Plasma osteopontin as a biomarker of prostate cancer aggression: relationship to risk category and treatment response.

BACKGROUND: High plasma osteopontin (OPN) has been linked to tumour hypoxia, metastasis, and poor prognosis. This study aims to assess whether plasma osteopontin was a biomarker of increasing progression within prostate cancer (PCa) prognostic groups and whether it reflected treatment response to local and systemic therapies. METHODS: Baseline OPN was determined in men with localised (n=199), locally recurrent (n=9) and castrate-resistant, metastatic PCa (CRPC-MET; n=37). Receiver-operating curves (ROC) were generated to describe the accuracy of OPN for distinguishing between localised risk groups or localised vs metastatic disease. We also measured OPN pre- and posttreatment, following radical prostatectomy, external beam radiotherapy (EBRT), androgen deprivation (AD) or taxane-based chemotherapy. RESULTS: The CRPC-MET patients had increased baseline values (mean 219; 56-513 ng ml(-1); P<0.0001) compared with the localised, non-metastatic group (mean 72; 12-438 ng ml(-1)). The area under the ROC to differentiate localised vs metastatic disease was improved when OPN was added to prostate-specific antigen (PSA) (0.943-0.969). Osteopontin neither distinguished high-risk PCa from other localised PCa nor correlated with serum PSA at baseline. Osteopontin levels reduced in low-risk patients after radical prostatectomy (P=0.005) and in CRPC-MET patients after chemotherapy (P=0.027), but not after EBRT or AD. CONCLUSION: Plasma OPN is as good as PSA at predicting treatment response in CRPC-MET patients after chemotherapy. Our data do not support the use of plasma OPN as a biomarker of increasing tumour burden within localised PCa.

Dynamic heterogeneity: rapid generation of metastatic variants in mouse B16 melanoma cells.

The ability of clonally derived lines of B16F1 and B16F10 melanoma cells to form experimental metastases in C57BL mice after intravenous injection was examined. Luria- Delbruck fluctuation analysis was applied to the results obtained with parallel subclones grown to small population sizes before testing for metastatic ability. The analysis demonstrated that variant cells capable of forming experimental metastases were generated in B16F1 cell populations at an effective rate of about 1.3 X 10(-5) per cell per generation while in B16F10 cell populations the effective rate of production was about 5 X 10(-5) per cell per generation. These results are consistent with a dynamic heterogeneity model of tumor progression. They suggest that the majority of cells in both lines are effectively nonmetastatic and that the higher metastatic ability of the B16F10 population may be due in part to a higher rate of generation of metastatic variants.

Genomic amplification of MET with boundaries within fragile site FRA7G and upregulation of MET pathways in esophageal adenocarcinoma.

Esophageal adenocarcinoma (EA) is characterized by a poor prognosis making the identification of clinically targetable proteins essential for improving patient outcome. We report the involvement of multiple alterations of the MET pathway in EA development and progression. Microarray analysis of Barrett's metaplasia, dysplasia, and EA revealed overexpression of the MET oncogene in EAs but only those with MET gene amplification. STS-amplification mapping revealed that the boundary of the MET amplicon in these EAs is defined by fragile site FRA7G. We also identified an amplicon at 11p13 that resulted in amplification and overexpression of CD44, a gene involved in MET autophosphorylation upon HGF stimulation. Tissue microarrays with phospho-MET-specific antibodies demonstrated a uniformly high abundance of MET activation in primary EA and cells metastatic to lymph nodes but to a lesser extent in a subset of metaplastic and dysplastic Barrett's samples. Increased expression of multiple genes in the MET pathway associated with invasive growth, for example, many MMPs and osteopontin, also was found in EAs. Treatment of EA-derived cell lines with geldanamycin, an inhibitor for tyrosine kinases including MET receptor kinase, reduced cell migration and induced EA cell apoptosis. The data indicate that upregulation of the MET pathway may contribute to the poor outcome of EA patients and that therapeutic agents targeting this pathway may help improve patient survival.

Cells transformed with a ts viral src mutant are temperature sensitive for in vivo growth.

Studies on ts mutants of avian sarcoma viruses have previously implicated the src gene product (pp60src) kinase function in in vitro transformation. The role of src in vivo, however, has not been clearly defined. Using a sensitive and quantitative assay that was developed in chicken embryos (Chambers et al., Cancer Res. 42:4018-4025, 1982), we tested the in vivo tumorigenic properties of cells transformed with LA23, an avian sarcoma virus that is temperature sensitive for in vitro transformation. We found that the in vivo growth ability of these cells was temperature sensitive and that this in vivo behavior correlated with the in vitro transformation behavior (growth in soft agar and saturation density).

