University of Groningen

On the elucidation of a tumour suppressor role of 3p in lung cancerElst, Arja ter

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2006

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Elst, A. T. (2006). On the elucidation of a tumour suppressor role of 3p in lung cancer. s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 22-06-2020

On the elucidation of a tumour suppressorrole of 3p in lung cancer

The study described in this thesis was supported by a grant (RUG 2000-2317) fromthe Dutch Cancer Society.Publication of this thesis was financially supported by:The Dutch Cancer Society and the University of Groningen.

, ter Elst, ArjaOn the elucidation of a tumour suppressor role of 3p in lung cancerISBN: 90-367-2435-X

Printed by Ponsen en Looijen b.v., Wageningen the Netherlands

RIJKSUNIVERSITEIT GRONINGEN

On the elucidation of a tumour suppressorrole of 3p in lung cancer

PROEFSCHRIFT

ter verkrijging van het doctoraat in deMedische Wetenschappen

aan de Rijksuniversiteit Groningenop gezag van de

Rector Magnificus, dr. F. Zwarts,in het openbaar te verdedigen op

woensdag 18 januari 2006om 14.45 uur

door

Arja ter Elst

geboren op 9 september 1974te Oldenzaal

Promotor Prof. dr. C.H.C.M. Buys

Beoordelingscommissie Prof. dr. S.L. NaylorProf. dr. S. ImrehProf. dr. H.J. Groen

ISBN 90-367-2434-1

Contents

Chapter 1 6Candidate lung tumour suppressor regions at the short arm ofchromosome 3. What evidence is there?

Chapter 2 46Transfection of a PAC contig covering the lung cancer criticalregion at 3p21.3 discloses the complexity of functional analysis ofhomozygous deletion regions

Chapter 3 80Micro-array expression analysis of a cell line stably transfectedwith 3p21.3 sequences indicates the occurrence of trans-regulationbetween originally neighbouring genes and regulatory DNAsequences

Chapter 4 96Analysis of a new homozygous deletion in the tumour suppressorregion at 3p12.3 reveals two novel intronic non-coding RNA genes

Chapter 5 128Summary and general discussion

Nederlandse samenvatting 136

Dankwoord 140

Chapter 1

Candidate lung tumour suppressor regions atthe short arm of chromosome 3. What

evidence is there?

Arja ter ElstCharles H.C.M. Buys

Department of Medical Genetics, University Medical Center Groningen, Groningen,The Netherlands

LUNG CANCER AND THE SHORT ARM OF CHROMOSOME 3

Lung cancer is the leading cause of cancer death among both men and women inthe western world. Consistent chromosomal aberrations occurring in lung tumoursmay provide a clue to the somatic genetic events leading to tumour development.Deletions of the short arm of chromosome 3 are a most common abnormality in lungcancer. They have been reported to occur in approximately 75% of non small celllung cancer (NSCLC) tumours and in up to 100% of small cell lung cancer (SCLC)tumours (reviewed in Kok et al., 1997; Zabarovsky et al., 2002). Such deletions havealso been found in the histological normal tissue surrounding tumours and inpreneoplastic and preinvasive lesions (reveiwed in Kok et al., 1997). In addition, 3pdeletions have been found in histologically normal tissue of about 50% of smokersand former smokers, not in control individuals (Wistuba et al., 1997). This suggeststhat losses at the short arm of chromosome 3 represent an early chromosomalchange in the development of lung tumours. Moreover, introduction by microcell-mediated chromosome transfer of a normal human chromosome 3 into a lungadenocarcinoma cell line, A549, resulted in a suppression of growth of thetransfected cell line in nude mice compared to the growth of the parental cell line(Satoh et al., 1993). The region of interest could be confined by the discovery ofoverlapping homozygous deletions in three different SCLC cell lines, NCI-H740,GLC20 and NCI-H1450, with a smallest overlap of 370 kb (Daly et al., 1993; Kok etal., 1994; Roche et al., 1996). When a cosmid contig was constructed for the 370 kbsmallest region of overlap, a search for genes by cDNA library screening and CpGisland identification revealed that the region was gene-rich (Wei et al., 1996).Sequencing of the whole region led by experimental and informatics methods to theidentification of 19 genes (Lerman and Minna, 2000). These genes will be discussedlater in this chapter.

Chapter 1

8

LOSS OF SEGMENTS OF THE SHORT ARM OF CHROMOSOME 3IN A VARIETY OF TUMOUR TYPES

Hemizygous and homozygous deletions of the short arm of chromosome 3 are alsofound in a multitude of other epithelial cancers, including renal cell carcinoma, headand neck carcinoma, nasopharyngeal carcinoma, malignant mesothelioma, anduterine cervix carcinoma (Kok et al., 1997). For lung cancer, three non-overlappingdeletion regions were described 3p25, 3p21.3 and 3p12-p14 (Hibi et al., 1992). Thenumber of distinct regions on the short arm of chromosome 3 that have beenimplicated in the development of tumours, has been expanded to seven in recentyears. These include the regions indicated in the following as 3p22 AP20, 3p21.3CER1 and CER2, 3p21 D3F15S2 region, 3p21.3 LUCA, 3p14 FHIT and 3p12ROBO1 (Fig. 1).

The 3p22 AP20 regionIn small cell lung cancer (SCLC) homozygous deletions at the AP20 region havebeen reported to occur in the large majority of cases (Senchenko et al., 2004). Thesmallest region of overlap of homozygous deletions in this region was found in abreast cancer cell line and in a renal cancer cell line (RCC) and mapped between themarkers D3S3623 and D3S1298 (Fig.1). This region contains four genes: APRG1,coding for AP20 region protein; ITGA9, coding for integrin alpha 9; CTDSPL,encoding CTD (carboxy-terminal domain, RNA polymerase II, polypeptide A) smallphosphatase-like and VILL, coding for villin-like. Altered expression of CTDSPL wasdetected in a panel of epithelial cancer biopsies and cell lines (Kashuba et al., 2004).In addition, clones from a CTDSPL-transfected RCC cell line and a CTDSPL-transfected SCLC cell line showed an inhibition of tumour growth in nude mice incomparison to the growth of the non-transfected parental cell lines (Kashuba et al.,2004).

The 3p21.3 CER1 and CER2 regionsSCID tumours caused by cell lines carrying a human chromosome 3 on a mousefibrosarcoma background show non-random elimination of 3p21.3 sequences thatare supposed to contain tumour suppressor genes (Kholodnyuk et al., 1997). Szeleset al. (1997) defined the genetic length of the eliminated region designated CER1,common elimination region 1, as 1.6 cM flanked by D3S1029 and D3S643. The

Introduction

9

physical size of CER1 was further restricted to approx. 1 Mb after the region wascovered with a PAC contig (Yang et al., 1999). Seventeen genes are located inCER1 (Fig. 1), whose telomeric border is at D3S3582, while its centromeric border isin the first intron of LRRC2 (Kiss et al., 2002). One of these 17 genes, LTF, codingfor lactoferrin, was tested for tumourigenicity by injecting SCID mice with clones froma mouse fibrosarcoma cell line transfected with a PAC containing the gene. The LTF

Figure 1. A map of the short arm of chromosome 3 giving the position of the genes and markers locatedin the region corresponding with the deletion.

Chapter 1

10

promoter sequences appeared to become hypermethylated and expression of LTFwas lost in derived tumours (Yang et al., 2003). Expression of LTF was also foundabsent in 18 of 37 SCLC cell lines, 11 of 43 NSCLC cell lines and 7 of 13 primaryNSCLC tumours (Iijima et al., 2005). When a human nonpapillary renal cellcarcinoma cell line was used as recipient of the human chromosome 3, CER1appeared to become also eliminated on a human background. The elimination regionis flanked by D3S3582 and CCR5, i.e. about 250 kb shorter than CER1 on a mousebackground (Kholodnyuk et al., 2002).

In addition, a second common elimination region was found, CER2, locatedbetween marker RH94338 and marker SHGC154057. In CER2 (Fig.1), seven geneshave been identified, including two genes coding for chemokine receptors(Kholodnyuk et al., 2002). It may be noted that in CER1 seven of the 17 genes areencoding chemokine receptors.

The 3p21.3 D3F15S2 regionUBE1L coding for ubiquitin-activating enzyme E1-like, was isolated from a regionconsidered by Kok et al. (1987) as most consistently reduced to hemizygosity inSCLC, the D3F15S2 locus. The gene was picked up by hybridisation of a lung cDNAlibrary with DNA from a human 3p21 fragment in a Chinese hamster-human hybrid(Carritt et al., 1992). UBE1L spans about 8.5 of genomic DNA. It has 26 exons andan open reading frame of 1009 nucleotides. The gene encodes a member of the E1ubiquitin-activating enzyme family, which are involved in the modification of proteinswith ubiquitin in order to target abnormal or short-lived proteins for degradation. ThemRNA concentration of UBE1L in SCLC cell lines was found to be 0.5%-3% of thatin normal lung tissue. No mutations or rearrangements of the remaining allele were,however, found in SCLC (Kok et al., 1993). UBE1L expression was found to beenhanced after treatment of an acute promyelocytic leukemia cell line with all-trans-retinoic acid (RA), which induces remission in acute promyelocytic leukemia. UBE1Lmight mediate degradation of the oncogenic PML/RA receptor of the t(15:17)rearrangement found in acute promyelocytic leukemia (Kitareewan et al., 2002).Treatment of immortalised human bronchial cells with RA also resulted in a higherexpression of UBE1L and cotransfection of UBE1L with CCND1 in these cellsresulted in a repression of CCND1 in a UBE1L dosage dependent manner (Pitha-Rowe et al., 2004). Since overexpression of Cyclin D1 is frequently observed intumours and may contribute to tumourigenesis, this is an interesting observation.

Introduction

11

The 3p21.3 LUCA regionAs already discussed this region contains 19 genes (Fig.1). Further homozygousdeletions of 3p21.3 were found by FISH in three uncultured lung squamous cellcarcinoma tumours, at marker D3S2968 (Todd et al., 1997) and by real-time PCR infour uncultured RCC tumours and four uncultured breast cancer tumours at markerD3S3874 (Senchenko et al., 2004). A mouse fibrosarcoma cell line, A9, containing 2Mb of the human chromosome 3 which included the 370-kb critical region, showedreduction of tumour growth in a tumourigenicity test (Killary et al., 1992).

Figure 2. A map of the LUCA region giving the positions of the genes located in the region. The map atthe right is a magnification of part of the map.

Chapter 1

12

A similar effect was observed for A9 cells containing a P1-phage with 80 kb from the370-kb critical region, but not for A9 cells containing P1 phages with the flankingDNA sequences (Todd et al., 1996).

Loss of 3p14.2, FHIT in lung cancerHomozygous deletions of 3p14.2 were frequently detected in several cancer celllines including renal cancer cell lines and lung cancer cell lines (Lisitsyn et al., 1995).Using exon amplification from cosmids covering the deleted region allowedidentification of the human FHIT gene, a member of the histidine triad gene family(Ohta et al., 1996). In approximately 50% of esophageal, stomach, and coloncarcinomas an abberrant transcript of FHIT was detected. To determine the role ofFHIT in lung cancer, cDNA of 59 primary lung tumours was sequenced. Aberranttranscripts were found in 80% of SCLC and 40% of NSCLC. In addition, loss of FHITalleles was found in 76% of lung tumours (Sozzi et al., 1996). Since then a numberof papers have described aberrant transcripts or loss of FHIT alleles in SCLC andNSCLC (Sozzi et al., 1998; Tseng et al., 1999; Zochbauer-Muller et al., 2000; Ho etal., 2002). FHIT promoter hypermethylation was detected in 36% of 120 primaryNSCLC tumours (Tomizawa et al., 2004). Presence of abnormal transcripts, in termsof frequency and variety, is, however, not cancer-specific, since abnormal transcriptshave also been found in normal lung tissue with the same frequency and variety(Tokuchi et al., 1999). In addition, the normal transcript of FHIT appeared to occur inrenal cell cancer- and lung cancer-derived cell lines, including a cell line with ahomozygous deletion in the FRA3B region (van den Berg et al., 1997). Nevertheless,a strong anti-tumourigenic effect of FHIT-transfected NSCLC cell lines has beenreported after injection into nude mice (Roz et al., 2002). In addition, a significantsuppression of both primary and metastatic lung tumour growth in nude mice wasobserved when treating the tumours with a DOTAP-FHIT complex (Ramesh et al.,2001).

Loss of 3p12, ROBO1A sub-microscopic deletion was found in the SCLC cell line U2020 (Rabbitts et al.,1990). This region was further characterised by (Latif et al., 1992) and was estimatedto be in the range of 4-7 MB. By cDNA library screening a gene called DUTT1(Deleted in U Twenty Twenty) was isolated from this homozygous deletion.(Sundaresan et al., 1998). The same gene was independently isolated, ROBO1, as

Introduction

13

the human homologue of the Drosophila gene Roundabout (Kidd et al., 1998).Cytogenetic analysis of the homozygous deletion revealed that genetic loss hadoccurred by complex rearrangements rather than a simple interstitial deletion ofchromosome 3 (Heppell-Parton et al., 1999). Mutations and promoterhypermethylation of ROBO1 appeared to be rare in NSCLC and SCLC primarytumours (Dallol et al., 2002). Mice homozygous for a deleted form of ROBO1,frequently die at birth due to respiratory failure because of delayed lung maturation(Xian et al., 2001), whereas heterozygous mice develop lymphomas and carcinomasin their second year of life with a 3-fold increase in incidence compared with controls(Xian et al., 2004). In addition to the large homozygous deletion found in the SCLCcell line U2020, a much smaller 3p12 homozygous deletion was found in the SCLCcell line GLC20 (Angeloni et al., this thesis), as a second deletion next to ahomozygous deletion in the LUCA region. By means of fiber-fluorescent in situhybridisation experiments by P1-clones from the region, the length of deletion wasfound to be approximately 110 to 130 kb. This 3p12 homozygously deleted region ofGLC20 affects exon 2 of ROBO1, causing the loss of amino acids 19-128 from theencoded protein. Two novel transcripts located in the second intron of ROBO1 werediscovered in this homozygous deletion. Based on their characteristics bothtranscripts do not seem to encode proteins, but represent non-coding RNAs.Possible miRNA target sites were discovered in the sequence of these RNAs. Inaddition, a computational identified miRNA (cand893 HS3 78768573-78768661 R) islocated in the 3p12 homozygously deleted region of GLC20. Target genes for thismiRNA have not yet been discovered.

Multiple regions on the short arm of chromosome 3 have been found deleted. Thismight indicate that deletions or mutations of a combination of genes from several ofthese regions are responsible for the development of lung cancer. In this review wefocus on genes or regulatory sequences in the 3p21.3 critical region (LUCA region).

GENES IN THE 3P21.3 CRITICAL REGION.

RBM6RBM6 codes for RNA binding motif 6. Its cloning, structure, expression in normaltissue and function have been reviewed by Lerman and Minna et al., (2000) andZabarovsky et al., (2002). More recently it was found that exon 5, which is excluded

Chapter 1

14

in a shorter transcript of RBM6, contains a RNP-1 RNA binding motif which might beimportant for tumour suppression (Sutherland et al., 2005). This doesnot seem inagreement with the results of Timmer et al. (1999), who found that in SCLC cell linesthe transcript with exon 5 was higher expressed than the shorter transcript, whereasin normal tissue the expression of both the shorter and the longer transcripts hadsimilar levels. That would attribute a tumour suppressor function to the shorterproduct lacking the RNP-1 RNA binding motif. Since RBM6 is well expressed in lungcancer cell lines and no mutations have been found RMB6 is not a likely tumoursuppressor candidate.

RBM5RBM5, codes for RNA binding motif 5. Its cloning, structure, expression in normaltissue and function have been reviewed by Lerman and Minna et al. (2000) andZabarovsky et al. (2002). Although RBM5 appeared to be well expressed in lungcancer cell lines, more recently Oh et al. (2002) reported a lower expression ofRMB5 in primary NSCLC tumours than in adjacent normal tissue in 9 out of 11 (82%)of the analysed cases. RBM5 has an alternative splice variant, which lacks exon 6and results, like the alternative splice variant of RMB6, in a truncated protein. Thisshorter variant appeared to be widely expressed at a low level in normal tissue, but isexpressed at increased levels in T-leukaemic cell lines (Mourtada-Maarabouni et al.,2003). Overexpression of this variant resulted in inhibition of CD95-mediatedapoptosis, whereas overexpression of the full-length form suppressed cellproliferation by inducing apoptosis and by extending the G1 phase of the cell cycle.RBM5 has been tested for tumour suppressor activity in several functional assays.Overexpression of the gene in two breast cancer cell lines (MCF-7 and HBL-100)and one human fibrosarcoma cell line (HT1080) resulted in suppression of colonyformation (Edamatsu et al., 2000; Oh et al., 2002). Overexpression of RBM5 in amouse fibrosarcoma cell line (A9), resulted in a significant suppression oftumourigenicity of the cells when injected into nude mice (Oh et al., 2002).Overexpression of RBM5 in Jurkat T lymphoblastic leukaemia cells rendered thecells more susceptible to the death-inducing ligand TRIAL in TNF-α and FAS-mediated apoptosis (Rintala-Maki and Sutherland, 2004). Also for the breast cancercell line MCF-7, a positive correlation was found for RBM5 overexpression and TNF-α susceptibility (Rintala-Maki et al., 2004). So far, RBM5 has never been tested for atumour suppressor function in lung cancer-derived cell lines.

Introduction

15

SEMA3FSEMA3F codes for Semaphorin 3F. Its cloning, structure, expression in normaltissue and function have been reviewed by Lerman and Minna et al. (2000) andZabarovsky et al. (2002). Apart from their role in the guidance of nerve growth conemigration, semaphorins may play a role in the cardiovascular system. It has beensuggested that SEMA3F can inhibit angiogenesis in vitro by competition withVEGF165 for binding to the neuropilin-1 receptor (Miao et al., 1999; Kessler et al.,2004). It has also been suggested that SEMA3F and VEGF have opposing effects onthe motility of primary tumour cells. By exposing two breast cancer cell lines tocultured supernatant of COS-7 cells transfected with SEMA3F, it was found thatSEMA3F inhibits cell spreading and membrane ruffling, activities associated withincreased cell migration and metastasis, whereas VEGF promotes these changes(Nasarre et al., 2003). For the same two breast cancer cell lines, it wasdemonstrated that SEMA3F inhibited migration of one of them, while in the otherbreast cancer cell line intercellular contacts were disrupted, and delocalisation of E-cadherin and beta-catenin took place (Nasarre et al., 2005). A SEMA3F transcriptwas detected in about 80% of the tested lung cancer cell lines. The typicalmembrane staining of the SEMA3F protein got, however, lost in 50% of the testedhigh-grade neuroendocrine tumours (see by Zabarovsky et al., (2002)). Morerecently, Lanteujoul et al., (2003) investigated the immunostaining of SEMA3F andVEGF in 50 preneoplastic lesions and 112 lung tumours. Immunostaining ofSEMA3F got lost in 88% of preneoplastic lesions analysed. The degree of lossappeared to correlate with tumour stage, whereas VEGF staining increased withtumour grade. Since both SEMA3F and VEGF are deregulated in lung cancer, acommon pathway might be involved in both breast cancer and lung cancer. Kusy etal. (2005b) characterised the SEMA3F promoter and tested 22 cancer cell lines,including 8 lung cancer cell lines, for promoter hypermethylation. This was found inall the lung cancer cell lines. In 15 primary lung tumours mostly partial methylationwas observed. Treatment with demethylating agents had no effect on SEMA3Fexpression. Treatment with a histone deacetylate inhibitor stimulated SEMA3Fexpression.

Contradicting results have been found with respect to a role of SEMA3F intumour suppression. Mouse fibrosarcoma (A9) cells and ovarian adenocarcinomacells transfected with SEMA3F showed a reduced tumour growth upon injection intonude mice, as well as an arrest in G2 in response to drugs such as taxol and

Chapter 1

16

adriamycin in vitro (Xiang et al., 2002). In addition, overexpression of SEMA3F inhighly metastatic melanoma cells (A375SM) blocked metastases after injection intonude mice (Bielenberg et al., 2004). In contrast, however, SEMA3F transfected intoNCI-H1299 or GLC45 lung cancer cell lines did not affect cell growth in vitro ortumour growth in vivo (Tomizawa et al., 2001; Xiang et al., 2002). When transfectedin a squamous carcinoma cell line (NCI-H157), SEMA3F inhibited tumour growthafter the transfected cells were transferred into the of nude rats. When SEMA3F wastransfected into a large cell carcinoma cell line (NCI-H460), the cells readily formedtumours in the trachea of in nude rats (Kusy et al., 2005a). The ability of SEMA3F toreduce tumour growth seems thus to be cell line-dependent. Altogether a role ofSEMA3F in lung cancer development is not yet clear.

GNAT1GNAT1 codes for a guanine nucleotide binding protein (G protein) alpha transducingactivity polypeptide. Its cloning, structure and what was known about its expressionand function has been reviewed by Lerman and Minna et al. (2000). Since GNAT1 isnot expressed in normal lung tissue (not in lung cancer cell lines) and no mutationshave been found, GNAT1 is not a likely tumour suppressor candidate.

SLC38A3SLC38A3 encodes solute carrier family 38, member 3. Its cloning, structure and whatwas known about its expression and function has been reviewed by Lerman andMinna et al. (2000). Since SLC38A3 is well expressed in normal lung tissue and lungcancer cell lines and no mutations have been found, SLC38A3 is not a likely tumoursuppressor candidate.

GNAI2GNAI2 encodes guanine nucleotide binding protein (G protein) alpha inhibitingactivity polypeptide 2. Its cloning, structure and what was known about its expressionin normal tissue and function has been reviewed by Lerman and Minna et al. (2000)and Zabarovsky et al. (2002). Two GNAI2 activating mutations have been found inhuman endocrine tumours (Lyons et al., 1990). Transfection of the activated form ofGNAI2, also known as Gip2, into rat fibroblast (Rat-1A) cells induced oncogenictransformation, but this was not the case in mouse embryonic fibroblast (NIH3T3)cells (Pace et al., 1991). In contrast, Hermouet et al., (1991) found that transfection

Introduction

17

of an activated mutant of GNAI2 in NIH 3T3 cells resulted in a reduced doublingtime, a decreased serum requirement and a somewhat anchorage-independentgrowth proliferation. Transfection of an inactivating mutant of GNAI2 slowed downthe growth of NIH3T3 cells. Nude mice injected with murine melanoma (K-1735) cellsexpressing this inactivating mutant showed a delayed tumour formation (Hermouet etal., 1996). Mice deficient for GNAI2, however, developed adenocarcinoma of thecolon (Rudolph et al., 1995). Since GNAI2 is well expressed in lung cancer cell linesand no mutations have been found in lung cancer cell lines, it is an unlikely tumoursuppressor candidate for lung cancer. The activating mutations of GNAI2 found inendocrine tumours and the induced growth after transfection of this constitutivelyactive GNAI2 rather classify it as a possible proto-oncogene.

SEMA3BSEMA3B codes for semaphorin 3B. Its cloning, structure and what was known aboutits expression and function has been reviewed by Lerman and Minna et al. (2000)and Zabarovsky et al. (2002). SEMA3B is the second semaphorin coding gene in the3p21.3 critical region. Like for SEMA3F, its receptors include neuropilin 1 and 2(Npn-1 and Npn-2) and plexins (Tamagnone et al., 1999; Takahashi et al., 1999;Rohm et al., 2000). Apart from its role in axon guidance, SEMA3B has beensuggested to be a direct transcriptional target of p53. SEMA3B might be involved inp53-dependent suppression of cell growth (Ochi et al., 2002).

The observed loss of expression in lung cancer cell lines and primarytumours as compared to normal lung tissue might be caused by hypermethylation ofthe SEMA3B promoter. Although in some reports a correlation was found betweenloss of expression, loss of heterozygosity in 3p21.3 and promoter hypermethylationof SEMA3B (Tomizawa et al., 2001; Kuroki et al., 2003), this could not be confirmedin a recent study of 138 primary non small cell lung tumours (Ito et al., 2005). Inaddition, in a study of 64 bronchial aspirates promoter hypermethylation of SEMA3Bwas found in bronchial aspirates of both tumour cases and non-tumour cases (Groteet al., 2005), suggesting non-tumour specific promoter hypermethylation of SEMA3B.A tumour suppressor function of SEMA3B has been claimed based upon resultsobtained in several cell lines. Wildtype SEMA3B transfected into the lung cancer cellline NCI-H1299 induced growth inhibition and apoptosis in vitro (Tomizawa et al.,2001; Castro-Rivera et al., 2004). This pro-apoptotic effect was significantlydecreased by vascular endothelial growth factor (VEGF165), competing with SEMA3B

Chapter 1

18

for their common receptor neuropilin (Castro-Rivera et al., 2004). In addition,overexpression of SEMA3B in the ovarian adenocarcinoma cell line HEY exhibited adiminished tumourigenicity upon injection into BALB/c nu/nu mice (Tse et al., 2002).Homozygous SEMA3B null mice of 12 months of age did not show any pathologicalabnormalities after histological analysis of several tissues including lung (van derWeyden et al., 2005). This might suggest a level of redundancy between class 3semaphorins. In lung cancer, the common heterozygous loss of a large part of 3p willeliminate one copy of both SEMA3F and SEMA3B. Occurrence of mutations in thesecond copy is, however, rare for both genes.

IFRD2IFRD2 codes for an interferon-related developmental regulator 2. Its cloning,structure and what was known about its expression and function has been reviewedby Lerman and Minna et al. (2000). The relatively low expression of IFRD2 in normallung tissue as well as in lung cancer cell lines and the lack of mutations make it anunlikely tumour suppressor candidate for lung cancer.

NAT6NAT6 codes for N-acetyltransferase 6. Its cloning, structure and what was knownabout its expression and function has been reviewed by Lerman and Minna et al.(2000). More recently, a highly complex organisation structure was found in humansand in mice for NAT6 together with HYAL1, HYAL3 and IFRD2. In mice, the secondexon of nat6 is located within the first intron of hyal3. Ifrd2 was found immediately 3’to hyal3. This dense organisation was accompanied by significant levels ofcotranscription of hyal1, nat6 and hyal3, leading to bicistronic mRNA products forhyal1 with nat6 and hyal1 with hyal3 (Shuttleworth et al., 2002). NAT6 expressionhas never been determined in lung cancer. But only four missense mutations weredetected in 78 lung cancer cell lines. A tumour suppression function for NAT6 in lungcancer is therefore not very likely.

HyaluronidasesWithin the 3p21.3 critical region three different hyaluronidase genes HYAL1, HYAL2and HYAL3, have been identified. Hyaluronidases are a family of enzymes crucial forthe spread of bacterial infections, for toxins present in various venoms and possiblyfor cancer progression (Lokeshwar et al., 2002). Hyaluronidases degrade hyaluronic

Introduction

19

acid (HA), which is present in body fluids, tissues and the extracellular matrix. It issynthesised by cells of mesenchymal origin in response to different stimuli (Li et al.,2000). HA keeps tissues hydrated and maintains osmotic balance and cartilageintegrity (Tammi et al., 2002). It can also actively regulate cell adhesion, migration,and proliferation by interacting with specific cell surface receptors (Turley et al.,2002). Hyaluronan chains have size-specific biological activities. High molecularweight HA has anti-inflammatory, immunosuppressive (McBride and Bard, 1979;Delmage et al., 1986) and anti-angiogenic (Feinberg and Beebe, 1983) properties.HYAL2 digestion of hyaluronan results in lower molecular weight hyaluronan of 20kDa which is a potent stimulator of inflammatory cytokines (Noble, 2002).Angiogenesis is also stimulated by these smaller fragments (Rooney et al., 1995;Trochon et al., 1997; Slevin et al., 1998). Dendritic cells are activated by 6-20 kDasized hyaluronan oligomers (Termeer et al., 2000; Termeer et al., 2003) HYAL1digestion of hyaluronan results in tetrasaccharides, which have an anti-apoptoticeffect and suppress cell death in cell cultures undergoing hyperthermia or serumstarvation (Xu et al., 2002). In tumour development, hyaluronan was found to expandupon hydration and opens up spaces for tumour cell migration (Lokeshwar et al.,2001). In concordance, hyaluronan was found upregulated in several cancersincluding breast cancer, colorectal cancer and lung cancer (Auvinen et al., 1997;Auvinen et al., 2000; Pirinen et al., 2001). Upregulation has been shown to correlatewith grade and metastatic potential (Zhang et al., 1995; Jojovic et al., 2002). Normalconnective tissue cells immediately adjacent to an invasive tumour could also beresponsible for the production of tumour-associated hyaluronan (Knudson et al.,1989). For hyaluronidase expression contradictory data has been described. In moststudies hyaluronidase expression was found elevated in tumour tissue (Wilkinson etal., 1996; Patel et al., 2002; Franzmann et al., 2003). In one study, however,hyaluronidase expression in tumour tissue was found to be reduced (Fiszer Szafarzand Szafarz, 1973).

HYAL3HYAL3 codes for hyaluronidase 3. Its cloning, structure and what was known aboutits expression and function has been reviewed by Lerman and Minna et al. (2000).For HYAL3 no precise function is, however, known yet. The lack of expression ofHYAL3 in lung tissue as well as in lung cancer cell lines make HYAL3 a less likelytumour suppressor candidate for lung cancer.

Chapter 1

20

HYAL1HYAL1 was found using a partial sequence derived from a protein purified fromserum to screen an EST database (Frost et al., 1997). The structure of HYAL1 andwhat was known about its expression and function has been reviewed by Lermanand Minna et al. (2000) and Zabarovsky et al. (2002). More recently, murine Hyal1and Hyal2 overexpression was found to enhance the sensitivity of murine L929fibroblasts to tumour necrosis factor mediated cell death (Chang, 2002), implicatingHYAL1 and HYAL2 in the TNF pathway. Although Lerman and Minna et al. (2000)found no expression of HYAL1 in 18 out of 20 lung cancer cell lines tested, including10 SCLC cell lines, Junker et al., (2003) found HYAL1 expression in all twenty SCLCcell lines tested. Hyaluronidase activity could, however, not be detected in these celllines, possibly due to aberrant splicing of the pre-mRNA, described earlier for headand neck squamous cell carcinomas (see review by Zabarovsky et al. (2002)).HYAL1 seems to have both a tumour-suppressive and a tumour-inducing effect.There are several reports on a tumour-suppressive effect in functional tests. Whentransfected into a subclone of the colon cancer cell line DHD-K12, HYAL1 showed asignificant growth reduction upon injection in BD-IX rats in comparison with the mocktransfected cell line (Jacobson et al., 2002). In addition, hyaluronidase administrationto SCID mice bearing human breast tumour xenografts, caused eradication ofhyaluronan and rapid reduction in tumour size (Shuster et al., 2002). Earlier,hyaluronidase was found to act as an anticarcinogenic agent in BALB/C mice(Pawlowski et al., 1979). Other authors have reported, however, a tumour-inducingeffect of HYAL1. Expression of hyaluronidase by tumour cells was found to induceangiogenesis in vivo (Liu et al., 1996). More recently, overexpression of HYAL1 wasshown to enhance the metastatic behaviour of a prostate cell line (Patel et al., 2002).HYAL1 secretion has been suggested to correlate with prostate cancer progression(Lokeshwar et al., 2001). Blocking of HYAL1 in a bladder cancer cell line resulted ina reduced tumour growth in nude mice and a reduced infiltration in skeletal musclewas observed compared with the bladder cancer cells transfected with vector only(Lokeshwar et al., 2005b). No effect was seen in NSCLC xenografts afterintratumoural injection of recombinant adenoviral vectors containing HYAL1 orHYAL2 (Ji et al., 2002). This has so far the only experiment in which HYAL1 hasbeen tested for a role in lung cancer. The contradicting findings about the role ofHYAL1 in cancer as resulting from these experiments were recently explained in astudy of Lokeshwar et al. (2005a) who describe that HYAL1 can function as either a

Introduction

21

tumour suppressor or a tumour promoter depending on the concentration of the geneproduct.

HYAL2HYAL2 codes for hyaluronidase 2. Its cloning, structure and what was known aboutits expression has been reviewed by Lerman and Minna et al. (2000). Hyaluronidase2 is a glucosylphosphatidylinositol (GPI) anchored cell-surface protein (Rai et al.,2001), with lysosomal hyaluronidase activity, degrading high molecular masshyaluronan (8000 kDa) to products of about 20 kDa (Lepperdinger et al., 1998) (seereview by Zabarovsky et al. (2002)). More recently, (2005) it was found that digestionof hyaluronan with increasing concentrations of soluble HYAL2 at the optimal pH of5.5 resulted in increasing degradation of the 20 kDa intermediate. In SV40-immortalised human bronchial cells HYAL2 is associated with the MST1 receptor,thereby negatively regulating MST1R signalling (Fig. 3). Hyal2 was shown to be avirus entry receptor of Jaagsiekte sheep retrovirus (JSRV), a virus which causescancer of the lower airways and alveoli in sheep (Rai et al., 2001). It mediates theentry of the highly oncogenic retrovirus JSRV into the cell (see review byZabarovsky et al. (2002)). Hyal2 has also been shown to be the receptor for enzooticnasal tumour virus, which induces nasal epithelial cancer in sheep (Dirks et al.,2002). The retrovirus binds to HYAL2, which is consequently degraded. The releaseof HYAL3 from MSTR1 activates the MSTR1 pathway and leads to constitutiveactivation of AKT and MAPK proteins and thereby to proliferation and promotion ofcell survival (Danilkovitch-Miagkova et al., 2003).

As for HYAL1, contradicting findings have also been published for a role ofHYAL2 in tumour suppression. HYAL2 is well expressed in all lung cancer cell linestested (Lerman and Minna, 2000). In non-Hodgkin lymphomas, however, the lowestlevel of HYAL2 expression was found in the most aggressive type of lymphoma,diffuse large cell lymphomas (Bertrand et al., 2005). Since the gene is wellexpressed in lung cancer cell lines and no mutations have been detected in 40 lungcancer cell lines tested, HYAL2 is not a likely tumour suppressor candidate for lungcancer.

Chapter 1

22

TUSC2TUSC2 codes for tumour suppressor candidate 2. Its cloning, structure and what wasknown about its expression and function has been reviewed by Lerman and Minna etal. (2000) and Zabarovsky et al. (2002). Although some mutations have been foundin TUSC2 these did not seem to impair the expression of the gene. More recently it

Figure 3. Model of JSRV-mediated transformation of human bronchial epithelial cells adapted fromDanilkovitch-Miagkova et al. (2003). A: Cells expressing MST1R as an inactive dimer, association ofHYAL2 with MST1R prevents MST1R activation. B: JSRV interacts with HYAL2 via the Env protein, whichleads to viral entry. C: Intracellular degradation of HYAL2. D: Possible MST1R conformational changesmay cause constitutive activation of oncogenic pathways.

Introduction

23

was postulated, however, that deficient post-translational modification could have aneffect on TUSC2 function. In human primary tumours and lung cancer cell lines adefect was observed in N-myristoylation of TUSC2 (Uno et al., 2004). N-myristoylation appears required for TUSC2 mediated tumour suppressor activity,since TUSC2 which is not myristoylated, is readily degraded. Some experimentalevidence has been obtained supporting a tumour suppressor function of TUSC2.TUSC2 transfection into NSCLC cell lines NCI-H1299 or NCI-H322 resulted in adramatically reduced colony formation compared to the mock transfected cell line(Kondo et al., 2001). Also the proliferation of several different NSCLC cell lines(A549, NCI-H1299, NCI-H358 and NCI-H460) transfected with TUSC2 wassignificantly reduced. Growth inhibition was also seen in vivo after intratumouralinjection of human xenografts with a TUSC2 adenoviral vector construct, as was asignificant inhibition of the development of A549 pulmonary metastases afterintravenous injection of the same construct (Ji et al., 2002). Growth inhibition andinhibition of pulmonary metastases could also be established by intratumouraladministration of a TUSC2 liposomal gene complex or intravenous injection of thiscomplex, respectively (Ito et al., 2004).

