Abstract

Metastasis is the most lethal attribute of a cancer. There is a critical need for markers that will distinguish accurately those histologic lesions and disseminated cells with a high probability of causing clinically important metastatic disease from those that will remain indolent. While the development of new diagnostic markers of metastasis was the initial motivation for many studies, the biologic approach used to identify metastasis-suppressor genes has provided surprising insights into the in vivo mechanisms regulating the formation of metastases. This review and perspective describes the evolving view of the mechanisms that regulate metastasis and the importance of metastasis-suppressor genes in this process. The known metastasis-suppressor proteins or genes and the microcell-mediated chromosomal transfer strategy used to identify many of them are reviewed. New evidence for the role of these metastasis-suppressor proteins or genes in regulating the growth of disseminated cancer cells at the secondary site, the potential for the identification of novel therapeutic targets, and the multidisciplinary approach needed to translate this information into clinical tools for the treatment of metastatic disease are discussed.

Clinical Problem: Predicting Metastatic Propensity

Our ability to detect and successfully treat localized cancers has improved appreciably in recent years. However, metastatic disease presents a continuing therapeutic challenge and is the most common cause of cancer-related death. Thus, there is an emphasis on the diagnosis of cancers at an early stage, when they are localized and most likely to be curable. Although screening for early-stage disease is logical, its utility is limited by the inability of conventional diagnostic and histologic parameters to predict accurately the true extent and prognosis of a substantial proportion of clinically localized cancers (13). This limitation is due, in part, to the inherent limitations and subjectivity of current grading and staging systems (4,5).

The incidence of disease recurrence in surgical patients treated for prostatic and breast cancers illustrates this problem particularly well. Although we have a wealth of clinical and biologic information on these diseases, a large percentage of apparently resectable and theoretically curable lesions is found to be more advanced at the time of resection than envisaged, resulting in a substantial failure rate after attempted curative surgery (68). In studies of prostate cancer patients (911), even when patient selection excludes men with factors predicting poor prognosis (e.g., poorly differentiated histology, high prostate-specific antigen [PSA] levels, and clinical suspicion of local invasion), the relapse rate after radical retroperitoneal prostatectomy has approached 20%–30%. Similarly, one third of surgical patients with lymph node-negative breast cancer will develop metastases, while the other two thirds, despite receiving no chemotherapy, will not (12). Even in patients with small tumors and tumor-negative lymph nodes (T1N0), there is a 15%–25% likelihood of distant metastases (8).

Since the current staging systems for breast and prostate cancers do not accurately identify those patients curable by regional treatment alone, the evaluation of additional parameters associated with the metastatic phenotype will be very important for the differentiation of patients curable by surgery alone from those requiring systemic therapy. For instance, men at high risk for relapse of prostate cancer can be identified [e.g., serum PSA level >10 ng/mL; clinical stage T1 or T2 with >50% of tissue at Gleason grade 4 (3,4) on biopsy or clinical stage T3 prostate cancer] and would be immediate candidates for adjuvant antimetastatic therapies if they existed (10,11,1316). Likewise, breast cancer patients with particularly poor prognoses can be identified by the detection of high microvessel counts concurrent with low expression of Nm23 and/or E-cadherin in the primary tumor (1217). In fact, these parameters are better prognostic biomarkers than the conventional analysis of tumor size and grade. The information obtained from the simultaneous evaluation of biomarkers such as these have the potential to lead to a reduction in the morbidity among those patients not requiring chemotherapy and possibly identify those patients requiring more aggressive therapies than indicated by current methods.

Overall, it is clear that there is a critical need for markers that will distinguish accurately those histologic lesions and disseminated cells that have a high probability of causing clinically important metastatic disease from those that will remain indolent (5,15). Concerns have been raised that “metastasis” has often occurred by the time of diagnosis of the primary tumor, the implication being that it is then too late for antimetastatic therapy to be of use (18). However, the mere spread of cancer cells into the vasculature or to a secondary site does not constitute metastasis. Development of clinically significant metastases requires that a cancer cell complete a series of well-defined steps, generally referred to as the metastatic cascade (13). If a cell fails to complete any one of these steps, overt metastases will not develop (1315).

The clinical importance of disseminated cancer cells (detected by sensitive methods such as reverse transcription–polymerase chain reaction [RT–PCR]) has become an issue of considerable interest (19). Several such studies (16,20) have reported the detection of tumor-derived cells in the circulation and bone marrow without future development of disease. Other reports have demonstrated an increased risk of disease recurrence in patients with bone marrow micrometastases both for prostate cancer [by the detection of messenger RNA transcripts for PSA (21)] and breast cancer [by the detection of cytokeratin-positive cells (22)]. Even in these later studies, however, the majority of patients with tumor cell-positive bone marrow samples did not actually develop recurrent disease, although the proportion with recurrence could increase given extended time for patient follow-up. The discrepancy regarding the clinical importance of disseminated cells is likely due to differences in the experimental approaches used to identify cells (i.e., RT–PCR versus immunohistochemical detection).

Tumor cell growth at the site of metastasis is an important clinical target, since cells must survive and proliferate to grow into overt, macroscopic metastases. The first step toward developing effective therapies to inhibit such growth is to identify the genes/proteins that regulate metastatic colonization. To this end, a growing number of laboratories are focusing translational research efforts on the discovery of genes that specifically regulate the metastatic ability of cancer cells. For example, several metastasis-promoting genes—including WDNM-1, WDNM-2, MMP11 (stromelysin-3), MTA1, and ERBB2—have been identified in association with the development of metastatic breast cancer (2327). One must keep in mind, however, that it takes the coordinated expression of many genes to allow the development of metastases (28,29). Thus, while it is relatively easy to demonstrate an association for a given gene with metastatic ability, it is difficult to prove that a particular gene is essential. On the other hand, it only takes one gene to block metastasis, since inability to complete any step of the metastatic cascade renders a cell nonmetastatic. Metastasis-suppressor genes suppress the formation of spontaneous, macroscopic metastases without affecting the growth rate of the primary tumor. It has now been more than 10 years since the discovery of the first metastasis-suppressor gene nm23 (NME1) (30). Since then, both in vitro and in vivo (e.g., animal) studies (15,3032) have documented the important role of the loss of metastasis-suppressor gene function in the acquisition of metastatic ability.

While the initial motivation for these studies was the development of new diagnostic markers of metastasis, the biologic approach used to identify metastasis-suppressor genes has provided surprising insights into the in vivo mechanisms regulating the formation of metastases. We anticipate that identifying the molecular pathways that regulate metastatic colonization and growth control at the secondary site will provide additional, potentially novel therapeutic targets for the treatment of metastatic disease. The purpose of this review is 1) to present the evolving view of the mechanisms that regulate metastasis, 2) to describe the functional strategy used to identify metastasis-suppressor genes and discuss important principles learned from these studies, 3) to document the known metastasis-suppressor genes and report new evidence that supports their role in the regulation of growth control at the secondary site, and 4) to discuss the multidisciplinary approach needed to translate metastasis-suppressor genes into clinical tools.

Regulation of Metastatic Propensity—Evolving Paradigms

Metastasis is defined as the formation of progressively growing secondary tumor foci at sites discontinuous from the primary lesion (15). This process is illustrated by the spontaneous hematogenous metastasis of tumor cells to the lung (Fig. 1, A). The formation of a primary tumor requires a cadre of molecular and cellular alterations that enable a cell(s) to circumvent normal growth control mechanisms as well as to manipulate its local environment (14). These changes include the development of a blood supply once the focus of transformed cells grows beyond a size that can be nourished by nutrient or metabolite diffusion (33,34). Tumor progression and the acquisition of metastatic competence require additional changes in gene expression (e.g., protein-degrading enzymes and adhesion molecules) that culminate in a malignant phenotype. After invasion into adjacent tissues, tumor cells disseminate via blood vasculature or lymphatics and travel individually or as emboli made up of tumor cells or tumor and host cells. At the secondary site, cells or emboli arrest either because of their physical size or by binding to specific molecules in particular organs or tissues (15,35). For disseminated cells to grow into overt metastases, they must survive and proliferate in the vasculature or in the surrounding tissue after extravasation. The formation of clinically important metastases depends on the completion of every step of this cascade, the last of which is metastatic colonization (Fig. 1) (14).

The presence of isolated cells at a secondary site represents a risk to the patient. Cells getting to the secondary site certainly have the potential to colonize; therefore, it is crucial not to ignore the presence of neoplastic cells anywhere. On the other hand, as we will show, the mere presence of cells does not necessarily mean that metastatic colonization will occur. The challenge is to determine how to discriminate between disseminated cells that will form overt metastases from those that will not.

Cancer metastasis, both clinically and experimentally, is known to be inefficient (36). In experimental models, fewer than 0.1% of cells injected into the circulation go on to form secondary tumors (15,37). While many factors contribute to the observed inefficiency of metastasis formation, those considered to be most important include the low survival rates of cells in the circulation and the low percentage of cells that successfully escape from the vasculature into surrounding tissues (18). At this time, there is some question as to whether postextravasational growth control or growth within a vessel are more predominant (38). This process has, for the most part, been studied using assays in which the number and kind of cells injected are known and the numbers and sizes of metastases formed are assessed (18,39). The processes that are responsible for metastatic efficiency in vivo remain hidden; thus, mechanistic paradigms have largely been based on logical inference rather than on direct observation. The development of new technologies has enabled researchers to test the possibility that cancer cell dissemination, arrest (nonspecific arrest and/or specific adhesion events), and growth at the secondary site are critical determinants in metastasis formation.

The ability to observe single cells in vivo has been greatly enhanced by improvements in intravital microscopy and the use of vital fluorescent dyes like green fluorescent protein (GFP) (18,40). Studies that couple these two powerful techniques have added greatly to our knowledge of the metastatic processes following tumor cell entry into circulatory compartments. The use of in vivo video microscopy allows for the direct observation of experimental metastasis over time (39). Cancer cells can be fluorescently labeled in vitro and then injected into an animal. The cells can then be viewed at different time points, by both fluorescence and oblique transillumination, in thin tissues or superficial (⩽50 μm) regions of thick tissues in vivo (39). Experiments using this technology have demonstrated that, in contrast to the long-held belief, the vast majority of cancer cells in the microcirculation manage not only to survive there but also to extravasate into the surrounding tissue within 1–2 days (41,42). Such studies have translated well into the clinical arena. Specifically, the vast majority of clinical studies using RT–PCR to detect prostate tumor cells in the peripheral circulation and bone marrow found no association between the detection of disseminated cells and treatment failure (16,20). Of interest, in a recent study of breast cancer patients (22), detection of cytokeratin-positive cancer cells in the bone marrow was associated with the development of overt metastases and death. The apparent difference between these two findings may be due to differences in study design (e.g., detection methods and markers used) or in factors that influence the growth of disseminated cancer cells at the metastatic site. Additional studies will be necessary to distinguish between these possibilities. Taken together, the clinical and experimental evidence supports the observation that dissemination from the primary tumor site is a frequent event. Furthermore, these independent and complementary studies strongly suggest that growth control of individual disseminated cells determines the efficiency of metastatic colonization.

