`
`European Patent Office
`
`Office european des brevets
`
`@ Publication number:
`
`0 403 156
`A1
`
`@
`
`EUROPEAN PATENT APPLICATION
`
`@ Application number: 90306178.6
`
`@ Date of filing: 07.06.90
`
`@ Int. c1.s: C07K 15/06, C12P 21/08,
`C12N 15/62, A61 K 39/395,
`C12N 5/20, A61K 47/48,
`C12N 5/10
`
`@ Priority: 07.06.89 US 362549
`25.05.90 us 529979
`
`@ Date of publication of application:
`19.12.90 Bulletin 90/51
`
`@ Designated Contracting States:
`AT BE CH DE DK ES FR GB GRIT Ll LU NL SE
`
`@ Applicant: GENZYME CORPORATION
`One Mountain Road
`Framingham, Massachusetts 01701(US)
`
`Applicant: BEHRINGWERKE
`Aktiengesellschaft
`Postfach 1140
`D-3550 Marburg 1 (DE)
`
`@ Inventor: Kurrle, Roland
`Schenkendorfweg 18
`D-3550 Marburg(DE)
`Inventor: Shearman, Clyde W.
`10, Pheasant Hill Road
`Bellingham, Massachussetts 02019(US)
`Inventor: Moore, Gordon P.
`24, Brookside Avenue
`Lexington, Massachusetts 02173(US)
`Inventor: Seiler, Fritz
`Oberer Eichweg 10
`D-3550 Marburg(DE)
`
`® Representative: Sheard, Andrew Gregory et al
`Kilburn & Strode 30, John Street
`London WC1N 2DD(GB)
`
`@ Improved monoclonal antibodies against the human alpha/beta t-cell receptor, their production and
`<(
`use.
`
`~@ Monoclonal antibodies against the human alpha/beta T-cell receptor are described along with pertinent
`<r- sequence information of the variable and constant regions and humanised and civilised versions of such
`. M antibodies.
`0
`~
`0
`c..
`w
`
`Xerox Copy Centre
`
`BIOEPIS EX. 1071
`Page 1
`
`
`
`EP 0 403 156 A1
`
`IMPROVED MONOCLONAL ANTIBODIES AGAINST THE HUMAN ALPHA/BETA T-CELL RECEPTOR, THEIR
`PRODUCTION AND USE
`
`s
`
`The instant application discloses useful monoclonal antibodies (mAbs) against an epitope within the
`constant region of the human alpha/beta T-cell receptor (TCR), their production and use for immunosup(cid:173)
`pressive therapy in organ and bone marrow transplantation, in treatment of autoimmune diseases and for
`therapeutic applications in immunoregulation.
`For decades Anti-leukocyte-antibodies for clinical and experimental use have been generated by
`immunizing, e.g. horses, rabbits, goats and rats with leukocytes, lymphocytes or subpopulations thereof or
`with various tumor cell lines. Specificity of such anti-leukocyte-globulin/Antithymocyte-globulin (ALG/ATG}
`preparations was usually obtained by careful selection of antigen sources or by absorption of undesired
`antibodies with different cell types such as erythrocytes, B-cells, selected cell lines, etc. This strategy,
`10 which resulted in high quality ALG/ATGs, requires a considerable expenditure of laboratory force, quality
`control, and the necessity to ascertain reproducible specificity from batch-to-batch. Within certain limita(cid:173)
`tions, those ALG/ATGs made it possible to study leukocyte differentiation, the cellular origin of leukemia
`and lymphoma, to define T-cell subpopulations, and even to prepare antiidiotypic antisera.
`The therapeutic efficacy of ALG/ATGs is well known, especially for immunosuppression in organ
`transplantation. In addition, ALG/ATGs have been used successfully to treat patients with aplastic anemia
`and for "purging" bone marrow cells within the context of bone marrow transplantation. Despite the success
`with ALG/ATG preparations it is accepted that even with a reasonable amount of laboratory effort, these
`polyclonal anti-leukocyte-antibodies might vary from batch-to-batch and the specificity of these antisera is
`limited.
