CA2145382A1 - Methods and compositions for cancer therapy and for prognosticating responses to cancer therapy - Google Patents

Abstract

A method for determining the efficacy of a therapeutic agent, in vitro, for a cancer expressing or overexpressing an onco-gene product is described. The method is particularly useful for determining the efficacy of therapeutic agents that have a binding affinity for cancer that express HER-2/neu. N24, N28 and N29 monoclonal antibodies are described which have been identified by this method. One or more of these antibodies can be used as a therapeutic agent in the treatment of breast, stomach, ovarian or salivary cancers.

Description

WO 93/03~1 PCl~/US92~07117 L~l~OI~ AND CO. ~ SITIONS FOR CAN OE R THERAPY AND FOR
PROGNOSTICATING K~O~S TO CAN OE R THERAPY
-Tecnnical Field -This invention relates generally to methods for selecting putative anti-cancer agents and for determ;ning the efficacy of such agents useful in the treatment of a cancer characterized by expression of a surface oncogene product. This invention further relates generally to compositions selected by such methods.

Back~round of the Invention This is a division of Application Serial No. 2,096,417 filed August 21, 1992.

The transformation of a normal cell into a malignant cell characteristically results, among other things, in the uncontrolled proliferation of the progeny cells, which exhibit immature, undifferentiated morphology, and expression or overexpression of oncogenes not normally expressed by normal, mature cells. It is the goal of cancer therapy to selectively ~ill or inhibit the uncontrolled growth of such malignant cells, while not adversely effecting normal cells.

Traditional chemotherapeutic agents are highly cytotoxic agents which preferably have greater affinity for malignant cells than normal cells or at least preferentially effect malignant cells based on their high rate of metabolic activity. Where an oncogene product unique to a malignant cell is expressed or overexpressed on its surface membrane, it may be used to target such malignant cells for destruction using chemothertpeutic agents designed to specifically interact with the oncogene product. Extremely precise methods of targeting malignant cells for destruction have become available with the advent of cytotoxic conjugates, consisting of a potent cytotoxin chemically lin~ed to an affinity molecule, such as a monoclonal antibody, having specificity for a unique protein produced by a malignant cell, such as a cell surface antigen.
Using immunocytochemical and molecular analyses, it i8 possible to precisely identify the composition and structure of an SUBSrlTUTE S~EET

` 21~5382 W093/03. . ~ PCT/USs2/07117 oncogenic protein and produce a monoclonal antibody which has the capacity to specifically bind the oncogenic protein, and thus, increase the accuracy of delivering the cytotoxin to the intended target cell.

Besides cytotoxic conjugates, it ha6 been proposed to use monoclonal antibodies which 6pecifically bind to the surface of a cancer cell. Anti-tumor effects of monoclonal antibodies may be achieved through the effector function of the antibody molecule through natural immunological respon6e to the antigen-antibody complex. In this respect, certain monoclonal antibodies have been fihown to result in a reduction of tumor size. Undesirably, however, other monoclonal antibodies which specifically bind to 6uch antigens on the surface of the malignant cell have no effect or, worse, actually accelerate the growth of the malignancy, even though such antibodies are 6pecific for the same malignant cell type and the same oncogene product as the antibodies that reduce tumor size. In view of the unpredictability of the effect, if any, of an antibody on malignant cells, it ha6 not been possible to determine, prior to starting therapy, whether one or more 6elected antibodies would react as anti-tumor agents or provide an accurate prognosis. Heretofore, it has not been possible to determine which antibody preparations, of a 6election of monoclonal ant;bo~ies (each of which is capable of specifically bin~jng an ~ GyO~iC protein) are tumor antagonists, and which are tumor agoni~ts that may undesirably accelerate proliferation of the malignancy. It would be de6irable to be able to deter~ine in an in vitro assay method which antibody preparation (or combination of antibodies) having specific affinity for an o--cGgol,e product, and how much thereof, would be predicted to inhibit the proliferation of malignant cell6 and provide a good prognosis for the patient. It would be de6irable to provide an in vitro method for ~LGyuo~icating the efficacy of a pLu~osed SUBSTITUTE SHEET

.~ ` 21g5382 W093/o~,.. PCTJUS92/07117 therapeutic agent (or combination of agents) and dosage thereof, which method is time-and cost-effective, as well as minimally traumatic to a cancer patient, so that the method may be practically employed in the great variety of cancer cases to be found among different patients.

Summary of The Invention We have discovered that in canceræ characterized by the presence of malignant cells which express or overexpress one or more membrane-associated, receptor-like oncogene proteins, malignant cells can be induced to terminally differentiate by ~m; n; stering an effective amount of a composition comprising an affinity molecule, such as a monoclonal antibody which is specific for an epitope on the extracellular ~n~?;n of the oncogene protein, and/or a ligand which is specific for the oncogene protein. In preferred embodiments of the present invention, the malignancy is one that is characterized by the expression or overexpression of at least the HER-2/neu oncogene. Among the cancers which characteristically express or overexpress HER-2/neu are certain breast, stomach, ovarian and salivary gland cancers.

Thus, a method of the present invention entails a method.
for determining/prognosticating the effectiveness of a therapeutic agent in the treatment of a cancer wherein malignant cells of the cancer express or ~e~e~Less an oncogene product, the method comprising the step of: (a) obt~;n;ng viable malignant cells which e~,ess or overexpress at least one on~ogcne product and dividing the same into at least first and secon~ portions; (b) treating the first portion comprising viable malignant cells with a sufficient quantity of a compo6ition comprising at least one compound having specific bin~;ng affinity for the oncogene product and contacting the SUBSTITUTE SHEET

21 15382~
wos3/o3 ~ - PCT/US92/07117 second portion with a eomposition which is devoid of the compound or cG...~o~,ds having specific binding affinity for the oncogene produet and ;nc1lh~ting the first and second portions in a physiologically acceptable medium for an amount of time sufficient to in~ce a percentage of the viable malignant cells of said first portion to terr;n~lly differentiate; and (c) comparing the percentage of cells in the first-portion which exhibit morphological evidence of said terminal differentiation to the percentage of cells in the 6econd portion which eAhibit morphological evidence of terminal differentiation, or, alternatively, comparing the average value across the first portion of one or more parameters indicative of ter~i~al differentiation with the average value of the same parameter(s~
aeross the second portion. The viable malignant cells may be obtained as a tissue biopsy, serum sample or other cell containing sample from a patient suffering from a malignancy.
In which case a therapeutic agent tailored to the patient may be sel-ected. Alternatively, the malignant cells may be those of an established transformed cell line derived from a malignant tissue, in which case the method of the present invention may be used as a general screening assay for selecting anti-cancer therapeutic agents effective against 6uch malignancy.

In accordance with certain aspects of the present invention, induction of ter~;n~l cell differentiation in malignant cells ~A~,essing or overeApressing HER-2/neu can be ~hown by an increased ~e.cenLage of treated cells which eApres6 a mature phenotype. For example, in the case of breast cancer, induction of differentiation in accordance with the present method may be determined by the prese~ce of milk comFone~t~
~uch as casein and lipid droplets in the treated cells. In accordance with other aspects of the pre6ent invention, induction of terminal differentiation in malignant cell6 SUBSTITUTE SHEET

- 21~5382 W093/03~ PCT/US92/07117 expressing or overexpressing HER2/neu can be shown by an increased percentage of cells that express ICAM-l (designated by the International Workshop on Human Leukocyte Differentiation Antigens as CD54) and/or E-cadherin (an 80kD
protein described in Wheelock et al., J. Cell. Biochem., 34:187-202 (1987), also known as "CAM 120/80"), and/or an increase in total nuclear area.
It has been found that a sample comprising malignant cells which' express or overexpress HER-2/neu, when treated with an affinity molecule which has specific bin~;ng affinity for the extracellular ~om-; n of the HER-2/neu product, results in ter~;nal cell differentiation and that this differentiation is correlated- to translocation of the HER-2/neu product from the surface membrane of a malignant cell to the cytoplasm or perinuclear region of the cell, and to a transient increase in the overall HER-2/neu content of the cell after which translocation the cells ceases to proliferate at rates characteristic of malignant cells. Thus, a monoclonal antibody preparation useful for the treatment of a malignancy characterized by HER-2/neu expression (or overexpression) can be selected based on its ability in a method of the invention to indl~ce in such malignant cells, translocation of HER-2/neu protein or the expression of other mature cell phenotypes as discussed below.

Additionally, we have found that, in at least some cancers characterized by the expression or overexpression of a membrane-associated, Lece~or-like oncogcne protein, contacting such malignant cells with a ligand specific for the membrane-associated protein results in the induction of terminal cell differentiation and conse~ently the appearance in such cells of mature pheno~y~e. In preferred aspects of the invention, the malignant cells express or overexpress the -SUBSTITUTE SHEET

2145382 ~
WOg3/Q ~ PCT/US92/07117 HER-2/neu product and ligands specific for the product are, for example glycoprotein gp30 (Lupu et al , Science, 249 1552-1555 (1990)), and neu differentiation factor (or "NDF" as described in Wen et al , Cell, 69 559-72 (1992)) Thus, one aspect of the present invention entails methods for selecting anti-cancer therapeutic agents, particularly monoclonal antibodies and ligands, and prognosticating their in vivo response to c~nrer therapy A detectable increase in terminal cell differentiation in malignant cells (e q , from a biopsy treated according to the method of the present invention) represents potential effectiveness of the composition in cancer therapy and provides a prognostic measure of the potential effectiveness of the therapy in vivo In another aspect of the present invention, monoclonal antibodies have been selected by application of the above-described method These antibodies also (1) are specific to the extracellular portion of the human HER2/neu product, (2) are capable of immunoprecipitating a single protein of 185 kD
from metabolically labeled HER2 cells, (3) do not react with human epidermal growth factor receptor ("EGFR") or with rat, p 185neu protein, and ~4) significantly inhibit the tumorigenic growth of HER2 cells in mice are described The antibodies N12, N24, and N29 have these properties and were described in Bacus et al , Cancer Res , 52 2580-89 (1992) N28, al80 describe~ by Bacus et al has the opposite effect on such cells N24, N28, N29 and N12 were deposited with the Collection Nationale de Cultures de ~icrorganismes, Institute Pasteur, 25 Rue du Docteur Roux, Paris, France, under the terms of Rule6 28 and 28a of the Eu~o~ean Patent Co..ven~ion as accession numbers I-1260, I-1261, I-1262 and I-1263 .a~e~ively on August 19, 1992 Ihese ant~ho~es, fragments or chimeric/humanized ~ersions thereof, can be used alone (or in combination with each other), and/or lin~ed to toYins to SUBSTITUTE SHEET

WOg3/0~,., PCT/US92/07117 form cytotoxic conjugates any and all of which can be used as therapeutic agents. In addition, these antibodies are useful in the above described prognostic methods.

