PDF Hosted at the Radboud Repository of the Radboud University Nijmegen

PDF hosted at the Radboud Repository of the Radboud University Nijmegen

The following full text is a publisher's version.

For additional information about this publication click this link. http://hdl.handle.net/2066/49938

Please be advised that this information was generated on 2021-10-01 and may be subject to change.

Cover design: Mieke Veltman-Hoedemaker ISBN: 90-9020104-1

Genetic analysis of neonatal and infantile germ cell tumours

Een wetenschappelijke proeve op het gebied van de Medische Wetenschappen

Proefschrift

ter verkrijging van de graad van doctor aan de Radboud Universiteit Nijmegen op gezag van de Rector Magnificus prof. dr. C.W.P.M. Blom, volgens besluit van het College van Decanen in het openbaar te verdedigen op woensdag 18 januari 2006 des namiddags om 1.30 uur precies

door

Imke Marike Veltman geboren op 16 juni 1976 te Hoensbroek Promotor: Prof dr. A. Geurts van Kessel

Manuscriptcommissie: Prof. dr. L.A.L.M Kiemeney, voorzitter Prof. dr. L.H.J. Looijenga (Erasmus MC-Universitair Medisch Centrum Rotterdam) Dr. W.J.A.J. Hendriks

Contents

List of abbreviations

Chapter 1 General introduction 11 APMIS (2003) 111: 152-160

Chapter 2 A novel case of infantile sacral teratoma and a 32 constitutional t(12;15)(q13;q25) pat Cancer Genetics Cytogenetics (2002) 136:17-22

Chapter 3 Fusion of the SUMO/Sentrin-specific protease 1 44 gene SENP1 and the embryonic polarity-related mesoderm development gene MESDC2 in a patient with an infantile teratoma and a constitutional t(12;15)(q13;q25) Human Molecular Genetics (2005) 14: 1955-1963

Chapter 4 Positional cloning of a constitutional t(8;12)(q23;q12) 64 in a patient with an infantile sacral teratoma

Chapter 5 Chromosomal breakpoint mapping by array CGH 78 using flow-sorted chromosomes Biotechniques (2003) 35:1066-70

Chapter 6 Identification of recurrent chromosomal aberrations 90 in germ cell tumors of neonates and infants using genomewide array-based comparative genomic hybridization Genes, Chromosomes & Cancer (2005) 43:367–376

Chapter 7 General discussion and future prospects 111

Colour section 124 Summary / Samenvatting 129 Dankwoord en curriculum vitae 139

LIST OF ABBREVIATIONS array CGH microarray-based comparative genomic hybridization CGH comparative genomic hybridization der(8) derivative 8 chromosome from the t(8;12)(q23;q12) der(12) derivative 12 chromosome from the t(12;15)(q13;q25) or t(8;12)(q23;q12) der(15) derivative 15 chromosome from the t(12;15)(q13;q25) EBV Epstein-Barr virus ER endoplasmatic reticulum FISH fluorescent in situ hybridization GCT(s) germ cell tumour(s) IP immunoprecipitation KCT kiemceltumor LCV large-scale copy number variation LOH loss of heterozygosity miRNAs microRNAs; small, functional non-coding RNAs NS non-seminoma ORF open reading frame PGC(s) primordial germ cell(s) 3’-RACE 3’-rapid amplification of cDNA ends RT-PCR reverse transcription-polymerase chain reaction SE seminoma TE teratoma TGCT(s) testicular germ cell tumour(s) WT wild-type YST(s) yolk sac tumour(s)

Genes and proteins involved in t(12;15)(q13;q25) MESDC2 MESDC2 gene (mesoderm development gene) MESDC2 MESDC2 protein MESDC2-SENP1 fusion gene of MESDC2 and SENP1 MESE protein derived from MESDC2-SENP1 SEME protein derived from SENP1-MESDC2 SENP1 SENP1 gene (SUMO/Sentrin specific protease 1) SENP1 SENP1 protein SENP1-MESDC2 fusion gene of SENP1 and MESDC2

General introduction

Adapted from: Germ cell tumours in neonates and infants: a distinct subgroup?

Imke M. Veltman Marga Schepens, Leendert H.J. Looijenga,

Louise C. Strong, APMIS (2003) 111: 152-160 Ad Geurts van Kessel

General introduction

GENETIC ALTERATIONS AND CANCER

In general cancer is caused by alterations in the control and activity of genes that regulate cellular growth, differentiation, death, adhesion and/or genomic stability. Overall, cancer is not caused by a single gene defect. Instead, an accumulation of multiple gene defects is required for a normal cell to progress through different stages of cancer initiation and progression. Thus, cancer development is a multi-step process. The genes involved in this process include so-called proto-oncogenes, tumour suppressor genes and genomic stability genes [1]. During tumourigenesis proto-oncogenes, which normally stimulate cellular growth, are hyper activated or activated under conditions in which the wild-type gene would not be activated. In contrast, tumour suppressor genes, normally involved in repression of cellular growth, are inactivated. Genomic stability genes, including DNA repair genes, are genes that normally limit the amount of genomic alterations. Inactivation of these latter genes results in a higher mutation rate in other genes (for example growth controlling genes). These gene alterations can occur at the genetic level (see below), including somatic and inherited changes and acquired transforming DNA or RNA viral sequences, or at the epigenetic level, i.e., DNA methylation, histon modification and/or chromatin remodelling [2-4]. Therefore, extensive research has been aimed at studying (epi)genetic defects in tumour cells, ranging from cytogenetic research at the chromosomal level to molecular research at the basepair level. The first tumour-specific chromosomal alteration was detected in 1960, i.e., the Philadelphia chromosome in human chronic myeloid leukaemia (CML) [5]. This Philadelphia chromosome is the product of a translocation between chromosomes 9 and 22, leading to disruption and fusion of the ABL and BCR genes, respectively. As a result of this fusion, the proto-oncogene ABL is activated. The discovery of this gene fusion has led to the development of one of the first effective targeted cancer therapies, using imatinib [6]. Imatinib is a tyrosine kinase inhibitor that blocks the tyrosine kinase activity of the oncogenic ABL protein. Next to balanced genomic changes such as chromosomal translocations, also numerical genomic changes resulting from gains or losses of specific chromosomal regions have been implicated in tumorigenesis. An explicit example of the latter is gain of chromosome 12p sequences in almost all adult testicular germ cell tumours [7]. Gain of (parts of) 12p has been suggested to play a role in progression of these tumours [8;9]. However, its exact role remains to be established and several candidate genes are currently under investigation [reviewed in 10]. Another example of a numerical genomic change is a

11 Chapter 1 homozygous deletion within the 10q23 region that led to the discovery of the PTEN (MMAC1) gene [11;12]. Loss of function of PTEN appears to be associated with advanced stages of multiple human cancers [13] and is associated with two hereditary cancer syndromes with overlapping clinical features, i.e., Cowden syndrome and Bannayan-Riley-Ruvalcaba syndrome [14]. Besides somatic chromosomal aberrations, also inherited or constitutional chromosomal aberrations have been instrumental for the identification of genes involved in the development of various types of cancer. Specific constitutional chromosomal deletions, for example, have led to the identification of the RB1 gene in retinoblastoma [15-18], the WT1 gene in Wilms’ tumour [18;19] , the VHL gene in renal cell cancer [20] and the APC gene in familial adenomatous polyposis [21-24]. In addition, inherited or constitutional chromosomal translocations have led to the identification of e.g. the FHIT gene in renal cell carcinoma [25], the WAVE3 gene in ganglioneuroblastoma [26], the CDC91L1 gene in bladder cancer [27], the ETV6- NTRK3 fusion gene in fibrosarcoma [28;29], and the NB4S gene in neuroblastoma [30]. These examples clearly illustrate that detailed knowledge on the genetic constitution of cells that are associated with tumour development is of major importance for the identification of the genes underlying this process. In addition, it has been shown that tumour-associated genetic alterations may be used for optimizing diagnostic, prognostic and therapeutic protocols. The (still ongoing) unravelling of the human genome, both at the sequence content level and at the technological level, greatly facilitates cancer research [31;32]. One particular technological advance, the development of DNA microarrays, has become invaluable for the comprehensive deciphering of cancer cells at both the genome (DNA) and the transcriptome (RNA) level [33-35].

GERM CELL DEVELOPMENT

Germ cells are the precursors of sperm and egg cells that are responsible for the transmission of genetic information (after meiotic recombination) from one generation to the next. During embryonic development the stem cells of the germ cell lineage, the primordial germ cells (PGCs), are specified outside the embryo and, subsequently, these cells migrate to the site of the future gonads (ovary or testis). The

12 General introduction current knowledge on the origin of the germ cell lineage as described below is mostly based on mouse studies. Development of an embryo from a one cell stage to a fully developed organism occurs in several predefined stages. An important stage is the gastrulation stage during which the three germinal layers are formed: ectoderm, mesoderm and endoderm. Eventually, these layers give rise to the different tissues of the embryo. During gastrulation proximal epiblast-derived cells migrate towards the primitive streak (Fig. 1A) and extend into the base of the future allantois (Fig. 1B and C) [36;37]. These cells are not yet restricted to the germ cell lineage and serve as precursors for extra-embryonic mesoderm, allantois and PGCs [38;39]. The first time that PGCs can be distinguished from other cell types, indicating that these PGCs are indeed germ cell lineage specific, is at the base of the allantois in the yolk sac (Fig. 1B). From this location the PGCs migrate along the hindgut to the genital ridges, the site of the future gonads (Fig. 1D). In case the PGCs arrive at an extragonadal site instead of at the genital ridges, they will die by BAX-dependent apoptosis [40].

Extra- Extra- embryonic embryonic mesoderm ectoderm Foregut Hindgut Genital Anterior Posterior Allantois ridge Allantois Hindgut bud

Genital Heart ridge Epiblast PGCs PGCs Primitive Visceral streak Primitive Yolk sac PGCs endoderm mesoderm streak AB C D

Figure 1. Model for the development and migration route of primordial germ cells in mouse embryos, adapted from Hogan et al. [41] and Sadler [42]. The epiblast will produce the embryo; extra-embryonic ectoderm and visceral endoderm are supportive tissues. A) Signals from the extra- embryonic ectoderm induce the proximal epiblast-derived cells to migrate towards the primitive streak and become common precursors of extra-embryonic mesoderm, allantois and primordial germ cells. B) Restriction to the germ cell lineage of a subgroup from the proximal epiblast-derived cells. C) Location of the primordial cells prior to migration into the embryo at the base of the allantois within the yolk sac. D) Migration route of the primordial germ cells through the embryo via the hindgut (posterior portion of the alimentary tract) into the newly forming genital ridges.

During migration, PGCs proliferate extensively from about 50-100 cells to approximately 50,000 cells at two days after entry into the genital ridges [43;44]. Shortly before and during the influx of PGCs, the genital ridges differentiate into indifferent gonads. Once the PGCs have arrived in the genital ridge, they become

13 Chapter 1 non-motile and are called gonocytes. Subsequently they enter the meiotic prophase (female gonocytes) or mitotic arrest (male gonocytes). The meiotic entry of the female gonocytes (now designated as oocytes) appears to be cell-autonomous as germ cells enter the meiotic prophase at about the same time, even if these cells are located outside the genital ridge [45]. In contrast, the entry into the spermatogenic pathway is induced by SRY, encoded by the Y chromosome, and several other factors [46]. Differentiation of the female oocytes is arrested at the last stage of meiotic prophase I and does not resume until the time of ovulation. Spermatogenesis, however, does not start before puberty and begins with a series of mitotic divisions followed by two successive meiotic divisions resulting in the formation of spermatids and spermatozoa. Interestingly, the oocyte pathway is considered the default pathway for all germ cells [39;45]. For a long time mammalian germ cells were thought to originate from predetermined cytoplasm, similar to germ cells of frogs, flies and worms. However, the mammalian germ cell lineage appeared to be induced by external signals from neighbouring cells [reviewed in 39]. These external signals include several extra- embryonic ectoderm-derived factors, such as the bone morphogenetic proteins BMP2, BMP4 and BMP8B [47-49]. During the processes of germ cell specification, migration and homing to the genital ridges many additional factors are involved that are either expressed by the (primordial) germ cells themselves or by their neighbouring cells. Several of these factors are currently under investigation [39;50;51].

HUMAN GERM CELL TUMOURS

The neoplastic counterparts of human germ cells are the human germ cell tumours (GCTs). These tumours constitute a heterogeneous group of neoplasms that can develop at any age, in both sexes, in the gonads and at several extragonadal sites along the midline of the body. Also, the clinical outcomes of GCTs are diverse, ranging from benign teratomas to highly malignant and clinically aggressive tumours such as embryonal carcinomas and choriocarcinomas. Phenotypically, one or more structures that appear during embryonic development are represented in these tumours with their germ cell (seminomas), embryonic (teratomas, embryonal carcinomas) and extra-embryonic (choriocarcinomas, yolk sac tumours) morphologies.

14 General introduction

Based on anatomical location, phenotype, age, and genotype the GCTs can be divided into five types: type I GCTs existing of teratomas and yolk sac tumours that primarily occur in neonates and infants, type II GCTs occurring as seminomatous and non-seminomatous GCTs; the seminomatous tumours are called seminomas in the testis, dysgerminomas in the ovary or dysgenetic gonads and germinomas in the brain, type III GCTs, the spermatocytic seminomas in elderly men, type IV GCTs, the dermoid cysts (mature cystic teratomas) of the ovary, and, recently added to this list, type V GCTs, the hydatidiform mole of the placenta/uterus [10] (Table 1). The recurrence risk of type I GCTs depends on primary site (with the coccygeal tumours being most at risk), histological grade, and completeness of the primary resection including the adjacent organ of origin (coccyx, ovary, testis etc.) [52;53]. The mature teratomas in this group are mostly benign and can be cured by surgery alone, however incomplete resection of teratomas imposes a risk for relapse [52]. In contrast to most metastatic cancers, metastatic testicular GCTs (TGCTs, type II) exhibit a high cure rate. Over 80% of patients with a metastatic TGCT can be cured using cisplatin-based combination chemotherapy [54]. Even when the tumour is spread to distant organs (such as lung) 74% of the TGCT patients have a 5-year survival rate [55]. Most spermatocytic seminomas (type III) and dermoid cysts (type IV) can be cured by resection. Chromosomal aberrations have frequently been found in GCTs (Table 1). For the adult testicular type II GCTs for example, gross numerical and structural abnormalities have been described by several groups [56-58]. In addition, these tumours show a consistent gain of 12p, either via the generation of one or more copies of an i(12p) chromosome or via alternative ways, including high copy number amplifications of defined regions of 12p [57-63], reviewed in [7]. Due to the low incidence of spermatocytic seminomas, these tumours are not very well characterised at the cytogenetic level. The few spermatocytic seminomas that have been examined (type III) were mainly aneuploid [64;65]. In addition, one study showed recurrent gains of chromosome 9, suggesting that this aberration may be characteristic for this latter subgroup [65]. Karyotypes from dermoid cysts of the ovary range from diploid, diploid/tetraploid to near triploid [66;67].

15 Chapter 1

Table 1. Characteristics of the five germ cell tumour types

Type Anatomical location Phenotype Originating cell Age DNA ploidy

I testis/ ovary/ sacral (immature) early PGC/ neonates and teratoma: diploid, region/ teratoma/ yolk gonocyte infants yolk sac tumour: retroperitoneum/ sac tumour aneuploid mediastinum/ neck/ midline brain/ other rare sites II testis seminoma/ early PGC/ >15 years Aneuploid (+/- non-seminoma gonocyte triploid) ovary dysgerminoma/ early PGC/ > 4 years aneuploid non-seminoma gonocyte dysgenetic gonad dysgerminoma/ early PGC/ congenital diploid/tetraploid non-seminoma gonocyte anterior mediastinum seminoma/ early PGC/ adolescents diploid/tri-tetraploid (thymus) non-seminoma gonocyte midline brain (pineal germinoma/ early PGC/ children (median diploid/tri-tetraploid gland/hypothalamus) non-seminoma gonocyte age 13 years) III testis spermatocytic spermatogonium/ > 50 years aneuploid seminoma spermatocyte IV ovary dermoid cyst oogonia/ oocyte children/adults (near) diploid, diploid/tetraploid, ~triploid V placenta/uterus hydatidiform empty ovum/ fertile period diploid mole spermatozoa Adapted from Oosterhuis and Looijenga [10]

GERM CELL TUMOURS IN NEONATES AND INFANTS

This thesis focuses on the type I GCTs of neonates and infants that comprises the predominantly benign teratomas and their malignant counterparts, the yolk sac tumours. The incidence of these type I GCTs is about 0.12 per 100,000 [68]. GCTs of neonates and infants differ in many aspects from the adult and elderly GCTs and it has been suggested [67;69-71] that they may constitute a separate pathogenetic subgroup based on site of occurrence, histology, clinical behaviour, cell of origin, and cytogenetics.

Occurrence and histology of infantile GCTs The majority of GCTs that occur under the age of 5 years develop at extragonadal sites, whereas more than 90% of GCT in adults develop in the gonads [53;70;72]. The most common extragonadal site for type I GCTs is the sacrococcygeal region [70]. The site of occurrence is dependent on age as well as on histology. Teratomas, for example, present as sacrococcygeal tumours at birth, whereas later in life they are

16 General introduction mainly present as gonadal tumours. In the ovaries teratomas predominantly occur between the ages of 9 to 15 years. Also at the histological level adult GCTs and infantile GCTs are distinct. Infantile GCTs mostly present as teratomas or yolk sac tumours (YSTs) [72-75], whereas adult GCTs mostly present as seminomas or non-seminomas. Interestingly, these non-seminomas can also contain teratoma and/or yolk sac tumour components. Genetically however, these components are quite different from their infantile histological counterparts, as they are predominantly aneuploid and typically contain a marker chromosome: i(12p). A study of 780 children (<15 years) with GCTs showed that age is a critical factor for the type of histology [52]. In this cohort the median age of children with a teratoma was 5 months (n=329). After the age of 5 months the main histology is shifted to YSTs with a median age of 2 years (n=260). Embryonal carcinomas, seminoma-like tumours and choriocarcinomas occur mainly later in life, after the age of 8 years. These observations were confirmed in an additional study of 1,157 childhood GCTs (<15 years) [70]. Type I teratomas may give rise to secondary malignancies [76]. Specifically, immature teratomas of the sacral region and the neck have been reported to recur as YST [72;77-79]. This notion is supported by animal models, i.e., the occurrence of teratomas and the recurrence of YSTs has also been reported in mice [80-82]. Based on observations that YSTs may be aneuploid and teratomas may be diploid (see below), the development of immature teratomas into YSTs is suggested to be accompanied by cytogenetic changes [83;84].

The cell of origin: different for infantile and adult GCTs? The term germ cell tumour implies that all subtypes develop from a common precursor cell. So far, this cell is thought to be the primordial germ cell (PGC). The occurrence of extragonadal GCTs may be explained by an aberrant migration pattern of PGCs during embryogenesis [85-87]. It cannot be excluded, however, that these cells may have a function at these extragonadal sites. The germ cell origin of some of the GCT subtypes is supported by the expression of the germ line-specific protein VASA or their specific pattern of genomic imprinting [10;88-91] (see below). In addition, these studies suggested that the different subtypes may not originate from a common precursor cell, but from germ cells of different developmental stages. Studies on the DNA content of infantile GCTs and adult TGCTs suggest that these two tumour types may have a different precursor cell. Adult TGCTs are usually aneuploid, with overall differences between non-seminomas and seminomas.

17 Chapter 1

Seminomas appear to be hypertriploid to tetraploid, whereas non-seminomas appear to be in the hypotriploid to peritriploid range (reviewed in [57;84;92;93]). Infantile sacral teratomas were found to be diploid, independent of the grade of immaturity [74;84;94]. Infantile YST were found to be aneuploid, although also diploid cases have been reported [74;95;96]. Next to ploidy differences between infantile and adult GCTs, there is also a disparity in the presence of carcinoma in situ (CIS) in these tumours. For adult TGCTs, CIS lesions are believed to provide the precursor cells [97-99]. For the other subtypes, however, this scenario has been questioned. Except for some isolated reports [100-102], infantile GCTs do not appear to be associated with CIS, at least not at a morphological level [103-105]. This again implies a difference in origin between adult and infantile types of GCT. Recently, Schneider et al. [88] reported a multipoint imprinting analysis of gonadal and extragonadal infantile GCTs. The term imprinting indicates that alleles of paternal and maternal origin are differentially expressed. The inherited imprinting status is erased in PGCs, supposedly during the migration from the yolk sac to the genital ridges, and may be imparted before they enter meiosis (reviewed in [106]). However, the exact time frame in which this occurs is not yet known. Schneider et al [88] determined the imprinting status of three well-known imprinted genes, H19, IGF-2 and SNRPN in a group of 42 infantile GCTs. By doing so, they deduced whether the cell of origin is a PGC before the entry into the genital ridges (PGC has the imprinting status of the original cell), or a premeiotic PGC with an erased imprinting status, which is measured by a biallelic expression of H19 and IGF-2 and no methylation of the imprinting control region (ICR) of the H19 or the SNRPN genes (loss of imprinting (LOI) of H19, IGF-2 and SNRPN). The SNRPN alleles turned out to be unmethylated in both gonadal and extragonadal infantile GCTs. In addition, gonadal and extragonadal infantile GCTs showed similar but lower frequencies of LOI of the H19 and IGF-2 genes [88;107] as compared with adult TGCTs [90]. Because the imprinting pattern of the latter genes differed between the infantile and adult GCTs, they suggested that initiation of at least some infantile GCTs may occur during germ cell migration, whereas the initiation of adult GCTs may occur at later stages of PGC development. Another possibility is that failure to erase the H19 and IGF-2 imprinting status during PGC migration may in fact reflect an early step in tumorigenesis.

18 General introduction

The differences in cell of origin and genomic imprinting pattern between the different germ cell tumour subtypes were recently reviewed by Oosterhuis and Looijenga [10] (Table 1). They suggested as cell of origin an early PGC/gonocyte for the type I tumours (with a biparental, partially erased imprinting pattern), a PGC/gonocyte for the type II tumours (with an erased imprinting pattern), a spermatogonium/spermatocyte for the type III tumours (with a partial paternal imprinting pattern) and an oogonium/oocyte for the type IV tumours (with a partial maternal imprinting pattern).

Recurrent chromosomal aberrations in infantile GCTs Next to the recurrent presence of an i(12p) marker chromosome and gain of parts of 12p in gonadal and extragonadal type II TGCTs, other frequent chromosomal aberrations have been encountered, such as specific gains for the chromosomes 7, 8, 12, and X and specific losses for the chromosomes 11, 13, 18, and Y [92;108;109], reviewed in [7;58]. For infantile GCTs the overall cytogenetic picture is different although these tumours have not been studied as extensively as the type II GCTs. The i(12p) marker chromosome is not a common feature in infantile GCTs and specific gains and losses as described for adult TGCTs are rare. Instead, the most prevalent chromosomal aberrations found in infantile GCTs are losses of 1p (more specific: 1p36) and 6q [69;71;79;95;110-115]. In addition, abnormalities of chromosome 3 (mainly gain of 3p), gains of 1q and 20q have been described as recurrent aberrations, as well as losses of (parts of) chromosome 4 [69;79;88;95;113;115;116]. All these aberrations were detected in malignant infantile GCTs, mainly YSTs with or without a teratoma component. Loss of 1p and loss of 6q have also been detected in several cases of adult TGCT. These losses were found in a lower frequency as compared to the infantile GCTs and they were observed mostly in non-seminomas, especially in tumours containing YST or embryonal carcinoma (EC) components [58;109;117-120]. Bussey et al. [111] suggested that loss of 1p is related to malignancy independent of the age of the patient. However, in their study most patients under the age of 5 years had teratomas and only a few had YSTs or a mixed tumour of YST and teratoma. So far, most chromosomal aberrations of infantile patients have been detected in the YST components. The observation that 1p and 6q losses occur in infantile GCTs with a YST component and that these losses occur in adult TGCTs containing YST or EC components, suggests that these anomalies may be related to the biological behaviour of these malignant tumour

19 Chapter 1

(sub)types. This phenomenon may be related to age as most 1p and 6q losses are observed in YSTs of infants. In contrast to the YSTs, most studies of benign infantile teratomas have indicated that they are diploid with normal karyotypes. In addition, in a classical CGH study using microdissected teratoma material no chromosomal imbalances were found. This latter observation may indicate that the absence of chromosomal aberrations in the cytogenetic studies is not due to cell culture artefacts [79]. Still, genetic aberrations may be too small to be detected with classical CGH or karyotyping methods. In a few cases, however, chromosomal aberrations were detected, including 1q abnormalities [121-124]. It is unknown whether these teratomas also contained YST components.

OBJECTIVES

In comparison with the most prevalent type of GCTs, i.e., TGCT of type II, information on the genetic constitution of the type I GCTs has remained scarce so far. The overall objectives of this thesis were to identify candidate genes that play a role in one or more of the developmental stages of type I GCTs and to identify chromosomal anomalies that may be used as diagnostic/prognostic markers for type I GCTs. Therefore, two approaches were employed: (1) positional cloning of two constitutional chromosome translocations in patients with type I GCTs and (2) genomic microarray analysis of acquired chromosomal aberrations in a selected group of sporadic type I GCTs.

The two constitutional chromosome translocations studied in this thesis, t(12;15)(q13;q25) and t(8;12)(q23;q12), are of particular interest since they both exhibit breakpoints in 12q11-q13, a suggested region with candidate genes for the development of human GCTs [116;117;123]. The motive for studying these two translocations was the initial mapping of both chromosome 12 breakpoints to the 12q13 region, suggesting that these breakpoints may disrupt the same gene within this region. The positional cloning of the respective translocation breakpoints and the identification of its corresponding genes are described in chapters 2, 3, and 4. Microarray-based comparative genomic hybridization (array CGH) was used to study the complete genomes of type I GCTs for copy number changes. Due to the relative high resolution of this technique, it allowed the direct identification of genomic regions (1) relevant for diagnosis and/or prognosis of type I GCTs and (2)

20 General introduction containing candidate genes relevant for type I GCT development (chapter 6). In addition, array CGH was used in conjunction with chromosome flow-sorting for the direct mapping of chromosomal breakpoints (chapter 5). In chapter 7 the implications of these studies for our understanding of type I GCT development are discussed and future prospects are provided.

REFERENCES

1. Vogelstein B, Kinzler KW (2004): Cancer genes and the pathways they control. Nat.Med. 10:789-799. 2. Laird PW (2005): Cancer epigenetics. Hum.Mol.Genet. 14 Suppl 1:R65-R76. 3. Klein G (2005): Epigenetics: surveillance team against cancer. Nature 434:150. 4. Fraga MF, Ballestar E, Villar-Garea A, Boix-Chornet M, Espada J, Schotta G, Bonaldi T, Haydon C, Ropero S, Petrie K et al.(2005): Loss of acetylation at Lys16 and trimethylation at Lys20 of histone H4 is a common hallmark of human cancer. Nat.Genet. 37:391-400. 5. Nowell PC, Hungerford DA (1960): Chromosome studies on normal and leukemic human leukocytes. J.Natl.Cancer Inst. 25:85-109. 6. Druker BJ, Talpaz M, Resta DJ, Peng B, Buchdunger E, Ford JM, Lydon NB, Kantarjian H, Capdeville R, Ohno-Jones S, Sawyers CL (2001): Efficacy and safety of a specific inhibitor of the BCR-ABL tyrosine kinase in chronic myeloid leukemia. N.Engl.J.Med. 344:1031-1037. 7. Looijenga LH, Zafarana G, Grygalewicz B, Summersgill B, Debiec-Rychter M, Veltman J, Schoenmakers EF, Rodriguez S, Jafer O, Clark J, et al. (2003): Role of gain of 12p in germ cell tumour development. APMIS 111:161-171. 8. Geurts van Kessel A, van Drunen E, de Jong B, Oosterhuis JW, Langeveld A, Mulder MP (1989): Chromosome 12q heterozygosity is retained in i(12p)-positive testicular germ cell tumor cells. Cancer Genet.Cytogenet. 40:129-134. 9. Rosenberg C, Van Gurp RJ, Geelen E, Oosterhuis JW, Looijenga LH (2000): Overrepresentation of the short arm of chromosome 12 is related to invasive growth of human testicular seminomas and nonseminomas. Oncogene 19:5858-5862. 10. Oosterhuis JW, Looijenga LH (2005): Testicular germ-cell tumours in a broader perspective. Nat.Rev.Cancer 5:210-222. 11. Li J, Yen C, Liaw D, Podsypanina K, Bose S, Wang SI, Puc J, Miliaresis C, Rodgers L, McCombie R, et al. (1997): PTEN, a putative protein tyrosine phosphatase gene mutated in human brain, breast, and prostate cancer. Science 275:1943-1947. 12. Steck PA, Pershouse MA, Jasser SA, Yung WK, Lin H, Ligon AH, Langford LA, Baumgard ML, Hattier T, Davis T, et al. (1997): Identification of a candidate tumour suppressor gene, MMAC1, at chromosome 10q23.3 that is mutated in multiple advanced cancers. Nat.Genet. 15:356-362. 13. Osaki M, Oshimura M, Ito H (2004): PI3K-Akt pathway: its functions and alterations in human cancer. Apoptosis. 9:667-676. 14. Bonneau D, Longy M (2000): Mutations of the human PTEN gene. Hum.Mutat. 16:109-122.

21 Chapter 1

15. Knudson AG, Jr. (1978): Retinoblastoma: a prototypic hereditary neoplasm. Semin.Oncol. 5:57- 60. 16. Friend SH, Bernards R, Rogelj S, Weinberg RA, Rapaport JM, Albert DM, Dryja TP (1986): A human DNA segment with properties of the gene that predisposes to retinoblastoma and osteosarcoma. Nature 323:643-646. 17. Lee WH, Bookstein R, Hong F, Young LJ, Shew JY, Lee EY (1987): Human retinoblastoma susceptibility gene: cloning, identification, and sequence. Science 235:1394-1399. 18. Fung YK, Murphree AL, T'Ang A, Qian J, Hinrichs SH, Benedict WF (1987): Structural evidence for the authenticity of the human retinoblastoma gene. Science 236:1657-1661. 19. Haber DA, Buckler AJ, Glaser T, Call KM, Pelletier J, Sohn RL, Douglass EC, Housman DE (1990): An internal deletion within an 11p13 zinc finger gene contributes to the development of Wilms' tumor. Cell 61:1257-1269. 20. Latif F, Tory K, Gnarra J, Yao M, Duh FM, Orcutt ML, Stackhouse T, Kuzmin I, Modi W, Geil L, . (1993): Identification of the von Hippel-Lindau disease tumor suppressor gene. Science 260:1317-1320. 21. Herrera L, Kakati S, Gibas L, Pietrzak E, Sandberg AA (1986): Gardner syndrome in a man with an interstitial deletion of 5q. Am.J.Med.Genet. 25:473-476. 22. Bodmer WF, Bailey CJ, Bodmer J, Bussey HJ, Ellis A, Gorman P, Lucibello FC, Murday VA, Rider SH, Scambler P, . (1987): Localization of the gene for familial adenomatous polyposis on chromosome 5. Nature 328:614-616. 23. Groden J, Thliveris A, Samowitz W, Carlson M, Gelbert L, Albertsen H, Joslyn G, Stevens J, Spirio L, Robertson M, . (1991): Identification and characterization of the familial adenomatous polyposis coli gene. Cell 66:589-600. 24. Kinzler KW, Nilbert MC, Su LK, Vogelstein B, Bryan TM, Levy DB, Smith KJ, Preisinger AC, Hedge P, McKechnie D, . (1991): Identification of FAP locus genes from chromosome 5q21. Science 253:661-665. 25. Ohta M, Inoue H, Cotticelli MG, Kastury K, Baffa R, Palazzo J, Siprashvili Z, Mori M, McCue P, Druck T, . (1996): The FHIT gene, spanning the chromosome 3p14.2 fragile site and renal carcinoma-associated t(3;8) breakpoint, is abnormal in digestive tract cancers. Cell 84:587-597. 26. Sossey-Alaoui K, Su G, Malaj E, Roe B, Cowell JK (2002): WAVE3, an actin-polymerization gene, is truncated and inactivated as a result of a constitutional t(1;13)(q21;q12) chromosome translocation in a patient with ganglioneuroblastoma. Oncogene 21:5967-5974. 27. Guo Z, Linn JF, Wu G, Anzick SL, Eisenberger CF, Halachmi S, Cohen Y, Fomenkov A, Hoque MO, Okami K, et al. (2004): CDC91L1 (PIG-U) is a newly discovered oncogene in human bladder cancer. Nat.Med. 10:374-381. 28. Wai DH, Knezevich SR, Lucas T, Jansen B, Kay RJ, Sorensen PH (2000): The ETV6-NTRK3 gene fusion encodes a chimeric protein tyrosine kinase that transforms NIH3T3 cells. Oncogene 19:906-915. 29. Knezevich SR, McFadden DE, Tao W, Lim JF, Sorensen PH (1998): A novel ETV6-NTRK3 gene fusion in congenital fibrosarcoma. Nat.Genet. 18:184-187. 30. Roberts T, Chernova O, Cowell JK (1998): NB4S, a member of the TBC1 domain family of genes, is truncated as a result of a constitutional t(1;10)(p22;q21) chromosome translocation in a patient with stage 4S neuroblastoma. Hum.Mol.Genet. 7:1169-1178.

