EA050318B1 - ANTI-LAG-3 ANTIBODY WITH INCREASED STABILITY - Google Patents

Abstract

The invention relates to isolated monoclonal antibodies that specifically bind LAG-3 and exhibit optimized functional properties compared to previously described anti-EAC-3 antibodies, such as the 25F7 antibody (US Patent Application 2011/0150892 A1). These antibodies comprise deleted deamidation sites while maintaining high binding affinity for human LAG-3 and physical (i.e., thermal and chemical) stability. Also described are nucleic acid molecules encoding the antibodies of the invention, expression vectors, host cells and methods for expressing the antibodies of the invention, as well as immunoconjugates, bispecific molecules and pharmaceutical compositions comprising these antibodies. The invention also relates to methods for detecting LAG-3, as well as to methods of treatment based on stimulating immune responses using the anti-EAC-3 antibody of the invention. Combination therapy is also described, in which the antibodies are administered together with at least one additional immunostimulatory antibody.

Description

State of the art

Therapeutic antibodies represent one of the fastest growing segments of the pharmaceutical industry. To maintain efficacy (i.e., potency) and minimize immunogenicity, antibodies and other protein drugs must be protected from physical and chemical degradation during manufacturing and storage. Indeed, one of the major challenges in developing antibody-based drugs is the potential immunogenic response upon administration to a subject, which may result in rapid clearance or even cause life-threatening side effects including anaphylactic shock. The immunogenicity of an antibody is influenced by various factors, such as its physiochemical properties (e.g., purity, stability, or solubility), clinical factors (e.g., dose, route of administration, disease heterogeneity, or patient characteristics), and concomitant treatment with other agents (Swann et al. (2008) Curr Opinion Immuol 20:493–499).

Antibody immunogenicity and/or loss of activity by antibodies is often associated with deamidation. Deamidation is a chemical degradation process that occurs spontaneously in proteins (e.g., antibodies). Deamidation removes the amide functional group from an amino acid residue such as asparagine and glutamine, thereby damaging its amide-containing side chains. This, in turn, causes structural and biological changes throughout the protein, thus creating heterogeneous forms of the antibody. Deamidation is one of the most common post-translational modifications that occur in recombinantly produced therapeutic antibodies. For example, heterogeneity in the heavy chain of the monoclonal antibody h1B4 (a humanized anti-CD18 antibody) due to deamidation during cell culture was noted by Tsai et al. (Pharm Res 10(11): 1580 (1993)). Furthermore, loss/reduction of biological activity due to deamidation has been a known problem in the past. For example, Kroon et al. characterized various deamidation sites in the therapeutic antibody OKT3 and reported that samples from a batch of OKT3 (ranging from 14 months to 3 years old) had activity less than 75% (Pharm Res 9(11): 1386 (1992), p. 1389, second column). Furthermore, OKT3 samples with large amounts of oxidized peptides in their maps had significantly reduced activity in an antigen binding efficiency assay (p. 1390, first column). The authors concluded that specific sites of chemical modification that occur during storage of OKT3 were identified by peptide mapping and correlated with the observed changes in the chemical assays and biological assays of the antibody (p. 1392, first column). Loss of biological activity has also been noted for a number of other deamidated therapeutic proteins, including recombinant human DNase (Cacia et al. (1993) J. Chromatogr. 634:229–239) and recombinant soluble CD4 (Teshima et al. (1991) Biochemistry 30:3916–3922).

Overall, deamidation presents a significant and unpredictable challenge to the pharmaceutical industry. The effort associated with monitoring the variability caused by deamidation in antibody therapeutics in particular, as well as FDA concerns about this variability, increases costs and delays clinical trials. Moreover, modifications associated with this problem, including changing conditions (e.g., temperature, pH, and cell type) associated with recombinant production and/or altering amino acids susceptible to deamidation (e.g., site-directed mutagenesis), can adversely affect stability and activity, particularly when changes are made to the complementarity-determining regions (CDRs) of the antibody. Accordingly, there is a need for more stable versions of therapeutic antibodies.

The essence of the invention

The present invention relates to isolated monoclonal antibodies (e.g., human monoclonal antibodies) that bind LAG-3 (e.g., human LAG-3) and have optimized physical stability compared to previously described anti-LAG-3 antibodies. In particular, the invention relates to a modified form of the 25F7 antibody (US 2011/0150892 A1) that has significantly increased thermal and chemical stability compared to the unmodified antibody. In particular, by altering the critical binding region of the heavy chain CDR2 domain of the 25F7 antibody, it was shown that the modified antibody exhibited significantly higher physical and thermal stability, reduced deamidation, higher thermal reversibility, and lower aggregation. At the same time, it was unexpectedly found that the modified antibody retains the same high binding affinity to human LAG-3 and functional activity of the unmodified antibody, including the ability to inhibit LAG-3 binding to class II major histocompatibility complex molecules and stimulate antigen-specific T-cell responses. The combined significant increase in stability and retention of binding efficiency/biological activity of the modified antibody was unexpected, especially in light of the criticality of CDR regions for antibody function.

The antibodies of the invention can be used for a variety of purposes, including detection of LAG-3.

- 1 050318 protein and stimulation of antigen-specific T-cell responses in subjects with tumors or viruses.

Accordingly, in one aspect, the invention provides an isolated monoclonal antibody (e.g., a human antibody) or antigen-binding portion thereof with a heavy chain variable region comprising the amino acid sequence of SEQ ID NO: 12. In another embodiment, the antibody also comprises a light chain variable region comprising the amino acid sequence of SEQ ID NO: 14. In another embodiment, the antibody or antigen-binding portion thereof comprises the CDR1, CDR2, and CDR3 regions of a heavy chain variable region comprising the amino acid sequence of SEQ ID NO: 12 (e.g., SEQ ID NOs: 15, 16, and 17, respectively). In another embodiment, the antibody also comprises the CDR1, CDR2, and CDR3 regions of a light chain variable region comprising the amino acid sequence of SEQ ID NO: 12 (e.g., SEQ ID NOs: 18, 19, and 20, respectively). In a preferred embodiment, the antibody has improved physical properties (i.e., thermal and chemical stability) compared to antibody 25F7, while maintaining at least the same binding affinity for human LAG-3 as 25F7. For example, the antibody has reduced sequence variability in the heavy chain CDR2 region due to deamidation compared to antibody 25F7, such as about 2.5% or less amino acid sequence modification after 12 weeks at 4°C (i.e., in real-time stability assays as described herein) and/or about 12.0% or less amino acid sequence modification after 12 weeks at 40°C (i.e., under accelerated stress conditions as described herein), while maintaining a binding affinity for human LAG-3 of at least a _KD of about 1x10' ⁷ M or less (more preferably a KD of 1x10' ⁸ M or less, a KD of 5x10'9 M or less, or a _{KD of} 1x10'9 M or less). In another embodiment, the antibody has a thermal reversibility of about 40% in PBS at pH 8.0.

In another embodiment, the antibody has a higher melting point (indicating higher overall stability in vivo) compared to the unmodified antibody (Krishnamurthy R and Manning MC (2002) Curr Pharm Biotechnol 1:361-71). In one embodiment, the antibody has a T _M1 (initial unfolding temperature) value of greater than 60°C, such as greater than 65°C or greater than 70°C. The melting point of the antibody can be determined by differential scanning calorimetry (Chen et al. (2003) Pharm Res 20:1952-60; Ghirlando et al. (1999) Immunol Lett 68:47-52) or circular dichroism (Murray et al. (2002) J. Chromatogr Sci 40:343-9).

In another embodiment, the antibody is characterized by its resistance to rapid degradation. Antibody degradation can be determined by capillary electrophoresis (CE) and MALDI-MS (Alexander AJ and Hughes DE (1995) Anal Chem 67:3626-32).

In another embodiment, the antibody has minimal aggregation properties, such as aggregation of 25% or less, such as 20% or less, 15% or less, 10% or less, 5% or less, or 4% or less. Aggregation can lead to the initiation of an undesirable immune response and/or altered or undesirable pharmacokinetic properties. Aggregation can be measured using a variety of methods, including size exclusion column (SEC), high performance liquid chromatography (HPLC), and light scattering.

In another embodiment, the antibody also has at least one of the following properties:

(a) binding to monkey LAG-3;

(b) no binding to mouse LAG-3;

(c) inhibition of LAG-3 binding to major histocompatibility complex class II molecules; and (d) stimulation of immune responses, in particular antigen-specific T-cell responses.

Preferably, the antibody has at least two of the properties (a), (b), (c) and (d). More preferably, the antibody has at least three of the properties (a), (b), (c) and (d). Even more preferably, the antibody has all four of the properties (a), (b), (c) and (d).

In another embodiment, the antibody stimulates an antigen-specific T cell response, such as the production of interleukin-2 (IL-2) in an antigen-specific T cell response. In other embodiments, the antibody stimulates an immune response, such as an anti-tumor response (e.g., inhibition of tumor growth in an in vivo xenograft tumor model) or an autoimmune response (e.g., development of diabetes in NOD mice).

In another embodiment, the antibody binds an epitope of human LAG-3 comprising the amino acid sequence PGHPLAPG (SEQ ID NO: 21). In another embodiment, the antibody binds an epitope of human LAG-3 comprising the amino acid sequence HPAAPSSW (SEQ ID NO: 22) or PAAPSSWG (SEQ ID NO: 23).

In other embodiments, the antibody stains pituitary tissue using immunohistochemistry or does not stain pituitary tissue using immunohistochemistry. Antibodies of the invention

- 2 050318 may be full-length antibodies, such as of the IgG1, IgG2 or IgG4 isotype, optionally with a serine-to-proline mutation in the hinge region of the heavy chain constant region (at a position corresponding to position 241 as described by Angal et al. (1993) Mol. Immunol. 30:105-108), such that disulfide bridge heterogeneity between the heavy chains is reduced or eliminated. In one aspect, the constant region isotype is IgG4 with a mutation at amino acid residues 228, such as S228P. Alternatively, the antibodies may be antibody fragments, such as Fab, Fab' or Fab' ₂ fragments or single chain antibodies.

In another aspect of the invention, the antibody (or antigen-binding portion thereof) is part of an immunoconjugate that includes a therapeutic agent, such as a cytotoxin or a radioactive isotope, linked to the antibody. In another aspect, the antibody is part of a bispecific molecule that includes a second functional group (e.g., a second antibody) that has a different binding specificity than said antibody or antigen-binding portion thereof. Also included within the scope of the invention are compositions comprising antibodies or antigen-binding portions thereof, immunoconjugates or bispecific molecules of the invention, optionally linked to a pharmaceutically acceptable carrier. Also included within the scope of the invention are nucleic acid molecules of the invention encoding antibodies or antigen-binding portions thereof (e.g., variable regions and/or CDRs), as well as expression vectors comprising these nucleic acids and host cells comprising these expression vectors. Also included within the scope of the invention are methods for producing anti-LAG-3 antibodies using host cells comprising these expression vectors, which may include the following steps: (i) expressing the antibody in the host cell and (ii) isolating the antibody from the host cell.

In another aspect, the invention relates to methods for stimulating immune responses using anti-LAG-3 antibodies of the invention. In one embodiment, the method comprises stimulating an antigen-specific T cell response by contacting T cells with an antibody of the invention such that the antigen-specific T cell response is stimulated. In a preferred embodiment, the production of interleukin-2 is stimulated by the antigen-specific T cell. In another embodiment, the subject is a subject with a tumor, wherein an immune response against the tumor is stimulated. In another embodiment, the subject is a subject with a virus, wherein an immune response against the virus is stimulated.

In another embodiment, the invention relates to a method of inhibiting tumor cell growth in a subject, comprising administering to the subject an antibody or antigen-binding portion thereof of the invention, such that tumor growth in the subject is inhibited. In another embodiment, the invention relates to a method of treating a viral infection in a subject, comprising administering to the subject an antibody or antigen-binding portion thereof of the invention, such that the viral infection is treated in the subject. In another embodiment, these methods comprise administering a composition, bispecific antibody, or immunoconjugate of the invention. In another embodiment, the invention relates to a method of stimulating an immune response in a subject, comprising administering to the subject an antibody or antigen-binding portion thereof of the invention and at least one additional immunostimulatory antibody, such as an anti-PD-1 antibody, an anti-PD-L1 antibody, and/or an anti-CTLA-4 antibody, such that an immune response is stimulated in the subject, for example, to inhibit tumor growth or to stimulate an antiviral response. In one embodiment, the additional immunostimulatory antibody is an anti-PD-1 antibody. In another embodiment, the additional immunostimulatory agent is an anti-PD-LI antibody. In another embodiment, the additional immunostimulatory agent is an anti-CTLA-4 antibody. In another embodiment, the antibody or antigen-binding portion thereof of the invention is administered with a cytokine (e.g., IL-2 and/or IL-21) or a costimulatory antibody (e.g., anti-CD137 and/or anti-GITR antibody). The antibodies can be, for example, human, chimeric, or humanized antibodies.

In another aspect, the invention relates to anti-LAG-3 antibodies and compositions of the invention for use in the above methods or for the manufacture of a medicament for use in the above methods (e.g. for treatment).

Other features and advantages of the present disclosure will be apparent from the following detailed description and examples, which should not be construed as limiting. The contents of all references, Genbank databases, patents and published patent applications cited in this application are herein incorporated by reference in their entirety.

Brief description of the drawings

Fig. 1A shows the nucleotide sequence (SEQ ID NO: 1) and amino acid sequence (SEQ ID NO: 2) of the heavy chain variable region of human monoclonal antibody 25F7. The CDR1 (SEQ ID NO: 5), CDR2 (SEQ ID NO: 6), and CDR3 (SEQ ID NO: 7) regions are shown by a broken line and the V, D, and J germline derivations are indicated. The CDR regions are shown by a broken line using the Kabat system (Kabat et al. (1991) Sequences of Proteins of Immunological Interest,

- 3 050318

5th ed., U.S. Department of Health and Human Services, NIH Publication No. 91-3242).

Fig. 1B shows the nucleotide sequence (SEQ ID NO: 3) and the amino acid sequence (SEQ ID NO: 4) of the light chain variable region kappa of human monoclonal antibody 25F7. The CDR1 (SEQ ID NO: 8), CDR2 (SEQ ID NO: 9), and CDR3 (SEQ ID NO: 10) regions are shown by a dashed line and the V and J germline derivations are indicated. The full-length amino acid sequences of the heavy and light chains of antibody 25F7 are shown in SEQ ID NOs: 32 and 34, respectively.

Fig. 2A shows the nucleotide sequence (SEQ ID NO: 12) of the variable region of the heavy chain of the monoclonal antibody LAG-3.5. CDR1 (SEQ ID NO: 15), CDR2 (SEQ ID NO: 16) and CDR3 (SEQ ID NO: 17) regions are outlined. The full-length amino acid sequences of the heavy and light chains of the LAG3.5 antibody are shown in SEQ ID NOs: 35 and 37, respectively.

Fig. 2B shows the nucleotide sequence (SEQ ID NO: 13) and the amino acid sequence (SEQ ID NO: 14) of the variable region of the kappa light chain of the monoclonal antibody LAG3.5. The CDR1 (SEQ ID NO: 18), CDR2 (SEQ ID NO: 19) and CDR3 (SEQ ID NO: 20) regions are outlined.

Figure 3 shows the amino acid sequences of the heavy chain variable region CDR2 of the LAG-3 variants LAG3.5 (SEQ ID NO: 42), LAG3.6 (SEQ ID NO: 43), LAG3.7 (SEQ ID NO: 44), and LAG3.8 (SEQ ID NO: 45) compared to the amino acid sequence of the heavy chain variable region CDR2 of the 25F7 antibody (LAG3.1) (SEQ ID NO: 41) and the corresponding human germline sequence (SEQ ID NO: 27). The heavy chain variable region CDR2 of LAG3.5 differs from the heavy chain variable region CDR2 of 25F7 by the arginine (R) at position 54 (versus asparagine (N)) and the serine (S) at position 56 (versus asparagine (N)). The remaining CDRs of LAG3.5 25F7 are identical. Figure 3 also shows SEQ ID NO: 40.

Fig. 4A and 4B are graphs showing the binding activity (EC ₅₀ and affinity, respectively) of LAG3.1 (25F7), LAG3.2, LAG3.5, LAG3.6, LAG3.7, and LAG3.8 antibodies to activated human CD4+ T cells. Fig. 4B shows SEQ ID NOS: 41, 42, 45, 44, and 43, respectively, in order.

Fig. 5A, B, C, D and E are graphs showing the thermal melting curves (i.e., thermal stability) of LAG3.1 (25F7), LAG3.5, LAG3.6, LAG3.7 and LAG3.8 antibodies, respectively.

Fig. 6A, B, C, D and E are graphs showing the thermal reversibility curves (i.e., thermal stability) of LAG3.1 (25F7), LAG3.5, LAG3.6, LAG3.7 and LAG3.8 antibodies, respectively.

Fig. 7 is a graph showing the binding activity of LAG3.1 (25F7) and LAG3.5 antibodies to activated human CD4+ T cells and antigen binding (Biacore).

Figure 8 shows the results of peptide mapping using mass spectrometry (chemical modifications/molecular stability) for LAG3.1 (25F7) and LAG3.5 antibodies, reflecting deamidation and isomerization after 5 days of culture under accelerated stress conditions as described here.

Fig. 8 shows SEQ ID NOs: 46-52, respectively, in order.

Fig. 9 is a graph comparing the hydrophobicity profiles of LAG3.1 (25F7) and LAG3.5 antibodies.

Fig. 10 A, B, C, and D are graphs comparing the affinity and physical stability (i.e., thermal and chemical stability) of LAG3.1 and LAG3.5 antibodies at 4°C and 40°C, both accelerated stress conditions, and real-time stability assays as described herein.

Fig. 11 A and B are graphs comparing the proportion of modification of amino acid sequences of LAG3.1 and LAG3.5 antibodies at 4°C and 40°C.

Detailed description of the invention

One aspect of the invention includes an isolated monoclonal antibody or antigen-binding portion thereof that binds human LAG-3, comprising heavy and light chain variable regions, wherein the heavy chain variable region comprises the CDR1, CDR2, and CDR3 regions of the heavy chain variable region of SEQ ID NO: 12.

In one aspect, the antibody or antigen-binding portion thereof is characterized in that the CDR1, CDR2, and CDR3 regions of the heavy chain comprise the amino acid sequences of SEQ ID NOs: 15, 16, and 17, respectively.

In one aspect, the antibody or antigen-binding portion thereof is characterized in that the light chain variable region comprises the CDR1, CDR2, and CDR3 regions of the light chain variable region of SEQ ID NO: 14.

In one aspect, the antibody or antigen-binding portion thereof is characterized in that the CDR1, CDR2, and CDR3 regions of the light chain comprise the amino acid sequences of SEQ ID NOs: 18, 19, and 20, respectively.

In one aspect, the antibody or antigen-binding portion thereof is characterized in that the variable

- 4 050318 heavy chain region comprises the amino acid sequence of SEQ ID NO: 12.

In one aspect, the antibody or antigen-binding portion thereof is characterized in that the light chain variable region comprises the amino acid sequence of SEQ ID NO: 14.

One aspect of the invention includes an isolated monoclonal antibody or antigen-binding portion thereof that binds human LAG-3 comprising the CDR1, CDR2, and CDR3 regions of the heavy and light chain comprising the amino acid sequences of SEQ ID NOs: 15, 16, 17 and SEQ ID NOs: 18, 19, and 20, respectively.

One aspect of the invention includes an isolated monoclonal antibody or antigen-binding portion thereof that binds human LAG-3, comprising heavy and light chain variable regions comprising the amino acid sequences of SEQ ID NOs: 12 and 14, respectively.

One aspect of the invention includes an antibody or antigen-binding portion thereof according to any of the preceding claims, having one or a combination of the following properties:

(a) monkey LAG-3 binding;

(b) lack of mouse LAG-3 binding;

(c) binding of LAG-3 to class II major histocompatibility complex molecules;

(d) inhibition of LAG-3 binding to major histocompatibility complex class II molecules; or (e) stimulation of an immune response.

In one aspect, the antibody or antigen-binding portion thereof stimulates the production of interleukin-2 (IL-2) in an antigen-specific T cell response.

In one aspect, the antibody or antigen-binding portion thereof stimulates an antitumor immune response.

In one aspect, the antibody or antigen-binding portion thereof binds an epitope of human LAG-3 comprising the amino acid sequence PGHPLAPG (SEQ ID NO: 21).

In one aspect, the antibody or antigen-binding portion thereof binds an epitope of human LAG-3 comprising the amino acid sequence HPAAPSSW (SEQ ID NO: 22) or PAAPSSWG (SEQ ID NO: 23).

In one aspect, the antibody or antigen-binding portion thereof binds human LAG-3 with a KD value of 0.27x10'9 M or less.

In one aspect, the antibody or antigen-binding portion thereof is a human, humanized, or chimeric antibody.

In one aspect, the antibody or antigen-binding portion thereof is of the IgG1, IgG2, or IgG4 isotype.

In one aspect, the antibody or antigen-binding portion thereof is an antibody fragment or a single chain antibody.

One aspect of the invention includes a bispecific molecule comprising an antibody or antigen-binding portion thereof and a second antibody or antigen-binding portion thereof.

One aspect of the invention includes an immunoconjugate comprising an antibody or antigen-binding portion thereof linked to a therapeutic agent.

One aspect of the invention includes an immunoconjugate, characterized in that the therapeutic agent is a cytotoxin or a radioactive isotope.

One aspect of the invention includes a composition comprising an antibody or antigen-binding portion thereof, a bispecific molecule or immunoconjugate, and a pharmaceutically acceptable carrier.

In one aspect, the composition further comprises an anticancer agent.

In one aspect, the composition is characterized in that the agent is an antibody or a chemotherapeutic agent.

One aspect of the invention includes an isolated nucleic acid encoding a variable region of a heavy and/or light chain of an antibody or an antigen-binding portion thereof as defined above.

One aspect of the invention includes an expression vector comprising a nucleic acid as defined above.

One aspect of the invention includes a host cell comprising an expression vector as defined above.

