Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Thorac Oncol. Author manuscript; available in PMC 2011 October 1.
Published in final edited form as:
PMCID: PMC3122268

Functional Heterogeneity of Osteopontin Isoforms in Non-Small Cell Lung Cancer



Osteopontin (OPN) is a multifunctional protein with an important but poorly understood role in non-small cell lung cancer (NSCLC) pathogenesis. Moreover, the role of the three known mRNA isoforms (OPNa, OPNb, and OPNc) has not been reported. We hypothesize that OPN isoforms play different roles in determining the metastatic potential of NSCLC.


We amplified mRNA for each OPN isoform in NSCLC tumors and matched normal lung. The functional impact of each isoform was evaluated by transfecting cDNA plasmids specific to each isoform into NSCLC cell lines and comparing behavior to empty vector controls in scratch closure, cell proliferation, soft-agar colony formation and Matrigel™ invasion assays. Gene array was used to evaluate differences in down-stream targets and was compared to a panel of markers epithelial-mesenchymal transition (EMT).


OPNa expression was increased in 91% of NSCLC tumors compared to matched lung. OPNa overexpression significantly increased activity in scratch closure, proliferation, soft-agar colony formation and Matrigel™ invasion assays compared to controls in all cell lines. OPNb overexpression produced a less significant modulation of function. OPNc overexpression significantly decreased activity in proliferation, colony formation and invasion assays compared to controls. Expression arrays revealed an increase in EMT with OPNa overexpression, but not OPNc. Differences were validated by quantitative RT-PCR.


Overexpression of the individual OPN isoforms in NSCLC results in divergent functional phenotypes. OPNa produced an aggressive phenotype while OPNc produced a more indolent phenotype. Exon 4 which is transcribed in OPNa but absent in OPNc may be central to this phenomenon and could serve as a target for isoform-specific inhibition of OPN in NSCLC.

Keywords: osteopontin, non-small cell lung cancer, mRNA isoforms, tumor progression, epithelial–mesenchymal transition


Osteopontin (OPN) is a multifunctional protein identified in a remarkable range of normal and pathologic contexts. It plays a key role in the mediation of immune cell recruitment, wound healing and tissue remodeling.1, 2 OPN’s diverse biologic functions involve cell adhesion, migration and invasion.3 It has two highly preserved central binding domains, the RGD and SVVYGLR sequences. OPN is a ligand for the αvβ integrins, and mediates adhesive cell-matrix interactions and signal transduction pathways through binding at these locations.

The importance of OPN in tumor dissemination is highlighted in gene transfer experiments, where transfection of OPN increases malignant phenotype4 and OPN knock-out with antisense oligonucleotides decreases malignant potential.5, 6 Structural heterogeneity may be responsible for varying functions of OPN from different cellular sources. Alternative gene splicing may provide an explanation for its diversity. Three distinct OPN cDNA isoforms have been identified in humans, but little is known about their function.7 Osteopontin-a (OPNa) represents the full-length cDNA, while osteopontin–b (OPNb) has a deletion at exon 5 and osteopontin-c (OPNc) has a deletion at exon 4 (Fig. 1).8

Figure 1
This is a schematic representation of the three human cDNA splice variants of OPN and GeneBank accession numbers. OPNa has six translated exons while OPNb has a deletion of exon five and OPNc has a deletion of exon four. The PCR primers amplify the region ...

Osteopontin is overexpressed by immunohistochemistry in non-small cell lung cancer (NSCLC) tumors compared to surrounding normal lung.9 NSCLC cell lines which overexpress OPN have increased in vitro migration and invasion.9 Elevated tumor OPN levels are associated with increased stage, lymph node involvement and poor long-term survival.911 These data indicate an important role for OPN in determining the metastatic potential in NSCLC; but, reports of OPN expression in clinical NSCLC samples and tumor models have not reported which isoforms are expressed. We hypothesize that the individual isoforms are not equally expressed in NSCLC tumors and have significantly different roles in determining the metastatic potential in NSCLC through alteration in downstream gene targets.

