MOLECULAR AND CELLULAR BIOLOGY, Mar. 1989, p. 1165-1172
`0270-7306/89/03 1165-08$02.00/0
`Copyright © 1989, American Society for Microbiology
`
`Vol. 9, No. 3
`
`p185¥£*? Monoclonal Antibody Has Antiproliferative Effects In
`Vitro and Sensitizes Human Breast TumorCells to
`Tumor Necrosis Factor
`
`ROBERTM. HUDZIAK,? GAIL D. LEWIS,2 MARCY WINGET,? BRIAN M. FENDLY,? H. MICHAEL SHEPARD,”
`AND AXEL ULLRICH?+*
`
`Departments of Developmental Biology,’ Pharmacological Sciences,” and Medicinal and Analytical Chemistry,°
`Genentech, Inc., 460 Point San Bruno Boulevard, South San Francisco, California 94080
`
`Received 3 October 1988/Accepted 8 December 1988
`
`The HER2/c-erbB-2 gene encodes the epidermal growth factor receptorlike human homolog of the rat neu
`oncogene. Amplification of this gene in primary breast carcinomas has been shown to correlate with poor
`clinical prognosis for certain cancer patients. We show here that a monoclonal antibody directed against the
`extracellular domain of p185‘7“*? specifically inhibits the growth of breast tumor-derived cell lines overex-
`pressing the HER2/c-erbB-2 gene product and prevents HER2/c-erbB-2-transformed NIH 3T3 cells from
`forming colonies in soft agar. Furthermore, resistance to the cytotoxic effect of tumor necrosis factor alpha,
`which has been shown to be a consequence of HER2/c-erbB-2 overexpression, is significantly reduced in the
`presence of this antibody.
`
`the human homolog of the rat proto-
`HER2!/c-erbB-2,
`oncogene neu (4, 34), encodes a 1,255-amino-acid glycopro-
`tein with extensive homology to the human epidermal
`growth factor (EGF) receptor (4, 21, 33, 34, 42). The
`HER2/c-erbB-2 gene product, p185‘7"*?, hasall of the struc-
`tural features and many of the functional properties of
`subclass I growth factor receptors (reviewed in references 43
`and 44),
`including cell surface location and an intrinsic
`tyrosine kinase activity. However, the ligand for this puta-
`tive growth factor receptor has not yet been identified.
`Amplification of the HER2/c-erbB-2 gene has been found
`in humansalivary gland and gastric tumor-derived cell lines
`(13, 34), as well as in mammary gland carcinomas(21, 22, 40,
`42). Slamonet al. (35) surveyed 189 primary breast adeno-
`carcinomas and determined that the HER2/c-erbB-2 gene
`was amplified in about 30% of the cases. Most importantly,
`HER2/c-erbB-2 amplification was correlated with a negative
`prognosis and high probability of relapse. Similar although
`less frequent amplification of the HER2/c-erbB-2 gene has
`been reported for gastric and colon adenocarcinomas(45,
`46). Experiments with NIH 3T3 cells also suggest a direct
`role for the overexpressed, structurally unaltered HER2/
`c-erbB-2 gene product p185"*? in neoplastic transforma-
`tion. High levels of HER2/c-erbB-2 gene expression attained
`by coamplification of the introduced gene with dihydrofolate
`reductase by methotrexate selection (18) or by using a strong
`promoter(6) was shownto transform NIH 3T3fibroblasts.
`Only cells with high levels of p1857"*? are transformed, i.e.,
`have an altered morphology, are anchorage independent,
`and will form tumors in athymic mice.
`Overexpression of p185”"”? may, furthermore, contribute
`to malignant tumor development by allowing tumorcells to
`evade one componentof the antitumor defenses of the body,
`the activated macrophage (17). Macrophages play an impor-
`tant role in immunesurveillance against neoplastic growth in
`vivo (1, 2, 38), and Urban et al. (39) have shown that tumor
`
`* Corresponding author.
`t Present address: Max-Planck-Institut fiir Biochemie, 8033 Mar-
`tinsried, Federal Republic of Germany.
`
`cells made resistant to macrophages display enhanced tu-
`morigenicity. Tumornecrosis factor alpha (TNF-a) has been
`shown to play a role in activated macrophage-mediated
`tumorcell killing in vitro (3, 11, 23, 29, 39). NIH 3T3 cells
`transformed by a transfected and amplified HER2/c-erbB-2
`cDNAshowincreasedresistance to the cytotoxic effects of
`activated macrophages or TNF-a in direct correlation with
`increased levels of p1857“"? expression. Furthermore,
`breast tumorcell lines with high levels of p185’7£*? exhibit
`resistance to TNF-a. Resistance to host antitumor defenses
`could facilitate the escape of cells from a primary tumor to
`establish metastases at distantsites.
