Cell, Vol. 41, 695-706, July 1985, Copyright © 1985 by MIT 0092-8674/85/070695-12 $02.00/0 Down-Modulation of an Oncogene Protein Product and Reversion of the Transformed Phenotype by Monoclonal Antibodies Jeffrey A. Drebin,* Victoria C. Link,*t David F. Stern,$ Robert A. Weinberg,~ and Mark i. Greene*t * Department of Pathology Harvard Medical School Boston, Massachusetts 02115 tDepartment of Medicine Tufts University School of Medicine Boston, Massachusetts 02111 ~Whitehead Institute for Biomedical Research Massachusetts Institute of Technology Cambridge, Massachusetts 02142 Summary Exposure of neu-oncogene-transformed NIH 3T3 cells to monoclonal antibodies reactive with the neu gene product, p185, results in the rapid and reversible loss of both cell-surface and total cellular p185. Although not directly cytotoxic, monoclonal anti-p185 antibody treatment causes neu-transformed NIH 3T3 cells to re- vert to a nontransformed phenotype, as determined by anchorage-independent growth. Isotype matched control antibodies of an unrelated specificity do not affect p185 levels or colony formation in soft agar by neu-transformed NIH 31"3 cells. Soft agar colony for- mation by NIH 3"1"3 cells transformed by ras oncogenes is not affected by anti-p185 antibody treatment. Anchorage-independent growth of cells from the ethylnitrosourea-induced rat neuroblastoma line in which neu was originally detected by DNA transfec- tion is also inhibited in the presence of anti-p185 monoclonal antibodies. Collectively, these results suggest that p185 is required to maintain transforma- tion induced by the neu oncogene. Introduction Malignant cells display a variety of in vitro characteristics that distinguish them from normal cells. These character- istics, collectively known as the transformed phenotype, include anchorage-independent growth, decreased se- rum requirements, rounded cellular morphology, in- creased hexose uptake, loss of microfilaments, increased plasminogen activator secretion, decreased cell surface fibronectin, and increased sensitivity to the drug ouabain (Freedman and Shin, 1974; Pollack et al., 1984; Noda et al., 1983). Anchorage-independent growth, as determined by the formation of colonies in soft agar, is the most reli- able parameter of the transformed phenotype because it is the phenotypic property most tightly linked with tumori- genic behavior in vivo (Freedman and Shin, 1974). The study of acutely transforming retroviruses has re- vealed a class of genes, called oncogenes, capable of rap- idly conferring the transformed phenotype on nontrans- formed cells (Bishop and Varmus, 1982). These retroviral oncogenes originated from cellular genes, called proto- oncogenes, that were transduced by retroviruses (Bishop, 1983). Proto-oncogenes have been highly conserved in evolution (Shilo and Weinberg, 1981), which has led to the suggestion that these genes may play critical roles in nor- mal cellular growth and development. It is thought that linkage to retroviral promoters, mutation of cellular se- quences in retroviral genomes, or both, result in activation of the malignant properties of proto-oncogenes trans- duced by retroviruses (Bishop, 1983). The malignant properties of cellular proto-oncogenes may also be activated by nonviral means. Within the past several years activated cellular oncogenes, capable of neoplastically transforming NIH 3T3 cells in DNA transfec- tion assays, have been identified in a variety of nonvirally induced tumors and tumor cell lines (Cooper, 1982; Land et al., 1983b). Furthermore, proto-oncogene rearrange- ments (Taub et al., 1982; de Klein et al., 1982), duplica- tions (Collins and Groudine, 1982), and aberrant expres- sion (Eva et al., 1982; Slamon et al., 1984) have been observed in a substantial fraction of tumors. Collectively these findings have led to the suggestion that genetic al- terations involving cellular proto-oncogenes may play a critical role in neoplastic transformation. A corollary of this hypothesis is that proteins encoded by activated cellular oncogenes may be specifically involved