• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Apr 27, 2004; 101(17): 6647–6652.
Published online Apr 15, 2004. doi:  10.1073/pnas.0401753101
PMCID: PMC404099
Medical Sciences

Requirement for Arf6 in breast cancer invasive activities


In most human breast cancer cell lines, there is a direct correlation between their in vivo invasive phenotypes and in vitro invasion activities. Here, we found that ADP-ribosylation factor 6 (Arf6) is localized at the invadopodia of the cultured breast cancer cells MDA-MB-231, and its suppression by a small-interfering RNA duplex effectively blocks the invasive activities of the cells, such as invadopodia formation, localized matrix degradation and Matrigel transmigration but not the cell-adhesion activity. We also found that the GTP hydrolysis-defective mutant Arf6(Q67L) and the GTP-binding defective mutant Arf6(T27N) both blocked these invasive activities but not cell adhesion, suggesting the necessity of continued activation and cycling of the Arf6 GTPase cycle in invasion. Among the different human breast cancer cell lines that we examined, cell lines with high invasive activities expressed higher amounts of Arf6 protein than those in weakly invasive and noninvasive cell lines, although no notable correlation was found between Arf6 mRNA expression levels and invasive activities. Moreover, Matrigel-transmigration activity of all of these invasive cells was blocked effectively by an Arf6 small-interfering RNA duplex. Hence, Arf6 appears to be an integral component of breast cancer invasive activities, and we propose that Arf6 and the intracellular machinery regulating Arf6 during invasion should be considered as therapeutic targets for the prevention of breast cancer invasion.

The metastatic potential of carcinomas constitutes a major cause of the poor prognosis of patients and correlates well with the invasive phenotype. Thus, an enormous amount of effort has been made to try to inhibit the invasive and migratory activities of carcinoma cells. For example, because matrix degradation by metalloproteases and serine proteases is essential for invasion, many inhibitors against these proteases have been developed. However, clinical trials for patients with late-stage cancers have so far shown these inhibitors to be largely ineffective in slowing tumor progression and metastasis. This phenomenon is attributed to the complicated and simultaneous involvement of different types of proteases in matrix degradation (1-4), as demonstrated in experimental systems in vitro (5). One of the best alternatives may, hence, be to target the molecular machinery involved in the more fundamental aspects of cancer invasion and migration. The small GTPase ADP-ribosylation factor 6 (Arf6) regulates membrane recycling and remodeling at the cell periphery, and it has been implicated in the higher orders of cellular functions (6). Here, we show that Arf6 plays an essential role in the invasive activities of human breast cancer cells. Possible roles of other Arf isoforms in cancer invasion and migration were examined also.

Materials and Methods

Cells and Small-Interfering RNA (siRNA)-Mediated Silencing of Protein Expression. Human breast cancer cell lines were obtained from the American Type Culture Collection. MDA-MB-231 cells were cultured in a 1:1 mixture of DMEM and RPMI 1640 supplemented with 10% FCS (HyClone) and 5% NuSerum (Becton Dickinson), as described (7). Other human breast cancer cell lines were cultured according to the manufacturer's instructions (American Type Culture Collection). A primary culture of human normal mammary gland cells was purchased from Cambrex (East Rutherford, NJ) and cultured according to the manufacturer's instructions. HeLa cells were cultured in DMEM supplemented with 10% FCS (HyClone). Silencing of protein expression was performed by using the siRNA technique, as described (8). Duplex oligonucleotides, 5′-GCACCGCAUUAUCA AUGACCGUU-3′ and 5′-CGGUCAUUGAUAAUGCGGUGCUU-3′, designed for Arf6 silencing were chemically synthesized and purified by Japan BioServices (Saitama, Japan). As a control, an siRNA duplex with an irrelevant sequence (5′-GCGCGCUUUGUAGGAUUCG-3′ and 5′-CGCGCGAAACAUCCUAAGC-3′; Dharmacon, Lafayette, CO) was used. Cells cultured in the growth medium were transfected with 25 nM oligonucleotide duplexes by using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions and incubated for 24 h before being subjected to analyses. Cell viability was measured by using a 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) colorimetric assay kit (Promega) according to the manufacturer's instructions.

