Which of the following are tumor suppressor genes?

Journal Article

Takashi Kohno,

1Biology Division, National Cancer Center Research Institute, 1-1, Tsukiji 5-chome, Chuo-ku, Tokyo 104-0045, Japan

Search for other works by this author on:

Jun Yokota1

1Biology Division, National Cancer Center Research Institute, 1-1, Tsukiji 5-chome, Chuo-ku, Tokyo 104-0045, Japan

Search for other works by this author on:

Published:

01 August 1999

  • PDF
  • Split View
    • Article contents
    • Figures & tables
    • Video
    • Audio
    • Supplementary Data
    • Email
    • Twitter
    • Facebook
    • More

Close

Navbar Search Filter Microsite Search Term Search

To date, only a limited number of tumor suppressor genes have been identified as being inactivated in lung cancer. The p53 and RB genes are frequently inactivated by genetic alterations such as chromosomal deletions and loss-of-function mutations, while the p16 gene is inactivated not only by genetic alterations but also by transcriptional silencing due to hypermethylation. Recently, it was shown that the FHIT gene encompassing the chromosomal fragile site, FRA3B, is also inactivated in a large proportion of lung cancers. Several lines of evidence indicate the presence of additional tumor suppressor genes involved in lung carcinogenesis. Lung cancer cells often show deletions at multiple chromosomal regions, and deletion mapping studies have defined more than 30 regions dispersed on 21 different chromosome arms as candidate tumor suppressor loci. Several chromosomal regions hypermethylated in lung cancer cells and a number of chromosomal fragile sites have been mapped to the regions deleted in lung cancer. These chromosomal loci can harbor unknown tumor suppressor genes inactivated in lung cancer. Studies on the inherited susceptibility to lung cancer in mice have also indicated the presence of additional tumor suppressor genes for lung cancer. Further analyses of these loci should elucidate how many tumor suppressor genes are involved in human lung carcinogenesis. Molecular and functional analyses of those genes will make it possible to fully understand the molecular mechanism of lung carcinogenesis.

Introduction

Over the past two decades, significant progress has been made toward understanding the molecular pathogenesis of human cancer by the identification and characterization of various cancer-related genes that are genetically and/or epigenetically altered in human cancers. Identification of these genes has been achieved through the molecular analysis of sporadic cancers as well as the study of families with an inherited predisposition to cancer. The interplay of these two approaches has led to the identification of cancer-related genes, including tens of tumor suppressor genes that are involved in the development of both hereditary and non-hereditary cancers. Tumor suppressor genes are defined as the genes inactivated in conformity to Knudson's `2-hit' model; that is, both alleles of a gene are inactivated by genetic alterations such as chromosomal deletions and loss-of-function mutations in the development of cancer [1]. Alternatively, recent studies have also demonstrated epigenetic alterations of tumor suppressor genes in human cancers resulting in the inactivation of their functions. In particular, transcriptional silencing associated with hypermethylation of CpG islands in the tumor suppressor genes is considered to be a significant mechanism [1,2].

Lung cancers can be divided into two histological groups, small cell lung carcinoma [SCLC] and non-small cell lung carcinoma [NSCLC] [3]. NSCLC is comprised of three major subtypes; adnenocarcinoma, squamous cell carcinoma and large cell carcinoma. Cytogenetic and molecular studies have revealed that multiple tumor suppressor genes are involved in lung carcinogenesis. In fact, lung cancer cells often show deletions at multiple chromosomal regions, and several tumor suppressor genes located in these regions are frequently inactivated genetically in lung cancer cells. Here, we run through the lists of tumor suppressor genes whose alterations have been detected in lung cancers and of candidate tumor suppressor loci deduced from the presence of chromosomal deletions. In addition, recent studies have shown the frequent hypermethylation of the p16 gene associated with its transcriptional silencing in lung cancer. Furthermore, inactivation of the FHIT gene, which is located at the common fragile site, FRA3B, at chromosome 3p14, has been detected in a large portion of lung cancer. These results indicate that chromosomal sites epigenetically altered in lung cancers, or genetically fragile, can harbor unknown tumor suppressor genes. Studies on the inherited susceptibility to chemical induction of lung carcinogenesis in mouse strains have also suggested the presence of additional tumor suppressor loci. In this review, we also comment on several putative tumor suppressor loci, which have been deduced from analyses of epigenetic alterations in lung cancers, chromosomal fragility and inherited susceptibility to lung cancers.

