WO2012018260A1

WO2012018260A1 - Epidermal growth factor receptor targeted immune therapy

Info

Publication number: WO2012018260A1
Application number: PCT/NL2011/050553
Authority: WO
Inventors: Dietmar Martin Werner Zaiss; Elisabeth Joanna Anna Maria Sijts
Original assignee: Universiteit Utrecht Holding B.V.
Priority date: 2010-08-05
Filing date: 2011-08-05
Publication date: 2012-02-09

Abstract

The disclosure relates among other to a method for the treatment of an individual suffering from a chronic infection by a pathogen and/or a tumor, said method comprising providing the individual in need thereof with antigen of said pathogen or tumor, said method further comprising administering an EGF-R inhibitor and/or a high affinity EGF-R signalling ligand to said individual. The disclosure also relates to a method for the treatment of an individual suffering from a disease associated with an over-active immune response, said method comprising administering to the individual in need thereof a low affinity EGF-R signalling ligand.

Description

Title: Epidermal growth factor receptor targeted immune therapy.

The invention relates to the field of immunology. In particular the invention relates to the modulation of the immune suppressive effects of regulatory T cells in an individual. The means and methods of the invention are among others useful for the treatment individuals suffering from a tumour, a chronic infection and/or an autoimmune disease.

The clinical introduction of monoclonal antibodies targeting tumour antigens has shown remarkable successes in the treatment of cancer. A number of aspects, such as the kinetics of the clinical responses (delayed on-set and delayed recurrence of the tumour after termination of treatment) or the proven induction of anti-tumour T cell responses in treated patients, have suggested that part of their clinical successes may be mediated by and be dependent on antibody treatment-induced T cell responses. One of the first monoclonal antibodies approved for clinical use were targeting the epidermal growth factor receptor (EGF-R). Their clinical success is assumed to be mediated by interference with EGF-R mediated growth signals on "growth factor addicted" tumours. However, there are also reports that the immune system is of relevance. The immune system can support the effect of antibody- mediated actions in different ways, such as through natural killer cell (NK- cell) mediated cell lysis (ADCC) or indirectly via induced antitumour T cell responses. In the present invention it was found that activated regulatory T cells in mice and men express the EGF-R and that their immune suppressive function is supported by specific EGF-R mediated signals. This novel EGF-R mediated immune regulatory effect represents an unexpected, pivotal mechanism that is affected in patients treated with drugs interfering with the EGF-R signalling pathway. The present invention provides evidence that targeting of the EGF-R signalling pathway can significantly enhance the immune mediated clearance of undesired cells and pathogens from individuals.

Tumour antigen targeting monoclonal antibodies, such as RetuxiMab (anti-CD20), TrastuzuMab (anti-Her2/neu) or CetuxiMab (anti-EGF-R), had striking successes upon introduction in the clinic, with significant clinical response rates and substantial survival advantages for patients suffering from specific cancers (21). While superficially, these successes are easy to explain, a number of observations suggested that the effects of these antibodies in fact were multi-layered. In recent years, it has become increasingly appreciated that observed clinical successes for a substantial part might be based on the action of antibody treatment-induced anti-tumour T cell responses (for review see Ferris RL et al. (20)).

Very fundamental aspects already question the idea that tumour targeting antibodies, like these targeting the EGF-R, function predominantly by direct effects on tumour survival (i.e. by induction of tumour cell death by blocking growth signals or by antibody induced NK-cell mediated cell lysis [ADCC]). One example is the fact that the targeted molecules usually are not only expressed on malignant cells but also on large number of normal cells. The adverse effects that should be experienced upon treatment, however, occur in a limited number of treated patients only. Also, antibody treatment is efficient only against a subset of tumour malignancies that (over-) express the targeted molecule. So is it well established that the EGF-R is over-expressed in a number of tumours, such as non- small- cell lung cancer (NSCLC), ovarian cancer, head and neck cancer, prostate cancer and colorectal cancer (CRC).

Nevertheless, mainly trial-and error based clinical trials have established that CetuxiMab treatment of NSCLC, head and neck cancer and CRC patients are successful in a relatively large number of patients, while treatment of, for example, ovarian and prostate cancer are effective in considerably fewer patients. We have now found that regulatory T (Treg) cells are direct targets of EGF-R targeted anti-tumour therapies. In a healthy individual approximately 5% of the Treg in the blood express detectable amount of EGF-R. This percentage is substantially increased in peritumoral Tregs. We have found that the RGF-R is active in regulating the immune-suppressive function of these EGF-R positive Treg cells. Moreover, we show that inhibition of signalling through the EGF-R, of EGF-R positive Tregs, stimulated the anti- tumour immune response in tumour carrying individuals, also when the tumour itself did not express detectable amounts of EGF-R.

Reducing the immune suppression by regulatory T cells as provided in the invention allows the immune system of the host to act more strongly to specific antigens present on the tumor or produced by the pathogen or pathogen infected cells in a chronic infection. The availability of specific antigens for initiating and/or boosting an existing immune response against the tumor or chronic infection may be enhanced in a method or use of the invention. The enhanced availability of such antigens, in combination with a method or use of the invention for reducing the immune suppressive effect of regulatory T cells of the invention, allows for a more effective immune response against the tumor and/or chronic infection.

The invention now provides a method for reducing the immune suppressive effect of EGF-R positive regulatory T cells in an individual suffering from a chronic infection and/or a tumour comprising administering to the individual in need thereof an effective amount of an EGF-R signalling inhibitor and/or a high affinity EGF-R ligand.

The individual is preferably a mammal, more preferably a primate or a rodent. In a particularly preferred embodiment said individual is a human. The invention further provides a method for the treatment of an individual suffering from a chronic infection by a pathogen and/or a tumor, said method comprising administering an EGF-R inhibitor and/or a high affinity EGF-R signalling ligand to said individual, said method further comprising providing the individual in need thereof with antigen of said pathogen or tumor, immune cells comprising and/or specific for said antigen, an antibody specific for said antigen or a combination thereof.

The invention further provides a method for the treatment of an individual suffering from a chronic infection and/or an epidermal growth factor receptor (EGF-R) negative tumor, said method comprising administering to the individual in need thereof an EGF-R inhibitor and/or a high affinity EGF-R ligand.

Said EGF-R inhibitor and/or high affinity EGF-R ligand preferably interacts with an EGF-R on regulatory T cells, thereby reducing the immune suppression by said regulatory T cells.

The invention further provides an EGF-R inhibitor and/or a high affinity ligand for use in reducing the immune suppressive effect of regulatory T cells, preferably in the treatment of an individual in need thereof. Prefereably said individual is suffering from a tumour or a chronic infection. Preferably said immune suppression is suppression of proliferation of effector cells, preferably CD4 positive effector cells.

Also provided is an EGF-R inhibitor or a high affinity EGF-R signalling ligand for use in the treatment of an individual suffering from a chronic infection and/or a tumor, wherein said treatment further comprises providing the individual in need thereof with antigen of said pathogen or tumor, immune cells comprising and/or specific for said antigen, an antibody specific for said antigen or a combination thereof. Also provided is an EGF-R inhibitor and/or a high affinity ligand for use in the treatment of individual suffering from a tumour and/or a chronic infection for obtaining a reduction in the immune suppressive effect of regulatory T cells in said individual. Treated regulatory T cells are preferably located in the vicinity of said tumor and/or chronic infection nidus. In the case of tumors such regulatory T cells are also referred to as peritumoral regulatory T cells. A regulatory T cell is said to be located in the vicinity of a tumor or chronic infection when it is present within a radius of 5 cm of the tumor or site of chronic infection (infection nidus). Preferably said regulatory T cell is located within a radius of 3 cm of the tumor or site of chronic infection, more preferably within a radius of 1 cm. The regulatory T cell can also be located within the tumor or the site of chronic infection.

Regulatory T cells located in the vicinity of the tumor or site of chronic infection can migrate to other parts of the body. This typically occurs via the draining lymph nodes. Once arrived back into the blood circulation they can migrate to the bone marrow or spleen and/or migrate back to the tumor or site of chronic infection. The interaction with immune cells to suppress the immune response can occur both in the vicinity of the tumor or site of chronic infection and in other parts of the body. A method of the invention can thus act throughout the body.

EGF-R is a member of the ErbB family of receptors. The family presently has four closely related receptor tyrosine kinases: EGF-R (ErbB-1), HER2/c-neu (ErbB- 2), Her 3 (ErbB-3) and Her 4 (ErbB-4). The human EGF-R has Gene ID: 1956 in genbank whereas the murine homologue has Gene ID: 13649. EGF-R has a number of specific ligands, including epidermal growth factor (EGF) and transforming growth factor alpha. A cell is said to be positive for EGF-R when EGF-R can be detected on the surface of the cell by an antibody that is specific for EGF-R.

Stimulation of the intrinsic tyrosine kinase activity of the epidermal growth factor receptor (EGF-R) induces a complex cascade of phosphorylation and activation events that determine a number of cell-fate decisions, such as proliferation or differentiation. Amphiregulin (AREG) is an EGF-like growth factor that binds the EGF-R. Although originally described as a growth factor that is produced by epithelial cells, more recent studies have shown that also different leuko- and lymphocytes, including activated Th2 cells, mast cells, eosinophiles and basophiles, produce AREG. In a previous study using AREG- deficient mice, we demonstrated that AREG, produced by Th2 cells, is involved in the efficient clearance of the parasite Trichuris muris. Remarkably, while establishing bone marrow (BM) chimeric mice, we found that wild-type (wt) recipient C57BL/6 mice that received bone marrow derived from AREG gene- deficient mice developed dermatitis (Fig 5). In the present invention we have identified an AREG responsive cell type that is important for transmitting the AREG function. We show that Forkhead box P3 (FoxP3) positive regulatory T cells (Treg cell) dampen CD4 T cell responses in infection models and mediate immune tolerization after BM transplantation. We also show that Treg cell function is impaired in the absence of AREG. Figure 2 and 3 depict

suppression of Treg and Treg function in vitro and in vivo respectively.

Immune- or immuno- suppression involves an act that reduces the activation or efficacy of the immune system in an individual. The immune suppressive effect of said regulatory T cells that is modulated is preferably directed towards said infection and/or tumour. Various tests are available to measure immune suppression by regulatory T cells. Results of such tests in vitro and in vivo are given in figures 2 and 3.

Regulatory T cells of the present invention are preferably Forkhead box P3 (FoxP3) positive T cells. Preferably said regulatory T cells are also CD4 positive T cells. Thus in a preferred embodiment said regulatory T cell is a CD4 positive and FoxP3 expressing regulatory T cell.