`Once the technique to produce mAbs was developed by Kohler and Milstein (Nature 225(1974)), it was
`possible to generate antibodies with much higher specificity as compared to ALG/ATG preparations. Since
`mAbs recognize not only distinct antigens, e.g. on cell surfaees, but also particular epitopes within such an
`antigen, they can be used with high efficacy to discriminate even between very similar cell populations, and
`to characterize the biochemical and functional aspects of the recognized antigen.
`mAbs have been most frequently and successfully used for immunosupp-ressive therapy in clinical
`organ transplantation. However, most mAbs have a broad immunosuppressive capacity, thus undesirably
`influencing functions of a wide spectrum of immune cells presumably not all involved in graft rejection.
`Among others, the monoclonal antibody OKT3, directed against mature human T cells, has been
`extensively used for the treatment of patients undergoing acute allograft rejection after kidney transplan-
`tation (Russell P.S., Colvin R.B., Cosimi; A.B.: Annu. Rev. Med. 35:63, (1984} and Cosimi A.B., Burton R.C.,
`Colvin, R.B. et al.: Transplantation 32:535, (1981 )). Moreover, OKT3 and rabbit complement were used for
`purging mature T-cells from donor marrow to prevent acute graf t-v-host disease (GVHD) in allogeneic bone
`marrow transplantation (Prentice, H. G., Blacklock, H.A., Janossy, G. et al.: Lancet 1:700, (1982} and
`Blacklock, H.A., Prentice, H.G., Gilmore, M.J. et ai.:Exp. Hematol. 11 :37, (1983)). Whereas OKT3 treatment
`35 seems to be effective in the prevention of GVHD in allogeneic bone marrow transplantation for acute
`leukemia, a combined in vitro/in vivo treatment with OKT3 failed to prevent GVHD during therapy for severe
`combined immunodeficiency (Hayward, A.R. et al.: Clin. Lab. Observ. 100:665, (1982)).
`Treatment of T-cells with OKT3 elicits several responses inconsistent with immune suppression
`including T-cell activation, production of immune mediators and T3-modulation. The T3-antigen complex
`recognized by CD3-mAbs (e.g. OKT3) is postulated to play a crucial role during T-cell activation. Functional
`studies indicate that the T3-antigen complex is involved in specific immune functions as molecules
`functionally and physically associated with the respective T-cell receptor. Under physiological conditions the
`mere binding of OKT3 to T-cells results in T-cell activation. When T-cells are activated in the presence of
`accessory cells, OKT3 is highly mitogenic forT-cells from all donors, whereas for anti-T3 mAbs of the lgG1
`isotype, nonresponsiveness caused by polymorphism in the accessory cell function has been described.
`Additionally, it has been reported that stimulation of human peripheral blood lymphocytes with OKT3
`induces the production of immune mediators such as interferon (alpha-IFN) and interleukin-2 (IL-2} (Pang,
`R.H., Yip, Y.K., Vilcek, J.; Cell lmmunol. 64:304, (1981) and Welte, K., Platzer E. Wang, C.Y., et al.: J.
`lmmunol. 131:2356, (1983)). One of the earliest events after the binding of OKT3 to the T-cell membrane is
`the modulation of the T3 complex. T3 modulation occurs under appropriate conditions in vitro as well as in
`vivo and this mechanism, among others, seems to be responsible for the "escape" of r-=-cells during in vivo
`therapy with OKT3. Also antigenic modulation seems to play a critical role in T-cell activation.
`- (cid:173)
`In light of undesired effects of OKT3 described above, there was a need to create new mAbs having a
`specificity against mature lymphocytes and suitable for clinical application. Still another aspect and goal
`
`15
`
`20
`
`25
`
`30
`
`40
`
`45
`
`so
`
`2
`
`BIOEPIS EX. 1071
`Page 2
`
`
`
`5
`
`25
`
`EP 0 403 156 A1
`
`3
`
`involves the coupling of such mAbs to cytotoxic agents (radioisotopes, toxins, enzymes, etc.) in order to
`improve their effectiveness in mediating cytolysis.
`Another goal involved the modification, via genetic manipulation, of the mAb in order to produce
`chimeric antibodies having mixed murine and human characteristics in order to improve their effectiveness
`and/or lower their immunogenicity in patients.
