Antibodies immunoreactive with heregulin-coupled HER3

Abstract

Antibodies which specifically bind heregulin-coupled HER3, at a site distinct from the heregulin binding site, are described. These antibodies are particularly useful in treating cancer.

Description

REFERENCE TO SUBMISSION OF SEQUENCE LISTING ON ASCII TEXT FILE

The content of the following submission on ASCII text file is incorporated herein by reference in its entirety: a computer readable form (CRF) of the Sequence Listing (file name: 388512012210CorrectedSeqList.txt, date recorded: Jul. 7, 2014, size: 9,147 bytes).

TECHNICAL FIELD

The invention relates to monoclonal antibodies with superior therapeutic value. The antibodies of the invention are particularly selective for cancer cells, and more particularly for cancer cells that secrete certain autocrine growth factors.

BACKGROUND ART

A family of receptors that mediate signaling through tyrosine kinases includes a subfamily of at least four members: HER1 also designated epidermal growth factor receptor (EGFR) is encoded by the ErbB1 gene; HER2 or HER2/neu encoded by the ErbB2 gene, HER3, encoded by the ErbB3 gene and HER4, encoded by ErbB4. Each of these receptors has a multiplicity of synonyms and in the present application, the “HER” terminology and ErbB terminology will be used interchangeably for all the factors, whether discussing protein or nucleic acids.

The primary natural ligand for the receptor encoded by ErbB3 is heregulin (HRG), a polypeptide that, when bound to HER3 induces a conformational change that promotes dimerization of HER3 with HER2 through the extracellular domains of each activating the signaling cascade, as shown in FIG. 1A. A similar dimerization is stimulated by the ligand for the HER1 receptor, Epidermal Growth Factor (EGF), which also activates a signaling cascade intracellularly.

Antibodies, including monoclonal antibodies to HER3 have been prepared. U.S. Pat. No. 5,480,968 discloses antibodies that bind specifically to ErbB3 and do not bind to ErbB2 or ErbB1. U.S. Pat. No. 5,968,511 discloses and claims antibodies that bind to HER3 and reduce heregulin-induced formation of an HER2-HER3 protein complex in cells that express both of these receptors or which antibodies increase the binding affinity of heregulin for ErbB3 protein or which reduce activation of the downstream signaling. Although only murine antibodies are prepared, humanized antibodies are also claimed.

Published U.S. application 2004/0197332 describes and claims anti-HER3 antibodies that downregulate the expression of HER3. PCT Publication WO2007/077028 describes antibodies that bind to HER3 that are produced in XenoMouse® and are thus fully human by sequence. Most of these antibodies are described as binding to the major ligand binding domain (L2) of the extracellular domain of HER3, but the binding is destroyed if the three-dimensional structure of the extracellular domain is disrupted. Further, the antibodies described in this publication compete with HRG for binding to HER3.

DISCLOSURE OF THE INVENTION

It has now been found possible to obtain antibodies that bind at a much higher affinity to HER3 when it is complexed with heregulin than to uncomplexed HER3. Such antibodies are particularly valuable since they are most effective in the context of tumor cells that secrete heregulin or analogous agonist peptides, thus stimulating the signaling cascade at a higher level than in normal cells. This enhances the specificity of the treatment to those tumor cells that are most aggressive in their invasive growth properties, while minimizing toxicity to normal cells that are not being intensively stimulated. Further, the novel antibodies are effective at blocking signaling in cells that over-express HER1 and HER3, providing a desirable “pan-HER” activity (Huang, Z, et al., Expert Opin Biol Ther. (2009) 9:97-110).

BRIEF DESCRIPTION OF THE DRAWINGS

FIGS. 1A and 1B are diagrams of the interaction of HER2 and HER3 in the presence of heregulin (FIG. 1A) and in the presence of both heregulin and either antibody 14B10 or antibody 1G4 (FIG. 1B).

FIGS. 2A and 2B are graphs showing a comparison of binding of 1G4 mAb and 14B10 mAb to the heregulin active complex with HER3 as measured by FortéBio® biosensor.

FIGS. 3A and 3B show the ability of various antibodies directed against HER3 to inhibit the growth of MCF-7 carcinoma cells and to inhibit tyrosine phosphorylation in these cells.

FIG. 4 shows the ability of various anti-HER3 antibodies to inhibit MCF-7 cell proliferation.

