Prb2/P130: a New Candidate for Retinoblastoma Tumor Formation

G De Falco1,2 and A Giordano1,2

1Department of Human Pathology and Oncology, University of Siena, Siena, Italy and 2Sbarro Institute for Cancer Research and Molecular Medicine, College of Science & Technology Temple University, Philadelphia, PA 19122, USA

Retinoblastoma is the most common primary intraocular alterations of tumor suppressor genes and proto- tumor in childhood. Mutations in both the alleles of the oncogenes that mayoccur if genomic integrityis RB1 gene represent the causative agent for the tumor to compromised owing to intrinsic factors or exogenous occur. It is becoming evident that, although these altera- agents. The mechanisms of inactivation of tumor tions represent key events in the genesis of retinoblastoma, suppressor genes and proto-oncogenes are different: they are not sufficient per se for the tumor to develop, and tumor suppressors are inactivated by‘loss of function other additional genetic or epigenetic alterations must mutations’, whereas proto-oncogenes are activated occur. A supportive role in the genesis of retinoblastoma through ‘gain of function mutations’. Tumor suppres- has recently been proposed for the RB1-related gene RB2/ sors are functionallyinactive in cancer cells, thus p130. Additionally, several other genetic alterations invol- determining the loss of control over cell proliferation; ving different chromosomes have been described as relevant on the other hand, proto-oncogenes are constitutively in the tumorigenic process. In this review we will analyse activated, leading to continuous signaling which acts current knowledge about the molecular mechanisms positivelyon cell growth. The different mechanism of involved in retinoblastoma, paying particular attention to inactivation means that when cancer depends on the the mechanisms of inactivation of the biological function of inactivation of a tumor suppressor gene, both the alleles the retinoblastoma family of proteins. of the tumor suppressor gene must be inactivated for the Oncogene (2006) 25, 5333–5340. doi:10.1038/sj.onc.1209614 tumor to occur, even though several lines of evidence indicate that this is not true in the case of haploid Keywords: cell cycle; cancer; RB family; retinoblastoma; insufficiency. Proto-oncogene-driven transformation, on pRb2/p130; tumor suppressors the contrary, requires the mutation of a single copy of the proto-oncogene which then becomes activated, leading cells toward transformation. The first tumor suppressor identified was RB1 (the retinoblastoma susceptibilitygene), mutations of which Introduction are responsible for a childhood retinal malignancy, the retinoblastoma (Friend et al., 1986; Fung et al., 1987; The fact that inherited abnormalities predispose some Lee et al., 1987a, b). The importance of hereditary individuals to cancer was first hypothesized more than factors for the occurrence of this pediatric cancer has one centuryago, but onlyin the past 20 yearshas it been been evident since 1821, when a familial aggregation of possible to identifyseveral genes responsible for retinoblastoma was observed (Kaelin, 1955). Initially, tumorigenesis (Li-Fraumeni syndrome). Cancer may all cases were considered to be inherited diseases, but it be considered a genetic disease, occurring when normal subsequentlybecame clear that there were non-inherited cells progressivelytransform through the acquisition of forms of retinoblastoma (Vogel, 1954). The problem mutations in the genome. When this happens, cells was unraveled in 1971, when Knudson postulated a continue to proliferate independentlyof regulatory model to explain the different origins of hereditaryand influences that control cell growth, whose regulation non-hereditaryretinoblastomas, much earlier than depends on negative and positive growth signals. cloning of the causative gene took place. Negative regulation is exerted bya familyof genes known as tumor suppressors, which regulate cell growth and homeostasis. The proto-oncogenes, whose function Retinoblastoma: the pediatric tumor is to promote cell proliferation, carryout positive regulation of cell growth. It has become increasingly Retinoblastoma is a rare malignant tumor of the eye evident that cancer is a genetic disease, requiring that originates in developing cells of the retina. It has an incidence of 1 in 20 000 live births in all human races Correspondence: Dr A Giordano, Sbarro Institute for Cancer (Suckling et al., 1982), and does not varyaccording to Research and Molecular Medicine, College of Science & Technology Temple University, Biolife Science Bldg. Suite 333, Philadelphia, geographical location or level of industrialization. PA 19122, USA. Diagnosis is possible on the basis of clinical symptoms, E-mail: [email protected] usuallywithin the first 5 yearsof age. Only10% pRb2/p130: a new candidate for retinoblastoma tumor G De Falco and A Giordano 5334 of affected children have a familial historyof retino- model to explain the difference between hereditaryand blastoma, therefore most of the time the tumors are