Enhanced Expression of Rab27A Gene by Breast Cancer Cells Promoting Invasiveness and the Metastasis Potential by Secretion of Insulin-Like Growth Factor-II

Invasion and metastasis, which represent the devastating attributes of cancer cells, are continuing therapeutic challenges and common causes of death for patients with cancer (1). However, it is clear that they are such complex processes involving several strategies, including mesenchymal movement, amoeboid locomotion, and migration through tissues, and so on, that the molecular mechanisms are still poorly understood (2, 3). Thus, an improved understanding of the molecular basis underlying cancer invasion and metastasis is essential to develop novel and more effective molecular targets for therapy.

In breast cancer, which is the leading malignancy in women (4), cytokines and their receptors perform significant functions on proliferation, invasion, metastasis, and resistance to drugs (5). The insulin-like growth factor system (IGF), which comprises a complex network of ligands (IGF-I and IGF-II), cognate receptors (type I, type II IGF receptors, and insulin receptor), IGF-binding proteins, and IGF-binding protein proteases, plays multiple roles in normal mammary gland and breast cancer (6-10). With the studies of cytokines, some proteins are found to collaborate with special cytokines in cancer cells. Rab proteins, members of the superfamily of Ras-related small GTPases, have distinct subcellular localization and are believed to control cellular events ranging from secretion and endocytosis to signal transduction and development (11-15). A recent genome analysis revealed that the Rab family is composed of 60 members in Homo sapiens (16). Rab27A is a unique member in Rab family for its specific implication in human genetic diseases (15). Loss-of-function mutations in human Rab27A gene result in Griscelli syndrome, a rare autosomal disorder characterized by combination of partial cutaneous albinism and severe immunodeficiency (17, 18). Its clinical picture seems to be a manifestation of defects in two specialized lysosome-related organelles, due to failure of distributing melanosome in melanocytes and releasing the contents of lytic granules in CTL (19). A mouse mutant, ashen (Rab27a^ash), has mutations in the Rab27A gene and exhibits the same phenotypic features as Griscelli syndrome patients (20-22). Provance et al. (23) have revealed the profound functions of Rab27A on the two lysosome-related organelles in normal and Griscelli syndrome or ashen cells. Melanocyte and CTL are only a subset of cell types expressing Rab27A, according to previous studies which indicated that Rab27A protein is widely expressed in specialized secretory cells, including exocrine (particularly in mucin- and zymogen-secreting cells), endocrine, ovarian, and hematopoietic cells, most of which undergo regulated exocytosis (24). Recent investigation revealed the roles of Rab27A in exocytosis of insulin and chromaffin granules in endocrine cells (25-27). Consistently, three families of Rab27A effectors have been recognized and thoroughly investigated (28).

In the present study, we systematically up-regulated and down-regulated the expression of Rab27A protein and tried to explore the roles of Rab27A in invasion and metastasis in breast cancer cells. We showed that Rab27A is critical to invasive and metastatic phenotypes of breast cancer cells by regulating the secretion of IGF-II, which modulates the expression of vascular endothelial growth factor (VEGF), cathepsin D, cyclin D1, urokinase-type plasminogen activator (uPA), matrix metalloproteinase-9 (MMP-9), and p16.

To investigate whether Rab27A is involved in invasion and metastasis in breast cancer cells, we analyzed Rab27A expression using reverse transcription-PCR (RT-PCR) and Western blot in a panel of breast cancer cell lines with different invasive and metastatic potentials comprising MCF-7, MDA-MB-231, MDA-MB-435, and MDA-MB-435HM (29). We found that Rab27A mRNA (Fig. 1A) but not other functionally intimate members of the Rab family, Rab3A, Rab7, Rab33A, and Rab37 (data not shown), increased as invasive and metastatic ability increased. Semiquantitative real-time RT-PCR analyses revealed that the levels of Rab27A mRNA in MDA-MB-231, MDA-MB-435, and MDA-MB-435HM were elevated 2.1-, 3.4-, and 6.9-fold, respectively, compared with that in MCF-7, which has the weakest invasive and metastatic potential (refs. 29, 30; Fig. 1B; P < 0.05 or P < 0.01). Western blot also showed that breast cancer cell lines differentially expressed the Rab27A protein, which was 2.83-, 4.9-, and 9.19-fold in MDA-MB-231, MDA-MB-435, and MDA-MB-435HM, respectively, compared with that in MCF-7 as determined by densitometric analyses (Fig. 1C). These data suggested that invasion and metastasis in breast cancer cells were likely to correlate with endogenous Rab27A expression. Rab27B, the unique homologue of Rab27A, was not expressed in all the four breast cancer cell lines (data not shown).