22 General introduction

31. Venter JC, Adams MD, Myers EW, Li PW, Mural RJ, Sutton GG, Smith HO, Yandell M, Evans CA, Holt RA, et al. (2001): The sequence of the human genome. Science 291:1304-1351. 32. Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, Baldwin J, Devon K, Dewar K, Doyle M, FitzHugh W, et al. (2001): Initial sequencing and analysis of the human genome. Nature 409:860-921. 33. Carpenter AE, Sabatini DM (2004): Systematic genome-wide screens of gene function. Nat.Rev.Genet. 5:11-22. 34. Pinkel D, Albertson DG (2005): Array comparative genomic hybridization and its applications in cancer. Nat.Genet. 37 Suppl:S11-S17. 35. Fire A, Xu S, Montgomery MK, Kostas SA, Driver SE, Mello CC (1998): Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 391:806-811. 36. Gardner RL, Rossant J (1979): Investigation of the fate of 4-5 day post-coitum mouse inner cell mass cells by blastocyst injection. J.Embryol.Exp.Morphol. 52:141-152. 37. McLaren A (1983): Primordial germ cells in mice. Bibl.Anat. 24:59-66. 38. Lawson KA, Hage WJ (1994): Clonal analysis of the origin of primordial germ cells in the mouse. Ciba Found.Symp. 182:68-84. 39. McLaren A (2003): Primordial germ cells in the mouse. Dev.Biol. 262:1-15. 40. Stallock J, Molyneaux K, Schaible K, Knudson CM, Wylie C (2003): The pro-apoptotic gene Bax is required for the death of ectopic primordial germ cells during their migration in the mouse embryo. Development 130:6589-6597. 41. Hogan B (2002): Developmental biology: decisions, decisions! Nature 418:282-283. 42. TW Sadler (2004): Langman's Medical Embryology, 9th edition. 43. Donovan PJ, Stott D, Cairns LA, Heasman J, Wylie CC (1986): Migratory and postmigratory mouse primordial germ cells behave differently in culture. Cell 44:831-838. 44. Tres LL, Rosselot C, Kierszenbaum AL (2004): Primordial germ cells: what does it take to be alive? Mol.Reprod.Dev. 68:1-4. 45. McLaren A (1995): Germ cells and germ cell sex. Philos.Trans.R.Soc.Lond B Biol.Sci. 350:229- 233. 46. Veitia RA, Salas-Cortes L, Ottolenghi C, Pailhoux E, Cotinot C, Fellous M (2001): Testis determination in mammals: more questions than answers. Mol.Cell Endocrinol. 179:3-16. 47. Lawson KA, Dunn NR, Roelen BA, Zeinstra LM, Davis AM, Wright CV, Korving JP, Hogan BL (1999): Bmp4 is required for the generation of primordial germ cells in the mouse embryo. Genes Dev. 13:424-436. 48. Ying Y, Zhao GQ (2001): Cooperation of endoderm-derived BMP2 and extraembryonic ectoderm-derived BMP4 in primordial germ cell generation in the mouse. Dev.Biol. 232:484- 492. 49. Ying Y, Qi X, Zhao GQ (2001): Induction of primordial germ cells from murine epiblasts by synergistic action of BMP4 and BMP8B signaling pathways. Proc.Natl.Acad.Sci.U.S.A 98:7858- 7862. 50. Molyneaux K, Wylie C (2004): Primordial germ cell migration. Int.J.Dev.Biol. 48:537-544. 51. Saitou M, Barton SC, Surani MA (2002): A molecular programme for the specification of germ cell fate in mice. Nature 418:293-300.

23 Chapter 1

52. Gobel U, Calaminus G, Engert J, Kaatsch P, Gadner H, Bokkerink JP, Hass RJ, Waag K, Blohm ME, Dippert S, et al. (1998): Teratomas in infancy and childhood. Med.Pediatr.Oncol. 31:8-15. 53. Gobel U, Schneider DT, Calaminus G, Haas RJ, Schmidt P, Harms D (2000): Germ-cell tumors in childhood and adolescence. GPOH MAKEI and the MAHO study groups. Ann.Oncol. 11:263-271. 54. Einhorn LH, Donohue J (1977): Cis-diamminedichloroplatinum, vinblastine, and bleomycin combination chemotherapy in disseminated testicular cancer. Ann.Intern.Med. 87:293-298. 55. Masters JR, Koberle B (2003): Curing metastatic cancer: lessons from testicular germ-cell tumours. Nat.Rev.Cancer 3:517-525. 56. Looijenga LH, Oosterhuis JW (1999): Pathogenesis of testicular germ cell tumours. Rev.Reprod. 4:90-100. 57. Sandberg AA, Meloni AM, Suijkerbuijk RF (1996): Reviews of chromosome studies in urological tumors. III. Cytogenetics and genes in testicular tumors. J.Urol. 155:1531-1556. 58. van Echten J, Oosterhuis JW, Looijenga LH, van de Pol M, Wiersema J, te Meerman GJ, Schaffordt KH, Sleijfer DT, de Jong B (1995): No recurrent structural abnormalities apart from i(12p) in primary germ cell tumors of the adult testis. Genes Chromosomes.Cancer 14:133-144. 59. Atkin NB, Baker MC (1982): Specific chromosome change, i(12p), in testicular tumours? Lancet 2:1349. 60. Geurts van Kessel A, Suijkerbuijk RF, Sinke RJ, Looijenga L, Oosterhuis JW, de Jong B (1993): Molecular cytogenetics of human germ cell tumours: i(12p) and related chromosomal anomalies. Eur.Urol. 23:23-28. 61. Suijkerbuijk RF, van de Veen AY, van Echten J, Buys CH, de Jong B, Oosterhuis JW, Warburton DA, Cassiman JJ, Schonk D, Geurts van Kessel A (1991): Demonstration of the genuine iso-12p character of the standard marker chromosome of testicular germ cell tumors and identification of further chromosome 12 aberrations by competitive in situ hybridization. Am.J.Hum.Genet. 48:269-273. 62. Suijkerbuijk RF, Sinke RJ, Meloni AM, Parrington JM, van Echten J, de Jong B, Oosterhuis JW, Sandberg AA, Geurts van Kessel A (1993): Overrepresentation of chromosome 12p sequences and karyotypic evolution in i(12p)-negative testicular germ-cell tumors revealed by fluorescence in situ hybridization. Cancer Genet.Cytogenet. 70:85-93. 63. Suijkerbuijk RF, Sinke RJ, Weghuis DE, Roque L, Forus A, Stellink F, Siepman A, van de KC, Soares J, Geurts van Kessel A (1994): Amplification of chromosome subregion 12p11.2-p12.1 in a metastasis of an i(12p)-negative seminoma: relationship to tumor progression? Cancer Genet.Cytogenet. 78:145-152. 64. Looijenga LH, Olie RA, van dG, I, van Sluijs FJ, Matoska J, Ploem-Zaaijer J, Knepfle C, Oosterhuis JW (1994): Seminomas of the canine testis. Counterpart of spermatocytic seminoma of men? Lab Invest 71:490-496. 65. Rosenberg C, Mostert MC, Schut TB, van de PM, van Echten J, de Jong B, Raap AK, Tanke H, Oosterhuis JW, Looijenga LH (1998): Chromosomal constitution of human spermatocytic seminomas: comparative genomic hybridization supported by conventional and interphase cytogenetics. Genes Chromosomes.Cancer 23:286-291. 66. Surti U, Hoffner L, Chakravarti A, Ferrell RE (1990): Genetics and biology of human ovarian teratomas. I. Cytogenetic analysis and mechanism of origin. Am.J.Hum.Genet. 47:635-643.

24 General introduction

67. Oosterhuis JW, Looijenga LH, van Echten J, de Jong B (1997): Chromosomal constitution and developmental potential of human germ cell tumors and teratomas. Cancer Genet.Cytogenet. 95:96-102. 68. Rescorla FJ (1999): Pediatric germ cell tumors. Semin.Surg.Oncol. 16:144-158. 69. Perlman EJ, Cushing B, Hawkins E, Griffin CA (1994): Cytogenetic analysis of childhood endodermal sinus tumors: a Pediatric Oncology Group study. Pediatr.Pathol. 14:695-708. 70. Schneider DT, Calaminus G, Koch S, Teske C, Schmidt P, Haas RJ, Harms D, Gobel U (2004): Epidemiologic analysis of 1,442 children and adolescents registered in the German germ cell tumor protocols. Pediatr.Blood Cancer 42:169-175. 71. van Echten J, Timmer A, van der Veen AY, Molenaar WM, de Jong B (2002): Infantile and adult testicular germ cell tumors. a different pathogenesis? Cancer Genet.Cytogenet. 135:57- 62. 72. Dehner LP (1983): Gonadal and extragonadal germ cell neoplasia of childhood. Hum.Pathol. 14:493-511. 73. Harms D, Janig U (1986): Germ cell tumours of childhood. Report of 170 cases including 59 pure and partial yolk-sac tumours. Virchows Arch.A Pathol.Anat.Histopathol. 409:223-239. 74. Hoffner L, Deka R, Chakravarti A, Surti U (1994): Cytogenetics and origins of pediatric germ cell tumors. Cancer Genet.Cytogenet. 74:54-58. 75. Valdiserri RO, Yunis EJ (1981): Sacrococcygeal teratomas: a review of 68 cases. Cancer 48:217- 221. 76. Isaacs H, Jr. (2004): Perinatal (fetal and neonatal) germ cell tumors. J.Pediatr.Surg. 39:1003- 1013. 77. Dehner LP, Mills A, Talerman A, Billman GF, Krous HF, Platz CE (1990): Germ cell neoplasms of head and neck soft tissues: a pathologic spectrum of teratomatous and endodermal sinus tumors. Hum.Pathol. 21:309-318. 78. Gilcrease MZ, Brandt ML, Hawkins EP (1995): Yolk sac tumor identified at autopsy after surgical excision of immature sacrococcygeal teratoma. J.Pediatr.Surg. 30:875-877. 79. Mostert M, Rosenberg C, Stoop H, Schuyer M, Timmer A, Oosterhuis J, Looijenga L (2000): Comparative genomic and in situ hybridization of germ cell tumors of the infantile testis. Lab Invest 80:1055-1064. 80. Stevens LC (1970): Experimental production of testicular teratomas in mice of strains 129, A/He, and their F1 hybrids. J.Natl.Cancer Inst. 44:923-929. 81. Sundstrom J, Pelliniemi LJ, Kuopio T, Verajankorva E, Frojdman K, Harley V, Salminen E, Pollanen P (1999): Characterization of the model for experimental testicular teratoma in 129/SvJ-mice. Br.J.Cancer 80:149-160. 82. van Berlo RJ, Oosterhuis JW, Schrijnemakers E, Schoots CJ, de Jong B, Damjanov I (1990): Yolk-sac carcinoma develops spontaneously as a late occurrence in slow- growing teratoid tumors produced from transplanted 7-day mouse embryos. Int.J.Cancer 45:153-155. 83. Baker BA, Frickey L, Yu IT, Hawkins EP, Cushing B, Perlman EJ (1998): DNA content of ovarian immature teratomas and malignant germ cell tumors. Gynecol.Oncol. 71:14-18. 84. Silver SA, Wiley JM, Perlman EJ (1994): DNA ploidy analysis of pediatric germ cell tumors. Mod.Pathol. 7:951-956.

25 Chapter 1

85. Chaganti RS, Rodriguez E, Mathew S (1994): Origin of adult male mediastinal germ-cell tumours. Lancet 343:1130-1132. 86. Hasle H, Jacobsen BB (1995): Origin of male mediastinal germ-cell tumours. Lancet 345:1046. 87. Wylie C (1999): Germ cells. Cell 96:165-174. 88. Schneider DT, Schuster AE, Fritsch MK, Hu J, Olson T, Lauer S, Gobel U, Perlman EJ (2001): Multipoint imprinting analysis indicates a common precursor cell for gonadal and nongonadal pediatric germ cell tumors. Cancer Res. 61:7268-7276. 89. Stoop H, van Gurp R, de Krijger R, Geurts van Kessel A, Koberle B, Oosterhuis W, Looijenga L (2001): Reactivity of germ cell maturation stage-specific markers in spermatocytic seminoma: diagnostic and etiological implications. Lab Invest 81:919-928. 90. Van Gurp RJ, Oosterhuis JW, Kalscheuer V, Mariman EC, Looijenga LH (1994): Biallelic expression of the H19 and IGF2 genes in human testicular germ cell tumors. J.Natl.Cancer Inst. 86:1070-1075. 91. Zeeman AM, Stoop H, Boter M, Gillis AJ, Castrillon DH, Oosterhuis JW, Looijenga LH (2002): VASA is a specific marker for both normal and malignant human germ cells. Lab Invest 82:159-166. 92. de Jong B, Oosterhuis JW, Castedo SM, Vos A, te Meerman GJ (1990): Pathogenesis of adult testicular germ cell tumors. A cytogenetic model. Cancer Genet.Cytogenet. 48:143-167. 93. Oosterhuis JW, Castedo SM, de Jong B, Cornelisse CJ, Dam A, Sleijfer DT, Schraffordt KH (1989): Ploidy of primary germ cell tumors of the testis. Pathogenetic and clinical relevance. Lab Invest 60:14-21. 94. Kruslin B, Visnjic A, Cizmic A, Tomicic I, Kos M, Jukic S, Seiwerth S (2000): DNA ploidy analysis and cell proliferation in congenital sacrococcygeal teratomas. Cancer 89:932-937. 95. Bussey KJ, Lawce HJ, Olson SB, Arthur DC, Kalousek DK, Krailo M, Giller R, Heifetz S, Womer R, Magenis RE (1999): Chromosome abnormalities of eighty-one pediatric germ cell tumors: sex- , age-, site-, and histopathology-related differences--a Children's Cancer Group study. Genes Chromosomes.Cancer 25:134-146. 96. Hoffner L, Shen-Schwarz S, Deka R, Chakravarti A, Surti U (1992): Genetics and biology of human ovarian teratomas. III. Cytogenetics and origins of malignant ovarian germ cell tumors. Cancer Genet.Cytogenet. 62:58-65. 97. Dieckmann KP, Skakkebaek NE (1999): Carcinoma in situ of the testis: review of biological and clinical features. Int.J.Cancer 83:815-822. 98. Rorth M, Rajpert-De Meyts E, Andersson L, Dieckmann KP, Fossa SD, Grigor KM, Hendry WF, Herr HW, Looijenga LH, Oosterhuis JW, Skakkebaek NE (2000): Carcinoma in situ in the testis. Scand.J.Urol.Nephrol.Suppl166-186. 99. Skakkebaek NE (1972): Possible carcinoma-in-situ of the testis. Lancet 2:516-517. 100. Hu LM, Phillipson J, Barsky SH (1992): Intratubular germ cell neoplasia in infantile yolk sac tumor. Verification by tandem repeat sequence in situ hybridization. Diagn.Mol.Pathol. 1:118- 128. 101. Renedo DE, Trainer TD (1994): Intratubular germ cell neoplasia (ITGCN) with p53 and PCNA expression and adjacent mature teratoma in an infant testis. An immunohistochemical and morphologic study with a review of the literature. Am.J.Surg.Pathol. 18:947-952.

26 General introduction

102. Stamp IM, Barlebo H, Rix M, Jacobsen GK (1993): Intratubular germ cell neoplasia in an infantile testis with immature teratoma. Histopathology 22:69-72. 103. Hawkins E, Heifetz SA, Giller R, Cushing B (1997): The prepubertal testis (prenatal and postnatal): its relationship to intratubular germ cell neoplasia: a combined Pediatric Oncology Group and Children's Cancer Study Group. Hum.Pathol. 28:404-410. 104. Jorgensen N, Muller J, Giwercman A, Visfeldt J, Moller H, Skakkebaek NE (1995): DNA content and expression of tumour markers in germ cells adjacent to germ cell tumours in childhood: probably a different origin for infantile and adolescent germ cell tumours. J.Pathol. 176:269-278. 105. Manivel JC, Simonton S, Wold LE, Dehner LP (1988): Absence of intratubular germ cell neoplasia in testicular yolk sac tumors in children. A histochemical and immunohistochemical study. Arch.Pathol.Lab Med. 112:641-645. 106. Mann JR (2001): Imprinting in the germ line. Stem Cells 19:287-294. 107. Ross JA, Schmidt PT, Perentesis JP, Davies SM (1999): Genomic imprinting of H19 and insulin- like growth factor-2 in pediatric germ cell tumors. Cancer 85:1389-1394. 108. Kraggerud SM, Skotheim RI, Szymanska J, Eknaes M, Fossa SD, Stenwig AE, Peltomaki P, Lothe RA (2002): Genome profiles of familial/bilateral and sporadic testicular germ cell tumors. Genes Chromosomes.Cancer 34:168-174. 109. Looijenga LH, Rosenberg C, Van Gurp RJ, Geelen E, Echten-Arends J, de Jong B, Mostert M, Wolter OJ (2000): Comparative genomic hybridization of microdissected samples from different stages in the development of a seminoma and a non-seminoma. J.Pathol. 191:187-192. 110. Hu J, Schuster AE, Fritsch MK, Schneider DT, Lauer S, Perlman EJ (2001): Deletion mapping of 6q21-26 and frequency of 1p36 deletion in childhood endodermal sinus tumors by microsatellite analysis. Oncogene 20:8042-8044. 111. Bussey KJ, Lawce HJ, Himoe E, Shu XO, Suijkerbuijk RF, Olson SB, Magenis RE (2001): Chromosomes 1 and 12 abnormalities in pediatric germ cell tumors by interphase fluorescence in situ hybridization. Cancer Genet.Cytogenet. 125:112-118. 112. Jenderny J, Koster E, Meyer A, Borchers O, Grote W, Harms D, Janig U (1995): Detection of chromosome aberrations in paraffin sections of seven gonadal yolk sac tumors of childhood. Hum.Genet. 96:644-650. 113. Schneider DT, Schuster AE, Fritsch MK, Calaminus G, Gobel U, Harms D, Lauer S, Olson T, Perlman EJ (2002): Genetic analysis of mediastinal nonseminomatous germ cell tumors in children and adolescents. Genes Chromosomes.Cancer 34:115-125. 114. Perlman EJ, Valentine MB, Griffin CA, Look AT (1996): Deletion of 1p36 in childhood endodermal sinus tumors by two-color fluorescence in situ hybridization: a pediatric oncology group study. Genes Chromosomes.Cancer 16:15-20. 115. Perlman EJ, Hu J, Ho D, Cushing B, Lauer S, Castleberry RP (2000): Genetic analysis of childhood endodermal sinus tumors by comparative genomic hybridization. J.Pediatr.Hematol.Oncol. 22:100-105. 116. Schneider DT, Schuster AE, Fritsch MK, Calaminus G, Harms D, Gobel U, Perlman EJ (2001): Genetic analysis of childhood germ cell tumors with comparative genomic hybridization. Klin.Padiatr. 213:204-211.

27 Chapter 1

117. Murty VV, Chaganti RS (1998): A genetic perspective of male germ cell tumors. Semin.Oncol. 25:133-144. 118. Samaniego F, Rodriguez E, Houldsworth J, Murty VV, Ladanyi M, Lele KP, Chen QG, Dmitrovsky E, Geller NL, Reuter V (1990): Cytogenetic and molecular analysis of human male germ cell tumors: chromosome 12 abnormalities and gene amplification. Genes Chromosomes.Cancer 1:289-300. 119. Rodriguez E, Mathew S, Reuter V, Ilson DH, Bosl GJ, Chaganti RS (1992): Cytogenetic analysis of 124 prospectively ascertained male germ cell tumors. Cancer Res. 52:2285-2291. 120. Mostert MM, van de PM, Olde WD, Suijkerbuijk RF, Geurts van Kessel A, van Echten J, Oosterhuis JW, Looijenga LH (1996): Comparative genomic hybridization of germ cell tumors of the adult testis: confirmation of karyotypic findings and identification of a 12p- amplicon. Cancer Genet.Cytogenet. 89:146-152. 121. Hecht F, Grix A, Jr., Hecht BK, Berger C, Bixenman H, Szucs S, O'Keeffe D, Finberg HJ (1984): Direct prenatal chromosome diagnosis of a malignancy. Cancer Genet.Cytogenet. 11:107-111. 122. Beverstock GC, Mollevanger P, Baaij M, Lind J, van Ieperen L, Bartelings MM, Teunissen K, Brandenburg H, Van Opstal D, Los F (1999): Nasopharyngeal teratoma and mosaic tetrasomy 1q detected at amniocentesis. A case report and review of the literature. Cancer Genet.Cytogenet. 115:11-18. 123. Schwartz S, Raffel LJ, Sun CC, Waters E (1992): An unusual mosaic karyotype detected through prenatal diagnosis with duplication of 1q and 19p and associated teratoma development. Teratology 46:399-404. 124. Scheres JM, de Pater JM, Stoutenbeek P, Wijmenga C, Rosenberg C, Pearson PL (1999): Isochromosome 1q as the sole chromosomal abnormality in two fetal teratomas. Possible trisomic or tetrasomic zygote rescue in fetal teratoma with an additional isochromosome 1q. Cancer Genet.Cytogenet. 115:1-10. 125. Murty VV, Houldsworth J, Baldwin S, Reuter V, Hunziker W, Besmer P, Bosl G, Chaganti RS (1992): Allelic deletions in the long arm of chromosome 12 identify sites of candidate tumor suppressor genes in male germ cell tumors. Proc.Natl.Acad.Sci.U.S.A 89:11006-11010.

A novel case of infantile sacral

teratoma and a constitutional

t(12;15) (q13;q25) pat

Imke Veltman Martien van Asseldonk

Marga Schepens

Hans Stoop Leendert Looijenga Cokkie Wouters

Lutgarde Govaerts

Ron Suijkerbuijk Cancer Genet. Cytogenet.

(2002) 136:17-22 Ad Geurts van Kessel

Chapter 2

ABSTRACT

Cytogenetic analysis of peripheral lymphocytes of an infantile patient with a sacral teratoma revealed a constitutional translocation (12;15) (q13;q25) pat. The same translocation was found in four additional relatives. Loss of heterozygosity analysis of the patients' tumor material showed retention of both translocation-derived chromosomes. Since allelic loss in the 12q13 region has been observed in germ cell tumors, we hypothesize that disregulation of gene(s) located at or near the 12q13 breakpoint may be related to the development of this sacral teratoma. As a first step towards the identification of this gene(s), a 12q13 genomic contig that spans the breakpoint has been constructed.

32 A novel case of infantile sacral teratoma and a constitutional t(12;15) (q13;q25)pat

INTRODUCTION

Germ cell tumors (GCTs) form a heterogeneous group of neoplasms that occur in the testis and the ovary and at extragonadal sites. All these neoplasms are believed to evolve from pluripotent progenitor cells. Gonadal germ cell tumors are generally thought to originate from cells belonging to the germ cell lineage [1;2]. However, for extragonadal tumors the origin is still not clear. These latter tumors may develop from germ cells that have migrated incorrectly during embryonic development or, alternatively, may arise from undifferentiated pluripotent embryonic stem cells that have escaped the influence of developmental organizers [3-5]. A classification of GCTs has been postulated, which was based on histology, anatomical localization, sex, and age of onset [6;7]. Based on this classification, the following four groups may be distinguished: (1) (immature) teratomas and yolk sac tumors of neonates and infants, (2) GCTs of the adult testicular type, (3) dermoid cysts (mature cystic teratomas) of the ovary and (4) spermatocytic seminomas in elderly men. Neoplasms of the first group are rare and are most often observed in female neonates in the sacral region. To a lesser extent, they appear at other extragonadal sites, in the ovary and in the testis. The most common site of occurrence of the second group is the testis of adolescent and young adult men. The dermoid cyst, the most frequent ovarian GCT, can be found in all age groups, but the median age is 30 years [8]. GCTs of the fourth group are rare and may only be found in the descended testis of elderly men [9]. Cytogenetic and molecular studies on GCTs have revealed several recurrent abnormalities that affect chromosome 12. As a hallmark of testicular GCTs (TGCT), an isochromosome of the short arm of 12, i(12p) has been identified [10;11;6]. This i(12p) chromosome is present in >80% of TGCTs, somewhat more frequent in non- seminomas than in seminomas [12]. This anomaly has also been found in several cases of extragonadal and ovarian GCTs [13-15]. For the regions 12q13 and 12q22, frequent loss of heterozygosity (>40%) was observed in all histological subtypes of male germ cell cancer [16]. In another study, frequent (44%) deletion of 12q13-q22 was found in non-seminomatous male GCTs and mixed male GCTs [13]. These findings were not confirmed by others, therefore, it is still debatable whether the 12q13 and 12q22 regions contain tumor suppressor gene(s). So far, several candidate genes have been postulated that may play a role in the initiation or progression of TGCTs, e.g., KRAS2, JAW1 and SOX5 on 12p [17]. The upregulated gene(s) present on i(12p) will probably not play a role in initiation, but

33 Chapter 2 rather in progression as suggested by Geurts van Kessel et al. [18] and, more recently, by Rosenberg et al. [19]. Through linkage analysis in a subgroup of familial TGCT cases, another locus has been suggested in the Xq27 region [20]. However, until now, no gene has unambiguously been identified to be involved, either in TGCT development or in any other form of germ cell cancer. The patient described in this report had an immature teratoma in the sacral region with a spread endodermal sinus tumor component. In addition, this patient had a constitutional chromosome translocation, t(12;15)(q13;q25) pat. Interestingly, the 12q13 breakpoint coincides with the region of allelic loss in male GCTs. In this study, we present the clinical data of the patient and the preliminary mapping of the 12q13 breakpoint within a genomic contig.

Figure 1. Lymphocyte-derived karyotype of the patient showing: 46,XX, t(12;15)(q13;q25).

MATERIAL AND METHODS

Karyotyping Peripheral blood lymphocytes were cultured in RPMI 1640 supplemented with 10% fetal calf serum and antibiotics for three days. Subsequently, the cells were harvested and metaphase spreads were prepared using standard procedures. Cytogenetic

34 A novel case of infantile sacral teratoma and a constitutional t(12;15) (q13;q25)pat analysis was carried out on trypsin-Giemsa stained slides and the karyotypes were described according to the International System for Human Cytogenetic Nomenclature [21].

B C

Figure 2. A) Section of the teratoma part of the tumor before microdissection, and B) after microdissection. C) Picture of the selected tumor sample used for DNA extraction.

Loss of heterozygosity analysis For loss of heterozygosity analysis, regions containing teratoma and normal tissue were selected from frozen tumor samples after microscopy. From these regions, tissue sections of 20 micron were used for microdissection of normal cells and tumor cells with a laser-capture system (Arcturus Pixcell II system) (Fig. 2). Directly after microdissection, the tissue samples were dissolved in TNE (10mM Tris, 400 mM NaCl, 2 mM EDTA pH 8.2), after which proteïnase K (10 mg/ml) and SDS (10%) was added. After an overnight incubation at 37°C, DNA was extracted with phenol/chloroform/ isoamylalcohol (25:24:1) (Merck) and precipitated with glycogen (10 mg/ml) (Boehringer). Pellets were dissolved in TE (Tris 10mM, EDTA 0.1 mM, pH 8.0). About 40 ng of the isolated DNA was used for PCR analysis in a total volume of 20 µl, containing: 2 µl of 10x PCR buffer without Mg (GIBCOBRL), 1.5 mM MgCl2 (GIBCOBRL), 0.25 mM dNTPs (GIBCOBRL), 0.1 µl fluorescent-labeled dUTP R110 (PE biosystems), 50 ng of each primer, 0.08 µl Taq polymerase (5 U/µl, GIBCOBRL). Primers specific for the following markers were used: VWF, D12S62

35 Chapter 2

(Genome Database accession number (GDB): 191682), D12S84 (GDB: 188005), GABRB3 (GDB: 188666). Samples were amplified for 30 cycles in a Perkin Elmer 2400 GeneAmp PCR System. Each cycle included: 0.15 seconds at 94°C, 0.15 seconds at 56°C (VWF, D12S62) or 60°C (D12S84, GABRB3) and 1 minute at 72°C. In the first cycle an extra denaturation step (94°C) was included for 1 minute. The fluorescent- labeled PCR products were purified using the Qiaquick PCR purification kit (Qiagen). After purification they were run on 6% polyacrylamide gels (Sequagel-6, National Diagnostics), using an ABI 373A (Applied biosystems, Perkin Elmer), and analysed with the GENESCAN ANALYSIS software package (Applied Biosystems, Perkin Elmer).

FISH analysis FISH analysis was carried out basically as described before [14]. on metaphase spreads of lymphocytes from the patient. The YAC clones (CEPH YAC library, GTC Leiden) used for FISH analysis were selected corresponding to markers from the Whitehead Contig WC12.1. These YAC clones were labeled with digoxigenin-11- dUTP using a nick-translation kit (Boehringer Mannheim). The labeled YACs were coprecipitated with excess of Cot-1 DNA and, prior to hybridization, preannealed at 37°C for 30 minutes. The slides with metaphase spread chromosomes were pre- treated with acetic acid according to standard procedures. This was followed by denaturation in 70% formamide, 2 x SSC, at 70°C for 2 minutes. Hybridizations were carried out at 37°C over 2-3 nights. For the YACs the following fluorochrome conjugates and antibodies were used: sheep-anti-digoxigenin-FITC (Boehringer Mannheim), donkey-anti-sheep-FITC (Jackson Immuno Research Laboratories), rabbit-anti-FITC (Dakopatts), mouse-anti-rabbit-FITC (Jackson Immuno Research Laboratories). Images were captured using a Zeiss epifluorescence microscope and a cooled CCD camera (Photometrics) and processed using the BDS-image software program (Biological Detection Systems).

RESULTS AND DISCUSSION

Clinical features of the patient At the age of six months, the patient, a female infant, had developed a swelling near the sacrum and a displacement of the anus. Macroscopic examination of the swelling

36 A novel case of infantile sacral teratoma and a constitutional t(12;15) (q13;q25)pat revealed the presence of a tumor. The tumor tissue was excised and, after pathological examination, classified as an immature teratoma (third grade) with a spread endodermal sinus tumor component. Further pathological data indicated that the tumor was not fully removed and despite additional chemotherapy the patient died. Additionally, the patient exhibited some dysmorphic features, i.e., obesitas and an upward slant.

Karyotyping of patient and relatives Peripheral blood lymphocytes were used for the preparation of metaphase spreads and karyotyping. This analysis revealed a diploid karyotype with a balanced translocation between the chromosomes 12 and 15, t(12;15)(q13;q25) (Fig. 1). To determine whether this translocation had occurred de novo, the parents were also karyotyped. The father appeared to be a carrier of the same translocation. Based on this observation, additional family members were examined cytogenetically. The grandfather, a paternal uncle and a sister of the patient also turned out to have the same t(12;15)(q13;q25). The patients’dysmorphic features were not seen in any of these translocation carriers nor did any of them develop a germ cell tumor.

200 tumor 150 tumor 100 100 50 50 216 194 220 203

normal 150 normal 100 100 50 50 A 216 C 194 220 203 300 400 tumor 300 tumor 200 200 100 100 180 157 190 169 40 200 normal 30 normal 150 20 100 10 50 D 180 B 157 190 169

Figure 3. Allele plots derived from tumor and normal tissue of the patient using four markers: A) D12S84, B) VWF, C) D12S62, D) GABRB3. Sizes of the alleles are indicated below the plots.

37 Chapter 2

LOH analysis of tumor tissue Given the fact that loss of translocation derivative chromosomes in renal cell tumors from patients with constitutional translocations has frequently been observed [22;23], we initiated a loss of heterozygosity (LOH) study on the tumor material of this patient. Therefore, tissue from the teratoma part of the tumor and normal tissue surrounding the tumor were microdissected (Fig. 2). DNA was extracted from these tissues and used for LOH analysis with known genetic markers. The markers VWF and D12S62 map on chromosome 12 above the breakpoint and, thus, are informative for the presence of der(12). D12S84, which maps distal to the breakpoint on chromosome 12, and GABRB3, a marker proximal to the breakpoint on chromosome 15, are informative for the presence of der(15). As shown in figure 3, similar allele patterns were observed for all four markers in both tumor and normal tissues. This result indicates that the tumor tissue of this patient is heterozygous for all four markers tested and thus both the der(12) and der(15) have been retained in the tumor cells. From these results we conclude that loss of either one of the translocation chromosomes is not associated with the development of this sacral teratoma.

Chromosome 12 YAC contig

58 cM WI-1851 D12S1663 893-A-3 WI-9296 881-D-8 CHCL.GATA91H06 WI-5295 WI-3757 944-A-5 WI-4540 q13.1 D12S85 754-A-1 WI-3295 WI-6754 WI-6377 WI-7025 WI-8712 919-G-8 WI-7600 WI-7321 WI-8355 WI-8280 IB3071 796-H-12 D12S347 67 cM D12S1629 WI-10864 A B

Figure 4. A) Diagram showing: chromosome 12, a series of q13.1 markers and the corresponding CEPH YACs. B) FISH analysis with YAC 919-G-8 on a lymphocyte derived metaphase spread from the patient. Positive signals (green) on both der(12) and der(15) and on the normal chromosome 12 are indicated. Chromosomes are shown in red.

38 A novel case of infantile sacral teratoma and a constitutional t(12;15) (q13;q25)pat

Mapping of the 12q13 breakpoint within a YAC contig Since the translocation breakpoint itself might be relevant for the development of the sacral teratoma, we initiated the physical mapping of the 12q13 breakpoint. Several YACs were selected within this region and used as probes for FISH analysis on metaphase spreads of peripheral blood lymphocytes from the patient. Figure 4A shows the selected YACs and their corresponding genetic markers. YAC 893-A-3, 881-D-8, 944-A-5 and 754-A-1 all displayed FISH signals on chromosome 12 and der(12), indicating that these YACs map centromeric to the breakpoint on chromosome 12. YAC 796-H-12 gave signals telomeric to the breakpoint on chromosome 12 and on der(15). In contrast, YAC 919-G-8 gave, besides on chromosome 12, signals on both der(12) and der(15), indicating that this YAC spans the 12q13 breakpoint (Fig. 4B). This YAC may serve as a starting point to identify the gene(s) involved. In order to explain the occurrence of dysmorphic features and a sacral teratoma in the patient, and lack of these anomalies in the other translocation- positive relatives including the father, one could speculate that additional submicroscopic changes (duplications or deletions) may be present in one of the translocation derivatives of the patient, leading to abnormal gene regulation and/or expression. These changes could have taken place due to an aberrant pairing and/or segregation of the translocation chromosomes during the first meiotic division of the primary spermatocytes, a phenomenon that has been observed in other translocation cases [24]. Another possibility is that the 12q13 allele on the normal chromosome 12 is mutated somatically and contributes to tumorigenesis through a classical two-hit scenario [25;26]. Currently we are investigating what gene(s) is involved in the translocation and whether this gene(s) is differently affected in the patient as compared to the other translocation carriers.

ACKNOWLEDGEMENTS

We would like to thank Marc J. Eleveld for technical support with the LOH studies.