nonhereditaryforms in 1971. This model, referred to as discovered when one or both eyes are already full of the ‘Two-hit hypothesis’, highlights that the phenotypic tumor, giving the pupil a ‘cat’s eye’ appearance due to a difference between these two forms of retinoblastoma white papillaryreflex (leukocoria) which maybe maybe explained bydifferent molecular mechanisms. accompanied bystrabismus. Retinoblastoma occurs in For the tumor to occur inactivation of both the alleles of both hereditaryand nonhereditaryforms: the hereditary the RB1 gene is necessary, but there is a different form accounts for 40% of patients and demonstrates an mechanism for this inactivation. Knudson hypothesized autosomal dominant inheritance; patients have retino- that heritable retinoblastoma tumors depend on a RB1 blastoma in both eyes (bilateral retinoblastoma), often gene mutation, which occurs in germinal cells and can showing more than one focus per eye (bilateral multi- therefore be inherited. Although this first mutation (M1) focal retinoblastoma) and a lifelong predisposition to renders people susceptible to cancer, in not sufficient per cancer throughout the body(Onadim et al., 1992). se for the tumor to develop. In fact, the normal allele About 60% of patients have retinoblastoma in onlyone still provides the required function of RB1 in the eye (unilateral retinoblastoma) and this represents a constitutional cells, in subjects carrying a germline sporadic disease, as there is no familyhistoryfor this RB1 mutation. Tumors will occur onlywhen an cancer. Nonhereditaryretinoblastoma consists of a additional, somatic mutation (M2) arises in retinal cells unifocal retinal tumor and involves no increased risk in the second copyof the gene (the second hit), thus of other cancers. Mutations of both the alleles of inactivating the function of the RB1 gene. In non- the RB1 gene are necessaryfor the disease to occur hereditaryretinoblastoma, both the events are supposed (Friend et al., 1986; Fung et al., 1987; Lee et al., to be somatic (Knudson, 1971; Knudson et al., 1973). 1987a, b). In patients with sporadic unilateral retino- The incidence of nonhereditaryforms depends upon a blastoma, the two RB1 mutations that initiate tumor first somatic mutation in the developing retina, growth development are somatic events and therefore they of the mutant clone, and a subsequent mutation in one cannot be inherited, and none of them are present in cell of that clone. This explains the high incidence of DNA from constitutional cells (Shimuzu et al., 1994; bilateral tumors in patients carrying germline mutations, Lohmann et al., 1997; Klutz et al., 1999). Patients and the younger ages at first appearance of tumor. In affected bybilateral retinoblastoma are heterozygous most hereditaryforms, there are new germinal muta- for an RB1 mutation, which was either inherited or tions and both parents are normal. In the hereditary occurred de novo in parental germline cells or during forms, therefore, M1 occurs in the germinal cells, embryonal development. During early childhood, these whereas M2 occurs in a retinal cell; in nonhereditary patients are predisposed to retinoblastoma because a tumors both M1 and M2, which inactivate both the RB1 mutation in the wild-type allele will lead to biallelic alleles, must arise in the same retinal cell (Figure 1). inactivation of the RB1 gene, therebyinitiating the Almost all kinds of RB1 mutations have been described, development of the tumor focus. In most families with including point mutations, insertions, deletions, trans- retinoblastoma, all members who have inherited a locations (for a review, see Lohmann, 1999). In mutation develop bilateral retinoblastoma. However, addition, epigenetic events such as promoter hyper- in some exceptional families unilateral retinoblastoma is methylation have been described as causative agents for frequent and some mutation carriers remain unaffected (low-penetrance retinoblastoma). In addition to retinoblastoma, carriers of an RB1 mutation also show a high M1 incidence of other tumors that are collectivelyknown as M1 second primaryneoplasms. These include bone and soft tissue sarcomas, malignant melanoma and neoplasms of M1 the brain and meninges (Eng et al., 1993). Diagnosis of M1 M1 patients with bilateral retinoblastoma is made earlier M1 M1 M2 than patients with unilateral retinoblastoma (11 and 22 M1 M2 months, respectively). Compared to other solid tumors, retinoblastomas are small at the time of the diagnosis M1 and can be successfullytreated in more than 95% of M1 M1 cases (de Sutter et al., 1987), although patients still require constant lifelong follow-up for earlydetection of recurrent tumors or detection of secondarytumors M1 (for a review about diagnosis and treatment of retino- M1 blastoma, see Balmer et al., in this issue). M1 Molecular basis of retinoblastoma Hereditary form Non-hereditary form Light was shed on the molecular mechanisms underlying Figure 1 Retinoblastoma tumors occur when two mutations retinoblastoma development when Knudson proposed a (M1 and M2) arise in the same cell.

