Lung cancer remains a worldwide major health challenge. Despite improvements in staging and the integrated application of surgery, radiotherapy, and chemotherapy, the 5-year survival rate for individuals with lung cancer is only about 15% (Zochbauer-Muller et al., 2002). Histologically, 80% of lung cancers are diagnosed as non-small-cell lung cancer (NSCLC), whereas the remaining 20% of cases are diagnosed as small-cell lung cancer (SCLC). On the basis of cell morphology, adenocarcinoma, and squamous cell carcinoma are the most common types of NSCLC. The current staging system for NSCLC is based upon the size and location of the primary tumor (T), involvement of regional lymph nodes (N), and presence of distant metastases (M) (Zochbauer-Muller et al., 2002). The standard treatment of patients with stage I NSCLC (T1-2, N0, M0) is resection of the primary tumor alone (no adjuvant therapy) (D'Amico et al., 2001). Survival for patients with stage I disease ranges between 40% and 70%, and the failure of treatment is due to distant recurrences (Harpole et al., 1995). This suggests that a significant number of patients with stage I NSCLC may actually be under-staged. Therefore, if correctly identified, these patients may benefit from adjuvant therapy in addition to resection, with a predictable improvement in the survival rates. Indeed, to identify patients with stage I NSCLC who might benefit from adjuvant therapy, investigators have attempted to identify factors predicting poor prognosis. These studies included analysis of performance status, histological subtype, size of the primary tumor, the degree of tumor differentiation, mitotic rate, and evidence of lymphatic or vascular invasion (Feld et al., 1984; D'Amico et al., 1999, 2000; Suzuki et al., 1999; Liu et al., 2001). However, all of these factors have failed, to date, to precisely identify a group of stage I patients who would benefit from adjuvant therapy.
Cigarette smoking remains the main risk factor for lung cancer, accounting for about 90% of the cases in men and 70% of the cases in women (Shopland, 1995). Over the past few years, our research group has investigated the possible involvement of several molecular mechanisms, such as cell-cycle and apoptosis regulators, oncogenes and tumor suppressor genes, cell adhesion molecules in the pathogenesis, and progression of lung cancer (Baldi et al., 1997; Esposito et al., 1997a,b, 2002; Caputi et al., 1998; Groeger et al., 2000, 2004).
The goal of this review is to summarize some of the most recent findings about the role of cell-cycle proteins in lung cancer pathogenesis and progression.
THE CELL-CYCLE MACHINERY AND LUNG CANCER
In the last decade, several studies have focused on the role of cell-cycle control in lung carcinogenesis. A precise regulation of the cell-cycle is a fundamental requirement for the homeostasis of a eukaryotic cell. The cell-cycle machinery is comprised of five sequential stages: G0, G1, S, G2, and M. Non-replicating cells reside in the quiescent state G0. A complicated balance of different signals, such as those produced by growth factors and growth factor inhibitor pathways, determines if the cell enters into the G1 phase of the cell-cycle. During the last decade, scientists successfully delved into the molecular machinery devoted to the fine regulation of the cell-cycle phases, identifying and characterizing several genes and gene products involved (Sherr, 1996). A key role is played by cell-cycle kinases (cdks), relatively small proteins with an apparent molecular mass between 33 and 43 kDa whose activity is regulated by their arrangement in a multimeric complex with larger proteins, called “cyclins,” after their cyclical expression and degradation during the cell-cycle. Different cdk-cyclin complexes, formed with clear-cut timing throughout the cell-cycle, together with their phosphorylation/dephosphorylation, efficiently regulate the activity of the multimeric holoenzyme.
Initiation of the cell-cycle via extracellular signals induces the transcription of several proteins including cyclin D, complexed with cdk4 (and cdk6), and leads to phosphorylation of the tumor suppressor protein pRb in the pRb–E2F complex. The phosphorylated form pRb is unable to bind to the E2F transcription complex, permitting the E2F-dependent transcription proteins to continue proliferation.