Genetic alterations of tumor suppressor genes

Chromosomal deletions in lung cancer

Karyotypic studies have revealed that lung cancers are often aneuploid with numerical and structural changes of chromosomes [4,5]. In SCLC, recurrent losses of chromosome arms 1p, 3p, 5q and 17p have been observed. In contrast, the karyotypes of NSCLC are often more complicated, and losses of chromosome materials are often due to both numerical and structural alterations. Such alterations have frequently been detected on chromosome arms 3p, 6q, 8p, 9p, 9q, 13q, 17p, 18q, 19p, 21q and 22q in NSCLC. Recent comparative genomic hybridization [CGH] analyses have confirmed the existence of various karyotypic imbalances identified by cytogenetic studies, and have revealed several recurrent abnormalities, such as 10q deletions in SCLC, that had not been recognized previously [4].

The results of molecular studies indicate that cytogenetic data underestimate the true extent of chromosome losses in lung cancers. Loss of heterozygosity [LOH] analyses have revealed the occurrence of chromosomal deletions in lung cancers with a much higher frequency than expected by karyotypic and CGH analyses. This is probably due to the fact that loss of a maternal or paternal allele in a tumor is often accompanied by a gain of the remaining allele, since it makes chromosome loss undetectable cytogenetically [6–8]. LOH on chromosome arms 3p, 13q and 17p is the most frequent alteration in both SCLC [>90%] and NSCLC [>70%] [9–12]. LOH on 5q, 10q and 22q in SCLC and that on 6q, 9p and 19p in NSCLC occurs in >60% of cases. Furthermore, LOH on 2q, 18q, 21q and 22q was also detected preferentially in advanced NSCLC [13,14]. A moderate frequency [40–60%] of LOH was observed on chromosomal arms 1p, 4q and 11p, while a low frequency [20–40%] of LOH was detected on almost all of the remaining chromosome arms [10–12,15,16]. Deletion mapping studies have defined >30 regions dispersed on 21 different chromosome arms as targets for deletions [Table I].

Tumor suppressor genes inactivated by chromosomal deletions

As described above, cytogenetic and molecular studies have revealed loss of multiple chromosomal segments in lung cancers as non-random events. If the majority of these segments encode tumor suppressor genes, individual lung tumors must have multiple genes inactivated by chromosome deletions to become clinically evident. However, only a limited number of tumor suppressor genes have been identified as `targets' for these chromosomal deletions [Table II]. The p53, RB and p16 genes have been well accepted as being the major targets for chromosome deletions in lung cancers [17–22]. Both alleles of these genes are frequently inactivated genetically in lung cancers.

p53 functions as a trancription factor in response to DNA damage, and induces expression of downstream genes including p21, MDM2 and BAX, which regulate cell cycle and/or apoptosis [23]. p53 is believed to play a role as a `guardian' maintaining the integrity of the genome by participating in the DNA damage checkpoints in the cell cycle. It is reported that inactivation of p53 leads to an increased frequency of mutations, chromosomal rearrangements and abnormal chromosomal segregations [24–26]. p53 mutations have been detected in preneoplastic lesions of the lung, suggesting that they occur early during lung carcinogenesis [27–29]. Therefore, loss of p53 function is likely to play a key role in the progression of preneoplastic cells to neoplastic cells by accelerating the acquisition of additional genetic alterations, including numerical and structural changes of chromosomes as well as mutations of critical genes. Recently, two functional and structural homologues of the p53 gene, p73 and p51 [also designated as p40, p63 and p73L], were identified. However, their genetic alterations seem to be infrequent in lung cancer [30–32].