Oncogene pRb2/p130: a new candidate for retinoblastoma tumor G De Falco and A Giordano 5335 RB1 functional inactivation (Sakai et al., 1991; Zesch- which ensures the correct folding and the stabilityof the nigk et al., 1997; Choy et al., 2002). protein (Lee et al., 1998). The integrityof these domains In this review, we will summarize the current knowl- seems to be important for pRb/p105 function, as edge about retinoblastoma, with particular emphasis on mutations occurring within this region appear to be the retinoblastoma familyof proteins, in order to tumorigenic. understand the molecular mechanisms underlying the The carboxyl-terminal domain of the pRb/p105 disease. protein is less characterized, but it is critical for growth suppression (Whitaker et al., 1998). This region contains a nuclear localization signal, which allows the shuttling of the protein from the cytoplasm to the nucleus and RB1: the gene and the protein structure also permits the physical transportation to the nucleus of proteins lacking a nuclear localization signal. The The RB1 gene consists of 27 exons scattered over 180 kb C-terminal domain has an intrinsic nonspecific DNA- at chromosome 13q14 (Toguchida et al., 1993). The gene binding activity(Lee et al., 1987a, b) and binds to the is transcribed in a 4.7 kb messenger RNA (Lee et al., oncoproteins c-abl and MDM2 (Whitaker et al., 1998). 1987b), whose open reading frame of 2.7 kb encodes a Mutations occurring in the C-terminal domain deter- ubiquitouslyexpressed protein of 928 residues (Lee mine low-penetrance retinoblastoma, suggesting the et al., 1987a). This protein migrates as a 110 kDa biological relevance of this domain (Bremner et al., phosphoprotein (Lee et al., 1987a), often referred to as 1997). pRB110 (pRb/p105), which has two paralogues, pRb2/ Although the A/B domains and the C-terminal region p130 and p107, with which it constitutes the retino- of pRb/p105 mediate interaction with most of its blastoma familyof proteins. These three proteins share associated proteins, there is evidence suggesting that sequence similarityand act on cell cycleregulation, but the N-terminal region of pRb/p105 is also important for theycannot be considered redundant, as their functions its tumor suppression function. The N-terminus is not are not completelyoverlapping (Claudio et al., 1994). completelycharacterized: like the C-terminus, this Structurally, the proteins belonging to the retino- region also binds to the pocket domain. This binding blastoma familycan be divided in to three domains: seems to be important because mutations in the the N-terminal domain, the T-antigen-binding region N-terminus of pRb/p105 abolish some functions exerted and the C-terminal domain; each with distinctive bythe pocket region (Qian et al., 1992; Shen et al., 1995; functional properties. The T-antigen-binding region Sellers et al., 1998). In addition, the integrityof the N- can be further divided in to three subdomains: the A, terminal region seems to be necessaryfor the pRb/p105 the spacer and the B regions. The homologyof the three protein to function; we mayhypothesize that the proteins resides in the A/B domain. N-terminal domain maystabilize an active pRb/p105 The A/B regions are conserved throughout the family conformational state. Another explanation is that the and are responsible for the binding to most of the Rb- N-terminal region binds to other proteins (Durfee et al., associated proteins (Chen et al., 1995). The spacer 1994; Chen et al., 1996; Sterner et al., 1995, 1996, 1998) region has a different sequence, which renders unique and maytherefore coordinate with other functions in each member of the retinoblastoma family. The spacer pRb/p105 tumor suppression. region is dispensable for pRb/p105 function, whereas this region is more conserved between p107 and pRb2/ p130 and mediates their interaction with specific cyclin/ cdk complexes (Ewen et al., 1992; Adams et al., 1996; Role of pRb/p105 in cell cycle control: transcriptional Lacyand Whyte,1997). repression Manybiological functions exerted bythe Rb proteins are owing to the integrityof the A/B domains, including To avoid the transmission of genetic alterations and to regulation of growth and differentiation, transcription permit cell replication onlywhen it is required, the cell regulation and the interaction with cellular and viral cycle is tightly regulated at the restriction points. Cells proteins, such as the adenovirus E1A, SV40T antigen, that need to enter the S phase will go through a the human papilloma virus E7 (for a review, see restriction point, which is controlled bypRb/p105 Kouzarides, 1995) and the HIV-1 Tat protein (De Falco during transition from earlyG1 to late G1/S phase et al., 2003). Crystallographic analysis of the pRb/p105 (Sherr, 1996). Passage through this point commits structure reveals multiple protein-binding interfaces in normal cells to DNA synthesis and cell division. Loss the A/B domains (Lee et al., 1998), suggesting that pRb/ of control of this restriction point, as a consequence of p105 can simultaneouslybind to several proteins, such loss of function of pRb/p105, does not interrupt cell as E2Fs and HDACs, through the E2F-binding site and cycle progression. Cells without proper control respond the LXCXE motif-binding site, respectively. There are improperlyto mitogenic signals, leading to abnormal also some proteins that have neither binding sites that cell proliferation in the neoplastic state. can still bind to this region (Chen and Wang, 2000). The regulation of the G1/S transition bypRb/p105 The A and B domains require each other to form a has been investigated intensively. pRb/p105 binds to the functional repressor motif (Chow and Dean, 1996), E2F familyof transcription factors, a group of because the A domain forms a scaffold to the B domain, sequence-specific DNA-binding proteins that regulate