The progression through G1 is also influenced by negative regulators of the cdk–cyclin complexes. These proteins belong to two different families: the INK4 family of proteins, and the kinase inhibitory protein (CIP/KIP) family (MacLachlan et al., 1995). The INK4 family of proteins, which inhibit cdk4 and cdk6, includes the p15, p16, p18, and p19 proteins, while the KIP family includes p21, p27, and p57 and is regulated by the tumor suppressor gene p53.
Normal progression through the S phase into the premitotic G2 phase, and through the G2/M checkpoint, is also regulated by specific cdk–cyclin complexes, such as the cdk2–cyclinA and cdk2–cyclinB dimers.
The p53 protein has been termed the guardian of the genome due to its role in initiating growth arrest or apoptosis during cellular proliferation by responding to the presence of damaged DNA within the cell. The p53 tumor suppressor gene is involved in cell-cycle checkpoints by virtue of its action as a transcription factor for several cell-cycle regulatory proteins, including the p21 gene (Kirsch and Kastan, 1998). On the other hand, proliferating cell nuclear antigen (PCNA) is involved in activation of DNA polymerase δ, a function required for DNA replication and repair (Bravo et al., 1987; Prelich et al., 1987). Moreover, the p53 to p21 pathway also inhibits DNA replication through p21's interaction with PCNA, without affecting PCNA' s DNA repair abilities (Li et al., 1994; Waga et al., 1994). The tumor suppressor protein p53 also regulates progression through the G1 checkpoint of the cell-cycle. In particular, p53 is activated in response to DNA damage and serves to arrest cell-cycle progression in G1 and hence allow time for DNA repair. It is recognized that p53 is a point of convergence of a complex network of signaling pathways that regulate its level in the cell. In turn, p53 binds to specific DNA sequences and transactivates a group of target genes (including the cell-cycle inhibitor p21Waf1/Cip1), thereby inhibiting cell proliferation and promoting apoptosis (Agarwal et al., 1998).
The retinoblastoma gene family consists of three members, the product of the retinoblastoma gene (pRb), which is one of the most studied tumor suppressor genes, and two related proteins, pRb2/p130 and p107, which have been shown to be structurally and functionally similar to pRb (Paggi et al., 1996). Sequence analysis of these two proteins show that they share large regions of homology with pRb, especially in two discontinuous domains which make up the “pocket region” (Ewen et al., 1991; Mayol et al., 1993). The RB family members are characterized by a peculiar steric conformation, the “pocket region,” which is responsible for most of the functional interactions that characterize the activity of these proteins in cell-cycle homeostasis.
Both pRb2/p130 and p107, like pRb, display growth suppressive properties, although the types of growth arrest mediated by the three pocket proteins are not identical. This suggests that, although the different members of the retinoblastoma gene family may complement each other, they are not fully redundant functionally (Zhu et al., 1993; Claudio et al., 1994). The RB pocket proteins play a critical role in G1/S progression, at least in part, through binding and inactivation of factors (e.g., E2Fs) that promote transcription of genes required for DNA replication (Lam and La Thangue, 1994). Although pRb2/p130 and p107, similar to pRb, interact with members of the E2F transcription family and have similar functional consequences, each pocket protein has a different temporal profile of interaction with different E2F/DP1 complexes. The binding of pRb2/p130 to these complexes is detected predominantly during G0 (Baldi et al., 1997; Caputi et al., 1998; Groeger et al., 2000; Esposito et al., 2002), while that of p107 is detected during the G1 and S phases (Shirodkar et al., 1992; Cobrinik et al., 1993; Zhu et al., 1993; Claudio et al., 1994; Hijmans et al., 1995; Jiang et al., 1995; Vairo et al., 1995).
Thus, it is possible to propose a simple model in which the three members of the retinoblastoma gene family bind to and modulate the activity of the E2F/DP complexes, as well as other transcription factors. In this model the binding is regulated by different upstream signals such as cyclin/cdk complexes or viral oncoproteins. The flexibility of this pathway could explain the distinct activities of the three pocket proteins in the regulation of cellular division and cellular differentiation. Active (underphosphorylated) pRb can be inactivated and induced to release transcription factors when it is hyperphosphorylated (in mid-late G1) by cyclin/cdk4,6 complexes. An example of the cell-cycle pathway is represented in Figure 1.
