Proc. Natl. Acad. Sci. USA
`Vol. 84, pp. 1159—7163, October 193?
`Cell Biology
`
`Increased expression of the putative growth factor receptor
`p185HERz causes transformation and tumorigenesis of NIH 3T3 cells
`
`ROBERT M. Huozmx“, JOSEPH Scntessruoeai, AND Axcr. ULLRICH“
`
`’Department of Developmental Biology. Genentech. Inc. 460 Point San Bruno Boulevard. Smith San Francisco. CA 91080; and lBiotechnology Research
`Center. Meloy Laboratories. 4 Research Conn, Rockville. MD 20850
`
`Communicated by Hilda! Koprowski. July 13. HS?
`
`The HERZ gene encodes a cell-stu'face glyco-
`ABSTRACT
`proteln with extensive homology to the epidermal growth factor
`receptor. Recently it was found to be amplified in about 30%
`of primary human breast malignancies. In experiments de-
`signed to assess the role of the HERZ gene in oncogenesls, we
`found that overexpression of unaltered HER2 coding sequences
`in NIH 3T3 cells resulted in cellular transformation and
`tumor-igenesis.
`
`The HER2 gene encodes a transmembrane glyc0protein with
`extensive structural homology to the human epidermal
`growth factor (EGF) receptor and the chicken oncogene
`v-erbB (1—3}. Chromosomal mapping and sequence compar-
`ison strongly suggest that the HERZ gene product and the
`ethylnitrosounea-activated, rat neuroblastoma oncogene neu
`represent species variants of the same polypeptide (4). The
`near oncogene encodes a 185-kDa cell-surface glycoprotein
`that possesses intrinsic tyrosine-specific kinase activity that
`is likely to be activated by an as yet unidentified ligand (5, 6).
`Comparison of the transforming neu oncogene sequence with
`its normal rat protooncogene counterpart suggested that a
`point mutation in the transmembrane domain resulting in
`substitution of a valine residue by glutamate unmasked the
`transforming potential of this putative growth factor receptor
`(7). Analogously. structural alterations have converted nor-
`mal genes coding for the receptors for macrophage colony-
`stimulating factor type 1 and EGF into v-fms (8) and v-erbB
`(9) oncogenes. respectively.
`Southern analysis of primary human tumors and estab-
`lished tumor-derived cell lines revealed amplification and in
`some cases rearrangement of the EGF receptor gene. Am-
`plification was particularly apparent in squamous carcinomas
`(10. 11) and glioblastomas (12). The HER? gene was also
`found to be amplified in a human salivary gland adenocarci-
`noma {3). a mammary gland carcinoma (2), and a gastric
`cancer cell line {13). Recently. Slamon et at. (14) demOn-
`strated that about 30% of primary human breast carcinoma
`tumors contained an amplified HERZ gene. Although a few
`sequence rearrangements were detected.
`in most tumors
`there were no obvious differences between amplified and
`normal HERZ genes. Furthermore, amplification of the
`HERZ gene correlated significantly with the prognosis of the
`disease and the probability of relapse.
`To investigate the significance of the correlation between
`overexpression and cellular transformation as it has been
`observed for protooncogenes c-mos (15) and c-Ha-rasi (16).
`we employed a HER2 expression vector and a selection
`scheme that permitted sequence amplification after transfec-
`tion of mouse NIH 3T3 cells. We report here that amplifi-
`cation of the unaltered HERZ gene in NIH 3T3 cells leads to
`overexpression of p185“ E“: as well as cellular transformation
`and tumor formation in athymic mice. These findings,
`in
`
`combination with the results of Slamon at at. (14), suggest
`that mere amplification of the HERZ gene and resulting
`overexpression of its product may play a crucial role in the
`genesis and development of some types of human cancer.
`
`MATERIALS AND METHODS
`
`Expression Plasmids. The mammalian expression vector
`CVN (17) contained expression units for mouse dihydrofolate
`reductase (DHFR) cDNA (18) and the bacterial neomycin
`phosphotransferase (neo) gene (19), both under simian virus
`40 early promoter control. Transcription of a 4.4-kilobase-
`pair Sal l—Dm I HERZ fragment containing the full-length
`HERZ coding region (I) was driven by the Rous sarcoma
`virus (RSV)
`long terminal repeat promoter (LTR). The
`poly(A) site was provided by the 3' untranslated sequence of
`the hepatitis B virus surface antigen gene (20}. The control
`CVN plasmid was identical but
`lacked cDNA sequences
`downstream from the RSV LTR.
