In vivo selection for metastasis promoting genesin the mouseKiranmai Gumireddy*, Fangxian Sun†, Andres J. Klein-Szanto‡, Jonathan M. Gibbins§, Phyllis A. Gimotty¶,Aleister J. Saunders�, Peter G. Schultz†**††, and Qihong Huang*††

*The Wistar Institute, 3601 Spruce Street, Philadelphia, PA 19104; †Genomics Institute of the Novartis Research Foundation, 10675 John Jay Hopkins Drive,San Diego, CA 92121; ‡Fox Chase Cancer Center, 333 Cottman Avenue, Philadelphia, PA 19111-2497; §School of Animal and Microbial Sciences, University ofReading, Whiteknights, Reading RG6 6AJ, United Kingdom; ¶Department of Biostatistics and Epidemiology, University of Pennsylvania, Philadelphia,PA 19104; �Department of Bioscience and Biotechnology, Drexel University, 3141 Chestnut Street, 316 Stratton Hall, Philadelphia, PA 19104; and**The Scripps Research Institute, 10550 North Torrey Pines Road, SR 202, La Jolla, CA 92037

Contributed by Peter G. Schultz, February 12, 2007 (sent for review December 1, 2006)

Here, we report the identification of a metastasis promoting factorby a forward genetic screen in mice. A retroviral cDNA library wasintroduced into the nonmetastatic cancer cell line 168FARN, whichwas then orthotopically transplanted into mouse mammary fatpads, followed by selection for cells that metastasize to the lung.The genes encoding the disulfide isomerase ERp5 and �-cateninwere found to promote breast cancer invasion and metastasis.Disulfide isomerases (thiol isomerases), which catalyze disulfidebond formation, reduction, and isomerization, have not previouslybeen implicated in cancer cell signaling and tumor metastasis.Overexpression of ERp5 promotes both in vitro migration andinvasion and in vivo metastasis of breast cancer cells. These effectswere shown to involve activation of ErbB2 and phosphoinositide3-kinase (PI3K) pathways through dimerization of ErbB2. Activa-tion of ErbB2 and PI3K subsequently stimulates RhoA and �-catenin, which mediate the migration and invasion of tumor cells.Inhibition of ErbB2 and PI3K reverses the phenotypes induced byERp5. Finally, ERp5 was shown to be up-regulated in humansurgical samples of invasive breast cancers. These data identify alink between disulfide isomerases and tumor development, andprovide a mechanism that modulates ErbB2 and PI3K signaling inthe promotion of cancer progression.

breast cancer invasion and metastasis � disulfide isomerase ERp5 �ErbB2/phosphoinositide 3-kinase � forward genetic screen �orthotopic animal model

Metastasis remains one of the most poorly understood pro-cesses in cancer biology despite extensive study (1). Metas-

tasis occurs by a multistep process requiring the coordinated actionof many genes (2–5): the escape of cancer cells from the primarytumor and entry into the blood stream (intravasation), survival inthe circulation, exit from capillaries into surrounding tissues at anew site (extravasation), initiation of growth to form micrometas-tases, and development of new blood vessels to form secondarytumors (2–5). Although a number of mechanisms have beenidentified that induce tumor cell migration and invasion, includingthe activation of cell surface receptors and small GTPases of theRho family (6–13), the systematic identification and characteriza-tion of additional genes that promote or suppress tumor invasionand metastasis should advance our understanding of the metastasisprocess and ultimately provide new therapeutic targets for thetreatment of cancer.

Forward genetic screens provide an unbiased approach to theidentification of genes that contribute to a phenotype of interest(14–16). Arrayed cDNA and RNAi libraries provide powerful toolsfor both gain-of-function and loss-of-function cell-based screens atthe genome-wide level (17–21). However, due to the complexity ofmetastasis, no cell-based screen accurately recapitulates the entiretumor dissemination process from the primary tumor to the growthof metastatic tumors in secondary organs. Thus, the application ofanimal models of metastasis to such genetic screens should facili-tate the systematic identification of genes that play critical roles in

this process. In this study, an in vivo selection system in the mousewas used to identify genes from a large cDNA library that com-plement a cell line defective in the early steps of metastasis. Twogene products, �-catenin and the disulfide isomerase ERp5, werefound to cause metastasis to the lung in an orthotopic breast cancermodel. Additional experiments demonstrated that ERp5 promotesbreast cancer cell migration and invasion through activation of theEGF receptor ErbB2/PI3K pathway and downstream signalingmolecules, including Akt and RhoA. In addition, ERp5 was foundto be up-regulated in invasive clinical breast cancer samples. Theseexperiments define a role for the disulfide isomerase ERp5 in tumorinvasion and metastasis and provide an additional mechanism formodulation of ErbB2/PI3K signaling.