Brief Description of the Drawinqs FIG. 1 depicts the binding of anti-HER2/neu monoclonal antibodies to HER2 cells.
FIG. 2 illustrates ;~-~moprecipitation of the HER2/neu protein by various monoclonal antibodies. Panel A illustrates immunoprecipitation with HER2 cells labeled with ~35S]methionine and Panel B a Kinase assay.
FIG. 3 illustrates the effect of various monoclonal antibodies on tumor growth in athymic mice. Panel A shows the effects of antibody treatment after 21 days post-inoculation. Panel B
illustrates the ~inetics of tumor growth in antibody-;nAuce~
athymic mice.
FIG. 4 depicts antibody-induced stimulation of tyrosine phosphorylation of the HER2/neu product. The autoradiograms of the SDS-gel separated proteins obt~;ne~ in two different tyrosine phosphorylation assays are shown. In Panel A HER2 cells labeled with 132P~orthophosphate were incubated with each antibody, and subjected to two consecutive immunoprecipitation steps with anti-phosphotyrosine and anti-HER2/neu ant;ho~ies. In Panel B, SKBR3 cells were first i~cllh~ted with various monoclonal antibodies and then subjected to two immunoprecipitation consecutive steps, followed by autophosphorylation with gamma ~32P]ATP and Mn2 .
FIG. S illustrates the effect of various monoclonal antibo~ies on the rate of turnover of the HER2/neu protein.
FIG. 6 depicts retardation of tumor growth by conjugates of antibodies N24 and N29 with ricin A.
FIG. 7 shows a graph of AU-565 cell numbers per square centimeter after six days of treatment with various concentrations of ligand gp30.

SUBSTITUTE SHEET

2145382) , ~ WO93/03l~l PCT/US92/07117 .

FIG. 8 shows a graph of total cellular HER-2/neu content in AU-565 cells over time as quantified by immunohistochemical stain optical density analysis. Open triangles represent a control sample, while open circles represent a sample treated with 6 ng/m~ of gp30 for 8iX days.
FIG. 9 shows the effect of ligand NDF over time on AU-565 cells as measured by cell number and nuclear area (A) and casein an lipid content (B), and on MDA-MB 453 cells as measured by cell number (C).

Description of Preferred Embodiments In accordance with the pre6ent invention, it has been discovered that certain affinity molecules which are capable of specifically b;nding to the extracellular domain of receptor-like o-co~one products, particularly the HER-2/neu oncogene product, have the capacity to induce malignant cells which e~Less or overexpres6 that product to terr;n~lly differentiate and cease unregulated proliferation. The n~K-2/neu on~ogene i8 a member of the erbB-2 oncogene family.
Administering 6uch affinity molecules to a patient suffering from a malignancy characterized by expression or overexpression of 6uch a product can be used therapeutically, alone or in conjunction with other therapies, to treat patients suffering from such a malignant disease. In accordance with the present invention as it pertains to in vitro assay methods for 6electing or determining the efficacy of an affinity molecule which is capable of cau6ing induction of terminal differentiation in a malignant cells of a patient suffering from the malignancy, a cell sample of a c~ncerous ti6sue having 6uch malignant cell6 which ~A~-e~6 or v~ê-~Le~s an oncogene product i6 obta~ned.

SUBSTITUTE SHEET
. .

21~5382 W093/03,.~ ~PCT/US92/07117 Preferably, the cell sample is a biopsy, and is suitably sized so that it may be divided into a plurality of portions for testing with one or more putative agents, at one or more concentrations. While the cell sample may be maintained for up to several days in a suitable maintenance medium, it is preferred to employ a cell sample within about 24 hours or less from the time it is excised.

The cell sample is divided into a first and second portion (each portion then may be further divided into a suitable number of representative aliquots), and the portions placed in individual sterile culture vessels (e.q., separate wells of a microtiter plate). The number of aliquots in a portion that are employed in an assay will be determined by the number of compounds and concentration thereof which are tested. Also, for tissue biopsies it is contemplated, to mince or otherwise disperse the cells so they can be cultured, to provide a suitable number of culture vessels having viable malignant cells grown from the biopsy as primary cultures. In this way, the number of malignant cells obta;nAhle for u~e in an assay method of the invention may be multiplied. It is preferred to have at least one aliquot of the biopsied tissue (or cells thereof) as a negative control (or second portion) which is not contacted with putative anti-cancer compound(s) so that the~
percentage of cells exhibiting evidence of term;nAl cell differentiation in the absence of the putative compounds(s) can be determ;ne~.

In accordance with the present invention, a monoclonal antibody or ligand (or a combination of these affinity molecules) can be added to the cultured biopsy after seeding.
It is preferred to allow the cells to acclimate to culture conditions for about one day after see~ing and then add the putative agent(s) to the re~ec~ive cultures in ~..0~.~(8) SUBSTITUTE SHEET

-- WOg3/Q 2 1 4 5 3 8 2- ~CT/US1V~71~7 sufficient to give a predetermined concentration of the agent.
Alternatively, a series of tubes of culture media, each of which is supplemented with a predetermined amount of one or more putative agents, can be used to seed the cells directly into the culture vessel.

The aliquots then are incubated for a period of time sufficient to cause induction of ter~in~l cell differentiation in at least a portion of the malignant cell~. Generally, a statistically significant percentage of cells (as compared to a negative control) exhibit evidence of terminal cell differentiation within about one to about seven days of i~ct~h~tion in the presen~e of a compound which has the capacity to inA~ce differentiation. Converl~ional incubation conditions for human and other mammalian cells are well known in the art.
Suitable incubation conditions include an incubation temperature of about 20 - 45C, more preferably about 37C, and a humidified atmosphere of air supplemented with about 5% -10~ CO2. Where inCl~hation times employed in the assay methods of the invention ~ aed about three or more days, it may be desirable to e~change spent culture medium in the rebpec~ive vessels for fresh culture medium, preferably supplemented with the same concentration of the putative agent.

While it is preferred to tailor the selection of affinity molecules for use as anti-cancer agents to individual patients by employing a cell sample from such patient in an assay method of the invention, the present invention also includes 6creening methods for determining the efficacy of affinity molecules such as monoclonal antibodies or ligands having 6pecificity for the HER-2/neu wherein cells of a transformed cell lines are used instead of biopsied tissue, for example. E~amples III and IV
below describe induction of terminal cell differentiation by ~nc~h~ating cells of well-~no~n, readily obt~nahle SUBSTITUTE ~

~` 2145382 - W093/03, PCT/US92/07117 .

transformed cell lines with monoclonal antibody preparations which are specific for a portion of the extracellular ~o~-;n of the HER-2/neu product.

Monoclonal antibodies which have specific binding affinity for certain regions on the extracellular dom~n of the B R-2/neu product are one type of affinity molecules which are capable of inducing malignant cells expressing or overexpressing HER-2/neu to undergo ter~inal cell differentiation. Importantly, it is a necessary, but not sufficient, condition that a monoclonal antibody be specific for an epitope on the extracellular domain of the HER-2/neu product. In other words, not all monoclonal antibodies which are able to specifically bind a region of the extracellular domain of HER-2/neu are able to induce differentiation. Some monoclonal antibodies that meet this first criterion have no effect or, worse, may have an agonistic effect on the proliferation of such malignant cells expressing HER-2/neu, such that their ~;ni6tration in vivo may undesirably promote growth of the malignancy. Also, a monoclonal antibody which is capable of inducing differentiation may have such an effect in one range of concentrations, but have an opposite, agonistic effect, at a different (i.e., higher or lower) concentration.
Thus, the present invention provides a method for deter~ining a preferred range of dosages of a therapeutic agent to be used in therapy.

Monoclonal antibodies that are capable of reacting with the HER-2/neu product are known in the art. Methods of making monoclonal antibodies generally also are well known in the art. See, generally, Harlow & Lane, Antibodies - A Laboratory Manual, Ch.s 5-6, Cold Spring Harbor (1988). With re~-pe~ to producing monoclonal ant~ho~es which are specific for the extracellular domain of HER-2/neu, briefly, an animal capable of producing an immune response to the antigen (e.q., HER-2/neu product) is injected with the antigen in a manner which will sue~ SHEET

-` wog3/o 2 1 4 5 3 8 2 ~usg2~0~ll7 result in an immune response. The antigen may be HER-2/neu product which has been isolated rom malignant cells which produce the protein, or the antigen may be produced by recombinant expression of the HER-2/neu gene (or a portion thereof which ~nco~es at least a portion of the extracellular domain) transformed or transfected as known in the art into in a suitable bacterial, yeast or mammalian host cell for the production of recombinant HER-2/neu product (or protein fragment thereo). Monoclonal antibodies may be produced from mouse lymphocytes by inje¢ting a mouse with a natural or synthetic protein (or part of a protein) or cell membranes derived from whole cell~. The immunized ~ni~-l naturally develops an immune response to the antigen and produces spleen cells which produce antiho~ies to various epitopes of the antigen, which then are fused with myeloma cells to form hybridomas. Clones with the desired antibody specificity are selected by their ability to (1) bind specifically to the extracellular domain of the ~ER-2/neu product and (2) induce terrin~l cell differentiation in viable malignant cells which express or over~ ess HER-2/neu. 8elected antibody-producing cell lines are e-lJ~ e~ by cor.ve~ional tissue culture techniques and monoclonal antihoAies may be routinely purified from the culture medium. Monoclonal antibodies which fulfill criterion (1) and (2) above appear to be able to mimic the action of a ligand for the HER-2/neu product. Chimeric or humanized forms of these antibodies are desirable for in vivo use. ~uch antibodies can be made in accordance with well known methods one of which i8 de~cribed in U.~. Patent No. 4,816,397 which is incorporated by reference.

