`a pattern of hybridization (Fig. IA) that
`differed both from that observed with
`DNA of normal human placenta and
`from that observed with the A431 squa-
`mous-cell carcinoma line, which con-
`tains amplified EGF receptor genes (7).
`In A431 DNA, four Eco RI fragments
`were detected that had increased signal
`intensities compared to those of corre-
`sponding fragments in placenta DNA
`(Fig. IA). In contrast, MAC117 DNA
`contained a single 6-kilobase pair (kbp)
`fragment, which appeared to be ampli-
`fied compared to corresponding frag-
`ments observed in both A431 and placen-
`ta DNA's (Fig. IA). These findings were
`consistent with the possibility that the
`MAC1 17 tumor contained an amplified
`DNA sequence related to, but distinct
`from, the cellular erbB proto-oncogene.
`To clone the 6-kbp fragment, we di-
`gested DNA from MAC1 17 with Eco RI,
`ligated it into bacteriophage AgtWES,
`packaged it in vitro, and transferred it to
`Escherichia coli strain BNN45 by infec-
`tion. A library of 4 x 105 bacteriophages
`was screened by plaque hybridization
`with radioactive v-erbB DNA. Ten of 14
`hybridizing phages contained a 6-kbp
`Eco RI fragment. Figure 2 shows the
`physical map of one of these phages,
`XMAC1 17, and pMAC1 17, a pUC 12 sub-
`clone containing a 2-kbp Bam HI frag-
`ment of XMAC1 17 that hybridized with
`v-erbB probes. The region of pMAC1 17
`to which v-erbB hybridized most in-
`tensely was flanked by Acc I and Nco I
`sites. Human repetitive sequences were
`also localized (Fig. 2, region demarcated
`by arrows).
`By digestion of pMAC1 17 with Bgl I
`and Bam HI, it was possible to generate
`a single-copy probe homologous to v-
`erbB. This probe detected a 6-kb Eco RI
`fragment that was amplified in MAC1 17
`DNA and possibly increased in A431
`cellular DNA relative to normal DNA
`(Fig. IB). The sizes of the fragments
`corresponded to the amplified 6-kb Eco
`RI fragment detected in MAC1 17 DNA
`by means of v-erbB (Fig. IA). Hybrid-
`ization to Southern blots containing seri-
`al dilutions of MACi 17 genomic DNA
`indicated an approximate amplification
`of 5- to 10-fold when compared to human
`placenta DNA.
`The nucleotide sequence of the por-
`tion of pMAC1 17 located between the
`Nco I and Acc I sites contained two
`regions of nucleotide sequence homolo-
`gous to v-erbB separated by 122 nucleo-
`tides (Fig. 3). These regions shared 69
`percent nucleotide sequence identity
`with both the v-erbB and the human
`EGF receptor gene. The predicted amino
`SCIENCE, VOL. 229
`
`Amplification of a Novel v-erbB-Related Gene in a
`Human Mammary Carcinoma
`
`Abstract. The cellular gene encoding the receptor for epidermal growth factor
`(EGF) has considerable homology to the oncogene ofavian erythroblastosis virus. In
`a human mammary carcinoma, a DNA sequence was identified that is related to v-
`erbB but ampljifed in a manner that appeared to distinguish itfrom the gene for the
`EGF receptor. Molecular cloning of this DNA segment and nucleotide sequence
`analysis revealed the presence of two putative exons in a DNA segment whose
`predicted amino acid sequence was closely related to, but different from, the
`corresponding sequence of the erbB/EGF receptor. Moreover, this DNA segment
`identified a 5-kilobase transcript distinct from the transcripts of the EGF receptor
`gene. Thus, a new member of the tyrosine kinase proto-oncogene family has been
`identified on the basis of its amplification in a human mammary carcinoma.
`growth factor receptors, we used the v-
`erbB gene to probe for related genes that
`might be candidates for other receptor
`coding sequences. We selected moderate
`stringency hybridization conditions un-
`der which different oncogenes of the
`tyrosine family did not cross-hybridize.
`Thus, any gene detected might be ex-
`pected to have a closer relationship to v-
`erbB than to other members of the tyro-
`sine kinase family.
