The EMBO Journal vol.6 no.3 pp.605-610. 1987
`
`erbB-2
`of the EGF receptor-related
`proto-oncogene
`Overexpr�ion
`in human mammary tumor cell lines by different molecular
`mechanisms
`
`Matthias ff.Kraus, Nicholas C.Popescu1, Suzanne
`C.Amsbaugb1 and C.Richter King
`Lab_<>ratory of Cellular and Molecular Biology, and 1 Laboratory of Biology.
`Nauonal Cancer Institute. Building 37, Room IE24. Bethesda. MD 20892.
`USA
`Communicated by J .Schlessinger
`Amplifacation of the erbB/EGF receptor and a structurally
`related gene, designated erbB-2, have previously been detected
`in a variety of human tumors. In a series of human mam­
`mary tumor cell lines, analysis of transcripts of these genes
`revealed elevated levels of one or the other in more than 60%
`of tumors analyzed. Eight cell lines demonstrated erbB-2
`mRNA levels ranging from 4- to 128-fold above those of nor­
`rnaJ controls. erbB-2 exp�ion was evaluated in comparison
`to the expr�ion level of actin observed in these cell lines.
`There was no evidence of an aberrantly sized erbB-2
`transcript in any of these lines. lmmunoblot analysis indicated
`elevation in levels of the 185-kd product of the erbB-2 gene
`exp� by these cells. In four lines erbB-2 gene amplifica­
`tion in the absence of an apparent gene rearrangement was
`demonstrated. In a represe.ntative cell line of this type, SK­
`BR-3, the amplified erbB-2 gene copies were located in an
`which demonstrated 4- to 8-fold overexpr�ion of erbB-i
`aberrant chromosomal location. Four additional cell lines
`mRNA, did not exhibit gene amplification. In a representative
`cell line of this type ZR-75-1, an apparently normal chromo­
`somal location was found for the erbB-2 gene. Our rmdings
`indicate that overexp�ion of the erbB-2 gene in mammary
`tumor cell lines is frequent and associated with different
`genetic abnormalities.
`Key words: erbB-2/gene amplification/growth factor receptor/
`mammary neoplasia/overexpression
`
`Introduction
`Cellular genes encoding effector molecules of growth regulation
`have been linked to the neoplastic process based on their
`homology to retroviraJ oncogenes. The c-sis proto-oncogene en­
`codes a chain of platelet-derived growth factor (POOF) (Doo­
`little et al., 1983; Waterfield er al., 1983), the v-erbB oncogene
`has been shown to encode a truncated fonn of the epidennal
`growth factor (EGF) receptor (Downward et al., 1984), and the
`c-jms proto-oncogene product is related to the receptor for mono­
`nuclear phagocyte growth factor (CSF-lR) (Sherr et al., 1985).
`More recently, we and others (King et al. , l 985a; Schechter et
`al., 1985; Semba er al., 1985; Coussens er al., 1985) have iden­
`tified a second cellular analogue of v-erbB in the human genome.
`This gene, designated erbB-2, is related to but distinct from the
`gene encoding the erbBIEGF receptor (EGFR). The predicted
`amino acid sequence of the erbB-2 gene reveals the structural
`features of a growth factor receptor molecule with close similarity
`to the EGF receptor (Coussens et al., 1985; Yamamoto et al.,
`1986a), including a cysteine-rich extracellular domain, transmem-
`
`. A d.ominan� cellu�ar transforming gene activated by point
`
`brane region and a highly conserved tyrosine kinase domain.
`Gene alterations affecting EGFR and erbB-2 occur in tumor
`cells .
`mutation m cherrucally induced rat neuroblastomas (Shih et al. ,
`1981), neu, is likely to be the rat homologue of human erbB-2
`based on comparative nucleotide sequence analysis and chromo­
`somal localization (Schechter et al. , 1985; Coussens et al. , 1985;
`Yamamoto et al., 1986a; Bargmann et al., 1986a). In human
`glioblastoma, amplification and rearrangement of the EGFR gene
`result in extensive expression of abnormal as well as norrnal­
`
`s.ized �As (Libermann et al., 1985). In addition, amplifica­
`
`tion without rearrangements affecting EGFR mRNA size is fre­
`quently found in cells derived from squamous cell carcinomas
`(Yamamoto et al., 1986b) and in two distinct mammary car­
`cinoma cell lines (King et al., l 985b; Filmus et al., 1985).
`Gene amplification of erbB-2 has been identified in a primary
`mammary adenocarcinoma (King er al., 1985a), as well as in
`a salivary gland adenocarcinorna (Semba et al., 1985). These
`
`�ndings have s�ggested the possibility that erbB-2 overexpres­
`
`s1on may contribute to neoplastic growth (King et al. , l 985a;
`Semba et al., 1985).
