Title of Invention

"AN ANTI CA6 GLYCOTOPE ANTIBODY OR EPITOPE BINDING FRAGMENT THEREOF"

Abstract A cell binding agent such as a monoclonal antibody and epitope-binding fragments thereof, including humanized or resurfaced versions of DS6, improved versions of DS6, polynucleotides encoding said agent and vectors encoding and expressing said agent, wherein said agent recognizes and binds the CA6 glycotope.
Full Text FIELD OF THE INVENTION
[02] The present invention is directed to a murine anti-CA6 glycotope monoclonal antibody, and humanized or resurfaced versions thereof. The present invention is also directed to epitope-binding fragments of the anti-CA6 glycotope monoclonal antibody, as well as to epitope-binding fragments of humanized or resurfaced versions of the anti-CA6 glycotope monoclonal antibody.
[03] The present invention is further directed to cytotoxic conjugates comprising a cell binding agent and a cytotoxic agent, therapeutic compositions comprising the conjugate, methods for using the conjugates in the inhibition of cell growth and the treatment of disease, and a kit comprising the cytotoxic conjugate. In particular, the cell binding agent is a monoclonal antibody, or epitope-binding fragment thereof, that recognizes and binds the CA6 glycotope or a humanized or resurfaced version thereof.
BACKGROUND OF THE INVENTION
[04] There have been numerous attempts to develop anti-cancer therapeutic agents that specifically destroy target cancer cells without harming surrounding, noncancerous cells and tissue. Such therapeutic agents have the potential to vastly improve the treatment of cancer in human patients.

[05] One promising approach has been to link cell binding agents, such as monoclonal antibodies, with cytotoxic drugs (Sela et al, in Immunoconjugates 189-216 (C. Vogel, ed. 1987); Ghose et al, in Targeted Drugs 1-22 (E. Goldberg, ed. 1983); Diener et al, in Antibody mediated delivery systems 1-23 (J. Rodwell, ed. 1988); Pietersz et al, in Antibody mediated delivery systems 25-53 (J. Rodwell, ed. 1988); Bumol et al, in Antibody mediated delivery systems 55-79 (J. Rodwell, ed. 1988). Depending on the selection of the cell binding agent, these cytotoxic conjugates can be designed to recognize and bind only specific types of cancerous cells, based on the expression profile of molecules expressed on the surface of such cells.
[06] Cytotoxic drugs such as methotrexate, daunorubicin, doxorubicin, vincristine, vinblastine, melphalan, mitomycin C, and chlorambucil have been used in such cytotoxic conjugates, linked to a variety of murine monoclonal antibodies. In some cases, the drug molecules were linked to the antibody molecules through an intermediary carrier molecule such as serum albumin (Garnett et al, 46 Cancer Res. 2407-2412 (1986); Ohkawa et al 23 Cancer Immunol. Immunother. 81-86 (1986); Endo et al, 47 Cancer Res. 1076-1080 (1980)), dextran (Hurwitz et al, 2 Appl. Biochem. 25-35 (1980); Manabi et al, 34 Biochem. Pharmacol. 289-291 (1985); Dillman et al, 46 Cancer Res. 4886-4891 (1986); Shoval et al, 85 Proc. Natl. Acad. Sci. 8276-8280 (1988)), or polyglutamic acid (Tsukada et al, 73 J. Natl. Cane. Inst. 721-729 (1984); Kato et al 27 J. Med. Chem. 1602-1607 (1984); Tsukada et al, 52 Br. J. Cancer 111-116 (1985)).
[07] As an example of one specific conjugate that has shown some promise, is the conjugate of the C242 antibody, directed against CanAg, an antigen expressed on colorectal and pancreatic tumors, and the maytansine derivative DM1 (Liu et al., Proc Natl Acad Sci USA, 93: 8618-8623 (1996)). In vitro evaluation of this conjugate

indicated that its binding affinity towards CanAg expressed on the cell surface was high with an apparent Kd value of 3 x 10-11 M, and its cytotoxic potency for CanAg-positive cells was high with an IC50 of 6 x 10-11 M. This cytotoxicity was antigen-dependent since it was blocked by an excess of non-conjugated antibody, and since antigen-negative cells were more than 100-fold less sensitive to the conjugate. Other examples of antibody-DMl conjugates with both high affinity towards respective target cells and high antigen-selective cytotoxicity include those of huN901, a humanized version of antibody against human CD56; huMy9-6, a humanized version of antibody against human CD33; huC242, a humanized version of antibody against the CanAg Mucl epitope; huJ591, a deimmunized antibody against PSMA; trastuzumab, a humanized antibody against Her2/neu; and bivatuzumab, a humanized antibody against CD44v6. [08] The development of additional cytotoxic conjugates that specifically recognize particular types of cancerous cells will be important in the continuing improvement of methods used to treat patients with cancer.
[09] To that end, the present invention is directed to the development of antibodies that recognize and bind molecules/receptors expressed on the surface of cancerous cells, and to the development of novel cytotoxic conjugates comprising cell binding agents, such as antibodies, and cytotoxic agents that specifically target the molecules/receptors expressed on the surface of cancerous cells. [10] More specifically, the present invention is directed to the characterization of a novel CA6 sialoglycotope on the Mucl mucin receptor expressed by cancerous cells, and to the provision of antibodies, preferably humanized antibodies, that recognize the novel CA6 sialoglycotope of the Mucl mucin and that may be used to inhibit the growth of a cell expressing the CA6 glycotope in the context of a cytotoxic agent.

SUMMARY OF THE INVENTION
[11] The present invention includes antibodies that specifically recognize and bind a novel CA6 sialoglycotope of the Mucl mucin receptor, or an epitope-binding fragment thereof. In another embodiment, the present invention includes a humanized antibody, or an epitope-binding fragment thereof, that recognizes the novel CA6 sialoglycotope ("the CA6 glycotope") of the Mucl mucin receptor.
[12] In preferred embodiments, the present invention includes the murine anti-CA6 monoclonal antibody DS6 ("the DS6 antibody"), and resurfaced or humanized versions of the DS6 antibody wherein surface-exposed residues of the antibody, or its epitope-binding fragments, are replaced in both light and heavy chains to more closely resemble known human antibody surfaces. The humanized antibodies and epitope-binding fragments thereof of the present invention have improved properties in that they are much less immunogenic (or completely non-immunogenic) in human subjects to which they are administered than fully murine versions. Thus, the humanized DS6 antibodies and epitope-binding fragments thereof of the present invention specifically recognize a novel sialoglycotope on the Mucl mucin receptor, i.e., the CA6 glycotope, while not being immunogenic to a human. The humanized antibodies and epitope-binding fragments thereof can be conjugated to a drug, such as a maytansinoid, to form a prodrug having specific cytotoxicity towards antigen-expressing cells by targeting the drug to the Mucl CA6 sialoglycotope. Cytotoxic conjugates comprising such antibodies and small, highly toxic drugs (e.g., maytansinoids, taxanes, and CC-1065 analogs) can thus be used as a therapeutic for treatment of tumors, such as breast and ovarian tumors.

[13] The humanized versions of the DS6 antibody of the present invention are fully characterized herein with respect to their respective amino acid sequences of both light and heavy chain variable regions, the DNA sequences of the genes for the light and heavy chain variable regions, the identification of the CDRs, the identification of their surface amino acids, and disclosure of a means for their expression in recombinant form.
[14] In one embodiment, there is provided a humanized DS6 antibody or an epitope-binding fragment thereof having a heavy chain including CDRs having amino acid sequences represented by SEQ ID NOS:l-3:
S Y N M H (SEQ ID NO : 1),
YIYPGNGATNYNQKFKG (SEQIDNO:2),
GDSVPFAY (SEQ ID NO : 3), and having a light chain that comprises CDRs having amino acid sequences represented by SEQ ID NOS:4-6:
SAHSSVSFMH (SEQ ID NO: 4),
STSSLAS (SEQ ID NO: 5),
QQRSSFPLT (SEQ ID NO: 6), [15] Also provided are humanized DS6 antibodies and epitope-binding fragments thereof having a light chain variable region that has an amino acid sequence that shares at least 90% sequence identity with an amino acid sequence represented by SEQ ID NO:7orSEQIDNO: 8:
QIVLTQSPAIMSASPGEKVTITCSAHSSVSFMHWFQQKPGTSPKLWIYST
SSLASGVPARFGGSGSGTSYSLTISRMEAEDAATYYCQQRSSFPLTFGAG
TKLELKR (SEQ ID NO:7)

EIVLTQSPATMSASPGERVTITCSAHSSVSFMHWFQQKPGTSPKLWIYST SSLASGVPARFGGSGSGTSYSLTISSMEAEDAATYYCQQRSSFPLTFGAG TKLELKR (SEQ ID N0:8) [16] Similarly, there are provided humanized DS6 antibodies and epitope-binding fragments thereof having a heavy chain variable region that has an amino acid sequence that shares at least 90% sequence identity with an amino acid sequence represented by SEQ ID NO:9, SEQ ID NO: 10, or SEQ ID NO: 11:
QAYLQQSGAELVRSGASVKMSCKASGYTFTSYNMHWVKQTPGQGLE WIGYIYPGNGATNYNQKFKGKATLTADPSSSTAYMQISSLTSEDSAVYF CARGDSVPFAYWGQGTLVTVSA (SEQ ID NO:9)
QAQLQVSGAEVVKPGASVKMSCKASGYTFTSYNMHWVKQTPGQGLE WIGYIYPGNGATNYNQKFQGKATLTADTSSSTAYMQISSLTSEDSAVYF CARGDSVPFAYWGQGTLVTVSA (SEQ ID NO: 10) QAQLQVSGAEVVKPGASVKMSCKASGYTFTSYNMHWVKQTPGQGLE WIGYIYPGNGATNYNQKFQGKATLTADPSSSTAYMQISSLTSEDSAVYF CARGDSVPFAYWGQGTLVTVSA (SEQ ID NO:l 1) [17] In another embodiment, humanized DS6 antibodies and epitope-binding fragments thereof are provided having a humanized or resurfaced light chain variable region having an amino acid sequence corresponding to SEQ ID NO: 8
EIVLTQSPATMSASPGERVTITCSAHSSVSFMHWFQQKPGTSPKLWIYST
SSLASGVPARFGGSGSGTSYSLTISSMEAEDAATYYCQQRSSFPLTFGAG
TKLELKR. (SEQIDNO:8)
[18] Similarly, humanized DS6 antibodies and epitope-binding fragments thereof are
provided having a humanized or resurfaced heavy chain variable region having an
amino acid sequence corresponding to SEQ ID NO: 10 or SEQ ID NO: 11, respectively:

QAQLQVSGAEVVKPGASVKMSCKASGYTFTSYNMHWVKQTPGQGLE WIGYIYPGNGATNYNQKFQGKATLTADTSSSTAYMQISSLTSEDSAVYF CARGDSVPFAYWGQGTLVTVSA. (SEQ ID NO: 10) QAQLQVSGAEVVKPGASVKMSCKASGYTFTSYNMHWVKQTPGQGLE WIGYIYPGNGATNYNQKFQGKATLTADPSSSTAYMQISSLTSEDSAVYF CARGDSVPFAYWGQGTLVTVSA. (SEQ ID NO:l 1) [19] The humanized DS6 antibodies and epitope-binding fragments thereof of the present invention can also include substitution in light and/or heavy chain amino acid residues at one or more positions defined by the starred residues in Table 1 which represent the murine surface framework residues found within 5 Angstroms of a CDR requiring change to a human residue. For example, the first amino acid residue Q in the murine sequence (SEQ ID NO:7) has been replaced by E (SEQ ID NO:8) to humanize the antibody. However, because of the proximity of this residue to a CDR, a back mutation to the murine residue Q may be required to maintain antibody affinity.
Table 1
(Table Removed)

[20] This is further shown in Table 2 where muDS6 variable region surface residues are shown aligned with the three most homologous human variable region surface residues. The amino acid residues in Table 1 correspond to the underlined amino acid residues in Table 2.

Table 2
(Table Removed)


[21] The present invention further provides cytotoxic conjugates comprising (1) a cell binding agent that recognizes and binds the CA6 glycotope, and (2) a cytotoxic agent. In the cytotoxic conjugates, the cell binding agent has a high affinity for the CA6 glycotope and the cytotoxic agent has a high degree of cytotoxicity for cells expressing the CA6 glycotope, such that the cytotoxic conjugates of the present invention form effective killing agents.
[22] In a preferred embodiment, the cell binding agent is an anti-CA6 antibody or an epitope-binding fragment thereof, more preferably a humanized anti-CA6 antibody or an epitope-binding fragment thereof, wherein a cytotoxic agent is covalently attached, directly or via a cleavable or non-cleavable linker, to the antibody or epitope-binding fragment thereof. In more preferred embodiments, the cell binding agent is the humanized DS6 antibody or an epitope-binding fragment thereof, and the cytotoxic agent is a taxol, a maytansinoid, CC-1065 or a CC-1065 analog. [23] In preferred embodiments of the invention, the cell binding agent is a humanized anti-CA6 antibody and the cytotoxic agent is a cytotoxic drug such as a maytansinoid or a taxane.

[24] More preferably, the cell binding agent is the humanized anti-CA6 antibody
DS6 and the cytotoxic agent is a maytansine compound, such as DM1 or DM4.
[25] The present invention also includes a method for inhibiting the growth of a cell
expressing the CA6 glycotope. In preferred embodiments, the method for inhibiting
growth of the cell expressing the CA6 glycotope takes place in vivo and results in the
death of the cell, although in vitro and ex vivo applications are also included.
[26] The present invention also provides a therapeutic composition comprising the
cytotoxic conjugate, and a pharmaceutically acceptable carrier or excipient.
[27] The present invention further includes a method of treating a subject having
cancer using the therapeutic composition. In preferred embodiments, the cytotoxic
conjugate comprises an anti-CA6 antibody and a cytotoxic agent. In more preferred
embodiments, the cytotoxic conjugate comprises a humanized DS6 antibody-DMl
conjugate, humanized DS6 antibody-DM4 or a humanized DS6 antibody-taxane
conjugate, and the conjugate is administered along with a pharmaceutically acceptable
carrier or excipient.
[28] The present invention also includes a kit comprising an anti-CA6 antibody-
cytotoxic agent conjugate and instructions for use. In preferred embodiments, the anti-
CA6 antibody is the humanized DS6 antibody, the cytotoxic agent is a maytansine
compound, such as DM1 or DM4, or a taxane, and the instructions are for using the
conjugates in the treatment of a subject having cancer. The kit may also include
components necessary for the preparation of a pharmaceutically acceptable
formulation, such a diluent if the conjugate is in a lyophilized state or concentrated
form, and for the administration of the formulation.
[29] The present invention also includes derivatives of antibodies that specifically
bind and recognize the CA6 glycotope. In preferred embodiments, the antibody

derivatives are prepared by resurfacing or humanizing antibodies that bind the CA6 glycotope, wherein the derivatives have decreased immunogenicity toward the host. [30] The present invention further provides for humanized antibodies or fragments thereof that are further labeled for use in research or diagnostic applications. In preferred embodiments, the label is a radiolabel, a fluorophore, a chromophore, an imaging agent or a metal ion.
[31] A method for diagnosis is also provided in which said labeled humanized antibodies or epitope-binding fragments thereof are administered to a subject suspected of having a cancer, and the distribution of the label within the body of the subject is measured or monitored.
[32] The present invention also provides methods for the treatment of a subject having a cancer by administering a humanized antibody conjugate of the present invention, either alone or in combination with other cytotoxic or therapeutic agents. The cancer can be one or more of, for example, breast cancer, colon cancer, ovarian carcinoma, endometrial cancer, osteosarcoma, cervical cancer, prostate cancer, lung cancer, synovial carcinoma, pancreatic cancer, a sarcoma or a carcinoma in which CA6 is expressed or other cancer yet to be determined in which CA6 glycotope is expressed predominantly.
[33] Unless otherwise stated, all references and patents cited herein are incorporated by reference.
BRIEF DESCRIPTION OF THE DRAWINGS
[34] Figure 1 shows the results of studies performed to determine the ability of the DS6 antibody to bind the surface of selected cancer cell lines. The fluorescence of cell lines incubated with the DS6 primary antibody and FITC conjugated anti-mouse IgG(H+L) secondary antibodies was measured by flow cytometry. The DS6 antibody

bound Caov-3 (Figure 1A) and T-47D (Figure IB) cells with an apparent Kd of 1.848 nM and 2.586 nM respectively. Antigen negative cell lines, SK-OV-3 (Figure IC) and Colo205 (Figure ID) demonstrated no antigen specific binding. [35] Figure 2 shows the results of dot blot analysis of epitope expression. Caov-3 (Figure 2A & Figure 2B), SKMEL28 (Figure 2C), and Colo205 (Figure 2D) cell lysates were individually spotted onto nitrocellulose membranes and then incubated individually with pronase, proteinase K, neuraminidase or periodic acid. The membranes were then immunoblotted with the DS6 antibody (Figure 2A), the CM1 antibody (Figure 2B), the R24 antibody (Figure 2C), or the C242 antibody (Figure 2D). [36] Figure 3 shows the results of a dot blot analysis of DS6 antigen expression. Caov-3 cell lysates were individually spotted onto PVDF membranes and then incubated in the presence of trifluoromethanesulfonic acid (TFMSA). The membranes were then immunoblotted with the CM1 antibody (1 & 2) or the DS6 antibody (3 & 4). [37] Figure 4 shows the results of glycotope analysis of the DS6 antigen. Caov-3 lysates pretreated with N-glycanase ("N-gly"), O-glycanase ("O-gly"), and/or sialidase ("S") were spotted onto nitrocellulose and then immunoblotted with the DS6 antibody or the CM1 antibody (Muc-1 VNTR).
[38] Figure 5 shows the results of western blot analysis of the DS6 antigen. Cell lysates were immunoprecipitated ("IP") and immunoblotted with the DS6 antibody. The antigen corresponds to a >250 kDa protein band observed in antigen-positive Caov-3 (Figure 5A and Figure 5B) and T47D (Figure 5C) cells. Antigen negative SK-OV-3 (Figure 5D) and Colo205 (Figure 5E) cell lines do not exhibit this band. After immunopreciptation, the Protein G beads of the Caov-3 cell lysates were incubated with (Figure 5A) neuraminidase ("N") or (Figure 5B) periodic acid ("PA"). Antibody ("a"), pre-IP ("Lys") and post-IP flow-through ("FT") lysate controls were run on the

same gel. Caov-3 immunoprecipitates were also incubated with N-glycanase ("N-gly"),
O-glycanase ("O-gly"), and/or sialidase ("S") (see Figure 5F), where the blot was
alternatively probed with biotinylated-DS6 and strepavidin-HRP.
[39] Figure 6 shows the results of immunoprecipitations and/or immunoblots of the
DS6 antibody and the CMl antibody on Caov-3 (Figure 6A) and HeLa (Figure 6B) cell
lysates. Overlapping CMl and DS6 western blot signals signify that the DS6 antigen is
on the Mucl protein. In HeLa lysates, the Mucl doublet results from Mucl expression
directed by distinct alleles differing in their number of tandem repeats.
[40] Figure 7 shows a DS6 antibody sandwich ELISA design (Figure 7A) and a
standard curve (Figure 7B). The standard curve was generated using known
concentrations of commercially available CA15-3 standards (where 1 CA15-3 unit = 1
DS6 unit).
[41] Figure 8 shows quantitative ELISA standard curves. The standard curves of the
detection antibody (streptavidin-HRP / biotin-DS6) signal (Figure 8C) were determined
using known concentrations of biotin-DS6 either captured by plated goat anti-mouse
IgG (Figure 8A) or bound directly onto the ELISA plate (Figure 8B).
[42] Figure 9 shows the cDNA and amino acid sequences of the light chain (Figure
9A) and heavy chain (Figure 9B) variable region for the murine DS6 antibody. The
three CDRs in each sequence are underlined (Kabat definitions).
[43] Figure 10 shows the light (Figure 10A) and heavy chain (Figure 10B) CDRs of
the murine DS6 antibody determined by Kabat definitions. The AbM modeling
software produces a slightly different definition for the heavy chain CDRs (Figure
IOC).
[44] Figure 11 shows the light chain ("muDS6LC") (residues 1 -95 of SEQ ID NO:7)
and heavy chain ("muDS6HC") (residues 1-98 of SEQ ID NO:9) amino acid sequences