Results and DiscussionA Forward Genetic Screen in an Orthotopic Mouse Model for Metas-tasis Promoting Genes. Our genetic screen for metastasis promotinggenes is based on orthotopic transplantation of tumor cells infectedwith a retroviral cDNA library into the mouse mammary fat pad,subsequent selection for metastasis to secondary organs, and finallyretrieval of genes that promote this process (Fig. 1A). An orthotopictransplantation mouse model was used (instead of a tail veininjection model) because transplanted tumor cells in the mammaryfat pad must complete all of the steps of metastasis to spread to thelung. 168FARN cells, originally isolated from a single mousemammary tumor, which arose spontaneously in a wild typeBALB/cJ mouse (22), were chosen for the initial screen. 168FARNcells are able to grow at the primary site but are not able to developmetastatic nodules when transplanted into the mammary fat pad(Fig. 1B). However, 168FARN cells develop lung metastasis wheninjected directly into the blood stream through the tail vein (Fig.1B), indicating that they are defective in the early steps of metastasis(i.e., tumor migration and invasion), but have full metastaticpotential once they reach the blood stream and lung. Therefore, thiscell line is especially suitable for the isolation of genes that promotethe early steps of tumor spread. Any gene that can complement themetastatic defects in this cell line should cause cells to metastasizefrom the mammary fat pad to the lung, which serves as a positiveselection system.

To carry out the screen, a retroviral cDNA library consisting ofcDNAs prepared from day 13.5 embryonic stage mice and ex-pressed behind the LTR promoter was used (23). The incorporationof cDNAs into the host genome after retroviral infection makes it

Author contributions: K.G., P.G.S., and Q.H. designed research; K.G., F.S., A.J.K.-S., and Q.H.performed research; K.G., J.M.G., P.A.G., A.J.S., and Q.H. contributed new reagents/analytictools; K.G., F.S., A.J.K.-S., and Q.H. analyzed data; and K.G., P.G.S., and Q.H. wrote thepaper.

The authors declare no conflict of interest.

Abbreviations: PI3K, phosphoinositide 3-kinase; DTNB, 5,5�-dithiobis(2-nitrobenzoic acid).

††To whom correspondence may be addressed. E-mail: schultz@scripps.edu or qhuang@wistar.org.

© 2007 by The National Academy of Sciences of the USA

6696–6701 � PNAS � April 17, 2007 � vol. 104 � no. 16 www.pnas.org�cgi�doi�10.1073�pnas.0701145104

Dow

nloa

ded

by g

uest

on

Sep

tem

ber

1, 2

020

possible to easily retrieve cDNAs from the positively selected cellsby PCR (23). 168FARN cells (1 � 106) that contain the cDNAlibrary were transplanted into the mouse mammary fat pad (Fig.1A). A proof-of-concept experiment using five BALB/cJ mice wasperformed. Of these five mice, two subsequently developed lungmetastasis in 7 weeks (Fig. 1C). Genes encoding �-catenin andERp5 were retrieved from the metastatic cells of these mice.�-Catenin is a major downstream effector of the Wnt pathway (24,25) that impairs epithelial cell differentiation and induces theepithelial-mesenchymal transition (26–29). In addition, nuclear�-catenin activity is up-regulated in various types of aggressivecancers (30–32). The isolation of a gene that is known to play animportant role in tumor metastasis validates the ability of thisapproach to identify bona fide metastasis promoting genes.

ERp5 is a member of the thiol isomerase family which isresponsible for the formation of native disulfide bonds in cellsurface and secreted proteins (33–37). These enzymes ensure thecorrect pairing of cysteine (Cys) residues in proteins by catalyzingthe formation and rearrangement of disulfide bonds. This processensures the correct folding of their substrates into native confor-mations (33–37). Recent studies have demonstrated that in additionto their localization in the endoplasmic reticulum, some thiolisomerases are also located on the cell surface where they functionin receptor activation and remodeling (38). For example, it has beenshown that ERp5 is recruited to the cell surface and plays a majorrole in platelet aggregation in response to platelet agonists (39).Despite extensive studies of the thiol isomerase family, little is

known regarding their roles in cellular signaling pathways andtumor progression.

To validate the metastasis-promoting activity of ERp5, the genewas cloned into a retroviral vector behind the LTR promoter andintroduced into 168FARN cells to generate a cell line that stablyexpresses ERp5. Overexpression of ERp5 was confirmed by im-munoblot with an ERp5 antibody (data not shown). These cellswere then transplanted into the mammary fat pad using the sameprocedure as in the initial screen. Lung metastasis nodules devel-oped after transplantation, demonstrating that ERp5 promotestumor metastasis in vivo, whereas 168FARN cells containing acontrol vector do not cause metastasis (Fig. 1D).

Two considerations may account for the lack of isolation of othermetastasis-promoting genes in this proof-of-concept study. ThecDNA library that was used had a titer of 107 to 108 cfu, which ismuch larger than the coverage of cDNAs that can be obtained withfive mice. Many genes, especially genes with low copy numbers,might be found using larger numbers of mice. The other possibilityis that the cDNA library was prepared from day 13.5 embryonicmice and contains only genes that are expressed during this stage.An ideal cDNA library for this type of study is one that contains allgenes in the genome, where each gene has equal representation inthe library. Such a library is now being generated for future studies.