EGF-R was one of the first tumour antigens to be successfully targeted in human cancer. Since this first use many different EGF-R signalling inhibitors have been developed. Among these are EGF-R specific antibodies and the EGF-R tyrosine kinase inhibitors that specifically inhibit the EGF-R signalling by inhibiting the tyrosine kinase activity of an active EGF-R.

Preferred but not limiting examples of EGF-R specific antibodies that inhibit EGF-R signalling and EGF-R specific tyrosine kinase inhibitors are described in Sridhar et al which is incorporated by reference herein (Sridhar et al, 2003: Lancet Vol. 4 pp: 397-406). In a preferred embodiment said EGF-R signalling inhibitor is an EGF-R specific antibody such as CetuxiMab, ABX-EGF,

EMD72000, MAb ICR62, hR3 and EGF-R binding derivates thereof (ref Sridhar lancet 2003). The different EGF-R antibodies can recognize different parts of the EGF-R. Such EGF-R specific antibodies have slightly different effects, however, for the present invention it is the EGF-R signalling blocking property that decreases the immune suppressive effect of Tregs.

A derivative of an EGF-R specific antibody as described herein above, is an antibody that has the same EGF-R specific binding properties as the reference antibody but differs there from in one or more amino acids, such as the constant region defining the isotype of the antibody. A derivative of an antibody as indicated herein above, thus retains the EGF-R specific binding part of the reference antibody. The antigen specific binding properties of antibodies have been determined to reside in the so-called complementarity determining regions (CDRs) of the antibody. The framework region and the constant region typically tolerate at least conservative amino acid substitution. Such substitutions are often performed to remove unwanted B- or T cell determinants from the reference antibody, a process that is sometimes also referred to as deimmunizing or veneering. The resultant antibody is often referred to as a "humanized" antibody, when the deimmunization/verneering has been carried out with respect to the human immune system.

As mentioned herein above, another preferred EGF-R signalling inhibitor is a tyrosine kinase inhibitor. None limiting preferred examples are gefitinib or erlotinib. (ref Shridhar lancet 2003). Gefitinib and erlotinib have the following respective chemical formulas "N-(3-chloro-4-fluoro-phenyl)-7- methoxy-6-(3-morpholin-4-ylpropoxy)quinazolin-4- amine" and "N-(3- ethynylphenyl)-6,7-bis(2-methoxyethoxy) quinazolin-4-amine". Indeed many other EGF-R specific tyrosine kinase inhibitors have been described see for instance US11852474; US 11912792; US12706675; US11636549. EGF-R specific antibodies and EGF-R specific tyrosine kinase inhibitors are presently the most used EGF-R signalling inhibitors in the clinic. Since the discovery of these inhibitors various other inhibitors have been developed. These included, fragments of antibodies that retain the EGF-R specific binding property of the original antibody, EGF-R specific single chain Fv-fragments, EGF-R specific monobodies, EGF-R specific nanobodies; EGF-R specific VHH, EGF-R specific Fab-fragments and EGF-R specific artificial binding proteins such as for example avimers, and the like. A preferred EGF-R specific nanobody is given in the examples.

An EGF-R inhibitor or high affinity EGF ligand can be used in the treatment of a variety of different tumors and/or chronic infections. Preferred tumors are squamous cell cancer, lung cancer including small-cell lung cancer, non-small cell lung cancer, adenocarcinoma of the lung, and squamous carcinoma of the lung, cancer of the peritoneum, hepatocellular cancer, gastric or stomach cancer including gastrointestinal cancer, pancreatic cancer, glioblastoma, cervical cancer, sarcoma, ovarian cancer, bladder cancer, hepatoma, breast cancer, colon cancer, colorectal cancer, endometrial or uterine carcinoma, salivary gland carcinoma, kidney or renal cancer, prostate cancer, vulval cancer, thyroid cancer, hepatic carcinoma and various types of head and neck cancer, lymphoma including T-cell lymphoma as well as B-cell lymphoma including Hogkin's lymphoma, low grade/follicular non- Hodgkin's lymphoma (NHL); small lymphocytic (SL) NHL; intermediate grade/ follicular NHL; intermediate grade diffuse NHL; high grade immunoblastic NHL; high grade lymphoblastic NHL; high grade small non- cleaved cell NHL; bulky disease NHL; mantle cell lymphoma; AIDS-related lymphoma; and

Waldenstrom's Macroglobulinemia; chronic lymphocytic leukemia (CLL); acute lymphoblastic leukemia (ALL); Hairy cell leukemia; chronic myeloblasts leukemia; and post-transplant lymphoproliferative disorder (PTLD), brain tumors, Meigs' syndrome, melanoma, mesothelioma, multiple myeloma, fibrosarcoma, osteosarcoma, and epidermoid carcinoma. Particularly preferred tumors are melanoma, prostate cancer, pancreatic cancer and lymphoma, preferably a lymphoma as listed herein above.

A treatment of the invention is in principle of use for any chronic infection wherein the infection persists in the individual due to the failure of the immune system to effectively mount an immune response against the pathogen. Preferred examples of suitable chronic infection are infections of viral origin such as preferably by human cytomegalovirus (HCMV), hepatitis B, C or D and Epstein Barr Virus (EBV). Another preferred example is an infection of bacterial origin. Preferred examples of a bacterial infection are a Chlamydia infection (preferably by Chlamydia trachomatis), and an infection with a mycobacterium, preferably by mycobacterium leprae or mycobacterium tuberculosis. Another preferred chronic infection is a parasitic infection, preferably an infection by Leishmania (Belkaid Y et al. Nature. 2002;

420(6915):502-7). Another preferred chronic infection is an infection by a fungus, preferably an Aspergillus infection.

As the invention has found that EGF-R positive regulatory T cells are instrumental in suppressing a host immune response against a tumor, a method of treatment or a medical use of the invention is preferably a treatment or use of a tumor that does not over-express EGF-R. A tumor is said to over-express EGF-R if the tumor or a metastasis thereof expresses on average more EGF-R per cell, or expresses EGF-R on average in a higher number of cells when compared to the tissue of origin of the tumor. Preferably the tumor or a metastasis thereof expresses on average at least 20%, more preferably at least 50% and particularly preferred at least 100% more EGF-R per cell, or expresses EGF-R on average in at least 20%, more preferably at least 50% and particularly preferred at least 100% more cells when compared to the tissue of origin of the tumor. In a particularly preferred embodiment the tumor that does not over-express EGF-R is an EGF-R negative tumor. A preferred method for determining whether EGF-R is expressed by a cell or a tumor is by means of immuno-histochemistry. Immuno-histochemistry utilizes directly or indirectly labelled antibodies to visualize biomolecule targets in biological samples. The technique uses the specificity of antibodies to their antigen to target labels to specific biomolecule targets in or on a cell or non- cellular structure in a biological sample. It allows visualisation of the distribution of the target molecule through the sample. Immuno- histochemistry can be used on tissue sections, cultured cell lines, or individual cells, and may be used to analyse the distribution of proteins, glycans, and small biological and non-biological molecules. Immuno-histochemistry can be used in combination with other, non-antibody methods of staining, preferably fluorescent staining, for example, use of DAPI to label DNA.

The antibody used for the immuno-histochemical studies can be conjugated to an enzyme that can catalyze a color-producing reaction, such as a peroxidise, or it can be tagged to a fluorophore. These compounds may be directly attached to the antibody with the given target specificity (i.e. the primary antibody) or be present on a further molecule with specificity for the primary antibody or associated label therewith. The further molecule can be an antibody that can bind to the primary antibody or it can be a molecule such as streptavidin that can bind to biotin, where the primary antibody is labelled with biotin.

A preferred method for determining the EGF-R status of a cell or tumor is given in the examples. Preferred methods are also described in Chung et al., which is hereby incorporated by reference (J Clin Oncol. 2005 Mar

20;23(9):1803-10). Chung et al. refers to various assays known in the art for determining EGF-R tumor status and describes a standard streptavidin-biotin- peroxidase immunostaining procedure.

Some methods, for instance, RT-PCR can be more sensitive than

Immuno-histochemistry. However, these are not preferred for the present invention. One reason is that for instance RT-PCR measure expression in a sample. A tumor sample is often not homogeneous so when expression is detected it is then not possible to determine with certainty from which cell this signal originated. So the preferred method for determining whether a cell or a tumor is EGF-R positive is by means of Immuno-histochemistry, preferably in a histological sample.

The individual is, as described, preferably further provided with an antigen of the pathogen or the tumor. The antigen is typically a protein or a polysaccharide. Lipids and nucleic acids are typically antigenic only when combined with a protein. The antigen can be a complete protein but also a part of the protein or polysaccharide that is an antigenic determinant recognized by the immune system. Such an antigenic determined is also referred to as an epitope. A protein antigenic determinant can be a linear epitope or a three dimensional epitope. A linear protein antigenic determinant comprises at least 5, preferably at least 7 and more preferably at least 10 consecutive amino acids of the complete protein. The enhanced availability of such antigens, in combination with a method or use of the invention for reducing the immune suppressive effect of regulatory T cells, allows for a more effective immune response against the tumor and/or chronic infection. Thus in a preferred embodiment a treatment is provided, wherein said treatment further comprises administering to said individual a further treatment for said chronic infection and/or tumor. Said further treatment preferably comprises a treatment that increases the availability of specific antigens for directing the immune response of the host to the tumor and/or chronic infection. The further treatment works best for tumors that previously have been found to be refractory to treatment with an EGF-R inhibitor such as an EGF-R specific antibody. In such cases a method of the invention will result in a higher number of individuals where the tumor responds to treatment when compared with treatment with the EGF-R inhibitor alone or the further treatment alone. In a preferred embodiment a method of the invention if performed on tumor of a type that is found to be refractory to treatment with an EGF-inhibitor.

Tumor types that are refractory to treatment with EGF-R are tumor types where, when viewed over at least 30 patients treated in a given trial, on average, less than 20 percent and preferably less than 10 percent of the individuals exhibit a statistically significant objective response upon treatment with the EGF-R inhibitor, preferably an EGF-R specific antibody alone. A patient is said to respond to EGF-R treatment when individuals exhibit a statistically significant objective response upon treatment with the EGF-R inhibitor and the primary tumor mass has shrunk by at least 50% over the treatment period, and/or when the metastasis tumor mass in the individual has shrunk by at least 50% over the treatment period.