It also has been found, surpri~ingly, that ligands for the HER-2/neu product are affinity molecules which are capable of inducing malignant cells ~A~es~ing or overexpressing HER-2/neu to undergo terminal cell differentiation. Examples of ~uch ligands include gp30 and NDF.

SUa~TlT~'TE S~ET

~ WOg3/0~,~ 2 1 ~ 5 3 8 2 PCT/USg2/07117 After treatment of the portions, the portions are analyzed for indicia of induced terminal differentiation.
Phenotypically, induced differentiation is evidenced by maturation markers including inhibition of cell growth, altered cytoplasmic and nuclear morphology, increased expression of cell adhesion markers (such as ICAM-l and/or E-cadherin) and, in malignant breast cells, enlargement of the nuclear size and synthesis of milk cn~ron~nts such as casein and lipids.
Surprisingly, it has been found that concomitantly with one or more of these mature phenotypic changes, the HER-2/neu protein translocates (or migrates) from the membrane to the cytoplasm and/or perinuclear regions of the cell, and that this translocation is additionally associated with a transient increase in total cellular HER-2/neu content. Translocation and a transient increase in total cellular HER-2/neu content serves as one indicator of terminal cell differentiation.

In particularly preferred emboAi~nts of the present invention, the response to antibody or ligand therapy in a patient having breast cancer or ovarian cancer is prognosticated by contacting a biopsied sample from said cancerous tissue with the monoclonal antibody or ligand selected for therapy for a predetermined time and deter~;n;ng, by immunohistochemical sta;ning techniques translocation of the HER-2/neu product from the cell membrane to the cytoplasm or perinuclear region of said cAncerous cell (or a transient increase in total HER-2/neu content), an increase in nuclear area and/or an increase in ICAM-l (and/or E-cA~herin) expression. StA;n~ samples may be analyzed for optical density ~alues which correspond to the &..O~u~S of stained cell constituents. Translocation may be deter~i nç~ by (1) a reduction of HER-2/neu in the surface, (2) an increase in HER-2/neu in the cytoplasm or perinuclear region, (3) a transient increase in the total HER-2/neu content, or a any SO~ I I I IJTE SHET

W093/037~ ~ 2 1 4 5 3 8 2 ~ PCT/USg~07117 combination of (1), (2) and (3). Nuclear area and expression of ICAM-l or E-cadherin may be measured by similar immunohistochemical techniques. The malignant cells treated in accordance with a method of the invention, in the presence of absence of a putative anti-cancer agent, then are examined to deterrine the percentage of cells which have been in~l~ce~ to differentiate.

This can be deter~;neA by comparing the percentage of treated cells containing HER-2/neu predominantly in the cytopla6m and/or perinuclear region as compared to the percentage of cells in a negative control showing such a distribution of HER-2/neu product. A decrease of HER-2/neu product in the 6urface membrane of the treated cells, alone, or in combination with an increase in the cytoplasm or perinuclear region or an increase in the total HER-2/neu content (as compared to untreated cells) can be used to indicate induction of ter~inal differentiation.

Preferably, the average amount of membrane-bound HER-2/neu per cell in the control population can be used as a test value in obtainjng cell percentages. The average is calculated rom a statistically 6ignificant number of cells in the control group. Then, the amount of membrane-bound HER-2/neu in individual control group cells i6 compared to the average, to determine what percentage of the population has a lower amount of membrane-bound HER-2/neu, and what percentage has a higher amount. Cells from the treated group are similarly examined to determine what percentage of cell6 evidence le6s membrane-bound HER-2/neu than the control group average, and what percentage evidence greater membrane-bound HER-2/neu than the control group average. Finally, a comparison can be made ~a~WeEn the ~a~anLage6 obt~ine~ for the control group, and the ~eLce~Lages obt~jne~ for the treated group. A statistically significant SU~SI~IUl~ SHET

~ ~ ` 21 ~5382 ~_ ~ W093/03. PCT/US92/07117 , -lS-increase in the percentage of cells in the treated group over the percentage of cells in the control group which have less membrane-bound HER-2/neu than the control group average indicates translocation of HER-2/neu. ~The same approach is followed for deter~;n;ng changes in nuclear area and cell adhesion molecule expression.) The amount of cytoplasmic HER-2/neu also can be eY~ined instead of the amount of membrane-bound HER-2/neu to obtain cell percentages as described above. A statistically significant increase in the percentage of cells in the treated group over the percentage of cells in the control group which have more cytoplasmic HER-2/neu than the control group average indicates translocation of HER-2/neu.

The total amount of cellular HER-2/neu also can be ~m;ned instead of membrane-bound cytoplasmic HER-2/neu to obtain cell percentages as described above. A statistically significant increase in the percentage of cells in the treated group over the percentage of cells in the control group which have more total cellular HER-2/neu than the control group average indicates translocation of HER-2/neu.
.

In an alternative embodiment, the average amount of HER-2/neu found in a sample of treated cells (by e~m;n~tion of optical density values after stAining) can be co~r~red to the average amount of HER-2/neu found in a sample of control cells to determine translocation. The amount compared may be only that which is membrane bo~ud, in which case a statistically significant decrease in stAining in the treated sample indicates translocation. Alternatively, the amount compared may be only cytoplasmic, or may be the total cellular content, in which cases any statistically significant increase in stAin;ng in the treated sample indicates translocation.

- SUBSTITUTE SHET

WO 93/037- 21 ~ 5 3 8 2 ` ~/US92/07117 -- The location and distribution of a cellular component, such as HER-2/neu protein, cel} adhesion molecule(63, or casein or lipid droplet6 can be determined imm~noh;stochemically. The cells of the biop6ied sample may be fixed in a fixative, such as paraormaldehyde, followed by treatment with an organic 601vent, such as acetone, formalin, or methanol, 60 as to render the cells permeable for immunohistological 6t~jnin~.
Methods of fixation are well within the skill of the art. 8ee, e.q., Bacus et al., Molec. Carcin., 3:350-62 (1990).

Where the presence and di6tribution of HER-2/neu and/or cell adhesion molecule(s) are to be determined, cells can be stained with an antibody 6pecific for the HER-2/neu product and/or cell adhesion molecule conjugated to a fluorescent dye, 6uch as fluorescein, rhodamine and the like. Where two or more different antibodies are conjugated to fluorescent dyes it i6 appropriate to conjugate each antibody to a fluorescent dye that fluoresces at di6tingui6hable wavelength6. The location and distribution of HER-2/neu and/or cell adhesion molecule(s) in the cell6 can be determined col-va~-ionally by fluorescence microscopy, and, optionally, confirmed by confocal microscopy.
Besides direct immunofluorescPnce st~ininq~ indirect antibody 6taining procedures which detect the presence of 6pecific antigen-antibody complexes, such a6 peroxidase-anti-peroxidase 6t~;ning proced~res or alkaline pho6phata6e 6t~inin~, may be u6ed to determine the distribution of HER-2/neu and/or cell adhe6ion molecule(s) in 6uch fixed cell6.

Mature phenotype ~Le6sion al60 can be u6ed to determine the extent of terminal cell differentiation in the fir6t portion of biop6y. For example, immature cPncerou6 human breast cells and mature cell6 (e.q., malignant cell6 which were ~ndu~e~ to differentiate) can be di6tin~6hs~ by the ability of the mature cells, but not the maliqnant cells, to produce human milk component6, includinq ca6ein and lipid6. The t SHET

W093/03~/~. PCT/~US92/07117 percentage of cells which have been caused to differentiate in a method of the invention may be determined by the presence of such milk components. Casein can be detected by known immr mohistochemical staining using anti-casein antibodies.
The presence of lipids may be detected by staining with a dye compound suitable for such detection, such as Oil Red O. See, e.q., Bacus _ al., Molec. Carcin., 3:350-62 (l990).

After stAin;ng, the location of the HER-2/neu protein, for example, can be determined and a g~alitative or quantitative analysis made of HER-2/neu migration (i.e., translocation). A
quantified measure of the amount of the protein per cell can be taken by digitizing microscope images of stained samples, and converting light intensity values of pixels of the digitized image to optical density values, which correspond to the amounts of stained protein. See, e.q., Bacus et al., Applied Optics, 26 3280-3293 (1987).

- In particular, quantification can be accomplished in the following manner. A cell culture sample is stained for the oncogene product, according to a staining procedure as described above, or some other staining procedure known in the art. The cell culture sample also is stained for DNA, such as by the Feulgen technique. The DNA stain should be distinguishable on the basis of the wavelength emitted (i.e., of different color from the stain for the protein(s) to permit differentiation be~ween the stains). Digitization of different filtered images of the single sample image through respectively different filters, one for each specific stain, allows an optical density value to be associated with each pixel of each filtered image in a computer system programmed to process the images. The optical density of the protein stain image(s) and the optical den6ity of the DNA stain image are summed by the computer.

SuB~ S'IIET

214~382 WOg3/03 J PCT/US~U07117 - The DNA stain is applied to another sample of the same cell culture, and a human operator interactively identifies individual cells to the computer, which calculates sums of optical densities for the individual cells so identified. This second image ~upplies the average DNA per cell. The previous sum of optical density from the first DNA stain image, representing the total DNA that was seen in that image, is divided by the average DNA per cell for the culture this yields the number of cells in the first portion. The ~um of optical density for the protein then is divided by this number of cells to yield the average protein content per cell. A reference control portion of a standard cell line, not necessarily related in any way to the cells from the sample, and in which DNA content and oncogene protein content per cell are known, can be stained with identical stains and used to calibrate optical density with the mass of staine~ material. A fuller understan~ing of protein quantification and nuclear area measurement can be obtained from U.S. Patent No. 4,175,860;
U.S. Patent No. 4,998,284; U.S. Patent No. 5,008,185; U.S.
Patent No. 5,106,283; and U.S. Patent No. 5,028,209, which are incorporated by reference.

.
The guantification of membrane-bound HER-2/neu (and cytoplasmic HER-2/neu) and/or cell adhesion molecule(s) preferably can be carried out by selecting for optical density summation only those pixels in the digitized images which correspond to the membrane (or the cytopla6m) or representative portions thereof. Pixel ~election can be carried out by automatic computer algorithm or by human interaction.