39 Chapter 2

REFERENCES

1. Looijenga LH, de Munnik H, Oosterhuis JW (1999): A molecular model for the development of germ cell cancer. Int J Cancer 83:809-814. 2. Looijenga LH, Oosterhuis JW (1999): Pathogenesis of testicular germ cell tumours. Rev Reprod 4:90-100. 3. Gonzalez-Crussi F (1982): Extragonadal teratomas. In atlas of tumor pathology, second series, fascicle 18. AFIP, Washington, D.C. 4. Hoffner L, Deka R, Chakravarti A, Surti U (1994): Cytogenetics and origins of pediatric germ cell tumors. Cancer Genet Cytogenet 74:54-8. 5. Sano K (1999): Pathogenesis of intracranial germ cell tumors reconsidered. J Neurosurg 90:258- 64. 6. Sandberg AA, Meloni AM, Suijkerbuijk RF (1996): Reviews of chromosome studies in urological tumors. III. Cytogenetics and genes in testicular tumors. J Urol 155:1531-56. 7. Oosterhuis JW, Looijenga LH, van Echten J, de Jong B (1997): Chromosomal constitution and developmental potential of human germ cell tumors and teratomas. Cancer Genet Cytogenet 95: 96-102. 8. Surti U, Hoffner L, Chakravarti A, Ferrell RE (1990): Genetics and biology of human ovarian teratomas. Am J Hum Genet 47:635-43. 9. Eble JN (1994): Spermatocytic seminoma. Hum Pathol 25:1035-42. 10. Atkin NB, Baker MC (1982): Specific chromosome change, i(12p), in testicular tumours? Lancet 2:1349. 11. Geurts van Kessel A, Suijkerbuijk R, de Jong B, Oosterhuis JW (1991): Molecular analysis of isochromosome 12p in testicular germ cell tumors. Recent Results Cancer Res 123:113-8. 12. van Echten J, Oosterhuis JW, Looijenga LH, van de Pol M, Wiersema J, te Meerman GJ, Schaffordt Koops H, Sleijfer DT, de Jong B (1995): No recurrent structural abnormalities in germ cell tumors of the adult testis apart from i(12p). Genes Chromosomes Cancer 14:133-44. 13. Samaniego F, Rodriguez E, Houldsworth J, Murty VV, Ladanyi M, Lele KP, Chen QG, Dmitrovsky E, Geller NL, Reuter V, Jhanwar SC, Bosl GJ, Chaganti RS (1990): Cytogenetic and molecular analysis of human male germ cell tumors: Chromosome 12 abnormalities and gene amplification. Genes Chromosomes Cancer 1:289-300. 14. Suijkerbuijk RF, Looijenga LH, de Jong B, Oosterhuis JW, Cassiman JJ, Geurts van Kessel A (1992): Verification of isochromosome 12p and identification of other chromosome 12 aberrations in gonadal and extragonadal human germ cell tumors by bicolor double fluorescence in situ hybridisation. Cancer Genet Cytogenet 63:8-16. 15. Riopel MA, Spellerberg A, Griffin CA, Perlman EJ (1998): Genetic analysis of ovarian germ cell tumors by comparative genomic hybridisation. Cancer Res 58:3105-10. 16. Murty VV, Houldsworth J, Baldwin S, Reuter V, Hunziker W, Besmer P, Bosl G, Chaganti RS (1992): Allelic deletions in the long arm of chromosome 12 identify sites of candidate tumor suppressor genes in male germ cell tumors. Proc Natl Acad Sci U S A 89:11006-10. 17. Mostert MC, Verkerk AJ, van de Pol M, Heighway J, Marynen P, Rosenberg C, Geurts van Kessel A, van Echten J, de Jong B, Oosterhuis JW, Looijenga LH (1998): Identification of the

40 A novel case of infantile sacral teratoma and a constitutional t(12;15) (q13;q25)pat

critical region of 12p over-representation in testicular germ cell tumors of adolescents and adults. Oncogene 16:2617-27. 18. Geurts van Kessel A, van Drunen E, de Jong B, Oosterhuis JW, Langeveld A, Mulder MP (1989): Chromosome 12q heterozygosity is retained in i(12p)-positive testicular germ cell tumor cells. Cancer Genet Cytogenet 40:129-34. 19. Rosenberg C, Van Gurp RJ, Geelen E, Oosterhuis JW, Looijenga LH (2000): Overrepresentation of the short arm of chromosome 12 is related to invasive growth of human testicular seminomas and nonseminomas. Oncogene 19:5858-5862. 20. Rapley EA, Crockford GP, Teare D, Biggs P, Seal S, Barfoot R, Edwards S, Hamoudi R, Heimdal K, Fossa SD, Tucker K, Donald J, Collins F, Friedlander M, Hogg D, Goss P, Heidenreich A, Ormiston W, Daly PA, Forman D, Oliver TD, Leahy M, Huddart R, Cooper CS, Bodmer JG, Easton DF, Stratton MR, Bishop DT (2000): Localisation to Xq27 of a susceptibility gene for testicular germ-cell tumours. Nat Genet 24:197-200. 21. ISCN (1995): An International System for Human Cytogenetic Nomenclature. F Mitelman, ed. Karger, Basel. 22. Knudson AG (1995): VHL Gene Mutation and Clear-Cell Renal Carcinomas. Cancer J Sci Am 1:180-181. 23. Bodmer D, Eleveld MJ, Ligtenberg MJ, Weterman MA, Janssen BA, Smeets DF, de Wit PE, van den Berg A, van den Berg E, Koolen MI, Geurts van Kessel A (1998): An alternative route for multistep tumorigenesis in a novel case of hereditary renal cell cancer and a t(2;3)(q35;q21) chromosome translocation. Am J Hum Genet 62:1475-83. 24. Smeets DF, Hamel BC, Nelen MR, Smeets HJ, Bollen JH, Smits AP, Ropers HH, van Oost BA (1992): Prader-Willi syndrome and Angelman syndrome in cousins from a family with a translocation between chromosomes 6 and 15. N Engl J Med 326:807-11. 25. Knudson AG (1993): Antioncogenes and human cancer. Proc Natl Acad Sci U S A 90:10914-21. 26. Knudson AG (1996): Hereditary cancer: two hits revisited. J Cancer Res Clin Oncol 122:135-40.

Fusion of the SUMO/Sentrin-specific protease 1 gene SENP1 and the

embryonic polarity-related mesoderm

development gene MESDC2 in a patient

with an infantile teratoma and a

constitutional t(12;15)(q13;q25)

Imke M. Veltman Lilian A. Vreede Jinke Cheng

Leendert H.J. Looijenga

Bert Janssen Eric F.P.M. Schoenmakers Human Molecular Genetics Edward T.H. Yeh (2005) 14: 1955-1963 Ad Geurts van Kessel

Chapter 3

ABSTRACT

Recently we identified a patient with an infantile sacrococcygeal teratoma and a constitutional t(12;15)(q13;q25). Here we show that, as a result of this chromosomal translocation, the SUMO/Sentrin-specific protease 1 gene (SENP1) on chromosome 12 and the embryonic polarity-related mesoderm development gene (MESDC2) on chromosome 15 are disrupted and fused. Both reciprocal SENP1-MESDC2 (SEME) and MESDC2-SENP1 (MESE) fusion genes are transcribed in tumor-derived cells and their open reading frames encode aberrant proteins. As a consequence of this, and in contrast to wild-type (WT) MESDC2, the translocation associated SEME protein is no longer targeted to the endoplasmatic reticulum, leading to a presumed loss of function as a chaperone for the WNT co-receptors LRP5 and/or LRP6. Ultimately, this might lead to abnormal development and/or routing of germ cell tumor precursor cells. SUMO, a post-translational modifier plays an important role in several cellular key processes and is cleaved from its substrates by WT SENP1. Using a PML desumoylation assay we found that translocation-associated MESE proteins exhibit desumoylation capacities similar to those observed for WT SENP1. We speculate that spatio-temporal disturbances in desumoylating activities during critical stages of embryonic development might have predisposed the patient. Together, the constitutional t(12;15)(q13;q25) translocation revealed two novel candidate genes for neonatal/infantile GCT development: MESDC2 and SENP1.

44 Fusion of SENP1 and MESDC2 in an infantile teratoma with constitutional t(12;15)(q13;q25)

INTRODUCTION

Human germ cell tumors (GCTs) can arise at various ages, in both sexes, at different body locations and vary with respect to clinical outcome, histology and genetic constitution [1-3]. GCTs are most prevalent in the testis of young adult males. We and others have suggested that GCTs of infants and neonates may represent a distinct subgroup [2-8]. These latter tumors are relatively rare, mainly arise in the sacral/sacrococcygeal region and, at a lower frequency, in the gonads and along the midline of the body [3]. How these tumors develop and from which precursor cells they originate is still a matter of debate. A common precursor cell type, however, has been suggested for infantile gonadal and extra-gonadal GCTs [9]. The main histological infantile subtypes include teratomas with mesodermal, ectodermal and endodermal components, and yolk sac tumors with yolk sac endodermal components. Teratoma development resembles embryogenesis in various aspects [10;11]. Therefore, these latter tumors have been used as model systems to study both early embryonic and tumorigenic processes [12;13]. Currently available information on the cytogenetic constitution of neonatal and infantile GCTs indicates that teratomas are mostly diploid with only minor non- recurrent chromosomal anomalies. In contrast, infantile yolk sac tumors exhibit several major recurrent chromosomal aberrations [7;8;15-18]. Overall, the most frequent cytogenetic anomaly in GCTs is the i(12p) chromosome. This marker chromosome is almost exclusively present in testicular GCTs of juveniles and adults [19]. Several recurrent genomic anomalies in both neonatal/infantile and juvenile/adult GCTs are currently under investigation [20-24]. So far, however, no gene that is causally related to GCT development has unambiguously been identified, not for the infantile type nor for the adult type.

Using linkage analysis, Rapley et al. [25] identified a locus on Xq27 for familial testicular germ cell cancer. Additional candidate regions harboring putative genes for familial GCT development include 16p13, 18q22-qter and 12q12-13 [26]. Together with the 12q22 region, the 12q12-13 region was found to be frequently deleted in sporadic GCTs as well [27-29]. Linkage and CGH studies typically result in the identification of relatively large tumor-related genomic intervals containing numerous candidate genes. Recurrent, apparently balanced chromosomal anomalies, however, provide an opportunity to overcome this problem. Constitutional chromosome translocations have been widely employed to identify candidate genes

45 Chapter 3 in various (hereditary) cancer syndromes [30-33]. Recently, we detected a patient with an infantile sacrococcygeal teratoma and a constitutional t(12;15)(q13;q25) [34]. Here we show that this translocation results in a fusion of the SUMO/Sentrin- specific protease 1 gene SENP1 on 12q13 and the embryonic polarity-related mesoderm development gene MESDC2 on 15q25. The putative implications of this novel gene fusion in the context of GCT development are discussed.

RESULTS

) Chr. 12 Breakpoint PAC contig Genes 5 ) ) 1 2 5 2 ; 1 1 1 . ( ( spanning YAC 2 r r 1 r ( h e e t c d d

; ; ; ; 1 2 3 4 HindIII >12kb - RP1-316M24 q13 RP1-130F5 MGC5576 919G8 4.1 kb- COL2A1 BglII RP3-197B17 ~10 kb- SENP1 6.7 kb- RP1-228P16 PFKM

A B

Figure 1. Mapping of the 12q13 breakpoint within the SENP1 gene. A) Diagram depicting chromosome 12, the breakpoint-spanning CEPH YAC clone 919G8, its corresponding PAC contig and the genes corresponding to the breakpoint-spanning PAC clone RP1-228p16. The breakpoint spanning clones and gene are depicted in gray. B) Southern blot analysis of the t(12;15)(q13;q25)- positive cell line (lane 1) and its derived somatic cell hybrids (lane 2: chromosome 12; lane 3: der(15); lane 4: der(12)) after digestion of the DNAs with HindIII or BglII. A PCR fragment obtained with primer pair 228p16-38 and located proximal to the breakpoint was used as a probe. Next to the WT chromosome 12 fragments (lanes 1 and 2: 4.1kb HindIII fragment and 6.7 kb BglII fragment) shifted fragments are readily seen in the original cell line (lane 1) and the der(15)-containing hybrid (lane 3: >12kb HindIII and ~11kb BglII fragments, respectively). A cross-hybridizing Chinese hamster fragment is marked by an arrowhead.

SENP1 is disrupted by the 12q13 breakpoint Previously, we identified a YAC clone (CEPH-919G8) that spans the constitutional t(12;15)(q13;q25) breakpoint at 12q13 [5]. In order to refine this breakpoint region, a PAC contig was assembled that corresponds to this YAC clone (Fig. 1A). FISH analysis on metaphase spreads of t(12;15)(q13;q25)-positive cells revealed that PAC

46 Fusion of SENP1 and MESDC2 in an infantile teratoma with constitutional t(12;15)(q13;q25) clones RP1-316M24, RP1-130F5 and RP3-197B17 map proximal to the breakpoint and that, in contrast, RP1-228P16 turned out to be a breakpoint spanning clone. This finding is in full concordance with our subsequent array CGH-based mapping data [35]. The complete sequence of RP1-228P16 is available in GenBank (accession no. AC004801) and, according to the May 2004 human genome assembly freeze (http://genome.ucsc.edu/), it encompasses four positional candidate genes: MGC5576 (a hypothetical gene), COL2A1, SENP1 and PFKM (Fig. 1A). To facilitate further refinement of the chromosomal breakpoint position in RP1-228P16 the t(12;15)(q13;q25)-associated chromosomes were segregated by generating Chinese hamster-human somatic cell hybrids. These hybrid cell lines with either the normal chromosome 12, the derivative chromosome 12 (der(12)) or the derivative chromosome 15 (der(15)) were employed for STS-content mapping using specific PCR primer sets from pre-selected regions within RP1-228P16. The original t(12;15)(q13;q25)-positive lymphoblastoid cell line was included as a positive control. By doing so, the breakpoint interval could be reduced to a 3.6 kb genomic interval (data not shown). In order to confirm these PCR results and to further narrow down the breakpoint region, Southern blot analysis was performed using HindIII and BglII digested DNAs extracted from the original and hybrid cell lines. A PCR fragment generated by one of the breakpoint-flanking primer sets (228p16-38) was used as probe. As expected, this probe hybridized to both HindIII and BglII wild-type (WT) fragments of respectively 4.1 and 6.7 kb in the t(12;15)(q13;q25)-positive cell line and the chromosome 12-positive hybrid cell line (Fig. 1B, lanes 1 and 2). In addition, aberrantly hybridizing fragments were observed in the HindIII (>12kb) and BglII (~10 kb) digested DNAs of the t(12;15)(q13;q25)-positive cell line and the der(15)-positive hybrid cell line (lanes 1 and 3, respectively). These shifted fragments were absent in the der(12)-positive hybrid cell line (lane 4), indicating that they originate from the der(15). Together, the PCR and Southern blot results reduced the 12q13 breakpoint region to a 2.4 kb genomic interval. Analysis of its sequence using the UCSC human genome browser (May 2004 freeze) revealed that it is located within the second intron of one of the four earlier mentioned candidate genes: SENP1 (Fig. 2A). SENP1 belongs to the family of SUMO/Sentrin-specific proteases which are known to be involved in the cleavage of SUMO proteins from their substrates [36]. The SENP1 gene is ~63 kb in size and encodes an mRNA of 4.7 kb (GenBank accession no. NM_014554). The May 2004 freeze of the UCSC human genome browser reveals several different mRNA splice variants that differ in their first exons, the length of

47 Chapter 3 their 3’-UTRs and/or the number of (additional) exons. The largest open reading frame (ORF) encodes a 643 amino acids protein that corresponds to the WT protein described by Gong et al. [36].

A SENP1

breakpoint MESDC2

breakpoint ) ol 5 B tr ;1 n 12 co t(

GGT TCCGGTTCGGACTTTGT/ TGAAATGGAAAGATGATGACATT GGTTCCGGTTCG------/ ------AAAGATGATGACATT exon1exon2 exon2 SENP1SENP1 MESDC2

l r ) l r ) l r ) l r o l u 5 ro u 5 ro u 5 ro u r ro o ;1 t o ;1 t o ;1 t o t t m 2 n 2 n 2 n n n u 1 o m 1 o m 1 o m co o C t t( -c tu t( -c tu t( -c tu + -c

SEME MESE SENP1 MESDC2

l ) l ) ro 5 r ro 5 r t ;1 u t ;1 u n 2 o n 2 o o 1 m o 1 m -c t( tu -c t( tu D

der(15) der(12)

Figure 2. Fusion of SENP1 on chromosome 12 and MESDC2 on chromosome 15. A) Diagram showing SENP1 and MESDC2 and the respective positions of the breakpoints in introns 2 and 1. B) 3’RACE products obtained from normal control and t(12;15)(q13;q25)-positive lymphoblastoid cell lines, and sequences of t(12;15)(q13;q25)-derived SENP1-MESDC2 fusion products. C) RT-PCR analysis of SENP1-MESDC2 (SEME), MESDC2-SENP1 (MESE), SENP1 and MESDC2 in the tumor sample, a positive control cell line (either the original t(12;15)(q13;q25)-positive lymphoblastoid cell line or, for MESDC2, a lymphoblastoid cell line from a healthy individual), and a negative control. D) PCR analysis of the fusion genes SENP1-MESDC2 (SEME) on der(15) and MESDC2-SENP1 (MESE) on der(12) in the paternal t(12;15)(q13;q25)-positive lymphoblastoid cell line, the patient-derived tumor sample, and a negative control.

SENP1 is fused to MESDC2 at 15q25 On the basis of the localization of the 12q13 breakpoint within SENP1, we set out to identify the breakpoint sequences present at 15q25. To this end 3’-rapid amplification of cDNA ends (3’RACE) analysis was performed using the SENP1 breakpoint

48 Fusion of SENP1 and MESDC2 in an infantile teratoma with constitutional t(12;15)(q13;q25) information as a starting point. cDNAs generated from the t(12;15)(q13;q25)-positive lymphoblastoid cell line and, as a control, cDNAs from an unrelated lymphoblastoid cell line without a t(12;15)(q13;q25) were used. Aberrant PCR-fragments only present in the t(12;15)(q13;q25)-positive cell line were cloned, sequenced and analyzed (Fig. 2B). This resulted in the identification of chimeric transcripts, consisting of SENP1 sequences fused to sequences from MESDC2, a gene located at the 15q25 breakpoint region. More specifically, the MESDC2 gene is located in BAC clone RP11-775C24, again completely in line with our array CGH-based mapping data [35]. The breakpoint in the MESDC2 gene appeared to be located within its first intron (Fig. 2A). MESDC2 (mesoderm development gene) is an embryonic polarity related gene [37;38] of about 14 kb in size and encodes a 4.2 kb mRNA corresponding to a deduced protein of 234 amino acids (GenBank accession no. D42039). Also, for this gene additional mRNAs that differ in length at their ‘3-UTR are reported in the May 2004 freeze of the UCSC human genome browser. The presence of SENP1-MESDC2 fusion transcripts (here referred to as SEME) in the t(12;15)(q13;q25)-positive cell line was confirmed by reverse transcription- polymerase chain reaction (RT-PCR). Similarly, the presence of SEME was confirmed in the primary tumor sample of the patient as well (Fig. 2C, first panel). Additionally, RT-PCR showed the presence of the reciprocal MESDC2-SENP1 transcript (referred to as MESE) and the WT SENP1 and MESDC2 transcripts in the primary tumor sample (Fig. 2C). The t(12;15)(q13;q25)-positive lymphoblastoid cell line that we used for mapping of the breakpoints was derived from the patient’s father. In order to verify whether the genomic breakpoints identified in this cell line are in conformity to those in the patient, we performed PCR analysis on genomic DNAs extracted from the tumor material of the patient and the t(12;15)(q13;q25)-positive cell line. In both cases similar PCR products were obtained (Fig. 2D) and their sequences revealed that both breakpoints were identical at the base-pair level, thereby ruling out breakpoint- flanking duplications and/or deletions in the patient due to incomplete pairing and unequal crossing over of the respective derivative chromosomes at the breakpoints during meiotic division. In addition, we sequenced the complete ORFs of SENP1 and MESDC2 from the tumor-derived DNA and no additional (somatic) mutations were detected.

49 Chapter 3

MESE has retained desumoylating activity The reciprocal MESE fusion transcript is composed of the first exon of MESDC2 followed by exons 3-18 of SENP1. Owing to the translocation, the original start codon of SENP1 [36] is lost and two alternative ORFs are predicted for this MESE transcript. Depending on which ATG is used as a start codon, the encoded protein represents either an aberrant MESDC2 protein with 71 in frame N-terminal amino acids followed by six additional amino acids or a protein that is almost completely composed of SENP1 except for the second amino acid at the N-terminus (here referred to as MESE 1.2). Using RT-PCR, we detected an additional MESE transcript containing an aberrant exon 1 of MESDC2 (lacking 199 bp at the 3’-end) and exons 3- 18 of SENP1 in the t(12;15)(q13;q25)-positive lymphoblastoid cell line. The predicted ORF of this transcript encodes a putative protein that lacks 8 amino acids at the N- terminus as compared with WT SENP1 (here referred to as MESE 11.4). The two aberrant SENP1 proteins (MESE 1.2 and 11.4) encoded by the largest predicted ORFs contain the known SENP1 functional domains, i.e., the nuclear localization signal NLS and the protease domain (Fig. 3) [36;39].

breakpoint NLS SENP1 protease domain

breakpoint MESDC2 signal peptide ER-retention signal

SEME ER-retention signal

NLS MESE protease domain Figure 3. Diagram showing the proteins encoded by WT SENP1 and MESDC2 genes and the SEME and MESE fusion genes. Functional domains and the respective breakpoints are marked.

SENP1 is a SUMO/Sentrin-specific protease that can cleave SUMO-1 from target proteins such as PML [36;40]. To examine whether the MESE proteins have retained this functional capacity, we performed desumoylation assays using PML as a target. Therefore, COS-1 cells were transfected with one or more of the following expression constructs: His-PML, Ha-SUMO-1, MESE-VSV (1.2 or 11.4), WT SENP1- VSV and/or a control empty vector (Fig. 4). His-PML proteins with or without a

50 Fusion of SENP1 and MESDC2 in an infantile teratoma with constitutional t(12;15)(q13;q25) conjugated Ha-SUMO-1 protein were purified from the cell lysates by immunoprecipitation (IP) using an anti-polyhistidine antibody. Multiple SUMO-1 associated PML isoforms were readily detected in the lysates of COS-1 cells expressing SUMO-1 and PML (Fig. 4; sumoylated PML, upper panel, lane 3). These isoforms were not detected in control IP samples of COS-1 cells transfected with empty vectors in conjunction with vectors without (lane 1) or with (lane 2) PML. Also, COS-1 cells expressing WT SENP1 together with SUMO-1 and PML did not show any sumoylated isoforms of PML (lane 4), indicating that the majority of PML protein is indeed desumoylated by SENP1. Similarly, IPs of COS-1 cells expressing MESE (11.4 or 1.2) together with PML and SUMO-1 (lane 5 and 6) did not show any sumoylated isoforms of PML, indicating that, although both MESE proteins were expressed at a lower level than WT SENP1 (Fig. 4; SENP1 variants), they still were fully capable of desumoylating PML. α-His-IP 13452 6

Sumoylated 150 - PML α-HA 100 -

75 - PML α-His

Total cell lysate SENP1 75 - α-VSV variants 50 -

Empty vector + + + - - - His-PML - + + + + + HA-SUMO-1 - - + + + + WT SENP1-VSV -- - + -- MESE-VSV (11.4) -- - - + - MESE-VSV (1.2) -- - - -+

Figure 4. PML desumoylation assay with VSV-tagged SENP1 and MESE (1.2 and 11.4) proteins. COS-1 cells were transfected with one or more of the following constructs: His-PML, Ha-SUMO-1, WT SENP1-VSV, MESE-VSV (11.4), MESE-VSV (1.2), or empty vector (PSG8). His-PML proteins were pulled down with the anti-polyhistidine antibody using IP (anti-polyhistidine IP) and subsequently detected on western blots: sumoylated His-PML with the anti-Ha antibody and desumoylated His-PML with the anti-polyhistidine antibody. SENP1 variants were visualized in the total cell lysates using an anti-VSV antibody. The molecular weights (in kilodaltons) from co-electrophoresed markers are indicated.

51 Chapter 3

Anomalous subcellular localization of SEME WT MESDC2 contains a putative N-terminal signal sequence and a non-conserved C- terminal ER-retention signal (Fig. 3). Previously, it was shown that a flag-tagged mouse homologue of MESDC2 co-localizes with the luminal ER protein calreticulin in COS-1 cells [38]. In this study, we independently confirmed this localization by showing that a VSV-tagged human MESDC2 co-localizes with the ER protein PDI in HeLa cells (Fig. 5A). The translocation-related SEME fusion transcript contains the first two exons of SENP1 and exon 2 and 3 of MESDC2. A predicted ORF encodes a MESDC2 protein that lacks 106 amino acids at the N-terminus when compared with WT MESDC2 (Fig. 3). In addition, we detected a splice variant of SEME without the SENP1 exon 2 in de tumor material (SEME-2). The predicted ORF of this splice variant lacks 71 N-terminal MESDC2 amino acids and, instead, contains 7 additional amino acids. Owing to the truncations, both SEME proteins lack the putative signal peptide that is required for transport to the ER. Indeed, exogeneous expression of VSV-tagged SEME in HeLa cells revealed a diffuse staining pattern throughout the cytoplasm and no co-localization with PDI was observed (Fig. 5B). Therefore, we conclude that SEME is not localized at or in the ER. Control empty vector cells failed to yield any signals using the anti-VSV and secondary antibodies (not shown).

A α-VSV α-PDI α-VSV + α-PDI

MESDC2-VSV

B α-VSV α-PDI α-VSV + α-PDI

SEME-VSV

Figure 5. Subcellular localization of A) MESDC2 and B) SEME, using C-terminal VSV tags and SV40 promoters, after transient expression in HeLa cells. The proteins were visualized using an anti-VSV antibody (green; left panels). Endogenous PDI, an ER resident protein, was visualized with an anti- PDI antibody (red; middle panels). Overlays are shown in the right panels.

52 Fusion of SENP1 and MESDC2 in an infantile teratoma with constitutional t(12;15)(q13;q25)

DISCUSSION

Recently, we presented a novel case of a sacrococcygeal teratoma from an infantile patient carrying a constitutional t(12;15)(q13;q25). LOH experiments confirmed the presence of both reciprocal derivative chromosomes in the tumor material [34]. In addition to this constitutional translocation, we identified one additional (somatic) cytogenetic aberration in this tumor, i.e., amplification of 20q and deletion of 20p [18]. Since the 12q13 region is recurrently affected in human germ cell tumors [27;28], we set out to characterize this constitutional t(12;15)(q13;q25) translocation in detail. Using positional cloning strategies, we found that the breakpoints on chromosome 12 and chromosome 15 disrupt two genes, i.e., SENP1 and MESDC2, respectively. This resulted in the formation of two reciprocal fusion genes: SENP1-MESDC2 on the der(15) and MESDC2-SENP1 on the der(12). Next to the respective WT genes, both fusion genes (here referred to as SEME and MESE) appeared to be transcribed in the tumor tissue, and functional ORFs were predicted from them. SUMO/Sentrin is an ubiquitin-like protein that covalently attaches to lysine residues in substrate proteins as a post-translational modification. In vertebrates three SUMO family members have been described, i.e., SUMO-1, -2 and -3 [41;42]. In contrast to ubiquitin, SUMO does not target proteins for proteasomal destruction. Instead, SUMO modification appears to be involved in the regulation of various cellular key processes, including transcriptional regulation, nuclear transport, maintenance of genome integrity, and signal transduction [43-45]. The SENP1 gene was first identified by Gong et al. [36] and encodes a cysteine-specific protease that deconjugates SUMO-1 from its target proteins. A recent in vitro study showed that SENP1 can also preprocess SUMO-1, -2 and -3 from their precursor forms [46]. So far, seven SUMO-related proteases have been identified and deposited in the Genbank database (SENP1-SENP7). These proteases are homologous at their C-termini, a region that contains a conserved protease domain, the HIS/ASP/CYS catalytic triad. On the basis of the differences in subcellular localization and mRNA expression patterns, substrate specificities have been suggested for the various SENP family members [39;47-50]. Interestingly, one of the family members, SENP6 (SUSP1) was recently found to be disrupted and fused to the TCBA1 gene in the lymphoma- derived cell line HT-1 [51]. Also, SENP1 was detected as one of the 30 overexpressed genes in thyroid oncocytic tumors, using a two-step differential RT-PCR assay [52]. Together with the present study these findings strengthen a possible role for this protease family in cancer development.

53 Chapter 3

Through a PML desumoylation assay we established that the two translocation-associated MESE proteins (differing in only eight or one amino acid(s) from WT SENP1, respectively) exhibited desumoylation capacities similar to that observed for WT SENP1. In addition, the MESDC2-SENP1 fusion gene is under the control of the MESDC2 promoter. Therefore, we speculate that spatio-temporal disturbances in desumoylating activities during critical stages of embryonic development may lead to an abnormal regulation of gene transcription, nuclear transport, genome integrity and/or signal transduction in GCT precursor cells. In addition to these promoter-swap related phenomena, it should be noted that N- terminal truncation of the SENP1 protein may have additional in vivo effects on its biological activity and/or its substrate specificity. Such disturbances may have predisposed our patient to GCT development. The mouse homologue of MESDC2 (aliases: Mesd in mouse, KIAA0081 in human, BOCA in fly) was first identified as a candidate gene present in the mouse proximal albino deletion interval that is critical for mesoderm differentiation and polarity during embryonic development [37]. Embryos homozygous for this deletion are lethal, fail to develop mesodermal structures and do not form a primitive streak [53]. Hsieh et al. [38] demonstrated by transgene rescue that Mesd is the gene that is responsible for the polarity defects and embryonic lethality in these mice. Subsequently, transient expression in COS-1 cells demonstrated that Mesd may act as a chaperone located in the ER that facilitates the membrane localization and prevents the aggregation of LDL receptor family members like the WNT co-receptors LRP5 and LRP6 [38;54]. In this study we show that, in contrast to WT MESDC2, the translocation-associated SEME protein is no longer restricted to the ER. Consequently, it may have lost its capacity to act as a chaperone for the LDL receptors LRP5 and/or LRP6. Like the MESDC2-SENP1 fusion gene, the SENP1- MESDC2 fusion gene also has undergone a promoter-swap. Again, such a swap may have spatio-temporal consequences during early embryogenesis, including an abnormal development and/or routing of GCT precursor cells. Obviously, additional studies are required to establish a definite role for desumoylation, LDL receptor aggregation or both, on the (de)regulation of GCT precursor cells. This novel constitutional chromosome translocation, however, appears to have revealed novel leads to candidate genes and pathways critical to the development of neonatal/infantile GCTs.

54 Fusion of SENP1 and MESDC2 in an infantile teratoma with constitutional t(12;15)(q13;q25)

MATERIALS AND METHODS

Patient material and somatic cell hybrids Peripheral blood lymphocytes were used (informed consent) to generate an immortal t(12;15)(q13;q25)-positive cell line after in vitro Epstein-Barr virus (EBV) transformation using standard procedures. For the preparation of somatic cell hybrids, these t(12;15)(q13;q25)-positive cells were fused with thymidine kinase– deficient Chinese hamster A3 cells as described before [55]. Subsequently, independent somatic cell hybrids were isolated after hypoxanthine-aminoterine- thymidine (HAT) selection and their human chromosomal constitution was established through karyotyping, FISH analysis and PCR-screening with primers: D12S63, D12S85, D12S1661 and D12S1701. Hybrid cell lines that contained the normal chromosome 12, the der(15) or the der(12), respectively, were selected for further analysis. No other chromosome 12 or chromosome 15 material was detected in these selected hybrid cell lines.

FISH analysis Metaphase spreads from the t(12;15)(q13;q25)-positive cell line and its derived somatic cell hybrids were used for FISH analysis. Probe labeling, slide preparation, and hybridization were performed essentially as described before [34;56]. The following clones were used for FISH analysis: RP1-316M24, RP1-130F5, RP3-197B17, RP1-228P16 (BACPAC Resources, Oakland, USA). A Zeiss epifluorescence microscope equipped with appropriate filters was used for visual examination of the slides. Digital images were captured using a high-performance cooled CCD camera (Photometrics) coupled to a Macintosh Quadra 950 computer and processed using the Oncor-image software (Gaithersburgh, MD, USA).

PCR and Southern blot analysis Genomic DNAs were isolated as described previously [57]. For mapping of the 12q13 breakpoint, genomic DNAs of the parental and hybrid cell lines were analyzed using different primer sets including those flanking the breakpoint, i.e.: 228p16-38: 228p16-38F; TCACTTCTAGTGGCTGCACC 228p16-38R; AAACCCATTCTTTCACTACAGC 228p16-41: 228p16-41F; GAGTTCTTTGAGAACAGTACG 228p16-41R; ATAGCACAGACATCTTAGAGC

55 Chapter 3

The PCR fragment of primer set 228p16-38 was subsequently labeled with [α- 32P]dCTP by random priming and used as a probe on Southern blots containing BglII- and HindIII-digested parental and hybrid cell-derived DNAs.

3’RACE and RT–PCR analyses Total RNA was isolated from the t(12;15)(q13;q25)-positive lymphoblastoid cell line and a control lymphoblastoid cell line using RNAzol B (Campro Scientific, Veenendaal, The Netherlands) according to the instructions of the manufacturer. RT- PCR was performed using Superscript II (Life Technologies, Gaithersburg, MD, USA), 5 μg of total RNA, the oligodT primer (AAGGATCCGTCGACATCTTTTTTTTTTTTTTTTT) and a random hexamer-mix (Invitrogen, Breda, The Netherlands), again according to the instructions of the manufacturer. 3’RACE analysis was carried out in two PCR steps. The first PCR was performed with a specific forward primer located in the first exon of SENP1 (GACTCTTCCGGTGCTGT) and a primer that hybridizes to the oligodT primer sequences (CUACUACUACUAAAGGATCCGTCGACATC). This PCR was carried out under standard conditions using 10 pmol of priming oligonucleotides. Samples were amplified during 30 cycles in a Perkin Elmer 2400 GeneAmp PCR System. Each cycle included: 1 minute at 94°C, 1 minute at 55°C and 3 minutes at 72°C. In the first cycle an extra denaturation step (94°C) was included for 1 minute and the last cycle was extended at 72°C for 5 minutes. Subsequently, the PCR sample was diluted 10 times and used for the second PCR step, a hemi-nested PCR. Therefore, the following primers were used: a nested forward primer located in the first exon of SENP1 (CATCATCATCATCTGTGAAGGCGGTTCC) and the reverse primer that hybridizes to the oligodT primer sequences. PCR analyzes (35 cycles) included: 30 seconds at 94°C, 30 seconds at 55°C and 3 minutes at 72°C. In the first cycle an extra denaturation step (94°C) was included for 2 minutes and the last cycle was extended at 72°C for 5 minutes. The PCR samples were loaded onto standard agarose-gels and bands were sliced out, purified by standard procedures, ligated into pGEM-T vectors (Promega, Leiden, The Netherlands) according to the instructions of the manufacturer, and sequenced using SP6-and T7-specific primers. For additional PCR analyzes the following primers were used: Sprot-exon5R; GAGGTCTTTCGGGTTTCGAGG Sprot-exon3R; AGCGAAAGCTGGTCCTCTGG Sprot-raceFnest; CGCATGGCAGCCGGTTCCG KIAA0081-exon3R; TTGCCCTTGTCTTGCTTTG

56 Fusion of SENP1 and MESDC2 in an infantile teratoma with constitutional t(12;15)(q13;q25)

KIAA0081-exon1F; TTGTGCCTCTGACCTGCTGC KIAA0081-exon2R; CCTTCTCAGTAGGGCTTCC.