Metastatic colonization is the lodging and subsequent growth of disseminated cancer cells to form clinically significant metastases (Fig. 1, B). To proliferate, surviving disseminated tumor cell(s) must be able to initiate cell-appropriate, context-dependent signaling cascades, which enable them to survive, enter the cell cycle, and divide. While disseminated cells are likely to be present in numerous organs, only certain environment(s) appear to allow their survival and subsequent growth (37,43,44). Intercellular interactions with the stroma and with other tumor cells are critical for tumor cell survival and involve the activation of adhesion-dependent survival pathways, such as those described for E-cadherin (45,46) and integrin molecules (47). Clusters of proliferating cells grow into lesions consisting of a few hundred that can be detected reliably by histologic methods. Cells within such microscopic lesions can receive oxygen and nutrients by diffusion. Progressive growth of microscopic lesions into overt or macroscopic metastases (>1 mm in diameter) requires that the fraction of proliferating cells exceed the fraction that are quiescent or apoptotic. This transition from microscopic to macroscopic metastasis has often been referred to as the switch to an angiogenic phenotype or the angiogenic switch (48). This terminology implies that microscopic metastases exist in one of two states: Either the lesion is angiogenic (forming new blood vessels), or it is not. However, the progression from a “microscopic lesion” to an overt metastasis is more accurately described in terms of growth control. Indeed, the interchangeable use of “angiogenesis” and “growth” has been a source of confusion. This progression may occur over a period of months or even years and is not necessarily dependent on new blood vessel formation. Vascularization is, in fact, a late step in metastatic colonization (49). Recent studies have shown that, in addition to the induction of classical neovascularization via endothelial cell recruitment, tumor cell masses can develop a blood supply by alternative means, such as the cooption of pre-existing host vessels (49) or by the formation of tumor channels, a process referred to as vascular mimicry (50). As we will describe in the following paragraphs, recent data from our laboratories suggest that a subset of metastasis-suppressor genes inhibits early steps in metastatic colonization, prior to the need for development or recruitment of vessels.

Identification of Metastasis-Suppressor Activity: a Functional Approach

Metastasis-suppressor genes suppress the formation of (spontaneous) macroscopic metastases. As their name implies, these genes are distinct from oncogenes, which promote cellular transformation, and tumor-suppressor genes, which suppress tumor growth. While the first metastasis-suppressor gene, nm23, was identified by a complementary DNA (cDNA) subtraction approach, the majority of metastasis-suppressor activities identified to date have been discovered using microcell-mediated chromosomal transfer (MMCT) (Table 1). The choice of the MMCT strategy was logical, since the existence of metastasis-suppressor genes was originally implicated by the results of somatic cell fusion studies, the precursor of MMCT (5154). The techniques for the generation of genetically stable somatic cell hybrids were developed in the early studies by Barski et al. [reviewed in (55)]. In most instances, fusion between malignant and normal cells results in hybrid cells that are suppressed in their tumorigenic capacity (56). Ichikawa et al. (57) were the first researchers to identify specific chromosomal losses associated with the reacquisition of metastatic ability. In their study, fusion of nonmetastatic with highly metastatic Dunning rat prostatic cancer cells resulted in nonmetastatic hybrids. More important, the tumorigenicity (e.g., tumor formation and latency period) and in vivo growth rates of the primary tumors of hybrid clones containing a full complement of rat chromosomes were not affected. At the experimental end point, none of the animals bearing hybrid tumors developed distant metastases. However, when the nonmetastatic primary tumors were serially passaged in vivo, animals occasionally developed distant metastases. Cytogenetic analysis of these metastatic revertants revealed a consistent loss of a copy of rat chromosome 2. This critical study suggested that the loss of specific chromosomes could increase the metastatic potential of prostate cancer cells without affecting growth rate or tumorigenicity.

The observation of a metastasis-suppression activity being associated with a specific chromosome coincided with the development of MMCT as a technique for the study of genes encoded by individual human chromosomes (5154,5861). In this approach, summarized in Fig. 2, well-characterized donor cells, carrying a single human chromosome tagged with a selectable marker or markers (e.g., neomycin phosphotransferase, etc.), are used to transfer the chromosome of interest into recipient cells (62). Briefly, donor cells are sequentially treated with Colcemid, to depolymerize microtubules, and cytochalasin-B, to depolymerize actin bundles. The treated cells are centrifuged, and the resulting pellet contains the microcells (63). Microcells are, in effect, micelles that contain a single chromosome or multiple chromosomes. To enrich for those containing a single chromosomes, the microcells are size fractionated by sequential filtration through polycarbonate membranes of decreasing pore size. Microcells become attached to recipient cells in the presence of phytohemagglutinin and then become fused with the addition of polyethylene glycol. Recipient cells containing human chromosomes are selected in G418-containing media and then characterized by molecular and cytogenetic methods, such as sequence tagged-site PCR, karyotyping, and fluorescence in situ hybridization (63,64). The complete characterization of the hybrids under study is critical, since it provides information on the addition and/or deletion of donor and recipient chromosomal material, as well as any rearrangements that may have occurred during MMCT.

Several laboratories have employed the technique of MMCT to test the functional significance of chromosomal alterations, such as loss of heterozygosity (LOH), observed in clinical samples. In addition, the use of MMCT, in combination with positional or expression-based cloning techniques, has allowed the functional identification of genes conveying phenotypes, such as senescence or tumor and metastasis suppression (15,65,66). A review of the literature shows that transfer of a given chromosome can have different phenotypic effects that are dependent on the characteristics of the recipient cell line. For example, the transfer of human chromosome 7 by MMCT into immortalized SUSM-1 fibroblast cells induces senescence (67), whereas transfer of the same chromosome into choriocarcinoma cells results in suppression of tumor growth in vivo (68). Such results have enabled the definition of complementation groups for particular chromosome functions. The potential outcomes of transferring a particular chromosome into highly metastatic cells are summarized in Fig. 2.

Studies using highly metastatic Dunning rat prostatic cells as the recipients for chromosomal transfer (63,64) showed that chromosomes 12 and 17 specifically suppressed the metastatic ability of these cells. The observed metastasis suppression had no effect on tumor growth rate. Of interest, in analogous studies of human prostate cancer cell lines, transfer of these chromosomes suppressed the cells' tumorigenicity (69,70). These findings could result from at least three alternative mechanisms. First, a given chromosomal region may encode a number of different genes, one or more of which may be active as a tumor-suppressor gene in human prostate cancer cells but be inactive or not expressed in rat prostate cancer cells. Second, genes may function as metastasis-suppressor genes when expressed in rat prostate cancer cells but may be inactive or not expressed in human prostate cancer cell lines. Third, gene(s) that lie in the same chromosomal region may have different functions, depending on the context (i.e., cell type) in which they are expressed.

In the third scenario, the effect of the gene product may be limited or determined by the recipient cells. We refer to this scenario as the “cellular hard-wiring” hypothesis.1 For example, human prostate cancer cell lines compared with Dunning rat prostate cancer cell lines are weakly metastatic in spontaneous metastasis assays (t162). These differences in their in vivo biologic activities could be the result of genetic differences between the tumor cells, or they could result from an epigenetic mechanism, such as differential tumor–stromal interactions. The nature of cellular interactions with the extracellular matrix can regulate tissue-specific gene expression, since cells form an elaborate three-dimensional network composed of the nuclear, cytoskeletal, and extracellular matrices (27,71). Thus, the differential effects of a given chromosome transferred into different cell types can be the result of differential expression of the genes on the chromosome as determined by the way a cell responds to its environment.

During the past decade, several human chromosomes have been functionally tested through the use of MMCT, and metastasis-suppressor activities have been reported on chromosomes 1, 6, 7, 8, 10, 11, 12, 16, and 17 (63,64,7282) (Table 1). Such functional studies, combined with positional and expression-based gene cloning techniques, have enabled the identification of KAI1, KISS-1, MKK4/SEK1, and BRMS1 as metastasis-suppressor genes.

Metastasis-Suppressor Genes

As discussed earlier, metastasis-suppressor genes suppress the formation of spontaneous, macroscopic metastases without affecting the growth rate of the primary tumor. To date, five genes, nm23 (NME1), KAI1, KiSS1, BrMS1, and MKK4 (MAP2K4), have been shown to meet the criteria of a metastasis-suppressor gene [see Table 2; (31,32,83119)]. The role of other genes, such as CD44 and maspin/PI5, in metastasis suppression is less well defined (102,120131). The potential mechanism of action of all of these genes has been inferred by analogy to other family members and observations in model systems. How these genes and their protein products function to suppress metastasis in vivo is the subject of enthusiastic study. Decreased expression of the suppressor gene is the key parameter determining metastatic potential and may occur by a variety of mechanisms, not necessarily LOH (32,91). To date, nm23 (NME1) and KAI1 are the best-characterized metastasis-suppressor genes.

nm23 (NME1)

The prototypical metastasis-suppressor gene, nm23, was identified in the murine K1735 melanoma by use of subtractive hybridization (a method to identify genes differentially expressed between two cell lines), and six human homologues have been identified (90). Loss of Nm23-H1 expression is associated with metastatic potential in many, but not all, late-stage tumors (91). Transfection of nm23-H1 cDNA into highly metastatic murine melanoma, rat mammary adenocarcinoma, and human breast cancer and melanoma cells reduces their invasiveness and metastatic ability in vivo (91). In cancers, such as lung, colon, prostate, etc. [reviewed in (87)], where no alterations in the expression pattern of Nm23-H1 are evident, it is possible that the biologic function of Nm23-H1 does not influence malignant progression in these cell types. Alternatively, its effects may be inhibited by alternate mechanisms. The mechanism of action for metastasis suppression by Nm23 still remains unknown; however, evidence suggests that it is phosphorylated and may be involved in a novel signaling pathway that, in turn, controls cell motility (84,87).