RASSF1RASSF1 codes for ras-association domain family 1 protein. Its cloning, structure andwhat was known about its expression has been reviewed by Lerman and Minna et al.(2000) and Zabarovsky et al. (2002). RASSF1 function has been extensively studiedand many proteins have been suggested to interact with the protein product of themajor transcript, RASSF1A. This protein product was suggested to block the cellcycle at the level of G1/S-phase transition by negatively regulating the accumulationof cyclin D1 (Shivakumar et al., 2002). More recently, it was suggested thatRASSF1A blocking of cyclin D1 accumulation was mediated through suppression ofthe c-jun-NH2-Kinase (JNK) pathway (Whang et al., 2005). RASSF1A was alsosuggested to interact with E4F1 (Fenton et al., 2004), an E1A-regulated transcriptionfactor which is associated with a number of cell cycle regulating proteins includingcyclin A, cyclin E and cyclin B (Fajas et al., 2000; Sandy et al., 2000).Overexpression of RASSF1A together with E4F1 resulted in a greater increase ofcells in the G1-phase than overexpression of RASSF1A alone (Fenton et al., 2004).The binding of RASSF1A to E4F1 has been suggested to enhance the inhibitoryeffect of E4F1 on Cyclin A2 expression (Ahmed-Choudhury et al., 2005). RASSF1A

Chapter 1

24

has also been suggested to interact with several different microtubulus associatedproteins including MAP1B and its close homologue VCY2IP1 (Dallol et al., 2004).Furthermore, RASSF1A seems to colocalise with the microtubuli and to stabilise themicrotubuli, which at mitosis results in metaphase arrest (Liu et al., 2002; Liu et al.,2003a). Stabilisation of microtubuli by RASSF1A was found to be mediated throughthe inhibitory effect of RASSF1A on the anaphase-promoting complex (APC).RASSF1A binds to Cdc20, thereby inhibiting binding of Cdc20 to APC, which leavesAPC inactive and the cells arrested in promethapase (Song et al., 2004; Castro etal., 2005). In addition, RASSF1A cotransfected with activated K-RAS blocks theability of oncogenic K-RAS to promote genomic instability (Vos et al., 2004).RASSF1A is therefore suggested to be an inhibitor of DNA synthesis, mitosis andcytokinesis. RASSF1A has also been suggested to heterodimerise with RASSF5 andthereby to bind indirectly to RAS-GTP (Khokhlatchev et al., 2002; Ortiz-Vega et al.,2002). In addition, it is proposed that RASSF5 and RASSF1A keep STK4 in aninhibited state, but have the ability to direct STK4 to specific cellular sites and/or co-localise it with upstream activators and/or substrates, thereby augmenting STK4apoptosis (Praskova et al., 2004). Furthermore, RASSF1A has been suggested toassociate with connector enhancer of kinase suppressor of Ras 1 (CNKSR1), amultidomain scaffold protein discovered in Drosophila, in which it is necessary forRAS activation of Raf kinase. Rabizadeh et al. (2004) have shown that coexpressionof CNKSR1 with RASSF1A greatly enhances CNKSR1-induced apoptosis, probablymediated by recruitment of STK4, which had already been shown to be associatedwith RASSF1A. A physical interaction was also found between RASSF1A andPMCA4b which encodes a plasma membrane calmodulin-dependent calciumATPase. Coexpression of RASSF1A and PMCA4b was shown to inhibit the EGF-dependent activation of the ERK pathway (Armesilla et al., 2004). One of the otherRASSF1 transcripts, RASSF1C, has been suggested to interact with IGFBP-5,coding for insulin-like growth factor binding protein 5. More recently, IGFBP-5 wassuggested to be able to stimulate cell proliferation mediated by activation of the p38MAP kinase and extracellular signal-regulated kinase (ERK)-1/2 pathways. SiRNA-mediated knockdown of RASSF1C was found to block the IGFBP-5-induced ERK1/2phosphorylation, thereby blocking cell proliferation (Amaar et al., 2005). Recently, amodel was proposed where RASSF1A and MOAP1 associate upon TNFαstimulation. Subsequently the complex promotes a conformational change of BAX,which mediates insertion of BAX into the mitochondrial membrane and ultimately

Introduction

25

cytochrome c release and apoptosis (Baksh et al., 2005). Figure 4 summarises theinteractions and pathways of RASSF1 transcripts.

The expression of RASSF1A is reduced in several cancers, including SCLC(Dammann et al., 2000), NSCLC and breast cancer (Burbee et al., 2001), gastriccancer (Byun et al., 2001), bladder cancer (Lee et al., 2001), thyroid cancer(Schagdarsurengin et al., 2002) and osteosarcoma (Lim et al., 2003). Only twoconfirmed somatic mutations have been found in over 200 samples of differenttumour tissues, including SCLC and breast cancer (Dammann et al., 2003). Incontrast, a high mutation rate, 17/23 (74%), was found for RASSF1A in primarynasopharyngeal carcinomas by the PCR-cloning-sequencing strategy (Pan et al.,2005). A rare polymorphism was found in codon 133 of exon 3 (alanine-serine) inthe germline of two cervical cancers, one nasopharyngeal cancer and in lung andbreast cancer cell lines (Yu et al., 2003). This polymorphism was found in 21% ofpatients with breast carcinoma and 24% of patients with fibroadenoma, whereas itwas only found in 3% of the controls (Schagdarsurengin et al., 2005). An alternativemechanism for loss of expression was found for RASSF1A. The promoter ofRASSF1A, but not of RASSF1C, is frequently hypermethylated in several cancersincluding lung cancer, breast cancer and renal cell cancer (Burbee et al., 2001;Dammann et al., 2001; Yoon et al., 2001). Table 1 summarises the methylationprofiles found for tumours tested for promoter hypermethylation.

A tumour suppressor function of RASSF1A has been tested in vitro in lungcancer cell lines, kidney cell lines and prostate cell lines. RASSF1A reinsertion led toreduced colony formation and/or anchorage-independent growth in soft agar(Dammann et al., 2000; Burbee et al., 2001; Dreijerink et al., 2001; Kuzmin et al.,2002; Chow et al., 2004; Li et al., 2004b). In addition, tumour suppressor activity hasalso been suggested by results obtained in several human lung cancer cell lines invivo (Burbee et al., 2001; Li et al., 2004b). In a mouse RASSF1A knockout model,heterozygous RASSF1A+/- mice and homozygous RASSF1A-/- mice were significantlymore tumour prone for spontaneous tumour formation as well as for chemicalinduction of tumours (Tommasi et al., 2005). In contrast, Suzuki et al. (2004) couldnot detect any growth effect in soft agar colony formation of NCI-H1299 (NSCLC)cells 14 days after DNA methyltransferase 1 (DNMT1) knockdown, although theyshowed that RASSF1A promoter methylation was reduced by 80% and RASSF1Aexpression was 60% increased.

Chapter 1

26

Figure 4. A summary of RASSF1 transcripts interactions and pathways adapted from Agathanggelou etal. (2005).

Introduction

27

Table 1. RASSF1A promoter hypermethylation in primary tumours

Tumour type primary tumours noncanceroustissue

reference

Small cell lung cancer 79% 22(28) (Dammann et al., 2001)72% 21(29) (Agathanggelou et al., 2001)~ 80% 34(43) (Toyooka et al., 2001)50% 4(8) (Honorio et al., 2003)

Non-small cell lung cancer 38% 22(58) (Dammann et al., 2000)34% 14(41) (Agathanggelou et al., 2001)30% 32(107) (Burbee et al., 2001)~ 30% 34(115) (Toyooka et al., 2001)32% 35(110) (Tomizawa et al., 2002)32% 65(204 (Kim et al., 2003)21% 5(21) (Honorio et al., 2003)43% 32(75) (Yanagawa et al., 2003)

Adenocarcinoma 47% 7(15) (Ramirez et al., 2003)39% 49(125) (Kim et al., 2003)55% 18(33) (Li et al., 2003)39% 28(72) (Tomizawa et al., 2004)38% 80(209) (Divine et al., 2004)50% 50(100) (Marsit et al., 2004)31% 31(101) (Ito et al., 2005)43% 13(30) (Ito et al., 2005)

Stage I ac 32% 35(110) (Tomizawa et al., 2002)Squamous cell carcinoma 24% 6(25) (Ramirez et al., 2003)

30% 32(107) (Zochbauer-Muller et al.,2003)

30% 25(84) (Kim et al., 2003)25% 5(20) (Li et al., 2003)41% 51(124) (Maruyama et al., 2004)13% 6(45) (Tomizawa et al., 2004)31% 22(71) (Marsit et al., 2004)

Large cell carcinoma 40% 4(10) (Ramirez et al., 2003)25% 3(12) (Li et al., 2003)

Lung cancer 60% 12(20) (Guo et al., 2004)Carcinoid cancer 50% 20(40) (Toyooka et al., 2001)Sputum from chronicsmokers

30% 4(13) (Honorio et al., 2003)

4% 3(73) (Zochbauer-Muller et al.,2003)

Sputum from former smokers 50% 1(2) (Honorio et al., 2003)Breast cancer 62% 28(45) (Dammann et al., 2001)

9% 4(44) (Agathanggelou et al., 2001)49% 19(39) (Burbee et al., 2001)56% 20(36) (Lehmann et al., 2002)58% 56(97) (Chen et al., 2003)81% 122(147) (Shinozaki et al., 2005)95% 38(40) 93% 37(40) (Yeo et al., 2005)68% 13(19) 7% 2(28) (Fackler et al., 2004)

Invasive breast cancer (IDC) 65% 11(17) (Honorio et al., 2003)70% 19(27) (Fackler et al., 2003)64% 29(45) (Pu et al., 2003)

Ductal carcinoma in situ(DCIS)

42% 5(12) (Honorio et al., 2003)

75% 33 (44) (Fackler et al., 2003)

Chapter 1

28

Table 1 continued

Tumour type primary tumours noncanceroustissue

Reference

In situ lobular carcinoma(LCIS)

62% 8(13) (Fackler et al., 2003)

ILC 84% 16(19) (Fackler et al., 2003)benign 34% 12(36) (Pu et al., 2003)In situ 62% 13(21) (Pu et al., 2003)primary 23% 6(23) (Muller et al., 2003)recurrent 80% 8 (10) (Muller et al., 2003)

56% 14(25) (Mehrotra et al., 2004)Metastasis bone 78% 7(9) (Mehrotra et al., 2004)Metastasis brain 67% 4(6) (Mehrotra et al., 2004)Metastasis lung 100% 10(10) (Mehrotra et al., 2004)Ovarian cancer 10% 2(21) (Agathanggelou et al., 2001)

40% 8(20) (Yoon et al., 2001)49% 20(41) (Toyooka et al., 2001)50% 25(50) (Ibanez de Caceres et al.,

2004)Invasive 30% 14(46) (Makarla et al., 2005)

Not invasive 0% 0(92) 13% 2(16) (Makarla et al., 2005)Cervical cancer 0% 0(22) (Agathanggelou et al., 2001)

17.5% 9(51) (Cohen et al., 2003)30% 11(33) SCC (Yu et al., 2003)12% 2(17) AC (Yu et al., 2003)10% 4(42) SCC (Kuzmin et al., 2003)24% 8(34) AC (Kuzmin et al., 2003)21% 4(19) ASC (Kuzmin et al., 2003)12.2% 10 (82) 41.2% 7(17) (Kang et al., 2005)

Malignant mesothelioma 32% 21(66) (Toyooka et al., 2002)0% 0(1) (Seidel et al., 2004)

phaeochromocytoma 22% 5(23) (Astuti et al., 2001)50% 13(26)

Renal cell carcinoma 56% 18(32) (Yoon et al., 2001)91% 39(43) (Dreijerink et al., 2001)

Papillary RCC 44% 12(27) (Morrissey et al., 2001)70% 14(20) (Dulaimi et al., 2004)100% 9(9) (Gonzalgo et al., 2004)

Clear Cell 46% 23(50) (Dulaimi et al., 2004)90% 9 (21) (Gonzalgo et al., 2004)

Wilm’s tumour 0% 0(1) (Dulaimi et al., 2004)88% 15(17) (Hoque et al., 2004)

CC-RCC 23% 32(138) (Morrissey et al., 2001)Chromophobe 17% 1(6) (Dulaimi et al., 2004)oncocytoma 14% 1(7) (Dulaimi et al., 2004)Collecting duct 60% 3(5) (Dulaimi et al., 2004)RCC unclassified 20% 1(5) (Dulaimi et al., 2004)TCC renal pelvis 33% 2(6) (Dulaimi et al., 2004)oncocytoma 25% 2(8) (Gonzalgo et al., 2004)Nasopharyngeal cancer 67% 14(21) (Lo et al., 2001)

50% 8(16) (Tong et al., 2002)67% 20(30) (Chang et al., 2003)

Mouth and throat rinsing fluid 33% 10(30) (Chang et al., 2003)Nasopharyngeal swab 37% 11(30) (Chang et al., 2003)Head and neck cancer 8% 6(80) (Hasegawa et al., 2002)

Introduction

29

Table 1 continued

Tumour type primary tumours noncanceroustissue

Reference

17% 4(24) (Hogg et al., 2002)15% 7(46) (Dong et al., 2003)10% 3 (32) 0% 0(32) (Maruya et al., 2004)

Betel-associated oralcarcinomaSquamous cell carcinoma 93% 25(27) (Tran et al., 2005)Verrucous carcinoma 100% 9(9) (Tran et al., 2005)Gastric cancer 43% 39(90) (Byun et al., 2001)

67% 14(21) (Kang et al., 2002)4% 2(56) (Kang et al., 2002)7,5% 6(80) (Kang et al., 2003)

MSI-H ~ 3% 1(36) (Kim et al., 2003)MSI-S 0% (Kim et al., 2003)Gastric stromal 40% 15(38) (House et al., 2003)Chronic gastritis 0.4% 1(268) (Kang et al., 2003)Bladder cancer 60% 33(55) (Lee et al., 2001)

35% 34(98) (Maruyama et al., 2001)48% 19(40) (Chan et al., 2003)

Low grade transitional cellcarcinoma (TCC)

47.5% 19(40) (Chan et al., 2003)

Carcinoma in situ 0% 0(6) (Chan et al., 2003)60% 76(127) 42% 16(37) (Friedrich et al., 2004)32% 116(351) (Marsit et al., 2005)

Biliary tract carcinomas 27% 10(37) (Tozawa et al., 2004)Prostate cancer 53% 54(101) (Maruyama et al., 2002)

100% 11(11) (Kuzmin et al., 2002)71% 37(52) (Liu et al., 2002)67% 59(90) (Woodson et al., 2004)96% 70(73) (Yegnasubramanian et al.,

2004)83% 20(24) (Woodson et al., 2004)78% 88 (113) 53% 19(36) (Florl et al., 2004)99.2% 117(118) (Jeronimo et al., 2004)

(high grade intraepithelialneoplasia

100% 38(38) (Jeronimo et al., 2004)

30% 3(10) (Woodson et al., 2004)(beging prostatic hyperplasia) 93.3 28(30) (Jeronimo et al., 2004)Colon cancer 12% 3(26) (Yoon et al., 2001)

20% 45(222) (van Engeland et al., 2002)45% 13(29) (Wagner et al., 2002)

High methyl donor intake 15% 9(61) (van Engeland et al., 2003)Low methyl donor intake 25% 15(61)Flat-type 81.3% 39(48) 49% 19(39) (Sakamoto et al., 2004)

3% 2(65) (Xu et al., 2004)Thyroid cancer 71% 27(38) (Schagdarsurengin et al.,

2002)Follicular 75% 9(12) (Xing et al., 2004)Follicular thyroid carcinoma 100% 4(4) (Nakamura et al., 2005)Follicular adenoma 33% 1(3) (Nakamura et al., 2005)papillary 20% 6(30) (Xing et al., 2004)Papillary thyroid carcinoma 26% 11(34) 0% 0(27) (Nakamura et al., 2005)Benign adenomas 44% 4(9) (Xing et al., 2004)

Chapter 1

30

Table 1 continued

Tumour type primary tumours noncanceroustissue

Reference

Medullary thyroid carcinoma 40% 2(5) (Nakamura et al., 2005)Anaplastic thyroid carcinoma 33% 4(12) (Nakamura et al., 2005)Hyalinizing trabecular 25% 5(20) (Nakamura et al., 2005)Pituitary adenomas 38% 20(52) (Qian et al., 2005)Salivary adenoid cystic 42% 25(60) (Li et al., 2005)Pediatric cancer10 different

40% 70(175) (Harada et al., 2002)

67% 16(24) 31% 4(13) (Wong et al., 2004)Neuroblastoma 55% 37(67) (Astuti et al., 2001)

70% 39(56) (Yang et al., 2004)Hodgkin’s lymphoma 65% 34(52) (Murray et al., 2004)Wilms tumour 73% 22(30) (Ehrlich et al., 2002)

54% 22(30) (Wagner et al., 2002)Esophageal squamous cellcarcinoma (slokdarm)

52% 25(48) (Kuroki et al., 2003)

51% 24(47) (Kuroki et al., 2003)Malignant melanomacutaneous 55% 24(55) (Spugnardi et al., 2003)

22% (Reifenberger et al., 2004)36% 9(25) (Furuta et al., 2004)

Cholangiocarcinoma 69% 9(13) (Wong et al., 2002)Extrahepaticcholangiocarcinoma

85% 28(33) (Chen et al., 2005)

Hepatocellular carcinoma 85% 70(82) (Zhang et al., 2002)93% 14(15) (Schagdarsurengin et al.,

2003)100% 29(29) 83% 24(29) (Yu et al., 2002)95% 41(43) 70% 16(23) (Zhong et al., 2003)67% 40(60) (Lee et al., 2003)

Testicular germ celltumours

71% 17(24) (Honorio et al., 2003)

0% 0(25) (Kawakami et al., 2003)Testicular malignantlymphomas

100% 3(3)

Male germ cell tumours 21.7% 20(92)83 primary

(Koul et al., 2002)

Differentiated nonseminoma 35.7% 25(70) (Koul et al., 2004)Primary brain tumourMedullablastoma 80% 22(27) (Lusher et al., 2002)

100% 5(5)93% 41(44) (Lindsey et al., 2004)

Glioblastoma multiforme 57% 12(21) (Balana et al., 2003)Serum 50% 13(26) (Balana et al., 2003)Gliomas (invasive, invariablyfata intracerebral tumours)

54% 25(46) (Horiguchi et al., 2003)

astrocytoma 69.8% 37(53) (Yu et al., 2004)ependymoma 85% 30(35) (Hamilton et al., 2005)Multiple myeloma 28% 9(32) (Ng et al., 2003)Muliple myelomaMonoclonal gammopathies

15% 4(29)14% 17 (113)

(Seidl et al., 2004)

Pancreatic tumours

Introduction

31

Table 1 continued

Tumour type primary tumours noncanceroustissue

Reference

PEN 75% 36(48) (House et al., 2003)83% 10(12) (Dammann et al., 2003)

adenocarcinoma 64% 29(45) (Dammann et al., 2003)Soft tissue sarcomas (Seidel et al., 2004)Liposarcoma 18% 4(22) (Seidel et al., 2004)Leiomyosarcoma 39% 7(18) (Seidel et al., 2004)MFH 6% 1(18) (Seidel et al., 2004)Rhabdomyosarcoma 0% 0(6) (Seidel et al., 2004)Neurogenic sarcoma 50% 3(6) (Seidel et al., 2004)Synovial sarcoma 33% 2(6) (Seidel et al., 2004)Fibrosarcoma 0% 0(3) (Seidel et al., 2004)Malignanthemangiopericytoma

0% 0(3) (Seidel et al., 2004)

ZMYND10ZMYND10, codes for zinc finger, MYND-type containing 10. Its cloning, structure andwhat was known about its expression and function has been reviewed by Lermanand Minna et al. (2000) and by Zabarovsky et al. (2002). The gene contains a Zincfinger MYND domain that is involved in specific protein-protein interactions (Lermanand Minna, 2000). More recently, Liu et al. (2003b) found that the domain may play arole in transcription regulation. The functional promoter of ZMYND10 was found tobecome activated by environmental stresses such as heat shock and was found tobe regulated by E2F, a transcription factor involved in the control of the cell cycle(Qiu et al., 2004). Although loss of expression of ZMYND10 was detected in 70% ofSCLC and NSCLC cell lines (Lerman and Minna, 2000) and in 78% of primarynasopharyngeal carcinomas (Liu et al., 2003b) only three missense mutations werefound in 61 lung cancer cell lines (Lerman and Minna, 2000) and no pathogenicmutations were found in 45 primary nasopharyngeal tumours and fivenasopharyngeal cell lines (Liu et al., 2003b). Since the second allele of ZMYND10 israrely inactivated by mutations, hypermethylation of the promoter might be analternative mechanism explaining the frequently found loss of expression.Methylation of the ZMYND10 promoter region CpG island was detected in 21/54(39%) of lung cancer cell lines, 3/7 (42%) of breast cancer cell lines, 3/6 (50%) ofkidney cancer cell lines, 6/7 (86%) of neuroblastoma cell lines and 4/5 (80%)nasopharyngeal carcinoma cell lines (Agathanggelou et al., 2003). In addition,

Chapter 1

32

ZMYND10 methylation was also detected in primary tumours: in SCLC 4/29 or 14%of cases (Agathanggelou et al., 2003), in NSCLC 26/145 or 19% (Agathanggelou etal., 2003) 41/138 or 30% (Ito et al., 2005) 68/160 or 43% (Marsit et al., 2005), inneuroblastomas 20/49 or 41% (Agathanggelou et al., 2003) and in nasopharyngealcarcinomas 17/23 or 74% (Liu et al., 2003b), 8/26 or 30.8 % (Chow et al., 2004) andin 19/29 or 66% (Qiu et al., 2004). Promoter hypermethylation of ZMYND10 seemedto be inversely correlated with smoking and occurred more often in adenocarcinomathan in squamous cell carcinoma (Marsit et al., 2005). The highest percentage ofpromoter hypermethylation was found in nasopharyngeal carcinomas, 74%. Inaddition, 7/29 (24%) of primary tumours of ZMYND10 were found homozygouslydeleted for ZMYND10, implying that in total up to 83% of the nasopharyngealtumours showed some aberration of ZMYND10 (Qiu et al., 2004).

ZMYND10 seemed to show tumour suppressor activity in vitro, sincetransfection of ZMYND10 resulted in a 49% and 75% reduced colony formation in anNSCLC cell line (NCI-H1299) and a neuroblastoma cell line (SK-N-SH), respectively(Agathanggelou et al., 2003). No tumour suppressor activity was found in vivo, sinceno significant growth inhibition was detected after intratumoural injection of anadenoviral vector containing ZMYND10 into a human NSCLC xenograft (Ji et al.,2002). Also after injection into a nude mice of a esophageal squamous cellcarcinoma cell line (SLMT-1) transfected with ZMYND10 no reduction of tumourgrowth was found as compared to the parental cell line (Yi Lo et al., 2005).

TUSC4TUSC4 codes for tumor suppressor candidate 4. Its cloning, structure and what wasknown about its expression has been reviewed by Lerman and Minna et al. (2000).The main spliceform of TUSC4 encodes a soluble protein with a bipartite nuclearlocalisation signal, a protein-binding domain, similarity to the MutS core domain anda newly identified nitrogen permease regulator 2 domain with unknown function (Li etal., 2004a). TUSC4 appeared to be expressed in lung cancer, although the 3’end ofTUSC4 was found homozygously deleted in 6 of the 19 cancer cell lines of differentorigin, including SCLC cell lines (U2020 and A549), a non-SCLC cell line (H647) andrenal cell carcinoma cell lines (HN4 and ACHN) (Li et al., 2004a). In five of the sixcell lines the homozygous deletion included RASSF1. Promoter hypermethylationwas not observed in an analysis of six nasopharyngeal carcinoma cell lines (Chow etal., 2004). There is evidence of tumour suppressor activity of this gene. After

Introduction

33

transfection of NSCLC cells with TUSC4 in an adenoviral vector, in vitro cellproliferation was significantly reduced. Growth inhibition of TUSC4 was also seen inhuman xenografts after intratumoural injection (Ji et al., 2002). In addition, Li et al.(2004a) using a gene inactivation test with a tetracycline inducible vector system,found a reduced colony formation in vitro and a delayed growth rate in vivo of aTUSC4-transfected RCC cell line (KRC/Y) and two SCLC cell lines (U2020 andA549).

CYB561D2CYB561D2 codes for cytochrome b-561 domain containing 2. Its cloning, structureand what was known about its expression and function has been reviewed byLerman and Minna et al. (2000). There is evidence of tumour suppressor activity ofCYB561D2, since cell proliferation of NSCLC cells transfected with CYB561D2 in anadenoviral vector, was significantly reduced and also after intratumoural injection ofthe same clones into xenografts tumour growth appeared inhibited (Ji et al., 2002).Perhaps, the abundant expression of CYB561D2 found in lung cancer cell lines mustbe explained by posttranslational modification at the protein level.

PL6PL6 codes for placental protein 6. Its cloning, structure and what was known aboutits expression has been reviewed by Lerman and Minna et al. (2000). Since PL6expression was reduced in SCLC cell lines, PL6 might be involved in small cell lungcancer.

CACNA2D2CACNA2D2 codes for calcium channel, voltage-dependent, alpha 2/delta subunit 2gene. Its cloning, structure and what was known about its expression and functionhas been reviewed by Lerman and Minna et al. (2000) and Zabarovsky et al. (2002).Although over 50% of lung cancer cell lines tested showed a reduced or absentexpression of CACNA2D2, particularly in NSCLCs, no mutations could be detectedin 60 lung cancer cell lines and 40 primary tumours (Lerman and Minna, 2000). Lossof CACNA2D2 expression is probably not mediated by promoter hypermethylation,since in only one primary glioma tumour promoter hypermethylation of CACNA2D2could be detected (Hesson et al., 2004). Promoter hypermethylation has, however,

Chapter 1

34

not been tested in lung tumours. CACNA2D2 was shown to possess tumoursuppressor activity. Cell proliferation of the NSCLC cell lines NCI-H358, NCI-H460,NCI-1299 and A549 transfected with CACNA2D2 in an adenoviral vector, wasreduced significantly. A significant suppression of tumour growth compared to controlgroups treated with PBS was found in NCI-H460 tumours treated with CACNA2D2 inan adenoviral vector. CACNA2D2 was also found capable of inducing apoptosis inthree out of four NSCLC cell lines tested. It was shown that the apoptotic effect isassociated with the regulation of cystosolic Ca2+ contents and the activation of themitochondrial pathway (Carboni et al., 2003).

A number of genes, RBM5, SEMA3F, SEMA3B, HYAL1, TUSC2, RASSF1,ZMYND10, TUSC4, CYB561D2, PL6 and CACNA2D2 from the 3p21.3 critical regionhave been implicated in lung cancer. Convincing evidence for an involvement of anyof these genes in the development of lung cancer is, however, lacking. In thefunctional tests for these genes cDNAs have mostly been cloned behind strong viralpromoters that cause an abundant production of a single protein which may welldisturb essential cellular processes and thereby affect growth and proliferation of thetumour cells, i.e. mimic tumour suppression. To demonstrate a tumour suppressorrole, an approach would be needed in which a functional effect is achieved at a lowgene dosage.

Introduction

35

REFERENCES

Agathanggelou A, Dallol A, Zochbauer-Muller S, Morrissey C, Honorio S, Hesson L, Martinsson T, FongKM, Kuo MJ, Yuen PW, Maher ER, Minna JD, Latif F. 2003. Epigenetic inactivation of thecandidate 3p21.3 suppressor gene BLU in human cancers. Oncogene. 22:1580-1588.

Ahmed-Choudhury J, Agathanggelou A, Fenton SL, Ricketts C, Clark GJ, Maher ER, Latif F. 2005.Transcriptional regulation of cyclin A2 by RASSF1A through the enhanced binding of p120E4F tothe cyclin A2 promoter. Cancer Res 65:2690-2697.

Amaar YG, Baylink DJ, Mohan S. 2005. Ras-Association Domain Family 1 Protein, RASSF1C, Is anIGFBP-5 Binding Partner and a Potential Regulator of Osteoblast Cell Proliferation. J Bone MinerRes 20:1430 -1439.

Armesilla AL, Williams JC, Buch MH, Pickard A, Emerson M, Cartwright EJ, Oceandy D, Vos MD, GilliesS, Clark GJ, Neyses L. 2004. Novel functional interaction between the plasma membrane Ca2+pump 4b and the proapoptotic tumor suppressor Ras-associated factor 1 (RASSF1). J Biol Chem279:31318-31328.

Auvinen PK, Parkkinen JJ, Johansson RT, Agren UM, Tammi RH, Eskelinen MJ, Kosma VM. 1997.Expression of hyaluronan in benign and malignant breast lesions. Int J Cancer 74:477-481.

Auvinen PK, Tammi R, Parkkinen J, Tammi M, Agren U, Johansson R, Hirvikoski P, Eskelinen M, KosmaVM. 2000. Hyaluronan in peritumoral stroma and malignant cells associates with breast cancerspreading and predicts survival. Am J Pathol 156: 529-536 .

Baksh S, Tommasi S, Fenton S, Yu VC, Martins LM, Pfeifer GP, Latif F, Downward J, Neel BG. 2005. Thetumor suppressor RASSF1A and MAP-1 link death receptor signaling to Bax conformationalchange and cell death. Mol Cell 18:637-650.

Bertrand P, Courel MN, Maingonnat C, Jardin F, Tilly H, Bastard C.2005. Expression of HYAL2 mRNA,hyaluronan and hyaluronidase in B-cell non-Hodgkin lymphoma: relationship with tumoraggressiveness. Int J Cancer 10:207-212.

Bielenberg DR, Hida Y, Shimizu A, Kaipainen A, Kreuter M, Kim CC, Klagsbrun M. 2004. Semaphorin 3F,a chemorepulsant for endothelial cells, induces a poorly vascularized, encapsulated,nonmetastatic tumor phenotype. J Clin Invest 114:1260-1271.

Burbee DG, Forgacs E, Zochbauer-Muller S, Shivakumar L, Fong K, Gao B, Randle D, Kondo M, VirmaniA, Bader S, Sekido Y, Latif F, Milchgrub S, Toyooka S, Gazdar AF, Lerman MI, Zabarovsky E,White M, Minna JD. 2001. Epigenetic inactivation of RASSF1A in lung and breast cancers andmalignant phenotype suppression. J Natl Cancer Inst 93:691-699.

Byun DS, Lee MG, Chae KS, Ryu BG, Chi SG. 2001. Frequent epigenetic inactivation of rassf1a byaberrant promoter hypermethylation in human gastric adenocarcinoma. Cancer Res 61:7034-7038.

Carboni GL, Gao B, Nishizaki M, Xu K, Minna JD, Roth JA, Ji L. 2003. CACNA2D2-mediated apoptosis inNSCLC cells is associated with alterations of the intracellular calcium signaling and disruption ofmitochondria membrane integrity. Oncogene 22:615-626.

Carritt B, Kok K, van den Berg A, Osinga J, Pilz A, Hofstra RMW, Davis MB, van der Veen AY, RabbittsPH, Gulati K, Buys CHCM. 1992. A gene from human chromosome region 3p21 with reducedexpression in small cell lung cancer. Cancer Res 52:1536-1541.

Castro A, Bernis C, Vigneron S, Labbe JC, Lorca T. 2005. The anaphase-promoting complex: a key factorin the regulation of cell cycle. Oncogene 24:314-325.

Castro-Rivera E, Ran S, Thorpe P, Minna JD. 2004. Semaphorin 3B (SEMA3B) induces apoptosis in lungand breast cancer, whereas VEGF165 antagonizes this effect. Proc Natl Acad Sci U S A101:11432-11437.

Chapter 1

36

Chang NS. 2002. Transforming growth factor-beta1 blocks the enhancement of tumor necrosis factorcytotoxicity by hyaluronidase Hyal-2 in L929 fibroblasts. BMC Cell Biol 3:8-16.

Chow LS, Lo KW, Kwong J, To KF, Tsang KS, Lam CW, Dammann R, Huang DP. 2004. RASSF1A is atarget tumor suppressor from 3p21.3 in nasopharyngeal carcinoma. Int J Cancer 109: 839-847.

Dallol A, Forgacs E, Martinez A, Sekido Y, Walker R, Kishida T, Rabbitts P, Maher ER, Minna JD, Latif F.2002. Tumour specific promoter region methylation of the human homologue of the DrosophilaRoundabout gene DUTT1 (ROBO1) in human cancers. Oncogene 21:3020-3028.

Dallol A, Agathanggelou A, Fenton SL, Ahmed-Choudhury J, Hesson L, Vos MD, Clark GJ, Downward J,Maher ER, Latif F.2004. RASSF1A interacts with microtubule-associated proteins and modulatesmicrotubule dynamics. Cancer Res 64:4112-4116.

Daly MC, Xiang RH, Buchhagen D, Hensel CH, Garcia DK, Killary AM, Minna JD, Naylor SL.1993. Ahomozygous deletion on chromosome 3 in a small cell lung cancer cell line correlates with aregion of tumor suppressor activity. Oncogene 8:1721-1729.

Dammann R, Li C, Yoon JH, Chin PL, Bates S, Pfeifer GP. 2000. Epigenetic inactivation of a RASassociation domain family protein from the lung tumour suppressor locus 3p21.3. Nat Genet25:315-319.

Dammann R, Takahashi T, Pfeifer GP. 2001. The CpG island of the novel tumor suppressor geneRASSF1A is intensely methylated in primary small cell lung carcinomas. Oncogene 20:3563-3567.

Dammann R, Schagdarsurengin U, Strunnikova M, Rastetter M, Seidel C, Liu L, Tommasi S, Pfeifer GP.2003. Epigenetic inactivation of the Ras-association domain family 1 (RASSF1A) gene and itsfunction in human carcinogenesis. Histol Histopathol 18:665-677.

Danilkovitch-Miagkova A, Duh FM, Kuzmin I, Angeloni D, Liu SL, Miller AD, Lerman MI. 2003.Hyaluronidase 2 negatively regulates RON receptor tyrosine kinase and mediates transformationof epithelial cells by jaagsiekte sheep retrovirus. Proc Natl Acad Sci U S A 100:4580-4585.

Delmage JM, Powars DR, Jaynes PK, Allerton SE. 1996. The selective suppression of immunogenicity byhyaluronic acid. Ann Clin Lab Sci 16:303-310.

Dirks C, Duh FM, Rai SK, Lerman MI, Miller AD. 2002. Mechanism of cell entry and transformation byenzootic nasal tumor virus. J Virol 76:2141-2149.

Dreijerink K, Braga E, Kuzmin I, Geil L, Duh FM, Angeloni D, Zbar B, Lerman MI, Stanbridge EJ, MinnaJD, Protopopov A, Li J, Kashuba V, Klein G, Zabarovsky ER. 2001. The candidate tumorsuppressor gene, RASSF1A, from human chromosome 3p21.3 is involved in kidneytumorigenesis. Proc Natl Acad Sci U S A 98:7504-7509.

Edamatsu H, Kaziro Y, Itoh H. 2000. LUCA15, a putative tumour suppressor gene encoding an RNA-binding nuclear protein, is down-regulated in ras-transformed Rat-1 cells. Genes Cells 5:849-858.

Fajas L, Paul C, Zugasti O, Le Cam L, Polanowska J, Fabbrizio E, Medema R, Vignais ML, Sardet C.2000. pRB binds to and modulates the transrepressing activity of the E1A-regulated transcriptionfactor p120E4F. Proc Natl Acad Sci U S A 97:7738-7743.

Feinberg RN, Beebe DC. 1983. Hyaluronate in vasculogenesis. Science 220:1177-1179.

Fenton SL, Dallol A, Agathanggelou A, Hesson L, Ahmed-Choudhury J, Baksh S, Sardet C, Dammann R,Minna JD, Downward J, Maher ER, Latif F. 2004. Identification of the E1A-regulated transcriptionfactor p120 E4F as an interacting partner of the RASSF1A candidate tumor suppressor gene.Cancer Res 64:102-107.

Fiszer Szafarz B, Szafarz D. 1973. Lysosomal hyaluronidase activity in normal rat liver and in chemicallyinduced hepatomas. Cancer Res 33:1104-1108.

Franzmann EJ, Schroeder GL, Goodwin WJ, Weed DT, Fisher P, Lokeshwar VB. 2003. Expression oftumor markers hyaluronic acid and hyaluronidase (HYAL1) in head and neck tumors. Int J Cancer

Introduction

37

106: 438-445.

Frost GI, Csoka TB, Wong T, Stern R. 1997. Purification, cloning, and expression of human plasmahyaluronidase. Biochem Biophys Res Commun 236:10-15.