Materials and Methods

Tumor specimens and RNA isolation

We evaluate m RNA expression of the OPN isoforms in 43 resected NSCLC tumors and matched normal lung. Tissue specimens were collected from consented patients undergoing resection of NSCLC at Wayne State University (WSU) from 1998 to 2005, with Institutional Review Board (IRB) approval. The specimens were transferred to NYU School of Medicine under a cooperative materials transfer agreement and the NYU IRB granted approval for use of de-identified specimens for biomarker discoveries. Tumor and normal tissues were homogenized and mRNA was isolated using RNEasy mini kit from Qiagen™ (QIAGEN, Valencia, CA).

NSCLC cell lines, H358, A549 and H460 were obtained from American Type Tissue Collection (Manassas, VA) and maintained in culture in Dulbecco's Modified Eagle Medium (DMEM) (Invitrogen, Grand Island, NY) with 10% fetal bovine serum (FBS) (Invitrogen, Grand Island, NY) at 37° and 5% CO2. Prior evaluation had determined that H358 had minimal native OPN mRNA expression and protein secretion, while A549 and H460 had high native mRNA expression of OPNa and OPNb and high protein secretion 12

RT-PCR primers were designed based upon the three NCBI GeneBank sequences to be inclusive of the regions of exon-4 and exon-5 at the N-terminus which are deleted from OPNc and OPNb respectively. The resulting primers (Table 1), amplify a 305 bp product from OPNa, a 263 bp product from OPNb, and a 224 bp product from OPNc in a single reaction, which are separated by electrophoresis (Fig. 1). RT-PCR amplification was carried out using Invitrogen PCR buffer system (Invitrogen, Carlsbad, CA). The reactions were performed with denaturing for 15 seconds at 94°, annealing for 30 seconds at 58°, and elongation for 30 seconds at 70° for 40 cycles; and final extension at 72° for 5 minutes. RT-PCR products were analyzed by Tris-acetate EDTA agarose (2.5% w/v) gel electrophesis. The amplification of peptidal prolyl isomerase A (PPIA) served as controls for equal loading and integrity of RT-PCR for this and all reactions.

Table 1
Primers used for Semi-quantitative PCR analysis.

OPN isoform expression constructs and transfection

pCMV2-based plasmid constructs expressing OPNa (NM_001040058), OPNb (NM_00058), OPNc (NM_001040060) and empty vector for control (Origene, Rockville, MD) were complexed with lipofectamine 2000 (Invitrogen, Carlsbad, CA) for transfection into NSCLC cell lines. Proof of transfection was evaluated by RT-PCR from a pooled population of cells. Pooled populations of transfected cells were used in scratch closure, cell proliferation, soft-agar colony formation, and Matrigel™ invasion assays. Comparisons were made between cells transfected with each isoform and empty pcmv vector. Assays were performed in triplicate, and the data were collected independently and expressed as mean ± SE. A two-sided Student’s t-test was used to evaluate data and differences were regarded as significant at p ≤ 0.05.

Scratch closure assay

2 × 105 transfected cells were plated in 3ml of serum-free DMEM in 6-well plates. A scratch was made through the bottom of the plate using a 1 ml sterile pipette tip 24 hours later. Plates were washed with PBS and the undersides of plates were marked. The scratch width was measured at six locations. Plates were incubated for 24 hours and scratch width measured at the same six points. Closure was calculated as a percent of initial scratch width.

Cell proliferation assay

3 × 103 transfected cells in 100µl of DMEN/10% FBS were plated in 96-well flat bottom plates in six replicates and incubated for 48 hours. 20µl of CellTiter-Blue reagent (Promega, Madison, WI) was added to each well and incubated for one hour. Optical density was measured at 560/590λ using the Universal Reader Victor3 (PerkinElmer, Waltham, MA) and compared to standard curve derived for each cell line.

Soft agar colony formation

Base agars were made by combing equal volumes of 1% agarose and 2×DMEM/10% FBS, both at 40°. 1.5ml of warm base agar was plated into 6-well plates and allowed to polymerize. The top agar was produced by combing equal volumes 5 ×103 transfected cells/ml in 1×DMEM/10% FBS with warm 0.7% agarose. 2 ml of the agarose cell mixture was placed on top of each base coat. Plates were incubated at 37°, 5% CO2. At 21 days colonies were stained with 0.5 ml of Crystal violet and counted under a dissecting microscope.