in the initiation and maintenance of the neoplastic state. The best evi- dence in support of this hypothesis is that mutants of Rous sarcoma virus (RSV) that are temperature sensitive for the ability to transform cells encode an aberrant oncogene product, pp60 src, which is temperature sensitive in its protein kinase activity (Sefton et al., 1980). This suggests that the pp60 src protein kinase activity is responsible for the neoplastic state in RSV-transformed cells. Although many activated cellular oncogenes detected by DNA transfection assays are related to retroviral ras on- cogenes (Der et al., 1982; Parada et al., 1982; Santos et al., 1982; Shimizu et al., 1983), transfection studies have identified a number of oncogenes that are distinct from those found in retroviruses (Cooper, 1982; Pulciani et al., 1982; Cooper et al., 1984; Lane et al., 1984; Padua et al., 1984). We have previously described one such oncogene, which has been isolated from several independent ethylnitrosourea-induced rat neuroblastomas (Shih et al., 1981). This oncogene, which we have termed neu, is related to, but distinct from, the erbB oncogene and its normal cellular homolog, the epidermal growth factor receptor gene (Schechter et al., 1984). The putative neu oncogene product is a 185 kilodalton (kd) phosphoprotein (p185) that is glycosylated and possesses intrinsic tyro- sine kinase activity (Padhy et al., 1982; D. F. S., unpub- lished data). We have developed monoclonal antibodies reactive with domains of the p185 molecule exposed on the surface of intact cells (Drebin et al., 1984). We de- scribe here the effects of these antibodies on p185 protein levels within, and on the transformed phenotype of, cells transformed by neu oncogenes.
`
`IMMUNOGEN 2070, pg. 1
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Cell 696 25C 20C 15C k. IOC 5C !; ii[il -45K i~'i![i!!i~!ii i~ii; il ~ii;iq -31 K ;4" ' i'iiiii,i -22K ;ii:~i,i'' -14K I I I I 6 6 6-- 0 1 3 10 :30 100 500 1000 ng El6.4 Figure 1. Specific Binding of Purified Monoclonal Antibody 7.16.4 to neu-Transformed NIH 3T3 Cells Samples containing 1 x 106 neu transformed NIH 3T3 cells (cell line B104-1-1; solid circles) or H-ras-transformed NIH 3T3 cells were in- cubated (cell line XHT-I-la; open circles) with the indicated amounts of purified antibody 7.16.4 and processed for immunofluorescent flow cytometric analysis as described in Experimental Procedures. (Inset) Coomassie blue staining of monoclonal antibody 7.16.4 (20 t~g) electro- phoresed on an SDS-polyacrylamide gel under reducing conditions. Only immunoglobulin heavy and light chain peptides are observed. Results Effects of Purified 7.16.4 Monoclonal Antibody on Cellular p185 Levels Monoclonal antibody 7.16.4 (IgG2a) was purified from hybridoma ascites fluid by ammonium sulfate precipita- tion and protein A-Sepharose affinity chromatography as described in Experimental Procedures. This resulted in an immunoglobulin preparation that was more than 95% pure as determined by SDS polyacrylamide gel elec- trophoresis and Coomassie blue staining (Figure 1, inset). The ability of this antibody preparation to bind specifically cell surface determinants of a neu-transfected NIH 3T3 cell line (B104-1-1), as quantitated by immunofluorescent flow cytometry, is shown in Figure 1. The binding of anti- body 7.16.4 to B104-1-1 cells saturates at 10-20 ng/106 cells (Figure 1, closed circles). This suggests that 2 x 10 s bind- ing sites exist on the B104-1-1 cell surface if each antibody molecule binds two p185 molecules. In contrast, no binding 100 8O 60 ..j 2O P / / / / / / / / / t Antibody / / Antibody Maintained I I I \, I 0 1 2 4 20 TIME (Hours/ Figure 2. Reversible Down-Modulation of Cell Surface p185 by Monoclonal Antibody 7.16.4 Purified antibody 7.16.4 (50 ,~g/dish) was added to B104-1-1 cell cultures at time 0. At the time points indicated, cells from antibody-treated dishes and untreated control dishes were processed for immunofluo- rescent flow cytometry as described in Experimental Procedures. Data are presented as the percentage of cell surface