Oncogene pRb2/p130: a new candidate for retinoblastoma tumor G De Falco and A Giordano 5336 transcription of genes required for entryinto S phase recognition motifs for phosphorylation. Hypopho- and DNA synthesis (Dyson, 1998; Nevins, 1998). Five sphorylated pRb/p105 binds to E2Fs and arrest cells of the six known members of the E2F familyare able to in G1. When cells need to proceed through the cell bind to pRb/p105. This interaction functionallyre- cycle, inactivation of pRb/p105 occurs through several presses transcription, so cells are blocked in the G0/G1 CDK-mediated phosphorylations, which start with phase and do not proceed through the cell cycle. This phosphorylation mediated by the cyclin D/CDK4,6 phenomenon depends on the phosphorylation status of complexes (Ewen et al., 1993; Kato et al., 1993), pRb/p105, which is in its active form when hypopho- occurring at the C-terminus of the protein, which then sphorylated and in this conformation it binds to E2Fs. becomes able to interact with the A/B domains and to The cell will remain quiescent until mitogenic signals displace some of the proteins bound to the pocket region induce pRb/p105 phosphorylation, which will inactivate (Harbour et al., 1999). Following cyclin D/CDK4,6 it and will allow the release of the E2Fs. Several models phosphorylation, associated proteins such as the have been postulated to explain transcriptional repres- HDACs that bind to pRb/p105 through their LXCXE sion mediated bypRb/p105. In the first one, pRb/p105 motif are dislodged (Harbour et al., 1999), but proteins represses transcription bybinding to the activation that do not have this motif, such as the E2Fs, remain domain of E2F, preventing its interaction with other bound to pRb/p105. At this point another complex, factors. Alternatively, pRb/p105 may be tethered to a cyclin E/CDK2, continues the phosphorylation process: promoter byE2F and can simultaneouslybind to the latter phosphorylation events unfold the A/B another activator, thus repressing two activators at the domain, thus releasing proteins bound to the pocket same time (Weintraub et al., 1995). Transcription may region, such as E2Fs, that can now bind to other also be activelyrepressed bypRb/p105, which maycarry complexes, allowing cell cycle progression (Harbour out this function byrecruiting other co-repressors et al., 1999; Zhang et al., 2000). (Buyse et al., 1995; Qian and Lee, 1995; Brehm et al., 1998; Luo et al., 1998; Magnaghi-Jaulin et al., 1998; Meloni et al., 1999). Manyof the co-repressors that facilitate active repression mediated bypRb/p105 act on RB1 gene mutation in retinoblastoma chromatin. A typical example is represented by histone deacetylases (HDACs), which remove acetyl groups Current knowledge indicates that mutations affecting from the histones, rendering them more attached to both alleles of the RB1 gene represent a prerequisite for DNA and giving the chromatin a less accessible the development of this tumor (Friend et al., 1986; Fung configuration (Brehm et al., 1998; Kadonaga, 1998; et al., 1987; Lee et al., 1987a, b), even though it is Luo et al., 1998; Magnaghi-Jaulin et al., 1998). pRb/ becoming evident that other genes, such as the RB1- p105 binds to HDAC1, 2 and 3 (Brehm et al., 1998; Luo related RB2/p130 gene, maybe relevant for the disease et al., 1998; Magnaghi-Jaulin et al., 1998; Lai et al., to develop. In addition, other alterations occurring at 1999; Dahiya et al., 2000). pRb/p105 mayrecruit different chromosomes maycontribute to the tumor- HDACs to DNA-bound E2F or other transcription igenesis of retinoblastoma. In this section, we will factors and alters the chromatin structure of its target summarize the current understanding about RB1 gene genes into a repressed-favored status. pRb/p105- mutations. mediated transcriptional repression mayalso involve Heterozygous carriers of oncogenic RB1 gene muta- alteration of a higher order chromatin structure, which tions show variable phenotypic expression. Patients may is regulated bynucleosome-remodeling complexes develop tumors in both eyes or only in one eye because (Trouche et al., 1997; Strobeck et al., 2000; Zhang of the variable expressivity. In addition, some carriers et al., 2000). Transcription regulation mediated bypRb/ show no retinoblastoma at all because the mutations p105 does not necessarilyinvolve HDAC activityand mayhave an incomplete penetrance. Almost all typesof chromatin remodeling, as pRb/p105 is able to repress the RB1 gene mutations have been identified, including transcription without the requirement of HDAC activity translocations, deletions, insertions, point mutations (Ross et al., 1999). and epigenetic alterations (hypermethylation of the CpG-islands in the RB1 promoter region) (for a review see Lohmann and Gallie, 2004). The majorityof germline mutations that have been Regulation of pRb/p105 activity identified in patients with hereditaryretinoblastoma are nonsense or frameshift mutations, mostlylocated in the pRb/p105 activityis regulated byphosphorylationin exons 1–25 of the RB1 gene (Sampieri et al., 2005). a cell cycle-dependent manner (Chen et al., 1989; These kinds of mutations will lead to abortive tran- De Caprio et al., 1989; Mihara et al., 1989). Hypo- scripts that will be degraded, leaving onlythe transcript phosphorylated forms of pRb/p105 predominate in of the normal allele. These mutations will have great earlyG1 phase and reappear during M phase, whereas impact on the phenotype because they completely hyperphopshorylated forms of pRb/p105 are present abolish the activityof one allele, therefore the affected from late G1 and throughout S, G2 and M phase (Chen patients will most probablydevelop bilateral retinoblas- et al., 1989; De Caprio et al., 1989; Mihara et al., 1989; toma. When a nonsense or frameshift mutation occurs Ludlow et al., 1990). The protein contains 16 CDK in the last two exons, the transcripts maymaintain some