FIG. 5 shows that activity by the tyrosine phosphorylation measure of activity is not predictive of anti-proliferative activity.

FIG. 6 shows the ability of various antibodies directed against HER3 to inhibit tyrosine proliferation in MB 468, a cell line that over-expresses HER1 and HER3 rather than HER2 and HER3.

FIG. 7 (SEQ ID NOS:1-12) shows the nucleotide sequence encoding 1G4 heavy chain (SEQ ID NO:1) and 1G4 light chain (SEQ ID NO:7) and deduced amino acid sequences (SEQ ID NO:2 and SEQ ID NO:8).

MODES OF CARRYING OUT THE INVENTION

The present invention provides monoclonal antibodies with unique specificities for the heregulin-bound active form of HER3, thus, preferentially inhibiting dimerization and downstream signaling in environments with high concentrations of heregulin. Such high concentrations characterize many tumor cells. By targeting cells bathed in a high local concentration of stimulatory ligand, toxicities arising from blocking signaling from lower concentrations in normal cells can be minimized (Bria, E, et al., Expert Opin Biol Ther. (2008) 8:1963-1971).

The use of antibodies that preferentially bind a HER3:heregulin complex is advantageous since they are most effective on cells, such as cancer cells that are dependent on the complex signaling for growth. This type of cell is characteristic of many tumors.

Thus, the invention is directed to monoclonal antibodies and pharmaceutical compositions thereof wherein these antibodies bind to heregulin activated HER3 with an affinity at least five times their affinity for HER3 not bound to heregulin, preferably ten times greater, at all values between 5 and 10, and at values greater than 10. Preferably, the affinity of the antibody will be at least that represented by a kD of 5 nM for HER3 itself, more preferably that represented by a kD of at least 2 nM and more preferably that represented by a kD of 100 pmol.

The monoclonal antibodies may take many forms, including chimeric forms wherein the variable regions are of one species and the constant regions of another; forms consisting only of the variable regions, single-chain forms; and the like. Thus, “antibodies” refers both to whole antibodies and to fragments thereof that exhibit the required immunospecificity. In some cases, this is spelled out, but if not, fragments are intended to be included in this term unless it is otherwise obvious from context.

The antibodies of the invention can be obtained by extensive screening of hybridomas or immortalized cells of laboratory animals such as rats, mice and rabbits immunized with appropriate immunogens. Appropriate immunogens include those exemplified below, as well as portions of the HER3 protein that have been treated with heregulin. The antibodies thus produced are screened for their ability to bind heregulin-bound HER3 differentially from HER3 itself. These antibodies can then be humanized using procedures now commercially available to obtain antibodies suitable for administration to humans.

In one particular example set forth below, an antibody with this desirable differential binding affinity, 1G4, was prepared. Humanized forms and human analogs of this antibody are included within the scope of this invention.

The hybridoma that produces IG4 was deposited with the American Type Culture Collection, 10801 University Boulevard, Manassas, Va. 20110, under the terms of the Budapest Treaty on 29 Apr. 2010. Upon issuance of a U.S. patent disclosing this hybridoma, all restrictions on this deposit will be irrevocably removed.

In addition, FIG. 7 shows the nucleotide sequence and amino acid sequences of the heavy chain and light chain of 1G4 respectively. The CDR of the heavy chain are: CDRH1:GYTFTDYVS (SEQ ID NO:4); CDRH2:IYPSGRY (SEQ ID NO:5); CDRH3:TRSLQRLRYFDV (SEQ ID NO:6). The CDR of the light chain are CDRL1:RASQSISDYLH (SEQ ID NO:10); CDRL2:YGSQSIS (SEQ ID NO:11); CDRL3:QQSNSWPLT (SEQ ID NO:12).

Alternatively, laboratory animals that produce human antibodies directly can be used as subjects to produce antibodies that are human. Thus, the immunogen can be administered to animals such as the XenoMouse® that will provide directly the human antibodies desired. These antibodies are screened in a manner similar to that employed in screening hybridomas or other immortalized rodent cells.

Typical procedures simply compare the measured affinity of the various antibodies in the screen with respect to HER3 and HER3 coupled to heregulin. A variety of assays is appropriate for this, and straightforward commercially available assays include those marketed as Biacore™ and FortéBio®.

Preferably, however, the proprietary technology CellSpot™ described in U.S. Pat. No. 7,413,868 incorporated herein by reference may be used. This permits millions of hybridomas and splenocytes or lymphocytes to be screened in practical time frames.