To characterize the localization of endogenous Rab27A in breast cancer cells, we did immunofluorescence and confocal microscopy on human breast cancer cells with antibody specific to Rab27A. As shown in Fig. 2, Rab27A was found to be targeted to diffuse in cytoplasm in three breast cancer cell lines and particularly concentrated in the perinuclear area in MDA-MB-231 and MDA-MB-435 cells. This perinuclear localization of a dot-like pattern is not similar to that of the ring-like pattern in parietal cells of the stomach (24). Meanwhile, we did not find any specific relationship in location between Rab27A and cytoskeletons (data not shown). It is likely that in breast cancer cells, Rab27A participates in secretion but does not link to cytoskeletons by its effectors, which was reported by Kuroda and Fukuda (31).

Rab27A expression may affect the invasive and metastatic phenotypes of breast cancer cells. To investigate this possibility, the Rab27A eukaryotic expression vector pcDNA3.1(+)-Rab27A was constructed and transfected into MDA-MB-231 and MDA-MB-435 cells and generated stable transfectants. For each cell line, two of the Rab27A transfectants, mock-transfected and parental cells in which Rab27A expression was detected by RT-PCR (data not shown) and Western blot (Fig. 3), were used for further studies.

To investigate whether Rab27A expression could affect breast cancer cell proliferation, we did flow cytometry analyses. The results showed that the percentage of cells in the S phase was dramatically increased from 40% to 70% and that in the G₀ and G₁ phase was accordingly decreased from ∼50% to 20% (P < 0.01) with Rab27A transfection, suggesting that Rab27A overexpression in breast cancer cells redistributes cell cycle and causes more cells to divide (data not shown).

To validate the effects of Rab27A on the invasive potential of breast cancer cells, we did in vitro invasion assay to quantify the invasion ability. We found that all cell lines tested were invasive through Matrigel after 12 hours; however, Rab27A transfectants of both MDA-MB-231 and MDA-MB-435 were highly invasive compared with parental or mock-transfected cells after 24 and 48 hours by culturing in Transwell (P < 0.01; data not shown). These results indicated that the invasive potential of breast cancer cells in vitro has close correlation with the level of Rab27A expression.

The progression of cancer is usually accompanied by the genetic underpinnings of cancer invasion and metastasis (32). To further investigate the molecular basis underlying the Rab27A-mediated invasive phenotype of breast cancer cells in vitro, we focused our attention on several extensively recognized invasive- and metastatic-associated genes. Semiquantitative real-time RT-PCR and Western blot analyses showed that Rab27A overexpression was constitutive with up-regulation of VEGF, cathepsin D, cyclin D1, uPA, and MMP-9, and with down-regulation of p16 in MDA-MB-231 and MDA-MB-435 cells (Fig. 4A-C; P < 0.05 or P < 0.01). In contrast, no significant difference in the expression of p21, transforming growth factor-α, cyclin A, cyclin E, MMP-1, MMP-2, tissue inhibitor of metalloproteinase-1 (TIMP-1), TIMP-2, uPA receptor, and basic fibroblast growth factor was found (data not shown).

At the same time, overexpressed Rab27A concomitantly increased the amount of vesicles in MDA-MB-231 and MDA-MB-435 cells (figure not shown), indicating that Rab27A may be a decisive regulator of vesicle formation in breast cancer cells.

