCs and GCs (NRIP1, SRPR, MBD6, JAK1, KIAA1009, JMJD1C, ADD3, INPPL1, SVIL, and HEXDC) The seven genes with mutation frequencies equal or higher than 15% and three additional genes (INPPL1, SVIL, HEXDC) were then screened in CRC and GC samples. Frameshift mutations were found in all the genes analyzed, in at least 65

67 one of the tumors. All mutations found were heterozygous and consisted of +1 or - 1 bp, as in figure 1, except for SRPR in a GC sample, where -2 bp mutations were also found. Three CRCs and 2 GCs did not show mutations for any of the repeats. For the CRC group, no associations were found between mutations and the classification of the patients as sporadic/lynch syndrome patients (data not available for GCs or ECs). Also no correlation was found between the mutations and the gender of the patients. Mutation frequencies higher than a cut-off value of 15% were found in SRPR, ADD3, MBD6 and NRIP1 genes (47%, 37%, 25% and 22%, respectively) in CRC samples; and in ADD3 (47%), SRPR (27%), SVIL (27%), JAK1 (20%) and INPPL1 (20%) in GC samples (Table II). Figure 2 shows the comparative profiles of the three types of tumors for the new target genes. After analyzing repeats in 10 genes, we found: 62 mutations (in 321 repeats) in EC patients, giving 1.9 mutations per patient on average; 53 mutations (in 289 repeats) in CRC giving 1.8 mutations per patient on average; 27 mutations (in 150 repeats) in GC (1.8 mutations per patient). Table II. Mononucleotide repeats analyzed and respective mutation frequencies found in the MSI-H colorectal and gastric tumor samples used; the results obtained for the MSI-H endometrial tumors are included for comparison. ND: not determined. Gene Exon Repeat CRC GC EC NRIP1 3 A8 22% (8/36) 13% (2/15) 34% (12/35) SRPR 4 A8 47% (14/30) 27% (4/15) 26% (8/31) MBD6 7/8 3XC7 25% (10/40) 13% (2/15) 24% (7/29) JAK1 5 A8 3% (1/30) 20% (3/15) 20% (7/35) KIAA T8+A7 ND 7% (1/15) 19% (5/27) JMJD1C 9 A8 3% (1/34) 0% (0/15) 15% (5/33) ADD3 14 A8 37% (11/30) 47% (7/15) 15% (5/34) INPPL1 26 C7 10% (3/30) 20% (3/15) 14% (5/35) SVIL 31 G7 14% (4/29) 27% (4/15) 11% (4/37) HEXDC 12 C7 3% (1/30) 7% (1/15) 11% (4/35) 66

68 Box 1: Function of the highly mutated proteins in EC NRIP1 (nuclear receptor-interacting protein 1) is a modulator of several, if not all, nuclear receptors (e.g. retinoic acid receptor, thyroid receptor, androgen receptor). It is also a known co-repressor of the estrogen-receptor (ER) pathway. Silencing of NRIP1 leads to growth advantages in breast cancer derived cell lines. SRPR encodes the signal recognition particle receptor subunit alpha ( docking protein ), which together with the SRP (signal recognition particle) ensures the correct targeting of the nascent secretory proteins to the endoplasmic reticulum membrane system (Janin et al., 1992). MBD6 (methyl-cpg binding domain protein 6) contains a methyl-cpg-binding domain (MBD) and is possible involvement in DNA methylationas are other MBD proteins. Another MBD protein, MBD4, is a known target gene in MSH-H tumors (Røyrvik et al., 2007). JAK1 (Janus kinase 1) is a protein-tyrosine kinase (PTK). It is a widely expressed membrane-associated phosphoprotein. Deregulation of the JAK-STAT signaling pathway has been described in a variety of cancers and immune disorders. Mutations in JAK1 have been reported in human leukemias and in several solid cancers (Jeong et al., 2008). Furthermore, the JAK/STAT3 pathway has been suggested as a new potential target for therapy of CRC (Xiong et al., 2008). KIAA1009 is a new microtubule-associated ATPase involved in cell division, a protein with essential role on chromosome segregation and mitotic spindle assembly. It is expressed throughout mitosis, and it is located at the pole of the mitotic spindle, associated with microtubules, and in the centrosome. The cell death induced by transfection with QN1/KIAA1009 sirna suggests that QN1/KIAA1009 protein is a potential target for novel antimitotic cancer therapies (Leon et al., 2006). JMJD1C (jumonji domain containing 1C), formerly TRIP8 (thyroid hormone receptor interactor 8) codes for a nuclear protein predicted to be a transcriptional regulator associated with nuclear thyroid hormone receptors. JMJD1C is believed to be a histone H3K9 demethylase, therefore playing a major role in histone code. ADD3 (adducin 3) is a membrane-cytoskeleton-associated protein that is involved in the assembly of the spectrin-actin network in erythrocytes and at sites of cell-cell contact in epithelial tissues. Not much is know about this protein. 67