ERp5 Promotes Tumor Cell Migration and Invasion in Vitro. Because168FARN cells are defective in the early steps of tumor metastasis,it is likely that ERp5 plays a role in initial tumor migration and

Fig. 1. Identification of ERp5 as a metastasis-promoting gene. (A) The scheme for the forward genetic screen. (B) Model establishment: 168FARN tumor cellgrowth in the primary site after transplantation (1 � 106 cells) to mammary fat pad in BALB/cJ mice (B Upper Left). Dissection of mice 7 weeks posttransplantationshows no metastasis in secondary organs (B Lower). 168FARN cells are able to form lung metastasis after tail vein injection (1 � 106 cells) (B Upper Right). (C andD) Screening and validation: Tumor cell growth in the primary site after injection of 168FARN cells (1 � 106 cells) containing a cDNA library in mammary fat pad.(C) Two of five mice developed lung metastasis. Transplantation of 168FARN cells stably overexpressing ERp5 in mammary fat pad leads to lung metastasis. (D)Mice were imaged 6–8 weeks after transplantation.

Gumireddy et al. PNAS � April 17, 2007 � vol. 104 � no. 16 � 6697

CELL

BIO

LOG

Y

Dow

nloa

ded

by g

uest

on

Sep

tem

ber

1, 2

020

invasion into neighboring tissues. The effect of ERp5 on theseinitial steps was therefore evaluated in a transwell cell migrationassay in vitro. Human ERp5 was cloned into a retroviral vector withits expression driven by the LTR promoter and was subsequentlyintroduced into human breast cancer cell lines MCF7 and MDA-MB-436 to establish cells stably expressing ERp5. Overexpressionof ERp5 in both cell lines was confirmed by immunoblot with anERp5 antibody (data not shown). Overexpression of ERp5 inMCF7 cells results in a migratory phenotype, whereas cells con-taining a control vector are nonmigratory. In addition, MDA-MB-436 cells that stably express ERp5 show more than an 80% increasein the number of migrated cells compared with the cells containinga control vector (Fig. 2A). Similar results were obtained in aninvasion assay in which cells must penetrate Matrigel during mi-

gration. The phenotype of MCF7 cells that overexpress ERp5changes from noninvasive to invasive, whereas MDA-MB-436 cellsoverexpressing ERp5 display more than a 60% increase in thenumber of invasive cells (Fig. 2B). Similar results were obtained in168FARN cells as well. Taken together, these results demonstratethat ERp5 promotes tumor cell migration and invasion.

It has been shown that some disulfide isomerases regulate cellsurface protein function through disulfide exchange at the cellsurface (38–41). The role of disulfide exchange in cell migrationand invasion was examined by treatment of cells with 5,5�-dithiobis(2-nitrobenzoic acid) (DTNB), an activated disulfidewhich reacts with free thiols and thereby inhibits cell surface thiolexchange (38–41). Treatment with 1 mM DTNB reduces migrationof MDA-MB-436 parent and ERp5-overexpressing cells by 40%and 80%, respectively. Furthermore, migration of MCF7 cells thatoverexpress ERp5 is reduced by �60% upon treatment with 1 mMDTNB (compared with DMSO-treated cells) (Fig. 2C). Similarly,1 mM DTNB inhibits invasion of MDA-MB-436 parent and ERp5overexpressing cells by 30% and 80%, respectively, and inhibitsinvasion of MCF7 cells overexpressing ERp5 by �80% (Fig. 2D).In contrast, 1 mM DTNB has no effect on cell growth (Fig. 2E). Theeffect of inhibiting cell surface disulfide exchange is more promi-nent in cells that overexpress ERp5, which may reflect a dominanteffect of thiol exchange on cell migration and invasion when ERp5is expressed at high levels. These results not only show that cellsurface disulfide exchange is critical for the activity of ERp5 on cellmigration and invasion, they also suggest that inhibition of disulfideexchange may be beneficial in the treatment of tumor metastasis.

Activation of the ErbB2 Pathway Is Required for the MetastaticActivity of ERp5. Because ERp5 is known to modulate cell surfacereceptors (38, 40), we next examined its effect on candidate genesknown to play an important role in breast cancer development.ErbB2 (HER2/NEU) is one such gene. It is a member of subclass Iof the receptor tyrosine kinase superfamily (42), which whenactivated leads to the stimulation of many cell signaling pathwaysincluding PI3K-AKT pathways, the mitogen-activated protein ki-nase pathway, and the SRC tyrosine kinase pathway (43, 44). Theexpression of ErbB2 is altered in many epithelial tumors such asbreast, ovarian, gastric, and non-small-cell lung cancers (45–48).ErbB2 expression has also been shown to be inversely correlatedwith the prognosis of breast cancer, indicating an important role intumor metastasis (45, 49, 50).