To determine whether Rab27A expression is critical to the invasive phenotype of breast cancer cells in vitro, we constructed RNA interference (RNAi) vectors targeting the Rab27A gene. First, we certified that the two fragments of RNAi effectively reduced the exogenous and endogenous Rab27A detected by Western blot in 293T cells (P < 0.01; data not shown). Then, we applied them on MDA-MB-231 and MDA-MB-435 parental cells. Approximately 60% to 65% of reductions in Rab27A mRNA and protein by 48 to 72 h after transfection were observed in RNAi-targeted MDA-MB-231 and MDA-MB-435 cells compared with parental and nonsilencing control (data not shown). Reduced Rab27A expression resulted in a significant decrease in the invasive potential of MDA-MB-231 and MDA-MB-435 parental cells in vitro (data not shown). With Rab27A RNAi, the expressions of VEGF, cathepsin D, cyclin D1, uPA, and MMP-9 in both mRNA and protein were down-regulated and those of p16 was up-regulated (Fig. 5A-C; P < 0.05 or P < 0.01). These results further verified that Rab27A affects the invasive potential in breast cancer cells in vitro by modulating the expression of VEGF, cathepsin D, cyclin D1, uPA, MMP-9, and p16.

Reduced Rab27A expression by RNAi also eliminated the concentrated perinuclear localization in MDA-MB-231 and MDA-MB-435 cell lines (data not shown).

To examine the effects of Rab27A overexpression on tumor growth and metastasis, we did the in vivo experiments using an orthotopic xenograft tumor model in the athymic mice. The results showed that Rab27A transfectants double the volume of primary tumor compared with mock-transfected or parental cells at the end of the in vivo assays (P < 0.05; data not shown). We also found Rab27A up-regulation induced much more hemorrhage and necrosis in primary tumors (data not shown). To study metastatic potential in vivo, the lungs of mice in both MDA-MB-231 and MDA-MB-435 groups were physically examined at autopsy and then subjected to microscopic examination for morphologic evidence of tumor cells by light microscopy on H&E-stained paraffin sections. We found that Rab27A overexpression dramatically increased the pulmonary metastasis in both MDA-MB-435 and MDA-MB-231 cells (P < 0.01; data not shown). Furthermore, we compared Rab27A expression both in the mRNA and protein levels in primary tumors and in pulmonary metastases. The results showed that the latter seems to express much more Rab27A than the former, especially in Rab27A-transfected groups (Fig. 6; P < 0.05 or P < 0.01), which is another evidence that a high level of Rab27A is necessary for acquisition of invasive and metastatic potentials of breast cancer cells in vivo. Additionally, similar effects of Rab27A on p16, VEGF, uPA, cathepsin D, cyclin D1, and MMP-9 expression noted in the in vitro studies were also observed in in vivo studies (data not shown). With Rab27A RNAi, the volume of primary tumor and the number of pulmonary metastasis were accordingly decreased (data not shown).

Rab27A is not found as a transcription factor. We speculate that Rab27A is involved in transporting or releasing proteins that can regulate the expression of other genes. To gain insight into the molecular basis, we first examined the expression of IGF, which has the homologous structure to insulin in human breast cancer cells. RT-PCR showed that the four breast cancer cell lines coincidentally express insulin receptor, IGF-II, type I and type II IGF receptors, but not IGF-I (Supplementary Data G). We then investigated the relationship between the expression of IGF-II and Rab27A and showed that IGF-II mRNA was not changed by Rab27A up-regulation or down-regulation in breast cancer cells (data not show). Finally, we detected IGF-II protein in cells by Western blot and in conditioned medium by ELISA at different time points. Interestingly, IGF-II protein in conditioned medium of either MDA-MB-231 or MDA-MB-435 was increased by Rab27A overexpression after 12 hours, and these disparities reached their peaks after 48 hours (P < 0.05 or P < 0.01; data not shown). In contrast, IGF-II protein detected by Western blot in both MDA-MB-231 and MDA-MB-435 cells was down-regulated by Rab27A overexpression accordingly after 24 hours but not after 12 hours (Fig. 7A and B; P < 0.05 or P < 0.01). These results suggested that Rab27A contributes to secretion but not expression of IGF-II in breast cancer cells. Whether IGF-II plays a key role in Rab27A pathway, we try to block the IGF-II signaling pathway in MDA-MB-231 and MDA-MB-435 cells and their Rab27A transfectants using IGF-IR inhibitor NVP-ADW742 (Novartis Pharma AG). Western blot showed that NVP-ADW742 efficiently decreased the expression of total Akt and especially phosphorylated Akt, which is the main molecule downstream of IGF-IR signaling pathway (Fig. 7C). With inhibition of IGF-IR, MDA-MB-231 and MDA-MB-435 cells and their Rab27A transfectants displayed decreased invasive potential in vitro (P < 0.05 or P < 0.01; data not shown). We also found that the inhibition of IGF-IR attenuated the effects of Rab27A on the expression of MMP-9, VEGF, cyclin D1, cathepsin D, uPA, and p16 to the levels in their parental cells (data not shown). We did coimmunostaining of Rab27A and IGF-II by immunofluorescence and confocal microscopy on MDA-MB-231 and MDA-MB-435 and found that Rab27A colocalized with IGF-II within these cells (data not shown).