for the murine DS6 antibody aligned with the germline sequences for the IgVicap4
(SEQ ID NO:23) and IgVh J558.41 (SEQ ID NO:24) genes. Grey indicates sequence
divergence.
[45] Figure 12 shows the ten light chain and heavy chain antibody sequences most
homologous to the murine DS6 (muDS6) light chain ("muDS6LC") and heavy chain
("muDS6HC") sequences that have solved structure files in the Brookhaven database.
Sequences are aligned in order of most to least homologous.
[46] Figure 13 shows surface accessibility data and calculations to predict which
framework residues of the murine DS6 antibody light chain variable region are surface
accessible. The positions with 25-35% average surface accessibility are marked (*??*)
and were subjected to the second round analysis. DS6 antibody light chain variable
region (Figure 13A) and heavy chain variable region (Figure 13B).
[47] Figure 14 shows the prDS6 vl.O mammalian expression plasmid map. This
plasmid was used to build and express the recombinant chimeric and humanized DS6
antibodies.
[48] Figure 15 shows amino acid sequences of murine ("muDS6") and humanized
("huDS6") (vl.O & vl.2) DS6 antibody light chain (Figure 15A) and heavy chain
(Figure 15B) variable domains.
[49] Figure 16 shows the cDNA and amino acid sequences of the light chain variable
region for the humanized DS6 antibody ("huDS6") (vl.O and vl.2).
[50] Figure 17 shows the cDNA and amino acid sequences of the heavy chain
variable region for the humanized DS6 antibody ("huDS6") vl.O (Figure 17A) and vl.2
(Figure 17B).
[51] Figure 18 shows flow cytometry binding curves of muDS6 and huDS6 clones
from an assay performed on WISH cells. The binding affinities of the (Figure 18A)

chimeric, vl.O, and vl.2 human DS6 clones (Kd = 3.15, 3.71, and 4.20 nM, respectively) are comparable to (Figure 18B) naked and biotinylated murine DS6 (Kd = 1.93 and 2.80 nM, respectively).
[52] Figure 19 shows the results of a competition binding assay of huDS6 antibodies with muDS6. (Figure 19A) WISH cells were incubated with biotin-muDS6 and streptavidin-DTAF producing a binding curve with an apparent Kd of 6.76 nM. (Figure 19B) Varying concentrations of naked muDS6, huDS6 v 1.0 and v 1.2 were combined with 2 nM of biotin-muDS6 and the strepavidin-DTAF secondary. [53] Figure 20 shows the results of a determination of the binding affinity of unconjugated DS6 antibody versus a DS6 antibody-DMl conjugate. The results demonstrated that DM1 conjugation does not adversely affect the binding affinity of the antibody. The apparent Kd of the DS6 antibody-DMl conjugate (3.902 nM) ("DS6-DM1") was slightly greater than the naked antibody (2.020 nM) ("DS6"). [54] Figure 21 shows the results of an indirect cell viability assay using the DS6 antibody in the presence or absence of the anti-mouse IgG (H+L) DM1 conjugate (2°Ab-DMl). Antigen-positive Caov-3 cells were killed in a DS6 antibody-dependent manner (IC50 = 424.9 pM) only in the presence of the secondary conjugate ("DS6 + 2°Ab-DMl").
[55] Figure 22 shows the results of a complement-dependent cytotoxicity (CDC) assay of the DS6 antibody and humanized DS6 antibody. The results demonstrated that there was no CDC mediated effect of the DS6 antibody or the humanized DS6 antibody (vl.O and vl.2) on HPAC (Figure 22A) and ZR-75-1 (Figure 22B) cells. [56] Figure 23 shows the results of an in vitro cytotoxicity assay of a DS6 antibody-DMl conjugate versus free maytansine. In a clonogenic assay, DS6 antigen-positive ovarian (Figure 23A), breast (Figure 23B), cervical (Figure 23C), and pancreatic

(Figure 23D) cancer cell lines were tested for cytotoxicity of continuous exposure to a DS6 antibody-DMl conjugate (left panels). These cell lines were similarly tested for maytansine sensitivity by a 72h exposure to free maytansine (right panels). The ovarian cancer cell lines tested were OVCAR5, TOV-21G, Caov-4 and Caov-3. The breast cancer cell lines tested were T47D, BT-20 and BT-483. The cervical cancer cell lines tested were KB, HeLa and WISH. The pancreatic cancer cell lines tested were HPAC, Hs766T and HPAF-II.
[57] Figure 24 shows the results of an in vitro cytotoxicity assay of a DS6 antibody-DMl conjugate. In a MTT cell viability assay, human ovarian (Figure 24A, Figure 24B & Figure 24C), breast (Figure 24D & Figure 24E), cervical (Figure 24F & Figure 24G), and pancreatic (Figure 24H & Figure 241) cancer cells were killed in a DS6 antibody-DMl conjugate-dependent manner. Naked DS6 did not adversely affect the growth of these cells, indicating that DM1 conjugation is required for the cytotoxicity. [58] Figure 25 A shows the results of an in vivo anti-tumor efficacy study of a DS6 antibody-DMl conjugate on established subcutaneous KB tumor xenografts. Tumor cells were inoculated on day 0, and the first treatment was given on day 6. Immunoconjugate treatments continued daily for a total of 5 doses. PBS control animals were euthanized once tumor volumes exceeded 1500 mm3. The conjugate was given at a dose of 150 or 225 ng/kg DM1, corresponding to antibody concentrations of 5.7 and 8.5 mg/kg respectively. The body weights (Figure 25B) of the mice were monitored during the course of the study.
[59] Figure 26 shows the results of an antitumor efficacy study of a DS6 antibody-DMl conjugate on established subcutaneous tumor xenografts. OVCAR5 (Figure 26A and Figure 26B), TOV-21G (Figure 26C and Figure 26D), HPAC (Figure 26E and Figure 26F), and HeLa (Figure 26G and Figure 26H) cells were inoculated on day 0,

and immunoconjugate treatments were given on day 6 and 13. PBS control animals were euthanized once tumor volumes exceeded 1000 mm3. The conjugate was given at a dose of 600 µg/kg DM1, corresponding to an antibody concentration 27.7 mg/kg. Tumor volume (Figure 26A, Figure 26C, Figure 26E, and Figure 26G) and body weight (Figure 26B, Figure 26D, Figure 26F, and Figure 26H) of the mice were monitored during the course of the study.
[60] Figure 27 shows the results of an in vivo efficacy study of a DS6 antibody-DMl conjugate on intraperitoneal OVCAR5 tumors. Tumor cells were injected intraperitoneally on day 0, and immunoconjugate treatments were given on day 6 and 13. Animals were euthanized once body weight loss exceeded 20%. [61] Figure 28 shows the flow cytometry binding curve from a study of the binding affinity of naked and taxane-conjugated DS6 antibody on HeLa cells. Taxane (MM1-202)-conjugation does not adversely affect the binding affinity of the antibody. The apparent Kd of the DS6-MM1-202 conjugate (1.24 nM) was slightly greater than the naked DS6 antibody (620 pM).
DETAILED DESCRIPTION OF THE INVENTION [62] The present invention provides, among other features, anti-CA6 monoclonal antibodies, anti-CA6 humanized antibodies, and fragments of the anti-CA6 antibodies. Each of the antibodies and antibody fragments of the present invention are designed to specifically recognize and bind the CA6 glycotope on the surface of a cell. CA6 is known to be expressed by many human tumors: 95% of serous ovarian carcinomas, 50% of endometrioid ovarian carcinomas, 50% of the neoplasms of the uterine cervix, 69% of the neoplasms of the endometrius, 80% of neoplasms of the vulva, 60% of breast carcinomas, 67% pancreatic tumors, and 48% of tumors of the urothelium, but is rarely expressed by normal human tissue.

[63] A report by Kearse et al., Int. J. Cancer 88(6):866-872 (2000) misidentified the protein on which the CA6 epitope is found as an 80 kDa protein having an N-linked carbohydrate containing the CA6 epitope when they used a hybridoma supernatant to characterize it. Using purified DS6 we have since demonstrated that the CA6 epitope is found on an O-linked carbohydrate of a greater than 250 kDa non-disulfide-linked glycoprotein. Furthermore, the glycoprotein was identified as the mucin, Mucl. Because different Mucl alleles have varying numbers of tandem repeats in the variable number tandem repeat (VNTR) domain cells often express two distinct Mucl proteins of different size (Taylor-Papadimitriou, Biochim. Biophys. Acta 1455(2-3):301-13 (1999). Because of differences in the number of repeats in the VNTR domain as well as differences in glycosylation the molecular weight of Mucl varies from cell line to cell line.
[64] The susceptibility of CA6 immunoreactivity to periodic acid indicates CA6 is a carbohydrate epitope "glycotope." The additional susceptibility of CA6 immunoreactivity to treatment with neuraminidase from Vibrio cholerae indicates that the CA6 epitope is a sialic acid dependent glycotope, thus a "sialoglycotope." [65] Details of the characterization of CA6 can be found in the Example 2 (see below). Additional details on CA6 may be found in WO 02/16401; Wennerberg et al., Am. J. Pathol. 143(4):1050-1054 (1993); Smith et al., Human Antibodies 9:61-65 (1999); Kearse et al., Int. J. Cancer 88(6):866-872 (2000); Smith et al., Int. J. Gynecol. Pathol. 20(3):260-6 (2001); and Smith et al., Appl. Immunohistochem. Mol. Morphol. 10(2): 152-8 (2002).
[66] The present invention also includes cytotoxic conjugates comprising two primary components. The first component is a cell binding agent that recognizes and binds the CA6 glycotope. The cell binding agent should recognize the CA6

sialoglycotope on Muc 1 with a high degree of specificity so that the cytotoxic
conjugates recognize and bind only the cells for which they are intended. A high
degree of specificity will allow the conjugates to act in a targeted fashion with little
side-effects resulting from non-specific binding.
[67] In another embodiment, the cell binding agent of the present invention also
recognizes the CA6 glycotope with a high degree of affinity so that the conjugates will
be in contact with the target cell for a sufficient period of time to allow the cytotoxic
drug portion of the conjugate to act on the cell, and/or to allow the conjugates sufficient
time in which to be internalized by the cell.
[68] In a preferred embodiment, the cytotoxic conjugates comprise an anti-CA6
antibody as the cell binding agent, more preferably the murine DS6 anti-CA6
monoclonal antibody. In a more preferred embodiment, the cytotoxic conjugates
comprises a humanized DS6 antibody or an epitope-binding fragment thereof. The
DS6 antibody is able to recognize CA6 with a high degree of specificity and directs the
cytotoxic agent to an abnormal cell or a tissue, such as cancer cells, in a targeted
fashion.
[69] The second component of the cytotoxic conjugates of the present invention is a
cytotoxic agent. In preferred embodiments, the cytotoxic agent is a taxol, a
maytansinoid such as DM1 or DM4, CC-1065 or a CC-1065 analog. In preferred
embodiments, the cell binding agents of the present invention are covalently attached,
directly or via a cleavable or non-cleavable linker, to the cytotoxic agent.
[70] The cell binding agents, cytotoxic agents, and linkers are discussed in more
detail below.
Cell Binding Agents

[71] The effectiveness of the compounds of the present invention as therapeutic agents depends on the careful selection of an appropriate cell binding agent. Cell binding agents may be of any kind presently known, or that become known and includes peptides and non-peptides. The cell binding agent may be any compound that can bind a cell, either in a specific or non-specific manner. Generally, these can be antibodies (especially monoclonal antibodies), lymphokines, hormones, growth factors, vitamins, nutrient-transport molecules (such as transferrin), or any other cell binding molecule or substance. [72] More specific examples of cell binding agents that can be used include:
(a) polyclonal antibodies;
(b) monoclonal antibodies;
(c) fragments of antibodies such as Fab, Fab', and F(ab')2, Fv (Parham, J. Immunol. 131:2895-2902 (1983); Spring et al. J. Immunol. 113:470-478 (1974); Nisonoff et al. Arch. Biochem. Biophys. 89:230-244 (I960));
(d) interferons (e.g. .alpha., .beta., .gamma.);
(e) lymphokines such as IL-2, IL-3, IL-4, IL-6;
(f) hormones such as insulin, TRH (thyrotropin releasing hormone), MSH (melanocyte-stimulating hormone), steroid hormones, such as androgens and estrogens;
(g) growth factors and colony-stimulating factors such as EGF, TGF-alpha, FGF, VEGF, G-CSF, M-CSF and GM-CSF (Burgess, Immunology Today 5:155-158 (1984));
(h) transferrin (O'Keefe et al. J. Biol. Chem. 260:932-937 (1985)); and (i) vitamins, such as folate.
Antibodies

[73] Selection of the appropriate cell binding agent is a matter of choice that depends upon the particular cell population that is to be targeted, but in general, antibodies are preferred if an appropriate one is available or can be prepared, more preferably a monoclonal antibody.
[74] Monoclonal antibody techniques allow for the production of extremely specific cell binding agents in the form of specific monoclonal antibodies. Particularly well known in the art are techniques for creating monoclonal antibodies produced by immunizing mice, rats, hamsters or any other mammal with the antigen of interest such as the intact target cell, antigens isolated from the target cell, whole virus, attenuated whole virus, and viral proteins such as viral coat proteins. Sensitized human cells can also be used. Another method of creating monoclonal antibodies is the use of phage libraries of scFv (single chain variable region), specifically human scFv (see e.g., Griffiths et al., U.S. Patent Nos. 5,885,793 and 5,969,108; McCafferty et al., WO 92/01047; Liming et al., WO 99/06587).
[75] A typical antibody is comprised of two identical heavy chains and two identical light chains that are joined by disulfide bonds. The variable region is a portion of the antibody heavy chains and light chains that differs in sequence among antibodies and that cooperates in the binding and specificity of each particular antibody for its antigen. Variability is not usually evenly distributed throughout antibody variable regions. It is typically concentrated within three segments of a variable region called complementarity-determining regions (CDRs) or hypervariable regions, both in the light chain and the heavy chain variable regions. The more highly conserved portions of the variable regions are called the framework regions. The variable regions of heavy and light chains comprise four framework regions, largely adopting a beta-sheet configuration, with each framework region connected by the three CDRs, which form

loops connecting the beta-sheet structure, and in some cases forming part of the beta-sheet structure. The CDRs in each chain are held in close proximity by the framework regions and, with the CDRs from the other chain, contribute to the formation of the antigen binding site of antibodies (E. A. Kabat et al. Sequences of Proteins of Immunological Interest, Fifth Edition, 1991, NTH).
[76] The constant region is a portion of the heavy chain. While not involved directly in binding an antibody to an antigen, it does exhibit various effector functions, such as participation of the antibody in antibody-dependent cellular toxicity. [77] A suitable monoclonal antibody for use in the present invention includes the murine DS6 monoclonal antibody (U.S. Patent No. 6,596,503; ATCC deposit number PTA-4449).
Humanized or Resurfaced DS6 Antibodies
[78] Preferably, a humanized anti-CA6 antibody is used as the cell binding agent of
the present invention. A preferred embodiment of such a humanized antibody is a
humanized DS6 antibody, or an epitope-binding fragment thereof.
[79] The goal of humanization is a reduction in the immunogenicity of a xenogenic
antibody, such as a murine antibody, for introduction into a human, while maintaining
the full antigen binding affinity and specificity of the antibody.
[80] Humanized antibodies may be produced using several technologies such as
resurfacing and CDR grafting. As used herein, the resurfacing technology uses a
combination of molecular modeling, statistical analysis and mutagenesis to alter the
non-CDR surfaces of antibody variable regions to resemble the surfaces of known
antibodies of the target host.

[81] Strategies and methods for the resurfacing of antibodies, and other methods for reducing immunogenicity of antibodies within a different host, are disclosed in US Patent 5,639,641 (Pedersen et al.), which is hereby incorporated in its entirety by reference. Briefly, in a preferred method, (1) position alignments of a pool of antibody heavy and light chain variable regions is generated to give a set of heavy and light chain variable region framework surface exposed positions wherein the alignment positions for all variable regions are at least about 98% identical; (2) a set of heavy and light chain variable region framework surface exposed amino acid residues is defined for a rodent antibody (or fragment thereof); (3) a set of heavy and light chain variable region framework surface exposed amino acid residues that is most closely identical to the set of rodent surface exposed amino acid residues is identified; (4) the set of heavy and light chain variable region framework surface exposed amino acid residues defined in step (2) is substituted with the set of heavy and light chain variable region framework surface exposed amino acid residues identified in step (3), except for those amino acid residues that are within 5 A of any atom of any residue of the complementarity-determining regions of the rodent antibody; and (5) the humanized rodent antibody having binding specificity is produced.
[82] Antibodies can be humanized using a variety of other techniques including CDR-grafting (EP 0 239 400; WO 91/09967; U.S. Pat. Nos. 5,530,101; and 5,585,089), veneering or resurfacing (EP 0 592 106; EP 0 519 596; Padlan E. A., 1991, Molecular Immunology 28(4/5):489-498; Studnicka G. M. et al., 1994, Protein Engineering 7(6):805-814; Roguska M.A. et al, 1994, PNAS 91:969-973), and chain shuffling (U.S. Pat. No. 5,565,332). Human antibodies can be made by a variety of methods known in the art including phage display methods. See also U.S. Pat. Nos. 4,444,887, 4,716,111, 5,545,806, and 5,814,318; and international patent application publication numbers WO