An antigen is herein defined as a molecule that, when introduced into the body of a healthy individual, triggers an immune response specific for the antigen in the individual. The antigen is typically a protein or a

polysaccharide. Lipids and nucleic acids are typically antigenic only when combined with a protein. The antigen can be a complete protein but also a part of the protein or polysaccharide that is an antigenic determinant recognized by the immune system. Such an antigenic determined is also referred to as an epitope. A protein antigenic determinant can be a linear epitope or a three dimensional epitope. A linear protein antigenic determinant comprises at least 5, preferably at least 7 and more preferably at least 10 consecutive amino acids of the complete protein. In a preferred embodiment said antigen is a tumor- specific antigen or a pathogen specific antigen.

The antigen can be provided to the individual in various ways. A preferred method is by means of a vaccine comprising said antigen, or a vaccine comprising a nucleic acid encoding said antigen. The vaccine may further comprise an adjuvant, excipient and/or other substances normally present in vaccines. Vaccines comprising nucleic acid encoding the antigen are also of interest. The nucleic acid typically contains an expression cassette for expression of the antigen in cells of the recipient of the vaccine. An expression cassette typically contains a promoter and other signals for producing an RNA copy encoding the antigen. When the nucleic acid is a translatable RNA then the expression cassette need only to have suitable translational start and stop signals. Non-limiting examples of nucleic acid vaccines are: DNA-vaccines comprising such an expression cassette and replicative nucleic acids

comprising such an expression cassette. Replicative nucleic acids typically utilize viral sequences for replicating nucleic acid contains the expression cassette. The replicative nucleic acid can be a viral vector, such as an adenoviral vector; vaccinia virus vector; a small pox virus vector, a retrovirus such as a lentivirus, adeno-associated virus, a vesicular stomatitis virus (VSV), a fowl-pox or a semliki forest virus vector. Viral vectors can be administered as virus particles, others, for instance semliki forest virus can be administered as nucleic acid. The replicative nucleic acid can also be a bacterial vector, such as a Listeria vector or a Salmonella vector.

In case of a pathogen, the antigen can also be provided to the individual by administering a medicament that is toxic for said pathogen to said individual. As a result a number of pathogens or cells of the pathogen will die. Antigen of the pathogen will be released and be available for triggering an immune response in the individual.

The individual may also be provided with immune cells containin; and/or specific for said antigen. Examples of suitable immune cells are:

antigen presenting cells comprising the antigen, preferably said antigen presenting cells comprises dendritic cells; cytotoxic T cells specific for said antigen; T-cells provided with a T-cell receptor that is specific for said antigen; and natural killer cells that are either ex vivo or in vitro directed toward specifically interacting with antigen of the pathogen and/or tumor. The individual may also be provided with a polyvalent peptide vaccine, or a vaccination with natural or artificial antigen-presenting cells, such as exosomes or micro-particles.

The individual may also be provided with an antibody specific for said antigen. When the individual is provided with an EGF-R specific antibody as EGF-R inhibitor it is preferred that the antibody mentioned in the previous sentence is a further antibody. Such further antibody is not an antibody specific for EGR-R. In a preferred embodiment said individual is provided with an antibody specific for a tumor antigen, preferably a tumor associated and/or tumor specific antigen. A tumor- specific antigen is only present on the tumor in the adult or child. A tumor -associated antigen is also present on some "normal" cells or cell types in the adult or child suffering from the tumor. In a preferred embodiment said further antibody is specific for CD 19, CD20, CD22, SLAM-F7, CD30, CD33, CD40, CD52, CD90, Her2/neu (Erb2B), Mucin, transmembrane glycoprotein NMB or vimentin. Non-limiting examples of such antibodies are: targeting CD20: RituxiMab, Tositumomab, IbritumoMab, OcrelizuMab, OfatumuMab, Veltuzumab; targeting CD19: BlinatumoMab or Taplitumomab; targeting CD22: Epratuzumab, targeting SLAM-F7 expressed myeloma: ElotuzuMab; targeting CD30: BrentuxiMab; targeting CD40:

Dacetuzumab or Lucatumumab; targeting CD52: AlemtuzuMab; targeting CD33: GemtuzuMab or Lintuzumab; targeting CD90: Galiximab; targeting Her2/neu (ErbB2): TrastuzuMab or ErtumaxoMab; targeting tumor-expressed mucin: ClivatuzuMab; targeting Transmembrane glycoprotein NMB on breast cancer cells or melanoma cells: GlembatumuMab; and targeting vimentin on glioma cells: PritumuMab. Like the other members of family of the ErbB family, the EGF-R is the cell surface receptor for a considerable number of members of the epidermal growth factor family. Ligands that bind to the EGF-R include but are not limited to EGF, TGF-alpha, HB-EGF, amphiregulin (AREG), betacellulin, epigen and epiregulin. Other ligands are for instance, virus proteins that bind the EGF-R. Tzahar et al describe a variety of virus encoded proteins that can bind and control signalling of EGF-R (Tzahar et al 1998, EMBO Vol 17, pp 5948-4963). These authors also show that the viral EGF-ligands induce potent EGF-R mediated signalling in spite of the fact that these ligands have 10- to 1000-fold lower binding affinity to the respective receptors. Thus different EGF-R ligands have different affinities for the receptor. Examples of high affinity ligands are EGF and TGF-alpha. An example of a low affinity ligand is AREG. In the present invention it was found that particularly the low affinity ligands stimulate the immune suppression by regulatory T cells. Vice versa, it has been found that high affinity ligands, although acting through the same receptor decrease the immune suppressive effect of the regulatory T cell thus treated. Without being bound by theory it is believed that a high affinity ligands and low affinity ligands induce different signals downstream of the EGF-R signalling cascade.

To determine whether an EGF-R ligand is a high affinity or a low affinity ligand reference is made to the method for determining binding specificities and affinities of EGF domains of ErbB receptors (Jones et al 1999, FEBS letters Vol 447, pp. 227-231). An EGF-R ligand is said to be a high affinity EGF-R ligand when the IC50 value determined using the method and cell line of Jones et al is 100 nM or less. An EGF-R ligand is said to be a low affinity EGF-R ligand when the IC50 value determined using the method of Jones et al is higher than 100 nM. In a preferred embodiment said low affinity ligand has an IC50 value of at least 1000 nM. The factor AREG is a low affinity ligand. The present invention shows that the suppression of an immune response exhibited by regulatory T cells can be modulated by modulating the activity of the EGF-R on these cells. Stimulation of the activity with a low affinity EGF-R ligand stimulates the immune suppressive effect exhibited by these cells. Vice versa, inhibition of stimulation of EGF-R decreases the immune suppressive effect exhibited by these cells. The invention thus provides a method for modulating the function and/or immune suppressive effect of a regulatory T cell, said method comprising providing an EGF-R positive regulatory T cell with an EGF-R inhibitor or an EGF-R ligand thereby modulating the function of said regulatory T cell. This finding can be used, for instance, to enhance an immune response in an individual, for instance in vaccinations where an effective immune response against the administered vaccine is desired. In this context, the provision of an EGF-R inhibitor can be seen as an adjuvant.

The capability to modulate the suppressive function of regulatory T cells can also be used to treat diseases that are the result of an over-active immune response. Thus the invention further provides a method for the treatment of an individual suffering from a disease associated with an over-active immune response, said method comprising administering to the individual in need thereof a low affinity EGF-R ligand. Over-active immune responses are typically seen in auto-immune diseases where the individuals own immune system is inadvertently active against one or more antigens that are normally present in an individual and should be recognized as "self- antigens". Preferred examples such autoimmune diseases are rheumatoid arthritis, multiple sclerosis, psoriasis or type I diabetis. Other preferred examples of diseases that are associated with an over-active immune response are allergies and contact hypersensitivities. Other diseases that are associated with an over-active immune response are graft versus host disease and host versus graft disease. These diseases are a serious side-effect of allogeneic transplantation. In these cases the immune system is often directed towards "non- self antigens". In case of graft versus host disease, cells of the graft act against "self- antigens" of the host. In relation to the graft the host cells are "non-self. In case of host versus graft disease, the host immune system acts against foreign antigens in the graft. Although the immune cells thus act more or less according to their inteneded function, the activity is not desired. Thus immune system is said to be over-active in the context of the present invention as the activity towards the graft and/or the host is undesired in individuals that have undergone the transplantation.

The EGF-like growth factor Amphiregulin (AREG) is a type II cytokine, expressed both by activated Th2- and mast cells. While analyzing the role of AREG in the immune system, we observed a dysregulated immune response in AREG gene- deficient mice. So develop, for example, bone marrow chimeric wt mice, transplanted with AREG gene- deficient bone marrow, a severe form of skin rejection six weeks after reconstitution; a finding that strongly suggests that the local functioning of regulatory T cells in the skin might be disrupted - a similar phenotype is, for example, observed when entry of FoxP3pos CD4 T cells into the skin is disrupted. Therefore, we analyzed the FoxP3pos CD4 T cell compartment in AREG gene- deficient and wt mice in more detail, however, detected similar frequencies of FoxP3pos CD4 T cells in lymph nodes and spleens of AREG gene- deficient and wt mice. Furthermore, however, we detected a small population of FoxP3pos CD4 T cells (about 5% of FoxP3pos CD4 T cells) that expressed the EGF-R. These EGF-Rpos FoxP3pos CD4 T cells were present both in wt and AREG-deficient mice. Remarkably, when analyzing the EGF-R expression of FoxP3pos CD4 T cells under inflammatory conditions, we noticed that the FoxP3pos CD4 T cell population that expressed the EGF-R increased to up to 50% in wt mice but still formed only about 5% of the FoxP3pos CD4 T cells in AREG gene-deficient mice; despite the fact that the overall expansion of FoxP3pos CD4 T cells under these inflammatory conditions was the same in both mouse strains.

Since mast cells are a major source of AREG under inflammatory conditions and have been shown to be "essential intermediaries in regulatory T cell tolerance", we transferred bone marrow derived mast cells (BMMC) into irradiated wt hosts together with bone marrow (BM) derived from AREG gene- deficient mice. Since those mice that receive wt BMMC did not develop dermatitis, while those that received AREG-/- BMMC did, we conclude that mast cell-derived AREG enables regulatory T cells in inflamed tissues to suppress local immune inflammation.

Potential applications

Application of the described technology is exemplified, but not limited to:

(i) AREG/EGF-R pathway as Treg target to modulate immune responses

(ii) The use of this pathway to modulate Inflammation in

autoimmune diseases

(iii) The use of this pathway to stimulate anti-tumor immune response

(iv) Use of anti-EGF-R (e.g. mAbs) or anti-AREG agents to modulate Treg controlled immune response.

