Several lines of evidence indicate that tumorigenesis in humans is a multi-step process, formally analogous to Darwinian evolution, in which a succession of genetic and epigenetic changes, each conferring a different type of growth advantage, leads to the progressive conversion of normal cells into cancer cells (Nowell, 1976). The accumulation of genetic changes liberates neoplastic cells from the homeostatic mechanisms that govern normal cell proliferation. Observations of human cancers and animal models implicate a limited number of molecular pathways, the disruption of which contributes to most cancers. In humans, at least four to six distinct somatic mutations are required to reach this state (Renan, 1993; Kinzler et al., 1996).
Several properties are shared by most of human tumours: self-sufficiency in growth signals, insensitivity to growth-inhibitory signals, evasion of programmed cell death (apoptosis), limitless replicative potential, sustained angiogenesis, and tissue invasion and metastasis (Hanahan and Weinberg, 2000). Each of these novel capabilities acquired during tumour development represents the successful breaking of an anti-cancer defense mechanism applied by cells and tissues.
Molecular studies have identified three groups of genes, which are frequently deregulated in cancer:
Development of gene-expression profiling as well as advances in tumour diagnosis revealed, however, a striking conceptual inconsistency in the prevailing multi-step model of tumour progression (Bernards and Weinberg, 2002). Gene-expression pattern of metastatic tumour cells is often strikingly similar to that of the cells confined to the primary tumour mass from which they were derived. Equally relevant are other studies in which the gene-expression profiles of the dominant populations of breast-cancer cells within a primary tumour mass have been used to predict, with 90% accuracy, whether the tumour will remain localized or whether the patient will experience metastases and disease relapse.
Based on these findings, it is suggested that a subset of the mutant alleles acquired by incipient tumour cells early in tumorigenesis confer not only the selected replicative advantage, but also, later in tumorigenesis, the tendency to metastasize. This tendency will become manifest only much later in tumour progression, in the context of yet other mutations that have struck the genomes of descendant cells.
This reasoning has important implications. First, genes and genetic changes specifically and exclusively involved in orchestrating the process of metastasis do not exist. Instead, it is the particular combination of genes that enables cells to create primary tumour mass that also empowers them to become metastatic. Second, because important components of the genotype of metastasis are already implanted in cells relatively early in tumorigenesis, even relatively small primary tumour cell populations may already have the ability to dispatch metastatic cells to distant sites in the body.
Moreover, several independent lines of evidence seem to support the idea that tumorigenesis is governed not only by the tumour cells per se, but also by the microenvironment (Chang and Werb, 2001). For example, tumour-associated fibroblasts can direct tumour progression (Olumi et al., 1999), and endothelial cells foster tumour angiogenesis (Carmeliet and Jain, 2000). Inflammatory cells might also promote tumour development, as shown by studies on skin tumorigenesis in K14-HPV16 transgenic mice that lack mast cells or metalloproteinase-9 (Coussens et al., 1999; Coussens et al., 2000).
Calcium (Ca2+) functions as a universal second messenger that plays a regulatory role in a great variety of cellular processes such as memory and transmission of nerve impulses, muscle contraction, secretion, cell motility and volume regulation, cell growth and differentiation, gene expression, cross-talk between different enzyme systems, apoptosis, and necrosis (Berridge et al., 2000). The Ca2+ signalling networks are composed of many molecular components including the large family of Ca2+-binding proteins characterized by the EF-hand structural motif (Kawasaki et al., 1998). Certain members, notably calbindin D28k and parvalbumin, serve as cytosolic Ca2+ buffers, whereas others, such as calmodulin, troponin C, and the S100 proteins, are Ca2+-dependent regulatory proteins.
S100 proteins represent the largest subgroup within the EF-hand protein family. They have received increasing attention in recent years due to their association with various human pathologies including cancer, neurodegenerative disorders, inflammation, and cardiomyopathy (Heizmann et al., 2002). Moreover, they have been of value in the diagnosis of these diseases. Twenty-two members have been identified so far. Unlike the ubiquitous calmodulin, most of them show cell- and tissue-specific expression.