KAI1

The localization of metastasis-suppressor activity to rat chromosome 2 in the cell fusion experiments by Ichikawa et al. (57) prompted the search for homologous metastasis-suppressor genes for human prostate cancer. The first of such genes identified was KAI1. MMCT was used to transfer human chromosome 11 into Dunning AT6.1 and AT3.1 rat prostate cancer cells, and the resulting microcell hybrids were assayed for metastasis suppression in immunodeficient mice (81). These studies led to the identification of the metastasis-suppressor gene KAI1, which maps to 11p11.2–p13 (101). The metastasis-suppressor activity of KAI1 was subsequently demonstrated by transfecting it into AT6.1 cells and assaying the metastatic ability of individual transfected control cell lines in severe combined immunodeficient (SCID) mice (101). Reports (101,110) suggest that expression of KAI1 decreases both the invasiveness and motility of cells in vitro. Additional studies show that KAI1 transfectants exhibit enhanced Ca2+-independent aggregation, indicating that KAI1 expression alters cell–cell interactions (109). The metastasis-suppressor activity of KAI1 was subsequently demonstrated by its transfection into AT6.1 cells and assaying the metastatic ability of individual transfected control cell lines in SCID mice (101). Lowered expression of KAI1 has also been associated with progression in a wide variety of cancers, including pancreatic, hepatocellular, bladder, breast, and non-small-cell lung cancers (31,133136), as well as esophageal cell carcinomas (137) and squamous and lymphoid neoplasms (138). These data suggest that KAI1 has a conserved metastasis-suppressor function. Furthermore, these studies demonstrate that metastasis-suppressor genes can be developed as clinical markers even before their biochemical mechanism of action has been elucidated.

Emerging Role of Metastasis-Suppressor Genes in the Regulation of Metastatic Growth

While it is tempting to speculate on the mechanism of action of genes shown in Table 2, examination of how genes, such as MKK4 or BRMS1, suppress metastasis will require construction of appropriate biochemical constructs and identification of in vitro conditions that will enable us to conduct meaningful biochemical and molecular studies. As a first step to accomplishing this goal, our laboratories have initiated studies designed to examine the step in the metastatic cascade inhibited by a chromosome or gene of interest. As an example of these studies, we will present recent work on the metastasis-suppressor activity encoded by chromosomes 17 and 6. These studies have brought us closer to defining mechanisms of metastasis suppression.

Chromosome 17

We have reported the identification of discontinuous portions of human chromosome 17 (D17S952 → D17S805, D17S930 → D17S797, and D17S944 → qter) that together suppress the metastatic ability of AT6.1 Dunning rat prostatic cancer cells when introduced via MMCT (63,80). PCR and Southern blot analyses demonstrated that three of the four markers on 17p13, including HIC1 and TP53, and 12 of the 13 markers in 17q21-23, including BRCA1 and the metastasis-suppressor gene NME1 (nm23), were not retained in this region (63). AT6.1 microcell hybrids containing this portion of chromosome 17 were tested in vivo in spontaneous metastasis assays. Spontaneous metastasis is measured by the ability of tumor cells to form a locally growing tumor at the site of injection and disseminate and grow at to secondary sites thereafter.

At the experimental end point, the number of overt surface metastases observed in the lungs from mice with AT6.1–17 tumors was reduced 15- to 30-fold compared with lungs from mice bearing parental AT6.1 tumors (63). This suppression could be due to the inhibition of any step within the metastatic cascade. We reasoned that examination of the biology of metastasis suppression would provide clues to the identity of genes responsible for suppression of metastatic growth. A series of in vivo experiments were conducted, and no evidence was found to suggest that there is a decrease in the number and/or viability of tumor cells colonizing the lung (80).

On the basis of these findings, we hypothesized that a gene or genes encoded by the suppressor region of chromosome 17 function by inhibiting the growth of metastases in the lung (139). To test this possibility, AT6.1–17 cells were transduced with a β-galactosidase reporter gene construct (AT6.1–17Tβgal cells) and tested in spontaneous metastasis assays (2). At the experimental end point, animals were killed and the excised lungs were stained for β-galactosidase expression. This approach allowed the visualization of microscopic AT6.1–17βgal surface metastases. Subcutaneous injection of AT6.1 parental cells resulted in the formation of a mean number of 97 overt surface metastases (detected using Bouin's fixation) per lung (Fig. 3, A; left). As expected, the number of overt macroscopic metastases after the subcutaneous injection of AT6.1–17-Tβgal cells was greatly reduced (Fig. 3, A; middle). In contrast, when lungs removed from the mice carrying AT6.1–17-Tβgal tumors were stained for β-galactosidase activity, numerous blue-staining microscopic metastases were observed (Fig. 3, A; right). Of interest, the mean number of AT6.1–17-Tβgal micrometastases (i.e., 62 ± 12 standard error [SE]) detected by this method is on the same order of magnitude as the mean number of macroscopic AT6.1 metastases (i.e., 97 ± 6 SE). These results demonstrate that AT6.1–17 cells do escape from the primary tumor and arrive in the lungs but do not form large metastatic foci (139). Development of overt metastases was associated with loss of the metastasis-suppressor region of chromosome 17 (139).

Because of the similarity between our findings to the angiostatin-mediated dormancy reported by Holmgren et al. (140), we investigated the possibility that AT6.1–17 primary tumors secrete a substance that suppresses the growth of its own metastases (139). For this experiment, 2 × 105 AT6.1–17 cells were injected subcutaneously into the flanks of SCID mice, which were then divided into two experimental groups. Once the tumors reached a volume of 1 cm3, they were surgically removed from the mice in the first group, while those in the second group were left intact, although a contralateral sham surgery was performed. It was anticipated that if the AT6.1–17 primary tumor secreted a substance like angiostatin, which suppresses the growth of its own metastases, then a substantial increase in the number of overt metastases should develop in the lungs of mice in which the primary tumors had been removed. However, after approximately 65 days after injection, the animals were killed and examination of the lungs from both groups showed no difference in the numbers of overt macrometastases (139). Thus, these studies found no evidence for an antiangiogenic mechanism in this model.

Taken together, our data suggested that AT6.1–17 cells escape from the primary tumor but are growth inhibited at the secondary site (139). If this is an early event, we predicted that viable, disseminated AT6.1 and AT6.1–17 cells should be present in the lung at very early time points. We found that viable cells could be harvested from the lungs of both AT6.1 and AT6.1–17 tumor bearers as early as 18 days after injection (Fig. 3, B). Our preliminary time-course data show that AT6.1–17 cells disseminate and lodge in the lungs but have an extended latent period as compared with AT6.1 parental cells.

Chromosome 6

On the basis of the high incidence of chromosome 6 abnormalities in late-stage human melanoma (141), we introduced an intact chromosome 6 into the highly metastatic C8161 human melanoma cells by MMCT. Parental cells formed tumors in every mouse given an intradermal injection of 1 × 106 cells, and more than 90% of the mice developed regional lymph node and lung metastases. In contrast, chromosome 6-C8161 hybrids (neo6/C8161) were still tumorigenic but completely suppressed for metastasis (142). Intravenous injection of neo6/C8161 cells also did not produce metastases. In a recent study (143), introduction of a version of a chromosome 6 with deletions on the long arm allowed refinement of the metastasis-suppressor locus to a 40-megabase (Mb) region represented by chromosomal bands 6q16.3–q23.

The mechanism of action for the metastasis-suppressor protein from the gene on chromosome 6 was studied using a variety of in vitro and in vivo techniques. The neo6/C8161 cells were still locally invasive, and cells were even detected in efferent vessels. This finding implied that the step(s) in the metastatic cascade inhibited by introduction of chromosome 6 occurred subsequent to intravasation. The identity of those steps was not further elucidated using in vitro assays mimicking adhesion, invasion, motility, or growth. No important differences between the metastatic and nonmetastatic cells were observed using the many in vitro assays (72,73,141,142).

For a better definition of the step(s) in metastasis blocked by addition of chromosome 6, cells that constitutively express GFP were engineered. GFP-tagged C8161 and neo6/C8161 cells were injected intravenously into athymic mice. C8161, as expected, formed overt metastases, but neo6/C8161 cells did not. Microscopic metastases (single cells or clusters of <10 cells) were observed in the lungs following neo6/C8161 cell injection, suggesting that these cells lodged in the lungs but failed to proliferate (144). For the determination of whether the fluorescing cells were viable, they were isolated from the lung up to 60 days after injection and grown in culture. On injection into the skin of athymic mice, the neo6/C8161 cells isolated from the lung grew at rates similar to those of previously injected neo6/C8161 cells. This result implies that the gene or genes on chromosome 6 interfere specifically with growth-regulatory responses in the lung but not in the skin.

From Gene Discovery to Clinical Utility

This review has focused on the identification and development of metastasis-suppressor genes as new additions to our molecular armamentarium. As translational researchers, our immediate goals are 1) to improve the ability of the pathologist to distinguish unambiguously malignant from indolent lesions and 2) to help the clinician differentiate tumors that are highly likely to metastasize from those that are not. The practical question, therefore, is: How can we use these genes, or the pathways that they regulate, to improve patient management? When the search for metastasis-suppressor genes was initiated in the late 1980s, the major challenge was the identification of candidate genes. Recently, however, there has been an explosion in the genetic information that is instantly available. Furthermore, because of the efforts of independent laboratories and cooperative efforts, such as the Cancer Genome Anatomy Project of the National Cancer Institute (Bethesda, MD), cancer transcriptomes and proteomes will soon be available (145,146). New technologies will continue to increase our ability to dissect molecular pathways in individual cells within human cancers. While this wealth of information will no doubt be of use, work from the groups of Bissell, Cunha, and Chung (147154) has clearly demonstrated that tissue structure determines, or at least greatly influences, gene expression and function. Thus, it may be extremely difficult to predict the importance of genes expressed in individual microdissected cancer cells to the biology of the intact tumor, the behavior of which is determined by complex interactions among a population of cells. The present challenge is to identify the genes that are functionally important in the acquisition of metastatic ability. Achieving this goal will require the use of well-characterized, in vivo (animal) models coupled with clinical correlative studies. It must be emphasized that in vitro models do not accurately reflect in vivo metastasis (155). Indeed, none of the metastasis-suppressor genes described herein could have been identified using traditional in vitro assays. Given the inherent variability and nonlinear behaviors of biologic systems, it is probable that no one model will prove to be adequate to separate out the contributions of the multiplicity of genes involved in the development of metastases. Thus, it is more advantageous to focus studies on a particular model and tease out important cellular pathways modulated by a particular gene of interest in that model and then to test and verify the importance of the target pathway in clinical disease as well as in additional model systems.