Eleven effectors, which are divided into three groups, are essential for Rab27A to function as a transporter (28). Our results manifested that melanophilin and SYTL4, but not other effectors, were expressed in the mRNA and protein levels in breast cancer cells (Supplementary data H-K). Using immunoprecipitation, we found that SYTL4 was coprecipitated with Rab27A in all four breast cancer cell lines but with melanophilin only in MCF-7 and MDA-MB-231 cells (data not shown), which suggests that Rab27A participates in some steps of IGF-II secretion in breast cancer cells by interacting with the effectors of SYTL4 and/or melanophilin, which may have the different function from the former. We presume that the reason Rab27A bands detected by anti-FLAG and anti-Rab27A antibodies double in Western blot is that whereas one portion of Rab27A binds with melanophilin, the other portion binds with SYTL4 in MCF-7 and MDA-MB-231 cell lines.

As the unique member of Rab protein family for its specific implication in human genetic diseases (15), Rab27A was first reported in 1997 (33). Because of Griscelli syndrome, the functions of Rab27A have been investigated in detail in melanocytes and CTLs, and recent studies showed that Rab27A is expressed and responsible for transportation and secretion of some cytokines in exocrine and endocrine cells (25-27). Paracrine and autocrine cytokines were reported in breast cancer cells (34). Although Rab27A contributes to transportation of melanosome in melanocytes (35, 36), the correlation between Rab27A expression and the phenotypes of other tumors still remains unknown.

In the present study, we showed that the expression of Rab27A but not other several Rab proteins, Rab3A, Rab7, Rab33A, and Rab37, which are functionally intimate with Rab27 (37), increased in human breast cancer cells as invasive and metastatic potential increased, which indicated possible involvement of Rab27A in invasive and metastatic phenotypes of breast cancer cells. We found that overexpression of the Rab27A protein resulted in a much higher percentage of cells in the S phase. These findings were consistent with our in vivo observations that up-regulation of Rab27A promoted the fast growth of primary tumors. Furthermore, we also showed that Rab27A expression positively regulated the invasive and pulmonary metastatic potentials of MDA-MB-231 and MDA-MB-435 cells in vitro and in vivo. All the pulmonary metastases expressed much more Rab27A mRNA and protein than the primary tumors, especially in the Rab27A transfectants, indicating that Rab27A may have a distinct role in invasion and metastasis of breast cancer cells.

Tolmachova et al. (24) observed that Rab27A is expressed in a comprehensive range of classic exocrine secretory cells and it is noticeable that the polarization of Rab27A is toward the apical/luminal side of the cells, coinciding with the location of secretory granules. In our results of immunofluorescence for Rab27A, Rab27A seems to be targeted to diffuse in cytoplasm in all three cell lines and in MDA-MB-231 and MDA-MB-435 cells particularly concentrated in the perinucleus.

The development and progression of breast cancer are also accompanied by genetic alterations of multiple oncogenes and tumor suppressor genes (32). In the present study, we showed that Rab27A positively modulated the expression of VEGF, cathepsin D, cyclin D1, uPA, and MMP-9 and negatively modulated p16. These results were further supported by the data obtained from RNAi targeting Rab27A in vitro.