`To further investigate the consequences ofalteration in
`HER2/c-erbB-2 gene expression in mammary gland neopla-
`sia and to facilitate investigation of the normalbiological role
`of the HER2/c-erbB-2 gene product, we have prepared
`monoclonal antibodies against the extracellular domain of
`p185£8?, One monoclonal antibody (4D5) was character-
`ized in more detail and was shown to inhibit
`in vitro
`proliferation of human breast tumor cells overexpressing
`p185£*? and, furthermore,
`to increase the sensitivity of
`these cells to the cytotoxic effects of TNF-a.
`
`MATERIALS AND METHODS
`Cells and cell culture. Human tumorcell lines were ob-
`tained from the American Type Culture Collection. The
`mouse fibroblast
`line NIH 3T3/HER2-349), expressing an
`amplified HER2/c-erbB-2 cDNA undersimian virus 40 early
`promotercontrol, and the vector-transfected controlcell line
`NIH 3T3/CVN have been described previously (18).
`Cells were cultured in a 1:1 mixture of Dulbecco modified
`Eagle medium and Ham nutrient mixture F-12 supplemented
`with 2 mM glutamine, 100 u of penicillin per ml, 100 pg of
`streptomycin per ml, and 10% serum. Human tumorcell
`lines were cultured with fetal bovine serum (GIBCO Labo-
`ratories, Grand Island, N.Y.); NIH 3T3 derivatives were
`cultured with calf serum (Hyclone Laboratories, Inc., Lo-
`gan, Utah.).
`Immunization. Female BALB/c mice were immunized
`with NIH 31T3/HER2-3,49) cells expressing high levels of
`
`1165
`
`(cid:43)(cid:82)(cid:86)(cid:83)(cid:76)(cid:85)(cid:68)(cid:3)(cid:89)(cid:17)(cid:3)(cid:42)(cid:72)(cid:81)(cid:72)(cid:81)(cid:87)(cid:72)(cid:70)(cid:75)(cid:3)
`Hospira v. Genentech
`(cid:44)(cid:51)(cid:53)(cid:21)(cid:19)(cid:20)(cid:26)(cid:16)(cid:19)(cid:19)(cid:27)(cid:19)(cid:24)(cid:3)
`IPR2017-00805
`(cid:42)(cid:72)(cid:81)(cid:72)(cid:81)(cid:87)(cid:72)(cid:70)(cid:75)(cid:3)(cid:40)(cid:91)(cid:75)(cid:76)(cid:69)(cid:76)(cid:87)(cid:3)(cid:21)(cid:19)(cid:24)(cid:24)
`Genentech Exhibit 2055
`
`

`

`1166
`
`HUDZIAK ET AL.
`
`Mo. CELL. BIOL.
`
`p185#*®?_ The cells were washed once with phosphate-
`buffered saline (PBS) and detached from the plate with PBS
`containing 25 mM EDTA. After low-speed centrifugation,
`the cells were suspended in cold PBS (2 x 10’ cells per ml).
`Each mouse wasinjected intraperitoneally with 0.5 mlof this
`cell suspension on weeks0, 2, 5, and 7.
`On weeks 9 and 13, 100 pl of a Triton X-100 membrane
`preparation of p1857#*?, partially purified by wheat germ
`agglutinin chromatography (700 pg of protein per ml) (25),
`was administered intraperitoneally. Three days before fu-
`sion, 100 yl of the enriched p185“"*? protein was adminis-
`tered intravenously.
`Fusion and screening. Mice with high antibodytiters as
`determined by immunoprecipitation of p185‘7"*? were sac-
`rificed, and their splenocytes were fused as described previ-
`ously (26). Spleen cells were mixed at a 4:1 ratio with the
`fusion partner, mouse myelomacell line X63-Ag8.653 (20),
`in the presence of 50% polyethylene glycol 4000. Fusedcells
`were plated at a density of 2 x 10° cells per well in 96-well
`microdilution plates. The hypoxanthine-azaserine (12) selec-
`tion for hybridomas was begun 24 h later. Beginning at day
`10 postfusion,
`supernatants from hybridoma-containing
`wells were tested for the presence of antibodies specific for
`p185#£®? by an enzyme-linked immunosorbent assay with
`the wheat germ agglutinin chromatography-purified p185"*?
`preparation (28). Enzyme-linked immunosorbent assay-pos-
`itive supernatants were confirmed by immunoprecipitation
`and cloned twice by limiting dilution.
`Large quantities of specific monoclonal antibodies were
`produced by preparation of ascites fluid; antibodies were
`then purified on protein A-Sepharose columns (Fermentech,
`Inc., Edinburgh, Scotland) and stored sterile in PBS at 4°C.
`Immunoprecipitations and antibodies. Cells were har-
`vested by trypsinization, counted in a Coulter counter
`(Coulter Electronics, Inc., Hialeah, Fla.), and plated 24 h
`before being harvested for analysis of p185"*? expression.