Technologic advances are enabling us to examine the metastatic process and the genes that regulate it in new ways. This ability has led us to re-evaluate fundamental concepts concerning the determinants of metastatic propensity. In the past, the escape of cells from the primary site was viewed as the rate-limiting step for the development of metastases. The clinical implication was that disseminated cancer cells were destined to grow into lethal metastases; thus, they were not a target for therapeutic intervention (18). Findings from clinical studies and basic research from several independent laboratories have shown that survival and subsequent growth of extravasated cancer cells at the secondary site may determine metastatic efficacy. These observations are driving our laboratories and others to reconsider the role of endothelial cell–tumor cell interactions in survival, signaling, and growth control cascades to develop new strategies for controlling the growth of disseminated cancer cells (39,45,156).

As metastasis researchers, we find ourselves in the midst of a revolution. In preparing this review, we considered the parallels between recent developments in our field and the development of the field of molecular biology. Much of early molecular biology was pursued by individuals who were not trained as biologists, but as physicists, such as Max Delbrück (157). We are respectful of the observations of Erwin Schrödinger, the father of statistical mechanics, who observed that, “all of the physical and chemical laws that are known to play an important part in the life of organisms are of the statistical kind. The behavior of such systems depends entirely on a large number of molecules that cooperate to form the observed function or phenotype” (158). Although this comment was made in regard to normal biologic processes, it is equally applicable to the multiple genetic changes that are required for the acquisition of metastatic ability. Metastasis is a complex, multigenic phenotype. As such, multiple markers will be needed for the accurate assessment of the metastatic ability of tumors and tumor cells. This need is highlighted by the tremendous impact of seemingly trivial experimental manipulations on the outcome of metastasis assays (155). Parallels have been drawn between the behavior of cancer cells and complex adaptive systems (159,160). As such, very small changes in initial conditions may produce an outcome of such great diversity as to appear random (159). Ultimately, we believe that, to translate our molecular findings into meaningful markers, we will have to go beyond our traditional areas of expertise and work with mathematicians, computational biologists, and others to take this revolution from bench to bedside.

Table 1.

Chromosomal regions identified by microcell-mediated chromosomal transfer that suppress metastases in vivo*

Chromosomal location Tumor type or site (reference Nos.) Cell lines tested (species of origin) In vitro phenotype† In vivo phenotype 
*ND = not determined, NE = not examined, GFP = green fluorescent protein, and mets. = metastases. 
†Motility was measured by micropipet motility assay or by migration toward a chemoattractant in Boyden chambers. Invasion was measured by migration through Matrigel. 
‡Rinker-Schaeffer CW: unpublished results. 
Chromosome 1 Melanoma (72) MelJuSo (human) ND ↓ Spontaneous mets. 
    ↓ Experimental mets. 
6q16.3–q23 Melanoma (73,74) C8161 (human) ↓ Motility ↓ Spontaneous mets.  
    ↓ Experimental mets. 
    Occasional single cells (detected by GFP tagging) which are growth suppressed but viable 
  MelJuSo (human) ND ↓ Spontaneous mets.  
    ↓ Experimental mets. 
Chromosome 6 Breast (75) MDA-MB-435 (human) ND NE spontaneous mets. 
7q21–22 and/or 7q31.2–32 Prostate (76) AT6.3 (rat) ND ↓ Spontaneous mets.  
    ↓ Experimental mets. 
8p21–p12 Prostate (77,78) AT6.2 (rat) ↓ Invasion ↓ Spontaneous mets.  
    NE experimental mets. 
10cen–10q23 Prostate (79) AT6.3 (rat) ND ↓ Spontaneous mets. 
11q13.1–13.2 Breast (75,80) MDA-MB-435 (human) ND ↓ Spontaneous mets. 
11pter–q14  R1564 (rat) ND NE spontaneous mets. 
11p11.2–13 Prostate (68) AT6.1 (rat) ND ↓ Spontaneous mets.  
  AT3.1 (rat) ND ↓ Spontaneous mets. 
12qcen–q13 and/or 12q24–ter Prostate (64) AT6.1 (rat) NE motility‡ ↓ Spontaneous mets.  
   NE invasion‡ No micrometastases observed at the experimental end point 
16q24.2 Prostate (82) AT6.1 (rat) ND ↓ Spontaneous mets. 
17p12–11.2 and/or 17cen–q12 Prostate (63) AT6.1 (rat) NE motility‡ ↓ Spontaneous mets.  
   NE invasion‡ Micrometastases observed at the experimental end point  
Chromosomal location Tumor type or site (reference Nos.) Cell lines tested (species of origin) In vitro phenotype† In vivo phenotype 
*ND = not determined, NE = not examined, GFP = green fluorescent protein, and mets. = metastases. 
†Motility was measured by micropipet motility assay or by migration toward a chemoattractant in Boyden chambers. Invasion was measured by migration through Matrigel. 
‡Rinker-Schaeffer CW: unpublished results. 
Chromosome 1 Melanoma (72) MelJuSo (human) ND ↓ Spontaneous mets. 
    ↓ Experimental mets. 
6q16.3–q23 Melanoma (73,74) C8161 (human) ↓ Motility ↓ Spontaneous mets.  
    ↓ Experimental mets. 
    Occasional single cells (detected by GFP tagging) which are growth suppressed but viable 
  MelJuSo (human) ND ↓ Spontaneous mets.  
    ↓ Experimental mets. 
Chromosome 6 Breast (75) MDA-MB-435 (human) ND NE spontaneous mets. 
7q21–22 and/or 7q31.2–32 Prostate (76) AT6.3 (rat) ND ↓ Spontaneous mets.  
    ↓ Experimental mets. 
8p21–p12 Prostate (77,78) AT6.2 (rat) ↓ Invasion ↓ Spontaneous mets.  
    NE experimental mets. 
10cen–10q23 Prostate (79) AT6.3 (rat) ND ↓ Spontaneous mets. 
11q13.1–13.2 Breast (75,80) MDA-MB-435 (human) ND ↓ Spontaneous mets. 
11pter–q14  R1564 (rat) ND NE spontaneous mets. 
11p11.2–13 Prostate (68) AT6.1 (rat) ND ↓ Spontaneous mets.  
  AT3.1 (rat) ND ↓ Spontaneous mets. 
12qcen–q13 and/or 12q24–ter Prostate (64) AT6.1 (rat) NE motility‡ ↓ Spontaneous mets.  
   NE invasion‡ No micrometastases observed at the experimental end point 
16q24.2 Prostate (82) AT6.1 (rat) ND ↓ Spontaneous mets. 
17p12–11.2 and/or 17cen–q12 Prostate (63) AT6.1 (rat) NE motility‡ ↓ Spontaneous mets.  
   NE invasion‡ Micrometastases observed at the experimental end point  

Table 2. Summary of metastasis-suppressor genes identified*

Fig. 1.

Development of spontaneous hematogenous metastases. Panel A: The development of spontaneous hematogenous metastases requires cancer cells to complete a well-defined series of steps. This figure is adapted from (13).Panel B: To form overt metastases, disseminated cells must complete additional steps at the metastatic site(s).

Fig. 1.

Development of spontaneous hematogenous metastases. Panel A: The development of spontaneous hematogenous metastases requires cancer cells to complete a well-defined series of steps. This figure is adapted from (13).Panel B: To form overt metastases, disseminated cells must complete additional steps at the metastatic site(s).

Fig. 2.

Identification of metastasis-suppressor activities by use of microcell-mediated chromosomal transfer. A9 donor cells containing a single human chromosome are used for the preparation of microcells that will be transferred to the recipient cell lines. Stable microcell hybrids are selected and characterized by molecular and cellular methods. To screen the microcell hybrids for the minimal metastasis-suppressor region, one employs the following techniques: FISH, karyotyping, and PCR amplification for human specific sequences using STS markers. Ultimately, the presence or absence of a suppressor region is determined by subcutaneous injection of the hybrids into the flanks of immunodeficient mice (63). Inclusion of a variety of controls is critical for the definition of metastasis-suppressor activity in vivo. The potential outcomes of in vivo studies using control, suppressed, and unsuppressed hybrids are illustrated at left. FISH = fluorescence in situ hybridization, PCR = polymerase chain reaction, STS = sequence-tagged site, and PEG = polyethylene glycol.

Fig. 2.

Identification of metastasis-suppressor activities by use of microcell-mediated chromosomal transfer. A9 donor cells containing a single human chromosome are used for the preparation of microcells that will be transferred to the recipient cell lines. Stable microcell hybrids are selected and characterized by molecular and cellular methods. To screen the microcell hybrids for the minimal metastasis-suppressor region, one employs the following techniques: FISH, karyotyping, and PCR amplification for human specific sequences using STS markers. Ultimately, the presence or absence of a suppressor region is determined by subcutaneous injection of the hybrids into the flanks of immunodeficient mice (63). Inclusion of a variety of controls is critical for the definition of metastasis-suppressor activity in vivo. The potential outcomes of in vivo studies using control, suppressed, and unsuppressed hybrids are illustrated at left. FISH = fluorescence in situ hybridization, PCR = polymerase chain reaction, STS = sequence-tagged site, and PEG = polyethylene glycol.

Fig. 3.