Alternatively, membrane bo~d HER-2/neu and/or cell adhesion molecule(~) can be quantified u~ing the above-described digitzed image analysis in conjunction with fixation and s~inina proce~l~res which do not make the membrane - i S~ lt SHET

214.~ 382 W093/o~, ~ PCT~US92/07117 permeable to the elements of the staining complex, and thus result exclusively in staining of membrane-bound product.
Briefly, for example, sample cells are fixed for 60 minutes at room temperature in 10% neutral buffered formalin. The murine monoclonal antibody TA-l (Applied Biotechnology, Cambridge, MA), which is directed to the membrane-external domain of HER-2/neu, is applied typically at a concentration o

2~g/mL. This fixation procedure does not make the cells permeable to the TA-l antibody. Second step antibodies and stains (e.q., goat anti-mouse antibodies conjugated to a fluorescent dye) are applied with the result that only membrane-bound HER-2/neu is stained. The amount of membrane-bound HER-2/neu per cell averaged over a sample of cells is determined as described above, by image analysis and using a Feulgen stain for DNA.

Alternatively, indicia of terminal differentiation in cells subject to the method of the present invention include morphological changes in cells which are characteristic of a mature cell type. In cases where the morphological change is dramatic, such as a fundamental qualitative change in the shape or structure of a cell (or its nucleus) as viewed through a microscope, a determination of the extent of cell differentiation may be made by exam;n;ng the cells under a microscope and counting the number if cells which exhibit qualitative morphological features associated with terminal cell differentiation. Malignant cells characteristically are compact and spherical with a similar nucleus which densely stains, whereas terminally differentiated cells characteristically are flattened, having a cytoplaæm which exhibits a delicate lacy appearance and a diffu6e nucleus. The percentage of cellæ displaying the latter morphological features may be used to quantify the extent of terminal cell differentiation induced by a putative therapeutic agent in a - { SUilSTITUTE SHET

2145382 ~
W093/03, ~ PCT~US92/07117 given portion of cells and consequently permit a prognosis relating to the effect of the putative therapeutic agent in the malignancy sought to be treated.

Moreover, quantitative morphological differences, 6uch as the change in the ratio of cytoplasmic area to nucleic area which can be quantified by computerized image analysis techniques as described above, can be used to delineate between immature and mature cells.

Cell proliferation i~ yet another measure of the extent of terminal cell differentiation. Immature cancer cells will proliferate indefinitely whereas mature cells will not. A
stabilization and reduction of cell population as compared to untreated control cells indicates substantial ter~;n~l cell differentiation. A marked difference in growth curves between treated and untreated portions also may indicate substantial terminal cell differentiation. Statistical methods for analyzing cell populations are well known in the art, and the aforementioned examples 6hould not be taken as a limitation of the methods which may be applied to determine aspects of terminal cell differentiation within the cell population.

The invention ifi illustrated in the following Examples.

EXAMPLE I

Monoclonal antibodies-to the HER-2/neu product were made by injecting Balb/c mice intraperitone~ly 3 times (2 week intervals) with 3 to 5 x lO6 viable SRBR3 human breast cancer cells in ~hc~hate buffered 6aline (~PBS"). 8pleen cells of mice which developed a D~Lohy immNne ~e~onse were isolated and fu6ed with NSO myeloma cells, using polyethylene glycol, and hybridomas were sele¢ted with HAT

SUBSIIlult SHET

~ W093/037~. 2 1 4 5 3~8 2 PCT/US92/07117 (hypoxathine/aminopterin/thymidine) medium. Hybridomas were screened for specific binding to recombinant HER-2/neu product expressed on the surface of fixed Chinese hamster o~ary (CH0) cells which had been transfected with an appropriate expression vector. Monoclonal antibodies specifically binding HER-2/neu products were detected with l25I-labeled goat anti-mouse F(ab')2 antibody. The antibodies that specifically bound to the transfected CHO cells were selected for further analysis using either an immuno-precipitation assay with [35s]methianine labeled cells, or immuno-precipitation followed by auto-phosphorylation in the presence of MnCl2 and ~ P]ATP. This ;m~lnization procedure elicited specific antibodies to the extracellular portion of the human HER-2/neu antigen. Four of the monoclonal antibody preparations, designated Nl2, N28, N24 and N29, were depositied as described above. NlO was provided by Dr. Yosef Yarden of the Weizmann Institute, Rehovot, Israel. Monoclonal antibodies Nl2, N24, N28 are of the IgGl subclass and the N29 monoclonal antibody preparation is of the IgG2 subclass.

EXAMPLE II

N29 was puriied from ascites fluid by ~mTo~ium sulfate precipitation (40% saturation) followed by chromatography on a Sepharose-~rotein A column. Fractions containing the antibody were obtaine~ by elution at low pH (50mM citric acid at pH
4.8). The antibody preparation was homogeneously pure as determined by gel electrophoresis under reducing conditions.
SDS-polyacrylamide gel-separated heavy and light c~;ns were transferred to polyvinylidene difluoride membranes and subjected to Edman degradation. Amino acid se~uences of the amino termini of both the heavy and the light rhain~ (20 amino acids of each) were thus obtained, as follows:

SU~SIIIUI~ SHET

: ~ 2145382 WO93~03741 PCT/US92/07117 - H. chain:
Glu Val Gln Leu Val Glu Ser Gly Gly Gly Leu Val Gln Pro Lys Gly Ser Leu Lys Leu L. chain:
Asp Ile Val Met Thr Gln Ser Gln Lys Phe Met Ser Thr Ser Val Val Asp Arg Ile Ser In order to characterize the antibodies Nl2, N24, and N29, the following experiments were performed.

HER2 cells were plated in 24-well plates and assayed at confluence. Confluent monolayers of the HER2 cells were incubated for an hour at 22C with various concentrations of antibodies in PBS containing 1% bovine serum albumin (BSA).
After w~sh;ng with the same buffer, the cells were incubated for 90 minutes with I125-labeled goat anti-mouse F(ab')2 to deter~ine bound antibodies. The cells then were washed, ~olubilized with 0.lM NaOH, and the radioactivity determined in a gamma counter. Control cells were incubated in the absence of the murine antibody and their background binding was subtracted.

FIG. l shows the bin~i ng of the five antibodies to the HER2=
cells: Nl0 (~), Nl2 (-), N24 (o), N28 (~) and N29 (O).
All specifically bound to cultured cells that express HER2/neu with different apparent affinities. N28 and N24 di~played the highest apparent affinity wherea6 Nl0 exhibited the lowest apparent affinity.
.
HER2 ce~ls were metabolically labeled with t35~]methionine and the cell lysates were separately 6ub~ected to an immunoprecipitation aggay with l0~g of each antibody.- A6 a control, an irrelevant antibody, anti-dinitrophenol (anti-DNP), was used. Proteins were SU~SlilUI~ SHET

; r ~ ~
~- l 2145382 W093/0374~ PCTrUS92~07117 separated on a SDS-7.5% polyacrylamide gel. The results are shown in FIG. 2A.

The immunoprecipitation assay was performed as described in Example I but with unlabeled cells. Prior to electrophoresis, the proteins from the cell lysate were labeled by autophosphorylation with gamma[32P]ATP and lOmM MnC12.
Autoradiograms are shown: NI (non-immune serum), Polycl (polyclonal anti-HER2/neu antibody). The results are shown in FIG. 2B.

All ;mm1lnoprecipitated a single protein of 185kD from metabolically labeled HER2 cells, as shown in FIG. 2A. This also was reflected in an in vitro kinase assay performed on the ;~m~lnoprecipitates (FIG. 2B). None reacted with EGFR or with the rat pl85neU. Western blot analysis of the HER-2/neu protein showed that only Nl2 and N29 were capable of reacting with the denatured form of the receptor. (See Table I). For the Western blot analyæis, HER2 cell lysates were separated by SDS-PAGE, transferred to nitrocellulose, and blotted with the antibodies, followed by detecting using horseradish peroxidase conjugated goat anti-mouse F(ab')2.
~ j .
The antibodies also were assayed for their ability to affect tumor growth of murine fibroblasts transformed by overexpression of HER2/neu in athymic mice. HER2 cells t3 x 106) were injected subcutaneously into CDl/nude mice. The antibodies or a control (an irrelevant antibody to dinitrophenol or PBS) were injected intraperitoneally, ~nto groups of 5 nude mice, on days 3, 7 and lO after tumor ;noc1~lation. ~Tùmor parameters were measured twice a week with callipers, and tumor volume was calculated according to the SuBslilul~ SHE~T

WO 93/03741 . 2 1 ~ 5 3 8 2 ~usg2~0~

C

o o~
~ al g s~ O o cr~ u~

,~ , o o o o o ~ ~ o E~ ~ o'~ '~
U
- ~ ~ CO o o _~ o o r~
h O ~ ~ O

.C
O o ~ 'D ~ ~

~1 + ~ ~ +
O O
~m Z
;2 ~ O ~ 0 N

-- - ~ll~lll~lt SHET

214~382 ~WO93/03~1 PCT/US92/07117 formula: tumor volume = length x width x height. In order to -validate volume measurements, the correlation between the tumor volume and tumor weight was determined on the day of animal killing. The results are shown in Table 1 (average tumor volume as percentage of control wherein "100" equals the control value; n=5, measured 21 days after tumor inoculation) and in FIG. 3.

FIG. 3A depicts tumor volumes of each group of mice, on day 21, post inoculation, after treatment. The tumorigenic growth of HER2 cells was significantly inhibited (P< 0.05 as calculated using the Anova and Duncan's multiple comparison test) in nude mice that were injected with N29, and N12, when compared with mice that received no antibody or the control anti-DNP antibody.

FIG. 38 depicts the kinetics of tumor growth in antibody treatment mice: control (O), N10 (~), N12 (-), N24 (o), N28 (~), N29 (O). It can be seen that the inhibitory effect of the antibodies persisted over 31 days after tumor injection. Antibodies N10 and N24 exhibited less efficient inhibition of tumor growth. In contrast, monoclonal antibody N28 consi~tently stimulated tumor growth. Essentially identical results were obtAine~ in three separate experiments.

To test the possibility that the effects seen in vivo are reflected in vitro, cell proliferation assay in culture and cytotoxicity assay with the antibodies were performed on SRBR3 human breast carcinoma cell line (from the American Type Culture Collection).