`For this study, we investigated the expression of erbB-2 and
`EGFR in 16 human mammary tumor cell lines. Our results in­
`dicate frequent overexpression of these proto-oncogenes that are
`related to growth factor receptors. Furthennore, analysis of the
`erbB-2 gene locus in these cell lines demonstrates that enhanced
`erbB-2 expression can occur in the presence or the absence of
`gene amplification, suggesting that different molecular mechan­
`isms result in overexpression of nonnal size erbB-2 mRNA in
`mammary tumor cells.
`
`Results
`Isolation of erbB-2 complementary DNA
`To allow a comprehensive anaJysis of erbB-2 mRNA and gene
`structure we isolated cDNAs with a complexity of over 4.5 kb
`from the mRNA (Figure IA). An oligo (dT) primed normal
`human fibroblast cDNA library (Okayama and Berg, 1983) was
`screened with a 0.8 kbp Accl DNA fragment from a genomic
`clone of erbB-2 (King et al., 1985a). The largest plasmid ob­
`tained, pMAC137, carried a 2-kbp insert comprising 1.5 kbp of
`3' coding infonnation and 3' untranslated sequence. The remain­
`ing coding infonnation upstream was obtained from three phage
`clones, AMAC30, AMACIO' and >.MAC14-l, identified in a ran­
`domly primed MCF-7 cDNA library (Walter et al., 1985; Figure
`IA). Nucleotide sequence analysis and restriction mapping of
`the entire cDNA indicated that its structure was the same as an
`isolate from normal human placenta (Coussens et al., 1985).
`Overexpression of erbB-2 or EGFR proto-oncogenes in human
`mammary tumor cell lines
`To assess the role of erbB-2 in human mammary neoplasia we
`compared the mRNA of 16 mammary tumor cell lines to nor­
`mal human fibroblasts, M413, and a human mammary epitheliaJ
`cell line, HBLIOO. Increased expression of an apparently nor-
`
`605
`
`1 of 6
`
`BI Exhibit 1046
`
`

`

`M.H.Kraus et al.
`
`A
`
`NN
`5· II
`
`St
`
`I
`
`811 I
`
`BllE PP Sp8P88
`
`11 II I ll 11
`
`88mEKSm811P
`
`� I )) I I I
`
`St
`
`E
`
`E
`
`E
`
`E
`
`3·
`
`pMAC137
`
`>..MAC30
`
`AMAC10·
`
`>..MAC14-1
`
`0
`
`B
`
`•
`
`2
`
`c
`
`3
`
`4
`
`5
`
`kbp
`
`probes
`
`Fig. l. (A) Isolation and restriction mapping of erbB-2 cDNA. Clone pMACl37 was isolated from an oligo (dTJ primed normal human fibroblast cDNA
`library (Okayama and Berg, 1983). Clones >.MAC30, AMAC 10' and >.MACl4-I were subsequently obtained from a randomly primed MCF-7 cDNA library
`(Walter et al . • 1985). Restriction sites: B = BamHI, Bil = BstELI. E = EcoRl. N = Ncol, P = Pstl. Sm = Smal, Sp = Spill and St = Sfld. (B) cDNA
`probes used in hybridization analysis.
`
`A
`
`B
`
`"' .... 0
`<') ..
`;.; <') "'
`a: "' ..
`I .... CD
`� m .... .... I
`CD <')
`<')
`m m
`I .. _,
`I .. < < <
`I
`I
`0
`::E ::E ::E
`... "' a: I- a a a a: I- m
`"' <') �
`I
`I
`::E (/) N m ::E ::E ::E N m I
`
`I
`
`•••
`
`28s -
`
`18s -
`
`"' I a: "'m - I .. "' ::E (/)
`
`I
`
`;;; "'
`"' .. .... .... I
`m ::E
`I .. <
`a: .... a
`N CD ::E
`
`"' .. .... 0
`"' "'
`I ... CD �
`"'
`m
`CD
`I .. _,
`I "' "' 0
`::E ::E
`a a a: .... m
`I
`< <
`::E ::E N m I
`
`•
`
`erbB-2
`
`• •
`
`relative amount
`
`1
`
`1 32 4 64 32 16 4 4 4
`•••••• •
`•••••••••
`actln ••••• e •• • •
`•••••••• •
`• •
`fl
`• •
`
`relative amount 1 II• 112 v2 V2 v• v2 1 11<
`1
`ovaraxpresslon 1 128 a 128 64 64 e 4 16 1
`
`of arbB-2
`
`Fig. 2. Overexpression of erbB-2 in human mammary rumor cell lines. (A) Northern blot analysis. Total cellular RNA (10 µg} of mammary rumor cell lines.
`normal human fibroblasts M413 and HBLIOO was hybridized with a cDNA probe derived from the 5' end of the erbB-2 coding region (Figure IB. probe a).