Examination of the mechanism of metastasis suppression by chromosome 17 and 6. A) Quantification of overt surface metastases and micrometastases. AT6.1 cells are highly metastatic rat prostate cancer cells. AT6.1–17-Tβgal cells contain the metastasis-suppressor region of human chromosome 17 and are tagged with a β-galactosidase reporter gene enabling the sensitive detection of microscopic metastases. The numbers of overt and microscopic metastases were determined by use of Bouin's fixation and β-galactosidase activity, respectively. At the experimental end point, lungs were removed from tumor-bearing animals. Left—lung from AT6.1 tumor-bearing animal stained with Bouin's solution; middle—lung from AT6.1–17-Tβgal tumor-bearing animal stained with Bouin's solution; and right—lung from AT6.1–17-Tβgal tumor-bearing animal stained for β-galactosidase activity. The average number of overt or microscopic metastases and standard error are shown below the panels. This figure is adapted from (139).B) A combination of techniques has been used to examine the time course of cancer cell dissemination and growth in suppressed AT6.1–17 cells as compared with metastatic AT6.1 parental cells. These data indicate that genes encoded by chromosome 17 inhibit a step in metastatic colonization. C) Photomicrographs of mouse lung following intravenous injection of green fluorescent protein-tagged C8161 and metastasis-suppressed neo6/C8161 cells (panel C3) are present. At 1 month, however, C8161 cells have proliferated to form macroscopic lung lesions (C2), but most neo6/C8161 cells have been cleared. Occasional single cells (C4, arrows) can be found in the lungs but fail to proliferate. These results imply that chromosome 6 suppresses metastasis by inhibiting the ability of C8161 cells to grow in the lung at an early stage of colonization (original magnification ×300). Data adapted from (144).

Fig. 3.

Examination of the mechanism of metastasis suppression by chromosome 17 and 6. A) Quantification of overt surface metastases and micrometastases. AT6.1 cells are highly metastatic rat prostate cancer cells. AT6.1–17-Tβgal cells contain the metastasis-suppressor region of human chromosome 17 and are tagged with a β-galactosidase reporter gene enabling the sensitive detection of microscopic metastases. The numbers of overt and microscopic metastases were determined by use of Bouin's fixation and β-galactosidase activity, respectively. At the experimental end point, lungs were removed from tumor-bearing animals. Left—lung from AT6.1 tumor-bearing animal stained with Bouin's solution; middle—lung from AT6.1–17-Tβgal tumor-bearing animal stained with Bouin's solution; and right—lung from AT6.1–17-Tβgal tumor-bearing animal stained for β-galactosidase activity. The average number of overt or microscopic metastases and standard error are shown below the panels. This figure is adapted from (139).B) A combination of techniques has been used to examine the time course of cancer cell dissemination and growth in suppressed AT6.1–17 cells as compared with metastatic AT6.1 parental cells. These data indicate that genes encoded by chromosome 17 inhibit a step in metastatic colonization. C) Photomicrographs of mouse lung following intravenous injection of green fluorescent protein-tagged C8161 and metastasis-suppressed neo6/C8161 cells (panel C3) are present. At 1 month, however, C8161 cells have proliferated to form macroscopic lung lesions (C2), but most neo6/C8161 cells have been cleared. Occasional single cells (C4, arrows) can be found in the lungs but fail to proliferate. These results imply that chromosome 6 suppresses metastasis by inhibiting the ability of C8161 cells to grow in the lung at an early stage of colonization (original magnification ×300). Data adapted from (144).

1In our work, the concept of cellular hardwiring refers to the work of Pienta and Coffey (71) and Pienta et al. (132).
Supported by Public Health Service (PHS) grant CA62168 (D. R. Welch) from the National Cancer Institute (NCI), National Institutes of Health (NIH), Department of Health and Human Services (DHHS); by DAMD1-96-6152 (D. R. Welch) from the Army Medical Research and Materiel Fund; by The National Foundation for Cancer Research (D. R. Welch); by The Jake Gittlen Memorial Golf Tournament (D. R. Welch); by PHS grants P20CA66132 and R29CA6948702 (C. W. Rinker-Schaeffer) from the NCI, NIH, DHHS; by Cancer Research Foundation Young Investigator Award (C. W. Rinker-Schaeffer); by the American Foundation for Urologic Disease (B. A. Yoshida and C. W. Rinker-Schaeffer), and by the University of Chicago RESCUE Fund (B. A. Yoshida, C. W. Rinker-Schaeffer).
We thank Ms. Zita Dubauskas and Mr. Edwin F. Schaeffer III for their editorial efforts. We also thank Dr. Charles Brendler of the University of Chicago for his enthusiastic support of this research.

References

1
Mckillop WJ, O'Sullivan B, Gospodarowicz M. The role of cancer staging in evidence-based medicine.
Cancer Prev Control
 
1998
;
2
:
269
–77.
2
Gospodarowicz M, Benedet L, Hutter RV, Fleming I, Henson DE, Sobin LH. History and international developments in cancer staging.
Cancer Prev Control
 
1998
;
2
:
262
–8.
3
Bostwick DG. Staging prostate cancer—1997: current methods and limitations.
Eur Urol
 
1997
;
32
:
2
–14.
4
Carter H, Partin AW. Diagnosis and staging of prostate cancer. In: Walsh PC, Vaughn ED, Wein AJ, editors. Campbell's urology. Philadelphia (PA): Saunders; 1998. p. 2519–38.
5
Isaacs JT. Molecular markers for prostate cancer metastasis. Developing diagnostic methods for predicting the aggressiveness of prostate cancer.
Am J Pathol
 
1997
;
150
:
1511
–21.
6
Schellhammer PF. Radical prostatectomy. Patterns of local failure and survival in 67 patients.
Urology
 
1988
;
31
:
191
–7.
7
Lerner SP, Seale-Hawkins C, Carlton CE Jr, Scardino PT. The risk of dying of prostate cancer in patients with clinically localized disease.
J Urol
 
1991
;
146
:
1040
–5.
8
Heimann R, Hellman S. Clinical progression of breast cancer malignant behavior: what to expect and when to expect it.
J Clin Oncol
 
2000
;
18
:
591
–9.
9
Catalona WJ, Dresner SM. Nerve-sparing radical prostatectomy: extraprostatic tumor extension and preservation of erectile function.
J Urol
 
1985
;
134
:
1149
–51.
10
Lu-Yao GL, McLerran D, Wasson J, Wennberg JE. An assessment of radical prostatectomy. Time trends, geographic variation, and outcomes. The Prostate Patient Outcomes Research Team.
JAMA
 
1993
;
269
:
2633
–6.
11
Voges GE, McNeal JE, Redwine EA, Freiha FS, Stamey TA. Morphologic analysis of surgical margins with positive findings in prostatectomy for adenocarcinoma of the prostate.
Cancer
 
1992
;
69
:
520
–6.
12
Heimann R, Lan F, McBride R, Hellman S. Separating favorable from unfavorable prognostic markers in breast cancer: the role of E-cadherin.
Nature
 
1980
;
283
:
139
–46.
13
Poste G, Fidler, IJ. The pathogenesis of cancer metastasis.
Nature
 
1980
;
283
:
139
–46.
14
MacDonald NJ, Steeg PS. Molecular basis of tumour metastasis.
Cancer Surv
 
1993
;
16
:
175
–99.
15
Welch DR, Rinker-Schaeffer CW. What defines a useful marker of metastasis in human cancer?
J Natl Cancer Inst
 
1999
;
91
:
1351
–3.
16
Melchior SW, Corey E, Ellis WJ, Ross AA, Layton TJ, Oswin MM, et al. Early tumor cell dissemination in patients with clinically localized carcinoma of the prostate.
Clin Cancer Res
 
1997
;
3
:
249
–56.
17
Heimann R, Ferguson DJ, Hellman S. The relationship between nm23, angiogenesis, and the metastatic proclivity of node-negative breast cancer.
Cancer Res
 
1998
;
58
:
2766
–71.
18
Chambers AF. The metastatic process: basic research and clinical implications.
Oncol Res
 
1999
;
11
:
161
–8.
19
Gomella LG, Raj GV, Moreno JG. Reverse transcriptase polymerase chain reaction for prostate specific antigen in the management of prostate cancer.
J Urol
 
1997
;
158
:
326
–37.
20
Christiano AP, Yoshida BA, Dubauskas Z, Sokoloff M, Rinker-Schaeffer CW. Development of markers of prostate cancer metastasis: review and perspective.
Urol Oncol
 
2000
;
5
:
217
–23.
21
Wood DP Jr, Banerjee M. Presence of circulating prostate cells in the bone marrow of patients undergoing radical prostatectomy is predictive of disease-free survival.
J Clin Oncol
 
1997
;
15
:
3451
–7.
22
Braun S, Pantel K, Muller P, Janni W, Hepp F, Kentenich CR, et al. Cytokeratin-positive cells in the bone marrow and survival of patients with stage I, II, or III breast cancer.
N Engl J Med
 
2000
;
342
:
525
–33.
23
Ebralidze A, Tulchinsky E, Grigorian M, Afanasyeva A, Senin V, Revazova E, et al. Isolation and characterization of a gene specifically expressed in different metastatic cells and whose deduced gene product has a high degree of homology to a Ca2+-binding protein family.
Genes Dev
 
1989
;
3
:
1086
–93.
24
Dear TN, Ramshaw IA, Kefford RF. Differential expression of a novel gene, WDNM1, in nonmetastatic rat mammary adenocarcinoma cells.
Cancer Res
 
1988
;
48
:
5203
–9.
25
Dear TN, McDonald DA, Kefford RF. Transcriptional down-regulation of a rat gene, WDNM2, in metastatic DMBA-8 cells.
Cancer Res
 
1989
;
49
:
5323
–8.
26
Basset P, Bellocq JP, Wolf C, Stoll I, Hutin P, Limacher JM, et al. A novel metalloproteinase gene specifically expressed in stromal cells of breast carcinomas.
Nature
 
1990
;
348
:
699
–704.
27
Fujita M, Spray DC, Choi H, Saez J, Jefferson DM, Hertzberg E, et al. Extracellular matrix regulation of cell–cell communication and tissue-specific gene expression in primary liver cultures.
Prog Clin Biol Res
 
1986
;
226
:
333
–60.
28
Fidler IJ, Radinsky R. Search for genes that suppress cancer metastasis [editorial].
J Natl Cancer Inst
 
1996
;
88
:
1700
–3.
29
Fidler IJ, Radinsky R. Genetic control of cancer metastasis [editorial].
J Natl Cancer Inst
 
1990
;
82
:
166
–8.
30
Steeg PS, Bevilacqua G, Kopper L, Thorgeirsson UP, Talmadge JE, Liotta LA, et al. Evidence for a novel gene associated with low tumor metastatic potential.
J Natl Cancer Inst
 