One aspect of the invention includes a method for producing an anti-LAG-3 antibody, comprising expressing the antibody in a host cell as defined above and isolating the antibody from the host cell.

One aspect of the invention includes a method for stimulating an immune response in a subject, comprising administering an antibody or antigen-binding portion thereof as defined above, a bispecific molecule as defined above, or an immunoconjugate as defined above, to the subject such that an immune response is stimulated in the subject.

In one aspect, the method is characterized in that the subject is a subject with a tumor, wherein an immune response against the tumor is stimulated.

- 5 050318

In one aspect, the method is characterized in that the subject is a subject with a virus, wherein an immune response against the virus is stimulated.

In one aspect, the method is characterized in that the immune response is an antigen-specific T-cell response such that the antigen-specific T-cell response is stimulated.

In one aspect, the method is characterized in that the production of interleukin-2 by an antigen-specific T cell is stimulated.

One aspect of the invention includes a method for inhibiting tumor cell growth in a subject, comprising administering to the subject an antibody or antigen-binding portion thereof as defined above, a bispecific as defined above, or an immunoconjugate as defined above such that tumor growth in the subject is inhibited.

One aspect of the invention includes a method of treating a viral infection in a subject, comprising administering to the subject an antibody or antigen-binding portion thereof, as defined above, a bispecific, as defined above, or an immunoconjugate, as defined above, such that the viral infection in the subject is treated.

In one aspect, the method also comprises administering at least one additional immunostimulatory antibody.

In one aspect, the method is characterized in that at least one immunostimulatory additional antibody is an anti-PD-1 antibody.

In one aspect, the method is characterized in that at least one additional immunostimulatory antibody is an anti-PD-L1 antibody.

In one aspect, the method is characterized in that at least one additional immunostimulatory antibody is an anti-CTLA-4 antibody.

One aspect of the invention includes the use of an antibody or antigen-binding portion thereof as defined above, a bispecific as defined above, or an immunoconjugate as defined above, for stimulating an immune response, optionally an antigen-specific T-cell response, or inhibiting tumor cell growth or treating a viral infection in a subject.

One aspect of the invention includes the use of an antibody or antigen-binding portion thereof as defined above, a bispecific as defined above, or an immunoconjugate as defined above, in the manufacture of a medicament for stimulating an immune response, optionally an antigen-specific T-cell response, or inhibiting tumor cell growth or treating a viral infection in a subject.

For a better understanding of the present description, certain terms are defined below. Additional definitions are provided throughout the detailed description. The terms 25F7, 25F7 antibody, LAG3.1 antibody and LAG3.1 mean the human LAG-3-specific antibody described in US 2011/0150892 A1. The nucleotide sequence (SEQ ID NO: 1) encoding the variable region of the heavy chain of 25F7 (LAG3.1) and the corresponding amino acid sequence (SEQ ID NO: 2) are shown in Fig. 1A (with the CDR sequences designated as SEQ ID NOs: 4, 5 and 7, respectively). The nucleotide sequence (SEQ ID NO: 3) encoding the variable region of the light chain of 25F7 (LAG3.1) and the corresponding amino acid sequence (SEQ ID NO: 4) are shown in Fig. 1B (with CDR sequences designated as SEQ ID NOs: 8, 9 and 10, respectively).

The term LAG-3 means lymphocyte activation gene 3. The term LAG-3 includes variants, isoforms, homologs, orthologs, and paralogs. For example, antibodies specific for a human LAG-3 protein may, in some cases, cross-react with a LAG-3 protein from a non-human species. In other embodiments, antibodies specific for a human LAG-3 protein may be completely specific for a human LAG-3 protein and may lack species or other types of cross-reactivity, or may cross-react with LAG-3 from some other species but not all other species (e.g., cross-react with monkey LAG-3 but not mouse LAG-3). The term human LAG-3 means a human LAG-3 sequence, such as the complete amino acid sequence of human LAG-3, which has the Genbank accession number NP_002277 (SEQ ID NO: 29). The term mouse LAG-3 means a mouse LAG-3 sequence, such as the complete amino acid sequence of mouse LAG-3, which has the Genbank number NP_032505. LAG-3 is also known in the art as, for example, CD223. The sequence of human LAG-3 may differ from human LAG-3, which has the Genbank number NP_002277, by having, for example, conserved mutations or mutations in non-conserved regions, wherein the LAG-3 has substantially the same biological function as the LAG-3, which has the Genbank number NP_002277. For example, the biological function of human LAG-3 is to have an epitope in the extracellular domain of LAG-3 that is specifically bound by an antibody of the present disclosure, or the biological function of human LAG-3 is to bind to MHC class II molecules.

The term monkey LAG-3 is intended to describe LAG-3 proteins expressed by Old and New World monkeys, including, but not limited to, cynomolgus monkey LAG-3 and rhesus macaque LAG-3.

- 6 050318

The corresponding amino acid sequence of monkey LAG-3 is the amino acid sequence of rhesus macaque LAG-3, which is also deposited in Genbank under accession number XM_001108923. Another corresponding amino acid sequence of monkey LAG-3 is the alternative sequence of rhesus macaque clone pa23-5, as described in US 2011/0150892 A1. The alternative rhesus sequence has one difference in the amino acid at position 419 compared to the sequence deposited in Genbank.

A particular human LAG-3 sequence will generally be at least 90% identical to the amino acid sequence of human LAG-3 Genbank NP_002277 and will contain amino acid residues that allow the amino acid sequence to be identified as being human when compared to the amino acid sequences of LAG-3 from other species (e.g., mouse). In certain cases, human LAG-3 may be at least 95%, or even at least 96%, 97%, 98%, or 99% identical in amino acid sequence to LAG-3 Genbank NP_002277. In certain embodiments, the human LAG-3 sequence will exhibit a difference of no more than 10 amino acids from the LAG-3 sequence of Genbank NP_002277. In certain embodiments, human LAG-3 may exhibit no more than 5, or even no more than 4, 3, 2, and 1 amino acid differences from the LAG-3 Genbank NP_002277 sequence. The percentage of identity may be determined as described herein.

The term immune response refers to the action of, for example, lymphocytes, antigen-presenting cells, phagocytic cells, granulocytes, and soluble macromolecules produced by the above cells or the liver (including antibodies, cytokines, and complement), which results in the selective killing, destruction, or removal from the human body of invasive pathogens, pathogen-infected cells or tissues, cancer cells, or, in the case of autoimmune or pathological inflammation, normal human cells or tissues.

Antigen-specific T cell responses refer to T cell-mediated responses that occur following stimulation of the T cell by an antigen for which the T cell has specificity. Non-limiting examples of T cell responses upon antigen-specific stimulation include proliferation and cytokine production (e.g., IL-2 production).

As used herein, the term "antibody" refers to whole antibodies and any antigen-binding fragment (i.e., antigen-binding portion) thereof, or individual chains thereof. Whole antibodies are glycoproteins comprising at least two heavy (H) chains and two light (L) chains linked together by disulfide bonds. Each heavy chain consists of a heavy chain variable region (hereinafter referred to as V _H ) and a heavy chain constant region. The heavy chain constant region consists of three domains, CH1, CH2, and _CH3 . Each light chain consists of a light chain variable region (hereinafter referred to as VL) and a light chain constant region. The light chain constant region consists of a single domain, CL, VH, and V _L. The regions can be further subdivided into regions of hypervariability, called complementarity determining regions (CDRs), interspersed with regions that are more conserved, called framework regions (FRs). Each VH and VL consists of three CDRs and four FRs arranged from amino terminus to carboxyl terminus in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4. The variable regions of the light and heavy chains contain the binding domain that interacts with antigen. The constant regions of antibodies can mediate the binding of immunoglobulin to host tissues or factors, including various cells of the immune system (e.g., effector cells) and the first component (C1q) of the classical complement system.

The term "antigen-binding portion of an antibody" (or simply "antibody portion") as used herein refers to one or more fragments of an antibody that retain the ability to specifically bind to an antigen (e.g., a LAG-3 protein). It has been shown that the antigen-binding function of an antibody can be performed by fragments of a full-length antibody. Examples of binding fragments encompassed by the term "antigen-binding portion of an antibody" include (i) the Fab fragment, a monovalent fragment consisting of the _VL , _VH , _CL , and _CH1 domains; (ii) the F(ab') ₂ fragment, a bivalent fragment comprising two Fab fragments linked by a disulfide bridge at the hinge region; (iii) the Fd fragment, consisting of the VH and CH1 domains; (iv) the Fv fragment, consisting of the VH and CH1 domains; (v) an Fv fragment consisting of the VL and VH domains of one arm of an antibody; (vi) a dAb fragment (Ward et al., (1989) Nature 341:544-546), which consists of the _VH domain; (vii) an isolated complementarity determining region (CDR); and (viii) a nanobody, a heavy chain variable region comprising one variable domain and two constant domains. Furthermore, although the two domains of the Fv fragment, VL and VH, are encoded by different genes, they can be joined recombinantly using a synthetic linker that enables them to be produced as a single protein chain in which the VL and VH regions are paired to form monovalent molecules (known as single-chain Fv (scFv); see, e.g., Bird et al. (1988) Science 242:423-426; and Huston et al. (1988) Proc. Natl. Acad. Sci. USA 85:5879-5883). These single-chain antibodies are also encompassed by the term antigen-binding portion of an antibody. These antibody fragments are produced using conventional techniques known to those skilled in the art,

- 7 050318 these fragments are then screened for utility in the same manner as intact antibodies.

As used herein, an isolated antibody means an antibody that is substantially free of other antibodies having different antigen specificities (e.g., an isolated antibody that specifically binds a LAG-3 protein is substantially free of antibodies that specifically bind antigens other than LAG-3 proteins). However, an isolated antibody that specifically binds a human LAG-3 protein may have cross-reactivity with other antigens, such as LAG-3 proteins from other species. Furthermore, an isolated antibody may be substantially free of other cellular material and/or chemical compounds.

As used herein, the terms monoclonal antibody or monoclonal antibody composition refer to a preparation of antibody molecules having the same molecular composition. A monoclonal antibody composition exhibits a single binding specificity and affinity for a particular epitope.

As used herein, the term "human antibody" includes antibodies having variable regions in which the framework and CDR regions are derived from human germline immunoglobulin sequences. Furthermore, if the antibody comprises a constant region, the constant region is also derived from human germline immunoglobulin sequences. Human antibodies of the invention may include amino acid residues not encoded by human germline immunoglobulin sequences (e.g., mutations introduced by random or site-directed mutagenesis in vitro, or somatic mutation in vivo). However, the term "human antibody" as used herein does not include antibodies in which CDR sequences derived from the germline of another mammalian species, such as a mouse, have been grafted onto human framework sequences.

The term human monoclonal antibody refers to antibodies exhibiting a single binding specificity that have variable regions in which the framework regions and CDR regions are derived from human germline immunoglobulin sequences. In one embodiment, human monoclonal antibodies are produced using a hybridoma that includes a B cell derived from a transgenic non-human animal, such as a transgenic mouse, having a genome comprising a human heavy chain transgene and a human light chain transgene fused to an immortalized cell.

The term "recombinant human antibody" as used herein includes all human antibodies that are produced, expressed, created, or isolated by recombinant methods, such as (a) antibodies isolated from an animal (e.g., a mouse) that is transgenic or transchromosomal for human immunoglobulin genes, or from a hybridoma derived therefrom (described below), (b) antibodies isolated from a host cell transformed to express a human antibody, such as from a transfectoma, (c) antibodies isolated from a recombinant combinatorial library of human antibodies, and (d) antibodies produced, expressed, created, or isolated by any other methods that involve splicing of human immunoglobulin gene sequences with other DNA sequences. These recombinant human antibodies include variable regions in which the framework and CDR regions are derived from human immunoglobulin germline sequences. However, in certain embodiments, these recombinant human antibodies may be subjected to in vitro mutagenesis (or, when an animal transgenic for human Ig sequences is used, in vivo somatic mutagenesis), and thus the amino acid sequences of both the VH and _VL regions of the recombinant antibodies are sequences that, although derived from or related to human germline VH and VL sequences, may not naturally exist in the human germline antibody repertoire in vivo.

The term isotype refers to the class of antibody (eg, IgM or IgG1) that is encoded by the heavy chain constant region genes.

The expressions "antibody recognizing an antigen" and "antibody specific for an antigen" are used herein interchangeably with the expression "antibody that specifically binds to an antigen."

The term human antibody derivative means any modified form of a human antibody, such as a conjugate of an antibody and another agent or antibody.

The term humanized antibody is intended to refer to antibodies in which CDR sequences derived from the germline of another mammalian species, such as a mouse, have been grafted onto human framework sequences. Additional framework modifications may be made to the human framework sequences.

The term chimeric antibody is intended to refer to antibodies in which the variable region sequences are derived from one species and the constant region sequences are derived from another species, such as an antibody in which the variable region sequences are

- 8 050318 are derived from a mouse antibody, while the constant region sequences are derived from a human antibody.

According to the meaning used herein, an antibody that specifically binds human LAG3 means an antibody that binds human LAG-3 protein (and optionally LAG-3 protein of one or more non-human species), but does not substantially bind to non-LAG-3 proteins. Preferably, the antibody binds human LAG-3 protein with high affinity, more particularly with a K _D value of 1x10' ⁷ M or less, more preferably 1x10' ⁸ M or less, more preferably 5x10'9 M or less, more preferably 1x10'9 M or less.

The term “substantially does not bind to a protein or cells,” as used herein, means “does not bind or does not bind with high affinity to a protein or cells,” i.e., binds to a protein or cells with a K _D of 1x10'6 M or higher, more preferably 1x10'5 M or higher, more preferably ^1x10'4 M or higher, more preferably 1x10'3 M or higher, and even more preferably 1x10'2 M or higher.

The term K _associ or K _a , as used herein, is intended to refer to the rate of association of a particular antibody-antigen interaction, while the term K _dis or K _d , as used herein, is intended to refer to the rate of dissociation of a particular antibody-antigen interaction. The term K _D , as used herein, is intended to refer to the dissociation constant, which is calculated from the ratio of K d to K _a (i.e., K d /K j ) and is expressed as a molar concentration (M). K _D values for antibodies can be determined using methods well known in the art. A preferred method for determining the K _D of an antibody is surface plasmon resonance, preferably using a biosensor system such as the Biacore® system.

The term "high affinity" for an IgG antibody means an antibody having a K _D for a target antigen of 1x10' ⁷ M or less, more preferably 5x10' ⁸ M or less, even more preferably 1x10' ⁸ M or less, even more preferably 5x10'9 M or less, and still more preferably 1x10'9 M or less. However, high binding affinity may differ for antibodies of other isotypes. For example, high binding affinity of the IgM isotype means an antibody with a K _D of 10' ⁶ M or less, more preferably 10' ⁷ M or less, more preferably 10' ⁸ M or less. The term "deamidation" means a chemical degradation process that occurs spontaneously in proteins (e.g., antibodies). Deamidation removes the amide functional group from an amino acid residue such as asparagine and glutamine, thereby damaging its amide-containing side chains. Specifically, the asparagine side chain attacks a nearby peptide group, forming a symmetrical succinimide derivative. The symmetry of the derivative results in two hydrolysis products, aspartate or isoaspartate. A similar reaction can occur in aspartate side chains, leading to partial conversion to isoaspartate. In the case of glutamine, the rate of deamidation is generally ten times slower than in the case of asparagine, but the mechanism is essentially the same, requiring only water molecules.

The term subject includes any animal, whether human or non-human. The term non-human animal includes all vertebrates, such as mammals, and non-mammals, such as non-human primates, sheep, dogs, cats, cows, horses, chickens, amphibians, and reptiles, although mammals are preferred, such as non-human primates, sheep, dogs, cats, cows, and horses.

Various aspects of the invention are described in more detail in the following subsections.

Anti-LAG-3 antibodies with increased stability and useful functional properties.

The antibodies of the invention specifically bind to human LAG-3 and have optimized stability compared to previously described anti-LAG-3 antibodies, in particular compared to the 25F7 (LAG3.1) antibody. This optimization includes reduced deamidation (e.g., increased chemical stability) and increased thermal refolding (e.g., increased physical stability), while maintaining high binding affinity for human LAG-3.

Methods for identifying deamidation sites are known in the art (see, for example, ion exchange, reverse phase, hydrophobic interaction chromatography and peptide mapping of proteolytic cleavage (HPLC-MS)). Suitable assays for measuring physical stability include, for example, melting point analysis and/or refolding of the antibody structure after denaturation (e.g., fractional reversibility, as described, for example, in Example 3, Section 3).

Binding to human LAG-3 can be assayed using one or more methods well known in the art. For example, the antibody can be tested using a flow cytometry-based assay in which the antibody is reacted with a cell line that expresses human LAG-3, such as CHO cells that have been transfected to express LAG-3 (e.g., human LAG-3 or simian (e.g., rhesus or cynomolgus) LAG-3 or mouse LAG-3) on their cell surface. Other suitable cells for use in a flow cytometry-based assay include anti-CD3-stimulated CD4+ activated

- 9 050318

T cells that express native LAG-3. Additionally or alternatively, antibody binding, including kinetic measures of binding (e.g., K _D value), can be tested using BIAcore assays. Other suitable binding assays include ELISA assays, such as those using recombinant LAG-3 protein.

The antibodies of the invention preferably bind to the human LAG-3 protein with a K _D value of 1x10' ⁷ M or less, more preferably 1x10' ⁸ M or less, 5x10'9 M or less, or 1x10' ⁹ M or less.

Typically, the antibody binds to LAG-3 in lymphoid tissues such as the tonsil, spleen, or thymus, which can be detected by immunohistochemistry. In one embodiment, the antibody stains pituitary tissue (e.g., accumulates in the pituitary gland), which is detected by immunohistochemistry. In another embodiment, the antibody does not stain pituitary tissue (i.e., does not accumulate in the pituitary gland), which is detected by immunohistochemistry.

Additional functional properties include cross-reactivity with LAG-3 of other species. For example, an antibody may bind to LAG-3 of a monkey (e.g., cynomolgus monkey, rhesus macaque), but does not substantially bind to LAG-3 of a mouse. Preferably, an antibody of the invention binds to human LAG-3 with high affinity.

Other functional properties include the ability of the antibody to stimulate an immune response, such as an antigen-specific T cell response. This can be determined, for example, by assaying the ability of the antibody to stimulate the production of interleukin-2 (IL-2) in an antigen-specific T cell response. In some embodiments, the antibody binds to human LAG-3 and stimulates an antigen-specific T cell response. In other embodiments, the antibody binds to human LAG-3 but does not stimulate an antigen-specific T cell response. Other means for assessing the ability of an antibody to stimulate an immune response include testing its ability to inhibit tumor growth, such as in an in vivo xenograft tumor model (see, for example, Example 6) or the ability to stimulate an autoimmune response, such as the ability to promote the development of an autoimmune disease in an autoimmune model, for example the ability to promote the development of diabetes in a NOD mouse model.

Preferred antibodies of the invention are human monoclonal antibodies. Additionally or alternatively, the antibodies may be, for example, chimeric or humanized monoclonal antibodies.

Monoclonal antibody LAG3.5

A preferred antibody of the invention is a human monoclonal antibody, LAG3.5, structurally and chemically characterized as described below in the Examples below. The VH amino acid sequence of LAG3.5 is shown in SEQ ID NO: 12 (FIG. 2A). The _VL amino acid sequence of LAG3.5 is shown in SEQ ID NO: 14 (FIG. 2B).

VH and _VL sequences (or CDR sequences) of other anti-LAG antibodies that bind human LAG-3 may be mixed and paired with the VH and VL sequences (or CDR sequences) of a LAG3.5 antibody. Preferably, when mixing and pairing the VH and VL chains (or CDRs within these chains), the VH sequence from a particular VH/VL pairing is replaced by a structurally similar V _H sequence. Likewise, the _VL sequence from a particular VH/VL pairing is replaced by a structurally similar VL sequence.

Accordingly, in one embodiment, the antibodies of the invention or antigen-binding portions thereof comprise:

(a) a heavy chain variable region comprising the amino acid sequence of SEQ ID NO: 12 (i.e., VH LAG3.5); and (b) a light chain variable region comprising the amino acid sequence of SEQ ID NO: 14 (i.e., V _L LAG3.5) or the VL of another anti-LAG3 antibody (i.e., other than LAG3.5);

The antibody specifically binds human LAG-3.

In another embodiment, the antibodies of the invention or antigen-binding portions thereof comprise:

(a) the CDR1, CDR2, and CDR3 regions of a heavy chain variable region comprising the amino acid sequence of SEQ ID NO: 12 (i.e., the CDRs of the LAG3.5 sequence, SEQ ID NOs: 15, 16, and 17, respectively); and (b) the CDR1, CDR2, and CDR3 regions of a light chain variable region comprising the amino acid sequences of SEQ ID NO: 14 (i.e., the CDRs of the LAG3.5 sequence, SEQ ID NOs: 18, 19, and 20, respectively) or the CDRs of another anti-LAG3 antibody (i.e., other than LAG3.5);

The antibody specifically binds human LAG-3.

In another embodiment, the antibody or antigen-binding portion thereof comprises the variable region CDR2 of the heavy chain of LAG3.5 combined with CDRs of other antibodies that bind human LAG-3, such as CDR1 and/or CDR3 from the variable region of the heavy chain and/or CDR1, CDR2 and/or CDR3 from the variable region of the light chain of another anti-LAG-3 antibody.