`Cells were lysed at 4°C with 0.8 ml of HNEGlysis buffer (18)
`per 100-mmplate. After 10 min, 1.6 ml of lysis dilution buffer
`(HNEG buffer with 1% bovine serum albumin and 0.1%
`Triton X-100) was addedto each plate, and the extracts were
`clarified by centrifugation at 12,000 x g for 5 min.
`Antibodies were added to the cell extracts and allowed to
`bind at 4°C for 2 to 4h. Immune complexes werecollected
`by adsorption to protein A-Sepharose beads for 20 min and
`washedthree times with 1 ml of HNEG buffer-0.1% Triton
`X-100. Autophosphorylation reactions were carried out for
`20 min at 4°C in 50 wl of HNEG washbuffer containing 5 mM
`MnCl, and 3 Ci of [y-72P]ATP (5,000 Ci/mmol, Amersham
`Corp., Arlington Heights,
`Ill.). The autophosphorylation
`reaction conditions have been described previously (18).
`Proteins were separated on sodium dodecyl sulfate (SDS)}-
`7.5% polyacrylamide gels and analyzed by autoradiography.
`The polyclonal antibody, G-H2CT17, recognizing the car-
`boxy-terminal 17 amino acids of p1857"®?, has been de-
`scribed previously (18). The anti-EGF receptor monoclonal
`antibody 108 (16) was provided by Joseph Schlessinger,
`Rorer Biotechnology, Inc.
`Fluorescence-activated cell sorting. SK-BR-3 human breast
`tumorcells overexpressing the HER2/c-erbB-2 gene(17, 22)
`or A431 human squamous carcinomacells overexpressing
`the EGFreceptor gene (14) were grown in T175 flasks. They
`were detached from the flasks by treatment with 25 mM
`EDTA-0.15 M NaCl, collected by low-speed centrifugation,
`and suspendedat 1 x 10° cells per ml in PBS—1% fetal bovine
`serum. One milliliter of each cell line was incubated with 10
`wg of either anti-HER2/c-erbB-2 monoclonal antibody (4D5)
`
`or a control antibody (40.1.H1) recognizing the hepatitis B
`surface antigen. The cells were washed twice and suspended
`on ice for 30 min in 1 ml of PBS-1% fetal bovine serum
`containing 10 pg of goat anti-mouse immunoglobulin G
`F(ab’), fragments conjugated with fluorescein isothiocyanate
`dye (Boehringer Mannheim Biochemicals,
`Indianapolis,
`Ind.). Unbound fluorescein dye was removed by two further
`washes. The cells were suspended at 2 x 10° per ml
`in
`PBS-1% fetal bovine serum and analyzed with an EPICS 753
`(Coulter) fluorescence-activated cell sorter. Fluorescein was
`excited by 300 mW of 488-nm argon laser light, and the
`emitted light was collected with a 525-nm band-passfilter
`with a 10-nm band width.
`Down-regulation assay. SK-BR-3 cells were plated at 1.5 x
`10° cells per 35-mm culture dish in normal medium. After a
`6-h period to allow attachment, the medium wasreplaced by
`1.5 ml of methionine-free labeling medium containing 150
`«Ci of [>°S]methionine per ml] and 2% dialyzed fetal bovine
`serum. The cells were metabolically labeled for 14 h and then
`chased with medium containing 2% dialyzed serum and
`unlabeled methionine. Either a control monoclonal antibody
`(40.1.H1) or anti-p185£*? (4D5) was added to a final con-
`centration of 2.5 pg/ml. At 0, 5, and 11 h, extracts were
`prepared with 0.3 ml of lysis solution and 0.6 ml of dilution
`buffer. The p185“£*? was immunoprecipitated with 2.5 pl of
`polyclonal antibody G-H2CT17. The washed immune com-
`plexes were dissolved in sample buffer, electrophoresed on a
`SDS-7.5% polyacrylamide gel, and analyzed by autoradiog-
`raphy. Each time point determination was performed in
`duplicate. Autoradiograph bandintensities were quantitated
`by using a scanner (Ambis Systems).
`Cell proliferation assays. The anti-p185"*? monoclonal
`antibodies were characterized by using the breast tumorcell
`line SK-BR-3. Cells were detached by using 0.25% (vol/vol)
`trypsin and suspended in complete medium at a density of 4
`
`CVN
`
`HER1
`
`HER2
`
`Go aS G6 Ofo 16. Bo
`vt
`wv
`—
`a
`wv
`—
`a
`+
`=
`1
`2 a4 5
`6
`[too
`9
`
` @ —
`#
`—-170
`
`FIG. 1. Specificity of monoclonal antibody 4DS5. Threecell lines,
`NIH 3T3/CVN, NIH 3T3/HERI-EGF receptor, and NIH 3T3/
`HER2-3499, were plated out at 2.0 x 10° in 100-mm culture dishes.