69 DISCUSSION In the present study we report seven new target genes for MSI-H endometrial cancer. Additionally, we show that most of those genes are also mutated in colorectal and gastric tumors, although with different frequencies. A mutational screening was performed on mononucleotide repeats in the coding sequence of genes that are expressed in normal human endometrial tissue. By using normal tissue, we aimed to include genes with a potential role in the normal maintenance of the endometrium and therefore theoretically the ones to be affected in disease context. More commonly, an approach of comparing tumor versus normal tissue would have been followed and down-regulated genes would have been selected; however, in that case, possibly the mutations reported in this study would not be found, since they are heterozygous and the gene can thus still be expressed. For the mutational screening we selected the genes with expression signals ten-fold higher than the background signal for further analysis. We are aware that a large number of candidate genes will in this way be excluded because of their expression low expression. Another reason that we have missed target genes is the fact that we selected only for specific repeat length. Type and length of the repeats are highly relevant for their mutation frequencies. Mononucleotide repeats are commonly considered the most MSI-H specific type of repeats and tracts with lengths between 6 and 10bp are usually taken. Considering the recent paper of Sammalkorpi et al 2007, to our knowledge the first study on mutation frequencies of intergenic repeats, we decided to not include (G)8 and C(8) repeats or longer, to avoid high background mutation frequencies interfering with the results. Known target genes like BAX and TGFβRII for instance have mononucleotide repeats longer than 9 bases and are therefore by definition not included in our screen. We have reasons to believe that this is a safe set-up of the experiment to avoid false positive candidates. These reasons imply that we only found a subset of all genes mutated in EC. Are the target genes found really involved in endometrial cancer development? To define a real target gene, criteria have been formulated (Duval and Hamelin, 2002). They consist of: (1) a high mutation frequency; (2) biallelic 68

70 inactivation of the gene by simultaneous alteration of the second allele s repeat tract or by point mutation or allelic loss; (3) possible involvement of the encoded protein in tumor development; (4) occurrence of mutations within the pathway in MSI-negative tumors; (5) in vitro or in vivo functional suppressor studies. These criteria are considered rather strict and some are controversial (Perucho, 2003). Due to this controversy and to the general lack of functional evidence proving the relevance of the candidate genes, usually a high mutation frequency (above a cut-off value of 12-15%) is taken as major criteria to classify a gene as a real target gene (Duval and Hamelin, 2002). When applying the 15% cut-off rule, 7 new target genes were identified: NRIP1, SRPR, MBD6, JAK1, KIAA1009, JMJD1C, and ADD3. To our knowledge these genes have never been reported before in MSI-H endometrial cancer. Bi-allelic mutations, the second criteria, were never found. All the mutations in this study were heterozygous. Whether however biallelic mutations are indeed necessary can be debated. Haploinsufficiency, caused by the loss of only one allele, is frequent finding in cancer. A good example is mono-allelic loss of PTEN, the main mutated gene in endometrial cancer (Nardella et al., 2008). The function of the encoded protein, and thereby its possible involvement in tumor development, is also an inclusion requirement for a real target gene. In Box 1 a short description of the proteins encoded by the newly identified target genes is given. Considering the mutation frequency and the function of the protein, NRIP1 came out of our study as the best candidate target gene for MSI-H EC. It was the highest mutated gene (34% of EC tumors) and it is a known co-repressor of the estrogen-receptor (ER) pathway. The ER is a very important pathway for endometrial tissue regulation, as the endometrium is a sex hormone responsive tissue, highly regulated by the concentrations of estrogens. The ER is a ligandactivated transcription factor from the nuclear receptor superfamily. Several estrogen-responsive genes have been described. Genetic alterations in ER and ER-responsive genes are thought to be key players in the development of hormone-dependent tumors (Notarnicola et al., 2001). Furthermore, the high exposure to estrogens is currently considered the major risk factor for EC. Moreover, approximately 80% of all sporadic EC tumors the endometrioid endometrial carcinomas - are estrogen-dependent carcinomas. In addition to this, 69