`DNA prepared from tissue of a human
`
`C. RICHTER KING
`MATTHIAS H. KRAus
`STUART A. AARoNsoN
`Laboratory of Cellular and
`Molecular Biology,
`National Cancer Institute,
`Bethesda, Maryland 20205
`
`A
`
`B
`
`C _C P
`
`v V
`co
`
`.C
`
`O
`
`Fig. 1. Detection of v-erbB- and pMAC 117-
`specific gene fragments in normal human pla-
`centa, A431 cells, or human mammary carci-
`noma MAC117. DNA (15 F.g) was cleaved
`with Eco RI, separated by electrophoresis in
`agarose gels, and transferred to nitrocellulose
`paper (18). Hybridization to the 32P-labeled
`probe (20) was conducted in a solution of 40
`percent formamide, 0.75M NaCl, 0.075M so-
`dium citrate, at 42°C (19). The v-erbB probe
`(A) was a mixture of the 0.5-kbp Bam HI-
`Bam HI fragment and 0.5-kbp Bam HI-Eco
`RI fragment of avian erythroblastosis proviral
`DNA. The pMAC1 17 probe (B) was a l-kbp
`Bgl I-Bam HI fragnent. After hybridization,
`the blots were washed first in 0.3M NaCl plus
`0.03M sodium citrate at room temperature,
`and then in 0.015M NaCl, 0.0015M sodium
`citrate, and 0.1 percent sodium dodecyl sul-
`fate at 42°C (A) or at 52°C (B). Hybridization
`was detected by autoradiography.
`
`The oncogenes of the acute transform-
`ing retroviruses have counterparts, des-
`ignated proto-oncogenes, that are con-
`served within the human genome (1).
`The human sis proto-oncogene encodes
`one major polypeptide chain of platelet-
`derived growth factor (PDGF) (2), and
`the erbB proto-oncogene appears to en-
`code the receptor for epidermal growth
`factor (EGF) (3). A number of other
`proto-oncogenes, like erbB, share nucle-
`otide sequence homology with the tyro-
`sine kinase-encoding src gene (4). The
`fact that cellular receptors for several
`growth factors or hormones, including
`the EGF receptor, possess this enzymat-
`ic activity suggests that other proto-on-
`cogenes may encode growth factor re-
`ceptors as well.
`Genetic alterations affecting proto-on-
`cogenes of the tyrosine kinase family can
`play a role in spontaneous tumor devel-
`opment. A specific translocation affect-
`ing the c-abl locus, for example, is asso-
`ciated with chronic myelogenous leuke-
`mia (5). Several recent studies have also
`documented amplification or rearrange-
`ment of the gene for the EGF receptor in
`certain human tumors (6) or tumor cell
`lines (7). We now report the detection
`and partial isolation of a gene that is a
`new member of the tyrosine kinase fam-
`ily and is amplified in a human mammary
`carcinoma. This gene is closely related
`to, but distinct from, the EGF receptor
`gene.
`The identification of additional mem-
`bers of some proto-oncogene families
`has emerged from findings of related
`sequences amplified sufficiently in a par-
`ticular tumor to allow detection (8). Be-
`cause of our interest in genes coding for
`974
`
`PETITIONER'S EXHIBITS
`
`Exhibit 1041 Page 1 of 3
`
`
`
`ail sequence of-these regions was 85
`percent homologous to two regions that
`are contiguous in the; EGF receptor se-
`quence (7). Furthermore, these two pu-
`tative coding regions of the MAC1 17
`sequence were each flanked by the AG
`and GT dinucleotides that border the
`exons of eukaryotic genes (9). These
`findings suggest that the sequence shown
`in Fig. 2 represents two exons, separated
`by an intron, of a gene related to the
`erbBIEGF receptor gene.
`The predicted amino acid sequence of
`the XMAC117 putative exons is homolo-
`gous to the corresponding sequences of
`several members of the tyrosine kinase
`family. The most striking homology was
`observed with the human EGF receptor
`or erbB (Fig. 3). In addition, we ob-
`served 42 percent to 52 percent homolo-
`gy with the predicted amino acid se-
`quences of other tyrosine kinase-encod-
`ing genes. At 25 percent of the positions
`there was identity among all the se-
`quences analyzed (Fig. 3). A tyrosine
`residue in the XMAC1 17 putative coding
`sequence, conserved among the tyrosine
`kinases analyzed, is the site of autophos-
`phorylation of the src protein (10).