Oncogene pRb2/p130: a new candidate for retinoblastoma tumor G De Falco and A Giordano 5337 residual biological activity, which will support that of pRb2/p130 has been postulated in the genesis of the normal allele, and this will result in a milder effect in neoplasms of the lung (Claudio et al., 2000), breast terms of phenotype. In some cases, however, carriers (Macaluso et al., 2003), ovary(D’Andrilli et al., 2004) with frameshift or nonsense mutations will not result in and non-Hodgkin’s lymphomas (Lazzi et al., 2002). the development of bilateral retinoblastoma because of In addition, inactivation of its biological function owing mosaicism, in which not all the cells will carrythe to interaction with viral oncoproteins has also been mutations and the phenotypic expression of the tumor described in mesothelioma (Mutti et al., 1998) and in will be less aggressive (Hall, 1988). AIDS-related lymphomas (De Falco et al., 2003). An Point mutations in intronic or exonic regions cause involvement for pRb2/p130 in retinoblastoma tumor aberrant splicing and abolish the biological function of development has recentlybeen hypothesized. It has been pRb/p105. These splice mutations are usuallyassociated demonstrated that pRb2/p130 expression often strongly with complete penetrance and have a great phenotypic decreases in primaryretinoblastoma tumors (Bellan impact. Missense mutations, on the other hand, will et al., 2002). In addition, pRb2/p130 expression seems to inﬂuence the phenotype to a variable degree depending correlate with the apoptotic index (AI) of the retino- on the residue that will be substituted and its location blastoma samples, as tumors that do not express pRb2/ within the protein. Generally, when these mutations p130 have a lower AI than tumors where pRb/p130 determine an alteration, that is in terms of electric expression is higher or normal (Bellan et al., 2002). This charge, folding of the protein, etc., which will inﬂuence suggests that pRb2/p130 mayexert a proapoptotic effect the biological activityof pRb/p105, the phenotypewill on the retinoblastoma tumor samples in vivo.In be more dramatic, compared to that resulting from addition, mutational analysis performed in both familial mutations that do not affect its biological activity. Most and sporadic retinoblastoma samples revealed that the missense mutations determine amino-acid changes in the RB2/p130 gene is often mutated in sporadic retinoblas- pocket region, whereas onlya few of them have been toma, and that the mutations occur mainlyin exon 1 described affecting the promoter gene of the RB1 gene (Tosi et al., 2005). Hypermethylation of the CpG islands (Lohmann, 1999). of the RB2/p130 gene has also been described as an Another cause of inactivation of the pRb/p105 alternative mechanism for inactivating the pRb2/p130 protein function is the hypermethylation of the CpG- rich islands in the promoter region of the RB1 gene, which is normallyunmethylated.These account for 10% of the retinoblastoma cases (Greger et al., 1989; a Ohtani-Fujita et al., 1997; Klutz et al., 1999) and cannot be inherited, because hypermethylation represents an RB2/p130 apoptosi epigenetic event. M1 M2 s

Role of the RB1-related RB2/p130 gene in the genesis of b retinoblastoma RB2/p130 Notwithstanding the fact that mutations of the RB1 apoptosi gene are common to all retinoblastomas, much evidence M1 M2 s indicates that loss of pRb/p105 from a developing retinal cell is not sufﬁcient for malignancy(Di Ciommo et al., 2000). The possibilitythat further mutations of other tumor suppressor genes could occur in sporadic c retinoblastoma is supported byrecent studies providing RB2/p130 direct evidence that loss of pRb/p105 function leads to genomic instabilityand predisposes to cancer by M1 M2 increasing the DNA mutation rate (Zheng et al., 2002). A possible candidate for contributing to retinoblastoma tumor formation is the RB1-related gene RB2/ RB2/p130 p130, which encodes the pRb2/p130 protein and is a RB2/p130 member of the retinoblastoma familyof proteins. The M1 M2 RB2 human gene maps to chromosome 16q12.2; the M1 M2 mRNA is 4.6 kb in length and the gene consists of 22 exons and spans over 50 kb of genomic DNA (for a Figure 2 A model to explain the role of the retinoblastoma (RB)- review see Claudio et al., 2002). Inactivation of its related protein pRb2/p130 in the genesis of RB. In (a) the cell biological function owing to genetic or epigenetic containing the M1 and M2 mutations posses a wild-type RB2/p130 gene, which leads this cell toward apoptosis. In (b) the mutated cell mechanisms or to physical interaction with viral contains an inactive pRb2/p130, which is not able to drive cell to oncoproteins has been described in neoplasms of apoptosis. This cell will therefore proliferate (c), allowing the different organs. In particular, an active role for transmission of M1 and M2 to daughter cells.