1988
;
80
:
200
–4.
31
Yang X, Welch DR, Phillips KK, Weissman BE, Wei LL. KAI1, a putative marker for metastatic potential in human breast cancer.
Cancer Lett
 
1997
;
119
:
149
–55.
32
Dong JT, Suzuki H, Pin SS, Bova GS, Schalken JA, Isaacs WB, et al. Down-regulation of the KAI1 metastasis suppressor gene during the progression of human prostatic cancer infrequently involves gene mutation or allelic loss.
Cancer Res
 
1996
;
56
:
4387
–90.
33
Folkman J. Angiogenesis and angiogenesis inhibition: an overview.
EXS
 
1997
;
79
:
1
–8.
34
Folkman J. The role of angiogenesis in tumor growth.
Semin Cancer Biol
 
1992
;
3
:
65
–71.
35
Pasqualini R, Ruoslahti E. Organ targeting in vivo using phage display peptide libraries.
Nature
 
1996
;
380
:
364
–6.
36
Fidler IJ. Review: biologic heterogeneity of cancer metastases.
Breast Cancer Res Treat
 
1987
;
9
:
17
–26.
37
Fidler IJ. Critical factors in the biology of human cancer metastasis: twenty- eighth G.H.A. Clowes memorial award lecture.
Cancer Res
 
1990
;
50
:
6130
–8.
38
Al-Mehdi AB, Tozawa K, Fisher AB, Shientag L, Lee A, Muschel RJ. Intravascular origin of metastasis from the proliferation of endothelium-attached tumor cells: a new model for metastasis.
Nat Med
 
2000
;
6
:
100
–2.
39
Chambers AF, MacDonald IC, Schmidt EE, Morris VL, Groom AC. Clinical targets for anti-metastasis therapy.
Adv Cancer Res
 
2000
;
79
:
91
–121.
40
Yang M, Jiang P, Sun FX, Hasegawa S, Baranov E, Chishima T, et al. A fluorescent orthotopic bone metastasis model of human prostate cancer.
Cancer Res
 
1999
;
59
:
781
–6.
41
Chambers AF, MacDonald IC, Schmidt EE, Koop S, Morris VL, Khokha R, et al. Steps in tumor metastasis: new concepts from intravital videomicroscopy.
Cancer Metastasis Rev
 
1995
;
14
:
279
–301.
42
MacDonald IC, Schmidt EE, Morris VL, Groom AC, Chambers AF. In vivo videomicroscopy of experimental hematogenous metastasis: cancer cell arrest, extravasation and migration. In: Wessels DR, editor. Motion analysis of living cells. New York (NY): John Wiley & Sons; 1998. p. 263–87.
43
Radinsky R, Fidler IJ. Regulation of tumor cell growth at organ-specific metastases.
In Vivo
 
1992
;
6
:
325
–31.
44
Fidler IJ. Critical determinants of melanoma metastasis.
J Investig Dermatol Symp Proc
 
1996
;
1
:
203
–8.
45
Day ML, Zhao X, Vallorosi CJ, Putzi M, Powell CT, Lin C, et al. E-cadherin mediates aggregation-dependent survival of prostate and mammary epithelial cells through the retinoblastoma cell cycle control pathway.
J Biol Chem
 
1999
;
274
:
9656
–64.
46
Guilford P. E-cadherin downregulation in cancer: fuel on the fire?
Mol Med Today
 
1999
;
5
:
172
–7.
47
Giancotti FG, Ruoslahti E. Integrin signaling.
Science
 
1999
;
285
:
1028
–32.
48
Arbiser JL, Moses MA, Fernandez CA, Ghiso N, Cao Y, Klauber N, et al. Oncogenic H-ras stimulates tumor angiogenesis by two distinct pathways.
Proc Natl Acad Sci U S A
 
1997
;
94
:
861
–6.
49
Holash J, Maisonpierre PC, Compton D, Boland P, Alexander CR, Zagzag D, et al. Vessel cooption, regression, and growth in tumors mediated by angiopoietins and VEGF.
Science
 
1999
;
284
:
1994
–8.
50
Maniotis AJ, Folberg R, Hess A, Seftor EA, Gardner LM, Pe'er J, et al. Vascular channel formation by human melanoma cells in vivo and in vitro: vasculogenic mimicry.
Am J Pathol
 
1999
;
155
:
739
–52.
51
Turpeenniemi-Hujanen T, Thorgeirsson UP, Hart IR, Grant SS, Liotta LA. Expression of collagenase IV (basement membrane collagenase) activity in murine tumor cell hybrids that differ in metastatic potential.
J Natl Cancer Inst
 
1985
;
75
:
99
–103.
52
Sidebottom E, Clark SR. Cell fusion segregates progressive growth from metastasis.
Br J Cancer
 
1983
;
47
:
399
–405.
53
Ramshaw IA, Carlsen S, Wang HC, Badenoch-Jones P. The use of cell fusion to analyse factors involved in tumour cell metastasis.
Int J Cancer
 
1983
;
32
:
471
–8.
54
Layton MG, Franks LM. Selective suppression of metastasis but not tumorigenicity of a mouse lung carcinoma by cell hybridization.
Int J Cancer
 
1986
;
37
:
723
–30.
55
Sager R. Genetic suppression of tumor formation.
Adv Cancer Res
 
1985
;
44
:
43
–68.
56
Harris H. The analysis of malignancy by cell fusion: the position in 1988.
Cancer Res
 
1988
;
48
:
3302
–6.
57
Ichikawa T, Ichikawa Y, Isaacs JT. Genetic factors and suppression of metastatic ability of prostatic cancer.
Cancer Res
 
1991
;
51
:
3788
–92.
58
Fournier RE, Ruddle FH. Microcell-mediated transfer of murine chromosomes into mouse, Chinese hamster, and human somatic cells.
Proc Natl Acad Sci U S A
 
1977
;
74
:
319
–23.
59
Ege T, Ringertz NR, Hamberg H, Sidebottom E. Preparation of microcells.
Methods Cell Biol
 
1977
;
15
:
339
–57.
60
Oshimura M. Lessons learned from studies on tumor suppression by microcell-mediated chromosome transfer.
Environ Health Perspect
 
1991
;
93
:
57
–8.
61
Saxon PJ, Srivatsan ES, Leipzig GV, Sameshima JH, Stanbridge EJ. Selective transfer of individual human chromosomes to recipient cells.
Mol Cell Biol
 
1985
;
5
:
140
–6.
62
Rinker-Schaeffer CW, Partin AW, Isaacs WB, Coffey DS, Isaacs JT. Molecular and cellular changes associated with the acquisition of metastatic ability by prostatic cancer cells.
Prostate
 
1994
;
25
:
249
–65.
63
Chekmareva MA, Hollowell CM, Smith RC, Davis EM, LeBeau MM, Rinker-Schaeffer CW. Localization of prostate cancer metastasis-suppressor activity on human chromosome 17.
Prostate
 
1997
;
33
:
271
–80.
64
Luu HH, Zagaja GP, Dubauskas Z, Chen SL, Smith RC, Watabe K, et al. Identification of a novel metastasis-suppressor region on human chromosome 12.
Cancer Res
 
1998
;
58
:
3561
–5.
65
Chang ZF. Regulatory mechanisms of replication growth limits in cellular senescence.
J Formos Med Assoc
 
1997
;
96
:
784
–91.
66
Vojta PJ, Futreal PA, Annab LA, Kato H, Pereira-Smith OM, Barrett JC. Evidence for two senescence loci on human chromosome 1.
Genes Chromosomes Cancer
 
1996
;
16
:
55
–63.
67
Nakabayashi K, Ogata T, Fujii M, Tahara H, Ide T, Wadhwa R, et al. Decrease in amplified telomeric sequences and induction of senescence markers by introduction of human chromosome 7 or its segments in SUSM-1.
Exp Cell Res
 
1997
;
235
:
345
–53.
68
Matsuda T, Sasaki M, Kato H, Yamada H, Cohen M, Barrett JC, et al. Human chromosome 7 carries a putative tumor suppressor gene(s) involved in choriocarcinoma.
Oncogene
 
1997
;
15
:
2773
–81.
69
Murakami YS, Brothman AR, Leach RJ, White RL. Suppression of malignant phenotype in a human prostate cancer cell line by fragments of normal chromosomal region 17q.
Cancer Res
 
1995
;
55
:
3389
–94.
70
Berube NG, Speevak MD, Chevrette M. Suppression of tumorigenicity of human prostate cancer cells by introduction of human chromosome del(12)(q13).
Cancer Res
 
1994
;
54
:
3077
–81.
71
Pienta KJ, Coffey DS. Nuclear–cytoskeletal interactions: evidence for physical connections between the nucleus and cell periphery and their alteration by transformation.
J Cell Biochem
 
1992
;
49
:
357
–65.
72
Miele ME, Robertson G, Lee JH, Coleman A, McGary CT, Fisher PB, et al. Metastasis suppressed, but tumorigenicity and local invasiveness unaffected, in the human melanoma cell line MelJuSo after introduction of human chromosomes 1 or 6.
Mol Carcinog
 
1996
;
15
:
284
–99.
73
You J, Miele ME, Dong C, Welch DR. Suppression of human melanoma metastasis by introduction of chromosome 6 may be partially due to inhibition of motility, but not to inhibition of invasion.
Biochem Biophys Res Commun
 
1995
;
208
:
476
–84.
74
Miele ME, De La Rosa A, Lee JH, Hicks DJ, Dennis JW, Steeg PS, et al. Suppression of human melanoma metastasis following introduction of chromosome 6 is independent of NME1 (Nm23).
Clin Exp Metastasis
 
1997
;
15
:
259
–65.
75
Phillips KK, Welch DR, Miele ME, Lee JH, Wei LL, Weissman BE. Suppression of MDA-MB-435 breast carcinoma cell metastasis following the introduction of human chromosome 11.
Cancer Res
 
1996
;
56
:
1222
–7.
76
Nihei N, Ohta S, Kuramochi H, Kugoh H, Oshimura M, Barrett JC, et al. Metastasis suppressor gene(s) for rat prostate cancer on the long arm of human chromosome 7.
Genes Chromosomes Cancer
 
1999
;
24
:
1
–8.
77
Nihei N, Ichikawa T, Kawana Y, Kuramochi H, Kugoh H, Oshimura M, et al. Mapping of metastasis suppressor gene(s) for rat prostate cancer on the short arm of human chromosome 8 by irradiated microcell-mediated chromosome transfer.
Genes Chromosomes Cancer
 
