MOLECULAR AND CELLULAR BIOLOGY, feb. 1991, p. 97�986
`0270-7306/911020979-08$02.00/0
`Copyright© 1991, American Society for Microbiology
`
`Vol. 11, No. 2
`
`Regulation of Phosphorylation of the c-erbB-2/HER2 Gene Product
`
`
`
`by a Monoclonal Antibody and Serum Growth Factor(s)
`in Human Mammary Carcinoma Cells
`RAKESH KUMAR,1* H. MICHAEL SHEPARD,2 AND JOHN MENDELSOHN1•3
`Laboratory of Receptor Biology, Memorial Sloan-Kettering Cancer Center,
`
`
`1275 York Avenue, New York,
`
`Biology, Genentech Inc., South San Francisco, California 940802;
`New York 100211; Developmental
`Medical College, New York, New York 100213
`and Cornell University
`
`Received 9 May 1990/Accepled 17 November 1990
`
`Monoclonal antibody (MAb) 4D5 was used to analyze the phosphorylation
`of p185H£RZ, the gene product of
`
`c-erbB-21HER2, lo SK-BR-3 cells. Culture in the continuous presence of 4D5 reduced the in vivo steady-state
`levels of p185HERZ phosphorylation
`by 80% in a dose-dependent manner, sugges ting that MAb 4D5 may have
`of p185HERZ . The observed MAb-mediated reduction of
`interfered with the activation of phosphorylation
`
`
`
`
`When cultures were accounted for by d.own-regulation. p185HERZ phosphorylation could not be completely
`
`
`
`phosphorylation grown under serum-free conditions, the steady-state levels of p185H£RZ were reduced by 56%,
`
`
`
`
`exposure With continuous and addition of 4D5 further lnhibited phosphorylation to 20% of steady-state levels.
`
`to increasing concentrations of newborn calf serum in these cultures, there was a linear increase in
`tyrosine-specific phospho� lation of p185HERZ, reaching a 5.4-fold
`10% newborn calf serum.
`increase with
`
`Phosphorylation of p185H R2 in the presence of newborn calf serum was not attributable
`
`to stimulation of the
`growth factor-a. Extension
`
`
`epidermal growth factor receptor by epidermal growth
`factor or by transforming
`mammary
`carcinoma ceU llnes, MDA-MB-453 and BT-474, also demon­
`of these observations to two other
`
`a significant capacity of serum to lnduce p185HERZ phosphorylation. The demonstrati
`strated
`on of antibody·
`mediated partial
`
`
`lnhibition of phosphorylation under serum-free conditions suggests that mammary carclnoma
`p185HER2 . Our observation
`cells may also produce and secrete a factor or factors which may activate
`that
`of p185H£RZ by newborn calf serum and by
`
`growtb-lnhibitory MAb 4D5 is able to reduce the phosphorylation
`
`factor(s) which uses phosphorylation of p185H£R2
`
`a ceUular-derived factor(s) suggests the existence or a growth
`as a signal transduction pathway to regulate ceU proliferation.
`
`Proto-oncogenes are a group of normal genes which play
`important roles in the regulation of cell proliferation and
`function (2, 5). Abnormalities in the expression, structure, or
`activity of proto-oncogene products contribute to the devel­
`opment and maintenance of the malignant phenotype in
`complex but important ways (36, 37, 46). Evidence that the
`gene products of several activated proto-oncogenes are
`either growth factors or growth factor receptors has sug­
`gested a possible link between proto-oncogenes and growth
`factors (20). For example, the receptor for macrophage
`colony-stimulating fa<:tor is identical to the product of c-fms
`(35), and c-erbB-1 encodes the receptor for epidermal growth
`factor (13) and transforming growth factor-a (TGF-a) (43).
`Growth factor receptors encoded by proto-oncogenes are
`transmembrane glycoproteins with intrinsic tyrosine-kinase
`activity (22). Receptor tyrosine kinases are activated by
`binding of their respective ligands, the growth factors (48).
`This activity is thought to be an integral part of signal
`transduction processes involved in the regulation of cell
`proliferation (21). Overexpression of some growth factor
`receptors has been shown to induce transformed properties
`in recipient cells (11, 32), possibly because of excessive
`activation of signal transduction mechanisms. Furthermore,
`a number of tumor cells with increased expression of growth
`factor receptors also produce ligands for these receptors
`(10).
`HER2 (also known as c-erbB-2 or c-neu), the human
`
`
`
`• Corresponding author.