Thus the invention further provides a use of an AREG/EGF-R pathway as Treg target to modulate an immune response in an individual. Further provided is the use of an AREG/EGF-R pathway to stimulate an antitumor response in an individual. Further provided is the use of an anti-EGF-R (e.g. mAbs) or anti-AREG agents to modulate a Treg controlled immune response. NOVELTY

We have for the first time demonstrated the expression of the EGF- receptor on CD4 T cells, including FoxP3 expressing regulatory T cells (Tregs). Activation of the EGF-R on Tregs by the EGF-like growth factor Amphiregulin enhances their suppressive capacity and is essential for their efficient functioning to suppress local inflammation in vivo.

This finding can open entirely novel approaches to regulate (enhance or suppress) immune responses. FoxP3pos CD4 T cells, so called regulatory T cells (Tregs), are well known to play an important role in immune-regulation preventing autoimmune diseases. On the other hand, however, they are also involved in the suppression of an efficient anti-cancer immuno- surveillance by contributing to a local immuno-protective environment around the tumor.

We recently reported the unexpected discovery that the immune system has adopted the EGF-R signalling pathway and uses the EGF-like growth factor, Amphiregulin (AREG), expressed by activated Th2 and mast cells, as a type II cytokine.

Most interestingly, mice gene- deficient for AREG display several characteristics indicative for malfunctioning of an immuno-regulatory pathway. We therefore analyzed the FoxP3pos CD4 T cell compartment in more detail and found that a subpopulation of CD4 T cells (in both human and mouse) expresses the EGF-R. This population expanded in infected wt but not AREG gene- deficient mice. Importantly, EGF-R expression could be restored by reconstitution of AREG gene- deficient mice with wt mast cells, which correlated with restored immune-regulatory capacities. In addition, treatment of Tregs with recombinant AREG significantly increased their suppressive capacity in standard in vitro suppression assays.

To test the relevance of this novel, AREG-mediated immuno-regulatory mechanism for tumor immunology, we tested a well-established vaccination strategy against B16 melanoma in AREG gene-deficient mice. While vaccination of wt mice with peptide loaded dendritic cells, 5 days after tumor injection, failed to protect the mice from tumor growth, AREG gene- deficient mice were protected by this treatment. Moreover, in a colitis model, where colitis is prevented by co-transfer of regulatory CD4 T cells, we find that co- transferred regulatory T cells are functional in wt RAG-1 gene- deficient mice but not in AREG x RAG-1 double gene- deficient mice. Importantly, we identified mast cells as the main source of AREG within inflamed tissues. So was, both in the colitis model and in a skin rejection model, inflammation prevented by reconstitution of the mice with wt - but not AREG-deficient bone marrow- derived mast cells.

Taken together, our findings imply a central role for the EGF-R signalling transduction pathway in the regulation of local inflammation and its potential as a target to modulate immune responses, in particular with respect to oncology.

Potential applications

Application of the described technology is exemplified, but not limited to:

(i) AREG/EGF-R pathway as Treg target to modulate immune responses

(ii) The use of this pathway to modulate Inflammation in

autoimmune diseases

(iii) The use of this pathway to stimulate anti-tumor immune response (iv) Use of anti-EGF-R (e.g. mAbs) or anti-AREG agents to modulate

Treg controlled immune response as an adjuvans in anti-tumor therapies, enhancing anti-tumor immunity. Thus further provided is the use of anti-EGF-R (e.g. mAbs) or anti-AREG agents to modulate Treg controlled immune response as an adjuvans in anti-tumor therapies, enhancing anti-tumor immunity.

NOVELTY

We have for the first time demonstrated the expression of the EGF- receptor on CD4 T cells, including FoxP3 expressing regulatory T cells (Tregs). Activation of the EGF-R on Tregs by the EGF-like growth factor Amphiregulin in in vitro suppression assays significantly enhances their suppressive capacity and is essential for their efficient functioning to suppress local inflammation in vivo.

This finding can explain a variety of clinical successes/effects of compounds that target the EGF-R and suggests novel intervention

mechanisms allowing modulating T cell immunity in general and more specifically in cancer therapy.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1: The EGF-R is expressed on mouse and human regulatory T- cells.

A&B) Regulatory T-cells were isolated based on the presence of CD25, mixed with splenocytes and cultured in the presence of 0.25 pg/ml activating anti- CD3 antibodies. A) Cells were stained for CD45.1, CD4, EGF-R and FoxP3 and analyzed by flow- cytometry. The orange line shows 0 hrs, the red line 19 hrs of activation and the blue line the isotype control for anti-EGF-R serum. B) % EGF-RP°^s regulatory T-cells over time of culture. C) Lymphocytes isolated from the MLN of uninfected or 28 days with Heligmosomoides bakeri infected mice or D) from the ectopic lymphoid structures surrounding B16 melanoma

(Shields JD, et al. Science. 2010 May 7;328(5979):749-52) were stained for CD4, the EGF-R and intracellularly for FoxP3. [C) n = 9 and 4 mice per group, P=0.0028; D) n = 10 and 17 mice per group, P=0.0003; both two-tailed Mann- Whitney test, bars are means + SEM]

E-G) PBMC of healthy controls (HC), juvenial arthritis (JIA) patients or mononucleated cells derived from the synovial fluid of JIA patients were stained for CD4, the EGF-R, Helios & FoxP3. E) The orange and red lines show EGF-R staining on PBMC from a healthy individual and a JIA patient, respectively. The blue line represents the isotype control for anti-EGF-R serum. [F) n=4 and 8 individuals per group, P=0.0286 two-tailed Mann- Whitney test] Figure 2: Amphiregulin enhances regulatory T-cells function in vitro

A) PBMC derived from healthy donors were stimulated with membrane-bound anti-CD3 antibodies and CD25P°^S/CD 127^neg CD4P°^s regulatory T-cells were added to suppress proliferation of CFSE-labeled CD4 T cells, in the presence or absence of 100 ng/ml recombinant AREG (Experiment performed in triplicates; P=0,0049 two-tailed Mann-Whitney test; bars are means +SEM). B) Regultory T-cells, i.e. in this case FoxP3/GFP CD4 T cells, were cultured for four days together with CFSE labelled CD45.1 expressing splenocytes at a ratio of 1:4 in the presence or absence of 100 ng/ml rec. AREG. T cells were activated with different amounts of soluble anti-CD3 and CFSE dilution within the CD45.1 expressing CD8 T cell population was analyzed by flow- cytometry. Proliferation was defined as the percentage of cells that had undergone at least one division. Triangles, inhibition in the presence of AREG; squares, inhibition in the absence of AREG. (Experiments performed in triplicates; P=0.0286 two-tailed Mann-Whitney test; bars are means +/-SEM)

Figure 3: Mast cell-derived Amphiregulin significantly enhances the capacity of regulatory T-cells cells to suppress the development of colitis in vivo.

RAG1-/- (light bars) or AREG-/- x RAGl-/- (dark bars) mice received 400,000 FACS-sorted naive CD4 T cells together with 150,000 CD25P°^S CD4P°^S regulatory T cells (A) or with increasing numbers of FOXP/GFPP°^s CD4 P°^S regulatory T cells (B). Development of colitis was measured six weeks later by histological score.⁹ [A) n = 6 per group, P=0.0020 B) n = 4 per group, P=0.028; both two-tailed Mann- Whitney test]. Bars are means +SEM.

C) 200,000 CD25P°^s CD4 P°^s regulatory T cells derived from wt or CD4-cre x EGF-R-flox/flox mice were co-transferred 400,000 naive CD4 T cells into RAGl-/- mice (n = 3 per group, P=0.05, one-tailed Mann-Whitney test). Bars are means +SEM.

D) RAG-1-/- x c-kit^{w sh} mice were reconstituted with BM-MC derived from either wt or AREG-/- mice. 200 000 FoxP3/GFP P°^s CD4 P°^s regulatory T cells were co-transferred with 400 000 naive CD4 and colitis scores were

determined six weeks after transfer (n = 5 per group, P=0.028, two-tailed Mann-Whitney test). Bars are means +SEM. Figure 4: Amphiregulin gene-deficient mice reject B16-F10 melanoma directly following post-transplantational immunization.

A&B) 7 000 B16-F10 cells were injected s.c. into AREG-/- mice (A&B) or c-kit^w- ^sh mice reconstituted with wt or AREG-/- BM-MC (C). On day 5 and 7 after tumor cell transfer, mice were immunized with BM-DC loaded with an immunogenic B16 melanoma epitope (TRP2i80-i8s) (A-C) or left un-immunized (no treatment A & B). As control, one mouse group received DC/pep and cyclophosphamide (A&B). Tumor size was determined over time (A) (squares represent non-immunized, triangles immunized and circles immunized and cyclophosphamide-treated animals) or 23 days after tumor transfer (B & C) [n = 5-7 mice per group, P = 0.016 (B) or n=3-7 mice per group, P = 0.032 (C); both two-tailed Mann- Whitney test]. Bars are means +SEM.

Figure 5: BM chimeric wt BL6-SJL mice reconstituted with AREG-/- BM develop an idiopathic form of dermatitis.

BL6-SJL mice (CD45.1) or AREG-/-were irradiated with lOGy X-rays and reconstituted with 1 x 10⁷ BM cells derived from BL6-SJL mice (CD45.1) or AREG-/- mice (B) (n= 4 mice per group) or c-kit^{w sh} mice (C) (n = 3

mice per group.) In D) BL6-SJL mice (CD45.1) were reconstituted with AREG- /- BM and 5 x 10⁶ BM-MC derived from either wt or AREG-/- mice were co- transferred, (n = 5 mice per group).

Figure 6: Similar frequencies of regulatory T-cells in AREG-/- and wt mice

MLN from wt C57BL/6 and AREG-/- mice were harvested, red blood cells depleted and stained for CD4 (GK1.5) and FoxP3 (FJK-16s). Frequencies of FoxP3+CD4 T cells were determined by flow- cytometry (n= 4 and 5 mice per group). Bars are means +SEM. Figure 7: Preferential expression of EGF-R by regulatory T-cells.

Effector T cells and regulatory T-cells from the blood of healthy individuals were FACS sorted based on non-overlapping expression markers, i.e. CD127^h¾^h versus CD127^low and CD25^h¾^h versus CD25^{very h}¾^h, and EGF-R expression was determined by quantitative RT-PCR.

Figure 8: AREG enhances regulatory T-cells function without influencing the proliferation of effector cells or regulatory T-cells survival.