--In the cell proliferation assay, SKBR3 cells were plated in 24-well plates 103 cells/well) and incubated for 48h in medium supplemented with 10% fetal calf serum. The amount of serum was then decreased to 5% and the indicated antibodies SUBSTITUTE SI~IEET

- ` 2145382, W O 93/03741 PC~r~US92/07~17 added at lo ~g/ml concentration. Five days later, the number of viable cells was determined. The results (in percentage) are shown in Table 1 wherein "100" equals the amount of cell proliferation for control treatment.

- Complement-dep~n~ent cytotoxicity ("CDC") assay of SRBR3 tumor cells was performed as follows: the SKBR3 tumor cells were incubated at 37C for 2 hours, in a volume of O.lml fetal calf serum, with 300 ~Ci of Na[52Cr]04 (New England Nuclear). At the end of the labeling period the cells were washed three times in PBS and 1.5 X 104 cells were plated in each well of a 96-well microtiter plate.

Various concentrations of the antibodies were added, and incubated with the cells for 1 hour, followed by the addition of human or rabbit complement and incubation for further

3 hours. Appropriate control wells cont~ining cells alone, cells with no antibody, or no complement, and cel}s lysed in 10% SDS were set up in parallel. The results also are shown in Table 1. Values represent tSlCrl04 release (determined in a gamma counter) from cells treated with the indicated antibodies (50~g/ml) as percentages of total cellular content of [51Crl. The means of triplicate determinations are given. Corrections were made for 6pontaneous release, in the absence of antibody and complement.

~ ntiho~y-mediated cell-depen~ent cytotoxicity ("ADCC") assay was performed as follows: the &KBR3 tumor cells were labeled with Nat51CrlO4 as described above. 5 X 103 cells in 25~1 were incubated for 1 hour with various concentrations of the antjbo~jes, and then for 5 hours with effector cells, human peripheral blood lymphocytes (O.lml, lymphocytes: tumor cells - 140:1), or with mouse spleno~Les (120:1). t51Crl release was determined a6 described above.
The results in Table 1 e~Less percentages of the SU~ TE S~EET

~ WOg3/03. . 2 1 ~ 5 ~ 8~2 PCT/US92/07117 antibody-mediated cell-depen~ent lysis of SKBR3 cells using 50 ~g/ml of each antibody in the assay.

Two different assays were employed to test the capacity of the monoclonal antibodies to elevate tyrosine phosphorylation of the HER-2/neu protein: HER2 ceils were metabolically labeled with t32P]orthophosphate, incubated with the antibodies and subjected to two consecutive immunoprecipitation steps with anti-phosphotyrosi~e and anti-HER-2/neu antibodies, as described by Yarden et al., Proc. Natl. Acad. Sci., 86:
3179-3183 (1989). Alternatively, SRBR3 cells were first incubated with the antibodies and then subjected to two consecutive immunoprecipitation steps, followed by an in vitro phosphorylation assay in the presence of gamma (32P)ATP and MnC12 .

The SRBR3 and HER2 cells were grown in a 24-well plate and labeled for 4h in Dulbecco's modified Eagle medium (DMEM) without phosphate, but in the presence of 1% dialyzed fetal calf serum ("FCS") and ~32P]orthopho6phate (0.5mCi/ml). The cells were washed with PBS and incubated for 15 min at 22C
with fresh medium cont~in;ng antibodies at a concentration of lO~g/ml. After w~h;ng, the cells were lysed in solubilization buffer (50mM Hepes, pH 7.5; 150mM NaCl; 10% (vol vol) glycerol; 1% Triton X; lmM EDTA; lmM EGFR; 1.5mM MgCL2;
2mM PMSF; 1% Aprotinin, 1% Leu~e~Lin (added just before use)) and the tyrosine phosphorylated HER2/neu protein was immunoprecipitated with an agarose-immobilized antibody to ~ho~hotyrosine (Hung et al., Proc. Natl. Acad. Sci., 84:
4408-4412 (1987)). The immuno-complexes were eluted with SUBSTITUTE StlEET

.

WOg3/03741 2 1 ~ 5 3 8 2 PCTrUS9~07117 solubilization buffer containing 50mM p-nitrophenylphosphate and subjected to immunoprecipitation with a rabbit polyclonal anti- B R-2/neu antibody, directed to the carboxy terminus of the protein.

According to the first assay, monolayers of HER2 cells were labeled with t32P]orthophosphate and then incubated for 15 min at 22C with lO~g/ml of each antibody. Tyrosine phosphorylated proteins were immunoprecipitated with an anti-phosphotyrosine antibody, followed by specific elution and a 6econd immunoprecipitation -step with the rabbit anti-HER-2/neu polyclonal antibody. The extent o induction of tyrosine phosphorylation of the HER-2/neu protein by the antibodies was determined by densitometry of autoradiograms.
The results are shown in Table I and in FIG. 4A.

In the second assay, 8RBR3 cells were first incubated with the antibodies, immunoprecipitated in two consecutive steps, as described above, and labeled by autophosphorylation with gamma t32P3ATP and Mn2 , The autoradiograms of the 8DS gel separated proteins are shown in FIG. ~B.

Similar results were obtained in both analyses: NB28 6ignificantly stimulated phosphorylation of the HER-2/Neu product on tyrosine residues, whereas the other antibodies displayed low or no activity (NlO) in living cell6.

- The interaction of ~e~e~or tyrosine ~in~6s with their re~e-~ive ligands i8 usually coupled to rapid endocytosis.
The potential of the antibo~ies to the human HER-2/neu protein to accelerate the ~u~uver of the receptor was tested. For thi6 purpose, HER2 cells were biosynthetically labeled with radioactive methionine, _nd then rha~e~ for v_rious periods of time with fresh medium that cont~ne~ different antjbo~ies. At Sll~SlllUl~ SHET

.
-WOg3/0374, 2 1 ~ 5 3 8 2 - PCT/US92/07117 the end of the chase period, the residual labeled protein were immunoprecipitated and analyzed by gel electrophoresis and autoradiography.

SKBR3 or HER2 cells were grown in 24-well plates to 80%
confluence, and then labeled for 16 hours at 37C with t35S]methionine (50~Ci/ml). After washing with PBS, the cells were incubated with fresh medium in the absence or presence of the antibodies (at a concentration of lO ~g/ml), for various periods of time. The cells then were washed and cell lysates were subjected to ;~m~lnoprecipitation with a rabbit polyclonal antibody to the HER2/neu protein. The results expressed as the half-life of the labeled protein (tl/2) are shown in Table I.

FIG. 5 shows the effect of the antibodies on the rate of turnover of the HER-2/neu product. HER2 cells were labeled with [35S]methionine in a 24-well plate and then chased for the indicated period of time with fresh medium that contained the indicated antibodies. Residual 35S-labeled HER-2/neu protein was subjected to immunoprecipitation with the rabbit polyclonal antibody and separated on a SDS-gel. Quantitative - ~analysis of receptor turnover is shown, as determined by measuring the densitometry of the autoradiogram. Control cells without antibody treatment (O), NlO antibody treated cells (~), Nl2 (-), N24 (o), N28 (~), and N29 treated cells (O). As shown in FIG. 5, all the antibodies accelerated, to different extents, the rate of turnover of the receptor, with antibody N29 being the most effective.

- -Conjugates of ricin A and antibodies N24 and N29 were prepared by covalent cro~slink;ng with the bifunctional reagent 8PDP (succinimidyl-3 2-pyridyldithiopropionate). Unbound ricin was separated by gel filteration - on Seph~ GlO0. The - SUB~lllUlt SHET
.

1 wog3/0374~ 21~382 PCT/US9~07117 conjugates were purified by passage on Blue Sepharose CL-6B
(removal of unsubstituted antibody).

The conjugates were assayed for their ability ~o affect tumor growth as described above, for this purpose, CDl nude female mice received HER2 tumor cells (3.2 x 106) injected subcutaneously. Eleven days later a single injection of Ricin A - antibody conjugate was injected intravenously.
RicinA-N24: 3.9~g Ricin A bound to 65~g o antibody. Ricin A-N29: 4~g Ricin A bound to 90~g of antibody. The tumor size was measured every 3-4 days, during 40 days. The results are depicted in FIG. 6, ~howing retardation of tumor growth by both conjugates Ricin A - N24 antibody and Ricin A - N29 antibody.

EXAMPLE III

Human breast cancer cell lines, AU-565, D A-MB 453 and MCF-7, are well known in the art and widely available. The AU-565 cell line overexpresses both HER-2~neu and EGFR, MDA-MB
453 cell~ overe~,es HER-2/neu; MCF-7 cells do not overexpress HER-2/neu. In each case cultured cells of the respective cell lines trypsinized, pelleted and 6eeded into four chamber ~lides (Nunc, Naper~ille, ILL) at 0.5 x 104. The AU-565 cells were obtained from Naval Biosciences Laboratory in Oakland, California. Cultures of MCF-7 cells (ATCC accession no. MCF-7 HTB 22) and MDA-MB 453 (ATCC accession no. 453 HTB 131) were obt~ned from the American Type Culture Collection in RG~hville, Maryland.

The cells were cultured in RPMI 1640 ~upplemented with 20%
fetal bo~ine ~erum, penicillin (l00~g/mL) and DL~ omycin (lOO~g/mL) in a humidifled 1n~h~tor with 8% CO2 in air at 37-C. One day after ~ ng, when the cells were approximately - SU~llllllt SHEI

- ` 2145382 W093/037.. i PCT/US92/07117 10%-20% confluent, the cell culture medium cells were supplemented with lO~g/mL of one of the following monoclonal antibody preparation having specificity for the extracellular d9~; n of the HER-2/neu protein: N12, N24, ~28 and N29, control IgG (an irrelevant IgG antibody). Also, PBS alone was added to certain control cultures as a control in which IgG was absent.
The cells were cultured for an additional 4 days and then ;ned to determine the efficacy of the respective monoclonal antibody preparation with respect to inducing the malignant breast cells to undergo terminal cell differentiation.
Differentiation was assayed by the percentage of cells producing lipid, cell numbers, nuclear area per cell (~m2) and the translocation of HER-2/neu as evidenced by total cellular content of the protein (where 100% expression equals amount HER-2/neu in sparsely growing untreated cells) and human identification of the location of staining in confocal microscopy. The results shown in Table 2 relate to the AU-565 cell line.
Table II
.
Ab Cell # HER-2/neu Nuclear % Cells % Cells -~ 104/Cm2 area lipid casein Cont. 6.0 103 100 12 20 IgG 6.3 84 101 7 20 N12 5.6 154 121 40 > 90 N24 7.1 152 147 52 > 90 N28 8.6 104 102 8 < 30 N29 4.8 160 154 55 >90 - ~ SUESTITUTE SHET

wo g3,0374l 2 1 4 5 3 8 2 ' PCT/US9V07tl7 - The data above indicates that monoclonal antibodies N29, N24 and N12 ~n~ceA the malignant breast cells to undergo differentiation and exhibit mature phenotypic traits, whereas the N28 antibody, which also has specific b;nA~ng affinity for a portion of the extracellular domain of the B R-2/neu product, actually promoted the tumorigenicity of the treated AU-565 cell6. Confocal micro6cope images showed that treatment of AU-565 cells with N28 antibody did not result in a translocation of the HER-2/neu protein from the membrane, while translocation from the membrane to the cytoplasm and perinuclear region of the cells wa6 ~emonstrated in AU-565 cells treated with the N29, N24 and N12 monoclonal antibodies.