Grote HJ, Schmiemann V, Geddert H, Rohr UP, Kappes R, Gabbert HE, Bocking A. 2005. Aberrantpromoter methylation of p16(INK4a), RARB2 and SEMA3B in bronchial aspirates from patientswith suspected lung cancer. Int J Cancer 116:720-725.

Heppell-Parton AC, Nacheva E, Carter NP, Bergh J, Ogilvie D, Rabbitts PH. 1999. Elucidation of themechanism of homozygous deletion of 3p12-13 in the U2020 cell line reveals the unexpectedinvolvement of other chromosomes. Cancer Genet Cytogenet 111:105-110.

Hermouet S, Merendino JJ, Jr., Gutkind JS, Spiegel AM. 1991. Activating and inactivating mutations of thealpha subunit of Gi2 protein have opposite effects on proliferation of NIH 3T3 cells. Proc Natl AcadSci U S A 88:10455-10459.

Hermouet S, Aznavoorian S, Spiegel AM. 1996. In vitro and in vivo growth inhibition of murine melanomaK-1735 cell by a dominant negative mutant alpha subunit of the Gi2 protein. Cell Signal 8:159-166.

Hesson L, Bieche I, Krex D, Criniere E, Hoang-Xuan K, Maher ER, Latif F. 2004. Frequent epigeneticinactivation of RASSF1A and BLU genes located within the critical 3p21.3 region in gliomas.Oncogene 23:2408-2419.

Hibi K, Takahashi T, Yamakawa K, Ueda R, Sekido Y, Ariyoshi Y, Suyama M, Takagi H, Nakamura Y.1992. Three distinct regions involved in 3p deletion in human lung cancer. Oncogene 7:445-449.

Ho WL, Chang JW, Tseng RC, Chen JT, Chen CY, Jou YS, Wang YC. 2002. Loss of heterozygosity at lociof candidate tumor suppressor genes in microdissected primary non-small cell lung cancer.Cancer Detect Prev 26:343-349.

Iijima H, Tomizawa Y, Iwasaki Y, Sato K, Sunaga N, Dobashi K, Saito R, Nakajima T, Minna JD, Mori M:Genetic and epigenetic inactivation of LTF gene at 3p21.3 in lung cancers. Int J Cancer 2005 Sep8 ; .: (2005).

Ito I, Ji L, Tanaka F, Saito Y, Gopalan B, Branch CD, Xu K, Atkinson EN, Bekele BN, Stephens LC, MinnaJD, Roth JA, Ramesh R. 2004. Liposomal vector mediated delivery of the 3p FUS1 genedemonstrates potent antitumor activity against human lung cancer in vivo. Cancer Gene Ther11:733-739.

Ito M, Ito G, Kondo M, Uchiyama M, Fukui T, Mori S, Yoshioka H, Ueda Y, Shimokata K, Sekido Y. 2005.Frequent inactivation of RASSF1A, BLU, and SEMA3B on 3p21.3 by promoter hypermethylationand allele loss in non-small cell lung cancer. Cancer Lett 225:131 -139.

Jacobson A, Rahmanian M, Rubin K, Heldin P. 2002. Expression of hyaluronan synthase 2 orhyaluronidase 1 differentially affect the growth rate of transplantable colon carcinoma cell tumors.Int J Cancer 102:212-219.

Ji L, Nishizaki M, Gao B, Burbee D, Kondo M, Kamibayashi C, Xu K, Yen N, Atkinson EN, Fang B, LermanMI, Roth JA, Minna JD. 2002. Expression of several genes in the human chromosome 3p21.3homozygous deletion region by an adenovirus vector results in tumor suppressor activities in vitroand in vivo. Cancer Res 62:2715-2720.

Jojovic M, Delpech B, Prehm P, Schumacher U. 2002. Expression of hyaluronate and hyaluronatesynthase in human primary tumours and their metastases in scid mice. Cancer Lett 188:181-189.

Junker N, Latini S, Petersen LN, Kristjansen PE. 2003. Expression and regulation patterns ofhyaluronidases in small cell lung cancer and glioma lines. Oncol Rep 10:609-616.

Kashuba VI, Li J, Wang F, Senchenko VN, Protopopov A, Malyukova A, Kutsenko AS, Kadyrova E,Zabarovska VI, Muravenko OV, Zelenin AV, Kisselev LL, Kuzmin I, Minna JD, Winberg G, ErnbergI, Braga E, Lerman MI, Klein G, Zabarovsky ER.2004. RBSP3 (HYA22) is a tumor suppressorgene implicated in major epithelial malignancies. Proc Natl Acad Sci U S A 101:4906 -4911.

Chapter 1

38

Kessler O, Shraga-Heled N, Lange T, Gutmann-Raviv N, Sabo E, Baruch L, Machluf M, Neufeld G. 2004.Semaphorin-3F is an inhibitor of tumor angiogenesis. Cancer Res 64:1008-1015.

Khokhlatchev A, Rabizadeh S, Xavier R, Nedwidek M, Chen T, Zhang X, Seed B, Avruch J. 2002.Identification of a novel ras-regulated proapoptotic pathway. Curr Biol 12:253-265.

Kholodnyuk I, Kost Alimova M, Kashuba V, Gizatulin R, Szeles A, Stanbridge EJ, Zabarovsky ER, Klein G,Imreh S. 1997. A 3p21.3 region is preferentially eliminated from human chromosome 3/mousemicrocell hybrids during tumor growth in SCID mice. Genes Chromosomes Cancer 18:200-211.

Kholodnyuk ID, Kost-Alimova M, Yang Y, Kiss H, Fedorova L, Klein G, Imreh S. 2002. The microcellhybrid-based "elimination test" identifies a 1-Mb putative tumor-suppressor region at 3p22.2-p22.1centromeric to the homozygous deletion region detected in lung cancer. Genes ChromosomesCancer 34:341-344.

Kidd T, Brose K, Mitchell KJ, Fetter RD, Tessier Lavigne M, Goodman CS, Tear G. 1998. Roundaboutcontrols axon crossing of the CNS midline and defines a novel subfamily of evolutionarilyconserved guidance receptors. Cell 92:205-215.

Killary AM, Wolf ME, Giambernardi TA, Naylor SL. 1992. Definition of a tumor suppressor locus withinhuman chromosome 3p21-p22. Proc Natl Acad Sci USA 89:10877-10881.

Kiss H, Yang Y, Kiss C, Andersson K, Klein G, Imreh S, Dumanski JP. 2002. The transcriptional map ofthe common eliminated region 1 (C3CER1) in 3p21.3. Eur J Hum Genet 10:52-61.

Kitareewan S, Pitha-Rowe I, Sekula D, Lowrey CH, Nemeth MJ, Golub TR, Freemantle SJ, Dmitrovsky E.2002. UBE1L is a retinoid target that triggers PML/RARalpha degradation and apoptosis in acutepromyelocytic leukemia. Proc Natl Acad Sci U S A 99:3806-3811.

Knudson W, Biswas C, Li XQ, Nemec RE, Toole BP. 1989. The role and regulation of tumour-associatedhyaluronan. Ciba Found Symp 143:150-159.

Kok K, Osinga J, Carritt B, Davis MB, van der Hout AH, van der Veen AY, Landsvater RM, de Leij LFMH,Berendsen HH, Postmus PE, Poppema S, Buys CHCM. 1987. Deletion of a DNA sequence at thechromosomal region 3p21 in all major types of lung cancer. Nature 330:578-581.

Kok K, van den Berg A, Buchhagen DL, Carritt B, Buys CHCM. 1993. A PCR-aided transcript titrationassay revealing very low expression of a gene at band 3p21 in 33 cells lines derived from all typesof lung cancer. Eur J Hum Genet 1:156-163.

Kok K, van den Berg A, Veldhuis PMJF, van der Veen AY, Franke M, Schoenmakers EFPM, HulsbeekMMF, van der Hout AH, de Leij L, van de Ven W, Buys CHCM. 1994. A homozygous deletion in asmall cell lung cancer cell line involving a 3p21 region with a marked instability in yeast artificialchromosomes. Cancer Res 54:4183-4187.

Kok K, Naylor SL, Buys CHCM. 1997. Deletions of the short arm of chromosome 3 in solid tumors and thesearch for suppressor genes. Adv Cancer Res 71:27-92.

Kondo M, Ji L, Kamibayashi C, Tomizawa Y, Randle D, Sekido Y, Yokota J, Kashuba V, Zabarovsky E,Kuzmin I, Lerman M, Roth J, Minna JD. 2001. Overexpression of candidate tumor suppressorgene FUS1 isolated from the 3p21.3 homozygous deletion region leads to G1 arrest and growthinhibition of lung cancer cells. Oncogene 20:6258-6262.

Kuroki T, Trapasso F, Yendamuri S, Matsuyama A, Alder H, Williams NN, Kaiser LR, Croce CM. 2003.Allelic Loss on Chromosome 3p21.3 and Promoter Hypermethylation of Semaphorin 3B in Non-Small Cell Lung Cancer. Cancer Res 63:3352-3355.

Kusy S, Nasarre P, Chan D, Potiron V, Meyronet D, Gemmill RM, Constantin B, Drabkin HA, Roche J.2005a. Selective suppression of in vivo tumorigenicity by semaphorin SEMA3F in lung cancercells. Neoplasia 7:457-465.

Kusy S, Potiron V, Zeng C, Franklin W, Brambilla E, Minna J, Drabkin HA, Roche J. 2005. Promotercharacterization of Semaphorin SEMA3F, a tumor suppressor gene. Biochim Biophys Acta 1730:

Introduction

39

66-76.

Kuzmin I, Gillespie JW, Protopopov A, Geil L, Dreijerink K, Yang Y, Vocke CD, Duh FM, Zabarovsky E,Minna JD, Rhim JS, Emmert-Buck MR, Linehan WM, Lerman MI. 2002. The RASSF1A TumorSuppressor Gene Is Inactivated in Prostate Tumors and Suppresses Growth of ProstateCarcinoma Cells. Cancer Res 62:3498-3502.

Lantuejoul S, Constantin B, Drabkin H, Brambilla C, Roche J, Brambilla E. 2003. Expression of VEGF,semaphorin SEMA3F, and their common receptors neuropilins NP1 and NP2 in preinvasivebronchial lesions, lung tumours, and cell lines. J Pathol 200:336-347.

Latif F, Tory K, Modi WS, Graziano SL, Gamble G, Douglas J, Heppell Parton AC, Rabbitts PH, Zbar B,Lerman MI: Molecular characterization of a large homozygous deletion in the small cell lungcancer cell line U2020: a strategy for cloning the putative tumor suppressor gene. GenesChromosomes Cancer 5:119-127 (1992).

Lee MG, Kim HY, Byun DS, Lee SJ, Lee CH, Kim JI, Chang SG, Chi SG. 2001. Frequent epigeneticinactivation of rassf1a in human bladder carcinoma. Cancer Res 61:6688-6692.

Lepperdinger G, Strobl B, Kreil G. 1998. HYAL2, a human gene expressed in many cells, encodes alysosomal hyaluronidase with a novel type of specificity. J Biol Chem 273:22466-22470.

Lerman MI, Minna JD. 2000. The 630-kb lung cancer homozygous deletion region on human chromosome3p21.3: identification and evaluation of the resident candidate tumor suppressor genes. TheInternational Lung Cancer Chromosome 3p21.3 Tumor Suppressor Gene Consortium. CancerRes 60:6116-6133.

Li J, Wang F, Haraldson K, Protopopov A, Duh FM, Geil L, Kuzmin I, Minna JD, Stanbridge E, Braga E,Kashuba VI, Klein G, Lerman MI, Zabarovsky ER. 2004a. Functional characterization of thecandidate tumor suppressor gene NPRL2/G21 located in 3p21.3C. Cancer Res 64:6438-6443.

Li J, Wang F, Protopopov A, Malyukova A, Kashuba V, Minna JD, Lerman MI, Klein G, Zabarovsky E.204b. Inactivation of RASSF1C during in vivo tumor growth identifies it as a tumor suppressorgene. Oncogene 23:5941-5949.

Li Y, Rahmanian M, Widstrom C, Lepperdinger G, Frost GI, Heldin P. 2000. Irradiation-induced expressionof hyaluronan (HA) synthase 2 and hyaluronidase 2 genes in rat lung tissue accompanies activeturnover of HA and induction of types I and III collagen gene expression. Am J Respir Cell Mol Biol23:411-418.

Lim S, Yang MH, Park JH, Nojima T, Hashimoto H, Unni KK, Park YK. 2003. Inactivation of the RASSF1Ain osteosarcoma. Oncol Rep 10:897-901.

Lisitsyn NA, Lisitsina NM, Dalbagni G, Barker P, Sanchez CA, Gnarra J, Linehan WM, Reid BJ, WiglerMH. 1995. Comparative genomic analysis of tumors: detection of DNA losses and amplification.Proc Natl Acad Sci U S A 92:151-155.

Liu D, Pearlman E, Diaconu E, Guo K, Mori H, Haqqi T, Markowitz S, Willson J, Sy MS. 1996. Expressionof hyaluronidase by tumor cells induces angiogenesis in vivo. Proc Natl Acad Sci U S A 93:7832-7837.

Liu L, Amy V, Liu G, McKeehan WL. 2002. Novel complex integrating mitochondria and the microtubularcytoskeleton with chromosome remodeling and tumor suppressor RASSF1 deduced by in silicohomology analysis, interaction cloning in yeast, and colocalization in cultured cells. In Vitro CellDev Biol Anim 38:582-594.

Liu L, Tommasi S, Lee DH, Dammann R, Pfeifer GP. 2003a. Control of microtubule stability by theRASSF1A tumor suppressor. Oncogene 22:8125-8136.

Liu XQ, Chen HK, Zhang XS, Pan ZG, Li A, Feng QS, Long QX, Wang XZ, Zeng YX. 2003b. Alterations ofBLU, a candidate tumor suppressor gene on chromosome 3p21.3, in human nasopharyngealcarcinoma. Int J Cancer 106:60-65.

Chapter 1

40

Lokeshwar VB, Rubinowicz D, Schroeder GL, Forgacs E, Minna JD, Block NL, Nadji M, Lokeshwar BL.2001. Stromal and epithelial expression of tumor markers hyaluronic acid and hyal1 hyaluronidasein prostate cancer. J Biol Chem 276:11922-11932.

Lokeshwar VB, Schroeder GL, Carey RI, Soloway MS, Iida N. 2002. Regulation of hyaluronidase activityby alternative mRNA splicing. J Biol Chem 277:33654-33663.

Lokeshwar VB, Cerwinka WH, Isoyama T, Lokeshwar BL. 2005. HYAL1 hyaluronidase in prostate cancer:a tumor promoter and suppressor. Cancer Res 65:7782-7789.

Lokeshwar VB, Cerwinka WH, Lokeshwar BL. 2005. HYAL1 hyaluronidase: a molecular determinant ofbladder tumor growth and invasion. Cancer Res 65:2243-2250.

Lyons J, Landis CA, Harsh G, Vallar L, Grunewald K, Feichtinger H, Duh QY, Clark OH, Kawasaki E,Bourne HR. 1990. Two G protein oncogenes in human endocrine tumors. Science 249:655-659.

Marsit CJ, Kim DH, Liu M, Hinds PW, Wiencke JK, Nelson HH, Kelsey KT. 2005. Hypermethylation ofRASSF1A and BLU tumor suppressor genes in non-small cell lung cancer: implications fortobacco smoking during adolescence. Int J Cancer 114:219-223.

McBride WH, Bard JB. 1979. Hyaluronidase-sensitive halos around adherent cells. Their role in blockinglymphocyte-mediated cytolysis. J Exp Med 149:507-515.

Miao HQ, Soker S, Feiner L, Alonso JL, Raper JA, Klagsbrun M. 1999. Neuropilin-1 mediates collapsin-1/semaphorin III inhibition of endothelial cell motility: functional competition of collapsin-1 andvascular endothelial growth factor-165. J Cell Biol 146:233-242.

Mourtada-Maarabouni M, Sutherland LC, Meredith JM, Williams GT. 2003. Simultaneous acceleration ofthe cell cycle and suppression of apoptosis by splice variant delta-6 of the candidate tumoursuppressor LUCA-15/RBM5. Genes Cells 8:109-119.

Nasarre P, Constantin B, Rouhaud L, Harnois T, Raymond G, Drabkin HA, Bourmeyster N, Roche J.2003. Semaphorin SEMA3F and VEGF Have Opposing Effects on Cell Attachment andSpreading. Neoplasia 5:83-92.

Nasarre P, Kusy S, Constantin B, Castellani V, Drabkin HA, Bagnard D, Roche J. 2005. SemaphorinSEMA3F has a repulsing activity on breast cancer cells and inhibits E-cadherin-mediated celladhesion. Neoplasia 7:180-189.

Noble PW. 2002. Hyaluronan and its catabolic products in tissue injury and repair. Matrix Biol 21:25-29.

Ochi K, Mori T, Toyama Y, Nakamura Y, Arakawa H. 2002. Identification of semaphorin3B as a directtarget of p53. Neoplasia 4:82-87.

Oh JJ, West AR, Fishbein MC, Slamon DJ. 2002. A candidate tumor suppressor gene, H37, from thehuman lung cancer tumor suppressor locus 3p21.3. Cancer Res 62:3207-3213.

Ohta M, Inoue H, Cotticelli MG, Kastury K, Baffa R, Palazzo J, Siprashvili Z, Mori M, Mccue P, Druck T,Croce CM, Huebner K. 1996. The FHIT gene, spanning the chromosome 3p14.2 fragile site acidrenal carcinoma-associated t(3;8) breakpoint, is abnormal in digestive tract cancers. Cell 84:587-597.

Ortiz-Vega S, Khokhlatchev A, Nedwidek M, Zhang XF, Dammann R, Pfeifer GP, Avruch J. 2002. Theputative tumor suppressor RASSF1A homodimerizes and heterodimerizes with the Ras-GTPbinding protein Nore1. Oncogene 21:1381-1390.

Pace AM, Wong YH, Bourne HR. 1991. A mutant alpha subunit of Gi2 induces neoplastic transformationof Rat-1 cells. Proc Natl Acad Sci U S A 88:7031-7035.

Pan ZG, Kashuba VI, Liu XQ, Shao JY, Zhang RH, Jiang JH, Guo C, Zabarovsky E, Ernberg I, Zeng YX.2005. High Frequency Somatic Mutations in RASSF1A in Nasopharyngeal Carcinoma. CancerBiol Ther 4.

Introduction

41

Patel S, Turner PR, Stubberfield C, Barry E, Rohlff CR, Stamps A, Tyson K, Terrett J, Box G, Eccles S,Page MJ. 2002. Hyaluronidase gene profiling and role of hyal-1 overexpression in an orthotopicmodel of prostate cancer. Int J Cancer 97:416-424.

Pawlowski A, Haberman HF, Menon IA. 1979. The effects of hyalurodinase upon tumor formation inBALB/c mice painted with 7,12-dimethylbenz-(a)anthracene. Int J Cancer 23:105-109.

Pirinen R, Tammi R, Tammi M, Hirvikoski P, Parkkinen JJ, Johansson R, Bohm J, Hollmen S, Kosma VM.2001. Prognostic value of hyaluronan expression in non-small-cell lung cancer: Increased stromalexpression indicates unfavorable outcome in patients with adenocarcinoma. Int J Cancer 95:12-17.

Pitha-Rowe I, Petty WJ, Feng Q, Koza-Taylor PH, Dimattia DA, Pinder L, Dragnev KH, Memoli N, MemoliV, Turi T, Beebe J, Kitareewan S, Dmitrovsky E. 2004. Microarray analyses uncover UBE1L as acandidate target gene for lung cancer chemoprevention. Cancer Res 64:8109-8115.

Praskova M, Khoklatchev A, Ortiz-Vega S, Avruch J. 2004. Regulation of MST1 kinase byautophosphorylation, by the growth inhibitory proteins, RASSF1 and NORE1 and by Ras. BiochemJ 381:453-462.

Qiu GH, Tan LK, Loh KS, Lim CY, Srivastava G, Tsai ST, Tsao SW, Tao Q. 2004. The candidate tumorsuppressor gene BLU, located at the commonly deleted region 3p21.3, is an E2F-regulated,stress-responsive gene and inactivated by both epigenetic and genetic mechanisms innasopharyngeal carcinoma. Oncogene 23:4793-4806.

Rabbitts P, Bergh J, Douglas J, Collins F, Waters J. 1990. A submicroscopic homozygous deletion at theD3S3 locus in a cell line isolated from a small cell lung carcinoma. Genes Chromosomes Cancer2:231-238.

Rabizadeh S, Xavier RJ, Ishiguro K, Bernabeortiz J, Lopez-Ilasaca M, Khokhlatchev A, Mollahan P, PfeiferGP, Avruch J, Seed B. 2004. The scaffold protein CNK1 interacts with the tumor suppressorRASSF1A and augments RASSF1A-induced cell death. J Biol Chem 279:29247-29254.

Rai SK, Duh FM, Vigdorovich V, Danilkovitch-Miagkova A, Lerman MI, Miller AD. 2001. Candidate tumorsuppressor HYAL2 is a glycosylphosphatidylinositol (GPI)-anchored cell-surface receptor forjaagsiekte sheep retrovirus, the envelope protein of which mediates oncogenic transformation.Proc Natl Acad Sci U S A 98:4443-4448.

Ramesh R, Saeki T, Templeton NS, Ji L, Stephens LC, Ito I, Wilson DR, Wu Z, Branch CD, Minna JD,Roth JA. 2001. Successful treatment of primary and disseminated human lung cancers bysystemic delivery of tumor suppressor genes using an improved liposome vector. Mol Ther 3:337-350.

Rintala-Maki ND, Abrasonis V, Burd M, Sutherland LC. 2004. Genetic instability of RBM5/LUCA-15/H37 inMCF-7 breast carcinoma sublines may affect susceptibility to apoptosis. Cell Biochem Funct22:307-313.

Rintala-Maki ND, Sutherland LC. 2004. LUCA-15/RBM5, a putative tumour suppressor, enhances multiplereceptor-initiated death signals. Apoptosis 9:475-484.

Roche J, Boldog F, Robinson M, Robinson L, VarellaGarcia M, Swanton M, Waggoner B, Fishel R,Franklin W, Gemmill R, Drabkin H. 1996. Distinct 3p21.3 deletions in lung cancer and identificationof a new human semaphorin. Oncogene 12:289-1297.

Rohm B, Ottemeyer A, Lohrum M, Puschel AW. 2000. Plexin/neuropilin complexes mediate repulsion bythe axonal guidance signal semaphorin 3A. Mech Dev 93:95-104.

Rooney P, Kumar S, Ponting J, Wang M. 1995. The role of hyaluronan in tumour neovascularization(review). Int J Cancer 60:632-636.

Roz L, Gramegna M, Ishii H, Croce CM, Sozzi G. 2002. Restoration of fragile histidine triad (FHIT)expression induces apoptosis and suppresses tumorigenicity in lung and cervical cancer cell lines.

Chapter 1

42

Proc Natl Acad Sci U S A 99:3615-3620.

Rudolph U, Finegold MJ, Rich SS, Harriman GR, Srinivasan Y, Brabet P, Boulay G, Bradley A,Birnbaumer L. 1995. Ulcerative colitis and adenocarcinoma of the colon in G alpha i2-deficientmice. Nat Genet 10:143-150.

Sandy P, Gostissa M, Fogal V, Cecco LD, Szalay K, Rooney RJ, Schneider C, Del Sal G. 2000. p53 isinvolved in the p120E4F-mediated growth arrest. Oncogene 19:188-199.

Satoh H, Lamb PW, Dong JT, Everitt J, Boreiko C, Oshimura M, Barrett JC. 1993. Suppression oftumorigenicity of A549 lung adenocarcinoma cells by human chromosomes 3 and 11 introducedvia microcell-mediated chromosome transfer. Mol Carcinog 7:157-164.

Schagdarsurengin U, Gimm O, Hoang-Vu C, Dralle H, Pfeifer GP, Dammann R. 2002. FrequentEpigenetic Silencing of the CpG Island Promoter of RASSF1A in Thyroid Carcinoma. Cancer Res62:3698-3701.

Schagdarsurengin U, Seidel C, Ulbrich EJ, Kolbl H, Dittmer J, Dammann R. 2005. A polymorphism atcodon 133 of the tumor suppressor RASSF1A is associated with tumorous alteration of the breast.Int J Oncol 27:185-19.

Senchenko VN, Liu J, Loginov W, Bazov I, Angeloni D, Seryogin Y, Ermilova V, Kazubskaya T,Garkavtseva R, Zabarovska VI, Kashuba VI, Kisselev LL, Minna JD, Lerman MI, Klein G, BragaEA, Zabarovsky ER. 2004. Discovery of frequent homozygous deletions in chromosome 3p21.3LUCA and AP20 regions in renal, lung and breast carcinomas. Oncogene 23:5719-5728.

Shivakumar L, Minna J, Sakamaki T, Pestell R, White MA. 2002. The RASSF1A tumor suppressor blockscell cycle progression and inhibits cyclin D1 accumulation. Mol Cell Biol 22:4309-4318.

Shuster S, Frost GI, Csoka AB, Formby B, Stern R. 2002. Hyaluronidase reduces human breast cancerxenografts in SCID mice. Int J Cancer 102:192-197.

Shuttleworth TL, Wilson MD, Wicklow BA, Wilkins JA, Triggs-Raine BL. 2002. Characterization of themurine hyaluronidase gene region reveals complex organization and cotranscription of Hyal1 withdownstream genes, Fus2 and Hyal3. J Biol Chem 277:23008-23018.

Slevin M, Krupinski J, Kumar S, Gaffney J. 1998. Angiogenic oligosaccharides of hyaluronan induceprotein tyrosine kinase activity in endothelial cells and activate a cytoplasmic signal transductionpathway resulting in proliferation. Lab Invest 78:987-1003.

Song MS, Song SJ, Ayad NG, Chang JS, Lee JH, Hong HK, Lee H, Choi N, Kim J, Kim H, Kim JW, ChoiEJ, Kirschner MW, Lim DS. 2004. The tumour suppressor RASSF1A regulates mitosis byinhibiting the APC-Cdc20 complex. Nat Cell Biol 6:129-137.

Sozzi G, Veronese ML, Negrini M, Baffa R, Cotticelli MG, Inoue H, Tornielli S, Pilotti S, DeGregorio L,Pastorino U, Pierotti MA, Ohta M, Huebner K, Croce CM. 1996. The FHIT gene at 3p14.2 isabnormal in lung cancer. Cell 85:17-26.

Sozzi G, Pastorino U, Moiraghi L, Tagliabue E, Pezzella F, Ghirelli C, Tornielli S, Sard L, Huebner K,Pierotti MA, Croce CM, Pilotti S. 1998. Loss of FHIT function in lung cancer and preinvasivebronchial lesions. Cancer Res 58:5032-5037.

Sundaresan V, Chung G, Heppell Parton A, Xiong J, Grundy C, Roberts I, James L, Cahn A, Bench A,Douglas J, Minna J, Sekido Y, Lerman M, Latif F, Bergh J, Li H, Lowe N, Ogilvie D, Rabbitts P:.1998. Homozygous deletions at 3p12 in breast and lung cancer. Oncogene 17:1723-1729.

Sutherland LC, Rintala-Maki ND, White RD, Morin CD. 2005. RNA binding motif (RBM) proteins: a novelfamily of apoptosis modulators? J Cell Biochem 94:5-24.

Suzuki M, Sunaga N, Shames DS, Toyooka S, Gazdar AF, Minna JD. 2004. RNA Interference-MediatedKnockdown of DNA Methyltransferase 1 Leads to Promoter Demethylation and Gene Re-Expression in Human Lung and Breast Cancer Cells. Cancer Res 64:3137-3143.

Introduction

43

Szeles A, Yang Y, Sandlund AM, Kholodnyuk I, Kiss H, KostAlimova M, Zabarovsky ER, Stanbridge E,Klein G, Imreh S. 1997. Human/mouse microcell hybrid based elimination test reduces theputative tumor suppressor regon at 3p21.3 to 1.6 cM. Genes Chromosomes Cancer 20:329-336.

Takahashi T, Fournier A, Nakamura F, Wang LH, Murakami Y, Kalb RG, Fujisawa H, Strittmatter SM.1999. Plexin-neuropilin-1 complexes form functional semaphorin-3A receptors. Cell 99:59-69.

Tamagnone L, Artigiani S, Chen H, He Z, Ming GI, Song H, Chedotal A, Winberg ML, Goodman CS, PooM, Tessier-Lavigne M, Comoglio PM. 1999. Plexins are a large family of receptors fortransmembrane, secreted, and GPI-anchored semaphorins in vertebrates. Cell 99:71-80.

Tammi MI, Day AJ, Turley EA. 2002. Hyaluronan and homeostasis: a balancing act. J Biol Chem277:4581-4584.

Termeer CC, Hennies J, Voith U, Ahrens T, Weiss JM, Prehm P, Simon JC. 2002. Oligosaccharides ofhyaluronan are potent activators of dendritic cells. J Immunol 165:1863-1870.

Termeer C, Sleeman JP, Simon JC. 2003. Hyaluronan--magic glue for the regulation of the immuneresponse? Trends Immunol 24:112-114.

Timmer T, Terpstra P, van den Berg A, Veldhuis PMJF, Ter Elst A, Voutsinas G, Hulsbeek MMF, DraaijersTG, Looman MWG, Kok K, Naylor SL, Buys CHCM. 1999. A comparison of genomic structuresand expression patterns of two closely related flanking genes in a critical lung cancer region3p21.3. Eur J Hum Genet 7:478-486.

Todd MC, Xiang RH, Garcia DK, Kerbacher KE, Moore SL, Hensel CH, Liu P, Siciliano MJ, Kok K, vanden Berg A, Veldhuis P, Buys CHCM, Killary AM, Naylor SL, Buys CHCM: An 80 Kb P1 clone fromchromosome 3p21.3 suppresses tumor growth in vivo. Oncogene 13:2387-2396 (1996).

Todd S, Franklin WA, Varella Garcia M, Kennedy T, Hilliker CE, Jr., Hahner L, Anderson M, Wiest JS,Drabkin HA, Gemmill RM. 1997. Homozygous deletions of human chromosome 3p in lung tumors.Cancer Res 57:1344-1352.

Tokuchi Y, Kobayashi Y, Hayashi S, Hayashi M, Tanimoto K, Hashimoto T, Nishida K, Ishikawa Y,Nakagawa K, Satoh Y, Yamamoto M, Tsuchiya E. 1999. Abnormal FHIT transcripts found in bothlung cancer and normal lung tissue. Genes Chromosomes Cancer 24:105-111.

Tomizawa Y, Sekido Y, Kondo M, Gao B, Yokota J, Roche J, Drabkin H, Lerman MI, Gazdar AF, MinnaJD. 2001. Inhibition of lung cancer cell growth and induction of apoptosis after reexpression of3p21.3 candidate tumor suppressor gene SEMA3B. Proc Natl Acad Sci U S A 98:13954-13959.

Tomizawa Y, Iijima H, Nomoto T, Iwasaki Y, Otani Y, Tsuchiya S, Saito R, Dobashi K, Nakajima T, MoriM:. 2004. Clinicopathological significance of aberrant methylation of RARbeta2 at 3p24, RASSF1Aat 3p21.3, and FHIT at 3p14.2 in patients with non-small cell lung cancer. Lung Cancer 46:305-312.

Tommasi S, Dammann R, Zhang Z, Wang Y, Liu L, Tsark WM, Wilczynski SP, Li J, You M, Pfeifer GP:.2005. Tumor susceptibility of Rassf1a knockout mice. Cancer Res 65:92-98.

Trochon V, Mabilat-Pragnon C, Bertrand P, Legrand Y, Soria J, Soria C, Delpech B, Lu H. 1997.Hyaluronectin blocks the stimulatory effect of hyaluronan-derived fragments on endothelial cellsduring angiogenesis in vitro. FEBS Lett 418:6-10.

Tse C, Xiang RH, Bracht T, Naylor SL. 2002. Human Semaphorin 3B (SEMA3B) located at chromosome3p21.3 suppresses tumor formation in an adenocarcinoma cell line. Cancer Res 62 :542-546.

Tseng JE, Kemp BL, Khuri FR, Kurie JM, Lee JS, Zhou X, Liu D, Hong WK, Mao L. 1999. Loss of Fhit isfrequent in stage I non-small cell lung cancer and in the lungs of chronic smokers. Cancer Res59:4798-4803.

Turley EA, Noble PW, Bourguignon LY. 2002. Signaling properties of hyaluronan receptors. J Biol Chem277:4589-4592.

Chapter 1

44

Uno F, Sasaki J, Nishizaki M, Carboni G, Xu K, Atkinson EN, Kondo M, Minna JD, Roth JA, Ji L. 2004.Myristoylation of the fus1 protein is required for tumor suppression in human lung cancer cells.Cancer Res 64:2969-2976.

van den Berg A, Draaijers TG, Kok K, Timmer T, van der Veen AY, Veldhuis PM, de Leij LFMH, GerhartzCD, Naylor SL, Smith DI, Buys CHCM. 1997. Normal FHIT transcripts in renal cell cancer- andlung cancer-derived cell lines, including a cell line with a homozygous deletion in the FRA3Bregion. Genes Chromosomes Cancer 19:220-227.

van der Weyden L, Adams DJ, Harris LW, Tannahill D, Arends MJ, Bradley A. 2005. Null and conditionalsemaphorin 3B alleles using a flexible puroDeltatk loxP/FRT vector. Genesis 41:171-178.

Vigdorovich V, Strong RK, Miller AD. 2005. Expression and characterization of a soluble, active form ofthe jaagsiekte sheep retrovirus receptor, Hyal2. J Virol 79:79-86.

Vos MD, Martinez A, Elam C, Dallol A, Taylor BJ, Latif F, Clark GJ. 2004. A role for the RASSF1A tumorsuppressor in the regulation of tubulin polymerization and genomic stability. Cancer Res 64:4244-4250.

Wei MH, Latif F, Bader S, Kashuba V, Chen JY, Duh FM, Sekido Y, Lee CC, Geil L, Kuzmin I, ZabarovskyE, Klein G, Zbar B, Minna JD, Lerman MI. 1996. Construction of a 600-kilobase cosmid clonecontig and generation of a transcriptional map surrounding the lung cancer tumor suppressor gene(TSG) locus on human chromosome 3p21.3: Progress toward the isolation of a lung cancer TSG.Cancer Res 56:1487-1492.

Whang YM, Kim YH, Kim JS, Yoo YD. 2005. RASSF1A suppresses the c-Jun-NH2-kinase pathway andinhibits cell cycle progression. Cancer Res 65:3682-3690.

Wilkinson CR, Bower LM, Warren C. 1996. The relationship between hyaluronidase activity and hyaluronicacid concentration in sera from normal controls and from patients with disseminated neoplasm.Clin Chim Acta 256:165-173.

Wistuba II, Lam S, Behrens C, Virmani AK, Fong KM, LeRiche J, Samet JM, Srivastava S, Minna JD,Gazdar AF. 1997. Molecular damage in the bronchial epithelium of current and former smokers. JNatl Cancer Inst 89:1366-1373.

Xian J, Clark KJ, Fordham R, Pannell R, Rabbitts TH, Rabbitts PH. 2001. Inadequate lung developmentand bronchial hyperplasia in mice with a targeted deletion in the Dutt1/Robo1 gene. Proc NatlAcad Sci U S A 98:15062-15066.

Xian J, Aitchison A, Bobrow L, Corbett G, Pannell R, Rabbitts T, Rabbitts P. 2004. Targeted disruption ofthe 3p12 gene, Dutt1/Robo1, predisposes mice to lung adenocarcinomas and lymphomas withmethylation of the gene promoter. Cancer Res 64:6432-6437.

Xiang R, Davalos AR, Hensel CH, Zhou XJ, Tse C, Naylor SL. 2002. Semaphorin 3F gene from human3p21.3 suppresses tumor formation in nude mice. Cancer Res 62:2637-2643.

Xu H, Ito T, Tawada A, Maeda H, Yamanokuchi H, Isahara K, Yoshida K, Uchiyama Y, Asari A. 2002.Effect of hyaluronan oligosaccharides on the expression of heat shock protein 72. J Biol Chem277:17308 -17314.

Yang Y, Kiss H, Kost-Alimova M, Kedra D, Fransson I, Seroussi E, Li J, Szeles A, Kholodnyuk I, ImrehMP, Fodor K, Hadlaczky G, Klein G, Dumanski JP, Imreh S. 1999. A 1-Mb PAC contig spanningthe common eliminated region 1 (CER1) in microcell hybrid-derived SCID tumors. Genomics62:147-155.