A) Regulatory T-cells and CFSE labeled CD45.1+ splenocytes were mixed at a ratio of 1:4 and co-cultured in the presence of 0.03 pg/ml activating anti-CD3 antibody for four days. At t:o and t:48 hrs medium only (red line) or medium containing 100 ng/ml recombinant AREG (blue line) was added to the wells. Picture depicts CFSE dilution of CD8 T cells.

B) CFSE labeled splenocytes were cultured in the presence of 0.03 pg/ml activating anti-CD3 antibody for four days in the presence or absence of 100 ng/ml recombinant AREG. Percentage of CD4 T cells that had devided for at least one time was determined by flowcytometry (n = 6 per group). Bars are means +SEM.

C) CD45.2+ Regulatory T-cells and CFSE labeled CD45.1+ splenocytes were mixed and co-cultured in the presence of activating anti CD3 antibodies for four days. The number of Regulatory T-cells per well was determined by flowcytometry. (n = 18 per group) Bars are mean +SEM.

D) CFSE labeled CD45.1+ splenocytes were cultured in the presence of 0.3 pg/ml activating antibody (left two columns) or mixed at a ratio of 1:4 with regulatory T-cells (right two columns) and cultured for four days. At t:0 and t:48 hrs medium only or medium containing 10 ng/ml recombinant TGFa was added to the wells. Bars are means +SEM. Figure 9: Transferred wt and EGF-R-/- CD25P°^s T cell populations contained similar frequencies of FOXP3P°^s CD4T cells

Splenocytes derived from CD4-cre x EGF-R flox/flox mice (red) and of wt (blue) were stained for CD4 (GK1.5), CD25 (PC61) and FoxP3 (FJK- 16s) and analyzed by flow- cytometry.

Figure 10:

A) B16 tumor cells lack EGF-R expression.

mRNA from B16-F10 tumor cells and lung tissue was purifed and quantitative RT-PCR was performed. Bar is mean of ratio GAPDH : EGF-R expression

+SEM. ND = non-detecable (no EGF-R signal). Gene-specific probes used were purchased from Applied Biosystems, (GAPDH: Mm99999915_gl and EGF-R: Mm00433023_ml)

B) wt mice can reject transplanted B16 tumors only, if tumor antigen- specific immunization is accompanied by cyclophosphamide

treatment.

10,000 B16-F10 were injected s.c. into wt C57BL/6 mice. On day 5 and 7 after tumor cell transfer, mice were immunized with BM-DC loaded with an immunogenic B16 melanoma epitope (TRP2 iso-iss) or left un-immunized. One mouse group was immunized and cyclophosphamide-treated. Tumor size was determined on day 21 after tumor transfer, (n = 7 mice per group; P = 0.0152 two-tailed Mann- Whitney). Bars are means +SEM

C) B16 tumor cells grow similarly well in wt and AREG-/- mice.

10,000 B16-F10 were injected s.c. into wt C57BL/6 or AREG-/- mice.

Tumor size was determined on day 21 after tumor transfer, (n = 5 mice per group). Bars are means +SEM

D) TRP2 iso-iss immunization is similarly efficient in wt and in AREG-/- mice.

10,000 B16-F10 were injected s.c. into wt C57BL/6 mice. On day 5 and 7 after tumor cell transfer, mice were immunized with BM-DC loaded with an immunogenic B16 melanoma epitope (TRP2 iso-iss). On day 21 after tumor transfer, mice were sacrificed and splenocytes were activated with TRP2 iso-iss and intracellular IFNy expression in CD8 T cells was determined by flow cytometry, (n = 9 mice per group) Bars are means +SEM.

Figure 11:

A) Mouse mast cells express AREG upon activation.

BM-MC were activated by FceR cross linking and AREG expression was measured by real time-PCR. cDNA levels were equalized to GAPDH

expression. Experiment was performed in triplicates. Bars are means +SEM. Gene-specific probes used were purchased from Applied Biosystems, (GAPDH: Mm99999915_gl and AREG: Mm00437583_ml)

B) In c-kit^{w sh} mice reconstituted with different BM-MC grow B16 tumors with similar characteristics.

7,000 B16-F10 cells were injected s.c. into c-kit^{w sh} mice reconstituted with wt or AREG-/- BM-MC.23 days after tumor transfer mice were killed and tumor size determined [n = 5 mice per group]. Bars are means +SEM

C) c-kit^{w sh} x RAG1-/- mice reconstituted with different BM-MC develop colitis with similar kinetics and severity.

RAG-1-/- x c-kit^{w sh} mice were reconstituted with BM-MC derived from either wt or AREG-/- mice. 400,000 naive CD4 T cells were transferred and colitis scores were determined six weeks after transfer (n = 3 mice per group). Bars are means +/-SEM.

D) All BM-MC reconstituted c-kit^w_sh mice contain similar numbers of mast cells surrounding the blood vessels.

c-kit^{w sh} mice were reconstituted with either wt or AREG-/- BM-MC. Three weeks after re-constitution 10,000 B16-F10 cells were injected. 23 days after tumor cell transfer, tumors were harvested and stained for mast cells using Csaba staining. Figure 12: The EGF-R specific kinase inhibitor Gefitinib enhances the efficiency of posttransplantational anti-melanoma immunization, and the success of chemotherapy.

B16-F10 cells were transferred s.c. into C57BL/6 wt mice and

A) 5 days after tumor transfer mice were then immunized with DC loaded with the immunogenic B16 melanoma epitope (TRP2 iso-iss) followed by i.p. injection of Gefitinib (10 mg /kg bodyweight) every other day. 23 days after tumor transfer mice were sacrificed and tumor size determined.

B) 17 days after tumor transfer mice were i.p. injected with the

chemotherapeuticum Oxaliplatin (5 mg /kg bodyweight) followed by i.p.

injection of Gefitinib (10 mg / kg bodyweight) every other day. 11 days after chemotherapy mice were sacrificed and tumor size determined. As control one group, in addition to chemotherapy and Gefitinib treatment also i.p injected with 200 pg/mouse CD8 T-cell depleting monoclonal antibodies (clone YTS- 169) two days after chemotherapy.

Figure 13: Screening for murine EGF-R blocking nanobodies.

huEGF-R- specific monomeric single domain antibodies, so called nanobodies, derived from a Lama immunoglobin library were tested in competition assays with 125-I-labelled EGF for binding to the mu EGF-R expressed on NIH-3T3 cells.

Figure 14: TRP2 vaccination and nanobodies

B16-F10 cells were transferred s.c. into C57BL/6 wt mice and 5 days after tumor transfer mice were then immunized with DC loaded with the

immunogenic B16 melanoma epitope (TRP2 iso-iss). The following day mice received an i.p. injection of purified in yeast produced anti-muEGF-R nanobody RR-359, coupled to a nanobody specific for murine albumin.

Assuming a half- life of 48 hrs for this bi-polar nanobody construct, the level of nanobody per mouse was kept above a concentration of 100 pg per mouse for one week after immunization. 21 days after tumor transfer mice were sacrificed and tumor size determined, (n = 5-7 mice per group; P = 0.0140 two- tailed Mann- Whitney). Bars are means +SEM. Figure 15: The EGF-like growth factor Amphiregulin (AREG) is a type II cytokine derived from Th2 and mast cell.

Bone marrow derived mast cells (left) or differentiated T cell lines (right) were activated either by Fcepsiion-R cross linking (BMMC) or through stimulation through the TCR (anti-CD3 antibody) and AREG expression was measured by real time-PCR or gene chip hybridization.

Figure 16: About five percent of FoxP3pos CD4 T cells express the EGF-R. This population increases under inflammatory conditions in wt but not in AREG gene-deficient mice.

Lymphocytes derived from the MLN of uninfected wt or AREG-/- B6 mice (left panel), uninfected or 10 d. with Trichuris muris infected wt mice (middle panel) or infected wt or AREG-/- mice (right panel) were stained for CD4, the EGF-R and intracellularly for FoxP3. Figure 17: AREG enhances the suppressive capacity of regulatory T cells in vitro.

PBMC derived from a healthy donor were stimulated with membrane-bound anti-CD3 antibodies and CD25P°^s/CD127^neg CD4 T cells were added at a ratio of 1:2 to suppress proliferation of CFSE-labeled CD4 (left panel) or CD8 (right panel) T cells, in the presence (bottom row) or absence (middle row) of recombinant AREG. Figure 18: AREG enhances the suppressive capacity of regulatory T cells in vivo.

7xl0³ B16-F10 cells were injected s.c. into C57BL/6 wt controls and AREG gene- deficient mice, and mice were vaccinated 5 days after tumor transfer with DC loaded with the immunodominant B16-F10-derived epitope (TRP2i8o-i8s). Tumor size was determined 21 days after tumor transfer.

Figure 19: AREG is essential and sufficient to ensure efficient regulatory T cell function.

CD45RB^low and CD45RB^h¾^h CD 4 T cells were purified from the spleens of

AREG-/- mice and lxlO⁵ cells of each population (right panel) or of CD45RB^h¾^h CD4 T cells only (middle panel) were transferred, per mouse, into RAG-/- mice on a wt B6 or an AREG-/- B6 background. Six weeks after transfer mice, were sacrificed and proximal colons were analyzed by histology.

Figure 20: Mast cell-derived AREG is essential and sufficient to ensure efficient regulatory T cell function.

wt BL6 mice were irradiated with 10 Gy and reconstituted with BM derived from AREG-/- mice. lxlO⁶ in vitro differentiated BM-MC derived from AREG-/- BM (left panel) or wt BM (right panel) were co-transferred and mice were sacrificed 6 or 10 weeks after BM reconstitution and EGF-R expression on FOXP3P°^s CD4 T cells in the draining inguinal lymph nodes was determined. (Of note: Since those mice that had received AREG-/- BM-MC had developed a severe form of dermatitis, they all had to be sacrificed already 6 weeks after transfer). Methods:

Cell isolation, staining and flow cytometry

Mice were sacrificed by cervical dislocation, spleens were excised, leukocytes were obtained by pressing through a 70-μΜ cell strainer (BD Biosciences), and red blood cells were removed by ammonium chloride lysis. Staining of surface markers with the indicated antibodies was performed in the presence of Fc block (2.4G2) for at least 20 minutes on ice. For intracellular staining, cells were fixed with 2% paraformaldehyde for 20 minutes at room temperature, and intracellular staining was performed in the presence of 0.5% saponine for 1 hour at 4 °C. Antibodies were purchased from eBioscience and BD- BioSciences [CD4 (GK1.5), CD8 (53-6.7), CD45.2 (104), CD45.1 (A20),

IFNgamma (XMG1.2), EGF-R (13)]. Samples were measured on a

FACSCantoII (BD Biosciences) and analyzed with FlowJo software (Tree Star).