Plasmid constructs On the basis of the published sequences of SENP1 [36] and MESDC2 (GenBank accession no.: KIAA0081, BC009210) [37;38] cDNAs of SENP1, SENP1-MESDC2 (SEME), MESDC2-SENP1 (MESE) were generated by RT-PCR on RNA extracted from the t(12;15)(q13;q25)-positive cell line using the oligodT primer, the random hexamer-mix and Sprotex1-EcoRI-F (CGGAATTCTCTTCCGGTGCTGTGAAGG), Sprotex18-NheI-R (CTAGCTAGCTAGGAGTTTTCGGTGGAGGATC), 5-BC009210-F (GGAATTCGGTAAGCGCGTCTAGGG). For the MESDC2 construct cDNA was derived from a clone (3637449) of the I.M.A.G.E. Consortium [58], using a primer specific for the SP6 promoter and 3-Mesdc2-R (GACTAGTGTCTTCTCTTTTATTCCCAGC). These cDNAs were cloned into a eukaryotic expression vector PSG8 [59] in frame with a VSV-tag at their 3’-end. His- PML and Ha-SUMO-1 constructs were described before [36;60].

Cellular localization assays HeLa cells were transfected directly on glass slides using FuGENE 6 (Roche Diagnostics, Mannheim, Germany) according to the instructions of the manufacturer. Immunofluorescence assays were performed essentially as described before [61]. The following antibodies were used for immunostaining: the monoclonal mouse anti- VSV antibody (1:1000) (Sigma Genosys, Cambridge, UK), the rabbit anti-PDI (1:10) [62], followed by appropriate FITC or Texas Red-conjugated goat-anti-mouse or goat- anti-rabbit antibodies (1:100) (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA). Images were recorded using a Zeiss LSM510 META confocal microscope running software release 3.2.

PML desumoylation assay COS-1 cells were transfected by nucleofection (Amaxa, Koeln, Germany) according to the instructions of the manufacturer with one or more of the following constructs (displayed in Fig. 4): His-PML, Ha-SUMO-1, MESE-VSV (11.4), MESE-VSV (1.2), WT SENP1-VSV or empty vector (PSG8). Approximately 24 hours after transfection the cells were washed in ice-cold PBS and lysed in ice-cold lysisbuffer (50 mM Hepes ph 7.5, 150 mM NaCl, 1.5 mM MgCl2, 1 mM EDTA, 0.5% NP-40, 0.3% Triton X-100) supplemented with protease inhibitors (100x dilution of 0,5 M DTE, 2M

57 Chapter 3 ethylmaleimide, complete cocktail (Roche Diagnostics, Mannheim, Germany), 100 mM PMSF). Before IP, aliquotes of the cell lysates were used to assay the expression levels from the different SENP1-VSV constructs. For IP-mediated extraction of the PML conjugates, a monoclonal anti-polyhistidine antibody (Sigma Genosys, Cambridge, UK) was added to the cell lysates (100x dilution) and mixed by tumbling for minimal 4 hours at 4˚C. Subsequently, protein A beads (Sigma Genosys, Cambridge, UK) were added (10% v/v) after washing with ice-cold lysisbuffer, and mixed by tumbling for an additional 1-2 hours at 4˚C. After several washes, the samples (before and after IP) were boiled for 5 minutes in SDS sample buffer before loading onto SDS polyacrylamide gels (7.5%) and, subsequently, transferred to PROTRAN nitrocellulose membranes (Schleicher & Schuell, Dassel, Germany). For immunodetection of the samples on the membranes, a mouse anti-Ha (1:1000), a mouse monoclonal anti-polyhistidine (1:3000) and an anti-VSV (1:10) (P5D4) antibody were used. As secondary antibodies a rabbit anti-mouse-HRP or a swine- anti-rabbit-HRP were used (1:1000) (DAKO, Denmark). Immunostaining was performed using chemoluminesence.

ACKNOWLEDGEMENTS

The authors thank Marga Schepens for expert technical assistance, Martien van Asseldonk for advice and support, Jack Fransen for antibodies, advice and support and Lutgarde Govaerts, Cokkie Wouters, Rosalyn Slater and Jan van Hemel for referring this case to us and for providing patient material.

REFERENCES

1. Looijenga, L.H., Oosterhuis, J.W. (1999): Pathogenesis of testicular germ cell tumours. Rev.Reprod. 4: 90-100. 2. Oosterhuis, J.W., Looijenga, L.H., van Echten, J., and de Jong, B. (1997): Chromosomal constitution and developmental potential of human germ cell tumors and teratomas. Cancer Genet.Cytogenet. 95: 96-102. 3. Schneider, D.T., Calaminus, G., Koch, S., Teske, C., Schmidt, P., Haas, R.J., Harms, D., and Gobel, U. (2004): Epidemiologic analysis of 1,442 children and adolescents registered in the German germ cell tumor protocols. Pediatr.Blood Cancer 42: 169-175. 4. van Echten, J., Timmer, A., van der Veen, A.Y., Molenaar, W.M., and de Jong, B. (2002): Infantile and adult testicular germ cell tumors. a different pathogenesis? Cancer Genet.Cytogenet. 135: 57-62.

58 Fusion of SENP1 and MESDC2 in an infantile teratoma with constitutional t(12;15)(q13;q25)

5. Veltman, I.M., Schepens, M.T., Looijenga, L.H.J., Strong, L.C., and Geurts van Kessel, A. (2003): Germ cell tumours in neonates and infants: a distinct subgroup? APMIS 111: 152-160. 6. Perlman, E.J., Valentine, M.B., Griffin, C.A., and Look, A.T. (1996): Deletion of 1p36 in childhood endodermal sinus tumors by two-color fluorescence in situ hybridization: a pediatric oncology group study. Genes Chromosomes.Cancer 16: 15-20. 7. Perlman, E.J., Hu, J., Ho, D., Cushing, B., Lauer, S., and Castleberry, R.P. (2000): Genetic analysis of childhood endodermal sinus tumors by comparative genomic hybridization. J.Pediatr.Hematol.Oncol. 22: 100-105. 8. Mostert, M., Rosenberg, C., Stoop, H., Schuyer, M., Timmer, A., Oosterhuis, J., and Looijenga, L. (2000): Comparative genomic and in situ hybridization of germ cell tumors of the infantile testis. Lab Invest. 80: 1055-1064. 9. Schneider, D.T., Schuster, A.E., Fritsch, M.K., Hu, J., Olson, T., Lauer, S., Gobel, U., and Perlman, E.J. (2001): Multipoint imprinting analysis indicates a common precursor cell for gonadal and nongonadal pediatric germ cell tumors. Cancer Res. 61: 7268-7276. 10. Damjanov, I. (1993): Teratocarcinoma: neoplastic lessons about normal embryogenesis. Int.J.Dev.Biol. 37: 39-46. 11. Andrews, P.W. (2002): From teratocarcinomas to embryonic stem cells. Philos.Trans.R.Soc.Lond B Biol.Sci. 357: 405-417. 12. Tzukerman, M., Rosenberg, T., Ravel, Y., Reiter, I., Coleman, R., and Skorecki, K. (2003): An experimental platform for studying growth and invasiveness of tumor cells within teratomas derived from human embryonic stem cells. Proc.Natl.Acad.Sci.U.S.A 100: 13507-13512. 13. Langa, F., Kress, C., Colucci-Guyon, E., Khun, H., Vandormael-Pournin, S., Huerre, M., and Babinet, C. (2000): Teratocarcinomas induced by embryonic stem (ES) cells lacking vimentin: an approach to study the role of vimentin in tumorigenesis. J.Cell Sci. 113: 3463-3472. 14. Bussey, K.J., Lawce, H.J., Olson, S.B., Arthur, D.C., Kalousek, D.K., Krailo, M., Giller, R., Heifetz, S., Womer, R., and Magenis, R.E. (1999): Chromosome abnormalities of eighty-one pediatric germ cell tumors: sex- , age-, site-, and histopathology-related differences - a children's cancer group study. Genes Chromosomes. Cancer 25: 134-146. 15. Schneider, D.T., Schuster, A.E., Fritsch, M.K., Calaminus, G., Gobel, U., Harms, D., Lauer, S., Olson, T., and Perlman, E.J. (2002): Genetic analysis of mediastinal nonseminomatous germ cell tumors in children and adolescents. Genes Chromosomes. Cancer 34: 115-125. 16. Hu, J., Schuster, A.E., Fritsch, M.K., Schneider, D.T., Lauer, S., and Perlman, E.J. (2001): Deletion mapping of 6q21-26 and frequency of 1p36 deletion in childhood endodermal sinus tumors by microsatellite analysis. Oncogene 20: 8042-8044. 17. Perlman, E.J., Cushing, B., Hawkins, E., and Griffin, C.A. (1994): Cytogenetic analysis of childhood endodermal sinus tumors: a pediatric oncology group study. Pediatr.Pathol. 14: 695-708. 18. Veltman, I., Veltman J., Janssen I., Hulsbergen-van de Kaa C., Oosterhuis W., Schneider D., Stoop H., Gillis A., Zahn S., Looijenga L. et al. (2005): Identification of recurrent chromosomal aberrations in germ cell tumors of neonates and infants using genome-wide array-based comparative genomic hybridization. Genes Chromosomes. Cancer 43:367–376. 19. Looijenga, L.H., Zafarana, G., Grygalewicz, B., Summersgill, B., Debiec-Rychter, M., Veltman, J., Schoenmakers, E.F., Rodriguez, S., Jafer, O., Clark, J. et al. (2003): Role of gain of 12p in germ cell tumour development. APMIS 111: 161-171. 20. Honorio, S., Agathanggelou, A., Wernert, N., Rothe, M., Maher, E.R., and Latif, F. (2003): Frequent epigenetic inactivation of the RASSF1A tumour suppressor gene in testicular tumours and distinct methylation profiles of seminoma and nonseminoma testicular germ cell tumours. Oncogene 22: 461-466. 21. Looijenga, L.H., de Leeuw, H., van Oorschot, M., Van Gurp, R.J., Stoop, H., Gillis, A.J., Gouveia Brazao, C.A., Weber, R.F., Kirkels, W.J., van Dijk, T. et al. (2003): Stem cell factor receptor (c-KIT) codon 816 mutations predict development of bilateral testicular germ-cell tumors. Cancer Res. 63: 7674-7678. 22. Schmidt, B.A., Rose, A., Steinhoff, C., Strohmeyer, T., Hartmann, M., and Ackermann, R. (2001): Up-regulation of cyclin-dependent kinase 4/cyclin D2 expression but down-regulation

59 Chapter 3

of cyclin-dependent kinase 2/cyclin E in testicular germ cell tumors. Cancer Res. 61: 4214- 4221. 23. Zafarana, G., Gillis, A.J., Van Gurp, R.J., Olsson, P.G., Elstrodt, F., Stoop, H., Millan, J.L., Oosterhuis, J.W., and Looijenga, L.H. (2002): Coamplification of DAD-R, SOX5, and EKI1 in human testicular seminomas, with specific overexpression of DAD-R, correlates with reduced levels of apoptosis and earlier clinical manifestation. Cancer Res. 62: 1822-1831. 24. Oosterhuis, J.W., Looijenga, L.H. (2003): Current views on the pathogenesis of testicular germ cell tumours and perspectives for future research: highlights of the 5th Copenhagen Workshop on Carcinoma in situ and Cancer of the Testis. APMIS 111: 280-289. 25. Rapley, E.A., Crockford, G.P., Teare, D., Biggs, P., Seal, S., Barfoot, R., Edwards, S., Hamoudi, R., Heimdal, K., Fossa, S.D. et al. (2000): Localization to Xq27 of a susceptibility gene for testicular germ-cell tumours. Nat.Genet. 24: 197-200. 26. Rapley, E.A., Crockford, G.P., Easton, D.F., Stratton, M.R., and Bishop, D.T. (2003): Localisation of susceptibility genes for familial testicular germ cell tumour. APMIS 111: 128- 133. 27. Murty, V.V., Houldsworth, J., Baldwin, S., Reuter, V., Hunziker, W., Besmer, P., Bosl, G., and Chaganti, R.S. (1992): Allelic deletions in the long arm of chromosome 12 identify sites of candidate tumor suppressor genes in male germ cell tumors. Proc.Natl.Acad.Sci.U.S.A 89: 11006-11010. 28. Samaniego, F., Rodriguez, E., Houldsworth, J., Murty, V.V., Ladanyi, M., Lele, K.P., Chen, Q.G., Dmitrovsky, E., Geller, N.L., and Reuter, V. (1990): Cytogenetic and molecular analysis of human male germ cell tumors: chromosome 12 abnormalities and gene amplification. Genes Chromosomes.Cancer 1: 289-300. 29. Rodriguez, E., Mathew, S., Reuter, V., Ilson, D.H., Bosl, G.J., and Chaganti, R.S. (1992): Cytogenetic analysis of 124 prospectively ascertained male germ cell tumors. Cancer Res. 52: 2285-2291. 30. Betts, D.R., Greiner, J., Feldges, A., Caflisch, U., and Niggli, F.K. (2001): Constitutional balanced chromosomal rearrangements and neoplasm in children. J.Pediatr.Hematol.Oncol. 23: 582-584. 31. Bodmer, D., van den Hurk, H.W., van Groningen, J.J., Eleveld, M.J., Martens, G.J., Weterman, M.A., and Geurts van Kessel, A. (2002): Understanding familial and non-familial renal cell cancer. Hum.Mol.Genet. 11: 2489-2498. 32. Ganly, P., McDonald, M., Spearing, R., and Morris, C.M. (2004): Constitutional t(5;7)(q11;p15) rearranged to acquire monosomy 7q and trisomy 1q in a patient with myelodysplastic syndrome transforming to acute myelocytic leukemia. Cancer Genet.Cytogenet. 149: 125-130. 33. Roberts, T., Chernova, O., and Cowell, J.K. (1998): NB4S, a member of the TBC1 domain family of genes, is truncated as a result of a constitutional t(1;10)(p22;q21) chromosome translocation in a patient with stage 4S neuroblastoma. Hum.Mol.Genet. 7: 1169-1178. 34. Veltman, I., van Asseldonk, M., Schepens, M., Stoop, H., Looijenga, L., Wouters, C., Govaerts, L., Suijkerbuijk, R., and Geurts van Kessel, A. (2002): A novel case of infantile sacral teratoma and a constitutional t(12;15)(q13;q25) pat. Cancer Genet.Cytogenet. 136: 17-22. 35. Veltman, I.M., Veltman, J.A., Arkesteijn, G., Janssen, I.M., Vissers, L.E., de Jong, P.J., Geurts van Kessel, A., and Schoenmakers, E.F. (2003): Chromosomal breakpoint mapping by arrayCGH using flow-sorted chromosomes. Biotechniques 35: 1066-1070. 36. Gong, L., Millas, S., Maul, G.G., and Yeh, E.T. (2000): Differential regulation of sentrinized proteins by a novel sentrin- specific protease. J.Biol.Chem. 275: 3355-3359. 37. Wines, M.E., Lee, L., Katari, M.S., Zhang, L., DeRossi, C., Shi, Y., Perkins, S., Feldman, M., McCombie, W.R., and Holdener, B.C. (2001): Identification of mesoderm development (mesd) candidate genes by comparative mapping and genome sequence analysis. Genomics 72: 88-98. 38. Hsieh, J.C., Lee, L., Zhang, L., Wefer, S., Brown, K., DeRossi, C., Wines, M.E., Rosenquist, T., and Holdener, B.C. (2003): Mesd encodes an LRP5/6 chaperone essential for specification of mouse embryonic polarity. Cell 112: 355-367. 39. Bailey, D., O'Hare, P. (2004): Characterization of the localization and proteolytic activity of the SUMO-specific protease, SENP1. J.Biol.Chem. 279: 692-703.

60 Fusion of SENP1 and MESDC2 in an infantile teratoma with constitutional t(12;15)(q13;q25)

40. Nefkens, I., Negorev, D.G., Ishov, A.M., Michaelson, J.S., Yeh, E.T., Tanguay, R.M., Muller, W.E., and Maul, G.G. (2003): Heat shock and Cd2+ exposure regulate PML and Daxx release from ND10 by independent mechanisms that modify the induction of heat-shock proteins 70 and 25 differently. J.Cell Sci. 116: 513-524. 41. Ayaydin, F., Dasso, M. (2004): Distinct in vivo dynamics of vertebrate SUMO paralogues. Mol.Biol.Cell 15: 5208-5218. 42. Kamitani, T., Kito, K., Nguyen, H.P., Fukuda-Kamitani, T., and Yeh, E.T. (1998): Characterization of a second member of the sentrin family of ubiquitin- like proteins. J.Biol.Chem. 273: 11349-11353. 43. Gill, G. (2004): SUMO and ubiquitin in the nucleus: different functions, similar mechanisms? Genes Dev. 18: 2046-2059. 44. Muller, S., Ledl, A., and Schmidt, D. (2004): SUMO: a regulator of gene expression and genome integrity. Oncogene 23: 1998-2008. 45. Yeh, E.T., Gong, L., and Kamitani, T. (2000): Ubiquitin-like proteins: new wines in new bottles. Gene 248: 1-14. 46. Xu, Z., Au, S.W. (2004): Mapping residues of SUMO precursors essential in differential maturation by SUMO-specific protease, SENP1. Biochem.J. 386:325-30. 47. Bachant, J., Alcasabas, A., Blat, Y., Kleckner, N., and Elledge, S.J. (2002): The SUMO-1 isopeptidase Smt4 is linked to centromeric cohesion through SUMO-1 modification of DNA topoisomerase II. Mol.Cell 9: 1169-1182. 48. Hang, J., Dasso, M. (2002): Association of the human SUMO-1 protease SENP2 with the nuclear pore. J.Biol.Chem. 277: 19961-19966. 49. Kim, K.I., Baek, S.H., Jeon, Y.J., Nishimori, S., Suzuki, T., Uchida, S., Shimbara, N., Saitoh, H., Tanaka, K., and Chung, C.H. (2000): A new SUMO-1-specific protease, SUSP1, that is highly expressed in reproductive organs. J.Biol.Chem. 275: 14102-14106. 50. Nishida, T., Tanaka, H., and Yasuda, H. (2000): A novel mammalian Smt3-specific isopeptidase 1 (SMT3IP1) localized in the nucleolus at interphase. Eur.J.Biochem. 267: 6423- 6427. 51. Tagawa, H., Miura, I., Suzuki, R., Suzuki, H., Hosokawa, Y., and Seto, M. (2002): Molecular cytogenetic analysis of the breakpoint region at 6q21-22 in T-cell lymphoma/leukemia cell lines. Genes Chromosomes.Cancer 34: 175-185. 52. Jacques, C., Baris, O., Prunier-Mirebeau, D., Savagner, F., Rodien, P., Rohmer, V., Franc, B., Guyetant, S., Malthiery, Y., and Reynier, P. (2005): Two-step differential expression analysis reveals a new set of genes involved in thyroid oncocytic tumors. J.Clin.Endocrinol.Metab 90:2314-20. 53. Holdener, B.C., Faust, C., Rosenthal, N.S., and Magnuson, T. (1994): msd is required for mesoderm induction in mice. Development 120: 1335-1346. 54. Herz, J., Marschang, P. (2003): Coaxing the LDL receptor family into the fold. Cell 112: 289- 292. 55. Geurts van Kessel, A., Tetteroo, P.A., dem Borne, A.E., Hagemeijer, A., and Bootsma, D. (1983): Expression of human myeloid-associated surface antigens in human-mouse myeloid cell hybrids. Proc.Natl.Acad.Sci.U.S.A 80: 3748-3752. 56. de Bruijn, D.R., Kater-Baats, E., Eleveld, M., Merkx, G., and Geurts van Kessel, A. (2001): Mapping and characterization of the mouse and human SS18 genes, two human SS18-like genes and a mouse Ss18 pseudogene. Cytogenet. Cell Genet. 92: 310-319. 57. Weterman, M.A., Wilbrink, M., and Geurts van Kessel, A. (1996): Fusion of the transcription factor TFE3 gene to a novel gene, PRCC, in t(X;1)(p11;q21)-positive papillary renal cell carcinomas. Proc.Natl.Acad.Sci.U.S.A 93: 15294-15298. 58. Bonaldo, M.F., Lennon, G., and Soares, M.B. (1996): Normalization and subtraction: two approaches to facilitate gene discovery. Genome Res. 6: 791-806. 59. Cuppen, E., Gerrits, H., Pepers, B., Wieringa, B., and Hendriks, W. (1998): PDZ motifs in PTP- BL and RIL bind to internal protein segments in the LIM domain protein RIL. Mol.Biol.Cell 9: 671-683.

61 Chapter 3

60. Kamitani, T., Kito, K., Nguyen, H.P., Wada, H., Fukuda-Kamitani, T., and Yeh, E.T. (1998): Identification of three major sentrinization sites in PML. J.Biol.Chem. 273: 26675-26682. 61. Weterman, M.J., van Groningen, J.J., Jansen, A., and Geurts van Kessel, A. (2000): Nuclear localization and transactivating capacities of the papillary renal cell carcinoma-associated TFE3 and PRCC (fusion) proteins. Oncogene 19: 69-74. 62. John, D.C., Grant, M.E., and Bulleid, N.J. (1993): Cell-free synthesis and assembly of prolyl 4- hydroxylase: the role of the beta-subunit (PDI) in preventing misfolding and aggregation of the alpha-subunit. EMBO J. 12: 1587-1595.

Positional cloning of a constitutional t(8;12)(q23;q12) in a patient with an infantile sacral teratoma

Imke Veltman

Lilian Vreede

Tom Harms

Harry Drabkin

Louise Strong

Eric Schoenmakers

Ad Geurts van Kessel

Chapter 4

ABSTRACT

Recently, we identified a patient with an infantile germ cell tumour (sacral teratoma) and a constitutional t(8;12)(q23;q12). Since the 12q11-q13 region is thought to harbour one or more genes relevant for germ cell tumour development, we set out to positionally clone this novel constitutional translocation. Using FISH and Southern blot analyses in conjunction with vectorette-PCR we were able to position the 8q23 breakpoint ~13 kb upstream of the EIF3S6 gene and the 12q12 breakpoint ~50kb downstream of the DKFZP434G1415 gene. As yet, no breakpoint-spanning (fusion) genes could be identified. Preliminary northern blot analyses revealed that the transcription levels of several positional candidate genes, including EIF3S6 and DKFZP434G1415, were not significantly affected by the translocation. We conclude that additional information is required to definitely establish the role of this constitutional translocation and its associated genes in the development of infantile sacral teratomas and/or other related germ cell tumours.

64 Positional cloning of a constitutional t(8;12)(q23;q12) in an infantile sacral teratoma

INTRODUCTION

Human germ cell tumours (GCTs) comprise a heterogeneous group of tumours that differ with respect to their age of onset, anatomical location, histology, chromosomal constitution and clinical behaviour. Based on these differences, human GCTs can be categorized into five different groups: (1) teratomas and yolk sac tumours (YSTs) in neonates and infants (2) GCTs of the adult testicular type, i.e., seminomas and non- seminomas, (3) spermatocytic seminomas in elderly men, (4) dermoid cysts (mature cystic teratomas) of the ovary, and (5) hydatidiform moles of the placenta/uterus [1;2]. The neonatal/infantile group of GCTs is relatively rare and these tumours mainly arise in the sacral/sacrococcygeal region. Previous molecular cytogenetic studies have shown that the teratomas of this group are predominantly diploid with only a few non-recurrent genomic anomalies [3]. In contrast, we and others found that the YSTs of this group frequently exhibit aneuploidies, including recurrent losses of 1p and 6q, and recurrent gains of 20q [4-8]. Testicular GCTs of juveniles and adults are the most frequently occurring GCTs with numerical and structural anomalies of chromosome 12, in particular i(12p), as their genetic hallmark [9-11]. Recently, we detected a 6 year old patient with a sacral teratoma and a constitutional t(8;12)(q23;q12) translocation [12]. The 12q12 breakpoint region in this patient is of particular interest since several studies have implied the presence of one or more genes relevant for GCT development within this region. Cytogenetic analysis of sixty-five male GCTs revealed eight cases with a deletion of the 12q11-q22 region and three cases with a translocation of the 12q11-q13 region [13;14]. In a follow-up study of forty-five male GCTs frequent losses of heterozygosity (LOHs) (>40%) were observed in the 12q11-q13 and 12q22 regions, respectively [15]. Allelic losses of the 12q11-q13 and 12q22 regions were also detected in an additional genetic study of male GCTs [16]. In the Mitelman Database of Chromosome Aberrations in Cancer (http://cgap.nci.nih.gov/Catalog) a total of nineteen GCTs with 12q11-q13 aberrations are listed, including seven cases with translocations, four cases with additions and six cases with deletions [17]. Interestingly, the 12q11-q13 abnormalities listed in the Mitelman Database, including those reported by Chaganti et al. [13-16] almost exclusively occur in non-seminomas or combined GCTs, with a highest incidence of teratomas (eight out of nineteen cases). Here, the positional cloning and molecular characterization of the t(8;12)(q23;q12)-associated breakpoint regions in a patient with an infantile sacral teratoma are presented.

65 Chapter 4

MATERIALS AND METHODS

Patient material and cell lines A male patient was diagnosed with a sacral teratoma at the age of 6 years. Subsequent cytogenetic analysis revealed the presence of a constitutional t(8;12)(q23;q12). A t(8;12)(q23;q12)-positive cell line was established after in vitro Epstein-Barr virus (EBV) transformation of patient-derived blood lymphocytes using standard procedures. Control lymphoblastoid cell lines were similarly derived from translocation-negative healthy donors.

FISH analysis FISH analyses were carried out using routine procedures as described before [18;19]. Metaphase spreads from the t(8;12)(q23;q12)-positive lymphoblastoid cell line were hybridized with the following probes on 12q12-q13: RP11-815o9, RP11-609i23, RP11- 510p12, RP11-73b8, RP11-350f4 (BACPAC Resources, Oakland, USA), and on 8q23: RP11-79e1, RP11-91k1, RP11-79o20, RP11-79c21, RP11-80d8, RP11-79f7 (BACPAC Resources, Oakland, USA). Chromosome 8- and 12-specific centromeric probes were used for reference. A Zeiss epifluorescence microscope equipped with appropriate filters was used for visual examination of the slides. Digital images were captured using a high-performance cooled CCD camera (Photometrics) coupled to a Macintosh Quadra 950 computer and processed using the Oncor-image software (Gaithersburgh, MD, USA).

Southern and Northern blot analyses DNA and RNA were isolated from the t(8;12)(q23;q12)-positive and control lymphoblastoid cell lines as described before [20]. Subsequently, the RNAs and DNAs were subjected to Southern and Northern blot analyses, respectively, using standard procedures [20]. For gene-specific breakpoint analysis, PCR fragments cloned into pGEMT vectors (Promega, Leiden, The Netherlands) were used as probes after labelling with [α-32P]dCTP by random priming. To this end, the following primer sets were employed: EIF3S6: EIF3S6-exon1F; 5’CAAGATGGCGGAGTACGAC 3’, EIF3S6-exon13R; 5’ATCTTGAGTTGCCCAGTTAGG 3’. DKFZP434G1415: DKFZP434G1415F1; 5’ACCGAGGCACTGTTATAGAAG 3’, DKFZP434G1415R1; 5’GTCCATAGTAATTCACAAAGCC 3’. PTK9: PTK9-F2; 5’GAGCAACTTGTGATTGGATC 3’,

66 Positional cloning of a constitutional t(8;12)(q23;q12) in an infantile sacral teratoma

PTK9-R1; 5’TTAAGGTCTTTCTGTCCCAC 3’. LOC51135/IRAK4: LOC51135-F1; 5’AGGAATAGAAGATGAACAAACC 3’, LOC51135-R1; 5’GTGAACCTTCTTAATGTCTGG 3’. In addition, genomic PCR fragments generated by the following primer sets were used as probes for detailed analysis of the 8q23 breakpoint: Chr8-107: Chr8-107F; 5’GCTGCCTTAAAACCATCGTC 3’, Chr8-107R; 5’GCCACTACTGTGGTGCACAT 3’. Chr8-105: Chr8-105F; 5’ACAGAGTTTGAAGGCTAAGTGACG 3’, Chr8-105R; 5’GATTGACAATTTGGGCCACC 3’.

Vectorette-PCR A vectorette-PCR procedure was employed essentially as described before [21], with minor modifications. Briefly, RsaI-digested DNA of the t(8;12)(q23;q12)-positive cell line was ligated o/n at 16°C to a phosphorylated Vectorette Top Strand Primer Blunt vector: (VTSP-blunt: 5’CAAGGAGAGGACGCTGTCTGTCGAAGGTAAGGAACGGACGAGAGAAG GGAGAG3’) and a phosphorylated Vectorette Bottom Strand Primer vector (VBSP: 5’CTCTCCCTTCTCGAATCGTAACCGTTCGTACGAGAATCGCTGTCCTCTCCTTG 3’). The resulting vectorette libraries were amplified in a hemi-nested PCR reaction, using a Universal Primer: UVP; 5’CGA ATC GTA ACC GTT CGT ACG AGA ATC GCT 3’, and sequence specific primers: 8q23-vect1F; 5’TCCTTGACAGAGTTTGAAGGCTAAGTGACG 3’, 8q23-vect2F; 5’TTGGATGCCCTGCCAAAGAACTTAGGATTG 3’. Subsequently, PCR fragments were separated on agarose gels, purified, ligated into a pGEMT vector (Promega, Leiden, The Netherlands), and sequenced. The genomic breakpoints were verified by PCR amplification using 12q12- and 8q23-specific primers: 12q12-1F; 5’TATGTGTTTTATCTCATGTGTGGC 3’, 8q23-1R; 5’TTGACTCTCCTGGTGTTCAGTTCTAG 3’, 8q23-1F;5’TCCTTGACAGAGTTTGAAGGCTAAGTGACG3’ and 12q12-1R; 5’AATAATACATGGTGGCTAAGAG 3’, respectively.

67 Chapter 4

RESULTS AND DISCUSSION

Fine mapping of the constitutional t(8;12)(q23;q12) translocation Previously, we identified CEPH893a3 as the yeast artificial clone (YAC) that spans the t(8;12)(q23;q12)-associated breakpoint in 12q12 (previously assigned to 12q13; Fig. 1A) [12]. In order to further refine the 12q12 breakpoint region, a BAC contig was assembled that corresponds to this YAC. In addition, a BAC contig was selected within the 8q23 region for the characterization of the 8q23 breakpoint (Fig. 1B). FISH analysis on metaphase spreads of the t(8;12)(q23;q12)-positive cell line revealed that BAC clones RP11-815o9, RP11-609i23, RP11-510p12 and RP11-73b8 map proximal to the 12q12 breakpoint. BAC clone RP11-350f4 appeared to span the 12q12 breakpoint, as this clone hybridized, next to the normal chromosome 12, to both translocation derivative chromosomes der(8) and der(12). Similarly, BAC clones RP11-79e1, RP11- 91k1, RP11-79o20 and RP11-79c21 were found to map proximal, and RP11-79f7 distal to the 8q23 breakpoint, respectively. The intermediate BAC clone RP11-80d8 turned out to span the breakpoint in 8q23. This was again revealed by split FISH signals on both der(8) and der(12), next to a signal on normal chromosome 8 (Fig. 1C). RP11-350f4 (GenBank accession no. AC016143) is a completely sequenced BAC and measures 55 kb. According to the May 2004 freeze of the UCSC human genome browser this BAC contains one predicted gene of which the mRNA (BC068502) encodes a hypothetical protein (DKFZP434G1415). RP11-80d8 (GenBank accession no. AC083931) encompasses three unordered stretches of DNA that were assembled in the May 2004 freeze of the UCSC human genome browser based on the completely sequenced clones kb1153c10 (GenBank accession no. AP001331) and RP11-35g22 (GenBank accession no. AC087620). Together, RP11-80d8 is approximately 152 kb in size and, as yet, it only contains one annotated gene, EIF3S6. This gene codes for the eukaryotic translation initiation factor 3 subunit 6, which binds to the 40S ribosome and promotes the binding of methionyl-tRNAi and mRNA. The EIF3S6 gene is also known as INT6 and, interestingly, its disruption by the mouse mammary tumour virus (MMTV) has been implicated in tumour development [22-24]. In addition, down-regulation of EIF3S6 has been observed in a considerable percentage of human breast and lung cancers [23].

68 Positional cloning of a constitutional t(8;12)(q23;q12) in an infantile sacral teratoma

Chr. 12 Breakpoint BAC contig Chr. 8 BAC contig spanning YAC

RP11-815o9 RP11-79e1

RP11-91k1 RP11-609i23 der(12) RP11-79o20 q12- q13 893a3 RP11-510p12 RP11-79c21 der(8)

RP11-80d8 RP11-73b8 8 RP11-80d8 q23 RP11-350f4 RP11-79f7

AB C

Figure 1. FISH mapping of the 12q12 and 8q23 breakpoints. A) Diagram depicting chromosome 12, the 12q12 breakpoint-spanning CEPH YAC clone 893a3 and its corresponding BAC contig. In grey the breakpoint spanning clones (893a3 and RP11-350f4) are shown. B) Diagram depicting chromosome 8 and the selected BAC contig within 8q23, including the breakpoint-spanning BAC clone RP11-80d8 (grey). C) FISH analysis on a t(8;12)(q23;q12)-positive metaphase spread, using RP11-80d8 as a probe. Positive signals (green) are visible on the normal chromosome 8, der(8) and der(12). The centromere of chromosome 8 is marked in red and the chromosomes are counterstained with DAPI (blue).