98/46645, WO 98/50433, WO 98/24893, WO 98/16654, WO 96/34096, WO 96/33735, and WO 91/10741 (said references incorporated by reference in their entireties). [83] In preferred embodiment, the present invention provides humanized antibodies or fragments thereof that recognizes a novel sialoglycotope (the CA6 glycotope) on the Mucl mucin. In another embodiment, the humanized antibodies or epitope-binding fragments thereof have the additional ability to inhibit growth of a cell expressing the CA6 glycotope.
[84] In more preferred embodiments, there are provided resurfaced or humanized versions of the DS6 antibody wherein surface-exposed residues of the antibody or its fragments are replaced in both light and heavy chains to more closely resemble known human antibody surfaces. The humanized DS6 antibodies or epitope-binding fragments thereof of the present invention have improved properties. For example, humanized DS6 antibodies or epitope-binding fragments thereof specifically recognize a novel sialoglycotope (the CA6 glycotope) on the Mucl mucin. More preferably, the humanized DS6 antibodies or epitope-binding fragments thereof have the additional ability to inhibit growth of a cell expressing the CA6 glycotope. The humanized antibody or an epitope-binding fragment thereof can be conjugated to a drug, such as a maytansinoid, to form a prodrug having specific cytotoxicity towards antigen-expressing cells by targeting the drug to the novel Mucl sialoglycotope, CA6. Cytotoxic conjugates comprising such antibodies and a small, highly toxic drug (e.g., maytansinoids, taxanes, and CC-1065 analogs) can be used as a therapeutic for treatment of tumors, such as breast and ovarian tumors.
[85] The humanized versions of the DS6 antibody are also fully characterized herein with respect to their respective amino acid sequences of both light and heavy chain variable regions, the DNA sequences of the genes for the light and heavy chain variable

regions, the identification of the CDRs, the identification of their surface amino acids, and disclosure of a means for their expression in recombinant form. [86] In one embodiment, there is provided a humanized antibody or epitope-binding fragment thereof having a heavy chain including CDRs having amino acid sequences represented by SEQ ID NOs:l-3:
SYNMH (SEQIDNO:l)
YIYPGNGATNYNQKFKG (SEQIDNO:2)
GDSVPFAY (SEQ ID NO:3) [87] When the heavy chain CDRs are determined by the AbM modeling software they are represented by SEQ ID NOs:20-22:
GYTFTSYNMH (SEQ ID NO:20)
YIYPGNGATN (SEQ ID NO:21)
GDSVPFAY (SEQ ID NO:22) [88] In the same embodiment, the humanized antibody or epitope-binding fragment thereof has a light chain that comprises CDRs having amino acid sequences represented by SEQ ID NOS:4-6:
SAHSSVSFMH (SEQ ID NO:4)
S T S S L A S (SEQ ID NO:5)
QQRSSFPLT (SEQIDNO:6) [89] Also provided are humanized antibodies and epitope-binding fragments thereof having a light chain variable region that has an amino acid sequence that shares at least 90% sequence identity with an amino acid sequence represented by SEQ ID NO:7 or SEQ ID NO:8:

QIVLTQSPAIMSASPGEKVTITCSAHSSVSFMHWFQQKPGTSPKLWIYST SSLASGVPARFGGSGSGTSYSLTISRMEAEDAATYYCQQRSSFPLTFGAG TKLELKR. (SEQ ID N0:7)
EIVLTQSPATMSASPGERVTITCSAHSSVSFMHWFQQKPGTSPKLWIYST SSLASGVPARFGGSGSGTSYSLTISSMEAEDAATYYCQQRSSFPLTFGAG TKLELKR. (SEQIDNO:8) [90] Similarly, there are provided humanized antibodies and epitope-binding fragments thereof having a heavy chain variable region that has an amino acid sequence that shares at least 90% sequence identity with an amino acid sequence represented by SEQ ID NO:9, SEQ ID NO: 10, or SEQ ID NO: 11:
QAYLQQSGAELVRSGASVKMSCKASGYTFTSYNMHWVKQTPGQGLE WIGYIYPGNGATNYNQKFKGKATLTADPSSSTAYMQISSLTSEDSAVYF CARGDSVPFAYWGQGTLVTVSA. (SEQ ID NO:9) QAQLQVSGAEVVKPGASVKMSCKASGYTFTSYNMHWVKQTPGQGLE WIGYIYPGNGATNYNQKFQGKATLTADTSSSTAYMQISSLTSEDSAVYF CARGDSVPFAYWGQGTLVTVSA. (SEQ ID NO: 10) QAQLQVSGAEVVKPGASVKMSCKASGYTFTSYNMHWVKQTPGQGLE WIGYIYPGNGATNYNQKFQGKATLTADPSSSTAYMQISSLTSEDSAVYF CARGDSVPFAYWGQGTLVTVSA. (SEQ ID NO: 11) [91] In another embodiment, humanized antibodies and epitope-binding fragments thereof are provided having a humanized or resurfaced light chain variable region having an amino acid sequence corresponding to SEQ ID NO: 8
EIVLTQSPATMSASPGERVTITCSAHSSVSFMHWFQQKPGTSPKLWIYST SSLASGVPARFGGSGSGTSYSLTISSMEAEDAATYYCQQRSSFPLTFGAG TKLELKR. (SEQIDNO:8)

[92] Similarly, humanized antibodies and epitope-binding fragments thereof are provided having a humanized or resurfaced heavy chain variable region having an amino acid sequence corresponding to SEQ ID NO: 10 or SEQ ID NO: 11:
QAQLQVSGAEVVKPGASVKMSCKASGYTFTSYNMHWVKQTPGQGLE WIGYIYPGNGATNYNQKFQGKATLTADTSSSTAYMQISSLTSEDSAVYF CARGDSVPFAYWGQGTLVTVSA. (SEQ ID NO: 10) QAQLQVSGAEVVKPGASVKMSCKASGYTFTSYNMHWVKQTPGQGLE WIGYIYPGNGATNYNQKFQGKATLTADPSSSTAYMQISSLTSEDSAVYF CARGDSVPFAYWGQGTLVTVSA. (SEQ ID NO:l 1) [93] The humanized antibodies and epitope-binding fragments thereof of the present invention can also include versions of light and/or heavy chain variable regions in which human surface amino acid residues in proximity to the CDRs are replaced by the corresponding muDS6 surface residues at one or more positions defined by the residues in Table 1 (Kabat numbering) marked with an asterisk in order to retain the binding affinity and specificity of muDS6.
Table 1
(Table Removed)


[94] The primary amino acid and DNA sequences of the DS6 antibody light and heavy chains, and of humanized versions thereof, are disclosed herein. However, the

scope of the present invention is not limited to antibodies and fragments comprising these sequences. Instead, all antibodies and fragments that specifically bind to CA6 as a unique tumor-specific glycotope on the Muc 1 receptor are included in the present invention. Preferably, the antibodies and fragments that specifically bind to CA6 also antagonize the biological activity of the receptor. More preferably, such antibodies further are substantially devoid of agonist activity. Thus, antibodies and antibody fragments of the present invention may differ from the DS6 antibody or the humanized derivatives thereof, in the amino acid sequences of their scaffold, CDRs, and/or light chain and heavy chain, and still fall within the scope of the present invention. [95] The CDRs of the DS6 antibody are identified by modeling and their molecular structures have been predicted. Again, while the CDRs are important for epitope recognition, they are not essential to the antibodies and fragments of the invention. Accordingly, antibodies and fragments are provided that have improved properties produced by, for example, affinity maturation of an antibody of the present invention. [96] The mouse light chain IgVic ap4 germline gene and heavy chain IgVh J558.41 germline gene from which DS6 was likely derived are shown in FIG. 11 aligned with the sequence of the DS6 antibody. The comparison identifies probable somatic mutations in the DS6 antibody, including several in the CDRs. [97] The sequence of the heavy chain and light chain variable region of the DS6 antibody, and the sequences of the CDRs of the DS6 antibody were not previously known and are set forth in Figures 9A and 9B. Such information can be used to
produce humanized versions of the DS6 antibody. Antibody Fragments

[98] The antibodies of the present invention include both the full length antibodies discussed above, as well as epitope-binding fragments. As used herein, "antibody fragments" include any portion of an antibody that retains the ability to bind to the epitope recognized by the full length antibody, generally termed "epitope-binding fragments." Examples of antibody fragments include, but are not limited to, Fab, Fab' and F(ab')2, Fd, single-chain Fvs (scFv), single-chain antibodies, disulfide-linked Fvs (dsFv) and fragments comprising either a VL or VH region. Epitope-binding fragments, including single-chain antibodies, may comprise the variable region(s) alone or in combination with the entirety or a portion of the following: hinge region, CH1, CH2, and CH3 domains.
[99] Such fragments may contain one or both Fab fragments or the F(ab')2 fragment. Preferably, the antibody fragments contain all six CDRs of the whole antibody, although fragments containing fewer than all of such regions, such as three, four or five CDRs, are also functional. Further, the fragments may be or may combine members of any one of the following immunoglobulin classes: IgG, IgM, IgA, IgD, or IgE, and the subclasses thereof.
[100] Fab and F(ab')2 fragments may be produced by proteolytic cleavage, using enzymes such as papain (Fab fragments) or pepsin (F(ab')2 fragments). [101] The single-chain FVs (scFvs) fragments are epitope-binding fragments that contain at least one fragment of an antibody heavy chain variable region (VH) linked to at least one fragment of an antibody light chain variable region (VL). The linker may be a short, flexible peptide selected to assure that the proper three-dimensional folding of the (VL) and (VH) regions occurs once they are linked so as to maintain the target molecule binding-specificity of the whole antibody from which the single-chain antibody fragment is derived. The carboxyl terminus of the (VL) or (VH) sequence may

be covalently linked by a linker to the amino acid terminus of a complementary (VL) or (VH) sequence.
[102] Single-chain antibody fragments of the present invention contain amino acid sequences having at least one of the variable or complementarity determining regions (CDRs) of the whole antibodies described in this specification, but are lacking some or all of the constant domains of those antibodies. These constant domains are not necessary for antigen binding, but constitute a major portion of the structure of whole antibodies. Single-chain antibody fragments may therefore overcome some of the problems associated with the use of antibodies containing a part or all of a constant domain. For example, single-chain antibody fragments tend to be free of undesired interactions between biological molecules and the heavy-chain constant region, or other unwanted biological activity. Additionally, single-chain antibody fragments are considerably smaller than whole antibodies and may therefore have greater capillary permeability than whole antibodies, allowing single-chain antibody fragments to localize and bind to target antigen-binding sites more efficiently. Also, antibody fragments can be produced on a relatively large scale in prokaryotic cells, thus facilitating their production. Furthermore, the relatively small size of single-chain antibody fragments makes them less likely to provoke an immune response in a recipient than whole antibodies.
[103] Single-chain antibody fragments may be generated by molecular cloning, antibody phage display library or similar techniques well known to the skilled artisan. These proteins may be produced, for example, in eukaryotic cells or prokaryotic cells, including bacteria. The epitope-binding fragments of the present invention can also be generated using various phage display methods known in the art. In phage display methods, functional antibody domains are displayed on the surface of phage particles

which carry the polynucleotide sequences encoding them. In particular, such phage can be utilized to display epitope-binding domains expressed from a repertoire or combinatorial antibody library (e.g., human or murine). Phage expressing an epitope-binding domain that binds the antigen of interest can be selected or identified with antigen, e.g., using labeled antigen bound or captured to a solid surface or bead. Phage used in these methods are typically filamentous phage including fd and Ml3 binding domains expressed from phage with Fab, Fv or disulfide-stabilized Fv antibody domains recombinantly fused to either the phage gene III or gene VIII protein. [104] Examples of phage display methods that can be used to make the epitope-binding fragments of the present invention include those disclosed in Brinkman et al., 1995, J. Immunol. Methods 182:41-50; Ames et al., 1995, J. Immunol. Methods 184:177-186; Kettleborough et al., 1994, Eur. J. Immunol. 24:952-958; Persic et al., 1997, Gene 187:9-18; Burton et al., 1994, Advances in Immunology 57:191-280; PCT application No. PCT/GB91/01134; PCT publications WO 90/02809; WO 91/10737; WO 92/01047; WO 92/18619; WO 93/11236; WO 95/15982; WO 95/20401; and U.S. Pat. Nos. 5,698,426; 5,223,409; 5,403,484; 5,580,717; 5,427,908; 5,750,753; 5,821,047; 5,571,698; 5,427,908; 5,516,637; 5,780,225; 5,658,727; 5,733,743 and 5,969,108; each of which is incorporated herein by reference in its entirety. [105] After phage selection, the regions of the phage encoding the fragments can be isolated and used to generate the epitope-binding fragments through expression in a chosen host, including mammalian cells, insect cells, plant cells, yeast, and bacteria, using recombinant DNA technology, e.g., as described in detail below. For example, techniques to recombinantly produce Fab, Fab' and F(ab')2 fragments can also be employed using methods known in the art such as those disclosed in PCT publication WO 92/22324; Mullinax et al., 1992, BioTechniques 12(6):864-869; Sawai et al., 1995,

AJRI 34:26-34; and Better et al., 1988, Science 240:1041-1043; said references incorporated by reference in their entireties. Examples of techniques which can be used to produce single-chain Fvs and antibodies include those described in U.S. Pat. Nos. 4,946,778 and 5,258,498; Huston et al., 1991, Methods in Enzymology 203:46-88; Shu et al., 1993, PNAS 90:7995-7999; Skerra et al., 1988, Science 240:1038-1040.
Functional Equivalents
[106] Also included within the scope of the invention are functional equivalents of the anti-CA6 antibody and the humanized anti-CA6 antibody. The term "functional equivalents" includes antibodies with homologous sequences, chimeric antibodies, artificial antibodies and modified antibodies, for example, wherein each functional equivalent is defined by its ability to bind to CA6. The skilled artisan will understand that there is an overlap in the group of molecules termed "antibody fragments" and the group termed "functional equivalents." Methods of producing functional equivalents are disclosed, for example, in PCT Application WO 93/21319, European Patent Application No. 239,400; PCT Application WO 89/09622; European Patent Application 338,745; and European Patent Application EP 332,424, which are incorporated in their respective entireties by reference.
[107] Antibodies with homologous sequences are those antibodies with amino acid sequences that have sequence homology with amino acid sequence of an anti-CA6 antibody and a humanized anti-CA6 antibody of the present invention. Preferably homology is with the amino acid sequence of the variable regions of the anti-CA6 antibody and humanized anti-CA6 antibody of the present invention. "Sequence homology" as applied to an amino acid sequence herein is defined as a sequence with at least about 90%, 91%, 92%, 93%, or 94% sequence homology, and more preferably at

least about 95%, 96%, 97%, 98%, or 99% sequence homology to another amino acid sequence, as determined, for example, by the FASTA search method in accordance with Pearson and Lipman, Proc. Natl. Acad. Sci. USA 85, 2444-2448 (1988). [108] As used herein, a chimeric antibody is one in which different portions of an antibody are derived from different animal species. For example, an antibody having a variable region derived from a murine monoclonal antibody paired with a human immunoglobulin constant region. Methods for producing chimeric antibodies are known in the art. See, e.g., Morrison, 1985, Science 229:1202; Oi et al., 1986, BioTechniques 4:214; Gillies et al., 1989, J. Immunol. Methods 125:191-202; U.S. Pat. Nos. 5,807,715; 4,816,567; and 4,816,397, which are incorporated herein by reference in their entireties.
[109] Humanized forms of chimeric antibodies are made by substituting the complementarity determining regions of, for example, a mouse antibody, into a human framework domain, e.g., see PCT Pub. No. W092/22653. Humanized chimeric antibodies preferably have constant regions and variable regions other than the complementarity determining regions (CDRs) derived substantially or exclusively from the corresponding human antibody regions and CDRs derived substantially or exclusively from a mammal other than a human.
[110] Artificial antibodies include scFv fragments, diabodies, triabodies, tetrabodies and mru (see reviews by Winter, G. and Milstein, C, 1991, Nature 349: 293-299; Hudson, P.J., 1999, Current Opinion in Immunology 11: 548-557), each of which has antigen-binding ability. In the single chain Fv fragment (scFv), the VH and VL domains of an antibody are linked by a flexible peptide. Typically, this linker peptide is about 15 amino acid residues long. If the linker is much smaller, for example 5 amino acids, diabodies are formed, which are bivalent scFv dimers. If the linker is reduced to less

than three amino acid residues, trimeric and tetrameric structures are formed that are called triabodies and tetrabodies. The smallest binding unit of an antibody is a CDR, typically the CDR2 of the heavy chain which has sufficient specific recognition and binding that it can be used separately. Such a fragment is called a molecular recognition unit or mru. Several such mrus can be linked together with short linker peptides, therefore forming an artificial binding protein with higher avidity than a single mru.
[Ill] The functional equivalents of the present application also include modified antibodies, e.g., antibodies modified by the covalent attachment of any type of molecule to the antibody. For example, modified antibodies include antibodies that have been modified, e.g., by glycosylation, acetylation, pegylation, phosphorylation, amidation, derivatization by known protecting/blocking groups, proteolytic cleavage, linkage to a cellular ligand or other protein, etc. The covalent attachment does not prevent the antibody from generating an anti-idiotypic response. These modifications may be carried out by known techniques, including, but not limited to, specific chemical cleavage, acetylation, formylation, metabolic synthesis of tunicamycin, etc. Additionally, the modified antibodies may contain one or more non-classical amino acids.
[112] Functional equivalents may be produced by interchanging different CDRs on different chains within different frameworks. Thus, for example, different classes of antibody are possible for a given set of CDRs by substitution of different heavy chains, whereby, for example, IgGl-4, IgM, IgAl-2, IgD, IgE antibody types and isotypes may be produced. Similarly, artificial antibodies within the scope of the invention may be produced by embedding a given set of CDRs within an entirely synthetic framework.

[113] Functional equivalents may be readily produced by mutation, deletion and/or insertion within the variable and/or constant region sequences that flank a particular set of CDRs, using a wide variety of methods known in the art. [114] The antibody fragments and functional equivalents of the present invention encompass those molecules with a detectable degree of binding to CA6, when compared to the DS6 antibody. A detectable degree of binding includes all values in the range of at least 10-100%, preferably at least 50%, 60% or 70%, more preferably at least 75%, 80%, 85%, 90%, 95% or 99% the binding ability of the murine DS6 antibody to CA6.
Improved Antibodies
[115] The CDRs are of primary importance for epitope recognition and antibody
binding. However, changes may be made to the residues that comprise the CDRs
without interfering with the ability of the antibody to recognize and bind its cognate
epitope. For example, changes that do not affect epitope recognition, yet increase the
binding affinity of the antibody for the epitope may be made.
[116] Thus, also included in the scope of the present invention are improved versions
of both the murine and humanized antibodies, which also specifically recognize and
bind CA6, preferably with increased affinity.
[117] Several studies have surveyed the effects of introducing one or more amino acid
changes at various positions in the sequence of an antibody, based on the knowledge of
the primary antibody sequence, on its properties such as binding and level of expression
(Yang, W. P. et al., 1995, J. Mol. Biol., 254, 392-403; Rader, C. et al., 1998, Proc. Natl.
Acad. Sci. USA, 95, 8910-8915; Vaughan, T. J. et al., 1998, Nature Biotechnology, 16,
535-539).