`M413 and HBLIOO cells contain erbB-2 specific mRNA detectable after longer autoradiographic exposures. (8) Quantitation of erbB-2 mRNA levels. Serial
`2-fold dilutions of total RNA were applied to nitrocellulose. Replicate filters were hybridized with either a erbB-2 cDNA probe (Figure IB, probe b) or
`human P·actin which served as control for RNA amounts present on the nitrocellulose filter. Relative amounts detected with each probe are indicated in
`comparison to the hybridization signals observed in normal human fibroblasts M413.
`
`606
`
`2 of 6
`
`BI Exhibit 1046
`
`

`

`Table I. Overexpression of erbB-2 and EGFR proto-oncogcnes in human
`mammary neoplasia
`
`A
`
`erbB-2
`
`EGFR
`
`Overexpression Gene
`Overexpression Gene
`of mRNA1
`amplification of mRNA•
`amplification
`
`M413
`HBLIOO
`I
`MCF-7
`SK-BR-3
`128
`BT474
`128
`MDA-MB361 64
`MDA-MB453 64
`ZR-75-1
`8
`ZR-75-30
`4
`MDA-M8175
`8
`BT483
`8
`BT20
`MDA-M8468
`
`I
`4-8
`4-8
`2-4
`2
`
`<I
`I
`<I
`I
`
`<I
`I
`
`<I
`I
`<I
`<I
`<I
`16
`32
`
`"Overexpression above normal fibroblast and HBLIOO.
`
`Overexpression of the erbB-2 gene in mammary tumor cells
`
`(")
`a:
`...
`CD
`I
`(")
`•
`:.::
`(/)
`<
`+- +
`
`MWX10�3
`- - 200
`
`- 97
`
`B
`
`(") ...
`I
`I ,.._
`a: It)
`I
`CD ,.._ u..
`I (.)
`I
`� a:
`(/) N �
`
`MW x 10-3
`- 200
`
`- 97
`
`I
`4-8
`32
`
`Fig. 3. Elevated erbB-2 protein levels in mammary rumor cell lines. 40 µ.g
`total cellular protein was separated by electrophoresis and transferred to
`nitrocellulose filters. The erbB-2 protein was detected with an antipeptide
`antibody coupled to m1 protein A. The specificity of antibody detection was
`determined by pre-incubation of the antibody with excess amounts of peptide
`prior to irnmunodetection. ( +) preincubation with peptide, (-) no peptide.
`In panel 8, nonspecific bands at 110 kd are observed in longer exposures of
`peptide-blocked irnmunoblots (panel A).
`
`mat size 5-kb transcript was detected in 8 of 16 tumor cell lines,
`when total cellular RNA was subjected to Northern blot analysis.
`Figure 2A shows the results using a cDNA probe comprising
`the coding sequences of the amino-tenninal extracellular domain
`of erb8-2 (Figure 18, probe a). These results of overexpression
`of normal-sized mRNA were confirmed by hybridization of po­
`ly(A) + selected RNA using several erb8-2-specific probes com­
`prising coding information for the transmembrane and tyrosine
`kinase domains. An aberrantly sized erb8-2 mRNA was not
`detected in any of the cell lines analyzed.
`To quantitate more precisely the amount of erb8-2 transcript
`in eight manunary tumor cell lines which overexpress erb8-2,
`serial 2-fold dilutions of total cellular RNA were subjected to
`dot blot analysis using human {3 actin as a control for the amount
`of RNA applied to the nitrocellulose filters. As shown in Figure
`28, the highest levels of erb8-2 mRNA, which ranged from 64-
`to 128-fold over that of our controls, were observed in the cell
`lines MDA-MB453, SK-8R-3, MDA-MB361 and 8T474. More­
`over, erb8-2 mRNA levels were increased 4- to 8-fold in four
`cell lines including 8T483, MDA-MB175, ZR-75-30 and
`ZR-75-1 (Table I).
`To detennine ifthe overexpression of erb8-2 mRNA resulted
`in a steady state increase of its encoded gene product, we
`developed a specific imrnunoblot assay. Antisera were raised
`against a synthetic peptide whose sequence corresponded to a
`portion of the putative erb8-2 tyrosine kinase domain. As this
`region is partially conserved between the encoded proteins of
`the EGFR and erb8-2 genes, we tested its specificity using A431
`and SK-BR-3 cell lines which overexpress EGFR or erbB-2
`mRNA, respectively. As shown in Figure 3A, a specific band
`of -185 kd was detected in extracts of SK-8R-3 but not in A431
`cells. Thjs band was not detected when the antibody was pre­
`incubated with the synthetic peptide corresponding to its antigen.
`The human erb8-2 and rat neu products have been reported to
`be glycoproteins of 185 kd (Akiyama et al. , 1986; Stem et al. ,
`1986).