Identification of a ras-activated enhancer in the mouse osteopontin promoter and its interaction with a putative ETS-related transcription factor whose activity correlates with the metastatic potential of the cell.

The role of RAS in transducing signals from an activated receptor into altered gene expression is becoming clear, though some links in the chain are still missing. Cells possessing activated RAS express higher levels of osteopontin (OPN), an alpha v beta 3 integrin-binding secreted phosphoprotein implicated in a number of developmental, physiological, and pathological processes. We report that in T24 H-ras-transformed NIH 3T3 cells enhanced transcription contributes to the increased expression of OPN. Transient transfection studies, DNA-protein binding assays, and methylation protection experiments have identified a novel ras-activated enhancer, distinct from known ras response elements, that appears responsible for part of the increase in OPN transcription in cells with an activated RAS. In electrophoretic mobility shift assays, the protein-binding motif GGAGGCAGG was found to be essential for the formation of several complexes, one of which (complex A) was generated at elevated levels by cell lines that are metastatic. Southwestern blotting and UV light cross-linking studies indicated the presence of several proteins able to interact with this sequence. The proteins that form these complexes have molecular masses estimated at approximately 16, 28, 32, 45, 80, and 100 kDa. Because the approximately 16-kDa protein was responsible for complex A formation, we have designated it MATF for metastasis-associated transcription factor. The GGANNNAGG motif is also found in some other promoters, suggesting that they may be similarly controlled by MATF.

Osteopontin expression in salivary gland tumours.

Osteopontin (OPN) is expressed in numerous carcinomas and plays a role in tumour development, invasion and metastasis. This study examines by immunohistochemistry the expression of OPN in normal salivary gland tissue and three types of salivary gland tumour: pleomorphic adenoma (PA), adenoid cystic carcinoma (ACC) and polymorphous low grade adenocarcinoma (PLGA). PAs and PLGAs demonstrated higher levels of OPN than normal salivary gland tissue, while ACC, although showing a trend towards increased OPN, was not significantly different. The results of this study indicate that OPN expression is present in normal salivary gland tissue, and is increased in certain salivary gland tumours, but further investigation is necessary to clarify its role.

Volume measurement variability in three-dimensional high-frequency ultrasound images of murine liver metastases.

The identification and quantification of tumour volume measurement variability is imperative for proper study design of longitudinal non-invasive imaging of pre-clinical mouse models of cancer. Measurement variability will dictate the minimum detectable volume change, which in turn influences the scheduling of imaging sessions and the interpretation of observed changes in tumour volume. In this paper, variability is quantified for tumour volume measurements from 3D high-frequency ultrasound images of murine liver metastases. Experimental B16F1 liver metastases were analysed in different size ranges including less than 1 mm3, 1-4 mm3, 4-8 mm3 and 8-70 mm3. The intra- and inter-observer repeatability was high over a large range of tumour volumes, but the coefficients of variation (COV) varied over the volume ranges. The minimum and maximum intra-observer COV were 4% and 14% for the 1-4 mm3 and <1 mm3 tumours, respectively. For tumour volumes measured by segmenting parallel planes, the maximum inter-slice distance that maintained acceptable measurement variability increased from 100 to 600 microm as tumour volume increased. Comparison of free breathing versus ventilated animals demonstrated that respiratory motion did not significantly change the measured volume. These results enable design of more efficient imaging studies by using the measured variability to estimate the time required to observe a significant change in tumour volume.

Changing views of the role of matrix metalloproteinases in metastasis.