[118] In these studies, equivalents of the primary antibody have been generated by changing the sequences of the heavy and light chain genes in the CDR1, CDR2, CDR3, or framework regions, using methods such as oligonucleotide-mediated site-directed mutagenesis, cassette mutagenesis, error-prone PCR, DNA shuffling, or mutator-strains of E. coli (Vaughan, T. J. et al., 1998, Nature Biotechnology, 16, 535-539; Adey, N. B. et al., 1996, Chapter 16, pp. 277-291, in "Phage Display of Peptides and Proteins", Eds. Kay, B. K. et al., Academic Press). These methods of changing the sequence of the primary antibody have resulted in improved affinities of the secondary antibodies (Gram, H. et al., 1992, Proc. Natl. Acad. Sci. USA, 89, 3576-3580; Boder, E. T. et al.,
2000, Proc. Natl. Acad. Sci. USA, 97, 10701-10705; Davies, J. and Riechmann, L., 1996, Immunotechnolgy, 2, 169-179; Thompson, J. et al., 1996, J. Mol. Biol., 256, 77-88; Short, M. K. et al., 2002, J. Biol. Chem., 277, 16365-16370; Furukawa, K. et al.,
2001, J. Biol. Chem., 276, 27622-27628).
[119] By a similar directed strategy of changing one or more amino acid residues of the antibody, the antibody sequences described in this invention can be used to develop anti-CA6 antibodies with improved functions, including improved affinity for CA6. [120] Improved antibodies also include those antibodies having improved characteristics that are prepared by the standard techniques of animal immunization, hybridoma formation and selection for antibodies with specific characteristics
Cytotoxic Agents
[121] The cytotoxic agent used in the cytotoxic conjugate of the present invention
may be any compound that results in the death of a cell, or induces cell death, or in
some manner decreases cell viability. Preferred cytotoxic agents include, for example,
maytansinoids and maytansinoid analogs, taxoids, CC-1065 and CC-1065 analogs,
dolastatin and dolastatin analogs, defined below. These cytotoxic agents are conjugated

ito the antibodies, antibodies fragments, functional equivalents, improved antibodies and their analogs as disclosed herein
[122] The cytotoxic conjugates may be prepared by in vitro methods. In order to link a drug or prodrug to the antibody, a linking group is used. Suitable linking groups are well known in the art and include disulfide groups, thioether groups, acid labile groups, photolabile groups, peptidase labile groups and esterase labile groups. Preferred linking groups are disulfide groups and thioether groups. For example, conjugates can be constructed using a disulfide exchange reaction or by forming a thioether bond between the antibody and the drug or prodrug.
Mavtansinoids
[123] Among the cytotoxic agents that may be used in the present invention to form a
cytotoxic conjugate, are mavtansinoids and maytansinoid analogs. Examples of
suitable mavtansinoids include maytansinol and maytansinol analogs. Mavtansinoids
are drugs that inhibit microtubule formation and that are highly toxic to mammalian
cells.
[124] Examples of suitable maytansinol analogues include those having a modified
aromatic ring and those having modifications at other positions. Such suitable
mavtansinoids are disclosed in U.S. Patent Nos. 4,424,219; 4,256,746; 4,294,757;
4,307,016; 4,313,946; 4,315,929; 4,331,598; 4,361,650; 4,362,663; 4,364,866;
4,450,254; 4,322,348; 4,371,533; 6,333,410; 5,475,092; 5,585,499; and 5,846,545.
[125] Specific examples of suitable analogues of maytansinol having a modified
aromatic ring include:
[126] (1) C-19-dechloro (U.S. Pat. No. 4,256,746) (prepared by LAH reduction of
ansamytocin P2);

[127] (2) C-20-hydroxy (or C-20-demethyl) +/-C-19-dechloro (U.S. Pat. Nos.
4,361,650 and 4,307,016) (prepared by demethylation using Streptomyces or
Actinomyces or dechlorination using LAH); and
[128] (3) C-20-demethoxy, C-20-acyloxy (-OCOR), +/-dechloro (U.S. Pat. No.
4,294,757) (prepared by acylation using acyl chlorides).
[129] Specific examples of suitable analogues of maytansinol having modifications of
other positions include:
[130] (1) C-9-SH (U.S. Pat. No. 4,424,219) (prepared by the reaction of maytansinol
with H2S or P2S5);
[131] (2) C-l4-alkoxymethyl (demethoxy/CH2OR) (U.S. Pat. No. 4,331,598);
[132] (3) C-14-hydroxymethyl or acyloxymethyl (CH2OH or CH2OAc) (U.S. Pat. No.
4,450,254) (prepared from Nocardia);
[133] (4) C-15-hydroxy/acyloxy (U.S. Pat. No. 4,364,866) (prepared by the
conversion of maytansinol by Streptomyces);
[134] (5) C-15-methoxy (U.S. Pat. Nos. 4,313,946 and 4,315,929) (isolated from
Trewia nudiflora);
[135] (6) C-18-N-demethyl (U.S. Pat. Nos. 4,362,663 and 4,322,348) (prepared by the
demethylation of maytansinol by Streptomyces); and
[136] (7) 4,5-deoxy (U.S. Pat. No. 4,371,533) (prepared by the titanium
trichloride/LAH reduction of maytansinol).
[137] In a preferred embodiment, the cytotoxic conjugates of the present invention
utilize the thiol-containing maytansinoid (DM1), formally termed N2 -deacetyl-N2 -(3-
mercapto-l-oxopropyl)-maytansine, as the cytotoxic agent. DM1 is represented by the
following structural formula (I):
(Formula Removed)


[138] In another preferred embodiment, the cytotoxic conjugates of the present invention utilize the thiol-containing maytansinoid N2 -deacetyl-N2 (4-methyl-4-mercapto-l-oxopentyl)-maytansine as the cytotoxic agent. DM4 is represented by the following structural formula (II):
(Formula Removed)


[139] In further embodiments of the invention, other maytansines, including thiol and disulfide-containing maytansinoids bearing a mono or di-alkyl substitution on the carbon atom bearing the sulfur atom, may be used. These include a maytansinoid having, at C-3, C-14 hydroxymethyl, C-15 hydroxy, or C-20 desmethyl, an acylated amino acid side chain with an acyl group bearing a hindered sulfhydryl group, wherein the carbon atom of the acyl group bearing the thiol functionality has one or two substituents, said substituents being CH3, C2H5, linear or branched alkyl or alkenyl having from 1 to 10 carbon atoms, cyclic alkyl or alkenyl having from 3 to 10 carbon

Atoms, phenyl, substituted phenyl, or heterocyclic aromatic or heterocycloalkyl radical, and further wherein one of the substituents can be H, and wherein the acyl group has a linear chain length of at least three carbon atoms between the carbonyl functionality and the sulfur atom. [140] Such additional maytansines include compounds represented by formula (III):
(Formula Removed)



wherein:
Y' represents
(CR7CR8)1(CR9=CR1o)pC=CqA0(CR5CR6)mDu(CRn=CR12)r(C=C)sBt(CR3CR4)„CR1R2S
Z,
wherein:
R1 and R2 are each independently CH3, C2H5, linear alkyl or alkenyl having from 1 to 10 carbon atoms, branched or cyclic alkyl or alkenyl having from 3 to 10 carbon atoms, phenyl, substituted phenyl or heterocyclic aromatic or heterocyclic radical, and in addition R2 can be H;
A, B, D are cycloalkyl or cycloalkenyl having 3-10 carbon atoms, simple or substituted aryl or heterocyclic aromatic or heterocyclic radical; R3, R4, R5, R6, R7, R8, R9, R10, R11, and R12 are each independently H, CH3, C2H5, linear alkyl or alkenyl having from 1 to 10 carbon atoms, branched or cyclic alkyl or alkenyl

having from 3 to 10 carbon atoms, phenyl, substituted phenyl or heterocyclic aromatic or heterocyclic radical;
1, m, n, o, p, q, r, s, and t are each independently 0 or an integer of from 1 to 5, provided that at least two of 1, m, n, o, p, q, r, s, t and u are not zero at any one time; and
Z is H, SR or -COR, wherein R is linear alkyl or alkenyl having from 1 to 10 carbon atoms, branched or cyclic alkyl or alkenyl having from 3 to 10 carbon atoms, or simple or substituted aryl or heterocyclic aromatic or heterocyclic radical. [141] Preferred embodiments of formula (III) include compounds of formula (III) wherein:
R1is methyl, R2 is H and Z is H.
R1 and R2 are methyl and Z is H.
R1 is methyl, R2 is H, and Z is -SCH3
R1 and R2 are methyl, and Z is -SCH3. [142] Such additional maytansines also include compounds represented by formula (IV-L), (IV-D), or (IV-D,L):
(Formula Removed)

(IV-D,L)
wherein:
Y represents (CR7CR8)1(CR5CR6) m(CR3CR4)nCR1R2SZ, wherein:
R1 and R2 are each independently CH3, C2H5, linear alkyl or alkenyl having from 1 to 10 carbon atoms, branched or cyclic alkyl or alkenyl having from 3 to 10

carbon atoms, phenyl, substituted phenyl, or heterocyclic aromatic or heterocycloalkyl radical, and in addition R2 can be H;
R3, R4, R5, R6, R7 and R8 are each independently H, CH3, C2H5, linear alkyl or alkenyl having from 1 to 10 carbon atoms, branched or cyclic alkyl or alkenyl having from 3 to 10 carbon atoms, phenyl, substituted phenyl, or heterocyclic aromatic or heterocyclic radical;
1, m and n are each independently an integer of from 1 to 5, and in addition n can be 0;
Z is H, SR or -COR wherein R is linear or branched alkyl or alkenyl having from 1 to 10 carbon atoms, cyclic alkyl or alkenyl having from 3 to 10 carbon atoms, or simple or substituted aryl or heterocyclic aromatic or heterocyclic radical; and
May represents a maytansinoid which bears the side chain at C-3, C-14 hydroxymethyl, C-15 hydroxy or C-20 desmethyl.
[143] Preferred embodiments of formulas (IV-L), (IV-D) and (IV-D,L) include compounds of formulas (IV-L), (IV-D) and (IV-D,L) wherein:
R1 is methyl, R2 is H, R5, R6, R7, and Rg are each H, 1 and m are each 1, n is 0, and Z is H.
R1and R2 are methyl, R5, R6, R7, R8 are each H, 1 and m are 1, n is 0, and Z is H.
R1is methyl, R2 is H, R5, R6, R7, and R8 are each H, 1 and m are each 1, n is 0, and Z is -SCH3.
R1 and R2 are methyl, R5, R6, R7, R8 are each H, 1 and m are 1, n is 0, and Z is -SCH3. [144] Preferably the cytotoxic agent is represented by formula (IV-L).

(145] Such additional maytansines also include compounds represented by formula
(V):
(Formula Removed)



wherein:
Y represents (CR7CR8)1(CR5CR6)m(CR3CR4)nCR1R2SZ, wherein:
R1 and R2 are each independently CH3, C2H5, linear alkyl or alkenyl having from 1 to 10 carbon atoms, branched or cyclic alkyl or alkenyl having from 3 to 10 carbon atoms, phenyl, substituted phenyl or heterocyclic aromatic or heterocyclic radical, and in addition R2 can be H;
R3, R4, R5, R6, R7 and R8 are each independently H, CH3, C2H5, linear alkyl or alkenyl having from 1 to 10 carbon atoms, branched or cyclic alkyl or alkenyl having from 3 to 10 carbon atoms, phenyl, substituted phenyl, or heterocyclic aromatic or heterocyclic radical;
1, m and n are each independently an integer of from 1 to 5, and in addition n can be 0; and

Z is H, SR or -COR, wherein R is linear alkyl or alkenyl having from 1 to 10 carbon atoms, branched or cyclic alkyl or alkenyl having from 3 to 10 carbon atoms, or simple or substituted aryl or heterocyclic aromatic or heterocycloalkyl radical. [146] Preferred embodiments of formula (V) include compounds of formula (V) wherein:
Rl is methyl, R2 is H, R5, R6, R7, and R8 are each H; 1 and m are each 1; n is 0;and Z is H.
R1 and R2 are methyl; R5, R6, R7, R8 are each H, 1 and m are 1; n is 0; and Z is H.
R1 is methyl, R2 is H, R5, R6, R7, and R8 are each H, 1 and m are each 1, n is 0, and Z is -SCH3.
R1 and R2 are methyl, R5, R6, R7, R8 are each H, 1 and m are 1, n is 0, and Z is -SCH3.
[147] Such additional maytansines further include compounds represented by formula (VI-L), (VI-D), or (VI-D,L):
(Formula Removed)
wherein:
Y2 represents (CR7CR8)1(CR5CR6)m(CR5CR6)m(CR3CR4)nCR1R2SZ2 wherein:
R1 and R2 are each independently CH3, C2H5, linear alkyl or alkenyl having from 1 to 10 carbon atoms, branched or cyclic alkyl or alkenyl having from 3 to 10

barbon atoms, phenyl, substituted phenyl or heterocyclic aromatic or heterocyclic radical, and in addition R2 can be H;
R3, R4, R5, R6, R7 and R8 are each independently H, CH3, C2H5, linear cyclic alkyl or alkenyl having from 1 to 10 carbon atoms, branched or cyclic alkyl or alkenyl having from 3 to 10 carbon atoms, phenyl, substituted phenyl or heterocyclic aromatic or heterocyclic radical;
1, m and n are each independently an integer of from 1 to 5, and in addition n can be 0;
Z2 is SR or COR, wherein R is linear alkyl or alkenyl having from 1 to 10 carbon atoms, branched or cyclic alkyl or alkenyl having from 3 to 10 carbon atoms, or simple or substituted aryl or heterocyclic aromatic or heterocyclic radical; and
May is a maytansinoid. [148] Such additional maytansines also include compounds represented by formula
(VII):

(Formula Removed)

wherein:


Y2' represents
(CR7CR8)1(CR9=CR10)p(C=C)qA0(CR5CR6)mDu(CR,1=CR12)r(C=C)sBt(CR3CR4)„CR1R2
SZ2)
wherein:
R1and R2 are each independently CH3, C2H5, linear branched or alkyl or alkenyl having from 1 to 10 carbon atoms, cyclic alkyl or alkenyl having from 3 to 10 carbon atoms, phenyl, substituted phenyl or heterocyclic aromatic or heterocyclic radical, and in addition R2 can be H;
A, B, and D each independently is cycloalkyl or cycloalkenyl having 3 to 10 carbon atoms, simple or substituted aryl, or heterocyclic aromatic or heterocyclic radical;
R3, R4, R5, R6, R7, R8, R9, R1o, R11, and R12 are each independently H, CH3, C2H5, linear alkyl or alkenyl having from 1 to 10 carbon atoms, branched or cyclic alkyl or alkenyl having from 3 to 10 carbon atoms, phenyl, substituted phenyl or heterocyclic aromatic or heterocyclic radical;
1, m, n, o, p, q, r, s, t, and u are each independently 0 or an integer of from 1 to 5, provided that at least two of 1, m, n, 0, p, q, r, s, t, and u are not zero at any one time; and
Z2 is SR or -COR, wherein R is linear alkyl or alkenyl having from 1 to 10 carbon atoms, branched or cyclic alkyl or alkenyl having from 3-10 carbon atoms, or simple or substituted aryl or heterocyclic aromatic or heterocyclic radical. [149] Preferred embodiments of formula (VII) include compounds of formula (VII) wherein: Rl is methyl and R2 is H.
[150] The above-mentioned maytansinoids can be conjugated to anti-CA6 antibody DS6, or a homologue or fragment thereof, wherein the antibody is linked to the

maytansinoid using the thiol or disulfide functionality that is present on the acyl group of an acylated amino acid side chain found at C-3, C-14 hydroxymethyl, C-15 hydroxy or C-20 desmethyl of the maytansinoid, and wherein the acyl group of the acylated amino acid side chain has its thiol or disulfide functionality located at a carbon atom that has one or two substituents, said substituents being CH3, C2H5, linear alkyl or alkenyl having from 1 to 10 carbon atoms, branched or cyclic alkyl or alkenyl having from 3 to 10 carbon atoms, phenyl, substituted phenyl or heterocyclic aromatic or heterocycloalkyl radical, and in addition one of the substituents can be H, and wherein the acyl group has a linear chain length of at least three carbon atoms between the carbonyl functionality and the sulfur atom.
[151] A preferred conjugate of the present invention is the one that comprises the anti-anti-CA6 antibody DS6, or a homologue or fragment thereof, conjugated to a maytansinoid of formula (VIII):
(Formula Removed)

wherein:
Y1' represents
(CR7CR8)1(CR9=CR1o)p(C=C)qA0(CR5CR6)mDu(CR11=CR12)r(C=C)sBt(CR3CR4)nCR1R2S

wherein:
A, B, and D, each independently is cycloalkyl or cycloalkenyl having 3-10 carbon atoms, simple or substituted aryl, or heterocyclic aromatic or heterocyclic radical;
R3, R4, R5, R6, R7, R8, R9, R10, R11, and R12 are each independently H, CH3, C2H5, linear alkyl or alkenyl having from 1 to 10 carbon atoms, branched or cyclic alkyl or alkenyl having from 3 to 10 carbon atoms, phenyl, substituted phenyl or heterocyclic aromatic or heterocycloalkyl radical; and
1, m, n, o, p, q, r, s, t and u are each independently 0 or an integer of from 1 to 5, provided that at least two of 1, m, n, o, p, q, r, s t and u are non-not zero at any one time.
[152] Preferably, R1is methyl and R2 is H or R1 and R2 are methyl. [153] An even more preferred conjugate of the present invention is the one that comprises the anti-CA6 antibody DS6, or a homologue or fragment thereof, conjugated to a maytansinoid of formula (IX-L), (IX-D), or (IX-D,L):
(Formula Removed)
wherein:
wherein:
R1 and R2 are each independently CH3, C2H5, linear alkyl or alkenyl having from 1 to 10 carbon atoms, branched or cyclic alkyl or alkenyl having from 3 to 10

carbon atoms, phenyl, substituted phenyl, heterocyclic aromatic or heterocyclic radical, and in addition R2 can be H;
R3, R4, R5, R6, R7 and R8 are each independently H, CH3, C2H5, linear alkyl or alkenyl having from 1 to 10 carbon atoms, branched or cyclic alkyl or alkenyl having from 3 to 10 carbon atoms, phenyl, substituted phenyl or heterocyclic aromatic or heterocyclic radical;
1, m and n are each independently an integer of from 1 to 5, and in addition n can be 0; and
May represents a maytansinol which bears the side chain at C-3, C-14 hydroxymethyl, C-15 hydroxy or C-20 desmethyl.
[154] Preferred embodiments of formulas (IX-L), (IX-D) and (IX-D,L) include compounds of formulas (IX-L), (IX-D) and (IX-D,L) wherein:
R1 is methyl and R2 is H or R1 and R2 are methyl,
R1 is methyl, R2 is H, R5, R6, R7 and R8 are each H; 1 and m are each 1; n is 0,
R1and R2 are methyl; R5, R6, R7 and R8are each H; 1 and m are 1; n is 0. [155] Preferably the cytotoxic agent is represented by formula (IX-L). [156] An further preferred conjugate of the present invention is the one that comprises the anti-CA6 antibody DS6, or a homologue or fragment thereof, conjugated to a maytansinoid of formula (X):


(Formula Removed)
wherein the substituents are as defined for formula (IX) above.
[157] Especially preferred are any of the above-described compounds, wherein R1 is
H, R2 is methyl, R5, R6, R7 and R8 are each H, 1 and m are each 1, and n is 0.
[158] Further especially preferred are any of the above-described compounds, wherein
R1 and R2 are methyl, R5, R6, R7, R8 are each H, 1 and m are 1, and n is 0
[159] Further, the Z-aminoacyl stereoisomer is preferred.
[160] Each of the maytansinoids taught in pending U.S. patent application number
10/849,136, filed May 20, 2004, may also be used in the cytotoxic conjugate of the
present invention. The entire disclosure of U.S. patent application number 10/849,136
is incorporated herein by reference.
Disulfide-containing linking groups
[161] In order to link the maytansinoid to a cell binding agent, such as the DS6
antibody, the maytansinoid comprises a linking moiety. The linking moiety contains a
chemical bond that allows for the release of fully active maytansinoids at a particular
site. Suitable chemical bonds are well known in the art and include disulfide bonds,

acid labile bonds, photolabile bonds, peptidase labile bonds and esterase labile bonds. Preferred are disulfide bonds.
[162] The linking moiety also comprises a reactive chemical group. In a preferred embodiment, the reactive chemical group can be covalently bound to the maytansinoid via a disulfide bond linking moiety.
[163] Particularly preferred reactive chemical groups are N-succinimidyl esters and N-sulfosuccinimidyl esters.
[164] Particularly preferred maytansinoids comprising a linking moiety that contains a reactive chemical group are C-3 esters of maytansinol and its analogs where the linking moiety contains a disulfide bond and the chemical reactive group comprises a N-succinimidyl or N-sulfosuccinimidyl ester.
[165] Many positions on maytansinoids can serve as the position to chemically link the linking moiety. For example, the C-3 position having a hydroxyl group, the C-14 position modified with hydroxymethyl, the C-15 position modified with hydroxy and the C-20 position having a hydroxy group are all expected to be useful. However the C-3 position is preferred and the C-3 position of maytansinol is especially preferred. [166] While the synthesis of esters of maytansinol having a linking moiety is described in terms of disulfide bond-containing linking moieties, one of skill in the art will understand that linking moieties with other chemical bonds (as described above) can also be used with the present invention, as can other maytansinoids. Specific examples of other chemical bonds include acid labile bonds, photolabile bonds, peptidase labile bonds and esterase labile bonds. The disclosure of U.S. Patent No. 5,208,020, incorporated herein, teaches the production of maytansinoids bearing such bonds.