`To estimate the relative amounts of erb8-2 protein in different
`mammary tumor cell lines, immunoblot analysis was conducted
`using equivalent amounts of total cellular protein. As shown in
`Figure 38, an intense band of protein was detected in extracts
`of SK-8R-3 and a less intense but readily detectable band in ex­
`tracts of ZR-75-l. No erb8-2 protein was detected in extracts
`
`A
`
`8
`
`.. -
`� i � :
`i � � �
`
`23 -
`8. -
`16 -
`•I -
`
`2 •
`
`-- -
`-----
`
`�B 2
`
`•
`
`relat "" a�ount
`
`1 4 � 8 4
`
`1
`
`, 2 1
`
`• • ••••• ••
`• •• • •
`
`Fig. 4. Gene amplification of erlJB-2 in mammary rumor cell lines. (A)
`Southern blot analysis. For each lane 10 µ.g genomic DNA were restricted
`with Xbal and hybridized with a probe comprising the entire coding region
`of t>rbB-2. Hind.Ill restriction fragments of lambda DNA served as mol. wt
`standards. (8) DNA dot-blot analysis. Genomic DNA (10 µ.g) digested with
`EcoRJ was applied in serial 2-fold dilutions to nitrocellulose filters. Filters
`were hybridized either with erbB-2 (Figure 1 B, probe b) or mos, which
`served as a control for ONA amounts applied to replicate nitrocellulose
`filters. Gene copy numbers of erbB-2 relative to M413 indicate the minimal
`extent of gene amplification detected in ONA from mammary rumor cell
`lines.
`
`of MCF-7, a mammary tumor cell line, that did not display
`overexpression of erb8-2 mRNA. We interpret these results to
`indicate that substantially more erb8-2 protein is found in both
`SK-BR-3 and ZR-75-1 than in MCF-7 cells where the amount
`of protein escapes the sensitivity of the assay. Dilution ex­
`periments suggest that SK-8R-3 contains between 5- and 10-fold
`more erb8-2 protein than does ZR-75-1 (data not shown).
`We also analyzed total cellular RNAs of the same mammary
`tumor cell lines for evidence of EGFR receptor mRNA overex­
`pression. Increased amounts of an apparently normal size EGFR
`
`607
`
`3 of 6
`
`BI Exhibit 1046
`
`

`

`M.H.Kraus et al.
`
`.i. ..
`
`,
`
`r
`� ... ,
`....
`'
`., �
`4',r; ., ..
`.
`.. ,
`...
`�� •
`
`"
`
`-
`
`'
`/
`
`,.
`
`ill
`
`,,
`.. I
`\ . ._ /
`' �
`
`�
`
`A
`
`B
`
`\,;
`J
`
`,
`
`Fig. 5. Representative metaphase from the SK-BR-3 carcinoma cell line
`after in situ hybridization with an erbB-2 cDNA probe, autoradiography and
`G-banding. (A) A chromosome spread exhibiting silver grains on a
`distinctive, highly rearranged chromosome (arrow). (B) The same spread
`after trypsin/EDTA treatment to produce G-bands. Detailed G-banding
`analysis did not indicate that the site of hybridization contains a
`homogeneously stained region (HSR).
`
`mRNA were observed in BT20 and MDA-MB468. These two
`cell Lines have previously been shown to contain amplified EGFR
`genes. EGFR transcripts were elevated 16-fold in BT20 and
`32-fold in MDA-MB468 above the level seen in normal human
`fibroblasts as determined by RNA dot-blot analysis (Table I).
`Genetic abnormalities associaJed with elevated erbB-2 expression
`To investigate alterations of the erbB-2 gene associated with
`mRNA overexpression, we examined Xbal-restricted high mol.
`wt DNA by Southern blot analysis using a probe comprising the
`entire coding sequence of erbB-2 (Figure IB, probe c). The nor­
`mal restriction pattern of Xbal fragments was detected in all DNA
`samples analyzed indicating that gene rearrangements in prox­
`imity of the erbB-2 coding region did not occur in these cell lines.
`When compared with normal human fibroblast DNA (Figure 4A,
`lane 1) the erbB-2-specific XbaI restriction fragments appeared
`clearly amplified in the cell lines SK-BR-3, BT474 and MDA­
`MB361 (Figure 4A, lanes 2-4). Similar results were obtained
`with restriction enzymes &oRJ and Sad (data not shown).
`Quantitation of erbB-2 gene copy number was accomplished
`using DNA dot-blot analysis. These studies revealed a 4- to 8-fold
`
`608
`
`erbB-2 gene amplification in SK-BR-3 and BT474 relative to nor­
`mal human DNA and a 2- to 4-fold erbB-2 gene amplification
`in MDA-MB361. In addition, a 2-fold erbB-2 gene amplifica­
`tion was identified for the cell line MDA-MB453 by DNA dot­
`blot analysis. Thus, gene amplification was associated with
`overexpression in the four rumor cell lines with the highest levels
`of erbB-2 mRNA (Table I). In contrast, gene amplification could
`not be detected by Southern blot analysis or DNA dot-blot
`analysis in the four tumor cell lines with erbB-2 transcript in­
`creased to intermediate levels.