`
`979
`
`homolog of the rat proto-oncogene neu (9), encodes a
`185-kDa transmembrane glycoprotein with intrinsic tyrosine
`kinase activity which is presumed to be the receptor for an
`as-yet-unident•fied ligand (3, 39). p185H£Ri also has homol­
`ogy to, but is distinct from, the epidermal growth factor
`receptor (EGF-R), which is the product of c-erbB-1. Both
`proteins have a cysteine-rich extracellular domain, a trans­
`membrane domain, and an intracellular tyrosine kinase (4,
`31, 47). In spite of sequence homology between c-erbB-2 and
`c-erbB-1, EGF does not bind to p185HERi (33). p185HERi has
`been shown to be overexpressed or amplified or both in a
`number of human malignancies: breast (45), ovarian (38),
`thyroid (1), lung (7), salivary gland (34), and stomach (50). In
`addition, p185HER2 is a potent oncogene capable of inducing
`transformation and tumorigenesis when overexpressed in
`NIH 3T3 cells (12, 19). Overexpression of p185HERi also
`induces tumor cell resistance to macrophage killing (15).
`Thus p185HERZ may have an important role in the develop­
`ment and maintenance of human tumors.
`These observations suggest that receptor-associated ty­
`rosine kinase activity of overexpressed proto-oncogene pro­
`tein products is important for the regulation of cell growth.
`We have developed a panel of monoclonal antibodies
`(MAbs) reactive with domains of the human EGF-R (23) and
`p185HERi (17, 18) in intact cells and have demonstrated
`antiproliferative effects of these antibodies in vitro (16, 18,
`23) and in vivo (29). Antibody 4D5, which is specifically
`directed against p185HERZ , exhibits strong antiproliferative
`activity on cultured human breast tumor cell lines which
`overexpress p185HERi (18). Since p185HERi is a receptor
`
`1 of 8
`
`BI Exhibit 1088
`
`

`

`980
`
`KUMAR ET AL.
`
`MOL. CELL. BIOL.
`
`2 3 4 5
`
`with intrinsic tyrosine kinase activity, we investigated the
`
`
`
`
`mod11la�i'.m of p185HERi phosphorylation
`by MAb 405. We
`
`report here that activation of phosphorylation of p185HERi
`
`by serum was reduced in the presence of an excess of MAb
`
`was mediated 4D5 and that MAb-sensitive phosphorylation
`
`by a growth factor or factors other than TGF-a or EGF.
`
`
`
`Furthermore, SK-BR-3 cell-conditioned medium contained a
`FIG. 1. Effect of MAb 405 on steady-state levels of pl85HERZ
`
`factor(s) that could activate p185HERi phosphorylation
`and
`phosphorylation in SK-BR-3 cells. Subconfluent cultures were la­
`
`
`was partially inhibited by MAb 405.
`beled with 32P; (400 µ.Ci in 1 ml of phosphate free F-12/DMEM
`supplemented with 5% newborn calf serum) in the continuous
`presence of different amounts of antibody for 15 h. Detergent
`MATERIALS AND METHODS
`extracts were made, and p185HERz was immunoprecipitated by using
`Materials. MAbs 405 (18) and 9G6 (44) were raised against
`MAb 9G6 and then resolved by 7% SDS-polyacrylamide gel elec­
`human pl85HER2• MAbs 528 and 225 bind to the human
`trophoresis (Materials and Methods). An autoradiogram resulting
`from 16 h of exposure of the dried gel is shown here. The arrow
`
`EGF-R (23). Antiphosphotyrosine MAb PY-69 was obtained
`indicates the position of"P-Jabeled pl8511e1u. Lane 1, Control cells;
`
`
`from ICN Biochemicals, Inc. Rabbit immunoglobulin to
`lanes 2 to 4, cells treated with MAb 405 at 30, 150, and 300 nM,
`
`mouse immunoglobulins G (RAM) was supplied by Accurate
`respectively; lane 5. cells treated with 400 nM F(ab) fragment of
`
`
`Chemicals, Westbury, N.Y. 32P; (carrier free; 28.5 Ci/nmol)
`MAb 405. The amounts (in counts per minute) of p18511F.Rz in each
`
`
`and 35S-labeled L-cysteine (1,030 Ci/mmol) were purchased
`lane were 4,453 (lane 1), 1,967 (lane 2), 1.785 (lane 3), 1,040 (lane 4),
`from New England Nuclear, Boston, Mass.
`and 335 (lane 5). Counts were corrected by subtracting the back­
`Cell lines and cell culture. Human breast
`tumor cell lines
`ground of 60 cpm. The results shown are representative of results in
`
`SK-BR-3, BT-474, and MOA-MB-453 were obtained from
`six different experiments.
`
`
`the American Type Culture Collection. The A431 human
`
`
`epidermal carcinoma cell line was originally supplied by
`Gordon Sato. All cell lines except MOA-MB-453 (which was
`the 185-kOa HER2 protein, resolved as described above,
`
`grown in L-15 medium) were maintained in Ham F-12-
`
`was excised out of the gel. 32P-labeled pl85HER.z in a gel slice
`0ulbecco modified Eagle medium (1:1, vol/vol) (F-12/
`
`
`was partially hydrolyzed with 200 µ.l of 6 N HCl at ll0°C for
`
`OMEM) supplemented with 10% fetal bovine serum.