cDNAs and Their Transfection. Hemagglutinin (HA)-tagged Arf cDNAs each in the pcDNA3 vector were gifts from K. Nakayama (Kyoto University). The Tac cDNA has been described (9). Wild-type and mutant Arf cDNAs that were used in this study are described in the legend to Fig. 3. Enhanced GFP (EGFP)-tagged Arf1 and Arf6 cDNAs, each in the pBabe vector (10), were constructed as follows. Each PCR-amplified Arf cDNA fragment was ligated into the XhoI site of pEGFP-N1 (Clontech), and the NheI-NotI fragment was then isolated, blunted, and ligated into the SnaBI site of the pBabePuro vector. The pBabe vector utilizes the murine leukemia virus LTR to transcribe cDNA sequences inserted in the vector. pEGFP-C1, encoding the EGFP protein, was purchased from Clontech. For invasion, migration, and adhesion assays of Arf cDNA transfected cells, 5 × 105 cells were cotransfected with 5 μg of HA-tagged Arf cDNA in the pcDNA vector and 0.2 μg of pEGFP-C1 by using Trans IT-LT1 (Mirus, Madison, WI) according to the manufacturer's instructions and incubated for 24 h in growth medium before being subjected to analyses. An empty pcDNA3 vector was used as a negative control (mock), instead of Arf cDNAs. Under these conditions, we observed by detecting the autofluorescence from EGFP and immunostaining cells with an anti-HA antibody that >95% of EGFP-positive cells also expressed HA-tagged Arf proteins. To achieve very low level expression, 5 × 105 cells were transfected with 0.3 μg of pBabe Arf6-EGFP or pBabe Arf1-EGFP by using Trans IT-LT1 and incubated for 24 h before analysis. For the rescue assay of Arf6 siRNA treatment, the mutant Arf6-EGFP cDNA in the pBabe vector was constructed by substituting the nucleotide within the siRNA target to 5′-TCATAGGATAATTAACGATAG-3′ (substituted nucleotides are italicized). These substitutions do not change the coding amino acids. We then cotransfected 5 × 105 cells with 2 μg of this cDNA together with 25 nM of the Arf6 siRNA oligonucleotide duplex by using Lipofectamine 2000 and incubated them for 24 h before analyses.

Fig. 3.
Possible requirement for the Arf6 GTPase cycle in the invasive activities of MDA-MB-231 cells. Cells were transiently cotransfected with pEGFP-C1 (for EGFP) and a plasmid each expressing wild-type, GTP hydrolysis-defective (Q67L for Arf6, and Q71L for ...

Immunoblotting. Antibodies against the following proteins were purchased from commercial sources: Arf6 (Santa Cruz Biotechnology), Arf (Affinity BioReagents, Golden, CO), β-actin (Sigma), Tac (7G7; Upstate Biotechnology), cortactin (Upstate Biotechnology, Lake Placid, NY), the HA sequence (Babco, Richmond, CA), MMP2 (Chemicon), and MMP9 (Santa Cruz Biotechnology). Immunoblot analyses of cell lysates were done after the separation of cell lysates (20 μg per lane) by SDS/PAGE, as described (11). Amounts of proteins were then measured by a GT8700 Scanner densitometer (Epson, Tokyo) using NIH IMAGE, version 1.63, software. Data are presented as mean ± SEM from at least three independent experiments.