Yang Y, Li J, Szeles A, Imreh MP, Kost-Alimova M, Kiss H, Kholodnyuk I, Fedorova L, Darai E, Klein G,Imreh S. 2004. Consistent downregulation of human lactoferrin gene, in the common eliminatedregion 1 on 3p21.3, following tumor growth in severe combined immunodeficient (SCID) mice.Cancer Lett 191:155-164.

Yi Lo PH, Chung Leung AC, Xiong W, Law S, Duh FM, Lerman MI, Stanbridge EJ, Lung ML. 2005.

Introduction

45

Expression of candidate chromosome 3p21.3 tumor suppressor genes and down-regulation ofBLU in some esophageal squamous cell carcinomas. Cancer Lett 7.

Yoon JH, Dammann R, Pfeifer GP. 2001. Hypermethylation of the CpG island of the RASSF1A gene inovarian and renal cell carcinomas. Int J Cancer 94:212-217.

Yu MY, Tong JH, Chan PK, Lee TL, Chan MW, Chan AW, Lo KW, To KF. 2003. Hypermethylation of thetumor suppressor gene RASSFIA and frequent concomitant loss of heterozygosity at 3p21 incervical cancers. Int J Cancer 105:204-209.

Zabarovsky ER, Lerman MI, Minna JD. 2002. Tumor suppressor genes on chromosome 3p involved in thepathogenesis of lung and other cancers. Oncogene 21:6915-6935.

Zhang L, Underhill CB, Chen L. 1995. Hyaluronan on the surface of tumor cells is correlated withmetastatic behavior. Cancer Res 55:428-433.

Zochbauer-Muller S, Wistuba II, Minna JD, Gazdar AF. 2000. Fragile histidine triad (FHIT) geneabnormalities in lung cancer. Clin Lung Cancer 2:141-145.

Chapter 2

Transfection of a PAC contig covering the lungcancer critical region at 3p21.3 discloses the

complexity of functional analysis ofhomozygous deletion regions

Arja ter ElstBea E. HiemstraPieter van der VliesWytske KammingaAnneke Y. van der VeenInge DavelaarGerard J. te MeermanFrans GerbensKlaas KokCharles H.C.M. Buys

Department of Medical Genetics, University Medical Center Groningen, Groningen,The Netherlands

ABSTRACT

A 3p21.3 region containing 19 genes is considered as a lung cancer critical regionwith tumour suppressor activity. We covered this region with overlapping P1 artificialchromosomes (PAC) and used these to transfect cells from an SCLC cell line whichreadily causes tumours in nude mice. We assume that in the PACs genes will mostlybe accompanied by their own promoter sequences. Per PAC we selected twodifferent transfectants with a low PAC copy number (1-6 copies) integrated at asingle genomic site. All integration sites appeared to be different. The selectedtransfectants were used in tumourigenicity tests of nude mice to investigate whetherthe integrated genes suppressed the tumour-inducing capacity of the original SCLCcell line. Two transfectants, each containing the same PAC from the centromeric partof the critical region, caused tumours significantly smaller than those caused bytransfectants containing other PACs and those caused by the parental SCLC cellline. We could demonstrate the occurrence in transfectants of PAC-specific geneexpression. Therefore, the reduced tumour growth can be attributed to theintroduced PAC gene content. Still, an alternative mechanism, explaining that one ofthe transfectants with only vector sequences integrated caused smaller tumours too,can also apply. In a large proportion of transfectants, namely, transfection andintegration of exogenous DNA appeared to induce a genome-wide instability whichmight involve chromosomal rearrangements interfering with cell growth andproliferation. The cellular phenotype resulting from chromosomal instability can bediverse. Therefore, it is possible that both newly introduced genes or regulatorysequences and chromosomal instability act as underlying mechanisms causing areduced tumour growth.

Chapter 2

48

INTRODUCTION

Deletions of the short arm of chromosome 3 are a most common abnormality in lungcancer. They have been reported to occur in approximately 75% of non small celllung cancer (NSCLC) tumours and in up to 100% of small cell lung cancer (SCLC)tumours (Kok et al., 1997; Zabarovsky et al., 2002). Such deletions have also beenfound in histologically normal tissue of about 50% of smokers and former smokers,not in control individuals (Wistuba et al., 1997). This suggests that losses at the shortarm of chromosome 3 represent an early chromosomal change in the developmentof lung tumours. For the 3p21.3 region, overlapping homozygous deletions havebeen found in three different lung cancer cell lines (Daly et al., 1993; Kok et al.,1994; Roche et al., 1996). The smallest region of overlap, the 3p21.3 critical region,contains 19 genes (Wei et al., 1996; Lerman and Minna, 2000). Virtually all thesegenes have been suggested to act as tumour suppressor genes (Hermouet et al.,1996; Ji et al., 2002; Oh et al., 2002; Tse et al., 2002; Xiang et al., 2002; Chow et al.,2004; Li et al., 2004). This has generally been based on the results of transfectingtumour cells with the cDNAs of these genes and injecting the transfected cellssubcutaneously into nude mice. It appeared that tumour growth became inhibitedcompared to the tumour growth caused by non-transfected tumour cells. The cDNAshave, however, mostly been cloned behind strong viral promoters that cause a greatoverexpression easily resulting in disturbing cellular processes to such an extent thatcell growth is severely affected. Instead of such cDNA constructs, we wanted to usePACs containing one or several of the genes that in the PAC will most likely beaccompanied by their own promoter or some other regulatory sequences. Bytransfecting each PAC into an SCLC cell line and subsequently testing thetransfected cells (transfectants) for tumourigenicity, we wanted to better define thelocation of the presumed tumour suppressing activity in the critical region. If suchactivity is present in normal lung cells, then it may be assumed that it will have beeninactivated in every SCLC tumour or cell line. For transfection experiments weselected in this study the SCLC cell line GLC45, which readily causes tumourdevelopment in nude mice and is growing attached to the bottom of culture flasks,thus greatly facilitating transfection. Like most SCLC cell lines, GLC45 has thegreater part of the short arm of chromosome 3 heterozygously deleted.

Complexity of functional analysis

49

MATERIAL AND METHODS

Cell lineThe SCLC cell line GLC45 (Xiang et al., 2002) was cultured in RPMI-1640 (GIBCO -Invitrogen Inc, Carlsbad, Ca, USA) supplemented with 15% foetal calf serum(BioWhittaker – Cambrex, Verviers, Belgium) and 0.1 mM sodium pyruvate, 0.1 mMHEPES, penicillin/streptomycin, glutamine and 250 µg/ml amphotericin B (all fromGIBCO – Invitrogen) at 37° C in a 5% CO2 incubator.

PAC library screeningBy hybridising different PCR products of cDNA sequences on the human femalePAC library RPCI-6 (Genome Technology Centre Leiden, The Netherlands), eightoverlapping PACs were obtained. The vector used for construction of the RCPI-6library, pPAC4, contains a blasticidin-S-methylase resistance marker for selection oftransfected cells.

Isolation of PAC insert ends. The PAC insert ends were obtained byvectorette-PCR (Riley et al., 1990). The different specific PAC vector primers usedwere pPAC4-A: 5’-ATGTTCATGTTCATGTCTCC and pPAC4-B: 5’-GGGTTGAAGGCTCTCAAGG; or sp6: 5’-ATTTAGGTGACACTATAG and T7: 5’-TAATACGACTCACTATAGGGAGA, in combination with the vectorette box primer224: 5’-CTCTAGATTCGGATCCTACGAGAATCGCT. The PCR products obtainedwere sequenced and insert-specific PCR primers were designed.

Transfection and selection of stable transfectantsGLC45 cells were passaged into 6-wells plates to a confluency of 60%-80%. Fortransfection, 2 µg of PAC DNA and 10 µg of DAC30 liposomal formulation(Eurogentec, Seraing, Belgium) were added using the protocol provided by themanufacturer. After 48 h, the cells were diluted 1:8 into RPMI-1640 supplementedwith blasticidin (final concentration 5 µg/ml) (Calbiochem, San Diego, Ca, USA).Clones still growing after 4-6 weeks were transferred to a 6-wells plate. After approx.8 weeks, clones were checked by FISH and PCR analyses.

FISH analysis of transfectantsFor metaphase preparation, cell cultures were treated with colcemid (GIBCO-Invitrogen) for 3-4 h (final concentration of 0.1 µg/ml). After aspirating the medium, a

Chapter 2

50

hypotonic shock was applied by adding an 83 mM KCl solution and leaving the flasksat 37° C for 30 min. Fixation was carried out with methanol-acetic acid (3:1, v/v).Slides were air-dried and stored at -20°C until use.

Probes were labelled by standard nick-translation with either biotin-16-dUTPor digoxigenin-11-dUTP. Metaphase spreads were pre-treated with RNase A(0.1mg/ml 2 × SSC) at 37°C for one hour, with pepsine (9.24 mA/100ml 1%HCl) at37°C for 10 minutes, with formaldehyde (2.7% in 1× PBS/MgCl2) at roomtemperature for 10 minutes and dehydrated in ethanol. Probes and slides were co-denatured in a PTC200 (Peltier Thermal Cycler; MJ Research, Waltham, MA) at 73ºC for 3 min. Hybridisation was carried out in a humidified chamber (70% formamide)at 37°C overnight. Excess probe was subsequently washed twice in 2 × SSC/0.1%Tween-20 at 37°C for 3 min, once in 0.1× SSC 60°C for 5 min and finally in 0.1×SSC at room temperature.

Immunodetection for PACs was accomplished by incubation with avidin-FITC(Zymed Invitrogen Inc., Carlsbad, CA), biotinylated-anti-avidin (Kordia, Leiden, TheNetherlands), and again avidin-FITC for 20 min at room temperature. Between theseincubations, the slides were washed three times in 4× SSC/0.05% Tween-20 at roomtemperature for 3 min. Afterwards slides were washed twice in 4× SSC/0.05%Tween-20 at room temperature for 3 min, once in 1 × PBS at room temperature for 5min and subsequently dehydrated in ethanol. Fluorochromes were diluted in blockingreagent for ELISA (5mg/ml 4 × SSC/0.05% Tween-20).

Immunodetection of whole chromosome paint were accomplished byincubation with streptavidin-Texas-Red (Invitrogen-Molecular Probes)/mousemonoclonal-anti-digoxygenin (Sigma), biotinylated-anti-streptavidin (Kordia)/rabbit-anti-mouse-IgG-digoxygenin (Sigma) and streptavidin-Texas-Red/goat-anti-rabbit-IgG-FITC (Sigma) at 37°C for 30 min. Between these incubations, the slides werewashed three times in 1× Tris NaCl/0.05% Tween-20 at room temperature for 3 min.Afterwards slides were washed twice in 1× Tris NaCl/0.05% Tween-20 at roomtemperature for 3 min, once in 1x PBS at room temperature for 5 min and thendehydrated in ethanol. Fluorochromes were diluted in blocking reagent (5mg/ml 1×Tris NACl/0.05% Tween-20). Slides were embedded in Vectashield : Vectashieldwith DAPI (1.5 µg/ml) (1:1, v/v).

Complexity of functional analysis

51

Selection criteria for transfectantsSingle clones were subjected to FISH analysis and pure clones were analysed forthe number of integration sites. The number of integrated PAC copies at theintegration site was estimated from comparison of the signal intensity at that site withthe intensity from the endogenous signal on the derivative chromosome 3, which hadalready been duplicated in the parental cell line. Transfectants with a low signalintensity at the integration site, indicative of integration of only a single PAC or veryfew copies, were selected for further experiments. For tumourigenicity tests, weselected two transfectants containing the PAC for five of the PACs overlapping theregion homozygously deleted in GLC20. In all but one case, these transfectants hada single site of integration and a PAC signal at that site showing an intensitycomparable to that of the signal caused by the endogenous sequence. The twotransfectants per PAC are designated transfectant-1 and transfectant-2. For twocoinciding PACs, only a single transfectant meeting the criteria could be found.These two transfectants were also considered as transfectant-1 and transfectant-2,as they contain more or less the same genomic sequence. For one PAC, no clonesmeeting the selection criteria could be obtained in two transfection experiments.

Southern blottingTo better determine the number of PAC copies at the integrated site we producedSouthern blots. Serial dilutions were made of genomic DNA (van der Hout et al.,1989) from the parental cell line and the derived transfectants and digested with PstI(Amersham Biosciences, Little Chalfont, Buckinghamshire, UK) (final concentration1U/µl) for approx. 6 h. Fragments were separated on a 0.8% gel for approx. 15 h andblotted on a nitro-cellulose filter for approx. 6 h. Purified PCR products from a geneof interest as well as from a control gene (see below) were radioactively labelled with[32P]dCTP by random priming using a random hexamer primer p(dN)6 (AmershamBiosciences). Both probes were hybridised together to the filter. After washing at highstringency, the filter was exposed to X-ray film. The film was developed and theapproximate level of amplification for each gene was calculated by comparison of thehybridisation signal intensities (on the X-ray films) of the labelled DNA from thetransfectants and the parental cell line, GLC45. To compensate for minor variationsin sample loading, signals from a control gene originating from the regionhomozygously deleted in GLC20 but not located on the integrated PAC were usedfor normalisation.

Chapter 2

52

Table 1. Primers used for PAC integration analysis of transfectants.

Primer pair Official genesymbol

Forward primer sequence5’-3’

Reverse primer sequence5’-3’

BLASTa Selection marker GCAATTCGTACGAAAACAGG ATCAACTCCCTACACATACC

D3S2968b CA-repeat GAAGCCTTTGTCAGAAGGAG AATGGAATTAGGAAACAGGACC

LUCA17a CA-repeat GTGGGTGGATCATCAGGTC GTGCCTACAATAGAGAGGGTA

LUCA11a CA-repeat GATCTCTGCCCAGGTTGAAG CGTTCATGCCTACACACCAC

LUCA13a TA-repeat TGGGGTCTTGCATACATTGC AGCCAAGACTGCTCCACTGT

LUCA19b AGAA-repeat GAGCCCAGTGCACTCCAG TTTCTCTCTCTTCCCTCATTCCT

LUCA22a CA-repeat TTGATCCTGACAACCCCAAT GGAAGGCATAACAGCCATTT

ISSCP4Ba RBM5 TATTGTTTGGGCTTTGGCTC TATCCTCTTCCTTGTGGCAC

ISSCP14a RBM5 ATAGGGGCACTAATTGTGGG GTCACGTCTGTGAATCTTGG

Hsema2a SEMA3F CTCTGTACTGGGAGAAGACC GAAGCCCAGAGAAGAAGACC

GNAI2gen1a GNAI2 GCATCTGCAACAACAAGTGG CGTCAAACACGACTGCACG

G17gen1a SLC38A3 CACAGGCAACTTCAGCCACG GACAGCGATGGACAGGTTGG

FUS1gen1a NAT6 CAATCGTCACCAAGAACGGG GGATCACAGGGAAATCCACG

101F6-5UTRa CYB561D2 GGAGGAAACCACCGCATCAGA CTCCGCAGAAAGGGCCATCG

BLU-3UTRa ZMYND10 TTGTGTCCTGGCAGCCCAGG GCTAGAGGAAGAGGGTCCCC

RASSF1-3UTRa RASSF1 CCCTTGGGTGACCTCTTGTA ATGCCTGCCTTATTCTGAGC

CACNA2D2-3UTRa CACNA2D2 GCAGCCTCAAGTCCTCGTCC TGCCAACCCTCCCACCAACC

PCYPAC2a pPAC4 vector-end GATGTTCATGTTCATGTCTCC GGGTTGAAGGCTCTCAAGG

76L5T7a PAC 76L5-end CTGTGAAACCTACACTACCC GGGTTGAAGGCTCTCAAGG

211D14Aa PAC 211D14-end TCCGGCTACTGATGTGGTGG GATGTTCATGTTCATGTCTCC

211D14Ba PAC 211D14-end GATGACATCACTGCAAACCC GGGTTGAAGGCTCTCAAGG

143C21Aa PAC 143C21-end TGCAACCTCCGCCTCTGGG GATGTTCATGTTCATGTCTCC

143C21Ba PAC 143C21-end GTAGCTCTGACTCCTGTCCC GGGTTGAAGGCTCTCAAGG

234N4Aa PAC 234N4-end GCTACTGTAAACATCCCAGC GATGTTCATGTTCATGTCTCC

234N4Ba PAC 234N4-end GTACAAGAGGTCACCCAAGG GGGTTGAAGGCTCTCAAGG

175O16Aa PAC175O16-end AATCGCTGTCCTCTCCTTCC GATGTTCATGTTCATGTCTCC

175O16Ba PAC175O16-end CTGGGTTCAAGTGATTCTCC GGGTTGAAGGCTCTCAAGG

134H6Ba PAC134H6-end GCCTTCCGCAGTGTTTGTG GGGTTGAAGGCTCTCAAGG

157G3T7a PAC 157G3-end CCCAGTCTCACCTTTTACC GGGTTGAAGGCTCTCAAGG

157G3SP6a PAC 157G3-end ACCCTGCAAACAGAACAGG GATGTTCATGTTCATGTCTCC

185A4-2B2a PAC 185A-end CCGGAAGCGCATCAGAAAG GGGTTGAAGGCTCTCAAGG

185A4SP62a PAC 185A4-end CTAGAGGATAGTCAGACCC GATGTTCATGTTCATGTCTCC

a This studyb The human genome database (GDB)c (Frengen et al., 2000)

Complexity of functional analysis

53

PCR analysis of transfectantsSince both the transfected PAC DNA and the endogenous GLC45 DNA are ofhuman origin, discrimination between these DNAs can only be made when theyshow a polymorphism. Table 1 shows the primers used for microsatellite markerscreening, SNP screening and PAC-end clone screening of the transfectants.Previously unknown markers were found with the programme Fileprocessor(Breedveld et al., 2002) using chromosome build 30 from the NCBI.

Table 2. RT-PCR primers used for expression analysis of transfectants.

Primer pair Official genesymbol

Forward primer sequence5’-3’

Reverse primer sequence 5’-3’

RBM6gen1 RBM6 CGTATCTATCGTTCCACACC TGCTAAATGGCGGATCAAGG

RBM5gen1 RBM5 AGAGCCCAAGCGCAAGAAGC TTTCGTCCCAAGCCAGAGCC

Hsema2 SEMA3F CTCTGTACTGGGAGAAGACC GAAGCCCAGAGAAGAAGACC

GNAT1gen2 GNAT1 ACCTCAGCATCTGTTTCCCG ATCTCCTTCACGTCGCGCC

G17gen1 SLC38A3 CACAGGCAACTTCAGCCACG GACAGCGATGGACAGGTTGG

GNAI2gen1 GNAI2 GCATCTGCAACAACAAGTGG CGTCAAACACGACTGCACG

Sema3B1 SEMA3B GCTGCATGCCTACAACCGCA TCTCGTCCCATGAGGTCTGC

skmc15gen1 IFRD2 TGGACCTAAGGGTGAGGAGC GGCTGATTTGGGTGCTAGGG

Luca3gen1 HYAL3 CAGTCCATTGGTGTGAGTGC AAGGCTTCCATCTGTCCTGG

Fus2gen2 NAT6 CAAAGGGTGCGGCTGCTGC GATCAGCTCCATCCGGTGTG

Hyal1gen1 HYAL1 TGGAGTGGTGCTCTGGGTGA CGACATTTGAACTCCACAGGC

Hyal2gen1 HYAL2 TACCTGGACGAGACACTTGC CGCCAATGGTAGAGATGAGG

Fus1gen1 TUSC2 CAATCGTCACCAAGAACGGG GGATCACAGGGAAATCCACG

RASSF1A RASSF1A ACACGTGGTGCGACCTCT GATGAAGCCTGTGTAAGAACCGTCCT

123F2gen1 RASSF1C GATCAAGGAGTACAATGCCC GTCATCATCCAACAGCTTCC

BLUgen1 ZMYND10 TGATCGCAGTGGAGATGTGG ATTCCATCAGCTCTGCCTGC

24gen1 TUSC4 TTCAGCAGAGGCAGATGTGG TTCAGCAGAGGCAGATGTGG

101F6gen1 CYB561D2 ACTAGCACGGCTGACGATGG AGACATAAGCACCGGGTGCC

PL6gen1 PL6 CACTTTCTTCCCTGAGATCC GGTCTTCCACTCTCTTCAGC

Alphagen1 CACNA2D2 TGTGTCATGCTTCACACACC ACTTCTCAAAGACGTCCTGC

Expression analysisTo determine the mRNA expression of the 19 genes from the region homozygouslydeleted in GLC20 and of the three genes located on pPAC4 in the transfectants usedfor tumourigenicity tests, we first isolated RNA using the RNeasy mini kit (Qiagen,Valencia, CA, USA). Replicate Northern blots were prepared using 15 µg of total

Chapter 2

54

RNA per lane. First strand cDNA was made with a Ready to go you-primed first-strand kit (Amersham Biosciences) using random hexamer primers p(dN)6(Amersham Biosciences). For each gene, double stranded cDNA probes wereproduced using gene-specific PCR primers (Table 2). These cDNA probes wereradioactively labelled with [32P]dCTP by random priming using random hexamerprimers p(dN)6 (Amersham Biosciences). Hybridisation was performed usingExpressHyb hybridisation solution (BD Biosciences, Mountain View, CA, USA) andpost-hybridisation washes were performed according to manufacturer’s instructions.

Methylation assay for RASSF1A and ZMYND10Genomic DNA was subjected to bisulfite treatment using the CpGenome DNAmodification kit (CHEMICON, Temecula, CA, USA) and the protocol provided. Thisconverts all unmethylated but not the methylated cytosines to uracil. Subsequently,PCR amplification was performed with primers specific for methylated versusunmethylated DNA for RASSF1A (Burbee et al., 2001) and ZMYND10 (Liu et al.,2003). PCR amplification for RASSF1A was performed for 40 cycles of denaturing at94°C, annealing at 59°C (unmethylated-specific PCR) and annealing at 64°C(methylated-specific PCR), and a final extension step at 72°C for 10 minutes. ForZMYND10 unmethylated and methylated-specific PCRs, a stepdown PCRprogramme was used with the annealing temperature decreasing from 64°C to 59°C.All other conditions were as for the PCR of RASSF1A. PCR products were analysedon agarose gels. Results given have been obtained in three independentexperiments, each starting at the bisulfite treatment step.

In vivo growth assayBefore starting the assays, transfectants were grown for four weeks in completemedium without blasticidin. Male athymic nude BALB/c nu/nu mice of 5-6 weeks ofage (Harlan, Indianapolis, IN, USA and Charles River Laboratories, Wilmington, MA,USA) were injected subcutaneously in both the left and the right dorsal flank with 1 x106 cells from single transfectants, washed and resuspended in PBS in a volume of100 µl supplemented with 100 µl of matrigel (BD Biosciences). Developing tumourswere measured with a calliper once a week. After 4 weeks, tumours were isolatedand their wet weight was determined.

Complexity of functional analysis

55

Experimental design for tumourigenicity testingSix mice were injected on their left side with a transfectant containing a first PAC andon their right side with another transfectant containing a second PAC. Another sixmice were injected with the same transfectants, but with the sites of injectionreversed. This procedure was repeated with a next pair of transfectants counting thesame PACs, i.e. another 12 mice were used (Fig. 1). Thus in each experiment, thetumourigenicity effect of an integrated PAC was tested in 24 mice. In total, ninetumourigenicity experiments were performed (Table 3).

Table 3. Complete tumourigenicity test design.

Experimenta Sample 1b Sample 2b

1 157G3 175O16/134H62 157G3 185A43 157G3 234N44 157G3 pPAC45 185A4 234N46 GLC45 pPAC47 185A4 211D148 185A4 GLC459 185A4 76L5

a per experiment 24 mice were usedb Sample 1 and 2 were per mouse injected in different flanks. The samples are transfectants withintegration of the indicated PAC

Figure 1. The transfection and tumourigenicity test design used in this study.Per PAC two different clones (transfectants) are selected. Per mouse two transfectants, each containinga different PAC, are tested. For each of the four combinations shown, six mice were used.

Chapter 2

56

StatisticsStatistical significance was determined by analysis of variance (ANOVA) withScheffe’s F test for contrasts, using the software programme SPSS 10.0.

BAC micro-arrayThe genome-wide microarray used in this study contains 6,465 large-insert clonesincluding both bacterial artificial chromosomes (BACs) and P1-based artificialchromosomes (PACs). The largest part of the clone collection used consisted of the1-Mb BAC collection obtained from Dr. Nigel Carter (Wellcome Trust SangerInstitute, UK) and the clones from the Human BAC Resource Consortium_1 Set (Dr.Pieter de Jong, Children’s Hospital Oakland Research Institute, USA). For thepositioning of the BACs relative to the human sequence we have used the May 2004human reference sequence (UCSC version hg17) that is based on NCBI Build 35.The BAC DNA isolation procedure was adapted from the protocol published at thewebsite of Dr. Reinhard Ullmann at the Max Planck Institute for Molecular Genetics,Berlin (http://www.molgen.mpg.de/~abt_rop/molecular_cytogenetics/BACIsolationProtocol.html) and carried out in 96-well plates. PCR amplification of the BAC andPAC clones was performed as described by Fiegler et al. (2003). All BACs werespotted in triplicate onto epoxy-coated slides (SCHOTT Nexterion, Mainz, Germany)by use of a MicroGrid II arrayer (BioRobotics, Cambridge, UK), and processed asrecommended by the manufacturer of the slides. C0t1 Human DNA (Roche, Basel,Switzerland) and Drosophila BAC DNA were spotted as controls.

Labelling of genomic DNAGenomic DNA (700 ng) was labelled with either Cy3- or Cy5-dUTP (Perkin Elmer,Langen, Germany) using the BioPrime DNA Labeling System (Invitrogen Inc.,Carlsbad, CA). Unincorporated nucleotides were removed by use of MicroSpin G-50columns (Amersham Biosciences). Test and reference DNAs were combined on aMicron YM30 column (Millipore, Etten-Leur, The Netherlands) and dissolved in 50 µlhybridisation buffer (supplied by SCHOTT Nexterion) supplemented with 5% dextransulphate, 600 µg C0t1 human DNA (Roche) and 3 mg yeast tRNA (Sigma, St. Louis,MO, USA). Reference DNAs (male or female) were pools of 20 different individuals.

Complexity of functional analysis

57

HybridisationSlides were processed according to the protocol of the manufacturer. Bothprehybridisation and hybridisation were carried out under a lifterslip (Erie Scientific)in a humidified chamber with a solution consisting of 10 µg/µl salmon sperm DNAand 5% dextran-sulphate in hybridisation solution (SCHOTT Nexterion). Forhybridisation, the hybridisation solution was supplemented with 5% dextran-sulphate,C0T-DNA (final concentration 4 mg/ml) (Invitrogen) and labelled DNA. Afterdenaturation of the DNA by heating the mixture at 100°C for 5 min, the hybridisationmix was immediately applied to the array. Hybridisation was performed at 65°C for40 hours. Post hybridisation washes were carried out as recommended by themanufacturer of the slides (SCHOTT Nexterion). Slides were dried by spinning at800 rpm at room temperature for 3 min.

Image AnalysisArrays were scanned using the Affymetrix 428 scanner (Affymetrix Inc., Carlsbad,CA). The resulting images were analysed with ImaGene software package 5.6(BioDiscovery Inc., Marina Del Rey, CA).

CGH data analysisTo determine significant DNA copy number changes in the transfectants comparedto the parental cell line, the software programmes CGHpro and CGH-Miner wereused. With CGHpro (Chen et al., 2005) the signal intensity ratio data werenormalised using printtip LOWESS (locally weighted scatterplot smoothing)normalisation to account for intensity-dependent and spatial effects. Next the datawere smoothed using the DNAcopy algorithm and the BAC genomic alignment wassegmented into regions with similar DNA copy numbers using Circular Binarysegmentation (A. Olshen, Circular Binary Segmentation for the analysis of Array-based DNA copy number Data). Regions with a mean log ratio < -0.1 or > 0.1between signal intensities were selected as aberrant. Data from all transfectantswere plotted and analysed simultaneously.

Chapter 2

58

RESULTS

Introduced PACs integrate into the DNA of the tumour cell lineThe 3p21.3 region which has been homozygously deleted in GLC20 was covered bya contig of eight overlapping PACs (Fig. 2). Each PAC was transfected into cellsfrom the SCLC cell line GLC45. Table 4 summarises the results of the transfectionand FISH analysis of the transfected clones. The number of clones obtained in thetransfection experiments differed for the different transfections and for the differentPACs. It appeared very difficult to obtain clones with integration of the vector only. Toverify the number of integrated PAC copies as estimated by FISH analysis, a numberof the transfected clones were subjected to Southern blot analysis. The copynumbers of the integrated PAC as calculated were similar to those estimated fromFISH analysis and ranged between 1 and 6. By PCR analysis using polymorphismsbetween the transfected PAC DNA and the endogenous GLC45 DNA, we coulddemonstrate integration of the PACs into the genome of the transfected clones. Theresults of these PCR analyses are shown in Table 5.

Table 4: Transfection and FISH results of PACs and the empty PAC vector.

Figure 2. PAC contig covering the 370 kb lung cancer critical region homozygously deleted at 3p21.3 inGLC20.PACs have been mapped by using their end-sequences and by analysing them with primers of the genesand markers located in the region.

Complexity of functional analysis

59

Table 4. Transfection and FISH results of PACs and the empty PAC vector

PAC Transfectionnumber

Wellsa Clonesb FISH c Not clonal d Transfectantsmeeting the

criteriae

76L5 1 34 142 53 293 22 12

Total 109 55 11 7 2211D14 1 39 21

Total 39 21 13 11 2143C21 1 20 4

2 71 36Total 91 40 4 3 0

234N4 1 2 22 5 43 35 74 31 19

Total 73 30 21 12 8175O16 1 30 27

Total 30 27 23 14 2134H6 1 54 30

Total 54 30 26 18 1157G3 1 21 14

2 43 13Total 64 27 20 15 2

185A4 1 14 102 26 113 50 20

Total 90 41 25 20 1pPAC4 1 0 0

2 ? 43 10 7

Total 11 11 8 4 4Total 561 279 151 104 22

a Number of wells (48 wells plate) in which cells are growing after 2-3 weeks.b Number of clones still growing after 4-6 weeks and transferred to a 6 wells plate.c Number of clones analysed with FISH.d Number of clones showing multiple cell lineages, according to FISH.e Pure transfectants with a single site of integration showing an intensity comparable to that of theendogenous signal.

Chapter 2

60

Table 5. Presence of PAC sequences in transfectants used in tumourigenicity tests

transfectant D3S2968(RBM6)

LUCA20(RBM5)

ISSCP4(RBM5)

ISSCP14(RBM5)

SEMA3F LUCA17(GNAT1)

SLC38A3

76L5-1 + +76L5-2 + +

211D14-1 + + + + +211D14-2 + + + + +234N4-1 + +234N4-2 + +

175O16-1 - +134H6-2 - -

transfectant GNAI2 LUCA13(HYAL2)

NAT6 RASSF13’UTR

ZMYND103’UTR

CYB561D105’UTR

211D14-1 +211D14-2 +234N4-1 + + +234N4-2 + + +

175O16-1 +134H6-2 +157G3-1 + + +157G3-2 + + +185A4-1 +185A4-2 +

transfectant CACNA2D23’UTR

LUCA11(CACNA2D2)

pPAC4-endTEL.

pPAC4-endCEN.

Blasticidin

76L5-1 + N.D1 +76L5-2 + N.D1 +

211D14-1 + + +211D14-2 + + +234N4-1 + + +234N4-2 + + +

175O16-1 N.D1 + +134H6-2 N.D1 + +157G3-1 + + + + +157G3-2 + + + + +185A4-1 + + + +185A4-2 + + + +

pPAC4-1 +pPAC4-2 +

Tests have only been applied for sequences present on the transfected PAC1 Could not be determined due to lack of unique sequence- absent+ present

Complexity of functional analysis

61

Genes on integrated PACs are being expressedThe expression level of the 3p21.3 genes was evaluated by semi-quantitative RT-PCR (Fig. 3 and Table 6) and Northern blot (not shown) analyses. Results of bothanalyses were in agreement with each other. For SEMA3F, expression could not bedetected, neither in the parental cell line GLC45, nor in any of the transfected clones.Expression of RASSF1A was detected in GLC45 and in the transfectants containingPAC 157G3 on which RASSF1 is present, but not in those clones that weretransfected with PACs 211D14 and 234N4. In these latter transfectants, however,promoter hypermethylation could be demonstrated, which was not the case inGLC45 and in the transfectants containing PAC157G3 (data not shown). The othergenes all showed endogenous expression in GLC45 and mostly a higher expressionin the transfectants with extra gene copies. Using polymorphisms in the 3’UTR or5’UTR of RASSF1A, ZMYND10 and CACNA2D2, we also investigated whetherthese genes showed a PAC-specific gene expression in the transfectants with PAC157G3. The results presented in Fig. 4, show that PAC-specific expression wasfound in the transfectants with RASSF1, ZMYND10 and CACNA2D2 integrated,albeit for CACNA2D2 in only one of the two transfectants analysed.

Chapter 2

62

Figure 3. Expression analysis of 3p21.3 genes in transfectants and parental cell line by semi-quantitativeRT-PCR.pPAC4, cDNA from two transfectants with the vector integrated; 76L5, cDNA from two transfectants withPAC 76L5 integrated; 211D14, cDNA from two transfectants with PAC 211D14 integrated; etc, GLC45, 1:cDNA, 2: DNA.

Complexity of functional analysis

63

Table 6. Gene expression of transfectants as assessed by semi-quantitative RT-PCR.

RBM6 RBM5 SEMA3F GNAT1 G17 GNAI2 SEMA3BGLC45 + + - - + + +

pPAC4-1 + + - - + + +pPAC4-2 + + - - + + +

76L5-1 + + - - + + +76L5-2 + + - - + + -

211D14-1 + - - + ++ ++ +211D14-2 + + - + ++ + ++

234N4-1 + + - + ++ ++ ++234N4-2 + + - + ++ ++ ++

175O16-1 + + - - + ++ ++134H6-2 + + - + + + +157G3-1 + + - - + + +157G3-2 + + - - + + +185A4-1 + + - - + + +185A4-2 + - - - + + +

IFRD2 HYAL3 NAT6 HYAL1 HYAL2 TUSC2 RASSF1AGLC45 + + + + + + +

pPAC4-1 + + + + + + +/-pPAC4-2 + + + + + + -

76L5-1 + + + + + + +/-76L5-2 + + + + + + -

211D14-1 + + ++ + + + -211D14-2 + + ++ + + + -

234N4-1 + ++ + + + + +/-234N4-2 + ++ ++ + + + +/-

175O16-1 + + + + + + +134H6-2 + + + + + + +157G3-1 + + + + + + +157G3-2 + + + + + + +185A4-1 + + + + + + +185A4-2 + + + + + + +

RASSF1C RASSF1F ZMYND10 TUSC4 CYB561D2 PL6 CACNA2D2GLC45 + + +/- +/- + + +

pPAC4-1 + + +/- +/- + + +pPAC4-2 + +/- +/- +/- + + +

76L5-1 + + - +/- + + +76L5-2 + - - +/- + + +

211D14-1 + - +/- +/- + + +211D14-2 + - - +/- + + +

234N4-1 + + +/- +/- + + +234N4-2 + + +/- +/- + + +

175O16-1 + + +/- +/- + + +134H6-2 + + +/- +/- + + +157G3-1 ++ + + + ++ ++ ++157G3-2 ++ + + + ++ ++ ++185A4-1 + + +/- + ++ ++ +185A4-2 + + +/- +/- ++ ++ ++

* grey bars show genes located on indicated PACs- No expression +/- low expression + moderate expression ++ high expression

Chapter 2

64

Transfectants with integration of a PAC from the centromeric part of the 3p21.3critical region show a significantly reduced tumour growth in nude mice. Fig. 5 showsthe number of nude mice injected with a transfectant containing a specific PAC andthe mean weight of the resulting tumours. The same parameters are also presentedfor the parental cell line. By making a scatterplot of the weights of all 418 tumours(Fig. 6) we checked whether the tumours growing on both flanks of a mouseinfluence each other’s growth. The slope of the regression line indicates that thetumours grew completely independent of each other.