Rab27A is not found as a transcriptional factor; thus, we speculate that some proteins will be mediators to link Rab27A and other genes that have effects on cellular behaviors such as IGFs. Recently, Rab27A was found to be involved in exocytosis of insulin in pancreatic β cells (27, 38, 39). IGF, which has a homologous structure to insulin, was thought to induce tumor division, invasion, and metastasis by regulating other gene expression (40). Our results revealed that IGF-II, type I and type II IGF receptors, and insulin receptor, but not IGF-I, were coincidentally expressed in four breast cancer cell lines. IGF-II released to cell culture supernatant was positively modulated by Rab27A expression. Accordingly, Rab27A affected IGF-II protein on the opposite side and did not change the mRNA of IGF-II in these cells, suggesting that Rab27A is involved in secretion but not expression of IGF-II. We also blocked the IGF-II signaling pathway using IGF-IR inhibitor NVP-ADW742, which was shown to have much more potent inhibitory effect against IGF-IR than insulin receptor, HER2, platelet-derived growth factor receptor, VEGF receptor 2, etc. (41). It was clearly shown that NVP-ADW742 efficiently decreased the expression of total Akt and especially phosphorylated Akt, which is the main molecule downstream of IGF-IR signaling pathway. With inhibition of IGF-IR, MDA-MB-231 and MDA-MB-435 cells and their Rab27A transfectants displayed decreased invasive potential in vitro. At the same time, we found colocalization of Rab27A and IGF-II in MDA-MB-231 and MDA-MB-435 cells, which suggests that Rab27A plays a key role in IGF-II pathway.

Cyclin D1 is overexpressed in up to 50% of primary breast cancers, in part due to amplification of the cyclin D1 gene, CCND1 (42). Hartmann et al. (43) also documented that IGF-II positively modulates cyclin D1 transcription induced by Sonic hedgehog in human medulloblastoma cells. In our experiment, the expression of cyclin D1 and a novel cell cycle–related tumor suppressor gene, p16, which has been highlighted as important in resistance to oncogenic transformation in mammary epithelium (44), were changed by Rab27A up-regulation or down-regulation in breast cancer cells.

The expression of MMP-9 was also evaluated using gelatin zymography in human non–small cell lung cancer cells treated by IGF-II (45). MMPs are a family of zinc-dependent neutral endopeptidases that play a key role in degrading the extracellular matrix and basement membrane in various cancers and therefore promotes metastasis and angiogenesis (42, 46). Increased numbers of blood vessels and hence increased angiogenic activity are essential for the growth and metastasis of cancer (47). Other than MMPs, VEGF and basic fibroblast growth factor are the two most potent factors that can stimulate angiogenesis and facilitate the malignant dissemination of cancer cells (48). Kwon et al. (49) observed that IGF-II mediated VEGF expression by the tyrosine kinase–Src–extracellular signal-regulated kinase 1/2 pathway and the phosphatidylinositol 3-kinase pathway in human keratinocytes. In addition, previous studies have shown that up-regulated expression of uPA, uPA receptor, and cathepsin D are associated with increased tumor cell invasion and metastasis in several malignancies, including breast cancer (50, 51). In our study, we found that MMP-9, VEGF, uPA, and cathepsin D expression, which can be up-regulated by IGF-II overexpression (52, 53), were positively modulated by Rab27A in MDA-MB-231 and MDA-MB-435 cells; however, no significantly different expression in TIMP-1, TIMP-2, MMP-1, MMP-2, basic fibroblast growth factor, and uPA receptor was observed. Taken together, these results indicate that Rab27A may be a decisive regulator of secretion of IGF-II, which regulates the expression of cyclin D1, VEGF, MMP-9, uPA, cathepsin D, and p16; and then affects cell cycle distribution, and the invasive and metastatic potentials in breast cancer cells.

Eleven putative Rab27A effectors that are classified into three distinct groups based on their structures have been reported in human and mice (28). The first group consists of the members of the synaptotagmin-like protein (Slp) family (Slp1/JFC1, Slp2-a, Slp3-a, Slp4/granuphilin, and Slp5) and rabphilin, all of which contain an NH₂-terminal Rab27A-binding domain (Slp homology domain) and COOH-terminal tandem C2 domains, which are putative phospholipid binding sites (26, 36, 54-56). The second group of Rab27A-binding proteins that contain an NH₂-terminal Rab27A-binding domain, the same as the Slp members, but lack tandem C2 domains consists of the members of the Slac2 family (Slac2-a/melanophilin, Slac2-b, and Slac2-c/MyRIP) and Noc2 (36, 54-57). The final group consists of only Munc13-4, which contains a novel Rab27A-binding domain distinct from the Slp homology domain and is a putative priming factor for exocytosis by certain immune cells (58, 59). Expression of several Rab27A effectors in one secreting cell type has also been reported (25, 26, 36). In our study, we detected that melanophilin and SYTL4 are expressed in four breast cancer cell lines. However, in the subsequent immunocoprecipitation, we confirmed that SYTL4, which was reported to be related to the expression of ER in breast cancer tissues (60), binds to Rab27A in all four breast cancer cell lines but binds to melanophilin only in MCF-7 and MDA-MB-231cells. These results revealed the complexity of the effectors of Rab27A in vivo, which may bind to different Rab proteins in different cell lines.