Invadopodia Formation and Localized Gelatin Matrix Degradation. Porcine skin gelatin (Type A with 300 Bloom; Sigma) was conjugated with Alexa-594 (Molecular Probes) by incubation at 37°C for 1 h according to the manufacturer's instructions. Coating of glass-bottomed dishes (MatTek, Ashland, MA) with the fluorescence-conjugated gelatin and the subsequent cross-linking of the gelatin with 0.5% glutaraldehyde were performed as described (12). Cells were replated onto the fluorescence-conjugated, cross-linked gelatin matrix-coated dishes in the growth medium and incubated for 16 h and then fixed in 4% paraformaldehyde for 20 min at room temperature. Immunostaining of fixed cells was performed as described (12). Cortactin was visualized by using an anticortactin antibody coupled with a Cy2-conjugated anti-mouse IgG antibody (Jackson ImmunoResearch), and F-actin was visualized with biotinylated phalloidin (Molecular Probes) and Cy5-conjugated streptavidin (Jackson ImmunoResearch). Fluorescence microscopy was performed by using an LSM 510 confocal laser-scanning microscope (Zeiss) and the associated software, as described (11). To determine the number of degrading cells and degraded areas for each experiment, we considered 100 random fields (containing at least 50 cells) at a ×63 magnification. Each figure shows representative results that were observed in a majority of the cells in at least three experiments. Error bars represent SEM.

Matrigel Invasion Assay. Matrigel invasion assay was performed by using Biocoat Matrigel chambers (Becton Dickinson), as described (13). Briefly, 1 × 105 cells were seeded on the upper wells of 24-well chambers in the absence of serum, in which the lower wells were filled with conditioned medium of NIH 3T3 cells cultured for 24 h in the absence of serum. After incubation for 6 h, cells invaded into Matrigel and migrated out onto the lower surface of the membrane were fixed in 4% paraformaldehyde. These cells were then stained with 1% crystal violet (for siRNA-treated cells) or subjected to identification of EGFP-positive cells by detecting the autofluorescence using LSM 510 laser scanning microscope (for Arf cDNA transfected cells). Data collection and presentations are the same as above.

Haptotactic Migration and Cell-Adhesion Assays. Haptotactic cell migration assay was performed in the absence of serum by using modified Boyden chambers (Transwell with 8-μm pores; Costar), in which the lower surface of the membrane was coated with collagen type I (10 μg/ml), as described (14). Cell-adhesion assay onto collagen type I-coated dishes (10 μg/ml) was also described (14).

Gelatin Zymography. We incubated 5 × 105 siRNA-treated cells in the absence of serum for 12 h, and the cultured medium was harvested and centrifuged at 10,000 × g for 2 min. The supernatants were then concentrated by using a Centricon centrifugal filter (10-kDa cut size; Millipore), separated by SDS/PAGE, and subjected to zymography for gelatin degradation, as described (15). Positions of MMP2, MMP9, and their active forms were identified with cell lysates run simultaneously on the same SDS/PAGE gels, as described (15).

Recycling of Tac. Internalization of Tac and its recycling back to the cell surface were measured as described (16) by using an anti-Tac antibody. Cell-associated Tac antigens were then detected by using Cy3-conjugated anti-mouse IgG antibodies. Transfection-positive cells were identified by the autofluorescence from EGFP or by immunolabeling of transfected proteins by their tags. Cells were scored as being blocked only if there was >80% inhibition of internalization or recycling back of Tac. We examined >100 transfection-positive cells in each experiment.

Real-Time PCR. Total cellular RNA was isolated by using TRIzol (Life Science, Arlington Heights, IL). Random-primed cDNAs, prepared from 2 μg of cellular RNAs by using the SuperScript first-strand cDNA synthesis kit (Invitrogen), were subjected to real-time PCR amplification analysis using the LightCycler FastStart DNA Master SYBR Green I (Roche) and the LightCycler (Roche), according to the manufacturer's instructions. PCR amplification conditions were as follows: 95°C for 10 min followed by 40 cycles of 95°C for 15 sec, 55°C for 5 sec, and 72°C for 5 sec. The following Arf6 primers were used: 5′-ATGGGGAAGGTGCTATCCAAAATC-3′ and 5′-GCAGTCCACTACGAAGATGAGACC-3′. We used β-actin primers (Takara Shuzo, Kyoto) as internal controls.