- 10 050318

Furthermore, it is well known in the art that the CDR3 domain, independently of the CDR1 and/or CDR2 domain(s), can alone determine the binding specificity of an antibody for its antigen, and that multiple antibodies with the same binding specificity can be predictably obtained based on a common CDR3 sequence. See, e.g., Klimka et al., British J. of Cancer 83(2):252–260 (2000); Beiboer et al., J. Mol. Biol. 296:833–849 (2000); Rader et al., Proc. Natl. Acad. Sci. U.S.A. 95:8910–8915 (1998); Barbas et al., J. Am. Chem. Soc. 116:2161–2162 (1994); Barbas et al., Proc. Natl. Acad. Sci. U.S.A. 92:2529-2533 (1995); Ditzel et al., J. Immunol. 157:739-749 (1996); Berezov et al., BIAjournal 8:Scientific Review 8 (2001); Igarashi et al., J. Biochem (Tokyo) 117:452-7 (1995); Bourgeois et al., J. Virol 72:807-10 (1998); Levi et al., Proc. Natl. Acad. Sci. U.S.A. 90:4374-8 (1993); Polymenis and Stoller, J. Immunol. 152:5218-5329 (1994) and Xu and Davis, Immunity 13:37-45 (2000). See also U.S. Patents 6,951,646; 6,914,128; 6,090,382; 6,818,216; 6,156,313; 6,827,925; 5,833,943; 5,762,905; and 5,760,185. Each of these publications is hereby incorporated by reference in its entirety.

Accordingly, in another embodiment, the antibodies of the invention comprise a CDR2 of the variable region of the heavy chain of LAG3.5 and at least a CDR3 of the variable region of the heavy and/or light chain of LAG3.5 (SEQ ID NO: 17 and/or 20) or a CDR3 of the variable region of the heavy and/or light chain of another LAG-3 antibody, wherein the antibody is capable of specifically binding to human LAG-3. These antibodies preferably (a) compete for binding; (b) retain functional characteristics; (c) bind to a single epitope; and/or (d) have the same binding affinity as HLAG3.5. In another embodiment, the antibodies may also comprise a CDR2 of the light chain variable region of LAG3.5 (SEQ ID NO: 17 and/or 20) or a CDR2 of the light chain variable region of another LAG-3 antibody, wherein the antibody is capable of specifically binding to human LAG-3. In another embodiment, the antibodies of the invention may also comprise a CDR1 of the heavy and/or light chain variable region of LAG3.5 (SEQ ID NO: 17 and/or 20) or a CDR1 of the heavy/or light chain variable region of another LAG-3 antibody, wherein the antibody is capable of specifically binding to human LAG-3.

Conservative modifications

In another embodiment, the antibodies of the invention comprise heavy and/or light chain variable region sequences CDR1, CDR2, and CDR3 sequences that differ from those of LAG3.5 by one or more conservative modifications. However, in a preferred embodiment, residues 54 and 56 of CDR2 V _H are maintained as arginine and serine, respectively (i.e., are not mutated). It is understood that such conservative modifications of the sequences can be made that do not result in a loss of antigen binding ability. See, e.g., Brummell et al. (1993) Biochem 32:1180-8; de Wildt et al. (1997) Prot. Eng. 10:835-41; Komissarov et al. (1997) J. Biol. Chem. 272-26864-26810; Hall et al. (1992) J. Immunol. 149:1605-12; Kelley and O'Connell (1993) Biochem. 32:6862-35; Adib-Conquy et al. (1998) Int. Immunol. 10:341-6 and Beers et al. (2000) Clin. Can. Res. 6:2835-43. Accordingly, in one embodiment, the antibody comprises a heavy chain variable region comprising CDR1, CDR2, and CDR3 sequences and/or a light chain variable region comprising CDR1, CDR2, and CDR3 sequences, wherein:

(a) the CDR1 sequence of the variable region of the heavy chain comprises SEQ ID NO: 15 and/or conservative modifications thereof, except for positions 54 and 56; and/or (b) the CDR3 sequence of the variable region of the heavy chain comprises SEQ ID NO: 17 and conservative modifications thereof; and/or (c) the CDR1 and/or CDR2 and/or CDR3 sequences of the variable region of the light chain comprise SEQ ID NO: 18 and/or SEQ ID NO: 19 and/or SEQ ID NO: 20 and/or conservative modifications thereof; and (d) the antibody specifically binds human LAG-3.

Additionally or alternatively, the antibody may have one or more of the following functional properties described above, such as high affinity binding to human LAG-3, binding to monkey LAG-3, lack of binding to mouse LAG-3, the ability to inhibit the binding of LAG-3 to MHC class II molecules, and/or the ability to stimulate antigen-specific T cell responses.

In various embodiments, the antibody can be, for example, a human antibody, a humanized antibody, or a chimeric antibody.

As used herein, the term conservative sequence modifications means amino acid modifications that do not significantly affect or alter the binding characteristics of an antibody comprising the amino acid sequence. These conservative modifications include amino acid substitutions, additions or deletions. Modifications can be introduced into an antibody of the invention by standard techniques known in the art, such as site-directed mutagenesis and PCR-mediated mutagenesis. Conservative amino acid substitutions are those in which an amino acid residue is replaced by another amino acid residue with a similar side chain. The art defines these

- 11 050318 families of amino acid residues having similar side chains. Such families include amino acids with basic side chains (e.g. lysine, arginine, histidine), acidic side chains (e.g. aspartic acid, glutamic acid), uncharged polar side chains (e.g. glycine, asparagine, glutamine, serine, threonine, tyrosine, cysteine, tryptophan), non-polar side chains (e.g. alanine, valine, leucine, isoleucine, proline, phenylalanine, methionine), beta-branched side chains (e.g. threonine, valine, isoleucine) and aromatic side chains (e.g. tyrosine, phenylalanine, tryptophan, histidine). Thus, one or more amino acid residues in the CDR regions of an antibody of the invention can be replaced with other amino acid residues of the same side chain family, and the altered antibody can be tested for retention of function (namely, the functions noted above) using the functional assays described herein.

Engineered and modified antibodies

Antibodies of the invention can be prepared using an antibody having one or more VH and/or V _L sequences of LAG3.5 as a starting material for constructing a modified antibody. The antibody can be engineered by modifying one or more residues in one or both variable regions (i.e., V _H and/or V _L ), such as in one or more CDR regions and/or in one or more framework regions. Additionally or alternatively, the antibody can be engineered by modifying residues in the constant region(s), such as to alter the effector function(s) of the antibody. In certain embodiments, CDR grafting can be used to construct the variable regions of antibodies. Antibodies interact with target antigens primarily through amino acid residues that are located in the six complementarity determining regions (CDRs) of the heavy and light chain. For this reason, amino acid sequences in the CDRs are more diverse between individual antibodies than sequences outside the CDRs. Because CDR sequences are responsible for most antibody-antigen interactions, it is possible to express recombinant antibodies that mimic the properties of specific naturally occurring antibodies by constructing expression vectors that incorporate CDR sequences from a specific naturally occurring antibody grafted onto framework sequences from another antibody with different properties (see, e.g., Riechmann et al. (1998) Nature 332:323–327; Jones et al. (1986) Nature 321:522–525; Queen et al. (1989) Proc. Natl. Acad. See. USA 86:10029–10033; U.S. Patents 5,225,539; 5,530,101; 5,585,089; 5,693,762; and 6,180,370).

Accordingly, another embodiment of the invention relates to an isolated monoclonal antibody, an antigen-binding portion thereof, comprising a heavy chain variable region comprising CDR1, CDR2 and CDR3 sequences comprising SEQ ID NO: 15, 16, 17, respectively, and/or a light chain variable region comprising CDR1, CDR2 and CDR3 sequences comprising SEQ ID NO: 18, 19, 20, respectively (i.e., CDR LAG3.5). While these antibodies comprise the VH and V _L CDR sequences of the monoclonal antibody LAG3.5, they may comprise other framework sequences.

These framework sequences can be obtained from publicly available DNA databases or published sources that include germline antibody gene sequences. For example, germline DNA sequences for the human heavy and light chain variable region genes can be found in the VBase human germline sequence database (available at www.mrc-cpe.cam.ac.uk/vbase) and in Kabat et al. (1991), cited above; Tomlinson et al. (1992) The Repertoire of Human Germline VH Sequences Reveals about Fifty Groups of V _H Segments with Different Hypervariable Loops J. Mol. Biol. 227:776–798; and Cox et al. (1994) A Directory of Human Germ-line V _H Segments Reveals a Strong Bias in their Usage Eur. J. Immunol. 24:827–836; the contents of each of which are incorporated herein by reference in their entirety. As another example, the germline DNA sequences of the human heavy and light chain variable region genes can be found in the Genbank database. For example, the following germline heavy chain sequences of the mouse HCo7 HuMAb are available from Genbank under the accession numbers: 1–69 (NG_0010109, NT_024637, & BC070333), 3–33 (NG_0010109 & NT_024637), and 3–7 (NG_0010109 & NT_024637). As another example, the following heavy chain germline sequences that comprise the mouse HCo12 HuMAb are available in GenBank under accession numbers: 1-69 (NG_0010109, NT_024637 & BC070333), 5-51 (NG_0010109 & NT_024637), 4-34 (NG_0010109 & NT_024637), 3-30.3 (CAJ556644) & 3-23 (AJ406678).

The protein sequences of the antibody are compared with the protein sequences contained in a compiled database using one of the sequence similarity search methods called Gapped BLAST (Altschul et al. (1997), supra), which is well known in the art.

Preferred framework sequences for use in antibodies of the invention are antibodies that are structurally similar to the framework sequences used in selected antibodies of the invention, for example, similar to the V _H 4-34 framework sequences.

- 12 050318 framework regions and/or V _K L6 framework sequences used in the preferred monoclonal antibodies of the invention. The CDR1, CDR2, and CDR3 of the VH sequences and the CDR1, CDR2, and CDR3 of the V _K sequences may be grafted into framework regions that have a sequence identical to a sequence found in the germline immunoglobulin gene from which the framework sequence is derived, or such CDR sequences may be grafted into framework regions that contain one or more mutations compared to the germline sequences. For example, it has been found that in some cases it is advantageous to mutate residues in framework regions to maintain or enhance the antigen-binding ability of the antibody (see, e.g., U.S. Patents 5,530,101; 5,585,089; 5,693,762; and 6,180,370).

Another type of variable region modification is the mutation of amino acid residues in CDR1, CDR2 and/or CDR3 of the VH and/or _VL regions to thereby improve one or more binding properties (e.g., affinity) of the antibody of interest. Site-directed mutagenesis or PCR-mediated mutagenesis may be used to introduce the mutation(s), and the effect on antibody binding or other functional property of interest may be assessed in the in vitro or in vivo assays described herein and in the examples section. Preferably, conservative modifications (discussed above) are introduced. The mutations may be amino acid substitutions, additions or deletions, but are preferably substitutions. Furthermore, typically no more than one, two, three, four or five residues in a CDR region are modified.

Accordingly, in another embodiment, the invention provides isolated anti-LAG-3 monoclonal antibodies or antigen-binding portions thereof, comprising a heavy chain variable region comprising: (a) a VH CDR1 region comprising SEQ ID NO: 15 or an amino acid sequence having one, two, three, four, or five amino acid substitutions, deletions, or additions compared to SEQ ID NO: 15; (b) a VH CDR2 region comprising SEQ ID NO: 16 or an amino acid sequence having one, two, three, four, or five amino acid substitutions, deletions, or additions compared to SEQ ID NO: 16 (preferably wherein positions 54 and 56 are the same as in SEQ ID NO: 16); (c) a V _H CDR3 region comprising SEQ ID NO: 17 or an amino acid sequence having one, two, three, four, or five amino acid substitutions, deletions, or additions compared to SEQ ID NO: 17; (d) a V _L CDR1 region comprising SEQ ID NO: 18 or an amino acid sequence having one, two, three, four, or five amino acid substitutions, deletions, or additions compared to SEQ ID NO: 18; (e) a V _L CDR2 region comprising SEQ ID NO: 19 or an amino acid sequence having one, two, three, four, or five amino acid substitutions, deletions, or additions compared to SEQ ID NO: 19; and (f) a V _L CDR3 region comprising SEQ ID NO: 20 or an amino acid sequence having one, two, three, four or five amino acid substitutions, deletions or additions compared to SEQ ID NO: 20. Engineered antibodies of the invention include antibodies in which modifications have been made to framework residues in the VH and/or VL, for example, to improve the properties of the antibody. Typically, such modifications to the framework region are made to reduce the immunogenicity of the antibody. For example, one approach is to mutate one or more framework residues back to the corresponding germline sequence residues. More particularly, an antibody that has been somatically mutated may comprise framework residues that differ from the germline sequence from which the antibody is derived. Such residues can be identified by comparing the framework sequences of the antibody to the germline sequences from which the antibody is derived.

Another type of framework modification involves mutating one or more residues in the framework region or even one or more CDR regions to remove T-cell epitopes, thereby reducing the potential immunogenicity of the antibody. This approach is also referred to as deimmunization and is described in more detail in U.S. Patent Publication 20030153043.

In addition to or as an alternative to modifications made in the framework or CDR regions, antibodies of the invention can be engineered to include modifications in the Fc region, typically to alter one or more functional properties of the antibody, such as serum half-life, complement fixation, Fc receptor binding, and/or antigen-dependent cellular cytotoxicity. Moreover, an antibody of the invention can also be chemically modified (e.g., one or more chemical groups can be attached to the antibody) or can be modified to alter its glycosylation, again to alter one or more functional properties of the antibody. Each of these embodiments is described in more detail below. The numbering of residues in the Fc region corresponds to the Kabat EU index numbering.

In a preferred embodiment, the antibody is an IgG4 isotype antibody comprising a serine to proline substitution mutation at a position corresponding to position 228 (S228P; EU index) in the hinge region of the heavy chain constant region. It was noted that this

- 13 050318 mutation eliminates heterogeneity of disulfide bridges between heavy chains in the hinge region (Angal et al. above; position 241 based on Kabat numbering system).

In one embodiment, the CH1 hinge region is modified such that the number of cysteine residues in the hinge region is altered, such as increased or decreased. This approach is also described in U.S. Patent 5,677,425. The number of cysteine residues in the CH1 hinge region is altered, such as to facilitate assembly of light and heavy chains or to increase or decrease the stability of the antibody.

In another embodiment, a mutation in the Fc hinge region of an antibody reduces the biological half-life of the antibody. More specifically, one or more amino acid mutations are introduced into the CH2-CH3 contact region of the Fc hinge domain such that binding of the antibody to Staphylococcyl protein A (SpA) is reduced compared to binding of the native Fc hinge domain of SpA. This approach is also described in U.S. Patent 6,165,745.

In another embodiment, the antibody is modified to increase its biological half-life. Various approaches are possible. For example, one or more of the following mutations can be introduced: T252L, T254S, T256F, as described in U.S. Patent 6,277,375. Alternatively, to increase the biological half-life, the antibody can be altered in the CH1 or CH2 region such that it contains a salvage receptor binding epitope formed from two loops of the CH2 domain of the Fc region of IgG, as described in U.S. Patents 5,869,046 and 6,121,022.

In other embodiments, the Fc region is altered by replacing at least one amino acid residue with a different amino acid residue to alter the effector function(s) of the antibody. For example, one or more amino acids selected from amino acid residues 234, 235, 236, 237, 297, 318, 320, and 322 can be replaced with a different amino acid residue such that the antibody has an altered affinity for the effector ligand but retains the antigen-binding ability of the original antibody. The effector ligand for which the affinity is altered can be, for example, an Fc receptor or a complement component C1. This approach is also described in U.S. Patents 5,624,821 and 5,648,260.

In another example, one or more amino acids selected from amino acid residues 329, 331, and 322 may be replaced with another amino acid residue such that the antibody has altered C1q binding and/or reduced or eliminated complement-dependent cytotoxicity (CDC). This approach is also described in U.S. Patent 6,194,551.

In another example, one or more amino acid residues at amino acid positions 231 and 239 are altered, thereby altering the ability of the antibody to fix complement. This approach is also described in PCT Publication WO 94/293 51.

In another example, the Fc region is modified to enhance the ability of the antibody to mediate antibody-dependent cellular cytotoxicity (ADCC) and/or to enhance the affinity of the antibody for the Fcy receptor by modifying one or more amino acids at the following positions: 238, 239, 248, 249, 252, 254, 255, 256, 258, 265, 267, 268, 269, 270, 272, 276, 278, 280, 283, 285, 286, 289, 290, 292, 293, 294, 295, 296, 298, 301, 303, 305, 307, 309, 312, 315, 320, 322, 324, 326, 327, 329, 330, 331, 333, 334, 335, 337, 338, 340, 360, 373, 376, 378, 382, 388, 389, 398, 414, 416, 419, 430, 434, 435, 437, 438 or 439. This approach is also described in PCT publication WO 00/42072. Furthermore, the binding sites on human IgG1 for FcyRI, FcyRII, FcyRIII, and FcRn have been mapped, and variants with improved binding have been described (see Shields et al. (2001) J. Biol. Chem. 276:6591–6604). Specific mutations at positions 256, 290, 298, 333, 334, and 339 have been shown to improve binding to FcyRIII. In addition, the following combined mutations have been shown to improve FcyRIII binding: T256A/S298A, S298A/E333A, S298A/K224A, and S298A/E333A/K334A.

In another embodiment, the glycosylation of the antibody is modified. For example, an aglycosylated antibody (i.e., an antibody without glycosylation) can be produced. Glycosylation can be altered to, for example, increase the affinity of the antibody for an antigen. Such carbohydrate modifications can be accomplished, for example, by altering one or more glycosylation sites in the antibody sequence. For example, one or more amino acid substitutions can be made that eliminate one or more glycosylation sites of the framework region of the variable region, thereby eliminating glycosylation at that site. This aglycosylation can increase the affinity of the antibody for the antigen. See, for example, U.S. Patents 5,714,350 and 6,350,861.

Additionally or alternatively, the antibody may be prepared such that it has an altered glycosylation pattern, such as a hypofucosylated antibody containing a reduced number of fucose residues or an antibody containing an increased number of biantennary GlcNac structures. Such altered glycosylation patterns have been shown to enhance the ADCC ability of antibodies. These carbohydrate modifications can be accomplished, for example, by expressing the antibody in a host cell with an altered glycosylation machinery. Cells with an altered glycosylation machinery have been described in the art and can be used as cells

- 14 050318 hosts for expressing therein recombinant antibodies of the invention, in order to thereby obtain an antibody with altered glycosylation. For example, in the cell lines Ms704, Ms705 and Ms709, the gene for fucosyltransferase, FUT8 (a(1,6)-fucosyltransferase), is absent, so that in the antibodies expressed in the Ms704, Ms705 and Ms709 cell lines, there is no fucose in the carbohydrates. FUTS ^-/- cell lines Ms704, Ms705 and Ms709 were generated by targeted disruption of the FUT8 gene in CHO/DG44 cells using two replacement vectors (see US Patent Publication 20040110704 and Yamane-Ohnuki et al. (2004) Biotechnol Bioeng 87:614-22). As another example, EP 1176195 describes a cell line with a functionally disrupted FUT8 gene, which encodes a fucosyltransferase, such that antibodies expressed in such a cell line exhibit hypofucosylation due to a reduction or elimination of the enzyme related to the a-1,6-linkage. EP 1 176 195 also describes cell lines that have low or no enzymatic activity for the addition of fucose to N-acetylglucosamine, which binds to the Fc region of an antibody, such as the rat myeloma cell line YB2/0 (ATCC CRL 1662). PCT Publication WO 03/035835 describes a variant of the CHO cell line, Lec13 cells, with a reduced ability to add fucose to Asn(297)-linked carbohydrates, which also results in hypofucosylation of antibodies expressed in the host cell (see also Shields et al. (2002) J. Biol. Chem. 277:26733-26740). Antibodies with a modified glycosylation profile can be produced in chicken eggs as described in PCT Publication WO 06/089231. Alternatively, antibodies with a modified glycosylation profile can be produced in plant cells such as Lemna. Methods for producing antibodies in a plant system are described in U.S. application Alston & Bird LLP Docket No. 040989/314911, filed August 11, 2006. PCT Publication No. WO 99/54342 describes cell lines engineered to express glycoprotein-modifying glycosyltransferases (e.g., β(1,4)-N-acetylglucosaminyltransferase III (GnTIII)) such that antibodies expressed in such engineered cell lines exhibit increased separation of the two halves of the GlcNac structures, resulting in increased ADCC activity of the antibodies (see also Umana et al. (1999) Nat. Biotech. 17:176-180). Alternatively, the fucose residues of the antibody can be cleaved using a fucosidase enzyme, for example fucosidase aL-fucosidase removes fucosyl residues from antibodies (Tarentino et al. (1975) Biochem. 14:5516-23).

Another modification of the antibodies of the invention is pegylation. An antibody can be pegylated, for example, to increase the biological half-life of the antibody (e.g., in serum). To pegylate an antibody, the antibody or fragment thereof is typically reacted with a polyethylene glycol (PEG), such as a reactive ester or aldehyde derivative of PEG, under conditions whereby one or more PEG groups are attached to the antibody or antibody fragment. Preferably, pegylation is accomplished via an acylation reaction or an alkylation reaction with a reactive PEG molecule (or a similar reactive water-soluble polymer). As used herein, the term polyethylene glycol encompasses any of the forms of PEG that have been used to derivatize other proteins, such as mono(C1-C10)alkoxy- or aryloxypolyethylene glycol or polyethylene glycol maleimide. In some embodiments, the antibody to be pegylated is an aglycosylated antibody. Methods for PEGylation of proteins are known in the art and can be applied to the antibodies of the invention. See, for example, EP 0 154 316 and EP 0 401 384.

Physical properties of antibodies

The antibodies of the invention can be characterized by their various physical properties to detect and/or distinguish their classes.

For example, antibodies may contain one or more glycosylation sites in the variable region of the heavy or light chain. These glycosylation sites may provide increased immunogenicity of the antibody or an altered pK of the antibody due to altered antigen binding (Marshall et al. (1972) Annu Rev Biochem 41:673–702; Gala and Morrison (2004) J Immunol 172:5489–94; Wallick et al (1988) J Exp Med 168: 1099–109; Spiro (2002) Glycobiology 12:43R–56R; Parekh et al. (1985) Nature 316:452–7; Mimura et al. (2000) Mol Immunol 37:697–706). Glycosylation has been shown to occur at motifs containing the N-X-S/T sequence. In some cases, it is preferable to have an anti-LAG-3 antibody without variable region glycosylation. This can be achieved either by selecting antibodies that do not contain a glycosylation motif in the variable region or by mutating residues in the glycosylation region.