[167] The synthesis of maytansinoids and maytansinoid derivatives having a disulfide
moiety that bears a reactive group is described in U.S. Patent Nos. 6, 441,163 and
6,333,410, and U.S. Application No. 10/161,651, each of which is herein incorporated
by reference.
[168] The reactive group-containing maytansinoids, such as DM1, are reacted with an
antibody, such as the DS6 antibody, to produce cytotoxic conjugates. These conjugates
may be purified by HPLC or by gel-filtration.
[169] Several excellent schemes for producing such antibody-maytansinoid
conjugates are provided in U.S. Patent No. 6,333,410, and U.S. Application Nos.
09/867,598, 10/161,651 and 10/024,290, each of which is incorporated herein in its
entirety.
[170] In general, a solution of an antibody in aqueous buffer may be incubated with a
molar excess of maytansinoids having a disulfide moiety that bears a reactive group.
The reaction mixture can be quenched by addition of excess amine (such as
ethanolamine, taurine, etc.). The maytansinoid-antibody conjugate may then be
purified by gel-filtration.
[171] The number of maytansinoid molecules bound per antibody molecule can be
determined by measuring spectrophotometrically the ratio of the absorbance at 252 nm
and 280 nm. An average of 1-10 maytansinoid molecules/antibody molecule is
preferred.
[172] Conjugates of antibodies with maytansinoid drugs can be evaluated for their
ability to suppress proliferation of various unwanted cell lines in vitro. For example,
cell lines such as the human epidermoid carcinoma line A-431, the human small cell
lung cancer cell line SW2, the human breast tumor line SKBR3 and the Burkitt's
lymphoma line Namalwa can easily be used for the assessment of cytotoxicity of these

compounds. Cells to be evaluated can be exposed to the compounds for 24 hours and the surviving fractions of cells measured in direct assays by known methods. IC50 values can then be calculated from the results of the assays.
PEG-containing linking groups
[173] Maytansinoids may also be linked to cell binding agents using PEG linking
groups, as set forth in U.S. Application No. 10/024,290. These PEG linking groups are
soluble both in water and in non-aqueous solvents, and can be used to join one or more
cytotoxic agents to a cell binding agent. Exemplary PEG linking groups include
hetero-bifunctional PEG linkers that bind to cytotoxic agents and cell binding agents at
opposite ends of the linkers through a functional sulfhydryl or disulfide group at one
end, and an active ester at the other end.
[174] As a general example of the synthesis of a cytotoxic conjugate using a PEG
linking group, reference is again made to U.S. Application No. 10/024,290 for specific
details. Synthesis begins with the reaction of one or more cytotoxic agents bearing a
reactive PEG moiety with a cell-binding agent, resulting in displacement of the terminal
active ester of each reactive PEG moiety by an amino acid residue of the cell binding
agent, to yield a cytotoxic conjugate comprising one or more cytotoxic agents
covalently bonded to a cell binding agent through a PEG linking group.
Taxanes
[175] The cytotoxic agent used in the cytotoxic conjugates according to the present
invention may also be a taxane or derivative thereof.
[176] Taxanes are a family of compounds that includes paclitaxel (Taxol), a cytotoxic
natural product, and docetaxel (Taxotere), a semi-synthetic derivative, two compounds
that are widely used in the treatment of cancer. Taxanes are mitotic spindle poisons

that inhibit the depolymerization of tubulin, resulting in cell death. While docetaxel and paclitaxel are useful agents in the treatment of cancer, their antitumor activity is limited because of their non-specific toxicity towards normal cells. Further, compounds like paclitaxel and docetaxel themselves are not sufficiently potent to be used in conjugates of cell binding agents.
[177] A preferred taxane for use in the preparation of cytotoxic conjugates is the taxane of formula (XI):
(Formula Removed)

[178] Methods for synthesizing taxanes that may be used in the cytotoxic conjugates
of the present invention, along with methods for conjugating the taxanes to cell binding
agents such as antibodies, are described in detail in U.S. Patent Nos. 5,416,064,
5,475,092, 6,340,701, 6,372,738 and 6,436,931, and in U.S. Application Nos.
10/024,290, 10/144,042, 10/207,814, 10/210,112 and 10/369,563.
CC-1065 Analogues
[179] The cytotoxic agent used in the cytotoxic conjugates according to the present
invention may also be CC-1065 or a derivative thereof.
[180] CC-1065 is a potent anti-tumor antibiotic isolated from the culture broth of
Streptomyces zelensis. CC-1065 is about 1000-fold more potent in vitro than are
commonly used anti-cancer drugs, such as doxorubicin, methotrexate and vincristine

(B.K. Bhuyan et al., Cancer Res., 42, 3532-3537 (1982)). CC-1065 and its analogs are disclosed in U.S. Patent Nos. 6,372,738, 6,340,701, 5,846,545 and 5,585,499. [181] The cytotoxic potency of CC-1065 has been correlated with its alkylating activity and its DNA-binding or DNA-intercalating activity. These two activities reside in separate parts of the molecule. Thus, the alkylating activity is contained in the cyclopropapyrroloindole (CPI) subunit and the DNA-binding activity resides in the two pyrroloindole subunits.
[182] Although CC-1065 has certain attractive features as a cytotoxic agent, it has limitations in therapeutic use. Administration of CC-1065 to mice caused a delayed hepatotoxicity leading to mortality on day 50 after a single intravenous dose of 12.5 ug/kg {V. L. Reynolds et al., J. Antibiotics, XXIX, 319-334 (1986)}. This has spurred efforts to develop analogs that do not cause delayed toxicity, and the synthesis of simpler analogs modeled on CC-1065 has been described {M.A. Warpehoski et al., J. Med. Chem., 31, 590-603 (1988)}.
[183] In another series of analogs, the CPI moiety was replaced by a cyclopropabenzindole (CBI) moiety {D.L. Boger et al., J. Org. Chem., 55, 5823-5833, (1990), D.L. Boger et al., BioOrg. Med. Chem. Lett., 1,115-120 (1991)}. These compounds maintain the high in vitro potency of the parental drug, without causing delayed toxicity in mice. Like CC-1065, these compounds are alkylating agents that bind to the minor groove of DNA in a covalent manner to cause cell death. However, clinical evaluation of the most promising analogs, Adozelesin and Carzelesin, has led to disappointing results {B.F. Foster et al., Investigational New Drugs, 13, 321-326 (1996); I. Wolff et al., Clin. Cancer Res., 2,1717-1723 (1996)}. These drugs display poor therapeutic effects because of their high systemic toxicity.

[184] The therapeutic efficacy of CC-1065 analogs can be greatly improved by changing the in vivo distribution through targeted delivery to the tumor site, resulting in lower toxicity to non-targeted tissues, and thus, lower systemic toxicity. In order to achieve this goal, conjugates of analogs and derivatives of CC-1065 with cell-binding agents that specifically target tumor cells have been described {US Patents; 5,475,092; 5,585,499; 5,846,545}. These conjugates typically display high target-specific cytotoxicity in vitro, and exceptional anti-tumor activity in human tumor xenograft models in mice {R.V. J. Chari et al., Cancer Res., 55,4079-4084 (1995)}. [185] Methods for synthesizing CC-1065 analogs that may be used in the cytotoxic conjugates of the present invention, along with methods for conjugating the analogs to cell binding agents such as antibodies, are described in detail in U.S. Patent Nos. 5,475,092, 5,846,545, 5,585,499, 6,534,660 and 6,586,618 and in U.S. Application Nos. 10/116,053 and 10/265,452.
Other Drugs
[186] Drugs such as methotrexate, daunorubicin, doxorubicin, vincristine, vinblastine, melphalan, mitomycin C, chlorambucil, calicheamicin, tubulysin and tubulysin analogs, duocarmycin and duocarmycin analogs, dolastatin and dolastatin analogs are also suitable for the preparation of conjugates of the present invention. The drug molecules can also be linked to the antibody molecules through an intermediary carrier molecule such as serum albumin. Doxarubicin and Danorubicin compounds, as described, for example, in U.S. Serial No. 09/740991, may also be useful cytotoxic agents.
Therapeutic Composition
[187] The present invention also provides a therapeutic composition comprising:

(a) an effective amount of one or more cytotoxic conjugate, and
(b) a pharmaceutically acceptable carrier.
[188] Similarly, the present invention provides a method for inhibiting the growth of
selected cell populations comprising contacting target cells, or tissue containing target
cells, with an effective amount of a cytotoxic conjugate, or therapeutic agent
comprising a cytotoxic conjugate, either alone or in combination with other cytotoxic or
therapeutic agents.
[189] The present invention also comprises a method for treating a subject having
cancer using the therapeutic composition of the present invention.
[190] Cytotoxic conjugates can be evaluated for in vitro potency and specificity by
methods previously described (see, e.g., R.V.J. Chari et al, Cancer Res. 55:4079-4084
(1995)). Anti-tumor activity can be evaluated in human tumor xenograft models in
mice by methods also previously described (see, e.g., Liu et al, Proc. Natl. Acad. Sci.
93:8618-8623(1996)).
[191] Suitable pharmaceutically-acceptable carriers are well known and can be
determined by those of ordinary skill in the art as the clinical situation warrants. As
used herein, carriers include diluents and excipients.
[192] Examples of suitable carriers, diluents and/or excipients include: (1) Dulbecco's
phosphate buffered saline, pH ~ 7.4, containing or not containing about 1 mg/ml to 25
mg/ml human serum albumin, (2) 0.9% saline (0.9% w/v sodium chloride (NaCl)), and
(3) 5% (w/v) dextrose; and may also contain an antioxidant such as tryptamine and a
stabilizing agent such as Tween 20.
[193] The method for inhibiting the growth of selected cell populations can be
practiced in vitro, in vivo, or ex vivo. As used herein, inhibiting growth means slowing

the growth of a cell, decreasing cell viability, causing the death of a cell, lysing a cell and inducing cell death, whether over a short or long period of time. [194] Examples of in vitro uses include treatments of autologous bone marrow prior to their transplant into the same patient in order to kill diseased or malignant cells; treatments of bone marrow prior to its transplantation in order to kill competent T cells and prevent graft-versus-host-disease (GVHD); treatments of cell cultures in order to kill all cells except for desired variants that do not express the target antigen; or to kill variants that express undesired antigen.
[195] The conditions of non-clinical in vitro use are readily determined by one of ordinary skill in the art.
[196] Examples of clinical ex vivo use are to remove tumor cells or lymphoid cells from bone marrow prior to autologous transplantation in cancer treatment or in treatment of autoimmune disease, or to remove T cells and other lymphoid cells from autologous or allogeneic bone marrow or tissue prior to transplant in order to prevent graft versus host disease (GVHD). Treatment can be carried out as follows. Bone marrow is harvested from the patient or other individual and then incubated in medium containing serum to which is added the cytotoxic agent of the invention. Concentrations range from about 10 uM to 1 pM, for about 30 minutes to about 48 hours at about 37°C. The exact conditions of concentration and time of incubation, i.e., the dose, are readily determined by one of ordinary skill in the art. After incubation the bone marrow cells are washed with medium containing serum and returned to the patient by i.v. infusion according to known methods. In circumstances where the patient receives other treatment such as a course of ablative chemotherapy or total-body irradiation between the time of harvest of the marrow and reinfusion of the treated cells, the treated marrow cells are stored frozen in liquid nitrogen using standard medical equipment.

[197] For clinical in vivo use, the cytotoxic conjugate of the invention will be supplied as solutions that are tested for sterility and for endotoxin levels. Examples of suitable protocols of cytotoxic conjugate administration are as follows. Conjugates are given weekly for 4 weeks as an i.v. bolus each week. Bolus doses are given in 50 to 100 ml of normal saline to which 5 to 10 ml of human serum albumin can be added. Dosages will be 10 µg to 100 mg per administration, i.v. (range of 100 ng to 1 mg/kg per day). More preferably, dosages will range from 50 µg to 30 mg. Most preferably, dosages will range from 1 mg to 20 mg. After four weeks of treatment, the patient can continue to receive treatment on a weekly basis. Specific clinical protocols with regard to route of administration, excipients, diluents, dosages, times, etc., can be determined by one of ordinary skill in the art as the clinical situation warrants.
[198] Examples of medical conditions that can be treated according to the in vivo or ex vivo methods of killing selected cell populations include malignancy of any type including, for example, cancer of the lung, breast, colon, prostate, kidney, pancreas, ovary, cervix and lymphatic organs, osteosarcoma, synovial carcinoma, a sarcoma or a carcinoma in which CA6 is expressed, and other cancers yet to be determined in which CA6 glycotope is expressed predominantly; autoimmune diseases, such as systemic lupus, rheumatoid arthritis, and multiple sclerosis; graft rejections, such as renal transplant rejection, liver transplant rejection, lung transplant rejection, cardiac transplant rejection, and bone marrow transplant rejection; graft versus host disease; viral infections, such as mV infection, HIV infection, AIDS, etc.; and parasite infections, such as giardiasis, amoebiasis, schistosomiasis, and others as determined by one of ordinary skill in the art.
Kit

[199] The present invention also includes kits, e.g., comprising a described cytotoxic conjugate and instructions for the use of the cytotoxic conjugate for killing of particular cell types. The instructions may include directions for using the cytotoxic conjugates in vitro, in vivo or ex vivo.
[200] Typically, the kit will have a compartment containing the cytotoxic conjugate. The cytotoxic conjugate may be in a lyophilized form, liquid form, or other form amendable to being included in a kit. The kit may also contain additional elements needed to practice the method described on the instructions in the kit, such a sterilized solution for reconstituting a lyophilized powder, additional agents for combining with the cytotoxic conjugate prior to administering to a patient, and tools that aid in administering the conjugate to a patient.
Additional Embodiments
[201] The present invention further provides for monoclonal antibodies, humanized antibodies and epitope-binding fragments thereof that are further labeled for use in research or diagnostic applications. In preferred embodiments, the label is a radiolabel, a fluorophore, a chromophore, an imaging agent or a metal ion. [202] A method for diagnosis is also provided in which said labeled antibodies or epitope-binding fragments thereof are administered to a subject suspected of having a cancer, and the distribution of the label within the body of the subject is measured or monitored.
Examples

[203] The broad scope of this invention is best understood with reference to the following examples, which are not intended to limit the invention to specific embodiments.
Example 1: Identification of antigen positive and negative cell lines by flow cytometry binding assays
[204] Flow cytometric analysis was used to localize the DS6 epitope, CA6, to the cell surface. Human cell lines were obtained from the American Type Culture Collection (ATCC) with the exception of OVCAR5 (Kearse et al., Int. J. Cancer 88(6):866-872 (2000)), OVCAR8 and IGROV1 cells (M. Seiden, Massachusetts General Hospital). All cells were grown in RPMI1640 supplemented with 4 mM L-glutamine, 50 U/ml penicillin, 50 u.g/ml streptomycin (Cambrex Bio Science, Rockland, ME) and 10% v/v fetal bovine serum (Atlas Biologicals, Fort Collins, CO), referred hereafter as culture media. Cells were maintained in a 37°C, 5% C02 humidified incubator. [205] Cells (1-2 x 10"5 cells/well) were incubated, on ice for 3-4 h, with serially diluted concentrations of the DS6 antibody prepared in FACS buffer (2% goat serum, RPMI) into 96-well plates. The cells were spun down in a table top centrifuge at 1500 rpm for 5 min at 4°C. After removing the media, the wells were then refilled with 150 µl of FACS buffer. The wash step was then repeated. FITC-labeled goat anti-mouse IgG (Jackson Immunoresearch) was diluted 1:100 to FACS buffer and incubated with the cells for 1 h on ice. The plate was covered in foil to prevent photobleaching of the signal. After two washes, the cells were fixed with 1% formaldehyde and analyzed on a flow cytometer.
[206] Predominantly, CA6 epitope was found in cell lines of ovarian, breast, cervical, and pancreatic origin (Table 3) as predicted from the tumor

iimmunohistochemistry. However, some cell lines of other tumor types exhibited limited CA6 expression. The DS6 antibody binds with an apparent KD of 135.6 pM (in PC-3 cells, Table 3). The maximum mean fluorescence (Table 3) of the binding curves (Figure 1) in the antigen positive cell lines are suggestive of the relative antigen density.
Table 3
(Table Removed)


'average maximum relative mean fluorescence
Example 2: Characterization of DS6 Epitope
[207] The properties of the DS6 antigen, CA6, were analyzed by immunoblotting the dot blots of CA6-positive cell lysates (Caov-3) that were digested with proteolytic (pronase and proteinase K) and/or glycolytic (neuraminidase and periodic acid) treatments. For positive controls, other antibodies recognizing a variety of epitope types were tested on lysates of antigen positive cell lines (Caov-3 and CMl; Colo205 and C242; SKMEL28 and R24). CMl is an antibody recognizing a protein epitope of the variable number tandem repeat domain (VNTR) of Muc-1 and thus, provides a control for a protein epitope. C242 binds to a novel colorectal cancer specific sialic acid-dependent glycotope on Muc-1 (CanAg) which provides a control for a glycotope

tan a protein. R24 binds to the GD3 ganglioside that is specific for melanoma and thus provides a control for a glycotope on a non-protein scaffold. [208] Caov-3, Colo205, and SKMEL28 cells were plated in 15 cm tissue culture plates. Culture media (30 mL/plate) was refreshed the day before lysis. A modified RIPA buffer (50 mM Tris-HCl pH 7.6, 150 mM NaCl, 5 mM EDTA, 1% NP40, 0.25% sodium deoxycholate), protease inhibitors (PMSF, Pepstatin A, Leupeptin, and Aprotinin), and PBS were pre-chilled on ice. After the culture media was aspirated from the plates, the cells were washed twice with 10 ml of chilled PBS. All of the subsequent steps were conducted on ice and/or in a 4°C cold room. After the last wash of PBS was aspirated, the cells were lysed in 1-2 mL of lysis buffer (RIPA buffer with freshly added protease inhibitors to a final concentration of 1 mM PMSF, 1 µM Pepstatin A, 10 µg/ml Leupeptin, and 2 µg/ml Aprotinin). The lysates were scraped off of the plates using a cell lifter and triturated by pipetting the suspensions up and down (5-10 times) with an 18G needle. The lysates were rotated for 10 min and then centrifuged in a microcentrifuge at maximum (13K rpm) for 10 min. The pellets were discarded and the supernatants were then assayed using a Bradford Protein assay kit (Biorad).
[209] The lysates (2 µl) were pipetted directly onto dry 0.2 urn nitrocellulose membranes. The spots were allowed to air dry for approximately 30 min. The membrane was sectioned into pieces that each contained a single spot. Spots were incubated in the presence of pronase (1 mg/ml enzyme, 50 mM Tris pH 7.5, 5 mM CaC12), proteinase K (1 mg/ml enzyme, 50 mM Tris pH 7.5, 5 mM CaCb), neuraminidase (20 mU/ml enzyme, 50 mM sodium acetate pH 5, 5 mM CaCk, 100 µg/ml BSA) or periodic acid (20 mM, 0.5M sodium acetate pH 5) for lh at 37°C. Reagents were purchased from Roche (enzymes) and VWR (periodic acid).