S100 proteins are small (10 to 12 kDa) acidic proteins that form homo- and heterodimers. They are characterized by a pair of helix-loop-helix (the EF-hand) motifs connected by a central hinge region. The two EF-hand structural motifs display different affinities for calcium. The C-terminal EF-hand contains the canonical Ca2+-binding loop consisting of 12 amino acids. The N-terminal EF-hand consists of 14 amino acids and is specific for S100 proteins. Upon Ca2+ binding S100 proteins undergo a conformational change required for target recognition and binding. Generally, the dimeric S100 proteins bind four Ca2+ per dimer. Besides Ca2+,a number of S100 proteins bind Zn2+ with a wide range of affinities. For S100B, S100A5, and S100A13 Cu2+ binding was reported (Nishikawa et al., 1997; Schäfer et al., 2000; Mandinova et al., 2003).
Another distinguishing feature of S100 proteins is that individual members are localized within specific cellular compartments from which some of them re-locate upon Ca2+ or Zn2+ activation (Davey et al., 2001). A signalling event is thus transduced in a temporal and spatial manner by specific targeting for each S100 protein. Furthermore, some S100 proteins are secreted from cells acting in a cytokine-like manner. The individual members are believed to utilize distinct pathways (ER-Golgi route, tubulin- or actin-dependent) for their translocation/ secretion into the extracellular space (Hsieh et al., 2002). S100B and S100A12 specifically bind to the surface receptor RAGE (receptor for advanced glycation endproduct) – a multiligand member of the immunoglobulin superfamily (Schmidt et al., 2000). The extracellular levels of S100B thereby play a crucial role in that nanomolar concentrations of S100B have trophic effects on cells whereas pathological levels (as found in Alzheimer’s patients) induce apoptosis (Huttunen et al., 2000).
Characteristic of the S100 protein family is that most S100 genes form a cluster on human chromosome 1q21, a region frequently involved in chromosomal rearrangements and deletions in human cancers (Schäfer et al., 1995; Weterman et al., 1996; Gendler et al., 1990).
S100 proteins were recently found to be reliable diagnostic markers for hypoxic brain damage and for monitoring the outcome after cardiac arrest (S100B; Böttiger et al., 2001), acute myocardial infarction (S100A1; Kiewitz et al., 2000), amyotrophic lateral sclerosis (S100A6; Hoyaux et al., 2000), for the classification of astrocytomas and glioblastomas (Camby et al., 2000; Camby et al., 1999), melanoma metastasis formation (S100B; Krähn et al., 2001), as prognostic indicators for gastric cancer (S100A4; Yonemura et al., 2000), laryngeal (S100A2; Lauriola et al., 2000) and esophageal squamous cell carcinomas (S100A4; Ninomiya et al., 2001), and for breast cancer (Platt-Higgins et al., 2001).
The structural organization of S100 genes is highly conserved both within an organism and in different species (Heizmann et al., 2002). A typical S100 gene consists of three exons whereby the first exon carries exclusively 5’ untranslated sequences. The second exon contains the ATG translation start codon and codes for the N-terminal EF-hand, and the third exon encodes the carboxy-terminal canonical EF-hand. Presently, 16 S100 genes are found in a tight gene cluster on human chromosome 1q21 within a genomic region of 260 kb ( Fig. 1 ).
|Fig. 1 The S100 gene cluster on human chromosome 1q21. Genes located in the cluster region are indicated as well as two commonly used genomic markers (D1S1664 and D1S2346). p and q indicate the short and the long arm of the chromosome, respectively.|
Four additional S100 genes are found on other human chromosomes: Xp22 (calbindin-D9K), 21q22 (S100B), 4p16 (S100P) and 5 (S100Z). Within the gene cluster, epidermal differentiation genes as well as a gene of unknown function called NICE2 interrupt the S100 genes. Furthermore, there are three proteins encoded in 1q21 that carry in the N-terminus an S100-like domain, namely trichohyalin, profilaggrin, and C1 or f10. The small distances between the genes on the chromosome and the phylogenetic tree indicate that S100 genes most likely originate from late gene duplication events. It is interesting that the clustered organization of the human genes seems to be evolutionarily conserved, at least in the mouse. In other species, S100 genes are less well characterized.