Overexpression of ERp5 did not affect ErbB2 expression levelsin MCF7 and MDA-MB-436 cells as determined by immunoblotanalysis (Fig. 3A). However, overexpression of ERp5 led to acti-vation of ErbB2 as determined by immunoprecipitation and West-ern blot analysis of cell extracts from MCF7 and MDA-MB-436cells. The level of ErbB2 phosphorylation is significantly higher inMCF7 and MDA-MB-436 cells expressing ERp5 than in cells withan empty vector (Fig. 3B). Similar results were obtained in168FARN cells as well. ErbB2 triggers signal transduction byligand-dependent heterodimerization or ligand-independent ho-modimerization (42, 51, 52). Consistent with this mechanism, ERp5expression induces dimerization of ErbB2 in both MCF7 andMDA-MB-436 cells (Fig. 3C). To determine whether ErbB2 isrequired for the activity of ERp5 in tumor cell migration andinvasion, a short hairpin RNA (shRNA) construct was used toknock down expression of ErbB2 in MCF7 and MDA-MB-436 cellsthat overexpress ERp5. Western blot analysis demonstrated thatthe shRNA is effective in suppressing ErbB2 when compared witha control shRNA (Fig. 3D). Migration and invasion assays werethen performed to determine whether ERp5 activity depends onErbB2 expression. The number of migrated MCF7 and MDA-MB-436 cells that express both ERp5 and the shRNA against ErbB2 isreduced by �90% and 85%, respectively, in comparison to cellsexpressing ERp5 and a control shRNA (Fig. 3D). Similar resultswere obtained in the invasion assay: invasive MCF7 and MDA-

Fig. 2. ERp5 promotes tumor cell migration and invasion in vitro. MDA-MB-436 and MCF7 cells were serum-starved for 24 h before the transwell assay.Complete medium was added to the bottom wells of the transwell chambers.Cells that migrated to the lower surface of the filters were fixed and counted12 h after they were added to the upper chamber. Stable ERp5 overexpressionin human breast cancer MDA-MB-436 cells causes 100% more migrated cells(A) and 60% more invaded cells (B) in a transwell assay compared with thesecells with a control vector. The expression of ERp5 in human breast cancerMCF7 cells causes a phenotype change from nonmigratory and noninvasive tosignificant migration (A) and invasion (B). Suppression of cell surface disulfideexchange blocks cell migration and invasion. Treatment of cells with 1 mMDTNB reduces migration of MDA-MB-436 cells by 40%, reduces migration ofMDA-MB-436 cells overexpressing ERp5 by �80%, and reduces migration ofMCF7 cells overexpressing ERp5 by �60% compared with DMSO treatment asa control (C). DTNB (1 mM) has a similar effect on cell invasion: MDA-MB-436cell invasion is inhibited by 30%; invasion by MDA-MB-436 cells overexpress-ing ERp5 is inhibited by 80%; and invasion by MCF7 cells overexpressing ERp5is inhibited by �80% (D). DTNB (1 mM) has no effect on cell growth (E).

6698 � www.pnas.org�cgi�doi�10.1073�pnas.0701145104 Gumireddy et al.

Dow

nloa

ded

by g

uest

on

Sep

tem

ber

1, 2

020

MB-436 cells that express both ERp5 and the shRNA againstErbB2 are reduced by �95% compared with cells that expressERp5 and a control shRNA (Fig. 3D). Interestingly, the effect ofErbB2 knockdown on migration and invasion in MCF7 and MDA-MB-436 cells overexpressing ERp5 is more prominent than in cellscontaining the control vector (Fig. 3 C and D). This may again bedue to a dominant effect of ERp5 on migration and invasion. Theseresults demonstrate that activation of the ErbB2 pathway is re-quired for the effects of ERp5 on tumor migration and invasion.

ERp5 Activates PI3K and Its Downstream Signaling Molecules RhoAand �-Catenin. Activation of ErbB signaling results in a number ofdownstream events, including activation of the PI3K signalingcascade. The effects of ERp5 on PI3K activity were thereforeexamined by measuring the phosphorylation status of the p85subunit of PI3K, which correlates with PI3K kinase activity in vivo(53, 54). p85 phosphorylation was determined by immunoprecipi-tation using a phosphotyrosine antibody followed by Western blotanalysis by using an antibody specific for p85. The level of activePI3K is significantly higher in MDA-MB-436 cells expressing ERp5than in those cells containing a control vector, whereas the totalamount of PI3K is similar in these samples, demonstrating thatPI3K is activated by ERp5 (Fig. 4A). The PI3K inhibitors LY294002and wortmannin were then used to examine whether PI3K isrequired for the effects of ERp5 on migration and invasion. Asshown in Fig. 4B, the addition of 10 �M LY294002 reducesmigration of MCF7 and MDA-MB-436 cells that stably expressERp5 by �90%; likewise, treatment with 0.1 �M wortmanninreduces migration by �85% in these cells (Fig. 4B) (these PI3Kinhibitors were not tested on MCF7 control cells because these cellshave minimal migration and invasion capacity in the absence ofERp5 overexpression). Both compounds have no effect on cellgrowth at the concentrations that were used in the experiments(data not shown). LY294002 also reduced invasion of MCF7 andMDA-MB-436 cells that stably express ERp5 by �85% and 95%,respectively; wortmannin had similar effects (Fig. 4B). Takentogether, these results strongly suggest that the PI3K pathway isactivated by ERp5 and is required for the induction of cell migrationand invasion by ERp5.