In a preferred embodiment, the antibodies do not contain asparagine isomerism sites. Deamidation of asparagine can occur in N-G or D-G sequences and lead to the formation of an isoaspartic acid residue, which causes the formation of a loop in the polypeptide chain or a decrease in its stability (the isaspartic acid effect).

Each antibody has a unique isoelectric point (pI), which typically ranges from pH 6 to 9.5. The pI of IgG1 antibodies typically ranges from pH 7-9.5, and the pI of IgG4 antibodies typically ranges from pH 6-8. It has been suggested that antibodies with pIs outside of

- 15 050318 normal range may exhibit some folding disorder and instability under in vivo conditions. Therefore, it is preferable to obtain anti-LAG-3 antibodies whose pI value is in the normal range. This can be achieved either by selecting an antibody with a pI in the normal range or by mutating charged surface residues.

Nucleic acid molecules encoding antibodies of the invention

In another aspect, the invention provides nucleic acid molecules that encode the heavy and/or light chain variable regions, or CDRs, of the antibodies of the invention. The nucleic acids may be present in whole cells, in a cell lysate, or in partially purified or substantially pure form. A nucleic acid is isolated or substantially pure if it is purified from other cellular components or other contaminants, such as other cellular nucleic acids or proteins, by standard techniques, including alkaline/SDS treatment, CsCl density gradient centrifugation, column chromatography, agarose gel electrophoresis, and other techniques well known in the art. See, Ausubel, et al., ed. (1987) Current Protocols in Molecular Biology, Greene Publishing and Wiley Interscience, New York. A nucleic acid of the invention may be, for example, DNA or RNA, and may or may not contain intron sequences. In a preferred embodiment, the nucleic acid is a cDNA molecule.

The nucleic acids of the invention can be prepared using standard molecular biology techniques. In the case of antibodies expressed in hybridomas (e.g., hybridomas prepared from transgenic mice having human immunoglobulin genes, as described in more detail below), cDNAs encoding the light and heavy chains of the antibody produced using the hybridoma can be prepared by conventional PCR amplification or cDNA cloning techniques. In the case of antibodies prepared from an immunoglobulin gene library (e.g., using phage display techniques), nucleic acid encoding such antibodies can be obtained from a gene library.

Preferred nucleic acids of the invention include nucleic acids that encode the VH and V _L sequences of the monoclonal antibody LAG3.5 (SEQ ID NOs: 12 and 14, respectively) or CDRs. Once DNA fragments encoding the VH and V _L segments are obtained, these DNA fragments can be further manipulated by standard recombinant DNA techniques, for example, to convert the variable region genes into full-length antibody chain genes, into Fab fragment genes, or into an scFv gene. In these manipulations, the DNA fragment encoding V _L or VH is operably linked to another DNA fragment encoding another protein, such as an antibody constant region or a flexible linker. The term operably linked, as used herein, is intended to mean that the two DNA fragments are linked such that the amino acid sequences encoded by the two DNA fragments remain in reading frame.

The isolated DNA encoding the VH region can be converted into a full-length heavy chain gene by operatively linking the VH-encoding DNA to another DNA molecule encoding the heavy chain constant regions ( _CH1 , CH2, and CH3). The sequences of the human heavy chain constant region genes are known in the art (see Kabat et al. (1991), supra), and DNA fragments comprising such regions can be obtained by standard PCR amplification. The heavy chain constant region can be an IgG1, IgG2, IgG3, IgG4, IgA, IgE, IgM, or IgD constant region, but is most preferably an IgG1 or IgG4 constant region. For a Fab fragment of the heavy chain gene, the VH-encoding DNA can be operatively linked to another DNA molecule encoding only the _CH1 constant region of the heavy chain.

The isolated DNA encoding the V _L region can be converted into a full-length light chain gene (as well as a Fab light chain gene) by operably linking the V L-encoding DNA to another DNA molecule encoding the light chain constant region CL. Sequences of human light chain constant region genes are known in the art (see, e.g., Kabat et al., supra), and DNA fragments including such regions can be obtained by standard PCR amplification.

In preferred embodiments, the light chain constant region may be a kappa or lambda constant region.

To generate the scFv gene, the V _H - and V L -encoding DNA fragments are operably linked to another fragment encoding a flexible linker, e.g., encoding the amino acid sequence (Gly4 -Ser) ₃ (SEQ ID NO: 28), so that the V _H and V _L sequences can be expressed as a continuous single-chain protein with the V H and V L regions connected by a flexible linker (see, e.g., Bird et al. (1988) Science 242:423-426; Huston et al. (1988) Proc. Natl. Acad. Sci. USA 85:5879-5883; McCafferty et al., (1990) Nature 348:552-554).

Obtaining monoclonal antibodies according to the invention

The monoclonal antibodies (mAbs) of the present invention can be prepared using the well-known somatic cell hybridization (hybridoma) method of Kohler and Milstein (1975) Nature 256:495. Other embodiments for preparing monoclonal antibodies include methods

- 16 050318 based on viral or oncogenic transformation of B-lymphocytes and phage display. Chimeric or humanized antibodies are also well known in the art. See, for example, U.S. Patents 4,816,567; 5,225,539; 5,530,101; 5,585,089; 5,693,762 and 6,180,370, the contents of which are incorporated herein by reference in their entirety.

In a preferred embodiment, the antibodies of the invention are human monoclonal antibodies. Such human monoclonal antibodies directed against human LAG-3 can be produced using transgenic or transchromosomal mice having portions of the human immune system instead of the mouse system. These transgenic or transchromosomal mice include mice referred to herein as HuMAb Mouse® and KM Mouse®, respectively, and are collectively referred to herein as human Ig mice.

HuMAb Mouse® (Medarex®, Inc.) contains human immunoglobulin gene miniloci that encode rearranged sequences of the human immunoglobulin heavy (μ and γ) and light chains, together with targeted mutations that inactivate the endogenous μ and κ chain loci (see, e.g., Lonberg, (1994) Nature 368(6474): 856–859).

Thus, these mice exhibit reduced expression of mouse IgM or κ, and in response to immunization, the introduced human heavy and light chain transgenes undergo class switching and somatic mutation to produce high-affinity human monoclonal IgGK (Lonberg et al. (1994), supra; described in Lonberg (1994) Handbook of Experimental Pharmacology 113:49-101; Lonberg, N. and Huszar, D. (1995) Intern. Rev. Immunol. 13: 65-93, and Harding and Lonberg (1995) Ann. NY. Acad. Sci. 764:536-546). The production and use of HuMAb Mouse®, as well as the genomic modifications present in these mice, are also described in Taylor et al. (1992) Nucleic Acids Research 20:62876295; Chen et al. (1993) International Immunology 5: 647-656; Tuaillon et al. (1993) Proc. Natl. Acad. Sci. USA 90:3720-3724; Choi et al. (1993) Nature Genetics 4:117-123; Chen et al. (1993) EMBO J. 12: 821-830; Tuaillon et al. (1994) J. Immunol. 152:2912-2920; Taylor et al. (1994) International Immunology 6: 579-591; and Fishwild et al. (1996) Nature Biotechnology 14: 845-851, the contents of which are incorporated herein by reference in their entirety. See also U.S. Patents 5,545,806; 5,569,825; 5,625,126; 5,633,425; 5,789,650; 5,877,397; 5,661,016; 5,814,318; 5,874,299; 5,770,429; and 5,545,807; PCT Publications WO 92/03918; WO 93/12227; WO 94/25585; WO 97/13852; WO 98/24884; WO 99/45962; and WO 01/14424, the contents of which are hereby incorporated by reference in their entirety.

In another embodiment, the human antibodies of the invention can be produced using mice having human immunoglobulin sequences in transgenes and transchromosomes, for example mice having a human heavy chain transgene and a human light chain transchromosome. These mice are referred to herein as KM mouse® and are described in detail in PCT Publication No. WO 02/43478. A modified form of such a mouse, which is also characterized by a homozygous disruption of the endogenous FcγRIIB receptor gene, is also described in PCT Publication No. WO 02/43478 and is referred to herein as KM/FCGR2D mouse®. In addition, mice having HCo7 or HCo12 heavy chain transgenes, or both, can be used.

Additional transgenic animal variants include Xenomouse (Abgenix, Inc., U.S. Patents 5,939,598; 6,075,181; 6,114,598; 6,150,584; and 6,162,963). Other variants include TC mice (Tomizuka et al. (2000) Proc. Natl. Acad. Sci. USA 97:722-727) and cows carrying human heavy and light chain transchromosomes (Kuroiwa et al. (2002) Nature Biotechnology 20:889-894; PCT Publication No. WO 02/092812). The contents of these patents and publications are incorporated herein by reference in their entirety.

In one embodiment, the human monoclonal antibodies of the invention are produced using phage display methods to screen libraries of human immunoglobulin genes. See, for example, U.S. Patents 5,223,409; 5,403,484; 5,571,698; 5,427,908; 5,580,717; 5,969,108; 6,172,197; 5,885,793; 6,521,404; 6,544,731; 6,555,313; 6,582,915; and 6,593,081, the disclosures of which are hereby incorporated by reference in their entirety. The human monoclonal antibodies of the invention can also be produced using SCID mice in which human immune cells have been reconstituted such that a human antibody response can be obtained upon immunization. See, for example, U.S. Patents 5,476,996 and 5,698,767, the contents of which are incorporated herein by reference in their entirety.

In another embodiment, anti-LAG-3 human antibodies are produced using phage display, wherein the phages comprise nucleic acids encoding antibodies produced in transgenic animals previously immunized with LAG-3. In a preferred embodiment, the transgenic animal is a HuMab, KM, or Kirin mouse. See, for example, U.S. Patent 6,794,132, the disclosure of which is incorporated herein by reference in its entirety.

Immunization of mice with human Ig

In one embodiment, mice with human Ig are immunized with a purified or enriched preparation of LAG-3 antigen, recombinant LAG-3 protein, or cells expressing LAG-3 protein. See, e.g., Lonberg et al. (1994), supra; Fishwild et al. (1996), supra; PCT Publication Nos. WO 98/24884 or WO 01/14424, the contents of which are incorporated herein by reference in their entirety. In a preferred embodiment, mice aged 6-16 weeks are immunized with 5-50 μg

- 17 050318

LAG-3 protein. Alternatively, a fusion polypeptide of a portion of LAG-3 and a portion that is not LAG-3 is used.

In one embodiment, transgenic mice are immunized intraperitoneally (I/P) or intravenously (I/V) with LAG-3 antigen in complete Freund's adjuvant, followed by I/P or I/V immunizations with antigen in incomplete Freund's adjuvant. In other embodiments, adjuvants other than Freund's adjuvant or whole cells without adjuvant are used. Plasma can be screened using ELISA, and cells from mice with sufficient levels of anti-LAG-3 human immunoglobulin titer can be used to produce fusion constructs.

Obtaining hybridomas producing human monoclonal antibodies according to the invention.

To obtain hybridomas producing human monoclonal antibodies of the invention, splenocytes and/or lymph node cells from immunized mice can be isolated and fused with a suitable immortalized cell line, such as a mouse myeloma cell line. The resulting hybridomas can be screened for the production of antigen-specific antibodies. The preparation of hybridomas is well known in the art. See, for example, Harlow and Lane (1988) Antibodies, A Laboratory Manual, Cold Spring Harbor Publications, New York.

Preparation of Transfectomas Producing Monoclonal Antibodies of the Invention The antibodies of the invention may also be produced in a host cell transfectoma, for example, using a combination of recombinant DNA techniques and gene transfection techniques, as is well known in the art (e.g., Morrison, S. (1985) Science 229:1202). In one embodiment, DNA encoding portions or full-length light and heavy chains, prepared by standard molecular biology techniques, is inserted into one or more expression vectors such that the genes are operably linked to transcriptional and translational control sequences. As used herein, the term operably linked means that the antibody gene is ligated into the vector such that the transcriptional and translational control sequences in the vector perform their function of regulating the transcription and translation of the antibody gene.

The term regulatory sequence includes promoters, enhancers, and other expression control elements (e.g., polyadenylation signals) that control transcription or translation of antibody chain genes. These regulatory sequences are described, for example, by Goeddel (Gene Expression Technology. Methods in Enzymology 185, Academic Press, San Diego, CA (1990)). Preferred regulatory sequences for expression in a mammalian host cell include viral elements that result in high levels of protein expression in mammalian cells, such as promoters and/or enhancers derived from cytomegalovirus (CMV), simian virus 40 (SV40), adenovirus (e.g., adenovirus major late promoter (AdMLP)) and polyoma virus. Alternatively, nonviral regulatory sequences such as the ubiquitin promoter or the β-globin promoter may be used. In addition, regulatory elements composed of sequences from different sources can be used, such as the SRa promoter system, which contains sequences from the SV40 early promoter, and the long terminal repeat of human T-cell leukemia virus type 1 (Takebe et al. (1988) Mol. Cell. Biol. 8:466-472). The expression vector and expression control sequences are chosen to be compatible with the host cell used.

The antibody light chain gene and the antibody heavy chain gene may be inserted into the same or different expression vectors. In preferred embodiments, variable regions are used to create full-length antibody genes of any isotype by inserting them into expression vectors that already encode the heavy chain constant region and the light chain constant region of the desired isotype, such that the V _H segment is operably linked to the _CH segment(s) in the vector and the V _L segment is operably linked to the CL segment in the vector. Additionally or alternatively, the recombinant expression vector may encode a signal peptide that provides for secretion of the antibody chain from the host cell. The antibody chain gene may be cloned into the vector such that the signal peptide is linked in frame to the amino terminus of the antibody chain gene. The signal peptide may be an immunoglobulin signal peptide or a heterologous signal peptide (i.e., a signal peptide from a non-immunoglobulin protein).

In addition to the antibody chain genes and regulatory sequences, the recombinant expression vectors of the invention may carry additional sequences, such as sequences that regulate replication of the vector in host cells (e.g., origins of replication) and selectable marker genes. A selectable marker gene provides for selection of host cells into which the vector has been introduced (see, e.g., U.S. Patents 4,399,216, 4,634,665, and 5,179,017). For example, in a typical case, a selectable marker gene confers resistance to drugs, such as G418, hygromycin, or methotrexate, on a host cell into which the vector has been introduced. Preferred selectable marker genes include the dihydrofolate reductase (DHFR) gene (for use in dhfr host cells with selection/amplification in the presence of methotrexate) and the neo gene (for selection in the presence of G418).

- 18 050318

For expression of the light and heavy chains, the expression vector(s) encoding the heavy and light chains are transfected into a host cell using standard techniques. The various forms of the term transfection cover a wide range of techniques commonly used to introduce exogenous DNA into a prokaryotic or eukaryotic host cell, such as electroporation, calcium phosphate precipitation, DEAE-dextran transfection, and the like. Although it is theoretically possible to express the antibodies of the invention in prokaryotic or eukaryotic host cells, it is most preferred to express the antibodies in eukaryotic cells, and most preferably in mammalian host cells, since such eukaryotic cells, and in particular mammalian cells rather than prokaryotic cells, are suitable for the assembly and secretion of a suitably folded and immunologically active antibody.

Preferred mammalian host cells for expressing the recombinant antibodies of the invention include Chinese hamster ovary cells (CHO cells) (including dhfr CHO cells as described in Urlaub and Chasin, (1980) Proc. Natl. Acad. Sci. USA 77:4216-4220, used with a DHFR selectable marker, such as described in R. J. Kaufman and P. A. Sharp (1982) J. Mol. Biol. 159:601-621), NSO myeloma cells, COS cells and SP2 cells. In particular, for use with NSO myeloma cells, another preferred expression system is the GS gene expression system as described in WO 87/04462, WO 89/01036 and EP 338,841. When recombinant expression vectors encoding antibody genes are introduced into mammalian host cells, the antibodies are produced by culturing the host cells for a period of time sufficient to allow expression of the antibody in the host cells, or, more preferably, secreting the antibody into the culture medium in which the host cells are grown. The antibodies can be isolated from the culture medium using standard protein purification techniques.

Immunoconjugates

The antibodies of the invention can be conjugated to a therapeutic agent to form an immunoconjugate, such as an antibody-drug conjugate (ADC). Suitable therapeutic agents include antimetabolites, alkylating agents, DNA minor groove binding agents, DNA intercalates, DNA cross-linking agents, histone acetylase inhibitors, nuclear export inhibitors, proteasome inhibitors, topoisomerase I or II inhibitors, heat shock protein inhibitors, tyrosine kinase inhibitors, antibiotics and antimitotic agents. In an ADC, the antibodies and the therapeutic agent are preferably conjugated via a cleavable linker, such as a peptidyl, disulfide or hydrazone linker. More preferably, the linker is a peptidyl linker such as Val-Cit, Ala-Val, Val-Ala-Val, Lys-Lys, Pro-Val-Gly-Val-Val (SEQ ID NO: 39), Ala-Asn-Val, Val-Leu-Lys, Ala-Ala-Asn, Cit-Cit, Val-Lys, Lys, Cit, Ser or Glu. ADCs can be prepared as described in U.S. Patents 7,087,600; 6,989,452 and 7,129,261; PCT Publications WO 02/096910; WO 07/038658; WO 07/051081; WO 07/059404; WO 08/083312; and WO 08/103693; U.S. Patent Publications 20060024317; 20060004081 and 20060247295, the disclosures of which are incorporated herein by reference.

Bispecific molecules

In another aspect, the present disclosure discloses bispecific molecules comprising one or more antibodies of the invention linked to at least one other functional molecule, such as another peptide or protein (e.g., another antibody or a ligand for a receptor) to produce a bispecific molecule that binds to at least two different binding sites or target molecules. Thus, as used herein, the term bispecific molecule includes molecules having three or more specificities. In a preferred embodiment, the bispecific molecule comprises a first binding specificity for LAG-3 and a second binding specificity for a priming molecule that captures cytotoxic effector cells that can kill a LAG-3-expressing target cell. Examples of suitable priming molecules are CD64, CD89, CD16 and CD3. See, for example, Kufer et al., TRENDS in Biotechnology, 22(5), 238–244 (2004).

In one embodiment, the bispecific molecule has, in addition to anti-Fc binding specificity and anti-LAG-3 binding specificity, a third specificity. The third specificity can be directed toward an enhancement factor (EF), such as a molecule that binds to a surface protein involved in cytotoxic activity and thereby enhances an immune response against a target cell. For example, an anti-enhancement factor can bind a cytotoxic T cell (e.g., via CD2, CD3, CD8, CD28, CD4, CD40, or ICAM-1) or another immune cell, resulting in an enhanced immune response against the target cell.

Bispecific molecules can have a variety of formats and sizes. At one end of the size spectrum, a bispecific molecule retains the traditional antibody format except that instead of having two binding sites with identical specificity, it has two binding sites, each with a different specificity.

At the other end of the spectrum are bispecific molecules consisting of two single-chain antibody fragments (scFv's) linked by a peptide chain - the so-called Bs(scFv) ₂ construct. Bispecific

- 19 050318 Intermediate-sized digital molecules comprise two different F(ab) fragments linked by a peptide linker. Bispecific molecules of this and other formats can be produced by genetic modification, somatic hybridization, or chemical means. See, e.g., Kufer et al., cited supra; Cao and Suresh, Bioconjugate Chemistry, 9 (6), 635-644 (1998); and van Spriel et al., Immunology Today, 21 (8), 391-397 (2000), and references cited therein.

Pharmaceutical compositions

In another aspect, the present disclosure includes a pharmaceutical composition comprising one or more antibodies of the present invention combined with a pharmaceutically acceptable carrier. The composition may optionally contain one or more additional pharmaceutically active ingredients, such as another antibody or a drug. The pharmaceutical compositions of the invention may also be administered as part of a combination therapy, for example, with another immunostimulatory agent, an anticancer agent, an antiviral agent, or a vaccine, such that the anti-LAG-3 antibody stimulates an immune response against the vaccine. The pharmaceutical composition may include any number of excipients. Excipients that may be used include carriers, surface-active agents, thickening or emulsifying agents, solid binders, dispersing or suspending agents, solvents, colorants, flavors, coatings, disintegrating agents, lubricants, sweeteners, preservatives, isotonic agents, and combinations thereof. The selection and use of suitable excipients is described in Gennaro, ed., Remington: The Science and Practice of Pharmacy, 20th Ed. (Lippincott Williams & Wilkins 2003), the disclosure of which is incorporated herein by reference.

Preferably, the pharmaceutical composition is suitable for intravenous, intramuscular, subcutaneous, parenteral, spinal or epidermal administration (e.g. by injection or infusion). Depending on the route of administration, the active compound may be coated with a material to protect it from the effects of acids and other natural conditions that may lead to its inactivation. The expression parenteral administration, as used herein, means routes of administration other than enteral and topical administration, typically by injection, and includes, without limitation, intravenous, intramuscular, intraarterial, intrathecal, intracapsular, intraorbital, intracardiac, intradermal, intraperitoneal, transtracheal, subcutaneous, intradermal, intraarticular, subcapsular, subarachnoid, intracapsular, epidural and epigastric injection and infusion. Alternatively, the antibody of the invention may be administered by a non-parenteral route, such as topical, epidermal or mucosal administration, for example intranasally, orally, vaginally, rectally, sublingually or topically. The pharmaceutical compositions of the invention may include pharmaceutically acceptable salts. The term pharmaceutically acceptable salt means a salt that retains the desired biological activity of the parent compound and does not introduce any undesirable toxicological effects. Examples of such salts include acid addition salts and base addition salts. Acid addition salts include salts derived from non-toxic inorganic acids such as hydrochloric, nitric, phosphoric, sulfuric, hydrobromic, hydroiodic, phosphorous and the like, as well as non-toxic organic acids such as aliphatic mono- and dicarboxylic acids, phenyl-substituted alkanoic acids, hydroxyalkanoic acids, aromatic acids, aliphatic and aromatic sulfonic acids and the like. Base addition salts include salts derived from alkaline earth metals such as sodium, potassium, magnesium, calcium and the like, as well as non-toxic organic amines such as K,K'-dibenzylethylenediamine, N-methylglucamine, chloroprocaine, choline, diethanolamine, ethylenediamine, procaine and the like.