Collectively, our results indicate that Rab27A plays a positive regulatory role in invasion and metastasis of human breast cancer cells. Rab27A has effects on the invasive and metastatic potentials in breast cancer cells by modulating the secretion of IGF-II, which regulates the expression of p16, VEGF, uPA, cathepsin D, cyclin D1, and MMP-9. In this process, Rab27A effectors, SYTL4, and melanophilin may be important participants by binding to Rab27A. These findings may lead to a new therapeutic strategy against breast cancer.

Human breast cancer cell lines MCF-7, MDA-MB-231, MDA-MB-435, and MDA-MB-435HM were used in this study. The former three cell lines were obtained from the American Type Culture Collection and grown in complete growth medium as recommended by the manufacturer. The last one was established from MDA-MB-435 using an in vivo stepwise selection scheme in our laboratory (61) and grown in the same medium to its parental cell line MDA-MB-435. All the cultured cells were maintained in a humidified 5% CO₂ atmosphere at 37°C. Female BALB/c-nu/nu nude mice, 4 to 6 wk old, were obtained from Shanghai Institute of Materia Medica, Chinese Academy of Sciences (Shanghai, China), and housed in laminar-flow cabinets under specific pathogen-free conditions with food and water ad libitum. All experiments on mice were conducted in accordance with the guidelines of NIH for Care and Use of Laboratory Animals. The study protocol was also approved by the Shanghai Medical Experimental Animal Care Committee.

Total RNA was isolated from cultured cells or tissues using the Trizol reagent (Invitrogen) according to the manufacturer's instruction. The semiquantitative real-time RT-PCR using SYBR green I to compare the relative expression of specific gene mRNA was done as described previously (62).

The primary antibody for IGF-II was obtained from Upstate. The primary antibodies for SYTL4 and melanophilin were produced as described previously (63, 64). Other primary antibodies were obtained from Santa Cruz Biotechnology (65, 66). For immunoblotting, proteins were resolved by SDS-PAGE and transferred to nitrocellulose membranes (Whatman Schleicher and Schuell). Immunoblots were developed in Western Lightning Chemiluminescence Reagent Plus (Perkin-Elmer Life and Analytical Sciences) and exposed to X-ray films (Eastman Kodak). When necessary, blots were stripped and reblotted with other antibodies. The level of each protein was normalized versus the corresponding protein level for internal α-tubulin. For immunoprecipitation, FLAG-tagged Rab27A from 1 × 10⁷ cells were immunoprecipitated with anti-FLAG M2 beads (Sigma) and were eluted with 0.8 mg FLAG peptide per milliliter in a total volume of 30 μL (67).

The entire open reading frame of the human Rab27A gene (GeneID 5873) was amplified from 293T cells by RT-PCR using the gene-specific primers (the upstream primer, 5′-GCTAAGCTTGGTGAACTACTGAGTTCTTC-3′, with an added HindIII site underlined; the downstream primer, 5′-GACGAATTCTGTCGCTTACTTGACTTCTC-3′, with an added EcoRI site underlined), and then subcloned in-frame into the multiple cloning site of pcDNA3.1(+) vector. MDA-MB-231 and MDA-MB-435 cells were transfected by Lipofectamine 2000 transfection reagent (Invitrogen) according to the manufacturer's instruction. After selection in the presence of 1,000 μg/mL Geneticin (G418 sulfate; Invitrogen) for 4 wk, the pcDNA3.1(+)-Rab27A–positive and the empty vector–positive colonies named as 231/Rab27A or 435/Rab27A and 231/vector or 435/vector were identified by RT-PCR and Western blotting analysis as described above. To express FLAG-tagged proteins, Rab27A cDNAs were properly inserted into pUHD30F.