`Cell Culture. NIH 3T3 cells were cultured in a 1:1 mixture
`of Dulbecco’s modified Eagle‘s medium and Ham's nutrient
`mixture F—12 supplemented with glutamine (2 mM), penicillin
`(100 unitsy’ml). streptomycin {100 ngfml). and 10% HyClone
`(Logan. Utah) calf serum in a humidified incubator under 5%
`C02 in air atmosphere.
`Transfections and Amplification. Plasmid DNA was intro—
`duced into mammalian cells by the calcium phosphate
`coprecipitation method (21). Half-confluent plates of cells (60
`mm) were exposed to 5 ,ug of plasmid DNA in 1 ml of
`precipitate for 6-8 hr. After a 20% (volfvol) glycerol shock
`(22}, the cells were fed with nonselective medium. Two days
`later. they were passaged into selective medium containing
`Geneticin (G418) at 400 ugfml.
`Clones were picked using glass cloning cylinders with
`petroleum jelly for the bottom seal. Colonies arising from
`transfected cells selected for growth in G418 were picked.
`expanded, and subcultured into medium containing 'i%
`dialyzed fetal bovine serum in place of 10%calfserum and the
`appropriate concentration of methotrexate for plasmid am-
`plification (23). The dialysis step removes trace amounts of
`purines and pyrimidines present in serum that decrease the
`efficiency of the methotrexate selection. To apply selective
`pressure. stepwise increasing concentrations of methotrex-
`ate were used with a final concentration of 400 nM. To avoid
`enriching for spontaneously transformed cells. cells were
`kept subc0nfluent. An additional control was to amplify the
`CVN neo-DHFR vector without the HERZ cDNA insert in
`the NIH 3T3 recipient cell line.
`Immunopreclpltations and Labeling. The G-HZCTI? anti-
`body recognizing the C-terminal 1? amino acids of HERZ was
`prepared in rabbits using a synthetic peptide conjugated with
`soybean trypsin inhibitor.
`
`The publication costs of this article were defrayed in part by page charge
`payment. This article must therefore be hereby marked "advertisement"
`in accordance with 18 U.S.C. §1734 solely to indicate this fact.
`
`AbbreViations: DHFR. dihydrofolate reductase; R, resistance: EGF.
`epidermal growth factor; RSV. Rous sarcoma virus; LTR.
`long
`terminal repeat.
`
`T159
`
`A
`
`1 of 5
`
`Celltrion, Inc., Exhibit 1047
`
`

`

`7160
`
`Cell Biology: Hudziak er of.
`
`Proc. Natl. Acad. Sci. USA 84 (1987)
`
`Cells were harvested by trypsinization and counted by
`Coulter Counter. and 1.5 X 10" cells were plated per 60-mrn
`culture dish. After 36 hr, the cells were lysed at 4°C with 0.4
`ml ofHNEG buffer per plate (50 mM Hepes. pH 1.53050 mM
`Nalel mM ESTA/10% glycerol) containing 1.0% Triton
`X-100 detergent and 1 mM phenylmethylsulfonyl fluoride.
`After 10 min, 0.8 ml of lysis dilution buffer (HNEG buffer
`with 1% bovine serum albumin and 0.1% Triton X-100) was
`added to each plate and the extracts were pelleted at 12,000
`x g for 5 min.
`HER2 antibody was added to the cell extracts. which were
`then incubated at 4°C for2 hr; this was followed by incubation
`with protein A-Sepharose beads for 20 min and three washes
`with 1 ml of HNEG buffer with 0.1% Triton X-100.
`Autophosphorylation reactions were can‘ied out at 4°C in 50
`,ul of HNEG wash buffer with 5 mM MnClz and 3 luCi of
`[r’zPMTP {Amersham, 5000 Cilmmol; 1 Ci = 3? GBq) for
`20 min. Proteins were separated on 1.5% NaDodSOrfpoly-
`acrylamide gels and analyzed by autoradiography.