The pharmaceutical compositions may be in the form of sterile aqueous solutions or dispersions. They may also be in the form of a microemulsion, liposome or other ordered structure suitable for high drug concentration.

The amount of active ingredient that can be combined with a carrier material to prepare a single dosage form will vary depending on the subject being treated and the particular route of administration, and will generally correspond to the amount of the composition that produces a therapeutic effect. Typically, on a one hundred percent basis, this amount will be in the range of from about 0.01% to about ninety-nine percent of the active ingredient, preferably from about 0.1% to about 70%, most preferably from about 1% to about 30% of the active ingredient in combination with a pharmaceutically acceptable carrier. Dosage regimens are adjusted to provide the optimum desired response (e.g., response to therapy). For example, a single bolus may be administered, or several divided doses may be administered over time, or the dose may be proportionally reduced or increased depending on the therapeutic situation. It is especially advantageous to formulate parenteral compositions in unit dosage form for ease of administration and uniformity of dosage.

The term dosage form as used herein means physically discrete units suitable as unitary dosages for administration.

- 20 050318 subjects to be treated; each unit contains a predetermined quantity of active compound calculated to provide the desired therapeutic effect together with the required pharmaceutical carrier. Alternatively, the antibody may be administered as a sustained release formulation, which requires less frequent administration.

For administration of the antibody, the dosage is in the range of about 0.0001 to 100 mg/kg, and more usually 0.01 to 5 mg/kg, of the host body weight. For example, the dosages can be 0.3 mg/kg body weight, 1 mg/kg body weight, 3 mg/kg body weight, 5 mg/kg body weight, or 10 mg/kg body weight, or in the range of 1-10 mg/kg. An exemplary treatment regimen includes administration once weekly, once every two weeks, once every three weeks, once every four weeks, once monthly, once every 3 months, or once every 3-6 months. Preferred dosage regimens for the anti-LAG-3 antibody of the invention include 1 mg/kg body weight or 3 mg/kg body weight when administered intravenously, wherein the antibody is administered using one of the following dosage regimens: (i) every four weeks for the first six doses, then every three months; (ii) every three weeks; (iii) 3 mg/kg body weight once and then 1 mg/kg body weight every three weeks. In some methods, the dosage is adjusted to achieve a plasma antibody concentration of about 1-1000 μg/mL, and in some methods about 25-300 μg/mL.

A therapeutically effective dose of an anti-LAG-3 antibody of the invention preferably results in a reduction in the severity of disease symptoms, an increase in the frequency and duration of disease-free periods, or a prevention of worsening or disability due to the disease. For example, for treating subjects with a tumor, a therapeutically effective dose preferably inhibits tumor growth by at least 20%, more preferably by at least 40%, even more preferably by at least 60%, and more preferably by at least 80% relative to the untreated subject. A therapeutically effective amount of a therapeutic compound may reduce tumor size or otherwise alleviate symptoms in a subject, which is typically a human or other mammal.

The pharmaceutical composition may be a controlled release formulation, including implants, transdermal patches, and microencapsulated delivery systems. Biodegradable biocompatible polymers such as ethylene vinyl acetate, polyanhydrides, polyglycolic acid, collagen, polyorthoesters, and polylactic acid may be used. See, for example, Sustained and Controlled Release Drug Delivery Systems, J.R. Robinson, ed., Marcel Dekker, Inc., New York, 1978.

Therapeutic compositions can be administered using medical devices such as needle-less hypodermic injection devices (e.g., U.S. Patents 5,399,163; 5,383,851; 5,312,335; 5,064,413; 4,941,880; 4,790,824; and 4,596,556); (2) microinfusion pumps (U.S. Patent 4,487,603); (3) transdermal devices (U.S. Patent 4,486,194); (4) infusion apparatuses (U.S. Patents 4,447,233 and 4,447,224); and (5) osmotic devices (U.S. Patents 4,439,196 and 4,475,196); the disclosures of which are incorporated herein by reference. In some embodiments, the human monoclonal antibodies of the invention can be formulated to ensure proper distribution in vivo. For example, to ensure that the therapeutic compounds of the invention cross the blood-brain barrier, they can be formulated into liposomes which can further include targeting moieties to ensure selective transport to specific cells or organs. See, for example, U.S. Patents 4,522,811; 5,374,548; 5,416,016; and 5,399,331; V. V. Ranade (1989) J. Clin. Pharmacol. 29:685; Umezawae et. al., (1988) Biochem. Biophys. Res. Commun. 153:1038; Bloeman et. al. (1995) FEBS Lett. 357:140; M. Owais et. al. (1995) Antimicrob. Agents Chemother. 39:180; Briscoe et al. (1995) Am. J. Physiol. 1233:134; Schreier et al. (1994) J. Biol. Chem. 269:9090; Keinanen and Laukkanen (1994) FEBS Lett. 346:123; and Killion and Fidler (1994) Immunomethods 4:273.

Applications and methods according to the invention

The antibodies (compositions, bispecific antibodies and immunoconjugates) of the present invention have a variety of in vitro and in vivo applications, including, for example, detecting LAG-3 or stimulating immune responses by blocking LAG-3. In a preferred embodiment, the antibodies are human antibodies. Such antibodies can be administered to cells in culture, in vitro or ex vivo, or to human subjects, for example, in vivo, to stimulate immunity in various situations. Accordingly, in one aspect, the invention relates to a method of modifying an immune response in a subject, comprising administering to the subject an antibody or antigen-binding portion thereof of the invention to modify an immune response in the subject. Preferably, the response is enhanced, stimulated or upregulated.

Preferred subjects include human patients in need of an immune response enhancement. The methods are preferably suitable for treating human patients having a disorder that can be treated by enhancing an immune response (e.g., a T cell-mediated immune response). In a particular embodiment, the methods are particularly suitable for treating cancer in vivo. To achieve antigen-specific immune stimulation, the anti-LAG-3 antibodies can

- 21 050318 co-administered with the antigen of interest or the antigen may already be present in the subject being treated (e.g., a subject with a tumor or a virus). When administering antibodies to LAG-3 together with another agent, the two agents may be administered in any order or simultaneously.

The invention also relates to methods for detecting the presence of an antigen in a sample or measuring the amount of human LAG-3 antigen, comprising contacting the sample and a control sample with a human monoclonal antibody or an antigen-binding portion thereof that specifically binds to human LAG-3 under conditions that provide for the formation of a complex between the antibody or portion thereof and human LAG-3. The formation of the complex is then detected, wherein the difference between the formation of the complex between the sample and the control sample indicates the presence of human LAG-3 antigen in the sample. Moreover, the anti-LAG-3 antibodies of the invention can be used to purify human LAG-3 by immunoaffinity purification.

In view of the ability of the anti-LAG-3 antibodies of the invention to inhibit the binding of LAG-3 to MHC class II molecules and to stimulate antigen-specific T-cell responses, the invention also relates to in vitro and in vivo methods of using the antibodies to stimulate, enhance or upregulate antigen-specific T-cell responses. For example, the invention relates to a method of stimulating an antigen-specific T-cell response, comprising contacting said T-cell with an antibody of the invention such that the antigen-specific T-cell response is stimulated. Any suitable indicator of an antigen-specific T-cell response can be used to measure the antigen-specific T-cell response. Non-limiting examples of these suitable indicators include increased T-cell proliferation in the presence of the antibody and/or increased cytokine production in the presence of the antibody. In a preferred embodiment, the production of interleukin-2 is stimulated by an antigen-specific T-cell.

The invention also relates to a method for stimulating an immune response (e.g., an antigen-specific T-cell response) in a subject, comprising administering an antibody of the invention to the subject such that the immune response (e.g., an antigen-specific T-cell response) of the subject is stimulated. In a preferred embodiment, the subject is a subject with a tumor, wherein an immune response against the tumor is stimulated. In another preferred embodiment, the subject is a subject with a virus, wherein an immune response against the virus is stimulated.

In another embodiment, the invention relates to a method for inhibiting tumor cell growth in a subject, comprising administering to the subject an antibody of the invention, such that tumor growth in the subject is inhibited. In another embodiment, the invention relates to methods for treating a viral infection in a subject, comprising administering to the subject an antibody of the invention, such that the viral infection in the subject is treated.

These and other methods of the invention are described in more detail below.

Cancer

Blockade of LAG-3 by antibodies can enhance the immune response of a patient to cancer cells. In one aspect, the present invention relates to treating a subject in vivo using an anti-LAG-3 antibody such that the growth of cancer tumors is inhibited. The Ahtu-LAG-3 antibody can be used alone to inhibit the growth of cancer tumors. Alternatively, the anti-LAG-3 antibody can be used in conjunction with other immunogenic agents, standard cancer treatments, or other antibodies, as described below.

Accordingly, in one embodiment, the invention provides a method of inhibiting tumor cell growth in a subject, comprising administering to the subject a therapeutically effective amount of an anti-LAG-3 antibody or an antigen-binding portion thereof. Preferably, the antibody is a human anti-LAG-3 antibody (such as any of the human anti-human LAG-3 antibodies described herein). Additionally or alternatively, the antibody may be a chimeric or humanized anti-LAG-3 antibody.

Preferred types of cancer tumors whose growth can be inhibited using the antibodies of the invention include tumors that are typically responsive to immunotherapy. Non-limiting examples of preferred types of cancer for treatment include melanoma (e.g., metastatic malignant melanoma), kidney cancer (e.g., hypernephroma), prostate cancer (e.g., hormone-unresponsive prostate adenocarcinoma), breast cancer, colon cancer, and lung cancer (e.g., non-small cell lung cancer). The invention also includes difficult-to-treat or recurrent malignancies whose growth can be inhibited using the antibodies of the invention.

Examples of other cancers that can be treated using the methods of the invention include bone cancer, pancreatic cancer, skin cancer, head and neck cancer, malignant melanoma of the skin or intraocular melanoma, uterine cancer, ovarian cancer, rectal cancer, anal cancer, gastric cancer, testicular cancer, fallopian tube carcinoma, endometrial carcinoma,

- 22 050318 Cervical cancer, vaginal carcinoma, vulvar carcinoma, Hodgkin's disease, non-Hodgkin's lymphoma, esophageal cancer, small bowel cancer, endocrine cancer, thyroid cancer, parathyroid cancer, adrenal cortex cancer, soft tissue sarcoma, urethral cancer, penile cancer, chronic or acute leukemia including acute myeloid leukemia, chronic myeloid leukemia, acute lymphoblastic leukemia, chronic lymphocytic leukemia, childhood solid tumors, lymphocytic lymphoma, bladder cancer, kidney or ureter cancer, renal pelvis carcinoma, central nervous system (CNS) tumor, primary CNS lymphoma, tumor angiogenesis, spinal tumor, brain stem glioma, pituitary adenoma, sarcoma Kaposi's cell carcinoma, epidermoid carcinoma, squamous cell carcinoma, T-cell lymphoma, environmentally induced malignancies, including asbestos-induced malignancies, and combinations of these cancers. The present invention can also be used to treat metastatic cancers, in particular metastatic cancers associated with PD-L1 expression (Iwai et al. (2005) Int. Immunol. 17:133-144).

Anti-LAG-3 antibodies may optionally be combined with an immunogenic agent such as cancer cells, purified tumor antigens (including recombinant proteins, peptides, and carbohydrate molecules), cells, and cells transfected with genes encoding immunostimulatory cytokines (He et al. (2004) J. Immunol. 173:4919-28). Non-limiting examples of tumor vaccines that may be used include melanoma antigen peptides such as gp100 peptides, MAGE antigens, Trp-2, MART1, and/or tyrosinase, or tumor cells transfected to express the cytokine GM-CSF (described further below).

Some human tumors, such as melanomas, have been found to be immunogenic. Anti-tumor responses in the host can be activated by raising the threshold for T-cell activation through LAG-3 blockade.

LAG-3 blockade appears to be more effective when combined with a vaccination protocol. Many experimental cancer vaccination strategies have been developed (see Rosenberg, S., 2000, Development of Cancer Vaccines, ASCO Educational Book Spring: 60-62; Logothetis, C., 2000, ASCO Educational Book Spring: 300-302; Khayat, D. 2000, ASCO Educational Book Spring: 414-428; Foon, K. 2000, ASCO Educational Book Spring: 730-738; see also Restifo, N. and Sznol, M., Cancer Vaccines, Ch. 61, pp. 3023-3043 in DeVita et al. (eds.), 1997, Cancer: Principles and Practice of Oncology, 5th ed.). In one such strategy, a vaccine is prepared using autologous or allogeneic tumor cells. Such cell-based vaccines have been shown to be most effective when tumor cells are transduced to express GM-CSF. GM-CSF has been shown to be a potent activator of antigen presentation for tumor vaccination (Dranoff et al. (1993) Proc. Natl. Acad. Sci U.S.A. 90: 3539-43).

Studies of gene expression and broad-spectrum gene expression patterns in various tumors have identified so-called tumor-specific antigens (Rosenberg, SA (1999) Immunity 10: 281-7). In many cases, tumor-specific antigens are differentiation antigens expressed in tumors and in the cell from which the tumor arose, such as the melanocyte antigen gp100, MAGE antigens, and Trp-2. Most importantly, many of these antigens have been shown to be targets of tumor-specific T cells found in the host. LAG-3 blockade can be used in conjunction with a panel of recombinant proteins and/or peptides expressed in the tumor to elicit an immune response to these proteins. These proteins are normally regarded as self-antigens by the immune system and are therefore tolerated. Tumor antigens may also include the protein telomerase, which is required for the synthesis of chromosome telomeres and is expressed in more than 85% of human cancers but only in a limited number of somatic tissues (Kim et al. (1994) Science 266: 2011–2013). (These somatic tissues may be protected from immune attack in a variety of ways.) Tumor antigens may also represent neoantigens expressed in cancer cells due to somatic mutations that alter the protein sequence or create fusion proteins between two unrelated sequences (e.g., bcr-abl in the Philadelphia chromosome) or the idiotype from B-cell tumors.

Other tumor vaccines may include proteins from viruses involved in human cancers, such as human papillomaviruses (HPV), hepatitis viruses (HBV and HCV), and Kaposi's sarcoma virus (KHSV). Another form of tumor-specific antigen that may be used in conjunction with LAG-3 blockade is purified heat shock proteins (HSPs) isolated from tumor tissue itself. Heat shock proteins contain fragments of proteins from tumor cells, and these HSPs are highly efficient in delivering to antigen-presenting cells to induce antitumor immunity (Suot & Srivastava (1995) Science 269:1585-1588; Tamura et al. (1997) Science 278:117-120).

Dendritic cells (DCs) are potent antigen-presenting cells that can be used to prime antigen-specific responses. DCs can be produced ex vivo and loaded with a variety of protein and peptide antigens, as well as tumor cell extracts (Nes

- 23 050318 tle et al. (1998) Nature Medicine 4: 328-332). DC can also be genetically transduced to express these tumor antigens. DC have also been fused directly to tumor cells for immunization purposes (Kugler et al. (2000) Nature Medicine 6:332-336). As a vaccination strategy, DC immunization can be effectively combined with LAG-3 blockade to activate more potent antitumor responses.

LAG-3 blockade can be combined with standard cancer treatments. LAG-3 blockade can be effectively combined with chemotherapeutic regimens. In these cases, it is possible to reduce the dose of the chemotherapeutic agent administered (Mokyr et al. (1998) Cancer Research 5%; 5301-5304). An example of such a combination is an anti-LAG-3 antibody in combination with decarbazine for the treatment of melanoma. Another example of such a combination is an anti-LAG-3 antibody in combination with interleukin-2 (IL-2) for the treatment of melanoma. The scientific rationale for the combined use of LAG-3 blockade and chemotherapy is that cell death, which is a consequence of the cytotoxic action of most chemotherapeutic compounds, should lead to increased levels of tumor antigen in the antigen presentation pathway. Other combination therapies that may synergize with LAG-3 blockade through cell death include radiation, surgery, and endocrine shutoff. Each of these protocols allows for the creation of a source of tumor antigen in the host. Angiogenesis inhibitors may also be combined with LAG-3 blockade. Inhibition of angiogenesis results in the death of tumor cells that may be delivering tumor antigen to the host antigen presentation pathway.

LAG-3 blocking antibodies can also be used in combination with bispecific antibodies that target Fca or Fcy receptor-expressing effector cells on tumor cells (see, e.g., U.S. Patents 5,922,845 and 5,837,243). Bispecific antibodies can be used to target two separate antigens. For example, Fc receptor/tumor antigen (e.g., Her-2/neu) bispecific antibodies have been used to target macrophages to tumor sites. Such targeting can more efficiently activate tumor-specific responses. The T cell arm of these responses can be enhanced by LAG-3 blockade. Alternatively, antigen can be delivered directly to DCs using bispecific antibodies that bind to a tumor antigen and a dendritic cell-specific cell surface marker.

Tumors evade immune surveillance through a variety of mechanisms. Many of these mechanisms can be overcome by inactivating proteins expressed by tumors that are immunosuppressive. These include, among others, TGF-β (Kehrl et al. (1986) J. Exp. Med. 163: 1037–1050), IL-10 (Howard & O'Garra (1992) Immunology Today 13: 198–200), and Fas ligand (Hahne et al. (1996) Science 274: 1363–1365). Antibodies to each of these agents can be used in combination with an anti-LAG-3 antibody to counteract the effects of the immunosuppressive agent and promote host immune responses against the tumor.

Other antibodies that activate the host immune response can be used in combination with the anti-LAG-3 antibody. These include molecules on the surface of dendritic cells that activate DC function and antigen presentation. Anti-CD40 antibodies can effectively substitute for T helper cell activity (Ridge et al. (1998) Nature 393: 474–478) and can be used in combination with anti-LAG-3 antibodies (Ito et al. (2000) Immunobiology 201 (5) 527–40). Activation of antibodies to T cell costimulatory molecules such as CTLA-4 (e.g., U.S. Patent 5,811,097), OX-40 (Weinberg et al. (2000) Immunol 164: 2160–2169), 4-1BB (Melero et al. (1997) Nature Medicine 3: 682–685 (1997), and ICOS (Hutloff et al. (1999) Nature 397: 262–266) may also provide increased levels of T cell activation.

Currently, bone marrow transplantation is used to treat various tumors of hematopoietic origin. While graft rejection is a consequence of this treatment, beneficial therapeutic effects can be achieved through the graft-versus-tumor reaction. LAG-3 blockade can be used to enhance the efficacy of transplanted donor tumor-specific T cells.

There are also several experimental treatment protocols that involve ex vivo activation and expansion of antigen-specific T cells and adoptive transfer of these cells into recipients to stimulate antigen-specific T cells against the tumor (Greenberg & Riddell (1999) Science 285: 546-51). Such methods may also be used to activate T cell responses to infectious agents such as CMV. Ex vivo activation in the presence of anti-LAG-3 antibodies may increase the abundance and activity of adoptively transferred T cells.

Infectious diseases

Other methods of the invention are used to treat patients infected with specific toxins or pathogens. Thus, another aspect of the invention relates to a method of treating an infectious disease in a subject, comprising administering to the subject an anti-LAG-3 antibody or

- 24 050318 its antigen-binding portion such that the subject is treated for said infectious disease. Preferably, the antibody is a human anti-LAG-3 antibody (e.g., any of the human anti-LAG-3 antibodies described herein). Additionally or alternatively, the antibody may be a chimeric or humanized antibody.

Similar to its use in tumors as discussed above, antibody-mediated LAG-3 blockade can be used alone or as an adjunct in combination with vaccines to stimulate immune responses to pathogens, toxins, and self-antigens. Examples of pathogens against which this therapeutic approach may be particularly effective include pathogens that are not eradicated by currently available vaccines or pathogens against which standard vaccines are not sufficiently effective. These include, but are not limited to, HIV, hepatitis viruses (A, B, and C), influenza viruses, herpes viruses, giardiasis, malaria, leishmaniasis, Staphylococcus aureus, and Pseudomonas aeruginosa. LAG-3 blockade is particularly useful against common infections with agents such as HIV that present altered antigens during the course of infection. These new epitopes are recognized as foreign at the time of administration of anti-LAG-3 antibodies, thus provoking a strong T-cell response that is not dampened by negative signals through LAG-3.

Some examples of pathogenic viruses that cause infections that can be treated by the methods of the invention include HIV, hepatitis virus (A, B or C), herpes virus (e.g., VZV, HSV-1, HAV-6, HSV-II and CMV, Epstein-Barr virus), adenovirus, influenza virus, flaviviruses, echovirus, rhinovirus, coxsackie virus, coronavirus, respiratory syncytial virus, mumps virus, rotavirus, measles virus, rubella virus, parvovirus, vaccinia virus, HTLV virus, dengue virus, papillomavirus, molluscum contagiosum virus, poliovirus, rabies virus, JC virus and arbovirus encephalitis virus.

Some examples of pathogenic bacteria that cause infections that can be treated by the methods of the invention include chlamydia, Rickettsia, mycobacteria, staphylococci, streptococci, pneumococci, meningococci and gonococci, Klebsiella, Proteus, Serratia, Pseudomonas, Legionella, diphtheria bacteria, Salmonella, bacilli, cholera, tetanus, botulism, anthrax, plague, leptospirosis and Lyme disease bacteria.

Some examples of pathogenic fungi that cause infections that can be treated by the methods of the invention include Candida (albicans, krusei, glabrata, tropicalis, etc.), Cryptococcus neoformans, Aspergillus (fumigatus, niger, etc.), the genus Mucorales (mucor, absidia, rhizophus), Sporothrix schenkii, Blastomyces dermatitidis, Paracoccidioides brasiliensis, Coccidioides immitis, and Histoplasma capsulatum.

Some examples of pathogenic parasites that cause infections that can be treated by the methods of the invention include Entamoeba histolytica, Balantidium coli, Naegleriafowleri, Acanthamoeba sp., Giardia lambia, Cryptosporidium sp., Pneumocystis carinii, Plasmodium vivax, Babesia microti, Trypanosoma brucei, Trypanosoma cruzi, Leishmania donovani, Toxoplasma gondii, Nippostrongylus brasiliensis.