`Chimeric antibodies offer an additional advantage over murine mAb with regards to immunogenicity in
`patients. A chimeric antibody would retain the affinity and specificity of the parental murine mAb and
`eliminate the patient immune response to the murine constant regions. A further refinement involves
`humanization of the variable regions. Only the complementarity determing regions and selected framework
`10 amino acids necessary for antigen binding are maintained murine. The remaining framework regions are
`converted to human sequences. The resulting mAb of the present invention is thus essentially a human
`antibody with a much lower immunogenicity in patients.
`In accordance with the various goals, aspects and principles of the present invention, we have made a
`useful mAb secreted by the hybridoma cell line designated as BMA 031 by immunizing mice with human
`75 peripheral blood T-lymphocytes separated by the E-rosette-technique (so called E+ -cells). The mAb
`secreted by SMA 031 (in the following likewise designated as BMA 031) is a murine monoclonal antibody of
`the lgG2b isotype with a specificity for the alpha- and beta-chain of the TCR/CD3 receptor complex. As
`compared to OKT3 or BMA 030 (both clustered as CD3-antibodies), BMA 031 only very weakly induces
`mediator release after binding to T-cells. It is highly effective in clinical application, e.g. kidney transplant for
`20 patients with increased immunological risk when given post or at the time of the transplant-operation. Since
`no major side effects were observed with SMA 031 even at doses of up to 50 mg/dose, it may be
`advantageously given at time of surgery (preferably via a first injection of 50 mg intravenously) followed by
`a second injection 48 hours after transplantation. Graft function was perfect in almost all cases.
`BMA 031 defines, therefore, a valuable epitope on the alpha-beta TCR distinct from the epitope on CD3
`recognized by OKT3 or by other mAbs against mature T-lymphocytes. In order to completely characterize
`this mAb and also to permit, by recombinant DNA techniques, exchange of the variable region frameworks
`outside of the hypervariable regions and exchange of the human constant region in place of the
`endogenous murine constant region, the DNA coding for the heavy and light chains of SMA 031 was cloned
`and sequenced.
`Principal aspects of the invention therefore involve the recognition and definition of the epitope on
`alpha/beta TCR and the monoclonal antibodies directed against this epitope.
`Preferred embodiments of the antibodies of the present invention are BMA 031 and those mAbs
`containing the same amino acid sequence at the hypervariable regions, or functional equivalents thereof.
`These equivalents include, for example, chimeric variants in which the mouse constant regions are replaced
`35 with human C-regions.
`A further aspect of the invention concerns the use of these antibodies in clinical application before,
`during or after transplant surgery, in bone marrow transplantation, in treatment of cancer (direct treatment of
`leukemic cells and indirect treatment of all types of cancer by activation of T-cell populations) and for
`therapeutic applications in immunoregulation.
`The antibodies of the present invention are also useful in the detection of immunocompetent T-cells.
`Additional aspects will become apparant upon study of the detailed description of preferred embodi(cid:173)
`ments set forth below.
`The description of the preferred embodiments refers by way of example to the accompanying drawings,
`wherein:
`Figure 1 graphically depicts the construction scheme of the BMA 031 genomic library;
`Figure 2 describes the probes used to screen the library;
`Figure 3 graphically sets forth, in linear fashion, the human constant region expression vectors;
`Figures 4A, 48 and 4C show results of competitive immunofluorescence assays with SMA 031
`Chimeric Antibodies.
`Figures 5A and 58 show results ofT-cell proliferation assays with BMA 031 Chimeric Antibodies.
`Figures 6A, 68 and 6C show results of ADCC assays with the BMA 031 Chimeric Antibodies.
`Figure 7 shows the results of competitive immunofluoresence assays with BMA-EUC1V3 antibody.