RB and p16INK4A/CDKN2A are involved in a signaling pathway for the regulation of the G1/S phase transition of the cell cycle [33]. Hypophosphorylated RB protein induces G1 arrest by binding and controlling cellular proteins essential for the G1/S phase transition, whereas RB protein phosphorylated by cyclin–CDK [cyclin dependent kinase] complexes lacks such an activity. CDK inhibitors, including p16, inhibit phosphorylation of RB protein by cyclin–CDK complexes. Several studies indicated that the RB and p16 genes are reciprocally inactivated in lung cancer cells, although the RB and p16 genes are preferentially altered in SCLC and NSCLC, respectively [34–36]. Due to genetic alterations, the signaling pathway via RB and p16 is estimated as being disturbed in 50–80% of lung cancers. Other RB family genes, p107 and p130, and other CDK inhibitor genes, p15INK4B/CDKN2B, p18INK4C, p19INK4D, p21CIP1, p27KIP1 and p57KIP2, are also involved in the regulation of the cell cycle. However, their genetic alterations are infrequent in lung cancer except for p15 [22,37–42]. The p15 gene is often co-deleted with the p16 gene, since they are closely located in the 9p21 region. However, since mutations of the p15 gene are rare, it may not be a target for 9p21 deletions in lung cancer [36]. The p16 locus expresses another mRNA transcript derived from alternative splicing, and a distinct polypeptide, p19ARF, is encoded in it. p19ARF functions as a growth suppressor that arrests the cell cycle in a p53-dependent manner. Since p19ARF is often co-inactivated with p16 genetically and/or epigenetically, p19ARF alterations may have a role in lung carcinogenesis [43,44].

SMAD2, SMAD4 and transforming growth factor β-type II receptor [TGFβRII] are downstream signaling mediators in a pathway of TGF-β1, a potent inhibitor of the proliferation of epithelial cells. This pathway is thought to be disturbed in nearly 50% of pancreatic cancers and ~20% of colorectal cancers by alterations of three mediator genes described above [45]. Lung cancer cell lines, especially SCLC cell lines, are often resistant to growth inhibition by TGF-β1 [44]. However, inactivation of SMAD2, SMAD4 and TGFβRII genes was observed only in a small subset of lung cancers [46–50]. Mutations of other genes involved in the pathway, SMAD1, SMAD3, SMAD5 and SMAD6, are also rare [51,52]. Therefore, TGF-β1 resistance of lung cancer cells can be caused by alterations of other components involved in the signaling pathway. The SMAD2 and SMAD4 genes are located at chromosome 18q21, while the TGFβRII gene is at chromosome 3p21–p22. Although these two regions are frequently deleted in lung cancer, other genes will probably be targets for deletions of these chromosomal regions.

The PTEN gene at chromosome 10q23 is also inactivated in a subset of lung cancers [53,54]. This gene was shown to encode a dual-specificity protein phosphatase. Recently, inactivation of the PPP2R1B gene at chromosome 11q23 was detected in 15% of lung cancers [55]. This gene encodes the β form of A subunit of the serine/threonine protein phosphatase 2A [PP2A]. The implication of PTEN and PPP2R1B alterations in lung carcinogenesis is unclear at present; however, it can be speculated that these protein phosphatases are involved in the suppression of cancer development by antagonizing protein kinases, many of which act as oncoproteins.

Other well-defined tumor suppressor genes including VHL [at 3p25], APC [at 5q21], NF2 [at 22q12] and PTC [at 9q22], which are frequently inactivated in other types of human cancers, seem to be rarely inactivated in lung cancer [56–59].

Putative tumor suppressor loci deduced from chromosomal deletions

Unknown tumor suppressor genes are considered to be present in the chromosomal regions frequently deleted in lung cancers [Table I]. To date, responsible genes for cancer-prone families have not been mapped in most of these regions by molecular epidemiological studies. Thus, at present, a large effort is being exclusively devoted to defining common regions of chromosomal deletions, and to identifying genes located in the regions, of which both alleles have been inactivated in lung cancers. The search for homozygously deleted chromosomal regions has been performed extensively in lung cancers, since the sizes of homozygously deleted regions are often smaller than those of hemizygously deleted ones. Indeed, several tumor suppressor genes, including RB, p16, SMAD4 and PTEN, have been isolated from the regions homozygously deleted in cancer cells. Until now, several genes located in the deleted chromosomal regions have been isolated as candidates for target genes. However, confirmation of these genes as tumor suppressor genes has proved difficult, since alterations in both alleles of these genes were rarely detected.