S100 proteins generally are involved in a large number of cellular activities such as signal transduction, cell differentiation, regulation of cell motility, transcription and cell cycle progression. Such activities can be expected since S100 proteins are thought to modulate the activity of target proteins in a Ca2+- (and possibly also in a Zn2+- and Cu2+-) dependent manner, thereby transferring the signal from the second messenger. Therefore, understanding the biological function of S100 proteins will crucially depend on the identification of their target proteins. During the last decade, a large number of such possible interactions have been described involving enzymes, cytoskeletal elements as well as transcription factors.
Apart from these intracellular functions, some S100 proteins like S100A8/S100A9, S100B, S100A4 and probably others can be secreted from cells, as noted above, and exibit cytokine-like extracellular functions. These include chemotactic activities related to inflammation (S100A8/A9 and A12), neurotrophic activities (S100B), and angiogenic effects (S100A4 and S100A13). In all cases, the mechanisms of secretion as well as the nature of high affinity surface receptors remain largely unknown. One candidate receptor to mediate at least some of the described extracellular functions is the RAGE, which is activated upon binding of S100A12, S100A13, S100P, and S100B (Hofmann et al., 1999; Huttunen et al., 2000; Arumugam et al., 2004; Hsieh et al., 2004). It is currently not known whether RAGE is a universal S100 receptor.
Generation of animal models has been initiated to study the physiological significance of S100 proteins. Ectopic overexpression in the mouse has been described for S100B and S100A4. In the case of S100B, enhanced expression in the brain led to hyperactivity associated with an impairment of hippocampal function (Gerlai and Roder, 1995). In contrast to this mild phenotype, expression of S100A4 in oncogene-bearing transgenic mice can induce metastasis of mammary tumours, suggesting that S100A4 plays an important role in the acquisition of the metastatic phenotype (Ambartsumian et al., 1996; Davies et al., 1996).
Inactivation through homologous recombination in mouse embryonic stem cells has been demonstrated for S100B, S100A8, and S100A1. While inactivation of S100B has no obvious consequences for life (Xiong et al., 2000), S100A8 null mice die via early resorption of the mouse embryo (Passey et al., 1999), a result that suggests a role for this protein in prevention of maternal rejection of the implanting embryo. S100A1 null mice have significantly reduced responses to acute and chronic hemodynamic stress that are associated with reduced cardiac calcium sensitivity (Du et al., 2002).
Since S100 proteins can form homo- and heterodimers and usually more than one S100 protein is expressed in a given cell type, functional redundancy or compensatory mechanisms might explain the lack of phenotype observed in some animal models.
Several S100 proteins are closely associated with human diseases ( Table 1 ).
A few S100 proteins have been proposed to play important roles in tumour progression and suppression and are recognized as potential tumour markers. The association of S100 proteins with cancer development originates in the finding that the evolutionary conserved gene cluster on human chromosome 1q21 is implicated in gene rearrangements during tumour development. As discussed below, up-regulation in tumours has been reported for S100A1, S100A4, S100A6, S100A7, S100A8, S100A9, S100A10, S100A16, S100B, and S100P, whereas S100A2 and S100A11 have been postulated to be tumour suppressor genes.
S100A1 and S100B are overexpressed in human cancers and have been suggested to play a role in the hyperactivation of Ndr kinase in melanomas (Millward et al., 1998). Inhibition of p53 activity by S100B could be one mechanism, whereby overexpressed S100B is involved in neoplastic transformation. Blood levels of S100B are used to monitor malignant melanomas (Krähn et al., 2001).
S100A4 has been implicated in invasion and metastasis. The prognostic significance of its selective expression in various cancers has been exploited. In gastric cancer the opposite expression of S100A4 in relation to a tumour suppressor E-cadherin was found to be a powerful aid in histological typing and in evaluating the metastatic potential/prognosis of patients with this type of cancer (Yonemura et al., 2000). It was also demonstrated that extracellular S100A4 could act as an angiogenic factor and might induce tumour progression via an extracellular route stimulating angiogenesis (Kriajevska et al., 2002).