`As used herein, the term "epitope" refers to the structure recognized by the monoclonal antibody BMA
`031, and is generally thought to be independent of the remaining portion of the antigen on which the
`55 epitope is located. It is presently unknown exactly how the epitope is formed structurally but it is anticipated
`that it may be formed by either (i) a part of the amino acid sequence of the antigen molecule; (ii) the three(cid:173)
`dimensional structure formed by non-contiguous amino acids within the same molecule; (iii) the three(cid:173)
`dimensional structure formed by various molecules within an antigen complex; or (iv) some combination
`
`30
`
`40
`
`45
`
`50
`
`BIOEPIS EX. 1071
`Page 3
`
`
`
`EP 0 403 156 A1
`
`5
`
`thereof. As used herein the term "monoclonal antibody (mAb)" means an antibody composition having a
`substantially homogeneous antibody population. It is not intended to be limited as regards to the source of
`the antibody or the manner in which it is made and in the most preferred embodiments is intended to
`include recombinant methods of manufacture.
`As used herein with respect to the exemplified BMA 031 antibody, the term "functional equivalent"
`means a monoclonal antibody that: (a) blocks the binding of BMA 031 and whose binding is blocked by
`bound BMA 031; (b) binds selectively to human T -cells having expressed the alpha/beta TCR but not
`having expressed TCR-gamma and/or delta-chains; (c) has one of the known isotypes; (d) binds to the
`same antigen as BMA 031 as determined by immunoprecipitation, western blotting or other biochemical
`10 analyses; (e) binds to the same antigen as determined by cells transfected by gene(s) for the alpha/beta
`TCR or segments thereof.
`
`Example 1. Immunization and cell fusion
`
`15
`
`Three Balb/c mice (female; age: 6-8 weeks) were twice immunized intraperitonally with 1.5 x 106 E+(cid:173)
`cells each. Peripheral blood human T-lymphocytes were separated by the E-rosette-technique (rosette(cid:173)
`formation with sheep red blood cells - E+ -cells) and were derived from the peripheral blood of a healthy
`donor whose donated blood routinely tested HIV-negative for more than one year after having given the
`20 blood for immunization purposes.
`Three days after the last immunization, the spleens of all three mice were removed, a single cell
`suspension was prepared and the lymphocytes were fused with the murine myeloma cell P3X63-Ag8.653
`(ATCC # CRL 1580) according to a standard fusion protocol. The myeloma cell P3X63-Ag8.653 is described
`-as an .immunoglobulin non-producer mutant derived from the original myeloma cell P3-X63-Ag8 (ATCC #
`25 TIB 9).
`. After fusion, cells (1 x 106 cells/well) were cultured in the presence of HAT-medium (Dulbecco's
`+
`modified Eagle's medium + 10% FCS 0.1mM hypoxanthine, 0.4 uM aminopterin, 16 iuM thymidine) to
`select for hybridoma cells.
`
`30
`
`Example 2. Isolation and characterization of the hybridoma clone BMA 031
`
`Supernatants of growing hybridoma cells (fusion number BW 242) were harvested routinely and tested
`for the presence of murine immunoglobulins in an ELISA test-system designed to measure murine lgG
`35 quantitatively. At the same time, the supernatants were tested for antibodies with specificity for human
`lymphocyte cell surface antigens.
`Within this selection process, single cells were picked from the original well (BW 242/1177) and were
`cultured separately. In subsequent steps these cells were repeatedly cloned by micromanipulation under
`microscopic control (BW 2421412). During three cloning cycles 1 00% of growing clones produced mon-
`40 oclonal antibodies with identical binding specificity and identical behavior with respect to biochemical
`criteria. One clone was selected and designated as BMA 031.
`Extensive analyses were carried out to define the specificity and the functional properties of the
`hybridoma clone, BMA 031.
`A master seed bank and a working cell bank were established starting with the hybridoma clone BMA
`45 031. The two cell banks underwent extensive examinations to assure absence of contamination with
`pathogens (mycoplasmatic, bacterial and viral infections). In addition, starting from the master seed bank,
`experiments were carried out which showed that even after the 50th culture passage detectable variations in
`antibody specificity were neither measurable nor could non-producer mutants be detected when analysed
`by single cell cloning and by calculating antibody production rates in bulk culture.
`
`50
`
`Example 3. Specificity of BMA 031
`
`Monoclonal antibodies directed to lymphocY_te cell surface antigens are usually ch~racterized by a
`55 binding assay. To analyse the specificity of BMA 031, cytofluorometric assay systems were used
`predominantly. In particular, binding assays were carried out as described below.