Cellular division is an ordered, regulated process involving multiple checkpoints assessing DNA integrity, cell size, and growth signals. Lung cancer is characterized by sequential accumulation of genetic and morphological changes leading to alteration in evasion of apoptosis, DNA repair and stability, tissue invasion, metastases, and angiogenesis. The major molecular abnormalities in the pathogenesis of lung cancer are summarized in Table 1. Several are key genes and pathways involved in the pathogenesis of lung cancer. Oncogenes contributing to the pathogenesis of lung cancer include mutated K-ras, c-myc, and overexpressed EGFR, BCL2, and cyclinD1 (Fong et al., 2003). Telomerase RNA and the catalytic component are expressed in almost all lung cancers providing a mechanism for cellular immortality. Loss of the function of different tumor suppressor genes (FHIT, RASFF1A, APC, p53, RB, and p16), as well as aberrant methylation resulting in loss of gene expression (APC, CDH13, DAPK, FHIT, MGMT, p16, RARβ, RASSF1A, SEMA3B and TIMP-3), are implicated in the pathogenesis of a number of lung cancers. Mitochondrial DNA, also undergoes mutations, in lung cancer, but the functional significance remain to be evaluated. The cell regulatory pathways that are deregulated in lung cancer include wnt/APC, EGFR/RAS, PP2a, and telomerase pathways, several genes involved in various DNA repair pathways, angiogenic, autocrine/paracrine growth regulatory, and apoptosis pathways. Apoptosis or programmed cell death, is an efficient cellular process in normal development. Fundamentally, there are two sub-pathways in apoptosis: the receptor-mediated pathway and the mitochondrial pathway (Fig. 2). The receptor-mediated pathway is due to activation of death receptors and growth factor receptors that activate the initiator caspases, the central machinery regulating apoptosis. The mitochondrial pathway is composed of members of the Bcl-2 family of proteins. Different studies show that genes expressing pro-apoptotic proteins are downregulated (APAF1, BAK1, CASP9, CASP10, CDKN2D, CYCS, NFKB1A, and PTEN) (Gazzeri et al., 1998; Vonlanthen et al., 1998; Milligan and Nopajaroonsri, 2001; Shivapurkar et al., 2002; Swisher et al., 2003; Singhal et al., 2003b), while genes expressing anti-apoptotic proteins are upregulated (AKT1, BCL2L1, CASP3, MAP3K14, MDM2, and PDPK1) (Reeve et al., 1996; Milligan and Nopajaroonsri, 2001; Eymin et al., 2002; Kandasamy and Srivastava, 2002; Singhal et al., 2003b). These studies are in accordance with the fact that lung cancer cells are more resistant to apoptosis than normal cells
Table 1. Main molecular abnormalities in the pathogenesis of lung cancer
Aberrant promoter methylation
Activation of proto-oncogenes, growth factors and receptors
Expression of telomerase activity
Expression of cellular immortality
Loss of function of tumor suppressor genes
Loss of components of apoptosis pathways
Potential loss of DNA repair mechanisms
Amongst the many pathways altered in lung cancer malignancy, the most criticals involve disruption of the normal cell-cycle regulation. The deregulation of growth enhancing and growth inhibiting signals can contribute to the alteration of cell-cycle control. Lung cancer cells develop the ability to bypass different checkpoints, especially at G1/S and G2/M. An incorrect and/or missing G1/S arrest of cell-cycle causes uncontrolled cellular proliferation.