fluorescence of antibody4reated cells as compared with the untreated control. There is clearly a rapid and persistent down-modulation of cell surface p185 in the presence of antibody (solid circles). Removal of antibody from the culture medium results in reexpression of p185 (open circle). of antibody 7.16.4 to the Ha-ras-transfected NIH 3T3 cell line XHT-I-la is detectable, even at 1000 ng/108 cells (Fig- ure 1, open circles). Untransfected NIH 3T3 cells behave similarly to ras-transfected cells in this assay (data not shown). Thus the purified 7.16.4 monoclonal antibody is specific in its binding to cells that contain the p185 product of the neu oncogene. The antibody binding studies described above were carried out on cells maintained at 4°C in order to prevent any reduction of cell surface p185 due to antigenic modu- lation. To assess whether antibody 7.16.4 could remove p185 from the cell surface, we examined p185 expression on B104-1-1 cells after the addition of antibody 7.16.4 to cell cultures maintained at 37°C. As shown in Figure 2, addi- tion of monoclonal 7.16.4 to cultured B104-1-1 cells causes rapid down-modulation of cell surface p185 expression. This modulation persists as long as cells are cultured in the presence of antibody, and is reversed when antibody is eliminated from the culture media (Figure 2). The ability of polyvalent and monoclonal antibodies to cause down-modulation of their cognate cell surface anti- gens has been described in a number of studies (Boyse et al., 1967; Edelman, 1976; Schreiner and Unanue, 1977; Baufnann and Doyle, 1980; Ritz et al., 1980; Levy and Miller, 1983; Carroll et al., 1984). Generally, divalent IgG molecules are capable of inducing antigenic modulation,
`
`IMMUNOGEN 2070, pg. 2
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Antibody-Mediated Reversion of Transformed Phenotype 697 1oo I 8£ 60 -4 40 o / ] I I I 5 20 t00 5oo Ug PRO TEIIV / DISH 100 8£ 6O -4 40 2O / t I I I 0 1 2 3 Lntact~ \\ I 2O TIME (Hours) Figure 3. Down-Modulation of p185 by Intact Antibody 7.16.4, but Not by Monovalent F(ab) Fragments (A) Varying amounts of antibody 7.16.4 or its F(ab) fragment were added to B104-1-1 cultures at time O. After 3 hr of incubation at 37°C, cells were removed from dishes, restained with saturating amounts of intact 7.16.4 or F(ab) fragments, and processed for flow cytometric analysis as de- scribed in Experimental Procedures. Results are presented as the per- centage of cellular fluorescence of antibody-7.16.4-treated cells as compared with untreated cells immunofluorescently stained by anti- body 7.16.4 (open circles), and the percentage of cellular fluorescence of F(ab)-treated cells as compared with untreated cells immunofluores- cently stained with F(ab) fragments (solid circles). (B) Antibody 7.16.4 or its F(ab) fragment (50/~g/dish) was added to whereas their monovalent F(ab) fragments are not. It is thought that down-modulation results from the ability of divalent antibodies to cross-link cell surface structures, which leads in turn to their internalization or shedding from the membrane; monovalent F(ab) fragments cannot cross-link cell surface antigens, and therefore do not in- duce antigenic modulation. As shown in Figure 3A, ex- posure of B104-1-1 cells to intact antibody 7.16.4 for 3 hr causes p185 modulation at doses as low as 5/~g/dish. In contrast, monovalent F(ab) fragments of antibody 7.16.4 bind to cell surface p185 but do not cause its down- modulation in this amount of time, even at doses of 500 /~g/dish. Figure 3B provides evidence that the lack of an effect of F(ab) fragments on p185 expression does not change with extended exposure to F(ab) fragments. We conclude that down-modulation of p185 by antibody 7.16.4 results from cross-linking of cell surface p185 molecules by the intact antibody molecule. Incubation of B104-1-1 cells with antibody 7.16.4 causes a significant decrease in cell surface p185 expression; antibody-treated cells display only 20% to 40% as much cell surface p185 as untreated cells. To determine whether this antibody-mediated