Epigenetic alterations of tumor suppressor genes

Recent studies have demonstrated that the CpG islands in the RB, p16, VHL and APC genes are frequently hypermethylated in a variety of human cancers, but not in the corresponding normal tissues [60,61]. These genes have been defined as tumor suppressor genes, since they are frequently inactivated genetically in human cancers. Since the hypermethylation has been shown to lead to the silencing of mRNA expression, this epigenetic alteration is considered to be a significant mechanism for inactivation of these tumor suppressor genes. The hypermethylation of the p16 gene is frequently observed in NSCLC, and ~30–50% of losses of p16 expression in NSCLC can be caused by hypermethylation of the 5′ regulatory region in the p16 gene [36,62,63]. Therefore, inactivation of tumor suppressor genes due to hypermethylation can play an important role in lung carcinogenesis. Several studies have suggested that a number of CpG islands at diverse chromosomal regions are hypermethylated in lung cancer cells [64–67]. Therefore, it would be worth defining the regions, which are commonly hypermethylated in lung cancer cells, because such sites may harbor novel tumor suppressor genes, which are inactivated both genetically and epigenetically in lung cancer. However, caution should be exerted when hypermethylation of CpG islands are observed in lung cancer cells, because some CpG islands are also methylated in a fraction of normal cells [68,69]. Therefore, it is possible that hypermethylation detected in lung cancer cells does not usually occur de novo during carcinogenesis, but reflects the methylation status of their precursor cells. In fact, it cannot be excluded that lung cancer cells with p16 hypermethylation have developed from precursor cells whose p16 gene was hypermethylated, since in situ methylation analysis is impossible at present. They might consist of infinitesimal fractions of normal lung cells, so that it would be difficult to detect p16 hypermethylation in those cells by the methods available.

Interestingly, the methylation of several chromosomal sites was shown to increase with age [68,69]. Furthermore, DNA hypermethylation is known to lead to chromatin condensation and late timing for the DNA replication that may underlie chromosomal instability [64,70,71]. Therefore, DNA hypermethylation may predispose inactivation of tumor suppressor genes by altering gene expression and/or triggering chromosomal deletions during the aging process. The pathogenetic significance of DNA hypermethylation needs to be elucidated in further studies.

Chromosomal fragile sites and tumor suppressor gene

The FHIT gene was identified from the homozygously deleted region at chromosome 3p14 in several human cancers, including lung cancer [72]. The discovery of the FHIT gene has drawn a great deal of attention, since the gene encompassed a common chromosomal fragile site, FRA3B. Common fragile sites have been shown to display a number of characteristics of unstable, highly recombinogenic DNA, and are preferential sites for chromosomal deletion and rearrangement [73]. Therefore, it has been hypothesized that genetic fragility of the FHIT locus predisposes to 3p14 deletions. FHIT is the first cancer-related gene, whose alterations are associated with chromosome instability. Expression of Fhit protein has been shown to be greatly reduced in a large fraction of lung cancers, and the abnormality was preferentially observed in tumors with LOH at 3p14 [74–76]. Therefore, the FHIT gene can be a target for chromosome 3p14 deletions. However, genetic alterations in the remaining alleles of the FHIT gene were rarely observed, although aberrant transcripts are often expressed in cancer cells resulting in the reduced expression of Fhit protein [77]. Mechanisms for the expression of aberrant FHIT transcripts are to be pursued.