`To examine the nature of any chromosomal abnormalities
`associated with overexpression we used in situ hybridization to
`localize the erbB-2 gene in two cell lines which either do or do
`not contain amplified gene copies. The erbB-2 gene has been
`mapped in normal human cells on chromosome 17ql 1.2-22
`(Schechter et al., 1985; Coussens et al., 1985; Fukushige et al.,
`1986). Mammary tumor cell line SK-BR-3 contains 4- to 8-fold
`gene amplification and 128-fold overexpression of erbB-2
`mRNA. In SK-BR-3, G-banding showed no copies of normal
`chromsome 17. In situ hybridization of the erbB-2 gene to these
`cells revealed accumulations of grains on two large abnormal
`marker chromosomes derived from complex rearrangements in­
`volving at least three chromosomes. An average of three grains
`was observed at each labelled site (Figure 5). These results in­
`dicate that the amplification of the erbB-2 gene occurs in an ab­
`normal chromosomal location and is not associated with either
`a homogeneously stained region or double minute chromosomes,
`abnormalities diagnostic for gene amplification (Biedler and
`Spengler, 1976; Levan et al., 1977). Mammary tumor cell line
`ZR-75-1 showed no evidence of gene amplification and an 8-fold
`overexpression of mRNA (Table I). Chromosome 17 was pre­
`sent in one or two copies per cell. Analysis of 50 cells after in
`situ hybridization with a erbB-2 cDNA probe revealed the largest
`accumulation of grains on chromosome 17 with 85 % of these
`clustered on chromosome bands 17ql 1.2-21, the normal loca­
`tion of the erbB-2 gene. This indicates that overexpression of
`the erbB-2 gene can occur in the absence of detectable structural
`abnormalities of chromosome 17.
`Mammary tumor cell lines overexpressing erbB-2 do not con­
`tain readily detectable transforming genes
`In chemically induced rat neuroblastomas, a point mutation within
`the transmembranous domain activates the rat homologue of
`erbB-2, neu, to transforming activity readily detectable in the
`NIH/3T3 transfection assay (Bargmann et al., 1986b). Previous
`transfection analysis of 21 mammary tumors and tumor cell lines
`did not reveal activation of erbB-2 as a transforming gene in
`human mammary neoplasia (Kraus et al., 1984). To investigate
`whether an activating lesion similar to the rat neu gene was
`associated with the overexpression of erbB-2 in human mam­
`mary tumor cell lines, we transfected genomic DNA of these
`cell lines into mouse NIH/3T3 cells. Under conditions where high
`mol. wt DNA from the cell line T24 which is known to contain
`an activated H-ras oncogene induced 4-8 foci/plate, genomic
`DNA from eight mammary tumor cell lines which overexpress
`erbB-2 did not induce detectable morphological transformation
`(Table II).
`
`Disc�ion
`Analysis of 16 human mammary tumor cell lines for the EGF
`receptor and the related erbB-2 gene revealed frequently increased
`transcript levels of either member of this family of growth fac­
`tor receptor genes. In BT20 and MDA-MB468 overexpression
`
`4 of 6
`
`BI Exhibit 1046
`
`

`

`Table 0. DNA transfection of human mammary tumor cell lines
`
`Overexpr�ion of the erbB-2 gene in mammary tumor cells
`
`Source of genomic DNA
`
`FFU/plate•
`
`T24 prep I
`T24 prep 2
`T24 prep 3
`SK-BR-3
`BT474
`MDA-MB361
`MDA-MB453
`ZR-75-1
`ZR-75-30
`BT483
`MDA-MB175
`
`3214
`19/4
`16/4
`0112
`0112
`0/2
`0/4
`0/8
`014
`0/4
`0/8
`
`1FFU/plate = focus-fonning units/number of transfected plates. 30 µ.g high
`mol. wt DNA were coprecipitated with calcium phosphate and transfected
`onto NIH/3T3 fibroblasts as previously described (Wigler et al., 1977).
`Genomic T24 DNA served as positive control for each assay.
`
`of a multistep process which confers selective growth advantage
`to a mammary tumor cell.
`In all cell lines examined we detected no abnormalities of
`erbB-2 gene structure by Southern blot hybridization. Moreover,
`in ZR-75-1, in which overexpression occurs without gene
`amplification, the erbB-2 gene was located at its normal site on
`chromosome 17. These results indicate that mechanisms of
`deregulation are unlikely to involve the type of rearrangements
`responsible for activation of the myc gene in Burkitt's lymphoma
`(Taub et al., 1982; Dalla-Favera et al., 1983). Deregulation of
`erbB-2 gene expression may therefore involve subtle changes in
`cis-acting control sequences, changes
`involving transacting
`regulatory elements, or changes which enhance mRNA stability.