Metastatic spread of cancer continues to be the greatest barrier to cancer cure. Understanding the molecular mechanisms of metastasis is crucial for the design and effective use of novel therapeutic strategies to combat metastases. One class of molecules that has been repeatedly implicated in metastasis is the matrix metalloproteinases (MMPs). In this review, we re-examine the evidence that MMPs are associated with metastasis and that they make a functional contribution to the process. Initially, it was believed that the major role of MMPs in metastasis was to facilitate the breakdown of physical barriers to metastasis, thus promoting invasion and entry into and out of blood or lymphatic vessels (intravasation, extravasation). However, recent evidence suggests that MMPs may have a more complex role in metastasis and that they may make important contributions at other steps in the metastatic process. Studies using intravital videomicroscopy, as well as experiments in which levels of MMPs or their inhibitors (tissue inhibitors of metalloproteinases [TIMPs]) are manipulated genetically or pharmacologically, suggest that MMPs are key regulators of growth of tumors, at both primary and metastatic sites. On the basis of this evidence, a new view of the functional role of MMPs in metastasis is presented, which suggests that MMPs are important in creating and maintaining an environment that supports the initiation and maintenance of growth of primary and metastatic tumors. Further clarification of the mechanisms by which MMPs regulate growth of primary and metastatic tumors will be important in the development of novel therapeutic strategies against metastases.

Clonal heterogeneity, experimental metastatic ability, and p21 expression in H-ras-transformed NIH 3T3 cells.

We examined individual clones of murine NIH 3T3 cells, transformed with the human bladder cancer (T24) H-ras oncogene, for p21 expression and for experimental metastatic ability in the immunodeficient chick embryo. We found that the clones were heterogeneous for both of these properties. In general, p21 expression was a good predictor of metastatic ability of the clones. Cells from poorly metastatic clones were passaged in the chick embryo metastasis assay to determine whether cells with increased metastatic ability could be selected. We found that the selected cells were more metastatic and that substantial increases in expression of p21 also accompanied this increase in metastatic ability. The relationship between p21 expression and metastatic ability appeared linear, with a high correlation coefficient (r = .85), suggesting that in this model system quantitative increases in metastatic properties can result from increased expression of the ras oncogene protein product p21.

Different patterns of gene expression in ras-resistant and ras-sensitive cells.

We have shown previously that nontumorigenic NIH 3T3 cells can be made tumorigenic and metastatic by transfection and expression of activated ras, whereas in LTA cells, which are tumorigenic but nonmetastatic, the degree of malignancy is not altered by ras. To investigate possible mechanisms of natural ras resistance, we compared the expression patterns of several genes thought to be involved in ras-induced metastatic progression in LTA (ras-resistant) and NIH 3T3 (ras-sensitive) cells, before and after constitutive expression of transfected T24-H-ras. We examined the expression of the nuclear "early-response" genes jun and fos and the "tumor-suppressor" retinoblastoma (Rb) gene, as well as genes involved in invasion (major excreted protein [MEP], tissue inhibitor of metalloproteinases [TIMP]), and cell adhesion (secreted phosphoprotein 1 [SPP1; also known as osteopontin]). We found distinct differences in both the basal and ras-induced levels of expression of most of these genes in LTA versus NIH 3T3 cells. High levels of MEP and low levels of TIMP were induced in ras-transfected NIH 3T3 cells, whereas LTA cells showed intermediate levels of MEP and high levels of TIMP that were only marginally affected by the expression of transfected ras. Similarly, SPP1 expression was strongly induced by ras in NIH 3T3 cells but was repressed by ras in LTA cells. Enzymogram assays for functional gelatinase activity showed an increase in 67-kd and 62-kd bands in NIH 3T3 cells in the presence of ras. LTA cells showed no gelatinolytic activity in the presence or absence of ras. Data from an in vitro assay for chemoinvasiveness showed a pattern as predicted from the expression of invasion-related genes; chemoinvasiveness in ras-transfected NIH 3T3 was greater than in LTA and ras-transfected LTA cells, which was greater than in NIH 3T3 cells. Differences in expression of the genes examined are believed to contribute to the ras responsiveness of NIH 3T3 cells and the ras resistance of LTA cells.

Early steps in hematogenous metastasis of B16F1 melanoma cells in chick embryos studied by high-resolution intravital videomicroscopy.

BACKGROUND: There are few techniques that permit direct observation of tumor metastasis. The ability to observe steps in this process as they occur in experimental animals would complement studies on molecular mechanisms. PURPOSE: We have developed a novel procedure using high-resolution intravital videomicroscopy to permit direct observation of cells as they arrest in the microcirculation, extravasate, and form micrometastases. We used this procedure to study early steps in experimental metastasis in immune-deficient chick embryos, permitting us to develop this technique in a relatively accessible respiratory organ and in the absence of host immune responses. Our goals were to develop techniques applicable to this host and to other hosts and to clarify the process of hematogenous tumor spread in this host. METHODS: We injected fluorescently labeled B16F1 melanoma cells into the circulation of 11- to 13-day chick embryos, and using intravital videomicroscopy, we observed the cells in the chorioallantoic membrane over time. RESULTS: The majority of injected cells were trapped initially in orifices to the chorioallantoic membrane capillary plexus or in tapering ends of arterioles leading to the plexus. During the first 2 hours, cells were found only in vessel lumina. After 8 hours, 83% of cells had extravasated, and the rest were in the process of extravasation. Cell shape changes and pseudopodial extensions were seen during extravasation and tumor development. Tumor cell division was seen only after extravasation. Tumors tended to develop near microvessels and were often wrapped around them. CONCLUSIONS: Intravital videomicroscopy can provide new information about steps in metastasis. This procedure is applicable to other hosts and can be used in future studies to test hypotheses about molecular mechanisms of tumor spread.