decrease in cell surface p185 results in lower steady state levels of the protein, we meta- bolically labeled B104-1-1 cells for 18 hr with 3sS-cysteine, and then added antibody to the cells for an additional 3 hr. The presence of labeled p185 was demonstrated by im- munoprecipitation of cell lysates with additional antibody followed by SDS-polyacrylamide gel electrophoresis. As shown in Figure 4A, cells incubated with a control IgG2a antibody (lane 4) display comparable amounts of p185 to cells incubated without antibody (lane 2). In contrast, cells incubated with anti-p185 antibody 7.16.4 contain markedly less total p185 (lane 3). The reduction in labeled p185 found in antibody-modulated cells is comparable to the reduction in cell surface p185--a 60% to 80% decrease. It is noteworthy that the major intracellular p185 precursor, which migrates slightly more rapidly than p185 (Figure 4A), is also precipitated by antibody 7.16.4 and does not appear to be affected by antibody-mediated down- modulation of mature p185. This suggests that antibody treatment does not inhibit the synthesis of p185 precursor proteins, and might act by causing loss from the cell of ma- ture p185. To determine whether the decrease in the steady state p185 level is accompanied by loss of the p185 protein from antibody-treated cells, we measured the effect of antibody treatment on the stability of p185. B104-1-1 cells that had been metabolically labeled for 18 hr with 35S-cysteine were washed free of isotope and incubated for 3 hr in the pres- ence or absence of antibody. The presence of labeled p185 was determined by immunoprecipitation and poly- acrylamide gel electrophoresis as described above. Pulse-chase studies, to be presented elsewhere, have shown that the p185 molecule has a half-life of well over 6 hr. As shown in Figure 4B, significant amounts of labeled B104-1-1 cultures at time 0. After varying amounts of time at 37°C, cells were removed from dishes and processed for flow cytometry. Results are presented as in A.
`
`IMMUNOGEN 2070, pg. 3
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Cell 698 I 2 3 4 -pl85 B104.1.1 t ,[ Bt04.f.f + l.ug "~ 16.4 NIH3T5 e Figure 5. Inhibition of Soft-Agar Colony Formation by B104-1-1 Cells in the Presence of Antibody 7.16.4 We plated 1 x 103 cells/dish in soft agar as described in Experimental Procedures. (A-C) Photographs of entire culture plates. (D-F) Repre- sentative colonies photographed at 40x magnification. (A,D) un- treated B104-1-1 cells, (B,E) B104-1-1 cells cultured with 1 ~g 7.16.4, (C,F) untreated NIH 3T3 cells. Solid bars in D-F indicate 200/~m. 1 2 54 567 , ~ ~ ~ i '~i~i p185 Figure 4. Effect of Antibody 7.16.4 on Steady State p185 Levels and on p185 Stability (A) S104-1-1 cells were metabolically labeled with 35S-cysteine for 18 hr and then were incubated in the absence or presence of antibody (25 ~g/dish) for 3 hr. The level of labeled p185 was analyzed by SDS- polyacrylamide gel electrophoresis following detergent lysis and im- munoprecipitation. Equal volumes of 3~S-labeled lysate were im- munoprecipitated in each group. Lane 1: Cells labeled in the absence of antibody, immunoprecipitated with a mixture of normal mouse and normal rabbit serum. Lane 2: Cells labeled in the absence of antibody, immunoprecipitated with antibody 7.16.4. Lane 3: cells labeled in the presence of antibody 7.16.4, immunoprecipitated with antibody 7.16.4. Lane 4: Cells labeled in the presence of control antibody 9BG5, im- munoprecipitated with antibody 7.16.4. (B) B104-1-1 cells were metabolically labeled with 3SS-cysteine for 18 hr and then incubated without label (chased) in the absence or presence of antibody for 3 hr. The level of p185 remaining was analyzed as de- scribed above. Equal amounts (counts per minute) of ~sS-labeled lysate were immunoprecipitated in each group; experimental groups were done in duplicate. Lane 1: Cells chased in the absence of antibody, im- munoprecipitated with a mixture of normal mouse and normal rabbit serum. Lanes 2,3: Cells chased in the absence of antibody, im- munoprecipitated