Membranes were washed (5 min) in T-TBS wash buffer (0.1% Tween 20, IX TBS), blocked in blocking buffer (3% BSA, T-TBS) for 2 h at room temperature, and incubated overnight with 2 µg/ml of primary antibody (i.e. DS6, CM1, C242, R24) in blocking buffer at 4°C. The membranes were washed three times for 5 min in T-TBS and then incubated in HRP-conjugated goat anti-mouse (or human) IgG secondary antibody (Jackson Immunoresearch; 1:2000 dilution in blocking buffer) for 1 h at room temperature. The immunoblots were washed three times and developed using an ECL system (Amersham).
[210] The immunoblots (Figure 2) of the digested control lysates showed that the CM1 signal was destroyed by the proteolytic treatments while the signals of the glycolytic digests were unaffected as would be expected for an antibody recognizing a protein epitope. The C242 signal was destroyed by either the proteolytic or glycolytic treatments as would be expected for an antibody recognizing a glycotope found on a protein. The R24 signal, unaffected by the proteolytic treatments, was abolished with neuraminidase or periodate treatments as expected for an antibody recognizing a ganglioside. The DS6 immunoblot of the digested Caov-3 lysate dot blots showed signal loss upon treatment with either the proteolytic and glycolytic compounds. Thus, like C242, DS6 binds to a carbohydrate epitope on a proteinaceous core. Furthermore, the signal in the DS6 immunoblot was sensitive to neuraminidase treatment. Therefore, CA6, like CanAg, is a sialic acid-dependent glycotope. [211] In order to confirm the carbohydrate nature of CA6, Caov-3 lysate was spotted onto PVDF membrane and treated with the chemical deglycosylating agent, trifluoromethane sulfonic acid (TFMSA), under nitrogen at ambient temperature for 5 minutes. The blot was washed with T-TBS and immunoblotted with either CM1 or DS6 (Figure 3). The DS6 signal was destroyed upon the acid treatment providing

further evidence that CA6 is a glycotope. The enhancement of the CM1 signal upon TFSMA treatment indicates that the acid treatment did not affect the protein on the filter and suggests that the glycolytic treatment unmasked the protein epitope recognized by CM1.
[212] To further elucidate the structure of the carbohydrate on which CA6 resides, dot blots were digested with N-glycanase, O-glycanase, and/or sialidase (Figure 4). Caov-3 cell lysates (100 µg, 30 ul) were incubated at 100°C for 5 min with 2.5 µ1 of denaturation buffer (Glyko) containing SDS and P-mercaptoethanol. The denatured lysates were then digested with 1 ul of N-glycanase, O-glycanase, and/or Sialidase A (Glyko) at 37°C for 1 h. The digested lysates were then spotted (2 µ1) onto nitrocellulose and immunoblotted as described above.
[213] N-glycanase had no apparent effect on the DS6 immunoblot signals. However, samples digested with sialidase produced no signal. Because O-glycanase cannot digest sialyated O-linked carbohydrates without pretreatment with sialidase, the DS6 signal of samples processed with O-glycanase alone would not be affected. N-glycanase, in contrast, does not require pretreatment with any glycosidic enzymes for activity. The fact that N-glycanase treatment does not affect the DS6 signal suggests that the CA6 epitope is most likely present on sialyated O-linked carbohydrate chains.
Example 3: Elucidation of the antigen on which the CA6 epitope is found [214] To identify the antigen on which the CA6 sialoglycotope is found, DS6 immunoprecipitates were analyzed by SDS-PAGE and Western blotting. Cell lysate supematants (1 mL/sample; 3-5 mg protein) were pre-cleared with Protein G beads (30 µl), equilibrated with 1 ml of RIPA buffer, for 1-2 h, with rotation, at 4°C. All of the subsequent steps were conducted on ice and/or in a 4°C cold room. The pre-cleared

Ibeads were spun down briefly (2-3 s) in a microcentrifuge. The pre-cleared supernatants were transferred to fresh tubes and incubated overnight with 2 µg of DS6, with rotation. Fresh, equilibrated Protein G beads (30 u.1) were added to the lysates and incubated for lh, with rotation. The bead-lysate suspensions were briefly spun down in a microcentrifuge and samples of the post-immunoprecipitation lysates were optionally taken. The beads were washed 5-10 times with 1 mL of RIP A buffer. [215] Immunoprecipitated DS6 samples were then digested with 30 u,l neuraminidase (20 mU neuraminidase (Roche), 50 mM sodium acetate pH 5, 5 mM CaCb, 100 µg/ml BSA) or 30µ1 periodic acid (20 mM periodic acid (VWR), 0.5M sodium acetate pH 5) for lh at 37°C. They were then resuspended in 30 ul of 2X sample loading buffer (containing P-mercaptoethanol). The beads were boiled for 5 min and the loading buffer supernatants were loaded onto 4-12% or 4-20% Tris-Glycine gels (Invitrogen). The gels were run in Laemmli electrophoresis running buffer at 125 V for 1.5 h. The gel samples were transferred, overnight at 20 mA, onto 0.2 urn nitrocellulose membranes (Invitrogen) using a Mini Trans-blot transfer apparatus (Biorad). Membranes were immunoblotted with DS6 as described above in Example 2. [216] Alternatively, the immunoprecipitated beads were first denatured and then enzymatically digested with N-glycanase, O-glycanase and/or sialidase A (Glyko). The beads were resuspended in 27 u.1 incubation buffer and 2 u.1 denaturation solution (Glyko) and incubated at 100°C for 5 minutes. After cooling to room temperature, detergent solution (2 ul) was added and the samples were incubated with 1 ul of N-glycanase, O-glycanase, and/or Sialidase A at 37°C for 4 h. After adding 5X sample loading buffer (7 µl), the samples were boiled for 5 min. The samples were subjected to SDS-PAGE and immunoblotted as described above.




(217] DS6 immunoprecipitates a >250 kDa protein band that can be seen in antigen positive cell lysates (Figure 5A, B, and C). In some cell lines (i.e. T-47D), a doublet is observed. The >250 kDa band was abolished in Caov-3 immunoprecipitates that were treated with neuraminidase or periodic acid (Figure 5 A and B) suggesting that the CA6 epitope resides on the >250 kDa band. The >250 kDa band was also shown to be insensitive to N-glycanase treatment of immunoprecipitates consistent with CA6 residing on an O-linked carbohydrate (Figure 5F). Further supporting that the 250 kDa band is the CA6 antigen is the fact that DS6 immunoprecipitates no such band from DS6 antigen negative cells (Figure 5D and E).
[218] Several lines of evidence suggested that the CA6 antigen was Muc 1. Because of the high molecular weight and the sensitivity to O-linked carbohydrate-specific glycolytic enzymes, it seemed likely that the CA6 antigen was a mucin. Mucin overexpression is well characterized in tumors particularly of the breast and ovary, consistent with the major tumor reactivities of DS6. Furthermore, CA6, like CanAg (a sialoglycotope on Mucl), is not susceptible to perchloric acid precipitation suggesting the CA6 antigen is heavily O-glycosylated. The observation that in some DS6 expressing cell lines, DS6 immunoprecipitated a doublet of >250 kDa suggested that the CA6 was Mucl. A hallmark of Mucl in humans is the presence of two distinct Mucl alleles differing in number of tandem repeats resulting in the expression of two Mucl proteins of different molecular weights.
[219] To test whether CA6 is found on Mucl, DS6 immunoprecipitates from Caov-3 lystate were subjected to SDS-PAGE and immunoblotted with either DS6 or a Mucl VNTR antibody, CM1. As can be seen in Figure 6A, CM1 reacts strongly with the >250 kDa band immunoprecipitated by DS6. In Figure 6B, DS6 and CM1 immunoprecipitates from HeLa cell lysate show the same >250 kDa doublet when

immunoblotted with either DS6 or CM1. These results indicate that the CA6 epitope is indeed located on the Muc-1 protein. The DS6 doublet seen in HeLa (and T-47D) cells can be explained by the fact that Muc-1 expression is directed by distinct alleles having differing number of tandem repeats.
[220] Although CM1 and DS6 bind to the same Muc-1 protein, they are distinct epitopes. Chemical deglycosylation of Caov-3 lysate dot blots by trifluoromethane sulfonic acid (TFMSA) abolished the DS6 signal (Figure 3). However, this same treatment enhanced the CM1 signal. Deglycosylation may have revealed hidden epitopes for the CM1 antibody. Furthermore, a comparison of the flow cytometry binding results of DS6 and CM1 (Table 4) demonstrates that the CA6 epitope does not exist on every cell expressing Mucl. It is interesting to note that the CA6 epitope is not expressed on Colo205 (Table 3), a cell line known to express high levels of the Mucl CanAg sialoglycotope.
Table 4
(Table Removed)
MMF = maximum mean relative fluorescence
Example 4: Quantitative analysis of shed CA6 epitope
[221] Because the CA6 epitope resides on Mucl, a molecule known to be shed into the blood stream in many cancer patients, a quantitative approach was undertaken in order to determine whether such levels would be prohibitive for DS6 antibody therapy.

Binding of circulating antibody to antigen is thought to lead to rapid clearance of immune complexes from the blood. If a significant portion of the administered antibody dose is rapidly removed from circulation the amount reaching the tumor is likely to be diminished resulting in decreased anti-tumor activity of an antibody therapeutic. When the antibody is conjugated to a highly potent cytotoxic compound the rapid clearance of conjugate could potentially increase non-specific toxicity. Thus, in the case of antibody-small drug conjugates such as DS6-DM1, high levels of shed antigen might be expected to both reduce the anti-tumor effect and increase the dose-limiting toxicity.
[222] Recent clinical trials of antibody therapeutics have yielded information as to the impact of shed antigen concentration on pharmacokinetics. For example, in clinical trials with trastuzumab (Herceptin), an antibody used for the treatment of her2/neu-expressing metastatic breast cancer, the pharmacokinetics of trastuzumab clearance was shown to be unaltered when the shed Her2/neu level was less than 500 ng/mL (Pegram etal., J. Clin. Oncol. 16(8):2659-71 (1998). Assuming a molecular weight of shed Her2/neu of 110,000 Daltons, a molar concentration shed Her2/neu below 4.5 nM appears to have little influence on the pharmacokinetics.
[223] In another example, a clinical trial with cantuzumab mertansine (huC242-DMl) indicated that there was no correlation with pretreatment shed CanAg (C242 epitope) levels and pharmacokinetics of antibody clearance (Tolcher et al., J. Clin. Oncol. 21(2):211-22 (2003). The CanAg epitope, similar to the CA6 epitope recognized by DS6, is a unique tumor-specific O-linked sialoglycotope on Mucl. However, the heterogeneous nature of the CanAg epitope makes it difficult to quantify in molar terms. In the general population Mucl alleles vary in length depending upon the number of tandem repeats in the variable number tandem repeat (VNTR) domain.

Several sites for O-linked glycosylation occur in each tandem repeat. Adding to the complexity of CanAg expression is the cell-to-cell variation in inherent glycosyl transferase activity. Thus a wide range of CanAg epitopes per Mucl molecule are possible even in a single patient. Moreover, the ratio of CanAg epitope per Mucl molecule will be different across a population of patients. For this reason, shed CanAg in serum samples is measured by sandwich ELISA where shed Mucl with CanAg epitope is captured by C242 and detected by a biotinylated C242/Streptavidin HRP system. The shed CanAg is quantified in standardized units (U) proportional to the number of epitopes per ml of serum rather than by a molar concentration of Mucl. By analogy, a similar situation occurs for the quantification of shed CA6 epitopes. In contrast, for trastuzumab there is only one epitope per shed her2/neu molecule vastly simplifying the quantification of shed antigen.
[224] In order to relate CA6 shed epitope levels to those found in clinical trials with trastuzumab and cantuzumab mertansine, a method for obtaining molar concentrations of complex shed epitopes such as sialoglycotopes on Mucl was developed. First, a simple sandwich ELISA assay for DS6 was established. A representation of the assay is shown in Figure 7A. DS6 was used to capture Mucl having CA6 epitope. Because each Mucl molecule has multiple CA6 epitopes, biotinylated DS6 was also used as the tracer antibody. Biotinylated DS6 bound to captured CA6 was detected by Streptavidin-HRP using ABTS as the substrate. CA6 epitope was captured from ovarian cancer patient serum or from standards which come from a commercially available Mucl test kit (CA15-3) used to monitor shed Mucl in breast cancer patients. DS6 units/ml were arbitrarily set equal to CA15-3 standards units/ml. [225] In Figure 7B is shown the results of the DS6 sandwich ELISA in which CA15-3 standards were used. The curve generated is very similar to that obtained with CA15-3

standards in the CA15-3 assay. In order to convert DS6 unit/ml to a molar concentration of CA6 a standard curve for biotinylated DS6 which converts signal to picograms of DS6 is required. Assuming a one-to-one stoichiometry between CA6 epitope and biotinylated DS6 antibody and a molecular weight of 160,000 Daltons for biotinylated DS6 the moles of CA6 captured per volume of sample added can be computed.
[226] In Figure 8 A and B are representations of two alternative means of generating a standard curve for biotinylated DS6. In Figure 8A, Goat anti-mouse IgG polyclonal antibody is used to capture biotinylated DS6 which is in turn detected in a manner identical to that used in the sandwich ELISA assay shown in Figure 7. In the method shown in Figure 8B biotinylated DS6 is plated directly onto the ELISA plate and detected as in Figure 8A. As seen in Figure 8C the biotinylated DS6 standard curves generated by each method are in good agreement.
[227] In Table 5 the analysis of ovarian cancer patient serum samples for various shed antigens is shown. CA125 ELISA is generally used to monitor the treatment of ovarian cancer patients by measuring shed CA125 units/ml. The CA125 status was provided with the serum samples. CA15-3 ELISA is generally used to monitor the treatment of breast cancer patients by measuring the units/ml of shed Mucl using capture and detections antibodies recognizing epitopes distinct from that recognized by DS6. In Table 5, CA15-3 is measured in ovarian cancer patients serum samples.

Table 5
(Table Removed)


1 determined by commercial ELISA kit
2 determined by commercial CA15-3 standards (1 CA15-3 U = 1 DS6 U)
3 goat anti-mouse IgG & biotin-DS6 standard curve
4 biotin-DS6 standard curve
[228] For the CA15-3 values reported in Table 5, a commercially available CA15-3 Enzyme Immuno Assay kit from CanAg Diagnostics was used. For the DS6 units/ml a standard curve was generated using the CA15-3 standards (from the CA15-3 Enzyme Immuno Assay kit from CanAg Diagnostics) in the DS6 sandwich ELISA. DS6 units/ml were arbitrarily set equal to CA15-3 units/ml. In the last two columns picomolar (pM) shed CA6 was calculated using the biotinylated DS6 standard curves shown in Figure 8C.
[229] For the quantitative analysis of CanAg levels, CanAg serum levels were those reported for patients participating in a cantuzumab mertansine clinical trial prior to treatment (Tolcher et al., J. Clin. Oncol. 21(2):211-22 (2003). An ELISA assay analogous to the one described for DS6 was used to make a CanAg standard curve using CanAg standards. C242 was used to capture the CanAg standards. Detection of

' captured CanAg was achieved using biotinylated C242 tracer followed by development with streptavidin-HRP using ABTS as substrate. A biotinylated-C242 standard curve was constructed as done for biotinylated DS6 allowing for the conversion of units/ml to a molar concentration of circulating CanAg epitopes. In Table 6 CanAg levels from cantuzumab mertansine clinical trial patients are reported along with the corresponding calculated molar concentrations of circulating CanAg.

Table 6
(Table Removed)


1 pretreatment levels of circulating CanAg measured by sandwich ELISA
2 goat anti-mouse IgG & biotin-C242 standard curve
3 biotin-C242 standard curve
[230] A comparison of the pM levels of shed CA6 in ovarian cancer patients with those calculated for shed CanAg in CanAg-positive cancer patients shows that in general shed CA6 levels are similar to shed CanAg levels. Furthermore, only 2 out of 16 ovarian cancer patients serum samples potentially have CA6 levels greater than

4.5nM, (serum samples 5 and 9 for which the signal was out of the range of the standard curve), the level above which altered herceptin pharmacokinetics was observed in clinical trials with Her2/neu-positive breast cancer patients. CanAg levels above 4.5nM were only seen in 3 out of 37 clinical trial patients. In this clinical trial there was no correlation with shed CanAg levels and more rapid clearance of cantuzumab mertansine. However, the patient with the highest CanAg level (31240 U/ml) was only sampled for 8 hours post-transfusion. These results indicate that certain epitopes of Mucl, such as CA6 and CanAg, while shed in cancer patients, are not shed at levels prohibitive for antibody therapeutic treatment.
Example 6: Cloning Murine DS6 Antibody Variable Regions.
[231] Murine monoclonal antibodies such as DS6 have limited utility in a clinical
setting because they are recognized as foreign by the human immune system. Patients
quickly develop human anti-mouse antibodies (HAMA) resulting in rapid clearance of
murine antibodies. For this reason, the variable region of murine DS6 (muDS6) was
resurfaced to produce humanized DS6 (huDS6) antibodies.
[232] The murine DS6 antibody variable regions were cloned by RT-PCR. Total
RNA was purified from a confluent T175 flask of DS6 hybridoma cells using the
Qiagen RNeasy miniprep kit. RNA concentrations were determined by UV
spectrophotometry and RT reactions were done with 4-5 ng total RNA using the Gibco
Superscript II kit and random hexamer primers.
[233] PCR reactions were performed with degenerate primers based on those
described in Wang Z et al., J Immunol Methods. Jan 13;233(l-2): 167-77 (2000). The
RT reaction mix was used directly for degenerate PCR reactions. The 3' light chain
primer, HindKL,

(TATAGAGCTCAAGCTTGGATGGTGGGAAGATGGATACAGTTGGTGC)(SEQ ID NO:25)
and 3' heavy chain primer, BamlgGl,
(GGAGGATCCATAGACAGATGGGGGTGTCGTTTTGGC) (SEQ ID NO:26) were used, and for the 5' end PCR primers were SaclMK
(GGGAGCTCGAYATTGTGMTSACMCARWCTMCA) (SEQ ID NO:27) for the light chain and an equal mix of EcoRlMHl
(CTTCCGGAATTCSARGTNMAGCTGSAGSAGTC) (SEQ ID NO:28) and EcoRlMH2 (CTTCCGGAATTCSARGTNMAGCTGSAGSAGTCWGG) (SEQ ID NO:29) for the heavy chain (mixed bases: H = A+T+C, S = G+C, Y = C+T, K = G+T, M = A+C, R = A+G, W = A+T, V = A+C+G, N = A+T+G+C). [234] PCR reactions were standard except they were supplemented with 10% DMSO (50µl reaction mixes contained final concentrations of IX reaction buffer (ROCHE), 2mM each dNTP, ImM each primer, 2 µl RT reaction, 5 µ1 DMSO, and 0.5µl Taq (ROCHE)). The PCR reactions were performed on an MJ research thermocycler using a program adapted from Wang Z et al., (JImmunol Methods. Jan 13;233( 1-2): 167-77 (2000)): 1) 94 °C, 3 min; 2) 94 °C, 15 sec; 3) 45 °C, 1 min; 4) 72 °C, 2 min; 5) cycle back to step #2 29 times; 6) finish with a final extension step at 72 °C for 10 min. The PCR products were cloned into pBluescript II SK+ (Stratagene) using restriction enzymes created by the PCR primers. Seqwright sequencing services sequenced the heavy and light chain clones.
[235] In order to confirm the 5' end cDNA sequences, additional PCR and cloning was done. The DS6 light chain and heavy chain cDNA sequences, determined from the degenerate PCR clones, were plugged into the NCBI's Blast search website and murine antibody sequences with signal sequence submitted were saved. PCR primers were

designed from these signal peptides using conserved stretches among the related DNA sequences. EcoRI restriction sites were added to the leader sequence primers (Table 7) and these were used in RT-PCR reactions as described above.
Table 7
DS6 Signal Sequence Degenerate Primers
(Table Removed)