Matrigel™ invasion assay

The BD Biocoat Matrigel™ invasion chambers were used in accordance with the manufacturer’s protocol (BD Sciences, Franklin Lakes, NJ). 1 × 105 transfected cells were seeded onto 8µM pore polycarbonate membrane inserts coated with a thin layer Matrigel™ basement membrane matrix diluted 1:100 in PBS. Plates were incubated for 48 hours. Cells that migrated through the Matrigel™ to the lower surface were fixed and stained with Giemsa solution and counted under microscope at 10X amplification. Ten fields from each membrane were counted.

Gene Expression Analysis

mRNA extracted from the pooled population of transfected cells used in the functional assays was analyzed using Affymetrix U133 plus-2 chips (Affymetrix Inc, Santa Clara, CA). Each probe was scanned twice (Hewlett-Packard Gene Array Scanner), the images overlaid, and the average intensities of each probe compiled and converted into expression file format using Affymetrix™ Gene Chip Operating System. We evaluated gene expression profiles in the A549 cell line with overexpression of OPNa and OPNc isoforms and empty vector controls in duplicate. Raw GeneChip data was normalized at the probe level using the robust multichip average algorithm,13 and filtered using Genespring 7.2 (Aligent Technologies, Palo Alto, CA). The differentially abundant mRNA between experimental cell lines and controls were filtered using an intersection of t-test results with p-value of 0.05 and false discovery rate set at 5%. Array results were compared against a panel of common downstream markers for epithelial-mesenchymal transition (EMT) to correlate functional results with downstream gene pathways.14 RT-PCR was performed using parameters described above to validate expression of snail-2, TGFb1-r, N-cadherin, vimentin, desmoplakin, cytokeratin -20, and e-cadherin. Primers are listed in Table 1.

OPN silencing with sh-RNA

sh-RNA plasmid 5’-GACCTGCCAGCAACCGAAG-3’ (Geneocopoeia, Rockville, MD) were complexed with lipofectamine 2000 (Invitrogen, Carlsbad, CA) for transfection into NSCLC cell lines. Proof of knock down was evaluated by RT-PCR from a pooled population of cells.

Real-time quantitative PCR analysis

The semi-quantitative expression levels were validated with quantitative PCR in four targets, TGFb1-r, N-cadherin, desmoplakin and cytokeratin-20. Super Script III first strand synthesis super mix (Invitrogen, Carlsbad, CA) was used to synthesize first strand cDNA. PCR was performed using 7900HT Fast Real-Time PCR System (Applied Biosystems, Carlsbad, CA). Samples were assayed in 10 µl-reaction mixture containing 2 µl cDNA, 0.5 µl of 5 µM gene specific primers (Table 2), 5 µl of 2 X SYBR green master mixes and 2 µl of molecular grade H2O. ACTB was used as a normalization control. The amplifications were performed for 40 cycles with annealing at 60°C for 1 minute. Samples were run in triplicates including negative controls. Relative quantification (2−ΔΔCT method) was performed to determine the change in gene expression levels.15

Table 2
Primers used for quantitative PCR analysis.


OPN mRNA expression is increased in 91% NSCLC tumors

OPN was overexpressed in the NSCLC tumors compared to the matched normal lung in 39 of 43 pairs (90.7%) and the dominant isoform expressed in all tumor samples was OPNa (Fig. 2). OPNc expression was detected in normal lung, but not in any of the tumor samples. The PCR products from OPNa from NSCLC tumor specimens were sequenced and match the OPNa NCBI GeneBank sequence NM_001040060.

Figure 2
OPN mRNA expression from NSCLC specimens (T) and paired normal lung (N) using primers which amplify each of the three OPN isoforms. OPN was overexpressed in 39 out of 43 (90.7%) of NSCLC specimens compared to matched normal lung and in all but specimen ...

Functional heterogeneity with individual OPN isoform overexpression in NSCLC cell lines

In all three cell lines, OPNa overexpression significantly increased activity in scratch closure and proliferation assays at 24 hours. OPNa overexpression significantly increased the size and number of colonies in soft-agar, and significantly increased the rate of invasion through Matrigel™ compared to controls (Fig 3a–d). Overexpression of OPNb produced a less significant modulation of function compared to OPNa in all assays and all cell lines (Fig. 3a–d). OPNc overexpression had no impact compared to control in scratch closure (Fig. 3a), but significantly decreased activity in proliferation, soft-agar colony formation, Matrigel™ invasion compared to controls (Fig. 3b–c). When Matrigel™ invasion results were normalized to differential rates of proliferation, the differences remained significant (Fig 3e).