with antibody 7.16.4. Lanes 4,5: Cells chased in the presence of 25 ~g/dish 7.16.4, immunoprecipitated with 7.16.4. Lanes 6,7: Cells chased in the presence of 25/~g/dish control antibody 9BG5, immunoprecipitated with antibody 7.16.4. p185 are precipitated from untreated B104-1-1 cells (lanes 2, 3) and from B104-1-1 cells incubated during the 3 hr chase with a control IgG2a antibody (lanes 6, 7). In con- trast, essentially no labeled p185 is precipitated from B104-1-1 cells incubated with antibody 7.16.4 during the 3 hr chase (lanes 4, 5). The relative lack of detectable p185 precursor in Figure 4B also demonstrates that significant synthesis of new labeled p185 did not occur during the 3 hr chase in the absence of label. Collectively, these results indicate that down-modulation of cell surface p185 is correlated with lower steady state levels of the p185 pro- tein and with an increased rate of destruction of the p185 molecule. Effect of Monoclonal Antibody 7.16.4 on the Anchorage-Independent Growth of neu-Transformed Cells Neoplastic cells display unusual properties in tissue cul- ture when compared with normal cells (Pollack, 1984). The most definitive in vitro characteristic that distin- guishes tumorigenic cells from nontumorigenic cells is their ability to form anchorage-independent colonies (Freedman and Shin, 1974). NIH 3T3 cells transformed by transfection with activated neu oncogenes (cell line B104- 1-1) grow well in soft agar, with 5% to 10% of the input cells yielding large (>0.5 ram) colonies after 14 days (Figure 5A). In contrast, normal NIH 3T3 cells develop only small clusters of cells, which grow slowly and then stop (Figure 5C). Less than 0.1% of NIH 3T3 cells plated in soft agar give rise to large colonies. The addition of 1 ~g of purified monoclonal 7.16.4 to B104-1-1 soft-agar cultures (0.17 Fg/ml) almost completely inhibits the formation of large colonies by neu oncogene transformants (Figure 5B) and causes the cells to form colonies having a morphology similar to normal NIH 3T3 cells (compare Figures 5D, 6E, 5F). The inhibition by anti- body 7.16.4 of B104-1-1 soft-agar colony formation is dose dependent (Figure 6). As little as 100 ng of antibody in- hibits B104-1-1 colony formation by more than 50%; larger amounts inhibit colony formation by more than 99%.
`
`IMMUNOGEN 2070, pg. 4
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Antibody-Mediated Reversion of Transformed Phenotype 699 12C IOC ~ 8C 6C A i Control 00t 01 10 10 :::>:x~:~:: :k~:~:~:~:~ ~{~ iiiiiiiiiiiii ~!~i~iiiiii iiiiiiii!ii :i.:!,: ,~ .k~::::~:~:~ ~ iiiiiiiiiii'i iii:iiiiiiiiiii i:ili!iiiii::! ~it:i;i~i~iil; iiiii!iiii!i iiiiiiiiiii .... ::i::i" i!i!i~i:ii:i:i !iiiii~!ii~! i~!ii!i!i!il :!:i~i:!:!iii :iild:i::i iiii]i!~!iii iiiiiiiiiiil ii::i:::ili;i ii:ii:i:i: iiiiiiiiiiii ili!iiii!!il ool ol 1.o lo L .pg 7116.4 ~ ~g NEI-026 ~ Figure 6. Dose-Dependent Inhibition of B104-1-1 Soft-Agar Colony Formation by Antibody 7.16.4 Soft agar cultures were prepared as described in Experimental Proce- dures and supplemented with varying amounts of antibody. Each group shows the mean of triplicate samples and the standard error of the mean. 120 100 ~ 80 \ ]_ iiiii!i!iiiiiil Control 0OI 01 10 i ± iii iii iiiil i:#ii!~i:~ ~!iiiiii!ilil iiiii!i!i~il iiiii!i!iiill ii!i!i!i!i!i iiiii iili iliiiiiiiiiii iiiiiiiiiii!iiiii!i!iiii!i iiiiiiiiiiii ii!iiii!i!!ii i!ii!!i!iiii! i!i!i~i!ili~i !iiiiiiiiiil ii!iiiiiiiiii iiiiiiliiii!] iiii!iiii!iii ~!ili!i!ilili iliiiiiii!ili iiiiiiiililil i{i!i!iii!iil !!i!i!i!iiiii Contr~ 0.01 O.i 1.0 iO NIH3T3 -- }104-1-1Cells --~ i XHT-I-~aCelIs ~ Cells (~g7.