The 12q12 and 8q23 breakpoints map outside known genes Southern blot analysis was performed to examine whether the 12q12 and 8q23 breakpoints disrupt one or both of the known candidate genes located on the breakpoint-spanning BAC clones RP11350f4 and RP1180d8, i.e., DKFZP434G1415 and EIF3S6, respectively. Therefore, RT-PCR fragments corresponding to these genes were labelled and hybridized to Southern blots containing cleaved DNAs extracted from the t(8;12)(q23;q12)-positive cell line and control cell lines derived from translocation-negative healthy donors. No aberrantly hybridizing fragments were detected on any of the blots containing different digests of the t(8;12)(q23;q12)- positive cell line as compared to the control cell lines (not shown). Since the above selected probes only partially covered the breakpoint- spanning BAC clones, additional probes were selected based on repeat-low sequences using the CENSOR database (http://www.girinst.org/Censor_Server.html) [25]. Using the chromosome 8-specific

69 Chapter 4

PCR fragment Chr8-107F/R as a probe on Southern blots, wild-type fragments were detected in both BglII and AseI-digested DNAs extracted from the t(8;12)(q23;q12)- positive and translocation-negative cell lines, as expected (Fig.2A: 7.5kb, 4.1kb and 3.4 kb, respectively). In addition, aberrantly hybridizing fragments of 11kb and 2kb were readily detected in the BglII and AseI-digested DNAs extracted from the t(8;12)(q23;q12)-positive cell line, respectively, indicating that the 8q23 breakpoint must be located within these fragments. In addition, using the chromosome 8-specific PCR fragment Chr8-105F/R as a probe, an aberrantly hybridizing fragment of 1.2 kb was detected in the RsaI-digested DNA extracted from the t(8;12)(q23;q12)-positive cell line. By combining these data, we could narrow down the 8q23 breakpoint to a genomic interval of ~1 kb. 1:t(8;12) 2:control

>11 kb 1:t(8;12) 2:control BglII 7.5 kb

4.1 kb ~1.2 kb 3.4 kb 1 kb AseI >2 kb RsaI-vectorette ~1.2 kb 1 kb RsaI

AB Figure 2. Southern blot and vectorette-PCR analysis of the 8q23 breakpoint region. A) Southern blots containing BglII- AseI- and RsaI-digested genomic DNAs (upper panel) of the t(8;12)(q23;q12)- positive cell line (lane 1) and a translocation-negative control cell line (lane 2), respectively, were hybridized with probe Chr8-107. Another Southern blot containing RsaI-digested DNA (lower panel) was hybridized with probe Chr8-105 (both PCR fragments). Next to the wild-type chromosome 8 fragments (lane 1: 7.5 kb BglII fragment, 4.1 and 3.4 kb AseI-fragments, and 1 kb RsaI fragment) aberrantly hybridizing fragments are readily detected in the t(8;12)(q23;q12)-positive cell line (lane 1: 11 kb BglII fragment, 2 kb AseI-fragment, and 1.2 kb RsaI fragment). B) Vectorette-PCR analysis using RsaI-digested genomic DNAs of the t(8;12)(q23;q12)-positive cell line (lane 1) and a translocation-negative control cell line (lane 2) as templates. A chromosome 8 wild-type fragment of 1 kb is present in both cell lines. In contrast, abnormally sized fragments of approximately 1.2 kb are present in the t(8;12)(q23;q12)-positive cell line

In order to confirm the above Southern blot data and to determine the exact positions of both the 8q23 and 12q12 breakpoints, we set out to molecularly clone the respective breakpoints using vectorette-PCR. Therefore, specific primers corresponding to the 8q23-associated ~1 kb breakpoint interval were used in conjunction with vectorette primers and RSAI-digested DNAs extracted from the t(8;12)(q23;q12)-positive and translocation-negative cell lines. Completely in line with the Southern blot data, PCR fragments of ~1.2 kb were generated only from the

70 Positional cloning of a constitutional t(8;12)(q23;q12) in an infantile sacral teratoma t(8;12)(q23;q12)-positive cell line, next to wild-type fragments of ~1 kb (Fig. 2B). Subsequent cloning and sequencing of this ~1.2 kb fragment indeed revealed a fusion of 8q23- and 12q12-derived genomic sequences (UCSC human genome browser, May 2004 freeze). From the thus obtained exact mapping data we conclude that the breakpoint in 12q12 is located ~50kb downstream of the candidate gene DKFZP434G1415 and that the breakpoint in 8q23 is located ~13 kb upstream of the candidate gene EIF3S6 (Fig. 3A). In addition, we conclude that the 12q12 breakpoint is located ~130 kb upstream of ADAMTS20 and that the 8q23 breakpoint is located ~183 kb upstream of KIAA0103, a putative gene that has been assigned to the neighbouring BAC clone RP11-79c21 (Fig. 1B and 3A).

12q12 8q23

ADAMTS20 RP11- KIAA0103 79c21 l 1 l 2 ) ro ro 12 nt nt 8; RP11-73b8 co co t( 1: 2: 3:

RP11- breakpoint 80d8 breakpoint RP11-350f4 EIF3S6 DKFZP434G1415 IRAK4 PTK9 EIF3S6 AB Figure 3. A) Diagram showing the 12q12 and 8q23 breakpoint-flanking BAC clones and their annotated genes. Directions of transcription are indicated by arrows and breakpoint-spanning BAC clones are marked in grey. B) Northern blot analysis of two translocation-negative control cell lines (lanes 1, 2) and the t(8;12)(q23;q12)-positive cell line (lane 3) using a EIF3S6-specific (exons 1-13) probe. For reference, the position of the 18S ribosomal RNA band is marked (arrowhead).

The respective breakpoint positions were confirmed by direct PCR amplification on DNA extracted from the t(8;12)(q23;q12)-positive cell line, using 12q12- and 8q23-specific primers. Subsequent cloning and sequencing of the reciprocal PCR fragments revealed that the translocation is balanced and that it is not associated with breakpoint-flanking deletions and/or duplications on either one of its derivative chromosomes. In addition, we examined whether the respective breakpoints were positioned within genomic regions that code for known microRNAs (miRNAs), using the May 2004 freeze of the UCSC human genome

71 Chapter 4 browser. These miRNAs represent a novel class of small, functional non-coding RNAs of 19-23 nucleotides that are cleaved from precursors of ~60 to ~110 nucleotides. They appear to be involved in several biological key processes such as cellular proliferation, death and stress resistance. Interestingly, recent literature has indicated that these miRNAs may be involved in tumourigenesis as well [26]. However, we found that the locations of both t(8;12)(q23;q12)-associated breakpoints were situated outside genomic regions that code for known miRNAs.

Normal transcription of candidate genes in cell lines The t(8;12)(q23;q12)-associated breakpoints do not appear to disrupt any of the currently known annotated and/or predicted candidate genes present on the respective breakpoint-spanning BAC clones. However, the expression of these genes, and/or genes located further up- or down-stream of the breakpoints, may still be altered through position effects [27]. Therefore, we set out to determine the transcription levels of several of these candidate genes by Northern blot analysis. Since EIF3S6 is situated closest to the breakpoint in 8q23 and serves as an interesting candidate gene (see above), its transcription levels were first determined in t(8;12)(q23;q12)-positive and translocation-negative cell lines using a EIF3S6-specific probe (Fig. 3B). No significant differences were observed, suggesting that in these lymphoblastoid cell lines the breakpoint does not appear to affect the transcription level of EIF3S6. Similarly, the transcription levels of the 12q12-related genes PTK9 and DKFZP434G1415 were determined. Again, no significant differences between the translocation-positive and -negative cell lines were observed (not shown). After similarly hybridizing the Northern blots with an IRAK4 probe, no transcripts could be detected in any of the cell lines tested. This latter finding may be due to a low expression level of this gene in lymphoblastoid cells. Unfortunately, no other tissues and/or tumour-derived samples of the patient were available for further analysis. Recently, we found that another gene located in the 12q12-q13 region, SENP1, was disrupted and fused to the MESDC2 gene in a patient with an infantile sacrococcygeal teratoma and a constitutional t(12;15)(q13;q25) [19;28]. Therefore, we examined the transcription level of SENP1 in the t(8;12)(q23;q12)-positive cell line as well. Again, no significant differences in transcription levels were detected between the t(8;12)(q23;q12)-positive and translocation-negative cell lines (not shown). In conclusion, we have detected a novel patient with an infantile sacral teratoma and a constitutional t(8;12)(q23;q12). Positional cloning of the breakpoints revealed that the translocation is balanced and that the respective breakpoints are

72 Positional cloning of a constitutional t(8;12)(q23;q12) in an infantile sacral teratoma located upstream of the EIF3S6 gene in 8q23 and downstream of the DKFZP434G1415 gene in 12q12. The transcription levels of positional candidate genes were not significantly affected by the translocation breakpoints in the lymphopblastoid cell lines tested so far. Additional studies are required to establish whether the constitutional t(8;12)(q23;q12) has predisposed this patient to the development of the infantile sacral teratoma and, if so, whether this information will be instrumental for the elucidation of its underlying molecular mechanism(s).

ACKNOWLEDGMENTS

The authors thank Ron Suijkerbuijk and Martien van Asseldonk for advice and support, and Marga Schepens for expert technical assistance.

REFERENCES

1. Oosterhuis JW, Looijenga LH (2005): Testicular germ-cell tumours in a broader perspective. Nat.Rev.Cancer 5:210-222. 2. Woodward PJ, Heidenreich A.,Looijenga L.H.J. et al. (2004): Testicular germ cell tumors. In: World Health Organization Classification of Tumours. Pathology and Genetics of the Urinary System and Male Genital Organs. Eble, ed. IARC Press, Lyon, 217-278. 3. Veltman I, Veltman J, Janssen I, Hulsbergen-van de Kaa C, Oosterhuis W, Schneider D, Stoop H, Gillis A, Zahn S, Looijenga L, Gobel U, Geurts van Kessel A (2005): Identification of recurrent chromosomal aberrations in germ cell tumors of neonates and infants using genomewide array-based comparative genomic hybridization. Genes Chromosomes Cancer 43:367-76. 4. Bussey KJ, Lawce HJ, Olson SB, Arthur DC, Kalousek DK, Krailo M, Giller R, Heifetz S, Womer R, Magenis RE (1999): Chromosome abnormalities of eighty-one pediatric germ cell tumors: sex- , age-, site-, and histopathology-related differences--a Children's Cancer Group study. Genes Chromosomes Cancer 25:134-146. 5. Hu J, Schuster AE, Fritsch MK, Schneider DT, Lauer S, Perlman EJ (2001): Deletion mapping of 6q21-26 and frequency of 1p36 deletion in childhood endodermal sinus tumors by microsatellite analysis. Oncogene 20:8042-8044. 6. Mostert M, Rosenberg C, Stoop H, Schuyer M, Timmer A, Oosterhuis J, Looijenga L (2000): Comparative genomic and in situ hybridization of germ cell tumors of the infantile testis. Lab Invest 80:1055-1064. 7. Perlman EJ, Hu J, Ho D, Cushing B, Lauer S, Castleberry RP (2000): Genetic analysis of childhood endodermal sinus tumors by comparative genomic hybridization. J.Pediatr.Hematol.Oncol. 22:100-105. 8. Schneider DT, Schuster AE, Fritsch MK, Calaminus G, Gobel U, Harms D, Lauer S, Olson T, Perlman EJ (2002): Genetic analysis of mediastinal nonseminomatous germ cell tumors in children and adolescents. Genes Chromosomes Cancer 34:115-125.

73 Chapter 4

9. Geurts van Kessel A, Suijkerbuijk RF, Sinke RJ, Looijenga L, Oosterhuis JW, de Jong B (1993): Molecular cytogenetics of human germ cell tumours: i(12p) and related chromosomal anomalies. Eur.Urol. 23:23-28. 10. Sandberg AA, Meloni AM, Suijkerbuijk RF (1996): Reviews of chromosome studies in urological tumors. III. Cytogenetics and genes in testicular tumors. J.Urol. 155:1531-1556. 11. Zafarana G, Grygalewicz B, Gillis AJ, Vissers LE, van d, V, Van Gurp RJ, Stoop H, Debiec- Rychter M, Oosterhuis JW, van Kessel AG, Schoenmakers EF, Looijenga LH, Veltman JA (2003): 12p-amplicon structure analysis in testicular germ cell tumors of adolescents and adults by array CGH. Oncogene 22:7695-7701. 12. Veltman IM, Schepens MT, Looijenga LHJ, Strong LC, Geurts van Kessel A (2003): Germ cell tumours in neonates and infants: a distinct subgroup? APMIS 111:152-160. 13. Samaniego F, Rodriguez E, Houldsworth J, Murty VV, Ladanyi M, Lele KP, Chen QG, Dmitrovsky E, Geller NL, Reuter V (1990): Cytogenetic and molecular analysis of human male germ cell tumors: chromosome 12 abnormalities and gene amplification. Genes Chromosomes Cancer 1:289-300. 14. Rodriguez E, Mathew S, Reuter V, Ilson DH, Bosl GJ, Chaganti RS (1992): Cytogenetic analysis of 124 prospectively ascertained male germ cell tumors. Cancer Res. 52:2285-2291. 15. Murty VV, Houldsworth J, Baldwin S, Reuter V, Hunziker W, Besmer P, Bosl G, Chaganti RS (1992): Allelic deletions in the long arm of chromosome 12 identify sites of candidate tumor suppressor genes in male germ cell tumors. Proc.Natl.Acad.Sci.U.S.A. 89:11006-11010. 16. Murty VV, Bosl GJ, Houldsworth J, Meyers M, Mukherjee AB, Reuter V, Chaganti RS (1994): Allelic loss and somatic differentiation in human male germ cell tumors. Oncogene 9:2245- 2251. 17. Mitelman F, Johansson B and Mertens F Eds. Mitelman Database of Chromosome Aberrations in Cancer. 2002. 18. de Bruijn DR, Kater-Baats E, Eleveld M, Merkx G, Geurts van Kessel A (2001): Mapping and characterization of the mouse and human SS18 genes, two human SS18-like genes and a mouse Ss18 pseudogene. Cytogenet.Cell Genet. 92:310-319. 19. Veltman I, van Asseldonk M, Schepens M, Stoop H, Looijenga L, Wouters C, Govaerts L, Suijkerbuijk R, Geurts van Kessel A (2002): A novel case of infantile sacral teratoma and a constitutional t(12;15)(q13;q25) pat. Cancer Genet.Cytogenet. 136:17-22. 20. Weterman MA, Wilbrink M, Geurts van Kessel A (1996): Fusion of the transcription factor TFE3 gene to a novel gene, PRCC, in t(X;1)(p11;q21)-positive papillary renal cell carcinomas. Proc.Natl.Acad.Sci.U.S.A. 93:15294-15298. 21. Kuiper RP, Schepens M, Thijssen J, van Asseldonk M, van den BE, Bridge J, Schuuring E, Schoenmakers EF, Geurts van Kessel A (2003): Upregulation of the transcription factor TFEB in t(6;11)(p21;q13)-positive renal cell carcinomas due to promoter substitution. Hum.Mol.Genet. 12:1661-1669. 22. Asano K, Merrick WC, Hershey JW (1997): The translation initiation factor eIF3-p48 subunit is encoded by int-6, a site of frequent integration by the mouse mammary tumor virus genome. J.Biol.Chem. 272:23477-23480. 23. Marchetti A, Buttitta F, Pellegrini S, Bertacca G, Callahan R (2001): Reduced expression of INT-6/eIF3-p48 in human tumors. Int.J.Oncol. 18:175-179. 24. Yen HC, Chang EC (2003): INT6--a link between the proteasome and tumorigenesis. Cell Cycle 2:81-83. 25. Jurka J, Klonowski P, Dagman V, Pelton P (1996): CENSOR--a program for identification and elimination of repetitive elements from DNA sequences. Comput.Chem. 20:119-121. 26. Calin GA, Liu CG, Sevignani C, Ferracin M, Felli N, Dumitru CD, Shimizu M, Cimmino A, Zupo S, Dono M, Dell'Aquila ML, Alder H, Rassenti L, Kipps TJ, Bullrich F, Negrini M, Croce CM (2004): MicroRNA profiling reveals distinct signatures in B cell chronic lymphocytic leukemias. Proc.Natl.Acad.Sci.U.S.A. 101:11755-11760. 27. Velagaleti GV, Bien-Willner GA, Northup JK, Lockhart LH, Hawkins JC, Jalal SM, Withers M, Lupski JR, Stankiewicz P (2005): Position effects due to chromosome breakpoints that map

74 Positional cloning of a constitutional t(8;12)(q23;q12) in an infantile sacral teratoma

approximately 900 Kb upstream and approximately 1.3 Mb downstream of SOX9 in two patients with campomelic dysplasia. Am.J.Hum.Genet. 76:652-662. 28. Veltman IM, Vreede LA, Cheng J, Looijenga LH, Janssen B, Schoenmakers EF, Yeh ET, Geurts van Kessel A (2005): Fusion of the SUMO/Sentrin-specific protease 1 gene SENP1 and the embryonic polarity-related mesoderm development gene MESDC2 in a patient with an infantile teratoma and a constitutional t(12;15)(q13;q25). Hum.Mol.Genet. 14:1955-63.

Chromosomal breakpoint mapping by

array CGH using flow-sorted

chromosomes

Imke M. Veltman

Joris A. Veltman Ger Arkesteijn

Irene M. Janssen

Lisenka E. Vissers Pieter J. de Jong

Biotechniques Ad Geurts van Kessel

(2003) 35:1066-1070 Eric F.P.M. Schoenmakers

Chapter 5

ABSTRACT

Despite the recent completion of the human genome project, the mapping of disease- related chromosomal translocation breakpoints and genes has remained laborious. Here, we describe a novel and rapid procedure to map such translocation breakpoints using flow-sorted chromosomes in combination with array-based comparative genomic hybridization. To test the feasibility of this approach, we used a t(12;15)(q13;q25)-positive cell line with known breakpoint positions as a model. The derivative 12 chromosomes were flow sorted, labeled and hybridized to a genome- wide array containing 3648 well-characterized human genomic clones. The exact locations of the breakpoints on both chromosome 12 and 15 could be determined in a single hybridization experiment. In addition, we have tested the minimal amount of material necessary to perform these experiments and show that it is possible to obtain highly reliable profiles using as little as 10,000 flow-sorted chromosomes.

78 Chromosomal breakpoint mapping by array CGH using flow-sorted chromosomes

INTRODUCTION

There is ample evidence that recurrent chromosomal abnormalities, including translocations, are causally related to abnormal development and tumorigenesis. Indeed, the molecular characterization of translocation breakpoints has instrumented in the identification of disease causing genes and the subsequent elucidation of the underlying molecular defects. In cancer, for example, more than 100 recurring chromosomal translocations are known to trigger the activation of various oncogenes (http://www.ehgonline.net/translocation.pdf). Despite the fact that novel technologies have recently expedited the laborious process of positional cloning, such an approach still requires a considerable investment in time and reagents. Current methods may include chromosomal flow sorting, (reverse) painting, spectral karyotyping and/or fluorescent in situ hybridization (FISH) [1-4]. These technologies, however, lack the resolution required for a rapid identification of breakpoint spanning genomic sequences and/or genes. In order to overcome these limitations, and taking advantage of recent technological developments, we developed an alternative method. Here we show that rapid and reliable high-resolution breakpoint mapping is feasible, using a combination of chromosome flow sorting and array-based comparative genomic hybridization, employing a genome-wide array containing 3648 genomic clones. As a proof of principle, we used a t(12;15)(q13;q25)-positive cell line with known breakpoint positions [5;6].

MATERIAL AND METHODS

Cell culture and preparation A t(12;15)(q13;q25)-containing cell line was established by Epstein-Barr-virus- transformation of peripheral blood lymphocytes, and cultured in RPMI 1640 medium supplemented with 10 % fetal calf serum, 2mM L-glutamine and antibiotics at 37°C in a humidified atmosphere containing 5% CO2. Cells in log phase were treated for 16 hours with colcemid (0.05 mg/ml final concentration) in order to arrest cells in metaphase. Metaphase chromosomes were subsequently isolated for flow sorting as described previously [1].

79 Chapter 5

Flow karyotyping and sorting Chromosomes were stained with Chromomycin-A3 (CA3) and Hoechst 33258 (Ho), and analyzed on a dual-laser beam flow cytometer (FACS-Vantage SE, Becton Dickinson, San Jose, Ca): CA3 was excited with an argon ion laser tuned at 458 nm at 100 mW. CA3 fluorescence was measured through a 550-nm longpass filter. Ho was excited with an argon ion laser tuned into the UV range (351 and 364 nm) at 125-mW laser power. Ho fluorescence was measured through two KV 408 filters. The system was triggered on the CA3 fluorescence signal. In total, 300,000 chromosomes were sorted from each cluster (Figure 1) and subsequently purified using the QIAamp purification kit (QIAGEN, Hilden, Germany). DNA obtained in this way was used for array-based CGH analyses. Due to the fact that these small amounts of DNA cannot be measured accurately, we have based the number of chromosome copies on the number that is indicated by the FACS.

Array-based comparative genomic hybridization (array CGH) Array CGH was performed essentially as described before [7;8] with minor modifications, including the production of a genome-wide BAC array, the use of an automated hybridization station and advanced normalization procedures [9]. In brief, the array used in this study contained 3648 colony-purified, FISH-verified, phage-tested BAC clones spotted in triplicate, most of which have been fully integrated in the human genome browser (http://genome.ucsc.edu/). Approximately 3000 of these clones cover the genome at a 1 Mb resolution. Additional clones included those mapping to known microdeletion hot-spots and to subtelomeric regions, as well as clones covering certain chromosome regions with a higher resolution than 1 Mb, including clones which were generated in the context of a previous positional cloning project performed on the cell line discussed in this paper [5]. The array CGH profile of the flow sorted chromosomes was established through hybridization of different amounts of cy3-dUTP labeled flow-sorted chromosomes combined with 500 ng cy5-dUTP labeled control genomic DNA. After scanning, fluorescence test over reference ratios were determined for each clone and, subsequently log2 transformed. These log2 ratios were centered to 0 by dividing the ratio for each clone by the mean of all ratios except for those mapping to the chromosomes 12 and 15. The latter approach was taken because there was hardly any signal in the cy3-channel for the clones not present on the derivative chromosome 12.

80 Chromosomal breakpoint mapping by array CGH using flow-sorted chromosomes

RESULTS AND DISCUSSION

A t(12;15)(q13;q25)-positive cell line with previously characterized breakpoint positions [5;6] was used for FACS mediated flow sorting of the derivative 12 chromosome [der(12)] and the normal chromosome 12 cluster (Figure 1). Due to the fact that the normal chromosome 12 overlaps with chromosomes 9, 10 and 11 in the flow karyogram, these chromosomes were sorted together.

der12 9-12 Hoechst 33258 fluorescence

Chromomycin-A3 fluorescence

Figure 1. Flow-karyogram of the t(12;15)(q13;q25)-positive cell line. The circles mark the selected areas that were flow-sorted: the der(12) chromosome and the chromosome 12 cluster, including chromosomes 9,10 and 11.

In an initial test we used 150,000 copies of the chromosome 9-12 cluster (test) labeled with cy3-dUTP in combination with 500 ng normal male genomic DNA (reference) labeled with cy5-dUTP and hybridized the mixture to the genome-wide BAC array. This pilot showed a marked hybridization of the clones mapping to the chromosomes 9-12 in the cy3 channel, whereas virtually no signals were measured for the clones mapping to the other chromosomes. Due to the fact that the normal cy5-labeled genomic reference DNA hybridized equally to all clones, the resulting log2-transformed test over reference values (log2 T/R values) for the clones mapping

81 Chapter 5 to the chromosomes 9-12 varied between 2 and 3, whereas the clones mapping outside these chromosomes were centered around 0 (results not shown). This initial test showed that this procedure allows a quick identification of the precise chromosomal regions selected by flow sorting onto the genome wide BAC array. Furthermore, this test also indicates that less flow-sorted chromosomes might be used in these array experiments. Accordingly, hybridizations were performed with 100,000 copies of der(12) labeled with cy3-dUTP (test) against 500 ng of normal male genomic DNA labeled with cy5-dUTP (reference) (Figure 2). The results showed a clear difference in T/R values between the clones that map to the der(12) and the clones that do not. The clones from 12pter through clone RP1-228p16 on 12q13.1 (48.4 Mb) revealed a mean log2 T/R value of 3.8, whereas the clones from RP11-480h15 on 12q13.1 (49.0 Mb) until 12qter displayed a log2 T/R value in the normal range (mean 0) (Figure 2B). This result indicates that the breakpoint is positioned in the region encompassing RP1-228p16 and RP11-480h15 and correlates perfectly with our recent positional cloning efforts, which revealed that clone RP1-228p16 contains the breakpoint-spanning gene [5]. For chromosome 15 a similar pattern was observed, with a normal log2 T/R value for all clones from 15q12 through clone RP11-60j21 on 15q25 (76.8 Mb), and a mean log2 T/R value of 4.1 for all clones from RP11-775c24 on 15q25 (77.2 Mb) until 15qter (Figure 2C). Again this result perfectly matches our recent data indicating that RP11-775c24 contains the breakpoint-spanning gene on chromosome 15 [5]. In order to define the minimal amount of chromosomes that is necessary to acquire a reliable array CGH profile, 50,000 der(12), 10,000 der(12) and 1,000 der(12) copies were hybridized as test DNA together with 500 ng of reference DNA to our genome-wide chip. The results showed that hybridizations with 50,000 and 10,000 derivative chromosomes gave profiles similar to Figure 2, whereas for 1,000 copies a clear-cut distinction between der(12) clones and non-der(12) clones was no longer possible. Therefore, the minimal amount of chromosome copies for a reliable array CGH profile lies in between 1,000 and 10,000 copies. The amount of 10.000 copies equals the chromosome copy number of autosomal chromosomes that are present in approximately 30 ng of genomic DNA from normal diploid human cells, assuming that a normal diploid human cell contains about 6 pg of DNA. Interestingly, in a typical array CGH experiment 500 ng of genomic DNA is routinely used. As an alternative for the use of even smaller numbers of flow-sorted chromosomes an amplification step may be included, as e.g. degenerate oligonucleotide primed (DOP) PCR (10,11). In a preliminary study we found that this approach is feasible using

82 Chromosomal breakpoint mapping by array CGH using flow-sorted chromosomes

1,000 copies as input in the PCR, although this needs further optimization (data not shown). Such amplifications may be required for other applications of this method, as for example delineation of the genomic content of unknown marker chromosomes in primary clinical samples. This study provides a further example of the power of the array CGH technology, which is rapidly becoming the method of choice for high-resolution screening of genomic copy number changes (8;12;13;14). Arrays with a genome-wide coverage at the 1-3 Mb resolution are becoming more readily available for these kind of applications (9;10;15). A novel method for rapid mapping of chromosomal translocation breakpoints based on a combination of chromosome flow sorting and array-based CGH was tested. As a proof of principle, derivative 12 chromosomes from a known t(12;15)(q13;q25)-positive cell line were flow-sorted and hybridized to a genome- wide CHIP containing 3648 well characterized genomic clones. In a single hybridization experiment, using as little as 10,000 copies of the der(12) chromosome, the breakpoints at both chromosome 12 and 15 were fine-mapped, fully in line with our recent positional cloning studies (5). For chromosome 12 the breakpoint was positioned in the interval: RP1-228p16 to RP11-480h15 (~600 kb interval) and for chromosome 15 in the interval: RP11-60j21 to RP11-775-c24 (~400 kb interval). This method will greatly facilitate the tedious procedure of positional cloning from balanced chromosomal anomalies. Although we show the applicability of this approach using a cytogenetically characterized translocation this approach can be extended to all undefined complex chromosomal anomalies. The only prerequisite for using this approach is the necessity to be able to flow-sort aberrant chromosomes separately from their (supposedly) reciprocal translocation products. Currently, we are in the process of preparing a full coverage (32K) genome-wide CHIP (http://bacpac.chori.org/pHumanMinSet.htm). Once available, we will be able to map translocation breakpoints to a single BAC clone in a single experiment, leading to direct gene identification.

83 Chapter 5

6.0 12 15 5.0 4.0 3.0 2.0

log2 T/R 1.0 0.0 -1.0 -2.0 12 15 A Clones ordered per chromosome by Mb position 6.0 5.0 4.0 3.0 2.0

log2 T/R 1.0 0.0 -1.0 0 20 40 60 80 100 120 140 -2.0 B Clones ordered for chromosome 12 by Mb position 6.0 5.0 4.0 3.0 2.0

log2 T/R 1.0 0.0 -1.0 0 20 40 60 80 100 -2.0 C Clones ordered for chromosome 15 by Mb position

Figure 2. Mapping of the t(12;15)(q13;q25) by array-based CGH analysis. Profiles are shown after hybridization of 100,000 copies of der(12) (test), cy3-dUTP labeled, versus 500 ng of normal male genomic DNA (reference), cy5-dUTP labeled. Dark squares represent mean normalized T/R values for each clone in a log2 scale and the vertical lines represent the standard deviations of the replicates per clone. A) The T/R values of the genomic clones, ordered from chromosome 1 to Y, and within each chromosome by Mb position as presented in the June 2002 human genome assembly freeze. The vertical lines separate the clones per chromosome. A clear distinction can be made between the clones that hybridize to der(12) (with high T/R values) and the clones that do not (low T/R values). B) A zoom-in showing the T/R values of the chromosome 12 clones, ordered by Mb position. Indicated are the clones surrounding the breakpoint. C) T/R values of the chromosome 15 clones, ordered by Mb position. Absence of clones from 0 to approximately 20 Mb is due to the difficulty of mapping, cloning and sequencing of this region. Indicated are the clones surrounding the breakpoint.

84 Chromosomal breakpoint mapping by array CGH using flow-sorted chromosomes

ACKNOWLEDGMENTS

We thank Kazutoyo Osoegawa and Chik On Choy (BACPAC Resources) for co- establishing the 3.7K BAC set and scientific interactions, Walter van der Vliet for array preparations and Huub Straatman for statistical analyses. Furthermore, we thank Lutgarde Govaerts (Erasmus University Rotterdam) for supplying us with the t(12;15)(q13;q25)-positive cell line.

REFERENCES

1. Arkesteijn, G., E. Jumelet, A. Hagenbeek, E. Smit, R. Slater and A. Martens. (1999):. Reverse chromosome painting for the identification of marker chromosomes and complex translocations in leukemia. Cytometry 35:117-124. 2. Griffin, D.K., D. Sanoudou, E. Adamski, C. McGiffert, P. O'Brien, J. Wienberg and M.A. Ferguson-Smith. (1998): Chromosome specific comparative genome hybridisation for determining the origin of intrachromosomal duplications. J Med Genet 35:37-41. 3. Heng, H., C. Ye, F. Yang, S. Ebrahim, G. Liu, S. Bremer, C. Thomas, J. Ye, T. Chen, C. Tuck- Muller, J. Yu, S. Krawetz and A. Johnson. (2003): Analysis of marker or complex chromosomal rearrangements present in pre- and post-natal karyotypes utilizing a combination of G- banding, spectral karyotyping and fluorescence in situ hybridization. Clin Genet 63:358-367. 4. Suijkerbuijk, R.F., D. Matthopoulos, L. Kearney, S. Monard, S. Dhut, F.E. Cotter, J. Herbergs, A. Geurts van Kessel and B.D. Young. (1992): Fluorescent in situ identification of human marker chromosomes using flow sorting and Alu element-mediated PCR. Genomics 13:355- 36210. 5. Veltman I.M., B.A. Janssen, L.A. Vreede, L.H.J. Looijenga, E.F.P.M. Schoenmakers and A. Geurts van Kessel. (2005): Fusion of SENP1 and MESD in a patient with a neonatal teratoma and a constitutional t(12;15)(q13;q25). Human Molecular Genetics 14: 1955-1963. 6. Veltman, I., M. van Asseldonk, M. Schepens, H. Stoop, L. Looijenga, C. Wouters, L. Govaerts, R. Suijkerbuijk and A. Geurts van Kessel. (2002): A novel case of infantile sacral teratoma and a constitutional t(12;15)(q13;q25) pat. Cancer Genet Cytogenet 136:17-22. 7. Veltman, J.A., Y. Jonkers, I. Nuijten, I. Janssen, D. Van, V, E. Huys, J. Vermeesch, G. Van Buggenhout, J.P. Fryns, R. Admiraal, P. Terhal, D. Lacombe, A. Geurts van Kessel, D. Smeets, E.F. Schoenmakers and C.M. Ravenswaaij-Arts. (2003): Definition of a Critical Region on Chromosome 18 for Congenital Aural Atresia by ArrayCGH. Am J Hum Genet 72:1578-1584. 8. Veltman, J.A., E.F. Schoenmakers, B.H. Eussen, I. Janssen, G. Merkx, B. van Cleef, C.M. van Ravenswaaij, H.G. Brunner, D. Smeets and A. Geurts van Kessel. (2002): High-throughput analysis of subtelomeric chromosome rearrangements by use of array-based comparative genomic hybridization. Am J Hum Genet 70:1269-1276. 9. Vissers LE, de Vries BB, Osoegawa K, Janssen IM, Feuth T, Choy CO, Straatman H, Van der Vliet W, Huys EH, van Rijk A, Smeets D, Ravenswaaij-Arts CM, Knoers NV, van der Burgt I, de Jong PJ, Brunner HG, Geurts van Kessel A, Schoenmakers EF, Veltman JA. (2003): Array- based comparative genomic hybridization for the genomewide detection of submicroscopic chromosomal abnormalities. Am J Hum Genet 73: 1261-1270. 10. Fiegler, H., P. Carr, E.J. Douglas, D.C. Burford, S. Hunt, J. Smith, D.Vetrie, P. Gorman, I.P. Tomlinson and N.P. Carter. (2003). DNA microarrays for comparative genomic hybridization

85 Chapter 5

based on DOP-PCR amplification of BAC and PAC clones. Genes Chromosomes Cancer 36:361-374. 11. Telenius, H., A.H. Pelmear, A. Tunnacliffe, N.P. Carter, A. Behmel, M.A. Ferguson-Smith, M. Nordenskjold, R. Pfragner and B.A. Ponder. (1992): Cytogenetic analysis by chromosome painting using DOP-PCR amplified flow-sorted chromosomes. Genes Chromosomes Cancer 4:257-263. 12. Pinkel, D., R. Segraves, D. Sudar, S. Clark, I. Poole, D. Kowbel, C. Collins, W.L. Kuo, C. Chen, Y. Zhai, S.H. Dairkee, B.M. Ljung, J.W. Gray and D.G. Albertson. (1998): High resolution analysis of DNA copy number variation using comparative genomic hybridization to microarrays. Nat Genet 20:207-211. 13. Pollack, J.R. and V.R. Iyer. (2002): Characterizing the physical genome. Nat Genet 32 Suppl:515-521. 14. Solinas-Toldo, S., S. Lampel, S. Stilgenbauer, J. Nickolenko, A. Benner, H. Dohner, T. Cremer and P. Lichter. (1997): Matrix-based comparative genomic hybridization: biochips to screen for genomic imbalances. Genes Chromosomes Cancer 20:399-407. 15. Snijders, A.M., N. Nowak, R. Segraves, S. Blackwood, N. Brown, J. Conroy, G. Hamilton, A.K. Hindle, B. Huey, K. Kimura, S. Law, K. Myambo, J. Palmer, B. Ylstra, J.P. Yue, J.W. Gray, A.N. Jain, D. Pinkel and D.G. Albertson. (2001): Assembly of microarrays for genome-wide measurement of DNA copy number. Nat Genet 29:263-264.