The weight of the tumours caused by transfectants with PAC 185A4appeared to be significantly smaller than the weight of the tumours caused bytransfectants containing the other PACs and those caused by the parental cell line(Fig. 5). Significance ranged from p= 0.010 for differences between mean weights oftumours caused by transfectants with PAC 185A4 and by transfectants with PAC76L5 to p<0.001 for differences in mean weight of tumours caused by transfectants

Figure 4. Representativeresults of sequence analysisfor specific PAC geneexpression.Sequence difference ofZMYND10 in GLC45compared to ZMYND10 inPAC 157G3. RNA from PACtransfectants contains >80%PAC-specific ZMYND10 RNAaccording to the sequenceratio C/T (90/10).

Complexity of functional analysis

65

with PAC 185A4 and by transfectants with PAC 157G3 and PACs 175O16 and134H6 or by the parental cell line.

Figure 5. Error bar plot of the mean weight of tumours caused by the parental cell line, transfectants witha PAC integrated and transfectants with vector only integrated.Black circles indicate the mean tumour weight per transfectant injected into 12-59 mice; the bars indicatethe standard error of the mean; N is the number of mice used.

Chapter 2

66

A direct comparison between transfectants with PAC 185A4 and the parentalcell line could be made in 24 mice injected on one side with the parental cell line andon the other side with either transfectant-1 containing PAC 185A4 (12 mice) ortransfectant-2 (also 12 mice). The mean weight of the tumours caused by theparental cell line differed significantly from the mean weight of the tumours causedby each of the two transfectants (p= 0.004 and p= 0.001, respectively). A comparisonof the weight of the tumours per mouse showed that 11 of 12 mice had smallertumours from each transfectant with PAC 185A4, whereas 1 mouse had smallertumours from the parental cell line GLC45 (Fig. 7a). PAC 185A4 is located at thecentromeric part of the region covering the deletion in GLC20 and contains twogenes, CYB561D2 and PL6.

Figure 6. Scatterplot of individual weights of all left- and right-sided tumours plotted against each other.The scatterplot shows a random spot pattern and the slope of the regression line indicates that the tumoursgrew completely independent of each other.

Complexity of functional analysis

67

Transfectants with vector only integration also show a significantly reducedtumour growth in nude miceAs a control we also introduced pPAC4 without a human insert into the parental cellline GLC45. The mean weight of the tumours formed by one of the transfectantscontaining vector sequences only was significantly lower (p = 0.016) than the weightof tumours resulting from injection of the parental cell line (Fig. 5).

A direct comparison between transfectants with vector only and the parentalcell line could be made in 24 mice on one side injected with cells from the parentalcell line and the other side with either transfectant-1 containing the empty vector (12mice) or transfectant-2 containing the vector (also 12 mice). The mean weight of thetumours caused by the parental cell line differed significantly from the mean weightof the tumours caused by transfectant-2 (p=0.000), but not from the mean weight of

Figure 7. Bar chart of a direct comparison between the weights of tumours caused by the parental cellline and by the transfectants with PAC 185A4 or vector only integrated (grey, transfectants-1; blacktransfectants-2).A. Direct in-mouse comparison between GLC45 (parental cell line) and transfectants with PAC 185A4.The weight of the tumour caused by GLC45 is subtracted from the weight of the tumour caused by atransfectant with PAC 185A4. Amongst 24 mice, 22 have smaller tumours from transfectants with PAC185A4.B. Direct in-mouse comparison between GLC45 (parental cell line) and transfectants with pPAC4(empty vector). The weight of the tumour caused by GLC45 is subtracted from the weight of the tumourcaused by a transfectant with pPAC4. Amongst 12 mice, 9 mice have smaller tumours from transfectant-1with pPAC4; from transfectant-2 all 12 mice have smaller tumours.

Chapter 2

68

the tumours caused by transfectant-1 (p=0.273). A comparison of the weight of thetumours per mouse showed that 9 of the 12 mice had smaller tumours fromtransfectant-1, whereas 3 mice had smaller tumours from the parental cell lineGLC45 (Fig. 7b). Transfectant-2 with vector only caused smaller tumours in 12 of the12 mice.

No tumour growth-related expression of the genes on the vectorTransfectants with integration of the approx. 20 kb vector only sequence showedFISH signal intensities similar to transfectants with approx. 100 kb PAC sequences.This might implie that there is a five-fold excess of vector sequences at anintegration site in comparison with the number of PAC sequences at an integrationsite. To investigate whether expression of the many copies of one or more of thethree vector genes SACBII, the EBV-oriP and the gene coding for blasticidin-S-methylase could cause the observed reduced tumour growth of the transfectants withvector sequences only, Northern blots with total RNA of all transfectants used in thetumourigenicity tests were hybridised with PCR products from these genes. Theexpression of the gene coding for blasticidine-S-methylase as seen on Northern blotsvaried over the transfectants (Fig. 8), but was not related to the size of the tumour.No expression was detected for SACBII and the EBV-oriP.

Figure 8. Northern blot ofSACBII, blasticidin-S-methylase and EBV-oriexpression in total RNA fromtransfectants and parentalcell line.No correlation can be seenbetween the expression ofthe genes and the tumourweight.

Complexity of functional analysis

69

No common genomic integration sitesTo investigate whether transfectants showed (a) preferred site(s) of integration wechecked the integration sites as shown by FISH analysis. Since gross chromosomalaberrations are already present in the parental cell line, chromosomes were groupedaccording to size and PAC integration sites were assigned to groups ofchromosomes. PAC integration sites were found spread over all groups and clearlycommon positions relative to a centromere could not be observed (Table 7). Notably,for each of the transfectants showing reduced tumour growth after injection into nudemice integration occurred at (a) different site(s) (Table 7). If the diminished tumourgrowth capacity of both transfectants with PAC 185A4 and of one of the tranfectantswith the vector only was due to interference of the integration event with theexpression of some oncogene, then a different oncogene would have been affectedin each of the three cases.

Table 7. PAC integration sites found by FISH analysis

Approximate position of signalChromosome groupdistal p at 2/3 p at ½ p prox. p centr. prox. q at 1/3 q at ½ q at 2/3 q distal q

A 2 1 3 1 2 7 1B 1 1 5 b

C 3b 1 4 3 2 a 4 3 6D 2 c 1

E/F 3 6 1 9 b

a including integration site of PAC 185A4 in transfectant-1b including integration sites of PAC 185A4 in transfectant-2c including integration site of pPAC4 in transfectant-2

Chapter 2

70

Transfection induces chromosomal instabilityIn a micro-array expression analysis of RNA from the transfectants with PAC 185A4and with vector only, an overrepresentation of differentially expressed genes locatedon the short arm of chromosome 17 was found in the vector only transfectant-1 incomparison to the parental cell line. Most of these genes had a lower expression intransfectant-1 than in the parental cell line, suggesting loss of the short arm ofchromosome 17 in the vector only transfectant-1 clone. Moreover, we observed withFISH analysis that in 8 of 51 transfectants analysed the original derivativechromosome 3 as present in the parental cell line GLC45 had been involved in sometranslocation event. This either indicates occurrence of chromosomal instability in thetransfectant clones or selection in cell cloning of representatives of minor sub-populations of cells of the parental cell line GLC45. To investigate the extent of achromosomal instability, we hybridised DNA from 8 transfectants with to a 6,500BAC array, using the parental cell line GLC45 as a reference. The 8 transfectantswere those with PACs 211D14, 234N4, 185A4 and those with vector only, thusincluding the three transfectant clones that caused a reduced tumour growth afters.c. injection into nude mice. We also included in this comparison two 4 weeks-subcultures of the parental cell line. In a comparison of GLC45 and thesesubcultures, no apparent gains or losses were found, indicating that GLC45 on thewhole remained stable. In the transfected clones, however, multiple gains and lossescould be seen (Fig. 9). Some of these had already occurred in subpopulations of theparental cell line, as was clear from whole chromosome paints, a.o. of chromosomes5, 8 and 15 in GLC45. However, a comparison of these as well as other paintedchromosomes with the results of Fig. 9 showed that many aberrations in thetransfectants had occurred de novo.

Complexity of functional analysis

71

Figure 9. Comparative analysis plot of array-CGH results for transfectants with different PACs and withvector only.Average log2 ratios were plotted for all clones based on chromosome position. Light grey, no aberrations;black, loss of coloured region from the transfectant as compared to the parental cell line; dark grey,gain/amplification of coloured region in the transfectant as compared to the parental cell line; white, noBACs available that met the proper criteria.

Chapter 2

72

DISCUSSION

Purpose of the studyWe wanted to introduce genes from the presumed lung cancer critical region at3p21.3 as genomic sequences together with their own promoter and some otherregulatory sequences into a lung cancer cell line in a very low copy number at asingle integration site and to investigate whether the integrated genes couldsuppress the tumour-inducing capacity of the original lung cancer cell line. By thisapproach to identify a gene’s tumour suppressor activity, we wanted to avoid theusual huge overexpression caused by introducing the gene as a cDNA behind astrong viral promoter. In that case, the resulting abundant production of a singleprotein may well disturb essential cellular processes and thereby affect growth andproliferation of the tumour cells, i.e. mimic tumour suppression. A genuine tumoursuppressor gene, however, would act as such also when present in low dosage.

Conditions for achieving the purposeIn the less artificial situation we wanted to create, successful introduction of arelatively long genomic segment into a cell, stable integration of that segment in lowcopy number into the cellular genome and expression of the newly introduced geneson the segment are essential conditions. Once FISH analysis showed that PAC DNAhad been integrated into the cellular genome, we could demonstrate presence in thecell of both PAC ends and of all PAC markers that we could check. We selectedthose transfectants that had the PAC DNA integrated at a single site which uponhybridisation with fluorescently labelled PAC DNA gave a signal with an intensitysimilar to that of the signal from the endogenous sequence. We could show that thecopy number of the PAC at the integration site varied from one to six, the highestcopy number giving more intense signals. For almost all of the introduced genes, wefound an endogenous expression in the parental cell line GLC45 and a higherexpression in the transfectants with an extra gene copy. In addition, by usingpolymorphisms present in the untranslated regions of genes located in thecentromeric part of the 3p21.3 critical region, we could demonstrate that this higherexpression was PAC-specific. Although for RASSF1A and ZMYND10, promoterhypermethylation is frequently found in a wide variety of tumours and cell lines,including SCLC and NSCLC (Agathanggelou et al., 2001; Burbee et al., 2001;Dreijerink et al., 2001; Dammann et al., 2001; Toyooka et al., 2001), this appeared to

Complexity of functional analysis

73

be not the case in the parental tumour cell line GLC45, nor in the transfectants withan extra RASSF1 or ZMYND10 gene copy. For four transfectants with one or theother of two PACs from the middle part of the 3p21.3 critical region we found,however, a diminished expression of RASSF1A. Since both transfectants showedexpression of the other large transcript of RASSF1, RASSF1C, it seemed unlikelythat RASSF1 was physically affected by the integration of the PACs. Indeed, wecould demonstrate hypermethylation of the RASSF1A promoter in thesetransfectants. Since neither in the parental cell line nor in transfectants containing anextra gene copy of SEMA3F expression of SEMA3F could be detected, we lookedfor possible miRNA target sites in its sequence. Using the DIANA algorithm(http://diana.pcbi.upenn.edu/cgi-bin/micro_t.cgi) a possible target site of miRNA hsa-mir-188 located at Xp11.22-p11.23 was found, predicted from homology with averified murine miRNA (Lagos-Quintana et al., 2003). We have no indication,however, for a possible SEMA3F silencing amplification of the X-chromosomalregion in GLC45.

Possible mechanisms explaining a reduced tumour growthIf a transfectant would show a significantly reduced tumour growth in nude mice,different mechanisms might be held accountable for that. It could be a consequenceof the expression of the newly introduced and integrated genes, in which case thesegenes would act as tumour suppressor genes. Alternatively, genes with anoncogenic activity at the site of integration may be disrupted and inactivated by theintegration event. Another mechanism might be that integration of a large DNAsegment at some position in a well-functioning genome or the expression of severaladditional genes causes a disturbance of the chromosomal stability, i.e. induces awidespread chromosomal instability interfering with the growth and proliferation ofcells.Newly introduced genesBoth transfectants containing PAC 185A4 from the centromeric part of 3p21.3 criticalregion caused tumours that were significantly smaller than those caused by theparental cell line. In one transfectant one to three copies of the PAC had beenintegrated together at a single site. Since we were not able to find anothertransfectant with a single integration site, we used as a second transfectant a clonewith one to three PAC copies at each of three different integration sites. PAC 185A4contains two genes. One is PL6, coding for placental protein 6, a putative

Chapter 2

74

endoplasmic reticulum protein with at least three transmembrane domains. The otheris CYB561D2, encoding cytochrome b-561 domain 2. Both genes show expressionin the two transfectants with integration of PAC 185A4. It has been reported that inSCLC cell lines expression of PL6 is slightly reduced, but mutations have not beenfound in the 38 cell lines and 40 tumour samples screened for mutations (Lermanand Minna, 2000). Tumour suppressor activity has not been tested. The CYB561D2product is responsible for recycling ascorbate for the generation of norepinephrine bydopamine-beta-hydroxylase in the chromaffin vesicles of the adrenal gland. Thegene is highly expressed in NSCLC lung cancer cell lines, but only moderately inSCLC cell lines. Mutations could not be found in 40 cell lines and 40 SCLC tumoursamples analysed for that purpose (Lerman and Minna, 2000). Cell proliferation invitro of NSCLC cells transfected with CYB561D2 in an adenoviral vector wassignificantly reduced and CYB561D2 also caused inhibition of tumour growth ofhuman xenografts. Although the functions presently attributed to the two genes donot seem to have an obvious relation to tumour suppression, based upon our resultsand what has been reported in the literature we can also not definitely exclude thesegenes as candidate tumour suppressor genes.

PAC 185A4 also contains an incomplete but large part of CACNA2D2, thegene encoding calcium channel voltage dependent alpha2/delta subunit 2. The thirdintron of CACNA2D2, which is located on PAC 185A4, harbours a miRNA (cand564HS 3, 50402640-50402729), included in a recently published list of computationallyidentified human miRNA genes (Berezikov et al., 2005). Although we could not findtarget genes with the DIANA algorithm, we found with BLAST a 20 bp overlapbetween pre-miRNA cand546 HS 3 and KCNQ3 encoding the potassium voltage-gated channel, KQT-like subfamily, member 3. This gene at 8q24 could be a miRNAtarget with its high degree of complementarity with miRNA cand 546. That wouldimply that when CACNA2D2, coding for a calcium channel subunit, would betranscribed, the miRNA would be transcribed and KCNQ3, a potassium channelsubunit, would be down regulated. There are no reports, however, on an involvementof KCNQ3 in oncogenic processes. Since many miRNAs have multiple targets, itcannot be excluded, however, that some other gene may be targeted which wouldhave a tumour promoting role.

Remarkably, also one of the transfectants with integration of vectorsequences only caused tumours that were significantly smaller than those caused bycells from the parental cell line. FISH analysis of transfectants with vector only

Complexity of functional analysis

75

showed signal intensities comparable to signal intensities of an integrated completePAC. Since the size of the vector is one fifth of the size of an entire PAC, there canbe a copy number of the integrated vector sequences which is fivefold higher thanthat of integrated PACs. We could not find any correlation, however, between theexpression of the few genes present in the vector sequence on the one hand andtumour size on the other hand. This makes it highly unlikely that the observeddiminished tumour growth could be attributed to vector gene expression.Gene disruption at integration siteDisruption of a gene with some oncogenic activity at the site of integration can alsocause reduction of a cell’s tumour growth capacity. Since all integrations of PACsand the vector had occurred at different sites, acceptance of gene disruption as anexplanatory mechanism for the smaller tumours found upon injection of the twotransfectants with PAC 185A4 and one of the transfectants with the vector only,would imply the assumption that each of at least the three integrations involved haddisrupted a gene and that each of these three genes would have some role in cellproliferation. Though this cannot be excluded, it doesn’t seem very likely.Induced chromosomal instabilityInduction of a widespread chromosomal instability as a consequence of theintroduction of exogenous DNA might by interfering with the growth and proliferationof the cells provide an explanation for the effect of both the transfectants with PAC185A4 and the transfectant with vector only sequences. The parental cell line showskaryotypic heterogeneity as demonstrated by whole chromosome painting for anumber of chromosomes. Though some of the chromosomal abnormalities found inthe transfectants will have their origin in this heterogeneity, we could demonstrate aclear increase in chromosomal instability when comparing genomic DNA from alltransfectants tested with that from the parental cell line. In this respect there was nodifference between the transfectants causing reduced tumour growth and the othertransfectants. Chromosomal instability has also been found after introduction andintegration of cDNAs of genes from various genomic regions and after integration ofthe vector only sequences pcDNA4.0, pcDNA3.1, pCi-neo and pUC19 from thosecDNA constructs, (Akli et al., 2004; Fraizer et al., 2004; Kim et al., 2005;Stavropoulou et al., 2005; Tomonaga et al., 2005). This is a strong argument againstconsidering our results as cell line-specific or selection marker-specific.

The effect of the induction of chromosomal instability as a consequence ofthe introduction into a cell of exogenous DNA depends on activation or inactivation of

Chapter 2

76

genes with a role in tumourigenesis through the deletion and duplication events thatrepresent the instability. The resulting cellular phenotype can be diverse. This mightexplain that in comparison with the parental tumour cell line, transfectants with PAC185A4 showed a reduced tumour growth in nude mice, whereas for transfectantswith PAC 157G3, which is completely overlapping PAC 185A4, no statisticallysignificant reduced tumour growth was observed. The possibility remains, however,that the newly introduced genes do exert a growth-inhibitory influence, which in thetransfectants with PAC 157G3 or in the tumours caused by these transfectants isoverruled by growth-promoting effects of the induced instability.

ACKNOWLEDGEMENTSThis work was supported by a grant from the Dutch cancer society RUG 2000-2317.We thank dr. Peter Terpstra for bioinformatics assistance, Hendrika Faber for hercontribution to the FISH analysis and Hans Bartels for zootechnical assistance.

Complexity of functional analysis

77

REFERENCES

Agathanggelou A, Honorio S, Macartney DP, Martinez A, Dallol A, Rader J, Fullwood P, Chauhan A,Walker R, Shaw JA, Hosoe S, Lerman MI, Minna JD, Maher ER, Latif F. 2001. Methylationassociated inactivation of RASSF1A from region 3p21.3 in lung, breast and ovarian tumours.Oncogene 20:1509-1518.

Akli S, Zheng PJ, Multani AS, Wingate HF, Pathak S, Zhang N, Tucker SL, Chang S, Keyomarsi K:. 2004.Tumor-specific low molecular weight forms of cyclin E induce genomic instability and resistance top21, p27, and antiestrogens in breast cancer. Cancer Res 64:3198 -3208.

Berezikov E, Guryev V, van de Belt J, Wienholds E, Plasterk RH, Cuppen E. 2005. Phylogeneticshadowing and computational identification of human microRNA genes. Cell 120:21-24.

Breedveld GJ, van Dongen JW, Danesino C, Guala A, Percy AK, Dure LS, Harper P, Lazarou LP, van derLinde H, Joosse M, Gruters A, MacDonald ME, de Vries BB, Arts WF, Oostra BA, Krude H,Heutink P. 2002. Mutations in TITF-1 are associated with benign hereditary chorea. Hum MolGenet 11:971 -979.

Burbee DG, Forgacs E, Zochbauer-Muller S, Shivakumar L, Fong K, Gao B, Randle D, Kondo M, VirmaniA, Bader S, Sekido Y, Latif F, Milchgrub S, Toyooka S, Gazdar AF, Lerman MI, Zabarovsky E,White M, Minna JD. 2001. Epigenetic inactivation of RASSF1A in lung and breast cancers andmalignant phenotype suppression. J Natl Cancer Inst 93:691-699.

Chen W, Erdogan F, Ropers HH, Lenzner S, Ullmann R. 2005. CGHPRO -- a comprehensive dataanalysis tool for array CGH. BMC Bioinformatics 6:85-91.

Chow LS, Lo KW, Kwong J, To KF, Tsang KS, Lam CW, Dammann R, Huang DP. 2004. RASSF1A is atarget tumor suppressor from 3p21.3 in nasopharyngeal carcinoma. Int J Cancer 109:839-847.

Daly MC, Xiang RH, Buchhagen D, Hensel CH, Garcia DK, Killary AM, Minna JD, Naylor SL. 1993. Ahomozygous deletion on chromosome 3 in a small cell lung cancer cell line correlates with aregion of tumor suppressor activity. Oncogene 8:1721-1729.

Dammann R, Takahashi T, Pfeifer GP. 2001. The CpG island of the novel tumor suppressor geneRASSF1A is intensely methylated in primary small cell lung carcinomas. Oncogene 20:3563-3567.

Dreijerink K, Braga E, Kuzmin I, Geil L, Duh FM, Angeloni D, Zbar B, Lerman MI, Stanbridge EJ, MinnaJD, Protopopov A, Li J, Kashuba V, Klein G, Zabarovsky ER. 2001. The candidate tumorsuppressor gene, RASSF1A, from human chromosome 3p21.3 is involved in kidneytumorigenesis. Proc Natl Acad Sci U S A 98:7504-7509.

Fiegler H, Carr P, Douglas EJ, Burford DC, Hunt S, Scott CE, Smith J, Vetrie D, Gorman P, Tomlinson IP,Carter NP. 2003. DNA microarrays for comparative genomic hybridization based on DOP-PCRamplification of BAC and PAC clones. Genes Chromosomes Cancer 36:361-374.

Fraizer GC, Diaz MF, Lee IL, Grossman HB, Sen S: Aurora-A/STK15/BTAK enhances chromosomalinstability in bladder cancer cells. Int J Oncol 2004 Dec ;25 (6 ):1631-1639 (2004).

Frengen E, Zhao B, Howe S, Weichenhan D, Osoegawa K, Gjernes E, Jessee J, Prydz H, Huxley C, deJong PJ. 2000. Modular bacterial artificial chromosome vectors for transfer of large inserts intomammalian cells. Genomics 68:118 -126.

Hermouet S, Aznavoorian S, Spiegel AM. 1996. In vitro and in vivo growth inhibition of murine melanomaK-1735 cell by a dominant negative mutant alpha subunit of the Gi2 protein. Cell Signal 8:159-166.

Ji L, Nishizaki M, Gao B, Burbee D, Kondo M, Kamibayashi C, Xu K, Yen N, Atkinson EN, Fang B, LermanMI, Roth JA, Minna JD. 2002. Expression of several genes in the human chromosome 3p21.3homozygous deletion region by an adenovirus vector results in tumor suppressor activities in vitroand in vivo. Cancer Res 62:2715 -2720.

Chapter 2

78

Kim D, Pemberton H, Stratford AL, Buelaert K, Watkinson JC, Lopes V, Franklyn JA, McCabe CJ. 2005.Pituitary tumour transforming gene (PTTG) induces genetic instability in thyroid cells. Oncogene24:4861 -4866.

Kok K, van den Berg A, Veldhuis PMJF, van der Veen AY, Franke M, Schoenmakers EFPM, HulsbeekMMF, van der Hout AH, de Leij LFMH, van de Ven W, Buys CHCM. 1994. A homozygous deletionin a small cell lung cancer cell line involving a 3p21 region with a marked instability in yeastartificial chromosomes. Cancer Res 54:4183-4187.

Kok K, Naylor SL, Buys CHCM. 1997. Deletions of the short arm of chromosome 3 in solid tumors and thesearch for suppressor genes. Adv Cancer Res 71:27-92.

Lagos-Quintana M, Rauhut R, Meyer J, Borkhardt A, Tuschl T: New microRNAs from mouse and human.RNA 2003 Feb ;9 (2):175 -183 (2003).

Lerman MI, Minna JD. 2000. The 630-kb lung cancer homozygous deletion region on human chromosome3p21.3: identification and evaluation of the resident candidate tumor suppressor genes. TheInternational Lung Cancer Chromosome 3p21.3 Tumor Suppressor Gene Consortium [In ProcessCitation]. Cancer Res 60:6116-6133.

Li J, Wang F, Protopopov A, Malyukova A, Kashuba V, Minna JD, Lerman MI, Klein G, Zabarovsky E:.2004. Inactivation of RASSF1C during in vivo tumor growth identifies it as a tumor suppressorgene. Oncogene 23:5941 -5949.

Liu XQ, Chen HK, Zhang XS, Pan ZG, Li A, Feng QS, Long QX, Wang XZ, Zeng YX. 2003. Alterations ofBLU, a candidate tumor suppressor gene on chromosome 3p21.3, in human nasopharyngealcarcinoma. Int J Cancer 106:60 -64.

Oh JJ, West AR, Fishbein MC, Slamon DJ. 2002. A candidate tumor suppressor gene, H37, from thehuman lung cancer tumor suppressor locus 3p21.3. Cancer Res 62:3207 -3213 .

Riley J, Butler R, Ogilvie D, Finniear R, Jenner D, Powell S, Anand R, Smith JC, Markham AF. 1990. Anovel, rapid method for the isolation of terminal sequences from yeast artificial chromosome (YAC)clones. Nucleic Acids Res 18:2887-2890.

Roche J, Boldog F, Robinson M, Robinson L, VarellaGarcia M, Swanton M, Waggoner B, Fishel R,Franklin W, Gemmill R, Drabkin H. 1996. Distinct 3p21.3 deletions in lung cancer and identificationof a new human semaphorin. Oncogene 12:1289-1297.

Stavropoulou V, Xie J, Henriksson M, Tomkinson B, Imreh S, Masucci MG. 2005. Mitotic infidelity andcentrosome duplication errors in cells overexpressing tripeptidyl-peptidase II. Cancer Res 65:1361-1368.

Tomonaga T, Matsushita K, Ishibashi M, Nezu M, Shimada H, Ochiai T, Yoda K, Nomura F. 2005.Centromere protein H is up-regulated in primary human colorectal cancer and its overexpressioninduces aneuploidy. Cancer Res 265 :4683 -4691.

Toyooka S, Toyooka KO, Maruyama R, Virmani AK, Girard L, Miyajima K, Harada K, Ariyoshi Y,Takahashi T, Sugio K, Brambilla E, Gilcrease M, Minna JD, Gazdar AF. 2001. DNA methylationprofiles of lung tumors. Mol Cancer Ther 1:61 -67.

Tse C, Xiang RH, Bracht T, Naylor SL. 2002. Human Semaphorin 3B (SEMA3B) located at chromosome3p21.3 suppresses tumor formation in an adenocarcinoma cell line. Cancer Res 62:542 -547.

van der Hout AH, Kok K, van der Veen AY, Osinga J, de Leij LFMH, Buys CHCM. 1989. Localization ofamplified c-myc and n-myc in small cell lung cancer cell lines. Cancer Genet Cytogenet 38:1-8.

Wei MH, Latif F, Bader S, Kashuba V, Chen JY, Duh FM, Sekido Y, Lee CC, Geil L, Kuzmin I, ZabarovskyE, Klein G, Zbar B, Minna JD, Lerman MI. 1996. Construction of a 600-kilobase cosmid clonecontig and generation of a transcriptional map surrounding the lung cancer tumor suppressor gene(TSG) locus on human chromosome 3p21.3: Progress toward the isolation of a lung cancer TSG.Cancer Res 56:1487-1492.

Complexity of functional analysis

79

Wistuba II, Lam S, Behrens C, Virmani AK, Fong KM, LeRiche J, Samet JM, Srivastava S, Minna JD,Gazdar AF. 1997. Molecular damage in the bronchial epithelium of current and former smokers. JNatl Cancer Inst 89:1366-1373.

Xiang R, Davalos AR, Hensel CH, Zhou XJ, Tse C, Naylor SL. 2002. Semaphorin 3F gene from human3p21.3 suppresses tumor formation in nude mice. Cancer Res 62:2637 -2643.

Zabarovsky ER, Lerman MI, Minna JD. 2002. Tumor suppressor genes on chromosome 3p involved in thepathogenesis of lung and other cancers. Oncogene 21:6915 -6935.

Chapter 3

Micro-array expression analysis of a cellline stably transfected with 3p21.3

sequences indicates the occurrence oftrans-regulation between originally

neighbouring genes and regulatory DNAsequences

Arja ter ElstBea E. HiemstraKrista A. KooiPieter van der VliesWytske KammingaKlaas KokCharles H.C.M. BuysFrans Gerbens

Department of Medical Genetics, University Medical Center Groningen, Groningen,The Netherlands

ABSTRACT

We have previously found a significant reduction of tumour growth upon injection innude mice of cells stably transfected with a PAC containing a centromeric sequenceof the 3p21.3 region commonly deleted in lung cancer, as compared to the tumourgrowth after injection of the parental cell line. To evaluate the transcriptionalconsequences of the introduction of this PAC into the cell line, we compared thegene expression levels of the transfectants with the gene expression levels of theparental cell line, using micro-array expression analysis. A complication, however, isthat injection of one of two transfectants with vector only sequences also led to asignificantly diminished tumour growth in nude mice and that array-CGH revealedthat all transfected clones tested had an increased chromosomal instabilitycompared to the parental cell line. When we compared the gene expression levels oftransfectant-1 containing the PAC with those of the parental cell line and did thesame for transfectant-2 containing the PAC and the parental cell line, we found thatonly a small proportion of significantly differentially expressed genes were present inboth comparisons. By comparing the gene expression data with the previouslyobtained array-CGH data, we were able to see a strong concordance of regions withDNA copy number changes –that varied between the transfectants- and regions withtranscriptional changes. This explains why only a small overlap occurred betweendifferentially expressed genes of both transfectants with the PAC. In addition, wefound that the 3p21.3 genes CACNA2D2 and TUSC4 were higher expressed intransfectants with the PAC than in the parental cell line. These two genes are notlocated on the PAC, but are normally in the 3p21.3 region neighbours of CYB561D2and PL6, that are located on the PAC. This suggests the occurrence of trans-regulation between genes and regulatory DNA sequences in the 3p21.3 region.

Chapter 3

82

INTRODUCTION

A 3p21.3 region containing 19 genes is considered as a lung cancer critical regionwith tumour suppressor activity (Wei ., 1996; Lerman and Minna, 2000). In aprevious study, we covered this region with overlapping P1 artificial chromosomes(PAC) and used these to transfect cells from GLC45, a small cell lung cancer(SCLC) cell line which readily causes tumours in nude mice. We found a significantreduction of tumour growth upon injection in nude mice of cells stably transfectedwith PAC 185A4, containing a centromeric part of the 3p21.3 critical region, ascompared to the tumour growth after injection of the parental cell line (A ter Elst, BEHiemstra, P van der Vlies, W Kamminga, AY van der Veen, I Davelaar, G J teMeerman, F Gerbens, K Kok, CHCM Buys, submitted). Injection of one of twotransfected clones with vector only sequences also led, however, to a significantreduction of tumour growth upon injection in nude mice. By array-CGH, we foundmultiple gains and losses in the transfectants with PAC185A4, as compared to theparental cell line. In both transfectants containing vector only sequences alsomultiple gains and losses were found, even though one of these transfectants did notshow a significantly reduced tumour growth upon injection into nude mice. Althoughthe transfectants had some gains and losses in common, the genomic profile oflosses and gains on the whole differed for each of the transfectants.

We wanted to evaluate the transcriptional consequences of the introductionof PAC 185A4. We, therefore, compared gene expression levels of the transfectantswith the gene expression level of the parental cell line GLC45, using micro-arrayexpression analysis.

Micro-array expression analysis

83

MATERIALS AND METHODS

Cell lineThe SCLC cell line GLC45 (Naylor et al., 1998) was cultured in RPMI-1640 (GIBCO -Invitrogen Inc, Carlsbad, Ca, USA) supplemented with 15% foetal calf serum and 0.1mM sodium pyruvate, 0.1 mM HEPES, penicillin/streptomycin, glutamine and 250µg/ml amphotericin B (all from GIBCO – Invitrogen) at 37° C in a 5% CO2 incubator.

RNA isolationTotal RNA was isolated with the RNeasy mini kit including DNAse treatment(Qiagen, Valencia, CA, USA) from the same cell population as used for injection ofnude mice in the tumourigenicity tests. The RNA yield and purity were determinedspectrophotometrically by measuring the absorbance at 260 nm and 280 nm. RNAintegrity was checked on 1% agarose gels.

Messenger-RNA amplification and Cy-dye couplingLinear amplification of mRNA was performed essentially according to a protocol ofthe Dutch Cancer Institute (www.nki.l.nl/nkidep/pa/microarray/protocols.html). Briefly,amplification started with first strand cDNA synthesis from 2 µg of total RNA, usingSuperscript II RT-polymerase (GIBCO) and a specific oligo(dT) primer containing a17 bp T7 polymerase recognition site (5'-GGCCAGTGAATTGTAATACGACTCACTATAGGGAGGCGGT-24-3') (Eurogentec, Seraing, Belgium). After second strandsynthesis, double-stranded cDNA was purified with the Qiaquick PCR purification kit(Qiagen) and the yield was determined spectrophotometrically. In vitro transcriptionwas performed with the T7 Megascript kit (Ambion, Huntingdon, UK) as described bythe manufacturer and aminoallyl-UTP (Ambion) was incorporated as described by ‘tHoen et al (2003). Amplified RNA (aRNA) was purified with the RNA clean upprotocol (Qiagen). Five µg of aRNA was labelled by coupling monoreactive Cyanine3 or Cyanine 5 fluorophores (Amersham Biosciences, Little Chalfont,Buckinghamshire, UK) to the aminoallyl-modified nucleotides. Labelled aRNA wasseparated from unincorporated Cyanine 3 or Cyanine 5 molecules using MicrospinG50 size exclusion columns (Millipore Corp, Amsterdam, The Netherlands) asdescribed by the manufacturer.

Experimental design for gene expression profiling. A balanced designincluding dye swap and self-self hybridisations was applied for micro-array based

Chapter 3

84

dual-colour gene expression analysis for both the transfectants with PAC185A4 andthe pPAC4 vector as described in Table 1.

Table 1. Experimental design of the dual-colour gene expression array analyses.

Experiment Slide # Cyanine 5 Cyanine 3PAC 185A4 1 GLC45 (230904)a GLC45 (230904)

2 GLC45 + PAC 185A4 transfectant-1 GLC45 + PAC 185A4 transfectant-13 GLC45 + PAC 185A4 transfectant-2 GLC45 + PAC 185A4 transfectant-24 GLC45 (230904) GLC45 + PAC 185A4 transfectant-15 GLC45 + PAC 185A4 transfectant-1 GLC45 (230904)6 GLC45 (230904) GLC45 + PAC 185A4 transfectant-27 GLC45 + PAC 185A4 transfectant-2 GLC45 (230904)

pPAC4 8 GLC45 (150704) a GLC45 (150704)9 GLC45 + pPAC4 transfectant-1 GLC45 + pPAC4 transfectant-110 GLC45 + pPAC4 transfectant-2 GLC45 + pPAC4 transfectant-211 GLC45 (150704) GLC45 + pPAC4 transfectant-112 GLC45 + pPAC4 transfectant-1 GLC45 (150704)13 GLC45 (150704) GLC45 + pPAC4 transfectant-214 GLC45 + pPAC4 transfectant-2 GLC45 (150704)

a Different subcultures of GLC45. By array-CGH no chromosomal alterations were found between both 4-week subcultures (our previous study).

Micro-array slides and hybridisationIn-house manufactured human oligonucleotide arrays were used containing theQiagen/operon 21,329 70-mer human gene-specific oligonucleotide set version 2.1extended with 4,000 negative and positive control features. The oligonucleotideswere printed in a concentration of 10 pM on Ultra-GAPS amino-silane coated slides(Corning BV. Life Sciences, New York, USA) using BioRobotics 10K quill pins withthe MicroGrid spotter (Isogen). Blocking, prehybridisation and hybridisation wereperformed as described by Hegde et al. (2000) with some modifications. In short,slides were blocked with ethanolamine at 52°C during 1 h. Prehybridisation wasdone with prewarmed prehybridisation buffer containing 0.5% bovine serum albumin(Sigma-Alldrich, St. Louis, MO, USA) at 52°C during 45 min Subsequently, the slideswere washed six times with preheated water (52°C), dried by centrifugation at 800rpm during 3 min and immediately used for hybridisation. The hybridisation sampleconsisted of the fluorescently labelled probe mixture and 30 µg poly-A (SigmaAlldrich) mixed with an equal volume of preheated (52°C) two times hybridisation

Micro-array expression analysis

85

buffer. This hybridisation sample was heated at 95°C for 3 min before it was appliedto the preheated slides. Hybridisation was performed in hybridisation chambers(Telechem International Inc, Sunnyvale, CA, USA) in a water bath at 52°C in thedark for approximately 48 h. Subsequently, slides were washed with 5 washsolutions under agitation: 1xSSC/0.2% SDS at 52ºC; 0.1xSSC/0.2% SDS at 52ºC;0.1xSSC at 52º; 0.1xSSC at room temperature and 0.01xSSC at room temperature.Each wash step lasted 5 minutes. Finally slides were dried by centrifugation at 800rpm during 3 min and scanned with an Affymetrix GMS428TM array scanner.