[236] Several individual light and heavy chain clones were sequenced to identify and avoid possible polymerase generated sequence errors. Only a single sequence was obtained for both the light chain and heavy chain RT-PCR clones. These sequences were sufficient to design primers that could amplify the murine DS6 light and heavy chain sequences extending into the signal sequence. The subsequent clones from these follow-up PCR reactions confirmed the 5' end sequences of the variable region that had been altered by the original degenerate primers. The cumulative results from the various cDNA clones provided the final murine DS6 light and heavy chain sequences presented in Figure 9. Using Kabat and AbM definitions, the three light chain and heavy chain CDRs were identified (Figures 9 and 10). A search of the NCBI IgBlast database indicates that the muDS6 antibody light chain variable region most likely derives from the murine IgVK ap4 germline gene while the heavy chain variable region most likely derives from the murine IgVh J558.41 germline gene (Figure 11).
Example 7: Determination of the Variable Region Surface Residues of DS6 Antibody

[237] The antibody resurfacing techniques described by Pedersen et al. (1994) and Roguska et al. (1996) begin by predicting the surface residues of the murine antibody variable sequences. A surface residue is defined as an amino acid that has at least 30% of its total surface area accessible to a water molecule. In the absence of a solved structure to find the surface residues for muDS6, we aligned the ten antibodies with the most homologous sequences in the set of 127 antibody structure files (Figure 12). The solvent accessibility for each Kabat position was averaged for these aligned sequences (Figures 13AandB).
[238] Surface positions with average accessibilities of between 25% and 35% were subjected to a second round of analysis by comparing a subset of antibodies containing two identical residues flanking on either side (Figures 13A and B). After the second round analysis, the 21 predicted surface residues for the muDS6 heavy chain were increased to 23, adding Tyr3 and Lys23 to the list of residues with predicted surface accessibility greater than 30%. In most of our resurfaced antibodies the Kabat definition of the heavy chain CDR1 is used, but for DS6 the AbM definition was inadvertently used during the calculations so the heavy chain residue T28 was not defined as a framework surface residue as it might otherwise have been. The number of light chain surface positions was reduced from 16 to 15 because the predicted surface accessibility of Ala80 was reduced from 30.5% to 27.8% in the second round analysis. Together, the muDS6 heavy and light chain variable sequences have 38 predicted surface accessible framework residues.
Example 8: Human antibody selection

[239] The surface positions of the murine DS6 variable region were compared to the corresponding positions in human antibody sequences in the Kabat database (Johnson G, WuTT. Nucleic Acids Res. Jan l;29(l):205-6 (2001)). The antibody database management software SR (Searle, 1998) was used to extract and align the surface residues from natural heavy and light chain human antibody pairs. The human antibody variable region surface with the most identical surface residues, with special consideration given to positions that come within 5 A of a CDR, was chosen to replace the murine DS6 antibody variable region surface residues.
Example 9: Expression vector for chimeric and humanized antibodies [240] The light and heavy chain paired sequences were cloned into a single mammalian expression vector. The PCR primers for the human variable sequences created restriction sites that allowed the human signal sequence to be added in the pBluescriptll cloning vector. The variable sequences could then be cloned into the mammalian expression plasmid with EcoRI and BsiWI or Hindlll and Apal for the light chain or heavy chain respectively (Figure 14). The light chain variable sequences were cloned in-frame onto the human IgKappa constant region and the heavy chain variable sequences were cloned into the human IgGammal constant region sequence. In the final expression plasmids, human CMV promoters drive the expression of both the light and heavy chain cDNA sequences.
Example 10: Identification of Residues That May Negatively Affect DS6 Activity [241] In most of the humanizations to date a molecular model of the subject antibody has been built to identify residues proximal to a CDR as potential problem residues. With an expanding number of resurfaced antibodies to work from, historical experience

is at least as effective at predicting problems as building a model, so no molecular model was built for DS6. Instead, the murine DS6 surface residues were compared with those of previously resurfaced antibodies and residues with low to high risk for affecting the antibody's binding activity were identified.
[242] Similar sets of residues are repeatedly identified as being within 5 A of a CDR in both the available solved antibody structures and the molecular models from previous humanizations. Using this data, Table 1 gives the murine DS6 residues that are likely proximal to and possibly within 5 A of a CDR. Many of these positions have also been changed in previous humanizations, but only heavy chain position 74 has ever resulted in a loss of binding activity. The murine residue was retained in this position in both huC242 and huB4 in order to conserve the binding activity of the murine antibody. On the other hand, this same position was changed to the corresponding human residue in humanized 6.2G5C6 without loss of activity (6.2G5C6 is the anti-IGFl-R antibody often referred to simply as anti-C6). While any of the residues in Table 1 could present a problem in the humanized antibody, the heavy chain residue P73 will be of particular concern due to previous experiences in this position.
Example 11: Selection of the Most Homologous Human Surface [243] Candidate human antibody surfaces for resurfacing muDS6 were pulled from the Kabat antibody sequence database using SR software. This software provides an interface to search only specified residue positions against the antibody database. To preserve the natural pairs, the surface residues of both the light and heavy chains were compared together. The most homologous human surfaces from the Kabat database were aligned in order of rank of sequence identity. The top 3 surfaces as aligned by the SR Kabat database software are given in Table 2. The surfaces were then compared to

identify which human surfaces would require the least changes to the residues identified in Table 1. The anti-Rh(D) antibody, 28E4 (Boucher et al, 1997), requires the least number of surface residue changes (11 total) and only 3 of these residues are included in the list of potential problem residues. Since the 28E4 antibody provides the most homologous human surface, it is the best candidate to resurface muDS6.
Example 12: Construction of the DNA Sequences for Humanized DS6 Antibodies [244] The 11 surface residue changes for DS6 were made using PCR mutagenesis. PCR mutagenesis was performed on the murine DS6 variable region cDNA clones to build the resurfaced, human DS6 gene. Humanization primer sets were designed to make the amino acid changes required for resurfaced DS6, shown below in Table 8.
Table 8
(Table Removed)


[245] PCR reactions were standard except they were supplemented with 10% DMSO (50 ul reaction mixes contained final concentrations of IX reaction buffer (ROCHE), 2 fflM each dNTP, 1 mM each primer, 100 ng template, 5 jo.1 DMSO, and 0.5 jil Taq

(ROCHE)). They were run on an MJ Research thermocycler with the following program: 1) 94 °C, 1 min; 2) 94 °C, 15 sec; 3) 55 °C, 1 min; 4) 72 °C, 1 min; 5) cycle back to step #2 29 times; 6) finish with a final extension step at 72 °C for 4 min. The PCR products were digested with their corresponding restriction enzymes and cloned into the pBluescript cloning vectors. Clones were sequenced to confirm the amino acid changes.
[246] Since changing heavy chain residue P73 has caused problems in the past, two versions of the heavy chain were built, one with the human 28E4 T73 and one retaining the murine P73. The other 10 surface residue were changed from murine to the human 28E4 residue in both versions of humanized DS6 (Table 2). Sticking with the usual naming method, the most human is version 1.0 since it has all 11 human surface residues. The heavy chain version retaining the murine P73 is named version 1.2 in case further versions are required so version 1.1 is reserved for a version containing the maximum number of murine residues. The amino acid sequences of the two humanized versions are shown aligned with the murine DS6 amino acid sequence in Figure 15A and B. Both of the humanized DS6 antibody genes were cloned into the antibody expression plasmid (Figure 14) for transient and stable transfections. The cDNA and amino acid sequences of the humanized versions vl.0 and vl.2 are light chain variable region are the same and shown in Figure 16. The heavy chain cDNA and amino acid sequences of the humanized versions vl.O and vl.2 are shown in Figure 17A and B.
Example 13: Expression and purification of huDS6 in CHO cells and affinity
measurements
[247] To determine whether the humanized DS6 versions retained the binding affinity
of muDS6 it was necessary to express and purify the antibodies. CHO cells were

transfected with the respective antibody expression plasmids. Because transient expression levels of huDS6 were very low, stable cell lines were selected. [248] CHODG44 cells (4.32 x 106 cells/plate) were seeded in 15 cm plates in nonselective media (Alpha MEM + nucleotides (Gibco), supplemented with 4 mM L-glutamine, 50 U/ml penicillin, 50µg/ml streptomycin, and 10% v/v FBS) and placed in a 37°C, 5% CO2 humidified incubator. The following day, the cells were transfected with the huDS6 vl.O and vl.2 expression plasmids using a modified version of the Qiagen recommended protocol for Polyfect Transfection. The non-selective media was aspirated from the cells. The plates were washed with 7 ml of pre-warmed (37°C) PBS and replenished with 20 ml of non-selective media. The plasmid DNA (11 µg) was diluted into 800 µ1 of Hybridoma SFM (Gibco). Then, 70µ1 of Polyfect (Qiagen) was added to the DNA/SFM mixture. The Polyfect mixture was then gently vortexed for several seconds and incubated for 10 min at ambient temperature. Non-selective media (2.7 ml) was added to the mixture. This final mixture was incubated with the plated cells for 24 h.
[249] The transfection mixture/media was removed from the plates and the cells were then trypsinized and counted. The cells were then plated in selective media (Alpha MEM -nucleotides, supplemented with 4 mM L-glutamine, 50 U/ml penicillin, 50 µg/ml streptomycin, 10% v/v FBS, 1.25 mg/ml G418) in 96 well plates (250 |j.l/well) at various densities (1800, 600, 200, and 67 cells/well). The cells were incubated for 2-3 weeks, supplementing media if necessary. Wells were screened for antibody production levels using a quantitative ELISA. An Immulon 2FTB 96 well plate was coated with goat anti-human IgG F(ab)2 antibody (Jackson Immunoresearch; 1 jag/well in 100 (4.1 50 mM sodium carbonate buffer pH 9.6) and incubated for 1.5 h at ambient temperature, with rocking. All subsequent steps were conducted at ambient

temperature. The wells were washed twice with T-TBS (0.1% Tween-20, TBS) and blocked with 200 nl of blocking buffer (1% BSA, T-TBS) for 1 h. The wells were washed twice with T-TBS. In a separate plate, the antibody standard, EMI64 (100 ng/ml), and culture supernatants were serially diluted (1:2 or 1:3) in blocking buffer. These dilutions (100 µ1) were transferred to the ELISA plate and incubated for 1 h. The wells were washed 3 times with T-TBS and incubated with 100 µ1 of goat anti-human IgG Fc-AP (Jackson ImmunoResearch) diluted 1:3000 in blocking buffer for 45 min. After 5 washes with T-TBS, the wells were developed using 100µl of PNPP development reagent (10 mg/ml PNPP (p-Nitrophenyl Phosphate, Disodium Salt; Pierce), 0.1 M diethanolamine pH 10.3 buffer) for 25 min. The absorbance at 405 nm was measured in an ELISA plate reader. Absorbance readings (of the culture supernatant) in the linear portion of the standard curve were used to determine the antibody levels.
[250] The highest producing clones, identified by the ELISA, were then expanded and frozen cell stocks were prepared. To produce sufficient amount of antibody to purify, cells were expanded onto 15 cm plates (~ 1 x 106 cells/plate) with 30 ml of selective media and incubated for 1 week. Culture supernatants were collected into 250 ml conical tubes, spun down in a tabletop centrifuge (2000 rpm, 5 min, 4°C), and then sterile-filtered through a 0.2 µm filter apparatus.
[251] For purification of DS6, pellets of NaOH were added to the filtered culture supernatants to a final pH of 8.0. A Hi Trap rProtein A column (Amersham) was equilibrated with 20-50 column volumes of binding buffer. The supernatant was loaded onto the column using a peristaltic pump. Then, the column was washed with 50 column volumes of binding buffer. The bound antibody was eluted off of the column using elution buffer (100mM acetic acid, 50 mM NaCl, pH 3) into tubes set in a

fraction collector. The eluted antibody was then neutralized using neutralization buffer (2 M K2HP04, pH 10.0) and then dialyzed overnight in PBS. The dialyzed antibody was filtered through a 0.2µm syringe filter. The absorbance at 280 nm was measured to determine the final protein concentration..
[252] The affinity of the purified hulgG was compared with muDS6 by flow cytometry. In the first set of experiments direct binding to a CA6-expressing cell line, WISH, was measured. As shown in Figure 18A the chimeric DS6, huDS6 vl.O, and huDS6 vl.2 show very similar affinities with apparent KDs of 3.15 nM, 3.71 nM, and 4.2 nM, respectively suggesting that resurfacing has not disrupted the CDRs. These affinities are only slightly lower than for muDS6 (Kd = 1.93 nM) shown in Figure 18B. To confirm that the huDS6 versions retains the affinity of muDS6, competitive binding experiments were conducted. The advantage of this format is that the same detection system is used for both murine and human antibodies; that is biotin-muDS6/streptavidin-DTAF. In some cases, the experimentally measured apparent Kd can vary simply because of the secondary reagent used to indirectly detect binding. This is illustrated by a comparison of the apparent Kd of biotin-DS6 with when Goat anti-Mouse-FITC is used (Kd = 2.80 nM; Figure 18B) or when streptavidin -DTAF is used (Kd = 6.76 nM; Figure 19A). The results of the competition binding assay comparing the ability of muDS6, huDS6 vl.O, and huDS6 vl.2 to compete with biotin-DS6 are shown in Figure 19B. The apparent EC50 are 12.18 nM, 37.07 nM, and 22.64 nM for muDS6, huDS6 vl.O, and huDS6 vl.2, respectively. These results indicate that resurfacing of muDS6 to produce a humanized DS6 causes little reduction in binding affinity.
Example 14: Preparation of DS6-DM1 cytotoxic conjugate

[253] The DS6 antibody (8 mg/ml) was modified using 8-fold molar excess of N-succinimidyl-4-(2-pyridyldithio) pentanoate (SPP) to introduce dithiopyridyl groups. The reaction was carried out in 95% v/v Buffer A (50 mM KPi, 50 mM NaCl, 2 mM EDTA, pH 6.5) and 5% v/v DMA for 2 h at room temperature. The slightly turgid reaction mixture was gel-filtered through a NAP or Sephadex G25 column (equilibrated in Buffer A). The degree of modification was determined by measuring the absorbance of the antibody at 280 nm and the DTT released 2-mercaptopyridine (Spy) at 280 and 343 nm. Modified DS6 was then conjugated at 2.5 mg Ab/mL using a 1.7-fold molar excess of N2 -deacetyl-N2 (3-mercapto-l-oxopropyl)-maytansine (L-DM1) over SPy. The reaction was carried out in Buffer A (97% v/v) with DMA (3% v/v). The reaction was incubated at room temperature overnight for -20 h. The opaque reaction mixture was centrifuged (1162 x g, 10 min) and the supernatant was then gel-filtered through a NAP-25 or S300 (Tandem 3, 3x 26/10 desalting columns, G25 medium) column equilibrated in Buffer B (1X PBS pH 6.5). The pellet was discarded. The conjugate was sterile-filtered using a 0.22 µm Millex-GV filter and was dialyzed in Buffer B with a Slide-A-Lyzer. The number of DM1 molecules linked per molecule of DS6 was determined by measuring the absorbance at both 252 nm and 280 nm of the filtered material. The DMl/Ab ratio was found to be 4.36 and the step yield of conjugated DS6 was 55%. The conjugated antibody concentration was 1.32mg/mL. The purified conjugate was biochemically characterized by size exclusion chromography (SEC) and found to be 92% monomer. Analysis of DM1 in the purified conjugated indicated that 99% was covalently bound to antibody. In Figure 20, flow cytometric binding of the DS6-DM1 conjugate and unmodified DS6 to Caov-3 cells shows that conjugation of DS6 results in only a slight loss of affinity.

Example 15: In vitro Cytotoxicity of DS6-DM1
[254] As a naked antibody, DS6 has shown no proliferative or growth inhibitive activity in cell cultures (Figure 21) However, when DS6 is incubated with cells in the presence of a DM1 conjugate to Goat anti-mouse IgG heavy and light chain, DS6 is very effective at targeting and delivering this conjugate to the cell resulting in indirect cytotoxicity (Figure 21). To further test the inherent activity of naked DS6, a complement-dependent cytotoxicity (CDC) assay using murine and humanized DS6 was conducted. HP AC and ZR-75-1 cells (25000 cells/well) were plated in 96 well plates, in the presence of 5% human or rabbit serum and various dilutions of murine or humanized DS6, in 200 jil of RHBP media (RPMI-1640, 0.1% BSA, 20mM HEPES (pH 7.2-7.4), 100 U/ml penicillin and 100 ug/ml streptomycin). The cells were incubated for 2 h at 37°C. Then Alamar Blue (10% of final concentration) reagent (Biosource) was added to the supernatant. The cells were incubated for 5-24 hrs before measuring fluorescence. Both murine and humanized DS6 had no effect in a complement-dependent cytotoxicity (CDC) assay (Figure 22) This suggests that the therapeutic application of DS6 would require the conjugation of a toxic effector molecule.
[255] The cytotoxicity of maytansinoid conjugated DS6 antibody was examined using 2 different assay formats in various DS6 positive cell lines. Clonogenic assays were conducted where cells (1000-2500 cells/well) were plated on 6-well plates in 2 ml of conjugate diluted in culture media. The cells were continuously exposed to the conjugate at several concentrations, generally between 3 x 10-11 M to 3 x 10"9 M, and were incubated in a 37°C, 6% CO2 humidified chamber for 5-9 days. The wells were washed with PBS and the colonies were stained with a 1% w/v crystal violet/10 % v/v formaldehyde/PBS solution. Unbound stain was then washed thoroughly from the

wells with distilled water, and the plates were allowed to dry. The colonies were counted using a Leica StereoZoom 4 dissecting microscope. [256] Plating efficiency (PE) was calculated as the number of colonies/number of cells plated. Surviving fraction was calculated as PE of treated cells/PE of non-treated cells. The IC50 concentration was determined by graphing the surviving fraction of cells vs. the molar concentration of the conjugate. In a clonogenic assay (Figure 23), DS6-DM1 was effective in killing Caov-3 cells with an estimated IC50 of 800 pM. Antigen negative cells, A375, were only slightly affected by the conjugate at a concentration of 3 x 10"9 M, the highest concentration of DS6-DM1 tested, demonstrating that the cell killing activity of the conjugate is directed specifically toward antigen-expressing cells. However, despite apparent sensitivity to maytansine, many of the other DS6 positive cell lines were not particularly sensitive to the immunoconjugate. All of the cervical cell lines (HeLa, KB, and WISH) were sensitive to the conjugate whereas only a select number of the ovarian and breast cell lines showed any cytotoxic affects. None of the pancreatic cell lines appeared to have been affected.
[257] In the MTT assay, cells were seeded in 96-well plates at a density of 1000-5000 cells/well. The cells were plated with serial dilutions of either naked DS6 or DS6-DM1 immunoconjugate in 200 µ1of culture media. The samples were run in triplicate. The cells and antibody/conjugate mixtures were then incubated for 2-7 d, at which time cell viability was assessed by an MTT ([3(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide)] assay. MTT (50 µg/well) was added to the culture supernatant and allowed to incubate for 3-4 h at 37°C. The media was removed and the MTT formazan was solubilized in DMSO (175 µl/well). The absorbance at 540-545 nm was measured. In a MTT cell viability assay (Figure 24C), the immunoconjugate was able