`The availability of cloned probes of
`the MAC1 17 gene made it possible to
`investigate its expression in a variety of
`cell types. The MAC1 17 probe detected
`a single 5-kb transcript in A431 cells
`(Fig. 4). Under the stringent conditions
`of hybridization utilized, this probe did
`not detect any of the three RNA species
`recognized by EGF receptor comple-
`mentary DNA. Thus, MAC117 repre-
`sents a new functional gene within the
`tyrosine kinase family, closely related
`to, but distinct from, the gene encoding
`the EGF receptor.
`There is precedent for the identifica-
`tion of genes related to known onco-
`genes on the basis of their amplification
`in human tumors. For example, the high
`degree of amplification of N-myc in cer-
`tain maligancies made it detectable by
`means of the myc gene as a molecular
`probe (8). In the present study, a five- to
`tenfold amplification of a v-erbB-related
`gene in the MAC1 17 mammary carcino-
`ma made it possible to identify this se-
`quence against a complex pattern of
`EGF receptor gene fiagments. Analysis
`of DNA from -ten additional mammary
`carcinomas has not revealed amplifica-
`tion of the MAC1 17 gene. However,
`extensive studies will be required to de-
`termine the frequency of MAC1 17 gene
`amplification in different human malig-
`nancies.
`The MAC1 17 coding sequence, as de-
`termined by nucleotide and predicted
`amino acid sequence, was most closely
`6 SEPTEMBER 1985
`
`related to eBEF receptor among
`known members of the tyrosine kinase
`family. The two genes are distinct, on
`the basis of sequence diversity and tran-
`script size. Detailed structural analysis
`of the complete coding sequence should
`give insights into the possible functions
`
`of-this v-erb-
`td gene. Neverthe-
`less, because of its close relationship to
`the sequence of the EGF receptor, it is
`possible to speculate that the MAC117
`coding sequence may also be derived
`from a gene encoding a growth factor
`receptor. An oncogene in a chemically
`
`AMAC117-
`
`B
`|
`
`BXh
`H
`
`B
`
`B BXbR
`g~t
`
`pMAC 1 17
`
`B,'
`
`I
`
`-
`
`N BgANN
`I:
`
`v-erbB
`-related
`
`%"B
`
`-
`
`I-~-
`
`1 kb
`
`0.5 kb
`
`0
`
`10O0
`
`200
`
`300
`
`80
`40
`20
`10
`70
`60
`90
`GTCTACATGGGTGCTTCCCATTCCAGGGGATGAGCTACCTGGAGGATGTGCGGCTCGTACACAGGGACTTGGCCGCTCGGAACGTGCTGGTCAAGAGTCC
`GlyHetSerTyrL.uGluAspValArqLeuValHi sArqAspL.uAlaAlaArqAsnValLeuValLysS.rPr
`
`30
`
`50
`
`CAACCATGTCAAAAT TACAGACT TCGGt;CT GGCTCGGCTGCTGGACA TTGACGAGACAGAGt ACCATGCAGA TGGGGGCAAGGT TAGGTGAAGGACCAAG
`oAsnHi sValLysIleThrAspPheGlyL.uAlaArqLeuLeuAspIl1AspGluThrGluTyrHi sAlaAspGlyGlyLys
`
`GAGCAGAGGAGGCTGGGTGGAGTGGTGTCTAGCCCATGGGAGAACTCTGAGTGGCCACCTCCCCACAACACACAGTTGGAGGACTTCCTCTTCTGCCCTC
`
`CCAGGTGCCCATCAAGTrGATGGCGCTGLAGTCCATTCTCCGCCGGCGGTTCACCCACCAGAGTuATGTGTGGAGTTATGGTGTGTGATSGeLGLTGTTe
`Va 1Pro I 1eL ysT rpMetAl1aL euGl uSerlI1 L euArqArqArqPhaThr i sGlnSerA spVal1TrpSorTyrGl y
`
`GGAGGGGTGGOTGAGGAGCCATGG
`400
`Fig. 2. Restriction-site map of XMAC1 17 and plasmid pMACl 17. A, Acc I; B, Bam HI; Bg, Bgl
`I; N, Nco I; R, Eco RI; X, Xba I; Xh, Xho I. The sites were located by electrophoretic analysis
`of the products of single and double digestion. Regions homologous to v-erbB or human
`repetitive sequences (region flanked by arrows) were located by Southern blot hybridization
`(18) with the v-erbB probe or total human DNA made radioactive by nick translation (20).