~.4) (~z16.41 Figure 7. Inhibition of the Anchorage-Independent Growth of neu Transfectant B104-1-1 Cells, but Not H-ras Transfectant XHT-I-la Cells, by Antibody 7.16.4 Soft-agar cultures were prepared as described in Experimental Proce- dures and supplemented with varying amounts of antibody. Each group shows the mean of triplicate samples and the standard error of the mean. To test the possibility that such inhibition results from toxic effects due to the presence of immunoglobulin in the soft agar, or from nonspecific effects of the binding of im- munoglobulin to the cell surface, we examined the effects of a control IgG2a antibody on the anchorage-indepen- dent growth of B104-1-1 cells. As shown in Figure 6, a mouse IgG2a monoclonal antibody that binds B104-1-1 cells via cell surface/]2-microgiobulin molecules does not inhibit the anchorage-independent growth of B104-1-1 cells. Thus, neither the presence of IgG2a protein in the agar layer, nor its binding to the cell surface, is sufficient to inhibit the anchorage-independent growth of these cells. The anchorage-independent growth of an indepen- dently derived neu oncogene transfectant, cell line B104- 1-2, is inhibited by antibody 7.16.4 to approximately the same degree as is cell line B104-1-1 (data not shown). In contrast antibody 7.16.4 has no effect on anchorage- independent growth of the H-ras-transfected NIH 3T3 line XHT-I-la, even at a 100-fold higher concentration (10/~g) than that which inhibits B104-1-1 cell anchorage-indepen- dent growth by more than 50% (Figure 7). We conclude that the ability of antibody 7.16.4 to inhibit neu-transformed NIH 3T3 soft agar colony formation is a function of the p185-specific antibody itself, and not of a toxic con- taminant that might copurify with it. Although repeated feeding of B104-1-1 soft agar cul- tures with antibody will continue to suppress colony for- mation, B104-1-1 cell cultures that are inhibited by anti- body exposure for 2 weeks will eventually develop large colonies if fed with antibody-free media (data not shown). This suggests that the antibody exerts a cytostatic effect, rather than an irreversible cytotoxic effect, on neu- transformed cells in soft agar. Furthermore, the antibody does not affect adherent growth of B104-1-1 cells in medium containing 10% fetal calf serum (data not shown). All experiments were performed using serum depleted of complement by heating at 56°C, and we have been unable to demonstrate lysis of B104-1-1 cells using antibody 7.16.4, or antibody plus complement, in standard 51Cr re- lease microcytotoxicity assays (data not shown). These results indicate that the effect of anti-p185 antibody on the anchorage-independent growth of B104-1-1 cells does not involve antibody-mediated cell killing. Inhibition of Anchorage-Independent Growth Requires Cross-linking of I)185 Molecules by Monoclonal Antibody The inhibition of anchorage-independent growth by anti- body 7.16.4 could result simply from its binding to cell sur- face p185, this affecting in turn the activity of the p185 mol- ecule. Thus the antibody might block a receptor site for some critical growth factor, or might induce a conforma- tional change in the p185 molecule that renders it unable to impart oncogenic signals to the cell. Alternatively, the ability of the antibody to inhibit anchorage-independent growth might stem from its ability to induce down- modulation of p185. The removal of p185 from its normal site of residence in the plasma membrane and the lower- ing of cellular p185 levels by antibody might deprive the neu-transfected cells of the protein that makes possible their anchorage-independent growth. These two alternatives can be distinguished by compar- ing the effects of intact divalent antibody and monovalent F(ab) antibody fragments on the anchorage-independent growth of neu-transformed cells. The intact antibody and
`
`IMMUNOGEN 2070, pg. 5
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Cell 700 \ 3C % 20 iii!i!ii~!ili!ii! iiiiii~iiii!iiiilill • • ~!!i!i!!