In vitro suppression assay

Human

CD4 positive cells were isolated from human PBMC and

CD4+CD25^h¾^hCD127^low T-cells were sorted and co-cultured with PBMC labeled with 2μΜ CFSE in anti-CD3 (clone OKT3) coated 96-wells. Cells were cultured for four days in RPMI medium supplemented with 10% FCS, 100 units/ml penicillin, 100 μg/ml streptomycin, and 2-mercaptoethanol, in the presence or absence of 100 ng/ml recombinant Amphiregulin (R&D BioSciences).

Proliferation of CD4 and CD8 positive cells was determined by measuring CFSE dilution using the FACS CANTO (BD Biosciences). Proliferation was defined as the percentage of cells that have undergone at least one division.

Mouse

FACS sorted CD45.2 expressing Treg cells (isolated either based on CD4 & CD25 expression or by GFP expression, if Treg cells were derived from Foxp3- GFP mice, Jackson, strain #006772) were added to CFSE labeled CD45.1 expressing splenocytes at different ratios. Cells were cultured in IMDM supplemented with 10% FCS, 2 mM L-glutamine, 100 units/ml penicillin, 100 pg/ml streptomycin, and 2-mercaptoethanol for four days, in the presence or absence of 100 ng/ml recombinant Amphiregulin. T-cells were activated with different amounts of soluble anti-CD3 (145-2C11; BD Pharmingen). To determine suppression of proliferation, CFSE dilution within the CD45.1 expressing T cell populations was analyzed by FACS. Proliferation was defined as the percentage of cells that have undergone at least one division.

Induced colitis mouse model

Ragl-/- mice were injected with 4xl0⁵ CD4+CD45RB^h¾^h cells to induce colitis.⁸ If not stated differently in the figure legend, 2xl0⁵ Tregs isolated from Foxp3- GFP mice (Jackson, strain #006772) were co-transferred and six weeks later the mice were sacrificed and colons scored by two independent experts in a blinded fashion according to Berg et al. ⁹, in brief:

Grade 0: no infiltration of mononuclear cells.

Grade 1: few foci of mononuclear cells, only slight depletion of goblet cells. Grade 2: many foci of mononuclear cells, infiltration in the lamina propria, however, not yet in the submucosa; diminished numbers of goblet cells.

Grade 3: strong infiltration, also in the submucosa; epithelial hyperplasia; number of goblet cells strongly diminished.

Grade 4: transmural infiltration of mononucleated cells; strong epithelial hyperplasia, goblet cell depletion.

Overall histological score per mouse is the sum of individual scores for the different segments of the colon (c. ascendens, c. transversum, c. descendens).

B16 melanoma vaccination

10 000 B16-F10 melanoma cells were transferred subcutaneously into the left flank of mice. Mice were then either immunized with TRP2i8o i88 peptide-loaded BM-DC on day 5 and 7 after transfer or left untreated. A control group received once on day 5 after transfer, an i.p. injection of 50 pg/kg body weight cyclophosphamide dissolved in 100 μΐ PBS to inactivate the Tregs. Differentiation and reconstitution of bone marrow derived mast cells

Bone marrow-derived mast cells (BM-MCs) were obtained from bone marrow that was cultured for 3 wks in the presence of pokeweed mitogen- stimulated spleen cell-conditioned medium as a source for IL-3. Non-adherent cells were passed once a week into new medium and purity of BM-MC population was determined by flow cytometry. Cultures contained a uniform cell population to over 94% positive for c-Kit and FceRIa. Mast cell- deficient c- kit^w'sh mice were injected i.v. with 5 x 10⁶ cultured BM-MC cells three weeks prior to start of the experiments. Mast cell reconstitution of the inflamed area was determined by Csaba staining and toluidine blue.

Examples

Example 1 Stimulation of the intrinsic tyrosine kinase activity of the epidermal growth factor receptor (EGF-R) induces a complex cascade of phosphorylation and activation events that determine a number of cell-fate decisions, such as proliferation or differentiation.¹ _' ² Amphiregulin (AREG) is an EGF-like growth factor that binds the EGF-R. Although originally described as a growth factor that is produced by epithelial cells, more recent studies have shown that also leuko- and lymphocyte populations, including activated Th2 cells,³ mast cells,⁴ _' ⁵ eosinophils and basophils, can produce AREG. In a previous study using AREG- deficient mice ³, we demonstrated that AREG, produced by Th2 cells, is involved in the efficient clearance of the parasite Trichuris muris. Infection by Trichuris muris is an acute infection that is cleared with a few weeks and the role of regulatory T-cells is supposed to be minor, at best.

Remarkably, however, while establishing BM chimeric mice, we found that wt recipient C57BL/6 mice that received bone marrow derived from AREG gene- deficient mice developed dermatitis (Fig 5). Since FoxP3 positive regulatory T cells are of crucial importance for immune tolerization after BM

transplantation, and since Treg- deficient mice have been shown to develop dermatitis, we postulated that Treg function could be impaired in the absence of Amphiregulin. Flow-cytometry analyses showed similar frequencies of FoxP3 expressing Tregs in the secondary lymphoid organs of AREG gene- deficient and wt mice (Fig 6), indicating that insufficient numbers of Tregs could not explain the observed lack of immune regulation in AREG-deficient mice. In the present invention, howver, we show that FoxP3 positive regulatory T cells are an important target through which AREG exerts its effect. Regulatory T-cells express the EGF-R:

To determine whether AREG might have a direct effect on Tregs, we measured EGF-R expression on Tregs by FACS analysis. While hardly any Tregs isolated from the spleen of mice expressed the EGF-R, up to 75% of Tregs gained EGF-R expression upon four days of in vitro culture in the presence of activating anti-CD3 antibodies (Fig 1A&B), indicating that Tregs gain EGF-R expression upon activation. To test whether Tregs express the EGF-R under inflammatory conditions in vivo, we analyzed EGF-R expression on Tregs derived from MLN of mice with a chronic worm infection. While in un-infected mice only about 5% of Tregs expressed the EGF-R, this population increased to more than 50% in worm-infected mice (Fig 1C). Similar observations were made in peri-tumoral regions of B16 melanoma tumors, where the frequency of EGF-R expressing Tregs was significantly higher than in the draining lymph nodes (Fig ID). Also in humans, EGF-R expressing Tregs were almost entirely absent in the blood of healthy individuals. Their frequencies, however, were significantly increased in the blood of arthritis patients and in particular at the site of inflammation, i.e. the synovial fluid (Fig 1E-G). To verify preferential EGF-R expression on Tregs, we sorted Tregs and activated CD4 effector T cells derived from healthy donors based on high expression of CD25 and the presence or absence of CD 127. Based on subsequent quantitative reverse transcription linked polymerase chain reaction (Q-PCR) on both cell populations, Tregs express at least 5 fold more EGF-R mRNA than effector T cells in healthy individuals (Fig. 7). Amphiregulin enhances regulatory T-cell functioning in vitro:

To address the functional relevance of EGF-R expression on Tregs, we first determined the effect of AREG on Tregs in in vitro suppression assays. As shown in Figure 2, the presence of AREG during the assay significantly enhanced the suppressive capacity of Tregs. Importantly, AREG had no influence on the overall proliferation or survival of the Tregs and did not influence the proliferation of effector cells (Fig 8), indicating that AREG directly enhances the suppressive capacity of Tregs. AREG induced effects were most prominent under conditions of only partial suppression and increased upon dilution of the stimulating anti-CD3 antibody concentration (Fig 2 B). Most interestingly, the presence of other EGF-R ligands, such as TGFalpha, that in contrast to AREG binds the EGF-R with high- affinity, did not enhance Treg function but in contrast actually diminished it (Figure 8 D).

Amphiregulin enhances regulatory T-cell functioning in vivo:

Next we determined the physiological relevance of AREG expression on

Treg function in vivo. To this end, we transferred na^'ive CD4 T cells in the presence or absence of Treg-cells into lymphopenic RAG 1- deficient or AREG x RAG 1- deficient mice.⁸ Colitis development as measured by histological score was determined six weeks after transfer.⁹ As shown in Figure 3A, co- transferred Tregs efficiently prevented the development of colitis in RAGl- deficient mice, while substantial inflammation could still be detected in AREG x RAG 1- deficient mice (Figure 3A). To explain the remaining inflammation, we transferred a fixed number of na^'ive CD4 T cells together with increasing amounts of Tregs into either RAG 1- deficient or AREG x RAG 1- deficient mice. Co-transferred Tregs decreased the severity of disease in a dose-dependent manner in both RAG 1- deficient and AREG x RAG 1- deficient mice, however, Tregs were significantly less suppressive in AREG x RAG 1- deficient mice (Fig 3B), demonstrating that AREG enhances the suppressive capacity of Tregs in vivo. To verify that AREG directly effects EGF-R expressing Tregs, we crossed EGF-R flox/flox mice onto a CD4-cre background and transferred sorted Tregs derived from these mice into RAGl gene- deficient mice. Both CD25+ population contained similar frequencies of FoxP3+ cells (Fig 9), still, as shown in Figure 3C, EGF-R gene- deficient Tregs were significantly less capable of suppressing the development of colitis than Tregs derived from wt mice. Post-transplantational anti-melanoma immunization is significantly more efficient in Amphiregulin gene-deficient than in C57BL/6 ivt mice:

It has been shown that Tregs constitute a mechanism by which tumors protect themselves against tumor- specific immune responses.¹⁰ We therefore tested the importance of AREG/EGF-R interaction in Treg functioning in a tumor model. For the B16 melanoma model it is well established that antitumor immunization can reject transplanted B16-F10 tumors in w t C57BL/6 mice, if Treg-cells are depleted prior to tumor transfer ¹¹ or when Treg are functionally inactivated, by injection of low-doses of cyclophosphamide.¹² _' ¹³ We therefore transferred B16-F10 melanoma cells that lack EGF-R expression (Fig 10) into wt C57BL/6 or AREG-deficient mice and immunized these mice with TRP2180- 188 tumor epitope-pulsed BM-DC, 5 and 7 days after tumor

transplantation. As shown in Figure 4A&B, AREG-deficient mice, unlike wt mice (Fig 10), efficiently reject transplanted B16-F10 tumors already upon immunization only. As BM-DC immunization induced similar anti TRP2iso-i88- specifc immune responses in wt as in AREG-deficient mice (Fig 10) and the B16 tumor grew with similar characteristics in wt C57BL/6 as in AREG- deficient mice (Fig 10), these data suggest that a diminished tumor

suppressive environment may explain for successful tumor rejection in DC/pep immunized AREG-deficient mice.