`
`
`
`1 h. Two portions (10 µ.l each) of the hydrolysate were taken
`Labeling or pl858ER2 with 321>1 and [3!S)cysteine. Cells (3 x
`
`
`for measurement of radioactivity in a liquid scintillation
`105) were plated in F-12/DMEM in each well of a six-well
`
`counter to determine the total incorporation of 32P into the
`pl85HERi receptor.
`
`dish. Twenty-four hours later, cultures were washed with
`
`The rest of the hydrolysate was dried,
`
`phosphate-free medium and incubated for up to 15 h in
`
`suspended in distilled water, and applied to a Oowex
`
`
`phosphate-free F-12/DMEM containing 0.4 mCi of 32P; per
`AG1-X8 column. The column was washed with distilled
`ml in the presence or absence of MAb and newborn calf
`
`
`were eluted water, and the absorbed 32P-labeled materials
`serum. At desired times, cells were harvested
`in 400 µ.l of
`
`
`with 0.5 N HCI and lyophilized. The recovery of radioactiv­
`lysis buffer (20 mM HEPES [N-2-hydroxyethylpiperazine­
`ity by this procedure was 78 to 85%. 32P-phosphoamino
`
`N' -2-ethanesulfonic acid; pH 7.5), 1% Triton X-100, 10%
`
`acids mixed with unlabeled carrier phosphoamino acids
`glycerol,
`
`1.5 mM magnesium chloride, 1 mM ethyleneglycol
`(phosphoserine, phosphothreonine, and phosphotyrosine
`
`
`bis-N,N,N' ,N'-tetraacetic acid, 0.1 mM phenylmethylsulfo­
`
`
`
`[1:1:1)) were analyzed by thin-layer electrophoresis as de­
`nyl fluoride, 10 µ.g of leupeptine
`per ml, 2 mM sodium
`scribed elsewhere (8).
`
`orthovanadate) at 4°C for 20 min. The lysate was centrifuged
`at 10,000 rpm in an Eppendorf microfuge for 10 min, and
`RESULTS
`then 60 µ.I of Pansorbin was added as described elsewhere
`
`MAb 405 reduces amount of 32P-labeled pl8SH£R2• MAb
`
`
`
`(42). For labeling with [35S]cysteine, the cells were washed
`
`with cysteine-free medium and refed with cysteine-free
`
`405 was used to investigate the regulation of phosphoryla­
`
`
`F-12/DMEM containing 0.15 mCi of [35S]cysteine per ml
`tion. SK-BR-3 cells, which have an amplified c-erbB-2 gene
`with or without 5% newborn calf serum.
`
`
`32P; in the (45), were cultured for 15 h in medium containing
`
`
`
`continuous presence of various concentrations of MAb 405.
`lmmunoprecipitation and SDS-polyacrylamide gel electro­
`phoresis. Aliquots (350 µ.l) of the cell lysates (or equal
`
`The pl85HERi from these cells was immunoprecipitated with
`amou nts of trichloroacetic acid-precifitable counts per
`another anti-pl85HERl MAb, 9G6, which recognizes a dis­
`tinct epitope of p185HER.z
`minute) containing 32P-labeled or [ 5S]cysteine-labeled
`
`, and resolved by SOS-polyacryl­
`
`
`pl85HERi were subjected to immunoprecipitation with 10 µ.g
`
`
`amide gel electrophoresis. Results of such an experiment are
`of MAb 9G6, 528, or PY-69 at 4°C for 2 h. Immune
`shown in Fig. 1. Treatment of cells with 405 reduced in vivo
`pl85HERi up to 80% in a
`
`
`complexes were collected by absorption to RAM-protein
`
`steady-state levels of 32P-labeled
`
`A-Sepharose beads at 4°C for 1 h. Beads were washed three
`
`
`dose-dependent manner (lanes 2 through 4). There was 49%
`by 150 nM MAb 405 in
`
`times with 1 ml of buffer (20 mM HEPES [pH 7 .5], 150 mM
`
`:: 8% reduction in phosphorylation
`NaCl, 0.1% Triton X-100, 10% glycerol, 2 mM sodium
`
`
`eight different experiments. When the F(ab) fragment of 405
`
`
`orthovanadate). Washed pellets were mixed with 40 µ.I of
`
`was used instead of intact antibody, comparable or greater
`pl85HERi was observed (lane 5). As
`sample loading buffer (10 mM Tris HCI [pH 6.8], 1% sodium
`reduction of 32P-labeled
`
`
`dodecyl sulfate [SOS], 0.2% 2-�-mercaptoethanol, 10% glyc­
`
`
`another MAb, a control, SK-BR-3 cells were incubated with
`erol, 0.001% bromophenol blue), heated at 95°C for 5 min,
`
`
`225 lgGI, specifically directed against the EGF-R. and there
`
`was no effect on the amount of 32P-labeled p185HERZ (un­
`
`
`and resolved on a 7% SOS-polyacrylamide slab gel (26). The
`
`
`efficiency of precipitating labeled receptor with MAb 9G6 is
`
`published data). The reduction in steady-state levels of
`80 to 90% when this procedure is used. Low-molecular-mass
`
`
`by 405 32P-labeled pl85HERi was not due to interference
`
`
`colored markers (Amersham Corp.) were used as standards.