The activation of downstream signaling molecules in the PI3Kpathway by ERp5 was examined next. RhoA is one of the down-stream targets of PI3K and activation of RhoA is known to promotecell migration (55). Affinity precipitation of active RhoA by Rho-tekin binding domain beads in the cell extracts of MCF7 andMDA-MB-436 cells overexpressing ERp5 revealed higher expres-sion levels of the active form of RhoA when compared with cellscontaining a control vector; the total RhoA levels are similar in the

presence or absence of ERp5 (Fig. 4C). Another major downstreamtarget of activated PI3K is the serine-threonine kinase Akt, whichis activated when it is phosphorylated at Thr308 and Ser473 (56, 57).As shown in Fig. 4C, the total protein levels of Akt are similar incells stably expressing either ERp5 or a control vector. However,the phosphorylated forms (active forms) of Akt at both amino acids308 and 473 are significantly increased in cells expressing ERp5(Fig. 4C). To further explore the downstream consequences ofPI3K/Akt activation, we examined the activation of �-catenin byAkt. Western blot analysis confirmed that the expression levels of�-catenin are significantly increased in MCF7 and MDA-MB-436cells stably expressing ERp5 (Fig. 4C). These results suggest thatactivation of the PI3K pathway is responsible for tumor cellmigration and invasion induced by ERp5 (Scheme 1).

ERp5 Is Up-Regulated in Invasive Clinical Breast Cancer Samples.Finally, tissue microarrays were used to examine the expression ofERp5 in clinical breast cancer samples at various stages. Immuno-staining of ERp5 showed that normal breast ducts as well as earlystages of breast cancer (including ductal carcinoma in situ and T1N0samples) exhibited little or no ERp5 expression. However, ERp5expression was increased in more advanced stages of invasive ductalcarcinoma (T2N0 and T2N1) (Fig. 5A): 60–100% of carcinoma cellswere stained exclusively in the cytoplasm. None of the normalbreast samples and only 4.3% of the T1N0 samples received 2�scores, whereas 21.7% of the T2N0 samples and 37.5% of the T2N1

Fig. 3. ERp5 activates ErbB2 pathway.(A) The expression level of total ErbB2does not change after ERp5 overexpres-sion in human breast cancer MDA-MB-436and MCF7 cells. (B) Immunoprecipitationof phosphorylated ErbB2 demonstratesErbB2 is activated by ERp5 overexpressionin MDA-MB-436 and MCF7 cells. (C) Immu-noblots under nonreducing conditionshow that ERP5 promotes dimerization ofErbB2 in MDA-MB-436 and MCF7 cells. (D)Knockdown of ErbB2 by a specific shRNAreduces the effect of ERp5 on migrationand invasion in MDA-MB-436 and MCF7cells.

Scheme 1. The molecular mechanisms of ERp5 in promoting tumor invasionand metastasis.

Gumireddy et al. PNAS � April 17, 2007 � vol. 104 � no. 16 � 6699

CELL

BIO

LOG

Y

Dow

nloa

ded

by g

uest

on

Sep

tem

ber

1, 2

020

samples scored 2� or higher (Fig. 5B), demonstrating that ERP5is up-regulated in more invasive breast cancer samples. The P valueof Fisher’s exact test is 0.0034, indicating that there is indeed an

association between the staining results and disease progression.These results further support the role of ERp5 in tumor progres-sion. Interestingly, macrophages were strongly stained and endo-thelial cells were also moderately stained with ERp5 antibody,which may correlate with their enhanced mobility relative to normalepithelial cells.

The immunostaining results with clinical breast cancers suggestthat tumors with no ErbB2 amplification can still be aggressive dueto the activation of ErbB2 by mechanisms other than gene ampli-fication. Therefore, the activation of the ErbB2 pathway, ratherthan the amplification or overexpression of ErbB2, may be anotherindication of the need for chemotherapies targeting ErbB2. Thefinding that tumor cells overexpressing ERp5 are sensitive to ErbB2and PI3K inhibition in migration and invasion assays (Figs. 3D and4B) suggests that chemotherapy targeting ErbB2 and PI3K may beeffective in patients in which ERp5 expression is up-regulated.

ConclusionIn summary, we have applied a novel gain-of-function in vivoselection system to identify the thiol isomerase ERp5 as an activatorin the early steps of metastasis. ERp5 promotes tumor cell migra-tion and invasion by activating the ErbB2/PI3K signaling pathway(Scheme 1). To our knowledge, this is the first report demonstratingthe amplification of ErbB2/PI3K signaling by a disulfide isomerase.The activation of ErbB2 by ERp5 leads to the activation of PI3Kand, subsequently, RhoA and �-catenin, which have been shown toplay important roles in tumor invasion and metastasis (Scheme 1).Together with the finding that �-catenin was also identified in thisforward genetic screen, these results demonstrate that the activationof ErbB2/PI3K pathway is critical in breast cancer metastasis,especially in the early stages of the process. Besides its location inthe endoplasmic reticulum, ERp5 also exists on the cell surfacebased on a previous publication (39) and our own observation(unpublished data). Because the majority of cell surface proteinscontains at least one disulfide bond, these proteins may requiredisulfide isomerases for their maturation. Alternatively, cell surfacedisulfide exchange may play an important role in the activation andregulation of cell signaling pathways (58). In either case, theselective inhibition of cell surface thiols may serve as a potentialtherapeutic approach to the suppression of tumor progression andmetastasis.