Suitable antibodies can also be obtained directly from human beings, since individuals harboring tumors, for example, that produce heregulin:HER3 complexes will generate antibodies to these tumor antigens. In general, antibodies immunospecific for tumor antigens are produced by such subjects. See, for example, Pavoni, E., et al., BMC Biotechnology (2007) 770:1-17. In addition, patients with autoimmune diseases such as lupus make antibodies to cell antigens not necessarily associated with tumors. Because of the large number of cells that can be sampled, even very rare antibodies can be obtained in this fashion.

When suitable cells secreting the desired antibodies have been identified in the screen, the antibodies can then be produced recombinantly using well known techniques. The nucleotide sequences encoding the antibodies are obtained from cells secreting them and suitable expression constructs prepared to transfect host cells for such recombinant production. The ability to produce such antibodies recombinantly permits variants such as single chain antibodies to be produced. A variety of cells can be used for such production including insect, mammalian and plant cells, as well as microorganism cultures.

The antibodies of the invention, and fragments thereof, in particular, human and humanized forms of them, are useful in treating cancers that are associated with heregulin-stimulated signaling. These include cancers of the breast, uterus, ovary, prostate, kidney, lung, pancreas, stomach, salivary gland, colon, colon-rectal, thyroid, bladder, skin, or any cancer exhibiting heregulin-stimulated proliferation. It may be useful to evaluate the cancer to be treated for its ability to secrete heregulin in connection with conducting treatment using the antibodies of the invention. This may be done by culturing cells from a biopsied cancer sample or by in situ testing of the tumor in vivo. In one method, this is done by contacting said tumor or biopsy thereof with the antibodies or fragments of the invention and determining the level of complex formed as compared to any complex formed with antibodies that bind to uncoupled HER3.

The subjects may be human or veterinary, including, domestic animals such as dogs and cats, farm animals such as pigs, cows and sheep; and laboratory model animals such as rats, mice and rabbits. Laboratory model animals may be particularly useful in evaluating the effect of the antibodies of the invention in the corresponding human tumors, and such use is included within the invention.

For use in treatment, the antibodies of the invention may also be coupled to cytotoxic agents and/or to anti-tumor drugs in general. Such drugs include, for example, platinum-based drugs, cell cycle inhibitors, natural products such as vincristine and various camptothecins.

For administration, the antibodies are formulated according to standard procedures for pharmaceutical compositions of antibodies such as those described in Remington's Pharmaceutical Sciences, Mack Publishing Company, Easton Pa., latest edition. These formulations may include delivery vehicles such as liposomes or micelles or may contain standard excipients, such as saline or saccharides.

Typically, the antibody compositions are administered by injection, in particular intravenous injection. However, any mode of administration that is workable with such compositions is included within the scope of the invention.

The production of the antibodies on a practical scale may include standard recombinant methods wherein the nucleotide sequences encoding the antibodies or portions thereof have been isolated and manipulated in standard procedures to obtain the active antibody compositions. Such production is typically in mammalian cell culture, insect cell culture or plant cell culture or may be by plants per se. The antibodies of the claimed characteristics are included within the scope of the inventions regardless of means of production.

The following examples are intended to illustrate but not to limit the invention.

Example 1

Production of 1G4 and Binding Assay

Seven murine hybridoma libraries were prepared by immunizing mice with 293 cells overexpressing HER3 extracellular domain (ECD) along with a polypeptide which consists essentially of the L2 domain of HER3 that had been produced in E. coli, and fusion of the immunized B cells with a murine myeloma by standard methods. About 400 million hybridoma cells were screened for immunoactivity with HER3 in the presence and absence of heregulin, of which about 700 were positive for binding to both by initial ELISA assay and about 70 by FACS analysis for binding to intact cells. The screening employs the assays in the above cited U.S. Pat. No. 7,413,868.