`
`
`with MAb 9G6 during the immunoprecipitation reaction, as
`
`Phosphoamino acid analysis. The band corresponding to
`
`
`immunoprecipitation performed with another polyclonal an-
`
`2 of 8
`
`BI Exhibit 1088
`
`

`

`VOL. 11, 1991
`
`MODULATION OF PHOSPHORYLATION OF pl85Ht:R�
`
`981
`
`Abs
`
`Pl85
`2 31
`___ ...
`
`EGF-R
`11 21 31 I
`
`FIG. 2. Specificity of the reduction of 31P-labeled pl85H£RZ by
`MAb in SK-BR-3 cells in the presence or absence of MAb 405.
`Subconfluent cells were labeled with 32P; for 15 h. The cells were
`lysed in 600 µI of extraction buffer and divided into two equal parts
`of 250 µI each. Immunoprecipitation was performed with anti-pl85
`MAb (lanes l to 3) or with anti-EGF-R MAb 528 (lanes I' to 3'). An
`autoradiogram of a dried gel is shown here. Lane 1 and 1'. Control;
`lanes 2 and 2', 30 nM MAb 405; lanes 3 and 3', 150 nM MAb 405.
`Counts per minute: lane l, 5,985; lane 2, 3,798; lane 3, 3,120; lane l'.
`853; lane 2', 779; lane 3'. 932. Abs, Antibodies.
`
`tibody (18) recognizing the carboxy-terminal 17 amino acids
`of pl85HERi gave similar results (unpublished data).
`Next, we examined the possibility of general inhibitory
`effects of MAb 405 on the steady-state levels of other
`32P-labeled receptor proteins by analyzing the amount of
`32P-labeled pl85HERZ and 32P-labeled EGF-R in the same
`experiment (Fig. 2). These results indicated that there was
`no reduction of 32P-labeled EGF-R during 15 h of treatment
`of SK-BR-3 cells with 150 nM MAb 405. which had reduced
`the amount of 32P-labeled pl8511£R by 48%.
`Analysis or reduction or p18SHERZ phosphorylation. The
`reduction of steady-state levels of 32P-labeled pl85Ht:Ri by
`MAh 4D5, shown in Fig. 1 and 2. could result from down­
`regulation of pl85HERi and/or interference in the activation
`of pl85Ht:Ri phosphorylation by a direct or indirect m.:cha­
`nism(s). In initial studies to explore these possibilities,
`parallel cultures of cells were metabolically labeled with
`[35S)cysteine or 32P;. During 11 h of concurrent incubation
`with MAb 405, there was a 45% reduction in 32P-labeled
`pl85HERZ (Fig. 3A) and only a 14% reduction in 35S-labeled
`pl85HER2 (Fig. 38). This suggests that the reduced 32P label
`in pl85Ht:Rz in the presence of MAb 405 can only partially
`be attributed to reduced pl85H£Rz content. Next, we per­
`formed a similar experiment comparing the capacities of the
`monovalent F(ab) fragment of MAb 405 and an intact MAb
`405 to affect the reduction of 35S-labeled pl8511£Rz. There
`was no change in 35S-labeled pl8511ERl in the presence of
`F(ab), but there was a 26% reduction caused by MAb 405
`(Fig. 3C, lanes 3 and 2, respectively). The results obtained in
`the immunoprecipitation experiments documented in Fig.
`3A through C were confirmed by immunoblotting (0). lmmu­
`noblotting of the 32P-labeled SK-BR-3 cell extracts used in
`Fig. 2 demonstrated only a marginal reduction in the content
`of pl8511ERz protein when cells were cultured in the presence
`of MAb 405 but a substantial reduction in the amount of
`32P-labeled p18511£Ri (Fig. 2). The expression ofEGF-R was
`not affected. Immunoblotting of similar unlabeled SK-BR-3
`extracts also demonstrated very little reduction in the con­
`tent of p18511ERi by MAb 405 (Fig. 30, experiment 2).
`These findings indicate that increased receptor catabolism
`induced by a MAb cannot fully account for the observed
`reduction in 32P labeling and show [with F(ab)) that reduced
`labeling is dissociated from reduced content of p185HER�
`.