p53 mutation, expression, and DNA ploidy in evolving gliomas: evidence for two pathways of progression.

BACKGROUND: Two lines of evidence indirectly implicate the tumor suppressor p53 (also known as TP53) gene in glioma development. First, germline mutations of the p53 gene are associated with increased susceptibility to glioma. Second, chromosome 17p deletions and p53 gene mutations are found frequently in sporadic gliomas of all malignancy stages. These observations suggest that mutations of the p53 gene may be early events in glioma development. PURPOSE: Our purpose was to analyze 15 low-grade astrocytic gliomas that progressed to higher-grade gliomas, examining the status of the p53 gene in both the initial and recurrent tumors. Also, we explored the relationships between p53 status, DNA ploidy, tumor grade, and patient survival. METHODS: Fifteen low-grade gliomas that recurred as tumors of higher grade 17-102 months after initial treatment (biopsy, resection, radiotherapy, or chemotherapy) were identified from hospital records of patients (eight male and seven female) aged 31-68 years. Pathologic diagnosis was re-evaluated. Polymerase chain reaction (PCR)-single-strand conformation polymorphism and DNA sequencing were performed on tissue samples from the initial and recurrent tumors of each patient, using oligonucleotide PCR primers directed to exons 5-9 of the p53 gene. p53 expression was determined by immunohistochemistry and DNA ploidy evaluated by DNA flow cytometry. RESULTS: Eight (53%) of fifteen tumors had p53 mutations in exons 5-9. Nine (64%) of fourteen were immunopositive initially, and eight of these were also immunopositive at recurrence. p53 gene status was significantly associated with p53 expression in the initial tumor (P = .02), and p53 expression at initial diagnosis was significantly related to tumor pathology at recurrence (P = .03). Patients with p53 mutant tumors survived nearly twice as long as those without mutations (median survival, 61 versus 33 months; P = .031). There was no significant difference in recurrence-free survival between patients with p53 mutant and nonmutant tumors (48 versus 33 months; P = .37), but there was a significant difference in postrecurrence survival (17 versus 2 months; P = .019). CONCLUSION: Low-grade tumors that recurred as anaplastic gliomas were characterized by p53 gene mutation, immunopositivity, and DNA non-diploidy. Low-grade tumors that recurred as glioblastomas generally had intact p53 genes and were immunonegative. These findings suggest that histologically indistinguishable, low-grade astrocytic gliomas that are destined to progress to higher grades, do so along two distinct clinicopathologic pathways (either stepwise to anaplastic glioma, then glioblastoma, or directly to glioblastoma) marked by the presence or absence of p53 mutation.

Selection for experimental metastatic ability of heterologous tumor cells in the chick embryo after DNA-mediated transfer.

The chick embryo is an immune-deficient host able to support growth of a wide variety of transformed cells. Since growth of normal cells is not observed, this system appears to be generally useful for investigating malignant properties of different cells. Recently, we developed a sensitive assay to quantitate and select for rodent cells able to survive and grow in embryonic chick organs following i.v. injection (Cancer Res., 42: 4018-4025, 1982). We envisage this assay as a model system for studying aspects of the metastatic process. We have used DNAs from murine and human melanoma cell lines (which grow well in chick embryos after i.v. injection) to transfect murine LTA cells (which do not grow in chicks after i.v. injection). From the transfected LTA cells, we were able to isolate clones which grow well in the chick after i.v. injection. Such clones were not observed in untransfected LTA cells or with LTA cells transfected with LTA DNA. These experiments clearly demonstrate the feasibility of using the chick embryo as a host system to study genes involved in growth control alteration of the sort seen in malignant transformation.