`
`4
`
`BIOEPIS EX. 1071
`Page 4
`
`
`
`EP 0 403 156 A1
`
`3.1 Analyses of the binding of BMA 031 with peripheral blood leukocytes
`
`5
`
`---
`Comparison of mAb reactivity with reference mAbs
`---'---- --
`--
`Defined subpopulations of PBLs were labelled either with a reference mAb and/or with BMA 031. If both
`mAbs have identical specificity they will. stain identical populations of cells. In these studies, the reference ·
`mAbs used were those which had been previously characterized in the Workshops for Human Leukocyte
`Differentiation Antigens 1., II. and Ill. (Paris, 1982; Boston, 1984; Oxford, 1986; Bernard, A.E.: Leukocyte
`10 Typing. Springer-Verlag (1984); Reinherz, E.L.E.: Leukocyte Typing II. Springer-Verlag (1986); McMichael,
`A.J.E.: Leukocyte Typing Ill. Oxford University Press (1987).
`
`Expression of the antigen on different leukocyte subpopulations
`
`15
`
`Peripheral blood leukocytes were analyzed on an Ortho Cytofluorograph 50H/2150 Computer-system
`modified for single-step analyses of whole blood. Cells were either directly labelled with BMA 031-FITC
`and/or with reference mAbs or alternatively stained in a second step with isothiocyanate-fluoresceinated
`rabbit anti-mouse lgG (lg-F(ab')2-FITC). By using cytofluorometric assays, it was discovered that binding of
`20 BMA 031 was only detected on cells of the T-cell lineage which express the alpha/beta T-cell receptor.
`BMA 031 does not react with cells which express the gamma/delta T-cell receptor. BMA 031 can therefore
`effectively be used to discriminate between alpha/beta-TCR + and gamma/delta-TCR + cells. Presence or
`absence of distinct TCR chains reflect the status of T-cell differentiation during T-cell ontogeny. In
`peripheral blood, molecules of the CD3-complex are predominantly expressed in association with the
`25 human alpha/beta-TCR (7). In healthy blood donors, the frequency of T~lymphocytes stained by BMA 031 is
`usually only 1-5% lower than that measured with CD3 mAbs (for normal frequencies of CD3 + cells see 1-3).
`As shown with cloned T-cells, this population of CD3 + BMA 031- cells express the gamma/delta TCR
`+
`instead of the alpha/beta TCR. In pathological situations, however, the frequency of CD3 BMA 031- T-cells
`can increase to 20% of the CD3 + cells in individual patients. On CD3 + BMA 031 + cells, binding of BMA
`so 031-FITC is blocked by OKT3 and vice versa in competitive immunofluorescence assays. Nevertheless, by
`analyzing such data in more detail (comparison of fluorescence-histograms) and by blocking studies with
`anti-idiotypic antisera it can be clearly shown that BMA 031 recognizes a different epitope than all known
`CD3-mAbs.
`
`35
`
`Example 4. Characterization of the functional properties of BMA 031
`
`BMA 031 is a murine monoclonal antibody of the lgG2b isotype. Due to the unique specificity and the
`isotype of BMA 031, this antibody triggers a specific pattern of biological functions. After binding toT-cells,
`40 BMA 031 does not induce T-cell proliferation comparable to CD3-mAbs of the lgG2a isotype (e.g. BMA 030,
`OKT3) in a three day thymidine incorporation assay nor does it induce Ca2 + influx in resting T-lymphocytes
`. In contrast to CD3-mAbs such as BMA 030 or OKT3, binding of BMA 031 to T-cells induces antigen
`modulation only weakly and triggers the release/production of cytokines to a very low extent.
`
`.45
`
`Example 5. Preparation of DNA and RNA from BMA 031
`
`For preparation of genomic DNA, approximately 1 x 109 cells were grown in T-flasks. DNA was
`prepared by
`lysis
`in SDS, digestion with proteinase K and RNAse A and sequential, gentle,
`50 phenol/chloroform extractions in high salt. Low density agarose gels indicated that the average length of the
`genomic DNA was greater than 50 kilobases (Kb); long length is important in creation of complete genomic
`clone libraries in lambda phage vectors. The yield was about 10 mg of DNA.