Oncogene pRb2/p130: a new candidate for retinoblastoma tumor G De Falco and A Giordano 5338 protein function in these tumors (Tosi et al., 2005). These and Yunis, 1995), monosomy17 or del(17p), i(17p) results suggest that the RB2/p130 gene maycooperate (Squire et al., 1984; Oliveros and Yunis, 1995), complete with RB1 in contributing to retinoblastoma tumor absence of pericentromeric heterochromatin on chromo- formation. A model for pRb2/p130 function in retino- some 9 (Sivakumaran et al., 1997) and loss of X or Y sex blastoma is shown in Figure 2. A role for pRb2/p130 in chromosomes (Pogosianz and Kuznetsova, 1986) have retinoblastoma maybe hypothesized: we mayspeculate also been described. The presence of abnormal gene that a functional pRb2/p130 is necessaryto address ampliﬁcations has also been discovered in retinoblasto- mutated cells to apoptosis and to maintain a high AI in ma (Lee et al., 1984; Arheden et al., 1988; Imbert et al., tumor cells. When pRb2/p130 function is lost, owing to 2001; Chen et al., 2002). Genetic alterations on mutations of the gene or to promoter hypermethylation, chromosomes 19, 20, 21, 22 and X, detected byloss of this mechanism of surveillance, which leads mutated cells heterozygosity, occur in retinoblastoma tumor samples to apoptosis is abolished, and this allows the transmission (Huang et al., 2003). of genetic alterations to daughter cells. The loss of pRb2/ p130 function maytherefore contribute to retinoblastoma tumor progression. Conclusions

Tumor suppressor gene cloning has improved our Other genetic alterations in retinoblastoma knowledge and understanding of the molecular mechanisms of cancer development. Genetic alterations of the Apart from genetic or epigenetic inactivation of the RB1 prototype of tumor suppressors, the RB1 gene, represent gene and allelic loss on chromosome 13q14, other the molecular hallmark of retinoblastoma, the most recurrent genomic alterations have frequentlybeen common childhood malignancy. Identiﬁcation of the discovered in retinoblastoma, including the isochromo- molecular mechanisms of inactivation of its biological some 6p (i6p) (Horsthemke, 1992) and trisomy1q function has shed some light on the genesis of cancer in (Cowell and Hogg, 1992). Tetrasomy6p rearrangement general. Recent discoveries have pointed out the role of presenting as i6p is considered to be almost exclusive to additional genetic or epigenetic alterations of other retinoblastoma (Murphree and Benedict, 1985) and markers that maybe involved in retinoblastoma tumor additional copies of chromosome 6p have been found formation. Among these, a role for the RB2/p130 gene in 25–60% of retinoblastomas (Squire et al., 1984; in retinoblastoma tumor formation has been emerging Turleau et al., 1995; Oliveros and Yunis, 1995). The recently, as mutations of this gene have been found presence of i6p mayrepresent an important genetic in retinoblastoma primarytumors. In addition, the change in the development or progression of retino- inactivation of its function correlates with a lower AI blastoma (Kusnetsova et al., 1982; Squire et al., 1984). of tumor cells, suggesting that its inactivation Additional copies of 6p might playa fundamental role maypromote tumor progression byacting on the in the malignancyof retinoblastoma (Cano et al., 1994) programmed cell death pathway. Additional genetic or and could alter the expression of genes of the major epigenetic alterations occurring on other chromosomes histocompatibilitycomplex on 6p21.3 (Horsthemke also seen to be relevant for a retinoblastoma tumor et al., 1989). Enhanced expression of genes mapped on to occur. Further studies to completelyelucidate 6p such as TNF-a could contribute to tumor develop- the molecular mechanisms underlying retinoblastoma ment or progression (Imbert et al., 2001). Although i6p tumorigenesis will be of help in designing new targets is the most frequent chromosomal alteration found in and novel therapeutic approaches to cure this disease. retinoblastomas with chromosome 6 abnormalities, it seems that onlythe duplication of the chromosomal region 6p21 is essential to the malignancyprocess of the Acknowledgements retinoblastoma disease (Imbert et al., 2001). Additional genetic alterations, such as monosomy16 We thank Katia Sampieri for critical comments on this paper or del(16q) (Pogosianz and Kuznetsova, 1986; Oliveros and Emma Thorleyfor manuscript revision.

References

Adams PD, Sellers WR, Sharma SK, Wu AD, Nalin CM, Cano J, Oliveros O, Yunis E. (1994). Cancer Genet Cytogenet Kaelin Jr WR. (1996). Mol Cell Biol 16: 6623–6633. 76: 112–115. Arheden K, Tommerup N, Mandhal H, Heim S, Winther J, Chen CF, Chen Y, Dai K, Chen PL, RileyDJ, Lee WH. Jensen OA et al. (1988). Cytogenet Cell Genet 48: 174–177. (1996). Mol Cell Biol 16: 4691–4699. Bellan C, De Falco G, Tosi GM, Lazzi S, Ferrari F, Morbini Chen CS, Bejcek BE, KerseyJH. (1995). Cytogenet Cell Genet G et al. (2002). Invest Ophthalmol Vis Sci 43: 3602–3608. 69: 260–265. Brehm A, Miska EA, McCance DJ, Reid JL, Bannister AJ, Chen D, Pajovic S, Duckett A, Brown VD, Squire JA, Gallie Kouzarides T. (1998). Nature 391: 597–601. BL. (2002). Cancer Res 62: 967–971. Bremner R, Du DC, Connolly-Wilson MJ, Bridge P, Ahmad Chen PL, ScullyP, Shew JY, Wang JY, Lee WH. (1989). Cell KF, Mostachﬁ H et al. (1997). Am J Hum Genet 61: 556–570. 58: 1193–1198. Buyse IM, Shao G, Huang S. (1995). Proc Natl Acad Sci USA Chen TT, Wang JY. (2000). Crit Rev Eukaryot Gene Exp 5: 92: 4467–4471. 79–95.