`Hybridization conditions were as described in Fig. IA. The nucleotide sequence of pMAC1 17
`between the Acc I site and the Nco I sites and regions of encoded amino acid sequence
`homologous to the EGF receptor are shown. The AG or GT dinucleotides flanking the putative
`coding regions are underlined. To determine the sequence, Nco I, Hinf I, and Sau % I
`fiagments were labeled at the 3' termini by means of the large fragment of E. coli DNA
`polymerase, separated into single strands by gel electrophoresis, and chemically degraded (21).
`
`Homol -
`ogy (%
`
`pMAC117
`Human EGF Receptor 85
`85
`v-erbB
`52
`v-src
`v-iBT
`51
`v-fis
`50
`Huiian Insulin
`42
`Receptor
`
`*N
`
`~~~~****
`GMSYLEDVRLVHRDLAARNVLVKSPNHYKITDFGLARLLDI
`*NR TItlQ XEGAG
`QGA
`RJ4NYflf I*4 ENL V
`ZASKCRYA G
`DI-M
`A.NAKKfCKIAHDFT GMSIDT--
`
`I--
`
`1%
`
`C
`u,
`
`s
`
`w < 288
`
`DETEtHA-DG-GK--VPIK*WALESILRRRFTHQSDVWSYGV
`
`EDW
`
`ITYT-FI
`NDS
`
`I
`
`P LAYNIM
`
`pMAC117
`Human EGF Receptor
`v-erbB
`v-src
`v-iaT
`v-EIn
`Huiiiii Insulirn Receptor
`Fig. 3 (left). Comparison of the putative encoded amino acid sequence in pMACI 17 with known
`tyrosine kinase sequences. Black regions represent homologous amino acids. Differing amino
`acid residues are shown in one-letter code (22). Amino acid positions conserved in all sequences
`are denoted by *. The tyrosine homologous to that autophosphorylated by the v-src protein (10)
`is shown by an arrow. The v-abl sequence contains a tyrosine residue in this region displaced by
`two positions. The amino acid sequences of human EGF receptor (7), v-erbB (4), v-src (23), v-
`abl (24), v-fms (4), and human insulin receptor (25) were aligned by the computer program
`described (26). The homology observed with the predicted amino acid sequences of v-yes and v-
`Fig. 4 (right). Detection of distinct
`fes was 51 percent and 48 percent, respectively.
`messenger RNA species derived from the XMAC1 17 gene and the human EGF receptor gene.
`Polyadenylated messenger RNA of A431 cells was separated by denaturing gel electrophoresis
`in formaldehyde (23), transferred to nitrocellulose (18), and hybridized under stringent
`conditions (50 percent formamide, 0.075M NaCl, 0.75M sodium citrate, at 42°C) with 32p-
`labeled probe from pMAC1 17 (Bgl I-Bam HI fragment) or human EGF receptor complementary
`DNA (PE7: 2-kb Cla I inserted fiagment). Filters were washed under conditions of high
`stringency (0.015M NaCl plus 0.0015M sodium citrate at 55°C). Hybridization was detected by
`autoradiography with exposure times of 4 hours for the pMACI 17 probe and 1 hour for the
`human EGF receptor probe.
`
`975
`
`PETITIONER'S EXHIBITS
`
`Exhibit 1041 Page 2 of 3
`
`
`
`induced rat neuroblastoma has been de-
`tected by DNA transfection analysis
`(11). This oncogene, designated neu, ap-
`pears to encode a protein immunologi-
`cally related to the EGF receptor (12).
`Whether the MAC1 17 coding sequence
`and neu represent the same or different
`cellular genes awaits further character-
`ization.