1996
;
17
:
260
–8.
78
Kuramochi H, Ichikawa T, Nihei N, Kawana Y, Suzuki H, Schalken JA, et al. Suppression of invasive ability of highly metastatic rat prostate cancer by introduction of human chromosome 8.
Prostate
 
1997
;
31
:
14
–20.
79
Nihei N, Ichikawa T, Kawana Y, Kuramochi H, Kugo H, Oshimura M, et al. Localization of metastasis suppressor gene(s) for rat prostatic cancer to the long arm of human chromosome 10.
Genes Chromosomes Cancer
 
1995
;
14
:
112
–9.
80
Rinker-Schaeffer CW, Hawkins AL, Ru N, Dong J, Stoica G, Griffin CA, et al. Differential suppression of mammary and prostate cancer metastasis by human chromosomes 17 and 11.
Cancer Res
 
1994
;
54
:
6249
–56.
81
Ichikawa T, Ichikawa Y, Dong J, Hawkins AL, Griffin CA, Isaacs WB, et al. Localization of metastasis suppressor gene(s) for prostatic cancer to the short arm of human chromosome 11.
Cancer Res
 
1992
;
52
:
3486
–90.
82
Mashimo T, Watabe M, Cuthbert AP, Newbold RF, Rinker-Schaeffer CW, Helfer E, et al. Human chromosome 16 suppresses metastasis but not tumorigenesis in rat prostatic tumor cells.
Cancer Res
 
1998
;
58
:
4572
–6.
83
Backer JM, Mendola CE, Kovesdi I, Fairhurst JL, O'Hara B, Eddy RL Jr, et al. Chromosomal localization and nucleoside diphosphate kinase activity of human metastasis-suppressor genes NM23-1 and NM23-2.
Oncogene
 
1993
;
8
:
497
–502.
84
Kantor JD, McCormick B, Steeg PS, Zetter BR. Inhibition of cell motility after nm23 transfection of human and murine tumor cells.
Cancer Res
 
1993
;
53
:
1971
–3.
85
Leone A, Flatow U, King CR, Sandeen MA, Margulies IM, Liotta LA, et al. Reduced tumor incidence, metastatic potential, and cytokine responsiveness of nm23-transfected melanoma cells.
Cell
 
1991
;
65
:
25
–35.
86
Parhar RS, Shi Y, Zou M, Farid NR, Ernst P, al-Sedairy ST. Effects of cytokine-mediated modulation of nm23 expression on the invasion and metastatic behavior of B16F10 melanoma cells.
Int J Cancer
 
1995
;
60
:
204
–10.
87
de la Rosa A, Williams RL, Steeg PS. Nm23/nucleoside diphosphate kinase: toward a structural and biochemical understanding of its biological functions.
Bioessays
 
1995
;
17
:
53
–62.
88
Baba H, Urano T, Okada K, Furukawa K, Nakayama E, Tanaka H, et al. Two isotypes of murine nm23/nucleoside diphosphate kinase, nm23-M1 and nm23-M2, are involved in metastatic suppression of a murine melanoma line.
Cancer Res
 
1995
;
55
:
1977
–81.
89
Lee JH, Welch DR. Identification of highly expressed genes in metastasis-suppressed chromosome 6/human malignant melanoma hybrid cells using subtractive hybridization and differential display.
Int J Cancer
 
1997
;
71
:
1035
–44.
90
Freije JM, MacDonald NJ, Steeg PS. Differential gene expression in tumor metastasis: Nm23. In: Gunthert U, Shlag PM, Birchmeier W, editors. Attempts to understand metastasis formation II: regulatory factors. Berlin (Germany): Springer-Verlag; 1996. p. 215–32.
91
Freije JM, MacDonald NJ, Steeg PS. Nm23 and tumour metastasis: basic and translational advances.
Biochem Soc Symp
 
1998
;
63
:
261
–71.
92
Leone A, Flatow U, VanHoutte K, Steeg PS. Transfection of human nm23-H1 into the human MDA-MB-435 breast carcinoma cell line: effects on tumor metastatic potential, colonization and enzymatic activity.
Oncogene
 
1993
;
8
:
2325
–33.
93
Bhujwalla ZM, Aboagye EO, Gillies RJ, Chacko VP, Mendola CE, Backer JM. Nm23-transfected MDA-MB-435 human breast carcinoma cells form tumors with altered phospholipid metabolism and pH: a 31P nuclear magnetic resonance study in vivo and in vitro.
Magn Reson Med
 
1999
;
41
:
897
–903.
94
Russell RL, Geisinger KR, Mehta RR, White WL, Shelton B, Kute TE. nm23—relationship to the metastatic potential of breast carcinoma cell lines, primary human xenografts, and lymph node negative breast carcinoma patients.
Cancer
 
1997
;
79
:
1158
–65.
95
Fukuda M, Ishii A, Yasutomo Y, Shimada N, Ishikawa N, Hanai N, et al. Decreased expression of nucleoside diphosphate kinase alpha isoform, an nm23-H2 gene homolog, is associated with metastatic potential of rat mammary-adenocarcinoma cells.
Int J Cancer
 
1996
;
65
:
531
–7.
96
Lim S, Lee HY, Lee H. Inhibition of colonization and cell–matrix adhesion after nm23-H1 transfection of human prostate carcinoma cells.
Cancer Lett
 
1998
;
133
:
143
–9.
97
Hsu S, Huang F, Wang L, Banerjee S, Winawer S, Friedman E. The role of nm23 in transforming growth factor beta 1-mediated adherence and growth arrest.
Cell Growth Differ
 
1994
;
5
:
909
–17.
98
Miyazaki H, Fukuda M, Ishijima Y, Takagi Y, Iimura T, Negishi A, et al. Overexpression of nm23-H2/NDP kinase B in a human oral squamous cell carcinoma cell line results in reduced metastasis, differentiated phenotype in the metastatic site, and growth factor-independent proliferative activity in culture.
Clin Cancer Res
 
1999
;
5
:
4301
–7.
99
Otsuki K, Alcalde RE, Matsumura T, Kimura N. Immunohistochemical analysis of nucleoside diphosphate kinases in oral squamous cell carcinomas.
Oncology
 
1997
;
54
:
63
–8.
100
Kawana Y, Komiya A, Ueda T, Nihei N, Kuramochi H, Suzuki H, et al. Location of KAI1 on the short arm of human chromosome 11 and frequency of allelic loss in advanced human prostate cancer. Prostate 997;32:205–13.
101
Dong JT, Lamb PW, Rinker-Schaeffer CW, Vukanovic J, Ichikawa T, Isaacs JT, et al. KAI1, a metastasis suppressor gene for prostate cancer on human chromosome 11p11.2.
Science
 
1995
;
268
:
884
–6.
102
Gao AC, Lou W, Dong JT, Isaacs JT. CD44 is a metastasis suppressor gene for prostatic cancer located on human chromosome 11p13.
Cancer Res
 
1997
;
57
:
846
–9.
103
Ueda T, Ichikawa T, Tamaru J, Mikata A, Akakura K, Akimoto S, et al. Expression of the KAI1 protein in benign prostatic hyperplasia and prostate cancer.
Am J Pathol
 
1996
;
149
:
1435
–40.
104
Rubinstein E, Le Naour F, Lagaudriere-Gesbert C, Billard M, Conjeaud H, Boucheix C. CD9, CD63, CD81, and CD82 are components of a surface tetraspan network connected to HLA-DR and VLA integrins.
Eur J Immunol
 
1996
;
26
:
2657
–65.
105
Berditchevski F, Odintsova E. Characterization of integrin-tetraspanin adhesion complexes: role of tetraspanins in integrin signaling.
J Cell Biol
 
1999
;
146
:
477
–92.
106
Maecker HT, Todd SC, Levy S. The tetraspanin superfamily: molecular facilitators.
FASEB J
 
1997
;
11
:
428
–42.
107
Phillips KK, White AE, Hicks DJ, Welch DR, Barrett JC, Wei LL, et al. Correlation between reduction of metastasis in the MDA-MB-435 model system and increased expression of the Kai-1 protein.
Mol Carcinog
 
1998
;
21
:
111
–20.
108
Wei LL, Yang X, Phillips K, Weissman BE, Welch DR. Analysis of KAI-1 mRNA expression in human breast cancer cell lines [abstract].
Proc Am Assoc Cancer Res
 
1996
;
37
:
76
.
109
Takaoka A, Hinoda Y, Sato S, Itoh F, Adachi M, Haregama M, et al. Reduced invasive and metastatic potentials of KAI1-transfected melanoma cells.
Jpn J Cancer Res
 
1998
;
89
:
397
–404.
110
Takaoka A, Hinoda Y, Sato S, Adachi Y, Itoh F, Adachi M, et al. Suppression of invasive properties of colon cancer cells by a metastasis suppressor KAI1 gene.
Oncogene
 
1998
;
16
:
1443
–53.
111
Lombardi DP, Geradts J, Foley JF, Chiao C, Lamb PW, Barrett JC. Loss of KAI1 expression in the progression of colorectal cancer.
Cancer Res
 
1999
;
59
:
5724
–31.
112
West A, Vojta PJ, Welch DR, Weissman BE. Chromosome localization and genomic structure of the KiSS-1 metastasis suppressor gene (KISS1).
Genomics
 
1998
;
54
:
145
–8.
113
Lee JH, Miele ME, Hicks DJ, Phillips KK, Trent JM, Weissman BE, et al. KiSS-1, a novel human malignant melanoma metastasis-suppressor gene [published erratum appears in J Natl Cancer Inst 1997;89:1549].
J Natl Cancer Inst
 
1996
;
88
:
1731
–7.
114
Lee JH, Welch DR. Suppression of metastasis in human breast carcinoma MDA-MB-435 cells after transfection with the metastasis suppressor gene, KiSS-1.
Cancer Res
 
1997
;
57
:
2384
–7.
115
Seraj MJ, Samant RS, Verderame MF, Welch DR. Functional evidence for a novel human breast carcinoma metastasis suppressor, BRMS1, encoded at chromosome 11q13.
Cancer Res
 
2000
;
60
:
2764
–9.
116
White RA, Hughes RT, Adkison LR, Bruns G, Zon LI. The gene encoding protein kinase SEK1 maps to mouse chromosome 11 and human chromosome 17.
Genomics
 