Micro-array data analysisFluorescent signal intensity data for each spot and for each fluorophore wereextracted from the scanned images of each micro-array slide using ImaGene version5.6 (BioDiscovery, El Segundo, California, USA). Signal intensity data were logtransformed and for each spot the Cyanine 5 signal intensity/Cyanine 3 signalintensity ratio was determined and subjected to print-tip lowess intensity dependentnormalisation using the Limma package from the Bioconductor project in R(http://bioinf.wehi.edu.au/limma). Since no dependency exists between both samples('t Hoen et al., 2004), normalised log-ratios were back transformed to log intensities.Further data analysis was performed using BRB ArrayTools v3.2 developed by Dr.Richard Simon and Amy Peng Lam (http://linus.nci.nih.gov/~brb/download.html).Basically, data was vigorously filtered to exclude control spots, empty spots, spotswith a high between pixel intensity variability and spots designated as bad by eye.Genes that had more than 50% missing data across all observations were excludedfrom the analysis. Genes significantly differentially expressed between transfectantsand the parental cell line were identified by a F-test using a randomised variancemodel and accounting for replicate readings of the same sample. Moreover, amultivariate permutation test (Reiner et al., 2003) was applied to account for a falsediscovery rate of no more than 10% in the set of significantly differentially expressedgenes. Regions with correlated expression biases for gene expression array dataand regions showing gains and losses for array-CGH data were determined usingthe cluster-along-chromosomes procedures as implemented in CGH-miner (Wang etal., 2005). The false discovery rate was restricted to 1% as compared to well-matched self-self hybridisations for expression array and three normal-normalhybridisations for array-CGH. Before clustering along individual chromosome arms,the signal intensities were smoothed with a moving average window of 7 gene-

Chapter 3

86

specific oligonucleotides for gene expression array data and 3 BACs for array-CGHdata.

Real-time RT-PCRFor quantitative real-time RT-PCR, first strand cDNA was generated by usingReady-To-GoTM You-Primed First-Strand Beads (Amersham Biosciences) and anoligo-dT primer (Eurogentec) as described by the manufacturer. Table 2 shows theprimers for each gene that were designed with the program Primer Express (AppliedBiosystems, Foster City, CA, USA). For each sample the PCR product yield wasanalysed in real time in triplicate in a final volume of 20 ul in a 384-well plate with theABI Prism 7700 Sequence Detection System (Applied Biosystems). Finally, PCRproducts were checked by electrophoresis in a 2% agarose gel. Real-time data wasanalysed with the Sequence Detection System software (SDS 2.1.1, AppliedBiosystems). For each gene the threshold cycle value (Ct) of every sample wasextracted from the amplification plots. The difference in expression between thetransfectant clone and the parental cell line sample was determined for each gene,using the ∆∆ median threshold cycle number (Ct) method. The corresponding foldchange in expression was calculated by the formula 2-∆∆Ct (Livak and Schmittgen,2001).

Table 2. Primer sequences for six genes located in the 3p21.3 critical region and forB2M used as a reference in real-time RT-PCR.

Primer pair Official Genesymbol

Forward primer sequence5’-3’

Reverse primer sequence5’-3’

QRT-CYB561D2 CYB561D2 GCCCTGTCCTCACCAGCTT GAGCTCATGGTTGGATCCTCTT

QRT-PL6 PL6 CTCCCCCGACGCTGTAACT TGGAGTAGCTGAGAGGATGTCAAG

QRT-CACNA2D2 CACNA2D2 CCACCGTTGCAGATTTCCTT TGGTAGATGAGGCCGTAGAGAAG

QRT-TUSC4-2 TUSC4 GCAAGAGGCATGTCTATCCTACGT CAGGGCTCAGGCTGCAGTATA

B2M B2M TGACTTTGTCACAGCCCAAGATA AATGCGGCATCTTCAAACCT

StatisticsStatistical significance was determined by analysis of variance (ANOVA) withScheffe’s F test, using the software programme SPSS.

Micro-array expression analysis

87

RESULTS

Gene expression analysisTo evaluate the transcriptional consequences of the introduction of PAC 185A4 andvector only sequences into GLC45 we compared the gene expression level of thetransfectants with the gene expression level of the parental cell line GLC45, usingmicro-array expression analysis. Since the genomic profile of losses and gainsdiffered for each of the transfectants, we decided to analyse each transfectantseparately, comparing it to the parental cell line. These analyses revealed that 502and 128 genes were significantly (p<0.0001) differentially expressed betweenGLC45 and transfectants-1 and 2 with PAC 185A4, respectively. For transfectantscontaining vector only sequences, 175 and 214 genes appeared to be significantly(p<0.0001) differentially expressed between the parental cell line and transfectants-1and 2, respectively.

Figure 1 shows an overlap of 45 genes between the differentially expressedgenes of transfectants-1 and 2 containing PAC 185A4 and an overlap of a more orless similar size of 37 genes between transfectants-1 and 2 with vector onlysequences. Of these 45 and 37 genes, 26 and 22, respectively, were either higherexpressed in both transfectants than in the parental cell line or lower expressed inboth transfectants. In our previous study a reduced tumour growth upon injection intonude mice, as compared to the tumour growth after injection of the parental cell linewas found for both transfectants with PAC 185A4 and for one of the transfectantswith vector only sequences. Therefore, we also determined the overlap ofdifferentially expressed genes between these three transfectants. Three genes,LOC145957, FOS and KYNU are similarly expressed in all three comparisons. FOSis higher expressed in each of the three transfectants, LOC145957 and KYNU arelower expressed in each of the three transfectants. Both FOS and KYNU did notoccur in the list of significantly differentially expressed genes of the transfectantwhich did not show a significantly reduced tumour growth upon injection into nudemice. LOC145957 did occur in this list, but not lower, but higher expressed in thistransfectant.

Chapter 3

88

Correspondence between DNA copy number alterations and gene expressionWe compared the previously obtained CGH-array data with the comparative geneexpression array data. Chromosomal plots revealed a strong concordance betweenregions with DNA copy number changes and regions with transcriptional changes(Fig. 2). A higher copy number corresponded with a higher expression, a lower copynumber with a lower expression.

Figure 1. Overlap of significantly (p<0.0001) differentially expressed genes from the comparisonsperformed in this study.Genes from each overlap are listed in the blocks. Genes in bold are either elevated or repressed in bothtransfectants as compared to the parental cell line GLC45. Genes that are underlined are part of agained region in both transfectants. A: Comparison of transfectant-1 with PAC 185A4 as compared to theparental cell line (502 genes) and transfectant-2 with PAC 185A4 as compared to the parental cell line(128 genes) B: Comparison of transfectant-1 with PAC vector only sequences as compared to theparental cell line (175 genes) and transfectant-2 with PAC vector only sequences as compared to theparental cell line (214 genes) C: Comparison of all transfectants that showed a reduction of tumourgrowth upon injection into nude mice, i.e. transfectants-1 and 2 containing PAC 185A4 and transfectant-2 containing vector only sequences.

Micro-array expression analysis

89

Figure 2. Correlated changes of genomic copy number and gene transcription in PAC transfectantscompared to the parental cell line GLC45.The panels show chromosomal profile log ration plots of array-CGH and corresponding array-based geneexpression data. Regions with significant changes in either the array-CGH data (gains and losses) orgene expression data as obtained by the cluster-along-chromosomes (CLAC) procedure are indicated ineach plot by horizontal bars.

Chapter 3

90

Expected higher expression of PAC 185A4 genesPAC 185A4 contains the genes PL6 and CYB561D2. In both transfectants with PAC185A4, CYB561D2 and PL6 were higher expressed than in the parental cell line(Fig. 3). CYB561D2 was 2.2 fold higher expressed in transfectant-1 with PAC 185A4and 4.5 fold higher expressed in transfectant-2. For PL6 the fold changes were 2.7and 6.7 for transfectants-1 and 2 with PAC 185A4, respectively compared to theirexpression in the parental cell line.

Higher expression of originally neighbouring genesAn analysis of the expression profiles of the transfectants with PAC 185A4 revealedthat the originally neighbouring 3p21.3 genes (Fig. 4), TUSC4 (coding for tumoursuppressor candidate 4) and CACNA2D2 (encoding calcium channel voltage-dependent, alpha 2/delta subunit 2) were each also higher expressed in onetransfectant with PAC 185A4 than in the parental cell line (Fig. 3).

Figure 3. Real-time RT-PCRvalidation of transcriptionalchanges in PAC transfectants ascompared to the parental cell lineGLC45 for four genes located inthe 3p21.3 critical region.Bar charts show fold changes withstandard deviation. A: Expressionarray B: Real-time RT-PCR.

Micro-array expression analysis

91

Validation of gene expression results for genes of the 3p21.3 critical region byreal time RT-PCRTo verify the gene expression levels for the genes of the 3p21.3 critical region by anindependent method, we performed real time RT-PCR analysis. For both of thegenes, CYB561D2 and PL6, located on PAC 185A4, and for both neighbouringgenes, CACNA2D2 and TUSC4, the results as obtained by either method werehighly concordant, as shown in Fig. 3.

Figure 4. A map of the genes,CYB561D2 and PL6 located on PAC185A4 and neighbouring genes,TUSC4 and CACNA2D2 partly locatedon PAC 185A4.The position of the genes and PACs isshown.

Chapter 3

92

DISCUSSION

To evaluate the transcriptional consequences of the introduction into the SCLC cellline GLC45 of PAC 185A4 containing two genes from the centromeric region of the3p21.3 lung cancer critical region, we compared the gene expression level of thetransfectants with the gene expression level of the parental cell line by micro-arrayexpression analysis. After analysing every transfectant separately and comparingthe significantly differential expressed genes of transfectants with PAC 185A4 witheach other we found an overlap of 45 genes. The overlap of significantly differentialexpressed genes between the transfectants with vector only was 37 genes. Thesesmall overlaps might be a consequence of the earlier found genomic instabilitycausing large differences between the genomic profiles of the transfectants.

To evaluate if indeed the genomic instability has a strong influence on theexpression of the genes we compared the gene expression array data with thepreviously obtained array-CGH data. We found that regions with DNA copy numberchanges completely overlapped with regions containing transcriptional changes. Ahigh correlation between DNA copy number and gene expression has already beensuggested before (Pollack et al., 1999), but positional association has so far onlybeen found for highly amplified regions (Hughes et al., 2000; Kauraniemi et al.,2001; Monni et al., 2001; Hyman et al., 2002). We found a highly consistentassociation between deletions and low-level amplifications with lower and highergene transcription, respectively, in individual comparisons between a parental cellline and derived transfectants.

As expected, also both genes located on PAC 185A4, PL6 and CYB561D2,belonged to the significantly differentially expressed genes of transfectants with PAC185A4 as compared to the parental cell line. Both genes were higher expressed inthe transfectants. In addition, we found that the two genes TUSC4 and CACNA2D2,that flank CYB561D2 and PL6, respectively, in the original 3p21.3 sequence, werealso significantly higher expressed in the transfectants containing PAC 185A4, ascompared to the parental cell line. Despite the low fold changes found in the arrayanalysis, we were able to confirm the significant differences in expression by real-time RT-PCR. None of these genes showed transcriptional changes in thetransfectants containing vector only sequences. Therefore, it is likely that theobserved higher expression of these genes is caused by sequences located on PAC185A4. Since in the two transfectants, PAC 185A4 has integrated into the GLC45

Micro-array expression analysis

93

genome at different sites, not on the short arm of chromosome 3, this increasedexpression of genes residing in the 3p21.3 lung cancer critical region, caused byintegrated PAC 185A4 sequences, must be due to trans-regulation.

Our results indicate that most of the differential gene expression found in thetransfectants is caused by genomic instability as compared to the parental cell line.Among the 26 genes with a similar change of expression that occur in the overlap of45 genes between the significantly expressed genes of the transfectants with PAC185A4, the large majority is not in regions showing a strong correlation betweenDNA copy number alterations and transcriptional changes. Among these may begenes whose expression is directly or indirectly regulated by sequences present onPAC 185A4.

ACKNOWLEDGEMENTSThis work was supported by a grant from the Dutch cancer society RUG 2000-2317.

Chapter 3

94

REFERENCES

't Hoen PA, de Kort F, van Ommen GJ, den Dunnen JT. 2003. Fluorescent labelling of cRNA formicroarray applications. Nucleic Acids Res 31:e20.

't Hoen PA, Turk R, Boer JM, Sterrenburg E, de Menezes RX, van Ommen GJ, den Dunnen JT. 2004.Intensity-based analysis of two-colour microarrays enables efficient and flexible hybridizationdesigns. Nucleic Acids Res 232:e41.

Hughes TR, Roberts CJ, Dai H, Jones AR, Meyer MR, Slade D, Burchard J, Dow S, Ward TR, Kidd MJ,Friend SH, Marton MJ. 2000. Widespread aneuploidy revealed by DNA microarray expressionprofiling. Nat Genet 25:333 -339.

Hyman E, Kauraniemi P, Hautaniemi S, Wolf M, Mousses S, Rozenblum E, Ringner M, Sauter G, MonniO, Elkahloun A, Kallioniemi OP, Kallioniemi A. 2002. Impact of DNA amplification on geneexpression patterns in breast cancer. Cancer Res 62:6240 -6244.

Kauraniemi P, Barlund M, Monni O, Kallioniemi A. 2001. New amplified and highly expressed genesdiscovered in the ERBB2 amplicon in breast cancer by cDNA microarrays. Cancer Res 61:8235 -8240.

Lerman MI, Minna JD. 2000. The 630-kb lung cancer homozygous deletion region on humanchromosome 3p21.3: identification and evaluation of the resident candidate tumor suppressorgenes. The International Lung Cancer Chromosome 3p21.3 Tumor Suppressor Gene ConsortiumCancer Res 60:6116-6133.

Livak KJ, Schmittgen TD. 2001. Analysis of relative gene expression data using real-time quantitativePCR and the 2(-Delta Delta C(T)) Method. Methods 25:402 -408.

Monni O, Barlund M, Mousses S, Kononen J, Sauter G, Heiskanen M, Paavola P, Avela K, Chen Y,Bittner ML, Kallioniemi A: Comprehensive copy number and gene expression profiling of the17q23 amplicon in human breast cancer. Proc Natl Acad Sci U S A 2001 May 8 ;98 (10):5711-5716 (2001).

Naylor SL, Davalos AR, Hensel CH, Xiang RH. 1998. Human semaphorin IV gene from 3p21.3suppresses tumor growth in nude mice and attenuates apoptosis. Am J Hum Genet 63:A80.

Pollack JR, Perou CM, Alizadeh AA, Eisen MB, Pergamenschikov A, Williams CF, Jeffrey SS, Botstein D,Brown PO. 1999. Genome-wide analysis of DNA copy-number changes using cDNA microarrays.Nat Genet 23:41-46.

Reiner A, Yekutieli D, Benjamini Y. 2003. Identifying differentially expressed genes using false discoveryrate controlling procedures. Bioinformatics 19:368 -375.

Wang P, Kim Y, Pollack J, Narasimhan B, Tibshirani R. 2005. A method for calling gains and losses inarray CGH data. Biostatistics 6:45 -58.

Wei MH, Latif F, Bader S, Kashuba V, Chen JY, Duh FM, Sekido Y, Lee CC, Geil L, Kuzmin I, ZabarovskyE, Klein G, Zbar B, Minna JD, Lerman MI. 1996. Construction of a 600-kilobase cosmid clonecontig and generation of a transcriptional map surrounding the lung cancer tumor suppressorgene (TSG) locus on human chromosome 3p21.3: Progress toward the isolation of a lung cancerTSG. Cancer Res 56:1487-1492.

Chapter 4

Analysis of a new homozygous deletion inthe tumour suppressor region at 3p12.3

reveals two novel intronic non-coding RNAgenes

Debora Angeloni*Arja ter Elst**Ming Hui Wei*Anneke Y. van der Veen**Tineke Timmer**Michael I. Lerman*Charles H.C.M. Buys**

*Laboratory of Immunobiology, CCR, NCI-F, Frederick, Maryland, USA**Department of Medical Genetics, University Medical Center Groningen, Groningen,The Netherlands

ABSTRACT

Homozygous deletions or loss of heterozygosity (LOH) at human chromosome 3p12band are consistent features of lung and other common malignancies suggesting thepresence of a tumour suppressor gene(s) (TSG) at this location. Only one gene,DUTT1 (Deleted in U Twenty Twenty) was so far cloned from the overlapping regiondeleted in several lung and breast cancer cell lines (U2020, NCI H2198, HCC38).DUTT1 is the human ortholog of the fly gene ROBO that has homology with NCAMproteins. Extensive analyses of DUTT1 in lung cancer did not reveal any mutations,suggesting that another gene(s) at this location could be associated with lung cancerinitiation and/or progression. We report here the discovery in the SCLC cell lineGLC20 of a new small, homozygous deletion nested in the known, overlapping,critical region. The deletion was PCR-characterised using several polymorphicmarkers and covered by three overlapping P1 phage clones. Fiber-FISHexperiments using those clones defined precisely the genomic location and the sizeof the deletion (approx. 130 kb). Comparative genomic sequence analysis revealedshort sequence elements highly conserved among mammalian genomes and in thechicken genome. The discovery of two EST clusters in the deletion led to theisolation of two non-coding RNA (ncRNA) genes. We further consider the possibilitythat these ncRNA and other highly conserved sequence elements may representmiRNA targets whose potential as novel 3p12 lung cancer TSG will be evaluatedthrough subsequent mutation and functional studies.

Chapter 4

98

INTRODUCTION

Loss of function of tumour suppressor genes (TSG) is a fundamental genetic changeinvolved in the origin and pathogenesis of human tumours (Knudson, 1971;Marshall, 1991). TSG have a recessive mode of action, therefore both copies needto be inactivated to produce a phenotypic effect. That frequently happens when oneallele is hit by a mutation or silenced through promoter region hypermethylation andthe other is lost due to a large chromosomal deletion. In some instances both allelesmight be inactivated by genetic loss as a consequence of a homozygous deletion.Sub-lethal homozygous deletions are frequently smaller than heterozygouscounterparts, a characteristic that makes them a useful tool for localising tumoursuppressor genes.

Cytogenetic and molecular deletion mapping studies have long implicatedchromosome bands 3p25-26, 3p21.3 and 3p12-14 as harbouring tumour suppressorgenes involved in multiple forms of human cancers including lung cancers (Whang-Peng et al., 1982; Kok et al., 1987; Kok et al., 1997; Zbar, 1989; Zabarovsky et al.,2002).

In 3p14.2 the FHIT gene is subject to homozygous deletions and alterationsof the mRNA in many sporadic cancers (Huebner et al., 1998). Frequent allele lossat the FHIT locus has been found in low-grade breast cancer. 3p12 is a particularlysignificant region (Lerman and Minna, 2000), as demonstrated also by functionalstudies of Lott et al. (1998) and Lovell et al. (1999). Rabbitts et al. (1990) reported ahomozygous deletion at the D3S3 locus in the U2020 cell line. It spans about 8 Mband is flanked by the microsatellite markers D3S1284 and D3S1276 (Latif et al.,1992; Drabkin et al., 1992). Some other nested or overlapping homozygousdeletions were reported in this region. Todd et al. (1997) reported a homozygousdeletion that overlaps the U2020 region and is flanked by microsatellite markersD3S1254 and D3S1776. Another overlapping homozygous deletion was found in thebreast cancer cell line HCC38. It spans about 5 Mb and is flanked by themicrosatellite markers D3S2537 and D3S2527 (Sundaresan et al., 1998a). In theSCLC cell line NCI H219x a much smaller deletion was found that contains themicrosatellite markers D3S1274, D3S2498, D3S4492 (telomere to centromere,Sundaresan et al., 1998a).

From this low gene-density region of chromosome 3, the DUTT1 (Deleted inU twenty twenty) gene was isolated (Sundaresan et al., 1998b). DUTT1 gene

Novel non-coding RNAs

99

expression is impaired by hypermethylation of the promoter in kidney and breastprimary cancers but to a much lesser extent in lung cancers.

In this paper we describe a new, homozygous deletion at 3p12 discovered inthe small cell lung cancer (SCLC) cell line GLC20. It affects exon 2 and flankingintrons of the ROBO1/DUTT1 gene. It spans approximately 130 kb around theD3S1274 microsatellite marker and partly overlaps with the NCI-H219x deletion. Inthe second intron of DUTT1 we discovered two novel transcripts with the sameorientation as DUTT1. Both are polyadenylated, show small ORFs without anyknown homologues or orthologues and likely do not encode proteins. Both areputatively noncoding (nc)RNA genes that will be evaluated for their role in lungcancer suppression. In the deleted region were also found some sequence elementshighly conserved between the human and chicken genome (Bejerano et al, 2004;Hillier et al.2004).

Chapter 4

100

MATERIALS AND METHODS

Human DNASigned, informed consent was obtained from all DNA donors in the study, accordingto the NCI institutional review board-approved protocol.

Cell Lines and DNALung cancer cell lines NCI-H750, NCI-H2198 and NCI-H1450 were obtained fromATCC (Manassas VA). U2020 DNA was kindly provided by Dr. Pamela Rabbits(MRC, Cambridge, UK). GLC20 is a SCLC cell line established from a primarytumour biopsy (De Leij et al., 1985) and known to harbour a circa 440 kbhomozygous deletion at 3p21.3 (Kok et al., 1994).

EST clonesEST clones were purchased from the I.M.A.G.E. Consortium (http://image.llnl.gov/).

PCRPCR primers were from BioServe Biotechnologies (Laurel, MD). PCR reactions wereperformed in a total reaction volume of 12.5 µl, containing 100 ng of genomic DNA,12.5 pmol of each primer, 200 µM dNTPs, 1.5 mM MgCl2. The PCR cycles were asfollow (with the appropriate annealing temperature, Ta, indicated in Table 1): 95°C, 5min; (95°C, 30 sec, Ta°C, 30 sec, 72°C, 30 sec) for 35 cycles; 72°C, 7 min. PCRproducts were run on 3% or 4 % NuSieve 3:1 agarose gel (Cambrex, Baltimore, MD)and stained with Ethidium Bromide (SIGMA, St. Louis, MO).

Southern BlotSouthern blot was done according to Sambrook et al. (1998). Briefly, genomic DNAsamples were digested overnight with EcoRI in the presence of 1% spermidine,precipitated with Na Acetate, re-suspended in TE buffer, load on 1% agarose geland run overnight at 30V. The gel was then denatured and DNA transferredovernight in NaOH 0.4 N. The membrane was washed in 2X SSC and dried for threehrs. at 80°C in a vacuum oven. Hybridisation was carried out at 60°C overnight.Washes were done in 2X SSC, 1% - 0.1% SDS at 65°C.

Novel non-coding RNAs

101

Table 1. List of PCR primers (with annealing temperature) designed on ESTs aroundmicrosatellite markers, microsatellite markers, P1-clone ends and cDNA sequencesof ROBO1.

EST Forward primer 5’-3’ Reverse primer 5’-3’ Ta (ºC)D81026 GGAATATAAAGCTGGAGACTGTGG ACGAGCTTACCTCCCATGTTAACT 62H77734 ATCCAGTGGAGAGCCATCTTTCTT CTTAAAGCCTGTATTGCAAGAGCC 64T02864 TCCTTATAGCAGTGTGGGAATG CAAAATATAAACCCAGATCCACC 62T02956 CAGGATGCTCAACCAGCAGGT GTTTTCTTCAAAGTAGCCAATCCGC 62T64897 TTCAATCAAGCTGTGGCATAGAGG AGTTCAAGGTGACAGTGAACTACG 62W91914 GCTCTTCATCACATTTTCCCATCG GTTCCCTGTCTTCCTATATTCCC 62Z1019 GTGCAACCCCTTATTCAGAATCC TATGTCTACCCCTGTTTCTGCTC 64

H51703 CTGTACTACAGGGAATCTCTC CTTCCTTTGGGTCTGTTCAG 60AW861295 GGGATAAAGCAGTCACACAG TCTCATCACATGGCCTGTTC 62

MicrosatelliteD3S1577 TCAAAAGTTGCATCGC TCCATTACAATCCCCTG 60D3S3681 GTGAGAACCATTTGGGGCAG GGCGAGCTATCTGTCAGGG 60D3S1604 CACCATTGTAAGAGGCTTCA AAATTGACGCATAAAATTGTG 62

D3S3 CAGAAGGACATATTCCCATTTG GCAGTTTCCTCTAGCTTTTACT 55D3S3049 AAAGACACAAGGGGTTTTAGG TTGCACATTCCATGAACATC 55D3S1274 TTATACATCAGTCTCTGGGAAACAC TACTGTGCATATAGGTTCCTGTGA 62D3S3507 TCCAGCCCTTATACCTACTCTC TGGAATCAAGACAAGACTGAAC 60D3S2563 ATATTTTAGCTGGACTCGGTGCTG TTGCAGACTCTGCTATTGGCCTG 60D3S2530 ACAGGCAATTGGTGAAGCATG GATTTATCCATGGCCTCTGCTC 62P1 Clone

P1-97 CCTGAGTTTGATTTGCATGTGTCT GAGCTACAGTTCAAGATGAGATTG 61P1-98 AATGAAATCTTCGAAGTTGC ATAGCATATATTGACAG 55P1-80 GTGATAGCATATATTGACAG TTTTAAGGAAAACCATCGCC 55P1-97 TGGTAGCGTGAAACTTGCCTACCAG CAGTGTGGAAAGTGGGAAGGTAGA 64P1-98 TGTCGGTTGTTTCAGCTCTGC GGGAGTCATTTTTCCCTCAGG 50P1-80 GAGGTGGTGGCTTGAAATGC ATGAGAACCCAGATGAC 60

ROBO1 cDNA71-259 ATCCTCTCTGCCCTTCTCTG ACACTCGCACGTCTTCTGGG 62

601-812 CACATTGTGAGGGCGCAC TCAGGGCAATTACTCGTCG 641021-1345 TCCCGTCTTCGTCAGGAAG TGGCTACTTCCAGCGATGC 641351-1490 CTTCGGGATGACTTCAGACA TTATCATCCAGTGGAGAGCC 621510-1620 CGAGGAGGAAAGCTCATGAT AGTCAGCTCGGCTACTTCAC 621623-1776 GCTCTTACTGAACTCCTAAA CGTAGAAATGCTGAGTTGAC 602341-2811 TAGTGCCCCATCAAAACCTG GGATCTGATATTTGGCTTGG 625291-5376 ATCTTCCACCACCTCCTGTG TAGAAGGGAGTTTTGGCACC 606371-6609 CAAACAATTCGAATGGGGTAG GGTCATTAAAAACATCCACTTG 62

Chapter 4

102

Northern BlotNorthern blot was performed with Clontech Human Multiple Tissue Northern Blots7760-1 (Clontech, Palo Alto, CA). Hybridisation was carried out at 42°C overnight.Washes were done first in 2X SSC, 1% SDS and then 2X SSC, 0.1% SDS at 65°C.

Probe labellingThe W91914 and H51703 cDNAs were excised with EcoRI/PacI from the modifiedpolylinker of the pT7T3D vector (Pharmacia, North Peapack, NJ) and radioactiveprobes were prepared by 32P-labeling with random primers (Rediprime DNALabelling System, Amersham, Arlington Heights, IL).

P1 library screeningThe P1 Human Library (Genome Systems, St. Louis, MO) was screened initially withPCR primers designed on the EST W91914 and three clones were isolated: P1-97,P1-98, P1-99. A second round of PCR, with primers designed on the Sp6-end of P1-98, produced the fourth clone P1-80.

Fiber Fluorescence In Situ Hybridisation (fiber-FISH)Preparations for fiber FISH analysis were obtained essentially according to Giles etal. (1997). GLC20 cells in culture were spun down by centrifugation at 1200 rpm for10 min and resuspended in distilled water to a concentration of a few, (1 to 5)10exp5, cells per ml. A volume of 100 µl was pipetted onto a microscope slidecoated with a 5% 3-aminopropylethoxy-silane in acetone. Coating was achieved byincubating the slide in that solution for 30 min, after which they were washed withdistilled water, allowed to air dry and stored at 4°C until use. The cell suspensionwas spread over the coated slide using the edge of a coverslip and dried with ahairdryer. Exposure of chromatin threads from the nuclei was obtained by applyingtwo drops of 50 µl from a 0.5% SDS, 50mM EDTA, 0.2M TrisHCl, pH 7.0lysissolution on 24mm x 60mm cover slips. The microscope slides were then put upsidedown (with the cells under) on top of the coverslips, slides were then turned and keptlike that for 30 sec (with the coverslip up). Coverslips were then gently slid off andpreparations were dried with a hairdryer.

Bicolor FISH analysis was performed on these preparations usingdifferentially labelled P1 phages and routine FISH procedures, essentially asdescribed by Driesen et al. (1991).

Novel non-coding RNAs

103

Sequencingsequencing reactions were done automatically (ABI 373 Stretch Automated DNASequencer, Applied Biosystems, Foster City CA).

Computational analysisWWW-based Servers and Databases were used to analyse genomic, cDNA andprotein sequences. Global and pairwise sequence alignments by BLAST athttp://www.ncbi.nlm.nih.gov/BLAST/ and BLAT: http://genome.ucsc.edu/cgibin/hgBlatMultiple sequence alignment with ClustalW at:http://www.ebi.ac.uk/clustalw/index.html Search of CpG island with CpG Plot at:http://www.ebi.ac.uk/emboss/cpgplot/index.html?Analysis of non-redundant sets of gene-oriented clusters: athttp://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=unigeneGenome browsers and annotations by The Human Genome Browser Gateway:http://genome.ucsc.edu/cgi-bin/hgGateway, Acembly: at:http://www.ncbi.nih.gov/IEB/Research/Acembly/index.html and Ensembl at:http://ensembl.org.Prediction of complete gene structures in genomic sequences (exons, introns,promotersand poly-adenylation signals) by:GenScan at: http://genome.dkfz-heidelberg.de/cgi-bin/GENSCAN/genscan.welcome.plAnalysis of protein features: Psort at http://psort.nibb.ac.jp and Pfam athttp://pfam.wustl.edu.Search of possible miRNA target sites by: http://diana.pcbi.upenn.edu/cgi-bin/micro_t.cgiAnalysis of known miRNA sequence at: The miRNA Registry,http://www.sanger.ac.uk/Software/Rfam/mirna/index.shtml

Chapter 4

104

RESULTS

Discovery of a new homozygous deletion in 3p12.3We performed PCR experiments on genomic DNA from a panel of lung cancer celllines with primers designed on seven ESTs (namely T02956, H77734, D81026,T02864, T64897, W91914 and Z41019, data not shown, summary in Table 2A) thatmap between the genetic markers D3S1274 (alias AFM154xa7, Z16684) andD3S1604 (alias AFM316vc1, Z24325). Among the cell lines analysed, only U2020and NCI H2198 were known to harbour a homozygous deletion at 3p12.AII markers, except T02956 and D81026, were found deleted in the U2020 cell line(data not shown, summary in Table 2A). W91914 was found deleted also in thesmall-cell lung cancer cell lines H2198 and GLC20 (Fig.1A), the latter not known toharbour deletions in this region of 3p. This serendipitous result was confirmed bySouthern blotting experiments (Fig.1B). The deletion was named ‘GLC20-3p12deletion.

Characterization of the deletion breakpointsThe genomic location and extension of the GLC20-3p12 deletion were definedthrough PCR experiments with primers designed on several microsatellite geneticmarkers (Table 2B, Fig. 2). The deletion was found to be located around theD3S1274 marker, both in NCI-H2198 and GLC20 cell lines (Fig. 2).To address the issues of how large the deletion is, we looked for genomic clones thatwould represent it entirely and performed a PCR screening of a Human P1 Library(Genome Systems, St. Louis, MO).

With PCR primers designed on the W91914 EST three clones were isolated,identified as P1-97, P1-98 and P1-99. A fourth clone, identified as P1-80, wasnecessary to cover the deletion in the GLC20 cell line. The fourth P1 clone wasisolated after a second PCR screening of the same P1 library, using primersdesigned on the Sp6-end of P1-98 (screening data not shown). Both ends of eachP1 clone were sequenced. PCR reactions performed with primers designed on bothends of each P1 clone (Fig. 3, Table 2C) allowed the construction of a contig asfollows (schematically represented in Fig. 5): P1- 97 contains the proximal boundaryof the deletion and is centromeric to P1-98 that overlaps only for a small region withP1-80. (P1-99 overlaps extensively with P1-97). P1- 80 contains the telomericboundary of the deletion.

Novel non-coding RNAs

105

Table 2. Results of PCR experiments of ESTs, markers, P1-clones and ROBO1sequences.A) ESTs designed around the microsatellite markers D3S1274 and D3S1604. EST

W91914 was found homozygously deleted in the NCI-H2198 small cell lungcancer cell line.

B) D3S genetic markers with regard to location of microsatellites in the P1-clonecontig and in the lung cancer cell lines NCI-H1450, NCI-H2198, GLC20 andU2020.

C) P1-clone ends used to build up the contig.D) CDNA sequences of the ROBO1/DUTT1 gene (nucleotides position as in

GenBank Z95705). Exon 2 was found deleted in the small cell lung cancer cellline GLC20.

U2020 H2198 H1450 GLC20 P197 P198 P199 P180A EST

D81026 + + nd nd nd nd nd ndH77734 - + nd nd nd nd nd ndT02864 - + nd nd nd nd nd ndT02956 + + nd nd nd nd nd ndT64897 - + nd nd nd nd nd ndW91914 - - + - + + + -Z1019 - + nd nd nd nd nd nd

H51703 - - nd - - - - +AW861295 - - nd - - + - -

B MicrosatelliteD3S1577 - + + + - - - -D3S3681 - + + + - - - -D3S1604 - + + + - - - -

D3S3 - + + + - - - -D3S3049 - - + - - + - -D3S1274 - - + - + + + -D3S3507 - + + + - - - -D3S2563 - + + + - - - -D3S2530 nd + + + - - - -

C P1 CloneP1-97 - - + + + - + -P1-98 - - + - - + - +P1-80 - - + - - + - +P1-97 - - + - + + + -P1-98 - - + - + + + -P1-80 - - + + - - - +

D ROBO1 cDNA71-259 - nd nd + - - - -

601-812 - nd nd + - - - -1021-1345 - nd nd - - + - +1351-1490 - nd nd + - - - -1510-1620 - nd nd + - - - -1623-1776 - nd nd + - - - -2341-2811 - nd nd + - - - -5291-5376 - nd nd + - - - -6371-6609 - nd nd + - - - -

Chapter 4

106

Other ESTs (Fig. 4 and Fig. 5) were positioned in the contig by PCR. ESTAW861295 was also shown by PCR to be homozygously deleted in GLC20-3p12. Itwas mapped on P1-98 (Fig. 4) but not investigated any further because a BLASTsearch did not retrieve other overlapping EST clones.

To corroborate the PCR data and verify the contig location with respect tothe centromeric and telomeric ends of 3p, the P1 clones were used for a fiber FISHexperiment on normal DNA and DNA from GLC20 cell line (Fig. 6), using referralclones indicated as P1-26 and P1-27, previously isolated by Latif et al., (1992).Seeding in BLAT the sequences of P1-97 Sp6-end and P1-80 T7-end (depositedunder GenBank Accession number DQ100613 and DQ100614 respectively), wewere able to retrieve their exact genomic location (respectively: 78.918,087-78.918,453 and 79.117,734-79.118,091; Human Genome Browser, May 2004, hg 17assembly, http://genome.ucsc.edu/cgi-bin/hgGateway) and measure the extension ofthe deletion, which must be less than 200.005 bp (i.e. the distance between the twoanchoring markers P1-80 T7-end and P1-97 Sp6-end). By means of Fiber-

Figure 1. A novel homozygous deletion at 3p12.3 inthe small-cell lung cancer cell lines GLC20 andH2198.A) Genomic PCR with W91914 primers on normal(CEPH) DNA (lane 1 and 2), H740 (lane 3), H1450(lane 4), GLC20 (lane 5), U2020 (lane 6), NCI-H2198(lane 7), 100-bp ladder (lane 8, Invitrogen). 3%NuSieve agarose gel (Cambrex). B) Southern blotperformed with 10 µg of genomic DNA digested withEcoRI. Probe: EST W91914. Lane 1: GLC20, lane 2and 3: normal individuals, lane 4: molecular marker(Invitrogen). The two normal individuals differ for anEcoRI polymorphism.