`Overexpression of proto-oncogenes
`can cause cell transformation in culture
`and may function in the development of
`human tumors. Amplification of a nor-
`mal ras gene or its increased expression
`under the control of a retroviral long
`terminal repeat (LTR) induces transfor-
`mation of NIH 3T3 cells (13). Expression
`of the normal human sisIPDGF-2 coding
`sequence in NIH 3T3 cells, which do not
`normally express their endogenous sis
`proto-oncogene, also leads to transfor-
`mation (14). In Burkitt lymphoma, a
`involving
`translocation
`chromosomal
`myc places its normal coding sequence
`under the control of an immunoglobulin
`gene regulatory sequence (15). The re-
`sulting alteration in myc expression is
`likely to be causally related to tumor
`development (16). The observation of
`amplification of myc or N-myc in more
`malignant phenotypes of certain tumors
`has supported the idea that overexpres-
`sion of these genes can contribute to the
`progression of such tumors (8, 17). The
`erbB/EGF receptor gene is amplified or
`overexpressed in certain tumors or tu-
`mor cell lines (6). The five- to tenfold
`amplification of our v-erbB-related gene
`in a mammary carcinoma suggests that
`increased expression of this gene may
`have provided a selective advantage to
`this tumor. The isolation of a new mem-
`ber of the tyrosine kinase gene family
`amplified in a human mammary carcino-
`ma provides an opportunity to investi-
`gate the potential role of this gene in
`human malignancy.
`Note added in proof: Recently, Semba
`et al. (28) independently detected a v-
`erbB-related gene that was amplified in a
`human salivary gland adenocarcinoma.
`Nucleotide sequence analysis of this
`gene indicates its identity to the MAC1 17
`gene in the regions compared.
`
`References and Notes
`1. J. M. Bishop, Annu. Rev. Biochem. 52, 301
`(1983); P. H. Duesberg, Nature (London) 304,
`219 (1983).
`2. R. F. Doolittle et al., Science 221, 275 (1983);
`M. D. Waterfield et al., Nature (London) 304, 35
`(1983).
`3. J. Downward et al., Nature (London) 307, 521
`(1984).
`4. A. Hampe, I. Laprevotle, F. Galibert, L. A.
`Fedele, C. J. Sherr, Cell 30, 775 (1982); N.
`Kitamura, A. Kitamura, K. Toyoshima, Y. Hir-
`ayama, M. Yoshida, Nature (London) 297, 205
`(1982); M. Shibuya and H. Hanafusa, Cel 30,
`787 (1982); T. Yamamoto et al., ibid. 35, 71
`(1983).
`
`976
`
`5. A. de Klein et al., Nature (London) 3W, 765
`(1982); S. J. Collins and M. T. Groudine, Proc.
`Natl. Acad. Sci. U.S.A. 80, 4813 (1983).
`6. T. A. Libermann et al., Nature (London) 313,
`144 (1985).
`7. A. Ullrich et al., ibid. 309, 418 (1984); Y. Xu et
`al., ibid., p. 809; C. R. Lin et al., Science 224,
`843 (1984).
`8. M. Schwab Nature (London) 305, 245 (1983); N.
`E. Kohl et al., Cell 35, 349 (1983).
`9. R. Breathnach and P. Chambon, Annu. Rev.
`Biochem. 50, 349 (1981).
`10. J. E. Smart et al., Proc. Natl. Acad. Sci. U.S.A.
`78, 6013 (1981).
`11. Alan L. Schechter et al., Nature (London) 312,
`513 (1984).
`12. J. A. Drebin, D. F. Stern, V. C. Link, R. A.
`Weinberg, M. I. Greene, ibid., p. 545.
`13. E. H. Chang, M. E. Furth, E. M. Scolnick, D.
`R. Lowy, ibid. 297, 479 (1982).
`14. A. Gazit et al., Cell 39, 89 (1984); M. F. Clarke
`et al., Nature (London) 308, 464 (1984).
`15. R. Taub et al., Proc. Natl. Acad. Sci. U.S.A.
`79, 7837 (1982).
`16. K. Nishikura et al., Science 224, 399 (1984).
`17. M. Schwab et al., Proc. Natl. Acad. Sci. U.S.A.
`81, 4940 (1984).
`18. E. M. Southern, J. Mol. Biol. 98, 503 (1975).
`19. G. M. Wahl, M. Stern, G. R. Stark, Proc. Natl.
`Acad. Sci. U.S.A. 76, 3683 (1979).
`
`20. P. W. J. Rigby, M. Dieckmann, C. Rhodes, P.
`Berg, J. Mol. Biol. 113, 237 (1977).
`21. A. M. Maxam and W. Gilbert, Proc. Natl.
`Acad. Sci. U.S.A. 74, 560 (1977).
`22. The following abbreviations were used for ami-
`no acids: A, alanine; C, cysteine, D, aspartic
`acid; E, glutamic acid; F. phenylalanine; G,
`glycine; H, histidine; I, isoleucine; K, lysine; L,
`leucine; M, methionine; N, asparagine; P, pro-
`line; Q, glutamine; R, arginine; S, serine; T,
`threonine; V, valine; W, tryptophan; Y, tyro-
`sine.