`At 24h, Triton X-100 lysates were prepared and divided into three
`portions. Either an irrelevant monoclonal antibody (6 pg of anti-
`hepatitis B virus surface antigen, 40.1.H1;
`lanes 1, 4, and 7),
`anti-p185““" monoclonal antibody 4D5 (6 yg; lanes 2, 5, and 8), or
`anti-EGF receptor monoclonal antibody 108 (6 jg; lanes 3, 6, and 9)
`was added and allowed to bind at 4°C for 4 h. The immune
`complexes were collected with 30 pl of protein A-Sepharose. Rabbit
`anti-mouse immunoglobulin (7 yg) was added to each 4D5 immuno-
`precipitation to improve the binding of this monoclonal antibody to
`the protein A-coated beads. Proteins were labeled by autophosphor-
`ylation and separated on an SDS-—7.5% polyacrylamide gel. The gel
`wasexposedto film at —70°C for 4 h with an intensifying screen. The
`arrows show the positions of proteins of M, 185,000 and 170,000.
`
`

`

`VoL. 9, 1989
`
`—G-H2CT17
`
`wG-H2CT17
`
`rm4D5
`
`185+
`170—
`
`FIG. 2. Binding of monoclonal antibody 4DS to unglycosylated
`receptor. NIH 3T3/HER2-349) cells were plated into two 100-mm
`plates at 2 x 10° cells per plate. After 14 h, the antibiotic tunicamy-
`cin was added to one plate at 3 pg/ml. After a further 5.5 h of
`incubation, Triton X-100 lysates were then prepared from each
`plate. Immunoprecipitations, the autophosphorylation reaction, and
`SDS-polyacrylamide gel electrophoresis were performed as de-
`scribed in the legend to Fig. 1. Lanes: 1, tunicamycin-treated cell
`lysate (one-third of a plate) immunoprecipitated with 2.5 pl of a
`polyclonal antibody directed against the C terminus of p185**?; 2,
`tunicamycin-treated cell lysate (one-third of a plate) immunoprecip-
`itated with 6 pg of 4DS; 3, untreated control lysate (one-third of a
`plate) immunoprecipitated with the polyclonal antibody. The arrows
`showthe locations of proteins of M, 185,000 and 170,000.
`
`x 10° cells per ml. Aliquots of 100 pl (4 x 10* cells) were
`plated into 96-well microdilution plates, the cells were al-
`lowed to adhere, and 100 pl of media alone or media
`containing monoclonal antibody (final concentration, 5 pg/
`ml) was then added. After 72 h, plates were washed twice
`with PBS (pH 7.5), stained with crystal violet (0.5% in
`methanol), and analyzed for relative cell proliferation as
`described previously (36).
`For assays in which monoclonal antibodies were com-
`bined with recombinant human TNF-a (5.0 x 107 U/mg;
`Genentech, Inc.), cells were plated and allowed to adhere as
`described above. Following cell adherence, contro] medium
`alone or medium containing monoclonal antibodies was
`added to a final concentration of 5 pg/ml. Cultures were
`incubated for another 4 h, and then increasing concentra-
`tions of TNF-a were added to a final volume of 200 ul.
`Following 72 h of incubation, the relative cell number was
`determined by crystal violet staining. Some samples were
`analyzed by crystal violet staining following cell adherence
`for determination ofthe initial cell number.
`
`RESULTS
`
`Specificity of monoclonal antibody 4D5. Monoclonal anti-
`bodies directed against the extracellular domain of p1857#*?
`were prepared by immunizing mice with NIH 3T3 cells
`transfected with a HER2/c-erbB-2 cDNA (HER2-3499) (17,
`18) and overexpressing the corresponding gene product,
`p185£®?, One antibody exhibited severalinteresting biolog-
`ical properties and was chosen for further characterization.
`Antibody 4D5 specifically immunoprecipitated a single *?P-
`
`MONOCLONAL ANTIBODIES AND p185#&*?
`
`1167
`
`Relative
`
`NumberofCells
`
`0.1
`
`1
`
`10
`
`102
`
`103
`
`104
`
`Relative FluorescenceIntensity
`FIG. 3. Fluorescence-activated cell sorter histograms of human
`tumor cells binding anti-p185 monoclonal antibody 4D5. ——,
`Binding by the control antibody, 40.1.H1, directed against
`the
`hepatitis B surface antigen; ------ , binding by the anti-HER2/c-
`erbB-2 antibody, 4D5. The antibodies were first allowed to react
`with the cell surface. After a wash step, bound antibody was labeled
`by addition of fluorescein-conjugated F(ab’), fragment of goat
`anti-mouse immunoglobulin G. (A) Binding of the antibodies to the
`humanbreast tumor line SK-BR-3, which contains an amplification
`of the HER2/c-erbB-2 gene and expresseshigh levels of the HER2/
`c-erbB-2 gene product p185"*?. (B) Binding of the same antibodies
`to the human squamous epithelial cell
`line A431. This cell
`line
`expresses low levels of mRNA for HER2/c-erbB-2 andhigh levels (2
`x 10° receptors per cell) of the EGF receptor.