In all of the above approaches, LAG-3 blockade can be combined with other forms of immunotherapy, such as cytokine treatment (e.g., interferons, GM-CSF, G-CSF, IL-2) or bispecific antibody therapy, which provides enhanced presentation of tumor antigens (see, e.g., Holliger (1993) Proc. Natl. Acad. Sci. USA 90:6444–6448; Poljak (1994) Structure 2:1121–1123).

Autoimmune reactions

Ahtu-LAG-3 antibodies can provoke and potentiate autoimmune responses. Indeed, the induction of antitumor responses using tumor cell- and peptide-based vaccines has revealed that many antitumor responses involve reactivities against self-antigens (van Elsas et al. (2001) J. Exp. Med. 194:481-489; Overwijk, et al. (1999) Proc. Natl. Acad. Sci. U.S.A. 96: 2982-2987; Hurwitz, (2000) supra; Rosenberg & White (1996) J. Immunother Emphasis Tumor Immunol 19 (1): 81-4). Thus, the use of anti-LAG-3 blockade in combination with various autologous proteins could be considered to design vaccination protocols to effectively generate immune responses against these autologous proteins for the treatment of diseases. For example, Alzheimer's disease involves the unwanted accumulation of Αβ peptide in amyloid deposits in the brain; anti-amyloid antibody responses are able to clear these amyloid deposits (Schenk et al., (1999) Nature 400: 173-177).

Other autologous proteins, such as IgE, may also be used as targets for the treatment of allergies and asthma, and TNFa may be used to treat rheumatoid arthritis. Finally, antibody responses to various hormones can be induced by the use of anti-LAG-3 antibodies. Neutralizing antibody responses against reproductive hormones may be used for contraception. Neutralizing antibody responses against hormones and other soluble factors essential for

- 25 050318 growth of certain types of tumors, can also be considered as possible targets for vaccination.

Methods similar to the above-described methods for using an anti-LAG-3 antibody can be used to induce therapeutic autoimmune responses for the treatment of patients having undesirable accumulation of other autoantigens such as amyloid deposits, including AB in Alzheimer's disease, cytokines such as TNFa, and IgE.

Vaccines

Ahtu-LAG-3 antibodies can be used to stimulate antigen-specific immune responses by co-administering the anti-LAG-3 antibody with an antigen of interest (e.g., a vaccine). Thus, in another aspect, the present invention relates to a method of stimulating an immune response to an antigen in a subject, comprising administering to said subject: (i) an antigen; and (ii) an anti-LAG-3 antibody or an antigen-binding portion thereof, such that an immune response to the antigen in the subject is enhanced. Preferably, the antibody is a human anti-LAG-3 antibody (e.g., any of the human anti-LAG-3 antibodies described herein). Additionally or alternatively, the antibody can be a chimeric or humanized antibody. The antigen can be, for example, a tumor antigen, a viral antigen, a bacterial antigen, or an antigen from a pathogen. Non-limiting examples of such antigens include the antigens described in the sections above, such as the tumor antigens (or tumor vaccines) described above, or antigens from viruses, bacteria, or other pathogens described above.

Suitable routes of administration of the antibody compositions (e.g., human monoclonal antibodies, polyspecific and bispecific molecules, or immunoconjugates) of the invention in vivo and in vitro are well known in the art and can be selected by those skilled in the art. For example, the antibody compositions can be administered by injection (e.g., intravenous or subcutaneous). Suitable doses of the molecules used will depend on the age and body weight of the subject and the concentration and/or composition of the antibody composition.

As described above, the human anti-LAG-3 antibodies of the invention can be co-administered with one or more other therapeutic agents, such as a cytotoxic agent, a radiotoxic agent, or an immunosuppressive agent. The antibody can be associated with the agent (as an immune complex) or it can be administered separately from the agent. In the latter case (in the case of separate administration), the antibody can be administered before, after, or together with said agent, or can be co-administered with other known therapies, such as anti-cancer therapy, such as radiation. Such therapeutic agents include, inter alia, anti-tumor agents such as doxorubicin (Adriamycin), cisplatin, bleomycin sulfate, carmustine, chlorambucil, dacarbazine, and cyclophosphamide hydroxyurea, which themselves are effective only at levels that are toxic or subtoxic to the patient. Cisplatin is administered intravenously at a 100 mg/mL dose once every four weeks, and adriamycin is administered intravenously at a 60-75 mg/mL dose once every 21 days. Co-administration of the human anti-LAG-3 antibodies or antigen-binding fragments thereof of the present invention with chemotherapeutic agents provides two anticancer agents that act through different mechanisms that exert cytotoxic effects on human tumor cells. Such co-administration may overcome problems caused by the development of drug resistance or a change in the antigenicity of tumor cells, making them non-reactive to the antibody.

Also included within the scope of the present invention are kits comprising antibody compositions of the invention (e.g., human antibodies, bispecific or multispecific molecules, or immunoconjugates) and instructions for use. The kit may also contain at least one additional reagent or one or more additional human antibodies of the invention (e.g., a human antibody having complementary activity that binds to an epitope in the LAG-3 antigen that is different from the epitope of the first human antibody). Kits typically include a label indicating the intended use of the contents of the kit. The term label includes any text or printed material affixed to or included in the packaging of the kit, or included in any other form with the kit.

Combination therapy

In another aspect, the invention relates to methods of combination therapy in which an anti-LAG-3 antibody (or antigen-binding portion thereof) of the present invention is co-administered with one or more additional antibodies that are effective in stimulating immune responses to further enhance, stimulate, or up-regulate immune responses in a subject. In one embodiment, the invention relates to a method of stimulating an immune response in a subject, comprising administering to the subject an anti-LAG-3 antibody and one or more additional immunostimulatory antibodies, such as an anti-PD1 antibody, an anti-PD-L1 antibody, and/or an anti-CTLA-4 antibody, such that an immune response is stimulated in the subject, for example, to inhibit tumor growth or stimulate an antiviral response. In another embodiment, an aHTu-LAG-3 antibody and an anti-PD-1 antibody are administered to the subject. In another embodiment, an anti-LAG-3 antibody and an anti-PD-L1 antibody are administered to the subject. In another embodiment, an anti-LAG-3 antibody and an anti-CTLA-4 antibody are administered to the subject. In one embodiment, the anti-LAG-3 antibody is a human antibody, such as those described herein. Alternatively, the anti-LAG-3 antibody can be, for example, a chimeric or humanized antibody (e.g., derived from a murine anti-LAG-3 mAb). In another embodiment, the at least one additional immunostimulatory antibody (e.g., an anti-PD1, anti-PD-L1, and/or anti-CTLA-4 antibody) is a human antibody. Alternatively, the at least one additional immunostimulatory antibody can be, for example, a chimeric or humanized antibody (e.g., derived from a murine αntu-PD-1, anti-PD-L1, and/or anti-CTLA-4 antibody).

In another embodiment, the invention relates to a method of treating a hyperproliferative disease (e.g., cancer), comprising administering an anti-LAG-3 antibody and an anti-CTLA-4 antibody. In other embodiments, the anti-LAG-3 antibody is administered at a subtherapeutic dose, the anti-CTLA-4 antibody is administered at a subtherapeutic dose, or both are administered at a subtherapeutic dose. In another embodiment, the present invention relates to a method of altering a side effect associated with the treatment of a hyperproliferative disease with an immunostimulatory agent, comprising administering an anti-LAG-3 antibody and a subtherapeutic dose of an anti-CTLA-4 antibody to a subject. In some embodiments, the subject is a human. In other embodiments, the anti-CTLA-4 antibody is a human monoclonal antibody 10D1 (described in PCT Publication No. WO 01/14424) and the anti-LAG-3 antibody is a human monoclonal antibody such as LAG3.5 described herein. Other anti-CTLA-4 antibodies within the methods of the present invention include, for example, those described in: WO 98/42752; WO 00/37504; U.S. Patent 6,207,156; Hurwitz et al. (1998) Proc. Natl. Acad. Sci. USA 95(17): 10067-10071; Camacho et al. (2004) J. Clin. Oncology 22(145): Abstract No. 2505 (CP-675206 antibody); and Mokyr et al. (1998) Cancer Res. 58:5301-5304. In some embodiments, the anti-CTLA-4 antibody binds to human CTLA-4 with a KD of 5x10'8 M or less, binds to human CTLA-4 with a KD of 1x10'8 M or less, binds to human CTLA-4 with a KD of 5x10'9 M or less, or binds to human CTLA-4 with a KD of 1x10'8 M to 1x10' ¹⁰ M or less. In another embodiment, the invention provides a method of treating a hyperproliferative disease (e.g., cancer) comprising administering a LAG-3 antibody and a PD-1 antibody to a subject. In other embodiments, the anti-LAG-3 antibody is administered at a subtherapeutic dose, the anti-PD-1 antibody is administered at a subtherapeutic dose, or both are administered at a subtherapeutic dose. In another embodiment, the present invention provides a method of altering a side effect associated with treating a hyperproliferative disease with an immunostimulatory agent, comprising administering an anti-LAG-3 antibody and a subtherapeutic dose of an anti-PD-1 antibody to a subject. In some embodiments, the subject is a human. In some embodiments, the anti-PD-1 antibody is a monoclonal antibody of human sequence and the anti-LAG-3 antibody is a monoclonal antibody of human sequence, such as LAG3.5 described herein. Examples of anti-PD-1 antibodies of human sequence include 17D8, 2D3, 4H1, 5C4, and 4A11, which are described in PCT Publication WO 06/121168. Other anti-PD-1 antibodies include, for example, lambrolizumab (WO2008/156712) and AMP514 (WO2010/027423, WO2010/027827, WO2010/027828, WO2010/098788). In some embodiments, the anti-PD-1 antibody binds to human PD-1 with a KD of 5x10'8 M or less, binds to human PD-1 with a KD of 1x10'8 M or less, binds to human PD-1 with a KD of 5x10'9 M or less, or binds to human PD-1 with a KD of 1x10'8 M to 1x10' ¹⁰ M or less.

In another embodiment, the present invention provides a method of treating a hyperproliferative disease (e.g., cancer), comprising administering a LAG-3 antibody and a PD-L1 antibody to a subject. In further embodiments, the anti-PD-1 antibody is administered at a subtherapeutic dose, the anti-PD-L1 antibody is administered at a subtherapeutic dose, or both are administered at a subtherapeutic dose. In another embodiment, the present invention provides a method of altering an adverse effect associated with the treatment of a hyperproliferative disease with an immunostimulatory agent, comprising administering an anti-LAG-3 antibody and a subtherapeutic dose of an anti-PD-L1 antibody to a subject. In some embodiments, the subject is a human. In other embodiments, the anti-PD-L1 antibody is a monoclonal antibody of human sequence and the anti-LAG-3 antibody is a monoclonal antibody of human sequence, such as LAG3.5 described herein. Examples of anti-PD-L1 antibodies with a human sequence include 3G10, 12A4, 10A5, 5F8, 10H10, 1B12, 7H1, 11E6, 12B7 and 13G4, which are described in PCT Publication No. WO 07/005874. Other anti-PD-L1 antibodies include, for example, MPDL3280A (RG7446) (WO 2010/077634), MEDI4736 (WO 2011/066389) and MDX1105 (WO 2007/005874).

- 27 050318

In some embodiments, the anti-PD-L1 antibody binds to human PD-L1 with a KD of ^5x10-8 M or less, binds to human PD-L1 with a KD of ^1x10-8 M or less, binds to human PD-L1 with a KD of ^5x10-9 M or less, or binds to human PD-L1 with a KD of ^1x10-8 M to ^1x10-10 M or less.

Blockade of LAG-3 and one or more second target antigens, such as CTLA-4 and/or PD-1 and/or PD-L1, by antibodies can stimulate an immune response against cancer cells in a patient. Cancer types whose growth can be inhibited by the antibodies of the present invention include cancer types that are typically sensitive to immunotherapy. Suitable examples of cancer types for treatment using the combination therapy of the present invention include the cancer types listed above in the description of anti-LAG-3 antibody monotherapy.

In some embodiments, the combination of therapeutic antibodies described herein can be administered simultaneously as a single composition in a pharmaceutically acceptable carrier, or simultaneously as separate compositions, wherein each antibody is in a pharmaceutically acceptable carrier. In another embodiment, the combination of therapeutic antibodies can be administered sequentially. For example, the anti-CTLA-4 antibody and the anti-LAG-3 antibody can be administered sequentially, such that the anti-CTLA-4 antibody is administered first and the anti-LAG-3 antibody is administered second, or the anti-LAG-3 antibody is administered first and the anti-CTLA-4 antibody is administered second. Additionally or alternatively, the anti-PD-1 antibody and the anti-LAG-3 antibody can be administered sequentially, such that the anti-PD-1 antibody is administered first and the anti-LAG-3 antibody is administered second, or the anti-LAG-3 antibody is administered first and the anti-PD-1 antibody is administered second. Additionally or alternatively, the anti-PD-L1 antibody and the anti-LAG-3 antibody may be administered sequentially such that the anti-PD-L1 antibody is administered first and the anti-LAG-3 antibody is administered second, or the anti-LAG-3 antibody is administered first and the anti-PD-L1 antibody is administered second.

In addition, if more than one dose of the combination therapy is administered sequentially, the order of sequential administration may be reversed or the same order of administration may be maintained at each administration time point, sequential administrations may be combined with co-administrations, or any combination thereof. For example, the first administration of the combination of an anti-CTLA-4 antibody and an anti-LAG-3 antibody may be simultaneous, the second administration may be sequential with the administration of anti-CTLA-4 followed by anti-LAG-3, and the third administration may be sequential with the administration of anti-LAG-3 followed by anti-CTLA-4, etc. Additionally or alternatively, the first administration of the combination of an anti-PD-1 antibody and an anti-LAG-3 antibody may be simultaneous, the second administration may be sequential with the administration of anti-PD-1 followed by anti-LAG-3, and the third administration may be sequential with the administration of anti-LAG-3 followed by anti-PD-1, etc. Additionally or alternatively, the first administration of the combination of anti-PD-L1 antibody and anti-LAG-3 antibody may be simultaneous, the second administration may be sequential with anti-PD-L1 followed by anti-LAG-3, and the third administration may be sequential with anti-LAG-3 followed by anti-PD-L1, etc. Another exemplary dosing regimen may include a first administration that is sequential, with anti-LAG-3 administered first and anti-CTLA-4 antibody (and/or anti-PD-1 and/or anti-PD-L1) administered second, wherein subsequent administrations may be simultaneous.

Optionally, the combination of anti-LAG-3 and one or more additional antibodies (e.g., anti-CTLA-4 and/or anti-PD-1 and/or anti-PD-L1 antibodies) can be further combined with an immunogenic agent, such as cancer cells, purified tumor antigens (including recombinant proteins, peptides, and carbohydrate molecules), cells, and cells transfected with genes encoding immunostimulatory cytokines (He et al. (2004) J. Immunol. 173:4919-28). Non-limiting examples of tumor vaccines that can be used include melanoma antigen peptides, such as gp100 peptides, MAGE antigens, Trp-2, MART1, and/or tyrosinase, or tumor cells transfected to express the cytokine GM-CSF (described further below). Combined blockade of LAG-3 and CTLA-4 and/or PD-1 and/or PD-L1 may also be combined with a vaccination protocol such as any of the vaccination protocols described in detail above, taking into account anti-LAG-3 antibody monotherapy.

Combined blockade of LAG-3 and CTLA-4 and/or PD-1 and/or PD-L1 may also be combined with standard cancer therapies. For example, combined blockade of LAG-3 and CTLA-4 and/or PD1 and/or PD-L1 may be effectively combined with chemotherapeutic regimens. In these cases, it is possible to reduce the dosage of the other chemotherapeutic agent administered with the combination of the present invention (Mokyr et al. (1998) Cancer Research 58: 5301-5304). An example of such a combination is a combination of anti-LAG-3 and anti-CTLA-4 and/or anti-PD-1 and/or anti-PD-L1 antibodies also in combination with decarbazine for the treatment of melanoma. Another example is the combination of anti-LAG-3 and anti-CTLA-4 and/or anti-PD-1 and/or anti-PD-L1 antibodies also in combination with interleukin-2 (IL-2) for the treatment of melanoma. The scientific rationale for the combined use of LAG-3 and CTLA-4 and/or PD-1 and/or PD-L1 blockade with chemotherapy is that cell death, which is a consequence of the cytotoxic action of most chemotherapeutic agents, is

- 28 050318 tic compounds should result in increased levels of tumor antigen in the antigen presentation pathway. Other combination therapies that may result in synergy with combined LAG-3 and CTLA-4 and/or PD-1 and/or PD-L1 blockade through cell death include radiation, surgery, and endocrine shutdown. Each of these protocols allows for the creation of a source of tumor antigen in the host. Angiogenesis inhibitors may also be combined with combined LAG-3 and CTLA-4 and/or PD-1 and/or PD-L1 blockade. Inhibition of angiogenesis results in the death of tumor cells, which may represent a source of tumor antigen delivered to the host antigen presentation pathway.

A combination of LAG-3 and CTLA-4 and/or PD-1 and/or PD-L1 blocking antibodies may also be used in combination with bispecific antibodies that target Fca or Fcy receptor expressing effector cells on tumor cells (see, e.g., U.S. Patents 5,922,845 and 5,837,243). Bispecific antibodies may be used to target two separate antigens. The efficacy of T cells in these responses will be increased by the use of combined LAG-3 and CTLA-4 and/or PD-1 and/or PD-L1 blockade.

As another example, combinations of anti-LAG-3 and anti-CTLA-4 and/or anti-PD-1 antibodies and/or anti-PD-L1 antibodies can be used in combination with anti-cancer antibodies such as Rituxan® (rituximab), Herceptin® (trastuzumab), Bexxar® (tositumomab), Zevalin® (ibritumomab), Campath® (alemtuzumab), Lymphocide® (eprtuzumab), Avastin® (bevacizumab), and Tarceva® (erlotinib), and the like. By way of example and without being bound by theory, treatment with an anti-cancer antibody and an anti-cancer antibody conjugated to a toxin can result in the death of cancer cells (e.g., tumor cells) that can enhance an immune response mediated by CTLA-4, PD-1, PD-L1, or LAG-3. In another embodiment, the treatment of a hyperproliferative disease (e.g., a cancer tumor) may include an anti-cancer antibody in combination with anti-LAG-3 and anti-CTLA-4 and/or anti-PD-1 and/or anti-PD-L1 antibodies, simultaneously or sequentially, or in any combination thereof, which can enhance anti-tumor immune responses of the host. Tumors evade immune surveillance through a variety of mechanisms. Many of these mechanisms can be overcome by inactivating proteins that are expressed by tumors and that are immunosuppressive. These include, among others, TGF-β (Kehrl et al. (1986) J. Exp. Med. 163: 1037-1050), IL-10 (Howard & O'Garra (1992) Immunology Today 13: 198-200), and Fas ligand (Hahne et al. (1996) Science 274: 1363-1365). In another example, antibodies to each of these agents can also be combined with a combination of anti-LAG-3 and anti-CTLA-4 and/or anti-PD-1 and/or anti-PD-L1 antibodies to counteract the effects of immunosuppressive agents and promote host immune responses against the tumor.

Other antibodies that may be used to activate the host immune responsiveness may additionally be used in combination with a combination of anti-LAG-3 and anti-CTLA4 and/or anti-PD-1 and/or anti-PD-L1 antibodies. These include molecules on the surface of dendritic cells that activate DC function and antigen presentation. Ahtu-CD40 antibodies (Ridge et al., supra) may be used in combination with a combination of anti-LAG-3 and anti-CTLA-4 and/or anti-PD-1 and/or anti-PD-LI antibodies (Ito et al., supra). Other activating antibodies against T-cell costimulatory molecules (Weinberg et al., supra, Melero et al. supra, Hutloff et al., supra) may also provide increased levels of T-cell activation.

As described above, bone marrow transplantation is currently used to treat various tumors of hematopoietic origin. Combined blockade of LAG-3 and CTLA-4 and/or PD-1 and/or PD-L1 can be used to increase the efficacy of transplanted donor tumor-specific T cells.

Several experimental treatment protocols involve ex vivo activation and expansion of antigen-specific T cells and adoptive transfer of these cells into recipients to stimulate antigen-specific T cells against the tumor (Greenberg & Riddell, supra). Such methods may be used to activate T cell responses to infectious agents such as CMV. Ex vivo activation in the presence of anti-LAG-3 and anti-CTLA-4 and/or anti-PD-1 and/or anti-PD-L1 antibodies is expected to increase the frequency and activity of adoptively transferred T cells.

In some embodiments, the present invention relates to a method of altering an adverse event associated with the treatment of a hyperproliferative disease with an immunostimulatory agent, comprising administering to a subject an anti-LAG-3 antibody and a subtherapeutic dose of an anti-CTLA4 and/or anti-PD-1 and/or anti-PD-L1 antibodies. For example, the methods of the present invention relate to methods of reducing the incidence of immunostimulatory antibody-induced colitis or diarrhea by administering a non-absorbable steroid to a patient. Since any patient who would receive an immunostimulatory therapeutic antibody is at risk of developing colitis or diarrhea induced by such an antibody, the entire patient population is suitable for therapy according to the methods of the present invention. Although steroids have been administered to treat inflammatory bowel disease (TBD) and prevent flares of IBD, they have not been used for pro

- 29 050318 lactic acid (reducing the incidence) of IBD in patients who did not have a diagnosis of IBD. Significant side effects associated with steroids, even non-absorbable steroids, were considered a barrier to prophylactic use.