`Abnormalities of erbB-2 and EGFR genes are not restricted
`to mammary tumor cells in culture, as we and others have iden­
`tified erbB-2 or EGFR gene amplification in several samples of
`mammary tumor tissue (King et al. , l 985a; Yokota et al. , 1986,
`and unpubHshed observation). The precise action of growth fac­
`tor receptor gene overexpression in the neoplastic process of
`of normal size EGF receptor gene transcripts is associated with
`mammary tumor cells has yet to be established. However, since
`tumorigenicity of mammary tumor cells can be enhanced by
`gene amplification (King et al., 1985b; Filmus et al., 1985). In­
`creased transcript levels of erbB-2 were detected in the presence
`steroid hormones or polypeptide growth factors (Kasid et al.,
`and absence of gene amplification, indicating that erbB-2 over­
`1985; Dickson et al., 1986), it is tempting to hypothesize that
`expression in human mammary tumor cell lines can be caused
`the overexpression of polypeptide growth factor receptors in­
`creases tumorigenicity by stimulation of the same or similar
`by different molecular mechanisms. The absence of aberrantly
`sized erbB-2 mRNA in Northern blot analysis suggests that a
`growth signalling pathways. Forty percent of mammary tumor
`normal size mRNA is overexpressed in mammary tumor cell lines
`cell lines analyzed did not exhibit overexpression of erbB-2 or
`rather than a rearranged form. Moreover, the fact that genomic
`EGFR. Heterogeneity of steroid receptor expression is of con­
`siderable therapeutic and prognostic value in human mammary
`DNA from these cell lines lacked the ability to transform
`NIH/3T3 cells by transfection indicates that point mutations
`neoplasia (McClelland et al., 1986; DeSombre et al., 1986). Our
`results strongly suggest an additional level of heterogeneity defin­
`similar to those that can activate the rat neu gene did not occur.
`These observations provide evidence that a structurally normal
`ed by overexpression of the erbB-2 or EGFR gene.
`In the cell line SK-BR-3, where we identified an amplified and
`coding sequence of erbB-2 is overexpressed as mRNA in human
`mammary tumor cell lines. Protein analysis of representative
`overexpressed erbB-2 gene, a myc gene amplification with over­
`expression has been previously reported (Kozbor and Croce,
`samples with erbB-2 overexpression suggests that elevated erbB-2
`transcript levels are translated into erbB-2 proteins.
`1984). The myc gene has been shown to cooperate with a ras
`Several lines of evidence link the overexpression of proto­
`gene in the transformation of primary fibroblasts (Land et al. ,
`1983). It will be of interest to determine whether overexpres­
`oncogenes to the neoplastic process. The increased transcription
`of normal coding sequences of either the human c-sis/PDGF gene
`sion of erbB-2 and myc can complement each other in promoting
`or the H-ras gene using a viral long terminal repeat promoter
`the growth of mammary tumor cells. In addition, systematic
`analysis of mammary tumors will determine whether overexpres­
`induces transformation of NIH/3T3 cells in culture (Chang er
`sion of erbB-2 or the EGFR gene can be associated with a par­
`al., 1982; Gazit et al., 1984). Moreover, in human tumors myc
`or N-myc amplification correlates with increased malignancy
`ticular clinical manifestation of mammary carcinoma.
`(Kohl et al., 1983; Schwab et al., 1983; Nau et al., 1984). Our
`results link erbB-2 overexpression to the neoplastic growth of
`mammary tumor cells.
`Gene amplification of multidrug resistance genes is observed
`in cells selected for the ability to grow in media containing cer­
`tian metabolic inhibitors (Alt et al., 1978). There is evidence
`that overexpression of these genes can precede gene amplifica­
`tion in the development of multidrug resistance (Shen et al. ,
`1986). Our observation of elevated erbB-2 transcript levels in
`the presence and absence of gene amplification may reflect a
`similar pathway. Overexpression of erbB-2 may confer an in­
`itial selective growth advantage to the tumor cell and subsequent
`gene amplification cause a further step where the selective growth
`advantage is enhanced and stabilized. erbB-2 overexpression is
`consistently higher in those samples with gene amplification when
`compared with cell lines lacking gene amplification. Interestingly,
`the amount of overexpression per gene copy is approximately
`constant in the cell lines that overexpress erbB-2 (Table I). This
`suggests that deregulated erbB-2 expression due to different
`molecular mechanisms in mammary tumor cell lines may be part
`
`RNA blotting
`Total cellular RNAs were isolated by CsCl-gradient centrifugation (Chirgwin et
`al., 1979). For Northern blot analysis 10 µg tOOll RNA were denatured and elecmr
`phoresed in I% formaldehyde gels (Lehrach et al., 1977). Following mild alkali
`degradation (50 mM NaOH at 24°C for 30 min) RNA was transferred to
`nitrocellulose filters using I M ammonium acetate as a convectant. RNA dot­
`blot analysis was conducted by applying serial 2-fold dilutions of total RNA in
`I M ammonium acetate to replicate nitrocellulose filters in a dot-blot manifold.