Materials and MethodsPlasmids and Reagents. Human and mouse ERp5 genes were clonedinto the pBabe-puromycin plasmid and expressed by the LTRpromoter. shRNA against ErbB2 and control shRNA were pur-chased from Open Biosystems (Huntsville, AL). Anti-ERp5 anti-body was prepared as previously described (39). Anti-human Akt,phospho-Akt-(Ser-473), phospho-Akt-(Thr-308), and PI3K inhibi-tor Ly294002 were purchased from Cell Signaling Technology(Beverly, MA); �-catenin was purchased from BD TransductionLaboratories (San Jose, CA); and ErbB2, phosphorylated tyrosine,clone 4G10, PI3-kinase p85, PI3K inhibitor wortmannin, and theRhoA activation assay kit were purchased from Upstate CellSignaling Solutions (Lake Placid, NY).

Cell Culture. The mouse breast cancer cell line 168FARN andhuman breast cancer cell line MCF7 and MDA-MB-436 werecultured in DMEM with 10% FBS. All cells were incubated at 37°Cwith 5% CO2.

Recombinant Retroviruses and Viral Infection. Virus production andinfection were performed as previously described (21).

Immunoblotting and Immunochemistry. Immunoblotting and immu-noprecipitation were performed as previously described (21). Rhoactivity was measured by affinity precipitation of GTP-Rho with

Fig. 4. ERp5 activates PI3K pathway. (A) Immunoprecipitation of phosphory-lated PI3K in MDA-MB-436 and MCF7 cells demonstrates that ERp5 overexpres-sion activates PI3K, whereas it has no effect on the expression level of total PI3K.(B) PI3K inhibitors Ly294002 (10 �M) and wortmannin (0.1 �M) reverse the effectof ERp5 in the migration and invasion assay in MDA-MB-436 and MCF7 cells. (C)Stable ERp5 overexpression in MDA-MB-436 and MCF7 cells leads to the activa-tion of RhoA and Akt but has no effect on the total protein expression level ofthese two proteins. The activation of Akt subsequently up-regulates �-catenin.

Fig. 5. ERp5 expression is up-regulated in the clinical tumor samples of invasiveductal carcinoma. (A) Immunostaining with an ERp5 antibody of the tumor tissuearray containing normal breast tissues and breast ductal carcinoma at variousstages was scored on a scale of 0 to 3: 0 represents no staining, and 3� representsthe strongest staining. ERp5 expression is elevated in T2N0 and T2N1 comparedwith normal breast duct and T1N0 tumors. (B) The percentage of tumor sampleson the array with different scores was tabulated. The percentage of tumorsamples with 2� scores is higher in invasive breast cancer: 21.7 and 37.5% of T2N0and T2N1 stage, respectively, compared with 0% in normal breast ductal struc-ture and 4.3% in T1N0 tumors.

6700 � www.pnas.org�cgi�doi�10.1073�pnas.0701145104 Gumireddy et al.

Dow

nloa

ded

by g

uest

on

Sep

tem

ber

1, 2

020

Rhotekin-agarose beads (Upstate Biotechnology, Lake Placid,NY) according to the manufacturer’s instructions.

Tissue microarrays of formalin-fixed, paraffin-embedded sam-ples of human breast epithelium representing normal as well asdifferent stages of tumor progression were prepared at Fox ChaseCancer Center. A total of 16 normal breast areas, 23 T1N0, 23T2N0, and 24 T2N1 breast cancers cases, two cores for each sample,represented by a total of 172 cores were used in the tissuemicroarrays. Immunostaining was performed as previously de-scribed (59). ERp5 antibody (1:100) was used. Cytoplasmic stainingwas scored according to the stain intensity from 0 (no stain), 1(marginal to moderate stain), 2 (moderate to intense stain), and 3(strong intense stain).

Invasion and Migration Assays. Matrigel invasion assays were per-formed as described (60) using transwell chambers (8-�M pore size;Costar). In some experiments, cells were preincubated with thePI3-kinase inhibitors LY294002 (100 �M), wortmannin (100 nM),DTNB (1 mM), or control medium for 30 min. After detachmentwith trypsin, cells were washed with PBS and resuspended inserum-free medium, and 250-�l cell suspension (2 � 105 cells/ml)was added to the upper chamber. DTNB (1 mM) was added to thecell suspension. Images of three different �10 fields were capturedfrom each membrane, and the number of invading cells wascounted. The mean of triplicate assays for each experimentalcondition was used. For migration assays, the same procedure wasperformed except that transwell chambers were not coated withmatrigel.