Mast cell-derived Amphiregulin enhances regulatory T-cell function in vivo:

Next we wanted to determine the physiologically relevant source of AREG that enhances Treg function in vivo. As Tregs and mast cells have been shown to cooperate ¹⁴ and activated mast cells express AREG ⁴>⁵ (Fig 11), we tested B16-F10 melanoma rejection upon immunization in mast cell deficient c-kit^{w sh} mice. Almost all tissues of this mouse strain are efficiently

reconstituted by in vitro differentiated BM mast cells (BM-MC) ¹⁵ and in tissues that are difficult to reconstitute at steady state, such as the lungs or brain, mast cells will infiltrate under inflammatory conditions.¹⁶ As shown in Figure 4C, c-kit^{w sh} mice that did not receive in vitro differentiated BM-MC or had been reconstituted with BM-MC derived from AREG gene- deficient mice efficiently controlled tumor growth following DC/pep immunization, while mice that had been reconstituted with BM-MC derived from wt C57BL/6 mice did not. Tumor growth was similar in all un-immunized c-kit^{w sh} mouse groups (Fig 11) and similar numbers of mast cells were found in the peri-tumoral regions (Fig 11) of c-kit^{w sh} mice reconstituted with BM-MC derived from either w t or AREG gene- deficient mice.

In the colitis model, we found that six weeks after co-transfer of naive

CD4 and Tregs, colitis scores in c-kit^{w sh} x RAG 1- deficient mice that had prior to transfer been reconstituted with BM-MC of wt mice were significantly lower than in mice that had not received BM-MC or had been reconstituted with BM- MC derived from AREG gene-deficient mice (Fig 3D). Transfer of na^'ive CD4 T cells only, led to similarly severe colitis in all c-kit^{w sh} x RAG 1- deficient mouse groups (Fig 11), indicating that neither mast cells nor mast cell derived AREG were necessary for the functioning or differentiation of colitogenic effector T cells. Taken together, these data indicate that mast cell-derived AREG is of crucial importance to enhance Treg function.

Our finding that mast cell- derived AREG is of crucial importance for the functioning of Tregs is in agreement with findings by others reporting crosstalk between mast cells and Tregs.¹⁴ Both Tregs and mast cells infiltrate sites of inflammation and, in the situation of developing tumors, accumulate in the peri-tumoral region,¹⁷ _' ¹⁸ Also we found that a particular high percentage of Tregs in the ectopic lymphoid structures at the periphery of tumors expressed the EGF-R (Figure ID). Since it has been shown that Tregs gain optimal suppressive capacity only by migrating through the site of inflammation,¹⁹ it is tempting to speculate that mast cell- derived AREG could explain for this observation. The novel EGF-R mediated link between mast cells and Tregs identified here may explain the clinical successes of EGF-R targeting medications, at least in part. Since a number of observations suggest that the clinical successes of EGF-R targeting treatments to a substantial extent is based on treatment-induced tumor- specific T cell responses,²⁰ it is easy to envision that an additional blockade of tumor-residential Tregs could indirectly enhance the efficacy of this treatment. Thus, in a wider setting, the addition of EGF-R targeting treatments could substantially enhance the efficacy of a number of therapeutic treatments that at the moment are hampered by Treg mediated immune modulation.

Example 2

Interference with the EGF-R signalling pathway enhances the

efficiency of melanoma post-transplantational immunization:

To test how our findings could be applied, we used our B16

immunization model in combination with Gefitinib, an EGF-R specific kinase inhibitor that is approved for clinical treatment of head and neck cancer. As shown in Figure 12A, Gefitinib treatment alone, similar to DC-peptide immunization alone, had no appreciable effect on the growth of the B16 melanoma. However, the combination of Gefitinib with the BM-DC vaccination diminished tumor growth significantly.

To test whether chemotherapy (CT) as such leads to antigen release and induction of CD8 T cell responses, capable of clearing the tumor when Treg cells are inactivated by blockage of their EGF-R signalling pathway, we combined Gefitinib treatment with CT. After B16F10 injection, we waited until clearly visible tumors had formed, then treated the mice once with 5mg/kg bodyweight Oxaliplatin and in the following days with Gefitinib, with a frequency of once in every two days. As shown in Figure 12B, mice that had received combination treatment showed a diminished tumor growth. Depletion of CD 8 T cells abrogated this effect, indicating that diminished tumor growth was dependent on CD8 T cells.

Although significant, the effects of combined CT and Gefitinib treatment on tumor growth were much more moderate than we had hoped for based on our results in AREG gene- deficient mice. Most likely, this is explained by the fact that Gefitinib was administered systemically every other day, while serum concentrations of this drug, with a serum half life of around 6 hrs, rapidly decline. In the clinic, Gefitinib is given daily in a tablet that slowly dissolves, which leads to a more constant high blood level in treated patients. To circumvent these complications, we used in the following experiments muEGF- R blocking antibodies that have a longer half- life in the circulation.

Establishment of murine EGF-R blocking nanobodies

Available EGF-R blocking antibodies target the human EGF-R and do not cross-react with the murine EGF-R. We therefore got in contact with the group of Paul van Bergen en Henegouwen of the Biology Department of the University of Utrecht, who had screened antibodies derived from Lama

Immunoglobulin libraries for their capacity to interact with the human EGF-R. Lamas express single domain antibodies, so called nanobodies. Like regular antibodies, these single monomeric variable antibody domains bind selectively to specific antigens, however, they are much smaller (only 15 kDa) and easy to express in large quantities in yeast or bacteria. Based on amino acid

alignments and 3D predictions of the human and murine EGF-R, a particular shared sequence homologous region of the EGF-R was chosen. Antibodies targeting this region block the activation of the EGF-R. Of those nanobodies that interact with this region of the human EGF-R, a number of different nanobodies were as shown in Figure 13 identified to also react with the murine EGF-R. In particular, the nanobody RR-359 was considered to be useful for our purposes and was used for the following experiments. EGF-R signalling pathway enhances the efficiency of melanoma post- transplantational immunization:

To test whether EGF-R blocking antibodies could have a superior effect on the efficacy of anti-tumor vaccinations, we used our B16

immunization model in combination with purified RR-359. To enhance the half- life of this murine EGF-R blocking nanobody in vivo, this nanobody was coupled to a nanobody that recognizes mouse albumin and this bi- specific antibody-construct was expressed in yeast. Nanobody-treatment was started one day after the immunization and for one week the concentration of the nanobody was kept above 100 microgram per mouse, assuming a half-life of 48 hrs of the bi-specific antibody construct in vivo. As shown in Figure 14, nanobody treatment alone, similar to DC-peptide immunization alone, had no appreciable effect on the growth of the B16 melanoma. However, the combination of murine EGF-R blocking nanobody with the BM-DC vaccination diminished tumor growth significantly.

Example 3

In the Lab of Tim Mosmann at the University of Rochester, we performed a differential display of genes expressed in different in vitro differentiated mouse T- helper cell cultures. Na^'ive CD4 T cells were

differentiated to Thl, Th2 and Thpp. Thpp are differentiated and form memory, but can be differentiated further towards Thl or Th2 cells upon reactivation under the appropriate conditions. As control groups, total spleen cells and isolated B-cells were used. One gene that showed a striking regulation upon stimulation through the TCR by CD3/CD28 antibodies was the EGF-like growth factor Amphiregulin (AREG). AREG was exclusively expressed by activated Th2 cells, while no other significant expression of any other EGF-R ligand was detected in this assay. Since we also could not detect significant AREG expression in the spleen of a na^'ive mouse, even after mitogenic activation, activated Th2 cells are the very only lymphocytes in the spleen that express AREG. Recently, also mast cells were described to express AREG upon activation ²> ³. In addition, it was shown that epithelial cells exposed to high levels of the pro-inflammatory cytokine TNFalpha express TGFalpha, another EGF-R ligand.

To elucidate the function of Th2-cell derived AREG, we analyzed AREG gene- deficient mice using the infection model, Trichuris muris. We specifically chose this infection model since infection induces strong Th2 responses, which are essential and sufficient for protection. Clearance of this pathogen is independent of B-cells, mast cells, macrophages or NKT cells. As shown in Zaiss et al. 2006, AREG gene-deficient mice had a significantly delayed clearance of the pathogen from the caecum. Efficient clearance of the pathogen was dependent on the expression of AREG in bone marrow (BM) derived cells, since BM chimeric mice could clear the pathogen efficiently only, if the irradiated host was reconstituted with BM cells capable of AREG expression. As the mechanism, we determined that Th2-derived AREG enhanced the proliferation of the epithelial cells in the caecum of infected mice. Remarkably, when establishing BM chimeric mouse, we found that irradiated w t hosts reconstituted with AREG gene-deficient BM consistently developed a severe form of dermatitis. Such tissue rejection is observed frequently in BM chimeric mice, mainly under conditions of local regulatory T cell dysfunction. Since also mice in which the entry of FOXP3P°^s CD4 T cells into the skin is disrupted develop dermatitis, we analyzed the FOXP3P°^s CD4 T cell compartment in AREG gene-deficient and w t mice in more detail. We found similar frequencies of FOXP3P°^s CD 4 T cells in the lymphoid organs of AREG gene- deficient as in w t mice; in addition, in both mouse strains we detected a small population FOXP3P°^s CD4 T cells (about 5% of FOXP3P°^s CD4 T cells) expressing the EGF-R. In particular the latter finding is unexpected; as it is assumed that BM- derived cells are the only cells that fail to express the EGF- R. Our finding has been confirmed by the group of R. Toes at the LUMC in Leiden, who find that also about 5% of FOXP3P°^s CD4 T cells in human blood express the EGF-R. When we analyzed EGF-R expression by FOXP3P°^s CD4 T cells under inflammatory conditions, we noticed that the EGF-RP°^S FOXP3P°^S CD4 T cell population had increased substantially (up to 50%) in wt mice, in contrast to in AREG gene-deficient mice where still only about 5% of the FOXP3P°^s CD4 T cells expressed the EGF-R. Total frequencies of FOXP3P°^s CD 4 T cells were the same in both mouse strains.

To determine the functional relevance of AREG signalling in regulatory T cells, we performed an in vitro T cell suppression assay in collaboration with the group of Paul Coffer from the UMC in Utrecht. As shown in Figure 17, the addition of AREG significantly (P=0.05, one-tailed Mann-Whitney) enhanced the suppressive capacity of added regulatory T cells.