The pRb protein is inactivated in more than 90% of SCLCs (Salgia and Skarin, 1998) as a result of different mechanisms including point mutations and abnormal mRNA expression (Gouyer et al., 1998). Changes in the other pocket proteins (p107 and p130) have been detected in only a few cases (Helin et al., 1997). In contrast to SCLC, the majority of NSCLC cases exhibit abnormalities in the upstream regulators of the pRb pathway, including inactivation of p16 (Kashiwabara et al., 1998; Tanaka et al., 1998) through different mechanisms (Gazzeri et al., 1998), reduced levels of p27Kip (Kawana et al., 1998; Yatabe et al., 1998), and enhanced expression of cyclin D1 (Marchetti et al., 1998). It is likely that inactivation of cdk4,6 inhibitors (p16) and overexpression of cyclin D1 bypass the pRb checkpoint, allowing progress through G1 into DNA synthesis (Driscoll et al., 1997; Shapiro et al., 1998). Immunohistochemical analyses of the RB gene product expression have been performed in malignant tissues from this human neoplasm (Higashiyama et al., 1994). Because the three members of the retinoblastoma protein family exhibit different growth suppressive properties, suggesting that they are not fully functionally redundant, our research group investigated their pattern of expression in a large group of specimens of lung cancer, using an immunohistochemical approach. These Rb family members displayed distinctive patterns when compared and contrasted with the different parameters. The highest percentage of undetectable levels in all the specimens examined and the tightest inverse correlation (P-value) with the histological grading and with PCNA expression in the most aggressive tumor types were found for pRb2/p130, which may suggest an important role for this protein in the pathogenesis and progression of lung cancer (Baldi et al., 1996). In fact, our immunohistochemical studies examining the expression profiles of RB family members in 235 specimens of lung cancer and the expression pattern of pRb2/p10 in 158 specimens of lung cancer have suggested an independent role of pRb2/p130 expression in the development and/or progression of human lung cancer (Baldi et al., 1996, 1997). Other studies of ours demonstrated a statistically significant inverse relationship between the histological grading and the expression of RB family members in squamous cell carcinomas (Minimo et al., 1999). Recently, we have demonstrated that Rb2/p130 gene exhibits oncosuppressor properties in vitro and in vivo in lung adenocarcinoma and modulates angiogenesis by inhibiting vascular endothelial growth factor expression (Claudio et al., 2001). Moreover, we were able to identify Rb2/p130 point mutations in primary lung cancer (Claudio et al., 2000) and to study the expression of pRb2/p130 target genes in a lung cancer cell line by microarray analysis (Russo et al., 2003). Today, several lung cancer microarray studies are available (Bhattacharjee et al., 2001; Garber et al., 2001; Goodwin et al., 2001; Nacht et al., 2001; Bangur et al., 2002; Beer et al., 2002; Chen et al., 2002, 2003; Gordon et al., 2002; Heighway et al., 2002; Miura et al., 2002; Wigle et al., 2002; Kikuchi et al., 2003; Powell et al., 2003). By using a computerized approach, it is possible to analyze in detail the presence of differential gene expression profiles of RB, Rb2/p130, and p107 in lung cancer from these microarray data (Table 2). All together, these studies strongly suggest an important role(s) of RB family members in lung cancer progression. Different studies showed that three genes downstream from the pRb pathway that support cell growth, E2F, DP-1, and HDAC, are overexpressed in lung adenocarcinomas (Chang and Szabo, 2002; Gorgoulis et al., 2002; Singhal et al., 2003a). The expression of these Rb-dependent genes is increased probably because upregulation of cyclin D1 and cyclin D2, key regulators of pRb phosphorylation, and downregulation of cyclin-dependent kinase inhibitors p15, p21, p19, and p57 (Kawamata et al., 1995; Rusin et al., 1996; Gazzeri et al., 1998; Keum et al., 1999; Jin et al., 2001; Shoji et al., 2002; Singhal et al., 2003a). The TGF-β/SMAD family pathway is an important regulator of several cyclin-dependent kinase inhibitors. Different studies reported that SMAD4, one of the proteins mediating the intracellular signaling of the TGF-β superfamily (Yanagisawa et al., 2000; Singhal et al., 2003a), is downregulated in lung adenocarcinomas.