`Transformation Assays. The efficiency of colony formation
`in soft agar (24) was determined by plating 25 .000 cells in 3
`ml of0.2% agar {Difco, “purified") over 4 ml of0.4% agar in
`a 60-mm dish. After 2—4 weeks, colonies of about 100 cells or
`more were counted.
`The plating efficiency of cell lines (25) in 1% calf serum was
`determined by plating equal numbers of cells into 100-mm
`plates with either 10% or 1% calf serum. Alter 2—3 weeks. the
`plates were stained with crystal violet and col0nies were
`counted.
`
`Mouse Tumorigenicity Assays. Athymic (unfair) mice were
`obtained from Charles River Breeding Laboratories. Control
`NIH 3T3 and NIH 3T3XCVN cells and experimental HERZ-
`3409 cells were harvested by trypsinization and counted with
`a Coulter Counter. They were then collected by low—speed
`centrifugation and resuspended in ice»cold phosphate-buff-
`ered saline to either 2.5 X 10", 5.0 X 10", or 1.0 X 10? cells
`per ml. Animals were injected subcutaneously with 0.1-ml
`volume of the cell suspensions. Tumor occurrence and size
`were monitored twice weekly.
`
`RESULTS
`
`For expression of HERZ sequences in NIH 3T3 cells, a cDNA
`coding for the entire 1255-amino acid polypeptide (1) was
`placed under transcriptional control of the RSV LTR. Tran-
`scriptional termination signals and a polytA} site were pro-
`vided by 3‘ sequences of the hepatitis virus surface antigen
`gene (20). in addition. the expression vector contained the
`neo resistance (neck) gene, which confers cellular resistance
`to the aminoglycoside antibiotic G418 (18) and therefore
`allows selection of primary transfectants, as well as the
`DHFR gene for methotrexate resistance, which was used to
`amplify transfected DNA sequences under selective pres-
`sure. Both drug resistance genes were under simian virus 40
`early promoter transcriptional control. Bacterial plasmid
`sequences. including an origin of replication and the gene for
`ampicillin resistance . allowed replication of the entire expres-
`sion plasmid in Escherichia coli.
`The transforming activity of HERZ sequences was initially
`tested using a conventional NIH 3T3 cell focus-formation
`assay. Under conditions that resulted in about 104 foci per pg
`of a v-fms viral construct, we were unable to detect any
`HER2Ltransforming activity. Because ofthe recently report~
`ed finding that about 30% of mammary carcinomas contain
`amplified HERZ gene sequences without apparent sequence
`rearrangements (14), we investigated whether amplification
`of an unaltered HERZ gene could transform mouse fibro-
`blasts. NIH 3T3 cells were transfected with the pCVNg'HERZ
`construct. An identical plasmid missing the HER2 expression
`module was used as a control. Four independent primary
`
`Table 1. Assay for growth in soft agar and 1% calf serum of
`HER2 primary and amplified cell lines
`
`Plating
`efficiency
`Soft agar
`HER2 gene copies
`in 1% calf
`colonies.
`per haploid
`serum. %
`no.
`genome. no.
`Cell line
`0.27
`0
`0
`NIH 3T3/CVN
`0
`0
`0
`NIH 3T3fCVNm
`3
`1
`3
`HERZ-l
`33
`424
`60
`HERZ-lm
`1.4
`0
`2
`HER2-3
`11.7
`836
`55
`HERZ-Jm
`0.2
`0
`4
`HER2-4
`49
`376
`90
`HER2-4m
`0.6
`0
`1
`HER2-BB
`50.2
`373
`131
`HERZ-B3m
`Two control lines were used. The first one was a NIH 3T3 line
`transfected with a plasmid containing only the net) and DHFR genes.
`The second control line contained the neo—DHFR plasmid and was
`amplified to resistance to 400 mM methotrexate. HERZ gene copy
`numbers were determined using a human DNA standard and
`densitorneter scanning of Southern hybridization autoradiograms.
`Equal cell numbers (25,000) were plated in soft agar and colonies
`were counted after 2—4 weeks. The plating efficiency in 1% calf
`serum is relative to the number of colonies arising when an equal
`aliquot was simultaneously plated in medium containing 10% calf
`serum.