Figure 3Figure 3Figure 3Figure 3Figure 3
Bar graphs depiction of results from functional assays with overexpression of the individual OPN isoforms in NSCLC cell lines. H358 (low native OPN expression) and A549 and H460 (high native OPN expression), transfected with empty vector for control or ...

OPNa overexpression induces EMT

We examined gene expression data from the A549 cell line with OPNa and OPNc overexpression compared to pcmv controls and compared it to a standard panel of makers associated with EMT16. OPNa overexpression resulted in a relative increase in expression of MMP-2, Snail -1, Snail-2, TGFb1-r, MMP-9, N-cadherin, ILK, vimentin, SOX-8 and SOX-9 and a relative decrease in the expression of cytokeratin 18, and 20, desmoplakin and E-cadherin, all consistent with activation of EMT pathways (Fig. 4a). We noted no significant change in expression of Twist, goosecoid, FoxC-2, Sox 10 or MMP-3. OPNc overexpression resulted in relative increase in expression of MMP-2, MMP-9 and Snail-1, but a relative decrease in Snail-2, TGFb1-r, N-cadherin, ILK, vimentin, Sox-8, Sox -9 and fibronectin. Cytokeratin 18, and occludin were down regulated but cytokeratin 20 and desmoplakin were up regulated, overall a far less consistent picture for EMT activation with OPNc overexpression. Relative differences in expression of the markers between OPNa and OPNc overexpressing cell lines and controls were analyzed by semi-quantitative RT-PCR and were similar across all three cell lines (Fig. 5).

Figure 4Figure 4
Waterfall depiction of changes in relative expression of a series of gene products associated with EMT with overexpression of OPN isoforms compared to controls. a) Overexpression of OPNa results in a pattern of gene regulation consistent with activation ...
Figure 5
Semi-quantitative PCR analysis of differences in relative expression in series of EMT markers discovered in gene arrays. Expression changes are validated in three NSCLC cell lines, each with overexpression of the of individual OPN isoforms. OPNa overexpression ...

Changes in expression of four markers were validated by quantitative PCR. CT-values for TGFb1-r and N-cadherin were higher in H358 controls than in H460 and A549, and were significantly decreased in all three cell lines with the overexpression of OPNa and increased with overexpression of OPNc, consistent with gene array expression data.(Fig 6a) Baseline CT-values for desmoplakin and cytokeratin-20 were lower in the H358 cell line than in H460 and A549 and were significantly increased with OPNa expression and decreased with OPNc expression across all cell lines.(Fig 7a)

Figure 6Figure 6
Quantitative PCR results validating difference in expression of TGFb1-r and N-cadherin. a) Bar graphs depict relative changes in mRNA levels of TGFb1-r and N-cadherin normalized to controls. Overexpression of OPNa resulted in a significant increase in ...
Figure 7Figure 7
Quantitative PCR results validating difference in expression of desmoplakin and cytokeratin-20. a) Bar graphs depict relative changes in mRNA levels of desmoplakin and cytokeratin-20 normalized to controls. Overexpression of OPNa resulted in a significant ...

Inhibition of native OPN expression by sh-RNA resulted in minimal changes in mRNA levels of TGFb1-r, N-cadherin, and desmoplakin in H358, the cell line with minimal native OPNa expression, but resulted in a significant decrease in TGFb1-r and N-cadherin mRNA levels and significant increase in desmoplakin and cytokeratin-20 in the A549 and H460 cell lines, each with high native OPNa expression. (Fig 6b and and7b7b).


Lung cancer is the world's leading cause of cancer death. A better understanding of the molecular pathogenesis is necessary to detect new biomarkers, develop new treatment strategies and improve survival. OPN is a ubiquitous protein whose functions center on cell migration and invasion. Hu noted an association between OPN overexpression and aggressive NSCLC phenotypes with increased in vitro cell invasion.9 The data presented here is unique because it identifies a single putative isoform, which is uniquely overexpressed and tied to malignant behavior. OPNa, the full length isoform, produced a more aggressive phenotype with increased migration, proliferation, anchorage independent growth and invasion when overexpressed. In contrast OPNc, which is defined by the deletion of exon 4 confers the opposite effect when overexpressed in NSCLC cell lines, decreasing proliferation, anchorage independent growth and invasion compared to controls. Importantly, OPNa is the isoform preferentially up regulated in NSCLC tumors (Fig. 2) and cell lines.12 The rate of overexpression is consistent with others who have noted increased OPN mRNA expression in 80% NSCLC tumors compared to normal lung 17, and recognize OPN as one of the most consistently overexpressed genes in NSCLC cDNA libraries18.