Results for the MDA-MB 453 cells line were similar to results for the AU-565 cell. The MCF-7 cell6, which did not overe~,ess HER-2/neu, were largely unaffected by the antibodies, e~'e~- that monoclonal antibody N29 increased the percentage of cells exhibiting lipid droplets.

Phenotype expression as a marker of terminal cell differentiation was measured by detecting the production of lipid droplets and casein, both of which are components of human milk. Lipid droplets were detected by a modified "Oil Red O in propylene glycol" method. D.C. ~heehan, Theory and Practice of Hi6totechnoloqY, p. 209, C.V. Mosby Company, St.
Louis, (2nd ed. 1980). For the lipid st~n~ng ~,ocad~re, the culture medium was remo~ed, the cells were rinsed with 0.05M
phosphate buffered ~aline, pH 7.6, and fixed by a quick dip in -20C methanol/acetone. After fixations, the slides on which the cells were gro~n were placed in absolute propylene glycol for 2 minutes at room temperature ~n an Oil Red O st~ini solution. The slide6 then were dipped in 85% isopropanol, rin6ed with deionized water, counterst~ n~ in Mayer's hematoxylin, blued in saturated lithium carbonate, and covered with glycerol jelly.
SUBSTITUTE SHEET

W093/0374l 2 1 4 5 3 8 2 PCT/US92/07117 .

The presence of casein was detected by histochemical staining with a mouse monoclonal antibody to human ~ or casein. After the medium was removed, cell slides were rinsed with PBS, and the cells were fixed in ethanol-formol solution at room temperature for 10 minutes. After nonspecific binding was blocked with 20% goat serum for 20 minutes at room temperature, the cells were incubated with the anti-casein (~
and ~) antibody (1:250 dilution) at room temperature for 60 minutes. The slides were then rinsed with 0.5M Tris-buffered saline (TBS), pH 7.6, and then incubated with biotinylated goat anti-mouse IgG ~Jackson Lahs, West Grove, PA) at 1:200 dilution for 30 minutes. The cells were rinsed with TBS, and streptomycin conjugated alkaline phosphatase (Jackson Labs) at 1:200 dilution was applied to the cells for 30 minutes. The cells were rinsed again with TBS and incubated for 15 minutes with CAS Red (Cell Analysis Systems, ~ rst~ ILL) as the chromogen. The cells were then counterstained with CAS DNA
stain (Cell Analysis Systems).

The localization of the HER-2/neu product ~i.e., translocation of HER-2/neu) was deter~ined using confocal microscopy after immunofluorescence staining. For deter~;nation of translocation, after the culture media was removed and the cells were rinsed with PBS, the cells were made=
permeable with 95% ethanol for 10 minutes. Following a TBS
rinse, the cells were post-fixed in 10~ neutral buffered formalin for 30 minutes. After a deionized water wash, the cells were st~ine~ for DNA with a Feulgen stain, whereupon they were rinsed well with TBS (pH 7.6). After a 20 minute block with 20% normal goat serum, one portion of the cells (the other portion served for an estimate of the average DNA content of the cells, described below) was ;ncl~h~ted with a polyclonal antibody to the C terminus of the HER-2/neu protein (o,.coye~e Kit from Cell Analysis Systems) for 60 minutes at room ~UES~ S'~:ET

1 W093/03741 2 1 ~ 5 3 8 2 t PCT/USg2/07tl7 temperature. The cells then were rinsed with TBS, and incubated with a first ljnkjng antibody, mouse anti-rabbit IgG
at a protein ~oncentrition of lOmg/L (JAckson Laboratories) for minutes. The dichlorotriazinyl amino fluorescein tDTAF]-conjugated goat anti-mouse IgG (Jackson Labs) was applied at a dilution of 1:100 for 30 minutes at room temperature. The cells then were rinsed with TBS, and coverslipped with gelvatol. Localization was deter~;ne~ using a Bio/Rad MRC-600 confocal scanning microscope adapted with a fluorescein filter. Confocal optical sections were recorded at 1~ intervals with 10 times averaging per image.

A CAS 200 Image Analyzer (Cell Analysis Systems~, a microscope-based, two color image analyzing system, was used in the quantification of the HER-2/neu protein. 80th solid state imaging ch~nnels of the CAS 200 Image Analyzer were used.
Digitized light intensity values were converted to optical density valueæ and added together, the result correspon~i n~ by the Lambert-Beer Absorption Law to the amounts of stained cell constituents. The two imaging ~h~nn~18 were specifically matched to the two components of the 6tains used. One chAnnel was used for guantifying the total DNA of the cells in the field following Feulgen staining with a DNA staining kit and the other for quantifying the total HER-2/neu protein of the cells in the field following immunostaining.

A separate preparation of cell6 from the same culture (the ~eco~ portion) was 6tA~ne~ only for DNA. A human operator identified individual cells to the apparatus, and optical densities of the pixels associated with each cell were summed. ~ummed optical densities for each cell as well as a count of the number of cells were ~.o~ c~. Thi6 ~upplied the total DNA amount per cell for the culture.

SUBSTITUTE SHEET

.

-~ 214~382 WO93/03741 PCT~US92/07117 Since the total DNA amount per cell was known from this second sample, the average total HER-2/neu protein per cell could be computed from the data of the first sample, which had been stained for both DNA and HER-2/neu. Sparsely growing AU-565 cells were used or calibrating the HER-2/neu protein content. The level of staining in such cells was defined as 100%. A complete description of this quantification is available in Bacus et al., Arch. Pathol. Lab. Med., 114:
164-169 (1990). Cell numbers were determined by hemocytometer chamber counting, and viability was monitored by trypan blue dye exclusion.

According to the method of treatment and analysis described above, the N29 antibody was found to be the best differentiation inducer. Treatment of AU-565 cells for four days with 10 ~g/mL N29 antibody doubled the proportion of cells with flat morphology, and increased the nuclear area of the cells on average to 154~m2 over the control cell nuclear area of lOO~m2. The fraction of morphologically mature AU-565 cells increased from 10-20% in the untreated cells to more than 90% in the cells treated with N29 antibody.
The fraction of 565 cells treated with N29 antibody which contained lipid droplets was 55%, compared to 12% in the untreated control. The fraction of N29-treated AU-565 cells staining positively for the presence of casein after four days was more than 90%, co~r~red to 20~ for the untreated control.
The population of N29-treated AU-565 cells was 4.8 x 10 untreated cells.

Incubation of AU-565 cells with N29 antibody resulted in a decrease in membrane st~;nin~ for HER-2/neu which was .. .
accompanied by diffuse cytoplasmic localization of the protein. Quantification of the st~inin~ revealed that the redistribution involved a transient increase in total cellular SUBSTITUTE S~EET

.

WOg3/037~ 2 1 4 5 3 8 ~J' PCT/US92/07117 HER-2/neu content. Confocal microscope images confirmed the immunohistochemical stain;ng results. The protein migrated from the membrane and localized in the ~y~oplasm and in particular the perinucleus upon treatment with N29 antibody.

Treatment of MDA-MB 453 eells with N29 antibody (data not shown) elicited a marked growth inhibition of 60%, and an increase in cells positive for differentiation markers: 90% of treated cells stained positively for lipid droplets, and 70% of treated cells stained positively for casein. Treatment of MCF-7 cells with N29 antibody increased the fraction containing lipid droplets and casein to about 90%, compared with 2% in the untreated control portion. N29 had only a small growth inhibition effect on MCF-7 cells.

EXAMPLE IV
. ~
Cells of the AU-565 cell line were treated with l~g/ml, 3~g/ml and lO~g/ml of the Nl2, N24, N28 or N29 antibodies for a period of four days as described in Example III. An unrelated IgG (lO~g/ml) was used as a control.
St~i n ing for lipid droplets and HER-2/neu, as well as quantification of HER-2/neu by optical density values, nuclear area (~m ) and determination o localization of the=
protein also were carried out as in Example III. The results are shown in Table III.

Again, N29 antibody ~howed the best efficacy in inducing differentiation. The N29 antibody preparation demonstrated a dose-depen~ent differentiation-inducing effect at co--c~ ration6 as low as l~g/ml. Again, ~parsely growing AU-565 cell6 were u~ed for calibrating the level of HER-2/neu in the cells. The level of ~t~ining in these cells was defined as 100%.

SUe~lLt~lt SH~

~`
wo g3/0374l 2 1 4 5 3 8 ~ PCT/US92/07117 ~- Table III
Ab Cell # HER-2/neu Nuclear ~ Cells Conc.
1o4/cm2 area lipid ~g/ml Control. 6.2 89 111.5 28 IgG 6.1 95 117.4 19 N29 4.8 119 144.2 53 N12 6.0 85 101.3 18 N24 5.6 102 136.0 48 N28 6.1 86 114.7 31 N29 4.1 136 166.5 73 3 N12 5.5 90 119.4 38 3 N24 4.9 105 156.1 57 3 N28 7.4 101 119.0 22 3 N29 3.6 124 167.2 >90 10 N12 5.3 104 117.3 61 10 N24 4.4 95 156.6 69 10 N28 8.0 117 117.0 17 10 EX~MPLE V

Human breast cancer cell line AU-565 was cultured as in Example III above, and incubated with the widely available TA-l monoclonal antibody. Incubation with the TA-l monoclonal antibody was initiated 24 hours after cell inoculation. From 15% - 20% of the cells in the control cultures exhibited a mature phenotype, characterized by large, lacy nuclei, and a spread cytoplasm conta;n;ng sizeable lipid droplets.
TnCllh~tion of AU-565 cells for 2 days with l~g/mL TA-l resulted in a three dimensional pattern of cell growth with an increased fraction of cells having mature phenotype. On the fourth day, the number of cells in the treated portion decreased by 60% relative to the control, and the fraction of mature cells increased from the range of 15 - 20% to the range of 50% - 60%.