Experimental metastatic ability of H-ras-transformed NIH3T3 cells.

We have used a quantitative "experimental" metastasis assay in the embryonic chick, an immunodeficient host, to examine in vivo growth properties of ras oncogene-transformed NIH3T3 cells. We found that two independently derived populations of NIH3T3 cells that had been morphologically transformed with the T24 human H-ras oncogene were able to grow in vivo following i.v. injection. Nontransformed control NIH3T3 cells with normal morphology did not grow in this assay. Spontaneously arising morphological transformants from control NIH3T3 cell populations were also tested and did not grow in this assay. We conclude that the H-ras gene can confer experimental metastatic ability on nonmetastatic NIH3T3 cells, that the ras gene alters the cells in some way beyond in vitro morphological transformation, and thus that the in vitro transformation assay detects only part of the malignant phenotype of these cells.

Wild-type p53 renders mouse astrocytes resistant to 1,3-Bis(2-chloroethyl)-1-nitrosourea despite the absence of a p53-dependent cell cycle arrest [corrected].

We examined the response of normal and p53-deficient mouse astrocytes to the alkylating agent 1,3-bis(2-chloroethyl)-l-nitrosourea (BCNU), a clinically useful DNA-damaging drug to which some human astrocytomas are resistant and some are sensitive. Astrocyte cultures were isolated from the cerebrums of wild-type, heterozygous, and knockout p53 neonatal mice and treated with various concentrations of BCNU. Wild-type p53 astrocytes were significantly more resistant to BCNU than were knockout p53 astrocytes, with heterozygous astrocytes exhibiting an intermediate level of resistance, Cell cycle analysis showed that wild-type p53 astrocytes treated with BCNU demonstrated a decline in the percentage of cells in G1 and an increase in the percentage of cells in G2. Similar cell cycle responses to BCNU occurred in knockout p53 astrocytes, suggesting that this effect was p53 independent. In contrast, G1 arrest was observed in wild-type astrocytes exposed to ionizing radiation and was not observed in knockout astrocytes, indicating a p53-dependent response. Our findings point to an as yet uncharacterized p53-associated mechanism of resistance to BCNU in mouse astrocytes.

A model system for studying metastasis using the embryonic chick.

An assay capable of recovering individual viable rodent cells localized in various organs of the chick embryo is described. This assay is based on the differential sensitivity of chick and rodent cells to the cytotoxic drug ouabain. Utilizing this assay, the potential of the chick embryo as a model system for studying metastasis was examined. Several cell lines were characterized in three ways: (a) ability to form local tumors after cell application onto the chorioallantoic membrane; (b) ability to form macro- or microscopic metastasis in the embryo from chorioallantoic membrane tumors; (c) experimental metastatic ability following i.v. injection into chorioallantoic membrane veins. These results were compared with the results obtained from the ouabain-plating assay. We conclude that this assay permits detection of viable metastatic cells even when tumors cannot be detected and helps to overcome the time constraints that have, in the past, limited the usefulness of the chick embryo in modeling metastasis.

Tumor heterogeneity and stability of the metastatic phenotype of mouse KHT sarcoma cells.

Heterogeneity in metastatic ability has been demonstrated in model systems for in vitro-cloned cell lines for a number of different tumors. We have examined the clonal diversity of mouse KHT sarcoma cells cloned either in vitro or in vivo by determining their ability to form lung colonies following i.v. injection into syngeneic mice. A wide range of metastatic ability was found in both the in vitro- and in vivo-isolated clones, suggesting that the diversity observed is not due to any selection occurring during in vitro growth. The stability of four in vitro-isolated clones, two of high metastatic and two of low metastatic ability, was then studied over a period of 3 to 4 months of growth in vitro. The phenotype of the highly metastatic cells remained relatively stable, declining only slightly over time. The clones with low metastatic ability, however, demonstrated a significant increase in ability to form lung colonies over the first 30 days in culture before becoming stable at levels approximately 10-fold higher than their original values. Even after this increase, however, there remained a difference of about a factor of 10 in the metastatic ability of the high and low pairs of clones. It was found that the number of lung colonies formed by all four cell lines was significantly increased when plastic microspheres were injected with the cell suspension. The cells with low metastatic ability were affected to a greater degree by the microspheres, resulting in the elimination of the difference between the four clones after 30 days in culture. This result suggests that the use of microspheres may provide a means to distinguish different cellular properties which affect the ability of cells to form lung metastases.