To determine the functional role of EGF-R expression on FoxP3pos CD4 T cells in vivo, we tested the rejection of established B16 tumors by C57BL/6 wt and AREG gene- deficient mice, following vaccination. In this model, FoxP3pos CD4 T cells have to be depleted prior to vaccination to ensure efficient tumor rejection 12. Tumors grew more rapidly in AREG gene- deficient than in wt mice (Of note: AREG gene- deficient mice with a tumor size of more than 1.6 cm - 3 out of 4 mice - had to be sacrificed already at day 19). However, while AREG gene- deficient mice were well able to control tumor growth after vaccination with peptide-loaded dendritic cells (DC), wt mice failed to reject the tumor following vaccination unless treated with cyclophosphamide in order to suppress FoxP3pos CD4 T cells (Figure 18).

To directly show that AREG influences the suppressive capacity of regulatory T cells in vivo, we tested regulatory T cell function in the well- established "T cell transfer induced colitis" model in RAG-1 gene- deficient mice that had been backcrossed on an AREG gene- deficient background. In this IBD model, colitis rapidly develops following transfer of naive (CD45RB^h¾^h) CD4 T cells into RAG gene- deficient mice. Co-transfer of FOXP3P°^s regulatory T cells prevents the development of IBD. As shown in Figure 19 co-transfer of

CD45RB^low CD4 T cells could prevent colitis in wt RAG-1 gene- deficient mice but not in AREG x RAG-1 double gene- deficient mice (Figure 19).

Since mast cell- deficient mouse strains develop "idiopathic" dermatitis; mast cells collaborate with FOXP3P°^S CD4 T cells in the local regulation of immune responses; and are a major source of AREG, we tested whether mast cells could be the source of AREG that enables regulatory T cells to suppress local immune responses. We established BM chimeric wt mice that, after irradiation, received wt or AREG gene-deficient BM-derived mast cells (BM-MC) together with BM of AREG gene- deficient mice. Mice that received wt BM-MC were protected against the development of dermatitis while mice that received AREG gene- deficient BM-MC were not. Remarkably, disease outcome directly correlated with EGF-R expression on the FOXP3P°^S CD4 T cells population in the inguinal, draining lymph node (Figure 20).

Similar results were seen in the T cell transfer colitis model, in mast cell deficient c-kit^{w sh} mice backcrossed onto a RAG-1 gene-deficient background. When regulatory T cells were co-transferred with naive CD4 T cells, colitis could be prevented in mice that, prior to transfer, had been reconstituted with wt BM-MC but not in mice that had received AREG gene-deficient BM-MC.

Thus: Signalling through the EGF-R on FOXP3P°^s CD 4 T cells, by mast cell-derived AREG, enables FOXP3P°^S CD4 T cells to suppress local immune responses. REFERENCES

1. Schlessinger, J. Cell signaling by receptor tyrosine kinases. Cell 103, 211-25 (2000).

2. Avraham, R. & Yarden, Y. Feedback regulation of EGFR signalling: decision making by early and delayed loops. Nat Rev Mol Cell Biol 12, 104-17 (2011).

3. Zaiss, D. M. et al. Amphiregulin, a TH2 cytokine enhancing resistance to nematodes. Science 314, 1746 (2006).

4. Wang, S. W. et al. Amphiregulin expression in human mast cells and its effect on the primary human lung fibroblasts. J Allergy Clin Immunol 115, 287-94 (2005).

5. Okumura, S., Sagara, H., Fukuda, T., Saito, H. & Okayama, Y.

FcepsilonRI-mediated amphiregulin production by human mast cells increases mucin gene expression in epithelial cells. J Allergy Clin

Immunol 115, 272-9 (2005).

6. Cobbold, S. P. et al. Immune privilege induced by regulatory T cells in transplantation tolerance. Immunol Rev 213, 239-55 (2006).

7. Sather, B. D. et al. Altering the distribution of Foxp3(+) regulatory T cells results in tissue-specific inflammatory disease. J Exp Med 204,

1335-47 (2007).

8. Powrie, F. et al. Inhibition of Thl responses prevents inflammatory bowel disease in scid mice reconstituted with CD45RBhi CD4+ T cells. Immunity 1, 553-62 (1994).

9. Berg, D. J. et al. Enterocolitis and colon cancer in interleukin-10- deficient mice are associated with aberrant cytokine production and CD4(+) THl-like responses. J Clin Invest 98, 1010-20 (1996).

10. Zou, W. Regulatory T cells, tumour immunity and immunotherapy. Nat Rev Immunol 6, 295-307 (2006).

11. Sutmuller, R. P. et al. Synergism of cytotoxic T lymphocyte-associated antigen 4 blockade and depletion of CD25(+) regulatory T cells in antitumor therapy reveals alternative pathways for suppression of autoreactive cytotoxic T lymphocyte responses. J Exp Med 194, 823-32 (2001).

12. North, R. J. Cyclophosphamide-facilitated adoptive immunotherapy of an established tumor depends on elimination of tumor-induced suppressor T cells. J Exp Med 155, 1063-74 (1982).

13. Matsushita, N., Pilon-Thomas, S. A., Martin, L. M. & Riker, A. I.

Comparative methodologies of regulatory T cell depletion in a murine melanoma model. J Immunol Methods 333, 167-79 (2008).

14. Lu, L. F. et al. Mast cells are essential intermediaries in regulatory T cell tolerance. Nature 442, 997-1002 (2006).

15. Grimbaldeston, M. A. et al. Mast cell- deficient W-sash c-kit mutant Kit W-sh/W-sh mice as a model for investigating mast cell biology in vivo. Am J Pathol 167, 835-48 (2005).

16. Hallgren, J. & Gurish, M. F. Pathways of murine mast cell development and trafficking: tracking the roots and routes of the mast cell. Immunol Rev 217, 8-18 (2007).

17. Aydin, O., Dusmez, D., Cinel, L., Doruk, E. & Kanik, A.

Immunohistological analysis of mast cell numbers in the intratumoral and peritumoral regions of prostate carcinoma compared to benign prostatic hyperplasia. Pathol Res Pract 198, 267-71 (2002).

18. Ju, M. J. et al. Combination of peritumoral mast cells and T regulatory cells predicts prognosis of hepatocellular carcinoma. Cancer Sci 100, 1267-74 (2009).

19. Zhang, N. et al. Regulatory T cells sequentially migrate from inflamed tissues to draining lymph nodes to suppress the alloimmune response. Immunity 30, 458-69 (2009).

20. Ferris, R. L., Jaffee, E. M. & Ferrone, S. Tumor antigen-targeted,

monoclonal antibody-based immunotherapy: clinical response, cellular immunity, and immunoescape. J Clin Oncol 28, 4390-9 (2010). Adams, G.P., and L.M. Weiner. Monoclonal antibody therapy of Nat Biotechnol. 23:1147-11.. (2005).

Claims

1. A method for the treatment of an individual suffering from a chronic infection by a pathogen and/or a tumor, said method comprising administering an EGF-R inhibitor and/or a high affinity EGF-R signalling ligand to said individual, said method further comprising providing the individual in need thereof with antigen of said pathogen or tumor, immune cells comprising and/or specific for said antigen, an antibody specific for said antigen or a combination thereof.

2. A method for the treatment of an individual suffering from a chronic infection by a pathogen, said method comprising administering an EGF-R inhibitor and/or a high affinity EGF-R signalling ligand to said individual.

3. A method according to claim 1 or claim 2, wherein said EGF-R inhibitor and/or high affinity EGF-R signalling ligand interacts with an EGF-R on regulatory T cells thereby reducing immune suppression by said regulatory T cell.

4. A method according to claim 1 or claim 3, wherein said tumor is an EGF-R negative tumor.

5. A method according to any one of claims 1, 3 or 4, wherein said tumor is a melanoma, prostate cancer, pancreatic cancer or a lymphoma.

6. A method according to any one of claims 1, 3-5, wherein antigen of said pathogen or tumor is provided to the individual by administering a vaccine comprising said antigen to said individual.

7. A method according to any one of claims 1, 3-6, wherein antigen of said pathogen is provided to the individual by administering a medicament that is toxic for said pathogen to said individual.

8. A method according to any one of claims 1, 3-7, wherein said individual is provided with an antibody specific for a tumor antigen.

9. A method according to any one of claims 1-8, wherein said EGF-R inhibitor is an EGF-R specific antibody such as CetuxiMab, ABX-EGF, EMD72000, MAb ICR62 and hR3.

10. A method according to any one of claims 1-8, wherein said EGF-R inhibitor is a tyrosine kinase inhibitor such as gefitinib or erlotinib.

11. A method according to any one of claims 1-8, wherein said high affinity EGF-R signalling ligand is epidermal growth factor (EGF), transforming growth factor alpha (TGFalpha) or heparin-binding EGF-like growth factor (HB-EGF).

12. An EGF-R inhibitor or a high affinity EGF-R signalling ligand, for use in reducing immune suppression by an EGF-R positive regulatory T cell, preferably in an individual in need thereof.

13. An EGF-R inhibitor or a high affinity EGF-R signalling ligand for use according to claim 12, in the treatment of an individual suffering from a chronic infection and/or a tumor.

14. An EGF-R inhibitor or a high affinity EGF-R signalling ligand for use according to claim 12 or claim 13, wherein said treatment further comprises providing the individual in need thereof with antigen of said pathogen or tumor, immune cells comprising and/or specific for said antigen, an antibody specific for said antigen or a combination thereof..

15. An EGF-R inhibitor or a high affinity EGF-R signalling ligand for use in the treatment of an individual suffering from a chronic infection and/or a tumor, wherein said treatment further comprises providing the individual in need thereof with antigen of said pathogen or tumor, immune cells comprising and/or specific for said antigen, an antibody specific for said antigen or a combination thereof.

16. A method for modulating the function of a regulatory T cell, said method comprising providing an EGF-R positive regulatory T cell with an EGF-R inhibitor or an EGF-R signalling ligand thereby modulating the function of said regulatory T cell.

17. A method for the treatment of an individual suffering from a disease associated with an over-active immune response, said method comprising administering to the individual in need thereof a low affinity EGF- R signalling ligand.

18. A method according to claim 17, wherein said disease associated with an over-active immune response comprises an auto-immune disease such as rheumatoid arthritis, multiple sclerosis or type I diabetis.

19. A method according to claim 17, wherein said disease associated with an over-active immune response is an allergy or a

contacthypersensitivity.

20. A method according to claim 17, wherein disease associated with an over-active immune response is graft versus host disease or host versus graft disease.