S~ET

W093/03~-. ^ ~' PCT/US9~07117 Immunohistochemical staining for lipid drople~s and nuclear area (~2) were performed as in Example III. Cell numbers were deterr;ne~ by hemocytometer chamber counting. Results are ~hown in Table IV:

Table IV

Ab Cell # Nuclear % Cells Conc.
104/cm2 area lipid ~g/ml Control 5.7 100 23 o.0 IgG S.l 96 16 1.0 TA-l 3.0 156 33 O.S
TA-l 2.3 160 48 1.0 EXAMPLE VI

A 30kD factor, gp30, secreted from MDA-MB-231 human breast cancer cells has been shown to be a ligand for the HER-2/neu product, a 18SkD transmembrane receptor (al~o known as pl85HER 2/neu) enoo~e~ by the HER-2/neu oncogene.

Briefly, gp30 can be isolated from the conditioned media of MDA-MB-231 cells low-affinity chromatography on a heparin-Sepharose column. Fraction~ cont~ining active gp30 may be detected by the ability of gp30 to bind EGFR on the cell membranes of A431 cell6 or MCF-7 cell6. Fractions cont~injng gp30 activity which are obt~ne~ after heparin-Serapho6e chromatography may be then chromatographed by rever~e-pha6e chromatography on a Bondapak C3 column equilibrated on O.OS%
trifluoroacetic acid and eluted with a 6tep gradient of acetonitrile and then rechromatographed, in a _ecQn~ round of LeveL8e-pha6e chromatography on the Ro~d~p~ C3 column (equilibrated in 0.05 trifluoroacetic ~cid), where elution i6 with a narrow gradient of acetonitrile.

SUBSrlTUTE SHEET

wo g3/03~- 2 1 4 5 3 8~2 PCT/US92/07117 The gp30 used in Example VI was dissolved in P8S and filtered. The protein concentration of the ligand solution was confirmed after the filteration step.

Malignant breast cells of each of the three cell lines, AU-565, MDA-MB 453 and MCF-7, were seeded and cultured as described in Example III. The culture media was supplemented with 0.0, 0.3, or 6.Ong/mL of gp30 instead of the monoclonal antibody preparations. The methods for determining the presence of lipids droplets, nuclear area (~m2) and cell number were all carried out as described in Example III.

Treatment of AU-565 cells with various doses of gp30 inhibited cell growth in a dose-dep~n~ent fashion, in the nanogram range. Treatment of AU-565 cells with 6ng/mL for four days resulted in about 40% growth inhibition. Treatment of MDA-M3 453 cells for four days with 6 ng/ml of the ligand gp30 resulted in 42%- growth inhibition compared to the untreated control. Similar treatment of MCF-7 cells resulted in no inhibition of growth. These results are shown in.Table V.

Table V
, ..~ ......
Cell Cell # Nuclear ~ Cells Conc.
Line lo4/cm2 area lipid ng/ml

4.0 96.0 15 0.0

5.3 162.0 28 0.3 ' 2.4 204.0 76 6.0 3.4 65.3 20 0.0 }i 2.9 77.8 62 0.3 2.0 113.1 84 6.0 - ~ S~ UT~ SHET

wog3/037~ 21~53~2 PCT/US92/07117 - Table V (cont.) 11.0 252.0 <1 0.0 11.0 251.0 <2 0.3 11.0 277.0 5 6.0 At the time of treatment, about 7~ of AU-565 cells, 10% of MDA-MB 453 cells, and less than 1% of MCF-7 cells for 4 days with 6ng/mL gp30 increased the fraction of cells having lipid droplets to 76%, whereas 15% of the control cells had lipid droplets. With res~ec~ to MDA-MB 453 cells, treatment for four days with 6 ng/mL gp30 increased the percentage of cells exhibiting lipid droplets to 84%, whereas 20% of the control cells exhibited lipid droplets. Similar treatment of MCF-7 cells (which do not express HER-2/neu) resulted in at most about 5% of the cells exhibiting lipid droplets, while less than 1% of control MCF-7 cells (O.Ong/mL gp30) exhibited lipid droplets.

AU-565 cell6 were also treated with gp30 at concentrations less than lng/ml. Surprisingly, treatment of these cells with a very low dose of gp30, less than lng/ml, resulted in stimulation of cell growth, as can be seen in FIG. 7. ~his - data is shown in Table VI:

Table VI

Cell # Nuclear % Cells Conc.
104Jcm2 area lipid ng/ml 8 94 17 0.00 19.3 93 14 0.03 12.6 101 16 0.1 - ----- 9.6 162 27 0.3 7.5 180 ~6 1.0 - 4.6 206 76 3.0 2.3 236 ~90 6.0 SUBSTITUlE S~IEET

Cells not treated with gp30 attained a cell density after six days of approximately 8 x 104/cm2. The results show that maximum growth stimulation occurs at a ligand con~sntration of about 0.03ng/mL for gp30, where a cell density of about 1.93 x 105/cm2 is attained. Thus, very low concentrations of gp30 appear to agonize malignant cell growth in cells which overexpress HER-2/neu.

With respect to inducing terminal cell differentiation in malignant cells expressing or overexpressing HER-2/neu, more than 90% of AU-565 cellæ treated with 6 ng/ml gp30 for 6 days evidenced mature morphology. Treatment of AU-565 cells for 6 days with 6ng/ml gp30 increased the fraction of cells having identical treatment conditions, the percentage of cells staining positively for casein increased to 90% (control = 30%).

o Unlike AU-565, MCF-7 cells (which do not express HER-2/neu) treated with 6 ng/ml gp30 did not show marked morphological differences compared to untreated cells.

As with induction of terminal differentiation by monoclonal - antibodies specific for the extracellular ~om-in of the HER-2/neu protein, induction of terminal differentiation by gp30 resulted in translocation of HER-2/neu protein from the membrane to the cytoplasm and perinuclear region of the cell.

During four days in culture, the cell surface of 80-90% of untreated AU-565 cells reacted with the antibody to the HER-2/neu protein, as shown by immunostaining as described in the previous Examples. The remaining cells, which had the morphology of mature cells, showed re~uce~ membrane st~inin~
and diffuse cytoplasm st~;n;ng. Treatment of AU-565 cells with ~. .
cQncentrations of gp30 which inhibited growth and induced differentiation markers (over lng/mL) caused a time depPn~Pnt .
- SUBSTITUlE S~IEET

~_ wog3/037~ 2 1 4 5 3 8 2 PCT/US92/07117 decrease in membrane staining, and a transient increase in total cellular HER-2/neu staining and cytoplasmic staining, as can be seen in Fig. 8. This diffuse immunost~;ning increased two- to three- fold during the initial two days of treatment, and decreased in the following two days.

However, treatment of AU-565 cells with 0.03 - 1 ng/mL gp30 did not change the immunost~inin~ pattern of HER-2/neu, which remained mainly membranous.

The immunostaining for HER-2/neu in MDA-MB 453 cells was less intense, yet the pattern and kinetics of staining, after a similar treatment, were similar to those observed in AU-565.

These results indicate that treatment of breast cancer cells with gp30, either inhibited or accelerated breast cancer cell growth, dep~nAing on the concentration of the ligand.
Ligand concentrations which resulted in cell growth inhibition induced cellular responses that resulted in cell differentiation and acquisition of mature phenotype, which was associated with translocation of of the HER-2/neu protein from the membrane to the perinuclear area.

EXAMPLE VI

Two human breast cancer cell lines, AU-565 and M OE -7, were treated with the chemical6 mycophenolic acid ~MPA), phorbol 12-myristate 13-Acetate (PMA) or retinoic acid (PA) which are known to induce maturation at low cnn~entrations in a variety of human cell types. m e cells were cultured as in Example III. Cells were inoculated into four chamber slides (Nunc) at 0.5 x 104 or 2 x 105 cells in lmL of medium per chamber or into lOOmm petri ~ishes at 5 x 104 cells in lOmL of medium.

SUBSTITUTE SHEET

W O 93/03741 2 1 4 ~ 3 8 2 PC~r/US92/07117 PMA and RA were dissolved in dimethylsulfoxide and stored at -70C. MPA was dissolved in 150 mM NaHC03. Treatment with MPA, PMA or RA was initiated 48 hours after cell inoculation.
Sparse cultures of the two cell lines were treated for four days with 9~M MPA, 1.6nM PMA or 2.5~M RA.

Three measures of cell differentiation were used. These include cell number, nuclear area (~m2) and precentage of cells expressing lipid droplets. The results are shown in Table VII:

Table VII

Treat. Cell # Nuclear % Cells Conc.
104/Cm2 area lipid ~M

Control 6.0 80 14 0.0 MPA 1.0 200 63 9.0 PMA 0.8 285 67 1.6 RA 2.1 220 97 2.5 Control lS.0 170 <1 0.0 MPA 2.7 163 5 9.0 PMA 12.5 167 6 1.6 RA 7.2 166 1 2.5 Cell proliferation was determined by counting cells in a hemocytometer chA~er. The cell count was monitored over four days.

Qualitative morphological appearance also characterized differentiation. Analysis of cell morphology in control cultures indicated that 70-80% of untreated, sparsely growing AU-565 cells had the morphology of immature cells, characterized by compact nuclei enclosed by a fine layer of cytoplasm. Another 10-20% displayed a morphology associated SUBSTITUTE S~IEET

W093/0374 2 1 4 5 3 ~ 2 ~ PCT/US92/07117 with mature cells, having large and lacy nuclei surrounded by sizeable flat cytoplasm~ Treatment of this cell line with MPA
increased the fraction of morphological appearance.