Oncogene pRb2/p130: a new candidate for retinoblastoma tumor G De Falco and A Giordano 5339 Chow KN, Dean DC. (1996). Mol Cell Biol 16: 4862–4868. Lee WH, Murphree AL, Benedict WF. (1984). Nature 309: ChoyKW, Pang CP, To KF, Yu CB, Ng JS, Lam DS. (2002). 458–460. Invest Ophthalmol Vis Sci 43: 1344–1349. Lee WH, Shew JY, Hong FD, SeryTW, Donoso LA, Young Claudio PP, Howard CM, Baldi A, De Luca A, Fu Y, LJ et al. (1987b). Nature 329: 642–645. Condorelli G et al. (1994). Cancer Res 54: 5556–5560. Lohmann DR. (1999). Hum Mutat 14: 283–288. Claudio PP, Howard CM, Pacilio C, Cinti C, Romano G, Lohmann DR, Gallie BL. (2004). Am J Med Genet Part C Minimo C et al. (2000). Cancer Res 60: 372–382. (Semin Med Genet) 129C: 23–28. Claudio PP, Tonini T, Giordano A. (2002). Genome Biol 3: Lohmann DR, Gerick M, Brandt B, Oelschlager U, Lorenz B, 3012.1–3012.9. Passarge E et al. (1997). Am J Hum Genet 61: 282–294. Cowell JK, Hogg A. (1992). Cancer Genet Cytogenet 64: Ludlow JW, Shon J, Pipas JM, Livingston DM, De Caprio JA. 1–11. (1990). Cell 60: 387–396. Dahiya A, Gavin MR, Luo RX, Dean DC. (2000). Mol Cell Luo RX, Postigo AA, Dean DC. (1998). Cell 92: 463–473. Biol 20: 6799–6805. Macaluso M, Cinti C, Russo G, Russo A, Giordano A. (2003). D’Andrilli G, Masciullo V, Bagella L, Tonini T, Minimo C, Oncogene 22: 3511–3517. Zannoni CF et al. (2004). Clin Cancer Res 10: 3098–3103. Magnaghi-Jaulin L, Groisman R, Naguibneva I, Robin P, De Caprio JA, Ludlow JW, Lynch D, Furukawa Y, Grifﬁn J, Lorain S, Le Villain JP et al. (1998). Nature 391: 601–605. Piwnica-Worms H et al. (1989). Cell 58: 1085–1095. Meloni AR, Smith EJ, Nevins JR. (1999). Proc Natl Acad Sci De Falco G, Bellan C, Lazzi S, Claudio PP, La Sala D, Cinti C USA 96: 9574–9579. et al. (2003). Oncogene 22: 6214–6219. Mihara K, Cao XR, Yen A, Chandler S, Driscoll B, Murphree De Sutter E, Havers W, Hopping W, Zeller G, Alberti W. AL et al. (1989). Science 246: 1300–1303. (1987). Ophthalmic Paediatr Genet 8: 85–88. Murphree AL, Benedict WF. (1985). Science 223: 1028–1033. Di Ciommo D, Gallie BL, Bremner R. (2000). Semin Cancer Mutti L, De Luca A, Claudio PP, Convertino G, Carbone M, Biol 10: 255–269. Giordano A. (1998). Dev Biol Stand 94: 47–53. Durfee T, Mancini MA, Jones D, Elledge SJ, Lee WH. (1994). Nevins JR. (1998). Cell Growth Differ 9: 585–593. J Cell Biol 127: 609–622. Ohtani-Fujita N, Dryja TP, Rapaport JM, Fujita T, Dyson N. (1998). Genes Dev 12: 2245–2262. Matsumura S, Ozasa K et al. (1997). Cancer Genet Eng C, Li FP, Abramson DH, Ellsworth RM, Wong FL, Cytogenet 98: 43–49. Goldman MB et al. (1993). JNatlCancerInst85: 1121–1128. Oliveros O, Yunis E. (1995). Cancer Genet Cytogenet 82: Ewen ME, Faha B, Harlow E, Livingston DM. (1992). Science 155–160. 255: 85–87. Onadim A, Hogg A, Baird PN, Cowell JK. (1992). Proc Natl Ewen ME, Sluss HK, Sherr CJ, Matsuhime H, Kato J, Acad Sci USA 89: 6177–6181. Livingston DM. (1993). Cell 73: 487–497. Pogosianz HE, Kuznetsova LE. (1986). Arch Geschwulstforsch Friend SH, Bernards R, Rogelj S, Weinberg RA, Rapaport 56: 135–143. JM, Albert DM et al. (1986). Nature 323: 643–646. Qian Y, LuckeyC, Horton L, Esser M, Templeton DJ. (1992). Fung YK, Murphree AL, T’Ang A, Qian J, Hinrichs SH, Mol Cell Biol 12: 5363–5372. Benedict WF. (1987). Science 236: 1657–1661. Qian YW, Lee EY. (1995). J Biol Chem 270: 25507–25513. Greger V, Passarge E, Hopping W, Messmer E, Horsthemke Ross JF, Liu X, Dynlacht BD. (1999). Mol Cell 3: B. (1989). Hum Genet 83: 155–158. 