`Next we addressed the possibility that the reduction in
`32P-labeled pl85HER2 associated with exposure to MAb 405
`could be related to a change in the level of expression of
`pl85HERi on the plasma membrane or to the extent of
`
`A.
`
`B.
`
`C.
`
`__.. -
`
`405
`
`+
`D.
`EXP l
`EGF-R
`Abs Pl85
`
`""""" - __.. "
`+
`
`2 3
`
`EXP. 2
`Pl85
`
`+
`+
`405 - +
`FIG. 3. Analysis of the reduction ofpl85H£RZ phosphorylation in
`SK-BR-3 cells treated with MAb 405. Cells were labeled with 32P;
`(A) or [3'S]cysteine (8) in the presence or absence of MAb 405 (150
`nM) for 11 h. Samples were prepared and separated as described in
`Materials and Methods. The autoradiogram shown here was ob­
`tained by 6 h of exp0sure. (Cl Cells were labeled with [3'SJcv�teine
`for 11 h in the presence of MAb 405 (150 nM, lane 2) or F(ab) <400
`nM. lane 3) or with culture medium (lane 1). Samples were prepared
`and immunoprecipitation was carried out as described in Materials
`and Methods. An autoradiogram of a dried gel is shown here.
`Quantitation of the pl85H£RZ bands was obtained by densitometric
`scanning (A through C) or by determining radioactivity associated
`with bnnds (A nnd B). Qunntitation by determining the radioactivity
`associated with pl85H£R1 bands in panels A and B gave results
`�imilar to tho�e with densitometric scanning, and there was a 27% :!:
`3% additional rec!t1c1in1' in 32P-labeled pl85H£R2 compared with
`3'S-labeled p18.'iH£Hi. ID) lmmunoblotting of pl85H£Ri and EGF-R
`proteins. In experiment 1 (Exp. I). 32P-labeled SK-BR-3 cell extracts
`(50 µg of pr<>teinl used in Fig. 2. lanes 1 and 2, were resolved on a
`7'n SDS-p0lyacrylamide gel and then immunoblotted with anti-Pl85
`M:\b 9G6 or anti-EGF-R polyclonal antibody RK-11. Experiment 2
`�hows the immunoblotting of unlabeled SK-BR-3 cell extracts
`µrepared following culture for 15 h with or without 30 nM MAb 405.
`Since some of the extracts used here were radiolabcled. immuno­
`blotted membranes were visualized by using a protein A-gold
`�11ne1111:ement kit \30). Abs, AntibvJ .. :s.
`
`down-regulation of receptor protein. First, we determined
`what fraction of the 35S-labeled p18511ERi is present on the
`cell surface at 37°C (Fig. 4A). In these experiments.
`p185ut:Rz expressed on the plasma membrane was identified
`by its capacity to bind MAb 405 prior to cell lysis. The
`results indicate that 19% ± 4% (average from three different
`experiments) of total 35S-labeled pl8511ERi is expressed on
`the cell surface under these experimental conditions; thus
`pl85HERz is available for down-regulation by MAb 405.
`Down-regulation of EGF-R has been shown to be dependent
`on temperature (41). To confirm that down-regulation of
`surface pi85Ht:R.z also is reduced at 4°C, experiments were
`performed to analyze the effect of temperature on the
`abundance of 35S-labeled p18511ERZ on the cell surface.
`Results indicated that at 4°C the amount of total l5S-labeled
`pl8511£Rz expressed on the surface increased to 35% ± 3%
`(data not shown) compared with 19% ± 4% of total 35S­
`labeled p18511£� at 37°C.
`In order to define the contribution of down-regulation to
`MAb-induced reduction in p18511ERZ phosphorylation, we
`
`3 of 8
`
`BI Exhibit 1088
`
`

`

`982
`
`KUMAR ET AL.
`
`MOL. CELL. BIOL.
`
`A. EXP.l
`I l 21
`
`EXP.2
`I 3 41
`
`B.