`Approximately 109 cells were also grown for RNA isolation. Cells were lysed and RNA extracted using
`guanidinium thiocyanate. RNA yield was about 1 0 mg and it appeared clean and undegraded on agarose
`+
`55 gels. Poly A RNA was prepared by binding total RNA to oligo dT cellulose. Yield was about 5%, i.e. about
`+
`500 ug of poly A mRNA.
`
`5
`
`BIOEPIS EX. 1071
`Page 5
`
`
`
`EP 0 403 156 A1
`
`Example ~ Sequencing SMA 031 heavy (H) and light (L) chains by primer extension of mRNA
`
`5
`
`Because of tne high abundance of specific mRNA, immunoglobulins can be sequenced by primer
`+
`extension of total poly A mRNA. This is also facilitated by their common constant regions which allow
`synthesis of "universal" primers with which to begin the extension. Subsequent primers can be made as
`sequence information is gathered. "Universal" H and L chain primers (Light chain "Universal" Primer:
`5 'TGGATGGTGGGMGATG3 '; Heavy chain "Universal" Primer: 5 'GGGGGCCAGTGGATAGAC3') were
`isotopically end labelled with polynucleotide kinase and hybridized individually to about 7 11.g aliquots of
`SMA 031 poly A+ mRNA. The hybridized product was extended using avian reverse transcriptase in the
`10 presence of dideoxy nucleoside triphosphates. Reaction products were separated on gradient acrylamide
`gels and autoradiographed. About 200 nucleotides (NT) of sequence were read from the initial extension.
`For both H and L chains, two additional primers were synthesized, each 17-18 NT in length, namely
`#149 (Light) 5 'AGGGACTCCAGMGCCA3'
`#180 (Light) 5 ' CTGGAGATGCMCATG3'
`15 #179 (Heavy) 5 'CTCCATGTAGGCTGTACP'
`#178 (Heavy) 5 'CCAGMGCCTTGCAGGA3 '
`The extension primers as described above generate complete, overlapping sequence (a total of about 440
`NT for each gene) specifying the H and L chain variable (V) regions.
`The sequences obtained are listed in Table 1A (Heavy) and 1 B (Light). Positions of the signal
`20 sequence, start of the mature protein, invariant cysteines, complementarity determining regions (CDR),
`joining region (J), and primers used in sequencing are indicated. These sequences were confirmed by DNA
`sequencing of isolated lambda clones (see Example 7 below).
`
`25 Example ?.:.. Construction and Screening of a Genomic Clone library
`
`The general strategy of the library construction is depicted in Figure 1 and Table 2 Parts A and B. Total
`genomic DNA was partially digested with the restriction enzyme Sau 3A. DNA from the digest in the size
`range 10-20 Kb was isolated by preparative gel electrophoresis and binding to glass-impregnated paper.
`30 The resulting material was ligated with phage "arms" from the lambda vector EMBL-3 (ATCC #37266).
`Ligated material was packaged into phage particles (cloning efficiency 1 x 106 pfu/l.l.g genomic DNA).
`Recombinant clones were plated at a density of 105 /plate. Replica filter lifts were performed on each plate
`and the nitrocellulose filter discs were processed for hybridization. These disks were hybridized in duplicate
`to 32p-labelled fragment probes derived from the intron between V and C exons of H or L chain (these
`35 probes are "universal" lg probes). Putative positive clones were confirmed and plaque-purified by up to four
`rounds of rescreening. Rescreening was carried out with radiolabelled, synthesized oligomers (17 - 33 NT)
`corresponding to hypervariable regions known from Example 6. The approximate relative location of the
`fragments and oligomers used as probes for screening are indicated in Figure 2.
`After exhaustive screening of the library, multiple positive lambda clones were isolated containing SMA
`40 031 H and L chain exons. As listed in Table 2, Part C, of 14 H chain clones initially identified, 4 were true
`positives; of 27 L chain clones, 7 were real. The genomic sequence of the H and L chains as determined by
`both Sanger and Maxam-Gilbert sequencing, are listed in Tables 3A and 38, respectively.