~i!!!.:,,~ " iiiiiil i~:iiiiiiii~ii:i !:!:::!,:i~:::: ~ - i!ii!i!!i!!ili!i i iiiiiiiiiiii!iiiii -iiiii!!iiii! I ili~iii~ilili~ili i!ii!iiiii!!!i i :iiiiiiii:ii::ii: iiiiiiiiiiiiiiiiiiiiil Control 10pg lug F{ob) 7164 X164 + lOpg F(cb) ii:iiil:iii::ii: ii!iiiiiii!!iiiiiiii! ii iiiii !!ili!! ii iiiiiiii:ilili ili ililiiii ! !iii! Anti Anti- Mouse Mouse IgG Ig+G lOng F(ab) Figure 8. Inhibition of B104-1-1 Anchorage-Independent Growth Re- quires Cross-linking of p185 by Monoclonal Antibody B104-1-1 cells were cultured in soft agar as described in Experimental Procedures, with the antibody mixtures shown. The anti-mouse IgG consisted of 12.5 ~l/dish of heat-inactivated goat anti-mouse IgG reac- tive with both heavy and light chains (Miles). Each group shows the mean of quadruplicate samples and the standard error of the mean. _100 18o 88 I2o i B104-I-t + [] Unlreoted Conlrol Cultures •+lpgz164 o CELL LINE: BIH-3 B/H 4 B/H 5 B/H 7 HaSV3T3 p185: + + + + p2f" - + + + + + Figure 9. Antibody 7.16.4 Does Not Inhibit the Anchorage-lndepen- dent Growth of Cells That Contain Both Activated neu and Activated ras Oncogenes Soft-agar cultures were performed as described in Experimental Procedures. Because different cell lines have different abilities to form anchorage-independent colonies, all results have been normalized. The untreated colony-forming ability of a particular cell line is repre- sented as 100% (stippled bars), and the number of colonies >0.5 mm formed in the presence of 1 /~g/dish antibody 7.16.4 is represented as the percentage relative to the untreated controls (solid bars). Results show the mean of quadruplicate samples and the standard error of the mean (also expressed as a percentage of the appropriate untreated control group). its F(ab) fragments compete for binding to the same site on the p185 molecule with less than a 10-fold difference in affinity (data not shown) and should be roughly equiva- lent in their ability to block access to critical sites of the cell surface p185 molecule. The equivalence of intact anti- body and F(ab) fragments in blocking ligand binding has been demonstrated utilizing anti-insulin receptor and anti- EGF receptor antibodies that directly block their respec- tive ligand binding sites (Kahn et al., 1978; Gill et al., 1984). Thus both intact antibody and F(ab) fragments should in- hibit anchorage-independent growth if this growth inhibi- tion results simply from immunoglobulin binding to cell surface p185. If the inhibition of anchorage-independent growth results from p185 cross-linking and down-modula- tion, however, we would expect that only intact antibody should affect soft-agar colony formation, since F(ab) frag- ments do not cause p185 down-modulation as described previously. Figure 8 presents evidence that cross-linking of cell sur- face p185 molecules is necessary to inhibit anchorage- independent growth. Monovalent F(ab) fragments have no direct effect on the anchorage-independent growth of B104-1-1 cells, and can even partially antagonize the abil- ity of intact antibody to inhibit soft-agar colony formation (Figure 8), presumably by preventing p185 cross-linking by the divalent antibody. However, if the monovalent F(ab) fragments are cross-linked by the addition of a second an- tibody (goat anti-mouse immunoglobulin) to soft-agar cul- tures, then they are able to inhibit soft-agar colony forma- tion (Figure 8). This demonstrates that muttivalent binding of antibody to cell surface p185 is required to inhibit soft- agar colony formation by neu-transformed cells. Because the requirements for antibody-mediated anchorage-inde- pendent growth inhibition parallel those for antibody-in- duced down-modulation of the p185 protein, we suggest that down-modulation of the p185 molecule is responsible for the loss of anchorage-independent growth observed when cells are grown in the presence of antibody 7.16.4. Effect of Antibody on Cells Containing Both Activated ras and Activated neu Oncogenes The data presented above suggest that down-modulation of the neu oncogene product, p185, is responsible for the loss of anchorage-independent growth when neu- transformed cells are cultured in the presence of anti-p185 monoclonal antibodies. In order to exclude any potential cytopathic effects of the antibody on p185-bearing cells, we examined its effect on the anchorage-independent growth of doubly transformed cells containing both neu and ras oncogenes. We reasoned that if the action of the antibody were exclusively to deprive the cell of the onco- genic signal provided by p185, the antibody should not af- fect the anchorage-independent growth of cells that ex- press both p185 and a second unrelated oncogene product that can itself induce anchorage-independent growth. To generate such doubly transformed cells, we in- fected the neu-transformed NIH 3T3 cell line B104-1-1 with Harvey sarcoma virus at a high multiplicity of infection (40 focus forming units per cell) and then cloned the cells at limiting dilution. Using this approach we obtained a num- ber of clonal cell lines (which we term B/H lines) that ex- press high levels both of cell surface p185 and of the intra-