In other embodiments, the combination blockade of LAG-3 and CTLA-4 and/or PD-1 and/or PD-L1 (i.e., anti-LAG-3 and anti-CTLA-4 and/or anti-PD-1 antibodies and/or anti-PD-L1 antibodies) can be further combined with the use of any non-absorbable steroid. As used herein, the term non-absorbable steroid means a glucocorticoid that exhibits extensive first-pass metabolism such that, upon completion of metabolism in the liver, the bioavailability of the steroid is low, i.e., less than about 20%. In one embodiment, the non-absorbable steroid is budesonide. Budesonide is a locally acting glucocorticosteroid that is extensively metabolized, primarily by the liver, following oral administration. ENTOCORT EC® (Astra-Zeneca) is a pH- and time-dependent oral formulation of budesonide designed to optimize drug delivery to the ileum and through the colon. ENTOCORT EC® is approved in the United States for the treatment of mild to moderate Crohn's disease involving the ileum and/or ascending colon. The usual oral dose of ENTOCORT EC® for the treatment of Crohn's disease is 6-9 mg/day. ENTOCORT EC® is released in the intestine prior to absorption and is retained in the intestinal mucosa. After passing through the target tissue of the intestinal mucosa, ENTOCORT EC® is extensively metabolized by the cytochrome P450 system in the liver to metabolites with negligible glucocorticoid activity. Thus, bioavailability is low (approximately 10%). This low bioavailability of budesonide results in an improved therapeutic index compared with other glucocorticoids with less extensive first-pass metabolism. Budesonide results in fewer side effects, including less hypothalamic-pituitary suppression, than systemically acting corticosteroids. However, chronic administration of ENTOCORT EC® may result in systemic glucocorticoid effects such as hypercortisolism and adrenal suppression. See PDR 58th ed. 2004; 608–610.

In other embodiments, the combination blockade of LAG-3 and CTLA-4 and/or PD-1 and/or PD-L1 (i.e., anti-LAG-3 and anti-CTLA-4 and/or anti-PD-1 and/or anti-PD-L1 immunostimulatory therapeutic antibodies) in combination with a non-absorbable steroid can be further combined with a salicylate. Salicylates include 5-ASA agents such as, for example: sulfasalazine (AZULFIDINE®, Pharmacia &UpJohn); olsalazine (DIPENTUM®, Pharmacia &UpJohn); balsalazide (COLAZAL®, Salix Pharmaceuticals, Inc.); and mesalamine (ASACOL®, Procter & Gamble Pharmaceuticals; PENTASA®, Shire US; CANASA®, Axcan Scandipharm, Inc.; ROWASA®, Solvay). According to the methods of the present invention, a salicylate administered in combination with anti-LAG-3 and anti-CTLA-4 and/or anti-PD-1 and/or anti-PD-L1 antibodies and a non-absorbable steroid can include any overlapping or sequential administration of the salicylate and the non-absorbable steroid for the purpose of reducing the incidence of colitis induced by the immunostimulatory antibodies. Thus, for example, methods for reducing the incidence of colitis induced by the immunostimulatory antibodies of the present invention include administering the salicylate and the non-absorbable steroid simultaneously or sequentially (e.g., the salicylate is administered 6 hours after the administration of the non-absorbable steroid), or any combination thereof. In addition, according to the present invention, the salicylate and the non-absorbable steroid can be administered by the same route (e.g., both are administered orally) or by different routes (e.g., the salicylate is administered orally and the non-absorbable steroid is administered rectally), which may differ from the route(s) used to administer the anti-LAG-3 and anti-CTLA-4 and/or anti-PD-1 and/or αntu-PD-L1 antibodies.

This description is further illustrated by the following examples, which should not be construed as limiting. The contents of all figures and references, the Genbank sequence, patents and published patent applications referenced in this description are incorporated herein by reference in their entirety. In particular, the disclosures of PCT publications WO 09/045957, WO 09/073533, WO 09/073546 and WO 09/054863 are incorporated herein by reference in their entirety.

Examples

Example 1. Design of LAG3.1 variants (antibody 25F7).

Variants of the above-described anti-LAG-3 antibody, 25F7, referred to herein as LAG3.1, were generated based on initial amino acid sequence analysis of the antibody for potential degradation sites. Expression by site-directed mutagenesis of the VH region of LAG3.1 was accomplished using the QuikChange II XL® Site-Directed Mutagenesis Kit (Agilent Technologies). The altered _VH regions were then subcloned into UCOE® vectors (EMD Millipore) containing the human IgG4-S228P constant region. Each of the different heavy chain vectors was cotransfected with a vector expressing the LAG3.1 kappa chain into CHO-S cells, and stable pools were selected for expression.

- 30 050318

Five potential deamidation motifs were identified in the variable heavy chain CDR2. These sites were located at positions 52, 54, 56, 58, and 60 of the variable heavy chain region of LAG3.1 (SEQ ID NO: 2) (see Fig. 1A). In particular, deamidation of the NG sequence in CDR2 V _H (SEQ ID NO: 6) was observed under all conditions, as was additional sequence isomerization. Deamidation of the starting material was approximately 10%. In addition, this NG sequence was found to be inconsistent with the germline sequence (see Fig. 3). However, the consensus germline sequence was a potential glycosylation site and was thus not present in the antibody variants.

Four variants (herein referred to as LAG3.5, LAG3.6, LAG3.7, and LAG3.8) were obtained that corresponded to two potential deamidation motifs (positions 54 and 56) as shown in Fig. 3. These variants were subjected to a state matrix as outlined in Table 1 below, and the following characteristics were analyzed: chemical and thermal stability (physical stability); (b) size exclusion chromatography (aggregation); (c) isoelectric focusing (IEF) gel (charge heterogeneity); (d) Biacore activity (binding and functional activity); and (e) peptide mapping by mass spectrometry (chemical modifications/molecular stability).

Table 1

Buffer Acetate (100 nM NaCl, 3% w/v mannitol, 0.03% Tween-20) Citrate (100 nM NaCl, 3% w/v mannitol, 0.03% Tween-20) rn 5.5, 6.0, 6.5, 7.0 5.5, 6.0, 6.5, 7.0 Temperature 4°C and 37°C 4°C and 37°C Time 0, 4, 8, 12 weeks 0, 4, 8, 12 weeks

Example 2. Characteristics of LAG-3 variants.

1. Binding of activated human CD4+ T cells.

To test the ability of the antibody variants to bind native human LAG-3 on the surface of activated human T cells, normal healthy donor peripheral blood mononuclear cells were stimulated on 15 cm tissue culture plates at a density of 2 x 10 ⁶ cells/mL with a combination of anti-CD3 (eBioscience, catalog #16-0037-85) and anti-CD28 (BD Bioscience, catalog #555725) antibodies present in solution at 5 μg/mL and 3 μg/mL, respectively. After 3 days of stimulation, the cells were harvested, washed once with 1x PFAE buffer (1x PBS + 2% FBS, 0.02% sodium azide, 2 mM Na EDTA), and resuspended in 1x PFAE staining buffer.

For the binding reaction, LAG3.1 variants were serially diluted with cold 1x PFAE buffer, then 50 μl of the diluted antibody solution was mixed with 50 μl of Fitc-labeled anti-human CD4 antibody (BD Bioscience, cat #555346) diluted 1:16 in 1x PFAE buffer. For the binding reaction, 100 μl of this diluted antibody mixture was added to 2 × 10 ⁵ cells and the mixture was cultured at 4°C for 30 min. The cells were then washed twice with 1x PFAE buffer. A 1:200 solution of PE-labeled Fcy-specific goat anti-human antibody (Jackson ImmunoResearch, cat #109-116-170) was added and the mixture was cultured for 30 min at 4°C, followed by washing twice with cold 1x PFAE buffer. After the final wash, 150 µl of cold 1x PFAE was added to each solution and antibody binding was analyzed by flow cytometry using a FACSCanto flow cytometer (BD Bioscience).

The results of the flow cytometric analysis are summarized in Fig. 4A, which is a graph showing the _EC50 value for antibody binding to activated human CD4 ⁺ T cells. Fig. 4B is a graph showing antibody binding to soluble human LAG-3/Fc antigen using BIACORE. As shown, the binding affinities of LAG3.5 and LAG3.8 are slightly lower than those of LAG3.1, while their dissociation constants are slightly higher than those of LAG3.1.

2. Physical stability.

Thermal stability and thermal denaturation of the variants were determined using Microcal VP-DSC. Specifically, each variant was diluted with PBS (Mediatech, catalog #21-040-CV, lot #21040139). The final sample concentration was 250 μg/mL after dilution with PBS. The sample was heated to 74°C, cooled to 25°C, and reheated to 74°C. PBS buffer was used as a zero control. The data were fitted to a 2-state model and curve-fitted using Origin software.

As indicated in Table 2 and shown in Fig. 5, LAG3.5 had a higher melting point TM2 than LAG3.1, indicating higher overall stability.

- 31 050318

Table 2

MAY Tsh1 (°C) Corresponds to CH2 and/or Fab domains Tsh2 (°C) Corresponds to CH3 and/or Fab domains LAG3.1 70.7 75.7 LAG3.5 70.5 76.3 LAG3.6 67.8 70.8 LAG3.7 69.4 73.5 LAG3.8 70.3 75.4

The refolding of antibodies after denaturation is the reciprocal of the long-term aggregation potential. Therefore, LAG-3 variants were also tested and compared for thermal reversibility. Specifically, antibodies were heated to 74°C and cooled to room temperature before being heated again to 74°C. The ratio of the area under the curves of the second and first thermograms allows an assessment of thermal reversibility, which is directly proportional to conformational reversibility.

As reported in Table 3 and shown in Fig. 6, LAG3.5 had significantly higher thermal reversibility than all other variants. It should be noted that the percentage reversibility of LAG3.5 (47%) was more than twice that of LAG3.1 (20%). Thermal reversibility was strongly correlated with long-term aggregation potential. Lower reversibility corresponds to higher potential aggregation. Based on this observation, LAG3.1 would potentially have significantly higher aggregation over time compared to LAG3.5. Likewise, all other variants could potentially have higher aggregation over time compared to LAG3.5.

Table 3

MAY Thermal reversibility (%) LAG3.1 20 LAG3.5 47 LAG3.6 0 LAG3.7 AND LAG3.8 26

3. Aggregation.

The variants were also tested for stability as protein aggregation value using standard size-exclusion HPLC (SEC-HPLC) according to the following protocol: antibody samples to be tested were diluted to 1.0 mg/mL in phosphate-buffered saline (PBS) and subjected to 10 μL of HPLC (Waters, model 2795). Separation was accomplished on a gel filtration column (TOSOH Bioscience, TSKgel G3000 SWx1, 7.8 mmx300 mm, product 08541) using a mobile phase of 0.1 M sodium phosphate, 0.15 M sodium chloride, 0.1 M sodium sulfate, pH 7.2. Analytical substance was detected by monitoring UV absorbance at 280 nm, and the percent peak area of the antibody composition was determined using Empower software. As shown in Table 4, LAG3.5 had significantly reduced aggregation compared to LAG3.1.

______________________________Table 4______________________________

Sample IgG monomer (% peak area) IgG aggregate (% peak area) LAG3.1 90 10 LAG3.5 96 4 LAG3.6 96 4 LAG3.7 95 5 LAG3.8 95 5

Example 3. Selection of options.

Based on the studies described above, the antibody variant LAG3.5 was selected for further analysis in light of its significantly increased physical and chemical stability compared to its unmodified form (LAG3.1), in particular its high ability to

- 32 050318 conformational refolding (thermal reversibility). This assay involved a two-step approach with (a) enhanced loading, (b) and (c) real-time stability assessment over 12 weeks. Specifically, LAG3.5 was cultured at 1.0 mg/mL in pH 8.0, 50 mM ammonium bicarbonate for 5 days at 40°C. The extent of modifications after 5 days, as well as the effect on activity and stability, were analyzed. The LAG3.5 variant was then subjected to real-time stability analysis in PBS for 12 weeks and further analysis. The results of these studies are described below.

1. Antigen binding.

As shown in Fig. 7 (and Table 5), no change in antigen binding was observed after 5 days. As also shown in Fig. 10A and B, LAG3.5 showed no change in antigen binding or physical stability after 12 weeks. In particular, LAG3.5 maintained higher affinity than LAG3.8 over the entire 12-week period at 4°C and 40°C.

Table 5

Clone ID Antigen Kdx ΙΟ' ⁹ (Μ) konx 10 ⁴ (1/Ms) _Koyakh ΙΟ' ⁴ (1/s) Lag3.1 PBS 0.21 166 3.44 pH8 0,20 184 3.61 Lag3.5 PBS 0.25 130 3.22 pH8 0,20 148 2.98 Lag3.8 PBS 0.25 147 3.68 pH8 0.25 162 4.02

2. Chemical modifications / Molecular stability.

Peptide mapping by mass spectrometry was used to analyze the chemical/molecular stability of LAG3.5 compared to LAG3.1. Specifically, the purified body was fragmented, alkylated, dialyzed and digested with trypsin (Promega cat. V5111) and GluC (Roche cat. 11047817001). The digestion products were analyzed by nano-LC with tandem mass spectrometry (Thermo Fisher LTQ Orbitrap).

As shown in Fig. 8, LAG3.1 had increased heterogeneity in VH compared to LAG3.5 when analyzing the stability enhancement at higher pH, which results in deamidation of asparagine residues (step 1). The mass change due to isomerization could not be determined under the current experimental conditions. Percentage change is expressed as the ratio of all changes to the original peak.

Furthermore, as shown in Fig. 11, LAG3.1 had increased heterogeneity in VH compared to LAG3.5 in a long-term real-time stability assay over 12 weeks at 4°C and 40°C (Stage 2).

3. Physical stability.

Thermal reversibility was measured in PBS and at pH 8.0. Under both conditions, LAG3.5 again exhibited approximately twofold increased folding compared to LAG3.1. Specifically, as shown in Tables 6–8, LAG3.5 exhibited a refolding rate of 43% compared to 18% for LAG3.1 in PBS. LAG3.5 also exhibited a refolding rate of 48% compared to 29% for LAG3.1 at pH 8.0.

Table 6 DSC:melting

MAb Condition Tml Tml Lag3.1 PBS 70.7 75.7 Lag3.1 pH8 70.4 75.6 Lag3.5 PBS 70.8 76.4 Lag3.5 pH8 70.5 76.3

Table 7 Fluorolog-2: unfolding

MaЬ/mutants Midpoint (M) Aggregation (Μ) Lag3.1PBS 1.99 — Lag3.1 pH8 2.08 - Lag3.5 PBS 1.86 — Lag3.5pH8 2,00 -

- 33 050318

Table 8 DSC: refolding

MAY % reversibility, PBS % reversibility, pH 8 Lag3.1 18 29 Lag3.5 43 48

4. Heterogeneity of charges.

To analyze charge heterogeneity, variants were analyzed using isoelectric focusing (IEF) with standard markers with isoelectric point values of 5.5 and 10.0 protein compared to LAG3.1. Briefly, antibody solutions were loaded onto a precast 1 mm thick IEF gel with isoelectric point values of 3-7 protein (Invitrogen, catalog number EC6648BOX) together with markers with isoelectric point values of 3-10 protein (SERVA, catalog number 39212). Electrophoresis was performed using IEF 3-7 cathode buffer (Invitrogen, catalog #LC5370) and IEF anode buffer (Invitrogen, catalog #LC5370) and an electric current of 100 V constant for 1 hour, 200 V constant for 1 hour, and 500 V constant for 30 minutes. IEF gels were stained with Coomassie blue to detect protein bands and destained with methanol-acetic acid solution. IEF gels were then analyzed using ImageQuant TL software. Based on this analysis (data not shown), LAG3.5 exhibited significantly less heterogeneity than LAG3.1.

5. HGV-HPLC.

For solubility analysis, the variants were analyzed using standard hydrophobic interaction chromatography (HICH-HPLC) according to the following protocol: 50 μL of 2 M ammonium sulfate was added to 50 μL of the antibody test sample at 1 mg/mL, then 80 μL of the test sample was loaded onto an HPLC column (Waters, model 2795) coupled in series with a HIC column (TOSOH Bioscience, ether-5PW TSK gel, 7.5 mm x 75 mm, product number 07573). The sample was eluted at a flow rate of 1.0 mL/min with a gradient of 100% buffer A (2 M ammonium sulfate, 0.1 M sodium phosphate, pH 7.0) to 100% buffer B (0.1 M sodium phosphate, pH 7.0) over 50 min. The antibody was detected by monitoring UV absorbance at 280 nm and the data were analyzed using Empower software. As shown in Fig. 9, the hydrophilicity of LAG3.5 ensured solubility at high ammonium sulfate concentrations.

Example 4: Reversal of T-cell-mediated immune response inhibition.

LAG3.5 activity was determined using a functional assay using an antigen-specific murine T cell hybridoma (3A9). The 3A9 hybridoma expresses a T cell receptor specific for a peptide from hen egg lysozyme (HEL48-62) and secretes IL-2 when co-cultured with peptide-matched MHC antigen-presenting cells (LK35.2). Since huLAG-3-Fc is able to bind to murine MHC class II B cell lines, expression of huLAG-3 in the 3A9 line may exert an inhibitory effect through interaction with class II in the presenting mouse line. Comparison of the peptide profile of the 3A9 source response with that of LAG-3-transduced 3A9 cells cocultured with MHC-matched antigen-presenting cells showed that human LAG-3 expression inhibited peptide responses compared to control 3A9 cells. This inhibition was reversed by LAG-3 blockade using LAG3.5. Thus, LAG3.5 demonstrated a blockade of LAG-3-mediated inhibition.

Example 5. Activation of T cells by LAG3.5.

The functional impact of LAG3.5 on primary T cells was analyzed using SEB superantigen-stimulated human peripheral blood mononuclear cell (PBMC) cultures. Whole PBMC were isolated from 18 human donors and stimulated for 72 h in either of two assay formats:

(i) a fixed amount of antibody (20 μg/ml) and serial dilutions of SEB or (ii) a fixed amount of SEB (85 ng/ml) and serial dilutions of antibody.

Secreted IL-2 levels, as a measure of T cell activity, were monitored by ELISA. Anti-PD-I antibody and ipilimumab were used as positive controls, and LAG3.5 activity in combination with anti-PD-1 or anti-CTLA-4 was also assessed for a subset of donors.

Increased IL-2 secretion was observed across a range of SEB concentrations in fifteen of eighteen donors treated with LAG3.5 alone compared with isotype control antibody treatment.

In most cases, stimulation levels were lower than those observed with anti-PD-i or ipilimumab. When adjusted for LAG3.5, the results of the two assay formats (described above) were consistent. Moreover, in 5 of the 6 donors studied, combining LAG3.5 with anti-PD-1 or ipilimumab resulted in stimulation levels higher than those observed with

- 34 050318 observed for the isotype control antibody combined with anti-PD-1 or ipilimumab. These data demonstrated that LAG3.5 can function within normal human T cell assays and can also activate responses mediated by inhibition of PD-1 and CTLA-4 function.

______________Brief list of sequences______________

SEQ ID NO: DESCRIPTION SUBSEQUENCE 1 V _H na 25F7 (LAG3.1)

>1408_LAG-3_403_25F7.1_VHl_NT

CAGGTGCAGCTACAGCAGTGGGGCGCAGGACTGTTGAAGCCTTCGGAGACCCTGTC CCTCACCTGCGCTGTCTATGGTGGGTCCTTCAGTGATTACTACTGGAACTGGATCCG CCAGCCCCCAGGGAAGGGGCTGGAGTGGATTGGGGAAATCAATCATAATGGAAACA CCAACTCCAACCCGTCCCTCAAGAGTCGAGTCACCCTATCACTAGACACGTCCAAGA ACCAGTTCTCCCTGAAGCTGAGGTCTGTGACCGCCGCGGACACGGCTGTGTATTACT GTGCGTTTGGATATAGTGACTACGAGTACAACTGGTTCGACCCCTGGGGCCAGGGA ACCCTGGTCACCGTCTCCTCA___________________________

2 V _H aa 25F7

>1408_LAG-3_403_25F7.1_VHl_AA

QVQLQQWGAGLLKPSETLSLTCAVYGGSFSDYYWNWIRQPPGKGLEWIGEINHNGNTN SNPSLKSRVTLSLDTSKNQFSLKLRSVTAADTAVYYCAFGYSDYEYNWFDPWGQGTLV TVSS________________________.___________________________________._______________________________________.

3 Vk n.a. 25F7

>1408_LAG-3_403_25F7.1_VKl_NT

GAAATTGTGTTGACACAGTCTCCAGCCACCCTGTCTTTGTCTCCAGGGGAAAGAGCC

ACCCTCTCCTGCAGGGCCAGTCAGAGTATTAGCAGCTACTTAGCCTGGTACCAACAG AAACCTGGCCAGCTCCCAGGCTCCTCATCTATGATGCATCCAACAGGGCCACTGGC

ATCCCAGCCAGGTTCAGTGGCAGTGGGTCTGGGACAGACTTCACTCTCACCATCAGC

AGCCTAGAGCCTGAAGATTTTGCAGTTTATTACTGTCAGCAGCGTAGCAACTGGCCT

CTCACTTTTGGCCAGGGGACCAACCTGGAGATCAAA___________________.