`
`labeled protein of M, 185,000 from NIH 3T3cells expressing
`p185/7£*? (Fig. 1, lane 8). This antibody did not cross-react
`with the human EGF receptor (HERI; Fig. 1, lane 5), even
`when overexpressed in a mouse NIH 3T3 background (Fig.
`1, lane 6). Furthermore, it did not immunoprecipitate any
`proteins from NIH 3T3 cells transfected with a control
`plasmid (pCVN) which expresses the neomycin resistance
`and dihydrofolate reductase genes only (Fig. 1, lane 2).
`To determine the nature of the epitope recognized by 4D5,
`NIH 3T3/HER2-3499 cells were treated with tunicamycin,
`which prevents addition of N-linked oligosaccharides to
`proteins (15, 41). Cells treated with this antibiotic for 5.5 h
`contained two proteins which were immunoprecipitated by a
`polyclonal antibody against the carboxy-terminal peptide of
`p185‘7*? (Fig. 2, lane 1). The polypeptide of 170,000 M,
`represents unglycosylated p1857"®”, The upper band ofca.
`185,000 M, comigrated with glycosylated p185£*? from
`untreated cells (Fig. 2, lane 3). Monoclonal antibody 4D5
`efficiently immunoprecipitated only the glycosylated form of
`p1857“*? (Fig. 2, lane 2). This experiment suggests either
`that the epitope recognized by 4D5 consists partly of carbo-
`hydrate, or, alternatively,
`that the antibody recognizes a
`conformation of the protein achieved only when it is glyco-
`sylated.
`
`

`

`1168
`
`HUDZIAK ET AL.
`
`MoL. CELL. BIOL.
`
`TABLE1. Inhibition of SK-BR-3 proliferation by anti-p185""*?
`monoclonal antibodies*
`
`&a ®Q
`
`a
`
`no
`
`O
`Oo
`
`Relative cell
`Monoclonal
`proliferation?
`antibody
`TOD ic cc cece cc eens e cence ene en ene ecececececegeeeseseseeneneaeaees 79.3 42.2
`24 ccc cece cee c ence ec eeececeeeeseneseneaeaeeeaeeeenenenenenes 79.5 + 4.4
`TDB oc cc cece cece cence eens ee eenene eens ne en ease eeeeneneneneere enone 83.8 + 5.9
`ADS ooo cece cece n ence ee cece ces et ete ence ea eeeeneeeeessee ee eee 44.2+ 4.4
`BEB oo ccc ccc cece sec ececnaeaeneneeeeecesessssceueceeseeseees 66.2 + 2.4
`GED ooo cece cece cea ececececneeeeeeeeeeegeeeeeeseseeeeeee 98.9 + 3.6
`
`TES occ cece ccc cece eae ceee eee eenenececedeaeseseseseeneenenegs 62.1 + 1.4
`3H4.....
`66.5 + 3.9
`
`2H] oo. cece ee ee ee rete tenees
`92.9 + 4.8
`“0.0
`1.0
`2.0
`3.0
`
`
`40.1.H1... debececeeseeeeeeeeeeeeneeeeensnetatenaee teeeees 105.8 + 3.8
`
`Antibody (ug/ml)
`AES cece ccc ec eee e cece cesses ee eneeneneneeeeseseee ene 94.7 + 2.8
`FIG. 5. Growth of SK-BR-3 cells in different concentrations of
`monoclonal antibody 4D5. The human breast tumor line SK-BR-3
`wasplated into 35-mm culture dishes at 20,000 cells per dish. Either
`0.1, 0.5, 1.0, or 3.0 wg of a control monoclonal antibody (40.1.H1,
`anti-hepatitis B surface antigen) or monoclonal 4D5 antibody per ml
`was added at the time of plating. After 8 days of growth, the plates
`were trypsinized and the cells were counted in a Coulter counter.
`Each concentration of antibody was plated and counted in duplicate,
`and the cell numbers were averaged.
`
` 2g
`
`w S
`
`o
`
`* SK-BR-3 breast tumor cells were plated as described in Materials and
`Methods. Following adherence, medium containing 5 pg of either anti-
`p185£*? or control monoclonal antibodies (40.1.H1 and 4F4) per ml were
`added.
`® Relative cell proliferation was determined by crystal violet staining of the
`monolayers after 72 h. Values are expressed as a percentage of results with
`untreated control cultures (100%).