Heterogeneous expression of the OPN isoforms has been noted in documented in human glioma and breast cancer cells.8, 19 In breast cancer OPNc is a unique marker for malignant tissue which is not seen in normal breast tissue, and correlates with tumor grade.20 A varied role for the isoforms is identified in breast cancer, where in vitro experiments demonstrate enhanced anchorage independent growth and a lack of calcium-dependant aggregation and cell adhesion with the incorporation of OPNc but not OPNa.21 This support the hypothesis that the isoforms have differing functions in malignancy.

This work also correlates the functional findings with gene array data where an increase in EMT markers was noted with OPNa overexpression, but not OPNc. EMT enables benign tumor cells the capacity to infiltrate surrounding tissue and ultimately metastasize to distant sites and is implicated in the conversion of early-tumors to invasive malignancies 22. The molecular basis of EMT involves changes in the expression, distribution and function of multiple proteins including integrins, vimentin, MMPs and cadherins 14.

The cadherins are a family of calcium-dependent cell adhesion molecules critical for the development and maintenance of epithelial architecture. E-cadherin maintains cell to cell interactions and its loss is a hallmark for metastatic progression of solid tumors.14 Conversely, N-cadherin stimulates tumor invasiveness and progression by providing tumor cells the ability to permeate tissues. It promotes adhesive interactions with host stroma and endothelium that facilitates tumor dissemination.23 Overexpression of OPNa in NSCLC cell lines resulted in the classic switch in cadherin class: loss of e-cadherin and gain of N-cadherin, which has been closely linked to both EMT and tumor progression.24 Knockdown of endogenous OPNa by sh-RNA, reversed the cadherin switch, increasing levels of E-cadherin and decreasing N-cadherin. Overexpression of OPNc, did not produce the same classic downstream pattern consistent with increased EMT and malignant potential.

The observations that OPNa overexpression in NSCLC increased malignant behavior in vitro and EMT pathways compared to controls coupled with the finding of a decrease in the same properties with OPNc overexpression is fortuitous since the only difference between OPNa and OPNc is the transcription of exon 4, a 27 amino acids sequence in the amino terminus of the protein. Although OPN's two centrally preserved integrin binding domains (RGD and SVVYGLR) are essential to its activity, these data implicate exon 4 as an important regulator in the malignant potential of NSCLC. This region should contain the structure that dictates function relevant to at least some of the demonstrated malignant properties conferred by OPNa in NSCLC. The gene array data provide important information on downstream gene targets and molecular pathways which are responsible for this differential function. Targeted small molecules and monoclonal antibody therapies are highly successful treatment programs in oncology and the structural similarities between the isoforms may make OPN an ideal target for such treatment strategies in NSCLC.


This project is supported in part by funds from the Early Detection Research Network NCI/NIH grant 5 U01 CA111295-02, an American Thoracic/Lungevity Lung Cancer research grant, the NIH-Loan Repayment Program, the Stephen Banner Lung Cancer Foundation, and grant 1UL1RR029893 from the National Center for Research Resources, NIH.


Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.