4 Vk a.a. 25F7

>1408_LAG-3_403_25F7.1_VKl_AA

EIVLTQSPATLSLSPGERATLSCRASQSISSYLAWYQQKPGQAPRLLIYDASNRATGIPAR

FSGSGSGTDFTLTISSLEPEDFAVYYCQQRSNWPLTFGQGTNLEIK______________

5 V _H CDR1 aa 25F7 DYYWN 6 V _H CDR2 aa 25F7 EINHNGNTNSNPSLKS 7 V _H CDR3 aa 25F7 GYSDYEYNWFDP 8 Vk CDR1 a.a. 25F7 RASQSISSYLA 9 Vk CDR2 a.a. 25F7 DASNRAT 10 Vk CDR3 a.a. 25F7 QQRSNWPLT 11 V _H na LAG3.5

V _H na LAG3.5 caggtgcagctacagcagtggggcgcaggactgttgaagccttcggagaccctgtccctcacctgcgctgtctatggtgggtccttcagtg attactactggaactggatccgccagcccccaggaaggggctggagtggattggggaaatcaatcatcgtggaagcaccaactccaacc cgtccctcaagagtcgagtcaccctatcactagacacgtccaagaaccagttctccctgaagctgaggtctgtgaccgccgcggacacggc tgtgtattactgtgcgtttggatatagtgactacgagtacaactggttcgacccctggggccagggaaccctggtcaccgtctcctca_^

12 V _H aa LAG3.5

V _H aa LAG3.5

QVQLQQWGAGLLKPSETLSLTCAVYGGSFSDYYWNWIRQPPGKGLEWIGE INHRGSTNSNPSLKSRVTLSLDTSKNQFSLKLRSVTAADTAVYYCAFGYS DYEYNWFDPWGQGTLVTVSS________________________

13 Vk n.a. LAG3.5

- 35 050318

Vk n.a. LAG3.5 gaaattgtgttgacacagtctccagccaccctgtctttgtctccaggggaaagagccaccctctcctgcagggccagtcagagtattagcag ctacttagcctggtaccaacagaaacctggccaggctcccaggctcctcatctatgatgcatccaacagggccactggcatcccagccagg ttcagtggcagtgggtctgggacagacttcactctcaccatcagcagcctagagcctgaagattttgcagtttattactgtcagcagcgtagca actggcctctcacttttggccaggggaccaacctggagatcaaa

14 Vk a.a. LAG3.5

Vk a.a. LAG3.5

EIVLTQSPATLSLSPGERATLSCRASQSISSYLAWYQQKPGQAPRLLIYD

ASNRATGIPARF SGSSGSGTDFTLTIS SLEPEDF AVYYCQQRSNWPLTFGQ

GTNLEIK

15 Vh CDR1 a.a. LAG3.5 DYYWN 16 V _H CDR2 aa LAG3.5 EINHRGSTNSNPSLKS 17 Vh CDR3 a.a. LAG3.5 GYSDYEYNWFDP 18 Vk CDR1 a.a. LAG3.5 RASQSISSYLA 19 Vk CDR2 a.a. LAG3.5 DASNRAT 20 Vk CDR3 a.a. LAG3.5 QQRSNWPLT 21 LAG-3 epitope PGHPLAPG 22 LAG-3 epitope HPAAPSSW 23 LAG-3 epitope PAAPSSWG 24 Vh CDR2 a.a. LAG3.6 EIIHSGSTNSNPSLKS 25 Vh CDR2 a.a. LAG3.7 EINHGGGTNSNPSLKS 26 Vh CDR2 a.a. LAG3.8 EINHIGNTNSNPSLKS 27 Vh CDR2 a.a.HUMAN GERMLINE GEINHSGSTNY 28 (Gly4 -8ег)з 29 LAG-3 human a.a.

human LAG-3 sequence a.a.

MWEAQFLGLLFLQPLWVAPVKPLQPGAEVPVVWAQEGAPAQLPCSPTIPLQDLSLLRR AGVTWQHQPDSGPPAAAPGHPLAPGPHPAAPSSWGPRPRRYTVLSVGPGGLRSGRLPL QPRVQLDERGRQRGDFSLWLRPARRADAGEYRAAVHLRDRALSCRLRLRLGQASMTA SPPGSLRASDWVILNCSFSRPDRPASVHWFRNRGQGRVPVRESPHHHLAESFLFLPQVSP MDSGPWGCILTYRDGFNVSIMYNLTVLGLEPPTPLTVYAGAGSRVGLPCRLPAGVGTRS FLTAKWTPPGGGPDLLVTGDNGDFTLRLEDVSQAQAGTYTCHIHLQEQQLNATVTLAII TVTPKSFGSPGSLGKLLCEVTPVSGQERFVWSSLDTPSQRSFSGPWLEAQEAQLLSQPW QCQLYQGERLLGAAVYFTELSSPGAQRSGRAPGALPAGHLLLFLTLGVLSLLLLVTGAF GFHLWRRQWRPRRFSALEQGIHPPQAQSKIEELEQEPEPEPEPEPEPEPEPEPEQL*

30 Vh CDR2 a.a. LAG3.2 VIWYDGSNKYYADSVKG 31 V _H LAG3.1 na

LAG3.1HC

CAGGTGCAGCTACAGCAGTGGGGCGCCAGGACTGTTGAAGCCTTCGGAGACCCTGTC

CCTCACCTGCGCTGTCTATGGTGGGTCCTTCAGTGATTACTACTGGAACTGGATCCG

CCAGCCCCCAGGGAAGGGGCTGGAGTGGATTGGGGAAATCAATCATAATGGAAACA

CCAACTCCAACCCGTCCCTCAAGAGTCGAGTCACCCTATCACTAGACACGTCCAAGA

ACCAGTTCTCCCTGAAGCTGAGGTCTGTGACCGCCGCGGACACGGCTGTGTATTACT

- 36 050318

GTGCGTTTGGATATAGTGACTACGAGTACAACTGGTTCGACCCCTGGGGCCAGGGA ACCCTGGTCACCGTCTCCTCAGCTAGCACCAAGGGCCCATCCGTCTTCCCCCTGGCG CCCTGCTCCAGGAGCACCTCCGAGAGCACAGCCGCCCTGGGCTGCCTGGTCAAGGA CTACTTCCCCGAACCGGTGACGGTGTCGTGGAACTCAGGCGCCCTGACCAGCGGCGT GCACACCTTCCCGGCTGTCCTACAGTCCTCAGGACTCTACTCCCTCAGCAGCGTGGT GACCGTGCCCTCCAGCAGCTTGGGCACGAAGACCTACACCTGCAACGTAGATCACA AGCCCAGCAACACCAAGGTGGACAAGAGAGTTGAGTCCAATATGGTCCCCCATGC CCACCATGCCCAGCACCTGAGTTCCTGGGGGGACCATCAGTCTTCCTGTTCCCCCCA AAACCCAAGGACACTCTCATGATCTCCCGGACCCCTGAGGTCACGTGCGTGGTGGTG GACGTGAGCCAGGAAGACCCCGAGGTCCAGTTCAACTGGTACGTGGATGGCGTGGA GGTGCATAATGCCAAGACAAAGCCGCGGGAGGAGCAGTTCAACAGCACGTACCGTG TGGTCAGCGTCCTCACCGTCCTGCACCAGGACTGGCTGAACGGCAAGGAGTACAAG TGCAAGGTCTCCAACAAAGGCCTCCCGTCCTCCATCGAGAAAACCATCTCCAAAGCC AAAGGGCAGCCCCGAGAGCCACAGGTGTACACCCTGCCCCCATCCCAGGAGGAGAT GACCAAGAACCAGGTCAGCCTGACCTGCCTGGTCAAAGGCTTCTACCCCAGCGACA TCGCCGTGGAGTGGGAGAGCAATGGGCAGCCGGAGAACAACTACAAGACCACGCCT CCCGTGCTGGACTCCGACGGCTCCTTCTTCCTCTACAGCAGGCTAACCGTGGACAAG AGCAGGTGGCAGGAGGGGAATGTCTTCTCATGCTCCGTGATGCATGAGGCTCTGCA CAACCACTACACACAGAAGAGCCTCTCCCTGTCTCTGGGTAAATGA_________

32 VhLAG3.1 a.a.

TRANSLATION\OF\LAG3.1HC

QVQLQQWGAGLLKPSETLSLTCAVYGGSFSDYYWNWIRQPPGKGLEWIGEINHNGNTN SNPSLKSRVTLSLDTSKNQFSLKLRSVTAADTAVYYCAFGYSDYEYNWFDPWGQGTLV TVSSASTKGPSVFPLAPCSRSTSESTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAV LQSSGLYSLSSVVTVPSSSLGTKTYTCNVDHKPSNTKVDKRVESKYGPPCPPCPAPEFLG GPSVFLFPPKPKDTLMISRTPEVTCVVVDVSQEDPEVQFNWYVDGVEVHNAKTKPREEQ FNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKGLPSSIEKTISKAKGQPREPQVYTLPPS QEEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSRLTVD KSRWQEGNVFSCSVMHEALHNHYTQKSLSLSLGK* ____________________

33 VlLAG3.1 n.a.

LAG3.1LC

GAAATTGTGTTGACACAGTCTCCAGCCACCCTGTCTTTGTCTCCAGGGGAAAGAGCC ACCCTCTCCTGCAGGGCCAGTCAGAGTATTAGCAGCTACTTAGCCTGGTACCAACAG AAACCTGGCCAGGCTCCCAGGCTCCTCATCTATGATGCATCCAACAGGGCCACTGGC ATCCCAGCCAGGTTCAGTGGCAGTGGGTCTGGGACAGACTTCACTCTCACCATCAGC AGCCTAGAGCCTGAAGATTTTGCAGTTTATTACTGTCAGCAGCGTAGCAACTGGCCT CTCACTTTTGGCCAGGGGACCAACCTGGAGATCAAACGTACGGTGGCTGCACCATCT GTCTTCATCTTCCCGCCATCTGATGAGCAGTTGAAATCTGGAACTGCCTCTGTTGTGT GCCTGCTGAATAACTTCTATCCCAGAGAGGCCAAAGTACAGTGGAAGGTGGATAAC GCCCTCCAATCGGGTAACTCCCAGGAGAGTGTCACAGAGCAGGACAGCAAGGACAG CACCTACAGCCTCAGCAGCACCCTGACGCTGAGCAAAGCAGACTACGAGAAACACA AAGTCTACGCCTGCGAAGTCACCCATCAGGGCCTGAGCTCGCCCGTCACAAAGAGC TTCAACAGGGGAGAGTGTTAG_________________________________

34 VlLAG3.1 a.a.

-37050318

TRANSLATION\OF\LAG3.1LC

EIVLTQSPATLSLSPGERATLSCRASQSISSYLAWYQQKPGQAPRLLIYDASNRATGIPAR FSGSGSGTDFTISSLEPEDFAVYYCQQRSNWPLTFGQGTNLEIKRTVAAPSVFIFPPSDE QLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLTLS KADYEKHKVYACEVTHQGLSSPVTKSFNRGEC*____________________

35 V _H LAG3.5 aa

LAG3.5 heavy chain sequence - complete

QVQLQQWGAGLLKPSETLSLTCAVYGGSFSDYYWNWIRQPPGKGLEWIGE INHRGSTNSNPSLKSRVTLSLDTSKNQFSLKLRSVTAADTAVYYCAFGYS DYEYNWFDPWGQGTLVTVSSASTKGPSVFPLAPCSRSTSESTAALGCLVK D YFPEPVT VSWNSGALTSGVHTFPAVLQS SGLYSLS S VVTVPS S SLGTKT YTCNVDHKPSNTKVDKRVESKYGPPCPPCPAPEFLGGPSVFLFPPKPKDT LMISRTPEVTCVVVDVSQEDPEVQFNWYVDGVEVHNAKTKPREEQFNSTY RVVS VLTVLHQDWLNGKEYKCKVSNKGLP S SIEKTISKAKGQPREPQ VYT LPPSQEEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDS DGSFFLYSRLTVDKSRWQEGNVFSCSVMHEALHNHYTQKSLSLSLGK*

36 V _H LAG3.5 na

LAG3.5 heavy chain sequence - full caggtgcagctacagcagtggggcgcaggactgttgaagccttcggagaccctgtccctcacctgcgctgtctatggtgggtccttcagtg attactactggaactggatccgccagcccccagggaaggggctggagtggattggggaaatcaatcatcgtggaagcaccaactccaacc cgtccctcaagagtcgagtcaccctatcactagacacgtccaagaaccagttctccctgaagctgaggtctgtgaccgccgcggacacggc tgtgtattactgtgcgtttggatatagtgactacgagtacaactggttcgacccctggggccagggaaccctggtcaccgtctcctcagctag caccaagggcccatccgtcttccccctggcgccctgctccaggagcacctccgagagcacagccgccctgggctgcctggtcaaggact acttccccgaaccggtgacggtgtcgtggaactcaggcgccctgaccagcggcgtgcacaccttcccggctgtcctacagtcctcaggact ctactccctcagcagcgtggtgaccgtgccctccagcagcttgggcacgaagacctacacctgcaacgtagatcacaagcccagcaacac caaggtggacaagagagttgagtccaaatatggtcccccatgcccaccatgcccagcacctgagttcctggggggaccatcagtcttcctg ttccccccaaaacccaaggacactctcatgatctcccggacccctgaggtcacgtgcgtggtggtggacgtgagccaggaagaccccga ggtccagttcaactggtacgtggatggcgtggaggtgcataatgccaagacaaagccgcgggaggagcagttcaacagcacgtaccgtg tggtcagcgtcctcaccgtcctgcaccaggactggctgaacggcaaggagtacaagtgcaaggtctccaacaaaggcctcccgtcctcca tcgagaaaaccatctccaaagccaaagggcagccccgagagccacaggtgtacaccctgcccccatcccaggaggagatgaccaagaa ccaggtcagcctgacctgcctggtcaaaggcttctaccccagcgacatcgccgtggagtggggagagcaatgggcagccggagaacaact acaagaccacgcctcccgtgctggactccgacggctccttcttcctctacagcaggctaaccgtggacaagagcaggtggcaggagggg aatgtcttctcatgctccgtgatgcatgaggctctgcacaaccactacacacagaagagcctctccctgtctctgggtaaatga_______

37 V _l LAG3.5 aa

LAG3.5 kappa chain sequence - complete

EIVLTQSPATLSLSPGERATLSCRASQSISSYLAWYQQKPGQAPRLLIYD ASNRATGIPARF SGSGSGTDFTLTIS SLEPEDF AVYYCQQRSNWPLTFGQ GTNLEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKV DNALQSGNSQESVTEQDSKDSTYSLSSTLTLSKADYEKHKVYACEVTHQG LSSPVTKSFNRGEC*___________________

38 V _l LAG3.5 na

Claims (37)

LAG3.5 - kappa chain sequence - completely gaaattgtgttgacacagtctccagccaccctgtctttgtctccaggggaaagagccaccctctcctgcagggccagtcagagtattagcag ctacttagcctggtaccaacagaaacctggccaggctcccaggctcctcatctatgatgcatccaacagggccactggcatcccagccagg ttcagtggcagtgggtctgggacagacttcactctcaccatcagcagcctagagcctgaagattttgcagtttattactgtcagcagcgtagca actggcctctcacttttggccaggggaccaacctggagatcaaacgtacggtggctgcaccatctgtcttcatcttcccgccatctgatgagc agttgaaatctggaactgcctctgttgtgtgcctgctgaataacttctatcccagagaggccaaagtacagtggaaggtggataacgccctcc aatcgggtaactcccaggagagtgtcacagagcaggacagcaaggacagcacctacagcctcagcagcaccctgacgctgagcaaagc agactacgagaaacacaaagtctacgcctgcgaagtcacccatcagggcctgagctcgcccgtcacaaagagcttcaacaggggagagt gttag CLAUSE OF THE INVENTION 1. A method of treating cancer in a subject, comprising administering to the subject a monoclonal antibody or an antigen-binding portion thereof that binds the human lymphocyte activation gene 3 (LAG-3) comprising heavy chain CDR1, CDR2, and CDR3 regions comprising the amino acid sequences of SEQ ID NOs: 15, 16, and 17, respectively, and light chain CDR1, CDR2, and CDR3 regions comprising the amino acid sequences of SEQ ID NOs: 18, 19, and 20, respectively. 2. The method according to claim 1, wherein the monoclonal antibody or antigen-binding portion thereof that binds human LAG-3 comprises variable regions of the heavy and light chains comprising the amino acid sequences of SEQ ID NOs: 12 and 14, respectively. 3. The method according to claim 1 or 2, wherein the monoclonal antibody or antigen-binding portion thereof that binds human LAG-3 exhibits one or a combination of the following properties: (a) binding to monkey LAG-3; (b) no binding to mouse LAG-3; (c) inhibition of LAG-3 binding to major histocompatibility complex (MHC) class II molecules; or (d) stimulation of the immune response. 4. The method according to any one of claims 1 to 3, wherein the monoclonal antibody or antigen-binding portion thereof that binds human LAG-3 stimulates the production of interleukin-2 (IL-2) in an antigen-specific T-cell response and/or stimulates an antitumor immune response. 5. The method according to any one of claims 1 to 4, wherein the monoclonal antibody or antigen-binding portion thereof that binds human LAG-3 is a full-length antibody. 6. The method according to any one of claims 1 to 5, wherein the monoclonal antibody or antigen-binding portion thereof that binds human LAG-3 is of the IgG1, IgG2 or IgG4 isotype. 7. The method according to any one of claims 1 to 6, wherein the monoclonal antibody or antigen-binding portion thereof that binds human LAG-3 is a full-length human IgG4 antibody that binds human LAG-3 with a KD of 0.27x10'9 M or less, as determined by surface plasmon resonance. 8. The method according to any one of claims 1 to 4, wherein the monoclonal antibody or antigen-binding portion thereof that binds human LAG-3 is an antibody fragment or a single-chain antibody. 9. The method according to any one of claims 1 to 8, wherein the monoclonal antibody or antigen-binding portion thereof that binds human LAG-3 is a human antibody, a humanized antibody or a chimeric antibody. 10. The method according to any one of claims 1 to 7 or 9, wherein the monoclonal antibody or antigen-binding portion thereof that binds human LAG-3 comprises a heavy and light chain comprising the amino acid sequences of SEQ ID NOs: 35 and 37, respectively. 11. The method according to any one of claims 1 to 10, wherein the monoclonal antibody or antigen-binding portion thereof that binds human LAG-3 is suitable for parenteral administration. 12. The method according to any one of claims 1 to 11, wherein the monoclonal antibody or antigen-binding portion thereof that binds human LAG-3 is suitable for intravenous, intramuscular, subcutaneous, parenteral, spinal or epidermal administration. 13. The method according to any one of claims 1-12, further comprising administering to the subject at least one additional immunostimulatory antibody or antigen-binding portion thereof. 14. The method according to claim 13, wherein at least one additional immunostimulatory antibody or antigen-binding portion thereof is a full-length antibody. 15. The method according to claim 13 or 14, wherein at least one additional immunostimulatory antibody or antigen-binding portion thereof is of the IgG1, IgG2 or IgG4 isotype. 16. The method according to claim 13, wherein at least one additional immunostimulatory antibody or antigen-binding portion thereof is an antibody fragment or a single-chain antibody. - 39 050318 lo. 17. The method according to any one of claims 13-16, wherein at least one additional immunostimulatory antibody or antigen-binding portion thereof is a human, humanized or chimeric antibody. 18. The method according to any one of claims 13-17, wherein at least one additional immunostimulatory antibody or antigen-binding portion thereof is an anti-PD-1 antibody, an anti-PDL1 antibody and/or an anti-CTLA-4 antibody or an antigen-binding portion thereof. 19. The method according to any one of claims 13-18, wherein at least one additional immunostimulatory antibody or antigen-binding portion thereof is an anti-PD-1 antibody or antigen-binding portion thereof. 20. The method according to claim 19, wherein the anti-PD-1 antibody or antigen-binding portion thereof comprises the CDR1, CDR2, and CDR3 regions of the heavy and light chains of the 5C4 antibody. 21. The method according to claim 19 or 20, wherein the anti-PD-1 antibody or antigen-binding portion thereof comprises the variable regions of the heavy and light chains of the 5C4 antibody. 22. The method according to claim 19, wherein the anti-PD-1 antibody or antigen-binding portion thereof comprises 17D8, 2D3, 4H1, 4A11, 5C4, lambrolizumab or AMP514 or an antigen-binding portion thereof. 23. The method according to any one of claims 13-18, wherein at least one additional immunostimulatory antibody or antigen-binding portion thereof is an anti-PD-L1 antibody or antigen-binding portion thereof. 24. The method according to claim 23, wherein the anti-PD-L1 antibody or antigen-binding portion comprises 3G10, 12A4, 10A5, 5F8, 10H10, 1B12, 7H1, 11E6, 12B7, 13G4, MPDL3280A (RG7446), MEDI4736 or MDX1105 or an antigen-binding portion thereof. 25. The method according to any one of claims 13-18, wherein at least one additional immunostimulatory antibody or antigen-binding portion thereof is an anti-CTLA-4 antibody or antigen-binding portion thereof. 26. The method of claim 25, wherein the anti-CTLA-4 antibody or antigen-binding portion thereof comprises ipilimumab or an antigen-binding portion thereof. 27. The method according to any one of claims 13-26, wherein at least one additional immunostimulatory antibody or antigen-binding portion thereof is suitable for parenteral administration. 28. The method according to any one of claims 13-27, wherein at least one additional immunostimulatory antibody or antigen-binding portion thereof is suitable for intravenous, intramuscular, subcutaneous, parenteral, spinal or epidermal administration. 29. The method according to any one of claims 13-28, wherein the monoclonal antibody or antigen-binding portion thereof that binds human LAG-3 and at least one immunostimulatory antibody or antigen-binding portion thereof are contained in a bispecific molecule. 30. The method according to any one of claims 13-29, wherein the monoclonal antibody or antigen-binding portion thereof that binds human LAG-3 and at least one immunostimulatory antibody or antigen-binding portion thereof are administered simultaneously as separate compositions with each antibody in a pharmaceutically acceptable carrier. 31. The method according to any one of claims 13-29, wherein the monoclonal antibody or antigen-binding portion thereof that binds human LAG-3 and at least one immunostimulatory antibody or antigen-binding portion thereof are administered simultaneously as a single composition in a pharmaceutically acceptable carrier. 32. The method according to any one of claims 13-29, wherein the monoclonal antibody or antigen-binding portion thereof that binds human LAG-3 and at least one immunostimulatory antibody or antigen-binding portion thereof are administered sequentially. 33. The method according to any one of claims 1 to 32, further comprising administering a chemotherapeutic agent. 34. The method according to any one of claims 1 to 33, further comprising administering an angiogenesis inhibitor. 35. The method according to any one of claims 1 to 34, wherein the cancer comprises melanoma, metastatic malignant melanoma, renal cell carcinoma, hypernephroma, prostate cancer, hormone-unresponsive prostate adenocarcinoma, breast cancer, colon cancer, lung cancer, or non-small cell lung cancer. 36. The method according to any one of claims 1 to 35, wherein the cancer is a difficult-to-treat or recurrent malignant tumor or metastatic cancer. 37. The method according to any one of claims 1 to 36, wherein the subject is a human. -