`
`4°C
`37°C
`2 3114 5
`
`- +
`
`- +
`
`405
`FIG. 4. (A) Quantitation of surface expression of 35S-labeled
`pl85HERz. Cells were labeled with C"S]cysteine for 11 h. At the end
`of incubation, some cultures were lysed in 500 µI of lysis buffer for
`the determination of total "S-labeled pl85HERZ by immunoprecipi­
`tation with 10 µg of 405 (Materials and Methods). For measuring the
`surface expression of 35S-labeled pl85HERZ, cultures were washed
`with phosphate-buffered saline and further incubated with F-12/
`OMEM-20 mM HEPES (pH 7 .5) containing 20 µg of high-affinity
`MAb 405 per ml for 1 h at 4°C. The cultures were washed, lysed in
`500 µI of extraction buffer, and processed for immunoprecipitation
`by adding RAM-protein A-Sepharose beads but no more MAb 405
`during the immunoprecipitation procedure. The results of two
`representative experiments are shown here. Lanes 1 and 3, Total
`35S-labeled pl85HERZ; lanes 2 and 4, 35S-labeled pl85HERZ on the cell
`surface. (B) Effect of MAb 405 on surface expression of pl85H£RZ in
`a temperature shift experiment. SK-BR-3 cells were first equili­
`brated with 32P; for 4 h at 37°C (lane 1) and then further incubated
`with or without MAb 405 for an additional 11 h (in the continuous
`presence of 32P;) either at 37°C (lanes 2 and 3) or at 4°C (lanes 4 and
`5). The autoradiogram shown here was obtained by exposing lanes
`1 to 3 for 24 h and lanes 4 and 5 for 96 h. Quantitation of the amount
`of 32P associated with pl85HERZ bands was obtained by densitomet­
`ric scanning of the autoradiogram and by determining the radioac­
`tivity associated with pl85HERZ bands (A and BJ.
`
`analyzed the etfec.t of incubation with MAb 405 at 4°C. In
`these studies, cells were first equilibrated with 32P1 for 4 h at
`37°C (Fig. 48, lane 1) and then maintained at 37°C (lanes 2
`and 3) or shifted to 4°C (lanes 4 and 5) for an additional 11 h,
`with or without MAb 405. A comparison of the labeled
`material in Janes 2 and 4 in Fig. 48 (ftuorographs exposed for
`24 and 96 h, respectively) showed a significant 85% reduc­
`tion in 32P labeling of pl85HERi during 11 h at 4°C compared
`with labeling at 37°C. However, incubation of cells a t 4°C uiu
`not prevent a further substantial MAb-mediated reduction in
`steady-state levels of pl85HERi phosphorylation: there was a
`34% decrease at 4°C (compare lanes 4 and 5) and a 51%
`decrease at 37°C (compare lanes 2 and 3). Taken together,
`these observations indicate that MAb-induced reduction in
`pl85HER phosphorylation cannot be completely accounted
`for by down-regulation.
`Experiments were performed to determine whether the
`F(ab) fragment might have the capacity to act as an agonist
`by activating tyrosine phosphorylation. The results in Fig. 5
`indicate that the addition of F(ab) for 15 min slight15 stimu­
`lated in vivo tyrosine phosphorylation of pl85HER
`in cul­
`tures labeled with 32P (Fig. 5, lane 2'). However, there was
`no activation in cultures exposed to F(ab) for a longer
`treatment of 60 min (Fig. 5, lane 3 '). The observation that the
`F(ab) fragment of 405 does not down-regulate the 35S-
`
`Antibody Pl85
`l 2 3
`
`-
`
`P-Tyr
`11 21 311
`
`FIG. 5. Partial agonist nature of F(ab). Subconftuem SK-BR-3
`cells were labeled with 32P; for 15 h. Some cultures were treated with
`400 nM F(ab) for the indicated times. The cells were lysed in 800 µI
`of extraction buffer. The lysates were divided into two equal parts of
`350 µI each and then immunoprecipitated with MAb 906 (lanes 1 to
`3) or with antiphosphotyrosine MAb PY-69 (lanes 1' to 3'). An
`autoradiogram resulting from a 1-h exposure of dried gel is shown
`here. Lane 1, Control; lane 2, F(ab) incubation for 15 min; lane 3,
`F(ab) incubation for 60 min.
`
`labeled pl85HER2 but can act for a short time as a partial
`agonist is interesting; however, we have not attempted to
`further characterize these properties in the present study.
`Activation of phosphorylation of pl858ER2 in presence or
`absence of newborn calf serum. Next, we investigated the
`possible source of the factor(s) that might stimulate
`pl85HER2 phosphorylation. As shown in Fig. 6A, culturing
`the cells in serum-free medium resulted in a steady-state
`level of phosphorylation of pl85HERi reduced 56% (lane 1)
`compared with that observed in the continuous presence of
`newborn calf serum (lane 3). The addition of MAb 405 in
`serum-free culture conditions further reduced pl85HERi
`
`A.
`
`2 3 4
`- . .. . -
`
`B.
`
`Serum
`
`405
`C.
`
`+ +
`+
`
`+
`
`•
`
`+ +
`
`y
`
`3
`2
`FIG. 6. (A) Detection of newborn calf serum-mediated phos­
`phorylntion of pl8SHERZ. Subconfluent SK-BR-3 cells were labeled
`with 32P1 in the culture medium without (lanes 1 and 2) or with (lanes
`3 and 4) 5% newborn calf serum for 15 h. Cultures analyzed in lanes
`2 and 4 also were continuously exposed to 150 nM MAb 405.
`Samples were prepared and immunoprecipitated for assaying the
`amount of pl85H ERi as described in the Materials and Methods.