Functional Screen in an Orthotopic Animal Model. To establishluciferase positive cancer cells, the luciferase gene from pGL3(Promega, San Luis Obispo, CA) was cloned into pBabe-

puromycin. 168FARN cells were infected with retroviruses con-taining luciferase to establish 168FARN-Luc cells that stably ex-press luciferase. To establish the animal model for the forwardgenetic screen, 1 � 106 168FARN-Luc cells were orthotopicallytransplanted into the mammary fat pads of 10-week-old BALB/cJmice or injected intravenously by tail vein. Mice bearing luciferasepositive tumors were imaged 7–8 weeks after transplantation on theXenogen IVIS system (Xenogen Corporation, Hopkinton, MA).Mice injected by tail vein were imaged 4–5 weeks after transplan-tation to monitor tumor growth in the lung. To perform the forwardgenetic screen, retroviruses containing a cDNA library were gen-erated as described (23). 168FARN-Luc cells were infected with theviruses containing the library at 1 multiplicity of infection.168FARN-Luc cells (1 � 106) containing the library were trans-planted into the mammary fat pad of each mouse. Lung metastasisnodules were isolated 6–7 weeks after transplantation. GenomicDNA of metastatic cells was isolated with the Genomic DNAPurification kit according to supplier’s instruction (Qiagen, Valen-cia, CA). PCRs were performed with the Expand High Fidelity kit(Roche Diagnostics, Indianapolis, IN) with primers as described(23). PCR products were cloned with a TA cloning kit (Invitrogen)and sequenced. To validate the hits from the screen, 168FARN-Luccells were infected with retroviruses containing the full-lengthmouse ERp5 behind the LTR promoter; 1 � 106 cells weretransplanted into the mammary fat pads of mice and imaged 6–8weeks after transplantation.

We thank Dr. Fred Miller for 168FARN cells and Dr. George Daley forthe cDNA library. Q.H. is supported by the Breast Cancer Alliance, thePardee Foundation, and the Commonwealth Universal Research En-hancement Program (Pennsylvania Department of Health). We alsoacknowledge support for this work by the Novartis Research Foundation.

1. Fidler IJ (2003) Nat Rev Cancer 3:1–6.2. Gupta GP, Massague J (2006) Cell 127:679–695.3. Steeg P (2006) Nat Med 12:895–904.4. Welch DR, Steeg PS, Rinker-Schaeffer CW (2000) Breast Cancer Res 2:408–416.5. Chambers AF, Naumov GN, Vantyghem SA, Tuck AB (2000) Breast Cancer Res

2:400–407.6. Hartwell KA, Muir B, Reinhardt F, Carpenter AE, Sgroi DC, Weinberg RA (2006)

Proc Natl Acad Sci USA 103:18969–18974.7. Yang J, Mani SA, Donaher JL, Ramaswamy S, Itzykson RA, Come C, Savagner P,

Gitelman I, Richardson A, Weinberg RA (2004) Cell 117:927–939.8. Kang Y, Siegel PM, Shu W, Drobnjak M, Kakonen SM, Cordon-Cardo C, Guise TA,

Massague J (2003) Cancer Cell 3:537–549.9. Clark EA, Golub TR, Lander ES, Hynes RO (2000) Nature 406:532–535.

10. Brinckerhoff CE, Matrisian LM (2002) Nat Rev Mol Cell Biol 3:207–214.11. Yu Q, Stamenkovic I (2004) Clin Exp Metastasis 21:235–242.12. Cairns RA, Khokha R, Hill RP (2003) Curr Mol Med 3:659–671.13. Cavallaro U, Christofori G (2004) Ann NY Acad Sci 1014:58–66.14. Grimm S (2004) Nat Rev Genet 5:179–189.15. Dasgupta R, Kaykas A, Moon RT, Perrimon N (2005) Science 308:826–832.16. Fraser AG, Kamath RS, Zipperlen P, Martinez-Campos M, Sohrmann M, Ahringer

J (2000) Nature 408:325–330.17. Friedman A, Perrimon N (2004) Curr Opin Genet Dev 14:470–476.18. Paddison PJ, Silva JM, Conklin DS, Schlabach M, Li M, Aruleba S, Balija V,

O’Shaughnessy A, Gnoj L, Scobie K, et al. (2004) Nature 428:427–431.19. Berns K, Hijmans EM, Mullenders J, Brummelkamp TR, Velds A, Heimerikx M,

Kerkhoven RM, Madiredjo M, Nijkamp W, Weigelt B, et al. (2004) Nature 428:431–437.

20. Kolfschoten IG, van Leeuwen B, Berns K, Mullenders J, Beijersbergen RL, BernardsR, Voorhoeve PM, Agami RA (2005) Cell 121:849–858.

21. Huang Q, Raya A, DeJesus P, Chao S, Quon KC, Caldwell JS, Chanda SK,Izpisua-Belmonte JC, Schultz PG (2004) Proc Natl Acad Sci USA 101:3456–3461.