siRNA-Mediated Suppression of Arf6 Blocks Invasive Activities of Breast Cancer Cells. We made oligonucleotides for human Arf6 siRNA, and we tested whether they inhibit the invasion and migration of MDA-MB-231 cells, a commonly used cell line for studying the molecular bases for human breast cancer cell invasion and metastasis (17). Treatment of cells with the oligonucleotides effectively blocked Arf6 expression to levels <10% of those in control cells (Fig. 1 A and B). Under these conditions of Arf6 siRNA treatment, cell viability was not affected (Fig. 1C). MDA-MB-231 cells produce invadopodia that extend into the matrix substratum, such as collagen and gelatin matrix (7, 12, 18). These cells actively reform invadopodia, leaving a cumulative record of matrix degradation. We found that both invadopodia formation and localized gelatin matrix degradation were effectively blocked by Arf6 siRNA treatment (Fig. 1 D-F). Invadopodia were identified by staining cells with their markers, cortactin and F-actin (7), and matrix degradation was assessed by seeding cells on a fluorescently labeled, cross-linked gelatin matrix and detecting spots where the fluorescence signals had disappeared, as described (12). Chemoinvasion activity, in which cells migrate toward chemoattractants by breaking through a barrier of Matrigel, was also effectively blocked by Arf6 siRNA treatment (Fig. 1G). We also found that Arf6 siRNA treatment inhibits the haptotactic migratory activity toward collagen (Fig. 1H), in which matrix degradation is not necessary. However, unlike the invasive activities, the rate of inhibition of migration was only ≈50%, suggesting nonidentical requirements for Arf6 in invasion and migration. On the other hand, Arf6 siRNA did not inhibit cell adhesion onto collagen (Fig. 1I), nor did it inhibit the secretion of active MMP2 and MMP9 (Fig. 1J). We also confirmed that effects of Arf6 siRNA can be restored to almost normal by the simultaneous expression of Arf6 rescue cDNA (data not shown).

Fig. 1.
Silencing Arf6 expression blocks the invasive activities of MDA-MB-231 cells. Cells were transfected with an siRNA duplex against Arf6 or an irrelevant sequence (irr) and cultured for 24 h before being subjected to analyses. Controls also included cells ...

Arf6 Localizes at Invadopodia of Breast Cancer Cells. Arf6 accumulates at Fcγ receptor-based phagocytic cups of macrophages, which extend out from the plasma membrane to engulf IgG-opsonized particles (11). There are striking similarities between phagocytic cups and invadopodia: invadopodia are not only structures that extend out of the plasma membrane, but in MDA-MB-231 cells, invadopodia have been shown to be sites where cells phagocytose degraded gelatin by means of internalization of integrin α3β1 bound to the matrices (18). We then examined whether Arf6 localizes to invadopodia. We expressed Arf6 tagged with EGFP, in a manner by which the expression levels were tuned carefully to be within a 2-fold excess of that of endogenous Arf6 to avoid the potential effects of overexpression (data not shown). We found that Arf6-EGFP also localizes within sites of the cumulative record of matrix degradation, in addition to the plasma membrane areas (Fig. 2A Upper). This localization was still observed when the focus was adjusted 0.7 μm beneath the cell bottom (Fig. 2 A Lower). Arf6-EGFP also colocalized with cortactin and F-actin at matrix-degradation sites (data not shown). On the other hand, Arf1-EGFP does not localize to matrix-degradation sites (Fig. 2B). We have tried in vain to stain invadopodia clearly with the currently available anti-Arf6 antibodies.

Fig. 2.
Arf6 localizes at invadopodia of MDA-MB-231 cells. Cells expressing Arf6-EGFP (A) or Arf1-EGFP (B) were cultured on a fluorescently labeled, cross-linked gelatin matrix for 16 h, as shown in Fig. 1. Degraded gelatin zones are observed as spots (red), ...