S100A5 has been postulated to be a marker of recurrence in WHO grade I meningiomas (Hancq et al., 2004).
S100A6 has been found overexpressed in human pancreatic adenocarcinomas (Logsdon et al., 2003) as well as in intrahepatic tumours, where it was activated by TNF-α and NFκB (Joo et al., 2003).
S100A7 (psoriasin) expression has been associated with psoriasiform hyperplasia, tumour progression in breast cancer and a worse prognosis in estrogen receptor-negative invasive ductal breast carcinomas (Emberley et al., 2003) as well as with gastric tumours (El-Rifai et al., 2002). In addition, S100A7 is a potential tumour marker for non-invasive follow-up of patients with urinary bladder squamous cell carcinoma (Ostergaard et al., 1999).
The two calgranulins – S100A8 and S100A9 – are differentially expressed at sites of acute and chronic inflammation. Recent reports, however, indicate that they are also overexpressed during skin carcinogenesis (Gebhardt et al., 2002), in poorly differentiated lung adenocarcinomas (Arai et al., 2001), gastric tumours (El-Rifai et al., 2002), and at the invasive margin of colorectal carcinomas (Stulik et al., 1999).
S100P expression has been noted in various cancer cell lines. It is associated with cellular immortalization in breast cancer cell lines (Guerreiro et al., 2000). In colon cancer cell lines, its expression is elevated in doxorubicin-resistant cells (Bertram et al., 1998). S100P is expressed in prostate cancer, where its expression is androgen-sensitive (Averboukh et al., 1996) and in pancreatic adenocarcinoma, where its expression has been localized to the neoplastic epithelium of pancreas (Logsdon et al., 2003). Furthermore, it was found that S100P expression correlates with decreased survival in patients with lung cancer (Beer et al., 2002).
S100A10 is an annexin 2 protein ligand and a key plasminogen receptor of the extracellular cell surface that is overexpressed in esophageal squamous cell carcinomas (Zhi et al., 2003), renal cell carcinomas (Teratani et al., 2002), and gastric tumours (Rifai et al., 2002). It was shown that S100A10 stimulates the conversion of plasminogen to plasmin on the tumour cell surface thereby contributing to the increased invasiveness of tumour cells (Zhang et al., 2004).
The recently identified S100A16 is up-regulated at the transcriptional level in tumours of the bladder, lung, thyroid gland, pancreas, and ovary (Marenholz and Heizmann, 2004).
By contrast, S100A2 is markedly down-regulated in breast tumour biopsies and can be re-expressed in mammary carcinoma cells by 5-azadeoxycytidine treatment (Wicki et al., 1997). A prognostic significance of S100A2 in laryngeal squamous-cell carcinoma has also been found allowing discrimination of high and low risk patients in the lymph-node negative subgroup to provide better therapy (Lauriola et al., 2000). Using DNA array technology, S100A2 was identified as differentially expressed in normal versus tumorigenic human bronchial epithelial cells (Feng et al., 2001).
A role for S100A11 as a tumour suppressor protein was postulated. This was based on its down-regulation in immortalized versus normal cells (Sakaguchi et al., 2000) and the observation that microinjection of an anti-S100A11 antibody into normal confluent quiescent cells induced DNA synthesis (Sakaguchi et al., 2003).
The primary aim of the study was to identify novel tumour-associated genes with potential application in the detection or treatment of cancer. The starting point was a collection of partial cDNA clones of yet unknown genes from a suppression subtractive hybridization (SSH) cDNA library preferentially representing genes that were down-regulated in the small cell lung carcinoma cell line as compared to normal human bronchial epithelial cells.
An unknown human transcript was selected for further characterization in view of its differential expression in tumour cell lines and in primary tumours, as well as its homology to a protein family known to be associated with tumorigenesis.
I aimed to elucidate the rationale for the differential expression of the genein tumours as well as to evaluate its potential clinical relevance. A further intention was to shed light on the mechanism of regulation of the gene in order to determine its association with malignant transformation.
|XDiML DTD Version 4.0||Zertifizierter Dokumentenserver|
der Humboldt-Universität zu Berlin