Identification of Recurrent

Chromosomal Aberrations in Germ

Cell Tumors of Neonates and Infants

Using Genomewide Array-Based

Comparative Genomic Hybridization

Imke Veltman

Joris Veltman

Irene Janssen Christina Hulsbergen-van de Kaa

Wolter Oosterhuis

Dominik Schneider Hans Stoop Ad Gillis

Susanne Zahn

Leendert Looijenga Ulrich Göbel Genes, Chromosomes & Cancer (2005) 43:367–376 Ad Geurts van Kessel

Chapter 6

ABSTRACT

Human germ cell tumors (GCTs) of neonates and infants constitute a heterogeneous group of neoplasms, including teratomas and yolk sac tumors with distinct clinical and epidemiologic features. As yet, little is known about the cytogenetic constitution of these tumors. We applied the recently developed genomewide array CGH technology to 24 germ cell tumors derived from patients under the age of 5 years. In addition, we included seven tumors derived from children and adolescents older than 5 years. In the series under the age of 5 years most teratomas displayed normal profiles, except for some minor recurrent aberrations. In contrast, the yolk sac tumors displayed recurrent losses of 1p35-pter, gains of 3p21-pter and gains of 20q13. In the GCTs of patients older than 5 years the main recurrent anomalies included gains of 12p and gains of whole chromosomes 7 and 8. In addition, gains of 1q32-qter, losses of 6q24-qter and losses of 18q21-qter regions were frequently observed in GCTs with different histologies, independent of age. We conclude that array CGH serves as a highly suitable method for the identification of recurrent chromosomal anomalies in GCTs of neonates and infants. The recurrent anomalies point to chromosomal regions that may harbor novel diagnostic/prognostic identifiers and genes relevant for the development of these neoplasms.

90 Array CGH profiles of neonatal/infantile GCTs

INTRODUCTION

Human germ cell tumors (GCTs) comprise a heterogeneous group of neoplasms exhibiting a high degree of epidemiologic, clinical, histologic and genetic diversity. They present at all ages, but are most frequently encountered as testicular GCTs in adolescent boys and young adult men. Previous genetic and epidemiologic studies have indicated that GCTs can be divided into four groups (1) teratomas and yolk sac tumors (YSTs) in neonates and infants, (2) GCTs of the adult testicular type, i.e., seminomas and non-seminomas, (3) dermoid cysts (mature cystic teratomas) of the ovary, and (4) spermatocytic seminomas in elderly men [5;18;22;30;45]. Several studies were aimed at identifying the precursor cells of these neoplasms and, although it is thought that most GCTs originate from germ cells, [29] suggested that extragonadal GCTs may originate from undifferentiated pluripotent stem cells. Other studies, however, have pointed at a cellular origin from transformed germ cells at different stages of development [3;28;33;35;38]. GCTs in adolescents and young adults primarily develop in the gonads, while the sacrococcygeal region is the main location for children under the age of 5 years [30]. Testicular GCTs in adult men with a teratoma component show almost always a malignant clinical course with metastatic spread. In contrast, (pure) mature teratomas in neonates and infants show clinical and histopathologic features of benign tumors, although they may recur after incomplete resection. Two extensive epidemiologic studies among children with GCTs revealed that age is a critical factor for histology and prognosis. The median age of children who present with a teratoma is 5 months. Above this age, YSTs become more prevalent and risk-adapted chemotherapy in addition to surgical resection may become necessary [5;30]. A suitable clinical diagnostic marker for the presence of potentially malignant YST components is an elevated level of serum α-fetoprotein (AFP). However, clinical detection of YST may be hampered by physiologically elevated AFP levels during infancy [2]. Histologically, some neonatal teratomas contain YST foci that may be detected using immunohistochemistry [8;9;15]. Testicular GCTs (group 2) have been studied most extensively at both the cytogenetic and the molecular levels. These studies revealed that, next to gross numerical and structural abnormalities, chromosome 12p sequences are overrepresented in almost all of these tumors, mostly due to the presence of one or more copies of an i(12p) chromosome. To date, i(12p) chromosomes and gain of the complete 12p arm have not been detected in GCTs of neonates and infants. Other

91 Chapter 6 recurrent genomic aberrations in testicular GCTs of adults include overrepresentations of (parts of) chromosomes 7, 8, 12 and X, and under- representations of (parts of) chromosomes 4, 5, 11, 13, 18 and Y [reviewed by 20]. As yet, little is known about the cytogenetic constitution of GCTs of neonates and infants. The sparsity of available information [reviewed by 40] indicates that in this patient group teratomas may be diploid with few, if any, cytogenetic aberrations [4;21;31]. In the malignant counterparts of these teratomas, i.e., YSTs, recurrent chromosomal aberrations have been observed, of which losses of the distal parts of 1p and 6q and gains of 20q appear to be the most frequent ones [4;13;21;23- 25;31;32;36]. Based on their site of occurrence, histology, presumed origin, and cytogenetics, we and others have suggested that GCTs of neonates and infants constitute a distinct subgroup [22;24;25;30;33;36;40]. In this study we aimed at the identification of genomic regions that are recurrently affected in GCT subtypes of neonates and infants in order to (i) refine the diagnosis and prognosis of these subtypes and (ii) generate tools for the identification of disease-associated genes. We used the recently developed microarray-based comparative genomic hybridization (array CGH) technology that, due to its high resolution, has rapidly become the method of choice for the molecular cytogenetic analysis of inherited and acquired disorders, including GCTs [10;41;43;44;46].

MATERIALS AND METHODS

Tumor samples In total 31 germ cell tumors (GCTs) of children under the age of 16 years were selected. These GCTs originated from different anatomic sites and presented with different histologies, mainly teratomas and yolk sac tumors (Table 1). The tumor samples were collected at the Department of Pathology, Erasmus MC (Rotterdam, The Netherlands) and the Department of Pathology, Radboud University Nijmegen Medical Centre (Nijmegen, The Netherlands). Additional tumor samples were derived from patients enrolled into a prospective trial for non-testicular GCTs in children (MAKEI 96, Düsseldorf, Germany). Fresh frozen tumor samples were characterized (histologically and morphologically) according to the MAKEI study guidelines [6] and F. Gonzalez-Crussi [7]. The samples were considered suitable for study when the percentage of tumor cells was > 80%, except for six cases which were

92 Array CGH profiles of neonatal/infantile GCTs either micro- or macro-dissected. For micro-dissection laser-capture systems (Leica AS LMD or Arcturus Pixcell II) were used to isolate the tumor components from cases 5, 8, 10, 11 and 25 [1;39]. In one case (case 3) macro-dissection was used to separate stroma from tumor components. Genomic DNA was isolated using the QIAamp purification kit (QIAGEN, Hilden, Germany) and applied for array CGH.

Array-based comparative genomic hybridization Array CGH was performed using a BAC microarray containing a total of 3569 colony-purified FISH-verified clones, including approximately 3200 clones obtained through a collaboration with the Children’s Hospital Oakland Research Institute, BACPAC Resources (Oakland, CA) and other groups, to cover the genome with a 800 kb resolution. This clone set represents a subset of a larger set that is currently being distributed by BACPAC Resources (http://bacpac.chori.org/). Our array has been extensively validated for the detection of copy number changes using normal versus normal control experiments as well as experiments in which cases with known micro-deletion syndromes were tested [42]. Details on BAC array preparation, probe labeling, hybridization, scanning and data analysis were also described by Vissers et al. [42]. In brief, equal amounts of patient and reference genomic DNAs were labeled by random priming with Cy3-dUTP or Cy5-dUTP. Labeled test and reference DNAs were mixed with Cot-1 DNA, co-precipitated and re-suspended in a hybridization solution. After denaturation of probe and target DNAs, hybridization and post hybridization washing procedures were performed using a GeneTac Hybridization Station (Genomic Solutions, Camebrigeshire, UK), according to the instructions of the manufacturer. Fluorescence intensity images were acquired using an Affymetrix 428 scanner, and analyzed by Genepix Pro 5.1 (Axon Instruments Inc., Foster City, CA). For each spot the median pixel intensity minus the median local background for both dyes were used to obtain a genomic copy number ratio. Data normalization was performed by using the software package SAS version 8.0 (SAS Institute, Cary, NC) per array subgrid by applying lowess curve fitting to predict the log2 transformed test over reference (T/R) values. All mapping information regarding clone locations, cytogenetic bands, and genomic contents were retrieved from the UCSC genome browser (April 2003/May 2004 freezes).

93 Chapter 6

Table 1. Characterization of Germ Cell Tumors

Sub- a b b Case group Age Sex Site Histology Copy number losses Copy number gains 1 1 <1 m sacrococcygeal mat. teratoma 21<1 f sacrococcygeal mat. teratoma d 3 1 <1 f sacrococcygeal mat. teratoma (14q11) 41<1 f sacrococcygeal mat. teratoma 5 1 <1 m testis mat.+ im. teratoma 61<1 f sacrococcygeal mat. teratoma 7 1 <1 m retroperitoneal mat. teratoma 8 1 <1 m sacrococcygeal mat.+ im. teratoma 91<1 f sacrococcygeal mat. teratoma 12p 10 1 <1 m testis mat. teratoma 11 1 <1 f neck mat.+ im. teratoma 12 1 <1 f sacrococcygeal mat.+ im. teratoma 20p 20q 13 1 2 m sacrococcygeal mat. teratoma 14 1 1 m sacrococcygeal mat. teratoma 15 1 <1 m kidney im. teratoma 16 1 <1 m sacrococcygeal im. teratoma 18, 20p 19 17 2 <1 m sacrococcygeal yolk sac tumor 1p22-p31, 12p, 14q32-qter, 16p13-pter, 19, 20q 15q25, 18q21-qter 18 2 1 f sacrococcygeal yolk sac tumor 19 2 1 f uterus/vagina yolk sac tumor 1p35-pter, 6q22-qter, 18q 20q13-qter 20 2 1 m testis yolk sac tumor 5, 6, 8 1q, 12p, 19q13-qter, d d 21 2 1 m testis yolk sac tumor 1p35-pter, 1p22-p31, (4p13) , 3p21-pter, (5q35) , 20q 7q11, 16p, 19, 20p 22 2 2 f sacrococcygeal yolk sac tumor 6q22-qter, 13pter-q21, 15q25 1q, 3p21-pter, 20q d 23 2 3 m mediastinum yolk sac tumor 1p35-pter, 4q21-q28, 5q12- 3p21-pter, (7q11) , 20q q14, 8p,18p11-pter, 20p 24 2 3 f sacrococcygeal yolk sac tumor 1p35-pter, 6q24-qter, 18q21- 1q, 20q qter, 19, 21q22-qter 25 3 6 f sacrococcygeal im. teratoma 6q16-qter 1q24-qter, 12 c 26 3 7 f ovary mixed tumor 18q 7,8, 12p, 19 27 3 8 f ovary mat. teratoma 28 3 11 f ovary mat. teratoma 12p 29 3 11 f ovary dysgerminoma 18q 1q, 7,8, 10, 12p13- pter, 12p11, 19, 21 30 3 12 f ovary im. teratoma 7,8, 12p, 21 31 3 16 f pelvis yolk sac tumor 1p, 4, 6q, 7q32-qter, 8p, 13 1q, 6p, 17q, 20q aIn years. bCopy number changes of genome regions in several BAC clones, depicted by chromosome bands, chromosomes arms, or whole chromosomes. cEmbryonal carcinoma, yolk sac tumor, teratoma. dEncompassing three BAC clones.

RESULTS

Patient classification and array CGH analysis Genomic DNA extracted from 24 GCTs of neonates and infants (0-5 years old) were analyzed with the array CGH technique for the presence of copy number alterations. Based on their histology, these tumors were subdivided into two subgroups (Table 1): subgroup 1 consisting of 16 teratomas (mature as well as immature types) and subgroup 2 consisting of eight yolk sac tumors (YSTs). Additionally, we included a

94 Array CGH profiles of neonatal/infantile GCTs third cohort of seven GCTs derived from patients above the age of 5 (range: 6-16 years) with different histologies, i.e., four teratomas, one YST, one dysgerminoma, and one mixed GCT with embryonal carcinoma, YST and teratoma components (subgroup 3). Most tumors from subgroups 1 and 2 originated from the sacrococcygeal region, whereas most tumors from the third subgroup originated from the ovary (Table 1). The genomewide microarray that we used contained 3569 BAC clones (spotted in triplicate) covering the entire human genome. Since sex-matches are required for the retrieval of reliable X/Y data, we only analyzed the clones mapping to the autosomes. Differentially labeled patient and normal control DNAs were hybridized to the arrays, whereas two sex-matched and four sex-mismatched normal versus normal hybridizations were included as controls. Figure 1 shows representative array CGH profiles of tumor samples selected from each subgroup. Note the differences between these three profiles: an almost normal profile in a teratoma of the first subgroup (Fig. 1A), multiple copy number changes including loss of 1p, loss of 20p and gain of 20q in a YST of the second subgroup (Fig. 1B), and multiple copy number changes including gain of 12p in a mixed GCT of the third subgroup (Fig. 1C).

Table 2. Recurrent Copy Number Changes of Genomic Regions >5 Mb a b Gain/loss Chromosome band Flanking clones Size (Mb) Subgroup 1 Subgroup 2 Subgroup 3 Gain 1q32-qter RP11-155f3-CTB-167k11 43 0 3 3 Gain 3p21-pter RP11-91p19-RP5-1186b18 47 0 3 0 Gain chr. 7 159 0 0 3 Gain chr. 8 146 0 0 3 Gain 12p13-pter RP11-174g6-CTB-124k20 15 0 0 5 Gain 12p11 RP11-38f11-RP11-78f16 2 0 0 5 Gain 19q13-qter RP11-2j15-qter 12 1 2 2 Gain 20q13 RP11-1e23-RP11-124d1 11 1 7 1 Loss 1p35-pter RP11-4p6-CTC-232b23 33 0 4 1 Loss 6q24-qter RP11-40o24-RP1-57h24 22 0 4 2 Loss 8p11-pter RP11-210f15-RP11-91j19 37 0 2 1 Loss 18q21-qter RP11-43k24-RP5-964m9 30 1 3 2 Loss 20p12 RP11-51o4-RP11-149o7 10 2 2 0 aDetermined by the minimal region of overlap. bClones that flank the minimal region of overlap.

95 Chapter 6 Case 1 2.0 1.5 1.0 0.5 0.0 -0.5 Log2 T/R Log2 -1.0 -1.5 122 -2.0 Clones ordered per chromosome by Mb position A Case 23 2.0 1.5 1.0 0.5 0.0 -0.5 Log2 T/R Log2 -1.0 -1.5 122 -2.0 Clones ordered per chromosome by Mb position B Case 26 2.0 1.5 1.0 0.5 0.0 -0.5 Log2 T/R Log2 -1.0 -1.5 122 -2.0 Clones ordered per chromosome by Mb position C

Figure 1. Genomic profiles of three germ cell tumors. Dark squares represent mean normalized T/R values for each clone in a log2 scale. The log2 T/R values of the genomic clones are ordered from chromosome 1 to 22, and within each chromosome by Mb position as presented in the UCSC genome browser (April 2003 freeze). The vertical dotted lines separate the clones per chromosome. A) An example of a subgroup 1 profile; a sacrococcygeal teratoma of a baby boy (case 1). Note the single deleted clone on 5p15 (RP11-88l18). B) An example of a subgroup 2 profile; a mediastinal yolk sac tumor of a 3 year old boy (case 23) with loss of 1pter-p35 and gain of 3pter-p21 and 20q. C) An example of a subgroup 3 profile, an ovarian mixed germ cell tumor of a 7 year old girl (case 26) with the typical gain on 12p co-occurring with a number of other imbalances; gain of chromosomes 7, 8 and 19 and loss of 18q.

96 Array CGH profiles of neonatal/infantile GCTs

Array CGH profiles in infantile teratomas Most teratomas of subgroup 1 showed normal array CGH profiles, independent of the tumor location (Table 1). In three cases (9, 12 and 16; Table 1) DNA copy number variations were detected. Interestingly, one of these profiles (case 12) again showed loss of 20p and gain of 20q (Fig. 2A). This observation suggests the presence of an i(20q) chromosome in this tumor sample. In addition, 20p is deleted in the profile of case 16, next to loss of chromosome 18 and gain of chromosome 19. The profile of case 9 showed loss of a complete 12p arm. Figure 3A displays a frequency plot of copy number changes in subgroup 1 based on the percentages of gains (log2 T/R >0.3) and losses (log2 T/R <-0.3) measured for each individual BAC clone on the array. Except for a few isolated clones, no recurrent gross (>5 Mb) copy number changes could be detected for this subgroup (Table 2). In contrast, small genomic regions (<5Mb) were frequently found to be imbalanced. In Table 3 the BAC clones are listed that showed gains or losses in at least four of the array CGH profiles (including subgroups 1, 2 and 3) and that did not coincide with the recurrent gross genomic changes presented in Table 2.

Case 12 Case 21 2.0 2.0 1.5 1.5 1.0 1.0 0.5 0.5 0.0 0.0 -0.5 -0.5 Log2 T/R Log2 T/R -1.0 -1.0 -1.5 -1.5 -2.0 -2.0 Chromosome 20 clones ordered by Mb position Chromosome 1 clones ordered by Mb position A C Case 24 Case 26 2.0 2.0 1.5 1.5 1.0 1.0 0.5 0.5 0.0 0.0 -0.5 -0.5 Log2 T/R Log2 T/R -1.0 -1.0 -1.5 -1.5 -2.0 -2.0 Chromosome 1 clones ordered by Mb position Chromosome 12 clones ordered by Mb position B D

Figure 2. Examples of the most frequent gross copy number changes. A) Chromosome 20 profile of a sacroccocygeal teratoma (case 12). B) Chromosome 1 profile of the infantile YST (case 24) showing the smallest region of overlap of the 1pter-p35. C) Chromosome 1 profile of the infantile YST (case 21) with a somatic (region 1) and a constitutional (region 2) deletion on 1p. D) Chromosome 12 profile of an ovarian mixed GCT (case 26).

97 Chapter 6

Recurrent genomic aberrations in infantile yolk sac tumors The genomic profiles of infantile YSTs (subgroup 2) revealed several frequently occurring gross (>5 Mb) chromosomal alterations (Fig. 3B). Seven cases showed gains and/or losses of multiple large chromosomal regions, and one case appeared to exhibit a normal profile (Table 1). The following recurrent gross aberrations (listed in Table 2) were detected in three or more cases: (1) loss of 1p35-1pter. This region was lost in four of eight cases, with the shortest region of deletion overlap in a sacrococcygeal YST from a 3-year-old girl (Fig. 2B; case 24, 33 Mb); (2) gain of 1q32- qter in three cases, with the shortest region of amplification overlap determined by the same case as that for the 1p35-pter region (Fig. 2B; case 24, 43 Mb); (3) gain of 3p21-pter, again observed in three of the cases, all of which showing amplified regions of similar size, the shortest region of overlap being 47 Mb; (4) loss of 6q24- qter in four cases, of which one profile showed loss of a complete chromosome 6 and with the shortest region of deletion overlap determined by case 24 (22 Mb); (5) loss of 18q21-qter in three cases, with the minimal region of deletion overlap (30 Mb) determined by cases 19 and 24; (6) gain of the complete or partial 20q arm, the most frequently observed anomaly in the infantile YSTs, with gain of 20q13 in seven of eight cases and the shortest region of overlap, 11 Mb, determined by case 19. Other, less frequent (< 3 cases), copy number changes in this subgroup were loss of 8p, loss of 15q25, gain of 19q13-qter, loss of 19 and loss of 20p (Table 1).

Loss of 1p35-pter in a yolk sac tumor and a constitutional del(1)(1p22-1p31) Subgroup 2 included a testicular YST of a 1-year-old boy (case 21) whose array CGH profile revealed two 1p regions with genomic loss (Fig. 2C, regions 1 and 2). This was confirmed by a replicate array CGH experiment and an independently performed whole-chromosome CGH analysis (not shown). One of these regions overlapped with the 1p35-pter region that was frequently found to be deleted in the samples of subgroup 2 (see above). The other region mapped to 1p22-p31. In addition, this boy suffered from mental retardation. Through conventional cytogenetic analysis of blood lymphocytes, a constitutional mosaicism for the del(1)(p22p31) was detected (not shown). The boundaries of the constitutional deletion were determined by BAC clones RP11-451c22 and RP11-114k16, encompassing a ~10-Mb region. Interestingly, loss of this region was also detected in one of the other infantile YSTs (subgroup 2; case 17) and it contains 49 genes, including those encoding MSH4 (involved in meiotic recombination), AK5 (adenylate kinase 5), FUBP1 (binds MYC promoter), PRKACB (kinase, catalytic subunit of AMPK), DLAD (DNase II beta), SPATA1

98 Array CGH profiles of neonatal/infantile GCTs

(spermatogenesis associated 1), SSX2IP (SSX2 interacting protein), NYD-SP29 (a presumed testis development protein), BCL10 (B-cell CLL/lymphoma 10) and CYR61 (promotes angiogenesis, cell adhesion, cell proliferation and chemotaxis).

Table 3. Recurrent Copy Number Changes of Genomic Regions <5 Mb e Chromosome Subgroup Subgroup Subgroup LCV a Gain/loss Clone name band 1 2 3 Genes detected Gain CTB-172i13 and 2qter 4 0 1 - B RP5-1011o17 Gain RP11-89a12 3p14.1 2 2 2 TAFA1 - Gain RP11-554b18 3p12.1 3 1 2 - - Gain RP11-80h24 3p12.1 3 0 1 - - Gain RP11-91a15 3p11 2 1 1 EPHA3 - Gain RP11-40d5 4q28 1 3 0 - - Gain RP11-467a2 5q13.2 3 0 1 TNPO1 - Gain RP11-17o4 9q34 4 1 1 AK130729 - Gain RP11-88n14 14q21 2 1 2 - B Gain RP1-149d14 17p11 2 4 2 - - LossRP11-11m9 4p15 1 1 3 - - Loss RP11-82m24 5p15 1 1 2 - - Loss RP11-29n3 5p15 1 1 2 CTNND2 - Loss RP11-135M13 5p15 0 1 3 FBXL7 - Loss RP11-88l18 5p15 9 3 4 - B b Loss CTD-2041d13 5q13 1 3 2 several genes A c Loss RP11-121g20 6p21 1 3 0 several genes - Loss RP11-4o1 9q31 1 1 2 SUSD1 - LossRP11-95j4 9q32 3 2 3 - - Loss RP11-180c18 10p15 2 1 1 ADARB2 - Loss RP11-115g22 15q13 3 0 1 CHRNA7 A Loss RP11-33b6 15q24 2 0 3 ARIH1, GOLGA6, BBS4 - d Loss RP11-368n21 16p11 1 2 2 several genes - Loss RP11-348E16 18q11 1 1 3 RBBP8 - Loss RP11-19l3 18q21 2 0 2 AK127467 - Loss RP11-135L16 22q13 0 3 2 PARVB, PARVG, AK074042, - AB051431 Loss RP11-93F4 22q13 2 1 1 TRMT1, GTSE1, CELSR1 - aBased on the UCSC humane genome browser (May 2004 freeze). bX83301, AK130833, SERF1A, SMN1, SMN2, BIRC1, GTF2H2. cCAPN11, BC062989, c6orf110, AL353957, SLC29A1, HSPCB, SLC35B2, NFKBIE, AK130739, AARSL, MGC33600, AK126839, AL833884, SPATS1. dAK023827, LAT1-3TM, AF271775, BC062756, SPN, QPRT. eLarge-scale copy number variations (LCVs), presented as such in (A) Sebat et al. [34] and/or (B) Iafrate et al. [14].

99 Chapter 6

Subgroup 1 0.8

0.6

0.4

0.2

-0.2

-0.4

%LOSS %GAIN %LOSS -0.6 1 22 -0.8 A Clones ordered per chromosome by Mb position

Subgroup 2 0.8

0.6

0.4

0.2

-0.2

-0.4

%LOSS %GAIN %LOSS -0.6 1 22 -0.8 B Clones ordered per chromosome by Mb position

Subgroup 3 0.8

0.6

0.4

0.2

-0.2

-0.4

% LOSS % GAIN % LOSS % -0.6 1 22 -0.8 C Clones ordered per chromosome by Mb position

Figure 3. Overview of the genomewide frequency of copy number changes. The percentage of copy number gains in black (log2 T/R >0.3) and losses in grey (log2 T/R <-0.3) from chromosome 1 to 22 ordered by Mb position as presented in the UCSC genome browser (April 2003 freeze). A) Overview of the infantile GCT profiles with a teratoma histology (age < 5 years); subgroup1. B) Overview of the infantile GCT profiles with a yolk sac histology (age < 5 years); subgroup2. C) Overview of the GCT profiles derived from children above the age of 5 years with mixed histology; subgroup3. From this figure it is clear that these groups differ markedly in their number of abnormalities as well as the genomic regions involved.

100 Array CGH profiles of neonatal/infantile GCTs

Recurrent genomic aberrations in GCTs from children over 5 years old The array CGH profiles of the older patients (subgroup 3; >5 years) again revealed gross recurrent genomic abnormalities (Fig. 3C). Four out of seven cases showed anomalies encompassing multiple large genomic regions. Two cases showed profiles with only one or three gross genomic aberrations and in one case the profile appeared to be normal (Table 1). The most prevalent abnormality was gain of chromosome 12p (present in five out of seven cases; Fig. 3C and 2D). Four of the profiles showed gain of the complete 12p arm (cases 25, 26, 28 and 30), whereas in one case (case 29) copy number gains were detected in two distinct regions on 12p (12p13-pter (15 Mb) and 12p11 (2 Mb), respectively). Another gross recurrent copy number change was gain of 1q32-qter, observed in three out of seven cases. Copy number loss of (a part of) 6q was detected in two cases, gains of the entire chromosomes 7 and 8 in three cases, loss of 18q in two cases, gain of a complete chromosome 19 in two cases and gain of a complete chromosome 21 in two cases.

DISCUSSION

We used the recently array CGH technique to analyze 24 archival GCTs from children under the age of 5 years, which we divided into two subgroups based on histology (teratoma, subgroup 1; and YST, subgroup 2). In addition, seven GCTs of patients older than 5 years were included (subgroup 3). The 5-year age cut-off was chosen on basis of previous studies that showed a clear age distribution in germ cell tumors with respect to histology and tumor location [5;30]. The aim of this study was to obtain detailed information on the cytogenetic constitution of the different GCTs of neonates and infants in order to (i) identify recurrent chromosomal aberrations with diagnostic and/or prognostic implications and (ii) identify genomic regions that may harbor genes relevant to tumor formation. In agreement with the results of previous cytogenetic and conventional CGH studies, we found no gross anomalies in most of the teratoma-derived profiles (subgroup 1). In contrast, most YST-derived profiles (subgroup 2) showed gross recurrent copy number changes. These recurrent changes included losses of 1p (1p35-pter), gains of 1q (1q32-qter), gains of 3p (3p21-pter), losses of 6q (6q24-qter), losses of 18q (18q21-qter) and gains of 20q (20q13), all of which had been detected previously [13;21;25;33]. Losses of (distal parts of) 1p and 6q were considered the

101 Chapter 6 most prevalent genomic aberrations in YSTs of neonates and infants. We detected losses of the distal parts of 1p (1p35-pter) in half these cases. A similar percentage was found in an ongoing independent LOH study, which includes one of the present cases (case 24). One patient (case 21) with a YST exhibited a constitutional 1p22-p31 deletion next to a somatic 1p35-pter deletion in the tumor. In addition, the patient was mentally retarded and the constitutional deletion may be causally related to the mental status of this patient. This deletion may also have acted as a predisposing factor for YST development in this patient. This idea was supported by a similar loss of 1p21-p31 in three infantile YST cases described previously [21] and one additional case of subgroup 2 with a 1p22-p31 deletion in this study. The most prevalent anomaly in subgroup 2 was gain of (parts of) 20q with a shortest region of overlap of 11 Mb at 20q13, containing multiple genes (~90). Gain of 20q or 20q11-q13 is a common anomaly in human cancers [12;16] and has also been reported in GCTs, both in the adult and infantile types [19;21;24;31]. In contrast, such chromosome 20 anomalies were not noted in a large cytogenetic study of childhood GCTs [4]. This may be a results of the experimental limitations, i.e., an i(20q) chromosome can easily be overlooked by conventional cytogenetic analyses. Interestingly, gain of 20q and loss of 20p was also observed in a teratoma profile of a tumor that contained both teratoma and YST components (case 12). Subgroup 3 (> 5 years) showed heterogeneous profiles, with gain of (parts of) 12p as main aberration (5 of 7 cases). Gain of (parts of) 12p together with gain of the complete chromosomes 7 and 8 were observed in three cases. These chromosomal anomalies are common in adult testicular GCTs (group 2; see introduction [20] and have also been detected in several malignant ovarian GCTs [17;26;37]. As most GCTs from children older than 5 years in this study were ovarian GCTs, our data support the hypothesis that ovarian and testicular GCTs may share a common pathogenic pathway. In accordance with the literature, the two minimal regions of overrepresentation on chromosome 12p mapped to 12p13-pter and 12p11, respectively [11;27;46;]. So far, gain of a part of 12p has only rarely been observed in neonatal or infantile GCTs [21]. In line with these observations, gain of 12p was detected only once (case 20) in the GCTs under the age of 5 years (subgroups 1 and 2) in this study. Two profiles of the third subgroup resembled profiles of subgroup 1 and 2. Similar to the profiles of subgroup 1, the profile of an ovarian teratoma of an 8 year old showed no gross aberrations (case 27). The array CGH profile of a pelvic YST from a 16-year-old male patient (case 31) resembled the infantile or neonatal YST

102 Array CGH profiles of neonatal/infantile GCTs profiles in several aspects, including loss of 1p35-pter, loss of 6q and gain of 20q. No gain of 12p was observed. Together, these observations suggest that this tumor may actually fit into subgroup 2. Notably, this patient had suffered from a sacrococcygeal YST during infancy. Therefore, the tumor examined in the present study can be considered a metachronous secondary tumor. Unfortunately, no material of the primary tumor was available for analysis. Nevertheless, this observation indicates that some tumor components that reflect the childhood type of GCT may have persisted until the metachronous tumor manifested itself. The fact that the array CGH profile diverts from the profile that should be considered appropriate for the patient’s age, strongly supports that GCTs that develop during infancy are biologically distinct from those that develop during adulthood. Gains of (parts of) 1q, losses of (parts of) 6q and losses of (parts of) 18q were detected in profiles of tumors from both subgroup 2 and 3, implying that these regions may be related to the pathogenesis of GCTs in both these subgroups. In addition to the gross recurrent copy number changes discussed above, some novel small regions (< 5 Mb) encompassing one or a few BACs, showed recurrent copy number alterations (Table 3). Some of these regions were affected only in specific subgroups, including 2qter and 9q34 in teratomas (subgroup 1), 4q28 and 6p21 in YSTs (subgroup 2) and 4p15, 15q24 and 18q11 in GCTs of children older than age 5 (subgroup 3). Other regions showed gains or losses in all GCT subgroups, including regions on 5p15, 5q13, 9q32, 17p11 and 22q13. Recently, it was found that large-scale copy number variations (LCVs, encompassing 100 Kb or more) may significantly contribute to the array of naturally occurring genomic polymorphisms [14;34]. On the basis of this information, we evaluated whether the recurrent anomalies detected in this study coincided with any of the LCVs reported in the previous articles. Indeed, we found that several of the small regions presented in Table 3 were described as such, including those on 2qter, 5p15, 5q13. In reverse, the other recurrent anomalies that we detected may represent tumor-related copy number changes. Taken together, we have concluded that genomewide array CGH technology is highly suitable for the detection of recurrent genomic anomalies in GCTs of neonates and infants. This study confirmed data from previous studies indicating that most neonatal and infantile teratomas do not exhibit gross genomic aberrations. Furthermore, we concluded that the genomes of the GCTs differ markedly between the two age-groups as presented here. For neonatal and infantile YSTs, regions 20q13, 1p35-pter and 3p21-pter and individual BAC clone regions may harbor novel

103 Chapter 6 diagnostic/prognostic identifiers and genes relevant to the development of these neoplasms.

ACKNOWLEDGMENTS

The authors thank Rolph Pfundt and Huub Straatman for advice and support, Eric Schoenmakers and Pieter de Jong for BAC clone collections, and Walter van der Vliet and Henry Dijkman for expert technical assistance. Dominik Schneider and Suzanne Zahn are recipients of a Max-Eder grant by the Deutsche Krebshilfe e.V.