`
`IMMUNOGEN 2070, pg. 6
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Antibody-Mediated Reversion of Transformed Phenotype 701 Antibody/Dish 0.01 pg Colonies >0.5 mm +_ S.E.M. 38.3+_1.2 2.5±05 0.1 ug 0.7 +_ 0.3 1 JJg 0.3£0.3 lOlJg O0 Figure 10. Inhibition of the Anchorage-Independent Growth of Rat Neuroblastoma Cell Line B104 by Antibody 7.16.4 Soft-agar cultures were performed as described in Experimental Procedures; because of slower growth of the B104 cell line in soft agar, colonies were counted after 21 days. A typical plate and the mean of triplicate samples plus or minus the standard error of the mean are shown for each group. cellular p21 protein encoded by the viral Ha-ras oncogene (data not shown). The presence of Harvey sarcoma virus in these cells was further indicated by the fact that they release infectious sarcoma virus into the culture superna- tant with a titer higher than 103 focus forming units/ml (data not shown). Each of these B/H cell lines thus con- tains both an activated neu oncogene and an activated ras oncogene. As shown in Figure 9, the anchorage-independent growth of four independent cell lines containing neu and ras oncogenes is completely resistant to the effects of anti- p185 antibody, as is the growth of an NIH 3T3 cell line transformed by Harvey sarcoma virus alone. In contrast, soft agar colony formation by the neu-transfected cell line B104-1-1 is highly sensitive to the presence of anti-p185 antibody, confirming earlier results. It appears that anti- p185 antibodies can inhibit anchorage-independent growth only when the p185 molecule is the exclusive cellu- lar source of an oncogenic signal. It should be stressed that p185 expression and sensitivity to antibody-induced modulation of these B/H lines do not differ significantly from that of the B104-1-1 cell line (data not shown). Thus these cells bind monoclonal anti-p185 antibody and as a result down-modulate their p185 in a manner analogous to the B104-1-1 cell line, but their anchorage-independent growth is not affected. This underscores the fact that the monoclonal antibody does not exert toxic effects on cells that express p185 on their surface; rather, it deprives them of a component necessary for anchorage-independent growth. Effect of Anti-p185 Antibody on the Anchorage- Independent Growth of the Rat Neuroblastoma from Which the neu Oncogene Was Originally Isolated An unresolved issue in the study of cellular oncogenes de- tected in transfection assays is whether these genes are involved in the creation of the malignant state of the can- cer cells in which they initially arise. Studies utilizing temperature-sensitive mutants of retroviruses have con- vincingly demonstrated that the retroviral oncogenes are responsible for both initiation and maintenance of the transformed state of retrovirus-induced tumors (Sefton et al., 1980; Bishop, 1983). However, in the case of tumors of nonviral etiology, the possibility remains that cellular proto-oncogene activation reflects a nonessential genetic heterogeneity arising late in tumor progression, and is not a critical early step in neoplastic transformation (Albino et al., 1984). If neu oncogenes were directly involved in the transformed state of the rat neuroblastomas in which they arose, we would expect anchorage-independent growth of these neuroblastomas to be inhibited in the presence of anti-p185 monoclonal antibodies. Figure 10 demonstrates that colony formation by the DNA donor rat neuroblastoma (cell line B104) is strikingly inhibited by the a