`DNA blorring
`For Southern blot analysis DNA samples (10 µg) were restricted with Xbal and
`subjected to electrophoresis in 0.8% agarose gels. DNAs were transferred to nitro­
`cellulose filters by the method of Southern (Southern, 1975). For dot-blot anaJ
`ysis
`DNAs were cleaved with EcoRJ, denatured with 0.1 M NaOH and neutralized
`_
`with I M ammonium acetate. Serial 2-fold dilutions of each sample were ap­
`plied to replicate nitrocellulose filters using a dot-blot manifold.
`
`Materials and methods
`Tumor cell lines
`Human mammary tumor cell Jines were obtained from ATCC (American Type
`Culrure Collection) and culrured according to the supplier's reconunendation.
`
`609
`
`5 of 6
`
`BI Exhibit 1046
`
`

`

`Kohl,N.E., Kunda,N., Schreck,R.R., Bruns,G .. Lan,S.A., Gilbert,G. and
`Alt,F.W. (1983) Cell, 35, 349-367.
`Kozbor,O. and Croce,C.M. (1984) Cancer Res., 44, 438-441.
`Kraus,M.H., Yuasa, Y. and Aaronson,S.A. ( 1984) Proc. Natl. Acad. Sci. USA,
`81, 5384-5388.
`Land,H., Parada,L.F. and Weinberg,R.A. (1983) Nature. 304, 596-602.
`Lehrach,H., Oiamond,K., Wozney,J.M. and Boedtker,H. (1977) Biochemistry,
`16, 4743-4751.
`Levan,A., Levan,G. and Mitclman,F. (1977) Hereditas, 86, 15-29.
`Libennann,T.A., Nusbaum,H.R., Razon,N., Kris,R., Lax,J., Soreg,H., Whittle.
`N., Waterfield,M.O., Ullrich.A. and Schlessinger,J. (1985) Nature, 313,
`144-147.
`Lowry,O.H., Rosebrough.NJ .. Farr,A.L. and Randall,R.J. (1951) J. Biol.
`Chem., 193, 265-275.
`McClelland,R.A., Berger,U., Miller,L.S., Powles,T.J., Jensen,E.V. and
`Coombes,R.C. (1986) Cancer Res. (Suppl.), 46, 4241s-4243s.
`Nau,M.M., Camey,O.N., Battey,]., Johnson,B., Linle,C., Gazdar,A. and Min­
`na,J.O. (1984) Curr. Top. Microbiol. /mmunol., 113, 172-177.
`Okayama,H. and Berg,P. (1983) Mo/. Cell. Biol., 3, 280-289.
`Popescu,N.C., Amsbaugh,S.C., OiPaolo,J.A., Tron.ick,S.R., Aaronson,S.A. and
`Swan,0.C. (1985a) Somat. Cell Mo/. Genet., 11, 149-155.
`Popescu,N.C., Amsbaugh,S.C., Swan,O.C. and OiPaolo,J.A. (1985b) Cytogenet.
`Cell Genet., 39, 73-74.
`Schechter,A.L, Hung,M.C .. Vaidyanathan,L., Weinberg,R.A., Yang-Feng,T.L.,
`Francke,U., Ullrich,A. and Coussens,L. (1985) Science, 229, 976-978.
`Schwab,M., Alitalo,V., Klempnauer,K.H., Varmus,H.E., Bishop,J.M.,
`Gilbert,F., Brodeur,G., Goldstein,M. and Trent,J. (1983) Nature, 305,
`245-248.
`Semba,K., Kamata,N., Toyoshima,K. and Yamamoto,T. (1985) Proc. Nm/. Acad.
`Sci. USA, 82, 6497-6501.
`Shen,O.W., Fojo,A., Chin,J.E., Roninson,l.B., Richert,N., Pastan,I. and Got·
`tesman,M.M. (1986) Science, 232, 643-645.
`Sherr,C.J., Renenmier,C.W., Sacca,R., Roussel,M.F., Look,A.T. and
`Stanley,E.R. (1985) Cell, 41, 665-676.
`Shih,C., Padhy,L., Murray,M. and Weinberg,R.A. (1981) Nature, 290,
`261-264.
`Southem,E.M. (1975) J. Mo/. Biol., 98, 503-517.
`Stem,O.F., Heffeman,P.A. and Weinberg,R.A. (1986) Mo/. Cell. Biol .• 6,
`1729-1740.
`Taub,R., IGrsch,I., Morton,C., Lenoir,G., Swan,0., Tronick,S., Aaronson,S.
`and Leder,P. (1982) Proc. Natl. Acad. Sci. USA, 97, 7837-7841.
`Walter,P., Green,S., Greene,G., Krust,A., Bomcrt,J.M., Jeltsch,J.M., Staub.A.,
`Jensen,E., Scrace,G., Waterfield,M. and Chambon,P. (1985) Proc. Nm/. Acad.