`
`45 Example 8. Vector construction for expression of chimeric BMA 031
`
`The H chain of chimeric SMA 031 was synthesized by a vector containing the human gamma 1
`constant region or a second vector containing the gamma 4 constant region. For the L chain, the vector
`contained the human kappa constant exon. Each vector also contained an upstream cloning site for the
`respective BMA V gene, drug selectable markers, and signals necessary to allow replication in bacteria.
`Restriction maps of these vectors are shown in Figure 3. The V exons of BMA 031 H and L chains were
`subcloned into the vectors described above.
`
`50
`
`55 Example ~ Transfection of chimeric BMA 031 gene into myeloma cells for expression
`
`The chimeric constructs described above were co-transfected into SP2/0 (ATCC #CRL 1581 ), a non(cid:173)
`immunoglobulin-producing murine hybridoma, by electroporation. After a 48 hour expression period, the
`
`6
`
`BIOEPIS EX. 1071
`Page 6
`
`
`
`EP 0 403 156 A1
`
`llg/ml) and
`resultant cells or, transfectomas, were placed in media containing both mycophenolic acid (1
`Geneticin
`(1 mg/ml) (Gibco). Growth of the transfectomas was apparent in about two weeks. The
`transfection efficiency in double selection was approximately 1 x 1 o-s. The proportion of drug-resistant
`clones secreting antibody was greater than 50% with the level of antibody secretion varying from about 1
`llg/ml to about 17 llg/ml.
`
`Example ~ Subcloning of Transfectomas
`
`The BMA 031-G1 and BMA 031-G4 cell lines were subcloned to eliminate genetic heterogeneity that
`may have arisen since the original transfection. The best subclone of each chimeric transfectoma was
`isolated and analyzed for antibody production. These final cell lines, BMA031-G1-1 and BMA 031-G4-1,
`produced antibody at a rate of 7 pg/cell/24 hours for lgG1 and 5 pg/cell/24 hours or lgG4. Saturated
`cultures accumulate antibody to 35 llg/ml for lgG1 and 151lg/ml for lgG4.
`
`5
`
`10
`
`15
`
`Example 11. Characterization of chimeric BMA 031
`
`3D
`
`25
`
`The antibodies, herein referred to as BMA 031-G1 and BAM 031-G4, secreted by BMA 031-G1-1 and
`20 BMA 031·G4-1, respectively, were tested to ensure that they were indeed BMA 031 mouse-human chimeric
`antibodies. Analysis by a series of ELISA assays showed that the antibodies contained human kappa and
`human gamma constant regions. Moreover, the antibodies did not react with antibodies directed against
`murine kappa or gamma constant regions. lsotyping reagents also confirmed that the chimeric antibodies
`were· of the lgG1 and lgG4 isotypes.
`The purified chimeric antibodies were also showri to be functionally active in an immobilized cell ELISA
`assay. Both chimeric antibodies bound to PBL-ALL cells in a similar fashion as the original murine BMA
`031. Directly labelled chimeric antibodies BMA 031-G1-FITC, and BMA 031-G4-FITC, were used in
`cytofluorometric assays systems to compare the specificity of these antibodies with murine BMA 031-FITC.
`A typical reactivity pattern of these antibodies with peripheral blood leukocytes is shown in Table 4. In
`competitive immunofluorescence assays, preincubation of human lymphocytes with chimeric BMA 031-G1
`or BMA 031-G4 blocked binding of murine BMA 031-FITC and vice versa (See Figure 4A-4C). All these
`data indicate that the specificity of chimeric BMA 031-G1 and BMA 031-G4 is identical to murine BMA 031.
`However, the functional properties of chimeric BMA 031-antibodies clearly differ from murine BMA 031
`or from CD3-mAbs such as BMA 030. It is well known that most CD3-mAbs (e.g. BMA 030) trigger T-cell
`35 proliferation even in low doses but have no mitogenic effect when used in high concentrations (See Figures
`5A and 58). This effect is known as "high dose inhibition".