`
`The binding of monoclonal antibody 4DS5 to human tumor
`cell
`lines was investigated by fluorescence-activated cell
`sorting (Fig. 3). This antibody was bound to the surface of
`cells expressing p185/7"**, Figure 3A shows the 160-fold
`increase in cellular fluorescence observed when 4D5 was
`added to SK-BR-3 breast adenocarcinomacells relative to a
`control monoclonal antibody. This cell
`line contains an
`amplified HER2/c-erbB-2 gene and expresses high levels of
`p1857£*? (17, 22). In contrast, the squamous carcinomacell
`line A431, which expresses about 2 x 10° EGFreceptors per
`cell (14) but only low levels of p1857“*? (4), exhibited only
`a twofold increase in fluorescence with 4DS5 (Fig. 3B) when
`compared with a control monoclonal antibody.
`The binding of 4DS correlated with the levels of p185/7"*?
`expressed by these twocell lines. SK-BR-3 cells, expressing
`high levels of p1857"*?, showed an 80-fold increase in
`relative fluorescence intensity compared with A431 cells.
`This experiment demonstrates that 4D5 specifically recog-
`nizes the extracellular domain of p1857"*?.
`
`100
`
`10
`
`Plate
`104Cellsper
`
`16
`
`0
`
`4
`
`8
`12
`Days of Culture
`FIG. 4. Growth curve of SK-BR-3 cells treated with anti-HER2/
`c-erbB-2 monoclonal antibody 4D5. Cells were plated into 35-mm
`culture dishes at 20,000 cells per plate in medium containing 2.5 pg
`
`of either control antibody (40.1.H1, anti-hepatitis B surface antigen)
`
`
`
`(Q) or anti-p1854"? antibody 4D5 (@) per ml. On the indicated
`days, cells were trypsinized and counted in a Coulter counter. The
`determination for each time point and each antibody was done in
`duplicate, and the counts were averaged. The arrow indicates the
`day the cells were refed with medium without antibodies.
`
`Effects on cell proliferation. We used the human mammary
`gland adenocarcinoma cell
`line, SK-BR-3,
`to determine
`whether monoclonal antibodies directed against the extra-
`cellular domain of p185’7“*? had any effect on the prolifera-
`tion of cell lines overexpressing this receptorlike protein.
`SK-BR-3 cells were coincubated with several HER2/c-erbB-
`2-specific monoclonal antibodies or with either of two dif-
`ferent control monoclonal antibodies (40.1.H1, directed
`against the hepatitis B surface antigen; 4F4, directed against
`recombinant human gammainterferon). Most anti-HER2/
`c-erbB-2 monoclonal antibodies which recognize the extra-
`cellular domain inhibited the growth of SK-BR-3 cells (Table
`
` Relative
`NumberofCells
`
`157.
`
`361
`
`175
`
`3713
`HER2
`
`231 MCF-7 SK-
`BR-3
`Cell Line
`FIG. 6. Screening of breast tumorcell lines for growth inhibition
`by monoclonal antibody 4D5. Each cell line was plated in 35-mm
`culture dishes at 20,000 cells per dish. Either a control monoclonal
`antibody (9F6, anti-human immunodeficiency virus gp120) or the
`anti-p185“*? monoclonal antibody 4D5 was added on day 0 to 2.5
`pg/ml. Becausethe different cell lines grow at different rates, the cell
`lines NIH 3T3/HER2-3499 and SK-BR-3 were counted after 6 days,
`cell lines MDA-MB-157, MDA-MB-231, and MCF-7 were counted
`after 9 days, and cell lines MDA-MB-175VII and MDA-MB-361
`were counted after 14 days. The difference in growth between cells
`treated with 4D5 and 40.1.H1 is expressed as the ratio of cell
`numbers with 4D5 versus a control monoclonalantibody, 9F6. Each
`cell line was assayed in duplicate for each antibody, and the counts
`were averaged.
`
`

`

`VOL. 9, 1989
`
`Ohr
`12-223"
`
`Shr
`4
`
`11hr
`6-77
`8
`
`5
`
`1169
`
`29
`
` MONOCLONAL ANTIBODIES AND p185s¥28?
`
`FIG. 8. Effect of antibody binding on p185£*? turnover. SK-
`BR-3 cells were labeled for 14 h with [°°S]methionine. The label was
`then chased with cold methionine and either an irrelevant monoclo-
`nal antibody (40.1.H1, anti-hepatitis B surface antigen) or 4DS was
`added to 2.5 g/ml. The cells on the plates were lysed at 0, 5, and 11
`h, and *S-labeled p1857“*? was quantitated by immunoprecipita-
`tion with the C-terminal specific polyclonal antibody. The 5- and
`11-h time point determinations were performed in duplicate for each
`of the two antibodies. Proteins were separated by SDS-polyacryl-
`amide gel electrophoresis. The fluor-treated gel was exposedto film
`for 4 h at room temperature. The arrow indicates the position of a
`protein of M, 185,000. Band intensities were quantitated by using an
`Ambis Systems scanner. Lanes; 1, 0 h; lanes 2 and 3, 40.1.H1 (5 h);
`lanes 4 and 5, 4D5 (5 h); lanes 6 and 7, 40.1.H1 (11 h); lanes 8 and
`9, 4D5 (11 h).