1996
;
34
:
430
–2.
117
Yoshida BA, Dubauskas Z, Chekmareva MA, Christiano TR, Stadler WM, Rinker-Schaeffer CW. Mitogen-activated protein kinase kinase 4/stress-activated protein/Erk kinase 1 (MKK4/SEK1), a prostate cancer metastasis suppressor gene encoded by human chromosome 17.
Cancer Res
 
1999
;
59
:
5483
–7.
118
Whitmarsh AJ, Davis RJ. Transcription factor AP-1 regulation by mitogen-activated protein kinase signal transduction pathways.
J Mol Med
 
1996
;
74
:
589
–607.
119
Ip YT, Davis RJ. Signal transduction by the c-Jun N-terminal kinase (JNK)—from inflammation to development.
Curr Opin Cell Biol
 
1998
;
10
:
205
–19.
120
Gunthert U, Stauder R, Mayer B, Terpe HJ, Finke L, Friedrichs K. Are CD44 variant isoforms involved in human tumour progression?
Cancer Surv
 
1995
;
24
:
19
–42.
121
Lou W, Krill D, Dhir R, Becich MJ, Dong JT, Frierson HF Jr, et al. Methylation of the CD44 metastasis suppressor gene in human prostate cancer.
Cancer Res
 
1999
;
59
:
2329
–31.
122
Noordzij MA, van Steenbrugge GJ, Schroder FH, Van der Kwast TH. Decreased expression of CD44 in metastatic prostate cancer.
Int J Cancer
 
1999
;
84
:
478
–83.
123
Nagabhushan M, Pretlow TG, Guo YJ, Amini SB, Pretlow TP, Sy MS. Altered expression of CD44 in human prostate cancer during progression.
Am J Clin Pathol
 
1996
;
106
:
647
–51.
124
De Marzo AM, Bradshaw C, Sauvageot J, Epstein JI, Miller GJ. CD44 and CD44v6 downregulation in clinical prostatic carcinoma: relation to Gleason grade and cytoarchitecture.
Prostate
 
1998
;
34
:
162
–8.
125
Noordzij MA, van Steenbrugge GJ, Verkaik NS, Schroder FH, van der Kwast TH. The prognostic value of CD44 isoforms in prostate cancer patients treated by radical prostatectomy.
Clin Cancer Res
 
1997
;
3
:
805
–15.
126
Kallakury BV, Sheehan CE, Ambros RA, Fisher HA, Kaufman RP Jr, Muraca PJ, et al. Correlation of p34cdc2 cyclin-dependent kinase overexpression, CD44s downregulation, and HER-2/neu oncogene amplification with recurrence in prostatic adenocarcinomas.
J Clin Oncol
 
1998
;
16
:
1302
–9.
127
Sneath RJ, Mangham DC. The normal structure and function of CD44 and its role in neoplasia.
Mol Pathol
 
1998
;
51
:
191
–200.
128
Schneider SS, Schick C, Fish KE, Miller E, Pena JC, Treter SD, et al. A serine proteinase inhibitor locus at 18q21.3 contains a tandem duplication of the human squamous cell carcinoma antigen gene.
Proc Natl Acad Sci U S A
 
1995
;
92
:
3147
–51.
129
Zou Z, Anisowicz A, Hendrix MJ, Thor A, Neveu M, Sheng S, et al. Maspin, a serpin with tumor-suppressing activity in human mammary epithelial cells.
Science
 
1994
;
263
:
526
–9.
130
Seftor RE, Seftor EA, Sheng S, Pemberton PA, Sager R, Hendrix MJ. Maspin suppresses the invasive phenotype of human breast carcinoma.
Cancer Res
 
1998
;
58
:
5681
–5.
131
Umekita Y, Hiipakka RA, Liao S. Rat and human maspins: structures, metastatic suppressor activity and mutation in prostate cancer cells.
Cancer Lett
 
1997
;
113
:
87
–93.
132
Pienta KJ, Partin AW, Coffey DS. Cancer as a disease of DNA organization and dynamic cell structure.
Cancer Res
 
1989
;
49
:
2525
–32.
133
Guo X, Friess H, Graber HU, Kashiwaga M, Zimmerman A, Korc M, et al. KAI1 expression is up-regulated in early pancreatic cancer and decreased in the presence of metastases.
Cancer Res
 
1996
;
56
:
4876
–80.
134
Guo XZ, Friess H, Di Mola FF, Heinecke JM, Abou-Shady M, Graber HU, et al. KAI1, a new metastasis suppressor gene, is reduced in metastatic hepatocellular carcinoma.
Hepatology
 
1998
;
28
:
1481
–8.
135
Yu Y, Yang JL, Markovic B, Jackson P, Yardley G, Barrett J, et al. Loss of KAI1 messenger RNA expression in both high-grade and invasive human bladder cancers.
Clin Cancer Res
 
1997
;
3
:
1045
–9.
136
Adachi M, Taki T, Ieki Y, Huang CL, Higashiyama M, Miyake M. Correlation of KAI1/CD82 gene expression with good prognosis in patients with non-small cell lung cancer.
Cancer Res
 
1996
;
56
:
1751
–5.
137
Uchida S, Shimada Y, Watanabe G, Li ZG, Hong T, Miyake M, et al. Motility-related protein (MRP-1/CD9) and KAI1/CD82 expression inversely correlate with lymph node metastasis in oesophageal squamous cell carcinoma.
Br J Cancer
 
1999
;
79
:
1168
–73.
138
Geradts J, Maynard R, Birrer MJ, Hendricks D, Abbondanzo SL, Fong KM, et al. Frequent loss of KAI1 expression in squamous and lymphoid neoplasms. An immunohistochemical study of archival tissues.
Am J Pathol
 
1999
;
154
:
1665
–71.
139
Chekmareva MA, Kadkhodaian MM, Hollowell CM, Kim H, Yoshida BA, Luu HH, et al. Chromosome 17-mediated dormancy of AT6.1 prostate cancer micrometastases.
Cancer Res
 
1998
;
58
:
4963
–9.
140
Holmgren L, O'Reilly MS, Folkman J. Dormancy of micrometastases: balanced proliferation and apoptosis in the presence of angiogenesis suppression.
Nat Med
 
1995
;
1
:
149
–53.
141
Welch DR, Goldberg SF. Molecular mechanisms controlling human melanoma progression and metastasis.
Pathobiology
 
1997
;
65
:
311
–30.
142
Welch DR, Chen P, Miele ME, McGary CT, Bower JM, Stanbridge EJ, et al. Microcell-mediated transfer of chromosome 6 into metastatic human C8161 melanoma cells suppresses metastasis but does not inhibit tumorigenicity.
Oncogene
 
1994
;
9
:
255
–62.
143
Miele ME, Jewett MD, Goldberg SF, Hyatt DL, Morelli C, Gualandi F, et al. A human melanoma metastasis-suppressor locus maps to 6q16.3-q23.
Int J Cancer
 
2000
;
86
:
524
–8.
144
Goldberg SF, Harms JF, Quon K, Welch DR. Metastasis-suppressed C8161 melanoma cells arrest in lung but fail to proliferate.
Clin Exp Metastasis
 
1999
;
17
:
601
–7.
145
Nelson PS, Ng WL, Schummer M, True LD, Liu AY, Bumgarner RE, et al. An expressed-sequence-tag database of the human prostate: sequence analysis of 1168 cDNA clones.
Genomics
 
1998
;
47
:
12
–25.
146
Hawkins V, Doll D, Bumgarner R, Smith T, Abajian C, Hood L, et al. PEDB: the Prostate Expression Database.
Nucleic Acids Res
 
1999
;
27
:
204
–8.
147
Bissell MJ, Weaver VM, Lelievre SA, Wang F, Petersen OW, Schmeichel KL. Tissue structure, nuclear organization, and gene expression in normal and malignant breast. Cancer Res 1999;59(7 Suppl):1757–1763s; discussion 1763s–1764s.
148
Boudreau N, Bissell MJ. Extracellular matrix signaling: integration of form and function in normal and malignant cells.
Curr Opin Cell Biol
 
1998
;
10
:
640
–6.
149
Hayward SW, Grossfeld GD, Tlsty TD, Cunha GR. Genetic and epigenetic influences in prostatic carcinogenesis (review).
Int J Oncol
 
1998
;
13
:
35
–47.
150
Baskin LS, Hayward SW, DiSandro MS, Li YW, Cunha GR. Epithelial–mesenchymal interactions in the bladder. Implications for bladder augmentation.
Adv Exp Med Biol
 
1999
;
462
:
49
–61.
151
Olumi AF, Grossfeld GD, Hayward SW, Carroll PR, Tlsty TD, Cunha GR. Carcinoma-associated fibroblasts direct tumor progression of initiated human prostatic epithelium.
Cancer Res
 
1999
;
59
:
5002
–11.
152
Chung LW, Davies R. Prostate epithelial differentiation is dictated by its surrounding stroma.
Mol Biol Rep
 
1996
;
23
:
13
–9.
153
Chung LW. The role of stromal-epithelial interaction in normal and malignant growth.
Cancer Surv
 
1995
;
23
:
33
–42.
154
Chung LW. Implications of stromal-epithelial interaction in human prostate cancer growth, progression and differentiation.
Semin Cancer Biol
 
1993
;
4
:
183
–92.
155
Welch DR. Technical considerations for studying cancer metastasis in vivo.
Clin Exp Metastasis
 
1997
;
15
:
272
–306.
156
Day ML, Foster RG, Day KC, Zhao X, Humphrey P, Swanson P, et al. Cell anchorage regulates apoptosis through the retinoblastoma tumor suppressor/E2F pathway.
J Biol Chem
 
1997
;
272
:
8125
–8.
157
Kendrew J. The thread of life; an introduction to molecular biology. Cambridge (MA): Harvard University Press; 1996.
158
Schrodinger E. What is life?: the physical aspect of the living cell; with, mind and matter: & autobiographical sketches. Cambridge paperbacks. Science. New York (NY): Cambridge University Press; 1992.
159
Coffey DS. Self-organization, complexity and chaos: the new biology for medicine.
Nat Med
 
1998
;
4
:
882
–5.
160
Schwab ED, Pienta KJ. Explaining aberrations of cell structure and cell signaling in cancer using complex adaptive systems.
Adv Molec Cell Biol
 
1997
;
24
:
207
–47.