`
`CHIS-resistant clones (HERZ-l. HERE-3. HER2-4. HERZ-
`BJ} were isolated. Cell
`lines containing amplified HERZ
`coding sequences were generated from these parental clones
`by culturing the cells in gradually increasing cancentrations
`of methotrexate up to 400 nM (HERZ-lm, HEM-3m,
`HEM-4400, HEM-384m). Southern hybridization analysis of
`parental and amplified cell lines demonstrated that the HER2
`cDNA copy number increased from 1—4 to 55—131 per haploid
`genome (Table 1}.
`To test whether gene amplification resulted in overexpres-
`sion of the HERZ gene product, cell
`lysates were im~
`munoprecipitated with an antibody against the C-terminal 1?
`amino acids of the HER2 sequence. As shown in Fig. 1,
`substantially increased levels of the p185”:Eu gene product
`were found in amplified cell lines relative to their parental
`C5418" transfectants. The parental cells had a normal mor-
`phology that was indistinguishable from NIH 3T3 cells.
`However, amplified cells had the typical refractile. spindle~
`
`l23456789
`
`200—
`
`FIG. 1. Quantitation of p185““2 in four primary . unamplified cell
`lines and lines derived from them by amplification to resistance to 400
`nM methotrexate. Lane I. neo-DH FR control; lanes 2 and 3. HERZ-l
`parent and amplified lines;
`lanes 4 and 5, HERE-3 parent and
`amplified lines; lanes 6 and ‘l'. HER2-4 parent and amplified lines;
`lanes 8 and 9. HER2-33 parent and amplified lines. Positions of the
`size markers myosin and ,B-galactosidase are indicated in kDa.
`
`A
`
`2 of 5
`
`Celltrion, Inc., Exhibit 1047
`
`

`

`3 of 5
`
`Celltrion, Inc., Exhibit 1047
`
`

`

`T162
`
`Cell Biology: Hudzialt er al.
`
`Proc. Natl. Acad. Sci. USA 84 (1987)
`
`Table 2. Tumongenicity testing of cell lines
`
`Cell line
`
`Cells injected, no.
`
`Mice with tumors)(
`mice injected
`
`NIH 3T3
`
`NIH 3T3/CVN
`
`HERZ-Jm
`
`0.25 X 10°
`0.5 X 10"
`1.0 x 10"
`0.25 x 10‘
`0.5 x10”
`1.0 x 10“
`0.25 x 10“i
`0.5
`x 10"'
`1.0 X 10".
`
`0/6
`Off:
`Off:
`0/6
`0/6
`we
`5/5
`6/6
`5,6
`
`The cell line HER2-3 and two control lines, NIH 3T3 cells. and a
`NIH 3T3 line Containing the control plasmid. were iniected subcu-
`taneously at three different dosages into nude mice. The time after
`injection before tumors became visible was dose related: 22 days
`(average) for l X 10" cells, 28 days (average) for 0.5 x 10‘ cells. and
`34 days (average) for 0.25 x 10" cells.
`
`injection of three different cell numbers per animal. As shown
`in Table 2,
`the overexpressing cell
`line HER2-3m was
`strongly tumorigenic at all dosages tested, whereas the
`control cell
`lines, NIH 3T3 cells and NIH 3T3 cells
`transfected with CVN vector without the HER2 insert, were
`both negative under the same conditions. Necropsy of three
`mice with well-established tumors failed to identify any
`metastasis. Cell lines reestablished from three excised tu-
`mors still expressed the G418“ phenotype, were resistant to
`methotrexate, and expressed high levels of pISSHERZ (not
`shown).
`
`DISCUSSION
`
`Amplification of the HER2 gene has been reported in a few
`primary human tumors (2, 3) and tumor-derived cell lines (13,
`26). Recently, Slamon et al. (14) found that the HERZ gene
`is amplified in 30% of primary breast tumors, a common
`human malignancy that affects about 7% of all American
`females. Notably, only 3/189 tumors surveyed showed any
`gross rearrangement of the HERZ gene. To assess the role of
`the HERZ gene in the neoplastic process, we characterized
`the transforming potential of the HERZ gene in an in vitro cell
`culture system.
`Expression of the full-length cDNA in NIH 3T3 cells did
`not
`lead to transformation as determined by a standard
`focus-forming transfection assay. However, HER2 overex-
`pression caused by gene amplification transformed these
`cells. Colonies that survived methotrexate selection were
`morphologically transformed and exhibited loss of contact
`inhibition. Such cells also grew in soft agar and would grow
`in 1% calf serum. Furthermore, cells transformed by HER2
`were tumorigenic in athymic mice.