`Sci. USA, 82, 7889-7893.
`Waterfield,M.O., Scracc,G.T .. Whinle,N., Stroobant,P., Jonson.A .. Wasteson,
`A., Westermark,B., Heldin,C.H., Huang,J.S. and Deuel,T.F. (1983) Nature,
`304, 35-39.
`Wiglcr,M., Silverstein,S., Lee.LS., Pellicer,A., Cheng,Y.C. and Axel,R. (1977)
`Cell, 11, 223-232.
`Yamamoto,T., lkawa,S., Akiyama,T., Semba,K., Nomura,N., Miyajima,N.,
`Saito,T. and Toyoshima,K. (1986a) Nature, 319, 230-234.
`Yamamoto,T., Kamata,N., Kawano,H., Shimizu,S., Kuroki,T., Toyoshima,K.,
`Rikimaru,K., Nomura,N., lshizaki,R., Pastan,J., Gamon,S. and Shimizu,N.
`(1986b) Cancer Res., 46, 414-416.
`Yokota,J .. Yamamoto,T., Toyoshima,K., Terada,M., Sugimura,T., Battifora,H.
`and Cline,M.J. (1986) Lancet, (i) (8484). 765-767.
`Received on Jm1uary 5, 1987
`Note added in proof
`After submission of this manuscript, a study by Slamon,O. et al., ((1987) Science,
`235, 177-182) indicated a correlation of erbB-2 gene amplification and reduced
`disease-free survival in breast cancer patients.
`
`M.H.Kra� et al.
`
`Hybridiwtion
`Hybridization was conducted at 42°C in 750 mM NaCl, 75 mM Na citrate and
`50% formamide. The filte.rs were washed under stringency conditions of 15 mM
`NaCl and 1.5 mM Na citrate at 50°C.
`lmmunoblotting
`One confluent 100 mm plate of each cell line was lysed in l % Triton, 0.1 % SOS,
`0.5% deoxycholate, 10 mM sodium phosphate pH 7.4, 0.1 M sodium chloride,
`I mM phenylmethyl sulfonyl fluoride, and centrifuged at 12 000 g for 10 min
`at 4 °C. The supernatant was tested for protein concentration using the method
`of Lowry (Lowry et al., 1951) and adjusted to I% SOS, 50 mM Tris-HCI pH
`6.8, 0.1 % (3 mercaptoethanol, 5% glycerol and 0.01 % bromophenol blue. After
`heat treatment (100°C for 5 min), 40 µg of each sample was subjected to poly­
`acrylamide gel (7%) electrophoresis and transferred to nitrocellulose. The nitro­
`cellulose was treated for I h in Blono solution (2 % nonfat dry milk, I% Triton,
`50 mM Tris-HCI pH 7.4, JO mM EOTA). The IgG fraction of an antipeptide
`antibody directed against amino acids 866-880 of the erbB-2 protein was con­
`ducted in Blono for 2 h. Blots were washed three times with Blotto solution for
`JO min each. Reaction with 2 x 1D5 c.p.m./ml of 1125 protein A was conducted
`for 20 min at room temperature. The blots were then washed three times for
`10 min with Blono solution and 20 times with water. Detection was accomplish­
`ed by autoradiography at -70°C.
`In situ hybridit.ation
`Exponentially growing cultures of ZR-75-1 and SK-BR-3 were exposed to col­
`cemid (0.05 µg/ml) and ethidium bromide (7 .5 µg/ml) for 4 and 2 h, respective­
`ly, before harvesting for chromosome preparation. In situ hybridization was carried
`out under stringent conditions as previously described (Harper and Saunders, 1981;
`Popescu et al., 1985a) using 3H-labelled erbB-2 cONA as probe. The hybridiz­
`ed slides were coated with nuclear track emulsion NTB-2, diluted I: I with H20
`and stored desiccated at 4°C for 14 days. G bands were produced as previously
`described (Popescu et al. , l 985b).
`
`Acknowledgements
`We thank H.Okayama, P.Chambon and R.Ralston for kindly providing cONA
`libraries, N.Ellmore for technical assistance, and S.A.Aaronson for continuous
`support and helpful discussions.
`
`References
`Akiyama,T., Sudo,C., Ogawara,H., Toyoshima,K. and Yamamoto,T. (1986)
`Science, 232, 1644-1646.
`Alt,F.W., Kellems,R.E., Bertino,J.R. and Schimke,R.T. (1978) J. Biol. Chem.,
`253, 1357-1370.
`Bargmann,C.J., Hung,M.C. and Weinberg,R.A. (1986a) Nature, 319, 226-230.
`Bargmann,C.J., Hung,M.C. and Weinberg,R.A. (1986b) Cell, 45, 649-657.
`Biedler,J.L. and Spengler,B.A. (1976) Science, 191, 185-187.
`Chang,E.H., Furth,M.E