`In contrast to murine BMA 031, both chimeric BMA 031-mAbs are highly effective in triggering T(cid:173)
`lymphocyte proliferation as measured in a 3 day or 6 day thymidine incorporation assay without showing
`any evidence of high dose inhibition effects even in much higher dose ranges. Two representative
`40 experiments are shown in Figures 5A and 58. In such experiments, the height of responsiveness and the
`optimal concentration for stimulation with BMA 030 or other CD3-mAbs may differ from blood donor to
`blood donor due to individual differences of the immune status.
`As described above (Example 4), in contrast to CD3-mAbs, binding of murine BMA 031 to human T(cid:173)
`lymphocytes usually results in the release of only low concentrations of immunomediators. Again, chimeric
`45 BM~ 031 antibodies differ in this respect from murine BMA 031 as well as from BMA 030 (CD3). In Table 5,
`a representative experiment is shown, where release of immunomediators is measured by induction of HLA(cid:173)
`Dr expression on COLO 205-cells.
`It is known from the literature that COLO 205 cells respond with an enhanced expression of HLA-DR
`antigens after incubation with recombinant immunomediators like gamma-Interferon, Tumor Necrosis Factor
`5o or supernatants of activated T-cells which contain such factors. By comparing the enhancement of HLA-DR
`expression induced by supernatants of activated T-cells with reference values generated by the use of
`recombinant factors such as gamma-Interferon, the amount of immunomediators ·in culture supernatants can
`be determined. The data summarized in Table 5 indicate that the kinetics and the amount of im(cid:173)
`munomediators released after binding of chimeric BM~ 031 antibodies differ from the effects tri_ggered by
`55 murine BMA 031 or the CD3-mAb BMA 030.
`Since the chimeric BMA 031 antibodies were able to interact efficiently with human Fe receptors in the
`T cell proliferation assays, there was a strong possibility that they would have high ADCC titers as well. To
`evaluate the ADCC capacity of these mAbs, they were compared to rabbit anti-GH-1 antiserum selected as
`
`7
`
`BIOEPIS EX. 1071
`Page 7
`
`
`
`EP 0 403 156 A1
`
`the best out of eight rabbit anti-human T cell globulins in ADCC capacity. The data of a representative
`experiment is shown in Figures 6A-6C. Even at low effector:target cell ratios (Figure 6A) or extremely low
`antibody concentrations (Figure 68, 6C), chimeric BMA 031 antibodies are highly potent in killing HPB-ALL
`cells. In contrast, murine BMA 031 is very poor at ADCC. Accordingly, if cytolysis is an important criteria,
`5 BMA 031 can be advantageously used conjugated to a cytotoxic moiety.
`
`Example 12. Production of "civilized" BMA 031 Antibodies
`
`70
`
`Determination of civilized BMA 031 amino acid sequence
`
`25
`
`so
`
`In the past, "humanization" has been associated with chimeric constructions in which murine V regions
`are expressed with human C regions. To avoid confusion, the term "civilized" is used herein for
`75 constructions of "humanized" V regions expressed with human C regions.
`To determine the optimal human sequence with which to civilize the murine BMA 031 antibody, the
`murine BMA 031 amino acid sequence was used to search the NBRF data base for the most homologous
`human antibody. Since molecular models of antibodies show strong interactions between the heavy and
`light chains, it was decided to use the heavy and light chains from the same human antibody. The human
`20 EU antibody turned out to be the best overall choice. The homology between the BMA 031 and EU FRs
`(#1-3) was 66% for the H chain and 63% for the L chain. The BMA 031 antibody uses JH3 and Jx5. These
`are most homologou·s to human JH4 and Jx4. A first generation civilized BMA 031 antibody would contain
`BMA 031 CDRs, EU FRs and homolgous human J regions. This antibody is referred to as BMA 031-
`EUCIV1 (Table 6A and 68).
`A refinement to this basic civilized version can advantage_ously be made in the sequence immediately
`before and after the CDRs. The CDRs are assigned based on sequence homology data (Kabat et al., 1987).
`Molecular models of antibodies have shown that the actual CDR loops can contain amino acids up to 4
`amino acids away from the "Kabat" CDRs. Therefore, maintaining at least the major amino acid differences
`{in size or charge) within 4 amino acids of the CDRs as murine may be beneficial. This antibody is referred
`to as BMA 031-EU