`Quantitation of the pl85 bands was obtained by densitometric
`scanning of the autoradiogram. (B) Control experiment showing
`effect of serum on the 32P1 labeling of EGF-Rs for 15 h in SK-BR-3
`cell cultures. Cell extracts were immunoprecipitated with anti­
`EGF-R MAb 528, which recognizes one distinct band with an
`approximate molecular mass of 170 kDa (arrow). (C) Two-dimen­
`sional thin-layer electrophoresis pattern of 32P-phosphoamino acids
`in a hydrolysate of the pl85HERZ immunoprecipitated in panel A. S,
`Phosphoserine; T, phosphothreonine; Y, phosphotyrosine. Number
`at lower left of each autoradiogram indicates the following culture
`conditions: l, with no serum; 2, with serum; 3. with serum and MAb
`405. Tyrosine phosphorylation in control cells was visualized
`faintly on the autoradiogram but reproduces poorly.
`
`4 of 8
`
`BI Exhibit 1088
`
`

`

`VOL. 11, 1991
`
`MODULATION OF PHOSPHORYLATION OF pl85HER2
`
`983
`
`Pl85
`2 3 4
`
`P-Tyr
`2' 31 4'
`
`11
`
`%Serum
`
`0
`
`2.5 5
`
`10
`
`0 2.5
`
`5
`
`10
`
`C.
`
`1 2 3 4
`
`/'
`
`500
`
`It)-
`co E
`
`a:c �8 300
`>--t- 0
`d.� 100
`
`10
`
`phosphorylation (Fig. 6A, lane 2) to 20% of the steady-state
`
`A. Antibody
`levels achieved in the absence of newborn calf serum (Fig.
`
`6A, lane l). Experiments were done to examine the capacity
`
`
`of newborn calf serum to stimulate tyrosine phosphorylation
`
`
`of p185H£Rl in short-term freatment. There was no increased
`
`
`activation of phosphorylation when serum-free cultures
`with newborn calf serum for 30 min at 37
`were supplemented
`or 4°C (data not shown).
`.
`To determine the specificity
`of the ca� city of newborn
`activation of p185 £R2 phosphoryla­
`calf seru m to stimulate
`
`
`tion in SK-BR-3 cells, we investigated the potential for
`B.
`
`serum activation of another closely related molecule, the
`
`EGF-R. There was no potentiating effect of newborn calf
`
`
`serum on phosphorylation of the EGF-R in SK-BR-3 cells
`(Fig. 6B).
`Hav� shown an increase in the steady-state
`levels of
`
`p185H phosphorylation induced by newborn calf serum
`./
`
`and its reduction by MAb 405, we determined the phos­
`phoamino acid content of p185H£Ri under these conditions
`
`
`
`by two-dimensional thin-layer electrophor�sis (Fig. 6C).
`0
`5
`p185H£R2 from Cells Cultured in the absence Of newborn Calf
`Percent Serum, VIV
`
`serum contained predominantly phosphoserine and phos­
`FIG. 7. (A) Detection of tyrosine-specific phosphoryl at.ion of
`
`
`photlireonine with little phosphotyrosine (Fig. 6C, blot l)�
`pl85HER.z by newborn calf scrum. Subconfluent SK-BR-3 cells were
`metabolically labeled with [35S]cysteine m the presence of dilferent
`
`The presence of some phosphorylation on tyrosine can be
`concentrations of serum for 15 h. Cells were lyscd in 650 µI of
`
`
`
`demonstrated by longer exposure of the autoradiogi-am but is
`
`
`not visualized well in the figure shown. Quantitation of the
`extraction buffer. The lysates were divided into two equal parts of
`
`relative amount of label in each amino acid was obtained by
`300 µI each and then immunoprecipitated with anti-p18SH£R2 MAb
`9G6 (lanes 1 to 4) or antiphosphotyrosine MAb PY-6.9 Oanes l' to
`
`scraping the ninhydrin-ideiltified spots from the thin-layer
`4'). Other details of the assay were as described in the legend to Fig.
`
`
`by 5% plate for liquid scintillation counting. Activation
`5. To detect the phosphotyrosine (P-Tyr) signal in cells cultured in
`n�wborn calf serum increased the total phosphoamino acid
`the absence of serum (lane l'), it was necessary to expose the
`
`content 2.2-fold, while for phosphotyrosine, the increase
`autoradiogram for 20 h, which resulted in overexposure of lanes l to
`
`
`was 3.9-fold (Fig. 6C, blot 2). Inhibition of newborn calf
`4. (8) To quantitate the data in panel A, the radioactivity associated
`
`
`
`
`serum-mediated Stimulation ofp185H£Rl phosphorylation by
`with p185H£R2 was determined by counting the excised bands in a
`
`
`MAb 405 resulted in a parall el reduction in the content of all
`liquid scintillation counter. The ratios of counts in p185HERz phos­
`
`three phosphoamino acids (Fig. 6C, blot 3).