71 it has been reported that NRIP1 is essential for female fertility in mice (White et al., 2000), and that mutations in NRIP1 may act as predisposing factor for human endometriosis (Caballero et al., 2005). We believe that it is very likely that NRIP1 mutations might result in functional differences at the ER-pathway level. We expect that inactivation of NRIP1 will interfere with the process of co-repression of the ER complex and lead to differences in the expression of estrogen-dependent genes. This could eventually be linked to tumors growth advantages. Most of the other six genes found highly mutated can by function also be coupled to the carcinogenic process. Two of the proteins are involved in chromatin remodeling (MBD6 and JMJD1C), JAK1 is likely involved in a pathway often found implicated in cancer in general, and KIAA1009 is essential in chromosome segregation and mitotic spindle assembly, a process which, when disturbed, will contribute to cancer development. Taking the mutation frequencies and the (known) function of the newly identified target genes we have reasons to suggest that for sure part of the seven genes do play a role in MSI-H EC development. Comparison of endometrial versus gastrointestinal tumors All the genes showed high mutation frequencies in at least one of the tumor types, except HEXDC, which reached the highest frequency of only 11%, in EC samples. Comparing the mutation frequencies found in the CRC and GC samples with the EC samples we observe some differences in the profile of target genes affected (table II and figure 1). We conclude that the target genes included in this study are involved both in EC and gastrointestinal carcinogenesis, although in a different order of mutation frequencies and therefore giving a different profile dependent on the tissue origin, as expected from previous studies on target gene profiles of MSI tumors (Duval et al., 2002). However, the differences found in the JAK1 repeat (3% in CRC; 20% in GC and EC) is quite striking, especially because JAK1 mutations have been reported in several solid cancers and the JAK/STAT3 pathway has even been suggested as a new potential target for therapy of CRC (Xiong et al., 2008). As this screening only looked for mutations in one mononucleotide repeat, which is only a very small part of the coding sequence of the gene, it can not be excluded that other mutations are 70

72 present and that the gene plays a more important role in CRC as well. Of course this holds true for all genes analyzed. Finding mutations in NRIP1, a protein clearly connected to estrogens and estrogen receptor signaling, in CRCs and GC seems surprising as the colon tissue is not typically hormone responsive. However, NRIP1 mutations in CRC and GC have been reported before, although at lower frequencies (under the threshold of 15%). Frameshift mutations were found in an A9 coding microsatellite, in 13% of MSI-H GCs and 7% of MSI-H CRC (Røyrvik et al., 2007). Moreover several findings in colorectal cancer support a hypothesis that high estrogen levels can have a protective effect. These findings have been used as an explanation for the gender bias observed on CRC incidence, with a lower incidence of the disease in women than in men. In particular, hormonal changes associated with pregnancy (McMichael and Potter, 1980), and hormone replacement therapy (HRT) have been associated with lower risk of CRC (Potter, 1995; Peipins et al., 1997; Chen et al., 1998; Kadiyska et al., 2007). It is interesting to notice that two other genes of our list, JMJD1C and SVIL encode proteins involved in the regulation of hormone receptors. JMJD1C is a transcription regulator of nuclear thyroid hormone receptors; SVIL has been described as an androgen-receptor (AR) co-regulator that can enhance AR transactivation in muscle and other cells (Ting et al., 2002). SVIL has already been linked to cancer, as it is underexpresssed in prostate cancer (Vanaja et al., 2006). CONCLUSIONS Future studies at the functional level are essential to elucidate how NRIP1 and the other genes are implicated in carcinogenesis, since even when mutations are found in genes with putative roles in tumor-related processes, the chance of having a bystander gene instead of a real target gene can not be discarded. However, with this study we propose 7 new genes, and in particular NRIP1 as novel target genes for MSI-H endometrial cancer. Our results also support the idea that MSI gastrointestinal and EC tumors present some differences in the profile of target genes affected but that there are also some genes affected at similar frequencies 71