`Selection of cells transfected with the HERZ cDNA for
`growth in low serum provided independent evidence that
`high-level expression of an unaltered HERZ gene product
`caused cellular transformation. Since DNA introduced into
`mammalian cells by transfection is more labile than genomic
`DNA (23, 2?), we reasoned that selection for a property
`demonstrated by pharmacologically amplified HERZ lines-—
`namely, growth in low serum—might lead to amplification or
`other changes resulting in overexpression of pISSHE”.
`Clones derived from an unamplifted HER2 line by this selection
`procedure appeared morphologically transformed and exhibit-
`ed elevated levels of plSSl'E’u.
`Taken together, the characteristic morphological changes,
`results of in vitro and in viva transformation and tumorigenic-
`ity assays, and the elevated levels of p185"“2 in cells
`selected for a transformed phenotype imply that high-level
`expression of HERZ results in transformation of NIH 3T3
`
`cells. Another member of the tyrosine kinase gene family was
`recently found to be amplified 4- to 8-fold in spontaneously
`arising foci of NIH 3T3 cells (28). Amplification of the met
`gene appears to be a frequent event and, similar to HER2
`amplification in mammary carcinomas, is rarely accompa-
`nied by gross rearrangements. These findings and the dis-
`covery of amplified EGF receptor genes in primary human
`tumors (10—12) suggest that ov'erexpression of other growth
`factor receptor genes will also lead to transformation and
`tumorigenesis. Whether susCeptibility to spontaneous ampli-
`fication is caused by characteristics of the genetic loci for
`HERZ. met and EGF receptor, or cell-specific selective
`advantages caused by receptor overexpression remains to be
`investigated.
`It
`is not yet clear whether transformation of cells
`overexpres sing a growth factor receptor gene is dependent on
`paracrine or autocrine stimulation by the appropriate ligand.
`For HERZ, this question cannot yet be addressed since its
`ligand has not been identified.
`In the case of the EGF
`receptor, however, we were able to demonstrate that primary
`human tumors and tumor-derived cell lines frequently ex-
`press mRNAs for both the receptor and transforming growth
`factor type a ligand (29). The finding that mere overexpres-
`sion of an intact growth factor receptor may subvert normal
`cellular growth control mechanisms and lead to tumor-igenic
`growth provides new potential for diagnostic approaches and
`therapeutic strategies for treatment of human malignancies.
`We express appreciation for the superb technical assistance of
`Thomas Doll in constructing the HERZ expression plasmids and Bill
`Lagrirnas for help with the mouse tumorigenicity assays. We are
`grateful to Jeanne Arch for her patience and skill
`in typing this
`manuscript and to Dr. Suzanne Pfeffer for her valuable editorial
`comments.
`
`1. Coussens, L., Yang-Feng,T. L... Liau, Y.-C., Chen, EL, Gray.
`A.. McGrath. J.. Seeburg, P. H., Liben'nann, T. A., Schles-
`singer, J.. Francke, U.. Levinson, A. & Ullrich, A. (1985)
`Science 230, 1132—1139.
`2. King. C. R., Kraus, M. H. & Aaronson. S. A. (1985) Science
`229, 974—976.
`3. Semba, K., Kamata, N., Toyoshima. K. & Yamamoto, T.
`(1985) Proc. Natl. Acad. Sci. USA 82, 6497—6501.
`4. Schechter. A. L... Hung, M.—C., Vaidyanathan, L., Weinberg.
`R.. Yang-Feng, T. l... Francke, U.. Ullrich, A. & Coussens.
`L. (1985) Science 229, 976-978.
`5. Akiyama. T.. Sudo, C.. Ogawara, H., Toyoshima, K. &
`Yamamoto. T. (1986) Science 232, 1644—1646.
`6. Stern, D. F.. Hefi'eman, P. A. & Weinberg, R. A. (1986) Mai.
`Cell. Biol. 6, 1729—1740.
`7. Bargmann, C. I.. Hung, M.-C. 3: Weinberg, R. A. (1986) Cell
`45, 649—657.
`8. Sherr. C. .l., Rettenmier. C. W., Sacca, R., Roussel, M. F..