195–205. Hall JG. (1988). Am J Hum Genet 43: 355–363. Sakai T, Ohtani N, McGee TL, Robbins PD, Dryja TP. Harbour JW, Luo RX, Dei Santi A, Postigo AA, Dean DC. (1991). Nature 353: 83–86. (1999). Cell 98: 859–869. Sampieri K, Hadjistilianou T, Mari F, speciale C, Mencerelli Horsthemke B. (1992). Cancer Genet Cytogenet 63: 1–7. MA, Cetta F et al. (2005). J Hum Genet 51: 209–216. Huang Q, ChoyKW, Cheung KF, Fu WL, Pang CP. (2003). Sellers WR, Novitch BG, Miyake S, Heith A, Otterson GA, Mol Vis 9: 502–507. Kaye FJ et al. (1998). Genes Dev 12: 95–106. Imbert I, Coignet LJA, Pellestor F. (2001). Cancer Genet Shen WJ, Kim HS, Tsai SY. (1995). J Biol Chem 270: Cytogenet 128: 141–147. 20525–20529. Kadonaga JT. (1998). Cell 92: 307–313. Sherr CJ. (1996). Science 274: 1672–1677. Kaelin A. (1955). Arch Julius Klaus-Stift 30: 442–485. Shimuzu T, Toguchida J, Kato MV, Kaneko A, Ishizaki K, Kato J, Matsuhijme H, Hiebert SW, Ewen ME, Sherr CJ. Sasaki MS. (1994). Am J Hum Genet 54: 793–800. (1993). Genes Dev 7: 331–342. Sivakumaran TA, Ghose S, Kumar H, Singha U, Kucheria K. Klutz M, Horsthemke B, Lohmann D. (1999). Am J Hum (1997). Acta Genet Med Gemellol 46: 193–198. Genet 64: 667–668. Squire J, Phillips RA, Boyce S, Godbout R, Rogers B, Gallie Knudson AG. (1971). Proc Natl Acad Sci USA 68: 820–823. BL. (1984). Hum Genet 66: 46–53. Knudson AG, Strong LC, Anderson DE. (1973). Prog Med Sterner JM, Dew-Knight S, Musahl C, Kornbluth S, Horowitz Genet 9: 113–158. JM. (1998). Mol Cell Biol 18: 2748–2757. Kouzarides T. (1995). Semin Cancer Biol 6: 91–98. Sterner JM, Murata Y, Kim HG, Kennet SB, Templeton DJ, Kusnetsova LE, Prigogina EL, Pogosianz HE, Belkina BM. Horowitz JM. (1995). J Biol Chem 270: 9281–9288. (1982). Hum Genet 61: 201–204. Sterner JM, Tao Y, Kennet SB, Kim HG, Horowitz JM. LacyS, WhyteP. (1997). Oncogene 14: 2395–2406. (1996). Cell Growth Differ 7: 53–64. Lai A, Lee JM, Yang WM, De Caprio JA, Kaelin Jr WG, Seto Strobeck MW, Knuden KE, Fribourg AF, De Cristofaro MF, E et al. (1999). Mol Cell Biol 19: 6632–6641. Weissman BE, Imbalzano AN et al. (2000). Proc Natl Acad Lazzi S, Bellan C, De Falco G, Cinti C, Ferrari F, Nyongo A Sci USA 97: 7748–7753. et al. (2002). Hum Pathol 33: 723–731. Suckling RD, Fitzgerald PH, Stewart J, Wells E. (1982). Br J Lee JO, Russo AA, Pavletich NP. (1998). Nature 391: 859–865. Cancer 46: 729–736. Lee WH, Bookstein R, hong F, Young LJ, Shew JY, Lee EY. Toguchida J, McGee TL, Paterson JC, Eagle JR, Tucjer S, (1987a). Science 235: 1394–1399. Yandell DW et al. (1993). Genomics 17: 535–543.

Oncogene pRb2/p130: a new candidate for retinoblastoma tumor G De Falco and A Giordano 5340 Tosi GM, Trimarchi C, Macaluso M, La Sala D, Ciccodicola Weintraub SJ, Chow KNB, Luo RX, Zhang SH, He S, Dean A, Lazzi S et al. (2005). Oncogene 24: 5827–5836. DC. (1995). Nature 375: 812–815. Trouche D, Le ChalonyC, Muchardt C, Yaniv M, Whitaker LL, Su H, Baskaran R, Knudsen ES, Wang JY. Kouzarides T. (1997). Proc Natl Acad Sci USA 94: (1998). Mol Cell Biol 18: 4032–4042. 11268–11273. Zeschnigk M, Schmitz B, Dittrich B, Buiting K, Horsthemke Turleau C, de GrouchyJ, Chavin-Colin F, Sunien C, B, Doerﬂer W. (1997). Hum Mol Genet 6: 387–395. Seger J, Schlienger P et al. (1995). Cancer Genet Cytogenet Zhang HS, Gavin M, Dahiya A, Postigo AA, Ma D, Luo RX 16: 321–334. et al. (2000). Cell 101: 79–89. Vogel F. (1954). Z menschl Vererb- u Konstitutionslehre 32: Zheng L, Flesken-Nikitin A, Chen PL, Lee WH. (2002). 308–336. Cancer Res 62: 2498–2502.

Oncogene