In vitro invasion assays were conducted with a Matrigel invasion chamber (BD Labware) as previously described with some modifications (68). Each well insert was coated with 100 μL of 1:3 dilution of Matrigel in serum-free culture medium. Then, 750 μL of medium containing 10% fetal bovine serum were added to the lower chamber as a chemoattractant, and 1 × 10⁵ cells in 500 μL of serum-free medium were added to the top of the Matrigel layer. The cells were incubated at 37°C for 6, 12, 24, or 48 h, respectively. The cell suspension was aspirated, and excess Matrigel was removed from the filter using a cotton swab. Invasion ability was assessed by counting the cells that had traveled across the filter and attached to the bottom side of the filter. Then, the filters were fixed in 10% formalin and stained with H&E. The cells that had invaded through the Matrigel and reached the lower surface of the filter were counted under a light microscope at a magnification of ×200.

Cells at the exponential growth phase were harvested and single-cell suspensions containing at least 1 × 10⁶ cells were made. The cells were treated following the standardized protocol, and cell cycle analyses were done by flow cytometry as described previously (69).

Generally, cells grown on glass coverslips were fixed in cold methanol for 10 min before immunofluorescence staining. Proper antibody or combinations were chosen for staining. Nuclear DNA was stained with 4,6-diamidino-2-phenylindole. Fluorescence images were captured with a cold charge-coupled device (SPOT II; Diagnostic Instruments) on an Olympus BX51 microscope or with a Leica TCS SP2 laser confocal microscope. For confocal microscopy with fixed cells, optical sections were scanned at 0.1- to 0.2-μm intervals when needed. Z-stack images were then formed by maximal projection.

To further show the role of Rab27A in progression of breast cancer cells, we used the microRNA interference technique to down-regulate Rab27A gene expression. The two pre-microRNA sequences that targeted the human Rab27A gene from the 516-bp sequence (5′-TGCTGTCCAGAAGCATCTCAATTGCTGTTTTGGCCACTGACTGACAGCAATTGATGCTTCTGGA-3′) and from the 611-bp sequence (5′-TGCTGTTAACTGATCCGTAGAGGCATGTTTTGGCCACTGACTGACATGCCTCTGGATCAGTTAA-3′) were designed by the Invitrogen RNAi Designer and ligated with pcDNA 6.2-GW/EmGFP-miR-lacZ (V49350-00, Invitrogen). We transfected the recombinant plasmids into 293T cells, which were expressed exogenous Rab27A tagged with FLAG to verify their function on suppression to the exogenous and endogenous expression of Rab27A. Then, we transfected the microRNAs into MDA-MB-231 and MDA-MB-435 cells, and their parental cells and pcDNA 6.2-GW/EmGFP-miR-lacZ vector–only transfectants were used as controls.

The tumorigenicity and spontaneous metastatic potential of the cell lines in the same orthotopic model described above were determined by injection into the mammary fat pad as described previously (70). The rate of primary tumor growth was determined by plotting the means of two orthogonal diameters of the tumors, measured twice per week. Mice were sacrificed and autopsied at 4 wk of postinoculation for MDA-MB-231 group and 8 wk for MDA-MB-435 group. To confirm the presence of lung metastases, sections were cut at 50-μm intervals and H&E staining was done. In this study, we count metastasis in the whole lung. Two independent pathologists calculated the number of metastasis.

The medium without serum was collected after 6, 12, 24, 36, and 48 h for IGF-II ELISA determinations as described below. The cells were taken through three freeze-thaw cycles and centrifuged; the supernatant was collected for determination of protein concentration and the cells were dissolved according to the protocol of RT-PCR or Western blot. IGF-II levels in the cell culture medium were measured using a Quantikine kit from R&D Diagnostics using the procedure provided by the supplier. Human recombinant IGF-II included in the kit was used to construct a standard curve and to obtain the absolute values of IGF-II protein content. The values were then normalized to the total protein concentration in each dish.

ANOVA and Student's t test were used to determine the statistical significance of differences between experimental groups, with P < 0.05 as the statistically significant level. Graphs were created with Excel software (Microsoft Office for Windows 2000).