Until now, 89 common fragile sites have been mapped on human chromosomes, and it is known that cancer-associated chromosomal deletions often involve chromosomal segments containing fragile sites [73]. Interestingly, a number of fragile sites are located in or near the chromosomal bands frequently deleted in lung cancers [Table I]. Therefore, it is possible that these fragile sites play a mechanistic role in the recurring chromosomal deletions, which lead to inactivation of tumor suppressor genes. It would be worth searching genes located at these fragile sites to identify target tumor suppressor genes for chromosomal deletions.

Genes involved in susceptibility to lung cancer

Familial aggregation of lung cancer has been suggested by statistical analyses and by case studies [78]. Familial aggregation of human cancers is partly attributable to inherited genetic defects. Up to the present, more than 10 tumor suppressor genes have been identified as being responsible for autosomal dominant hereditary cancer syndromes. To identify genes involved in susceptibility to lung cancer, we recruited families with lung cancer clustering from 1068 families by selecting patients with lung cancer as probands [79]. However, familial clustering of lung cancer was not evident in lung cancer patients. The p53, RB and p16 tumor suppressor genes, which are responsible for Li-Fraumeni syndrome, familial retinoblastoma and familial melanoma [1,2], are frequently mutated somatically in sporadic lung cancers. However, the result indicated that germ-line mutations of the p53, RB and p16 genes are not likely to contribute to susceptibility to lung cancer. Therefore, genetic factors may not contribute greatly to the development of human lung cancer. However, it remains possible that individuals exhibit differences in lung cancer susceptibility due to genetic polymorphisms in certain genes, including tumor suppressor genes, oncogenes, DNA repair genes and genes involved in metabolizing carcinogens, as indicated by previous studies [80–84].

Recently, a possibly polygenic nature of the inherited predisposition to lung cancer has been envisaged by studies on mouse strains with an inherited susceptibility to chemical induction of lung carcinogenesis. More than 20 putative loci for genetic modifiers for lung carcinogen resistance, including the Pas [pulmonary adenoma susceptibility] and Par [pulmonary adenoma resistance] loci, have been mapped on mouse chromosomes [85,86]. Based on synteny between mouse and human chromosomes, the map positions of Pas7, Pas9, Par1 and Par3 correspond to human chromosomal regions, 18q21, 9p21, 17q11–q23 and 14q11–q24, respectively, which are often deleted in human lung cancers. Therefore, these genetic modifier loci can harbor genes, whose human homologues function as tumor suppressors in human lung carcinogenesis.

Genes inactivated by microsatellite instability

Recent studies on genetic alterations in colorectal cancers have indicated that there are two levels of genomic instability in colorectal cancer cells [8,87,88], and that they are associated with an alternative genetic pathway in colorectal carcinogenesis. One is the chromosomal instability [CIN] observed in ~85% of colorectal cancers. CINs are thought to result from abnormalities in genes involved in cell-cycle checkpoints, such as p53, ATM, hBUB1 and hBUBR1. Colorectal cancers with the CIN phenotype are usually aneuploid, and carry gross chromosomal changes including chromosomal deletions leading to inactivation of tumor suppressor genes. The remaining colorectal cancers show the other phenotype, microsatellite instability [MIN/MSI]/replication error [RER] phenotype. This phenotype, represented by changes in the number of short-tandem DNA repeats [microsatellite], is thought to result from defects in mismatch repair genes. These tumors are, in general, diploid with few chromosomal changes. Instead, both alleles of some genes, including TGFβRII, BAX and insulin-like growth factor II receptor [IGFIIR], are often inactivated by frameshift mutations of microsatellites in their coding sequences [89]. Therefore, MIN/RER tumors are considered to be developed though a genetic pathway distinct from CIN tumors, and rapid accumulation of mutations in multiple genes including the ones described above is considered to be critical for the development of tumors. Therefore, the genes containing microsatellites in the coding sequences can be tumor suppressors preferentially inactivated in MIN/RER tumors.

MIN/RER is observed not only in colorectal cancers but also in several types of human cancers, including endometrial cancers and gastric cancers. MIN/RER is also observed in lung cancer, preferentially in advanced NSCLC [90]. However, it has been shown that MIN/RER at multiple loci occurs in

Bài Viết Liên Quan

Chủ Đề