Phenotype expression as a marker of terminal cell differentiation was measured by detecting the production of lipid droplets and casein, both of which are components of human milk. Lipid droplets were detected by a modify "Oil Red O in propylene glycol" method, as described in previous Examples.

The presence of casein was detected by histochemical staining with a human antibody to human casein.

At the time of treatment, about 5% of the cells in the sparse AU-565 cultures and less than 1% of the cells in the MDF-7 cultures contained small lipid droplets. Treatment of AU-565 cultures with MPA or PMA increased the fraction of cells containing the lipid droplets in a time depen~ent manner to 60-70%. Treatment with RA increased this fraction to more than 90%. Moreover, the lipid droplets in the treated cells were visibly larger than those observed in untreated cells by more than five-fold.

Unlike the AU-565 cells, the MCF-7 cultures treated with MPA, PMA, or RA showed only a small increase in the fraction of cell& contA;n;ng the large lipid droplets: up to about 5% of the cell6 in cultures treated with MPA or PMA had little or no increase over controls in cultures treated with RA.

..... ~., - .

SUBSTITUTE S~EET

i WOg3/03741 - PCT/US92/07117 Four days after treatment of the two cell lines was begun, the control cultures contained less than 2% of cells that reacted positively with the anti-casein antibody. Treatment of AU-565 cultures with either MPA or RA increased this to 70 to 80%, and treatment with PMA to about 90%. Treatment of the MCF-7 cultures with MPA or RA also increased the percentage of cells staining positively for casein. PMA, even at high doses, had little or no effect on the MCF-7 cell fractions staining positively for this protein.

The HER-2/neu protein was detected by specific antibodies, as described in Example III, and translocation and quantification were performed as also described in Example III.

During four days in culture, the cell surface membrane of 80-90% of untreated AU-565 cells reacted with the two antibodies. The remaining cells, which had the morphology of mature cells, showed reduced membrane staining but increased diffusive cytoplasmic staining. Treatment of AU-565 cells with MPA, PMA or RA caused a time dependent decrease in cell surface membrane concentration of HER-2/neu, and a two to three-fold increase in cytoplasmic concentration of the protein. The ;r~nostaining in the untreated MCF-7 cells was about one-tenth that in untreated AU-565 cells. However the pattern and kinetics of immunost~;nin~ after treatment with MPA or RA were similar to those observed for AU-565 cells. PMA, which did not induce differentiation markers in the MCF-7 cells, did not cause a change in the pattern of immunostaining with these antibodies.

EXAMPLE VII
. . .
The biological effects of NDF determined on AU-565 cells in a manner similar to that described above in Example VI for SU8Sr1TUTE SHEET

~ wos3/o374 2 1 4 5 3 8 2 ~ i PCT/US9~07117 gp30. Briefly, AU-565 cells (0.4 x 104) were inoculated into culture disheæ in 1 ml of medium supplemented with 10% serum.
Twenty-four hour~ later NDF was added st the indicated ron~entrations, and the cells were analy~ed after four additional days. Cell numbers were determined, and nuclear area was estimated by an imagining system after DNA staining with Feulgen. The numbers given are the calculated averages from ten microscope fields (40 x magnification). The results are shown in FIG. 9A.

AU-565 cells were treated as above and then stained for casein and lipids as described above. The average fractions of cells st~ine~ positively for lipids (closed circles) and casein (open circles) were determined by counting stained cells in ten microscope fields (40 x magnification). The variation among fields did not exceed 15%. The results are shown in FIG. 9B.

MDS-MB 453 (105) cells were inoculated into multiwell culture dishes and after 24 hours their medium was replaced with serum-free medium. This was supplemented with 5 ng/ml EGF
(squares) of 5 ng/ml NDF (circles). Control cultures (triangles) received no growth factor. The dishes were then incubated at 37C, and on the indicated days cell numbers were determined in duplicate cultures. The averages and their ranges (vertical bars) are shown in FIG. 9C.

EX~MPLE VIII

In addition to the expression of ~ER-2/neu protein in cancer cells, The expression of ICAM-l al~o correlates with ~n~l~red diferentiation upon treatment. Briefly, AU-565 cells were treated with NDF, N29 or N28 at a co~-c~ ration of lO~g/mL in a manner similar to that de w ribed in Example III. A monoclonal antibody against ICAM-l (Becton Dir~in~cn SUBSTITUTE SHEET
.

~ ; 2145382 T~mllnocytometry Systems, San Jose, CA) was applied to treated cells. Expression of IQ M-l, HER-2/neu, cell number, percent of cells stained for lipid droplets and nuclear area (~m2) were determ;ne~ by methods described in Example III. Units for ICAM-l and HER-2/neu expression are arbitrary with "1"
representing the value of expresion in control cells as measured by image cytometry. The results are shown in Table VIII.

Table VIII

Treat-. Cell # Nuclear % Cells ICAM HER-2/neu 104/cm2 area lipid CONTROL 6 73.3 13 NDF 3.5 187 73 3.3 1.3 N29 3.1 156 66 1.7 1.4 N28 7.05 85 15 1.2 1.2 Thus, the methods of the present invention provide a powerful prognostic tool for predicting the effectiveness of a cancer therapy using monoclonal antibodies or ligands which induce differentiation of cancer cells. The methods of the present invention also provide for the screpni ng of putative anti-cancer agents for the determination of efficacy of the agent in treatment of a malignancy. Monoclonal antibodies and ligands identified in accordance with the present- invention induce the expression of mature phenotype and terminal cell differentiation, thereby inhibiting the growth af a malignancy. Additionally, the methods of the invention provide for the determination of beneficial does of, and/or improved combinations of, such therapeutic agents. Finally, the methods of the present invention are easily performed, and are therefore time andnd cost-effective, as well as minimally traumatic to a cancer patient.
... .
While the invention has been described with some specificity, modifications apparent to those with ordinary skill in the art may be made without departing from the scope of the invention.

SUBSTITUTE SHEET

Claims (25)

Claims

1. A method for determining, in vitro, the effectiveness of a therapeutic agent for treatment of a cancer, wherein malignant cells of the cancer express or overexpress an oncogene product, the method comprising the step of:
(a) obtaining viable malignant cells which express or overexpress at least one oncogene product and dividing the same into at least first and second portions;
(b) treating the first portion comprising viable malignant cells with a sufficient quantity of said agent comprising at least one compound having specific binding affinity for the oncogene product and contacting the second portion with a composition which is devoid of the compound or compounds having specific binding affinity for the oncogene product;
(c) incubating the first and second portions in a physiologically acceptable medium for an amount of time sufficient to induce a percentage of the viable malignant cells of said first portion to terminally differentiate; and (d) comparing the percentage of cells in the first portion which exhibit evidence of said terminal differentiation to the percentage of cells in the second portion which exhibit morphological evidence of terminal differentiation.

2. The method of claim 1 wherein the viable malignant cells are obtained from a patient.

3. The method of claim 2 wherein the cells are obtained from a tissue biopsy.

4. The method of claim 1 wherein the oncogene product is expressed on the extracellular domain of a cells and the agent has a binding affinity for the extracellular domain of the oncogene product.

5. The method of claim 4 wherein the oncogene product is HER-2/neu protein.

6. The method of claim 5 wherein the agent comprises at least one monoclonal antibody.

7. The method of claim 5 wherein the agent comprises at least one ligand.

8. The method of claim 1 wherein evidence of terminal differentiation is measured by translocation of the oncogene product from the surface of a cell to the cytoplasm of the cell.

9. The method of claim 1 wherein evidence of terminal differentiation is measured by changes in the expression of one or more cell adhesion molecules.

10. The method of claim 8 or 9 wherein translocation is determined immunohistochemically with one or more labeled antibodies for the oncogene product.

11. The method of claim 1 wherein morphological evidence of terminal differentiation is measured by changes in nuclear area.

12. The method of claim 9 wherein the cell adhesion molecule is ICAM-1.

13. The method of claim 1 wherein morphological evidence of terminal differentiation is measured by one or more of the changes described in claims 8, 9 or 11.

14. The method of claim 1 wherein the cancer is breast cancer and terminal differentiation is measured by production of one or more milk components.

lS. The method of claim 14 wherein the milk component is lipid droplets.

16. A method for determining, in vitro, the effectiveness of a therapeutic agent for treatment of a cancer, wherein malignant cells of the cancer express or overexpress HER-2/neu product, the method comprising the step of:
(a) obtaining viable malignant cells rom a breast, stomach, ovarian or salivary tissue biopsy and dividing the same into at least first and second portions;
(b) treating the first portion comprising viable malignant cells with a sufficient quantity of said agent comprising at least one compound having specific binding affinity for HER-2/neu product and contacting the second portion with a composition which is devoid of the compound or compounds having specific binding affinity for HER-2/neu product;
(c) incubating the first and second portions in a physiologically acceptable medium for an amount of time sufficient to induce a percentage of the viable malignant cells of said first portion to terminally differentiate; and (d) comparing the percentage of cells in the first portion which exhibit translocation of HER-2/neu product to the percentage of cells in the second portion which exhibit translocation of HER-2/neu product.

17. The method of claim 6 or 18 wherein the compound is selected from the group consisting of monoclonal antibodies produced by the hybridomas deposited as I-1260, I-1261, I-1262 and I-1263.

18. The method of claim 7 or 18 wherein the compound is selected from the group consisting of gp30 and neu differentiation factor.

19. A monoclonal antibody or fragment thereof selected from the group consisting of monoclonal antibodies produced by the hybridomas deposited as I-1260, I-1261, I-1262 and I-1263.

20. A chimeric human-murine monoclonal antibody according to claim 19.

21. A humanized human-murine monoclonal antibody according to claim 19.

22. A monoclonal antibody according to claim 19 conjugated to a toxin.

23. The antibody of claim 22 wherein the toxin is Ricin A.

24. A method for treating as cancer which expresses or overexpresses HER-2/neu product comprising the step of administering to a patient in a physiologically acceptable carrier a therapeutically effective amount of one or more monoclonal antibodies according to any of claims 19, 20, 21 or 22.

25. A monoclonal antibody or fragment thereof having substantially the same cellular reactivity and efficacy on AU-565 cells as the monoclonal antibody produced by the hybridoma deposited as I-1262.