Evidence for Requirement of Continued Activation of the Arf6 GTPase Cycle in Cancer Invasion. To obtain insight into how Arf6 is involved in invasion, we next expressed its mutant cDNAs. We found that both the GTP hydrolysis-defective Arf6(Q67L) and the GTP binding-defective Arf6(T27N) mutants blocked invadopodia formation, localized matrix degradation, and Matrigel chemoinvasion (Fig. 3 A-C). Overexpression of wild-type Arf6 also blocked these activities (Fig. 3 A-C), in cells in which expression was >20-fold higher than that of endogenous Arf6 (Fig. 3F). Cell viability was not affected under these conditions (data not shown). To examine whether exogenous expression of wild-type Arf6 always gives rise to an inhibitory effect, we isolated MDA-MB-231 cells stably expressing exogenous wild-type Arf6. We found that cell clones expressing exogenous Arf6 at levels only 2- to 3-fold higher than that of endogenous Arf6 exhibited almost similar or slightly augmented activities of invasiveness (Fig. 3 G-I). Cells stably expressing wild-type Arf6 at much higher amounts, as achieved by the transient expression of Arf6, were not obtained.

The inhibition of Arf6 function by both Arf6(Q67L) and Arf6(T27N) has been reported, such as regarding recycling of the transferrin receptor (19), Tac (16), and also Fcγ receptor-based phagocytosis (11, 20). Overexpression of wild-type Arf6 also inhibits transferrin receptor recycling (19) and Fcγ receptor-based phagocytosis (11, 20). It has, hence, been suggested that the continued activation and cycling of the Arf6 GTPase is necessary for such receptor recycling and phagocytosis. Fcγ receptor-based phagocytic cups and cancer invadopodia have many similarities, as mentioned earlier. In the case of Fcγ receptor phagocytosis, Arf6(Q67L), Arf6(T27N), and wild-type Arf6 overexpression have been shown to all block pseudopod extension at the initial stage of phagocytosis, similar to what we observed in the inhibition of invadopodia formation. On the other hand, in the case of recycling of the transferrin receptor and Tac, Arf6(Q67L) decreased the rate of endocytosis, whereas Arf6(T27N) inhibited these molecules from recycling back to the cell surface. We then examined whether expression of these Arf6 mutants in MDA-MB-231 cells results in similar effects as when expressed in other cells. To do this, we expressed Tac together with Arf6 in MDA-MB-231 cells. As shown in Fig. 4, similar manners of inhibition of internalization and recycling back of Tac were observed in MDA-MB-231 cells, as in HeLa cells (16). However, because MDA-MB-231 cells incorporated fairly large amounts of biotinylated transferrin even at 4°C, we could not assess precisely the effects of Arf6 mutants on the cellular dynamics of transferrin receptors (data not shown).

Fig. 4.
Effects of Arf6 mutants on Tac recycling in MDA-MB-231 cells. MDA-MB-231 cells were transfected with Tac cDNA, together with Arf6(Q67L) or Arf6(T27N) cDNAs as described, and subjected to internalization (A) and recycling-back (B) assays of Tac antigen. ...

We also examined the effects of the Arf6 mutants on haptotactic migration, which does not require matrix degradation. We found that although Arf6(Q67L) and Arf6(T27N) also inhibited this migration, the rates of inhibition were significantly less compared with the inhibition observed for the invasions (Fig. 3D). Moreover, unlike in the case of the invasions, overexpression of wild-type Arf6 did not block haptotactic migration (Fig. 3D). We have no explanation for why high levels of Arf6 overexpression blocked cancer invasion whereas it did not block haptotactic migration. This result was, however, reproducible. Because Arf6 siRNA inhibits haptotactic migration only partially while it substantially blocks invasion, this finding also suggests the unequal involvement of Arf6 in migration and invasion of breast cancer cells.