to effectively kill Caov-3 cells with an estimated IC50 of 1.61 nM. The wells with the highest concentrations of conjugate contained no viable cells as compared to naked antibody which had no effect (Figures 21 and 24).
[258] The results of the MTT assays on the other cell lines were slightly different (Figure 24A, B, and D-I). In many cases, although some cytotoxicity was seen, the conjugate was unable to completely kill the entire cell population (with the exception of WISH cells). BT-20, OVCAR5, and HPAC cells were particularly resistant: in the highest conjugate concentration (32 nM) wells, over 50% of the cells were still viable.
Example 16: In vivo Conjugate Anti-tumor Activity.
[259] To demonstrate the in vivo activity of the DS6-DM1 conjugate, human tumor xenografts were established in SCID mice. A subcutaneous model of the human cervical carcinoma cell-line, KB, was developed. KB cells were grown in vitro, collected, and 5 x 106 cells in a 100 uL of serum free medium were injected under the right shoulder of each mouse and allowed to grow for 6 days to an average tumor volume of 144 ± 125 mm3 at which time drug treatment was initiated. Mice were given either PBS, conjugate at 150 µg/kg DM1, or conjugate at 225 µg/kg DM1 (2 mice per group) intravenously every day for 5 days. Toxic responses were monitored daily during the treatment. Tumor volumes (Figure 25A) and corresponding body weights (Figure 25B) were monitored throughout the study.
[260] The KB tumors treated with PBS control grew rapidly with a doubling time of about 4 days. In contrast, both groups of mice treated with conjugate exhibit complete tumor regression 14 days and 18 days after treatment initiation for the 225 µg/kg and 150 µg/kg dose groups, respectively. At the 150µg/kg dose the tumor delay was approximately 70 days. Treatment at 225µg/kg resulted in cures as there was no

evidence of tumor recurrence at the termination of the study on day 120. As seen in Figure 25B the mice in the 150 µg/kg group showed no weight loss indicating that the dose was well tolerated. At the higher dose the mice experience only a temporary 3% reduction in body weight. During the 5-day treatment course, mice exhibited no visible signs of toxicity. Taken together, this study demonstrates that DS6-DM1 treatment can cure mice of KB xenograft tumors at a non-toxic dose.
[261] DS6-DM1 activity was further tested on a panel of subcutaneous xenograft models (see Figure 26). The tumor cell-lines used to make xenografts displayed a range of in vitro maytansine sensitivities and CA6 epitope densities (Table 9 below). OVCAR5 cells and TOV-21G are ovarian tumor cell lines; HPAC is a pancreatic tumor cell line; HeLa is a cervical tumor cell line. OVCAR5 and TOV-21G cells have low surface CA6 expression; HeLa cells have an intermediate level of surface CA6 expression; HPAC cells have a high CA6 density of surface expression. TOV-21G and HPAC cells are maytansine sensitive; OVCAR5 and HeLa cells are 2-7-fold less maytansine sensitive.

Table 9
(Table Removed)


'average maximum relative mean fluorescence
[262] The 4 cell-lines were grown in vitro, collected, and 1 xlO7 cells in a 100µL of serum free medium were injected under the right shoulder of each mouse (6 mice per model) and allowed to grow for 6 days to an average tumor volume of 57.6 ±6.7 and 90.2 ±13.4 mm3 for the test and control groups respectively of OVCAR5, 147.1 ±29.6 and 176.2 ±18.9 mm3 for the test and control groups respectively of HPAC, 194.3 ±37.2 and 201.7 ±71.7 mm3 for the test and control groups respectively of HeLa, and 96.6 ±22.8 and 155.6 ±13.4 mm3 for the test and control groups respectively of TOV-21G, at which time drug treatment was initiated. For each model three control mice were treated with two weekly doses of PBS and three test mice were treated with two weekly doses of conjugate (600 µg/kg DM1) intravenously. Toxic responses were monitored daily during the treatment and tumor volumes and body weights were monitored throughout the study. The conjugate efficacy for the various models is shown graphically in Figure 26A, C, E, and G and the corresponding body weights are plotted in Figure 26B, D, F, and H. OVCAR5, TOV-21G, and HPAC cell-lines form

aggressive tumors as can be seen in the PBS controls for each model. The HeLa model had about a 3 week lag period before beginning exponential growth. In all models, DS6-DM1 conjugate treatment resulted in a complete tumor regression in all mice. For the TOV-21G, HP AC, and HeLa models the mice remain tumor-free at day 61. In the OVCAR5 model tumors recurred at about day 45 after tumor inoculation. Thus, DS6-DM1 treatment in this model results in a tumor growth delay of approximately 34 days. The growth delay is significant as OVCAR5 cells are less maytansine-sensitive and have low CA6 epitope expression. In models where either the CA6 epitope density is higher or the model has greater maytansine sensitivity, the tumor regression is more robust. It is important to note that only 2 doses were administered. Clearly the dosing schedule used in this study was not toxic to the mice as no weight loss was observed. It is likely that cures could be achieved with additional or higher conjugate doses. [263] Human ovarian cancer is largely a disease of the peritoneum. OVCAR5 cells grow aggressively as an intraperitoneal (IP) model in SCID mice forming tumor nodules and producing ascites in a manner similar to human disease. To demonstrate activity in an IP model, DS6-DM1 was used to treat mice bearing OVCAR5 IP tumors (Figure 27). OVCAR5 cells were grown in vitro, harvested and 1 x 107 cells in 100 u,L of serum free medium were injected intraperitoneally. Tumors were allowed to grow for 6 days at which time treatment was initiated. Mice were treated weekly for 2 weeks with either PBS or DS6-DM1 conjugate at a dose of 600 ug/kg DM1 and monitored for weight loss resulting from peritoneal disease. By day 28, the PBS group of mice had lost greater than 20% body weight and were euthanized. The treated group was sacrificed at day 45 after exceeding 20% body weight loss. This study demonstrates that DS6-DM1 is able to delay tumor growth in the aggressive OVCAR5 IP model despite the fact that OVCAR5 cells are less sensitive to maytansine and have few CA6

epitopes per cell. Because the dosing schedule used elicited no visible signs of toxicity, it is likely that additional and higher doses could be used to achieve further tumor growth delay or cures.
Example 17: Synthesis and Characterization of DS6-SPP-MM1-202 Taxoid Cytotoxic Conjugate
[264] DS6 was modified with the JV-sulfosuccinimidyl 4-nitro-2-pyridyl-pentanoate (SSNPP) linker. To 50 mg of DS6 Ab in 90% Buffer A, 10% DMA was added 10 equivalents of SSNPP in DMA. The final concentration of Ab was 8 mg/ml. The reaction was stirred for 4 hours at room temperature, then purified by G25 chromatography. The extent of antibody modification was measured spectophotometrically using the absorbance at 280 nm (antibody) and 325 (linker) and found to have 3.82 linkers/antibody. Recovery of the antibody was 43.3 mg giving an 87% yield. Conjugation of DS6-nitroSPP was conjugated with Taxoid MMl-202 (1812 P.16). Conjugation was carried out on a 42 mg scale in 90% Buffer A, 10% DM1. The taxoid was added in 4 aliquots of 0.43 eq/Linker (each aliquot) over a period of about 20 hours. By this time the reaction had turned noticeably cloudy. After G25 purification the resulting conjugate, recovered in about 64% yield had about 4.3 taxoids/Ab and about 1 equivalent of unreacted linker left. To quench unreacted linker, 1 equivalent of cysteine/unreacted linker was added to the conjugate with stirring overnight. A definite yellowish tinge was noticeable upon cysteine addition indicating release of thiopyridine. The reaction solution was then dialyzed in Buffer B/0.01% Tween 20 followed by further dialysis in Buffer B alone over several days. The final conjugate had 2.86 drugs/antibody. The antibody recovery was 14.7 mg, giving a 35% yield

overall. Conjugate was further biochemically characterized by SEC and found to have 89% monomer, 10.5% dimer and 0.5% higher molecular weight aggregate. [265] The results of a flow cytometry analysis comparing the binding of DS6-SPP-MM1-202 taxoid versus DS6 antibody on HeLa cells is shown in Figure 28. The results indicate that DS6 retains binding activity when conjugated to a taxane.





WE CLAIM:
1. An anti CA6 glycotope antibody or epitope-binding fragment thereof comprising
(a) at least one heavy chain variable region or fragment thereof comprises three
sequential complementarity-determining regions having amino acid sequences
represented by SEQ ID NOS: 1-3 and
(b) at least one light chain variable region or fragment thereof comprises three
sequential complementarity-determining regions having amino acid
sequences represented by SEQ ID NOS: 4-6.
2. The antibody or epitope-binding fragment thereof as claimed in claim 1, wherein said light chain variable region or fragment thereof has at least 90% sequence identity to an amino acid sequence represented by SEQ ID NO. 7 and SEQ ID NO. 8.
3. The antibody or epitope-binding fragment thereof as claimed in claim 2, wherein said light chain variable region or fragment thereof has at least 95% sequence identity to the amino acid sequence of SEQ ID NO : 7 and SEQ ID NO : 8.
4. The antibody or epitope-binding fragment thereof as claimed in claim 2, wherein said light chain variable region or fragment thereof has the amino acid sequence of SEQ ID NO : 7 and SEQ ID NO : 8.
5. The antibody or epitope-binding fragment thereof as claimed in claim 1, wherein said heavy chain variable region or fragment thereof has at least 90% sequence identity to an amino acid sequence selected from a group consisting of SEQ ID NO : 9, SEQ ID NO : 10, and SEQ ID NO: 11.
6. The antibody or epitope-binding fragment thereof as claimed in claim 5, wherein said heavy chain variable region or fragment thereof has at least 95% sequence identity to the amino acid sequence of SEQ ID NO : 9, SEQ ID NO: 10, and SEQ ID NO : 11.
7. The antibody or epitope-binding fragment thereof as claimed in claim 5, wherein said heavy chain variable region or fragment thereof has the amino acid sequence of SEQ ID NO : 9, SEQ ID NO : 10, andSEQ ID NO: 11.
8. The antibody or an epitope-binding fragment as claimed in any one of claims 1, 2, or 5 is encoded in a polynucleotide.

9. A polynucleotide of claim 8, encoding a light or heavy chain of an antibody or an epitope-binding fragment of any one of claims 1, 2, 3 or 6.
10. The antibody or an epitope-binding fragment as claimed in claim 8 encoded in a vector.

11. The vector of claim 10, wherein said vector is an expression vector that expresses said
antibody or epitope-binding fragment.
12. The anti CA6 glycotope antibody or an epitope-binding fragment thereof of claim 1 that binds to CA6 glycotope, produced by a method of resurfacing comprising:
(a) generating position alignments from relative accessibility distributions from X-ray crystallographic structures of a pool of rodent and human antibody heavy and light chain variable regions to yield a set of heavy and light chain variable region framework surface exposed positions, wherein the alignment positions for all rodent and human variable regions are at least about 98% identical;
(b)defining for a rodent antibody or epitope-binding fragment thereof a set of heavy and light chain variable region framework surface exposed amino acid residues using said set of heavy and light chain variable region framework surface exposed positions generated in said step (a);
(c) identifying from human antibody amino acid sequences a set of heavy and light chain variable region framework surface exposed amino acid residues that is most closely identical to said set of rodent surface exposed amino acid residues defined in said step (b);
(d)substituting, in the variable region framework amino acid sequence of said rodent antibody or epitope-binding fragment thereof, said set of heavy and light chain variable region framework surface exposed amino acid residues defined in said step (b) with said set of heavy and light chain variable region framework surface exposed amino acid residues identified in said step (c);
(e) constructing three-dimensional models of said variable region of said rodent antibody or epitope-binding fragment thereof and of said variable region of said rodent antibody or epitope-binding fragment thereof resulting from the substituting specified in said step (d);
(f) comparing said three-dimensional models constructed in said step (e), and identifying any amino acid residues from said sets identified in said steps (b) or (c), that are within 5 A of any atom of any residue of the complementarity determining regions of said rodent antibody or epitope-binding fragment thereof;

(g)changing any residues identified in said step (f) from the human to the original
rodent amino acid residue to thereby define a humanizing set of surface exposed
amino acid residues;
(h) replacing the set of rodent antibody variable region framework surface exposed
amino acid residues defined in said step (b) with the humanizing set of variable
region framework surface exposed amino acid residues defined in said step (g); and
(i) producing said humanized rodent antibody or epitope-binding fragment thereof that
binds to the CA6 glycotope.
13. The antibody or epitope-binding fragment thereof of claim12wherein said antibody
is humanized rodent antibody or epitope-binding fragment thereof.
14. The antibody or epitope-binding fragment thereof as claimed in claim 13 wherein
said epitope-binding fragment is selected from the group consisting of a Fab
fragment, a Fab' fragment, a F(ab') 2 fragment, a Fd fragment, a single-chain Fvs
(scFv) fragment, a single-chain antibody, a disulfide-linked Fvs (sdFv) fragment and
a fragment comprising either a VL or VH domain.
15The antibody or epitope-binding fragment thereof as claimed in claim 13, wherein
said surface exposed amino acid residues are those residues whose solvent
accessibility is above 30%.
The antibody or epitope-binding fragment thereof as claimed in claim13 consists of
16 at least one complementarity-determining region having an amino acid sequence
selected from the group consisting of SEQ ID NOS: 1 -6.
17. The antibody or epitope-binding fragment thereof as claimed in claim 13, consists of
at least one heavy chain variable region or fragment thereof and at least one light
chain variable region or fragment thereof, wherein said at least one heavy chain
variable region or fragment thereof comprises three sequential complementarity-
determining regions having amino acid sequences represented by SEQ ID NOS: 1-3
and wherein said at least one light chain variable region or fragment thereof
comprises three sequential complementarity-determining regions having amino acid
sequences represented by SEQ ID NOS: 4-6.
The antibody or epitope-binding fragment thereof as claimed in claim13 consists of a light chain variable region having an amino acid sequence of SEQ ID NO : 7 and
SEQ ID NO:8.

19. The antibody or epitope-binding fragment thereof as claimed in claim 13, consists of a heavy chain variable region having an amino acid sequence selected from the group consisting of SEQ ID NO : 9, SEQ ID NO : 10 and SEQ ID NO : 11.. 20. An anti CA6 glycotope antibody or epitope-binding fragment thereof as claimed in any of the preceding claim, that binds to a cell expressing CA6 glycotope, said improved antibody or epitope-binding fragment thereof produced by: (a) providing a DNA polynucleotide encoding an antibody or epitope-binding fragment thereof comprising at least one sequence selected from the group consisting of SEQ ID NOs : 1 to 11.
(b) introducing at least one nucleotide mutation, deletion or insertion into said
DNA polynucleotide of (a) such that at least one residue of the amino acid
sequence encoded by said DNA polynucleotide is changed ;
(c) expressing an antibody or antibody fragment encoded by the DNA
polynucleotide of (b); and
(d) screening said expressed antibody or antibody fragment for an improvement in
binding affinity to a cell expressing CA6 glycotope, thereby producing an
improved antibody or epitope-binding fragment thereof.
Anti CA6 glycotope antibody or epitope-binding fragment thereof as claimed in claim
21 20 wherein said at least one nucleotide mutation, deletion or insertion is made by a
method selected from the group consisting of oligonucleotide-mediated site-directed
mutagenesis, cassette mutagenesis, error-prone PCR, DNA shuffling and the use of

Documents:

707-delnp-2006-pct-332.pdf

707-delnp-2006-pct-401.pdf

707-delnp-2006-pct-402.pdf

770-DELNP-2006-Abstract-(14-10-2011).pdf

770-DELNP-2006-Abstract-(29-05-2009).pdf

770-DELNP-2006-Abstract-(30-04-2009).pdf

770-delnp-2006-abstract.pdf

770-DELNP-2006-Claims-(14-10-2011).pdf

770-DELNP-2006-Claims-(29-05-2009).pdf

770-DELNP-2006-Claims-(30-04-2009).pdf

770-delnp-2006-claims.pdf

770-DELNP-2006-Correspondence Others-(14-10-2011).pdf

770-DELNP-2006-Correspondence-Others-(29-03-2011).pdf

770-DELNP-2006-Correspondence-Others-(29-05-2009).pdf

770-DELNP-2006-Correspondence-Others-(30-04-2009).pdf

770-delnp-2006-correspondence-others.pdf

770-delnp-2006-correspondence-po.pdf

770-DELNP-2006-Description (Complete)-(14-10-2011).pdf

770-DELNP-2006-Description (Complete)-(29-05-2009).pdf

770-DELNP-2006-Description (Complete)-(30-04-2009).pdf

770-delnp-2006-description (complete).pdf

770-DELNP-2006-Drawings-(14-10-2011).pdf

770-DELNP-2006-Drawings-(29-05-2009).pdf

770-DELNP-2006-Drawings-(30-04-2009).pdf

770-delnp-2006-drawings.pdf

770-DELNP-2006-Form-1-(29-05-2009).pdf

770-DELNP-2006-Form-1-(30-04-2009).pdf

770-delnp-2006-form-1.pdf

770-delnp-2006-form-13(30-04-2009).pdf

770-DELNP-2006-Form-13-(29-03-2011).pdf

770-delnp-2006-form-13-(30-04-2009).pdf

770-delnp-2006-form-13.pdf

770-delnp-2006-form-18.pdf

770-DELNP-2006-Form-2-(14-10-2011).pdf

770-DELNP-2006-Form-2-(29-05-2009).pdf

770-DELNP-2006-Form-2-(30-04-2009).pdf

770-delnp-2006-form-2.pdf

770-DELNP-2006-Form-26-(30-04-2009).pdf

770-delnp-2006-form-26.pdf

770-DELNP-2006-Form-3-(30-04-2009).pdf

770-delnp-2006-form-3.pdf

770-delnp-2006-form-5.pdf

770-delnp-2006-pct-101.pdf

770-delnp-2006-pct-210.pdf

770-delnp-2006-pct-237.pdf

770-delnp-2006-pct-304.pdf

770-DELNP-2006-Petition-137-(30-04-2009).pdf


Patent Number 234811
Indian Patent Application Number 770/DELNP/2006
PG Journal Number 31/2009
Publication Date 31-Jul-2009
Grant Date 15-Jun-2009
Date of Filing 15-Feb-2006
Name of Patentee IMMUNOGEN, INC.
Applicant Address 128 SIDNEY STREET, CAMBRIDGE, MA 02139-4239, USA.
Inventors:
# Inventor's Name Inventor's Address
1 PAYNE, GILLIAN 21 OAK VALE ROAD, WABAN, MA 02468, USA.
2 CHUN, PHILIP 181 PLEASANT STREET #25,MALDEN, MA 02148, USA
3 TAVARES, DANIEL, J. 27 SYLVESTER RD, NATICK, MA 01760, USA
PCT International Classification Number C07H 21/04
PCT International Application Number PCT/US2004/023340
PCT International Filing date 2003-07-21
PCT Conventions:
# PCT Application Number Date of Convention Priority Country
1 60/448,447 2003-07-21 U.S.A.