73 among the different types of tumors. More importantly, the present study suggests that there might exist a stronger link between hormones and MSI than thought so far, and that genes of hormone-related pathways should be considered important candidates when searching for new target genes of MSI tumors Mutation Frequency (%) MSI-H EC MSI-H CRC MSI-H GC 0 NRIP1 SRPR MBD6 JAK1 KIAA1009 JMJD1C ADD3 INPPL1 SVIL HEXDC Figure 2. Distribution of mutation frequencies found in MSI-H endometrial (EC), colorectal (CRC) and gastric carcinomas (GC), for the 10 most mutated target genes. ACKNOWLEDGEMENTS This work was supported by the Portuguese Foundation for Science and Technology ( Fundação para a Ciência e a Tecnologia ), Portugal (Grant ref.:sfrh/bd/18832/2004) and by the European Community (FP LIFESCIHEALTH-5, proposal No ). 72

74 REFERENCES Aaltonen LA, Peltomäki P, Leach FS, et al. Clues to the pathogenesis of familial colorectal cancer. Science 1993;260(5109): Boland CR, Thibodeau SN, Hamilton SR, et al. A National Cancer Institute Workshop on Microsatellite Instability for cancer detection and familial predisposition: development of international criteria for the determination of microsatellite instability in colorectal cancer. Cancer Res 1998;58(22): Caballero V, Ruiz R, Sainz JA, et al. Preliminary molecular genetic analysis of the Receptor Interacting Protein 140 (RIP140) in women affected by endometriosis. J Exp Clin Assist Reprod 2005;2:11. Chen MJ, Longnecker MP, Morgenstern H, et al. Recent use of hormone replacement therapy and the prevalence of colorectal adenomas. Cancer Epidemiol Biomarkers Prev 1998;7(3): Duval A, Gayet J, Zhou XP, et al. Frequent frameshift mutations of the TCF-4 gene in colorectal cancers with microsatellite instability. Cancer Res 1999;59(17): Duval A, Hamelin R. Mutations at coding repeat sequences in mismatch repair-deficient human cancers: toward a new concept of target genes for instability. Cancer Res 2002;62(9): Duval A, Reperant M, Compoint A, et al. Target gene mutation profile differs between gastrointestinal and endometrial tumors with mismatch repair deficiency. Cancer Res 2002;62(6): Gurin CC, Federici MG, Kang L, et al. Causes and consequences of microsatellite instability in endometrial carcinoma. Cancer Res 1999;59(2): Hegde P, Qi R, Abernathy K, et al. A concise guide to cdna microarray analysis. Biotechniques 2000;29(3):548-50, 552-4, 556 passim. Ionov Y, Peinado MA, Malkhosyan S, et al. Ubiquitous somatic mutations in simple repeated sequences reveal a new mechanism for colonic carcinogenesis. Nature 1993;363(6429): Janin N, Delattre O, Lipinski M. PCR-assisted localization of the human SRPR gene. Hum Genet 1992;88(5): Jeong EG, Kim MS, Nam HK, et al. Somatic mutations of JAK1 and JAK3 in acute leukemias and solid cancers. Clin Cancer Res. 2008;14(12):