Comparison with Other Classes of Arf Isoforms. Besides Arf6, there are five isoforms of Arf in mammalian tissues (Arf1-5) (21). They are subclassified as class I (Arf1-3), class II (Arf4 and Arf5), and class III (Arf6), based on their sequence similarity (21). Class I Arfs are involved primarily in secretion and Golgi function (21, 22), whereas the function of the class II Arfs is still largely unknown. The possible involvement of Arf isoforms other than Arf6 in cancer cell invasion has been suggested by the use of brefeldin A (23, 24), which inhibits the functions of Arfs with the exception of Arf6, by blocking several guanine nucleotide exchange factor (GEF) activities (25). We next investigated the specific part, if any, played by Arf6 among the Arf isoforms in cancer cell invasion. For this purpose, we used Arf1 and Arf5, as representing the class I and class II Arfs, respectively. Forced expression of wild-type Arf1, as well as expression of the GTP-hydrolysis defective or GTP-binding defective mutants, all significantly blocked the invasive activities, whereas none of the Arf1 cDNAs exerted any inhibitory effects on haptotactic migration (Fig. 3 A-D). On the other hand, expression of Arf5 or its mutants did not block invasion or haptotactic migration (Fig. 3 A-D). None of the Arf isoforms nor the mutants blocked cell adhesion (Fig. 3E). Therefore, among the different classes of Arfs, Arf6 is unique in participating in both invasive and migratory activities of breast cancer cells.

Arf6 in Other Human Breast Cancer Cells. We finally examined whether our findings described above applies to other human breast cancer cell lines with different invasive activities (7, 13). We also used a primary culture of human normal mammary gland cells, which did not show any invasive activities in in vitro cultures. We found significantly higher levels Arf6 protein expression in those cell lines which are known to exhibit high invasive activities in the Matrigel invasion and chemotaxis assay, as compared with weak and noninvasive cells (Fig. 5A). However, Arf6 mRNA levels in these cells, assessed by a quantitative method, did not correlate with the enhanced protein expression (Fig. 5A). We also observed that Arf6 siRNA treatment of these highly invasive cancer cells also effectively blocked their invasive activities (Fig. 5B). In all of these cell lines, siRNA treatment suppressed Arf6 protein expression to levels <10% of those in untreated cells without affecting their immediate cell viability (data not shown).

Fig. 5.
Arf6 expression in different breast cancer cell lines and inhibition of their invasion by an Arf6 siRNA duplex. (A) Cell lysates, as indicated, were subjected to immunobloting analysis using antibodies for Arf6, pan-Arf, and β-actin (Upper). The ...


In this article, we show that Arf6 is an integral component necessary for the invasive activities of MDA-MB-231 cells. Moreover, our analysis on other representative, well characterized cancer cell lines support the idea that Arf6 may be involved more generally in the invasive phenotypes of human breast cancer. Our study using MDA-MB-231 cells also suggests that the continued activation and cycling of the Arf6 GTPase cycle are necessary for the invasive activities of breast cancer cells, as has been proposed for receptor recycling and phagocytosis. In the case of Fcγ receptor phagocytosis, continued activation of the Arf6 GTPase cycle is thought to be required for membrane mass supply for extension, perhaps by facilitating endosomal recycling (11, 20, 26), whereas intercommunication of Arf6 with Rac1 may also play a role in phagocytosis (27). Given the close similarity between Fcγ receptor phagocytic cups and invadopodia, it is conceivable that continued Arf6 GTPase cycling may be similarly required for plasma membrane extension during invadopodia formation. However, the detailed mechanisms of how Arf6 is involved in Fcγ receptor phagocytosis have not been well established. Likewise, we have yet to analyze the precise molecular mechanisms by which Arf6 is involved in the invasive activities of breast cancer cells.

We observed higher Arf6 protein expression in highly invasive breast cancers than in weakly invasive or noninvasive breast cancers and normal mammary epithelial cells, although Arf6 mRNA levels do not correlate with invasiveness. Consistent with our observations, gene-profiling analyses of primary breast tumors have not implicated enhanced expression of Arf6 mRNA with the progression of malignancy (28). Thus, a posttranscriptional process may be involved in the regulation of Arf6 protein expression in cancer cells, rather than the simple regulation of its mRNA levels. Possibly related to this notion, cellular protein levels of Arf6 appear to be regulated tightly by its ubiquitination and its subsequent degradation (I. Kobayashi and H.S., unpublished data). To examine whether higher levels of Arf6 protein expression indeed correlate with tumor malignancy and acquisition of invasive phenotypes, Arf6 antibodies that can be applied to immunohistochemical analyses of clinical specimens will be required.