`Look. A. T. & Stanley, E. R. (1985) Cell 4|, 665—676.
`9. Ullrich. A.. Schlessinger. J. & Waterlield, M. D. (1984) Na-
`ture (London) 301, 521—527.
`10. Hunts,
`.I., Ueda, M., Ottawa, 8., Abe, 0.. Pastan,
`Shimizu, N. (1985) Conn 76, 663—666.
`11. Cowley. (3., Smith, J. A.. Gusterson. B.. Hendler. F. &
`Deanne. B. (1984) in Cancer Cells, eds. Levine, A., Vande
`Woude, G., TOPP. W. & Watson. J. D. (Cold Spring Harbor
`Laboratory. Cold Spring Harbor, NY), pp. 5—10.
`12. Libennann. T. A.. Nusbaum. H. R., Razon, N., Kris, R..
`Lax, I., Soreq, H., Whittle. N., Waterfield. M. D.. Ullrich. A.
`& SchleSsinger, J. (1985) Nature (London) 313. 144-147.
`13. Fukushige, S.-l., Matsubara, K.-l., Yoshida, M., Sasaki, M.,
`Suzuki. T.. Semba. K.. Toyoshima. K. a Yamamoto, T.
`(1986) Mai. Cell. Biol. 6, 955—958.
`14. Slamon. D. J., Clark, G. M., Wong, S. (3., Levin, W..|.,
`Ullrich, A. & McGuire. W. L. (198?) Science 235, 1?7—182.
`15. Blair, D. G., Oskarssot'l, M., Wood, T. (3.. McClements,
`W. L., Fischinger, P. J. & Vande Woude, G. G. (1981) Sci-
`ence 2I2, 941—943.
`16. Pulciani. S., Santos, E.. Long. L. K., Sorrentino. V. &
`Barbacid, M. (1985) Mol. Cell. Biol. 5, 2836-2841.
`
`I. d:
`
`4 of 5
`
`Celltrion, Inc., Exhibit 1047
`
`

`

`Cell Biology: Hudziak er al.
`
`Proc. Natl. Acad. Sci. USA 84 {I987}
`
`7163
`
`17.
`
`18.
`
`19.
`
`20.
`
`21.
`
`22.
`23.
`
`Riedel. H.. Schlessinger. J. & Ullrich. A. (1987) Science 236,
`197—200.
`Alt. F. W.. Kellems. R. D.. Bertino. .I. R. & Schimke. R. T.
`(1978} J. Biol. Chem. 253, 1357—1370.
`Colbcre-Garapin. F.. Horodniceanu. F.. Kourilsky. P. &
`Garapin. A.-C. (1981) J. Mol. Biol. 150. 1—14.
`Simonsen. C. C. & Levinson. A. D. (1983} Mai. Ce". Biol. 3.
`2250—2253.
`Graham. F. L. & van dcr Eb. A.J.
`456—461.
`Frosi. E. & Williams. 1. (EMS) Virology 91, 39—50.
`Kaufman, RJ. & Sharp. P. A.
`(1982) J'. Mol. Biol. 159,
`601—621.
`
`(191F131 Virology 52.
`
`24.
`
`25.
`
`26.
`
`23.
`
`29.
`
`.I. (1973) The Molecular Biology of Tumor Viruses
`Tooze.
`(Cold Spfing Harbor Laboratory. Cold Spring Harbor. NY}.
`van der Hoorn. F. A. & Muller. V. (1985} Mai. Ceh'. Biol. 5,
`2204—2211.
`Kraus. M. H.. Popescu. N. C.. Amsbaugh. S. C.
`C. R. (198?) EMBO J. 6. 605-610.
`Jolly. D. J.. Willis. R. C. & Friedmann. T. (1986) Mai. Ce”.
`Biol. 6, 1141—1147.
`Cooper. C. S.. Tempest, P. R.. Beckman, M. P.. Heldin,
`C.—H.
`.I. & Brookes. P. (19860133480 J. 5, 2623—2623.
`Derynck. R., Goeddel. D. V.. Ullrich. A.. Gutterman. J. U..
`Williams. R. D.. Bringrnan. T. S. & Berger. W. H.
`(1987)
`Cancer Res. 47. 707412.
`
`8L King.
`
`5 of 5
`
`Celltrion, Inc., Exhibit 1047
`
`