75 Kadiyska T, Yakulov T, Kaneva R, et al. Vitamin D and estrogen receptor gene polymorphisms and the risk of colorectal cancer in Bulgaria. Int J Colorectal Dis 2007;22(4): Leon A, Omri B, Gely A, et al. QN1/KIAA1009: a new essential protein for chromosome segregation and mitotic spindle assembly. Oncogene 2006;25(13): McMichael AJ, Potter JD: Reproduction, endogenous and exogenous sex hormones, and colon cancer: a review and hypothesis. J Natl Cancer Inst 1980;65: Myeroff LL, Parsons R, Kim SJ, et al. A transforming growth factor beta receptor type II gene mutation common in colon and gastric but rare in endometrial cancers with microsatellite instability. Cancer Res 1995;55(23): Nardella C, Chen Z, Salmena L, et al. Aberrant Rheb-mediated mtorc1 activation and Pten haploinsufficiency are cooperative oncogenic events. Genes Dev. 2008;22(16): Notarnicola M, Gristina R, Messa C, et al. Oestrogen receptors and microsatellite instability in colorectal carcinoma patients. Cancer Lett. 2001;168(1): Peipins LA, Newman B, Sandler RS. Reproductive history, use of exogenous hormones, and risk of colorectal adenomas. Cancer Epidemiol Biomarkers Prev 1997;6(9): Peltomäki P, Lothe RA, Aaltonen LA, et al. Microsatellite instability is associated with tumors that characterize the hereditary non-polyposis colorectal carcinoma syndrome. Cancer Res 1993;53: Perucho M. Tumors with microsatellite instability: many mutations, targets and paradoxes. Oncogene 2003;22(15): Potter JD. Hormones and Colon cancer. J Natl Cancer Inst 1995;87(14): Røyrvik EC, Ahlquist T, Rognes T, et al. Slip slidin' away: a duodecennial review of targeted genes in mismatch repair deficient colorectal cancer. Crit Rev Oncog 2007;13(3): Sammalkorpi H, Alhopuro P, Lehtonen R, et al. Background mutation frequency in microsatelliteunstable colorectal cancer. Cancer Res 2007;67(12): Schwartz S Jr, Yamamoto H, Navarro M, et al. Frameshift mutations at mononucleotide repeats in caspase-5 and other target genes in endometrial and gastrointestinal cancer of the microsatellite mutator phenotype. Cancer Res 1999;59(12):

76 Semba S, Ouyang H, Han SY, et al. Analysis of the candidate target genes for mutation in microsatellite instability-positive cancers of the colorectum, stomach, and endometrium. Int J Oncol 2000;16(4): 't Hoen PA, Turk R, Boer JM, et al. Intensity-based analysis of two-colour microarrays enables efficient and flexible hybridization designs. Nucleic Acids Res 2004;32(4):e41. Thibodeau SN, Bren G, Schaid D. Microsatellite instability in cancer of the proximal colon. Science 1993;260(5109): Ting HJ, Yeh S, Nishimura K, et al. Supervillin associates with androgen receptor and modulates its transcriptional activity. Proc Natl Acad Sci U S A 2002;99(2): Vanaja DK, Ballman KV, Morlan BW, et al. PDLIM4 repression by hypermethylation as a potential biomarker for prostate cancer. Clin Cancer Res 2006;12(4): Vilkki S, Launonen V, Karhu A, et al. Screening for microsatellite instability target genes in colorectal cancers. J Med Genet 2002;39(11): White R, Leonardsson G, Rosewell I, et al. The nuclear receptor co-repressor nrip1 (RIP140) is essential for female fertility. Nat Med 2000;6(12): Xiong H, Zhang ZG, Tian XQ, et al. Inhibition of JAK1, 2/STAT3 signaling induces apoptosis, cell cycle arrest, and reduces tumor cell invasion in colorectal cancer cells. Neoplasia 2008;10(3):

77 76

78 CHAPTER 5 Estrogen, MSI and Lynch Syndrome-Associated Tumors Ana M. Ferreira 1,2, Helga Westers 1, André Albergaria 2, Raquel Seruca 2,3, Robert M. W. Hofstra 1 1 Department of Genetics, University Medical Centre Groningen, University of Groningen, Groningen, the Netherlands. 2 Institute of Molecular Pathology and Immunology of the University of Porto, Porto, Portugal. 3 Faculdade de Medicina da Universidade do Porto, Portugal. Under review 77