Our results indicate that Arf6 can be considered as a target for developing therapeutics preventing breast cancer-cell invasion. However, Arf6 is expressed ubiquitously in adult tissues (29). Thus, its direct manipulation may cause serious side effects on normal tissues, unless methods to deliver the siRNAs or inhibitors

selectively to cancer cells are developed. However, we showed also that Arf6(Q67L) and Arf6(T27N) both block cancer cell invasion and migration effectively. Therefore, given that Arf6 itself cannot be a direct target, GEFs and GTPase-activating proteins (GAPs) for Arf6 can alternatively be considered as targets for therapeutics. Normal breast epithelia in adults do not exhibit an invasive phenotype, although they possess Arf6. It is, thus, conceivable that the invasive activity shown by breast carcinomas might be reminiscent of mechanisms used by epithelia in other stages such as during embryonic development, or used in other types of cells. The number of genes coding for Arf GEFs and Arf GAPs in humans is significantly larger than that for the Arf isoforms and Arf-like factors, and more than one type of GEF or GAP has been shown to act on Arf6 (6, 30, 31). Arf6 is versatile (6), and different GAPs and GEFs act on Arf6, perhaps depending on its different functions. We have already identified an ArfGAP that localizes to invadopodia of cultured breast cancer cells and is essential for invasion (Y.O. and H.S., unpublished data). This GAP is highly expressed in invasive ductal carcinomas in clinical specimens of human breast cancer, although its expression is low in ductal carcinomas in situ and in adult normal mammary epithelia (Y.O., S.H., and H.S., unpublished data).

In conclusion, we propose that Arf6 plays an essential role in the invasive phenotype of human breast cancer cells. Further identification and characterization of its auxiliary proteins, including GEFs, which are abnormally used in breast carcinomas for their invasion, could possibly greatly facilitate the development of therapeutics for the prevention of breast cancer cell invasion and perhaps also for preventing metastasis.


We thank Manami Hiraishi and Yumiko Shibata for technical assistance; Mayumi Yoneda for secretarial work; Ken-ichi Nakayama for Arf cDNAs; and Helena Akiko Popiel for critical reading of the manuscript. This work was supported in part by grants-in-aid from the Ministry of Education, Science, Sports, and Culture of Japan (MESSC) and by grants from Takeda Pharmaceutical and the Uehara Memorial Life Science Foundation. Y.O. is a recipient of a MESSC Studentship of the 21st Century Center of Excellence Program. The Osaka Bioscience Institute was funded in commemoration of the centenary of the municipal government of Osaka City and is supported by Osaka City.


Abbreviations: Arf, ADP-ribosylation factor; siRNA, small-interfering RNA; EGFP, enhanced GFP; HA, hemagglutinin; GAP, GTPase-activating protein; GEF, guanine nucleotide exchange factor.