`
`effects were achieved by using a concentration of between
`0.5 and 1 pg/ml.
`The effect of 4DS on the proliferation of six additional
`breast tumorcell lines, as well as mouse NIH 3T3fibroblasts
`transformed by p185/7“"? overexpression (NIH 3T3/HER2-
`3499), Was tested in monolayer growth assays. Cells were
`plated at low density in medium containing 2.5 pg of either a
`control antibody or 4DS per ml. When the cultures ap-
`proached confluency, cells were removed with trypsin and
`counted. 4DS did not have any significant effect on the
`growth of the MCF-7, MDA-MB-157, MDA-MB-231, or
`NIH 3T3/HER2-3,49) cell
`lines (Fig. 6). It did, however,
`significantly affect the growth of the cell lines MDA-MB-361
`(58% of control) and MDA-MB-175-VII (52% of control),
`which express high levels of p185/7#*? (17).
`Interestingly, monoclonal antibody 4DS5 had no effect on
`the monolayer growth of the NIH 3T3/HER2-345 cell line.
`However,
`it completely prevented colony formation by
`these cells in soft agar (Fig. 7), a property which had been
`induced by HER2/c-erbB-2 amplification (18). In the pres-
`ence of 200 ng of a control monoclonal antibody (antitissue
`factor, TC-C8) per ml, 116 (average of two plates) soft-agar
`colonies were counted, while the samecells plated simulta-
`neously into soft agar containing 200 ng of 4DS per ml did not
`yield any colonies.
`Monoclonal antibody 4D5 down-regulates p185#*?, To
`determine whether the antiproliferative effect of 4D5 was
`due to enhanced degradation of p185"*?, we measuredits
`rate of turnover in the presence or absence of antibody.
`p1857“®? was metabolically labeled by culturing SK-BR-3
`cells for 14 h in the presence of [>°S]methionine. Cells were
`then chased for various times, and either a control antibody
`or 4D5 was addedat the beginning of the chase period. At 0,
`5, and 11 h, cells were lysed and p185”"*? levels were
`assayed by immunoprecipitation and SDS-polyacrylamide
`gel electrophoresis. p1857“*? is degraded morerapidly after
`exposure of SK-BR-3 cells to 4DS5 (Fig. 8). Densitometric
`evaluation of the data showed that the p185’7“*?half-life of
`
`FIG. 7. Inhibition of anchorage-independent growth of NIH 3T3/
`HER2-34o9 cells by 4D5. Cells (20,000 per 60-mm plate) were plated
`in 0.2% soft agar over a 0.4% agar base. After 3 weeks, the plates
`were photographed at x100 magnification by using a Nikon micro-
`scope with phase-contrast optics. (a) HER2-3499 cells plated in agar
`containing 200 ng of a control antibody (TF-C8) per ml. (b) The same
`cells plated in agar containing 200 ng of 4D5 per ml.
`
`1). Maximum inhibition was obtained with monoclonal anti-
`body 4DS, which inhibited cellular proliferation by 56%. The
`control antibodies had nosignificant effect on cell growth.
`Figure 4 compares the growth of SK-BR-3 cells in the
`presence of either a control antibody, 40.1.H1, or the
`anti-p1857"*? antibody. Proliferation of the cells was inhib-
`ited when antibody 4D5 was present. The generation time
`increased from 3.2 to 12.2 days. To determine whether 4D5
`treatment was cytostatic or cytotoxic, antibody was re-
`moved by medium change11 daysafter treatment. Thecells
`resumed growth at a nearly normalrate, suggesting that the
`antibody affected cell growth rather than cell viability. The
`dose-response curve (Fig. 5) showed that a concentration of
`200 ng/ml
`inhibited growth by 50%, whereas maximum
`
`

`

`1170
`
`HUDZIAK ET AL.
`
`MOL. CELL. BIOL.
`
`A. SK-BR-3
`
`B. SK-BR-3
`
`C. MDA-MB-175-VIl
`
`1.5
`
`1.0
`
`0.5
`
`—O Oe
`
`1.0
`
`0.5
`
`
`
`RelativeCellProliferation
`
`
`
`
`0.0
`tumor cells to the cytotoxic effects of TNF-a. Cells were plated in 96-well
`FIG. 9. Monoclonal antibody 4D5 sensitizes breast
`microdilution plates (4 x 10* cells per well for SK-BR-3, MDA-MB-175-VII, and MDA-MB-231; 10* cells per well for HBL-100 and T24) and
`allowed to adhere for 2 h. Anti-HER2/c-erbB-2 monoclonal antibody 4D5 (5 wg/ml) or anti-hepatitis B surface antigen monoclonal antibody
`40.1.H1 (5 pg/ml) was then added for a 4-h incubation prior to the addition of TNF-« to a final concentration of 10* units/ml. After 72 h, the
`monolayers were washed twice with PBS and stained with crystal violet dye for determination of relative cell proliferation. In addition, some
`cell monolayers were stained with crystal violet following adherence in order to determine the initial cell density for comparison with cell
`densities measured after 72 h. The symbols denoteinitial cell density (I), untreated (control) cells (@), cells treated with TNF-a (i), 4DS
`
`(ZB), TNF-a plus 4D5 (0), 40.1.1 (38:9; or TNF-a plus 40.1.1 (440%).