`photyrosine over total counts in p185HERJ were plotted as a percent­
`age of control with 0% newborn calf serum against the concentration
`Tyrosine phosphorylation of pl8SH£R2. To further quanti­
`of newborn calf serum used in the culture medium. (C) Depletion of
`
`
`
`tate tlie relative increase in the steady-state phosphotyrosine
`
`content of p185H£R2 induced by newborn calf serum, cells
`the activator(s) of pl85H£R2 phosphorylation in serum. SK-BR-3
`cells were labeled with 32P1 in the absence or presence of scrum for
`
`
`were metabolically labeled with [35S]cysteine and assayed
`15 h. Lane 1, Control without scrum; lane 2, medium with 5%
`
`
`for the steady:state phosphotyrosine content of pl85H£Ri by
`newborn calf serum; lane 3, medium with 5% newborn calf serum
`
`using antiphosphotyrosine MAb PY-69. MAb PY-69 was
`depleted of factor(s) by three repetitive adsorptions of 4 h each on
`
`specific for phosphotyrosine in the immunoprecipitation
`SK-BR-3 cells at 4"C; lane 4, control medium adsorbed on A431
`
`reaction; i.e., we were able to show competion for binding
`cells. Cell extracts were prepared and pl85H£R2 was assayed as
`
`and not with phosphoserine in
`with cold
`phosphotyrosine
`described in Materials and Methods.
`
`EGF-R (data not shown). As
`experiments with 35S-labeled
`in Fig. 7A, lanes l' to 4', the amount of phosphor­
`illustrated
`ylation of pl85H£Ri on tyrosine
`
`increased with the concen­
`pleted of such a factor(s). In these experiments, phosphate­
`
`
`tration of newborn calf serum in the culture medium. The
`
`free medium containing 5% newborn calf serum was treated
`
`
`
`level of activation of tyrosine phosphorylation in serum-free
`
`
`by repetitive absorption with SK-BR-3 cells (three times, for
`medium was 18% of that observed in 10% newborn calf
`4 h each time, at 4°C). Culture medium treated in an identical
`
`serum, and in medium containing 2.5% serum, tyrosine
`
`manner by adsorption with A431 cells, which do not express
`Figure 7C
`
`
`phosphorylation was 35% of that observed with a serum
`
`high levels of p185HER2, was used as control.
`
`
`concentration of 10%. As a control, equal amounts of labeled
`shows that there was a 53% reduction in p185H£R2 phosphor­
`
`cell extracts were immunoprecipitated with anti-pl85H£R2
`
`ylation in SK-BR-3 cells cultured in the presence of medium
`MAb 9G6 (Fig. 7 A, lanes 1 through 4). There was no
`with SK-BR-3 cells Oane 3) compared with
`preadsorbed
`significant effect of serum on the levels of 35S-labeled
`untreated medium (lane 2), and there was only a 16%
`in p185HERz phosphorylation
`
`pl85H£R2 in the cells. To quantitate these results, the ratios
`reduction
`when SK-BR-3 cell
`
`
`of phosphotyroslne-associated counts to total counts asso­
`
`
`cultures were supplemented with medium preadsorbed with
`ciated wilh pl85H£R2 are presented in Fig. 78, which shows
`A431 cells (Fig. 7C, lane 4).
`Activation factor(s) In newborn calf serum was not TGF-a
`
`
`
`a dose-dependent increase of up to 5.4-fold with 10% new­
`born calf serum in the culture medium.
`or EGF. p185H£Ri has been shown to be phosphorylated
`Partial depletion of activating factor from newborn calf
`e� sure to EGF or TGF-a,
`when EGF-Rs are activated by
`which are not ligands for p185H (24, 40). To determine
`
`serum. Since our results indicated the presence of some
`for pl85H£R2 phosphorylation
`
`activating factor(s)
`in new­
`
`
`whether the newborn calf serum-mediated 2.2-fold enhance­
`
`
`ment of pl85H£Ri phosphorylation resulted from the activity
`
`born calf serum, we sought confirmation of this observation
`
`
`by determining whether newborn calf serum could be de-
`
`ofTGF-a or EGF, we used anti-EGF-R MAb 528, which has
`
`5 of 8
`
`BI Exhibit 1088
`
`

`

`984
`
`KUMAR ET AL.
`
`"'
`·E ::>
` 5
`.i5
`l:;
`� a..
`!2
`!: 3
`'O Cll
`l 2
`... 0 c..> c:
`d. Pl
`
`>-l?
`
`4
`
`+ + + + +
`+
`
`+
`
`
`growth factors transduce signals that activate cell prolifera­
`tion (6, 22, 27, 48). Support for this model is provided by our
`experiments demonstrating growth inhibition
`of cells ov