Novel non-coding RNAs

107

Fluorescent In Situ Hybridisation experiments it was subsequently found to beapproximately 110 to 130 kb (see legend of Fig. 6)The genomic sequence of the deleted region contained between the two anchoringmarkers, was deposited under provisional GenBank Accession numberbankit729386.

Figure 2. PCR experiments toinvestigate the position of severalmicrosatellite markers with respect tothe GLC20-3p12 homozygous deletionand the four P1 clones that cover thedeletion. Microsatellites are ordered (topto bottom) from the more telomeric tothe more centromeric. 4% NuSieveagarose gel stained with EthidiumBromide. M.M.: molecular weight marker(100 bp ladder or 1 Kb ladder,Invitrogen). PCR primers and conditionsare listed in Table 1B.

Chapter 4

108

Figure 3. PCR experiments withprimers designed on the ends ofeach P1 clone were performed tobuild a P1 clone contig and to findout which clone would cover thedeletion boundaries. The small-cell lung cancer cell lines GLC20,U2020 and NCI-H2198 wereinvestigated. 4% NuSieve agarosegel stained with Ethidium Bromide.M.M.: molecular weight marker(100 bp ladder or 1 Kb ladder,Invitrogen). PCR primers andconditions are listed in Table 1C.

Novel non-coding RNAs

109

Figure 4. PCR experiments withprimers designed on the ESTsAW861295 and H51703. BothESTs are deleted in the GLC20-3p12 homozygous deletion andrepresented in the P1 contig. 4%NuSieve agarose gel stained withEthidium Bromide. M.M.: molecularweight marker (100 bp ladder or 1Kb ladder, Invitrogen). PCRprimers and conditions are listed inTable 1A.

Chapter 4

110

Figure 6. Fiber - Fluorescent Is Situ Hybridisationwith P1 clones. The yellow spots result frommerging of green and red labelling and showregions of overlapping between two clones Thereferral clone P1-27 is centromeric to P1 -97.Therefore the contig is oriented with P1-97 as themost proximal and P1-80 as the most distal(telomeric) clone. A) P1-97 and P1-98 showing anoverlap of an estimated size of about 20 kb whenhybridised to fibers from an EBV-transformedlymphoblastoid cell line. B) P1-98 and P1-80hardly showing an overlap when hybridised tofibers from the same cell line. C) size comparisonbetween the part of P1-97 hybridising to fibersfrom the cell line GLC20 and the cohybridisedcosmid cosD8 (Kok et al., 1995) showing that thepart of P1-97 extending beyond the deletion issomewhat larger than the 40 kb cosmid insert. D)P1-97 and P1-80 hybridised together with themore centromeric P1-27 to fibers from thelymphoblastoid cell line, showing the orientationof P1-97 and P1-80 with respect to thecentromere and the 3p telomere. Sincecohybridisation of P1-80 and cosD8 to fibers fromGLC20 gave a picture very similar to C (i.e. bothP1 phages extend about 45 kb beyond thedeletion) the size of the deletion will be about 3 xthe P1 insert size of 80 kb (cf. Fig. 3), minus theoverlaps of together about 20 kb, minus theextending P1 parts of about 2 x 45 kb. Thedeletion size will thus be approx. 130 kb.

Figure 5. Drawing that summarises all results obtained with PCR and fiber-FISH experiments, showingthe genomic location of the GLC20-3p12 homozygous deletion. Coloured rectangles: sequenced ends ofP1 clones. Coloured circles: positive results of specific PCR experiments indicating ‘anchoring points’ ofsequences to one another, used to build up the contig.

Novel non-coding RNAs

111

Status of the DUTT1 gene in the GLC20-3p12 homozygous deletionUsing a set of PCR primers (Table 2D, Fig.7) designed on the exonic sequence ofthe ROBO1/DUTT1 gene, we investigated to what extent the GLC20-3p12homozygous deletion affects the genomic structure of this gene in the GLC20 cellline. We found that the exon corresponding to cDNA residues 1021-1345 of DUTT1(variant 2, GenBank Z95705, NM_133631) is lost (Fig. 7) due to the small deletionthat spreads across exon 2 and is comprised entirely between introns 1 and 2 (Fig.5). These data were confirmed by RT-PCR on DUTT1 cDNA prepared from GLC20cells. The deletion causes the loss of amino acids 19-128 (as described for the cellline H219x by Sundaresan et al., 1998b). These data were also confirmed by PCRon YAC clones 912A11, 15HC9 and 35AH8 (data not shown). The coverage of thenew deletion in terms of BAC clones was determined in silico with electronic PCRusing the P1-ends sequence as probes (Fig. 5). When we compared ourexperimental results with the Human Genome Browser database (May 2004, hg 17assembly, http://genome.ucsc.edu/), we found a perfect match for the location of allmarkers. The new homozygous deletion in GLC20-3p12 is nested into but muchsmaller than the 8 Mb deletion of the U2020 cell line or the 5 Mb of the HCC38 cellline. It partly overlaps with H219x whose exact size was not determined but that isknown to be internal to DUTT1 (Sundaresan et al., 1998a).

The new deletion harbours two previously unknown transcripts –Computational analysisSeveral ESTs (some of which adenylated) map in the region hit by the new deletionin 3p12.3. We focused on two clusters and picked one from each, namely W91914and H51703, for further analysis. Several SAGE entries support these ESTs thatoriginated from different cDNA libraries. Seeding the sequence of W91914 andH51703 in BLAST (http://www.ncbi.nlm.nih.gov/BLAST/) against the human ESTdatabase, we were able to assemble two separate clusters and analyse theirsequence with GenScan (http://genes.mit.edu/GENSCAN.html) looking for predictedexons.

Cluster of W91914The cluster corresponding to the initial probe W91914 comprises now eight ESTs(namely: W91914, W94988, H90477, H90421, AI078492, AA639329, AA777646,CR740005) isolated from Fetal Liver and Spleen, Head and Neck carcinoma and

Chapter 4

112

Figure 7. Exon 2 of the DUTT1 gene(cDNA 1021-1345, as in GenBankZ95705) is homozygously deleted inthe GLC20-3p12 deletion. PCRexperiments with primers (listed inTable 1D) designed on some exons atthe 5’-end of DUTT1. The GLC20-3p12 homozygous deletion bridgesacross exon 2 to introns 1 and 2. 4%NuSieve agarose gel stained withEthidium Bromide. M.M.: molecularweight marker (100 bp ladder or 1 Kbladder, Invitrogen).

Novel non-coding RNAs

113

Mammary carcinoma libraries. Noteworthy, this cluster aligns (see at the HumanGenome Browser, May 2004, http://genome.ucsc.edu/cgi-bin/hgGateway,position:chr3:78,485,247-79,099,496) with the full-length mRNA BC017743 (isolatedby Mammalian Gene Collection Program, Imanishi et al., 2004).

A Genscan analysis (http://genes.mit.edu/GENSCAN.html) of this mRNA,predicts a putative exon with ORF of 49 amino acids, but the lack of homologues ororthologues in other species suggests that this transcript unlikely codes for a protein.Acembly (http://www.ncbi.nih.gov/IEB/Research/Acembly/index.html) annotated thegene as a single-exon transcript of 3.1 kb. Interestingly, the alignment of BC017743sequence with the chicken genome produces two hits on Gallus gallus chromosome1, one of which shares 86% identity over 353 residues. Given the more and morerecognised importance of microRNAs (miRNAs) in cancer biology (Xu et al., 2004;McManus, 2003), we looked in the BC017743 sequence for possible miRNA targetsites, which would suggest the existence of a posttranscriptional control mechanismacting on this transcript. The DIANA algorithm (http://diana.pcbi.upenn.edu/cgi-bin/micro_t.cgi) predicted two possible targe t sites in the sequence (Fig. 8A), bothscoring in the high-confidence range. Both miRNAs belong to a group of miRNAgenes that were experimentally identified and their expression analysed by Northernblot. They have a similar sequence originating from different loci: has-miR-17-5pderives from a locus on chromosome 13 and has-miR-106a from a locus onchromosome X (Mourelatos et al., 2002; Dostie et al., 2003; Kasashima et al., 2004;Suh et al., 2004).

Cluster of H51703The cluster assembled around the sequence of H51703 is now represented byseventeen ESTs (H51703, T69773, T70759, T84499, H40323, H40377, R83269,AA668381, AA669442, BI598464, BM993003, BX105987, BE062088, BF746150,BE061843, BF746204, CA440361) isolated from Liver and Spleen, Hypothalamus,Lung carcinoma and Metastatic Lung Chondrosarcoma libraries. Again, it issignificant that this cluster aligns (see the Human Genome Browser, May 2004,http://genome.ucsc.edu/cgi-bin/hgGateway, position: chr3: 78,485,247-79,099,496)with a full-length transcript, namely BC043430 (isolated by The Mammalian GeneCollection Program, Imanishi et al., 2004). Genscan analysis(http://genes.mit.edu/GENSCAN.html) does not predict any exon and the possible,small ORFs do not show similarities with known homologues or orthologues.

Chapter 4

114

Acembly (http://www.ncbi.nih.gov/IEB/Research/Acembly/index.html) annotated thistranscript as a putative single-exon gene of about 1.9 kb, with an ORF of 46 aminoacids, without known similarities. However, two bovine ESTs, AV601957andCR454939, align in correspondence of BC043430, with 92% and 81% identityrespectively (in a region free of repetitive DNA). The DIANA algorithm(http://diana.pcbi.upenn.edu/cgi-bin/micro_t.cgi) predicted four putative miRNA targetsites in the BC043430 sequence (Fig. 8B). Those miRNAs were previously identifiedthrough homology with known miRNAs (Houbaviy et al., 2003; Weber, 2005).

A

B

Figure 8. Human microRNAs that putatively hit target sequences on (A) BC017743 and (B) BC043430transcripts (adapted from a window of the DIANA algorithm results (http://diana.pcbi.upenn.edu/cgi-bin/micro_t.cgi).

Novel non-coding RNAs

115

Conservation analysis of the deleted sequenceAligning the sequence of the GLC20-3p12 deletion with the chicken genome(Bejerano et al, 2004; Hillier et al., 2004, Chicken Genome Browser Gateway,February 2004 assembly), we identified seven blocks of highly conserved sequence.Four blocks contain Alu repeats. Two of those show 93% identity between humanand chicken and find homology with the 7SL chicken gene on chicken chromosome5. Two others blocks, almost identical to the previous ones, hit chicken chromosome2 and perhaps represent another copy of the 7SL gene, located on this otherchromosome. Finally, three highly conserved regions hit chicken chromosome 1(numbers refer to the human sequence - Human Genome Browser, May 2004 hg 17assembly): Seq5: 69 bp (79000588 - 79000654), 84% identity hum/chick Seq6: 353bp (78954642 - 78954984), 86% identity hum/chick Seq7: 243 bp (79044143 -79044377), 81% identity hum/chick. These regions do not contain repetitiveelements. Interestingly seq6 aligns inside the non-coding transcript BC017743.Understanding whether seq5 and seq7 are actually transcribed and thereforepossibly encode functional RNA, would require further experimentation.

Expression studiesTo independently verify whether the two clusters of ESTs are indeed transcribed, weperformed Northern blot experiments using W91914 and H51703 cDNA as probes.In our conditions, the W91914 probe lighted two bands of about 2.4 and 3 kb in alltissues tested (Clontech MTN blot, Palo Alto, CA) except kidney and pancreas,(Fig.9A). In the same conditions, the H51703 probe lighted a band of about 2 kb inpancreas, one of about 3 kb in skeletal muscle and one of about 1.35 kb in liver(Fig.9B). RT-PCR experiments performed with primers designed on W91914 andH51703 also confirmed that these sequences are actually expressed. H51703 wasamplified from all cDNA samples prepared from various organs (Clontech, Palo AltoCA), (Fig.9C). W91914 was amplified from fetal lung cDNA (Clontech, Palo Alto CA),Fig.9D. By means of RT-PCR experiments using primers designed across splicesites, all cDNA samples were tested to be free of genomic DNA (data not shown).

Chapter 4

116

Figure 9. Expression studies of ESTs W91914 andH51703, chosen as representative of two clusters ofESTs located in the region affected by the GCL20-3p12 homozygous deletion.A and B, Northern blots with MTN blots (Clontech,Palo Alto, CA). Probes: W91914 (A) and H51703(B). Arrows indicate bands highlighted by therespective probes. C and D, RT-PCR with cDNAs(Clontech, Palo Alto, CA) from various organs. H:heart, Br: brain, Pl: placenta, Lu: lung, Li: liver, SM:skeletal muscle, K: kidney, Pc: pancreas. PCRprimers for H51703 (223 bp) in C and for W91914(215 bp) in D. 3% NuSieve agarose gel, Cambrex,Ethidium Bromide stained. Ladder: 100 bp marker(Invitrogen).

Novel non-coding RNAs

117

DISCUSSION

Detection of homozygous deletions and microsatellite mapping to identify allele lossand deletions are still the most powerful method to localise putative TSG. Interstitialdeletions in the low gene-density chromosome region 3p12 were reported in lungand other malignancies (Daly et al., 1991; Ganly et al., 1992; Pandis et al., 1993).Rabbitts et al. (1990) reported a homozygous deletion at the locus D3S3 in theU2020 cell line. Chen et al. (1994) reported a homozygous deletion in this sameregion in breast cancer. Sundaresan et al. (1998a) found in lung and breast cancertwo more overlapping homozygous deletions harboured in this region.

The U2020 region has been strongly suspected to harbour a tumoursuppressor gene since the mid 1990s when karyotype analysis had shown thatdeletions in 3p12 are the only evident abnormality in cells cultured from normalbronchial epithelial cells of lung cancer patients (Sundaresan et al., 1995). Sanchezet al. (1994) showed that the introduction of two centromeric fragments of 3p(encompassing 3p12-q24 and 3p14-q11) into a highly malignant renal cell carcinoma(RCC) cell line resulted in a dramatic suppression of tumour growth in athymic nudemice, suggesting that a locus in this region controls the growth of RCC cells byinducing rapid cell death in vivo. Lott et al. (1998) and Lovell et al. (1999) showedthat a fragment of human chromosome 3 overlapping with U2020 deletion mediatesrapid cell death and tumour growth suppression of RCC cells in vivo, whereas adeletion of this region is associated with immortalisation of human uroepithelial cells(Vieten et al., 1998).

Detailed mapping showed that the U2020 deletion is about 8 Mb in size(Drabkin et al., 1992; Latif et al., 1992) and harbours the smaller deletions HCC38and H219x (Sundaresan et al., 1998a). In the U2020 region, Sundaresan et al.(1998b) identified and cloned a gene that is disrupted by these deletions: DUTT1(also known as ROBO1, independently isolated by Kidd et al., 1998).DUTT1/ROBO1 is an integral membrane protein. It is an axon guidance/celladhesion receptor whose best-characterised function deals with regulating thedecision by axons to cross the central nervous system midline. DUTT1 is widelyexpressed as an about 8 Kb mRNA and two transcript variants are known that differfor their 5’ terminus (start site and first exon). The H219x deletion hits exon 2(encoding the first Ig domain) of DUTT1-short variant. Introduction of this mutation inthe mouse germ line (Xian et al., 2001) generates animals that in the homozygous

Chapter 4

118

state frequently die at birth of respiratory failure due to functional immaturity of lungs.Survivors acquire bronchial epithelial abnormalities similar to those involved in earlystages of lung cancer and die in the first year of life (Xian et al., 2001). Heterozygousmice grow normally but develop spontaneously lymphomas and carcinomas in theirsecond year of life with a 3-fold increase in incidence compared with controls (Xianet al., 2004). Invasive lung adenocarcinoma is by far the predominant carcinoma. Inaddition to the mutant allele, loss of heterozygosity analysis indicates that thesetumours retain the structurally normal allele but with substantial methylation of thegene's promoter (Xian et al., 2004).

Dallol et al., (2002) found that DUTT1 promoter is hypermethylated in 19% ofprimary invasive breast carcinomas, 18% of primary clear cell renal cell carcinomasand in 4% of primary NSCLC tumours. In addition, 80% of breast and 75% of clearcell renal cell carcinomas (CC-RCC) showing DUTT1 hypermethylation also showedallelic loss at 3p12 suggesting that DUTT1/ROBO1 is a classic tumour suppressorgene requiring inactivation of both alleles to elicit tumourigenesis. However, anextensive mutation analysis of DUTT1 in lung, breast and kidney cancer did notretrieve inactivating mutations (Dallol et al., 2002).

Here we report the identification in the lung cancer cell line GLC20 of a newhomozygous deletion at 3p12.3 that spans about 110 to 130 kb (perhaps thesmallest described so far in this region) and hits the second exon of DUTT1.Moreover, using molecular biology and bioinformatics methods, we identified in thisregion two novel putative genes that reside in the second intron of DUTT1 andtherefore are also hit by the deletion. Both transcripts do not show obvious splicingsignals however in our hands each one decorated slightly different bands in Northernblots. Both show a polyA tail and very small ORFs that comparative genomeanalyses suggest are unlikely to encode proteins. On the bases of thesecharacteristics they are possibly mRNA-like non-coding RNAs (Tupy et al., 2005;Erdmann et al., 2000). Transcribed by RNA polymerase II, in absence of proteinproducts, these types of RNAs serve as riboregulators or regulators of expression ofrelated genes (Typy et al., 2005; Numata et al., 2003; Erdmann et al., 2002). Basedon their sequence, BC017743 and BC043430 seemingly lack homologues ororthologues but this should not be a discrediting criterion as it was reported for otherknown genes and ncRNAs (Conrad et al., 2002; Weber et al., 2005). Also, somebiologically important ncRNA families show that, inside each group, conservation ofsecondary structure has a higher significance compared to conservation of primary

Novel non-coding RNAs

119

sequence (Weinberg and Ruzzo, 2004).Up to now, very few cases of independent transcription units embedded

inside gene introns were described. The genes EV12A and EV12B are encoded byone intron of the human neurofibromatosis type 1 (NF1) gene and are transcribed inthe opposite direction. However their products are functionally unrelated to NF1. TheAch transporter gene (transcribed from the first intron of the rat ChAT gene, Bejaninet al., 1994) and the Saitohin gene (transcribed from intron 9 of the human tau gene,Conrad et al., 2002) are different as both their respective products are transcribed inthe same orientation and also are functionally related to the longer gene inside whichthey are harboured. The casistics is bound to increase as it was recently reported(Reis et al., 2005) that 233,303 clusters of ESTs are totally contained within intronicregions. Both BC017743 and BC043430 are transcribed in the same direction ofDUTT1 suggesting that their product could be co-regulated and possibly related toDUTT1 function. The fact that both are transcribed in the same orientation as DUTT1suggests also that the hypermethylation that affects DUTT1 promoter in somecancers (Dallol et al., 2002) might concomitantly deregulate their own transcription. Itwill be interesting to verify whether both transcripts are actually under translationalcontrol by the miRNAs whose putative target sites were predicted in their sequence.Since the loss of BC017743 and BC043430 was found associated with a tumourphenotype, one might argue that both transcripts exert a tumour suppressor action,possibly mediated by the miRNA action. In fact, several cases are now known (Calinet al., 2005; Xu et al., 2004; McManus, 2003) of changes in the expression level ofmiRNAs that may affect the control of cell growth or survival and therefore areinvolved with cancer onset or progression. Noteworthy, hsa-mir-17-5p originatesfrom 13q31.3 (http://www.ensembl.org/Homo_sapiens/contigview?highlight=&chr=13&vc_start=90700860&vc_end=90900943), a region whose LOH has beeninvolved with breast cancer progression (Eiriksdottir et al., 1998). Therefore onemight speculate that LOH at the miRNA locus (e.g. hsa-mir-17-5p) might obtain thesame effect of LOH at the miRNA target site (conceivably BC017743 and BC043430in the case of lung, breast or kidney cancer). Moreover, hsa-mir-17-5p was shown tobe down-regulated following differentiation induced by TPA treatment of HL-60promyelocytic leukemia cells (Kasashima et al., 2004). This suggests that miRNA-induced control of cell differentiation might consist of two distinct mechanisms: I) amechanism of gene silencing through up-regulation of miRNAs, II) a mechanism ofgene activation through termination of miRNA-regulated gene silencing.

Chapter 4

120

Similarly, it would be interesting to investigate whether LOH at 11q23 in lung cancerinvolves not only loss of the TSLC1 gene that resides there (Kuramochi et al., 2001)but also loss of the locus encoding hsa-mir-34b, the one that putatively hitsBC043430 (as well as other targets in the genome probably: it was shown, forexample, that reduced levels of hsa-mir-15a and hsa-mir-16 are a trait shared bydifferent forms of lymphomas and leukemias; Calin et al., 2002; Eis et al., 2005). Asituation possibly symmetrical to what described in the GLC20-3p12 deletion, isgiven by the accumulation of the ncRNA BIC (Eis et al., 2005). BIC transcript ispolyadenylated, has short putative ORFs that are not conserved. Most likely BICdoes not encode a protein. However, a phylogenetically conserved region of BICwas indeed shown to encode a miRNA, mir-155, whose accumulation is stronglycorrelated with an aggressive B cell neoplasm (Eis et al., 2005). The GLC20-3p12deletion might also harbour miRNA loci. Recently, a list of computationally identifiedhuman miRNA genes was reported, some of whom are encoded at 3p12 (Berezikovet al., 2005). One of them (cand893 HS3, 78.768.573- 78.768.661 R, whose closestexperimentally identified miRNA is M. musculus mmu-mir 297, Houbaviy et al., 2003)is harboured inside the deletion. Further studies are necessary to investigatewhether DUTT1 itself is a target for this miRNA, as it was suggested that intronicmiRNAs might exert a role in regulating fast cell transitions in response to externalstimuli, perhaps modifying the expression of genes that harbour them, bypassingprotein synthesis (Ying and Lin, 2004).

Thanks to the recently published draft of the chicken genome, we were ableto analyse the degree of conservation of the non-coding sequence around the exon2 of DUTT1. This region contains at least three blocks of extremely conserved, non-repetitive sequence, spanning respectively 353, 69 and 243 bp, with a degree ofidentity from 81 to 86% between the human and chicken genome, whose functionalmeaning was not investigated in this work.

In conclusion, it is interesting to note that the two smallest, partly overlappinghomozygous deletions described at 3p12.3 (that is H219x, Sundaresan et al., 1998a,and the one described here) remove, besides one exon of DUTT1, two ncRNAgenes. BC043430 and BC017743 are the first ncRNA transcripts/genes found in ahomozygous cancer deletion affecting the 3p12.3 region. The presence of two highlyconserved intronic ncRNAs of the ROBO1/DUTT1 gene, along with several shortnon-coding regions highly conserved between human and chicken genomes

Novel non-coding RNAs

121

(Bejerano et al., 2004; Hillier et al., 2004), suggests that the effect of the 3p12deletion might be complex. Moreover, it must be mentioned that another element ofcomplexity in determining the full tumour phenotype is introduced by the presence inGLC20 cells of another homozygous deletion, at 3p21.3 (De Leij et al., 1985; Kok etal., 1997). In fact, with respect to the short arm of chromosome 3, it was suggestedfor several types of tumour that multiple deletions, coexisting but with different 3plocations, most likely have a synergistic effect in driving tumourigenesis (Van denBerg et al., 1997; Senchenko et al., 2003).

In further studies the two ncRNA genes, BC043430 and BC017743, need tobe evaluated as novel 3p12 lung, breast, and kidney TSG through mutation andfunctional studies. Similarly, the highly conserved, short sequence elements will beevaluated as potential ncRNA loci.

ACKNOWLEDGEMENTS

DAA, MHW and MIL were funded in toto with funds from the National CancerInstitute, National Institutes of Health. The content of the publication does notnecessarily reflect the views or policies of the Department of Health and HumanServices, nor does mention of trade names, commercial products, or organisationsimply endorsement by the U.S. Government. The study was also supported bygrants to CHCMB from the Dutch Cancer Society (Grants RUG 94-834 and RUG2000-2317) and from the J.K. de Cock Stichting (Grant 99-07). We are grateful toAndrei Kouranov (Department of Genetics, School of Medicine, University ofPennsylvania, Philadelphia, USA) for his kind help with interpreting the DIANAalgorithm results. Special thanks to each and every LIB members.

Chapter 4

122

REFERENCES

Bejanin S, Cervini R, Mallet J, Berrard S. 1994. A unique gene organization for two cholinergic markers,choline acetyltransferase and a putative vesicular transporter of acetylcholine. J Biol Chem.269:21944-7.

Bejerano G, Pheasant M, Makunin I, Stephen S, Kent WJ, Mattick JS, Haussler D. 2004. Ultraconservedelements in the human genome. Science 304:1321-5.

Berezikov E, Guryev V, van de Belt J, Wienholds E, Plasterk RH, Cuppen E. 2005. Phylogeneticshadowing and computational identification of human microRNA genes. Cell 120:21-4.

Calin GA, Dumitru CD, Shimizu M, Bichi R, Zupo S, Noch E, Aldler H, Rattan S, Keating M, Rai K,Rassenti L, Kipps T, Negrini M, Bullrich F, Croce CM. 2002. Frequent deletions and down-regulation of micro- RNA genes miR15 and miR16 at 13q14 in chronic lymphocytic leukemia. ProcNatl Acad Sci U S A. 99:15524-9.

Chen LC, Matsumura K, Deng G, Kurisu W, Ljung BM, Lerman MI, Waldman FM, Smith HS. 1994.Deletion of two separate regions on chromosome 3p in breast cancers. Cancer Res. 54:3021-4.

Conrad C, Vianna C, Freeman M, Davies P. 2002. A polymorphic gene nested within an intron of the taugene: implications for Alzheimer's disease. Proc Natl Acad Sci U S A. 99:7751-6.

Dallol A, Forgacs E, Martinez A, Sekido Y, Walker R, Kishida T, Rabbitts P, Maher ER, Minna JD, Latif F.2002. Tumour specific promoter region methylation of the human homologue of the DrosophilaRoundabout gene DUTT1 (ROBO1) in human cancers. Oncogene. 21:3020-8.

Daly MC, Douglas JB, Bleehen NM, Hastleton P, Twentyman PR, Sundaresan V, Carritt B, Bergh J,Rabbitts PH. 1991. An unusually proximal deletion on the short arm of chromosome 3 in a patientwith small cell lung cancer. Genomics. 9:113-9.

de Leij LFMH, Postmus PE, Buys CHCM, Elema JD, Ramaekers F, Poppema S, Brouwer M, van derVeen AY, Mesander G, The TH. 1985. Characterization of three new variant type cell lines derivedfrom small cell carcinoma of the lung. Cancer Res. 45:6024-33.

Dostie J, Mourelatos Z, Yang M, Sharma A, Dreyfuss G. 2003. Numerous microRNPs in neuronal cellscontaining novel microRNAs. RNA. 9:180-6.

Drabkin HA, Mendez MJ, Rabbitts PH, Varkony T, Bergh J, Schlessinger J, Erickson P, Gemmill RM.1992. Characterization of the submicroscopic deletion in the small-cell lung carcinoma (SCLC) cellline U2020. Genes Chromosomes Cancer. 5:67-74.

Driesen MS, Dauwerse JG, Wapenaar MG, Wapenaar MG, Meerhoek EJ, Mollevanger P, Chen KL,Fishbeck KH, van Ommen GJ. 1991. Generation and fluorescent in situ hybridisation mapping ofyeast artificial chromosomes of 1p, 17p, and 19q from a hybrid cell line by high density screeningof an amplified library. Genomics 11: 1079-1087.

Eiriksdottir G, Johannesdottir G, Ingvarsson S, Bjornsdottir IB, Jonasson JG, Agnarsson BA, HallgrimssonJ, Gudmundsson J, Egilsson V, Sigurdsson H, Barkardottir RB. 1998. Mapping loss ofheterozygosity at chromosome 13q: loss at 13q12-q13 is associated with breast tumourprogression and poor prognosis. Eur J Cancer. 34:2076-81. Eis PS, Tam W, Sun L, Chadburn A,Li Z, Gomez MF, Lund E, Dahlberg JE. 2005. Accumulation of miR-155 and BIC RNA in human Bcell lymphomas. Proc Natl Acad Sci U S A. 102:3627-32.

Erdmann VA, Szymanski M, Hochberg A, Groot N, Barciszewski J. 2002. Non-coding, mRNA-like RNAsdatabase Y2K. Nucleic Acids Res. 28:197-200.

Ganly PS, Jarad N, Rudd RM, Rabbitts PH. 1992. PCR-based RFLP analysis allows genotyping of theshort arm of chromosome 3 in small biopsies from patients with lung cancer. Genomics. 12:221-8.

Novel non-coding RNAs

123

Giles RH, Petrij F, Dauwerse HHG, Den Hollander AI, Lushnikova T, van Ommen GJ, Goodman RH,Deaven LL, Doggett NA, Peters DJ, Breuning MH. 1997. Construction of a 1.2-Mbcontigsurrounding, and molecular analysis of, the human CREB-binding protein (CBP/CREBBP)gene on chromosome 16p13.3. Genomics 42: 96-114.

Hillier LW, Miller W, Birney E, Warren W, Hardison RC, Ponting CP, Bork P, Burt DW, Groenen MA,Delany ME, Dodgson JB, Chinwalla AT, Cliften PF, Clifton SW, Delehaunty KD, Fronick C, FultonRS, Graves TA, Kremitzki C, Layman D, Magrini V, McPherson JD, Miner TL, Minx P, Nash WE,Nhan MN, Nelson JO, Oddy LG, Pohl CS, Randall-Maher J, Smith SM, Wallis JW, Yang SP,Romanov MN, Rondelli CM, Paton B, Smith J, Morrice D, Daniels L, Tempest HG, Robertson L,Masabanda JS, Griffin DK, Vignal A, Fillon V, Jacobbson L, Kerje S, Andersson L, CrooijmansRP, Aerts J, van der Poel JJ, Ellegren H, Caldwell RB, Hubbard SJ, Grafham DV, Kierzek AM,McLaren SR, Overton IM, Arakawa H, Beattie KJ, Bezzubov Y, Boardman PE, Bonfield JK,Croning MD, Davies RM, Francis MD, Humphray SJ, Scott CE, Taylor RG, Tickle C, Brown WR,Rogers J, Buerstedde JM, Wilson SA, Stubbs L, Ovcharenko I, Gordon L, Lucas S, Miller MM,Inoko H, Shiina T, Kaufman J, Salomonsen J, Skjoedt K, Wong GK, WangJ, Liu B, Wang J, Yu J,Yang H, Nefedov M, Koriabine M, Dejong PJ, Goodstadt L, Webber C, Dickens NJ, Letunic I,Suyama M, Torrents D, von Mering C, Zdobnov EM, Makova K, Nekrutenko A, Elnitski L, EswaraP, King DC, Yang S, Tyekucheva S, Radakrishnan A, Harris RS, Chiaromonte F, Taylor J, He J,Rijnkels M, Griffiths-Jones S, Ureta-Vidal A, Hoffman MM, Severin J, Searle SM, Law AS, SpeedD, Waddington D, Cheng Z, Tuzun E, Eichler E, Bao Z, Flicek P, Shteynberg DD, Brent MR, ByeJM, Huckle EJ, Chatterji S, Dewey C, Pachter L, Kouranov A, Mourelatos Z, Hatzigeorgiou AG,Paterson AH, Ivarie R, Brandstrom M, Axelsson E, Backstrom N, Berlin S, Webster MT, PourquieO, Reymond A, Ucla C, Antonarakis SE, Long M, Emerson JJ, Betran E, Dupanloup I,Kaessmann H, Hinrichs AS, Bejerano G, Furey TS, Harte RA, Raney B, Siepel A, Kent WJ,Haussler D, Eyras E, Castelo R, Abril JF, Castellano S, Camara F, Parra G, Guigo R, Bourque G,Tesler G, Pevzner PA, Smit A, Fulton LA, Mardis ER, Wilson RK. (The International ChickenGenome Sequencing Consortium). 2004. Sequence and comparative analysis of the chickengenome provide unique perspectives on vertebrate evolution. Nature. 432:695-716.

Houbaviy HB, Murray MF, Sharp PA. 2003. Embryonic stem cell-specific MicroRNAs. Dev Cell. 5:351-8.

Huebner K, Garrison PN, Barnes LD, Croce CM. 1998. The role of the FHIT/FRA3B locus in cancer. AnnuRev Genet.;32:7-31.

Imanishi T, Itoh T, Suzuki Y, O'Donovan C, Fukuchi S, Koyanagi KO, Barrero RA, Tamura T, Yamaguchi-Kabata Y, Tanino M, Yura K, Miyazaki S, Ikeo K, Homma K, Kasprzyk A, Nishikawa T, HirakawaM, Thierry-Mieg J, Thierry-Mieg D, Ashurst J, Jia L, Nakao M, Thomas MA, Mulder N,Karavidopoulou Y, Jin L, Kim S, Yasuda T, Lenhard B, Eveno E, Suzuki Y, Yamasaki C, TakedaJ, Gough C, Hilton P, Fujii Y, Sakai H, Tanaka S, Amid C, Bellgard M, Bonaldo Mde F, Bono H,Bromberg SK, Brookes AJ, Bruford E, Carninci P, Chelala C, Couillault C, de Souza SJ, DebilyMA, Devignes MD, Dubchak I, Endo T, Estreicher A, Eyras E, Fukami-Kobayashi K, Gopinath GR,Graudens E, Hahn Y, Han M, Han ZG, Hanada K, Hanaoka H, Harada E, Hashimoto K, Hinz U,Hirai M, Hishiki T, Hopkinson I, Imbeaud S, Inoko H, Kanapin A, Kaneko Y, Kasukawa T, Kelso J,Kersey P, Kikuno R, Kimura K, Korn B, Kuryshev V, Makalowska I, Makino T, Mano S, Mariage-Samson R, Mashima J, Matsuda H, Mewes HW, Minoshima S, Nagai K, Nagasaki H, Nagata N,Nigam R, Ogasawara O, Ohara O, Ohtsubo M, Okada N, Okido T, Oota S, Ota M, Ota T, OtsukiT, Piatier-Tonneau D, Poustka A, Ren SX, Saitou N, Sakai K, Sakamoto S, Sakate R, Schupp I,Servant F, Sherry S, Shiba R, Shimizu N, Shimoyama M, Simpson AJ, Soares B, Steward C,Suwa M, Suzuki M, Takahashi A, Tamiya G, Tanaka H, Taylor T, Terwilliger JD, Unneberg P,Veeramachaneni V, Watanabe S, Wilming L, Yasuda N, Yoo HS, Stodolsky M, Makalowski W, GoM, Nakai K, Takagi T, Kanehisa M, Sakaki Y, Quackenbush J, Okazaki Y, Hayashizaki Y, Hide W,Chakraborty R, Nishikawa K, Sugawara H, Tateno Y, Chen Z, Oishi M, Tonellato P, Apweiler R,Okubo K, Wagner L, Wiemann S, Strausberg RL, Isogai T, Auffray C, Nomura N, Gojobori T,Sugano S. 2004. Integrative annotation of 21,037 human genes validated by full-length cDNAclones. PLoS Biol. 2:e162.

Kasashima K, Nakamura Y, Kozu T. 2004. Altered expression profiles of microRNAs during TPA-induceddifferentiation of HL-60 cells. Biochem Biophys Res Commun. 322:403-10.

Chapter 4

124

Kidd T, Brose K, Mitchell KJ, Fetter RD, Tessier-Lavigne M, Goodman CS, Tear G. 1998. Roundaboutcontrols axon crossing of the CNS midline and defines a novel subfamily of evolutionarilyconserved guidance receptors. Cell. 92:205-15.

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

Kok K, Osinga J, Carritt B, Davis MB, van der Hout AH, van der Veen AY, Landsvater RM, de Leij LF,Berendsen HH, Postmus PE, et al. 1987. Deletion of a DNA sequence at the chromosomal region3p21 in all major types of lung cancer. Nature. 330:578-81.