`
`7 h decreased to 5 h in the presence of antibody (data not
`shown).
`Monoclonal antibody 4D5 enhances TNF-a cytotoxicity.
`The addition of certain growth factors to tumorcells has
`been shown to increase their resistance to the cytotoxic
`effects of TNF-a (37). A prediction based on these findings
`would be that expression of oncogenes that mimic or replace
`growth factor receptor function mayalso increase the resis-
`tance ofcells to this cytokine. Recently, it was shownthat
`overexpression of the putative growth factor
`receptor
`p1857£*? in NIH 3T3 cells caused an increase in the resis-
`tance of these cells to TNF-a (17). Furthermore, breast
`tumorcell lines with high levels of p185’7“*? also exhibited
`TNF-e resistance.
`To further investigate the mechanism by which the 4D5
`antibodyinhibited cell growth, we investigated the response
`of three breast tumorcell lines to TNF-a in the presence or
`absence of this antibody. If the anti-p185’7“"? monoclonal
`antibody 4D5 inhibited proliferation of breast tumorcells by
`interfering with the signalling functions of p185‘7"*?, addi-
`tion of this antibody would be expected to enhance the
`sensitivity of tumor cells to TNF-a. Both SK-BR-3 (Fig. 9A)
`and MDA-MB-175-VII (Fig. 9C) were growth inhibited by
`both the monoclonal antibody 4D5 (5 pg/ml; 50% and 25%
`inhibition, respectively) and high concentrations of TNF-a
`
`(Q x 10* units/ml; 50% and 60% inhibition, respectively).
`However, the combination of TNF-a and monoclonal anti-
`body 4D5 reduced the SK-BR-3 and MDA-MB-175-VII
`tumor cell number to a level below that
`initially plated,
`indicating the induction of a cytotoxic response. In a sepa-
`rate experiment, SK-BR-3 cell viability was determined
`directly by using trypan blue dye exclusion, yielding identi-
`cal results to those described above that were obtained by
`using crystal violet staining (data not shown). A control
`monoclonal antibody, 40.1.H1, did not
`inhibit SK-BR-3
`breast tumorcell proliferation, nor did it induce an enhanced
`sensitivity of this cell line to the cytotoxic effects of TNF-a
`(Fig. 9B). In addition, the growth of the breast tumorcell line
`MDA-MB-231, which does not express detectable levels of
`p185/7£”? (17), was unaffected by monoclonal antibody 4DS5,
`and the growthinhibition seen with the combination of 4D5
`and TNF-a wassimilar to that observed with TNF-a alone
`(Fig. 9D). Furthermore, neither HBL-100 (30), a nontrans-
`formed but immortalized human breast epithelial cell line
`(Fig. 9E), nor T24 (27), a human bladder carcinomacell line
`(Fig. 9F), expressed high levels of p1857"*? (data not
`shown), and neither demonstrated growth inhibition by 4D5
`or an enhanced growth-inhibitory or cytotoxic response to
`the combination of TNF-a and monoclonal antibody 4D5.
`These results demonstrate that only tumor cells which
`
`

`

`VOL. 9, 1989
`
`MONOCLONAL ANTIBODIES AND p85?
`
`1171
`
`overexpress p185“*? will become sensitized to the cyto-
`toxic effects of TNF-a by antibody 4DS.
`
`DISCUSSION
`
`the
`We have prepared monoclonal antibodies against
`extracellular domain of the HER2/c-erbB-2 gene product,
`p185#"*? | and have found that one of these, 4D5, strongly
`inhibits the growth of several breast tumorcell lines and
`furthermore sensitizes p185/7"*?-overexpressing breast car-
`cinoma cell lines SK-BR-3 and MDA-MB-175-VII to the
`cytotoxic effects of TNF-a. Monoclonal antibody 4DS5 is
`specific for p1857"*? and showsno cross-reactivity with the
`closely related human EGF receptor expressed in mouse
`fibroblasts. Of six mammary carcinoma cell lines tested,
`only the three lines which express high levels of p185/77*?
`(SK-BR3, MBA-MB-175, and MDA-MD-361 [17]) were
`growth inhibited, and 4D5 did not inhibit the proliferation of
`a nontransformed human breast epithelial cell line, HBL-
`100, or the bladder carcinomacell line T24.
`In the presenceofthe antibody, th