`23. H. D. Lehrach, D. Diamond, J. M. Wozney, H.
`Boedtker, Biochemistry 16, 4743 (1977).
`24. A. P. Czernilofsky et al., Nature (London) 287,
`193 (1980).
`25. E. P. Reddy, M. J. Smith, A. Srinivasan, Proc.
`Natl. Acad. Sci. U.S.A. 80, 3623 (1983).
`26. A. Ullrich et al., Nature (London) 313, 756
`(1985).
`27. D. J. Lipman and W. R. Pearson, Science 227,
`1435 (1985).
`28. K. Semba, N. Kamata, K. Toyshima, T. Yama-
`moto, Proc. Natl. Acad. Sci. U.S.A., in press.
`29. We thank I. Pastan and G. Merlino for providing
`the human EGF complementary DNA clone
`PE7 and P. di Fiore for the polyadenylated A431
`RNA.
`22 April 1985; accepted 3 June 1985
`
`The neu Gene: An erbB-Homologous Gene Distinct from and
`Unlinked to the Gene Encoding the EGF Receptor
`Abstract. The neu oncogene, identified in ethylnitrosourea-induced rat neuroglio-
`blastomas, had strong homology with the erbB gene that encodes the epidermal
`growthfactor receptor. This homology was limited to the region oferbB encoding the
`tyrosine kinase domain. It was concluded that the neu gene is a distinct novel gene,
`as it is not coamplified with sequences encoding the EGF receptor in the genome of
`the A431 tumor line and it maps to human chromosome 17.
`
`ALAN L. SCHECHTER
`MIEN-CHIE HUNG
`LALITHA VAIDYANATHAN
`ROBERT A. WEINBERG
`Whitehead Institute for Biomedical
`Research and Department of Biology,
`Massachusetts Institute of Technology,
`Cambridge 02142
`TERESA L. YANG-FENG
`UTA FRANCKE
`Department of Human Genetics,
`Yale University School of Medicine,
`New Haven, Connecticut 06510
`AXEL ULLRICH
`LISA COUSSENS
`Department of Molecular Biology,
`Genentech, Inc.,
`San Francisco, California 94080
`
`Rat neuroglioblastomas induced by
`exposure in utero to ethylnitrosourea
`frequently carry an oncogene detectable
`upon transfection into NIH 3T3 mouse
`cells (1). This oncogene (which we have
`termed neu) was found to be related to c-
`erbB (2), a gene that encodes the recep-
`tor for epidermal growth factor (EGF-r)
`(3). The neu oncogene induces the syn-
`thesis of a tumor antigen, p185, which is
`serologically related to the EGF-r (2).
`Southern blot analysis of rat DNA
`after Eco RI digestion revealed at least
`two erbB-homologous segments, one of
`
`which contained the neu oncogene in a
`biologically active form (2). It remained
`unclear whether the same or other DNA
`segments encode the EGF-r. Other anal-
`ysis uncovered differences between the
`products of the two genes. While poly-
`clonal sera to the EGF-r recognized
`p185, monoclonal antibodies to p185 did
`not react with the EGF-r. Moreover,
`there was an apparent molecular weight
`difference of 15,000 daltons between the
`two proteins (2).
`These data raised several possibilities
`regarding the relationship between the
`neu and c-erbB genes. The neu oncogene
`might be a mutated allele of the normal c-
`erbB gene, or it might be derived from a
`normal gene, the sequences of which
`overlap with those of c-erbB. Alterna-
`tively, the neu oncogene might have aris-
`en from a gene that is totally separate
`and distinct from erbB.
`We used three subclones of human c-
`erbB complementary DNA (cDNA) (4)
`and a 0.7-kilobase (kb) subclone of the v-
`erbB oncogene that had been transduced
`by the genome of avian erythroblastosis
`virus (5) for these studies. All of the neu
`oncogene lies within a 34-kb Eco RI
`segment that is present in the genomes of
`normal and tumor rat cells as well as in
`mouse NIH 3T3 cells that have acquired
`a neu oncogene via transfection (2). We
`SCIENCE, VOL. 229
`
`PETITIONER'S EXHIBITS
`
`Exhibit 1041 Page 3 of 3