Table 2. RB, Rb2/p130 and p107 differential expressions in detail, analyzed from microarray data published from 2001 to 2003 available in Oncomine database
Adjusted P-values were calculated using the Bonferroni correction, to account for multiple hypothesis testing, ensuring a conservative estimate of significance. Border-line P-values are shown in bold.
Tissue type: others
Tissue type: carcinoid
Tissue type: others
Tissue type: squamous cell lung carcinoma
Squamous cell carcinoma
Cluster group: 2, 3
Cluster group: 1
Squamous cell carcinoma
The progression through S phase is principally regulated by the cyclin A/cdk2 complex. The overexpression of cyclin A has been consistently shown to be a negative prognostic indicator of lung cancer (Volm et al., 1997; Dobashi et al., 1998). Recently, overexpression of cyclin F, which is involved in significant increase in the G2 population, was found in lung adenocarcinomas (Singhal et al., 2003a). The cell division cycle phosphatase gene (cdc25A), one of the rate-limiting mechanisms for G1 progression into S phase, is frequently overexpressed in NSCLC (Mailand et al., 2000). The gene cdk10 is essential for cellular proliferation, and its effect is also exerted in the G2–M transition. Only recently, this gene was found overexpressed in lung adenocarcinomas (Singhal et al., 2003a).
Cyclin B1/cdc2 is the classic M phase-promoting factor that drives entry into mitosis. Cyclin B1 and cdc2 are overexpressed in lung cancer (Soria et al., 2000; Singhal et al., 2003a). The G2 checkpoint is monitored for genomic integrity before mitosis. Failure at this checkpoint results in genomic instability, predisposing cells to neoplastic transformation.
The cascade of events that monitors DNA damage originates with the ATM, ATR, and p53 gene products. Following DNA damage, a series of events is initiated that ends with phosphorylation, hence inactivation, of cdc2. As a result, the chk2 protein is phosphorylated resulting in phosphorylation of cdc25c, which then binds 14-3-3. Once cdc25c binds 14-3-3-σ, it is exported to the cytoplasm where it sequesters and loses its ability to activate cdc2. Different studies showed that chk2 expression is reduced in NSCLC (Matsuoka et al., 2001; Miller et al., 2002).
The fundamental importance of p53 in lung cancer is highlighted by the frequency of its mutations, 80% in SCLC and 50% in NSCLC (Agarwal et al., 1998). Once p53 gene is deleted or mutated, cells become susceptible to DNA damage and dysregulated cell growth. There are numerous reports that investigated the association of p53 abnormalities with the prognosis of NSCLC patients; however, the results are discordant as to whether p53 abnormalities influence prognosis. A recent meta-analysis of 43 published articles has revealed that the negative prognostic effect of p53 alterations was highly significant in patients with adenocarcinomas but not in patients with squamous cell carcinomas.
p73, a gene structurally similar to p53, activates the promoters of several p53-responsive genes participating in cell-cycle control, apoptosis, DNA repair, and inhibits cell growth in a p53-like manner by inducing G1 cell-cycle arrest or apoptosis. Higher p73 expression levels in lung tumor tissues compared with adjacent normal tissues are associated with p53 mutations, suggesting a role of p73 in compensating for the loss of p53 function (Yokomizo et al., 1999; Kang et al., 2000). However, overexpression of wild-type p73 may also have some p53-independent functions either as an oncogene or as a tumor suppressor gene in lung carcinogenesis, suggesting an important role in the development of lung cancer.
p16 is part of the p16-cyclin D1-CDK4-RB pathway that is important to controlling the G1–S transition of the cell-cycle and functionally altered or mutated lung cancer. p16 abnormalities are frequently found in NSCLCs but are rare in SCLCs. Perhaps 30%–50% of early stage primary NSCLCs do not express p16. The p16 locus also encodes a second alternative reading frame protein, p14ARF. By binding to the MDM2-p53 complex, p14ARF prevents p53 degradation, thereby leading to p53 activation. Loss of p14 expression is more frequently found in lung tumors with neuroendocrine features. However, aberrant methylation of the p14 promoter region do not occur frequently in NSCLCs. Interestingly, lung tumours have developed distinct ways of interfering with the two different products from a single genetic locus, each of which functions in a distinct growth regulatory pathway. Moreover, the specific mutational targets differ according to lung cancer subtype.