REFERENCES

1. Bernsen MR, Dijkman HB, de Vries E, Figdor CG, Ruiter DJ, Adema GJ, van Muijen GN. (1998): Identification of multiple mRNA and DNA sequences from small tissue samples isolated by laser-assisted microdissection. Lab Invest 78: 1267-1273. 2. Blohm ME, Vesterling-Horner D, Calaminus G, Göbel U. (1998): Alpha 1-fetoprotein (AFP) reference values in infants up to 2 years of age. Pediatr Hematol Oncol 15: 135-142. 3. Bussey KJ, Lawce HJ, Himoe E, Shu XO, Heerema NA, Perlman EJ, Olson SB, Magenis RE. (2001): SNRPN methylation patterns in germ cell tumors as a reflection of primordial germ cell development. Genes Chromosomes Cancer 32: 342-352. 4. Bussey KJ, Lawce HJ, Olson SB, Arthur DC, Kalousek DK, Krailo M, Giller R, Heifetz S, Womer R, Magenis RE. (1999):. Chromosome abnormalities of eighty-one pediatric germ cell tumors: sex- , age-, site-, and histopathology-related differences--a Children's Cancer Group study. Genes Chromosomes Cancer 25: 134-146. 5. Göbel U, Calaminus G, Engert J, Kaatsch P, Gadner H, Bokkerink JP, Hass RJ, Waag K, Blohm ME, Dippert S, Teske C, Harms D. (1998): Teratomas in infancy and childhood. Med Pediatr Oncol 31: 8-15. 6. Göbel U, Schneider DT, Calaminus G, Haas RJ, Schmidt P, Harms D. (2000): Germ-cell tumors in childhood and adolescence. GPOH MAKEI and the MAHO study groups. Ann Oncol 11: 263-271. 7. Gonzalez-Crussi F. 1982. Extragonadal teratomas. 2nd series, Fascicle 18. Washington DC: Armed forces Institute of Pathology. 8. Harms D, Janig U. (1986): Germ cell tumours of childhood. Report of 170 cases including 59 pure and partial yolk-sac tumours. Virchows Arch A Pathol Anat Histopathol 409: 223-239. 9. Hawkins E, Issacs H, Cushing B, Rogers P. (1993): Occult malignancy in neonatal sacrococcygeal teratomas. A report from a Combined Pediatric Oncology Group and Children's Cancer Group study. Am J Pediatr Hematol Oncol 15: 406-409. 10. Heidenblad M, Schoenmakers EF, Jonson T, Gorunova L, Veltman JA, Geurts van Kessel A, Hoglund M. (2004): Genome-wide array-based comparative genomic hybridization reveals multiple amplification targets and novel homozygous deletions in pancreatic carcinoma cell lines. Cancer Res 64: 3052-3059. 11. Henegariu O, Heerema NA, Thurston V, Jung SH, Pera M, Vance GH. (2004): Characterization of gains, losses, and regional amplification in testicular germ cell tumor cell lines by comparative genomic hybridization: Cancer Genet Cytogenet 148: 14-20.

104 Array CGH profiles of neonatal/infantile GCTs

12. Hodgson JG, Chin K, Collins C, Gray JW. (2003): Genome amplification of chromosome 20 in breast cancer: Breast Cancer Res Treat 78: 337-345. 13. Hu J, Schuster AE, Fritsch MK, Schneider DT, Lauer S, Perlman EJ. (2001): Deletion mapping of 6q21-26 and frequency of 1p36 deletion in childhood endodermal sinus tumors by microsatellite analysis. Oncogene 20: 8042-8044. 14. Iafrate AJ, Feuk L, Rivera MN, Listewnik ML, Donahoe PK, Qi Y, Scherer SW, Lee C. (2004): Detection of large-scale variation in the human genome. Nat Genet 36: 949-951. 15. Isaacs H, Jr. (2004): Perinatal (fetal and neonatal) germ cell tumors. J Pediatr Surg 39: 1003- 1013. 16. Kimura Y, Noguchi T, Kawahara K, Kashima K, Daa T, Yokoyama S.(2004): Genetic alterations in 102 primary gastric cancers by comparative genomic hybridization: gain of 20q and loss of 18q are associated with tumor progression. Mod Pathol 17: 1328-1337. 17. Kraggerud SM, Szymanska J, Abeler VM, Kaern J, Eknaes M, Heim S, Teixeira MR, Trope CG, Peltomaki P, Lothe RA. (2000): DNA copy number changes in malignant ovarian germ cell tumors. Cancer Res 60: 3025-3030. 18. Looijenga LH and Oosterhuis JW. (1999): Pathogenesis of testicular germ cell tumours. Rev Reprod 4: 90-100. 19. Looijenga LH, Rosenberg C, Van Gurp RJ, Geelen E, Echten-Arends J, de Jong B, Mostert M, Oosterhuis JW. (2000): Comparative genomic hybridization of microdissected samples from different stages in the development of a seminoma and a non-seminoma. J Pathol 191: 187-192. 20. Looijenga LH, Zafarana G, Grygalewicz B, Summersgill B, Debiec-Rychter M, Veltman J, Schoenmakers EF, Rodriguez S, Jafer O, Clark J, Geurts van Kessel A, Shipley J, Van Gurp RJ, Gillis AJ, Oosterhuis JW. (2003): Role of gain of 12p in germ cell tumour development. APMIS 111: 161-171. 21. Mostert M, Rosenberg C, Stoop H, Schuyer M, Timmer A, Oosterhuis J, Looijenga L. (2000): Comparative genomic and in situ hybridization of germ cell tumors of the infantile testis. Lab Invest 80: 1055-1064. 22. Oosterhuis JW, Looijenga LH, van Echten J, de Jong B. (1997): Chromosomal constitution and developmental potential of human germ cell tumors and teratomas. Cancer Genet Cytogenet 95: 96-102. 23. Perlman EJ, Cushing B, Hawkins E, Griffin CA. (1994): Cytogenetic analysis of childhood endodermal sinus tumors: a Pediatric Oncology Group study. Pediatr Pathol 14: 695-708. 24. Perlman EJ, Hu J, Ho D, Cushing B, Lauer S, Castleberry RP. (2000): Genetic analysis of childhood endodermal sinus tumors by comparative genomic hybridization. J Pediatr Hematol Oncol 22: 100-105. 25. Perlman EJ, Valentine MB, Griffin CA, Look AT. (1996): Deletion of 1p36 in childhood endodermal sinus tumors by two-color fluorescence in situ hybridization: a pediatric oncology group study. Genes Chromosomes Cancer 16: 15-20. 26. Riopel MA, Spellerberg A, Griffin CA, Perlman EJ. (1998): Genetic analysis of ovarian germ cell tumors by comparative genomic hybridization. Cancer Res 58: 3105-3110. 27. Rodriguez S, Jafer O, Goker H, Summersgill BM, Zafarana G, Gillis AJ, Van Gurp RJ, Oosterhuis JW, Lu YJ, Huddart R, Cooper CS, Clark J, Looijenga LH, Shipley JM. (2003): Expression profile of genes from 12p in testicular germ cell tumors of adolescents and adults associated with i(12p) and amplification at 12p11.2-p12.1. Oncogene 22: 1880-1891. 28. Ross JA, Schmidt PT, Perentesis JP, Davies SM. (1999): Genomic imprinting of H19 and insulin-like growth factor-2 in pediatric germ cell tumors. Cancer 85: 1389-1394. 29. Sano K. (1999): Pathogenesis of intracranial germ cell tumors reconsidered. J Neurosurg 90: 258-264. 30. Schneider DT, Calaminus G, Koch S, Teske C, Schmidt P, Haas RJ, Harms D, Göbel U. (2004): Epidemiologic analysis of 1,442 children and adolescents registered in the German germ cell tumor protocols. Pediatr Blood Cancer 42: 169-175. 31. Schneider DT, Schuster AE, Fritsch MK, Calaminus G, Göbel U, Harms D, Lauer S, Olson T, Perlman EJ. (2002): Genetic analysis of mediastinal nonseminomatous germ cell tumors in children and adolescents. Genes Chromosomes Cancer 34: 115-125.

105 Chapter 6

32. Schneider DT, Schuster AE, Fritsch MK, Calaminus G, Harms D, Göbel U, Perlman EJ. (2001a): Genetic analysis of childhood germ cell tumors with comparative genomic hybridization. Klin Padiatr 213: 204-211. 33. Schneider DT, Schuster AE, Fritsch MK, Hu J, Olson T, Lauer S, Göbel U, Perlman EJ. (2001b): Multipoint imprinting analysis indicates a common precursor cell for gonadal and nongonadal pediatric germ cell tumors. Cancer Res 61: 7268-7276. 34. Sebat J, Lakshmi B, Troge J, Alexander J, Young J, Lundin P, Maner S, Massa H, Walker M, Chi M, Navin N, Lucito R, Healy J, Hicks J, Ye K, Reiner A, Gilliam TC, Trask B, Patterson N, Zetterberg A, Wigler M. (2004):. Large-scale copy number polymorphism in the human genome. Science 305: 525-528. 35. Stoop H, van Gurp R, de Krijger R, Geurts van Kessel A, Koberle B, Oosterhuis W, Looijenga L. (2001): Reactivity of germ cell maturation stage-specific markers in spermatocytic seminoma: diagnostic and etiological implications. Lab Invest 81: 919-928. 36. van Echten J, Timmer A, van der Veen AY, Molenaar WM, de Jong B. (2002): Infantile and adult testicular germ cell tumors. a different pathogenesis? Cancer Genet Cytogenet 135: 57- 62. 37. van Echten J, van Doorn LC, van der Linden HC, van der Veen AY, Burger CW, de Jong B. (1998): Cytogenetics of a malignant ovarian germ-cell tumor. Int J Cancer 77: 217-218. 38. Van Gurp RJ, Oosterhuis JW, Kalscheuer V, Mariman EC, Looijenga LH. (1994): Biallelic expression of the H19 and IGF2 genes in human testicular germ cell tumors. J Natl Cancer Inst 86: 1070-1075. 39. Veltman I, van Asseldonk M, Schepens M, Stoop H, Looijenga L, Wouters C, Govaerts L, Suijkerbuijk R, Geurts van Kessel A. (2002): A novel case of infantile sacral teratoma and a constitutional t(12;15)(q13;q25) pat. Cancer Genet Cytogenet 136:17-22. 40. Veltman IM, Schepens MT, Looijenga LHJ, Strong LC, Geurts van Kessel A. (2003): Germ cell tumours in neonates and infants: a distinct subgroup? APMIS 111: 152-160. 41. Veltman JA, Fridlyand J, Pejavar S, Olshen AB, Korkola JE, de Vries S, Carroll P, Kuo WL, Pinkel D, Albertson D, Cordon-Cardo C, Jain AN, Waldman FM. (2003): Array-based comparative genomic hybridization for genome-wide screening of DNA copy number in bladder tumors. Cancer Res 63: 2872-2880. 42. Vissers LE, de Vries BB, Osoegawa K, Janssen IM, Feuth T, Choy CO, Straatman H, Van der Vliet W, Huys EH, van Rijk A, Smeets D, Ravenswaaij-Arts CM, Knoers NV, van der Burgt I, de Jong PJ, Brunner HG, Geurts van Kessel A, Schoenmakers EF, Veltman JA. (2003): Array- based comparative genomic hybridization for the genomewide detection of submicroscopic chromosomal abnormalities. Am J Hum Genet 73: 1261-1270. 43. Vissers LE, van Ravenswaaij CM, Admiraal R, Hurst JA, de Vries BB, Janssen IM, van der Vliet WA, Huys EH, de Jong PJ, Hamel BC, Schoenmakers EF, Brunner HG, Veltman JA, Geurts van Kessel A. (2004): Mutations in a new member of the chromodomain gene family cause CHARGE syndrome. Nat Genet 36: 955-957. 44. Wilhelm M, Veltman JA, Olshen AB, Jain AN, Moore DH, Presti JC Jr, Kovacs G, Waldman FM. (2002): Array-based comparative genomic hybridization for the differential diagnosis of renal cell cancer. Cancer Res 62: 957-960. 45. Woodward PJ, Heidenreich A, Looijenga LHJ et al. (2004): Testicular germ cell tumors. In: Eble JN, Sauter G, Epstein JI and Sesterhenn IA, editors. World Health Organization Classification of Tumours. Pathology and Genetics of the Urinary System and Male Genital Organs. Lyon: IARC Press. p 217-278. 46. Zafarana G, Grygalewicz B, Gillis AJ, Vissers LE, van der Vliet WA, Van Gurp RJ, Stoop H, Debiec-Rychter M, Oosterhuis JW, Geurts van Kessel A, Schoenmakers EF, Looijenga LH, Veltman JA. (2003): 12p-amplicon structure analysis in testicular germ cell tumors of adolescents and adults by array CGH. Oncogene 22: 7695-7701

106

General discussion and future prospects

GENERAL DISCUSSION

Many human tumours are characterized by non-random chromosomal aberrations [1]. These aberrations have been instrumental for (i) improvement of diagnosis and prognosis of various tumour (sub)types [2;3] and (ii) to gain insight in the genetic mechanisms that underlie the development of different tumour (sub)types [4;5]. Moreover, in some cases detailed (cyto)genetic knowledge has resulted in the design and subsequent clinical application of efficient and specifically targeted cancer therapies [6;7]. Tumour-associated chromosomal aberrations may occur as inherited/constitutional anomalies or as anomalies that are somatically acquired by (precursor) cells during the processes of tumour initiation and/or progression. Also in human germ cell tumours (GCTs), non-random chromosomal aberrations have been documented [8]. Until recently, however, only limited information was available on the (cyto)genetic constitution of neonatal and infantile GCTs [chapter 1, [9-15]. In this thesis, both constitutional and acquired chromosomal aberrations were studied aiming at the identification of diagnostic/prognostic identifiers and genes that play a role in the development of this latter group of GCTs. To this end, advanced positional cloning and microarray-based genomic profiling strategies were used.

Constitutional aberrations The 12q11-q13 region has been suggested as a candidate region that may harbour one or more genes relevant for the development of GCTs [16-18]. Interestingly, this region was also involved in two constitutional chromosome translocations in patients with infantile GCTs: t(12;15)(q13;q25) and t(8;12)(q23;q12), respectively. Therefore, we set out to positionally clone the breakpoints and genes associated with these translocations [chapters 2, 3 and 4]. As a result of the constitutional t(12;15)(q13;q25) translocation, the SENP1 gene on chromosome 12 and the MESDC2 gene on chromosome 15 were disrupted and fused [chapter 3]. Both reciprocal fusion genes were present and transcribed in the tumour cells and their open reading frames predicted aberrant proteins. Current knowledge on the function of wild-type SENP1 indicates that this protein can modify the precursor form of the small ubiquitin-related modifier 1 (SUMO-1) and can deconjugate mature SUMO-1 from its target proteins [19-22]. Although in vitro studies have shown that SENP1 can function as a SUMO-2 and SUMO-3 protease as well [20;22], a recent mouse study indicated that SENP1 primarily acts as a SUMO-1

111 Chapter 7 protease [19]. As yet, it is unknown whether SENP1 can also act as a protease for SUMO-4 [23]. The recent mouse study also revealed that SENP1 plays a critical role during embryogenesis, i.e., a homozygous retroviral insertion into the first intron of the mouse ortholog of SENP1 (SuPr-2) caused a 50-fold reduction in SuPr-2 transcription and an increase in sumoylated proteins and unprocessed SUMO-1. Together, these anomalies resulted in an embryonic lethal phenotype due to placental failure [19]. SUMO-1 is a post-translational modifier and, depending on the affected target protein, this modification may alter its sub-cellular localization, its activity, its stability and/or its protein-protein interaction network [24;25]. Consequently, these modifications may affect various cellular key processes, including regulation of transcription, maintenance of genome integrity and/or signal transduction. In chapter 3 we showed that the two MESDC2-SENP1 derived SENP1 proteins (MESEs) are still able to desumoylate target proteins such as PML in a desumoylation assay. Recently, other groups showed that, next to PML, SENP1 is also able to deconjugate SUMO-1 from HDAC1 [26], ERM [27], SF1 [28], C-Jun [29] and IE1 [20]. It is expected that in the future more SENP1 target proteins will be added to this still growing list. Therefore, we assume that SENP1 may affect a plethora of cellular processes. Based on the MESDC2-SENP1-associated promoter swap [chapter 3], we suggest that the normal spatio-temporal expression pattern of SENP1 may have been altered in the patient. However, it remains to be established which cellular processes were altered, and how these altered processes contributed to the development of the infantile GCT. Additional support for a putative role of sentrin-specific proteases in tumour development comes from two recent findings: (i) one of the SENP family members, SENP6, is also disrupted and fused to a gene (TCBA1) in the lymphoma-derived cell line HT1 [30] and (ii) SENP1 may be involved in thyroid oncocytic adenomas [31]. Transient expression studies showed that MESD (the MESDC2 mouse homolog) acts as an ER chaperone for several LDL receptors, including the WNT co- receptors LRP5 and LRP6 [32;33]. In addition, MESD null mutations led to an embryonic lethal phenotype, resulting from a lack of primitive streak formation and mesoderm induction due to defective patterning of the proximal epiblast [32;34;35]. Hsieh et al. suggested that this patterning defect might be caused by anomalous WNT signalling resulting from an incorrect intracellular routing of the LRP5 and LRP6 co-receptors [32]. This hypothesis was supported by a recent study in which LRP5/LRP6 double knockout mice exhibited similar phenotypic features as MESD mutant mice [36]. We found that the t(12;15)(q13;q25)-associated N-terminally

112 General discussion and future prospects truncated MESDC2 protein (SEME) is no longer targeted to the intracellular ER compartment [chapter 3], indicating that this protein may have lost its capacity to act as an ER chaperone for the LDL receptor family members. During embryonic development, germ cells originate from the proximal epiblast [chapter 1], a structure that showed a patterning defect in the MESD mutant mice. Based on these observations and the finding that one MESDC2 wild-type allele is still present in the tumour cells [chapters 2 and 3], we suggest that MESDC2 haplo-insufficiency and/or anomalous subcellular localization of SEME proteins may lead to an abnormal development and/or routing of uncommitted primordial germ cells, the presumed precursors of infantile GCTs. At the chromosome architectural level, it is of interest to note that SENP1 is located within the FRA12A region on chromosome 12 [37]. FRA12A is a rare, folate- sensitive fragile site that coincides with YAC clone 919G8, which contains SENP1 next to 41 other genes [37, chapter 2]. As fragile sites are prone to DNA double- strand break formation, FRA12A may underlie the occurrence of the t(12;15)(q13;q25) translocation. Previously, fragile sites have been suggested to be involved in the pathogenesis of human tumours [38;39]. A well-known example of the latter is FRA3B, a fragile site that coincides with the FHIT gene which is disrupted by the familial renal cell carcinoma-associated t(3;8)(p14;q24) translocation [39]. So far, however, no association with tumour development and FRA12A has been documented. FRA12A has been associated with mental retardation and congenital anomalies, including some distinct dysmorphic facial features [reviewed in 37]. It should be noted that, next to the GCT, the t(12;15)(q13;q25)-positive patient also showed some dysmorphic features, including an upward slant and obesitas [chapter 2]. Although these latter features do not overlap with those mentioned above, it cannot be excluded that the constitutional t(12;15)(q13;q25) may have contributed to the occurrence of these dysmorphic features. The chromosome 12 breakpoint of the constitutional t(8;12)(q23;q12) translocation appeared to map in a region that, as yet, is devoid of any annotated genes and/or known miRNA-coding sequences (May 2004 freeze of the UCSC human genome browser) [chapter 4]. Moreover, this breakpoint maps ~4 Mb from the SENP1 gene, which indicates that SENP1 may not be affected by the t(8;12)(q23;q12) translocation. This idea was substantiated by our preliminary transcription assays, which failed to show any significant differences between t(8;12)(q23;q12)-positive and -negative cells. Similar to the 12q12 breakpoint, the 8q23 breakpoint was mapped outside known genes and/or miRNA-coding sequences,

113 Chapter 7 and our preliminary transcription data again suggested that positional candidate genes flanking the breakpoints are not affected. However, the possibility remains that t(8;12)(q23;q12) may affect the expression of more distant genes and/or regulatory elements that may play a role in the initiation/progression of the infantile GCT. As a result of the Humane Genome Project, several new technologies were developed that may facilitate the search for positional candidate genes. One of these technologies is microarray-based comparative genomic hybridization (array CGH) [40;41]. We used array CGH in combination with flow-sorted derivative 12 chromosomes to map the t(12;15)(q13;q25)-associated breakpoints in a single hybridization experiment. By doing so, we found that the results obtained [chapter 5] fully correspond to those obtained through our positional cloning efforts [chapter 3], thereby indicating that this novel integrated array CGH technology serves as a very efficient high-throughput tool to map chromosomal aberrations and to identify its associated candidate genes.

Acquired aberrations At the start of this thesis-work (cyto)genetic information on neonatal and infantile GCTs was scarce [chapter 1]. Therefore, a detailed analysis was initiated to screen this group of GCTs for the presence of acquired chromosomal aberrations using the array CGH technology [chapter 6]. Such aberrations may (i) serve as diagnostic/prognostic identifiers and (ii) point at regions harbouring genes underlying the development of these tumours. The GCTs included in our study were classified as type I according to the criteria outlined in chapter 1 [8;42]. Among these GCTs, a subgroup consisting of yolk sac tumours (YSTs) exhibited recurrent losses of 1p35-pter and 6q24-qter and recurrent gains of 1q32-qter, 3p21-pter and 20q13. Similar data were reported by others in type I YSTs using alternative methods such as conventional karyotyping and/or conventional chromosomal CGH [9-15]. In contrast to the YSTs, a subgroup of type I teratomas failed to exhibit any gross recurrent genomic aberrations, even after detailed analysis with our genome-wide BAC arrays [chapter 6]. This result agrees with several other studies reported to date [9;10;13-15] and may explain why these tumours usually behave in a benign fashion and rarely progress to malignancy. Instead, minute genetic alterations, epigenetic changes and/or microenvironmental factors (stroma) may play a role in the development of these teratomas. Currently, this latter option is being investigated by several groups for different tumour types, including GCTs [43-45]. Microsatellite

114 General discussion and future prospects analysis showed that the stroma cells adjacent to metastatic mature type II teratomas (after chemotherapy) may exhibit similar genomic anomalies as the tumour component itself, suggesting that these stroma cells may actually be derived from the same progenitor cells [46]. In the constitutional t(12;15)(q13;q25)-positive infantile GCT (see above), a YST component was present next to a teratoma component. By using array CGH we found that this tumour, next to the translocation, exhibited a gain of 20q and a loss of 20p [chapter 6, case 12]. As gain of 20q was also detected in almost all infantile/neonatal YSTs [chapter 6], we speculate that the chromosome 20 anomalies observed in the former case may reflect the development of a malignant endodermal sinus (YST) component (i.e., tumour progession). Type II GCTs exhibit genomic profiles distinct from those observed in the type I GCTs. An overview of the most frequent recurrent genomic aberrations encountered in these two types of GCTs is depicted in Table 1. Some aberrations, however, were detected in both tumour types, i.e., gains of (parts of) 1q, and losses of (parts of) 1p, 6q and 18q, although we found that type I GCTs exhibited relatively higher percentages of losses of (parts of) 1p and 6q (in ~50% of the type I GCTs studied). Also, an occasional type I YST showed gain of (a part of) 12p (notably 12p13) [chapter 6; see also 10].

Table 1. Recurrent genomic aberrations in type I and II GCTs. The recurrent aberrations in type I GCTs were based on chapter 6 and [9-15], and those in type II GCTs were based on a recent review of Oosterhuis et al. [8].

Histological Type of GCT Copy number losses Copy number gains component YST 1p, 4(q), 6q 1q, 3(p), 20q Type I TE - - SE, NS (including Type II 1p, 11, 13, 18 and Y 7,8, 12p, 21 and X YST and TE) YST = yolk sac tumour, TE = teratoma, SE = seminoma, NS = non-seminoma.

Some of the recurrent genomic anomalies [Table 1] are not restricted to GCTs but also occur in other, distinct, tumour types. This latter observation suggests that the genomic regions involved may harbour genes that affect a broader spectrum of target tissues, or reflect unstable regions that are prone to chromosomal aberrations. In addition, some of the recurrent genomic anomalies that we detected in the type I GCTs, i.e., loss of 1p36 and gain of 20q13, have also been associated with advanced

115 Chapter 7 stages of tumour development [47-51]. Therefore, these regions may harbour progression-related genes (see also above). Based on occurrence, clinical outcome, histology and genetic constitution we [chapter 1 and 6] and others [11;42;52;53] have suggested that neonatal/infantile GCTs may constitute a distinct pathogenic subgroup. This hypothesis was further substantiated by the genomic profiles that we observed in the subgroup of GCTs above the age of 5 years [chapter 6]. Based on Table 1, these profiles suggest a mix of type I and type II GCTs. Compared to testicular type II GCTs (> 15 years), ovarian type II GCTs usually occur at a younger age (> 4 years) [chapter 1 and [8;42]. As the above-described GCT subgroup mainly consisted of ovarian GCTs, this might explain the mix of type I and type II GCTs within this subgroup (> 5 years). The recurrent aberrations as depicted in Table 1 may, therefore, serve as additional diagnostic tools for this subgroup. Taken together, we conclude that array-based genomic profiles may serve as excellent diagnostic/prognostic identifiers for neonatal and infantile GCTs and, in addition, may point at genomic regions that may harbour genes [see chapter 6] relevant for the initiation and/or progression of these tumours.

Future prospects

Through the constitutional t(12;15)(q13;q25) two novel candidate genes, SENP1 and MESDC2, were identified that, as a result of their disruption and fusion, may have contributed to the development of the infantile GCT. Obviously, additional studies are required (i) to clarify the role of the respective fusion genes in the development of this GCT, and (ii) to assess whether the genes involved play a role in the development of sporadic GCTs and/or normal germ cells. Such information may also explain why other family members carrying the same constitutional translocation did not develop any GCT. In order to assess the role of SENP1 and MESDC2 in sporadic GCTs, we suggest a screening of multiple tumour samples for mutations, copy number changes and aberrant expression patterns, including those of downstream targets. For analyses at the protein level, the availability of suitable antibodies, tumour samples and matched normal control tissues will be essential. Type I GCTs are complex tumours capable of forming highly differentiated structures, resembling embryogenesis in several aspects. Studies in mice revealed that disturbances in germ cell development during embryogenesis may predispose to GCTs [54-56]. In order to completely comprehend the development of type I GCTs

116 General discussion and future prospects from a precursor cell to a full blown tumour an integrated multi-disciplinary approach is suggested. Such an approach would include studies on (i) factors involved in formation, routing and homing of normal germ cells and in re- programming these cells, (ii) DNA, mRNA and protein constitution of various tumour stages, and (iii) interactions of the surrounding tissues with tumour cells and/or their precursors. Several of these approaches have already been initiated, including studies on normal germ cell development [57, for review], imprinting studies to search for the progenitor cell [13] and the (cyto)genetic studies [chapter 6 and 9-15], including our array CGH study in which we analyzed the type I GCT genome at a 1 Mb resolution [chapter 6]. Over the past few years the microarray technology has rapidly evolved and several new applications are becoming available [58] that, e.g., will provide ploidy information at a high resolution. These include (i) 32k BAC arrays with a tiling-path resolution of the complete human genome [59;60], (ii) SNP arrays that allow a detailed screening for the presence of uniparental disomies, polymorphisms and copy number changes that may predispose to GCT development [61], and (iii) various oligonucleotide arrays that allow high resolution DNA copy number and/or gene expression analysis [62-64]. Additionally, high- density cDNA or oligonucleotide arrays can be used for simultaneous, integrated analyses of expression and genomic anomalies [64-67]. For all above assays the availability of well-defined and well-preserved tissues and/or cells is crucial, next to matched control tissues and/or cells for reference. Once defined, candidate genes can be tested for their role in the initiation and/or progression of type I GCTs using cell- and animal model systems. For these purposes, several suitable mouse models are currently available [55;56;68-70] and reviewed in [8]. Since these mice develop teratomas that closely resemble human type I teratomas, they may be of particular relevance for unravelling the in vivo process of type I GCT development.

REFERENCES

1. Mitelman Database of Chromosome Aberrations in Cancer (2005). Mitelman F, Johansson B and Mertens F. http://cgap.nci.nih.gov/Chromosomes/Mitelman. 2. Fletcher CDM, Unni KK, Mertens F (2002): World Health Organization classification of tumours. Tumours of Soft Tissue and Bone. IARC Press, Lyon, France. 3. Jaffe ES, Lee Harris N, Stein H, Vardiman JW (2001): World Health Organization classification of tumours. Tumours of Haematopoietic and Lymphoid Tissues. IARC Press, Lyon, France. 4. Vogelstein B, Kinzler KW (2004): Cancer genes and the pathways they control. Nat.Med. 10:789-799.

117 Chapter 7

5. Weir B, Zhao X, Meyerson M (2004): Somatic alterations in the human cancer genome. Cancer Cell 6:433-438. 6. Druker BJ, Talpaz M, Resta DJ, Peng B, Buchdunger E, Ford JM, Lydon NB, Kantarjian H, Capdeville R, Ohno-Jones S, Sawyers CL (2001): Efficacy and safety of a specific inhibitor of the BCR-ABL tyrosine kinase in chronic myeloid leukemia. N.Engl.J.Med. 344:1031-1037. 7. Savli H, Sirma S, Nagy B, Aktan M, Dincol G, Ozbek U (2003): Real-time PCR analysis of the apoptosis related genes in ATRA treated APL t(15;17) patients. Exp.Mol.Med. 35:454-459. 8. Oosterhuis JW, Looijenga LH (2005): Testicular germ-cell tumours in a broader perspective. Nat.Rev.Cancer 5:210-222. 9. Bussey KJ, Lawce HJ, Olson SB, Arthur DC, Kalousek DK, Krailo M, Giller R, Heifetz S, Womer R, Magenis RE (1999): Chromosome abnormalities of eighty-one pediatric germ cell tumors: sex- , age-, site-, and histopathology-related differences--a Children's Cancer Group study. Genes Chromosomes.Cancer 25:134-146. 10. Mostert M, Rosenberg C, Stoop H, Schuyer M, Timmer A, Oosterhuis J, Looijenga L (2000): Comparative genomic and in situ hybridization of germ cell tumors of the infantile testis. Lab Invest 80:1055-1064. 11. Perlman EJ, Cushing B, Hawkins E, Griffin CA (1994): Cytogenetic analysis of childhood endodermal sinus tumors: a Pediatric Oncology Group study. Pediatr.Pathol. 14:695-708. 12. Perlman EJ, Hu J, Ho D, Cushing B, Lauer S, Castleberry RP (2000): Genetic analysis of childhood endodermal sinus tumors by comparative genomic hybridization. J.Pediatr.Hematol.Oncol. 22:100-105. 13. Schneider DT, Schuster AE, Fritsch MK, Hu J, Olson T, Lauer S, Gobel U, Perlman EJ (2001): Multipoint imprinting analysis indicates a common precursor cell for gonadal and nongonadal pediatric germ cell tumors. Cancer Res. 61:7268-7276. 14. Schneider DT, Schuster AE, Fritsch MK, Calaminus G, Harms D, Gobel U, Perlman EJ (2001): Genetic analysis of childhood germ cell tumors with comparative genomic hybridization. Klin.Padiatr. 213:204-211. 15. Schneider DT, Schuster AE, Fritsch MK, Calaminus G, Gobel U, Harms D, Lauer S, Olson T, Perlman EJ (2002): Genetic analysis of mediastinal nonseminomatous germ cell tumors in children and adolescents. Genes Chromosomes.Cancer 34:115-125. 16. Murty VV, Houldsworth J, Baldwin S, Reuter V, Hunziker W, Besmer P, Bosl G, Chaganti RS (1992): Allelic deletions in the long arm of chromosome 12 identify sites of candidate tumor suppressor genes in male germ cell tumors. Proc.Natl.Acad.Sci.U.S.A 89:11006-11010. 17. Rodriguez E, Mathew S, Reuter V, Ilson DH, Bosl GJ, Chaganti RS (1992): Cytogenetic analysis of 124 prospectively ascertained male germ cell tumors. Cancer Res. 52:2285-2291. 18. Samaniego F, Rodriguez E, Houldsworth J, Murty VV, Ladanyi M, Lele KP, Chen QG, Dmitrovsky E, Geller NL, Reuter V (1990): Cytogenetic and molecular analysis of human male germ cell tumors: chromosome 12 abnormalities and gene amplification. Genes Chromosomes.Cancer 1:289-300. 19. Yamaguchi T, Sharma P, Athanasiou M, Kumar A, Yamada S, Kuehn MR (2005): Mutation of SENP1/SuPr-2 Reveals an Essential Role for Desumoylation in Mouse Development. Mol.Cell Biol. 25:5171-5182. 20. Gravel A, Dion V, Cloutier N, Gosselin J, Flamand L (2004): Characterization of human herpesvirus 6 variant B immediate-early 1 protein modifications by small ubiquitin-related modifiers. J.Gen.Virol. 85:1319-1328. 21. Johnson ES (2004): Protein modification by SUMO. Annu.Rev.Biochem. 73:355-382. 22. Xu Z, Au SW (2005): Mapping residues of SUMO precursors essential in differential maturation by SUMO-specific protease, SENP1. Biochem.J. 386:325-330. 23. Bohren KM, Nadkarni V, Song JH, Gabbay KH, Owerbach D (2004): A M55V polymorphism in a novel SUMO gene (SUMO-4) differentially activates heat shock transcription factors and is associated with susceptibility to type I diabetes mellitus. J.Biol.Chem. 279:27233-27238. 24. Dohmen RJ (2004): SUMO protein modification. Biochim.Biophys.Acta 1695:113-131. 25. Hay RT (2005): SUMO: a history of modification. Mol.Cell 18:1-12.