Adhesion of metastatic, ras-transformed NIH 3T3 cells to osteopontin, fibronectin, and laminin.

We previously reported that H-ras-induced metastatic ability in murine NIH 3T3 cells is accompanied by increased expression of osteopontin (OPN). OPN is a secreted phosphoprotein that contains a GRGDS amino acid sequence, suggesting adhesive function, but the function of OPN in tumor cells remains poorly understood. Here we report that PAP2 cells (ras-transformed, metastatic NIH 3T3 cells) adhere and spread on OPN-coated substrates, while NIH 3T3 cells adhere and spread poorly on OPN. A similar pattern was seen for adhesion to laminin, while both cell lines adhered equally well to fibronectin. Adhesive interactions to OPN, laminin, and fibronectin were specific and were blocked by GRGDS (but not control GRGESP) peptides. The kinetics of adhesion to all three substrates was examined. Maximum adhesion was observed at 30-60 min, with reduced adhesion thereafter. We also purified metabolically labeled [32P]OPN secreted by PAP2 cells. Labeled OPN bound better in solution to PAP2 cells than to NIH 3T3 cells, and binding to both cell lines was blocked by GRGDS peptides, results that are consistent with the adhesion and spreading of these cells to OPN-coated substrates. Malignant PAP2 cells thus not only secrete increased levels of OPN, relative to NIH 3T3 cells, but also adhere better to this protein. While the target of OPN secreted by tumor cells is not known, our results raise the possibility that tumor cells that secrete OPN may also bind this protein and that this binding may function in autocrine-type signal transduction important to malignancy.

Differential expression of drug resistance genes and chemosensitivity in glial cell lineages correlate with differential response of oligodendrogliomas and astrocytomas to chemotherapy.

The two principal subtypes of glial neoplasms, astrocytomas and oligodendrogliomas, exhibit striking differences in response to chemotherapy. This differential chemosensitivity might be explained by the specific genetic alterations causing gliomas but could also be attributable to specific properties intrinsic to the cells from which gliomas arise. To examine the possibility that chemosensitivity might be associated with lineage-specific properties of potential ancestors of these tumors, we explored: (a) the expression of drug resistance genes in rat glial cells; (b) the sensitivity of rat glial subtypes to the bifunctional alkylating agent, 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU); and (c) the effect of O6-methylguanine-DNA methyltransferase (MGMT) and glutathione modulation on resistance to BCNU. Astrocytes, O-2A progenitors, and oligodendrocytes each displayed a unique pattern of expression of six drug resistance genes: MGMT, GST mu, GST pi,p53, MDR, and MT. Oligodendrocytes were more sensitive to BCNU than either astrocytes or O-2A progenitors. The increased resistance of astrocytes in comparison to oligodendrocytes was modulated, at least in part, by both O6-benzylguanine (BG) and DL-buthionine-(S,R)-sulfoximine, suggesting a role for both MGMT and glutathione in the resistance of astrocytes to BCNU. The sensitivity of O-2A progenitors to BCNU following BG pretreatment is virtually indistinguishable from that of oligodendrocytes depleted of MGMT, suggesting that the down-regulation of MGMT is sufficient to account for the increased sensitivity of oligodendrocyte lineage cells to BCNU as they differentiate. These experiments provide support for the hypothesis that properties of glial cells retained in gliomas may contribute to the differential chemosensitivity of glial neoplasms.

Reduced malignancy of ras-transformed NIH 3T3 cells expressing antisense osteopontin RNA.

Osteopontin (OPN) is a secreted, calcium-binding phosphoprotein that frequently has been associated with the transformed phenotype. To clarify the function of OPN in tumor cells, we designed experiments to: (a) express antisense OPN RNA in murine PAP2 cells (metastatic, ras-transformed NIH 3T3 cells) and (b) examine the effects of antisense OPN expression on the tumorigenic and metastatic properties of the cells. PAP2 cells were transfected with pNMH-asOPN, an inducible, mammalian expression vector that can generate antisense OPN RNA complementary to the OPN mRNA. Two clones have been identified that expressed antisense OPN RNA in vitro. While reduced OPN protein secretion was not detected when the cells were grown in vitro, the in vivo expression of antisense OPN RNA was associated with reduced tumorigenicity. Tumors that did arise, with greatly extended lag time, had lost expression of antisense OPN RNA in vivo, suggesting that antisense OPN RNA expression was associated with reduced tumorigenicity of these cells.