1. Ashkar S, Weber GF, Panoutsakopoulou V, et al. Eta-1 (osteopontin): an early component of type-1 (cell-mediated) immunity. Science. 2000;287(5454):860–864. [PubMed]
2. Liaw L, Birk DE, Ballas CB, Whitsitt JS, Davidson JM, Hogan BL. Altered wound healing in mice lacking a functional osteopontin gene (spp1) J Clin Invest. 1998;101(7):1468–1478. [PMC free article] [PubMed]
3. Das R, Philip S, Mahabeleshwar GH, Bulbule A, Kundu GC. Osteopontin: it's role in regulation of cell motility and nuclear factor kappa B-mediated urokinase type plasminogen activator expression. IUBMB life. 2005;57(6):441–447. [PubMed]
4. Denhardt DT, Guo X. Osteopontin: a protein with diverse functions. Faseb J. 1993;7(15):1475–1482. [PubMed]
5. Behrend EI, Craig AM, Wilson SM, Denhardt DT, Chambers AF. Reduced malignancy of ras-transformed NIH 3T3 cells expressing antisense osteopontin RNA. Cancer Res. 1994;54(3):832–837. [PubMed]
6. Gardner HA, Berse B, Senger DR. Specific reduction in osteopontin synthesis by antisense RNA inhibits the tumorigenicity of transformed Rat1 fibroblasts. Oncogene. 1994;9(8):2321–2326. [PubMed]
7. O'Regan A, Berman JS. Osteopontin: a key cytokine in cell-mediated and granulomatous inflammation. Int J Exp Pathol. 2000;81(6):373–390. [PubMed]
8. Young MF, Kerr JM, Termine JD, et al. cDNA cloning, mRNA distribution and heterogeneity, chromosomal location, and RFLP analysis of human osteopontin (OPN) Genomics. 1990;7(4):491–502. [PubMed]
9. Hu Z, Lin D, Yuan J, et al. Overexpression of osteopontin is associated with more aggressive phenotypes in human non-small cell lung cancer. Clin Cancer Res. 2005;11(13):4646–4652. [PubMed]
10. Shijubo N, Uede T, Kon S, et al. Vascular endothelial growth factor and osteopontin in stage I lung adenocarcinoma. Am J Respir Crit Care Med. 1999;160(4):1269–1273. [PubMed]
11. Donati V, Boldrini L, Dell'Omodarme M, et al. Osteopontin expression and prognostic significance in non-small cell lung cancer. Clin Cancer Res. 2005;11(18):6459–6465. [PubMed]
12. Blasberg JD, Goparaju CM, Pass HI, Donington JS. Lung cancer osteopontin isoforms exhibit angiogenic functional heterogeneity. The Journal of thoracic and cardiovascular surgery. 2009 [PMC free article] [PubMed]
13. Irizarry RA, Bolstad BM, Collin F, Cope LM, Hobbs B, Speed TP. Summaries of Affymetrix GeneChip probe level data. Nucleic acids research. 2003;31(4):e15. [PMC free article] [PubMed]
14. Lee JM, Dedhar S, Kalluri R, Thompson EW. The epithelial-mesenchymal transition: new insights in signaling, development, and disease. The Journal of cell biology. 2006;172(7):973–981. [PMC free article] [PubMed]
15. Livak KJ, Schmittgen TD. Methods. 4. Vol. 25. San Diego, Calif: 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method; pp. 402–408. [PubMed]
16. Le QT, Chen E, Salim A, et al. An evaluation of tumor oxygenation and gene expression in patients with early stage non-small cell lung cancers. Clin Cancer Res. 2006;12(5):1507–1514. [PubMed]
17. Zhang J, Takahashi K, Takahashi F, et al. Differential osteopontin expression in lung cancer. Cancer Lett. 2001;171(2):215–222. [PubMed]
18. Bogenrieder T, Herlyn M. Axis of evil: molecular mechanisms of cancer metastasis. Oncogene. 2003;22(42):6524–6536. [PubMed]
19. Saitoh Y, Kuratsu J, Takeshima H, Yamamoto S, Ushio Y. Expression of osteopontin in human glioma. Its correlation with the malignancy. Lab Invest. 1995;72(1):55–63. [PubMed]
20. Mirza M, Shaughnessy E, Hurley JK, et al. Osteopontin-c is a selective marker of breast cancer. International journal of cancer. 2008;122(4):889–897. [PubMed]
21. He B, Mirza M, Weber GF. An osteopontin splice variant induces anchorage independence in human breast cancer cells. Oncogene. 2006;25(15):2192–2202. [PubMed]
22. Yilmaz M, Christofori G, Lehembre F. Distinct mechanisms of tumor invasion and metastasis. Trends in molecular medicine. 2007;13(12):535–541. [PubMed]
23. Hulit J, Suyama K, Chung S, et al. N-cadherin signaling potentiates mammary tumor metastasis via enhanced extracellular signal-regulated kinase activation. Cancer Res. 2007;67(7):3106–3116. [PubMed]
24. Agiostratidou G, Hulit J, Phillips GR, Hazan RB. Differential cadherin expression: potential markers for epithelial to mesenchymal transformation during tumor progression. Journal of mammary gland biology and neoplasia. 2007;12(2–3):127–133. [PubMed]