79 ABSTRACT Estrogens play a major role in the biology of hormonally responsive tissues but also in the normal physiology of various non-typical hormone responsive tissues. In disease, estrogens have been associated with tumor development, in particular with tumors such as breast, endometrium, ovary and prostate. In this paper we will review the molecular mechanisms by which estrogens are involved in cancer development, with a special focus in Lynch syndromerelated tumors. Further, we discuss the role estrogens might have on cell proliferation and apoptosis, how estrogens metabolites can induce DNA damage, and we discuss a possible connection between estrogens and changes in DNA (hypo- and hyper-) methylation. In this review we will also address the protective effect that high levels of estrogen have in MMRrelated neoplasias. 78

80 INTRODUCTION The most common types of cancer worldwide occur in hormonally responsive tissues, such as breast, endometrium, ovary and prostate. Tumors occurring in these tissues show strong associations with the exposure to exogenous or endogenous steroidal hormones. Estrogens are a group of steroid compounds which are present in both men and women; however, their levels are significantly higher in women of reproductive age. There are three types of estrogens of which 17β-estradiol is the most potent one as has the highest affinity for its receptors. It is produced in high amounts in pre-menopausal women by the ovary. The second endogenous but less potent estrogen is estrone. It is produced from androstenedione, the immediate precursor of estrone. The third estrogen is estriol, a metabolite of estradiol. It is mainly produced by the placenta during pregnancy and is found in lower concentrations than estradiol and estrone in non-pregnant women (Chen et al., 2008). Estrogens act through the estrogen receptors (ERs). ERs are ligandactivated transcription factors that have several domains that can bind estrogens and activate transcription of several estrogen-responsive genes (see Figure 1) (Notarnicola et al., 2001). There are two receptor isoforms, ERα and ERβ (Tsai & O Malley, 1994; Hall et al, 2001). When estrogen binds to these receptors, the receptors dimerize, go to the nucleus and bind to specific DNA sequences, the consensus estrogen response elements (EREs) of ER-responsive genes (Klein- Hitpass et al., 1989). The receptors may form ERα (αα) or ERβ (ββ) homodimers or ERαβ (αβ) heterodimers (Li X et al., 2004). The activation of ER is influenced by a set of different co-activators, enzymes, and co-repressors. These factors influence the assembly of the transcriptional complex and the subsequent transcription of the ER-responsive genes. This is called the canonical pathway of ER. A non-canonical pathway of ER has also been described, in which genes are activated without having ERE-like sequences. This non-classical mechanism accounts for the transcriptional activation of approximately one-third of all estrogen responsive genes (Huang et al., 2004). 79

81 Alternative mechanisms without DNA binding have also been described. DNA binding proteins such as specificity protein 1 (SP1) are activated by the direct binding of ER (Velarde et al., 2007), for a schematic representation see Figures 1 and 3. Estrogens play a major role in controlling the menstrual cycle, pregnancy, thus female reproduction. However, estrogens are not only important for the biology of hormonally responsive tissues; they also play an important role in bone strengthening and cholesterol metabolism, and have an influence on the central nervous system and the gastrointestinal physiology (Roy & Liehr, 1999; Nilsson & Gustafsson, 2001). On one hand ER signaling plays an important role in many normal physiological processes, on the other hand several studies have shown that estrogens and their metabolites are also involved in tumor development. In this review we will address the different possible mechanisms by which estrogens can be involved in tumor development and in particular, we will focus on how the hormone can be involved in the development of Lynch syndrome-related neoplasias showing microsatellite instability. Estrogens Non-genomic Genomic Membrane or cytoplasmic ERs Non ER-membrane bound receptors/ adapters Estrogen Response Elements (ERE) dependent signaling regulated by coactivators and repressors ERE (Nuclear) ERs Estrogen response elements independent signaling by binding to DNA bound transcription factors directly Activation of signaling routes such as: MAPK, P13K, PKA,PKC / binding of transcription factors to the active ERs Binding of coactivators and co-repressors (e.g. NRIP1) Transcription factors (e.g. AP-1 / PS2) Proliferation Growth Differentation Angiogenesis Apoptosis Target gene expression (e.g. VEGF) Figure 1. Mechanisms of action of estrogens. 80