Kok K, van den Berg A, Veldhuis PMJF, van der Veen AY, Franke M, Schoenmakers EFPM, HulsbeekMMF, van der Hout AH, de Leij L, van der Ven W, Buys CHCM. 1994. A homozygous deletion in asmall cell lung cancer cell line involving a 3p21 region with a marked instability in yeast artificialchromosomes. Cancer Res. 54:4183-4187.

Kok K, Van den Berg A, Veldhuis PMJF, Franke M, Terpstra P, Buys CHCM. 1995. The genomic structureof the human UBE1L gene. Gene Expression. 4:163-175.

Kok K, Naylor SL, Buys CHCM. 1997. Deletions of the short arm of chromosome 3 in solid tumors and thesearch for suppressor genes. Adv Cancer Res. 71:27-92.

Kuramochi M, Fukuhara H, Nobukuni T, Kanbe T, Maruyama T, Ghosh HP, Pletcher M, Isomura M,Onizuka M, Kitamura T, Sekiya T, Reeves RH, Murakami Y. 2001. TSLC1 is a tumor-suppressorgene in human non-small-cell lung cancer.

Latif F, Tory K, Modi WS, Graziano SL, Gamble G, Douglas J, Heppell-Parton AC,Rabbitts PH, Zbar B,Lerman MI. 1992. Molecular characterization of a large homozygous deletion in the small cell lungcancer cell line U2020: a strategy for cloning the putative tumor suppressor gene. GenesChromosomes Cancer. 5:119-27.

Lerman MI, Minna JD. 2000. The 630-kb lung cancer homozygous deletion region on humanchromosome 3p21.3: identification and evaluation of the resident candidate tumor suppressorgenes. The International Lung Cancer Chromosome 3p21.3 Tumor SuppressorGene Consortium.Cancer Res. 60:6116-33.

Lott ST, Lovell M, Naylor SL, Killary AM. 1998. Physical and functional mapping of a tumor suppressorlocus for renal cell carcinoma within chromosome 3p12. Cancer Res. 58:3533-7.

Lovell M, Lott ST, Wong P, El-Naggar A, Tucker S, Killary AM. 1999. The genetic locus NRC-1 withinchromosome 3p12 mediates tumor suppression in renal cell carcinoma independently ofhistological type, tumor microenvironment, and VHL mutation. Cancer Res. 59:2182-9.

Marshall E. 1991. Breast cancer: stalemate in the war on cancer. Science. 254:1719-20.

McManus MT. 2003. MicroRNAs and cancer. Semin Cancer Biol. 13:253-8.

Mourelatos Z, Dostie J, Paushkin S, Sharma A, Charroux B, Abel L, Rappsilber J, Mann M, Dreyfuss G.2002. miRNPs: a novel class of ribonucleoproteins containing numerous icroRNAs. Genes Dev.16:720-8.

Numata K, Kanai A, Saito R, Kondo S, Adachi J, Wilming LG, Hume DA, Hayashizaki Y, Tomita M;RIKEN GER Group; GSL Members. 2003. Identification of putative noncoding RNAs among theRIKEN mouse full-length cDNA collection. Genome Res. 13:1301-6.

Pandis N, Jin Y, Limon J, Bardi G, Idvall I, Mandahl N, Mitelman F, Heim S. 1993. Interstitial deletion ofthe short arm of chromosome 3 as a primary chromosome abnormality in carcinomas of thebreast. Genes Chromosomes Cancer. 6:151-5.

Rabbitts P, Bergh J, Douglas J, Collins F, Waters J. 1990. A submicroscopic homozygous deletion at theD3S3 locus in a cell line isolated from a small cell lung carcinoma. Genes Chromosomes Cancer.2:231-8.

Novel non-coding RNAs

125

Reis EM, Louro R, Nakaya HI, Verjovski-Almeida S. 2005. As antisense RNA gets intronic. OMICS. 9:2-12.

Sambrook J, Fritsch EF, Maniatis T 1998. Molecular Cloning, A Laboratory Manual. Cold Spring HarborLaboratory Press.

Sanchez Y, el-Naggar A, Pathak S, Killary AM. 1994. A tumor suppressor locus within 3p14-p12 mediatesrapid cell death of renal cell carcinoma in vivo. Proc Natl Acad Sci U S A. 91:3383-7.

Senchenko V, Liu J, Braga E, Mazurenko N, Loginov W, Seryogin Y, Bazov I, Protopopov A, Kisseljov FL,Kashuba V, Lerman MI, Klein G, Zabarovsky ER. 2003. Deletion mapping using quantitative real-time PCR identifies two distinct 3p21.3 regions affected in most cervical carcinomas. Oncogene.22:2984-92.

Suh MR, Lee Y, Kim JY, Kim SK, Moon SH, Lee JY, Cha KY, Chung HM, Yoon HS, Moon SY, Kim VN,Kim KS. 2004. Human embryonic stem cells express a unique set of microRNAs.

Dev Biol. 270:488-98.

Sundaresan V, Heppell-Parton A, Coleman N, Miozzo M, Sozzi G, Ball R, Cary N, Hasleton P, Fowler W,Rabbitts P. 1995. Somatic genetic changes in lung cancer and precancerous lesions. Ann Oncol.6 Suppl 1:27-31; discussion 31-2.

Sundaresan V, Chung G, Heppell-Parton A, Xiong J, Grundy C, Roberts I, James L, Cahn A, Bench A,Douglas J, Minna J, Sekido Y, Lerman M, Latif F, Bergh J, Li H, Lowe N, Ogilvie D, Rabbitts P.1998a. Homozygous deletions at 3p12 in breast and lung cancer. Oncogene. 17:1723-9.

Sundaresan V, Roberts I, Bateman A, Bankier A, Sheppard M, Hobbs C, Xiong J, Minna J, Latif F,Lerman M, Rabbitts P. 1998b. The DUTT1 gene, a novel NCAM family member is expressed indeveloping murine neural tissues and has an unusually broad pattern of expression. Mol CellNeurosci. 11:29-35.

Todd S, Franklin WA, Varella-Garcia M, Kennedy T, Hilliker CE Jr, Hahner L, Anderson M, Wiest JS,Drabkin HA, Gemmill RM. 1997. Homozygous deletions of human chromosome 3p in lung tumors.Cancer Res. 57:1344-52.

Tupy JL, Bailey AM, Dailey G, Evans-Holm M, Siebel CW, Misra S, Celniker SE, Rubin GM. 2005.Identification of putative noncoding polyadenylated transcripts in Drosophila melanogaster.ProcNatl Acad Sci U S A. 102:5495-500.

Van den Berg A, Dijkhuizen T, Draaijers TG, Hulsbeek MMF, Maher ER, Van den Berg E, Störkel S, BuysCHCM. 1997. Analysis of multiple renal cell adenomas and carcinomas suggests allelic loss at3p21 to be a prerequisite for malignant development. Genes Chromosomes Cancer. 19: 228-232.

Vieten L, Belair CD, Savelieva L, Julicher K, Brocker F, Bardenheuer W, Schutte J, Opalka B, ReznikoffCA. 1998. Minimal deletion of 3p13-->14.2 associated with immortalization of human uroepithelialcells. Genes Chromosomes Cancer. 21:39-48.

Weber MJ. 2005. New human and mouse microRNA genes found by homology search. FEBS J. 272:59-73.

Weinberg Z, Ruzzo WL. 2004. Exploiting conserved structure for faster annotation of non-coding RNAswithout loss of accuracy. Bioinformatics. 20 Suppl 1:I334-I341.

Whang-Peng J, Kao-Shan CS, Lee EC, Bunn PA, Carney DN, Gazdar AF, Minna JD. 1982. Specificchromosome defect associated with human small-cell lung cancer; deletion 3p(14-23). Science.215:181-2.

Xian J, Clark KJ, Fordham R, Pannell R, Rabbitts TH, Rabbitts PH. 2001. Inadequate lung developmentand bronchial hyperplasia in mice with a targeted deletion in the DUTT1/Robo1 gene. Proc NatlAcad Sci U S A. 98:15062-6.

Xian J, Aitchison A, Bobrow L, Corbett G, Pannell R, Rabbitts T, Rabbitts P. 2004. Targeted disruption of

Chapter 4

126

the 3p12 gene, DUTT1/Robo1, predisposes mice to lung adenocarcinomas and lymphomas withmethylation of the gene promoter. Cancer Res. 64:6432-7.

Xu P, Guo M, Hay BA. 2004. MicroRNAs and the regulation of cell death. Trends Genet. 20:617-24.

Ying SY, Lin SL. 2005. Intronic microRNAs. Biochem Biophys Res Commun. 326:515-20.

Zabarovsky ER, Lerman MI, Minna JD. 2002. Tumor suppressor genes on chromosome 3p involved in thepathogenesis of lung and other cancers. Oncogene. 21:6915-35.

Zbar B. 1989. Chromosomal deletions in lung cancer and renal cancer. Important Adv Oncol. 41-60.

Chapter 5

Summary and general discussion

SUMMARY

Deletions of the short arm of chromosome 3 are a most common abnormality in lungcancer. They have been reported to occur in approximately 75% of non small celllung cancer (NSCLC) tumours and in up to 100% of small cell lung cancer (SCLC)tumours. For lung cancer, three non-overlapping deletion regions have beenidentified at 3p25, 3p21.3 and 3p12-p14. In recent years, the number of distinctregions on the short arm of chromosome 3 that have been implicated in thedevelopment of tumours, has been expanded to seven. These include the 3p22AP20 region, the 3p21.3 CER1 and CER2 regions, the 3p21 D3F15S2 region, the3p21.3 LUCA region, the 3p14 FHIT region and the 3p12 ROBO1 region.Chapter 1 of this thesis gives a review of the literature on these deletions. In smallcell lung cancer (SCLC) the AP20 region has been reported to show homozygousdeletions in the large majority of cases. The region contains four genes. One ofthese genes, CTDSPL, showed inhibition of tumour growth upon subcutaneousinjection into nude mice. The CER1 and CER2 regions have been discovered by anapproach called the elimination test. They represent regions commonly eliminatedfrom a human chromosome 3 introduced into a cancer cell of murine or human originand are supposed to contain tumour suppressor genes. A number of genes havebeen identified in these regions, but none of them have been implicated in lungcancer. The region most consistently reduced to hemizygosity in SCLC is theD3F15S2 region. A gene, UBE1L, coding for ubiquitin-activating enzyme E1-like wasisolated from this region. The mRNA concentration of UBE1L in SCLC cell lines wasfound to be 0.5%-3% of that in normal lung tissue, making it a possible tumoursuppressor candidate for lung cancer. The region has, however, been furtherconfined by the discovery of overlapping homozygous deletions in three differentSCLC cell lines, NCI-H740, GLC20, NCI-H1450, the smallest overlap of thedeletions being 370 kb. In this region 19 genes were discovered. Although several ofthese genes, RBM5, SEMA3F, SEMA3B, HYAL1, TUSC2, RASSF1, ZMYND10,TUSC4, CYB561D2, PL6 and CACNA2D2, have been implicated in lung cancer,convincing evidence for one of these genes to be involved in the development oflung cancer is lacking. Homozygous deletions of 3p14.2 have also frequently beendetected in lung cancer cell lines. Although abnormal transcripts of FHIT were foundin normal tissue and normal transcripts in tumours, FHIT appeared to have tumoursuppressing abilities upon injection into nude mice. A sub-microscopic deletion was

Chapter 5

130

found in the SCLC cell line U2020. One gene was isolated from this region, ROBO1.Mutations and promoter hypermethylation of ROBO1 were found to be rare in lungcancer primary tumours.The 3p21 LUCA region (3p21.3 critical region) has been completely overlapped withP1-derived artificial chromosomes (PACs). In Chapter 2 cells from the SCLC cell lineGLC45 were stably transfected with these PACs and resulting clones with low copynumber integration at a single genomic site were used in tumourigenicity tests ofnude mice to investigate whether the integrated genes suppressed the tumour-inducing capacity of the original SCLC cell line. Two transfectants, each containingPAC 185A4 from the centromeric part of the critical region, caused tumourssignificantly smaller than those caused by transfectants containing other PACs andthose caused by the parental SCLC cell line. Although we could demonstrate theoccurrence in transfectants of PAC-specific gene expression and might, therefore,attribute the reduced tumour growth to the introduced PAC gene content, analternative explanation might apply. This can also explain that one of thetransfectants with only vector sequences integrated, also caused smaller tumours. Ina large proportion of transfectants, namely, transfection and integration ofexogenous DNA appeared to induce a genome-wide instability, which can involvechromosomal rearrangements interfering with cell growth and proliferation. Since thecellular phenotype resulting from chromosomal instability can be diverse, it is hard todiscriminate between expression of newly introduced genes and chromosomalinstability as the underlying mechanism causing a reduced tumour growth.In Chapter 3 we have evaluated the transcriptional consequences of the introductionof PAC 185A4 into the SCLC cell line. For this we used micro-array expressionanalysis to compare the gene expression levels of the transfectants with the geneexpression levels of the parental cell line. A complication, however, is that injectionof one of two transfectants with vector only sequences also led to a significantlydiminished tumour growth in nude mice and that array-CGH revealed that alltransfected clones had increased chromosomal instability compared to the parentalcell line. When we compared the gene expression levels of transfectant-1 containingPAC 185A4 with those of the parental cell line and did the same for transfectant-2containing PAC 185A4 and the parental cell line, we found that only a smallproportion of significantly differentially expressed genes were present in bothcomparisons. By comparing the gene expression data with the previously obtainedarray-CGH data, we were able to see a strong concordance of regions with DNA

Summary and General Discussion

131

copy number changes –that varied between the transfectants- and regions withtranscriptional changes. This explains why only a small overlap occurred betweendifferentially expressed genes of both transfectants with the PAC. In addition, wefound that the 3p21.3 genes CACNA2D2 and TUSC4 were higher expressed intransfectants with PAC 185A4 than in the parental cell line. These two genes are notlocated on PAC 185A4, but are normally the 3p21.3 neighbours of CYB561D2 andPL6, that are located on PAC 185A4. This suggests the occurrence of trans-regulation between genes and regulatory DNA sequences in the 3p21.3 region.In Chapter 4 we found a homozygous deletion in the 3p12 region in the SCLC cellline GLC20 in addition to the 3p21.3 homozygous deletion of this cell line. By meansof fiber-fluorescent in situ hybridisation experiments using P1-clones from the region,we found the length of the deletion to be approximately 110 kb-130 kb. Thehomozygously deleted region affects exon 2 of ROBO1, which causes the loss ofamino acids 19-128 in the protein product. We discovered two novel transcripts inthe second intron of ROBO1. Based on their characteristics both transcripts do notseem to encode proteins, but are possibly non-coding RNAs. We also found apossible miRNA target site in the sequences of these RNAs. In addition, we found acomputationally identified miRNA for which we have, however, not yet discoveredpossible target genes. Further studies will be needed to investigate the role of thesenon-coding RNAs.

Chapter 5

132

GENERAL DISCUSSION

Deletions of the short arm of chromosome 3 are a most common abnormality in lungcancer. Several non-overlapping regions of loss have been described in lung cancer.This has also been found for renal cell cancer, where three regions, 3p25, 3p21 and3p12-14, on the short arm of chromosome 3 may be involved. When the short arm ofchromosome 3 is deleted in renal cell carcinomas, the 3p21 region is alwaysinvolved. Sometimes these losses occur together with allelic losses of the 3p25region, and sometimes together with losses of the 3p12-p14 region. This indicatesthat loss of either 3p25 or 3p12-p14 might not be sufficient and that loss of 3p21 isneeded in order to develop carcinomas (van den Berg, thesis Groningen 1996). Theregion most commonly showing deletions in lung cancer is the 3p21.3 critical region.Since also in lung cancer multiple deletion regions have been identified acombination of genes from several of these regions might be responsible for thedevelopment of lung cancer. In addition, multiple genes in a single deletion regionmay be involved. Common regulators could be located within a critical region,thereby regulating several genes simultaneously. Genes located outside deletedregions might also be regulated by sequences located in the 3p21.3 critical region orother deleted regions. The existence of so many possibilities may explain why,despite all the functional assays conducted for genes from the 3p21.3 critical region,no convincing evidence has been presented for a single tumour suppressor gene asbeing responsible for lung cancer.Contradicting results have been obtained in functional assays for several genes.Experimental differences might be held accountable for that since the growth rate orproliferation of tumour cells might be very different in artificial environments. It couldbe that in in vitro experiments such as colony formation in soft agar or growth ratedetermination, the results obtained differ from those for the same gene in in vivoexperiments. In our experiments we saw no correlation between the growth of thecells in culture and the tumour growth rate. All our results are, therefore, based on invivo experiments.The genomic context of a parental cell line used for transfections might also have aninfluence on the function of the gene tested. Differences in cell lines used might,therefore, cause discrepancies in experimental outcome. To achieve a matchbetween the model and the disease we have used in our experiment the humansmall cell lung cancer cell line GLC45.

Summary and General Discussion

133

Another reason for experimental differences could be the lack of statistical power ofin vivo experiments. For our experiments, the number of mice necessary to achievestatistically significant results was calculated in advance. We had to use three to fourtimes as many mice as in other studies in order to demonstrate statisticalsignificance in a number of comparisons using Scheffe’s F-test for multiplecomparisons. By injecting cells into both flanks of a mouse and demonstrating thatthe tumours growing on each side did not influence each other, we could on theother hand restrict the number of mice to be used.The rate of expression of a supposed tumour suppressor gene can vary betweenexperiments. Usually cDNA of genes are used in functional assays. The cDNAshave mostly been cloned behind strong viral promoters that cause a greatoverexpression. The resulting abundant production of a single protein may welldisturb essential cellular processes and thereby affect growth and proliferation of thetumour cells, i.e. mimic tumour suppression. A genuine tumour suppressor gene,however, would act as such also when present in low dosage. We were able to avoidthe huge overexpression of the gene of interest by transfecting overlapping PACscontaining multiple genes from the 3p21.3 critical region into the parental cell line. Byselecting clones with a minimal PAC integration we were able to assess the tumoursuppressor function of the genes of interest in a low dosage. In addition, by usingPACs instead of cDNAs a possible tumour suppressor activity of the genes wouldpresumably be under control of its own promoter sequences.Our study suggests that genes or regulatory sequences located in the centromericpart of the 3p21.3 critical region have tumour suppressor activity. To what extent theobserved chromosomal instability interferes with this tumour suppressor activity is,however, not clear. A better understanding of the role of the centromeric part of the3p21.3 critical region in lung tumour suppression can be obtained by tumourigenicitytests of a larger number of transfectants with integrations of this same part of thecritical region. These transfectants can also be tested with expression arrays, to gaininsight in the expression profile caused by genes or regulatory sequences located onPAC 185A4 as opposed to the expression profile caused by the chromosomalinstability. The observed trans-regulation between sequences located on PAC185A4 and originally neighbouring genes should be confirmed by independentexperiments. To define the role of the newly found non-coding RNAs in lung cancerthese RNAs should be analysed in a number of cancer cell lines and normal tissues.In a way the situation in 3p21.3 in lung cancer resembles that of 13q14.2 in B-cell

Chapter 5

134

CLL, where several groups succeeded to delimit the critical region as defined byhomozygous deletions, but none could come up with a plausible candidate gene.Eventually, a critical role has been attributed to two miRNAs in the critical region(Calin et al., 2002; Calin et al., 2004; Calin et al., 2005; Chen et al., 2005) It may wellbe that regulatory RNAs, such as miRNAs and other non-coding RNAs, will also turnout to play a critical role in lung cancer.

Summary and General Discussion

135

REFERENCES

Calin GA, Dumitru CD, Shimizu M, Bichi R, Zupo S, Noch E, Aldler H, Rattan S, Keating M, Rai K,Rassenti L, Kipps T, Negrini M, Bullrich F, Croce CM. 2002. Frequent deletions and down-regulation of micro- RNA genes miR15 and miR16 at 13q14 in chronic lymphocytic leukemia. ProcNatl Acad Sci U S A 99:15524-9.

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. MicroRNAprofiling reveals distinct signatures in B cell chronic lymphocytic leukemias. Proc Natl Acad Sci US A. 101:11755-60.

Calin GA, Ferracin M, Cimmino A, Di Leva G, Shimizu M, Wojcik SE, Iorio MV, Visone R, Sever NI, FabbriM, Iuliano R, Palumbo T, Pichiorri F, Roldo C, Garzon R, Sevignani C, Rassenti L, Alder H, VoliniaS, Liu CG, Kipps TJ, Negrini M, Croce CM. 2005. A microRNA signature assocciated withprognosis and progression in chronic lymphocytic leukemia. N Engl J Med. 353:1793-801.

Chen CZ. 2005. MicroRNAs as oncogenes and tumour suppressors. N Engl J Med. 353:1768-771.

Nederlandse samenvatting

Nederlandse samenvatting

137

Het ontstaan van kanker

Wanneer de cel zich deelt, wordt het DNA, dat alle erfelijke informatie bevat,gekopieerd en vervolgens verdeeld over beide dochtercellen. Bij het kopiërenkunnen er fouten ontstaan, mutaties genoemd. Wanneer deze mutaties in een genontstaan, kan de functie van het genproduct veranderen. De chromosomen zijn intweevoud aanwezig waardoor ook alle genen in tweevoud aanwezig zijn. Als in éénvan beide kopieën een mutatie ontstaat, is nog één functionerend gen aanwezig,waardoor er meestal nog voldoende genproduct wordt aangemaakt om de celnormaal te laten functioneren. Wanneer echter in beide genen een mutatie optreedt,kan het genproduct niet meer worden gemaakt en kan dit gevolgen hebben voor hetfunctioneren van de cel. Als het desbetreffende genproduct als functie het remmenvan de celdeling heeft, kan de celdeling verstoord raken, waardoor ongebreideldeceldeling plaatsvindt. Het gen dat codeert voor een dergelijk genproduct wordt eentumorsuppressor-gen genoemd. Daarnaast komen oncogenen voor diegroeibevorderend werken. In tumoren zijn door genomische veranderingen ofmutaties ofwel oncogenen geactiveerd, d.w.z. dat een oncogen meer product maaktdan normaal, ofwel zijn beide tumorsuppressor-genen uitgeschakeld. Dit laatstegebeurt vaak doordat één gen is uitgeschakeld door het verlies van een groot stukchromosoom, een deletie genoemd. Het andere gen wordt dan vaak uitgeschakelddoor een mutatie in het gen zelf, maar ook dat kan gebeuren door een deletie.Wanneer op beide chromosomen een groot stuk van hetzelfde gebied weg is,spreekt men van een homozygote deletie. Genen in deze gebieden zijn per definitiecompleet uitgeschakeld. Deze homozygote deleties zijn vaak kleiner dan gebiedendie op één van beide chromosomen gedeleteerd zijn. In de gebieden diehomozygoot gedeleteerd raken wordt vaak verondersteld dat er tumorsuppressor-genen liggen.

Het onderzoek

Hoewel voor een aantal typen kankers tumorsuppressor-genen bekend zijn, is hettumorsuppressor-gen dat verantwoordelijk is voor het ontstaan van longkanker nogniet ontdekt. Het eerste hoofdstuk beschrijft wat de literatuur hierover vermeldt.Beschreven wordt dat in longkanker tumoren vaak een gedeelte van de korte armvan één van de chromosomen 3 gedeleteerd is. Er worden zeven van dergelijke

138

gebieden beschreven die niet met elkaar overlappen, te weten de 3p22 AP20 regio,de 3p21.3 CER1 en CER2 regio’s, de 3p21 D3F15S2 regio, de 3p21.3 LUCA regio,de 3p14 FHIT regio en de 3p12 ROBO1 regio. Voor geen van de genen in al dezegebieden is echter overtuigend bewijs gevonden voor een tumorsuppressor rol inlongkanker.

In dit onderzoek is met behulp van functionele testen in muizen meer specifiekgekeken naar de rol van de 3p21.3 LUCA regio. Hiervoor is het gebied overdekt metgrote, elkaar overlappende fragmenten DNA die in bacteriën worden gekweekt(PACs). In hoofdstuk 2 worden deze PACs ingebracht in een longkankercellijn,GLC45. De PACs blijken vervolgens te integreren in de chromosomen van de cellijn.Klonen van deze cellijn waarin een PAC geïntegreerd is, zijn onderhuids ingespotenin naakte muizen waarin ze tumoren vormen. De tumorgroei van de klonen isvergeleken met de tumorgroei van de ouderlijke cellijn, GLC45. Tumoren dieontstonden uit twee klonen met elk één en dezelfde PAC (PAC 185A4) bleven nainspuiten in naakte muizen significant kleiner dan de tumoren ontstaan uit deouderlijke cellijn. De genen CYB561D2 en PL6 of regulatoire volgorden op de PACzouden verantwoordelijk kunnen zijn voor de geremde tumorgroei. Dit biedt echtergeen verklaring voor de eveneens gevonden verminderde tumorgroei van één vande klonen waarin alleen de lege PAC-vector geïntegreerd was. In alle klonen werdevenwel na het inbrengen van de PAC in de cel en het integreren van de PAC in hetchromosoom een sterke toename van chromosomale veranderingen gezien tenopzichte van de ouderlijke cellijn. Gedeelten van chromosomen bleken in klonen ofgedeleteerd of gedupliceerd ten opzichte van de ouderlijke cellijn. Hierdoor kan deverminderde tumorgroei zowel veroorzaakt worden door nieuw ingebrachte genen ofregulatoire volgorden als door de opgetreden instabiliteit.

In hoofdstuk 3 wordt de consequentie voor mRNA productie geëvalueerd naintroductie van PAC 185A4 in the longkankercellijn, GLC45. Hiervoor werd eenanalyse toegepast waarbij in één keer mRNAs van ongeveer 21.000 genengescreend kunnen worden. De hoeveelheid mRNA per gen in beide klonen van delongkankercellijn met PAC 185A4 werd vergeleken met de hoeveelheid mRNA pergen in de ouderlijke longkankercellijn. Met behulp van een significantie analysewerden die genen opgezocht waarvan ten opzichte van de ouderlijke cellijnsignificant meer of minder mRNA in de afzonderlijke klonen aanwezig was. Maar eenklein percentage van de genen kwam voor in beide genen-lijsten. Dit kon verklaard

Nederlandse samenvatting

139

worden door de sterke correlatie die werd gevonden tussen de eerder gevondendeleties en duplicaties op chromosoom niveau en de hogere en lagere mRNAhoeveelheden op gen niveau. Waar in de kloon ten opzichte van de ouderlijke cellijneen deletie op chromosoom niveau werd gevonden was de mRNA hoeveelheid voorde genen in dat gebied over het algemeen verlaagd. Voor een duplicatie was demRNA hoeveelheid over het algemeen verhoogd. Een opmerkelijke bevinding washet feit dat de mRNA hoeveelheid van twee genen, TUSC4 en CACNA2D2,afkomstig uit de gebieden die het gebied in PAC 185A4 flankeren, in de klonenverhoogd was ten opzichte van de ouderlijke cellijn. Het lijkt of deze genenaangestuurd worden door volgorden aanwezig op PAC 185A4. Doordat integratievan PAC 185A4 niet op chromosoom 3, maar op andere chromosomen gebeurde,worden deze genen kennelijk op afstand gereguleerd door volgorden op PAC 185A4.

In hoofdstuk 4 wordt een homozygote deletie beschreven in de kleincelligelongkanker cellijn GLC20. De homozygote deletie is gevonden in de 3p12 band, dusnaast de homozygote deletie die in die cellijn gevonden is in de 3p21.3 LUCA regio.De grootte van de deletie werd met behulp van fiber-FISH technieken bepaald op110 kb -130 kb. Een klein gedeelte van het gen ROBO1 blijkt gedeleteerd. In de niet-coderende volgorden van ROBO1 werden twee nieuwe RNA’s gevonden. DezeRNA’s die niet voor eiwitten coderen, worden niet-coderende RNA’s genoemd. Zekunnen betrokken zijn bij de uitschakeling van andere genen. In deze RNA’s werdenmogelijk ook gebieden gevonden waar kleine niet-coderende RNA’s (miRNA’s)aanhechten die de eerst genoemde RNA’s remmen of afbreken. Volgorden van eendergelijke kleine niet-coderende RNA werd tevens gevonden in het gebied dathomozygoot gedeleteerd is.

Hoewel het tumorsuppressor-gen verantwoordelijk voor het ontstaan van longkankernog niet gevonden is, verschaffen de resultaten van dit onderzoek aanwijzingen dathet LUCA gebied is ingeperkt naar twee mogelijke tumorsuppressor-genen dan welregulatoire volgorden in een gebied van ongeveer 100 kb. Doordat de instabiliteit inalle klonen verschillend is, zou met het testen van meer klonen waarin PAC 185A4 isgeïntegreerd in naakte muizen en d.m.v. mRNA analyses, een duidelijkeronderscheid gemaakt kunnen worden tussen wat het effect van het inbrengen vannieuwe DNA volgorden is en wat het gevolg is van toegenomen chromosomaleinstabiliteit.

Dankwoord

Dankwoord

141

Mijn proefschrift is klaar! En waar anderen misschien alleen maar blij zijn dat heteindelijk af is, kijk ik toch met enige weemoed terug op een ontzettend leuke enleerzame tijd. Maar, ik kan nu wel eindelijk alle mensen bedanken die mij al die jarengeholpen hebben met het tot stand komen van dit boekje.Om te beginnen Charles, mijn promotor. Ik wil je allereerst bedanken voor hetvertrouwen dat je in mij stelde door me aan te nemen op deze functie. Het bleekuiteindelijk geen makkelijk project, maar ik vind dat we ons er goed doorheenhebben geslagen. Ik wil je ook bedanken voor de onderhoudende gesprekken die wehebben gehad over de verschillende aspecten van het onderzoek en voor decomplimenten die je dikwijls gaf. Hierdoor kon ik met meer zelfvertrouwen deverschillende taken tegemoet zien. Ook het geduld waarmee je mijn stukken nakeekwas ongelofelijk. Ik denk dat je menigmaal tenenkrommend naar mijn punten enkomma’s hebt gekeken. Eén ding is zeker: ik heb er een hoop van opgestoken.I would like to thank Prof. dr. S.L. Naylor, Prof. dr. S. Imreh and Prof. dr. Harry Groenfor their willingness to be a member of the promotion committee. I would also like tothank all co-authors of the articles in my thesis.Frans, ook al ben je officieel geen co-promotor, je hebt me enorm geholpen met hetdiscussiëren over het longkanker onderzoek, je speelde graag de advocaat van deduivel zoals je dat zelf zei. Ik kon ook altijd bij jou terecht met alle “vreemde”resultaten en ideeën en je gaf mij regelmatig advies over hoe ik de experimenten hetbeste kon uitvoeren. Daarnaast heb je me natuurlijk ook enorm geholpen bij deexpressie-array analyse, die zonder jou nooit tot een artikel had geleid.Dan natuurlijk mijn eerste analiste en paranimf Bea. Wij hebben het samen gedaan,ook met jou heb ik veel gediscussieerd over ons onderzoek en de experimenten, endat kwam vooral omdat je altijd met mij mee hebt gedacht. Natuurlijk heb je ook heelveel praktisch werk voor me verricht, waarbij hele dagen kweken natuurlijk niet altijdde leukste bezigheid was. Je hebt me mijn grenzen laten verleggen doorbijvoorbeeld te zeggen dat ik best zelf de muizen dood kon maken. Je bent ook echteen vriendin van me geworden en om met een vriendin samen op een project tewerken is natuurlijk heel prettig. Daarvoor wil ik je heel graag bedanken.Wytske, mijn eerste stagiaire en later mijn tweede analiste, je was de perfectestagiaire, leergierig, heel erg netjes (voorbeeld voor mij en Bea) en daarbij ook nogeens heel gezellig. Je hebt voor mij het onmogelijke gepresteerd door stabieleklonen te verkrijgen met de lege PAC vector. Daarnaast heb je natuurlijk ook heel

veel werk voor me verzet en was het altijd gezellig om samen met jou, of met z’ndrieën, naar de muizen te gaan kijken.Dan wil ik graag alle mensen van de FISH-groep bedanken en in het bijzonderAnneke, Inge en Hendrika. Jullie hebben in totaal 151 klonen voor mij gehybridiseerden bekeken. En daarnaast nog menig ander proefje voor mij verricht. Ik heb onzesamenwerking altijd als erg prettig ervaren en zonder jullie had ik deze proeven nooitkunnen voltooien.Gerard te Meerman, jou wil ik bedanken voor de statistiek van de muizenproeven ende expressie experimenten, en daarbij ook voor de goede raad met betrekking tothet onderzoek. Ik heb onder andere het aantal muizen met de helft kunnenverminderen door jouw advies om ze aan beide kanten met een PAC in te spuiten.Peter Terpstra wil ik bedanken voor het aanleveren van de marker-sequenties voorhet checken van de klonen.Ook wil ik natuurlijk de mensen van het Centraal Dieren Laboratorium bedanken, inhet bijzonder Hans Bartels, Sylvia, Ar en Hester, voor de medewerking en degezelligheid. Ik heb altijd met veel plezier proeven bij jullie gedaan.Klaas, jou wil ik bedanken voor de adviezen die je mij gaf met betrekking tot hetonderzoek, menig experiment was niet gelukt zonder jouw inbreng. Daarbij heb jeme geholpen met de analyse van de array-CGH uitkomsten. Pieter, jou ik wil ikbedanken voor het hybridiseren van de array-CGH’s, eigenlijk moest het altijd metspoed, de uitkomsten en de snelheid heb ik zeer gewaardeerd. Ook wil ik Kristabedanken, je hebt voor mij de expressie-arrays gehybridiseerd en later ook nog real-time RT-PCR gedaan. De hoge correlatie die we vonden tussen beide technieken ligtin belangrijke mate aan jouw capaciteiten als analist.Verder wil ik ‘de rest’ van mijn collega’s bij de Medische Genetica bedanken, die mijnverblijf tot een feest hebben gemaakt. Een paar mensen wil ik graag bij naamnoemen. Chantal, Petra, Marga, Jantine en Ana, mijn collega AIO’s en anderekamer/labgenoten. Het was erg gezellig om met jullie een lab/kamer te delen. Ana, Ialways had a great time with you, walking to and from the hospital en of courseeating there. Sometimes we laughed so much during lunch we almost forgot to eat.Yvonne wil ik ook bedanken voor het bijvullen van de snoeppot en het prettigereisgezelschap naar de ASHG en de ESHG. Bart en Jos voor het weer leegmakenvan de snoeppot en natuurlijk voor de antwoorden die ik kreeg op al mijn vragen. Hetis jammer dat er vaak te weinig wordt gedaan met de kennis van de mensen bij dediagnostiek (en andersom). Tegen Rein wil ik zeggen: “HET BOEKJE IS NU ECHT

Dankwoord

143

AF!!” Maar ik wil jou ook bedanken voor de gesprekken die we voerden en ik hoopdat we die voort kunnen zetten. Het secretariaat: Mentje, Ria en Ineke, jullie wil ikbedanken voor het meedenken en het traceren van Charles als ik die weer eensnodig had.Dan zijn er natuurlijk ook nog een heleboel ex-collega’s die ik hier wil noemen. In hetbijzonder, Tineke Timmer en Patrick Veldhuis, voor de gedegen opleiding die ik hebgehad van stagiaire naar analist.

I would like to thank Dr. Debora Angeloni and Dr. Michael Lerman of the Laboratoryof Immunobiology, CCR, NCI-F, Frederick, Maryland, USA, for the intensive co-operation during the stay of Dr. Debora Angeloni at the lab of Dr. Michael Lerman aswell as after her return to Italy.

En natuurlijk wil ik de mensen buiten het werk bedanken, mijn volleybal vriendinnenen trainers, voor de broodnodige afleiding. En in het bijzonder Gwenny, volgens mijwerd je helemaal gek van al mijn vragen over het lay-outen, maar uiteindelijk is hetallemaal gelukt.

Lieve papa en mama, jullie hebben me altijd gesteund in mijn keuzes en zonder datextra zetje wat ik af en toe nodig had, was ik nooit zo ver gekomen. Heel erg bedanktvoor alles. Frederike, mijn zusje en tweede paranimf, door jouw advies heb ikgesolliciteerd op deze AIO baan en het was de beste keuze die ik kon maken. Heelerg bedankt hiervoor.En dan natuurlijk Edwin, lieve Ed, met de kans ontzettend afgezaagd te klinken, wilik jou bedanken voor alles wat je voor me gedaan hebt. Je hebt me heel veel werkuit handen genomen en je stond altijd voor me klaar.