118 General discussion and future prospects

26. Cheng J, Wang D, Wang Z, Yeh ET (2004): SENP1 enhances androgen receptor-dependent transcription through desumoylation of histone deacetylase 1. Mol.Cell Biol. 24:6021-6028. 27. Degerny C, Monte D, Beaudoin C, Jaffray E, Portois L, Hay RT, de Launoit Y, Baert JL (2005): SUMO-modification of the Ets-related transcription factor ERM inhibits its transcriptional activity. J.Biol.Chem. 280:24330-8. 28. Lee MB, Lebedeva LA, Suzawa M, Wadekar SA, Desclozeaux M, Ingraham HA (2005): The DEAD-box protein DP103 (Ddx20 or Gemin-3) represses orphan nuclear receptor activity via SUMO modification. Mol.Cell Biol. 25:1879-1890. 29. Cheng J, Perkins ND, Yeh ET (2005): Differential regulation of c-Jun-dependent transcription by SUMO-specific proteases. J.Biol.Chem. 280:14492-8. 30. Tagawa H, Miura I, Suzuki R, Suzuki H, Hosokawa Y, Seto M (2002): Molecular cytogenetic analysis of the breakpoint region at 6q21-22 in T-cell lymphoma/leukemia cell lines. Genes Chromosomes.Cancer 34:175-185. 31. Jacques C, Baris O, Prunier-Mirebeau D, Savagner F, Rodien P, Rohmer V, Franc B, Guyetant S, Malthiery Y, Reynier P (2004): Two-step differential expression analysis reveals a new set of genes involved in thyroid oncocytic tumors. J.Clin.Endocrinol.Metab. 90:2314-20. 32. Hsieh JC, Lee L, Zhang L, Wefer S, Brown K, DeRossi C, Wines ME, Rosenquist T, Holdener BC (2003): Mesd encodes an LRP5/6 chaperone essential for specification of mouse embryonic polarity. Cell 112:355-367. 33. Culi J, Mann RS (2003): Boca, an endoplasmic reticulum protein required for wingless signaling and trafficking of LDL receptor family members in Drosophila. Cell 112:343-354. 34. Holdener BC, Faust C, Rosenthal NS, Magnuson T (1994): Msd is required for mesoderm induction in mice. Development 120:1335-1346. 35. Wines ME, Shi Y, Lindor M, Holdener BC (2000): Physical localization of the mesoderm development (mesd) functional region. Genomics 68:322-329. 36. Kelly OG, Pinson KI, Skarnes WC (2004): The Wnt co-receptors Lrp5 and Lrp6 are essential for gastrulation in mice. Development 131:2803-2815. 37. Winnepenninckx B, Reyniers E, Bossuyt P, Smits A, Wauters J, Kooy F (2001): Precise mapping of the fragile site FRA12A on chromosome 12q13.1. GeneScreen 1:131-137. 38. Finnis M, Dayan S, Hobson L, Chenevix-Trench G, Friend K, Ried K, Venter D, Woollatt E, Baker E, Richards RI (2005): Common chromosomal fragile site FRA16D mutation in cancer cells. Hum.Mol.Genet. 14:1341-1349. 39. Ohta M, Inoue H, Cotticelli MG, Kastury K, Baffa R, Palazzo J, Siprashvili Z, Mori M, McCue P, Druck T, . (1996): The FHIT gene, spanning the chromosome 3p14.2 fragile site and renal carcinoma-associated t(3;8) breakpoint, is abnormal in digestive tract cancers. Cell 84:587-597. 40. Albertson DG, Pinkel D (2003): Genomic microarrays in human genetic disease and cancer. Hum.Mol.Genet. 12 Spec No 2:R145-R152. 41. Pinkel D, Albertson DG (2005): Array comparative genomic hybridization and its applications in cancer. Nat.Genet. 37 Suppl:S11-S17. 42. Schneider DT, Calaminus G, Koch S, Teske C, Schmidt P, Haas RJ, Harms D, Gobel U (2004): Epidemiologic analysis of 1,442 children and adolescents registered in the German germ cell tumor protocols. Pediatr.Blood Cancer 42:169-175. 43. Bhowmick NA, Moses HL (2005): Tumor-stroma interactions. Curr.Opin.Genet.Dev. 15:97- 101. 44. Jodele S, Chantrain CF, Blavier L, Lutzko C, Crooks GM, Shimada H, Coussens LM, Declerck YA (2005): The contribution of bone marrow-derived cells to the tumor vasculature in neuroblastoma is matrix metalloproteinase-9 dependent. Cancer Res. 65:3200-3208. 45. Orimo A, Gupta PB, Sgroi DC, Arenzana-Seisdedos F, Delaunay T, Naeem R, Carey VJ, Richardson AL, Weinberg RA (2005): Stromal fibroblasts present in invasive human breast carcinomas promote tumor growth and angiogenesis through elevated SDF-1/CXCL12 secretion. Cell 121:335-348. 46. Brandli DW, Ulbright TM, Foster RS, Cummings OW, Zhang S, Sweeney CJ, Eble JN, Cheng L (2003): Stroma adjacent to metastatic mature teratoma after chemotherapy for testicular germ

119 Chapter 7

cell tumors is derived from the same progenitor cells as the teratoma. Cancer Res. 63:6063- 6068. 47. Hodgson JG, Chin K, Collins C, Gray JW (2003): Genome amplification of chromosome 20 in breast cancer. Breast Cancer Res.Treat. 78:337-345. 48. Kimura Y, Noguchi T, Kawahara K, Kashima K, Daa T, Yokoyama S (2004): Genetic alterations in 102 primary gastric cancers by comparative genomic hybridization: gain of 20q and loss of 18q are associated with tumor progression. Mod.Pathol. 17:1328-1337. 49. Murakami M, Hashimoto N, Takahashi Y, Hosokawa Y, Inazawa J, Mineura K (2003): A consistent region of deletion on 1p36 in meningiomas: identification and relation to malignant progression. Cancer Genet.Cytogenet. 140:99-106. 50. Iuchi T, Namba H, Iwadate Y, Shishikura T, Kageyama H, Nakamura Y, Ohira M, Yamaura A, Osato K, Sakiyama S, Nakagawara A (2002): Identification of the small interstitial deletion at chromosome band 1p34-p35 and its association with poor outcome in oligodendroglial tumors. Genes Chromosomes Cancer 35:170-175. 51. White PS, Thompson PM, Gotoh T, Okawa ER, Igarashi J, Kok M, Winter C, Gregory SG, Hogarty MD, Maris JM, Brodeur GM (2005): Definition and characterization of a region of 1p36.3 consistently deleted in neuroblastoma. Oncogene 24:2684-2694. 52. Oosterhuis JW, Looijenga LH, van Echten J, de Jong B (1997): Chromosomal constitution and developmental potential of human germ cell tumors and teratomas. Cancer Genet.Cytogenet. 95:96-102. 53. van Echten J, Timmer A, van der Veen AY, Molenaar WM, de Jong B (2002): Infantile and adult testicular germ cell tumors. a different pathogenesis? Cancer Genet.Cytogenet. 135:57- 62. 54. Donovan PJ, de Miguel MP (2003): Turning germ cells into stem cells. Curr.Opin.Genet.Dev. 13:463-471. 55. Kimura T, Suzuki A, Fujita Y, Yomogida K, Lomeli H, Asada N, Ikeuchi M, Nagy A, Mak TW, Nakano T (2003): Conditional loss of PTEN leads to testicular teratoma and enhances embryonic germ cell production. Development 130:1691-1700. 56. Youngren KK, Coveney D, Peng X, Bhattacharya C, Schmidt LS, Nickerson ML, Lamb BT, Deng JM, Behringer RR, Capel B, Rubin EM, Nadeau JH, Matin A (2005): The Ter mutation in the dead end gene causes germ cell loss and testicular germ cell tumours. Nature 435:360-364. 57. McLaren A (2003): Primordial germ cells in the mouse. Dev.Biol. 262:1-15. 58. The Chipping Forecast III (2005). Nat.Genet. 37: S1-S45. 59. Ishkanian AS, Malloff CA, Watson SK, DeLeeuw RJ, Chi B, Coe BP, Snijders A, Albertson DG, Pinkel D, Marra MA, Ling V, MacAulay C, Lam WL (2004): A tiling resolution DNA microarray with complete coverage of the human genome. Nat.Genet. 36:299-303. 60. de Vries B, Pfundt R, Leisink M, Koolen D, Vissers L, Janssen I, van Reijmersdal S, Nillissen W, Huys E, de Leeuw N, Smeets D, Sistermans E, Feuth T, van Ravenswaaij-Arts C, Geurts van Kessel A, Schoenmakers E, Brunner H, Veltman J. Diagnostic genome profiling in mental retardation. Am. J. Hum. Genet. (in press). 61. Syvanen AC (2005): Toward genome-wide SNP genotyping. Nat.Genet. 37 Suppl:S5-S10. 62. Barrett MT, Scheffer A, Ben Dor A, Sampas N, Lipson D, Kincaid R, Tsang P, Curry B, Baird K, Meltzer PS, Yakhini Z, Bruhn L, Laderman S (2004): Comparative genomic hybridization using oligonucleotide microarrays and total genomic DNA. Proc.Natl.Acad.Sci.U.S.A 101:17765-17770. 63. Dhami P, Coffey AJ, Abbs S, Vermeesch JR, Dumanski JP, Woodward KJ, Andrews RM, Langford C, Vetrie D (2005): Exon array CGH: detection of copy-number changes at the resolution of individual exons in the human genome. Am.J.Hum.Genet. 76:750-762. 64. McIntyre A, Summersgill B, Jafer O, Rodriguez S, Zafarana G, Oosterhuis JW, Gillis AJ, Looijenga L, Cooper C, Huddart R, Clark J, Shipley J (2004): Defining minimum genomic regions of imbalance involved in testicular germ cell tumors of adolescents and adults through genome wide microarray analysis of cDNA clones. Oncogene. 23:9142-7. 65. Mantripragada KK, Buckley PG, de Stahl TD, Dumanski JP (2004): Genomic microarrays in the spotlight. Trends Genet. 20:87-94.

120 General discussion and future prospects

66. Gorringe KL, Boussioutas A, Bowtell DD (2005): Novel regions of chromosomal amplification at 6p21, 5p13, and 12q14 in gastric cancer identified by array comparative genomic hybridization. Genes Chromosomes Cancer 42:247-259. 67. Tsuda H, Birrer MJ, Ito YM, Ohashi Y, Lin M, Lee C, Wong WH, Rao PH, Lau CC, Berkowitz RS, Wong KK, Mok SC (2004): Identification of DNA copy number changes in microdissected serous ovarian cancer tissue using a cDNA microarray platform. Cancer Genet.Cytogenet. 155:97-107. 68. Fujishita T, Doi Y, Sonoshita M, Hiai H, Oshima M, Huebner K, Croce CM, Taketo MM (2004): Development of spontaneous tumours and intestinal lesions in Fhit gene knockout mice. Br.J.Cancer 91:1571-1574. 69. Stevens LC (1970): Experimental production of testicular teratomas in mice of strains 129, A/He, and their F1 hybrids. J.Natl.Cancer Inst. 44:923-929. 70. Walt H, Oosterhuis JW, Stevens LC (1993): Experimental testicular germ cell tumorigenesis in mouse strains with and without spontaneous tumours differs from development of germ cell tumours of the adult human testis. Int.J.Androl 16:267-271.

121

Colour section

This section contains the colour images of the figures listed below.

Chapter 2 Figure 4 38 Chapter 3 Figure 5 52 Chapter 4 Figure 1 69

Chromosome 12 YAC contig

Figure 4 (chapter 2). A) Diagram showing: chromosome 12, a series of q13.1 markers and the corresponding CEPH YACs. B) FISH analysis with YAC 919-G-8 on a lymphocyte derived metaphase spread from the patient. Positive signals (green) on both der(12) and der(15) and on the normal chromosome 12 are indicated. Chromosomes are shown in red.

124 Colour section

A α-VSV α-PDI α-VSV + α-PDI

MESDC2-VSV

B α-VSV α-PDI α-VSV + α-PDI

SEME-VSV

Figure 5 (chapter 3). Subcellular localization of A) MESDC2 and B) SEME, using C-terminal VSV tags and SV40 promoters, after transient expression in HeLa cells. The proteins were visualized using an anti-VSV antibody (green; left panels). Endogenous PDI, an ER resident protein, was visualized with an anti-PDI antibody (red; middle panels). Overlays are shown in the right panels.

Chr. 12 Breakpoint BAC contig Chr. 8 BAC contig spanning YAC

RP11-815o9 RP11-79e1

RP11-91k1 RP11-609i23 der(12) RP11-79o20 q12- q13 893a3 RP11-510p12 RP11-79c21 der(8)

RP11-80d8 RP11-73b8 8 RP11-80d8 q23 RP11-350f4 RP11-79f7

AB C Figure 1 (chapter 4). FISH mapping of the 12q12 and 8q23 breakpoints. A) Diagram depicting chromosome 12, the 12q12 breakpoint-spanning CEPH YAC clone 893a3 and its corresponding BAC contig. In grey the breakpoint spanning clones (893a3 and RP11-350f4) are shown. B) Diagram depicting chromosome 8 and the selected BAC contig within 8q23, including the breakpoint-spanning BAC clone RP11-80d8 (grey). C) FISH analysis on a t(8;12)(q23;q12)-positive metaphase spread, using RP11-80d8 as a probe. Positive signals (green) are visible on the normal chromosome 8, der(8) and der(12). The centromere of chromosome 8 is marked in red and the chromosomes are counterstained with DAPI (blue).

125

Summary / Samenvatting

Summary

SUMMARY

Human germ cell tumours (GCTs) can be classified into five distinct types, based on differences in anatomical location, histology, clinical outcome, age and genotype (chapter 1). The first type, the type I GCTs primarily occur in neonates and infants under the age of five years and include teratomas and yolk sac tumours with distinct clinical and epidemiologic features. Until recently, the type I GCTs were poorly characterised at the genetic level. The objectives of this thesis were (i) to identify candidate genes involved in the development of type I GCTs and (ii) to identify chromosomal anomalies that may serve as diagnostic/prognostic markers for this tumour type. To these ends, we positionally cloned two constitutional chromosome translocations in patients with type I GCTs and breakpoints in the 12q11-q13 region, a region thought to harbour one or more genes involved in GCT development. In addition, we analysed a selected series of sporadic type I GCTs for the presence of recurrent chromosomal anomalies using a microarray-based comparative genomic hybridisation (array CGH) technology. In chapter 2 clinical and cytogenetic data are reported of a novel case with an infantile sacral teratoma and a constitutional t(12;15)(q13;q25). Loss of heterozygosity analysis revealed that both translocation-derived chromosomes were retained in the patient's tumour. In addition, we assembled a 12q13 genomic contig that spans the breakpoint. This genomic contig served as a starting point for the final elucidation of the translocation breakpoint-associated sequences. In chapter 3 the positional cloning of the 12q13 and 15q25 breakpoints is described in further detail. This effort resulted in the identification of two breakpoint-spanning genes: the SUMO/Sentrin-specific protease 1 gene SENP1 on chromosome 12 and the embryonic polarity-related mesoderm development gene MESDC2 on chromosome 15. In addition, we found that through the translocation both genes were disrupted and fused. The resulting reciprocal SENP1-MESDC2 (SEME) and MESDC2-SENP1 (MESE) fusion genes were transcribed in tumour-derived cells and their open reading frames encode aberrant proteins. The translocation-associated SEME protein contained most of the MESDC2 protein sequence except for the N-terminal part, which includes the signal peptide. This explains why, in contrast to wild-type MESDC2, SEME is no longer targeted to the endoplasmatic reticulum, thereby leading to a presumed loss of function as a chaperone for the WNT co-receptors LRP5 and/or LRP6. Wild-type SENP1 is a protease that can cleave SUMO-1 from its substrates and pre-process the precursor

129 Summary form of SUMO-1, a post-translational modifier that plays an important role in several cellular key processes. The translocation-associated MESE proteins, containing almost the complete protein sequence of SENP1, were found to exhibit desumoylation capacities similar to those observed for wild-type SENP1, as assessed by using a PML desumoylation assay. From these results we hypothesize that a disturbed function of MESDC2 (SEME), MESDC2 haplo-insufficiency and/or spatio- temporal disturbances in desumoylating activities during critical stages of embryonic development might have predisposed the patient to the development of its teratoma (chapters 3 and 7). In chapter 4 we reported the positional cloning of a second 12q11-q13- associated breakpoint in a patient with a type I GCT (sacral teratoma) and a constitutional t(8;12)(q23;q12). The 12q12 breakpoint appeared to be located ~50kb downstream of the DKFZP434G1415 gene and ~4 Mb upstream of the SENP1 gene. The 8q23 breakpoint was subsequently found to be located ~13 kb upstream of the EIF3S6 gene. Preliminary expression analyses revealed that the transcription levels of several positional candidate genes, including DKFZP434G1415, SENP1 and EIF3S6, were not significantly affected by the translocation. As yet, however, it cannot be excluded that these translocation breakpoints affect the expression of more distant genes and/or regulatory elements that may have played a role in the development of the sacral teratoma. In chapter 5 a novel and rapid procedure is presented for the mapping of genomic translocation breakpoints using flow sorted chromosomes in combination with array CGH. To test the feasibility of this approach, the t(12;15)(q13;q25)-positive cell line described in chapter 3 was used as a model. The derivative 12 chromosomes were flow sorted, labelled and hybridized to a genome-wide array containing 3.6 thousand well-characterized human genomic clones. By doing so, we found that in a single hybridization experiment the genomic locations of the breakpoints on both chromosomes 12 and 15 could be determined. The array CGH data obtained were in complete accordance with those reported in chapter 3. In order to gain further insight into the genetic constitution of type I GCTs, we applied array CGH to 24 GCTs derived from patients under the age of 5 years and 7 GCTs from patients above the age of 5 years (chapter 6). Three subgroups were defined that could be distinguished by their genomic profiles: 1) teratomas of patients at ages <5 years displaying normal profiles, except for some minor recurrent aberrations, 2) yolk sac tumours of patients at ages <5 years displaying recurrent losses of 1p35–pter and gains of 3p21–pter and of 20q13 and 3) GCTs of patients at

130 Summary ages > 5 years displaying recurrent gains of 12p, chromosomes 7 and 8. Independent of age and histology, gains of the 1q32–qter region and losses of the 6q24–qter and 18q21–qter regions were frequent events in these GCTs. From these results we conclude that array-based genomic profiles may be used as excellent diagnostic/prognostic identifiers for type I GCTs. In addition, these profiles point at genomic regions that may harbour genes relevant for the initiation and/or progression of these tumours (chapters 6 and 7). Next to the above-mentioned results and conclusions, some future prospects are discussed in chapter 7. We propose that the development of type I GCTs may be elucidated by using a multi-disciplinary approach, including the analysis of normal germ cells, neoplastic germ cells and tissue environmental interactions of both these cells.

131 Samenvatting

SAMENVATTING

Humane kiemceltumoren (KCTen) kunnen op basis van verschillen in anatomische locatie, histologie, klinisch verloop, leeftijd en genotype ingedeeld worden in 5 groepen (hoofdstuk 1). De eerste groep, de type I KCTen, zijn zeldzame tumoren die voornamelijk voorkomen bij pasgeborenen en kinderen tot vijf jaar. Histologisch gezien bestaat deze groep uit teratomen en dooierzaktumoren, met onderlinge klinische en epidemiologische verschillen. Teratomen zijn relatief goedaardige tumoren, waarin gedifferentieerde structuren als bot-, haar- en tandstructuren kunnen voorkomen. Dooierzaktumoren hebben een grotere neiging tot uitzaaien en vertonen minder differentiatie dan teratomen. Tot voor kort was nog niet veel bekend over de genetische samenstelling van de type I KCTen. De doelstellingen van dit proefschrift waren (i) het opsporen van kandidaat- genen die mogelijk een rol spelen bij het ontstaan en/of de ontwikkeling van type I KCTen en (ii) het identificeren van chromosomale afwijkingen die als diagnostische /prognostische parameters zouden kunnen dienen voor dit tumortype. Daartoe werden twee constitutionele chromosomale translocaties in kaart gebracht in patiënten met type I KCTen. De translocaties van beide patiënten lieten breuken in 12q11-q13 zien, een chromosomaal gebied waarvan verondersteld wordt dat deze een of meer genen bevat voor KCT ontwikkeling. Daarnaast werd een groep sporadische type I KCTen geselecteerd en met behulp van array CGH onderzocht op het voorkomen van frequente chromosomale afwijkingen. In hoofdstuk 2 werden de klinische en cytogenetische data van een kind met een sacraal teratoom en een constitutionele t(12;15)(q13;q25) beschreven. Beide translocatie-geassocieerde chromosomen bleken behouden te zijn in de tumor. Verder werden in dit hoofdstuk genomische fragmenten geïdentificeerd die het breukpunt in het 12q13 gebied omspannen. Deze fragmenten werden als startpunt gebruikt voor de definitieve opheldering van de DNA sequenties waarin de translocatie breukpunten zich bevinden. De positionele klonering van de 12q13 en 15q25 breukpunten werd in detail beschreven in hoofdstuk 3. Dit resulteerde in de identificatie van twee breukpunt-omspannende genen: het SUMO/Sentrin-specifieke protease 1 gen SENP1 op chromosoom 12 en het polariteit-gerelateerde mesoderm ontwikkelingsgen MESDC2 op chromosoom 15. Door de translocatie bleken beide genen gebroken en aan elkaar gefuseerd te zijn. De resulterende SENP1-MESDC2 (SEME) en MESDC2-SENP1 (MESE) fusiegenen kwamen tot expressie in de

132 Samenvatting tumorcellen en coderen voor afwijkende eiwitten. Het translocatie-geassocieerde SEME eiwit bevat een groot deel van de MESDC2 eiwitsequentie, behalve het N- terminale deel dat het signaalpeptide bevat. Dit verklaart waarom, in tegenstelling tot het wildtype MESDC2, SEME niet in het endoplasmatisch reticulum terecht komt. Deze afwijkende locatie zou mogelijk kunnen leiden tot een verlies van de functie van het SEME eiwit als chaperonne voor de WNT receptors LRP5 en/of LRP6. Wildtype SENP1 is een protease dat 1) SUMO-1 kan afklieven van zijn substraateiwitten en 2) de voorloper van SUMO-1 kan modificeren. SUMO-1 is een post-translationele modificeerder die een belangrijke rol speelt in verschillende essentiële cellulaire processen. De translocatie-geassocieerde MESE eiwitten bevatten bijna de gehele sequentie van SENP1. Uit een PML desumoyleringsanalyse bleek dat de MESE eiwitten vergelijkbare desumoylerings-capaciteiten bezitten als het wildtype SENP1 eiwit. Op basis van bovenstaande resultaten stellen wij voor dat tijdens kritische fases van de embryonale ontwikkeling een verstoorde functie van MESDC2 (SEME), MESDC2 haplo-insufficiëntie en/of spatio-temporele verstoringen in desumoylerings-activiteiten mogelijk hebben geleid tot de ontwikkeling van het teratoom bij de patiënt (hoofdstukken 3 en 7). In hoofdstuk 4 werd de positionele klonering beschreven van een tweede 12q11-q13-geassocieerde breuk in een patiënt met een type I KCT (sacraal teratoom) en een constitutionele t(8;12)(q23;q12). Het 12q12 breukpunt bleek ~50 Kb distaal van het DKFZP434G1415 gen en ~4 Mb proximaal van het SENP1 gen te liggen. Het 8q23 breukpunt kon vervolgens worden gelokaliseerd en bleek ~13 Kb proximaal van het EIF3S6 gen te liggen. Voorlopige expressie analyses lieten zien dat de transcriptie niveaus van verschillende positionele kandidaat-genen, waaronder DKFZP434G1415, SENP1 en EIF3S6, niet significant beïnvloed werden door de translocatie. Op dit moment kan niet worden uitgesloten dat de translocatie de expressie heeft beïnvloed van andere verderop gelegen genen en/of hun regulatoire elementen en, via deze weg, de ontwikkeling van het sacrale teratoom. In hoofdstuk 5 werd een nieuwe en snelle manier voorgesteld om genomische translocatie-breupunten te lokaliseren met behulp van gesorteerde chromosomen in combinatie met de array CGH techniek. Om de effectiviteit van deze methode te testen werd de t(12;15)(q13;q25)-positieve cellijn uit hoofdstuk 3 gebruikt als model. De derivaat 12 chromosomen werden gesorteerd, gelabeld en gehybridiseerd op een genoomwijde array met daarop 3.6 duizend goed gekarakteriseerde humane genomische clones. Op deze manier werd vastgesteld dat één hybridisatie experiment voldoende is om de breukpunten op zowel chromosoom 12 als

133 Samenvatting chromosoom 15 exact te bepalen. De array CGH data kwamen volledig overeen met de data zoals gerapporteerd in hoofdstuk 3. Om meer inzicht te krijgen in de genetische samenstelling van type I KCTen werden vierentwintig van deze tumoren uit patiënten onder de 5 jaar en zeven van deze tumoren uit patiënten boven de 5 jaar geanalyseerd met behulp van de array CGH techniek (hoofdstuk 6). Daarbij werden drie subgroepen gedefinieerd die onderscheiden konden worden op basis van hun genomische profielen: 1) teratomen van patiënten onder de 5 jaar met normale profielen, op een paar kleine terugkomende afwijkingen na, 2) dooierzaktumoren van patiënten onder de 5 jaar met de volgende frequente afwijkingen: verlies van 1p35-pter en winst van 3p21-pter en 20q13 en 3) tumoren van patiënten boven de 5 jaar met frequente winst van 12p en de chromosomen 7 en 8. Verder werd geconstateerd dat, onafhankelijk van leeftijd of histologie, winst van 1q32-qter en verlies van 6q24-qter en 18q21-qter frequente afwijkingen zijn in deze KCTen. Op basis van deze gegevens hebben we geconcludeerd dat genomische array CGH profielen geschikt zijn voor het identificeren van diagnostische en/of prognostische markers voor de type I KCTen en dat deze profielen duiden op genomische regio’s die genen kunnen bevatten die relevant zijn voor de initiatie en/of progressie van deze tumoren. Naast de bovengenoemde resultaten en conclusies werden in hoofdstuk 7 enkele toekomstperspectieven besproken, met als aanbeveling om de ontwikkeling van type I KCTen te bestuderen via een multidisciplinaire aanpak. Een dergelijke aanpak zou de analyse van normale kiemcellen, ontaarde kiemcellen en de interacties tussen normale cellen, tumorcellen en hun omliggende weefsels kunnen omvatten.

134

Dankwoord en curriculum vitae

Dankwoord

DANKWOORD

De laatste woorden staan op papier, het boekje is af! Dit is het moment om iedereen te bedanken die mij op een of andere manier heeft geholpen en er mede voor heeft gezorgd dat het boekje is zoals het nu is.

Ad Geurts van Kessel, mijn promotor, wil ik als eerste bedanken voor de zinvolle discussies, het razendsnelle nakijken van manuscripten en de vrijheid die je me gaf om me wetenschappelijk en persoonlijk te ontwikkelen. Het steuntje in de rug in het begin van mijn promotieonderzoek heeft er toe bijgedragen dat ik doorgezet heb en uiteindelijk beide breukpunten in kaart heb gebracht, ook hiervoor bedankt!

Han Brunner, bedankt dat ik onderzoek mocht doen op jouw afdeling. Al zal het tussen-de-middag schaatsen er niet meer bij zijn, graag hoor ik wanneer het ijs in Nijmegen weer dik genoeg is! Eric Schoenmakers (EricS) bedankt voor jouw inbreng, al je ideeën en praktische tips. Martien van Asseldonk, al ben je maar even mijn begeleider geweest, jou wil ik bedanken voor je hulp tijdens de beginfase van mijn promotieonderzoek: een goed begin is het halve werk.

Halverwege mijn promotieonderzoek kwam jij, Lilian Vreede, het kiemceltumor- ‘groepje’ uitbreiden. Je hebt bergen werk verzet, je was altijd super georganiseerd en een goede discussiepartner. Bovenal was het erg gezellig om met jou samen te werken, bedankt voor alles!

Het researchlab van Cytogenetica, waar ik al die tijd heb gewerkt wil ik bedanken voor alle praktische tips, kritische vragen, gezellige film en spel avondjes en de goede sfeer op het lab. In de loop van de tijd hebben er verschillende mensen gewerkt: Anita, AnkeD, AnkeR, Daniëlle, Diederik, Ellen, ErikH, Helma, Jan, José, Marc, Marga, Marieke, Marjan, Monique, Nuno, Roland, Ron. Van de FISH en Chips: Corine, Gerard, Huub, Irene (bedankt voor al je hulp met de arrays!), Judith, Lisenka, Rolph, Simon, Walter. Bert, jij bedankt voor het maken van de hybride cellijnen; in die tijd een must voor het ophelderen van chromosomale breukpunten.

Verder wil ik uiteraard de hele afdeling Antropogenetica bedanken. In de loop van de tijd hebben verschillende mensen mij geholpen met praktische zaken, computerproblemen, tips, en ideeën. Ook de gezellige uitjes, lunchdates en nog veel

139 Dankwoord meer hebben natuurlijk meegeholpen aan een goed verloop van mijn promotieonderzoek, allemaal bedankt!

Naast de (ex)collega’s van Antropogenetica zijn er verschillende mensen binnen en buiten Nijmegen waarmee ik met veel plezier heb samengewerkt en daarvoor wil bedanken.

Leendert Looijenga, jou wil ik bedanken voor het feit dat ik altijd even kon e-mailen of komen overleggen als ik met een specifieke kiemceltumor-vraag zat. Onze samenwerking heeft dan ook geresulteerd in een paar mooie artikelen. En omdat er nog genoeg vragen overblijven, doet het me goed dat ik in jouw groep verder kan werken aan de kiemceltumoren. Tevens wil ik jou en Wolter Oosterhuis bedanken voor onze discussie in het begin van mijn promotieonderzoek; uiteindelijk heeft dit hoofdstuk 6 opgeleverd. Van mijn huidige lab wil ik ook Hans Stoop, Ad Gillis en Ruud van Gurp bedanken voor kleuringen, microdissecties en DNA en RNA isolaties.

Dominik Schneider, vielen Dank für die Bereitstellung der Patientenproben und die angenehme Kooperation. Ich freue mich, dass wir auch in Zukunft weiter zusammen arbeiten werden. Switching from German to English, I would like to thank Dr. Yeh, who allowed me to do some ‘SENP1’ work in his lab. Also, I would like to thank Jinke Cheng for his help in the lab. Sometimes my ‘Dutch-English’ and your ‘Chinese-English’ did not completely match, but finally we always understood each other, so also thanks for your patience!

Christine Hulsbergen-van de Kaa, bedankt voor het uitzoeken van de verschillende Nijmeegse kiemceltumoren en Henry Dijkman voor je hulp bij de microdissecties. Jack Fransen wil ik bedanken voor zijn hulp met de confocale microscoop.

Voor het informeren over de t(12;15)(q13;q25) en het ter beschikking stellen van patiëntmateriaal bedank ik Lutgarde Govaerts, Cokkie Wouters, Rosalyn Slater and Jan van Hemel. For referring the t(8;12)(q23;q12) to us and for providing patient material, I thank Louise Strong and Harry Drabkin.

140 Dankwoord

Naast collega’s zijn er mijn vrienden en familie die ik bedank voor al jullie steun en vooral jullie begrip in de afgelopen periode wanneer ik weer eens aan mijn boekje moest werken.

Alma, Sieg en Cin, snel gaan we weer eens swingen! Alma (‘zusje’), jij gaf al het goede voorbeeld, ik ben dan ook erg blij met jou als paranimf. Ciska, heerlijk waren de hardlooprondjes samen, waarbij we al kletsend ongemerkt steeds langer liepen. Uiteraard ook de schaatsers van Lacustris bedankt voor de gezellige schaatslessen, de warme chocomel avonden, de trainingsweken in collalbo, feestjes en het IUT!

Lieve Pa, Ma, Rinse, Marga, Kim, Joris en Mirian, ik prijs mij super gelukkig met een familie zoals jullie! Het is altijd heerlijk om weer bij jullie in het Zuiden of waar ook ter wereld te zijn. Pa en ma, zonder jullie had dit boekje er zeker niet gelegen en al helemaal niet zo mooi. Mam, het is je weer gelukt; de kaft is prachtig, bedankt! Pa, ook figuur 1 mag er wezen, bedankt. Het is speciaal een broer als collega te hebben en publiceren met je broer heeft al helemaal iets extra’s. Joris bedankt voor je kritische blik en al je goede tips!

Erik, je hebt aardig wat geduld en begrip opgebracht de afgelopen tijd. Bedankt voor al je lieve steun. Nu hebben we echt tijd om samen leuke dingen te doen en zijn de weekenden weer voor ons!

141 Curriculum vitae

142 Curriculum vitae

CURRICULUM VITAE

Imke Veltman werd op 16 juni 1976 geboren in Hoensbroek. In 1994 behaalde zij haar VWO diploma aan het Sint Jans College te Hoensbroek. Aansluitend begon zij met de opleiding Bioprocestechnologie aan de Wageningen Universiteit, waar zij in 1999 afstudeerde in de moleculair-cellulaire richting. Tijdens deze studie deed zij moleculair-biologisch onderzoek bij de vakgroep Moleculaire Biologie van de Wageningen Universiteit (dr. J. Carette, dr. ir. J. Wellink). Vervolgens deed zij immunologisch onderzoek bij de afdeling ‘Molecular Microbiology and Immunology’ van de Oregon Health and Sciences University te Portland (OR, VS) (dr. G. Wiens, Prof. dr. M. Rittenberg) en bij de afdeling Immunologie van het Erasmus MC-Universitair Medisch Centrum Rotterdam (dr. ir. M. Nawijn, Prof. dr. H. Savenkoul). Na haar afstuderen in 1999 is zij in dienst getreden als assistent in opleiding (AIO) bij de afdeling Antropogenetica van de Radboud Universiteit te Nijmegen (Prof. dr. A. Geurts van Kessel). Het onderzoek dat zij hier heeft uitgevoerd heeft geresulteerd in dit proefschrift. Sinds januari 2005 is zij werkzaam als postdoc op de afdeling Pathologie van het Erasmus MC-Universitair Medisch Centrum Rotterdam.

Publications Wiens GD, Lekkerkerker A, Veltman I, Rittenberg MB. (2001): Mutation of a single conserved residue in VH complementarity-determining region 2 results in a severe Ig secretion defect. J Immunol 167: 2179-86.

Veltman I, van Asseldonk M, Schepens M, Stoop H, Looijenga L, Wouters C, Govaerts L, Suijkerbuijk R, Geurts van Kessel A. (2002): A novel case of infantile sacral teratoma and a constitutional t(12;15)(q13;q25) pat. Cancer Genet Cytogenet 136: 17-22.

Veltman IM, Schepens MT, Looijenga LH, Strong LC, Geurts van Kessel A. (2003): Germ cell tumours in neonates and infants: a distinct subgroup? APMIS 111: 152-60.

Veltman IM, Veltman JA, Arkesteijn G, Janssen IM, Vissers LE, de Jong PJ, Geurts van Kessel A, Schoenmakers EF. (2003): Chromosomal breakpoint mapping by arrayCGH using flow-sorted chromosomes. Biotechniques 35: 1066-70.

Veltman, I., Veltman J., Janssen I., Hulsbergen-van de Kaa C., Oosterhuis W., Schneider D., Stoop H., Gillis A., Zahn S., Looijenga L. et al. (2005): Identification of recurrent chromosomal aberrations in germ cell tumors of neonates and infants using genome-wide array-based comparative genomic hybridization. Genes Chromosomes. Cancer 43:367–376.

Veltman IM, Vreede LA, Cheng J, Looijenga LHJ, Janssen B, Schoenmakers EFPM, Yeh ETH, Geurts van Kessel A. (2005): Fusion of the SUMO/Sentrin-specific protease 1 gene SENP1 and the embryonic polarity-related mesoderm development gene MESDC2 in a patient with an infantile teratoma and a constitutional t(12;15)(q13;q25). Human Molecular Genetics 14: 1955-1963.

143