Of these, six of these murine antibodies were assayed quantitatively for their binding affinities to HER3 ECD according to the commercially available FortéBio® method. In this method, ECD at 10-15 μg/ml was captured onto an amine-reactive surface in 10 mM MES, pH 6.0 at 25° C. Binding capacity is tested by treating the surface with unlabeled antibody. As unlabeled antibody accumulates on the surface, the optical characteristics of the surface are changed, thus allowing measurement of mass accumulation without requiring labeling. This is a modification of the Biacore™ style analysis. The antibodies were tested at 67 nM as the highest concentration and diluted in a three-fold dilution series. Response data for the various concentrations were individually and globally fitted to determine the binding affinities. The results for binding to HER3 ECD are summarized in Table 1

TABLE 1
Binding to HER3 ECD
	kD (pM) Average	Stdev
	14B10	18	4
	1G4	2,810	2,660
	P1G1	115	89
	C27.1	281	168
	C31.1	349	185
	A28	2,140	2,500
	heregulin	165,000	111,000

As shown, 14B10 showed the highest affinity but all of the six antibodies bound in the pM or nM range.

The binding assay was repeated for two of the antibodies, 1G4 and 14B10 using the same FortéBio® method. HER3 ECD at 10-15 μg/ml was again captured onto the amine surface in the same buffer. Heregulin at 10 μg/ml was added to form the HER3:heregulin complex on the surface and the antibodies tested at 67 nM once again using three-fold dilution series and the data were fitted to determine binding constants. These are listed in Table 2

TABLE 2
Binding to HER3-Heregulin Complex
	kD (pM) Average	Stdev
	1G4	327	113
	14B10	34	15

Antibody 14B10 bound to HER3:heregulin complex at about the same affinity as to HER3. Antibody 1 G4 bound to HER3 alone with an affinity of 2.8 nM, but to the HER3:heregulin complex with an affinity of 327 pM. In both cases, the binding affinity for HER3 was tighter than that of heregulin itself which binds to HER3 ECD with an affinity of 165 nM.

A diagram contrasting the apparent mode of action of these antibodies is shown in FIG. 1B.

FIGS. 2A and 2B show a comparison between 1G4 and 14B10 in their ability to bind the complex of HER3 ECD with heregulin. As shown in FIG. 2A, 1G4 is able to bind the complex differentially, whereas 14B10 (FIG. 2B) does not.

Example 2

Effect on Tumor Cells

In addition, the ability of various antibodies to inhibit phosphorylation of tyrosine in HER3 and to inhibit cell growth in MCF-7 breast carcinoma cells cultured with heregulin was tested at various antibody concentrations.

To assess the ability of the antibodies to inhibit downstream phosphorylation, MCF-7 cells were maintained in log phase and for testing the medium was aspirated and the cells were rinsed, 2-3 ml of trypsin was added and the mixture incubated at 37° C. for 5 minutes. Trypsinization was stopped by adding 10-12 ml of growth medium and the cells were suspended for counting using a Becton Dickenson Vi-CELL™ cell counter and then plated at 10⁴ cells/well in 96-well plates at 100 μl/well. The plates were centrifuged at 960 rpm for 5 minutes and incubated at 37° C. for 4 hours, after which the medium was replaced with serum-free medium at 100 μl/well and the cells returned to the incubator for 3 days.

For the assay, purified antibodies at 1 μg/ml were added to each well and incubated at 37° C. for 1 hour, after which 20 nM heregulin 1b was added for 10 minutes to some of the wells. The media were aspirated from the cell surface and lysis buffer was added for determination of HER3 phosphorylation.

For this determination, a 96-well Greiner plate was coated with 2 μg/ml capture antibody in PBS at 100 μl/well and incubated at 4° C. overnight. The plate was then washed with PBST, blocked with 200 μl/well 3% BSA and PBS, washed twice with 300 μl PBST. Eighty μl of the cell lysate prepared in the previous paragraph was added to each well and incubated at room temperature with a rotator for 2 hours. The plate was then washed with PBST and then incubated with anti-phosphotyrosine labeled with horseradish peroxidase at 1:2000 for 2 hours. After washing, LumiGlo™ peroxidase solution KPL catalog number 54-61-00 was added at 100 μl/well and the extent of phosphorylation determined spectraphotometrically. The results are shown in FIG. 3A. As shown in these results, the IC₅₀ for 1G4 is about 10 ng/ml; whereas that for two other antibodies, E7 and C31, was much higher.

To assess effects on cell proliferation, MCF-7 cells were grown in flasks until confluent, then trypsinized and collected in growing media to stop trypsinization. The cells were then washed with starving media 3-4 times.