22. Aslakson CJ, Miller FR (1992) Cancer Res 52:1399–1405.23. Koh EY, Chen T, Daley GQ (2002) Nucleic Acids Res 30:e142p1–e142p7.24. Korinek V, Barker N, Morin PJ, van Wichen D, de Weger R, Kinzler KW, Vogelstein

B, Clevers H (1997) Science 275:1784–1787.25. Bienz M, Clevers H (2000) Cell 103:311–320.26. Eger A, Stockinger A, Schaffhauser B, Beug H, Foisner R (2000) J Cell Biol

148:173–188.27. Morali OG, Delmas V, Moore R, Jeanney C, Thiery JP, Laure L (2001) Oncogene

20:4942–4950.28. Mariadason JM, Bordonaro M, Aslam F, Shi L, Kuraguchi M, Velcich A, Augenlicht

LH (2001) Cancer Res 61:3465–3471.29. Naishiro Y, Yamada T, Takaoka AS, Hayashi R, Hasegawa F, Imai K, Hirohashi S

(2001) Cancer Res 61:2751–2758.

30. Brabletz T, Hermann K, Jung A, Faller G, Kirchner T (2000) Am J Pathol156:865–870.

31. Gamallo C, Palacios J, Moreno G, Calvo de Mora J, Suarez A, Armas A (1999) Am JPathol 155:527–536.

32. Zheng Z, Pan J, Chu B, Wong Y, Cheung AL, Tsao S (1999) Human Pathol30:458–466.

33. Ellgaard L, Ruddock LW (2005) EMBO Rep 6:28–32.34. Ferrari DM, Soling H (1999) Biochem J 339:1–10.35. Sevier CS, Kaiser CA (2002) Nat Rev Mol Cell Biol 3:836–846.36. Woycechowsky KJ, Raines RT (2000) Curr Opin Chem Biol 4:533–539.37. Bulleid NJ (2003) Curr Biol 13:R380.38. Turano C, Coppari S, Altieri F, Ferraro A (2002) J Cell Physiol 193:154–163.39. Jordan PA, Stevens JM, Hubbard GP, Barrett NE, Sage T, Authi KS, Gibbins JM

(2005) Blood 105:1500–1507.40. Ellerman DA, Myles DG, Primakoff P (2006) Dev Cell 10(6):831–837.41. Markovic I, Stantchev TS, Fields KH, Tiffany LJ, Tomic M, Weiss CD, Broder CC,

Strebel K, Clouse KA (2004) Blood 103:1586–1594.42. Hynes NE, Lane HA (2005) Nat Rev Cancer 5:341–354.43. Schlessinger J (2004) Science 306:1506–1507.44. Yarden Y, Silwkowski MX (2001) Nat Rev Mol Cell Biol 2:127–137.45. Slamon DJ, Clark GM, Wong SG, Levin WJ, Ullrich A, McGuire WL (1987) Science

235:177–182.46. Falck VG, Gullick WJ (1989) J Pathol 159:107–111.47. Tateishi M, Ishida T, Mitsudomi T, Kaneko S, Sugimachi K (1991) Eur J Cancer

27:1372–1375.48. Stenman G, Sandros J, Nordkvist A, Mark J, Sahlin P (1991) Genes Chromosomes Cancer

3:128–135.49. Yu D, Wang SS, Dulski KM, Tsai CM, Nicolson GL, Hung MC (1994) Cancer Res

54:3260–3266.50. Li YM, Pan Y, Wei Y, Cheng X, Zhou BP, Tan M, Zhou X, Xia W, Hortobagyi GN,

Yu D, et al. (2004) Cancer Cell 6:459–469.51. Weiner DB, Liu J, Cohen JA, Williams WV, Greene MI (1989) Nature 339:230–231.52. Schlessinger J (2002) Cell 110:669–672.53. Fruman DA, Meyers RE, Cantley LC (1998) Annu Rev Biochem 67:481–507.54. Pleiman CM, Hertz WM, Cambier JC (1994) Science 263:1609–1612.55. Raftopoulou M, Hall A (2004) Dev Biol 265:23–32.56. Rane MJ, Coxon PY, Powell DW, Webster R, Klein JB, Pierce W, Ping P, McLeish

KR (2001) J Biol Chem 276:3517–3523.57. Testa JR, Bellacosa A (2001) Proc Natl Acad Sci USA 98:10983–10985.58. Jordan P, Gibbins JM (2006) Antioxid Redox Signal 8:312–324.59. Lopez de Cicco R, Bassi DE, Zucker S, Seidah, NG, Klein-Szanto AJ (2005) Cancer

Res 65:4162–4171.60. Shaw LM, Rabinovitz I, Wang HH, Toker A, Mercurio AM (1997) Cell 91:949–960.

Gumireddy et al. PNAS � April 17, 2007 � vol. 104 � no. 16 � 6701

CELL

BIO

LOG

Y

Dow

nloa

ded

by g

uest

on

Sep

tem

ber

1, 2

020