While several of the factors involved in regulating cell-cycle control have been investigated in lung cancer, few studies have examined multiple factors in the same tumor series. Our research group recently set up a study to evaluate the expression of p53, p21, p16, and PCNA proteins in a large series of NSCLCs to assess the integrity of cell-cycle checkpoints in these tumors, to evaluate the co-expression of these proteins and, finally, to examine the relationship between these cell-cycle regulators and the clinicopathological features of NSCLCs, including their ability to predict survival in NSCLC patients. Numerous checkpoint proteins have been examined so far in lung cancer, but few studies have investigated multiple factors in the same tumors. We have analyzed the expression of four key proteins involved in cell-cycle checkpoints in a 68 well-characterized NSCLCs using immunohistochemestry (Esposito et al., 2004) (Fig. 3). All of the cell-cycle associated proteins examined were present in the nuclei of tumour cells.
When we focused on the possible correlations between clinicopathological data and expression of cell-cycle proteins, we found an inverse correlation between the status of lymph nodes and the expression of p21 (P = 0.015) and p16 (P = 0.008) proteins. These data might indicate a possible role of these two proteins in lung cancer progression. Moreover, no correlation has been identified between p16, p21, and p53 expression.
When we investigated the correlation between different protein expressions and survival using univariate analysis, we found that all the cell-cycle markers analyzed except for PCNA had a statistically significant correlation with survival (Fig. 4). This result is in agreement with numerous data published about the cell-cycle checkpoints investigated in this paper and lung cancer (Esposito et al., 1997a; Caputi et al., 1998; Groeger et al., 1999; Mitsudomi et al., 2000; Zhou et al., 2001; Kaye, 2002; Shoji et al., 2002). As expected, the lymph nodes status and clinical tumor stage were significantly correlated with survival as well.
Surprisingly, when we performed multivariate analysis, the only immunohistochemical parameter that proved to influence overall survival was p16. This result is in agreement with the proposed hypothesis that the great majority of lung cancer samples have inactivated the RB/p16 tumor suppressor pathway (Zochbauer-Muller et al., 2002). Among the clinical parameters, tumor staging was the only factor to influence survival in multivariate analysis.
Finally, we grouped the lung cancer specimens based on p21 and p16 status. Interestingly, we found that the group of lung cancer specimens that were both p21- and p16-negative displayed a significantly shorter overall survival. Numerous data from the literature suggest the existence of a functional collaboration between distinct CDK inhibitor genes (Franklin et al., 2000). Indeed, it has been recently demonstrated that cell-cycle inhibition by p16 is associated with a post-transcriptional induction of p21 and a strong inhibition of cyclin E-cdk2 kinase activity (Mitra et al., 1999). Moreover, it has been shown that members of the p21 family of proteins promote the association of D-type cyclins with CDKs by counteracting the effects of p16 molecules (Parry et al., 1999). It has been therefore proposed that functional cooperation between different cell-cycle inhibitor proteins constitutes another level of regulation in cell growth control and tumor suppression.
Taking into account the extensive functional network constituted by the cell-cycle regulator proteins, it appears evident that the knowledge of the expression levels of these factors, and their co-regulation, may be important in predicting patient clinical response to therapy. Nevertheless, targeting multiple checkpoint proteins may represent a good therapeutic strategy for the development of new molecular treatments for lung cancer. These data strongly suggest further studies to be performed in order to investigate the simultaneous expression of numerous cell-cycle regulators in lung cancer.
This study was supported by NIH grants and the Sbarro Health Research Organization to A.G. Giuseppe Russo acknowledges the Ph.D. program: “Diagnostic, quantitative, and molecular pathology” of the University of Siena, Italy.