1. Friedl, P. & Wolf, K. (2003) Nat. Rev. Cancer 3, 362-374. [PubMed]
2. Zucker, S., Cao, J. & Chen, W. T. (2000) Oncogene 19, 6642-6650. [PubMed]
3. Coussens, L. M., Fingleton, B. & Matrisian, L. M. (2002) Science 295, 2387-2392. [PubMed]
4. Overall, C. M. & Lopez-Otin, C. (2002) Nat. Rev. Cancer 2, 657-672. [PubMed]
5. Coopman, P., Do, M., Thompson, E. & Mueller, S. (1998) Clin. Cancer Res. 4, 507-515. [PubMed]
6. Donaldson, J. (2003) J. Biol. Chem. 278, 41573-41576. [PubMed]
7. Bowden, E. T., Barth, M., Thomas, D., Glazer, R. I. & Mueller, S. C. (1999) Oncogene 18, 4440-4449. [PubMed]
8. Elbashir, S. M., Harborth, J., Lendeckel, W., Yalcin, A., Weber, K. & Tuschl, T. (2001) Nature 411, 494-498. [PubMed]
9. Nikaido, T., Shimizu, A., Ishida, N., Sabe, H., Teshigawara, K., Maeda, M., Uchiyama, T., Yodoi, J. & Honjo, T. (1984) Nature 311, 631-635. [PubMed]
10. Morgenstern, J. P. & Land, H. (1990) Nucleic Acids Res. 18, 3587-3596. [PMC free article] [PubMed]
11. Uchida, H., Kondo, A., Yoshimura, Y., Mazaki, Y. & Sabe, H. (2001) J. Exp. Med. 193, 955-966. [PMC free article] [PubMed]
12. Bowden, E. T., Coopman, P. J. & Mueller, S. C. (2001) Methods Cell Biol. 63, 613-627. [PubMed]
13. Thompson, E. W., Paik, S., Brunner, N., Sommers, C. L., Zugmaier, G., Clarke, R., Shima, T. B., Torri, J., Donahue, S., Lippman, M. E., et al. (1992) J. Cell. Physiol. 150, 534-544. [PubMed]
14. Kondo, A., Hashimoto, S., Yano, H., Nagayama, K., Mazaki, Y. & Sabe, H. (2000) Mol. Biol. Cell 11, 1315-1327. [PMC free article] [PubMed]
15. Yoneda, T., Sasaki, A., Dunstan, C., Williams, P. J., Bauss, F., De Clerck, Y. A. & Mundy, G. R. (1997) J. Clin. Invest. 99, 2509-2517. [PMC free article] [PubMed]
16. Radhakrishna, H. & Donaldson, J. G. (1997) J. Cell Biol. 139, 49-61. [PMC free article] [PubMed]
17. Kang, Y., Siegel, P. M., Shu, W., Drobnjak, M., Kakonen, S. M., Cordon-Cardo, C., Guise, T. A. & Massague, J. (2003) Cancer Cells 3, 537-549.
18. Coopman, P., Thomas, D., Gehlsen, K. & Mueller, S. (1996) Mol. Biol. Cell 7, 1789-1804. [PMC free article] [PubMed]
19. D'Souza-Schorey, C., Li, G., Colombo, M. I. & Stahl, P. D. (1995) Science 267, 1175-1178. [PubMed]
20. Zhang, Q., Cox, D., Tseng, C. C., Donaldson, J. G. & Greenberg, S. (1998) J. Biol. Chem. 273, 19977-19981. [PubMed]
21. Moss, J. & Vaughan, M. (1998) J. Biol. Chem. 273, 21431-21434. [PubMed]
22. Lippincott-Schwartz, J., Cole, N. B. & Donaldson, J. G. (1998) Histochem. Cell Biol. 109, 449-462. [PubMed]
23. Kang, T., Nagase, H. & Pei, D. (2002) Cancer Res. 62, 675-681. [PubMed]
24. Ho, W., Exton, J. H. & Williger, B. T. (2003) FEBS Lett. 537, 91-95. [PubMed]
25. Chardin, P. & McCormick, F. (1999) Cell 97, 153-155. [PubMed]
26. Niedergang, F., Colucci-Guyon, E., Dubois, T., Raposo, G. & Chavrier, P. (2003) J. Cell Biol. 161, 1143-1150. [PMC free article] [PubMed]
27. Aderem, A. & Underhill, D. M. (1999) Annu. Rev. Immunol. 17, 593-623. [PubMed]
28. van 't Veer, L. J., Dai, H., van de Vijver, M. J., He, Y. D., Hart, A. A., Mao, M., Peterse, H. L., van der Kooy, K., Marton, M. J., Witteveen, A. T., et al. (2002) Nature 415, 530-536. [PubMed]
29. Yang, C. Z., Heimberg, H., D'Souza-Schorey, C., Mueckler, M. M. & Stahl, P. D. (1998) J. Biol. Chem. 273, 4006-4011. [PubMed]
30. Sabe, H. (2003) in ARF Family GTPases, ed. Kahn, R. (Kluwer, Dordrecht, The Netherlands), pp. 185-207.
31. Jackson, T. R., Kearns, B. G. & Theibert, A. B. (2000) Trends Biochem. Sci. 25, 489-495. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...