Sterile 96-well black clear-bottom plates were provided with 50 μl of starving media per well and incubated for 15 minutes at 37° C. The antibodies to be tested were diluted as desired in starving medium and seeded at 5,000 cells/50 μl into individual wells. The cells were incubated with antibody for 30 minutes at 37° C. and then 3 nM HRG added. The cells were then cultured for 5-6 days and cell proliferation measured by adding fluorescent dye Resazurin™ at 1:10 and incubated for 3 hours at 37° C. The plates were read on a fluorescent plate reader at 531/590 nm The results are shown in FIG. 3B.

As shown, 1G4 had an EC₅₀ between 0.1 μg/ml and 1 μg/ml; only 14B10 had a lower EC₅₀.

As shown in FIG. 4, several of the antibodies including 1 G4 were also able to block migration of cells stimulated with heregulin.

It is important to note that inhibition of tyrosine phosphorylation does not guarantee or evaluate inhibition of proliferation, as shown in FIG. 5. Prior art teachings cited above have assumed that the two activities would be tightly correlated. By surveying 400 million hybridoma cells from seven libraries, a more comprehensive view of the activities was obtained, and lack of correlation established.

Example 3

Effect on Tumor Cells Expressing HER1

The novel antibodies also inhibit tyrosine proliferation in a cell line that over-expresses HER1 and HER3, with no detectable HER2: MB468, human breast adenocarcinoma (Moasser, M. M., et al., Cancer Res (2001) 61:7184-7188). The cells were plated at 5,000 cells in 100 μl in DMEM/F12 50/50 plus 0.02% BSA-Transferrin. Under these conditions, titration of heregulin stimulation of proliferation established that 3 nM achieves 80% of the maximal stimulation. Antibody inhibition was thereby measureable along the linear dose/response part of the stimulation curve. The cells were cultured for 5 days and cell proliferation measured by adding fluorescent dye Resazurin™ as in the previous Example. The results are shown in FIG. 6. As shown, IG4, 14B10 and A28 have similar EC₅₀'s at approximately 10⁻³-10⁻² μg/ml. P1G1, however, has an EC₅₀ of about 10⁻¹ μg/ml. Activity of the antibodies was then verified at 30 nM heregulin stimulation.

Claims

1. A method to treat HER3-expressing cancer in a subject which method comprises administering to a subject in need of such treatment an effective amount of antibodies or fragments which bind to HER3 not complexed to heregulin, and which bind to HER3:heregulin complex with greater affinity as compared to binding uncomplexed HER3.

2. The method of claim 1 wherein the antibodies or fragments contain a heavy chain that comprises CDR1:GYTFTDYVIS (SEQ ID NO:4), CDR2:IYPGSGRY (SEQ ID NO:5) and CDR3:TRSLQRLRYFDV (SEQ ID NO:6).

3. The method of claim 2 wherein the antibodies or fragments further contain a light chain that comprises CDR1:RASQSISDYLH (SEQ ID NO:10), CDR2:YGSQSIS (SEQ ID NO:11) and CDR3:QQSNSWPLT (SEQ ID NO:12).

4. The method of claim 1 wherein the subject is human and said antibodies or fragments are human or humanized.

5. The method of claim 2 wherein the subject is human and said antibodies or fragments are human or humanized.

6. The method of claim 3 wherein the subject is human and said antibodies or fragments are human or humanized.

7. A method to treat HER3-expressing cancer in a subject which method comprises administering to a subject in need of such treatment an effective amount of a conjugate of a monoclonal antibody or fragment thereof which binds to HER3 not complexed to heregulin, and binds to HER3:heregulin complex with greater affinity as compared to binding uncomplexed HER3, said antibody or fragment coupled to a cytotoxic or chemotherapeutic agent.

8. The method of claim 7 wherein the antibody or fragment contains a heavy chain that comprises CDR1:GYTFTDYVIS (SEQ ID NO:4), CDR2:IYPGSGRY (SEQ ID NO:5) and CDR3:TRSLQRLRYFDV (SEQ ID NO:6).

9. The method of claim 8 wherein the antibody or fragment further contains a light chain that comprises CDR1:RASQSISDYLH (SEQ ID NO:10), CDR2:YGSQSIS (SEQ ID NO:11) and CDR3:QQSNSWPLT (SEQ ID NO:12).

10. The method of claim 7 wherein the subject is human and said antibody or fragment is human or humanized.

11. The method of claim 8 wherein the subject is human and said antibody or fragment is human or humanized.

12. The method of claim 9 wherein the subject is human and said antibody or fragment is human or humanized.